xfs
[Top] [All Lists]

Re: [PATCH] xfs_db: check on-disk structure sizes

To: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Subject: Re: [PATCH] xfs_db: check on-disk structure sizes
From: "Darrick J. Wong" <darrick.wong@xxxxxxxxxx>
Date: Fri, 15 Jan 2016 12:56:29 -0800
Cc: Eric Sandeen <sandeen@xxxxxxxxxxx>, xfs@xxxxxxxxxxx
Delivered-to: xfs@xxxxxxxxxxx
In-reply-to: <20160113074640.GA21939@xxxxxxxxxxxxx>
References: <20160111234644.GB7831@xxxxxxxxxxxxxxxx> <20160112140122.GC12156@xxxxxxxxxxxxxxx> <20160113012945.GC2455@xxxxxxxxxxxxxxxx> <5695C01F.7030807@xxxxxxxxxxx> <20160113054730.GQ10456@dastard> <10021348-43A2-465E-BC00-E0A0882FFD22@xxxxxxxxxxx> <20160113074640.GA21939@xxxxxxxxxxxxx>
User-agent: Mutt/1.5.21 (2010-09-15)
On Tue, Jan 12, 2016 at 11:46:40PM -0800, Christoph Hellwig wrote:
> On Wed, Jan 13, 2016 at 12:02:36AM -0600, Eric Sandeen wrote:
> > > As I've pointed out previously to Darrick: xfstests:/tests/xfs/122
> > > 
> > > Make that build again, update it.
> > > 
> > Oh, I went looking for that and missed it somehow, thought it had been 
> > removed.   Ok then!
> 
> The real issue is that build environment change alignments.  32 vs 64
> bit builds are obvious, but on some architectures different ABIs have
> different alignments (we had some fun with ARM in that regard),
> nevermind the equivalents to IRIX n32 popping up everywhere these days
> that make things complicated.
> 
> I'd really love to have Darrick's check in xfs_format.h as an opt-in
> if a build time assert is provided - that way every user can check it
> doesn't screw up the structures.

I've fixed xfs/122, so I suppose we no longer need this to end up in
xfsprogs.  The kernel-side patch can stick around in fs/xfs/ without
touching libxfs.

(New patches out soonish.)

--D

> 
> _______________________________________________
> xfs mailing list
> xfs@xxxxxxxxxxx
> http://oss.sgi.com/mailman/listinfo/xfs

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