From 43159c9ec156e7363ba24528fced7a9d7b2f3134 Mon Sep 17 00:00:00 2001 From: David Howells Date: Mon, 7 Oct 2019 10:58:28 +0100 Subject: rxrpc: Fix call ref leak commit c48fc11b69e95007109206311b0187a3090591f3 upstream. When sendmsg() finds a call to continue on with, if the call is in an inappropriate state, it doesn't release the ref it just got on that call before returning an error. This causes the following symptom to show up with kasan: BUG: KASAN: use-after-free in rxrpc_send_keepalive+0x8a2/0x940 net/rxrpc/output.c:635 Read of size 8 at addr ffff888064219698 by task kworker/0:3/11077 where line 635 is: whdr.epoch = htonl(peer->local->rxnet->epoch); The local endpoint (which cannot be pinned by the call) has been released, but not the peer (which is pinned by the call). Fix this by releasing the call in the error path. Fixes: 37411cad633f ("rxrpc: Fix potential NULL-pointer exception") Reported-by: syzbot+d850c266e3df14da1d31@syzkaller.appspotmail.com Signed-off-by: David Howells Signed-off-by: Greg Kroah-Hartman --- net/rxrpc/sendmsg.c | 1 + 1 file changed, 1 insertion(+) (limited to 'net/rxrpc/sendmsg.c') diff --git a/net/rxrpc/sendmsg.c b/net/rxrpc/sendmsg.c index 016e293681b8..a980b49d7a4f 100644 --- a/net/rxrpc/sendmsg.c +++ b/net/rxrpc/sendmsg.c @@ -586,6 +586,7 @@ int rxrpc_do_sendmsg(struct rxrpc_sock *rx, struct msghdr *msg, size_t len) case RXRPC_CALL_SERVER_PREALLOC: case RXRPC_CALL_SERVER_SECURING: case RXRPC_CALL_SERVER_ACCEPTING: + rxrpc_put_call(call, rxrpc_call_put); ret = -EBUSY; goto error_release_sock; default: -- cgit v1.2.3