Closed Bug 1392911 Opened 8 years ago Closed 4 months ago

startup Crash @ GlyphDataElement<T>::GetExistingGlyphs

Categories

(Core :: Graphics, defect, P3)

Unspecified
Windows
defect

Tracking

()

RESOLVED WORKSFORME
Tracking Status
firefox-esr52 --- wontfix
firefox55 --- wontfix
firefox56 --- fix-optional
firefox57 --- fix-optional

People

(Reporter: wsmwk, Unassigned)

Details

(Keywords: crash, regression, regressionwindow-wanted, Whiteboard: [gfx-noted][tbird crash])

Crash Data

This is now a top crash for thunderbird 56 beta 3 56.0b3 bp-c13fe2c8-8f08-4976-99d7-3f3390170821 Windows 0 dwrite.dll GlyphDataElement<GlyphBitmapLayout, GlyphBitmapRasterizationParameters, GlyphBitmapRasterizationState>::GetExistingGlyphs(IBaseCacheContext&, CachedElementData const&, CacheReferenceList&, unsigned int, unsigned int, unsigned short const*, GlyphBitmapLayout const**) 1 dwrite.dll GlyphDataElement<GlyphBitmapLayout, GlyphBitmapRasterizationParameters, GlyphBitmapRasterizationState>::GetGlyphData(IBaseCacheContext&, CacheReferenceList&, FontFaceElement const&, GlyphBitmapRasterizationParameters const&, unsigned int, unsigned short const*, GlyphBitmapLayout const**) 2 dwrite.dll GetGlyphBitmaps(IBaseCacheContext&, FontFaceElement const&, FontRasterizer::Transform const&, unsigned char, float, unsigned short, unsigned short, FontRasterizer::RenderingFlags, unsigned short const*, unsigned int, GlyphBitmapLayout const**, CacheReferenceList&) 3 dwrite.dll GlyphRunAnalysis::GlyphRunAnalysis(ClientSideCacheContext&, FontFaceElement const&, DWRITE_FONT_SIMULATIONS, DWRITE_GLYPH_RUN const*, float, DWRITE_MATRIX const*, DWRITE_RENDERING_MODE, DWRITE_MEASURING_MODE, DWRITE_TEXT_ANTIALIAS_MODE, float, float, unsigned char*) 4 dwrite.dll DWriteGlyphRunAnalysis::DWriteGlyphRunAnalysis(DWRITE_GLYPH_RUN const*, float, DWRITE_MATRIX const*, DWRITE_RENDERING_MODE, DWRITE_MEASURING_MODE, bool, float, float, unsigned char*) 5 dwrite.dll DWriteGlyphRunAnalysis::Create(DWRITE_GLYPH_RUN const*, float, DWRITE_MATRIX const*, DWRITE_RENDERING_MODE, DWRITE_MEASURING_MODE, bool, float, float) 6 dwrite.dll DWriteFactory::CreateGlyphRunAnalysis(DWRITE_GLYPH_RUN const*, float, DWRITE_MATRIX const*, DWRITE_RENDERING_MODE, DWRITE_MEASURING_MODE, float, float, IDWriteGlyphRunAnalysis**) 7 xul.dll SkScalerContext_DW::getBoundingBox(SkGlyph*, DWRITE_RENDERING_MODE, DWRITE_TEXTURE_TYPE, tagRECT*) gfx/skia/skia/src/ports/SkScalerContext_win_dw.cpp:454 8 xul.dll SkScalerContext_DW::generateMetrics(SkGlyph*) gfx/skia/skia/src/ports/SkScalerContext_win_dw.cpp:545 9 xul.dll SkScalerContext::getMetrics(SkGlyph*) gfx/skia/skia/src/core/SkScalerContext.cpp:108 bp-c88aa4b4-d710-4592-8928-d49f70170822 [Mac] GlyphBitmapLayout::ReadFromCache
Component: General → Graphics
Product: Thunderbird → Core
Whiteboard: [gfx-noted]
reporter of these bp-bca3a35f-1783-4cb5-8889-421530170903 bp-bb4b69ac-fdd9-43bc-87bd-10e3d0170903 bp-ac43fa0c-1085-4620-aab5-4d1510170824 bp-3278394e-264f-4e4d-a18b-eafec0170824 states "sometimes, but not always, flipping the char set to unicode [from Western, in Thunderbird View, Text Encoding] can result in ability to view the email without crashing" I got a sample mail, but was unable to reproduce crashing
40+% ja locale. No longer topcrash m_kato can you make use of these crashes? bp-792a9c56-530d-4b86-954c-16e970180220 受信トレイから返信を押すと止まってしまいます。 bp-f1abcb83-bdfc-4111-9a7f-7b1fa0180225 返信できません
Flags: needinfo?(m_kato)
Whiteboard: [gfx-noted] → [gfx-noted][tbird crash]
(In reply to Wayne Mery (:wsmwk) from comment #3) > 40+% ja locale. No longer topcrash > > m_kato can you make use of these crashes? > bp-792a9c56-530d-4b86-954c-16e970180220 受信トレイから返信を押すと止まってしまいます。 > bp-f1abcb83-bdfc-4111-9a7f-7b1fa0180225 返信できません No, this crash might depend on font or font cache that uses direct write.
Flags: needinfo?(m_kato)
These seem to be reporting as EXCEPTION_IN_PAGE_ERROR_READ which means we (or the OS) mmap'd a (font?) file and then a disk error causes the OS to kill out process. https://bugzilla.mozilla.org/show_bug.cgi?id=771138#c4 gives some thoughts on this.
Looks like a user is experiencing this crash in Firefox 61 at https://support.mozilla.org/en-US/questions/1226351
QA Whiteboard: qa-not-actionable

More resent reports;
Version 89.0 = Build ID 20210527174632 (2021-05-27)

Tue, 22 Jun 2021 01:41:21 GMT
bp-1fce2514-da15-4ff1-9e4c-7d0510210622
Signature: GlyphBitmapLayout::ReadFromCache
++++++++++++++++++++++++++++++++++
Version 90.0.2 =Build ID 20210721174149 (2021-07-21)

Tue, 10 Aug 2021 19:37:03 GMT
bp-e888ef81-491c-42ae-9878-124970210810
Signature: GlyphBitmapLayout::ReadFromCache

Crashing deep inside dwrite.dll with a device error:

Crash Reason: EXCEPTION_IN_PAGE_ERROR_READ / STATUS_DEVICE_DATA_ERROR

looks to me more like there may be a faulty/unreliable disk or interface in the machine than an actionable Firefox bug.

(In reply to FredMcD from comment #7)

More resent reports;
bp-1fce2514-da15-4ff1-9e4c-7d0510210622
Signature: GlyphBitmapLayout::ReadFromCache
bp-e888ef81-491c-42ae-9878-124970210810
Signature: GlyphBitmapLayout::ReadFromCache

(In reply to Jonathan Kew (:jfkthame) from comment #8)

looks to me more like there may be a faulty/unreliable disk or interface in the machine than an actionable Firefox bug.

The user reports Microsoft ScanDisk fixed the issue (lots of bad sectors).

Since the crash volume is low (less than 5 per week), the severity is downgraded to S3. Feel free to change it back if you think the bug is still critical.

For more information, please visit auto_nag documentation.

Severity: critical → S3

only one Thunderbird crash in 6 months. good enough for WFM

Status: NEW → RESOLVED
Closed: 4 months ago
Resolution: --- → WORKSFORME
You need to log in before you can comment on or make changes to this bug.