Mercurial > hg
view tests/test-narrow-exchange-merges.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 | a2a6e724d61a |
children | 7e5be4a7cda7 |
line wrap: on
line source
$ . "$TESTDIR/narrow-library.sh" create full repo $ hg init master $ cd master $ cat >> .hg/hgrc <<EOF > [narrow] > serveellipses=True > EOF $ mkdir inside $ echo 1 > inside/f $ hg commit -Aqm 'initial inside' $ mkdir outside $ echo 1 > outside/f $ hg commit -Aqm 'initial outside' $ echo 2a > outside/f $ hg commit -Aqm 'outside 2a' $ echo 3 > inside/f $ hg commit -Aqm 'inside 3' $ echo 4a > outside/f $ hg commit -Aqm 'outside 4a' $ hg update '.~3' 2 files updated, 0 files merged, 0 files removed, 0 files unresolved $ echo 2b > outside/f $ hg commit -Aqm 'outside 2b' $ echo 3 > inside/f $ hg commit -Aqm 'inside 3' $ echo 4b > outside/f $ hg commit -Aqm 'outside 4b' $ hg update '.~3' 2 files updated, 0 files merged, 0 files removed, 0 files unresolved $ echo 2c > outside/f $ hg commit -Aqm 'outside 2c' $ echo 3 > inside/f $ hg commit -Aqm 'inside 3' $ echo 4c > outside/f $ hg commit -Aqm 'outside 4c' $ hg update '.~3' 2 files updated, 0 files merged, 0 files removed, 0 files unresolved $ echo 2d > outside/f $ hg commit -Aqm 'outside 2d' $ echo 3 > inside/f $ hg commit -Aqm 'inside 3' $ echo 4d > outside/f $ hg commit -Aqm 'outside 4d' $ hg update -r 'desc("outside 4a")' 1 files updated, 0 files merged, 0 files removed, 0 files unresolved $ hg merge -r 'desc("outside 4b")' 2>&1 | egrep -v '(warning:|incomplete!)' merging outside/f 0 files updated, 0 files merged, 0 files removed, 1 files unresolved use 'hg resolve' to retry unresolved file merges or 'hg merge --abort' to abandon $ echo 5 > outside/f $ rm outside/f.orig $ hg resolve --mark outside/f (no more unresolved files) $ hg commit -m 'merge a/b 5' $ echo 6 > outside/f $ hg commit -Aqm 'outside 6' $ hg merge -r 'desc("outside 4c")' 2>&1 | egrep -v '(warning:|incomplete!)' merging outside/f 0 files updated, 0 files merged, 0 files removed, 1 files unresolved use 'hg resolve' to retry unresolved file merges or 'hg merge --abort' to abandon $ echo 7 > outside/f $ rm outside/f.orig $ hg resolve --mark outside/f (no more unresolved files) $ hg commit -Aqm 'merge a/b/c 7' $ echo 8 > outside/f $ hg commit -Aqm 'outside 8' $ hg merge -r 'desc("outside 4d")' 2>&1 | egrep -v '(warning:|incomplete!)' merging outside/f 0 files updated, 0 files merged, 0 files removed, 1 files unresolved use 'hg resolve' to retry unresolved file merges or 'hg merge --abort' to abandon $ echo 9 > outside/f $ rm outside/f.orig $ hg resolve --mark outside/f (no more unresolved files) $ hg commit -Aqm 'merge a/b/c/d 9' $ echo 10 > outside/f $ hg commit -Aqm 'outside 10' $ echo 11 > inside/f $ hg commit -Aqm 'inside 11' $ echo 12 > outside/f $ hg commit -Aqm 'outside 12' $ hg log -G -T '{rev} {node|short} {desc}\n' @ 21 8d874d57adea outside 12 | o 20 7ef88b4dd4fa inside 11 | o 19 2a20009de83e outside 10 | o 18 3ac1f5779de3 merge a/b/c/d 9 |\ | o 17 38a9c2f7e546 outside 8 | | | o 16 094aa62fc898 merge a/b/c 7 | |\ | | o 15 f29d083d32e4 outside 6 | | | | | o 14 2dc11382541d merge a/b 5 | | |\ o | | | 13 27d07ef97221 outside 4d | | | | o | | | 12 465567bdfb2d inside 3 | | | | o | | | 11 d1c61993ec83 outside 2d | | | | | o | | 10 56859a8e33b9 outside 4c | | | | | o | | 9 bb96a08b062a inside 3 | | | | | o | | 8 b844052e7b3b outside 2c |/ / / | | o 7 9db2d8fcc2a6 outside 4b | | | | | o 6 6418167787a6 inside 3 | | | +---o 5 77344f344d83 outside 2b | | | o 4 9cadde08dc9f outside 4a | | | o 3 019ef06f125b inside 3 | | | o 2 75e40c075a19 outside 2a |/ o 1 906d6c682641 initial outside | o 0 9f8e82b51004 initial inside Now narrow clone this and get a hopefully correct graph $ cd .. $ hg clone --narrow ssh://user@dummy/master narrow --include inside requesting all changes adding changesets adding manifests adding file changes added 14 changesets with 3 changes to 1 files new changesets *:* (glob) updating to branch default 1 files updated, 0 files merged, 0 files removed, 0 files unresolved $ cd narrow To make updating the tests easier, we print the emitted nodes sorted. This makes it easier to identify when the same node structure has been emitted, just in a different order. $ hg log -T '{if(ellipsis,"...")}{node|short} {p1node|short} {p2node|short} {desc}\n' | sort ...094aa62fc898 6418167787a6 bb96a08b062a merge a/b/c 7 ...2a20009de83e 019ef06f125b 3ac1f5779de3 outside 10 ...3ac1f5779de3 465567bdfb2d 094aa62fc898 merge a/b/c/d 9 ...75e40c075a19 9f8e82b51004 000000000000 outside 2a ...77344f344d83 9f8e82b51004 000000000000 outside 2b ...8d874d57adea 7ef88b4dd4fa 000000000000 outside 12 ...b844052e7b3b 9f8e82b51004 000000000000 outside 2c ...d1c61993ec83 9f8e82b51004 000000000000 outside 2d 019ef06f125b 75e40c075a19 000000000000 inside 3 465567bdfb2d d1c61993ec83 000000000000 inside 3 6418167787a6 77344f344d83 000000000000 inside 3 7ef88b4dd4fa 2a20009de83e 000000000000 inside 11 9f8e82b51004 000000000000 000000000000 initial inside bb96a08b062a b844052e7b3b 000000000000 inside 3 But seeing the graph is also nice: $ hg log -G -T '{if(ellipsis,"...")}{node|short} {desc}\n' @ ...8d874d57adea outside 12 | o 7ef88b4dd4fa inside 11 | o ...2a20009de83e outside 10 |\ | o ...3ac1f5779de3 merge a/b/c/d 9 | |\ | | o ...094aa62fc898 merge a/b/c 7 | | |\ | o | | 465567bdfb2d inside 3 | | | | | o | | ...d1c61993ec83 outside 2d | | | | | | | o bb96a08b062a inside 3 | | | | | +---o ...b844052e7b3b outside 2c | | | | | o 6418167787a6 inside 3 | | | | | o ...77344f344d83 outside 2b | |/ o | 019ef06f125b inside 3 | | o | ...75e40c075a19 outside 2a |/ o 9f8e82b51004 initial inside