Bug 48552 – Mouse events never reach Flash in WebKit2 on Mac (asserts in Debug builds)

• Status: RESOLVED
• Resolution: FIXED
• Component: WebKit2

[Adam Roben (:aroben)]

To reproduce:

1. Go to http://youtube.com/ and try to interact with any video

The video doesn't respond to any mouse events. In Debug builds, you'll get an assertion in ArgumentDecoder::decodeBytes:

ASSERT(size == bufferSize);

size is 56, bufferSize is 52

[Adam Roben (:aroben)]

This happens when handling a HandleMouseEnterEvent. Presumably the issue is that WebMouseEvent is 52 bytes in 32-bit builds, and 56 bytes in 64-bit builds. Perhaps this is due to different padding?

[Adam Roben (:aroben)]

<rdar://problem/8606571>

[Adam Roben (:aroben)]

Seems like we should add a WebEvent.cpp, or perhaps even WebMouseEvent.cpp/h!

[Adam Roben (:aroben)]

Created attachment 72415 [details]
Stop using encode/decodeBytes for WebEvent subclasses

[Anders Carlsson]

Comment on attachment 72415 [details]
Stop using encode/decodeBytes for WebEvent subclasses

View in context: https://bugs.webkit.org/attachment.cgi?id=72415&action=review

> WebKit2/Shared/WebKeyboardEvent.cpp:60
> +                                t.m_nativeVirtualKeyCode, t.m_isAutoRepeat, t.m_isKeypad, t.m_isSystemKey));

Not sure if it is, but this should all be on one line!

[Early Warning System Bot]

Attachment 72415 [details] did not build on qt:
Build output: http://queues.webkit.org/results/4853082

[Adam Roben (:aroben)]

Committed r70954: <http://trac.webkit.org/changeset/70954>

[Csaba Osztrogonác]

(In reply to comment #6)
> Attachment 72415 [details] did not build on qt:
> Build output: http://queues.webkit.org/results/4853082

Guys, it wasn't a fair-play game ... Qt EWS was red, but
you committed the wrong patch, and the Qt build is broken 
since 1,5 hours ago. :-S

[Csaba Osztrogonác]

(In reply to comment #8)
> (In reply to comment #6)
> > Attachment 72415 [details] [details] did not build on qt:
> > Build output: http://queues.webkit.org/results/4853082
> 
> Guys, it wasn't a fair-play game ... Qt EWS was red, but
> you committed the wrong patch, and the Qt build is broken 
> since 1,5 hours ago. :-S

Kling fixed it in http://trac.webkit.org/changeset/70964

[Adam Roben (:aroben)]

(In reply to comment #9)
> (In reply to comment #8)
> > (In reply to comment #6)
> > > Attachment 72415 [details] [details] [details] did not build on qt:
> > > Build output: http://queues.webkit.org/results/4853082
> > 
> > Guys, it wasn't a fair-play game ... Qt EWS was red, but
> > you committed the wrong patch, and the Qt build is broken 
> > since 1,5 hours ago. :-S
> 
> Kling fixed it in http://trac.webkit.org/changeset/70964

Sorry for the trouble. I've been paying attention to the Qt-EWS in general, and find it extremely useful. I guess I wasn't thinking on this one. Thanks for cleaning up my mess!