Bug 134327 – [iOS][WK2] Protect WKWebView's delayUpdateVisibleContentRects in case of reentry

• Status: RESOLVED
• Resolution: LATER
• Component: New Bugs

[Benjamin Poulain]

[iOS][WK2] Protect WKWebView's delayUpdateVisibleContentRects in case of reentry

[Benjamin Poulain]

Created attachment 233870 [details]
Patch

[Simon Fraser (smfr)]

Comment on attachment 233870 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=233870&action=review

> Source/WebKit2/UIProcess/API/Cocoa/WKWebView.mm:188
> +    unsigned _delayUpdateVisibleContentRectsCounter;

int please to make underflow less devastating?

> Source/WebKit2/UIProcess/API/Cocoa/WKWebView.mm:636
> +    --_delayUpdateVisibleContentRectsCounter;

Please assert that it's > 0

> Source/WebKit2/UIProcess/API/Cocoa/WKWebView.mm:-731
> -    _delayUpdateVisibleContentRects = NO;

Set to 0 here?

[Build Bot]

Comment on attachment 233870 [details]
Patch

Attachment 233870 [details] did not pass mac-wk2-ews (mac-wk2):
Output: http://webkit-queues.appspot.com/results/6471168152305664

New failing tests:
media/W3C/audio/canPlayType/canPlayType_application_octet_stream_with_codecs_1.html

[Build Bot]

Created attachment 233886 [details]
Archive of layout-test-results from webkit-ews-12 for mac-mountainlion-wk2

The attached test failures were seen while running run-webkit-tests on the mac-wk2-ews.
Bot: webkit-ews-12  Port: mac-mountainlion-wk2  Platform: Mac OS X 10.8.5

[Benjamin Poulain]

I'll keep this out of the tree. We haven't discovered the case where this would be necessary and the boolean is safer than the counter (the side effect is too many updates)