Age | Commit message (Collapse) | Author |
|
Signed-off-by: Rajagopal Venkat <rajagopal.venkat@linaro.org>
|
|
This patch adds stubs for features that are not supported
by Andriod. An header file which defines all stubs is
included only for Android builds.
Signed-off-by: Rajagopal Venkat <rajagopal.venkat@linaro.org>
|
|
This patch adds following minor changes to prepare powertop
to support Android platform.
- Add missing HAVE_CONFIG_H conditional check.
- remove un-used ethtool_cmd_speed_set and ethtool_cmd_speed
functions.
- Minimize dependency on exception handling in catch blocks.
These changes will not affect powertop functionality.
Signed-off-by: Rajagopal Venkat <rajagopal.venkat@linaro.org>
|
|
Remove unneded escaping backslashes from CSS.
|
|
|
|
abstract_cpu: do not call total_pstate_time() in validate()
Signed-off-by: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
|
|
This patch add more ui navigation features allowing for the scrolling of
PowerTop content for each tab. This is done by controlling the location
of the winpad, and refreshing the terminal window when needed. To do
this, the navigation keys have been updated as follows.
"Arrow Keys" now scroll the tab windows
"Page up/down" scroll tab windows up and down
"TAB" cycles the next window tab
"SHIFT+TAB" cycles the previous window tab
"ENTER" toggles tunables
"SPACE BAR" toggles tunables
"r KEY" refresh results and resets view of tab window.
Known issues:
There is no handler for terminal re-sizing. The workaround is to use "r
KEY" to refresh if terminal windows size is changed mid session.
|
|
|
|
|
|
Fix timer and work perf events timestamp tracing by
- considering event exit timestamp relative to measurement first
timestamp for events which entry timestamp is not recorded.
Currently these events exit timestamp itself is considered as
usage period resulting in false usage report.
- clearing event timestamps from global map at the end of each
measurement to avoid collision with earlier recorded timestamps.
Signed-off-by: Rajagopal Venkat <rajagopal.venkat@linaro.org>
|
|
|
|
|
|
entry for --quiet, which was missing.
Signed-off-by: Joerg Mayer <jmpt@loplof.de>
|
|
Signed-off-by: Joerg Mayer <jmpt@loplof.de>
|
|
Signed-off-by: Joerg Mayer <jmpt@loplof.de>
|
|
Signed-off-by: Joerg Mayer <jmpt@loplof.de>
|
|
Signed-off-by: Joerg Mayer <jmpt@loplof.de>
|
|
|
|
|
|
It is desired that PowerTOP not use ifdef's for the majority.
It is understood that if may just need to stub some code out,
but this should be a rare case. To clean up code there are
three defines identified
-- DISABLE_NCURSES
-- DISABLE_TRYCATCH
-- DISABLE_I18N
that are not really needed. They were put in the code originally
to quickly support android. After evaluation it has occurred to
us that there are better ways of dealing this distribution specific issues.
It is preferred for the project, that stubs be created by those how need
or care about such issues, and should be a separate stub.
|
|
|
|
|
|
|
|
|
|
paramerters.
Removed the global_fixed_paramerters as it is no longer a valid flag.
Also removed teh global_power_overide from main section that loads parameters, as it will
not be vaild at this point.
|
|
|
|
As of the 3.5 kernel, the Intel GPUs report their C states (power gating) via sysfs.
This patch will show them as part of the C state tab, arranged like a core inside package 0
(which matches physical topology)
Changes since v1: incorporated feedback from Sergey
|
|
In a hybrid system (CPU and GPU), it's common (well, for Intel graphics)
to not have statistics for the GPU P states, while we do have statistics
for the GPU C states.
This patch puts all the plumbing in place to allow for cores that
don't have P states, to not show up in the P state displays
|
|
the ARM guys complained that their human-readable C state names didn't have
numbers in them, and that as a result, the output is all messed up.
Using the "linux_name" instead is only a partial solution; it messes up the x86
side of the logic.
This patch fixes the logic to make the code use the human readable logic first,
but if there's no numbers there, fall back to the Linux name.
In addition, the patch allows callers to specify the line directly, overriding
both sets of logic.
|
|
AM_SILENT_RULES makes the output for "make" much more readable; it also
allows us to see warnings much more clearly.
|
|
Fix hrtimer event names by adding "timer:" prefix.
|
|
|
|
|
|
|
|
Change search order for libnl in configure.ac:
first try to use libnl-3, then libnl-2, then libnl-1.
|
|
|
|
|
|
This will allow me to keep update and align with the libary code until it is usable without
building it local.
|
|
|
|
|
|
in older kernels. sys/syscall.h will not be correct.
|
|
meet all needs as wanted.
|
|
disable pci if not supported by host platform. When
libpci is not found, define a macro which will be
used for conditional compilation.
Signed-off-by: Rajagopal Venkat <rajagopal.venkat@linaro.org>
|
|
pci you will not.
libpci has been made an optional lib, but for thoes who have pci, a warning has been added.
this warning is to let people know that PowerTOP was configured with pci support.
|
|
disable pci if not supported by host platform. When
libpci is not found, define a macro which will be
used for conditional compilation.
Signed-off-by: Rajagopal Venkat <rajagopal.venkat@linaro.org>
|
|
Get rid of TRUE and FALSE ncurses' defines used for boolean variables
initialization. This fix also helps build without ncurses support.
|
|
Remove unneded zlib dependency from configure.ac
|
|
Out-of-tree builds are a very useful feature which should be supported
by all programs which use GNU automake and the configure / make build.
For powertop, only a very small fix was needed to make it work.
Without that fix, powertop.css is not found when a build is not
started from the source tree, so csstoh fails and creates no css.h.
Signed-off-by: Stefan Weil <sw@weilnetz.de>
|
|
exit(0) is the same as returning EXIT_SUCCESS which is a bad
idea after a severe failure. It will for example result in
wrong behaviour of make (make won't stop after such failures).
Returning EXIT_FAILURE fixes this.
Signed-off-by: Stefan Weil <sw@weilnetz.de>
|
|
Not all architectures include lines with "vendor_id\t" or
"processor\t". ARM Linux without SMP support includes
neither of these two entries. With SMP support, there are
"processor\t" entries for each CPU.
Set vendor to an empty string initially, so there is a default value.
Assume CPU id 0 when no number was set while encountering the first
"bogomips\t" line. If there are more such lines without a correct
number, only the first CPU is handled. This should not be needed
for ARM, but makes the code more robust because it avoids calling
handle_one_cpu twice with number == 0.
Signed-off-by: Stefan Weil <sw@weilnetz.de>
|