metag: hugetlbpage: Remove null pointer checks that could never happen
authorRickard Strandqvist <rickard_strandqvist@spectrumdigital.se>
Thu, 29 May 2014 23:11:58 +0000 (01:11 +0200)
committerJames Hogan <james.hogan@imgtec.com>
Mon, 16 Jun 2014 15:22:44 +0000 (16:22 +0100)
There are inconsistent NULL checks in
hugetlb_get_unmapped_area_existing(). The last conditional block in the
loop checks vma, but vma is dereferenced after that.

Due to the checks above however (bearing in mind after_huge is either 0
or 1) vma is guaranteed not to be NULL here, so remove that part of the
check.

Signed-off-by: Rickard Strandqvist <rickard_strandqvist@spectrumdigital.se>
[james.hogan@imgtec.com: remove brackets and expand commit message]
Signed-off-by: James Hogan <james.hogan@imgtec.com>
arch/metag/mm/hugetlbpage.c

index 3c52fa6..3c32075 100644 (file)
@@ -173,7 +173,7 @@ new_search:
                                mm->context.part_huge = 0;
                        return addr;
                }
-               if (vma && (vma->vm_flags & MAP_HUGETLB)) {
+               if (vma->vm_flags & MAP_HUGETLB) {
                        /* space after a huge vma in 2nd level page table? */
                        if (vma->vm_end & HUGEPT_MASK) {
                                after_huge = 1;