summaryrefslogtreecommitdiffstats
path: root/fs/ext4/super.c
diff options
context:
space:
mode:
authorTheodore Ts'o <tytso@mit.edu>2009-05-01 08:50:38 -0400
committerTheodore Ts'o <tytso@mit.edu>2009-05-01 08:50:38 -0400
commit8df9675f8b498d0bfa1f0b5b06f56bf1ff366dd5 (patch)
tree38fd56a82049f50b4d774af47b9d39f116071755 /fs/ext4/super.c
parent9ca92389c5312a51e819c15c762f0abdc7f3129b (diff)
downloadop-kernel-dev-8df9675f8b498d0bfa1f0b5b06f56bf1ff366dd5.zip
op-kernel-dev-8df9675f8b498d0bfa1f0b5b06f56bf1ff366dd5.tar.gz
ext4: Avoid races caused by on-line resizing and SMP memory reordering
Ext4's on-line resizing adds a new block group and then, only at the last step adjusts s_groups_count. However, it's possible on SMP systems that another CPU could see the updated the s_group_count and not see the newly initialized data structures for the just-added block group. For this reason, it's important to insert a SMP read barrier after reading s_groups_count and before reading any (for example) the new block group descriptors allowed by the increased value of s_groups_count. Unfortunately, we rather blatently violate this locking protocol documented in fs/ext4/resize.c. Fortunately, (1) on-line resizes happen relatively rarely, and (2) it seems rare that the filesystem code will immediately try to use just-added block group before any memory ordering issues resolve themselves. So apparently problems here are relatively hard to hit, since ext3 has been vulnerable to the same issue for years with no one apparently complaining. Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Diffstat (limited to 'fs/ext4/super.c')
-rw-r--r--fs/ext4/super.c3
1 files changed, 1 insertions, 2 deletions
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 68c3a44..fcd7b24 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -3557,9 +3557,8 @@ static int ext4_statfs(struct dentry *dentry, struct kstatfs *buf)
if (test_opt(sb, MINIX_DF)) {
sbi->s_overhead_last = 0;
} else if (sbi->s_blocks_last != ext4_blocks_count(es)) {
- ext4_group_t ngroups = sbi->s_groups_count, i;
+ ext4_group_t i, ngroups = ext4_get_groups_count(sb);
ext4_fsblk_t overhead = 0;
- smp_rmb();
/*
* Compute the overhead (FS structures). This is constant
OpenPOWER on IntegriCloud