Open Bug 1429514 Opened 6 years ago Updated 2 years ago

Assertion failed at /builds/worker/workspace/build/src/gfx/cairo/cairo/src/cairo-hash.c:196: hash_table->live_entries == 0

Categories

(Core :: Graphics, defect, P3)

59 Branch
defect

Tracking

()

Tracking Status
firefox59 --- affected

People

(Reporter: jkratzer, Unassigned)

References

(Blocks 1 open bug)

Details

(Keywords: assertion, testcase, Whiteboard: [gfx-noted])

Attachments

(2 files)

Attached file trigger.html
Non-fatal assertion found while fuzzing mozilla-central rev d5f42a23909e.
Flags: in-testsuite?
Attached file output.log
I reproduced in a debug build based off of e4107773cffb:

> [Parent 4561, Main Thread] WARNING: Fonts still alive while shutting down gfxFontCache: 'mFonts.Count() == 0', file /moz/gecko-dev/gfx/thebes/gfxFont.cpp, line 217
> Assertion failed at /moz/gecko-dev/gfx/cairo/cairo/src/cairo-hash.c:196: hash_table->live_entries == 0

I also hit a font related crash which may or not be related:

> firefox: fccache.c:534: FcCacheFini: Assertion `fcCacheChains[i] == NULL' failed.
Priority: -- → P3
Whiteboard: [gfx-noted]
Preliminary investigation shows this is not because of a leak of the font itself, but because there is still an nsPresContext holding an nsDeviceContext which points down the chain nsFontCache -> nsFontMetrics -> gfxFontGroup which hangs onto our gfxFont, so gfxFontCache can never expire it.

I'm not really familiar with nsPresContext/nsDeviceContext or what pins that in place, so can't really recommend how to triage this better, but I'm not sure it's a Graphics issue, per se. Jonathan, ideas?
Flags: needinfo?(jfkthame)
This looks a lot like bug 1384585; see discussion there. I don't have any fresh ideas on how to get to the root of it, though.
Flags: needinfo?(jfkthame)
See Also: → 1712410
Severity: normal → S3
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: