1490288 - Crash in <name omitted> | mozilla::layers::InputQueue::ConfirmDragBlock

Status: RESOLVED DUPLICATE of bug 1457603

(Core :: Panning and Zooming, defect)

Description

[Marcia Knous [:marcia]]

This bug was filed from the Socorro interface and is
report bp-4491c398-ee37-45f7-a5d9-bd0bd0180911.
=============================================================

Seen while looking at nightly crash data - 5 Mac crashes so far which started using 20180908100402: https://bit.ly/2CHaEbK

Possible regression range based on Build ID: https://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=409f3966645a651c8cb8d6d1e2f3fa995b69c739&tochange=4df1ba9c741f1cb6661d9f205ad0fedf91d31925

Top 10 frames of crashing thread:

0 XUL <name omitted> gfx/layers/apz/src/InputBlockState.cpp:72
1 XUL mozilla::layers::InputQueue::ConfirmDragBlock gfx/layers/apz/src/InputQueue.cpp:721
2 XUL mozilla::layers::APZCTreeManager::StartScrollbarDrag gfx/layers/apz/src/APZCTreeManager.cpp:793
3 XUL mozilla::detail::RunnableMethodImpl<RefPtr<mozilla::layers::APZCTreeManager>, void  xpcom/threads/nsThreadUtils.h:1178
4 XUL nsThread::ProcessNextEvent xpcom/threads/nsThread.cpp:1161
5 XUL nsBaseAppShell::NativeEventCallback xpcom/threads/nsThreadUtils.cpp:461
6 XUL nsAppShell::ProcessGeckoEvents widget/cocoa/nsAppShell.mm:461
7 CoreFoundation __CFRUNLOOP_IS_CALLING_OUT_TO_A_SOURCE0_PERFORM_FUNCTION__ 
8 CoreFoundation __CFRunLoopDoSources0 
9 CoreFoundation __CFRunLoopRun 

=============================================================

Comment 1

[Marcia Knous [:marcia]]

One comment from a web developer "Firefox Nightly crashes when I change the responsive sizes"

14 Mac crashes/12 installs in the last 7 days and is the #3 top Mac crash on nightly. ni on :dbolter to try to find a triage owner for this crash.

Comment 2

[Botond Ballo [:botond]]

(In reply to Marcia Knous [:marcia - needinfo? me] from comment #0)
> 0 XUL <name omitted> gfx/layers/apz/src/InputBlockState.cpp:72

Not sure why function name is omitted in the top stack frame, but based on the line number this is the same diagnostic assert firing as in bug 1457603.

Comment 3

[Botond Ballo [:botond]]

It would be very helpful if people running into this would post STR in bug 1457603 (it's not clear to me what "change the responsive sizes" means).