RE: [for-5.0 PATCH 03/11] migration: introduce icount field for snapshots

["Pavel Dovgalyuk" <dovgaluk@ispras.ru>]

> From: Alex Bennée [mailto:alex.bennee@linaro.org]
> Kevin Wolf <kwolf@redhat.com> writes:
> 
> > Am 23.12.2019 um 10:47 hat Pavel Dovgalyuk geschrieben:
> >> From: Pavel Dovgalyuk <Pavel.Dovgaluk@ispras.ru>
> >>
> >> Saving icount as a parameters of the snapshot allows navigation between
> >> them in the execution replay scenario.
> >> This information can be used for finding a specific snapshot for proceeding
> >> the recorded execution to the specific moment of the time.
> >> E.g., 'reverse step' action (introduced in one of the following patches)
> >> needs to load the nearest snapshot which is prior to the current moment
> >> of time.
> >>
> >> Signed-off-by: Pavel Dovgalyuk <Pavel.Dovgaluk@ispras.ru>
> >> Acked-by: Markus Armbruster <armbru@redhat.com>
> >
> > Acked-by: Kevin Wolf <kwolf@redhat.com>
> 
> Apologies my mailer ignored my replay-all:
> 
> This commit breaks when of the iotests for me:
> 
>  git bisect run /bin/sh -c "cd builds/all && make -j4 \
>      && cd tests/qemu-iotests && ./check -qcow2 267"
> 
> 
>    List of snapshots present on all disks:
>   -ID        TAG                 VM SIZE                DATE       VM CLOCK
>   +ID        TAG               VM SIZE                DATE     VM CLOCK     ICOUNT
> 
> But I've also seen:
> 
>   ERROR:/home/.../qemu.git/replay/replay-events.c:80:replay_flush_events:
>      assertion failed: (replay_mutex_locked())

Thank you, I've updated the code.
I also added a patch for fixing the test output.

Pavel Dovgalyuk