Roland,
You are right that the idr implementation introduces insignificant change in
performance. I made the version with idr and semaphores usage and I see a
minimal change comparing to hash table. Now only a shared page is used instead
of kmalloc and copy_to_user.
I simplified changes to
From: Mitko Haralanov mi...@qlogic.com
Hold the IB link at DISABLED until we get the correct TX
settings on mezz boards.
Signed-off-by: Mike Marciniszyn mike.marcinis...@qlogic.com
---
drivers/infiniband/hw/qib/qib_iba7322.c | 30 +++---
1 files changed, 11
+ qp = idr_read_qp(cmd.qp_handle, file-ucontext);
+ if (!qp)
+ goto out_raw_qp;
+
+ if (qp-qp_type == IB_QPT_RAW_ETH) {
+ resp.bad_wr = 0;
+ ret = qp-device-post_send(qp, NULL, NULL);
This looks odd to me and can definitely confuse someone
MAXBURST cannot exceed 256B for on-chip queues. With a 512B MAXBURST,
we can lock up the chip.
Signed-off-by: Steve Wise sw...@opengridcomputing.com
Cc: sta...@kernel.org
---
drivers/infiniband/hw/cxgb4/qp.c |4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git
The PF passed to FW was 0 causing PCI failures in an SR-IOV environment.
Signed-off-by: Steve Wise sw...@opengridcomputing.com
Cc: sta...@kernel.org
---
drivers/infiniband/hw/cxgb4/cm.c |3 ++-
1 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/drivers/infiniband/hw/cxgb4/cm.c
On 08/04/2010 12:44 PM, Roland Dreier wrote:
thanks, applied
Hey Roland,
Do you plan to merge the other 3 patches in this series?
https://patchwork.kernel.org/patch/110153/
https://patchwork.kernel.org/patch/110154/
https://patchwork.kernel.org/patch/110155/
Thanks,
Steve.
--
To
As some of you read from my blog, today was my last day at Cisco. I
will be continuing to help maintain the Linux RDMA stack. The main
change in the forseeable future is that I will no longer be reachable
at any @cisco.com email address; for the moment, this email
(rol...@kernel.org) is the best