[webkit-reviews] review denied: [Bug 66747] There is no facility for profiling how the write barrier is used : [Attachment 104801] the patch (fix added file)

[bugzilla-daemon at webkit.org]

Geoffrey Garen <ggaren at apple.com> has denied Filip Pizlo <fpizlo at apple.com>'s
request for review:
Bug 66747: There is no facility for profiling how the write barrier is used
https://bugs.webkit.org/show_bug.cgi?id=66747

Attachment 104801: the patch (fix added file)
https://bugs.webkit.org/attachment.cgi?id=104801&action=review

------- Additional Comments from Geoffrey Garen <ggaren at apple.com>
View in context: https://bugs.webkit.org/attachment.cgi?id=104801&action=review


Looks good, but I think the C++ counter is not quite right.

> Source/JavaScriptCore/runtime/WriteBarrier.h:151
>      // when some basic types aren't yet completely instantiated
>      void setEarlyValue(JSGlobalData&, const JSCell* owner, T* value)
>      {
> +#if ENABLE(WRITE_BARRIER_PROFILING)
> +	   WriteBarrierCounters::usesWithBarrierFromCpp.count();
> +#endif
>	   this->m_cell = reinterpret_cast<JSCell*>(value);
>	   Heap::writeBarrier(owner, this->m_cell);

I think you want to put the counter inside Heap::writeBarrier, and not
setEarlyValue. Heap::writeBarrier is the central funnel that will catch all C++
access.