843551 - crash [@ -[NativeMenuItemTarget menuItemHit:] ] and [ @0x0 | -[NativeMenuItemTarget menuItemHit:] ]

Status: RESOLVED WORKSFORME

(Core :: Widget: Cocoa, defect, P5)

Description

[Wayne Mery (:wsmwk)]

+++ This bug was initially created as a clone of Bug #670914 +++

crash signature is back.  Came back pretty soon after bug 670914 was fixed. But I haven't nailed down when it started back up.
[@ -[NativeMenuItemTarget menuItemHit:] ]
[@0x0 | -[NativeMenuItemTarget menuItemHit:] ]

most or all of the following crashes have email of reporter

bp-e32749be-75b9-4cd9-a6fe-b32372121026 TB14
bp-35669d59-ce6b-4414-a779-d44732121003 TB15 
bp-8a4973f3-02bf-450f-a771-e0fe52120601 TB15 alpha
bp-baabc474-a7e4-4535-9c56-db0122130208 TB19
0	XUL	-[NativeMenuItemTarget menuItemHit:]	widget/cocoa/nsMenuBarX.mm:838
1	AppKit	-[NSApplication sendAction:to:from:]	
2	AppKit	-[NSMenuItem _corePerformAction]	
3	AppKit	-[NSCarbonMenuImpl performActionWithHighlightingForItemAtIndex:]	
4	AppKit	-[NSMenu _internalPerformActionForItemAtIndex:]	
5	AppKit	-[NSCarbonMenuImpl _carbonCommandProcessEvent:handlerCallRef:]	

bp-0ce13cc2-e8d9-48ab-8745-1c7262130218 TB17
0		@0xffffff0300000088	
1	XUL	-[NativeMenuItemTarget menuItemHit:]	widget/cocoa/nsMenuBarX.mm:840
2	AppKit	-[NSApplication sendAction:to:from:]	
3	AppKit	-[NSMenuItem _corePerformAction]	
4	AppKit	-[NSCarbonMenuImpl performActionWithHighlightingForItemAtIndex:]	
5	AppKit	-[NSMenu _internalPerformActionForItemAtIndex:]	
6	AppKit	-[NSCarbonMenuImpl _carbonCommandProcessEvent:handlerCallRef:]

Comment 1

[Steven Michaud [:smichaud] (Retired)]

Please try to find when these crashes started happening again.  If they're high enough volume, that might give us a regression range.

I'm pretty sure my fix for bug 670914 did make them go away temporarily, and that something else must have changed to bring them back.

Comment 2

[Wayne Mery (:wsmwk)]

Steven, based on the following I would say the crash was never totally gone...

bug 728964 fixed in TB 11, 12, 13 circa 2012-02-29.  TB11 beta shipped March 13. two month crash count for -[NativeMenuItemTarget menuItemHit:] ending..
4/5/2012 2,700
6/5/2012   921
8/5/2012   704
10/5/2012  685
12/5/2012  773
so your patch had a definite positive impact.  so no regression based on the raw crash numbers


also, signature nsMenuUtilsX::DispatchCommandTo exists in TB 11, 12, 13 and prior. And currently exists in TB17 bp-98e5cc6c-2b29-4492-ab4a-3901c2130222
bp-0088993d-d600-43a3-8bbe-255242120704 TB13.0.1
bp-a593e8b9-c065-432d-81f3-033812120615 TB12.0.1
bp-25c58e50-c8f0-4ab1-9006-8be402120618 TB11

Comment 3

[Wayne Mery (:wsmwk)]

bp-61cf925f-7033-487b-bc78-a99082121230 "I tried to close Thunderbird, without success, several times, and tried Apple-escape. Then, Firefox died,,, that's o.k., since I mainly use Safari…."

all of the following diffrent users say  clicked option and it crashed
bp-f3c045b9-83ae-443f-81c2-a8ace2130203 
bp-7798ffd8-57bb-4081-b917-269652121228
bp-8673b7a4-608d-4251-b585-71c3d2121205
bp-3d5a019b-dee0-4d95-8847-1cf062121219

HTH

Comment 4

[Wayne Mery (:wsmwk)]

Perhaps regression per comment 1. But I'm not so sure. The crash rate did go down in the time frame that bug 670914 was fixed in release, but I think it only reduced by about 75%.  

n.b. bug 670914 comment 8

There are also firefox crashes, whose comments are similar.
bp-74681681-f8d1-484a-b81b-9c3ad2130306 quitting firefox
bp-1eb41f54-a707-4070-8bf9-691d62130213 tried to open preferences

Comment 5

[Steven Michaud [:smichaud] (Retired)]

Wayne, you may be right that these crashes never went away after my fixes for bug 704866 and (especially) bug 728964 landed.  But they did disappear entirely for at least two weeks after my patch for bug 728964 landed (see bug 670914 comment #15).

Bug 728964 was a recent regression (which was quickly fixed because it was reproducible).  But before that regression, crashes with similar signatures had been around for years, at a pretty low level.  And I'd also spent many hours trying to figure them out, without much success.

Frankly, they're not worth spending more time on unless they become huge topcrashers or some way is found to reproduce them reliably.

Comment 6

[Virtual_ManPL [:Virtual] 🇵🇱 - (please needinfo? me - so I will see your comment/reply/question/etc.)]

I'm marking this bug as WORKSFORME as bug crashlog signature didn't appear from a long time (over half year) [except some obsolete <42 versions, no crashes starting since 42 version].