From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mx0b-001b2d01.pphosted.com ([148.163.158.5]:48942 "EHLO mx0a-001b2d01.pphosted.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1726445AbgIWIhN (ORCPT ); Wed, 23 Sep 2020 04:37:13 -0400 From: Julian Wiedmann Subject: [PATCH net-next 5/9] s390/qeth: tighten ucast IP locking Date: Wed, 23 Sep 2020 10:36:56 +0200 Message-Id: <20200923083700.44624-6-jwi@linux.ibm.com> In-Reply-To: <20200923083700.44624-1-jwi@linux.ibm.com> References: <20200923083700.44624-1-jwi@linux.ibm.com> List-ID: To: David Miller , Jakub Kicinski Cc: netdev , linux-s390 , Heiko Carstens , Ursula Braun , Karsten Graul , Julian Wiedmann The programming of ucast IPs via qeth_l3_modify_ip() is driven independently from any of our typical locking mechanisms (eg. detaching the netdevice, or holding the conf_mutex). So when we inspect the card state to check whether the required cmd IO should be deferred, there is no protection against concurrent state changes. But by slightly re-ordering the teardown sequence, we can rely on the ip_lock to sufficiently serialize things: 1. when running concurrently to qeth_l3_set_online(), any instance of qeth_l3_modify_ip() that aquires the ip_lock _after_ qeth_l3_recover_ip() will observe the state as CARD_STATE_SOFTSETUP and not defer the IO. 2. when running concurrently to qeth_l3_set_offline(), any instance of qeth_l3_modify_ip() that aquires the ip_lock _after_ qeth_l3_clear_ip_htable() will observe the state as CARD_STATE_DOWN and defer the IO. These guarantees in mind, we can now drop the conf_mutex from the qeth_l3_modify_rxip_vipa() wrapper. Signed-off-by: Julian Wiedmann --- drivers/s390/net/qeth_l3_main.c | 9 ++------- 1 file changed, 2 insertions(+), 7 deletions(-) diff --git a/drivers/s390/net/qeth_l3_main.c b/drivers/s390/net/qeth_l3_main.c index 876a21d451f5..0815b64c9797 100644 --- a/drivers/s390/net/qeth_l3_main.c +++ b/drivers/s390/net/qeth_l3_main.c @@ -591,7 +591,6 @@ int qeth_l3_modify_rxip_vipa(struct qeth_card *card, bool add, const u8 *ip, enum qeth_prot_versions proto) { struct qeth_ipaddr addr; - int rc; qeth_l3_init_ipaddr(&addr, type, proto); if (proto == QETH_PROT_IPV4) @@ -599,11 +598,7 @@ int qeth_l3_modify_rxip_vipa(struct qeth_card *card, bool add, const u8 *ip, else memcpy(&addr.u.a6.addr, ip, 16); - mutex_lock(&card->conf_mutex); - rc = qeth_l3_modify_ip(card, &addr, add); - mutex_unlock(&card->conf_mutex); - - return rc; + return qeth_l3_modify_ip(card, &addr, add); } int qeth_l3_modify_hsuid(struct qeth_card *card, bool add) @@ -1161,9 +1156,9 @@ static void qeth_l3_stop_card(struct qeth_card *card) qeth_diags_trace(card, QETH_DIAGS_CMD_TRACE_DISABLE); if (card->state == CARD_STATE_SOFTSETUP) { + card->state = CARD_STATE_DOWN; qeth_l3_clear_ip_htable(card, 1); qeth_clear_ipacmd_list(card); - card->state = CARD_STATE_DOWN; } qeth_qdio_clear_card(card, 0); -- 2.17.1