[Webkit-unassigned] [Bug 196533] [META] Undefined behavior bugs

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=196533

--- Comment #11 from Filip Pizlo <fpizlo at apple.com> ---
Simple question: has anyone perf tested JSC with -fwrapv?  I believe both clang and gcc support it.

If that is perf-neutral, then we won't have to fix any more signed overflow bugs, ever.

I'm not sure I understand the argument that says that running a sanitizer and fixing those bugs one-off is superior to just simply adding one compiler flag.

If that is successful, maybe we can just ask for an option in clang to mask shift amounts.  I could write an llvm pass to do that in an afternoon.

-- 
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/20190411/902d1c9f/attachment.html>