Mercurial > jhg

comparison design.txt @ 200:114c9fe7b643

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Performance optimization: reduce memory ParentWalker hogs
author Artem Tikhomirov <tikhomirov.artem@gmail.com>
date Wed, 20 Apr 2011 21:14:51 +0200
parents 33a7d76f067b
children 706bcc7cfee4
comparison
equal deleted inserted replaced
199:f4fa4456fa50 200:114c9fe7b643
101 101
102 IndexEntry(int offset, int baseRevision) got replaced with int[] arrays (offsets - optional) 102 IndexEntry(int offset, int baseRevision) got replaced with int[] arrays (offsets - optional)
103 for 69338 revisions from cpython repo 1109408 bytes reduced to 277368 bytes with the new int[] version. 103 for 69338 revisions from cpython repo 1109408 bytes reduced to 277368 bytes with the new int[] version.
104 I.e. total for changelog+manifest is 1,5 Mb+ gain 104 I.e. total for changelog+manifest is 1,5 Mb+ gain
105 105
106 ParentWalker got arrays (Nodeid[] and int[]) instead of HashMap/LinkedHashSet. This change saves, per revision:
107 was: LinkedHashSet$Entry:32 + HashMap$Entry:24 + HashMap.entries[]:4 (in fact, up to 8, given entries size is power of 2, and 69000+
108 elements in cpython test repo resulted in entries[131072].
109 total: (2 HashMaps) 32+(24+4)*2 = 88 bytes
110 now: Nodeid[]:4 , int[]:4 bytes per entry. arrays of exact revlog size
111 total: (4 Nodeid[], 1 int[]) 4*4 + 4 = 20 bytes
112 for cpython test repo with 69338 revisions, 1 387 224 instead of 4 931 512 bytes. Mem usage (TaskManager) ~50 Mb when 10000 revs read
113
106 <<<<< 114 <<<<<
107 115
108 Tests: 116 Tests:
109 DataAccess - readBytes(length > memBufferSize, length*2 > memBufferSize) - to check impl is capable to read huge chunks of data, regardless of own buffer size 117 DataAccess - readBytes(length > memBufferSize, length*2 > memBufferSize) - to check impl is capable to read huge chunks of data, regardless of own buffer size
110 118