[webkit-reviews] review denied: [Bug 12931] ARM crash due to non-aligned memory access : [Attachment 13433] Updated per comments

bugzilla-request-daemon at macosforge.org bugzilla-request-daemon at macosforge.org
Thu Mar 1 10:21:36 PST 2007


Darin Adler <darin at apple.com> has denied Darin Adler <darin at apple.com>'s
request for review:
Bug 12931: ARM crash due to non-aligned memory access
http://bugs.webkit.org/show_bug.cgi?id=12931

Attachment 13433: Updated per comments
http://bugs.webkit.org/attachment.cgi?id=13433&action=edit

------- Additional Comments from Darin Adler <darin at apple.com>
Looks fine.

+	 const uint16_t* strChars = reinterpret_cast<const
uint16_t*>(str->characters());
+	 const uint16_t* bufChars = reinterpret_cast<const uint16_t*>(buf.s);

These should just be const UChar* instead of const uint16_t*, and then you
won't need reinterpret_cast.

I think in the future we might want to reverse the if and enable the fast
32-bit-at-a-time path only on certain platforms. The ARM case can be the normal
case.

In fact, I don't think the ARM side needs a comment. Instead the comment should
be on the 32-bit side. Something like:

    // On platforms where it's safe to do so, it's faster to go 4 bytes at a
time.

review- because of the unneeded reinterpret_cast.



More information about the webkit-reviews mailing list