browse:news:globo crashes on Android One |
|||||
Issue descriptionThis story started failing consistently on https://luci-milo.appspot.com/buildbot/chromium.perf/Android%20One%20Perf%20%282%29 Chrome crashes with: I 2.230s Main Stack Trace: I 2.230s Main RELADDR FUNCTION FILE:LINE I 2.230s Main 0003d980 tgkill+12 /system/lib/libc.so I 2.230s Main 00016db1 pthread_kill+52 /system/lib/libc.so I 2.230s Main 000179c7 raise+10 /system/lib/libc.so I 2.231s Main 00014169 __libc_android_abort+36 /system/lib/libc.so I 2.231s Main 000124f0 abort+4 /system/lib/libc.so I 2.231s Main 00216461 art::Runtime::Abort()+160 /system/lib/libart.so I 2.231s Main 000a6d25 art::LogMessage::~LogMessage()+1312 /system/lib/libart.so I 2.231s Main 00223cb7 art::Thread::AssertNoPendingException() const+354 /system/lib/libart.so I 2.231s Main 000d713b art::ClassLinker::FindClass(art::Thread*, char const*, art::Handle<art::mirror::ClassLoader>)+14 /system/lib/libart.so I 2.231s Main 001cc56d art::JNI::FindClass(_JNIEnv*, char const*)+788 /system/lib/libart.so I 2.231s Main 00001f2d jniRegisterNativeMethods+20 /system/lib/libnativehelper.so I 2.231s Main 0005f815 <unknown> /system/lib/libandroid_runtime.so I 2.231s Main 000608e9 android::AndroidRuntime::startReg(_JNIEnv*)+28 /system/lib/libandroid_runtime.so I 2.231s Main 00060a2d android::AndroidRuntime::start(char const*, android::Vector<android::String8> const&)+284 /system/lib/libandroid_runtime.so I 2.231s Main 00001675 <unknown> /system/bin/app_process32 I 2.231s Main 000123a1 __libc_init+44 /system/lib/libc.so I 2.231s Main 000017b0 <unknown> /system/bin/app_process32 https://luci-logdog.appspot.com/v/?s=chrome%2Fbb%2Fchromium.perf%2FAndroid_One_Perf__2_%2F6303%2F%2B%2Frecipes%2Fsteps%2Fsystem_health.common_mobile%2F0%2Fstdout ssid can you have a look to triage and maybe disable if no quick fix is found? Interestingly, the error seems to happen on system_health.common_mobile and *not* on memory_mobile.
,
Apr 25 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/4c78ae4f85fb2e25e10ca92abd86269fc4d38643 commit 4c78ae4f85fb2e25e10ca92abd86269fc4d38643 Author: ssid <ssid@chromium.org> Date: Tue Apr 25 02:00:44 2017 [system_health] Reduce the number of article reads in globo story The renderer OOMs on system_health.common benchmark on Android One device because of tracing and other metrics being enabled. So, reduce the number of article reads. BUG=714650 R=nednguyen@google.com,perezju@chromium.org Review-Url: https://codereview.chromium.org/2842593002 Cr-Commit-Position: refs/heads/master@{#466868} [modify] https://crrev.com/4c78ae4f85fb2e25e10ca92abd86269fc4d38643/tools/perf/page_sets/system_health/browsing_stories.py
,
Apr 25 2017
,
May 2 2017
We're seeing crashes again on system_health.common_mobile. For example: https://luci-logdog.appspot.com/v/?s=chrome%2Fbb%2Fchromium.perf%2FAndroid_One_Perf__2_%2F6381%2F%2B%2Frecipes%2Fsteps%2Fsystem_health.common_mobile%2F0%2Fstdout Screenshot ("Aw, Snap!"): https://console.developers.google.com/m/cloudstorage/b/chrome-telemetry-output/o/profiler-file-id_10-2017-05-02_04-45-5136397.png
,
May 2 2017
Also interesting that: a) the story does not fail on system_health.memory_mobile (where it is run 3 times per build), e.g.: b) the story does not fail on system_health.common_mobile.reference (run on the very So b) definitely points to some regression in Chrome?
,
May 2 2017
* the end b) should have read: "run on the very same device". For the moment I've sent out CL to disable the story.
,
May 2 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/718f39723b40178893aed49600c24773c0ed5a88 commit 718f39723b40178893aed49600c24773c0ed5a88 Author: perezju <perezju@chromium.org> Date: Tue May 02 17:13:04 2017 [tools/perf] Disable browse:news:globo Story failing on Android One bot. TBR=nednguyen@google.com,ssid@chromium.org BUG=714650 Review-Url: https://codereview.chromium.org/2858603002 Cr-Commit-Position: refs/heads/master@{#468691} [modify] https://crrev.com/718f39723b40178893aed49600c24773c0ed5a88/tools/perf/page_sets/system_health/browsing_stories.py
,
Nov 3 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/f32bd811a08c7e5769666711fdc190de46f21378 commit f32bd811a08c7e5769666711fdc190de46f21378 Author: Annie Sullivan <sullivan@chromium.org> Date: Fri Nov 03 12:00:14 2017 Disable browse:news:globo story on v8.runtime_stats.browsing_mobile. It's already disabled in v8.browsing_mobile, and I think this was just an oversight. Bug: 714650 Change-Id: Ic699826a60bff0403cce6a616453622d9492bb58 Reviewed-on: https://chromium-review.googlesource.com/751964 Reviewed-by: Ned Nguyen <nednguyen@google.com> Commit-Queue: Ned Nguyen <nednguyen@google.com> Cr-Commit-Position: refs/heads/master@{#513749} [modify] https://crrev.com/f32bd811a08c7e5769666711fdc190de46f21378/tools/perf/benchmarks/v8_browsing.py
,
Dec 13 2017
,
Dec 14 2017
,
Jul 19
📍 Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/1116e485a40000
,
Jul 19
📍 Job complete. See results below. https://pinpoint-dot-chromeperf.appspot.com/job/1116e485a40000
,
Aug 14
Looks like this is still failing flakily in the last perf try run.
,
Aug 14
I think Android One bot (upstream) no longer exists? Except for memory mobile benchmarks (which still run on a downstream Android One bot), we can probably remove all Android One disables from expectations file.
,
Aug 24
📍 Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/16a440c5640000
,
Aug 24
📍 Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/12f87c75640000
,
Aug 24
📍 Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/15beb2d9640000
,
Aug 24
📍 Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/16f87c75640000
,
Aug 24
Kicked off try jobs for the same patch on a bunch of non-Android-One platforms to make sure it's working there.
,
Aug 24
📍 Job complete. See results below. https://pinpoint-dot-chromeperf.appspot.com/job/16f87c75640000
,
Aug 24
📍 Job complete. See results below. https://pinpoint-dot-chromeperf.appspot.com/job/16a440c5640000
,
Aug 24
📍 Job complete. See results below. https://pinpoint-dot-chromeperf.appspot.com/job/15beb2d9640000
,
Aug 24
📍 Job complete. See results below. https://pinpoint-dot-chromeperf.appspot.com/job/12f87c75640000
,
Sep 18
This appears to be failing for all android bots, should we consider removing the story since this bug is over a year old?
,
Sep 18
Some of the failed pinpoint jobs had "build" failures, before even attempting to run the test. I'll retry one of those.
,
Sep 18
📍 Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/14eb5560e40000
,
Sep 18
📍 Job complete. See results below. https://pinpoint-dot-chromeperf.appspot.com/job/14eb5560e40000
,
Sep 18
Ah! The problem is "Patch failure". Sounds like charlie's original CL to re-enable the story needs to be rebased now.
,
Sep 19
📍 Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/10d8df97640000
,
Sep 19
📍 Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/13acf7ff640000
,
Sep 19
📍 Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/15fc9a3f640000
,
Sep 19
📍 Pinpoint job started. https://pinpoint-dot-chromeperf.appspot.com/job/142ed170e40000
,
Sep 19
I went ahead and reuploaded the CL synced to HEAD. I also kicked off four try jobs - the one that you tried to kick off earlier (system_health.common_mobile), along with the same tryjob on each of the different benchmarks where this story was disabled (sh.memory_mobile, v8.browsing_mobile, and v8.browsing_mobile-future).
,
Sep 20
📍 Job complete. See results below. https://pinpoint-dot-chromeperf.appspot.com/job/13acf7ff640000
,
Sep 20
📍 Job complete. See results below. https://pinpoint-dot-chromeperf.appspot.com/job/15fc9a3f640000
,
Sep 20
📍 Job complete. See results below. https://pinpoint-dot-chromeperf.appspot.com/job/142ed170e40000
,
Sep 20
📍 Job complete. See results below. https://pinpoint-dot-chromeperf.appspot.com/job/10d8df97640000 |
|||||
►
Sign in to add a comment |
|||||
Comment 1 by ssid@chromium.org
, Apr 24 2017