From 0d737a8cfd8341d6ba8215443a3ddfe5ddd4e51a Mon Sep 17 00:00:00 2001 From: Marcelo Ricardo Leitner Date: Tue, 11 Aug 2015 19:22:21 -0300 Subject: [PATCH] dlm: fix race while closing connections When a connection have issues DLM may need to close it. Therefore we should also cancel pending workqueues for such connection at that time, and not just when dlm is not willing to use this connection anymore. Also, if we don't clear CF_CONNECT_PENDING flag, the error handling routines won't be able to re-connect as lowcomms_connect_sock() will check for it. Signed-off-by: Marcelo Ricardo Leitner Signed-off-by: David Teigland --- fs/dlm/lowcomms.c | 29 +++++++++++++++-------------- 1 file changed, 15 insertions(+), 14 deletions(-) diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c index bc04f5e3af7a..749deb3b69b2 100644 --- a/fs/dlm/lowcomms.c +++ b/fs/dlm/lowcomms.c @@ -514,17 +514,24 @@ static void make_sockaddr(struct sockaddr_storage *saddr, uint16_t port, } /* Close a remote connection and tidy up */ -static void close_connection(struct connection *con, bool and_other) +static void close_connection(struct connection *con, bool and_other, + bool tx, bool rx) { - mutex_lock(&con->sock_mutex); + clear_bit(CF_CONNECT_PENDING, &con->flags); + clear_bit(CF_WRITE_PENDING, &con->flags); + if (tx && cancel_work_sync(&con->swork)) + log_print("canceled swork for node %d", con->nodeid); + if (rx && cancel_work_sync(&con->rwork)) + log_print("canceled rwork for node %d", con->nodeid); + mutex_lock(&con->sock_mutex); if (con->sock) { sock_release(con->sock); con->sock = NULL; } if (con->othercon && and_other) { /* Will only re-enter once. */ - close_connection(con->othercon, false); + close_connection(con->othercon, false, true, true); } if (con->rx_page) { __free_page(con->rx_page); @@ -902,7 +909,7 @@ out_resched: out_close: mutex_unlock(&con->sock_mutex); if (ret != -EAGAIN) { - close_connection(con, false); + close_connection(con, false, true, false); /* Reconnect when there is something to send */ } /* Don't return success if we really got EOF */ @@ -1622,7 +1629,7 @@ out: send_error: mutex_unlock(&con->sock_mutex); - close_connection(con, false); + close_connection(con, false, false, true); lowcomms_connect_sock(con); return; @@ -1654,15 +1661,9 @@ int dlm_lowcomms_close(int nodeid) log_print("closing connection to node %d", nodeid); con = nodeid2con(nodeid, 0); if (con) { - clear_bit(CF_CONNECT_PENDING, &con->flags); - clear_bit(CF_WRITE_PENDING, &con->flags); set_bit(CF_CLOSE, &con->flags); - if (cancel_work_sync(&con->swork)) - log_print("canceled swork for node %d", nodeid); - if (cancel_work_sync(&con->rwork)) - log_print("canceled rwork for node %d", nodeid); + close_connection(con, true, true, true); clean_one_writequeue(con); - close_connection(con, true); } spin_lock(&dlm_node_addrs_spin); @@ -1745,7 +1746,7 @@ static void stop_conn(struct connection *con) static void free_conn(struct connection *con) { - close_connection(con, true); + close_connection(con, true, true, true); if (con->othercon) kmem_cache_free(con_cache, con->othercon); hlist_del(&con->list); @@ -1816,7 +1817,7 @@ fail_unlisten: dlm_allow_conn = 0; con = nodeid2con(0,0); if (con) { - close_connection(con, false); + close_connection(con, false, true, true); kmem_cache_free(con_cache, con); } fail_destroy: -- 2.20.1