changelog: disable delta chains
This patch disables delta chains on changelogs. After this patch, new
entries on changelogs - including existing changelogs - will be stored
as the fulltext of that data (likely compressed). No delta computation
will be performed.
An overview of delta chains and data justifying this change follows.
Revlogs try to store entries as a delta against a previous entry (either
a parent revision in the case of generaldelta or the previous physical
revision when not using generaldelta). Most of the time this is the
correct thing to do: it frequently results in less CPU usage and smaller
storage.
Delta chains are most effective when the base revision being deltad
against is similar to the current data. This tends to occur naturally
for manifests and file data, since only small parts of each tend to
change with each revision. Changelogs, however, are a different story.
Changelog entries represent changesets/commits. And unless commits in a
repository are homogonous (same author, changing same files, similar
commit messages, etc), a delta from one entry to the next tends to be
relatively large compared to the size of the entry. This means that
delta chains tend to be short. How short? Here is the full vs delta
revision breakdown on some real world repos:
Repo % Full % Delta Max Length
hg 45.8 54.2 6
mozilla-central 42.4 57.6 8
mozilla-unified 42.5 57.5 17
pypy 46.1 53.9 6
python-zstandard 46.1 53.9 3
(I threw in python-zstandard as an example of a repo that is homogonous.
It contains a small Python project with changes all from the same
author.)
Contrast this with the manifest revlog for these repos, where 99+% of
revisions are deltas and delta chains run into the thousands.
So delta chains aren't as useful on changelogs. But even a short delta
chain may provide benefits. Let's measure that.
Delta chains may require less CPU to read revisions if the CPU time
spent reading smaller deltas is less than the CPU time used to
decompress larger individual entries. We can measure this via
`hg perfrevlog -c -d 1` to iterate a revlog to resolve each revision's
fulltext. Here are the results of that command on a repo using delta
chains in its changelog and on a repo without delta chains:
hg (forward)
! wall 0.407008 comb 0.410000 user 0.410000 sys 0.000000 (best of 25)
! wall 0.390061 comb 0.390000 user 0.390000 sys 0.000000 (best of 26)
hg (reverse)
! wall 0.515221 comb 0.520000 user 0.520000 sys 0.000000 (best of 19)
! wall 0.400018 comb 0.400000 user 0.390000 sys 0.010000 (best of 25)
mozilla-central (forward)
! wall 4.508296 comb 4.490000 user 4.490000 sys 0.000000 (best of 3)
! wall 4.370222 comb 4.370000 user 4.350000 sys 0.020000 (best of 3)
mozilla-central (reverse)
! wall 5.758995 comb 5.760000 user 5.720000 sys 0.040000 (best of 3)
! wall 4.346503 comb 4.340000 user 4.320000 sys 0.020000 (best of 3)
mozilla-unified (forward)
! wall 4.957088 comb 4.950000 user 4.940000 sys 0.010000 (best of 3)
! wall 4.660528 comb 4.650000 user 4.630000 sys 0.020000 (best of 3)
mozilla-unified (reverse)
! wall 6.119827 comb 6.110000 user 6.090000 sys 0.020000 (best of 3)
! wall 4.675136 comb 4.670000 user 4.670000 sys 0.000000 (best of 3)
pypy (forward)
! wall 1.231122 comb 1.240000 user 1.230000 sys 0.010000 (best of 8)
! wall 1.164896 comb 1.160000 user 1.160000 sys 0.000000 (best of 9)
pypy (reverse)
! wall 1.467049 comb 1.460000 user 1.460000 sys 0.000000 (best of 7)
! wall 1.160200 comb 1.170000 user 1.160000 sys 0.010000 (best of 9)
The data clearly shows that it takes less wall and CPU time to resolve
revisions when there are no delta chains in the changelogs, regardless
of the direction of traversal. Furthermore, not using a delta chain
means that fulltext resolution in reverse is as fast as iterating
forward. So not using delta chains on the changelog is a clear CPU win
for reading operations.
An example of a user-visible operation showing this speed-up is revset
evaluation. Here are results for
`hg perfrevset 'author(gps) or author(mpm)'`:
hg
! wall 1.655506 comb 1.660000 user 1.650000 sys 0.010000 (best of 6)
! wall 1.612723 comb 1.610000 user 1.600000 sys 0.010000 (best of 7)
mozilla-central
! wall 17.629826 comb 17.640000 user 17.600000 sys 0.040000 (best of 3)
! wall 17.311033 comb 17.300000 user 17.260000 sys 0.040000 (best of 3)
What about 00changelog.i size?
Repo Delta Chains No Delta Chains
hg 7,033,250 6,976,771
mozilla-central 82,978,748 81,574,623
mozilla-unified 88,112,349 86,702,162
pypy 20,740,699 20,659,741
The data shows that removing delta chains from the changelog makes the
changelog smaller.
Delta chains are also used during changegroup generation. This
operation essentially converts a series of revisions to one large
delta chain. And changegroup generation is smart: if the delta in
the revlog matches what the changegroup is emitting, it will reuse
the delta instead of recalculating it. We can measure the impact
removing changelog delta chains has on changegroup generation via
`hg perfchangegroupchangelog`:
hg
! wall 1.589245 comb 1.590000 user 1.590000 sys 0.000000 (best of 7)
! wall 1.788060 comb 1.790000 user 1.790000 sys 0.000000 (best of 6)
mozilla-central
! wall 17.382585 comb 17.380000 user 17.340000 sys 0.040000 (best of 3)
! wall 20.161357 comb 20.160000 user 20.120000 sys 0.040000 (best of 3)
mozilla-unified
! wall 18.722839 comb 18.720000 user 18.680000 sys 0.040000 (best of 3)
! wall 21.168075 comb 21.170000 user 21.130000 sys 0.040000 (best of 3)
pypy
! wall 4.828317 comb 4.830000 user 4.820000 sys 0.010000 (best of 3)
! wall 5.415455 comb 5.420000 user 5.410000 sys 0.010000 (best of 3)
The data shows eliminating delta chains makes the changelog part of
changegroup generation slower. This is expected since we now have to
compute deltas for revisions where we could recycle the delta before.
It is worth putting this regression into context of overall changegroup
times. Here is the rough total CPU time spent in changegroup generation
for various repos while using delta chains on the changelog:
Repo CPU Time (s) CPU Time w/ compression
hg 4.50 7.05
mozilla-central 111.1 222.0
pypy 28.68 75.5
Before compression, removing delta chains from the changegroup adds
~4.4% overhead to hg changegroup generation, 1.3% to mozilla-central,
and 2.0% to pypy. When you factor in zlib compression, these percentages
are roughly divided by 2.
While the increased CPU usage for changegroup generation is unfortunate,
I think it is acceptable because the percentage is small, server
operators (those likely impacted most by this) have other mechanisms
to mitigate CPU consumption (namely reducing zlib compression level and
pre-generated clone bundles), and because there is room to optimize this
in the future. For example, we could use the nullid as the base revision,
effectively encoding the full revision for each entry in the changegroup.
When doing this, `hg perfchangegroupchangelog` nearly halves:
mozilla-unified
! wall 21.168075 comb 21.170000 user 21.130000 sys 0.040000 (best of 3)
! wall 11.196461 comb 11.200000 user 11.190000 sys 0.010000 (best of 3)
This looks very promising as a future optimization opportunity.
It's worth that the changes in test-acl.t to the changegroup part size.
This is because revision 6 in the changegroup had a delta chain of
length 2 before and after this patch the base revision is nullrev.
When the base revision is nullrev, cg2packer.deltaparent() hardcodes
the *previous* revision from the changegroup as the delta parent.
This caused the delta in the changegroup to switch base revisions,
the delta to change, and the size to change accordingly. While the
size increased in this case, I think sizes will remain the same
on average, as the delta base for changelog revisions doesn't matter
too much (as this patch shows). So, I don't consider this a regression.
#require git
make git commits repeatable
$ echo "[core]" >> $HOME/.gitconfig
$ echo "autocrlf = false" >> $HOME/.gitconfig
$ GIT_AUTHOR_NAME='test'; export GIT_AUTHOR_NAME
$ GIT_AUTHOR_EMAIL='test@example.org'; export GIT_AUTHOR_EMAIL
$ GIT_AUTHOR_DATE='1234567891 +0000'; export GIT_AUTHOR_DATE
$ GIT_COMMITTER_NAME="$GIT_AUTHOR_NAME"; export GIT_COMMITTER_NAME
$ GIT_COMMITTER_EMAIL="$GIT_AUTHOR_EMAIL"; export GIT_COMMITTER_EMAIL
$ GIT_COMMITTER_DATE="$GIT_AUTHOR_DATE"; export GIT_COMMITTER_DATE
$ GIT_CONFIG_NOSYSTEM=1; export GIT_CONFIG_NOSYSTEM
root hg repo
$ hg init t
$ cd t
$ echo a > a
$ hg add a
$ hg commit -m a
$ cd ..
new external git repo
$ mkdir gitroot
$ cd gitroot
$ git init -q
$ echo g > g
$ git add g
$ git commit -q -m g
add subrepo clone
$ cd ../t
$ echo 's = [git]../gitroot' > .hgsub
$ git clone -q ../gitroot s
$ hg add .hgsub
$ hg commit -m 'new git subrepo'
$ hg debugsub
path s
source ../gitroot
revision da5f5b1d8ffcf62fb8327bcd3c89a4367a6018e7
record a new commit from upstream from a different branch
$ cd ../gitroot
$ git checkout -q -b testing
$ echo gg >> g
$ git commit -q -a -m gg
$ cd ../t/s
$ git pull -q >/dev/null 2>/dev/null
$ git checkout -q -b testing origin/testing >/dev/null
$ cd ..
$ hg status --subrepos
M s/g
$ hg commit -m 'update git subrepo'
$ hg debugsub
path s
source ../gitroot
revision 126f2a14290cd5ce061fdedc430170e8d39e1c5a
make $GITROOT pushable, by replacing it with a clone with nothing checked out
$ cd ..
$ git clone gitroot gitrootbare --bare -q
$ rm -rf gitroot
$ mv gitrootbare gitroot
clone root
$ cd t
$ hg clone . ../tc 2> /dev/null
updating to branch default
cloning subrepo s from $TESTTMP/gitroot
3 files updated, 0 files merged, 0 files removed, 0 files unresolved
$ cd ../tc
$ hg debugsub
path s
source ../gitroot
revision 126f2a14290cd5ce061fdedc430170e8d39e1c5a
update to previous substate
$ hg update 1 -q
$ cat s/g
g
$ hg debugsub
path s
source ../gitroot
revision da5f5b1d8ffcf62fb8327bcd3c89a4367a6018e7
clone root, make local change
$ cd ../t
$ hg clone . ../ta 2> /dev/null
updating to branch default
cloning subrepo s from $TESTTMP/gitroot
3 files updated, 0 files merged, 0 files removed, 0 files unresolved
$ cd ../ta
$ echo ggg >> s/g
$ hg status --subrepos
M s/g
$ hg diff --subrepos
diff --git a/s/g b/s/g
index 089258f..85341ee 100644
--- a/s/g
+++ b/s/g
@@ -1,2 +1,3 @@
g
gg
+ggg
$ hg commit --subrepos -m ggg
committing subrepository s
$ hg debugsub
path s
source ../gitroot
revision 79695940086840c99328513acbe35f90fcd55e57
clone root separately, make different local change
$ cd ../t
$ hg clone . ../tb 2> /dev/null
updating to branch default
cloning subrepo s from $TESTTMP/gitroot
3 files updated, 0 files merged, 0 files removed, 0 files unresolved
$ cd ../tb/s
$ hg status --subrepos
$ echo f > f
$ hg status --subrepos
? s/f
$ hg add .
adding f
$ git add f
$ cd ..
$ hg status --subrepos
A s/f
$ hg commit --subrepos -m f
committing subrepository s
$ hg debugsub
path s
source ../gitroot
revision aa84837ccfbdfedcdcdeeedc309d73e6eb069edc
user b push changes
$ hg push 2>/dev/null
pushing to $TESTTMP/t (glob)
pushing branch testing of subrepo s
searching for changes
adding changesets
adding manifests
adding file changes
added 1 changesets with 1 changes to 1 files
user a pulls, merges, commits
$ cd ../ta
$ hg pull
pulling from $TESTTMP/t (glob)
searching for changes
adding changesets
adding manifests
adding file changes
added 1 changesets with 1 changes to 1 files (+1 heads)
(run 'hg heads' to see heads, 'hg merge' to merge)
$ hg merge 2>/dev/null
subrepository s diverged (local revision: 7969594, remote revision: aa84837)
(M)erge, keep (l)ocal [working copy] or keep (r)emote [merge rev]? m
pulling subrepo s from $TESTTMP/gitroot
0 files updated, 0 files merged, 0 files removed, 0 files unresolved
(branch merge, don't forget to commit)
$ hg st --subrepos s
A s/f
$ cat s/f
f
$ cat s/g
g
gg
ggg
$ hg commit --subrepos -m 'merge'
committing subrepository s
$ hg status --subrepos --rev 1:5
M .hgsubstate
M s/g
A s/f
$ hg debugsub
path s
source ../gitroot
revision f47b465e1bce645dbf37232a00574aa1546ca8d3
$ hg push 2>/dev/null
pushing to $TESTTMP/t (glob)
pushing branch testing of subrepo s
searching for changes
adding changesets
adding manifests
adding file changes
added 2 changesets with 2 changes to 1 files
make upstream git changes
$ cd ..
$ git clone -q gitroot gitclone
$ cd gitclone
$ echo ff >> f
$ git commit -q -a -m ff
$ echo fff >> f
$ git commit -q -a -m fff
$ git push origin testing 2>/dev/null
make and push changes to hg without updating the subrepo
$ cd ../t
$ hg clone . ../td 2>&1 | egrep -v '^Cloning into|^done\.'
updating to branch default
cloning subrepo s from $TESTTMP/gitroot
checking out detached HEAD in subrepo s
check out a git branch if you intend to make changes
3 files updated, 0 files merged, 0 files removed, 0 files unresolved
$ cd ../td
$ echo aa >> a
$ hg commit -m aa
$ hg push
pushing to $TESTTMP/t (glob)
searching for changes
adding changesets
adding manifests
adding file changes
added 1 changesets with 1 changes to 1 files
sync to upstream git, distribute changes
$ cd ../ta
$ hg pull -u -q
$ cd s
$ git pull -q >/dev/null 2>/dev/null
$ cd ..
$ hg commit -m 'git upstream sync'
$ hg debugsub
path s
source ../gitroot
revision 32a343883b74769118bb1d3b4b1fbf9156f4dddc
$ hg push -q
$ cd ../tb
$ hg pull -q
$ hg update 2>/dev/null
pulling subrepo s from $TESTTMP/gitroot
2 files updated, 0 files merged, 0 files removed, 0 files unresolved
$ hg debugsub
path s
source ../gitroot
revision 32a343883b74769118bb1d3b4b1fbf9156f4dddc
create a new git branch
$ cd s
$ git checkout -b b2
Switched to a new branch 'b2'
$ echo a>a
$ git add a
$ git commit -qm 'add a'
$ cd ..
$ hg commit -m 'add branch in s'
pulling new git branch should not create tracking branch named 'origin/b2'
(issue3870)
$ cd ../td/s
$ git remote set-url origin $TESTTMP/tb/s
$ git branch --no-track oldtesting
$ cd ..
$ hg pull -q ../tb
$ hg up
From $TESTTMP/tb/s
* [new branch] b2 -> origin/b2
Previous HEAD position was f47b465... merge
Switched to a new branch 'b2'
pulling subrepo s from $TESTTMP/tb/s
1 files updated, 0 files merged, 0 files removed, 0 files unresolved
update to a revision without the subrepo, keeping the local git repository
$ cd ../t
$ hg up 0
0 files updated, 0 files merged, 2 files removed, 0 files unresolved
$ ls -a s
.
..
.git
$ hg up 2
2 files updated, 0 files merged, 0 files removed, 0 files unresolved
$ ls -a s
.
..
.git
g
archive subrepos
$ cd ../tc
$ hg pull -q
$ hg archive --subrepos -r 5 ../archive 2>/dev/null
pulling subrepo s from $TESTTMP/gitroot
$ cd ../archive
$ cat s/f
f
$ cat s/g
g
gg
ggg
$ hg -R ../tc archive --subrepo -r 5 -X ../tc/**f ../archive_x 2>/dev/null
$ find ../archive_x | sort | grep -v pax_global_header
../archive_x
../archive_x/.hg_archival.txt
../archive_x/.hgsub
../archive_x/.hgsubstate
../archive_x/a
../archive_x/s
../archive_x/s/g
$ hg -R ../tc archive -S ../archive.tgz --prefix '.' 2>/dev/null
$ tar -tzf ../archive.tgz | sort | grep -v pax_global_header
.hg_archival.txt
.hgsub
.hgsubstate
a
s/g
create nested repo
$ cd ..
$ hg init outer
$ cd outer
$ echo b>b
$ hg add b
$ hg commit -m b
$ hg clone ../t inner 2> /dev/null
updating to branch default
cloning subrepo s from $TESTTMP/gitroot
3 files updated, 0 files merged, 0 files removed, 0 files unresolved
$ echo inner = inner > .hgsub
$ hg add .hgsub
$ hg commit -m 'nested sub'
nested commit
$ echo ffff >> inner/s/f
$ hg status --subrepos
M inner/s/f
$ hg commit --subrepos -m nested
committing subrepository inner
committing subrepository inner/s (glob)
nested archive
$ hg archive --subrepos ../narchive
$ ls ../narchive/inner/s | grep -v pax_global_header
f
g
relative source expansion
$ cd ..
$ mkdir d
$ hg clone t d/t 2> /dev/null
updating to branch default
cloning subrepo s from $TESTTMP/gitroot
3 files updated, 0 files merged, 0 files removed, 0 files unresolved
Don't crash if the subrepo is missing
$ hg clone t missing -q
$ cd missing
$ rm -rf s
$ hg status -S
$ hg sum | grep commit
commit: 1 subrepos
$ hg push -q
abort: subrepo s is missing (in subrepo s)
[255]
$ hg commit --subrepos -qm missing
abort: subrepo s is missing (in subrepo s)
[255]
#if symlink
Don't crash if subrepo is a broken symlink
$ ln -s broken s
$ hg status -S
$ hg push -q
abort: subrepo s is missing (in subrepo s)
[255]
$ hg commit --subrepos -qm missing
abort: subrepo s is missing (in subrepo s)
[255]
$ rm s
#endif
$ hg update -C 2> /dev/null
cloning subrepo s from $TESTTMP/gitroot
1 files updated, 0 files merged, 0 files removed, 0 files unresolved
$ hg sum | grep commit
commit: (clean)
Don't crash if the .hgsubstate entry is missing
$ hg update 1 -q
$ hg rm .hgsubstate
$ hg commit .hgsubstate -m 'no substate'
nothing changed
[1]
$ hg tag -l nosubstate
$ hg manifest
.hgsub
.hgsubstate
a
$ hg status -S
R .hgsubstate
$ hg sum | grep commit
commit: 1 removed, 1 subrepos (new branch head)
$ hg commit -m 'restore substate'
nothing changed
[1]
$ hg manifest
.hgsub
.hgsubstate
a
$ hg sum | grep commit
commit: 1 removed, 1 subrepos (new branch head)
$ hg update -qC nosubstate
$ ls s
g
issue3109: false positives in git diff-index
$ hg update -q
$ touch -t 200001010000 s/g
$ hg status --subrepos
$ touch -t 200001010000 s/g
$ hg sum | grep commit
commit: (clean)
Check hg update --clean
$ cd $TESTTMP/ta
$ echo > s/g
$ cd s
$ echo c1 > f1
$ echo c1 > f2
$ git add f1
$ cd ..
$ hg status -S
M s/g
A s/f1
? s/f2
$ ls s
f
f1
f2
g
$ hg update --clean
1 files updated, 0 files merged, 0 files removed, 0 files unresolved
$ hg status -S
? s/f1
? s/f2
$ ls s
f
f1
f2
g
Sticky subrepositories, no changes
$ cd $TESTTMP/ta
$ hg id -n
7
$ cd s
$ git rev-parse HEAD
32a343883b74769118bb1d3b4b1fbf9156f4dddc
$ cd ..
$ hg update 1 > /dev/null 2>&1
$ hg id -n
1
$ cd s
$ git rev-parse HEAD
da5f5b1d8ffcf62fb8327bcd3c89a4367a6018e7
$ cd ..
Sticky subrepositories, file changes
$ touch s/f1
$ cd s
$ git add f1
$ cd ..
$ hg id -n
1+
$ cd s
$ git rev-parse HEAD
da5f5b1d8ffcf62fb8327bcd3c89a4367a6018e7
$ cd ..
$ hg update 4
subrepository s diverged (local revision: da5f5b1, remote revision: aa84837)
(M)erge, keep (l)ocal [working copy] or keep (r)emote [destination]? m
subrepository sources for s differ
use (l)ocal source (da5f5b1) or (r)emote source (aa84837)? l
0 files updated, 0 files merged, 0 files removed, 0 files unresolved
$ hg id -n
4+
$ cd s
$ git rev-parse HEAD
da5f5b1d8ffcf62fb8327bcd3c89a4367a6018e7
$ cd ..
$ hg update --clean tip > /dev/null 2>&1
Sticky subrepository, revision updates
$ hg id -n
7
$ cd s
$ git rev-parse HEAD
32a343883b74769118bb1d3b4b1fbf9156f4dddc
$ cd ..
$ cd s
$ git checkout aa84837ccfbdfedcdcdeeedc309d73e6eb069edc
Previous HEAD position was 32a3438... fff
HEAD is now at aa84837... f
$ cd ..
$ hg update 1
subrepository s diverged (local revision: 32a3438, remote revision: da5f5b1)
(M)erge, keep (l)ocal [working copy] or keep (r)emote [destination]? m
subrepository sources for s differ (in checked out version)
use (l)ocal source (32a3438) or (r)emote source (da5f5b1)? l
1 files updated, 0 files merged, 0 files removed, 0 files unresolved
$ hg id -n
1+
$ cd s
$ git rev-parse HEAD
aa84837ccfbdfedcdcdeeedc309d73e6eb069edc
$ cd ..
Sticky subrepository, file changes and revision updates
$ touch s/f1
$ cd s
$ git add f1
$ git rev-parse HEAD
aa84837ccfbdfedcdcdeeedc309d73e6eb069edc
$ cd ..
$ hg id -n
1+
$ hg update 7
subrepository s diverged (local revision: 32a3438, remote revision: 32a3438)
(M)erge, keep (l)ocal [working copy] or keep (r)emote [destination]? m
subrepository sources for s differ
use (l)ocal source (32a3438) or (r)emote source (32a3438)? l
1 files updated, 0 files merged, 0 files removed, 0 files unresolved
$ hg id -n
7+
$ cd s
$ git rev-parse HEAD
aa84837ccfbdfedcdcdeeedc309d73e6eb069edc
$ cd ..
Sticky repository, update --clean
$ hg update --clean tip 2>/dev/null
1 files updated, 0 files merged, 0 files removed, 0 files unresolved
$ hg id -n
7
$ cd s
$ git rev-parse HEAD
32a343883b74769118bb1d3b4b1fbf9156f4dddc
$ cd ..
Test subrepo already at intended revision:
$ cd s
$ git checkout 32a343883b74769118bb1d3b4b1fbf9156f4dddc
HEAD is now at 32a3438... fff
$ cd ..
$ hg update 1
Previous HEAD position was 32a3438... fff
HEAD is now at da5f5b1... g
2 files updated, 0 files merged, 0 files removed, 0 files unresolved
$ hg id -n
1
$ cd s
$ git rev-parse HEAD
da5f5b1d8ffcf62fb8327bcd3c89a4367a6018e7
$ cd ..
Test forgetting files, not implemented in git subrepo, used to
traceback
#if no-windows
$ hg forget 'notafile*'
notafile*: No such file or directory
[1]
#else
$ hg forget 'notafile'
notafile: * (glob)
[1]
#endif
$ cd ..
Test sanitizing ".hg/hgrc" in subrepo
$ cd t
$ hg tip -q
7:af6d2edbb0d3
$ hg update -q -C af6d2edbb0d3
$ cd s
$ git checkout -q -b sanitize-test
$ mkdir .hg
$ echo '.hg/hgrc in git repo' > .hg/hgrc
$ mkdir -p sub/.hg
$ echo 'sub/.hg/hgrc in git repo' > sub/.hg/hgrc
$ git add .hg sub
$ git commit -qm 'add .hg/hgrc to be sanitized at hg update'
$ git push -q origin sanitize-test
$ cd ..
$ grep ' s$' .hgsubstate
32a343883b74769118bb1d3b4b1fbf9156f4dddc s
$ hg commit -qm 'commit with git revision including .hg/hgrc'
$ hg parents -q
8:3473d20bddcf
$ grep ' s$' .hgsubstate
c4069473b459cf27fd4d7c2f50c4346b4e936599 s
$ cd ..
$ hg -R tc pull -q
$ hg -R tc update -q -C 3473d20bddcf 2>&1 | sort
warning: removing potentially hostile 'hgrc' in '$TESTTMP/tc/s/.hg' (glob)
warning: removing potentially hostile 'hgrc' in '$TESTTMP/tc/s/sub/.hg' (glob)
$ cd tc
$ hg parents -q
8:3473d20bddcf
$ grep ' s$' .hgsubstate
c4069473b459cf27fd4d7c2f50c4346b4e936599 s
$ test -f s/.hg/hgrc
[1]
$ test -f s/sub/.hg/hgrc
[1]
$ cd ..
additional test for "git merge --ff" route:
$ cd t
$ hg tip -q
8:3473d20bddcf
$ hg update -q -C af6d2edbb0d3
$ cd s
$ git checkout -q testing
$ mkdir .hg
$ echo '.hg/hgrc in git repo' > .hg/hgrc
$ mkdir -p sub/.hg
$ echo 'sub/.hg/hgrc in git repo' > sub/.hg/hgrc
$ git add .hg sub
$ git commit -qm 'add .hg/hgrc to be sanitized at hg update (git merge --ff)'
$ git push -q origin testing
$ cd ..
$ grep ' s$' .hgsubstate
32a343883b74769118bb1d3b4b1fbf9156f4dddc s
$ hg commit -qm 'commit with git revision including .hg/hgrc'
$ hg parents -q
9:ed23f7fe024e
$ grep ' s$' .hgsubstate
f262643c1077219fbd3858d54e78ef050ef84fbf s
$ cd ..
$ cd tc
$ hg update -q -C af6d2edbb0d3
$ test -f s/.hg/hgrc
[1]
$ test -f s/sub/.hg/hgrc
[1]
$ cd ..
$ hg -R tc pull -q
$ hg -R tc update -q -C ed23f7fe024e 2>&1 | sort
warning: removing potentially hostile 'hgrc' in '$TESTTMP/tc/s/.hg' (glob)
warning: removing potentially hostile 'hgrc' in '$TESTTMP/tc/s/sub/.hg' (glob)
$ cd tc
$ hg parents -q
9:ed23f7fe024e
$ grep ' s$' .hgsubstate
f262643c1077219fbd3858d54e78ef050ef84fbf s
$ test -f s/.hg/hgrc
[1]
$ test -f s/sub/.hg/hgrc
[1]
Test that sanitizing is omitted in meta data area:
$ mkdir s/.git/.hg
$ echo '.hg/hgrc in git metadata area' > s/.git/.hg/hgrc
$ hg update -q -C af6d2edbb0d3
checking out detached HEAD in subrepo s
check out a git branch if you intend to make changes
check differences made by most recent change
$ cd s
$ cat > foobar << EOF
> woopwoop
>
> foo
> bar
> EOF
$ git add foobar
$ cd ..
$ hg diff --subrepos
diff --git a/s/foobar b/s/foobar
new file mode 100644
index 0000000..8a5a5e2
--- /dev/null
+++ b/s/foobar
@@ -0,0 +1,4 @@
+woopwoop
+
+foo
+bar
$ hg commit --subrepos -m "Added foobar"
committing subrepository s
created new head
$ hg diff -c . --subrepos --nodates
diff -r af6d2edbb0d3 -r 255ee8cf690e .hgsubstate
--- a/.hgsubstate
+++ b/.hgsubstate
@@ -1,1 +1,1 @@
-32a343883b74769118bb1d3b4b1fbf9156f4dddc s
+fd4dbf828a5b2fcd36b2bcf21ea773820970d129 s
diff --git a/s/foobar b/s/foobar
new file mode 100644
index 0000000..8a5a5e2
--- /dev/null
+++ b/s/foobar
@@ -0,0 +1,4 @@
+woopwoop
+
+foo
+bar
check output when only diffing the subrepository
$ hg diff -c . --subrepos s
diff --git a/s/foobar b/s/foobar
new file mode 100644
index 0000000..8a5a5e2
--- /dev/null
+++ b/s/foobar
@@ -0,0 +1,4 @@
+woopwoop
+
+foo
+bar
check output when diffing something else
$ hg diff -c . --subrepos .hgsubstate --nodates
diff -r af6d2edbb0d3 -r 255ee8cf690e .hgsubstate
--- a/.hgsubstate
+++ b/.hgsubstate
@@ -1,1 +1,1 @@
-32a343883b74769118bb1d3b4b1fbf9156f4dddc s
+fd4dbf828a5b2fcd36b2bcf21ea773820970d129 s
add new changes, including whitespace
$ cd s
$ cat > foobar << EOF
> woop woop
>
> foo
> bar
> EOF
$ echo foo > barfoo
$ git add barfoo
$ cd ..
$ hg diff --subrepos --ignore-all-space
diff --git a/s/barfoo b/s/barfoo
new file mode 100644
index 0000000..257cc56
--- /dev/null
+++ b/s/barfoo
@@ -0,0 +1* @@ (glob)
+foo
$ hg diff --subrepos s/foobar
diff --git a/s/foobar b/s/foobar
index 8a5a5e2..bd5812a 100644
--- a/s/foobar
+++ b/s/foobar
@@ -1,4 +1,4 @@
-woopwoop
+woop woop
foo
bar
execute a diffstat
the output contains a regex, because git 1.7.10 and 1.7.11
change the amount of whitespace
$ hg diff --subrepos --stat
\s*barfoo |\s*1 + (re)
\s*foobar |\s*2 +- (re)
2 files changed, 2 insertions\(\+\), 1 deletions?\(-\) (re)
adding an include should ignore the other elements
$ hg diff --subrepos -I s/foobar
diff --git a/s/foobar b/s/foobar
index 8a5a5e2..bd5812a 100644
--- a/s/foobar
+++ b/s/foobar
@@ -1,4 +1,4 @@
-woopwoop
+woop woop
foo
bar
adding an exclude should ignore this element
$ hg diff --subrepos -X s/foobar
diff --git a/s/barfoo b/s/barfoo
new file mode 100644
index 0000000..257cc56
--- /dev/null
+++ b/s/barfoo
@@ -0,0 +1* @@ (glob)
+foo
moving a file should show a removal and an add
$ hg revert --all
reverting subrepo ../gitroot
$ cd s
$ git mv foobar woop
$ cd ..
$ hg diff --subrepos
diff --git a/s/foobar b/s/foobar
deleted file mode 100644
index 8a5a5e2..0000000
--- a/s/foobar
+++ /dev/null
@@ -1,4 +0,0 @@
-woopwoop
-
-foo
-bar
diff --git a/s/woop b/s/woop
new file mode 100644
index 0000000..8a5a5e2
--- /dev/null
+++ b/s/woop
@@ -0,0 +1,4 @@
+woopwoop
+
+foo
+bar
$ rm s/woop
revert the subrepository
$ hg revert --all
reverting subrepo ../gitroot
$ hg status --subrepos
? s/barfoo
? s/foobar.orig
$ mv s/foobar.orig s/foobar
$ hg revert --no-backup s
reverting subrepo ../gitroot
$ hg status --subrepos
? s/barfoo
revert moves orig files to the right place
$ echo 'bloop' > s/foobar
$ hg revert --all --verbose --config 'ui.origbackuppath=.hg/origbackups'
reverting subrepo ../gitroot
creating directory: $TESTTMP/tc/.hg/origbackups (glob)
saving current version of foobar as $TESTTMP/tc/.hg/origbackups/foobar.orig (glob)
$ ls .hg/origbackups
foobar.orig
$ rm -rf .hg/origbackups
show file at specific revision
$ cat > s/foobar << EOF
> woop woop
> fooo bar
> EOF
$ hg commit --subrepos -m "updated foobar"
committing subrepository s
$ cat > s/foobar << EOF
> current foobar
> (should not be visible using hg cat)
> EOF
$ hg cat -r . s/foobar
woop woop
fooo bar (no-eol)
$ hg cat -r "parents(.)" s/foobar > catparents
$ mkdir -p tmp/s
$ hg cat -r "parents(.)" --output tmp/%% s/foobar
$ diff tmp/% catparents
$ hg cat -r "parents(.)" --output tmp/%s s/foobar
$ diff tmp/foobar catparents
$ hg cat -r "parents(.)" --output tmp/%d/otherfoobar s/foobar
$ diff tmp/s/otherfoobar catparents
$ hg cat -r "parents(.)" --output tmp/%p s/foobar
$ diff tmp/s/foobar catparents
$ hg cat -r "parents(.)" --output tmp/%H s/foobar
$ diff tmp/255ee8cf690ec86e99b1e80147ea93ece117cd9d catparents
$ hg cat -r "parents(.)" --output tmp/%R s/foobar
$ diff tmp/10 catparents
$ hg cat -r "parents(.)" --output tmp/%h s/foobar
$ diff tmp/255ee8cf690e catparents
$ rm tmp/10
$ hg cat -r "parents(.)" --output tmp/%r s/foobar
$ diff tmp/10 catparents
$ mkdir tmp/tc
$ hg cat -r "parents(.)" --output tmp/%b/foobar s/foobar
$ diff tmp/tc/foobar catparents
cleanup
$ rm -r tmp
$ rm catparents
add git files, using either files or patterns
$ echo "hsss! hsssssssh!" > s/snake.python
$ echo "ccc" > s/c.c
$ echo "cpp" > s/cpp.cpp
$ hg add s/snake.python s/c.c s/cpp.cpp
$ hg st --subrepos s
M s/foobar
A s/c.c
A s/cpp.cpp
A s/snake.python
? s/barfoo
$ hg revert s
reverting subrepo ../gitroot
$ hg add --subrepos "glob:**.python"
adding s/snake.python (glob)
$ hg st --subrepos s
A s/snake.python
? s/barfoo
? s/c.c
? s/cpp.cpp
? s/foobar.orig
$ hg revert s
reverting subrepo ../gitroot
$ hg add --subrepos s
adding s/barfoo (glob)
adding s/c.c (glob)
adding s/cpp.cpp (glob)
adding s/foobar.orig (glob)
adding s/snake.python (glob)
$ hg st --subrepos s
A s/barfoo
A s/c.c
A s/cpp.cpp
A s/foobar.orig
A s/snake.python
$ hg revert s
reverting subrepo ../gitroot
make sure everything is reverted correctly
$ hg st --subrepos s
? s/barfoo
? s/c.c
? s/cpp.cpp
? s/foobar.orig
? s/snake.python
$ hg add --subrepos --exclude "path:s/c.c"
adding s/barfoo (glob)
adding s/cpp.cpp (glob)
adding s/foobar.orig (glob)
adding s/snake.python (glob)
$ hg st --subrepos s
A s/barfoo
A s/cpp.cpp
A s/foobar.orig
A s/snake.python
? s/c.c
$ hg revert --all -q
.hgignore should not have influence in subrepos
$ cat > .hgignore << EOF
> syntax: glob
> *.python
> EOF
$ hg add .hgignore
$ hg add --subrepos "glob:**.python" s/barfoo
adding s/snake.python (glob)
$ hg st --subrepos s
A s/barfoo
A s/snake.python
? s/c.c
? s/cpp.cpp
? s/foobar.orig
$ hg revert --all -q
.gitignore should have influence,
except for explicitly added files (no patterns)
$ cat > s/.gitignore << EOF
> *.python
> EOF
$ hg add s/.gitignore
$ hg st --subrepos s
A s/.gitignore
? s/barfoo
? s/c.c
? s/cpp.cpp
? s/foobar.orig
$ hg st --subrepos s --all
A s/.gitignore
? s/barfoo
? s/c.c
? s/cpp.cpp
? s/foobar.orig
I s/snake.python
C s/f
C s/foobar
C s/g
$ hg add --subrepos "glob:**.python"
$ hg st --subrepos s
A s/.gitignore
? s/barfoo
? s/c.c
? s/cpp.cpp
? s/foobar.orig
$ hg add --subrepos s/snake.python
$ hg st --subrepos s
A s/.gitignore
A s/snake.python
? s/barfoo
? s/c.c
? s/cpp.cpp
? s/foobar.orig
correctly do a dry run
$ hg add --subrepos s --dry-run
adding s/barfoo (glob)
adding s/c.c (glob)
adding s/cpp.cpp (glob)
adding s/foobar.orig (glob)
$ hg st --subrepos s
A s/.gitignore
A s/snake.python
? s/barfoo
? s/c.c
? s/cpp.cpp
? s/foobar.orig
error given when adding an already tracked file
$ hg add s/.gitignore
s/.gitignore already tracked!
[1]
$ hg add s/g
s/g already tracked!
[1]
removed files can be re-added
removing files using 'rm' or 'git rm' has the same effect,
since we ignore the staging area
$ hg ci --subrepos -m 'snake'
committing subrepository s
$ cd s
$ rm snake.python
(remove leftover .hg so Mercurial doesn't look for a root here)
$ rm -rf .hg
$ hg status --subrepos --all .
R snake.python
? barfoo
? c.c
? cpp.cpp
? foobar.orig
C .gitignore
C f
C foobar
C g
$ git rm snake.python
rm 'snake.python'
$ hg status --subrepos --all .
R snake.python
? barfoo
? c.c
? cpp.cpp
? foobar.orig
C .gitignore
C f
C foobar
C g
$ touch snake.python
$ cd ..
$ hg add s/snake.python
$ hg status -S
M s/snake.python
? .hgignore
? s/barfoo
? s/c.c
? s/cpp.cpp
? s/foobar.orig
$ hg revert --all -q
make sure we show changed files, rather than changed subtrees
$ mkdir s/foo
$ touch s/foo/bwuh
$ hg add s/foo/bwuh
$ hg commit -S -m "add bwuh"
committing subrepository s
$ hg status -S --change .
M .hgsubstate
A s/foo/bwuh
? s/barfoo
? s/c.c
? s/cpp.cpp
? s/foobar.orig
? s/snake.python.orig
test for Git CVE-2016-3068
$ hg init malicious-subrepository
$ cd malicious-subrepository
$ echo "s = [git]ext::sh -c echo% pwned:% \$PWNED_MSG% >pwned.txt" > .hgsub
$ git init s
Initialized empty Git repository in $TESTTMP/tc/malicious-subrepository/s/.git/
$ cd s
$ git commit --allow-empty -m 'empty'
[master (root-commit) 153f934] empty
$ cd ..
$ hg add .hgsub
$ hg commit -m "add subrepo"
$ cd ..
$ rm -f pwned.txt
$ unset GIT_ALLOW_PROTOCOL
$ PWNED_MSG="your git is too old or mercurial has regressed" hg clone \
> malicious-subrepository malicious-subrepository-protected
Cloning into '$TESTTMP/tc/malicious-subrepository-protected/s'... (glob)
fatal: transport 'ext' not allowed
updating to branch default
cloning subrepo s from ext::sh -c echo% pwned:% $PWNED_MSG% >pwned.txt
abort: git clone error 128 in s (in subrepo s)
[255]
$ f -Dq pwned.txt
pwned.txt: file not found
whitelisting of ext should be respected (that's the git submodule behaviour)
$ rm -f pwned.txt
$ env GIT_ALLOW_PROTOCOL=ext PWNED_MSG="you asked for it" hg clone \
> malicious-subrepository malicious-subrepository-clone-allowed
Cloning into '$TESTTMP/tc/malicious-subrepository-clone-allowed/s'... (glob)
fatal: Could not read from remote repository.
Please make sure you have the correct access rights
and the repository exists.
updating to branch default
cloning subrepo s from ext::sh -c echo% pwned:% $PWNED_MSG% >pwned.txt
abort: git clone error 128 in s (in subrepo s)
[255]
$ f -Dq pwned.txt
pwned: you asked for it