It's unlikely to ever occur, but if there were already a lease set on
the file then we could end up getting back a different pointer on a
successful setlease attempt than the one we allocated. If that happens,
the one we allocated could leak.
In practice, I don't think this will happen due to the fact that we only
try to set up the lease once per nfs4_file, but this error handling is a
bit more correct given the current lease API.
Cc: J. Bruce Fields <bfields@xxxxxxxxxxxx>
Signed-off-by: Jeff Layton <jlayton@xxxxxxxxxxxxxxx>
Reviewed-by: Christoph Hellwig <hch@xxxxxx>
---
fs/nfsd/nfs4state.c | 7 +++++--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index fd5ff4b17292..29fac18d9102 100644
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -3774,7 +3774,7 @@ static struct file_lock *nfs4_alloc_init_lease(struct nfs4_file *fp, int flag)
static int nfs4_setlease(struct nfs4_delegation *dp)
{
struct nfs4_file *fp = dp->dl_stid.sc_file;
- struct file_lock *fl;
+ struct file_lock *fl, *ret;
struct file *filp;
int status = 0;
@@ -3788,11 +3788,14 @@ static int nfs4_setlease(struct nfs4_delegation *dp)
return -EBADF;
}
fl->fl_file = filp;
- status = vfs_setlease(filp, fl->fl_type, &fl);
+ ret = fl;
+ status = vfs_setlease(filp, fl->fl_type, &ret);
if (status) {
locks_free_lock(fl);
goto out_fput;
}
+ if (ret != fl)
+ locks_free_lock(fl);
spin_lock(&state_lock);
spin_lock(&fp->fi_lock);
/* Did the lease get broken before we took the lock? */
--
1.9.3
--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
