Issue metadata
Sign in to add a comment
|
1.9%-154.2% regression in loading.mobile at 483149:483555 |
||||||||||||||||||||
Issue descriptionSee the link to graphs below.
,
Jul 4 2017
Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8975010452454869664
,
Jul 4 2017
=== Auto-CCing suspected CL author reed@google.com === Hi reed@google.com, the bisect results pointed to your CL, please take a look at the results. === BISECT JOB RESULTS === Perf regression found with culprit Suspected Commit Author : Mike Reed Commit : 638b500582b0ffb30843452c6895f775be3b31dd Date : Thu Jun 29 16:15:07 2017 Subject: use rasterpipeline for non-N32 formats Bisect Details Configuration: android_one_perf_bisect Benchmark : smoothness.sync_scroll.key_mobile_sites_smooth Metric : mean_input_event_latency/http___cuteoverload.com Change : 326.54% | 49.8046666667 -> 212.439 Revision Result N chromium@483330 49.8047 +- 8.31555 6 good chromium@483354 54.8638 +- 19.0181 6 good chromium@483366 49.0682 +- 6.2751 6 good chromium@483372 49.8418 +- 4.5861 6 good chromium@483375 48.273 +- 4.79646 6 good chromium@483377 51.432 +- 8.07612 6 good chromium@483378 244.28 +- 319.94 6 bad <-- chromium@483426 212.439 +- 222.145 6 bad To Run This Test src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=http...cuteoverload.com smoothness.sync_scroll.key_mobile_sites_smooth More information on addressing performance regressions: http://g.co/ChromePerformanceRegressions Debug information about this bisect: https://chromeperf.appspot.com/buildbucket_job_status/8975010452454869664 For feedback, file a bug with component Speed>Bisection
,
Jul 27 2017
Explictly assigning. A CL you landed tripped one of the speed metrics we measure in the lab. If this is the first time this has happened to one of your CLs, or if it's been a while, please read: https://chromium.googlesource.com/chromium/src/+/master/docs/speed/addressing_performance_regressions.md We're looking for one of the following: 1. Justification via explanation 2. Plan to revert or fix 3. Angry rage throwing of equipment at my head Just be aware that I'm trained in trumpet playing and First Aid and am not afraid to use it. Note: This was a bulk edit message and not very personal.
,
Sep 15 2017
reed: any chance to investigate this?
,
Sep 18 2017
,
Sep 18 2017
,
Sep 18 2017
The change was part of a long series related to the 'raster pipeline' design - reduce code size - take advantage of advanced SIMD instructions (sometimes pre-compiled) - enable broader set of pixel formats : half-float, color-correct, RGBA/BGRA swizzles The particular flagged CL can not be reverted, as it is tied to a long series of related changes. What we can do next is, with guidance on relative priorities, profile specific tests and see what opportunities can be found to speed them up.
,
Jan 5 2018
reed: any update on the work in #8?
,
Apr 4 2018
This is way too old now to make sense of or do anything about directly. Please consider this a retroactive WontFix circa Sept. 2017. |
|||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||
Comment 1 by rmcilroy@chromium.org
, Jul 4 2017