[Webkit-unassigned] [Bug 125280] New: REGRESSION(r134984): Fix readonly and type attribute selector incorrect style sharing

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Wed Dec 4 23:11:31 PST 2013


https://bugs.webkit.org/show_bug.cgi?id=125280

           Summary: REGRESSION(r134984): Fix readonly and type attribute
                    selector incorrect style sharing
           Product: WebKit
           Version: 528+ (Nightly build)
          Platform: Unspecified
        OS/Version: Unspecified
            Status: NEW
          Keywords: BlinkMergeCandidate
          Severity: Normal
          Priority: P2
         Component: CSS
        AssignedTo: webkit-unassigned at lists.webkit.org
        ReportedBy: rniwa at webkit.org
                CC: koivisto at iki.fi, simon.fraser at apple.com,
                    akling at apple.com


Merge https://chromium.googlesource.com/chromium/blink/+/b857a7d6c5813dc5d9f10c7eece0ac3d58afeafe
or come up with a better fix

RuleSet has a special case in isCommonAttributeSelectorAttribute for
allowing style sharing on elements that match "type" or "readonly"
attribute selectors as long as they have the same value. This was
added long ago in r96393 and worked based on the assumption that
there was logic in canShareStyleWithElement that actually checked
the value of the type and readonly attributes on the style sharing
candidates.

Later in r123730 the type attribute handling was regressed when it
was moved to only check the value for <input>. The readonly attribute
was later regressed in the same way in r134984.

This patch restores the checks those two patches moved but retains the
switch from getAttribute to fastGetAttribute for HTML which was likely
most of the performance improvement and fixes the bug where elements
would incorrectly share style even if one of them matched an attribute
selector.

Unfortunately this does mean that we need to call the slow geAttribute
in SVG documents which may be a minor regression. We can work to
improve the performance of SVG in the future once we get the
correctness back.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.



More information about the webkit-unassigned mailing list