xfs
[Top] [All Lists]

Re: sparse (was Re: [PATCH] xfs: fix brown paperbag quota-aware stafs bu

To: Nathan Scott <nathans@xxxxxxx>
Subject: Re: sparse (was Re: [PATCH] xfs: fix brown paperbag quota-aware stafs bug)
From: Christoph Hellwig <hch@xxxxxx>
Date: Thu, 13 Jul 2006 10:21:43 +0200
Cc: Christoph Hellwig <hch@xxxxxx>, xfs@xxxxxxxxxxx
In-reply-to: <20060713134927.D1761066@xxxxxxxxxxxxxxxxxxxxxxxx>
References: <20060710165539.GA26673@xxxxxx> <20060711102341.C1702118@xxxxxxxxxxxxxxxxxxxxxxxx> <20060713134927.D1761066@xxxxxxxxxxxxxxxxxxxxxxxx>
Sender: xfs-bounce@xxxxxxxxxxx
User-agent: Mutt/1.3.28i
On Thu, Jul 13, 2006 at 01:49:27PM +1000, Nathan Scott wrote:
> On Tue, Jul 11, 2006 at 10:23:42AM +1000, Nathan Scott wrote:
> > On Mon, Jul 10, 2006 at 06:55:39PM +0200, Christoph Hellwig wrote:
> > ...
> > > (I wish someone at SGI would run sparse once in a while..)
> > 
> > Indeed, my bad.  I'm much more motivated to do so now, thanks. :)
> 
> Do you see this kind of warning from sparse, for debug builds?
> 
>   CHECK   fs/xfs/xfs_ialloc_btree.c
> fs/xfs/xfs_alloc_btree.c:922:4: error: too long token expansion
> 
> Looks like an assert (guess it expands to a fairly long string,
> after cpp-style expansion) - is there any way to make sparse not
> warn/error on these?

Yes, although I only see them with recent sparse versions.  There's
also a lot 'foo could be static' warnings with debug builds because
STATIC is defined to nothing there.  And quite a few warnings because
of the uncmpleteted endianess annotations and the iovec mess..  So
it's generally not so useful to look at all the warnings for now but
only at the diff to a previous version, carefully looking at the changes
and making sure the warnings get less, not more..


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