Crash on canalplus.com [@ mozilla::MediaDecoder::SetCDMProxy ]
Categories
(Core :: Audio/Video: Playback, defect, P3)
Tracking
()
People
(Reporter: mayankleoboy1, Assigned: aosmond)
References
Details
Crash Data
Attachments
(3 files)
I was on particularly slow network.
I was on canalplus, testing stuff. I had opened the page for one of the sport streams..
While a stream loads, I triggered a "Kill GPU process" from about:support
AR: Crash
https://crash-stats.mozilla.org/report/index/a24418de-83e1-4fb0-896b-410620240725#tab-bugzilla
https://crash-stats.mozilla.org/report/index/1a82b08c-500b-40bf-895f-ca5970240725#tab-bugzilla
Can repro intermittently
Reporter | ||
Comment 1•9 months ago
|
||
Reporter | ||
Updated•9 months ago
|
Reporter | ||
Updated•9 months ago
|
Reporter | ||
Comment 2•9 months ago
|
||
Most of the time the tab recovers after killing the gpu process. But on these 2 occasions, the tab crashed.
Comment 3•9 months ago
|
||
The severity field is not set for this bug.
:jimm, could you have a look please?
For more information, please visit BugBot documentation.
Updated•8 months ago
|
Updated•2 months ago
|
Assignee | ||
Comment 4•13 days ago
|
||
Based on the crash reports in nightly/release, I don't believe this necessarily requires a GPU process crash to trigger, or at least it isn't obvious that is required.
Assignee | ||
Comment 5•13 days ago
|
||
I've been looking at this but I'm not sure how it can happen. Given there are only two CDMProxy types, my assumption is that the switchover to the other state machine type fails, but the failure paths don't seem very likely. Any ideas alwu?
Assignee | ||
Comment 6•13 days ago
|
||
I'll continue to investigate in the meantime.
Assignee | ||
Comment 7•13 days ago
|
||
If the prefs were changed to something other than 1 or 2, I see that we return false, but the CDM may or not match the state machine:
https://searchfox.org/mozilla-central/rev/19764d620c02025bdcc8d1f3c4fcf5a580407a01/dom/media/ExternalEngineStateMachine.cpp#1418
Can't tell from the crash reports if this is the case for those users.
Assignee | ||
Comment 8•13 days ago
|
||
This patch makes it so that we more gracefully handle unsupported or
disabled CDM proxy types. It will diagnostic assert but barring that now
return a more appropriate error code instead of attempting to use the
CDM with the state machine.
Updated•13 days ago
|
Comment 10•12 days ago
|
||
bugherder |
Assignee | ||
Comment 11•10 days ago
|
||
Comment on attachment 9478642 [details]
Bug 1909826 - Improve MediaDecoder::SetCDMProxy to better support invalid CDM proxy types.
Beta/Release Uplift Approval Request
- User impact if declined/Reason for urgency: Fixes another MFCDM related crash
- Is this code covered by automated tests?: Yes
- Has the fix been verified in Nightly?: Yes
- Needs manual test from QE?: No
- If yes, steps to reproduce:
- List of other uplifts needed: None
- Risk to taking this patch: Low
- Why is the change risky/not risky? (and alternatives if risky): Just checks for more errors to handle them properly
- String changes made/needed:
- Is Android affected?: No
Assignee | ||
Updated•10 days ago
|
Assignee | ||
Comment 13•9 days ago
|
||
This patch makes it so that we more gracefully handle unsupported or
disabled CDM proxy types. It will diagnostic assert but barring that now
return a more appropriate error code instead of attempting to use the
CDM with the state machine.
Original Revision: https://phabricator.services.mozilla.com/D245284
Updated•9 days ago
|
Updated•9 days ago
|
Updated•9 days ago
|
Comment 14•9 days ago
|
||
uplift |
Updated•9 days ago
|
Updated•8 days ago
|
Description
•