Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma

Pull rdma updates from Jason Gunthorpe:
 "This is significantly bug fixes and general cleanups. The noteworthy
  new features are fairly small:

   - XRC support for HNS and improves RQ operations

   - Bug fixes and updates for hns, mlx5, bnxt_re, hfi1, i40iw, rxe, siw
     and qib

   - Quite a few general cleanups on spelling, error handling, static
     checker detections, etc

   - Increase the number of device ports supported beyond 255. High port
     count software switches now exist

   - Several bug fixes for rtrs

   - mlx5 Device Memory support for host controlled atomics

   - Report SRQ tables through to rdma-tool"

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma: (145 commits)
  IB/qib: Remove redundant assignment to ret
  RDMA/nldev: Add copy-on-fork attribute to get sys command
  RDMA/bnxt_re: Fix a double free in bnxt_qplib_alloc_res
  RDMA/siw: Fix a use after free in siw_alloc_mr
  IB/hfi1: Remove redundant variable rcd
  RDMA/nldev: Add QP numbers to SRQ information
  RDMA/nldev: Return SRQ information
  RDMA/restrack: Add support to get resource tracking for SRQ
  RDMA/nldev: Return context information
  RDMA/core: Add CM to restrack after successful attachment to a device
  RDMA/cma: Skip device which doesn't support CM
  RDMA/rxe: Fix a bug in rxe_fill_ip_info()
  RDMA/mlx5: Expose private query port
  RDMA/mlx4: Remove an unused variable
  RDMA/mlx5: Fix type assignment for ICM DM
  IB/mlx5: Set right RoCE l3 type and roce version while deleting GID
  RDMA/i40iw: Fix error unwinding when i40iw_hmc_sd_one fails
  RDMA/cxgb4: add missing qpid increment
  IB/ipoib: Remove unnecessary struct declaration
  RDMA/bnxt_re: Get rid of custom module reference counting
  ...

20 files changed, 1198 insertions, 758 deletions

diff --git a/drivers/infiniband/hw/mlx5/Makefile b/drivers/infiniband/hw/mlx5/Makefile
index b4c009bb0db6..f43380106bd0 100644
--- a/drivers/infiniband/hw/mlx5/Makefile
+++ b/drivers/infiniband/hw/mlx5/Makefile
@@ -6,6 +6,7 @@ mlx5_ib-y := ah.o \
 	     cong.o \
 	     counters.o \
 	     cq.o \
+	     dm.o \
 	     doorbell.o \
 	     gsi.o \
 	     ib_virt.o \
diff --git a/drivers/infiniband/hw/mlx5/cmd.c b/drivers/infiniband/hw/mlx5/cmd.c
index 234f29912ba9..a8db8a051170 100644
--- a/drivers/infiniband/hw/mlx5/cmd.c
+++ b/drivers/infiniband/hw/mlx5/cmd.c
@@ -47,107 +47,6 @@ int mlx5_cmd_query_cong_params(struct mlx5_core_dev *dev, int cong_point,
 	return mlx5_cmd_exec_inout(dev, query_cong_params, in, out);
 }
 
-int mlx5_cmd_alloc_memic(struct mlx5_dm *dm, phys_addr_t *addr,
-			 u64 length, u32 alignment)
-{
-	struct mlx5_core_dev *dev = dm->dev;
-	u64 num_memic_hw_pages = MLX5_CAP_DEV_MEM(dev, memic_bar_size)
-					>> PAGE_SHIFT;
-	u64 hw_start_addr = MLX5_CAP64_DEV_MEM(dev, memic_bar_start_addr);
-	u32 max_alignment = MLX5_CAP_DEV_MEM(dev, log_max_memic_addr_alignment);
-	u32 num_pages = DIV_ROUND_UP(length, PAGE_SIZE);
-	u32 out[MLX5_ST_SZ_DW(alloc_memic_out)] = {};
-	u32 in[MLX5_ST_SZ_DW(alloc_memic_in)] = {};
-	u32 mlx5_alignment;
-	u64 page_idx = 0;
-	int ret = 0;
-
-	if (!length || (length & MLX5_MEMIC_ALLOC_SIZE_MASK))
-		return -EINVAL;
-
-	/* mlx5 device sets alignment as 64*2^driver_value
-	 * so normalizing is needed.
-	 */
-	mlx5_alignment = (alignment < MLX5_MEMIC_BASE_ALIGN) ? 0 :
-			 alignment - MLX5_MEMIC_BASE_ALIGN;
-	if (mlx5_alignment > max_alignment)
-		return -EINVAL;
-
-	MLX5_SET(alloc_memic_in, in, opcode, MLX5_CMD_OP_ALLOC_MEMIC);
-	MLX5_SET(alloc_memic_in, in, range_size, num_pages * PAGE_SIZE);
-	MLX5_SET(alloc_memic_in, in, memic_size, length);
-	MLX5_SET(alloc_memic_in, in, log_memic_addr_alignment,
-		 mlx5_alignment);
-
-	while (page_idx < num_memic_hw_pages) {
-		spin_lock(&dm->lock);
-		page_idx = bitmap_find_next_zero_area(dm->memic_alloc_pages,
-						      num_memic_hw_pages,
-						      page_idx,
-						      num_pages, 0);
-
-		if (page_idx < num_memic_hw_pages)
-			bitmap_set(dm->memic_alloc_pages,
-				   page_idx, num_pages);
-
-		spin_unlock(&dm->lock);
-
-		if (page_idx >= num_memic_hw_pages)
-			break;
-
-		MLX5_SET64(alloc_memic_in, in, range_start_addr,
-			   hw_start_addr + (page_idx * PAGE_SIZE));
-
-		ret = mlx5_cmd_exec_inout(dev, alloc_memic, in, out);
-		if (ret) {
-			spin_lock(&dm->lock);
-			bitmap_clear(dm->memic_alloc_pages,
-				     page_idx, num_pages);
-			spin_unlock(&dm->lock);
-
-			if (ret == -EAGAIN) {
-				page_idx++;
-				continue;
-			}
-
-			return ret;
-		}
-
-		*addr = dev->bar_addr +
-			MLX5_GET64(alloc_memic_out, out, memic_start_addr);
-
-		return 0;
-	}
-
-	return -ENOMEM;
-}
-
-void mlx5_cmd_dealloc_memic(struct mlx5_dm *dm, phys_addr_t addr, u64 length)
-{
-	struct mlx5_core_dev *dev = dm->dev;
-	u64 hw_start_addr = MLX5_CAP64_DEV_MEM(dev, memic_bar_start_addr);
-	u32 num_pages = DIV_ROUND_UP(length, PAGE_SIZE);
-	u32 in[MLX5_ST_SZ_DW(dealloc_memic_in)] = {};
-	u64 start_page_idx;
-	int err;
-
-	addr -= dev->bar_addr;
-	start_page_idx = (addr - hw_start_addr) >> PAGE_SHIFT;
-
-	MLX5_SET(dealloc_memic_in, in, opcode, MLX5_CMD_OP_DEALLOC_MEMIC);
-	MLX5_SET64(dealloc_memic_in, in, memic_start_addr, addr);
-	MLX5_SET(dealloc_memic_in, in, memic_size, length);
-
-	err =  mlx5_cmd_exec_in(dev, dealloc_memic, in);
-	if (err)
-		return;
-
-	spin_lock(&dm->lock);
-	bitmap_clear(dm->memic_alloc_pages,
-		     start_page_idx, num_pages);
-	spin_unlock(&dm->lock);
-}
-
 void mlx5_cmd_destroy_tir(struct mlx5_core_dev *dev, u32 tirn, u16 uid)
 {
 	u32 in[MLX5_ST_SZ_DW(destroy_tir_in)] = {};
diff --git a/drivers/infiniband/hw/mlx5/cmd.h b/drivers/infiniband/hw/mlx5/cmd.h
index 88ea6ef8f2cb..66c96292ed43 100644
--- a/drivers/infiniband/hw/mlx5/cmd.h
+++ b/drivers/infiniband/hw/mlx5/cmd.h
@@ -41,9 +41,6 @@ int mlx5_cmd_dump_fill_mkey(struct mlx5_core_dev *dev, u32 *mkey);
 int mlx5_cmd_null_mkey(struct mlx5_core_dev *dev, u32 *null_mkey);
 int mlx5_cmd_query_cong_params(struct mlx5_core_dev *dev, int cong_point,
 			       void *out);
-int mlx5_cmd_alloc_memic(struct mlx5_dm *dm, phys_addr_t *addr,
-			 u64 length, u32 alignment);
-void mlx5_cmd_dealloc_memic(struct mlx5_dm *dm, phys_addr_t addr, u64 length);
 int mlx5_cmd_dealloc_pd(struct mlx5_core_dev *dev, u32 pdn, u16 uid);
 void mlx5_cmd_destroy_tir(struct mlx5_core_dev *dev, u32 tirn, u16 uid);
 void mlx5_cmd_destroy_tis(struct mlx5_core_dev *dev, u32 tisn, u16 uid);
diff --git a/drivers/infiniband/hw/mlx5/cong.c b/drivers/infiniband/hw/mlx5/cong.c
index b9291e482428..0b61df52332a 100644
--- a/drivers/infiniband/hw/mlx5/cong.c
+++ b/drivers/infiniband/hw/mlx5/cong.c
@@ -267,7 +267,7 @@ static void mlx5_ib_set_cc_param_mask_val(void *field, int offset,
 	}
 }
 
-static int mlx5_ib_get_cc_params(struct mlx5_ib_dev *dev, u8 port_num,
+static int mlx5_ib_get_cc_params(struct mlx5_ib_dev *dev, u32 port_num,
 				 int offset, u32 *var)
 {
 	int outlen = MLX5_ST_SZ_BYTES(query_cong_params_out);
@@ -304,7 +304,7 @@ alloc_err:
 	return err;
 }
 
-static int mlx5_ib_set_cc_params(struct mlx5_ib_dev *dev, u8 port_num,
+static int mlx5_ib_set_cc_params(struct mlx5_ib_dev *dev, u32 port_num,
 				 int offset, u32 var)
 {
 	int inlen = MLX5_ST_SZ_BYTES(modify_cong_params_in);
@@ -397,7 +397,7 @@ static const struct file_operations dbg_cc_fops = {
 	.read	= get_param,
 };
 
-void mlx5_ib_cleanup_cong_debugfs(struct mlx5_ib_dev *dev, u8 port_num)
+void mlx5_ib_cleanup_cong_debugfs(struct mlx5_ib_dev *dev, u32 port_num)
 {
 	if (!mlx5_debugfs_root ||
 	    !dev->port[port_num].dbg_cc_params ||
@@ -409,7 +409,7 @@ void mlx5_ib_cleanup_cong_debugfs(struct mlx5_ib_dev *dev, u8 port_num)
 	dev->port[port_num].dbg_cc_params = NULL;
 }
 
-void mlx5_ib_init_cong_debugfs(struct mlx5_ib_dev *dev, u8 port_num)
+void mlx5_ib_init_cong_debugfs(struct mlx5_ib_dev *dev, u32 port_num)
 {
 	struct mlx5_ib_dbg_cc_params *dbg_cc_params;
 	struct mlx5_core_dev *mdev;
diff --git a/drivers/infiniband/hw/mlx5/counters.c b/drivers/infiniband/hw/mlx5/counters.c
index 084652e2b15a..e365341057cb 100644
--- a/drivers/infiniband/hw/mlx5/counters.c
+++ b/drivers/infiniband/hw/mlx5/counters.c
@@ -139,7 +139,7 @@ static int mlx5_ib_create_counters(struct ib_counters *counters,
 
 
 static const struct mlx5_ib_counters *get_counters(struct mlx5_ib_dev *dev,
-						   u8 port_num)
+						   u32 port_num)
 {
 	return is_mdev_switchdev_mode(dev->mdev) ? &dev->port[0].cnts :
 						   &dev->port[port_num].cnts;
@@ -154,7 +154,7 @@ static const struct mlx5_ib_counters *get_counters(struct mlx5_ib_dev *dev,
  * device port combination in switchdev and non switchdev mode of the
  * parent device.
  */
-u16 mlx5_ib_get_counters_id(struct mlx5_ib_dev *dev, u8 port_num)
+u16 mlx5_ib_get_counters_id(struct mlx5_ib_dev *dev, u32 port_num)
 {
 	const struct mlx5_ib_counters *cnts = get_counters(dev, port_num);
 
@@ -162,7 +162,7 @@ u16 mlx5_ib_get_counters_id(struct mlx5_ib_dev *dev, u8 port_num)
 }
 
 static struct rdma_hw_stats *mlx5_ib_alloc_hw_stats(struct ib_device *ibdev,
-						    u8 port_num)
+						    u32 port_num)
 {
 	struct mlx5_ib_dev *dev = to_mdev(ibdev);
 	const struct mlx5_ib_counters *cnts;
@@ -236,13 +236,13 @@ free:
 
 static int mlx5_ib_get_hw_stats(struct ib_device *ibdev,
 				struct rdma_hw_stats *stats,
-				u8 port_num, int index)
+				u32 port_num, int index)
 {
 	struct mlx5_ib_dev *dev = to_mdev(ibdev);
 	const struct mlx5_ib_counters *cnts = get_counters(dev, port_num - 1);
 	struct mlx5_core_dev *mdev;
 	int ret, num_counters;
-	u8 mdev_port_num;
+	u32 mdev_port_num;
 
 	if (!stats)
 		return -EINVAL;
diff --git a/drivers/infiniband/hw/mlx5/counters.h b/drivers/infiniband/hw/mlx5/counters.h
index 1aa30c2f3f4d..6bcaaa52e2b2 100644
--- a/drivers/infiniband/hw/mlx5/counters.h
+++ b/drivers/infiniband/hw/mlx5/counters.h
@@ -13,5 +13,5 @@ void mlx5_ib_counters_cleanup(struct mlx5_ib_dev *dev);
 void mlx5_ib_counters_clear_description(struct ib_counters *counters);
 int mlx5_ib_flow_counters_set_data(struct ib_counters *ibcounters,
 				   struct mlx5_ib_create_flow *ucmd);
-u16 mlx5_ib_get_counters_id(struct mlx5_ib_dev *dev, u8 port_num);
+u16 mlx5_ib_get_counters_id(struct mlx5_ib_dev *dev, u32 port_num);
 #endif /* _MLX5_IB_COUNTERS_H */
diff --git a/drivers/infiniband/hw/mlx5/devx.c b/drivers/infiniband/hw/mlx5/devx.c
index 07b8350929cd..a0b677accd96 100644
--- a/drivers/infiniband/hw/mlx5/devx.c
+++ b/drivers/infiniband/hw/mlx5/devx.c
@@ -2185,27 +2185,69 @@ static int devx_umem_get(struct mlx5_ib_dev *dev, struct ib_ucontext *ucontext,
 	return 0;
 }
 
+static unsigned int devx_umem_find_best_pgsize(struct ib_umem *umem,
+					       unsigned long pgsz_bitmap)
+{
+	unsigned long page_size;
+
+	/* Don't bother checking larger page sizes as offset must be zero and
+	 * total DEVX umem length must be equal to total umem length.
+	 */
+	pgsz_bitmap &= GENMASK_ULL(max_t(u64, order_base_2(umem->length),
+					 PAGE_SHIFT),
+				   MLX5_ADAPTER_PAGE_SHIFT);
+	if (!pgsz_bitmap)
+		return 0;
+
+	page_size = ib_umem_find_best_pgoff(umem, pgsz_bitmap, U64_MAX);
+	if (!page_size)
+		return 0;
+
+	/* If the page_size is less than the CPU page size then we can use the
+	 * offset and create a umem which is a subset of the page list.
+	 * For larger page sizes we can't be sure the DMA  list reflects the
+	 * VA so we must ensure that the umem extent is exactly equal to the
+	 * page list. Reduce the page size until one of these cases is true.
+	 */
+	while ((ib_umem_dma_offset(umem, page_size) != 0 ||
+		(umem->length % page_size) != 0) &&
+		page_size > PAGE_SIZE)
+		page_size /= 2;
+
+	return page_size;
+}
+
 static int devx_umem_reg_cmd_alloc(struct mlx5_ib_dev *dev,
 				   struct uverbs_attr_bundle *attrs,
 				   struct devx_umem *obj,
 				   struct devx_umem_reg_cmd *cmd)
 {
+	unsigned long pgsz_bitmap;
 	unsigned int page_size;
 	__be64 *mtt;
 	void *umem;
+	int ret;
 
 	/*
-	 * We don't know what the user intends to use this umem for, but the HW
-	 * restrictions must be met. MR, doorbell records, QP, WQ and CQ all
-	 * have different requirements. Since we have no idea how to sort this
-	 * out, only support PAGE_SIZE with the expectation that userspace will
-	 * provide the necessary alignments inside the known PAGE_SIZE and that
-	 * FW will check everything.
+	 * If the user does not pass in pgsz_bitmap then the user promises not
+	 * to use umem_offset!=0 in any commands that allocate on top of the
+	 * umem.
+	 *
+	 * If the user wants to use a umem_offset then it must pass in
+	 * pgsz_bitmap which guides the maximum page size and thus maximum
+	 * object alignment inside the umem. See the PRM.
+	 *
+	 * Users are not allowed to use IOVA here, mkeys are not supported on
+	 * umem.
 	 */
-	page_size = ib_umem_find_best_pgoff(
-		obj->umem, PAGE_SIZE,
-		__mlx5_page_offset_to_bitmask(__mlx5_bit_sz(umem, page_offset),
-					      0));
+	ret = uverbs_get_const_default(&pgsz_bitmap, attrs,
+			MLX5_IB_ATTR_DEVX_UMEM_REG_PGSZ_BITMAP,
+			GENMASK_ULL(63,
+				    min(PAGE_SHIFT, MLX5_ADAPTER_PAGE_SHIFT)));
+	if (ret)
+		return ret;
+
+	page_size = devx_umem_find_best_pgsize(obj->umem, pgsz_bitmap);
 	if (!page_size)
 		return -EINVAL;
 
@@ -2791,6 +2833,8 @@ DECLARE_UVERBS_NAMED_METHOD(
 			   UA_MANDATORY),
 	UVERBS_ATTR_FLAGS_IN(MLX5_IB_ATTR_DEVX_UMEM_REG_ACCESS,
 			     enum ib_access_flags),
+	UVERBS_ATTR_CONST_IN(MLX5_IB_ATTR_DEVX_UMEM_REG_PGSZ_BITMAP,
+			     u64),
 	UVERBS_ATTR_PTR_OUT(MLX5_IB_ATTR_DEVX_UMEM_REG_OUT_ID,
 			    UVERBS_ATTR_TYPE(u32),
 			    UA_MANDATORY));
diff --git a/drivers/infiniband/hw/mlx5/dm.c b/drivers/infiniband/hw/mlx5/dm.c
new file mode 100644
index 000000000000..094bf85589db
--- /dev/null
+++ b/drivers/infiniband/hw/mlx5/dm.c
@@ -0,0 +1,587 @@
+// SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB
+/*
+ * Copyright (c) 2021, Mellanox Technologies inc. All rights reserved.
+ */
+
+#include <rdma/uverbs_std_types.h>
+#include "dm.h"
+
+#define UVERBS_MODULE_NAME mlx5_ib
+#include <rdma/uverbs_named_ioctl.h>
+
+static int mlx5_cmd_alloc_memic(struct mlx5_dm *dm, phys_addr_t *addr,
+				u64 length, u32 alignment)
+{
+	struct mlx5_core_dev *dev = dm->dev;
+	u64 num_memic_hw_pages = MLX5_CAP_DEV_MEM(dev, memic_bar_size)
+					>> PAGE_SHIFT;
+	u64 hw_start_addr = MLX5_CAP64_DEV_MEM(dev, memic_bar_start_addr);
+	u32 max_alignment = MLX5_CAP_DEV_MEM(dev, log_max_memic_addr_alignment);
+	u32 num_pages = DIV_ROUND_UP(length, PAGE_SIZE);
+	u32 out[MLX5_ST_SZ_DW(alloc_memic_out)] = {};
+	u32 in[MLX5_ST_SZ_DW(alloc_memic_in)] = {};
+	u32 mlx5_alignment;
+	u64 page_idx = 0;
+	int ret = 0;
+
+	if (!length || (length & MLX5_MEMIC_ALLOC_SIZE_MASK))
+		return -EINVAL;
+
+	/* mlx5 device sets alignment as 64*2^driver_value
+	 * so normalizing is needed.
+	 */
+	mlx5_alignment = (alignment < MLX5_MEMIC_BASE_ALIGN) ? 0 :
+			 alignment - MLX5_MEMIC_BASE_ALIGN;
+	if (mlx5_alignment > max_alignment)
+		return -EINVAL;
+
+	MLX5_SET(alloc_memic_in, in, opcode, MLX5_CMD_OP_ALLOC_MEMIC);
+	MLX5_SET(alloc_memic_in, in, range_size, num_pages * PAGE_SIZE);
+	MLX5_SET(alloc_memic_in, in, memic_size, length);
+	MLX5_SET(alloc_memic_in, in, log_memic_addr_alignment,
+		 mlx5_alignment);
+
+	while (page_idx < num_memic_hw_pages) {
+		spin_lock(&dm->lock);
+		page_idx = bitmap_find_next_zero_area(dm->memic_alloc_pages,
+						      num_memic_hw_pages,
+						      page_idx,
+						      num_pages, 0);
+
+		if (page_idx < num_memic_hw_pages)
+			bitmap_set(dm->memic_alloc_pages,
+				   page_idx, num_pages);
+
+		spin_unlock(&dm->lock);
+
+		if (page_idx >= num_memic_hw_pages)
+			break;
+
+		MLX5_SET64(alloc_memic_in, in, range_start_addr,
+			   hw_start_addr + (page_idx * PAGE_SIZE));
+
+		ret = mlx5_cmd_exec_inout(dev, alloc_memic, in, out);
+		if (ret) {
+			spin_lock(&dm->lock);
+			bitmap_clear(dm->memic_alloc_pages,
+				     page_idx, num_pages);
+			spin_unlock(&dm->lock);
+
+			if (ret == -EAGAIN) {
+				page_idx++;
+				continue;
+			}
+
+			return ret;
+		}
+
+		*addr = dev->bar_addr +
+			MLX5_GET64(alloc_memic_out, out, memic_start_addr);
+
+		return 0;
+	}
+
+	return -ENOMEM;
+}
+
+void mlx5_cmd_dealloc_memic(struct mlx5_dm *dm, phys_addr_t addr,
+			    u64 length)
+{
+	struct mlx5_core_dev *dev = dm->dev;
+	u64 hw_start_addr = MLX5_CAP64_DEV_MEM(dev, memic_bar_start_addr);
+	u32 num_pages = DIV_ROUND_UP(length, PAGE_SIZE);
+	u32 in[MLX5_ST_SZ_DW(dealloc_memic_in)] = {};
+	u64 start_page_idx;
+	int err;
+
+	addr -= dev->bar_addr;
+	start_page_idx = (addr - hw_start_addr) >> PAGE_SHIFT;
+
+	MLX5_SET(dealloc_memic_in, in, opcode, MLX5_CMD_OP_DEALLOC_MEMIC);
+	MLX5_SET64(dealloc_memic_in, in, memic_start_addr, addr);
+	MLX5_SET(dealloc_memic_in, in, memic_size, length);
+
+	err =  mlx5_cmd_exec_in(dev, dealloc_memic, in);
+	if (err)
+		return;
+
+	spin_lock(&dm->lock);
+	bitmap_clear(dm->memic_alloc_pages,
+		     start_page_idx, num_pages);
+	spin_unlock(&dm->lock);
+}
+
+void mlx5_cmd_dealloc_memic_op(struct mlx5_dm *dm, phys_addr_t addr,
+			       u8 operation)
+{
+	u32 in[MLX5_ST_SZ_DW(modify_memic_in)] = {};
+	struct mlx5_core_dev *dev = dm->dev;
+
+	MLX5_SET(modify_memic_in, in, opcode, MLX5_CMD_OP_MODIFY_MEMIC);
+	MLX5_SET(modify_memic_in, in, op_mod, MLX5_MODIFY_MEMIC_OP_MOD_DEALLOC);
+	MLX5_SET(modify_memic_in, in, memic_operation_type, operation);
+	MLX5_SET64(modify_memic_in, in, memic_start_addr, addr - dev->bar_addr);
+
+	mlx5_cmd_exec_in(dev, modify_memic, in);
+}
+
+static int mlx5_cmd_alloc_memic_op(struct mlx5_dm *dm, phys_addr_t addr,
+				   u8 operation, phys_addr_t *op_addr)
+{
+	u32 out[MLX5_ST_SZ_DW(modify_memic_out)] = {};
+	u32 in[MLX5_ST_SZ_DW(modify_memic_in)] = {};
+	struct mlx5_core_dev *dev = dm->dev;
+	int err;
+
+	MLX5_SET(modify_memic_in, in, opcode, MLX5_CMD_OP_MODIFY_MEMIC);
+	MLX5_SET(modify_memic_in, in, op_mod, MLX5_MODIFY_MEMIC_OP_MOD_ALLOC);
+	MLX5_SET(modify_memic_in, in, memic_operation_type, operation);
+	MLX5_SET64(modify_memic_in, in, memic_start_addr, addr - dev->bar_addr);
+
+	err = mlx5_cmd_exec_inout(dev, modify_memic, in, out);
+	if (err)
+		return err;
+
+	*op_addr = dev->bar_addr +
+		   MLX5_GET64(modify_memic_out, out, memic_operation_addr);
+	return 0;
+}
+
+static int add_dm_mmap_entry(struct ib_ucontext *context,
+			     struct mlx5_user_mmap_entry *mentry, u8 mmap_flag,
+			     size_t size, u64 address)
+{
+	mentry->mmap_flag = mmap_flag;
+	mentry->address = address;
+
+	return rdma_user_mmap_entry_insert_range(
+		context, &mentry->rdma_entry, size,
+		MLX5_IB_MMAP_DEVICE_MEM << 16,
+		(MLX5_IB_MMAP_DEVICE_MEM << 16) + (1UL << 16) - 1);
+}
+
+static void mlx5_ib_dm_memic_free(struct kref *kref)
+{
+	struct mlx5_ib_dm_memic *dm =
+		container_of(kref, struct mlx5_ib_dm_memic, ref);
+	struct mlx5_ib_dev *dev = to_mdev(dm->base.ibdm.device);
+
+	mlx5_cmd_dealloc_memic(&dev->dm, dm->base.dev_addr, dm->base.size);
+	kfree(dm);
+}
+
+static int copy_op_to_user(struct mlx5_ib_dm_op_entry *op_entry,
+			   struct uverbs_attr_bundle *attrs)
+{
+	u64 start_offset;
+	u16 page_idx;
+	int err;
+
+	page_idx = op_entry->mentry.rdma_entry.start_pgoff & 0xFFFF;
+	start_offset = op_entry->op_addr & ~PAGE_MASK;
+	err = uverbs_copy_to(attrs, MLX5_IB_ATTR_DM_MAP_OP_ADDR_RESP_PAGE_INDEX,
+			     &page_idx, sizeof(page_idx));
+	if (err)
+		return err;
+
+	return uverbs_copy_to(attrs,
+			      MLX5_IB_ATTR_DM_MAP_OP_ADDR_RESP_START_OFFSET,
+			      &start_offset, sizeof(start_offset));
+}
+
+static int map_existing_op(struct mlx5_ib_dm_memic *dm, u8 op,
+			   struct uverbs_attr_bundle *attrs)
+{
+	struct mlx5_ib_dm_op_entry *op_entry;
+
+	op_entry = xa_load(&dm->ops, op);
+	if (!op_entry)
+		return -ENOENT;
+
+	return copy_op_to_user(op_entry, attrs);
+}
+
+static int UVERBS_HANDLER(MLX5_IB_METHOD_DM_MAP_OP_ADDR)(
+	struct uverbs_attr_bundle *attrs)
+{
+	struct ib_uobject *uobj = uverbs_attr_get_uobject(
+		attrs, MLX5_IB_ATTR_DM_MAP_OP_ADDR_REQ_HANDLE);
+	struct mlx5_ib_dev *dev = to_mdev(uobj->context->device);
+	struct ib_dm *ibdm = uobj->object;
+	struct mlx5_ib_dm_memic *dm = to_memic(ibdm);
+	struct mlx5_ib_dm_op_entry *op_entry;
+	int err;
+	u8 op;
+
+	err = uverbs_copy_from(&op, attrs, MLX5_IB_ATTR_DM_MAP_OP_ADDR_REQ_OP);
+	if (err)
+		return err;
+
+	if (!(MLX5_CAP_DEV_MEM(dev->mdev, memic_operations) & BIT(op)))
+		return -EOPNOTSUPP;
+
+	mutex_lock(&dm->ops_xa_lock);
+	err = map_existing_op(dm, op, attrs);
+	if (!err || err != -ENOENT)
+		goto err_unlock;
+
+	op_entry = kzalloc(sizeof(*op_entry), GFP_KERNEL);
+	if (!op_entry)
+		goto err_unlock;
+
+	err = mlx5_cmd_alloc_memic_op(&dev->dm, dm->base.dev_addr, op,
+				      &op_entry->op_addr);
+	if (err) {
+		kfree(op_entry);
+		goto err_unlock;
+	}
+	op_entry->op = op;
+	op_entry->dm = dm;
+
+	err = add_dm_mmap_entry(uobj->context, &op_entry->mentry,
+				MLX5_IB_MMAP_TYPE_MEMIC_OP, dm->base.size,
+				op_entry->op_addr & PAGE_MASK);
+	if (err) {
+		mlx5_cmd_dealloc_memic_op(&dev->dm, dm->base.dev_addr, op);
+		kfree(op_entry);
+		goto err_unlock;
+	}
+	/* From this point, entry will be freed by mmap_free */
+	kref_get(&dm->ref);
+
+	err = copy_op_to_user(op_entry, attrs);
+	if (err)
+		goto err_remove;
+
+	err = xa_insert(&dm->ops, op, op_entry, GFP_KERNEL);
+	if (err)
+		goto err_remove;
+	mutex_unlock(&dm->ops_xa_lock);
+
+	return 0;
+
+err_remove:
+	rdma_user_mmap_entry_remove(&op_entry->mentry.rdma_entry);
+err_unlock:
+	mutex_unlock(&dm->ops_xa_lock);
+
+	return err;
+}
+
+static struct ib_dm *handle_alloc_dm_memic(struct ib_ucontext *ctx,
+					   struct ib_dm_alloc_attr *attr,
+					   struct uverbs_attr_bundle *attrs)
+{
+	struct mlx5_dm *dm_db = &to_mdev(ctx->device)->dm;
+	struct mlx5_ib_dm_memic *dm;
+	u64 start_offset;
+	u16 page_idx;
+	int err;
+	u64 address;
+
+	if (!MLX5_CAP_DEV_MEM(dm_db->dev, memic))
+		return ERR_PTR(-EOPNOTSUPP);
+
+	dm = kzalloc(sizeof(*dm), GFP_KERNEL);
+	if (!dm)
+		return ERR_PTR(-ENOMEM);
+
+	dm->base.type = MLX5_IB_UAPI_DM_TYPE_MEMIC;
+	dm->base.size = roundup(attr->length, MLX5_MEMIC_BASE_SIZE);
+	dm->base.ibdm.device = ctx->device;
+
+	kref_init(&dm->ref);
+	xa_init(&dm->ops);
+	mutex_init(&dm->ops_xa_lock);
+	dm->req_length = attr->length;
+
+	err = mlx5_cmd_alloc_memic(dm_db, &dm->base.dev_addr,
+				   dm->base.size, attr->alignment);
+	if (err) {
+		kfree(dm);
+		return ERR_PTR(err);
+	}
+
+	address = dm->base.dev_addr & PAGE_MASK;
+	err = add_dm_mmap_entry(ctx, &dm->mentry, MLX5_IB_MMAP_TYPE_MEMIC,
+				dm->base.size, address);
+	if (err) {
+		mlx5_cmd_dealloc_memic(dm_db, dm->base.dev_addr, dm->base.size);
+		kfree(dm);
+		return ERR_PTR(err);
+	}
+
+	page_idx = dm->mentry.rdma_entry.start_pgoff & 0xFFFF;
+	err = uverbs_copy_to(attrs, MLX5_IB_ATTR_ALLOC_DM_RESP_PAGE_INDEX,
+			     &page_idx, sizeof(page_idx));
+	if (err)
+		goto err_copy;
+
+	start_offset = dm->base.dev_addr & ~PAGE_MASK;
+	err = uverbs_copy_to(attrs,
+			     MLX5_IB_ATTR_ALLOC_DM_RESP_START_OFFSET,
+			     &start_offset, sizeof(start_offset));
+	if (err)
+		goto err_copy;
+
+	return &dm->base.ibdm;
+
+err_copy:
+	rdma_user_mmap_entry_remove(&dm->mentry.rdma_entry);
+	return ERR_PTR(err);
+}
+
+static enum mlx5_sw_icm_type get_icm_type(int uapi_type)
+{
+	return uapi_type == MLX5_IB_UAPI_DM_TYPE_STEERING_SW_ICM ?
+		       MLX5_SW_ICM_TYPE_STEERING :
+		       MLX5_SW_ICM_TYPE_HEADER_MODIFY;
+}
+
+static struct ib_dm *handle_alloc_dm_sw_icm(struct ib_ucontext *ctx,
+					    struct ib_dm_alloc_attr *attr,
+					    struct uverbs_attr_bundle *attrs,
+					    int type)
+{
+	struct mlx5_core_dev *dev = to_mdev(ctx->device)->mdev;
+	enum mlx5_sw_icm_type icm_type = get_icm_type(type);
+	struct mlx5_ib_dm_icm *dm;
+	u64 act_size;
+	int err;
+
+	dm = kzalloc(sizeof(*dm), GFP_KERNEL);
+	if (!dm)
+		return ERR_PTR(-ENOMEM);
+
+	dm->base.type = type;
+	dm->base.ibdm.device = ctx->device;
+
+	if (!capable(CAP_SYS_RAWIO) || !capable(CAP_NET_RAW)) {
+		err = -EPERM;
+		goto free;
+	}
+
+	if (!(MLX5_CAP_FLOWTABLE_NIC_RX(dev, sw_owner) ||
+	      MLX5_CAP_FLOWTABLE_NIC_TX(dev, sw_owner) ||
+	      MLX5_CAP_FLOWTABLE_NIC_RX(dev, sw_owner_v2) ||
+	      MLX5_CAP_FLOWTABLE_NIC_TX(dev, sw_owner_v2))) {
+		err = -EOPNOTSUPP;
+		goto free;
+	}
+
+	/* Allocation size must a multiple of the basic block size
+	 * and a power of 2.
+	 */
+	act_size = round_up(attr->length, MLX5_SW_ICM_BLOCK_SIZE(dev));
+	act_size = roundup_pow_of_two(act_size);
+
+	dm->base.size = act_size;
+	err = mlx5_dm_sw_icm_alloc(dev, icm_type, act_size, attr->alignment,
+				   to_mucontext(ctx)->devx_uid,
+				   &dm->base.dev_addr, &dm->obj_id);
+	if (err)
+		goto free;
+
+	err = uverbs_copy_to(attrs, MLX5_IB_ATTR_ALLOC_DM_RESP_START_OFFSET,
+			     &dm->base.dev_addr, sizeof(dm->base.dev_addr));
+	if (err) {
+		mlx5_dm_sw_icm_dealloc(dev, icm_type, dm->base.size,
+				       to_mucontext(ctx)->devx_uid,
+				       dm->base.dev_addr, dm->obj_id);
+		goto free;
+	}
+	return &dm->base.ibdm;
+free:
+	kfree(dm);
+	return ERR_PTR(err);
+}
+
+struct ib_dm *mlx5_ib_alloc_dm(struct ib_device *ibdev,
+			       struct ib_ucontext *context,
+			       struct ib_dm_alloc_attr *attr,
+			       struct uverbs_attr_bundle *attrs)
+{
+	enum mlx5_ib_uapi_dm_type type;
+	int err;
+
+	err = uverbs_get_const_default(&type, attrs,
+				       MLX5_IB_ATTR_ALLOC_DM_REQ_TYPE,
+				       MLX5_IB_UAPI_DM_TYPE_MEMIC);
+	if (err)
+		return ERR_PTR(err);
+
+	mlx5_ib_dbg(to_mdev(ibdev), "alloc_dm req: dm_type=%d user_length=0x%llx log_alignment=%d\n",
+		    type, attr->length, attr->alignment);
+
+	switch (type) {
+	case MLX5_IB_UAPI_DM_TYPE_MEMIC:
+		return handle_alloc_dm_memic(context, attr, attrs);
+	case MLX5_IB_UAPI_DM_TYPE_STEERING_SW_ICM:
+		return handle_alloc_dm_sw_icm(context, attr, attrs, type);
+	case MLX5_IB_UAPI_DM_TYPE_HEADER_MODIFY_SW_ICM:
+		return handle_alloc_dm_sw_icm(context, attr, attrs, type);
+	default:
+		return ERR_PTR(-EOPNOTSUPP);
+	}
+}
+
+static void dm_memic_remove_ops(struct mlx5_ib_dm_memic *dm)
+{
+	struct mlx5_ib_dm_op_entry *entry;
+	unsigned long idx;
+
+	mutex_lock(&dm->ops_xa_lock);
+	xa_for_each(&dm->ops, idx, entry) {
+		xa_erase(&dm->ops, idx);
+		rdma_user_mmap_entry_remove(&entry->mentry.rdma_entry);
+	}
+	mutex_unlock(&dm->ops_xa_lock);
+}
+
+static void mlx5_dm_memic_dealloc(struct mlx5_ib_dm_memic *dm)
+{
+	dm_memic_remove_ops(dm);
+	rdma_user_mmap_entry_remove(&dm->mentry.rdma_entry);
+}
+
+static int mlx5_dm_icm_dealloc(struct mlx5_ib_ucontext *ctx,
+			       struct mlx5_ib_dm_icm *dm)
+{
+	enum mlx5_sw_icm_type type = get_icm_type(dm->base.type);
+	struct mlx5_core_dev *dev = to_mdev(dm->base.ibdm.device)->mdev;
+	int err;
+
+	err = mlx5_dm_sw_icm_dealloc(dev, type, dm->base.size, ctx->devx_uid,
+				     dm->base.dev_addr, dm->obj_id);
+	if (!err)
+		kfree(dm);
+	return 0;
+}
+
+static int mlx5_ib_dealloc_dm(struct ib_dm *ibdm,
+			      struct uverbs_attr_bundle *attrs)
+{
+	struct mlx5_ib_ucontext *ctx = rdma_udata_to_drv_context(
+		&attrs->driver_udata, struct mlx5_ib_ucontext, ibucontext);
+	struct mlx5_ib_dm *dm = to_mdm(ibdm);
+
+	switch (dm->type) {
+	case MLX5_IB_UAPI_DM_TYPE_MEMIC:
+		mlx5_dm_memic_dealloc(to_memic(ibdm));
+		return 0;
+	case MLX5_IB_UAPI_DM_TYPE_STEERING_SW_ICM:
+	case MLX5_IB_UAPI_DM_TYPE_HEADER_MODIFY_SW_ICM:
+		return mlx5_dm_icm_dealloc(ctx, to_icm(ibdm));
+	default:
+		return -EOPNOTSUPP;
+	}
+}
+
+static int UVERBS_HANDLER(MLX5_IB_METHOD_DM_QUERY)(
+	struct uverbs_attr_bundle *attrs)
+{
+	struct ib_dm *ibdm =
+		uverbs_attr_get_obj(attrs, MLX5_IB_ATTR_QUERY_DM_REQ_HANDLE);
+	struct mlx5_ib_dm *dm = to_mdm(ibdm);
+	struct mlx5_ib_dm_memic *memic;
+	u64 start_offset;
+	u16 page_idx;
+	int err;
+
+	if (dm->type != MLX5_IB_UAPI_DM_TYPE_MEMIC)
+		return -EOPNOTSUPP;
+
+	memic = to_memic(ibdm);
+	page_idx = memic->mentry.rdma_entry.start_pgoff & 0xFFFF;
+	err = uverbs_copy_to(attrs, MLX5_IB_ATTR_QUERY_DM_RESP_PAGE_INDEX,
+			     &page_idx, sizeof(page_idx));
+	if (err)
+		return err;
+
+	start_offset = memic->base.dev_addr & ~PAGE_MASK;
+	err =  uverbs_copy_to(attrs, MLX5_IB_ATTR_QUERY_DM_RESP_START_OFFSET,
+			      &start_offset, sizeof(start_offset));
+	if (err)
+		return err;
+
+	return uverbs_copy_to(attrs, MLX5_IB_ATTR_QUERY_DM_RESP_LENGTH,
+			      &memic->req_length,
+			      sizeof(memic->req_length));
+}
+
+void mlx5_ib_dm_mmap_free(struct mlx5_ib_dev *dev,
+			  struct mlx5_user_mmap_entry *mentry)
+{
+	struct mlx5_ib_dm_op_entry *op_entry;
+	struct mlx5_ib_dm_memic *mdm;
+
+	switch (mentry->mmap_flag) {
+	case MLX5_IB_MMAP_TYPE_MEMIC:
+		mdm = container_of(mentry, struct mlx5_ib_dm_memic, mentry);
+		kref_put(&mdm->ref, mlx5_ib_dm_memic_free);
+		break;
+	case MLX5_IB_MMAP_TYPE_MEMIC_OP:
+		op_entry = container_of(mentry, struct mlx5_ib_dm_op_entry,
+					mentry);
+		mdm = op_entry->dm;
+		mlx5_cmd_dealloc_memic_op(&dev->dm, mdm->base.dev_addr,
+					  op_entry->op);
+		kfree(op_entry);
+		kref_put(&mdm->ref, mlx5_ib_dm_memic_free);
+		break;
+	default:
+		WARN_ON(true);
+	}
+}
+
+DECLARE_UVERBS_NAMED_METHOD(
+	MLX5_IB_METHOD_DM_QUERY,
+	UVERBS_ATTR_IDR(MLX5_IB_ATTR_QUERY_DM_REQ_HANDLE, UVERBS_OBJECT_DM,
+			UVERBS_ACCESS_READ, UA_MANDATORY),
+	UVERBS_ATTR_PTR_OUT(MLX5_IB_ATTR_QUERY_DM_RESP_START_OFFSET,
+			    UVERBS_ATTR_TYPE(u64), UA_MANDATORY),
+	UVERBS_ATTR_PTR_OUT(MLX5_IB_ATTR_QUERY_DM_RESP_PAGE_INDEX,
+			    UVERBS_ATTR_TYPE(u16), UA_MANDATORY),
+	UVERBS_ATTR_PTR_OUT(MLX5_IB_ATTR_QUERY_DM_RESP_LENGTH,
+			    UVERBS_ATTR_TYPE(u64), UA_MANDATORY));
+
+ADD_UVERBS_ATTRIBUTES_SIMPLE(
+	mlx5_ib_dm, UVERBS_OBJECT_DM, UVERBS_METHOD_DM_ALLOC,
+	UVERBS_ATTR_PTR_OUT(MLX5_IB_ATTR_ALLOC_DM_RESP_START_OFFSET,
+			    UVERBS_ATTR_TYPE(u64), UA_MANDATORY),
+	UVERBS_ATTR_PTR_OUT(MLX5_IB_ATTR_ALLOC_DM_RESP_PAGE_INDEX,
+			    UVERBS_ATTR_TYPE(u16), UA_OPTIONAL),
+	UVERBS_ATTR_CONST_IN(MLX5_IB_ATTR_ALLOC_DM_REQ_TYPE,
+			     enum mlx5_ib_uapi_dm_type, UA_OPTIONAL));
+
+DECLARE_UVERBS_NAMED_METHOD(
+	MLX5_IB_METHOD_DM_MAP_OP_ADDR,
+	UVERBS_ATTR_IDR(MLX5_IB_ATTR_DM_MAP_OP_ADDR_REQ_HANDLE,
+			UVERBS_OBJECT_DM,
+			UVERBS_ACCESS_READ,
+			UA_MANDATORY),
+	UVERBS_ATTR_PTR_IN(MLX5_IB_ATTR_DM_MAP_OP_ADDR_REQ_OP,
+			   UVERBS_ATTR_TYPE(u8),
+			   UA_MANDATORY),
+	UVERBS_ATTR_PTR_OUT(MLX5_IB_ATTR_DM_MAP_OP_ADDR_RESP_START_OFFSET,
+			    UVERBS_ATTR_TYPE(u64),
+			    UA_MANDATORY),
+	UVERBS_ATTR_PTR_OUT(MLX5_IB_ATTR_DM_MAP_OP_ADDR_RESP_PAGE_INDEX,
+			    UVERBS_ATTR_TYPE(u16),
+			    UA_OPTIONAL));
+
+DECLARE_UVERBS_GLOBAL_METHODS(UVERBS_OBJECT_DM,
+			      &UVERBS_METHOD(MLX5_IB_METHOD_DM_MAP_OP_ADDR),
+			      &UVERBS_METHOD(MLX5_IB_METHOD_DM_QUERY));
+
+const struct uapi_definition mlx5_ib_dm_defs[] = {
+	UAPI_DEF_CHAIN_OBJ_TREE(UVERBS_OBJECT_DM, &mlx5_ib_dm),
+	UAPI_DEF_CHAIN_OBJ_TREE_NAMED(UVERBS_OBJECT_DM),
+	{},
+};
+
+const struct ib_device_ops mlx5_ib_dev_dm_ops = {
+	.alloc_dm = mlx5_ib_alloc_dm,
+	.dealloc_dm = mlx5_ib_dealloc_dm,
+	.reg_dm_mr = mlx5_ib_reg_dm_mr,
+};
diff --git a/drivers/infiniband/hw/mlx5/dm.h b/drivers/infiniband/hw/mlx5/dm.h
new file mode 100644
index 000000000000..9674a80d8d70
--- /dev/null
+++ b/drivers/infiniband/hw/mlx5/dm.h
@@ -0,0 +1,68 @@
+/* SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB */
+/*
+ * Copyright (c) 2021, Mellanox Technologies inc. All rights reserved.
+ */
+
+#ifndef _MLX5_IB_DM_H
+#define _MLX5_IB_DM_H
+
+#include "mlx5_ib.h"
+
+extern const struct ib_device_ops mlx5_ib_dev_dm_ops;
+extern const struct uapi_definition mlx5_ib_dm_defs[];
+
+struct mlx5_ib_dm {
+	struct ib_dm		ibdm;
+	u32			type;
+	phys_addr_t		dev_addr;
+	size_t			size;
+};
+
+struct mlx5_ib_dm_op_entry {
+	struct mlx5_user_mmap_entry	mentry;
+	phys_addr_t			op_addr;
+	struct mlx5_ib_dm_memic		*dm;
+	u8				op;
+};
+
+struct mlx5_ib_dm_memic {
+	struct mlx5_ib_dm           base;
+	struct mlx5_user_mmap_entry mentry;
+	struct xarray               ops;
+	struct mutex                ops_xa_lock;
+	struct kref                 ref;
+	size_t                      req_length;
+};
+
+struct mlx5_ib_dm_icm {
+	struct mlx5_ib_dm      base;
+	u32                    obj_id;
+};
+
+static inline struct mlx5_ib_dm *to_mdm(struct ib_dm *ibdm)
+{
+	return container_of(ibdm, struct mlx5_ib_dm, ibdm);
+}
+
+static inline struct mlx5_ib_dm_memic *to_memic(struct ib_dm *ibdm)
+{
+	return container_of(ibdm, struct mlx5_ib_dm_memic, base.ibdm);
+}
+
+static inline struct mlx5_ib_dm_icm *to_icm(struct ib_dm *ibdm)
+{
+	return container_of(ibdm, struct mlx5_ib_dm_icm, base.ibdm);
+}
+
+struct ib_dm *mlx5_ib_alloc_dm(struct ib_device *ibdev,
+			       struct ib_ucontext *context,
+			       struct ib_dm_alloc_attr *attr,
+			       struct uverbs_attr_bundle *attrs);
+void mlx5_ib_dm_mmap_free(struct mlx5_ib_dev *dev,
+			  struct mlx5_user_mmap_entry *mentry);
+void mlx5_cmd_dealloc_memic(struct mlx5_dm *dm, phys_addr_t addr,
+			    u64 length);
+void mlx5_cmd_dealloc_memic_op(struct mlx5_dm *dm, phys_addr_t addr,
+			       u8 operation);
+
+#endif /* _MLX5_IB_DM_H */
diff --git a/drivers/infiniband/hw/mlx5/fs.c b/drivers/infiniband/hw/mlx5/fs.c
index 01370d9a871a..2fc6a60c4e77 100644
--- a/drivers/infiniband/hw/mlx5/fs.c
+++ b/drivers/infiniband/hw/mlx5/fs.c
@@ -1528,8 +1528,8 @@ static struct mlx5_ib_flow_handler *raw_fs_rule_add(
 		dst_num++;
 	}
 
-	handler = _create_raw_flow_rule(dev, ft_prio, dst, fs_matcher,
-					flow_context, flow_act,
+	handler = _create_raw_flow_rule(dev, ft_prio, dst_num ? dst : NULL,
+					fs_matcher, flow_context, flow_act,
 					cmd_in, inlen, dst_num);
 
 	if (IS_ERR(handler)) {
@@ -1885,8 +1885,9 @@ static int get_dests(struct uverbs_attr_bundle *attrs,
 		else
 			*dest_id = mqp->raw_packet_qp.rq.tirn;
 		*dest_type = MLX5_FLOW_DESTINATION_TYPE_TIR;
-	} else if (fs_matcher->ns_type == MLX5_FLOW_NAMESPACE_EGRESS ||
-		   fs_matcher->ns_type == MLX5_FLOW_NAMESPACE_RDMA_TX) {
+	} else if ((fs_matcher->ns_type == MLX5_FLOW_NAMESPACE_EGRESS ||
+		    fs_matcher->ns_type == MLX5_FLOW_NAMESPACE_RDMA_TX) &&
+		   !(*flags & MLX5_IB_ATTR_CREATE_FLOW_FLAGS_DROP)) {
 		*dest_type = MLX5_FLOW_DESTINATION_TYPE_PORT;
 	}
 
diff --git a/drivers/infiniband/hw/mlx5/ib_rep.c b/drivers/infiniband/hw/mlx5/ib_rep.c
index db5de720bb12..b25e0b33a11a 100644
--- a/drivers/infiniband/hw/mlx5/ib_rep.c
+++ b/drivers/infiniband/hw/mlx5/ib_rep.c
@@ -29,7 +29,7 @@ mlx5_ib_set_vport_rep(struct mlx5_core_dev *dev, struct mlx5_eswitch_rep *rep)
 static int
 mlx5_ib_vport_rep_load(struct mlx5_core_dev *dev, struct mlx5_eswitch_rep *rep)
 {
-	int num_ports = mlx5_eswitch_get_total_vports(dev);
+	u32 num_ports = mlx5_eswitch_get_total_vports(dev);
 	const struct mlx5_ib_profile *profile;
 	struct mlx5_ib_dev *ibdev;
 	int vport_index;
@@ -110,7 +110,7 @@ struct net_device *mlx5_ib_get_rep_netdev(struct mlx5_eswitch *esw,
 
 struct mlx5_flow_handle *create_flow_rule_vport_sq(struct mlx5_ib_dev *dev,
 						   struct mlx5_ib_sq *sq,
-						   u16 port)
+						   u32 port)
 {
 	struct mlx5_eswitch *esw = dev->mdev->priv.eswitch;
 	struct mlx5_eswitch_rep *rep;
diff --git a/drivers/infiniband/hw/mlx5/ib_rep.h b/drivers/infiniband/hw/mlx5/ib_rep.h
index ce1dcb105dbd..9c55e5c528b4 100644
--- a/drivers/infiniband/hw/mlx5/ib_rep.h
+++ b/drivers/infiniband/hw/mlx5/ib_rep.h
@@ -16,7 +16,7 @@ int mlx5r_rep_init(void);
 void mlx5r_rep_cleanup(void);
 struct mlx5_flow_handle *create_flow_rule_vport_sq(struct mlx5_ib_dev *dev,
 						   struct mlx5_ib_sq *sq,
-						   u16 port);
+						   u32 port);
 struct net_device *mlx5_ib_get_rep_netdev(struct mlx5_eswitch *esw,
 					  u16 vport_num);
 #else /* CONFIG_MLX5_ESWITCH */
@@ -25,7 +25,7 @@ static inline void mlx5r_rep_cleanup(void) {}
 static inline
 struct mlx5_flow_handle *create_flow_rule_vport_sq(struct mlx5_ib_dev *dev,
 						   struct mlx5_ib_sq *sq,
-						   u16 port)
+						   u32 port)
 {
 	return NULL;
 }
diff --git a/drivers/infiniband/hw/mlx5/ib_virt.c b/drivers/infiniband/hw/mlx5/ib_virt.c
index 46b2d370fb3f..f2f62875d072 100644
--- a/drivers/infiniband/hw/mlx5/ib_virt.c
+++ b/drivers/infiniband/hw/mlx5/ib_virt.c
@@ -48,7 +48,7 @@ static inline u32 mlx_to_net_policy(enum port_state_policy mlx_policy)
 	}
 }
 
-int mlx5_ib_get_vf_config(struct ib_device *device, int vf, u8 port,
+int mlx5_ib_get_vf_config(struct ib_device *device, int vf, u32 port,
 			  struct ifla_vf_info *info)
 {
 	struct mlx5_ib_dev *dev = to_mdev(device);
@@ -91,7 +91,7 @@ static inline enum port_state_policy net_to_mlx_policy(int policy)
 }
 
 int mlx5_ib_set_vf_link_state(struct ib_device *device, int vf,
-			      u8 port, int state)
+			      u32 port, int state)
 {
 	struct mlx5_ib_dev *dev = to_mdev(device);
 	struct mlx5_core_dev *mdev = dev->mdev;
@@ -119,7 +119,7 @@ out:
 }
 
 int mlx5_ib_get_vf_stats(struct ib_device *device, int vf,
-			 u8 port, struct ifla_vf_stats *stats)
+			 u32 port, struct ifla_vf_stats *stats)
 {
 	int out_sz = MLX5_ST_SZ_BYTES(query_vport_counter_out);
 	struct mlx5_core_dev *mdev;
@@ -149,7 +149,8 @@ ex:
 	return err;
 }
 
-static int set_vf_node_guid(struct ib_device *device, int vf, u8 port, u64 guid)
+static int set_vf_node_guid(struct ib_device *device, int vf, u32 port,
+			    u64 guid)
 {
 	struct mlx5_ib_dev *dev = to_mdev(device);
 	struct mlx5_core_dev *mdev = dev->mdev;
@@ -172,7 +173,8 @@ static int set_vf_node_guid(struct ib_device *device, int vf, u8 port, u64 guid)
 	return err;
 }
 
-static int set_vf_port_guid(struct ib_device *device, int vf, u8 port, u64 guid)
+static int set_vf_port_guid(struct ib_device *device, int vf, u32 port,
+			    u64 guid)
 {
 	struct mlx5_ib_dev *dev = to_mdev(device);
 	struct mlx5_core_dev *mdev = dev->mdev;
@@ -195,7 +197,7 @@ static int set_vf_port_guid(struct ib_device *device, int vf, u8 port, u64 guid)
 	return err;
 }
 
-int mlx5_ib_set_vf_guid(struct ib_device *device, int vf, u8 port,
+int mlx5_ib_set_vf_guid(struct ib_device *device, int vf, u32 port,
 			u64 guid, int type)
 {
 	if (type == IFLA_VF_IB_NODE_GUID)
@@ -206,7 +208,7 @@ int mlx5_ib_set_vf_guid(struct ib_device *device, int vf, u8 port,
 	return -EINVAL;
 }
 
-int mlx5_ib_get_vf_guid(struct ib_device *device, int vf, u8 port,
+int mlx5_ib_get_vf_guid(struct ib_device *device, int vf, u32 port,
 			struct ifla_vf_guid *node_guid,
 			struct ifla_vf_guid *port_guid)
 {
diff --git a/drivers/infiniband/hw/mlx5/mad.c b/drivers/infiniband/hw/mlx5/mad.c
index 652c6ccf1881..ec242a5a17a3 100644
--- a/drivers/infiniband/hw/mlx5/mad.c
+++ b/drivers/infiniband/hw/mlx5/mad.c
@@ -42,7 +42,7 @@ enum {
 	MLX5_IB_VENDOR_CLASS2 = 0xa
 };
 
-static bool can_do_mad_ifc(struct mlx5_ib_dev *dev, u8 port_num,
+static bool can_do_mad_ifc(struct mlx5_ib_dev *dev, u32 port_num,
 			   struct ib_mad *in_mad)
 {
 	if (in_mad->mad_hdr.mgmt_class != IB_MGMT_CLASS_SUBN_LID_ROUTED &&
@@ -52,7 +52,7 @@ static bool can_do_mad_ifc(struct mlx5_ib_dev *dev, u8 port_num,
 }
 
 static int mlx5_MAD_IFC(struct mlx5_ib_dev *dev, int ignore_mkey,
-			int ignore_bkey, u8 port, const struct ib_wc *in_wc,
+			int ignore_bkey, u32 port, const struct ib_wc *in_wc,
 			const struct ib_grh *in_grh, const void *in_mad,
 			void *response_mad)
 {
@@ -147,12 +147,12 @@ static void pma_cnt_assign(struct ib_pma_portcounters *pma_cnt,
 			     vl_15_dropped);
 }
 
-static int process_pma_cmd(struct mlx5_ib_dev *dev, u8 port_num,
+static int process_pma_cmd(struct mlx5_ib_dev *dev, u32 port_num,
 			   const struct ib_mad *in_mad, struct ib_mad *out_mad)
 {
 	struct mlx5_core_dev *mdev;
 	bool native_port = true;
-	u8 mdev_port_num;
+	u32 mdev_port_num;
 	void *out_cnt;
 	int err;
 
@@ -216,7 +216,7 @@ done:
 	return err;
 }
 
-int mlx5_ib_process_mad(struct ib_device *ibdev, int mad_flags, u8 port_num,
+int mlx5_ib_process_mad(struct ib_device *ibdev, int mad_flags, u32 port_num,
 			const struct ib_wc *in_wc, const struct ib_grh *in_grh,
 			const struct ib_mad *in, struct ib_mad *out,
 			size_t *out_mad_size, u16 *out_mad_pkey_index)
@@ -444,7 +444,7 @@ out:
 	return err;
 }
 
-int mlx5_query_mad_ifc_pkey(struct ib_device *ibdev, u8 port, u16 index,
+int mlx5_query_mad_ifc_pkey(struct ib_device *ibdev, u32 port, u16 index,
 			    u16 *pkey)
 {
 	struct ib_smp *in_mad  = NULL;
@@ -473,7 +473,7 @@ out:
 	return err;
 }
 
-int mlx5_query_mad_ifc_gids(struct ib_device *ibdev, u8 port, int index,
+int mlx5_query_mad_ifc_gids(struct ib_device *ibdev, u32 port, int index,
 			    union ib_gid *gid)
 {
 	struct ib_smp *in_mad  = NULL;
@@ -513,7 +513,7 @@ out:
 	return err;
 }
 
-int mlx5_query_mad_ifc_port(struct ib_device *ibdev, u8 port,
+int mlx5_query_mad_ifc_port(struct ib_device *ibdev, u32 port,
 			    struct ib_port_attr *props)
 {
 	struct mlx5_ib_dev *dev = to_mdev(ibdev);
diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index 7a7f6ccd02a5..6d1dd09a4388 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -34,6 +34,7 @@
 #include "ib_rep.h"
 #include "cmd.h"
 #include "devx.h"
+#include "dm.h"
 #include "fs.h"
 #include "srq.h"
 #include "qp.h"
@@ -42,6 +43,7 @@
 #include "counters.h"
 #include <linux/mlx5/accel.h>
 #include <rdma/uverbs_std_types.h>
+#include <rdma/uverbs_ioctl.h>
 #include <rdma/mlx5_user_ioctl_verbs.h>
 #include <rdma/mlx5_user_ioctl_cmds.h>
 #include <rdma/ib_umem_odp.h>
@@ -100,7 +102,7 @@ mlx5_port_type_cap_to_rdma_ll(int port_type_cap)
 }
 
 static enum rdma_link_layer
-mlx5_ib_port_link_layer(struct ib_device *device, u8 port_num)
+mlx5_ib_port_link_layer(struct ib_device *device, u32 port_num)
 {
 	struct mlx5_ib_dev *dev = to_mdev(device);
 	int port_type_cap = MLX5_CAP_GEN(dev->mdev, port_type);
@@ -109,7 +111,7 @@ mlx5_ib_port_link_layer(struct ib_device *device, u8 port_num)
 }
 
 static int get_port_state(struct ib_device *ibdev,
-			  u8 port_num,
+			  u32 port_num,
 			  enum ib_port_state *state)
 {
 	struct ib_port_attr attr;
@@ -124,7 +126,7 @@ static int get_port_state(struct ib_device *ibdev,
 
 static struct mlx5_roce *mlx5_get_rep_roce(struct mlx5_ib_dev *dev,
 					   struct net_device *ndev,
-					   u8 *port_num)
+					   u32 *port_num)
 {
 	struct net_device *rep_ndev;
 	struct mlx5_ib_port *port;
@@ -154,7 +156,7 @@ static int mlx5_netdev_event(struct notifier_block *this,
 {
 	struct mlx5_roce *roce = container_of(this, struct mlx5_roce, nb);
 	struct net_device *ndev = netdev_notifier_info_to_dev(ptr);
-	u8 port_num = roce->native_port_num;
+	u32 port_num = roce->native_port_num;
 	struct mlx5_core_dev *mdev;
 	struct mlx5_ib_dev *ibdev;
 
@@ -233,7 +235,7 @@ done:
 }
 
 static struct net_device *mlx5_ib_get_netdev(struct ib_device *device,
-					     u8 port_num)
+					     u32 port_num)
 {
 	struct mlx5_ib_dev *ibdev = to_mdev(device);
 	struct net_device *ndev;
@@ -261,8 +263,8 @@ out:
 }
 
 struct mlx5_core_dev *mlx5_ib_get_native_port_mdev(struct mlx5_ib_dev *ibdev,
-						   u8 ib_port_num,
-						   u8 *native_port_num)
+						   u32 ib_port_num,
+						   u32 *native_port_num)
 {
 	enum rdma_link_layer ll = mlx5_ib_port_link_layer(&ibdev->ib_dev,
 							  ib_port_num);
@@ -296,7 +298,7 @@ struct mlx5_core_dev *mlx5_ib_get_native_port_mdev(struct mlx5_ib_dev *ibdev,
 	return mdev;
 }
 
-void mlx5_ib_put_native_port_mdev(struct mlx5_ib_dev *ibdev, u8 port_num)
+void mlx5_ib_put_native_port_mdev(struct mlx5_ib_dev *ibdev, u32 port_num)
 {
 	enum rdma_link_layer ll = mlx5_ib_port_link_layer(&ibdev->ib_dev,
 							  port_num);
@@ -452,7 +454,7 @@ static int translate_eth_proto_oper(u32 eth_proto_oper, u16 *active_speed,
 						active_width);
 }
 
-static int mlx5_query_port_roce(struct ib_device *device, u8 port_num,
+static int mlx5_query_port_roce(struct ib_device *device, u32 port_num,
 				struct ib_port_attr *props)
 {
 	struct mlx5_ib_dev *dev = to_mdev(device);
@@ -462,7 +464,7 @@ static int mlx5_query_port_roce(struct ib_device *device, u8 port_num,
 	enum ib_mtu ndev_ib_mtu;
 	bool put_mdev = true;
 	u32 eth_prot_oper;
-	u8 mdev_port_num;
+	u32 mdev_port_num;
 	bool ext;
 	int err;
 
@@ -498,7 +500,7 @@ static int mlx5_query_port_roce(struct ib_device *device, u8 port_num,
 	translate_eth_proto_oper(eth_prot_oper, &props->active_speed,
 				 &props->active_width, ext);
 
-	if (!dev->is_rep && mlx5_is_roce_enabled(mdev)) {
+	if (!dev->is_rep && dev->mdev->roce.roce_en) {
 		u16 qkey_viol_cntr;
 
 		props->port_cap_flags |= IB_PORT_CM_SUP;
@@ -549,19 +551,19 @@ out:
 	return err;
 }
 
-static int set_roce_addr(struct mlx5_ib_dev *dev, u8 port_num,
+static int set_roce_addr(struct mlx5_ib_dev *dev, u32 port_num,
 			 unsigned int index, const union ib_gid *gid,
 			 const struct ib_gid_attr *attr)
 {
-	enum ib_gid_type gid_type = IB_GID_TYPE_ROCE;
+	enum ib_gid_type gid_type;
 	u16 vlan_id = 0xffff;
 	u8 roce_version = 0;
 	u8 roce_l3_type = 0;
 	u8 mac[ETH_ALEN];
 	int ret;
 
+	gid_type = attr->gid_type;
 	if (gid) {
-		gid_type = attr->gid_type;
 		ret = rdma_read_gid_l2_fields(attr, &vlan_id, &mac[0]);
 		if (ret)
 			return ret;
@@ -573,7 +575,7 @@ static int set_roce_addr(struct mlx5_ib_dev *dev, u8 port_num,
 		break;
 	case IB_GID_TYPE_ROCE_UDP_ENCAP:
 		roce_version = MLX5_ROCE_VERSION_2;
-		if (ipv6_addr_v4mapped((void *)gid))
+		if (gid && ipv6_addr_v4mapped((void *)gid))
 			roce_l3_type = MLX5_ROCE_L3_TYPE_IPV4;
 		else
 			roce_l3_type = MLX5_ROCE_L3_TYPE_IPV6;
@@ -600,7 +602,7 @@ static int mlx5_ib_del_gid(const struct ib_gid_attr *attr,
 			   __always_unused void **context)
 {
 	return set_roce_addr(to_mdev(attr->device), attr->port_num,
-			     attr->index, NULL, NULL);
+			     attr->index, NULL, attr);
 }
 
 __be16 mlx5_get_roce_udp_sport_min(const struct mlx5_ib_dev *dev,
@@ -1267,7 +1269,7 @@ static int translate_max_vl_num(struct ib_device *ibdev, u8 vl_hw_cap,
 	return 0;
 }
 
-static int mlx5_query_hca_port(struct ib_device *ibdev, u8 port,
+static int mlx5_query_hca_port(struct ib_device *ibdev, u32 port,
 			       struct ib_port_attr *props)
 {
 	struct mlx5_ib_dev *dev = to_mdev(ibdev);
@@ -1335,7 +1337,7 @@ out:
 	return err;
 }
 
-int mlx5_ib_query_port(struct ib_device *ibdev, u8 port,
+int mlx5_ib_query_port(struct ib_device *ibdev, u32 port,
 		       struct ib_port_attr *props)
 {
 	unsigned int count;
@@ -1380,13 +1382,13 @@ int mlx5_ib_query_port(struct ib_device *ibdev, u8 port,
 	return ret;
 }
 
-static int mlx5_ib_rep_query_port(struct ib_device *ibdev, u8 port,
+static int mlx5_ib_rep_query_port(struct ib_device *ibdev, u32 port,
 				  struct ib_port_attr *props)
 {
 	return mlx5_query_port_roce(ibdev, port, props);
 }
 
-static int mlx5_ib_rep_query_pkey(struct ib_device *ibdev, u8 port, u16 index,
+static int mlx5_ib_rep_query_pkey(struct ib_device *ibdev, u32 port, u16 index,
 				  u16 *pkey)
 {
 	/* Default special Pkey for representor device port as per the
@@ -1396,7 +1398,7 @@ static int mlx5_ib_rep_query_pkey(struct ib_device *ibdev, u8 port, u16 index,
 	return 0;
 }
 
-static int mlx5_ib_query_gid(struct ib_device *ibdev, u8 port, int index,
+static int mlx5_ib_query_gid(struct ib_device *ibdev, u32 port, int index,
 			     union ib_gid *gid)
 {
 	struct mlx5_ib_dev *dev = to_mdev(ibdev);
@@ -1415,13 +1417,13 @@ static int mlx5_ib_query_gid(struct ib_device *ibdev, u8 port, int index,
 
 }
 
-static int mlx5_query_hca_nic_pkey(struct ib_device *ibdev, u8 port,
+static int mlx5_query_hca_nic_pkey(struct ib_device *ibdev, u32 port,
 				   u16 index, u16 *pkey)
 {
 	struct mlx5_ib_dev *dev = to_mdev(ibdev);
 	struct mlx5_core_dev *mdev;
 	bool put_mdev = true;
-	u8 mdev_port_num;
+	u32 mdev_port_num;
 	int err;
 
 	mdev = mlx5_ib_get_native_port_mdev(dev, port, &mdev_port_num);
@@ -1442,7 +1444,7 @@ static int mlx5_query_hca_nic_pkey(struct ib_device *ibdev, u8 port,
 	return err;
 }
 
-static int mlx5_ib_query_pkey(struct ib_device *ibdev, u8 port, u16 index,
+static int mlx5_ib_query_pkey(struct ib_device *ibdev, u32 port, u16 index,
 			      u16 *pkey)
 {
 	switch (mlx5_get_vport_access_method(ibdev)) {
@@ -1486,12 +1488,12 @@ static int mlx5_ib_modify_device(struct ib_device *ibdev, int mask,
 	return err;
 }
 
-static int set_port_caps_atomic(struct mlx5_ib_dev *dev, u8 port_num, u32 mask,
+static int set_port_caps_atomic(struct mlx5_ib_dev *dev, u32 port_num, u32 mask,
 				u32 value)
 {
 	struct mlx5_hca_vport_context ctx = {};
 	struct mlx5_core_dev *mdev;
-	u8 mdev_port_num;
+	u32 mdev_port_num;
 	int err;
 
 	mdev = mlx5_ib_get_native_port_mdev(dev, port_num, &mdev_port_num);
@@ -1520,7 +1522,7 @@ out:
 	return err;
 }
 
-static int mlx5_ib_modify_port(struct ib_device *ibdev, u8 port, int mask,
+static int mlx5_ib_modify_port(struct ib_device *ibdev, u32 port, int mask,
 			       struct ib_port_modify *props)
 {
 	struct mlx5_ib_dev *dev = to_mdev(ibdev);
@@ -1929,7 +1931,7 @@ uar_done:
 	print_lib_caps(dev, context->lib_caps);
 
 	if (mlx5_ib_lag_should_assign_affinity(dev)) {
-		u8 port = mlx5_core_native_port_num(dev->mdev) - 1;
+		u32 port = mlx5_core_native_port_num(dev->mdev) - 1;
 
 		atomic_set(&context->tx_port_affinity,
 			   atomic_add_return(
@@ -2087,14 +2089,11 @@ static void mlx5_ib_mmap_free(struct rdma_user_mmap_entry *entry)
 	struct mlx5_user_mmap_entry *mentry = to_mmmap(entry);
 	struct mlx5_ib_dev *dev = to_mdev(entry->ucontext->device);
 	struct mlx5_var_table *var_table = &dev->var_table;
-	struct mlx5_ib_dm *mdm;
 
 	switch (mentry->mmap_flag) {
 	case MLX5_IB_MMAP_TYPE_MEMIC:
-		mdm = container_of(mentry, struct mlx5_ib_dm, mentry);
-		mlx5_cmd_dealloc_memic(&dev->dm, mdm->dev_addr,
-				       mdm->size);
-		kfree(mdm);
+	case MLX5_IB_MMAP_TYPE_MEMIC_OP:
+		mlx5_ib_dm_mmap_free(dev, mentry);
 		break;
 	case MLX5_IB_MMAP_TYPE_VAR:
 		mutex_lock(&var_table->bitmap_lock);
@@ -2219,19 +2218,6 @@ free_bfreg:
 	return err;
 }
 
-static int add_dm_mmap_entry(struct ib_ucontext *context,
-			     struct mlx5_ib_dm *mdm,
-			     u64 address)
-{
-	mdm->mentry.mmap_flag = MLX5_IB_MMAP_TYPE_MEMIC;
-	mdm->mentry.address = address;
-	return rdma_user_mmap_entry_insert_range(
-			context, &mdm->mentry.rdma_entry,
-			mdm->size,
-			MLX5_IB_MMAP_DEVICE_MEM << 16,
-			(MLX5_IB_MMAP_DEVICE_MEM << 16) + (1UL << 16) - 1);
-}
-
 static unsigned long mlx5_vma_to_pgoff(struct vm_area_struct *vma)
 {
 	unsigned long idx;
@@ -2333,206 +2319,6 @@ static int mlx5_ib_mmap(struct ib_ucontext *ibcontext, struct vm_area_struct *vm
 	return 0;
 }
 
-static inline int check_dm_type_support(struct mlx5_ib_dev *dev,
-					u32 type)
-{
-	switch (type) {
-	case MLX5_IB_UAPI_DM_TYPE_MEMIC:
-		if (!MLX5_CAP_DEV_MEM(dev->mdev, memic))
-			return -EOPNOTSUPP;
-		break;
-	case MLX5_IB_UAPI_DM_TYPE_STEERING_SW_ICM:
-	case MLX5_IB_UAPI_DM_TYPE_HEADER_MODIFY_SW_ICM:
-		if (!capable(CAP_SYS_RAWIO) ||
-		    !capable(CAP_NET_RAW))
-			return -EPERM;
-
-		if (!(MLX5_CAP_FLOWTABLE_NIC_RX(dev->mdev, sw_owner) ||
-		      MLX5_CAP_FLOWTABLE_NIC_TX(dev->mdev, sw_owner) ||
-		      MLX5_CAP_FLOWTABLE_NIC_RX(dev->mdev, sw_owner_v2) ||
-		      MLX5_CAP_FLOWTABLE_NIC_TX(dev->mdev, sw_owner_v2)))
-			return -EOPNOTSUPP;
-		break;
-	}
-
-	return 0;
-}
-
-static int handle_alloc_dm_memic(struct ib_ucontext *ctx,
-				 struct mlx5_ib_dm *dm,
-				 struct ib_dm_alloc_attr *attr,
-				 struct uverbs_attr_bundle *attrs)
-{
-	struct mlx5_dm *dm_db = &to_mdev(ctx->device)->dm;
-	u64 start_offset;
-	u16 page_idx;
-	int err;
-	u64 address;
-
-	dm->size = roundup(attr->length, MLX5_MEMIC_BASE_SIZE);
-
-	err = mlx5_cmd_alloc_memic(dm_db, &dm->dev_addr,
-				   dm->size, attr->alignment);
-	if (err)
-		return err;
-
-	address = dm->dev_addr & PAGE_MASK;
-	err = add_dm_mmap_entry(ctx, dm, address);
-	if (err)
-		goto err_dealloc;
-
-	page_idx = dm->mentry.rdma_entry.start_pgoff & 0xFFFF;
-	err = uverbs_copy_to(attrs,
-			     MLX5_IB_ATTR_ALLOC_DM_RESP_PAGE_INDEX,
-			     &page_idx,
-			     sizeof(page_idx));
-	if (err)
-		goto err_copy;
-
-	start_offset = dm->dev_addr & ~PAGE_MASK;
-	err = uverbs_copy_to(attrs,
-			     MLX5_IB_ATTR_ALLOC_DM_RESP_START_OFFSET,
-			     &start_offset, sizeof(start_offset));
-	if (err)
-		goto err_copy;
-
-	return 0;
-
-err_copy:
-	rdma_user_mmap_entry_remove(&dm->mentry.rdma_entry);
-err_dealloc:
-	mlx5_cmd_dealloc_memic(dm_db, dm->dev_addr, dm->size);
-
-	return err;
-}
-
-static int handle_alloc_dm_sw_icm(struct ib_ucontext *ctx,
-				  struct mlx5_ib_dm *dm,
-				  struct ib_dm_alloc_attr *attr,
-				  struct uverbs_attr_bundle *attrs,
-				  int type)
-{
-	struct mlx5_core_dev *dev = to_mdev(ctx->device)->mdev;
-	u64 act_size;
-	int err;
-
-	/* Allocation size must a multiple of the basic block size
-	 * and a power of 2.
-	 */
-	act_size = round_up(attr->length, MLX5_SW_ICM_BLOCK_SIZE(dev));
-	act_size = roundup_pow_of_two(act_size);
-
-	dm->size = act_size;
-	err = mlx5_dm_sw_icm_alloc(dev, type, act_size, attr->alignment,
-				   to_mucontext(ctx)->devx_uid, &dm->dev_addr,
-				   &dm->icm_dm.obj_id);
-	if (err)
-		return err;
-
-	err = uverbs_copy_to(attrs,
-			     MLX5_IB_ATTR_ALLOC_DM_RESP_START_OFFSET,
-			     &dm->dev_addr, sizeof(dm->dev_addr));
-	if (err)
-		mlx5_dm_sw_icm_dealloc(dev, type, dm->size,
-				       to_mucontext(ctx)->devx_uid, dm->dev_addr,
-				       dm->icm_dm.obj_id);
-
-	return err;
-}
-
-struct ib_dm *mlx5_ib_alloc_dm(struct ib_device *ibdev,
-			       struct ib_ucontext *context,
-			       struct ib_dm_alloc_attr *attr,
-			       struct uverbs_attr_bundle *attrs)
-{
-	struct mlx5_ib_dm *dm;
-	enum mlx5_ib_uapi_dm_type type;
-	int err;
-
-	err = uverbs_get_const_default(&type, attrs,
-				       MLX5_IB_ATTR_ALLOC_DM_REQ_TYPE,
-				       MLX5_IB_UAPI_DM_TYPE_MEMIC);
-	if (err)
-		return ERR_PTR(err);
-
-	mlx5_ib_dbg(to_mdev(ibdev), "alloc_dm req: dm_type=%d user_length=0x%llx log_alignment=%d\n",
-		    type, attr->length, attr->alignment);
-
-	err = check_dm_type_support(to_mdev(ibdev), type);
-	if (err)
-		return ERR_PTR(err);
-
-	dm = kzalloc(sizeof(*dm), GFP_KERNEL);
-	if (!dm)
-		return ERR_PTR(-ENOMEM);
-
-	dm->type = type;
-
-	switch (type) {
-	case MLX5_IB_UAPI_DM_TYPE_MEMIC:
-		err = handle_alloc_dm_memic(context, dm,
-					    attr,
-					    attrs);
-		break;
-	case MLX5_IB_UAPI_DM_TYPE_STEERING_SW_ICM:
-		err = handle_alloc_dm_sw_icm(context, dm,
-					     attr, attrs,
-					     MLX5_SW_ICM_TYPE_STEERING);
-		break;
-	case MLX5_IB_UAPI_DM_TYPE_HEADER_MODIFY_SW_ICM:
-		err = handle_alloc_dm_sw_icm(context, dm,
-					     attr, attrs,
-					     MLX5_SW_ICM_TYPE_HEADER_MODIFY);
-		break;
-	default:
-		err = -EOPNOTSUPP;
-	}
-
-	if (err)
-		goto err_free;
-
-	return &dm->ibdm;
-
-err_free:
-	kfree(dm);
-	return ERR_PTR(err);
-}
-
-int mlx5_ib_dealloc_dm(struct ib_dm *ibdm, struct uverbs_attr_bundle *attrs)
-{
-	struct mlx5_ib_ucontext *ctx = rdma_udata_to_drv_context(
-		&attrs->driver_udata, struct mlx5_ib_ucontext, ibucontext);
-	struct mlx5_core_dev *dev = to_mdev(ibdm->device)->mdev;
-	struct mlx5_ib_dm *dm = to_mdm(ibdm);
-	int ret;
-
-	switch (dm->type) {
-	case MLX5_IB_UAPI_DM_TYPE_MEMIC:
-		rdma_user_mmap_entry_remove(&dm->mentry.rdma_entry);
-		return 0;
-	case MLX5_IB_UAPI_DM_TYPE_STEERING_SW_ICM:
-		ret = mlx5_dm_sw_icm_dealloc(dev, MLX5_SW_ICM_TYPE_STEERING,
-					     dm->size, ctx->devx_uid, dm->dev_addr,
-					     dm->icm_dm.obj_id);
-		if (ret)
-			return ret;
-		break;
-	case MLX5_IB_UAPI_DM_TYPE_HEADER_MODIFY_SW_ICM:
-		ret = mlx5_dm_sw_icm_dealloc(dev, MLX5_SW_ICM_TYPE_HEADER_MODIFY,
-					     dm->size, ctx->devx_uid, dm->dev_addr,
-					     dm->icm_dm.obj_id);
-		if (ret)
-			return ret;
-		break;
-	default:
-		return -EOPNOTSUPP;
-	}
-
-	kfree(dm);
-
-	return 0;
-}
-
 static int mlx5_ib_alloc_pd(struct ib_pd *ibpd, struct ib_udata *udata)
 {
 	struct mlx5_ib_pd *pd = to_mpd(ibpd);
@@ -2779,7 +2565,7 @@ static void delay_drop_handler(struct work_struct *work)
 static void handle_general_event(struct mlx5_ib_dev *ibdev, struct mlx5_eqe *eqe,
 				 struct ib_event *ibev)
 {
-	u8 port = (eqe->data.port.port >> 4) & 0xf;
+	u32 port = (eqe->data.port.port >> 4) & 0xf;
 
 	switch (eqe->sub_type) {
 	case MLX5_GENERAL_SUBTYPE_DELAY_DROP_TIMEOUT:
@@ -2795,7 +2581,7 @@ static void handle_general_event(struct mlx5_ib_dev *ibdev, struct mlx5_eqe *eqe
 static int handle_port_change(struct mlx5_ib_dev *ibdev, struct mlx5_eqe *eqe,
 			      struct ib_event *ibev)
 {
-	u8 port = (eqe->data.port.port >> 4) & 0xf;
+	u32 port = (eqe->data.port.port >> 4) & 0xf;
 
 	ibev->element.port_num = port;
 
@@ -3152,7 +2938,7 @@ static u32 get_core_cap_flags(struct ib_device *ibdev,
 	return ret;
 }
 
-static int mlx5_port_immutable(struct ib_device *ibdev, u8 port_num,
+static int mlx5_port_immutable(struct ib_device *ibdev, u32 port_num,
 			       struct ib_port_immutable *immutable)
 {
 	struct ib_port_attr attr;
@@ -3180,7 +2966,7 @@ static int mlx5_port_immutable(struct ib_device *ibdev, u8 port_num,
 	return 0;
 }
 
-static int mlx5_port_rep_immutable(struct ib_device *ibdev, u8 port_num,
+static int mlx5_port_rep_immutable(struct ib_device *ibdev, u32 port_num,
 				   struct ib_port_immutable *immutable)
 {
 	struct ib_port_attr attr;
@@ -3252,7 +3038,7 @@ static void mlx5_eth_lag_cleanup(struct mlx5_ib_dev *dev)
 	}
 }
 
-static int mlx5_add_netdev_notifier(struct mlx5_ib_dev *dev, u8 port_num)
+static int mlx5_add_netdev_notifier(struct mlx5_ib_dev *dev, u32 port_num)
 {
 	int err;
 
@@ -3266,7 +3052,7 @@ static int mlx5_add_netdev_notifier(struct mlx5_ib_dev *dev, u8 port_num)
 	return 0;
 }
 
-static void mlx5_remove_netdev_notifier(struct mlx5_ib_dev *dev, u8 port_num)
+static void mlx5_remove_netdev_notifier(struct mlx5_ib_dev *dev, u32 port_num)
 {
 	if (dev->port[port_num].roce.nb.notifier_call) {
 		unregister_netdevice_notifier(&dev->port[port_num].roce.nb);
@@ -3300,7 +3086,7 @@ static void mlx5_disable_eth(struct mlx5_ib_dev *dev)
 	mlx5_nic_vport_disable_roce(dev->mdev);
 }
 
-static int mlx5_ib_rn_get_params(struct ib_device *device, u8 port_num,
+static int mlx5_ib_rn_get_params(struct ib_device *device, u32 port_num,
 				 enum rdma_netdev_t type,
 				 struct rdma_netdev_alloc_params *params)
 {
@@ -3352,7 +3138,7 @@ static const struct file_operations fops_delay_drop_timeout = {
 static void mlx5_ib_unbind_slave_port(struct mlx5_ib_dev *ibdev,
 				      struct mlx5_ib_multiport_info *mpi)
 {
-	u8 port_num = mlx5_core_native_port_num(mpi->mdev) - 1;
+	u32 port_num = mlx5_core_native_port_num(mpi->mdev) - 1;
 	struct mlx5_ib_port *port = &ibdev->port[port_num];
 	int comps;
 	int err;
@@ -3398,7 +3184,7 @@ static void mlx5_ib_unbind_slave_port(struct mlx5_ib_dev *ibdev,
 
 	err = mlx5_nic_vport_unaffiliate_multiport(mpi->mdev);
 
-	mlx5_ib_dbg(ibdev, "unaffiliated port %d\n", port_num + 1);
+	mlx5_ib_dbg(ibdev, "unaffiliated port %u\n", port_num + 1);
 	/* Log an error, still needed to cleanup the pointers and add
 	 * it back to the list.
 	 */
@@ -3412,14 +3198,14 @@ static void mlx5_ib_unbind_slave_port(struct mlx5_ib_dev *ibdev,
 static bool mlx5_ib_bind_slave_port(struct mlx5_ib_dev *ibdev,
 				    struct mlx5_ib_multiport_info *mpi)
 {
-	u8 port_num = mlx5_core_native_port_num(mpi->mdev) - 1;
+	u32 port_num = mlx5_core_native_port_num(mpi->mdev) - 1;
 	int err;
 
 	lockdep_assert_held(&mlx5_ib_multiport_mutex);
 
 	spin_lock(&ibdev->port[port_num].mp.mpi_lock);
 	if (ibdev->port[port_num].mp.mpi) {
-		mlx5_ib_dbg(ibdev, "port %d already affiliated.\n",
+		mlx5_ib_dbg(ibdev, "port %u already affiliated.\n",
 			    port_num + 1);
 		spin_unlock(&ibdev->port[port_num].mp.mpi_lock);
 		return false;
@@ -3455,12 +3241,12 @@ unbind:
 
 static int mlx5_ib_init_multiport_master(struct mlx5_ib_dev *dev)
 {
-	int port_num = mlx5_core_native_port_num(dev->mdev) - 1;
+	u32 port_num = mlx5_core_native_port_num(dev->mdev) - 1;
 	enum rdma_link_layer ll = mlx5_ib_port_link_layer(&dev->ib_dev,
 							  port_num + 1);
 	struct mlx5_ib_multiport_info *mpi;
 	int err;
-	int i;
+	u32 i;
 
 	if (!mlx5_core_is_mp_master(dev->mdev) || ll != IB_LINK_LAYER_ETHERNET)
 		return 0;
@@ -3523,10 +3309,10 @@ static int mlx5_ib_init_multiport_master(struct mlx5_ib_dev *dev)
 
 static void mlx5_ib_cleanup_multiport_master(struct mlx5_ib_dev *dev)
 {
-	int port_num = mlx5_core_native_port_num(dev->mdev) - 1;
+	u32 port_num = mlx5_core_native_port_num(dev->mdev) - 1;
 	enum rdma_link_layer ll = mlx5_ib_port_link_layer(&dev->ib_dev,
 							  port_num + 1);
-	int i;
+	u32 i;
 
 	if (!mlx5_core_is_mp_master(dev->mdev) || ll != IB_LINK_LAYER_ETHERNET)
 		return;
@@ -3539,7 +3325,8 @@ static void mlx5_ib_cleanup_multiport_master(struct mlx5_ib_dev *dev)
 				kfree(dev->port[i].mp.mpi);
 				dev->port[i].mp.mpi = NULL;
 			} else {
-				mlx5_ib_dbg(dev, "unbinding port_num: %d\n", i + 1);
+				mlx5_ib_dbg(dev, "unbinding port_num: %u\n",
+					    i + 1);
 				mlx5_ib_unbind_slave_port(dev, dev->port[i].mp.mpi);
 			}
 		}
@@ -3816,20 +3603,6 @@ DECLARE_UVERBS_NAMED_OBJECT(MLX5_IB_OBJECT_UAR,
 			    &UVERBS_METHOD(MLX5_IB_METHOD_UAR_OBJ_DESTROY));
 
 ADD_UVERBS_ATTRIBUTES_SIMPLE(
-	mlx5_ib_dm,
-	UVERBS_OBJECT_DM,
-	UVERBS_METHOD_DM_ALLOC,
-	UVERBS_ATTR_PTR_OUT(MLX5_IB_ATTR_ALLOC_DM_RESP_START_OFFSET,
-			    UVERBS_ATTR_TYPE(u64),
-			    UA_MANDATORY),
-	UVERBS_ATTR_PTR_OUT(MLX5_IB_ATTR_ALLOC_DM_RESP_PAGE_INDEX,
-			    UVERBS_ATTR_TYPE(u16),
-			    UA_OPTIONAL),
-	UVERBS_ATTR_CONST_IN(MLX5_IB_ATTR_ALLOC_DM_REQ_TYPE,
-			     enum mlx5_ib_uapi_dm_type,
-			     UA_OPTIONAL));
-
-ADD_UVERBS_ATTRIBUTES_SIMPLE(
 	mlx5_ib_flow_action,
 	UVERBS_OBJECT_FLOW_ACTION,
 	UVERBS_METHOD_FLOW_ACTION_ESP_CREATE,
@@ -3851,10 +3624,10 @@ static const struct uapi_definition mlx5_ib_defs[] = {
 	UAPI_DEF_CHAIN(mlx5_ib_flow_defs),
 	UAPI_DEF_CHAIN(mlx5_ib_qos_defs),
 	UAPI_DEF_CHAIN(mlx5_ib_std_types_defs),
+	UAPI_DEF_CHAIN(mlx5_ib_dm_defs),
 
 	UAPI_DEF_CHAIN_OBJ_TREE(UVERBS_OBJECT_FLOW_ACTION,
 				&mlx5_ib_flow_action),
-	UAPI_DEF_CHAIN_OBJ_TREE(UVERBS_OBJECT_DM, &mlx5_ib_dm),
 	UAPI_DEF_CHAIN_OBJ_TREE(UVERBS_OBJECT_DEVICE, &mlx5_ib_query_context),
 	UAPI_DEF_CHAIN_OBJ_TREE_NAMED(MLX5_IB_OBJECT_VAR,
 				UAPI_DEF_IS_OBJ_SUPPORTED(var_is_supported)),
@@ -3891,8 +3664,6 @@ static int mlx5_ib_stage_init_init(struct mlx5_ib_dev *dev)
 		dev->port[i].roce.last_port_state = IB_PORT_DOWN;
 	}
 
-	mlx5_ib_internal_fill_odp_caps(dev);
-
 	err = mlx5_ib_init_multiport_master(dev);
 	if (err)
 		return err;
@@ -4032,12 +3803,6 @@ static const struct ib_device_ops mlx5_ib_dev_xrc_ops = {
 	INIT_RDMA_OBJ_SIZE(ib_xrcd, mlx5_ib_xrcd, ibxrcd),
 };
 
-static const struct ib_device_ops mlx5_ib_dev_dm_ops = {
-	.alloc_dm = mlx5_ib_alloc_dm,
-	.dealloc_dm = mlx5_ib_dealloc_dm,
-	.reg_dm_mr = mlx5_ib_reg_dm_mr,
-};
-
 static int mlx5_ib_init_var_table(struct mlx5_ib_dev *dev)
 {
 	struct mlx5_core_dev *mdev = dev->mdev;
@@ -4160,7 +3925,7 @@ static int mlx5_ib_roce_init(struct mlx5_ib_dev *dev)
 	struct mlx5_core_dev *mdev = dev->mdev;
 	enum rdma_link_layer ll;
 	int port_type_cap;
-	u8 port_num = 0;
+	u32 port_num = 0;
 	int err;
 
 	port_type_cap = MLX5_CAP_GEN(mdev, port_type);
@@ -4173,7 +3938,7 @@ static int mlx5_ib_roce_init(struct mlx5_ib_dev *dev)
 
 		/* Register only for native ports */
 		err = mlx5_add_netdev_notifier(dev, port_num);
-		if (err || dev->is_rep || !mlx5_is_roce_enabled(mdev))
+		if (err || dev->is_rep || !mlx5_is_roce_init_enabled(mdev))
 			/*
 			 * We don't enable ETH interface for
 			 * 1. IB representors
@@ -4197,7 +3962,7 @@ static void mlx5_ib_roce_cleanup(struct mlx5_ib_dev *dev)
 	struct mlx5_core_dev *mdev = dev->mdev;
 	enum rdma_link_layer ll;
 	int port_type_cap;
-	u8 port_num;
+	u32 port_num;
 
 	port_type_cap = MLX5_CAP_GEN(mdev, port_type);
 	ll = mlx5_port_type_cap_to_rdma_ll(port_type_cap);
@@ -4710,7 +4475,7 @@ static int mlx5r_probe(struct auxiliary_device *adev,
 	dev->mdev = mdev;
 	dev->num_ports = num_ports;
 
-	if (ll == IB_LINK_LAYER_ETHERNET && !mlx5_is_roce_enabled(mdev))
+	if (ll == IB_LINK_LAYER_ETHERNET && !mlx5_is_roce_init_enabled(mdev))
 		profile = &raw_eth_profile;
 	else
 		profile = &pf_profile;
diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h
index 88cc26e008fc..e9a3f34a30b8 100644
--- a/drivers/infiniband/hw/mlx5/mlx5_ib.h
+++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h
@@ -166,6 +166,7 @@ enum mlx5_ib_mmap_type {
 	MLX5_IB_MMAP_TYPE_VAR = 2,
 	MLX5_IB_MMAP_TYPE_UAR_WC = 3,
 	MLX5_IB_MMAP_TYPE_UAR_NC = 4,
+	MLX5_IB_MMAP_TYPE_MEMIC_OP = 5,
 };
 
 struct mlx5_bfreg_info {
@@ -406,7 +407,7 @@ struct mlx5_ib_qp_base {
 struct mlx5_ib_qp_trans {
 	struct mlx5_ib_qp_base	base;
 	u16			xrcdn;
-	u8			alt_port;
+	u32			alt_port;
 	u8			atomic_rd_en;
 	u8			resp_depth;
 };
@@ -453,7 +454,7 @@ struct mlx5_ib_dct {
 
 struct mlx5_ib_gsi_qp {
 	struct ib_qp *rx_qp;
-	u8 port_num;
+	u32 port_num;
 	struct ib_qp_cap cap;
 	struct ib_cq *cq;
 	struct mlx5_ib_gsi_wr *outstanding_wrs;
@@ -490,7 +491,7 @@ struct mlx5_ib_qp {
 	struct mutex		mutex;
 	/* cached variant of create_flags from struct ib_qp_init_attr */
 	u32			flags;
-	u8			port;
+	u32			port;
 	u8			state;
 	int			max_inline_data;
 	struct mlx5_bf	        bf;
@@ -547,11 +548,6 @@ static inline const struct mlx5_umr_wr *umr_wr(const struct ib_send_wr *wr)
 	return container_of(wr, struct mlx5_umr_wr, wr);
 }
 
-struct mlx5_shared_mr_info {
-	int mr_id;
-	struct ib_umem		*umem;
-};
-
 enum mlx5_ib_cq_pr_flags {
 	MLX5_IB_CQ_PR_FLAGS_CQE_128_PAD	= 1 << 0,
 };
@@ -623,20 +619,6 @@ struct mlx5_user_mmap_entry {
 	u32 page_idx;
 };
 
-struct mlx5_ib_dm {
-	struct ib_dm		ibdm;
-	phys_addr_t		dev_addr;
-	u32			type;
-	size_t			size;
-	union {
-		struct {
-			u32	obj_id;
-		} icm_dm;
-		/* other dm types specific params should be added here */
-	};
-	struct mlx5_user_mmap_entry mentry;
-};
-
 #define MLX5_IB_MTT_PRESENT (MLX5_IB_MTT_READ | MLX5_IB_MTT_WRITE)
 
 #define MLX5_IB_DM_MEMIC_ALLOWED_ACCESS (IB_ACCESS_LOCAL_WRITE   |\
@@ -654,47 +636,69 @@ struct mlx5_ib_dm {
 	atomic64_add(value, &((mr)->odp_stats.counter_name))
 
 struct mlx5_ib_mr {
-	struct ib_mr		ibmr;
-	void			*descs;
-	dma_addr_t		desc_map;
-	int			ndescs;
-	int			data_length;
-	int			meta_ndescs;
-	int			meta_length;
-	int			max_descs;
-	int			desc_size;
-	int			access_mode;
-	unsigned int		page_shift;
-	struct mlx5_core_mkey	mmkey;
-	struct ib_umem	       *umem;
-	struct mlx5_shared_mr_info	*smr_info;
-	struct list_head	list;
-	struct mlx5_cache_ent  *cache_ent;
-	u32 out[MLX5_ST_SZ_DW(create_mkey_out)];
-	struct mlx5_core_sig_ctx    *sig;
-	void			*descs_alloc;
-	int			access_flags; /* Needed for rereg MR */
-
-	struct mlx5_ib_mr      *parent;
-	/* Needed for IB_MR_TYPE_INTEGRITY */
-	struct mlx5_ib_mr      *pi_mr;
-	struct mlx5_ib_mr      *klm_mr;
-	struct mlx5_ib_mr      *mtt_mr;
-	u64			data_iova;
-	u64			pi_iova;
-
-	/* For ODP and implicit */
-	struct xarray		implicit_children;
+	struct ib_mr ibmr;
+	struct mlx5_core_mkey mmkey;
+
+	/* User MR data */
+	struct mlx5_cache_ent *cache_ent;
+	struct ib_umem *umem;
+
+	/* This is zero'd when the MR is allocated */
 	union {
-		struct list_head elm;
-		struct work_struct work;
-	} odp_destroy;
-	struct ib_odp_counters	odp_stats;
-	bool			is_odp_implicit;
+		/* Used only while the MR is in the cache */
+		struct {
+			u32 out[MLX5_ST_SZ_DW(create_mkey_out)];
+			struct mlx5_async_work cb_work;
+			/* Cache list element */
+			struct list_head list;
+		};
 
-	struct mlx5_async_work  cb_work;
+		/* Used only by kernel MRs (umem == NULL) */
+		struct {
+			void *descs;
+			void *descs_alloc;
+			dma_addr_t desc_map;
+			int max_descs;
+			int ndescs;
+			int desc_size;
+			int access_mode;
+
+			/* For Kernel IB_MR_TYPE_INTEGRITY */
+			struct mlx5_core_sig_ctx *sig;
+			struct mlx5_ib_mr *pi_mr;
+			struct mlx5_ib_mr *klm_mr;
+			struct mlx5_ib_mr *mtt_mr;
+			u64 data_iova;
+			u64 pi_iova;
+			int meta_ndescs;
+			int meta_length;
+			int data_length;
+		};
+
+		/* Used only by User MRs (umem != NULL) */
+		struct {
+			unsigned int page_shift;
+			/* Current access_flags */
+			int access_flags;
+
+			/* For User ODP */
+			struct mlx5_ib_mr *parent;
+			struct xarray implicit_children;
+			union {
+				struct work_struct work;
+			} odp_destroy;
+			struct ib_odp_counters odp_stats;
+			bool is_odp_implicit;
+		};
+	};
 };
 
+/* Zero the fields in the mr that are variant depending on usage */
+static inline void mlx5_clear_mr(struct mlx5_ib_mr *mr)
+{
+	memset(mr->out, 0, sizeof(*mr) - offsetof(struct mlx5_ib_mr, out));
+}
+
 static inline bool is_odp_mr(struct mlx5_ib_mr *mr)
 {
 	return IS_ENABLED(CONFIG_INFINIBAND_ON_DEMAND_PAGING) && mr->umem &&
@@ -822,7 +826,7 @@ struct mlx5_roce {
 	atomic_t		tx_port_affinity;
 	enum ib_port_state last_port_state;
 	struct mlx5_ib_dev	*dev;
-	u8			native_port_num;
+	u32			native_port_num;
 };
 
 struct mlx5_ib_port {
@@ -837,7 +841,7 @@ struct mlx5_ib_dbg_param {
 	int			offset;
 	struct mlx5_ib_dev	*dev;
 	struct dentry		*dentry;
-	u8			port_num;
+	u32			port_num;
 };
 
 enum mlx5_ib_dbg_cc_types {
@@ -1063,6 +1067,7 @@ struct mlx5_ib_dev {
 	struct mutex			slow_path_mutex;
 	struct ib_odp_caps	odp_caps;
 	u64			odp_max_size;
+	struct mutex		odp_eq_mutex;
 	struct mlx5_ib_pf_eq	odp_pf_eq;
 
 	struct xarray		odp_mkeys;
@@ -1170,11 +1175,6 @@ static inline struct mlx5_ib_srq *to_mibsrq(struct mlx5_core_srq *msrq)
 	return container_of(msrq, struct mlx5_ib_srq, msrq);
 }
 
-static inline struct mlx5_ib_dm *to_mdm(struct ib_dm *ibdm)
-{
-	return container_of(ibdm, struct mlx5_ib_dm, ibdm);
-}
-
 static inline struct mlx5_ib_mr *to_mmr(struct ib_mr *ibmr)
 {
 	return container_of(ibmr, struct mlx5_ib_mr, ibmr);
@@ -1268,8 +1268,7 @@ struct mlx5_ib_mr *mlx5_ib_alloc_implicit_mr(struct mlx5_ib_pd *pd,
 					     struct ib_udata *udata,
 					     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);
-void mlx5_ib_fence_dmabuf_mr(struct mlx5_ib_mr *mr);
+void mlx5_ib_free_odp_mr(struct mlx5_ib_mr *mr);
 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);
@@ -1285,7 +1284,7 @@ int mlx5_ib_map_mr_sg_pi(struct ib_mr *ibmr, struct scatterlist *data_sg,
 			 int data_sg_nents, unsigned int *data_sg_offset,
 			 struct scatterlist *meta_sg, int meta_sg_nents,
 			 unsigned int *meta_sg_offset);
-int mlx5_ib_process_mad(struct ib_device *ibdev, int mad_flags, u8 port_num,
+int mlx5_ib_process_mad(struct ib_device *ibdev, int mad_flags, u32 port_num,
 			const struct ib_wc *in_wc, const struct ib_grh *in_grh,
 			const struct ib_mad *in, struct ib_mad *out,
 			size_t *out_mad_size, u16 *out_mad_pkey_index);
@@ -1300,13 +1299,13 @@ int mlx5_query_mad_ifc_vendor_id(struct ib_device *ibdev,
 				 u32 *vendor_id);
 int mlx5_query_mad_ifc_node_desc(struct mlx5_ib_dev *dev, char *node_desc);
 int mlx5_query_mad_ifc_node_guid(struct mlx5_ib_dev *dev, __be64 *node_guid);
-int mlx5_query_mad_ifc_pkey(struct ib_device *ibdev, u8 port, u16 index,
+int mlx5_query_mad_ifc_pkey(struct ib_device *ibdev, u32 port, u16 index,
 			    u16 *pkey);
-int mlx5_query_mad_ifc_gids(struct ib_device *ibdev, u8 port, int index,
+int mlx5_query_mad_ifc_gids(struct ib_device *ibdev, u32 port, int index,
 			    union ib_gid *gid);
-int mlx5_query_mad_ifc_port(struct ib_device *ibdev, u8 port,
+int mlx5_query_mad_ifc_port(struct ib_device *ibdev, u32 port,
 			    struct ib_port_attr *props);
-int mlx5_ib_query_port(struct ib_device *ibdev, u8 port,
+int mlx5_ib_query_port(struct ib_device *ibdev, u32 port,
 		       struct ib_port_attr *props);
 void mlx5_ib_populate_pas(struct ib_umem *umem, size_t page_size, __be64 *pas,
 			  u64 access_flags);
@@ -1317,8 +1316,6 @@ int mlx5_mr_cache_cleanup(struct mlx5_ib_dev *dev);
 
 struct mlx5_ib_mr *mlx5_mr_cache_alloc(struct mlx5_ib_dev *dev,
 				       unsigned int entry, int access_flags);
-void mlx5_mr_cache_free(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr);
-int mlx5_mr_cache_invalidate(struct mlx5_ib_mr *mr);
 
 int mlx5_ib_check_mr_status(struct ib_mr *ibmr, u32 check_mask,
 			    struct ib_mr_status *mr_status);
@@ -1332,18 +1329,13 @@ int mlx5_ib_create_rwq_ind_table(struct ib_rwq_ind_table *ib_rwq_ind_table,
 				 struct ib_rwq_ind_table_init_attr *init_attr,
 				 struct ib_udata *udata);
 int mlx5_ib_destroy_rwq_ind_table(struct ib_rwq_ind_table *wq_ind_table);
-struct ib_dm *mlx5_ib_alloc_dm(struct ib_device *ibdev,
-			       struct ib_ucontext *context,
-			       struct ib_dm_alloc_attr *attr,
-			       struct uverbs_attr_bundle *attrs);
-int mlx5_ib_dealloc_dm(struct ib_dm *ibdm, struct uverbs_attr_bundle *attrs);
 struct ib_mr *mlx5_ib_reg_dm_mr(struct ib_pd *pd, struct ib_dm *dm,
 				struct ib_dm_mr_attr *attr,
 				struct uverbs_attr_bundle *attrs);
 
 #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING
-void mlx5_ib_internal_fill_odp_caps(struct mlx5_ib_dev *dev);
 int mlx5_ib_odp_init_one(struct mlx5_ib_dev *ibdev);
+int mlx5r_odp_create_eq(struct mlx5_ib_dev *dev, struct mlx5_ib_pf_eq *eq);
 void mlx5_ib_odp_cleanup_one(struct mlx5_ib_dev *ibdev);
 int __init mlx5_ib_odp_init(void);
 void mlx5_ib_odp_cleanup(void);
@@ -1357,12 +1349,12 @@ int mlx5_ib_advise_mr_prefetch(struct ib_pd *pd,
 int mlx5_ib_init_odp_mr(struct mlx5_ib_mr *mr);
 int mlx5_ib_init_dmabuf_mr(struct mlx5_ib_mr *mr);
 #else /* CONFIG_INFINIBAND_ON_DEMAND_PAGING */
-static inline void mlx5_ib_internal_fill_odp_caps(struct mlx5_ib_dev *dev)
+static inline int mlx5_ib_odp_init_one(struct mlx5_ib_dev *ibdev) { return 0; }
+static inline int mlx5r_odp_create_eq(struct mlx5_ib_dev *dev,
+				      struct mlx5_ib_pf_eq *eq)
 {
-	return;
+	return 0;
 }
-
-static inline int mlx5_ib_odp_init_one(struct mlx5_ib_dev *ibdev) { return 0; }
 static inline void mlx5_ib_odp_cleanup_one(struct mlx5_ib_dev *ibdev) {}
 static inline int mlx5_ib_odp_init(void) { return 0; }
 static inline void mlx5_ib_odp_cleanup(void)				    {}
@@ -1397,22 +1389,22 @@ int __mlx5_ib_add(struct mlx5_ib_dev *dev,
 		  const struct mlx5_ib_profile *profile);
 
 int mlx5_ib_get_vf_config(struct ib_device *device, int vf,
-			  u8 port, struct ifla_vf_info *info);
+			  u32 port, struct ifla_vf_info *info);
 int mlx5_ib_set_vf_link_state(struct ib_device *device, int vf,
-			      u8 port, int state);
+			      u32 port, int state);
 int mlx5_ib_get_vf_stats(struct ib_device *device, int vf,
-			 u8 port, struct ifla_vf_stats *stats);
-int mlx5_ib_get_vf_guid(struct ib_device *device, int vf, u8 port,
+			 u32 port, struct ifla_vf_stats *stats);
+int mlx5_ib_get_vf_guid(struct ib_device *device, int vf, u32 port,
 			struct ifla_vf_guid *node_guid,
 			struct ifla_vf_guid *port_guid);
-int mlx5_ib_set_vf_guid(struct ib_device *device, int vf, u8 port,
+int mlx5_ib_set_vf_guid(struct ib_device *device, int vf, u32 port,
 			u64 guid, int type);
 
 __be16 mlx5_get_roce_udp_sport_min(const struct mlx5_ib_dev *dev,
 				   const struct ib_gid_attr *attr);
 
-void mlx5_ib_cleanup_cong_debugfs(struct mlx5_ib_dev *dev, u8 port_num);
-void mlx5_ib_init_cong_debugfs(struct mlx5_ib_dev *dev, u8 port_num);
+void mlx5_ib_cleanup_cong_debugfs(struct mlx5_ib_dev *dev, u32 port_num);
+void mlx5_ib_init_cong_debugfs(struct mlx5_ib_dev *dev, u32 port_num);
 
 /* GSI QP helper functions */
 int mlx5_ib_create_gsi(struct ib_pd *pd, struct mlx5_ib_qp *mqp,
@@ -1435,10 +1427,10 @@ void mlx5_ib_free_bfreg(struct mlx5_ib_dev *dev, struct mlx5_bfreg_info *bfregi,
 			int bfregn);
 struct mlx5_ib_dev *mlx5_ib_get_ibdev_from_mpi(struct mlx5_ib_multiport_info *mpi);
 struct mlx5_core_dev *mlx5_ib_get_native_port_mdev(struct mlx5_ib_dev *dev,
-						   u8 ib_port_num,
-						   u8 *native_port_num);
+						   u32 ib_port_num,
+						   u32 *native_port_num);
 void mlx5_ib_put_native_port_mdev(struct mlx5_ib_dev *dev,
-				  u8 port_num);
+				  u32 port_num);
 
 extern const struct uapi_definition mlx5_ib_devx_defs[];
 extern const struct uapi_definition mlx5_ib_flow_defs[];
diff --git a/drivers/infiniband/hw/mlx5/mr.c b/drivers/infiniband/hw/mlx5/mr.c
index db05b0e0a8d7..4388afeff251 100644
--- a/drivers/infiniband/hw/mlx5/mr.c
+++ b/drivers/infiniband/hw/mlx5/mr.c
@@ -42,6 +42,7 @@
 #include <rdma/ib_umem.h>
 #include <rdma/ib_umem_odp.h>
 #include <rdma/ib_verbs.h>
+#include "dm.h"
 #include "mlx5_ib.h"
 
 /*
@@ -119,8 +120,6 @@ mlx5_ib_create_mkey_cb(struct mlx5_ib_dev *dev,
 				create_mkey_callback, context);
 }
 
-static void clean_mr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr);
-static void dereg_mr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr);
 static int mr_cache_max_order(struct mlx5_ib_dev *dev);
 static void queue_adjust_cache_locked(struct mlx5_cache_ent *ent);
 
@@ -590,6 +589,8 @@ struct mlx5_ib_mr *mlx5_mr_cache_alloc(struct mlx5_ib_dev *dev,
 		ent->available_mrs--;
 		queue_adjust_cache_locked(ent);
 		spin_unlock_irq(&ent->lock);
+
+		mlx5_clear_mr(mr);
 	}
 	mr->access_flags = access_flags;
 	return mr;
@@ -615,42 +616,20 @@ static struct mlx5_ib_mr *get_cache_mr(struct mlx5_cache_ent *req_ent)
 			ent->available_mrs--;
 			queue_adjust_cache_locked(ent);
 			spin_unlock_irq(&ent->lock);
-			break;
+			mlx5_clear_mr(mr);
+			return mr;
 		}
 		queue_adjust_cache_locked(ent);
 		spin_unlock_irq(&ent->lock);
 	}
-
-	if (!mr)
-		req_ent->miss++;
-
-	return mr;
-}
-
-static void detach_mr_from_cache(struct mlx5_ib_mr *mr)
-{
-	struct mlx5_cache_ent *ent = mr->cache_ent;
-
-	mr->cache_ent = NULL;
-	spin_lock_irq(&ent->lock);
-	ent->total_mrs--;
-	spin_unlock_irq(&ent->lock);
+	req_ent->miss++;
+	return NULL;
 }
 
-void mlx5_mr_cache_free(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr)
+static void mlx5_mr_cache_free(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr)
 {
 	struct mlx5_cache_ent *ent = mr->cache_ent;
 
-	if (!ent)
-		return;
-
-	if (mlx5_mr_cache_invalidate(mr)) {
-		detach_mr_from_cache(mr);
-		destroy_mkey(dev, mr);
-		kfree(mr);
-		return;
-	}
-
 	spin_lock_irq(&ent->lock);
 	list_add_tail(&mr->list, &ent->head);
 	ent->available_mrs++;
@@ -993,8 +972,6 @@ static struct mlx5_ib_mr *alloc_cacheable_mr(struct ib_pd *pd,
 
 	mr->ibmr.pd = pd;
 	mr->umem = umem;
-	mr->access_flags = access_flags;
-	mr->desc_size = sizeof(struct mlx5_mtt);
 	mr->mmkey.iova = iova;
 	mr->mmkey.size = umem->length;
 	mr->mmkey.pd = to_mpd(pd)->pdn;
@@ -1028,7 +1005,7 @@ static void *mlx5_ib_alloc_xlt(size_t *nents, size_t ent_size, gfp_t gfp_mask)
 	 */
 	might_sleep();
 
-	gfp_mask |= __GFP_ZERO;
+	gfp_mask |= __GFP_ZERO | __GFP_NORETRY;
 
 	/*
 	 * If the system already has a suitable high order page then just use
@@ -1505,7 +1482,7 @@ static struct ib_mr *create_real_mr(struct ib_pd *pd, struct ib_umem *umem,
 		 */
 		err = mlx5_ib_update_mr_pas(mr, MLX5_IB_UPD_XLT_ENABLE);
 		if (err) {
-			dereg_mr(dev, mr);
+			mlx5_ib_dereg_mr(&mr->ibmr, NULL);
 			return ERR_PTR(err);
 		}
 	}
@@ -1524,6 +1501,9 @@ static struct ib_mr *create_user_odp_mr(struct ib_pd *pd, u64 start, u64 length,
 	if (!IS_ENABLED(CONFIG_INFINIBAND_ON_DEMAND_PAGING))
 		return ERR_PTR(-EOPNOTSUPP);
 
+	err = mlx5r_odp_create_eq(dev, &dev->odp_pf_eq);
+	if (err)
+		return ERR_PTR(err);
 	if (!start && length == U64_MAX) {
 		if (iova != 0)
 			return ERR_PTR(-EINVAL);
@@ -1562,7 +1542,7 @@ static struct ib_mr *create_user_odp_mr(struct ib_pd *pd, u64 start, u64 length,
 	return &mr->ibmr;
 
 err_dereg_mr:
-	dereg_mr(dev, mr);
+	mlx5_ib_dereg_mr(&mr->ibmr, NULL);
 	return ERR_PTR(err);
 }
 
@@ -1659,19 +1639,19 @@ struct ib_mr *mlx5_ib_reg_user_mr_dmabuf(struct ib_pd *pd, u64 offset,
 	return &mr->ibmr;
 
 err_dereg_mr:
-	dereg_mr(dev, mr);
+	mlx5_ib_dereg_mr(&mr->ibmr, NULL);
 	return ERR_PTR(err);
 }
 
 /**
- * mlx5_mr_cache_invalidate - Fence all DMA on the MR
+ * revoke_mr - Fence all DMA on the MR
  * @mr: The MR to fence
  *
  * Upon return the NIC will not be doing any DMA to the pages under the MR,
- * and any DMA inprogress will be completed. Failure of this function
+ * and any DMA in progress will be completed. Failure of this function
  * indicates the HW has failed catastrophically.
  */
-int mlx5_mr_cache_invalidate(struct mlx5_ib_mr *mr)
+static int revoke_mr(struct mlx5_ib_mr *mr)
 {
 	struct mlx5_umr_wr umrwr = {};
 
@@ -1765,7 +1745,7 @@ static int umr_rereg_pas(struct mlx5_ib_mr *mr, struct ib_pd *pd,
 	 * with it. This ensure the change is atomic relative to any use of the
 	 * MR.
 	 */
-	err = mlx5_mr_cache_invalidate(mr);
+	err = revoke_mr(mr);
 	if (err)
 		return err;
 
@@ -1844,7 +1824,7 @@ struct ib_mr *mlx5_ib_rereg_user_mr(struct ib_mr *ib_mr, int flags, u64 start,
 		 * Only one active MR can refer to a umem at one time, revoke
 		 * the old MR before assigning the umem to the new one.
 		 */
-		err = mlx5_mr_cache_invalidate(mr);
+		err = revoke_mr(mr);
 		if (err)
 			return ERR_PTR(err);
 		umem = mr->umem;
@@ -1931,7 +1911,7 @@ err:
 static void
 mlx5_free_priv_descs(struct mlx5_ib_mr *mr)
 {
-	if (mr->descs) {
+	if (!mr->umem && mr->descs) {
 		struct ib_device *device = mr->ibmr.device;
 		int size = mr->max_descs * mr->desc_size;
 		struct mlx5_ib_dev *dev = to_mdev(device);
@@ -1943,69 +1923,82 @@ mlx5_free_priv_descs(struct mlx5_ib_mr *mr)
 	}
 }
 
-static void clean_mr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr)
+int mlx5_ib_dereg_mr(struct ib_mr *ibmr, struct ib_udata *udata)
 {
-	if (mr->sig) {
+	struct mlx5_ib_mr *mr = to_mmr(ibmr);
+	struct mlx5_ib_dev *dev = to_mdev(ibmr->device);
+	int rc;
+
+	/*
+	 * Any async use of the mr must hold the refcount, once the refcount
+	 * goes to zero no other thread, such as ODP page faults, prefetch, any
+	 * UMR activity, etc can touch the mkey. Thus it is safe to destroy it.
+	 */
+	if (IS_ENABLED(CONFIG_INFINIBAND_ON_DEMAND_PAGING) &&
+	    refcount_read(&mr->mmkey.usecount) != 0 &&
+	    xa_erase(&mr_to_mdev(mr)->odp_mkeys, mlx5_base_mkey(mr->mmkey.key)))
+		mlx5r_deref_wait_odp_mkey(&mr->mmkey);
+
+	if (ibmr->type == IB_MR_TYPE_INTEGRITY) {
+		xa_cmpxchg(&dev->sig_mrs, mlx5_base_mkey(mr->mmkey.key), ibmr,
+			   NULL, GFP_KERNEL);
+
+		if (mr->mtt_mr) {
+			rc = mlx5_ib_dereg_mr(&mr->mtt_mr->ibmr, NULL);
+			if (rc)
+				return rc;
+			mr->mtt_mr = NULL;
+		}
+		if (mr->klm_mr) {
+			rc = mlx5_ib_dereg_mr(&mr->klm_mr->ibmr, NULL);
+			if (rc)
+				return rc;
+			mr->klm_mr = NULL;
+		}
+
 		if (mlx5_core_destroy_psv(dev->mdev,
 					  mr->sig->psv_memory.psv_idx))
 			mlx5_ib_warn(dev, "failed to destroy mem psv %d\n",
 				     mr->sig->psv_memory.psv_idx);
-		if (mlx5_core_destroy_psv(dev->mdev,
-					  mr->sig->psv_wire.psv_idx))
+		if (mlx5_core_destroy_psv(dev->mdev, mr->sig->psv_wire.psv_idx))
 			mlx5_ib_warn(dev, "failed to destroy wire psv %d\n",
 				     mr->sig->psv_wire.psv_idx);
-		xa_erase(&dev->sig_mrs, mlx5_base_mkey(mr->mmkey.key));
 		kfree(mr->sig);
 		mr->sig = NULL;
 	}
 
+	/* Stop DMA */
+	if (mr->cache_ent) {
+		if (revoke_mr(mr)) {
+			spin_lock_irq(&mr->cache_ent->lock);
+			mr->cache_ent->total_mrs--;
+			spin_unlock_irq(&mr->cache_ent->lock);
+			mr->cache_ent = NULL;
+		}
+	}
 	if (!mr->cache_ent) {
-		destroy_mkey(dev, mr);
-		mlx5_free_priv_descs(mr);
+		rc = destroy_mkey(to_mdev(mr->ibmr.device), mr);
+		if (rc)
+			return rc;
 	}
-}
 
-static void dereg_mr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr)
-{
-	struct ib_umem *umem = mr->umem;
+	if (mr->umem) {
+		bool is_odp = is_odp_mr(mr);
 
-	/* Stop all DMA */
-	if (is_odp_mr(mr))
-		mlx5_ib_fence_odp_mr(mr);
-	else if (is_dmabuf_mr(mr))
-		mlx5_ib_fence_dmabuf_mr(mr);
-	else
-		clean_mr(dev, mr);
-
-	if (umem) {
-		if (!is_odp_mr(mr))
-			atomic_sub(ib_umem_num_pages(umem),
+		if (!is_odp)
+			atomic_sub(ib_umem_num_pages(mr->umem),
 				   &dev->mdev->priv.reg_pages);
-		ib_umem_release(umem);
+		ib_umem_release(mr->umem);
+		if (is_odp)
+			mlx5_ib_free_odp_mr(mr);
 	}
 
-	if (mr->cache_ent)
+	if (mr->cache_ent) {
 		mlx5_mr_cache_free(dev, mr);
-	else
+	} else {
+		mlx5_free_priv_descs(mr);
 		kfree(mr);
-}
-
-int mlx5_ib_dereg_mr(struct ib_mr *ibmr, struct ib_udata *udata)
-{
-	struct mlx5_ib_mr *mmr = to_mmr(ibmr);
-
-	if (ibmr->type == IB_MR_TYPE_INTEGRITY) {
-		dereg_mr(to_mdev(mmr->mtt_mr->ibmr.device), mmr->mtt_mr);
-		dereg_mr(to_mdev(mmr->klm_mr->ibmr.device), mmr->klm_mr);
 	}
-
-	if (is_odp_mr(mmr) && to_ib_umem_odp(mmr->umem)->is_implicit_odp) {
-		mlx5_ib_free_implicit_mr(mmr);
-		return 0;
-	}
-
-	dereg_mr(to_mdev(ibmr->device), mmr);
-
 	return 0;
 }
 
@@ -2177,10 +2170,10 @@ err_free_descs:
 	destroy_mkey(dev, mr);
 	mlx5_free_priv_descs(mr);
 err_free_mtt_mr:
-	dereg_mr(to_mdev(mr->mtt_mr->ibmr.device), mr->mtt_mr);
+	mlx5_ib_dereg_mr(&mr->mtt_mr->ibmr, NULL);
 	mr->mtt_mr = NULL;
 err_free_klm_mr:
-	dereg_mr(to_mdev(mr->klm_mr->ibmr.device), mr->klm_mr);
+	mlx5_ib_dereg_mr(&mr->klm_mr->ibmr, NULL);
 	mr->klm_mr = NULL;
 err_destroy_psv:
 	if (mlx5_core_destroy_psv(dev->mdev, mr->sig->psv_memory.psv_idx))
diff --git a/drivers/infiniband/hw/mlx5/odp.c b/drivers/infiniband/hw/mlx5/odp.c
index b103555b1f5d..782b2af8f211 100644
--- a/drivers/infiniband/hw/mlx5/odp.c
+++ b/drivers/infiniband/hw/mlx5/odp.c
@@ -181,64 +181,29 @@ void mlx5_odp_populate_xlt(void *xlt, size_t idx, size_t nentries,
 	}
 }
 
-static void dma_fence_odp_mr(struct mlx5_ib_mr *mr)
-{
-	struct ib_umem_odp *odp = to_ib_umem_odp(mr->umem);
-
-	/* Ensure mlx5_ib_invalidate_range() will not touch the MR any more */
-	mutex_lock(&odp->umem_mutex);
-	if (odp->npages) {
-		mlx5_mr_cache_invalidate(mr);
-		ib_umem_odp_unmap_dma_pages(odp, ib_umem_start(odp),
-					    ib_umem_end(odp));
-		WARN_ON(odp->npages);
-	}
-	odp->private = NULL;
-	mutex_unlock(&odp->umem_mutex);
-
-	if (!mr->cache_ent) {
-		mlx5_core_destroy_mkey(mr_to_mdev(mr)->mdev, &mr->mmkey);
-		WARN_ON(mr->descs);
-	}
-}
-
 /*
  * This must be called after the mr has been removed from implicit_children.
  * NOTE: The MR does not necessarily have to be
  * empty here, parallel page faults could have raced with the free process and
  * added pages to it.
  */
-static void free_implicit_child_mr(struct mlx5_ib_mr *mr, bool need_imr_xlt)
+static void free_implicit_child_mr_work(struct work_struct *work)
 {
+	struct mlx5_ib_mr *mr =
+		container_of(work, struct mlx5_ib_mr, odp_destroy.work);
 	struct mlx5_ib_mr *imr = mr->parent;
 	struct ib_umem_odp *odp_imr = to_ib_umem_odp(imr->umem);
 	struct ib_umem_odp *odp = to_ib_umem_odp(mr->umem);
-	unsigned long idx = ib_umem_start(odp) >> MLX5_IMR_MTT_SHIFT;
 
 	mlx5r_deref_wait_odp_mkey(&mr->mmkey);
 
-	if (need_imr_xlt) {
-		mutex_lock(&odp_imr->umem_mutex);
-		mlx5_ib_update_xlt(mr->parent, idx, 1, 0,
-				   MLX5_IB_UPD_XLT_INDIRECT |
-				   MLX5_IB_UPD_XLT_ATOMIC);
-		mutex_unlock(&odp_imr->umem_mutex);
-	}
-
-	dma_fence_odp_mr(mr);
-
-	mr->parent = NULL;
-	mlx5_mr_cache_free(mr_to_mdev(mr), mr);
-	ib_umem_odp_release(odp);
-}
-
-static void free_implicit_child_mr_work(struct work_struct *work)
-{
-	struct mlx5_ib_mr *mr =
-		container_of(work, struct mlx5_ib_mr, odp_destroy.work);
-	struct mlx5_ib_mr *imr = mr->parent;
+	mutex_lock(&odp_imr->umem_mutex);
+	mlx5_ib_update_xlt(mr->parent, ib_umem_start(odp) >> MLX5_IMR_MTT_SHIFT,
+			   1, 0,
+			   MLX5_IB_UPD_XLT_INDIRECT | MLX5_IB_UPD_XLT_ATOMIC);
+	mutex_unlock(&odp_imr->umem_mutex);
+	mlx5_ib_dereg_mr(&mr->ibmr, NULL);
 
-	free_implicit_child_mr(mr, true);
 	mlx5r_deref_odp_mkey(&imr->mmkey);
 }
 
@@ -352,7 +317,7 @@ const struct mmu_interval_notifier_ops mlx5_mn_ops = {
 	.invalidate = mlx5_ib_invalidate_range,
 };
 
-void mlx5_ib_internal_fill_odp_caps(struct mlx5_ib_dev *dev)
+static void internal_fill_odp_caps(struct mlx5_ib_dev *dev)
 {
 	struct ib_odp_caps *caps = &dev->odp_caps;
 
@@ -455,8 +420,10 @@ static struct mlx5_ib_mr *implicit_get_child_mr(struct mlx5_ib_mr *imr,
 
 	ret = mr = mlx5_mr_cache_alloc(
 		mr_to_mdev(imr), MLX5_IMR_MTT_CACHE_ENTRY, imr->access_flags);
-	if (IS_ERR(mr))
-		goto out_umem;
+	if (IS_ERR(mr)) {
+		ib_umem_odp_release(odp);
+		return mr;
+	}
 
 	mr->ibmr.pd = imr->ibmr.pd;
 	mr->ibmr.device = &mr_to_mdev(imr)->ib_dev;
@@ -506,9 +473,7 @@ static struct mlx5_ib_mr *implicit_get_child_mr(struct mlx5_ib_mr *imr,
 out_lock:
 	xa_unlock(&imr->implicit_children);
 out_mr:
-	mlx5_mr_cache_free(mr_to_mdev(imr), mr);
-out_umem:
-	ib_umem_odp_release(odp);
+	mlx5_ib_dereg_mr(&mr->ibmr, NULL);
 	return ret;
 }
 
@@ -531,8 +496,8 @@ struct mlx5_ib_mr *mlx5_ib_alloc_implicit_mr(struct mlx5_ib_pd *pd,
 
 	imr = mlx5_mr_cache_alloc(dev, MLX5_IMR_KSM_CACHE_ENTRY, access_flags);
 	if (IS_ERR(imr)) {
-		err = PTR_ERR(imr);
-		goto out_umem;
+		ib_umem_odp_release(umem_odp);
+		return imr;
 	}
 
 	imr->ibmr.pd = &pd->ibpd;
@@ -562,93 +527,22 @@ struct mlx5_ib_mr *mlx5_ib_alloc_implicit_mr(struct mlx5_ib_pd *pd,
 	return imr;
 out_mr:
 	mlx5_ib_err(dev, "Failed to register MKEY %d\n", err);
-	mlx5_mr_cache_free(dev, imr);
-out_umem:
-	ib_umem_odp_release(umem_odp);
+	mlx5_ib_dereg_mr(&imr->ibmr, NULL);
 	return ERR_PTR(err);
 }
 
-void mlx5_ib_free_implicit_mr(struct mlx5_ib_mr *imr)
+void mlx5_ib_free_odp_mr(struct mlx5_ib_mr *mr)
 {
-	struct ib_umem_odp *odp_imr = to_ib_umem_odp(imr->umem);
-	struct mlx5_ib_dev *dev = mr_to_mdev(imr);
 	struct mlx5_ib_mr *mtt;
 	unsigned long idx;
 
-	xa_erase(&dev->odp_mkeys, mlx5_base_mkey(imr->mmkey.key));
-	/*
-	 * All work on the prefetch list must be completed, xa_erase() prevented
-	 * new work from being created.
-	 */
-	mlx5r_deref_wait_odp_mkey(&imr->mmkey);
-	/*
-	 * At this point it is forbidden for any other thread to enter
-	 * pagefault_mr() on this imr. It is already forbidden to call
-	 * pagefault_mr() on an implicit child. Due to this additions to
-	 * implicit_children are prevented.
-	 * In addition, any new call to destroy_unused_implicit_child_mr()
-	 * may return immediately.
-	 */
-
 	/*
-	 * Fence the imr before we destroy the children. This allows us to
-	 * skip updating the XLT of the imr during destroy of the child mkey
-	 * the imr points to.
+	 * If this is an implicit MR it is already invalidated so we can just
+	 * delete the children mkeys.
 	 */
-	mlx5_mr_cache_invalidate(imr);
-
-	xa_for_each(&imr->implicit_children, idx, mtt) {
-		xa_erase(&imr->implicit_children, idx);
-		free_implicit_child_mr(mtt, false);
-	}
-
-	mlx5_mr_cache_free(dev, imr);
-	ib_umem_odp_release(odp_imr);
-}
-
-/**
- * mlx5_ib_fence_odp_mr - Stop all access to the ODP MR
- * @mr: to fence
- *
- * On return no parallel threads will be touching this MR and no DMA will be
- * active.
- */
-void mlx5_ib_fence_odp_mr(struct mlx5_ib_mr *mr)
-{
-	/* Prevent new page faults and prefetch requests from succeeding */
-	xa_erase(&mr_to_mdev(mr)->odp_mkeys, mlx5_base_mkey(mr->mmkey.key));
-
-	/* Wait for all running page-fault handlers to finish. */
-	mlx5r_deref_wait_odp_mkey(&mr->mmkey);
-
-	dma_fence_odp_mr(mr);
-}
-
-/**
- * mlx5_ib_fence_dmabuf_mr - Stop all access to the dmabuf MR
- * @mr: to fence
- *
- * On return no parallel threads will be touching this MR and no DMA will be
- * active.
- */
-void mlx5_ib_fence_dmabuf_mr(struct mlx5_ib_mr *mr)
-{
-	struct ib_umem_dmabuf *umem_dmabuf = to_ib_umem_dmabuf(mr->umem);
-
-	/* Prevent new page faults and prefetch requests from succeeding */
-	xa_erase(&mr_to_mdev(mr)->odp_mkeys, mlx5_base_mkey(mr->mmkey.key));
-
-	mlx5r_deref_wait_odp_mkey(&mr->mmkey);
-
-	dma_resv_lock(umem_dmabuf->attach->dmabuf->resv, NULL);
-	mlx5_mr_cache_invalidate(mr);
-	umem_dmabuf->private = NULL;
-	ib_umem_dmabuf_unmap_pages(umem_dmabuf);
-	dma_resv_unlock(umem_dmabuf->attach->dmabuf->resv);
-
-	if (!mr->cache_ent) {
-		mlx5_core_destroy_mkey(mr_to_mdev(mr)->mdev, &mr->mmkey);
-		WARN_ON(mr->descs);
+	xa_for_each(&mr->implicit_children, idx, mtt) {
+		xa_erase(&mr->implicit_children, idx);
+		mlx5_ib_dereg_mr(&mtt->ibmr, NULL);
 	}
 }
 
@@ -1637,20 +1531,24 @@ enum {
 	MLX5_IB_NUM_PF_DRAIN	= 64,
 };
 
-static int
-mlx5_ib_create_pf_eq(struct mlx5_ib_dev *dev, struct mlx5_ib_pf_eq *eq)
+int mlx5r_odp_create_eq(struct mlx5_ib_dev *dev, struct mlx5_ib_pf_eq *eq)
 {
 	struct mlx5_eq_param param = {};
-	int err;
+	int err = 0;
 
+	mutex_lock(&dev->odp_eq_mutex);
+	if (eq->core)
+		goto unlock;
 	INIT_WORK(&eq->work, mlx5_ib_eq_pf_action);
 	spin_lock_init(&eq->lock);
 	eq->dev = dev;
 
 	eq->pool = mempool_create_kmalloc_pool(MLX5_IB_NUM_PF_DRAIN,
 					       sizeof(struct mlx5_pagefault));
-	if (!eq->pool)
-		return -ENOMEM;
+	if (!eq->pool) {
+		err = -ENOMEM;
+		goto unlock;
+	}
 
 	eq->wq = alloc_workqueue("mlx5_ib_page_fault",
 				 WQ_HIGHPRI | WQ_UNBOUND | WQ_MEM_RECLAIM,
@@ -1661,7 +1559,7 @@ mlx5_ib_create_pf_eq(struct mlx5_ib_dev *dev, struct mlx5_ib_pf_eq *eq)
 	}
 
 	eq->irq_nb.notifier_call = mlx5_ib_eq_pf_int;
-	param = (struct mlx5_eq_param) {
+	param = (struct mlx5_eq_param){
 		.irq_index = 0,
 		.nent = MLX5_IB_NUM_PF_EQE,
 	};
@@ -1677,21 +1575,27 @@ mlx5_ib_create_pf_eq(struct mlx5_ib_dev *dev, struct mlx5_ib_pf_eq *eq)
 		goto err_eq;
 	}
 
+	mutex_unlock(&dev->odp_eq_mutex);
 	return 0;
 err_eq:
 	mlx5_eq_destroy_generic(dev->mdev, eq->core);
 err_wq:
+	eq->core = NULL;
 	destroy_workqueue(eq->wq);
 err_mempool:
 	mempool_destroy(eq->pool);
+unlock:
+	mutex_unlock(&dev->odp_eq_mutex);
 	return err;
 }
 
 static int
-mlx5_ib_destroy_pf_eq(struct mlx5_ib_dev *dev, struct mlx5_ib_pf_eq *eq)
+mlx5_ib_odp_destroy_eq(struct mlx5_ib_dev *dev, struct mlx5_ib_pf_eq *eq)
 {
 	int err;
 
+	if (!eq->core)
+		return 0;
 	mlx5_eq_disable(dev->mdev, eq->core, &eq->irq_nb);
 	err = mlx5_eq_destroy_generic(dev->mdev, eq->core);
 	cancel_work_sync(&eq->work);
@@ -1735,6 +1639,8 @@ int mlx5_ib_odp_init_one(struct mlx5_ib_dev *dev)
 {
 	int ret = 0;
 
+	internal_fill_odp_caps(dev);
+
 	if (!(dev->odp_caps.general_caps & IB_ODP_SUPPORT))
 		return ret;
 
@@ -1748,8 +1654,7 @@ int mlx5_ib_odp_init_one(struct mlx5_ib_dev *dev)
 		}
 	}
 
-	ret = mlx5_ib_create_pf_eq(dev, &dev->odp_pf_eq);
-
+	mutex_init(&dev->odp_eq_mutex);
 	return ret;
 }
 
@@ -1758,7 +1663,7 @@ void mlx5_ib_odp_cleanup_one(struct mlx5_ib_dev *dev)
 	if (!(dev->odp_caps.general_caps & IB_ODP_SUPPORT))
 		return;
 
-	mlx5_ib_destroy_pf_eq(dev, &dev->odp_pf_eq);
+	mlx5_ib_odp_destroy_eq(dev, &dev->odp_pf_eq);
 }
 
 int mlx5_ib_odp_init(void)
diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c
index f5a52a6fae43..9282eb10bfae 100644
--- a/drivers/infiniband/hw/mlx5/qp.c
+++ b/drivers/infiniband/hw/mlx5/qp.c
@@ -67,7 +67,7 @@ struct mlx5_modify_raw_qp_param {
 	struct mlx5_rate_limit rl;
 
 	u8 rq_q_ctr_id;
-	u16 port;
+	u32 port;
 };
 
 static void get_cqs(enum ib_qp_type qp_type,
@@ -3146,6 +3146,19 @@ enum {
 	MLX5_PATH_FLAG_COUNTER	= 1 << 2,
 };
 
+static int mlx5_to_ib_rate_map(u8 rate)
+{
+	static const int rates[] = { IB_RATE_PORT_CURRENT, IB_RATE_56_GBPS,
+				     IB_RATE_25_GBPS,	   IB_RATE_100_GBPS,
+				     IB_RATE_200_GBPS,	   IB_RATE_50_GBPS,
+				     IB_RATE_400_GBPS };
+
+	if (rate < ARRAY_SIZE(rates))
+		return rates[rate];
+
+	return rate - MLX5_STAT_RATE_OFFSET;
+}
+
 static int ib_to_mlx5_rate_map(u8 rate)
 {
 	switch (rate) {
@@ -4485,7 +4498,7 @@ static void to_rdma_ah_attr(struct mlx5_ib_dev *ibdev,
 	rdma_ah_set_path_bits(ah_attr, MLX5_GET(ads, path, mlid));
 
 	static_rate = MLX5_GET(ads, path, stat_rate);
-	rdma_ah_set_static_rate(ah_attr, static_rate ? static_rate - 5 : 0);
+	rdma_ah_set_static_rate(ah_attr, mlx5_to_ib_rate_map(static_rate));
 	if (MLX5_GET(ads, path, grh) ||
 	    ah_attr->type == RDMA_AH_ATTR_TYPE_ROCE) {
 		rdma_ah_set_grh(ah_attr, NULL, MLX5_GET(ads, path, flow_label),
diff --git a/drivers/infiniband/hw/mlx5/std_types.c b/drivers/infiniband/hw/mlx5/std_types.c
index 16145fda68d0..c0ddf7b3c6e2 100644
--- a/drivers/infiniband/hw/mlx5/std_types.c
+++ b/drivers/infiniband/hw/mlx5/std_types.c
@@ -7,6 +7,8 @@
 #include <rdma/mlx5_user_ioctl_cmds.h>
 #include <rdma/mlx5_user_ioctl_verbs.h>
 #include <linux/mlx5/driver.h>
+#include <linux/mlx5/eswitch.h>
+#include <linux/mlx5/vport.h>
 #include "mlx5_ib.h"
 
 #define UVERBS_MODULE_NAME mlx5_ib
@@ -23,6 +25,174 @@ static int UVERBS_HANDLER(MLX5_IB_METHOD_PD_QUERY)(
 			      &mpd->pdn, sizeof(mpd->pdn));
 }
 
+static int fill_vport_icm_addr(struct mlx5_core_dev *mdev, u16 vport,
+			       struct mlx5_ib_uapi_query_port *info)
+{
+	u32 out[MLX5_ST_SZ_DW(query_esw_vport_context_out)] = {};
+	u32 in[MLX5_ST_SZ_DW(query_esw_vport_context_in)] = {};
+	bool sw_owner_supp;
+	u64 icm_rx;
+	u64 icm_tx;
+	int err;
+
+	sw_owner_supp = MLX5_CAP_ESW_FLOWTABLE_FDB(mdev, sw_owner) ||
+			MLX5_CAP_ESW_FLOWTABLE_FDB(mdev, sw_owner_v2);
+
+	if (vport == MLX5_VPORT_UPLINK) {
+		icm_rx = MLX5_CAP64_ESW_FLOWTABLE(mdev,
+			sw_steering_uplink_icm_address_rx);
+		icm_tx = MLX5_CAP64_ESW_FLOWTABLE(mdev,
+			sw_steering_uplink_icm_address_tx);
+	} else {
+		MLX5_SET(query_esw_vport_context_in, in, opcode,
+			 MLX5_CMD_OP_QUERY_ESW_VPORT_CONTEXT);
+		MLX5_SET(query_esw_vport_context_in, in, vport_number, vport);
+		MLX5_SET(query_esw_vport_context_in, in, other_vport, true);
+
+		err = mlx5_cmd_exec_inout(mdev, query_esw_vport_context, in,
+					  out);
+
+		if (err)
+			return err;
+
+		icm_rx = MLX5_GET64(
+			query_esw_vport_context_out, out,
+			esw_vport_context.sw_steering_vport_icm_address_rx);
+
+		icm_tx = MLX5_GET64(
+			query_esw_vport_context_out, out,
+			esw_vport_context.sw_steering_vport_icm_address_tx);
+	}
+
+	if (sw_owner_supp && icm_rx) {
+		info->vport_steering_icm_rx = icm_rx;
+		info->flags |=
+			MLX5_IB_UAPI_QUERY_PORT_VPORT_STEERING_ICM_RX;
+	}
+
+	if (sw_owner_supp && icm_tx) {
+		info->vport_steering_icm_tx = icm_tx;
+		info->flags |=
+			MLX5_IB_UAPI_QUERY_PORT_VPORT_STEERING_ICM_TX;
+	}
+
+	return 0;
+}
+
+static int fill_vport_vhca_id(struct mlx5_core_dev *mdev, u16 vport,
+			      struct mlx5_ib_uapi_query_port *info)
+{
+	size_t out_sz = MLX5_ST_SZ_BYTES(query_hca_cap_out);
+	u32 in[MLX5_ST_SZ_DW(query_hca_cap_in)] = {};
+	void *out;
+	int err;
+
+	out = kzalloc(out_sz, GFP_KERNEL);
+	if (!out)
+		return -ENOMEM;
+
+	MLX5_SET(query_hca_cap_in, in, opcode, MLX5_CMD_OP_QUERY_HCA_CAP);
+	MLX5_SET(query_hca_cap_in, in, other_function, true);
+	MLX5_SET(query_hca_cap_in, in, function_id, vport);
+	MLX5_SET(query_hca_cap_in, in, op_mod,
+		 MLX5_SET_HCA_CAP_OP_MOD_GENERAL_DEVICE |
+		 HCA_CAP_OPMOD_GET_CUR);
+
+	err = mlx5_cmd_exec(mdev, in, sizeof(in), out, out_sz);
+	if (err)
+		goto out;
+
+	info->vport_vhca_id = MLX5_GET(query_hca_cap_out, out,
+				       capability.cmd_hca_cap.vhca_id);
+
+	info->flags |= MLX5_IB_UAPI_QUERY_PORT_VPORT_VHCA_ID;
+out:
+	kfree(out);
+	return err;
+}
+
+static int fill_switchdev_info(struct mlx5_ib_dev *dev, u32 port_num,
+			       struct mlx5_ib_uapi_query_port *info)
+{
+	struct mlx5_core_dev *mdev = dev->mdev;
+	struct mlx5_eswitch_rep *rep;
+	int err;
+
+	rep = dev->port[port_num - 1].rep;
+	if (!rep)
+		return -EOPNOTSUPP;
+
+	info->vport = rep->vport;
+	info->flags |= MLX5_IB_UAPI_QUERY_PORT_VPORT;
+
+	if (rep->vport != MLX5_VPORT_UPLINK) {
+		err = fill_vport_vhca_id(mdev, rep->vport, info);
+		if (err)
+			return err;
+	}
+
+	info->esw_owner_vhca_id = MLX5_CAP_GEN(mdev, vhca_id);
+	info->flags |= MLX5_IB_UAPI_QUERY_PORT_ESW_OWNER_VHCA_ID;
+
+	err = fill_vport_icm_addr(mdev, rep->vport, info);
+	if (err)
+		return err;
+
+	if (mlx5_eswitch_vport_match_metadata_enabled(mdev->priv.eswitch)) {
+		info->reg_c0.value = mlx5_eswitch_get_vport_metadata_for_match(
+			mdev->priv.eswitch, rep->vport);
+		info->reg_c0.mask = mlx5_eswitch_get_vport_metadata_mask();
+		info->flags |= MLX5_IB_UAPI_QUERY_PORT_VPORT_REG_C0;
+	}
+
+	return 0;
+}
+
+static int UVERBS_HANDLER(MLX5_IB_METHOD_QUERY_PORT)(
+	struct uverbs_attr_bundle *attrs)
+{
+	struct mlx5_ib_uapi_query_port info = {};
+	struct mlx5_ib_ucontext *c;
+	struct mlx5_ib_dev *dev;
+	u32 port_num;
+	int ret;
+
+	if (uverbs_copy_from(&port_num, attrs,
+			     MLX5_IB_ATTR_QUERY_PORT_PORT_NUM))
+		return -EFAULT;
+
+	c = to_mucontext(ib_uverbs_get_ucontext(attrs));
+	if (IS_ERR(c))
+		return PTR_ERR(c);
+	dev = to_mdev(c->ibucontext.device);
+
+	if (!rdma_is_port_valid(&dev->ib_dev, port_num))
+		return -EINVAL;
+
+	if (mlx5_eswitch_mode(dev->mdev) == MLX5_ESWITCH_OFFLOADS) {
+		ret = fill_switchdev_info(dev, port_num, &info);
+		if (ret)
+			return ret;
+	}
+
+	return uverbs_copy_to_struct_or_zero(attrs, MLX5_IB_ATTR_QUERY_PORT, &info,
+					     sizeof(info));
+}
+
+DECLARE_UVERBS_NAMED_METHOD(
+	MLX5_IB_METHOD_QUERY_PORT,
+	UVERBS_ATTR_PTR_IN(MLX5_IB_ATTR_QUERY_PORT_PORT_NUM,
+			   UVERBS_ATTR_TYPE(u32), UA_MANDATORY),
+	UVERBS_ATTR_PTR_OUT(
+		MLX5_IB_ATTR_QUERY_PORT,
+		UVERBS_ATTR_STRUCT(struct mlx5_ib_uapi_query_port,
+				   reg_c0),
+		UA_MANDATORY));
+
+ADD_UVERBS_METHODS(mlx5_ib_device,
+		   UVERBS_OBJECT_DEVICE,
+		   &UVERBS_METHOD(MLX5_IB_METHOD_QUERY_PORT));
+
 DECLARE_UVERBS_NAMED_METHOD(
 	MLX5_IB_METHOD_PD_QUERY,
 	UVERBS_ATTR_IDR(MLX5_IB_ATTR_QUERY_PD_HANDLE,
@@ -41,5 +211,8 @@ const struct uapi_definition mlx5_ib_std_types_defs[] = {
 	UAPI_DEF_CHAIN_OBJ_TREE(
 		UVERBS_OBJECT_PD,
 		&mlx5_ib_pd),
+	UAPI_DEF_CHAIN_OBJ_TREE(
+		UVERBS_OBJECT_DEVICE,
+		&mlx5_ib_device),
 	{},
 };