From: Nikita Danilov <nikita@clusterfs.com>
Date: Wed, 9 Aug 2006 17:53:47 +0000 (-0400)
Subject: NFS: Fix a potential deadlock in nfs_release_page
X-Git-Tag: v2.6.18-rc5~58^2~14
X-Git-Url: https://err.no/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=ddeff520f02b92128132c282c350fa72afffb84a;p=linux-2.6

NFS: Fix a potential deadlock in nfs_release_page

nfs_wb_page() waits on request completion and, as a result, is not safe to be
called from nfs_release_page() invoked by VM scanner as part of GFP_NOFS
allocation. Fix possible deadlock by analyzing gfp mask and refusing to
release page if __GFP_FS is not set.

Signed-off-by: Nikita Danilov <danilov@gmail.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
(cherry picked from 374d969debfb290bafcb41d28918dc6f7e43ce31 commit)
---

diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index cc2b874ad5..48e892880d 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -312,7 +312,13 @@ static void nfs_invalidate_page(struct page *page, unsigned long offset)
 
 static int nfs_release_page(struct page *page, gfp_t gfp)
 {
-	return !nfs_wb_page(page->mapping->host, page);
+	if (gfp & __GFP_FS)
+		return !nfs_wb_page(page->mapping->host, page);
+	else
+		/*
+		 * Avoid deadlock on nfs_wait_on_request().
+		 */
+		return 0;
 }
 
 const struct address_space_operations nfs_file_aops = {