From: Christoph Hellwig <hch@lst.de>
Date: Wed, 14 Aug 2019 09:59:24 +0200
Subject: nouveau: simplify nouveau_dmem_migrate_to_ram
Git-commit: bfe69ef94aeeae76531b8f37799b8ac6fca56957
Patch-mainline: v5.4-rc1
References: jsc#SLE-16387, bsc#1152489

Factor the main copy page to ram routine out into a helper that acts on
a single page and which doesn't require the nouveau_dmem_fault
structure for argument passing.  Also remove the loop over multiple
pages as we only handle one at the moment, although the structure of
the main worker function makes it relatively easy to add multi page
support back if needed in the future.  But at least for now this avoid
the needed to dynamically allocate memory for the dma addresses in
what is essentially the page fault path.

Link: https://lore.kernel.org/r/20190814075928.23766-7-hch@lst.de
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Ralph Campbell <rcampbell@nvidia.com>
Tested-by: Ralph Campbell <rcampbell@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
Acked-by: Thomas Zimmermann <tzimmermann@suse.de>
---
 drivers/gpu/drm/nouveau/nouveau_dmem.c |  161 ++++++++-------------------------
 1 file changed, 40 insertions(+), 121 deletions(-)

--- a/drivers/gpu/drm/nouveau/nouveau_dmem.c
+++ b/drivers/gpu/drm/nouveau/nouveau_dmem.c
@@ -86,13 +86,6 @@ static inline struct nouveau_dmem *page_
 	return container_of(page->pgmap, struct nouveau_dmem, pagemap);
 }
 
-struct nouveau_dmem_fault {
-	struct nouveau_drm *drm;
-	struct nouveau_fence *fence;
-	dma_addr_t *dma;
-	unsigned long npages;
-};
-
 struct nouveau_migrate {
 	struct vm_area_struct *vma;
 	struct nouveau_drm *drm;
@@ -146,130 +139,55 @@ static void nouveau_dmem_fence_done(stru
 	}
 }
 
-static void
-nouveau_dmem_fault_alloc_and_copy(struct vm_area_struct *vma,
-				  const unsigned long *src_pfns,
-				  unsigned long *dst_pfns,
-				  unsigned long start,
-				  unsigned long end,
-				  struct nouveau_dmem_fault *fault)
+static vm_fault_t nouveau_dmem_fault_copy_one(struct nouveau_drm *drm,
+		struct vm_fault *vmf, struct migrate_vma *args,
+		dma_addr_t *dma_addr)
 {
-	struct nouveau_drm *drm = fault->drm;
 	struct device *dev = drm->dev->dev;
-	unsigned long addr, i, npages = 0;
-	nouveau_migrate_copy_t copy;
-	int ret;
-
-
-	/* First allocate new memory */
-	for (addr = start, i = 0; addr < end; addr += PAGE_SIZE, i++) {
-		struct page *dpage, *spage;
-
-		dst_pfns[i] = 0;
-		spage = migrate_pfn_to_page(src_pfns[i]);
-		if (!spage || !(src_pfns[i] & MIGRATE_PFN_MIGRATE))
-			continue;
-
-		dpage = alloc_page_vma(GFP_HIGHUSER, vma, addr);
-		if (!dpage) {
-			dst_pfns[i] = MIGRATE_PFN_ERROR;
-			continue;
-		}
-		lock_page(dpage);
-
-		dst_pfns[i] = migrate_pfn(page_to_pfn(dpage)) |
-			      MIGRATE_PFN_LOCKED;
-		npages++;
-	}
+	struct page *dpage, *spage;
 
-	/* Allocate storage for DMA addresses, so we can unmap later. */
-	fault->dma = kmalloc(sizeof(*fault->dma) * npages, GFP_KERNEL);
-	if (!fault->dma)
-		goto error;
-
-	/* Copy things over */
-	copy = drm->dmem->migrate.copy_func;
-	for (addr = start, i = 0; addr < end; addr += PAGE_SIZE, i++) {
-		struct page *spage, *dpage;
-
-		dpage = migrate_pfn_to_page(dst_pfns[i]);
-		if (!dpage || dst_pfns[i] == MIGRATE_PFN_ERROR)
-			continue;
-
-		spage = migrate_pfn_to_page(src_pfns[i]);
-		if (!spage || !(src_pfns[i] & MIGRATE_PFN_MIGRATE)) {
-			dst_pfns[i] = MIGRATE_PFN_ERROR;
-			__free_page(dpage);
-			continue;
-		}
-
-		fault->dma[fault->npages] =
-			dma_map_page_attrs(dev, dpage, 0, PAGE_SIZE,
-					   PCI_DMA_BIDIRECTIONAL,
-					   DMA_ATTR_SKIP_CPU_SYNC);
-		if (dma_mapping_error(dev, fault->dma[fault->npages])) {
-			dst_pfns[i] = MIGRATE_PFN_ERROR;
-			__free_page(dpage);
-			continue;
-		}
-
-		ret = copy(drm, 1, NOUVEAU_APER_HOST,
-				fault->dma[fault->npages++],
-				NOUVEAU_APER_VRAM,
-				nouveau_dmem_page_addr(spage));
-		if (ret) {
-			dst_pfns[i] = MIGRATE_PFN_ERROR;
-			__free_page(dpage);
-			continue;
-		}
-	}
-
-	nouveau_fence_new(drm->dmem->migrate.chan, false, &fault->fence);
-
-	return;
-
-error:
-	for (addr = start, i = 0; addr < end; addr += PAGE_SIZE, ++i) {
-		struct page *page;
-
-		if (!dst_pfns[i] || dst_pfns[i] == MIGRATE_PFN_ERROR)
-			continue;
+	spage = migrate_pfn_to_page(args->src[0]);
+	if (!spage || !(args->src[0] & MIGRATE_PFN_MIGRATE))
+		return 0;
 
-		page = migrate_pfn_to_page(dst_pfns[i]);
-		dst_pfns[i] = MIGRATE_PFN_ERROR;
-		if (page == NULL)
-			continue;
+	dpage = alloc_page_vma(GFP_HIGHUSER, vmf->vma, vmf->address);
+	if (!dpage)
+		return VM_FAULT_SIGBUS;
+	lock_page(dpage);
 
-		__free_page(page);
-	}
-}
+	*dma_addr = dma_map_page(dev, dpage, 0, PAGE_SIZE, DMA_BIDIRECTIONAL);
+	if (dma_mapping_error(dev, *dma_addr))
+		goto error_free_page;
 
-static void
-nouveau_dmem_fault_finalize_and_map(struct nouveau_dmem_fault *fault)
-{
-	struct nouveau_drm *drm = fault->drm;
+	if (drm->dmem->migrate.copy_func(drm, 1, NOUVEAU_APER_HOST, *dma_addr,
+			NOUVEAU_APER_VRAM, nouveau_dmem_page_addr(spage)))
+		goto error_dma_unmap;
 
-	nouveau_dmem_fence_done(&fault->fence);
+	args->dst[0] = migrate_pfn(page_to_pfn(dpage)) | MIGRATE_PFN_LOCKED;
+	return 0;
 
-	while (fault->npages--) {
-		dma_unmap_page(drm->dev->dev, fault->dma[fault->npages],
-			       PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
-	}
-	kfree(fault->dma);
+error_dma_unmap:
+	dma_unmap_page(dev, *dma_addr, PAGE_SIZE, DMA_BIDIRECTIONAL);
+error_free_page:
+	__free_page(dpage);
+	return VM_FAULT_SIGBUS;
 }
 
 static vm_fault_t nouveau_dmem_migrate_to_ram(struct vm_fault *vmf)
 {
 	struct nouveau_dmem *dmem = page_to_dmem(vmf->page);
-	unsigned long src[1] = {0}, dst[1] = {0};
+	struct nouveau_drm *drm = dmem->drm;
+	struct nouveau_fence *fence;
+	unsigned long src = 0, dst = 0;
+	dma_addr_t dma_addr = 0;
+	vm_fault_t ret;
 	struct migrate_vma args = {
 		.vma		= vmf->vma,
 		.start		= vmf->address,
 		.end		= vmf->address + PAGE_SIZE,
-		.src		= src,
-		.dst		= dst,
+		.src		= &src,
+		.dst		= &dst,
 	};
-	struct nouveau_dmem_fault fault = { .drm = dmem->drm };
 
 	/*
 	 * FIXME what we really want is to find some heuristic to migrate more
@@ -281,16 +199,17 @@ static vm_fault_t nouveau_dmem_migrate_t
 	if (!args.cpages)
 		return 0;
 
-	nouveau_dmem_fault_alloc_and_copy(args.vma, src, dst, args.start,
-			args.end, &fault);
-	migrate_vma_pages(&args);
-	nouveau_dmem_fault_finalize_and_map(&fault);
+	ret = nouveau_dmem_fault_copy_one(drm, vmf, &args, &dma_addr);
+	if (ret || dst == 0)
+		goto done;
 
+	nouveau_fence_new(dmem->migrate.chan, false, &fence);
+	migrate_vma_pages(&args);
+	nouveau_dmem_fence_done(&fence);
+	dma_unmap_page(drm->dev->dev, dma_addr, PAGE_SIZE, DMA_BIDIRECTIONAL);
+done:
 	migrate_vma_finalize(&args);
-	if (dst[0] == MIGRATE_PFN_ERROR)
-		return VM_FAULT_SIGBUS;
-
-	return 0;
+	return ret;
 }
 
 static const struct dev_pagemap_ops nouveau_dmem_pagemap_ops = {