NFSv4.1: Fix a race in pNFS layoutcommit

We need to clear the NFS_LSEG_LAYOUTCOMMIT bits atomically with the
NFS_INO_LAYOUTCOMMIT bit, otherwise we may end up with situations
where the two are out of sync.
The first half of the problem is to ensure that pnfs_layoutcommit_inode
clears the NFS_LSEG_LAYOUTCOMMIT bit through pnfs_list_write_lseg.
We still need to keep the reference to those segments until the RPC call
is finished, so in order to make it clear _where_ those references come
from, we add a helper pnfs_list_write_lseg_done() that cleans up after
pnfs_list_write_lseg.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Acked-by: Benny Halevy <bhalevy@tonian.com>
Cc: stable@vger.kernel.org

2 files changed, 18 insertions, 15 deletions

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index b2671cb0f901..6ccdd4fd9b59 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -6416,22 +6416,8 @@ nfs4_layoutcommit_done(struct rpc_task *task, void *calldata)
 static void nfs4_layoutcommit_release(void *calldata)
 {
         struct nfs4_layoutcommit_data *data = calldata;
-        struct pnfs_layout_segment *lseg, *tmp;
-        unsigned long *bitlock = &NFS_I(data->args.inode)->flags;
 
         pnfs_cleanup_layoutcommit(data);
-        /* Matched by references in pnfs_set_layoutcommit */
-        list_for_each_entry_safe(lseg, tmp, &data->lseg_list, pls_lc_list) {
-                list_del_init(&lseg->pls_lc_list);
-                if (test_and_clear_bit(NFS_LSEG_LAYOUTCOMMIT,
-                                       &lseg->pls_flags))
-                        pnfs_put_lseg(lseg);
-        }
-
-        clear_bit_unlock(NFS_INO_LAYOUTCOMMITTING, bitlock);
-        smp_mb__after_clear_bit();
-        wake_up_bit(bitlock, NFS_INO_LAYOUTCOMMITTING);
-
         put_rpccred(data->cred);
         kfree(data);
 }
diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c
index 48ac5aad6258..3d900916fd41 100644
--- a/fs/nfs/pnfs.c
+++ b/fs/nfs/pnfs.c
@@ -1746,11 +1746,27 @@ static void pnfs_list_write_lseg(struct inode *inode, struct list_head *listp)
 
         list_for_each_entry(lseg, &NFS_I(inode)->layout->plh_segs, pls_list) {
                 if (lseg->pls_range.iomode == IOMODE_RW &&
-                    test_bit(NFS_LSEG_LAYOUTCOMMIT, &lseg->pls_flags))
+                    test_and_clear_bit(NFS_LSEG_LAYOUTCOMMIT, &lseg->pls_flags))
                         list_add(&lseg->pls_lc_list, listp);
         }
 }
 
+static void pnfs_list_write_lseg_done(struct inode *inode, struct list_head *listp)
+{
+        struct pnfs_layout_segment *lseg, *tmp;
+        unsigned long *bitlock = &NFS_I(inode)->flags;
+
+        /* Matched by references in pnfs_set_layoutcommit */
+        list_for_each_entry_safe(lseg, tmp, listp, pls_lc_list) {
+                list_del_init(&lseg->pls_lc_list);
+                pnfs_put_lseg(lseg);
+        }
+
+        clear_bit_unlock(NFS_INO_LAYOUTCOMMITTING, bitlock);
+        smp_mb__after_clear_bit();
+        wake_up_bit(bitlock, NFS_INO_LAYOUTCOMMITTING);
+}
+
 void pnfs_set_lo_fail(struct pnfs_layout_segment *lseg)
 {
         pnfs_layout_io_set_failed(lseg->pls_layout, lseg->pls_range.iomode);
@@ -1795,6 +1811,7 @@ void pnfs_cleanup_layoutcommit(struct nfs4_layoutcommit_data *data)
 
         if (nfss->pnfs_curr_ld->cleanup_layoutcommit)
                 nfss->pnfs_curr_ld->cleanup_layoutcommit(data);
+        pnfs_list_write_lseg_done(data->args.inode, &data->lseg_list);
 }
 
 /*