view tests/test-merge-combination.t @ 44363:f7459da77f23

nodemap: introduce an option to use mmap to read the nodemap mapping

The performance and memory benefit is much greater if we don't have to copy all
the data in memory for each information. So we introduce an option (on by
default) to read the data using mmap.

This changeset is the last one definition the API for index support nodemap
data. (they have to be able to use the mmaping).

Below are some benchmark comparing the best we currently have in 5.3 with the
final step of this series (using the persistent nodemap implementation in
Rust). The benchmark run `hg perfindex` with various revset and the following
variants:

Before:

 * do not use the persistent nodemap
 * use the CPython implementation of the index for nodemap
 * use mmapping of the changelog index

After:

 * use the MixedIndex Rust code, with the NodeTree object for nodemap access
   (still in review)
 * use the persistent nodemap data from disk
 * access the persistent nodemap data through mmap
 * use mmapping of the changelog index

The persistent nodemap greatly speed up  most operation on very large
repositories. Some of the previously very fast lookup end up a bit slower because
the persistent nodemap has to be setup. However the absolute slowdown is very
small and won't matters in the big picture.

Here are some numbers (in seconds) for the reference copy of mozilla-try:

Revset            Before   After     abs-change speedup

-10000:           0.004622 0.005532   0.000910  ×   0.83
-10:              0.000050 0.000132   0.000082  ×   0.37
tip               0.000052 0.000085   0.000033  ×   0.61
0 + (-10000:)     0.028222 0.005337  -0.022885  ×   5.29
0                 0.023521 0.000084  -0.023437  × 280.01
(-10000:) + 0     0.235539 0.005308  -0.230231  ×  44.37
(-10:) + :9       0.232883 0.000180  -0.232703  ×1293.79
(-10000:) + (:99) 0.238735 0.005358  -0.233377  ×  44.55
:99 + (-10000:)   0.317942 0.005593  -0.312349  ×  56.84
:9 + (-10:)       0.313372 0.000179  -0.313193  ×1750.68
:9                0.316450 0.000143  -0.316307  ×2212.93


On smaller repositories, the cost of nodemap related operation is not as big, so
the win is much more modest. Yet it helps shaving a handful of millisecond here
and there.

Here are some numbers (in seconds) for the reference copy of mercurial:

Revset            Before   After     abs-change speedup
-10:              0.000065 0.000097   0.000032  × 0.67
tip               0.000063 0.000078   0.000015  × 0.80
0                 0.000561 0.000079  -0.000482  × 7.10
-10000:           0.004609 0.003648  -0.000961  × 1.26
0 + (-10000:)     0.005023 0.003715  -0.001307  × 1.35
(-10:) + :9       0.002187 0.000108  -0.002079  ×20.25
(-10000:) + 0     0.006252 0.003716  -0.002536  × 1.68
(-10000:) + (:99) 0.006367 0.003707  -0.002660  × 1.71
:9 + (-10:)       0.003846 0.000110  -0.003736  ×34.96
:9                0.003854 0.000099  -0.003755  ×38.92
:99 + (-10000:)   0.007644 0.003778  -0.003866  × 2.02

Differential Revision: https://phab.mercurial-scm.org/D7894

This file shows what hg says are "modified" files for a merge commit
(hg log -T {files}), somewhat exhaustively.
It shows merges that involves files contents changing, and merges that
involve executable bit changing, but not merges with multiple or zero
merge ancestors, nor copies/renames, and nor identical file contents
with different filelog revisions.

genmerges is the workhorse. Given:
- a range function describing the possible values for file a
- a isgood function to filter out uninteresting combination
- a createfile function to actually write the values for file a on the
filesystem
it print a series of lines that look like: abcd C: output of -T {files}
describing the file a at respectively the base, p2, p1, merge
revision. "C" indicates that hg merge had conflicts.
  $ genmerges () {
  >   for base in `range` -; do
  >     for r1 in `range $base` -; do
  >       for r2 in `range $base $r1` -; do
  >         for m in `range $base $r1 $r2` -; do
  >           line="$base$r1$r2$m"
  >           isgood $line || continue
  >           hg init repo
  >           cd repo
  >           make_commit () {
  >             v=$1; msg=$2; file=$3;
  >             if [ $v != - ]; then
  >               createfile $v
  >             else
  >               if [ -f a ]
  >               then rm a
  >               else touch $file
  >               fi
  >             fi
  >             hg commit -q -Am $msg || exit 123
  >           }
  >           echo foo > foo
  >           make_commit $base base b
  >           make_commit $r1 r1 c
  >           hg up -r 0 -q
  >           make_commit $r2 r2 d
  >           hg merge -q -r 1 > ../output 2>&1
  >           if [ $? -ne 0 ]; then rm -f *.orig; hg resolve -m --all -q; fi
  >           if [ -s ../output ]; then conflicts=" C"; else conflicts="  "; fi
  >           make_commit $m m e
  >           if [ $m = $r1 ] && [ $m = $r2 ]
  >           then expected=
  >           elif [ $m = $r1 ]
  >           then if [ $base = $r2 ]
  >                then expected=
  >                else expected=a
  >                fi
  >           elif [ $m = $r2 ]
  >           then if [ $base = $r1 ]
  >                then expected=
  >                else expected=a
  >                fi
  >           else expected=a
  >           fi
  >           got=`hg log -r 3 --template '{files}\n' | tr -d 'e '`
  >           if [ "$got" = "$expected" ]
  >           then echo "$line$conflicts: agree on \"$got\""
  >           else echo "$line$conflicts: hg said \"$got\", expected \"$expected\""
  >           fi
  >           cd ../
  >           rm -rf repo
  >         done
  >       done
  >     done
  >   done
  > }

All the merges of various file contents.

  $ range () {
  >   max=0
  >   for i in $@; do
  >     if [ $i = - ]; then continue; fi
  >     if [ $i -gt $max ]; then max=$i; fi
  >   done
  >   $TESTDIR/seq.py `expr $max + 1`
  > }
  $ isgood () { true; }
  $ createfile () {
  >   if [ -f a ] && [ "`cat a`" = $1 ]
  >   then touch $file
  >   else echo $v > a
  >   fi
  > }

  $ genmerges
  1111  : agree on ""
  1112  : agree on "a"
  111-  : agree on "a"
  1121  : agree on "a"
  1122  : agree on ""
  1123  : agree on "a"
  112-  : agree on "a"
  11-1  : hg said "", expected "a"
  11-2  : agree on "a"
  11--  : agree on ""
  1211  : agree on "a"
  1212  : agree on ""
  1213  : agree on "a"
  121-  : agree on "a"
  1221  : agree on "a"
  1222  : agree on ""
  1223  : agree on "a"
  122-  : agree on "a"
  1231 C: agree on "a"
  1232 C: agree on "a"
  1233 C: agree on "a"
  1234 C: agree on "a"
  123- C: agree on "a"
  12-1 C: agree on "a"
  12-2 C: hg said "", expected "a"
  12-3 C: agree on "a"
  12-- C: agree on "a"
  1-11  : hg said "", expected "a"
  1-12  : agree on "a"
  1-1-  : agree on ""
  1-21 C: agree on "a"
  1-22 C: hg said "", expected "a"
  1-23 C: agree on "a"
  1-2- C: agree on "a"
  1--1  : agree on "a"
  1--2  : agree on "a"
  1---  : agree on ""
  -111  : agree on ""
  -112  : agree on "a"
  -11-  : agree on "a"
  -121 C: agree on "a"
  -122 C: agree on "a"
  -123 C: agree on "a"
  -12- C: agree on "a"
  -1-1  : agree on ""
  -1-2  : agree on "a"
  -1--  : agree on "a"
  --11  : agree on ""
  --12  : agree on "a"
  --1-  : agree on "a"
  ---1  : agree on "a"
  ----  : agree on ""

All the merges of executable bit.

  $ range () {
  >   max=a
  >   for i in $@; do
  >     if [ $i = - ]; then continue; fi
  >     if [ $i > $max ]; then max=$i; fi
  >   done
  >   if [ $max = a ]; then echo f; else echo f x; fi
  > }
  $ isgood () { case $line in *f*x*) true;; *) false;; esac; }
  $ createfile () {
  >   if [ -f a ] && (([ -x a ] && [ $v = x ]) || (! [ -x a ] && [ $v != x ]))
  >   then touch $file
  >   else touch a; if [ $v = x ]; then chmod +x a; else chmod -x a; fi
  >   fi
  > }

#if execbit
  $ genmerges
  fffx  : agree on "a"
  ffxf  : agree on "a"
  ffxx  : agree on ""
  ffx-  : agree on "a"
  ff-x  : hg said "", expected "a"
  fxff  : hg said "", expected "a"
  fxfx  : hg said "a", expected ""
  fxf-  : agree on "a"
  fxxf  : agree on "a"
  fxxx  : agree on ""
  fxx-  : agree on "a"
  fx-f  : hg said "", expected "a"
  fx-x  : hg said "", expected "a"
  fx--  : hg said "", expected "a"
  f-fx  : agree on "a"
  f-xf  : agree on "a"
  f-xx  : hg said "", expected "a"
  f-x-  : agree on "a"
  f--x  : agree on "a"
  -ffx  : agree on "a"
  -fxf C: agree on "a"
  -fxx C: hg said "", expected "a"
  -fx- C: agree on "a"
  -f-x  : hg said "", expected "a"
  --fx  : agree on "a"
#endif

Files modified or cleanly merged, with no greatest common ancestors:

  $ hg init repo; cd repo
  $ touch a0 b0; hg commit -qAm 0
  $ hg up -qr null; touch a1 b1; hg commit -qAm 1
  $ hg merge -qr 0; rm b*; hg commit -qAm 2
  $ hg log -r . -T '{files}\n'
  b0 b1
  $ cd ../
  $ rm -rf repo

A few cases of criss-cross merges involving deletions (listing all
such merges is probably too much). Both gcas contain $files, so we
expect the final merge to behave like a merge with a single gca
containing $files.

  $ hg init repo; cd repo
  $ files="c1 u1 c2 u2"
  $ touch $files; hg commit -qAm '0 root'
  $ for f in $files; do echo f > $f; done; hg commit -qAm '1 gca1'
  $ hg up -qr0; hg revert -qr 1 --all; hg commit -qAm '2 gca2'
  $ hg up -qr 1; hg merge -qr 2; rm *1; hg commit -qAm '3 p1'
  $ hg up -qr 2; hg merge -qr 1; rm *2; hg commit -qAm '4 p2'
  $ hg merge -qr 3; echo f > u1; echo f > u2; rm -f c1 c2
  $ hg commit -qAm '5 merge with two gcas'
  $ hg log -r . -T '{files}\n' # expecting u1 u2
  
  $ cd ../
  $ rm -rf repo