Re: [PATCH] nfs41: make sure nfs server return right ca_maxresponsesize_cached

[Mi Jinlong <mijinlong@cn.fujitsu.com>]

J. Bruce Fields:
> On Thu, Mar 10, 2011 at 05:43:37PM +0800, Mi Jinlong wrote:
>> According to rfc5661, 
>>   
>>   ca_maxresponsesize_cached:
>>
>>      Like ca_maxresponsesize, but the maximum size of a reply that
>>      will be stored in the reply cache (Section 2.10.6.1).  For each
>>      channel, the server MAY decrease this value, but MUST NOT
>>      increase it.
> 
> I wonder what the reason for that is--I can't see why the client would
> object to the server increasing this.  Oh well, my as well observe the
> MUST NOT; applied.

  I just read the RFC and got it.

  Agree with you.

--
thanks,
Mi Jinlong

> 
> --b.
> 
>>  
>> the latest kernel(2.6.38-rc8) may increase the value for ignoring 
>> request's ca_maxresponsesize_cached value. We should not ignore it.
>>
>> Signed-off-by: Mi Jinlong <mijinlong@cn.fujitsu.com>
>> ---
>>  fs/nfsd/nfs4state.c |    3 ++-
>>  1 files changed, 2 insertions(+), 1 deletions(-)
>>
>> diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
>> index 54b60bf..96edd02 100644
>> --- a/fs/nfsd/nfs4state.c
>> +++ b/fs/nfsd/nfs4state.c
>> @@ -608,7 +608,8 @@ static void init_forechannel_attrs(struct nfsd4_channel_attrs *new, struct nfsd4
>>  	u32 maxrpc = nfsd_serv->sv_max_mesg;
>>  
>>  	new->maxreqs = numslots;
>> -	new->maxresp_cached = slotsize + NFSD_MIN_HDR_SEQ_SZ;
>> +	new->maxresp_cached = min_t(u32, req->maxresp_cached, 
>> +					slotsize + NFSD_MIN_HDR_SEQ_SZ);
>>  	new->maxreq_sz = min_t(u32, req->maxreq_sz, maxrpc);
>>  	new->maxresp_sz = min_t(u32, req->maxresp_sz, maxrpc);
>>  	new->maxops = min_t(u32, req->maxops, NFSD_MAX_OPS_PER_COMPOUND);
>> -- 
>> 1.7.4.1
>>
>>
>>