drm/vmwgfx: Support huge page faults

With vmwgfx dirty-tracking we need a specialized huge_fault
callback. Implement and hook it up.

Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Cc: "Jérôme Glisse" <jglisse@redhat.com>
Cc: "Christian König" <christian.koenig@amd.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Thomas Hellstrom (VMware) <thomas_os@shipmail.org>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Acked-by: Christian König <christian.koenig@amd.com>

drivers/gpu/drm/vmwgfx/vmwgfx_drv.h

@@ -1402,6 +1402,10 @@ void vmw_bo_dirty_unmap(struct vmw_buffer_object *vbo,
 			pgoff_t start, pgoff_t end);
 vm_fault_t vmw_bo_vm_fault(struct vm_fault *vmf);
 vm_fault_t vmw_bo_vm_mkwrite(struct vm_fault *vmf);
+#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+vm_fault_t vmw_bo_vm_huge_fault(struct vm_fault *vmf,
+				enum page_entry_size pe_size);
+#endif
 
 /**
  * VMW_DEBUG_KMS - Debug output for kernel mode-setting

drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c

@@ -473,7 +473,7 @@ vm_fault_t vmw_bo_vm_fault(struct vm_fault *vmf)
 	 * a lot of unnecessary write faults.
 	 */
 	if (vbo->dirty && vbo->dirty->method == VMW_BO_DIRTY_MKWRITE)
-		prot = vma->vm_page_prot;
+		prot = vm_get_page_prot(vma->vm_flags & ~VM_SHARED);
 	else
 		prot = vm_get_page_prot(vma->vm_flags);
 
@@ -486,3 +486,75 @@ vm_fault_t vmw_bo_vm_fault(struct vm_fault *vmf)
 
 	return ret;
 }
+
+#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+vm_fault_t vmw_bo_vm_huge_fault(struct vm_fault *vmf,
+				enum page_entry_size pe_size)
+{
+	struct vm_area_struct *vma = vmf->vma;
+	struct ttm_buffer_object *bo = (struct ttm_buffer_object *)
+	    vma->vm_private_data;
+	struct vmw_buffer_object *vbo =
+		container_of(bo, struct vmw_buffer_object, base);
+	pgprot_t prot;
+	vm_fault_t ret;
+	pgoff_t fault_page_size;
+	bool write = vmf->flags & FAULT_FLAG_WRITE;
+	bool is_cow_mapping =
+		(vma->vm_flags & (VM_SHARED | VM_MAYWRITE)) == VM_MAYWRITE;
+
+	switch (pe_size) {
+	case PE_SIZE_PMD:
+		fault_page_size = HPAGE_PMD_SIZE >> PAGE_SHIFT;
+		break;
+#ifdef CONFIG_HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD
+	case PE_SIZE_PUD:
+		fault_page_size = HPAGE_PUD_SIZE >> PAGE_SHIFT;
+		break;
+#endif
+	default:
+		WARN_ON_ONCE(1);
+		return VM_FAULT_FALLBACK;
+	}
+
+	/* Always do write dirty-tracking and COW on PTE level. */
+	if (write && (READ_ONCE(vbo->dirty) || is_cow_mapping))
+		return VM_FAULT_FALLBACK;
+
+	ret = ttm_bo_vm_reserve(bo, vmf);
+	if (ret)
+		return ret;
+
+	if (vbo->dirty) {
+		pgoff_t allowed_prefault;
+		unsigned long page_offset;
+
+		page_offset = vmf->pgoff -
+			drm_vma_node_start(&bo->base.vma_node);
+		if (page_offset >= bo->num_pages ||
+		    vmw_resources_clean(vbo, page_offset,
+					page_offset + PAGE_SIZE,
+					&allowed_prefault)) {
+			ret = VM_FAULT_SIGBUS;
+			goto out_unlock;
+		}
+
+		/*
+		 * Write protect, so we get a new fault on write, and can
+		 * split.
+		 */
+		prot = vm_get_page_prot(vma->vm_flags & ~VM_SHARED);
+	} else {
+		prot = vm_get_page_prot(vma->vm_flags);
+	}
+
+	ret = ttm_bo_vm_fault_reserved(vmf, prot, 1, fault_page_size);
+	if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))
+		return ret;
+
+out_unlock:
+	dma_resv_unlock(bo->base.resv);
+
+	return ret;
+}
+#endif

drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c

@@ -34,7 +34,10 @@ int vmw_mmap(struct file *filp, struct vm_area_struct *vma)
 		.page_mkwrite = vmw_bo_vm_mkwrite,
 		.fault = vmw_bo_vm_fault,
 		.open = ttm_bo_vm_open,
-		.close = ttm_bo_vm_close
+		.close = ttm_bo_vm_close,
+#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+		.huge_fault = vmw_bo_vm_huge_fault,
+#endif
 	};
 	struct drm_file *file_priv = filp->private_data;
 	struct vmw_private *dev_priv = vmw_priv(file_priv->minor->dev);