[PATCH] nfs: fix resource leak and warning in svc_destroy on callback init error

[PATCH] nfs: fix resource leak and warning in svc_destroy on callback init error

[Hruday Rao]

From: Rahul Rao <shreeramrahul94@gmail.com>

During NFSv4 callback initialization, if creating the PF_INET TCP
listener succeeds but the subsequent PF_INET6 listener creation fails
with an error other than -EAFNOSUPPORT, the function bails out without
cleaning up the previously created IPv4 transport.

This leaves serv->sv_permsocks non-empty when svc_destroy() is called
during error unwinding, triggering a svc_destroy warning.

Fix this by invoking svc_xprt_destroy_all(serv, net, true) to properly
tear down active transports upon initialization failure.

Signed-off-by: Rahul Rao <shreeramrahul94@gmail.com>
---
 fs/nfs/callback.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c
index bc282b744f34..3d2ad5297297 100644
--- a/fs/nfs/callback.c
+++ b/fs/nfs/callback.c
@@ -61,9 +61,11 @@ static int nfs4_callback_up_net(struct svc_serv *serv, struct net *net)
 		dprintk("NFS: Callback listener port = %u (af %u, net %x)\n",
 			nn->nfs_callback_tcpport6, PF_INET6, net->ns.inum);
 	} else if (ret != -EAFNOSUPPORT)
-		goto out_err;
+		goto out_destroy_xprt;
 	return 0;
 
+out_destroy_xprt:
+	svc_xprt_destroy_all(serv, net, true);
 out_err:
 	return (ret) ? ret : -ENOMEM;
 }
-- 
2.43.0