view tests/test-hup.t @ 22040:122fa73657c6

shelve: do not retract phase boundary by hand We rely on the internal mechanism to commit the changeset in the right state. This is similar to what the mq extension is doing. This is an important change as we plan to move phase movement with the transaction. Avoiding phase movement from high level code will avoid them the burden of transaction handling. It is also important to limit the need for transaction handling as this limits the odds of people messing up. Most common expected mess-up is to use a different transaction for changesets creation and phase adjustment.
author Pierre-Yves David <pierre-yves.david@fb.com>
date Tue, 05 Aug 2014 14:37:45 -0700
parents 9fae01831dea
children 7a9cbb315d84
line wrap: on
line source

Test hangup signal in the middle of transaction

  $ "$TESTDIR/hghave" serve fifo || exit 80
  $ hg init
  $ mkfifo p
  $ hg serve --stdio < p 1>out 2>&1 &
  $ P=$!

Do test while holding fifo open

  $ (
  > echo lock
  > echo addchangegroup
  > start=`date +%s`
  > # 10 second seems much enough to let the server catch up
  > deadline=`expr $start + 10`
  > while [ ! -s .hg/store/journal ]; do
  >     sleep 0;
  >     if [ `date +%s` -gt $deadline ]; then
  >         echo "transaction did not start after 10 seconds" >&2;
  >         exit 1;
  >     fi
  > done
  > kill -HUP $P
  > ) > p

  $ wait
  $ cat out
  0
  0
  adding changesets
  transaction abort!
  rollback completed
  killed!

  $ echo .hg/* .hg/store/*
  .hg/00changelog.i .hg/journal.bookmarks .hg/journal.branch .hg/journal.desc .hg/journal.dirstate .hg/requires .hg/store .hg/store/00changelog.i .hg/store/00changelog.i.a .hg/store/journal.phaseroots