[PATCH 0/2] NFSD: fix handling of unsupported nfsv4 ops

[PATCH 0/2] NFSD: fix handling of unsupported nfsv4 ops

[Benny Halevy]

Both patches fix issues I introduced in 2.6.27.
The first one is more crucial to get to the 2.6.2{7,8} stable streams
but while at it, I'd pull both.
This was triggered in connectathon with the Sun OpenSolaris
client trying to read extended attrs (when copying a file).
This translates to a NFS4 OP_OPENATTR which caused our server to barf.

[PATCH 1/2] NFSD: provide encode routine for OP_OPENATTR
[PATCH 2/2] NFSD: return nfsv4 error code nfserr_notsupp rather than nfsv[23]'s nfserr_opnotsupp

Benny

Re: [PATCH 0/2] NFSD: fix handling of unsupported nfsv4 ops

[J. Bruce Fields]

On Thu, Feb 26, 2009 at 12:37:22PM -0800, Benny Halevy wrote:
> Both patches fix issues I introduced in 2.6.27.
> The first one is more crucial to get to the 2.6.2{7,8} stable streams
> but while at it, I'd pull both.
> This was triggered in connectathon with the Sun OpenSolaris
> client trying to read extended attrs (when copying a file).
> This translates to a NFS4 OP_OPENATTR which caused our server to barf.

Thanks!  But, sorry, somehow I didn't get both of these (probably a
screwup on my end); would you mind resending?

Also, have you checked to make sure there aren't any other holes in our
list of operations?

--b.

> 
> [PATCH 1/2] NFSD: provide encode routine for OP_OPENATTR
> [PATCH 2/2] NFSD: return nfsv4 error code nfserr_notsupp rather than nfsv[23]'s nfserr_opnotsupp
> 
> Benny

Re: [PATCH 0/2] NFSD: fix handling of unsupported nfsv4 ops

[Benny Halevy]

On Mar. 04, 2009, 23:02 +0200, "J. Bruce Fields" <bfields@fieldses.org> wrote:
> On Thu, Feb 26, 2009 at 12:37:22PM -0800, Benny Halevy wrote:
>> Both patches fix issues I introduced in 2.6.27.
>> The first one is more crucial to get to the 2.6.2{7,8} stable streams
>> but while at it, I'd pull both.
>> This was triggered in connectathon with the Sun OpenSolaris
>> client trying to read extended attrs (when copying a file).
>> This translates to a NFS4 OP_OPENATTR which caused our server to barf.
> 
> Thanks!  But, sorry, somehow I didn't get both of these (probably a
> screwup on my end); would you mind resending?

Just did.

> 
> Also, have you checked to make sure there aren't any other holes in our
> list of operations?

Yup.  OPENATTR was the only hole I found.

Benny

> 
> --b.
> 
>> [PATCH 1/2] NFSD: provide encode routine for OP_OPENATTR
>> [PATCH 2/2] NFSD: return nfsv4 error code nfserr_notsupp rather than nfsv[23]'s nfserr_opnotsupp
>>
>> Benny

Re: [PATCH 0/2] NFSD: fix handling of unsupported nfsv4 ops

[J. Bruce Fields]

On Wed, Mar 04, 2009 at 04:02:00PM -0500, bfields wrote:
> On Thu, Feb 26, 2009 at 12:37:22PM -0800, Benny Halevy wrote:
> > Both patches fix issues I introduced in 2.6.27.
> > The first one is more crucial to get to the 2.6.2{7,8} stable streams
> > but while at it, I'd pull both.
> > This was triggered in connectathon with the Sun OpenSolaris
> > client trying to read extended attrs (when copying a file).
> > This translates to a NFS4 OP_OPENATTR which caused our server to barf.
> 
> Thanks!  But, sorry, somehow I didn't get both of these (probably a
> screwup on my end); would you mind resending?

Both applied now, thanks.--b.