Add support for eglGetFrameTimestamps |
|||||||
Issue descriptionNow that eglGetFrameTimestamps is live in Android O, hook it up to LatencyInfo and add trace events. This will help debug latency issues, performance bottlenecks, and enable better metrics.
,
Sep 28 2017
,
Oct 5 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/f4356922d03f30a608b805df60f6f35dea0ee079 commit f4356922d03f30a608b805df60f6f35dea0ee079 Author: Brian Anderson <brianderson@chromium.org> Date: Thu Oct 05 22:16:19 2017 egl: Update khronos EGL headers. Copy latest khronos EGL headers to pick up eglGetFrameTimstampsANDROID. Also copy over existing chromium-specific edits. BUG=769415 Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel Change-Id: Ibd9b185d568c9e714e664e7380e15138f30fc3d2 Reviewed-on: https://chromium-review.googlesource.com/703076 Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Brian Anderson <brianderson@chromium.org> Cr-Commit-Position: refs/heads/master@{#506880} [modify] https://crrev.com/f4356922d03f30a608b805df60f6f35dea0ee079/gpu/ipc/service/image_transport_surface_android.cc [modify] https://crrev.com/f4356922d03f30a608b805df60f6f35dea0ee079/third_party/khronos/EGL/egl.h [modify] https://crrev.com/f4356922d03f30a608b805df60f6f35dea0ee079/third_party/khronos/EGL/eglext.h [modify] https://crrev.com/f4356922d03f30a608b805df60f6f35dea0ee079/third_party/khronos/EGL/eglplatform.h [modify] https://crrev.com/f4356922d03f30a608b805df60f6f35dea0ee079/third_party/khronos/README.chromium
,
Oct 6 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/1cf8f23156a1caf64158f3d5b10b7ff14226c1b5 commit 1cf8f23156a1caf64158f3d5b10b7ff14226c1b5 Author: Brian Anderson <brianderson@chromium.org> Date: Fri Oct 06 02:36:40 2017 gl: Add bindings for eglGetFrameTimestampsANDROID Bug: 769415 Test: Trace events that use this extension look good. Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel Change-Id: I3f051d4f86894f9029774e73b1327262c514fd53 Reviewed-on: https://chromium-review.googlesource.com/685402 Commit-Queue: Brian Anderson <brianderson@chromium.org> Reviewed-by: Kenneth Russell <kbr@chromium.org> Cr-Commit-Position: refs/heads/master@{#506954} [modify] https://crrev.com/1cf8f23156a1caf64158f3d5b10b7ff14226c1b5/ui/gl/generate_bindings.py [modify] https://crrev.com/1cf8f23156a1caf64158f3d5b10b7ff14226c1b5/ui/gl/gl_bindings_api_autogen_egl.h [modify] https://crrev.com/1cf8f23156a1caf64158f3d5b10b7ff14226c1b5/ui/gl/gl_bindings_autogen_egl.cc [modify] https://crrev.com/1cf8f23156a1caf64158f3d5b10b7ff14226c1b5/ui/gl/gl_bindings_autogen_egl.h
,
Oct 6 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/6c1f5306e33a3a9ca1b427e3341f73aba6b7d77e commit 6c1f5306e33a3a9ca1b427e3341f73aba6b7d77e Author: Brian Anderson <brianderson@chromium.org> Date: Fri Oct 06 02:47:48 2017 egl: Use khronos version of __APPLE__ section in eglplatform.h Bug: 769415 Change-Id: I8e16a2a35286144eaa2d97b4d5f7cf239a00a8e5 Reviewed-on: https://chromium-review.googlesource.com/703739 Reviewed-by: Corentin Wallez <cwallez@chromium.org> Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Brian Anderson <brianderson@chromium.org> Cr-Commit-Position: refs/heads/master@{#506957} [modify] https://crrev.com/6c1f5306e33a3a9ca1b427e3341f73aba6b7d77e/third_party/khronos/EGL/eglplatform.h
,
Oct 6 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/2f2c446254ddc743af534c08b342725887ec89b3 commit 2f2c446254ddc743af534c08b342725887ec89b3 Author: Hiroki Nakagawa <nhiroki@chromium.org> Date: Fri Oct 06 03:16:27 2017 Revert "egl: Use khronos version of __APPLE__ section in eglplatform.h" This reverts commit 6c1f5306e33a3a9ca1b427e3341f73aba6b7d77e. Reason for revert: Speculative revert for build failures on the Mac bot https://build.chromium.org/p/chromium/builders/Mac/builds/33456 Original change's description: > egl: Use khronos version of __APPLE__ section in eglplatform.h > > Bug: 769415 > Change-Id: I8e16a2a35286144eaa2d97b4d5f7cf239a00a8e5 > Reviewed-on: https://chromium-review.googlesource.com/703739 > Reviewed-by: Corentin Wallez <cwallez@chromium.org> > Reviewed-by: Kenneth Russell <kbr@chromium.org> > Commit-Queue: Brian Anderson <brianderson@chromium.org> > Cr-Commit-Position: refs/heads/master@{#506957} TBR=brianderson@chromium.org,kbr@chromium.org,cwallez@chromium.org Change-Id: Ia96b247a9080be24a8c53dca6a7033aed43eb40f No-Presubmit: true No-Tree-Checks: true No-Try: true Bug: 769415 Reviewed-on: https://chromium-review.googlesource.com/704334 Reviewed-by: Hiroki Nakagawa <nhiroki@chromium.org> Commit-Queue: Hiroki Nakagawa <nhiroki@chromium.org> Cr-Commit-Position: refs/heads/master@{#506958} [modify] https://crrev.com/2f2c446254ddc743af534c08b342725887ec89b3/third_party/khronos/EGL/eglplatform.h
,
Oct 6 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/822c07c6ef35e5b1cbb36fdabf5fe9a276c49aba commit 822c07c6ef35e5b1cbb36fdabf5fe9a276c49aba Author: Brian Anderson <brianderson@chromium.org> Date: Fri Oct 06 18:39:05 2017 gl: Add TraceEvents for eglGetFrameTimestamps. Sorts the events by timestamp and records: 1) A trace event for the entire range. 2) Nested ranges within (1) of the remaining events, which provide a nice visual representation of a frames events and durations, making anomalous frames very apparent. 3) Nested instant events within (1) for a more detailed description of the event. Bug: 769415 Test: Traces look niice. Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel Change-Id: Iaff6867f7f9eea48d3a456264e36978da9991abb Reviewed-on: https://chromium-review.googlesource.com/685409 Commit-Queue: Brian Anderson <brianderson@chromium.org> Reviewed-by: oysteine <oysteine@chromium.org> Reviewed-by: Kenneth Russell <kbr@chromium.org> Cr-Commit-Position: refs/heads/master@{#507125} [modify] https://crrev.com/822c07c6ef35e5b1cbb36fdabf5fe9a276c49aba/base/trace_event/common/trace_event_common.h [modify] https://crrev.com/822c07c6ef35e5b1cbb36fdabf5fe9a276c49aba/ui/gl/gl_surface_egl.cc [modify] https://crrev.com/822c07c6ef35e5b1cbb36fdabf5fe9a276c49aba/ui/gl/gl_surface_egl.h
,
Oct 9 2017
Change r507125 seems to break WebVR with a GPU process crash, and reverting that change fixes it. How to reproduce: - open https://webvr.info/samples/03-vr-presentation.html on Android - press "Enter VR" button" 10-09 11:06:37.135 9347 9369 E libEGL : eglGetFrameTimestampsANDROID:2371 error 3002 (EGL_BAD_ACCESS) 10-09 11:06:37.136 9347 9369 F libc : stack corruption detected (-fstack-protector) 10-09 11:06:37.136 9347 9369 F libc : Fatal signal 6 (SIGABRT), code -6 in tid 9369 (CrGpuMain), pid 9347 (ileged_process0) 10-09 11:06:37.158 9752 9785 W YouTube : ECatcher disabled: level: ERROR, category: logging, message: Failed delayed event dispatch, no dispatchers. 10-09 11:06:37.158 9752 9785 W YouTube : java.lang.Exception 10-09 11:06:37.158 9752 9785 W YouTube : at xer.a(SourceFile:3) 10-09 11:06:37.158 9752 9785 W YouTube : at xcj.c(SourceFile:97) 10-09 11:06:37.158 9752 9785 W YouTube : at xcr.run(Unknown Source:2) 10-09 11:06:37.158 9752 9785 W YouTube : at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:457) 10-09 11:06:37.158 9752 9785 W YouTube : at java.util.concurrent.FutureTask.run(FutureTask.java:266) 10-09 11:06:37.158 9752 9785 W YouTube : at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:301) 10-09 11:06:37.158 9752 9785 W YouTube : at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1162) 10-09 11:06:37.158 9752 9785 W YouTube : at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:636) 10-09 11:06:37.158 9752 9785 W YouTube : at qux.run(SourceFile:24) 10-09 11:06:37.158 9752 9785 W YouTube : at java.lang.Thread.run(Thread.java:764) 10-09 11:06:37.230 9825 9825 I crash_dump32: obtaining output fd from tombstoned, type: kDebuggerdTombstone 10-09 11:06:37.231 743 743 I /system/bin/tombstoned: received crash request for pid 9347 10-09 11:06:37.238 9825 9825 I crash_dump32: performing dump of process 9347 (target tid = 9369) 10-09 11:06:37.238 9825 9825 F DEBUG : *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** 10-09 11:06:37.238 9825 9825 F DEBUG : Build fingerprint: 'google/marlin/marlin:8.1.0/OPM1.170928.003/4367109:userdebug/dev-keys' 10-09 11:06:37.238 9825 9825 F DEBUG : Revision: '0' 10-09 11:06:37.239 9825 9825 F DEBUG : ABI: 'arm' 10-09 11:06:37.239 9825 9825 F DEBUG : pid: 9347, tid: 9369, name: CrGpuMain >>> org.chromium.chrome:privileged_process0 <<< 10-09 11:06:37.239 9825 9825 F DEBUG : signal 6 (SIGABRT), code -6 (SI_TKILL), fault addr -------- 10-09 11:06:37.247 9825 9825 F DEBUG : Abort message: 'stack corruption detected (-fstack-protector)' 10-09 11:06:37.247 9825 9825 F DEBUG : r0 00000000 r1 00002499 r2 00000006 r3 00000008 10-09 11:06:37.247 9825 9825 F DEBUG : r4 00002483 r5 00002499 r6 c9e77e3c r7 0000010c 10-09 11:06:37.247 9825 9825 F DEBUG : r8 00000000 r9 dc233c40 sl 00000000 fp ffffffff 10-09 11:06:37.247 9825 9825 F DEBUG : ip c74d43df sp c9e77e28 lr e523fbdf pc e52395ec cpsr 20070030 10-09 11:06:37.252 9825 9825 F DEBUG : 10-09 11:06:37.252 9825 9825 F DEBUG : backtrace: 10-09 11:06:37.253 9825 9825 F DEBUG : #00 pc 0001a5ec /system/lib/libc.so (abort+63) 10-09 11:06:37.253 9825 9825 F DEBUG : #01 pc 00049269 /system/lib/libc.so (__stack_chk_fail+8) 10-09 11:06:37.253 9825 9825 F DEBUG : #02 pc 0099caeb /data/app/org.chromium.chrome-5eGKbYgsE0idX9RehfmE8A==/base.apk (offset 0xfe3000) 10-09 11:06:37.253 9825 9825 F DEBUG : #03 pc fffffffd <unknown> Translated to: Stack Trace: RELADDR FUNCTION FILE:LINE 0001a5ec abort+63 /system/lib/libc.so 00049269 __stack_chk_fail+8 /system/lib/libc.so 0099cafb gl::NativeViewGLSurfaceEGL::TraceSwapEvents(unsigned long long) /usr/local/google/home/klausw/chromium2/src/ui/gl/gl_surface_egl.cc:1138 00008b86 <unknown>
,
Oct 9 2017
Re comment #8: I have what I think is a fix for that here: https://chromium-review.googlesource.com/c/chromium/src/+/706546
,
Oct 9 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/0e7fb2f9e40213ddc37e5ee457399159eec8903a commit 0e7fb2f9e40213ddc37e5ee457399159eec8903a Author: Brian Anderson <brianderson@chromium.org> Date: Mon Oct 09 22:48:28 2017 gl: Protect against unexpected calls to NativeViewGLSurfaceEGL::Initialize. Extra calls to Initialize() would append timestamps supported by eglGetFrameTimestampsANDROID without resetting the original vector, which would result in stack overflow in TraceSwapEvents once the size of the supported vector is greater than the number of supportable timestamps. This patch resets the vectors in Initialize() to avoid the double counting. It also clamps the supported count in TraceSwapEvents just in case. Bug: 772743, 769415 TBR: kbr@chromium.org Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel Change-Id: I944b506c3022c4f6b373685dab6acf0d1ce87af0 Reviewed-on: https://chromium-review.googlesource.com/706546 Reviewed-by: Brian Anderson <brianderson@chromium.org> Commit-Queue: Brian Anderson <brianderson@chromium.org> Cr-Commit-Position: refs/heads/master@{#507515} [modify] https://crrev.com/0e7fb2f9e40213ddc37e5ee457399159eec8903a/ui/gl/gl_surface_egl.cc
,
Oct 9 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/17e5bbe6b2bf33a075db985bce9dcae991fd99e5 commit 17e5bbe6b2bf33a075db985bce9dcae991fd99e5 Author: Brian Anderson <brianderson@chromium.org> Date: Mon Oct 09 23:13:17 2017 egl: Use official khronos version of eglext.h The remaining chromium-specific extension definitions aren't actually used from this header. Bug: 769415 Test: Compiles Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel Change-Id: I097c93e456e6e7c8b89f15dab61aed281e7ab8db Reviewed-on: https://chromium-review.googlesource.com/703734 Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Brian Anderson <brianderson@chromium.org> Cr-Commit-Position: refs/heads/master@{#507524} [modify] https://crrev.com/17e5bbe6b2bf33a075db985bce9dcae991fd99e5/third_party/khronos/EGL/eglext.h
,
Oct 9 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/3fe13ca8f244535aee53771e6e8545810e9ac5e3 commit 3fe13ca8f244535aee53771e6e8545810e9ac5e3 Author: Brian Anderson <brianderson@chromium.org> Date: Mon Oct 09 23:17:37 2017 egl: Use khronos version of __APPLE__ section in eglplatform.h Bug: 769415 Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel Change-Id: I5586b1ff0405fddbf950fe3d2ca877594ea39d95 Reviewed-on: https://chromium-review.googlesource.com/705877 Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Brian Anderson <brianderson@chromium.org> Cr-Commit-Position: refs/heads/master@{#507526} [modify] https://crrev.com/3fe13ca8f244535aee53771e6e8545810e9ac5e3/gpu/gles2_conform_support/egl/display.cc [modify] https://crrev.com/3fe13ca8f244535aee53771e6e8545810e9ac5e3/third_party/khronos/EGL/eglplatform.h
,
Oct 9 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/c5dff3b05d1cfa98e705130d2b037b6ef696ce6e commit c5dff3b05d1cfa98e705130d2b037b6ef696ce6e Author: Vincent Scheib <scheib@chromium.org> Date: Mon Oct 09 23:35:09 2017 Revert "egl: Use official khronos version of eglext.h" This reverts commit 17e5bbe6b2bf33a075db985bce9dcae991fd99e5. Reason for revert: Compile failure FAILED: obj/ui/gl/gl/gl_image_dxgi.obj ninja -t msvc -e environment.x64 -- C:\b\c\goma_client/gomacc.exe "c:\b\c\win_toolchain\vs_files\9bc7ccbf9f4bd50d4a3bd185e8ca94ff1618de0b\vc\tools\msvc\14.11.25503\bin\hostx64\x64/cl.exe" /nologo /showIncludes @obj/ui/gl/gl/gl_image_dxgi.obj.rsp /c ../../ui/gl/gl_image_dxgi.cc /Foobj/ui/gl/gl/gl_image_dxgi.obj /Fd"obj/ui/gl/gl_cc.pdb" ../../ui/gl/gl_image_dxgi.cc(116): error C2065: 'EGL_D3D_TEXTURE_SUBRESOURCE_ID_ANGLE': undeclared identifier on build: https://build.chromium.org/p/chromium.win/builders/Win%20x64%20Builder%20%28dbg%29/builds/58992 Original change's description: > egl: Use official khronos version of eglext.h > > The remaining chromium-specific extension definitions aren't > actually used from this header. > > Bug: 769415 > Test: Compiles > Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel > Change-Id: I097c93e456e6e7c8b89f15dab61aed281e7ab8db > Reviewed-on: https://chromium-review.googlesource.com/703734 > Reviewed-by: Kenneth Russell <kbr@chromium.org> > Commit-Queue: Brian Anderson <brianderson@chromium.org> > Cr-Commit-Position: refs/heads/master@{#507524} TBR=brianderson@chromium.org,kbr@chromium.org Change-Id: I14baa9ee558aa92b057325cb4ec5d95a69575655 No-Presubmit: true No-Tree-Checks: true No-Try: true Bug: 769415 Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel Reviewed-on: https://chromium-review.googlesource.com/708054 Reviewed-by: Vincent Scheib <scheib@chromium.org> Commit-Queue: Vincent Scheib <scheib@chromium.org> Cr-Commit-Position: refs/heads/master@{#507535} [modify] https://crrev.com/c5dff3b05d1cfa98e705130d2b037b6ef696ce6e/third_party/khronos/EGL/eglext.h
,
Nov 2 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/5971d3d542606026db209ed86dd40568d36e81b8 commit 5971d3d542606026db209ed86dd40568d36e81b8 Author: Brian Anderson <brianderson@chromium.org> Date: Thu Nov 02 18:02:40 2017 gl: Work around sync_merge kernel bug with eglGetFrameTimestamps. Discard timestamps of 0. Although technically valid, they are unexpected in practice and are likely a result of a bug in newer versions of the linux kernel. Test: Trace events on the Pixel 2 don't start at 0. Bug: 769415 Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel Change-Id: Ie20b8cf82f7a3b634420460b4d4ec66bc69a5e5d Reviewed-on: https://chromium-review.googlesource.com/750201 Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Brian Anderson <brianderson@chromium.org> Cr-Commit-Position: refs/heads/master@{#513549} [modify] https://crrev.com/5971d3d542606026db209ed86dd40568d36e81b8/ui/gl/gl_surface_egl.cc
,
Nov 3 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/0f5a741df316bfe38448177b647b4ac7d934e76c commit 0f5a741df316bfe38448177b647b4ac7d934e76c Author: Brian Anderson <brianderson@chromium.org> Date: Fri Nov 03 00:58:51 2017 gl: Only use eglGetFrametimestamps when tracing. The existing code attepts to do this but failed to dereference a pointer, causing the tracing logic to always be executed since the pointer was always not NULL. Bug: 769415 Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel Change-Id: I497b87c51d4e0012213d7f1231693dc5d04cf38c Reviewed-on: https://chromium-review.googlesource.com/748029 Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Brian Anderson <brianderson@chromium.org> Cr-Commit-Position: refs/heads/master@{#513640} [modify] https://crrev.com/0f5a741df316bfe38448177b647b4ac7d934e76c/ui/gl/gl_surface_egl.cc
,
Nov 4 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/776b28a0b3efd650371a23949e1e08a097ceda31 commit 776b28a0b3efd650371a23949e1e08a097ceda31 Author: Brian Anderson <brianderson@chromium.org> Date: Sat Nov 04 00:35:21 2017 gl: Make eglGetFrameTimestamps opt-in. eglGetFrameTimestamps can't be used with WebVR surfaces since they aren't consumed by SurfaceFlinger. Attempting to enable it for WebVR surfaces results in logcat spam. There isn't an easy way to detect WebVR surfaces service-side, so this patch requires some client-side changes. Client side: Adds a |enable_swap_timestamps| surface attribute that is only set for the default onscreen surfaces. Service side: Adds SupportsSwapTimestamps and SetEnableSwapTimestamps to GLSurface, which are properly overriden by NativeViewGLSurfaceEGL based on eglGetFrameTimestamps support. Test: No more logcat spam with WebVR. Bug: 769415 Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel Change-Id: Idb7314aceb2e84e8b886d881dc262fb1b9c7e1ba Reviewed-on: https://chromium-review.googlesource.com/748106 Reviewed-by: Daniel Cheng <dcheng@chromium.org> Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Brian Anderson <brianderson@chromium.org> Cr-Commit-Position: refs/heads/master@{#513992} [modify] https://crrev.com/776b28a0b3efd650371a23949e1e08a097ceda31/content/browser/renderer_host/compositor_impl_android.cc [modify] https://crrev.com/776b28a0b3efd650371a23949e1e08a097ceda31/gpu/command_buffer/common/gles2_cmd_utils.h [modify] https://crrev.com/776b28a0b3efd650371a23949e1e08a097ceda31/gpu/ipc/common/gpu_command_buffer_traits_multi.h [modify] https://crrev.com/776b28a0b3efd650371a23949e1e08a097ceda31/gpu/ipc/service/gpu_command_buffer_stub.cc [modify] https://crrev.com/776b28a0b3efd650371a23949e1e08a097ceda31/ui/gl/gl_surface.cc [modify] https://crrev.com/776b28a0b3efd650371a23949e1e08a097ceda31/ui/gl/gl_surface.h [modify] https://crrev.com/776b28a0b3efd650371a23949e1e08a097ceda31/ui/gl/gl_surface_egl.cc [modify] https://crrev.com/776b28a0b3efd650371a23949e1e08a097ceda31/ui/gl/gl_surface_egl.h
,
Nov 4 2017
In two builds on the new Quadro P400 bots after this landed: https://ci.chromium.org/buildbot/chromium.gpu.fyi/Win7%20Release%20%28NVIDIA%20Quadro%20P400%29/151 https://chromium-swarm.appspot.com/task?id=399da18f782ed110&refresh=10&show_raw=1 and: https://ci.chromium.org/buildbot/chromium.gpu.fyi/Win7%20Release%20%28NVIDIA%29/31981 https://chromium-swarm.appspot.com/task?id=399deea2efdc8d10&refresh=10&show_raw=1 WebglConformance_conformance2_textures_canvas_sub_rectangle_tex_2d_rgb565_rgb_unsigned_byte started failing in the webgl2_conformance_gl_tests suite. The crash stack in the GPU process is suspicious and involves files touched by the CL above, even though ANGLE doesn't support the extension. It's identical between the two builds: Last event: a14.1140: Break instruction exception - code 80000003 (first/second chance not available) debugger time: Fri Nov 3 20:41:01.598 2017 (UTC - 7:00) ChildEBP RetAddr Args to Child 002fe84c 67689487 6969e020 00000114 02866b81 chrome_child!base::debug::BreakDebugger+0xc 002fe874 667670be 02821820 002fe8c0 002fe8c4 chrome_child!?Run@?$Invoker@U?$BindState@P6AXPBDHV?$BasicStringPiece@V?$basic_string@DU?$char_traits@D@std@@V?$allocator@D@2@@std@@@base@@1@Z$$V@internal@base@@$$A6AXPBDHV?$BasicStringPiece@V?$basic_string@DU?$char_traits@D@std@@V?$allocator@D@2@@std@@@3@1@Z@internal@base@@SAXPAVBindStateBase@23@$$QAPBD$$QAH$$QAV?$BasicStringPiece@V?$basic_string@DU?$char_traits@D@std@@V?$allocator@D@2@@std@@@3@3@Z+0x25 002fed08 673155fa 00000003 002fed20 69205174 chrome_child!logging::LogMessage::~LogMessage+0x41e 002fedd4 6731528d 028698d8 002fedec 667f076b chrome_child!gl::GLContextEGL::IsCurrent+0xa8 002feeec 6840919c 028698d8 02867ffc 00736290 chrome_child!gl::GLContextEGL::MakeCurrent+0x61 002fef30 6840e429 02867f68 0869c9f0 002fef4c chrome_child!gpu::GpuCommandBufferStub::Destroy+0x128 002fef40 6840de37 0284e630 002fef68 68408e4a chrome_child!gpu::GpuCommandBufferStub::~GpuCommandBufferStub+0x47 002fef4c 68408e4a 00000001 0284e644 0284e630 chrome_child!gpu::GpuCommandBufferStub::`scalar deleting destructor'+0xb 002fef68 6840820d 0284d07c 002fef8c 675a3982 chrome_child!gpu::GpuChannel::~GpuChannel+0x34 002fef74 675a3982 00000001 0284d078 02855f10 chrome_child!gpu::GpuChannel::`scalar deleting destructor'+0xb 002fef8c 68402e2a 002fefa0 420b58a2 0284d078 chrome_child!std::list<std::pair<int const ,std::unique_ptr<gpu::GpuChannel,std::default_delete<gpu::GpuChannel> > >,std::allocator<std::pair<int const ,std::unique_ptr<gpu::GpuChannel,std::default_delete<gpu::GpuChannel> > > > >::clear+0x2e 002fefa4 68402cc9 002fefc0 02855f10 0284d078 chrome_child!std::_Hash<std::_Umap_traits<int,std::unique_ptr<gpu::GpuChannel,std::default_delete<gpu::GpuChannel> >,std::_Uhash_compare<int,std::hash<int>,std::equal_to<int> >,std::allocator<std::pair<int const ,std::unique_ptr<gpu::GpuChannel,std::default_delete<gpu::GpuChannel> > > >,0> >::clear+0x1a 002fefb8 67553dc7 028570b8 002fefb8 002fefb8 chrome_child!gpu::GpuChannelManager::~GpuChannelManager+0xd 002ff0a8 6754f55b 0283f21c 002ff188 6667494b chrome_child!viz::GpuServiceImpl::~GpuServiceImpl+0x1f7 002ff0b4 6667494b 00000001 7769715c 76a2865d chrome_child!viz::GpuServiceImpl::`scalar deleting destructor'+0xb 002ff188 67d7704c 02856620 420b58a2 0283f100 chrome_child!viz::VizMainImpl::~VizMainImpl+0x115 002ff1a0 67d778d7 002ff3b0 002ff280 676ea436 chrome_child!content::GpuChildThread::~GpuChildThread+0xec 002ff1ac 676ea436 00000001 69fa7f90 6678a212 chrome_child!content::GpuChildThread::`scalar deleting destructor'+0xb 002ff280 67d764af 02835940 01400284 027e2538 chrome_child!content::ChildProcess::~ChildProcess+0xac 002ff474 6675ad7a 002ff560 027e2d20 00000000 chrome_child!content::GpuMain+0x38f 002ff544 6675b2e2 002ff578 002ff560 002ff7cc chrome_child!content::RunNamedProcessTypeMain+0x10a 002ff640 66760cd3 002ff590 00000003 002ff760 chrome_child!content::ContentMainRunnerImpl::Run+0x116 002ff750 6675ac49 002ff75c 002ff760 693f34e0 chrome_child!service_manager::Main+0x2a3 002ff790 65831120 002ff7b4 002ff79c 002ff798 chrome_child!content::ContentMain+0x31 Rather than suppress this one test failure I'm going to do a speculative revert of the patch above. If it clears up the failure then it seems something's wrong with it. If not I'll reland it. (The tryservers have been switched to the P400 bots and so if there is a real persistent failure it will show up in the tryjobs the next time.)
,
Nov 4 2017
I was wrong, since the failure was in webgl2_conformance_gl_tests and not webgl2_conformance_tests, this isn't tested on the win_optional_gpu_tests_rel so wouldn't show up on the tryservers. (We don't have the capacity to run the WebGL 2.0 tests against CLs in multiple configurations.)
,
Nov 4 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/1d6cfdd22a5890b0d1bf429d2e9b04a0b39a8507 commit 1d6cfdd22a5890b0d1bf429d2e9b04a0b39a8507 Author: Kenneth Russell <kbr@chromium.org> Date: Sat Nov 04 06:15:38 2017 Revert "gl: Make eglGetFrameTimestamps opt-in." This reverts commit 776b28a0b3efd650371a23949e1e08a097ceda31. Reason for revert: speculative revert for GPU process crash: see https://bugs.chromium.org/p/chromium/issues/detail?id=769415#c17 . Original change's description: > gl: Make eglGetFrameTimestamps opt-in. > > eglGetFrameTimestamps can't be used with WebVR surfaces since they > aren't consumed by SurfaceFlinger. Attempting to enable it for WebVR > surfaces results in logcat spam. > > There isn't an easy way to detect WebVR surfaces service-side, so > this patch requires some client-side changes. > > Client side: > Adds a |enable_swap_timestamps| surface attribute that is only set > for the default onscreen surfaces. > > Service side: > Adds SupportsSwapTimestamps and SetEnableSwapTimestamps to GLSurface, > which are properly overriden by NativeViewGLSurfaceEGL based on > eglGetFrameTimestamps support. > > Test: No more logcat spam with WebVR. > Bug: 769415 > Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel > Change-Id: Idb7314aceb2e84e8b886d881dc262fb1b9c7e1ba > Reviewed-on: https://chromium-review.googlesource.com/748106 > Reviewed-by: Daniel Cheng <dcheng@chromium.org> > Reviewed-by: Kenneth Russell <kbr@chromium.org> > Commit-Queue: Brian Anderson <brianderson@chromium.org> > Cr-Commit-Position: refs/heads/master@{#513992} TBR=dcheng@chromium.org,brianderson@chromium.org,kbr@chromium.org,klausw@chromium.org Change-Id: I992667635217406a77e820a52aee6d3066d20d2c No-Presubmit: true No-Tree-Checks: true No-Try: true Bug: 769415 Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel Reviewed-on: https://chromium-review.googlesource.com/754127 Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Kenneth Russell <kbr@chromium.org> Cr-Commit-Position: refs/heads/master@{#514034} [modify] https://crrev.com/1d6cfdd22a5890b0d1bf429d2e9b04a0b39a8507/content/browser/renderer_host/compositor_impl_android.cc [modify] https://crrev.com/1d6cfdd22a5890b0d1bf429d2e9b04a0b39a8507/gpu/command_buffer/common/gles2_cmd_utils.h [modify] https://crrev.com/1d6cfdd22a5890b0d1bf429d2e9b04a0b39a8507/gpu/ipc/common/gpu_command_buffer_traits_multi.h [modify] https://crrev.com/1d6cfdd22a5890b0d1bf429d2e9b04a0b39a8507/gpu/ipc/service/gpu_command_buffer_stub.cc [modify] https://crrev.com/1d6cfdd22a5890b0d1bf429d2e9b04a0b39a8507/ui/gl/gl_surface.cc [modify] https://crrev.com/1d6cfdd22a5890b0d1bf429d2e9b04a0b39a8507/ui/gl/gl_surface.h [modify] https://crrev.com/1d6cfdd22a5890b0d1bf429d2e9b04a0b39a8507/ui/gl/gl_surface_egl.cc [modify] https://crrev.com/1d6cfdd22a5890b0d1bf429d2e9b04a0b39a8507/ui/gl/gl_surface_egl.h
,
Nov 6 2017
The build in which the revert landed was: https://ci.chromium.org/buildbot/chromium.gpu.fyi/Win7%20Release%20%28NVIDIA%29/31983 Shard: https://chromium-swarm.appspot.com/task?id=39a83dab25ed3a10&refresh=10&show_raw=1 and the test still failed. Re-landing (reverting the revert).
,
Nov 6 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/698351b0ed14eb7e162e90f57a615c3f714ff1f8 commit 698351b0ed14eb7e162e90f57a615c3f714ff1f8 Author: Kenneth Russell <kbr@chromium.org> Date: Mon Nov 06 06:58:09 2017 Revert "Revert "gl: Make eglGetFrameTimestamps opt-in."" This reverts commit 1d6cfdd22a5890b0d1bf429d2e9b04a0b39a8507. Reason for revert: not the cause of the test failure; see https://bugs.chromium.org/p/chromium/issues/detail?id=769415#c20 . Original change's description: > Revert "gl: Make eglGetFrameTimestamps opt-in." > > This reverts commit 776b28a0b3efd650371a23949e1e08a097ceda31. > > Reason for revert: speculative revert for GPU process crash: see https://bugs.chromium.org/p/chromium/issues/detail?id=769415#c17 . > > Original change's description: > > gl: Make eglGetFrameTimestamps opt-in. > > > > eglGetFrameTimestamps can't be used with WebVR surfaces since they > > aren't consumed by SurfaceFlinger. Attempting to enable it for WebVR > > surfaces results in logcat spam. > > > > There isn't an easy way to detect WebVR surfaces service-side, so > > this patch requires some client-side changes. > > > > Client side: > > Adds a |enable_swap_timestamps| surface attribute that is only set > > for the default onscreen surfaces. > > > > Service side: > > Adds SupportsSwapTimestamps and SetEnableSwapTimestamps to GLSurface, > > which are properly overriden by NativeViewGLSurfaceEGL based on > > eglGetFrameTimestamps support. > > > > Test: No more logcat spam with WebVR. > > Bug: 769415 > > Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel > > Change-Id: Idb7314aceb2e84e8b886d881dc262fb1b9c7e1ba > > Reviewed-on: https://chromium-review.googlesource.com/748106 > > Reviewed-by: Daniel Cheng <dcheng@chromium.org> > > Reviewed-by: Kenneth Russell <kbr@chromium.org> > > Commit-Queue: Brian Anderson <brianderson@chromium.org> > > Cr-Commit-Position: refs/heads/master@{#513992} > > TBR=dcheng@chromium.org,brianderson@chromium.org,kbr@chromium.org,klausw@chromium.org > > Change-Id: I992667635217406a77e820a52aee6d3066d20d2c > No-Presubmit: true > No-Tree-Checks: true > No-Try: true > Bug: 769415 > Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel > Reviewed-on: https://chromium-review.googlesource.com/754127 > Reviewed-by: Kenneth Russell <kbr@chromium.org> > Commit-Queue: Kenneth Russell <kbr@chromium.org> > Cr-Commit-Position: refs/heads/master@{#514034} TBR=dcheng@chromium.org,brianderson@chromium.org,kbr@chromium.org,klausw@chromium.org # Not skipping CQ checks because original CL landed > 1 day ago. Bug: 769415 Change-Id: Ifd5155f0058010ade2e4e4276eb2621a1704feb5 Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel Reviewed-on: https://chromium-review.googlesource.com/754118 Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Kenneth Russell <kbr@chromium.org> Cr-Commit-Position: refs/heads/master@{#514095} [modify] https://crrev.com/698351b0ed14eb7e162e90f57a615c3f714ff1f8/content/browser/renderer_host/compositor_impl_android.cc [modify] https://crrev.com/698351b0ed14eb7e162e90f57a615c3f714ff1f8/gpu/command_buffer/common/gles2_cmd_utils.h [modify] https://crrev.com/698351b0ed14eb7e162e90f57a615c3f714ff1f8/gpu/ipc/common/gpu_command_buffer_traits_multi.h [modify] https://crrev.com/698351b0ed14eb7e162e90f57a615c3f714ff1f8/gpu/ipc/service/gpu_command_buffer_stub.cc [modify] https://crrev.com/698351b0ed14eb7e162e90f57a615c3f714ff1f8/ui/gl/gl_surface.cc [modify] https://crrev.com/698351b0ed14eb7e162e90f57a615c3f714ff1f8/ui/gl/gl_surface.h [modify] https://crrev.com/698351b0ed14eb7e162e90f57a615c3f714ff1f8/ui/gl/gl_surface_egl.cc [modify] https://crrev.com/698351b0ed14eb7e162e90f57a615c3f714ff1f8/ui/gl/gl_surface_egl.h
,
Nov 10 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/d98d5728326275faa69c03c20ac2cd48e0b1897a commit d98d5728326275faa69c03c20ac2cd48e0b1897a Author: Brian Anderson <brianderson@chromium.org> Date: Fri Nov 10 20:05:54 2017 gl: Use heap in NativeViewGLSurfaceEGL::TraceSwapEvents The std::min negating the DCHECK just above it is an anti pattern. Remove the std::min requirement by using heap allocated vectors and strings that can vary in size. Keep the DCHECK since it would be unexpected that we call eglGetFrameTimestampsANDROID with more timestamps than it can support. Too bad MSVC doesn't support C99's variable length arrays and C++ doesn't yet support something like std::dynarray. Bug: 772743, 769415 Cq-Include-Trybots: master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel Change-Id: I4e96cce8ba1ff81a69e3cd2778e565ddda7d1d98 Reviewed-on: https://chromium-review.googlesource.com/758629 Reviewed-by: Kenneth Russell <kbr@chromium.org> Commit-Queue: Brian Anderson <brianderson@chromium.org> Cr-Commit-Position: refs/heads/master@{#515653} [modify] https://crrev.com/d98d5728326275faa69c03c20ac2cd48e0b1897a/ui/gl/gl_surface_egl.cc
,
Nov 17 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/feaa47c68a99a0817c2b945801fdc9d2873fc6b6 commit feaa47c68a99a0817c2b945801fdc9d2873fc6b6 Author: Brian Anderson <brianderson@chromium.org> Date: Fri Nov 17 23:40:21 2017 gpu: Remove gpu's dependency on ui/latency_info. Add a gfx::SwapResponse that contains the swap result and the timing information for every swap. Let the client add those timestamps to cached LatencyInfos rather than passing LatencyInfo to the GPU service to modify and send back. Behaviorally, this patch is a no-op, but the intent is to: 1) Simplify later patches that will add timestamps to the SwapResponse. 2) Step towards removing PassThroughImageTransportSurface. 3) Remove gpu's dependency on ui/latency_info, which was a layering smell. Bug: 769415 Cq-Include-Trybots: master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.android:android_optional_gpu_tests_rel;master.tryserver.chromium.linux:linux_optional_gpu_tests_rel;master.tryserver.chromium.mac:mac_optional_gpu_tests_rel;master.tryserver.chromium.win:win_optional_gpu_tests_rel Change-Id: I9e1ac97d32600d13ee05467d70fad5674e1355f0 Reviewed-on: https://chromium-review.googlesource.com/767572 Commit-Queue: Brian Anderson <brianderson@chromium.org> Reviewed-by: Robert Sesek <rsesek@chromium.org> Reviewed-by: danakj <danakj@chromium.org> Reviewed-by: Kenneth Russell <kbr@chromium.org> Reviewed-by: Antoine Labour <piman@chromium.org> Cr-Commit-Position: refs/heads/master@{#517614} [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/cc/test/test_context_support.cc [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/cc/test/test_context_support.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/components/viz/service/display/output_surface.cc [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/components/viz/service/display/output_surface.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/components/viz/service/display_embedder/display_output_surface.cc [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/components/viz/service/display_embedder/display_output_surface.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/content/browser/compositor/gpu_browser_compositor_output_surface.cc [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/content/browser/compositor/gpu_browser_compositor_output_surface.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/content/browser/compositor/gpu_output_surface_mac.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/content/browser/compositor/gpu_output_surface_mac.mm [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/content/browser/compositor/gpu_surfaceless_browser_compositor_output_surface.cc [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/content/browser/compositor/gpu_surfaceless_browser_compositor_output_surface.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/content/browser/renderer_host/compositor_impl_android.cc [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/content/test/BUILD.gn [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/BUILD.gn [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/command_buffer/client/client_test_helper.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/command_buffer/client/context_support.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/command_buffer/client/gles2_implementation.cc [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/command_buffer/client/gles2_implementation.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/command_buffer/client/gpu_control.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/command_buffer/tests/decoder_perftest.cc [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/command_buffer/tests/gl_manager.cc [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/command_buffer/tests/gl_manager.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/gles2_conform_support/egl/context.cc [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/gles2_conform_support/egl/context.h [delete] https://crrev.com/c3f2c900f54d80b6ad801abe20d5c7731a6186e3/gpu/ipc/DEPS [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/client/BUILD.gn [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/client/command_buffer_proxy_impl.cc [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/client/command_buffer_proxy_impl.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/client/gpu_channel_host.cc [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/client/gpu_channel_host.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/common/BUILD.gn [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/common/flush_params.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/common/gpu_messages.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/common/gpu_param_traits_macros.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/in_process_command_buffer.cc [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/in_process_command_buffer.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/service/BUILD.gn [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/service/direct_composition_surface_win_unittest.cc [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/service/gpu_channel.cc [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/service/gpu_command_buffer_stub.cc [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/service/gpu_command_buffer_stub.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/service/gpu_vsync_provider_unittest_win.cc [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/service/image_transport_surface_delegate.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/service/image_transport_surface_overlay_mac.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/service/image_transport_surface_overlay_mac.mm [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/service/pass_through_image_transport_surface.cc [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/gpu/ipc/service/pass_through_image_transport_surface.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/ppapi/proxy/ppapi_command_buffer_proxy.cc [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/ppapi/proxy/ppapi_command_buffer_proxy.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/ui/gfx/ipc/gfx_param_traits_macros.h [modify] https://crrev.com/feaa47c68a99a0817c2b945801fdc9d2873fc6b6/ui/gfx/swap_result.h
,
Nov 30 2017
,
Dec 1 2017
,
Jun 27 2018
|
|||||||
►
Sign in to add a comment |
|||||||
Comment 1 by briander...@chromium.org
, Sep 27 2017