##// END OF EJS Templates
exchange: improve computation of relevant markers for large repos...
exchange: improve computation of relevant markers for large repos Compute the candidate nodes with relevant markers directly from keys of the predecessors/successors/children dictionaries of obsstore. This is faster than iterating over all nodes directly. This test could be further improved for repositories with relative few markers compared to the repository size, but this is no longer hot already. With the current loop structure, the obshashrange use works as well as before as it passes lists with a single node. Adjust the interface by allowing revision lists as well as node lists. This helps cases that computes ancestors as it reduces the materialisation cost. Use this in _pushdiscoveryobsmarker and _getbundleobsmarkerpart. Improve the latter further by directly using ancestors(). Performance benchmarks show notable and welcome improvement to no-op push and pull (that would also apply to other push/pull). This apply to push and pull done without evolve. ### push/pull Benchmark parameter # bin-env-vars.hg.flavor = default # benchmark.variants.explicit-rev = none # benchmark.variants.protocol = ssh # benchmark.variants.revs = none ## benchmark.name = hg.command.pull # data-env-vars.name = mercurial-devel-2024-03-22-zstd-sparse-revlog before: 5.968537 seconds after: 5.668507 seconds (-5.03%, -0.30) # data-env-vars.name = tryton-devel-2024-03-22-zstd-sparse-revlog before: 1.446232 seconds after: 0.835553 seconds (-42.23%, -0.61) # data-env-vars.name = netbsd-src-draft-2024-09-19-zstd-sparse-revlog before: 5.777412 seconds after: 2.523454 seconds (-56.32%, -3.25) ## benchmark.name = hg.command.push # data-env-vars.name = mercurial-devel-2024-03-22-zstd-sparse-revlog before: 6.155501 seconds after: 5.885072 seconds (-4.39%, -0.27) # data-env-vars.name = tryton-devel-2024-03-22-zstd-sparse-revlog before: 1.491054 seconds after: 0.934882 seconds (-37.30%, -0.56) # data-env-vars.name = netbsd-src-draft-2024-09-19-zstd-sparse-revlog before: 5.902494 seconds after: 2.957644 seconds (-49.89%, -2.94) There is not notable different in these result using the "rust" flavor instead of the "default". The performance impact on the same operation when using evolve were also tested and no impact was noted.

File last commit:

r52756:f4733654 default
r52789:8583d138 default
Show More
osutil.py
115 lines | 3.8 KiB | text/x-python | PythonLexer
# osutil.py - CFFI version of osutil.c
#
# Copyright 2016 Maciej Fijalkowski <fijall@gmail.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 annotations
import os
import stat as statmod
from ..pure.osutil import *
from .. import pycompat
if pycompat.isdarwin:
from . import _osutil # pytype: disable=import-error
ffi = _osutil.ffi
lib = _osutil.lib
listdir_batch_size = 4096
# tweakable number, only affects performance, which chunks
# of bytes do we get back from getattrlistbulk
attrkinds = [None] * 20 # we need the max no for enum VXXX, 20 is plenty
attrkinds[lib.VREG] = statmod.S_IFREG
attrkinds[lib.VDIR] = statmod.S_IFDIR
attrkinds[lib.VLNK] = statmod.S_IFLNK
attrkinds[lib.VBLK] = statmod.S_IFBLK
attrkinds[lib.VCHR] = statmod.S_IFCHR
attrkinds[lib.VFIFO] = statmod.S_IFIFO
attrkinds[lib.VSOCK] = statmod.S_IFSOCK
class stat_res:
def __init__(self, st_mode, st_mtime, st_size):
self.st_mode = st_mode
self.st_mtime = st_mtime
self.st_size = st_size
tv_sec_ofs = ffi.offsetof("struct timespec", "tv_sec")
buf = ffi.new("char[]", listdir_batch_size)
def listdirinternal(dfd, req, stat, skip):
ret = []
while True:
r = lib.getattrlistbulk(dfd, req, buf, listdir_batch_size, 0)
if r == 0:
break
if r == -1:
raise OSError(ffi.errno, os.strerror(ffi.errno))
cur = ffi.cast("val_attrs_t*", buf)
for i in range(r):
lgt = cur.length
assert lgt == ffi.cast('uint32_t*', cur)[0]
ofs = cur.name_info.attr_dataoffset
str_lgt = cur.name_info.attr_length
base_ofs = ffi.offsetof('val_attrs_t', 'name_info')
name = bytes(
ffi.buffer(
ffi.cast("char*", cur) + base_ofs + ofs, str_lgt - 1
)
)
tp = attrkinds[cur.obj_type]
if name == b"." or name == b"..":
continue
if skip == name and tp == statmod.S_ISDIR:
return []
if stat:
mtime = cur.mtime.tv_sec
mode = (cur.accessmask & ~lib.S_IFMT) | tp
ret.append(
(
name,
tp,
stat_res(
st_mode=mode,
st_mtime=mtime,
st_size=cur.datalength,
),
)
)
else:
ret.append((name, tp))
cur = ffi.cast(
"val_attrs_t*", int(ffi.cast("intptr_t", cur)) + lgt
)
return ret
def listdir(path, stat=False, skip=None):
req = ffi.new("struct attrlist*")
req.bitmapcount = lib.ATTR_BIT_MAP_COUNT
req.commonattr = (
lib.ATTR_CMN_RETURNED_ATTRS
| lib.ATTR_CMN_NAME
| lib.ATTR_CMN_OBJTYPE
| lib.ATTR_CMN_ACCESSMASK
| lib.ATTR_CMN_MODTIME
)
req.fileattr = lib.ATTR_FILE_DATALENGTH
dfd = lib.open(path, lib.O_RDONLY, 0)
if dfd == -1:
raise OSError(ffi.errno, os.strerror(ffi.errno))
try:
ret = listdirinternal(dfd, req, stat, skip)
finally:
try:
lib.close(dfd)
except BaseException:
pass # we ignore all the errors from closing, not
# much we can do about that
return ret