lkml.org 
[lkml]   [2006]   [Mar]   [6]   [last100]   RSS Feed
Views: [more markup]  [less markup]  [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Subject[PATCH] hugetlb_no_page might break hugetlb quota
From"Zhang, Yanmin" <>
DateMon, 06 Mar 2006 14:21:36 +0800
From: Zhang Yanmin <yanmin.zhang@intel.com>

In function hugetlb_no_page, backout path always calls hugetlb_put_quota.
It's incorrect when find_lock_page gets the page or the new page is added
into page cache.

In addition, if the vma->vm_flags doesn't include VM_SHARED, the quota
shouldn't be decreased.

My patch against 2.6.16-rc5-mm2 fixes it.

Signed-off-by: Zhang Yanmin <yanmin.zhang@intel.com>

---
diff -Nraup linux-2.6.16-rc5-mm2/mm/hugetlb.c linux-2.6.16-rc5-mm2_quota/mm/hugetlb.c
--- linux-2.6.16-rc5-mm2/mm/hugetlb.c	2006-03-06 18:48:18.000000000 +0800
+++ linux-2.6.16-rc5-mm2_quota/mm/hugetlb.c	2006-03-06 18:48:58.000000000 +0800
@@ -562,11 +562,12 @@ int hugetlb_no_page(struct mm_struct *mm
 retry:
 	page = find_lock_page(mapping, idx);
 	if (!page) {
-		if (hugetlb_get_quota(mapping))
+		if (vma->vm_flags & VM_SHARED && hugetlb_get_quota(mapping))
 			goto out;
 		page = alloc_huge_page(vma, address);
 		if (!page) {
-			hugetlb_put_quota(mapping);
+			if (vma->vm_flags & VM_SHARED)
+				hugetlb_put_quota(mapping);
 			ret = VM_FAULT_OOM;
 			goto out;
 		}
@@ -613,7 +614,6 @@ out:
 
 backout:
 	spin_unlock(&mm->page_table_lock);
-	hugetlb_put_quota(mapping);
 	unlock_page(page);
 	put_page(page);
 	goto out;

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2006-03-06 12:10    [from the cache]
©2003-2008