Issue metadata
Sign in to add a comment
|
24.7kb regression in resource_sizes (MonochromePublic.apk) at 479580:479583 |
||||||||||||||||||||
Issue descriptionCaused by “Revert "Upgrade gvr library to 1.60.1"” Commit: 2f7ee461cdd933ac5ed72f737b9ea65c5b19a28c Link to size graph: https://chromeperf.appspot.com/report?sid=a097e74b1aa288511afb4cb616efe0f95ba4d347ad61d5e835072f23450938ba&num_points=10&rev=479583 Debugging size regressions is documented at: https://chromium.googlesource.com/chromium/src/+/master/tools/perf/docs/apk_size_regressions.md#Debugging-Apk-Size-Increase Seems something got changed with proguard and it's not excluding correctly in VR land. This is pure dex count, so a large number of methods are no longer being correctly removed by proguard.
,
Jun 15 2017
Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8976687916788023680
,
Jun 15 2017
hmm. Interesting. I would expect my patch increased binary size instead of shrink. But anyway, I have put my patch in CQ again. It should be in TOT soon.
,
Jun 15 2017
=== BISECT JOB RESULTS ===
NO Perf regression found, tests failed to produce values
Bisect Details
Configuration: android_nexus7_perf_bisect
Benchmark : resource_sizes
Metric : MonochromePublic.apk_Specifics/normalized apk size
To Run This Test
src/build/android/resource_sizes.py --chromium-output-directory {CHROMIUM_OUTPUT_DIR} --chartjson {CHROMIUM_OUTPUT_DIR}/apks/MonochromePublic.apk
Debug Info
https://chromeperf.appspot.com/buildbucket_job_status/8976687916788023680
Is this bisect wrong?
https://chromeperf.appspot.com/bad_bisect?try_job_id=5439602501877760
| O O | Visit http://www.chromium.org/developers/speed-infra/perf-bug-faq
| X | for more information addressing perf regression bugs. For feedback,
| / \ | file a bug with component Speed>Bisection. Thank you!
,
Jun 15 2017
Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8976683478210497200
,
Jun 15 2017
=== BISECT JOB RESULTS ===
NO Perf regression found, tests failed to produce values
Bisect Details
Configuration: android_nexus7_perf_bisect
Benchmark : resource_sizes
Metric : MonochromePublic.apk_Specifics/normalized apk size
To Run This Test
src/build/android/resource_sizes.py --chromium-output-directory {CHROMIUM_OUTPUT_DIR} --chartjson {CHROMIUM_OUTPUT_DIR}/apks/MonochromePublic.apk
Debug Info
https://chromeperf.appspot.com/buildbucket_job_status/8976683478210497200
Is this bisect wrong?
https://chromeperf.appspot.com/bad_bisect?try_job_id=5439602501877760
| O O | Visit http://www.chromium.org/developers/speed-infra/perf-bug-faq
| X | for more information addressing perf regression bugs. For feedback,
| / \ | file a bug with component Speed>Bisection. Thank you!
,
Jun 16 2017
"regression" in this case just means "metric moved in the bad direction". It was an increase. I know you've already looked into minimizing the overhead from the first land, so closing. |
|||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||
Comment 1 by wnwen@chromium.org
, Jun 15 2017