[webkit-reviews] review denied: [Bug 102828] r134080 causes heap problem on linux systems where PAGESIZE != 4096 : [Attachment 175259] proposed patch

[bugzilla-daemon at webkit.org]

Mark Hahnenberg <mhahnenberg at apple.com> has denied Balazs Kilvady
<kilvadyb at homejinni.com>'s request for review:
Bug 102828: r134080 causes heap problem on linux systems where PAGESIZE != 4096
https://bugs.webkit.org/show_bug.cgi?id=102828

Attachment 175259: proposed patch
https://bugs.webkit.org/attachment.cgi?id=175259&action=review

------- Additional Comments from Mark Hahnenberg <mhahnenberg at apple.com>
View in context: https://bugs.webkit.org/attachment.cgi?id=175259&action=review


I'm somewhat confused as to why a different page size makes things segfault.
Where exactly did it segfault? I wouldn't be opposed to making the Region size
variable, e.g. 16 * WTF::pageSize() or something like that.

> Source/JavaScriptCore/heap/MarkStack.cpp:48
> +#if OS(LINUX) && COMPILER(GCC)

Instead of putting a bunch of nasty ifdefs everywhere, maybe you could use
WTF::pageSize()?