Summary: | Move cookie flushing SPI from WKProcessPool to WKHTTPCookieStore | ||||||||
---|---|---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Alex Christensen <achristensen> | ||||||
Component: | New Bugs | Assignee: | Alex Christensen <achristensen> | ||||||
Status: | RESOLVED FIXED | ||||||||
Severity: | Normal | CC: | beidson, cdumez, webkit-bug-importer | ||||||
Priority: | P2 | Keywords: | InRadar | ||||||
Version: | WebKit Nightly Build | ||||||||
Hardware: | Unspecified | ||||||||
OS: | Unspecified | ||||||||
Attachments: |
|
Description
Alex Christensen
2020-09-14 12:49:27 PDT
Created attachment 408737 [details]
Patch
Comment on attachment 408737 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=408737&action=review > Source/WebKit/UIProcess/API/Cocoa/WKHTTPCookieStorePrivate.h:30 > - (void)_setCookieAcceptPolicy:(NSHTTPCookieAcceptPolicy)policy completionHandler:(void (^)())completionHandler WK_API_AVAILABLE(macos(WK_MAC_TBA), ios(WK_IOS_TBA)); This also needs a void inside the () to be able to be compiled from ObjC (not ObjC++). I'll add when committing this. Comment on attachment 408737 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=408737&action=review > Source/WebKit/ChangeLog:11 > + rdar://problem/68749055 tracks adoption in Safari, which I will land within an hour of landing this to avoid cookie regressions. rdar://problem/68872711 is the correct radar. Will fix before landing. Comment on attachment 408737 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=408737&action=review > Source/WebKit/NetworkProcess/cocoa/NetworkProcessCocoa.mm:228 > + platformSyncAllCookies(WTFMove(completionHandler)); Since this is now on the HTTPCookieStore (which is associated with a single data store AFAIK), don't we want to sync cookies for the cookie store 's session only instead of doing it for ALL sessions? Created attachment 408746 [details]
Patch
|