tests/test-strip-branch-cache.t
author Arun Kulshreshtha <akulshreshtha@janestreet.com>
Tue, 30 Aug 2022 15:29:55 -0400
changeset 49491 c6a1beba27e9
parent 49071 9caf23927d04
child 51492 2e8a88e5809f
permissions -rw-r--r--
bisect: avoid copying ancestor list for non-merge commits During a bisection, hg needs to compute a list of all ancestors for every candidate commit. This is accomplished via a bottom-up traversal of the set of candidates, during which each revision's ancestor list is populated using the ancestor list of its parent(s). Previously, this involved copying the entire list, which could be very long in if the bisection range was large. To help improve this, we can observe that each candidate commit is visited exactly once, at which point its ancestor list is copied into its children's lists and then dropped. In the case of non-merge commits, a commit's ancestor list consists exactly of its parent's list plus itself. This means that we can trivially reuse the parent's existing list for one of its non-merge children, which avoids copying entirely if that commit is the parent's only child. This makes bisections over linear ranges of commits much faster. During some informal testing in the large publicly-available `mozilla-central` repository, this noticeably sped up bisections over large ranges of history: Setup: $ cd mozilla-central $ hg bisect --reset $ hg bisect --good 0 $ hg log -r tip -T '{rev}\n' 628417 Test: $ time hg bisect --bad tip --noupdate Before: real 3m35.927s user 3m35.553s sys 0m0.319s After: real 1m41.142s user 1m40.810s sys 0m0.285s

Define helpers.

  $ hg_log () { hg log -G -T "{rev}:{node|short}"; }
  $ commit () { echo "foo - ${2:-$1}" > $1; hg commit -Aqm "Edited $1"; }
  $ strip() { hg --config extensions.strip= strip -q -r "$1" ; }

Setup hg repo.

  $ hg init repo
  $ cd repo
  $ touch x; hg add x; hg commit -m "initial"
  $ hg clone -q . ../clone
  $ commit a

  $ cd ../clone

  $ commit b

  $ hg pull -q ../repo

  $ cat .hg/cache/branch2-visible
  222ae9789a75703f9836e44de7db179cbfd420ee 2
  a3498d6e39376d2456425dd8c692367bdbf00fa2 o default
  222ae9789a75703f9836e44de7db179cbfd420ee o default

  $ hg_log
  o  2:222ae9789a75
  |
  | @  1:a3498d6e3937
  |/
  o  0:7ab0a3bd758a
  

  $ strip '1:'

The branchmap cache is not adjusted on strip.
Now mentions a changelog entry that has been stripped.

  $ cat .hg/cache/branch2-visible
  222ae9789a75703f9836e44de7db179cbfd420ee 2
  a3498d6e39376d2456425dd8c692367bdbf00fa2 o default
  222ae9789a75703f9836e44de7db179cbfd420ee o default

  $ commit c

Not adjusted on commit, either.

  $ cat .hg/cache/branch2-visible
  222ae9789a75703f9836e44de7db179cbfd420ee 2
  a3498d6e39376d2456425dd8c692367bdbf00fa2 o default
  222ae9789a75703f9836e44de7db179cbfd420ee o default

On pull we end up with the same tip, and so wrongly reuse the invalid cache and crash.

  $ hg pull ../repo 2>&1 | grep 'ValueError:'
  ValueError: node a3498d6e39376d2456425dd8c692367bdbf00fa2 does not exist (known-bad-output !)