From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <xfs-bounce@oss.sgi.com>
Received: with ECARTIS (v1.0.0; list xfs); Sun, 27 Apr 2008 17:09:23 -0700 (PDT)
Received: from larry.melbourne.sgi.com (larry.melbourne.sgi.com [134.14.52.130])
	by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with SMTP id m3S095Ch008112
	for <xfs@oss.sgi.com>; Sun, 27 Apr 2008 17:09:06 -0700
Message-ID: <481515C8.6020601@sgi.com>
Date: Mon, 28 Apr 2008 10:09:44 +1000
From: Timothy Shimmin <tes@sgi.com>
MIME-Version: 1.0
Subject: Re: [PATCH, XFSQA] Don't run 175-177 if DMAPI is not supported
References: <20080423013802.GJ103491721@sgi.com> <480EDE04.5080003@sgi.com> <20080424065822.GY103491721@sgi.com>
In-Reply-To: <20080424065822.GY103491721@sgi.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Sender: xfs-bounce@oss.sgi.com
Errors-to: xfs-bounce@oss.sgi.com
List-Id: xfs
To: David Chinner <dgc@sgi.com>
Cc: xfs-dev <xfs-dev@sgi.com>, xfs-oss <xfs@oss.sgi.com>

David Chinner wrote:
> On Wed, Apr 23, 2008 at 04:58:12PM +1000, Timothy Shimmin wrote:
>> Hi,
>>
>> _test_punch already does a umount $SCRATCH_MNT.
>> However, IMHO, it would probably be nicer if one could
>> use a _require_dmapi or really it may be simpler if
>> we could roll it into common.dmapi
>> so one just needs to include common.dmapi and
>> it would complain at that point if the kernel
>> didn't support it.
> 
> Sure, but we don't have that, and I'm not about to spend the time to
> convert >20 tests to something like this. Right now I just want to
> remove all the bloody silly test failures on mainline kernels.
> 
> All the other dmapi tests "not run" just fine, I want these to do
> the same thing. Is that good enough to check in right now?
> 
> Cheers,
> 
> Dave.

Sure, that's fine.

I was really just suggesting to put:

+# test that we have DMAPI support
+_dmapi_scratch_mount
+unmount $SCRATCH_MNT > /dev/null 2>&1

directly into common.dmapi - not as a function
but just as code to always execute.
As long as there is no output under normal circumstances from
this, then all dmapi tests should not require any change
and it is simpler (all they need is to include common.dmapi
which most do).
There may be an extra mount in some of the
tests (for those that want to keep it mounted) but who cares.
Why include common.dmapi if you don't require it to be supported :)

--Tim