linux-kernel - [PATCH 3.12 19/40] nfs: add memory barriers around NFS_INO_INVALID_DATA and NFS_INO

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <2d0270ca61c101531d9663fcb1d1b93d0cc9a05b.1396469191.git.jslaby@suse.cz>
Date:	Wed,  2 Apr 2014 22:13:05 +0200
From:	Jiri Slaby <jslaby@...e.cz>
To:	stable@...r.kernel.org
Cc:	linux-kernel@...r.kernel.org, Jeff Layton <jlayton@...hat.com>,
	Trond Myklebust <trond.myklebust@...marydata.com>,
	Jiri Slaby <jslaby@...e.cz>
Subject: [PATCH 3.12 19/40] nfs: add memory barriers around NFS_INO_INVALID_DATA and NFS_INO_INVALIDATING

From: Jeff Layton <jlayton@...hat.com>

3.12-stable review patch.  If anyone has any objections, please let me know.

===============

commit 4db72b40fdbc706f8957e9773ae73b1574b8c694 upstream.

If the setting of NFS_INO_INVALIDATING gets reordered to before the
clearing of NFS_INO_INVALID_DATA, then another task may hit a race
window where both appear to be clear, even though the inode's pages are
still in need of invalidation. Fix this by adding the appropriate memory
barriers.

Signed-off-by: Jeff Layton <jlayton@...hat.com>
Signed-off-by: Trond Myklebust <trond.myklebust@...marydata.com>
Signed-off-by: Jiri Slaby <jslaby@...e.cz>
---
 fs/nfs/dir.c   | 14 +++++++++++---
 fs/nfs/inode.c |  1 +
 fs/nfs/write.c |  1 +
 3 files changed, 13 insertions(+), 3 deletions(-)

diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 4d9f098614bc..140280623348 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -276,6 +276,15 @@ out_eof:
 	return -EBADCOOKIE;
 }
 
+static bool
+nfs_readdir_inode_mapping_valid(struct nfs_inode *nfsi)
+{
+	if (nfsi->cache_validity & (NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA))
+		return false;
+	smp_rmb();
+	return !test_bit(NFS_INO_INVALIDATING, &nfsi->flags);
+}
+
 static
 int nfs_readdir_search_for_cookie(struct nfs_cache_array *array, nfs_readdir_descriptor_t *desc)
 {
@@ -289,9 +298,8 @@ int nfs_readdir_search_for_cookie(struct nfs_cache_array *array, nfs_readdir_des
 			struct nfs_open_dir_context *ctx = desc->file->private_data;
 
 			new_pos = desc->current_index + i;
-			if (ctx->attr_gencount != nfsi->attr_gencount
-			    || (nfsi->cache_validity & (NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA))
-			    || test_bit(NFS_INO_INVALIDATING, &nfsi->flags)) {
+			if (ctx->attr_gencount != nfsi->attr_gencount ||
+			    !nfs_readdir_inode_mapping_valid(nfsi)) {
 				ctx->duped = 0;
 				ctx->attr_gencount = nfsi->attr_gencount;
 			} else if (new_pos < desc->ctx->pos) {
diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index 12c0fa8ae004..fdeeb28f287b 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -1060,6 +1060,7 @@ int nfs_revalidate_mapping(struct inode *inode, struct address_space *mapping)
 	}
 
 	set_bit(NFS_INO_INVALIDATING, bitlock);
+	smp_wmb();
 	nfsi->cache_validity &= ~NFS_INO_INVALID_DATA;
 	spin_unlock(&inode->i_lock);
 	trace_nfs_invalidate_mapping_enter(inode);
diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index d69e789f1908..c6aa89f92558 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -915,6 +915,7 @@ static bool nfs_write_pageuptodate(struct page *page, struct inode *inode)
 		goto out;
 	if (nfsi->cache_validity & (NFS_INO_INVALID_DATA|NFS_INO_REVAL_PAGECACHE))
 		return false;
+	smp_rmb();
 	if (test_bit(NFS_INO_INVALIDATING, &nfsi->flags))
 		return false;
 out:
-- 
1.9.1

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