view tests/test-push.t @ 51681:522b4d729e89

mmap: populate the mapping by default

Without pre-population, accessing all data through a mmap can result in many
pagefault, reducing performance significantly. If the mmap is prepopulated, the
performance can no longer get slower than a full read.
(See benchmark number below)

In some cases were very few data is read, prepopulating can be overkill and
slower than populating on access (through page fault). So that behavior can be
controlled when the caller can pre-determine the best behavior.
(See benchmark number below)

In addition, testing with populating in a secondary thread yield great result
combining the best of each approach. This might be implemented in later
changesets.

In all cases, using mmap has a great effect on memory usage when many processes
run in parallel on the same machine.

### Benchmarks

# What did I run

A couple of month back I ran a large benchmark campaign to assess the impact of
various approach for using mmap with the revlog (and other files), it
highlighted a few benchmarks that capture the impact of the changes well. So to
validate this change I checked the following:

- log command displaying various revisions
  (read the changelog index)
- log command displaying the patch of listed revisions
  (read the changelog index, the manifest index and a few files indexes)
- unbundling a few revisions
  (read and write changelog, manifest and few files indexes, and walk the graph
  to update some cache)
- pushing a few revisions
  (read and write changelog, manifest and few files indexes, walk the graph to
  update some cache, performs various accesses locally and remotely during
  discovery)

Benchmarks were run using the default module policy (c+py) and the rust one. No
significant difference were found between the two implementation, so we will
present result using the default policy (unless otherwise specified).

I ran them on a few repositories :

- mercurial: a "public changeset only" copy of mercurial from 2018-08-01 using
  zstd compression and sparse-revlog

- pypy: a copy of pypy from 2018-08-01 using zstd compression and sparse-revlog

- netbeans: a copy of netbeans from 2018-08-01 using zstd compression and
  sparse-revlog

- mozilla-try: a copy of mozilla-try from 2019-02-18 using zstd compression and
  sparse-revlog

- mozilla-try persistent-nodemap: Same as the above but with a persistent
  nodemap. Used for the log --patch benchmark only

# Results

For the smaller repositories (mercurial, pypy), the impact of mmap is almost
imperceptible, other cost dominating the operation. The impact of prepopulating
is undiscernible in the benchmark we ran.

For larger repositories the benchmark support explanation given above:

On netbeans, the log can be about 1% faster without repopulation (for a
difference < 100ms) but unbundle becomes a bit slower, even when small.

### data-env-vars.name                             = netbeans-2018-08-01-zstd-sparse-revlog
  # benchmark.name                                 = hg.command.unbundle
  # benchmark.variants.issue6528                   = disabled
  # benchmark.variants.reuse-external-delta-parent = yes
  # benchmark.variants.revs                        = any-1-extra-rev
  # benchmark.variants.source                      = unbundle
  # benchmark.variants.verbosity                   = quiet
with-populate: 0.240157
no-populate:   0.265087  (+10.38%, +0.02)
  # benchmark.variants.revs                        = any-100-extra-rev
with-populate: 1.459518
no-populate:   1.481290   (+1.49%, +0.02)
 ## benchmark.name                                 = hg.command.push
  # benchmark.variants.explicit-rev                = none
  # benchmark.variants.issue6528                   = disabled
  # benchmark.variants.protocol                    = ssh
  # benchmark.variants.reuse-external-delta-parent = yes
  # benchmark.variants.revs                        = any-1-extra-rev
with-populate: 0.771919
no-populate:   0.792025   (+2.60%, +0.02)
  # benchmark.variants.revs                        = any-100-extra-rev
with-populate: 1.459518
no-populate:   1.481290   (+1.49%, +0.02)


For mozilla-try, the "slow down" from pre-populate for small `hg log` is more
visible, but still small in absolute time. (using rust value for the persistent
nodemap value to be relevant).

### data-env-vars.name            = mozilla-try-2019-02-18-ds2-pnm
  # benchmark.name                = hg.command.log
  # bin-env-vars.hg.flavor        = rust
  # benchmark.variants.patch      = yes
  # benchmark.variants.limit-rev  = 1
with-populate: 0.237813
no-populate:   0.229452   (-3.52%, -0.01)
  # benchmark.variants.limit-rev  = 10
  # benchmark.variants.patch      = yes
with-populate: 1.213578
no-populate:   1.205189
### data-env-vars.name            = mozilla-try-2019-02-18-zstd-sparse-revlog
  # benchmark.variants.limit-rev  = 1000
  # benchmark.variants.patch      = no
  # benchmark.variants.rev        = tip
with-populate: 0.198607
no-populate:   0.195038   (-1.80%, -0.00)

However pre-populating provide a significant boost on more complex operations
like unbundle or push:

### data-env-vars.name                             = mozilla-try-2019-02-18-zstd-sparse-revlog
  # benchmark.name                                 = hg.command.push
  # benchmark.variants.explicit-rev                = none
  # benchmark.variants.issue6528                   = disabled
  # benchmark.variants.protocol                    = ssh
  # benchmark.variants.reuse-external-delta-parent = yes
  # benchmark.variants.revs                        = any-1-extra-rev
with-populate: 4.798632
no-populate:   4.953295   (+3.22%, +0.15)
  # benchmark.variants.revs                        = any-100-extra-rev
with-populate: 4.903618
no-populate:   5.014963   (+2.27%, +0.11)
 ## benchmark.name                                 = hg.command.unbundle
  # benchmark.variants.revs                        = any-1-extra-rev
with-populate: 1.423411
no-populate:   1.585365  (+11.38%, +0.16)
  # benchmark.variants.revs                        = any-100-extra-rev
with-populate: 1.537909
no-populate:   1.688489   (+9.79%, +0.15)

==================================
Basic testing for the push command
==================================

Testing of the '--rev' flag
===========================

  $ hg init test-revflag
  $ hg -R test-revflag unbundle "$TESTDIR/bundles/remote.hg"
  adding changesets
  adding manifests
  adding file changes
  added 9 changesets with 7 changes to 4 files (+1 heads)
  new changesets bfaf4b5cbf01:916f1afdef90 (9 drafts)
  (run 'hg heads' to see heads, 'hg merge' to merge)

  $ for i in 0 1 2 3 4 5 6 7 8; do
  >    echo
  >    hg init test-revflag-"$i"
  >    hg -R test-revflag push -r "$i" test-revflag-"$i"
  >    hg -R test-revflag-"$i" verify -q
  > done
  
  pushing to test-revflag-0
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  added 1 changesets with 1 changes to 1 files
  
  pushing to test-revflag-1
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  added 2 changesets with 2 changes to 1 files
  
  pushing to test-revflag-2
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  added 3 changesets with 3 changes to 1 files
  
  pushing to test-revflag-3
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  added 4 changesets with 4 changes to 1 files
  
  pushing to test-revflag-4
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  added 2 changesets with 2 changes to 1 files
  
  pushing to test-revflag-5
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  added 3 changesets with 3 changes to 1 files
  
  pushing to test-revflag-6
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  added 4 changesets with 5 changes to 2 files
  
  pushing to test-revflag-7
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  added 5 changesets with 6 changes to 3 files
  
  pushing to test-revflag-8
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  added 5 changesets with 5 changes to 2 files

  $ cd test-revflag-8

  $ hg pull ../test-revflag-7
  pulling from ../test-revflag-7
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  added 4 changesets with 2 changes to 3 files (+1 heads)
  new changesets c70afb1ee985:faa2e4234c7a
  (run 'hg heads' to see heads, 'hg merge' to merge)

  $ hg verify -q

  $ cd ..

Test server side validation during push
=======================================

  $ hg init test-validation
  $ cd test-validation

  $ cat > .hg/hgrc <<EOF
  > [server]
  > validate=1
  > EOF

  $ echo alpha > alpha
  $ echo beta > beta
  $ hg addr
  adding alpha
  adding beta
  $ hg ci -m 1

  $ cd ..
  $ hg clone test-validation test-validation-clone
  updating to branch default
  2 files updated, 0 files merged, 0 files removed, 0 files unresolved

#if reporevlogstore

Test spurious filelog entries:

  $ cd test-validation-clone
  $ echo blah >> beta
  $ cp .hg/store/data/beta.i tmp1
  $ hg ci -m 2
  $ cp .hg/store/data/beta.i tmp2
  $ hg -q rollback
  $ mv tmp2 .hg/store/data/beta.i
  $ echo blah >> beta
  $ hg ci -m '2 (corrupt)'

Expected to fail:

  $ hg verify -q
   beta@1: dddc47b3ba30 not in manifests
  not checking dirstate because of previous errors
  1 integrity errors encountered!
  (first damaged changeset appears to be 1)
  [1]

  $ hg push
  pushing to $TESTTMP/test-validation
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  transaction abort!
  rollback completed
  abort: received spurious file revlog entry
  [255]

  $ hg -q rollback
  $ mv tmp1 .hg/store/data/beta.i
  $ echo beta > beta

Test missing filelog entries:

  $ cp .hg/store/data/beta.i tmp
  $ echo blah >> beta
  $ hg ci -m '2 (corrupt)'
  $ mv tmp .hg/store/data/beta.i

Expected to fail:

  $ hg verify -q
   beta@1: manifest refers to unknown revision dddc47b3ba30
  not checking dirstate because of previous errors
  1 integrity errors encountered!
  (first damaged changeset appears to be 1)
  [1]

  $ hg push
  pushing to $TESTTMP/test-validation
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  transaction abort!
  rollback completed
  abort: missing file data for beta:dddc47b3ba30e54484720ce0f4f768a0f4b6efb9 - run hg verify
  [255]

  $ cd ..

#endif

Test push hook locking
=====================

  $ hg init 1

  $ echo '[ui]' >> 1/.hg/hgrc
  $ echo 'timeout = 10' >> 1/.hg/hgrc

  $ echo foo > 1/foo
  $ hg --cwd 1 ci -A -m foo
  adding foo

  $ hg clone 1 2
  updating to branch default
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved

  $ hg clone 2 3
  updating to branch default
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved

  $ cat <<EOF > $TESTTMP/debuglocks-pretxn-hook.sh
  > hg debuglocks
  > true
  > EOF
  $ echo '[hooks]' >> 2/.hg/hgrc
  $ echo "pretxnchangegroup.a = sh $TESTTMP/debuglocks-pretxn-hook.sh" >> 2/.hg/hgrc
  $ echo 'changegroup.push = hg push -qf ../1' >> 2/.hg/hgrc

  $ echo bar >> 3/foo
  $ hg --cwd 3 ci -m bar

  $ hg --cwd 3 push ../2 --config devel.legacy.exchange=bundle1
  pushing to ../2
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  lock:  user *, process * (*s) (glob)
  wlock: free
  added 1 changesets with 1 changes to 1 files

  $ hg --cwd 1 --config extensions.strip= strip tip -q
  $ hg --cwd 2 --config extensions.strip= strip tip -q
  $ hg --cwd 3 push ../2 # bundle2+
  pushing to ../2
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  lock:  user *, process * (*s) (glob)
  wlock: user *, process * (*s) (glob)
  added 1 changesets with 1 changes to 1 files

Test bare push with multiple race checking options
--------------------------------------------------

  $ hg init test-bare-push-no-concurrency
  $ hg init test-bare-push-unrelated-concurrency
  $ hg -R test-revflag push -r 0 test-bare-push-no-concurrency --config server.concurrent-push-mode=strict
  pushing to test-bare-push-no-concurrency
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  added 1 changesets with 1 changes to 1 files
  $ hg -R test-revflag push -r 0 test-bare-push-unrelated-concurrency --config server.concurrent-push-mode=check-related
  pushing to test-bare-push-unrelated-concurrency
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  added 1 changesets with 1 changes to 1 files

SEC: check for unsafe ssh url

  $ cat >> $HGRCPATH << EOF
  > [ui]
  > ssh = sh -c "read l; read l; read l"
  > EOF

  $ hg -R test-revflag push 'ssh://-oProxyCommand=touch${IFS}owned/path'
  pushing to ssh://-oProxyCommand%3Dtouch%24%7BIFS%7Downed/path
  abort: potentially unsafe url: 'ssh://-oProxyCommand=touch${IFS}owned/path'
  [255]
  $ hg -R test-revflag push 'ssh://%2DoProxyCommand=touch${IFS}owned/path'
  pushing to ssh://-oProxyCommand%3Dtouch%24%7BIFS%7Downed/path
  abort: potentially unsafe url: 'ssh://-oProxyCommand=touch${IFS}owned/path'
  [255]
  $ hg -R test-revflag push 'ssh://fakehost|touch${IFS}owned/path'
  pushing to ssh://fakehost%7Ctouch%24%7BIFS%7Downed/path
  abort: no suitable response from remote hg
  [255]
  $ hg -R test-revflag push 'ssh://fakehost%7Ctouch%20owned/path'
  pushing to ssh://fakehost%7Ctouch%20owned/path
  abort: no suitable response from remote hg
  [255]

  $ [ ! -f owned ] || echo 'you got owned'

Test `commands.push.require-revs`
---------------------------------

  $ hg clone -q test-revflag test-require-revs-source
  $ hg init test-require-revs-dest
  $ cd test-require-revs-source
  $ cat >> .hg/hgrc << EOF
  > [paths]
  > default = ../test-require-revs-dest
  > [commands]
  > push.require-revs=1
  > EOF
  $ hg push
  pushing to $TESTTMP/test-require-revs-dest
  abort: no revisions specified to push
  (did you mean "hg push -r ."?)
  [10]
  $ hg push -r 0
  pushing to $TESTTMP/test-require-revs-dest
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  added 1 changesets with 1 changes to 1 files
  $ hg bookmark -r 0 push-this-bookmark
(test that -B (bookmark) works for specifying "revs")
  $ hg push -B push-this-bookmark
  pushing to $TESTTMP/test-require-revs-dest
  searching for changes
  no changes found
  exporting bookmark push-this-bookmark
  [1]
(test that -b (branch)  works for specifying "revs")
  $ hg push -b default
  pushing to $TESTTMP/test-require-revs-dest
  searching for changes
  abort: push creates new remote head [0-9a-f]+ (re)
  (merge or see 'hg help push' for details about pushing new heads)
  [20]
(demonstrate that even though we don't have anything to exchange, we're still
showing the error)
  $ hg push
  pushing to $TESTTMP/test-require-revs-dest
  abort: no revisions specified to push
  (did you mean "hg push -r ."?)
  [10]
  $ hg push --config paths.default:pushrev=0
  pushing to $TESTTMP/test-require-revs-dest
  searching for changes
  no changes found
  [1]