Closed Bug 1775460 Opened 2 years ago Closed 2 years ago

Intermittent TEST-UNEXPECTED-CRASH | /mediacapture-fromelement/HTMLCanvasElement-getImageData-noframe.html | expected OK

Categories

(Core :: Audio/Video: Recording, defect, P5)

defect

Tracking

()

RESOLVED DUPLICATE of bug 1743143

People

(Reporter: intermittent-bug-filer, Unassigned)

Details

(Keywords: intermittent-failure)

Filed by: ctuns [at] mozilla.com
Parsed log: https://treeherder.mozilla.org/logviewer?job_id=382133334&repo=mozilla-central
Full log: https://firefox-ci-tc.services.mozilla.com/api/queue/v1/task/MWzmLcMQSza-nZkmT7wSPQ/runs/0/artifacts/public/logs/live_backing.log


[task 2022-06-22T10:59:58.517Z] 10:59:58     INFO - TEST-START | /mediacapture-fromelement/HTMLCanvasElement-getImageData-noframe.html
[task 2022-06-22T11:00:01.450Z] 11:00:01     INFO - PID 16166 | ==16166==ERROR: ThreadSanitizer failed to allocate 0xf698005c4000 (271132701507584) bytes at address 81000001000 (errno: 12)
[task 2022-06-22T11:00:01.502Z] 11:00:01     INFO - PID 16166 | Exiting due to channel error.
[task 2022-06-22T11:00:01.503Z] 11:00:01     INFO - PID 16166 | Exiting due to channel error.
[task 2022-06-22T11:00:01.503Z] 11:00:01     INFO - PID 16166 | Crash Annotation GraphicsCriticalError: |[C0][GFX1-]: Receive IPC close with reason=AbnormalShutdown (t=5.04441) Exiting due to channel error.
[task 2022-06-22T11:00:01.503Z] 11:00:01     INFO - IOError on command, setting status to CRASH
[task 2022-06-22T11:00:01.505Z] 11:00:01     INFO - TEST-UNEXPECTED-CRASH | /mediacapture-fromelement/HTMLCanvasElement-getImageData-noframe.html | expected OK
[task 2022-06-22T11:00:01.506Z] 11:00:01     INFO - TEST-INFO took 2986ms
[task 2022-06-22T11:00:01.514Z] 11:00:01     INFO - PID 16166 | Exiting due to channel error.
[task 2022-06-22T11:00:01.514Z] 11:00:01     INFO - PID 16166 | Crash Annotation GraphicsCriticalError: |[C0][GFX1-]: Receive IPC close with reason=AbnormalShutdown (t=6.72763) Crash Annotation GraphicsCriticalError: |[C0][GFX1-]: Receive IPC close with reason=AbnormalShutdown (t=18.6286) [GFX1-]: Receive IPC close with reason=AbnormalShutdown
[task 2022-06-22T11:00:01.514Z] 11:00:01     INFO - PID 16166 | Exiting due to channel error.
[task 2022-06-22T11:00:01.514Z] 11:00:01     INFO - PID 16166 | Crash Annotation GraphicsCriticalError: |[C0][GFX1-]: Receive IPC close with reason=AbnormalShutdown (t=19.5498) Exiting due to channel error.
[task 2022-06-22T11:00:02.541Z] 11:00:02     INFO - Browser exited with return code -6
[task 2022-06-22T11:00:02.542Z] 11:00:02     INFO - Closing logging queue
[task 2022-06-22T11:00:02.542Z] 11:00:02     INFO - queue closed
[task 2022-06-22T11:00:02.553Z] 11:00:02     INFO - INFO | runtests.py | TSan using symbolizer at /builds/worker/workspace/build/application/firefox/llvm-symbolizer
[task 2022-06-22T11:00:02.553Z] 11:00:02     INFO - Application command: /builds/worker/workspace/build/application/firefox/firefox --marionette about:blank -profile /tmp/tmps_0erbk_
[task 2022-06-22T11:00:02.572Z] 11:00:02     INFO - PID 16468 | 1655895600606	Marionette	INFO	Marionette enabled
[task 2022-06-22T11:00:02.577Z] 11:00:02     INFO - PID 16468 | 1655895600625	Marionette	INFO	Listening on port 36818
[task 2022-06-22T11:00:02.581Z] 11:00:02     INFO - Starting runner
[task 2022-06-22T11:00:12.523Z] 11:00:12     INFO - PID 16468 | console.warn: SearchSettings: "get: No settings file exists, new profile?" (new NotFoundError("Could not open the file at /tmp/tmpv3wsro7x/search.json.mozlz4", (void 0)))
[task 2022-06-22T11:00:20.171Z] 11:00:20     INFO - TEST-START | /mediacapture-fromelement/capture.html
Status: NEW → RESOLVED
Closed: 2 years ago
Resolution: --- → INCOMPLETE
Status: RESOLVED → REOPENED
Resolution: INCOMPLETE → ---
Status: REOPENED → RESOLVED
Closed: 2 years ago2 years ago
Resolution: --- → INCOMPLETE
Status: RESOLVED → REOPENED
Resolution: INCOMPLETE → ---
Status: REOPENED → RESOLVED
Closed: 2 years ago2 years ago
Resolution: --- → INCOMPLETE
Status: RESOLVED → REOPENED
Resolution: INCOMPLETE → ---

This bug seems to be covering several different types of crashes.

The initial report here was a crash with "ThreadSanitizer failed to allocate 0xf698005c4000 (271132701507584)" which is a version of bug 1773612, which is not specific to any one test. We should probably set that aside and focus on the later crashes here.

The more recent crashes (comment 7 and comment 9) are in [@ _g_log_abort], crashing deep in X11/gdk/glib code, inside of our call to mozilla::gl::GLXLibrary::fQueryServerString from mozilla::gl::GLXLibrary::EnsureInitialized.

Here's a pruned backtrace from https://treeherder.mozilla.org/logviewer?job_id=397725438&repo=autoland&lineNumber=16645 (comment 6's log):

[task 2022-11-25T15:52:06.211Z] 15:52:06     INFO - PROCESS-CRASH | /mediacapture-fromelement/HTMLCanvasElement-getImageData-noframe.html | application crashed [@ _g_log_abort]
[...]
[task 2022-11-25T15:52:06.215Z] 15:52:06     INFO - Crash reason:  SIGTRAP / SI_KERNEL
[task 2022-11-25T15:52:06.215Z] 15:52:06     INFO - Crash address: 0x0
[task 2022-11-25T15:52:06.215Z] 15:52:06     INFO - Process uptime: not available
[task 2022-11-25T15:52:06.215Z] 15:52:06     INFO - 
[task 2022-11-25T15:52:06.215Z] 15:52:06     INFO - Thread 21 CanvasRenderer (crashed)
[task 2022-11-25T15:52:06.215Z] 15:52:06     INFO -  0  libglib-2.0.so.0!_g_log_abort [gmessages.c : 583 + 0x0]
[task 2022-11-25T15:52:06.216Z] 15:52:06     INFO -  1  libglib-2.0.so.0!g_log_writer_default [gmessages.c : 2735 + 0xb]
[task 2022-11-25T15:52:06.217Z] 15:52:06     INFO -  2  libglib-2.0.so.0!g_log_structured_array [gmessages.c : 1970 + 0xe]
[task 2022-11-25T15:52:06.217Z] 15:52:06     INFO -  3  libglib-2.0.so.0!g_log_structured_standard [gmessages.c : 2027 + 0x4]
[task 2022-11-25T15:52:06.217Z] 15:52:06     INFO -  4  libgdk-3.so.0 + 0x60c40
[task 2022-11-25T15:52:06.230Z] 15:52:06     INFO -  5  libgdk-3.so.0 + 0x6dac2
[task 2022-11-25T15:52:06.230Z] 15:52:06     INFO -  6  libX11.so.6!<unknown in 24168452d079d48767ebadd480872381d71faa.debug> + 0x29835e
[task 2022-11-25T15:52:06.231Z] 15:52:06     INFO -  7  libX11.so.6!_XError [XlibInt.c : 1434 + 0x7]
[task 2022-11-25T15:52:06.231Z] 15:52:06     INFO -  8  libX11.so.6!handle_error [xcb_io.c : 199 + 0xa]
[task 2022-11-25T15:52:06.232Z] 15:52:06     INFO -  9  libX11.so.6!handle_response [xcb_io.c : 320 + 0x4]
[task 2022-11-25T15:52:06.233Z] 15:52:06     INFO - 10  libX11.so.6!_XReply [xcb_io.c : 630 + 0xf]
[task 2022-11-25T15:52:06.233Z] 15:52:06     INFO - 11  libX11.so.6!XSync [Sync.c : 44 + 0x11]
[task 2022-11-25T15:52:06.234Z] 15:52:06     INFO - 12  libGLX_mesa.so.0 + 0x45135
[task 2022-11-25T15:52:06.234Z] 15:52:06     INFO - 13  libX11.so.6!<unknown in 24168452d079d48767ebadd480872381d71faa.debug> + 0x297d36
[task 2022-11-25T15:52:06.234Z] 15:52:06     INFO - 14  libGLX_mesa.so.0 + 0x37f37
[task 2022-11-25T15:52:06.234Z] 15:52:06     INFO - 15  libGLX_mesa.so.0 + 0x335c3
[task 2022-11-25T15:52:06.234Z] 15:52:06     INFO - 16  libGLX_mesa.so.0 + 0x33a30
[task 2022-11-25T15:52:06.235Z] 15:52:06     INFO - 17  libxul.so!mozilla::gl::GLXLibrary::fQueryServerString(_XDisplay*, int, int) const [GLXLibrary.h:81a6f7f802a09e72798d4157f6096e674a746bb9 : 131 + 0xb]
[task 2022-11-25T15:52:06.235Z] 15:52:06     INFO - 18  libxul.so!mozilla::gl::GLXLibrary::EnsureInitialized(_XDisplay*) [GLContextProviderGLX.cpp:81a6f7f802a09e72798d4157f6096e674a746bb9 : 187 + 0x12]
[task 2022-11-25T15:52:06.235Z] 15:52:06     INFO - 19  libGLX_mesa.so.0 + 0x5d8ef
[task 2022-11-25T15:52:06.235Z] 15:52:06     INFO - 20  libGL.so.1 + 0x4167f

We have a few other bugs filed on intermittent crashes at [@ _g_log_abort]: bug 1759954, bug 1694636, bug 1770828.

Likely these are all the same issue (possibly an external bug).

Aha, and bug 1743143 seems to be the general bug for [@ _g_log_abort] crashes in any test.

--> Duping this that bug, since that seems to be what's going on in most of the failures here (all but comment 0, which is bug 1743143).

Status: REOPENED → RESOLVED
Closed: 2 years ago2 years ago
Duplicate of bug: 1743143
Resolution: --- → DUPLICATE
You need to log in before you can comment on or make changes to this bug.