[Webkit-unassigned] [Bug 129429] [Win32][LLINT] Crash when running JSC stress tests.

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=129429





--- Comment #33 from peavo at outlook.com  2014-03-04 14:05:21 PST ---
(In reply to comment #32)
> > Can we avoid updating the stack bounds?
> 
> I don't think so :(. On Windows, a thread can switch to another fiber at any time.
> 
> > I reverted this in the latest patch, but I see now that it introduces more stress test errors.
> > Probably because it gives different results after stack memory is precommitted.
> 
> What kind of errors?

Seems like stack overflow errors.
I have debugged when we enter the StackBounds::initialize() after precommitting,
and the result is different compared to the initial call, because we have precommitted memory.
There is now committed memory before and after the guard page, which produces different bounds result.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.