Mercurial > hg
view contrib/perf.py @ 20182:04036798ebed
branches: avoid unnecessary changectx.branch() calls
This requires reading from the changelog, which can be costly over NFS.
Note that this does not totally remove reading from the changelog; we
still do that when calling changectx.closesbranch(). That call will be
removed in a later patch.
Running hg branches on the PyPy repo (with 996) over a busy NFS server,
before this change:
$ time hg --profile branches > /dev/null
CallCount Recursive Total(s) Inline(s) module:lineno(function)
2042 0 2.2827 2.2827 <open>
2036 0 0.9840 0.9840 <method 'close' of 'file' objects>
2036 0 0.0464 0.0464 <method 'read' of 'file' objects>
5233 0 0.1985 0.0453 mercurial.repoview:161(changelog)
10462 0 0.0791 0.0314 mercurial.changelog:133(tip)
5233 0 0.0388 0.0176 mercurial.localrepo:26(__get__)
10462 0 0.0250 0.0126 <len>
5233 0 0.0059 0.0039 mercurial.repoview:112(filterrevs)
10462 0 0.0029 0.0029 <hash>
2034 0 0.0444 0.0444 <method 'seek' of 'file' objects>
5340 0 0.0390 0.0390 mercurial.revlog:296(rev)
2582 0 0.0371 0.0371 <zlib.decompress>
3155 0 0.1963 0.0366 mercurial.context:202(__init__)
3155 0 0.1238 0.0306 mercurial.repoview:161(changelog)
3155 0 0.0261 0.0080 mercurial.changelog:183(rev)
9465 0 0.0061 0.0061 <isinstance>
1096 0 0.0023 0.0023 <binascii.unhexlify>
4251 0 0.0014 0.0014 <len>
2059 0 3.7341 0.0332 mercurial.changelog:270(read)
2059 0 3.6304 0.0307 mercurial.revlog:907(revision)
2057 0 0.0262 0.0137 mercurial.changelog:28(decodeextra)
4118 0 0.0094 0.0094 <method 'split' of 'str' objects>
4118 0 0.0270 0.0048 mercurial.encoding:61(tolocal)
2059 0 0.0040 0.0040 <method 'index' of 'str' objects>
10462 0 0.0791 0.0314 mercurial.changelog:133(tip)
10462 0 0.0289 0.0207 mercurial.changelog:190(node)
10462 0 0.0188 0.0091 <len>
52433 20932 0.0478 0.0310 <len>
20932 0 0.0221 0.0168 mercurial.revlog:262(__len__)
2059 0 3.6304 0.0307 mercurial.revlog:907(revision)
real 0m4.361s
user 0m0.986s
sys 0m0.237s
After this change:
$ time hg --profile branches > /dev/null
CallCount Recursive Total(s) Inline(s) module:lineno(function)
1069 0 1.1098 1.1098 <open>
1063 0 0.4865 0.4865 <method 'close' of 'file' objects>
4122 0 0.1811 0.0404 mercurial.repoview:161(changelog)
8240 0 0.0712 0.0272 mercurial.changelog:133(tip)
4122 0 0.0378 0.0177 mercurial.localrepo:26(__get__)
8240 0 0.0221 0.0115 <len>
4122 0 0.0057 0.0033 mercurial.repoview:112(filterrevs)
8240 0 0.0025 0.0025 <hash>
3029 0 0.1979 0.0371 mercurial.context:202(__init__)
3029 0 0.1278 0.0310 mercurial.repoview:161(changelog)
3029 0 0.0230 0.0081 mercurial.changelog:183(rev)
9087 0 0.0061 0.0061 <isinstance>
1096 0 0.0026 0.0026 <binascii.unhexlify>
4125 0 0.0014 0.0014 <len>
4229 0 0.0337 0.0337 mercurial.revlog:296(rev)
1061 0 0.0296 0.0296 <method 'seek' of 'file' objects>
1063 0 0.0292 0.0292 <method 'read' of 'file' objects>
8240 0 0.0712 0.0272 mercurial.changelog:133(tip)
8240 0 0.0271 0.0196 mercurial.changelog:190(node)
8240 0 0.0169 0.0083 <len>
40476 16488 0.0422 0.0271 <len>
16488 0 0.0193 0.0152 mercurial.revlog:262(__len__)
1342 0 0.0241 0.0241 <zlib.decompress>
9445 0 0.0336 0.0224 mercurial.changelog:190(node)
9445 0 0.0112 0.0112 mercurial.revlog:317(node)
1074 0 1.9102 0.0224 mercurial.changelog:270(read)
1074 0 1.8397 0.0202 mercurial.revlog:907(revision)
1073 0 0.0187 0.0099 mercurial.changelog:28(decodeextra)
2148 0 0.0061 0.0061 <method 'split' of 'str' objects>
2148 0 0.0184 0.0034 mercurial.encoding:61(tolocal)
real 0m2.402s
user 0m0.735s
sys 0m0.177s
author | Brodie Rao <brodie@sf.io> |
---|---|
date | Fri, 15 Nov 2013 23:18:08 -0500 |
parents | 74aea4be8e78 |
children | d67a7758da6d |
line wrap: on
line source
# perf.py - performance test routines '''helper extension to measure performance''' from mercurial import cmdutil, scmutil, util, commands, obsolete from mercurial import repoview, branchmap, merge, copies import time, os, sys cmdtable = {} command = cmdutil.command(cmdtable) def timer(func, title=None): results = [] begin = time.time() count = 0 while True: ostart = os.times() cstart = time.time() r = func() cstop = time.time() ostop = os.times() count += 1 a, b = ostart, ostop results.append((cstop - cstart, b[0] - a[0], b[1]-a[1])) if cstop - begin > 3 and count >= 100: break if cstop - begin > 10 and count >= 3: break if title: sys.stderr.write("! %s\n" % title) if r: sys.stderr.write("! result: %s\n" % r) m = min(results) sys.stderr.write("! wall %f comb %f user %f sys %f (best of %d)\n" % (m[0], m[1] + m[2], m[1], m[2], count)) @command('perfwalk') def perfwalk(ui, repo, *pats): try: m = scmutil.match(repo[None], pats, {}) timer(lambda: len(list(repo.dirstate.walk(m, [], True, False)))) except Exception: try: m = scmutil.match(repo[None], pats, {}) timer(lambda: len([b for a, b, c in repo.dirstate.statwalk([], m)])) except Exception: timer(lambda: len(list(cmdutil.walk(repo, pats, {})))) @command('perfannotate') def perfannotate(ui, repo, f): fc = repo['.'][f] timer(lambda: len(fc.annotate(True))) @command('perfstatus', [('u', 'unknown', False, 'ask status to look for unknown files')]) def perfstatus(ui, repo, **opts): #m = match.always(repo.root, repo.getcwd()) #timer(lambda: sum(map(len, repo.dirstate.status(m, [], False, False, # False)))) timer(lambda: sum(map(len, repo.status(**opts)))) @command('perfaddremove') def perfaddremove(ui, repo): try: oldquiet = repo.ui.quiet repo.ui.quiet = True timer(lambda: scmutil.addremove(repo, dry_run=True)) finally: repo.ui.quiet = oldquiet def clearcaches(cl): # behave somewhat consistently across internal API changes if util.safehasattr(cl, 'clearcaches'): cl.clearcaches() elif util.safehasattr(cl, '_nodecache'): from mercurial.node import nullid, nullrev cl._nodecache = {nullid: nullrev} cl._nodepos = None @command('perfheads') def perfheads(ui, repo): cl = repo.changelog def d(): len(cl.headrevs()) clearcaches(cl) timer(d) @command('perftags') def perftags(ui, repo): import mercurial.changelog import mercurial.manifest def t(): repo.changelog = mercurial.changelog.changelog(repo.sopener) repo.manifest = mercurial.manifest.manifest(repo.sopener) repo._tags = None return len(repo.tags()) timer(t) @command('perfancestors') def perfancestors(ui, repo): heads = repo.changelog.headrevs() def d(): for a in repo.changelog.ancestors(heads): pass timer(d) @command('perfancestorset') def perfancestorset(ui, repo, revset): revs = repo.revs(revset) heads = repo.changelog.headrevs() def d(): s = repo.changelog.ancestors(heads) for rev in revs: rev in s timer(d) @command('perfdirs') def perfdirs(ui, repo): dirstate = repo.dirstate 'a' in dirstate def d(): dirstate.dirs() del dirstate._dirs timer(d) @command('perfdirstate') def perfdirstate(ui, repo): "a" in repo.dirstate def d(): repo.dirstate.invalidate() "a" in repo.dirstate timer(d) @command('perfdirstatedirs') def perfdirstatedirs(ui, repo): "a" in repo.dirstate def d(): "a" in repo.dirstate._dirs del repo.dirstate._dirs timer(d) @command('perfdirstatewrite') def perfdirstatewrite(ui, repo): ds = repo.dirstate "a" in ds def d(): ds._dirty = True ds.write() timer(d) @command('perfmergecalculate', [('r', 'rev', '.', 'rev to merge against')]) def perfmergecalculate(ui, repo, rev): wctx = repo[None] rctx = scmutil.revsingle(repo, rev, rev) ancestor = wctx.ancestor(rctx) # we don't want working dir files to be stat'd in the benchmark, so prime # that cache wctx.dirty() def d(): # acceptremote is True because we don't want prompts in the middle of # our benchmark merge.calculateupdates(repo, wctx, rctx, ancestor, False, False, False, acceptremote=True) timer(d) @command('perfpathcopies', [], "REV REV") def perfpathcopies(ui, repo, rev1, rev2): ctx1 = scmutil.revsingle(repo, rev1, rev1) ctx2 = scmutil.revsingle(repo, rev2, rev2) def d(): copies.pathcopies(ctx1, ctx2) timer(d) @command('perfmanifest', [], 'REV') def perfmanifest(ui, repo, rev): ctx = scmutil.revsingle(repo, rev, rev) t = ctx.manifestnode() def d(): repo.manifest._mancache.clear() repo.manifest._cache = None repo.manifest.read(t) timer(d) @command('perfchangeset') def perfchangeset(ui, repo, rev): n = repo[rev].node() def d(): repo.changelog.read(n) #repo.changelog._cache = None timer(d) @command('perfindex') def perfindex(ui, repo): import mercurial.revlog mercurial.revlog._prereadsize = 2**24 # disable lazy parser in old hg n = repo["tip"].node() def d(): cl = mercurial.revlog.revlog(repo.sopener, "00changelog.i") cl.rev(n) timer(d) @command('perfstartup') def perfstartup(ui, repo): cmd = sys.argv[0] def d(): os.system("HGRCPATH= %s version -q > /dev/null" % cmd) timer(d) @command('perfparents') def perfparents(ui, repo): nl = [repo.changelog.node(i) for i in xrange(1000)] def d(): for n in nl: repo.changelog.parents(n) timer(d) @command('perflookup') def perflookup(ui, repo, rev): timer(lambda: len(repo.lookup(rev))) @command('perfrevrange') def perfrevrange(ui, repo, *specs): revrange = scmutil.revrange timer(lambda: len(revrange(repo, specs))) @command('perfnodelookup') def perfnodelookup(ui, repo, rev): import mercurial.revlog mercurial.revlog._prereadsize = 2**24 # disable lazy parser in old hg n = repo[rev].node() cl = mercurial.revlog.revlog(repo.sopener, "00changelog.i") def d(): cl.rev(n) clearcaches(cl) timer(d) @command('perflog', [('', 'rename', False, 'ask log to follow renames')]) def perflog(ui, repo, **opts): ui.pushbuffer() timer(lambda: commands.log(ui, repo, rev=[], date='', user='', copies=opts.get('rename'))) ui.popbuffer() @command('perfmoonwalk') def perfmoonwalk(ui, repo): """benchmark walking the changelog backwards This also loads the changelog data for each revision in the changelog. """ def moonwalk(): for i in xrange(len(repo), -1, -1): ctx = repo[i] ctx.branch() # read changelog data (in addition to the index) timer(moonwalk) @command('perftemplating') def perftemplating(ui, repo): ui.pushbuffer() timer(lambda: commands.log(ui, repo, rev=[], date='', user='', template='{date|shortdate} [{rev}:{node|short}]' ' {author|person}: {desc|firstline}\n')) ui.popbuffer() @command('perfcca') def perfcca(ui, repo): timer(lambda: scmutil.casecollisionauditor(ui, False, repo.dirstate)) @command('perffncacheload') def perffncacheload(ui, repo): s = repo.store def d(): s.fncache._load() timer(d) @command('perffncachewrite') def perffncachewrite(ui, repo): s = repo.store s.fncache._load() def d(): s.fncache._dirty = True s.fncache.write() timer(d) @command('perffncacheencode') def perffncacheencode(ui, repo): s = repo.store s.fncache._load() def d(): for p in s.fncache.entries: s.encode(p) timer(d) @command('perfdiffwd') def perfdiffwd(ui, repo): """Profile diff of working directory changes""" options = { 'w': 'ignore_all_space', 'b': 'ignore_space_change', 'B': 'ignore_blank_lines', } for diffopt in ('', 'w', 'b', 'B', 'wB'): opts = dict((options[c], '1') for c in diffopt) def d(): ui.pushbuffer() commands.diff(ui, repo, **opts) ui.popbuffer() title = 'diffopts: %s' % (diffopt and ('-' + diffopt) or 'none') timer(d, title) @command('perfrevlog', [('d', 'dist', 100, 'distance between the revisions')], "[INDEXFILE]") def perfrevlog(ui, repo, file_, **opts): from mercurial import revlog dist = opts['dist'] def d(): r = revlog.revlog(lambda fn: open(fn, 'rb'), file_) for x in xrange(0, len(r), dist): r.revision(r.node(x)) timer(d) @command('perfrevset', [('C', 'clear', False, 'clear volatile cache between each call.')], "REVSET") def perfrevset(ui, repo, expr, clear=False): """benchmark the execution time of a revset Use the --clean option if need to evaluate the impact of build volatile revisions set cache on the revset execution. Volatile cache hold filtered and obsolete related cache.""" def d(): if clear: repo.invalidatevolatilesets() repo.revs(expr) timer(d) @command('perfvolatilesets') def perfvolatilesets(ui, repo, *names): """benchmark the computation of various volatile set Volatile set computes element related to filtering and obsolescence.""" repo = repo.unfiltered() def getobs(name): def d(): repo.invalidatevolatilesets() obsolete.getrevs(repo, name) return d allobs = sorted(obsolete.cachefuncs) if names: allobs = [n for n in allobs if n in names] for name in allobs: timer(getobs(name), title=name) def getfiltered(name): def d(): repo.invalidatevolatilesets() repoview.filteredrevs(repo, name) return d allfilter = sorted(repoview.filtertable) if names: allfilter = [n for n in allfilter if n in names] for name in allfilter: timer(getfiltered(name), title=name) @command('perfbranchmap', [('f', 'full', False, 'Includes build time of subset'), ]) def perfbranchmap(ui, repo, full=False): """benchmark the update of a branchmap This benchmarks the full repo.branchmap() call with read and write disabled """ def getbranchmap(filtername): """generate a benchmark function for the filtername""" if filtername is None: view = repo else: view = repo.filtered(filtername) def d(): if full: view._branchcaches.clear() else: view._branchcaches.pop(filtername, None) view.branchmap() return d # add filter in smaller subset to bigger subset possiblefilters = set(repoview.filtertable) allfilters = [] while possiblefilters: for name in possiblefilters: subset = branchmap.subsettable.get(name) if subset not in possiblefilters: break else: assert False, 'subset cycle %s!' % possiblefilters allfilters.append(name) possiblefilters.remove(name) # warm the cache if not full: for name in allfilters: repo.filtered(name).branchmap() # add unfiltered allfilters.append(None) oldread = branchmap.read oldwrite = branchmap.branchcache.write try: branchmap.read = lambda repo: None branchmap.write = lambda repo: None for name in allfilters: timer(getbranchmap(name), title=str(name)) finally: branchmap.read = oldread branchmap.branchcache.write = oldwrite