[PATCH] hugetlb: check for brk() entering a hugepage region
authorDavid Gibson <david@gibson.dropbear.id.au>
Tue, 14 Nov 2006 10:03:38 +0000 (02:03 -0800)
committerLinus Torvalds <torvalds@woody.osdl.org>
Tue, 14 Nov 2006 17:09:27 +0000 (09:09 -0800)
Unlike mmap(), the codepath for brk() creates a vma without first checking
that it doesn't touch a region exclusively reserved for hugepages.  On
powerpc, this can allow it to create a normal page vma in a hugepage
region, causing oopses and other badness.

Add a test to prevent this.  With this patch, brk() will simply fail if it
attempts to move the break into a hugepage reserved region.

Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Cc: Adam Litke <agl@us.ibm.com>
Cc: Hugh Dickins <hugh@veritas.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
mm/mmap.c

index bdace87d7c019a75c0dbb26e29548bbd888869ed..2526463c99a786afc0758010faa4e2482178d350 100644 (file)
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -1880,6 +1880,10 @@ unsigned long do_brk(unsigned long addr, unsigned long len)
        if ((addr + len) > TASK_SIZE || (addr + len) < addr)
                return -EINVAL;
 
+       error = is_hugepage_only_range(current->mm, addr, len);
+       if (error)
+               return error;
+
        flags = VM_DATA_DEFAULT_FLAGS | VM_ACCOUNT | mm->def_flags;
 
        error = arch_mmap_check(addr, len, flags);