Summary: | Pack bitfields in InlineBox for Windows | ||||||
---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Ryosuke Niwa <rniwa> | ||||
Component: | New Bugs | Assignee: | Ryosuke Niwa <rniwa> | ||||
Status: | RESOLVED FIXED | ||||||
Severity: | Normal | CC: | darin, eric, kling, koivisto, krit, mitz, morrita, ojan, tkent, tony, webkit.review.bot | ||||
Priority: | P2 | ||||||
Version: | 528+ (Nightly build) | ||||||
Hardware: | Unspecified | ||||||
OS: | Unspecified | ||||||
Attachments: |
|
Description
Ryosuke Niwa
2012-03-29 00:44:39 PDT
Created attachment 134519 [details]
Initial attempt
Comment on attachment 134519 [details]
Initial attempt
Looks good.
Comment on attachment 134519 [details] Initial attempt Clearing flags on attachment: 134519 Committed r112512: <http://trac.webkit.org/changeset/112512> All reviewed patches have been landed. Closing bug. Sadly, this change appears to have had no effect on page cycler results: http://build.chromium.org/p/chromium.webkit/builders/Vista%20Perf/builds/9443 Do we know from COMPILE_ASSERTS that we definitely made these smaller on Win? (In reply to comment #6) > Do we know from COMPILE_ASSERTS that we definitely made these smaller on Win? Yes. The InlineBox's size was reduced from 40 bytes to 36 bytes. So that's about 10% reduction in its size. Maybe the size of InlineBox doesn't matter as much. It's possible that we're over-optimzing in this area. Of course, it might also mean that we don't have good tests. |