[Webkit-unassigned] [Bug 103128] LLInt C Loop backend should work on big endian platforms both 32 and 64 bit

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=103128

Tobias Netzel <tobias.netzel at gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|---                         |DUPLICATE

--- Comment #46 from Tobias Netzel <tobias.netzel at gmail.com> ---
(In reply to comment #45)
> I didn't know this was still unfixed upstream. We have a similar patch in
> QtWebKit to fix big-endian builds.

I just now found bug 137020. That bug got committed the same day it was filed - and this one was waiting here for years.

My patch solves quite some more issues that apparently don't exist anymore in the current trunk - but they do in the safari-600.5-branch that I'm currently stuck at.
If I ever catch up with trunk again I'll see what's still necessary and file some new bugs for that - at least I'll need compatibility with 32 bit bools.

There still are potentially interesting pieces for you like the change to Source/JavaScriptCore/llint/LLIntSlowPaths.cpp which might be needed when using CLoop only (https://bugs.webkit.org/attachment.cgi?id=248357&action=diff#a/tags/Safari-600.5.10/Source/JavaScriptCore/llint/LLIntSlowPaths.cpp_sec1). In debug builds I got occasional assertions without this.
Or the changes I made to Source/JavaScriptCore/runtime/DataView.h in order to make the DataView object behave more endian independent (https://bugs.webkit.org/attachment.cgi?id=248357&action=diff#a/tags/Safari-600.5.10/Source/JavaScriptCore/runtime/DataView.h_sec1).
Regarding my change to Source/JavaScriptCore/runtime/Arguments.h I hope it isn't necessary anymore but in case of yet more occasional assertions in JavaScriptCore it might help (https://bugs.webkit.org/attachment.cgi?id=248357&action=diff#a/tags/Safari-600.5.10/Source/JavaScriptCore/runtime/Arguments.h_sec1). Again in debug builds I got occasional assertions without this.

*** This bug has been marked as a duplicate of bug 137020 ***

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20150310/bac675a3/attachment-0002.html>