[Webkit-unassigned] [Bug 230241] Fix crash in 32 bits due to not enough scratch registers available

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=230241

--- Comment #9 from Caio Lima <ticaiolima at gmail.com> ---
Comment on attachment 438171
  --> https://bugs.webkit.org/attachment.cgi?id=438171
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=438171&action=review

I like this version much more. I’m just a bit afraid we are giving up on cache on cases where we still could do it. Could you run some benchmarks to verify if we are not regressing anything?

> Source/JavaScriptCore/bytecode/AccessCase.cpp:140
> +        requiredRegisters *= 2;

This is a bit conservative, because we can have `baseRegs` be only payload register, since we know somehow it’s a cell. IIRC, such info will be available on stubInfo.

-- 
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/20210915/dba5f73b/attachment-0001.htm>