[webkit-reviews] review denied: [Bug 37734] Small HashSet::remove optimization : [Attachment 53579] idea

[bugzilla-daemon at webkit.org]

Alexey Proskuryakov <ap at webkit.org> has denied Jędrzej Nowacki
<jedrzej.nowacki at nokia.com>'s request for review:
Bug 37734: Small HashSet::remove optimization
https://bugs.webkit.org/show_bug.cgi?id=37734

Attachment 53579: idea
https://bugs.webkit.org/attachment.cgi?id=53579&action=review

------- Additional Comments from Alexey Proskuryakov <ap at webkit.org>
+#if CHECK_HASHTABLE_CONSISTENCY
	 if (it.m_impl == m_impl.end())
	     return;
+#endif

What is the reason to keep this code when CHECK_HASHTABLE_CONSISTENCY is
enabled?

I don't know what the original idea for sharing responsibilities between
HashSet and implementation classes was, but it seems more logical to remove the
check from HashTable::removeWithoutEntryConsistencyCheck. All callers currently
have it.

> There are autotests for this code, that I can use?

Regression tests for WebKit are executed by run-webkit-tests script, please see
<http://webkit.org/coding/contributing.html> for more detail.

r- for the lack of ChangeLog