linux/net/sctp/associola.c

1699 lines
48 KiB
C
Raw Normal View History

/* SCTP kernel implementation
* (C) Copyright IBM Corp. 2001, 2004
* Copyright (c) 1999-2000 Cisco, Inc.
* Copyright (c) 1999-2001 Motorola, Inc.
* Copyright (c) 2001 Intel Corp.
* Copyright (c) 2001 La Monte H.P. Yarroll
*
* This file is part of the SCTP kernel implementation
*
* This module provides the abstraction for an SCTP association.
*
* This SCTP implementation is free software;
* you can redistribute it and/or modify it under the terms of
* the GNU General Public License as published by
* the Free Software Foundation; either version 2, or (at your option)
* any later version.
*
* This SCTP implementation is distributed in the hope that it
* will be useful, but WITHOUT ANY WARRANTY; without even the implied
* ************************
* warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
* See the GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with GNU CC; see the file COPYING. If not, see
* <http://www.gnu.org/licenses/>.
*
* Please send any bug reports or fixes you make to the
* email address(es):
* lksctp developers <linux-sctp@vger.kernel.org>
*
* Written or modified by:
* La Monte H.P. Yarroll <piggy@acm.org>
* Karl Knutson <karl@athena.chicago.il.us>
* Jon Grimm <jgrimm@us.ibm.com>
* Xingang Guo <xingang.guo@intel.com>
* Hui Huang <hui.huang@nokia.com>
* Sridhar Samudrala <sri@us.ibm.com>
* Daisy Chang <daisyc@us.ibm.com>
* Ryan Layer <rmlayer@us.ibm.com>
* Kevin Gao <kevin.gao@intel.com>
*/
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
#include <linux/types.h>
#include <linux/fcntl.h>
#include <linux/poll.h>
#include <linux/init.h>
#include <linux/slab.h>
#include <linux/in.h>
#include <net/ipv6.h>
#include <net/sctp/sctp.h>
#include <net/sctp/sm.h>
/* Forward declarations for internal functions. */
static void sctp_select_active_and_retran_path(struct sctp_association *asoc);
static void sctp_assoc_bh_rcv(struct work_struct *work);
static void sctp_assoc_free_asconf_acks(struct sctp_association *asoc);
static void sctp_assoc_free_asconf_queue(struct sctp_association *asoc);
/* 1st Level Abstractions. */
/* Initialize a new association from provided memory. */
static struct sctp_association *sctp_association_init(struct sctp_association *asoc,
const struct sctp_endpoint *ep,
const struct sock *sk,
sctp_scope_t scope,
gfp_t gfp)
{
struct net *net = sock_net(sk);
struct sctp_sock *sp;
int i;
sctp_paramhdr_t *p;
int err;
/* Retrieve the SCTP per socket area. */
sp = sctp_sk((struct sock *)sk);
/* Discarding const is appropriate here. */
asoc->ep = (struct sctp_endpoint *)ep;
asoc->base.sk = (struct sock *)sk;
sctp_endpoint_hold(asoc->ep);
sock_hold(asoc->base.sk);
/* Initialize the common base substructure. */
asoc->base.type = SCTP_EP_TYPE_ASSOCIATION;
/* Initialize the object handling fields. */
atomic_set(&asoc->base.refcnt, 1);
/* Initialize the bind addr area. */
sctp_bind_addr_init(&asoc->base.bind_addr, ep->base.bind_addr.port);
asoc->state = SCTP_STATE_CLOSED;
asoc->cookie_life = ms_to_ktime(sp->assocparams.sasoc_cookie_life);
asoc->user_frag = sp->user_frag;
/* Set the association max_retrans and RTO values from the
* socket values.
*/
asoc->max_retrans = sp->assocparams.sasoc_asocmaxrxt;
asoc->pf_retrans = net->sctp.pf_retrans;
asoc->rto_initial = msecs_to_jiffies(sp->rtoinfo.srto_initial);
asoc->rto_max = msecs_to_jiffies(sp->rtoinfo.srto_max);
asoc->rto_min = msecs_to_jiffies(sp->rtoinfo.srto_min);
/* Initialize the association's heartbeat interval based on the
* sock configured value.
*/
asoc->hbinterval = msecs_to_jiffies(sp->hbinterval);
/* Initialize path max retrans value. */
asoc->pathmaxrxt = sp->pathmaxrxt;
/* Initialize default path MTU. */
asoc->pathmtu = sp->pathmtu;
/* Set association default SACK delay */
asoc->sackdelay = msecs_to_jiffies(sp->sackdelay);
asoc->sackfreq = sp->sackfreq;
/* Set the association default flags controlling
* Heartbeat, SACK delay, and Path MTU Discovery.
*/
asoc->param_flags = sp->param_flags;
/* Initialize the maximum number of new data packets that can be sent
* in a burst.
*/
asoc->max_burst = sp->max_burst;
/* initialize association timers */
asoc->timeouts[SCTP_EVENT_TIMEOUT_T1_COOKIE] = asoc->rto_initial;
asoc->timeouts[SCTP_EVENT_TIMEOUT_T1_INIT] = asoc->rto_initial;
asoc->timeouts[SCTP_EVENT_TIMEOUT_T2_SHUTDOWN] = asoc->rto_initial;
/* sctpimpguide Section 2.12.2
* If the 'T5-shutdown-guard' timer is used, it SHOULD be set to the
* recommended value of 5 times 'RTO.Max'.
*/
asoc->timeouts[SCTP_EVENT_TIMEOUT_T5_SHUTDOWN_GUARD]
= 5 * asoc->rto_max;
asoc->timeouts[SCTP_EVENT_TIMEOUT_SACK] = asoc->sackdelay;
asoc->timeouts[SCTP_EVENT_TIMEOUT_AUTOCLOSE] = sp->autoclose * HZ;
/* Initializes the timers */
for (i = SCTP_EVENT_TIMEOUT_NONE; i < SCTP_NUM_TIMEOUT_TYPES; ++i)
setup_timer(&asoc->timers[i], sctp_timer_events[i],
(unsigned long)asoc);
/* Pull default initialization values from the sock options.
* Note: This assumes that the values have already been
* validated in the sock.
*/
asoc->c.sinit_max_instreams = sp->initmsg.sinit_max_instreams;
asoc->c.sinit_num_ostreams = sp->initmsg.sinit_num_ostreams;
asoc->max_init_attempts = sp->initmsg.sinit_max_attempts;
asoc->max_init_timeo =
msecs_to_jiffies(sp->initmsg.sinit_max_init_timeo);
/* Set the local window size for receive.
* This is also the rcvbuf space per association.
* RFC 6 - A SCTP receiver MUST be able to receive a minimum of
* 1500 bytes in one SCTP packet.
*/
if ((sk->sk_rcvbuf/2) < SCTP_DEFAULT_MINWINDOW)
asoc->rwnd = SCTP_DEFAULT_MINWINDOW;
else
asoc->rwnd = sk->sk_rcvbuf/2;
asoc->a_rwnd = asoc->rwnd;
/* Use my own max window until I learn something better. */
asoc->peer.rwnd = SCTP_DEFAULT_MAXWINDOW;
/* Initialize the receive memory counter */
atomic_set(&asoc->rmem_alloc, 0);
init_waitqueue_head(&asoc->wait);
asoc->c.my_vtag = sctp_generate_tag(ep);
asoc->c.my_port = ep->base.bind_addr.port;
asoc->c.initial_tsn = sctp_generate_tsn(ep);
asoc->next_tsn = asoc->c.initial_tsn;
asoc->ctsn_ack_point = asoc->next_tsn - 1;
asoc->adv_peer_ack_point = asoc->ctsn_ack_point;
asoc->highest_sacked = asoc->ctsn_ack_point;
asoc->last_cwr_tsn = asoc->ctsn_ack_point;
/* ADDIP Section 4.1 Asconf Chunk Procedures
*
* When an endpoint has an ASCONF signaled change to be sent to the
* remote endpoint it should do the following:
* ...
* A2) a serial number should be assigned to the chunk. The serial
* number SHOULD be a monotonically increasing number. The serial
* numbers SHOULD be initialized at the start of the
* association to the same value as the initial TSN.
*/
asoc->addip_serial = asoc->c.initial_tsn;
INIT_LIST_HEAD(&asoc->addip_chunk_list);
INIT_LIST_HEAD(&asoc->asconf_ack_list);
/* Make an empty list of remote transport addresses. */
INIT_LIST_HEAD(&asoc->peer.transport_addr_list);
/* RFC 2960 5.1 Normal Establishment of an Association
*
* After the reception of the first data chunk in an
* association the endpoint must immediately respond with a
* sack to acknowledge the data chunk. Subsequent
* acknowledgements should be done as described in Section
* 6.2.
*
* [We implement this by telling a new association that it
* already received one packet.]
*/
asoc->peer.sack_needed = 1;
sctp: be more restrictive in transport selection on bundled sacks It was noticed recently that when we send data on a transport, its possible that we might bundle a sack that arrived on a different transport. While this isn't a major problem, it does go against the SHOULD requirement in section 6.4 of RFC 2960: An endpoint SHOULD transmit reply chunks (e.g., SACK, HEARTBEAT ACK, etc.) to the same destination transport address from which it received the DATA or control chunk to which it is replying. This rule should also be followed if the endpoint is bundling DATA chunks together with the reply chunk. This patch seeks to correct that. It restricts the bundling of sack operations to only those transports which have moved the ctsn of the association forward since the last sack. By doing this we guarantee that we only bundle outbound saks on a transport that has received a chunk since the last sack. This brings us into stricter compliance with the RFC. Vlad had initially suggested that we strictly allow only sack bundling on the transport that last moved the ctsn forward. While this makes sense, I was concerned that doing so prevented us from bundling in the case where we had received chunks that moved the ctsn on multiple transports. In those cases, the RFC allows us to select any of the transports having received chunks to bundle the sack on. so I've modified the approach to allow for that, by adding a state variable to each transport that tracks weather it has moved the ctsn since the last sack. This I think keeps our behavior (and performance), close enough to our current profile that I think we can do this without a sysctl knob to enable/disable it. Signed-off-by: Neil Horman <nhorman@tuxdriver.com> CC: Vlad Yaseivch <vyasevich@gmail.com> CC: David S. Miller <davem@davemloft.net> CC: linux-sctp@vger.kernel.org Reported-by: Michele Baldessari <michele@redhat.com> Reported-by: sorin serban <sserban@redhat.com> Acked-by: Vlad Yasevich <vyasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-06-30 05:04:26 +02:00
asoc->peer.sack_generation = 1;
/* Assume that the peer will tell us if he recognizes ASCONF
* as part of INIT exchange.
* The sctp_addip_noauth option is there for backward compatibility
* and will revert old behavior.
*/
if (net->sctp.addip_noauth)
asoc->peer.asconf_capable = 1;
/* Create an input queue. */
sctp_inq_init(&asoc->base.inqueue);
sctp_inq_set_th_handler(&asoc->base.inqueue, sctp_assoc_bh_rcv);
/* Create an output queue. */
sctp_outq_init(asoc, &asoc->outqueue);
if (!sctp_ulpq_init(&asoc->ulpq, asoc))
goto fail_init;
/* Assume that peer would support both address types unless we are
* told otherwise.
*/
asoc->peer.ipv4_address = 1;
if (asoc->base.sk->sk_family == PF_INET6)
asoc->peer.ipv6_address = 1;
INIT_LIST_HEAD(&asoc->asocs);
asoc->default_stream = sp->default_stream;
asoc->default_ppid = sp->default_ppid;
asoc->default_flags = sp->default_flags;
asoc->default_context = sp->default_context;
asoc->default_timetolive = sp->default_timetolive;
asoc->default_rcv_context = sp->default_rcv_context;
/* AUTH related initializations */
INIT_LIST_HEAD(&asoc->endpoint_shared_keys);
err = sctp_auth_asoc_copy_shkeys(ep, asoc, gfp);
if (err)
goto fail_init;
asoc->active_key_id = ep->active_key_id;
/* Save the hmacs and chunks list into this association */
if (ep->auth_hmacs_list)
memcpy(asoc->c.auth_hmacs, ep->auth_hmacs_list,
ntohs(ep->auth_hmacs_list->param_hdr.length));
if (ep->auth_chunk_list)
memcpy(asoc->c.auth_chunks, ep->auth_chunk_list,
ntohs(ep->auth_chunk_list->param_hdr.length));
/* Get the AUTH random number for this association */
p = (sctp_paramhdr_t *)asoc->c.auth_random;
p->type = SCTP_PARAM_RANDOM;
p->length = htons(sizeof(sctp_paramhdr_t) + SCTP_AUTH_RANDOM_LENGTH);
get_random_bytes(p+1, SCTP_AUTH_RANDOM_LENGTH);
return asoc;
fail_init:
sock_put(asoc->base.sk);
sctp_endpoint_put(asoc->ep);
return NULL;
}
/* Allocate and initialize a new association */
struct sctp_association *sctp_association_new(const struct sctp_endpoint *ep,
const struct sock *sk,
sctp_scope_t scope,
gfp_t gfp)
{
struct sctp_association *asoc;
asoc = kzalloc(sizeof(*asoc), gfp);
if (!asoc)
goto fail;
if (!sctp_association_init(asoc, ep, sk, scope, gfp))
goto fail_init;
SCTP_DBG_OBJCNT_INC(assoc);
net: sctp: rework debugging framework to use pr_debug and friends We should get rid of all own SCTP debug printk macros and use the ones that the kernel offers anyway instead. This makes the code more readable and conform to the kernel code, and offers all the features of dynamic debbuging that pr_debug() et al has, such as only turning on/off portions of debug messages at runtime through debugfs. The runtime cost of having CONFIG_DYNAMIC_DEBUG enabled, but none of the debug statements printing, is negligible [1]. If kernel debugging is completly turned off, then these statements will also compile into "empty" functions. While we're at it, we also need to change the Kconfig option as it /now/ only refers to the ifdef'ed code portions in outqueue.c that enable further debugging/tracing of SCTP transaction fields. Also, since SCTP_ASSERT code was enabled with this Kconfig option and has now been removed, we transform those code parts into WARNs resp. where appropriate BUG_ONs so that those bugs can be more easily detected as probably not many people have SCTP debugging permanently turned on. To turn on all SCTP debugging, the following steps are needed: # mount -t debugfs none /sys/kernel/debug # echo -n 'module sctp +p' > /sys/kernel/debug/dynamic_debug/control This can be done more fine-grained on a per file, per line basis and others as described in [2]. [1] https://www.kernel.org/doc/ols/2009/ols2009-pages-39-46.pdf [2] Documentation/dynamic-debug-howto.txt Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-06-28 19:49:40 +02:00
pr_debug("Created asoc %p\n", asoc);
return asoc;
fail_init:
kfree(asoc);
fail:
return NULL;
}
/* Free this association if possible. There may still be users, so
* the actual deallocation may be delayed.
*/
void sctp_association_free(struct sctp_association *asoc)
{
struct sock *sk = asoc->base.sk;
struct sctp_transport *transport;
struct list_head *pos, *temp;
int i;
/* Only real associations count against the endpoint, so
* don't bother for if this is a temporary association.
*/
if (!list_empty(&asoc->asocs)) {
list_del(&asoc->asocs);
/* Decrement the backlog value for a TCP-style listening
* socket.
*/
if (sctp_style(sk, TCP) && sctp_sstate(sk, LISTENING))
sk->sk_ack_backlog--;
}
/* Mark as dead, so other users can know this structure is
* going away.
*/
asoc->base.dead = true;
/* Dispose of any data lying around in the outqueue. */
sctp_outq_free(&asoc->outqueue);
/* Dispose of any pending messages for the upper layer. */
sctp_ulpq_free(&asoc->ulpq);
/* Dispose of any pending chunks on the inqueue. */
sctp_inq_free(&asoc->base.inqueue);
sctp_tsnmap_free(&asoc->peer.tsn_map);
/* Free ssnmap storage. */
sctp_ssnmap_free(asoc->ssnmap);
/* Clean up the bound address list. */
sctp_bind_addr_free(&asoc->base.bind_addr);
/* Do we need to go through all of our timers and
* delete them? To be safe we will try to delete all, but we
* should be able to go through and make a guess based
* on our state.
*/
for (i = SCTP_EVENT_TIMEOUT_NONE; i < SCTP_NUM_TIMEOUT_TYPES; ++i) {
if (del_timer(&asoc->timers[i]))
sctp_association_put(asoc);
}
/* Free peer's cached cookie. */
kfree(asoc->peer.cookie);
kfree(asoc->peer.peer_random);
kfree(asoc->peer.peer_chunks);
kfree(asoc->peer.peer_hmacs);
/* Release the transport structures. */
list_for_each_safe(pos, temp, &asoc->peer.transport_addr_list) {
transport = list_entry(pos, struct sctp_transport, transports);
list_del_rcu(pos);
sctp_unhash_transport(transport);
sctp_transport_free(transport);
}
asoc->peer.transport_count = 0;
sctp_asconf_queue_teardown(asoc);
/* Free pending address space being deleted */
kfree(asoc->asconf_addr_del_pending);
/* AUTH - Free the endpoint shared keys */
sctp_auth_destroy_keys(&asoc->endpoint_shared_keys);
/* AUTH - Free the association shared key */
sctp_auth_key_put(asoc->asoc_shared_key);
sctp_association_put(asoc);
}
/* Cleanup and free up an association. */
static void sctp_association_destroy(struct sctp_association *asoc)
{
net: sctp: rework debugging framework to use pr_debug and friends We should get rid of all own SCTP debug printk macros and use the ones that the kernel offers anyway instead. This makes the code more readable and conform to the kernel code, and offers all the features of dynamic debbuging that pr_debug() et al has, such as only turning on/off portions of debug messages at runtime through debugfs. The runtime cost of having CONFIG_DYNAMIC_DEBUG enabled, but none of the debug statements printing, is negligible [1]. If kernel debugging is completly turned off, then these statements will also compile into "empty" functions. While we're at it, we also need to change the Kconfig option as it /now/ only refers to the ifdef'ed code portions in outqueue.c that enable further debugging/tracing of SCTP transaction fields. Also, since SCTP_ASSERT code was enabled with this Kconfig option and has now been removed, we transform those code parts into WARNs resp. where appropriate BUG_ONs so that those bugs can be more easily detected as probably not many people have SCTP debugging permanently turned on. To turn on all SCTP debugging, the following steps are needed: # mount -t debugfs none /sys/kernel/debug # echo -n 'module sctp +p' > /sys/kernel/debug/dynamic_debug/control This can be done more fine-grained on a per file, per line basis and others as described in [2]. [1] https://www.kernel.org/doc/ols/2009/ols2009-pages-39-46.pdf [2] Documentation/dynamic-debug-howto.txt Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-06-28 19:49:40 +02:00
if (unlikely(!asoc->base.dead)) {
WARN(1, "Attempt to destroy undead association %p!\n", asoc);
return;
}
sctp_endpoint_put(asoc->ep);
sock_put(asoc->base.sk);
if (asoc->assoc_id != 0) {
spin_lock_bh(&sctp_assocs_id_lock);
idr_remove(&sctp_assocs_id, asoc->assoc_id);
spin_unlock_bh(&sctp_assocs_id_lock);
}
WARN_ON(atomic_read(&asoc->rmem_alloc));
kfree(asoc);
SCTP_DBG_OBJCNT_DEC(assoc);
}
/* Change the primary destination address for the peer. */
void sctp_assoc_set_primary(struct sctp_association *asoc,
struct sctp_transport *transport)
{
int changeover = 0;
/* it's a changeover only if we already have a primary path
* that we are changing
*/
if (asoc->peer.primary_path != NULL &&
asoc->peer.primary_path != transport)
changeover = 1 ;
asoc->peer.primary_path = transport;
/* Set a default msg_name for events. */
memcpy(&asoc->peer.primary_addr, &transport->ipaddr,
sizeof(union sctp_addr));
/* If the primary path is changing, assume that the
* user wants to use this new path.
*/
if ((transport->state == SCTP_ACTIVE) ||
(transport->state == SCTP_UNKNOWN))
asoc->peer.active_path = transport;
/*
* SFR-CACC algorithm:
* Upon the receipt of a request to change the primary
* destination address, on the data structure for the new
* primary destination, the sender MUST do the following:
*
* 1) If CHANGEOVER_ACTIVE is set, then there was a switch
* to this destination address earlier. The sender MUST set
* CYCLING_CHANGEOVER to indicate that this switch is a
* double switch to the same destination address.
*
* Really, only bother is we have data queued or outstanding on
* the association.
*/
if (!asoc->outqueue.outstanding_bytes && !asoc->outqueue.out_qlen)
return;
if (transport->cacc.changeover_active)
transport->cacc.cycling_changeover = changeover;
/* 2) The sender MUST set CHANGEOVER_ACTIVE to indicate that
* a changeover has occurred.
*/
transport->cacc.changeover_active = changeover;
/* 3) The sender MUST store the next TSN to be sent in
* next_tsn_at_change.
*/
transport->cacc.next_tsn_at_change = asoc->next_tsn;
}
/* Remove a transport from an association. */
void sctp_assoc_rm_peer(struct sctp_association *asoc,
struct sctp_transport *peer)
{
struct list_head *pos;
struct sctp_transport *transport;
net: sctp: rework debugging framework to use pr_debug and friends We should get rid of all own SCTP debug printk macros and use the ones that the kernel offers anyway instead. This makes the code more readable and conform to the kernel code, and offers all the features of dynamic debbuging that pr_debug() et al has, such as only turning on/off portions of debug messages at runtime through debugfs. The runtime cost of having CONFIG_DYNAMIC_DEBUG enabled, but none of the debug statements printing, is negligible [1]. If kernel debugging is completly turned off, then these statements will also compile into "empty" functions. While we're at it, we also need to change the Kconfig option as it /now/ only refers to the ifdef'ed code portions in outqueue.c that enable further debugging/tracing of SCTP transaction fields. Also, since SCTP_ASSERT code was enabled with this Kconfig option and has now been removed, we transform those code parts into WARNs resp. where appropriate BUG_ONs so that those bugs can be more easily detected as probably not many people have SCTP debugging permanently turned on. To turn on all SCTP debugging, the following steps are needed: # mount -t debugfs none /sys/kernel/debug # echo -n 'module sctp +p' > /sys/kernel/debug/dynamic_debug/control This can be done more fine-grained on a per file, per line basis and others as described in [2]. [1] https://www.kernel.org/doc/ols/2009/ols2009-pages-39-46.pdf [2] Documentation/dynamic-debug-howto.txt Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-06-28 19:49:40 +02:00
pr_debug("%s: association:%p addr:%pISpc\n",
__func__, asoc, &peer->ipaddr.sa);
/* If we are to remove the current retran_path, update it
* to the next peer before removing this peer from the list.
*/
if (asoc->peer.retran_path == peer)
sctp_assoc_update_retran_path(asoc);
/* Remove this peer from the list. */
list_del_rcu(&peer->transports);
/* Remove this peer from the transport hashtable */
sctp_unhash_transport(peer);
/* Get the first transport of asoc. */
pos = asoc->peer.transport_addr_list.next;
transport = list_entry(pos, struct sctp_transport, transports);
/* Update any entries that match the peer to be deleted. */
if (asoc->peer.primary_path == peer)
sctp_assoc_set_primary(asoc, transport);
if (asoc->peer.active_path == peer)
asoc->peer.active_path = transport;
if (asoc->peer.retran_path == peer)
asoc->peer.retran_path = transport;
if (asoc->peer.last_data_from == peer)
asoc->peer.last_data_from = transport;
/* If we remove the transport an INIT was last sent to, set it to
* NULL. Combined with the update of the retran path above, this
* will cause the next INIT to be sent to the next available
* transport, maintaining the cycle.
*/
if (asoc->init_last_sent_to == peer)
asoc->init_last_sent_to = NULL;
/* If we remove the transport an SHUTDOWN was last sent to, set it
* to NULL. Combined with the update of the retran path above, this
* will cause the next SHUTDOWN to be sent to the next available
* transport, maintaining the cycle.
*/
if (asoc->shutdown_last_sent_to == peer)
asoc->shutdown_last_sent_to = NULL;
/* If we remove the transport an ASCONF was last sent to, set it to
* NULL.
*/
if (asoc->addip_last_asconf &&
asoc->addip_last_asconf->transport == peer)
asoc->addip_last_asconf->transport = NULL;
/* If we have something on the transmitted list, we have to
* save it off. The best place is the active path.
*/
if (!list_empty(&peer->transmitted)) {
struct sctp_transport *active = asoc->peer.active_path;
struct sctp_chunk *ch;
/* Reset the transport of each chunk on this list */
list_for_each_entry(ch, &peer->transmitted,
transmitted_list) {
ch->transport = NULL;
ch->rtt_in_progress = 0;
}
list_splice_tail_init(&peer->transmitted,
&active->transmitted);
/* Start a T3 timer here in case it wasn't running so
* that these migrated packets have a chance to get
* retransmitted.
*/
if (!timer_pending(&active->T3_rtx_timer))
if (!mod_timer(&active->T3_rtx_timer,
jiffies + active->rto))
sctp_transport_hold(active);
}
asoc->peer.transport_count--;
sctp_transport_free(peer);
}
/* Add a transport address to an association. */
struct sctp_transport *sctp_assoc_add_peer(struct sctp_association *asoc,
const union sctp_addr *addr,
const gfp_t gfp,
const int peer_state)
{
struct net *net = sock_net(asoc->base.sk);
struct sctp_transport *peer;
struct sctp_sock *sp;
unsigned short port;
sp = sctp_sk(asoc->base.sk);
/* AF_INET and AF_INET6 share common port field. */
port = ntohs(addr->v4.sin_port);
net: sctp: rework debugging framework to use pr_debug and friends We should get rid of all own SCTP debug printk macros and use the ones that the kernel offers anyway instead. This makes the code more readable and conform to the kernel code, and offers all the features of dynamic debbuging that pr_debug() et al has, such as only turning on/off portions of debug messages at runtime through debugfs. The runtime cost of having CONFIG_DYNAMIC_DEBUG enabled, but none of the debug statements printing, is negligible [1]. If kernel debugging is completly turned off, then these statements will also compile into "empty" functions. While we're at it, we also need to change the Kconfig option as it /now/ only refers to the ifdef'ed code portions in outqueue.c that enable further debugging/tracing of SCTP transaction fields. Also, since SCTP_ASSERT code was enabled with this Kconfig option and has now been removed, we transform those code parts into WARNs resp. where appropriate BUG_ONs so that those bugs can be more easily detected as probably not many people have SCTP debugging permanently turned on. To turn on all SCTP debugging, the following steps are needed: # mount -t debugfs none /sys/kernel/debug # echo -n 'module sctp +p' > /sys/kernel/debug/dynamic_debug/control This can be done more fine-grained on a per file, per line basis and others as described in [2]. [1] https://www.kernel.org/doc/ols/2009/ols2009-pages-39-46.pdf [2] Documentation/dynamic-debug-howto.txt Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-06-28 19:49:40 +02:00
pr_debug("%s: association:%p addr:%pISpc state:%d\n", __func__,
asoc, &addr->sa, peer_state);
/* Set the port if it has not been set yet. */
if (0 == asoc->peer.port)
asoc->peer.port = port;
/* Check to see if this is a duplicate. */
peer = sctp_assoc_lookup_paddr(asoc, addr);
if (peer) {
/* An UNKNOWN state is only set on transports added by
* user in sctp_connectx() call. Such transports should be
* considered CONFIRMED per RFC 4960, Section 5.4.
*/
if (peer->state == SCTP_UNKNOWN) {
peer->state = SCTP_ACTIVE;
}
return peer;
}
peer = sctp_transport_new(net, addr, gfp);
if (!peer)
return NULL;
sctp_transport_set_owner(peer, asoc);
/* Initialize the peer's heartbeat interval based on the
* association configured value.
*/
peer->hbinterval = asoc->hbinterval;
/* Set the path max_retrans. */
peer->pathmaxrxt = asoc->pathmaxrxt;
/* And the partial failure retrans threshold */
peer->pf_retrans = asoc->pf_retrans;
/* Initialize the peer's SACK delay timeout based on the
* association configured value.
*/
peer->sackdelay = asoc->sackdelay;
peer->sackfreq = asoc->sackfreq;
/* Enable/disable heartbeat, SACK delay, and path MTU discovery
* based on association setting.
*/
peer->param_flags = asoc->param_flags;
sctp_transport_route(peer, NULL, sp);
/* Initialize the pmtu of the transport. */
if (peer->param_flags & SPP_PMTUD_DISABLE) {
if (asoc->pathmtu)
peer->pathmtu = asoc->pathmtu;
else
peer->pathmtu = SCTP_DEFAULT_MAXSEGMENT;
}
/* If this is the first transport addr on this association,
* initialize the association PMTU to the peer's PMTU.
* If not and the current association PMTU is higher than the new
* peer's PMTU, reset the association PMTU to the new peer's PMTU.
*/
if (asoc->pathmtu)
asoc->pathmtu = min_t(int, peer->pathmtu, asoc->pathmtu);
else
asoc->pathmtu = peer->pathmtu;
net: sctp: rework debugging framework to use pr_debug and friends We should get rid of all own SCTP debug printk macros and use the ones that the kernel offers anyway instead. This makes the code more readable and conform to the kernel code, and offers all the features of dynamic debbuging that pr_debug() et al has, such as only turning on/off portions of debug messages at runtime through debugfs. The runtime cost of having CONFIG_DYNAMIC_DEBUG enabled, but none of the debug statements printing, is negligible [1]. If kernel debugging is completly turned off, then these statements will also compile into "empty" functions. While we're at it, we also need to change the Kconfig option as it /now/ only refers to the ifdef'ed code portions in outqueue.c that enable further debugging/tracing of SCTP transaction fields. Also, since SCTP_ASSERT code was enabled with this Kconfig option and has now been removed, we transform those code parts into WARNs resp. where appropriate BUG_ONs so that those bugs can be more easily detected as probably not many people have SCTP debugging permanently turned on. To turn on all SCTP debugging, the following steps are needed: # mount -t debugfs none /sys/kernel/debug # echo -n 'module sctp +p' > /sys/kernel/debug/dynamic_debug/control This can be done more fine-grained on a per file, per line basis and others as described in [2]. [1] https://www.kernel.org/doc/ols/2009/ols2009-pages-39-46.pdf [2] Documentation/dynamic-debug-howto.txt Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-06-28 19:49:40 +02:00
pr_debug("%s: association:%p PMTU set to %d\n", __func__, asoc,
asoc->pathmtu);
sctp: Prevent uninitialized memory access valgrind reports uninizialized memory accesses when running sctp inside the network simulation cradle simulator: Conditional jump or move depends on uninitialised value(s) at 0x570E34A: sctp_assoc_sync_pmtu (associola.c:1324) by 0x57427DA: sctp_packet_transmit (output.c:403) by 0x5710EFF: sctp_outq_flush (outqueue.c:824) by 0x5710B88: sctp_outq_uncork (outqueue.c:701) by 0x5745262: sctp_cmd_interpreter (sm_sideeffect.c:1548) by 0x57444B7: sctp_side_effects (sm_sideeffect.c:976) by 0x5744460: sctp_do_sm (sm_sideeffect.c:945) by 0x572157D: sctp_primitive_ASSOCIATE (primitive.c:94) by 0x5725C04: __sctp_connect (socket.c:1094) by 0x57297DC: sctp_connect (socket.c:3297) Conditional jump or move depends on uninitialised value(s) at 0x575D3A5: mod_timer (timer.c:630) by 0x5752B78: sctp_cmd_hb_timers_start (sm_sideeffect.c:555) by 0x5754133: sctp_cmd_interpreter (sm_sideeffect.c:1448) by 0x5753607: sctp_side_effects (sm_sideeffect.c:976) by 0x57535B0: sctp_do_sm (sm_sideeffect.c:945) by 0x571E9AE: sctp_endpoint_bh_rcv (endpointola.c:474) by 0x573347F: sctp_inq_push (inqueue.c:104) by 0x572EF93: sctp_rcv (input.c:256) by 0x5689623: ip_local_deliver_finish (ip_input.c:230) by 0x5689759: ip_local_deliver (ip_input.c:268) by 0x5689CAC: ip_rcv_finish (dst.h:246) #1 is due to "if (t->pmtu_pending)". 8a4794914f9cf2681235ec2311e189fe307c28c7 "[SCTP] Flag a pmtu change request" suggests it should be initialized to 0. #2 is the heartbeat timer 'expires' value, which is uninizialised, but test by mod_timer(). T3_rtx_timer seems to be affected by the same problem, so initialize it, too. Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: Vlad Yasevich <vladislav.yasevich@hp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-07-19 08:04:39 +02:00
peer->pmtu_pending = 0;
asoc->frag_point = sctp_frag_point(asoc, asoc->pathmtu);
/* The asoc->peer.port might not be meaningful yet, but
* initialize the packet structure anyway.
*/
sctp_packet_init(&peer->packet, peer, asoc->base.bind_addr.port,
asoc->peer.port);
/* 7.2.1 Slow-Start
*
* o The initial cwnd before DATA transmission or after a sufficiently
* long idle period MUST be set to
* min(4*MTU, max(2*MTU, 4380 bytes))
*
* o The initial value of ssthresh MAY be arbitrarily high
* (for example, implementations MAY use the size of the
* receiver advertised window).
*/
peer->cwnd = min(4*asoc->pathmtu, max_t(__u32, 2*asoc->pathmtu, 4380));
/* At this point, we may not have the receiver's advertised window,
* so initialize ssthresh to the default value and it will be set
* later when we process the INIT.
*/
peer->ssthresh = SCTP_DEFAULT_MAXWINDOW;
peer->partial_bytes_acked = 0;
peer->flight_size = 0;
peer->burst_limited = 0;
/* Set the transport's RTO.initial value */
peer->rto = asoc->rto_initial;
sctp: Add support to per-association statistics via a new SCTP_GET_ASSOC_STATS call The current SCTP stack is lacking a mechanism to have per association statistics. This is an implementation modeled after OpenSolaris' SCTP_GET_ASSOC_STATS. Userspace part will follow on lksctp if/when there is a general ACK on this. V4: - Move ipackets++ before q->immediate.func() for consistency reasons - Move sctp_max_rto() at the end of sctp_transport_update_rto() to avoid returning bogus RTO values - return asoc->rto_min when max_obs_rto value has not changed V3: - Increase ictrlchunks in sctp_assoc_bh_rcv() as well - Move ipackets++ to sctp_inq_push() - return 0 when no rto updates took place since the last call V2: - Implement partial retrieval of stat struct to cope for future expansion - Kill the rtxpackets counter as it cannot be precise anyway - Rename outseqtsns to outofseqtsns to make it clearer that these are out of sequence unexpected TSNs - Move asoc->ipackets++ under a lock to avoid potential miscounts - Fold asoc->opackets++ into the already existing asoc check - Kill unneeded (q->asoc) test when increasing rtxchunks - Do not count octrlchunks if sending failed (SCTP_XMIT_OK != 0) - Don't count SHUTDOWNs as SACKs - Move SCTP_GET_ASSOC_STATS to the private space API - Adjust the len check in sctp_getsockopt_assoc_stats() to allow for future struct growth - Move association statistics in their own struct - Update idupchunks when we send a SACK with dup TSNs - return min_rto in max_rto when RTO has not changed. Also return the transport when max_rto last changed. Signed-off: Michele Baldessari <michele@acksyn.org> Acked-by: Vlad Yasevich <vyasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-12-01 05:49:42 +01:00
sctp_max_rto(asoc, peer);
/* Set the peer's active state. */
peer->state = peer_state;
/* Attach the remote transport to our asoc. */
list_add_tail_rcu(&peer->transports, &asoc->peer.transport_addr_list);
asoc->peer.transport_count++;
/* Add this peer into the transport hashtable */
sctp_hash_transport(peer);
/* If we do not yet have a primary path, set one. */
if (!asoc->peer.primary_path) {
sctp_assoc_set_primary(asoc, peer);
asoc->peer.retran_path = peer;
}
if (asoc->peer.active_path == asoc->peer.retran_path &&
peer->state != SCTP_UNCONFIRMED) {
asoc->peer.retran_path = peer;
}
return peer;
}
/* Delete a transport address from an association. */
void sctp_assoc_del_peer(struct sctp_association *asoc,
const union sctp_addr *addr)
{
struct list_head *pos;
struct list_head *temp;
struct sctp_transport *transport;
list_for_each_safe(pos, temp, &asoc->peer.transport_addr_list) {
transport = list_entry(pos, struct sctp_transport, transports);
if (sctp_cmp_addr_exact(addr, &transport->ipaddr)) {
/* Do book keeping for removing the peer and free it. */
sctp_assoc_rm_peer(asoc, transport);
break;
}
}
}
/* Lookup a transport by address. */
struct sctp_transport *sctp_assoc_lookup_paddr(
const struct sctp_association *asoc,
const union sctp_addr *address)
{
struct sctp_transport *t;
/* Cycle through all transports searching for a peer address. */
list_for_each_entry(t, &asoc->peer.transport_addr_list,
transports) {
if (sctp_cmp_addr_exact(address, &t->ipaddr))
return t;
}
return NULL;
}
/* Remove all transports except a give one */
void sctp_assoc_del_nonprimary_peers(struct sctp_association *asoc,
struct sctp_transport *primary)
{
struct sctp_transport *temp;
struct sctp_transport *t;
list_for_each_entry_safe(t, temp, &asoc->peer.transport_addr_list,
transports) {
/* if the current transport is not the primary one, delete it */
if (t != primary)
sctp_assoc_rm_peer(asoc, t);
}
}
/* Engage in transport control operations.
* Mark the transport up or down and send a notification to the user.
* Select and update the new active and retran paths.
*/
void sctp_assoc_control_transport(struct sctp_association *asoc,
struct sctp_transport *transport,
sctp_transport_cmd_t command,
sctp_sn_error_t error)
{
struct sctp_ulpevent *event;
struct sockaddr_storage addr;
int spc_state = 0;
bool ulp_notify = true;
/* Record the transition on the transport. */
switch (command) {
case SCTP_TRANSPORT_UP:
/* If we are moving from UNCONFIRMED state due
* to heartbeat success, report the SCTP_ADDR_CONFIRMED
* state to the user, otherwise report SCTP_ADDR_AVAILABLE.
*/
if (SCTP_UNCONFIRMED == transport->state &&
SCTP_HEARTBEAT_SUCCESS == error)
spc_state = SCTP_ADDR_CONFIRMED;
else
spc_state = SCTP_ADDR_AVAILABLE;
/* Don't inform ULP about transition from PF to
* active state and set cwnd to 1 MTU, see SCTP
* Quick failover draft section 5.1, point 5
*/
if (transport->state == SCTP_PF) {
ulp_notify = false;
transport->cwnd = asoc->pathmtu;
}
transport->state = SCTP_ACTIVE;
break;
case SCTP_TRANSPORT_DOWN:
/* If the transport was never confirmed, do not transition it
* to inactive state. Also, release the cached route since
* there may be a better route next time.
*/
if (transport->state != SCTP_UNCONFIRMED)
transport->state = SCTP_INACTIVE;
else {
dst_release(transport->dst);
transport->dst = NULL;
ulp_notify = false;
}
spc_state = SCTP_ADDR_UNREACHABLE;
break;
case SCTP_TRANSPORT_PF:
transport->state = SCTP_PF;
ulp_notify = false;
break;
default:
return;
}
/* Generate and send a SCTP_PEER_ADDR_CHANGE notification
* to the user.
*/
if (ulp_notify) {
memset(&addr, 0, sizeof(struct sockaddr_storage));
memcpy(&addr, &transport->ipaddr,
transport->af_specific->sockaddr_len);
event = sctp_ulpevent_make_peer_addr_change(asoc, &addr,
0, spc_state, error, GFP_ATOMIC);
if (event)
sctp_ulpq_tail_event(&asoc->ulpq, event);
}
/* Select new active and retran paths. */
sctp_select_active_and_retran_path(asoc);
}
/* Hold a reference to an association. */
void sctp_association_hold(struct sctp_association *asoc)
{
atomic_inc(&asoc->base.refcnt);
}
/* Release a reference to an association and cleanup
* if there are no more references.
*/
void sctp_association_put(struct sctp_association *asoc)
{
if (atomic_dec_and_test(&asoc->base.refcnt))
sctp_association_destroy(asoc);
}
/* Allocate the next TSN, Transmission Sequence Number, for the given
* association.
*/
__u32 sctp_association_get_next_tsn(struct sctp_association *asoc)
{
/* From Section 1.6 Serial Number Arithmetic:
* Transmission Sequence Numbers wrap around when they reach
* 2**32 - 1. That is, the next TSN a DATA chunk MUST use
* after transmitting TSN = 2*32 - 1 is TSN = 0.
*/
__u32 retval = asoc->next_tsn;
asoc->next_tsn++;
asoc->unack_data++;
return retval;
}
/* Compare two addresses to see if they match. Wildcard addresses
* only match themselves.
*/
int sctp_cmp_addr_exact(const union sctp_addr *ss1,
const union sctp_addr *ss2)
{
struct sctp_af *af;
af = sctp_get_af_specific(ss1->sa.sa_family);
if (unlikely(!af))
return 0;
return af->cmp_addr(ss1, ss2);
}
/* Return an ecne chunk to get prepended to a packet.
* Note: We are sly and return a shared, prealloced chunk. FIXME:
* No we don't, but we could/should.
*/
struct sctp_chunk *sctp_get_ecne_prepend(struct sctp_association *asoc)
{
if (!asoc->need_ecne)
return NULL;
/* Send ECNE if needed.
* Not being able to allocate a chunk here is not deadly.
*/
return sctp_make_ecne(asoc, asoc->last_ecne_tsn);
}
/*
* Find which transport this TSN was sent on.
*/
struct sctp_transport *sctp_assoc_lookup_tsn(struct sctp_association *asoc,
__u32 tsn)
{
struct sctp_transport *active;
struct sctp_transport *match;
struct sctp_transport *transport;
struct sctp_chunk *chunk;
__be32 key = htonl(tsn);
match = NULL;
/*
* FIXME: In general, find a more efficient data structure for
* searching.
*/
/*
* The general strategy is to search each transport's transmitted
* list. Return which transport this TSN lives on.
*
* Let's be hopeful and check the active_path first.
* Another optimization would be to know if there is only one
* outbound path and not have to look for the TSN at all.
*
*/
active = asoc->peer.active_path;
list_for_each_entry(chunk, &active->transmitted,
transmitted_list) {
if (key == chunk->subh.data_hdr->tsn) {
match = active;
goto out;
}
}
/* If not found, go search all the other transports. */
list_for_each_entry(transport, &asoc->peer.transport_addr_list,
transports) {
if (transport == active)
continue;
list_for_each_entry(chunk, &transport->transmitted,
transmitted_list) {
if (key == chunk->subh.data_hdr->tsn) {
match = transport;
goto out;
}
}
}
out:
return match;
}
/* Is this the association we are looking for? */
struct sctp_transport *sctp_assoc_is_match(struct sctp_association *asoc,
struct net *net,
const union sctp_addr *laddr,
const union sctp_addr *paddr)
{
struct sctp_transport *transport;
if ((htons(asoc->base.bind_addr.port) == laddr->v4.sin_port) &&
(htons(asoc->peer.port) == paddr->v4.sin_port) &&
net_eq(sock_net(asoc->base.sk), net)) {
transport = sctp_assoc_lookup_paddr(asoc, paddr);
if (!transport)
goto out;
if (sctp_bind_addr_match(&asoc->base.bind_addr, laddr,
sctp_sk(asoc->base.sk)))
goto out;
}
transport = NULL;
out:
return transport;
}
/* Do delayed input processing. This is scheduled by sctp_rcv(). */
static void sctp_assoc_bh_rcv(struct work_struct *work)
{
struct sctp_association *asoc =
container_of(work, struct sctp_association,
base.inqueue.immediate);
struct net *net = sock_net(asoc->base.sk);
struct sctp_endpoint *ep;
struct sctp_chunk *chunk;
struct sctp_inq *inqueue;
int state;
sctp_subtype_t subtype;
int error = 0;
/* The association should be held so we should be safe. */
ep = asoc->ep;
inqueue = &asoc->base.inqueue;
sctp_association_hold(asoc);
while (NULL != (chunk = sctp_inq_pop(inqueue))) {
state = asoc->state;
subtype = SCTP_ST_CHUNK(chunk->chunk_hdr->type);
/* SCTP-AUTH, Section 6.3:
* The receiver has a list of chunk types which it expects
* to be received only after an AUTH-chunk. This list has
* been sent to the peer during the association setup. It
* MUST silently discard these chunks if they are not placed
* after an AUTH chunk in the packet.
*/
if (sctp_auth_recv_cid(subtype.chunk, asoc) && !chunk->auth)
continue;
/* Remember where the last DATA chunk came from so we
* know where to send the SACK.
*/
if (sctp_chunk_is_data(chunk))
asoc->peer.last_data_from = chunk->transport;
sctp: Add support to per-association statistics via a new SCTP_GET_ASSOC_STATS call The current SCTP stack is lacking a mechanism to have per association statistics. This is an implementation modeled after OpenSolaris' SCTP_GET_ASSOC_STATS. Userspace part will follow on lksctp if/when there is a general ACK on this. V4: - Move ipackets++ before q->immediate.func() for consistency reasons - Move sctp_max_rto() at the end of sctp_transport_update_rto() to avoid returning bogus RTO values - return asoc->rto_min when max_obs_rto value has not changed V3: - Increase ictrlchunks in sctp_assoc_bh_rcv() as well - Move ipackets++ to sctp_inq_push() - return 0 when no rto updates took place since the last call V2: - Implement partial retrieval of stat struct to cope for future expansion - Kill the rtxpackets counter as it cannot be precise anyway - Rename outseqtsns to outofseqtsns to make it clearer that these are out of sequence unexpected TSNs - Move asoc->ipackets++ under a lock to avoid potential miscounts - Fold asoc->opackets++ into the already existing asoc check - Kill unneeded (q->asoc) test when increasing rtxchunks - Do not count octrlchunks if sending failed (SCTP_XMIT_OK != 0) - Don't count SHUTDOWNs as SACKs - Move SCTP_GET_ASSOC_STATS to the private space API - Adjust the len check in sctp_getsockopt_assoc_stats() to allow for future struct growth - Move association statistics in their own struct - Update idupchunks when we send a SACK with dup TSNs - return min_rto in max_rto when RTO has not changed. Also return the transport when max_rto last changed. Signed-off: Michele Baldessari <michele@acksyn.org> Acked-by: Vlad Yasevich <vyasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-12-01 05:49:42 +01:00
else {
SCTP_INC_STATS(net, SCTP_MIB_INCTRLCHUNKS);
sctp: Add support to per-association statistics via a new SCTP_GET_ASSOC_STATS call The current SCTP stack is lacking a mechanism to have per association statistics. This is an implementation modeled after OpenSolaris' SCTP_GET_ASSOC_STATS. Userspace part will follow on lksctp if/when there is a general ACK on this. V4: - Move ipackets++ before q->immediate.func() for consistency reasons - Move sctp_max_rto() at the end of sctp_transport_update_rto() to avoid returning bogus RTO values - return asoc->rto_min when max_obs_rto value has not changed V3: - Increase ictrlchunks in sctp_assoc_bh_rcv() as well - Move ipackets++ to sctp_inq_push() - return 0 when no rto updates took place since the last call V2: - Implement partial retrieval of stat struct to cope for future expansion - Kill the rtxpackets counter as it cannot be precise anyway - Rename outseqtsns to outofseqtsns to make it clearer that these are out of sequence unexpected TSNs - Move asoc->ipackets++ under a lock to avoid potential miscounts - Fold asoc->opackets++ into the already existing asoc check - Kill unneeded (q->asoc) test when increasing rtxchunks - Do not count octrlchunks if sending failed (SCTP_XMIT_OK != 0) - Don't count SHUTDOWNs as SACKs - Move SCTP_GET_ASSOC_STATS to the private space API - Adjust the len check in sctp_getsockopt_assoc_stats() to allow for future struct growth - Move association statistics in their own struct - Update idupchunks when we send a SACK with dup TSNs - return min_rto in max_rto when RTO has not changed. Also return the transport when max_rto last changed. Signed-off: Michele Baldessari <michele@acksyn.org> Acked-by: Vlad Yasevich <vyasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-12-01 05:49:42 +01:00
asoc->stats.ictrlchunks++;
if (chunk->chunk_hdr->type == SCTP_CID_SACK)
asoc->stats.isacks++;
}
if (chunk->transport)
chunk->transport->last_time_heard = ktime_get();
/* Run through the state machine. */
error = sctp_do_sm(net, SCTP_EVENT_T_CHUNK, subtype,
state, ep, asoc, chunk, GFP_ATOMIC);
/* Check to see if the association is freed in response to
* the incoming chunk. If so, get out of the while loop.
*/
if (asoc->base.dead)
break;
/* If there is an error on chunk, discard this packet. */
if (error && chunk)
chunk->pdiscard = 1;
}
sctp_association_put(asoc);
}
/* This routine moves an association from its old sk to a new sk. */
void sctp_assoc_migrate(struct sctp_association *assoc, struct sock *newsk)
{
struct sctp_sock *newsp = sctp_sk(newsk);
struct sock *oldsk = assoc->base.sk;
/* Delete the association from the old endpoint's list of
* associations.
*/
list_del_init(&assoc->asocs);
/* Decrement the backlog value for a TCP-style socket. */
if (sctp_style(oldsk, TCP))
oldsk->sk_ack_backlog--;
/* Release references to the old endpoint and the sock. */
sctp_endpoint_put(assoc->ep);
sock_put(assoc->base.sk);
/* Get a reference to the new endpoint. */
assoc->ep = newsp->ep;
sctp_endpoint_hold(assoc->ep);
/* Get a reference to the new sock. */
assoc->base.sk = newsk;
sock_hold(assoc->base.sk);
/* Add the association to the new endpoint's list of associations. */
sctp_endpoint_add_asoc(newsp->ep, assoc);
}
/* Update an association (possibly from unexpected COOKIE-ECHO processing). */
void sctp_assoc_update(struct sctp_association *asoc,
struct sctp_association *new)
{
struct sctp_transport *trans;
struct list_head *pos, *temp;
/* Copy in new parameters of peer. */
asoc->c = new->c;
asoc->peer.rwnd = new->peer.rwnd;
asoc->peer.sack_needed = new->peer.sack_needed;
net: sctp: inherit auth_capable on INIT collisions Jason reported an oops caused by SCTP on his ARM machine with SCTP authentication enabled: Internal error: Oops: 17 [#1] ARM CPU: 0 PID: 104 Comm: sctp-test Not tainted 3.13.0-68744-g3632f30c9b20-dirty #1 task: c6eefa40 ti: c6f52000 task.ti: c6f52000 PC is at sctp_auth_calculate_hmac+0xc4/0x10c LR is at sg_init_table+0x20/0x38 pc : [<c024bb80>] lr : [<c00f32dc>] psr: 40000013 sp : c6f538e8 ip : 00000000 fp : c6f53924 r10: c6f50d80 r9 : 00000000 r8 : 00010000 r7 : 00000000 r6 : c7be4000 r5 : 00000000 r4 : c6f56254 r3 : c00c8170 r2 : 00000001 r1 : 00000008 r0 : c6f1e660 Flags: nZcv IRQs on FIQs on Mode SVC_32 ISA ARM Segment user Control: 0005397f Table: 06f28000 DAC: 00000015 Process sctp-test (pid: 104, stack limit = 0xc6f521c0) Stack: (0xc6f538e8 to 0xc6f54000) [...] Backtrace: [<c024babc>] (sctp_auth_calculate_hmac+0x0/0x10c) from [<c0249af8>] (sctp_packet_transmit+0x33c/0x5c8) [<c02497bc>] (sctp_packet_transmit+0x0/0x5c8) from [<c023e96c>] (sctp_outq_flush+0x7fc/0x844) [<c023e170>] (sctp_outq_flush+0x0/0x844) from [<c023ef78>] (sctp_outq_uncork+0x24/0x28) [<c023ef54>] (sctp_outq_uncork+0x0/0x28) from [<c0234364>] (sctp_side_effects+0x1134/0x1220) [<c0233230>] (sctp_side_effects+0x0/0x1220) from [<c02330b0>] (sctp_do_sm+0xac/0xd4) [<c0233004>] (sctp_do_sm+0x0/0xd4) from [<c023675c>] (sctp_assoc_bh_rcv+0x118/0x160) [<c0236644>] (sctp_assoc_bh_rcv+0x0/0x160) from [<c023d5bc>] (sctp_inq_push+0x6c/0x74) [<c023d550>] (sctp_inq_push+0x0/0x74) from [<c024a6b0>] (sctp_rcv+0x7d8/0x888) While we already had various kind of bugs in that area ec0223ec48a9 ("net: sctp: fix sctp_sf_do_5_1D_ce to verify if we/peer is AUTH capable") and b14878ccb7fa ("net: sctp: cache auth_enable per endpoint"), this one is a bit of a different kind. Giving a bit more background on why SCTP authentication is needed can be found in RFC4895: SCTP uses 32-bit verification tags to protect itself against blind attackers. These values are not changed during the lifetime of an SCTP association. Looking at new SCTP extensions, there is the need to have a method of proving that an SCTP chunk(s) was really sent by the original peer that started the association and not by a malicious attacker. To cause this bug, we're triggering an INIT collision between peers; normal SCTP handshake where both sides intent to authenticate packets contains RANDOM; CHUNKS; HMAC-ALGO parameters that are being negotiated among peers: ---------- INIT[RANDOM; CHUNKS; HMAC-ALGO] ----------> <------- INIT-ACK[RANDOM; CHUNKS; HMAC-ALGO] --------- -------------------- COOKIE-ECHO --------------------> <-------------------- COOKIE-ACK --------------------- RFC4895 says that each endpoint therefore knows its own random number and the peer's random number *after* the association has been established. The local and peer's random number along with the shared key are then part of the secret used for calculating the HMAC in the AUTH chunk. Now, in our scenario, we have 2 threads with 1 non-blocking SEQ_PACKET socket each, setting up common shared SCTP_AUTH_KEY and SCTP_AUTH_ACTIVE_KEY properly, and each of them calling sctp_bindx(3), listen(2) and connect(2) against each other, thus the handshake looks similar to this, e.g.: ---------- INIT[RANDOM; CHUNKS; HMAC-ALGO] ----------> <------- INIT-ACK[RANDOM; CHUNKS; HMAC-ALGO] --------- <--------- INIT[RANDOM; CHUNKS; HMAC-ALGO] ----------- -------- INIT-ACK[RANDOM; CHUNKS; HMAC-ALGO] --------> ... Since such collisions can also happen with verification tags, the RFC4895 for AUTH rather vaguely says under section 6.1: In case of INIT collision, the rules governing the handling of this Random Number follow the same pattern as those for the Verification Tag, as explained in Section 5.2.4 of RFC 2960 [5]. Therefore, each endpoint knows its own Random Number and the peer's Random Number after the association has been established. In RFC2960, section 5.2.4, we're eventually hitting Action B: B) In this case, both sides may be attempting to start an association at about the same time but the peer endpoint started its INIT after responding to the local endpoint's INIT. Thus it may have picked a new Verification Tag not being aware of the previous Tag it had sent this endpoint. The endpoint should stay in or enter the ESTABLISHED state but it MUST update its peer's Verification Tag from the State Cookie, stop any init or cookie timers that may running and send a COOKIE ACK. In other words, the handling of the Random parameter is the same as behavior for the Verification Tag as described in Action B of section 5.2.4. Looking at the code, we exactly hit the sctp_sf_do_dupcook_b() case which triggers an SCTP_CMD_UPDATE_ASSOC command to the side effect interpreter, and in fact it properly copies over peer_{random, hmacs, chunks} parameters from the newly created association to update the existing one. Also, the old asoc_shared_key is being released and based on the new params, sctp_auth_asoc_init_active_key() updated. However, the issue observed in this case is that the previous asoc->peer.auth_capable was 0, and has *not* been updated, so that instead of creating a new secret, we're doing an early return from the function sctp_auth_asoc_init_active_key() leaving asoc->asoc_shared_key as NULL. However, we now have to authenticate chunks from the updated chunk list (e.g. COOKIE-ACK). That in fact causes the server side when responding with ... <------------------ AUTH; COOKIE-ACK ----------------- ... to trigger a NULL pointer dereference, since in sctp_packet_transmit(), it discovers that an AUTH chunk is being queued for xmit, and thus it calls sctp_auth_calculate_hmac(). Since the asoc->active_key_id is still inherited from the endpoint, and the same as encoded into the chunk, it uses asoc->asoc_shared_key, which is still NULL, as an asoc_key and dereferences it in ... crypto_hash_setkey(desc.tfm, &asoc_key->data[0], asoc_key->len) ... causing an oops. All this happens because sctp_make_cookie_ack() called with the *new* association has the peer.auth_capable=1 and therefore marks the chunk with auth=1 after checking sctp_auth_send_cid(), but it is *actually* sent later on over the then *updated* association's transport that didn't initialize its shared key due to peer.auth_capable=0. Since control chunks in that case are not sent by the temporary association which are scheduled for deletion, they are issued for xmit via SCTP_CMD_REPLY in the interpreter with the context of the *updated* association. peer.auth_capable was 0 in the updated association (which went from COOKIE_WAIT into ESTABLISHED state), since all previous processing that performed sctp_process_init() was being done on temporary associations, that we eventually throw away each time. The correct fix is to update to the new peer.auth_capable value as well in the collision case via sctp_assoc_update(), so that in case the collision migrated from 0 -> 1, sctp_auth_asoc_init_active_key() can properly recalculate the secret. This therefore fixes the observed server panic. Fixes: 730fc3d05cd4 ("[SCTP]: Implete SCTP-AUTH parameter processing") Reported-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Tested-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Cc: Vlad Yasevich <vyasevich@gmail.com> Acked-by: Vlad Yasevich <vyasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-07-22 15:22:45 +02:00
asoc->peer.auth_capable = new->peer.auth_capable;
asoc->peer.i = new->peer.i;
sctp_tsnmap_init(&asoc->peer.tsn_map, SCTP_TSN_MAP_INITIAL,
asoc->peer.i.initial_tsn, GFP_ATOMIC);
/* Remove any peer addresses not present in the new association. */
list_for_each_safe(pos, temp, &asoc->peer.transport_addr_list) {
trans = list_entry(pos, struct sctp_transport, transports);
if (!sctp_assoc_lookup_paddr(new, &trans->ipaddr)) {
sctp_assoc_rm_peer(asoc, trans);
continue;
}
if (asoc->state >= SCTP_STATE_ESTABLISHED)
sctp_transport_reset(trans);
}
/* If the case is A (association restart), use
* initial_tsn as next_tsn. If the case is B, use
* current next_tsn in case data sent to peer
* has been discarded and needs retransmission.
*/
if (asoc->state >= SCTP_STATE_ESTABLISHED) {
asoc->next_tsn = new->next_tsn;
asoc->ctsn_ack_point = new->ctsn_ack_point;
asoc->adv_peer_ack_point = new->adv_peer_ack_point;
/* Reinitialize SSN for both local streams
* and peer's streams.
*/
sctp_ssnmap_clear(asoc->ssnmap);
/* Flush the ULP reassembly and ordered queue.
* Any data there will now be stale and will
* cause problems.
*/
sctp_ulpq_flush(&asoc->ulpq);
/* reset the overall association error count so
* that the restarted association doesn't get torn
* down on the next retransmission timer.
*/
asoc->overall_error_count = 0;
} else {
/* Add any peer addresses from the new association. */
list_for_each_entry(trans, &new->peer.transport_addr_list,
transports) {
if (!sctp_assoc_lookup_paddr(asoc, &trans->ipaddr))
sctp_assoc_add_peer(asoc, &trans->ipaddr,
GFP_ATOMIC, trans->state);
}
asoc->ctsn_ack_point = asoc->next_tsn - 1;
asoc->adv_peer_ack_point = asoc->ctsn_ack_point;
if (!asoc->ssnmap) {
/* Move the ssnmap. */
asoc->ssnmap = new->ssnmap;
new->ssnmap = NULL;
}
if (!asoc->assoc_id) {
/* get a new association id since we don't have one
* yet.
*/
sctp_assoc_set_id(asoc, GFP_ATOMIC);
}
}
/* SCTP-AUTH: Save the peer parameters from the new associations
* and also move the association shared keys over
*/
kfree(asoc->peer.peer_random);
asoc->peer.peer_random = new->peer.peer_random;
new->peer.peer_random = NULL;
kfree(asoc->peer.peer_chunks);
asoc->peer.peer_chunks = new->peer.peer_chunks;
new->peer.peer_chunks = NULL;
kfree(asoc->peer.peer_hmacs);
asoc->peer.peer_hmacs = new->peer.peer_hmacs;
new->peer.peer_hmacs = NULL;
sctp_auth_asoc_init_active_key(asoc, GFP_ATOMIC);
}
/* Update the retran path for sending a retransmitted packet.
net: sctp: rework multihoming retransmission path selection to rfc4960 Problem statement: 1) both paths (primary path1 and alternate path2) are up after the association has been established i.e., HB packets are normally exchanged, 2) path2 gets inactive after path_max_retrans * max_rto timed out (i.e. path2 is down completely), 3) now, if a transmission times out on the only surviving/active path1 (any ~1sec network service impact could cause this like a channel bonding failover), then the retransmitted packets are sent over the inactive path2; this happens with partial failover and without it. Besides not being optimal in the above scenario, a small failure or timeout in the only existing path has the potential to cause long delays in the retransmission (depending on RTO_MAX) until the still active path is reselected. Further, when the T3-timeout occurs, we have active_patch == retrans_path, and even though the timeout occurred on the initial transmission of data, not a retransmit, we end up updating retransmit path. RFC4960, section 6.4. "Multi-Homed SCTP Endpoints" states under 6.4.1. "Failover from an Inactive Destination Address" the following: Some of the transport addresses of a multi-homed SCTP endpoint may become inactive due to either the occurrence of certain error conditions (see Section 8.2) or adjustments from the SCTP user. When there is outbound data to send and the primary path becomes inactive (e.g., due to failures), or where the SCTP user explicitly requests to send data to an inactive destination transport address, before reporting an error to its ULP, the SCTP endpoint should try to send the data to an alternate __active__ destination transport address if one exists. When retransmitting data that timed out, if the endpoint is multihomed, it should consider each source-destination address pair in its retransmission selection policy. When retransmitting timed-out data, the endpoint should attempt to pick the most divergent source-destination pair from the original source-destination pair to which the packet was transmitted. Note: Rules for picking the most divergent source-destination pair are an implementation decision and are not specified within this document. So, we should first reconsider to take the current active retransmission transport if we cannot find an alternative active one. If all of that fails, we can still round robin through unkown, partial failover, and inactive ones in the hope to find something still suitable. Commit 4141ddc02a92 ("sctp: retran_path update bug fix") broke that behaviour by selecting the next inactive transport when no other active transport was found besides the current assoc's peer.retran_path. Before commit 4141ddc02a92, we would have traversed through the list until we reach our peer.retran_path again, and in case that is still in state SCTP_ACTIVE, we would take it and return. Only if that is not the case either, we take the next inactive transport. Besides all that, another issue is that transports in state SCTP_UNKNOWN could be preferred over transports in state SCTP_ACTIVE in case a SCTP_ACTIVE transport appears after SCTP_UNKNOWN in the transport list yielding a weaker transport state to be used in retransmission. This patch mostly reverts 4141ddc02a92, but also rewrites this function to introduce more clarity and strictness into the code. A strict priority of transport states is enforced in this patch, hence selection is active > unkown > partial failover > inactive. Fixes: 4141ddc02a92 ("sctp: retran_path update bug fix") Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Gui Jianfeng <guijianfeng@cn.fujitsu.com> Acked-by: Vlad Yasevich <yasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-20 20:51:06 +01:00
* See also RFC4960, 6.4. Multi-Homed SCTP Endpoints:
*
* When there is outbound data to send and the primary path
* becomes inactive (e.g., due to failures), or where the
* SCTP user explicitly requests to send data to an
* inactive destination transport address, before reporting
* an error to its ULP, the SCTP endpoint should try to send
* the data to an alternate active destination transport
* address if one exists.
*
* When retransmitting data that timed out, if the endpoint
* is multihomed, it should consider each source-destination
* address pair in its retransmission selection policy.
* When retransmitting timed-out data, the endpoint should
* attempt to pick the most divergent source-destination
* pair from the original source-destination pair to which
* the packet was transmitted.
*
* Note: Rules for picking the most divergent source-destination
* pair are an implementation decision and are not specified
* within this document.
*
* Our basic strategy is to round-robin transports in priorities
* according to sctp_trans_score() e.g., if no such
net: sctp: rework multihoming retransmission path selection to rfc4960 Problem statement: 1) both paths (primary path1 and alternate path2) are up after the association has been established i.e., HB packets are normally exchanged, 2) path2 gets inactive after path_max_retrans * max_rto timed out (i.e. path2 is down completely), 3) now, if a transmission times out on the only surviving/active path1 (any ~1sec network service impact could cause this like a channel bonding failover), then the retransmitted packets are sent over the inactive path2; this happens with partial failover and without it. Besides not being optimal in the above scenario, a small failure or timeout in the only existing path has the potential to cause long delays in the retransmission (depending on RTO_MAX) until the still active path is reselected. Further, when the T3-timeout occurs, we have active_patch == retrans_path, and even though the timeout occurred on the initial transmission of data, not a retransmit, we end up updating retransmit path. RFC4960, section 6.4. "Multi-Homed SCTP Endpoints" states under 6.4.1. "Failover from an Inactive Destination Address" the following: Some of the transport addresses of a multi-homed SCTP endpoint may become inactive due to either the occurrence of certain error conditions (see Section 8.2) or adjustments from the SCTP user. When there is outbound data to send and the primary path becomes inactive (e.g., due to failures), or where the SCTP user explicitly requests to send data to an inactive destination transport address, before reporting an error to its ULP, the SCTP endpoint should try to send the data to an alternate __active__ destination transport address if one exists. When retransmitting data that timed out, if the endpoint is multihomed, it should consider each source-destination address pair in its retransmission selection policy. When retransmitting timed-out data, the endpoint should attempt to pick the most divergent source-destination pair from the original source-destination pair to which the packet was transmitted. Note: Rules for picking the most divergent source-destination pair are an implementation decision and are not specified within this document. So, we should first reconsider to take the current active retransmission transport if we cannot find an alternative active one. If all of that fails, we can still round robin through unkown, partial failover, and inactive ones in the hope to find something still suitable. Commit 4141ddc02a92 ("sctp: retran_path update bug fix") broke that behaviour by selecting the next inactive transport when no other active transport was found besides the current assoc's peer.retran_path. Before commit 4141ddc02a92, we would have traversed through the list until we reach our peer.retran_path again, and in case that is still in state SCTP_ACTIVE, we would take it and return. Only if that is not the case either, we take the next inactive transport. Besides all that, another issue is that transports in state SCTP_UNKNOWN could be preferred over transports in state SCTP_ACTIVE in case a SCTP_ACTIVE transport appears after SCTP_UNKNOWN in the transport list yielding a weaker transport state to be used in retransmission. This patch mostly reverts 4141ddc02a92, but also rewrites this function to introduce more clarity and strictness into the code. A strict priority of transport states is enforced in this patch, hence selection is active > unkown > partial failover > inactive. Fixes: 4141ddc02a92 ("sctp: retran_path update bug fix") Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Gui Jianfeng <guijianfeng@cn.fujitsu.com> Acked-by: Vlad Yasevich <yasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-20 20:51:06 +01:00
* transport with state SCTP_ACTIVE exists, round-robin through
* SCTP_UNKNOWN, etc. You get the picture.
*/
net: sctp: rework multihoming retransmission path selection to rfc4960 Problem statement: 1) both paths (primary path1 and alternate path2) are up after the association has been established i.e., HB packets are normally exchanged, 2) path2 gets inactive after path_max_retrans * max_rto timed out (i.e. path2 is down completely), 3) now, if a transmission times out on the only surviving/active path1 (any ~1sec network service impact could cause this like a channel bonding failover), then the retransmitted packets are sent over the inactive path2; this happens with partial failover and without it. Besides not being optimal in the above scenario, a small failure or timeout in the only existing path has the potential to cause long delays in the retransmission (depending on RTO_MAX) until the still active path is reselected. Further, when the T3-timeout occurs, we have active_patch == retrans_path, and even though the timeout occurred on the initial transmission of data, not a retransmit, we end up updating retransmit path. RFC4960, section 6.4. "Multi-Homed SCTP Endpoints" states under 6.4.1. "Failover from an Inactive Destination Address" the following: Some of the transport addresses of a multi-homed SCTP endpoint may become inactive due to either the occurrence of certain error conditions (see Section 8.2) or adjustments from the SCTP user. When there is outbound data to send and the primary path becomes inactive (e.g., due to failures), or where the SCTP user explicitly requests to send data to an inactive destination transport address, before reporting an error to its ULP, the SCTP endpoint should try to send the data to an alternate __active__ destination transport address if one exists. When retransmitting data that timed out, if the endpoint is multihomed, it should consider each source-destination address pair in its retransmission selection policy. When retransmitting timed-out data, the endpoint should attempt to pick the most divergent source-destination pair from the original source-destination pair to which the packet was transmitted. Note: Rules for picking the most divergent source-destination pair are an implementation decision and are not specified within this document. So, we should first reconsider to take the current active retransmission transport if we cannot find an alternative active one. If all of that fails, we can still round robin through unkown, partial failover, and inactive ones in the hope to find something still suitable. Commit 4141ddc02a92 ("sctp: retran_path update bug fix") broke that behaviour by selecting the next inactive transport when no other active transport was found besides the current assoc's peer.retran_path. Before commit 4141ddc02a92, we would have traversed through the list until we reach our peer.retran_path again, and in case that is still in state SCTP_ACTIVE, we would take it and return. Only if that is not the case either, we take the next inactive transport. Besides all that, another issue is that transports in state SCTP_UNKNOWN could be preferred over transports in state SCTP_ACTIVE in case a SCTP_ACTIVE transport appears after SCTP_UNKNOWN in the transport list yielding a weaker transport state to be used in retransmission. This patch mostly reverts 4141ddc02a92, but also rewrites this function to introduce more clarity and strictness into the code. A strict priority of transport states is enforced in this patch, hence selection is active > unkown > partial failover > inactive. Fixes: 4141ddc02a92 ("sctp: retran_path update bug fix") Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Gui Jianfeng <guijianfeng@cn.fujitsu.com> Acked-by: Vlad Yasevich <yasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-20 20:51:06 +01:00
static u8 sctp_trans_score(const struct sctp_transport *trans)
{
switch (trans->state) {
case SCTP_ACTIVE:
return 3; /* best case */
case SCTP_UNKNOWN:
return 2;
case SCTP_PF:
return 1;
default: /* case SCTP_INACTIVE */
return 0; /* worst case */
}
net: sctp: rework multihoming retransmission path selection to rfc4960 Problem statement: 1) both paths (primary path1 and alternate path2) are up after the association has been established i.e., HB packets are normally exchanged, 2) path2 gets inactive after path_max_retrans * max_rto timed out (i.e. path2 is down completely), 3) now, if a transmission times out on the only surviving/active path1 (any ~1sec network service impact could cause this like a channel bonding failover), then the retransmitted packets are sent over the inactive path2; this happens with partial failover and without it. Besides not being optimal in the above scenario, a small failure or timeout in the only existing path has the potential to cause long delays in the retransmission (depending on RTO_MAX) until the still active path is reselected. Further, when the T3-timeout occurs, we have active_patch == retrans_path, and even though the timeout occurred on the initial transmission of data, not a retransmit, we end up updating retransmit path. RFC4960, section 6.4. "Multi-Homed SCTP Endpoints" states under 6.4.1. "Failover from an Inactive Destination Address" the following: Some of the transport addresses of a multi-homed SCTP endpoint may become inactive due to either the occurrence of certain error conditions (see Section 8.2) or adjustments from the SCTP user. When there is outbound data to send and the primary path becomes inactive (e.g., due to failures), or where the SCTP user explicitly requests to send data to an inactive destination transport address, before reporting an error to its ULP, the SCTP endpoint should try to send the data to an alternate __active__ destination transport address if one exists. When retransmitting data that timed out, if the endpoint is multihomed, it should consider each source-destination address pair in its retransmission selection policy. When retransmitting timed-out data, the endpoint should attempt to pick the most divergent source-destination pair from the original source-destination pair to which the packet was transmitted. Note: Rules for picking the most divergent source-destination pair are an implementation decision and are not specified within this document. So, we should first reconsider to take the current active retransmission transport if we cannot find an alternative active one. If all of that fails, we can still round robin through unkown, partial failover, and inactive ones in the hope to find something still suitable. Commit 4141ddc02a92 ("sctp: retran_path update bug fix") broke that behaviour by selecting the next inactive transport when no other active transport was found besides the current assoc's peer.retran_path. Before commit 4141ddc02a92, we would have traversed through the list until we reach our peer.retran_path again, and in case that is still in state SCTP_ACTIVE, we would take it and return. Only if that is not the case either, we take the next inactive transport. Besides all that, another issue is that transports in state SCTP_UNKNOWN could be preferred over transports in state SCTP_ACTIVE in case a SCTP_ACTIVE transport appears after SCTP_UNKNOWN in the transport list yielding a weaker transport state to be used in retransmission. This patch mostly reverts 4141ddc02a92, but also rewrites this function to introduce more clarity and strictness into the code. A strict priority of transport states is enforced in this patch, hence selection is active > unkown > partial failover > inactive. Fixes: 4141ddc02a92 ("sctp: retran_path update bug fix") Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Gui Jianfeng <guijianfeng@cn.fujitsu.com> Acked-by: Vlad Yasevich <yasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-20 20:51:06 +01:00
}
net: sctp: improve sctp_select_active_and_retran_path selection In function sctp_select_active_and_retran_path(), we walk the transport list in order to look for the two most recently used ACTIVE transports (trans_pri, trans_sec). In case we didn't find anything ACTIVE, we currently just camp on a possibly PF or INACTIVE transport that is primary path; this behavior actually dates back to linux-history tree of the very early days of lksctp, and can yield a behavior that chooses suboptimal transport paths. Instead, be a bit more clever by reusing and extending the recently introduced sctp_trans_elect_best() handler. In case both transports are evaluated to have the same score resulting from their states, break the tie by looking at: 1) transport patch error count 2) last_time_heard value from each transport. This is analogous to Nishida's Quick Failover draft [1], section 5.1, 3: The sender SHOULD avoid data transmission to PF destinations. When all destinations are in either PF or Inactive state, the sender MAY either move the destination from PF to active state (and transmit data to the active destination) or the sender MAY transmit data to a PF destination. In the former scenario, (i) the sender MUST NOT notify the ULP about the state transition, and (ii) MUST NOT clear the destination's error counter. It is recommended that the sender picks the PF destination with least error count (fewest consecutive timeouts) for data transmission. In case of a tie (multiple PF destinations with same error count), the sender MAY choose the last active destination. Thus for sctp_select_active_and_retran_path(), we keep track of the best, if any, transport that is in PF state and in case no ACTIVE transport has been found (hence trans_{pri,sec} is NULL), we select the best out of the three: current primary_path and retran_path as well as a possible PF transport. The secondary may still camp on the original primary_path as before. The change in sctp_trans_elect_best() with a more fine grained tie selection also improves at the same time path selection for sctp_assoc_update_retran_path() in case of non-ACTIVE states. [1] http://tools.ietf.org/html/draft-nishida-tsvwg-sctp-failover-05 Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-11 18:19:31 +02:00
static struct sctp_transport *sctp_trans_elect_tie(struct sctp_transport *trans1,
struct sctp_transport *trans2)
{
if (trans1->error_count > trans2->error_count) {
return trans2;
} else if (trans1->error_count == trans2->error_count &&
ktime_after(trans2->last_time_heard,
trans1->last_time_heard)) {
return trans2;
} else {
return trans1;
}
}
net: sctp: rework multihoming retransmission path selection to rfc4960 Problem statement: 1) both paths (primary path1 and alternate path2) are up after the association has been established i.e., HB packets are normally exchanged, 2) path2 gets inactive after path_max_retrans * max_rto timed out (i.e. path2 is down completely), 3) now, if a transmission times out on the only surviving/active path1 (any ~1sec network service impact could cause this like a channel bonding failover), then the retransmitted packets are sent over the inactive path2; this happens with partial failover and without it. Besides not being optimal in the above scenario, a small failure or timeout in the only existing path has the potential to cause long delays in the retransmission (depending on RTO_MAX) until the still active path is reselected. Further, when the T3-timeout occurs, we have active_patch == retrans_path, and even though the timeout occurred on the initial transmission of data, not a retransmit, we end up updating retransmit path. RFC4960, section 6.4. "Multi-Homed SCTP Endpoints" states under 6.4.1. "Failover from an Inactive Destination Address" the following: Some of the transport addresses of a multi-homed SCTP endpoint may become inactive due to either the occurrence of certain error conditions (see Section 8.2) or adjustments from the SCTP user. When there is outbound data to send and the primary path becomes inactive (e.g., due to failures), or where the SCTP user explicitly requests to send data to an inactive destination transport address, before reporting an error to its ULP, the SCTP endpoint should try to send the data to an alternate __active__ destination transport address if one exists. When retransmitting data that timed out, if the endpoint is multihomed, it should consider each source-destination address pair in its retransmission selection policy. When retransmitting timed-out data, the endpoint should attempt to pick the most divergent source-destination pair from the original source-destination pair to which the packet was transmitted. Note: Rules for picking the most divergent source-destination pair are an implementation decision and are not specified within this document. So, we should first reconsider to take the current active retransmission transport if we cannot find an alternative active one. If all of that fails, we can still round robin through unkown, partial failover, and inactive ones in the hope to find something still suitable. Commit 4141ddc02a92 ("sctp: retran_path update bug fix") broke that behaviour by selecting the next inactive transport when no other active transport was found besides the current assoc's peer.retran_path. Before commit 4141ddc02a92, we would have traversed through the list until we reach our peer.retran_path again, and in case that is still in state SCTP_ACTIVE, we would take it and return. Only if that is not the case either, we take the next inactive transport. Besides all that, another issue is that transports in state SCTP_UNKNOWN could be preferred over transports in state SCTP_ACTIVE in case a SCTP_ACTIVE transport appears after SCTP_UNKNOWN in the transport list yielding a weaker transport state to be used in retransmission. This patch mostly reverts 4141ddc02a92, but also rewrites this function to introduce more clarity and strictness into the code. A strict priority of transport states is enforced in this patch, hence selection is active > unkown > partial failover > inactive. Fixes: 4141ddc02a92 ("sctp: retran_path update bug fix") Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Gui Jianfeng <guijianfeng@cn.fujitsu.com> Acked-by: Vlad Yasevich <yasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-20 20:51:06 +01:00
static struct sctp_transport *sctp_trans_elect_best(struct sctp_transport *curr,
struct sctp_transport *best)
{
net: sctp: improve sctp_select_active_and_retran_path selection In function sctp_select_active_and_retran_path(), we walk the transport list in order to look for the two most recently used ACTIVE transports (trans_pri, trans_sec). In case we didn't find anything ACTIVE, we currently just camp on a possibly PF or INACTIVE transport that is primary path; this behavior actually dates back to linux-history tree of the very early days of lksctp, and can yield a behavior that chooses suboptimal transport paths. Instead, be a bit more clever by reusing and extending the recently introduced sctp_trans_elect_best() handler. In case both transports are evaluated to have the same score resulting from their states, break the tie by looking at: 1) transport patch error count 2) last_time_heard value from each transport. This is analogous to Nishida's Quick Failover draft [1], section 5.1, 3: The sender SHOULD avoid data transmission to PF destinations. When all destinations are in either PF or Inactive state, the sender MAY either move the destination from PF to active state (and transmit data to the active destination) or the sender MAY transmit data to a PF destination. In the former scenario, (i) the sender MUST NOT notify the ULP about the state transition, and (ii) MUST NOT clear the destination's error counter. It is recommended that the sender picks the PF destination with least error count (fewest consecutive timeouts) for data transmission. In case of a tie (multiple PF destinations with same error count), the sender MAY choose the last active destination. Thus for sctp_select_active_and_retran_path(), we keep track of the best, if any, transport that is in PF state and in case no ACTIVE transport has been found (hence trans_{pri,sec} is NULL), we select the best out of the three: current primary_path and retran_path as well as a possible PF transport. The secondary may still camp on the original primary_path as before. The change in sctp_trans_elect_best() with a more fine grained tie selection also improves at the same time path selection for sctp_assoc_update_retran_path() in case of non-ACTIVE states. [1] http://tools.ietf.org/html/draft-nishida-tsvwg-sctp-failover-05 Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-11 18:19:31 +02:00
u8 score_curr, score_best;
if (best == NULL || curr == best)
net: sctp: rework multihoming retransmission path selection to rfc4960 Problem statement: 1) both paths (primary path1 and alternate path2) are up after the association has been established i.e., HB packets are normally exchanged, 2) path2 gets inactive after path_max_retrans * max_rto timed out (i.e. path2 is down completely), 3) now, if a transmission times out on the only surviving/active path1 (any ~1sec network service impact could cause this like a channel bonding failover), then the retransmitted packets are sent over the inactive path2; this happens with partial failover and without it. Besides not being optimal in the above scenario, a small failure or timeout in the only existing path has the potential to cause long delays in the retransmission (depending on RTO_MAX) until the still active path is reselected. Further, when the T3-timeout occurs, we have active_patch == retrans_path, and even though the timeout occurred on the initial transmission of data, not a retransmit, we end up updating retransmit path. RFC4960, section 6.4. "Multi-Homed SCTP Endpoints" states under 6.4.1. "Failover from an Inactive Destination Address" the following: Some of the transport addresses of a multi-homed SCTP endpoint may become inactive due to either the occurrence of certain error conditions (see Section 8.2) or adjustments from the SCTP user. When there is outbound data to send and the primary path becomes inactive (e.g., due to failures), or where the SCTP user explicitly requests to send data to an inactive destination transport address, before reporting an error to its ULP, the SCTP endpoint should try to send the data to an alternate __active__ destination transport address if one exists. When retransmitting data that timed out, if the endpoint is multihomed, it should consider each source-destination address pair in its retransmission selection policy. When retransmitting timed-out data, the endpoint should attempt to pick the most divergent source-destination pair from the original source-destination pair to which the packet was transmitted. Note: Rules for picking the most divergent source-destination pair are an implementation decision and are not specified within this document. So, we should first reconsider to take the current active retransmission transport if we cannot find an alternative active one. If all of that fails, we can still round robin through unkown, partial failover, and inactive ones in the hope to find something still suitable. Commit 4141ddc02a92 ("sctp: retran_path update bug fix") broke that behaviour by selecting the next inactive transport when no other active transport was found besides the current assoc's peer.retran_path. Before commit 4141ddc02a92, we would have traversed through the list until we reach our peer.retran_path again, and in case that is still in state SCTP_ACTIVE, we would take it and return. Only if that is not the case either, we take the next inactive transport. Besides all that, another issue is that transports in state SCTP_UNKNOWN could be preferred over transports in state SCTP_ACTIVE in case a SCTP_ACTIVE transport appears after SCTP_UNKNOWN in the transport list yielding a weaker transport state to be used in retransmission. This patch mostly reverts 4141ddc02a92, but also rewrites this function to introduce more clarity and strictness into the code. A strict priority of transport states is enforced in this patch, hence selection is active > unkown > partial failover > inactive. Fixes: 4141ddc02a92 ("sctp: retran_path update bug fix") Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Gui Jianfeng <guijianfeng@cn.fujitsu.com> Acked-by: Vlad Yasevich <yasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-20 20:51:06 +01:00
return curr;
net: sctp: improve sctp_select_active_and_retran_path selection In function sctp_select_active_and_retran_path(), we walk the transport list in order to look for the two most recently used ACTIVE transports (trans_pri, trans_sec). In case we didn't find anything ACTIVE, we currently just camp on a possibly PF or INACTIVE transport that is primary path; this behavior actually dates back to linux-history tree of the very early days of lksctp, and can yield a behavior that chooses suboptimal transport paths. Instead, be a bit more clever by reusing and extending the recently introduced sctp_trans_elect_best() handler. In case both transports are evaluated to have the same score resulting from their states, break the tie by looking at: 1) transport patch error count 2) last_time_heard value from each transport. This is analogous to Nishida's Quick Failover draft [1], section 5.1, 3: The sender SHOULD avoid data transmission to PF destinations. When all destinations are in either PF or Inactive state, the sender MAY either move the destination from PF to active state (and transmit data to the active destination) or the sender MAY transmit data to a PF destination. In the former scenario, (i) the sender MUST NOT notify the ULP about the state transition, and (ii) MUST NOT clear the destination's error counter. It is recommended that the sender picks the PF destination with least error count (fewest consecutive timeouts) for data transmission. In case of a tie (multiple PF destinations with same error count), the sender MAY choose the last active destination. Thus for sctp_select_active_and_retran_path(), we keep track of the best, if any, transport that is in PF state and in case no ACTIVE transport has been found (hence trans_{pri,sec} is NULL), we select the best out of the three: current primary_path and retran_path as well as a possible PF transport. The secondary may still camp on the original primary_path as before. The change in sctp_trans_elect_best() with a more fine grained tie selection also improves at the same time path selection for sctp_assoc_update_retran_path() in case of non-ACTIVE states. [1] http://tools.ietf.org/html/draft-nishida-tsvwg-sctp-failover-05 Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-11 18:19:31 +02:00
score_curr = sctp_trans_score(curr);
score_best = sctp_trans_score(best);
/* First, try a score-based selection if both transport states
* differ. If we're in a tie, lets try to make a more clever
* decision here based on error counts and last time heard.
*/
if (score_curr > score_best)
return curr;
else if (score_curr == score_best)
return sctp_trans_elect_tie(curr, best);
else
return best;
net: sctp: rework multihoming retransmission path selection to rfc4960 Problem statement: 1) both paths (primary path1 and alternate path2) are up after the association has been established i.e., HB packets are normally exchanged, 2) path2 gets inactive after path_max_retrans * max_rto timed out (i.e. path2 is down completely), 3) now, if a transmission times out on the only surviving/active path1 (any ~1sec network service impact could cause this like a channel bonding failover), then the retransmitted packets are sent over the inactive path2; this happens with partial failover and without it. Besides not being optimal in the above scenario, a small failure or timeout in the only existing path has the potential to cause long delays in the retransmission (depending on RTO_MAX) until the still active path is reselected. Further, when the T3-timeout occurs, we have active_patch == retrans_path, and even though the timeout occurred on the initial transmission of data, not a retransmit, we end up updating retransmit path. RFC4960, section 6.4. "Multi-Homed SCTP Endpoints" states under 6.4.1. "Failover from an Inactive Destination Address" the following: Some of the transport addresses of a multi-homed SCTP endpoint may become inactive due to either the occurrence of certain error conditions (see Section 8.2) or adjustments from the SCTP user. When there is outbound data to send and the primary path becomes inactive (e.g., due to failures), or where the SCTP user explicitly requests to send data to an inactive destination transport address, before reporting an error to its ULP, the SCTP endpoint should try to send the data to an alternate __active__ destination transport address if one exists. When retransmitting data that timed out, if the endpoint is multihomed, it should consider each source-destination address pair in its retransmission selection policy. When retransmitting timed-out data, the endpoint should attempt to pick the most divergent source-destination pair from the original source-destination pair to which the packet was transmitted. Note: Rules for picking the most divergent source-destination pair are an implementation decision and are not specified within this document. So, we should first reconsider to take the current active retransmission transport if we cannot find an alternative active one. If all of that fails, we can still round robin through unkown, partial failover, and inactive ones in the hope to find something still suitable. Commit 4141ddc02a92 ("sctp: retran_path update bug fix") broke that behaviour by selecting the next inactive transport when no other active transport was found besides the current assoc's peer.retran_path. Before commit 4141ddc02a92, we would have traversed through the list until we reach our peer.retran_path again, and in case that is still in state SCTP_ACTIVE, we would take it and return. Only if that is not the case either, we take the next inactive transport. Besides all that, another issue is that transports in state SCTP_UNKNOWN could be preferred over transports in state SCTP_ACTIVE in case a SCTP_ACTIVE transport appears after SCTP_UNKNOWN in the transport list yielding a weaker transport state to be used in retransmission. This patch mostly reverts 4141ddc02a92, but also rewrites this function to introduce more clarity and strictness into the code. A strict priority of transport states is enforced in this patch, hence selection is active > unkown > partial failover > inactive. Fixes: 4141ddc02a92 ("sctp: retran_path update bug fix") Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Gui Jianfeng <guijianfeng@cn.fujitsu.com> Acked-by: Vlad Yasevich <yasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-20 20:51:06 +01:00
}
net: sctp: rework multihoming retransmission path selection to rfc4960 Problem statement: 1) both paths (primary path1 and alternate path2) are up after the association has been established i.e., HB packets are normally exchanged, 2) path2 gets inactive after path_max_retrans * max_rto timed out (i.e. path2 is down completely), 3) now, if a transmission times out on the only surviving/active path1 (any ~1sec network service impact could cause this like a channel bonding failover), then the retransmitted packets are sent over the inactive path2; this happens with partial failover and without it. Besides not being optimal in the above scenario, a small failure or timeout in the only existing path has the potential to cause long delays in the retransmission (depending on RTO_MAX) until the still active path is reselected. Further, when the T3-timeout occurs, we have active_patch == retrans_path, and even though the timeout occurred on the initial transmission of data, not a retransmit, we end up updating retransmit path. RFC4960, section 6.4. "Multi-Homed SCTP Endpoints" states under 6.4.1. "Failover from an Inactive Destination Address" the following: Some of the transport addresses of a multi-homed SCTP endpoint may become inactive due to either the occurrence of certain error conditions (see Section 8.2) or adjustments from the SCTP user. When there is outbound data to send and the primary path becomes inactive (e.g., due to failures), or where the SCTP user explicitly requests to send data to an inactive destination transport address, before reporting an error to its ULP, the SCTP endpoint should try to send the data to an alternate __active__ destination transport address if one exists. When retransmitting data that timed out, if the endpoint is multihomed, it should consider each source-destination address pair in its retransmission selection policy. When retransmitting timed-out data, the endpoint should attempt to pick the most divergent source-destination pair from the original source-destination pair to which the packet was transmitted. Note: Rules for picking the most divergent source-destination pair are an implementation decision and are not specified within this document. So, we should first reconsider to take the current active retransmission transport if we cannot find an alternative active one. If all of that fails, we can still round robin through unkown, partial failover, and inactive ones in the hope to find something still suitable. Commit 4141ddc02a92 ("sctp: retran_path update bug fix") broke that behaviour by selecting the next inactive transport when no other active transport was found besides the current assoc's peer.retran_path. Before commit 4141ddc02a92, we would have traversed through the list until we reach our peer.retran_path again, and in case that is still in state SCTP_ACTIVE, we would take it and return. Only if that is not the case either, we take the next inactive transport. Besides all that, another issue is that transports in state SCTP_UNKNOWN could be preferred over transports in state SCTP_ACTIVE in case a SCTP_ACTIVE transport appears after SCTP_UNKNOWN in the transport list yielding a weaker transport state to be used in retransmission. This patch mostly reverts 4141ddc02a92, but also rewrites this function to introduce more clarity and strictness into the code. A strict priority of transport states is enforced in this patch, hence selection is active > unkown > partial failover > inactive. Fixes: 4141ddc02a92 ("sctp: retran_path update bug fix") Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Gui Jianfeng <guijianfeng@cn.fujitsu.com> Acked-by: Vlad Yasevich <yasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-20 20:51:06 +01:00
void sctp_assoc_update_retran_path(struct sctp_association *asoc)
{
struct sctp_transport *trans = asoc->peer.retran_path;
struct sctp_transport *trans_next = NULL;
net: sctp: rework multihoming retransmission path selection to rfc4960 Problem statement: 1) both paths (primary path1 and alternate path2) are up after the association has been established i.e., HB packets are normally exchanged, 2) path2 gets inactive after path_max_retrans * max_rto timed out (i.e. path2 is down completely), 3) now, if a transmission times out on the only surviving/active path1 (any ~1sec network service impact could cause this like a channel bonding failover), then the retransmitted packets are sent over the inactive path2; this happens with partial failover and without it. Besides not being optimal in the above scenario, a small failure or timeout in the only existing path has the potential to cause long delays in the retransmission (depending on RTO_MAX) until the still active path is reselected. Further, when the T3-timeout occurs, we have active_patch == retrans_path, and even though the timeout occurred on the initial transmission of data, not a retransmit, we end up updating retransmit path. RFC4960, section 6.4. "Multi-Homed SCTP Endpoints" states under 6.4.1. "Failover from an Inactive Destination Address" the following: Some of the transport addresses of a multi-homed SCTP endpoint may become inactive due to either the occurrence of certain error conditions (see Section 8.2) or adjustments from the SCTP user. When there is outbound data to send and the primary path becomes inactive (e.g., due to failures), or where the SCTP user explicitly requests to send data to an inactive destination transport address, before reporting an error to its ULP, the SCTP endpoint should try to send the data to an alternate __active__ destination transport address if one exists. When retransmitting data that timed out, if the endpoint is multihomed, it should consider each source-destination address pair in its retransmission selection policy. When retransmitting timed-out data, the endpoint should attempt to pick the most divergent source-destination pair from the original source-destination pair to which the packet was transmitted. Note: Rules for picking the most divergent source-destination pair are an implementation decision and are not specified within this document. So, we should first reconsider to take the current active retransmission transport if we cannot find an alternative active one. If all of that fails, we can still round robin through unkown, partial failover, and inactive ones in the hope to find something still suitable. Commit 4141ddc02a92 ("sctp: retran_path update bug fix") broke that behaviour by selecting the next inactive transport when no other active transport was found besides the current assoc's peer.retran_path. Before commit 4141ddc02a92, we would have traversed through the list until we reach our peer.retran_path again, and in case that is still in state SCTP_ACTIVE, we would take it and return. Only if that is not the case either, we take the next inactive transport. Besides all that, another issue is that transports in state SCTP_UNKNOWN could be preferred over transports in state SCTP_ACTIVE in case a SCTP_ACTIVE transport appears after SCTP_UNKNOWN in the transport list yielding a weaker transport state to be used in retransmission. This patch mostly reverts 4141ddc02a92, but also rewrites this function to introduce more clarity and strictness into the code. A strict priority of transport states is enforced in this patch, hence selection is active > unkown > partial failover > inactive. Fixes: 4141ddc02a92 ("sctp: retran_path update bug fix") Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Gui Jianfeng <guijianfeng@cn.fujitsu.com> Acked-by: Vlad Yasevich <yasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-20 20:51:06 +01:00
/* We're done as we only have the one and only path. */
if (asoc->peer.transport_count == 1)
return;
/* If active_path and retran_path are the same and active,
* then this is the only active path. Use it.
*/
if (asoc->peer.active_path == asoc->peer.retran_path &&
asoc->peer.active_path->state == SCTP_ACTIVE)
return;
net: sctp: rework multihoming retransmission path selection to rfc4960 Problem statement: 1) both paths (primary path1 and alternate path2) are up after the association has been established i.e., HB packets are normally exchanged, 2) path2 gets inactive after path_max_retrans * max_rto timed out (i.e. path2 is down completely), 3) now, if a transmission times out on the only surviving/active path1 (any ~1sec network service impact could cause this like a channel bonding failover), then the retransmitted packets are sent over the inactive path2; this happens with partial failover and without it. Besides not being optimal in the above scenario, a small failure or timeout in the only existing path has the potential to cause long delays in the retransmission (depending on RTO_MAX) until the still active path is reselected. Further, when the T3-timeout occurs, we have active_patch == retrans_path, and even though the timeout occurred on the initial transmission of data, not a retransmit, we end up updating retransmit path. RFC4960, section 6.4. "Multi-Homed SCTP Endpoints" states under 6.4.1. "Failover from an Inactive Destination Address" the following: Some of the transport addresses of a multi-homed SCTP endpoint may become inactive due to either the occurrence of certain error conditions (see Section 8.2) or adjustments from the SCTP user. When there is outbound data to send and the primary path becomes inactive (e.g., due to failures), or where the SCTP user explicitly requests to send data to an inactive destination transport address, before reporting an error to its ULP, the SCTP endpoint should try to send the data to an alternate __active__ destination transport address if one exists. When retransmitting data that timed out, if the endpoint is multihomed, it should consider each source-destination address pair in its retransmission selection policy. When retransmitting timed-out data, the endpoint should attempt to pick the most divergent source-destination pair from the original source-destination pair to which the packet was transmitted. Note: Rules for picking the most divergent source-destination pair are an implementation decision and are not specified within this document. So, we should first reconsider to take the current active retransmission transport if we cannot find an alternative active one. If all of that fails, we can still round robin through unkown, partial failover, and inactive ones in the hope to find something still suitable. Commit 4141ddc02a92 ("sctp: retran_path update bug fix") broke that behaviour by selecting the next inactive transport when no other active transport was found besides the current assoc's peer.retran_path. Before commit 4141ddc02a92, we would have traversed through the list until we reach our peer.retran_path again, and in case that is still in state SCTP_ACTIVE, we would take it and return. Only if that is not the case either, we take the next inactive transport. Besides all that, another issue is that transports in state SCTP_UNKNOWN could be preferred over transports in state SCTP_ACTIVE in case a SCTP_ACTIVE transport appears after SCTP_UNKNOWN in the transport list yielding a weaker transport state to be used in retransmission. This patch mostly reverts 4141ddc02a92, but also rewrites this function to introduce more clarity and strictness into the code. A strict priority of transport states is enforced in this patch, hence selection is active > unkown > partial failover > inactive. Fixes: 4141ddc02a92 ("sctp: retran_path update bug fix") Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Gui Jianfeng <guijianfeng@cn.fujitsu.com> Acked-by: Vlad Yasevich <yasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-20 20:51:06 +01:00
/* Iterate from retran_path's successor back to retran_path. */
for (trans = list_next_entry(trans, transports); 1;
trans = list_next_entry(trans, transports)) {
/* Manually skip the head element. */
if (&trans->transports == &asoc->peer.transport_addr_list)
continue;
if (trans->state == SCTP_UNCONFIRMED)
continue;
trans_next = sctp_trans_elect_best(trans, trans_next);
/* Active is good enough for immediate return. */
if (trans_next->state == SCTP_ACTIVE)
break;
net: sctp: rework multihoming retransmission path selection to rfc4960 Problem statement: 1) both paths (primary path1 and alternate path2) are up after the association has been established i.e., HB packets are normally exchanged, 2) path2 gets inactive after path_max_retrans * max_rto timed out (i.e. path2 is down completely), 3) now, if a transmission times out on the only surviving/active path1 (any ~1sec network service impact could cause this like a channel bonding failover), then the retransmitted packets are sent over the inactive path2; this happens with partial failover and without it. Besides not being optimal in the above scenario, a small failure or timeout in the only existing path has the potential to cause long delays in the retransmission (depending on RTO_MAX) until the still active path is reselected. Further, when the T3-timeout occurs, we have active_patch == retrans_path, and even though the timeout occurred on the initial transmission of data, not a retransmit, we end up updating retransmit path. RFC4960, section 6.4. "Multi-Homed SCTP Endpoints" states under 6.4.1. "Failover from an Inactive Destination Address" the following: Some of the transport addresses of a multi-homed SCTP endpoint may become inactive due to either the occurrence of certain error conditions (see Section 8.2) or adjustments from the SCTP user. When there is outbound data to send and the primary path becomes inactive (e.g., due to failures), or where the SCTP user explicitly requests to send data to an inactive destination transport address, before reporting an error to its ULP, the SCTP endpoint should try to send the data to an alternate __active__ destination transport address if one exists. When retransmitting data that timed out, if the endpoint is multihomed, it should consider each source-destination address pair in its retransmission selection policy. When retransmitting timed-out data, the endpoint should attempt to pick the most divergent source-destination pair from the original source-destination pair to which the packet was transmitted. Note: Rules for picking the most divergent source-destination pair are an implementation decision and are not specified within this document. So, we should first reconsider to take the current active retransmission transport if we cannot find an alternative active one. If all of that fails, we can still round robin through unkown, partial failover, and inactive ones in the hope to find something still suitable. Commit 4141ddc02a92 ("sctp: retran_path update bug fix") broke that behaviour by selecting the next inactive transport when no other active transport was found besides the current assoc's peer.retran_path. Before commit 4141ddc02a92, we would have traversed through the list until we reach our peer.retran_path again, and in case that is still in state SCTP_ACTIVE, we would take it and return. Only if that is not the case either, we take the next inactive transport. Besides all that, another issue is that transports in state SCTP_UNKNOWN could be preferred over transports in state SCTP_ACTIVE in case a SCTP_ACTIVE transport appears after SCTP_UNKNOWN in the transport list yielding a weaker transport state to be used in retransmission. This patch mostly reverts 4141ddc02a92, but also rewrites this function to introduce more clarity and strictness into the code. A strict priority of transport states is enforced in this patch, hence selection is active > unkown > partial failover > inactive. Fixes: 4141ddc02a92 ("sctp: retran_path update bug fix") Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Gui Jianfeng <guijianfeng@cn.fujitsu.com> Acked-by: Vlad Yasevich <yasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-20 20:51:06 +01:00
/* We've reached the end, time to update path. */
if (trans == asoc->peer.retran_path)
break;
}
asoc->peer.retran_path = trans_next;
net: sctp: rework multihoming retransmission path selection to rfc4960 Problem statement: 1) both paths (primary path1 and alternate path2) are up after the association has been established i.e., HB packets are normally exchanged, 2) path2 gets inactive after path_max_retrans * max_rto timed out (i.e. path2 is down completely), 3) now, if a transmission times out on the only surviving/active path1 (any ~1sec network service impact could cause this like a channel bonding failover), then the retransmitted packets are sent over the inactive path2; this happens with partial failover and without it. Besides not being optimal in the above scenario, a small failure or timeout in the only existing path has the potential to cause long delays in the retransmission (depending on RTO_MAX) until the still active path is reselected. Further, when the T3-timeout occurs, we have active_patch == retrans_path, and even though the timeout occurred on the initial transmission of data, not a retransmit, we end up updating retransmit path. RFC4960, section 6.4. "Multi-Homed SCTP Endpoints" states under 6.4.1. "Failover from an Inactive Destination Address" the following: Some of the transport addresses of a multi-homed SCTP endpoint may become inactive due to either the occurrence of certain error conditions (see Section 8.2) or adjustments from the SCTP user. When there is outbound data to send and the primary path becomes inactive (e.g., due to failures), or where the SCTP user explicitly requests to send data to an inactive destination transport address, before reporting an error to its ULP, the SCTP endpoint should try to send the data to an alternate __active__ destination transport address if one exists. When retransmitting data that timed out, if the endpoint is multihomed, it should consider each source-destination address pair in its retransmission selection policy. When retransmitting timed-out data, the endpoint should attempt to pick the most divergent source-destination pair from the original source-destination pair to which the packet was transmitted. Note: Rules for picking the most divergent source-destination pair are an implementation decision and are not specified within this document. So, we should first reconsider to take the current active retransmission transport if we cannot find an alternative active one. If all of that fails, we can still round robin through unkown, partial failover, and inactive ones in the hope to find something still suitable. Commit 4141ddc02a92 ("sctp: retran_path update bug fix") broke that behaviour by selecting the next inactive transport when no other active transport was found besides the current assoc's peer.retran_path. Before commit 4141ddc02a92, we would have traversed through the list until we reach our peer.retran_path again, and in case that is still in state SCTP_ACTIVE, we would take it and return. Only if that is not the case either, we take the next inactive transport. Besides all that, another issue is that transports in state SCTP_UNKNOWN could be preferred over transports in state SCTP_ACTIVE in case a SCTP_ACTIVE transport appears after SCTP_UNKNOWN in the transport list yielding a weaker transport state to be used in retransmission. This patch mostly reverts 4141ddc02a92, but also rewrites this function to introduce more clarity and strictness into the code. A strict priority of transport states is enforced in this patch, hence selection is active > unkown > partial failover > inactive. Fixes: 4141ddc02a92 ("sctp: retran_path update bug fix") Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Gui Jianfeng <guijianfeng@cn.fujitsu.com> Acked-by: Vlad Yasevich <yasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-20 20:51:06 +01:00
pr_debug("%s: association:%p updated new path to addr:%pISpc\n",
__func__, asoc, &asoc->peer.retran_path->ipaddr.sa);
}
static void sctp_select_active_and_retran_path(struct sctp_association *asoc)
{
struct sctp_transport *trans, *trans_pri = NULL, *trans_sec = NULL;
net: sctp: improve sctp_select_active_and_retran_path selection In function sctp_select_active_and_retran_path(), we walk the transport list in order to look for the two most recently used ACTIVE transports (trans_pri, trans_sec). In case we didn't find anything ACTIVE, we currently just camp on a possibly PF or INACTIVE transport that is primary path; this behavior actually dates back to linux-history tree of the very early days of lksctp, and can yield a behavior that chooses suboptimal transport paths. Instead, be a bit more clever by reusing and extending the recently introduced sctp_trans_elect_best() handler. In case both transports are evaluated to have the same score resulting from their states, break the tie by looking at: 1) transport patch error count 2) last_time_heard value from each transport. This is analogous to Nishida's Quick Failover draft [1], section 5.1, 3: The sender SHOULD avoid data transmission to PF destinations. When all destinations are in either PF or Inactive state, the sender MAY either move the destination from PF to active state (and transmit data to the active destination) or the sender MAY transmit data to a PF destination. In the former scenario, (i) the sender MUST NOT notify the ULP about the state transition, and (ii) MUST NOT clear the destination's error counter. It is recommended that the sender picks the PF destination with least error count (fewest consecutive timeouts) for data transmission. In case of a tie (multiple PF destinations with same error count), the sender MAY choose the last active destination. Thus for sctp_select_active_and_retran_path(), we keep track of the best, if any, transport that is in PF state and in case no ACTIVE transport has been found (hence trans_{pri,sec} is NULL), we select the best out of the three: current primary_path and retran_path as well as a possible PF transport. The secondary may still camp on the original primary_path as before. The change in sctp_trans_elect_best() with a more fine grained tie selection also improves at the same time path selection for sctp_assoc_update_retran_path() in case of non-ACTIVE states. [1] http://tools.ietf.org/html/draft-nishida-tsvwg-sctp-failover-05 Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-11 18:19:31 +02:00
struct sctp_transport *trans_pf = NULL;
/* Look for the two most recently used active transports. */
list_for_each_entry(trans, &asoc->peer.transport_addr_list,
transports) {
net: sctp: improve sctp_select_active_and_retran_path selection In function sctp_select_active_and_retran_path(), we walk the transport list in order to look for the two most recently used ACTIVE transports (trans_pri, trans_sec). In case we didn't find anything ACTIVE, we currently just camp on a possibly PF or INACTIVE transport that is primary path; this behavior actually dates back to linux-history tree of the very early days of lksctp, and can yield a behavior that chooses suboptimal transport paths. Instead, be a bit more clever by reusing and extending the recently introduced sctp_trans_elect_best() handler. In case both transports are evaluated to have the same score resulting from their states, break the tie by looking at: 1) transport patch error count 2) last_time_heard value from each transport. This is analogous to Nishida's Quick Failover draft [1], section 5.1, 3: The sender SHOULD avoid data transmission to PF destinations. When all destinations are in either PF or Inactive state, the sender MAY either move the destination from PF to active state (and transmit data to the active destination) or the sender MAY transmit data to a PF destination. In the former scenario, (i) the sender MUST NOT notify the ULP about the state transition, and (ii) MUST NOT clear the destination's error counter. It is recommended that the sender picks the PF destination with least error count (fewest consecutive timeouts) for data transmission. In case of a tie (multiple PF destinations with same error count), the sender MAY choose the last active destination. Thus for sctp_select_active_and_retran_path(), we keep track of the best, if any, transport that is in PF state and in case no ACTIVE transport has been found (hence trans_{pri,sec} is NULL), we select the best out of the three: current primary_path and retran_path as well as a possible PF transport. The secondary may still camp on the original primary_path as before. The change in sctp_trans_elect_best() with a more fine grained tie selection also improves at the same time path selection for sctp_assoc_update_retran_path() in case of non-ACTIVE states. [1] http://tools.ietf.org/html/draft-nishida-tsvwg-sctp-failover-05 Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-11 18:19:31 +02:00
/* Skip uninteresting transports. */
if (trans->state == SCTP_INACTIVE ||
net: sctp: improve sctp_select_active_and_retran_path selection In function sctp_select_active_and_retran_path(), we walk the transport list in order to look for the two most recently used ACTIVE transports (trans_pri, trans_sec). In case we didn't find anything ACTIVE, we currently just camp on a possibly PF or INACTIVE transport that is primary path; this behavior actually dates back to linux-history tree of the very early days of lksctp, and can yield a behavior that chooses suboptimal transport paths. Instead, be a bit more clever by reusing and extending the recently introduced sctp_trans_elect_best() handler. In case both transports are evaluated to have the same score resulting from their states, break the tie by looking at: 1) transport patch error count 2) last_time_heard value from each transport. This is analogous to Nishida's Quick Failover draft [1], section 5.1, 3: The sender SHOULD avoid data transmission to PF destinations. When all destinations are in either PF or Inactive state, the sender MAY either move the destination from PF to active state (and transmit data to the active destination) or the sender MAY transmit data to a PF destination. In the former scenario, (i) the sender MUST NOT notify the ULP about the state transition, and (ii) MUST NOT clear the destination's error counter. It is recommended that the sender picks the PF destination with least error count (fewest consecutive timeouts) for data transmission. In case of a tie (multiple PF destinations with same error count), the sender MAY choose the last active destination. Thus for sctp_select_active_and_retran_path(), we keep track of the best, if any, transport that is in PF state and in case no ACTIVE transport has been found (hence trans_{pri,sec} is NULL), we select the best out of the three: current primary_path and retran_path as well as a possible PF transport. The secondary may still camp on the original primary_path as before. The change in sctp_trans_elect_best() with a more fine grained tie selection also improves at the same time path selection for sctp_assoc_update_retran_path() in case of non-ACTIVE states. [1] http://tools.ietf.org/html/draft-nishida-tsvwg-sctp-failover-05 Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-11 18:19:31 +02:00
trans->state == SCTP_UNCONFIRMED)
continue;
net: sctp: improve sctp_select_active_and_retran_path selection In function sctp_select_active_and_retran_path(), we walk the transport list in order to look for the two most recently used ACTIVE transports (trans_pri, trans_sec). In case we didn't find anything ACTIVE, we currently just camp on a possibly PF or INACTIVE transport that is primary path; this behavior actually dates back to linux-history tree of the very early days of lksctp, and can yield a behavior that chooses suboptimal transport paths. Instead, be a bit more clever by reusing and extending the recently introduced sctp_trans_elect_best() handler. In case both transports are evaluated to have the same score resulting from their states, break the tie by looking at: 1) transport patch error count 2) last_time_heard value from each transport. This is analogous to Nishida's Quick Failover draft [1], section 5.1, 3: The sender SHOULD avoid data transmission to PF destinations. When all destinations are in either PF or Inactive state, the sender MAY either move the destination from PF to active state (and transmit data to the active destination) or the sender MAY transmit data to a PF destination. In the former scenario, (i) the sender MUST NOT notify the ULP about the state transition, and (ii) MUST NOT clear the destination's error counter. It is recommended that the sender picks the PF destination with least error count (fewest consecutive timeouts) for data transmission. In case of a tie (multiple PF destinations with same error count), the sender MAY choose the last active destination. Thus for sctp_select_active_and_retran_path(), we keep track of the best, if any, transport that is in PF state and in case no ACTIVE transport has been found (hence trans_{pri,sec} is NULL), we select the best out of the three: current primary_path and retran_path as well as a possible PF transport. The secondary may still camp on the original primary_path as before. The change in sctp_trans_elect_best() with a more fine grained tie selection also improves at the same time path selection for sctp_assoc_update_retran_path() in case of non-ACTIVE states. [1] http://tools.ietf.org/html/draft-nishida-tsvwg-sctp-failover-05 Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-11 18:19:31 +02:00
/* Keep track of the best PF transport from our
* list in case we don't find an active one.
*/
if (trans->state == SCTP_PF) {
trans_pf = sctp_trans_elect_best(trans, trans_pf);
continue;
}
/* For active transports, pick the most recent ones. */
if (trans_pri == NULL ||
ktime_after(trans->last_time_heard,
trans_pri->last_time_heard)) {
trans_sec = trans_pri;
trans_pri = trans;
} else if (trans_sec == NULL ||
ktime_after(trans->last_time_heard,
trans_sec->last_time_heard)) {
trans_sec = trans;
}
}
/* RFC 2960 6.4 Multi-Homed SCTP Endpoints
*
* By default, an endpoint should always transmit to the primary
* path, unless the SCTP user explicitly specifies the
* destination transport address (and possibly source transport
* address) to use. [If the primary is active but not most recent,
* bump the most recently used transport.]
*/
if ((asoc->peer.primary_path->state == SCTP_ACTIVE ||
asoc->peer.primary_path->state == SCTP_UNKNOWN) &&
asoc->peer.primary_path != trans_pri) {
trans_sec = trans_pri;
trans_pri = asoc->peer.primary_path;
}
/* We did not find anything useful for a possible retransmission
* path; either primary path that we found is the the same as
* the current one, or we didn't generally find an active one.
*/
if (trans_sec == NULL)
trans_sec = trans_pri;
/* If we failed to find a usable transport, just camp on the
net: sctp: fix suboptimal edge-case on non-active active/retrans path selection In SCTP, selection of active (T.ACT) and retransmission (T.RET) transports is being done whenever transport control operations (UP, DOWN, PF, ...) are engaged through sctp_assoc_control_transport(). Commits 4c47af4d5eb2 ("net: sctp: rework multihoming retransmission path selection to rfc4960") and a7288c4dd509 ("net: sctp: improve sctp_select_active_and_retran_path selection") have both improved it towards a more fine-grained and optimal path selection. Currently, the selection algorithm for T.ACT and T.RET is as follows: 1) Elect the two most recently used ACTIVE transports T1, T2 for T.ACT, T.RET, where T.ACT<-T1 and T1 is most recently used 2) In case primary path T.PRI not in {T1, T2} but ACTIVE, set T.ACT<-T.PRI and T.RET<-T1 3) If only T1 is ACTIVE from the set, set T.ACT<-T1 and T.RET<-T1 4) If none is ACTIVE, set T.ACT<-best(T.PRI, T.RET, T3) where T3 is the most recently used (if avail) in PF, set T.RET<-T.PRI Prior to above commits, 4) was simply a camp on T.ACT<-T.PRI and T.RET<-T.PRI, ignoring possible paths in PF. Camping on T.PRI is still slightly suboptimal as it can lead to the following scenario: Setup: <A> <B> T1: p1p1 (10.0.10.10) <==> .'`) <==> p1p1 (10.0.10.12) <= T.PRI T2: p1p2 (10.0.10.20) <==> (_ . ) <==> p1p2 (10.0.10.22) net.sctp.rto_min = 1000 net.sctp.path_max_retrans = 2 net.sctp.pf_retrans = 0 net.sctp.hb_interval = 1000 T.PRI is permanently down, T2 is put briefly into PF state (e.g. due to link flapping). Here, the first time transmission is sent over PF path T2 as it's the only non-INACTIVE path, but the retransmitted data-chunks are sent over the INACTIVE path T1 (T.PRI), which is not good. After the patch, it's choosing better transports in both cases by modifying step 4): 4) If none is ACTIVE, set T.ACT_new<-best(T.ACT_old, T3) where T3 is the most recently used (if avail) in PF, set T.RET<-T.ACT_new This will still select a best possible path in PF if available (which can also include T.PRI/T.RET), and set both T.ACT/T.RET to it. In case sctp_assoc_control_transport() *just* put T.ACT_old into INACTIVE as it transitioned from ACTIVE->PF->INACTIVE and stays in INACTIVE just for a very short while before going back ACTIVE, it will guarantee that this path will be reselected for T.ACT/T.RET since T3 (PF) is not available. Previously, this was not possible, as we would only select between T.PRI and T.RET, and a possible T3 would be NULL due to the fact that we have just transitioned T3 in sctp_assoc_control_transport() from PF->INACTIVE and would select a suboptimal path when T.PRI/T.RET have worse properties. In the case that T.ACT_old permanently went to INACTIVE during this transition and there's no PF path available, plus T.PRI and T.RET are INACTIVE as well, we would now camp on T.ACT_old, but if everything is being INACTIVE there's really not much we can do except hoping for a successful HB to bring one of the transports back up again and, thus cause a new selection through sctp_assoc_control_transport(). Now both tests work fine: Case 1: 1. T1 S(ACTIVE) T.ACT T2 S(ACTIVE) T.RET 2. T1 S(ACTIVE) T.ACT, T.RET T2 S(PF) 3. T1 S(ACTIVE) T.ACT, T.RET T2 S(INACTIVE) 5. T1 S(PF) T.ACT, T.RET T2 S(INACTIVE) [ 5.1 T1 S(INACTIVE) T.ACT, T.RET T2 S(INACTIVE) ] 6. T1 S(ACTIVE) T.ACT, T.RET T2 S(INACTIVE) 7. T1 S(ACTIVE) T.ACT T2 S(ACTIVE) T.RET Case 2: 1. T1 S(ACTIVE) T.ACT T2 S(ACTIVE) T.RET 2. T1 S(PF) T2 S(ACTIVE) T.ACT, T.RET 3. T1 S(INACTIVE) T2 S(ACTIVE) T.ACT, T.RET 5. T1 S(INACTIVE) T2 S(PF) T.ACT, T.RET [ 5.1 T1 S(INACTIVE) T2 S(INACTIVE) T.ACT, T.RET ] 6. T1 S(INACTIVE) T2 S(ACTIVE) T.ACT, T.RET 7. T1 S(ACTIVE) T.ACT T2 S(ACTIVE) T.RET Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> Acked-by: Vlad Yasevich <vyasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-08-22 13:03:30 +02:00
* active or pick a PF iff it's the better choice.
*/
if (trans_pri == NULL) {
net: sctp: fix suboptimal edge-case on non-active active/retrans path selection In SCTP, selection of active (T.ACT) and retransmission (T.RET) transports is being done whenever transport control operations (UP, DOWN, PF, ...) are engaged through sctp_assoc_control_transport(). Commits 4c47af4d5eb2 ("net: sctp: rework multihoming retransmission path selection to rfc4960") and a7288c4dd509 ("net: sctp: improve sctp_select_active_and_retran_path selection") have both improved it towards a more fine-grained and optimal path selection. Currently, the selection algorithm for T.ACT and T.RET is as follows: 1) Elect the two most recently used ACTIVE transports T1, T2 for T.ACT, T.RET, where T.ACT<-T1 and T1 is most recently used 2) In case primary path T.PRI not in {T1, T2} but ACTIVE, set T.ACT<-T.PRI and T.RET<-T1 3) If only T1 is ACTIVE from the set, set T.ACT<-T1 and T.RET<-T1 4) If none is ACTIVE, set T.ACT<-best(T.PRI, T.RET, T3) where T3 is the most recently used (if avail) in PF, set T.RET<-T.PRI Prior to above commits, 4) was simply a camp on T.ACT<-T.PRI and T.RET<-T.PRI, ignoring possible paths in PF. Camping on T.PRI is still slightly suboptimal as it can lead to the following scenario: Setup: <A> <B> T1: p1p1 (10.0.10.10) <==> .'`) <==> p1p1 (10.0.10.12) <= T.PRI T2: p1p2 (10.0.10.20) <==> (_ . ) <==> p1p2 (10.0.10.22) net.sctp.rto_min = 1000 net.sctp.path_max_retrans = 2 net.sctp.pf_retrans = 0 net.sctp.hb_interval = 1000 T.PRI is permanently down, T2 is put briefly into PF state (e.g. due to link flapping). Here, the first time transmission is sent over PF path T2 as it's the only non-INACTIVE path, but the retransmitted data-chunks are sent over the INACTIVE path T1 (T.PRI), which is not good. After the patch, it's choosing better transports in both cases by modifying step 4): 4) If none is ACTIVE, set T.ACT_new<-best(T.ACT_old, T3) where T3 is the most recently used (if avail) in PF, set T.RET<-T.ACT_new This will still select a best possible path in PF if available (which can also include T.PRI/T.RET), and set both T.ACT/T.RET to it. In case sctp_assoc_control_transport() *just* put T.ACT_old into INACTIVE as it transitioned from ACTIVE->PF->INACTIVE and stays in INACTIVE just for a very short while before going back ACTIVE, it will guarantee that this path will be reselected for T.ACT/T.RET since T3 (PF) is not available. Previously, this was not possible, as we would only select between T.PRI and T.RET, and a possible T3 would be NULL due to the fact that we have just transitioned T3 in sctp_assoc_control_transport() from PF->INACTIVE and would select a suboptimal path when T.PRI/T.RET have worse properties. In the case that T.ACT_old permanently went to INACTIVE during this transition and there's no PF path available, plus T.PRI and T.RET are INACTIVE as well, we would now camp on T.ACT_old, but if everything is being INACTIVE there's really not much we can do except hoping for a successful HB to bring one of the transports back up again and, thus cause a new selection through sctp_assoc_control_transport(). Now both tests work fine: Case 1: 1. T1 S(ACTIVE) T.ACT T2 S(ACTIVE) T.RET 2. T1 S(ACTIVE) T.ACT, T.RET T2 S(PF) 3. T1 S(ACTIVE) T.ACT, T.RET T2 S(INACTIVE) 5. T1 S(PF) T.ACT, T.RET T2 S(INACTIVE) [ 5.1 T1 S(INACTIVE) T.ACT, T.RET T2 S(INACTIVE) ] 6. T1 S(ACTIVE) T.ACT, T.RET T2 S(INACTIVE) 7. T1 S(ACTIVE) T.ACT T2 S(ACTIVE) T.RET Case 2: 1. T1 S(ACTIVE) T.ACT T2 S(ACTIVE) T.RET 2. T1 S(PF) T2 S(ACTIVE) T.ACT, T.RET 3. T1 S(INACTIVE) T2 S(ACTIVE) T.ACT, T.RET 5. T1 S(INACTIVE) T2 S(PF) T.ACT, T.RET [ 5.1 T1 S(INACTIVE) T2 S(INACTIVE) T.ACT, T.RET ] 6. T1 S(INACTIVE) T2 S(ACTIVE) T.ACT, T.RET 7. T1 S(ACTIVE) T.ACT T2 S(ACTIVE) T.RET Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> Acked-by: Vlad Yasevich <vyasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-08-22 13:03:30 +02:00
trans_pri = sctp_trans_elect_best(asoc->peer.active_path, trans_pf);
trans_sec = trans_pri;
}
/* Set the active and retran transports. */
asoc->peer.active_path = trans_pri;
asoc->peer.retran_path = trans_sec;
}
net: sctp: rework multihoming retransmission path selection to rfc4960 Problem statement: 1) both paths (primary path1 and alternate path2) are up after the association has been established i.e., HB packets are normally exchanged, 2) path2 gets inactive after path_max_retrans * max_rto timed out (i.e. path2 is down completely), 3) now, if a transmission times out on the only surviving/active path1 (any ~1sec network service impact could cause this like a channel bonding failover), then the retransmitted packets are sent over the inactive path2; this happens with partial failover and without it. Besides not being optimal in the above scenario, a small failure or timeout in the only existing path has the potential to cause long delays in the retransmission (depending on RTO_MAX) until the still active path is reselected. Further, when the T3-timeout occurs, we have active_patch == retrans_path, and even though the timeout occurred on the initial transmission of data, not a retransmit, we end up updating retransmit path. RFC4960, section 6.4. "Multi-Homed SCTP Endpoints" states under 6.4.1. "Failover from an Inactive Destination Address" the following: Some of the transport addresses of a multi-homed SCTP endpoint may become inactive due to either the occurrence of certain error conditions (see Section 8.2) or adjustments from the SCTP user. When there is outbound data to send and the primary path becomes inactive (e.g., due to failures), or where the SCTP user explicitly requests to send data to an inactive destination transport address, before reporting an error to its ULP, the SCTP endpoint should try to send the data to an alternate __active__ destination transport address if one exists. When retransmitting data that timed out, if the endpoint is multihomed, it should consider each source-destination address pair in its retransmission selection policy. When retransmitting timed-out data, the endpoint should attempt to pick the most divergent source-destination pair from the original source-destination pair to which the packet was transmitted. Note: Rules for picking the most divergent source-destination pair are an implementation decision and are not specified within this document. So, we should first reconsider to take the current active retransmission transport if we cannot find an alternative active one. If all of that fails, we can still round robin through unkown, partial failover, and inactive ones in the hope to find something still suitable. Commit 4141ddc02a92 ("sctp: retran_path update bug fix") broke that behaviour by selecting the next inactive transport when no other active transport was found besides the current assoc's peer.retran_path. Before commit 4141ddc02a92, we would have traversed through the list until we reach our peer.retran_path again, and in case that is still in state SCTP_ACTIVE, we would take it and return. Only if that is not the case either, we take the next inactive transport. Besides all that, another issue is that transports in state SCTP_UNKNOWN could be preferred over transports in state SCTP_ACTIVE in case a SCTP_ACTIVE transport appears after SCTP_UNKNOWN in the transport list yielding a weaker transport state to be used in retransmission. This patch mostly reverts 4141ddc02a92, but also rewrites this function to introduce more clarity and strictness into the code. A strict priority of transport states is enforced in this patch, hence selection is active > unkown > partial failover > inactive. Fixes: 4141ddc02a92 ("sctp: retran_path update bug fix") Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Gui Jianfeng <guijianfeng@cn.fujitsu.com> Acked-by: Vlad Yasevich <yasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-20 20:51:06 +01:00
struct sctp_transport *
sctp_assoc_choose_alter_transport(struct sctp_association *asoc,
struct sctp_transport *last_sent_to)
{
/* If this is the first time packet is sent, use the active path,
* else use the retran path. If the last packet was sent over the
* retran path, update the retran path and use it.
*/
net: sctp: rework multihoming retransmission path selection to rfc4960 Problem statement: 1) both paths (primary path1 and alternate path2) are up after the association has been established i.e., HB packets are normally exchanged, 2) path2 gets inactive after path_max_retrans * max_rto timed out (i.e. path2 is down completely), 3) now, if a transmission times out on the only surviving/active path1 (any ~1sec network service impact could cause this like a channel bonding failover), then the retransmitted packets are sent over the inactive path2; this happens with partial failover and without it. Besides not being optimal in the above scenario, a small failure or timeout in the only existing path has the potential to cause long delays in the retransmission (depending on RTO_MAX) until the still active path is reselected. Further, when the T3-timeout occurs, we have active_patch == retrans_path, and even though the timeout occurred on the initial transmission of data, not a retransmit, we end up updating retransmit path. RFC4960, section 6.4. "Multi-Homed SCTP Endpoints" states under 6.4.1. "Failover from an Inactive Destination Address" the following: Some of the transport addresses of a multi-homed SCTP endpoint may become inactive due to either the occurrence of certain error conditions (see Section 8.2) or adjustments from the SCTP user. When there is outbound data to send and the primary path becomes inactive (e.g., due to failures), or where the SCTP user explicitly requests to send data to an inactive destination transport address, before reporting an error to its ULP, the SCTP endpoint should try to send the data to an alternate __active__ destination transport address if one exists. When retransmitting data that timed out, if the endpoint is multihomed, it should consider each source-destination address pair in its retransmission selection policy. When retransmitting timed-out data, the endpoint should attempt to pick the most divergent source-destination pair from the original source-destination pair to which the packet was transmitted. Note: Rules for picking the most divergent source-destination pair are an implementation decision and are not specified within this document. So, we should first reconsider to take the current active retransmission transport if we cannot find an alternative active one. If all of that fails, we can still round robin through unkown, partial failover, and inactive ones in the hope to find something still suitable. Commit 4141ddc02a92 ("sctp: retran_path update bug fix") broke that behaviour by selecting the next inactive transport when no other active transport was found besides the current assoc's peer.retran_path. Before commit 4141ddc02a92, we would have traversed through the list until we reach our peer.retran_path again, and in case that is still in state SCTP_ACTIVE, we would take it and return. Only if that is not the case either, we take the next inactive transport. Besides all that, another issue is that transports in state SCTP_UNKNOWN could be preferred over transports in state SCTP_ACTIVE in case a SCTP_ACTIVE transport appears after SCTP_UNKNOWN in the transport list yielding a weaker transport state to be used in retransmission. This patch mostly reverts 4141ddc02a92, but also rewrites this function to introduce more clarity and strictness into the code. A strict priority of transport states is enforced in this patch, hence selection is active > unkown > partial failover > inactive. Fixes: 4141ddc02a92 ("sctp: retran_path update bug fix") Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Gui Jianfeng <guijianfeng@cn.fujitsu.com> Acked-by: Vlad Yasevich <yasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-20 20:51:06 +01:00
if (last_sent_to == NULL) {
return asoc->peer.active_path;
net: sctp: rework multihoming retransmission path selection to rfc4960 Problem statement: 1) both paths (primary path1 and alternate path2) are up after the association has been established i.e., HB packets are normally exchanged, 2) path2 gets inactive after path_max_retrans * max_rto timed out (i.e. path2 is down completely), 3) now, if a transmission times out on the only surviving/active path1 (any ~1sec network service impact could cause this like a channel bonding failover), then the retransmitted packets are sent over the inactive path2; this happens with partial failover and without it. Besides not being optimal in the above scenario, a small failure or timeout in the only existing path has the potential to cause long delays in the retransmission (depending on RTO_MAX) until the still active path is reselected. Further, when the T3-timeout occurs, we have active_patch == retrans_path, and even though the timeout occurred on the initial transmission of data, not a retransmit, we end up updating retransmit path. RFC4960, section 6.4. "Multi-Homed SCTP Endpoints" states under 6.4.1. "Failover from an Inactive Destination Address" the following: Some of the transport addresses of a multi-homed SCTP endpoint may become inactive due to either the occurrence of certain error conditions (see Section 8.2) or adjustments from the SCTP user. When there is outbound data to send and the primary path becomes inactive (e.g., due to failures), or where the SCTP user explicitly requests to send data to an inactive destination transport address, before reporting an error to its ULP, the SCTP endpoint should try to send the data to an alternate __active__ destination transport address if one exists. When retransmitting data that timed out, if the endpoint is multihomed, it should consider each source-destination address pair in its retransmission selection policy. When retransmitting timed-out data, the endpoint should attempt to pick the most divergent source-destination pair from the original source-destination pair to which the packet was transmitted. Note: Rules for picking the most divergent source-destination pair are an implementation decision and are not specified within this document. So, we should first reconsider to take the current active retransmission transport if we cannot find an alternative active one. If all of that fails, we can still round robin through unkown, partial failover, and inactive ones in the hope to find something still suitable. Commit 4141ddc02a92 ("sctp: retran_path update bug fix") broke that behaviour by selecting the next inactive transport when no other active transport was found besides the current assoc's peer.retran_path. Before commit 4141ddc02a92, we would have traversed through the list until we reach our peer.retran_path again, and in case that is still in state SCTP_ACTIVE, we would take it and return. Only if that is not the case either, we take the next inactive transport. Besides all that, another issue is that transports in state SCTP_UNKNOWN could be preferred over transports in state SCTP_ACTIVE in case a SCTP_ACTIVE transport appears after SCTP_UNKNOWN in the transport list yielding a weaker transport state to be used in retransmission. This patch mostly reverts 4141ddc02a92, but also rewrites this function to introduce more clarity and strictness into the code. A strict priority of transport states is enforced in this patch, hence selection is active > unkown > partial failover > inactive. Fixes: 4141ddc02a92 ("sctp: retran_path update bug fix") Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Gui Jianfeng <guijianfeng@cn.fujitsu.com> Acked-by: Vlad Yasevich <yasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-20 20:51:06 +01:00
} else {
if (last_sent_to == asoc->peer.retran_path)
sctp_assoc_update_retran_path(asoc);
net: sctp: rework multihoming retransmission path selection to rfc4960 Problem statement: 1) both paths (primary path1 and alternate path2) are up after the association has been established i.e., HB packets are normally exchanged, 2) path2 gets inactive after path_max_retrans * max_rto timed out (i.e. path2 is down completely), 3) now, if a transmission times out on the only surviving/active path1 (any ~1sec network service impact could cause this like a channel bonding failover), then the retransmitted packets are sent over the inactive path2; this happens with partial failover and without it. Besides not being optimal in the above scenario, a small failure or timeout in the only existing path has the potential to cause long delays in the retransmission (depending on RTO_MAX) until the still active path is reselected. Further, when the T3-timeout occurs, we have active_patch == retrans_path, and even though the timeout occurred on the initial transmission of data, not a retransmit, we end up updating retransmit path. RFC4960, section 6.4. "Multi-Homed SCTP Endpoints" states under 6.4.1. "Failover from an Inactive Destination Address" the following: Some of the transport addresses of a multi-homed SCTP endpoint may become inactive due to either the occurrence of certain error conditions (see Section 8.2) or adjustments from the SCTP user. When there is outbound data to send and the primary path becomes inactive (e.g., due to failures), or where the SCTP user explicitly requests to send data to an inactive destination transport address, before reporting an error to its ULP, the SCTP endpoint should try to send the data to an alternate __active__ destination transport address if one exists. When retransmitting data that timed out, if the endpoint is multihomed, it should consider each source-destination address pair in its retransmission selection policy. When retransmitting timed-out data, the endpoint should attempt to pick the most divergent source-destination pair from the original source-destination pair to which the packet was transmitted. Note: Rules for picking the most divergent source-destination pair are an implementation decision and are not specified within this document. So, we should first reconsider to take the current active retransmission transport if we cannot find an alternative active one. If all of that fails, we can still round robin through unkown, partial failover, and inactive ones in the hope to find something still suitable. Commit 4141ddc02a92 ("sctp: retran_path update bug fix") broke that behaviour by selecting the next inactive transport when no other active transport was found besides the current assoc's peer.retran_path. Before commit 4141ddc02a92, we would have traversed through the list until we reach our peer.retran_path again, and in case that is still in state SCTP_ACTIVE, we would take it and return. Only if that is not the case either, we take the next inactive transport. Besides all that, another issue is that transports in state SCTP_UNKNOWN could be preferred over transports in state SCTP_ACTIVE in case a SCTP_ACTIVE transport appears after SCTP_UNKNOWN in the transport list yielding a weaker transport state to be used in retransmission. This patch mostly reverts 4141ddc02a92, but also rewrites this function to introduce more clarity and strictness into the code. A strict priority of transport states is enforced in this patch, hence selection is active > unkown > partial failover > inactive. Fixes: 4141ddc02a92 ("sctp: retran_path update bug fix") Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Gui Jianfeng <guijianfeng@cn.fujitsu.com> Acked-by: Vlad Yasevich <yasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-02-20 20:51:06 +01:00
return asoc->peer.retran_path;
}
}
/* Update the association's pmtu and frag_point by going through all the
* transports. This routine is called when a transport's PMTU has changed.
*/
void sctp_assoc_sync_pmtu(struct sock *sk, struct sctp_association *asoc)
{
struct sctp_transport *t;
__u32 pmtu = 0;
if (!asoc)
return;
/* Get the lowest pmtu of all the transports. */
list_for_each_entry(t, &asoc->peer.transport_addr_list,
transports) {
if (t->pmtu_pending && t->dst) {
sctp_transport_update_pmtu(sk, t, dst_mtu(t->dst));
t->pmtu_pending = 0;
}
if (!pmtu || (t->pathmtu < pmtu))
pmtu = t->pathmtu;
}
if (pmtu) {
asoc->pathmtu = pmtu;
asoc->frag_point = sctp_frag_point(asoc, pmtu);
}
net: sctp: rework debugging framework to use pr_debug and friends We should get rid of all own SCTP debug printk macros and use the ones that the kernel offers anyway instead. This makes the code more readable and conform to the kernel code, and offers all the features of dynamic debbuging that pr_debug() et al has, such as only turning on/off portions of debug messages at runtime through debugfs. The runtime cost of having CONFIG_DYNAMIC_DEBUG enabled, but none of the debug statements printing, is negligible [1]. If kernel debugging is completly turned off, then these statements will also compile into "empty" functions. While we're at it, we also need to change the Kconfig option as it /now/ only refers to the ifdef'ed code portions in outqueue.c that enable further debugging/tracing of SCTP transaction fields. Also, since SCTP_ASSERT code was enabled with this Kconfig option and has now been removed, we transform those code parts into WARNs resp. where appropriate BUG_ONs so that those bugs can be more easily detected as probably not many people have SCTP debugging permanently turned on. To turn on all SCTP debugging, the following steps are needed: # mount -t debugfs none /sys/kernel/debug # echo -n 'module sctp +p' > /sys/kernel/debug/dynamic_debug/control This can be done more fine-grained on a per file, per line basis and others as described in [2]. [1] https://www.kernel.org/doc/ols/2009/ols2009-pages-39-46.pdf [2] Documentation/dynamic-debug-howto.txt Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-06-28 19:49:40 +02:00
pr_debug("%s: asoc:%p, pmtu:%d, frag_point:%d\n", __func__, asoc,
asoc->pathmtu, asoc->frag_point);
}
/* Should we send a SACK to update our peer? */
static inline bool sctp_peer_needs_update(struct sctp_association *asoc)
{
struct net *net = sock_net(asoc->base.sk);
switch (asoc->state) {
case SCTP_STATE_ESTABLISHED:
case SCTP_STATE_SHUTDOWN_PENDING:
case SCTP_STATE_SHUTDOWN_RECEIVED:
case SCTP_STATE_SHUTDOWN_SENT:
if ((asoc->rwnd > asoc->a_rwnd) &&
((asoc->rwnd - asoc->a_rwnd) >= max_t(__u32,
(asoc->base.sk->sk_rcvbuf >> net->sctp.rwnd_upd_shift),
asoc->pathmtu)))
return true;
break;
default:
break;
}
return false;
}
Revert "net: sctp: Fix a_rwnd/rwnd management to reflect real state of the receiver's buffer" This reverts commit ef2820a735f7 ("net: sctp: Fix a_rwnd/rwnd management to reflect real state of the receiver's buffer") as it introduced a serious performance regression on SCTP over IPv4 and IPv6, though a not as dramatic on the latter. Measurements are on 10Gbit/s with ixgbe NICs. Current state: [root@Lab200slot2 ~]# iperf3 --sctp -4 -c 192.168.241.3 -V -l 1452 -t 60 iperf version 3.0.1 (10 January 2014) Linux Lab200slot2 3.14.0 #1 SMP Thu Apr 3 23:18:29 EDT 2014 x86_64 Time: Fri, 11 Apr 2014 17:56:21 GMT Connecting to host 192.168.241.3, port 5201 Cookie: Lab200slot2.1397238981.812898.548918 [ 4] local 192.168.241.2 port 38616 connected to 192.168.241.3 port 5201 Starting Test: protocol: SCTP, 1 streams, 1452 byte blocks, omitting 0 seconds, 60 second test [ ID] Interval Transfer Bandwidth [ 4] 0.00-1.09 sec 20.8 MBytes 161 Mbits/sec [ 4] 1.09-2.13 sec 10.8 MBytes 86.8 Mbits/sec [ 4] 2.13-3.15 sec 3.57 MBytes 29.5 Mbits/sec [ 4] 3.15-4.16 sec 4.33 MBytes 35.7 Mbits/sec [ 4] 4.16-6.21 sec 10.4 MBytes 42.7 Mbits/sec [ 4] 6.21-6.21 sec 0.00 Bytes 0.00 bits/sec [ 4] 6.21-7.35 sec 34.6 MBytes 253 Mbits/sec [ 4] 7.35-11.45 sec 22.0 MBytes 45.0 Mbits/sec [ 4] 11.45-11.45 sec 0.00 Bytes 0.00 bits/sec [ 4] 11.45-11.45 sec 0.00 Bytes 0.00 bits/sec [ 4] 11.45-11.45 sec 0.00 Bytes 0.00 bits/sec [ 4] 11.45-12.51 sec 16.0 MBytes 126 Mbits/sec [ 4] 12.51-13.59 sec 20.3 MBytes 158 Mbits/sec [ 4] 13.59-14.65 sec 13.4 MBytes 107 Mbits/sec [ 4] 14.65-16.79 sec 33.3 MBytes 130 Mbits/sec [ 4] 16.79-16.79 sec 0.00 Bytes 0.00 bits/sec [ 4] 16.79-17.82 sec 5.94 MBytes 48.7 Mbits/sec (etc) [root@Lab200slot2 ~]# iperf3 --sctp -6 -c 2001:db8:0:f101::1 -V -l 1400 -t 60 iperf version 3.0.1 (10 January 2014) Linux Lab200slot2 3.14.0 #1 SMP Thu Apr 3 23:18:29 EDT 2014 x86_64 Time: Fri, 11 Apr 2014 19:08:41 GMT Connecting to host 2001:db8:0:f101::1, port 5201 Cookie: Lab200slot2.1397243321.714295.2b3f7c [ 4] local 2001:db8:0:f101::2 port 55804 connected to 2001:db8:0:f101::1 port 5201 Starting Test: protocol: SCTP, 1 streams, 1400 byte blocks, omitting 0 seconds, 60 second test [ ID] Interval Transfer Bandwidth [ 4] 0.00-1.00 sec 169 MBytes 1.42 Gbits/sec [ 4] 1.00-2.00 sec 201 MBytes 1.69 Gbits/sec [ 4] 2.00-3.00 sec 188 MBytes 1.58 Gbits/sec [ 4] 3.00-4.00 sec 174 MBytes 1.46 Gbits/sec [ 4] 4.00-5.00 sec 165 MBytes 1.39 Gbits/sec [ 4] 5.00-6.00 sec 199 MBytes 1.67 Gbits/sec [ 4] 6.00-7.00 sec 163 MBytes 1.36 Gbits/sec [ 4] 7.00-8.00 sec 174 MBytes 1.46 Gbits/sec [ 4] 8.00-9.00 sec 193 MBytes 1.62 Gbits/sec [ 4] 9.00-10.00 sec 196 MBytes 1.65 Gbits/sec [ 4] 10.00-11.00 sec 157 MBytes 1.31 Gbits/sec [ 4] 11.00-12.00 sec 175 MBytes 1.47 Gbits/sec [ 4] 12.00-13.00 sec 192 MBytes 1.61 Gbits/sec [ 4] 13.00-14.00 sec 199 MBytes 1.67 Gbits/sec (etc) After patch: [root@Lab200slot2 ~]# iperf3 --sctp -4 -c 192.168.240.3 -V -l 1452 -t 60 iperf version 3.0.1 (10 January 2014) Linux Lab200slot2 3.14.0+ #1 SMP Mon Apr 14 12:06:40 EDT 2014 x86_64 Time: Mon, 14 Apr 2014 16:40:48 GMT Connecting to host 192.168.240.3, port 5201 Cookie: Lab200slot2.1397493648.413274.65e131 [ 4] local 192.168.240.2 port 50548 connected to 192.168.240.3 port 5201 Starting Test: protocol: SCTP, 1 streams, 1452 byte blocks, omitting 0 seconds, 60 second test [ ID] Interval Transfer Bandwidth [ 4] 0.00-1.00 sec 240 MBytes 2.02 Gbits/sec [ 4] 1.00-2.00 sec 239 MBytes 2.01 Gbits/sec [ 4] 2.00-3.00 sec 240 MBytes 2.01 Gbits/sec [ 4] 3.00-4.00 sec 239 MBytes 2.00 Gbits/sec [ 4] 4.00-5.00 sec 245 MBytes 2.05 Gbits/sec [ 4] 5.00-6.00 sec 240 MBytes 2.01 Gbits/sec [ 4] 6.00-7.00 sec 240 MBytes 2.02 Gbits/sec [ 4] 7.00-8.00 sec 239 MBytes 2.01 Gbits/sec With the reverted patch applied, the SCTP/IPv4 performance is back to normal on latest upstream for IPv4 and IPv6 and has same throughput as 3.4.2 test kernel, steady and interval reports are smooth again. Fixes: ef2820a735f7 ("net: sctp: Fix a_rwnd/rwnd management to reflect real state of the receiver's buffer") Reported-by: Peter Butler <pbutler@sonusnet.com> Reported-by: Dongsheng Song <dongsheng.song@gmail.com> Reported-by: Fengguang Wu <fengguang.wu@intel.com> Tested-by: Peter Butler <pbutler@sonusnet.com> Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Matija Glavinic Pecotic <matija.glavinic-pecotic.ext@nsn.com> Cc: Alexander Sverdlin <alexander.sverdlin@nsn.com> Cc: Vlad Yasevich <vyasevich@gmail.com> Acked-by: Vlad Yasevich <vyasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-04-14 21:45:17 +02:00
/* Increase asoc's rwnd by len and send any window update SACK if needed. */
void sctp_assoc_rwnd_increase(struct sctp_association *asoc, unsigned int len)
{
struct sctp_chunk *sack;
struct timer_list *timer;
Revert "net: sctp: Fix a_rwnd/rwnd management to reflect real state of the receiver's buffer" This reverts commit ef2820a735f7 ("net: sctp: Fix a_rwnd/rwnd management to reflect real state of the receiver's buffer") as it introduced a serious performance regression on SCTP over IPv4 and IPv6, though a not as dramatic on the latter. Measurements are on 10Gbit/s with ixgbe NICs. Current state: [root@Lab200slot2 ~]# iperf3 --sctp -4 -c 192.168.241.3 -V -l 1452 -t 60 iperf version 3.0.1 (10 January 2014) Linux Lab200slot2 3.14.0 #1 SMP Thu Apr 3 23:18:29 EDT 2014 x86_64 Time: Fri, 11 Apr 2014 17:56:21 GMT Connecting to host 192.168.241.3, port 5201 Cookie: Lab200slot2.1397238981.812898.548918 [ 4] local 192.168.241.2 port 38616 connected to 192.168.241.3 port 5201 Starting Test: protocol: SCTP, 1 streams, 1452 byte blocks, omitting 0 seconds, 60 second test [ ID] Interval Transfer Bandwidth [ 4] 0.00-1.09 sec 20.8 MBytes 161 Mbits/sec [ 4] 1.09-2.13 sec 10.8 MBytes 86.8 Mbits/sec [ 4] 2.13-3.15 sec 3.57 MBytes 29.5 Mbits/sec [ 4] 3.15-4.16 sec 4.33 MBytes 35.7 Mbits/sec [ 4] 4.16-6.21 sec 10.4 MBytes 42.7 Mbits/sec [ 4] 6.21-6.21 sec 0.00 Bytes 0.00 bits/sec [ 4] 6.21-7.35 sec 34.6 MBytes 253 Mbits/sec [ 4] 7.35-11.45 sec 22.0 MBytes 45.0 Mbits/sec [ 4] 11.45-11.45 sec 0.00 Bytes 0.00 bits/sec [ 4] 11.45-11.45 sec 0.00 Bytes 0.00 bits/sec [ 4] 11.45-11.45 sec 0.00 Bytes 0.00 bits/sec [ 4] 11.45-12.51 sec 16.0 MBytes 126 Mbits/sec [ 4] 12.51-13.59 sec 20.3 MBytes 158 Mbits/sec [ 4] 13.59-14.65 sec 13.4 MBytes 107 Mbits/sec [ 4] 14.65-16.79 sec 33.3 MBytes 130 Mbits/sec [ 4] 16.79-16.79 sec 0.00 Bytes 0.00 bits/sec [ 4] 16.79-17.82 sec 5.94 MBytes 48.7 Mbits/sec (etc) [root@Lab200slot2 ~]# iperf3 --sctp -6 -c 2001:db8:0:f101::1 -V -l 1400 -t 60 iperf version 3.0.1 (10 January 2014) Linux Lab200slot2 3.14.0 #1 SMP Thu Apr 3 23:18:29 EDT 2014 x86_64 Time: Fri, 11 Apr 2014 19:08:41 GMT Connecting to host 2001:db8:0:f101::1, port 5201 Cookie: Lab200slot2.1397243321.714295.2b3f7c [ 4] local 2001:db8:0:f101::2 port 55804 connected to 2001:db8:0:f101::1 port 5201 Starting Test: protocol: SCTP, 1 streams, 1400 byte blocks, omitting 0 seconds, 60 second test [ ID] Interval Transfer Bandwidth [ 4] 0.00-1.00 sec 169 MBytes 1.42 Gbits/sec [ 4] 1.00-2.00 sec 201 MBytes 1.69 Gbits/sec [ 4] 2.00-3.00 sec 188 MBytes 1.58 Gbits/sec [ 4] 3.00-4.00 sec 174 MBytes 1.46 Gbits/sec [ 4] 4.00-5.00 sec 165 MBytes 1.39 Gbits/sec [ 4] 5.00-6.00 sec 199 MBytes 1.67 Gbits/sec [ 4] 6.00-7.00 sec 163 MBytes 1.36 Gbits/sec [ 4] 7.00-8.00 sec 174 MBytes 1.46 Gbits/sec [ 4] 8.00-9.00 sec 193 MBytes 1.62 Gbits/sec [ 4] 9.00-10.00 sec 196 MBytes 1.65 Gbits/sec [ 4] 10.00-11.00 sec 157 MBytes 1.31 Gbits/sec [ 4] 11.00-12.00 sec 175 MBytes 1.47 Gbits/sec [ 4] 12.00-13.00 sec 192 MBytes 1.61 Gbits/sec [ 4] 13.00-14.00 sec 199 MBytes 1.67 Gbits/sec (etc) After patch: [root@Lab200slot2 ~]# iperf3 --sctp -4 -c 192.168.240.3 -V -l 1452 -t 60 iperf version 3.0.1 (10 January 2014) Linux Lab200slot2 3.14.0+ #1 SMP Mon Apr 14 12:06:40 EDT 2014 x86_64 Time: Mon, 14 Apr 2014 16:40:48 GMT Connecting to host 192.168.240.3, port 5201 Cookie: Lab200slot2.1397493648.413274.65e131 [ 4] local 192.168.240.2 port 50548 connected to 192.168.240.3 port 5201 Starting Test: protocol: SCTP, 1 streams, 1452 byte blocks, omitting 0 seconds, 60 second test [ ID] Interval Transfer Bandwidth [ 4] 0.00-1.00 sec 240 MBytes 2.02 Gbits/sec [ 4] 1.00-2.00 sec 239 MBytes 2.01 Gbits/sec [ 4] 2.00-3.00 sec 240 MBytes 2.01 Gbits/sec [ 4] 3.00-4.00 sec 239 MBytes 2.00 Gbits/sec [ 4] 4.00-5.00 sec 245 MBytes 2.05 Gbits/sec [ 4] 5.00-6.00 sec 240 MBytes 2.01 Gbits/sec [ 4] 6.00-7.00 sec 240 MBytes 2.02 Gbits/sec [ 4] 7.00-8.00 sec 239 MBytes 2.01 Gbits/sec With the reverted patch applied, the SCTP/IPv4 performance is back to normal on latest upstream for IPv4 and IPv6 and has same throughput as 3.4.2 test kernel, steady and interval reports are smooth again. Fixes: ef2820a735f7 ("net: sctp: Fix a_rwnd/rwnd management to reflect real state of the receiver's buffer") Reported-by: Peter Butler <pbutler@sonusnet.com> Reported-by: Dongsheng Song <dongsheng.song@gmail.com> Reported-by: Fengguang Wu <fengguang.wu@intel.com> Tested-by: Peter Butler <pbutler@sonusnet.com> Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Matija Glavinic Pecotic <matija.glavinic-pecotic.ext@nsn.com> Cc: Alexander Sverdlin <alexander.sverdlin@nsn.com> Cc: Vlad Yasevich <vyasevich@gmail.com> Acked-by: Vlad Yasevich <vyasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-04-14 21:45:17 +02:00
if (asoc->rwnd_over) {
if (asoc->rwnd_over >= len) {
asoc->rwnd_over -= len;
} else {
asoc->rwnd += (len - asoc->rwnd_over);
asoc->rwnd_over = 0;
}
} else {
asoc->rwnd += len;
}
Revert "net: sctp: Fix a_rwnd/rwnd management to reflect real state of the receiver's buffer" This reverts commit ef2820a735f7 ("net: sctp: Fix a_rwnd/rwnd management to reflect real state of the receiver's buffer") as it introduced a serious performance regression on SCTP over IPv4 and IPv6, though a not as dramatic on the latter. Measurements are on 10Gbit/s with ixgbe NICs. Current state: [root@Lab200slot2 ~]# iperf3 --sctp -4 -c 192.168.241.3 -V -l 1452 -t 60 iperf version 3.0.1 (10 January 2014) Linux Lab200slot2 3.14.0 #1 SMP Thu Apr 3 23:18:29 EDT 2014 x86_64 Time: Fri, 11 Apr 2014 17:56:21 GMT Connecting to host 192.168.241.3, port 5201 Cookie: Lab200slot2.1397238981.812898.548918 [ 4] local 192.168.241.2 port 38616 connected to 192.168.241.3 port 5201 Starting Test: protocol: SCTP, 1 streams, 1452 byte blocks, omitting 0 seconds, 60 second test [ ID] Interval Transfer Bandwidth [ 4] 0.00-1.09 sec 20.8 MBytes 161 Mbits/sec [ 4] 1.09-2.13 sec 10.8 MBytes 86.8 Mbits/sec [ 4] 2.13-3.15 sec 3.57 MBytes 29.5 Mbits/sec [ 4] 3.15-4.16 sec 4.33 MBytes 35.7 Mbits/sec [ 4] 4.16-6.21 sec 10.4 MBytes 42.7 Mbits/sec [ 4] 6.21-6.21 sec 0.00 Bytes 0.00 bits/sec [ 4] 6.21-7.35 sec 34.6 MBytes 253 Mbits/sec [ 4] 7.35-11.45 sec 22.0 MBytes 45.0 Mbits/sec [ 4] 11.45-11.45 sec 0.00 Bytes 0.00 bits/sec [ 4] 11.45-11.45 sec 0.00 Bytes 0.00 bits/sec [ 4] 11.45-11.45 sec 0.00 Bytes 0.00 bits/sec [ 4] 11.45-12.51 sec 16.0 MBytes 126 Mbits/sec [ 4] 12.51-13.59 sec 20.3 MBytes 158 Mbits/sec [ 4] 13.59-14.65 sec 13.4 MBytes 107 Mbits/sec [ 4] 14.65-16.79 sec 33.3 MBytes 130 Mbits/sec [ 4] 16.79-16.79 sec 0.00 Bytes 0.00 bits/sec [ 4] 16.79-17.82 sec 5.94 MBytes 48.7 Mbits/sec (etc) [root@Lab200slot2 ~]# iperf3 --sctp -6 -c 2001:db8:0:f101::1 -V -l 1400 -t 60 iperf version 3.0.1 (10 January 2014) Linux Lab200slot2 3.14.0 #1 SMP Thu Apr 3 23:18:29 EDT 2014 x86_64 Time: Fri, 11 Apr 2014 19:08:41 GMT Connecting to host 2001:db8:0:f101::1, port 5201 Cookie: Lab200slot2.1397243321.714295.2b3f7c [ 4] local 2001:db8:0:f101::2 port 55804 connected to 2001:db8:0:f101::1 port 5201 Starting Test: protocol: SCTP, 1 streams, 1400 byte blocks, omitting 0 seconds, 60 second test [ ID] Interval Transfer Bandwidth [ 4] 0.00-1.00 sec 169 MBytes 1.42 Gbits/sec [ 4] 1.00-2.00 sec 201 MBytes 1.69 Gbits/sec [ 4] 2.00-3.00 sec 188 MBytes 1.58 Gbits/sec [ 4] 3.00-4.00 sec 174 MBytes 1.46 Gbits/sec [ 4] 4.00-5.00 sec 165 MBytes 1.39 Gbits/sec [ 4] 5.00-6.00 sec 199 MBytes 1.67 Gbits/sec [ 4] 6.00-7.00 sec 163 MBytes 1.36 Gbits/sec [ 4] 7.00-8.00 sec 174 MBytes 1.46 Gbits/sec [ 4] 8.00-9.00 sec 193 MBytes 1.62 Gbits/sec [ 4] 9.00-10.00 sec 196 MBytes 1.65 Gbits/sec [ 4] 10.00-11.00 sec 157 MBytes 1.31 Gbits/sec [ 4] 11.00-12.00 sec 175 MBytes 1.47 Gbits/sec [ 4] 12.00-13.00 sec 192 MBytes 1.61 Gbits/sec [ 4] 13.00-14.00 sec 199 MBytes 1.67 Gbits/sec (etc) After patch: [root@Lab200slot2 ~]# iperf3 --sctp -4 -c 192.168.240.3 -V -l 1452 -t 60 iperf version 3.0.1 (10 January 2014) Linux Lab200slot2 3.14.0+ #1 SMP Mon Apr 14 12:06:40 EDT 2014 x86_64 Time: Mon, 14 Apr 2014 16:40:48 GMT Connecting to host 192.168.240.3, port 5201 Cookie: Lab200slot2.1397493648.413274.65e131 [ 4] local 192.168.240.2 port 50548 connected to 192.168.240.3 port 5201 Starting Test: protocol: SCTP, 1 streams, 1452 byte blocks, omitting 0 seconds, 60 second test [ ID] Interval Transfer Bandwidth [ 4] 0.00-1.00 sec 240 MBytes 2.02 Gbits/sec [ 4] 1.00-2.00 sec 239 MBytes 2.01 Gbits/sec [ 4] 2.00-3.00 sec 240 MBytes 2.01 Gbits/sec [ 4] 3.00-4.00 sec 239 MBytes 2.00 Gbits/sec [ 4] 4.00-5.00 sec 245 MBytes 2.05 Gbits/sec [ 4] 5.00-6.00 sec 240 MBytes 2.01 Gbits/sec [ 4] 6.00-7.00 sec 240 MBytes 2.02 Gbits/sec [ 4] 7.00-8.00 sec 239 MBytes 2.01 Gbits/sec With the reverted patch applied, the SCTP/IPv4 performance is back to normal on latest upstream for IPv4 and IPv6 and has same throughput as 3.4.2 test kernel, steady and interval reports are smooth again. Fixes: ef2820a735f7 ("net: sctp: Fix a_rwnd/rwnd management to reflect real state of the receiver's buffer") Reported-by: Peter Butler <pbutler@sonusnet.com> Reported-by: Dongsheng Song <dongsheng.song@gmail.com> Reported-by: Fengguang Wu <fengguang.wu@intel.com> Tested-by: Peter Butler <pbutler@sonusnet.com> Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Matija Glavinic Pecotic <matija.glavinic-pecotic.ext@nsn.com> Cc: Alexander Sverdlin <alexander.sverdlin@nsn.com> Cc: Vlad Yasevich <vyasevich@gmail.com> Acked-by: Vlad Yasevich <vyasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-04-14 21:45:17 +02:00
/* If we had window pressure, start recovering it
* once our rwnd had reached the accumulated pressure
* threshold. The idea is to recover slowly, but up
* to the initial advertised window.
*/
if (asoc->rwnd_press && asoc->rwnd >= asoc->rwnd_press) {
int change = min(asoc->pathmtu, asoc->rwnd_press);
asoc->rwnd += change;
asoc->rwnd_press -= change;
}
Revert "net: sctp: Fix a_rwnd/rwnd management to reflect real state of the receiver's buffer" This reverts commit ef2820a735f7 ("net: sctp: Fix a_rwnd/rwnd management to reflect real state of the receiver's buffer") as it introduced a serious performance regression on SCTP over IPv4 and IPv6, though a not as dramatic on the latter. Measurements are on 10Gbit/s with ixgbe NICs. Current state: [root@Lab200slot2 ~]# iperf3 --sctp -4 -c 192.168.241.3 -V -l 1452 -t 60 iperf version 3.0.1 (10 January 2014) Linux Lab200slot2 3.14.0 #1 SMP Thu Apr 3 23:18:29 EDT 2014 x86_64 Time: Fri, 11 Apr 2014 17:56:21 GMT Connecting to host 192.168.241.3, port 5201 Cookie: Lab200slot2.1397238981.812898.548918 [ 4] local 192.168.241.2 port 38616 connected to 192.168.241.3 port 5201 Starting Test: protocol: SCTP, 1 streams, 1452 byte blocks, omitting 0 seconds, 60 second test [ ID] Interval Transfer Bandwidth [ 4] 0.00-1.09 sec 20.8 MBytes 161 Mbits/sec [ 4] 1.09-2.13 sec 10.8 MBytes 86.8 Mbits/sec [ 4] 2.13-3.15 sec 3.57 MBytes 29.5 Mbits/sec [ 4] 3.15-4.16 sec 4.33 MBytes 35.7 Mbits/sec [ 4] 4.16-6.21 sec 10.4 MBytes 42.7 Mbits/sec [ 4] 6.21-6.21 sec 0.00 Bytes 0.00 bits/sec [ 4] 6.21-7.35 sec 34.6 MBytes 253 Mbits/sec [ 4] 7.35-11.45 sec 22.0 MBytes 45.0 Mbits/sec [ 4] 11.45-11.45 sec 0.00 Bytes 0.00 bits/sec [ 4] 11.45-11.45 sec 0.00 Bytes 0.00 bits/sec [ 4] 11.45-11.45 sec 0.00 Bytes 0.00 bits/sec [ 4] 11.45-12.51 sec 16.0 MBytes 126 Mbits/sec [ 4] 12.51-13.59 sec 20.3 MBytes 158 Mbits/sec [ 4] 13.59-14.65 sec 13.4 MBytes 107 Mbits/sec [ 4] 14.65-16.79 sec 33.3 MBytes 130 Mbits/sec [ 4] 16.79-16.79 sec 0.00 Bytes 0.00 bits/sec [ 4] 16.79-17.82 sec 5.94 MBytes 48.7 Mbits/sec (etc) [root@Lab200slot2 ~]# iperf3 --sctp -6 -c 2001:db8:0:f101::1 -V -l 1400 -t 60 iperf version 3.0.1 (10 January 2014) Linux Lab200slot2 3.14.0 #1 SMP Thu Apr 3 23:18:29 EDT 2014 x86_64 Time: Fri, 11 Apr 2014 19:08:41 GMT Connecting to host 2001:db8:0:f101::1, port 5201 Cookie: Lab200slot2.1397243321.714295.2b3f7c [ 4] local 2001:db8:0:f101::2 port 55804 connected to 2001:db8:0:f101::1 port 5201 Starting Test: protocol: SCTP, 1 streams, 1400 byte blocks, omitting 0 seconds, 60 second test [ ID] Interval Transfer Bandwidth [ 4] 0.00-1.00 sec 169 MBytes 1.42 Gbits/sec [ 4] 1.00-2.00 sec 201 MBytes 1.69 Gbits/sec [ 4] 2.00-3.00 sec 188 MBytes 1.58 Gbits/sec [ 4] 3.00-4.00 sec 174 MBytes 1.46 Gbits/sec [ 4] 4.00-5.00 sec 165 MBytes 1.39 Gbits/sec [ 4] 5.00-6.00 sec 199 MBytes 1.67 Gbits/sec [ 4] 6.00-7.00 sec 163 MBytes 1.36 Gbits/sec [ 4] 7.00-8.00 sec 174 MBytes 1.46 Gbits/sec [ 4] 8.00-9.00 sec 193 MBytes 1.62 Gbits/sec [ 4] 9.00-10.00 sec 196 MBytes 1.65 Gbits/sec [ 4] 10.00-11.00 sec 157 MBytes 1.31 Gbits/sec [ 4] 11.00-12.00 sec 175 MBytes 1.47 Gbits/sec [ 4] 12.00-13.00 sec 192 MBytes 1.61 Gbits/sec [ 4] 13.00-14.00 sec 199 MBytes 1.67 Gbits/sec (etc) After patch: [root@Lab200slot2 ~]# iperf3 --sctp -4 -c 192.168.240.3 -V -l 1452 -t 60 iperf version 3.0.1 (10 January 2014) Linux Lab200slot2 3.14.0+ #1 SMP Mon Apr 14 12:06:40 EDT 2014 x86_64 Time: Mon, 14 Apr 2014 16:40:48 GMT Connecting to host 192.168.240.3, port 5201 Cookie: Lab200slot2.1397493648.413274.65e131 [ 4] local 192.168.240.2 port 50548 connected to 192.168.240.3 port 5201 Starting Test: protocol: SCTP, 1 streams, 1452 byte blocks, omitting 0 seconds, 60 second test [ ID] Interval Transfer Bandwidth [ 4] 0.00-1.00 sec 240 MBytes 2.02 Gbits/sec [ 4] 1.00-2.00 sec 239 MBytes 2.01 Gbits/sec [ 4] 2.00-3.00 sec 240 MBytes 2.01 Gbits/sec [ 4] 3.00-4.00 sec 239 MBytes 2.00 Gbits/sec [ 4] 4.00-5.00 sec 245 MBytes 2.05 Gbits/sec [ 4] 5.00-6.00 sec 240 MBytes 2.01 Gbits/sec [ 4] 6.00-7.00 sec 240 MBytes 2.02 Gbits/sec [ 4] 7.00-8.00 sec 239 MBytes 2.01 Gbits/sec With the reverted patch applied, the SCTP/IPv4 performance is back to normal on latest upstream for IPv4 and IPv6 and has same throughput as 3.4.2 test kernel, steady and interval reports are smooth again. Fixes: ef2820a735f7 ("net: sctp: Fix a_rwnd/rwnd management to reflect real state of the receiver's buffer") Reported-by: Peter Butler <pbutler@sonusnet.com> Reported-by: Dongsheng Song <dongsheng.song@gmail.com> Reported-by: Fengguang Wu <fengguang.wu@intel.com> Tested-by: Peter Butler <pbutler@sonusnet.com> Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Matija Glavinic Pecotic <matija.glavinic-pecotic.ext@nsn.com> Cc: Alexander Sverdlin <alexander.sverdlin@nsn.com> Cc: Vlad Yasevich <vyasevich@gmail.com> Acked-by: Vlad Yasevich <vyasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-04-14 21:45:17 +02:00
pr_debug("%s: asoc:%p rwnd increased by %d to (%u, %u) - %u\n",
__func__, asoc, len, asoc->rwnd, asoc->rwnd_over,
asoc->a_rwnd);
/* Send a window update SACK if the rwnd has increased by at least the
* minimum of the association's PMTU and half of the receive buffer.
* The algorithm used is similar to the one described in
* Section 4.2.3.3 of RFC 1122.
*/
Revert "net: sctp: Fix a_rwnd/rwnd management to reflect real state of the receiver's buffer" This reverts commit ef2820a735f7 ("net: sctp: Fix a_rwnd/rwnd management to reflect real state of the receiver's buffer") as it introduced a serious performance regression on SCTP over IPv4 and IPv6, though a not as dramatic on the latter. Measurements are on 10Gbit/s with ixgbe NICs. Current state: [root@Lab200slot2 ~]# iperf3 --sctp -4 -c 192.168.241.3 -V -l 1452 -t 60 iperf version 3.0.1 (10 January 2014) Linux Lab200slot2 3.14.0 #1 SMP Thu Apr 3 23:18:29 EDT 2014 x86_64 Time: Fri, 11 Apr 2014 17:56:21 GMT Connecting to host 192.168.241.3, port 5201 Cookie: Lab200slot2.1397238981.812898.548918 [ 4] local 192.168.241.2 port 38616 connected to 192.168.241.3 port 5201 Starting Test: protocol: SCTP, 1 streams, 1452 byte blocks, omitting 0 seconds, 60 second test [ ID] Interval Transfer Bandwidth [ 4] 0.00-1.09 sec 20.8 MBytes 161 Mbits/sec [ 4] 1.09-2.13 sec 10.8 MBytes 86.8 Mbits/sec [ 4] 2.13-3.15 sec 3.57 MBytes 29.5 Mbits/sec [ 4] 3.15-4.16 sec 4.33 MBytes 35.7 Mbits/sec [ 4] 4.16-6.21 sec 10.4 MBytes 42.7 Mbits/sec [ 4] 6.21-6.21 sec 0.00 Bytes 0.00 bits/sec [ 4] 6.21-7.35 sec 34.6 MBytes 253 Mbits/sec [ 4] 7.35-11.45 sec 22.0 MBytes 45.0 Mbits/sec [ 4] 11.45-11.45 sec 0.00 Bytes 0.00 bits/sec [ 4] 11.45-11.45 sec 0.00 Bytes 0.00 bits/sec [ 4] 11.45-11.45 sec 0.00 Bytes 0.00 bits/sec [ 4] 11.45-12.51 sec 16.0 MBytes 126 Mbits/sec [ 4] 12.51-13.59 sec 20.3 MBytes 158 Mbits/sec [ 4] 13.59-14.65 sec 13.4 MBytes 107 Mbits/sec [ 4] 14.65-16.79 sec 33.3 MBytes 130 Mbits/sec [ 4] 16.79-16.79 sec 0.00 Bytes 0.00 bits/sec [ 4] 16.79-17.82 sec 5.94 MBytes 48.7 Mbits/sec (etc) [root@Lab200slot2 ~]# iperf3 --sctp -6 -c 2001:db8:0:f101::1 -V -l 1400 -t 60 iperf version 3.0.1 (10 January 2014) Linux Lab200slot2 3.14.0 #1 SMP Thu Apr 3 23:18:29 EDT 2014 x86_64 Time: Fri, 11 Apr 2014 19:08:41 GMT Connecting to host 2001:db8:0:f101::1, port 5201 Cookie: Lab200slot2.1397243321.714295.2b3f7c [ 4] local 2001:db8:0:f101::2 port 55804 connected to 2001:db8:0:f101::1 port 5201 Starting Test: protocol: SCTP, 1 streams, 1400 byte blocks, omitting 0 seconds, 60 second test [ ID] Interval Transfer Bandwidth [ 4] 0.00-1.00 sec 169 MBytes 1.42 Gbits/sec [ 4] 1.00-2.00 sec 201 MBytes 1.69 Gbits/sec [ 4] 2.00-3.00 sec 188 MBytes 1.58 Gbits/sec [ 4] 3.00-4.00 sec 174 MBytes 1.46 Gbits/sec [ 4] 4.00-5.00 sec 165 MBytes 1.39 Gbits/sec [ 4] 5.00-6.00 sec 199 MBytes 1.67 Gbits/sec [ 4] 6.00-7.00 sec 163 MBytes 1.36 Gbits/sec [ 4] 7.00-8.00 sec 174 MBytes 1.46 Gbits/sec [ 4] 8.00-9.00 sec 193 MBytes 1.62 Gbits/sec [ 4] 9.00-10.00 sec 196 MBytes 1.65 Gbits/sec [ 4] 10.00-11.00 sec 157 MBytes 1.31 Gbits/sec [ 4] 11.00-12.00 sec 175 MBytes 1.47 Gbits/sec [ 4] 12.00-13.00 sec 192 MBytes 1.61 Gbits/sec [ 4] 13.00-14.00 sec 199 MBytes 1.67 Gbits/sec (etc) After patch: [root@Lab200slot2 ~]# iperf3 --sctp -4 -c 192.168.240.3 -V -l 1452 -t 60 iperf version 3.0.1 (10 January 2014) Linux Lab200slot2 3.14.0+ #1 SMP Mon Apr 14 12:06:40 EDT 2014 x86_64 Time: Mon, 14 Apr 2014 16:40:48 GMT Connecting to host 192.168.240.3, port 5201 Cookie: Lab200slot2.1397493648.413274.65e131 [ 4] local 192.168.240.2 port 50548 connected to 192.168.240.3 port 5201 Starting Test: protocol: SCTP, 1 streams, 1452 byte blocks, omitting 0 seconds, 60 second test [ ID] Interval Transfer Bandwidth [ 4] 0.00-1.00 sec 240 MBytes 2.02 Gbits/sec [ 4] 1.00-2.00 sec 239 MBytes 2.01 Gbits/sec [ 4] 2.00-3.00 sec 240 MBytes 2.01 Gbits/sec [ 4] 3.00-4.00 sec 239 MBytes 2.00 Gbits/sec [ 4] 4.00-5.00 sec 245 MBytes 2.05 Gbits/sec [ 4] 5.00-6.00 sec 240 MBytes 2.01 Gbits/sec [ 4] 6.00-7.00 sec 240 MBytes 2.02 Gbits/sec [ 4] 7.00-8.00 sec 239 MBytes 2.01 Gbits/sec With the reverted patch applied, the SCTP/IPv4 performance is back to normal on latest upstream for IPv4 and IPv6 and has same throughput as 3.4.2 test kernel, steady and interval reports are smooth again. Fixes: ef2820a735f7 ("net: sctp: Fix a_rwnd/rwnd management to reflect real state of the receiver's buffer") Reported-by: Peter Butler <pbutler@sonusnet.com> Reported-by: Dongsheng Song <dongsheng.song@gmail.com> Reported-by: Fengguang Wu <fengguang.wu@intel.com> Tested-by: Peter Butler <pbutler@sonusnet.com> Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Matija Glavinic Pecotic <matija.glavinic-pecotic.ext@nsn.com> Cc: Alexander Sverdlin <alexander.sverdlin@nsn.com> Cc: Vlad Yasevich <vyasevich@gmail.com> Acked-by: Vlad Yasevich <vyasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-04-14 21:45:17 +02:00
if (sctp_peer_needs_update(asoc)) {
asoc->a_rwnd = asoc->rwnd;
net: sctp: rework debugging framework to use pr_debug and friends We should get rid of all own SCTP debug printk macros and use the ones that the kernel offers anyway instead. This makes the code more readable and conform to the kernel code, and offers all the features of dynamic debbuging that pr_debug() et al has, such as only turning on/off portions of debug messages at runtime through debugfs. The runtime cost of having CONFIG_DYNAMIC_DEBUG enabled, but none of the debug statements printing, is negligible [1]. If kernel debugging is completly turned off, then these statements will also compile into "empty" functions. While we're at it, we also need to change the Kconfig option as it /now/ only refers to the ifdef'ed code portions in outqueue.c that enable further debugging/tracing of SCTP transaction fields. Also, since SCTP_ASSERT code was enabled with this Kconfig option and has now been removed, we transform those code parts into WARNs resp. where appropriate BUG_ONs so that those bugs can be more easily detected as probably not many people have SCTP debugging permanently turned on. To turn on all SCTP debugging, the following steps are needed: # mount -t debugfs none /sys/kernel/debug # echo -n 'module sctp +p' > /sys/kernel/debug/dynamic_debug/control This can be done more fine-grained on a per file, per line basis and others as described in [2]. [1] https://www.kernel.org/doc/ols/2009/ols2009-pages-39-46.pdf [2] Documentation/dynamic-debug-howto.txt Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-06-28 19:49:40 +02:00
pr_debug("%s: sending window update SACK- asoc:%p rwnd:%u "
"a_rwnd:%u\n", __func__, asoc, asoc->rwnd,
asoc->a_rwnd);
sack = sctp_make_sack(asoc);
if (!sack)
return;
asoc->peer.sack_needed = 0;
sctp_outq_tail(&asoc->outqueue, sack);
/* Stop the SACK timer. */
timer = &asoc->timers[SCTP_EVENT_TIMEOUT_SACK];
if (del_timer(timer))
sctp_association_put(asoc);
}
}
Revert "net: sctp: Fix a_rwnd/rwnd management to reflect real state of the receiver's buffer" This reverts commit ef2820a735f7 ("net: sctp: Fix a_rwnd/rwnd management to reflect real state of the receiver's buffer") as it introduced a serious performance regression on SCTP over IPv4 and IPv6, though a not as dramatic on the latter. Measurements are on 10Gbit/s with ixgbe NICs. Current state: [root@Lab200slot2 ~]# iperf3 --sctp -4 -c 192.168.241.3 -V -l 1452 -t 60 iperf version 3.0.1 (10 January 2014) Linux Lab200slot2 3.14.0 #1 SMP Thu Apr 3 23:18:29 EDT 2014 x86_64 Time: Fri, 11 Apr 2014 17:56:21 GMT Connecting to host 192.168.241.3, port 5201 Cookie: Lab200slot2.1397238981.812898.548918 [ 4] local 192.168.241.2 port 38616 connected to 192.168.241.3 port 5201 Starting Test: protocol: SCTP, 1 streams, 1452 byte blocks, omitting 0 seconds, 60 second test [ ID] Interval Transfer Bandwidth [ 4] 0.00-1.09 sec 20.8 MBytes 161 Mbits/sec [ 4] 1.09-2.13 sec 10.8 MBytes 86.8 Mbits/sec [ 4] 2.13-3.15 sec 3.57 MBytes 29.5 Mbits/sec [ 4] 3.15-4.16 sec 4.33 MBytes 35.7 Mbits/sec [ 4] 4.16-6.21 sec 10.4 MBytes 42.7 Mbits/sec [ 4] 6.21-6.21 sec 0.00 Bytes 0.00 bits/sec [ 4] 6.21-7.35 sec 34.6 MBytes 253 Mbits/sec [ 4] 7.35-11.45 sec 22.0 MBytes 45.0 Mbits/sec [ 4] 11.45-11.45 sec 0.00 Bytes 0.00 bits/sec [ 4] 11.45-11.45 sec 0.00 Bytes 0.00 bits/sec [ 4] 11.45-11.45 sec 0.00 Bytes 0.00 bits/sec [ 4] 11.45-12.51 sec 16.0 MBytes 126 Mbits/sec [ 4] 12.51-13.59 sec 20.3 MBytes 158 Mbits/sec [ 4] 13.59-14.65 sec 13.4 MBytes 107 Mbits/sec [ 4] 14.65-16.79 sec 33.3 MBytes 130 Mbits/sec [ 4] 16.79-16.79 sec 0.00 Bytes 0.00 bits/sec [ 4] 16.79-17.82 sec 5.94 MBytes 48.7 Mbits/sec (etc) [root@Lab200slot2 ~]# iperf3 --sctp -6 -c 2001:db8:0:f101::1 -V -l 1400 -t 60 iperf version 3.0.1 (10 January 2014) Linux Lab200slot2 3.14.0 #1 SMP Thu Apr 3 23:18:29 EDT 2014 x86_64 Time: Fri, 11 Apr 2014 19:08:41 GMT Connecting to host 2001:db8:0:f101::1, port 5201 Cookie: Lab200slot2.1397243321.714295.2b3f7c [ 4] local 2001:db8:0:f101::2 port 55804 connected to 2001:db8:0:f101::1 port 5201 Starting Test: protocol: SCTP, 1 streams, 1400 byte blocks, omitting 0 seconds, 60 second test [ ID] Interval Transfer Bandwidth [ 4] 0.00-1.00 sec 169 MBytes 1.42 Gbits/sec [ 4] 1.00-2.00 sec 201 MBytes 1.69 Gbits/sec [ 4] 2.00-3.00 sec 188 MBytes 1.58 Gbits/sec [ 4] 3.00-4.00 sec 174 MBytes 1.46 Gbits/sec [ 4] 4.00-5.00 sec 165 MBytes 1.39 Gbits/sec [ 4] 5.00-6.00 sec 199 MBytes 1.67 Gbits/sec [ 4] 6.00-7.00 sec 163 MBytes 1.36 Gbits/sec [ 4] 7.00-8.00 sec 174 MBytes 1.46 Gbits/sec [ 4] 8.00-9.00 sec 193 MBytes 1.62 Gbits/sec [ 4] 9.00-10.00 sec 196 MBytes 1.65 Gbits/sec [ 4] 10.00-11.00 sec 157 MBytes 1.31 Gbits/sec [ 4] 11.00-12.00 sec 175 MBytes 1.47 Gbits/sec [ 4] 12.00-13.00 sec 192 MBytes 1.61 Gbits/sec [ 4] 13.00-14.00 sec 199 MBytes 1.67 Gbits/sec (etc) After patch: [root@Lab200slot2 ~]# iperf3 --sctp -4 -c 192.168.240.3 -V -l 1452 -t 60 iperf version 3.0.1 (10 January 2014) Linux Lab200slot2 3.14.0+ #1 SMP Mon Apr 14 12:06:40 EDT 2014 x86_64 Time: Mon, 14 Apr 2014 16:40:48 GMT Connecting to host 192.168.240.3, port 5201 Cookie: Lab200slot2.1397493648.413274.65e131 [ 4] local 192.168.240.2 port 50548 connected to 192.168.240.3 port 5201 Starting Test: protocol: SCTP, 1 streams, 1452 byte blocks, omitting 0 seconds, 60 second test [ ID] Interval Transfer Bandwidth [ 4] 0.00-1.00 sec 240 MBytes 2.02 Gbits/sec [ 4] 1.00-2.00 sec 239 MBytes 2.01 Gbits/sec [ 4] 2.00-3.00 sec 240 MBytes 2.01 Gbits/sec [ 4] 3.00-4.00 sec 239 MBytes 2.00 Gbits/sec [ 4] 4.00-5.00 sec 245 MBytes 2.05 Gbits/sec [ 4] 5.00-6.00 sec 240 MBytes 2.01 Gbits/sec [ 4] 6.00-7.00 sec 240 MBytes 2.02 Gbits/sec [ 4] 7.00-8.00 sec 239 MBytes 2.01 Gbits/sec With the reverted patch applied, the SCTP/IPv4 performance is back to normal on latest upstream for IPv4 and IPv6 and has same throughput as 3.4.2 test kernel, steady and interval reports are smooth again. Fixes: ef2820a735f7 ("net: sctp: Fix a_rwnd/rwnd management to reflect real state of the receiver's buffer") Reported-by: Peter Butler <pbutler@sonusnet.com> Reported-by: Dongsheng Song <dongsheng.song@gmail.com> Reported-by: Fengguang Wu <fengguang.wu@intel.com> Tested-by: Peter Butler <pbutler@sonusnet.com> Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Cc: Matija Glavinic Pecotic <matija.glavinic-pecotic.ext@nsn.com> Cc: Alexander Sverdlin <alexander.sverdlin@nsn.com> Cc: Vlad Yasevich <vyasevich@gmail.com> Acked-by: Vlad Yasevich <vyasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-04-14 21:45:17 +02:00
/* Decrease asoc's rwnd by len. */
void sctp_assoc_rwnd_decrease(struct sctp_association *asoc, unsigned int len)
{
int rx_count;
int over = 0;
if (unlikely(!asoc->rwnd || asoc->rwnd_over))
pr_debug("%s: association:%p has asoc->rwnd:%u, "
"asoc->rwnd_over:%u!\n", __func__, asoc,
asoc->rwnd, asoc->rwnd_over);
if (asoc->ep->rcvbuf_policy)
rx_count = atomic_read(&asoc->rmem_alloc);
else
rx_count = atomic_read(&asoc->base.sk->sk_rmem_alloc);
/* If we've reached or overflowed our receive buffer, announce
* a 0 rwnd if rwnd would still be positive. Store the
* the potential pressure overflow so that the window can be restored
* back to original value.
*/
if (rx_count >= asoc->base.sk->sk_rcvbuf)
over = 1;
if (asoc->rwnd >= len) {
asoc->rwnd -= len;
if (over) {
asoc->rwnd_press += asoc->rwnd;
asoc->rwnd = 0;
}
} else {
asoc->rwnd_over = len - asoc->rwnd;
asoc->rwnd = 0;
}
pr_debug("%s: asoc:%p rwnd decreased by %d to (%u, %u, %u)\n",
__func__, asoc, len, asoc->rwnd, asoc->rwnd_over,
asoc->rwnd_press);
}
/* Build the bind address list for the association based on info from the
* local endpoint and the remote peer.
*/
int sctp_assoc_set_bind_addr_from_ep(struct sctp_association *asoc,
sctp_scope_t scope, gfp_t gfp)
{
int flags;
/* Use scoping rules to determine the subset of addresses from
* the endpoint.
*/
flags = (PF_INET6 == asoc->base.sk->sk_family) ? SCTP_ADDR6_ALLOWED : 0;
if (asoc->peer.ipv4_address)
flags |= SCTP_ADDR4_PEERSUPP;
if (asoc->peer.ipv6_address)
flags |= SCTP_ADDR6_PEERSUPP;
return sctp_bind_addr_copy(sock_net(asoc->base.sk),
&asoc->base.bind_addr,
&asoc->ep->base.bind_addr,
scope, gfp, flags);
}
/* Build the association's bind address list from the cookie. */
int sctp_assoc_set_bind_addr_from_cookie(struct sctp_association *asoc,
struct sctp_cookie *cookie,
gfp_t gfp)
{
int var_size2 = ntohs(cookie->peer_init->chunk_hdr.length);
int var_size3 = cookie->raw_addr_list_len;
__u8 *raw = (__u8 *)cookie->peer_init + var_size2;
return sctp_raw_to_bind_addrs(&asoc->base.bind_addr, raw, var_size3,
asoc->ep->base.bind_addr.port, gfp);
}
/* Lookup laddr in the bind address list of an association. */
int sctp_assoc_lookup_laddr(struct sctp_association *asoc,
const union sctp_addr *laddr)
{
int found = 0;
if ((asoc->base.bind_addr.port == ntohs(laddr->v4.sin_port)) &&
sctp_bind_addr_match(&asoc->base.bind_addr, laddr,
sctp_sk(asoc->base.sk)))
found = 1;
return found;
}
/* Set an association id for a given association */
int sctp_assoc_set_id(struct sctp_association *asoc, gfp_t gfp)
{
mm, page_alloc: distinguish between being unable to sleep, unwilling to sleep and avoiding waking kswapd __GFP_WAIT has been used to identify atomic context in callers that hold spinlocks or are in interrupts. They are expected to be high priority and have access one of two watermarks lower than "min" which can be referred to as the "atomic reserve". __GFP_HIGH users get access to the first lower watermark and can be called the "high priority reserve". Over time, callers had a requirement to not block when fallback options were available. Some have abused __GFP_WAIT leading to a situation where an optimisitic allocation with a fallback option can access atomic reserves. This patch uses __GFP_ATOMIC to identify callers that are truely atomic, cannot sleep and have no alternative. High priority users continue to use __GFP_HIGH. __GFP_DIRECT_RECLAIM identifies callers that can sleep and are willing to enter direct reclaim. __GFP_KSWAPD_RECLAIM to identify callers that want to wake kswapd for background reclaim. __GFP_WAIT is redefined as a caller that is willing to enter direct reclaim and wake kswapd for background reclaim. This patch then converts a number of sites o __GFP_ATOMIC is used by callers that are high priority and have memory pools for those requests. GFP_ATOMIC uses this flag. o Callers that have a limited mempool to guarantee forward progress clear __GFP_DIRECT_RECLAIM but keep __GFP_KSWAPD_RECLAIM. bio allocations fall into this category where kswapd will still be woken but atomic reserves are not used as there is a one-entry mempool to guarantee progress. o Callers that are checking if they are non-blocking should use the helper gfpflags_allow_blocking() where possible. This is because checking for __GFP_WAIT as was done historically now can trigger false positives. Some exceptions like dm-crypt.c exist where the code intent is clearer if __GFP_DIRECT_RECLAIM is used instead of the helper due to flag manipulations. o Callers that built their own GFP flags instead of starting with GFP_KERNEL and friends now also need to specify __GFP_KSWAPD_RECLAIM. The first key hazard to watch out for is callers that removed __GFP_WAIT and was depending on access to atomic reserves for inconspicuous reasons. In some cases it may be appropriate for them to use __GFP_HIGH. The second key hazard is callers that assembled their own combination of GFP flags instead of starting with something like GFP_KERNEL. They may now wish to specify __GFP_KSWAPD_RECLAIM. It's almost certainly harmless if it's missed in most cases as other activity will wake kswapd. Signed-off-by: Mel Gorman <mgorman@techsingularity.net> Acked-by: Vlastimil Babka <vbabka@suse.cz> Acked-by: Michal Hocko <mhocko@suse.com> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Cc: Christoph Lameter <cl@linux.com> Cc: David Rientjes <rientjes@google.com> Cc: Vitaly Wool <vitalywool@gmail.com> Cc: Rik van Riel <riel@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-11-07 01:28:21 +01:00
bool preload = gfpflags_allow_blocking(gfp);
int ret;
/* If the id is already assigned, keep it. */
if (asoc->assoc_id)
return 0;
if (preload)
idr_preload(gfp);
spin_lock_bh(&sctp_assocs_id_lock);
/* 0 is not a valid assoc_id, must be >= 1 */
ret = idr_alloc_cyclic(&sctp_assocs_id, asoc, 1, 0, GFP_NOWAIT);
spin_unlock_bh(&sctp_assocs_id_lock);
if (preload)
idr_preload_end();
if (ret < 0)
return ret;
asoc->assoc_id = (sctp_assoc_t)ret;
return 0;
}
/* Free the ASCONF queue */
static void sctp_assoc_free_asconf_queue(struct sctp_association *asoc)
{
struct sctp_chunk *asconf;
struct sctp_chunk *tmp;
list_for_each_entry_safe(asconf, tmp, &asoc->addip_chunk_list, list) {
list_del_init(&asconf->list);
sctp_chunk_free(asconf);
}
}
/* Free asconf_ack cache */
static void sctp_assoc_free_asconf_acks(struct sctp_association *asoc)
{
struct sctp_chunk *ack;
struct sctp_chunk *tmp;
list_for_each_entry_safe(ack, tmp, &asoc->asconf_ack_list,
transmitted_list) {
list_del_init(&ack->transmitted_list);
sctp_chunk_free(ack);
}
}
/* Clean up the ASCONF_ACK queue */
void sctp_assoc_clean_asconf_ack_cache(const struct sctp_association *asoc)
{
struct sctp_chunk *ack;
struct sctp_chunk *tmp;
/* We can remove all the entries from the queue up to
* the "Peer-Sequence-Number".
*/
list_for_each_entry_safe(ack, tmp, &asoc->asconf_ack_list,
transmitted_list) {
if (ack->subh.addip_hdr->serial ==
htonl(asoc->peer.addip_serial))
break;
list_del_init(&ack->transmitted_list);
sctp_chunk_free(ack);
}
}
/* Find the ASCONF_ACK whose serial number matches ASCONF */
struct sctp_chunk *sctp_assoc_lookup_asconf_ack(
const struct sctp_association *asoc,
__be32 serial)
{
struct sctp_chunk *ack;
/* Walk through the list of cached ASCONF-ACKs and find the
* ack chunk whose serial number matches that of the request.
*/
list_for_each_entry(ack, &asoc->asconf_ack_list, transmitted_list) {
net: sctp: fix panic on duplicate ASCONF chunks When receiving a e.g. semi-good formed connection scan in the form of ... -------------- INIT[ASCONF; ASCONF_ACK] -------------> <----------- INIT-ACK[ASCONF; ASCONF_ACK] ------------ -------------------- COOKIE-ECHO --------------------> <-------------------- COOKIE-ACK --------------------- ---------------- ASCONF_a; ASCONF_b -----------------> ... where ASCONF_a equals ASCONF_b chunk (at least both serials need to be equal), we panic an SCTP server! The problem is that good-formed ASCONF chunks that we reply with ASCONF_ACK chunks are cached per serial. Thus, when we receive a same ASCONF chunk twice (e.g. through a lost ASCONF_ACK), we do not need to process them again on the server side (that was the idea, also proposed in the RFC). Instead, we know it was cached and we just resend the cached chunk instead. So far, so good. Where things get nasty is in SCTP's side effect interpreter, that is, sctp_cmd_interpreter(): While incoming ASCONF_a (chunk = event_arg) is being marked !end_of_packet and !singleton, and we have an association context, we do not flush the outqueue the first time after processing the ASCONF_ACK singleton chunk via SCTP_CMD_REPLY. Instead, we keep it queued up, although we set local_cork to 1. Commit 2e3216cd54b1 changed the precedence, so that as long as we get bundled, incoming chunks we try possible bundling on outgoing queue as well. Before this commit, we would just flush the output queue. Now, while ASCONF_a's ASCONF_ACK sits in the corked outq, we continue to process the same ASCONF_b chunk from the packet. As we have cached the previous ASCONF_ACK, we find it, grab it and do another SCTP_CMD_REPLY command on it. So, effectively, we rip the chunk->list pointers and requeue the same ASCONF_ACK chunk another time. Since we process ASCONF_b, it's correctly marked with end_of_packet and we enforce an uncork, and thus flush, thus crashing the kernel. Fix it by testing if the ASCONF_ACK is currently pending and if that is the case, do not requeue it. When flushing the output queue we may relink the chunk for preparing an outgoing packet, but eventually unlink it when it's copied into the skb right before transmission. Joint work with Vlad Yasevich. Fixes: 2e3216cd54b1 ("sctp: Follow security requirement of responding with 1 packet") Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Signed-off-by: Vlad Yasevich <vyasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-10-09 22:55:32 +02:00
if (sctp_chunk_pending(ack))
continue;
if (ack->subh.addip_hdr->serial == serial) {
sctp_chunk_hold(ack);
return ack;
}
}
return NULL;
}
void sctp_asconf_queue_teardown(struct sctp_association *asoc)
{
/* Free any cached ASCONF_ACK chunk. */
sctp_assoc_free_asconf_acks(asoc);
/* Free the ASCONF queue. */
sctp_assoc_free_asconf_queue(asoc);
/* Free any cached ASCONF chunk. */
if (asoc->addip_last_asconf)
sctp_chunk_free(asoc->addip_last_asconf);
}