hugetlb: debit quota in alloc_huge_page
Adam Litke [Thu, 15 Nov 2007 00:59:39 +0000 (16:59 -0800)]
Now that quota is credited by free_huge_page(), calls to hugetlb_get_quota()
seem out of place.  The alloc/free API is unbalanced because we handle the
hugetlb_put_quota() but expect the caller to open-code hugetlb_get_quota().
Move the get inside alloc_huge_page to clean up this disparity.

This patch has been kept apart from the previous patch because of the somewhat
dodgy ERR_PTR() use herein.  Moving the quota logic means that
alloc_huge_page() has two failure modes.  Quota failure must result in a
SIGBUS while a standard allocation failure is OOM.  Unfortunately, ERR_PTR()
doesn't like the small positive errnos we have in VM_FAULT_* so they must be
negated before they are used.

Does anyone take issue with the way I am using PTR_ERR.  If so, what are your
thoughts on how to clean this up (without needing an if,else if,else block at
each alloc_huge_page() callsite)?

Signed-off-by: Adam Litke <agl@us.ibm.com>
Cc: Ken Chen <kenchen@google.com>
Cc: Andy Whitcroft <apw@shadowen.org>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: David Gibson <hermes@gibson.dropbear.id.au>
Cc: William Lee Irwin III <wli@holomorphy.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/hugetlb.c

index 3992bd5..bc12b0a 100644 (file)
@@ -388,6 +388,10 @@ static struct page *alloc_huge_page(struct vm_area_struct *vma,
                                    unsigned long addr)
 {
        struct page *page;
+       struct address_space *mapping = vma->vm_file->f_mapping;
+
+       if (hugetlb_get_quota(mapping))
+               return ERR_PTR(-VM_FAULT_SIGBUS);
 
        if (vma->vm_flags & VM_MAYSHARE)
                page = alloc_huge_page_shared(vma, addr);
@@ -395,9 +399,10 @@ static struct page *alloc_huge_page(struct vm_area_struct *vma,
                page = alloc_huge_page_private(vma, addr);
        if (page) {
                set_page_refcounted(page);
-               set_page_private(page, (unsigned long) vma->vm_file->f_mapping);
-       }
-       return page;
+               set_page_private(page, (unsigned long) mapping);
+               return page;
+       } else
+               return ERR_PTR(-VM_FAULT_OOM);
 }
 
 static int __init hugetlb_init(void)
@@ -737,15 +742,13 @@ static int hugetlb_cow(struct mm_struct *mm, struct vm_area_struct *vma,
                set_huge_ptep_writable(vma, address, ptep);
                return 0;
        }
-       if (hugetlb_get_quota(vma->vm_file->f_mapping))
-               return VM_FAULT_SIGBUS;
 
        page_cache_get(old_page);
        new_page = alloc_huge_page(vma, address);
 
-       if (!new_page) {
+       if (IS_ERR(new_page)) {
                page_cache_release(old_page);
-               return VM_FAULT_OOM;
+               return -PTR_ERR(new_page);
        }
 
        spin_unlock(&mm->page_table_lock);
@@ -789,12 +792,9 @@ retry:
                size = i_size_read(mapping->host) >> HPAGE_SHIFT;
                if (idx >= size)
                        goto out;
-               if (hugetlb_get_quota(mapping))
-                       goto out;
                page = alloc_huge_page(vma, address);
-               if (!page) {
-                       hugetlb_put_quota(mapping);
-                       ret = VM_FAULT_OOM;
+               if (IS_ERR(page)) {
+                       ret = -PTR_ERR(page);
                        goto out;
                }
                clear_huge_page(page, address);