Mercurial > hg
view tests/test-rebase-mq.t @ 25757:4d1382fd96ff
context: write dirstate out explicitly at the end of markcommitted
To detect change of a file without redundant comparison of file
content, dirstate recognizes a file as certainly clean, if:
(1) it is already known as "normal",
(2) dirstate entry for it has valid (= not "-1") timestamp, and
(3) mode, size and timestamp of it on the filesystem are as same as
ones expected in dirstate
This works as expected in many cases, but doesn't in the corner case
that changing a file keeps mode, size and timestamp of it on the
filesystem.
The timetable below shows steps in one of typical such situations:
---- ----------------------------------- ----------------
timestamp of "f"
----------------
dirstate file-
time action mem file system
---- ----------------------------------- ---- ----- -----
* *** ***
- 'hg transplant REV1 REV2 ...'
- transplanting REV1
....
N
- change "f", but keep size N
(via 'patch.patch()')
- 'dirstate.normal("f")' N ***
(via 'repo.commit()')
- transplanting REV2
- change "f", but keep size N
(via 'patch.patch()')
- aborted while patching
N+1
- release wlock
- 'dirstate.write()' N N N
- 'hg status' shows "r1" as "clean" N N N
---- ----------------------------------- ---- ----- -----
The most important point is that 'dirstate.write()' is executed at N+1
or later. This causes writing dirstate timestamp N of "f" out
successfully. If it is executed at N, 'parsers.pack_dirstate()'
replaces timestamp N with "-1" before actual writing dirstate out.
This issue can occur when 'hg transplant' satisfies conditions below:
- multiple revisions to be transplanted change the same file
- those revisions don't change mode and size of the file, and
- the 2nd or later revision of them fails after changing the file
The root cause of this issue is that files are changed without
flushing in-memory dirstate changes via 'repo.commit()' (even though
omitting 'dirstate.normallookup()' on files changed by 'patch.patch()'
for efficiency also causes this issue).
To detect changes of files correctly, this patch writes in-memory
dirstate changes out explicitly after marking files as clean in
'committablectx.markcommitted()', which is invoked via
'repo.commit()'.
After this change, timetable is changed as below:
---- ----------------------------------- ----------------
timestamp of "f"
----------------
dirstate file-
time action mem file system
---- ----------------------------------- ---- ----- -----
* *** ***
- 'hg transplant REV1 REV2 ...'
- transplanting REV1
....
N
- change "f", but keep size N
(via 'patch.patch()')
- 'dirstate.normal("f")' N ***
(via 'repo.commit()')
----------------------------------- ---- ----- -----
- 'dirsttate.write()' -1 -1
----------------------------------- ---- ----- -----
- transplanting REV2
- change "f", but keep size N
(via 'patch.patch()')
- aborted while patching
N+1
- release wlock
- 'dirstate.write()' -1 -1 N
- 'hg status' shows "r1" as "clean" -1 -1 N
---- ----------------------------------- ---- ----- -----
To reproduce this issue in tests certainly, this patch emulates some
timing critical actions as below:
- change "f" at N
'patch.patch()' with 'fakepatchtime.py' explicitly changes mtime
of patched files to "2000-01-01 00:00" (= N).
- 'dirstate.write()' via 'repo.commit()' at N
'fakedirstatewritetime.py' forces 'pack_dirstate()' to use
"2000-01-01 00:00" as "now", only if 'pack_dirstate()' is invoked
via 'committablectx.markcommitted()'.
- 'dirstate.write()' via releasing wlock at N+1 (or "not at N")
'pack_dirstate()' via releasing wlock uses actual timestamp at
runtime as "now", and it should be different from the "2000-01-01
00:00" of "f".
BTW, this patch doesn't test cases below, even though 'patch.patch()'
is used similarly in these cases:
1. failure of 'hg import' or 'hg qpush'
2. success of 'hg import', 'hg qpush' or 'hg transplant'
Case (1) above doesn't cause this kind of issue, because:
- if patching is aborted by conflicts, changed files are committed
changed files are marked as CLEAN, even though they are partially
patched.
- otherwise, dirstate are fully restored by 'dirstateguard'
For example in timetable above, timestamp of "f" in .hg/dirstate
is restored to -1 (or less than N), and subsequent 'hg status' can
detect changes correctly.
Case (2) always causes 'repo.status()' invocation via 'repo.commit()'
just after changing files inside same wlock scope.
---- ----------------------------------- ----------------
timestamp of "f"
----------------
dirstate file-
time action mem file system
---- ----------------------------------- ---- ----- -----
N *** ***
- make file "f" clean N
- execute 'hg foobar'
....
- 'dirstate.normal("f")' N ***
(e.g. via dirty check
or previous 'repo.commit()')
- change "f", but keep size N
- 'repo.status()' (*1)
(via 'repo.commit()')
---- ----------------------------------- ---- ----- -----
At a glance, 'repo.status()' at (*1) seems to cause similar issue (=
"changed files are treated as clean"), but actually doesn't.
'dirstate._lastnormaltime' should be N at (*1) above, because
'dirstate.normal()' via dirty check is finished at N.
Therefore, "f" changed at N (= 'dirstate._lastnormaltime') is forcibly
treated as "unsure" at (*1), and changes are detected as expected (see
'dirstate.status()' for detail).
If 'hg import' is executed with '--no-commit', 'repo.status()' isn't
invoked just after changing files inside same wlock scope.
But preceding 'dirstate.normal()' is invoked inside another wlock
scope via 'cmdutil.bailifchanged()', and in-memory changes should be
flushed at the end of that scope.
Therefore, timestamp N of clean "f" should be replaced by -1, if
'dirstate.write()' is invoked at N. It means that condition of this
issue isn't satisfied.
author | FUJIWARA Katsunori <foozy@lares.dti.ne.jp> |
---|---|
date | Wed, 08 Jul 2015 17:01:09 +0900 |
parents | aa4a1672583e |
children | ef1eb6df7071 |
line wrap: on
line source
$ cat >> $HGRCPATH <<EOF > [extensions] > rebase= > mq= > > [mq] > plain=true > > [alias] > tglog = log -G --template "{rev}: '{desc}' tags: {tags}\n" > EOF $ hg init a $ cd a $ hg qinit -c $ echo c1 > f $ hg add f $ hg ci -m C1 $ echo r1 > f $ hg ci -m R1 $ hg up -q 0 $ hg qnew f.patch -d '1 0' $ echo mq1 > f $ hg qref -m P0 $ hg qnew f2.patch $ echo mq2 > f $ hg qref -m P1 -d '2 0' $ hg tglog @ 3: 'P1' tags: f2.patch qtip tip | o 2: 'P0' tags: f.patch qbase | | o 1: 'R1' tags: |/ o 0: 'C1' tags: qparent Rebase - try to rebase on an applied mq patch: $ hg rebase -s 1 -d 3 abort: cannot rebase onto an applied mq patch [255] Rebase - same thing, but mq patch is default dest: $ hg up -q 1 $ hg rebase abort: cannot rebase onto an applied mq patch [255] $ hg up -q qtip Rebase - generate a conflict: $ hg rebase -s 2 -d 1 rebasing 2:3504f44bffc0 "P0" (f.patch qbase) merging f warning: conflicts during merge. merging f incomplete! (edit conflicts, then use 'hg resolve --mark') unresolved conflicts (see hg resolve, then hg rebase --continue) [1] Fix the 1st conflict: $ echo mq1r1 > f $ hg resolve -m f (no more unresolved files) $ hg rebase -c rebasing 2:3504f44bffc0 "P0" (f.patch qbase) rebasing 3:929394423cd3 "P1" (f2.patch qtip tip) merging f warning: conflicts during merge. merging f incomplete! (edit conflicts, then use 'hg resolve --mark') unresolved conflicts (see hg resolve, then hg rebase --continue) [1] Fix the 2nd conflict: $ echo mq1r1mq2 > f $ hg resolve -m f (no more unresolved files) $ hg rebase -c already rebased 2:3504f44bffc0 "P0" (f.patch qbase) as ebe9914c0d1c rebasing 3:929394423cd3 "P1" (f2.patch qtip) saved backup bundle to $TESTTMP/a/.hg/strip-backup/3504f44bffc0-30595b40-backup.hg (glob) $ hg tglog @ 3: 'P1' tags: f2.patch qtip tip | o 2: 'P0' tags: f.patch qbase | o 1: 'R1' tags: qparent | o 0: 'C1' tags: $ hg up -q qbase $ cat f mq1r1 $ cat .hg/patches/f.patch # HG changeset patch # User test # Date 1 0 # Thu Jan 01 00:00:01 1970 +0000 # Node ID ebe9914c0d1c3f60096e952fa4dbb3d377dea3ab # Parent bac9ed9960d8992bcad75864a879fa76cadaf1b0 P0 diff -r bac9ed9960d8 -r ebe9914c0d1c f --- a/f Thu Jan 01 00:00:00 1970 +0000 +++ b/f Thu Jan 01 00:00:01 1970 +0000 @@ -1,1 +1,1 @@ -r1 +mq1r1 Update to qtip: $ hg up -q qtip $ cat f mq1r1mq2 $ cat .hg/patches/f2.patch # HG changeset patch # User test # Date 2 0 # Thu Jan 01 00:00:02 1970 +0000 # Node ID 462012cf340c97d44d62377c985a423f6bb82f07 # Parent ebe9914c0d1c3f60096e952fa4dbb3d377dea3ab P1 diff -r ebe9914c0d1c -r 462012cf340c f --- a/f Thu Jan 01 00:00:01 1970 +0000 +++ b/f Thu Jan 01 00:00:02 1970 +0000 @@ -1,1 +1,1 @@ -mq1r1 +mq1r1mq2 Adding one git-style patch and one normal: $ hg qpop -a popping f2.patch popping f.patch patch queue now empty $ rm -fr .hg/patches $ hg qinit -c $ hg up -q 0 $ hg qnew --git f_git.patch -d '3 0' $ echo mq1 > p $ hg add p $ hg qref --git -m 'P0 (git)' $ hg qnew f.patch -d '4 0' $ echo mq2 > p $ hg qref -m P1 $ hg qci -m 'save patch state' $ hg qseries -s f_git.patch: P0 (git) f.patch: P1 $ hg -R .hg/patches manifest .hgignore f.patch f_git.patch series $ cat .hg/patches/f_git.patch Date: 3 0 P0 (git) diff --git a/p b/p new file mode 100644 --- /dev/null +++ b/p @@ -0,0 +1,1 @@ +mq1 $ cat .hg/patches/f.patch Date: 4 0 P1 diff -r ???????????? p (glob) --- a/p ??? ??? ?? ??:??:?? ???? ????? (glob) +++ b/p ??? ??? ?? ??:??:?? ???? ????? (glob) @@ -1,1 +1,1 @@ -mq1 +mq2 Rebase the applied mq patches: $ hg rebase -s 2 -d 1 rebasing 2:0c587ffcb480 "P0 (git)" (f_git.patch qbase) rebasing 3:c7f18665e4bc "P1" (f.patch qtip tip) saved backup bundle to $TESTTMP/a/.hg/strip-backup/0c587ffcb480-0ea5695f-backup.hg (glob) $ hg qci -m 'save patch state' $ hg qseries -s f_git.patch: P0 (git) f.patch: P1 $ hg -R .hg/patches manifest .hgignore f.patch f_git.patch series $ cat .hg/patches/f_git.patch # HG changeset patch # User test # Date 3 0 # Thu Jan 01 00:00:03 1970 +0000 # Node ID 12d9f6a3bbe560dee50c7c454d434add7fb8e837 # Parent bac9ed9960d8992bcad75864a879fa76cadaf1b0 P0 (git) diff --git a/p b/p new file mode 100644 --- /dev/null +++ b/p @@ -0,0 +1,1 @@ +mq1 $ cat .hg/patches/f.patch # HG changeset patch # User test # Date 4 0 # Thu Jan 01 00:00:04 1970 +0000 # Node ID c77a2661c64c60d82f63c4f7aefd95b3a948a557 # Parent 12d9f6a3bbe560dee50c7c454d434add7fb8e837 P1 diff -r 12d9f6a3bbe5 -r c77a2661c64c p --- a/p Thu Jan 01 00:00:03 1970 +0000 +++ b/p Thu Jan 01 00:00:04 1970 +0000 @@ -1,1 +1,1 @@ -mq1 +mq2 $ cd .. Rebase with guards $ hg init foo $ cd foo $ echo a > a $ hg ci -Am a adding a Create mq repo with guarded patches foo and bar and empty patch: $ hg qinit $ echo guarded > guarded $ hg add guarded $ hg qnew guarded $ hg qnew empty-important -m 'important commit message' -d '1 0' $ echo bar > bar $ hg add bar $ hg qnew bar -d '2 0' $ echo foo > foo $ hg add foo $ hg qnew foo $ hg qpop -a popping foo popping bar popping empty-important popping guarded patch queue now empty $ hg qguard guarded +guarded $ hg qguard bar +baz $ hg qguard foo +baz $ hg qselect baz number of unguarded, unapplied patches has changed from 1 to 3 $ hg qpush bar applying empty-important patch empty-important is empty applying bar now at: bar $ hg qguard -l guarded: +guarded empty-important: unguarded bar: +baz foo: +baz $ hg tglog @ 2: 'imported patch bar' tags: bar qtip tip | o 1: 'important commit message' tags: empty-important qbase | o 0: 'a' tags: qparent Create new head to rebase bar onto: $ hg up -C 0 0 files updated, 0 files merged, 1 files removed, 0 files unresolved $ echo b > b $ hg add b $ hg ci -m b created new head $ hg up -C 2 1 files updated, 0 files merged, 1 files removed, 0 files unresolved $ echo a >> a $ hg qref $ hg tglog @ 3: '[mq]: bar' tags: bar qtip tip | | o 2: 'b' tags: | | o | 1: 'important commit message' tags: empty-important qbase |/ o 0: 'a' tags: qparent Rebase bar (make sure series order is preserved and empty-important also is removed from the series): $ hg qseries guarded empty-important bar foo $ [ -f .hg/patches/empty-important ] $ hg -q rebase -d 2 note: rebase of 1:0aaf4c3af7eb created no changes to commit $ hg qseries guarded bar foo $ [ -f .hg/patches/empty-important ] [1] $ hg qguard -l guarded: +guarded bar: +baz foo: +baz $ hg tglog @ 2: '[mq]: bar' tags: bar qbase qtip tip | o 1: 'b' tags: qparent | o 0: 'a' tags: $ cd ..