Use the real time inside the InspectorPerformanceAgent for *_duration |
||
Issue descriptionTo get correct metrics through Performance.getMetrics, we should use the real time inside the InspectorPerformanceAgent for *_duration and treat it as the cpu ticks.
,
May 4 2018
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/425817138ff95e438278b1d359c389b44ca998d1 commit 425817138ff95e438278b1d359c389b44ca998d1 Author: Marc Treib <treib@chromium.org> Date: Fri May 04 11:58:10 2018 Revert "Using non-override time in InspectorPerformanceAgent.cpp." This reverts commit 79b28587446e62bf42c9ee569d13c1463463d3c1. Reason for revert: performance-return-real-time.js fails on MSAN: https://ci.chromium.org/buildbot/chromium.webkit/WebKit%20Linux%20Trusty%20MSAN/ Original change's description: > Using non-override time in InspectorPerformanceAgent.cpp. > > Right now the InspectorPerformanceAgent returns incorrect performance metrics > (such as LayoutDuration, ScriptDuration, etc.) if there is an time override in > place. The InspectorPerformanceAgent needs to return real time elapsed. > > Per offline discussion with Pavel(pfeldman@), the reason we don't expose the > non-override time API in the web platform (<blink>/platform/time.h) is that time > override is to abstract the web platform from the time. The web platform should > just deal with time and not know about whether the time is overridden. Exposing > the non-override time API in the very core part of blink will leak the override > aspect and hence compromising the whole design > > BUG: 830033 > Change-Id: If3140892ac60bc853ef5f98f7b478a98ae332806 > Reviewed-on: https://chromium-review.googlesource.com/988833 > Reviewed-by: Kentaro Hara <haraken@chromium.org> > Reviewed-by: Pavel Feldman <pfeldman@chromium.org> > Commit-Queue: Johnny Ding <jnd@chromium.org> > Cr-Commit-Position: refs/heads/master@{#556012} TBR=dgozman@chromium.org,jnd@chromium.org,haraken@chromium.org,pfeldman@chromium.org,jochen@chromium.org Change-Id: Ic6f8feb1c77147e8b18fad12a95b1761cd13c2a0 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/1044145 Reviewed-by: Marc Treib <treib@chromium.org> Commit-Queue: Marc Treib <treib@chromium.org> Cr-Commit-Position: refs/heads/master@{#556027} [delete] https://crrev.com/0a3600c1bf36d80734992e8998447ac3edff997c/third_party/WebKit/LayoutTests/inspector-protocol/emulation/performance-return-real-time-expected.txt [delete] https://crrev.com/0a3600c1bf36d80734992e8998447ac3edff997c/third_party/WebKit/LayoutTests/inspector-protocol/emulation/performance-return-real-time.js [modify] https://crrev.com/425817138ff95e438278b1d359c389b44ca998d1/third_party/blink/renderer/core/inspector/DEPS [modify] https://crrev.com/425817138ff95e438278b1d359c389b44ca998d1/third_party/blink/renderer/core/inspector/inspector_performance_agent.cc [modify] https://crrev.com/425817138ff95e438278b1d359c389b44ca998d1/third_party/blink/renderer/core/inspector/inspector_performance_agent.h [modify] https://crrev.com/425817138ff95e438278b1d359c389b44ca998d1/third_party/blink/tools/audit_non_blink_usage.py
,
May 8 2018
The test script in inspector-protocol/emulation/performance-return-real-time.js is slow on MSAN mode. It takes less 0.1 second on normal mode, but may takes 6-7 seconds on MSAn mode. Will mark it TIMEOUT in MSANExpectations.
,
May 9 2018
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/d2ca7833e5f8992b1e8ccdfd164de0b7d37b67b2 commit d2ca7833e5f8992b1e8ccdfd164de0b7d37b67b2 Author: Johnny(Jianning) Ding <jnd@chromium.org> Date: Wed May 09 19:00:44 2018 Reland "Using non-override time in InspectorPerformanceAgent.cpp." This is a reland of 79b28587446e62bf42c9ee569d13c1463463d3c1 Original change's description: > Using non-override time in InspectorPerformanceAgent.cpp. > > Right now the InspectorPerformanceAgent returns incorrect performance metrics > (such as LayoutDuration, ScriptDuration, etc.) if there is an time override in > place. The InspectorPerformanceAgent needs to return real time elapsed. > > Per offline discussion with Pavel(pfeldman@), the reason we don't expose the > non-override time API in the web platform (<blink>/platform/time.h) is that time > override is to abstract the web platform from the time. The web platform should > just deal with time and not know about whether the time is overridden. Exposing > the non-override time API in the very core part of blink will leak the override > aspect and hence compromising the whole design > > BUG: 830033 > Change-Id: If3140892ac60bc853ef5f98f7b478a98ae332806 > Reviewed-on: https://chromium-review.googlesource.com/988833 > Reviewed-by: Kentaro Hara <haraken@chromium.org> > Reviewed-by: Pavel Feldman <pfeldman@chromium.org> > Commit-Queue: Johnny Ding <jnd@chromium.org> > Cr-Commit-Position: refs/heads/master@{#556012} Change-Id: I6f04a582d63376b3d6eeb80e27615c455dba4a61 Reviewed-on: https://chromium-review.googlesource.com/1045566 Commit-Queue: Johnny Ding <jnd@chromium.org> Reviewed-by: Pavel Feldman <pfeldman@chromium.org> Reviewed-by: Kentaro Hara <haraken@chromium.org> Cr-Commit-Position: refs/heads/master@{#557267} [add] https://crrev.com/d2ca7833e5f8992b1e8ccdfd164de0b7d37b67b2/third_party/WebKit/LayoutTests/inspector-protocol/emulation/performance-return-real-time-expected.txt [add] https://crrev.com/d2ca7833e5f8992b1e8ccdfd164de0b7d37b67b2/third_party/WebKit/LayoutTests/inspector-protocol/emulation/performance-return-real-time.js [modify] https://crrev.com/d2ca7833e5f8992b1e8ccdfd164de0b7d37b67b2/third_party/blink/renderer/core/inspector/DEPS [modify] https://crrev.com/d2ca7833e5f8992b1e8ccdfd164de0b7d37b67b2/third_party/blink/renderer/core/inspector/inspector_performance_agent.cc [modify] https://crrev.com/d2ca7833e5f8992b1e8ccdfd164de0b7d37b67b2/third_party/blink/renderer/core/inspector/inspector_performance_agent.h [modify] https://crrev.com/d2ca7833e5f8992b1e8ccdfd164de0b7d37b67b2/third_party/blink/tools/audit_non_blink_usage.py
,
May 30 2018
|
||
►
Sign in to add a comment |
||
Comment 1 by bugdroid1@chromium.org
, May 4 2018