A couple of i_dir_seq fixes for fs/dcache.c
v2 cover A couple of i_dir_seq fixes for fs/dcache.c
Hi again,
This is a resend of the patches I previously posted here:
v1: http://lkml.kernel.org/r/[email protected]
v2: https://lkml.org/lkml/2018/2/19/360
Since this fixes a livelock that we're seeing in practice, I'd really like to get these merged if possible.
Thanks,
Will
--->8
Will Deacon (2):
fs: dcache: Avoid livelock between d_alloc_parallel and __d_add
fs: dcache: Use READ_ONCE when accessing i_dir_seq
fs/dcache.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-)
Signature · 1 line
2.1.4
PATCH RESEND v2 1/2 fs: dcache: Avoid livelock between d_alloc_parallel and __d_add
If d_alloc_parallel runs concurrently with __d_add, it is possible for d_alloc_parallel to continuously retry whilst i_dir_seq has been incremented to an odd value by __d_add:
CPU0:__d_add
n = start_dir_add(dir); cmpxchg(&dir->i_dir_seq, n, n + 1) == n
CPU1:d_alloc_parallel
retry:
seq = smp_load_acquire(&parent->d_inode->i_dir_seq) & ~1; hlist_bl_lock(b); bit_spin_lock(0, (unsigned long *)b); // Always succeeds
CPU0:
__d_lookup_done(dentry) hlist_bl_lock bit_spin_lock(0, (unsigned long *)b); // Never succeeds
CPU1:
if (unlikely(parent->d_inode->i_dir_seq != seq)) {
hlist_bl_unlock(b);
goto retry;
}Since the simple bit_spin_lock used to implement hlist_bl_lock does not provide any fairness guarantees, then CPU1 can starve CPU0 of the lock and prevent it from reaching end_dir_add(dir), therefore CPU1 cannot exit its retry loop because the sequence number always has the bottom bit set.
This patch resolves the livelock by not taking hlist_bl_lock in d_alloc_parallel if the sequence counter is odd, since any subsequent masked comparison with i_dir_seq will fail anyway.
--- fs/dcache.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-)
fs/dcache.c +7 -1
@@ -2474,7 +2474,7 @@ struct dentry *d_alloc_parallel(struct dentry *parent, retry:rcu_read_lock (); - seq =smp_load_acquire (&parent->d_inode->i_dir_seq) & ~1 ; + seq =smp_load_acquire (&parent->d_inode->i_dir_seq); r_seq =read_seqbegin (&rename_lock); dentry =__d_lookup_rcu (parent, name, &d_seq);if (unlikely (dentry)) {
@@ -2495,6 +2495,12 @@ struct dentry *d_alloc_parallel(struct dentry *parent,rcu_read_unlock (); goto retry; } + +if (unlikely (seq &1 )) { +rcu_read_unlock (); + goto retry; + } +hlist_bl_lock (b);if (unlikely (parent->d_inode->i_dir_seq != seq)) {hlist_bl_unlock (b);
Signature · 1 line
2.1.4
PATCH RESEND v2 2/2 fs: dcache: Use READ_ONCE when accessing i_dir_seq
In reply to Will Deaconi_dir_seq is subject to concurrent modification by a cmpxchg or store-release operation, so ensure that the relaxed access in d_alloc_parallel uses READ_ONCE.
--- fs/dcache.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
fs/dcache.c +1 -1
@@ -2502,7 +2502,7 @@ struct dentry *d_alloc_parallel(struct dentry *parent, }hlist_bl_lock (b); -if (unlikely (parent->d_inode->i_dir_seq != seq)) { +if (unlikely (READ_ONCE (parent->d_inode->i_dir_seq) != seq)) {hlist_bl_unlock (b);rcu_read_unlock (); goto retry;
Signature · 1 line
2.1.4