memremap: add a migrate_to_ram method to struct dev_pagemap_ops

This replaces the hacky ->fault callback, which is currently directly
called from common code through a hmm specific data structure as an
exercise in layering violations.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Ralph Campbell <rcampbell@nvidia.com>
Reviewed-by: Jason Gunthorpe <jgg@mellanox.com>
Reviewed-by: Dan Williams <dan.j.williams@intel.com>
Tested-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>

include/linux/hmm.h

@@ -692,11 +692,6 @@ struct hmm_devmem_ops {
  * chunk, as an optimization. It must, however, prioritize the faulting address
  * over all the others.
  */
-typedef vm_fault_t (*dev_page_fault_t)(struct vm_area_struct *vma,
-				unsigned long addr,
-				const struct page *page,
-				unsigned int flags,
-				pmd_t *pmdp);
 
 struct hmm_devmem {
 	struct completion		completion;
@@ -707,7 +702,6 @@ struct hmm_devmem {
 	struct dev_pagemap		pagemap;
 	const struct hmm_devmem_ops	*ops;
 	struct percpu_ref		ref;
-	dev_page_fault_t		page_fault;
 };
 
 /*

include/linux/memremap.h

@@ -80,6 +80,12 @@ struct dev_pagemap_ops {
 	 * Wait for refcount in struct dev_pagemap to be idle and reap it.
 	 */
 	void (*cleanup)(struct dev_pagemap *pgmap);
+
+	/*
+	 * Used for private (un-addressable) device memory only.  Must migrate
+	 * the page back to a CPU accessible page.
+	 */
+	vm_fault_t (*migrate_to_ram)(struct vm_fault *vmf);
 };
 
 /**

include/linux/swapops.h

@@ -129,12 +129,6 @@ static inline struct page *device_private_entry_to_page(swp_entry_t entry)
 {
 	return pfn_to_page(swp_offset(entry));
 }
-
-vm_fault_t device_private_entry_fault(struct vm_area_struct *vma,
-		       unsigned long addr,
-		       swp_entry_t entry,
-		       unsigned int flags,
-		       pmd_t *pmdp);
 #else /* CONFIG_DEVICE_PRIVATE */
 static inline swp_entry_t make_device_private_entry(struct page *page, bool write)
 {
@@ -164,15 +158,6 @@ static inline struct page *device_private_entry_to_page(swp_entry_t entry)
 {
 	return NULL;
 }
-
-static inline vm_fault_t device_private_entry_fault(struct vm_area_struct *vma,
-				     unsigned long addr,
-				     swp_entry_t entry,
-				     unsigned int flags,
-				     pmd_t *pmdp)
-{
-	return VM_FAULT_SIGBUS;
-}
 #endif /* CONFIG_DEVICE_PRIVATE */
 
 #ifdef CONFIG_MIGRATION

kernel/memremap.c

@@ -11,7 +11,6 @@
 #include <linux/types.h>
 #include <linux/wait_bit.h>
 #include <linux/xarray.h>
-#include <linux/hmm.h>
 
 static DEFINE_XARRAY(pgmap_array);
 #define SECTION_MASK ~((1UL << PA_SECTION_SHIFT) - 1)
@@ -46,36 +45,6 @@ static int devmap_managed_enable_get(struct device *dev, struct dev_pagemap *pgm
 }
 #endif /* CONFIG_DEV_PAGEMAP_OPS */
 
-#if IS_ENABLED(CONFIG_DEVICE_PRIVATE)
-vm_fault_t device_private_entry_fault(struct vm_area_struct *vma,
-		       unsigned long addr,
-		       swp_entry_t entry,
-		       unsigned int flags,
-		       pmd_t *pmdp)
-{
-	struct page *page = device_private_entry_to_page(entry);
-	struct hmm_devmem *devmem;
-
-	devmem = container_of(page->pgmap, typeof(*devmem), pagemap);
-
-	/*
-	 * The page_fault() callback must migrate page back to system memory
-	 * so that CPU can access it. This might fail for various reasons
-	 * (device issue, device was unsafely unplugged, ...). When such
-	 * error conditions happen, the callback must return VM_FAULT_SIGBUS.
-	 *
-	 * Note that because memory cgroup charges are accounted to the device
-	 * memory, this should never fail because of memory restrictions (but
-	 * allocation of regular system page might still fail because we are
-	 * out of memory).
-	 *
-	 * There is a more in-depth description of what that callback can and
-	 * cannot do, in include/linux/memremap.h
-	 */
-	return devmem->page_fault(vma, addr, page, flags, pmdp);
-}
-#endif /* CONFIG_DEVICE_PRIVATE */
-
 static void pgmap_array_delete(struct resource *res)
 {
 	xa_store_range(&pgmap_array, PHYS_PFN(res->start), PHYS_PFN(res->end),
@@ -193,6 +162,10 @@ void *devm_memremap_pages(struct device *dev, struct dev_pagemap *pgmap)
 			WARN(1, "Device private memory not supported\n");
 			return ERR_PTR(-EINVAL);
 		}
+		if (!pgmap->ops || !pgmap->ops->migrate_to_ram) {
+			WARN(1, "Missing migrate_to_ram method\n");
+			return ERR_PTR(-EINVAL);
+		}
 		break;
 	case MEMORY_DEVICE_FS_DAX:
 		if (!IS_ENABLED(CONFIG_ZONE_DEVICE) ||

mm/hmm.c

@@ -1366,15 +1366,12 @@ static void hmm_devmem_ref_kill(struct dev_pagemap *pgmap)
 	percpu_ref_kill(pgmap->ref);
 }
 
-static vm_fault_t hmm_devmem_fault(struct vm_area_struct *vma,
+static vm_fault_t hmm_devmem_migrate_to_ram(struct vm_fault *vmf)
-			    unsigned long addr,
-			    const struct page *page,
-			    unsigned int flags,
-			    pmd_t *pmdp)
 {
-	struct hmm_devmem *devmem = page->pgmap->data;
+	struct hmm_devmem *devmem = vmf->page->pgmap->data;
 
-	return devmem->ops->fault(devmem, vma, addr, page, flags, pmdp);
+	return devmem->ops->fault(devmem, vmf->vma, vmf->address, vmf->page,
+			vmf->flags, vmf->pmd);
 }
 
 static void hmm_devmem_free(struct page *page, void *data)
@@ -1388,6 +1385,7 @@ static const struct dev_pagemap_ops hmm_pagemap_ops = {
 	.page_free		= hmm_devmem_free,
 	.kill			= hmm_devmem_ref_kill,
 	.cleanup		= hmm_devmem_ref_exit,
+	.migrate_to_ram		= hmm_devmem_migrate_to_ram,
 };
 
 /*
@@ -1438,7 +1436,6 @@ struct hmm_devmem *hmm_devmem_add(const struct hmm_devmem_ops *ops,
 	devmem->pfn_first = devmem->resource->start >> PAGE_SHIFT;
 	devmem->pfn_last = devmem->pfn_first +
 			   (resource_size(devmem->resource) >> PAGE_SHIFT);
-	devmem->page_fault = hmm_devmem_fault;
 
 	devmem->pagemap.type = MEMORY_DEVICE_PRIVATE;
 	devmem->pagemap.res = *devmem->resource;

mm/memory.c

@@ -2748,13 +2748,8 @@ vm_fault_t do_swap_page(struct vm_fault *vmf)
 			migration_entry_wait(vma->vm_mm, vmf->pmd,
 					     vmf->address);
 		} else if (is_device_private_entry(entry)) {
-			/*
+			vmf->page = device_private_entry_to_page(entry);
-			 * For un-addressable device memory we call the pgmap
+			ret = vmf->page->pgmap->ops->migrate_to_ram(vmf);
-			 * fault handler callback. The callback must migrate
-			 * the page back to some CPU accessible page.
-			 */
-			ret = device_private_entry_fault(vma, vmf->address, entry,
-						 vmf->flags, vmf->pmd);
 		} else if (is_hwpoison_entry(entry)) {
 			ret = VM_FAULT_HWPOISON;
 		} else {