[webkit-reviews] review canceled: [Bug 93742] HTML Parser should produce 8bit substrings for inline style and script elements : [Attachment 158598] Updated Patch using OS(DARWIN) and intrinsics

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Wed Aug 15 15:34:01 PDT 2012


Benjamin Poulain <benjamin at webkit.org> has canceled Michael Saboff
<msaboff at apple.com>'s request for review:
Bug 93742: HTML Parser should produce 8bit substrings for inline style and
script elements
https://bugs.webkit.org/show_bug.cgi?id=93742

Attachment 158598: Updated Patch using OS(DARWIN) and intrinsics
https://bugs.webkit.org/attachment.cgi?id=158598&action=review

------- Additional Comments from Benjamin Poulain <benjamin at webkit.org>
View in context: https://bugs.webkit.org/attachment.cgi?id=158598&action=review


> Source/WTF/wtf/text/ASCIIFastPath.h:114
> +    // See Intel documentation for how the shuffle control mask works.

I don't think that comment adds much :)

> Source/WTF/wtf/text/ASCIIFastPath.h:120
> +    const __m128i xmmShuf1 = {0x0e0c0a0806040200ll, 0x8080808080808080ll};
> +    const __m128i xmmShuf2 = {0x8080808080808080ll, 0x0e0c0a0806040200ll};

I think mm_setr_pi16 would be clearer for the masks.

> Source/WTF/wtf/text/ASCIIFastPath.h:137
> +		   __m128i first8LChars = _mm_shuffle_epi8(first8UChars,
xmmShuf1);
> +		   __m128i second8LChars = _mm_shuffle_epi8(second8UChars,
xmmShuf2);
> +		   __m128i all16LChars = _mm_or_si128(first8LChars,
second8LChars);

Could mm_packus_epi16 do the trick?

I clear the flag while you try regular packing.


More information about the webkit-reviews mailing list