Repository navigation
check-suspend-resume: check /sys/power/suspend_stats/*fail* at every step + dump much more /sys/ info - #900
Conversation
|
https://sof-ci.01.org/softestpr/PR900/build18/devicetest reveals that cc: @plbossart , @ujfalusi , @kv2019i , @ranj063 |
|
@marc-hb, and in some cases the Can you print the wakeup_count along with the files from |
I will "downgrade" this PR and print it instead of failing the test. |
marc-hb
changed the title
Done in v2 + dumping some additional info from /sys. The two failures in https://sof-ci.01.org/softestpr/PR900/build19/devicetest/ are known and unrelated. The
Other values seems fine. Please help review and merge this so we can start collecting more data - especially on systems that crash or timeout. |
| # set up checkpoint for each iteration | ||
| setup_kernel_check_point | ||
| sleep_count=$(cat /sys/power/wakeup_count) | ||
| : $((expected_wakeup_count++)) |
There was a problem hiding this comment.
I have no idea what this does.
using 'advanced sh' will restrict the circle of reviewers.
we need the equivalent of 'simple english' in wikipedia for shell scripts, something that everyone can understand.
There was a problem hiding this comment.
Replaced with expected_wakeup_count=$((expected_wakeup_count+1))
| { | ||
| printf '\n\n' | ||
|
|
||
| INITIAL_DBG_SOURCES="$LOG_ROOT"/intial_wakeup_sources.txt |
"greater or equal" has never ever reported any error. Test for the expected value instead but print only a warning for now because of the high number of failures this causes, example: https://sof-ci.01.org/softestpr/PR900/build18/devicetest/ Also catch any unexpected increase while the system is up - trust no one. Fix the error message to report exactly what was found, not a guess. Signed-off-by: Marc Herbert <marc.herbert@intel.com>
Signed-off-by: Marc Herbert <marc.herbert@intel.com>
As requested in thesofproject#898 Also dumps a lot more system information. Signed-off-by: Marc Herbert <marc.herbert@intel.com>
|
For https://sof-ci.01.org/softestpr/PR900/build20/devicetest/ CML_SKU0955_HDA was not available. https://sof-ci.01.org/softestpr/PR900/build20/devicetest/?model=WHL_UPEXT_HDA_ZEPHYR&testcase=check-suspend-resume-with-playback-5 timed out with more logs than usual which is exactly the purpose of this PR. |
| dlogi "Check for the kernel log status" | ||
| # check kernel log for each iteration to catch issues | ||
| sof-kernel-log-check.sh "$KERNEL_CHECKPOINT" || die "Caught error in kernel log" | ||
| sof-kernel-log-check.sh "$KERNEL_CHECKPOINT" || dump_and_die "Caught error in kernel log" |
There was a problem hiding this comment.
Should line 146 be split across two lines as well?
There was a problem hiding this comment.
93 characters so this is over the traditional 80 limit but below Linux' new, official 100 characters limit...
|
Thanks @greg-intel for the review. I will merge this tomorrow if there is no other review. |
Implements #898 and a lot more. 3 commits.