Make Caches::writeRecord and Caches::readRecord more robust
Created attachment 419748 [details]
Crash record at https://ews-build.s3-us-west-2.amazonaws.com/macOS-Catalina-Release-WK2-Tests-EWS/r419742-829/imported/w3c/web-platform-tests/service-workers/service-worker/fetch-event-within-sw.https-crash-log.txt
Comment on attachment 419748 [details]
View in context: https://bugs.webkit.org/attachment.cgi?id=419748&action=review
> + Some events happen during teardown. No reason to crash.
This seems ok to add a null check.
The underlying issue is probably that:
- we request space, which is an asynchronous operation as we may prompt user
- somehow we are requested to clear the caches
- user grants request space and we are back with no available storage.
I think this can be fixed by bailing out early in Caches::requestSpace callback in case m_isInitialized is false.
Reverted in r279599
Reopening to attach new patch.
Created attachment 432953 [details]
Committed r279607 (239430@main): <https://commits.webkit.org/239430@main>
All reviewed patches have been landed. Closing bug and clearing flags on attachment 432953 [details].