RDMA/uverbs: Allow drivers to create a new HW object during rereg_mr

mlx5 has an ugly flow where it tries to allocate a new MR and replace the
existing MR in the same memory during rereg. This is very complicated and
buggy. Instead of trying to replace in-place inside the driver, provide
support from uverbs to change the entire HW object assigned to a handle
during rereg_mr.

Since destroying a MR is allowed to fail (ie if a MW is pointing at it)
and can't be detected in advance, the algorithm creates a completely new
uobject to hold the new MR and swaps the IDR entries of the two objects.

The old MR in the temporary IDR entry is destroyed, and if it fails
rereg_mr succeeds and destruction is deferred to FD release. This
complexity is why this cannot live in a driver safely.

Link: https://lore.kernel.org/r/20201130075839.278575-4-leon@kernel.org
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>

6 files changed, 35 insertions, 32 deletions

diff --git a/drivers/infiniband/hw/hns/hns_roce_device.h b/drivers/infiniband/hw/hns/hns_roce_device.h
index a5c6bb073569..70ae37bad77e 100644
--- a/drivers/infiniband/hw/hns/hns_roce_device.h
+++ b/drivers/infiniband/hw/hns/hns_roce_device.h
@@ -1223,9 +1223,10 @@ struct ib_mr *hns_roce_get_dma_mr(struct ib_pd *pd, int acc);
 struct ib_mr *hns_roce_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 				   u64 virt_addr, int access_flags,
 				   struct ib_udata *udata);
-int hns_roce_rereg_user_mr(struct ib_mr *mr, int flags, u64 start, u64 length,
-			   u64 virt_addr, int mr_access_flags, struct ib_pd *pd,
-			   struct ib_udata *udata);
+struct ib_mr *hns_roce_rereg_user_mr(struct ib_mr *mr, int flags, u64 start,
+				     u64 length, u64 virt_addr,
+				     int mr_access_flags, struct ib_pd *pd,
+				     struct ib_udata *udata);
 struct ib_mr *hns_roce_alloc_mr(struct ib_pd *pd, enum ib_mr_type mr_type,
 				u32 max_num_sg);
 int hns_roce_map_mr_sg(struct ib_mr *ibmr, struct scatterlist *sg, int sg_nents,
diff --git a/drivers/infiniband/hw/hns/hns_roce_mr.c b/drivers/infiniband/hw/hns/hns_roce_mr.c
index 87e2e6236c69..98671925debf 100644
--- a/drivers/infiniband/hw/hns/hns_roce_mr.c
+++ b/drivers/infiniband/hw/hns/hns_roce_mr.c
@@ -328,9 +328,10 @@ static int rereg_mr_trans(struct ib_mr *ibmr, int flags,
 	return ret;
 }
 
-int hns_roce_rereg_user_mr(struct ib_mr *ibmr, int flags, u64 start, u64 length,
-			   u64 virt_addr, int mr_access_flags, struct ib_pd *pd,
-			   struct ib_udata *udata)
+struct ib_mr *hns_roce_rereg_user_mr(struct ib_mr *ibmr, int flags, u64 start,
+				     u64 length, u64 virt_addr,
+				     int mr_access_flags, struct ib_pd *pd,
+				     struct ib_udata *udata)
 {
 	struct hns_roce_dev *hr_dev = to_hr_dev(ibmr->device);
 	struct ib_device *ib_dev = &hr_dev->ib_dev;
@@ -341,11 +342,11 @@ int hns_roce_rereg_user_mr(struct ib_mr *ibmr, int flags, u64 start, u64 length,
 	int ret;
 
 	if (!mr->enabled)
-		return -EINVAL;
+		return ERR_PTR(-EINVAL);
 
 	mailbox = hns_roce_alloc_cmd_mailbox(hr_dev);
 	if (IS_ERR(mailbox))
-		return PTR_ERR(mailbox);
+		return ERR_CAST(mailbox);
 
 	mtpt_idx = key_to_hw_index(mr->key) & (hr_dev->caps.num_mtpts - 1);
 	ret = hns_roce_cmd_mbox(hr_dev, 0, mailbox->dma, mtpt_idx, 0,
@@ -390,12 +391,12 @@ int hns_roce_rereg_user_mr(struct ib_mr *ibmr, int flags, u64 start, u64 length,
 
 	hns_roce_free_cmd_mailbox(hr_dev, mailbox);
 
-	return 0;
+	return NULL;
 
 free_cmd_mbox:
 	hns_roce_free_cmd_mailbox(hr_dev, mailbox);
 
-	return ret;
+	return ERR_PTR(ret);
 }
 
 int hns_roce_dereg_mr(struct ib_mr *ibmr, struct ib_udata *udata)
diff --git a/drivers/infiniband/hw/mlx4/mlx4_ib.h b/drivers/infiniband/hw/mlx4/mlx4_ib.h
index 58df06492d69..78c9bb79ec75 100644
--- a/drivers/infiniband/hw/mlx4/mlx4_ib.h
+++ b/drivers/infiniband/hw/mlx4/mlx4_ib.h
@@ -908,10 +908,10 @@ int mlx4_ib_steer_qp_alloc(struct mlx4_ib_dev *dev, int count, int *qpn);
 void mlx4_ib_steer_qp_free(struct mlx4_ib_dev *dev, u32 qpn, int count);
 int mlx4_ib_steer_qp_reg(struct mlx4_ib_dev *mdev, struct mlx4_ib_qp *mqp,
 			 int is_attach);
-int mlx4_ib_rereg_user_mr(struct ib_mr *mr, int flags,
-			  u64 start, u64 length, u64 virt_addr,
-			  int mr_access_flags, struct ib_pd *pd,
-			  struct ib_udata *udata);
+struct ib_mr *mlx4_ib_rereg_user_mr(struct ib_mr *mr, int flags, u64 start,
+				    u64 length, u64 virt_addr,
+				    int mr_access_flags, struct ib_pd *pd,
+				    struct ib_udata *udata);
 int mlx4_ib_gid_index_to_real_index(struct mlx4_ib_dev *ibdev,
 				    const struct ib_gid_attr *attr);
 
diff --git a/drivers/infiniband/hw/mlx4/mr.c b/drivers/infiniband/hw/mlx4/mr.c
index 426fed005d53..50becc0e4b62 100644
--- a/drivers/infiniband/hw/mlx4/mr.c
+++ b/drivers/infiniband/hw/mlx4/mr.c
@@ -456,10 +456,10 @@ err_free:
 	return ERR_PTR(err);
 }
 
-int mlx4_ib_rereg_user_mr(struct ib_mr *mr, int flags,
-			  u64 start, u64 length, u64 virt_addr,
-			  int mr_access_flags, struct ib_pd *pd,
-			  struct ib_udata *udata)
+struct ib_mr *mlx4_ib_rereg_user_mr(struct ib_mr *mr, int flags, u64 start,
+				    u64 length, u64 virt_addr,
+				    int mr_access_flags, struct ib_pd *pd,
+				    struct ib_udata *udata)
 {
 	struct mlx4_ib_dev *dev = to_mdev(mr->device);
 	struct mlx4_ib_mr *mmr = to_mmr(mr);
@@ -472,9 +472,8 @@ int mlx4_ib_rereg_user_mr(struct ib_mr *mr, int flags,
 	 * race exists.
 	 */
 	err =  mlx4_mr_hw_get_mpt(dev->dev, &mmr->mmr, &pmpt_entry);
-
 	if (err)
-		return err;
+		return ERR_PTR(err);
 
 	if (flags & IB_MR_REREG_PD) {
 		err = mlx4_mr_hw_change_pd(dev->dev, *pmpt_entry,
@@ -542,8 +541,9 @@ int mlx4_ib_rereg_user_mr(struct ib_mr *mr, int flags,
 
 release_mpt_entry:
 	mlx4_mr_hw_put_mpt(dev->dev, pmpt_entry);
-
-	return err;
+	if (err)
+		return ERR_PTR(err);
+	return NULL;
 }
 
 static int
diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h
index 718e59fce006..ab84d4efbda3 100644
--- a/drivers/infiniband/hw/mlx5/mlx5_ib.h
+++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h
@@ -1254,9 +1254,9 @@ struct mlx5_ib_mr *mlx5_ib_alloc_implicit_mr(struct mlx5_ib_pd *pd,
 					     int access_flags);
 void mlx5_ib_free_implicit_mr(struct mlx5_ib_mr *mr);
 void mlx5_ib_fence_odp_mr(struct mlx5_ib_mr *mr);
-int mlx5_ib_rereg_user_mr(struct ib_mr *ib_mr, int flags, u64 start,
-			  u64 length, u64 virt_addr, int access_flags,
-			  struct ib_pd *pd, struct ib_udata *udata);
+struct ib_mr *mlx5_ib_rereg_user_mr(struct ib_mr *ib_mr, int flags, u64 start,
+				    u64 length, u64 virt_addr, int access_flags,
+				    struct ib_pd *pd, struct ib_udata *udata);
 int mlx5_ib_dereg_mr(struct ib_mr *ibmr, struct ib_udata *udata);
 struct ib_mr *mlx5_ib_alloc_mr(struct ib_pd *pd, enum ib_mr_type mr_type,
 			       u32 max_num_sg);
diff --git a/drivers/infiniband/hw/mlx5/mr.c b/drivers/infiniband/hw/mlx5/mr.c
index b6116f6d065d..778cc08e17ad 100644
--- a/drivers/infiniband/hw/mlx5/mr.c
+++ b/drivers/infiniband/hw/mlx5/mr.c
@@ -1620,9 +1620,10 @@ static int rereg_umr(struct ib_pd *pd, struct mlx5_ib_mr *mr,
 	return err;
 }
 
-int mlx5_ib_rereg_user_mr(struct ib_mr *ib_mr, int flags, u64 start,
-			  u64 length, u64 virt_addr, int new_access_flags,
-			  struct ib_pd *new_pd, struct ib_udata *udata)
+struct ib_mr *mlx5_ib_rereg_user_mr(struct ib_mr *ib_mr, int flags, u64 start,
+				    u64 length, u64 virt_addr,
+				    int new_access_flags, struct ib_pd *new_pd,
+				    struct ib_udata *udata)
 {
 	struct mlx5_ib_dev *dev = to_mdev(ib_mr->device);
 	struct mlx5_ib_mr *mr = to_mmr(ib_mr);
@@ -1638,10 +1639,10 @@ int mlx5_ib_rereg_user_mr(struct ib_mr *ib_mr, int flags, u64 start,
 		    start, virt_addr, length, access_flags);
 
 	if (!mr->umem)
-		return -EINVAL;
+		return ERR_PTR(-EINVAL);
 
 	if (is_odp_mr(mr))
-		return -EOPNOTSUPP;
+		return ERR_PTR(-EOPNOTSUPP);
 
 	if (flags & IB_MR_REREG_TRANS) {
 		addr = virt_addr;
@@ -1717,14 +1718,14 @@ int mlx5_ib_rereg_user_mr(struct ib_mr *ib_mr, int flags, u64 start,
 
 	set_mr_fields(dev, mr, len, access_flags);
 
-	return 0;
+	return NULL;
 
 err:
 	ib_umem_release(mr->umem);
 	mr->umem = NULL;
 
 	clean_mr(dev, mr);
-	return err;
+	return ERR_PTR(err);
 }
 
 static int