view tests/test-casecollision-merge.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)

#require icasefs

################################
test for branch merging
################################

test for rename awareness of case-folding collision check:

(1) colliding file is one renamed from collided file:
this is also case for issue3370.

  $ hg init branch_merge_renaming
  $ cd branch_merge_renaming

  $ echo a > a
  $ hg add a
  $ echo b > b
  $ hg add b
  $ hg commit -m '#0'
  $ hg tag -l A
  $ hg rename a tmp
  $ hg rename tmp A
  $ hg commit -m '#1'
  $ hg tag -l B
  $ hg update -q 0
  $ touch x
  $ hg add x
  $ hg commit -m '#2'
  created new head
  $ hg tag -l C

  $ hg merge -q
  $ hg status -A
  M A
  R a
  C b
  C x

  $ hg update -q --clean 1
  $ hg merge -q
  $ hg status -A
  M x
  C A
  C b
  $ hg commit -m '(D)'
  $ hg tag -l D

additional test for issue3452:

| this assumes the history below.
|
|  (A) -- (C) -- (E) -------
|      \      \             \
|       \      \             \
|         (B) -- (D) -- (F) -- (G)
|
|   A: add file 'a'
|   B: rename from 'a' to 'A'
|   C: add 'x' (or operation other than modification of 'a')
|   D: merge C into B
|   E: modify 'a'
|   F: modify 'A'
|   G: merge E into F
|
| issue3452 occurs when (B) is recorded before (C)

  $ hg update -q --clean C
  $ echo "modify 'a' at (E)" > a
  $ echo "modify 'b' at (E)" > b
  $ hg commit -m '(E)'
  created new head
  $ hg tag -l E

  $ hg update -q --clean D
  $ echo "modify 'A' at (F)" > A
  $ hg commit -m '(F)'
  $ hg tag -l F

  $ hg merge -q --tool internal:other E
  $ hg status -A
  M A
    a
  M b
  C x
  $ cat A
  modify 'a' at (E)

test also the case that (B) is recorded after (C), to prevent
regression by changes in the future.

to avoid unexpected (successful) behavior by filelog unification,
target file is not 'a'/'A' but 'b'/'B' in this case.

  $ hg update -q --clean A
  $ hg rename b tmp
  $ hg rename tmp B
  $ hg commit -m '(B1)'
  created new head
  $ hg tag -l B1

  $ hg merge -q C
  $ hg status -A
  M x
  C B
  C a
  $ hg commit -m '(D1)'
  $ hg tag -l D1

  $ echo "modify 'B' at (F1)" > B
  $ hg commit -m '(F1)'
  $ hg tag -l F1

  $ hg merge -q --tool internal:other E
  $ hg status -A
  M B
    b
  M a
  C x
  $ cat B
  modify 'b' at (E)

  $ cd ..

(2) colliding file is not related to collided file

  $ hg init branch_merge_collding
  $ cd branch_merge_collding

  $ echo a > a
  $ hg add a
  $ hg commit -m '#0'
  $ hg remove a
  $ hg commit -m '#1'
  $ echo A > A
  $ hg add A
  $ hg commit -m '#2'
  $ hg update --clean 0
  1 files updated, 0 files merged, 1 files removed, 0 files unresolved
  $ echo x > x
  $ hg add x
  $ hg commit -m '#3'
  created new head
  $ echo 'modified at #4' > a
  $ hg commit -m '#4'

  $ hg merge
  abort: case-folding collision between [aA] and [Aa] (re)
  [20]
  $ hg parents --template '{rev}\n'
  4
  $ hg status -A
  C a
  C x
  $ cat a
  modified at #4

  $ hg update --clean 2
  1 files updated, 0 files merged, 2 files removed, 0 files unresolved
  $ hg merge
  abort: case-folding collision between [aA] and [Aa] (re)
  [20]
  $ hg parents --template '{rev}\n'
  2
  $ hg status -A
  C A
  $ cat A
  A

test for deletion awareness of case-folding collision check (issue3648):
revision '#3' doesn't change 'a', so 'a' should be recognized as
safely removed in merging between #2 and #3.

  $ hg update --clean 3
  2 files updated, 0 files merged, 1 files removed, 0 files unresolved
  $ hg merge 2
  1 files updated, 0 files merged, 1 files removed, 0 files unresolved
  (branch merge, don't forget to commit)
  $ hg status -A
  M A
  R a
  C x

  $ hg update --clean 2
  1 files updated, 0 files merged, 1 files removed, 0 files unresolved
  $ hg merge 3
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  (branch merge, don't forget to commit)
  $ hg status -A
  M x
  C A

  $ cd ..

Prepare for tests of directory case-folding collisions

  $ hg init directory-casing
  $ cd directory-casing
  $ touch 0 # test: file without directory
  $ mkdir 0a
  $ touch 0a/f
  $ mkdir aA
  $ touch aA/a
  $ hg ci -Aqm0

Directory/file case-folding collision:

  $ hg up -q null
  $ touch 00 # test: starts as '0'
  $ mkdir 000 # test: starts as '0'
  $ touch 000/f
  $ touch Aa # test: collision with 'aA/a'
  $ hg ci -Aqm1

  $ hg merge 0
  abort: case-folding collision between Aa and directory of aA/a
  [20]
(note: no collision between 0 and 00 or 000/f)

Directory case-folding collision:

  $ hg up -qC null
  $ hg --config extensions.purge= purge
  $ mkdir 0A0
  $ touch 0A0/f # test: starts as '0a'
  $ mkdir Aa
  $ touch Aa/b # test: collision with 'aA/a'
  $ hg ci -Aqm2

  $ hg merge 0
  3 files updated, 0 files merged, 0 files removed, 0 files unresolved
  (branch merge, don't forget to commit)

  $ cd ..

################################
test for linear updates
################################

test for rename awareness of case-folding collision check:

(1) colliding file is one renamed from collided file

  $ hg init linearupdate_renameaware_1
  $ cd linearupdate_renameaware_1

  $ echo a > a
  $ hg add a
  $ hg commit -m '#0'
  $ hg rename a tmp
  $ hg rename tmp A
  $ hg commit -m '#1'

  $ hg update 0
  1 files updated, 0 files merged, 1 files removed, 0 files unresolved

  $ echo 'this is added line' >> a
  $ hg update 1
  merging a and A to A
  0 files updated, 1 files merged, 0 files removed, 0 files unresolved
  $ hg status -A
  M A
  $ cat A
  a
  this is added line

  $ cd ..

(2) colliding file is not related to collided file

  $ hg init linearupdate_renameaware_2
  $ cd linearupdate_renameaware_2

  $ echo a > a
  $ hg add a
  $ hg commit -m '#0'
  $ hg remove a
  $ hg commit -m '#1'
  $ echo A > A
  $ hg add A
  $ hg commit -m '#2'

  $ hg update 0
  1 files updated, 0 files merged, 1 files removed, 0 files unresolved
  $ hg parents --template '{rev}\n'
  0
  $ hg status -A
  C a
  $ cat A
  a
  $ hg up -qC 2

  $ hg update --check 0
  1 files updated, 0 files merged, 1 files removed, 0 files unresolved
  $ hg parents --template '{rev}\n'
  0
  $ hg status -A
  C a
  $ cat a
  a

  $ hg update --clean 2
  1 files updated, 0 files merged, 1 files removed, 0 files unresolved
  $ hg parents --template '{rev}\n'
  2
  $ hg status -A
  C A
  $ cat A
  A

  $ cd ..

(3) colliding file is not related to collided file: added in working dir

  $ hg init linearupdate_renameaware_3
  $ cd linearupdate_renameaware_3

  $ echo a > a
  $ hg add a
  $ hg commit -m '#0'
  $ hg rename a b
  $ hg commit -m '#1'
  $ hg update 0
  1 files updated, 0 files merged, 1 files removed, 0 files unresolved

  $ echo B > B
  $ hg add B
  $ hg status
  A B
  $ hg update
  abort: case-folding collision between [bB] and [Bb] (re)
  [20]

  $ hg update --check
  abort: uncommitted changes
  [20]

  $ hg update --clean
  1 files updated, 0 files merged, 1 files removed, 0 files unresolved
  $ hg parents --template '{rev}\n'
  1
  $ hg status -A
  C b
  $ cat b
  a

  $ cd ..