linux-next: manual merge of the rdma tree with the origin tree

linux-next: manual merge of the rdma tree with the origin tree

[Mark Brown]

[-- Attachment #1: Type: text/plain, Size: 2316 bytes --]

Hi all,

Today's linux-next merge of the rdma tree got conflicts in:

  drivers/infiniband/hw/mana/qp.c
  drivers/infiniband/hw/mlx4/srq.c

between commits:

  c54c7e4cb679c ("RDMA/mlx4: Fix resource leak on error in mlx4_ib_create_srq()")
  6aaa978c6b621 ("RDMA/mana: Fix error unwind in mana_ib_create_qp_rss()")

from the origin tree and commits:

  271746ec81325 ("RDMA: Convert drivers using min to ib_respond_udata()")
  22d072b3bcc59 ("RDMA: Convert drivers using sizeof() to ib_respond_udata()")
  23a45206d20de ("RDMA: Use proper driver data response structs instead of open coding")

from the rdma tree.

I fixed it up (see below) and can carry the fix as necessary. This
is now fixed as far as linux-next is concerned, but any non trivial
conflicts should be mentioned to your upstream maintainer when your tree
is submitted for merging.  You may also want to consider cooperating
with the maintainer of the conflicting tree to minimise any particularly
complex conflicts.

diff --cc drivers/infiniband/hw/mana/qp.c
index 0fbcf449c134b,c50a1eafee58c..0000000000000
--- a/drivers/infiniband/hw/mana/qp.c
+++ b/drivers/infiniband/hw/mana/qp.c
@@@ -212,13 -210,9 +212,9 @@@ static int mana_ib_create_qp_rss(struc
  	if (ret)
  		goto fail;
  
- 	ret = ib_copy_to_udata(udata, &resp, sizeof(resp));
- 	if (ret) {
- 		ibdev_dbg(&mdev->ib_dev,
- 			  "Failed to copy to udata create rss-qp, %d\n",
- 			  ret);
+ 	ret = ib_respond_udata(udata, resp);
+ 	if (ret)
 -		goto fail;
 +		goto err_disable_vport_rx;
- 	}
  
  	kfree(mana_ind_table);
  
diff --cc drivers/infiniband/hw/mlx4/srq.c
index 767840736d583,0b4df4f48ca14..0000000000000
--- a/drivers/infiniband/hw/mlx4/srq.c
+++ b/drivers/infiniband/hw/mlx4/srq.c
@@@ -191,11 -191,15 +191,15 @@@ int mlx4_ib_create_srq(struct ib_srq *i
  	srq->msrq.event = mlx4_ib_srq_event;
  	srq->ibsrq.ext.xrc.srq_num = srq->msrq.srqn;
  
- 	if (udata)
- 		if (ib_copy_to_udata(udata, &srq->msrq.srqn, sizeof (__u32))) {
- 			err = -EFAULT;
+ 	if (udata) {
+ 		struct mlx4_ib_create_srq_resp uresp = {
+ 			.srqn = srq->msrq.srqn
+ 		};
+ 
+ 		err = ib_respond_udata(udata, uresp);
+ 		if (err)
 -			goto err_wrid;
 +			goto err_srq;
- 		}
+ 	}
  
  	init_attr->attr.max_wr = srq->msrq.max - 1;
  

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

linux-next: manual merge of the rdma tree with the origin tree

[Mark Brown]

Hi all,

Today's linux-next merge of the rdma tree got a conflict in:

  drivers/infiniband/hw/mana/cq.c

between commit:

  887bfe5986396 ("RDMA/mana_ib: check cqe length for kernel CQs")

from the origin tree and commit:

  f972bde7326e9 ("RDMA/mana_ib: Take CQ type from the device type")

from the rdma tree.

I fixed it up (see below) and can carry the fix as necessary. This
is now fixed as far as linux-next is concerned, but any non trivial
conflicts should be mentioned to your upstream maintainer when your tree
is submitted for merging.  You may also want to consider cooperating
with the maintainer of the conflicting tree to minimise any particularly
complex conflicts.

diff --cc drivers/infiniband/hw/mana/cq.c
index 7600412b0739f,2dce1b6771150..0000000000000
--- a/drivers/infiniband/hw/mana/cq.c
+++ b/drivers/infiniband/hw/mana/cq.c
@@@ -55,11 -54,6 +54,10 @@@ int mana_ib_create_cq(struct ib_cq *ibc
  							  ibucontext);
  		doorbell = mana_ucontext->doorbell;
  	} else {
- 		is_rnic_cq = true;
 +		if (attr->cqe > U32_MAX / COMP_ENTRY_SIZE / 2 + 1) {
 +			ibdev_dbg(ibdev, "CQE %d exceeding limit\n", attr->cqe);
 +			return -EINVAL;
 +		}
  		buf_size = MANA_PAGE_ALIGN(roundup_pow_of_two(attr->cqe * COMP_ENTRY_SIZE));
  		cq->cqe = buf_size / COMP_ENTRY_SIZE;
  		err = mana_ib_create_kernel_queue(mdev, buf_size, GDMA_CQ, &cq->queue);

linux-next: manual merge of the rdma tree with the origin tree

[broonie]

Hi all,

Today's linux-next merge of the rdma tree got conflicts in:

  drivers/infiniband/hw/hns/hns_roce_hw_v2.c
  drivers/infiniband/hw/hns/hns_roce_main.c

between commit:

  45baad7dd98f4 ("RDMA/hns: Remove the num_qpc_timer variable")

from the origin tree and commits:

  29dc063596772 ("RDMA/hns: Remove redundant 'use_lowmem' argument from hns_roce_init_hem_table()")
  5436272c8cf4e ("RDMA/hns: Remove redundant 'num_mtt_segs' and 'max_extend_sg'")

from the rdma tree.

I fixed it up (see below) and can carry the fix as necessary. This
is now fixed as far as linux-next is concerned, but any non trivial
conflicts should be mentioned to your upstream maintainer when your tree
is submitted for merging.  You may also want to consider cooperating
with the maintainer of the conflicting tree to minimise any particularly
complex conflicts.

diff --cc drivers/infiniband/hw/hns/hns_roce_hw_v2.c
index c780646bd60ac,2d0192057d1af..0000000000000
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
@@@ -2272,7 -2270,7 +2270,6 @@@ static int hns_roce_query_pf_caps(struc
  	caps->max_sq_inline	     = le16_to_cpu(resp_a->max_sq_inline);
  	caps->max_rq_sg		     = le16_to_cpu(resp_a->max_rq_sg);
  	caps->max_rq_sg = roundup_pow_of_two(caps->max_rq_sg);
- 	caps->max_extend_sg	     = le32_to_cpu(resp_a->max_extend_sg);
 -	caps->num_qpc_timer	     = le16_to_cpu(resp_a->num_qpc_timer);
  	caps->max_srq_sges	     = le16_to_cpu(resp_a->max_srq_sges);
  	caps->max_srq_sges = roundup_pow_of_two(caps->max_srq_sges);
  	caps->num_aeq_vectors	     = resp_a->num_aeq_vectors;
diff --cc drivers/infiniband/hw/hns/hns_roce_main.c
index 4ccb217b2841d,53c53c20360d1..0000000000000
--- a/drivers/infiniband/hw/hns/hns_roce_main.c
+++ b/drivers/infiniband/hw/hns/hns_roce_main.c
@@@ -725,10 -734,10 +734,10 @@@ static int hns_roce_init_hem(struct hns
  		ret = hns_roce_init_hem_table(hr_dev, &hr_dev->qpc_timer_table,
  					      HEM_TYPE_QPC_TIMER,
  					      hr_dev->caps.qpc_timer_entry_sz,
- 					      hr_dev->caps.qpc_timer_bt_num, 1);
 -					      hr_dev->caps.num_qpc_timer);
++					      hr_dev->caps.qpc_timer_bt_num);
  		if (ret) {
  			dev_err(dev,
- 				"Failed to init QPC timer memory, aborting.\n");
+ 				"failed to init QPC timer memory, aborting.\n");
  			goto err_unmap_ctx;
  		}
  	}