IndexedDB 2.0: Send operations off to the server in batches instead of one at a time.
authorbeidson@apple.com <beidson@apple.com@268f45cc-cd09-0410-ab3c-d52691b4dbfc>
Wed, 30 Nov 2016 22:59:28 +0000 (22:59 +0000)
committerbeidson@apple.com <beidson@apple.com@268f45cc-cd09-0410-ab3c-d52691b4dbfc>
Wed, 30 Nov 2016 22:59:28 +0000 (22:59 +0000)
commit75e4453e1526f5b93a2d15bd4de759b7fa65e0d1
tree34fc1ba980517dc51a3792153a1cee5f348c3b68
parent517f0bad55d1b2ee8978bd9566075a0d268ff7cc
IndexedDB 2.0: Send operations off to the server in batches instead of one at a time.
https://bugs.webkit.org/show_bug.cgi?id=165221

Reviewed by Myles C. Maxfield.

No new tests (No observable behavior change).

On a profile of "Lots of writes to an IndexedDB", timer scheduling/firing presented as over 10% of the time.

This patch negates much of that.

* Modules/indexeddb/IDBTransaction.cpp:
(WebCore::IDBTransaction::pendingOperationTimerFired): Send 100 operations off to the server
  before spinning the runloop, instead of only 1.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@209157 268f45cc-cd09-0410-ab3c-d52691b4dbfc
Source/WebCore/ChangeLog
Source/WebCore/Modules/indexeddb/IDBTransaction.cpp