1272492 - BadIDChoice X11 errors

Status: RESOLVED WORKSFORME

(Core :: Graphics, defect, P3)

Description

[Karl Tomlinson (:karlt)]

I've observed a couple of BadIDChoice X11 errors in beta builds over the last
few months.  The first was probably February.

[3917] ###!!! ABORT: RenderCreatePicture: BadIDChoice (invalid resource ID chosen for this connection); 2 requests ago: file /mnt/ssd1/portage/www-client/firefox-45.0_beta1/work/firefox-45.0b1/toolkit/xre/nsX11ErrorHandler.cpp, line 157

[Parent 619] ###!!! ABORT: X_CreatePixmap: BadIDChoice (invalid resource ID chosen for this connection); 2 requests ago: file /mnt/ssd1/portage/www-client/firefox-46.0_beta9/work/firefox-46.0b9/toolkit/xre/nsX11ErrorHandler.cpp, line 157

I'm not the only person.  Uptimes are reasonably long.
https://crash-stats.mozilla.org/signature/?product=Firefox&app_notes=~BadIDChoice&signature=mozalloc_abort+%7C+NS_DebugBreak+%7C+X11Error&_columns=date&_columns=product&_columns=version&_columns=build_id&_columns=platform&_columns=reason&_columns=address&_columns=uptime&page=1

A couple of better crash stacks:  They are off main thread
https://crash-stats.mozilla.org/report/index/9decf629-16f5-4cc6-a427-bc8292160508
https://crash-stats.mozilla.org/report/index/bf833b12-a0f1-433c-ab0f-7caf72160504

It usually seems to be off main thread, but not always:
https://crash-stats.mozilla.org/report/index/ed3b004c-5377-4586-bb4c-c79632160509

Comment 1

[Karl Tomlinson (:karlt)]

Attachment: record resource id on X11 error

I'll run with this patch to see if I can catch what the resource id is.

If we want to land a patch, a slight more efficient approach would be to use
XSetErrorHandler to override GDK's handler and chain up.  That wouldn't be
called during gdk_x11_display_error_trap_push/pop.

Comment 2

[Karl Tomlinson (:karlt)]

A recent report in a different single thread app:
http://stackoverflow.com/questions/34748158/badidchoice-render-in-python-3-3-and-tk-tcl-displayed-on-x

Will see if I can catch X resource usage if/when this happens again.

Comment 3

[BugBot [:suhaib / :marco/ :calixte]]

Crash volume for signature 'mozalloc_abort | NS_DebugBreak | X11Error':
 - nightly (version 50): 14 crashes from 2016-06-06.
 - aurora  (version 49): 9 crashes from 2016-06-07.
 - beta    (version 48): 10 crashes from 2016-06-06.
 - release (version 47): 334 crashes from 2016-05-31.
 - esr     (version 45): 37 crashes from 2016-04-07.

Crash volume on the last weeks:
             Week N-1   Week N-2   Week N-3   Week N-4   Week N-5   Week N-6   Week N-7
 - nightly          0          0          0          1          1          7          5
 - aurora           0          0          0          0          0          9          0
 - beta             0          0          0          0          0          1          9
 - release          0          0          0          0        121        170         43
 - esr              2          3          2          1          1          6          3

Affected platform: Linux

Comment 4

[BugBot [:suhaib / :marco/ :calixte]]

Closing because no crashes reported for 12 weeks.