streamclone: use backgroundfilecloser (issue4889)
Closing files that have been appended to is slow on Windows/NTFS.
CloseHandle() calls on this platform often take 1-10ms - and that's
on my i7-6700K Skylake processor with a modern and fast SSD. Contrast
with other I/O operations, such as writing data, which take <100us.
This means that creating/appending thousands of files can add
significant overhead. For example, cloning mozilla-central creates
~232,000 revlog files. Assuming 1ms per CloseHandle(), that yields
232s (3:52) of wall time waiting for file closes!
The impact of this overhead can be measured most directly when applying
stream clone bundles. Applying these files is effectively uncompressing
a tar archive (read: it's very fast).
Using a RAM disk (read: no I/O wait), the difference in wall time for a
`hg debugapplystreamclonebundle` for a ~1731 MB mozilla-central bundle
between Windows and Linux from the same machine is drastic:
Linux: ~12.8s (128MB/s)
Windows: ~352.0s (4.7MB/s)
Windows is ~27.5x slower. Yikes!
After this patch:
Linux: ~12.8s (128MB/s)
Windows: ~102.1s (16.1MB/s)
Windows is now ~3.4x faster. Unfortunately, it is still ~8x slower than
Linux. Profiling reveals a few hot code paths that could likely be
improved. But those are for other patches.
This patch introduces test-clone-uncompressed.t because existing tests
of `clone --uncompressed` are scattered about and adding a variation for
background thread closing to e.g. test-http.t doesn't feel correct.
$ . "$TESTDIR/histedit-helpers.sh"
$ cat >> $HGRCPATH <<EOF
> [extensions]
> histedit=
> EOF
$ initrepo ()
> {
> hg init $1
> cd $1
> for x in a b c d e f ; do
> echo $x$x$x$x$x > $x
> hg add $x
> done
> hg ci -m 'Initial commit'
> for x in a b c d e f ; do
> echo $x > $x
> hg ci -m $x
> done
> echo 'I can haz no commute' > e
> hg ci -m 'does not commute with e'
> cd ..
> }
$ initrepo r
$ cd r
Initial generation of the command files
$ EDITED="$TESTTMP/editedhistory"
$ hg log --template 'pick {node|short} {rev} {desc}\n' -r 3 >> $EDITED
$ hg log --template 'pick {node|short} {rev} {desc}\n' -r 4 >> $EDITED
$ hg log --template 'fold {node|short} {rev} {desc}\n' -r 7 >> $EDITED
$ hg log --template 'pick {node|short} {rev} {desc}\n' -r 5 >> $EDITED
$ hg log --template 'pick {node|short} {rev} {desc}\n' -r 6 >> $EDITED
$ cat $EDITED
pick 65a9a84f33fd 3 c
pick 00f1c5383965 4 d
fold 39522b764e3d 7 does not commute with e
pick 7b4e2f4b7bcd 5 e
pick 500cac37a696 6 f
log before edit
$ hg log --graph
@ changeset: 7:39522b764e3d
| tag: tip
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: does not commute with e
|
o changeset: 6:500cac37a696
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: f
|
o changeset: 5:7b4e2f4b7bcd
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: e
|
o changeset: 4:00f1c5383965
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: d
|
o changeset: 3:65a9a84f33fd
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: c
|
o changeset: 2:da6535b52e45
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: b
|
o changeset: 1:c1f09da44841
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: a
|
o changeset: 0:1715188a53c7
user: test
date: Thu Jan 01 00:00:00 1970 +0000
summary: Initial commit
edit the history
$ hg histedit 3 --commands $EDITED 2>&1 | fixbundle
2 files updated, 0 files merged, 0 files removed, 0 files unresolved
merging e
warning: conflicts while merging e! (edit, then use 'hg resolve --mark')
Fix up the change (fold 39522b764e3d)
(hg histedit --continue to resume)
fix up
$ echo 'I can haz no commute' > e
$ hg resolve --mark e
(no more unresolved files)
continue: hg histedit --continue
$ cat > cat.py <<EOF
> import sys
> print open(sys.argv[1]).read()
> print
> print
> EOF
$ HGEDITOR="python cat.py" hg histedit --continue 2>&1 | fixbundle | grep -v '2 files removed'
2 files updated, 0 files merged, 0 files removed, 0 files unresolved
d
***
does not commute with e
HG: Enter commit message. Lines beginning with 'HG:' are removed.
HG: Leave message empty to abort commit.
HG: --
HG: user: test
HG: branch 'default'
HG: changed d
HG: changed e
2 files updated, 0 files merged, 0 files removed, 0 files unresolved
merging e
warning: conflicts while merging e! (edit, then use 'hg resolve --mark')
Fix up the change (pick 7b4e2f4b7bcd)
(hg histedit --continue to resume)
just continue this time
$ hg revert -r 'p1()' e
$ hg resolve --mark e
(no more unresolved files)
continue: hg histedit --continue
$ hg histedit --continue 2>&1 | fixbundle
7b4e2f4b7bcd: empty changeset
log after edit
$ hg log --graph
@ changeset: 5:d9cf42e54966
| tag: tip
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: f
|
o changeset: 4:10486af2e984
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: d
|
o changeset: 3:65a9a84f33fd
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: c
|
o changeset: 2:da6535b52e45
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: b
|
o changeset: 1:c1f09da44841
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: a
|
o changeset: 0:1715188a53c7
user: test
date: Thu Jan 01 00:00:00 1970 +0000
summary: Initial commit
contents of e
$ hg cat e
I can haz no commute
manifest
$ hg manifest
a
b
c
d
e
f
$ cd ..
Repeat test using "roll", not "fold". "roll" folds in changes but drops message
$ initrepo r2
$ cd r2
Initial generation of the command files
$ EDITED="$TESTTMP/editedhistory.2"
$ hg log --template 'pick {node|short} {rev} {desc}\n' -r 3 >> $EDITED
$ hg log --template 'pick {node|short} {rev} {desc}\n' -r 4 >> $EDITED
$ hg log --template 'roll {node|short} {rev} {desc}\n' -r 7 >> $EDITED
$ hg log --template 'pick {node|short} {rev} {desc}\n' -r 5 >> $EDITED
$ hg log --template 'pick {node|short} {rev} {desc}\n' -r 6 >> $EDITED
$ cat $EDITED
pick 65a9a84f33fd 3 c
pick 00f1c5383965 4 d
roll 39522b764e3d 7 does not commute with e
pick 7b4e2f4b7bcd 5 e
pick 500cac37a696 6 f
log before edit
$ hg log --graph
@ changeset: 7:39522b764e3d
| tag: tip
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: does not commute with e
|
o changeset: 6:500cac37a696
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: f
|
o changeset: 5:7b4e2f4b7bcd
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: e
|
o changeset: 4:00f1c5383965
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: d
|
o changeset: 3:65a9a84f33fd
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: c
|
o changeset: 2:da6535b52e45
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: b
|
o changeset: 1:c1f09da44841
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: a
|
o changeset: 0:1715188a53c7
user: test
date: Thu Jan 01 00:00:00 1970 +0000
summary: Initial commit
edit the history
$ hg histedit 3 --commands $EDITED 2>&1 | fixbundle
2 files updated, 0 files merged, 0 files removed, 0 files unresolved
merging e
warning: conflicts while merging e! (edit, then use 'hg resolve --mark')
Fix up the change (roll 39522b764e3d)
(hg histedit --continue to resume)
fix up
$ echo 'I can haz no commute' > e
$ hg resolve --mark e
(no more unresolved files)
continue: hg histedit --continue
$ hg histedit --continue 2>&1 | fixbundle | grep -v '2 files removed'
2 files updated, 0 files merged, 0 files removed, 0 files unresolved
2 files updated, 0 files merged, 0 files removed, 0 files unresolved
merging e
warning: conflicts while merging e! (edit, then use 'hg resolve --mark')
Fix up the change (pick 7b4e2f4b7bcd)
(hg histedit --continue to resume)
just continue this time
$ hg revert -r 'p1()' e
$ hg resolve --mark e
(no more unresolved files)
continue: hg histedit --continue
$ hg histedit --continue 2>&1 | fixbundle
7b4e2f4b7bcd: empty changeset
log after edit
$ hg log --graph
@ changeset: 5:e7c4f5d4eb75
| tag: tip
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: f
|
o changeset: 4:803d1bb561fc
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: d
|
o changeset: 3:65a9a84f33fd
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: c
|
o changeset: 2:da6535b52e45
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: b
|
o changeset: 1:c1f09da44841
| user: test
| date: Thu Jan 01 00:00:00 1970 +0000
| summary: a
|
o changeset: 0:1715188a53c7
user: test
date: Thu Jan 01 00:00:00 1970 +0000
summary: Initial commit
contents of e
$ hg cat e
I can haz no commute
manifest
$ hg manifest
a
b
c
d
e
f
description is taken from rollup target commit
$ hg log --debug --rev 4
changeset: 4:803d1bb561fceac3129ec778db9da249a3106fc3
phase: draft
parent: 3:65a9a84f33fdeb1ad5679b3941ec885d2b24027b
parent: -1:0000000000000000000000000000000000000000
manifest: 4:b068a323d969f22af1296ec6a5ea9384cef437ac
user: test
date: Thu Jan 01 00:00:00 1970 +0000
files: d e
extra: branch=default
extra: histedit_source=00f1c53839651fa5c76d423606811ea5455a79d0,39522b764e3d26103f08bd1fa2ccd3e3d7dbcf4e
description:
d
done with repo r2
$ cd ..