Mercurial > hg
view tests/test-mq-qclone-http.t @ 38732:be4984261611
merge: mark file gets as not thread safe (issue5933)
In default installs, this has the effect of disabling the thread-based
worker on Windows when manifesting files in the working directory. My
measurements have shown that with revlog-based repositories, Mercurial
spends a lot of CPU time in revlog code resolving file data. This ends
up incurring a lot of context switching across threads and slows down
`hg update` operations when going from an empty working directory to
the tip of the repo.
On mozilla-unified (246,351 files) on an i7-6700K (4+4 CPUs):
before: 487s wall
after: 360s wall (equivalent to worker.enabled=false)
cpus=2: 379s wall
Even with only 2 threads, the thread pool is still slower.
The introduction of the thread-based worker (02b36e860e0b) states that
it resulted in a "~50%" speedup for `hg sparse --enable-profile` and
`hg sparse --disable-profile`. This disagrees with my measurement
above. I theorize a few reasons for this:
1) Removal of files from the working directory is I/O - not CPU - bound
and should benefit from a thread pool (unless I/O is insanely fast
and the GIL release is near instantaneous). So tests like `hg sparse
--enable-profile` may exercise deletion throughput and aren't good
benchmarks for worker tasks that are CPU heavy.
2) The patch was authored by someone at Facebook. The results were
likely measured against a repository using remotefilelog. And I
believe that revision retrieval during working directory updates with
remotefilelog will often use a remote store, thus being I/O and not
CPU bound. This probably resulted in an overstated performance gain.
Since there appears to be a need to enable the thread-based worker with
some stores, I've made the flagging of file gets as thread safe
configurable. I've made it experimental because I don't want to formalize
a boolean flag for this option and because this attribute is best
captured against the store implementation. But we don't have a proper
store API for this yet. I'd rather cross this bridge later.
It is possible there are revlog-based repositories that do benefit from
a thread-based worker. I didn't do very comprehensive testing. If there
are, we may want to devise a more proper algorithm for whether to use
the thread-based worker, including possibly config options to limit the
number of threads to use. But until I see evidence that justifies
complexity, simplicity wins.
Differential Revision: https://phab.mercurial-scm.org/D3963
author | Gregory Szorc <gregory.szorc@gmail.com> |
---|---|
date | Wed, 18 Jul 2018 09:49:34 -0700 |
parents | b4b7427b5786 |
children |
line wrap: on
line source
hide outer repo $ hg init $ echo "[extensions]" >> $HGRCPATH $ echo "mq=" >> $HGRCPATH $ mkdir webdir $ cd webdir $ hg init a $ hg --cwd a qinit -c $ echo a > a/a $ hg --cwd a ci -A -m a adding a $ echo b > a/b $ hg --cwd a addremove adding b $ hg --cwd a qnew -f b.patch $ hg --cwd a qcommit -m b.patch $ hg --cwd a log --template "{desc}\n" [mq]: b.patch a $ hg --cwd a/.hg/patches log --template "{desc}\n" b.patch $ root=`pwd` $ cd .. test with recursive collection $ cat > collections.conf <<EOF > [paths] > /=$root/** > EOF $ hg serve -p $HGPORT -d --pid-file=hg.pid --webdir-conf collections.conf \ > -A access-paths.log -E error-paths-1.log $ cat hg.pid >> $DAEMON_PIDS $ get-with-headers.py localhost:$HGPORT '?style=raw' 200 Script output follows /a/ /a/.hg/patches/ $ hg qclone http://localhost:$HGPORT/a b requesting all changes adding changesets adding manifests adding file changes added 2 changesets with 2 changes to 2 files new changesets cb9a9f314b8b:184916345baa requesting all changes adding changesets adding manifests adding file changes added 1 changesets with 3 changes to 3 files new changesets 4052ceaa8c4e updating to branch default 3 files updated, 0 files merged, 0 files removed, 0 files unresolved 1 files updated, 0 files merged, 0 files removed, 0 files unresolved $ hg --cwd b log --template "{desc}\n" a $ hg --cwd b qpush -a applying b.patch now at: b.patch $ hg --cwd b log --template "{desc}\n" imported patch b.patch a test with normal collection $ cat > collections1.conf <<EOF > [paths] > /=$root/* > EOF $ hg serve -p $HGPORT1 -d --pid-file=hg.pid --webdir-conf collections1.conf \ > -A access-paths.log -E error-paths-1.log $ cat hg.pid >> $DAEMON_PIDS $ get-with-headers.py localhost:$HGPORT1 '?style=raw' 200 Script output follows /a/ /a/.hg/patches/ $ hg qclone http://localhost:$HGPORT1/a c requesting all changes adding changesets adding manifests adding file changes added 2 changesets with 2 changes to 2 files new changesets cb9a9f314b8b:184916345baa requesting all changes adding changesets adding manifests adding file changes added 1 changesets with 3 changes to 3 files new changesets 4052ceaa8c4e updating to branch default 3 files updated, 0 files merged, 0 files removed, 0 files unresolved 1 files updated, 0 files merged, 0 files removed, 0 files unresolved $ hg --cwd c log --template "{desc}\n" a $ hg --cwd c qpush -a applying b.patch now at: b.patch $ hg --cwd c log --template "{desc}\n" imported patch b.patch a test with old-style collection $ cat > collections2.conf <<EOF > [collections] > $root=$root > EOF $ hg serve -p $HGPORT2 -d --pid-file=hg.pid --webdir-conf collections2.conf \ > -A access-paths.log -E error-paths-1.log $ cat hg.pid >> $DAEMON_PIDS $ get-with-headers.py localhost:$HGPORT2 '?style=raw' 200 Script output follows /a/ /a/.hg/patches/ $ hg qclone http://localhost:$HGPORT2/a d requesting all changes adding changesets adding manifests adding file changes added 2 changesets with 2 changes to 2 files new changesets cb9a9f314b8b:184916345baa requesting all changes adding changesets adding manifests adding file changes added 1 changesets with 3 changes to 3 files new changesets 4052ceaa8c4e updating to branch default 3 files updated, 0 files merged, 0 files removed, 0 files unresolved 1 files updated, 0 files merged, 0 files removed, 0 files unresolved $ hg --cwd d log --template "{desc}\n" a $ hg --cwd d qpush -a applying b.patch now at: b.patch $ hg --cwd d log --template "{desc}\n" imported patch b.patch a test --mq works and uses correct repository config $ hg --cwd d outgoing --mq comparing with http://localhost:$HGPORT2/a/.hg/patches searching for changes no changes found [1] $ hg --cwd d log --mq --template '{rev} {desc|firstline}\n' 0 b.patch $ killdaemons.py