> >>
> >
> > So the code as it stands violates the RFC by sending 2 chunks?
>
> The Linux client is non-compliant because it sends two chunks. A compliant
> server
ignores the
> second chunk, so this is functionally harmless (but a waste of resources).
>
> The Linux server is non-compliant beca
On May 14, 2014, at 2:11 PM, Steve Wise wrote:
>
>>>
If you print ch_count, it is 2 for NFS WRITEs from a Linux client,
no matter how large the write payload is. Therefore I think the check
as it is written is not particularly useful.
>>>
>>> Why are there 2?
>>
>> The first c
> >
> >> If you print ch_count, it is 2 for NFS WRITEs from a Linux client,
> >> no matter how large the write payload is. Therefore I think the check
> >> as it is written is not particularly useful.
> >
> > Why are there 2?
>
> The first chunk lists the pages the server is to read, and the seco
On May 14, 2014, at 10:26 AM, Steve Wise wrote:
> On 5/13/2014 4:44 PM, Chuck Lever wrote:
>>
>>> I'm not sure what "segment count” you are talking about?
>> AFAICT a chunk is a list of segments. Thus one chunk can reference
>> many pages, one per segment.
>
> Ok yes. rdma_read_chunk_frmr()/r
On 5/13/2014 4:44 PM, Chuck Lever wrote:
+static int rdma_read_chunks(struct svcxprt_rdma *xprt,
+ struct rpcrdma_msg *rmsgp,
+ struct svc_rqst *rqstp,
+ struct svc_rdma_op_ctxt *head)
{
- struct ib_send_wr read_w
ernel.org; Tom
>> Tucker
>> Subject: Re: [PATCH V2 RFC 2/3] svcrdma: Recvfrom changes
>>
>> Hi Steve-
>>
>> Some random review comments, see below.
>>
>
>
>
>>> +static int rdma_read_chunks(struct svcxprt_rdma *xprt,
>>&
> -Original Message-
> From: Chuck Lever [mailto:chuckle...@gmail.com]
> Sent: Tuesday, May 13, 2014 1:22 PM
> To: Steve Wise
> Cc: J. Bruce Fields; Linux NFS Mailing List; linux-rdma@vger.kernel.org; Tom
> Tucker
> Subject: Re: [PATCH V2 RFC 2/3] svcrdma: Recvfrom
Hi Steve-
Some random review comments, see below.
On May 6, 2014, at 1:46 PM, Steve Wise wrote:
> From: Tom Tucker
>
> Based on device support, RDMA read target sgls are fast-registered,
> or composed using the local dma lkey or a dma_mr lkey. A given NFS
> Write chunk list will be split in
From: Tom Tucker
Based on device support, RDMA read target sgls are fast-registered,
or composed using the local dma lkey or a dma_mr lkey. A given NFS
Write chunk list will be split into a set of rdma reads based on the
limitations of the device.
Signed-off-by: Tom Tucker
Signed-off-by: Steve