Summary: | [ Catalina WK2 Release ] http/tests/IndexedDB/storage-limit-1.https.html is a flaky failure | ||||||||
---|---|---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Truitt Savell <tsavell> | ||||||
Component: | New Bugs | Assignee: | Sihui Liu <sihui_liu> | ||||||
Status: | RESOLVED FIXED | ||||||||
Severity: | Normal | CC: | alecflett, beidson, ews-watchlist, ggaren, jsbell, sihui_liu, webkit-bot-watchers-bugzilla, webkit-bug-importer | ||||||
Priority: | P2 | Keywords: | InRadar | ||||||
Version: | WebKit Nightly Build | ||||||||
Hardware: | Unspecified | ||||||||
OS: | Unspecified | ||||||||
Attachments: |
|
Description
Truitt Savell
2020-06-17 07:32:23 PDT
Marked this test as failing while it is investigated: https://trac.webkit.org/changeset/263148/webkit Created attachment 425317 [details]
Patch
Comment on attachment 425317 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=425317&action=review > Source/WebCore/Modules/indexeddb/server/SQLiteIDBBackingStore.cpp:1275 > + String dbFilename = fullDatabasePath(); > + if (FileSystem::fileExists(dbFilename)) > + return 0; Was this supposed to be !FileSystem::fileExists()? I believe the behavior of creating the file and then checking its version would return the most up-to-date version, rather than zero, right? Should we mimic that behavior here, instead of returning zero? (In reply to Geoffrey Garen from comment #4) > Comment on attachment 425317 [details] > Patch > > View in context: > https://bugs.webkit.org/attachment.cgi?id=425317&action=review > > > Source/WebCore/Modules/indexeddb/server/SQLiteIDBBackingStore.cpp:1275 > > + String dbFilename = fullDatabasePath(); > > + if (FileSystem::fileExists(dbFilename)) > > + return 0; > > Was this supposed to be !FileSystem::fileExists()? ah right! will change > > I believe the behavior of creating the file and then checking its version > would return the most up-to-date version, rather than zero, right? Should we > mimic that behavior here, instead of returning zero? The up-to-date version number is stored in the database file. If it does not exist, we don't have the version number. Our old behavior is creating a new db, setting its version to currentMetadataVersion (1), and returning its version. We can directly returning 1 in this case, but 0 seems to be what other browsers use Created attachment 425326 [details]
Patch
Comment on attachment 425326 [details]
Patch
r=me
> > Was this supposed to be !FileSystem::fileExists()? > > ah right! will change Looks like tests caught this too. Nice. > Our old behavior is creating a new db, setting its version to > currentMetadataVersion (1), and returning its version. We can directly > returning 1 in this case, but 0 seems to be what other browsers use Sounds good -- nice to match other browsers. Committed r275583: <https://commits.webkit.org/r275583> All reviewed patches have been landed. Closing bug and clearing flags on attachment 425326 [details]. |