stable.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	stable@vger.kernel.org, Ursula Braun <ubraun@linux.ibm.com>,
	Karsten Graul <kgraul@linux.ibm.com>,
	"David S. Miller" <davem@davemloft.net>
Subject: [PATCH 5.5 45/46] net/smc: fix cleanup for linkgroup setup failures
Date: Tue,  7 Apr 2020 12:22:16 +0200	[thread overview]
Message-ID: <20200407101504.120133480@linuxfoundation.org> (raw)
In-Reply-To: <20200407101459.502593074@linuxfoundation.org>

From: Ursula Braun <ubraun@linux.ibm.com>

commit 51e3dfa8906ace90c809235b3d3afebc166b6433 upstream.

If an SMC connection to a certain peer is setup the first time,
a new linkgroup is created. In case of setup failures, such a
linkgroup is unusable and should disappear. As a first step the
linkgroup is removed from the linkgroup list in smc_lgr_forget().

There are 2 problems:
smc_listen_decline() might be called before linkgroup creation
resulting in a crash due to calling smc_lgr_forget() with
parameter NULL.
If a setup failure occurs after linkgroup creation, the connection
is never unregistered from the linkgroup, preventing linkgroup
freeing.

This patch introduces an enhanced smc_lgr_cleanup_early() function
which
* contains a linkgroup check for early smc_listen_decline()
  invocations
* invokes smc_conn_free() to guarantee unregistering of the
  connection.
* schedules fast linkgroup removal of the unusable linkgroup

And the unused function smcd_conn_free() is removed from smc_core.h.

Fixes: 3b2dec2603d5b ("net/smc: restructure client and server code in af_smc")
Fixes: 2a0674fffb6bc ("net/smc: improve abnormal termination of link groups")
Signed-off-by: Ursula Braun <ubraun@linux.ibm.com>
Signed-off-by: Karsten Graul <kgraul@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

---
 net/smc/af_smc.c   |   25 +++++++++++++++----------
 net/smc/smc_core.c |   12 ++++++++++++
 net/smc/smc_core.h |    2 +-
 3 files changed, 28 insertions(+), 11 deletions(-)

--- a/net/smc/af_smc.c
+++ b/net/smc/af_smc.c
@@ -512,15 +512,18 @@ static int smc_connect_decline_fallback(
 static int smc_connect_abort(struct smc_sock *smc, int reason_code,
 			     int local_contact)
 {
+	bool is_smcd = smc->conn.lgr->is_smcd;
+
 	if (local_contact == SMC_FIRST_CONTACT)
-		smc_lgr_forget(smc->conn.lgr);
-	if (smc->conn.lgr->is_smcd)
+		smc_lgr_cleanup_early(&smc->conn);
+	else
+		smc_conn_free(&smc->conn);
+	if (is_smcd)
 		/* there is only one lgr role for SMC-D; use server lock */
 		mutex_unlock(&smc_server_lgr_pending);
 	else
 		mutex_unlock(&smc_client_lgr_pending);
 
-	smc_conn_free(&smc->conn);
 	smc->connect_nonblock = 0;
 	return reason_code;
 }
@@ -1091,7 +1094,6 @@ static void smc_listen_out_err(struct sm
 	if (newsmcsk->sk_state == SMC_INIT)
 		sock_put(&new_smc->sk); /* passive closing */
 	newsmcsk->sk_state = SMC_CLOSED;
-	smc_conn_free(&new_smc->conn);
 
 	smc_listen_out(new_smc);
 }
@@ -1102,12 +1104,13 @@ static void smc_listen_decline(struct sm
 {
 	/* RDMA setup failed, switch back to TCP */
 	if (local_contact == SMC_FIRST_CONTACT)
-		smc_lgr_forget(new_smc->conn.lgr);
+		smc_lgr_cleanup_early(&new_smc->conn);
+	else
+		smc_conn_free(&new_smc->conn);
 	if (reason_code < 0) { /* error, no fallback possible */
 		smc_listen_out_err(new_smc);
 		return;
 	}
-	smc_conn_free(&new_smc->conn);
 	smc_switch_to_fallback(new_smc);
 	new_smc->fallback_rsn = reason_code;
 	if (reason_code && reason_code != SMC_CLC_DECL_PEERDECL) {
@@ -1170,16 +1173,18 @@ static int smc_listen_ism_init(struct sm
 			    new_smc->conn.lgr->vlan_id,
 			    new_smc->conn.lgr->smcd)) {
 		if (ini->cln_first_contact == SMC_FIRST_CONTACT)
-			smc_lgr_forget(new_smc->conn.lgr);
-		smc_conn_free(&new_smc->conn);
+			smc_lgr_cleanup_early(&new_smc->conn);
+		else
+			smc_conn_free(&new_smc->conn);
 		return SMC_CLC_DECL_SMCDNOTALK;
 	}
 
 	/* Create send and receive buffers */
 	if (smc_buf_create(new_smc, true)) {
 		if (ini->cln_first_contact == SMC_FIRST_CONTACT)
-			smc_lgr_forget(new_smc->conn.lgr);
-		smc_conn_free(&new_smc->conn);
+			smc_lgr_cleanup_early(&new_smc->conn);
+		else
+			smc_conn_free(&new_smc->conn);
 		return SMC_CLC_DECL_MEM;
 	}
 
--- a/net/smc/smc_core.c
+++ b/net/smc/smc_core.c
@@ -162,6 +162,18 @@ static void smc_lgr_unregister_conn(stru
 	conn->lgr = NULL;
 }
 
+void smc_lgr_cleanup_early(struct smc_connection *conn)
+{
+	struct smc_link_group *lgr = conn->lgr;
+
+	if (!lgr)
+		return;
+
+	smc_conn_free(conn);
+	smc_lgr_forget(lgr);
+	smc_lgr_schedule_free_work_fast(lgr);
+}
+
 /* Send delete link, either as client to request the initiation
  * of the DELETE LINK sequence from server; or as server to
  * initiate the delete processing. See smc_llc_rx_delete_link().
--- a/net/smc/smc_core.h
+++ b/net/smc/smc_core.h
@@ -296,6 +296,7 @@ struct smc_clc_msg_accept_confirm;
 struct smc_clc_msg_local;
 
 void smc_lgr_forget(struct smc_link_group *lgr);
+void smc_lgr_cleanup_early(struct smc_connection *conn);
 void smc_lgr_terminate(struct smc_link_group *lgr, bool soft);
 void smc_port_terminate(struct smc_ib_device *smcibdev, u8 ibport);
 void smc_smcd_terminate(struct smcd_dev *dev, u64 peer_gid,
@@ -316,7 +317,6 @@ int smc_vlan_by_tcpsk(struct socket *clc
 
 void smc_conn_free(struct smc_connection *conn);
 int smc_conn_create(struct smc_sock *smc, struct smc_init_info *ini);
-void smcd_conn_free(struct smc_connection *conn);
 void smc_lgr_schedule_free_work_fast(struct smc_link_group *lgr);
 int smc_core_init(void);
 void smc_core_exit(void);



  parent reply	other threads:[~2020-04-07 10:25 UTC|newest]

Thread overview: 50+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-04-07 10:21 [PATCH 5.5 00/46] 5.5.16-rc1 review Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 01/46] ipv4: fix a RCU-list lock in fib_triestat_seq_show Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 02/46] net: dsa: ksz: Select KSZ protocol tag Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 03/46] net, ip_tunnel: fix interface lookup with no key Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 04/46] sctp: fix possibly using a bad saddr with a given dst Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 05/46] sctp: fix refcount bug in sctp_wfree Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 06/46] net: macb: Fix handling of fixed-link node Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 07/46] nvme-rdma: Avoid double freeing of async event data Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 08/46] staging: wfx: fix warning about freeing in-use mutex during device unregister Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 09/46] drm/amdgpu: add fbdev suspend/resume on gpu reset Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 10/46] drm/amd/display: Add link_rate quirk for Apple 15" MBP 2017 Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 11/46] drm/bochs: downgrade pci_request_region failure from error to warning Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 12/46] initramfs: restore default compression behavior Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 13/46] drm/amdgpu: fix typo for vcn1 idle check Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 14/46] tools/power turbostat: Fix gcc build warnings Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 15/46] tools/power turbostat: Fix missing SYS_LPI counter on some Chromebooks Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 16/46] tools/power turbostat: Fix 32-bit capabilities warning Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 17/46] padata: fix uninitialized return value in padata_replace() Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 18/46] brcmfmac: abort and release host after error Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 19/46] bpf: Fix tnum constraints for 32-bit comparisons Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 20/46] XArray: Fix xa_find_next for large multi-index entries Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 21/46] misc: rtsx: set correct pcr_ops for rts522A Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 22/46] misc: pci_endpoint_test: Fix to support > 10 pci-endpoint-test devices Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 23/46] misc: pci_endpoint_test: Avoid using module parameter to determine irqtype Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 24/46] PCI: sysfs: Revert "rescan" file renames Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 25/46] coresight: do not use the BIT() macro in the UAPI header Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 26/46] mei: me: add cedar fork device ids Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 27/46] nvmem: check for NULL reg_read and reg_write before dereferencing Greg Kroah-Hartman
2020-04-07 10:21 ` [PATCH 5.5 28/46] nvmem: sprd: Fix the block lock operation Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.5 29/46] extcon: axp288: Add wakeup support Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.5 30/46] power: supply: axp288_charger: Add special handling for HP Pavilion x2 10 Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.5 31/46] Revert "dm: always call blk_queue_split() in dm_process_bio()" Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.5 32/46] ALSA: hda/ca0132 - Add Recon3Di quirk to handle integrated sound on EVGA X99 Classified motherboard Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.5 33/46] soc: mediatek: knows_txdone needs to be set in Mediatek CMDQ helper Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.5 34/46] net/mlx5e: kTLS, Fix wrong value in record tracker enum Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.5 35/46] iwlwifi: consider HE capability when setting LDPC Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.5 36/46] iwlwifi: yoyo: dont add TLV offset when reading FIFOs Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.5 37/46] iwlwifi: dbg: dont abort if sending DBGC_SUSPEND_RESUME fails Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.5 38/46] rxrpc: Fix sendmsg(MSG_WAITALL) handling Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.5 39/46] IB/hfi1: Ensure pq is not left on waitlist Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.5 40/46] sched: act: count in the size of action flags bitfield Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.5 41/46] tcp: fix TFO SYNACK undo to avoid double-timestamp-undo Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.5 42/46] i2c: i801: Do not add ICH_RES_IO_SMI for the iTCO_wdt device Greg Kroah-Hartman
2020-04-07 11:13   ` Wolfram Sang
2020-04-07 14:48     ` Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.5 43/46] net: genetlink: return the error code when attribute parsing fails Greg Kroah-Hartman
2020-04-07 10:22 ` [PATCH 5.5 44/46] net: Fix Tx hash bound checking Greg Kroah-Hartman
2020-04-07 10:22 ` Greg Kroah-Hartman [this message]
2020-04-07 10:22 ` [PATCH 5.5 46/46] padata: always acquire cpu_hotplug_lock before pinst->lock Greg Kroah-Hartman
2020-04-07 12:37 ` [PATCH 5.5 00/46] 5.5.16-rc1 review Jon Hunter

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200407101504.120133480@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=davem@davemloft.net \
    --cc=kgraul@linux.ibm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=stable@vger.kernel.org \
    --cc=ubraun@linux.ibm.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).