Never say never...

754 files changed, 2436 insertions, 4872 deletions

diff --git a/SRC/cbbcsd.f b/SRC/cbbcsd.f
index 9db7b9ee..91254a91 100644
--- a/SRC/cbbcsd.f
+++ b/SRC/cbbcsd.f
@@ -282,8 +282,7 @@
 *> \verbatim
 *>          LRWORK is INTEGER
 *>          The dimension of the array RWORK. LRWORK >= MAX(1,8*Q).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LRWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal size of the RWORK array,
 *>          returns this value as the first entry of the work array, and
@@ -298,20 +297,16 @@
 *>          > 0:  if CBBCSD did not converge, INFO specifies the number
 *>                of nonzero entries in PHI, and B11D, B11E, etc.,
 *>                contain the partially reduced matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Reference
 *>  =========
-*> \endverbatim
-*> \verbatim
+*>
 *>  [1] Brian D. Sutton. Computing the complete CS decomposition. Numer.
 *>      Algorithms, 50(1):33-65, 2009.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  TOLMUL  REAL, default = MAX(10,MIN(100,EPS**(-1/8)))
 *>          TOLMUL controls the convergence criterion of the QR loop.
 *>          Angles THETA(i), PHI(i) are rounded to 0 or PI/2 when they
diff --git a/SRC/cbdsqr.f b/SRC/cbdsqr.f
index e8ea175d..c9423b93 100644
--- a/SRC/cbdsqr.f
+++ b/SRC/cbdsqr.f
@@ -180,12 +180,10 @@
 *>                elements of a bidiagonal matrix which is orthogonally
 *>                similar to the input matrix B;  if INFO = i, i
 *>                elements of E have not converged to zero.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  TOLMUL  REAL, default = max(10,min(100,EPS**(-1/8)))
 *>          TOLMUL controls the convergence criterion of the QR loop.
 *>          If it is positive, TOLMUL*EPS is the desired relative
@@ -200,8 +198,7 @@
 *>          Default is to lose at either one eighth or 2 of the
 *>             available decimal digits in each computed singular value
 *>             (whichever is smaller).
-*> \endverbatim
-*> \verbatim
+*>
 *>  MAXITR  INTEGER, default = 6
 *>          MAXITR controls the maximum number of passes of the
 *>          algorithm through its inner loop. The algorithms stops
diff --git a/SRC/cgbrfs.f b/SRC/cgbrfs.f
index 4fad09c9..1bff7c13 100644
--- a/SRC/cgbrfs.f
+++ b/SRC/cgbrfs.f
@@ -181,12 +181,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/cgbrfsx.f b/SRC/cgbrfsx.f
index fa6ee264..7448f1e4 100644
--- a/SRC/cgbrfsx.f
+++ b/SRC/cgbrfsx.f
@@ -256,37 +256,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -295,8 +289,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -307,14 +300,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -322,26 +313,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -352,8 +339,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -372,8 +358,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0
@@ -384,8 +369,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -395,8 +379,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/cgbsvx.f b/SRC/cgbsvx.f
index 41b0df4d..890112da 100644
--- a/SRC/cgbsvx.f
+++ b/SRC/cgbsvx.f
@@ -151,14 +151,12 @@
 *>          The j-th column of A is stored in the j-th column of the
 *>          array AB as follows:
 *>          AB(KU+1+i-j,j) = A(i,j) for max(1,j-KU)<=i<=min(N,j+kl)
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'F' and EQUED is not 'N', then A must have been
 *>          equilibrated by the scaling factors in R and/or C.  AB is not
 *>          modified if FACT = 'F' or 'N', or if FACT = 'E' and
 *>          EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if EQUED .ne. 'N', A is scaled as follows:
 *>          EQUED = 'R':  A := diag(R) * A
 *>          EQUED = 'C':  A := A * diag(C)
@@ -181,12 +179,10 @@
 *>          and the multipliers used during the factorization are stored
 *>          in rows KL+KU+2 to 2*KL+KU+1.  If EQUED .ne. 'N', then AFB is
 *>          the factored form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AFB is an output argument and on exit
 *>          returns details of the LU factorization of A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then AFB is an output argument and on exit
 *>          returns details of the LU factorization of the equilibrated
 *>          matrix A (see the description of AB for the form of the
@@ -206,13 +202,11 @@
 *>          contains the pivot indices from the factorization A = L*U
 *>          as computed by CGBTRF; row i of the matrix was interchanged
 *>          with row IPIV(i).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains the pivot indices from the factorization A = L*U
 *>          of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then IPIV is an output argument and on exit
 *>          contains the pivot indices from the factorization A = L*U
 *>          of the equilibrated matrix A.
diff --git a/SRC/cgbsvxx.f b/SRC/cgbsvxx.f
index 483fa9c3..3684e81b 100644
--- a/SRC/cgbsvxx.f
+++ b/SRC/cgbsvxx.f
@@ -180,14 +180,12 @@
 *>     The j-th column of A is stored in the j-th column of the
 *>     array AB as follows:
 *>     AB(KU+1+i-j,j) = A(i,j) for max(1,j-KU)<=i<=min(N,j+kl)
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'F' and EQUED is not 'N', then AB must have been
 *>     equilibrated by the scaling factors in R and/or C.  AB is not
 *>     modified if FACT = 'F' or 'N', or if FACT = 'E' and
 *>     EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>     On exit, if EQUED .ne. 'N', A is scaled as follows:
 *>     EQUED = 'R':  A := diag(R) * A
 *>     EQUED = 'C':  A := A * diag(C)
@@ -210,13 +208,11 @@
 *>     and the multipliers used during the factorization are stored
 *>     in rows KL+KU+2 to 2*KL+KU+1.  If EQUED .ne. 'N', then AFB is
 *>     the factored form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then AF is an output argument and on exit
 *>     returns the factors L and U from the factorization A = P*L*U
 *>     of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'E', then AF is an output argument and on exit
 *>     returns the factors L and U from the factorization A = P*L*U
 *>     of the equilibrated matrix A (see the description of A for
@@ -236,13 +232,11 @@
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     as computed by SGETRF; row i of the matrix was interchanged
 *>     with row IPIV(i).
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then IPIV is an output argument and on exit
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'E', then IPIV is an output argument and on exit
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     of the equilibrated matrix A.
@@ -382,37 +376,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -421,8 +409,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -433,14 +420,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -448,26 +433,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -478,8 +459,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -498,8 +478,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0
@@ -510,8 +489,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -521,8 +499,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/cgbtf2.f b/SRC/cgbtf2.f
index 0b9a514e..b01bad78 100644
--- a/SRC/cgbtf2.f
+++ b/SRC/cgbtf2.f
@@ -76,8 +76,7 @@
 *>          The j-th column of A is stored in the j-th column of the
 *>          array AB as follows:
 *>          AB(kl+ku+1+i-j,j) = A(i,j) for max(1,j-ku)<=i<=min(m,j+kl)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, details of the factorization: U is stored as an
 *>          upper triangular band matrix with KL+KU superdiagonals in
 *>          rows 1 to KL+KU+1, and the multipliers used during the
diff --git a/SRC/cgbtrf.f b/SRC/cgbtrf.f
index 7d319ebf..17a238b6 100644
--- a/SRC/cgbtrf.f
+++ b/SRC/cgbtrf.f
@@ -76,8 +76,7 @@
 *>          The j-th column of A is stored in the j-th column of the
 *>          array AB as follows:
 *>          AB(kl+ku+1+i-j,j) = A(i,j) for max(1,j-ku)<=i<=min(m,j+kl)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, details of the factorization: U is stored as an
 *>          upper triangular band matrix with KL+KU superdiagonals in
 *>          rows 1 to KL+KU+1, and the multipliers used during the
diff --git a/SRC/cgees.f b/SRC/cgees.f
index b0ee5e2a..0c5199c9 100644
--- a/SRC/cgees.f
+++ b/SRC/cgees.f
@@ -143,8 +143,7 @@
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.  LWORK >= max(1,2*N).
 *>          For good performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cgeesx.f b/SRC/cgeesx.f
index 37d41962..c4c24f32 100644
--- a/SRC/cgeesx.f
+++ b/SRC/cgeesx.f
@@ -184,8 +184,7 @@
 *>          that an error is only returned if LWORK < max(1,2*N), but if
 *>          SENSE = 'E' or 'V' or 'B' this may not be large enough.
 *>          For good performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates upper bound on the optimal size of the
 *>          array WORK, returns this value as the first entry of the WORK
diff --git a/SRC/cgeev.f b/SRC/cgeev.f
index 51bd22dc..bad77c57 100644
--- a/SRC/cgeev.f
+++ b/SRC/cgeev.f
@@ -139,8 +139,7 @@
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.  LWORK >= max(1,2*N).
 *>          For good performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cgeevx.f b/SRC/cgeevx.f
index 76a481e5..8f37a566 100644
--- a/SRC/cgeevx.f
+++ b/SRC/cgeevx.f
@@ -89,8 +89,7 @@
 *>                 to make the rows and columns of A more equal in
 *>                 norm. Do not permute;
 *>          = 'B': Both diagonally scale and permute A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Computed reciprocal condition numbers will be for the matrix
 *>          after balancing and/or permuting. Permuting does not change
 *>          condition numbers (in exact arithmetic), but balancing does.
@@ -120,8 +119,7 @@
 *>          = 'E': Computed for eigenvalues only;
 *>          = 'V': Computed for right eigenvectors only;
 *>          = 'B': Computed for eigenvalues and right eigenvectors.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If SENSE = 'E' or 'B', both left and right eigenvectors
 *>          must also be computed (JOBVL = 'V' and JOBVR = 'V').
 *> \endverbatim
@@ -248,8 +246,7 @@
 *>          LWORK >= max(1,2*N), and if SENSE = 'V' or 'B',
 *>          LWORK >= N*N+2*N.
 *>          For good performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cgegs.f b/SRC/cgegs.f
index 78c42fe2..90d2e345 100644
--- a/SRC/cgegs.f
+++ b/SRC/cgegs.f
@@ -124,8 +124,7 @@
 *>          The non-negative real scalars beta that define the
 *>          eigenvalues of GNEP.  BETA(j) = T(j,j), the diagonal element
 *>          of the triangular factor T.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Together, the quantities alpha = ALPHA(j) and beta = BETA(j)
 *>          represent the j-th eigenvalue of the matrix pair (A,B), in
 *>          one of the forms lambda = alpha/beta or mu = beta/alpha.
@@ -176,8 +175,7 @@
 *>          blocksizes (for CGEQRF, CUNMQR, and CUNGQR.)  Then compute:
 *>          NB  -- MAX of the blocksizes for CGEQRF, CUNMQR, and CUNGQR;
 *>          the optimal LWORK is N*(NB+1).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cgegv.f b/SRC/cgegv.f
index 29f1b99e..69fc8295 100644
--- a/SRC/cgegv.f
+++ b/SRC/cgegv.f
@@ -200,8 +200,7 @@
 *>          blocksizes (for CGEQRF, CUNMQR, and CUNGQR.)  Then compute:
 *>          NB  -- MAX of the blocksizes for CGEQRF, CUNMQR, and CUNGQR;
 *>          The optimal LWORK is  MAX( 2*N, N*(NB+1) ).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cgehrd.f b/SRC/cgehrd.f
index 40b88a64..363c9a84 100644
--- a/SRC/cgehrd.f
+++ b/SRC/cgehrd.f
@@ -100,8 +100,7 @@
 *>          The length of the array WORK.  LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is the
 *>          optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cgelqf.f b/SRC/cgelqf.f
index 3239113d..f88e5257 100644
--- a/SRC/cgelqf.f
+++ b/SRC/cgelqf.f
@@ -89,8 +89,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,M).
 *>          For optimum performance LWORK >= M*NB, where NB is the
 *>          optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cgels.f b/SRC/cgels.f
index 3d1a9dba..29bb2aba 100644
--- a/SRC/cgels.f
+++ b/SRC/cgels.f
@@ -149,8 +149,7 @@
 *>          For optimal performance,
 *>          LWORK >= max( 1, MN + max( MN, NRHS )*NB ).
 *>          where MN = min(M,N) and NB is the optimum block size.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cgelsd.f b/SRC/cgelsd.f
index 039adf25..4ec0862d 100644
--- a/SRC/cgelsd.f
+++ b/SRC/cgelsd.f
@@ -159,8 +159,7 @@
 *>              2 * M + M * NRHS
 *>          if M is less than N, the code will execute correctly.
 *>          For good performance, LWORK should generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the array WORK and the
 *>          minimum sizes of the arrays RWORK and IWORK, and returns
diff --git a/SRC/cgelss.f b/SRC/cgelss.f
index 740be762..e2540780 100644
--- a/SRC/cgelss.f
+++ b/SRC/cgelss.f
@@ -141,8 +141,7 @@
 *>          The dimension of the array WORK. LWORK >= 1, and also:
 *>          LWORK >=  2*min(M,N) + max(M,N,NRHS)
 *>          For good performance, LWORK should generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cgelsy.f b/SRC/cgelsy.f
index 2d5d7cdf..8e8261c6 100644
--- a/SRC/cgelsy.f
+++ b/SRC/cgelsy.f
@@ -169,8 +169,7 @@
 *>          where NB is an upper bound on the blocksize returned
 *>          by ILAENV for the routines CGEQP3, CTZRZF, CTZRQF, CUNMQR,
 *>          and CUNMRZ.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cgeqlf.f b/SRC/cgeqlf.f
index abe6472a..a3db5bbc 100644
--- a/SRC/cgeqlf.f
+++ b/SRC/cgeqlf.f
@@ -92,8 +92,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cgeqp3.f b/SRC/cgeqp3.f
index abec72b5..7c1dc195 100644
--- a/SRC/cgeqp3.f
+++ b/SRC/cgeqp3.f
@@ -101,8 +101,7 @@
 *>          The dimension of the array WORK. LWORK >= N+1.
 *>          For optimal performance LWORK >= ( N+1 )*NB, where NB
 *>          is the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cgeqrf.f b/SRC/cgeqrf.f
index db420ede..cc76e06e 100644
--- a/SRC/cgeqrf.f
+++ b/SRC/cgeqrf.f
@@ -90,8 +90,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cgeqrfp.f b/SRC/cgeqrfp.f
index 7662fdd2..d734cc4c 100644
--- a/SRC/cgeqrfp.f
+++ b/SRC/cgeqrfp.f
@@ -90,8 +90,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cgerfs.f b/SRC/cgerfs.f
index 62a677cf..f06fd213 100644
--- a/SRC/cgerfs.f
+++ b/SRC/cgerfs.f
@@ -162,12 +162,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/cgerfsx.f b/SRC/cgerfsx.f
index 48b27d11..e43b36c6 100644
--- a/SRC/cgerfsx.f
+++ b/SRC/cgerfsx.f
@@ -231,37 +231,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -270,8 +264,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -282,14 +275,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -297,26 +288,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -327,8 +314,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -347,8 +333,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0
@@ -359,8 +344,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -370,8 +354,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/cgesdd.f b/SRC/cgesdd.f
index 8d2f308e..994428f2 100644
--- a/SRC/cgesdd.f
+++ b/SRC/cgesdd.f
@@ -174,8 +174,7 @@
 *>          if JOBZ = 'S' or 'A',
 *>                LWORK >= min(M,N)*min(M,N)+2*min(M,N)+max(M,N).
 *>          For good performance, LWORK should generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, a workspace query is assumed.  The optimal
 *>          size for the WORK array is calculated and stored in WORK(1),
 *>          and no other work except argument checking is performed.
diff --git a/SRC/cgesvd.f b/SRC/cgesvd.f
index d270d979..9abb7632 100644
--- a/SRC/cgesvd.f
+++ b/SRC/cgesvd.f
@@ -82,8 +82,7 @@
 *>                  vectors) are overwritten on the array A;
 *>          = 'N':  no rows of V**H (no right singular vectors) are
 *>                  computed.
-*> \endverbatim
-*> \verbatim
+*>
 *>          JOBVT and JOBU cannot both be 'O'.
 *> \endverbatim
 *>
@@ -172,8 +171,7 @@
 *>          The dimension of the array WORK.
 *>          LWORK >=  MAX(1,2*MIN(M,N)+MAX(M,N)).
 *>          For good performance, LWORK should generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cgesvx.f b/SRC/cgesvx.f
index 660e9118..17dc4ade 100644
--- a/SRC/cgesvx.f
+++ b/SRC/cgesvx.f
@@ -138,8 +138,7 @@
 *>          not 'N', then A must have been equilibrated by the scaling
 *>          factors in R and/or C.  A is not modified if FACT = 'F' or
 *>          'N', or if FACT = 'E' and EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if EQUED .ne. 'N', A is scaled as follows:
 *>          EQUED = 'R':  A := diag(R) * A
 *>          EQUED = 'C':  A := A * diag(C)
@@ -159,13 +158,11 @@
 *>          contains the factors L and U from the factorization
 *>          A = P*L*U as computed by CGETRF.  If EQUED .ne. 'N', then
 *>          AF is the factored form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AF is an output argument and on exit
 *>          returns the factors L and U from the factorization A = P*L*U
 *>          of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then AF is an output argument and on exit
 *>          returns the factors L and U from the factorization A = P*L*U
 *>          of the equilibrated matrix A (see the description of A for
@@ -185,13 +182,11 @@
 *>          contains the pivot indices from the factorization A = P*L*U
 *>          as computed by CGETRF; row i of the matrix was interchanged
 *>          with row IPIV(i).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains the pivot indices from the factorization A = P*L*U
 *>          of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then IPIV is an output argument and on exit
 *>          contains the pivot indices from the factorization A = P*L*U
 *>          of the equilibrated matrix A.
diff --git a/SRC/cgesvxx.f b/SRC/cgesvxx.f
index 32996803..543a52c5 100644
--- a/SRC/cgesvxx.f
+++ b/SRC/cgesvxx.f
@@ -168,8 +168,7 @@
 *>     not 'N', then A must have been equilibrated by the scaling
 *>     factors in R and/or C.  A is not modified if FACT = 'F' or
 *>     'N', or if FACT = 'E' and EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>     On exit, if EQUED .ne. 'N', A is scaled as follows:
 *>     EQUED = 'R':  A := diag(R) * A
 *>     EQUED = 'C':  A := A * diag(C)
@@ -189,13 +188,11 @@
 *>     contains the factors L and U from the factorization
 *>     A = P*L*U as computed by CGETRF.  If EQUED .ne. 'N', then
 *>     AF is the factored form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then AF is an output argument and on exit
 *>     returns the factors L and U from the factorization A = P*L*U
 *>     of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'E', then AF is an output argument and on exit
 *>     returns the factors L and U from the factorization A = P*L*U
 *>     of the equilibrated matrix A (see the description of A for
@@ -215,13 +212,11 @@
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     as computed by CGETRF; row i of the matrix was interchanged
 *>     with row IPIV(i).
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then IPIV is an output argument and on exit
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'E', then IPIV is an output argument and on exit
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     of the equilibrated matrix A.
@@ -361,37 +356,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -400,8 +389,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -412,14 +400,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -427,26 +413,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -457,8 +439,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -477,8 +458,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0
@@ -489,8 +469,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -500,8 +479,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/cgetri.f b/SRC/cgetri.f
index cce96ed4..feb3a381 100644
--- a/SRC/cgetri.f
+++ b/SRC/cgetri.f
@@ -84,8 +84,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,N).
 *>          For optimal performance LWORK >= N*NB, where NB is
 *>          the optimal blocksize returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cgges.f b/SRC/cgges.f
index 719b52de..723c61fd 100644
--- a/SRC/cgges.f
+++ b/SRC/cgges.f
@@ -108,8 +108,7 @@
 *>          to the top left of the Schur form.
 *>          An eigenvalue ALPHA(j)/BETA(j) is selected if
 *>          SELCTG(ALPHA(j),BETA(j)) is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note that a selected complex eigenvalue may no longer satisfy
 *>          SELCTG(ALPHA(j),BETA(j)) = .TRUE. after ordering, since
 *>          ordering may change the value of complex eigenvalues
@@ -171,8 +170,7 @@
 *>          generalized eigenvalues.  ALPHA(j), j=1,...,N  and  BETA(j),
 *>          j=1,...,N  are the diagonals of the complex Schur form (A,B)
 *>          output by CGGES. The  BETA(j) will be non-negative real.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note: the quotients ALPHA(j)/BETA(j) may easily over- or
 *>          underflow, and BETA(j) may even be zero.  Thus, the user
 *>          should avoid naively computing the ratio alpha/beta.
@@ -220,8 +218,7 @@
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.  LWORK >= max(1,2*N).
 *>          For good performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cggesx.f b/SRC/cggesx.f
index 5715f98b..d2073a19 100644
--- a/SRC/cggesx.f
+++ b/SRC/cggesx.f
@@ -182,8 +182,7 @@
 *>          generalized eigenvalues.  ALPHA(j) and BETA(j),j=1,...,N  are
 *>          the diagonals of the complex Schur form (S,T).  BETA(j) will
 *>          be non-negative real.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note: the quotients ALPHA(j)/BETA(j) may easily over- or
 *>          underflow, and BETA(j) may even be zero.  Thus, the user
 *>          should avoid naively computing the ratio alpha/beta.
@@ -254,8 +253,7 @@
 *>          Note also that an error is only returned if
 *>          LWORK < MAX(1,2*N), but if SENSE = 'E' or 'V' or 'B' this may
 *>          not be large enough.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the bound on the optimal size of the WORK
 *>          array and the minimum size of the IWORK array, returns these
@@ -282,8 +280,7 @@
 *>          The dimension of the array WORK.
 *>          If SENSE = 'N' or N = 0, LIWORK >= 1, otherwise
 *>          LIWORK >= N+2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the bound on the optimal size of the
 *>          WORK array and the minimum size of the IWORK array, returns
diff --git a/SRC/cggev.f b/SRC/cggev.f
index 75a36609..69cb6fc2 100644
--- a/SRC/cggev.f
+++ b/SRC/cggev.f
@@ -121,8 +121,7 @@
 *>          BETA is COMPLEX array, dimension (N)
 *>          On exit, ALPHA(j)/BETA(j), j=1,...,N, will be the
 *>          generalized eigenvalues.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note: the quotients ALPHA(j)/BETA(j) may easily over- or
 *>          underflow, and BETA(j) may even be zero.  Thus, the user
 *>          should avoid naively computing the ratio alpha/beta.
@@ -178,8 +177,7 @@
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.  LWORK >= max(1,2*N).
 *>          For good performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cggevx.f b/SRC/cggevx.f
index 20ca5fb3..011ff1f6 100644
--- a/SRC/cggevx.f
+++ b/SRC/cggevx.f
@@ -158,8 +158,7 @@
 *>          BETA is COMPLEX array, dimension (N)
 *>          On exit, ALPHA(j)/BETA(j), j=1,...,N, will be the generalized
 *>          eigenvalues.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note: the quotient ALPHA(j)/BETA(j) ) may easily over- or
 *>          underflow, and BETA(j) may even be zero.  Thus, the user
 *>          should avoid naively computing the ratio ALPHA/BETA.
@@ -289,8 +288,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,2*N).
 *>          If SENSE = 'E', LWORK >= max(1,4*N).
 *>          If SENSE = 'V' or 'B', LWORK >= max(1,2*N*N+2*N).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cggglm.f b/SRC/cggglm.f
index 22363f0c..16d74fc3 100644
--- a/SRC/cggglm.f
+++ b/SRC/cggglm.f
@@ -130,8 +130,7 @@
 *> \param[out] Y
 *> \verbatim
 *>          Y is COMPLEX array, dimension (P)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, X and Y are the solutions of the GLM problem.
 *> \endverbatim
 *>
@@ -148,8 +147,7 @@
 *>          For optimum performance, LWORK >= M+min(N,P)+max(N,P)*NB,
 *>          where NB is an upper bound for the optimal blocksizes for
 *>          CGEQRF, CGERQF, CUNMQR and CUNMRQ.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cgghrd.f b/SRC/cgghrd.f
index e30fbf58..46c6b285 100644
--- a/SRC/cgghrd.f
+++ b/SRC/cgghrd.f
@@ -101,8 +101,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          ILO and IHI mark the rows and columns of A which are to be
 *>          reduced.  It is assumed that A is already upper triangular
 *>          in rows and columns 1:ILO-1 and IHI+1:N.  ILO and IHI are
diff --git a/SRC/cgglse.f b/SRC/cgglse.f
index fe3df1c8..4a62fd4d 100644
--- a/SRC/cgglse.f
+++ b/SRC/cgglse.f
@@ -141,8 +141,7 @@
 *>          For optimum performance LWORK >= P+min(M,N)+max(M,N)*NB,
 *>          where NB is an upper bound for the optimal blocksizes for
 *>          CGEQRF, CGERQF, CUNMQR and CUNMRQ.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cggsvd.f b/SRC/cggsvd.f
index 8522da23..2280d3fc 100644
--- a/SRC/cggsvd.f
+++ b/SRC/cggsvd.f
@@ -170,8 +170,7 @@
 *> \param[out] L
 *> \verbatim
 *>          L is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, K and L specify the dimension of the subblocks
 *>          described in Purpose.
 *>          K + L = effective numerical rank of (A**H,B**H)**H.
@@ -213,8 +212,7 @@
 *> \param[out] BETA
 *> \verbatim
 *>          BETA is REAL array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, ALPHA and BETA contain the generalized singular
 *>          value pairs of A and B;
 *>            ALPHA(1:K) = 1,
@@ -300,12 +298,10 @@
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value.
 *>          > 0:  if INFO = 1, the Jacobi-type procedure failed to
 *>                converge.  For further details, see subroutine CTGSJA.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  TOLA    REAL
 *>  TOLB    REAL
 *>          TOLA and TOLB are the thresholds to determine the effective
diff --git a/SRC/cggsvp.f b/SRC/cggsvp.f
index bcb30940..dd98adc1 100644
--- a/SRC/cggsvp.f
+++ b/SRC/cggsvp.f
@@ -144,8 +144,7 @@
 *> \param[in] TOLB
 *> \verbatim
 *>          TOLB is REAL
-*> \endverbatim
-*> \verbatim
+*>
 *>          TOLA and TOLB are the thresholds to determine the effective
 *>          numerical rank of matrix B and a subblock of A. Generally,
 *>          they are set to
@@ -163,8 +162,7 @@
 *> \param[out] L
 *> \verbatim
 *>          L is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, K and L specify the dimension of the subblocks
 *>          described in Purpose section.
 *>          K + L = effective numerical rank of (A**H,B**H)**H.
diff --git a/SRC/cgtrfs.f b/SRC/cgtrfs.f
index 34fc2297..de8ed77d 100644
--- a/SRC/cgtrfs.f
+++ b/SRC/cgtrfs.f
@@ -185,12 +185,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/cgtsvx.f b/SRC/cgtsvx.f
index cae1b36c..27667e1c 100644
--- a/SRC/cgtsvx.f
+++ b/SRC/cgtsvx.f
@@ -136,8 +136,7 @@
 *>          If FACT = 'F', then DLF is an input argument and on entry
 *>          contains the (n-1) multipliers that define the matrix L from
 *>          the LU factorization of A as computed by CGTTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then DLF is an output argument and on exit
 *>          contains the (n-1) multipliers that define the matrix L from
 *>          the LU factorization of A.
@@ -149,8 +148,7 @@
 *>          If FACT = 'F', then DF is an input argument and on entry
 *>          contains the n diagonal elements of the upper triangular
 *>          matrix U from the LU factorization of A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then DF is an output argument and on exit
 *>          contains the n diagonal elements of the upper triangular
 *>          matrix U from the LU factorization of A.
@@ -161,8 +159,7 @@
 *>          DUF is or output) COMPLEX array, dimension (N-1)
 *>          If FACT = 'F', then DUF is an input argument and on entry
 *>          contains the (n-1) elements of the first superdiagonal of U.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then DUF is an output argument and on exit
 *>          contains the (n-1) elements of the first superdiagonal of U.
 *> \endverbatim
@@ -173,8 +170,7 @@
 *>          If FACT = 'F', then DU2 is an input argument and on entry
 *>          contains the (n-2) elements of the second superdiagonal of
 *>          U.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then DU2 is an output argument and on exit
 *>          contains the (n-2) elements of the second superdiagonal of
 *>          U.
@@ -186,8 +182,7 @@
 *>          If FACT = 'F', then IPIV is an input argument and on entry
 *>          contains the pivot indices from the LU factorization of A as
 *>          computed by CGTTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains the pivot indices from the LU factorization of A;
 *>          row i of the matrix was interchanged with row IPIV(i).
diff --git a/SRC/cgttrf.f b/SRC/cgttrf.f
index 492422fb..17d40172 100644
--- a/SRC/cgttrf.f
+++ b/SRC/cgttrf.f
@@ -59,8 +59,7 @@
 *>          DL is COMPLEX array, dimension (N-1)
 *>          On entry, DL must contain the (n-1) sub-diagonal elements of
 *>          A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, DL is overwritten by the (n-1) multipliers that
 *>          define the matrix L from the LU factorization of A.
 *> \endverbatim
@@ -69,8 +68,7 @@
 *> \verbatim
 *>          D is COMPLEX array, dimension (N)
 *>          On entry, D must contain the diagonal elements of A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, D is overwritten by the n diagonal elements of the
 *>          upper triangular matrix U from the LU factorization of A.
 *> \endverbatim
@@ -80,8 +78,7 @@
 *>          DU is COMPLEX array, dimension (N-1)
 *>          On entry, DU must contain the (n-1) super-diagonal elements
 *>          of A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, DU is overwritten by the (n-1) elements of the first
 *>          super-diagonal of U.
 *> \endverbatim
diff --git a/SRC/chbev.f b/SRC/chbev.f
index dc2dbacc..49c0a344 100644
--- a/SRC/chbev.f
+++ b/SRC/chbev.f
@@ -80,8 +80,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AB is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the first
 *>          superdiagonal and the diagonal of the tridiagonal matrix T
diff --git a/SRC/chbevd.f b/SRC/chbevd.f
index 9308a47c..42b1a565 100644
--- a/SRC/chbevd.f
+++ b/SRC/chbevd.f
@@ -89,8 +89,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AB is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the first
 *>          superdiagonal and the diagonal of the tridiagonal matrix T
@@ -140,8 +139,7 @@
 *>          If N <= 1,               LWORK must be at least 1.
 *>          If JOBZ = 'N' and N > 1, LWORK must be at least N.
 *>          If JOBZ = 'V' and N > 1, LWORK must be at least 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK, RWORK and
 *>          IWORK arrays, returns these values as the first entries of
@@ -164,8 +162,7 @@
 *>          If JOBZ = 'N' and N > 1, LRWORK must be at least N.
 *>          If JOBZ = 'V' and N > 1, LRWORK must be at least
 *>                        1 + 5*N + 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LRWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
@@ -185,8 +182,7 @@
 *>          The dimension of array IWORK.
 *>          If JOBZ = 'N' or N <= 1, LIWORK must be at least 1.
 *>          If JOBZ = 'V' and N > 1, LIWORK must be at least 3 + 5*N .
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
diff --git a/SRC/chbevx.f b/SRC/chbevx.f
index a10b7c54..74cc1b48 100644
--- a/SRC/chbevx.f
+++ b/SRC/chbevx.f
@@ -95,8 +95,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AB is overwritten by values generated during the
 *>          reduction to tridiagonal form.
 *> \endverbatim
@@ -156,24 +155,20 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing AB to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*SLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
 *>          eigenvectors did not converge, try setting ABSTOL to
 *>          2*SLAMCH('S').
-*> \endverbatim
-*> \verbatim
+*>
 *>          See "Computing Small Singular Values of Bidiagonal Matrices
 *>          with Guaranteed High Relative Accuracy," by Demmel and
 *>          Kahan, LAPACK Working Note #3.
diff --git a/SRC/chbgst.f b/SRC/chbgst.f
index 469a021a..7ca26620 100644
--- a/SRC/chbgst.f
+++ b/SRC/chbgst.f
@@ -94,8 +94,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(ka+1+i-j,j) = A(i,j) for max(1,j-ka)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+ka).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the transformed matrix X**H*A*X, stored in the same
 *>          format as A.
 *> \endverbatim
diff --git a/SRC/chbgv.f b/SRC/chbgv.f
index b3cf165b..fcb39ee6 100644
--- a/SRC/chbgv.f
+++ b/SRC/chbgv.f
@@ -90,8 +90,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(ka+1+i-j,j) = A(i,j) for max(1,j-ka)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+ka).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AB are destroyed.
 *> \endverbatim
 *>
@@ -110,8 +109,7 @@
 *>          as follows:
 *>          if UPLO = 'U', BB(kb+1+i-j,j) = B(i,j) for max(1,j-kb)<=i<=j;
 *>          if UPLO = 'L', BB(1+i-j,j)    = B(i,j) for j<=i<=min(n,j+kb).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the factor S from the split Cholesky factorization
 *>          B = S**H*S, as returned by CPBSTF.
 *> \endverbatim
diff --git a/SRC/chbgvd.f b/SRC/chbgvd.f
index 6f48693b..68eed18f 100644
--- a/SRC/chbgvd.f
+++ b/SRC/chbgvd.f
@@ -101,8 +101,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(ka+1+i-j,j) = A(i,j) for max(1,j-ka)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+ka).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AB are destroyed.
 *> \endverbatim
 *>
@@ -121,8 +120,7 @@
 *>          as follows:
 *>          if UPLO = 'U', BB(kb+1+i-j,j) = B(i,j) for max(1,j-kb)<=i<=j;
 *>          if UPLO = 'L', BB(1+i-j,j)    = B(i,j) for j<=i<=min(n,j+kb).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the factor S from the split Cholesky factorization
 *>          B = S**H*S, as returned by CPBSTF.
 *> \endverbatim
@@ -169,8 +167,7 @@
 *>          If N <= 1,               LWORK >= 1.
 *>          If JOBZ = 'N' and N > 1, LWORK >= N.
 *>          If JOBZ = 'V' and N > 1, LWORK >= 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK, RWORK and
 *>          IWORK arrays, returns these values as the first entries of
@@ -191,8 +188,7 @@
 *>          If N <= 1,               LRWORK >= 1.
 *>          If JOBZ = 'N' and N > 1, LRWORK >= N.
 *>          If JOBZ = 'V' and N > 1, LRWORK >= 1 + 5*N + 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LRWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
@@ -212,8 +208,7 @@
 *>          The dimension of array IWORK.
 *>          If JOBZ = 'N' or N <= 1, LIWORK >= 1.
 *>          If JOBZ = 'V' and N > 1, LIWORK >= 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
diff --git a/SRC/chbgvx.f b/SRC/chbgvx.f
index fb40c44f..451ffc35 100644
--- a/SRC/chbgvx.f
+++ b/SRC/chbgvx.f
@@ -105,8 +105,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(ka+1+i-j,j) = A(i,j) for max(1,j-ka)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+ka).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AB are destroyed.
 *> \endverbatim
 *>
@@ -125,8 +124,7 @@
 *>          as follows:
 *>          if UPLO = 'U', BB(kb+1+i-j,j) = B(i,j) for max(1,j-kb)<=i<=j;
 *>          if UPLO = 'L', BB(1+i-j,j)    = B(i,j) for j<=i<=min(n,j+kb).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the factor S from the split Cholesky factorization
 *>          B = S**H*S, as returned by CPBSTF.
 *> \endverbatim
@@ -161,8 +159,7 @@
 *> \param[in] VU
 *> \verbatim
 *>          VU is REAL
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='V', the lower and upper bounds of the interval to
 *>          be searched for eigenvalues. VL < VU.
 *>          Not referenced if RANGE = 'A' or 'I'.
@@ -176,8 +173,7 @@
 *> \param[in] IU
 *> \verbatim
 *>          IU is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='I', the indices (in ascending order) of the
 *>          smallest and largest eigenvalues to be returned.
 *>          1 <= IL <= IU <= N, if N > 0; IL = 1 and IU = 0 if N = 0.
@@ -191,17 +187,14 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing AP to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*SLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
diff --git a/SRC/chbtrd.f b/SRC/chbtrd.f
index fe68cef6..1e2ec2d8 100644
--- a/SRC/chbtrd.f
+++ b/SRC/chbtrd.f
@@ -114,8 +114,7 @@
 *>          Q is COMPLEX array, dimension (LDQ,N)
 *>          On entry, if VECT = 'U', then Q must contain an N-by-N
 *>          matrix X; if VECT = 'N' or 'V', then Q need not be set.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit:
 *>          if VECT = 'V', Q contains the N-by-N unitary matrix Q;
 *>          if VECT = 'U', Q contains the product X*Q;
diff --git a/SRC/cheev.f b/SRC/cheev.f
index 03d4436d..fc702c1f 100644
--- a/SRC/cheev.f
+++ b/SRC/cheev.f
@@ -103,8 +103,7 @@
 *>          The length of the array WORK.  LWORK >= max(1,2*N-1).
 *>          For optimal efficiency, LWORK >= (NB+1)*N,
 *>          where NB is the blocksize for CHETRD returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cheevd.f b/SRC/cheevd.f
index 583914cb..817ed7b6 100644
--- a/SRC/cheevd.f
+++ b/SRC/cheevd.f
@@ -113,8 +113,7 @@
 *>          If N <= 1,                LWORK must be at least 1.
 *>          If JOBZ  = 'N' and N > 1, LWORK must be at least N + 1.
 *>          If JOBZ  = 'V' and N > 1, LWORK must be at least 2*N + N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK, RWORK and
 *>          IWORK arrays, returns these values as the first entries of
@@ -137,8 +136,7 @@
 *>          If JOBZ  = 'N' and N > 1, LRWORK must be at least N.
 *>          If JOBZ  = 'V' and N > 1, LRWORK must be at least
 *>                         1 + 5*N + 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LRWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
@@ -159,8 +157,7 @@
 *>          If N <= 1,                LIWORK must be at least 1.
 *>          If JOBZ  = 'N' and N > 1, LIWORK must be at least 1.
 *>          If JOBZ  = 'V' and N > 1, LIWORK must be at least 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
diff --git a/SRC/cheevr.f b/SRC/cheevr.f
index cbe9bfdf..64b2513a 100644
--- a/SRC/cheevr.f
+++ b/SRC/cheevr.f
@@ -187,22 +187,18 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing A to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          See "Computing Small Singular Values of Bidiagonal Matrices
 *>          with Guaranteed High Relative Accuracy," by Demmel and
 *>          Kahan, LAPACK Working Note #3.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If high relative accuracy is important, set ABSTOL to
 *>          SLAMCH( 'Safe minimum' ).  Doing so will guarantee that
 *>          eigenvalues are computed to high relative accuracy when
@@ -272,8 +268,7 @@
 *>          For optimal efficiency, LWORK >= (NB+1)*N,
 *>          where NB is the max of the blocksize for CHETRD and for
 *>          CUNMTR as returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK, RWORK and
 *>          IWORK arrays, returns these values as the first entries of
@@ -292,8 +287,7 @@
 *> \verbatim
 *>          LRWORK is INTEGER
 *>          The length of the array RWORK.  LRWORK >= max(1,24*N).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LRWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
@@ -312,8 +306,7 @@
 *> \verbatim
 *>          LIWORK is INTEGER
 *>          The dimension of the array IWORK.  LIWORK >= max(1,10*N).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
diff --git a/SRC/cheevx.f b/SRC/cheevx.f
index 2073733b..f2d41f68 100644
--- a/SRC/cheevx.f
+++ b/SRC/cheevx.f
@@ -131,24 +131,20 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing A to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*SLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
 *>          eigenvectors did not converge, try setting ABSTOL to
 *>          2*SLAMCH('S').
-*> \endverbatim
-*> \verbatim
+*>
 *>          See "Computing Small Singular Values of Bidiagonal Matrices
 *>          with Guaranteed High Relative Accuracy," by Demmel and
 *>          Kahan, LAPACK Working Note #3.
@@ -205,8 +201,7 @@
 *>          For optimal efficiency, LWORK >= (NB+1)*N,
 *>          where NB is the max of the blocksize for CHETRD and for
 *>          CUNMTR as returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/chegs2.f b/SRC/chegs2.f
index c335b06c..dd09df1c 100644
--- a/SRC/chegs2.f
+++ b/SRC/chegs2.f
@@ -82,8 +82,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the transformed matrix, stored in the
 *>          same format as A.
 *> \endverbatim
diff --git a/SRC/chegst.f b/SRC/chegst.f
index 32bbabdf..400252f5 100644
--- a/SRC/chegst.f
+++ b/SRC/chegst.f
@@ -82,8 +82,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the transformed matrix, stored in the
 *>          same format as A.
 *> \endverbatim
diff --git a/SRC/chegv.f b/SRC/chegv.f
index 73c0b729..4c5dbeb2 100644
--- a/SRC/chegv.f
+++ b/SRC/chegv.f
@@ -84,8 +84,7 @@
 *>          upper triangular part of the matrix A.  If UPLO = 'L',
 *>          the leading N-by-N lower triangular part of A contains
 *>          the lower triangular part of the matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if JOBZ = 'V', then if INFO = 0, A contains the
 *>          matrix Z of eigenvectors.  The eigenvectors are normalized
 *>          as follows:
@@ -110,8 +109,7 @@
 *>          contains the upper triangular part of the matrix B.
 *>          If UPLO = 'L', the leading N-by-N lower triangular part of B
 *>          contains the lower triangular part of the matrix B.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO <= N, the part of B containing the matrix is
 *>          overwritten by the triangular factor U or L from the Cholesky
 *>          factorization B = U**H*U or B = L*L**H.
@@ -141,8 +139,7 @@
 *>          The length of the array WORK.  LWORK >= max(1,2*N-1).
 *>          For optimal efficiency, LWORK >= (NB+1)*N,
 *>          where NB is the blocksize for CHETRD returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/chegvd.f b/SRC/chegvd.f
index 53215629..88e9de1a 100644
--- a/SRC/chegvd.f
+++ b/SRC/chegvd.f
@@ -92,8 +92,7 @@
 *>          upper triangular part of the matrix A.  If UPLO = 'L',
 *>          the leading N-by-N lower triangular part of A contains
 *>          the lower triangular part of the matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if JOBZ = 'V', then if INFO = 0, A contains the
 *>          matrix Z of eigenvectors.  The eigenvectors are normalized
 *>          as follows:
@@ -118,8 +117,7 @@
 *>          upper triangular part of the matrix B.  If UPLO = 'L',
 *>          the leading N-by-N lower triangular part of B contains
 *>          the lower triangular part of the matrix B.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO <= N, the part of B containing the matrix is
 *>          overwritten by the triangular factor U or L from the Cholesky
 *>          factorization B = U**H*U or B = L*L**H.
@@ -150,8 +148,7 @@
 *>          If N <= 1,                LWORK >= 1.
 *>          If JOBZ  = 'N' and N > 1, LWORK >= N + 1.
 *>          If JOBZ  = 'V' and N > 1, LWORK >= 2*N + N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK, RWORK and
 *>          IWORK arrays, returns these values as the first entries of
@@ -172,8 +169,7 @@
 *>          If N <= 1,                LRWORK >= 1.
 *>          If JOBZ  = 'N' and N > 1, LRWORK >= N.
 *>          If JOBZ  = 'V' and N > 1, LRWORK >= 1 + 5*N + 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LRWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
@@ -194,8 +190,7 @@
 *>          If N <= 1,                LIWORK >= 1.
 *>          If JOBZ  = 'N' and N > 1, LIWORK >= 1.
 *>          If JOBZ  = 'V' and N > 1, LIWORK >= 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
diff --git a/SRC/chegvx.f b/SRC/chegvx.f
index 0b191152..ca395bc4 100644
--- a/SRC/chegvx.f
+++ b/SRC/chegvx.f
@@ -138,8 +138,7 @@
 *> \param[in] VU
 *> \verbatim
 *>          VU is REAL
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='V', the lower and upper bounds of the interval to
 *>          be searched for eigenvalues. VL < VU.
 *>          Not referenced if RANGE = 'A' or 'I'.
@@ -153,8 +152,7 @@
 *> \param[in] IU
 *> \verbatim
 *>          IU is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='I', the indices (in ascending order) of the
 *>          smallest and largest eigenvalues to be returned.
 *>          1 <= IL <= IU <= N, if N > 0; IL = 1 and IU = 0 if N = 0.
@@ -238,8 +236,7 @@
 *>          The length of the array WORK.  LWORK >= max(1,2*N).
 *>          For optimal efficiency, LWORK >= (NB+1)*N,
 *>          where NB is the blocksize for CHETRD returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cherfs.f b/SRC/cherfs.f
index d4c4efa0..38afe80b 100644
--- a/SRC/cherfs.f
+++ b/SRC/cherfs.f
@@ -168,12 +168,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/cherfsx.f b/SRC/cherfsx.f
index d52eec1b..33a58533 100644
--- a/SRC/cherfsx.f
+++ b/SRC/cherfsx.f
@@ -218,37 +218,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -257,8 +251,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -269,14 +262,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -284,26 +275,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -314,8 +301,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -334,8 +320,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0
@@ -346,8 +331,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -357,8 +341,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/chesv.f b/SRC/chesv.f
index e34c334a..ffc0a76a 100644
--- a/SRC/chesv.f
+++ b/SRC/chesv.f
@@ -85,8 +85,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the block diagonal matrix D and the
 *>          multipliers used to obtain the factor U or L from the
 *>          factorization A = U*D*U**H or A = L*D*L**H as computed by
@@ -140,8 +139,7 @@
 *>          CHETRF.
 *>          for LWORK < N, TRS will be done with Level BLAS 2
 *>          for LWORK >= N, TRS will be done with Level BLAS 3
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/chesvx.f b/SRC/chesvx.f
index f6dd7b71..cf38ff88 100644
--- a/SRC/chesvx.f
+++ b/SRC/chesvx.f
@@ -136,8 +136,7 @@
 *>          contains the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
 *>          A = U*D*U**H or A = L*D*L**H as computed by CHETRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AF is an output argument and on exit
 *>          returns the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
@@ -163,8 +162,7 @@
 *>          is a 2-by-2 diagonal block.  If UPLO = 'L' and IPIV(k) =
 *>          IPIV(k+1) < 0, then rows and columns k+1 and -IPIV(k) were
 *>          interchanged and D(k:k+1,k:k+1) is a 2-by-2 diagonal block.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains details of the interchanges and the block structure
 *>          of D, as determined by CHETRF.
@@ -237,8 +235,7 @@
 *>          The length of WORK.  LWORK >= max(1,2*N), and for best
 *>          performance, when FACT = 'N', LWORK >= max(1,2*N,N*NB), where
 *>          NB is the optimal blocksize for CHETRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/chesvxx.f b/SRC/chesvxx.f
index bc134549..9c6ac832 100644
--- a/SRC/chesvxx.f
+++ b/SRC/chesvxx.f
@@ -166,8 +166,7 @@
 *>     N-by-N lower triangular part of A contains the lower
 *>     triangular part of the matrix A, and the strictly upper
 *>     triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>     On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>     diag(S)*A*diag(S).
 *> \endverbatim
@@ -185,8 +184,7 @@
 *>     contains the block diagonal matrix D and the multipliers
 *>     used to obtain the factor U or L from the factorization A =
 *>     U*D*U**T or A = L*D*L**T as computed by SSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then AF is an output argument and on exit
 *>     returns the block diagonal matrix D and the multipliers
 *>     used to obtain the factor U or L from the factorization A =
@@ -212,8 +210,7 @@
 *>     diagonal block.  If UPLO = 'L' and IPIV(k) = IPIV(k+1) < 0,
 *>     then rows and columns k+1 and -IPIV(k) were interchanged
 *>     and D(k:k+1,k:k+1) is a 2-by-2 diagonal block.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then IPIV is an output argument and on exit
 *>     contains details of the interchanges and the block
 *>     structure of D, as determined by CHETRF.
@@ -324,37 +321,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -363,8 +354,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -375,14 +365,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -390,26 +378,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -420,8 +404,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -440,8 +423,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0
@@ -452,8 +434,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -463,8 +444,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/cheswapr.f b/SRC/cheswapr.f
index 03d3188f..bb4a17f3 100644
--- a/SRC/cheswapr.f
+++ b/SRC/cheswapr.f
@@ -61,8 +61,7 @@
 *>          A is COMPLEX array, dimension (LDA,N)
 *>          On entry, the NB diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by CSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/chetf2.f b/SRC/chetf2.f
index 982fcb49..76d62511 100644
--- a/SRC/chetf2.f
+++ b/SRC/chetf2.f
@@ -76,8 +76,7 @@
 *>          leading n-by-n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L (see below for further details).
 *> \endverbatim
diff --git a/SRC/chetrd.f b/SRC/chetrd.f
index 870f60b3..0c1927f6 100644
--- a/SRC/chetrd.f
+++ b/SRC/chetrd.f
@@ -118,8 +118,7 @@
 *>          The dimension of the array WORK.  LWORK >= 1.
 *>          For optimum performance LWORK >= N*NB, where NB is the
 *>          optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/chetrf.f b/SRC/chetrf.f
index 33ed8bc2..7f14b7d6 100644
--- a/SRC/chetrf.f
+++ b/SRC/chetrf.f
@@ -75,8 +75,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L (see below for further details).
 *> \endverbatim
diff --git a/SRC/chetri.f b/SRC/chetri.f
index d2681dfb..4c916ce7 100644
--- a/SRC/chetri.f
+++ b/SRC/chetri.f
@@ -64,8 +64,7 @@
 *>          A is COMPLEX array, dimension (LDA,N)
 *>          On entry, the block diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by CHETRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (Hermitian) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/chetri2.f b/SRC/chetri2.f
index 3af108ac..02fd41c3 100644
--- a/SRC/chetri2.f
+++ b/SRC/chetri2.f
@@ -65,8 +65,7 @@
 *>          A is COMPLEX array, dimension (LDA,N)
 *>          On entry, the NB diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by CHETRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/chetri2x.f b/SRC/chetri2x.f
index 2b32ea64..a065a6b5 100644
--- a/SRC/chetri2x.f
+++ b/SRC/chetri2x.f
@@ -64,8 +64,7 @@
 *>          A is COMPLEX array, dimension (LDA,N)
 *>          On entry, the NNB diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by CHETRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/chfrk.f b/SRC/chfrk.f
index dcd698b9..a289f828 100644
--- a/SRC/chfrk.f
+++ b/SRC/chfrk.f
@@ -68,16 +68,13 @@
 *>           On  entry,   UPLO  specifies  whether  the  upper  or  lower
 *>           triangular  part  of the  array  C  is to be  referenced  as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'U' or 'u'   Only the  upper triangular part of  C
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'L' or 'l'   Only the  lower triangular part of  C
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -86,14 +83,11 @@
 *>          TRANS is CHARACTER*1
 *>           On entry,  TRANS  specifies the operation to be performed as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              TRANS = 'N' or 'n'   C := alpha*A*A**H + beta*C.
-*> \endverbatim
-*> \verbatim
+*>
 *>              TRANS = 'C' or 'c'   C := alpha*A**H*A + beta*C.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/chgeqz.f b/SRC/chgeqz.f
index e5776d8f..52cf95dd 100644
--- a/SRC/chgeqz.f
+++ b/SRC/chgeqz.f
@@ -180,8 +180,7 @@
 *>          The real non-negative scalars beta that define the
 *>          eigenvalues of GNEP.  BETA(i) = P(i,i) in the generalized
 *>          Schur factorization.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Together, the quantities alpha = ALPHA(j) and beta = BETA(j)
 *>          represent the j-th eigenvalue of the matrix pair (A,B), in
 *>          one of the forms lambda = alpha/beta or mu = beta/alpha.
@@ -235,8 +234,7 @@
 *> \verbatim
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.  LWORK >= max(1,N).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/chpev.f b/SRC/chpev.f
index 457c87a3..16d72408 100644
--- a/SRC/chpev.f
+++ b/SRC/chpev.f
@@ -72,8 +72,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AP is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the diagonal
 *>          and first superdiagonal of the tridiagonal matrix T overwrite
diff --git a/SRC/chpevd.f b/SRC/chpevd.f
index 902ce0f6..e5b446fa 100644
--- a/SRC/chpevd.f
+++ b/SRC/chpevd.f
@@ -81,8 +81,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AP is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the diagonal
 *>          and first superdiagonal of the tridiagonal matrix T overwrite
@@ -126,8 +125,7 @@
 *>          If N <= 1,               LWORK must be at least 1.
 *>          If JOBZ = 'N' and N > 1, LWORK must be at least N.
 *>          If JOBZ = 'V' and N > 1, LWORK must be at least 2*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the required sizes of the WORK, RWORK and
 *>          IWORK arrays, returns these values as the first entries of
@@ -149,8 +147,7 @@
 *>          If JOBZ = 'N' and N > 1, LRWORK must be at least N.
 *>          If JOBZ = 'V' and N > 1, LRWORK must be at least
 *>                    1 + 5*N + 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LRWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the required sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
@@ -170,8 +167,7 @@
 *>          The dimension of array IWORK.
 *>          If JOBZ  = 'N' or N <= 1, LIWORK must be at least 1.
 *>          If JOBZ  = 'V' and N > 1, LIWORK must be at least 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the required sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
diff --git a/SRC/chpevx.f b/SRC/chpevx.f
index 254e4654..4634b340 100644
--- a/SRC/chpevx.f
+++ b/SRC/chpevx.f
@@ -86,8 +86,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AP is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the diagonal
 *>          and first superdiagonal of the tridiagonal matrix T overwrite
@@ -130,24 +129,20 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing AP to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*SLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
 *>          eigenvectors did not converge, try setting ABSTOL to
 *>          2*SLAMCH('S').
-*> \endverbatim
-*> \verbatim
+*>
 *>          See "Computing Small Singular Values of Bidiagonal Matrices
 *>          with Guaranteed High Relative Accuracy," by Demmel and
 *>          Kahan, LAPACK Working Note #3.
diff --git a/SRC/chpgst.f b/SRC/chpgst.f
index a183d760..1368e71a 100644
--- a/SRC/chpgst.f
+++ b/SRC/chpgst.f
@@ -80,8 +80,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the transformed matrix, stored in the
 *>          same format as A.
 *> \endverbatim
diff --git a/SRC/chpgv.f b/SRC/chpgv.f
index f956d95f..faa9b782 100644
--- a/SRC/chpgv.f
+++ b/SRC/chpgv.f
@@ -84,8 +84,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AP are destroyed.
 *> \endverbatim
 *>
@@ -97,8 +96,7 @@
 *>          is stored in the array BP as follows:
 *>          if UPLO = 'U', BP(i + (j-1)*j/2) = B(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', BP(i + (j-1)*(2*n-j)/2) = B(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the triangular factor U or L from the Cholesky
 *>          factorization B = U**H*U or B = L*L**H, in the same storage
 *>          format as B.
diff --git a/SRC/chpgvd.f b/SRC/chpgvd.f
index a0623cca..b873f5a5 100644
--- a/SRC/chpgvd.f
+++ b/SRC/chpgvd.f
@@ -93,8 +93,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AP are destroyed.
 *> \endverbatim
 *>
@@ -106,8 +105,7 @@
 *>          is stored in the array BP as follows:
 *>          if UPLO = 'U', BP(i + (j-1)*j/2) = B(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', BP(i + (j-1)*(2*n-j)/2) = B(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the triangular factor U or L from the Cholesky
 *>          factorization B = U**H*U or B = L*L**H, in the same storage
 *>          format as B.
@@ -149,8 +147,7 @@
 *>          If N <= 1,               LWORK >= 1.
 *>          If JOBZ = 'N' and N > 1, LWORK >= N.
 *>          If JOBZ = 'V' and N > 1, LWORK >= 2*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the required sizes of the WORK, RWORK and
 *>          IWORK arrays, returns these values as the first entries of
@@ -171,8 +168,7 @@
 *>          If N <= 1,               LRWORK >= 1.
 *>          If JOBZ = 'N' and N > 1, LRWORK >= N.
 *>          If JOBZ = 'V' and N > 1, LRWORK >= 1 + 5*N + 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LRWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the required sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
@@ -192,8 +188,7 @@
 *>          The dimension of array IWORK.
 *>          If JOBZ  = 'N' or N <= 1, LIWORK >= 1.
 *>          If JOBZ  = 'V' and N > 1, LIWORK >= 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the required sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
diff --git a/SRC/chpgvx.f b/SRC/chpgvx.f
index 6dce19a2..a1d0b904 100644
--- a/SRC/chpgvx.f
+++ b/SRC/chpgvx.f
@@ -98,8 +98,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AP are destroyed.
 *> \endverbatim
 *>
@@ -111,8 +110,7 @@
 *>          is stored in the array BP as follows:
 *>          if UPLO = 'U', BP(i + (j-1)*j/2) = B(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', BP(i + (j-1)*(2*n-j)/2) = B(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the triangular factor U or L from the Cholesky
 *>          factorization B = U**H*U or B = L*L**H, in the same storage
 *>          format as B.
@@ -126,8 +124,7 @@
 *> \param[in] VU
 *> \verbatim
 *>          VU is REAL
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='V', the lower and upper bounds of the interval to
 *>          be searched for eigenvalues. VL < VU.
 *>          Not referenced if RANGE = 'A' or 'I'.
@@ -141,8 +138,7 @@
 *> \param[in] IU
 *> \verbatim
 *>          IU is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='I', the indices (in ascending order) of the
 *>          smallest and largest eigenvalues to be returned.
 *>          1 <= IL <= IU <= N, if N > 0; IL = 1 and IU = 0 if N = 0.
@@ -156,17 +152,14 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing AP to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*SLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
@@ -199,8 +192,7 @@
 *>          The eigenvectors are normalized as follows:
 *>          if ITYPE = 1 or 2, Z**H*B*Z = I;
 *>          if ITYPE = 3, Z**H*inv(B)*Z = I.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If an eigenvector fails to converge, then that column of Z
 *>          contains the latest approximation to the eigenvector, and the
 *>          index of the eigenvector is returned in IFAIL.
diff --git a/SRC/chprfs.f b/SRC/chprfs.f
index 900e3037..863f6aee 100644
--- a/SRC/chprfs.f
+++ b/SRC/chprfs.f
@@ -156,12 +156,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/chpsv.f b/SRC/chpsv.f
index 85360d54..6f3ff6d7 100644
--- a/SRC/chpsv.f
+++ b/SRC/chpsv.f
@@ -83,8 +83,7 @@
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
 *>          A = U*D*U**H or A = L*D*L**H as computed by CHPTRF, stored as
diff --git a/SRC/chpsvx.f b/SRC/chpsvx.f
index efe7671b..3fe0aea5 100644
--- a/SRC/chpsvx.f
+++ b/SRC/chpsvx.f
@@ -128,8 +128,7 @@
 *>          to obtain the factor U or L from the factorization
 *>          A = U*D*U**H or A = L*D*L**H as computed by CHPTRF, stored as
 *>          a packed triangular matrix in the same storage format as A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AFP is an output argument and on exit
 *>          contains the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
@@ -150,8 +149,7 @@
 *>          is a 2-by-2 diagonal block.  If UPLO = 'L' and IPIV(k) =
 *>          IPIV(k+1) < 0, then rows and columns k+1 and -IPIV(k) were
 *>          interchanged and D(k:k+1,k:k+1) is a 2-by-2 diagonal block.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains details of the interchanges and the block structure
 *>          of D, as determined by CHPTRF.
diff --git a/SRC/chptrf.f b/SRC/chptrf.f
index fb0d03cc..e78d98e6 100644
--- a/SRC/chptrf.f
+++ b/SRC/chptrf.f
@@ -70,8 +70,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L, stored as a packed triangular
 *>          matrix overwriting A (see below for further details).
diff --git a/SRC/chptri.f b/SRC/chptri.f
index b052d48c..69dbc560 100644
--- a/SRC/chptri.f
+++ b/SRC/chptri.f
@@ -65,8 +65,7 @@
 *>          On entry, the block diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by CHPTRF,
 *>          stored as a packed triangular matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (Hermitian) inverse of the original
 *>          matrix, stored as a packed triangular matrix. The j-th column
 *>          of inv(A) is stored in the array AP as follows:
diff --git a/SRC/chseqr.f b/SRC/chseqr.f
index f54a9307..b1e2423c 100644
--- a/SRC/chseqr.f
+++ b/SRC/chseqr.f
@@ -82,8 +82,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>           It is assumed that H is already upper triangular in rows
 *>           and columns 1:ILO-1 and IHI+1:N. ILO and IHI are normally
 *>           set by a previous call to CGEBAL, and then passed to ZGEHRD
@@ -102,8 +101,7 @@
 *>           Schur form). If INFO = 0 and JOB = 'E', the contents of
 *>           H are unspecified on exit.  (The output value of H when
 *>           INFO.GT.0 is given under the description of INFO below.)
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unlike earlier versions of CHSEQR, this subroutine may
 *>           explicitly H(i,j) = 0 for i.GT.j and j = 1, 2, ... ILO-1
 *>           or j = IHI+1, IHI+2, ... N.
@@ -162,8 +160,7 @@
 *>           may be required for optimal performance.  A workspace
 *>           query is recommended to determine the optimal workspace
 *>           size.
-*> \endverbatim
-*> \verbatim
+*>
 *>           If LWORK = -1, then CHSEQR does a workspace query.
 *>           In this case, CHSEQR checks the input parameters and
 *>           estimates the optimal workspace size for the given
@@ -182,42 +179,33 @@
 *>                the eigenvalues.  Elements 1:ilo-1 and i+1:n of WR
 *>                and WI contain those eigenvalues which have been
 *>                successfully computed.  (Failures are rare.)
-*> \endverbatim
-*> \verbatim
+*>
 *>                If INFO .GT. 0 and JOB = 'E', then on exit, the
 *>                remaining unconverged eigenvalues are the eigen-
 *>                values of the upper Hessenberg matrix rows and
 *>                columns ILO through INFO of the final, output
 *>                value of H.
-*> \endverbatim
-*> \verbatim
+*>
 *>                If INFO .GT. 0 and JOB   = 'S', then on exit
-*> \endverbatim
-*> \verbatim
+*>
 *>           (*)  (initial value of H)*U  = U*(final value of H)
-*> \endverbatim
-*> \verbatim
+*>
 *>                where U is a unitary matrix.  The final
 *>                value of  H is upper Hessenberg and triangular in
 *>                rows and columns INFO+1 through IHI.
-*> \endverbatim
-*> \verbatim
+*>
 *>                If INFO .GT. 0 and COMPZ = 'V', then on exit
-*> \endverbatim
-*> \verbatim
+*>
 *>                  (final value of Z)  =  (initial value of Z)*U
-*> \endverbatim
-*> \verbatim
+*>
 *>                where U is the unitary matrix in (*) (regard-
 *>                less of the value of JOB.)
-*> \endverbatim
-*> \verbatim
+*>
 *>                If INFO .GT. 0 and COMPZ = 'I', then on exit
 *>                      (final value of Z)  = U
 *>                where U is the unitary matrix in (*) (regard-
 *>                less of the value of JOB.)
-*> \endverbatim
-*> \verbatim
+*>
 *>                If INFO .GT. 0 and COMPZ = 'N', then Z is not
 *>                accessed.
 *> \endverbatim
diff --git a/SRC/cla_gbamv.f b/SRC/cla_gbamv.f
index 85f84ab2..3a22bca6 100644
--- a/SRC/cla_gbamv.f
+++ b/SRC/cla_gbamv.f
@@ -63,13 +63,11 @@
 *>          TRANS is INTEGER
 *>           On entry, TRANS specifies the operation to be performed as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>             BLAS_NO_TRANS      y := alpha*abs(A)*abs(x) + beta*abs(y)
 *>             BLAS_TRANS         y := alpha*abs(A**T)*abs(x) + beta*abs(y)
 *>             BLAS_CONJ_TRANS    y := alpha*abs(A**T)*abs(x) + beta*abs(y)
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -169,8 +167,7 @@
 *>           On entry, INCY specifies the increment for the elements of
 *>           Y. INCY must not be zero.
 *>           Unchanged on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Level 2 Blas routine.
 *> \endverbatim
 *>
diff --git a/SRC/cla_geamv.f b/SRC/cla_geamv.f
index 65ffac34..09eb5b05 100644
--- a/SRC/cla_geamv.f
+++ b/SRC/cla_geamv.f
@@ -64,13 +64,11 @@
 *>          TRANS is INTEGER
 *>           On entry, TRANS specifies the operation to be performed as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>             BLAS_NO_TRANS      y := alpha*abs(A)*abs(x) + beta*abs(y)
 *>             BLAS_TRANS         y := alpha*abs(A**T)*abs(x) + beta*abs(y)
 *>             BLAS_CONJ_TRANS    y := alpha*abs(A**T)*abs(x) + beta*abs(y)
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -158,8 +156,7 @@
 *>           On entry, INCY specifies the increment for the elements of
 *>           Y. INCY must not be zero.
 *>           Unchanged on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Level 2 Blas routine.
 *> \endverbatim
 *>
diff --git a/SRC/cla_heamv.f b/SRC/cla_heamv.f
index 359219f4..ea7acd0a 100644
--- a/SRC/cla_heamv.f
+++ b/SRC/cla_heamv.f
@@ -63,16 +63,13 @@
 *>           On entry, UPLO specifies whether the upper or lower
 *>           triangular part of the array A is to be referenced as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = BLAS_UPPER   Only the upper triangular part of A
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = BLAS_LOWER   Only the lower triangular part of A
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/cla_herfsx_extended.f b/SRC/cla_herfsx_extended.f
index 11e84fb1..91f58993 100644
--- a/SRC/cla_herfsx_extended.f
+++ b/SRC/cla_herfsx_extended.f
@@ -200,37 +200,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -239,8 +233,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     This subroutine is only responsible for setting the second field
 *>     above.
 *>     See Lapack Working Note 165 for further details and extra
@@ -254,14 +247,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -269,26 +260,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -299,8 +286,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     This subroutine is only responsible for setting the second field
 *>     above.
 *>     See Lapack Working Note 165 for further details and extra
diff --git a/SRC/cla_porfsx_extended.f b/SRC/cla_porfsx_extended.f
index 7deea24e..ac8428c6 100644
--- a/SRC/cla_porfsx_extended.f
+++ b/SRC/cla_porfsx_extended.f
@@ -192,37 +192,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -231,8 +225,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     This subroutine is only responsible for setting the second field
 *>     above.
 *>     See Lapack Working Note 165 for further details and extra
@@ -246,14 +239,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -261,26 +252,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -291,8 +278,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     This subroutine is only responsible for setting the second field
 *>     above.
 *>     See Lapack Working Note 165 for further details and extra
diff --git a/SRC/cla_syamv.f b/SRC/cla_syamv.f
index 77ffc50e..c738954a 100644
--- a/SRC/cla_syamv.f
+++ b/SRC/cla_syamv.f
@@ -64,16 +64,13 @@
 *>           On entry, UPLO specifies whether the upper or lower
 *>           triangular part of the array A is to be referenced as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = BLAS_UPPER   Only the upper triangular part of A
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = BLAS_LOWER   Only the lower triangular part of A
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/cla_syrfsx_extended.f b/SRC/cla_syrfsx_extended.f
index c8b07130..44a90b34 100644
--- a/SRC/cla_syrfsx_extended.f
+++ b/SRC/cla_syrfsx_extended.f
@@ -200,37 +200,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -239,8 +233,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     This subroutine is only responsible for setting the second field
 *>     above.
 *>     See Lapack Working Note 165 for further details and extra
@@ -254,14 +247,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -269,26 +260,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -299,8 +286,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     This subroutine is only responsible for setting the second field
 *>     above.
 *>     See Lapack Working Note 165 for further details and extra
diff --git a/SRC/clahef.f b/SRC/clahef.f
index 572b8323..34c4bcaf 100644
--- a/SRC/clahef.f
+++ b/SRC/clahef.f
@@ -113,8 +113,7 @@
 *>          Details of the interchanges and the block structure of D.
 *>          If UPLO = 'U', only the last KB elements of IPIV are set;
 *>          if UPLO = 'L', only the first KB elements are set.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If IPIV(k) > 0, then rows and columns k and IPIV(k) were
 *>          interchanged and D(k,k) is a 1-by-1 diagonal block.
 *>          If UPLO = 'U' and IPIV(k) = IPIV(k-1) < 0, then rows and
diff --git a/SRC/clahqr.f b/SRC/clahqr.f
index 4b09a708..8e32a189 100644
--- a/SRC/clahqr.f
+++ b/SRC/clahqr.f
@@ -147,22 +147,19 @@
 *>                  per eigenvalue; elements i+1:ihi of W contain
 *>                  those eigenvalues which have been successfully
 *>                  computed.
-*> \endverbatim
-*> \verbatim
+*>
 *>                  If INFO .GT. 0 and WANTT is .FALSE., then on exit,
 *>                  the remaining unconverged eigenvalues are the
 *>                  eigenvalues of the upper Hessenberg matrix
 *>                  rows and columns ILO thorugh INFO of the final,
 *>                  output value of H.
-*> \endverbatim
-*> \verbatim
+*>
 *>                  If INFO .GT. 0 and WANTT is .TRUE., then on exit
 *>          (*)       (initial value of H)*U  = U*(final value of H)
 *>                  where U is an orthognal matrix.    The final
 *>                  value of H is upper Hessenberg and triangular in
 *>                  rows and columns INFO+1 through IHI.
-*> \endverbatim
-*> \verbatim
+*>
 *>                  If INFO .GT. 0 and WANTZ is .TRUE., then on exit
 *>                      (final value of Z)  = (initial value of Z)*U
 *>                  where U is the orthogonal matrix in (*)
diff --git a/SRC/clals0.f b/SRC/clals0.f
index 0bc0b122..3bc79e3f 100644
--- a/SRC/clals0.f
+++ b/SRC/clals0.f
@@ -102,8 +102,7 @@
 *>          SQRE is INTEGER
 *>         = 0: the lower block is an NR-by-NR square matrix.
 *>         = 1: the lower block is an NR-by-(NR+1) rectangular matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>         The bidiagonal matrix has row dimension N = NL + NR + 1,
 *>         and column dimension M = N + SQRE.
 *> \endverbatim
diff --git a/SRC/clanhf.f b/SRC/clanhf.f
index 81816af1..29fcbd16 100644
--- a/SRC/clanhf.f
+++ b/SRC/clanhf.f
@@ -83,12 +83,10 @@
 *>          UPLO is CHARACTER
 *>            On entry, UPLO specifies whether the RFP matrix A came from
 *>            an upper or lower triangular matrix as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>            UPLO = 'U' or 'u' RFP A came from an upper triangular
 *>            matrix
-*> \endverbatim
-*> \verbatim
+*>
 *>            UPLO = 'L' or 'l' RFP A came from a  lower triangular
 *>            matrix
 *> \endverbatim
diff --git a/SRC/claqgb.f b/SRC/claqgb.f
index be826c38..642ff10a 100644
--- a/SRC/claqgb.f
+++ b/SRC/claqgb.f
@@ -77,8 +77,7 @@
 *>          The j-th column of A is stored in the j-th column of the
 *>          array AB as follows:
 *>          AB(ku+1+i-j,j) = A(i,j) for max(1,j-ku)<=i<=min(m,j+kl)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the equilibrated matrix, in the same storage format
 *>          as A.  See EQUED for the form of the equilibrated matrix.
 *> \endverbatim
@@ -130,18 +129,15 @@
 *>                  by diag(C).
 *>          = 'B':  Both row and column equilibration, i.e., A has been
 *>                  replaced by diag(R) * A * diag(C).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if row or column scaling
 *>  should be done based on the ratio of the row or column scaling
 *>  factors.  If ROWCND < THRESH, row scaling is done, and if
 *>  COLCND < THRESH, column scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if row scaling
 *>  should be done based on the absolute size of the largest matrix
 *>  element.  If AMAX > LARGE or AMAX < SMALL, row scaling is done.
diff --git a/SRC/claqge.f b/SRC/claqge.f
index 54003de7..902dec0d 100644
--- a/SRC/claqge.f
+++ b/SRC/claqge.f
@@ -112,18 +112,15 @@
 *>                  by diag(C).
 *>          = 'B':  Both row and column equilibration, i.e., A has been
 *>                  replaced by diag(R) * A * diag(C).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if row or column scaling
 *>  should be done based on the ratio of the row or column scaling
 *>  factors.  If ROWCND < THRESH, row scaling is done, and if
 *>  COLCND < THRESH, column scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if row scaling
 *>  should be done based on the absolute size of the largest matrix
 *>  element.  If AMAX > LARGE or AMAX < SMALL, row scaling is done.
diff --git a/SRC/claqhb.f b/SRC/claqhb.f
index fbc2d99d..562acf34 100644
--- a/SRC/claqhb.f
+++ b/SRC/claqhb.f
@@ -75,8 +75,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**H *U or A = L*L**H of the band
 *>          matrix A, in the same storage format as A.
@@ -113,17 +112,14 @@
 *>          = 'N':  No equilibration.
 *>          = 'Y':  Equilibration was done, i.e., A has been replaced by
 *>                  diag(S) * A * diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if scaling should be done
 *>  based on the ratio of the scaling factors.  If SCOND < THRESH,
 *>  scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if scaling should
 *>  be done based on the absolute size of the largest matrix element.
 *>  If AMAX > LARGE or AMAX < SMALL, scaling is done.
diff --git a/SRC/claqhe.f b/SRC/claqhe.f
index 5019df8e..fceb3f74 100644
--- a/SRC/claqhe.f
+++ b/SRC/claqhe.f
@@ -69,8 +69,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if EQUED = 'Y', the equilibrated matrix:
 *>          diag(S) * A * diag(S).
 *> \endverbatim
@@ -106,17 +105,14 @@
 *>          = 'N':  No equilibration.
 *>          = 'Y':  Equilibration was done, i.e., A has been replaced by
 *>                  diag(S) * A * diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if scaling should be done
 *>  based on the ratio of the scaling factors.  If SCOND < THRESH,
 *>  scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if scaling should
 *>  be done based on the absolute size of the largest matrix element.
 *>  If AMAX > LARGE or AMAX < SMALL, scaling is done.
diff --git a/SRC/claqhp.f b/SRC/claqhp.f
index a2e9cd0d..c4d8f0aa 100644
--- a/SRC/claqhp.f
+++ b/SRC/claqhp.f
@@ -67,8 +67,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the equilibrated matrix:  diag(S) * A * diag(S), in
 *>          the same storage format as A.
 *> \endverbatim
@@ -98,17 +97,14 @@
 *>          = 'N':  No equilibration.
 *>          = 'Y':  Equilibration was done, i.e., A has been replaced by
 *>                  diag(S) * A * diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if scaling should be done
 *>  based on the ratio of the scaling factors.  If SCOND < THRESH,
 *>  scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if scaling should
 *>  be done based on the absolute size of the largest matrix element.
 *>  If AMAX > LARGE or AMAX < SMALL, scaling is done.
diff --git a/SRC/claqr0.f b/SRC/claqr0.f
index 31730620..be73778e 100644
--- a/SRC/claqr0.f
+++ b/SRC/claqr0.f
@@ -98,8 +98,7 @@
 *>           .FALSE., then the contents of H are unspecified on exit.
 *>           (The output value of H when INFO.GT.0 is given under the
 *>           description of INFO below.)
-*> \endverbatim
-*> \verbatim
+*>
 *>           This subroutine may explicitly set H(i,j) = 0 for i.GT.j and
 *>           j = 1, 2, ... ILO-1 or j = IHI+1, IHI+2, ... N.
 *> \endverbatim
@@ -164,8 +163,7 @@
 *>           is sufficient, but LWORK typically as large as 6*N may
 *>           be required for optimal performance.  A workspace query
 *>           to determine the optimal workspace size is recommended.
-*> \endverbatim
-*> \verbatim
+*>
 *>           If LWORK = -1, then CLAQR0 does a workspace query.
 *>           In this case, CLAQR0 checks the input parameters and
 *>           estimates the optimal workspace size for the given
diff --git a/SRC/claqr1.f b/SRC/claqr1.f
index 7c46c6f1..8b468a65 100644
--- a/SRC/claqr1.f
+++ b/SRC/claqr1.f
@@ -76,8 +76,7 @@
 *> \param[in] S2
 *> \verbatim
 *>          S2 is COMPLEX
-*> \endverbatim
-*> \verbatim
+*>
 *>          S1 and S2 are the shifts defining K in (*) above.
 *> \endverbatim
 *>
diff --git a/SRC/claqr2.f b/SRC/claqr2.f
index a0ee3890..5fc0555c 100644
--- a/SRC/claqr2.f
+++ b/SRC/claqr2.f
@@ -239,8 +239,7 @@
 *>          The dimension of the work array WORK.  LWORK = 2*NW
 *>          suffices, but greater efficiency may result from larger
 *>          values of LWORK.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; CLAQR2
 *>          only estimates the optimal workspace size for the given
 *>          values of N, NW, KTOP and KBOT.  The estimate is returned
diff --git a/SRC/claqr3.f b/SRC/claqr3.f
index 49badd30..185be7ed 100644
--- a/SRC/claqr3.f
+++ b/SRC/claqr3.f
@@ -236,8 +236,7 @@
 *>          The dimension of the work array WORK.  LWORK = 2*NW
 *>          suffices, but greater efficiency may result from larger
 *>          values of LWORK.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; CLAQR3
 *>          only estimates the optimal workspace size for the given
 *>          values of N, NW, KTOP and KBOT.  The estimate is returned
diff --git a/SRC/claqr4.f b/SRC/claqr4.f
index 59104227..d7a96d8f 100644
--- a/SRC/claqr4.f
+++ b/SRC/claqr4.f
@@ -106,8 +106,7 @@
 *>           .FALSE., then the contents of H are unspecified on exit.
 *>           (The output value of H when INFO.GT.0 is given under the
 *>           description of INFO below.)
-*> \endverbatim
-*> \verbatim
+*>
 *>           This subroutine may explicitly set H(i,j) = 0 for i.GT.j and
 *>           j = 1, 2, ... ILO-1 or j = IHI+1, IHI+2, ... N.
 *> \endverbatim
@@ -172,8 +171,7 @@
 *>           is sufficient, but LWORK typically as large as 6*N may
 *>           be required for optimal performance.  A workspace query
 *>           to determine the optimal workspace size is recommended.
-*> \endverbatim
-*> \verbatim
+*>
 *>           If LWORK = -1, then CLAQR4 does a workspace query.
 *>           In this case, CLAQR4 checks the input parameters and
 *>           estimates the optimal workspace size for the given
diff --git a/SRC/claqsb.f b/SRC/claqsb.f
index 30fa6b08..22d92c12 100644
--- a/SRC/claqsb.f
+++ b/SRC/claqsb.f
@@ -75,8 +75,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**H *U or A = L*L**H of the band
 *>          matrix A, in the same storage format as A.
@@ -113,17 +112,14 @@
 *>          = 'N':  No equilibration.
 *>          = 'Y':  Equilibration was done, i.e., A has been replaced by
 *>                  diag(S) * A * diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if scaling should be done
 *>  based on the ratio of the scaling factors.  If SCOND < THRESH,
 *>  scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if scaling should
 *>  be done based on the absolute size of the largest matrix element.
 *>  If AMAX > LARGE or AMAX < SMALL, scaling is done.
diff --git a/SRC/claqsp.f b/SRC/claqsp.f
index 91f88693..31f68d14 100644
--- a/SRC/claqsp.f
+++ b/SRC/claqsp.f
@@ -67,8 +67,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the equilibrated matrix:  diag(S) * A * diag(S), in
 *>          the same storage format as A.
 *> \endverbatim
@@ -98,17 +97,14 @@
 *>          = 'N':  No equilibration.
 *>          = 'Y':  Equilibration was done, i.e., A has been replaced by
 *>                  diag(S) * A * diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if scaling should be done
 *>  based on the ratio of the scaling factors.  If SCOND < THRESH,
 *>  scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if scaling should
 *>  be done based on the absolute size of the largest matrix element.
 *>  If AMAX > LARGE or AMAX < SMALL, scaling is done.
diff --git a/SRC/claqsy.f b/SRC/claqsy.f
index 12ec9145..80b46968 100644
--- a/SRC/claqsy.f
+++ b/SRC/claqsy.f
@@ -69,8 +69,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if EQUED = 'Y', the equilibrated matrix:
 *>          diag(S) * A * diag(S).
 *> \endverbatim
@@ -106,17 +105,14 @@
 *>          = 'N':  No equilibration.
 *>          = 'Y':  Equilibration was done, i.e., A has been replaced by
 *>                  diag(S) * A * diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if scaling should be done
 *>  based on the ratio of the scaling factors.  If SCOND < THRESH,
 *>  scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if scaling should
 *>  be done based on the absolute size of the largest matrix element.
 *>  If AMAX > LARGE or AMAX < SMALL, scaling is done.
diff --git a/SRC/clascl.f b/SRC/clascl.f
index 07d29967..d5039e97 100644
--- a/SRC/clascl.f
+++ b/SRC/clascl.f
@@ -86,8 +86,7 @@
 *> \param[in] CTO
 *> \verbatim
 *>          CTO is REAL
-*> \endverbatim
-*> \verbatim
+*>
 *>          The matrix A is multiplied by CTO/CFROM. A(I,J) is computed
 *>          without over/underflow if the final result CTO*A(I,J)/CFROM
 *>          can be represented without over/underflow.  CFROM must be
diff --git a/SRC/clasyf.f b/SRC/clasyf.f
index 23f8d7a2..0d9a4dff 100644
--- a/SRC/clasyf.f
+++ b/SRC/clasyf.f
@@ -113,8 +113,7 @@
 *>          Details of the interchanges and the block structure of D.
 *>          If UPLO = 'U', only the last KB elements of IPIV are set;
 *>          if UPLO = 'L', only the first KB elements are set.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If IPIV(k) > 0, then rows and columns k and IPIV(k) were
 *>          interchanged and D(k,k) is a 1-by-1 diagonal block.
 *>          If UPLO = 'U' and IPIV(k) = IPIV(k-1) < 0, then rows and
diff --git a/SRC/clatbs.f b/SRC/clatbs.f
index 9c94705c..1e7dcce5 100644
--- a/SRC/clatbs.f
+++ b/SRC/clatbs.f
@@ -137,15 +137,13 @@
 *> \param[in,out] CNORM
 *> \verbatim
 *>          CNORM is or output) REAL array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'Y', CNORM is an input argument and CNORM(j)
 *>          contains the norm of the off-diagonal part of the j-th column
 *>          of A.  If TRANS = 'N', CNORM(j) must be greater than or equal
 *>          to the infinity-norm, and if TRANS = 'T' or 'C', CNORM(j)
 *>          must be greater than or equal to the 1-norm.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'N', CNORM is an output argument and CNORM(j)
 *>          returns the 1-norm of the offdiagonal part of the j-th column
 *>          of A.
diff --git a/SRC/clatps.f b/SRC/clatps.f
index 6aad2fe9..5c36c905 100644
--- a/SRC/clatps.f
+++ b/SRC/clatps.f
@@ -125,15 +125,13 @@
 *> \param[in,out] CNORM
 *> \verbatim
 *>          CNORM is or output) REAL array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'Y', CNORM is an input argument and CNORM(j)
 *>          contains the norm of the off-diagonal part of the j-th column
 *>          of A.  If TRANS = 'N', CNORM(j) must be greater than or equal
 *>          to the infinity-norm, and if TRANS = 'T' or 'C', CNORM(j)
 *>          must be greater than or equal to the 1-norm.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'N', CNORM is an output argument and CNORM(j)
 *>          returns the 1-norm of the offdiagonal part of the j-th column
 *>          of A.
diff --git a/SRC/clatrs.f b/SRC/clatrs.f
index 7098388d..675550f9 100644
--- a/SRC/clatrs.f
+++ b/SRC/clatrs.f
@@ -133,15 +133,13 @@
 *> \param[in,out] CNORM
 *> \verbatim
 *>          CNORM is or output) REAL array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'Y', CNORM is an input argument and CNORM(j)
 *>          contains the norm of the off-diagonal part of the j-th column
 *>          of A.  If TRANS = 'N', CNORM(j) must be greater than or equal
 *>          to the infinity-norm, and if TRANS = 'T' or 'C', CNORM(j)
 *>          must be greater than or equal to the 1-norm.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'N', CNORM is an output argument and CNORM(j)
 *>          returns the 1-norm of the offdiagonal part of the j-th column
 *>          of A.
diff --git a/SRC/clatzm.f b/SRC/clatzm.f
index e0a15e89..d0915dfb 100644
--- a/SRC/clatzm.f
+++ b/SRC/clatzm.f
@@ -107,8 +107,7 @@
 *>                         (M,1)   if SIDE = 'R'
 *>          On entry, the n-vector C1 if SIDE = 'L', or the m-vector C1
 *>          if SIDE = 'R'.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the first row of P*C if SIDE = 'L', or the first
 *>          column of C*P if SIDE = 'R'.
 *> \endverbatim
@@ -120,8 +119,7 @@
 *>                         (LDC, N-1) if SIDE = 'R'
 *>          On entry, the (m - 1) x n matrix C2 if SIDE = 'L', or the
 *>          m x (n - 1) matrix C2 if SIDE = 'R'.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, rows 2:m of P*C if SIDE = 'L', or columns 2:m of C*P
 *>          if SIDE = 'R'.
 *> \endverbatim
diff --git a/SRC/cpbrfs.f b/SRC/cpbrfs.f
index d5018944..9ac4f70a 100644
--- a/SRC/cpbrfs.f
+++ b/SRC/cpbrfs.f
@@ -165,12 +165,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/cpbsv.f b/SRC/cpbsv.f
index b100b97c..627cfc36 100644
--- a/SRC/cpbsv.f
+++ b/SRC/cpbsv.f
@@ -90,8 +90,7 @@
 *>          if UPLO = 'U', AB(KD+1+i-j,j) = A(i,j) for max(1,j-KD)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(N,j+KD).
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**H*U or A = L*L**H of the band
 *>          matrix A, in the same storage format as A.
diff --git a/SRC/cpbsvx.f b/SRC/cpbsvx.f
index 9af0418b..05c99895 100644
--- a/SRC/cpbsvx.f
+++ b/SRC/cpbsvx.f
@@ -145,8 +145,7 @@
 *>          if UPLO = 'U', AB(KD+1+i-j,j) = A(i,j) for max(1,j-KD)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(N,j+KD).
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>          diag(S)*A*diag(S).
 *> \endverbatim
@@ -165,13 +164,11 @@
 *>          factorization A = U**H*U or A = L*L**H of the band matrix
 *>          A, in the same storage format as A (see AB).  If EQUED = 'Y',
 *>          then AFB is the factored form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AFB is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**H*U or A = L*L**H.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then AFB is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**H*U or A = L*L**H of the equilibrated
diff --git a/SRC/cpbtf2.f b/SRC/cpbtf2.f
index b5d63c2b..883edfaa 100644
--- a/SRC/cpbtf2.f
+++ b/SRC/cpbtf2.f
@@ -81,8 +81,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**H *U or A = L*L**H of the band
 *>          matrix A, in the same storage format as A.
diff --git a/SRC/cpbtrf.f b/SRC/cpbtrf.f
index 7a5abaec..8a11a5c7 100644
--- a/SRC/cpbtrf.f
+++ b/SRC/cpbtrf.f
@@ -76,8 +76,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**H*U or A = L*L**H of the band
 *>          matrix A, in the same storage format as A.
diff --git a/SRC/cpftrf.f b/SRC/cpftrf.f
index 996d9e51..647d596c 100644
--- a/SRC/cpftrf.f
+++ b/SRC/cpftrf.f
@@ -82,8 +82,7 @@
 *>          of lower packed A. The LDA of RFP A is (N+1)/2 when TRANSR =
 *>          'C'. When TRANSR is 'N' the LDA is N+1 when N is even and N
 *>          is odd. See the Note below for more details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization RFP A = U**H*U or RFP A = L*L**H.
 *> \endverbatim
@@ -96,27 +95,22 @@
 *>          > 0:  if INFO = i, the leading minor of order i is not
 *>                positive definite, and the factorization could not be
 *>                completed.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Further Notes on RFP Format:
 *>  ============================
-*> \endverbatim
-*> \verbatim
+*>
 *>  We first consider Standard Packed Format when N is even.
 *>  We give an example where N = 6.
-*> \endverbatim
-*> \verbatim
+*>
 *>     AP is Upper             AP is Lower
-*> \endverbatim
-*> \verbatim
+*>
 *>   00 01 02 03 04 05       00
 *>      11 12 13 14 15       10 11
 *>         22 23 24 25       20 21 22
 *>            33 34 35       30 31 32 33
 *>               44 45       40 41 42 43 44
 *>                  55       50 51 52 53 54 55
-*> \endverbatim
-*> \verbatim
+*>
 *>  Let TRANSR = 'N'. RFP holds AP as follows:
 *>  For UPLO = 'U' the upper trapezoid A(0:5,0:2) consists of the last
 *>  three columns of AP upper. The lower triangle A(4:6,0:2) consists of
@@ -126,19 +120,16 @@
 *>  conjugate-transpose of the last three columns of AP lower.
 *>  To denote conjugate we place -- above the element. This covers the
 *>  case N even and TRANSR = 'N'.
-*> \endverbatim
-*> \verbatim
+*>
 *>         RFP A                   RFP A
-*> \endverbatim
-*> \verbatim
+*>
 *>                                -- -- --
 *>        03 04 05                33 43 53
 *>                                   -- --
 *>        13 14 15                00 44 54
 *>                                      --
 *>        23 24 25                10 11 55
-*> \endverbatim
-*> \verbatim
+*>
 *>        33 34 35                20 21 22
 *>        --
 *>        00 44 45                30 31 32
@@ -146,37 +137,30 @@
 *>        01 11 55                40 41 42
 *>        -- -- --
 *>        02 12 22                50 51 52
-*> \endverbatim
-*> \verbatim
+*>
 *>  Now let TRANSR = 'C'. RFP A in both UPLO cases is just the conjugate-
 *>  transpose of RFP A above. One therefore gets:
-*> \endverbatim
-*> \verbatim
+*>
 *>           RFP A                   RFP A
-*> \endverbatim
-*> \verbatim
+*>
 *>     -- -- -- --                -- -- -- -- -- --
 *>     03 13 23 33 00 01 02    33 00 10 20 30 40 50
 *>     -- -- -- -- --                -- -- -- -- --
 *>     04 14 24 34 44 11 12    43 44 11 21 31 41 51
 *>     -- -- -- -- -- --                -- -- -- --
 *>     05 15 25 35 45 55 22    53 54 55 22 32 42 52
-*> \endverbatim
-*> \verbatim
+*>
 *>  We next  consider Standard Packed Format when N is odd.
 *>  We give an example where N = 5.
-*> \endverbatim
-*> \verbatim
+*>
 *>     AP is Upper                 AP is Lower
-*> \endverbatim
-*> \verbatim
+*>
 *>   00 01 02 03 04              00
 *>      11 12 13 14              10 11
 *>         22 23 24              20 21 22
 *>            33 34              30 31 32 33
 *>               44              40 41 42 43 44
-*> \endverbatim
-*> \verbatim
+*>
 *>  Let TRANSR = 'N'. RFP holds AP as follows:
 *>  For UPLO = 'U' the upper trapezoid A(0:4,0:2) consists of the last
 *>  three columns of AP upper. The lower triangle A(3:4,0:1) consists of
@@ -186,31 +170,25 @@
 *>  conjugate-transpose of the last two   columns of AP lower.
 *>  To denote conjugate we place -- above the element. This covers the
 *>  case N odd  and TRANSR = 'N'.
-*> \endverbatim
-*> \verbatim
+*>
 *>         RFP A                   RFP A
-*> \endverbatim
-*> \verbatim
+*>
 *>                                   -- --
 *>        02 03 04                00 33 43
 *>                                      --
 *>        12 13 14                10 11 44
-*> \endverbatim
-*> \verbatim
+*>
 *>        22 23 24                20 21 22
 *>        --
 *>        00 33 34                30 31 32
 *>        -- --
 *>        01 11 44                40 41 42
-*> \endverbatim
-*> \verbatim
+*>
 *>  Now let TRANSR = 'C'. RFP A in both UPLO cases is just the conjugate-
 *>  transpose of RFP A above. One therefore gets:
-*> \endverbatim
-*> \verbatim
+*>
 *>           RFP A                   RFP A
-*> \endverbatim
-*> \verbatim
+*>
 *>     -- -- --                   -- -- -- -- -- --
 *>     02 12 22 00 01             00 10 20 30 40 50
 *>     -- -- -- --                   -- -- -- -- --
diff --git a/SRC/cpftri.f b/SRC/cpftri.f
index 3b172c13..5937effc 100644
--- a/SRC/cpftri.f
+++ b/SRC/cpftri.f
@@ -76,8 +76,7 @@
 *>          of lower packed A. The LDA of RFP A is (N+1)/2 when TRANSR =
 *>          'C'. When TRANSR is 'N' the LDA is N+1 when N is even and N
 *>          is odd. See the Note below for more details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the Hermitian inverse of the original matrix, in the
 *>          same storage format.
 *> \endverbatim
diff --git a/SRC/cporfs.f b/SRC/cporfs.f
index 7eedf5c4..580ab10f 100644
--- a/SRC/cporfs.f
+++ b/SRC/cporfs.f
@@ -159,12 +159,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/cporfsx.f b/SRC/cporfsx.f
index c95aba5d..2ba6cefa 100644
--- a/SRC/cporfsx.f
+++ b/SRC/cporfsx.f
@@ -210,37 +210,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -249,8 +243,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -261,14 +254,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -276,26 +267,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -306,8 +293,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -326,8 +312,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0
@@ -338,8 +323,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -349,8 +333,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/cposv.f b/SRC/cposv.f
index 58652b1e..4b73f6c0 100644
--- a/SRC/cposv.f
+++ b/SRC/cposv.f
@@ -82,8 +82,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization A = U**H*U or A = L*L**H.
 *> \endverbatim
diff --git a/SRC/cposvx.f b/SRC/cposvx.f
index e5477d88..ef3d591d 100644
--- a/SRC/cposvx.f
+++ b/SRC/cposvx.f
@@ -140,8 +140,7 @@
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.  A is not modified if
 *>          FACT = 'F' or 'N', or if FACT = 'E' and EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>          diag(S)*A*diag(S).
 *> \endverbatim
@@ -160,14 +159,12 @@
 *>          factorization A = U**H*U or A = L*L**H, in the same storage
 *>          format as A.  If EQUED .ne. 'N', then AF is the factored form
 *>          of the equilibrated matrix diag(S)*A*diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AF is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**H*U or A = L*L**H of the original
 *>          matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then AF is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**H*U or A = L*L**H of the equilibrated
diff --git a/SRC/cposvxx.f b/SRC/cposvxx.f
index 7550ea5f..fb1c6df8 100644
--- a/SRC/cposvxx.f
+++ b/SRC/cposvxx.f
@@ -167,8 +167,7 @@
 *>     the strictly upper triangular part of A is not referenced.  A is
 *>     not modified if FACT = 'F' or 'N', or if FACT = 'E' and EQUED =
 *>     'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>     On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>     diag(S)*A*diag(S).
 *> \endverbatim
@@ -187,14 +186,12 @@
 *>     factorization A = U**T*U or A = L*L**T, in the same storage
 *>     format as A.  If EQUED .ne. 'N', then AF is the factored
 *>     form of the equilibrated matrix diag(S)*A*diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then AF is an output argument and on exit
 *>     returns the triangular factor U or L from the Cholesky
 *>     factorization A = U**T*U or A = L*L**T of the original
 *>     matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'E', then AF is an output argument and on exit
 *>     returns the triangular factor U or L from the Cholesky
 *>     factorization A = U**T*U or A = L*L**T of the equilibrated
@@ -313,37 +310,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -352,8 +343,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -364,14 +354,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -379,26 +367,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -409,8 +393,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -429,8 +412,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0
@@ -441,8 +423,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -452,8 +433,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/cpotf2.f b/SRC/cpotf2.f
index a2eed9d8..a9f5efef 100644
--- a/SRC/cpotf2.f
+++ b/SRC/cpotf2.f
@@ -74,8 +74,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization A = U**H *U  or A = L*L**H.
 *> \endverbatim
diff --git a/SRC/cpotrf.f b/SRC/cpotrf.f
index 21189b2e..0d20bb0d 100644
--- a/SRC/cpotrf.f
+++ b/SRC/cpotrf.f
@@ -72,8 +72,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization A = U**H*U or A = L*L**H.
 *> \endverbatim
diff --git a/SRC/cpprfs.f b/SRC/cpprfs.f
index a6db5697..a5c26651 100644
--- a/SRC/cpprfs.f
+++ b/SRC/cpprfs.f
@@ -147,12 +147,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/cppsv.f b/SRC/cppsv.f
index cde5f27f..5aa477de 100644
--- a/SRC/cppsv.f
+++ b/SRC/cppsv.f
@@ -81,8 +81,7 @@
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization A = U**H*U or A = L*L**H, in the same storage
 *>          format as A.
diff --git a/SRC/cppsvx.f b/SRC/cppsvx.f
index fce18b70..b9051a6e 100644
--- a/SRC/cppsvx.f
+++ b/SRC/cppsvx.f
@@ -138,8 +138,7 @@
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
 *>          See below for further details.  A is not modified if
 *>          FACT = 'F' or 'N', or if FACT = 'E' and EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>          diag(S)*A*diag(S).
 *> \endverbatim
@@ -152,14 +151,12 @@
 *>          factorization A = U**H*U or A = L*L**H, in the same storage
 *>          format as A.  If EQUED .ne. 'N', then AFP is the factored
 *>          form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AFP is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**H * U or A = L * L**H of the original
 *>          matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then AFP is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**H*U or A = L*L**H of the equilibrated
diff --git a/SRC/cpptrf.f b/SRC/cpptrf.f
index 933d1e89..e99f0339 100644
--- a/SRC/cpptrf.f
+++ b/SRC/cpptrf.f
@@ -69,8 +69,7 @@
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**H*U or A = L*L**H, in the same
 *>          storage format as A.
diff --git a/SRC/cpptri.f b/SRC/cpptri.f
index 70e0840b..dcfbae5e 100644
--- a/SRC/cpptri.f
+++ b/SRC/cpptri.f
@@ -65,8 +65,7 @@
 *>          array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = U(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = L(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the upper or lower triangle of the (Hermitian)
 *>          inverse of A, overwriting the input factor U or L.
 *> \endverbatim
diff --git a/SRC/cpstf2.f b/SRC/cpstf2.f
index 7cea72c6..4ec6ea79 100644
--- a/SRC/cpstf2.f
+++ b/SRC/cpstf2.f
@@ -79,8 +79,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization as above.
 *> \endverbatim
diff --git a/SRC/cpstrf.f b/SRC/cpstrf.f
index 8da607ec..976fd5e9 100644
--- a/SRC/cpstrf.f
+++ b/SRC/cpstrf.f
@@ -79,8 +79,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization as above.
 *> \endverbatim
diff --git a/SRC/cptrfs.f b/SRC/cptrfs.f
index abd4fa53..ce1277db 100644
--- a/SRC/cptrfs.f
+++ b/SRC/cptrfs.f
@@ -159,12 +159,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/cspmv.f b/SRC/cspmv.f
index bc5a9cc6..7d3539a2 100644
--- a/SRC/cspmv.f
+++ b/SRC/cspmv.f
@@ -53,16 +53,13 @@
 *>           On entry, UPLO specifies whether the upper or lower
 *>           triangular part of the matrix A is supplied in the packed
 *>           array AP as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'U' or 'u'   The upper triangular part of A is
 *>                                  supplied in AP.
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'L' or 'l'   The lower triangular part of A is
 *>                                  supplied in AP.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/cspr.f b/SRC/cspr.f
index 124174b0..eb282570 100644
--- a/SRC/cspr.f
+++ b/SRC/cspr.f
@@ -53,16 +53,13 @@
 *>           On entry, UPLO specifies whether the upper or lower
 *>           triangular part of the matrix A is supplied in the packed
 *>           array AP as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'U' or 'u'   The upper triangular part of A is
 *>                                  supplied in AP.
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'L' or 'l'   The lower triangular part of A is
 *>                                  supplied in AP.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/csprfs.f b/SRC/csprfs.f
index d75faf7d..b84474d2 100644
--- a/SRC/csprfs.f
+++ b/SRC/csprfs.f
@@ -156,12 +156,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/cspsv.f b/SRC/cspsv.f
index 06c168f7..ebcde720 100644
--- a/SRC/cspsv.f
+++ b/SRC/cspsv.f
@@ -83,8 +83,7 @@
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
 *>          A = U*D*U**T or A = L*D*L**T as computed by CSPTRF, stored as
diff --git a/SRC/cspsvx.f b/SRC/cspsvx.f
index 3728234b..2c7be9d5 100644
--- a/SRC/cspsvx.f
+++ b/SRC/cspsvx.f
@@ -128,8 +128,7 @@
 *>          to obtain the factor U or L from the factorization
 *>          A = U*D*U**T or A = L*D*L**T as computed by CSPTRF, stored as
 *>          a packed triangular matrix in the same storage format as A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AFP is an output argument and on exit
 *>          contains the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
@@ -150,8 +149,7 @@
 *>          is a 2-by-2 diagonal block.  If UPLO = 'L' and IPIV(k) =
 *>          IPIV(k+1) < 0, then rows and columns k+1 and -IPIV(k) were
 *>          interchanged and D(k:k+1,k:k+1) is a 2-by-2 diagonal block.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains details of the interchanges and the block structure
 *>          of D, as determined by CSPTRF.
diff --git a/SRC/csptrf.f b/SRC/csptrf.f
index d5d8af77..2842bd4a 100644
--- a/SRC/csptrf.f
+++ b/SRC/csptrf.f
@@ -71,8 +71,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L, stored as a packed triangular
 *>          matrix overwriting A (see below for further details).
diff --git a/SRC/csptri.f b/SRC/csptri.f
index 2e9c0ab9..0981022b 100644
--- a/SRC/csptri.f
+++ b/SRC/csptri.f
@@ -65,8 +65,7 @@
 *>          On entry, the block diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by CSPTRF,
 *>          stored as a packed triangular matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix, stored as a packed triangular matrix. The j-th column
 *>          of inv(A) is stored in the array AP as follows:
diff --git a/SRC/cstedc.f b/SRC/cstedc.f
index 9b7aad51..54c5d1d7 100644
--- a/SRC/cstedc.f
+++ b/SRC/cstedc.f
@@ -119,8 +119,7 @@
 *>          Note that for COMPZ = 'V', then if N is less than or
 *>          equal to the minimum divide size, usually 25, then LWORK need
 *>          only be 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK, RWORK and
 *>          IWORK arrays, returns these values as the first entries of
@@ -148,8 +147,7 @@
 *>          Note that for COMPZ = 'I' or 'V', then if N is less than or
 *>          equal to the minimum divide size, usually 25, then LRWORK
 *>          need only be max(1,2*(N-1)).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LRWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
@@ -175,8 +173,7 @@
 *>          Note that for COMPZ = 'I' or 'V', then if N is less than or
 *>          equal to the minimum divide size, usually 25, then LIWORK
 *>          need only be 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
diff --git a/SRC/cstegr.f b/SRC/cstegr.f
index fb085184..1222baf1 100644
--- a/SRC/cstegr.f
+++ b/SRC/cstegr.f
@@ -111,8 +111,7 @@
 *> \param[in] VU
 *> \verbatim
 *>          VU is REAL
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='V', the lower and upper bounds of the interval to
 *>          be searched for eigenvalues. VL < VU.
 *>          Not referenced if RANGE = 'A' or 'I'.
@@ -126,8 +125,7 @@
 *> \param[in] IU
 *> \verbatim
 *>          IU is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='I', the indices (in ascending order) of the
 *>          smallest and largest eigenvalues to be returned.
 *>          1 <= IL <= IU <= N, if N > 0.
diff --git a/SRC/cstein.f b/SRC/cstein.f
index 2a232d76..691a764a 100644
--- a/SRC/cstein.f
+++ b/SRC/cstein.f
@@ -153,16 +153,13 @@
 *>          > 0: if INFO = i, then i eigenvectors failed to converge
 *>               in MAXITS iterations.  Their indices are stored in
 *>               array IFAIL.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  MAXITS  INTEGER, default = 5
 *>          The maximum number of iterations performed.
-*> \endverbatim
-*> \verbatim
+*>
 *>  EXTRA   INTEGER, default = 2
 *>          The number of iterations performed after norm growth
 *>          criterion is satisfied, should be at least 1.
diff --git a/SRC/cstemr.f b/SRC/cstemr.f
index 1c6f8074..4e605e77 100644
--- a/SRC/cstemr.f
+++ b/SRC/cstemr.f
@@ -159,8 +159,7 @@
 *> \param[in] VU
 *> \verbatim
 *>          VU is REAL
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='V', the lower and upper bounds of the interval to
 *>          be searched for eigenvalues. VL < VU.
 *>          Not referenced if RANGE = 'A' or 'I'.
@@ -174,8 +173,7 @@
 *> \param[in] IU
 *> \verbatim
 *>          IU is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='I', the indices (in ascending order) of the
 *>          smallest and largest eigenvalues to be returned.
 *>          1 <= IL <= IU <= N, if N > 0.
diff --git a/SRC/csymv.f b/SRC/csymv.f
index 5cc548db..ee7caedf 100644
--- a/SRC/csymv.f
+++ b/SRC/csymv.f
@@ -53,16 +53,13 @@
 *>           On entry, UPLO specifies whether the upper or lower
 *>           triangular part of the array A is to be referenced as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'U' or 'u'   Only the upper triangular part of A
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'L' or 'l'   Only the lower triangular part of A
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/csyr.f b/SRC/csyr.f
index 4d265f91..b6b1ed75 100644
--- a/SRC/csyr.f
+++ b/SRC/csyr.f
@@ -53,16 +53,13 @@
 *>           On entry, UPLO specifies whether the upper or lower
 *>           triangular part of the array A is to be referenced as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'U' or 'u'   Only the upper triangular part of A
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'L' or 'l'   Only the lower triangular part of A
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/csyrfs.f b/SRC/csyrfs.f
index 03a5539d..bd070ffb 100644
--- a/SRC/csyrfs.f
+++ b/SRC/csyrfs.f
@@ -168,12 +168,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/csyrfsx.f b/SRC/csyrfsx.f
index 97465907..216d39f2 100644
--- a/SRC/csyrfsx.f
+++ b/SRC/csyrfsx.f
@@ -219,37 +219,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -258,8 +252,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -270,14 +263,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -285,26 +276,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -315,8 +302,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -335,8 +321,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0
@@ -347,8 +332,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -358,8 +342,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/csysv.f b/SRC/csysv.f
index 8e28a936..f9233e40 100644
--- a/SRC/csysv.f
+++ b/SRC/csysv.f
@@ -85,8 +85,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the block diagonal matrix D and the
 *>          multipliers used to obtain the factor U or L from the
 *>          factorization A = U*D*U**T or A = L*D*L**T as computed by
@@ -140,8 +139,7 @@
 *>          CSYTRF.
 *>          for LWORK < N, TRS will be done with Level BLAS 2
 *>          for LWORK >= N, TRS will be done with Level BLAS 3
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/csysvx.f b/SRC/csysvx.f
index cfbc5689..34b8688d 100644
--- a/SRC/csysvx.f
+++ b/SRC/csysvx.f
@@ -136,8 +136,7 @@
 *>          contains the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
 *>          A = U*D*U**T or A = L*D*L**T as computed by CSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AF is an output argument and on exit
 *>          returns the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
@@ -163,8 +162,7 @@
 *>          is a 2-by-2 diagonal block.  If UPLO = 'L' and IPIV(k) =
 *>          IPIV(k+1) < 0, then rows and columns k+1 and -IPIV(k) were
 *>          interchanged and D(k:k+1,k:k+1) is a 2-by-2 diagonal block.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains details of the interchanges and the block structure
 *>          of D, as determined by CSYTRF.
@@ -237,8 +235,7 @@
 *>          The length of WORK.  LWORK >= max(1,2*N), and for best
 *>          performance, when FACT = 'N', LWORK >= max(1,2*N,N*NB), where
 *>          NB is the optimal blocksize for CSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/csysvxx.f b/SRC/csysvxx.f
index 6d68bea5..652e08b8 100644
--- a/SRC/csysvxx.f
+++ b/SRC/csysvxx.f
@@ -168,8 +168,7 @@
 *>     N-by-N lower triangular part of A contains the lower
 *>     triangular part of the matrix A, and the strictly upper
 *>     triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>     On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>     diag(S)*A*diag(S).
 *> \endverbatim
@@ -187,8 +186,7 @@
 *>     contains the block diagonal matrix D and the multipliers
 *>     used to obtain the factor U or L from the factorization A =
 *>     U*D*U**T or A = L*D*L**T as computed by SSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then AF is an output argument and on exit
 *>     returns the block diagonal matrix D and the multipliers
 *>     used to obtain the factor U or L from the factorization A =
@@ -214,8 +212,7 @@
 *>     diagonal block.  If UPLO = 'L' and IPIV(k) = IPIV(k+1) < 0,
 *>     then rows and columns k+1 and -IPIV(k) were interchanged
 *>     and D(k:k+1,k:k+1) is a 2-by-2 diagonal block.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then IPIV is an output argument and on exit
 *>     contains details of the interchanges and the block
 *>     structure of D, as determined by SSYTRF.
@@ -326,37 +323,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -365,8 +356,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -377,14 +367,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -392,26 +380,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -422,8 +406,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -442,8 +425,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0
@@ -454,8 +436,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -465,8 +446,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/csyswapr.f b/SRC/csyswapr.f
index e072e20f..ec98d5d2 100644
--- a/SRC/csyswapr.f
+++ b/SRC/csyswapr.f
@@ -61,8 +61,7 @@
 *>          A is COMPLEX array, dimension (LDA,N)
 *>          On entry, the NB diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by CSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/csytf2.f b/SRC/csytf2.f
index 2c58a02e..11c0d219 100644
--- a/SRC/csytf2.f
+++ b/SRC/csytf2.f
@@ -76,8 +76,7 @@
 *>          leading n-by-n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L (see below for further details).
 *> \endverbatim
diff --git a/SRC/csytrf.f b/SRC/csytrf.f
index 289d0766..f3206b38 100644
--- a/SRC/csytrf.f
+++ b/SRC/csytrf.f
@@ -75,8 +75,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L (see below for further details).
 *> \endverbatim
@@ -111,8 +110,7 @@
 *>          LWORK is INTEGER
 *>          The length of WORK.  LWORK >=1.  For best performance
 *>          LWORK >= N*NB, where NB is the block size returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/csytri.f b/SRC/csytri.f
index afe2caa5..c41a0615 100644
--- a/SRC/csytri.f
+++ b/SRC/csytri.f
@@ -64,8 +64,7 @@
 *>          A is COMPLEX array, dimension (LDA,N)
 *>          On entry, the block diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by CSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/csytri2.f b/SRC/csytri2.f
index 80714b24..92f9ec7c 100644
--- a/SRC/csytri2.f
+++ b/SRC/csytri2.f
@@ -65,8 +65,7 @@
 *>          A is COMPLEX array, dimension (LDA,N)
 *>          On entry, the NB diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by CSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/csytri2x.f b/SRC/csytri2x.f
index d12f1faa..fd8c32b2 100644
--- a/SRC/csytri2x.f
+++ b/SRC/csytri2x.f
@@ -64,8 +64,7 @@
 *>          A is COMPLEX array, dimension (LDA,N)
 *>          On entry, the NNB diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by CSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/ctfsm.f b/SRC/ctfsm.f
index b36ee28a..8b6e15c1 100644
--- a/SRC/ctfsm.f
+++ b/SRC/ctfsm.f
@@ -68,14 +68,11 @@
 *>          SIDE is CHARACTER*1
 *>           On entry, SIDE specifies whether op( A ) appears on the left
 *>           or right of X as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              SIDE = 'L' or 'l'   op( A )*X = alpha*B.
-*> \endverbatim
-*> \verbatim
+*>
 *>              SIDE = 'R' or 'r'   X*op( A ) = alpha*B.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -86,8 +83,7 @@
 *>           an upper or lower triangular matrix as follows:
 *>           UPLO = 'U' or 'u' RFP A came from an upper triangular matrix
 *>           UPLO = 'L' or 'l' RFP A came from a  lower triangular matrix
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -96,14 +92,11 @@
 *>          TRANS is CHARACTER*1
 *>           On entry, TRANS  specifies the form of op( A ) to be used
 *>           in the matrix multiplication as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              TRANS  = 'N' or 'n'   op( A ) = A.
-*> \endverbatim
-*> \verbatim
+*>
 *>              TRANS  = 'C' or 'c'   op( A ) = conjg( A' ).
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -112,15 +105,12 @@
 *>          DIAG is CHARACTER*1
 *>           On entry, DIAG specifies whether or not RFP A is unit
 *>           triangular as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              DIAG = 'U' or 'u'   A is assumed to be unit triangular.
-*> \endverbatim
-*> \verbatim
+*>
 *>              DIAG = 'N' or 'n'   A is not assumed to be unit
 *>                                  triangular.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/ctftri.f b/SRC/ctftri.f
index a55283aa..ea606490 100644
--- a/SRC/ctftri.f
+++ b/SRC/ctftri.f
@@ -85,8 +85,7 @@
 *>          elements of lower packed A. The LDA of RFP A is (N+1)/2 when
 *>          TRANSR = 'C'. When TRANSR is 'N' the LDA is N+1 when N is
 *>          even and N is odd. See the Note below for more details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the (triangular) inverse of the original matrix, in
 *>          the same storage format.
 *> \endverbatim
diff --git a/SRC/ctgsen.f b/SRC/ctgsen.f
index 9a5e24b7..b9a17569 100644
--- a/SRC/ctgsen.f
+++ b/SRC/ctgsen.f
@@ -154,8 +154,7 @@
 *> \param[out] BETA
 *> \verbatim
 *>          BETA is COMPLEX array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          The diagonal elements of A and B, respectively,
 *>          when the pair (A,B) has been reduced to generalized Schur
 *>          form.  ALPHA(i)/BETA(i) i=1,...,N are the generalized
@@ -213,8 +212,7 @@
 *> \param[out] PR
 *> \verbatim
 *>          PR is REAL
-*> \endverbatim
-*> \verbatim
+*>
 *>          If IJOB = 1, 4 or 5, PL, PR are lower bounds on the
 *>          reciprocal  of the norm of "projections" onto left and right
 *>          eigenspace with respect to the selected cluster.
@@ -247,8 +245,7 @@
 *>          The dimension of the array WORK. LWORK >=  1
 *>          If IJOB = 1, 2 or 4, LWORK >=  2*M*(N-M)
 *>          If IJOB = 3 or 5, LWORK >=  4*M*(N-M)
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
@@ -267,8 +264,7 @@
 *>          The dimension of the array IWORK. LIWORK >= 1.
 *>          If IJOB = 1, 2 or 4, LIWORK >=  N+2;
 *>          If IJOB = 3 or 5, LIWORK >= MAX(N+2, 2*M*(N-M));
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal size of the IWORK array,
 *>          returns this value as the first entry of the IWORK array, and
diff --git a/SRC/ctgsyl.f b/SRC/ctgsyl.f
index 0066823b..79efe853 100644
--- a/SRC/ctgsyl.f
+++ b/SRC/ctgsyl.f
@@ -232,8 +232,7 @@
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK. LWORK > = 1.
 *>          If IJOB = 1 or 2 and TRANS = 'N', LWORK >= max(1,2*M*N).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/ctrexc.f b/SRC/ctrexc.f
index ee090a3b..9722600e 100644
--- a/SRC/ctrexc.f
+++ b/SRC/ctrexc.f
@@ -96,8 +96,7 @@
 *> \param[in] ILST
 *> \verbatim
 *>          ILST is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          Specify the reordering of the diagonal elements of T:
 *>          The element with row index IFST is moved to row ILST by a
 *>          sequence of transpositions between adjacent elements.
diff --git a/SRC/ctrsen.f b/SRC/ctrsen.f
index 2d0b70c0..ef22122b 100644
--- a/SRC/ctrsen.f
+++ b/SRC/ctrsen.f
@@ -161,8 +161,7 @@
 *>          If JOB = 'N', LWORK >= 1;
 *>          if JOB = 'E', LWORK = max(1,M*(N-M));
 *>          if JOB = 'V' or 'B', LWORK >= max(1,2*M*(N-M)).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/ctrti2.f b/SRC/ctrti2.f
index 74f05a9f..5c11fd78 100644
--- a/SRC/ctrti2.f
+++ b/SRC/ctrti2.f
@@ -78,8 +78,7 @@
 *>          triangular part of A is not referenced.  If DIAG = 'U', the
 *>          diagonal elements of A are also not referenced and are
 *>          assumed to be 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the (triangular) inverse of the original matrix, in
 *>          the same storage format.
 *> \endverbatim
diff --git a/SRC/ctzrzf.f b/SRC/ctzrzf.f
index 36506b8e..95945e18 100644
--- a/SRC/ctzrzf.f
+++ b/SRC/ctzrzf.f
@@ -95,8 +95,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,M).
 *>          For optimum performance LWORK >= M*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cunbdb.f b/SRC/cunbdb.f
index 50e08769..1f14986f 100644
--- a/SRC/cunbdb.f
+++ b/SRC/cunbdb.f
@@ -234,8 +234,7 @@
 *> \verbatim
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK. LWORK >= M-Q.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cuncsd.f b/SRC/cuncsd.f
index f7fdff3a..ad81209c 100644
--- a/SRC/cuncsd.f
+++ b/SRC/cuncsd.f
@@ -252,8 +252,7 @@
 *> \verbatim
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the work array, and no error
@@ -275,8 +274,7 @@
 *> \verbatim
 *>          LRWORK is INTEGER
 *>          The dimension of the array RWORK.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LRWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the RWORK array, returns
 *>          this value as the first entry of the work array, and no error
@@ -295,12 +293,10 @@
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value.
 *>          > 0:  CBBCSD did not converge. See the description of RWORK
 *>                above for details.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Reference
 *>  =========
-*> \endverbatim
-*> \verbatim
+*>
 *>  [1] Brian D. Sutton. Computing the complete CS decomposition. Numer.
 *>      Algorithms, 50(1):33-65, 2009.
 *> \endverbatim
diff --git a/SRC/cungbr.f b/SRC/cungbr.f
index ed4b2463..13e333ce 100644
--- a/SRC/cungbr.f
+++ b/SRC/cungbr.f
@@ -129,8 +129,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,min(M,N)).
 *>          For optimum performance LWORK >= min(M,N)*NB, where NB
 *>          is the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cunghr.f b/SRC/cunghr.f
index 966c4ecf..8288e163 100644
--- a/SRC/cunghr.f
+++ b/SRC/cunghr.f
@@ -58,8 +58,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          ILO and IHI must have the same values as in the previous call
 *>          of CGEHRD. Q is equal to the unit matrix except in the
 *>          submatrix Q(ilo+1:ihi,ilo+1:ihi).
@@ -99,8 +98,7 @@
 *>          The dimension of the array WORK. LWORK >= IHI-ILO.
 *>          For optimum performance LWORK >= (IHI-ILO)*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cunglq.f b/SRC/cunglq.f
index 028fcfbc..8b263bf0 100644
--- a/SRC/cunglq.f
+++ b/SRC/cunglq.f
@@ -99,8 +99,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,M).
 *>          For optimum performance LWORK >= M*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cungql.f b/SRC/cungql.f
index 4fe19daf..59510b6c 100644
--- a/SRC/cungql.f
+++ b/SRC/cungql.f
@@ -100,8 +100,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is the
 *>          optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cungqr.f b/SRC/cungqr.f
index 00612f74..71b9294f 100644
--- a/SRC/cungqr.f
+++ b/SRC/cungqr.f
@@ -100,8 +100,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is the
 *>          optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cungrq.f b/SRC/cungrq.f
index 76ee4058..f6cca67d 100644
--- a/SRC/cungrq.f
+++ b/SRC/cungrq.f
@@ -100,8 +100,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,M).
 *>          For optimum performance LWORK >= M*NB, where NB is the
 *>          optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cungtr.f b/SRC/cungtr.f
index cb682813..f31fe245 100644
--- a/SRC/cungtr.f
+++ b/SRC/cungtr.f
@@ -95,8 +95,7 @@
 *>          The dimension of the array WORK. LWORK >= N-1.
 *>          For optimum performance LWORK >= (N-1)*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cunmbr.f b/SRC/cunmbr.f
index 420804b4..785e540c 100644
--- a/SRC/cunmbr.f
+++ b/SRC/cunmbr.f
@@ -168,8 +168,7 @@
 *>          For optimum performance LWORK >= max(1,N*NB) if SIDE = 'L',
 *>          and LWORK >= max(1,M*NB) if SIDE = 'R', where NB is the
 *>          optimal blocksize. (NB = 0 if M = 0 or N = 0.)
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cunmhr.f b/SRC/cunmhr.f
index 191ca113..4fffdeaf 100644
--- a/SRC/cunmhr.f
+++ b/SRC/cunmhr.f
@@ -87,8 +87,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          ILO and IHI must have the same values as in the previous call
 *>          of CGEHRD. Q is equal to the unit matrix except in the
 *>          submatrix Q(ilo+1:ihi,ilo+1:ihi).
@@ -151,8 +150,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cunmlq.f b/SRC/cunmlq.f
index 9c914027..76740467 100644
--- a/SRC/cunmlq.f
+++ b/SRC/cunmlq.f
@@ -142,8 +142,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cunmql.f b/SRC/cunmql.f
index 817955b3..72147b56 100644
--- a/SRC/cunmql.f
+++ b/SRC/cunmql.f
@@ -142,8 +142,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cunmqr.f b/SRC/cunmqr.f
index 4e01bd52..cbeb19ae 100644
--- a/SRC/cunmqr.f
+++ b/SRC/cunmqr.f
@@ -142,8 +142,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cunmrq.f b/SRC/cunmrq.f
index 0198cb80..093ae0b3 100644
--- a/SRC/cunmrq.f
+++ b/SRC/cunmrq.f
@@ -142,8 +142,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cunmrz.f b/SRC/cunmrz.f
index f2441b56..42c6e3a2 100644
--- a/SRC/cunmrz.f
+++ b/SRC/cunmrz.f
@@ -149,8 +149,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/cunmtr.f b/SRC/cunmtr.f
index 9169891f..a4247d17 100644
--- a/SRC/cunmtr.f
+++ b/SRC/cunmtr.f
@@ -143,8 +143,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >=M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dbbcsd.f b/SRC/dbbcsd.f
index dbc28641..23aaf533 100644
--- a/SRC/dbbcsd.f
+++ b/SRC/dbbcsd.f
@@ -282,8 +282,7 @@
 *> \verbatim
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK. LWORK >= MAX(1,8*Q).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal size of the WORK array,
 *>          returns this value as the first entry of the work array, and
@@ -298,20 +297,16 @@
 *>          > 0:  if DBBCSD did not converge, INFO specifies the number
 *>                of nonzero entries in PHI, and B11D, B11E, etc.,
 *>                contain the partially reduced matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Reference
 *>  =========
-*> \endverbatim
-*> \verbatim
+*>
 *>  [1] Brian D. Sutton. Computing the complete CS decomposition. Numer.
 *>      Algorithms, 50(1):33-65, 2009.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  TOLMUL  DOUBLE PRECISION, default = MAX(10,MIN(100,EPS**(-1/8)))
 *>          TOLMUL controls the convergence criterion of the QR loop.
 *>          Angles THETA(i), PHI(i) are rounded to 0 or PI/2 when they
diff --git a/SRC/dbdsqr.f b/SRC/dbdsqr.f
index 96abe542..cab83f3b 100644
--- a/SRC/dbdsqr.f
+++ b/SRC/dbdsqr.f
@@ -187,12 +187,10 @@
 *>                   elements of a bidiagonal matrix which is orthogonally
 *>                   similar to the input matrix B;  if INFO = i, i
 *>                   elements of E have not converged to zero.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  TOLMUL  DOUBLE PRECISION, default = max(10,min(100,EPS**(-1/8)))
 *>          TOLMUL controls the convergence criterion of the QR loop.
 *>          If it is positive, TOLMUL*EPS is the desired relative
@@ -207,8 +205,7 @@
 *>          Default is to lose at either one eighth or 2 of the
 *>             available decimal digits in each computed singular value
 *>             (whichever is smaller).
-*> \endverbatim
-*> \verbatim
+*>
 *>  MAXITR  INTEGER, default = 6
 *>          MAXITR controls the maximum number of passes of the
 *>          algorithm through its inner loop. The algorithms stops
diff --git a/SRC/dgbrfs.f b/SRC/dgbrfs.f
index d2930ef7..2dba93a0 100644
--- a/SRC/dgbrfs.f
+++ b/SRC/dgbrfs.f
@@ -180,12 +180,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/dgbrfsx.f b/SRC/dgbrfsx.f
index 5a331ffe..86839c82 100644
--- a/SRC/dgbrfsx.f
+++ b/SRC/dgbrfsx.f
@@ -256,37 +256,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -295,8 +289,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -307,14 +300,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -322,26 +313,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -352,8 +339,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -372,8 +358,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0D+0
@@ -384,8 +369,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -395,8 +379,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/dgbsvx.f b/SRC/dgbsvx.f
index 5ee6be0f..e33e31a2 100644
--- a/SRC/dgbsvx.f
+++ b/SRC/dgbsvx.f
@@ -150,14 +150,12 @@
 *>          The j-th column of A is stored in the j-th column of the
 *>          array AB as follows:
 *>          AB(KU+1+i-j,j) = A(i,j) for max(1,j-KU)<=i<=min(N,j+kl)
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'F' and EQUED is not 'N', then A must have been
 *>          equilibrated by the scaling factors in R and/or C.  AB is not
 *>          modified if FACT = 'F' or 'N', or if FACT = 'E' and
 *>          EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if EQUED .ne. 'N', A is scaled as follows:
 *>          EQUED = 'R':  A := diag(R) * A
 *>          EQUED = 'C':  A := A * diag(C)
@@ -180,12 +178,10 @@
 *>          and the multipliers used during the factorization are stored
 *>          in rows KL+KU+2 to 2*KL+KU+1.  If EQUED .ne. 'N', then AFB is
 *>          the factored form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AFB is an output argument and on exit
 *>          returns details of the LU factorization of A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then AFB is an output argument and on exit
 *>          returns details of the LU factorization of the equilibrated
 *>          matrix A (see the description of AB for the form of the
@@ -205,13 +201,11 @@
 *>          contains the pivot indices from the factorization A = L*U
 *>          as computed by DGBTRF; row i of the matrix was interchanged
 *>          with row IPIV(i).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains the pivot indices from the factorization A = L*U
 *>          of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then IPIV is an output argument and on exit
 *>          contains the pivot indices from the factorization A = L*U
 *>          of the equilibrated matrix A.
diff --git a/SRC/dgbsvxx.f b/SRC/dgbsvxx.f
index 116fc3ed..f35aa518 100644
--- a/SRC/dgbsvxx.f
+++ b/SRC/dgbsvxx.f
@@ -178,14 +178,12 @@
 *>     The j-th column of A is stored in the j-th column of the
 *>     array AB as follows:
 *>     AB(KU+1+i-j,j) = A(i,j) for max(1,j-KU)<=i<=min(N,j+kl)
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'F' and EQUED is not 'N', then AB must have been
 *>     equilibrated by the scaling factors in R and/or C.  AB is not
 *>     modified if FACT = 'F' or 'N', or if FACT = 'E' and
 *>     EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>     On exit, if EQUED .ne. 'N', A is scaled as follows:
 *>     EQUED = 'R':  A := diag(R) * A
 *>     EQUED = 'C':  A := A * diag(C)
@@ -208,13 +206,11 @@
 *>     and the multipliers used during the factorization are stored
 *>     in rows KL+KU+2 to 2*KL+KU+1.  If EQUED .ne. 'N', then AFB is
 *>     the factored form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then AF is an output argument and on exit
 *>     returns the factors L and U from the factorization A = P*L*U
 *>     of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'E', then AF is an output argument and on exit
 *>     returns the factors L and U from the factorization A = P*L*U
 *>     of the equilibrated matrix A (see the description of A for
@@ -234,13 +230,11 @@
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     as computed by DGETRF; row i of the matrix was interchanged
 *>     with row IPIV(i).
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then IPIV is an output argument and on exit
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'E', then IPIV is an output argument and on exit
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     of the equilibrated matrix A.
@@ -380,37 +374,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -419,8 +407,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -431,14 +418,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -446,26 +431,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -476,8 +457,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -496,8 +476,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0D+0
@@ -505,8 +484,7 @@
 *>                    computed.
 *>            = 1.0 : Use the extra-precise refinement algorithm.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -516,8 +494,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/dgbtf2.f b/SRC/dgbtf2.f
index 6907f1d1..174c1712 100644
--- a/SRC/dgbtf2.f
+++ b/SRC/dgbtf2.f
@@ -76,8 +76,7 @@
 *>          The j-th column of A is stored in the j-th column of the
 *>          array AB as follows:
 *>          AB(kl+ku+1+i-j,j) = A(i,j) for max(1,j-ku)<=i<=min(m,j+kl)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, details of the factorization: U is stored as an
 *>          upper triangular band matrix with KL+KU superdiagonals in
 *>          rows 1 to KL+KU+1, and the multipliers used during the
diff --git a/SRC/dgbtrf.f b/SRC/dgbtrf.f
index 5e9e9d8a..76f57cb4 100644
--- a/SRC/dgbtrf.f
+++ b/SRC/dgbtrf.f
@@ -76,8 +76,7 @@
 *>          The j-th column of A is stored in the j-th column of the
 *>          array AB as follows:
 *>          AB(kl+ku+1+i-j,j) = A(i,j) for max(1,j-ku)<=i<=min(m,j+kl)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, details of the factorization: U is stored as an
 *>          upper triangular band matrix with KL+KU superdiagonals in
 *>          rows 1 to KL+KU+1, and the multipliers used during the
diff --git a/SRC/dgeesx.f b/SRC/dgeesx.f
index 3900b5d7..0f648b56 100644
--- a/SRC/dgeesx.f
+++ b/SRC/dgeesx.f
@@ -209,8 +209,7 @@
 *>          returned if LWORK < max(1,3*N), but if SENSE = 'E' or 'V' or
 *>          'B' this may not be large enough.
 *>          For good performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates upper bounds on the optimal sizes of the
 *>          arrays WORK and IWORK, returns these values as the first
@@ -232,8 +231,7 @@
 *>          Note that SDIM*(N-SDIM) <= N*N/4. Note also that an error is
 *>          only returned if LIWORK < 1, but if SENSE = 'V' or 'B' this
 *>          may not be large enough.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates upper bounds on the optimal sizes of
 *>          the arrays WORK and IWORK, returns these values as the first
diff --git a/SRC/dgeev.f b/SRC/dgeev.f
index b548908f..e1ceb714 100644
--- a/SRC/dgeev.f
+++ b/SRC/dgeev.f
@@ -156,8 +156,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,3*N), and
 *>          if JOBVL = 'V' or JOBVR = 'V', LWORK >= 4*N.  For good
 *>          performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dgeevx.f b/SRC/dgeevx.f
index 1c022882..fbcff637 100644
--- a/SRC/dgeevx.f
+++ b/SRC/dgeevx.f
@@ -89,8 +89,7 @@
 *>                 to make the rows and columns of A more equal in
 *>                 norm. Do not permute;
 *>          = 'B': Both diagonally scale and permute A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Computed reciprocal condition numbers will be for the matrix
 *>          after balancing and/or permuting. Permuting does not change
 *>          condition numbers (in exact arithmetic), but balancing does.
@@ -120,8 +119,7 @@
 *>          = 'E': Computed for eigenvalues only;
 *>          = 'V': Computed for right eigenvectors only;
 *>          = 'B': Computed for eigenvalues and right eigenvectors.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If SENSE = 'E' or 'B', both left and right eigenvectors
 *>          must also be computed (JOBVL = 'V' and JOBVR = 'V').
 *> \endverbatim
@@ -265,8 +263,7 @@
 *>          LWORK >= max(1,2*N), and if JOBVL = 'V' or JOBVR = 'V',
 *>          LWORK >= 3*N.  If SENSE = 'V' or 'B', LWORK >= N*(N+6).
 *>          For good performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dgegs.f b/SRC/dgegs.f
index c443fd69..cd10b6dd 100644
--- a/SRC/dgegs.f
+++ b/SRC/dgegs.f
@@ -182,8 +182,7 @@
 *>          blocksizes (for DGEQRF, DORMQR, and DORGQR.)  Then compute:
 *>          NB  -- MAX of the blocksizes for DGEQRF, DORMQR, and DORGQR
 *>          The optimal LWORK is  2*N + N*(NB+1).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dgegv.f b/SRC/dgegv.f
index 0d41f81f..b22b7e58 100644
--- a/SRC/dgegv.f
+++ b/SRC/dgegv.f
@@ -171,8 +171,7 @@
 *>             u(j) = VL(:,j) + i*VL(:,j+1)
 *>          and
 *>            u(j+1) = VL(:,j) - i*VL(:,j+1).
-*> \endverbatim
-*> \verbatim
+*>
 *>          Each eigenvector is scaled so that its largest component has
 *>          abs(real part) + abs(imag. part) = 1, except for eigenvectors
 *>          corresponding to an eigenvalue with alpha = beta = 0, which
@@ -198,8 +197,7 @@
 *>            x(j) = VR(:,j) + i*VR(:,j+1)
 *>          and
 *>            x(j+1) = VR(:,j) - i*VR(:,j+1).
-*> \endverbatim
-*> \verbatim
+*>
 *>          Each eigenvector is scaled so that its largest component has
 *>          abs(real part) + abs(imag. part) = 1, except for eigenvalues
 *>          corresponding to an eigenvalue with alpha = beta = 0, which
@@ -230,8 +228,7 @@
 *>          NB  -- MAX of the blocksizes for DGEQRF, DORMQR, and DORGQR;
 *>          The optimal LWORK is:
 *>              2*N + MAX( 6*N, N*(NB+1) ).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dgehd2.f b/SRC/dgehd2.f
index b69d1dc9..e2ab7ffa 100644
--- a/SRC/dgehd2.f
+++ b/SRC/dgehd2.f
@@ -55,8 +55,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          It is assumed that A is already upper triangular in rows
 *>          and columns 1:ILO-1 and IHI+1:N. ILO and IHI are normally
 *>          set by a previous call to DGEBAL; otherwise they should be
diff --git a/SRC/dgehrd.f b/SRC/dgehrd.f
index 2076e757..fdf8f6b2 100644
--- a/SRC/dgehrd.f
+++ b/SRC/dgehrd.f
@@ -55,8 +55,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          It is assumed that A is already upper triangular in rows
 *>          and columns 1:ILO-1 and IHI+1:N. ILO and IHI are normally
 *>          set by a previous call to DGEBAL; otherwise they should be
diff --git a/SRC/dgels.f b/SRC/dgels.f
index fb016acb..ac76f110 100644
--- a/SRC/dgels.f
+++ b/SRC/dgels.f
@@ -150,8 +150,7 @@
 *>          For optimal performance,
 *>          LWORK >= max( 1, MN + max( MN, NRHS )*NB ).
 *>          where MN = min(M,N) and NB is the optimum block size.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dgelsd.f b/SRC/dgelsd.f
index 9641b055..4f40929c 100644
--- a/SRC/dgelsd.f
+++ b/SRC/dgelsd.f
@@ -162,8 +162,7 @@
 *>          tree (usually about 25), and
 *>             NLVL = MAX( 0, INT( LOG_2( MIN( M,N )/(SMLSIZ+1) ) ) + 1 )
 *>          For good performance, LWORK should generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dgelss.f b/SRC/dgelss.f
index e889d35c..3be272d8 100644
--- a/SRC/dgelss.f
+++ b/SRC/dgelss.f
@@ -140,8 +140,7 @@
 *>          The dimension of the array WORK. LWORK >= 1, and also:
 *>          LWORK >= 3*min(M,N) + max( 2*min(M,N), max(M,N), NRHS )
 *>          For good performance, LWORK should generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dgelsy.f b/SRC/dgelsy.f
index 4a69b167..2aa8a49f 100644
--- a/SRC/dgelsy.f
+++ b/SRC/dgelsy.f
@@ -168,8 +168,7 @@
 *>          where NB is an upper bound on the blocksize returned
 *>          by ILAENV for the routines DGEQP3, DTZRZF, STZRQF, DORMQR,
 *>          and DORMRZ.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dgeqp3.f b/SRC/dgeqp3.f
index 1de642c3..30f2f321 100644
--- a/SRC/dgeqp3.f
+++ b/SRC/dgeqp3.f
@@ -99,8 +99,7 @@
 *>          The dimension of the array WORK. LWORK >= 3*N+1.
 *>          For optimal performance LWORK >= 2*N+( N+1 )*NB, where NB
 *>          is the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dgeqrf.f b/SRC/dgeqrf.f
index 10c11222..50254dc4 100644
--- a/SRC/dgeqrf.f
+++ b/SRC/dgeqrf.f
@@ -90,8 +90,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dgeqrfp.f b/SRC/dgeqrfp.f
index e3d6f14d..07ce01b1 100644
--- a/SRC/dgeqrfp.f
+++ b/SRC/dgeqrfp.f
@@ -90,8 +90,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dgerfs.f b/SRC/dgerfs.f
index 2984616d..62485c0a 100644
--- a/SRC/dgerfs.f
+++ b/SRC/dgerfs.f
@@ -161,12 +161,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/dgerfsx.f b/SRC/dgerfsx.f
index 7cf79ab1..523cf340 100644
--- a/SRC/dgerfsx.f
+++ b/SRC/dgerfsx.f
@@ -231,37 +231,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -270,8 +264,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -282,14 +275,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -297,26 +288,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -327,8 +314,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -347,8 +333,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0D+0
@@ -359,8 +344,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -370,8 +354,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/dgesvd.f b/SRC/dgesvd.f
index 8ba44666..1b8e0f86 100644
--- a/SRC/dgesvd.f
+++ b/SRC/dgesvd.f
@@ -81,8 +81,7 @@
 *>                  vectors) are overwritten on the array A;
 *>          = 'N':  no rows of V**T (no right singular vectors) are
 *>                  computed.
-*> \endverbatim
-*> \verbatim
+*>
 *>          JOBVT and JOBU cannot both be 'O'.
 *> \endverbatim
 *>
@@ -179,8 +178,7 @@
 *>             - PATH 1t (N much larger than M, JOBVT='N')
 *>          LWORK >= MAX(1,3*MIN(M,N)+MAX(M,N),5*MIN(M,N)) for the other paths
 *>          For good performance, LWORK should generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dgesvj.f b/SRC/dgesvj.f
index 8dced567..6e0bfdd3 100644
--- a/SRC/dgesvj.f
+++ b/SRC/dgesvj.f
@@ -138,8 +138,7 @@
 *>                 values in SVA(1:N)) and V is still a decomposition of the
 *>                 input matrix A in the sense that the residual
 *>                 ||A-SCALE*U*SIGMA*V^T||_2 / ||A||_2 is small.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If JOBU .EQ. 'N' :
 *>                 If INFO .EQ. 0 :
 *>                 Note that the left singular vectors are 'for free' in the
diff --git a/SRC/dgesvx.f b/SRC/dgesvx.f
index f67834a0..798ff9f6 100644
--- a/SRC/dgesvx.f
+++ b/SRC/dgesvx.f
@@ -137,8 +137,7 @@
 *>          not 'N', then A must have been equilibrated by the scaling
 *>          factors in R and/or C.  A is not modified if FACT = 'F' or
 *>          'N', or if FACT = 'E' and EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if EQUED .ne. 'N', A is scaled as follows:
 *>          EQUED = 'R':  A := diag(R) * A
 *>          EQUED = 'C':  A := A * diag(C)
@@ -158,13 +157,11 @@
 *>          contains the factors L and U from the factorization
 *>          A = P*L*U as computed by DGETRF.  If EQUED .ne. 'N', then
 *>          AF is the factored form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AF is an output argument and on exit
 *>          returns the factors L and U from the factorization A = P*L*U
 *>          of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then AF is an output argument and on exit
 *>          returns the factors L and U from the factorization A = P*L*U
 *>          of the equilibrated matrix A (see the description of A for
@@ -184,13 +181,11 @@
 *>          contains the pivot indices from the factorization A = P*L*U
 *>          as computed by DGETRF; row i of the matrix was interchanged
 *>          with row IPIV(i).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains the pivot indices from the factorization A = P*L*U
 *>          of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then IPIV is an output argument and on exit
 *>          contains the pivot indices from the factorization A = P*L*U
 *>          of the equilibrated matrix A.
diff --git a/SRC/dgesvxx.f b/SRC/dgesvxx.f
index fc25b6e7..88411a3c 100644
--- a/SRC/dgesvxx.f
+++ b/SRC/dgesvxx.f
@@ -166,8 +166,7 @@
 *>     not 'N', then A must have been equilibrated by the scaling
 *>     factors in R and/or C.  A is not modified if FACT = 'F' or
 *>     'N', or if FACT = 'E' and EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>     On exit, if EQUED .ne. 'N', A is scaled as follows:
 *>     EQUED = 'R':  A := diag(R) * A
 *>     EQUED = 'C':  A := A * diag(C)
@@ -187,13 +186,11 @@
 *>     contains the factors L and U from the factorization
 *>     A = P*L*U as computed by DGETRF.  If EQUED .ne. 'N', then
 *>     AF is the factored form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then AF is an output argument and on exit
 *>     returns the factors L and U from the factorization A = P*L*U
 *>     of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'E', then AF is an output argument and on exit
 *>     returns the factors L and U from the factorization A = P*L*U
 *>     of the equilibrated matrix A (see the description of A for
@@ -213,13 +210,11 @@
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     as computed by DGETRF; row i of the matrix was interchanged
 *>     with row IPIV(i).
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then IPIV is an output argument and on exit
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'E', then IPIV is an output argument and on exit
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     of the equilibrated matrix A.
@@ -359,37 +354,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -398,8 +387,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -410,14 +398,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -425,26 +411,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -455,8 +437,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -475,8 +456,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0D+0
@@ -484,8 +464,7 @@
 *>                    computed.
 *>            = 1.0 : Use the extra-precise refinement algorithm.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -495,8 +474,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/dgetri.f b/SRC/dgetri.f
index e13b2a46..82b32cc0 100644
--- a/SRC/dgetri.f
+++ b/SRC/dgetri.f
@@ -84,8 +84,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,N).
 *>          For optimal performance LWORK >= N*NB, where NB is
 *>          the optimal blocksize returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dgges.f b/SRC/dgges.f
index bc2c69ae..44b8f6b1 100644
--- a/SRC/dgges.f
+++ b/SRC/dgges.f
@@ -116,8 +116,7 @@
 *>          SELCTG(ALPHAR(j),ALPHAI(j),BETA(j)) is true; i.e. if either
 *>          one of a complex conjugate pair of eigenvalues is selected,
 *>          then both complex eigenvalues are selected.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note that in the ill-conditioned case, a selected complex
 *>          eigenvalue may no longer satisfy SELCTG(ALPHAR(j),ALPHAI(j),
 *>          BETA(j)) = .TRUE. after ordering. INFO is to be set to N+2
@@ -189,8 +188,7 @@
 *>          If ALPHAI(j) is zero, then the j-th eigenvalue is real; if
 *>          positive, then the j-th and (j+1)-st eigenvalues are a
 *>          complex conjugate pair, with ALPHAI(j+1) negative.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note: the quotients ALPHAR(j)/BETA(j) and ALPHAI(j)/BETA(j)
 *>          may easily over- or underflow, and BETA(j) may even be zero.
 *>          Thus, the user should avoid naively computing the ratio.
@@ -239,8 +237,7 @@
 *>          The dimension of the array WORK.
 *>          If N = 0, LWORK >= 1, else LWORK >= 8*N+16.
 *>          For good performance , LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dggesx.f b/SRC/dggesx.f
index b123c08b..4d6d6830 100644
--- a/SRC/dggesx.f
+++ b/SRC/dggesx.f
@@ -204,8 +204,7 @@
 *>          If ALPHAI(j) is zero, then the j-th eigenvalue is real; if
 *>          positive, then the j-th and (j+1)-st eigenvalues are a
 *>          complex conjugate pair, with ALPHAI(j+1) negative.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note: the quotients ALPHAR(j)/BETA(j) and ALPHAI(j)/BETA(j)
 *>          may easily over- or underflow, and BETA(j) may even be zero.
 *>          Thus, the user should avoid naively computing the ratio.
@@ -277,8 +276,7 @@
 *>          Note also that an error is only returned if
 *>          LWORK < max( 8*N, 6*N+16), but if SENSE = 'E' or 'V' or 'B'
 *>          this may not be large enough.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the bound on the optimal size of the WORK
 *>          array and the minimum size of the IWORK array, returns these
@@ -299,8 +297,7 @@
 *>          The dimension of the array IWORK.
 *>          If SENSE = 'N' or N = 0, LIWORK >= 1, otherwise
 *>          LIWORK >= N+6.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the bound on the optimal size of the
 *>          WORK array and the minimum size of the IWORK array, returns
diff --git a/SRC/dggev.f b/SRC/dggev.f
index afbb67f8..35278177 100644
--- a/SRC/dggev.f
+++ b/SRC/dggev.f
@@ -129,8 +129,7 @@
 *>          the j-th eigenvalue is real; if positive, then the j-th and
 *>          (j+1)-st eigenvalues are a complex conjugate pair, with
 *>          ALPHAI(j+1) negative.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note: the quotients ALPHAR(j)/BETA(j) and ALPHAI(j)/BETA(j)
 *>          may easily over- or underflow, and BETA(j) may even be zero.
 *>          Thus, the user should avoid naively computing the ratio
@@ -192,8 +191,7 @@
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.  LWORK >= max(1,8*N).
 *>          For good performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dggevx.f b/SRC/dggevx.f
index 6da2fe3a..0f4f86b0 100644
--- a/SRC/dggevx.f
+++ b/SRC/dggevx.f
@@ -169,8 +169,7 @@
 *>          the j-th eigenvalue is real; if positive, then the j-th and
 *>          (j+1)-st eigenvalues are a complex conjugate pair, with
 *>          ALPHAI(j+1) negative.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note: the quotients ALPHAR(j)/BETA(j) and ALPHAI(j)/BETA(j)
 *>          may easily over- or underflow, and BETA(j) may even be zero.
 *>          Thus, the user should avoid naively computing the ratio
@@ -314,8 +313,7 @@
 *>          LWORK >= max(1,6*N).
 *>          If SENSE = 'E' or 'B', LWORK >= max(1,10*N).
 *>          If SENSE = 'V' or 'B', LWORK >= 2*N*N+8*N+16.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dggglm.f b/SRC/dggglm.f
index 8fb2f5d5..945ab59f 100644
--- a/SRC/dggglm.f
+++ b/SRC/dggglm.f
@@ -130,8 +130,7 @@
 *> \param[out] Y
 *> \verbatim
 *>          Y is DOUBLE PRECISION array, dimension (P)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, X and Y are the solutions of the GLM problem.
 *> \endverbatim
 *>
@@ -148,8 +147,7 @@
 *>          For optimum performance, LWORK >= M+min(N,P)+max(N,P)*NB,
 *>          where NB is an upper bound for the optimal blocksizes for
 *>          DGEQRF, SGERQF, DORMQR and SORMRQ.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dgghrd.f b/SRC/dgghrd.f
index 57b1fd2c..5296059f 100644
--- a/SRC/dgghrd.f
+++ b/SRC/dgghrd.f
@@ -104,8 +104,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          ILO and IHI mark the rows and columns of A which are to be
 *>          reduced.  It is assumed that A is already upper triangular
 *>          in rows and columns 1:ILO-1 and IHI+1:N.  ILO and IHI are
diff --git a/SRC/dgglse.f b/SRC/dgglse.f
index 908f7a5d..0f5ac04b 100644
--- a/SRC/dgglse.f
+++ b/SRC/dgglse.f
@@ -141,8 +141,7 @@
 *>          For optimum performance LWORK >= P+min(M,N)+max(M,N)*NB,
 *>          where NB is an upper bound for the optimal blocksizes for
 *>          DGEQRF, SGERQF, DORMQR and SORMRQ.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dggsvd.f b/SRC/dggsvd.f
index fc8a6d46..fc439389 100644
--- a/SRC/dggsvd.f
+++ b/SRC/dggsvd.f
@@ -170,8 +170,7 @@
 *> \param[out] L
 *> \verbatim
 *>          L is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, K and L specify the dimension of the subblocks
 *>          described in Purpose.
 *>          K + L = effective numerical rank of (A**T,B**T)**T.
@@ -213,8 +212,7 @@
 *> \param[out] BETA
 *> \verbatim
 *>          BETA is DOUBLE PRECISION array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, ALPHA and BETA contain the generalized singular
 *>          value pairs of A and B;
 *>            ALPHA(1:K) = 1,
@@ -296,12 +294,10 @@
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value.
 *>          > 0:  if INFO = 1, the Jacobi-type procedure failed to
 *>                converge.  For further details, see subroutine DTGSJA.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  TOLA    DOUBLE PRECISION
 *>  TOLB    DOUBLE PRECISION
 *>          TOLA and TOLB are the thresholds to determine the effective
diff --git a/SRC/dggsvp.f b/SRC/dggsvp.f
index bc1a0ac1..aa82939e 100644
--- a/SRC/dggsvp.f
+++ b/SRC/dggsvp.f
@@ -143,8 +143,7 @@
 *> \param[in] TOLB
 *> \verbatim
 *>          TOLB is DOUBLE PRECISION
-*> \endverbatim
-*> \verbatim
+*>
 *>          TOLA and TOLB are the thresholds to determine the effective
 *>          numerical rank of matrix B and a subblock of A. Generally,
 *>          they are set to
@@ -162,8 +161,7 @@
 *> \param[out] L
 *> \verbatim
 *>          L is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, K and L specify the dimension of the subblocks
 *>          described in Purpose section.
 *>          K + L = effective numerical rank of (A**T,B**T)**T.
diff --git a/SRC/dgtrfs.f b/SRC/dgtrfs.f
index ae68f79e..897b81a5 100644
--- a/SRC/dgtrfs.f
+++ b/SRC/dgtrfs.f
@@ -184,12 +184,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/dgtsv.f b/SRC/dgtsv.f
index 3c242860..58cb98bd 100644
--- a/SRC/dgtsv.f
+++ b/SRC/dgtsv.f
@@ -66,8 +66,7 @@
 *>          DL is DOUBLE PRECISION array, dimension (N-1)
 *>          On entry, DL must contain the (n-1) sub-diagonal elements of
 *>          A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, DL is overwritten by the (n-2) elements of the
 *>          second super-diagonal of the upper triangular matrix U from
 *>          the LU factorization of A, in DL(1), ..., DL(n-2).
@@ -77,8 +76,7 @@
 *> \verbatim
 *>          D is DOUBLE PRECISION array, dimension (N)
 *>          On entry, D must contain the diagonal elements of A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, D is overwritten by the n diagonal elements of U.
 *> \endverbatim
 *>
@@ -87,8 +85,7 @@
 *>          DU is DOUBLE PRECISION array, dimension (N-1)
 *>          On entry, DU must contain the (n-1) super-diagonal elements
 *>          of A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, DU is overwritten by the (n-1) elements of the first
 *>          super-diagonal of U.
 *> \endverbatim
diff --git a/SRC/dgtsvx.f b/SRC/dgtsvx.f
index 69f2a7bc..e7d2d1b5 100644
--- a/SRC/dgtsvx.f
+++ b/SRC/dgtsvx.f
@@ -135,8 +135,7 @@
 *>          If FACT = 'F', then DLF is an input argument and on entry
 *>          contains the (n-1) multipliers that define the matrix L from
 *>          the LU factorization of A as computed by DGTTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then DLF is an output argument and on exit
 *>          contains the (n-1) multipliers that define the matrix L from
 *>          the LU factorization of A.
@@ -148,8 +147,7 @@
 *>          If FACT = 'F', then DF is an input argument and on entry
 *>          contains the n diagonal elements of the upper triangular
 *>          matrix U from the LU factorization of A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then DF is an output argument and on exit
 *>          contains the n diagonal elements of the upper triangular
 *>          matrix U from the LU factorization of A.
@@ -160,8 +158,7 @@
 *>          DUF is or output) DOUBLE PRECISION array, dimension (N-1)
 *>          If FACT = 'F', then DUF is an input argument and on entry
 *>          contains the (n-1) elements of the first superdiagonal of U.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then DUF is an output argument and on exit
 *>          contains the (n-1) elements of the first superdiagonal of U.
 *> \endverbatim
@@ -172,8 +169,7 @@
 *>          If FACT = 'F', then DU2 is an input argument and on entry
 *>          contains the (n-2) elements of the second superdiagonal of
 *>          U.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then DU2 is an output argument and on exit
 *>          contains the (n-2) elements of the second superdiagonal of
 *>          U.
@@ -185,8 +181,7 @@
 *>          If FACT = 'F', then IPIV is an input argument and on entry
 *>          contains the pivot indices from the LU factorization of A as
 *>          computed by DGTTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains the pivot indices from the LU factorization of A;
 *>          row i of the matrix was interchanged with row IPIV(i).
diff --git a/SRC/dgttrf.f b/SRC/dgttrf.f
index 11dfc855..173c1a76 100644
--- a/SRC/dgttrf.f
+++ b/SRC/dgttrf.f
@@ -59,8 +59,7 @@
 *>          DL is DOUBLE PRECISION array, dimension (N-1)
 *>          On entry, DL must contain the (n-1) sub-diagonal elements of
 *>          A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, DL is overwritten by the (n-1) multipliers that
 *>          define the matrix L from the LU factorization of A.
 *> \endverbatim
@@ -69,8 +68,7 @@
 *> \verbatim
 *>          D is DOUBLE PRECISION array, dimension (N)
 *>          On entry, D must contain the diagonal elements of A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, D is overwritten by the n diagonal elements of the
 *>          upper triangular matrix U from the LU factorization of A.
 *> \endverbatim
@@ -80,8 +78,7 @@
 *>          DU is DOUBLE PRECISION array, dimension (N-1)
 *>          On entry, DU must contain the (n-1) super-diagonal elements
 *>          of A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, DU is overwritten by the (n-1) elements of the first
 *>          super-diagonal of U.
 *> \endverbatim
diff --git a/SRC/dhgeqz.f b/SRC/dhgeqz.f
index 27c22f1e..849b76cc 100644
--- a/SRC/dhgeqz.f
+++ b/SRC/dhgeqz.f
@@ -255,8 +255,7 @@
 *> \verbatim
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.  LWORK >= max(1,N).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dhsein.f b/SRC/dhsein.f
index 9e056686..b59fd91f 100644
--- a/SRC/dhsein.f
+++ b/SRC/dhsein.f
@@ -125,8 +125,7 @@
 *> \param[in] WI
 *> \verbatim
 *>          WI is DOUBLE PRECISION array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On entry, the real and imaginary parts of the eigenvalues of
 *>          H; a complex conjugate pair of eigenvalues must be stored in
 *>          consecutive elements of WR and WI.
diff --git a/SRC/dhseqr.f b/SRC/dhseqr.f
index 8e17443f..9e665906 100644
--- a/SRC/dhseqr.f
+++ b/SRC/dhseqr.f
@@ -83,8 +83,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>           It is assumed that H is already upper triangular in rows
 *>           and columns 1:ILO-1 and IHI+1:N. ILO and IHI are normally
 *>           set by a previous call to DGEBAL, and then passed to ZGEHRD
@@ -107,8 +106,7 @@
 *>           contents of H are unspecified on exit.  (The output value of
 *>           H when INFO.GT.0 is given under the description of INFO
 *>           below.)
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unlike earlier versions of DHSEQR, this subroutine may
 *>           explicitly H(i,j) = 0 for i.GT.j and j = 1, 2, ... ILO-1
 *>           or j = IHI+1, IHI+2, ... N.
@@ -128,8 +126,7 @@
 *> \param[out] WI
 *> \verbatim
 *>          WI is DOUBLE PRECISION array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>           The real and imaginary parts, respectively, of the computed
 *>           eigenvalues. If two eigenvalues are computed as a complex
 *>           conjugate pair, they are stored in consecutive elements of
@@ -180,8 +177,7 @@
 *>           may be required for optimal performance.  A workspace
 *>           query is recommended to determine the optimal workspace
 *>           size.
-*> \endverbatim
-*> \verbatim
+*>
 *>           If LWORK = -1, then DHSEQR does a workspace query.
 *>           In this case, DHSEQR checks the input parameters and
 *>           estimates the optimal workspace size for the given
@@ -200,42 +196,33 @@
 *>                the eigenvalues.  Elements 1:ilo-1 and i+1:n of WR
 *>                and WI contain those eigenvalues which have been
 *>                successfully computed.  (Failures are rare.)
-*> \endverbatim
-*> \verbatim
+*>
 *>                If INFO .GT. 0 and JOB = 'E', then on exit, the
 *>                remaining unconverged eigenvalues are the eigen-
 *>                values of the upper Hessenberg matrix rows and
 *>                columns ILO through INFO of the final, output
 *>                value of H.
-*> \endverbatim
-*> \verbatim
+*>
 *>                If INFO .GT. 0 and JOB   = 'S', then on exit
-*> \endverbatim
-*> \verbatim
+*>
 *>           (*)  (initial value of H)*U  = U*(final value of H)
-*> \endverbatim
-*> \verbatim
+*>
 *>                where U is an orthogonal matrix.  The final
 *>                value of H is upper Hessenberg and quasi-triangular
 *>                in rows and columns INFO+1 through IHI.
-*> \endverbatim
-*> \verbatim
+*>
 *>                If INFO .GT. 0 and COMPZ = 'V', then on exit
-*> \endverbatim
-*> \verbatim
+*>
 *>                  (final value of Z)  =  (initial value of Z)*U
-*> \endverbatim
-*> \verbatim
+*>
 *>                where U is the orthogonal matrix in (*) (regard-
 *>                less of the value of JOB.)
-*> \endverbatim
-*> \verbatim
+*>
 *>                If INFO .GT. 0 and COMPZ = 'I', then on exit
 *>                      (final value of Z)  = U
 *>                where U is the orthogonal matrix in (*) (regard-
 *>                less of the value of JOB.)
-*> \endverbatim
-*> \verbatim
+*>
 *>                If INFO .GT. 0 and COMPZ = 'N', then Z is not
 *>                accessed.
 *> \endverbatim
diff --git a/SRC/dla_gbamv.f b/SRC/dla_gbamv.f
index a913da81..d1da1278 100644
--- a/SRC/dla_gbamv.f
+++ b/SRC/dla_gbamv.f
@@ -62,13 +62,11 @@
 *>          TRANS is INTEGER
 *>           On entry, TRANS specifies the operation to be performed as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>             BLAS_NO_TRANS      y := alpha*abs(A)*abs(x) + beta*abs(y)
 *>             BLAS_TRANS         y := alpha*abs(A**T)*abs(x) + beta*abs(y)
 *>             BLAS_CONJ_TRANS    y := alpha*abs(A**T)*abs(x) + beta*abs(y)
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -168,8 +166,7 @@
 *>           On entry, INCY specifies the increment for the elements of
 *>           Y. INCY must not be zero.
 *>           Unchanged on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Level 2 Blas routine.
 *> \endverbatim
 *>
diff --git a/SRC/dla_geamv.f b/SRC/dla_geamv.f
index 77ca7341..b13d4756 100644
--- a/SRC/dla_geamv.f
+++ b/SRC/dla_geamv.f
@@ -62,13 +62,11 @@
 *>          TRANS is INTEGER
 *>           On entry, TRANS specifies the operation to be performed as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>             BLAS_NO_TRANS      y := alpha*abs(A)*abs(x) + beta*abs(y)
 *>             BLAS_TRANS         y := alpha*abs(A**T)*abs(x) + beta*abs(y)
 *>             BLAS_CONJ_TRANS    y := alpha*abs(A**T)*abs(x) + beta*abs(y)
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -157,8 +155,7 @@
 *>           On entry, INCY specifies the increment for the elements of
 *>           Y. INCY must not be zero.
 *>           Unchanged on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Level 2 Blas routine.
 *> \endverbatim
 *>
diff --git a/SRC/dla_porfsx_extended.f b/SRC/dla_porfsx_extended.f
index 2c5c1bac..c7b68a57 100644
--- a/SRC/dla_porfsx_extended.f
+++ b/SRC/dla_porfsx_extended.f
@@ -192,37 +192,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -231,8 +225,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     This subroutine is only responsible for setting the second field
 *>     above.
 *>     See Lapack Working Note 165 for further details and extra
@@ -246,14 +239,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -261,26 +252,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -291,8 +278,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     This subroutine is only responsible for setting the second field
 *>     above.
 *>     See Lapack Working Note 165 for further details and extra
diff --git a/SRC/dla_syamv.f b/SRC/dla_syamv.f
index aa5fd29f..21e50be8 100644
--- a/SRC/dla_syamv.f
+++ b/SRC/dla_syamv.f
@@ -62,16 +62,13 @@
 *>           On entry, UPLO specifies whether the upper or lower
 *>           triangular part of the array A is to be referenced as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = BLAS_UPPER   Only the upper triangular part of A
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = BLAS_LOWER   Only the lower triangular part of A
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/dla_syrfsx_extended.f b/SRC/dla_syrfsx_extended.f
index f5915cb0..131144db 100644
--- a/SRC/dla_syrfsx_extended.f
+++ b/SRC/dla_syrfsx_extended.f
@@ -200,37 +200,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -239,8 +233,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     This subroutine is only responsible for setting the second field
 *>     above.
 *>     See Lapack Working Note 165 for further details and extra
@@ -254,14 +247,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -269,26 +260,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -299,8 +286,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     This subroutine is only responsible for setting the second field
 *>     above.
 *>     See Lapack Working Note 165 for further details and extra
diff --git a/SRC/dlaed4.f b/SRC/dlaed4.f
index 5ac5d2f0..d0c6bd13 100644
--- a/SRC/dlaed4.f
+++ b/SRC/dlaed4.f
@@ -106,24 +106,19 @@
 *>          INFO is INTEGER
 *>         = 0:  successful exit
 *>         > 0:  if INFO = 1, the updating process failed.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  Logical variable ORGATI (origin-at-i?) is used for distinguishing
 *>  whether D(i) or D(i+1) is treated as the origin.
-*> \endverbatim
-*> \verbatim
+*>
 *>            ORGATI = .true.    origin at i
 *>            ORGATI = .false.   origin at i+1
-*> \endverbatim
-*> \verbatim
+*>
 *>   Logical variable SWTCH3 (switch-for-3-poles?) is for noting
 *>   if we are working with THREE poles!
-*> \endverbatim
-*> \verbatim
+*>
 *>   MAXIT is the maximum number of iterations allowed for each
 *>   eigenvalue.
 *> \endverbatim
diff --git a/SRC/dlagtf.f b/SRC/dlagtf.f
index 48382de5..eb4e25d8 100644
--- a/SRC/dlagtf.f
+++ b/SRC/dlagtf.f
@@ -67,8 +67,7 @@
 *> \verbatim
 *>          A is DOUBLE PRECISION array, dimension (N)
 *>          On entry, A must contain the diagonal elements of T.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, A is overwritten by the n diagonal elements of the
 *>          upper triangular matrix U of the factorization of T.
 *> \endverbatim
@@ -84,8 +83,7 @@
 *>          B is DOUBLE PRECISION array, dimension (N-1)
 *>          On entry, B must contain the (n-1) super-diagonal elements of
 *>          T.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, B is overwritten by the (n-1) super-diagonal
 *>          elements of the matrix U of the factorization of T.
 *> \endverbatim
@@ -95,8 +93,7 @@
 *>          C is DOUBLE PRECISION array, dimension (N-1)
 *>          On entry, C must contain the (n-1) sub-diagonal elements of
 *>          T.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, C is overwritten by the (n-1) sub-diagonal elements
 *>          of the matrix L of the factorization of T.
 *> \endverbatim
@@ -128,11 +125,9 @@
 *>          an interchange occurred at the kth step of the elimination,
 *>          then IN(k) = 1, otherwise IN(k) = 0. The element IN(n)
 *>          returns the smallest positive integer j such that
-*> \endverbatim
-*> \verbatim
+*>
 *>             abs( u(j,j) ).le. norm( (T - lambda*I)(j) )*TOL,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where norm( A(j) ) denotes the sum of the absolute values of
 *>          the jth row of the matrix A. If no such j exists then IN(n)
 *>          is returned as zero. If IN(n) is returned as positive, then a
diff --git a/SRC/dlagts.f b/SRC/dlagts.f
index 695b64c4..691d7a7a 100644
--- a/SRC/dlagts.f
+++ b/SRC/dlagts.f
@@ -129,8 +129,7 @@
 *>          is the relative machine precision, but if TOL is supplied as
 *>          non-positive, then it is reset to eps*max( abs( u(i,j) ) ).
 *>          If  JOB .gt. 0  then TOL is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, TOL is changed as described above, only if TOL is
 *>          non-positive on entry. Otherwise TOL is unchanged.
 *> \endverbatim
diff --git a/SRC/dlahqr.f b/SRC/dlahqr.f
index e9dd56a9..8ed28bbc 100644
--- a/SRC/dlahqr.f
+++ b/SRC/dlahqr.f
@@ -159,22 +159,19 @@
 *>                  per eigenvalue; elements i+1:ihi of WR and WI
 *>                  contain those eigenvalues which have been
 *>                  successfully computed.
-*> \endverbatim
-*> \verbatim
+*>
 *>                  If INFO .GT. 0 and WANTT is .FALSE., then on exit,
 *>                  the remaining unconverged eigenvalues are the
 *>                  eigenvalues of the upper Hessenberg matrix rows
 *>                  and columns ILO thorugh INFO of the final, output
 *>                  value of H.
-*> \endverbatim
-*> \verbatim
+*>
 *>                  If INFO .GT. 0 and WANTT is .TRUE., then on exit
 *>          (*)       (initial value of H)*U  = U*(final value of H)
 *>                  where U is an orthognal matrix.    The final
 *>                  value of H is upper Hessenberg and triangular in
 *>                  rows and columns INFO+1 through IHI.
-*> \endverbatim
-*> \verbatim
+*>
 *>                  If INFO .GT. 0 and WANTZ is .TRUE., then on exit
 *>                      (final value of Z)  = (initial value of Z)*U
 *>                  where U is the orthogonal matrix in (*)
diff --git a/SRC/dlals0.f b/SRC/dlals0.f
index da2b64e2..728584f2 100644
--- a/SRC/dlals0.f
+++ b/SRC/dlals0.f
@@ -101,8 +101,7 @@
 *>          SQRE is INTEGER
 *>         = 0: the lower block is an NR-by-NR square matrix.
 *>         = 1: the lower block is an NR-by-(NR+1) rectangular matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>         The bidiagonal matrix has row dimension N = NL + NR + 1,
 *>         and column dimension M = N + SQRE.
 *> \endverbatim
diff --git a/SRC/dlaqgb.f b/SRC/dlaqgb.f
index c7866610..d742512a 100644
--- a/SRC/dlaqgb.f
+++ b/SRC/dlaqgb.f
@@ -76,8 +76,7 @@
 *>          The j-th column of A is stored in the j-th column of the
 *>          array AB as follows:
 *>          AB(ku+1+i-j,j) = A(i,j) for max(1,j-ku)<=i<=min(m,j+kl)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the equilibrated matrix, in the same storage format
 *>          as A.  See EQUED for the form of the equilibrated matrix.
 *> \endverbatim
@@ -129,18 +128,15 @@
 *>                  by diag(C).
 *>          = 'B':  Both row and column equilibration, i.e., A has been
 *>                  replaced by diag(R) * A * diag(C).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if row or column scaling
 *>  should be done based on the ratio of the row or column scaling
 *>  factors.  If ROWCND < THRESH, row scaling is done, and if
 *>  COLCND < THRESH, column scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if row scaling
 *>  should be done based on the absolute size of the largest matrix
 *>  element.  If AMAX > LARGE or AMAX < SMALL, row scaling is done.
diff --git a/SRC/dlaqge.f b/SRC/dlaqge.f
index f424cc06..3fce578a 100644
--- a/SRC/dlaqge.f
+++ b/SRC/dlaqge.f
@@ -111,18 +111,15 @@
 *>                  by diag(C).
 *>          = 'B':  Both row and column equilibration, i.e., A has been
 *>                  replaced by diag(R) * A * diag(C).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if row or column scaling
 *>  should be done based on the ratio of the row or column scaling
 *>  factors.  If ROWCND < THRESH, row scaling is done, and if
 *>  COLCND < THRESH, column scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if row scaling
 *>  should be done based on the absolute size of the largest matrix
 *>  element.  If AMAX > LARGE or AMAX < SMALL, row scaling is done.
diff --git a/SRC/dlaqr0.f b/SRC/dlaqr0.f
index 29e0492d..985dd9f0 100644
--- a/SRC/dlaqr0.f
+++ b/SRC/dlaqr0.f
@@ -102,8 +102,7 @@
 *>           .FALSE., then the contents of H are unspecified on exit.
 *>           (The output value of H when INFO.GT.0 is given under the
 *>           description of INFO below.)
-*> \endverbatim
-*> \verbatim
+*>
 *>           This subroutine may explicitly set H(i,j) = 0 for i.GT.j and
 *>           j = 1, 2, ... ILO-1 or j = IHI+1, IHI+2, ... N.
 *> \endverbatim
@@ -180,8 +179,7 @@
 *>           is sufficient, but LWORK typically as large as 6*N may
 *>           be required for optimal performance.  A workspace query
 *>           to determine the optimal workspace size is recommended.
-*> \endverbatim
-*> \verbatim
+*>
 *>           If LWORK = -1, then DLAQR0 does a workspace query.
 *>           In this case, DLAQR0 checks the input parameters and
 *>           estimates the optimal workspace size for the given
@@ -223,10 +221,8 @@
 *>
 *>                If INFO .GT. 0 and WANTZ is .FALSE., then Z is not
 *>                accessed.
-*> \endverbatim
-*> \verbatim
-*> \endverbatim
-*> \verbatim
+*>
+*>
 *>     Based on contributions by
 *>        Karen Braman and Ralph Byers, Department of Mathematics,
 *>        University of Kansas, USA
diff --git a/SRC/dlaqr2.f b/SRC/dlaqr2.f
index 580cda40..8af7c6a1 100644
--- a/SRC/dlaqr2.f
+++ b/SRC/dlaqr2.f
@@ -248,8 +248,7 @@
 *>          The dimension of the work array WORK.  LWORK = 2*NW
 *>          suffices, but greater efficiency may result from larger
 *>          values of LWORK.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; DLAQR2
 *>          only estimates the optimal workspace size for the given
 *>          values of N, NW, KTOP and KBOT.  The estimate is returned
diff --git a/SRC/dlaqr3.f b/SRC/dlaqr3.f
index e09def41..63c98c62 100644
--- a/SRC/dlaqr3.f
+++ b/SRC/dlaqr3.f
@@ -245,8 +245,7 @@
 *>          The dimension of the work array WORK.  LWORK = 2*NW
 *>          suffices, but greater efficiency may result from larger
 *>          values of LWORK.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; DLAQR3
 *>          only estimates the optimal workspace size for the given
 *>          values of N, NW, KTOP and KBOT.  The estimate is returned
diff --git a/SRC/dlaqr4.f b/SRC/dlaqr4.f
index e5c1973a..2cfbc81b 100644
--- a/SRC/dlaqr4.f
+++ b/SRC/dlaqr4.f
@@ -109,8 +109,7 @@
 *>           .FALSE., then the contents of H are unspecified on exit.
 *>           (The output value of H when INFO.GT.0 is given under the
 *>           description of INFO below.)
-*> \endverbatim
-*> \verbatim
+*>
 *>           This subroutine may explicitly set H(i,j) = 0 for i.GT.j and
 *>           j = 1, 2, ... ILO-1 or j = IHI+1, IHI+2, ... N.
 *> \endverbatim
@@ -187,8 +186,7 @@
 *>           is sufficient, but LWORK typically as large as 6*N may
 *>           be required for optimal performance.  A workspace query
 *>           to determine the optimal workspace size is recommended.
-*> \endverbatim
-*> \verbatim
+*>
 *>           If LWORK = -1, then DLAQR4 does a workspace query.
 *>           In this case, DLAQR4 checks the input parameters and
 *>           estimates the optimal workspace size for the given
diff --git a/SRC/dlaqsb.f b/SRC/dlaqsb.f
index 85093cca..4c49597c 100644
--- a/SRC/dlaqsb.f
+++ b/SRC/dlaqsb.f
@@ -74,8 +74,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**T*U or A = L*L**T of the band
 *>          matrix A, in the same storage format as A.
@@ -112,17 +111,14 @@
 *>          = 'N':  No equilibration.
 *>          = 'Y':  Equilibration was done, i.e., A has been replaced by
 *>                  diag(S) * A * diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if scaling should be done
 *>  based on the ratio of the scaling factors.  If SCOND < THRESH,
 *>  scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if scaling should
 *>  be done based on the absolute size of the largest matrix element.
 *>  If AMAX > LARGE or AMAX < SMALL, scaling is done.
diff --git a/SRC/dlaqsp.f b/SRC/dlaqsp.f
index 262a6092..9dddd358 100644
--- a/SRC/dlaqsp.f
+++ b/SRC/dlaqsp.f
@@ -66,8 +66,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the equilibrated matrix:  diag(S) * A * diag(S), in
 *>          the same storage format as A.
 *> \endverbatim
@@ -97,17 +96,14 @@
 *>          = 'N':  No equilibration.
 *>          = 'Y':  Equilibration was done, i.e., A has been replaced by
 *>                  diag(S) * A * diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if scaling should be done
 *>  based on the ratio of the scaling factors.  If SCOND < THRESH,
 *>  scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if scaling should
 *>  be done based on the absolute size of the largest matrix element.
 *>  If AMAX > LARGE or AMAX < SMALL, scaling is done.
diff --git a/SRC/dlaqsy.f b/SRC/dlaqsy.f
index 87b0459f..0b141c2e 100644
--- a/SRC/dlaqsy.f
+++ b/SRC/dlaqsy.f
@@ -68,8 +68,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if EQUED = 'Y', the equilibrated matrix:
 *>          diag(S) * A * diag(S).
 *> \endverbatim
@@ -105,17 +104,14 @@
 *>          = 'N':  No equilibration.
 *>          = 'Y':  Equilibration was done, i.e., A has been replaced by
 *>                  diag(S) * A * diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if scaling should be done
 *>  based on the ratio of the scaling factors.  If SCOND < THRESH,
 *>  scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if scaling should
 *>  be done based on the absolute size of the largest matrix element.
 *>  If AMAX > LARGE or AMAX < SMALL, scaling is done.
diff --git a/SRC/dlarrd.f b/SRC/dlarrd.f
index cd30cca5..2cfdd268 100644
--- a/SRC/dlarrd.f
+++ b/SRC/dlarrd.f
@@ -279,12 +279,10 @@
 *>                                        floating-point arithmetic.
 *>                        Cure: Increase the PARAMETER "FUDGE",
 *>                              recompile, and try again.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  FUDGE   DOUBLE PRECISION, default = 2
 *>          A "fudge factor" to widen the Gershgorin intervals.  Ideally,
 *>          a value of 1 should work, but on machines with sloppy
@@ -292,8 +290,7 @@
 *>          publicly released versions should be large enough to handle
 *>          the worst machine around.  Note that this has no effect
 *>          on accuracy of the solution.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Based on contributions by
 *>     W. Kahan, University of California, Berkeley, USA
 *>     Beresford Parlett, University of California, Berkeley, USA
diff --git a/SRC/dlarre.f b/SRC/dlarre.f
index 41111ad0..946e0c49 100644
--- a/SRC/dlarre.f
+++ b/SRC/dlarre.f
@@ -249,8 +249,7 @@
 *>          < 0:  One of the called subroutines signaled an internal problem.
 *>                Needs inspection of the corresponding parameter IINFO
 *>                for further information.
-*> \endverbatim
-*> \verbatim
+*>
 *>          =-1:  Problem in DLARRD.
 *>          = 2:  No base representation could be found in MAXTRY iterations.
 *>                Increasing MAXTRY and recompilation might be a remedy.
diff --git a/SRC/dlarrk.f b/SRC/dlarrk.f
index 6814a2fe..b04dfcad 100644
--- a/SRC/dlarrk.f
+++ b/SRC/dlarrk.f
@@ -120,12 +120,10 @@
 *>          INFO is INTEGER
 *>          = 0:       Eigenvalue converged
 *>          = -1:      Eigenvalue did NOT converge
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  FUDGE   DOUBLE PRECISION, default = 2
 *>          A "fudge factor" to widen the Gershgorin intervals.
 *> \endverbatim
diff --git a/SRC/dlartg.f b/SRC/dlartg.f
index ace79c84..82f301b6 100644
--- a/SRC/dlartg.f
+++ b/SRC/dlartg.f
@@ -78,8 +78,7 @@
 *> \verbatim
 *>          R is DOUBLE PRECISION
 *>          The nonzero component of the rotated vector.
-*> \endverbatim
-*> \verbatim
+*>
 *>  This version has a few statements commented out for thread safety
 *>  (machine parameters are computed on each entry). 10 feb 03, SJH.
 *> \endverbatim
diff --git a/SRC/dlartgp.f b/SRC/dlartgp.f
index fbd23621..7935ded5 100644
--- a/SRC/dlartgp.f
+++ b/SRC/dlartgp.f
@@ -76,8 +76,7 @@
 *> \verbatim
 *>          R is DOUBLE PRECISION
 *>          The nonzero component of the rotated vector.
-*> \endverbatim
-*> \verbatim
+*>
 *>  This version has a few statements commented out for thread safety
 *>  (machine parameters are computed on each entry). 10 feb 03, SJH.
 *> \endverbatim
diff --git a/SRC/dlascl.f b/SRC/dlascl.f
index c8be9939..f7748b47 100644
--- a/SRC/dlascl.f
+++ b/SRC/dlascl.f
@@ -86,8 +86,7 @@
 *> \param[in] CTO
 *> \verbatim
 *>          CTO is DOUBLE PRECISION
-*> \endverbatim
-*> \verbatim
+*>
 *>          The matrix A is multiplied by CTO/CFROM. A(I,J) is computed
 *>          without over/underflow if the final result CTO*A(I,J)/CFROM
 *>          can be represented without over/underflow.  CFROM must be
diff --git a/SRC/dlasd1.f b/SRC/dlasd1.f
index 4e2b6a06..d8626c2c 100644
--- a/SRC/dlasd1.f
+++ b/SRC/dlasd1.f
@@ -97,8 +97,7 @@
 *>          SQRE is INTEGER
 *>         = 0: the lower block is an NR-by-NR square matrix.
 *>         = 1: the lower block is an NR-by-(NR+1) rectangular matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>         The bidiagonal matrix has row dimension N = NL + NR + 1,
 *>         and column dimension M = N + SQRE.
 *> \endverbatim
diff --git a/SRC/dlasd2.f b/SRC/dlasd2.f
index 9cc0bfc2..0565d363 100644
--- a/SRC/dlasd2.f
+++ b/SRC/dlasd2.f
@@ -71,8 +71,7 @@
 *>          SQRE is INTEGER
 *>         = 0: the lower block is an NR-by-NR square matrix.
 *>         = 1: the lower block is an NR-by-(NR+1) rectangular matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>         The bidiagonal matrix has N = NL + NR + 1 rows and
 *>         M = N + SQRE >= N columns.
 *> \endverbatim
@@ -236,8 +235,7 @@
 *>         2 : non-zero in the lower half only
 *>         3 : dense
 *>         4 : deflated
-*> \endverbatim
-*> \verbatim
+*>
 *>         On exit, it is an array of dimension 4, with COLTYP(I) being
 *>         the dimension of the I-th type columns.
 *> \endverbatim
diff --git a/SRC/dlasd3.f b/SRC/dlasd3.f
index 6b25a93b..19045915 100644
--- a/SRC/dlasd3.f
+++ b/SRC/dlasd3.f
@@ -75,8 +75,7 @@
 *>          SQRE is INTEGER
 *>         = 0: the lower block is an NR-by-NR square matrix.
 *>         = 1: the lower block is an NR-by-(NR+1) rectangular matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>         The bidiagonal matrix has N = NL + NR + 1 rows and
 *>         M = N + SQRE >= N columns.
 *> \endverbatim
@@ -175,8 +174,7 @@
 *>         contains non-zero entries only at and below (or after) NL+2;
 *>         and the third is dense. The first column of U and the row of
 *>         VT are treated separately, however.
-*> \endverbatim
-*> \verbatim
+*>
 *>         The rows of the singular vectors found by DLASD4
 *>         must be likewise permuted before the matrix multiplies can
 *>         take place.
diff --git a/SRC/dlasd4.f b/SRC/dlasd4.f
index a577d632..79a0b0b1 100644
--- a/SRC/dlasd4.f
+++ b/SRC/dlasd4.f
@@ -114,24 +114,19 @@
 *>          INFO is INTEGER
 *>         = 0:  successful exit
 *>         > 0:  if INFO = 1, the updating process failed.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  Logical variable ORGATI (origin-at-i?) is used for distinguishing
 *>  whether D(i) or D(i+1) is treated as the origin.
-*> \endverbatim
-*> \verbatim
+*>
 *>            ORGATI = .true.    origin at i
 *>            ORGATI = .false.   origin at i+1
-*> \endverbatim
-*> \verbatim
+*>
 *>  Logical variable SWTCH3 (switch-for-3-poles?) is for noting
 *>  if we are working with THREE poles!
-*> \endverbatim
-*> \verbatim
+*>
 *>  MAXIT is the maximum number of iterations allowed for each
 *>  eigenvalue.
 *> \endverbatim
diff --git a/SRC/dlasd6.f b/SRC/dlasd6.f
index 5f6e0ff4..5f34d79d 100644
--- a/SRC/dlasd6.f
+++ b/SRC/dlasd6.f
@@ -118,8 +118,7 @@
 *>          SQRE is INTEGER
 *>         = 0: the lower block is an NR-by-NR square matrix.
 *>         = 1: the lower block is an NR-by-(NR+1) rectangular matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>         The bidiagonal matrix has row dimension N = NL + NR + 1,
 *>         and column dimension M = N + SQRE.
 *> \endverbatim
@@ -239,12 +238,10 @@
 *>         On exit, DIFR(I, 1) is the distance between I-th updated
 *>         (undeflated) singular value and the I+1-th (undeflated) old
 *>         singular value.
-*> \endverbatim
-*> \verbatim
+*>
 *>         If ICOMPQ = 1, DIFR(1:K,2) is an array containing the
 *>         normalizing factors for the right singular vector matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>         See DLASD8 for details on DIFL and DIFR.
 *> \endverbatim
 *>
diff --git a/SRC/dlasd7.f b/SRC/dlasd7.f
index dbee953a..de432ffa 100644
--- a/SRC/dlasd7.f
+++ b/SRC/dlasd7.f
@@ -83,8 +83,7 @@
 *>          SQRE is INTEGER
 *>         = 0: the lower block is an NR-by-NR square matrix.
 *>         = 1: the lower block is an NR-by-(NR+1) rectangular matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>         The bidiagonal matrix has
 *>         N = NL + NR + 1 rows and
 *>         M = N + SQRE >= N columns.
diff --git a/SRC/dlasd8.f b/SRC/dlasd8.f
index b2eade7c..4873c5a2 100644
--- a/SRC/dlasd8.f
+++ b/SRC/dlasd8.f
@@ -111,8 +111,7 @@
 *>                   dimension ( K ) if ICOMPQ = 0.
 *>          On exit, DIFR(I,1) = D(I) - DSIGMA(I+1), DIFR(K,1) is not
 *>          defined and will not be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If ICOMPQ = 1, DIFR(1:K,2) is an array containing the
 *>          normalizing factors for the right singular vector matrix.
 *> \endverbatim
diff --git a/SRC/dlasdq.f b/SRC/dlasdq.f
index 98f0f7a9..36b8c4c2 100644
--- a/SRC/dlasdq.f
+++ b/SRC/dlasdq.f
@@ -72,8 +72,7 @@
 *>        = 0: then the input matrix is N-by-N.
 *>        = 1: then the input matrix is N-by-(N+1) if UPLU = 'U' and
 *>             (N+1)-by-N if UPLU = 'L'.
-*> \endverbatim
-*> \verbatim
+*>
 *>        The bidiagonal matrix has
 *>        N = NL + NR + 1 rows and
 *>        M = N + SQRE >= N columns.
diff --git a/SRC/dlaset.f b/SRC/dlaset.f
index 873b4db5..12468694 100644
--- a/SRC/dlaset.f
+++ b/SRC/dlaset.f
@@ -82,13 +82,11 @@
 *> \verbatim
 *>          A is DOUBLE PRECISION array, dimension (LDA,N)
 *>          On exit, the leading m-by-n submatrix of A is set as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>          if UPLO = 'U', A(i,j) = ALPHA, 1<=i<=j-1, 1<=j<=n,
 *>          if UPLO = 'L', A(i,j) = ALPHA, j+1<=i<=m, 1<=j<=n,
 *>          otherwise,     A(i,j) = ALPHA, 1<=i<=m, 1<=j<=n, i.ne.j,
-*> \endverbatim
-*> \verbatim
+*>
 *>          and, for all UPLO, A(i,i) = BETA, 1<=i<=min(m,n).
 *> \endverbatim
 *>
diff --git a/SRC/dlasq3.f b/SRC/dlasq3.f
index 9660d528..2711c157 100644
--- a/SRC/dlasq3.f
+++ b/SRC/dlasq3.f
@@ -161,8 +161,7 @@
 *> \param[in,out] TAU
 *> \verbatim
 *>          TAU is DOUBLE PRECISION
-*> \endverbatim
-*> \verbatim
+*>
 *>         These are passed as arguments in order to save their values
 *>         between calls to DLASQ3.
 *> \endverbatim
diff --git a/SRC/dlasyf.f b/SRC/dlasyf.f
index 0476da3a..ae4b252b 100644
--- a/SRC/dlasyf.f
+++ b/SRC/dlasyf.f
@@ -112,8 +112,7 @@
 *>          Details of the interchanges and the block structure of D.
 *>          If UPLO = 'U', only the last KB elements of IPIV are set;
 *>          if UPLO = 'L', only the first KB elements are set.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If IPIV(k) > 0, then rows and columns k and IPIV(k) were
 *>          interchanged and D(k,k) is a 1-by-1 diagonal block.
 *>          If UPLO = 'U' and IPIV(k) = IPIV(k-1) < 0, then rows and
diff --git a/SRC/dlatbs.f b/SRC/dlatbs.f
index 5101b7fd..c24841f6 100644
--- a/SRC/dlatbs.f
+++ b/SRC/dlatbs.f
@@ -136,15 +136,13 @@
 *> \param[in,out] CNORM
 *> \verbatim
 *>          CNORM is or output) DOUBLE PRECISION array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'Y', CNORM is an input argument and CNORM(j)
 *>          contains the norm of the off-diagonal part of the j-th column
 *>          of A.  If TRANS = 'N', CNORM(j) must be greater than or equal
 *>          to the infinity-norm, and if TRANS = 'T' or 'C', CNORM(j)
 *>          must be greater than or equal to the 1-norm.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'N', CNORM is an output argument and CNORM(j)
 *>          returns the 1-norm of the offdiagonal part of the j-th column
 *>          of A.
diff --git a/SRC/dlatps.f b/SRC/dlatps.f
index c818f131..6b28c24e 100644
--- a/SRC/dlatps.f
+++ b/SRC/dlatps.f
@@ -123,15 +123,13 @@
 *> \param[in,out] CNORM
 *> \verbatim
 *>          CNORM is or output) DOUBLE PRECISION array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'Y', CNORM is an input argument and CNORM(j)
 *>          contains the norm of the off-diagonal part of the j-th column
 *>          of A.  If TRANS = 'N', CNORM(j) must be greater than or equal
 *>          to the infinity-norm, and if TRANS = 'T' or 'C', CNORM(j)
 *>          must be greater than or equal to the 1-norm.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'N', CNORM is an output argument and CNORM(j)
 *>          returns the 1-norm of the offdiagonal part of the j-th column
 *>          of A.
diff --git a/SRC/dlatrs.f b/SRC/dlatrs.f
index b0772c08..b347f825 100644
--- a/SRC/dlatrs.f
+++ b/SRC/dlatrs.f
@@ -132,15 +132,13 @@
 *> \param[in,out] CNORM
 *> \verbatim
 *>          CNORM is or output) DOUBLE PRECISION array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'Y', CNORM is an input argument and CNORM(j)
 *>          contains the norm of the off-diagonal part of the j-th column
 *>          of A.  If TRANS = 'N', CNORM(j) must be greater than or equal
 *>          to the infinity-norm, and if TRANS = 'T' or 'C', CNORM(j)
 *>          must be greater than or equal to the 1-norm.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'N', CNORM is an output argument and CNORM(j)
 *>          returns the 1-norm of the offdiagonal part of the j-th column
 *>          of A.
diff --git a/SRC/dlatzm.f b/SRC/dlatzm.f
index 121b2bc2..7bcc3c36 100644
--- a/SRC/dlatzm.f
+++ b/SRC/dlatzm.f
@@ -107,8 +107,7 @@
 *>                         (M,1)   if SIDE = 'R'
 *>          On entry, the n-vector C1 if SIDE = 'L', or the m-vector C1
 *>          if SIDE = 'R'.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the first row of P*C if SIDE = 'L', or the first
 *>          column of C*P if SIDE = 'R'.
 *> \endverbatim
@@ -120,8 +119,7 @@
 *>                         (LDC, N-1) if SIDE = 'R'
 *>          On entry, the (m - 1) x n matrix C2 if SIDE = 'L', or the
 *>          m x (n - 1) matrix C2 if SIDE = 'R'.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, rows 2:m of P*C if SIDE = 'L', or columns 2:m of C*P
 *>          if SIDE = 'R'.
 *> \endverbatim
diff --git a/SRC/dorbdb.f b/SRC/dorbdb.f
index f949f87a..a6aae378 100644
--- a/SRC/dorbdb.f
+++ b/SRC/dorbdb.f
@@ -234,8 +234,7 @@
 *> \verbatim
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK. LWORK >= M-Q.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dorcsd.f b/SRC/dorcsd.f
index 8916fb40..1345bbbb 100644
--- a/SRC/dorcsd.f
+++ b/SRC/dorcsd.f
@@ -255,8 +255,7 @@
 *> \verbatim
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the work array, and no error
@@ -275,12 +274,10 @@
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value.
 *>          > 0:  DBBCSD did not converge. See the description of WORK
 *>                above for details.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Reference
 *>  =========
-*> \endverbatim
-*> \verbatim
+*>
 *>  [1] Brian D. Sutton. Computing the complete CS decomposition. Numer.
 *>      Algorithms, 50(1):33-65, 2009.
 *> \endverbatim
diff --git a/SRC/dorgbr.f b/SRC/dorgbr.f
index f92e0545..c65bc073 100644
--- a/SRC/dorgbr.f
+++ b/SRC/dorgbr.f
@@ -129,8 +129,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,min(M,N)).
 *>          For optimum performance LWORK >= min(M,N)*NB, where NB
 *>          is the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dorghr.f b/SRC/dorghr.f
index 9e6dd87c..738ec8f8 100644
--- a/SRC/dorghr.f
+++ b/SRC/dorghr.f
@@ -58,8 +58,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          ILO and IHI must have the same values as in the previous call
 *>          of DGEHRD. Q is equal to the unit matrix except in the
 *>          submatrix Q(ilo+1:ihi,ilo+1:ihi).
@@ -99,8 +98,7 @@
 *>          The dimension of the array WORK. LWORK >= IHI-ILO.
 *>          For optimum performance LWORK >= (IHI-ILO)*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dorglq.f b/SRC/dorglq.f
index e9c42b1e..19d83458 100644
--- a/SRC/dorglq.f
+++ b/SRC/dorglq.f
@@ -99,8 +99,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,M).
 *>          For optimum performance LWORK >= M*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dorgql.f b/SRC/dorgql.f
index d34ad36e..2baeb533 100644
--- a/SRC/dorgql.f
+++ b/SRC/dorgql.f
@@ -100,8 +100,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is the
 *>          optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dorgqr.f b/SRC/dorgqr.f
index 9a6a0319..5c9b5bf1 100644
--- a/SRC/dorgqr.f
+++ b/SRC/dorgqr.f
@@ -100,8 +100,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is the
 *>          optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dorgrq.f b/SRC/dorgrq.f
index 1c8573c8..bfb319e1 100644
--- a/SRC/dorgrq.f
+++ b/SRC/dorgrq.f
@@ -100,8 +100,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,M).
 *>          For optimum performance LWORK >= M*NB, where NB is the
 *>          optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dorgtr.f b/SRC/dorgtr.f
index c0f2245a..dff9bf44 100644
--- a/SRC/dorgtr.f
+++ b/SRC/dorgtr.f
@@ -95,8 +95,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,N-1).
 *>          For optimum performance LWORK >= (N-1)*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dormbr.f b/SRC/dormbr.f
index fb310879..523e0959 100644
--- a/SRC/dormbr.f
+++ b/SRC/dormbr.f
@@ -166,8 +166,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dormhr.f b/SRC/dormhr.f
index 4dbb2d3e..c2a66452 100644
--- a/SRC/dormhr.f
+++ b/SRC/dormhr.f
@@ -86,8 +86,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          ILO and IHI must have the same values as in the previous call
 *>          of DGEHRD. Q is equal to the unit matrix except in the
 *>          submatrix Q(ilo+1:ihi,ilo+1:ihi).
@@ -150,8 +149,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dormlq.f b/SRC/dormlq.f
index 0b56dcd2..482b676c 100644
--- a/SRC/dormlq.f
+++ b/SRC/dormlq.f
@@ -141,8 +141,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dormql.f b/SRC/dormql.f
index 7501f2ba..68940442 100644
--- a/SRC/dormql.f
+++ b/SRC/dormql.f
@@ -141,8 +141,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dormqr.f b/SRC/dormqr.f
index 58517e03..011c5665 100644
--- a/SRC/dormqr.f
+++ b/SRC/dormqr.f
@@ -141,8 +141,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dormrq.f b/SRC/dormrq.f
index 73c17a7b..e4e1beb6 100644
--- a/SRC/dormrq.f
+++ b/SRC/dormrq.f
@@ -141,8 +141,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dormrz.f b/SRC/dormrz.f
index ae41d933..52d3e11b 100644
--- a/SRC/dormrz.f
+++ b/SRC/dormrz.f
@@ -149,8 +149,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dormtr.f b/SRC/dormtr.f
index 29488835..32852b62 100644
--- a/SRC/dormtr.f
+++ b/SRC/dormtr.f
@@ -142,8 +142,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dpbrfs.f b/SRC/dpbrfs.f
index c29c2a65..4440aaef 100644
--- a/SRC/dpbrfs.f
+++ b/SRC/dpbrfs.f
@@ -165,12 +165,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/dpbstf.f b/SRC/dpbstf.f
index 130ba513..4d7429ef 100644
--- a/SRC/dpbstf.f
+++ b/SRC/dpbstf.f
@@ -82,8 +82,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor S from the split Cholesky
 *>          factorization A = S**T*S. See Further Details.
 *> \endverbatim
diff --git a/SRC/dpbsv.f b/SRC/dpbsv.f
index 6a5fbd08..b22b8f8f 100644
--- a/SRC/dpbsv.f
+++ b/SRC/dpbsv.f
@@ -90,8 +90,7 @@
 *>          if UPLO = 'U', AB(KD+1+i-j,j) = A(i,j) for max(1,j-KD)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(N,j+KD).
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**T*U or A = L*L**T of the band
 *>          matrix A, in the same storage format as A.
diff --git a/SRC/dpbsvx.f b/SRC/dpbsvx.f
index 9483a7b9..bb1cf552 100644
--- a/SRC/dpbsvx.f
+++ b/SRC/dpbsvx.f
@@ -146,8 +146,7 @@
 *>          if UPLO = 'U', AB(KD+1+i-j,j) = A(i,j) for max(1,j-KD)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(N,j+KD).
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>          diag(S)*A*diag(S).
 *> \endverbatim
@@ -166,13 +165,11 @@
 *>          factorization A = U**T*U or A = L*L**T of the band matrix
 *>          A, in the same storage format as A (see AB).  If EQUED = 'Y',
 *>          then AFB is the factored form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AFB is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**T*U or A = L*L**T.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then AFB is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**T*U or A = L*L**T of the equilibrated
diff --git a/SRC/dpbtf2.f b/SRC/dpbtf2.f
index cafc15d2..898be3f4 100644
--- a/SRC/dpbtf2.f
+++ b/SRC/dpbtf2.f
@@ -81,8 +81,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**T*U or A = L*L**T of the band
 *>          matrix A, in the same storage format as A.
diff --git a/SRC/dpbtrf.f b/SRC/dpbtrf.f
index 0ac4dddd..575e5de3 100644
--- a/SRC/dpbtrf.f
+++ b/SRC/dpbtrf.f
@@ -76,8 +76,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**T*U or A = L*L**T of the band
 *>          matrix A, in the same storage format as A.
diff --git a/SRC/dpftrf.f b/SRC/dpftrf.f
index ccef6e09..477bbb06 100644
--- a/SRC/dpftrf.f
+++ b/SRC/dpftrf.f
@@ -82,8 +82,7 @@
 *>          of lower packed A. The LDA of RFP A is (N+1)/2 when TRANSR =
 *>          'T'. When TRANSR is 'N' the LDA is N+1 when N is even and N
 *>          is odd. See the Note below for more details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization RFP A = U**T*U or RFP A = L*L**T.
 *> \endverbatim
diff --git a/SRC/dpftri.f b/SRC/dpftri.f
index 0d2f65c7..f173f419 100644
--- a/SRC/dpftri.f
+++ b/SRC/dpftri.f
@@ -76,8 +76,7 @@
 *>          of lower packed A. The LDA of RFP A is (N+1)/2 when TRANSR =
 *>          'T'. When TRANSR is 'N' the LDA is N+1 when N is even and N
 *>          is odd. See the Note below for more details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the symmetric inverse of the original matrix, in the
 *>          same storage format.
 *> \endverbatim
diff --git a/SRC/dporfs.f b/SRC/dporfs.f
index 43f99569..5a213ce7 100644
--- a/SRC/dporfs.f
+++ b/SRC/dporfs.f
@@ -159,12 +159,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/dporfsx.f b/SRC/dporfsx.f
index 941cc747..3772991e 100644
--- a/SRC/dporfsx.f
+++ b/SRC/dporfsx.f
@@ -211,37 +211,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -250,8 +244,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -262,14 +255,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -277,26 +268,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -307,8 +294,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -327,8 +313,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0D+0
@@ -339,8 +324,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -350,8 +334,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/dposv.f b/SRC/dposv.f
index 427d979d..1e8963fd 100644
--- a/SRC/dposv.f
+++ b/SRC/dposv.f
@@ -82,8 +82,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization A = U**T*U or A = L*L**T.
 *> \endverbatim
diff --git a/SRC/dposvx.f b/SRC/dposvx.f
index 110cc2b5..55dff5b1 100644
--- a/SRC/dposvx.f
+++ b/SRC/dposvx.f
@@ -141,8 +141,7 @@
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.  A is not modified if
 *>          FACT = 'F' or 'N', or if FACT = 'E' and EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>          diag(S)*A*diag(S).
 *> \endverbatim
@@ -161,14 +160,12 @@
 *>          factorization A = U**T*U or A = L*L**T, in the same storage
 *>          format as A.  If EQUED .ne. 'N', then AF is the factored form
 *>          of the equilibrated matrix diag(S)*A*diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AF is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**T*U or A = L*L**T of the original
 *>          matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then AF is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**T*U or A = L*L**T of the equilibrated
diff --git a/SRC/dposvxx.f b/SRC/dposvxx.f
index 3b72faa6..8c3ff0fb 100644
--- a/SRC/dposvxx.f
+++ b/SRC/dposvxx.f
@@ -168,8 +168,7 @@
 *>     the strictly upper triangular part of A is not referenced.  A is
 *>     not modified if FACT = 'F' or 'N', or if FACT = 'E' and EQUED =
 *>     'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>     On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>     diag(S)*A*diag(S).
 *> \endverbatim
@@ -188,14 +187,12 @@
 *>     factorization A = U**T*U or A = L*L**T, in the same storage
 *>     format as A.  If EQUED .ne. 'N', then AF is the factored
 *>     form of the equilibrated matrix diag(S)*A*diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then AF is an output argument and on exit
 *>     returns the triangular factor U or L from the Cholesky
 *>     factorization A = U**T*U or A = L*L**T of the original
 *>     matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'E', then AF is an output argument and on exit
 *>     returns the triangular factor U or L from the Cholesky
 *>     factorization A = U**T*U or A = L*L**T of the equilibrated
@@ -314,37 +311,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -353,8 +344,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -365,14 +355,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -380,26 +368,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -410,8 +394,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -430,8 +413,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0D+0
@@ -439,8 +421,7 @@
 *>                    computed.
 *>            = 1.0 : Use the extra-precise refinement algorithm.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -450,8 +431,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/dpotf2.f b/SRC/dpotf2.f
index 0658a4d9..4604659e 100644
--- a/SRC/dpotf2.f
+++ b/SRC/dpotf2.f
@@ -74,8 +74,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization A = U**T *U  or A = L*L**T.
 *> \endverbatim
diff --git a/SRC/dpotrf.f b/SRC/dpotrf.f
index 4feb0539..0e7010d1 100644
--- a/SRC/dpotrf.f
+++ b/SRC/dpotrf.f
@@ -72,8 +72,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization A = U**T*U or A = L*L**T.
 *> \endverbatim
diff --git a/SRC/dpprfs.f b/SRC/dpprfs.f
index 4735169f..1f0c58c0 100644
--- a/SRC/dpprfs.f
+++ b/SRC/dpprfs.f
@@ -147,12 +147,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/dppsv.f b/SRC/dppsv.f
index 9369955a..bea94b3e 100644
--- a/SRC/dppsv.f
+++ b/SRC/dppsv.f
@@ -81,8 +81,7 @@
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization A = U**T*U or A = L*L**T, in the same storage
 *>          format as A.
diff --git a/SRC/dppsvx.f b/SRC/dppsvx.f
index 2d782f2e..a4031f9e 100644
--- a/SRC/dppsvx.f
+++ b/SRC/dppsvx.f
@@ -138,8 +138,7 @@
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
 *>          See below for further details.  A is not modified if
 *>          FACT = 'F' or 'N', or if FACT = 'E' and EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>          diag(S)*A*diag(S).
 *> \endverbatim
@@ -153,14 +152,12 @@
 *>          factorization A = U**T*U or A = L*L**T, in the same storage
 *>          format as A.  If EQUED .ne. 'N', then AFP is the factored
 *>          form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AFP is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**T * U or A = L * L**T of the original
 *>          matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then AFP is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**T * U or A = L * L**T of the equilibrated
diff --git a/SRC/dpptrf.f b/SRC/dpptrf.f
index 5f629567..e6f05e6c 100644
--- a/SRC/dpptrf.f
+++ b/SRC/dpptrf.f
@@ -69,8 +69,7 @@
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**T*U or A = L*L**T, in the same
 *>          storage format as A.
diff --git a/SRC/dpptri.f b/SRC/dpptri.f
index 90b94e17..10f77ce0 100644
--- a/SRC/dpptri.f
+++ b/SRC/dpptri.f
@@ -65,8 +65,7 @@
 *>          array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = U(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = L(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the upper or lower triangle of the (symmetric)
 *>          inverse of A, overwriting the input factor U or L.
 *> \endverbatim
diff --git a/SRC/dpstf2.f b/SRC/dpstf2.f
index 5566589a..feae5e92 100644
--- a/SRC/dpstf2.f
+++ b/SRC/dpstf2.f
@@ -78,8 +78,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization as above.
 *> \endverbatim
diff --git a/SRC/dpstrf.f b/SRC/dpstrf.f
index ebcd76cc..0842d78b 100644
--- a/SRC/dpstrf.f
+++ b/SRC/dpstrf.f
@@ -78,8 +78,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization as above.
 *> \endverbatim
diff --git a/SRC/dptrfs.f b/SRC/dptrfs.f
index 00543716..cbbcdd3a 100644
--- a/SRC/dptrfs.f
+++ b/SRC/dptrfs.f
@@ -139,12 +139,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/dsbev.f b/SRC/dsbev.f
index 3971a023..dc490fcf 100644
--- a/SRC/dsbev.f
+++ b/SRC/dsbev.f
@@ -79,8 +79,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AB is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the first
 *>          superdiagonal and the diagonal of the tridiagonal matrix T
diff --git a/SRC/dsbevd.f b/SRC/dsbevd.f
index 6ff74c95..c4e6aa90 100644
--- a/SRC/dsbevd.f
+++ b/SRC/dsbevd.f
@@ -88,8 +88,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AB is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the first
 *>          superdiagonal and the diagonal of the tridiagonal matrix T
@@ -141,8 +140,7 @@
 *>          If JOBZ  = 'N' and N > 2, LWORK must be at least 2*N.
 *>          If JOBZ  = 'V' and N > 2, LWORK must be at least
 *>                         ( 1 + 5*N + 2*N**2 ).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK and IWORK
 *>          arrays, returns these values as the first entries of the WORK
@@ -162,8 +160,7 @@
 *>          The dimension of the array LIWORK.
 *>          If JOBZ  = 'N' or N <= 1, LIWORK must be at least 1.
 *>          If JOBZ  = 'V' and N > 2, LIWORK must be at least 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK and
 *>          IWORK arrays, returns these values as the first entries of
diff --git a/SRC/dsbevx.f b/SRC/dsbevx.f
index 8ab0d415..1bd8403c 100644
--- a/SRC/dsbevx.f
+++ b/SRC/dsbevx.f
@@ -94,8 +94,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AB is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the first
 *>          superdiagonal and the diagonal of the tridiagonal matrix T
@@ -159,24 +158,20 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing AB to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*DLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
 *>          eigenvectors did not converge, try setting ABSTOL to
 *>          2*DLAMCH('S').
-*> \endverbatim
-*> \verbatim
+*>
 *>          See "Computing Small Singular Values of Bidiagonal Matrices
 *>          with Guaranteed High Relative Accuracy," by Demmel and
 *>          Kahan, LAPACK Working Note #3.
diff --git a/SRC/dsbgst.f b/SRC/dsbgst.f
index 096f5c5e..d7974b6c 100644
--- a/SRC/dsbgst.f
+++ b/SRC/dsbgst.f
@@ -93,8 +93,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(ka+1+i-j,j) = A(i,j) for max(1,j-ka)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+ka).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the transformed matrix X**T*A*X, stored in the same
 *>          format as A.
 *> \endverbatim
diff --git a/SRC/dsbgv.f b/SRC/dsbgv.f
index 7400e2b3..96be241d 100644
--- a/SRC/dsbgv.f
+++ b/SRC/dsbgv.f
@@ -89,8 +89,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(ka+1+i-j,j) = A(i,j) for max(1,j-ka)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+ka).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AB are destroyed.
 *> \endverbatim
 *>
@@ -109,8 +108,7 @@
 *>          as follows:
 *>          if UPLO = 'U', BB(kb+1+i-j,j) = B(i,j) for max(1,j-kb)<=i<=j;
 *>          if UPLO = 'L', BB(1+i-j,j)    = B(i,j) for j<=i<=min(n,j+kb).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the factor S from the split Cholesky factorization
 *>          B = S**T*S, as returned by DPBSTF.
 *> \endverbatim
diff --git a/SRC/dsbgvd.f b/SRC/dsbgvd.f
index 555a2d15..901e7cda 100644
--- a/SRC/dsbgvd.f
+++ b/SRC/dsbgvd.f
@@ -98,8 +98,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(ka+1+i-j,j) = A(i,j) for max(1,j-ka)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+ka).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AB are destroyed.
 *> \endverbatim
 *>
@@ -118,8 +117,7 @@
 *>          as follows:
 *>          if UPLO = 'U', BB(ka+1+i-j,j) = B(i,j) for max(1,j-kb)<=i<=j;
 *>          if UPLO = 'L', BB(1+i-j,j)    = B(i,j) for j<=i<=min(n,j+kb).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the factor S from the split Cholesky factorization
 *>          B = S**T*S, as returned by DPBSTF.
 *> \endverbatim
@@ -166,8 +164,7 @@
 *>          If N <= 1,               LWORK >= 1.
 *>          If JOBZ = 'N' and N > 1, LWORK >= 3*N.
 *>          If JOBZ = 'V' and N > 1, LWORK >= 1 + 5*N + 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK and IWORK
 *>          arrays, returns these values as the first entries of the WORK
@@ -187,8 +184,7 @@
 *>          The dimension of the array IWORK.
 *>          If JOBZ  = 'N' or N <= 1, LIWORK >= 1.
 *>          If JOBZ  = 'V' and N > 1, LIWORK >= 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK and
 *>          IWORK arrays, returns these values as the first entries of
diff --git a/SRC/dsbgvx.f b/SRC/dsbgvx.f
index 55d4c88b..1b783710 100644
--- a/SRC/dsbgvx.f
+++ b/SRC/dsbgvx.f
@@ -104,8 +104,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(ka+1+i-j,j) = A(i,j) for max(1,j-ka)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+ka).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AB are destroyed.
 *> \endverbatim
 *>
@@ -124,8 +123,7 @@
 *>          as follows:
 *>          if UPLO = 'U', BB(ka+1+i-j,j) = B(i,j) for max(1,j-kb)<=i<=j;
 *>          if UPLO = 'L', BB(1+i-j,j)    = B(i,j) for j<=i<=min(n,j+kb).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the factor S from the split Cholesky factorization
 *>          B = S**T*S, as returned by DPBSTF.
 *> \endverbatim
@@ -160,8 +158,7 @@
 *> \param[in] VU
 *> \verbatim
 *>          VU is DOUBLE PRECISION
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='V', the lower and upper bounds of the interval to
 *>          be searched for eigenvalues. VL < VU.
 *>          Not referenced if RANGE = 'A' or 'I'.
@@ -175,8 +172,7 @@
 *> \param[in] IU
 *> \verbatim
 *>          IU is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='I', the indices (in ascending order) of the
 *>          smallest and largest eigenvalues to be returned.
 *>          1 <= IL <= IU <= N, if N > 0; IL = 1 and IU = 0 if N = 0.
@@ -190,17 +186,14 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing A to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*DLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
diff --git a/SRC/dsbtrd.f b/SRC/dsbtrd.f
index 094f0296..b97a9076 100644
--- a/SRC/dsbtrd.f
+++ b/SRC/dsbtrd.f
@@ -114,8 +114,7 @@
 *>          Q is DOUBLE PRECISION array, dimension (LDQ,N)
 *>          On entry, if VECT = 'U', then Q must contain an N-by-N
 *>          matrix X; if VECT = 'N' or 'V', then Q need not be set.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit:
 *>          if VECT = 'V', Q contains the N-by-N orthogonal matrix Q;
 *>          if VECT = 'U', Q contains the product X*Q;
diff --git a/SRC/dsfrk.f b/SRC/dsfrk.f
index c0532648..fcd7555a 100644
--- a/SRC/dsfrk.f
+++ b/SRC/dsfrk.f
@@ -68,16 +68,13 @@
 *>           On  entry, UPLO specifies whether the upper or lower
 *>           triangular part of the array C is to be referenced as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'U' or 'u'   Only the upper triangular part of C
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'L' or 'l'   Only the lower triangular part of C
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -86,14 +83,11 @@
 *>          TRANS is CHARACTER*1
 *>           On entry, TRANS specifies the operation to be performed as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              TRANS = 'N' or 'n'   C := alpha*A*A**T + beta*C.
-*> \endverbatim
-*> \verbatim
+*>
 *>              TRANS = 'T' or 't'   C := alpha*A**T*A + beta*C.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/dspev.f b/SRC/dspev.f
index dcb978c5..484f50cc 100644
--- a/SRC/dspev.f
+++ b/SRC/dspev.f
@@ -70,8 +70,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AP is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the diagonal
 *>          and first superdiagonal of the tridiagonal matrix T overwrite
diff --git a/SRC/dspevd.f b/SRC/dspevd.f
index b57b1a67..679a65db 100644
--- a/SRC/dspevd.f
+++ b/SRC/dspevd.f
@@ -80,8 +80,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AP is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the diagonal
 *>          and first superdiagonal of the tridiagonal matrix T overwrite
@@ -127,8 +126,7 @@
 *>          If JOBZ = 'N' and N > 1, LWORK must be at least 2*N.
 *>          If JOBZ = 'V' and N > 1, LWORK must be at least
 *>                                                 1 + 6*N + N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the required sizes of the WORK and IWORK
 *>          arrays, returns these values as the first entries of the WORK
@@ -148,8 +146,7 @@
 *>          The dimension of the array IWORK.
 *>          If JOBZ  = 'N' or N <= 1, LIWORK must be at least 1.
 *>          If JOBZ  = 'V' and N > 1, LIWORK must be at least 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the required sizes of the WORK and
 *>          IWORK arrays, returns these values as the first entries of
diff --git a/SRC/dspevx.f b/SRC/dspevx.f
index e7b6452b..9d31549f 100644
--- a/SRC/dspevx.f
+++ b/SRC/dspevx.f
@@ -85,8 +85,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AP is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the diagonal
 *>          and first superdiagonal of the tridiagonal matrix T overwrite
@@ -129,24 +128,20 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing AP to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*DLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
 *>          eigenvectors did not converge, try setting ABSTOL to
 *>          2*DLAMCH('S').
-*> \endverbatim
-*> \verbatim
+*>
 *>          See "Computing Small Singular Values of Bidiagonal Matrices
 *>          with Guaranteed High Relative Accuracy," by Demmel and
 *>          Kahan, LAPACK Working Note #3.
diff --git a/SRC/dspgst.f b/SRC/dspgst.f
index 89d2199d..d9e7f70a 100644
--- a/SRC/dspgst.f
+++ b/SRC/dspgst.f
@@ -80,8 +80,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the transformed matrix, stored in the
 *>          same format as A.
 *> \endverbatim
diff --git a/SRC/dspgv.f b/SRC/dspgv.f
index 8223a5a4..74c7b5d2 100644
--- a/SRC/dspgv.f
+++ b/SRC/dspgv.f
@@ -85,8 +85,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AP are destroyed.
 *> \endverbatim
 *>
@@ -98,8 +97,7 @@
 *>          is stored in the array BP as follows:
 *>          if UPLO = 'U', BP(i + (j-1)*j/2) = B(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', BP(i + (j-1)*(2*n-j)/2) = B(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the triangular factor U or L from the Cholesky
 *>          factorization B = U**T*U or B = L*L**T, in the same storage
 *>          format as B.
diff --git a/SRC/dspgvd.f b/SRC/dspgvd.f
index 0694561f..fb7d3e91 100644
--- a/SRC/dspgvd.f
+++ b/SRC/dspgvd.f
@@ -93,8 +93,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AP are destroyed.
 *> \endverbatim
 *>
@@ -106,8 +105,7 @@
 *>          is stored in the array BP as follows:
 *>          if UPLO = 'U', BP(i + (j-1)*j/2) = B(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', BP(i + (j-1)*(2*n-j)/2) = B(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the triangular factor U or L from the Cholesky
 *>          factorization B = U**T*U or B = L*L**T, in the same storage
 *>          format as B.
@@ -149,8 +147,7 @@
 *>          If N <= 1,               LWORK >= 1.
 *>          If JOBZ = 'N' and N > 1, LWORK >= 2*N.
 *>          If JOBZ = 'V' and N > 1, LWORK >= 1 + 6*N + 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the required sizes of the WORK and IWORK
 *>          arrays, returns these values as the first entries of the WORK
@@ -170,8 +167,7 @@
 *>          The dimension of the array IWORK.
 *>          If JOBZ  = 'N' or N <= 1, LIWORK >= 1.
 *>          If JOBZ  = 'V' and N > 1, LIWORK >= 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the required sizes of the WORK and
 *>          IWORK arrays, returns these values as the first entries of
diff --git a/SRC/dspgvx.f b/SRC/dspgvx.f
index f83eb174..178e0d1c 100644
--- a/SRC/dspgvx.f
+++ b/SRC/dspgvx.f
@@ -98,8 +98,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AP are destroyed.
 *> \endverbatim
 *>
@@ -111,8 +110,7 @@
 *>          is stored in the array BP as follows:
 *>          if UPLO = 'U', BP(i + (j-1)*j/2) = B(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', BP(i + (j-1)*(2*n-j)/2) = B(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the triangular factor U or L from the Cholesky
 *>          factorization B = U**T*U or B = L*L**T, in the same storage
 *>          format as B.
@@ -126,8 +124,7 @@
 *> \param[in] VU
 *> \verbatim
 *>          VU is DOUBLE PRECISION
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='V', the lower and upper bounds of the interval to
 *>          be searched for eigenvalues. VL < VU.
 *>          Not referenced if RANGE = 'A' or 'I'.
@@ -141,8 +138,7 @@
 *> \param[in] IU
 *> \verbatim
 *>          IU is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='I', the indices (in ascending order) of the
 *>          smallest and largest eigenvalues to be returned.
 *>          1 <= IL <= IU <= N, if N > 0; IL = 1 and IU = 0 if N = 0.
@@ -156,17 +152,14 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing A to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*DLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
@@ -199,8 +192,7 @@
 *>          The eigenvectors are normalized as follows:
 *>          if ITYPE = 1 or 2, Z**T*B*Z = I;
 *>          if ITYPE = 3, Z**T*inv(B)*Z = I.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If an eigenvector fails to converge, then that column of Z
 *>          contains the latest approximation to the eigenvector, and the
 *>          index of the eigenvector is returned in IFAIL.
diff --git a/SRC/dsprfs.f b/SRC/dsprfs.f
index e00f4af4..1641fe26 100644
--- a/SRC/dsprfs.f
+++ b/SRC/dsprfs.f
@@ -155,12 +155,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/dspsv.f b/SRC/dspsv.f
index a80049f2..e55c4b50 100644
--- a/SRC/dspsv.f
+++ b/SRC/dspsv.f
@@ -83,8 +83,7 @@
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
 *>          A = U*D*U**T or A = L*D*L**T as computed by DSPTRF, stored as
diff --git a/SRC/dspsvx.f b/SRC/dspsvx.f
index 33256f38..865492eb 100644
--- a/SRC/dspsvx.f
+++ b/SRC/dspsvx.f
@@ -128,8 +128,7 @@
 *>          to obtain the factor U or L from the factorization
 *>          A = U*D*U**T or A = L*D*L**T as computed by DSPTRF, stored as
 *>          a packed triangular matrix in the same storage format as A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AFP is an output argument and on exit
 *>          contains the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
@@ -150,8 +149,7 @@
 *>          is a 2-by-2 diagonal block.  If UPLO = 'L' and IPIV(k) =
 *>          IPIV(k+1) < 0, then rows and columns k+1 and -IPIV(k) were
 *>          interchanged and D(k:k+1,k:k+1) is a 2-by-2 diagonal block.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains details of the interchanges and the block structure
 *>          of D, as determined by DSPTRF.
diff --git a/SRC/dsptrf.f b/SRC/dsptrf.f
index 2e1c25f1..b5c420d9 100644
--- a/SRC/dsptrf.f
+++ b/SRC/dsptrf.f
@@ -70,8 +70,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L, stored as a packed triangular
 *>          matrix overwriting A (see below for further details).
diff --git a/SRC/dsptri.f b/SRC/dsptri.f
index a9a8b7e4..03a8dffa 100644
--- a/SRC/dsptri.f
+++ b/SRC/dsptri.f
@@ -65,8 +65,7 @@
 *>          On entry, the block diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by DSPTRF,
 *>          stored as a packed triangular matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix, stored as a packed triangular matrix. The j-th column
 *>          of inv(A) is stored in the array AP as follows:
diff --git a/SRC/dstebz.f b/SRC/dstebz.f
index 095ea368..a8933108 100644
--- a/SRC/dstebz.f
+++ b/SRC/dstebz.f
@@ -93,8 +93,7 @@
 *> \param[in] VU
 *> \verbatim
 *>          VU is DOUBLE PRECISION
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='V', the lower and upper bounds of the interval to
 *>          be searched for eigenvalues.  Eigenvalues less than or equal
 *>          to VL, or greater than VU, will not be returned.  VL < VU.
@@ -109,8 +108,7 @@
 *> \param[in] IU
 *> \verbatim
 *>          IU is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='I', the indices (in ascending order) of the
 *>          smallest and largest eigenvalues to be returned.
 *>          1 <= IL <= IU <= N, if N > 0; IL = 1 and IU = 0 if N = 0.
@@ -125,8 +123,7 @@
 *>          determined to lie in an interval whose width is ABSTOL or
 *>          less.  If ABSTOL is less than or equal to zero, then ULP*|T|
 *>          will be used, where |T| means the 1-norm of T.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*DLAMCH('S'), not zero.
 *> \endverbatim
@@ -229,19 +226,16 @@
 *>                                        floating-point arithmetic.
 *>                        Cure: Increase the PARAMETER "FUDGE",
 *>                              recompile, and try again.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  RELFAC  DOUBLE PRECISION, default = 2.0e0
 *>          The relative tolerance.  An interval (a,b] lies within
 *>          "relative tolerance" if  b-a < RELFAC*ulp*max(|a|,|b|),
 *>          where "ulp" is the machine precision (distance from 1 to
 *>          the next larger floating point number.)
-*> \endverbatim
-*> \verbatim
+*>
 *>  FUDGE   DOUBLE PRECISION, default = 2
 *>          A "fudge factor" to widen the Gershgorin intervals.  Ideally,
 *>          a value of 1 should work, but on machines with sloppy
diff --git a/SRC/dstedc.f b/SRC/dstedc.f
index a689ac49..541b1229 100644
--- a/SRC/dstedc.f
+++ b/SRC/dstedc.f
@@ -125,8 +125,7 @@
 *>          Note that for COMPZ = 'I' or 'V', then if N is less than or
 *>          equal to the minimum divide size, usually 25, then LWORK need
 *>          only be max(1,2*(N-1)).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
@@ -151,8 +150,7 @@
 *>          Note that for COMPZ = 'I' or 'V', then if N is less than or
 *>          equal to the minimum divide size, usually 25, then LIWORK
 *>          need only be 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal size of the IWORK array,
 *>          returns this value as the first entry of the IWORK array, and
diff --git a/SRC/dstegr.f b/SRC/dstegr.f
index 1d1d1b72..e5f58353 100644
--- a/SRC/dstegr.f
+++ b/SRC/dstegr.f
@@ -111,8 +111,7 @@
 *> \param[in] VU
 *> \verbatim
 *>          VU is DOUBLE PRECISION
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='V', the lower and upper bounds of the interval to
 *>          be searched for eigenvalues. VL < VU.
 *>          Not referenced if RANGE = 'A' or 'I'.
@@ -126,8 +125,7 @@
 *> \param[in] IU
 *> \verbatim
 *>          IU is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='I', the indices (in ascending order) of the
 *>          smallest and largest eigenvalues to be returned.
 *>          1 <= IL <= IU <= N, if N > 0.
diff --git a/SRC/dstein.f b/SRC/dstein.f
index c4250de5..34abb7f6 100644
--- a/SRC/dstein.f
+++ b/SRC/dstein.f
@@ -145,16 +145,13 @@
 *>          > 0: if INFO = i, then i eigenvectors failed to converge
 *>               in MAXITS iterations.  Their indices are stored in
 *>               array IFAIL.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  MAXITS  INTEGER, default = 5
 *>          The maximum number of iterations performed.
-*> \endverbatim
-*> \verbatim
+*>
 *>  EXTRA   INTEGER, default = 2
 *>          The number of iterations performed after norm growth
 *>          criterion is satisfied, should be at least 1.
diff --git a/SRC/dstemr.f b/SRC/dstemr.f
index c0513d6e..ad044dbe 100644
--- a/SRC/dstemr.f
+++ b/SRC/dstemr.f
@@ -142,8 +142,7 @@
 *> \param[in] VU
 *> \verbatim
 *>          VU is DOUBLE PRECISION
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='V', the lower and upper bounds of the interval to
 *>          be searched for eigenvalues. VL < VU.
 *>          Not referenced if RANGE = 'A' or 'I'.
@@ -157,8 +156,7 @@
 *> \param[in] IU
 *> \verbatim
 *>          IU is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='I', the indices (in ascending order) of the
 *>          smallest and largest eigenvalues to be returned.
 *>          1 <= IL <= IU <= N, if N > 0.
diff --git a/SRC/dstevd.f b/SRC/dstevd.f
index 396910b2..31f98ac5 100644
--- a/SRC/dstevd.f
+++ b/SRC/dstevd.f
@@ -111,8 +111,7 @@
 *>          If JOBZ  = 'N' or N <= 1 then LWORK must be at least 1.
 *>          If JOBZ  = 'V' and N > 1 then LWORK must be at least
 *>                         ( 1 + 4*N + N**2 ).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK and IWORK
 *>          arrays, returns these values as the first entries of the WORK
@@ -132,8 +131,7 @@
 *>          The dimension of the array IWORK.
 *>          If JOBZ  = 'N' or N <= 1 then LIWORK must be at least 1.
 *>          If JOBZ  = 'V' and N > 1 then LIWORK must be at least 3+5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK and
 *>          IWORK arrays, returns these values as the first entries of
diff --git a/SRC/dstevr.f b/SRC/dstevr.f
index 54cc96da..a266fbd2 100644
--- a/SRC/dstevr.f
+++ b/SRC/dstevr.f
@@ -160,22 +160,18 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing A to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          See "Computing Small Singular Values of Bidiagonal Matrices
 *>          with Guaranteed High Relative Accuracy," by Demmel and
 *>          Kahan, LAPACK Working Note #3.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If high relative accuracy is important, set ABSTOL to
 *>          DLAMCH( 'Safe minimum' ).  Doing so will guarantee that
 *>          eigenvalues are computed to high relative accuracy when
@@ -242,8 +238,7 @@
 *> \verbatim
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.  LWORK >= max(1,20*N).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK and IWORK
 *>          arrays, returns these values as the first entries of the WORK
@@ -262,8 +257,7 @@
 *> \verbatim
 *>          LIWORK is INTEGER
 *>          The dimension of the array IWORK.  LIWORK >= max(1,10*N).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK and
 *>          IWORK arrays, returns these values as the first entries of
diff --git a/SRC/dstevx.f b/SRC/dstevx.f
index b00e0b39..af02662d 100644
--- a/SRC/dstevx.f
+++ b/SRC/dstevx.f
@@ -121,24 +121,20 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less
 *>          than or equal to zero, then  EPS*|T|  will be used in
 *>          its place, where |T| is the 1-norm of the tridiagonal
 *>          matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*DLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
 *>          eigenvectors did not converge, try setting ABSTOL to
 *>          2*DLAMCH('S').
-*> \endverbatim
-*> \verbatim
+*>
 *>          See "Computing Small Singular Values of Bidiagonal Matrices
 *>          with Guaranteed High Relative Accuracy," by Demmel and
 *>          Kahan, LAPACK Working Note #3.
diff --git a/SRC/dsyev.f b/SRC/dsyev.f
index 8b962c27..d30d4d0f 100644
--- a/SRC/dsyev.f
+++ b/SRC/dsyev.f
@@ -101,8 +101,7 @@
 *>          The length of the array WORK.  LWORK >= max(1,3*N-1).
 *>          For optimal efficiency, LWORK >= (NB+2)*N,
 *>          where NB is the blocksize for DSYTRD returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dsyevd.f b/SRC/dsyevd.f
index 2be0f4fe..7a8b139e 100644
--- a/SRC/dsyevd.f
+++ b/SRC/dsyevd.f
@@ -117,8 +117,7 @@
 *>          If JOBZ = 'N' and N > 1, LWORK must be at least 2*N+1.
 *>          If JOBZ = 'V' and N > 1, LWORK must be at least
 *>                                                1 + 6*N + 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK and IWORK
 *>          arrays, returns these values as the first entries of the WORK
@@ -139,8 +138,7 @@
 *>          If N <= 1,                LIWORK must be at least 1.
 *>          If JOBZ  = 'N' and N > 1, LIWORK must be at least 1.
 *>          If JOBZ  = 'V' and N > 1, LIWORK must be at least 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK and
 *>          IWORK arrays, returns these values as the first entries of
diff --git a/SRC/dsyevr.f b/SRC/dsyevr.f
index 266d0392..76833157 100644
--- a/SRC/dsyevr.f
+++ b/SRC/dsyevr.f
@@ -185,22 +185,18 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing A to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          See "Computing Small Singular Values of Bidiagonal Matrices
 *>          with Guaranteed High Relative Accuracy," by Demmel and
 *>          Kahan, LAPACK Working Note #3.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If high relative accuracy is important, set ABSTOL to
 *>          DLAMCH( 'Safe minimum' ).  Doing so will guarantee that
 *>          eigenvalues are computed to high relative accuracy when
@@ -271,8 +267,7 @@
 *>          For optimal efficiency, LWORK >= (NB+6)*N,
 *>          where NB is the max of the blocksize for DSYTRD and DORMTR
 *>          returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
@@ -289,8 +284,7 @@
 *> \verbatim
 *>          LIWORK is INTEGER
 *>          The dimension of the array IWORK.  LIWORK >= max(1,10*N).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal size of the IWORK array,
 *>          returns this value as the first entry of the IWORK array, and
diff --git a/SRC/dsyevx.f b/SRC/dsyevx.f
index a9698052..90861c04 100644
--- a/SRC/dsyevx.f
+++ b/SRC/dsyevx.f
@@ -130,24 +130,20 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing A to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*DLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
 *>          eigenvectors did not converge, try setting ABSTOL to
 *>          2*DLAMCH('S').
-*> \endverbatim
-*> \verbatim
+*>
 *>          See "Computing Small Singular Values of Bidiagonal Matrices
 *>          with Guaranteed High Relative Accuracy," by Demmel and
 *>          Kahan, LAPACK Working Note #3.
@@ -204,8 +200,7 @@
 *>          For optimal efficiency, LWORK >= (NB+3)*N,
 *>          where NB is the max of the blocksize for DSYTRD and DORMTR
 *>          returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dsygs2.f b/SRC/dsygs2.f
index 4da6bc54..2b7bf1df 100644
--- a/SRC/dsygs2.f
+++ b/SRC/dsygs2.f
@@ -82,8 +82,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the transformed matrix, stored in the
 *>          same format as A.
 *> \endverbatim
diff --git a/SRC/dsygst.f b/SRC/dsygst.f
index 3dbfa82e..792a8076 100644
--- a/SRC/dsygst.f
+++ b/SRC/dsygst.f
@@ -82,8 +82,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the transformed matrix, stored in the
 *>          same format as A.
 *> \endverbatim
diff --git a/SRC/dsygv.f b/SRC/dsygv.f
index 29f5ca55..dc5e0591 100644
--- a/SRC/dsygv.f
+++ b/SRC/dsygv.f
@@ -83,8 +83,7 @@
 *>          upper triangular part of the matrix A.  If UPLO = 'L',
 *>          the leading N-by-N lower triangular part of A contains
 *>          the lower triangular part of the matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if JOBZ = 'V', then if INFO = 0, A contains the
 *>          matrix Z of eigenvectors.  The eigenvectors are normalized
 *>          as follows:
@@ -109,8 +108,7 @@
 *>          contains the upper triangular part of the matrix B.
 *>          If UPLO = 'L', the leading N-by-N lower triangular part of B
 *>          contains the lower triangular part of the matrix B.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO <= N, the part of B containing the matrix is
 *>          overwritten by the triangular factor U or L from the Cholesky
 *>          factorization B = U**T*U or B = L*L**T.
@@ -140,8 +138,7 @@
 *>          The length of the array WORK.  LWORK >= max(1,3*N-1).
 *>          For optimal efficiency, LWORK >= (NB+2)*N,
 *>          where NB is the blocksize for DSYTRD returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dsygvd.f b/SRC/dsygvd.f
index 28e4d1cc..0fde04c4 100644
--- a/SRC/dsygvd.f
+++ b/SRC/dsygvd.f
@@ -91,8 +91,7 @@
 *>          upper triangular part of the matrix A.  If UPLO = 'L',
 *>          the leading N-by-N lower triangular part of A contains
 *>          the lower triangular part of the matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if JOBZ = 'V', then if INFO = 0, A contains the
 *>          matrix Z of eigenvectors.  The eigenvectors are normalized
 *>          as follows:
@@ -117,8 +116,7 @@
 *>          upper triangular part of the matrix B.  If UPLO = 'L',
 *>          the leading N-by-N lower triangular part of B contains
 *>          the lower triangular part of the matrix B.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO <= N, the part of B containing the matrix is
 *>          overwritten by the triangular factor U or L from the Cholesky
 *>          factorization B = U**T*U or B = L*L**T.
@@ -149,8 +147,7 @@
 *>          If N <= 1,               LWORK >= 1.
 *>          If JOBZ = 'N' and N > 1, LWORK >= 2*N+1.
 *>          If JOBZ = 'V' and N > 1, LWORK >= 1 + 6*N + 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK and IWORK
 *>          arrays, returns these values as the first entries of the WORK
@@ -171,8 +168,7 @@
 *>          If N <= 1,                LIWORK >= 1.
 *>          If JOBZ  = 'N' and N > 1, LIWORK >= 1.
 *>          If JOBZ  = 'V' and N > 1, LIWORK >= 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK and
 *>          IWORK arrays, returns these values as the first entries of
diff --git a/SRC/dsygvx.f b/SRC/dsygvx.f
index b045a9a5..6f022477 100644
--- a/SRC/dsygvx.f
+++ b/SRC/dsygvx.f
@@ -97,8 +97,7 @@
 *>          upper triangular part of the matrix A.  If UPLO = 'L',
 *>          the leading N-by-N lower triangular part of A contains
 *>          the lower triangular part of the matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the lower triangle (if UPLO='L') or the upper
 *>          triangle (if UPLO='U') of A, including the diagonal, is
 *>          destroyed.
@@ -118,8 +117,7 @@
 *>          upper triangular part of the matrix B.  If UPLO = 'L',
 *>          the leading N-by-N lower triangular part of B contains
 *>          the lower triangular part of the matrix B.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO <= N, the part of B containing the matrix is
 *>          overwritten by the triangular factor U or L from the Cholesky
 *>          factorization B = U**T*U or B = L*L**T.
@@ -165,19 +163,16 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing C to tridiagonal form, where C is the symmetric
 *>          matrix of the standard symmetric problem to which the
 *>          generalized problem is transformed.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*DLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
@@ -210,8 +205,7 @@
 *>          The eigenvectors are normalized as follows:
 *>          if ITYPE = 1 or 2, Z**T*B*Z = I;
 *>          if ITYPE = 3, Z**T*inv(B)*Z = I.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If an eigenvector fails to converge, then that column of Z
 *>          contains the latest approximation to the eigenvector, and the
 *>          index of the eigenvector is returned in IFAIL.
@@ -239,8 +233,7 @@
 *>          The length of the array WORK.  LWORK >= max(1,8*N).
 *>          For optimal efficiency, LWORK >= (NB+3)*N,
 *>          where NB is the blocksize for DSYTRD returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dsyrfs.f b/SRC/dsyrfs.f
index e2582e4a..9333713c 100644
--- a/SRC/dsyrfs.f
+++ b/SRC/dsyrfs.f
@@ -167,12 +167,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/dsyrfsx.f b/SRC/dsyrfsx.f
index efab12d1..14e64b16 100644
--- a/SRC/dsyrfsx.f
+++ b/SRC/dsyrfsx.f
@@ -219,37 +219,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -258,8 +252,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -270,14 +263,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -285,26 +276,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -315,8 +302,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -335,8 +321,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0D+0
@@ -347,8 +332,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -358,8 +342,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/dsysv.f b/SRC/dsysv.f
index 1d5630a5..f4d2acfc 100644
--- a/SRC/dsysv.f
+++ b/SRC/dsysv.f
@@ -85,8 +85,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the block diagonal matrix D and the
 *>          multipliers used to obtain the factor U or L from the
 *>          factorization A = U*D*U**T or A = L*D*L**T as computed by
@@ -140,8 +139,7 @@
 *>          DSYTRF.
 *>          for LWORK < N, TRS will be done with Level BLAS 2
 *>          for LWORK >= N, TRS will be done with Level BLAS 3
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dsysvx.f b/SRC/dsysvx.f
index ef9733d8..7b92f4e8 100644
--- a/SRC/dsysvx.f
+++ b/SRC/dsysvx.f
@@ -135,8 +135,7 @@
 *>          contains the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
 *>          A = U*D*U**T or A = L*D*L**T as computed by DSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AF is an output argument and on exit
 *>          returns the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
@@ -162,8 +161,7 @@
 *>          is a 2-by-2 diagonal block.  If UPLO = 'L' and IPIV(k) =
 *>          IPIV(k+1) < 0, then rows and columns k+1 and -IPIV(k) were
 *>          interchanged and D(k:k+1,k:k+1) is a 2-by-2 diagonal block.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains details of the interchanges and the block structure
 *>          of D, as determined by DSYTRF.
@@ -236,8 +234,7 @@
 *>          The length of WORK.  LWORK >= max(1,3*N), and for best
 *>          performance, when FACT = 'N', LWORK >= max(1,3*N,N*NB), where
 *>          NB is the optimal blocksize for DSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dsysvxx.f b/SRC/dsysvxx.f
index de61c010..84953d35 100644
--- a/SRC/dsysvxx.f
+++ b/SRC/dsysvxx.f
@@ -167,8 +167,7 @@
 *>     N-by-N lower triangular part of A contains the lower
 *>     triangular part of the matrix A, and the strictly upper
 *>     triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>     On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>     diag(S)*A*diag(S).
 *> \endverbatim
@@ -186,8 +185,7 @@
 *>     contains the block diagonal matrix D and the multipliers
 *>     used to obtain the factor U or L from the factorization A =
 *>     U*D*U**T or A = L*D*L**T as computed by DSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then AF is an output argument and on exit
 *>     returns the block diagonal matrix D and the multipliers
 *>     used to obtain the factor U or L from the factorization A =
@@ -213,8 +211,7 @@
 *>     diagonal block.  If UPLO = 'L' and IPIV(k) = IPIV(k+1) < 0,
 *>     then rows and columns k+1 and -IPIV(k) were interchanged
 *>     and D(k:k+1,k:k+1) is a 2-by-2 diagonal block.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then IPIV is an output argument and on exit
 *>     contains details of the interchanges and the block
 *>     structure of D, as determined by DSYTRF.
@@ -325,37 +322,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -364,8 +355,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -376,14 +366,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -391,26 +379,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -421,8 +405,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -441,8 +424,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0D+0
@@ -450,8 +432,7 @@
 *>                    computed.
 *>            = 1.0 : Use the extra-precise refinement algorithm.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -461,8 +442,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/dsyswapr.f b/SRC/dsyswapr.f
index 8526ef89..8b0bd6af 100644
--- a/SRC/dsyswapr.f
+++ b/SRC/dsyswapr.f
@@ -61,8 +61,7 @@
 *>          A is DOUBLE PRECISION array, dimension (LDA,N)
 *>          On entry, the NB diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by DSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/dsytf2.f b/SRC/dsytf2.f
index c68ad395..3bdf48c8 100644
--- a/SRC/dsytf2.f
+++ b/SRC/dsytf2.f
@@ -76,8 +76,7 @@
 *>          leading n-by-n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L (see below for further details).
 *> \endverbatim
diff --git a/SRC/dsytrf.f b/SRC/dsytrf.f
index fcf06879..07341b1a 100644
--- a/SRC/dsytrf.f
+++ b/SRC/dsytrf.f
@@ -75,8 +75,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L (see below for further details).
 *> \endverbatim
@@ -111,8 +110,7 @@
 *>          LWORK is INTEGER
 *>          The length of WORK.  LWORK >=1.  For best performance
 *>          LWORK >= N*NB, where NB is the block size returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dsytri.f b/SRC/dsytri.f
index 5da539b2..da96cbe7 100644
--- a/SRC/dsytri.f
+++ b/SRC/dsytri.f
@@ -64,8 +64,7 @@
 *>          A is DOUBLE PRECISION array, dimension (LDA,N)
 *>          On entry, the block diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by DSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/dsytri2.f b/SRC/dsytri2.f
index 6ee2f281..c34c8f43 100644
--- a/SRC/dsytri2.f
+++ b/SRC/dsytri2.f
@@ -65,8 +65,7 @@
 *>          A is DOUBLE PRECISION array, dimension (LDA,N)
 *>          On entry, the NB diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by DSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/dsytri2x.f b/SRC/dsytri2x.f
index 38aa6d0e..71a79851 100644
--- a/SRC/dsytri2x.f
+++ b/SRC/dsytri2x.f
@@ -64,8 +64,7 @@
 *>          A is DOUBLE PRECISION array, dimension (LDA,N)
 *>          On entry, the NNB diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by DSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/dtfsm.f b/SRC/dtfsm.f
index 1d94232f..57e174f5 100644
--- a/SRC/dtfsm.f
+++ b/SRC/dtfsm.f
@@ -68,14 +68,11 @@
 *>          SIDE is CHARACTER*1
 *>           On entry, SIDE specifies whether op( A ) appears on the left
 *>           or right of X as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              SIDE = 'L' or 'l'   op( A )*X = alpha*B.
-*> \endverbatim
-*> \verbatim
+*>
 *>              SIDE = 'R' or 'r'   X*op( A ) = alpha*B.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -86,8 +83,7 @@
 *>           an upper or lower triangular matrix as follows:
 *>           UPLO = 'U' or 'u' RFP A came from an upper triangular matrix
 *>           UPLO = 'L' or 'l' RFP A came from a  lower triangular matrix
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -96,14 +92,11 @@
 *>          TRANS is CHARACTER*1
 *>           On entry, TRANS  specifies the form of op( A ) to be used
 *>           in the matrix multiplication as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              TRANS  = 'N' or 'n'   op( A ) = A.
-*> \endverbatim
-*> \verbatim
+*>
 *>              TRANS  = 'T' or 't'   op( A ) = A'.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -112,15 +105,12 @@
 *>          DIAG is CHARACTER*1
 *>           On entry, DIAG specifies whether or not RFP A is unit
 *>           triangular as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              DIAG = 'U' or 'u'   A is assumed to be unit triangular.
-*> \endverbatim
-*> \verbatim
+*>
 *>              DIAG = 'N' or 'n'   A is not assumed to be unit
 *>                                  triangular.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/dtftri.f b/SRC/dtftri.f
index 127ceb63..03c8bca6 100644
--- a/SRC/dtftri.f
+++ b/SRC/dtftri.f
@@ -86,8 +86,7 @@
 *>          elements of lower packed A. The LDA of RFP A is (N+1)/2 when
 *>          TRANSR = 'T'. When TRANSR is 'N' the LDA is N+1 when N is
 *>          even and N is odd. See the Note below for more details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the (triangular) inverse of the original matrix, in
 *>          the same storage format.
 *> \endverbatim
diff --git a/SRC/dtgevc.f b/SRC/dtgevc.f
index fbe3b814..7b5553c3 100644
--- a/SRC/dtgevc.f
+++ b/SRC/dtgevc.f
@@ -146,13 +146,11 @@
 *>          if HOWMNY = 'S', the left eigenvectors of (S,P) specified by
 *>                      SELECT, stored consecutively in the columns of
 *>                      VL, in the same order as their eigenvalues.
-*> \endverbatim
-*> \verbatim
+*>
 *>          A complex eigenvector corresponding to a complex eigenvalue
 *>          is stored in two consecutive columns, the first holding the
 *>          real part, and the second the imaginary part.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Not referenced if SIDE = 'R'.
 *> \endverbatim
 *>
@@ -169,8 +167,7 @@
 *>          On entry, if SIDE = 'R' or 'B' and HOWMNY = 'B', VR must
 *>          contain an N-by-N matrix Z (usually the orthogonal matrix Z
 *>          of right Schur vectors returned by DHGEQZ).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if SIDE = 'R' or 'B', VR contains:
 *>          if HOWMNY = 'A', the matrix X of right eigenvectors of (S,P);
 *>          if HOWMNY = 'B' or 'b', the matrix Z*X;
@@ -178,8 +175,7 @@
 *>                      specified by SELECT, stored consecutively in the
 *>                      columns of VR, in the same order as their
 *>                      eigenvalues.
-*> \endverbatim
-*> \verbatim
+*>
 *>          A complex eigenvector corresponding to a complex eigenvalue
 *>          is stored in two consecutive columns, the first holding the
 *>          real part and the second the imaginary part.
diff --git a/SRC/dtgexc.f b/SRC/dtgexc.f
index 9580f3ce..6a4fa366 100644
--- a/SRC/dtgexc.f
+++ b/SRC/dtgexc.f
@@ -169,8 +169,7 @@
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.
 *>          LWORK >= 1 when N <= 1, otherwise LWORK >= 4*N + 16.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dtgsen.f b/SRC/dtgsen.f
index 831c9c73..e3e2110b 100644
--- a/SRC/dtgsen.f
+++ b/SRC/dtgsen.f
@@ -164,8 +164,7 @@
 *> \param[out] BETA
 *> \verbatim
 *>          BETA is DOUBLE PRECISION array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, (ALPHAR(j) + ALPHAI(j)*i)/BETA(j), j=1,...,N, will
 *>          be the generalized eigenvalues.  ALPHAR(j) + ALPHAI(j)*i
 *>          and BETA(j),j=1,...,N  are the diagonals of the complex Schur
@@ -228,8 +227,7 @@
 *> \param[out] PR
 *> \verbatim
 *>          PR is DOUBLE PRECISION
-*> \endverbatim
-*> \verbatim
+*>
 *>          If IJOB = 1, 4 or 5, PL, PR are lower bounds on the
 *>          reciprocal of the norm of "projections" onto left and right
 *>          eigenspaces with respect to the selected cluster.
@@ -262,8 +260,7 @@
 *>          The dimension of the array WORK. LWORK >=  4*N+16.
 *>          If IJOB = 1, 2 or 4, LWORK >= MAX(4*N+16, 2*M*(N-M)).
 *>          If IJOB = 3 or 5, LWORK >= MAX(4*N+16, 4*M*(N-M)).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
@@ -282,8 +279,7 @@
 *>          The dimension of the array IWORK. LIWORK >= 1.
 *>          If IJOB = 1, 2 or 4, LIWORK >=  N+6.
 *>          If IJOB = 3 or 5, LIWORK >= MAX(2*M*(N-M), N+6).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal size of the IWORK array,
 *>          returns this value as the first entry of the IWORK array, and
diff --git a/SRC/dtgsja.f b/SRC/dtgsja.f
index c8a3b476..fdf5d29b 100644
--- a/SRC/dtgsja.f
+++ b/SRC/dtgsja.f
@@ -185,8 +185,7 @@
 *> \param[in] L
 *> \verbatim
 *>          L is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          K and L specify the subblocks in the input matrices A and B:
 *>          A23 = A(K+1:MIN(K+L,M),N-L+1:N) and B13 = B(1:L,N-L+1:N)
 *>          of A and B, whose GSVD is going to be computed by DTGSJA.
@@ -229,8 +228,7 @@
 *> \param[in] TOLB
 *> \verbatim
 *>          TOLB is DOUBLE PRECISION
-*> \endverbatim
-*> \verbatim
+*>
 *>          TOLA and TOLB are the convergence criteria for the Jacobi-
 *>          Kogbetliantz iteration procedure. Generally, they are the
 *>          same as used in the preprocessing step, say
@@ -246,8 +244,7 @@
 *> \param[out] BETA
 *> \verbatim
 *>          BETA is DOUBLE PRECISION array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, ALPHA and BETA contain the generalized singular
 *>          value pairs of A and B;
 *>            ALPHA(1:K) = 1,
diff --git a/SRC/dtgsna.f b/SRC/dtgsna.f
index e4a1b387..3818668b 100644
--- a/SRC/dtgsna.f
+++ b/SRC/dtgsna.f
@@ -203,8 +203,7 @@
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK. LWORK >= max(1,N).
 *>          If JOB = 'V' or 'B' LWORK >= 2*N*(N+2)+16.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dtgsyl.f b/SRC/dtgsyl.f
index 19fa8acf..b59acfec 100644
--- a/SRC/dtgsyl.f
+++ b/SRC/dtgsyl.f
@@ -234,8 +234,7 @@
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK. LWORK > = 1.
 *>          If IJOB = 1 or 2 and TRANS = 'N', LWORK >= max(1,2*M*N).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/dtrexc.f b/SRC/dtrexc.f
index 3d0f0a80..6623e1ab 100644
--- a/SRC/dtrexc.f
+++ b/SRC/dtrexc.f
@@ -104,8 +104,7 @@
 *> \param[in,out] ILST
 *> \verbatim
 *>          ILST is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          Specify the reordering of the diagonal blocks of T.
 *>          The block with row index IFST is moved to row ILST, by a
 *>          sequence of transpositions between adjacent blocks.
diff --git a/SRC/dtrsen.f b/SRC/dtrsen.f
index b0234b58..8e79e909 100644
--- a/SRC/dtrsen.f
+++ b/SRC/dtrsen.f
@@ -136,8 +136,7 @@
 *> \param[out] WI
 *> \verbatim
 *>          WI is DOUBLE PRECISION array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          The real and imaginary parts, respectively, of the reordered
 *>          eigenvalues of T. The eigenvalues are stored in the same
 *>          order as on the diagonal of T, with WR(i) = T(i,i) and, if
@@ -186,8 +185,7 @@
 *>          If JOB = 'N', LWORK >= max(1,N);
 *>          if JOB = 'E', LWORK >= max(1,M*(N-M));
 *>          if JOB = 'V' or 'B', LWORK >= max(1,2*M*(N-M)).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
@@ -206,8 +204,7 @@
 *>          The dimension of the array IWORK.
 *>          If JOB = 'N' or 'E', LIWORK >= 1;
 *>          if JOB = 'V' or 'B', LIWORK >= max(1,M*(N-M)).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal size of the IWORK array,
 *>          returns this value as the first entry of the IWORK array, and
diff --git a/SRC/dtrti2.f b/SRC/dtrti2.f
index 0e359456..bba6491c 100644
--- a/SRC/dtrti2.f
+++ b/SRC/dtrti2.f
@@ -78,8 +78,7 @@
 *>          triangular part of A is not referenced.  If DIAG = 'U', the
 *>          diagonal elements of A are also not referenced and are
 *>          assumed to be 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the (triangular) inverse of the original matrix, in
 *>          the same storage format.
 *> \endverbatim
diff --git a/SRC/dtzrzf.f b/SRC/dtzrzf.f
index 73084758..9661e370 100644
--- a/SRC/dtzrzf.f
+++ b/SRC/dtzrzf.f
@@ -95,8 +95,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,M).
 *>          For optimum performance LWORK >= M*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/ieeeck.f b/SRC/ieeeck.f
index 1f536a77..06d0f515 100644
--- a/SRC/ieeeck.f
+++ b/SRC/ieeeck.f
@@ -62,8 +62,7 @@
 *>          Must contain the value 1.0
 *>          This is passed to prevent the compiler from optimizing
 *>          away this code.
-*> \endverbatim
-*> \verbatim
+*>
 *>  RETURN VALUE:  INTEGER
 *>          = 0:  Arithmetic failed to produce the correct answers
 *>          = 1:  Arithmetic produced the correct answers
diff --git a/SRC/iparmq.f b/SRC/iparmq.f
index f49e0d04..9f6d0c86 100644
--- a/SRC/iparmq.f
+++ b/SRC/iparmq.f
@@ -44,21 +44,18 @@
 *>          ISPEC is integer scalar
 *>              ISPEC specifies which tunable parameter IPARMQ should
 *>              return.
-*> \endverbatim
-*> \verbatim
+*>
 *>              ISPEC=12: (INMIN)  Matrices of order nmin or less
 *>                        are sent directly to xLAHQR, the implicit
 *>                        double shift QR algorithm.  NMIN must be
 *>                        at least 11.
-*> \endverbatim
-*> \verbatim
+*>
 *>              ISPEC=13: (INWIN)  Size of the deflation window.
 *>                        This is best set greater than or equal to
 *>                        the number of simultaneous shifts NS.
 *>                        Larger matrices benefit from larger deflation
 *>                        windows.
-*> \endverbatim
-*> \verbatim
+*>
 *>              ISPEC=14: (INIBL) Determines when to stop nibbling and
 *>                        invest in an (expensive) multi-shift QR sweep.
 *>                        If the aggressive early deflation subroutine
@@ -73,12 +70,10 @@
 *>                        IPARMQ(ISPEC=14) greater than or equal to 100
 *>                        prevents TTQRE from skipping a multi-shift
 *>                        QR sweep.
-*> \endverbatim
-*> \verbatim
+*>
 *>              ISPEC=15: (NSHFTS) The number of simultaneous shifts in
 *>                        a multi-shift QR iteration.
-*> \endverbatim
-*> \verbatim
+*>
 *>              ISPEC=16: (IACC22) IPARMQ is set to 0, 1 or 2 with the
 *>                        following meanings.
 *>                        0:  During the multi-shift QR sweep,
diff --git a/SRC/sbbcsd.f b/SRC/sbbcsd.f
index 4007c609..9099d023 100644
--- a/SRC/sbbcsd.f
+++ b/SRC/sbbcsd.f
@@ -282,8 +282,7 @@
 *> \verbatim
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK. LWORK >= MAX(1,8*Q).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal size of the WORK array,
 *>          returns this value as the first entry of the work array, and
@@ -298,20 +297,16 @@
 *>          > 0:  if SBBCSD did not converge, INFO specifies the number
 *>                of nonzero entries in PHI, and B11D, B11E, etc.,
 *>                contain the partially reduced matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Reference
 *>  =========
-*> \endverbatim
-*> \verbatim
+*>
 *>  [1] Brian D. Sutton. Computing the complete CS decomposition. Numer.
 *>      Algorithms, 50(1):33-65, 2009.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  TOLMUL  REAL, default = MAX(10,MIN(100,EPS**(-1/8)))
 *>          TOLMUL controls the convergence criterion of the QR loop.
 *>          Angles THETA(i), PHI(i) are rounded to 0 or PI/2 when they
diff --git a/SRC/sbdsqr.f b/SRC/sbdsqr.f
index b37102c4..5d7fbcf9 100644
--- a/SRC/sbdsqr.f
+++ b/SRC/sbdsqr.f
@@ -187,12 +187,10 @@
 *>                   elements of a bidiagonal matrix which is orthogonally
 *>                   similar to the input matrix B;  if INFO = i, i
 *>                   elements of E have not converged to zero.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  TOLMUL  REAL, default = max(10,min(100,EPS**(-1/8)))
 *>          TOLMUL controls the convergence criterion of the QR loop.
 *>          If it is positive, TOLMUL*EPS is the desired relative
@@ -207,8 +205,7 @@
 *>          Default is to lose at either one eighth or 2 of the
 *>             available decimal digits in each computed singular value
 *>             (whichever is smaller).
-*> \endverbatim
-*> \verbatim
+*>
 *>  MAXITR  INTEGER, default = 6
 *>          MAXITR controls the maximum number of passes of the
 *>          algorithm through its inner loop. The algorithms stops
diff --git a/SRC/sgbrfs.f b/SRC/sgbrfs.f
index 13357c84..17c0993f 100644
--- a/SRC/sgbrfs.f
+++ b/SRC/sgbrfs.f
@@ -180,12 +180,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/sgbrfsx.f b/SRC/sgbrfsx.f
index de8754be..8a5e17f2 100644
--- a/SRC/sgbrfsx.f
+++ b/SRC/sgbrfsx.f
@@ -256,37 +256,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -295,8 +289,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -307,14 +300,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -322,26 +313,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -352,8 +339,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -372,8 +358,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0
@@ -384,8 +369,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -395,8 +379,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/sgbsvx.f b/SRC/sgbsvx.f
index aa142e00..81fbaa4f 100644
--- a/SRC/sgbsvx.f
+++ b/SRC/sgbsvx.f
@@ -150,14 +150,12 @@
 *>          The j-th column of A is stored in the j-th column of the
 *>          array AB as follows:
 *>          AB(KU+1+i-j,j) = A(i,j) for max(1,j-KU)<=i<=min(N,j+kl)
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'F' and EQUED is not 'N', then A must have been
 *>          equilibrated by the scaling factors in R and/or C.  AB is not
 *>          modified if FACT = 'F' or 'N', or if FACT = 'E' and
 *>          EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if EQUED .ne. 'N', A is scaled as follows:
 *>          EQUED = 'R':  A := diag(R) * A
 *>          EQUED = 'C':  A := A * diag(C)
@@ -180,12 +178,10 @@
 *>          and the multipliers used during the factorization are stored
 *>          in rows KL+KU+2 to 2*KL+KU+1.  If EQUED .ne. 'N', then AFB is
 *>          the factored form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AFB is an output argument and on exit
 *>          returns details of the LU factorization of A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then AFB is an output argument and on exit
 *>          returns details of the LU factorization of the equilibrated
 *>          matrix A (see the description of AB for the form of the
@@ -205,13 +201,11 @@
 *>          contains the pivot indices from the factorization A = L*U
 *>          as computed by SGBTRF; row i of the matrix was interchanged
 *>          with row IPIV(i).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains the pivot indices from the factorization A = L*U
 *>          of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then IPIV is an output argument and on exit
 *>          contains the pivot indices from the factorization A = L*U
 *>          of the equilibrated matrix A.
diff --git a/SRC/sgbsvxx.f b/SRC/sgbsvxx.f
index 93c7e343..ddf4e602 100644
--- a/SRC/sgbsvxx.f
+++ b/SRC/sgbsvxx.f
@@ -180,14 +180,12 @@
 *>     The j-th column of A is stored in the j-th column of the
 *>     array AB as follows:
 *>     AB(KU+1+i-j,j) = A(i,j) for max(1,j-KU)<=i<=min(N,j+kl)
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'F' and EQUED is not 'N', then AB must have been
 *>     equilibrated by the scaling factors in R and/or C.  AB is not
 *>     modified if FACT = 'F' or 'N', or if FACT = 'E' and
 *>     EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>     On exit, if EQUED .ne. 'N', A is scaled as follows:
 *>     EQUED = 'R':  A := diag(R) * A
 *>     EQUED = 'C':  A := A * diag(C)
@@ -210,13 +208,11 @@
 *>     and the multipliers used during the factorization are stored
 *>     in rows KL+KU+2 to 2*KL+KU+1.  If EQUED .ne. 'N', then AFB is
 *>     the factored form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then AF is an output argument and on exit
 *>     returns the factors L and U from the factorization A = P*L*U
 *>     of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'E', then AF is an output argument and on exit
 *>     returns the factors L and U from the factorization A = P*L*U
 *>     of the equilibrated matrix A (see the description of A for
@@ -236,13 +232,11 @@
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     as computed by SGETRF; row i of the matrix was interchanged
 *>     with row IPIV(i).
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then IPIV is an output argument and on exit
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'E', then IPIV is an output argument and on exit
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     of the equilibrated matrix A.
@@ -382,37 +376,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -421,8 +409,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -433,14 +420,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -448,26 +433,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -478,8 +459,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -498,8 +478,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0
@@ -510,8 +489,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -521,8 +499,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/sgbtf2.f b/SRC/sgbtf2.f
index 6d46dba5..871bca5d 100644
--- a/SRC/sgbtf2.f
+++ b/SRC/sgbtf2.f
@@ -76,8 +76,7 @@
 *>          The j-th column of A is stored in the j-th column of the
 *>          array AB as follows:
 *>          AB(kl+ku+1+i-j,j) = A(i,j) for max(1,j-ku)<=i<=min(m,j+kl)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, details of the factorization: U is stored as an
 *>          upper triangular band matrix with KL+KU superdiagonals in
 *>          rows 1 to KL+KU+1, and the multipliers used during the
diff --git a/SRC/sgbtrf.f b/SRC/sgbtrf.f
index c1b951ae..9add5355 100644
--- a/SRC/sgbtrf.f
+++ b/SRC/sgbtrf.f
@@ -76,8 +76,7 @@
 *>          The j-th column of A is stored in the j-th column of the
 *>          array AB as follows:
 *>          AB(kl+ku+1+i-j,j) = A(i,j) for max(1,j-ku)<=i<=min(m,j+kl)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, details of the factorization: U is stored as an
 *>          upper triangular band matrix with KL+KU superdiagonals in
 *>          rows 1 to KL+KU+1, and the multipliers used during the
diff --git a/SRC/sgees.f b/SRC/sgees.f
index 50df3aff..8b564da5 100644
--- a/SRC/sgees.f
+++ b/SRC/sgees.f
@@ -168,8 +168,7 @@
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.  LWORK >= max(1,3*N).
 *>          For good performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sgeesx.f b/SRC/sgeesx.f
index ff064a52..dba6cc93 100644
--- a/SRC/sgeesx.f
+++ b/SRC/sgeesx.f
@@ -209,8 +209,7 @@
 *>          returned if LWORK < max(1,3*N), but if SENSE = 'E' or 'V' or
 *>          'B' this may not be large enough.
 *>          For good performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates upper bounds on the optimal sizes of the
 *>          arrays WORK and IWORK, returns these values as the first
@@ -232,8 +231,7 @@
 *>          Note that SDIM*(N-SDIM) <= N*N/4. Note also that an error is
 *>          only returned if LIWORK < 1, but if SENSE = 'V' or 'B' this
 *>          may not be large enough.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates upper bounds on the optimal sizes of
 *>          the arrays WORK and IWORK, returns these values as the first
diff --git a/SRC/sgeev.f b/SRC/sgeev.f
index 1b124658..68f666cb 100644
--- a/SRC/sgeev.f
+++ b/SRC/sgeev.f
@@ -156,8 +156,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,3*N), and
 *>          if JOBVL = 'V' or JOBVR = 'V', LWORK >= 4*N.  For good
 *>          performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sgeevx.f b/SRC/sgeevx.f
index 5c78c837..bd3c2238 100644
--- a/SRC/sgeevx.f
+++ b/SRC/sgeevx.f
@@ -89,8 +89,7 @@
 *>                 to make the rows and columns of A more equal in
 *>                 norm. Do not permute;
 *>          = 'B': Both diagonally scale and permute A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Computed reciprocal condition numbers will be for the matrix
 *>          after balancing and/or permuting. Permuting does not change
 *>          condition numbers (in exact arithmetic), but balancing does.
@@ -120,8 +119,7 @@
 *>          = 'E': Computed for eigenvalues only;
 *>          = 'V': Computed for right eigenvectors only;
 *>          = 'B': Computed for eigenvalues and right eigenvectors.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If SENSE = 'E' or 'B', both left and right eigenvectors
 *>          must also be computed (JOBVL = 'V' and JOBVR = 'V').
 *> \endverbatim
@@ -265,8 +263,7 @@
 *>          LWORK >= max(1,2*N), and if JOBVL = 'V' or JOBVR = 'V',
 *>          LWORK >= 3*N.  If SENSE = 'V' or 'B', LWORK >= N*(N+6).
 *>          For good performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sgegs.f b/SRC/sgegs.f
index 570f18e6..21b46fec 100644
--- a/SRC/sgegs.f
+++ b/SRC/sgegs.f
@@ -182,8 +182,7 @@
 *>          blocksizes (for SGEQRF, SORMQR, and SORGQR.)  Then compute:
 *>          NB  -- MAX of the blocksizes for SGEQRF, SORMQR, and SORGQR
 *>          The optimal LWORK is  2*N + N*(NB+1).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sgegv.f b/SRC/sgegv.f
index 6830dfc3..435316b7 100644
--- a/SRC/sgegv.f
+++ b/SRC/sgegv.f
@@ -171,8 +171,7 @@
 *>             u(j) = VL(:,j) + i*VL(:,j+1)
 *>          and
 *>            u(j+1) = VL(:,j) - i*VL(:,j+1).
-*> \endverbatim
-*> \verbatim
+*>
 *>          Each eigenvector is scaled so that its largest component has
 *>          abs(real part) + abs(imag. part) = 1, except for eigenvectors
 *>          corresponding to an eigenvalue with alpha = beta = 0, which
@@ -198,8 +197,7 @@
 *>            x(j) = VR(:,j) + i*VR(:,j+1)
 *>          and
 *>            x(j+1) = VR(:,j) - i*VR(:,j+1).
-*> \endverbatim
-*> \verbatim
+*>
 *>          Each eigenvector is scaled so that its largest component has
 *>          abs(real part) + abs(imag. part) = 1, except for eigenvalues
 *>          corresponding to an eigenvalue with alpha = beta = 0, which
@@ -230,8 +228,7 @@
 *>          NB  -- MAX of the blocksizes for SGEQRF, SORMQR, and SORGQR;
 *>          The optimal LWORK is:
 *>              2*N + MAX( 6*N, N*(NB+1) ).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sgehd2.f b/SRC/sgehd2.f
index 2282b4cf..af00d4c5 100644
--- a/SRC/sgehd2.f
+++ b/SRC/sgehd2.f
@@ -55,8 +55,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          It is assumed that A is already upper triangular in rows
 *>          and columns 1:ILO-1 and IHI+1:N. ILO and IHI are normally
 *>          set by a previous call to SGEBAL; otherwise they should be
diff --git a/SRC/sgehrd.f b/SRC/sgehrd.f
index 3afadab1..f4268235 100644
--- a/SRC/sgehrd.f
+++ b/SRC/sgehrd.f
@@ -55,8 +55,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          It is assumed that A is already upper triangular in rows
 *>          and columns 1:ILO-1 and IHI+1:N. ILO and IHI are normally
 *>          set by a previous call to SGEBAL; otherwise they should be
@@ -101,8 +100,7 @@
 *>          The length of the array WORK.  LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is the
 *>          optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sgels.f b/SRC/sgels.f
index 3df887db..a892941e 100644
--- a/SRC/sgels.f
+++ b/SRC/sgels.f
@@ -150,8 +150,7 @@
 *>          For optimal performance,
 *>          LWORK >= max( 1, MN + max( MN, NRHS )*NB ).
 *>          where MN = min(M,N) and NB is the optimum block size.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sgelsd.f b/SRC/sgelsd.f
index 52f6b730..8e7f5eab 100644
--- a/SRC/sgelsd.f
+++ b/SRC/sgelsd.f
@@ -162,8 +162,7 @@
 *>          tree (usually about 25), and
 *>             NLVL = MAX( 0, INT( LOG_2( MIN( M,N )/(SMLSIZ+1) ) ) + 1 )
 *>          For good performance, LWORK should generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the array WORK and the
 *>          minimum size of the array IWORK, and returns these values as
diff --git a/SRC/sgelss.f b/SRC/sgelss.f
index 590361a3..8550b415 100644
--- a/SRC/sgelss.f
+++ b/SRC/sgelss.f
@@ -140,8 +140,7 @@
 *>          The dimension of the array WORK. LWORK >= 1, and also:
 *>          LWORK >= 3*min(M,N) + max( 2*min(M,N), max(M,N), NRHS )
 *>          For good performance, LWORK should generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sgelsy.f b/SRC/sgelsy.f
index d07f2f73..4f77b700 100644
--- a/SRC/sgelsy.f
+++ b/SRC/sgelsy.f
@@ -168,8 +168,7 @@
 *>          where NB is an upper bound on the blocksize returned
 *>          by ILAENV for the routines SGEQP3, STZRZF, STZRQF, SORMQR,
 *>          and SORMRZ.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sgeqp3.f b/SRC/sgeqp3.f
index b4dc7be3..2a2497e3 100644
--- a/SRC/sgeqp3.f
+++ b/SRC/sgeqp3.f
@@ -99,8 +99,7 @@
 *>          The dimension of the array WORK. LWORK >= 3*N+1.
 *>          For optimal performance LWORK >= 2*N+( N+1 )*NB, where NB
 *>          is the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sgeqrf.f b/SRC/sgeqrf.f
index 1517a0d9..2d2499da 100644
--- a/SRC/sgeqrf.f
+++ b/SRC/sgeqrf.f
@@ -90,8 +90,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is 
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sgeqrfp.f b/SRC/sgeqrfp.f
index 99de605e..9fbf1d07 100644
--- a/SRC/sgeqrfp.f
+++ b/SRC/sgeqrfp.f
@@ -90,8 +90,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is 
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sgerfs.f b/SRC/sgerfs.f
index dd01b847..cd0220da 100644
--- a/SRC/sgerfs.f
+++ b/SRC/sgerfs.f
@@ -161,12 +161,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/sgerfsx.f b/SRC/sgerfsx.f
index 1989949e..12f0a1b5 100644
--- a/SRC/sgerfsx.f
+++ b/SRC/sgerfsx.f
@@ -231,37 +231,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -270,8 +264,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -282,14 +275,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -297,26 +288,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -327,8 +314,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -347,8 +333,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0
@@ -359,8 +344,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -370,8 +354,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/sgesvd.f b/SRC/sgesvd.f
index ddcc3cb3..2362c4ac 100644
--- a/SRC/sgesvd.f
+++ b/SRC/sgesvd.f
@@ -81,8 +81,7 @@
 *>                  vectors) are overwritten on the array A;
 *>          = 'N':  no rows of V**T (no right singular vectors) are
 *>                  computed.
-*> \endverbatim
-*> \verbatim
+*>
 *>          JOBVT and JOBU cannot both be 'O'.
 *> \endverbatim
 *>
@@ -179,8 +178,7 @@
 *>             - PATH 1t (N much larger than M, JOBVT='N')
 *>          LWORK >= MAX(1,3*MIN(M,N)+MAX(M,N),5*MIN(M,N)) for the other paths
 *>          For good performance, LWORK should generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sgesvj.f b/SRC/sgesvj.f
index 00fb8350..902dff94 100644
--- a/SRC/sgesvj.f
+++ b/SRC/sgesvj.f
@@ -224,8 +224,7 @@
 *>                 The singular values of A are SCALE*SVA(1:N), and this
 *>                 factored representation is due to the fact that some of the
 *>                 singular values of A might underflow or overflow.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If INFO .GT. 0 :
 *>          the procedure SGESVJ did not converge in the given number of
 *>          iterations (sweeps) and SCALE*SVA(1:N) may not be accurate.
diff --git a/SRC/sgesvx.f b/SRC/sgesvx.f
index 7aebb2b0..88940fbe 100644
--- a/SRC/sgesvx.f
+++ b/SRC/sgesvx.f
@@ -137,8 +137,7 @@
 *>          not 'N', then A must have been equilibrated by the scaling
 *>          factors in R and/or C.  A is not modified if FACT = 'F' or
 *>          'N', or if FACT = 'E' and EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if EQUED .ne. 'N', A is scaled as follows:
 *>          EQUED = 'R':  A := diag(R) * A
 *>          EQUED = 'C':  A := A * diag(C)
@@ -158,13 +157,11 @@
 *>          contains the factors L and U from the factorization
 *>          A = P*L*U as computed by SGETRF.  If EQUED .ne. 'N', then
 *>          AF is the factored form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AF is an output argument and on exit
 *>          returns the factors L and U from the factorization A = P*L*U
 *>          of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then AF is an output argument and on exit
 *>          returns the factors L and U from the factorization A = P*L*U
 *>          of the equilibrated matrix A (see the description of A for
@@ -184,13 +181,11 @@
 *>          contains the pivot indices from the factorization A = P*L*U
 *>          as computed by SGETRF; row i of the matrix was interchanged
 *>          with row IPIV(i).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains the pivot indices from the factorization A = P*L*U
 *>          of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then IPIV is an output argument and on exit
 *>          contains the pivot indices from the factorization A = P*L*U
 *>          of the equilibrated matrix A.
diff --git a/SRC/sgesvxx.f b/SRC/sgesvxx.f
index 0db0b03a..a1636350 100644
--- a/SRC/sgesvxx.f
+++ b/SRC/sgesvxx.f
@@ -168,8 +168,7 @@
 *>     not 'N', then A must have been equilibrated by the scaling
 *>     factors in R and/or C.  A is not modified if FACT = 'F' or
 *>     'N', or if FACT = 'E' and EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>     On exit, if EQUED .ne. 'N', A is scaled as follows:
 *>     EQUED = 'R':  A := diag(R) * A
 *>     EQUED = 'C':  A := A * diag(C)
@@ -189,13 +188,11 @@
 *>     contains the factors L and U from the factorization
 *>     A = P*L*U as computed by SGETRF.  If EQUED .ne. 'N', then
 *>     AF is the factored form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then AF is an output argument and on exit
 *>     returns the factors L and U from the factorization A = P*L*U
 *>     of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'E', then AF is an output argument and on exit
 *>     returns the factors L and U from the factorization A = P*L*U
 *>     of the equilibrated matrix A (see the description of A for
@@ -215,13 +212,11 @@
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     as computed by SGETRF; row i of the matrix was interchanged
 *>     with row IPIV(i).
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then IPIV is an output argument and on exit
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'E', then IPIV is an output argument and on exit
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     of the equilibrated matrix A.
@@ -361,37 +356,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -400,8 +389,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -412,14 +400,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -427,26 +413,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -457,8 +439,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -477,8 +458,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0
@@ -489,8 +469,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -500,8 +479,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/sgetri.f b/SRC/sgetri.f
index 2b1bfb04..1c1e340a 100644
--- a/SRC/sgetri.f
+++ b/SRC/sgetri.f
@@ -84,8 +84,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,N).
 *>          For optimal performance LWORK >= N*NB, where NB is
 *>          the optimal blocksize returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sgges.f b/SRC/sgges.f
index e0c7f4f4..7bba3308 100644
--- a/SRC/sgges.f
+++ b/SRC/sgges.f
@@ -116,8 +116,7 @@
 *>          SELCTG(ALPHAR(j),ALPHAI(j),BETA(j)) is true; i.e. if either
 *>          one of a complex conjugate pair of eigenvalues is selected,
 *>          then both complex eigenvalues are selected.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note that in the ill-conditioned case, a selected complex
 *>          eigenvalue may no longer satisfy SELCTG(ALPHAR(j),ALPHAI(j),
 *>          BETA(j)) = .TRUE. after ordering. INFO is to be set to N+2
@@ -189,8 +188,7 @@
 *>          If ALPHAI(j) is zero, then the j-th eigenvalue is real; if
 *>          positive, then the j-th and (j+1)-st eigenvalues are a
 *>          complex conjugate pair, with ALPHAI(j+1) negative.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note: the quotients ALPHAR(j)/BETA(j) and ALPHAI(j)/BETA(j)
 *>          may easily over- or underflow, and BETA(j) may even be zero.
 *>          Thus, the user should avoid naively computing the ratio.
@@ -239,8 +237,7 @@
 *>          The dimension of the array WORK.
 *>          If N = 0, LWORK >= 1, else LWORK >= max(8*N,6*N+16).
 *>          For good performance , LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sggesx.f b/SRC/sggesx.f
index 009d4d6b..c0b8a54c 100644
--- a/SRC/sggesx.f
+++ b/SRC/sggesx.f
@@ -204,8 +204,7 @@
 *>          If ALPHAI(j) is zero, then the j-th eigenvalue is real; if
 *>          positive, then the j-th and (j+1)-st eigenvalues are a
 *>          complex conjugate pair, with ALPHAI(j+1) negative.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note: the quotients ALPHAR(j)/BETA(j) and ALPHAI(j)/BETA(j)
 *>          may easily over- or underflow, and BETA(j) may even be zero.
 *>          Thus, the user should avoid naively computing the ratio.
@@ -277,8 +276,7 @@
 *>          Note also that an error is only returned if
 *>          LWORK < max( 8*N, 6*N+16), but if SENSE = 'E' or 'V' or 'B'
 *>          this may not be large enough.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the bound on the optimal size of the WORK
 *>          array and the minimum size of the IWORK array, returns these
@@ -299,8 +297,7 @@
 *>          The dimension of the array IWORK.
 *>          If SENSE = 'N' or N = 0, LIWORK >= 1, otherwise
 *>          LIWORK >= N+6.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the bound on the optimal size of the
 *>          WORK array and the minimum size of the IWORK array, returns
diff --git a/SRC/sggev.f b/SRC/sggev.f
index 8efd5c05..07fbd3b3 100644
--- a/SRC/sggev.f
+++ b/SRC/sggev.f
@@ -129,8 +129,7 @@
 *>          the j-th eigenvalue is real; if positive, then the j-th and
 *>          (j+1)-st eigenvalues are a complex conjugate pair, with
 *>          ALPHAI(j+1) negative.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note: the quotients ALPHAR(j)/BETA(j) and ALPHAI(j)/BETA(j)
 *>          may easily over- or underflow, and BETA(j) may even be zero.
 *>          Thus, the user should avoid naively computing the ratio
@@ -192,8 +191,7 @@
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.  LWORK >= max(1,8*N).
 *>          For good performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sggevx.f b/SRC/sggevx.f
index c07a8398..daa51c4d 100644
--- a/SRC/sggevx.f
+++ b/SRC/sggevx.f
@@ -169,8 +169,7 @@
 *>          the j-th eigenvalue is real; if positive, then the j-th and
 *>          (j+1)-st eigenvalues are a complex conjugate pair, with
 *>          ALPHAI(j+1) negative.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note: the quotients ALPHAR(j)/BETA(j) and ALPHAI(j)/BETA(j)
 *>          may easily over- or underflow, and BETA(j) may even be zero.
 *>          Thus, the user should avoid naively computing the ratio
@@ -314,8 +313,7 @@
 *>          LWORK >= max(1,6*N).
 *>          If SENSE = 'E', LWORK >= max(1,10*N).
 *>          If SENSE = 'V' or 'B', LWORK >= 2*N*N+8*N+16.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sggglm.f b/SRC/sggglm.f
index 1b107819..d276de90 100644
--- a/SRC/sggglm.f
+++ b/SRC/sggglm.f
@@ -130,8 +130,7 @@
 *> \param[out] Y
 *> \verbatim
 *>          Y is REAL array, dimension (P)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, X and Y are the solutions of the GLM problem.
 *> \endverbatim
 *>
@@ -148,8 +147,7 @@
 *>          For optimum performance, LWORK >= M+min(N,P)+max(N,P)*NB,
 *>          where NB is an upper bound for the optimal blocksizes for
 *>          SGEQRF, SGERQF, SORMQR and SORMRQ.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sgghrd.f b/SRC/sgghrd.f
index bc2dc485..fa54857d 100644
--- a/SRC/sgghrd.f
+++ b/SRC/sgghrd.f
@@ -104,8 +104,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          ILO and IHI mark the rows and columns of A which are to be
 *>          reduced.  It is assumed that A is already upper triangular
 *>          in rows and columns 1:ILO-1 and IHI+1:N.  ILO and IHI are
diff --git a/SRC/sgglse.f b/SRC/sgglse.f
index efc3a476..6a1bf5e5 100644
--- a/SRC/sgglse.f
+++ b/SRC/sgglse.f
@@ -141,8 +141,7 @@
 *>          For optimum performance LWORK >= P+min(M,N)+max(M,N)*NB,
 *>          where NB is an upper bound for the optimal blocksizes for
 *>          SGEQRF, SGERQF, SORMQR and SORMRQ.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sggsvd.f b/SRC/sggsvd.f
index 9b0fc544..08633242 100644
--- a/SRC/sggsvd.f
+++ b/SRC/sggsvd.f
@@ -170,8 +170,7 @@
 *> \param[out] L
 *> \verbatim
 *>          L is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, K and L specify the dimension of the subblocks
 *>          described in Purpose.
 *>          K + L = effective numerical rank of (A**T,B**T)**T.
@@ -213,8 +212,7 @@
 *> \param[out] BETA
 *> \verbatim
 *>          BETA is REAL array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, ALPHA and BETA contain the generalized singular
 *>          value pairs of A and B;
 *>            ALPHA(1:K) = 1,
@@ -296,12 +294,10 @@
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value.
 *>          > 0:  if INFO = 1, the Jacobi-type procedure failed to
 *>                converge.  For further details, see subroutine STGSJA.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  TOLA    REAL
 *>  TOLB    REAL
 *>          TOLA and TOLB are the thresholds to determine the effective
diff --git a/SRC/sggsvp.f b/SRC/sggsvp.f
index e9b37e82..ae481213 100644
--- a/SRC/sggsvp.f
+++ b/SRC/sggsvp.f
@@ -143,8 +143,7 @@
 *> \param[in] TOLB
 *> \verbatim
 *>          TOLB is REAL
-*> \endverbatim
-*> \verbatim
+*>
 *>          TOLA and TOLB are the thresholds to determine the effective
 *>          numerical rank of matrix B and a subblock of A. Generally,
 *>          they are set to
@@ -162,8 +161,7 @@
 *> \param[out] L
 *> \verbatim
 *>          L is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, K and L specify the dimension of the subblocks
 *>          described in Purpose section.
 *>          K + L = effective numerical rank of (A**T,B**T)**T.
diff --git a/SRC/sgtrfs.f b/SRC/sgtrfs.f
index 7168ca18..98c22c77 100644
--- a/SRC/sgtrfs.f
+++ b/SRC/sgtrfs.f
@@ -184,12 +184,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/sgtsv.f b/SRC/sgtsv.f
index 59a13c53..7c16fd42 100644
--- a/SRC/sgtsv.f
+++ b/SRC/sgtsv.f
@@ -66,8 +66,7 @@
 *>          DL is REAL array, dimension (N-1)
 *>          On entry, DL must contain the (n-1) sub-diagonal elements of
 *>          A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, DL is overwritten by the (n-2) elements of the
 *>          second super-diagonal of the upper triangular matrix U from
 *>          the LU factorization of A, in DL(1), ..., DL(n-2).
@@ -77,8 +76,7 @@
 *> \verbatim
 *>          D is REAL array, dimension (N)
 *>          On entry, D must contain the diagonal elements of A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, D is overwritten by the n diagonal elements of U.
 *> \endverbatim
 *>
@@ -87,8 +85,7 @@
 *>          DU is REAL array, dimension (N-1)
 *>          On entry, DU must contain the (n-1) super-diagonal elements
 *>          of A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, DU is overwritten by the (n-1) elements of the first
 *>          super-diagonal of U.
 *> \endverbatim
diff --git a/SRC/sgtsvx.f b/SRC/sgtsvx.f
index 226823f6..766e6834 100644
--- a/SRC/sgtsvx.f
+++ b/SRC/sgtsvx.f
@@ -135,8 +135,7 @@
 *>          If FACT = 'F', then DLF is an input argument and on entry
 *>          contains the (n-1) multipliers that define the matrix L from
 *>          the LU factorization of A as computed by SGTTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then DLF is an output argument and on exit
 *>          contains the (n-1) multipliers that define the matrix L from
 *>          the LU factorization of A.
@@ -148,8 +147,7 @@
 *>          If FACT = 'F', then DF is an input argument and on entry
 *>          contains the n diagonal elements of the upper triangular
 *>          matrix U from the LU factorization of A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then DF is an output argument and on exit
 *>          contains the n diagonal elements of the upper triangular
 *>          matrix U from the LU factorization of A.
@@ -160,8 +158,7 @@
 *>          DUF is or output) REAL array, dimension (N-1)
 *>          If FACT = 'F', then DUF is an input argument and on entry
 *>          contains the (n-1) elements of the first superdiagonal of U.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then DUF is an output argument and on exit
 *>          contains the (n-1) elements of the first superdiagonal of U.
 *> \endverbatim
@@ -172,8 +169,7 @@
 *>          If FACT = 'F', then DU2 is an input argument and on entry
 *>          contains the (n-2) elements of the second superdiagonal of
 *>          U.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then DU2 is an output argument and on exit
 *>          contains the (n-2) elements of the second superdiagonal of
 *>          U.
@@ -185,8 +181,7 @@
 *>          If FACT = 'F', then IPIV is an input argument and on entry
 *>          contains the pivot indices from the LU factorization of A as
 *>          computed by SGTTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains the pivot indices from the LU factorization of A;
 *>          row i of the matrix was interchanged with row IPIV(i).
diff --git a/SRC/sgttrf.f b/SRC/sgttrf.f
index 38a6be11..d85d51e7 100644
--- a/SRC/sgttrf.f
+++ b/SRC/sgttrf.f
@@ -59,8 +59,7 @@
 *>          DL is REAL array, dimension (N-1)
 *>          On entry, DL must contain the (n-1) sub-diagonal elements of
 *>          A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, DL is overwritten by the (n-1) multipliers that
 *>          define the matrix L from the LU factorization of A.
 *> \endverbatim
@@ -69,8 +68,7 @@
 *> \verbatim
 *>          D is REAL array, dimension (N)
 *>          On entry, D must contain the diagonal elements of A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, D is overwritten by the n diagonal elements of the
 *>          upper triangular matrix U from the LU factorization of A.
 *> \endverbatim
@@ -80,8 +78,7 @@
 *>          DU is REAL array, dimension (N-1)
 *>          On entry, DU must contain the (n-1) super-diagonal elements
 *>          of A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, DU is overwritten by the (n-1) elements of the first
 *>          super-diagonal of U.
 *> \endverbatim
diff --git a/SRC/shgeqz.f b/SRC/shgeqz.f
index 279bd84f..e573fcc9 100644
--- a/SRC/shgeqz.f
+++ b/SRC/shgeqz.f
@@ -255,8 +255,7 @@
 *> \verbatim
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.  LWORK >= max(1,N).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/shsein.f b/SRC/shsein.f
index 5678fedf..0ac9bf6b 100644
--- a/SRC/shsein.f
+++ b/SRC/shsein.f
@@ -125,8 +125,7 @@
 *> \param[in] WI
 *> \verbatim
 *>          WI is REAL array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On entry, the real and imaginary parts of the eigenvalues of
 *>          H; a complex conjugate pair of eigenvalues must be stored in
 *>          consecutive elements of WR and WI.
diff --git a/SRC/shseqr.f b/SRC/shseqr.f
index 8c7113bb..0cf4a1e3 100644
--- a/SRC/shseqr.f
+++ b/SRC/shseqr.f
@@ -83,8 +83,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>           It is assumed that H is already upper triangular in rows
 *>           and columns 1:ILO-1 and IHI+1:N. ILO and IHI are normally
 *>           set by a previous call to SGEBAL, and then passed to ZGEHRD
@@ -107,8 +106,7 @@
 *>           contents of H are unspecified on exit.  (The output value of
 *>           H when INFO.GT.0 is given under the description of INFO
 *>           below.)
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unlike earlier versions of SHSEQR, this subroutine may
 *>           explicitly H(i,j) = 0 for i.GT.j and j = 1, 2, ... ILO-1
 *>           or j = IHI+1, IHI+2, ... N.
@@ -128,8 +126,7 @@
 *> \param[out] WI
 *> \verbatim
 *>          WI is REAL array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>           The real and imaginary parts, respectively, of the computed
 *>           eigenvalues. If two eigenvalues are computed as a complex
 *>           conjugate pair, they are stored in consecutive elements of
@@ -180,8 +177,7 @@
 *>           may be required for optimal performance.  A workspace
 *>           query is recommended to determine the optimal workspace
 *>           size.
-*> \endverbatim
-*> \verbatim
+*>
 *>           If LWORK = -1, then SHSEQR does a workspace query.
 *>           In this case, SHSEQR checks the input parameters and
 *>           estimates the optimal workspace size for the given
diff --git a/SRC/sla_gbamv.f b/SRC/sla_gbamv.f
index d423ed7d..42579b9b 100644
--- a/SRC/sla_gbamv.f
+++ b/SRC/sla_gbamv.f
@@ -62,13 +62,11 @@
 *>          TRANS is INTEGER
 *>           On entry, TRANS specifies the operation to be performed as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>             BLAS_NO_TRANS      y := alpha*abs(A)*abs(x) + beta*abs(y)
 *>             BLAS_TRANS         y := alpha*abs(A**T)*abs(x) + beta*abs(y)
 *>             BLAS_CONJ_TRANS    y := alpha*abs(A**T)*abs(x) + beta*abs(y)
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -168,8 +166,7 @@
 *>           On entry, INCY specifies the increment for the elements of
 *>           Y. INCY must not be zero.
 *>           Unchanged on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Level 2 Blas routine.
 *> \endverbatim
 *>
diff --git a/SRC/sla_geamv.f b/SRC/sla_geamv.f
index 7df9e8a4..b8376cad 100644
--- a/SRC/sla_geamv.f
+++ b/SRC/sla_geamv.f
@@ -62,13 +62,11 @@
 *>          TRANS is INTEGER
 *>           On entry, TRANS specifies the operation to be performed as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>             BLAS_NO_TRANS      y := alpha*abs(A)*abs(x) + beta*abs(y)
 *>             BLAS_TRANS         y := alpha*abs(A**T)*abs(x) + beta*abs(y)
 *>             BLAS_CONJ_TRANS    y := alpha*abs(A**T)*abs(x) + beta*abs(y)
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -157,8 +155,7 @@
 *>           On entry, INCY specifies the increment for the elements of
 *>           Y. INCY must not be zero.
 *>           Unchanged on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Level 2 Blas routine.
 *> \endverbatim
 *>
diff --git a/SRC/sla_porfsx_extended.f b/SRC/sla_porfsx_extended.f
index 8a1c11a7..a7d7d429 100644
--- a/SRC/sla_porfsx_extended.f
+++ b/SRC/sla_porfsx_extended.f
@@ -190,37 +190,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -229,8 +223,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     This subroutine is only responsible for setting the second field
 *>     above.
 *>     See Lapack Working Note 165 for further details and extra
@@ -243,14 +236,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -258,26 +249,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -288,8 +275,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     This subroutine is only responsible for setting the second field
 *>     above.
 *>     See Lapack Working Note 165 for further details and extra
diff --git a/SRC/sla_syamv.f b/SRC/sla_syamv.f
index 9cf66d57..02e91503 100644
--- a/SRC/sla_syamv.f
+++ b/SRC/sla_syamv.f
@@ -62,16 +62,13 @@
 *>           On entry, UPLO specifies whether the upper or lower
 *>           triangular part of the array A is to be referenced as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = BLAS_UPPER   Only the upper triangular part of A
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = BLAS_LOWER   Only the lower triangular part of A
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/sla_syrfsx_extended.f b/SRC/sla_syrfsx_extended.f
index 63fecc83..3894d317 100644
--- a/SRC/sla_syrfsx_extended.f
+++ b/SRC/sla_syrfsx_extended.f
@@ -198,37 +198,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -237,8 +231,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     This subroutine is only responsible for setting the second field
 *>     above.
 *>     See Lapack Working Note 165 for further details and extra
@@ -251,14 +244,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -266,26 +257,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -296,8 +283,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     This subroutine is only responsible for setting the second field
 *>     above.
 *>     See Lapack Working Note 165 for further details and extra
diff --git a/SRC/slaed4.f b/SRC/slaed4.f
index 95ab425a..f365c555 100644
--- a/SRC/slaed4.f
+++ b/SRC/slaed4.f
@@ -106,24 +106,19 @@
 *>          INFO is INTEGER
 *>         = 0:  successful exit
 *>         > 0:  if INFO = 1, the updating process failed.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  Logical variable ORGATI (origin-at-i?) is used for distinguishing
 *>  whether D(i) or D(i+1) is treated as the origin.
-*> \endverbatim
-*> \verbatim
+*>
 *>            ORGATI = .true.    origin at i
 *>            ORGATI = .false.   origin at i+1
-*> \endverbatim
-*> \verbatim
+*>
 *>   Logical variable SWTCH3 (switch-for-3-poles?) is for noting
 *>   if we are working with THREE poles!
-*> \endverbatim
-*> \verbatim
+*>
 *>   MAXIT is the maximum number of iterations allowed for each
 *>   eigenvalue.
 *> \endverbatim
diff --git a/SRC/slagtf.f b/SRC/slagtf.f
index 8bfef3ba..e8c7c685 100644
--- a/SRC/slagtf.f
+++ b/SRC/slagtf.f
@@ -67,8 +67,7 @@
 *> \verbatim
 *>          A is REAL array, dimension (N)
 *>          On entry, A must contain the diagonal elements of T.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, A is overwritten by the n diagonal elements of the
 *>          upper triangular matrix U of the factorization of T.
 *> \endverbatim
@@ -84,8 +83,7 @@
 *>          B is REAL array, dimension (N-1)
 *>          On entry, B must contain the (n-1) super-diagonal elements of
 *>          T.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, B is overwritten by the (n-1) super-diagonal
 *>          elements of the matrix U of the factorization of T.
 *> \endverbatim
@@ -95,8 +93,7 @@
 *>          C is REAL array, dimension (N-1)
 *>          On entry, C must contain the (n-1) sub-diagonal elements of
 *>          T.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, C is overwritten by the (n-1) sub-diagonal elements
 *>          of the matrix L of the factorization of T.
 *> \endverbatim
@@ -128,11 +125,9 @@
 *>          an interchange occurred at the kth step of the elimination,
 *>          then IN(k) = 1, otherwise IN(k) = 0. The element IN(n)
 *>          returns the smallest positive integer j such that
-*> \endverbatim
-*> \verbatim
+*>
 *>             abs( u(j,j) ).le. norm( (T - lambda*I)(j) )*TOL,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where norm( A(j) ) denotes the sum of the absolute values of
 *>          the jth row of the matrix A. If no such j exists then IN(n)
 *>          is returned as zero. If IN(n) is returned as positive, then a
diff --git a/SRC/slagts.f b/SRC/slagts.f
index d25f3248..3afaae72 100644
--- a/SRC/slagts.f
+++ b/SRC/slagts.f
@@ -129,8 +129,7 @@
 *>          is the relative machine precision, but if TOL is supplied as
 *>          non-positive, then it is reset to eps*max( abs( u(i,j) ) ).
 *>          If  JOB .gt. 0  then TOL is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, TOL is changed as described above, only if TOL is
 *>          non-positive on entry. Otherwise TOL is unchanged.
 *> \endverbatim
diff --git a/SRC/slahqr.f b/SRC/slahqr.f
index 3f2213ea..6cb3a6aa 100644
--- a/SRC/slahqr.f
+++ b/SRC/slahqr.f
@@ -159,22 +159,19 @@
 *>                  per eigenvalue; elements i+1:ihi of WR and WI
 *>                  contain those eigenvalues which have been
 *>                  successfully computed.
-*> \endverbatim
-*> \verbatim
+*>
 *>                  If INFO .GT. 0 and WANTT is .FALSE., then on exit,
 *>                  the remaining unconverged eigenvalues are the
 *>                  eigenvalues of the upper Hessenberg matrix rows
 *>                  and columns ILO thorugh INFO of the final, output
 *>                  value of H.
-*> \endverbatim
-*> \verbatim
+*>
 *>                  If INFO .GT. 0 and WANTT is .TRUE., then on exit
 *>          (*)       (initial value of H)*U  = U*(final value of H)
 *>                  where U is an orthognal matrix.    The final
 *>                  value of H is upper Hessenberg and triangular in
 *>                  rows and columns INFO+1 through IHI.
-*> \endverbatim
-*> \verbatim
+*>
 *>                  If INFO .GT. 0 and WANTZ is .TRUE., then on exit
 *>                      (final value of Z)  = (initial value of Z)*U
 *>                  where U is the orthogonal matrix in (*)
diff --git a/SRC/slals0.f b/SRC/slals0.f
index 5e64889d..b0911444 100644
--- a/SRC/slals0.f
+++ b/SRC/slals0.f
@@ -101,8 +101,7 @@
 *>          SQRE is INTEGER
 *>         = 0: the lower block is an NR-by-NR square matrix.
 *>         = 1: the lower block is an NR-by-(NR+1) rectangular matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>         The bidiagonal matrix has row dimension N = NL + NR + 1,
 *>         and column dimension M = N + SQRE.
 *> \endverbatim
diff --git a/SRC/slaqgb.f b/SRC/slaqgb.f
index 6eb64275..886ab225 100644
--- a/SRC/slaqgb.f
+++ b/SRC/slaqgb.f
@@ -76,8 +76,7 @@
 *>          The j-th column of A is stored in the j-th column of the
 *>          array AB as follows:
 *>          AB(ku+1+i-j,j) = A(i,j) for max(1,j-ku)<=i<=min(m,j+kl)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the equilibrated matrix, in the same storage format
 *>          as A.  See EQUED for the form of the equilibrated matrix.
 *> \endverbatim
@@ -129,18 +128,15 @@
 *>                  by diag(C).
 *>          = 'B':  Both row and column equilibration, i.e., A has been
 *>                  replaced by diag(R) * A * diag(C).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if row or column scaling
 *>  should be done based on the ratio of the row or column scaling
 *>  factors.  If ROWCND < THRESH, row scaling is done, and if
 *>  COLCND < THRESH, column scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if row scaling
 *>  should be done based on the absolute size of the largest matrix
 *>  element.  If AMAX > LARGE or AMAX < SMALL, row scaling is done.
diff --git a/SRC/slaqge.f b/SRC/slaqge.f
index db1385fe..b11e1ded 100644
--- a/SRC/slaqge.f
+++ b/SRC/slaqge.f
@@ -111,18 +111,15 @@
 *>                  by diag(C).
 *>          = 'B':  Both row and column equilibration, i.e., A has been
 *>                  replaced by diag(R) * A * diag(C).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if row or column scaling
 *>  should be done based on the ratio of the row or column scaling
 *>  factors.  If ROWCND < THRESH, row scaling is done, and if
 *>  COLCND < THRESH, column scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if row scaling
 *>  should be done based on the absolute size of the largest matrix
 *>  element.  If AMAX > LARGE or AMAX < SMALL, row scaling is done.
diff --git a/SRC/slaqr0.f b/SRC/slaqr0.f
index 942674e6..94c1d3d3 100644
--- a/SRC/slaqr0.f
+++ b/SRC/slaqr0.f
@@ -102,8 +102,7 @@
 *>           .FALSE., then the contents of H are unspecified on exit.
 *>           (The output value of H when INFO.GT.0 is given under the
 *>           description of INFO below.)
-*> \endverbatim
-*> \verbatim
+*>
 *>           This subroutine may explicitly set H(i,j) = 0 for i.GT.j and
 *>           j = 1, 2, ... ILO-1 or j = IHI+1, IHI+2, ... N.
 *> \endverbatim
@@ -180,8 +179,7 @@
 *>           is sufficient, but LWORK typically as large as 6*N may
 *>           be required for optimal performance.  A workspace query
 *>           to determine the optimal workspace size is recommended.
-*> \endverbatim
-*> \verbatim
+*>
 *>           If LWORK = -1, then SLAQR0 does a workspace query.
 *>           In this case, SLAQR0 checks the input parameters and
 *>           estimates the optimal workspace size for the given
diff --git a/SRC/slaqr2.f b/SRC/slaqr2.f
index 12f0145b..3c96a201 100644
--- a/SRC/slaqr2.f
+++ b/SRC/slaqr2.f
@@ -248,8 +248,7 @@
 *>          The dimension of the work array WORK.  LWORK = 2*NW
 *>          suffices, but greater efficiency may result from larger
 *>          values of LWORK.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; SLAQR2
 *>          only estimates the optimal workspace size for the given
 *>          values of N, NW, KTOP and KBOT.  The estimate is returned
diff --git a/SRC/slaqr3.f b/SRC/slaqr3.f
index 264ca70b..4841606f 100644
--- a/SRC/slaqr3.f
+++ b/SRC/slaqr3.f
@@ -245,8 +245,7 @@
 *>          The dimension of the work array WORK.  LWORK = 2*NW
 *>          suffices, but greater efficiency may result from larger
 *>          values of LWORK.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; SLAQR3
 *>          only estimates the optimal workspace size for the given
 *>          values of N, NW, KTOP and KBOT.  The estimate is returned
diff --git a/SRC/slaqr4.f b/SRC/slaqr4.f
index e77e1e21..cd261201 100644
--- a/SRC/slaqr4.f
+++ b/SRC/slaqr4.f
@@ -109,8 +109,7 @@
 *>           .FALSE., then the contents of H are unspecified on exit.
 *>           (The output value of H when INFO.GT.0 is given under the
 *>           description of INFO below.)
-*> \endverbatim
-*> \verbatim
+*>
 *>           This subroutine may explicitly set H(i,j) = 0 for i.GT.j and
 *>           j = 1, 2, ... ILO-1 or j = IHI+1, IHI+2, ... N.
 *> \endverbatim
@@ -187,8 +186,7 @@
 *>           is sufficient, but LWORK typically as large as 6*N may
 *>           be required for optimal performance.  A workspace query
 *>           to determine the optimal workspace size is recommended.
-*> \endverbatim
-*> \verbatim
+*>
 *>           If LWORK = -1, then SLAQR4 does a workspace query.
 *>           In this case, SLAQR4 checks the input parameters and
 *>           estimates the optimal workspace size for the given
diff --git a/SRC/slaqsb.f b/SRC/slaqsb.f
index bfdb6e93..2a5ffb46 100644
--- a/SRC/slaqsb.f
+++ b/SRC/slaqsb.f
@@ -74,8 +74,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**T*U or A = L*L**T of the band
 *>          matrix A, in the same storage format as A.
@@ -112,17 +111,14 @@
 *>          = 'N':  No equilibration.
 *>          = 'Y':  Equilibration was done, i.e., A has been replaced by
 *>                  diag(S) * A * diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if scaling should be done
 *>  based on the ratio of the scaling factors.  If SCOND < THRESH,
 *>  scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if scaling should
 *>  be done based on the absolute size of the largest matrix element.
 *>  If AMAX > LARGE or AMAX < SMALL, scaling is done.
diff --git a/SRC/slaqsp.f b/SRC/slaqsp.f
index 22f33056..7facd59d 100644
--- a/SRC/slaqsp.f
+++ b/SRC/slaqsp.f
@@ -66,8 +66,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the equilibrated matrix:  diag(S) * A * diag(S), in
 *>          the same storage format as A.
 *> \endverbatim
@@ -97,17 +96,14 @@
 *>          = 'N':  No equilibration.
 *>          = 'Y':  Equilibration was done, i.e., A has been replaced by
 *>                  diag(S) * A * diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if scaling should be done
 *>  based on the ratio of the scaling factors.  If SCOND < THRESH,
 *>  scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if scaling should
 *>  be done based on the absolute size of the largest matrix element.
 *>  If AMAX > LARGE or AMAX < SMALL, scaling is done.
diff --git a/SRC/slaqsy.f b/SRC/slaqsy.f
index 36420a7a..19390482 100644
--- a/SRC/slaqsy.f
+++ b/SRC/slaqsy.f
@@ -68,8 +68,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if EQUED = 'Y', the equilibrated matrix:
 *>          diag(S) * A * diag(S).
 *> \endverbatim
@@ -105,17 +104,14 @@
 *>          = 'N':  No equilibration.
 *>          = 'Y':  Equilibration was done, i.e., A has been replaced by
 *>                  diag(S) * A * diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if scaling should be done
 *>  based on the ratio of the scaling factors.  If SCOND < THRESH,
 *>  scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if scaling should
 *>  be done based on the absolute size of the largest matrix element.
 *>  If AMAX > LARGE or AMAX < SMALL, scaling is done.
diff --git a/SRC/slarrd.f b/SRC/slarrd.f
index fbb70bdd..43b261e8 100644
--- a/SRC/slarrd.f
+++ b/SRC/slarrd.f
@@ -279,12 +279,10 @@
 *>                                        floating-point arithmetic.
 *>                        Cure: Increase the PARAMETER "FUDGE",
 *>                              recompile, and try again.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  FUDGE   REAL            , default = 2
 *>          A "fudge factor" to widen the Gershgorin intervals.  Ideally,
 *>          a value of 1 should work, but on machines with sloppy
@@ -292,8 +290,7 @@
 *>          publicly released versions should be large enough to handle
 *>          the worst machine around.  Note that this has no effect
 *>          on accuracy of the solution.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Based on contributions by
 *>     W. Kahan, University of California, Berkeley, USA
 *>     Beresford Parlett, University of California, Berkeley, USA
diff --git a/SRC/slarre.f b/SRC/slarre.f
index d09862a7..1c468d5c 100644
--- a/SRC/slarre.f
+++ b/SRC/slarre.f
@@ -249,8 +249,7 @@
 *>          < 0:  One of the called subroutines signaled an internal problem.
 *>                Needs inspection of the corresponding parameter IINFO
 *>                for further information.
-*> \endverbatim
-*> \verbatim
+*>
 *>          =-1:  Problem in SLARRD.
 *>          = 2:  No base representation could be found in MAXTRY iterations.
 *>                Increasing MAXTRY and recompilation might be a remedy.
diff --git a/SRC/slarrk.f b/SRC/slarrk.f
index f7d8fd8e..aaf5d63c 100644
--- a/SRC/slarrk.f
+++ b/SRC/slarrk.f
@@ -120,12 +120,10 @@
 *>          INFO is INTEGER
 *>          = 0:       Eigenvalue converged
 *>          = -1:      Eigenvalue did NOT converge
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  FUDGE   REAL            , default = 2
 *>          A "fudge factor" to widen the Gershgorin intervals.
 *> \endverbatim
diff --git a/SRC/slartg.f b/SRC/slartg.f
index bf94b6f4..f024f919 100644
--- a/SRC/slartg.f
+++ b/SRC/slartg.f
@@ -78,8 +78,7 @@
 *> \verbatim
 *>          R is REAL
 *>          The nonzero component of the rotated vector.
-*> \endverbatim
-*> \verbatim
+*>
 *>  This version has a few statements commented out for thread safety
 *>  (machine parameters are computed on each entry). 10 feb 03, SJH.
 *> \endverbatim
diff --git a/SRC/slartgp.f b/SRC/slartgp.f
index c08a8469..03b84b68 100644
--- a/SRC/slartgp.f
+++ b/SRC/slartgp.f
@@ -76,8 +76,7 @@
 *> \verbatim
 *>          R is REAL
 *>          The nonzero component of the rotated vector.
-*> \endverbatim
-*> \verbatim
+*>
 *>  This version has a few statements commented out for thread safety
 *>  (machine parameters are computed on each entry). 10 feb 03, SJH.
 *> \endverbatim
diff --git a/SRC/slascl.f b/SRC/slascl.f
index 20827b06..7ad115b5 100644
--- a/SRC/slascl.f
+++ b/SRC/slascl.f
@@ -86,8 +86,7 @@
 *> \param[in] CTO
 *> \verbatim
 *>          CTO is REAL
-*> \endverbatim
-*> \verbatim
+*>
 *>          The matrix A is multiplied by CTO/CFROM. A(I,J) is computed
 *>          without over/underflow if the final result CTO*A(I,J)/CFROM
 *>          can be represented without over/underflow.  CFROM must be
diff --git a/SRC/slasd1.f b/SRC/slasd1.f
index 808ebe27..98711d10 100644
--- a/SRC/slasd1.f
+++ b/SRC/slasd1.f
@@ -97,8 +97,7 @@
 *>          SQRE is INTEGER
 *>         = 0: the lower block is an NR-by-NR square matrix.
 *>         = 1: the lower block is an NR-by-(NR+1) rectangular matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>         The bidiagonal matrix has row dimension N = NL + NR + 1,
 *>         and column dimension M = N + SQRE.
 *> \endverbatim
diff --git a/SRC/slasd2.f b/SRC/slasd2.f
index 8118b524..0baaaa16 100644
--- a/SRC/slasd2.f
+++ b/SRC/slasd2.f
@@ -71,8 +71,7 @@
 *>          SQRE is INTEGER
 *>         = 0: the lower block is an NR-by-NR square matrix.
 *>         = 1: the lower block is an NR-by-(NR+1) rectangular matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>         The bidiagonal matrix has N = NL + NR + 1 rows and
 *>         M = N + SQRE >= N columns.
 *> \endverbatim
@@ -236,8 +235,7 @@
 *>         2 : non-zero in the lower half only
 *>         3 : dense
 *>         4 : deflated
-*> \endverbatim
-*> \verbatim
+*>
 *>         On exit, it is an array of dimension 4, with COLTYP(I) being
 *>         the dimension of the I-th type columns.
 *> \endverbatim
diff --git a/SRC/slasd3.f b/SRC/slasd3.f
index 106c1fc9..ba89d980 100644
--- a/SRC/slasd3.f
+++ b/SRC/slasd3.f
@@ -75,8 +75,7 @@
 *>          SQRE is INTEGER
 *>         = 0: the lower block is an NR-by-NR square matrix.
 *>         = 1: the lower block is an NR-by-(NR+1) rectangular matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>         The bidiagonal matrix has N = NL + NR + 1 rows and
 *>         M = N + SQRE >= N columns.
 *> \endverbatim
@@ -175,8 +174,7 @@
 *>         contains non-zero entries only at and below (or after) NL+2;
 *>         and the third is dense. The first column of U and the row of
 *>         VT are treated separately, however.
-*> \endverbatim
-*> \verbatim
+*>
 *>         The rows of the singular vectors found by SLASD4
 *>         must be likewise permuted before the matrix multiplies can
 *>         take place.
diff --git a/SRC/slasd4.f b/SRC/slasd4.f
index e730f76e..ee96ca14 100644
--- a/SRC/slasd4.f
+++ b/SRC/slasd4.f
@@ -114,24 +114,19 @@
 *>          INFO is INTEGER
 *>         = 0:  successful exit
 *>         > 0:  if INFO = 1, the updating process failed.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  Logical variable ORGATI (origin-at-i?) is used for distinguishing
 *>  whether D(i) or D(i+1) is treated as the origin.
-*> \endverbatim
-*> \verbatim
+*>
 *>            ORGATI = .true.    origin at i
 *>            ORGATI = .false.   origin at i+1
-*> \endverbatim
-*> \verbatim
+*>
 *>  Logical variable SWTCH3 (switch-for-3-poles?) is for noting
 *>  if we are working with THREE poles!
-*> \endverbatim
-*> \verbatim
+*>
 *>  MAXIT is the maximum number of iterations allowed for each
 *>  eigenvalue.
 *> \endverbatim
diff --git a/SRC/slasd6.f b/SRC/slasd6.f
index 53c1f156..26561f33 100644
--- a/SRC/slasd6.f
+++ b/SRC/slasd6.f
@@ -118,8 +118,7 @@
 *>          SQRE is INTEGER
 *>         = 0: the lower block is an NR-by-NR square matrix.
 *>         = 1: the lower block is an NR-by-(NR+1) rectangular matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>         The bidiagonal matrix has row dimension N = NL + NR + 1,
 *>         and column dimension M = N + SQRE.
 *> \endverbatim
@@ -239,12 +238,10 @@
 *>         On exit, DIFR(I, 1) is the distance between I-th updated
 *>         (undeflated) singular value and the I+1-th (undeflated) old
 *>         singular value.
-*> \endverbatim
-*> \verbatim
+*>
 *>         If ICOMPQ = 1, DIFR(1:K,2) is an array containing the
 *>         normalizing factors for the right singular vector matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>         See SLASD8 for details on DIFL and DIFR.
 *> \endverbatim
 *>
diff --git a/SRC/slasd7.f b/SRC/slasd7.f
index 12da5cae..4f5e4615 100644
--- a/SRC/slasd7.f
+++ b/SRC/slasd7.f
@@ -83,8 +83,7 @@
 *>          SQRE is INTEGER
 *>         = 0: the lower block is an NR-by-NR square matrix.
 *>         = 1: the lower block is an NR-by-(NR+1) rectangular matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>         The bidiagonal matrix has
 *>         N = NL + NR + 1 rows and
 *>         M = N + SQRE >= N columns.
diff --git a/SRC/slasd8.f b/SRC/slasd8.f
index 74798899..21dc6e04 100644
--- a/SRC/slasd8.f
+++ b/SRC/slasd8.f
@@ -111,8 +111,7 @@
 *>                   dimension ( K ) if ICOMPQ = 0.
 *>          On exit, DIFR(I,1) = D(I) - DSIGMA(I+1), DIFR(K,1) is not
 *>          defined and will not be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If ICOMPQ = 1, DIFR(1:K,2) is an array containing the
 *>          normalizing factors for the right singular vector matrix.
 *> \endverbatim
diff --git a/SRC/slasdq.f b/SRC/slasdq.f
index ca266921..1157a482 100644
--- a/SRC/slasdq.f
+++ b/SRC/slasdq.f
@@ -72,8 +72,7 @@
 *>        = 0: then the input matrix is N-by-N.
 *>        = 1: then the input matrix is N-by-(N+1) if UPLU = 'U' and
 *>             (N+1)-by-N if UPLU = 'L'.
-*> \endverbatim
-*> \verbatim
+*>
 *>        The bidiagonal matrix has
 *>        N = NL + NR + 1 rows and
 *>        M = N + SQRE >= N columns.
diff --git a/SRC/slaset.f b/SRC/slaset.f
index 9532c444..ff5820e9 100644
--- a/SRC/slaset.f
+++ b/SRC/slaset.f
@@ -82,13 +82,11 @@
 *> \verbatim
 *>          A is REAL array, dimension (LDA,N)
 *>          On exit, the leading m-by-n submatrix of A is set as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>          if UPLO = 'U', A(i,j) = ALPHA, 1<=i<=j-1, 1<=j<=n,
 *>          if UPLO = 'L', A(i,j) = ALPHA, j+1<=i<=m, 1<=j<=n,
 *>          otherwise,     A(i,j) = ALPHA, 1<=i<=m, 1<=j<=n, i.ne.j,
-*> \endverbatim
-*> \verbatim
+*>
 *>          and, for all UPLO, A(i,i) = BETA, 1<=i<=min(m,n).
 *> \endverbatim
 *>
diff --git a/SRC/slasq3.f b/SRC/slasq3.f
index cbe8ce42..6f285836 100644
--- a/SRC/slasq3.f
+++ b/SRC/slasq3.f
@@ -161,8 +161,7 @@
 *> \param[in,out] TAU
 *> \verbatim
 *>          TAU is REAL
-*> \endverbatim
-*> \verbatim
+*>
 *>         These are passed as arguments in order to save their values
 *>         between calls to SLASQ3.
 *> \endverbatim
diff --git a/SRC/slasyf.f b/SRC/slasyf.f
index 33f60c8f..832f6a24 100644
--- a/SRC/slasyf.f
+++ b/SRC/slasyf.f
@@ -112,8 +112,7 @@
 *>          Details of the interchanges and the block structure of D.
 *>          If UPLO = 'U', only the last KB elements of IPIV are set;
 *>          if UPLO = 'L', only the first KB elements are set.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If IPIV(k) > 0, then rows and columns k and IPIV(k) were
 *>          interchanged and D(k,k) is a 1-by-1 diagonal block.
 *>          If UPLO = 'U' and IPIV(k) = IPIV(k-1) < 0, then rows and
diff --git a/SRC/slatbs.f b/SRC/slatbs.f
index 4cb166b4..09745c2e 100644
--- a/SRC/slatbs.f
+++ b/SRC/slatbs.f
@@ -136,15 +136,13 @@
 *> \param[in,out] CNORM
 *> \verbatim
 *>          CNORM is or output) REAL array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'Y', CNORM is an input argument and CNORM(j)
 *>          contains the norm of the off-diagonal part of the j-th column
 *>          of A.  If TRANS = 'N', CNORM(j) must be greater than or equal
 *>          to the infinity-norm, and if TRANS = 'T' or 'C', CNORM(j)
 *>          must be greater than or equal to the 1-norm.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'N', CNORM is an output argument and CNORM(j)
 *>          returns the 1-norm of the offdiagonal part of the j-th column
 *>          of A.
diff --git a/SRC/slatps.f b/SRC/slatps.f
index 1a64ce7b..9e84b1db 100644
--- a/SRC/slatps.f
+++ b/SRC/slatps.f
@@ -123,15 +123,13 @@
 *> \param[in,out] CNORM
 *> \verbatim
 *>          CNORM is or output) REAL array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'Y', CNORM is an input argument and CNORM(j)
 *>          contains the norm of the off-diagonal part of the j-th column
 *>          of A.  If TRANS = 'N', CNORM(j) must be greater than or equal
 *>          to the infinity-norm, and if TRANS = 'T' or 'C', CNORM(j)
 *>          must be greater than or equal to the 1-norm.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'N', CNORM is an output argument and CNORM(j)
 *>          returns the 1-norm of the offdiagonal part of the j-th column
 *>          of A.
diff --git a/SRC/slatrs.f b/SRC/slatrs.f
index 72e44a46..59bf0eca 100644
--- a/SRC/slatrs.f
+++ b/SRC/slatrs.f
@@ -132,15 +132,13 @@
 *> \param[in,out] CNORM
 *> \verbatim
 *>          CNORM is or output) REAL array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'Y', CNORM is an input argument and CNORM(j)
 *>          contains the norm of the off-diagonal part of the j-th column
 *>          of A.  If TRANS = 'N', CNORM(j) must be greater than or equal
 *>          to the infinity-norm, and if TRANS = 'T' or 'C', CNORM(j)
 *>          must be greater than or equal to the 1-norm.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'N', CNORM is an output argument and CNORM(j)
 *>          returns the 1-norm of the offdiagonal part of the j-th column
 *>          of A.
diff --git a/SRC/slatzm.f b/SRC/slatzm.f
index 178c9cee..40d48643 100644
--- a/SRC/slatzm.f
+++ b/SRC/slatzm.f
@@ -107,8 +107,7 @@
 *>                         (M,1)   if SIDE = 'R'
 *>          On entry, the n-vector C1 if SIDE = 'L', or the m-vector C1
 *>          if SIDE = 'R'.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the first row of P*C if SIDE = 'L', or the first
 *>          column of C*P if SIDE = 'R'.
 *> \endverbatim
@@ -120,8 +119,7 @@
 *>                         (LDC, N-1) if SIDE = 'R'
 *>          On entry, the (m - 1) x n matrix C2 if SIDE = 'L', or the
 *>          m x (n - 1) matrix C2 if SIDE = 'R'.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, rows 2:m of P*C if SIDE = 'L', or columns 2:m of C*P
 *>          if SIDE = 'R'.
 *> \endverbatim
diff --git a/SRC/sorbdb.f b/SRC/sorbdb.f
index 9f0428c6..a298ec85 100644
--- a/SRC/sorbdb.f
+++ b/SRC/sorbdb.f
@@ -234,8 +234,7 @@
 *> \verbatim
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK. LWORK >= M-Q.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sorcsd.f b/SRC/sorcsd.f
index 1051a680..154342ca 100644
--- a/SRC/sorcsd.f
+++ b/SRC/sorcsd.f
@@ -255,8 +255,7 @@
 *> \verbatim
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the work array, and no error
@@ -275,12 +274,10 @@
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value.
 *>          > 0:  SBBCSD did not converge. See the description of WORK
 *>                above for details.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Reference
 *>  =========
-*> \endverbatim
-*> \verbatim
+*>
 *>  [1] Brian D. Sutton. Computing the complete CS decomposition. Numer.
 *>      Algorithms, 50(1):33-65, 2009.
 *> \endverbatim
diff --git a/SRC/sorgbr.f b/SRC/sorgbr.f
index 2c467008..2dbb47f2 100644
--- a/SRC/sorgbr.f
+++ b/SRC/sorgbr.f
@@ -129,8 +129,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,min(M,N)).
 *>          For optimum performance LWORK >= min(M,N)*NB, where NB
 *>          is the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sorghr.f b/SRC/sorghr.f
index 80087ebc..3bfa6847 100644
--- a/SRC/sorghr.f
+++ b/SRC/sorghr.f
@@ -58,8 +58,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          ILO and IHI must have the same values as in the previous call
 *>          of SGEHRD. Q is equal to the unit matrix except in the
 *>          submatrix Q(ilo+1:ihi,ilo+1:ihi).
@@ -99,8 +98,7 @@
 *>          The dimension of the array WORK. LWORK >= IHI-ILO.
 *>          For optimum performance LWORK >= (IHI-ILO)*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sorglq.f b/SRC/sorglq.f
index 736f3ceb..1fa482dc 100644
--- a/SRC/sorglq.f
+++ b/SRC/sorglq.f
@@ -99,8 +99,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,M).
 *>          For optimum performance LWORK >= M*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sorgql.f b/SRC/sorgql.f
index 707b71c6..fa0130e9 100644
--- a/SRC/sorgql.f
+++ b/SRC/sorgql.f
@@ -100,8 +100,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is the
 *>          optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sorgqr.f b/SRC/sorgqr.f
index 02241fea..524fe8a9 100644
--- a/SRC/sorgqr.f
+++ b/SRC/sorgqr.f
@@ -100,8 +100,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is the
 *>          optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sorgrq.f b/SRC/sorgrq.f
index 053f1b19..2784dcfc 100644
--- a/SRC/sorgrq.f
+++ b/SRC/sorgrq.f
@@ -100,8 +100,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,M).
 *>          For optimum performance LWORK >= M*NB, where NB is the
 *>          optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sorgtr.f b/SRC/sorgtr.f
index d2c99aaf..f7842156 100644
--- a/SRC/sorgtr.f
+++ b/SRC/sorgtr.f
@@ -95,8 +95,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,N-1).
 *>          For optimum performance LWORK >= (N-1)*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sormbr.f b/SRC/sormbr.f
index f225c663..36a80d58 100644
--- a/SRC/sormbr.f
+++ b/SRC/sormbr.f
@@ -167,8 +167,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sormhr.f b/SRC/sormhr.f
index 88fd9687..ca87aa29 100644
--- a/SRC/sormhr.f
+++ b/SRC/sormhr.f
@@ -87,8 +87,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          ILO and IHI must have the same values as in the previous call
 *>          of SGEHRD. Q is equal to the unit matrix except in the
 *>          submatrix Q(ilo+1:ihi,ilo+1:ihi).
@@ -151,8 +150,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sormlq.f b/SRC/sormlq.f
index 31b5a4cb..10797f47 100644
--- a/SRC/sormlq.f
+++ b/SRC/sormlq.f
@@ -142,8 +142,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sormql.f b/SRC/sormql.f
index 8e891a47..825fbdd0 100644
--- a/SRC/sormql.f
+++ b/SRC/sormql.f
@@ -142,8 +142,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sormqr.f b/SRC/sormqr.f
index b3fd9d3d..127ed656 100644
--- a/SRC/sormqr.f
+++ b/SRC/sormqr.f
@@ -142,8 +142,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sormrq.f b/SRC/sormrq.f
index 97bb34d3..0b1e7ad8 100644
--- a/SRC/sormrq.f
+++ b/SRC/sormrq.f
@@ -142,8 +142,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sormrz.f b/SRC/sormrz.f
index dcfb55a8..2b3d48f7 100644
--- a/SRC/sormrz.f
+++ b/SRC/sormrz.f
@@ -149,8 +149,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/sormtr.f b/SRC/sormtr.f
index 3b7fbf46..a309f0c9 100644
--- a/SRC/sormtr.f
+++ b/SRC/sormtr.f
@@ -143,8 +143,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/spbrfs.f b/SRC/spbrfs.f
index 2c8dea1a..5ce6964e 100644
--- a/SRC/spbrfs.f
+++ b/SRC/spbrfs.f
@@ -165,12 +165,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/spbsv.f b/SRC/spbsv.f
index cae04b48..37c076ca 100644
--- a/SRC/spbsv.f
+++ b/SRC/spbsv.f
@@ -90,8 +90,7 @@
 *>          if UPLO = 'U', AB(KD+1+i-j,j) = A(i,j) for max(1,j-KD)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(N,j+KD).
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**T*U or A = L*L**T of the band
 *>          matrix A, in the same storage format as A.
diff --git a/SRC/spbsvx.f b/SRC/spbsvx.f
index 6b13e133..9c5c9e76 100644
--- a/SRC/spbsvx.f
+++ b/SRC/spbsvx.f
@@ -146,8 +146,7 @@
 *>          if UPLO = 'U', AB(KD+1+i-j,j) = A(i,j) for max(1,j-KD)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(N,j+KD).
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>          diag(S)*A*diag(S).
 *> \endverbatim
@@ -166,13 +165,11 @@
 *>          factorization A = U**T*U or A = L*L**T of the band matrix
 *>          A, in the same storage format as A (see AB).  If EQUED = 'Y',
 *>          then AFB is the factored form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AFB is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**T*U or A = L*L**T.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then AFB is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**T*U or A = L*L**T of the equilibrated
diff --git a/SRC/spbtf2.f b/SRC/spbtf2.f
index 1012ed4d..79c63c73 100644
--- a/SRC/spbtf2.f
+++ b/SRC/spbtf2.f
@@ -81,8 +81,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**T*U or A = L*L**T of the band
 *>          matrix A, in the same storage format as A.
diff --git a/SRC/spbtrf.f b/SRC/spbtrf.f
index 4d4acc1b..5fbe3add 100644
--- a/SRC/spbtrf.f
+++ b/SRC/spbtrf.f
@@ -76,8 +76,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**T*U or A = L*L**T of the band
 *>          matrix A, in the same storage format as A.
diff --git a/SRC/spftrf.f b/SRC/spftrf.f
index e01d171f..9902166a 100644
--- a/SRC/spftrf.f
+++ b/SRC/spftrf.f
@@ -82,8 +82,7 @@
 *>          of lower packed A. The LDA of RFP A is (N+1)/2 when TRANSR =
 *>          'T'. When TRANSR is 'N' the LDA is N+1 when N is even and N
 *>          is odd. See the Note below for more details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization RFP A = U**T*U or RFP A = L*L**T.
 *> \endverbatim
diff --git a/SRC/spftri.f b/SRC/spftri.f
index c5bb5118..14dff1ac 100644
--- a/SRC/spftri.f
+++ b/SRC/spftri.f
@@ -76,8 +76,7 @@
 *>          of lower packed A. The LDA of RFP A is (N+1)/2 when TRANSR =
 *>          'T'. When TRANSR is 'N' the LDA is N+1 when N is even and N
 *>          is odd. See the Note below for more details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the symmetric inverse of the original matrix, in the
 *>          same storage format.
 *> \endverbatim
diff --git a/SRC/sporfs.f b/SRC/sporfs.f
index 43a7adfc..8d837ed1 100644
--- a/SRC/sporfs.f
+++ b/SRC/sporfs.f
@@ -159,12 +159,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/sporfsx.f b/SRC/sporfsx.f
index bc0df788..8cca00e6 100644
--- a/SRC/sporfsx.f
+++ b/SRC/sporfsx.f
@@ -211,37 +211,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -250,8 +244,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -262,14 +255,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -277,26 +268,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -307,8 +294,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -327,8 +313,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0
@@ -339,8 +324,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -350,8 +334,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/sposv.f b/SRC/sposv.f
index 89a5f16f..8b6394a0 100644
--- a/SRC/sposv.f
+++ b/SRC/sposv.f
@@ -82,8 +82,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization A = U**T*U or A = L*L**T.
 *> \endverbatim
diff --git a/SRC/sposvx.f b/SRC/sposvx.f
index cdc16d7b..f8ec2392 100644
--- a/SRC/sposvx.f
+++ b/SRC/sposvx.f
@@ -141,8 +141,7 @@
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.  A is not modified if
 *>          FACT = 'F' or 'N', or if FACT = 'E' and EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>          diag(S)*A*diag(S).
 *> \endverbatim
@@ -161,14 +160,12 @@
 *>          factorization A = U**T*U or A = L*L**T, in the same storage
 *>          format as A.  If EQUED .ne. 'N', then AF is the factored form
 *>          of the equilibrated matrix diag(S)*A*diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AF is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**T*U or A = L*L**T of the original
 *>          matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then AF is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**T*U or A = L*L**T of the equilibrated
diff --git a/SRC/sposvxx.f b/SRC/sposvxx.f
index 6f31c410..f04e61eb 100644
--- a/SRC/sposvxx.f
+++ b/SRC/sposvxx.f
@@ -168,8 +168,7 @@
 *>     the strictly upper triangular part of A is not referenced.  A is
 *>     not modified if FACT = 'F' or 'N', or if FACT = 'E' and EQUED =
 *>     'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>     On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>     diag(S)*A*diag(S).
 *> \endverbatim
@@ -188,14 +187,12 @@
 *>     factorization A = U**T*U or A = L*L**T, in the same storage
 *>     format as A.  If EQUED .ne. 'N', then AF is the factored
 *>     form of the equilibrated matrix diag(S)*A*diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then AF is an output argument and on exit
 *>     returns the triangular factor U or L from the Cholesky
 *>     factorization A = U**T*U or A = L*L**T of the original
 *>     matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'E', then AF is an output argument and on exit
 *>     returns the triangular factor U or L from the Cholesky
 *>     factorization A = U**T*U or A = L*L**T of the equilibrated
@@ -314,37 +311,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -353,8 +344,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -365,14 +355,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -380,26 +368,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -410,8 +394,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -430,8 +413,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0
@@ -442,8 +424,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -453,8 +434,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/spotf2.f b/SRC/spotf2.f
index d1b6453f..9cf510e7 100644
--- a/SRC/spotf2.f
+++ b/SRC/spotf2.f
@@ -74,8 +74,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization A = U**T *U  or A = L*L**T.
 *> \endverbatim
diff --git a/SRC/spotrf.f b/SRC/spotrf.f
index c010bd76..865fcca0 100644
--- a/SRC/spotrf.f
+++ b/SRC/spotrf.f
@@ -72,8 +72,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization A = U**T*U or A = L*L**T.
 *> \endverbatim
diff --git a/SRC/spprfs.f b/SRC/spprfs.f
index 5b25215a..53fa3b8a 100644
--- a/SRC/spprfs.f
+++ b/SRC/spprfs.f
@@ -147,12 +147,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/sppsv.f b/SRC/sppsv.f
index 7245392d..192ff675 100644
--- a/SRC/sppsv.f
+++ b/SRC/sppsv.f
@@ -81,8 +81,7 @@
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
 *>          See below for further details.  
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization A = U**T*U or A = L*L**T, in the same storage
 *>          format as A.
diff --git a/SRC/sppsvx.f b/SRC/sppsvx.f
index cfa1a948..9841eaae 100644
--- a/SRC/sppsvx.f
+++ b/SRC/sppsvx.f
@@ -138,8 +138,7 @@
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
 *>          See below for further details.  A is not modified if
 *>          FACT = 'F' or 'N', or if FACT = 'E' and EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>          diag(S)*A*diag(S).
 *> \endverbatim
@@ -153,14 +152,12 @@
 *>          factorization A = U**T*U or A = L*L**T, in the same storage
 *>          format as A.  If EQUED .ne. 'N', then AFP is the factored
 *>          form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AFP is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**T * U or A = L * L**T of the original
 *>          matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then AFP is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**T * U or A = L * L**T of the equilibrated
diff --git a/SRC/spptrf.f b/SRC/spptrf.f
index 89766442..70ffdc5d 100644
--- a/SRC/spptrf.f
+++ b/SRC/spptrf.f
@@ -69,8 +69,7 @@
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**T*U or A = L*L**T, in the same
 *>          storage format as A.
diff --git a/SRC/spptri.f b/SRC/spptri.f
index 092ef9af..444ca85d 100644
--- a/SRC/spptri.f
+++ b/SRC/spptri.f
@@ -65,8 +65,7 @@
 *>          array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = U(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = L(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the upper or lower triangle of the (symmetric)
 *>          inverse of A, overwriting the input factor U or L.
 *> \endverbatim
diff --git a/SRC/spstf2.f b/SRC/spstf2.f
index 81efae7d..3d43eb70 100644
--- a/SRC/spstf2.f
+++ b/SRC/spstf2.f
@@ -78,8 +78,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization as above.
 *> \endverbatim
diff --git a/SRC/spstrf.f b/SRC/spstrf.f
index 4fe1ca10..72377370 100644
--- a/SRC/spstrf.f
+++ b/SRC/spstrf.f
@@ -78,8 +78,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization as above.
 *> \endverbatim
diff --git a/SRC/sptrfs.f b/SRC/sptrfs.f
index c109007e..402c80ed 100644
--- a/SRC/sptrfs.f
+++ b/SRC/sptrfs.f
@@ -139,12 +139,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/ssbev.f b/SRC/ssbev.f
index 780bab5e..82e0da37 100644
--- a/SRC/ssbev.f
+++ b/SRC/ssbev.f
@@ -79,8 +79,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AB is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the first
 *>          superdiagonal and the diagonal of the tridiagonal matrix T
diff --git a/SRC/ssbevd.f b/SRC/ssbevd.f
index bdd4e957..64645f72 100644
--- a/SRC/ssbevd.f
+++ b/SRC/ssbevd.f
@@ -88,8 +88,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AB is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the first
 *>          superdiagonal and the diagonal of the tridiagonal matrix T
@@ -141,8 +140,7 @@
 *>          If JOBZ  = 'N' and N > 2, LWORK must be at least 2*N.
 *>          If JOBZ  = 'V' and N > 2, LWORK must be at least
 *>                         ( 1 + 5*N + 2*N**2 ).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK and IWORK
 *>          arrays, returns these values as the first entries of the WORK
@@ -162,8 +160,7 @@
 *>          The dimension of the array LIWORK.
 *>          If JOBZ  = 'N' or N <= 1, LIWORK must be at least 1.
 *>          If JOBZ  = 'V' and N > 2, LIWORK must be at least 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK and
 *>          IWORK arrays, returns these values as the first entries of
diff --git a/SRC/ssbevx.f b/SRC/ssbevx.f
index 8e50f4c6..28cf7a3a 100644
--- a/SRC/ssbevx.f
+++ b/SRC/ssbevx.f
@@ -94,8 +94,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AB is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the first
 *>          superdiagonal and the diagonal of the tridiagonal matrix T
@@ -159,24 +158,20 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing AB to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*SLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
 *>          eigenvectors did not converge, try setting ABSTOL to
 *>          2*SLAMCH('S').
-*> \endverbatim
-*> \verbatim
+*>
 *>          See "Computing Small Singular Values of Bidiagonal Matrices
 *>          with Guaranteed High Relative Accuracy," by Demmel and
 *>          Kahan, LAPACK Working Note #3.
diff --git a/SRC/ssbgst.f b/SRC/ssbgst.f
index 903884f8..2bd5dd0d 100644
--- a/SRC/ssbgst.f
+++ b/SRC/ssbgst.f
@@ -93,8 +93,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(ka+1+i-j,j) = A(i,j) for max(1,j-ka)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+ka).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the transformed matrix X**T*A*X, stored in the same
 *>          format as A.
 *> \endverbatim
diff --git a/SRC/ssbgv.f b/SRC/ssbgv.f
index 2916d66d..2f79900c 100644
--- a/SRC/ssbgv.f
+++ b/SRC/ssbgv.f
@@ -89,8 +89,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(ka+1+i-j,j) = A(i,j) for max(1,j-ka)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+ka).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AB are destroyed.
 *> \endverbatim
 *>
@@ -109,8 +108,7 @@
 *>          as follows:
 *>          if UPLO = 'U', BB(kb+1+i-j,j) = B(i,j) for max(1,j-kb)<=i<=j;
 *>          if UPLO = 'L', BB(1+i-j,j)    = B(i,j) for j<=i<=min(n,j+kb).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the factor S from the split Cholesky factorization
 *>          B = S**T*S, as returned by SPBSTF.
 *> \endverbatim
diff --git a/SRC/ssbgvd.f b/SRC/ssbgvd.f
index c133b6c5..28b96642 100644
--- a/SRC/ssbgvd.f
+++ b/SRC/ssbgvd.f
@@ -98,8 +98,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(ka+1+i-j,j) = A(i,j) for max(1,j-ka)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+ka).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AB are destroyed.
 *> \endverbatim
 *>
@@ -118,8 +117,7 @@
 *>          as follows:
 *>          if UPLO = 'U', BB(ka+1+i-j,j) = B(i,j) for max(1,j-kb)<=i<=j;
 *>          if UPLO = 'L', BB(1+i-j,j)    = B(i,j) for j<=i<=min(n,j+kb).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the factor S from the split Cholesky factorization
 *>          B = S**T*S, as returned by SPBSTF.
 *> \endverbatim
@@ -166,8 +164,7 @@
 *>          If N <= 1,               LWORK >= 1.
 *>          If JOBZ = 'N' and N > 1, LWORK >= 3*N.
 *>          If JOBZ = 'V' and N > 1, LWORK >= 1 + 5*N + 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK and IWORK
 *>          arrays, returns these values as the first entries of the WORK
@@ -187,8 +184,7 @@
 *>          The dimension of the array IWORK.
 *>          If JOBZ  = 'N' or N <= 1, LIWORK >= 1.
 *>          If JOBZ  = 'V' and N > 1, LIWORK >= 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK and
 *>          IWORK arrays, returns these values as the first entries of
diff --git a/SRC/ssbgvx.f b/SRC/ssbgvx.f
index ee291c69..747b9a6f 100644
--- a/SRC/ssbgvx.f
+++ b/SRC/ssbgvx.f
@@ -104,8 +104,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(ka+1+i-j,j) = A(i,j) for max(1,j-ka)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+ka).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AB are destroyed.
 *> \endverbatim
 *>
@@ -124,8 +123,7 @@
 *>          as follows:
 *>          if UPLO = 'U', BB(ka+1+i-j,j) = B(i,j) for max(1,j-kb)<=i<=j;
 *>          if UPLO = 'L', BB(1+i-j,j)    = B(i,j) for j<=i<=min(n,j+kb).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the factor S from the split Cholesky factorization
 *>          B = S**T*S, as returned by SPBSTF.
 *> \endverbatim
@@ -160,8 +158,7 @@
 *> \param[in] VU
 *> \verbatim
 *>          VU is REAL
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='V', the lower and upper bounds of the interval to
 *>          be searched for eigenvalues. VL < VU.
 *>          Not referenced if RANGE = 'A' or 'I'.
@@ -175,8 +172,7 @@
 *> \param[in] IU
 *> \verbatim
 *>          IU is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='I', the indices (in ascending order) of the
 *>          smallest and largest eigenvalues to be returned.
 *>          1 <= IL <= IU <= N, if N > 0; IL = 1 and IU = 0 if N = 0.
@@ -190,17 +186,14 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing A to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*SLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
diff --git a/SRC/ssbtrd.f b/SRC/ssbtrd.f
index 9ca9559e..51e0d200 100644
--- a/SRC/ssbtrd.f
+++ b/SRC/ssbtrd.f
@@ -114,8 +114,7 @@
 *>          Q is REAL array, dimension (LDQ,N)
 *>          On entry, if VECT = 'U', then Q must contain an N-by-N
 *>          matrix X; if VECT = 'N' or 'V', then Q need not be set.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit:
 *>          if VECT = 'V', Q contains the N-by-N orthogonal matrix Q;
 *>          if VECT = 'U', Q contains the product X*Q;
diff --git a/SRC/ssfrk.f b/SRC/ssfrk.f
index 895b1135..f2754021 100644
--- a/SRC/ssfrk.f
+++ b/SRC/ssfrk.f
@@ -68,16 +68,13 @@
 *>           On  entry, UPLO specifies whether the upper or lower
 *>           triangular part of the array C is to be referenced as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'U' or 'u'   Only the upper triangular part of C
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'L' or 'l'   Only the lower triangular part of C
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -86,14 +83,11 @@
 *>          TRANS is CHARACTER*1
 *>           On entry, TRANS specifies the operation to be performed as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              TRANS = 'N' or 'n'   C := alpha*A*A**T + beta*C.
-*> \endverbatim
-*> \verbatim
+*>
 *>              TRANS = 'T' or 't'   C := alpha*A**T*A + beta*C.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/sspev.f b/SRC/sspev.f
index aeea31ee..e87bee32 100644
--- a/SRC/sspev.f
+++ b/SRC/sspev.f
@@ -70,8 +70,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AP is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the diagonal
 *>          and first superdiagonal of the tridiagonal matrix T overwrite
diff --git a/SRC/sspevd.f b/SRC/sspevd.f
index 19e37223..3abbc72a 100644
--- a/SRC/sspevd.f
+++ b/SRC/sspevd.f
@@ -80,8 +80,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AP is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the diagonal
 *>          and first superdiagonal of the tridiagonal matrix T overwrite
@@ -126,8 +125,7 @@
 *>          If JOBZ = 'N' and N > 1, LWORK must be at least 2*N.
 *>          If JOBZ = 'V' and N > 1, LWORK must be at least
 *>                                                 1 + 6*N + N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the required sizes of the WORK and IWORK
 *>          arrays, returns these values as the first entries of the WORK
@@ -147,8 +145,7 @@
 *>          The dimension of the array IWORK.
 *>          If JOBZ  = 'N' or N <= 1, LIWORK must be at least 1.
 *>          If JOBZ  = 'V' and N > 1, LIWORK must be at least 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the required sizes of the WORK and
 *>          IWORK arrays, returns these values as the first entries of
diff --git a/SRC/sspevx.f b/SRC/sspevx.f
index d8415bbc..bb059b7f 100644
--- a/SRC/sspevx.f
+++ b/SRC/sspevx.f
@@ -85,8 +85,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AP is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the diagonal
 *>          and first superdiagonal of the tridiagonal matrix T overwrite
@@ -129,24 +128,20 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing AP to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*SLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
 *>          eigenvectors did not converge, try setting ABSTOL to
 *>          2*SLAMCH('S').
-*> \endverbatim
-*> \verbatim
+*>
 *>          See "Computing Small Singular Values of Bidiagonal Matrices
 *>          with Guaranteed High Relative Accuracy," by Demmel and
 *>          Kahan, LAPACK Working Note #3.
diff --git a/SRC/sspgst.f b/SRC/sspgst.f
index d0ff02e3..d011afc7 100644
--- a/SRC/sspgst.f
+++ b/SRC/sspgst.f
@@ -80,8 +80,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the transformed matrix, stored in the
 *>          same format as A.
 *> \endverbatim
diff --git a/SRC/sspgv.f b/SRC/sspgv.f
index 5e3c4af5..27be82a4 100644
--- a/SRC/sspgv.f
+++ b/SRC/sspgv.f
@@ -85,8 +85,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AP are destroyed.
 *> \endverbatim
 *>
@@ -98,8 +97,7 @@
 *>          is stored in the array BP as follows:
 *>          if UPLO = 'U', BP(i + (j-1)*j/2) = B(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', BP(i + (j-1)*(2*n-j)/2) = B(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the triangular factor U or L from the Cholesky
 *>          factorization B = U**T*U or B = L*L**T, in the same storage
 *>          format as B.
diff --git a/SRC/sspgvd.f b/SRC/sspgvd.f
index c9edcf57..a03b3582 100644
--- a/SRC/sspgvd.f
+++ b/SRC/sspgvd.f
@@ -93,8 +93,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AP are destroyed.
 *> \endverbatim
 *>
@@ -106,8 +105,7 @@
 *>          is stored in the array BP as follows:
 *>          if UPLO = 'U', BP(i + (j-1)*j/2) = B(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', BP(i + (j-1)*(2*n-j)/2) = B(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the triangular factor U or L from the Cholesky
 *>          factorization B = U**T*U or B = L*L**T, in the same storage
 *>          format as B.
@@ -149,8 +147,7 @@
 *>          If N <= 1,               LWORK >= 1.
 *>          If JOBZ = 'N' and N > 1, LWORK >= 2*N.
 *>          If JOBZ = 'V' and N > 1, LWORK >= 1 + 6*N + 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the required sizes of the WORK and IWORK
 *>          arrays, returns these values as the first entries of the WORK
@@ -170,8 +167,7 @@
 *>          The dimension of the array IWORK.
 *>          If JOBZ  = 'N' or N <= 1, LIWORK >= 1.
 *>          If JOBZ  = 'V' and N > 1, LIWORK >= 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the required sizes of the WORK and
 *>          IWORK arrays, returns these values as the first entries of
diff --git a/SRC/sspgvx.f b/SRC/sspgvx.f
index c135485f..2da83251 100644
--- a/SRC/sspgvx.f
+++ b/SRC/sspgvx.f
@@ -98,8 +98,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AP are destroyed.
 *> \endverbatim
 *>
@@ -111,8 +110,7 @@
 *>          is stored in the array BP as follows:
 *>          if UPLO = 'U', BP(i + (j-1)*j/2) = B(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', BP(i + (j-1)*(2*n-j)/2) = B(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the triangular factor U or L from the Cholesky
 *>          factorization B = U**T*U or B = L*L**T, in the same storage
 *>          format as B.
@@ -126,8 +124,7 @@
 *> \param[in] VU
 *> \verbatim
 *>          VU is REAL
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='V', the lower and upper bounds of the interval to
 *>          be searched for eigenvalues. VL < VU.
 *>          Not referenced if RANGE = 'A' or 'I'.
@@ -141,8 +138,7 @@
 *> \param[in] IU
 *> \verbatim
 *>          IU is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='I', the indices (in ascending order) of the
 *>          smallest and largest eigenvalues to be returned.
 *>          1 <= IL <= IU <= N, if N > 0; IL = 1 and IU = 0 if N = 0.
@@ -156,17 +152,14 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing A to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*SLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
@@ -199,8 +192,7 @@
 *>          The eigenvectors are normalized as follows:
 *>          if ITYPE = 1 or 2, Z**T*B*Z = I;
 *>          if ITYPE = 3, Z**T*inv(B)*Z = I.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If an eigenvector fails to converge, then that column of Z
 *>          contains the latest approximation to the eigenvector, and the
 *>          index of the eigenvector is returned in IFAIL.
diff --git a/SRC/ssprfs.f b/SRC/ssprfs.f
index 2bf5ef41..8d27b39b 100644
--- a/SRC/ssprfs.f
+++ b/SRC/ssprfs.f
@@ -155,12 +155,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/sspsv.f b/SRC/sspsv.f
index f790cf4d..72fb0620 100644
--- a/SRC/sspsv.f
+++ b/SRC/sspsv.f
@@ -83,8 +83,7 @@
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
 *>          A = U*D*U**T or A = L*D*L**T as computed by SSPTRF, stored as
diff --git a/SRC/sspsvx.f b/SRC/sspsvx.f
index a1d66e2c..aa821581 100644
--- a/SRC/sspsvx.f
+++ b/SRC/sspsvx.f
@@ -128,8 +128,7 @@
 *>          to obtain the factor U or L from the factorization
 *>          A = U*D*U**T or A = L*D*L**T as computed by SSPTRF, stored as
 *>          a packed triangular matrix in the same storage format as A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AFP is an output argument and on exit
 *>          contains the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
@@ -150,8 +149,7 @@
 *>          is a 2-by-2 diagonal block.  If UPLO = 'L' and IPIV(k) =
 *>          IPIV(k+1) < 0, then rows and columns k+1 and -IPIV(k) were
 *>          interchanged and D(k:k+1,k:k+1) is a 2-by-2 diagonal block.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains details of the interchanges and the block structure
 *>          of D, as determined by SSPTRF.
diff --git a/SRC/ssptrf.f b/SRC/ssptrf.f
index 72525c91..9f127306 100644
--- a/SRC/ssptrf.f
+++ b/SRC/ssptrf.f
@@ -70,8 +70,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L, stored as a packed triangular
 *>          matrix overwriting A (see below for further details).
diff --git a/SRC/ssptri.f b/SRC/ssptri.f
index 0bc33a75..b17a0730 100644
--- a/SRC/ssptri.f
+++ b/SRC/ssptri.f
@@ -65,8 +65,7 @@
 *>          On entry, the block diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by SSPTRF,
 *>          stored as a packed triangular matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix, stored as a packed triangular matrix. The j-th column
 *>          of inv(A) is stored in the array AP as follows:
diff --git a/SRC/sstebz.f b/SRC/sstebz.f
index 7d1ace0a..41c1045f 100644
--- a/SRC/sstebz.f
+++ b/SRC/sstebz.f
@@ -93,8 +93,7 @@
 *> \param[in] VU
 *> \verbatim
 *>          VU is REAL
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='V', the lower and upper bounds of the interval to
 *>          be searched for eigenvalues.  Eigenvalues less than or equal
 *>          to VL, or greater than VU, will not be returned.  VL < VU.
@@ -109,8 +108,7 @@
 *> \param[in] IU
 *> \verbatim
 *>          IU is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='I', the indices (in ascending order) of the
 *>          smallest and largest eigenvalues to be returned.
 *>          1 <= IL <= IU <= N, if N > 0; IL = 1 and IU = 0 if N = 0.
@@ -125,8 +123,7 @@
 *>          determined to lie in an interval whose width is ABSTOL or
 *>          less.  If ABSTOL is less than or equal to zero, then ULP*|T|
 *>          will be used, where |T| means the 1-norm of T.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*SLAMCH('S'), not zero.
 *> \endverbatim
@@ -229,19 +226,16 @@
 *>                                        floating-point arithmetic.
 *>                        Cure: Increase the PARAMETER "FUDGE",
 *>                              recompile, and try again.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  RELFAC  REAL, default = 2.0e0
 *>          The relative tolerance.  An interval (a,b] lies within
 *>          "relative tolerance" if  b-a < RELFAC*ulp*max(|a|,|b|),
 *>          where "ulp" is the machine precision (distance from 1 to
 *>          the next larger floating point number.)
-*> \endverbatim
-*> \verbatim
+*>
 *>  FUDGE   REAL, default = 2
 *>          A "fudge factor" to widen the Gershgorin intervals.  Ideally,
 *>          a value of 1 should work, but on machines with sloppy
diff --git a/SRC/sstedc.f b/SRC/sstedc.f
index d02a4849..5f689f06 100644
--- a/SRC/sstedc.f
+++ b/SRC/sstedc.f
@@ -124,8 +124,7 @@
 *>          Note that for COMPZ = 'I' or 'V', then if N is less than or
 *>          equal to the minimum divide size, usually 25, then LWORK need
 *>          only be max(1,2*(N-1)).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
@@ -150,8 +149,7 @@
 *>          Note that for COMPZ = 'I' or 'V', then if N is less than or
 *>          equal to the minimum divide size, usually 25, then LIWORK
 *>          need only be 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal size of the IWORK array,
 *>          returns this value as the first entry of the IWORK array, and
diff --git a/SRC/sstegr.f b/SRC/sstegr.f
index 792ec3b9..a00ed992 100644
--- a/SRC/sstegr.f
+++ b/SRC/sstegr.f
@@ -111,8 +111,7 @@
 *> \param[in] VU
 *> \verbatim
 *>          VU is REAL
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='V', the lower and upper bounds of the interval to
 *>          be searched for eigenvalues. VL < VU.
 *>          Not referenced if RANGE = 'A' or 'I'.
@@ -126,8 +125,7 @@
 *> \param[in] IU
 *> \verbatim
 *>          IU is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='I', the indices (in ascending order) of the
 *>          smallest and largest eigenvalues to be returned.
 *>          1 <= IL <= IU <= N, if N > 0.
diff --git a/SRC/sstein.f b/SRC/sstein.f
index aecec548..d2228479 100644
--- a/SRC/sstein.f
+++ b/SRC/sstein.f
@@ -145,16 +145,13 @@
 *>          > 0: if INFO = i, then i eigenvectors failed to converge
 *>               in MAXITS iterations.  Their indices are stored in
 *>               array IFAIL.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  MAXITS  INTEGER, default = 5
 *>          The maximum number of iterations performed.
-*> \endverbatim
-*> \verbatim
+*>
 *>  EXTRA   INTEGER, default = 2
 *>          The number of iterations performed after norm growth
 *>          criterion is satisfied, should be at least 1.
diff --git a/SRC/sstemr.f b/SRC/sstemr.f
index b2c13a81..e09c068a 100644
--- a/SRC/sstemr.f
+++ b/SRC/sstemr.f
@@ -142,8 +142,7 @@
 *> \param[in] VU
 *> \verbatim
 *>          VU is REAL
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='V', the lower and upper bounds of the interval to
 *>          be searched for eigenvalues. VL < VU.
 *>          Not referenced if RANGE = 'A' or 'I'.
@@ -157,8 +156,7 @@
 *> \param[in] IU
 *> \verbatim
 *>          IU is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='I', the indices (in ascending order) of the
 *>          smallest and largest eigenvalues to be returned.
 *>          1 <= IL <= IU <= N, if N > 0.
diff --git a/SRC/sstevd.f b/SRC/sstevd.f
index 07095df5..d2b423e7 100644
--- a/SRC/sstevd.f
+++ b/SRC/sstevd.f
@@ -111,8 +111,7 @@
 *>          If JOBZ  = 'N' or N <= 1 then LWORK must be at least 1.
 *>          If JOBZ  = 'V' and N > 1 then LWORK must be at least
 *>                         ( 1 + 4*N + N**2 ).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK and IWORK
 *>          arrays, returns these values as the first entries of the WORK
@@ -132,8 +131,7 @@
 *>          The dimension of the array IWORK.
 *>          If JOBZ  = 'N' or N <= 1 then LIWORK must be at least 1.
 *>          If JOBZ  = 'V' and N > 1 then LIWORK must be at least 3+5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK and
 *>          IWORK arrays, returns these values as the first entries of
diff --git a/SRC/sstevr.f b/SRC/sstevr.f
index 44192fa2..c55a9f2d 100644
--- a/SRC/sstevr.f
+++ b/SRC/sstevr.f
@@ -160,22 +160,18 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing A to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          See "Computing Small Singular Values of Bidiagonal Matrices
 *>          with Guaranteed High Relative Accuracy," by Demmel and
 *>          Kahan, LAPACK Working Note #3.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If high relative accuracy is important, set ABSTOL to
 *>          SLAMCH( 'Safe minimum' ).  Doing so will guarantee that
 *>          eigenvalues are computed to high relative accuracy when
@@ -242,8 +238,7 @@
 *> \verbatim
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.  LWORK >= 20*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK and IWORK
 *>          arrays, returns these values as the first entries of the WORK
@@ -262,8 +257,7 @@
 *> \verbatim
 *>          LIWORK is INTEGER
 *>          The dimension of the array IWORK.  LIWORK >= 10*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK and
 *>          IWORK arrays, returns these values as the first entries of
diff --git a/SRC/sstevx.f b/SRC/sstevx.f
index 215a6c30..42dd8d94 100644
--- a/SRC/sstevx.f
+++ b/SRC/sstevx.f
@@ -121,24 +121,20 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less
 *>          than or equal to zero, then  EPS*|T|  will be used in
 *>          its place, where |T| is the 1-norm of the tridiagonal
 *>          matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*SLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
 *>          eigenvectors did not converge, try setting ABSTOL to
 *>          2*SLAMCH('S').
-*> \endverbatim
-*> \verbatim
+*>
 *>          See "Computing Small Singular Values of Bidiagonal Matrices
 *>          with Guaranteed High Relative Accuracy," by Demmel and
 *>          Kahan, LAPACK Working Note #3.
diff --git a/SRC/ssyev.f b/SRC/ssyev.f
index 05f5219a..82ee4614 100644
--- a/SRC/ssyev.f
+++ b/SRC/ssyev.f
@@ -101,8 +101,7 @@
 *>          The length of the array WORK.  LWORK >= max(1,3*N-1).
 *>          For optimal efficiency, LWORK >= (NB+2)*N,
 *>          where NB is the blocksize for SSYTRD returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/ssyevd.f b/SRC/ssyevd.f
index bde96fc2..2c188a23 100644
--- a/SRC/ssyevd.f
+++ b/SRC/ssyevd.f
@@ -117,8 +117,7 @@
 *>          If JOBZ = 'N' and N > 1, LWORK must be at least 2*N+1.
 *>          If JOBZ = 'V' and N > 1, LWORK must be at least 
 *>                                                1 + 6*N + 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK and IWORK
 *>          arrays, returns these values as the first entries of the WORK
@@ -139,8 +138,7 @@
 *>          If N <= 1,                LIWORK must be at least 1.
 *>          If JOBZ  = 'N' and N > 1, LIWORK must be at least 1.
 *>          If JOBZ  = 'V' and N > 1, LIWORK must be at least 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK and
 *>          IWORK arrays, returns these values as the first entries of
diff --git a/SRC/ssyevr.f b/SRC/ssyevr.f
index 8bd78f17..1ee75462 100644
--- a/SRC/ssyevr.f
+++ b/SRC/ssyevr.f
@@ -185,22 +185,18 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing A to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          See "Computing Small Singular Values of Bidiagonal Matrices
 *>          with Guaranteed High Relative Accuracy," by Demmel and
 *>          Kahan, LAPACK Working Note #3.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If high relative accuracy is important, set ABSTOL to
 *>          SLAMCH( 'Safe minimum' ).  Doing so will guarantee that
 *>          eigenvalues are computed to high relative accuracy when
@@ -271,8 +267,7 @@
 *>          For optimal efficiency, LWORK >= (NB+6)*N,
 *>          where NB is the max of the blocksize for SSYTRD and SORMTR
 *>          returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK and IWORK
 *>          arrays, returns these values as the first entries of the WORK
@@ -290,8 +285,7 @@
 *> \verbatim
 *>          LIWORK is INTEGER
 *>          The dimension of the array IWORK.  LIWORK >= max(1,10*N).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK and
 *>          IWORK arrays, returns these values as the first entries of
diff --git a/SRC/ssyevx.f b/SRC/ssyevx.f
index be1c70c3..389f6f37 100644
--- a/SRC/ssyevx.f
+++ b/SRC/ssyevx.f
@@ -130,24 +130,20 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing A to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*SLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
 *>          eigenvectors did not converge, try setting ABSTOL to
 *>          2*SLAMCH('S').
-*> \endverbatim
-*> \verbatim
+*>
 *>          See "Computing Small Singular Values of Bidiagonal Matrices
 *>          with Guaranteed High Relative Accuracy," by Demmel and
 *>          Kahan, LAPACK Working Note #3.
@@ -204,8 +200,7 @@
 *>          For optimal efficiency, LWORK >= (NB+3)*N,
 *>          where NB is the max of the blocksize for SSYTRD and SORMTR
 *>          returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/ssygs2.f b/SRC/ssygs2.f
index 99a8894e..53e4d929 100644
--- a/SRC/ssygs2.f
+++ b/SRC/ssygs2.f
@@ -82,8 +82,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the transformed matrix, stored in the
 *>          same format as A.
 *> \endverbatim
diff --git a/SRC/ssygst.f b/SRC/ssygst.f
index eb827ab0..29afe23f 100644
--- a/SRC/ssygst.f
+++ b/SRC/ssygst.f
@@ -82,8 +82,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the transformed matrix, stored in the
 *>          same format as A.
 *> \endverbatim
diff --git a/SRC/ssygv.f b/SRC/ssygv.f
index 4111749f..79dd4bba 100644
--- a/SRC/ssygv.f
+++ b/SRC/ssygv.f
@@ -83,8 +83,7 @@
 *>          upper triangular part of the matrix A.  If UPLO = 'L',
 *>          the leading N-by-N lower triangular part of A contains
 *>          the lower triangular part of the matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if JOBZ = 'V', then if INFO = 0, A contains the
 *>          matrix Z of eigenvectors.  The eigenvectors are normalized
 *>          as follows:
@@ -109,8 +108,7 @@
 *>          contains the upper triangular part of the matrix B.
 *>          If UPLO = 'L', the leading N-by-N lower triangular part of B
 *>          contains the lower triangular part of the matrix B.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO <= N, the part of B containing the matrix is
 *>          overwritten by the triangular factor U or L from the Cholesky
 *>          factorization B = U**T*U or B = L*L**T.
@@ -140,8 +138,7 @@
 *>          The length of the array WORK.  LWORK >= max(1,3*N-1).
 *>          For optimal efficiency, LWORK >= (NB+2)*N,
 *>          where NB is the blocksize for SSYTRD returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/ssygvd.f b/SRC/ssygvd.f
index b1798af4..28b90816 100644
--- a/SRC/ssygvd.f
+++ b/SRC/ssygvd.f
@@ -91,8 +91,7 @@
 *>          upper triangular part of the matrix A.  If UPLO = 'L',
 *>          the leading N-by-N lower triangular part of A contains
 *>          the lower triangular part of the matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if JOBZ = 'V', then if INFO = 0, A contains the
 *>          matrix Z of eigenvectors.  The eigenvectors are normalized
 *>          as follows:
@@ -117,8 +116,7 @@
 *>          upper triangular part of the matrix B.  If UPLO = 'L',
 *>          the leading N-by-N lower triangular part of B contains
 *>          the lower triangular part of the matrix B.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO <= N, the part of B containing the matrix is
 *>          overwritten by the triangular factor U or L from the Cholesky
 *>          factorization B = U**T*U or B = L*L**T.
@@ -149,8 +147,7 @@
 *>          If N <= 1,               LWORK >= 1.
 *>          If JOBZ = 'N' and N > 1, LWORK >= 2*N+1.
 *>          If JOBZ = 'V' and N > 1, LWORK >= 1 + 6*N + 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK and IWORK
 *>          arrays, returns these values as the first entries of the WORK
@@ -171,8 +168,7 @@
 *>          If N <= 1,                LIWORK >= 1.
 *>          If JOBZ  = 'N' and N > 1, LIWORK >= 1.
 *>          If JOBZ  = 'V' and N > 1, LIWORK >= 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK and
 *>          IWORK arrays, returns these values as the first entries of
diff --git a/SRC/ssygvx.f b/SRC/ssygvx.f
index 3fe5ad0f..ffe4bf53 100644
--- a/SRC/ssygvx.f
+++ b/SRC/ssygvx.f
@@ -97,8 +97,7 @@
 *>          upper triangular part of the matrix A.  If UPLO = 'L',
 *>          the leading N-by-N lower triangular part of A contains
 *>          the lower triangular part of the matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the lower triangle (if UPLO='L') or the upper
 *>          triangle (if UPLO='U') of A, including the diagonal, is
 *>          destroyed.
@@ -118,8 +117,7 @@
 *>          upper triangular part of the matrix B.  If UPLO = 'L',
 *>          the leading N-by-N lower triangular part of B contains
 *>          the lower triangular part of the matrix B.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO <= N, the part of B containing the matrix is
 *>          overwritten by the triangular factor U or L from the Cholesky
 *>          factorization B = U**T*U or B = L*L**T.
@@ -165,19 +163,16 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing C to tridiagonal form, where C is the symmetric
 *>          matrix of the standard symmetric problem to which the
 *>          generalized problem is transformed.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*DLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
@@ -210,8 +205,7 @@
 *>          The eigenvectors are normalized as follows:
 *>          if ITYPE = 1 or 2, Z**T*B*Z = I;
 *>          if ITYPE = 3, Z**T*inv(B)*Z = I.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If an eigenvector fails to converge, then that column of Z
 *>          contains the latest approximation to the eigenvector, and the
 *>          index of the eigenvector is returned in IFAIL.
@@ -239,8 +233,7 @@
 *>          The length of the array WORK.  LWORK >= max(1,8*N).
 *>          For optimal efficiency, LWORK >= (NB+3)*N,
 *>          where NB is the blocksize for SSYTRD returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/ssyrfs.f b/SRC/ssyrfs.f
index d154f68f..f5a06401 100644
--- a/SRC/ssyrfs.f
+++ b/SRC/ssyrfs.f
@@ -167,12 +167,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/ssyrfsx.f b/SRC/ssyrfsx.f
index b0febebc..67830e4e 100644
--- a/SRC/ssyrfsx.f
+++ b/SRC/ssyrfsx.f
@@ -219,37 +219,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -258,8 +252,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -270,14 +263,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -285,26 +276,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -315,8 +302,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -335,8 +321,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0
@@ -347,8 +332,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -358,8 +342,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/ssysv.f b/SRC/ssysv.f
index 4fce6de3..baf78b94 100644
--- a/SRC/ssysv.f
+++ b/SRC/ssysv.f
@@ -85,8 +85,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the block diagonal matrix D and the
 *>          multipliers used to obtain the factor U or L from the
 *>          factorization A = U*D*U**T or A = L*D*L**T as computed by
@@ -140,8 +139,7 @@
 *>          SSYTRF.
 *>          for LWORK < N, TRS will be done with Level BLAS 2
 *>          for LWORK >= N, TRS will be done with Level BLAS 3
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/ssysvx.f b/SRC/ssysvx.f
index f1315002..9494b18d 100644
--- a/SRC/ssysvx.f
+++ b/SRC/ssysvx.f
@@ -135,8 +135,7 @@
 *>          contains the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
 *>          A = U*D*U**T or A = L*D*L**T as computed by SSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AF is an output argument and on exit
 *>          returns the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
@@ -162,8 +161,7 @@
 *>          is a 2-by-2 diagonal block.  If UPLO = 'L' and IPIV(k) =
 *>          IPIV(k+1) < 0, then rows and columns k+1 and -IPIV(k) were
 *>          interchanged and D(k:k+1,k:k+1) is a 2-by-2 diagonal block.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains details of the interchanges and the block structure
 *>          of D, as determined by SSYTRF.
@@ -236,8 +234,7 @@
 *>          The length of WORK.  LWORK >= max(1,3*N), and for best
 *>          performance, when FACT = 'N', LWORK >= max(1,3*N,N*NB), where
 *>          NB is the optimal blocksize for SSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/ssysvxx.f b/SRC/ssysvxx.f
index 3824bebb..aa45f3fa 100644
--- a/SRC/ssysvxx.f
+++ b/SRC/ssysvxx.f
@@ -167,8 +167,7 @@
 *>     N-by-N lower triangular part of A contains the lower
 *>     triangular part of the matrix A, and the strictly upper
 *>     triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>     On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>     diag(S)*A*diag(S).
 *> \endverbatim
@@ -186,8 +185,7 @@
 *>     contains the block diagonal matrix D and the multipliers
 *>     used to obtain the factor U or L from the factorization A =
 *>     U*D*U**T or A = L*D*L**T as computed by SSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then AF is an output argument and on exit
 *>     returns the block diagonal matrix D and the multipliers
 *>     used to obtain the factor U or L from the factorization A =
@@ -213,8 +211,7 @@
 *>     diagonal block.  If UPLO = 'L' and IPIV(k) = IPIV(k+1) < 0,
 *>     then rows and columns k+1 and -IPIV(k) were interchanged
 *>     and D(k:k+1,k:k+1) is a 2-by-2 diagonal block.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then IPIV is an output argument and on exit
 *>     contains details of the interchanges and the block
 *>     structure of D, as determined by SSYTRF.
@@ -325,37 +322,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -364,8 +355,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -376,14 +366,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -391,26 +379,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -421,8 +405,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -441,8 +424,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0
@@ -453,8 +435,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -464,8 +445,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/ssyswapr.f b/SRC/ssyswapr.f
index 2ebdfd2a..2f4f85c1 100644
--- a/SRC/ssyswapr.f
+++ b/SRC/ssyswapr.f
@@ -61,8 +61,7 @@
 *>          A is REAL array, dimension (LDA,N)
 *>          On entry, the NB diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by SSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/ssytf2.f b/SRC/ssytf2.f
index b3c77310..2ac172a3 100644
--- a/SRC/ssytf2.f
+++ b/SRC/ssytf2.f
@@ -76,8 +76,7 @@
 *>          leading n-by-n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L (see below for further details).
 *> \endverbatim
diff --git a/SRC/ssytrf.f b/SRC/ssytrf.f
index 23a4869d..7d33f857 100644
--- a/SRC/ssytrf.f
+++ b/SRC/ssytrf.f
@@ -75,8 +75,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L (see below for further details).
 *> \endverbatim
@@ -111,8 +110,7 @@
 *>          LWORK is INTEGER
 *>          The length of WORK.  LWORK >=1.  For best performance
 *>          LWORK >= N*NB, where NB is the block size returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/ssytri.f b/SRC/ssytri.f
index a0bc5000..ed93bdd0 100644
--- a/SRC/ssytri.f
+++ b/SRC/ssytri.f
@@ -64,8 +64,7 @@
 *>          A is REAL array, dimension (LDA,N)
 *>          On entry, the block diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by SSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/ssytri2.f b/SRC/ssytri2.f
index 30cc2bb7..021fd63a 100644
--- a/SRC/ssytri2.f
+++ b/SRC/ssytri2.f
@@ -65,8 +65,7 @@
 *>          A is REAL array, dimension (LDA,N)
 *>          On entry, the NB diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by SSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/ssytri2x.f b/SRC/ssytri2x.f
index 09e0da46..860366bb 100644
--- a/SRC/ssytri2x.f
+++ b/SRC/ssytri2x.f
@@ -64,8 +64,7 @@
 *>          A is REAL array, dimension (LDA,N)
 *>          On entry, the NNB diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by SSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/stfsm.f b/SRC/stfsm.f
index 51ad0e7b..e2639546 100644
--- a/SRC/stfsm.f
+++ b/SRC/stfsm.f
@@ -68,14 +68,11 @@
 *>          SIDE is CHARACTER*1
 *>           On entry, SIDE specifies whether op( A ) appears on the left
 *>           or right of X as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              SIDE = 'L' or 'l'   op( A )*X = alpha*B.
-*> \endverbatim
-*> \verbatim
+*>
 *>              SIDE = 'R' or 'r'   X*op( A ) = alpha*B.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -86,8 +83,7 @@
 *>           an upper or lower triangular matrix as follows:
 *>           UPLO = 'U' or 'u' RFP A came from an upper triangular matrix
 *>           UPLO = 'L' or 'l' RFP A came from a  lower triangular matrix
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -96,14 +92,11 @@
 *>          TRANS is CHARACTER*1
 *>           On entry, TRANS  specifies the form of op( A ) to be used
 *>           in the matrix multiplication as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              TRANS  = 'N' or 'n'   op( A ) = A.
-*> \endverbatim
-*> \verbatim
+*>
 *>              TRANS  = 'T' or 't'   op( A ) = A'.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -112,15 +105,12 @@
 *>          DIAG is CHARACTER*1
 *>           On entry, DIAG specifies whether or not RFP A is unit
 *>           triangular as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              DIAG = 'U' or 'u'   A is assumed to be unit triangular.
-*> \endverbatim
-*> \verbatim
+*>
 *>              DIAG = 'N' or 'n'   A is not assumed to be unit
 *>                                  triangular.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/stftri.f b/SRC/stftri.f
index e2db8abc..15dfee70 100644
--- a/SRC/stftri.f
+++ b/SRC/stftri.f
@@ -86,8 +86,7 @@
 *>          elements of lower packed A. The LDA of RFP A is (N+1)/2 when
 *>          TRANSR = 'T'. When TRANSR is 'N' the LDA is N+1 when N is
 *>          even and N is odd. See the Note below for more details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the (triangular) inverse of the original matrix, in
 *>          the same storage format.
 *> \endverbatim
diff --git a/SRC/stgevc.f b/SRC/stgevc.f
index 2bcabf22..d382322a 100644
--- a/SRC/stgevc.f
+++ b/SRC/stgevc.f
@@ -146,13 +146,11 @@
 *>          if HOWMNY = 'S', the left eigenvectors of (S,P) specified by
 *>                      SELECT, stored consecutively in the columns of
 *>                      VL, in the same order as their eigenvalues.
-*> \endverbatim
-*> \verbatim
+*>
 *>          A complex eigenvector corresponding to a complex eigenvalue
 *>          is stored in two consecutive columns, the first holding the
 *>          real part, and the second the imaginary part.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Not referenced if SIDE = 'R'.
 *> \endverbatim
 *>
@@ -169,8 +167,7 @@
 *>          On entry, if SIDE = 'R' or 'B' and HOWMNY = 'B', VR must
 *>          contain an N-by-N matrix Z (usually the orthogonal matrix Z
 *>          of right Schur vectors returned by SHGEQZ).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if SIDE = 'R' or 'B', VR contains:
 *>          if HOWMNY = 'A', the matrix X of right eigenvectors of (S,P);
 *>          if HOWMNY = 'B' or 'b', the matrix Z*X;
@@ -178,8 +175,7 @@
 *>                      specified by SELECT, stored consecutively in the
 *>                      columns of VR, in the same order as their
 *>                      eigenvalues.
-*> \endverbatim
-*> \verbatim
+*>
 *>          A complex eigenvector corresponding to a complex eigenvalue
 *>          is stored in two consecutive columns, the first holding the
 *>          real part and the second the imaginary part.
diff --git a/SRC/stgexc.f b/SRC/stgexc.f
index 5d4ffb6d..154aa2cb 100644
--- a/SRC/stgexc.f
+++ b/SRC/stgexc.f
@@ -169,8 +169,7 @@
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.
 *>          LWORK >= 1 when N <= 1, otherwise LWORK >= 4*N + 16.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/stgsen.f b/SRC/stgsen.f
index d3189ee7..43e413b2 100644
--- a/SRC/stgsen.f
+++ b/SRC/stgsen.f
@@ -164,8 +164,7 @@
 *> \param[out] BETA
 *> \verbatim
 *>          BETA is REAL array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, (ALPHAR(j) + ALPHAI(j)*i)/BETA(j), j=1,...,N, will
 *>          be the generalized eigenvalues.  ALPHAR(j) + ALPHAI(j)*i
 *>          and BETA(j),j=1,...,N  are the diagonals of the complex Schur
@@ -228,8 +227,7 @@
 *> \param[out] PR
 *> \verbatim
 *>          PR is REAL
-*> \endverbatim
-*> \verbatim
+*>
 *>          If IJOB = 1, 4 or 5, PL, PR are lower bounds on the
 *>          reciprocal of the norm of "projections" onto left and right
 *>          eigenspaces with respect to the selected cluster.
@@ -261,8 +259,7 @@
 *>          The dimension of the array WORK. LWORK >=  4*N+16.
 *>          If IJOB = 1, 2 or 4, LWORK >= MAX(4*N+16, 2*M*(N-M)).
 *>          If IJOB = 3 or 5, LWORK >= MAX(4*N+16, 4*M*(N-M)).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
@@ -281,8 +278,7 @@
 *>          The dimension of the array IWORK. LIWORK >= 1.
 *>          If IJOB = 1, 2 or 4, LIWORK >=  N+6.
 *>          If IJOB = 3 or 5, LIWORK >= MAX(2*M*(N-M), N+6).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal size of the IWORK array,
 *>          returns this value as the first entry of the IWORK array, and
diff --git a/SRC/stgsja.f b/SRC/stgsja.f
index 3dbd1b87..a53e4a82 100644
--- a/SRC/stgsja.f
+++ b/SRC/stgsja.f
@@ -185,8 +185,7 @@
 *> \param[in] L
 *> \verbatim
 *>          L is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          K and L specify the subblocks in the input matrices A and B:
 *>          A23 = A(K+1:MIN(K+L,M),N-L+1:N) and B13 = B(1:L,N-L+1:N)
 *>          of A and B, whose GSVD is going to be computed by STGSJA.
@@ -229,8 +228,7 @@
 *> \param[in] TOLB
 *> \verbatim
 *>          TOLB is REAL
-*> \endverbatim
-*> \verbatim
+*>
 *>          TOLA and TOLB are the convergence criteria for the Jacobi-
 *>          Kogbetliantz iteration procedure. Generally, they are the
 *>          same as used in the preprocessing step, say
@@ -246,8 +244,7 @@
 *> \param[out] BETA
 *> \verbatim
 *>          BETA is REAL array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, ALPHA and BETA contain the generalized singular
 *>          value pairs of A and B;
 *>            ALPHA(1:K) = 1,
diff --git a/SRC/stgsna.f b/SRC/stgsna.f
index 6d80e97d..03f6a9b0 100644
--- a/SRC/stgsna.f
+++ b/SRC/stgsna.f
@@ -203,8 +203,7 @@
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK. LWORK >= max(1,N).
 *>          If JOB = 'V' or 'B' LWORK >= 2*N*(N+2)+16.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/stgsyl.f b/SRC/stgsyl.f
index 52353777..ac8edab4 100644
--- a/SRC/stgsyl.f
+++ b/SRC/stgsyl.f
@@ -234,8 +234,7 @@
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK. LWORK > = 1.
 *>          If IJOB = 1 or 2 and TRANS = 'N', LWORK >= max(1,2*M*N).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/strexc.f b/SRC/strexc.f
index 507e3924..8609b9c4 100644
--- a/SRC/strexc.f
+++ b/SRC/strexc.f
@@ -104,8 +104,7 @@
 *> \param[in,out] ILST
 *> \verbatim
 *>          ILST is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          Specify the reordering of the diagonal blocks of T.
 *>          The block with row index IFST is moved to row ILST, by a
 *>          sequence of transpositions between adjacent blocks.
diff --git a/SRC/strsen.f b/SRC/strsen.f
index fad507a2..877de3ce 100644
--- a/SRC/strsen.f
+++ b/SRC/strsen.f
@@ -137,8 +137,7 @@
 *> \param[out] WI
 *> \verbatim
 *>          WI is REAL array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          The real and imaginary parts, respectively, of the reordered
 *>          eigenvalues of T. The eigenvalues are stored in the same
 *>          order as on the diagonal of T, with WR(i) = T(i,i) and, if
@@ -187,8 +186,7 @@
 *>          If JOB = 'N', LWORK >= max(1,N);
 *>          if JOB = 'E', LWORK >= max(1,M*(N-M));
 *>          if JOB = 'V' or 'B', LWORK >= max(1,2*M*(N-M)).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
@@ -207,8 +205,7 @@
 *>          The dimension of the array IWORK.
 *>          If JOB = 'N' or 'E', LIWORK >= 1;
 *>          if JOB = 'V' or 'B', LIWORK >= max(1,M*(N-M)).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal size of the IWORK array,
 *>          returns this value as the first entry of the IWORK array, and
diff --git a/SRC/strti2.f b/SRC/strti2.f
index d8f7cde5..911a6ed5 100644
--- a/SRC/strti2.f
+++ b/SRC/strti2.f
@@ -78,8 +78,7 @@
 *>          triangular part of A is not referenced.  If DIAG = 'U', the
 *>          diagonal elements of A are also not referenced and are
 *>          assumed to be 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the (triangular) inverse of the original matrix, in
 *>          the same storage format.
 *> \endverbatim
diff --git a/SRC/stzrzf.f b/SRC/stzrzf.f
index 886eeaf7..2b9999bf 100644
--- a/SRC/stzrzf.f
+++ b/SRC/stzrzf.f
@@ -95,8 +95,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,M).
 *>          For optimum performance LWORK >= M*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zbbcsd.f b/SRC/zbbcsd.f
index ab54ebb6..7911a4a9 100644
--- a/SRC/zbbcsd.f
+++ b/SRC/zbbcsd.f
@@ -282,8 +282,7 @@
 *> \verbatim
 *>          LRWORK is INTEGER
 *>          The dimension of the array RWORK. LRWORK >= MAX(1,8*Q).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LRWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal size of the RWORK array,
 *>          returns this value as the first entry of the work array, and
@@ -298,20 +297,16 @@
 *>          > 0:  if ZBBCSD did not converge, INFO specifies the number
 *>                of nonzero entries in PHI, and B11D, B11E, etc.,
 *>                contain the partially reduced matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Reference
 *>  =========
-*> \endverbatim
-*> \verbatim
+*>
 *>  [1] Brian D. Sutton. Computing the complete CS decomposition. Numer.
 *>      Algorithms, 50(1):33-65, 2009.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  TOLMUL  DOUBLE PRECISION, default = MAX(10,MIN(100,EPS**(-1/8)))
 *>          TOLMUL controls the convergence criterion of the QR loop.
 *>          Angles THETA(i), PHI(i) are rounded to 0 or PI/2 when they
diff --git a/SRC/zbdsqr.f b/SRC/zbdsqr.f
index fafad639..0e2bda02 100644
--- a/SRC/zbdsqr.f
+++ b/SRC/zbdsqr.f
@@ -180,12 +180,10 @@
 *>                elements of a bidiagonal matrix which is orthogonally
 *>                similar to the input matrix B;  if INFO = i, i
 *>                elements of E have not converged to zero.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  TOLMUL  DOUBLE PRECISION, default = max(10,min(100,EPS**(-1/8)))
 *>          TOLMUL controls the convergence criterion of the QR loop.
 *>          If it is positive, TOLMUL*EPS is the desired relative
@@ -200,8 +198,7 @@
 *>          Default is to lose at either one eighth or 2 of the
 *>             available decimal digits in each computed singular value
 *>             (whichever is smaller).
-*> \endverbatim
-*> \verbatim
+*>
 *>  MAXITR  INTEGER, default = 6
 *>          MAXITR controls the maximum number of passes of the
 *>          algorithm through its inner loop. The algorithms stops
diff --git a/SRC/zcposv.f b/SRC/zcposv.f
index 54d7b5b2..ee010e43 100644
--- a/SRC/zcposv.f
+++ b/SRC/zcposv.f
@@ -107,12 +107,10 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note that the imaginary parts of the diagonal
 *>          elements need not be set and are assumed to be zero.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if iterative refinement has been successfully used
 *>          (INFO.EQ.0 and ITER.GE.0, see description below), then A is
 *>          unchanged, if double precision factorization has been used
diff --git a/SRC/zgbrfs.f b/SRC/zgbrfs.f
index 69644b53..25a04c9a 100644
--- a/SRC/zgbrfs.f
+++ b/SRC/zgbrfs.f
@@ -181,12 +181,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/zgbrfsx.f b/SRC/zgbrfsx.f
index 93e7797d..ef608537 100644
--- a/SRC/zgbrfsx.f
+++ b/SRC/zgbrfsx.f
@@ -256,37 +256,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -295,8 +289,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -307,14 +300,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -322,26 +313,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -352,8 +339,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -372,8 +358,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0D+0
@@ -384,8 +369,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -395,8 +379,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/zgbsvx.f b/SRC/zgbsvx.f
index 9e026149..cbe43b20 100644
--- a/SRC/zgbsvx.f
+++ b/SRC/zgbsvx.f
@@ -151,14 +151,12 @@
 *>          The j-th column of A is stored in the j-th column of the
 *>          array AB as follows:
 *>          AB(KU+1+i-j,j) = A(i,j) for max(1,j-KU)<=i<=min(N,j+kl)
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'F' and EQUED is not 'N', then A must have been
 *>          equilibrated by the scaling factors in R and/or C.  AB is not
 *>          modified if FACT = 'F' or 'N', or if FACT = 'E' and
 *>          EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if EQUED .ne. 'N', A is scaled as follows:
 *>          EQUED = 'R':  A := diag(R) * A
 *>          EQUED = 'C':  A := A * diag(C)
@@ -181,12 +179,10 @@
 *>          and the multipliers used during the factorization are stored
 *>          in rows KL+KU+2 to 2*KL+KU+1.  If EQUED .ne. 'N', then AFB is
 *>          the factored form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AFB is an output argument and on exit
 *>          returns details of the LU factorization of A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then AFB is an output argument and on exit
 *>          returns details of the LU factorization of the equilibrated
 *>          matrix A (see the description of AB for the form of the
@@ -206,13 +202,11 @@
 *>          contains the pivot indices from the factorization A = L*U
 *>          as computed by ZGBTRF; row i of the matrix was interchanged
 *>          with row IPIV(i).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains the pivot indices from the factorization A = L*U
 *>          of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then IPIV is an output argument and on exit
 *>          contains the pivot indices from the factorization A = L*U
 *>          of the equilibrated matrix A.
diff --git a/SRC/zgbsvxx.f b/SRC/zgbsvxx.f
index 390ec7fa..955df209 100644
--- a/SRC/zgbsvxx.f
+++ b/SRC/zgbsvxx.f
@@ -178,14 +178,12 @@
 *>     The j-th column of A is stored in the j-th column of the
 *>     array AB as follows:
 *>     AB(KU+1+i-j,j) = A(i,j) for max(1,j-KU)<=i<=min(N,j+kl)
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'F' and EQUED is not 'N', then AB must have been
 *>     equilibrated by the scaling factors in R and/or C.  AB is not
 *>     modified if FACT = 'F' or 'N', or if FACT = 'E' and
 *>     EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>     On exit, if EQUED .ne. 'N', A is scaled as follows:
 *>     EQUED = 'R':  A := diag(R) * A
 *>     EQUED = 'C':  A := A * diag(C)
@@ -208,13 +206,11 @@
 *>     and the multipliers used during the factorization are stored
 *>     in rows KL+KU+2 to 2*KL+KU+1.  If EQUED .ne. 'N', then AFB is
 *>     the factored form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then AF is an output argument and on exit
 *>     returns the factors L and U from the factorization A = P*L*U
 *>     of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'E', then AF is an output argument and on exit
 *>     returns the factors L and U from the factorization A = P*L*U
 *>     of the equilibrated matrix A (see the description of A for
@@ -234,13 +230,11 @@
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     as computed by DGETRF; row i of the matrix was interchanged
 *>     with row IPIV(i).
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then IPIV is an output argument and on exit
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'E', then IPIV is an output argument and on exit
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     of the equilibrated matrix A.
@@ -380,37 +374,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -419,8 +407,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -431,14 +418,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -446,26 +431,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -476,8 +457,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -496,8 +476,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0D+0
@@ -505,8 +484,7 @@
 *>                    computed.
 *>            = 1.0 : Use the extra-precise refinement algorithm.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -516,8 +494,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/zgbtf2.f b/SRC/zgbtf2.f
index ed68af22..c5f12329 100644
--- a/SRC/zgbtf2.f
+++ b/SRC/zgbtf2.f
@@ -76,8 +76,7 @@
 *>          The j-th column of A is stored in the j-th column of the
 *>          array AB as follows:
 *>          AB(kl+ku+1+i-j,j) = A(i,j) for max(1,j-ku)<=i<=min(m,j+kl)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, details of the factorization: U is stored as an
 *>          upper triangular band matrix with KL+KU superdiagonals in
 *>          rows 1 to KL+KU+1, and the multipliers used during the
diff --git a/SRC/zgbtrf.f b/SRC/zgbtrf.f
index f9c31f36..ee891551 100644
--- a/SRC/zgbtrf.f
+++ b/SRC/zgbtrf.f
@@ -76,8 +76,7 @@
 *>          The j-th column of A is stored in the j-th column of the
 *>          array AB as follows:
 *>          AB(kl+ku+1+i-j,j) = A(i,j) for max(1,j-ku)<=i<=min(m,j+kl)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, details of the factorization: U is stored as an
 *>          upper triangular band matrix with KL+KU superdiagonals in
 *>          rows 1 to KL+KU+1, and the multipliers used during the
diff --git a/SRC/zgebrd.f b/SRC/zgebrd.f
index f6ea3069..6a2881e0 100644
--- a/SRC/zgebrd.f
+++ b/SRC/zgebrd.f
@@ -126,8 +126,7 @@
 *>          The length of the array WORK.  LWORK >= max(1,M,N).
 *>          For optimum performance LWORK >= (M+N)*NB, where NB
 *>          is the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zgees.f b/SRC/zgees.f
index 02f6497f..27083007 100644
--- a/SRC/zgees.f
+++ b/SRC/zgees.f
@@ -143,8 +143,7 @@
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.  LWORK >= max(1,2*N).
 *>          For good performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zgeesx.f b/SRC/zgeesx.f
index 2b92751f..2181038b 100644
--- a/SRC/zgeesx.f
+++ b/SRC/zgeesx.f
@@ -184,8 +184,7 @@
 *>          that an error is only returned if LWORK < max(1,2*N), but if
 *>          SENSE = 'E' or 'V' or 'B' this may not be large enough.
 *>          For good performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates upper bound on the optimal size of the
 *>          array WORK, returns this value as the first entry of the WORK
diff --git a/SRC/zgeev.f b/SRC/zgeev.f
index c0466f6b..73f1b27a 100644
--- a/SRC/zgeev.f
+++ b/SRC/zgeev.f
@@ -139,8 +139,7 @@
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.  LWORK >= max(1,2*N).
 *>          For good performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zgeevx.f b/SRC/zgeevx.f
index abf6789c..3e14a9ba 100644
--- a/SRC/zgeevx.f
+++ b/SRC/zgeevx.f
@@ -89,8 +89,7 @@
 *>                 to make the rows and columns of A more equal in
 *>                 norm. Do not permute;
 *>          = 'B': Both diagonally scale and permute A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Computed reciprocal condition numbers will be for the matrix
 *>          after balancing and/or permuting. Permuting does not change
 *>          condition numbers (in exact arithmetic), but balancing does.
@@ -120,8 +119,7 @@
 *>          = 'E': Computed for eigenvalues only;
 *>          = 'V': Computed for right eigenvectors only;
 *>          = 'B': Computed for eigenvalues and right eigenvectors.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If SENSE = 'E' or 'B', both left and right eigenvectors
 *>          must also be computed (JOBVL = 'V' and JOBVR = 'V').
 *> \endverbatim
@@ -248,8 +246,7 @@
 *>          LWORK >= max(1,2*N), and if SENSE = 'V' or 'B',
 *>          LWORK >= N*N+2*N.
 *>          For good performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zgegs.f b/SRC/zgegs.f
index f164962d..3859be54 100644
--- a/SRC/zgegs.f
+++ b/SRC/zgegs.f
@@ -124,8 +124,7 @@
 *>          The non-negative real scalars beta that define the
 *>          eigenvalues of GNEP.  BETA(j) = T(j,j), the diagonal element
 *>          of the triangular factor T.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Together, the quantities alpha = ALPHA(j) and beta = BETA(j)
 *>          represent the j-th eigenvalue of the matrix pair (A,B), in
 *>          one of the forms lambda = alpha/beta or mu = beta/alpha.
@@ -176,8 +175,7 @@
 *>          blocksizes (for ZGEQRF, ZUNMQR, and CUNGQR.)  Then compute:
 *>          NB  -- MAX of the blocksizes for ZGEQRF, ZUNMQR, and CUNGQR;
 *>          the optimal LWORK is N*(NB+1).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zgegv.f b/SRC/zgegv.f
index 9c8dc6c0..f12cbdcb 100644
--- a/SRC/zgegv.f
+++ b/SRC/zgegv.f
@@ -200,8 +200,7 @@
 *>          blocksizes (for ZGEQRF, ZUNMQR, and ZUNGQR.)  Then compute:
 *>          NB  -- MAX of the blocksizes for ZGEQRF, ZUNMQR, and ZUNGQR;
 *>          The optimal LWORK is  MAX( 2*N, N*(NB+1) ).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zgehd2.f b/SRC/zgehd2.f
index 67f24e6f..6a8ae738 100644
--- a/SRC/zgehd2.f
+++ b/SRC/zgehd2.f
@@ -55,8 +55,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          It is assumed that A is already upper triangular in rows
 *>          and columns 1:ILO-1 and IHI+1:N. ILO and IHI are normally
 *>          set by a previous call to ZGEBAL; otherwise they should be
diff --git a/SRC/zgehrd.f b/SRC/zgehrd.f
index ff81af1e..c546808e 100644
--- a/SRC/zgehrd.f
+++ b/SRC/zgehrd.f
@@ -55,8 +55,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          It is assumed that A is already upper triangular in rows
 *>          and columns 1:ILO-1 and IHI+1:N. ILO and IHI are normally
 *>          set by a previous call to ZGEBAL; otherwise they should be
@@ -101,8 +100,7 @@
 *>          The length of the array WORK.  LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is the
 *>          optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zgels.f b/SRC/zgels.f
index d0ea510e..05cdfe6f 100644
--- a/SRC/zgels.f
+++ b/SRC/zgels.f
@@ -149,8 +149,7 @@
 *>          For optimal performance,
 *>          LWORK >= max( 1, MN + max( MN, NRHS )*NB ).
 *>          where MN = min(M,N) and NB is the optimum block size.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zgelsd.f b/SRC/zgelsd.f
index 2a55ade4..ef7064a9 100644
--- a/SRC/zgelsd.f
+++ b/SRC/zgelsd.f
@@ -159,8 +159,7 @@
 *>              2*M + M*NRHS
 *>          if M is less than N, the code will execute correctly.
 *>          For good performance, LWORK should generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the array WORK and the
 *>          minimum sizes of the arrays RWORK and IWORK, and returns
diff --git a/SRC/zgelss.f b/SRC/zgelss.f
index 097e3767..6e628463 100644
--- a/SRC/zgelss.f
+++ b/SRC/zgelss.f
@@ -141,8 +141,7 @@
 *>          The dimension of the array WORK. LWORK >= 1, and also:
 *>          LWORK >=  2*min(M,N) + max(M,N,NRHS)
 *>          For good performance, LWORK should generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zgelsy.f b/SRC/zgelsy.f
index 3aaf57ba..c0126773 100644
--- a/SRC/zgelsy.f
+++ b/SRC/zgelsy.f
@@ -169,8 +169,7 @@
 *>          where NB is an upper bound on the blocksize returned
 *>          by ILAENV for the routines ZGEQP3, ZTZRZF, CTZRQF, ZUNMQR,
 *>          and ZUNMRZ.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zgeqlf.f b/SRC/zgeqlf.f
index 40ab0b2a..beaaf5e1 100644
--- a/SRC/zgeqlf.f
+++ b/SRC/zgeqlf.f
@@ -92,8 +92,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zgeqp3.f b/SRC/zgeqp3.f
index a39c87a0..e98296a6 100644
--- a/SRC/zgeqp3.f
+++ b/SRC/zgeqp3.f
@@ -101,8 +101,7 @@
 *>          The dimension of the array WORK. LWORK >= N+1.
 *>          For optimal performance LWORK >= ( N+1 )*NB, where NB
 *>          is the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zgeqrf.f b/SRC/zgeqrf.f
index 81047ea2..e11b19d8 100644
--- a/SRC/zgeqrf.f
+++ b/SRC/zgeqrf.f
@@ -90,8 +90,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zgeqrfp.f b/SRC/zgeqrfp.f
index 936eced3..bd084038 100644
--- a/SRC/zgeqrfp.f
+++ b/SRC/zgeqrfp.f
@@ -90,8 +90,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zgerfs.f b/SRC/zgerfs.f
index 5892610c..a720ac95 100644
--- a/SRC/zgerfs.f
+++ b/SRC/zgerfs.f
@@ -162,12 +162,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/zgerfsx.f b/SRC/zgerfsx.f
index 405b66d2..4e5aaa19 100644
--- a/SRC/zgerfsx.f
+++ b/SRC/zgerfsx.f
@@ -231,37 +231,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -270,8 +264,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -282,14 +275,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -297,26 +288,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -327,8 +314,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -347,8 +333,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0D+0
@@ -359,8 +344,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -370,8 +354,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/zgesdd.f b/SRC/zgesdd.f
index 74b40920..42c53ca6 100644
--- a/SRC/zgesdd.f
+++ b/SRC/zgesdd.f
@@ -174,8 +174,7 @@
 *>          if JOBZ = 'S' or 'A',
 *>                LWORK >= min(M,N)*min(M,N)+2*min(M,N)+max(M,N).
 *>          For good performance, LWORK should generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, a workspace query is assumed.  The optimal
 *>          size for the WORK array is calculated and stored in WORK(1),
 *>          and no other work except argument checking is performed.
diff --git a/SRC/zgesvd.f b/SRC/zgesvd.f
index f2b02165..3609a381 100644
--- a/SRC/zgesvd.f
+++ b/SRC/zgesvd.f
@@ -82,8 +82,7 @@
 *>                  vectors) are overwritten on the array A;
 *>          = 'N':  no rows of V**H (no right singular vectors) are
 *>                  computed.
-*> \endverbatim
-*> \verbatim
+*>
 *>          JOBVT and JOBU cannot both be 'O'.
 *> \endverbatim
 *>
@@ -172,8 +171,7 @@
 *>          The dimension of the array WORK.
 *>          LWORK >=  MAX(1,2*MIN(M,N)+MAX(M,N)).
 *>          For good performance, LWORK should generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zgesvx.f b/SRC/zgesvx.f
index 833c7076..48b29df3 100644
--- a/SRC/zgesvx.f
+++ b/SRC/zgesvx.f
@@ -138,8 +138,7 @@
 *>          not 'N', then A must have been equilibrated by the scaling
 *>          factors in R and/or C.  A is not modified if FACT = 'F' or
 *>          'N', or if FACT = 'E' and EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if EQUED .ne. 'N', A is scaled as follows:
 *>          EQUED = 'R':  A := diag(R) * A
 *>          EQUED = 'C':  A := A * diag(C)
@@ -159,13 +158,11 @@
 *>          contains the factors L and U from the factorization
 *>          A = P*L*U as computed by ZGETRF.  If EQUED .ne. 'N', then
 *>          AF is the factored form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AF is an output argument and on exit
 *>          returns the factors L and U from the factorization A = P*L*U
 *>          of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then AF is an output argument and on exit
 *>          returns the factors L and U from the factorization A = P*L*U
 *>          of the equilibrated matrix A (see the description of A for
@@ -185,13 +182,11 @@
 *>          contains the pivot indices from the factorization A = P*L*U
 *>          as computed by ZGETRF; row i of the matrix was interchanged
 *>          with row IPIV(i).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains the pivot indices from the factorization A = P*L*U
 *>          of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then IPIV is an output argument and on exit
 *>          contains the pivot indices from the factorization A = P*L*U
 *>          of the equilibrated matrix A.
diff --git a/SRC/zgesvxx.f b/SRC/zgesvxx.f
index aa5be45d..e7ee1295 100644
--- a/SRC/zgesvxx.f
+++ b/SRC/zgesvxx.f
@@ -168,8 +168,7 @@
 *>     not 'N', then A must have been equilibrated by the scaling
 *>     factors in R and/or C.  A is not modified if FACT = 'F' or
 *>     'N', or if FACT = 'E' and EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>     On exit, if EQUED .ne. 'N', A is scaled as follows:
 *>     EQUED = 'R':  A := diag(R) * A
 *>     EQUED = 'C':  A := A * diag(C)
@@ -189,13 +188,11 @@
 *>     contains the factors L and U from the factorization
 *>     A = P*L*U as computed by ZGETRF.  If EQUED .ne. 'N', then
 *>     AF is the factored form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then AF is an output argument and on exit
 *>     returns the factors L and U from the factorization A = P*L*U
 *>     of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'E', then AF is an output argument and on exit
 *>     returns the factors L and U from the factorization A = P*L*U
 *>     of the equilibrated matrix A (see the description of A for
@@ -215,13 +212,11 @@
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     as computed by ZGETRF; row i of the matrix was interchanged
 *>     with row IPIV(i).
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then IPIV is an output argument and on exit
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     of the original matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'E', then IPIV is an output argument and on exit
 *>     contains the pivot indices from the factorization A = P*L*U
 *>     of the equilibrated matrix A.
@@ -361,37 +356,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -400,8 +389,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -412,14 +400,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -427,26 +413,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -457,8 +439,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -477,8 +458,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0D+0
@@ -486,8 +466,7 @@
 *>                    computed.
 *>            = 1.0 : Use the extra-precise refinement algorithm.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -497,8 +476,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/zgetri.f b/SRC/zgetri.f
index 8605ef13..8e929bfd 100644
--- a/SRC/zgetri.f
+++ b/SRC/zgetri.f
@@ -84,8 +84,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,N).
 *>          For optimal performance LWORK >= N*NB, where NB is
 *>          the optimal blocksize returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zgges.f b/SRC/zgges.f
index 403d5f6a..5c1456ef 100644
--- a/SRC/zgges.f
+++ b/SRC/zgges.f
@@ -108,8 +108,7 @@
 *>          to the top left of the Schur form.
 *>          An eigenvalue ALPHA(j)/BETA(j) is selected if
 *>          SELCTG(ALPHA(j),BETA(j)) is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note that a selected complex eigenvalue may no longer satisfy
 *>          SELCTG(ALPHA(j),BETA(j)) = .TRUE. after ordering, since
 *>          ordering may change the value of complex eigenvalues
@@ -171,8 +170,7 @@
 *>          generalized eigenvalues.  ALPHA(j), j=1,...,N  and  BETA(j),
 *>          j=1,...,N  are the diagonals of the complex Schur form (A,B)
 *>          output by ZGGES. The  BETA(j) will be non-negative real.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note: the quotients ALPHA(j)/BETA(j) may easily over- or
 *>          underflow, and BETA(j) may even be zero.  Thus, the user
 *>          should avoid naively computing the ratio alpha/beta.
@@ -220,8 +218,7 @@
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.  LWORK >= max(1,2*N).
 *>          For good performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zggesx.f b/SRC/zggesx.f
index b80b1e23..e7f8329e 100644
--- a/SRC/zggesx.f
+++ b/SRC/zggesx.f
@@ -182,8 +182,7 @@
 *>          generalized eigenvalues.  ALPHA(j) and BETA(j),j=1,...,N  are
 *>          the diagonals of the complex Schur form (S,T).  BETA(j) will
 *>          be non-negative real.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note: the quotients ALPHA(j)/BETA(j) may easily over- or
 *>          underflow, and BETA(j) may even be zero.  Thus, the user
 *>          should avoid naively computing the ratio alpha/beta.
@@ -254,8 +253,7 @@
 *>          Note also that an error is only returned if
 *>          LWORK < MAX(1,2*N), but if SENSE = 'E' or 'V' or 'B' this may
 *>          not be large enough.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the bound on the optimal size of the WORK
 *>          array and the minimum size of the IWORK array, returns these
@@ -282,8 +280,7 @@
 *>          The dimension of the array IWORK.
 *>          If SENSE = 'N' or N = 0, LIWORK >= 1, otherwise
 *>          LIWORK >= N+2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the bound on the optimal size of the
 *>          WORK array and the minimum size of the IWORK array, returns
diff --git a/SRC/zggev.f b/SRC/zggev.f
index eaf61ffe..155b19ce 100644
--- a/SRC/zggev.f
+++ b/SRC/zggev.f
@@ -121,8 +121,7 @@
 *>          BETA is COMPLEX*16 array, dimension (N)
 *>          On exit, ALPHA(j)/BETA(j), j=1,...,N, will be the
 *>          generalized eigenvalues.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note: the quotients ALPHA(j)/BETA(j) may easily over- or
 *>          underflow, and BETA(j) may even be zero.  Thus, the user
 *>          should avoid naively computing the ratio alpha/beta.
@@ -178,8 +177,7 @@
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.  LWORK >= max(1,2*N).
 *>          For good performance, LWORK must generally be larger.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zggevx.f b/SRC/zggevx.f
index 001067e1..a02451af 100644
--- a/SRC/zggevx.f
+++ b/SRC/zggevx.f
@@ -158,8 +158,7 @@
 *>          BETA is COMPLEX*16 array, dimension (N)
 *>          On exit, ALPHA(j)/BETA(j), j=1,...,N, will be the generalized
 *>          eigenvalues.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Note: the quotient ALPHA(j)/BETA(j) ) may easily over- or
 *>          underflow, and BETA(j) may even be zero.  Thus, the user
 *>          should avoid naively computing the ratio ALPHA/BETA.
@@ -289,8 +288,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,2*N).
 *>          If SENSE = 'E', LWORK >= max(1,4*N).
 *>          If SENSE = 'V' or 'B', LWORK >= max(1,2*N*N+2*N).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zggglm.f b/SRC/zggglm.f
index bbfd4f72..fd48a9db 100644
--- a/SRC/zggglm.f
+++ b/SRC/zggglm.f
@@ -130,8 +130,7 @@
 *> \param[out] Y
 *> \verbatim
 *>          Y is COMPLEX*16 array, dimension (P)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, X and Y are the solutions of the GLM problem.
 *> \endverbatim
 *>
@@ -148,8 +147,7 @@
 *>          For optimum performance, LWORK >= M+min(N,P)+max(N,P)*NB,
 *>          where NB is an upper bound for the optimal blocksizes for
 *>          ZGEQRF, ZGERQF, ZUNMQR and ZUNMRQ.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zgghrd.f b/SRC/zgghrd.f
index 309749f2..a9ae228f 100644
--- a/SRC/zgghrd.f
+++ b/SRC/zgghrd.f
@@ -101,8 +101,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          ILO and IHI mark the rows and columns of A which are to be
 *>          reduced.  It is assumed that A is already upper triangular
 *>          in rows and columns 1:ILO-1 and IHI+1:N.  ILO and IHI are
diff --git a/SRC/zgglse.f b/SRC/zgglse.f
index 068e1a74..56629ca6 100644
--- a/SRC/zgglse.f
+++ b/SRC/zgglse.f
@@ -141,8 +141,7 @@
 *>          For optimum performance LWORK >= P+min(M,N)+max(M,N)*NB,
 *>          where NB is an upper bound for the optimal blocksizes for
 *>          ZGEQRF, CGERQF, ZUNMQR and CUNMRQ.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zggsvd.f b/SRC/zggsvd.f
index 1210f102..0928f177 100644
--- a/SRC/zggsvd.f
+++ b/SRC/zggsvd.f
@@ -169,8 +169,7 @@
 *> \param[out] L
 *> \verbatim
 *>          L is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, K and L specify the dimension of the subblocks
 *>          described in Purpose.
 *>          K + L = effective numerical rank of (A**H,B**H)**H.
@@ -212,8 +211,7 @@
 *> \param[out] BETA
 *> \verbatim
 *>          BETA is DOUBLE PRECISION array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, ALPHA and BETA contain the generalized singular
 *>          value pairs of A and B;
 *>            ALPHA(1:K) = 1,
@@ -299,12 +297,10 @@
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value.
 *>          > 0:  if INFO = 1, the Jacobi-type procedure failed to
 *>                converge.  For further details, see subroutine ZTGSJA.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  TOLA    DOUBLE PRECISION
 *>  TOLB    DOUBLE PRECISION
 *>          TOLA and TOLB are the thresholds to determine the effective
diff --git a/SRC/zggsvp.f b/SRC/zggsvp.f
index 137ca5ed..005f361a 100644
--- a/SRC/zggsvp.f
+++ b/SRC/zggsvp.f
@@ -144,8 +144,7 @@
 *> \param[in] TOLB
 *> \verbatim
 *>          TOLB is DOUBLE PRECISION
-*> \endverbatim
-*> \verbatim
+*>
 *>          TOLA and TOLB are the thresholds to determine the effective
 *>          numerical rank of matrix B and a subblock of A. Generally,
 *>          they are set to
@@ -163,8 +162,7 @@
 *> \param[out] L
 *> \verbatim
 *>          L is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, K and L specify the dimension of the subblocks
 *>          described in Purpose section.
 *>          K + L = effective numerical rank of (A**H,B**H)**H.
diff --git a/SRC/zgtrfs.f b/SRC/zgtrfs.f
index 77e34d81..e6902eb9 100644
--- a/SRC/zgtrfs.f
+++ b/SRC/zgtrfs.f
@@ -185,12 +185,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/zgtsvx.f b/SRC/zgtsvx.f
index aeb65a88..9bf49032 100644
--- a/SRC/zgtsvx.f
+++ b/SRC/zgtsvx.f
@@ -136,8 +136,7 @@
 *>          If FACT = 'F', then DLF is an input argument and on entry
 *>          contains the (n-1) multipliers that define the matrix L from
 *>          the LU factorization of A as computed by ZGTTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then DLF is an output argument and on exit
 *>          contains the (n-1) multipliers that define the matrix L from
 *>          the LU factorization of A.
@@ -149,8 +148,7 @@
 *>          If FACT = 'F', then DF is an input argument and on entry
 *>          contains the n diagonal elements of the upper triangular
 *>          matrix U from the LU factorization of A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then DF is an output argument and on exit
 *>          contains the n diagonal elements of the upper triangular
 *>          matrix U from the LU factorization of A.
@@ -161,8 +159,7 @@
 *>          DUF is or output) COMPLEX*16 array, dimension (N-1)
 *>          If FACT = 'F', then DUF is an input argument and on entry
 *>          contains the (n-1) elements of the first superdiagonal of U.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then DUF is an output argument and on exit
 *>          contains the (n-1) elements of the first superdiagonal of U.
 *> \endverbatim
@@ -173,8 +170,7 @@
 *>          If FACT = 'F', then DU2 is an input argument and on entry
 *>          contains the (n-2) elements of the second superdiagonal of
 *>          U.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then DU2 is an output argument and on exit
 *>          contains the (n-2) elements of the second superdiagonal of
 *>          U.
@@ -186,8 +182,7 @@
 *>          If FACT = 'F', then IPIV is an input argument and on entry
 *>          contains the pivot indices from the LU factorization of A as
 *>          computed by ZGTTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains the pivot indices from the LU factorization of A;
 *>          row i of the matrix was interchanged with row IPIV(i).
diff --git a/SRC/zgttrf.f b/SRC/zgttrf.f
index 18fafe4f..cd46a004 100644
--- a/SRC/zgttrf.f
+++ b/SRC/zgttrf.f
@@ -59,8 +59,7 @@
 *>          DL is COMPLEX*16 array, dimension (N-1)
 *>          On entry, DL must contain the (n-1) sub-diagonal elements of
 *>          A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, DL is overwritten by the (n-1) multipliers that
 *>          define the matrix L from the LU factorization of A.
 *> \endverbatim
@@ -69,8 +68,7 @@
 *> \verbatim
 *>          D is COMPLEX*16 array, dimension (N)
 *>          On entry, D must contain the diagonal elements of A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, D is overwritten by the n diagonal elements of the
 *>          upper triangular matrix U from the LU factorization of A.
 *> \endverbatim
@@ -80,8 +78,7 @@
 *>          DU is COMPLEX*16 array, dimension (N-1)
 *>          On entry, DU must contain the (n-1) super-diagonal elements
 *>          of A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, DU is overwritten by the (n-1) elements of the first
 *>          super-diagonal of U.
 *> \endverbatim
diff --git a/SRC/zhbev.f b/SRC/zhbev.f
index 2341fd72..4d7f6f27 100644
--- a/SRC/zhbev.f
+++ b/SRC/zhbev.f
@@ -80,8 +80,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AB is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the first
 *>          superdiagonal and the diagonal of the tridiagonal matrix T
diff --git a/SRC/zhbevd.f b/SRC/zhbevd.f
index 6458086a..cf745abb 100644
--- a/SRC/zhbevd.f
+++ b/SRC/zhbevd.f
@@ -89,8 +89,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AB is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the first
 *>          superdiagonal and the diagonal of the tridiagonal matrix T
@@ -140,8 +139,7 @@
 *>          If N <= 1,               LWORK must be at least 1.
 *>          If JOBZ = 'N' and N > 1, LWORK must be at least N.
 *>          If JOBZ = 'V' and N > 1, LWORK must be at least 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK, RWORK and
 *>          IWORK arrays, returns these values as the first entries of
@@ -164,8 +162,7 @@
 *>          If JOBZ = 'N' and N > 1, LRWORK must be at least N.
 *>          If JOBZ = 'V' and N > 1, LRWORK must be at least
 *>                        1 + 5*N + 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LRWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
@@ -185,8 +182,7 @@
 *>          The dimension of array IWORK.
 *>          If JOBZ = 'N' or N <= 1, LIWORK must be at least 1.
 *>          If JOBZ = 'V' and N > 1, LIWORK must be at least 3 + 5*N .
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
diff --git a/SRC/zhbevx.f b/SRC/zhbevx.f
index 38786534..16cb51e9 100644
--- a/SRC/zhbevx.f
+++ b/SRC/zhbevx.f
@@ -95,8 +95,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AB is overwritten by values generated during the
 *>          reduction to tridiagonal form.
 *> \endverbatim
@@ -156,24 +155,20 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing AB to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*DLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
 *>          eigenvectors did not converge, try setting ABSTOL to
 *>          2*DLAMCH('S').
-*> \endverbatim
-*> \verbatim
+*>
 *>          See "Computing Small Singular Values of Bidiagonal Matrices
 *>          with Guaranteed High Relative Accuracy," by Demmel and
 *>          Kahan, LAPACK Working Note #3.
diff --git a/SRC/zhbgst.f b/SRC/zhbgst.f
index 2903db6d..82c1df9c 100644
--- a/SRC/zhbgst.f
+++ b/SRC/zhbgst.f
@@ -94,8 +94,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(ka+1+i-j,j) = A(i,j) for max(1,j-ka)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+ka).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the transformed matrix X**H*A*X, stored in the same
 *>          format as A.
 *> \endverbatim
diff --git a/SRC/zhbgv.f b/SRC/zhbgv.f
index 7fbb2aa3..108c5403 100644
--- a/SRC/zhbgv.f
+++ b/SRC/zhbgv.f
@@ -90,8 +90,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(ka+1+i-j,j) = A(i,j) for max(1,j-ka)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+ka).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AB are destroyed.
 *> \endverbatim
 *>
@@ -110,8 +109,7 @@
 *>          as follows:
 *>          if UPLO = 'U', BB(kb+1+i-j,j) = B(i,j) for max(1,j-kb)<=i<=j;
 *>          if UPLO = 'L', BB(1+i-j,j)    = B(i,j) for j<=i<=min(n,j+kb).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the factor S from the split Cholesky factorization
 *>          B = S**H*S, as returned by ZPBSTF.
 *> \endverbatim
diff --git a/SRC/zhbgvd.f b/SRC/zhbgvd.f
index 4cb21953..77f37dc3 100644
--- a/SRC/zhbgvd.f
+++ b/SRC/zhbgvd.f
@@ -101,8 +101,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(ka+1+i-j,j) = A(i,j) for max(1,j-ka)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+ka).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AB are destroyed.
 *> \endverbatim
 *>
@@ -121,8 +120,7 @@
 *>          as follows:
 *>          if UPLO = 'U', BB(kb+1+i-j,j) = B(i,j) for max(1,j-kb)<=i<=j;
 *>          if UPLO = 'L', BB(1+i-j,j)    = B(i,j) for j<=i<=min(n,j+kb).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the factor S from the split Cholesky factorization
 *>          B = S**H*S, as returned by ZPBSTF.
 *> \endverbatim
@@ -169,8 +167,7 @@
 *>          If N <= 1,               LWORK >= 1.
 *>          If JOBZ = 'N' and N > 1, LWORK >= N.
 *>          If JOBZ = 'V' and N > 1, LWORK >= 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK, RWORK and
 *>          IWORK arrays, returns these values as the first entries of
@@ -191,8 +188,7 @@
 *>          If N <= 1,               LRWORK >= 1.
 *>          If JOBZ = 'N' and N > 1, LRWORK >= N.
 *>          If JOBZ = 'V' and N > 1, LRWORK >= 1 + 5*N + 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LRWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
@@ -212,8 +208,7 @@
 *>          The dimension of array IWORK.
 *>          If JOBZ = 'N' or N <= 1, LIWORK >= 1.
 *>          If JOBZ = 'V' and N > 1, LIWORK >= 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
diff --git a/SRC/zhbgvx.f b/SRC/zhbgvx.f
index fc9e112e..c5bbfb86 100644
--- a/SRC/zhbgvx.f
+++ b/SRC/zhbgvx.f
@@ -105,8 +105,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(ka+1+i-j,j) = A(i,j) for max(1,j-ka)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+ka).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AB are destroyed.
 *> \endverbatim
 *>
@@ -125,8 +124,7 @@
 *>          as follows:
 *>          if UPLO = 'U', BB(kb+1+i-j,j) = B(i,j) for max(1,j-kb)<=i<=j;
 *>          if UPLO = 'L', BB(1+i-j,j)    = B(i,j) for j<=i<=min(n,j+kb).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the factor S from the split Cholesky factorization
 *>          B = S**H*S, as returned by ZPBSTF.
 *> \endverbatim
@@ -161,8 +159,7 @@
 *> \param[in] VU
 *> \verbatim
 *>          VU is DOUBLE PRECISION
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='V', the lower and upper bounds of the interval to
 *>          be searched for eigenvalues. VL < VU.
 *>          Not referenced if RANGE = 'A' or 'I'.
@@ -176,8 +173,7 @@
 *> \param[in] IU
 *> \verbatim
 *>          IU is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='I', the indices (in ascending order) of the
 *>          smallest and largest eigenvalues to be returned.
 *>          1 <= IL <= IU <= N, if N > 0; IL = 1 and IU = 0 if N = 0.
@@ -191,17 +187,14 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing AP to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*DLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
diff --git a/SRC/zhbtrd.f b/SRC/zhbtrd.f
index 167b01b6..0fa7f4ea 100644
--- a/SRC/zhbtrd.f
+++ b/SRC/zhbtrd.f
@@ -114,8 +114,7 @@
 *>          Q is COMPLEX*16 array, dimension (LDQ,N)
 *>          On entry, if VECT = 'U', then Q must contain an N-by-N
 *>          matrix X; if VECT = 'N' or 'V', then Q need not be set.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit:
 *>          if VECT = 'V', Q contains the N-by-N unitary matrix Q;
 *>          if VECT = 'U', Q contains the product X*Q;
diff --git a/SRC/zheev.f b/SRC/zheev.f
index a7cdfceb..3a0c65c2 100644
--- a/SRC/zheev.f
+++ b/SRC/zheev.f
@@ -103,8 +103,7 @@
 *>          The length of the array WORK.  LWORK >= max(1,2*N-1).
 *>          For optimal efficiency, LWORK >= (NB+1)*N,
 *>          where NB is the blocksize for ZHETRD returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zheevd.f b/SRC/zheevd.f
index 04ba3314..20d932a9 100644
--- a/SRC/zheevd.f
+++ b/SRC/zheevd.f
@@ -113,8 +113,7 @@
 *>          If N <= 1,                LWORK must be at least 1.
 *>          If JOBZ  = 'N' and N > 1, LWORK must be at least N + 1.
 *>          If JOBZ  = 'V' and N > 1, LWORK must be at least 2*N + N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK, RWORK and
 *>          IWORK arrays, returns these values as the first entries of
@@ -137,8 +136,7 @@
 *>          If JOBZ  = 'N' and N > 1, LRWORK must be at least N.
 *>          If JOBZ  = 'V' and N > 1, LRWORK must be at least
 *>                         1 + 5*N + 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LRWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
@@ -159,8 +157,7 @@
 *>          If N <= 1,                LIWORK must be at least 1.
 *>          If JOBZ  = 'N' and N > 1, LIWORK must be at least 1.
 *>          If JOBZ  = 'V' and N > 1, LIWORK must be at least 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
diff --git a/SRC/zheevr.f b/SRC/zheevr.f
index 980d22d8..c59f26c0 100644
--- a/SRC/zheevr.f
+++ b/SRC/zheevr.f
@@ -187,22 +187,18 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing A to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          See "Computing Small Singular Values of Bidiagonal Matrices
 *>          with Guaranteed High Relative Accuracy," by Demmel and
 *>          Kahan, LAPACK Working Note #3.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If high relative accuracy is important, set ABSTOL to
 *>          DLAMCH( 'Safe minimum' ).  Doing so will guarantee that
 *>          eigenvalues are computed to high relative accuracy when
@@ -272,8 +268,7 @@
 *>          For optimal efficiency, LWORK >= (NB+1)*N,
 *>          where NB is the max of the blocksize for ZHETRD and for
 *>          ZUNMTR as returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK, RWORK and
 *>          IWORK arrays, returns these values as the first entries of
@@ -292,8 +287,7 @@
 *> \verbatim
 *>          LRWORK is INTEGER
 *>          The length of the array RWORK.  LRWORK >= max(1,24*N).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LRWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
@@ -312,8 +306,7 @@
 *> \verbatim
 *>          LIWORK is INTEGER
 *>          The dimension of the array IWORK.  LIWORK >= max(1,10*N).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
diff --git a/SRC/zheevx.f b/SRC/zheevx.f
index 3b96e40f..02bac469 100644
--- a/SRC/zheevx.f
+++ b/SRC/zheevx.f
@@ -131,24 +131,20 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing A to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*DLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
 *>          eigenvectors did not converge, try setting ABSTOL to
 *>          2*DLAMCH('S').
-*> \endverbatim
-*> \verbatim
+*>
 *>          See "Computing Small Singular Values of Bidiagonal Matrices
 *>          with Guaranteed High Relative Accuracy," by Demmel and
 *>          Kahan, LAPACK Working Note #3.
@@ -205,8 +201,7 @@
 *>          For optimal efficiency, LWORK >= (NB+1)*N,
 *>          where NB is the max of the blocksize for ZHETRD and for
 *>          ZUNMTR as returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zhegs2.f b/SRC/zhegs2.f
index 4a40b25c..eaa65340 100644
--- a/SRC/zhegs2.f
+++ b/SRC/zhegs2.f
@@ -82,8 +82,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the transformed matrix, stored in the
 *>          same format as A.
 *> \endverbatim
diff --git a/SRC/zhegst.f b/SRC/zhegst.f
index 42da2982..dd95053d 100644
--- a/SRC/zhegst.f
+++ b/SRC/zhegst.f
@@ -82,8 +82,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the transformed matrix, stored in the
 *>          same format as A.
 *> \endverbatim
diff --git a/SRC/zhegv.f b/SRC/zhegv.f
index 8f3c647c..f51a7df0 100644
--- a/SRC/zhegv.f
+++ b/SRC/zhegv.f
@@ -84,8 +84,7 @@
 *>          upper triangular part of the matrix A.  If UPLO = 'L',
 *>          the leading N-by-N lower triangular part of A contains
 *>          the lower triangular part of the matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if JOBZ = 'V', then if INFO = 0, A contains the
 *>          matrix Z of eigenvectors.  The eigenvectors are normalized
 *>          as follows:
@@ -110,8 +109,7 @@
 *>          contains the upper triangular part of the matrix B.
 *>          If UPLO = 'L', the leading N-by-N lower triangular part of B
 *>          contains the lower triangular part of the matrix B.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO <= N, the part of B containing the matrix is
 *>          overwritten by the triangular factor U or L from the Cholesky
 *>          factorization B = U**H*U or B = L*L**H.
@@ -141,8 +139,7 @@
 *>          The length of the array WORK.  LWORK >= max(1,2*N-1).
 *>          For optimal efficiency, LWORK >= (NB+1)*N,
 *>          where NB is the blocksize for ZHETRD returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zhegvd.f b/SRC/zhegvd.f
index 5736fab8..ca7c8435 100644
--- a/SRC/zhegvd.f
+++ b/SRC/zhegvd.f
@@ -92,8 +92,7 @@
 *>          upper triangular part of the matrix A.  If UPLO = 'L',
 *>          the leading N-by-N lower triangular part of A contains
 *>          the lower triangular part of the matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if JOBZ = 'V', then if INFO = 0, A contains the
 *>          matrix Z of eigenvectors.  The eigenvectors are normalized
 *>          as follows:
@@ -118,8 +117,7 @@
 *>          upper triangular part of the matrix B.  If UPLO = 'L',
 *>          the leading N-by-N lower triangular part of B contains
 *>          the lower triangular part of the matrix B.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO <= N, the part of B containing the matrix is
 *>          overwritten by the triangular factor U or L from the Cholesky
 *>          factorization B = U**H*U or B = L*L**H.
@@ -150,8 +148,7 @@
 *>          If N <= 1,                LWORK >= 1.
 *>          If JOBZ  = 'N' and N > 1, LWORK >= N + 1.
 *>          If JOBZ  = 'V' and N > 1, LWORK >= 2*N + N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK, RWORK and
 *>          IWORK arrays, returns these values as the first entries of
@@ -172,8 +169,7 @@
 *>          If N <= 1,                LRWORK >= 1.
 *>          If JOBZ  = 'N' and N > 1, LRWORK >= N.
 *>          If JOBZ  = 'V' and N > 1, LRWORK >= 1 + 5*N + 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LRWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
@@ -194,8 +190,7 @@
 *>          If N <= 1,                LIWORK >= 1.
 *>          If JOBZ  = 'N' and N > 1, LIWORK >= 1.
 *>          If JOBZ  = 'V' and N > 1, LIWORK >= 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
diff --git a/SRC/zhegvx.f b/SRC/zhegvx.f
index 3eb5f7f3..331f6bbe 100644
--- a/SRC/zhegvx.f
+++ b/SRC/zhegvx.f
@@ -138,8 +138,7 @@
 *> \param[in] VU
 *> \verbatim
 *>          VU is DOUBLE PRECISION
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='V', the lower and upper bounds of the interval to
 *>          be searched for eigenvalues. VL < VU.
 *>          Not referenced if RANGE = 'A' or 'I'.
@@ -153,8 +152,7 @@
 *> \param[in] IU
 *> \verbatim
 *>          IU is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='I', the indices (in ascending order) of the
 *>          smallest and largest eigenvalues to be returned.
 *>          1 <= IL <= IU <= N, if N > 0; IL = 1 and IU = 0 if N = 0.
@@ -238,8 +236,7 @@
 *>          The length of the array WORK.  LWORK >= max(1,2*N).
 *>          For optimal efficiency, LWORK >= (NB+1)*N,
 *>          where NB is the blocksize for ZHETRD returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zherfs.f b/SRC/zherfs.f
index 176744a4..eee0f291 100644
--- a/SRC/zherfs.f
+++ b/SRC/zherfs.f
@@ -168,12 +168,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/zherfsx.f b/SRC/zherfsx.f
index d26362a8..0a5e13c7 100644
--- a/SRC/zherfsx.f
+++ b/SRC/zherfsx.f
@@ -218,37 +218,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -257,8 +251,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -269,14 +262,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -284,26 +275,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -314,8 +301,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -334,8 +320,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0D+0
@@ -346,8 +331,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -357,8 +341,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/zhesv.f b/SRC/zhesv.f
index 6fddcf6b..837d60e8 100644
--- a/SRC/zhesv.f
+++ b/SRC/zhesv.f
@@ -85,8 +85,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the block diagonal matrix D and the
 *>          multipliers used to obtain the factor U or L from the
 *>          factorization A = U*D*U**H or A = L*D*L**H as computed by
@@ -140,8 +139,7 @@
 *>          ZHETRF.
 *>          for LWORK < N, TRS will be done with Level BLAS 2
 *>          for LWORK >= N, TRS will be done with Level BLAS 3
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zhesvx.f b/SRC/zhesvx.f
index ac993176..86f88665 100644
--- a/SRC/zhesvx.f
+++ b/SRC/zhesvx.f
@@ -136,8 +136,7 @@
 *>          contains the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
 *>          A = U*D*U**H or A = L*D*L**H as computed by ZHETRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AF is an output argument and on exit
 *>          returns the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
@@ -163,8 +162,7 @@
 *>          is a 2-by-2 diagonal block.  If UPLO = 'L' and IPIV(k) =
 *>          IPIV(k+1) < 0, then rows and columns k+1 and -IPIV(k) were
 *>          interchanged and D(k:k+1,k:k+1) is a 2-by-2 diagonal block.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains details of the interchanges and the block structure
 *>          of D, as determined by ZHETRF.
@@ -237,8 +235,7 @@
 *>          The length of WORK.  LWORK >= max(1,2*N), and for best
 *>          performance, when FACT = 'N', LWORK >= max(1,2*N,N*NB), where
 *>          NB is the optimal blocksize for ZHETRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zhesvxx.f b/SRC/zhesvxx.f
index c180cbed..01063ade 100644
--- a/SRC/zhesvxx.f
+++ b/SRC/zhesvxx.f
@@ -168,8 +168,7 @@
 *>     N-by-N lower triangular part of A contains the lower
 *>     triangular part of the matrix A, and the strictly upper
 *>     triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>     On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>     diag(S)*A*diag(S).
 *> \endverbatim
@@ -187,8 +186,7 @@
 *>     contains the block diagonal matrix D and the multipliers
 *>     used to obtain the factor U or L from the factorization A =
 *>     U*D*U**T or A = L*D*L**T as computed by DSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then AF is an output argument and on exit
 *>     returns the block diagonal matrix D and the multipliers
 *>     used to obtain the factor U or L from the factorization A =
@@ -214,8 +212,7 @@
 *>     diagonal block.  If UPLO = 'L' and IPIV(k) = IPIV(k+1) < 0,
 *>     then rows and columns k+1 and -IPIV(k) were interchanged
 *>     and D(k:k+1,k:k+1) is a 2-by-2 diagonal block.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then IPIV is an output argument and on exit
 *>     contains details of the interchanges and the block
 *>     structure of D, as determined by ZHETRF.
@@ -326,37 +323,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -365,8 +356,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -377,14 +367,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -392,26 +380,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -422,8 +406,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -442,8 +425,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0D+0
@@ -451,8 +433,7 @@
 *>                    computed.
 *>            = 1.0 : Use the extra-precise refinement algorithm.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -462,8 +443,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/zheswapr.f b/SRC/zheswapr.f
index 58720bcd..80b01358 100644
--- a/SRC/zheswapr.f
+++ b/SRC/zheswapr.f
@@ -61,8 +61,7 @@
 *>          A is COMPLEX*16 array, dimension (LDA,N)
 *>          On entry, the NB diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by CSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/zhetf2.f b/SRC/zhetf2.f
index 68e95261..d28ba0a9 100644
--- a/SRC/zhetf2.f
+++ b/SRC/zhetf2.f
@@ -76,8 +76,7 @@
 *>          leading n-by-n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L (see below for further details).
 *> \endverbatim
diff --git a/SRC/zhetrd.f b/SRC/zhetrd.f
index 3eeb574a..e2b1ae76 100644
--- a/SRC/zhetrd.f
+++ b/SRC/zhetrd.f
@@ -118,8 +118,7 @@
 *>          The dimension of the array WORK.  LWORK >= 1.
 *>          For optimum performance LWORK >= N*NB, where NB is the
 *>          optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zhetrf.f b/SRC/zhetrf.f
index e6bf0f34..38c84d0d 100644
--- a/SRC/zhetrf.f
+++ b/SRC/zhetrf.f
@@ -75,8 +75,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L (see below for further details).
 *> \endverbatim
diff --git a/SRC/zhetri.f b/SRC/zhetri.f
index 299f88d8..d7e69fed 100644
--- a/SRC/zhetri.f
+++ b/SRC/zhetri.f
@@ -64,8 +64,7 @@
 *>          A is COMPLEX*16 array, dimension (LDA,N)
 *>          On entry, the block diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by ZHETRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (Hermitian) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/zhetri2.f b/SRC/zhetri2.f
index 2ec6a01b..a79684e6 100644
--- a/SRC/zhetri2.f
+++ b/SRC/zhetri2.f
@@ -65,8 +65,7 @@
 *>          A is COMPLEX*16 array, dimension (LDA,N)
 *>          On entry, the NB diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by ZHETRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/zhetri2x.f b/SRC/zhetri2x.f
index 9398a2d6..789b88d5 100644
--- a/SRC/zhetri2x.f
+++ b/SRC/zhetri2x.f
@@ -64,8 +64,7 @@
 *>          A is COMPLEX*16 array, dimension (LDA,N)
 *>          On entry, the NNB diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by ZHETRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/zhfrk.f b/SRC/zhfrk.f
index 6d7b8898..22375210 100644
--- a/SRC/zhfrk.f
+++ b/SRC/zhfrk.f
@@ -68,16 +68,13 @@
 *>           On  entry,   UPLO  specifies  whether  the  upper  or  lower
 *>           triangular  part  of the  array  C  is to be  referenced  as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'U' or 'u'   Only the  upper triangular part of  C
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'L' or 'l'   Only the  lower triangular part of  C
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -86,14 +83,11 @@
 *>          TRANS is CHARACTER*1
 *>           On entry,  TRANS  specifies the operation to be performed as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              TRANS = 'N' or 'n'   C := alpha*A*A**H + beta*C.
-*> \endverbatim
-*> \verbatim
+*>
 *>              TRANS = 'C' or 'c'   C := alpha*A**H*A + beta*C.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/zhgeqz.f b/SRC/zhgeqz.f
index 30a9f865..9630edaa 100644
--- a/SRC/zhgeqz.f
+++ b/SRC/zhgeqz.f
@@ -180,8 +180,7 @@
 *>          The real non-negative scalars beta that define the
 *>          eigenvalues of GNEP.  BETA(i) = P(i,i) in the generalized
 *>          Schur factorization.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Together, the quantities alpha = ALPHA(j) and beta = BETA(j)
 *>          represent the j-th eigenvalue of the matrix pair (A,B), in
 *>          one of the forms lambda = alpha/beta or mu = beta/alpha.
@@ -235,8 +234,7 @@
 *> \verbatim
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.  LWORK >= max(1,N).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zhpev.f b/SRC/zhpev.f
index 333b38e9..11cc330e 100644
--- a/SRC/zhpev.f
+++ b/SRC/zhpev.f
@@ -72,8 +72,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AP is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the diagonal
 *>          and first superdiagonal of the tridiagonal matrix T overwrite
diff --git a/SRC/zhpevd.f b/SRC/zhpevd.f
index 5a47db45..0b035545 100644
--- a/SRC/zhpevd.f
+++ b/SRC/zhpevd.f
@@ -81,8 +81,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AP is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the diagonal
 *>          and first superdiagonal of the tridiagonal matrix T overwrite
@@ -126,8 +125,7 @@
 *>          If N <= 1,               LWORK must be at least 1.
 *>          If JOBZ = 'N' and N > 1, LWORK must be at least N.
 *>          If JOBZ = 'V' and N > 1, LWORK must be at least 2*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the required sizes of the WORK, RWORK and
 *>          IWORK arrays, returns these values as the first entries of
@@ -150,8 +148,7 @@
 *>          If JOBZ = 'N' and N > 1, LRWORK must be at least N.
 *>          If JOBZ = 'V' and N > 1, LRWORK must be at least
 *>                    1 + 5*N + 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LRWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the required sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
@@ -171,8 +168,7 @@
 *>          The dimension of array IWORK.
 *>          If JOBZ  = 'N' or N <= 1, LIWORK must be at least 1.
 *>          If JOBZ  = 'V' and N > 1, LIWORK must be at least 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the required sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
diff --git a/SRC/zhpevx.f b/SRC/zhpevx.f
index 7ee20df3..05590799 100644
--- a/SRC/zhpevx.f
+++ b/SRC/zhpevx.f
@@ -86,8 +86,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, AP is overwritten by values generated during the
 *>          reduction to tridiagonal form.  If UPLO = 'U', the diagonal
 *>          and first superdiagonal of the tridiagonal matrix T overwrite
@@ -130,24 +129,20 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing AP to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*DLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
 *>          eigenvectors did not converge, try setting ABSTOL to
 *>          2*DLAMCH('S').
-*> \endverbatim
-*> \verbatim
+*>
 *>          See "Computing Small Singular Values of Bidiagonal Matrices
 *>          with Guaranteed High Relative Accuracy," by Demmel and
 *>          Kahan, LAPACK Working Note #3.
diff --git a/SRC/zhpgst.f b/SRC/zhpgst.f
index 782d640c..41640849 100644
--- a/SRC/zhpgst.f
+++ b/SRC/zhpgst.f
@@ -80,8 +80,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the transformed matrix, stored in the
 *>          same format as A.
 *> \endverbatim
diff --git a/SRC/zhpgv.f b/SRC/zhpgv.f
index 6934af12..67f8674b 100644
--- a/SRC/zhpgv.f
+++ b/SRC/zhpgv.f
@@ -84,8 +84,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AP are destroyed.
 *> \endverbatim
 *>
@@ -97,8 +96,7 @@
 *>          is stored in the array BP as follows:
 *>          if UPLO = 'U', BP(i + (j-1)*j/2) = B(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', BP(i + (j-1)*(2*n-j)/2) = B(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the triangular factor U or L from the Cholesky
 *>          factorization B = U**H*U or B = L*L**H, in the same storage
 *>          format as B.
diff --git a/SRC/zhpgvd.f b/SRC/zhpgvd.f
index b999ba9d..76cf7d82 100644
--- a/SRC/zhpgvd.f
+++ b/SRC/zhpgvd.f
@@ -93,8 +93,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AP are destroyed.
 *> \endverbatim
 *>
@@ -106,8 +105,7 @@
 *>          is stored in the array BP as follows:
 *>          if UPLO = 'U', BP(i + (j-1)*j/2) = B(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', BP(i + (j-1)*(2*n-j)/2) = B(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the triangular factor U or L from the Cholesky
 *>          factorization B = U**H*U or B = L*L**H, in the same storage
 *>          format as B.
@@ -149,8 +147,7 @@
 *>          If N <= 1,               LWORK >= 1.
 *>          If JOBZ = 'N' and N > 1, LWORK >= N.
 *>          If JOBZ = 'V' and N > 1, LWORK >= 2*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the required sizes of the WORK, RWORK and
 *>          IWORK arrays, returns these values as the first entries of
@@ -171,8 +168,7 @@
 *>          If N <= 1,               LRWORK >= 1.
 *>          If JOBZ = 'N' and N > 1, LRWORK >= N.
 *>          If JOBZ = 'V' and N > 1, LRWORK >= 1 + 5*N + 2*N**2.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LRWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the required sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
@@ -192,8 +188,7 @@
 *>          The dimension of array IWORK.
 *>          If JOBZ  = 'N' or N <= 1, LIWORK >= 1.
 *>          If JOBZ  = 'V' and N > 1, LIWORK >= 3 + 5*N.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the required sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
diff --git a/SRC/zhpgvx.f b/SRC/zhpgvx.f
index 284aaa7d..42821bde 100644
--- a/SRC/zhpgvx.f
+++ b/SRC/zhpgvx.f
@@ -98,8 +98,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2*n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the contents of AP are destroyed.
 *> \endverbatim
 *>
@@ -111,8 +110,7 @@
 *>          is stored in the array BP as follows:
 *>          if UPLO = 'U', BP(i + (j-1)*j/2) = B(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', BP(i + (j-1)*(2*n-j)/2) = B(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the triangular factor U or L from the Cholesky
 *>          factorization B = U**H*U or B = L*L**H, in the same storage
 *>          format as B.
@@ -126,8 +124,7 @@
 *> \param[in] VU
 *> \verbatim
 *>          VU is DOUBLE PRECISION
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='V', the lower and upper bounds of the interval to
 *>          be searched for eigenvalues. VL < VU.
 *>          Not referenced if RANGE = 'A' or 'I'.
@@ -141,8 +138,7 @@
 *> \param[in] IU
 *> \verbatim
 *>          IU is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='I', the indices (in ascending order) of the
 *>          smallest and largest eigenvalues to be returned.
 *>          1 <= IL <= IU <= N, if N > 0; IL = 1 and IU = 0 if N = 0.
@@ -156,17 +152,14 @@
 *>          An approximate eigenvalue is accepted as converged
 *>          when it is determined to lie in an interval [a,b]
 *>          of width less than or equal to
-*> \endverbatim
-*> \verbatim
+*>
 *>                  ABSTOL + EPS *   max( |a|,|b| ) ,
-*> \endverbatim
-*> \verbatim
+*>
 *>          where EPS is the machine precision.  If ABSTOL is less than
 *>          or equal to zero, then  EPS*|T|  will be used in its place,
 *>          where |T| is the 1-norm of the tridiagonal matrix obtained
 *>          by reducing AP to tridiagonal form.
-*> \endverbatim
-*> \verbatim
+*>
 *>          Eigenvalues will be computed most accurately when ABSTOL is
 *>          set to twice the underflow threshold 2*DLAMCH('S'), not zero.
 *>          If this routine returns with INFO>0, indicating that some
@@ -199,8 +192,7 @@
 *>          The eigenvectors are normalized as follows:
 *>          if ITYPE = 1 or 2, Z**H*B*Z = I;
 *>          if ITYPE = 3, Z**H*inv(B)*Z = I.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If an eigenvector fails to converge, then that column of Z
 *>          contains the latest approximation to the eigenvector, and the
 *>          index of the eigenvector is returned in IFAIL.
diff --git a/SRC/zhprfs.f b/SRC/zhprfs.f
index 8312fb95..214fc50a 100644
--- a/SRC/zhprfs.f
+++ b/SRC/zhprfs.f
@@ -156,12 +156,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/zhpsv.f b/SRC/zhpsv.f
index ef5b665c..2e6cc1b8 100644
--- a/SRC/zhpsv.f
+++ b/SRC/zhpsv.f
@@ -83,8 +83,7 @@
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
 *>          A = U*D*U**H or A = L*D*L**H as computed by ZHPTRF, stored as
diff --git a/SRC/zhpsvx.f b/SRC/zhpsvx.f
index 62372598..95b1ecb7 100644
--- a/SRC/zhpsvx.f
+++ b/SRC/zhpsvx.f
@@ -128,8 +128,7 @@
 *>          to obtain the factor U or L from the factorization
 *>          A = U*D*U**H or A = L*D*L**H as computed by ZHPTRF, stored as
 *>          a packed triangular matrix in the same storage format as A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AFP is an output argument and on exit
 *>          contains the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
@@ -150,8 +149,7 @@
 *>          is a 2-by-2 diagonal block.  If UPLO = 'L' and IPIV(k) =
 *>          IPIV(k+1) < 0, then rows and columns k+1 and -IPIV(k) were
 *>          interchanged and D(k:k+1,k:k+1) is a 2-by-2 diagonal block.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains details of the interchanges and the block structure
 *>          of D, as determined by ZHPTRF.
diff --git a/SRC/zhptrf.f b/SRC/zhptrf.f
index d5f7df25..97007053 100644
--- a/SRC/zhptrf.f
+++ b/SRC/zhptrf.f
@@ -70,8 +70,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L, stored as a packed triangular
 *>          matrix overwriting A (see below for further details).
diff --git a/SRC/zhptri.f b/SRC/zhptri.f
index 43b2eb82..7cc6a139 100644
--- a/SRC/zhptri.f
+++ b/SRC/zhptri.f
@@ -65,8 +65,7 @@
 *>          On entry, the block diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by ZHPTRF,
 *>          stored as a packed triangular matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (Hermitian) inverse of the original
 *>          matrix, stored as a packed triangular matrix. The j-th column
 *>          of inv(A) is stored in the array AP as follows:
diff --git a/SRC/zhseqr.f b/SRC/zhseqr.f
index 838d54bd..00856d8b 100644
--- a/SRC/zhseqr.f
+++ b/SRC/zhseqr.f
@@ -82,8 +82,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>           It is assumed that H is already upper triangular in rows
 *>           and columns 1:ILO-1 and IHI+1:N. ILO and IHI are normally
 *>           set by a previous call to ZGEBAL, and then passed to ZGEHRD
@@ -102,8 +101,7 @@
 *>           Schur form). If INFO = 0 and JOB = 'E', the contents of
 *>           H are unspecified on exit.  (The output value of H when
 *>           INFO.GT.0 is given under the description of INFO below.)
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unlike earlier versions of ZHSEQR, this subroutine may
 *>           explicitly H(i,j) = 0 for i.GT.j and j = 1, 2, ... ILO-1
 *>           or j = IHI+1, IHI+2, ... N.
@@ -162,8 +160,7 @@
 *>           may be required for optimal performance.  A workspace
 *>           query is recommended to determine the optimal workspace
 *>           size.
-*> \endverbatim
-*> \verbatim
+*>
 *>           If LWORK = -1, then ZHSEQR does a workspace query.
 *>           In this case, ZHSEQR checks the input parameters and
 *>           estimates the optimal workspace size for the given
@@ -182,42 +179,33 @@
 *>                the eigenvalues.  Elements 1:ilo-1 and i+1:n of WR
 *>                and WI contain those eigenvalues which have been
 *>                successfully computed.  (Failures are rare.)
-*> \endverbatim
-*> \verbatim
+*>
 *>                If INFO .GT. 0 and JOB = 'E', then on exit, the
 *>                remaining unconverged eigenvalues are the eigen-
 *>                values of the upper Hessenberg matrix rows and
 *>                columns ILO through INFO of the final, output
 *>                value of H.
-*> \endverbatim
-*> \verbatim
+*>
 *>                If INFO .GT. 0 and JOB   = 'S', then on exit
-*> \endverbatim
-*> \verbatim
+*>
 *>           (*)  (initial value of H)*U  = U*(final value of H)
-*> \endverbatim
-*> \verbatim
+*>
 *>                where U is a unitary matrix.  The final
 *>                value of  H is upper Hessenberg and triangular in
 *>                rows and columns INFO+1 through IHI.
-*> \endverbatim
-*> \verbatim
+*>
 *>                If INFO .GT. 0 and COMPZ = 'V', then on exit
-*> \endverbatim
-*> \verbatim
+*>
 *>                  (final value of Z)  =  (initial value of Z)*U
-*> \endverbatim
-*> \verbatim
+*>
 *>                where U is the unitary matrix in (*) (regard-
 *>                less of the value of JOB.)
-*> \endverbatim
-*> \verbatim
+*>
 *>                If INFO .GT. 0 and COMPZ = 'I', then on exit
 *>                      (final value of Z)  = U
 *>                where U is the unitary matrix in (*) (regard-
 *>                less of the value of JOB.)
-*> \endverbatim
-*> \verbatim
+*>
 *>                If INFO .GT. 0 and COMPZ = 'N', then Z is not
 *>                accessed.
 *> \endverbatim
diff --git a/SRC/zla_gbamv.f b/SRC/zla_gbamv.f
index b369f1de..1b9fa777 100644
--- a/SRC/zla_gbamv.f
+++ b/SRC/zla_gbamv.f
@@ -63,13 +63,11 @@
 *>          TRANS is INTEGER
 *>           On entry, TRANS specifies the operation to be performed as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>             BLAS_NO_TRANS      y := alpha*abs(A)*abs(x) + beta*abs(y)
 *>             BLAS_TRANS         y := alpha*abs(A**T)*abs(x) + beta*abs(y)
 *>             BLAS_CONJ_TRANS    y := alpha*abs(A**T)*abs(x) + beta*abs(y)
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -169,8 +167,7 @@
 *>           On entry, INCY specifies the increment for the elements of
 *>           Y. INCY must not be zero.
 *>           Unchanged on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Level 2 Blas routine.
 *> \endverbatim
 *>
diff --git a/SRC/zla_geamv.f b/SRC/zla_geamv.f
index 949c7e8e..cbf1516a 100644
--- a/SRC/zla_geamv.f
+++ b/SRC/zla_geamv.f
@@ -64,13 +64,11 @@
 *>          TRANS is INTEGER
 *>           On entry, TRANS specifies the operation to be performed as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>             BLAS_NO_TRANS      y := alpha*abs(A)*abs(x) + beta*abs(y)
 *>             BLAS_TRANS         y := alpha*abs(A**T)*abs(x) + beta*abs(y)
 *>             BLAS_CONJ_TRANS    y := alpha*abs(A**T)*abs(x) + beta*abs(y)
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -158,8 +156,7 @@
 *>           On entry, INCY specifies the increment for the elements of
 *>           Y. INCY must not be zero.
 *>           Unchanged on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Level 2 Blas routine.
 *> \endverbatim
 *>
diff --git a/SRC/zla_heamv.f b/SRC/zla_heamv.f
index ed87815f..616c3864 100644
--- a/SRC/zla_heamv.f
+++ b/SRC/zla_heamv.f
@@ -63,16 +63,13 @@
 *>           On entry, UPLO specifies whether the upper or lower
 *>           triangular part of the array A is to be referenced as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = BLAS_UPPER   Only the upper triangular part of A
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = BLAS_LOWER   Only the lower triangular part of A
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/zla_herfsx_extended.f b/SRC/zla_herfsx_extended.f
index aeaeb9d1..bd075d97 100644
--- a/SRC/zla_herfsx_extended.f
+++ b/SRC/zla_herfsx_extended.f
@@ -200,37 +200,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -239,8 +233,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     This subroutine is only responsible for setting the second field
 *>     above.
 *>     See Lapack Working Note 165 for further details and extra
@@ -254,14 +247,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -269,26 +260,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -299,8 +286,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     This subroutine is only responsible for setting the second field
 *>     above.
 *>     See Lapack Working Note 165 for further details and extra
diff --git a/SRC/zla_porfsx_extended.f b/SRC/zla_porfsx_extended.f
index 930ce6ac..53eaefc2 100644
--- a/SRC/zla_porfsx_extended.f
+++ b/SRC/zla_porfsx_extended.f
@@ -192,37 +192,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -231,8 +225,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     This subroutine is only responsible for setting the second field
 *>     above.
 *>     See Lapack Working Note 165 for further details and extra
@@ -246,14 +239,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -261,26 +252,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -291,8 +278,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     This subroutine is only responsible for setting the second field
 *>     above.
 *>     See Lapack Working Note 165 for further details and extra
diff --git a/SRC/zla_syamv.f b/SRC/zla_syamv.f
index 8aead800..3d1e6918 100644
--- a/SRC/zla_syamv.f
+++ b/SRC/zla_syamv.f
@@ -64,16 +64,13 @@
 *>           On entry, UPLO specifies whether the upper or lower
 *>           triangular part of the array A is to be referenced as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = BLAS_UPPER   Only the upper triangular part of A
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = BLAS_LOWER   Only the lower triangular part of A
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/zla_syrfsx_extended.f b/SRC/zla_syrfsx_extended.f
index 3ceb9da0..ae0c7ae6 100644
--- a/SRC/zla_syrfsx_extended.f
+++ b/SRC/zla_syrfsx_extended.f
@@ -200,37 +200,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -239,8 +233,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     This subroutine is only responsible for setting the second field
 *>     above.
 *>     See Lapack Working Note 165 for further details and extra
@@ -254,14 +247,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -269,26 +260,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * slamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * slamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * slamch('Epsilon') to determine if the error
@@ -299,8 +286,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     This subroutine is only responsible for setting the second field
 *>     above.
 *>     See Lapack Working Note 165 for further details and extra
diff --git a/SRC/zlahef.f b/SRC/zlahef.f
index 4a056757..04650bda 100644
--- a/SRC/zlahef.f
+++ b/SRC/zlahef.f
@@ -113,8 +113,7 @@
 *>          Details of the interchanges and the block structure of D.
 *>          If UPLO = 'U', only the last KB elements of IPIV are set;
 *>          if UPLO = 'L', only the first KB elements are set.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If IPIV(k) > 0, then rows and columns k and IPIV(k) were
 *>          interchanged and D(k,k) is a 1-by-1 diagonal block.
 *>          If UPLO = 'U' and IPIV(k) = IPIV(k-1) < 0, then rows and
diff --git a/SRC/zlahqr.f b/SRC/zlahqr.f
index 73964b9f..4c955ce2 100644
--- a/SRC/zlahqr.f
+++ b/SRC/zlahqr.f
@@ -145,22 +145,19 @@
 *>                  per eigenvalue; elements i+1:ihi of W contain
 *>                  those eigenvalues which have been successfully
 *>                  computed.
-*> \endverbatim
-*> \verbatim
+*>
 *>                  If INFO .GT. 0 and WANTT is .FALSE., then on exit,
 *>                  the remaining unconverged eigenvalues are the
 *>                  eigenvalues of the upper Hessenberg matrix
 *>                  rows and columns ILO thorugh INFO of the final,
 *>                  output value of H.
-*> \endverbatim
-*> \verbatim
+*>
 *>                  If INFO .GT. 0 and WANTT is .TRUE., then on exit
 *>          (*)       (initial value of H)*U  = U*(final value of H)
 *>                  where U is an orthognal matrix.    The final
 *>                  value of H is upper Hessenberg and triangular in
 *>                  rows and columns INFO+1 through IHI.
-*> \endverbatim
-*> \verbatim
+*>
 *>                  If INFO .GT. 0 and WANTZ is .TRUE., then on exit
 *>                      (final value of Z)  = (initial value of Z)*U
 *>                  where U is the orthogonal matrix in (*)
diff --git a/SRC/zlals0.f b/SRC/zlals0.f
index ae421c9f..c7a68817 100644
--- a/SRC/zlals0.f
+++ b/SRC/zlals0.f
@@ -102,8 +102,7 @@
 *>          SQRE is INTEGER
 *>         = 0: the lower block is an NR-by-NR square matrix.
 *>         = 1: the lower block is an NR-by-(NR+1) rectangular matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>         The bidiagonal matrix has row dimension N = NL + NR + 1,
 *>         and column dimension M = N + SQRE.
 *> \endverbatim
diff --git a/SRC/zlanhf.f b/SRC/zlanhf.f
index 635fee2e..7e9e136d 100644
--- a/SRC/zlanhf.f
+++ b/SRC/zlanhf.f
@@ -83,12 +83,10 @@
 *>          UPLO is CHARACTER
 *>            On entry, UPLO specifies whether the RFP matrix A came from
 *>            an upper or lower triangular matrix as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>            UPLO = 'U' or 'u' RFP A came from an upper triangular
 *>            matrix
-*> \endverbatim
-*> \verbatim
+*>
 *>            UPLO = 'L' or 'l' RFP A came from a  lower triangular
 *>            matrix
 *> \endverbatim
diff --git a/SRC/zlaqgb.f b/SRC/zlaqgb.f
index 88b144a5..927ec89f 100644
--- a/SRC/zlaqgb.f
+++ b/SRC/zlaqgb.f
@@ -77,8 +77,7 @@
 *>          The j-th column of A is stored in the j-th column of the
 *>          array AB as follows:
 *>          AB(ku+1+i-j,j) = A(i,j) for max(1,j-ku)<=i<=min(m,j+kl)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the equilibrated matrix, in the same storage format
 *>          as A.  See EQUED for the form of the equilibrated matrix.
 *> \endverbatim
@@ -130,18 +129,15 @@
 *>                  by diag(C).
 *>          = 'B':  Both row and column equilibration, i.e., A has been
 *>                  replaced by diag(R) * A * diag(C).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if row or column scaling
 *>  should be done based on the ratio of the row or column scaling
 *>  factors.  If ROWCND < THRESH, row scaling is done, and if
 *>  COLCND < THRESH, column scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if row scaling
 *>  should be done based on the absolute size of the largest matrix
 *>  element.  If AMAX > LARGE or AMAX < SMALL, row scaling is done.
diff --git a/SRC/zlaqge.f b/SRC/zlaqge.f
index 60134dbc..857b088b 100644
--- a/SRC/zlaqge.f
+++ b/SRC/zlaqge.f
@@ -112,18 +112,15 @@
 *>                  by diag(C).
 *>          = 'B':  Both row and column equilibration, i.e., A has been
 *>                  replaced by diag(R) * A * diag(C).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if row or column scaling
 *>  should be done based on the ratio of the row or column scaling
 *>  factors.  If ROWCND < THRESH, row scaling is done, and if
 *>  COLCND < THRESH, column scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if row scaling
 *>  should be done based on the absolute size of the largest matrix
 *>  element.  If AMAX > LARGE or AMAX < SMALL, row scaling is done.
diff --git a/SRC/zlaqhb.f b/SRC/zlaqhb.f
index 7a4acfb1..c96a1081 100644
--- a/SRC/zlaqhb.f
+++ b/SRC/zlaqhb.f
@@ -75,8 +75,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**H *U or A = L*L**H of the band
 *>          matrix A, in the same storage format as A.
@@ -113,17 +112,14 @@
 *>          = 'N':  No equilibration.
 *>          = 'Y':  Equilibration was done, i.e., A has been replaced by
 *>                  diag(S) * A * diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if scaling should be done
 *>  based on the ratio of the scaling factors.  If SCOND < THRESH,
 *>  scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if scaling should
 *>  be done based on the absolute size of the largest matrix element.
 *>  If AMAX > LARGE or AMAX < SMALL, scaling is done.
diff --git a/SRC/zlaqhe.f b/SRC/zlaqhe.f
index c63fbca7..1028cdca 100644
--- a/SRC/zlaqhe.f
+++ b/SRC/zlaqhe.f
@@ -69,8 +69,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if EQUED = 'Y', the equilibrated matrix:
 *>          diag(S) * A * diag(S).
 *> \endverbatim
@@ -106,17 +105,14 @@
 *>          = 'N':  No equilibration.
 *>          = 'Y':  Equilibration was done, i.e., A has been replaced by
 *>                  diag(S) * A * diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if scaling should be done
 *>  based on the ratio of the scaling factors.  If SCOND < THRESH,
 *>  scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if scaling should
 *>  be done based on the absolute size of the largest matrix element.
 *>  If AMAX > LARGE or AMAX < SMALL, scaling is done.
diff --git a/SRC/zlaqhp.f b/SRC/zlaqhp.f
index 23b5b4cd..f2ae7067 100644
--- a/SRC/zlaqhp.f
+++ b/SRC/zlaqhp.f
@@ -67,8 +67,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the equilibrated matrix:  diag(S) * A * diag(S), in
 *>          the same storage format as A.
 *> \endverbatim
@@ -98,17 +97,14 @@
 *>          = 'N':  No equilibration.
 *>          = 'Y':  Equilibration was done, i.e., A has been replaced by
 *>                  diag(S) * A * diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if scaling should be done
 *>  based on the ratio of the scaling factors.  If SCOND < THRESH,
 *>  scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if scaling should
 *>  be done based on the absolute size of the largest matrix element.
 *>  If AMAX > LARGE or AMAX < SMALL, scaling is done.
diff --git a/SRC/zlaqr0.f b/SRC/zlaqr0.f
index dc2f81b4..71f45987 100644
--- a/SRC/zlaqr0.f
+++ b/SRC/zlaqr0.f
@@ -78,8 +78,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>           It is assumed that H is already upper triangular in rows
 *>           and columns 1:ILO-1 and IHI+1:N and, if ILO.GT.1,
 *>           H(ILO,ILO-1) is zero. ILO and IHI are normally set by a
@@ -100,8 +99,7 @@
 *>           .FALSE., then the contents of H are unspecified on exit.
 *>           (The output value of H when INFO.GT.0 is given under the
 *>           description of INFO below.)
-*> \endverbatim
-*> \verbatim
+*>
 *>           This subroutine may explicitly set H(i,j) = 0 for i.GT.j and
 *>           j = 1, 2, ... ILO-1 or j = IHI+1, IHI+2, ... N.
 *> \endverbatim
@@ -166,8 +164,7 @@
 *>           is sufficient, but LWORK typically as large as 6*N may
 *>           be required for optimal performance.  A workspace query
 *>           to determine the optimal workspace size is recommended.
-*> \endverbatim
-*> \verbatim
+*>
 *>           If LWORK = -1, then ZLAQR0 does a workspace query.
 *>           In this case, ZLAQR0 checks the input parameters and
 *>           estimates the optimal workspace size for the given
diff --git a/SRC/zlaqr1.f b/SRC/zlaqr1.f
index 6239c161..461d6048 100644
--- a/SRC/zlaqr1.f
+++ b/SRC/zlaqr1.f
@@ -76,8 +76,7 @@
 *> \param[in] S2
 *> \verbatim
 *>          S2 is COMPLEX*16
-*> \endverbatim
-*> \verbatim
+*>
 *>          S1 and S2 are the shifts defining K in (*) above.
 *> \endverbatim
 *>
diff --git a/SRC/zlaqr2.f b/SRC/zlaqr2.f
index 00cdb3d0..7e3b5d0a 100644
--- a/SRC/zlaqr2.f
+++ b/SRC/zlaqr2.f
@@ -240,8 +240,7 @@
 *>          The dimension of the work array WORK.  LWORK = 2*NW
 *>          suffices, but greater efficiency may result from larger
 *>          values of LWORK.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; ZLAQR2
 *>          only estimates the optimal workspace size for the given
 *>          values of N, NW, KTOP and KBOT.  The estimate is returned
diff --git a/SRC/zlaqr3.f b/SRC/zlaqr3.f
index 425fa9fa..864221ea 100644
--- a/SRC/zlaqr3.f
+++ b/SRC/zlaqr3.f
@@ -237,8 +237,7 @@
 *>          The dimension of the work array WORK.  LWORK = 2*NW
 *>          suffices, but greater efficiency may result from larger
 *>          values of LWORK.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; ZLAQR3
 *>          only estimates the optimal workspace size for the given
 *>          values of N, NW, KTOP and KBOT.  The estimate is returned
diff --git a/SRC/zlaqr4.f b/SRC/zlaqr4.f
index 43279b98..94042bdb 100644
--- a/SRC/zlaqr4.f
+++ b/SRC/zlaqr4.f
@@ -105,8 +105,7 @@
 *>           .FALSE., then the contents of H are unspecified on exit.
 *>           (The output value of H when INFO.GT.0 is given under the
 *>           description of INFO below.)
-*> \endverbatim
-*> \verbatim
+*>
 *>           This subroutine may explicitly set H(i,j) = 0 for i.GT.j and
 *>           j = 1, 2, ... ILO-1 or j = IHI+1, IHI+2, ... N.
 *> \endverbatim
@@ -171,8 +170,7 @@
 *>           is sufficient, but LWORK typically as large as 6*N may
 *>           be required for optimal performance.  A workspace query
 *>           to determine the optimal workspace size is recommended.
-*> \endverbatim
-*> \verbatim
+*>
 *>           If LWORK = -1, then ZLAQR4 does a workspace query.
 *>           In this case, ZLAQR4 checks the input parameters and
 *>           estimates the optimal workspace size for the given
diff --git a/SRC/zlaqsb.f b/SRC/zlaqsb.f
index 7f647116..81029b8f 100644
--- a/SRC/zlaqsb.f
+++ b/SRC/zlaqsb.f
@@ -75,8 +75,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**H *U or A = L*L**H of the band
 *>          matrix A, in the same storage format as A.
@@ -113,17 +112,14 @@
 *>          = 'N':  No equilibration.
 *>          = 'Y':  Equilibration was done, i.e., A has been replaced by
 *>                  diag(S) * A * diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if scaling should be done
 *>  based on the ratio of the scaling factors.  If SCOND < THRESH,
 *>  scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if scaling should
 *>  be done based on the absolute size of the largest matrix element.
 *>  If AMAX > LARGE or AMAX < SMALL, scaling is done.
diff --git a/SRC/zlaqsp.f b/SRC/zlaqsp.f
index c2ca10bb..edea5462 100644
--- a/SRC/zlaqsp.f
+++ b/SRC/zlaqsp.f
@@ -67,8 +67,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the equilibrated matrix:  diag(S) * A * diag(S), in
 *>          the same storage format as A.
 *> \endverbatim
@@ -98,17 +97,14 @@
 *>          = 'N':  No equilibration.
 *>          = 'Y':  Equilibration was done, i.e., A has been replaced by
 *>                  diag(S) * A * diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if scaling should be done
 *>  based on the ratio of the scaling factors.  If SCOND < THRESH,
 *>  scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if scaling should
 *>  be done based on the absolute size of the largest matrix element.
 *>  If AMAX > LARGE or AMAX < SMALL, scaling is done.
diff --git a/SRC/zlaqsy.f b/SRC/zlaqsy.f
index 05a6a7ed..dc4e8e91 100644
--- a/SRC/zlaqsy.f
+++ b/SRC/zlaqsy.f
@@ -69,8 +69,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if EQUED = 'Y', the equilibrated matrix:
 *>          diag(S) * A * diag(S).
 *> \endverbatim
@@ -106,17 +105,14 @@
 *>          = 'N':  No equilibration.
 *>          = 'Y':  Equilibration was done, i.e., A has been replaced by
 *>                  diag(S) * A * diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  THRESH is a threshold value used to decide if scaling should be done
 *>  based on the ratio of the scaling factors.  If SCOND < THRESH,
 *>  scaling is done.
-*> \endverbatim
-*> \verbatim
+*>
 *>  LARGE and SMALL are threshold values used to decide if scaling should
 *>  be done based on the absolute size of the largest matrix element.
 *>  If AMAX > LARGE or AMAX < SMALL, scaling is done.
diff --git a/SRC/zlascl.f b/SRC/zlascl.f
index 6fff6e13..55ab2ded 100644
--- a/SRC/zlascl.f
+++ b/SRC/zlascl.f
@@ -86,8 +86,7 @@
 *> \param[in] CTO
 *> \verbatim
 *>          CTO is DOUBLE PRECISION
-*> \endverbatim
-*> \verbatim
+*>
 *>          The matrix A is multiplied by CTO/CFROM. A(I,J) is computed
 *>          without over/underflow if the final result CTO*A(I,J)/CFROM
 *>          can be represented without over/underflow.  CFROM must be
diff --git a/SRC/zlasyf.f b/SRC/zlasyf.f
index 83d5665f..8a1b39a9 100644
--- a/SRC/zlasyf.f
+++ b/SRC/zlasyf.f
@@ -113,8 +113,7 @@
 *>          Details of the interchanges and the block structure of D.
 *>          If UPLO = 'U', only the last KB elements of IPIV are set;
 *>          if UPLO = 'L', only the first KB elements are set.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If IPIV(k) > 0, then rows and columns k and IPIV(k) were
 *>          interchanged and D(k,k) is a 1-by-1 diagonal block.
 *>          If UPLO = 'U' and IPIV(k) = IPIV(k-1) < 0, then rows and
diff --git a/SRC/zlatbs.f b/SRC/zlatbs.f
index 1d1521c4..464d47f2 100644
--- a/SRC/zlatbs.f
+++ b/SRC/zlatbs.f
@@ -137,15 +137,13 @@
 *> \param[in,out] CNORM
 *> \verbatim
 *>          CNORM is or output) DOUBLE PRECISION array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'Y', CNORM is an input argument and CNORM(j)
 *>          contains the norm of the off-diagonal part of the j-th column
 *>          of A.  If TRANS = 'N', CNORM(j) must be greater than or equal
 *>          to the infinity-norm, and if TRANS = 'T' or 'C', CNORM(j)
 *>          must be greater than or equal to the 1-norm.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'N', CNORM is an output argument and CNORM(j)
 *>          returns the 1-norm of the offdiagonal part of the j-th column
 *>          of A.
diff --git a/SRC/zlatps.f b/SRC/zlatps.f
index ee56c5fc..92607131 100644
--- a/SRC/zlatps.f
+++ b/SRC/zlatps.f
@@ -125,15 +125,13 @@
 *> \param[in,out] CNORM
 *> \verbatim
 *>          CNORM is or output) DOUBLE PRECISION array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'Y', CNORM is an input argument and CNORM(j)
 *>          contains the norm of the off-diagonal part of the j-th column
 *>          of A.  If TRANS = 'N', CNORM(j) must be greater than or equal
 *>          to the infinity-norm, and if TRANS = 'T' or 'C', CNORM(j)
 *>          must be greater than or equal to the 1-norm.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'N', CNORM is an output argument and CNORM(j)
 *>          returns the 1-norm of the offdiagonal part of the j-th column
 *>          of A.
diff --git a/SRC/zlatrs.f b/SRC/zlatrs.f
index 35c6158a..ad603df7 100644
--- a/SRC/zlatrs.f
+++ b/SRC/zlatrs.f
@@ -133,15 +133,13 @@
 *> \param[in,out] CNORM
 *> \verbatim
 *>          CNORM is or output) DOUBLE PRECISION array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'Y', CNORM is an input argument and CNORM(j)
 *>          contains the norm of the off-diagonal part of the j-th column
 *>          of A.  If TRANS = 'N', CNORM(j) must be greater than or equal
 *>          to the infinity-norm, and if TRANS = 'T' or 'C', CNORM(j)
 *>          must be greater than or equal to the 1-norm.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If NORMIN = 'N', CNORM is an output argument and CNORM(j)
 *>          returns the 1-norm of the offdiagonal part of the j-th column
 *>          of A.
diff --git a/SRC/zlatzm.f b/SRC/zlatzm.f
index cc6e1e39..231b5e53 100644
--- a/SRC/zlatzm.f
+++ b/SRC/zlatzm.f
@@ -107,8 +107,7 @@
 *>                         (M,1)   if SIDE = 'R'
 *>          On entry, the n-vector C1 if SIDE = 'L', or the m-vector C1
 *>          if SIDE = 'R'.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the first row of P*C if SIDE = 'L', or the first
 *>          column of C*P if SIDE = 'R'.
 *> \endverbatim
@@ -120,8 +119,7 @@
 *>                         (LDC, N-1) if SIDE = 'R'
 *>          On entry, the (m - 1) x n matrix C2 if SIDE = 'L', or the
 *>          m x (n - 1) matrix C2 if SIDE = 'R'.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, rows 2:m of P*C if SIDE = 'L', or columns 2:m of C*P
 *>          if SIDE = 'R'.
 *> \endverbatim
diff --git a/SRC/zpbrfs.f b/SRC/zpbrfs.f
index ea70883c..5d975af0 100644
--- a/SRC/zpbrfs.f
+++ b/SRC/zpbrfs.f
@@ -165,12 +165,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/zpbstf.f b/SRC/zpbstf.f
index 97a0596d..0e48bfb1 100644
--- a/SRC/zpbstf.f
+++ b/SRC/zpbstf.f
@@ -82,8 +82,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor S from the split Cholesky
 *>          factorization A = S**H*S. See Further Details.
 *> \endverbatim
diff --git a/SRC/zpbsv.f b/SRC/zpbsv.f
index 466dca6c..38a3822e 100644
--- a/SRC/zpbsv.f
+++ b/SRC/zpbsv.f
@@ -90,8 +90,7 @@
 *>          if UPLO = 'U', AB(KD+1+i-j,j) = A(i,j) for max(1,j-KD)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(N,j+KD).
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**H *U or A = L*L**H of the band
 *>          matrix A, in the same storage format as A.
diff --git a/SRC/zpbsvx.f b/SRC/zpbsvx.f
index af34cf18..a7739ee7 100644
--- a/SRC/zpbsvx.f
+++ b/SRC/zpbsvx.f
@@ -145,8 +145,7 @@
 *>          if UPLO = 'U', AB(KD+1+i-j,j) = A(i,j) for max(1,j-KD)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(N,j+KD).
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>          diag(S)*A*diag(S).
 *> \endverbatim
@@ -165,13 +164,11 @@
 *>          factorization A = U**H *U or A = L*L**H of the band matrix
 *>          A, in the same storage format as A (see AB).  If EQUED = 'Y',
 *>          then AFB is the factored form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AFB is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**H *U or A = L*L**H.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then AFB is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**H *U or A = L*L**H of the equilibrated
diff --git a/SRC/zpbtf2.f b/SRC/zpbtf2.f
index 24c83d35..3b291174 100644
--- a/SRC/zpbtf2.f
+++ b/SRC/zpbtf2.f
@@ -81,8 +81,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**H *U or A = L*L**H of the band
 *>          matrix A, in the same storage format as A.
diff --git a/SRC/zpbtrf.f b/SRC/zpbtrf.f
index 5290f030..202dd4c0 100644
--- a/SRC/zpbtrf.f
+++ b/SRC/zpbtrf.f
@@ -76,8 +76,7 @@
 *>          as follows:
 *>          if UPLO = 'U', AB(kd+1+i-j,j) = A(i,j) for max(1,j-kd)<=i<=j;
 *>          if UPLO = 'L', AB(1+i-j,j)    = A(i,j) for j<=i<=min(n,j+kd).
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**H*U or A = L*L**H of the band
 *>          matrix A, in the same storage format as A.
diff --git a/SRC/zpftrf.f b/SRC/zpftrf.f
index 218f615b..e4e116e6 100644
--- a/SRC/zpftrf.f
+++ b/SRC/zpftrf.f
@@ -82,8 +82,7 @@
 *>          of lower packed A. The LDA of RFP A is (N+1)/2 when TRANSR =
 *>          'C'. When TRANSR is 'N' the LDA is N+1 when N is even and N
 *>          is odd. See the Note below for more details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization RFP A = U**H*U or RFP A = L*L**H.
 *> \endverbatim
@@ -96,27 +95,22 @@
 *>          > 0:  if INFO = i, the leading minor of order i is not
 *>                positive definite, and the factorization could not be
 *>                completed.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Further Notes on RFP Format:
 *>  ============================
-*> \endverbatim
-*> \verbatim
+*>
 *>  We first consider Standard Packed Format when N is even.
 *>  We give an example where N = 6.
-*> \endverbatim
-*> \verbatim
+*>
 *>     AP is Upper             AP is Lower
-*> \endverbatim
-*> \verbatim
+*>
 *>   00 01 02 03 04 05       00
 *>      11 12 13 14 15       10 11
 *>         22 23 24 25       20 21 22
 *>            33 34 35       30 31 32 33
 *>               44 45       40 41 42 43 44
 *>                  55       50 51 52 53 54 55
-*> \endverbatim
-*> \verbatim
+*>
 *>  Let TRANSR = 'N'. RFP holds AP as follows:
 *>  For UPLO = 'U' the upper trapezoid A(0:5,0:2) consists of the last
 *>  three columns of AP upper. The lower triangle A(4:6,0:2) consists of
@@ -126,19 +120,16 @@
 *>  conjugate-transpose of the last three columns of AP lower.
 *>  To denote conjugate we place -- above the element. This covers the
 *>  case N even and TRANSR = 'N'.
-*> \endverbatim
-*> \verbatim
+*>
 *>         RFP A                   RFP A
-*> \endverbatim
-*> \verbatim
+*>
 *>                                -- -- --
 *>        03 04 05                33 43 53
 *>                                   -- --
 *>        13 14 15                00 44 54
 *>                                      --
 *>        23 24 25                10 11 55
-*> \endverbatim
-*> \verbatim
+*>
 *>        33 34 35                20 21 22
 *>        --
 *>        00 44 45                30 31 32
@@ -146,37 +137,30 @@
 *>        01 11 55                40 41 42
 *>        -- -- --
 *>        02 12 22                50 51 52
-*> \endverbatim
-*> \verbatim
+*>
 *>  Now let TRANSR = 'C'. RFP A in both UPLO cases is just the conjugate-
 *>  transpose of RFP A above. One therefore gets:
-*> \endverbatim
-*> \verbatim
+*>
 *>           RFP A                   RFP A
-*> \endverbatim
-*> \verbatim
+*>
 *>     -- -- -- --                -- -- -- -- -- --
 *>     03 13 23 33 00 01 02    33 00 10 20 30 40 50
 *>     -- -- -- -- --                -- -- -- -- --
 *>     04 14 24 34 44 11 12    43 44 11 21 31 41 51
 *>     -- -- -- -- -- --                -- -- -- --
 *>     05 15 25 35 45 55 22    53 54 55 22 32 42 52
-*> \endverbatim
-*> \verbatim
+*>
 *>  We next  consider Standard Packed Format when N is odd.
 *>  We give an example where N = 5.
-*> \endverbatim
-*> \verbatim
+*>
 *>     AP is Upper                 AP is Lower
-*> \endverbatim
-*> \verbatim
+*>
 *>   00 01 02 03 04              00
 *>      11 12 13 14              10 11
 *>         22 23 24              20 21 22
 *>            33 34              30 31 32 33
 *>               44              40 41 42 43 44
-*> \endverbatim
-*> \verbatim
+*>
 *>  Let TRANSR = 'N'. RFP holds AP as follows:
 *>  For UPLO = 'U' the upper trapezoid A(0:4,0:2) consists of the last
 *>  three columns of AP upper. The lower triangle A(3:4,0:1) consists of
@@ -186,31 +170,25 @@
 *>  conjugate-transpose of the last two   columns of AP lower.
 *>  To denote conjugate we place -- above the element. This covers the
 *>  case N odd  and TRANSR = 'N'.
-*> \endverbatim
-*> \verbatim
+*>
 *>         RFP A                   RFP A
-*> \endverbatim
-*> \verbatim
+*>
 *>                                   -- --
 *>        02 03 04                00 33 43
 *>                                      --
 *>        12 13 14                10 11 44
-*> \endverbatim
-*> \verbatim
+*>
 *>        22 23 24                20 21 22
 *>        --
 *>        00 33 34                30 31 32
 *>        -- --
 *>        01 11 44                40 41 42
-*> \endverbatim
-*> \verbatim
+*>
 *>  Now let TRANSR = 'C'. RFP A in both UPLO cases is just the conjugate-
 *>  transpose of RFP A above. One therefore gets:
-*> \endverbatim
-*> \verbatim
+*>
 *>           RFP A                   RFP A
-*> \endverbatim
-*> \verbatim
+*>
 *>     -- -- --                   -- -- -- -- -- --
 *>     02 12 22 00 01             00 10 20 30 40 50
 *>     -- -- -- --                   -- -- -- -- --
diff --git a/SRC/zpftri.f b/SRC/zpftri.f
index 88b860f5..35a6fe62 100644
--- a/SRC/zpftri.f
+++ b/SRC/zpftri.f
@@ -76,8 +76,7 @@
 *>          of lower packed A. The LDA of RFP A is (N+1)/2 when TRANSR =
 *>          'C'. When TRANSR is 'N' the LDA is N+1 when N is even and N
 *>          is odd. See the Note below for more details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the Hermitian inverse of the original matrix, in the
 *>          same storage format.
 *> \endverbatim
diff --git a/SRC/zporfs.f b/SRC/zporfs.f
index c7f8102e..8d56c397 100644
--- a/SRC/zporfs.f
+++ b/SRC/zporfs.f
@@ -159,12 +159,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/zporfsx.f b/SRC/zporfsx.f
index 0ecc4bf3..a1fb9b00 100644
--- a/SRC/zporfsx.f
+++ b/SRC/zporfsx.f
@@ -210,37 +210,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -249,8 +243,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -261,14 +254,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -276,26 +267,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -306,8 +293,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -326,8 +312,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0D+0
@@ -338,8 +323,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -349,8 +333,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/zposv.f b/SRC/zposv.f
index d79906ac..683f2870 100644
--- a/SRC/zposv.f
+++ b/SRC/zposv.f
@@ -82,8 +82,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization A = U**H *U or A = L*L**H.
 *> \endverbatim
diff --git a/SRC/zposvx.f b/SRC/zposvx.f
index 5d732daf..cf60118d 100644
--- a/SRC/zposvx.f
+++ b/SRC/zposvx.f
@@ -140,8 +140,7 @@
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.  A is not modified if
 *>          FACT = 'F' or 'N', or if FACT = 'E' and EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>          diag(S)*A*diag(S).
 *> \endverbatim
@@ -160,14 +159,12 @@
 *>          factorization A = U**H *U or A = L*L**H, in the same storage
 *>          format as A.  If EQUED .ne. 'N', then AF is the factored form
 *>          of the equilibrated matrix diag(S)*A*diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AF is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**H *U or A = L*L**H of the original
 *>          matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then AF is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**H *U or A = L*L**H of the equilibrated
diff --git a/SRC/zposvxx.f b/SRC/zposvxx.f
index c4a5700c..20a334a0 100644
--- a/SRC/zposvxx.f
+++ b/SRC/zposvxx.f
@@ -167,8 +167,7 @@
 *>     the strictly upper triangular part of A is not referenced.  A is
 *>     not modified if FACT = 'F' or 'N', or if FACT = 'E' and EQUED =
 *>     'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>     On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>     diag(S)*A*diag(S).
 *> \endverbatim
@@ -187,14 +186,12 @@
 *>     factorization A = U**T*U or A = L*L**T, in the same storage
 *>     format as A.  If EQUED .ne. 'N', then AF is the factored
 *>     form of the equilibrated matrix diag(S)*A*diag(S).
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then AF is an output argument and on exit
 *>     returns the triangular factor U or L from the Cholesky
 *>     factorization A = U**T*U or A = L*L**T of the original
 *>     matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'E', then AF is an output argument and on exit
 *>     returns the triangular factor U or L from the Cholesky
 *>     factorization A = U**T*U or A = L*L**T of the equilibrated
@@ -313,37 +310,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -352,8 +343,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -364,14 +354,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -379,26 +367,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -409,8 +393,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -429,8 +412,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0D+0
@@ -438,8 +420,7 @@
 *>                    computed.
 *>            = 1.0 : Use the extra-precise refinement algorithm.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -449,8 +430,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/zpotf2.f b/SRC/zpotf2.f
index 27ba410f..2abd1b32 100644
--- a/SRC/zpotf2.f
+++ b/SRC/zpotf2.f
@@ -74,8 +74,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization A = U**H *U  or A = L*L**H.
 *> \endverbatim
diff --git a/SRC/zpotrf.f b/SRC/zpotrf.f
index 7c7eb7d5..23106172 100644
--- a/SRC/zpotrf.f
+++ b/SRC/zpotrf.f
@@ -72,8 +72,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization A = U**H *U or A = L*L**H.
 *> \endverbatim
diff --git a/SRC/zpprfs.f b/SRC/zpprfs.f
index e417a136..01977a93 100644
--- a/SRC/zpprfs.f
+++ b/SRC/zpprfs.f
@@ -147,12 +147,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/zppsv.f b/SRC/zppsv.f
index 9f42c1f9..030bd5ec 100644
--- a/SRC/zppsv.f
+++ b/SRC/zppsv.f
@@ -81,8 +81,7 @@
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization A = U**H*U or A = L*L**H, in the same storage
 *>          format as A.
diff --git a/SRC/zppsvx.f b/SRC/zppsvx.f
index 52803cb4..f7fec165 100644
--- a/SRC/zppsvx.f
+++ b/SRC/zppsvx.f
@@ -138,8 +138,7 @@
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
 *>          See below for further details.  A is not modified if
 *>          FACT = 'F' or 'N', or if FACT = 'E' and EQUED = 'N' on exit.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>          diag(S)*A*diag(S).
 *> \endverbatim
@@ -152,14 +151,12 @@
 *>          factorization A = U**H*U or A = L*L**H, in the same storage
 *>          format as A.  If EQUED .ne. 'N', then AFP is the factored
 *>          form of the equilibrated matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AFP is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**H * U or A = L * L**H of the original
 *>          matrix A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'E', then AFP is an output argument and on exit
 *>          returns the triangular factor U or L from the Cholesky
 *>          factorization A = U**H * U or A = L * L**H of the equilibrated
diff --git a/SRC/zpptrf.f b/SRC/zpptrf.f
index 886418ad..eab3d046 100644
--- a/SRC/zpptrf.f
+++ b/SRC/zpptrf.f
@@ -69,8 +69,7 @@
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the triangular factor U or L from the
 *>          Cholesky factorization A = U**H*U or A = L*L**H, in the same
 *>          storage format as A.
diff --git a/SRC/zpptri.f b/SRC/zpptri.f
index 1d7f6181..53350477 100644
--- a/SRC/zpptri.f
+++ b/SRC/zpptri.f
@@ -65,8 +65,7 @@
 *>          array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = U(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = L(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the upper or lower triangle of the (Hermitian)
 *>          inverse of A, overwriting the input factor U or L.
 *> \endverbatim
diff --git a/SRC/zpstf2.f b/SRC/zpstf2.f
index f1080e25..472c706f 100644
--- a/SRC/zpstf2.f
+++ b/SRC/zpstf2.f
@@ -79,8 +79,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization as above.
 *> \endverbatim
diff --git a/SRC/zpstrf.f b/SRC/zpstrf.f
index cc3887f1..4be6d127 100644
--- a/SRC/zpstrf.f
+++ b/SRC/zpstrf.f
@@ -79,8 +79,7 @@
 *>          leading n by n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the factor U or L from the Cholesky
 *>          factorization as above.
 *> \endverbatim
diff --git a/SRC/zptrfs.f b/SRC/zptrfs.f
index 300bc2af..16e71ca5 100644
--- a/SRC/zptrfs.f
+++ b/SRC/zptrfs.f
@@ -159,12 +159,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/zspmv.f b/SRC/zspmv.f
index c7b096ae..4daf3024 100644
--- a/SRC/zspmv.f
+++ b/SRC/zspmv.f
@@ -53,16 +53,13 @@
 *>           On entry, UPLO specifies whether the upper or lower
 *>           triangular part of the matrix A is supplied in the packed
 *>           array AP as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'U' or 'u'   The upper triangular part of A is
 *>                                  supplied in AP.
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'L' or 'l'   The lower triangular part of A is
 *>                                  supplied in AP.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/zspr.f b/SRC/zspr.f
index 9ba27bf7..977d3265 100644
--- a/SRC/zspr.f
+++ b/SRC/zspr.f
@@ -53,16 +53,13 @@
 *>           On entry, UPLO specifies whether the upper or lower
 *>           triangular part of the matrix A is supplied in the packed
 *>           array AP as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'U' or 'u'   The upper triangular part of A is
 *>                                  supplied in AP.
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'L' or 'l'   The lower triangular part of A is
 *>                                  supplied in AP.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/zsprfs.f b/SRC/zsprfs.f
index cf055fba..8985b8c0 100644
--- a/SRC/zsprfs.f
+++ b/SRC/zsprfs.f
@@ -156,12 +156,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/zspsv.f b/SRC/zspsv.f
index 1f928db8..fa16bfd3 100644
--- a/SRC/zspsv.f
+++ b/SRC/zspsv.f
@@ -83,8 +83,7 @@
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
 *>          See below for further details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
 *>          A = U*D*U**T or A = L*D*L**T as computed by ZSPTRF, stored as
diff --git a/SRC/zspsvx.f b/SRC/zspsvx.f
index ee83aa87..b28194dc 100644
--- a/SRC/zspsvx.f
+++ b/SRC/zspsvx.f
@@ -128,8 +128,7 @@
 *>          to obtain the factor U or L from the factorization
 *>          A = U*D*U**T or A = L*D*L**T as computed by ZSPTRF, stored as
 *>          a packed triangular matrix in the same storage format as A.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AFP is an output argument and on exit
 *>          contains the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
@@ -150,8 +149,7 @@
 *>          is a 2-by-2 diagonal block.  If UPLO = 'L' and IPIV(k) =
 *>          IPIV(k+1) < 0, then rows and columns k+1 and -IPIV(k) were
 *>          interchanged and D(k:k+1,k:k+1) is a 2-by-2 diagonal block.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains details of the interchanges and the block structure
 *>          of D, as determined by ZSPTRF.
diff --git a/SRC/zsptrf.f b/SRC/zsptrf.f
index ec9722c8..2cbb295b 100644
--- a/SRC/zsptrf.f
+++ b/SRC/zsptrf.f
@@ -71,8 +71,7 @@
 *>          is stored in the array AP as follows:
 *>          if UPLO = 'U', AP(i + (j-1)*j/2) = A(i,j) for 1<=i<=j;
 *>          if UPLO = 'L', AP(i + (j-1)*(2n-j)/2) = A(i,j) for j<=i<=n.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L, stored as a packed triangular
 *>          matrix overwriting A (see below for further details).
diff --git a/SRC/zsptri.f b/SRC/zsptri.f
index f5904398..4400a3f3 100644
--- a/SRC/zsptri.f
+++ b/SRC/zsptri.f
@@ -65,8 +65,7 @@
 *>          On entry, the block diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by ZSPTRF,
 *>          stored as a packed triangular matrix.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix, stored as a packed triangular matrix. The j-th column
 *>          of inv(A) is stored in the array AP as follows:
diff --git a/SRC/zstedc.f b/SRC/zstedc.f
index d1872a87..88c6d32f 100644
--- a/SRC/zstedc.f
+++ b/SRC/zstedc.f
@@ -119,8 +119,7 @@
 *>          Note that for COMPZ = 'V', then if N is less than or
 *>          equal to the minimum divide size, usually 25, then LWORK need
 *>          only be 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal sizes of the WORK, RWORK and
 *>          IWORK arrays, returns these values as the first entries of
@@ -149,8 +148,7 @@
 *>          Note that for COMPZ = 'I' or 'V', then if N is less than or
 *>          equal to the minimum divide size, usually 25, then LRWORK
 *>          need only be max(1,2*(N-1)).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LRWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
@@ -176,8 +174,7 @@
 *>          Note that for COMPZ = 'I' or 'V', then if N is less than or
 *>          equal to the minimum divide size, usually 25, then LIWORK
 *>          need only be 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal sizes of the WORK, RWORK
 *>          and IWORK arrays, returns these values as the first entries
diff --git a/SRC/zstegr.f b/SRC/zstegr.f
index 5067e9d1..c8e61aef 100644
--- a/SRC/zstegr.f
+++ b/SRC/zstegr.f
@@ -111,8 +111,7 @@
 *> \param[in] VU
 *> \verbatim
 *>          VU is DOUBLE PRECISION
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='V', the lower and upper bounds of the interval to
 *>          be searched for eigenvalues. VL < VU.
 *>          Not referenced if RANGE = 'A' or 'I'.
@@ -126,8 +125,7 @@
 *> \param[in] IU
 *> \verbatim
 *>          IU is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='I', the indices (in ascending order) of the
 *>          smallest and largest eigenvalues to be returned.
 *>          1 <= IL <= IU <= N, if N > 0.
diff --git a/SRC/zstein.f b/SRC/zstein.f
index acf1be25..894b56c0 100644
--- a/SRC/zstein.f
+++ b/SRC/zstein.f
@@ -153,16 +153,13 @@
 *>          > 0: if INFO = i, then i eigenvectors failed to converge
 *>               in MAXITS iterations.  Their indices are stored in
 *>               array IFAIL.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  MAXITS  INTEGER, default = 5
 *>          The maximum number of iterations performed.
-*> \endverbatim
-*> \verbatim
+*>
 *>  EXTRA   INTEGER, default = 2
 *>          The number of iterations performed after norm growth
 *>          criterion is satisfied, should be at least 1.
diff --git a/SRC/zstemr.f b/SRC/zstemr.f
index e57b45bb..233aa44d 100644
--- a/SRC/zstemr.f
+++ b/SRC/zstemr.f
@@ -159,8 +159,7 @@
 *> \param[in] VU
 *> \verbatim
 *>          VU is DOUBLE PRECISION
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='V', the lower and upper bounds of the interval to
 *>          be searched for eigenvalues. VL < VU.
 *>          Not referenced if RANGE = 'A' or 'I'.
@@ -174,8 +173,7 @@
 *> \param[in] IU
 *> \verbatim
 *>          IU is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          If RANGE='I', the indices (in ascending order) of the
 *>          smallest and largest eigenvalues to be returned.
 *>          1 <= IL <= IU <= N, if N > 0.
diff --git a/SRC/zsymv.f b/SRC/zsymv.f
index cc6cbb46..03d9567f 100644
--- a/SRC/zsymv.f
+++ b/SRC/zsymv.f
@@ -53,16 +53,13 @@
 *>           On entry, UPLO specifies whether the upper or lower
 *>           triangular part of the array A is to be referenced as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'U' or 'u'   Only the upper triangular part of A
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'L' or 'l'   Only the lower triangular part of A
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/zsyr.f b/SRC/zsyr.f
index 170d12d5..adb4be39 100644
--- a/SRC/zsyr.f
+++ b/SRC/zsyr.f
@@ -53,16 +53,13 @@
 *>           On entry, UPLO specifies whether the upper or lower
 *>           triangular part of the array A is to be referenced as
 *>           follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'U' or 'u'   Only the upper triangular part of A
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>              UPLO = 'L' or 'l'   Only the lower triangular part of A
 *>                                  is to be referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/zsyrfs.f b/SRC/zsyrfs.f
index 1329bb12..2ed039d9 100644
--- a/SRC/zsyrfs.f
+++ b/SRC/zsyrfs.f
@@ -168,12 +168,10 @@
 *>          INFO is INTEGER
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
-*> \endverbatim
-*> \verbatim
+*>
 *>  ITMAX is the maximum number of steps of iterative refinement.
 *> \endverbatim
 *>
diff --git a/SRC/zsyrfsx.f b/SRC/zsyrfsx.f
index 9ff79ab4..ae66d58c 100644
--- a/SRC/zsyrfsx.f
+++ b/SRC/zsyrfsx.f
@@ -219,37 +219,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -258,8 +252,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -270,14 +263,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -285,26 +276,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -315,8 +302,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -335,8 +321,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0D+0
@@ -347,8 +332,7 @@
 *>                    compilation environment does not support DOUBLE
 *>                    PRECISION.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -358,8 +342,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/zsysv.f b/SRC/zsysv.f
index 245130ed..5e9270fa 100644
--- a/SRC/zsysv.f
+++ b/SRC/zsysv.f
@@ -85,8 +85,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the block diagonal matrix D and the
 *>          multipliers used to obtain the factor U or L from the
 *>          factorization A = U*D*U**T or A = L*D*L**T as computed by
@@ -140,8 +139,7 @@
 *>          ZSYTRF.
 *>          for LWORK < N, TRS will be done with Level BLAS 2
 *>          for LWORK >= N, TRS will be done with Level BLAS 3
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zsysvx.f b/SRC/zsysvx.f
index ad906aa5..addd8fe6 100644
--- a/SRC/zsysvx.f
+++ b/SRC/zsysvx.f
@@ -136,8 +136,7 @@
 *>          contains the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
 *>          A = U*D*U**T or A = L*D*L**T as computed by ZSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then AF is an output argument and on exit
 *>          returns the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L from the factorization
@@ -163,8 +162,7 @@
 *>          is a 2-by-2 diagonal block.  If UPLO = 'L' and IPIV(k) =
 *>          IPIV(k+1) < 0, then rows and columns k+1 and -IPIV(k) were
 *>          interchanged and D(k:k+1,k:k+1) is a 2-by-2 diagonal block.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If FACT = 'N', then IPIV is an output argument and on exit
 *>          contains details of the interchanges and the block structure
 *>          of D, as determined by ZSYTRF.
@@ -237,8 +235,7 @@
 *>          The length of WORK.  LWORK >= max(1,2*N), and for best
 *>          performance, when FACT = 'N', LWORK >= max(1,2*N,N*NB), where
 *>          NB is the optimal blocksize for ZSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zsysvxx.f b/SRC/zsysvxx.f
index d9ceef41..0394fee7 100644
--- a/SRC/zsysvxx.f
+++ b/SRC/zsysvxx.f
@@ -168,8 +168,7 @@
 *>     N-by-N lower triangular part of A contains the lower
 *>     triangular part of the matrix A, and the strictly upper
 *>     triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>     On exit, if FACT = 'E' and EQUED = 'Y', A is overwritten by
 *>     diag(S)*A*diag(S).
 *> \endverbatim
@@ -187,8 +186,7 @@
 *>     contains the block diagonal matrix D and the multipliers
 *>     used to obtain the factor U or L from the factorization A =
 *>     U*D*U**T or A = L*D*L**T as computed by DSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then AF is an output argument and on exit
 *>     returns the block diagonal matrix D and the multipliers
 *>     used to obtain the factor U or L from the factorization A =
@@ -214,8 +212,7 @@
 *>     diagonal block.  If UPLO = 'L' and IPIV(k) = IPIV(k+1) < 0,
 *>     then rows and columns k+1 and -IPIV(k) were interchanged
 *>     and D(k:k+1,k:k+1) is a 2-by-2 diagonal block.
-*> \endverbatim
-*> \verbatim
+*>
 *>     If FACT = 'N', then IPIV is an output argument and on exit
 *>     contains details of the interchanges and the block
 *>     structure of D, as determined by DSYTRF.
@@ -326,37 +323,31 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     normwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Normwise relative error in the ith solution vector:
 *>             max_j (abs(XTRUE(j,i) - X(j,i)))
 *>            ------------------------------
 *>                  max_j abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the type of error information as described
 *>     below. There currently are up to three pieces of information
 *>     returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_NORM(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_NORM(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated normwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -365,8 +356,7 @@
 *>              appropriately scaled matrix Z.
 *>              Let Z = S*A, where S scales each row by a power of the
 *>              radix so all absolute row sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -377,14 +367,12 @@
 *>     For each right-hand side, this array contains information about
 *>     various error bounds and condition numbers corresponding to the
 *>     componentwise relative error, which is defined as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>     Componentwise relative error in the ith solution vector:
 *>                    abs(XTRUE(j,i) - X(j,i))
 *>             max_j ----------------------
 *>                         abs(X(j,i))
-*> \endverbatim
-*> \verbatim
+*>
 *>     The array is indexed by the right-hand side i (on which the
 *>     componentwise relative error depends), and the type of error
 *>     information as described below. There currently are up to three
@@ -392,26 +380,22 @@
 *>     componentwise accuracy is not requested (PARAMS(3) = 0.0), then
 *>     ERR_BNDS_COMP is not accessed.  If N_ERR_BNDS .LT. 3, then at most
 *>     the first (:,N_ERR_BNDS) entries are returned.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The first index in ERR_BNDS_COMP(i,:) corresponds to the ith
 *>     right-hand side.
-*> \endverbatim
-*> \verbatim
+*>
 *>     The second index in ERR_BNDS_COMP(:,err) contains the following
 *>     three fields:
 *>     err = 1 "Trust/don't trust" boolean. Trust the answer if the
 *>              reciprocal condition number is less than the threshold
 *>              sqrt(n) * dlamch('Epsilon').
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 2 "Guaranteed" error bound: The estimated forward error,
 *>              almost certainly within a factor of 10 of the true error
 *>              so long as the next entry is greater than the threshold
 *>              sqrt(n) * dlamch('Epsilon'). This error bound should only
 *>              be trusted if the previous boolean is true.
-*> \endverbatim
-*> \verbatim
+*>
 *>     err = 3  Reciprocal condition number: Estimated componentwise
 *>              reciprocal condition number.  Compared with the threshold
 *>              sqrt(n) * dlamch('Epsilon') to determine if the error
@@ -422,8 +406,7 @@
 *>              current right-hand side and S scales each row of
 *>              A*diag(x) by a power of the radix so all absolute row
 *>              sums of Z are approximately 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>     See Lapack Working Note 165 for further details and extra
 *>     cautions.
 *> \endverbatim
@@ -442,8 +425,7 @@
 *>     that entry will be filled with default value used for that
 *>     parameter.  Only positions up to NPARAMS are accessed; defaults
 *>     are used for higher-numbered parameters.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITREF_I = 1) : Whether to perform iterative
 *>            refinement or not.
 *>         Default: 1.0D+0
@@ -451,8 +433,7 @@
 *>                    computed.
 *>            = 1.0 : Use the extra-precise refinement algorithm.
 *>              (other values are reserved for future use)
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_ITHRESH_I = 2) : Maximum number of residual
 *>            computations allowed for refinement.
 *>         Default: 10
@@ -462,8 +443,7 @@
 *>                     Gaussian elimination, the guarantees in
 *>                     err_bnds_norm and err_bnds_comp may no longer be
 *>                     trustworthy.
-*> \endverbatim
-*> \verbatim
+*>
 *>       PARAMS(LA_LINRX_CWISE_I = 3) : Flag determining if the code
 *>            will attempt to find a solution with small componentwise
 *>            relative error in the double-precision algorithm.  Positive
diff --git a/SRC/zsyswapr.f b/SRC/zsyswapr.f
index c8b8b499..dc243023 100644
--- a/SRC/zsyswapr.f
+++ b/SRC/zsyswapr.f
@@ -61,8 +61,7 @@
 *>          A is COMPLEX*16 array, dimension (LDA,N)
 *>          On entry, the NB diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by ZSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/zsytf2.f b/SRC/zsytf2.f
index 267499b3..225b676a 100644
--- a/SRC/zsytf2.f
+++ b/SRC/zsytf2.f
@@ -76,8 +76,7 @@
 *>          leading n-by-n lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L (see below for further details).
 *> \endverbatim
diff --git a/SRC/zsytrf.f b/SRC/zsytrf.f
index 4a0de5b1..dd206451 100644
--- a/SRC/zsytrf.f
+++ b/SRC/zsytrf.f
@@ -75,8 +75,7 @@
 *>          leading N-by-N lower triangular part of A contains the lower
 *>          triangular part of the matrix A, and the strictly upper
 *>          triangular part of A is not referenced.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the block diagonal matrix D and the multipliers used
 *>          to obtain the factor U or L (see below for further details).
 *> \endverbatim
@@ -111,8 +110,7 @@
 *>          LWORK is INTEGER
 *>          The length of WORK.  LWORK >=1.  For best performance
 *>          LWORK >= N*NB, where NB is the block size returned by ILAENV.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zsytri.f b/SRC/zsytri.f
index 0f83351f..09cbb2cc 100644
--- a/SRC/zsytri.f
+++ b/SRC/zsytri.f
@@ -64,8 +64,7 @@
 *>          A is COMPLEX*16 array, dimension (LDA,N)
 *>          On entry, the block diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by ZSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/zsytri2.f b/SRC/zsytri2.f
index 18a49da2..dc8decb4 100644
--- a/SRC/zsytri2.f
+++ b/SRC/zsytri2.f
@@ -65,8 +65,7 @@
 *>          A is COMPLEX*16 array, dimension (LDA,N)
 *>          On entry, the NB diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by ZSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/zsytri2x.f b/SRC/zsytri2x.f
index d289f1fc..83456834 100644
--- a/SRC/zsytri2x.f
+++ b/SRC/zsytri2x.f
@@ -64,8 +64,7 @@
 *>          A is COMPLEX*16 array, dimension (LDA,N)
 *>          On entry, the NNB diagonal matrix D and the multipliers
 *>          used to obtain the factor U or L as computed by ZSYTRF.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, if INFO = 0, the (symmetric) inverse of the original
 *>          matrix.  If UPLO = 'U', the upper triangular part of the
 *>          inverse is formed and the part of A below the diagonal is not
diff --git a/SRC/ztfsm.f b/SRC/ztfsm.f
index af2fc7d2..1e1f00e2 100644
--- a/SRC/ztfsm.f
+++ b/SRC/ztfsm.f
@@ -68,14 +68,11 @@
 *>          SIDE is CHARACTER*1
 *>           On entry, SIDE specifies whether op( A ) appears on the left
 *>           or right of X as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              SIDE = 'L' or 'l'   op( A )*X = alpha*B.
-*> \endverbatim
-*> \verbatim
+*>
 *>              SIDE = 'R' or 'r'   X*op( A ) = alpha*B.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -86,8 +83,7 @@
 *>           an upper or lower triangular matrix as follows:
 *>           UPLO = 'U' or 'u' RFP A came from an upper triangular matrix
 *>           UPLO = 'L' or 'l' RFP A came from a  lower triangular matrix
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -96,14 +92,11 @@
 *>          TRANS is CHARACTER*1
 *>           On entry, TRANS  specifies the form of op( A ) to be used
 *>           in the matrix multiplication as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              TRANS  = 'N' or 'n'   op( A ) = A.
-*> \endverbatim
-*> \verbatim
+*>
 *>              TRANS  = 'C' or 'c'   op( A ) = conjg( A' ).
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
@@ -112,15 +105,12 @@
 *>          DIAG is CHARACTER*1
 *>           On entry, DIAG specifies whether or not RFP A is unit
 *>           triangular as follows:
-*> \endverbatim
-*> \verbatim
+*>
 *>              DIAG = 'U' or 'u'   A is assumed to be unit triangular.
-*> \endverbatim
-*> \verbatim
+*>
 *>              DIAG = 'N' or 'n'   A is not assumed to be unit
 *>                                  triangular.
-*> \endverbatim
-*> \verbatim
+*>
 *>           Unchanged on exit.
 *> \endverbatim
 *>
diff --git a/SRC/ztftri.f b/SRC/ztftri.f
index 78512abc..f94c530d 100644
--- a/SRC/ztftri.f
+++ b/SRC/ztftri.f
@@ -85,8 +85,7 @@
 *>          elements of lower packed A. The LDA of RFP A is (N+1)/2 when
 *>          TRANSR = 'C'. When TRANSR is 'N' the LDA is N+1 when N is
 *>          even and N is odd. See the Note below for more details.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the (triangular) inverse of the original matrix, in
 *>          the same storage format.
 *> \endverbatim
diff --git a/SRC/ztgsen.f b/SRC/ztgsen.f
index 2fe42930..66455442 100644
--- a/SRC/ztgsen.f
+++ b/SRC/ztgsen.f
@@ -154,8 +154,7 @@
 *> \param[out] BETA
 *> \verbatim
 *>          BETA is COMPLEX*16 array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          The diagonal elements of A and B, respectively,
 *>          when the pair (A,B) has been reduced to generalized Schur
 *>          form.  ALPHA(i)/BETA(i) i=1,...,N are the generalized
@@ -213,8 +212,7 @@
 *> \param[out] PR
 *> \verbatim
 *>          PR is DOUBLE PRECISION
-*> \endverbatim
-*> \verbatim
+*>
 *>          If IJOB = 1, 4 or 5, PL, PR are lower bounds on the
 *>          reciprocal  of the norm of "projections" onto left and right
 *>          eigenspace with respect to the selected cluster.
@@ -247,8 +245,7 @@
 *>          The dimension of the array WORK. LWORK >=  1
 *>          If IJOB = 1, 2 or 4, LWORK >=  2*M*(N-M)
 *>          If IJOB = 3 or 5, LWORK >=  4*M*(N-M)
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
@@ -267,8 +264,7 @@
 *>          The dimension of the array IWORK. LIWORK >= 1.
 *>          If IJOB = 1, 2 or 4, LIWORK >=  N+2;
 *>          If IJOB = 3 or 5, LIWORK >= MAX(N+2, 2*M*(N-M));
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LIWORK = -1, then a workspace query is assumed; the
 *>          routine only calculates the optimal size of the IWORK array,
 *>          returns this value as the first entry of the IWORK array, and
diff --git a/SRC/ztgsja.f b/SRC/ztgsja.f
index 6beaf63d..7567e9df 100644
--- a/SRC/ztgsja.f
+++ b/SRC/ztgsja.f
@@ -186,8 +186,7 @@
 *> \param[in] L
 *> \verbatim
 *>          L is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          K and L specify the subblocks in the input matrices A and B:
 *>          A23 = A(K+1:MIN(K+L,M),N-L+1:N) and B13 = B(1:L,,N-L+1:N)
 *>          of A and B, whose GSVD is going to be computed by ZTGSJA.
@@ -230,8 +229,7 @@
 *> \param[in] TOLB
 *> \verbatim
 *>          TOLB is DOUBLE PRECISION
-*> \endverbatim
-*> \verbatim
+*>
 *>          TOLA and TOLB are the convergence criteria for the Jacobi-
 *>          Kogbetliantz iteration procedure. Generally, they are the
 *>          same as used in the preprocessing step, say
@@ -247,8 +245,7 @@
 *> \param[out] BETA
 *> \verbatim
 *>          BETA is DOUBLE PRECISION array, dimension (N)
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, ALPHA and BETA contain the generalized singular
 *>          value pairs of A and B;
 *>            ALPHA(1:K) = 1,
@@ -335,8 +332,7 @@
 *>          = 0:  successful exit
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value.
 *>          = 1:  the procedure does not converge after MAXIT cycles.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Internal Parameters
 *>  ===================
 *>
diff --git a/SRC/ztgsyl.f b/SRC/ztgsyl.f
index 9eefb571..da509d69 100644
--- a/SRC/ztgsyl.f
+++ b/SRC/ztgsyl.f
@@ -232,8 +232,7 @@
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK. LWORK > = 1.
 *>          If IJOB = 1 or 2 and TRANS = 'N', LWORK >= max(1,2*M*N).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/ztrexc.f b/SRC/ztrexc.f
index 8f00002e..aa5e8e79 100644
--- a/SRC/ztrexc.f
+++ b/SRC/ztrexc.f
@@ -96,8 +96,7 @@
 *> \param[in] ILST
 *> \verbatim
 *>          ILST is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          Specify the reordering of the diagonal elements of T:
 *>          The element with row index IFST is moved to row ILST by a
 *>          sequence of transpositions between adjacent elements.
diff --git a/SRC/ztrsen.f b/SRC/ztrsen.f
index b40776a4..be3e73de 100644
--- a/SRC/ztrsen.f
+++ b/SRC/ztrsen.f
@@ -161,8 +161,7 @@
 *>          If JOB = 'N', LWORK >= 1;
 *>          if JOB = 'E', LWORK = max(1,M*(N-M));
 *>          if JOB = 'V' or 'B', LWORK >= max(1,2*M*(N-M)).
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/ztrti2.f b/SRC/ztrti2.f
index 40f8e464..79710a6e 100644
--- a/SRC/ztrti2.f
+++ b/SRC/ztrti2.f
@@ -78,8 +78,7 @@
 *>          triangular part of A is not referenced.  If DIAG = 'U', the
 *>          diagonal elements of A are also not referenced and are
 *>          assumed to be 1.
-*> \endverbatim
-*> \verbatim
+*>
 *>          On exit, the (triangular) inverse of the original matrix, in
 *>          the same storage format.
 *> \endverbatim
diff --git a/SRC/ztzrzf.f b/SRC/ztzrzf.f
index 20fca661..f6ec0316 100644
--- a/SRC/ztzrzf.f
+++ b/SRC/ztzrzf.f
@@ -95,8 +95,7 @@
 *>          The dimension of the array WORK.  LWORK >= max(1,M).
 *>          For optimum performance LWORK >= M*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zunbdb.f b/SRC/zunbdb.f
index 36a5c13b..0db8e984 100644
--- a/SRC/zunbdb.f
+++ b/SRC/zunbdb.f
@@ -234,8 +234,7 @@
 *> \verbatim
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK. LWORK >= M-Q.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zuncsd.f b/SRC/zuncsd.f
index 208d2030..e5c1371c 100644
--- a/SRC/zuncsd.f
+++ b/SRC/zuncsd.f
@@ -252,8 +252,7 @@
 *> \verbatim
 *>          LWORK is INTEGER
 *>          The dimension of the array WORK.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the work array, and no error
@@ -275,8 +274,7 @@
 *> \verbatim
 *>          LRWORK is INTEGER
 *>          The dimension of the array RWORK.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LRWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the RWORK array, returns
 *>          this value as the first entry of the work array, and no error
@@ -295,12 +293,10 @@
 *>          < 0:  if INFO = -i, the i-th argument had an illegal value.
 *>          > 0:  ZBBCSD did not converge. See the description of RWORK
 *>                above for details.
-*> \endverbatim
-*> \verbatim
+*>
 *>  Reference
 *>  =========
-*> \endverbatim
-*> \verbatim
+*>
 *>  [1] Brian D. Sutton. Computing the complete CS decomposition. Numer.
 *>      Algorithms, 50(1):33-65, 2009.
 *> \endverbatim
diff --git a/SRC/zungbr.f b/SRC/zungbr.f
index f2343e81..7cfa7c6d 100644
--- a/SRC/zungbr.f
+++ b/SRC/zungbr.f
@@ -129,8 +129,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,min(M,N)).
 *>          For optimum performance LWORK >= min(M,N)*NB, where NB
 *>          is the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zunghr.f b/SRC/zunghr.f
index a09b7233..7a558501 100644
--- a/SRC/zunghr.f
+++ b/SRC/zunghr.f
@@ -58,8 +58,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          ILO and IHI must have the same values as in the previous call
 *>          of ZGEHRD. Q is equal to the unit matrix except in the
 *>          submatrix Q(ilo+1:ihi,ilo+1:ihi).
@@ -99,8 +98,7 @@
 *>          The dimension of the array WORK. LWORK >= IHI-ILO.
 *>          For optimum performance LWORK >= (IHI-ILO)*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zunglq.f b/SRC/zunglq.f
index 44e1f032..8b744975 100644
--- a/SRC/zunglq.f
+++ b/SRC/zunglq.f
@@ -99,8 +99,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,M).
 *>          For optimum performance LWORK >= M*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zungql.f b/SRC/zungql.f
index 2d2529b7..6e854bb4 100644
--- a/SRC/zungql.f
+++ b/SRC/zungql.f
@@ -100,8 +100,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is the
 *>          optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zungqr.f b/SRC/zungqr.f
index 76e40e72..a268fedf 100644
--- a/SRC/zungqr.f
+++ b/SRC/zungqr.f
@@ -100,8 +100,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,N).
 *>          For optimum performance LWORK >= N*NB, where NB is the
 *>          optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zungrq.f b/SRC/zungrq.f
index 8ad3a783..5e163bd4 100644
--- a/SRC/zungrq.f
+++ b/SRC/zungrq.f
@@ -100,8 +100,7 @@
 *>          The dimension of the array WORK. LWORK >= max(1,M).
 *>          For optimum performance LWORK >= M*NB, where NB is the
 *>          optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zungtr.f b/SRC/zungtr.f
index 29aedd16..049ec0b6 100644
--- a/SRC/zungtr.f
+++ b/SRC/zungtr.f
@@ -95,8 +95,7 @@
 *>          The dimension of the array WORK. LWORK >= N-1.
 *>          For optimum performance LWORK >= (N-1)*NB, where NB is
 *>          the optimal blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zunmbr.f b/SRC/zunmbr.f
index d4dd6cce..35e62246 100644
--- a/SRC/zunmbr.f
+++ b/SRC/zunmbr.f
@@ -167,8 +167,7 @@
 *>          For optimum performance LWORK >= max(1,N*NB) if SIDE = 'L',
 *>          and LWORK >= max(1,M*NB) if SIDE = 'R', where NB is the
 *>          optimal blocksize. (NB = 0 if M = 0 or N = 0.)
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zunmhr.f b/SRC/zunmhr.f
index af931394..77f77cf3 100644
--- a/SRC/zunmhr.f
+++ b/SRC/zunmhr.f
@@ -86,8 +86,7 @@
 *> \param[in] IHI
 *> \verbatim
 *>          IHI is INTEGER
-*> \endverbatim
-*> \verbatim
+*>
 *>          ILO and IHI must have the same values as in the previous call
 *>          of ZGEHRD. Q is equal to the unit matrix except in the
 *>          submatrix Q(ilo+1:ihi,ilo+1:ihi).
@@ -150,8 +149,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zunmlq.f b/SRC/zunmlq.f
index 01480168..c87e6bd3 100644
--- a/SRC/zunmlq.f
+++ b/SRC/zunmlq.f
@@ -141,8 +141,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zunmql.f b/SRC/zunmql.f
index e5ae1cdb..67365d78 100644
--- a/SRC/zunmql.f
+++ b/SRC/zunmql.f
@@ -141,8 +141,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zunmqr.f b/SRC/zunmqr.f
index d9c88a32..0dd18abc 100644
--- a/SRC/zunmqr.f
+++ b/SRC/zunmqr.f
@@ -141,8 +141,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zunmrq.f b/SRC/zunmrq.f
index 1ff05157..e0ac87b7 100644
--- a/SRC/zunmrq.f
+++ b/SRC/zunmrq.f
@@ -141,8 +141,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zunmrz.f b/SRC/zunmrz.f
index 4c62208f..c1366aae 100644
--- a/SRC/zunmrz.f
+++ b/SRC/zunmrz.f
@@ -149,8 +149,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >= M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error
diff --git a/SRC/zunmtr.f b/SRC/zunmtr.f
index cd806a5d..a8566fe7 100644
--- a/SRC/zunmtr.f
+++ b/SRC/zunmtr.f
@@ -142,8 +142,7 @@
 *>          For optimum performance LWORK >= N*NB if SIDE = 'L', and
 *>          LWORK >=M*NB if SIDE = 'R', where NB is the optimal
 *>          blocksize.
-*> \endverbatim
-*> \verbatim
+*>
 *>          If LWORK = -1, then a workspace query is assumed; the routine
 *>          only calculates the optimal size of the WORK array, returns
 *>          this value as the first entry of the WORK array, and no error