bootmem: Fix __free_pages_bootmem() to use @order properly
authorTejun Heo <tj@kernel.org>
Tue, 12 Jul 2011 07:58:06 +0000 (09:58 +0200)
committerH. Peter Anvin <hpa@linux.intel.com>
Wed, 13 Jul 2011 23:35:56 +0000 (16:35 -0700)
a226f6c899 (FRV: Clean up bootmem allocator's page freeing algorithm)
separated out __free_pages_bootmem() from free_all_bootmem_core().
__free_pages_bootmem() takes @order argument but it assumes @order is
either 0 or ilog2(BITS_PER_LONG).  Note that all the current users
match that assumption and this doesn't cause actual problems.

Fix it by using 1 << order instead of BITS_PER_LONG.

Signed-off-by: Tejun Heo <tj@kernel.org>
Link: http://lkml.kernel.org/r/1310457490-3356-3-git-send-email-tj@kernel.org
Cc: David Howells <dhowells@redhat.com>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
mm/page_alloc.c

index 9119faae6e6a94b0134f64268c03d05d985b3884..b6da6ed818a8fffd8f7f26bca1a0ab09028965ff 100644 (file)
@@ -705,10 +705,10 @@ void __meminit __free_pages_bootmem(struct page *page, unsigned int order)
                int loop;
 
                prefetchw(page);
-               for (loop = 0; loop < BITS_PER_LONG; loop++) {
+               for (loop = 0; loop < (1 << order); loop++) {
                        struct page *p = &page[loop];
 
-                       if (loop + 1 < BITS_PER_LONG)
+                       if (loop + 1 < (1 << order))
                                prefetchw(p + 1);
                        __ClearPageReserved(p);
                        set_page_count(p, 0);