dma: fix order calculation in dma_mark_declared_memory_occupied()

get_order() takes byte-sized input, not a page-granular one.

Irrespective of this fix I'm inclined to believe that this doesn't work
right anyway - bitmap_allocate_region() has an implicit assumption of
'pos' being suitable for 'order', which this function doesn't seem to
enforce (and since it's being called with a byte-granular value there's no
reason to believe that the callers would make sure device_addr is passed
accordingly - it's also not documented that way).

Signed-off-by: Jan Beulich <jbeulich@novell.com>
Cc: James E.J. Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Dmitry Baryshkov <dbaryshkov@gmail.com>
Cc: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

kernel/dma-coherent.c

@@ -77,15 +77,14 @@ void *dma_mark_declared_memory_occupied(struct device *dev,
 {
 	struct dma_coherent_mem *mem = dev->dma_mem;
 	int pos, err;
-	int pages = (size + (device_addr & ~PAGE_MASK) + PAGE_SIZE - 1);
 
-	pages >>= PAGE_SHIFT;
+	size += device_addr & ~PAGE_MASK;
 
 	if (!mem)
 		return ERR_PTR(-EINVAL);
 
 	pos = (device_addr - mem->device_base) >> PAGE_SHIFT;
-	err = bitmap_allocate_region(mem->bitmap, pos, get_order(pages));
+	err = bitmap_allocate_region(mem->bitmap, pos, get_order(size));
 	if (err != 0)
 		return ERR_PTR(err);
 	return mem->virt_base + (pos << PAGE_SHIFT);