[PATCH 0/7] RDMA/i40iw: Fixes for i40iw driver

[PATCH 0/7] RDMA/i40iw: Fixes for i40iw driver

[Tatyana Nikolova]

This patch series includes a few fixes for the original
Intel X722 iWARP driver i40iw.

It is based on Doug Ledford's https://github.com/dledford/linux.git
branch: k.o/for-4.6-rc. 

It needs to be applied after the patch series submitted by Mustafa Ismail - 
"[PATCH V3 00/16] RDMA/Intel X722 iWARP driver update"

Chien Tin Tung (1):
  RDMA/i40iw: Correct STag mask to min of 14 bits

Mohammad Khan (1):
  RDMA/i40iw: Fix for a NOP WQE size

Shiraz Saleem (3):
  RDMA/i40iw: Fixes for WQE alignment
  RDMA/i40iw: Fix for the size of kernel mode SQ
  RDMA/i40iw: Fix for using one sge for RDMA READ

Tatyana Nikolova (2):
  RDMA/i40iw: Fix for checking if the QP is destroyed
  RDMA/i40iw: Fix for removing quad hash entries

 drivers/infiniband/hw/i40iw/i40iw_cm.c     |  4 +++-
 drivers/infiniband/hw/i40iw/i40iw_d.h      |  2 ++
 drivers/infiniband/hw/i40iw/i40iw_hw.c     |  3 ++-
 drivers/infiniband/hw/i40iw/i40iw_status.h |  1 +
 drivers/infiniband/hw/i40iw/i40iw_uk.c     | 26 ++++++++++++++++++++++++--
 drivers/infiniband/hw/i40iw/i40iw_verbs.c  | 10 +++++++---
 6 files changed, 39 insertions(+), 7 deletions(-)

-- 
2.7.4

--
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 1/7] RDMA/i40iw: Fixes for WQE alignment

[Tatyana Nikolova]

From: Shiraz Saleem <shiraz.saleem-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>

Invalidation after every WQE write is changed to invalidate
only if required. NOPs are padded so that WQE writes are
aligned to 64B boundary.

Signed-off-by: Tatyana Nikolova <tatyana.e.nikolova-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Faisal Latif <faisal.latif-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
 drivers/infiniband/hw/i40iw/i40iw_d.h  |  2 ++
 drivers/infiniband/hw/i40iw/i40iw_uk.c | 18 ++++++++++++++++--
 2 files changed, 18 insertions(+), 2 deletions(-)

diff --git a/drivers/infiniband/hw/i40iw/i40iw_d.h b/drivers/infiniband/hw/i40iw/i40iw_d.h
index e8951a7..bd942da 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_d.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_d.h
@@ -1512,6 +1512,8 @@ enum i40iw_alignment {
 	I40IW_SD_BUF_ALIGNMENT =	0x100
 };
 
+#define I40IW_WQE_SIZE_64	64
+
 #define I40IW_QP_WQE_MIN_SIZE	32
 #define I40IW_QP_WQE_MAX_SIZE	128
 
diff --git a/drivers/infiniband/hw/i40iw/i40iw_uk.c b/drivers/infiniband/hw/i40iw/i40iw_uk.c
index 9e3a700..6e0e327 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_uk.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_uk.c
@@ -162,6 +162,17 @@ u64 *i40iw_qp_get_next_send_wqe(struct i40iw_qp_uk *qp,
 		if (!*wqe_idx)
 			qp->swqe_polarity = !qp->swqe_polarity;
 	}
+
+	if (((*wqe_idx & 3) == 1) && (wqe_size == I40IW_WQE_SIZE_64)) {
+		i40iw_nop_1(qp);
+		I40IW_RING_MOVE_HEAD(qp->sq_ring, ret_code);
+		if (ret_code)
+			return NULL;
+		*wqe_idx = I40IW_RING_GETCURRENT_HEAD(qp->sq_ring);
+		if (!*wqe_idx)
+			qp->swqe_polarity = !qp->swqe_polarity;
+	}
+
 	for (i = 0; i < wqe_size / I40IW_QP_WQE_MIN_SIZE; i++) {
 		I40IW_RING_MOVE_HEAD(qp->sq_ring, ret_code);
 		if (ret_code)
@@ -172,8 +183,11 @@ u64 *i40iw_qp_get_next_send_wqe(struct i40iw_qp_uk *qp,
 
 	peek_head = I40IW_RING_GETCURRENT_HEAD(qp->sq_ring);
 	wqe_0 = qp->sq_base[peek_head].elem;
-	if (peek_head & 0x3)
-		wqe_0[3] = LS_64(!qp->swqe_polarity, I40IWQPSQ_VALID);
+
+	if (((peek_head & 3) == 1) || ((peek_head & 3) == 3)) {
+		if (RS_64(wqe_0[3], I40IWQPSQ_VALID) != !qp->swqe_polarity)
+			wqe_0[3] = LS_64(!qp->swqe_polarity, I40IWQPSQ_VALID);
+	}
 
 	qp->sq_wrtrk_array[*wqe_idx].wrid = wr_id;
 	qp->sq_wrtrk_array[*wqe_idx].wr_len = total_size;
-- 
2.7.4

--
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 2/7] RDMA/i40iw: Correct STag mask to min of 14 bits

[Tatyana Nikolova]

From: Chien Tin Tung <Chien.Tin.Tung-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>

STag index mask is calculated incorrectly, missing
the 14 bits minimum requirement. Add max macro to use
either # of MRs or 14 bits in the mask size calculation.

Signed-off-by: Tatyana Nikolova <Tatyana.E.Nikolova-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Faisal Latif <faisal.latif-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
 drivers/infiniband/hw/i40iw/i40iw_hw.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/infiniband/hw/i40iw/i40iw_hw.c b/drivers/infiniband/hw/i40iw/i40iw_hw.c
index 615e115..3ee0cad 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_hw.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_hw.c
@@ -107,7 +107,8 @@ u32 i40iw_initialize_hw_resources(struct i40iw_device *iwdev)
 
 	spin_lock_init(&iwdev->resource_lock);
 	spin_lock_init(&iwdev->qptable_lock);
-	mrdrvbits = 24 - get_count_order(iwdev->max_mr);
+	/* stag index mask has a minimum of 14 bits */
+	mrdrvbits = 24 - max(get_count_order(iwdev->max_mr), 14);
 	iwdev->mr_stagmask = ~(((1 << mrdrvbits) - 1) << (32 - mrdrvbits));
 	return 0;
 }
-- 
2.7.4

--
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 3/7] RDMA/i40iw: Fix for a NOP WQE size

[Tatyana Nikolova]

From: Mohammad Khan <mohammad.a.khan-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>

Fix for filling in the WQE size for NOP

Signed-off-by: Tatyana Nikolova <tatyana.e.nikolova-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Faisal Latif <faisal.latif-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
 drivers/infiniband/hw/i40iw/i40iw_uk.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/infiniband/hw/i40iw/i40iw_uk.c b/drivers/infiniband/hw/i40iw/i40iw_uk.c
index 6e0e327..2cd9091 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_uk.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_uk.c
@@ -56,6 +56,9 @@ static enum i40iw_status_code i40iw_nop_1(struct i40iw_qp_uk *qp)
 
 	wqe_idx = I40IW_RING_GETCURRENT_HEAD(qp->sq_ring);
 	wqe = qp->sq_base[wqe_idx].elem;
+
+	qp->sq_wrtrk_array[wqe_idx].wqe_size = I40IW_QP_WQE_MIN_SIZE;
+
 	peek_head = (qp->sq_ring.head + 1) % qp->sq_ring.size;
 	wqe_0 = qp->sq_base[peek_head].elem;
 	if (peek_head)
-- 
2.7.4

--
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 4/7] RDMA/i40iw: Fix for the size of kernel mode SQ

[Tatyana Nikolova]

From: Shiraz Saleem <shiraz.saleem-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>

Fix to calculate the SQ size based on the max
frag_count, requested by the application instead
of overwriting it with the max supported frag_count

Signed-off-by: Tatyana Nikolova <tatyana.e.nikolova-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Faisal Latif <faisal.latif-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
 drivers/infiniband/hw/i40iw/i40iw_verbs.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
index bf4e1e3..2d832c7 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
@@ -520,8 +520,6 @@ static int i40iw_setup_kmode_qp(struct i40iw_device *iwdev,
 	enum i40iw_status_code status;
 	struct i40iw_qp_uk_init_info *ukinfo = &info->qp_uk_init_info;
 
-	ukinfo->max_sq_frag_cnt = I40IW_MAX_WQ_FRAGMENT_COUNT;
-
 	sq_size = i40iw_qp_roundup(ukinfo->sq_size + 1);
 	rq_size = i40iw_qp_roundup(ukinfo->rq_size + 1);
 
-- 
2.7.4

--
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 5/7] RDMA/i40iw: Fix for using one sge for RDMA READ

[Tatyana Nikolova]

From: Shiraz Saleem <shiraz.saleem-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>

A check is added to validate the requested sge number.
iWARP doesn't support multiple sg elements for
RDMA READ work requests.

Signed-off-by: Tatyana Nikolova <tatyana.e.nikolova-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Faisal Latif <faisal.latif-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
 drivers/infiniband/hw/i40iw/i40iw_verbs.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
index 2d832c7..45f70f5 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
@@ -74,7 +74,7 @@ static int i40iw_query_device(struct ib_device *ibdev,
 	props->max_cqe = iwdev->max_cqe;
 	props->max_mr = iwdev->max_mr;
 	props->max_pd = iwdev->max_pd;
-	props->max_sge_rd = 1;
+	props->max_sge_rd = I40IW_MAX_SGE_RD;
 	props->max_qp_rd_atom = I40IW_MAX_IRD_SIZE;
 	props->max_qp_init_rd_atom = props->max_qp_rd_atom;
 	props->atomic_cap = IB_ATOMIC_NONE;
@@ -2117,6 +2117,10 @@ static int i40iw_post_send(struct ib_qp *ibqp,
 			inv_stag = true;
 			/* fall-through*/
 		case IB_WR_RDMA_READ:
+			if (ib_wr->num_sge > I40IW_MAX_SGE_RD) {
+				err = -EINVAL;
+				break;
+			}
 			info.op_type = I40IW_OP_TYPE_RDMA_READ;
 			info.op.rdma_read.rem_addr.tag_off = rdma_wr(ib_wr)->remote_addr;
 			info.op.rdma_read.rem_addr.stag = rdma_wr(ib_wr)->rkey;
-- 
2.7.4

--
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 5/7] RDMA/i40iw: Fix for using one sge for RDMA READ

[Leon Romanovsky]

[-- Attachment #1: Type: text/plain, Size: 2001 bytes --]

On Fri, Apr 22, 2016 at 02:14:27PM -0500, Tatyana Nikolova wrote:
> From: Shiraz Saleem <shiraz.saleem-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> 
> A check is added to validate the requested sge number.
> iWARP doesn't support multiple sg elements for
> RDMA READ work requests.
> 
> Signed-off-by: Tatyana Nikolova <tatyana.e.nikolova-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> Signed-off-by: Faisal Latif <faisal.latif-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> ---
>  drivers/infiniband/hw/i40iw/i40iw_verbs.c | 6 +++++-
>  1 file changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
> index 2d832c7..45f70f5 100644
> --- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
> +++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
> @@ -74,7 +74,7 @@ static int i40iw_query_device(struct ib_device *ibdev,
>  	props->max_cqe = iwdev->max_cqe;
>  	props->max_mr = iwdev->max_mr;
>  	props->max_pd = iwdev->max_pd;
> -	props->max_sge_rd = 1;
> +	props->max_sge_rd = I40IW_MAX_SGE_RD;
>  	props->max_qp_rd_atom = I40IW_MAX_IRD_SIZE;
>  	props->max_qp_init_rd_atom = props->max_qp_rd_atom;
>  	props->atomic_cap = IB_ATOMIC_NONE;
> @@ -2117,6 +2117,10 @@ static int i40iw_post_send(struct ib_qp *ibqp,
>  			inv_stag = true;
>  			/* fall-through*/
                        ^^^^^^^^^^^^^^^^
Do you support multiple SGE for RDMA WRITE?


>  		case IB_WR_RDMA_READ:
> +			if (ib_wr->num_sge > I40IW_MAX_SGE_RD) {
> +				err = -EINVAL;
> +				break;
> +			}
>  			info.op_type = I40IW_OP_TYPE_RDMA_READ;
>  			info.op.rdma_read.rem_addr.tag_off = rdma_wr(ib_wr)->remote_addr;
>  			info.op.rdma_read.rem_addr.stag = rdma_wr(ib_wr)->rkey;
> -- 
> 2.7.4
> 
> --
> 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

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

RE: [PATCH 5/7] RDMA/i40iw: Fix for using one sge for RDMA READ

[Nikolova, Tatyana E]

Hi Leon,

We support multiple sges for RDMA WRITE. In the patch, the fall-through case above RDMA READ is RDMA READ WITH INVALIDATE.

Thanks,
Tatyana

-----Original Message-----
From: Leon Romanovsky [mailto:leon-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org] 
Sent: Sunday, April 24, 2016 2:35 AM
To: Nikolova, Tatyana E <tatyana.e.nikolova-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Cc: dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org; linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org; e1000-rdma-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org; Latif, Faisal <faisal.latif-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Subject: Re: [PATCH 5/7] RDMA/i40iw: Fix for using one sge for RDMA READ

On Fri, Apr 22, 2016 at 02:14:27PM -0500, Tatyana Nikolova wrote:
> From: Shiraz Saleem <shiraz.saleem-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> 
> A check is added to validate the requested sge number.
> iWARP doesn't support multiple sg elements for RDMA READ work 
> requests.
> 
> Signed-off-by: Tatyana Nikolova <tatyana.e.nikolova-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> Signed-off-by: Faisal Latif <faisal.latif-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> ---
>  drivers/infiniband/hw/i40iw/i40iw_verbs.c | 6 +++++-
>  1 file changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c 
> b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
> index 2d832c7..45f70f5 100644
> --- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
> +++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
> @@ -74,7 +74,7 @@ static int i40iw_query_device(struct ib_device *ibdev,
>  	props->max_cqe = iwdev->max_cqe;
>  	props->max_mr = iwdev->max_mr;
>  	props->max_pd = iwdev->max_pd;
> -	props->max_sge_rd = 1;
> +	props->max_sge_rd = I40IW_MAX_SGE_RD;
>  	props->max_qp_rd_atom = I40IW_MAX_IRD_SIZE;
>  	props->max_qp_init_rd_atom = props->max_qp_rd_atom;
>  	props->atomic_cap = IB_ATOMIC_NONE;
> @@ -2117,6 +2117,10 @@ static int i40iw_post_send(struct ib_qp *ibqp,
>  			inv_stag = true;
>  			/* fall-through*/
                        ^^^^^^^^^^^^^^^^ Do you support multiple SGE for RDMA WRITE?


>  		case IB_WR_RDMA_READ:
> +			if (ib_wr->num_sge > I40IW_MAX_SGE_RD) {
> +				err = -EINVAL;
> +				break;
> +			}
>  			info.op_type = I40IW_OP_TYPE_RDMA_READ;
>  			info.op.rdma_read.rem_addr.tag_off = rdma_wr(ib_wr)->remote_addr;
>  			info.op.rdma_read.rem_addr.stag = rdma_wr(ib_wr)->rkey;
> --
> 2.7.4
> 
> --
> 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
--
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 5/7] RDMA/i40iw: Fix for using one sge for RDMA READ

[Leon Romanovsky]

[-- Attachment #1: Type: text/plain, Size: 3022 bytes --]

On Tue, Apr 26, 2016 at 03:27:46PM +0000, Nikolova, Tatyana E wrote:
> Hi Leon,
> 
> We support multiple sges for RDMA WRITE. In the patch, the fall-through case above RDMA READ is RDMA READ WITH INVALIDATE.

I believe that we are looking on the different trees.
In official tree, the case above RDMA_READ is RDMA_WRITE.

> 
> Thanks,
> Tatyana
> 
> -----Original Message-----
> From: Leon Romanovsky [mailto:leon-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org] 
> Sent: Sunday, April 24, 2016 2:35 AM
> To: Nikolova, Tatyana E <tatyana.e.nikolova-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> Cc: dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org; linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org; e1000-rdma-5NWGOfrQmneHXe+LvDLADg@public.gmane.orgrceforge.net; Latif, Faisal <faisal.latif-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> Subject: Re: [PATCH 5/7] RDMA/i40iw: Fix for using one sge for RDMA READ
> 
> On Fri, Apr 22, 2016 at 02:14:27PM -0500, Tatyana Nikolova wrote:
> > From: Shiraz Saleem <shiraz.saleem-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> > 
> > A check is added to validate the requested sge number.
> > iWARP doesn't support multiple sg elements for RDMA READ work 
> > requests.
> > 
> > Signed-off-by: Tatyana Nikolova <tatyana.e.nikolova-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> > Signed-off-by: Faisal Latif <faisal.latif-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
> > ---
> >  drivers/infiniband/hw/i40iw/i40iw_verbs.c | 6 +++++-
> >  1 file changed, 5 insertions(+), 1 deletion(-)
> > 
> > diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c 
> > b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
> > index 2d832c7..45f70f5 100644
> > --- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
> > +++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
> > @@ -74,7 +74,7 @@ static int i40iw_query_device(struct ib_device *ibdev,
> >  	props->max_cqe = iwdev->max_cqe;
> >  	props->max_mr = iwdev->max_mr;
> >  	props->max_pd = iwdev->max_pd;
> > -	props->max_sge_rd = 1;
> > +	props->max_sge_rd = I40IW_MAX_SGE_RD;
> >  	props->max_qp_rd_atom = I40IW_MAX_IRD_SIZE;
> >  	props->max_qp_init_rd_atom = props->max_qp_rd_atom;
> >  	props->atomic_cap = IB_ATOMIC_NONE;
> > @@ -2117,6 +2117,10 @@ static int i40iw_post_send(struct ib_qp *ibqp,
> >  			inv_stag = true;
> >  			/* fall-through*/
>                         ^^^^^^^^^^^^^^^^ Do you support multiple SGE for RDMA WRITE?
> 
> 
> >  		case IB_WR_RDMA_READ:
> > +			if (ib_wr->num_sge > I40IW_MAX_SGE_RD) {
> > +				err = -EINVAL;
> > +				break;
> > +			}
> >  			info.op_type = I40IW_OP_TYPE_RDMA_READ;
> >  			info.op.rdma_read.rem_addr.tag_off = rdma_wr(ib_wr)->remote_addr;
> >  			info.op.rdma_read.rem_addr.stag = rdma_wr(ib_wr)->rkey;
> > --
> > 2.7.4
> > 
> > --
> > 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

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: [PATCH 5/7] RDMA/i40iw: Fix for using one sge for RDMA READ

[Doug Ledford]

[-- Attachment #1: Type: text/plain, Size: 3258 bytes --]

On 04/27/2016 02:25 AM, Leon Romanovsky wrote:
> On Tue, Apr 26, 2016 at 03:27:46PM +0000, Nikolova, Tatyana E wrote:
>> Hi Leon,
>>
>> We support multiple sges for RDMA WRITE. In the patch, the fall-through case above RDMA READ is RDMA READ WITH INVALIDATE.
> 
> I believe that we are looking on the different trees.
> In official tree, the case above RDMA_READ is RDMA_WRITE.

Theses patches presumably are on top of the 16 previous patches to the
i40iw driver.

>>
>> Thanks,
>> Tatyana
>>
>> -----Original Message-----
>> From: Leon Romanovsky [mailto:leon-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org] 
>> Sent: Sunday, April 24, 2016 2:35 AM
>> To: Nikolova, Tatyana E <tatyana.e.nikolova-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
>> Cc: dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org; linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org; e1000-rdma@lists.sourceforge.net; Latif, Faisal <faisal.latif-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
>> Subject: Re: [PATCH 5/7] RDMA/i40iw: Fix for using one sge for RDMA READ
>>
>> On Fri, Apr 22, 2016 at 02:14:27PM -0500, Tatyana Nikolova wrote:
>>> From: Shiraz Saleem <shiraz.saleem-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
>>>
>>> A check is added to validate the requested sge number.
>>> iWARP doesn't support multiple sg elements for RDMA READ work 
>>> requests.
>>>
>>> Signed-off-by: Tatyana Nikolova <tatyana.e.nikolova-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
>>> Signed-off-by: Faisal Latif <faisal.latif-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
>>> ---
>>>  drivers/infiniband/hw/i40iw/i40iw_verbs.c | 6 +++++-
>>>  1 file changed, 5 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c 
>>> b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
>>> index 2d832c7..45f70f5 100644
>>> --- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
>>> +++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
>>> @@ -74,7 +74,7 @@ static int i40iw_query_device(struct ib_device *ibdev,
>>>  	props->max_cqe = iwdev->max_cqe;
>>>  	props->max_mr = iwdev->max_mr;
>>>  	props->max_pd = iwdev->max_pd;
>>> -	props->max_sge_rd = 1;
>>> +	props->max_sge_rd = I40IW_MAX_SGE_RD;
>>>  	props->max_qp_rd_atom = I40IW_MAX_IRD_SIZE;
>>>  	props->max_qp_init_rd_atom = props->max_qp_rd_atom;
>>>  	props->atomic_cap = IB_ATOMIC_NONE;
>>> @@ -2117,6 +2117,10 @@ static int i40iw_post_send(struct ib_qp *ibqp,
>>>  			inv_stag = true;
>>>  			/* fall-through*/
>>                         ^^^^^^^^^^^^^^^^ Do you support multiple SGE for RDMA WRITE?
>>
>>
>>>  		case IB_WR_RDMA_READ:
>>> +			if (ib_wr->num_sge > I40IW_MAX_SGE_RD) {
>>> +				err = -EINVAL;
>>> +				break;
>>> +			}
>>>  			info.op_type = I40IW_OP_TYPE_RDMA_READ;
>>>  			info.op.rdma_read.rem_addr.tag_off = rdma_wr(ib_wr)->remote_addr;
>>>  			info.op.rdma_read.rem_addr.stag = rdma_wr(ib_wr)->rkey;
>>> --
>>> 2.7.4
>>>
>>> --
>>> 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


-- 
Doug Ledford <dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
              GPG KeyID: 0E572FDD



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

[PATCH 6/7] RDMA/i40iw: Fix for checking if the QP is destroyed

[Tatyana Nikolova]

Fix for checking if the QP associated with a completion
has been destroyed while processing CQ elements.
If that is the case, move the CQ head to the next element
and continue completion processing.

Signed-off-by: Tatyana Nikolova <tatyana.e.nikolova-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Faisal Latif <faisal.latif-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
 drivers/infiniband/hw/i40iw/i40iw_status.h | 1 +
 drivers/infiniband/hw/i40iw/i40iw_uk.c     | 5 +++++
 drivers/infiniband/hw/i40iw/i40iw_verbs.c  | 2 ++
 3 files changed, 8 insertions(+)

diff --git a/drivers/infiniband/hw/i40iw/i40iw_status.h b/drivers/infiniband/hw/i40iw/i40iw_status.h
index b0110c1..91c4217 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_status.h
+++ b/drivers/infiniband/hw/i40iw/i40iw_status.h
@@ -95,6 +95,7 @@ enum i40iw_status_code {
 	I40IW_ERR_INVALID_MAC_ADDR = -65,
 	I40IW_ERR_BAD_STAG      = -66,
 	I40IW_ERR_CQ_COMPL_ERROR = -67,
+	I40IW_ERR_QUEUE_DESTROYED = -68
 
 };
 #endif
diff --git a/drivers/infiniband/hw/i40iw/i40iw_uk.c b/drivers/infiniband/hw/i40iw/i40iw_uk.c
index 2cd9091..e35faea8 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_uk.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_uk.c
@@ -802,6 +802,10 @@ static enum i40iw_status_code i40iw_cq_poll_completion(struct i40iw_cq_uk *cq,
 	info->is_srq = (bool)RS_64(qword3, I40IWCQ_SRQ);
 
 	qp = (struct i40iw_qp_uk *)(unsigned long)comp_ctx;
+	if (!qp) {
+		ret_code = I40IW_ERR_QUEUE_DESTROYED;
+		goto exit;
+	}
 	wqe_idx = (u32)RS_64(qword3, I40IW_CQ_WQEIDX);
 	info->qp_handle = (i40iw_qp_handle)(unsigned long)qp;
 
@@ -859,6 +863,7 @@ static enum i40iw_status_code i40iw_cq_poll_completion(struct i40iw_cq_uk *cq,
 
 	ret_code = 0;
 
+exit:
 	if (!ret_code &&
 	    (info->comp_status == I40IW_COMPL_STATUS_FLUSHED))
 		if (pring && (I40IW_RING_MORE_WORK(*pring)))
diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
index 45f70f5..eaa79c9 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
@@ -2259,6 +2259,8 @@ static int i40iw_poll_cq(struct ib_cq *ibcq,
 		ret = ukcq->ops.iw_cq_poll_completion(ukcq, &cq_poll_info, true);
 		if (ret == I40IW_ERR_QUEUE_EMPTY) {
 			break;
+		} else if (ret == I40IW_ERR_QUEUE_DESTROYED) {
+			continue;
 		} else if (ret) {
 			if (!cqe_count)
 				cqe_count = -1;
-- 
2.7.4

--
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 7/7] RDMA/i40iw: Fix for removing quad hash entries

[Tatyana Nikolova]

Fix for removing a quad hash entry when the
corresponding quad hash entry hasn't been added,
which is the case in loopback connections

Signed-off-by: Tatyana Nikolova <Tatyana.E.Nikolova-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Faisal Latif <faisal.latif-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
 drivers/infiniband/hw/i40iw/i40iw_cm.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/infiniband/hw/i40iw/i40iw_cm.c b/drivers/infiniband/hw/i40iw/i40iw_cm.c
index 8cb4b87..d2fa725 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_cm.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_cm.c
@@ -3729,6 +3729,7 @@ int i40iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param)
 	struct sockaddr_in *raddr;
 	struct sockaddr_in6 *laddr6;
 	struct sockaddr_in6 *raddr6;
+	bool qhash_set = false;
 	int apbvt_set = 0;
 	enum i40iw_status_code status;
 
@@ -3787,6 +3788,7 @@ int i40iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param)
 					    true);
 		if (status)
 			return -EINVAL;
+		qhash_set = true;
 	}
 	status = i40iw_manage_apbvt(iwdev, cm_info.loc_port, I40IW_MANAGE_APBVT_ADD);
 	if (status) {
@@ -3814,7 +3816,7 @@ int i40iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param)
 		cm_node->ord_size = 1;
 
 	cm_node->apbvt_set = apbvt_set;
-	cm_node->qhash_set = true;
+	cm_node->qhash_set = qhash_set;
 	iwqp->cm_node = cm_node;
 	cm_node->iwqp = iwqp;
 	iwqp->cm_id = cm_id;
-- 
2.7.4

--
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 0/7] RDMA/i40iw: Fixes for i40iw driver

[Doug Ledford]

[-- Attachment #1: Type: text/plain, Size: 1446 bytes --]

On 04/22/2016 03:14 PM, Tatyana Nikolova wrote:
> This patch series includes a few fixes for the original
> Intel X722 iWARP driver i40iw.
> 
> It is based on Doug Ledford's https://github.com/dledford/linux.git
> branch: k.o/for-4.6-rc. 
> 
> It needs to be applied after the patch series submitted by Mustafa Ismail - 
> "[PATCH V3 00/16] RDMA/Intel X722 iWARP driver update"

I've taken both that set and this set into my for-next branch.

> Chien Tin Tung (1):
>   RDMA/i40iw: Correct STag mask to min of 14 bits
> 
> Mohammad Khan (1):
>   RDMA/i40iw: Fix for a NOP WQE size
> 
> Shiraz Saleem (3):
>   RDMA/i40iw: Fixes for WQE alignment
>   RDMA/i40iw: Fix for the size of kernel mode SQ
>   RDMA/i40iw: Fix for using one sge for RDMA READ
> 
> Tatyana Nikolova (2):
>   RDMA/i40iw: Fix for checking if the QP is destroyed
>   RDMA/i40iw: Fix for removing quad hash entries
> 
>  drivers/infiniband/hw/i40iw/i40iw_cm.c     |  4 +++-
>  drivers/infiniband/hw/i40iw/i40iw_d.h      |  2 ++
>  drivers/infiniband/hw/i40iw/i40iw_hw.c     |  3 ++-
>  drivers/infiniband/hw/i40iw/i40iw_status.h |  1 +
>  drivers/infiniband/hw/i40iw/i40iw_uk.c     | 26 ++++++++++++++++++++++++--
>  drivers/infiniband/hw/i40iw/i40iw_verbs.c  | 10 +++++++---
>  6 files changed, 39 insertions(+), 7 deletions(-)
> 


-- 
Doug Ledford <dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
              GPG KeyID: 0E572FDD



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