kernel-fxtec-pro1x/drivers/infiniband
Steve Wise 97cb7e40c6 RDMA/ucma: Allow tuning the max listen backlog
For iWARP connections, the connect request is carried in a TCP payload
on an already established TCP connection.  So if the ucma's backlog is
full, the connection request is transmitted and acked at the TCP level
by the time the connect request gets dropped in the ucma.  The end
result is the connection gets rejected by the iWARP provider.
Further, a 32 node 256NP OpenMPI job will generate > 128 connect
requests on some ranks.

This patch increases the default max backlog to 1024, and adds a
sysctl variable so the backlog can be adjusted at run time.

Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-10-23 13:41:40 -07:00
..
core RDMA/ucma: Allow tuning the max listen backlog 2010-10-23 13:41:40 -07:00
hw RDMA/cxgb3: Turn off RX coalescing for iWARP connections 2010-09-27 09:28:55 -07:00
ulp Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband 2010-08-07 17:08:02 -07:00
Kconfig IB/qib: Add new qib driver for QLogic PCIe InfiniBand adapters 2010-05-23 21:44:54 -07:00
Makefile IB/qib: Add new qib driver for QLogic PCIe InfiniBand adapters 2010-05-23 21:44:54 -07:00