Crash in [@ nvwgf2um_cfg.dll | CDecodeContext::BeginFrame]
Categories
(Core :: Audio/Video: Playback, defect, P3)
Tracking
()
People
(Reporter: yoasif, Unassigned)
References
Details
(Keywords: crash, Whiteboard: [media-crash])
Crash Data
Crash report: https://crash-stats.mozilla.org/report/index/67ff1748-4223-4b82-be7c-90bef0201214
Reason: EXCEPTION_ACCESS_VIOLATION_READ
Top 10 frames of crashing thread:
0 nvwgf2um_cfg.dll nvwgf2um_cfg.dll@0xdbcf72
1 nvwgf2um_cfg.dll nvwgf2um_cfg.dll@0xdbce59
2 nvwgf2um_cfg.dll nvwgf2um_cfg.dll@0xdbc31a
3 nvwgf2um_cfg.dll nvwgf2um_cfg.dll@0xdbc35f
4 nvwgf2um_cfg.dll nvwgf2um_cfg.dll@0xdc17f7
5 nvwgf2um_cfg.dll nvwgf2um_cfg.dll@0xdc161a
6 nvwgf2um_cfg.dll nvwgf2um_cfg.dll@0xdd045f
7 nvwgf2um_cfg.dll nvwgf2um_cfg.dll@0xe5b56a
8 nvwgf2um_cfg.dll nvwgf2um_cfg.dll@0xe5bb67
9 nvwgf2um_cfg.dll nvwgf2um_cfg.dll@0xdd3379
User posted here: https://www.reddit.com/r/firefox/comments/kd5w7f/firefox_83_crashing_when_scrolling_too_far_down/
Sometimes when I am actively not doing work, I scroll through r/all. I sometimes get pretty far down and the tab invariably crashes. Is this a Firefox issue or a computer hardware issue? The error report says it's an EXCEPTION_ACCESS_VIOLATION_READ crash, but I don't really know what that means.
I am on windows 10 x86 with Firefox 83.
Looks like bug 1374148 but user is using WR.
|
||
Comment 1•3 years ago
|
||
This identified the crashing dll - nvwgf2um_cfg.dll - you can search a search engine and find that it is Nvidia.
This seems like an Nvidia driver crash. Not sure what Firefox should do here. Hand this to gfx to see if we need to do something or not.
|
||
Comment 2•3 years ago
|
||
Do we need to block webrender in the same cases that we blocked D3D in bug 1374148?
|
||
Comment 3•3 years ago
|
||
This looks like a media crash inside of DXVA. Any blocking should likely happen there.
Forwarding to the media team.
See bug 1374148 as an example of how to block DLLs, if that's what's needed here -- I don't know, please investigate!
It looks like the majority of these crashes are happening on SeaMonkey, which I don't know what to make of. I'm not opposed to blacklisting some of the crashier drivers here, but it's not clear to me what the mechanism to do that is following our remote decoding changes. I'll hold the NI until I'm able to investigate further.
Given the low Fx crash rates and no crashes on recent releases at this time, I'm reluctant to block here. If we were to block, I suppose the current mechanism is here -- which would also apply to our out of process decoders.
I'm not clear on how effective blocking would be for SeaMonkey, as it sounds like they're based on ESR 60 at this time, and I don't know if we can uplift that far. This sounds like something they may wish to block specifically on their branch.
I'll move this to media so the appropriate people can be pinged if the rate goes up.
|
||
Comment 7•2 years ago
|
||
Closing because no crashes reported for 12 weeks.
Description
•