[PATCH rdma-rc 0/3] mlx5 and NLDEV fixes

[PATCH rdma-rc 0/3] mlx5 and NLDEV fixes

[Leon Romanovsky]

Hi Doug,

There are three patches targeted for this -RC, while mlx5 patches can
easily go to -next, the nldev patch better to be forwarded to Linus in
this cycle.

Because, we are in -rc7 now, I would really appreciate to hear your
plans regarding nldev patch.

The patches are available in the git repository at:
  git.kernel.org/pub/scm/linux/kernel/git/leon/linux-rdma.git tags/rdma-rc-2017-10-30

	Thanks
---------------------------------------

Leon Romanovsky (1):
  RDMA/nldev: Enforce device index check for port callback

Majd Dibbiny (2):
  IB/mlx5: Assign send CQ and recv CQ of UMR QP
  IB/mlx5: Fix RoCE Address Path fields

 drivers/infiniband/core/nldev.c   | 4 +++-
 drivers/infiniband/hw/mlx5/main.c | 2 ++
 drivers/infiniband/hw/mlx5/qp.c   | 8 ++++++--
 3 files changed, 11 insertions(+), 3 deletions(-)

--
2.14.2

--
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 rdma-rc 1/3] IB/mlx5: Assign send CQ and recv CQ of UMR QP

[Leon Romanovsky]

From: Majd Dibbiny <majd-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>

The UMR's QP is created by calling mlx5_ib_create_qp directly, and
therefore the send CQ and the recv CQ on the ibqp weren't assigned.

Assign them right after calling the mlx5_ib_create_qp to assure
that any access to those pointers will work as expected and won't
crash the system as might happen as part of reset flow.

Fixes: e126ba97dba9 ("mlx5: Add driver for Mellanox Connect-IB adapters")
Signed-off-by: Majd Dibbiny <majd-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
Reviewed-by: Yishai Hadas <yishaih-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
Signed-off-by: Leon Romanovsky <leon-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
---
 drivers/infiniband/hw/mlx5/main.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index 552f7bd4ecc3..5aff1e33d984 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -3097,6 +3097,8 @@ static int create_umr_res(struct mlx5_ib_dev *dev)
 	qp->real_qp    = qp;
 	qp->uobject    = NULL;
 	qp->qp_type    = MLX5_IB_QPT_REG_UMR;
+	qp->send_cq    = init_attr->send_cq;
+	qp->recv_cq    = init_attr->recv_cq;
 
 	attr->qp_state = IB_QPS_INIT;
 	attr->port_num = 1;
-- 
2.14.2

--
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 rdma-rc 2/3] IB/mlx5: Fix RoCE Address Path fields

[Leon Romanovsky]

From: Majd Dibbiny <majd-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>

When working over a RoCE network, the UDP source port should be set only
for statically connected QPs (RC, UC and XRC).

Fixes: 2811ba51b049 ("IB/mlx5: Add RoCE fields to Address Vector")
Signed-off-by: Majd Dibbiny <majd-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
Reviewed-by: Yishai Hadas <yishaih-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
Signed-off-by: Leon Romanovsky <leon-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
---
 drivers/infiniband/hw/mlx5/qp.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c
index acb79d3a4f1d..a63060b93a43 100644
--- a/drivers/infiniband/hw/mlx5/qp.c
+++ b/drivers/infiniband/hw/mlx5/qp.c
@@ -2283,8 +2283,12 @@ static int mlx5_set_path(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp,
 		if (err)
 			return err;
 		memcpy(path->rmac, ah->roce.dmac, sizeof(ah->roce.dmac));
-		path->udp_sport = mlx5_get_roce_udp_sport(dev, port,
-							  grh->sgid_index);
+		if (qp->ibqp.qp_type == IB_QPT_RC ||
+		    qp->ibqp.qp_type == IB_QPT_UC ||
+		    qp->ibqp.qp_type == IB_QPT_XRC_INI ||
+		    qp->ibqp.qp_type == IB_QPT_XRC_TGT)
+			path->udp_sport = mlx5_get_roce_udp_sport(dev, port,
+								  grh->sgid_index);
 		path->dci_cfi_prio_sl = (sl & 0x7) << 4;
 		if (gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP)
 			path->ecn_dscp = (grh->traffic_class >> 2) & 0x3f;
-- 
2.14.2

--
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 rdma-rc 3/3] RDMA/nldev: Enforce device index check for port callback

[Leon Romanovsky]

From: Leon Romanovsky <leonro-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>

IB device index is nldev's handler and it should be always checked.

Cc: stable-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
Fixes: c3f66f7b0052 ("RDMA/netlink: Implement nldev port doit callback")
Signed-off-by: Leon Romanovsky <leonro-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
---
 drivers/infiniband/core/nldev.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/infiniband/core/nldev.c b/drivers/infiniband/core/nldev.c
index 3ba24c428c3b..2fae850a3eff 100644
--- a/drivers/infiniband/core/nldev.c
+++ b/drivers/infiniband/core/nldev.c
@@ -214,7 +214,9 @@ static int nldev_port_get_doit(struct sk_buff *skb, struct nlmsghdr *nlh,

 	err = nlmsg_parse(nlh, 0, tb, RDMA_NLDEV_ATTR_MAX - 1,
 			  nldev_policy, extack);
-	if (err || !tb[RDMA_NLDEV_ATTR_PORT_INDEX])
+	if (err ||
+	    !tb[RDMA_NLDEV_ATTR_DEV_INDEX] ||
+	    !tb[RDMA_NLDEV_ATTR_PORT_INDEX])
 		return -EINVAL;

 	index = nla_get_u32(tb[RDMA_NLDEV_ATTR_DEV_INDEX]);
--
2.14.2

--
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 rdma-rc 3/3] RDMA/nldev: Enforce device index check for port callback

[Leon Romanovsky]

From: Leon Romanovsky <leonro@mellanox.com>

IB device index is nldev's handler and it should be always checked.

Cc: stable@vger.kernel.org
Fixes: c3f66f7b0052 ("RDMA/netlink: Implement nldev port doit callback")
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
---
 drivers/infiniband/core/nldev.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/infiniband/core/nldev.c b/drivers/infiniband/core/nldev.c
index 3ba24c428c3b..2fae850a3eff 100644
--- a/drivers/infiniband/core/nldev.c
+++ b/drivers/infiniband/core/nldev.c
@@ -214,7 +214,9 @@ static int nldev_port_get_doit(struct sk_buff *skb, struct nlmsghdr *nlh,

 	err = nlmsg_parse(nlh, 0, tb, RDMA_NLDEV_ATTR_MAX - 1,
 			  nldev_policy, extack);
-	if (err || !tb[RDMA_NLDEV_ATTR_PORT_INDEX])
+	if (err ||
+	    !tb[RDMA_NLDEV_ATTR_DEV_INDEX] ||
+	    !tb[RDMA_NLDEV_ATTR_PORT_INDEX])
 		return -EINVAL;

 	index = nla_get_u32(tb[RDMA_NLDEV_ATTR_DEV_INDEX]);
--
2.14.2

Re: [PATCH rdma-rc 0/3] mlx5 and NLDEV fixes

[Doug Ledford]

[-- Attachment #1: Type: text/plain, Size: 1054 bytes --]

On Mon, 2017-10-30 at 14:23 +0200, Leon Romanovsky wrote:
> Hi Doug,
> 
> There are three patches targeted for this -RC, while mlx5 patches can
> easily go to -next, the nldev patch better to be forwarded to Linus in
> this cycle.
> 
> Because, we are in -rc7 now, I would really appreciate to hear your
> plans regarding nldev patch.
> 
> The patches are available in the git repository at:
>   git.kernel.org/pub/scm/linux/kernel/git/leon/linux-rdma.git tags/rdma-rc-2017-10-30
> 
> 	Thanks
> ---------------------------------------
> 
> Leon Romanovsky (1):
>   RDMA/nldev: Enforce device index check for port callback

This went into -rc7 direct from you to Linus, so I dropped it here.

> Majd Dibbiny (2):
>   IB/mlx5: Assign send CQ and recv CQ of UMR QP
>   IB/mlx5: Fix RoCE Address Path fields

But I went ahead and took these two.  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

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 833 bytes --]