Re: 2.6.21-mm2

[Andrew Morton <akpm@linux-foundation.org>]

On Thu, 10 May 2007 18:11:35 +0200 Marc Dietrich <Marc.Dietrich@ap.physik.uni-giessen.de> wrote:

> 
> Hi,
> 
> Am Wednesday 09 May 2007 10:23 schrieb Andrew Morton:
> > ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.21/2.6.21
> >-mm2/
> >
> >
> > - git-acpi was temporarily dropped due to git conflicts which I didn't feel
> >   like fixing.
> >
> > - So was git-kvm
> >
> > - A handful of IDE patches were dropped due to compilation failures.  Bart
> >   may have fixed this since I last downloaded his tree..
> >
> > - The adaptive readahead patches were dropped.  We're expecting a new, much
> >   simpler version of these.
> 
> since -mm1 I cannot create files on nfs4 shares. 
> 
> #me@localhost:/home: touch t
> touch: setting times of `t': Input/output error
> #me@localhost:/home: dmesg|tail -n1
> decode_attr_group: reply buffer overflowed in line 2676.
> 
> This problem was reported before in a different thread, but no reaction....
> 
> http://marc.info/?l=linux-kernel&m=117837974628893&w=2
> 

We had one regression there which was fixed by the below patch.

That fix is now in Linus's git tree but it was not present in 2.6.21-mm2.

So can you please retest either current -linus, or 2.6.21-mm2 with this patch?

commit 6ce7dc940701cf3fde3c6e826a696b333092cbb1
Author: Chuck Lever <chuck.lever@oracle.com>
Date:   Tue May 8 18:23:28 2007 -0400

    NFS: NFS client underestimates how large an NFSv4 SETATTR reply can be
    
    The maximum size of an NFSv4 SETATTR compound reply should include the
    GETATTR operation that we send.
    
    Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
    Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>

diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index b8c28f2..f1e2b8c 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -224,7 +224,8 @@ #define NFS4_enc_setattr_sz     (compoun
                                 encode_getattr_maxsz)
 #define NFS4_dec_setattr_sz     (compound_decode_hdr_maxsz + \
                                 decode_putfh_maxsz + \
-                                op_decode_hdr_maxsz + 3)
+                                op_decode_hdr_maxsz + 3 + \
+                                nfs4_fattr_maxsz)
 #define NFS4_enc_fsinfo_sz	(compound_encode_hdr_maxsz + \
 				encode_putfh_maxsz + \
 				encode_fsinfo_maxsz)