Issue metadata
Sign in to add a comment
|
rasterize_and_record_micro.polymer.reference failure on chromium.perf |
||||||||||||||||||||
Issue descriptionRevision range first seen: 420969:421077 Link to failing step log: https://build.chromium.org/p/chromium.perf/builders/Android%20Nexus5%20Perf%20%282%29/builds/3947/steps/rasterize_and_record_micro.polymer.reference/logs/stdio Traceback (most recent call last): File "/b/rr/tmp4xYzAV/w/src/third_party/catapult/telemetry/telemetry/internal/story_runner.py", line 86, in _RunStoryAndProcessErrorIfNeeded state.RunStory(results) File "/b/rr/tmp4xYzAV/w/src/third_party/catapult/common/py_trace_event/py_trace_event/trace_event_impl/decorators.py", line 52, in traced_function return func(*args, **kwargs) File "/b/rr/tmp4xYzAV/w/src/third_party/catapult/telemetry/telemetry/page/shared_page_state.py", line 306, in RunStory self._current_page, self._current_tab, results) File "/b/rr/tmp4xYzAV/w/src/third_party/catapult/common/py_trace_event/py_trace_event/trace_event_impl/decorators.py", line 52, in traced_function return func(*args, **kwargs) File "/b/rr/tmp4xYzAV/w/src/tools/perf/measurements/rasterize_and_record_micro.py", line 89, in ValidateAndMeasurePage data['record_time_subsequence_caching_disabled_ms'] KeyError: 'record_time_subsequence_caching_disabled_ms' cc-ing authors from catapult roll in case anything stands out: 2016-09-26 eyaich Outputting chart json for disabled tests. 2016-09-26 alexandermont Add load and after_load time to power metrics. 2016-09-26 benjhayden Improve some docs, comments, and tests. 2016-09-26 benjhayden Remove font-size:12px from tr-ui-b-table. 2016-09-26 petrcermak [tracing] Use 'count' unit in memory-infra code 2016-09-26 petrcermak [tracing] Enable setting context group in MemoryColumn.fromRows
,
Sep 28 2016
Alas -- how can I check the Chrome version of the reference build?
,
Sep 28 2016
,
Sep 28 2016
It is more likely it is this change of mine: https://codereview.chromium.org/2352783004/ where I assumed, due to the age of the TODOs removed, that surely the reference Chrome version had the noted parameter present, but I must have underestimated the age of reference builds, at least for some benchmarks.
,
Sep 28 2016
cc-ing aiolos, who is working on updating the reference build for android. I have a CL in CQ to disable this until the ref build can be rolled: https://codereview.chromium.org/2376113002/
,
Sep 28 2016
Believe only the one line for subsequence caching is cause, as it was added more recently than the other TODOs, in this change from Jan 2016: https://codereview.chromium.org/1552693002 I will send a small change to restore that default-to-zero bit.
,
Sep 28 2016
https://codereview.chromium.org/2375213002/ sent for review.
,
Sep 29 2016
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/ff30596cef625e4438c6875f7273d5f323367389 commit ff30596cef625e4438c6875f7273d5f323367389 Author: wkorman <wkorman@chromium.org> Date: Thu Sep 29 00:31:21 2016 Restore default-to-zero for subsequence caching record time metric. http://crrev.com/1552693002 landed as r367465 and added record_time_subsequence_caching_disabled_ms. The current Android reference build is 48.0.2564.95 which is built at r359700. This change is a partial revert of http://crrev.com/2352783004. I've checked the other TODOs I removed and those look to be fine, with the next-newest being http://crrev.com/1104433005 as r326586 which should be present in the Android reference build. BUG= 651200 Review-Url: https://codereview.chromium.org/2375213002 Cr-Commit-Position: refs/heads/master@{#421694} [modify] https://crrev.com/ff30596cef625e4438c6875f7273d5f323367389/tools/perf/measurements/rasterize_and_record_micro.py
,
Sep 30 2016
I see a successful build for today: https://build.chromium.org/p/chromium.perf/builders/Android%20Nexus5%20Perf%20%282%29/builds/3973 so presuming this is fixed, please reopen if I'm off base. |
|||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||
Comment 1 by nedngu...@google.com
, Sep 28 2016