Mercurial > hg
view mercurial/verify.py @ 26379:39d643252b9f
revlog: use existing file handle when reading during _addrevision
_addrevision() may need to read from revlogs as part of computing
deltas. Previously, we would flush existing file handles and open
a new, short-lived file handle to perform the reading.
If we have an existing file handle, it seems logical to reuse it
for reading instead of opening a new file handle. This patch
makes that the new behavior.
After this patch, revlog files are only reopened when adding
revisions if the revlog is switched from inline to non-inline.
On Linux when unbundling a bundle of the mozilla-central repo, this
patch has the following impact on system call counts:
Call Before After Delta
write 827,639 673,390 -154,249
open 700,103 684,089 -16,014
read 74,489 74,489 0
fstat 493,924 461,896 -32,028
close 249,131 233,117 -16,014
stat 242,001 242,001 0
lstat 18,676 18,676 0
lseek 20,268 20,268 0
ioctl 14,652 13,173 -1,479
TOTAL 3,180,758 2,930,679 -250,079
It's worth noting that many of the open() calls fail due to missing
files. That's why there are many more open() calls than close().
Despite the significant system call reduction, this change does not
seem to have a significant performance impact on Linux.
On Windows 10 (not a VM, on a SSD), this patch appears to reduce
unbundle time for mozilla-central from ~960s to ~920s. This isn't
as significant as I was hoping. But a decrease it is nonetheless.
Still, Windows unbundle performance is still >2x slower than Linux.
Despite the lack of significant gains, fewer system calls is fewer
system calls. If nothing else, this will narrow the focus of potential
areas to optimize in the future.
author | Gregory Szorc <gregory.szorc@gmail.com> |
---|---|
date | Sun, 27 Sep 2015 16:08:18 -0700 |
parents | d21d1774c73b |
children | 56b2bcea2529 |
line wrap: on
line source
# verify.py - repository integrity checking for Mercurial # # Copyright 2006, 2007 Matt Mackall <mpm@selenic.com> # # This software may be used and distributed according to the terms of the # GNU General Public License version 2 or any later version. from __future__ import absolute_import import os from .i18n import _ from .node import ( nullid, short, ) from . import ( error, revlog, util, ) def verify(repo): lock = repo.lock() try: return _verify(repo) finally: lock.release() def _normpath(f): # under hg < 2.4, convert didn't sanitize paths properly, so a # converted repo may contain repeated slashes while '//' in f: f = f.replace('//', '/') return f def _verify(repo): repo = repo.unfiltered() mflinkrevs = {} filelinkrevs = {} filenodes = {} revisions = 0 badrevs = set() errors = [0] warnings = [0] ui = repo.ui cl = repo.changelog mf = repo.manifest lrugetctx = util.lrucachefunc(repo.changectx) if not repo.url().startswith('file:'): raise util.Abort(_("cannot verify bundle or remote repos")) def err(linkrev, msg, filename=None): if linkrev is not None: badrevs.add(linkrev) else: linkrev = '?' msg = "%s: %s" % (linkrev, msg) if filename: msg = "%s@%s" % (filename, msg) ui.warn(" " + msg + "\n") errors[0] += 1 def exc(linkrev, msg, inst, filename=None): if isinstance(inst, KeyboardInterrupt): ui.warn(_("interrupted")) raise if not str(inst): inst = repr(inst) err(linkrev, "%s: %s" % (msg, inst), filename) def warn(msg): ui.warn(msg + "\n") warnings[0] += 1 def checklog(obj, name, linkrev): if not len(obj) and (havecl or havemf): err(linkrev, _("empty or missing %s") % name) return d = obj.checksize() if d[0]: err(None, _("data length off by %d bytes") % d[0], name) if d[1]: err(None, _("index contains %d extra bytes") % d[1], name) if obj.version != revlog.REVLOGV0: if not revlogv1: warn(_("warning: `%s' uses revlog format 1") % name) elif revlogv1: warn(_("warning: `%s' uses revlog format 0") % name) def checkentry(obj, i, node, seen, linkrevs, f): lr = obj.linkrev(obj.rev(node)) if lr < 0 or (havecl and lr not in linkrevs): if lr < 0 or lr >= len(cl): msg = _("rev %d points to nonexistent changeset %d") else: msg = _("rev %d points to unexpected changeset %d") err(None, msg % (i, lr), f) if linkrevs: if f and len(linkrevs) > 1: try: # attempt to filter down to real linkrevs linkrevs = [l for l in linkrevs if lrugetctx(l)[f].filenode() == node] except Exception: pass warn(_(" (expected %s)") % " ".join(map(str, linkrevs))) lr = None # can't be trusted try: p1, p2 = obj.parents(node) if p1 not in seen and p1 != nullid: err(lr, _("unknown parent 1 %s of %s") % (short(p1), short(node)), f) if p2 not in seen and p2 != nullid: err(lr, _("unknown parent 2 %s of %s") % (short(p2), short(node)), f) except Exception as inst: exc(lr, _("checking parents of %s") % short(node), inst, f) if node in seen: err(lr, _("duplicate revision %d (%d)") % (i, seen[node]), f) seen[node] = i return lr if os.path.exists(repo.sjoin("journal")): ui.warn(_("abandoned transaction found - run hg recover\n")) revlogv1 = cl.version != revlog.REVLOGV0 if ui.verbose or not revlogv1: ui.status(_("repository uses revlog format %d\n") % (revlogv1 and 1 or 0)) havecl = len(cl) > 0 havemf = len(mf) > 0 ui.status(_("checking changesets\n")) refersmf = False seen = {} checklog(cl, "changelog", 0) total = len(repo) for i in repo: ui.progress(_('checking'), i, total=total, unit=_('changesets')) n = cl.node(i) checkentry(cl, i, n, seen, [i], "changelog") try: changes = cl.read(n) if changes[0] != nullid: mflinkrevs.setdefault(changes[0], []).append(i) refersmf = True for f in changes[3]: filelinkrevs.setdefault(_normpath(f), []).append(i) except Exception as inst: refersmf = True exc(i, _("unpacking changeset %s") % short(n), inst) ui.progress(_('checking'), None) ui.status(_("checking manifests\n")) seen = {} if refersmf: # Do not check manifest if there are only changelog entries with # null manifests. checklog(mf, "manifest", 0) total = len(mf) for i in mf: ui.progress(_('checking'), i, total=total, unit=_('manifests')) n = mf.node(i) lr = checkentry(mf, i, n, seen, mflinkrevs.get(n, []), "manifest") if n in mflinkrevs: del mflinkrevs[n] else: err(lr, _("%s not in changesets") % short(n), "manifest") try: for f, fn in mf.readdelta(n).iteritems(): if not f: err(lr, _("file without name in manifest")) elif f != "/dev/null": # ignore this in very old repos filenodes.setdefault(_normpath(f), {}).setdefault(fn, lr) except Exception as inst: exc(lr, _("reading manifest delta %s") % short(n), inst) ui.progress(_('checking'), None) ui.status(_("crosschecking files in changesets and manifests\n")) total = len(mflinkrevs) + len(filelinkrevs) + len(filenodes) count = 0 if havemf: for c, m in sorted([(c, m) for m in mflinkrevs for c in mflinkrevs[m]]): count += 1 if m == nullid: continue ui.progress(_('crosschecking'), count, total=total) err(c, _("changeset refers to unknown manifest %s") % short(m)) mflinkrevs = None # del is bad here due to scope issues for f in sorted(filelinkrevs): count += 1 ui.progress(_('crosschecking'), count, total=total) if f not in filenodes: lr = filelinkrevs[f][0] err(lr, _("in changeset but not in manifest"), f) if havecl: for f in sorted(filenodes): count += 1 ui.progress(_('crosschecking'), count, total=total) if f not in filelinkrevs: try: fl = repo.file(f) lr = min([fl.linkrev(fl.rev(n)) for n in filenodes[f]]) except Exception: lr = None err(lr, _("in manifest but not in changeset"), f) ui.progress(_('crosschecking'), None) ui.status(_("checking files\n")) storefiles = set() for f, f2, size in repo.store.datafiles(): if not f: err(None, _("cannot decode filename '%s'") % f2) elif size > 0 or not revlogv1: storefiles.add(_normpath(f)) fncachewarned = False files = sorted(set(filenodes) | set(filelinkrevs)) total = len(files) for i, f in enumerate(files): ui.progress(_('checking'), i, item=f, total=total) try: linkrevs = filelinkrevs[f] except KeyError: # in manifest but not in changelog linkrevs = [] if linkrevs: lr = linkrevs[0] else: lr = None try: fl = repo.file(f) except error.RevlogError as e: err(lr, _("broken revlog! (%s)") % e, f) continue for ff in fl.files(): try: storefiles.remove(ff) except KeyError: warn(_(" warning: revlog '%s' not in fncache!") % ff) fncachewarned = True checklog(fl, f, lr) seen = {} rp = None for i in fl: revisions += 1 n = fl.node(i) lr = checkentry(fl, i, n, seen, linkrevs, f) if f in filenodes: if havemf and n not in filenodes[f]: err(lr, _("%s not in manifests") % (short(n)), f) else: del filenodes[f][n] # verify contents try: l = len(fl.read(n)) rp = fl.renamed(n) if l != fl.size(i): if len(fl.revision(n)) != fl.size(i): err(lr, _("unpacked size is %s, %s expected") % (l, fl.size(i)), f) except error.CensoredNodeError: # experimental config: censor.policy if ui.config("censor", "policy", "abort") == "abort": err(lr, _("censored file data"), f) except Exception as inst: exc(lr, _("unpacking %s") % short(n), inst, f) # check renames try: if rp: if lr is not None and ui.verbose: ctx = lrugetctx(lr) found = False for pctx in ctx.parents(): if rp[0] in pctx: found = True break if not found: warn(_("warning: copy source of '%s' not" " in parents of %s") % (f, ctx)) fl2 = repo.file(rp[0]) if not len(fl2): err(lr, _("empty or missing copy source revlog %s:%s") % (rp[0], short(rp[1])), f) elif rp[1] == nullid: ui.note(_("warning: %s@%s: copy source" " revision is nullid %s:%s\n") % (f, lr, rp[0], short(rp[1]))) else: fl2.rev(rp[1]) except Exception as inst: exc(lr, _("checking rename of %s") % short(n), inst, f) # cross-check if f in filenodes: fns = [(lr, n) for n, lr in filenodes[f].iteritems()] for lr, node in sorted(fns): err(lr, _("%s in manifests not found") % short(node), f) ui.progress(_('checking'), None) for f in storefiles: warn(_("warning: orphan revlog '%s'") % f) ui.status(_("%d files, %d changesets, %d total revisions\n") % (len(files), len(cl), revisions)) if warnings[0]: ui.warn(_("%d warnings encountered!\n") % warnings[0]) if fncachewarned: ui.warn(_('hint: run "hg debugrebuildfncache" to recover from ' 'corrupt fncache\n')) if errors[0]: ui.warn(_("%d integrity errors encountered!\n") % errors[0]) if badrevs: ui.warn(_("(first damaged changeset appears to be %d)\n") % min(badrevs)) return 1