Age | Commit message (Collapse) | Author |
|
Signed-off-by: Naresh Kamboju <naresh.kamboju@linaro.org>
|
|
Signed-off-by: Naresh Kamboju <naresh.kamboju@linaro.org>
|
|
Fixes: https://bugs.linaro.org/show_bug.cgi?id=3852
Skips known LTP timing issue: select04 on qemu32 and qemu64.
Change-Id: Ie5a151763bb911dbce6d71cc5222e3a5497ebf7d
Signed-off-by: Rafael David Tinoco <rafael.tinoco@linaro.org>
|
|
|
|
skipping kselftest known failure rtnetlink.sh on all devices and all branches.
Ref:
https://bugs.linaro.org/show_bug.cgi?id=3834
Change-Id: I43683c3bebd775af542daf9096dff53a87492afc
Signed-off-by: Naresh Kamboju <naresh.kamboju@linaro.org>
|
|
Skip the known LTP CVE failures
- cve-2011-2183
- cve-2014-0196
- cve-2015-7550
- cve-2016-7117
Change-Id: I9a0b0e19fe866be06408b27b8f8e5cf7ab03fb6a
Signed-off-by: Naresh Kamboju <naresh.kamboju@linaro.org>
|
|
* Unify 'route' output with '-n' argument.
* On CentOS/Fedora, dhclient exits with non-zero when it is already
running. pkill dhclient before the following dhclient test.
* Save downloaded file to output directory.
Change-Id: I0f828486e4820f69a66ad08c95befe364184f83f
Signed-off-by: Chase Qi <chase.qi@linaro.org>
|
|
|
|
|
|
to determine if 'lava-test-case' exist.
Busybox doesn't have 'command'.
Change-Id: Ia808e29eca1467b909e27d84e14730c6204f8ec2
|
|
pritee_test_utility is Microsoft PlayReady Interface
for Trusted Execution Environment test utility
Change-Id: I2171db1bb538e59593d8e73413b575758f3d0d2d
|
|
bpf test case test_tunnel.sh is causing kernel crash on x86
running linux-next branch.
Change-Id: Icea556afdd0a1358a8917af5d25a2c580189167f
Signed-off-by: Naresh Kamboju <naresh.kamboju@linaro.org>
|
|
The final kill command in this test can often fail, eg. when
running on a BusyBox system.
Instead of exiting with an error, ignore it so the test results
get passed back to LAVA.
Change-Id: Ieeb9bfd73b7bf70a827a13a25b28f4f960124f21
Signed-off-by: Ryan Harkin <ryan.harkin@linaro.org>
|
|
|
|
Backport patch fixed fcntl35 and fcntl35_64 on 4.4 kernel.
so unskip these tests.
Re: Backport of pipe.c patch to 4.4.y that fixes LTP fcntl35 test error
https://www.spinics.net/lists/stable/msg239475.html
https://www.spinics.net/lists/stable/msg239479.html
Bug link:
https://bugs.linaro.org/show_bug.cgi?id=3378
Change-Id: I8692e731d7200c910aa82f9de01bd351e4c713f9
Signed-off-by: Naresh Kamboju <naresh.kamboju@linaro.org>
|
|
Remove git commit ID from concatenated name of the metric collected by
LISA.
Change-Id: Id5d936a3a14ae242d9944beda3bf8aab0f028627
Signed-off-by: Milosz Wasilewski <milosz.wasilewski@linaro.org>
|
|
Change-Id: Ic8890255e83e2fbd519e9130bdb804772508a7ef
Signed-off-by: Naresh Kamboju <naresh.kamboju@linaro.org>
|
|
kselftest net reuseport_bpf test case use to fail and now this been
fixed upstream. so unskip this test now.
Ref:
https://bugs.linaro.org/show_bug.cgi?id=3502
LKFT: linux-mainline: x86: kselftest NET reuseport_bpf failed
Change-Id: I068ebfa3df0143efd045f45578b79e807b5ebe7c
Signed-off-by: Naresh Kamboju <naresh.kamboju@linaro.org>
|
|
Removing skips from next. The skips for next that remains is if the test
hangs the target or crashes in a non recoverable way.
Change-Id: I8ec69211578bec3d9403112c852179399c93ea61
Signed-off-by: Anders Roxell <anders.roxell@linaro.org>
|
|
This may be resolved due to
http://lkml.iu.edu/hypermail/linux/kernel/1801.1/06039.html
Related bugs:
- https://bugs.linaro.org/show_bug.cgi?id=3297
- https://bugs.linaro.org/show_bug.cgi?id=3304
- https://bugs.linaro.org/show_bug.cgi?id=3612
Change-Id: Idf37892499692fe1a65b64488a29eff355047fa9
Signed-off-by: Dan Rue <dan.rue@linaro.org>
|
|
to work around the timed out problem when run cts opengl test
did not find the root cause yet, so set searchwindowsize to 1024
might not very accurate.
When tested on local side, it could help to finish CtsDeqpTestCases
module less than 90 minutes, so this should help on the opgengl timed
out problem on lava
Change-Id: I9507bc09e1336ed7f41514ba241d9e939ea074e4
Signed-off-by: Yongqin Liu <yongqin.liu@linaro.org>
|
|
Lava-dispatcher slows the execution when is too many output ([1] vs [2])
in the stdout so only redirect to a file in order to speed-up, same in
piglit related to rev 36daa3ece5.
[1] https://validation.linaro.org/scheduler/job/1776006
[2] https://validation.linaro.org/scheduler/job/1794400
Change-Id: I8aa0d43b43362765946f9e4b2c89a507ea223e69
Signed-off-by: Aníbal Limón <anibal.limon@linaro.org>
|
|
The piglit tool support regex exclusion (-x) to avoid run certain tests
but in order to exclude a high number of tests [1] isn't usable because
needs to be passed into the command line and there are limits in the
amount of data that you can pass using cmdline.
In my use case i want to exclude 212 failing tests [1].
[1]
https://qa-reports.linaro.org/qcomlt/openembedded-rpb-rocko/build/146/testrun/1787216/
Change-Id: I4b96e3c2540a4ba14bc086b5eb7856d67bc2b8de
Signed-off-by: Aníbal Limón <anibal.limon@linaro.org>
|
|
|
|
|
|
gfortran on CentOS 7 is provided in the gcc-gfortran package.
Change-Id: Ie7c5efd427ea1a9f4148697b1c57dbfcc4e39844
Signed-off-by: Chase Qi <chase.qi@linaro.org>
|
|
When docker not pre-installed, install docker-ce from upstream as it is
already available on ARM now. Refer to: https://get.docker.com/
Change-Id: If6b7d953e1770f847be8b9b85a56c66bf01d356b
Signed-off-by: Chase Qi <chase.qi@linaro.org>
|
|
There are some gst-validate tests that fails depending on the board
so add a way to use a file stored in a git repository to ignore the
result.
The gst-validate-launcher dosen't provide a command line option to
exclude certain tests by design depends on write your own testsuite.py
module so this is the better implementation for now.
Change-Id: I8e05de42892a58c1ccb344eb72932217de23921d
Signed-off-by: Aníbal Limón <anibal.limon@linaro.org>
|
|
|
|
The -e option works to exclude certain ptests for execution.
Change-Id: I644e45a8ab15ecc7712f04547cab3dcd61bb961f
Signed-off-by: Aníbal Limón <anibal.limon@linaro.org>
|
|
The scripts in this commit are meant to be used as a post-processing
step for results obtained with Workload Automation v3 (WA) running with
ARM Energy Proble (AEP) connected. LISA provides additional
measurements/metrics compared to pure WA results.
The script allows to use LISA from custom repository as well as use
custom branch/tag/commit from this source. lisa.sh executed python
script pointed to by LISA_SCRIPT variable. It is assumed that this
script can be executed in LISA environment. It is expected that
executing the script produces CSV files with the following structure:
<empty>,<empty>,<empty>,<metric_name>,<metric_name>,...
<empty>,<empty>,<empty>,<metric_1>,<metric_2>,...
test,tag,kernel,<empty>,<empty>,...
<test_name>,<tag>,<kernel>,<result_1>,<result2>,...
...
First 3 rows are concatenated to form single test name that is recorded in
LAVA or similar execution environment.
Change-Id: I78b52de1933ab99c0869d37ac8b1db561f5ec185
Signed-off-by: Milosz Wasilewski <milosz.wasilewski@linaro.org>
|
|
This commit adds support for installing and configuring ARM Energy Probe
(AEP) devices. These devices are connected to Device Under Test in order
to measure energy consumption at the connection point.
AEP is currently only suppored in WA scripts that execute on AOSP. There
is no support for devices running Linux.
Scripts allow to use custom repositories and branches for installation
and configuration.
Change-Id: Ieb9a485f1d269f108c1305c373d66e422136a2b0
Signed-off-by: Milosz Wasilewski <milosz.wasilewski@linaro.org>
|
|
|
|
Add a new test to run badblocks in read-only mode.
By default, the test will run on the block device where the
rootfs is mounted.
Callers can either set BADBLOCKS_DEVICE variable to test a different
device, or pass the "-d <device>" parameter into badblocks.sh
Additional parameters can also be passed by either setting the
BADBLOCKS_PARAMS variable or passing "-p <params>" to badblocks.sh.
For example, passing the additional parameter "-n" will ask badblocks to
run a read-write test. Note: this will fail is the block device is
mounted, eg. for the rootfs.
Change-Id: Ifa04d78cddb2070110cb7c6eb465f9364f281f6c
Signed-off-by: Ryan Harkin <ryan.harkin@linaro.org>
|
|
WA v2 and WA v3 use different variable names for installing
plugins/extensions. This enables using both v2 and v3 syntaxes for AEP.
Change-Id: I6e2408918baa1a773c71df9984b57c4e43ced63f
Signed-off-by: Milosz Wasilewski <milosz.wasilewski@linaro.org>
|
|
Change-Id: Id0bbfa3916d470178fde521752ee60c9836b7de3
Signed-off-by: Vishal Bhoj <vishal.bhoj@linaro.org>
|
|
Few jobs are submmitted with boot-<kernel_version>-<commit_id>.img
format.
Change-Id: Ibd9a22df83ed8b855731e424be8f3a8ed66d299f
Signed-off-by: Vishal Bhoj <vishal.bhoj@linaro.org>
|
|
The following known failures test cases skipped on qemu_arm.
- fork13
- msgctl10
- msgctl11
- pselect01
- pselect01_64
- quotactl01
- readahead02
Change-Id: I9c889d690c3d4d016953dfee298db86ad5fdb916
Signed-off-by: Naresh Kamboju <naresh.kamboju@linaro.org>
|
|
* Migrated from android/memtester.yaml
* Add memory size and test iteration params
* Rewrite test parser to catch all tests
Change-Id: Ie534e88ab125fe9f6c57f99afb68ad1d42932d42
Signed-off-by: Chase Qi <chase.qi@linaro.org>
|
|
LTP sched tests hackbench crashing the system so skip them.
Ref:
https://bugs.linaro.org/show_bug.cgi?id=3777
Change-Id: I90f52eb8f09c93d1b59a735a3610924b657a662b
Signed-off-by: Naresh Kamboju <naresh.kamboju@linaro.org>
|
|
Openembedded RPB images by default comes with busybox and no LSB support
so some tests like lsb_release and lscpu fails.
Change-Id: I48000a4fdb5c2ba5bf707f635e48b204345a7ca3
Signed-off-by: Aníbal Limón <anibal.limon@linaro.org>
|
|
WaRP7 has a V4L2 compliant camera driver. This test runs the compliance
tests against that driver.
Taken from the man page:
"The v4l2-compliance tool is used to test video4linux devices, either
video, vbi, radio or swradio, both input and output. It attempts to test
almost all aspects of a V4L2 device and it covers almost all V4L2
ioctls. It has very good support for video capture and output, VBI
capture and output and (software) radio tuning and transmitting."
"Note that v4l2-compliance not only tests for compliance against the
V4L2 API, but also whether the driver is using all the correct
frameworks. These frameworks often automatically provide ioctls that are
strictly speaking optional, but that come for free if you use those
frameworks. By requiring their presence the v4l2-compliance utility will
enforce their use."
Change-Id: Ide6de6b954eeb6edd1f19c9c49164d05582c772f
Signed-off-by: Ryan Harkin <ryan.harkin@linaro.org>
|
|
|
|
|
|
Skip the following LTP sched sub tests on qemu arm and arm64
- pth_str01
- pth_str02
- pth_str03
- time-schedule01
Change-Id: I65c30fe0ad02a9bcd5a91c47eb619c7dc20cc817
Signed-off-by: Naresh Kamboju <naresh.kamboju@linaro.org>
|
|
due to bugs in fastboot protocol in hikey UEFI,
we get error reported on host while running
fastboot boot boot.img due to which the script
exits which should not happen. The target reports
error but still boots the downloaded image.
as a result, at next reboot in the test the script
won't be running to provide boot.img
Here are such jobs:
https://lkft.validation.linaro.org/scheduler/job/194938#L6099
https://lkft.validation.linaro.org/scheduler/job/195058#L5872
Change-Id: I3f10973272ea793e5cc2d5ce118092837174ee0d
Signed-off-by: Vishal Bhoj <vishal.bhoj@linaro.org>
|
|
When kselftest sources upgrade to 4.16 we have seen new failure.
Those failure are recorded as bug and investigated in the meanwhile
these failed test cases would be skipped.
Change-Id: If51daf2693fb146e5ac52fdc6e7d4ae8329be349
Signed-off-by: Naresh Kamboju <naresh.kamboju@linaro.org>
|
|
|
|
|
|
Slow machines need more timeout. Default is 5 minutes and multiply
TIMEOUT_MULTIPLIER factor.
Change-Id: Ia26f8b7419ae509fdb939a3cb02a4f8f5947a25a
Signed-off-by: Chase Qi <chase.qi@linaro.org>
|