If an NPObject invoke() call returns false, something went wrong and an exception should be thrown in V8. Currently we fail silently. Error message should probably match Firefox ("Error calling method on NPObject!").
This issue appears to affect JSC also, so I'm fixing it there too.
Created attachment 41132 [details] patch + layout test
V8 part looks good.
Comment on attachment 41132 [details] patch + layout test Thanks for the patch.
Comment on attachment 41132 [details] patch + layout test Let commit bot land it
Comment on attachment 41132 [details] patch + layout test Clearing flags on attachment: 41132 Committed r49796: <http://trac.webkit.org/changeset/49796>
All reviewed patches have been landed. Closing bug.
Test fails on Snow Leopard: http://build.webkit.org/results/SnowLeopard%20Intel%20Release%20(Tests)/r49797%20(1179)/results.html We need to come up with a fix (or skip on SL) or roll this out.
(In reply to comment #8) > Test fails on Snow Leopard: > http://build.webkit.org/results/SnowLeopard%20Intel%20Release%20(Tests)/r49797%20(1179)/results.html > > We need to come up with a fix (or skip on SL) or roll this out. I don't think I'm going to be able to fix this immediately (not least because I don't have SL), so I'm happy to either skip or roll out, whichever people prefer.
(In reply to comment #9) > (In reply to comment #8) > > Test fails on Snow Leopard: > > http://build.webkit.org/results/SnowLeopard%20Intel%20Release%20(Tests)/r49797%20(1179)/results.html > > > > We need to come up with a fix (or skip on SL) or roll this out. > > I don't think I'm going to be able to fix this immediately (not least because I > don't have SL), so I'm happy to either skip or roll out, whichever people > prefer. Per conversation in #webkit, closing this bug. See SL-specific issue at https://bugs.webkit.org/show_bug.cgi?id=30528.