| commit | 0f86eb4d63c92bf2bbf9de5f73fa2d429454412e | |
| author | Jan Kundrát <jkt@kde.org> | |
| Sun, 12 Apr 2015 23:59:34 +0000 (13 01:59 +0200) | ||
| committer | Jan Kundrát <jkt@kde.org> | |
| Wed, 22 Apr 2015 11:35:42 +0000 (22 13:35 +0200) | ||
| tree | bfe620783de8282b2d11dab6d6e240a805f40027 | treesnapshot (tar.gz zip) |
| parent | f062be94ae8e1503599f9a53e7c27fe8cb86296e | commitdiff |
Threading: Reserve enough space for QHash
Found by Intel's V-Tune, thanks. A non-scientific test suggests savings
of up to 10% in testThreadingPerformance/testSortingPerformance, but I
suspect that I'm doing something wrong here. Anyway, given that the
final size of the hash table is known in advance, there's no point in
not making use of it, IMHO.
Change-Id: Id3d2bf1f9e91f4cda6bf4da4a59bf6c0e20e1834
Found by Intel's V-Tune, thanks. A non-scientific test suggests savings
of up to 10% in testThreadingPerformance/testSortingPerformance, but I
suspect that I'm doing something wrong here. Anyway, given that the
final size of the hash table is known in advance, there's no point in
not making use of it, IMHO.
Change-Id: Id3d2bf1f9e91f4cda6bf4da4a59bf6c0e20e1834
| src/Imap/Model/ThreadingMsgListModel.cpp | diffblobblamehistory |