RE: [PATCH v2 9/9] svcrdma: Use new CQ API for RPC-over-RDMA server send CQs

["Steve Wise" <swise-7bPotxP6k4+P2YhJcF5u+vpXobYPEAuW@public.gmane.org>]

> -----Original Message-----
> From: linux-rdma-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org [mailto:linux-rdma-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org] On Behalf Of Chuck Lever
> Sent: Monday, February 08, 2016 11:26 AM
> To: linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org; linux-nfs-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> Subject: [PATCH v2 9/9] svcrdma: Use new CQ API for RPC-over-RDMA server send CQs
> 
> Calling ib_poll_cq() to sort through WCs during a completion is a
> common pattern amongst RDMA consumers. Since commit 14d3a3b2498e
> ("IB: add a proper completion queue abstraction"), WC sorting can
> be handled by the IB core.
> 
> By converting to this new API, svcrdma is made a better neighbor to
> other RDMA consumers, as it allows the core to schedule the delivery
> of completions more fairly amongst all active consumers.
> 
> This new API also aims each completion at a function that is
> specific to the WR's opcode. Thus the ctxt->wr_op field and the
> switch in process_context is replaced by a set of methods that
> handle each completion type.
> 
> The server's rdma_stat_sq_poll and rdma_stat_sq_prod metrics are no
> longer updated.
> 
> As a clean up, the cq_event_handler, the dto_tasklet, and all
> associated locking is removed, as they are no longer referenced or
> used.
> 
> Signed-off-by: Chuck Lever <chuck.lever-QHcLZuEGTsvQT0dZR+AlfA@public.gmane.org>
> ---

<snip>

> @@ -310,26 +311,26 @@ int rdma_read_chunk_frmr(struct svcxprt_rdma *xprt,
> 
>  	/* Prepare RDMA_READ */
>  	memset(&read_wr, 0, sizeof(read_wr));
> +	ctxt->cqe.done = svc_rdma_wc_read;
> +	read_wr.wr.wr_cqe = &ctxt->cqe;
>  	read_wr.wr.send_flags = IB_SEND_SIGNALED;
>  	read_wr.rkey = rs_handle;
>  	read_wr.remote_addr = rs_offset;
>  	read_wr.wr.sg_list = ctxt->sge;
>  	read_wr.wr.num_sge = 1;
>  	if (xprt->sc_dev_caps & SVCRDMA_DEVCAP_READ_W_INV) {
> -		read_wr.wr.opcode = IB_WR_RDMA_READ_WITH_INV;
> -		read_wr.wr.wr_id = (unsigned long)ctxt;
>  		read_wr.wr.ex.invalidate_rkey = ctxt->frmr->mr->lkey;

Hey Chuck, I found this bug by testing.  You mistakenly removed setting the wr opcode for READ_WITH_INV above.

This fixed it:

@@ -319,6 +319,7 @@ int rdma_read_chunk_frmr(struct svcxprt_rdma *xprt,
        read_wr.wr.sg_list = ctxt->sge;
        read_wr.wr.num_sge = 1;
        if (xprt->sc_dev_caps & SVCRDMA_DEVCAP_READ_W_INV) {
+               read_wr.wr.opcode = IB_WR_RDMA_READ_WITH_INV;
                read_wr.wr.ex.invalidate_rkey = ctxt->frmr->mr->lkey;
        } else {
                read_wr.wr.opcode = IB_WR_RDMA_READ;

And with this patch, it tests out over iWARP/cxgb4.

Tested-by: Steve Wise <swise-7bPotxP6k4+P2YhJcF5u+vpXobYPEAuW@public.gmane.org>


--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html