[PATCH for-next 0/2] IB/hfi1,rdmavt: patches for next 06/17/2017

[PATCH for-next 0/2] IB/hfi1,rdmavt: patches for next 06/17/2017

[Dennis Dalessandro]

Hi Doug,

Here are the next set of patches for 4.13. These are two by Kaike. One fixes an
overflow for the QP timeout and the other extends our qp_stats to print info on
the ack_queue.

Patches can can also be found in my GitHub repo at:
https://github.com/ddalessa/kernel/tree/for-4.13
---

Kaike Wan (2):
      IB/rdmavt: Setting of QP timeout can overflow jiffies computation
      IB/hfi1: Add receiving queue info to qp_stats


 drivers/infiniband/hw/hfi1/qp.c   |    6 +++++-
 drivers/infiniband/sw/rdmavt/qp.c |    4 +---
 include/rdma/rdmavt_qp.h          |   14 ++++++++++++++
 3 files changed, 20 insertions(+), 4 deletions(-)

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

[PATCH for-next 1/2] IB/rdmavt: Setting of QP timeout can overflow jiffies computation

[Dennis Dalessandro]

From: Kaike Wan <kaike.wan-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>

Current computation of qp->timeout_jiffies in rvt_modify_qp() will cause
overflow due to the fact that the input to the function usecs_to_jiffies
is only 32-bit ( unsigned int). Overflow will occur when attr->timeout is
equal to or greater than 30. The consequence is unnecessarily excessive
retry and thus degradation of the system performance.

This patch fixes the problem by limiting the input to 5-bit and calling
usecs_to_jiffies() before multiplying the scaling factor.

Reviewed-by: Mike Marciniszyn <mike.marciniszyn-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Kaike Wan <kaike.wan-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
 drivers/infiniband/sw/rdmavt/qp.c |    4 +---
 include/rdma/rdmavt_qp.h          |   14 ++++++++++++++
 2 files changed, 15 insertions(+), 3 deletions(-)

diff --git a/drivers/infiniband/sw/rdmavt/qp.c b/drivers/infiniband/sw/rdmavt/qp.c
index 2ce0928..17b2eb0 100644
--- a/drivers/infiniband/sw/rdmavt/qp.c
+++ b/drivers/infiniband/sw/rdmavt/qp.c
@@ -1284,9 +1284,7 @@ int rvt_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
 
 	if (attr_mask & IB_QP_TIMEOUT) {
 		qp->timeout = attr->timeout;
-		qp->timeout_jiffies =
-			usecs_to_jiffies((4096UL * (1UL << qp->timeout)) /
-				1000UL);
+		qp->timeout_jiffies = rvt_timeout_to_jiffies(qp->timeout);
 	}
 
 	if (attr_mask & IB_QP_QKEY)
diff --git a/include/rdma/rdmavt_qp.h b/include/rdma/rdmavt_qp.h
index 13f43b3..07e2fff 100644
--- a/include/rdma/rdmavt_qp.h
+++ b/include/rdma/rdmavt_qp.h
@@ -647,6 +647,20 @@ static inline u32 rvt_div_mtu(struct rvt_qp *qp, u32 len)
 	return len >> qp->log_pmtu;
 }
 
+/**
+ * rvt_timeout_to_jiffies - Convert a ULP timeout input into jiffies
+ * @timeout - timeout input(0 - 31).
+ *
+ * Return a timeout value in jiffies.
+ */
+static inline unsigned long rvt_timeout_to_jiffies(u8 timeout)
+{
+	if (timeout > 31)
+		timeout = 31;
+
+	return usecs_to_jiffies(1U << timeout) * 4096UL / 1000UL;
+}
+
 extern const int  ib_rvt_state_ops[];
 
 struct rvt_dev_info;

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

[PATCH for-next 2/2] IB/hfi1: Add receiving queue info to qp_stats

[Dennis Dalessandro]

From: Kaike Wan <kaike.wan-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>

This patch adds qp->s_ack_queue indices and size to qp_stats printout.
This information will provide information about the receiving side.

Reviewed-by: Mike Marciniszyn <mike.marciniszyn-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Kaike Wan <kaike.wan-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
 drivers/infiniband/hw/hfi1/qp.c |    6 +++++-
 1 files changed, 5 insertions(+), 1 deletions(-)

diff --git a/drivers/infiniband/hw/hfi1/qp.c b/drivers/infiniband/hw/hfi1/qp.c
index e91be05..f1e4bff 100644
--- a/drivers/infiniband/hw/hfi1/qp.c
+++ b/drivers/infiniband/hw/hfi1/qp.c
@@ -601,7 +601,7 @@ void qp_iter_print(struct seq_file *s, struct qp_iter *iter)
 	wqe = rvt_get_swqe_ptr(qp, qp->s_last);
 	send_context = qp_to_send_context(qp, priv->s_sc);
 	seq_printf(s,
-		   "N %d %s QP %x R %u %s %u %u %u f=%x %u %u %u %u %u %u SPSN %x %x %x %x %x RPSN %x (%u %u %u %u %u %u %u) RQP %x LID %x SL %u MTU %u %u %u %u %u SDE %p,%u SC %p,%u SCQ %u %u PID %d\n",
+		   "N %d %s QP %x R %u %s %u %u %u f=%x %u %u %u %u %u %u SPSN %x %x %x %x %x RPSN %x S(%u %u %u %u %u %u %u) R(%u %u %u) RQP %x LID %x SL %u MTU %u %u %u %u %u SDE %p,%u SC %p,%u SCQ %u %u PID %d\n",
 		   iter->n,
 		   qp_idle(qp) ? "I" : "B",
 		   qp->ibqp.qp_num,
@@ -624,6 +624,10 @@ void qp_iter_print(struct seq_file *s, struct qp_iter *iter)
 		   qp->s_last, qp->s_acked, qp->s_cur,
 		   qp->s_tail, qp->s_head, qp->s_size,
 		   qp->s_avail,
+		   /* ack_queue ring pointers, size */
+		   qp->s_tail_ack_queue, qp->r_head_ack_queue,
+		   HFI1_MAX_RDMA_ATOMIC,
+		   /* remote QP info  */
 		   qp->remote_qpn,
 		   rdma_ah_get_dlid(&qp->remote_ah_attr),
 		   rdma_ah_get_sl(&qp->remote_ah_attr),

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

Re: [PATCH for-next 1/2] IB/rdmavt: Setting of QP timeout can overflow jiffies computation

[Doug Ledford]

[-- Attachment #1.1: Type: text/plain, Size: 1124 bytes --]

On 6/17/2017 1:37 PM, Dennis Dalessandro wrote:
> From: Kaike Wan <kaike.wan-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> 
> Current computation of qp->timeout_jiffies in rvt_modify_qp() will cause
> overflow due to the fact that the input to the function usecs_to_jiffies
> is only 32-bit ( unsigned int). Overflow will occur when attr->timeout is
> equal to or greater than 30. The consequence is unnecessarily excessive
> retry and thus degradation of the system performance.
> 
> This patch fixes the problem by limiting the input to 5-bit and calling
> usecs_to_jiffies() before multiplying the scaling factor.
> 
> Reviewed-by: Mike Marciniszyn <mike.marciniszyn-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> Signed-off-by: Kaike Wan <kaike.wan-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> Signed-off-by: Dennis Dalessandro <dennis.dalessandro-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>

This was accepted into 4.13-rc, thanks.

-- 
Doug Ledford <dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
    GPG Key ID: B826A3330E572FDD
    Key fingerprint = AE6B 1BDA 122B 23B4 265B  1274 B826 A333 0E57 2FDD


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 884 bytes --]

Re: [PATCH for-next 2/2] IB/hfi1: Add receiving queue info to qp_stats

[Doug Ledford]

On Sat, 2017-06-17 at 10:37 -0700, Dennis Dalessandro wrote:
> From: Kaike Wan <kaike.wan-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> 
> This patch adds qp->s_ack_queue indices and size to qp_stats
> printout.
> This information will provide information about the receiving side.
> 
> Reviewed-by: Mike Marciniszyn <mike.marciniszyn-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> Signed-off-by: Kaike Wan <kaike.wan-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> Signed-off-by: Dennis Dalessandro <dennis.dalessandro-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>

This patch didn't look important enough to warrant a for-rc submission,
so I applied it to for-next instead.  That completes this 2 patch
series, with one in -rc and one in -next.  Thanks.

-- 
Doug Ledford <dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
    GPG KeyID: B826A3330E572FDD
    Key fingerprint = AE6B 1BDA 122B 23B4 265B  1274 B826 A333 0E57 2FDD

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