dma-remap: align the size in dma_common_*_remap()

Running a guest with a virtio-iommu protecting virtio devices
is broken since commit 515e5b6d ("dma-mapping: use vmap insted
of reimplementing it"). Before the conversion, the size was
page aligned in __get_vm_area_node(). Doing so fixes the
regression.

Fixes: 515e5b6d ("dma-mapping: use vmap insted of reimplementing it")
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>

kernel/dma/remap.c

@@ -24,7 +24,8 @@ void *dma_common_pages_remap(struct page **pages, size_t size,
 {
 	void *vaddr;
 
-	vaddr = vmap(pages, size >> PAGE_SHIFT, VM_DMA_COHERENT, prot);
+	vaddr = vmap(pages, PAGE_ALIGN(size) >> PAGE_SHIFT,
+		     VM_DMA_COHERENT, prot);
 	if (vaddr)
 		find_vm_area(vaddr)->pages = pages;
 	return vaddr;
@@ -37,7 +38,7 @@ void *dma_common_pages_remap(struct page **pages, size_t size,
 void *dma_common_contiguous_remap(struct page *page, size_t size,
 			pgprot_t prot, const void *caller)
 {
-	int count = size >> PAGE_SHIFT;
+	int count = PAGE_ALIGN(size) >> PAGE_SHIFT;
 	struct page **pages;
 	void *vaddr;
 	int i;