view tests/test-clone-pull-corruption.t @ 40479:197f092b2cd9 stable 4.8.2

server: always close http socket if responding with an error (issue6033)

It's possible for hgweb to respond _very_ early with an error if we're
catching certain types of errors. When we do, we need to tell the client
the socket is toast when there's a POST involved because otherwise there
can be lingering POST data on the socket that will confuse any future
requests on the socket. This manifested as a flaky failure on Linux in an
lfs extension test and a reliable failure on FreeBSD. With this patch
applied, test-lfs-serve-access.t now passes for me on FreeBSD.

Differential Revision: https://phab.mercurial-scm.org/D5498

Corrupt an hg repo with a pull started during an aborted commit
Create two repos, so that one of them can pull from the other one.

  $ hg init source
  $ cd source
  $ touch foo
  $ hg add foo
  $ hg ci -m 'add foo'
  $ hg clone . ../corrupted
  updating to branch default
  1 files updated, 0 files merged, 0 files removed, 0 files unresolved
  $ echo >> foo
  $ hg ci -m 'change foo'

Add a hook to wait 5 seconds and then abort the commit

  $ cd ../corrupted
  $ echo "[hooks]" >> .hg/hgrc
  $ echo 'pretxncommit = sh -c "sleep 5; exit 1"' >> .hg/hgrc

start a commit...

  $ touch bar
  $ hg add bar
  $ hg ci -m 'add bar' &

... and start a pull while the commit is still running

  $ sleep 1
  $ hg pull ../source 2>/dev/null
  pulling from ../source
  transaction abort!
  rollback completed
  abort: pretxncommit hook exited with status 1
  searching for changes
  adding changesets
  adding manifests
  adding file changes
  added 1 changesets with 1 changes to 1 files
  new changesets 52998019f625
  (run 'hg update' to get a working copy)

see what happened

  $ wait
  $ hg verify
  checking changesets
  checking manifests
  crosschecking files in changesets and manifests
  checking files
  checked 2 changesets with 2 changes to 1 files

  $ cd ..