xfs
[Top] [All Lists]

Re: xfstests: more 214 updates

To: Eric Sandeen <sandeen@xxxxxxxxxxx>
Subject: Re: xfstests: more 214 updates
From: Dave Chinner <david@xxxxxxxxxxxxx>
Date: Mon, 29 Oct 2012 08:35:51 +1100
Cc: Rich Johnston <rjohnston@xxxxxxx>, xfs-oss <xfs@xxxxxxxxxxx>
In-reply-to: <508AC251.1020308@xxxxxxxxxxx>
References: <4B4D02E5.7040306@xxxxxxxxxx> <508ABF90.6080102@xxxxxxx> <508AC251.1020308@xxxxxxxxxxx>
User-agent: Mutt/1.5.21 (2010-09-15)
On Fri, Oct 26, 2012 at 12:03:13PM -0500, Eric Sandeen wrote:
> On 10/26/12 11:51 AM, Rich Johnston wrote:
> > On 01/12/2010 05:16 PM, Eric Sandeen wrote:
> >> A few more updates for the 214 fallocate test, since ext4 is
> >> having some fallocate trouble these days.
> >>
> >> First off, updated 214.out was missing from the previous commit
> >> to this file.
> >>
> >> Second, run each test against a new filename and don't remove
> >> it, so that the post-test fs check will have more work to do.
> >>
> >> Third, and a newly-found corruption case for ext4.
> >>
> >> Signed-off-by: Eric Sandeen <sandeen@xxxxxxxxxx>
> >>
> >> ---
> >>
> >>
> >> diff --git a/214 b/214 index a71680a..0cfb23e 100755 --- a/214
> >> +++ b/214
> > 
> > 
> > Looks good
> > 
> > Reviewed-by: Rich Johnston <rjohnston@xxxxxxx>
> > 
> > Eric,
> > 
> > This patch has been committed to
> > git://oss.sgi.com/xfs/cmds/xfstests, master branch, commit ID
> > e2e226715.
> > 
> > Thanks --Rich
> 
> Rich, I appreciate you guys searching for lost patches.
> 
> I think we need to be very careful about grabbing 2yo changes off
> the list and applying them to today's tree; if you have any
> questions at all about anything, please don't hesitate to bring up
> a question on the list.

Actaully, I think it would be better to grab all the old patches and
post them to the list again for review and discussion before
checking them in.  There may have been a reason they didn't get
picked up in the first place (e.g. a discussion on IRC rather than
the mainling list), and there may be other reasons the patches need
modification now.

e.g:

> Just as an example: hopefully the new test case on this change
> doesn't cause 214 to start failing on any fs?  We usually try to
> not change tests to add new failure cases if the test passed
> before.  If it's passing on all filesystems today, it's probably
> fine to add the case.  If not, it should probably be moved to a
> new test.

That was less of a concern 2 years ago, but is something we tend to
avoid now.

FWIW, the question of "should I repost something" is answered by
asking whether anyone would be surprised by the commit - nobody
should ever be surprised.  Anything that isn't recently posted needs
to be looked at again before committing, if only to let people know
it was missed and needs to be looked at again....

Cheers,

Dave.
-- 
Dave Chinner
david@xxxxxxxxxxxxx

<Prev in Thread] Current Thread [Next in Thread>