diff options
| author | Tomeu Vizoso <tomeu.vizoso@collabora.com> | 2019-05-07 12:57:58 +0200 |
|---|---|---|
| committer | Tomeu Vizoso <tomeu.vizoso@collabora.com> | 2019-05-10 07:37:30 +0200 |
| commit | a94cf20051b66ee37149563bf967dbacec7ee0e5 (patch) | |
| tree | 8ea5ac93097ee9758b2e01c56f40c3eff29ddb9e | |
| parent | 0d0823638feb2fa7dc9fa9d5e6302c628c1a8805 (diff) | |
panfrost: ci: Don't restart Weston
There doesn't seem to actually be any noticeably memory leaks on Weston
when running dEQP. We do seem to leak quiet a bit in the client, so we
still have to run the dEQP runner in batches.
This removes the risk of Weston not restarting properly and introducing
spurious failures.
Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Reviewed-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
| -rw-r--r-- | src/gallium/drivers/panfrost/ci/deqp-runner.sh | 9 |
1 files changed, 1 insertions, 8 deletions
diff --git a/src/gallium/drivers/panfrost/ci/deqp-runner.sh b/src/gallium/drivers/panfrost/ci/deqp-runner.sh index 992560f8cd4..68d9bc1fc87 100644 --- a/src/gallium/drivers/panfrost/ci/deqp-runner.sh +++ b/src/gallium/drivers/panfrost/ci/deqp-runner.sh @@ -38,14 +38,7 @@ tail -f /tmp/result.txt & while [ -s /tmp/case-list.txt ]; do head -$BATCH_SIZE /tmp/case-list.txt > /tmp/next-batch.txt ./deqp-gles2 $DEQP_OPTIONS --deqp-log-filename=/dev/null --deqp-caselist-file=/tmp/next-batch.txt >> /tmp/result.txt - deqp_status=$? - - kill $(pidof weston) - sleep 1 # Give some time for Weston to release the VT - weston --tty=7 & - sleep 1 # Give some time for Weston to start up - - if [ $deqp_status -ne 0 ]; then + if [ $? -ne 0 ]; then # Continue from the subtest after the failing one crashed_test=$(grep "Test case" /tmp/result.txt | tail -1 | sed "s/Test case '\(.*\)'\.\./\1/") sed -i "0,/^$crashed_test$/d" /tmp/case-list.txt |