linux/drivers/net/cxgb3
Steve Wise f8b0dfd152 RDMA/cxgb3: Support peer-2-peer connection setup
Open MPI, Intel MPI and other applications don't respect the iWARP
requirement that the client (active) side of the connection send the
first RDMA message.  This class of application connection setup is
called peer-to-peer.  Typically once the connection is setup, _both_
sides want to send data.

This patch enables supporting peer-to-peer over the chelsio RNIC by
enforcing this iWARP requirement in the driver itself as part of RDMA
connection setup.

Connection setup is extended, when the peer2peer module option is 1,
such that the MPA initiator will send a 0B Read (the RTR) just after
connection setup.  The MPA responder will suspend SQ processing until
the RTR message is received and reply-to.

In the longer term, this will be handled in a standardized way by
enhancing the MPA negotiation so peers can indicate whether they
want/need the RTR and what type of RTR (0B read, 0B write, or 0B send)
should be sent.  This will be done by standardizing a few bits of the
private data in order to negotiate all this.  However this patch
enables peer-to-peer applications now and allows most of the required
firmware and driver changes to be done and tested now.

Design:

 - Add a module option, peer2peer, to enable this mode.

 - New firmware support for peer-to-peer mode:

	- a new bit in the rdma_init WR to tell it to do peer-2-peer
	  and what form of RTR message to send or expect.

	- process _all_ preposted recvs before moving the connection
	  into rdma mode.

	- passive side: defer completing the rdma_init WR until all
	  pre-posted recvs are processed.  Suspend SQ processing until
	  the RTR is received.

	- active side: expect and process the 0B read WR on offload TX
	  queue. Defer completing the rdma_init WR until all
	  pre-posted recvs are processed.  Suspend SQ processing until
	  the 0B read WR is processed from the offload TX queue.

 - If peer2peer is set, driver posts 0B read request on offload TX
   queue just after posting the rdma_init WR to the offload TX queue.

 - Add CQ poll logic to ignore unsolicitied read responses.

Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
2008-04-29 13:46:52 -07:00
..
adapter.h net: Remove unnecessary inclusions of asm/semaphore.h 2008-04-18 22:15:50 -04:00
ael1002.c cxgb3 - Fix direct XAUI support 2007-06-20 19:16:58 -04:00
common.h annotate cxgb3 2008-01-28 15:10:30 -08:00
cxgb3_ctl_defs.h cxgb3 - Expose HW memory page info 2007-10-10 16:50:49 -07:00
cxgb3_defs.h cxgb3 - tighten checks on TID values 2007-10-10 16:50:49 -07:00
cxgb3_ioctl.h cxgb3 - private ioctl cleanup 2007-02-27 04:27:12 -05:00
cxgb3_main.c cxgb3: Fix __must_check warning with dev_dbg. 2008-04-17 15:31:32 -04:00
cxgb3_offload.c annotate cxgb3 (ab)uses of skb->priority/skb->csum 2008-03-26 00:18:46 -04:00
cxgb3_offload.h cxgb3 - Fix dev->priv usage 2007-08-31 07:29:08 -04:00
firmware_exports.h drivers/net/cxgb3: trim trailing whitespace 2008-01-28 15:04:13 -08:00
l2t.c [NET] NEIGHBOUR: Remove unpopular neigh_is_connected(). 2008-03-04 15:18:23 +09:00
l2t.h
Makefile
mc5.c cxgb3: Remove incorrect __devinit annotations 2008-02-03 04:28:35 -08:00
regs.h cxgb3 - parity initialization for T3C adapters. 2008-01-28 15:07:22 -08:00
sge_defs.h cxgb3 - Set the CQ_ERR bit in CQ contexts. 2007-10-10 16:51:05 -07:00
sge.c cxgb3: Fix lockdep problems with sge.reg_lock 2008-03-25 23:42:05 -04:00
t3_cpl.h
t3_hw.c cxgb3: Remove incorrect __devinit annotations 2008-02-03 04:28:35 -08:00
t3cdev.h net: Remove unnecessary inclusions of asm/semaphore.h 2008-04-18 22:15:50 -04:00
version.h RDMA/cxgb3: Support peer-2-peer connection setup 2008-04-29 13:46:52 -07:00
vsc8211.c
xgmac.c drivers/net/cxgb3: trim trailing whitespace 2008-01-28 15:04:13 -08:00