Bug 23829

Summary:

Make createElement non-virtual

Product:

WebKit

Reporter:

Julien Chaffraix <jchaffraix>

Component:

DOM

Assignee:

Julien Chaffraix <jchaffraix>

Status:

RESOLVED WONTFIX

Severity:

Normal

Priority:

Version:

528+ (Nightly build)

Hardware:

All

OS:

All

Attachments:

Description	Flags
Proposed fix: inline HTMLDocument::createElement into Document's	ap: review-

Julien Chaffraix

Reported 2009-02-08 01:07:57 PST

As mentioned in https://bugs.webkit.org/show_bug.cgi?id=14506#c9, it would be nice to remove the forked version in HTMLDocument. It is the only other specialization so we can make createElement non-virtual afterwards. Patch forthcoming.

Attachments
Proposed fix: inline HTMLDocument::createElement into Document's (5.17 KB, patch) 2009-02-08 01:32 PST, Julien Chaffraix	ap: review-	Details Formatted Diff Diff
View All Add attachment proposed patch, testcase, etc.

Julien Chaffraix

Comment 1 2009-02-08 01:32:33 PST

Created attachment 27461 [details] Proposed fix: inline HTMLDocument::createElement into Document's

Alexey Proskuryakov

Comment 2 2009-02-09 04:48:14 PST

Comment on attachment 27461 [details] Proposed fix: inline HTMLDocument::createElement into Document's r=me

Alexey Proskuryakov

Comment 3 2009-02-09 04:56:35 PST

Comment on attachment 27461 [details] Proposed fix: inline HTMLDocument::createElement into Document's I've just had a second thought after noticing that isHTMLDocument() is virtual. This patch doesn't cut back on virtual dispatches in most cases, but adds a branch, so I don't see how it is an improvement. Taking back the r=me, sorry.

Julien Chaffraix

Comment 4 2009-03-17 13:21:19 PDT

After a second thought, it goes against some of the latest changes. Also making HTMLDocument() non virtual to offset the cost of the introduced branch is not really an amelioration in term of code cleanliness. Closing this bug as WONTFIX.

Note You need to log in before you can comment on or make changes to this bug.