| Summary: | Improve error handling in sendWithAsyncReply() | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Product: | WebKit | Reporter: | Chris Dumez <cdumez> | ||||||
| Component: | WebKit2 | Assignee: | Chris Dumez <cdumez> | ||||||
| Status: | RESOLVED FIXED | ||||||||
| Severity: | Normal | CC: | achristensen, darin, kkinnunen, webkit-bug-importer | ||||||
| Priority: | P2 | Keywords: | InRadar | ||||||
| Version: | WebKit Nightly Build | ||||||||
| Hardware: | Unspecified | ||||||||
| OS: | Unspecified | ||||||||
| Attachments: |
|
||||||||
|
Description
Chris Dumez
2021-10-29 14:52:19 PDT
Created attachment 442866 [details]
Patch
Comment on attachment 442866 [details]
Patch
Should we test this?
(In reply to Darin Adler from comment #2) > Comment on attachment 442866 [details] > Patch > > Should we test this? We'd need a way to keep the IPC::Connection object alive after we've invalidated it. If the IPC::Connection object gets destroyed promptly, then there is no issue since its destructor will call the async reply handlers. I noticed this when working on another patch where I mistakenly leaked a Connection object and one test was timing out because the sendWithAsyncReply() completion handler was never getting called. I will see if there is an easy way to write a test for this but it is not super trivial. Created attachment 442879 [details]
Patch
(In reply to Chris Dumez from comment #4) > Created attachment 442879 [details] > Patch I added an API test. Committed r285086 (243728@main): <https://commits.webkit.org/243728@main> All reviewed patches have been landed. Closing bug and clearing flags on attachment 442879 [details]. |