linux/net/dccp/minisocks.c
Eric Dumazet 5fa411e855 tcp/dccp: block bh before arming time_wait timer
[ Upstream commit cfac7f836a ]

Maciej Żenczykowski reported some panics in tcp_twsk_destructor()
that might be caused by the following bug.

timewait timer is pinned to the cpu, because we want to transition
timwewait refcount from 0 to 4 in one go, once everything has been
initialized.

At the time commit ed2e923945 ("tcp/dccp: fix timewait races in timer
handling") was merged, TCP was always running from BH habdler.

After commit 5413d1babe ("net: do not block BH while processing
socket backlog") we definitely can run tcp_time_wait() from process
context.

We need to block BH in the critical section so that the pinned timer
has still its purpose.

This bug is more likely to happen under stress and when very small RTO
are used in datacenter flows.

Fixes: 5413d1babe ("net: do not block BH while processing socket backlog")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Maciej Żenczykowski <maze@google.com>
Acked-by: Maciej Żenczykowski <maze@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-12-17 15:07:57 +01:00

273 lines
7.5 KiB
C

/*
* net/dccp/minisocks.c
*
* An implementation of the DCCP protocol
* Arnaldo Carvalho de Melo <acme@conectiva.com.br>
*
* This program 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 of the License, or (at your option) any later version.
*/
#include <linux/dccp.h>
#include <linux/gfp.h>
#include <linux/kernel.h>
#include <linux/skbuff.h>
#include <linux/timer.h>
#include <net/sock.h>
#include <net/xfrm.h>
#include <net/inet_timewait_sock.h>
#include "ackvec.h"
#include "ccid.h"
#include "dccp.h"
#include "feat.h"
struct inet_timewait_death_row dccp_death_row = {
.sysctl_max_tw_buckets = NR_FILE * 2,
.hashinfo = &dccp_hashinfo,
};
EXPORT_SYMBOL_GPL(dccp_death_row);
void dccp_time_wait(struct sock *sk, int state, int timeo)
{
struct inet_timewait_sock *tw;
tw = inet_twsk_alloc(sk, &dccp_death_row, state);
if (tw != NULL) {
const struct inet_connection_sock *icsk = inet_csk(sk);
const int rto = (icsk->icsk_rto << 2) - (icsk->icsk_rto >> 1);
#if IS_ENABLED(CONFIG_IPV6)
if (tw->tw_family == PF_INET6) {
tw->tw_v6_daddr = sk->sk_v6_daddr;
tw->tw_v6_rcv_saddr = sk->sk_v6_rcv_saddr;
tw->tw_ipv6only = sk->sk_ipv6only;
}
#endif
/* Get the TIME_WAIT timeout firing. */
if (timeo < rto)
timeo = rto;
tw->tw_timeout = DCCP_TIMEWAIT_LEN;
if (state == DCCP_TIME_WAIT)
timeo = DCCP_TIMEWAIT_LEN;
/* tw_timer is pinned, so we need to make sure BH are disabled
* in following section, otherwise timer handler could run before
* we complete the initialization.
*/
local_bh_disable();
inet_twsk_schedule(tw, timeo);
/* Linkage updates. */
__inet_twsk_hashdance(tw, sk, &dccp_hashinfo);
inet_twsk_put(tw);
local_bh_enable();
} else {
/* Sorry, if we're out of memory, just CLOSE this
* socket up. We've got bigger problems than
* non-graceful socket closings.
*/
DCCP_WARN("time wait bucket table overflow\n");
}
dccp_done(sk);
}
struct sock *dccp_create_openreq_child(const struct sock *sk,
const struct request_sock *req,
const struct sk_buff *skb)
{
/*
* Step 3: Process LISTEN state
*
* (* Generate a new socket and switch to that socket *)
* Set S := new socket for this port pair
*/
struct sock *newsk = inet_csk_clone_lock(sk, req, GFP_ATOMIC);
if (newsk != NULL) {
struct dccp_request_sock *dreq = dccp_rsk(req);
struct inet_connection_sock *newicsk = inet_csk(newsk);
struct dccp_sock *newdp = dccp_sk(newsk);
newdp->dccps_role = DCCP_ROLE_SERVER;
newdp->dccps_hc_rx_ackvec = NULL;
newdp->dccps_service_list = NULL;
newdp->dccps_service = dreq->dreq_service;
newdp->dccps_timestamp_echo = dreq->dreq_timestamp_echo;
newdp->dccps_timestamp_time = dreq->dreq_timestamp_time;
newicsk->icsk_rto = DCCP_TIMEOUT_INIT;
INIT_LIST_HEAD(&newdp->dccps_featneg);
/*
* Step 3: Process LISTEN state
*
* Choose S.ISS (initial seqno) or set from Init Cookies
* Initialize S.GAR := S.ISS
* Set S.ISR, S.GSR from packet (or Init Cookies)
*
* Setting AWL/AWH and SWL/SWH happens as part of the feature
* activation below, as these windows all depend on the local
* and remote Sequence Window feature values (7.5.2).
*/
newdp->dccps_iss = dreq->dreq_iss;
newdp->dccps_gss = dreq->dreq_gss;
newdp->dccps_gar = newdp->dccps_iss;
newdp->dccps_isr = dreq->dreq_isr;
newdp->dccps_gsr = dreq->dreq_gsr;
/*
* Activate features: initialise CCIDs, sequence windows etc.
*/
if (dccp_feat_activate_values(newsk, &dreq->dreq_featneg)) {
sk_free_unlock_clone(newsk);
return NULL;
}
dccp_init_xmit_timers(newsk);
__DCCP_INC_STATS(DCCP_MIB_PASSIVEOPENS);
}
return newsk;
}
EXPORT_SYMBOL_GPL(dccp_create_openreq_child);
/*
* Process an incoming packet for RESPOND sockets represented
* as an request_sock.
*/
struct sock *dccp_check_req(struct sock *sk, struct sk_buff *skb,
struct request_sock *req)
{
struct sock *child = NULL;
struct dccp_request_sock *dreq = dccp_rsk(req);
bool own_req;
/* TCP/DCCP listeners became lockless.
* DCCP stores complex state in its request_sock, so we need
* a protection for them, now this code runs without being protected
* by the parent (listener) lock.
*/
spin_lock_bh(&dreq->dreq_lock);
/* Check for retransmitted REQUEST */
if (dccp_hdr(skb)->dccph_type == DCCP_PKT_REQUEST) {
if (after48(DCCP_SKB_CB(skb)->dccpd_seq, dreq->dreq_gsr)) {
dccp_pr_debug("Retransmitted REQUEST\n");
dreq->dreq_gsr = DCCP_SKB_CB(skb)->dccpd_seq;
/*
* Send another RESPONSE packet
* To protect against Request floods, increment retrans
* counter (backoff, monitored by dccp_response_timer).
*/
inet_rtx_syn_ack(sk, req);
}
/* Network Duplicate, discard packet */
goto out;
}
DCCP_SKB_CB(skb)->dccpd_reset_code = DCCP_RESET_CODE_PACKET_ERROR;
if (dccp_hdr(skb)->dccph_type != DCCP_PKT_ACK &&
dccp_hdr(skb)->dccph_type != DCCP_PKT_DATAACK)
goto drop;
/* Invalid ACK */
if (!between48(DCCP_SKB_CB(skb)->dccpd_ack_seq,
dreq->dreq_iss, dreq->dreq_gss)) {
dccp_pr_debug("Invalid ACK number: ack_seq=%llu, "
"dreq_iss=%llu, dreq_gss=%llu\n",
(unsigned long long)
DCCP_SKB_CB(skb)->dccpd_ack_seq,
(unsigned long long) dreq->dreq_iss,
(unsigned long long) dreq->dreq_gss);
goto drop;
}
if (dccp_parse_options(sk, dreq, skb))
goto drop;
child = inet_csk(sk)->icsk_af_ops->syn_recv_sock(sk, skb, req, NULL,
req, &own_req);
if (child) {
child = inet_csk_complete_hashdance(sk, child, req, own_req);
goto out;
}
DCCP_SKB_CB(skb)->dccpd_reset_code = DCCP_RESET_CODE_TOO_BUSY;
drop:
if (dccp_hdr(skb)->dccph_type != DCCP_PKT_RESET)
req->rsk_ops->send_reset(sk, skb);
inet_csk_reqsk_queue_drop(sk, req);
out:
spin_unlock_bh(&dreq->dreq_lock);
return child;
}
EXPORT_SYMBOL_GPL(dccp_check_req);
/*
* Queue segment on the new socket if the new socket is active,
* otherwise we just shortcircuit this and continue with
* the new socket.
*/
int dccp_child_process(struct sock *parent, struct sock *child,
struct sk_buff *skb)
{
int ret = 0;
const int state = child->sk_state;
if (!sock_owned_by_user(child)) {
ret = dccp_rcv_state_process(child, skb, dccp_hdr(skb),
skb->len);
/* Wakeup parent, send SIGIO */
if (state == DCCP_RESPOND && child->sk_state != state)
parent->sk_data_ready(parent);
} else {
/* Alas, it is possible again, because we do lookup
* in main socket hash table and lock on listening
* socket does not protect us more.
*/
__sk_add_backlog(child, skb);
}
bh_unlock_sock(child);
sock_put(child);
return ret;
}
EXPORT_SYMBOL_GPL(dccp_child_process);
void dccp_reqsk_send_ack(const struct sock *sk, struct sk_buff *skb,
struct request_sock *rsk)
{
DCCP_BUG("DCCP-ACK packets are never sent in LISTEN/RESPOND state");
}
EXPORT_SYMBOL_GPL(dccp_reqsk_send_ack);
int dccp_reqsk_init(struct request_sock *req,
struct dccp_sock const *dp, struct sk_buff const *skb)
{
struct dccp_request_sock *dreq = dccp_rsk(req);
spin_lock_init(&dreq->dreq_lock);
inet_rsk(req)->ir_rmt_port = dccp_hdr(skb)->dccph_sport;
inet_rsk(req)->ir_num = ntohs(dccp_hdr(skb)->dccph_dport);
inet_rsk(req)->acked = 0;
dreq->dreq_timestamp_echo = 0;
/* inherit feature negotiation options from listening socket */
return dccp_feat_clone_list(&dp->dccps_featneg, &dreq->dreq_featneg);
}
EXPORT_SYMBOL_GPL(dccp_reqsk_init);