Mon, 28 Jan 2013 19:05:35 +0900 parsers: fix memleak of revlog cache entries on strip stable
Yuya Nishihara <yuya@tcha.org> [Mon, 28 Jan 2013 19:05:35 +0900] rev 18504
parsers: fix memleak of revlog cache entries on strip Since 12a852c7c763, raw_length can be reduced on strip, but corresponding cache entries still have refcount. They are not dereferenced by _index_clearcache(), and never freed. To reproduce the problem, run "hg pull" and "hg strip null" several times in the same process.
Wed, 30 Jan 2013 17:32:17 +0100 hgweb: fix navigation label (issue3792) stable
Pierre-Yves David <pierre-yves.david@logilab.fr> [Wed, 30 Jan 2013 17:32:17 +0100] rev 18503
hgweb: fix navigation label (issue3792) Latest refactoring (6da1e979340a) was buggy and used a variable from another loop. Tests are run on repo too small to cache that.
Tue, 29 Jan 2013 20:03:51 +0100 profiling: add documentation of lsprof 'sort' and 'nested' stable
Mads Kiilerich <madski@unity3d.com> [Tue, 29 Jan 2013 20:03:51 +0100] rev 18502
profiling: add documentation of lsprof 'sort' and 'nested'
Tue, 29 Jan 2013 17:01:41 +0100 OS X: try cheap ascii .lower() in normcase before making full unicode dance stable
Mads Kiilerich <madski@unity3d.com> [Tue, 29 Jan 2013 17:01:41 +0100] rev 18501
OS X: try cheap ascii .lower() in normcase before making full unicode dance This is similar to what is done in encoding.lower, introduced in c481761033bd. This has been seen making 'hg up' and 'hg st' in a 50000+ files repo 13% faster. This might make Mercurial slightly slower for users who mainly use non-ASCII filenames. That is a reasonable trade-off.
(0) -10000 -3000 -1000 -300 -100 -30 -10 -4 +4 +10 +30 +100 +300 +1000 +3000 +10000 +30000 tip