RESOLVED FIXED Bug 220469
[macOS] -[WKWebView acceptsFirstMouse:] sometimes crashes in IPC::Connection::createSyncMessageEncoder 
https://bugs.webkit.org/show_bug.cgi?id=220469
Summary [macOS] -[WKWebView acceptsFirstMouse:] sometimes crashes in IPC::Connection:...
Wenson Hsieh
Reported 2021-01-08 10:45:54 PST
<rdar://problem/72319199>
Attachments
Patch (15.72 KB, patch)
2021-01-08 11:11 PST, Wenson Hsieh 		no flags
DetailsFormatted DiffDiff
Patch (4.57 KB, patch)
2021-01-08 11:59 PST, Wenson Hsieh 		no flags
DetailsFormatted DiffDiff
Show Obsolete (1) View All Add attachment proposed patch, testcase, etc.
Wenson Hsieh
Comment 1 2021-01-08 11:11:54 PST
Created attachment 417280 [details] Patch
Chris Dumez
Comment 2 2021-01-08 11:43:06 PST
Comment on attachment 417280 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=417280&action=review > Source/WebKit/UIProcess/WebPageProxy.cpp:10346 > +bool WebPageProxy::canSendSyncMessage() const Maybe we can update sendSync() in MessageSender.h to null-check messageSenderConnection() instead of asserting it?
Wenson Hsieh
Comment 3 2021-01-08 11:44:42 PST
(In reply to Chris Dumez from comment #2) > Comment on attachment 417280 [details] > Patch > > View in context: > https://bugs.webkit.org/attachment.cgi?id=417280&action=review > > > Source/WebKit/UIProcess/WebPageProxy.cpp:10346 > > +bool WebPageProxy::canSendSyncMessage() const > > Maybe we can update sendSync() in MessageSender.h to null-check > messageSenderConnection() instead of asserting it? Yes, that would work too (I was just under the impression that we wanted to keep the assertion). I'll update the patch to just null check instead.
Chris Dumez
Comment 4 2021-01-08 11:45:57 PST
(In reply to Wenson Hsieh from comment #3) > (In reply to Chris Dumez from comment #2) > > Comment on attachment 417280 [details] > > Patch > > > > View in context: > > https://bugs.webkit.org/attachment.cgi?id=417280&action=review > > > > > Source/WebKit/UIProcess/WebPageProxy.cpp:10346 > > > +bool WebPageProxy::canSendSyncMessage() const > > > > Maybe we can update sendSync() in MessageSender.h to null-check > > messageSenderConnection() instead of asserting it? > > Yes, that would work too (I was just under the impression that we wanted to > keep the assertion). > > I'll update the patch to just null check instead. I don't think so. I think this crash just proved that this assertion does not hold since in theory you can do a sync IPC while the process is still launching.
Wenson Hsieh
Comment 5 2021-01-08 11:59:28 PST
Created attachment 417286 [details] Patch
Chris Dumez
Comment 6 2021-01-08 12:00:25 PST
Comment on attachment 417286 [details] Patch r=me, assuming the bots are happy.
EWS
Comment 7 2021-01-08 14:35:33 PST
Committed r271322: <https://trac.webkit.org/changeset/271322> All reviewed patches have been landed. Closing bug and clearing flags on attachment 417286 [details].
Note You need to log in before you can comment on or make changes to this bug.