view tests/test-log-bookmark.t @ 46607:e9901d01d135

revlog: add a mechanism to verify expected file position before appending

If someone uses `hg debuglocks`, or some non-hg process writes to the .hg
directory without respecting the locks, or if the repo's on a networked
filesystem, it's possible for the revlog code to write out corrupted data.

The form of this corruption can vary depending on what data was written and how
that happened. We are in the "networked filesystem" case (though I've had users
also do this to themselves with the "`hg debuglocks`" scenario), and most often
see this with the changelog. What ends up happening is we produce two items
(let's call them rev1 and rev2) in the .i file that have the same linkrev,
baserev, and offset into the .d file, while the data in the .d file is appended
properly. rev2's compressed_size is accurate for rev2, but when we go to
decompress the data in the .d file, we use the offset that's recorded in the
index file, which is the same as rev1, and attempt to decompress
rev2.compressed_size bytes of rev1's data. This usually does not succeed. :)

When using inline data, this also fails, though I haven't investigated why too
closely. This shows up as a "patch decode" error. I believe what's happening
there is that we're basically ignoring the offset field, getting the data
properly, but since baserev != rev, it thinks this is a delta based on rev
(instead of a full text) and can't actually apply it as such.

For now, I'm going to make this an optional component and default it to entirely
off. I may increase the default severity of this in the future, once I've
enabled it for my users and we gain more experience with it. Luckily, most of my
users have a versioned filesystem and can roll back to before the corruption has
been written, it's just a hassle to do so and not everyone knows how (so it's a
support burden). Users on other filesystems will not have that luxury, and this
can cause them to have a corrupted repository that they are unlikely to know how
to resolve, and they'll see this as a data-loss event. Refusing to create the
corruption is a much better user experience.

This mechanism is not perfect. There may be false-negatives (racy writes that
are not detected). There should not be any false-positives (non-racy writes that
are detected as such). This is not a mechanism that makes putting a repo on a
networked filesystem "safe" or "supported", just *less* likely to cause
corruption.

Differential Revision: https://phab.mercurial-scm.org/D9952

Test 'hg log' with a bookmark


Create the repository

  $ hg init Test-D8973
  $ cd Test-D8973
  $ echo "bar" > foo.txt
  $ hg add foo.txt
  $ hg commit -m "Add foo in 'default'"


Add a bookmark for topic X

  $ hg branch -f sebhtml
  marked working directory as branch sebhtml
  (branches are permanent and global, did you want a bookmark?)

  $ hg bookmark sebhtml/99991-topic-X
  $ hg up sebhtml/99991-topic-X
  0 files updated, 0 files merged, 0 files removed, 0 files unresolved

  $ echo "X" > x.txt
  $ hg add x.txt
  $ hg commit -m "Add x.txt in 'sebhtml/99991-topic-X'"

  $ hg log -B sebhtml/99991-topic-X
  changeset:   1:29f39dea9bf9
  branch:      sebhtml
  bookmark:    sebhtml/99991-topic-X
  tag:         tip
  user:        test
  date:        Thu Jan 01 00:00:00 1970 +0000
  summary:     Add x.txt in 'sebhtml/99991-topic-X'
  

Add a bookmark for topic Y

  $ hg update default
  0 files updated, 0 files merged, 1 files removed, 0 files unresolved
  (leaving bookmark sebhtml/99991-topic-X)

  $ echo "Y" > y.txt
  $ hg add y.txt
  $ hg branch -f sebhtml
  marked working directory as branch sebhtml
  $ hg bookmark sebhtml/99992-topic-Y
  $ hg up sebhtml/99992-topic-Y
  0 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ hg commit -m "Add y.txt in 'sebhtml/99992-topic-Y'"
  created new head

  $ hg log -B sebhtml/99992-topic-Y
  changeset:   2:11df7969cf8d
  branch:      sebhtml
  bookmark:    sebhtml/99992-topic-Y
  tag:         tip
  parent:      0:eaea25376a59
  user:        test
  date:        Thu Jan 01 00:00:00 1970 +0000
  summary:     Add y.txt in 'sebhtml/99992-topic-Y'
  

The log of topic Y does not interfere with the log of topic X

  $ hg log -B sebhtml/99991-topic-X
  changeset:   1:29f39dea9bf9
  branch:      sebhtml
  bookmark:    sebhtml/99991-topic-X
  user:        test
  date:        Thu Jan 01 00:00:00 1970 +0000
  summary:     Add x.txt in 'sebhtml/99991-topic-X'
  

Merge topics Y and X in the default branch

  $ hg update default
  0 files updated, 0 files merged, 1 files removed, 0 files unresolved
  (leaving bookmark sebhtml/99992-topic-Y)

  $ hg bookmark
     sebhtml/99991-topic-X     1:29f39dea9bf9
     sebhtml/99992-topic-Y     2:11df7969cf8d

  $ hg merge sebhtml/99992-topic-Y
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  (branch merge, don't forget to commit)

  $ hg commit -m "Merge branch 'sebhtml/99992-topic-Y' into 'default'"

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

  $ hg merge sebhtml/99991-topic-X
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  (branch merge, don't forget to commit)

  $ hg commit -m "Merge branch 'sebhtml/99991-topic-X' into 'default'"


Check the log of topic X, topic Y, and default branch

  $ hg log -B sebhtml/99992-topic-Y

  $ hg log -B sebhtml/99991-topic-X

  $ hg log -b default
  changeset:   4:c26ba8c1e1cb
  tag:         tip
  parent:      3:2189f3fb90d6
  parent:      1:29f39dea9bf9
  user:        test
  date:        Thu Jan 01 00:00:00 1970 +0000
  summary:     Merge branch 'sebhtml/99991-topic-X' into 'default'
  
  changeset:   3:2189f3fb90d6
  parent:      0:eaea25376a59
  parent:      2:11df7969cf8d
  user:        test
  date:        Thu Jan 01 00:00:00 1970 +0000
  summary:     Merge branch 'sebhtml/99992-topic-Y' into 'default'
  
  changeset:   0:eaea25376a59
  user:        test
  date:        Thu Jan 01 00:00:00 1970 +0000
  summary:     Add foo in 'default'
  

Set up multiple bookmarked heads:

  $ hg bookmark merged-head
  $ hg up 1
  0 files updated, 0 files merged, 1 files removed, 0 files unresolved
  (leaving bookmark merged-head)
  $ echo "Z" > z.txt
  $ hg ci -Am 'Add Z'
  adding z.txt
  $ hg bookmark topic-Z

  $ hg log -GT '{rev}: {branch}, {bookmarks}\n'
  @  5: sebhtml, topic-Z
  |
  | o  4: default, merged-head
  |/|
  | o    3: default,
  | |\
  | | o  2: sebhtml, sebhtml/99992-topic-Y
  | |/
  o |  1: sebhtml, sebhtml/99991-topic-X
  |/
  o  0: default,
  

Multiple revisions under bookmarked head:

  $ hg log -GT '{rev}: {branch}, {bookmarks}\n' -B merged-head
  o    4: default, merged-head
  |\
  | ~
  o    3: default,
  |\
  ~ ~

Follows multiple bookmarks:

  $ hg log -GT '{rev}: {branch}, {bookmarks}\n' -B merged-head -B topic-Z
  @  5: sebhtml, topic-Z
  |
  ~
  o    4: default, merged-head
  |\
  | ~
  o    3: default,
  |\
  ~ ~

Filter by bookmark and branch:

  $ hg log -GT '{rev}: {branch}, {bookmarks}\n' -B merged-head -B topic-Z -b default
  o    4: default, merged-head
  |\
  | ~
  o    3: default,
  |\
  ~ ~


Unknown bookmark:

  $ hg log -B unknown
  abort: bookmark 'unknown' does not exist
  [255]

Shouldn't accept string-matcher syntax:

  $ hg log -B 're:.*'
  abort: bookmark 're:.*' does not exist
  [255]