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.
$ hg init
Revision 0:
$ echo "unchanged" > unchanged
$ echo "remove me" > remove
$ echo "copy me" > copy
$ echo "move me" > move
$ for i in 1 2 3 4 5 6 7 8 9; do
> echo "merge ok $i" >> zzz1_merge_ok
> done
$ echo "merge bad" > zzz2_merge_bad
$ hg ci -Am "revision 0"
adding copy
adding move
adding remove
adding unchanged
adding zzz1_merge_ok
adding zzz2_merge_bad
Revision 1:
$ hg rm remove
$ hg mv move moved
$ hg cp copy copied
$ echo "added" > added
$ hg add added
$ echo "new first line" > zzz1_merge_ok
$ hg cat zzz1_merge_ok >> zzz1_merge_ok
$ echo "new last line" >> zzz2_merge_bad
$ hg ci -m "revision 1"
Local changes to revision 0:
$ hg co 0
4 files updated, 0 files merged, 3 files removed, 0 files unresolved
$ echo "new last line" >> zzz1_merge_ok
$ echo "another last line" >> zzz2_merge_bad
$ hg diff --nodates | grep "^[+-][^<>]"
--- a/zzz1_merge_ok
+++ b/zzz1_merge_ok
+new last line
--- a/zzz2_merge_bad
+++ b/zzz2_merge_bad
+another last line
$ hg st
M zzz1_merge_ok
M zzz2_merge_bad
Local merge with bad merge tool:
$ HGMERGE=false hg co
merging zzz1_merge_ok
merging zzz2_merge_bad
merging zzz2_merge_bad failed!
3 files updated, 1 files merged, 2 files removed, 1 files unresolved
use 'hg resolve' to retry unresolved file merges
[1]
$ hg resolve -m
(no more unresolved files)
$ hg co 0
merging zzz1_merge_ok
merging zzz2_merge_bad
warning: conflicts while merging zzz2_merge_bad! (edit, then use 'hg resolve --mark')
2 files updated, 1 files merged, 3 files removed, 1 files unresolved
use 'hg resolve' to retry unresolved file merges
[1]
$ hg diff --nodates | grep "^[+-][^<>]"
--- a/zzz1_merge_ok
+++ b/zzz1_merge_ok
+new last line
--- a/zzz2_merge_bad
+++ b/zzz2_merge_bad
+another last line
+=======
$ hg st
M zzz1_merge_ok
M zzz2_merge_bad
? zzz2_merge_bad.orig
Local merge with conflicts:
$ hg resolve -m
(no more unresolved files)
$ hg co
merging zzz1_merge_ok
merging zzz2_merge_bad
warning: conflicts while merging zzz2_merge_bad! (edit, then use 'hg resolve --mark')
3 files updated, 1 files merged, 2 files removed, 1 files unresolved
use 'hg resolve' to retry unresolved file merges
[1]
$ hg resolve -m
(no more unresolved files)
$ hg co 0 --config 'ui.origbackuppath=.hg/origbackups'
merging zzz1_merge_ok
merging zzz2_merge_bad
warning: conflicts while merging zzz2_merge_bad! (edit, then use 'hg resolve --mark')
2 files updated, 1 files merged, 3 files removed, 1 files unresolved
use 'hg resolve' to retry unresolved file merges
[1]
Are orig files from the last commit where we want them?
$ ls .hg/origbackups
zzz2_merge_bad.orig
$ hg diff --nodates | grep "^[+-][^<>]"
--- a/zzz1_merge_ok
+++ b/zzz1_merge_ok
+new last line
--- a/zzz2_merge_bad
+++ b/zzz2_merge_bad
+another last line
+=======
+=======
+new last line
+=======
$ hg st
M zzz1_merge_ok
M zzz2_merge_bad
? zzz2_merge_bad.orig
Local merge without conflicts:
$ hg revert zzz2_merge_bad
$ hg resolve -m
(no more unresolved files)
$ hg co
merging zzz1_merge_ok
4 files updated, 1 files merged, 2 files removed, 0 files unresolved
$ hg diff --nodates | grep "^[+-][^<>]"
--- a/zzz1_merge_ok
+++ b/zzz1_merge_ok
+new last line
$ hg st
M zzz1_merge_ok
? zzz2_merge_bad.orig