[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


Geoffrey Garen <ggaren at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
 Attachment #225791|review?                     |review-
               Flag|                            |




--- Comment #27 from Geoffrey Garen <ggaren at apple.com>  2014-03-04 12:07:45 PST ---
(From update of attachment 225791)
View in context: https://bugs.webkit.org/attachment.cgi?id=225791&action=review

Looking good, but I think this needs one more pass.

> Source/JavaScriptCore/runtime/VM.cpp:763
> +// when needed, see http://support.microsoft.com/kb/100775

"." at the end of the sentence, please.

> Source/JavaScriptCore/runtime/VM.cpp:778
> +#if PLATFORM(WIN) && ENABLE(LLINT)

I don't think we want to check ENABLE(LLINT) here. For example, if you disabled LLInt and enabled JIT, you would still need preCommitStackMemory. So, let's just check PLATFORM(WIN).

> Source/JavaScriptCore/runtime/VM.cpp:803
> +#if PLATFORM(WIN) && ENABLE(LLINT)
> +    if (lastStackLimit != m_stackLimit)
> +        preCommitStackMemory(m_stackLimit);
> +#endif

Looks good.

> Source/WTF/wtf/WTFThreadData.h:99
>          // We need to always get a fresh StackBounds from the OS due to how fibers work.
>          // See https://bugs.webkit.org/show_bug.cgi?id=102411
> -#if OS(WINDOWS)
> +#if OS(WINDOWS) && !ENABLE(LLINT)
>          m_stackBounds = StackBounds::currentThreadStackBounds();
>  #endif
>          return m_stackBounds;

I don't think we want to check ENABLE(LLINT) here. A Windows client can use fibers regardless of LLInt.

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