phases: rework the logic of _pushdiscoveryphase to bound complexity
This rework the various graph traversal in _pushdiscoveryphase to keep the
complexity in check.
This is done though a couple of things:
- first, limiting the space we have to explore, for example, if we are not in
publishing push, we don't need to consider remote draft roots that are also
draft locally, as there is nothing to be moved there.
- avoid unbounded descendant computation, and use the faster "rev between"
computation.
This provide a massive boost to performance when exchanging with repository with
a massive amount of draft, like mozilla-try:
### data-env-vars.name = mozilla-try-2023-03-22-zstd-sparse-revlog
# benchmark.name = hg.command.push
# bin-env-vars.hg.flavor = default
# bin-env-vars.hg.py-re2-module = default
# benchmark.variants.explicit-rev = all-out-heads
# benchmark.variants.issue6528 = disabled
# benchmark.variants.protocol = ssh
# benchmark.variants.reuse-external-delta-parent = default
## benchmark.variants.revs = any-1-extra-rev
before: 20.346590 seconds
after: 11.232059 seconds (-38.15%, -7.48 seconds)
## benchmark.variants.revs = any-100-extra-rev
before: 24.752051 seconds
after: 15.367412 seconds (-37.91%, -9.38 seconds)
After this changes, the push operation is still quite too slow. Some of this
can be attributed to general phases slowness (reading all the roots from disk
for example) and other know slowness (not using persistent-nodemap, branchmap,
tags, etc. We are also working on them, but with this series, phase discovery
during push no longer showing up in profile and this is a pretty nice and bit
low-hanging fruit out of the way.
### (same case as the above)
# benchmark.variants.revs = any-1-extra-rev
pre-%ln-change: 44.235070
this-changeset: 11.232059 seconds (-74.61%, -33.00 seconds)
# benchmark.variants.revs = any-100-extra-rev
pre-%ln-change: 49.234697
this-changeset: 15.367412 seconds (-68.79%, -33.87 seconds)
Note that with this change, the `hg push` performance is now much closer to the
`hg pull` performance, even it still lagging behind a bit. (and the overall
performance are still too slow).
### data-env-vars.name = mozilla-try-2023-03-22-ds2-pnm
# benchmark.variants.explicit-rev = all-out-heads
# benchmark.variants.issue6528 = disabled
# benchmark.variants.protocol = ssh
# benchmark.variants.pulled-delta-reuse-policy = default
# bin-env-vars.hg.flavor = rust
## benchmark.variants.revs = any-1-extra-rev
hg.command.pull: 6.517450
hg.command.push: 11.219888
## benchmark.variants.revs = any-100-extra-rev
hg.command.pull: 10.160991
hg.command.push: 14.251107
### data-env-vars.name = mozilla-try-2023-03-22-zstd-sparse-revlog
# bin-env-vars.hg.py-re2-module = default
# benchmark.variants.explicit-rev = all-out-heads
# benchmark.variants.issue6528 = disabled
# benchmark.variants.protocol = ssh
# benchmark.variants.pulled-delta-reuse-policy = default
## bin-env-vars.hg.flavor = default
## benchmark.variants.revs = any-1-extra-rev
hg.command.pull: 8.577772
hg.command.push: 11.232059
## bin-env-vars.hg.flavor = default
## benchmark.variants.revs = any-100-extra-rev
hg.command.pull: 13.152976
hg.command.push: 15.367412
## bin-env-vars.hg.flavor = rust
## benchmark.variants.revs = any-1-extra-rev
hg.command.pull: 8.731982
hg.command.push: 11.178751
## bin-env-vars.hg.flavor = rust
## benchmark.variants.revs = any-100-extra-rev
hg.command.pull: 13.184236
hg.command.push: 15.620843
# Rust index does not support creating new flags dynamically
#if no-rust
# Create server
$ hg init server
$ cd server
$ cat >> .hg/hgrc << EOF
> [extensions]
> extension=$TESTDIR/flagprocessorext.py
> EOF
$ cd ../
# Clone server and enable extensions
$ hg clone -q server client
$ cd client
$ cat >> .hg/hgrc << EOF
> [extensions]
> extension=$TESTDIR/flagprocessorext.py
> EOF
# Commit file that will trigger the noop extension
$ echo '[NOOP]' > noop
$ hg commit -Aqm "noop"
# Commit file that will trigger the base64 extension
$ echo '[BASE64]' > base64
$ hg commit -Aqm 'base64'
# Commit file that will trigger the gzip extension
$ echo '[GZIP]' > gzip
$ hg commit -Aqm 'gzip'
# Commit file that will trigger noop and base64
$ echo '[NOOP][BASE64]' > noop-base64
$ hg commit -Aqm 'noop+base64'
# Commit file that will trigger noop and gzip
$ echo '[NOOP][GZIP]' > noop-gzip
$ hg commit -Aqm 'noop+gzip'
# Commit file that will trigger base64 and gzip
$ echo '[BASE64][GZIP]' > base64-gzip
$ hg commit -Aqm 'base64+gzip'
# Commit file that will trigger base64, gzip and noop
$ echo '[BASE64][GZIP][NOOP]' > base64-gzip-noop
$ hg commit -Aqm 'base64+gzip+noop'
# TEST: ensure the revision data is consistent
$ hg cat noop
[NOOP]
$ hg debugdata noop 0
[NOOP]
$ hg cat -r . base64
[BASE64]
$ hg debugdata base64 0
W0JBU0U2NF0K (no-eol)
$ hg cat -r . gzip
[GZIP]
$ hg debugdata gzip 0
x\x9c\x8bv\x8f\xf2\x0c\x88\xe5\x02\x00\x08\xc8\x01\xfd (no-eol) (esc)
$ hg cat -r . noop-base64
[NOOP][BASE64]
$ hg debugdata noop-base64 0
W05PT1BdW0JBU0U2NF0K (no-eol)
$ hg cat -r . noop-gzip
[NOOP][GZIP]
$ hg debugdata noop-gzip 0
x\x9c\x8b\xf6\xf3\xf7\x0f\x88\x8dv\x8f\xf2\x0c\x88\xe5\x02\x00\x1dH\x03\xf1 (no-eol) (esc)
$ hg cat -r . base64-gzip
[BASE64][GZIP]
$ hg debugdata base64-gzip 0
eJyLdnIMdjUziY12j/IMiOUCACLBBDo= (no-eol)
$ hg cat -r . base64-gzip-noop
[BASE64][GZIP][NOOP]
$ hg debugdata base64-gzip-noop 0
eJyLdnIMdjUziY12j/IMiI328/cPiOUCAESjBi4= (no-eol)
# Push to the server
$ hg push
pushing to $TESTTMP/server
searching for changes
adding changesets
adding manifests
adding file changes
added 7 changesets with 7 changes to 7 files
Ensure the data got to the server OK
$ cd ../server
$ hg cat -r 6e48f4215d24 noop
[NOOP]
$ hg debugdata noop 0
[NOOP]
$ hg cat -r 6e48f4215d24 base64
[BASE64]
$ hg debugdata base64 0
W0JBU0U2NF0K (no-eol)
$ hg cat -r 6e48f4215d24 gzip
[GZIP]
$ hg debugdata gzip 0
x\x9c\x8bv\x8f\xf2\x0c\x88\xe5\x02\x00\x08\xc8\x01\xfd (no-eol) (esc)
$ hg cat -r 6e48f4215d24 noop-base64
[NOOP][BASE64]
$ hg debugdata noop-base64 0
W05PT1BdW0JBU0U2NF0K (no-eol)
$ hg cat -r 6e48f4215d24 noop-gzip
[NOOP][GZIP]
$ hg debugdata noop-gzip 0
x\x9c\x8b\xf6\xf3\xf7\x0f\x88\x8dv\x8f\xf2\x0c\x88\xe5\x02\x00\x1dH\x03\xf1 (no-eol) (esc)
$ hg cat -r 6e48f4215d24 base64-gzip
[BASE64][GZIP]
$ hg debugdata base64-gzip 0
eJyLdnIMdjUziY12j/IMiOUCACLBBDo= (no-eol)
$ hg cat -r 6e48f4215d24 base64-gzip-noop
[BASE64][GZIP][NOOP]
$ hg debugdata base64-gzip-noop 0
eJyLdnIMdjUziY12j/IMiI328/cPiOUCAESjBi4= (no-eol)
# Initialize new client (not cloning) and setup extension
$ cd ..
$ hg init client2
$ cd client2
$ cat >> .hg/hgrc << EOF
> [paths]
> default = $TESTTMP/server
> [extensions]
> extension=$TESTDIR/flagprocessorext.py
> EOF
# Pull from server and update to latest revision
$ hg pull default
pulling from $TESTTMP/server
requesting all changes
adding changesets
adding manifests
adding file changes
added 7 changesets with 7 changes to 7 files
new changesets 07b1b9442c5b:6e48f4215d24
(run 'hg update' to get a working copy)
$ hg update
7 files updated, 0 files merged, 0 files removed, 0 files unresolved
# TEST: ensure the revision data is consistent
$ hg cat noop
[NOOP]
$ hg debugdata noop 0
[NOOP]
$ hg cat -r . base64
[BASE64]
$ hg debugdata base64 0
W0JBU0U2NF0K (no-eol)
$ hg cat -r . gzip
[GZIP]
$ hg debugdata gzip 0
x\x9c\x8bv\x8f\xf2\x0c\x88\xe5\x02\x00\x08\xc8\x01\xfd (no-eol) (esc)
$ hg cat -r . noop-base64
[NOOP][BASE64]
$ hg debugdata noop-base64 0
W05PT1BdW0JBU0U2NF0K (no-eol)
$ hg cat -r . noop-gzip
[NOOP][GZIP]
$ hg debugdata noop-gzip 0
x\x9c\x8b\xf6\xf3\xf7\x0f\x88\x8dv\x8f\xf2\x0c\x88\xe5\x02\x00\x1dH\x03\xf1 (no-eol) (esc)
$ hg cat -r . base64-gzip
[BASE64][GZIP]
$ hg debugdata base64-gzip 0
eJyLdnIMdjUziY12j/IMiOUCACLBBDo= (no-eol)
$ hg cat -r . base64-gzip-noop
[BASE64][GZIP][NOOP]
$ hg debugdata base64-gzip-noop 0
eJyLdnIMdjUziY12j/IMiI328/cPiOUCAESjBi4= (no-eol)
# TEST: ensure a missing processor is handled
$ echo '[FAIL][BASE64][GZIP][NOOP]' > fail-base64-gzip-noop
$ hg commit -Aqm 'fail+base64+gzip+noop'
abort: missing processor for flag '0x1'
[50]
$ rm fail-base64-gzip-noop
# TEST: ensure we cannot register several flag processors on the same flag
$ cat >> .hg/hgrc << EOF
> [extensions]
> extension=$TESTDIR/flagprocessorext.py
> duplicate=$TESTDIR/flagprocessorext.py
> EOF
$ hg debugrebuilddirstate
Traceback (most recent call last):
File "*/mercurial/extensions.py", line *, in _runextsetup (glob) (no-pyoxidizer !)
File "mercurial.extensions", line *, in _runextsetup (glob) (pyoxidizer !)
extsetup(ui)
File "*/tests/flagprocessorext.py", line *, in extsetup (glob)
flagutil.addflagprocessor( (py38 !)
validatehash, (no-py38 !)
File "*/mercurial/revlogutils/flagutil.py", line *, in addflagprocessor (glob) (no-pyoxidizer !)
File "mercurial.revlogutils.flagutil", line *, in addflagprocessor (glob) (pyoxidizer !)
insertflagprocessor(flag, processor, flagprocessors)
File "*/mercurial/revlogutils/flagutil.py", line *, in insertflagprocessor (glob) (no-pyoxidizer !)
File "mercurial.revlogutils.flagutil", line *, in insertflagprocessor (glob) (pyoxidizer !)
raise error.Abort(msg)
mercurial.error.Abort: cannot register multiple processors on flag '0x8'.
*** failed to set up extension duplicate: cannot register multiple processors on flag '0x8'.
$ hg st 2>&1 | grep -E 'cannot register multiple processors|flagprocessorext'
File "*/tests/flagprocessorext.py", line *, in extsetup (glob)
mercurial.error.Abort: cannot register multiple processors on flag '0x8'.
*** failed to set up extension duplicate: cannot register multiple processors on flag '0x8'.
File "*/tests/flagprocessorext.py", line *, in b64decode (glob)
$ cd ..
# TEST: bundle repo
$ hg init bundletest
$ cd bundletest
$ cat >> .hg/hgrc << EOF
> [extensions]
> flagprocessor=$TESTDIR/flagprocessorext.py
> EOF
$ for i in 0 single two three 4; do
> echo '[BASE64]a-bit-longer-'$i > base64
> hg commit -m base64-$i -A base64
> done
$ hg update 2 -q
$ echo '[BASE64]a-bit-longer-branching' > base64
$ hg commit -q -m branching
#if repobundlerepo
$ hg bundle --base 1 bundle.hg
4 changesets found
$ hg --config extensions.strip= strip -r 2 --no-backup --force -q
$ hg -R bundle.hg log --stat -T '{rev} {desc}\n' base64
5 branching
base64 | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
4 base64-4
base64 | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
3 base64-three
base64 | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
2 base64-two
base64 | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
1 base64-single
base64 | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
0 base64-0
base64 | 1 +
1 files changed, 1 insertions(+), 0 deletions(-)
$ hg bundle -R bundle.hg --base 1 bundle-again.hg -q
$ hg -R bundle-again.hg log --stat -T '{rev} {desc}\n' base64
5 branching
base64 | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
4 base64-4
base64 | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
3 base64-three
base64 | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
2 base64-two
base64 | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
1 base64-single
base64 | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
0 base64-0
base64 | 1 +
1 files changed, 1 insertions(+), 0 deletions(-)
$ rm bundle.hg bundle-again.hg
#endif
# TEST: hg status
$ hg status
$ hg diff
#endif