RDMA/umem: Combine contiguous PAGE_SIZE regions in SGEs

Combine contiguous regions of PAGE_SIZE pages into single scatter list
entry while building the scatter table for a umem. This minimizes the
number of the entries in the scatter list and reduces the DMA mapping
overhead, particularly with the IOMMU.

Set default max_seg_size in core for IB devices to 2G and do not combine
if we exceed this limit.

Also, purge npages in struct ib_umem as we now DMA map the umem SGL with
sg_nents and npage computation is not needed. Drivers should now be using
ib_umem_num_pages(), so fix the last stragglers.

Move npages tracking to ib_umem_odp as ODP drivers still need it.

Suggested-by: Jason Gunthorpe <[email protected]>
Reviewed-by: Michael J. Ruhl <[email protected]>
Reviewed-by: Ira Weiny <[email protected]>
Acked-by: Adit Ranadive <[email protected]>
Signed-off-by: Shiraz Saleem <[email protected]>
Tested-by: Gal Pressman <[email protected]>
Tested-by: Selvin Xavier <[email protected]>
Signed-off-by: Jason Gunthorpe <[email protected]>

drivers/infiniband/core/device.c

@@ -1089,6 +1089,9 @@ static void setup_dma_device(struct ib_device *device)
 		WARN_ON_ONCE(!parent);
 		device->dma_device = parent;
 	}
+	/* Setup default max segment size for all IB devices */
+	dma_set_max_seg_size(device->dma_device, SZ_2G);
+
 }
 
 /*

drivers/infiniband/core/umem.c

@@ -39,25 +39,22 @@
 #include <linux/export.h>
 #include <linux/hugetlb.h>
 #include <linux/slab.h>
+#include <linux/pagemap.h>
 #include <rdma/ib_umem_odp.h>
 
 #include "uverbs.h"
 
-
 static void __ib_umem_release(struct ib_device *dev, struct ib_umem *umem, int dirty)
 {
-	struct scatterlist *sg;
+	struct sg_page_iter sg_iter;
 	struct page *page;
-	int i;
 
 	if (umem->nmap > 0)
-		ib_dma_unmap_sg(dev, umem->sg_head.sgl,
-				umem->npages,
+		ib_dma_unmap_sg(dev, umem->sg_head.sgl, umem->sg_nents,
 				DMA_BIDIRECTIONAL);
 
-	for_each_sg(umem->sg_head.sgl, sg, umem->npages, i) {
-
-		page = sg_page(sg);
+	for_each_sg_page(umem->sg_head.sgl, &sg_iter, umem->sg_nents, 0) {
+		page = sg_page_iter_page(&sg_iter);
 		if (!PageDirty(page) && umem->writable && dirty)
 			set_page_dirty_lock(page);
 		put_page(page);
@@ -66,6 +63,69 @@ static void __ib_umem_release(struct ib_device *dev, struct ib_umem *umem, int d
 	sg_free_table(&umem->sg_head);
 }
 
+/* ib_umem_add_sg_table - Add N contiguous pages to scatter table
+ *
+ * sg: current scatterlist entry
+ * page_list: array of npage struct page pointers
+ * npages: number of pages in page_list
+ * max_seg_sz: maximum segment size in bytes
+ * nents: [out] number of entries in the scatterlist
+ *
+ * Return new end of scatterlist
+ */
+static struct scatterlist *ib_umem_add_sg_table(struct scatterlist *sg,
+						struct page **page_list,
+						unsigned long npages,
+						unsigned int max_seg_sz,
+						int *nents)
+{
+	unsigned long first_pfn;
+	unsigned long i = 0;
+	bool update_cur_sg = false;
+	bool first = !sg_page(sg);
+
+	/* Check if new page_list is contiguous with end of previous page_list.
+	 * sg->length here is a multiple of PAGE_SIZE and sg->offset is 0.
+	 */
+	if (!first && (page_to_pfn(sg_page(sg)) + (sg->length >> PAGE_SHIFT) ==
+		       page_to_pfn(page_list[0])))
+		update_cur_sg = true;
+
+	while (i != npages) {
+		unsigned long len;
+		struct page *first_page = page_list[i];
+
+		first_pfn = page_to_pfn(first_page);
+
+		/* Compute the number of contiguous pages we have starting
+		 * at i
+		 */
+		for (len = 0; i != npages &&
+			      first_pfn + len == page_to_pfn(page_list[i]);
+		     len++)
+			i++;
+
+		/* Squash N contiguous pages from page_list into current sge */
+		if (update_cur_sg &&
+		    ((max_seg_sz - sg->length) >= (len << PAGE_SHIFT))) {
+			sg_set_page(sg, sg_page(sg),
+				    sg->length + (len << PAGE_SHIFT), 0);
+			update_cur_sg = false;
+			continue;
+		}
+
+		/* Squash N contiguous pages into next sge or first sge */
+		if (!first)
+			sg = sg_next(sg);
+
+		(*nents)++;
+		sg_set_page(sg, first_page, len << PAGE_SHIFT, 0);
+		first = false;
+	}
+
+	return sg;
+}
+
 /**
  * ib_umem_get - Pin and DMA map userspace memory.
  *
@@ -93,7 +153,7 @@ struct ib_umem *ib_umem_get(struct ib_udata *udata, unsigned long addr,
 	int ret;
 	int i;
 	unsigned long dma_attrs = 0;
-	struct scatterlist *sg, *sg_list_start;
+	struct scatterlist *sg;
 	unsigned int gup_flags = FOLL_WRITE;
 
 	if (!udata)
@@ -190,7 +250,7 @@ struct ib_umem *ib_umem_get(struct ib_udata *udata, unsigned long addr,
 	if (!umem->writable)
 		gup_flags |= FOLL_FORCE;
 
-	sg_list_start = umem->sg_head.sgl;
+	sg = umem->sg_head.sgl;
 
 	while (npages) {
 		down_read(&mm->mmap_sem);
@@ -203,28 +263,29 @@ struct ib_umem *ib_umem_get(struct ib_udata *udata, unsigned long addr,
 			goto umem_release;
 		}
 
-		umem->npages += ret;
 		cur_base += ret * PAGE_SIZE;
 		npages   -= ret;
 
+		sg = ib_umem_add_sg_table(sg, page_list, ret,
+			dma_get_max_seg_size(context->device->dma_device),
+			&umem->sg_nents);
+
 		/* Continue to hold the mmap_sem as vma_list access
 		 * needs to be protected.
 		 */
-		for_each_sg(sg_list_start, sg, ret, i) {
+		for (i = 0; i < ret && umem->hugetlb; i++) {
 			if (vma_list && !is_vm_hugetlb_page(vma_list[i]))
 				umem->hugetlb = 0;
-
-			sg_set_page(sg, page_list[i], PAGE_SIZE, 0);
 		}
-		up_read(&mm->mmap_sem);
 
-		/* preparing for next loop */
-		sg_list_start = sg;
+		up_read(&mm->mmap_sem);
 	}
 
+	sg_mark_end(sg);
+
 	umem->nmap = ib_dma_map_sg_attrs(context->device,
 				  umem->sg_head.sgl,
-				  umem->npages,
+				  umem->sg_nents,
 				  DMA_BIDIRECTIONAL,
 				  dma_attrs);
 
@@ -320,8 +381,8 @@ int ib_umem_copy_from(void *dst, struct ib_umem *umem, size_t offset,
 		return -EINVAL;
 	}
 
-	ret = sg_pcopy_to_buffer(umem->sg_head.sgl, umem->npages, dst, length,
-				 offset + ib_umem_offset(umem));
+	ret = sg_pcopy_to_buffer(umem->sg_head.sgl, ib_umem_num_pages(umem),
+				 dst, length, offset + ib_umem_offset(umem));
 
 	if (ret < 0)
 		return ret;

drivers/infiniband/core/umem_odp.c

@@ -526,7 +526,7 @@ static int ib_umem_odp_map_dma_single_page(
 		}
 		umem_odp->dma_list[page_index] = dma_addr | access_mask;
 		umem_odp->page_list[page_index] = page;
-		umem->npages++;
+		umem_odp->npages++;
 	} else if (umem_odp->page_list[page_index] == page) {
 		umem_odp->dma_list[page_index] |= access_mask;
 	} else {
@@ -752,7 +752,7 @@ void ib_umem_odp_unmap_dma_pages(struct ib_umem_odp *umem_odp, u64 virt,
 			}
 			umem_odp->page_list[idx] = NULL;
 			umem_odp->dma_list[idx] = 0;
-			umem->npages--;
+			umem_odp->npages--;
 		}
 	}
 	mutex_unlock(&umem_odp->umem_mutex);

drivers/infiniband/hw/mlx5/odp.c

@@ -288,7 +288,7 @@ void mlx5_ib_invalidate_range(struct ib_umem_odp *umem_odp, unsigned long start,
 
 	ib_umem_odp_unmap_dma_pages(umem_odp, start, end);
 
-	if (unlikely(!umem->npages && mr->parent &&
+	if (unlikely(!umem_odp->npages && mr->parent &&
 		     !umem_odp->dying)) {
 		WRITE_ONCE(umem_odp->dying, 1);
 		atomic_inc(&mr->parent->num_leaf_free);

drivers/infiniband/hw/vmw_pvrdma/pvrdma_mr.c

@@ -119,7 +119,7 @@ struct ib_mr *pvrdma_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 	union pvrdma_cmd_resp rsp;
 	struct pvrdma_cmd_create_mr *cmd = &req.create_mr;
 	struct pvrdma_cmd_create_mr_resp *resp = &rsp.create_mr_resp;
-	int ret;
+	int ret, npages;
 
 	if (length == 0 || length > dev->dsr->caps.max_mr_size) {
 		dev_warn(&dev->pdev->dev, "invalid mem region length\n");
@@ -133,9 +133,10 @@ struct ib_mr *pvrdma_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 		return ERR_CAST(umem);
 	}
 
-	if (umem->npages < 0 || umem->npages > PVRDMA_PAGE_DIR_MAX_PAGES) {
+	npages = ib_umem_num_pages(umem);
+	if (npages < 0 || npages > PVRDMA_PAGE_DIR_MAX_PAGES) {
 		dev_warn(&dev->pdev->dev, "overflow %d pages in mem region\n",
-			 umem->npages);
+			 npages);
 		ret = -EINVAL;
 		goto err_umem;
 	}
@@ -150,7 +151,7 @@ struct ib_mr *pvrdma_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 	mr->mmr.size = length;
 	mr->umem = umem;
 
-	ret = pvrdma_page_dir_init(dev, &mr->pdir, umem->npages, false);
+	ret = pvrdma_page_dir_init(dev, &mr->pdir, npages, false);
 	if (ret) {
 		dev_warn(&dev->pdev->dev,
 			 "could not allocate page directory\n");
@@ -167,7 +168,7 @@ struct ib_mr *pvrdma_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 	cmd->length = length;
 	cmd->pd_handle = to_vpd(pd)->pd_handle;
 	cmd->access_flags = access_flags;
-	cmd->nchunks = umem->npages;
+	cmd->nchunks = npages;
 	cmd->pdir_dma = mr->pdir.dir_dma;
 
 	ret = pvrdma_cmd_post(dev, &req, &rsp, PVRDMA_CMD_CREATE_MR_RESP);

include/rdma/ib_umem.h

@@ -53,7 +53,7 @@ struct ib_umem {
 	struct work_struct	work;
 	struct sg_table sg_head;
 	int             nmap;
-	int             npages;
+	unsigned int    sg_nents;
 };
 
 /* Returns the offset of the umem start relative to the first page. */

include/rdma/ib_umem_odp.h

@@ -69,6 +69,7 @@ struct ib_umem_odp {
 
 	int notifiers_seq;
 	int notifiers_count;
+	int npages;
 
 	/* Tree tracking */
 	struct umem_odp_node	interval_tree;