| Summary: | REGRESSION[267763] RELEASE_ASSERT in allNetworkProcesses hit when registering legacy protocol handlers before instantiating any WebKit objects | ||||||
|---|---|---|---|---|---|---|---|
| Product: | WebKit | Reporter: | Alex Christensen <achristensen> | ||||
| Component: | New Bugs | Assignee: | Alex Christensen <achristensen> | ||||
| Status: | RESOLVED FIXED | ||||||
| Severity: | Normal | CC: | beidson, cdumez, ggaren, thorton, webkit-bug-importer | ||||
| Priority: | P2 | Keywords: | InRadar | ||||
| Version: | WebKit Nightly Build | ||||||
| Hardware: | Unspecified | ||||||
| OS: | Unspecified | ||||||
| See Also: | https://bugs.webkit.org/show_bug.cgi?id=216041 | ||||||
| Attachments: |
|
||||||
|
Description
Alex Christensen
2020-09-30 12:42:24 PDT
Created attachment 410147 [details]
Patch
Comment on attachment 410147 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=410147&action=review > Source/WebKit/ChangeLog:12 > + This unfortunately can't be tested with our current infrastructure because TestsController's constructor calls initializeMainThread. I verified that removing the initializeMainThread call from TestsController::TestsController causes me to hit the issue in the radar and this fixes it. Comment on attachment 410147 [details]
Patch
Shouldn't we have initialized WebKit2 when we created the WebProcessPool? If not, what about all the other functions here, like registerURLSchemeAsCORSEnabled?
WebKit2 is initialized when making a WebProcessPool. The problem here is registerGlobalURLSchemeAsHavingCustomProtocolHandlers is static, so no WebProcessPool is instantiated. |