Re: [PATCH 1/3] xfs: support realtime/log device setup changes in config sections

[Eryu Guan <eguan@redhat.com>]

On Fri, Feb 05, 2016 at 08:04:55AM +1100, Dave Chinner wrote:
> On Tue, Feb 02, 2016 at 12:36:33PM +0800, Eryu Guan wrote:
> > On Thu, Jan 21, 2016 at 12:43:11PM +1100, Dave Chinner wrote:
> > > From: Dave Chinner <dchinner@redhat.com>
> > > 
> > > Currently changing the devices used by "USE_EXTERNAL" environmental
> > > variable is not supported by the config section parsing. Add the
> > > functionality so that we can use config sections to test external
> > > device configs successfully.
> > > 
> > > This required tracking down a bug in _check_xfs_filesystem() which
> > > was causing a log device to be passed to a test device without an
> > > external log device. This was caused by an uninitialised variable in
> > > the function. I also added full output file removals to the first
> > > couple of generic tests that were failing, because that's where the
> > > check failure output ends up in this case.
> > > 
> > > Signed-off-by: Dave Chinner <dchinner@redhat.com>
> > 
> > Looks good to me. And I ran some tests on NFS/CIFS/XFS/overlay and the
> > new _check_device function worked as expected. Just one white space
> > issue below
> > 
> > > ---
> > >  common/config     | 63 +++++++++++++++++++++++++++++++++++++++----------------
> > >  common/dmflakey   |  3 ++-
> > >  common/rc         |  1 +
> > >  tests/generic/002 |  2 ++
> > >  tests/generic/004 |  2 ++
> > >  5 files changed, 52 insertions(+), 19 deletions(-)
> > > 
> > > diff --git a/common/config b/common/config
> > > index cb34fd7..477753e 100644
> > > --- a/common/config
> > > +++ b/common/config
> > > @@ -416,6 +416,30 @@ if [ -f "$HOST_OPTIONS" ]; then
> > >  	fi
> > >  fi
> > >  
> > > +_check_device()
> > > +{
> > > +	local name=$1
> > > +	local dev_needed=$2
> > > +	local dev=$3
> > > +
> > > +	if [ -z "$dev" ]; then
> > > +		if [ "$dev_needed" == "required" ]; then 
> > 
> > Above line introduced trailing white space.
> 
> Fixed, so can I add your reviewed-by?

Sure, and I'll review the other two soon (got interrupted last time I
reviewed).

Thanks,
Eryu