view tests/test-rebase-mq-skip.t @ 26457:7e81305092a0

demandimport: replace more references to _demandmod instances

_demandmod instances may be referenced by multiple importing modules.
Before this patch, the _demandmod instance only maintained a reference
to its first consumer when using the "from X import Y" syntax. This is
because we only created a single _demandmod instance (attached to the
parent X module). If multiple modules A and B performed
"from X import Y", we'd produce a single _demandmod instance
"demandmod" with the following references:

  X.Y = <demandmod>
  A.Y = <demandmod>
  B.Y = <demandmod>

The locals from the first consumer (A) would be stored in <demandmod1>.
When <demandmod1> was loaded, we'd look at the locals for the first
consumer and replace the symbol, if necessary. This resulted in state:

  X.Y = <module>
  A.Y = <module>
  B.Y = <demandmod>

B's reference to Y wasn't updated and was still using the proxy object
because we just didn't record that B had a reference to <demandmod> that
needed updating!

With this patch, we add support for tracking which modules in addition
to the initial importer have a reference to the _demandmod instance and
we replace those references at module load time.

In the case of posix.py, this fixes an issue where the "encoding" module
was being proxied, resulting in hundreds of thousands of
__getattribute__ lookups on the _demandmod instance during dirstate
operations on mozilla-central, speeding up execution by many
milliseconds. There are likely several other operation that benefit from
this change as well.

The new mechanism isn't perfect: references in locals (not globals) may
likely linger. So, if there is an import inside a function and a symbol
from that module is used in a hot loop, we could have unwanted overhead
from proxying through _demandmod. Non-global imports are discouraged
anyway. So hopefully this isn't a big deal in practice. We could
potentially deploy a code checker that bans use of attribute lookups of
function-level-imported modules inside loops.

This deficiency in theory could be avoided by storing the set of globals
and locals dicts to update in the _demandmod instance. However, I tried
this and it didn't work. One reason is that some globals are _demandmod
instances. We could work around this, but it's a bit more work. There
also might be other module import foo at play. The solution as
implemented is better than what we had and IMO is good enough for the
time being.

It's worth noting that this sub-optimal behavior was made worse by the
introduction of absolute_import and its recommended "from . import X"
syntax for importing modules from the "mercurial" package. If we ever
wrote performance tests, measuring the amount of module imports and
__getattribute__ proxy calls through _demandmod instances would be
something I'd have it check.

This emulates the effects of an hg pull --rebase in which the remote repo
already has one local mq patch

  $ cat >> $HGRCPATH <<EOF
  > [extensions]
  > rebase=
  > mq=
  > 
  > [phases]
  > publish=False
  > 
  > [alias]
  > tglog = log -G --template "{rev}: '{desc}' tags: {tags}\n"
  > EOF


  $ hg init a
  $ cd a
  $ hg qinit -c

  $ echo c1 > c1
  $ hg add c1
  $ hg ci -m C1

  $ echo r1 > r1
  $ hg add r1
  $ hg ci -m R1

  $ hg up -q 0

  $ hg qnew p0.patch -d '1 0'
  $ echo p0 > p0
  $ hg add p0
  $ hg qref -m P0

  $ hg qnew p1.patch -d '2 0'
  $ echo p1 > p1
  $ hg add p1
  $ hg qref -m P1

  $ hg export qtip > p1.patch

  $ hg up -q -C 1

  $ hg import p1.patch
  applying p1.patch

  $ rm p1.patch

  $ hg up -q -C qtip

  $ hg rebase -v
  rebasing 2:13a46ce44f60 "P0" (p0.patch qbase)
  resolving manifests
  removing p0
  getting r1
  resolving manifests
  getting p0
  committing files:
  p0
  committing manifest
  committing changelog
  rebasing 3:148775c71080 "P1" (p1.patch qtip)
  resolving manifests
  note: rebase of 3:148775c71080 created no changes to commit
  rebase merging completed
  updating mq patch p0.patch to 5:9ecc820b1737
  $TESTTMP/a/.hg/patches/p0.patch (glob)
  2 changesets found
  uncompressed size of bundle content:
       344 (changelog)
       284 (manifests)
       109  p0
       109  p1
  saved backup bundle to $TESTTMP/a/.hg/strip-backup/13a46ce44f60-5da6ecfb-backup.hg (glob)
  2 changesets found
  uncompressed size of bundle content:
       399 (changelog)
       284 (manifests)
       109  p0
       109  p1
  adding branch
  adding changesets
  adding manifests
  adding file changes
  added 2 changesets with 2 changes to 2 files
  rebase completed
  1 revisions have been skipped

  $ hg tglog
  @  3: 'P0' tags: p0.patch qbase qtip tip
  |
  o  2: 'P1' tags: qparent
  |
  o  1: 'R1' tags:
  |
  o  0: 'C1' tags:
  
  $ cd ..


  $ hg init b
  $ cd b
  $ hg qinit -c

  $ for i in r0 r1 r2 r3 r4 r5 r6;
  > do
  >     echo $i > $i
  >     hg ci -Am $i
  > done
  adding r0
  adding r1
  adding r2
  adding r3
  adding r4
  adding r5
  adding r6

  $ hg qimport -r 1:tip

  $ hg up -q 0

  $ for i in r1 r3 r7 r8;
  > do
  >     echo $i > $i
  >     hg ci -Am branch2-$i
  > done
  adding r1
  created new head
  adding r3
  adding r7
  adding r8

  $ echo somethingelse > r4
  $ hg ci -Am branch2-r4
  adding r4

  $ echo r6 > r6
  $ hg ci -Am branch2-r6
  adding r6

  $ hg up -q qtip

  $ HGMERGE=internal:fail hg rebase
  rebasing 1:b4bffa6e4776 "r1" (1.diff qbase)
  note: rebase of 1:b4bffa6e4776 created no changes to commit
  rebasing 2:c0fd129beb01 "r2" (2.diff)
  rebasing 3:6ff5b8feed8e "r3" (3.diff)
  note: rebase of 3:6ff5b8feed8e created no changes to commit
  rebasing 4:094320fec554 "r4" (4.diff)
  unresolved conflicts (see hg resolve, then hg rebase --continue)
  [1]

  $ HGMERGE=internal:local hg resolve --all
  (no more unresolved files)

  $ hg rebase --continue
  already rebased 1:b4bffa6e4776 "r1" (1.diff qbase) as 057f55ff8f44
  already rebased 2:c0fd129beb01 "r2" (2.diff) as 1660ab13ce9a
  already rebased 3:6ff5b8feed8e "r3" (3.diff) as 1660ab13ce9a
  rebasing 4:094320fec554 "r4" (4.diff)
  note: rebase of 4:094320fec554 created no changes to commit
  rebasing 5:681a378595ba "r5" (5.diff)
  rebasing 6:512a1f24768b "r6" (6.diff qtip)
  note: rebase of 6:512a1f24768b created no changes to commit
  saved backup bundle to $TESTTMP/b/.hg/strip-backup/b4bffa6e4776-b9bfb84d-backup.hg (glob)

  $ hg tglog
  @  8: 'r5' tags: 5.diff qtip tip
  |
  o  7: 'r2' tags: 2.diff qbase
  |
  o  6: 'branch2-r6' tags: qparent
  |
  o  5: 'branch2-r4' tags:
  |
  o  4: 'branch2-r8' tags:
  |
  o  3: 'branch2-r7' tags:
  |
  o  2: 'branch2-r3' tags:
  |
  o  1: 'branch2-r1' tags:
  |
  o  0: 'r0' tags:
  

  $ cd ..