[webkit-reviews] review canceled: [Bug 81332] Cache support for OpenTypeVerticalData : [Attachment 137552] reflected changes from Kenichi's review (removed bit-field)

[bugzilla-daemon at webkit.org]

Koji Ishii <kojiishi at gmail.com> has canceled Koji Ishii <kojiishi at gmail.com>'s
request for review:
Bug 81332: Cache support for OpenTypeVerticalData
https://bugs.webkit.org/show_bug.cgi?id=81332

Attachment 137552: reflected changes from Kenichi's review (removed bit-field)
https://bugs.webkit.org/attachment.cgi?id=137552&action=review

------- Additional Comments from Koji Ishii <kojiishi at gmail.com>
Sorry for back and forth but after more thoughts, I found one of my claim was
invalid.　There are 3 cases here:
1. No leading "@"
2. One leading "@"
3. Two or more leading "@"
99% of the cases run the code for 1. Within the rest of 1%, 99% run the code
for 2.
So adding machine cycles to case 1 in order to make case 3 faster doesn't make
sense, but adding to case 2 isn't a big deal, especially if it can save from
stack overflow on really huge number of leading "@".
I was mistaken between case 1　and 2, sorry about that, and now the new patch
uses a loop within case 2 to remove the leading "@". Thank you for the review
to find this.