fix: mtrace timestamp convertion #1342
Conversation
ekurdybx
requested review from
a team,
golowanow,
lgirdwood and
marc-hb
as code owners
ekurdybx
force-pushed
the
fix-mtrace-timestamp-convertion
branch
from
9c9eca3 to
d15d154
Compare
New OS has different timestamp format in mtrace file. This fix makes sof_per_analyzer.py handle both formats correctly. Signed-off-by: Emilia Kurdybelska <emiliax.kurdybelska@intel.com>
ekurdybx
force-pushed
the
fix-mtrace-timestamp-convertion
branch
from
d15d154 to
6aa732f
Compare
kv2019i
left a comment
kv2019i
left a comment
There was a problem hiding this comment.
Change looks good, but maybe add a reference to CONFIG_LOG_OUTPUT_FORMAT_LINUX_TIMESTAMP and CONFIG_LOG_OUTPUT_FORMAT_TIME_TIMESTAMP. This PR adds support for the latter.
| span_end_pos = match_obj.span()[1] | ||
| trace_lvl = line[span_end_pos - 4: span_end_pos - 1] | ||
| timestamp = float(line[span_end_pos - 19: span_end_pos - 7].strip()) | ||
| try: |
There was a problem hiding this comment.
The difference is caused by the Zephyr log timestamp format. It seems in some new targets, the time format is not set (default Zephyr format is used) and this breaks test. I now submitted a pull request to make the Linux format the default in Linux overlay thesofproject/sof#10501 .
I think it's still ok to have support for both widely used Zephyr timestamp formats in sof-test, but maybe a comment either in a comment or in the git commit would be good that the difference is caused by Zephyr build options and whether CONFIG_LOG_OUTPUT_FORMAT_LINUX_TIMESTAMP is set or not.
2 participants
New OS has different timestamp format in mtrace file. This fix makes sof_per_analyzer.py handle both formats correctly.