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

Pull rdma updates from Doug Ledford:
 "This is my initial round of 4.4 merge window patches.  There are a few
  other things I wish to get in for 4.4 that aren't in this pull, as
  this represents what has gone through merge/build/run testing and not
  what is the last few items for which testing is not yet complete.

   - "Checksum offload support in user space" enablement
   - Misc cxgb4 fixes, add T6 support
   - Misc usnic fixes
   - 32 bit build warning fixes
   - Misc ocrdma fixes
   - Multicast loopback prevention extension
   - Extend the GID cache to store and return attributes of GIDs
   - Misc iSER updates
   - iSER clustering update
   - Network NameSpace support for rdma CM
   - Work Request cleanup series
   - New Memory Registration API"

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dledford/rdma: (76 commits)
  IB/core, cma: Make __attribute_const__ declarations sparse-friendly
  IB/core: Remove old fast registration API
  IB/ipath: Remove fast registration from the code
  IB/hfi1: Remove fast registration from the code
  RDMA/nes: Remove old FRWR API
  IB/qib: Remove old FRWR API
  iw_cxgb4: Remove old FRWR API
  RDMA/cxgb3: Remove old FRWR API
  RDMA/ocrdma: Remove old FRWR API
  IB/mlx4: Remove old FRWR API support
  IB/mlx5: Remove old FRWR API support
  IB/srp: Dont allocate a page vector when using fast_reg
  IB/srp: Remove srp_finish_mapping
  IB/srp: Convert to new registration API
  IB/srp: Split srp_map_sg
  RDS/IW: Convert to new memory registration API
  svcrdma: Port to new memory registration API
  xprtrdma: Port to new memory registration API
  iser-target: Port to new memory registration API
  IB/iser: Port to new fast registration API
  ...

5 files changed, 253 insertions, 222 deletions

diff --git a/drivers/infiniband/hw/mlx5/cq.c b/drivers/infiniband/hw/mlx5/cq.c
index 2d0dbbf38ceb..3dfd287256d6 100644
--- a/drivers/infiniband/hw/mlx5/cq.c
+++ b/drivers/infiniband/hw/mlx5/cq.c
@@ -109,8 +109,8 @@ static enum ib_wc_opcode get_umr_comp(struct mlx5_ib_wq *wq, int idx)
 	case IB_WR_LOCAL_INV:
 		return IB_WC_LOCAL_INV;
 
-	case IB_WR_FAST_REG_MR:
-		return IB_WC_FAST_REG_MR;
+	case IB_WR_REG_MR:
+		return IB_WC_REG_MR;
 
 	default:
 		pr_warn("unknown completion status\n");
diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index 68508d528ba0..7e97cb55a6bf 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -1425,8 +1425,7 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev)
 	dev->ib_dev.detach_mcast	= mlx5_ib_mcg_detach;
 	dev->ib_dev.process_mad		= mlx5_ib_process_mad;
 	dev->ib_dev.alloc_mr		= mlx5_ib_alloc_mr;
-	dev->ib_dev.alloc_fast_reg_page_list = mlx5_ib_alloc_fast_reg_page_list;
-	dev->ib_dev.free_fast_reg_page_list  = mlx5_ib_free_fast_reg_page_list;
+	dev->ib_dev.map_mr_sg		= mlx5_ib_map_mr_sg;
 	dev->ib_dev.check_mr_status	= mlx5_ib_check_mr_status;
 	dev->ib_dev.get_port_immutable  = mlx5_port_immutable;
 
diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h
index 22123b79d550..633347260b79 100644
--- a/drivers/infiniband/hw/mlx5/mlx5_ib.h
+++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h
@@ -245,6 +245,7 @@ enum mlx5_ib_qp_flags {
 };
 
 struct mlx5_umr_wr {
+	struct ib_send_wr		wr;
 	union {
 		u64			virt_addr;
 		u64			offset;
@@ -257,6 +258,11 @@ struct mlx5_umr_wr {
 	u32				mkey;
 };
 
+static inline struct mlx5_umr_wr *umr_wr(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;
@@ -313,6 +319,11 @@ enum mlx5_ib_mtt_access_flags {
 
 struct mlx5_ib_mr {
 	struct ib_mr		ibmr;
+	void			*descs;
+	dma_addr_t		desc_map;
+	int			ndescs;
+	int			max_descs;
+	int			desc_size;
 	struct mlx5_core_mr	mmr;
 	struct ib_umem	       *umem;
 	struct mlx5_shared_mr_info	*smr_info;
@@ -324,12 +335,7 @@ struct mlx5_ib_mr {
 	struct mlx5_create_mkey_mbox_out out;
 	struct mlx5_core_sig_ctx    *sig;
 	int			live;
-};
-
-struct mlx5_ib_fast_reg_page_list {
-	struct ib_fast_reg_page_list	ibfrpl;
-	__be64			       *mapped_page_list;
-	dma_addr_t			map;
+	void			*descs_alloc;
 };
 
 struct mlx5_ib_umr_context {
@@ -358,20 +364,6 @@ enum {
 	MLX5_FMR_BUSY,
 };
 
-struct mlx5_ib_fmr {
-	struct ib_fmr			ibfmr;
-	struct mlx5_core_mr		mr;
-	int				access_flags;
-	int				state;
-	/* protect fmr state
-	 */
-	spinlock_t			lock;
-	u64				wrid;
-	struct ib_send_wr		wr[2];
-	u8				page_shift;
-	struct ib_fast_reg_page_list	page_list;
-};
-
 struct mlx5_cache_ent {
 	struct list_head	head;
 	/* sync access to the cahce entry
@@ -456,11 +448,6 @@ static inline struct mlx5_ib_dev *to_mdev(struct ib_device *ibdev)
 	return container_of(ibdev, struct mlx5_ib_dev, ib_dev);
 }
 
-static inline struct mlx5_ib_fmr *to_mfmr(struct ib_fmr *ibfmr)
-{
-	return container_of(ibfmr, struct mlx5_ib_fmr, ibfmr);
-}
-
 static inline struct mlx5_ib_cq *to_mcq(struct ib_cq *ibcq)
 {
 	return container_of(ibcq, struct mlx5_ib_cq, ibcq);
@@ -501,11 +488,6 @@ static inline struct mlx5_ib_mr *to_mmr(struct ib_mr *ibmr)
 	return container_of(ibmr, struct mlx5_ib_mr, ibmr);
 }
 
-static inline struct mlx5_ib_fast_reg_page_list *to_mfrpl(struct ib_fast_reg_page_list *ibfrpl)
-{
-	return container_of(ibfrpl, struct mlx5_ib_fast_reg_page_list, ibfrpl);
-}
-
 struct mlx5_ib_ah {
 	struct ib_ah		ibah;
 	struct mlx5_av		av;
@@ -573,15 +555,9 @@ int mlx5_ib_dereg_mr(struct ib_mr *ibmr);
 struct ib_mr *mlx5_ib_alloc_mr(struct ib_pd *pd,
 			       enum ib_mr_type mr_type,
 			       u32 max_num_sg);
-struct ib_fast_reg_page_list *mlx5_ib_alloc_fast_reg_page_list(struct ib_device *ibdev,
-							       int page_list_len);
-void mlx5_ib_free_fast_reg_page_list(struct ib_fast_reg_page_list *page_list);
-struct ib_fmr *mlx5_ib_fmr_alloc(struct ib_pd *pd, int acc,
-				 struct ib_fmr_attr *fmr_attr);
-int mlx5_ib_map_phys_fmr(struct ib_fmr *ibfmr, u64 *page_list,
-		      int npages, u64 iova);
-int mlx5_ib_unmap_fmr(struct list_head *fmr_list);
-int mlx5_ib_fmr_dealloc(struct ib_fmr *ibfmr);
+int mlx5_ib_map_mr_sg(struct ib_mr *ibmr,
+		      struct scatterlist *sg,
+		      int sg_nents);
 int mlx5_ib_process_mad(struct ib_device *ibdev, int mad_flags, u8 port_num,
 			const struct ib_wc *in_wc, const struct ib_grh *in_grh,
 			const struct ib_mad_hdr *in, size_t in_mad_size,
diff --git a/drivers/infiniband/hw/mlx5/mr.c b/drivers/infiniband/hw/mlx5/mr.c
index 54a15b5d336d..ec8993a7b3be 100644
--- a/drivers/infiniband/hw/mlx5/mr.c
+++ b/drivers/infiniband/hw/mlx5/mr.c
@@ -687,7 +687,7 @@ static void prep_umr_reg_wqe(struct ib_pd *pd, struct ib_send_wr *wr,
 			     int access_flags)
 {
 	struct mlx5_ib_dev *dev = to_mdev(pd->device);
-	struct mlx5_umr_wr *umrwr = (struct mlx5_umr_wr *)&wr->wr.fast_reg;
+	struct mlx5_umr_wr *umrwr = umr_wr(wr);
 
 	sg->addr = dma;
 	sg->length = ALIGN(sizeof(u64) * n, 64);
@@ -715,7 +715,7 @@ static void prep_umr_reg_wqe(struct ib_pd *pd, struct ib_send_wr *wr,
 static void prep_umr_unreg_wqe(struct mlx5_ib_dev *dev,
 			       struct ib_send_wr *wr, u32 key)
 {
-	struct mlx5_umr_wr *umrwr = (struct mlx5_umr_wr *)&wr->wr.fast_reg;
+	struct mlx5_umr_wr *umrwr = umr_wr(wr);
 
 	wr->send_flags = MLX5_IB_SEND_UMR_UNREG | MLX5_IB_SEND_UMR_FAIL_IF_FREE;
 	wr->opcode = MLX5_IB_WR_UMR;
@@ -752,7 +752,8 @@ static struct mlx5_ib_mr *reg_umr(struct ib_pd *pd, struct ib_umem *umem,
 	struct device *ddev = dev->ib_dev.dma_device;
 	struct umr_common *umrc = &dev->umrc;
 	struct mlx5_ib_umr_context umr_context;
-	struct ib_send_wr wr, *bad;
+	struct mlx5_umr_wr umrwr;
+	struct ib_send_wr *bad;
 	struct mlx5_ib_mr *mr;
 	struct ib_sge sg;
 	int size;
@@ -798,14 +799,14 @@ static struct mlx5_ib_mr *reg_umr(struct ib_pd *pd, struct ib_umem *umem,
 		goto free_pas;
 	}
 
-	memset(&wr, 0, sizeof(wr));
-	wr.wr_id = (u64)(unsigned long)&umr_context;
-	prep_umr_reg_wqe(pd, &wr, &sg, dma, npages, mr->mmr.key, page_shift,
-			 virt_addr, len, access_flags);
+	memset(&umrwr, 0, sizeof(umrwr));
+	umrwr.wr.wr_id = (u64)(unsigned long)&umr_context;
+	prep_umr_reg_wqe(pd, &umrwr.wr, &sg, dma, npages, mr->mmr.key,
+			 page_shift, virt_addr, len, access_flags);
 
 	mlx5_ib_init_umr_context(&umr_context);
 	down(&umrc->sem);
-	err = ib_post_send(umrc->qp, &wr, &bad);
+	err = ib_post_send(umrc->qp, &umrwr.wr, &bad);
 	if (err) {
 		mlx5_ib_warn(dev, "post send failed, err %d\n", err);
 		goto unmap_dma;
@@ -851,8 +852,8 @@ int mlx5_ib_update_mtt(struct mlx5_ib_mr *mr, u64 start_page_index, int npages,
 	int size;
 	__be64 *pas;
 	dma_addr_t dma;
-	struct ib_send_wr wr, *bad;
-	struct mlx5_umr_wr *umrwr = (struct mlx5_umr_wr *)&wr.wr.fast_reg;
+	struct ib_send_wr *bad;
+	struct mlx5_umr_wr wr;
 	struct ib_sge sg;
 	int err = 0;
 	const int page_index_alignment = MLX5_UMR_MTT_ALIGNMENT / sizeof(u64);
@@ -917,26 +918,26 @@ int mlx5_ib_update_mtt(struct mlx5_ib_mr *mr, u64 start_page_index, int npages,
 		dma_sync_single_for_device(ddev, dma, size, DMA_TO_DEVICE);
 
 		memset(&wr, 0, sizeof(wr));
-		wr.wr_id = (u64)(unsigned long)&umr_context;
+		wr.wr.wr_id = (u64)(unsigned long)&umr_context;
 
 		sg.addr = dma;
 		sg.length = ALIGN(npages * sizeof(u64),
 				MLX5_UMR_MTT_ALIGNMENT);
 		sg.lkey = dev->umrc.pd->local_dma_lkey;
 
-		wr.send_flags = MLX5_IB_SEND_UMR_FAIL_IF_FREE |
+		wr.wr.send_flags = MLX5_IB_SEND_UMR_FAIL_IF_FREE |
 				MLX5_IB_SEND_UMR_UPDATE_MTT;
-		wr.sg_list = &sg;
-		wr.num_sge = 1;
-		wr.opcode = MLX5_IB_WR_UMR;
-		umrwr->npages = sg.length / sizeof(u64);
-		umrwr->page_shift = PAGE_SHIFT;
-		umrwr->mkey = mr->mmr.key;
-		umrwr->target.offset = start_page_index;
+		wr.wr.sg_list = &sg;
+		wr.wr.num_sge = 1;
+		wr.wr.opcode = MLX5_IB_WR_UMR;
+		wr.npages = sg.length / sizeof(u64);
+		wr.page_shift = PAGE_SHIFT;
+		wr.mkey = mr->mmr.key;
+		wr.target.offset = start_page_index;
 
 		mlx5_ib_init_umr_context(&umr_context);
 		down(&umrc->sem);
-		err = ib_post_send(umrc->qp, &wr, &bad);
+		err = ib_post_send(umrc->qp, &wr.wr, &bad);
 		if (err) {
 			mlx5_ib_err(dev, "UMR post send failed, err %d\n", err);
 		} else {
@@ -1122,16 +1123,17 @@ static int unreg_umr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr)
 {
 	struct umr_common *umrc = &dev->umrc;
 	struct mlx5_ib_umr_context umr_context;
-	struct ib_send_wr wr, *bad;
+	struct mlx5_umr_wr umrwr;
+	struct ib_send_wr *bad;
 	int err;
 
-	memset(&wr, 0, sizeof(wr));
-	wr.wr_id = (u64)(unsigned long)&umr_context;
-	prep_umr_unreg_wqe(dev, &wr, mr->mmr.key);
+	memset(&umrwr.wr, 0, sizeof(umrwr));
+	umrwr.wr.wr_id = (u64)(unsigned long)&umr_context;
+	prep_umr_unreg_wqe(dev, &umrwr.wr, mr->mmr.key);
 
 	mlx5_ib_init_umr_context(&umr_context);
 	down(&umrc->sem);
-	err = ib_post_send(umrc->qp, &wr, &bad);
+	err = ib_post_send(umrc->qp, &umrwr.wr, &bad);
 	if (err) {
 		up(&umrc->sem);
 		mlx5_ib_dbg(dev, "err %d\n", err);
@@ -1151,6 +1153,52 @@ error:
 	return err;
 }
 
+static int
+mlx5_alloc_priv_descs(struct ib_device *device,
+		      struct mlx5_ib_mr *mr,
+		      int ndescs,
+		      int desc_size)
+{
+	int size = ndescs * desc_size;
+	int add_size;
+	int ret;
+
+	add_size = max_t(int, MLX5_UMR_ALIGN - ARCH_KMALLOC_MINALIGN, 0);
+
+	mr->descs_alloc = kzalloc(size + add_size, GFP_KERNEL);
+	if (!mr->descs_alloc)
+		return -ENOMEM;
+
+	mr->descs = PTR_ALIGN(mr->descs_alloc, MLX5_UMR_ALIGN);
+
+	mr->desc_map = dma_map_single(device->dma_device, mr->descs,
+				      size, DMA_TO_DEVICE);
+	if (dma_mapping_error(device->dma_device, mr->desc_map)) {
+		ret = -ENOMEM;
+		goto err;
+	}
+
+	return 0;
+err:
+	kfree(mr->descs_alloc);
+
+	return ret;
+}
+
+static void
+mlx5_free_priv_descs(struct mlx5_ib_mr *mr)
+{
+	if (mr->descs) {
+		struct ib_device *device = mr->ibmr.device;
+		int size = mr->max_descs * mr->desc_size;
+
+		dma_unmap_single(device->dma_device, mr->desc_map,
+				 size, DMA_TO_DEVICE);
+		kfree(mr->descs_alloc);
+		mr->descs = NULL;
+	}
+}
+
 static int clean_mr(struct mlx5_ib_mr *mr)
 {
 	struct mlx5_ib_dev *dev = to_mdev(mr->ibmr.device);
@@ -1170,6 +1218,8 @@ static int clean_mr(struct mlx5_ib_mr *mr)
 		mr->sig = NULL;
 	}
 
+	mlx5_free_priv_descs(mr);
+
 	if (!umred) {
 		err = destroy_mkey(dev, mr);
 		if (err) {
@@ -1259,6 +1309,14 @@ struct ib_mr *mlx5_ib_alloc_mr(struct ib_pd *pd,
 	if (mr_type == IB_MR_TYPE_MEM_REG) {
 		access_mode = MLX5_ACCESS_MODE_MTT;
 		in->seg.log2_page_size = PAGE_SHIFT;
+
+		err = mlx5_alloc_priv_descs(pd->device, mr,
+					    ndescs, sizeof(u64));
+		if (err)
+			goto err_free_in;
+
+		mr->desc_size = sizeof(u64);
+		mr->max_descs = ndescs;
 	} else if (mr_type == IB_MR_TYPE_SIGNATURE) {
 		u32 psv_index[2];
 
@@ -1315,6 +1373,7 @@ err_destroy_psv:
 			mlx5_ib_warn(dev, "failed to destroy wire psv %d\n",
 				     mr->sig->psv_wire.psv_idx);
 	}
+	mlx5_free_priv_descs(mr);
 err_free_sig:
 	kfree(mr->sig);
 err_free_in:
@@ -1324,48 +1383,6 @@ err_free:
 	return ERR_PTR(err);
 }
 
-struct ib_fast_reg_page_list *mlx5_ib_alloc_fast_reg_page_list(struct ib_device *ibdev,
-							       int page_list_len)
-{
-	struct mlx5_ib_fast_reg_page_list *mfrpl;
-	int size = page_list_len * sizeof(u64);
-
-	mfrpl = kmalloc(sizeof(*mfrpl), GFP_KERNEL);
-	if (!mfrpl)
-		return ERR_PTR(-ENOMEM);
-
-	mfrpl->ibfrpl.page_list = kmalloc(size, GFP_KERNEL);
-	if (!mfrpl->ibfrpl.page_list)
-		goto err_free;
-
-	mfrpl->mapped_page_list = dma_alloc_coherent(ibdev->dma_device,
-						     size, &mfrpl->map,
-						     GFP_KERNEL);
-	if (!mfrpl->mapped_page_list)
-		goto err_free;
-
-	WARN_ON(mfrpl->map & 0x3f);
-
-	return &mfrpl->ibfrpl;
-
-err_free:
-	kfree(mfrpl->ibfrpl.page_list);
-	kfree(mfrpl);
-	return ERR_PTR(-ENOMEM);
-}
-
-void mlx5_ib_free_fast_reg_page_list(struct ib_fast_reg_page_list *page_list)
-{
-	struct mlx5_ib_fast_reg_page_list *mfrpl = to_mfrpl(page_list);
-	struct mlx5_ib_dev *dev = to_mdev(page_list->device);
-	int size = page_list->max_page_list_len * sizeof(u64);
-
-	dma_free_coherent(&dev->mdev->pdev->dev, size, mfrpl->mapped_page_list,
-			  mfrpl->map);
-	kfree(mfrpl->ibfrpl.page_list);
-	kfree(mfrpl);
-}
-
 int mlx5_ib_check_mr_status(struct ib_mr *ibmr, u32 check_mask,
 			    struct ib_mr_status *mr_status)
 {
@@ -1406,3 +1423,39 @@ int mlx5_ib_check_mr_status(struct ib_mr *ibmr, u32 check_mask,
 done:
 	return ret;
 }
+
+static int mlx5_set_page(struct ib_mr *ibmr, u64 addr)
+{
+	struct mlx5_ib_mr *mr = to_mmr(ibmr);
+	__be64 *descs;
+
+	if (unlikely(mr->ndescs == mr->max_descs))
+		return -ENOMEM;
+
+	descs = mr->descs;
+	descs[mr->ndescs++] = cpu_to_be64(addr | MLX5_EN_RD | MLX5_EN_WR);
+
+	return 0;
+}
+
+int mlx5_ib_map_mr_sg(struct ib_mr *ibmr,
+		      struct scatterlist *sg,
+		      int sg_nents)
+{
+	struct mlx5_ib_mr *mr = to_mmr(ibmr);
+	int n;
+
+	mr->ndescs = 0;
+
+	ib_dma_sync_single_for_cpu(ibmr->device, mr->desc_map,
+				   mr->desc_size * mr->max_descs,
+				   DMA_TO_DEVICE);
+
+	n = ib_sg_to_pages(ibmr, sg, sg_nents, mlx5_set_page);
+
+	ib_dma_sync_single_for_device(ibmr->device, mr->desc_map,
+				      mr->desc_size * mr->max_descs,
+				      DMA_TO_DEVICE);
+
+	return n;
+}
diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c
index 6f521a3418e8..307bdbca8938 100644
--- a/drivers/infiniband/hw/mlx5/qp.c
+++ b/drivers/infiniband/hw/mlx5/qp.c
@@ -64,7 +64,7 @@ static const u32 mlx5_ib_opcode[] = {
 	[IB_WR_ATOMIC_FETCH_AND_ADD]		= MLX5_OPCODE_ATOMIC_FA,
 	[IB_WR_SEND_WITH_INV]			= MLX5_OPCODE_SEND_INVAL,
 	[IB_WR_LOCAL_INV]			= MLX5_OPCODE_UMR,
-	[IB_WR_FAST_REG_MR]			= MLX5_OPCODE_UMR,
+	[IB_WR_REG_MR]				= MLX5_OPCODE_UMR,
 	[IB_WR_MASKED_ATOMIC_CMP_AND_SWP]	= MLX5_OPCODE_ATOMIC_MASKED_CS,
 	[IB_WR_MASKED_ATOMIC_FETCH_AND_ADD]	= MLX5_OPCODE_ATOMIC_MASKED_FA,
 	[MLX5_IB_WR_UMR]			= MLX5_OPCODE_UMR,
@@ -1838,9 +1838,9 @@ static __always_inline void set_raddr_seg(struct mlx5_wqe_raddr_seg *rseg,
 static void set_datagram_seg(struct mlx5_wqe_datagram_seg *dseg,
 			     struct ib_send_wr *wr)
 {
-	memcpy(&dseg->av, &to_mah(wr->wr.ud.ah)->av, sizeof(struct mlx5_av));
-	dseg->av.dqp_dct = cpu_to_be32(wr->wr.ud.remote_qpn | MLX5_EXTENDED_UD_AV);
-	dseg->av.key.qkey.qkey = cpu_to_be32(wr->wr.ud.remote_qkey);
+	memcpy(&dseg->av, &to_mah(ud_wr(wr)->ah)->av, sizeof(struct mlx5_av));
+	dseg->av.dqp_dct = cpu_to_be32(ud_wr(wr)->remote_qpn | MLX5_EXTENDED_UD_AV);
+	dseg->av.key.qkey.qkey = cpu_to_be32(ud_wr(wr)->remote_qkey);
 }
 
 static void set_data_ptr_seg(struct mlx5_wqe_data_seg *dseg, struct ib_sge *sg)
@@ -1896,22 +1896,24 @@ static __be64 sig_mkey_mask(void)
 	return cpu_to_be64(result);
 }
 
-static void set_frwr_umr_segment(struct mlx5_wqe_umr_ctrl_seg *umr,
-				 struct ib_send_wr *wr, int li)
+static void set_reg_umr_seg(struct mlx5_wqe_umr_ctrl_seg *umr,
+				struct mlx5_ib_mr *mr)
 {
-	memset(umr, 0, sizeof(*umr));
-
-	if (li) {
-		umr->mkey_mask = cpu_to_be64(MLX5_MKEY_MASK_FREE);
-		umr->flags = 1 << 7;
-		return;
-	}
+	int ndescs = mr->ndescs;
 
-	umr->flags = (1 << 5); /* fail if not free */
-	umr->klm_octowords = get_klm_octo(wr->wr.fast_reg.page_list_len);
+	memset(umr, 0, sizeof(*umr));
+	umr->flags = MLX5_UMR_CHECK_NOT_FREE;
+	umr->klm_octowords = get_klm_octo(ndescs);
 	umr->mkey_mask = frwr_mkey_mask();
 }
 
+static void set_linv_umr_seg(struct mlx5_wqe_umr_ctrl_seg *umr)
+{
+	memset(umr, 0, sizeof(*umr));
+	umr->mkey_mask = cpu_to_be64(MLX5_MKEY_MASK_FREE);
+	umr->flags = 1 << 7;
+}
+
 static __be64 get_umr_reg_mr_mask(void)
 {
 	u64 result;
@@ -1952,7 +1954,7 @@ static __be64 get_umr_update_mtt_mask(void)
 static void set_reg_umr_segment(struct mlx5_wqe_umr_ctrl_seg *umr,
 				struct ib_send_wr *wr)
 {
-	struct mlx5_umr_wr *umrwr = (struct mlx5_umr_wr *)&wr->wr.fast_reg;
+	struct mlx5_umr_wr *umrwr = umr_wr(wr);
 
 	memset(umr, 0, sizeof(*umr));
 
@@ -1987,29 +1989,31 @@ static u8 get_umr_flags(int acc)
 		MLX5_PERM_LOCAL_READ | MLX5_PERM_UMR_EN;
 }
 
-static void set_mkey_segment(struct mlx5_mkey_seg *seg, struct ib_send_wr *wr,
-			     int li, int *writ)
+static void set_reg_mkey_seg(struct mlx5_mkey_seg *seg,
+			     struct mlx5_ib_mr *mr,
+			     u32 key, int access)
 {
-	memset(seg, 0, sizeof(*seg));
-	if (li) {
-		seg->status = MLX5_MKEY_STATUS_FREE;
-		return;
-	}
+	int ndescs = ALIGN(mr->ndescs, 8) >> 1;
 
-	seg->flags = get_umr_flags(wr->wr.fast_reg.access_flags) |
-		     MLX5_ACCESS_MODE_MTT;
-	*writ = seg->flags & (MLX5_PERM_LOCAL_WRITE | IB_ACCESS_REMOTE_WRITE);
-	seg->qpn_mkey7_0 = cpu_to_be32((wr->wr.fast_reg.rkey & 0xff) | 0xffffff00);
+	memset(seg, 0, sizeof(*seg));
+	seg->flags = get_umr_flags(access) | MLX5_ACCESS_MODE_MTT;
+	seg->qpn_mkey7_0 = cpu_to_be32((key & 0xff) | 0xffffff00);
 	seg->flags_pd = cpu_to_be32(MLX5_MKEY_REMOTE_INVAL);
-	seg->start_addr = cpu_to_be64(wr->wr.fast_reg.iova_start);
-	seg->len = cpu_to_be64(wr->wr.fast_reg.length);
-	seg->xlt_oct_size = cpu_to_be32((wr->wr.fast_reg.page_list_len + 1) / 2);
-	seg->log2_page_size = wr->wr.fast_reg.page_shift;
+	seg->start_addr = cpu_to_be64(mr->ibmr.iova);
+	seg->len = cpu_to_be64(mr->ibmr.length);
+	seg->xlt_oct_size = cpu_to_be32(ndescs);
+	seg->log2_page_size = ilog2(mr->ibmr.page_size);
+}
+
+static void set_linv_mkey_seg(struct mlx5_mkey_seg *seg)
+{
+	memset(seg, 0, sizeof(*seg));
+	seg->status = MLX5_MKEY_STATUS_FREE;
 }
 
 static void set_reg_mkey_segment(struct mlx5_mkey_seg *seg, struct ib_send_wr *wr)
 {
-	struct mlx5_umr_wr *umrwr = (struct mlx5_umr_wr *)&wr->wr.fast_reg;
+	struct mlx5_umr_wr *umrwr = umr_wr(wr);
 
 	memset(seg, 0, sizeof(*seg));
 	if (wr->send_flags & MLX5_IB_SEND_UMR_UNREG) {
@@ -2028,21 +2032,14 @@ static void set_reg_mkey_segment(struct mlx5_mkey_seg *seg, struct ib_send_wr *w
 				       mlx5_mkey_variant(umrwr->mkey));
 }
 
-static void set_frwr_pages(struct mlx5_wqe_data_seg *dseg,
-			   struct ib_send_wr *wr,
-			   struct mlx5_core_dev *mdev,
-			   struct mlx5_ib_pd *pd,
-			   int writ)
+static void set_reg_data_seg(struct mlx5_wqe_data_seg *dseg,
+			     struct mlx5_ib_mr *mr,
+			     struct mlx5_ib_pd *pd)
 {
-	struct mlx5_ib_fast_reg_page_list *mfrpl = to_mfrpl(wr->wr.fast_reg.page_list);
-	u64 *page_list = wr->wr.fast_reg.page_list->page_list;
-	u64 perm = MLX5_EN_RD | (writ ? MLX5_EN_WR : 0);
-	int i;
+	int bcount = mr->desc_size * mr->ndescs;
 
-	for (i = 0; i < wr->wr.fast_reg.page_list_len; i++)
-		mfrpl->mapped_page_list[i] = cpu_to_be64(page_list[i] | perm);
-	dseg->addr = cpu_to_be64(mfrpl->map);
-	dseg->byte_count = cpu_to_be32(ALIGN(sizeof(u64) * wr->wr.fast_reg.page_list_len, 64));
+	dseg->addr = cpu_to_be64(mr->desc_map);
+	dseg->byte_count = cpu_to_be32(ALIGN(bcount, 64));
 	dseg->lkey = cpu_to_be32(pd->ibpd.local_dma_lkey);
 }
 
@@ -2224,22 +2221,22 @@ static int mlx5_set_bsf(struct ib_mr *sig_mr,
 	return 0;
 }
 
-static int set_sig_data_segment(struct ib_send_wr *wr, struct mlx5_ib_qp *qp,
-				void **seg, int *size)
+static int set_sig_data_segment(struct ib_sig_handover_wr *wr,
+				struct mlx5_ib_qp *qp, void **seg, int *size)
 {
-	struct ib_sig_attrs *sig_attrs = wr->wr.sig_handover.sig_attrs;
-	struct ib_mr *sig_mr = wr->wr.sig_handover.sig_mr;
+	struct ib_sig_attrs *sig_attrs = wr->sig_attrs;
+	struct ib_mr *sig_mr = wr->sig_mr;
 	struct mlx5_bsf *bsf;
-	u32 data_len = wr->sg_list->length;
-	u32 data_key = wr->sg_list->lkey;
-	u64 data_va = wr->sg_list->addr;
+	u32 data_len = wr->wr.sg_list->length;
+	u32 data_key = wr->wr.sg_list->lkey;
+	u64 data_va = wr->wr.sg_list->addr;
 	int ret;
 	int wqe_size;
 
-	if (!wr->wr.sig_handover.prot ||
-	    (data_key == wr->wr.sig_handover.prot->lkey &&
-	     data_va == wr->wr.sig_handover.prot->addr &&
-	     data_len == wr->wr.sig_handover.prot->length)) {
+	if (!wr->prot ||
+	    (data_key == wr->prot->lkey &&
+	     data_va == wr->prot->addr &&
+	     data_len == wr->prot->length)) {
 		/**
 		 * Source domain doesn't contain signature information
 		 * or data and protection are interleaved in memory.
@@ -2273,8 +2270,8 @@ static int set_sig_data_segment(struct ib_send_wr *wr, struct mlx5_ib_qp *qp,
 		struct mlx5_stride_block_ctrl_seg *sblock_ctrl;
 		struct mlx5_stride_block_entry *data_sentry;
 		struct mlx5_stride_block_entry *prot_sentry;
-		u32 prot_key = wr->wr.sig_handover.prot->lkey;
-		u64 prot_va = wr->wr.sig_handover.prot->addr;
+		u32 prot_key = wr->prot->lkey;
+		u64 prot_va = wr->prot->addr;
 		u16 block_size = sig_attrs->mem.sig.dif.pi_interval;
 		int prot_size;
 
@@ -2326,16 +2323,16 @@ static int set_sig_data_segment(struct ib_send_wr *wr, struct mlx5_ib_qp *qp,
 }
 
 static void set_sig_mkey_segment(struct mlx5_mkey_seg *seg,
-				 struct ib_send_wr *wr, u32 nelements,
+				 struct ib_sig_handover_wr *wr, u32 nelements,
 				 u32 length, u32 pdn)
 {
-	struct ib_mr *sig_mr = wr->wr.sig_handover.sig_mr;
+	struct ib_mr *sig_mr = wr->sig_mr;
 	u32 sig_key = sig_mr->rkey;
 	u8 sigerr = to_mmr(sig_mr)->sig->sigerr_count & 1;
 
 	memset(seg, 0, sizeof(*seg));
 
-	seg->flags = get_umr_flags(wr->wr.sig_handover.access_flags) |
+	seg->flags = get_umr_flags(wr->access_flags) |
 				   MLX5_ACCESS_MODE_KLM;
 	seg->qpn_mkey7_0 = cpu_to_be32((sig_key & 0xff) | 0xffffff00);
 	seg->flags_pd = cpu_to_be32(MLX5_MKEY_REMOTE_INVAL | sigerr << 26 |
@@ -2346,7 +2343,7 @@ static void set_sig_mkey_segment(struct mlx5_mkey_seg *seg,
 }
 
 static void set_sig_umr_segment(struct mlx5_wqe_umr_ctrl_seg *umr,
-				struct ib_send_wr *wr, u32 nelements)
+				u32 nelements)
 {
 	memset(umr, 0, sizeof(*umr));
 
@@ -2357,37 +2354,37 @@ static void set_sig_umr_segment(struct mlx5_wqe_umr_ctrl_seg *umr,
 }
 
 
-static int set_sig_umr_wr(struct ib_send_wr *wr, struct mlx5_ib_qp *qp,
+static int set_sig_umr_wr(struct ib_send_wr *send_wr, struct mlx5_ib_qp *qp,
 			  void **seg, int *size)
 {
-	struct mlx5_ib_mr *sig_mr = to_mmr(wr->wr.sig_handover.sig_mr);
+	struct ib_sig_handover_wr *wr = sig_handover_wr(send_wr);
+	struct mlx5_ib_mr *sig_mr = to_mmr(wr->sig_mr);
 	u32 pdn = get_pd(qp)->pdn;
 	u32 klm_oct_size;
 	int region_len, ret;
 
-	if (unlikely(wr->num_sge != 1) ||
-	    unlikely(wr->wr.sig_handover.access_flags &
-		     IB_ACCESS_REMOTE_ATOMIC) ||
+	if (unlikely(wr->wr.num_sge != 1) ||
+	    unlikely(wr->access_flags & IB_ACCESS_REMOTE_ATOMIC) ||
 	    unlikely(!sig_mr->sig) || unlikely(!qp->signature_en) ||
 	    unlikely(!sig_mr->sig->sig_status_checked))
 		return -EINVAL;
 
 	/* length of the protected region, data + protection */
-	region_len = wr->sg_list->length;
-	if (wr->wr.sig_handover.prot &&
-	    (wr->wr.sig_handover.prot->lkey != wr->sg_list->lkey  ||
-	     wr->wr.sig_handover.prot->addr != wr->sg_list->addr  ||
-	     wr->wr.sig_handover.prot->length != wr->sg_list->length))
-		region_len += wr->wr.sig_handover.prot->length;
+	region_len = wr->wr.sg_list->length;
+	if (wr->prot &&
+	    (wr->prot->lkey != wr->wr.sg_list->lkey  ||
+	     wr->prot->addr != wr->wr.sg_list->addr  ||
+	     wr->prot->length != wr->wr.sg_list->length))
+		region_len += wr->prot->length;
 
 	/**
 	 * KLM octoword size - if protection was provided
 	 * then we use strided block format (3 octowords),
 	 * else we use single KLM (1 octoword)
 	 **/
-	klm_oct_size = wr->wr.sig_handover.prot ? 3 : 1;
+	klm_oct_size = wr->prot ? 3 : 1;
 
-	set_sig_umr_segment(*seg, wr, klm_oct_size);
+	set_sig_umr_segment(*seg, klm_oct_size);
 	*seg += sizeof(struct mlx5_wqe_umr_ctrl_seg);
 	*size += sizeof(struct mlx5_wqe_umr_ctrl_seg) / 16;
 	if (unlikely((*seg == qp->sq.qend)))
@@ -2433,38 +2430,52 @@ static int set_psv_wr(struct ib_sig_domain *domain,
 	return 0;
 }
 
-static int set_frwr_li_wr(void **seg, struct ib_send_wr *wr, int *size,
-			  struct mlx5_core_dev *mdev, struct mlx5_ib_pd *pd, struct mlx5_ib_qp *qp)
+static int set_reg_wr(struct mlx5_ib_qp *qp,
+		      struct ib_reg_wr *wr,
+		      void **seg, int *size)
 {
-	int writ = 0;
-	int li;
+	struct mlx5_ib_mr *mr = to_mmr(wr->mr);
+	struct mlx5_ib_pd *pd = to_mpd(qp->ibqp.pd);
 
-	li = wr->opcode == IB_WR_LOCAL_INV ? 1 : 0;
-	if (unlikely(wr->send_flags & IB_SEND_INLINE))
+	if (unlikely(wr->wr.send_flags & IB_SEND_INLINE)) {
+		mlx5_ib_warn(to_mdev(qp->ibqp.device),
+			     "Invalid IB_SEND_INLINE send flag\n");
 		return -EINVAL;
+	}
 
-	set_frwr_umr_segment(*seg, wr, li);
+	set_reg_umr_seg(*seg, mr);
 	*seg += sizeof(struct mlx5_wqe_umr_ctrl_seg);
 	*size += sizeof(struct mlx5_wqe_umr_ctrl_seg) / 16;
 	if (unlikely((*seg == qp->sq.qend)))
 		*seg = mlx5_get_send_wqe(qp, 0);
-	set_mkey_segment(*seg, wr, li, &writ);
+
+	set_reg_mkey_seg(*seg, mr, wr->key, wr->access);
 	*seg += sizeof(struct mlx5_mkey_seg);
 	*size += sizeof(struct mlx5_mkey_seg) / 16;
 	if (unlikely((*seg == qp->sq.qend)))
 		*seg = mlx5_get_send_wqe(qp, 0);
-	if (!li) {
-		if (unlikely(wr->wr.fast_reg.page_list_len >
-			     wr->wr.fast_reg.page_list->max_page_list_len))
-			return	-ENOMEM;
 
-		set_frwr_pages(*seg, wr, mdev, pd, writ);
-		*seg += sizeof(struct mlx5_wqe_data_seg);
-		*size += (sizeof(struct mlx5_wqe_data_seg) / 16);
-	}
+	set_reg_data_seg(*seg, mr, pd);
+	*seg += sizeof(struct mlx5_wqe_data_seg);
+	*size += (sizeof(struct mlx5_wqe_data_seg) / 16);
+
 	return 0;
 }
 
+static void set_linv_wr(struct mlx5_ib_qp *qp, void **seg, int *size)
+{
+	set_linv_umr_seg(*seg);
+	*seg += sizeof(struct mlx5_wqe_umr_ctrl_seg);
+	*size += sizeof(struct mlx5_wqe_umr_ctrl_seg) / 16;
+	if (unlikely((*seg == qp->sq.qend)))
+		*seg = mlx5_get_send_wqe(qp, 0);
+	set_linv_mkey_seg(*seg);
+	*seg += sizeof(struct mlx5_mkey_seg);
+	*size += sizeof(struct mlx5_mkey_seg) / 16;
+	if (unlikely((*seg == qp->sq.qend)))
+		*seg = mlx5_get_send_wqe(qp, 0);
+}
+
 static void dump_wqe(struct mlx5_ib_qp *qp, int idx, int size_16)
 {
 	__be32 *p = NULL;
@@ -2578,7 +2589,6 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 {
 	struct mlx5_wqe_ctrl_seg *ctrl = NULL;  /* compiler warning */
 	struct mlx5_ib_dev *dev = to_mdev(ibqp->device);
-	struct mlx5_core_dev *mdev = dev->mdev;
 	struct mlx5_ib_qp *qp = to_mqp(ibqp);
 	struct mlx5_ib_mr *mr;
 	struct mlx5_wqe_data_seg *dpseg;
@@ -2627,7 +2637,6 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 		switch (ibqp->qp_type) {
 		case IB_QPT_XRC_INI:
 			xrc = seg;
-			xrc->xrc_srqn = htonl(wr->xrc_remote_srq_num);
 			seg += sizeof(*xrc);
 			size += sizeof(*xrc) / 16;
 			/* fall through */
@@ -2636,8 +2645,8 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 			case IB_WR_RDMA_READ:
 			case IB_WR_RDMA_WRITE:
 			case IB_WR_RDMA_WRITE_WITH_IMM:
-				set_raddr_seg(seg, wr->wr.rdma.remote_addr,
-					      wr->wr.rdma.rkey);
+				set_raddr_seg(seg, rdma_wr(wr)->remote_addr,
+					      rdma_wr(wr)->rkey);
 				seg += sizeof(struct mlx5_wqe_raddr_seg);
 				size += sizeof(struct mlx5_wqe_raddr_seg) / 16;
 				break;
@@ -2654,22 +2663,16 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 				next_fence = MLX5_FENCE_MODE_INITIATOR_SMALL;
 				qp->sq.wr_data[idx] = IB_WR_LOCAL_INV;
 				ctrl->imm = cpu_to_be32(wr->ex.invalidate_rkey);
-				err = set_frwr_li_wr(&seg, wr, &size, mdev, to_mpd(ibqp->pd), qp);
-				if (err) {
-					mlx5_ib_warn(dev, "\n");
-					*bad_wr = wr;
-					goto out;
-				}
+				set_linv_wr(qp, &seg, &size);
 				num_sge = 0;
 				break;
 
-			case IB_WR_FAST_REG_MR:
+			case IB_WR_REG_MR:
 				next_fence = MLX5_FENCE_MODE_INITIATOR_SMALL;
-				qp->sq.wr_data[idx] = IB_WR_FAST_REG_MR;
-				ctrl->imm = cpu_to_be32(wr->wr.fast_reg.rkey);
-				err = set_frwr_li_wr(&seg, wr, &size, mdev, to_mpd(ibqp->pd), qp);
+				qp->sq.wr_data[idx] = IB_WR_REG_MR;
+				ctrl->imm = cpu_to_be32(reg_wr(wr)->key);
+				err = set_reg_wr(qp, reg_wr(wr), &seg, &size);
 				if (err) {
-					mlx5_ib_warn(dev, "\n");
 					*bad_wr = wr;
 					goto out;
 				}
@@ -2678,7 +2681,7 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 
 			case IB_WR_REG_SIG_MR:
 				qp->sq.wr_data[idx] = IB_WR_REG_SIG_MR;
-				mr = to_mmr(wr->wr.sig_handover.sig_mr);
+				mr = to_mmr(sig_handover_wr(wr)->sig_mr);
 
 				ctrl->imm = cpu_to_be32(mr->ibmr.rkey);
 				err = set_sig_umr_wr(wr, qp, &seg, &size);
@@ -2706,7 +2709,7 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 					goto out;
 				}
 
-				err = set_psv_wr(&wr->wr.sig_handover.sig_attrs->mem,
+				err = set_psv_wr(&sig_handover_wr(wr)->sig_attrs->mem,
 						 mr->sig->psv_memory.psv_idx, &seg,
 						 &size);
 				if (err) {
@@ -2728,7 +2731,7 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 				}
 
 				next_fence = MLX5_FENCE_MODE_INITIATOR_SMALL;
-				err = set_psv_wr(&wr->wr.sig_handover.sig_attrs->wire,
+				err = set_psv_wr(&sig_handover_wr(wr)->sig_attrs->wire,
 						 mr->sig->psv_wire.psv_idx, &seg,
 						 &size);
 				if (err) {
@@ -2752,8 +2755,8 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 			switch (wr->opcode) {
 			case IB_WR_RDMA_WRITE:
 			case IB_WR_RDMA_WRITE_WITH_IMM:
-				set_raddr_seg(seg, wr->wr.rdma.remote_addr,
-					      wr->wr.rdma.rkey);
+				set_raddr_seg(seg, rdma_wr(wr)->remote_addr,
+					      rdma_wr(wr)->rkey);
 				seg  += sizeof(struct mlx5_wqe_raddr_seg);
 				size += sizeof(struct mlx5_wqe_raddr_seg) / 16;
 				break;
@@ -2780,7 +2783,7 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 				goto out;
 			}
 			qp->sq.wr_data[idx] = MLX5_IB_WR_UMR;
-			ctrl->imm = cpu_to_be32(wr->wr.fast_reg.rkey);
+			ctrl->imm = cpu_to_be32(umr_wr(wr)->mkey);
 			set_reg_umr_segment(seg, wr);
 			seg += sizeof(struct mlx5_wqe_umr_ctrl_seg);
 			size += sizeof(struct mlx5_wqe_umr_ctrl_seg) / 16;