RESOLVED FIXED Bug 233698
Persist NavigationPreloadState in service worker registration database
https://bugs.webkit.org/show_bug.cgi?id=233698
Summary Persist NavigationPreloadState in service worker registration database
youenn fablet
Reported 2021-12-01 05:13:55 PST
Persist NavigationPreloadState in service worker registration database
Attachments
Patch (23.63 KB, patch)
2021-12-01 05:29 PST, youenn fablet
no flags
Patch (24.20 KB, patch)
2021-12-02 00:13 PST, youenn fablet
ews-feeder: commit-queue-
Patch (25.60 KB, patch)
2021-12-02 01:41 PST, youenn fablet
no flags
Patch for landing (25.12 KB, patch)
2021-12-02 07:44 PST, youenn fablet
no flags
Fixing path in API tests (28.65 KB, patch)
2021-12-03 05:22 PST, youenn fablet
no flags
youenn fablet
Comment 1 2021-12-01 05:29:38 PST
youenn fablet
Comment 2 2021-12-02 00:13:38 PST
youenn fablet
Comment 3 2021-12-02 01:41:03 PST
Radar WebKit Bug Importer
Comment 4 2021-12-02 02:50:01 PST
Chris Dumez
Comment 5 2021-12-02 07:16:32 PST
Comment on attachment 445691 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=445691&action=review > Source/WebCore/workers/service/NavigationPreloadState.h:36 > + static NavigationPreloadState defaultValue() { return { false, "true"_s }; } Maybe we should do `String headerValue { "true"_s };` below and use the default constructor instead of this defaultValue()? Or do we need to distinguish default-constructed objects somehow? > Source/WebCore/workers/service/server/RegistrationStore.h:43 > +struct NavigationPreloadState; Seems unnecessary? It isn't used in this header.
youenn fablet
Comment 6 2021-12-02 07:42:17 PST
(In reply to Chris Dumez from comment #5) > Comment on attachment 445691 [details] > Patch > > View in context: > https://bugs.webkit.org/attachment.cgi?id=445691&action=review > > > Source/WebCore/workers/service/NavigationPreloadState.h:36 > > + static NavigationPreloadState defaultValue() { return { false, "true"_s }; } > > Maybe we should do `String headerValue { "true"_s };` below and use the > default constructor instead of this defaultValue()? Or do we need to > distinguish default-constructed objects somehow? We use default constructed NavigationPreloadState when installing context data. It is better to keep it smaller since we are IPCing. Or we could add a specific InstallingServiceWorkerContextData without this parameter. > > > Source/WebCore/workers/service/server/RegistrationStore.h:43 > > +struct NavigationPreloadState; > > Seems unnecessary? It isn't used in this header. Right, forgot to remove it.
youenn fablet
Comment 7 2021-12-02 07:44:55 PST
Created attachment 445720 [details] Patch for landing
youenn fablet
Comment 8 2021-12-03 05:22:11 PST
Created attachment 445845 [details] Fixing path in API tests
EWS
Comment 9 2021-12-03 06:01:22 PST
Committed r286488 (244827@main): <https://commits.webkit.org/244827@main> All reviewed patches have been landed. Closing bug and clearing flags on attachment 445845 [details].
Note You need to log in before you can comment on or make changes to this bug.