Crash in [@ libhwui.so@0x30afa]
Categories
(Firefox for Android Graveyard :: General, defect, P2)
Tracking
(firefox66 affected, firefox67 affected, firefox68 affected)
People
(Reporter: marcia, Unassigned)
Details
(Keywords: crash, Whiteboard: [media-q2])
Crash Data
This bug is for crash report bp-970bc9c3-763e-4ab8-94c3-5222a0190403.
LG phones running API 27 are crashing: https://bit.ly/2IdKm1Y. On 68 nightly, this is the #7 overall crash and filing since it didn't have a bug. Some of the comments mention video crashing.
Top 10 frames of crashing thread:
0 libc.so libc.so@0x1a6fa
1 libhwui.so libhwui.so@0x30afa
2 libhwui.so libhwui.so@0x30afa
3 libhwui.so libhwui.so@0x30afa
4 liblog.so liblog.so@0x6593
5 libhwui.so libhwui.so@0x9d37d
6 liblog.so liblog.so@0x65ae
7 dalvik-main space (region space) (deleted) dalvik-main space @0xdb26163
8 ashmem (deleted) ashmem @0xcddf52
9 dalvik-main space (region space) (deleted) dalvik-main space @0xd99524d
Reporter | ||
Comment 1•6 years ago
|
||
The crash graph does show an increase in crashes since November. This continues to only affect LG devices running API 27. Between the 66.0 releases so far we have over 1000 crashes.
Updated•6 years ago
|
Comment 2•6 years ago
|
||
Looking at the crash reports it's difficult to understand what's happening. The comments aren't leading to steps to reproduce. The crash stats don't have symbols either.
:marcia - do you have access to URLs associated to the crashes by chance?
Reporter | ||
Comment 3•6 years ago
|
||
There are lots of URLs listed - most appear to be general browsing. I can add some if you think it would be useful.
As far as no symbols, this is fairly common with mobile crashes AFAIK.
Top crashing phones are LM-Q710(FGN), LG-M700 and LM-Q710.FG. We probably need to get one of the phones to reproduce, unless Kevin or someone else can reproduce in an emulator.
I looked at some of the individual logcats, and there are appears to be a few different things going on. On some devices touch events are showing up in the logcat.
Reporter | ||
Comment 4•6 years ago
|
||
Sorina - Do you know if we any of the three LG devices that I called out in Comment 3?
- LG Stylo 4
- LG Q6
- LG Q7
Also some correlations to Adreno:
- (66.07% in signature vs 06.87% overall) adapter_device_id = Adreno (TM) 506 [66.37% vs 13.16% if adapter_vendor_id = Qualcomm]
Comment 5•6 years ago
|
||
(In reply to Marcia Knous [:marcia - needinfo? me] from comment #4)
Sorina - Do you know if we any of the three LG devices that I called out in Comment 3?
- LG Stylo 4
- LG Q6
- LG Q7
Also some correlations to Adreno:
- (66.07% in signature vs 06.87% overall) adapter_device_id = Adreno (TM) 506 [66.37% vs 13.16% if adapter_vendor_id = Qualcomm]
Unfortunately, we don't have the LG devices listed above. I leave the needinfo on me and also on Andrei to check if we have devices with Qualcomm Adreno 506.
Updated•6 years ago
|
Comment 6•6 years ago
|
||
Hello, please note that I couldn't find any device mentioned in Comment 3, the closest one I could find was a Samsung Galaxy J7(GPU- MALI-T830) but I couldn't reproduce the issue.
Note that the correlations mentioned in Comment 4 regarding Adreno are that the Device is Samsung Galaxy J7 din 2017(GPU - Qualcomm Adreno ™ 506).
Also, note that on the emulators that I have available I have only Google Devices.
During my tests I will try to investigate more this issue on Samsung Galaxy J7(GPU- MALI-T830).
Reporter | ||
Comment 7•6 years ago
|
||
This continues to be a top 10 crash in 67 release and is seen in 68 as well. Top crashing devices are all LG, with LM-Q710(FGN) being the top crashing device. Adding the stalled keyword since without devices it doesn't seem we can do much here.
Reporter | ||
Comment 8•6 years ago
|
||
No crashes in 68.2.1. Resolving this one as WFM. We can reopen if crashes are seen again.
Comment 9•6 years ago
|
||
Since the bug is closed, the stalled keyword is now meaningless.
For more information, please visit auto_nag documentation.
Assignee | ||
Updated•4 years ago
|
Description
•