IndexedDB: Simplify transaction timers and event tracking
Created attachment 175536 [details] Patch
NOTE: http://crrev.com/11280128 needs to land first. (There some Chromium public WebKit API cleanup that can be done after this lands, as well as some cleanup inside Chromium itself.)
dgrogan@, alecflett@ - please take a look once Tryptophan levels have diminished to a reasonable level?
Created attachment 175541 [details] Patch
Updated patch - removed the Chromium public WebKit API impls. Just leaves an ASSERT_NOT_REACHED() impl in WebIDBTransaction.h
Now that turkey day has passed - alecflett@, dgrogan@ - please take a look?
Comment on attachment 175541 [details] Patch Other than a comment request, this looks great. LGTM View in context: https://bugs.webkit.org/attachment.cgi?id=175541&action=review > Source/WebCore/Modules/indexeddb/IDBRequest.cpp:504 > if (m_transaction) { Now I'm wondering when we do and don't have an m_transaction - do you know? If so, add a comment..
(In reply to comment #7) > (From update of attachment 175541 [details]) > Other than a comment request, this looks great. LGTM > > View in context: https://bugs.webkit.org/attachment.cgi?id=175541&action=review > > > Source/WebCore/Modules/indexeddb/IDBRequest.cpp:504 > > if (m_transaction) { > > Now I'm wondering when we do and don't have an m_transaction - do you know? If so, add a comment.. The IDBFactory methods (open/deleteDatabase/getDatabaseNames) are the only cases. I'll add a comment.
Created attachment 176298 [details] Patch
tony@ - r?
Created attachment 176343 [details] Patch for landing
Comment on attachment 176343 [details] Patch for landing Clearing flags on attachment: 176343 Committed r135927: <http://trac.webkit.org/changeset/135927>
All reviewed patches have been landed. Closing bug.
26% improvement: http://build.chromium.org/f/chromium/perf/chromium-rel-win7-webkit/indexeddb/report.html?rev=170079&graph=basic_insert&trace=t&history=150