Page hangs consistently upon device rotation |
|||||||||||||||
Issue descriptionChrome Version: 61.0.3132.0 canary OS: Android 7.1.2 What steps will reproduce the problem? (1) https://www.techwalls.com/differences-between-iphone-6-6-plus-models/ (2) (3) What is the expected result? No hang What happens instead? Hang. Have to kill chrome. Please use labels and text to provide additional information. For graphics-related bugs, please copy/paste the contents of the about:gpu page at the end of this report.
,
Jun 17 2017
I sent feedback with the magic keyword "lazerbadger" under my wfh@google.com account.
,
Jun 19 2017
This still repro on today's canary wfh@?
,
Jun 20 2017
We are not able to repro this issue on Chrome:61.0.3132.0,61.0.3136.3 Device:Pixel/NJH47D/7.1.2,Pixel XL/NHG47L/7.1.2 and Samsung Galaxy S8(SM-G950U)/NRD90M/7.0.0
,
Jun 20 2017
I realized I was one revision behind so updated to 61.0.3136.3 and I can still reproduce. I filed another feedback with the secret keyword 'lazerbadger'.
,
Jun 20 2017
,
Jun 21 2017
BR is attached @ https://b.corp.google.com/issues/62828196 I reviewed the logs briefly and it looks like Chrome was ANRed due to not Chrome not responding to an ActivityManager touch event within 500ms - unfortunately no great details that I could see in the log, but I didn't spend a huge amount of time there. I did try to repro and couldn't. I attempted to diff wfh@'s variations with my own, and toggled a few items that were dissimilar, but even then could not repro. Since this is resulting in an ANR, will place in the stability sheriff queue to see if someone has insights. Since this is 10/10 reproducible will mark as stable blocking.
,
Jun 21 2017
Not much to go on here with it being not reproducible outside of Will's device. Anyone know how to find a pulse feedback report from a keyword "lazerbadger"?
,
Jun 21 2017
wnwen@, did you look at the BR attached to c#8? I don't think the feedback report is going to give you anything the BR won't. Have you tried toggling any of the experiments listed to see if any of those could be to blame?
,
Jun 22 2017
Relevant logcat:
06-20 19:33:22.938 1083 1083 V NotificationService: pkg=com.google.android.talk canInterrupt=false intercept=true
06-20 19:34:40.957 1083 1141 E BatteryStatsService: no controller energy info supplied
06-20 19:34:41.497 1083 1083 V NotificationService: pkg=com.google.android.talk canInterrupt=false intercept=true
06-20 19:34:54.327 1083 1083 V NotificationService: pkg=com.google.android.talk canInterrupt=false intercept=true
06-20 19:35:11.453 1083 11842 W InputMethodManagerService: Window already focused, ignoring focus gain of: com.android.internal.view.IInputMethodClient$Stub$Proxy@a4ad8e5 attribute=null, token = android.os.BinderProxy@527d140
06-20 19:35:11.462 1083 1948 I ActivityManager: Setting hasTopUi=false for pid=1550
06-20 19:35:48.217 1083 2525 D ConnectivityService: Returning unblocked NetworkInfo to uid=10077
06-20 19:37:23.544 1083 1138 I ProcessStatsService: Prepared write state in 12ms
06-20 19:37:23.586 1083 11842 I ActivityManager: Start proc 3095:com.gyft.android/u0a140 for broadcast com.gyft.android/.services.sync.receiver.StartPayPalProcessorReceiver
06-20 19:37:24.147 1083 27882 I ActivityManager: Killing 468:com.google.android.apps.plus/u0a137 (adj 906): empty #17
06-20 19:37:24.168 1083 3903 D ActivityManager: cleanUpApplicationRecord -- 468
06-20 19:38:47.604 1083 11028 I ActivityManager: Setting hasTopUi=true for pid=1550
06-20 19:38:50.720 1083 27882 I ActivityManager: START u0 {cmp=com.google.android.gms/.update.SystemUpdateActivity} from uid 10019 on display 0
06-20 19:38:50.725 1083 1948 W InputMethodManagerService: Window already focused, ignoring focus gain of: com.android.internal.view.IInputMethodClient$Stub$Proxy@696c23d attribute=null, token = android.os.BinderProxy@527d140
06-20 19:38:50.741 1083 27882 W ActivityManager: startActivity called from non-Activity context; forcing Intent.FLAG_ACTIVITY_NEW_TASK for: Intent { cmp=com.google.android.gms/.update.SystemUpdateActivity }
06-20 19:38:50.757 1083 3900 E ActivityManager: applyOptionsLocked: Unknown animationType=0
06-20 19:38:50.848 1083 1083 V NotificationService: pkg=com.google.android.gms canInterrupt=false intercept=true
06-20 19:38:51.054 1083 1448 I ActivityManager: Setting hasTopUi=false for pid=1550
06-20 19:39:13.139 1083 11843 E BatteryStatsService: no controller energy info supplied
06-20 19:39:13.286 1083 1141 E BatteryStatsService: no controller energy info supplied
06-20 19:39:13.301 1083 1141 E BatteryStatsService: modem info is invalid: ModemActivityInfo{ mTimestamp=0 mSleepTimeMs=0 mIdleTimeMs=0 mTxTimeMs[]=[0, 0, 0, 0, 0] mRxTimeMs=0 mEnergyUsed=0}
06-20 19:39:28.059 1083 1439 I ActivityManager: START u0 {act=android.intent.action.MAIN cat=[android.intent.category.HOME] flg=0x10200000 cmp=com.google.android.apps.nexuslauncher/.NexusLauncherActivity (has extras)} from uid 1000 on display 0
06-20 19:39:28.157 1083 1083 V NotificationService: pkg=com.google.android.gms canInterrupt=false intercept=true
06-20 19:39:28.658 1083 3899 I ActivityManager: START u0 {act=android.intent.action.MAIN cat=[android.intent.category.LAUNCHER] flg=0x10200000 cmp=com.chrome.canary/com.google.android.apps.chrome.Main bnds=[306,2112][582,2392] (has extras)} from uid 100
06-20 19:39:28.703 1083 1445 D ConnectivityService: Returning unblocked NetworkInfo to uid=10118
06-20 19:39:28.738 1083 12741 I ActivityManager: START u0 {act=android.intent.action.MAIN cat=[android.intent.category.LAUNCHER] flg=0x14002000 cmp=com.chrome.canary/org.chromium.chrome.browser.ChromeTabbedActivity bnds=[306,2112][582,2392] (has extras)}
06-20 19:39:31.696 1083 1159 D LuxLevels: bright hysteresis constant= 0.1, threshold=46.533157, lux=42.302868
06-20 19:39:31.696 1083 1159 D LuxLevels: dark hysteresis constant= 0.2, threshold=33.842297, lux=42.302868
06-20 19:39:34.073 1083 1159 D LuxLevels: bright hysteresis constant= 0.1, threshold=36.57123, lux=33.246574
06-20 19:39:34.073 1083 1159 D LuxLevels: dark hysteresis constant= 0.2, threshold=26.59726, lux=33.246574
06-20 19:39:35.340 1083 11843 I ActivityManager: Start proc 3264:com.chrome.canary:sandboxed_process8/u0i111 for service com.chrome.canary/org.chromium.content.app.SandboxedProcessService8
06-20 19:39:38.075 1083 27883 I ActivityManager: Process com.chrome.canary:sandboxed_process7 (pid 25257) has died
06-20 19:39:38.075 1083 27883 D ActivityManager: cleanUpApplicationRecord -- 25257
06-20 19:39:38.075 1083 27883 W ActivityManager: Scheduling restart of crashed service com.chrome.canary/org.chromium.content.app.SandboxedProcessService7 in 1000ms
06-20 19:39:50.437 1083 1159 D LuxLevels: bright hysteresis constant= 0.1, threshold=42.914143, lux=39.012856
06-20 19:39:50.437 1083 1159 D LuxLevels: dark hysteresis constant= 0.2, threshold=31.210285, lux=39.012856
06-20 19:39:51.715 1083 1439 I WindowManager: Input event dispatching timed out sending to com.chrome.canary/org.chromium.chrome.browser.ChromeTabbedActivity. Reason: Waiting to send non-key event because the touched window has not finished processing
06-20 19:39:55.265 1083 1138 E ActivityManager: ANR in com.chrome.canary (com.chrome.canary/org.chromium.chrome.browser.ChromeTabbedActivity)
06-20 19:39:55.265 1083 1138 E ActivityManager: PID: 11799
06-20 19:39:55.265 1083 1138 E ActivityManager: Reason: Input dispatching timed out (Waiting to send non-key event because the touched window has not finished processing certain input events that were delivered to it over 500.0ms ago. Wait queue length
06-20 19:39:55.265 1083 1138 E ActivityManager: Load: 7.15 / 6.75 / 7.07
06-20 19:39:55.265 1083 1138 E ActivityManager: CPU usage from 38801ms to 0ms ago (2017-06-20 19:39:12.934 to 2017-06-20 19:39:51.735):
06-20 19:39:55.265 1083 1138 E ActivityManager: 22% 15675/com.google.android.gms: 19% user + 3.1% kernel / faults: 17790 minor 315 major
06-20 19:39:55.265 1083 1138 E ActivityManager: 16% 11799/com.chrome.canary: 12% user + 3.2% kernel / faults: 29204 minor 423 major
06-20 19:39:55.265 1083 1138 E ActivityManager: 19% 1083/system_server: 13% user + 5.4% kernel / faults: 8760 minor 42 major
06-20 19:39:55.265 1083 1138 E ActivityManager: 6.2% 2336/com.google.android.gms.persistent: 4.6% user + 1.6% kernel / faults: 6687 minor 11 major
06-20 19:39:55.265 1083 1138 E ActivityManager: 2.9% 505/surfaceflinger: 1.2% user + 1.6% kernel / faults: 384 minor
06-20 19:39:55.265 1083 1138 E ActivityManager: 0.7% 11866/com.chrome.canary:privileged_process0: 0.4% user + 0.3% kernel / faults: 9270 minor 30 major
06-20 19:39:55.265 1083 1138 E ActivityManager: 1.1% 613/kschedfreq:0: 0% user + 1.1% kernel
06-20 19:39:55.265 1083 1138 E ActivityManager: 1% 16419/com.google.android.googlequicksearchbox:search: 0.8% user + 0.2% kernel / faults: 9437 minor 10 major
06-20 19:39:55.265 1083 1138 E ActivityManager: 0.7% 118/kswapd0: 0% user + 0.7% kernel
06-20 19:39:55.265 1083 1138 E ActivityManager: 0.7% 2386/mdss_fb0: 0% user + 0.7% kernel
06-20 19:39:55.265 1083 1138 E ActivityManager: 0.1% 649/mediadrmserver: 0% user + 0% kernel / faults: 2480 minor
06-20 19:39:55.265 1083 1138 E ActivityManager: 0.6% 506/audioserver: 0.2% user + 0.3% kernel / faults: 171 minor
06-20 19:39:55.265 1083 1138 E ActivityManager: 0.7% 30864/kworker/u8:14: 0% user + 0.7% kernel
06-20 19:39:55.265 1083 1138 E ActivityManager: 0.5% 1503/com.google.android.inputmethod.latin: 0.4% user + 0.1% kernel / faults: 5543 minor 31 major
06-20 19:39:55.265 1083 1138 E ActivityManager: 0.6% 7/rcu_preempt: 0% user + 0.6% kernel
06-20 19:39:55.265 1083 1138 E ActivityManager: 0% 1343/com.google.android.gms.unstable: 0% user + 0% kernel / faults: 5528 minor
06-20 19:39:55.265 1083 1138 E ActivityManager: 0.5% 1550/com.android.systemui: 0.4% user + 0.1% kernel / faults: 1407 minor 4 major
06-20 19:39:55.265 1083 1138 E ActivityManager: 0.6% 28738/com.google.android.talk: 0.4% user + 0.2% kernel / faults: 1179 minor 13 major
06-20 19:39:55.265 1083 1138 E ActivityManager: 0.6% 2338/kschedfreq:2: 0% user + 0.6% kernel
06-20 19:39:55.265 1083 1138 E ActivityManager: 0.3% 269/irq/480-synapti: 0% user + 0.3% kernel
06-20 19:39:55.265 1083 1138 E ActivityManager: 0.1% 14794/kworker/u8:4: 0% user + 0.1% kernel
06-20 19:39:55.265 1083 1138 E ActivityManager: 0.5% 3/ksoftirqd/0: 0% user + 0.5% kernel
06-20 19:39:55.265 1083 1138 E ActivityManager: 0.3% 2552/com.google.android.apps.nexuslauncher: 0.2% user + 0.1% kernel / faults: 3662 minor 28 major
06-20 19:39:55.265 1083 1138 E ActivityManager: 0.4% 30861/kworker/u8:9: 0% user + 0.4% kernel
06-20 19:39:55.265 1083 1138 E ActivityManager: 0.4% 28785/kworker/u8:16: 0% user + 0.4% kernel
0
,
Jun 22 2017
Haven't worked with variations before, sorry Alex. Perhaps test team has some insights?
,
Jun 25 2017
I have a repro for this with latest canary Just load a long Reddit thread e.g. https://pay.reddit.com/r/politics/comments/6iu431/megathread_senate_aca_repeal_draft_unveiled/ Then rotate phone and rotate back and the page completely hangs
,
Jun 27 2017
dropping JS label as it's not clear that this is a JS hang. wfh@, did your chrome upload a crash (about:crashes)?
,
Jun 28 2017
Unable to reproduce on Pixel/NHG47N and Marlin/NHG47N. Tried with both the steps mentioned in comment #0 and comment #13
,
Jun 28 2017
#13 is still happening on my pixel XL for canary (first screenshot) I am doing repro steps 1. load https://pay.reddit.com/r/politics/comments/6iu431/megathread_senate_aca_repeal_draft_unveiled/ 2. rotate device 3. rotate device back again. 4. Chrome hangs and displays ANR as per screenshots. Chrome version: 61.0.3142.0 OS: Android 7.1.2; Pixel XL Build/NJH47D variations are: 6a89113b-8842e68f 16e0dd70-3f4a17df e81a9d4f-ca7d8d80 da89714-9b69700a 6c18ba9d-3d98b302 c68ab9a3-6edc92c7 b130ecb8-2e32ee7e 6025934e-3f4a17df d175604-65bced95 27af3425-49d6ce1 236d5d9e-347fc86a bd007046-f23d1dea 47e5d3db-3d47f4f4 e036a196-65bced95 77207729-e0e232cc f0a371b8-65bced95 776de70c-e0278d3d 5ca89f9-f23d1dea 6c7c7e88-f23d1dea 57f575bb-f23d1dea 68812885-f23d1dea 3f140034-17bbd7d0 2ce29c0a-3f4a17df 287f29ba-365ad45c 28bf4731-3f4a17df 176d07dc-65bced95 9773d3bd-1410f10 93731dca-803f8fc4 81328e71-f23d1dea 99144bc3-4da47e09 6e434f55-3f4a17df 9e5c75f1-fff58f2c b2972962-ca7d8d80 f79cb77b-3f4a17df 27219e67-b2047178 23a898eb-3f2db5fb 9e46b3e5-4a60b24c e856c60b-f267a6f9 4ea303a6-53a3c5c2 f66ea88e-4eb6d224 93fa4c19-65bced95 d2247fcb-ca7d8d80 2697ea25-ca7d8d80 f56e0452-f23d1dea 65a86f01-ca7d8d80 b2f0086-b4530445 ef25c1eb-f23d1dea 4f864ce1-3d47f4f4 db5efd10-88e62167 1354da85-23b43aed 494d8760-3d47f4f4 3ac60855-3ec2a267 4442aae2-e1cc0f14 ed1d377-e1cc0f14 75f0f0a0-d7f6b13c e7e71889-4ad60575 f5fff3a2-f23d1dea 644b8345-726d8ace 9c7f6723-65bced95 828a5926-c6c0a780 6ba22d84-3f4a17df 64fbb017-cf4f6ead I can repro this on another Android device (61.0.3142.0; Android 7.1.1; Pixel Build/NOF26V) (second screenshot) so I'm not sure why you can't repro.
,
Jun 29 2017
I'm this week's stability sheriff. I can't reproduce the issue using my pixel (7.1.2) and nexus 6p. The relevant code is here - it is poll waiting for something but it does not clearly show what it's waiting for. "main" prio=5 tid=1 Native | group="main" sCount=1 dsCount=0 obj=0x74df6f48 self=0x7c3b695a00 | sysTid=1083 nice=-2 cgrp=default sched=0/0 handle=0x7c3f675a98 | state=S schedstat=( 81320557418 84227442035 216294 ) utm=5887 stm=2243 core=1 HZ=100 | stack=0x7feded4000-0x7feded6000 stackSize=8MB | held mutexes= kernel: (couldn't read /proc/self/task/1083/stack) native: #00 pc 000000000006a32c /system/lib64/libc.so (__epoll_pwait+8) native: #01 pc 000000000001e09c /system/lib64/libc.so (epoll_pwait+64) native: #02 pc 0000000000017f98 /system/lib64/libutils.so (_ZN7android6Looper9pollInnerEi+156) native: #03 pc 0000000000017e4c /system/lib64/libutils.so (_ZN7android6Looper8pollOnceEiPiS1_PPv+60) native: #04 pc 00000000000f0c98 /system/lib64/libandroid_runtime.so (_ZN7android18NativeMessageQueue8pollOnceEP7_JNIEnvP8_jobjecti+48) native: #05 pc 00000000008a14f0 /system/framework/arm64/boot-framework.oat (Java_android_os_MessageQueue_nativePollOnce__JI+140) at android.os.MessageQueue.nativePollOnce(Native method) at android.os.MessageQueue.next(MessageQueue.java:323) at android.os.Looper.loop(Looper.java:136) at com.android.server.SystemServer.run(SystemServer.java:355) at com.android.server.SystemServer.main(SystemServer.java:222) at java.lang.reflect.Method.invoke!(Native method) at com.android.internal.os.ZygoteInit$MethodAndArgsCaller.run(ZygoteInit.java:889) at com.android.internal.os.ZygoteInit.main(ZygoteInit.java:779) There isn't much information here, so I think we need to find a way to reproduce the issues in a consistent way.
,
Jul 12 2017
wfh@, find me when you're back in the office so we can discuss this.
,
Jul 31 2017
removing from sheriff queue since it's not reproducible. Please add back if the issue still exists.
,
Jul 31 2017
Chrome Canary 62.0.3168.3 page in #0 still hangs consistently. See screenshot 1. page in #13 still hangs on a rotate. See screenshot 2. I'm happy to repro in person...
,
Jul 31 2017
it seems I was running an old version of Canary in #21 - I am now running 62.0.3172.0 and I can repro both issues exactly the same (not re-posting the same screenshots)
,
Aug 1 2017
candrada@, can you please sync up with wfh@ in MTV for a demo and perhaps a bisect?
,
Aug 9 2017
worked with wfh@ to try reproing the bug. Thanks to wfh@, he even tried to replicate the same things as his devices on the test device to see if the issue repros there. Unfortunately it still wasn't reproducible. Note, on his devices where the bug repros, it did repro on a 'clean install' of Chrome Dev and not even signed in to Chrome with his account.
,
Aug 19 2017
tedchoc@, aelias@ picking on you randomly here, sorry - I legit have no idea who'd be better to debug. If we have something that's 100% reproducible for wfh@ on his device can we locally build an instrumented APK to try to figure out WTF is going on here? Or suggest someone who could help with this? M61 is rapidly approaching and I don't like having a 100% repro ANR in our queue.
,
Aug 24 2017
To update on this bug, I can still reliably repro on Canary - 62.0.3194.0 Dev - 62.0.3193.3 Beta - 61.0.3163.51 I cannot repro on Stable - 60.0.3112.107 I can repro on both my devices: 1. Android 7.1.2; Pixel XL Build/NJH47D 2. Android 7.1.1; Pixel Build/NOF26V It repros for a fresh install and does not matter if sync is turned on or off. At amineer's suggestion I tried disabling Chrome Home and that did not make any difference.
,
Aug 25 2017
I sat with wfh@ and we bisected on their phone using per-CL builds. Here's what we've got: C 1278.455s Main You are looking for a change made after 478953(GOOD), but before 478954(BAD). Culprit: https://chromium.googlesource.com/chromium/src/+/c6c1d15d250e1c51893cf2abc8a242a8bcf801ea
,
Aug 25 2017
We looked at wfh@'s accessibility services that were enabled, and it was only Tasker (https://play.google.com/store/apps/details?id=net.dinglisch.android.taskerm&hl=en). Disabling Tasker fixed the issue.
,
Aug 25 2017
,
Aug 25 2017
From the bug report, this is the stack of the hang. JVM is holding a condition variable lock, wat?! "main" prio=5 tid=1 Native | group="main" sCount=1 dsCount=0 obj=0x74816000 self=0xe7505400 | sysTid=4346 nice=-10 cgrp=default sched=0/0 handle=0xea17e534 | state=S schedstat=( 35996128027 15195134397 61778 ) utm=3150 stm=448 core=3 HZ=100 | stack=0xff05b000-0xff05d000 stackSize=8MB | held mutexes= kernel: (couldn't read /proc/self/task/4346/stack) native: #00 pc 00017418 /system/lib/libc.so (syscall+28) native: #01 pc 000b6ec1 /system/lib/libart.so (_ZN3art17ConditionVariable16WaitHoldingLocksEPNS_6ThreadE+92) native: #02 pc 003f4e8b /system/lib/libart.so (_ZN3artL12GoToRunnableEPNS_6ThreadE+230) native: #03 pc 003f4d7d /system/lib/libart.so (_ZN3art12JniMethodEndEjPNS_6ThreadE+8) native: #04 pc 0006c809 /data/app/com.chrome.canary-1/oat/arm/base.odex (Java_org_chromium_content_browser_accessibility_WebContentsAccessibility_nativePopulateAccessibilityEvent__JLandroid_view_accessibility_AccessibilityEvent_2II+148) at org.chromium.content.browser.accessibility.WebContentsAccessibility.nativePopulateAccessibilityEvent(Native method) at org.chromium.content.browser.accessibility.WebContentsAccessibility.buildAccessibilityEvent(WebContentsAccessibility.java:273) at org.chromium.content.browser.accessibility.WebContentsAccessibility.sendAccessibilityEvent(WebContentsAccessibility.java:263) at org.chromium.content.browser.accessibility.WebContentsAccessibility.handleContentChanged(WebContentsAccessibility.java:308) at org.chromium.base.SystemMessageHandler.nativeDoRunLoopOnce(Native method) at org.chromium.base.SystemMessageHandler.handleMessage(SystemMessageHandler.java:7) at android.os.Handler.dispatchMessage(Handler.java:102) at android.os.Looper.loop(Looper.java:154) at android.app.ActivityThread.main(ActivityThread.java:6121) at java.lang.reflect.Method.invoke!(Native method) at com.android.internal.os.ZygoteInit$MethodAndArgsCaller.run(ZygoteInit.java:889) at com.android.internal.os.ZygoteInit.main(ZygoteInit.java:779)
,
Aug 25 2017
It's possible Tasker is doing something strange here, but it has 1m+ users, so I expect I won't be the only one who hits this issue when M-61 goes to Stable.
,
Aug 25 2017
I can't reproduce yet. Chrome 62.0.3193.3 Nexus 5X, OMR1 MASTER 4186978 I tried with either Tasker or TalkBack running. I loaded the long Reddit comment thread. With TalkBack running, it definitely takes a couple of seconds to recover after rotating, but no ANR and things run smoothly after a brief pause. That's basically what I'd expect to happen for such a large page. With Tasker running I don't detect any difference, and if I visit chrome://accessibility I don't see accessibility enabled - which is what I'd hope for. Is it possible this was fixed on trunk? Let me install a 61 build next, but other ideas would be welcome.
,
Aug 25 2017
see #26 - I can still repro on latest Canary. I just re-verified it still happens on 62.0.3196.0. the best page for a reliable repro is in comment #13 - that is what we used for the bisect: rotate and then rotate back. When I was testing on a new device for comment #24, I also installed tasker and even copied my tasker configurations onto the test device, and I was unable to repro, so I'm not sure what specifically it is about tasker or the tasker configuration that causes the repro. Re: "That's basically what I'd expect to happen for such a large page." - it didn't happen before the CL in #27 so something definitely changed...
,
Aug 25 2017
So there is no chrome native stack in that stack trace. This is from the dex file, meaning it's coming from java. native: #04 pc 0006c809 /data/app/com.chrome.canary-1/oat/arm/base.odex So this is hanging in JniMethodEnd in the jvm, and by that method name, it's hanging when returning from a java->native call, and have already exited the chromium part of the stack. I honestly don't know how that can happen..
,
Aug 25 2017
Or alternative theory... it's not actually a deadlock and it's just middle of doing a lot of work for accessibility? wfh: can you reproduce the anr a few more times, and each time after the "app not responding" dialog comes up, adb pull /data/anr/traces.txt and attach the traces.txt files here?
,
Aug 25 2017
Over to dmazzoni@ since he's been looking today. jinsukkim@, PTAL ASAP as well.
,
Aug 25 2017
@askatte provided a test device and @wfh configured it to repro the bug. Working on it now.
,
Aug 25 2017
,
Aug 26 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/400e90126213561c94c377ee786ab34d5b4ffe40 commit 400e90126213561c94c377ee786ab34d5b4ffe40 Author: Dominic Mazzoni <dmazzoni@chromium.org> Date: Sat Aug 26 01:04:52 2017 Fix Android optimization in SendLocationChangeEvents BrowserAccessibilityManagerAndroid::SendLocationChangeEvents contains an optimization (387488, crbug.com/588889 ) that avoids sending too many events when a lot of objects on the screen change their bounding box at the same time - for example, when rotating the screen. This optimization was accidentally broken by r478954, which accidentally dropped the "return" at the end of the conditional, causing it to run the efficient version and then the slow version. Bug: 734353 Change-Id: I3f8abe9b6d4188257c5cc316f71f5ef07f345ecd Reviewed-on: https://chromium-review.googlesource.com/636169 Reviewed-by: Paul Miller <paulmiller@chromium.org> Reviewed-by: David Tseng <dtseng@chromium.org> Commit-Queue: Dominic Mazzoni <dmazzoni@chromium.org> Cr-Commit-Position: refs/heads/master@{#497622} [modify] https://crrev.com/400e90126213561c94c377ee786ab34d5b4ffe40/content/browser/accessibility/browser_accessibility_manager_android.cc
,
Aug 26 2017
,
Aug 26 2017
This bug requires manual review: We are only 9 days from stable. Please contact the milestone owner if you have questions. Owners: amineer@(Android), cmasso@(iOS), ketakid@(ChromeOS), govind@(Desktop) For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
Aug 27 2017
Sorry for the mistake in r478954 - glad that it is being taken care of.
,
Aug 29 2017
Merge approved for M61 branch 3163.
,
Aug 29 2017
I'll merge now
,
Aug 29 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/92a2ea4d7e1da500eb636430b3dcb3e270a11ed0 commit 92a2ea4d7e1da500eb636430b3dcb3e270a11ed0 Author: Dominic Mazzoni <dmazzoni@chromium.org> Date: Tue Aug 29 17:10:25 2017 Merge to M61: Fix Android optimization in SendLocationChangeEvents BrowserAccessibilityManagerAndroid::SendLocationChangeEvents contains an optimization (387488, crbug.com/588889 ) that avoids sending too many events when a lot of objects on the screen change their bounding box at the same time - for example, when rotating the screen. This optimization was accidentally broken by r478954, which accidentally dropped the "return" at the end of the conditional, causing it to run the efficient version and then the slow version. (cherry picked from commit 400e90126213561c94c377ee786ab34d5b4ffe40) Bug: 734353 Change-Id: I3f8abe9b6d4188257c5cc316f71f5ef07f345ecd Reviewed-on: https://chromium-review.googlesource.com/636169 Reviewed-by: Paul Miller <paulmiller@chromium.org> Reviewed-by: David Tseng <dtseng@chromium.org> Commit-Queue: Dominic Mazzoni <dmazzoni@chromium.org> Cr-Original-Commit-Position: refs/heads/master@{#497622} Reviewed-on: https://chromium-review.googlesource.com/641662 Reviewed-by: Dominic Mazzoni <dmazzoni@chromium.org> Cr-Commit-Position: refs/branch-heads/3163@{#972} Cr-Branched-From: ff259bab28b35d242e10186cd63af7ed404fae0d-refs/heads/master@{#488528} [modify] https://crrev.com/92a2ea4d7e1da500eb636430b3dcb3e270a11ed0/content/browser/accessibility/browser_accessibility_manager_android.cc
,
Aug 29 2017
Assuming this is fixed given we're good on trunk & branch, marking as such. Reopen if more work is required. Thanks so much for reporting, wfh@ and fixing, dmazzoni@!
,
Aug 29 2017
yes, thank you so much for the fix! I can no longer reproduce on Chrome Canary 62.0.3199.3 (fix landed in 3198). |
|||||||||||||||
►
Sign in to add a comment |
|||||||||||||||
Comment 1 by wfh@chromium.org
, Jun 17 2017