leandrof/linux
1
0
Fork 0
forked from Minki/linux
Code Issues Pull Requests Packages Projects Releases Wiki Activity
cf61f5b015
linux/net/llc/llc_conn.c

1051 lines
28 KiB
C
Raw Normal View History

Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/*
* llc_conn.c - Driver routines for connection component.
*
* Copyright (c) 1997 by Procom Technology, Inc.
* 2001-2003 by Arnaldo Carvalho de Melo <acme@conectiva.com.br>
*
* This program can be redistributed or modified under the terms of the
* GNU General Public License as published by the Free Software Foundation.
* This program is distributed without any warranty or implied warranty
* of merchantability or fitness for a particular purpose.
*
* See the GNU General Public License for more details.
*/
#include <linux/init.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 08:04:11 +00:00
#include <linux/slab.h>
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#include <net/llc_sap.h>
#include <net/llc_conn.h>
#include <net/sock.h>
[TCP]: Move the tcp sock states to net/tcp_states.h Lots of places just needs the states, not even linux/tcp.h, where this enum was, needs it. This speeds up development of the refactorings as less sources are rebuilt when things get moved from net/tcp.h. Signed-off-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2005-08-10 03:08:28 +00:00
#include <net/tcp_states.h>
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#include <net/llc_c_ev.h>
#include <net/llc_c_ac.h>
#include <net/llc_c_st.h>
#include <net/llc_pdu.h>
#if 0
#define dprintk(args...) printk(KERN_DEBUG args)
#else
#define dprintk(args...)
#endif
static int llc_find_offset(int state, int ev_type);
llc: properly handle dev_queue_xmit() return value llc_conn_send_pdu() pushes the skb into write queue and calls llc_conn_send_pdus() to flush them out. However, the status of dev_queue_xmit() is not returned to caller, in this case, llc_conn_state_process(). llc_conn_state_process() needs hold the skb no matter success or failure, because it still uses it after that, therefore we should hold skb before dev_queue_xmit() when that skb is the one being processed by llc_conn_state_process(). For other callers, they can just pass NULL and ignore the return value as they are. Reported-by: Noam Rathaus <noamr@beyondsecurity.com> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-26 22:08:33 +00:00
static int llc_conn_send_pdus(struct sock *sk, struct sk_buff *skb);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
static int llc_conn_service(struct sock *sk, struct sk_buff *skb);
static int llc_exec_conn_trans_actions(struct sock *sk,
struct llc_conn_state_trans *trans,
struct sk_buff *ev);
static struct llc_conn_state_trans *llc_qualify_conn_ev(struct sock *sk,
struct sk_buff *skb);
/* Offset table on connection states transition diagram */
static int llc_offset_table[NBR_CONN_STATES][NBR_CONN_EV];
[LLC]: Add sysctl support for the LLC timeouts Signed-off-by: Jochen Friedrich <jochen@scram.de> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 07:30:44 +00:00
int sysctl_llc2_ack_timeout = LLC2_ACK_TIME * HZ;
int sysctl_llc2_p_timeout = LLC2_P_TIME * HZ;
int sysctl_llc2_rej_timeout = LLC2_REJ_TIME * HZ;
int sysctl_llc2_busy_timeout = LLC2_BUSY_TIME * HZ;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/**
* llc_conn_state_process - sends event to connection state machine
* @sk: connection
* @skb: occurred event
*
* Sends an event to connection state machine. After processing event
* (executing it's actions and changing state), upper layer will be
* indicated or confirmed, if needed. Returns 0 for success, 1 for
* failure. The socket lock has to be held before calling this function.
*/
int llc_conn_state_process(struct sock *sk, struct sk_buff *skb)
{
int rc;
[LLC]: Fix the accept path Borrowing the structure of TCP/IP for this. On the receive of new connections I was bh_lock_socking the _new_ sock, not the listening one, duh, now it survives the ssh connections storm I've been using to test this specific bug. Also fixes send side skb sock accounting. Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 10:57:21 +00:00
struct llc_sock *llc = llc_sk(skb->sk);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct llc_conn_state_ev *ev = llc_conn_ev(skb);
/*
* We have to hold the skb, because llc_conn_service will kfree it in
* the sending path and we need to look at the skb->cb, where we encode
* llc_conn_state_ev.
*/
skb_get(skb);
ev->ind_prim = ev->cfm_prim = 0;
[LLC]: Fix the accept path Borrowing the structure of TCP/IP for this. On the receive of new connections I was bh_lock_socking the _new_ sock, not the listening one, duh, now it survives the ssh connections storm I've been using to test this specific bug. Also fixes send side skb sock accounting. Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 10:57:21 +00:00
/*
* Send event to state machine
*/
rc = llc_conn_service(skb->sk, skb);
[LLC]: Help the compiler with likely/unlikely, saving some more bytes Signed-off-by: Jochen Friedrich <jochen@scram.de> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 06:59:22 +00:00
if (unlikely(rc != 0)) {
net: replace remaining __FUNCTION__ occurrences __FUNCTION__ is gcc-specific, use __func__ Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-06 04:47:47 +00:00
printk(KERN_ERR "%s: llc_conn_service failed\n", __func__);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto out_kfree_skb;
}
[LLC]: Help the compiler with likely/unlikely, saving some more bytes Signed-off-by: Jochen Friedrich <jochen@scram.de> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 06:59:22 +00:00
if (unlikely(!ev->ind_prim && !ev->cfm_prim)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* indicate or confirm not required */
[NET]: Kill skb->list Remove the "list" member of struct sk_buff, as it is entirely redundant. All SKB list removal callers know which list the SKB is on, so storing this in sk_buff does nothing other than taking up some space. Two tricky bits were SCTP, which I took care of, and two ATM drivers which Francois Romieu <romieu@fr.zoreil.com> fixed up. Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Francois Romieu <romieu@fr.zoreil.com>
2005-08-10 02:25:21 +00:00
if (!skb->next)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto out_kfree_skb;
goto out_skb_put;
}
[LLC]: Help the compiler with likely/unlikely, saving some more bytes Signed-off-by: Jochen Friedrich <jochen@scram.de> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 06:59:22 +00:00
if (unlikely(ev->ind_prim && ev->cfm_prim)) /* Paranoia */
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
skb_get(skb);
switch (ev->ind_prim) {
case LLC_DATA_PRIM:
[LLC]: Do better struct sock accounting on skbs Signed-off-by: Jochen Friedrich <jochen@scram.de> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 07:40:59 +00:00
llc_save_primitive(sk, skb, LLC_DATA_PRIM);
if (unlikely(sock_queue_rcv_skb(sk, skb))) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/*
* shouldn't happen
*/
printk(KERN_ERR "%s: sock_queue_rcv_skb failed!\n",
net: replace remaining __FUNCTION__ occurrences __FUNCTION__ is gcc-specific, use __func__ Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-06 04:47:47 +00:00
__func__);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
kfree_skb(skb);
}
break;
[LLC]: Fix the accept path Borrowing the structure of TCP/IP for this. On the receive of new connections I was bh_lock_socking the _new_ sock, not the listening one, duh, now it survives the ssh connections storm I've been using to test this specific bug. Also fixes send side skb sock accounting. Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 10:57:21 +00:00
case LLC_CONN_PRIM:
[LLC]: Do better struct sock accounting on skbs Signed-off-by: Jochen Friedrich <jochen@scram.de> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 07:40:59 +00:00
/*
[LLC]: Fix the accept path Borrowing the structure of TCP/IP for this. On the receive of new connections I was bh_lock_socking the _new_ sock, not the listening one, duh, now it survives the ssh connections storm I've been using to test this specific bug. Also fixes send side skb sock accounting. Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 10:57:21 +00:00
* Can't be sock_queue_rcv_skb, because we have to leave the
* skb->sk pointing to the newly created struct sock in
* llc_conn_handler. -acme
[LLC]: Do better struct sock accounting on skbs Signed-off-by: Jochen Friedrich <jochen@scram.de> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 07:40:59 +00:00
*/
[LLC]: Fix the accept path Borrowing the structure of TCP/IP for this. On the receive of new connections I was bh_lock_socking the _new_ sock, not the listening one, duh, now it survives the ssh connections storm I've been using to test this specific bug. Also fixes send side skb sock accounting. Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 10:57:21 +00:00
skb_queue_tail(&sk->sk_receive_queue, skb);
sk->sk_state_change(sk);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
break;
case LLC_DISC_PRIM:
sock_hold(sk);
if (sk->sk_type == SOCK_STREAM &&
sk->sk_state == TCP_ESTABLISHED) {
sk->sk_shutdown = SHUTDOWN_MASK;
sk->sk_socket->state = SS_UNCONNECTED;
sk->sk_state = TCP_CLOSE;
if (!sock_flag(sk, SOCK_DEAD)) {
sock_set_flag(sk, SOCK_DEAD);
[LLC]: fix llc_ui_recvmsg, making it behave like tcp_recvmsg In fact it is an exact copy of the parts that makes sense to LLC :-) Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 11:29:08 +00:00
sk->sk_state_change(sk);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
}
kfree_skb(skb);
sock_put(sk);
break;
case LLC_RESET_PRIM:
/*
* FIXME:
* RESET is not being notified to upper layers for now
*/
net: replace remaining __FUNCTION__ occurrences __FUNCTION__ is gcc-specific, use __func__ Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-06 04:47:47 +00:00
printk(KERN_INFO "%s: received a reset ind!\n", __func__);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
kfree_skb(skb);
break;
default:
if (ev->ind_prim) {
printk(KERN_INFO "%s: received unknown %d prim!\n",
net: replace remaining __FUNCTION__ occurrences __FUNCTION__ is gcc-specific, use __func__ Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-06 04:47:47 +00:00
__func__, ev->ind_prim);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
kfree_skb(skb);
}
/* No indication */
break;
}
switch (ev->cfm_prim) {
case LLC_DATA_PRIM:
if (!llc_data_accept_state(llc->state))
sk->sk_write_space(sk);
else
rc = llc->failed_data_req = 1;
break;
case LLC_CONN_PRIM:
if (sk->sk_type == SOCK_STREAM &&
sk->sk_state == TCP_SYN_SENT) {
if (ev->status) {
sk->sk_socket->state = SS_UNCONNECTED;
sk->sk_state = TCP_CLOSE;
} else {
sk->sk_socket->state = SS_CONNECTED;
sk->sk_state = TCP_ESTABLISHED;
}
sk->sk_state_change(sk);
}
break;
case LLC_DISC_PRIM:
sock_hold(sk);
if (sk->sk_type == SOCK_STREAM && sk->sk_state == TCP_CLOSING) {
sk->sk_socket->state = SS_UNCONNECTED;
sk->sk_state = TCP_CLOSE;
sk->sk_state_change(sk);
}
sock_put(sk);
break;
case LLC_RESET_PRIM:
/*
* FIXME:
* RESET is not being notified to upper layers for now
*/
net: replace remaining __FUNCTION__ occurrences __FUNCTION__ is gcc-specific, use __func__ Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-06 04:47:47 +00:00
printk(KERN_INFO "%s: received a reset conf!\n", __func__);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
break;
default:
if (ev->cfm_prim) {
printk(KERN_INFO "%s: received unknown %d prim!\n",
net: replace remaining __FUNCTION__ occurrences __FUNCTION__ is gcc-specific, use __func__ Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-06 04:47:47 +00:00
__func__, ev->cfm_prim);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
break;
}
goto out_skb_put; /* No confirmation */
}
out_kfree_skb:
kfree_skb(skb);
out_skb_put:
kfree_skb(skb);
return rc;
}
llc: properly handle dev_queue_xmit() return value llc_conn_send_pdu() pushes the skb into write queue and calls llc_conn_send_pdus() to flush them out. However, the status of dev_queue_xmit() is not returned to caller, in this case, llc_conn_state_process(). llc_conn_state_process() needs hold the skb no matter success or failure, because it still uses it after that, therefore we should hold skb before dev_queue_xmit() when that skb is the one being processed by llc_conn_state_process(). For other callers, they can just pass NULL and ignore the return value as they are. Reported-by: Noam Rathaus <noamr@beyondsecurity.com> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-26 22:08:33 +00:00
int llc_conn_send_pdu(struct sock *sk, struct sk_buff *skb)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
/* queue PDU to send to MAC layer */
skb_queue_tail(&sk->sk_write_queue, skb);
llc: properly handle dev_queue_xmit() return value llc_conn_send_pdu() pushes the skb into write queue and calls llc_conn_send_pdus() to flush them out. However, the status of dev_queue_xmit() is not returned to caller, in this case, llc_conn_state_process(). llc_conn_state_process() needs hold the skb no matter success or failure, because it still uses it after that, therefore we should hold skb before dev_queue_xmit() when that skb is the one being processed by llc_conn_state_process(). For other callers, they can just pass NULL and ignore the return value as they are. Reported-by: Noam Rathaus <noamr@beyondsecurity.com> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-26 22:08:33 +00:00
return llc_conn_send_pdus(sk, skb);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/**
* llc_conn_rtn_pdu - sends received data pdu to upper layer
* @sk: Active connection
* @skb: Received data frame
*
* Sends received data pdu to upper layer (by using indicate function).
* Prepares service parameters (prim and prim_data). calling indication
* function will be done in llc_conn_state_process.
*/
void llc_conn_rtn_pdu(struct sock *sk, struct sk_buff *skb)
{
struct llc_conn_state_ev *ev = llc_conn_ev(skb);
ev->ind_prim = LLC_DATA_PRIM;
}
/**
* llc_conn_resend_i_pdu_as_cmd - resend all all unacknowledged I PDUs
* @sk: active connection
* @nr: NR
* @first_p_bit: p_bit value of first pdu
*
* Resend all unacknowledged I PDUs, starting with the NR; send first as
* command PDU with P bit equal first_p_bit; if more than one send
* subsequent as command PDUs with P bit equal zero (0).
*/
void llc_conn_resend_i_pdu_as_cmd(struct sock *sk, u8 nr, u8 first_p_bit)
{
struct sk_buff *skb;
struct llc_pdu_sn *pdu;
u16 nbr_unack_pdus;
struct llc_sock *llc;
u8 howmany_resend = 0;
llc_conn_remove_acked_pdus(sk, nr, &nbr_unack_pdus);
if (!nbr_unack_pdus)
goto out;
/*
* Process unack PDUs only if unack queue is not empty; remove
* appropriate PDUs, fix them up, and put them on mac_pdu_q.
*/
llc = llc_sk(sk);
while ((skb = skb_dequeue(&llc->pdu_unack_q)) != NULL) {
pdu = llc_pdu_sn_hdr(skb);
llc_pdu_set_cmd_rsp(skb, LLC_PDU_CMD);
llc_pdu_set_pf_bit(skb, first_p_bit);
skb_queue_tail(&sk->sk_write_queue, skb);
first_p_bit = 0;
llc->vS = LLC_I_GET_NS(pdu);
howmany_resend++;
}
if (howmany_resend > 0)
llc->vS = (llc->vS + 1) % LLC_2_SEQ_NBR_MODULO;
/* any PDUs to re-send are queued up; start sending to MAC */
llc: properly handle dev_queue_xmit() return value llc_conn_send_pdu() pushes the skb into write queue and calls llc_conn_send_pdus() to flush them out. However, the status of dev_queue_xmit() is not returned to caller, in this case, llc_conn_state_process(). llc_conn_state_process() needs hold the skb no matter success or failure, because it still uses it after that, therefore we should hold skb before dev_queue_xmit() when that skb is the one being processed by llc_conn_state_process(). For other callers, they can just pass NULL and ignore the return value as they are. Reported-by: Noam Rathaus <noamr@beyondsecurity.com> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-26 22:08:33 +00:00
llc_conn_send_pdus(sk, NULL);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
out:;
}
/**
* llc_conn_resend_i_pdu_as_rsp - Resend all unacknowledged I PDUs
* @sk: active connection.
* @nr: NR
* @first_f_bit: f_bit value of first pdu.
*
* Resend all unacknowledged I PDUs, starting with the NR; send first as
* response PDU with F bit equal first_f_bit; if more than one send
* subsequent as response PDUs with F bit equal zero (0).
*/
void llc_conn_resend_i_pdu_as_rsp(struct sock *sk, u8 nr, u8 first_f_bit)
{
struct sk_buff *skb;
u16 nbr_unack_pdus;
struct llc_sock *llc = llc_sk(sk);
u8 howmany_resend = 0;
llc_conn_remove_acked_pdus(sk, nr, &nbr_unack_pdus);
if (!nbr_unack_pdus)
goto out;
/*
* Process unack PDUs only if unack queue is not empty; remove
* appropriate PDUs, fix them up, and put them on mac_pdu_q
*/
while ((skb = skb_dequeue(&llc->pdu_unack_q)) != NULL) {
struct llc_pdu_sn *pdu = llc_pdu_sn_hdr(skb);
llc_pdu_set_cmd_rsp(skb, LLC_PDU_RSP);
llc_pdu_set_pf_bit(skb, first_f_bit);
skb_queue_tail(&sk->sk_write_queue, skb);
first_f_bit = 0;
llc->vS = LLC_I_GET_NS(pdu);
howmany_resend++;
}
if (howmany_resend > 0)
llc->vS = (llc->vS + 1) % LLC_2_SEQ_NBR_MODULO;
/* any PDUs to re-send are queued up; start sending to MAC */
llc: properly handle dev_queue_xmit() return value llc_conn_send_pdu() pushes the skb into write queue and calls llc_conn_send_pdus() to flush them out. However, the status of dev_queue_xmit() is not returned to caller, in this case, llc_conn_state_process(). llc_conn_state_process() needs hold the skb no matter success or failure, because it still uses it after that, therefore we should hold skb before dev_queue_xmit() when that skb is the one being processed by llc_conn_state_process(). For other callers, they can just pass NULL and ignore the return value as they are. Reported-by: Noam Rathaus <noamr@beyondsecurity.com> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-26 22:08:33 +00:00
llc_conn_send_pdus(sk, NULL);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
out:;
}
/**
* llc_conn_remove_acked_pdus - Removes acknowledged pdus from tx queue
* @sk: active connection
* nr: NR
* how_many_unacked: size of pdu_unack_q after removing acked pdus
*
* Removes acknowledged pdus from transmit queue (pdu_unack_q). Returns
* the number of pdus that removed from queue.
*/
int llc_conn_remove_acked_pdus(struct sock *sk, u8 nr, u16 *how_many_unacked)
{
int pdu_pos, i;
struct sk_buff *skb;
struct llc_pdu_sn *pdu;
int nbr_acked = 0;
struct llc_sock *llc = llc_sk(sk);
int q_len = skb_queue_len(&llc->pdu_unack_q);
if (!q_len)
goto out;
skb = skb_peek(&llc->pdu_unack_q);
pdu = llc_pdu_sn_hdr(skb);
/* finding position of last acked pdu in queue */
pdu_pos = ((int)LLC_2_SEQ_NBR_MODULO + (int)nr -
(int)LLC_I_GET_NS(pdu)) % LLC_2_SEQ_NBR_MODULO;
for (i = 0; i < pdu_pos && i < q_len; i++) {
skb = skb_dequeue(&llc->pdu_unack_q);
llc: remove some pointless conditionals before kfree_skb() Remove some pointless conditionals before kfree_skb(). Signed-off-by: Wei Yongjun <yjwei@cn.fujitsu.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-02-25 00:42:22 +00:00
kfree_skb(skb);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
nbr_acked++;
}
out:
*how_many_unacked = skb_queue_len(&llc->pdu_unack_q);
return nbr_acked;
}
/**
* llc_conn_send_pdus - Sends queued PDUs
* @sk: active connection
llc: properly handle dev_queue_xmit() return value llc_conn_send_pdu() pushes the skb into write queue and calls llc_conn_send_pdus() to flush them out. However, the status of dev_queue_xmit() is not returned to caller, in this case, llc_conn_state_process(). llc_conn_state_process() needs hold the skb no matter success or failure, because it still uses it after that, therefore we should hold skb before dev_queue_xmit() when that skb is the one being processed by llc_conn_state_process(). For other callers, they can just pass NULL and ignore the return value as they are. Reported-by: Noam Rathaus <noamr@beyondsecurity.com> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-26 22:08:33 +00:00
* @hold_skb: the skb held by caller, or NULL if does not care
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
*
llc: properly handle dev_queue_xmit() return value llc_conn_send_pdu() pushes the skb into write queue and calls llc_conn_send_pdus() to flush them out. However, the status of dev_queue_xmit() is not returned to caller, in this case, llc_conn_state_process(). llc_conn_state_process() needs hold the skb no matter success or failure, because it still uses it after that, therefore we should hold skb before dev_queue_xmit() when that skb is the one being processed by llc_conn_state_process(). For other callers, they can just pass NULL and ignore the return value as they are. Reported-by: Noam Rathaus <noamr@beyondsecurity.com> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-26 22:08:33 +00:00
* Sends queued pdus to MAC layer for transmission. When @hold_skb is
* NULL, always return 0. Otherwise, return 0 if @hold_skb is sent
* successfully, or 1 for failure.
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
*/
llc: properly handle dev_queue_xmit() return value llc_conn_send_pdu() pushes the skb into write queue and calls llc_conn_send_pdus() to flush them out. However, the status of dev_queue_xmit() is not returned to caller, in this case, llc_conn_state_process(). llc_conn_state_process() needs hold the skb no matter success or failure, because it still uses it after that, therefore we should hold skb before dev_queue_xmit() when that skb is the one being processed by llc_conn_state_process(). For other callers, they can just pass NULL and ignore the return value as they are. Reported-by: Noam Rathaus <noamr@beyondsecurity.com> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-26 22:08:33 +00:00
static int llc_conn_send_pdus(struct sock *sk, struct sk_buff *hold_skb)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct sk_buff *skb;
llc: properly handle dev_queue_xmit() return value llc_conn_send_pdu() pushes the skb into write queue and calls llc_conn_send_pdus() to flush them out. However, the status of dev_queue_xmit() is not returned to caller, in this case, llc_conn_state_process(). llc_conn_state_process() needs hold the skb no matter success or failure, because it still uses it after that, therefore we should hold skb before dev_queue_xmit() when that skb is the one being processed by llc_conn_state_process(). For other callers, they can just pass NULL and ignore the return value as they are. Reported-by: Noam Rathaus <noamr@beyondsecurity.com> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-26 22:08:33 +00:00
int ret = 0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
while ((skb = skb_dequeue(&sk->sk_write_queue)) != NULL) {
struct llc_pdu_sn *pdu = llc_pdu_sn_hdr(skb);
if (LLC_PDU_TYPE_IS_I(pdu) &&
!(skb->dev->flags & IFF_LOOPBACK)) {
struct sk_buff *skb2 = skb_clone(skb, GFP_ATOMIC);
skb_queue_tail(&llc_sk(sk)->pdu_unack_q, skb);
if (!skb2)
break;
llc: properly handle dev_queue_xmit() return value llc_conn_send_pdu() pushes the skb into write queue and calls llc_conn_send_pdus() to flush them out. However, the status of dev_queue_xmit() is not returned to caller, in this case, llc_conn_state_process(). llc_conn_state_process() needs hold the skb no matter success or failure, because it still uses it after that, therefore we should hold skb before dev_queue_xmit() when that skb is the one being processed by llc_conn_state_process(). For other callers, they can just pass NULL and ignore the return value as they are. Reported-by: Noam Rathaus <noamr@beyondsecurity.com> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-26 22:08:33 +00:00
dev_queue_xmit(skb2);
} else {
bool is_target = skb == hold_skb;
int rc;
if (is_target)
skb_get(skb);
rc = dev_queue_xmit(skb);
if (is_target)
ret = rc;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
}
llc: properly handle dev_queue_xmit() return value llc_conn_send_pdu() pushes the skb into write queue and calls llc_conn_send_pdus() to flush them out. However, the status of dev_queue_xmit() is not returned to caller, in this case, llc_conn_state_process(). llc_conn_state_process() needs hold the skb no matter success or failure, because it still uses it after that, therefore we should hold skb before dev_queue_xmit() when that skb is the one being processed by llc_conn_state_process(). For other callers, they can just pass NULL and ignore the return value as they are. Reported-by: Noam Rathaus <noamr@beyondsecurity.com> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-03-26 22:08:33 +00:00
return ret;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/**
* llc_conn_service - finds transition and changes state of connection
* @sk: connection
* @skb: happened event
*
* This function finds transition that matches with happened event, then
* executes related actions and finally changes state of connection.
* Returns 0 for success, 1 for failure.
*/
static int llc_conn_service(struct sock *sk, struct sk_buff *skb)
{
int rc = 1;
struct llc_sock *llc = llc_sk(sk);
struct llc_conn_state_trans *trans;
if (llc->state > NBR_CONN_STATES)
goto out;
rc = 0;
trans = llc_qualify_conn_ev(sk, skb);
if (trans) {
rc = llc_exec_conn_trans_actions(sk, trans, skb);
if (!rc && trans->next_state != NO_STATE_CHANGE) {
llc->state = trans->next_state;
if (!llc_data_accept_state(llc->state))
sk->sk_state_change(sk);
}
}
out:
return rc;
}
/**
* llc_qualify_conn_ev - finds transition for event
* @sk: connection
* @skb: happened event
*
* This function finds transition that matches with happened event.
* Returns pointer to found transition on success, %NULL otherwise.
*/
static struct llc_conn_state_trans *llc_qualify_conn_ev(struct sock *sk,
struct sk_buff *skb)
{
struct llc_conn_state_trans **next_trans;
llc: Make llc_conn_ev_qfyr_t function pointer arrays const It's better when function pointer arrays aren't modifiable. Net change from original: $ size net/llc/built-in.o.* text data bss dec hex filename 61065 12886 1344 75295 1261f net/llc/built-in.o.new 47113 27030 1344 75487 126df net/llc/built-in.o.old Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-12-10 17:43:57 +00:00
const llc_conn_ev_qfyr_t *next_qualifier;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
struct llc_conn_state_ev *ev = llc_conn_ev(skb);
struct llc_sock *llc = llc_sk(sk);
struct llc_conn_state *curr_state =
&llc_conn_state_table[llc->state - 1];
/* search thru events for this state until
* list exhausted or until no more
*/
for (next_trans = curr_state->transitions +
llc_find_offset(llc->state - 1, ev->type);
(*next_trans)->ev; next_trans++) {
if (!((*next_trans)->ev)(sk, skb)) {
/* got POSSIBLE event match; the event may require
* qualification based on the values of a number of
* state flags; if all qualifications are met (i.e.,
* if all qualifying functions return success, or 0,
* then this is THE event we're looking for
*/
for (next_qualifier = (*next_trans)->ev_qualifiers;
next_qualifier && *next_qualifier &&
!(*next_qualifier)(sk, skb); next_qualifier++)
/* nothing */;
if (!next_qualifier || !*next_qualifier)
/* all qualifiers executed successfully; this is
* our transition; return it so we can perform
* the associated actions & change the state
*/
return *next_trans;
}
}
return NULL;
}
/**
* llc_exec_conn_trans_actions - executes related actions
* @sk: connection
* @trans: transition that it's actions must be performed
* @skb: event
*
* Executes actions that is related to happened event. Returns 0 for
* success, 1 to indicate failure of at least one action.
*/
static int llc_exec_conn_trans_actions(struct sock *sk,
struct llc_conn_state_trans *trans,
struct sk_buff *skb)
{
int rc = 0;
llc: Make function pointer arrays const It's better when function pointer arrays aren't modifiable. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-12-10 16:18:50 +00:00
const llc_conn_action_t *next_action;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
for (next_action = trans->ev_actions;
next_action && *next_action; next_action++) {
int rc2 = (*next_action)(sk, skb);
if (rc2 == 2) {
rc = rc2;
break;
} else if (rc2)
rc = 1;
}
return rc;
}
llc: convert the socket list to RCU locking For the reclamation phase we use the SLAB_DESTROY_BY_RCU mechanism, which require some extra checks in the lookup code: a) If the current socket was released, reallocated & inserted in another list it will short circuit the iteration for the current list, thus we need to restart the lookup. b) If the current socket was released, reallocated & inserted in the same list we just need to recheck it matches the look-up criteria and if not we can skip to the next element. In this case there is no need to restart the lookup, since sockets are inserted at the start of the list and the worst that will happen is that we will iterate throught some of the list elements more then once. Note that the /proc and multicast delivery was not yet converted to RCU, it still uses spinlocks for protection. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:02 +00:00
static inline bool llc_estab_match(const struct llc_sap *sap,
const struct llc_addr *daddr,
const struct llc_addr *laddr,
const struct sock *sk)
{
struct llc_sock *llc = llc_sk(sk);
return llc->laddr.lsap == laddr->lsap &&
llc->daddr.lsap == daddr->lsap &&
llc: Use normal etherdevice.h tests Convert the llc_<foo> static inlines to the equivalents from etherdevice.h and remove the llc_<foo> static inline functions. llc_mac_null -> is_zero_ether_addr llc_mac_multicast -> is_multicast_ether_addr llc_mac_match -> ether_addr_equal Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-09-01 20:11:55 +00:00
ether_addr_equal(llc->laddr.mac, laddr->mac) &&
ether_addr_equal(llc->daddr.mac, daddr->mac);
llc: convert the socket list to RCU locking For the reclamation phase we use the SLAB_DESTROY_BY_RCU mechanism, which require some extra checks in the lookup code: a) If the current socket was released, reallocated & inserted in another list it will short circuit the iteration for the current list, thus we need to restart the lookup. b) If the current socket was released, reallocated & inserted in the same list we just need to recheck it matches the look-up criteria and if not we can skip to the next element. In this case there is no need to restart the lookup, since sockets are inserted at the start of the list and the worst that will happen is that we will iterate throught some of the list elements more then once. Note that the /proc and multicast delivery was not yet converted to RCU, it still uses spinlocks for protection. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:02 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/**
[LLC]: Fix the accept path Borrowing the structure of TCP/IP for this. On the receive of new connections I was bh_lock_socking the _new_ sock, not the listening one, duh, now it survives the ssh connections storm I've been using to test this specific bug. Also fixes send side skb sock accounting. Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 10:57:21 +00:00
* __llc_lookup_established - Finds connection for the remote/local sap/mac
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* @sap: SAP
* @daddr: address of remote LLC (MAC + SAP)
* @laddr: address of local LLC (MAC + SAP)
*
* Search connection list of the SAP and finds connection using the remote
* mac, remote sap, local mac, and local sap. Returns pointer for
* connection found, %NULL otherwise.
[LLC]: Fix the accept path Borrowing the structure of TCP/IP for this. On the receive of new connections I was bh_lock_socking the _new_ sock, not the listening one, duh, now it survives the ssh connections storm I've been using to test this specific bug. Also fixes send side skb sock accounting. Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 10:57:21 +00:00
* Caller has to make sure local_bh is disabled.
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
*/
[LLC]: Fix the accept path Borrowing the structure of TCP/IP for this. On the receive of new connections I was bh_lock_socking the _new_ sock, not the listening one, duh, now it survives the ssh connections storm I've been using to test this specific bug. Also fixes send side skb sock accounting. Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 10:57:21 +00:00
static struct sock *__llc_lookup_established(struct llc_sap *sap,
struct llc_addr *daddr,
struct llc_addr *laddr)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct sock *rc;
llc: convert the socket list to RCU locking For the reclamation phase we use the SLAB_DESTROY_BY_RCU mechanism, which require some extra checks in the lookup code: a) If the current socket was released, reallocated & inserted in another list it will short circuit the iteration for the current list, thus we need to restart the lookup. b) If the current socket was released, reallocated & inserted in the same list we just need to recheck it matches the look-up criteria and if not we can skip to the next element. In this case there is no need to restart the lookup, since sockets are inserted at the start of the list and the worst that will happen is that we will iterate throught some of the list elements more then once. Note that the /proc and multicast delivery was not yet converted to RCU, it still uses spinlocks for protection. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:02 +00:00
struct hlist_nulls_node *node;
llc: replace the socket list with a local address based hash For the cases where a lot of interfaces are used in conjunction with a lot of LLC sockets bound to the same SAP, the iteration of the socket list becomes prohibitively expensive. Replacing the list with a a local address based hash significantly improves the bind and listener lookup operations as well as the datagram delivery. Connected sockets delivery is also improved, but this patch does not address the case where we have lots of sockets with the same local address connected to different remote addresses. In order to keep the socket sanity checks alive and fast a socket counter was added to the SAP structure. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:05 +00:00
int slot = llc_sk_laddr_hashfn(sap, laddr);
struct hlist_nulls_head *laddr_hb = &sap->sk_laddr_hash[slot];
llc: convert the socket list to RCU locking For the reclamation phase we use the SLAB_DESTROY_BY_RCU mechanism, which require some extra checks in the lookup code: a) If the current socket was released, reallocated & inserted in another list it will short circuit the iteration for the current list, thus we need to restart the lookup. b) If the current socket was released, reallocated & inserted in the same list we just need to recheck it matches the look-up criteria and if not we can skip to the next element. In this case there is no need to restart the lookup, since sockets are inserted at the start of the list and the worst that will happen is that we will iterate throught some of the list elements more then once. Note that the /proc and multicast delivery was not yet converted to RCU, it still uses spinlocks for protection. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:02 +00:00
rcu_read_lock();
again:
llc: replace the socket list with a local address based hash For the cases where a lot of interfaces are used in conjunction with a lot of LLC sockets bound to the same SAP, the iteration of the socket list becomes prohibitively expensive. Replacing the list with a a local address based hash significantly improves the bind and listener lookup operations as well as the datagram delivery. Connected sockets delivery is also improved, but this patch does not address the case where we have lots of sockets with the same local address connected to different remote addresses. In order to keep the socket sanity checks alive and fast a socket counter was added to the SAP structure. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:05 +00:00
sk_nulls_for_each_rcu(rc, node, laddr_hb) {
llc: convert the socket list to RCU locking For the reclamation phase we use the SLAB_DESTROY_BY_RCU mechanism, which require some extra checks in the lookup code: a) If the current socket was released, reallocated & inserted in another list it will short circuit the iteration for the current list, thus we need to restart the lookup. b) If the current socket was released, reallocated & inserted in the same list we just need to recheck it matches the look-up criteria and if not we can skip to the next element. In this case there is no need to restart the lookup, since sockets are inserted at the start of the list and the worst that will happen is that we will iterate throught some of the list elements more then once. Note that the /proc and multicast delivery was not yet converted to RCU, it still uses spinlocks for protection. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:02 +00:00
if (llc_estab_match(sap, daddr, laddr, rc)) {
mm: Rename SLAB_DESTROY_BY_RCU to SLAB_TYPESAFE_BY_RCU A group of Linux kernel hackers reported chasing a bug that resulted from their assumption that SLAB_DESTROY_BY_RCU provided an existence guarantee, that is, that no block from such a slab would be reallocated during an RCU read-side critical section. Of course, that is not the case. Instead, SLAB_DESTROY_BY_RCU only prevents freeing of an entire slab of blocks. However, there is a phrase for this, namely "type safety". This commit therefore renames SLAB_DESTROY_BY_RCU to SLAB_TYPESAFE_BY_RCU in order to avoid future instances of this sort of confusion. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Christoph Lameter <cl@linux.com> Cc: Pekka Enberg <penberg@kernel.org> Cc: David Rientjes <rientjes@google.com> Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: <linux-mm@kvack.org> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Vlastimil Babka <vbabka@suse.cz> [ paulmck: Add comments mentioning the old name, as requested by Eric Dumazet, in order to help people familiar with the old name find the new one. ] Acked-by: David Rientjes <rientjes@google.com>
2017-01-18 10:53:44 +00:00
/* Extra checks required by SLAB_TYPESAFE_BY_RCU */
net: convert sock.sk_refcnt from atomic_t to refcount_t refcount_t type and corresponding API should be used instead of atomic_t when the variable is used as a reference counter. This allows to avoid accidental refcounter overflows that might lead to use-after-free situations. This patch uses refcount_inc_not_zero() instead of atomic_inc_not_zero_hint() due to absense of a _hint() version of refcount API. If the hint() version must be used, we might need to revisit API. Signed-off-by: Elena Reshetova <elena.reshetova@intel.com> Signed-off-by: Hans Liljestrand <ishkamiel@gmail.com> Signed-off-by: Kees Cook <keescook@chromium.org> Signed-off-by: David Windsor <dwindsor@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-06-30 10:08:01 +00:00
if (unlikely(!refcount_inc_not_zero(&rc->sk_refcnt)))
llc: convert the socket list to RCU locking For the reclamation phase we use the SLAB_DESTROY_BY_RCU mechanism, which require some extra checks in the lookup code: a) If the current socket was released, reallocated & inserted in another list it will short circuit the iteration for the current list, thus we need to restart the lookup. b) If the current socket was released, reallocated & inserted in the same list we just need to recheck it matches the look-up criteria and if not we can skip to the next element. In this case there is no need to restart the lookup, since sockets are inserted at the start of the list and the worst that will happen is that we will iterate throught some of the list elements more then once. Note that the /proc and multicast delivery was not yet converted to RCU, it still uses spinlocks for protection. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:02 +00:00
goto again;
if (unlikely(llc_sk(rc)->sap != sap ||
!llc_estab_match(sap, daddr, laddr, rc))) {
sock_put(rc);
continue;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto found;
}
}
rc = NULL;
llc: replace the socket list with a local address based hash For the cases where a lot of interfaces are used in conjunction with a lot of LLC sockets bound to the same SAP, the iteration of the socket list becomes prohibitively expensive. Replacing the list with a a local address based hash significantly improves the bind and listener lookup operations as well as the datagram delivery. Connected sockets delivery is also improved, but this patch does not address the case where we have lots of sockets with the same local address connected to different remote addresses. In order to keep the socket sanity checks alive and fast a socket counter was added to the SAP structure. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:05 +00:00
/*
* if the nulls value we got at the end of this lookup is
* not the expected one, we must restart lookup.
* We probably met an item that was moved to another chain.
*/
if (unlikely(get_nulls_value(node) != slot))
goto again;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
found:
llc: convert the socket list to RCU locking For the reclamation phase we use the SLAB_DESTROY_BY_RCU mechanism, which require some extra checks in the lookup code: a) If the current socket was released, reallocated & inserted in another list it will short circuit the iteration for the current list, thus we need to restart the lookup. b) If the current socket was released, reallocated & inserted in the same list we just need to recheck it matches the look-up criteria and if not we can skip to the next element. In this case there is no need to restart the lookup, since sockets are inserted at the start of the list and the worst that will happen is that we will iterate throught some of the list elements more then once. Note that the /proc and multicast delivery was not yet converted to RCU, it still uses spinlocks for protection. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:02 +00:00
rcu_read_unlock();
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return rc;
}
[LLC]: Fix the accept path Borrowing the structure of TCP/IP for this. On the receive of new connections I was bh_lock_socking the _new_ sock, not the listening one, duh, now it survives the ssh connections storm I've been using to test this specific bug. Also fixes send side skb sock accounting. Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 10:57:21 +00:00
struct sock *llc_lookup_established(struct llc_sap *sap,
struct llc_addr *daddr,
struct llc_addr *laddr)
{
struct sock *sk;
local_bh_disable();
sk = __llc_lookup_established(sap, daddr, laddr);
local_bh_enable();
return sk;
}
llc: convert the socket list to RCU locking For the reclamation phase we use the SLAB_DESTROY_BY_RCU mechanism, which require some extra checks in the lookup code: a) If the current socket was released, reallocated & inserted in another list it will short circuit the iteration for the current list, thus we need to restart the lookup. b) If the current socket was released, reallocated & inserted in the same list we just need to recheck it matches the look-up criteria and if not we can skip to the next element. In this case there is no need to restart the lookup, since sockets are inserted at the start of the list and the worst that will happen is that we will iterate throught some of the list elements more then once. Note that the /proc and multicast delivery was not yet converted to RCU, it still uses spinlocks for protection. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:02 +00:00
static inline bool llc_listener_match(const struct llc_sap *sap,
const struct llc_addr *laddr,
const struct sock *sk)
{
struct llc_sock *llc = llc_sk(sk);
return sk->sk_type == SOCK_STREAM && sk->sk_state == TCP_LISTEN &&
llc->laddr.lsap == laddr->lsap &&
llc: Use normal etherdevice.h tests Convert the llc_<foo> static inlines to the equivalents from etherdevice.h and remove the llc_<foo> static inline functions. llc_mac_null -> is_zero_ether_addr llc_mac_multicast -> is_multicast_ether_addr llc_mac_match -> ether_addr_equal Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-09-01 20:11:55 +00:00
ether_addr_equal(llc->laddr.mac, laddr->mac);
llc: convert the socket list to RCU locking For the reclamation phase we use the SLAB_DESTROY_BY_RCU mechanism, which require some extra checks in the lookup code: a) If the current socket was released, reallocated & inserted in another list it will short circuit the iteration for the current list, thus we need to restart the lookup. b) If the current socket was released, reallocated & inserted in the same list we just need to recheck it matches the look-up criteria and if not we can skip to the next element. In this case there is no need to restart the lookup, since sockets are inserted at the start of the list and the worst that will happen is that we will iterate throught some of the list elements more then once. Note that the /proc and multicast delivery was not yet converted to RCU, it still uses spinlocks for protection. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:02 +00:00
}
llc: replace the socket list with a local address based hash For the cases where a lot of interfaces are used in conjunction with a lot of LLC sockets bound to the same SAP, the iteration of the socket list becomes prohibitively expensive. Replacing the list with a a local address based hash significantly improves the bind and listener lookup operations as well as the datagram delivery. Connected sockets delivery is also improved, but this patch does not address the case where we have lots of sockets with the same local address connected to different remote addresses. In order to keep the socket sanity checks alive and fast a socket counter was added to the SAP structure. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:05 +00:00
static struct sock *__llc_lookup_listener(struct llc_sap *sap,
struct llc_addr *laddr)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct sock *rc;
llc: convert the socket list to RCU locking For the reclamation phase we use the SLAB_DESTROY_BY_RCU mechanism, which require some extra checks in the lookup code: a) If the current socket was released, reallocated & inserted in another list it will short circuit the iteration for the current list, thus we need to restart the lookup. b) If the current socket was released, reallocated & inserted in the same list we just need to recheck it matches the look-up criteria and if not we can skip to the next element. In this case there is no need to restart the lookup, since sockets are inserted at the start of the list and the worst that will happen is that we will iterate throught some of the list elements more then once. Note that the /proc and multicast delivery was not yet converted to RCU, it still uses spinlocks for protection. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:02 +00:00
struct hlist_nulls_node *node;
llc: replace the socket list with a local address based hash For the cases where a lot of interfaces are used in conjunction with a lot of LLC sockets bound to the same SAP, the iteration of the socket list becomes prohibitively expensive. Replacing the list with a a local address based hash significantly improves the bind and listener lookup operations as well as the datagram delivery. Connected sockets delivery is also improved, but this patch does not address the case where we have lots of sockets with the same local address connected to different remote addresses. In order to keep the socket sanity checks alive and fast a socket counter was added to the SAP structure. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:05 +00:00
int slot = llc_sk_laddr_hashfn(sap, laddr);
struct hlist_nulls_head *laddr_hb = &sap->sk_laddr_hash[slot];
llc: convert the socket list to RCU locking For the reclamation phase we use the SLAB_DESTROY_BY_RCU mechanism, which require some extra checks in the lookup code: a) If the current socket was released, reallocated & inserted in another list it will short circuit the iteration for the current list, thus we need to restart the lookup. b) If the current socket was released, reallocated & inserted in the same list we just need to recheck it matches the look-up criteria and if not we can skip to the next element. In this case there is no need to restart the lookup, since sockets are inserted at the start of the list and the worst that will happen is that we will iterate throught some of the list elements more then once. Note that the /proc and multicast delivery was not yet converted to RCU, it still uses spinlocks for protection. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:02 +00:00
rcu_read_lock();
again:
llc: replace the socket list with a local address based hash For the cases where a lot of interfaces are used in conjunction with a lot of LLC sockets bound to the same SAP, the iteration of the socket list becomes prohibitively expensive. Replacing the list with a a local address based hash significantly improves the bind and listener lookup operations as well as the datagram delivery. Connected sockets delivery is also improved, but this patch does not address the case where we have lots of sockets with the same local address connected to different remote addresses. In order to keep the socket sanity checks alive and fast a socket counter was added to the SAP structure. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:05 +00:00
sk_nulls_for_each_rcu(rc, node, laddr_hb) {
llc: convert the socket list to RCU locking For the reclamation phase we use the SLAB_DESTROY_BY_RCU mechanism, which require some extra checks in the lookup code: a) If the current socket was released, reallocated & inserted in another list it will short circuit the iteration for the current list, thus we need to restart the lookup. b) If the current socket was released, reallocated & inserted in the same list we just need to recheck it matches the look-up criteria and if not we can skip to the next element. In this case there is no need to restart the lookup, since sockets are inserted at the start of the list and the worst that will happen is that we will iterate throught some of the list elements more then once. Note that the /proc and multicast delivery was not yet converted to RCU, it still uses spinlocks for protection. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:02 +00:00
if (llc_listener_match(sap, laddr, rc)) {
mm: Rename SLAB_DESTROY_BY_RCU to SLAB_TYPESAFE_BY_RCU A group of Linux kernel hackers reported chasing a bug that resulted from their assumption that SLAB_DESTROY_BY_RCU provided an existence guarantee, that is, that no block from such a slab would be reallocated during an RCU read-side critical section. Of course, that is not the case. Instead, SLAB_DESTROY_BY_RCU only prevents freeing of an entire slab of blocks. However, there is a phrase for this, namely "type safety". This commit therefore renames SLAB_DESTROY_BY_RCU to SLAB_TYPESAFE_BY_RCU in order to avoid future instances of this sort of confusion. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Christoph Lameter <cl@linux.com> Cc: Pekka Enberg <penberg@kernel.org> Cc: David Rientjes <rientjes@google.com> Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: <linux-mm@kvack.org> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Vlastimil Babka <vbabka@suse.cz> [ paulmck: Add comments mentioning the old name, as requested by Eric Dumazet, in order to help people familiar with the old name find the new one. ] Acked-by: David Rientjes <rientjes@google.com>
2017-01-18 10:53:44 +00:00
/* Extra checks required by SLAB_TYPESAFE_BY_RCU */
net: convert sock.sk_refcnt from atomic_t to refcount_t refcount_t type and corresponding API should be used instead of atomic_t when the variable is used as a reference counter. This allows to avoid accidental refcounter overflows that might lead to use-after-free situations. This patch uses refcount_inc_not_zero() instead of atomic_inc_not_zero_hint() due to absense of a _hint() version of refcount API. If the hint() version must be used, we might need to revisit API. Signed-off-by: Elena Reshetova <elena.reshetova@intel.com> Signed-off-by: Hans Liljestrand <ishkamiel@gmail.com> Signed-off-by: Kees Cook <keescook@chromium.org> Signed-off-by: David Windsor <dwindsor@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-06-30 10:08:01 +00:00
if (unlikely(!refcount_inc_not_zero(&rc->sk_refcnt)))
llc: convert the socket list to RCU locking For the reclamation phase we use the SLAB_DESTROY_BY_RCU mechanism, which require some extra checks in the lookup code: a) If the current socket was released, reallocated & inserted in another list it will short circuit the iteration for the current list, thus we need to restart the lookup. b) If the current socket was released, reallocated & inserted in the same list we just need to recheck it matches the look-up criteria and if not we can skip to the next element. In this case there is no need to restart the lookup, since sockets are inserted at the start of the list and the worst that will happen is that we will iterate throught some of the list elements more then once. Note that the /proc and multicast delivery was not yet converted to RCU, it still uses spinlocks for protection. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:02 +00:00
goto again;
if (unlikely(llc_sk(rc)->sap != sap ||
!llc_listener_match(sap, laddr, rc))) {
sock_put(rc);
continue;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto found;
}
}
rc = NULL;
llc: replace the socket list with a local address based hash For the cases where a lot of interfaces are used in conjunction with a lot of LLC sockets bound to the same SAP, the iteration of the socket list becomes prohibitively expensive. Replacing the list with a a local address based hash significantly improves the bind and listener lookup operations as well as the datagram delivery. Connected sockets delivery is also improved, but this patch does not address the case where we have lots of sockets with the same local address connected to different remote addresses. In order to keep the socket sanity checks alive and fast a socket counter was added to the SAP structure. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:05 +00:00
/*
* if the nulls value we got at the end of this lookup is
* not the expected one, we must restart lookup.
* We probably met an item that was moved to another chain.
*/
if (unlikely(get_nulls_value(node) != slot))
goto again;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
found:
llc: convert the socket list to RCU locking For the reclamation phase we use the SLAB_DESTROY_BY_RCU mechanism, which require some extra checks in the lookup code: a) If the current socket was released, reallocated & inserted in another list it will short circuit the iteration for the current list, thus we need to restart the lookup. b) If the current socket was released, reallocated & inserted in the same list we just need to recheck it matches the look-up criteria and if not we can skip to the next element. In this case there is no need to restart the lookup, since sockets are inserted at the start of the list and the worst that will happen is that we will iterate throught some of the list elements more then once. Note that the /proc and multicast delivery was not yet converted to RCU, it still uses spinlocks for protection. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:02 +00:00
rcu_read_unlock();
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
return rc;
}
llc: replace the socket list with a local address based hash For the cases where a lot of interfaces are used in conjunction with a lot of LLC sockets bound to the same SAP, the iteration of the socket list becomes prohibitively expensive. Replacing the list with a a local address based hash significantly improves the bind and listener lookup operations as well as the datagram delivery. Connected sockets delivery is also improved, but this patch does not address the case where we have lots of sockets with the same local address connected to different remote addresses. In order to keep the socket sanity checks alive and fast a socket counter was added to the SAP structure. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:05 +00:00
/**
* llc_lookup_listener - Finds listener for local MAC + SAP
* @sap: SAP
* @laddr: address of local LLC (MAC + SAP)
*
* Search connection list of the SAP and finds connection listening on
* local mac, and local sap. Returns pointer for parent socket found,
* %NULL otherwise.
* Caller has to make sure local_bh is disabled.
*/
static struct sock *llc_lookup_listener(struct llc_sap *sap,
struct llc_addr *laddr)
{
static struct llc_addr null_addr;
struct sock *rc = __llc_lookup_listener(sap, laddr);
if (!rc)
rc = __llc_lookup_listener(sap, &null_addr);
return rc;
}
[LLC]: Fix the accept path Borrowing the structure of TCP/IP for this. On the receive of new connections I was bh_lock_socking the _new_ sock, not the listening one, duh, now it survives the ssh connections storm I've been using to test this specific bug. Also fixes send side skb sock accounting. Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 10:57:21 +00:00
static struct sock *__llc_lookup(struct llc_sap *sap,
struct llc_addr *daddr,
struct llc_addr *laddr)
{
struct sock *sk = __llc_lookup_established(sap, daddr, laddr);
return sk ? : llc_lookup_listener(sap, laddr);
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/**
* llc_data_accept_state - designates if in this state data can be sent.
* @state: state of connection.
*
* Returns 0 if data can be sent, 1 otherwise.
*/
u8 llc_data_accept_state(u8 state)
{
return state != LLC_CONN_STATE_NORMAL && state != LLC_CONN_STATE_BUSY &&
state != LLC_CONN_STATE_REJ;
}
/**
[LLC]: Mark llc_find_next_offset as __init, saving some more bytes Signed-off-by: Jochen Friedrich <jochen@scram.de> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 06:57:55 +00:00
* llc_find_next_offset - finds offset for next category of transitions
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* @state: state table.
* @offset: start offset.
*
* Finds offset of next category of transitions in transition table.
* Returns the start index of next category.
*/
[LLC]: Mark llc_find_next_offset as __init, saving some more bytes Signed-off-by: Jochen Friedrich <jochen@scram.de> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 06:57:55 +00:00
static u16 __init llc_find_next_offset(struct llc_conn_state *state, u16 offset)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
u16 cnt = 0;
struct llc_conn_state_trans **next_trans;
for (next_trans = state->transitions + offset;
(*next_trans)->ev; next_trans++)
++cnt;
return cnt;
}
/**
* llc_build_offset_table - builds offset table of connection
*
* Fills offset table of connection state transition table
* (llc_offset_table).
*/
void __init llc_build_offset_table(void)
{
struct llc_conn_state *curr_state;
int state, ev_type, next_offset;
for (state = 0; state < NBR_CONN_STATES; state++) {
curr_state = &llc_conn_state_table[state];
next_offset = 0;
for (ev_type = 0; ev_type < NBR_CONN_EV; ev_type++) {
llc_offset_table[state][ev_type] = next_offset;
[LLC]: Mark llc_find_next_offset as __init, saving some more bytes Signed-off-by: Jochen Friedrich <jochen@scram.de> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 06:57:55 +00:00
next_offset += llc_find_next_offset(curr_state,
next_offset) + 1;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
}
}
/**
* llc_find_offset - finds start offset of category of transitions
* @state: state of connection
* @ev_type: type of happened event
*
* Finds start offset of desired category of transitions. Returns the
* desired start offset.
*/
static int llc_find_offset(int state, int ev_type)
{
int rc = 0;
/* at this stage, llc_offset_table[..][2] is not important. it is for
* init_pf_cycle and I don't know what is it.
*/
switch (ev_type) {
case LLC_CONN_EV_TYPE_PRIM:
rc = llc_offset_table[state][0]; break;
case LLC_CONN_EV_TYPE_PDU:
rc = llc_offset_table[state][4]; break;
case LLC_CONN_EV_TYPE_SIMPLE:
rc = llc_offset_table[state][1]; break;
case LLC_CONN_EV_TYPE_P_TMR:
case LLC_CONN_EV_TYPE_ACK_TMR:
case LLC_CONN_EV_TYPE_REJ_TMR:
case LLC_CONN_EV_TYPE_BUSY_TMR:
rc = llc_offset_table[state][3]; break;
}
return rc;
}
/**
* llc_sap_add_socket - adds a socket to a SAP
* @sap: SAP
* @sk: socket
*
llc: replace the socket list with a local address based hash For the cases where a lot of interfaces are used in conjunction with a lot of LLC sockets bound to the same SAP, the iteration of the socket list becomes prohibitively expensive. Replacing the list with a a local address based hash significantly improves the bind and listener lookup operations as well as the datagram delivery. Connected sockets delivery is also improved, but this patch does not address the case where we have lots of sockets with the same local address connected to different remote addresses. In order to keep the socket sanity checks alive and fast a socket counter was added to the SAP structure. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:05 +00:00
* This function adds a socket to the hash tables of a SAP.
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
*/
void llc_sap_add_socket(struct llc_sap *sap, struct sock *sk)
{
llc: use a device based hash table to speed up multicast delivery This patch adds a per SAP device based hash table to solve the multicast delivery scalability issue when we have large number of interfaces and a large number of sockets bound to the same SAP. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:04 +00:00
struct llc_sock *llc = llc_sk(sk);
struct hlist_head *dev_hb = llc_sk_dev_hash(sap, llc->dev->ifindex);
llc: replace the socket list with a local address based hash For the cases where a lot of interfaces are used in conjunction with a lot of LLC sockets bound to the same SAP, the iteration of the socket list becomes prohibitively expensive. Replacing the list with a a local address based hash significantly improves the bind and listener lookup operations as well as the datagram delivery. Connected sockets delivery is also improved, but this patch does not address the case where we have lots of sockets with the same local address connected to different remote addresses. In order to keep the socket sanity checks alive and fast a socket counter was added to the SAP structure. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:05 +00:00
struct hlist_nulls_head *laddr_hb = llc_sk_laddr_hash(sap, &llc->laddr);
llc: use a device based hash table to speed up multicast delivery This patch adds a per SAP device based hash table to solve the multicast delivery scalability issue when we have large number of interfaces and a large number of sockets bound to the same SAP. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:04 +00:00
[LLC]: Use refcounting with struct llc_sap Signed-off-by: Jochen Friedrich <jochen@scram.de> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 07:43:05 +00:00
llc_sap_hold(sap);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
llc_sk(sk)->sap = sap;
llc: use a device based hash table to speed up multicast delivery This patch adds a per SAP device based hash table to solve the multicast delivery scalability issue when we have large number of interfaces and a large number of sockets bound to the same SAP. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:04 +00:00
spin_lock_bh(&sap->sk_lock);
llc: set SOCK_RCU_FREE in llc_sap_add_socket() WHen an llc sock is added into the sk_laddr_hash of an llc_sap, it is not marked with SOCK_RCU_FREE. This causes that the sock could be freed while it is still being read by __llc_lookup_established() with RCU read lock. sock is refcounted, but with RCU read lock, nothing prevents the readers getting a zero refcnt. Fix it by setting SOCK_RCU_FREE in llc_sap_add_socket(). Reported-by: syzbot+11e05f04c15e03be5254@syzkaller.appspotmail.com Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-10-11 18:15:13 +00:00
sock_set_flag(sk, SOCK_RCU_FREE);
llc: replace the socket list with a local address based hash For the cases where a lot of interfaces are used in conjunction with a lot of LLC sockets bound to the same SAP, the iteration of the socket list becomes prohibitively expensive. Replacing the list with a a local address based hash significantly improves the bind and listener lookup operations as well as the datagram delivery. Connected sockets delivery is also improved, but this patch does not address the case where we have lots of sockets with the same local address connected to different remote addresses. In order to keep the socket sanity checks alive and fast a socket counter was added to the SAP structure. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:05 +00:00
sap->sk_count++;
sk_nulls_add_node_rcu(sk, laddr_hb);
llc: use a device based hash table to speed up multicast delivery This patch adds a per SAP device based hash table to solve the multicast delivery scalability issue when we have large number of interfaces and a large number of sockets bound to the same SAP. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:04 +00:00
hlist_add_head(&llc->dev_hash_node, dev_hb);
llc: convert the socket list to RCU locking For the reclamation phase we use the SLAB_DESTROY_BY_RCU mechanism, which require some extra checks in the lookup code: a) If the current socket was released, reallocated & inserted in another list it will short circuit the iteration for the current list, thus we need to restart the lookup. b) If the current socket was released, reallocated & inserted in the same list we just need to recheck it matches the look-up criteria and if not we can skip to the next element. In this case there is no need to restart the lookup, since sockets are inserted at the start of the list and the worst that will happen is that we will iterate throught some of the list elements more then once. Note that the /proc and multicast delivery was not yet converted to RCU, it still uses spinlocks for protection. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:02 +00:00
spin_unlock_bh(&sap->sk_lock);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/**
* llc_sap_remove_socket - removes a socket from SAP
* @sap: SAP
* @sk: socket
*
llc: replace the socket list with a local address based hash For the cases where a lot of interfaces are used in conjunction with a lot of LLC sockets bound to the same SAP, the iteration of the socket list becomes prohibitively expensive. Replacing the list with a a local address based hash significantly improves the bind and listener lookup operations as well as the datagram delivery. Connected sockets delivery is also improved, but this patch does not address the case where we have lots of sockets with the same local address connected to different remote addresses. In order to keep the socket sanity checks alive and fast a socket counter was added to the SAP structure. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:05 +00:00
* This function removes a connection from the hash tables of a SAP if
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* the connection was in this list.
*/
void llc_sap_remove_socket(struct llc_sap *sap, struct sock *sk)
{
llc: use a device based hash table to speed up multicast delivery This patch adds a per SAP device based hash table to solve the multicast delivery scalability issue when we have large number of interfaces and a large number of sockets bound to the same SAP. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:04 +00:00
struct llc_sock *llc = llc_sk(sk);
llc: convert the socket list to RCU locking For the reclamation phase we use the SLAB_DESTROY_BY_RCU mechanism, which require some extra checks in the lookup code: a) If the current socket was released, reallocated & inserted in another list it will short circuit the iteration for the current list, thus we need to restart the lookup. b) If the current socket was released, reallocated & inserted in the same list we just need to recheck it matches the look-up criteria and if not we can skip to the next element. In this case there is no need to restart the lookup, since sockets are inserted at the start of the list and the worst that will happen is that we will iterate throught some of the list elements more then once. Note that the /proc and multicast delivery was not yet converted to RCU, it still uses spinlocks for protection. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:02 +00:00
spin_lock_bh(&sap->sk_lock);
sk_nulls_del_node_init_rcu(sk);
llc: use a device based hash table to speed up multicast delivery This patch adds a per SAP device based hash table to solve the multicast delivery scalability issue when we have large number of interfaces and a large number of sockets bound to the same SAP. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:04 +00:00
hlist_del(&llc->dev_hash_node);
llc: replace the socket list with a local address based hash For the cases where a lot of interfaces are used in conjunction with a lot of LLC sockets bound to the same SAP, the iteration of the socket list becomes prohibitively expensive. Replacing the list with a a local address based hash significantly improves the bind and listener lookup operations as well as the datagram delivery. Connected sockets delivery is also improved, but this patch does not address the case where we have lots of sockets with the same local address connected to different remote addresses. In order to keep the socket sanity checks alive and fast a socket counter was added to the SAP structure. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:05 +00:00
sap->sk_count--;
llc: convert the socket list to RCU locking For the reclamation phase we use the SLAB_DESTROY_BY_RCU mechanism, which require some extra checks in the lookup code: a) If the current socket was released, reallocated & inserted in another list it will short circuit the iteration for the current list, thus we need to restart the lookup. b) If the current socket was released, reallocated & inserted in the same list we just need to recheck it matches the look-up criteria and if not we can skip to the next element. In this case there is no need to restart the lookup, since sockets are inserted at the start of the list and the worst that will happen is that we will iterate throught some of the list elements more then once. Note that the /proc and multicast delivery was not yet converted to RCU, it still uses spinlocks for protection. Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-12-26 11:51:02 +00:00
spin_unlock_bh(&sap->sk_lock);
[LLC]: Use refcounting with struct llc_sap Signed-off-by: Jochen Friedrich <jochen@scram.de> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 07:43:05 +00:00
llc_sap_put(sap);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
/**
* llc_conn_rcv - sends received pdus to the connection state machine
* @sk: current connection structure.
* @skb: received frame.
*
* Sends received pdus to the connection state machine.
*/
llc: "foo* bar" should be "foo *bar" Signed-off-by: Weilong Chen <chenweilong@huawei.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-20 03:14:59 +00:00
static int llc_conn_rcv(struct sock *sk, struct sk_buff *skb)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct llc_conn_state_ev *ev = llc_conn_ev(skb);
ev->type = LLC_CONN_EV_TYPE_PDU;
ev->reason = 0;
return llc_conn_state_process(sk, skb);
}
[LLC]: Fix the accept path Borrowing the structure of TCP/IP for this. On the receive of new connections I was bh_lock_socking the _new_ sock, not the listening one, duh, now it survives the ssh connections storm I've been using to test this specific bug. Also fixes send side skb sock accounting. Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 10:57:21 +00:00
static struct sock *llc_create_incoming_sock(struct sock *sk,
struct net_device *dev,
struct llc_addr *saddr,
struct llc_addr *daddr)
{
[NET] NETNS: Omit sock->sk_net without CONFIG_NET_NS. Introduce per-sock inlines: sock_net(), sock_net_set() and per-inet_timewait_sock inlines: twsk_net(), twsk_net_set(). Without CONFIG_NET_NS, no namespace other than &init_net exists. Let's explicitly define them to help compiler optimizations. Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
2008-03-25 17:26:21 +00:00
struct sock *newsk = llc_sk_alloc(sock_net(sk), sk->sk_family, GFP_ATOMIC,
net: Pass kern from net_proto_family.create to sk_alloc In preparation for changing how struct net is refcounted on kernel sockets pass the knowledge that we are creating a kernel socket from sock_create_kern through to sk_alloc. Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-05-09 02:09:13 +00:00
sk->sk_prot, 0);
[LLC]: Fix the accept path Borrowing the structure of TCP/IP for this. On the receive of new connections I was bh_lock_socking the _new_ sock, not the listening one, duh, now it survives the ssh connections storm I've been using to test this specific bug. Also fixes send side skb sock accounting. Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 10:57:21 +00:00
struct llc_sock *newllc, *llc = llc_sk(sk);
if (!newsk)
goto out;
newllc = llc_sk(newsk);
memcpy(&newllc->laddr, daddr, sizeof(newllc->laddr));
memcpy(&newllc->daddr, saddr, sizeof(newllc->daddr));
newllc->dev = dev;
dev_hold(dev);
llc_sap_add_socket(llc->sap, newsk);
llc_sap_hold(llc->sap);
out:
return newsk;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
void llc_conn_handler(struct llc_sap *sap, struct sk_buff *skb)
{
struct llc_addr saddr, daddr;
struct sock *sk;
llc_pdu_decode_sa(skb, saddr.mac);
llc_pdu_decode_ssap(skb, &saddr.lsap);
llc_pdu_decode_da(skb, daddr.mac);
llc_pdu_decode_dsap(skb, &daddr.lsap);
[LLC]: Fix the accept path Borrowing the structure of TCP/IP for this. On the receive of new connections I was bh_lock_socking the _new_ sock, not the listening one, duh, now it survives the ssh connections storm I've been using to test this specific bug. Also fixes send side skb sock accounting. Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 10:57:21 +00:00
sk = __llc_lookup(sap, &saddr, &daddr);
if (!sk)
goto drop;
bh_lock_sock(sk);
/*
* This has to be done here and not at the upper layer ->accept
* method because of the way the PROCOM state machine works:
* it needs to set several state variables (see, for instance,
* llc_adm_actions_2 in net/llc/llc_c_st.c) and send a packet to
* the originator of the new connection, and this state has to be
* in the newly created struct sock private area. -acme
*/
if (unlikely(sk->sk_state == TCP_LISTEN)) {
struct sock *newsk = llc_create_incoming_sock(sk, skb->dev,
&saddr, &daddr);
if (!newsk)
goto drop_unlock;
skb_set_owner_r(skb, newsk);
} else {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/*
[LLC]: Fix the accept path Borrowing the structure of TCP/IP for this. On the receive of new connections I was bh_lock_socking the _new_ sock, not the listening one, duh, now it survives the ssh connections storm I've been using to test this specific bug. Also fixes send side skb sock accounting. Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 10:57:21 +00:00
* Can't be skb_set_owner_r, this will be done at the
* llc_conn_state_process function, later on, when we will use
* skb_queue_rcv_skb to send it to upper layers, this is
* another trick required to cope with how the PROCOM state
* machine works. -acme
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
*/
net/llc: avoid BUG_ON() in skb_orphan() It seems nobody used LLC since linux-3.12. Fortunately fuzzers like syzkaller still know how to run this code, otherwise it would be no fun. Setting skb->sk without skb->destructor leads to all kinds of bugs, we now prefer to be very strict about it. Ideally here we would use skb_set_owner() but this helper does not exist yet, only CAN seems to have a private helper for that. Fixes: 376c7311bdb6 ("net: add a temporary sanity check in skb_orphan()") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Andrey Konovalov <andreyknvl@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-02-12 22:03:52 +00:00
skb_orphan(skb);
sock_hold(sk);
[LLC]: Fix the accept path Borrowing the structure of TCP/IP for this. On the receive of new connections I was bh_lock_socking the _new_ sock, not the listening one, duh, now it survives the ssh connections storm I've been using to test this specific bug. Also fixes send side skb sock accounting. Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 10:57:21 +00:00
skb->sk = sk;
net/llc: avoid BUG_ON() in skb_orphan() It seems nobody used LLC since linux-3.12. Fortunately fuzzers like syzkaller still know how to run this code, otherwise it would be no fun. Setting skb->sk without skb->destructor leads to all kinds of bugs, we now prefer to be very strict about it. Ideally here we would use skb_set_owner() but this helper does not exist yet, only CAN seems to have a private helper for that. Fixes: 376c7311bdb6 ("net: add a temporary sanity check in skb_orphan()") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Andrey Konovalov <andreyknvl@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-02-12 22:03:52 +00:00
skb->destructor = sock_efree;
[LLC]: Do better struct sock accounting on skbs Signed-off-by: Jochen Friedrich <jochen@scram.de> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 07:40:59 +00:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (!sock_owned_by_user(sk))
llc_conn_rcv(sk, skb);
else {
net: replace remaining __FUNCTION__ occurrences __FUNCTION__ is gcc-specific, use __func__ Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-06 04:47:47 +00:00
dprintk("%s: adding to backlog...\n", __func__);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
llc_set_backlog_type(skb, LLC_PACKET);
net: add a limit parameter to sk_add_backlog() sk_add_backlog() & sk_rcvqueues_full() hard coded sk_rcvbuf as the memory limit. We need to make this limit a parameter for TCP use. No functional change expected in this patch, all callers still using the old sk_rcvbuf limit. Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Neal Cardwell <ncardwell@google.com> Cc: Tom Herbert <therbert@google.com> Cc: Maciej Żenczykowski <maze@google.com> Cc: Yuchung Cheng <ycheng@google.com> Cc: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi> Cc: Rick Jones <rick.jones2@hp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2012-04-22 23:34:26 +00:00
if (sk_add_backlog(sk, skb, sk->sk_rcvbuf))
llc: use limited socket backlog Make llc adapt to the limited socket backlog change. Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net> Signed-off-by: Zhu Yi <yi.zhu@intel.com> Acked-by: Eric Dumazet <eric.dumazet@gmail.com> Acked-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-03-04 18:01:43 +00:00
goto drop_unlock;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
[LLC]: Fix the accept path Borrowing the structure of TCP/IP for this. On the receive of new connections I was bh_lock_socking the _new_ sock, not the listening one, duh, now it survives the ssh connections storm I've been using to test this specific bug. Also fixes send side skb sock accounting. Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 10:57:21 +00:00
out:
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
bh_unlock_sock(sk);
sock_put(sk);
return;
drop:
kfree_skb(skb);
[LLC]: Fix the accept path Borrowing the structure of TCP/IP for this. On the receive of new connections I was bh_lock_socking the _new_ sock, not the listening one, duh, now it survives the ssh connections storm I've been using to test this specific bug. Also fixes send side skb sock accounting. Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 10:57:21 +00:00
return;
drop_unlock:
kfree_skb(skb);
goto out;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
#undef LLC_REFCNT_DEBUG
#ifdef LLC_REFCNT_DEBUG
static atomic_t llc_sock_nr;
#endif
/**
* llc_backlog_rcv - Processes rx frames and expired timers.
* @sk: LLC sock (p8022 connection)
* @skb: queued rx frame or event
*
* This function processes frames that has received and timers that has
* expired during sending an I pdu (refer to data_req_handler). frames
* queue by llc_rcv function (llc_mac.c) and timers queue by timer
* callback functions(llc_c_ac.c).
*/
static int llc_backlog_rcv(struct sock *sk, struct sk_buff *skb)
{
int rc = 0;
struct llc_sock *llc = llc_sk(sk);
[LLC]: Help the compiler with likely/unlikely, saving some more bytes Signed-off-by: Jochen Friedrich <jochen@scram.de> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 06:59:22 +00:00
if (likely(llc_backlog_type(skb) == LLC_PACKET)) {
if (likely(llc->state > 1)) /* not closed */
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
rc = llc_conn_rcv(sk, skb);
else
goto out_kfree_skb;
} else if (llc_backlog_type(skb) == LLC_EVENT) {
/* timer expiration event */
[LLC]: Help the compiler with likely/unlikely, saving some more bytes Signed-off-by: Jochen Friedrich <jochen@scram.de> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 06:59:22 +00:00
if (likely(llc->state > 1)) /* not closed */
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
rc = llc_conn_state_process(sk, skb);
else
goto out_kfree_skb;
} else {
net: replace remaining __FUNCTION__ occurrences __FUNCTION__ is gcc-specific, use __func__ Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-06 04:47:47 +00:00
printk(KERN_ERR "%s: invalid skb in backlog\n", __func__);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
goto out_kfree_skb;
}
out:
return rc;
out_kfree_skb:
kfree_skb(skb);
goto out;
}
/**
* llc_sk_init - Initializes a socket with default llc values.
* @sk: socket to initialize.
*
* Initializes a socket with default llc values.
*/
llc: "foo* bar" should be "foo *bar" Signed-off-by: Weilong Chen <chenweilong@huawei.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-12-20 03:14:59 +00:00
static void llc_sk_init(struct sock *sk)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
struct llc_sock *llc = llc_sk(sk);
llc->state = LLC_CONN_STATE_ADM;
llc->inc_cntr = llc->dec_cntr = 2;
llc->dec_step = llc->connect_step = 1;
net: LLC: Convert timers to use timer_setup() In preparation for unconditionally passing the struct timer_list pointer to all timer callbacks, switch to using the new timer_setup() and from_timer() to pass the timer pointer explicitly. Cc: "David S. Miller" <davem@davemloft.net> Cc: Eric Dumazet <edumazet@google.com> Cc: Hans Liljestrand <ishkamiel@gmail.com> Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> Cc: "Reshetova, Elena" <elena.reshetova@intel.com> Cc: netdev@vger.kernel.org Signed-off-by: Kees Cook <keescook@chromium.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-24 08:45:48 +00:00
timer_setup(&llc->ack_timer.timer, llc_conn_ack_tmr_cb, 0);
[LLC]: Add sysctl support for the LLC timeouts Signed-off-by: Jochen Friedrich <jochen@scram.de> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 07:30:44 +00:00
llc->ack_timer.expire = sysctl_llc2_ack_timeout;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
net: LLC: Convert timers to use timer_setup() In preparation for unconditionally passing the struct timer_list pointer to all timer callbacks, switch to using the new timer_setup() and from_timer() to pass the timer pointer explicitly. Cc: "David S. Miller" <davem@davemloft.net> Cc: Eric Dumazet <edumazet@google.com> Cc: Hans Liljestrand <ishkamiel@gmail.com> Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> Cc: "Reshetova, Elena" <elena.reshetova@intel.com> Cc: netdev@vger.kernel.org Signed-off-by: Kees Cook <keescook@chromium.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-24 08:45:48 +00:00
timer_setup(&llc->pf_cycle_timer.timer, llc_conn_pf_cycle_tmr_cb, 0);
[LLC]: Add sysctl support for the LLC timeouts Signed-off-by: Jochen Friedrich <jochen@scram.de> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 07:30:44 +00:00
llc->pf_cycle_timer.expire = sysctl_llc2_p_timeout;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
net: LLC: Convert timers to use timer_setup() In preparation for unconditionally passing the struct timer_list pointer to all timer callbacks, switch to using the new timer_setup() and from_timer() to pass the timer pointer explicitly. Cc: "David S. Miller" <davem@davemloft.net> Cc: Eric Dumazet <edumazet@google.com> Cc: Hans Liljestrand <ishkamiel@gmail.com> Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> Cc: "Reshetova, Elena" <elena.reshetova@intel.com> Cc: netdev@vger.kernel.org Signed-off-by: Kees Cook <keescook@chromium.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-24 08:45:48 +00:00
timer_setup(&llc->rej_sent_timer.timer, llc_conn_rej_tmr_cb, 0);
[LLC]: Add sysctl support for the LLC timeouts Signed-off-by: Jochen Friedrich <jochen@scram.de> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 07:30:44 +00:00
llc->rej_sent_timer.expire = sysctl_llc2_rej_timeout;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
net: LLC: Convert timers to use timer_setup() In preparation for unconditionally passing the struct timer_list pointer to all timer callbacks, switch to using the new timer_setup() and from_timer() to pass the timer pointer explicitly. Cc: "David S. Miller" <davem@davemloft.net> Cc: Eric Dumazet <edumazet@google.com> Cc: Hans Liljestrand <ishkamiel@gmail.com> Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> Cc: "Reshetova, Elena" <elena.reshetova@intel.com> Cc: netdev@vger.kernel.org Signed-off-by: Kees Cook <keescook@chromium.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-10-24 08:45:48 +00:00
timer_setup(&llc->busy_state_timer.timer, llc_conn_busy_tmr_cb, 0);
[LLC]: Add sysctl support for the LLC timeouts Signed-off-by: Jochen Friedrich <jochen@scram.de> Signed-off-by: Arnaldo Carvalho de Melo <acme@mandriva.com>
2005-09-22 07:30:44 +00:00
llc->busy_state_timer.expire = sysctl_llc2_busy_timeout;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
llc->n2 = 2; /* max retransmit */
llc->k = 2; /* tx win size, will adjust dynam */
llc->rw = 128; /* rx win size (opt and equal to
[NET] LLC: Fix whitespace errors. Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-02-09 14:25:01 +00:00
* tx_win of remote LLC) */
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
skb_queue_head_init(&llc->pdu_unack_q);
sk->sk_backlog_rcv = llc_backlog_rcv;
}
/**
* llc_sk_alloc - Allocates LLC sock
* @family: upper layer protocol family
* @priority: for allocation (%GFP_KERNEL, %GFP_ATOMIC, etc)
*
* Allocates a LLC sock and initializes it. Returns the new LLC sock
* or %NULL if there's no memory available for one
*/
net: Pass kern from net_proto_family.create to sk_alloc In preparation for changing how struct net is refcounted on kernel sockets pass the knowledge that we are creating a kernel socket from sock_create_kern through to sk_alloc. Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-05-09 02:09:13 +00:00
struct sock *llc_sk_alloc(struct net *net, int family, gfp_t priority, struct proto *prot, int kern)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
{
net: Pass kern from net_proto_family.create to sk_alloc In preparation for changing how struct net is refcounted on kernel sockets pass the knowledge that we are creating a kernel socket from sock_create_kern through to sk_alloc. Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-05-09 02:09:13 +00:00
struct sock *sk = sk_alloc(net, family, priority, prot, kern);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
if (!sk)
goto out;
llc_sk_init(sk);
sock_init_data(NULL, sk);
#ifdef LLC_REFCNT_DEBUG
atomic_inc(&llc_sock_nr);
printk(KERN_DEBUG "LLC socket %p created in %s, now we have %d alive\n", sk,
net: replace remaining __FUNCTION__ occurrences __FUNCTION__ is gcc-specific, use __func__ Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-06 04:47:47 +00:00
__func__, atomic_read(&llc_sock_nr));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#endif
out:
return sk;
}
llc: delete timers synchronously in llc_sk_free() The connection timers of an llc sock could be still flying after we delete them in llc_sk_free(), and even possibly after we free the sock. We could just wait synchronously here in case of troubles. Note, I leave other call paths as they are, since they may not have to wait, at least we can change them to synchronously when needed. Also, move the code to net/llc/llc_conn.c, which is apparently a better place. Reported-by: <syzbot+f922284c18ea23a8e457@syzkaller.appspotmail.com> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-19 19:25:38 +00:00
void llc_sk_stop_all_timers(struct sock *sk, bool sync)
{
struct llc_sock *llc = llc_sk(sk);
if (sync) {
del_timer_sync(&llc->pf_cycle_timer.timer);
del_timer_sync(&llc->ack_timer.timer);
del_timer_sync(&llc->rej_sent_timer.timer);
del_timer_sync(&llc->busy_state_timer.timer);
} else {
del_timer(&llc->pf_cycle_timer.timer);
del_timer(&llc->ack_timer.timer);
del_timer(&llc->rej_sent_timer.timer);
del_timer(&llc->busy_state_timer.timer);
}
llc->ack_must_be_send = 0;
llc->ack_pf = 0;
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/**
* llc_sk_free - Frees a LLC socket
* @sk - socket to free
*
* Frees a LLC socket
*/
void llc_sk_free(struct sock *sk)
{
struct llc_sock *llc = llc_sk(sk);
llc->state = LLC_CONN_OUT_OF_SVC;
/* Stop all (possibly) running timers */
llc: delete timers synchronously in llc_sk_free() The connection timers of an llc sock could be still flying after we delete them in llc_sk_free(), and even possibly after we free the sock. We could just wait synchronously here in case of troubles. Note, I leave other call paths as they are, since they may not have to wait, at least we can change them to synchronously when needed. Also, move the code to net/llc/llc_conn.c, which is apparently a better place. Reported-by: <syzbot+f922284c18ea23a8e457@syzkaller.appspotmail.com> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2018-04-19 19:25:38 +00:00
llc_sk_stop_all_timers(sk, true);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#ifdef DEBUG_LLC_CONN_ALLOC
net: replace remaining __FUNCTION__ occurrences __FUNCTION__ is gcc-specific, use __func__ Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-06 04:47:47 +00:00
printk(KERN_INFO "%s: unackq=%d, txq=%d\n", __func__,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
skb_queue_len(&llc->pdu_unack_q),
skb_queue_len(&sk->sk_write_queue));
#endif
skb_queue_purge(&sk->sk_receive_queue);
skb_queue_purge(&sk->sk_write_queue);
skb_queue_purge(&llc->pdu_unack_q);
#ifdef LLC_REFCNT_DEBUG
net: convert sock.sk_refcnt from atomic_t to refcount_t refcount_t type and corresponding API should be used instead of atomic_t when the variable is used as a reference counter. This allows to avoid accidental refcounter overflows that might lead to use-after-free situations. This patch uses refcount_inc_not_zero() instead of atomic_inc_not_zero_hint() due to absense of a _hint() version of refcount API. If the hint() version must be used, we might need to revisit API. Signed-off-by: Elena Reshetova <elena.reshetova@intel.com> Signed-off-by: Hans Liljestrand <ishkamiel@gmail.com> Signed-off-by: Kees Cook <keescook@chromium.org> Signed-off-by: David Windsor <dwindsor@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-06-30 10:08:01 +00:00
if (refcount_read(&sk->sk_refcnt) != 1) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
printk(KERN_DEBUG "Destruction of LLC sock %p delayed in %s, cnt=%d\n",
net: convert sock.sk_refcnt from atomic_t to refcount_t refcount_t type and corresponding API should be used instead of atomic_t when the variable is used as a reference counter. This allows to avoid accidental refcounter overflows that might lead to use-after-free situations. This patch uses refcount_inc_not_zero() instead of atomic_inc_not_zero_hint() due to absense of a _hint() version of refcount API. If the hint() version must be used, we might need to revisit API. Signed-off-by: Elena Reshetova <elena.reshetova@intel.com> Signed-off-by: Hans Liljestrand <ishkamiel@gmail.com> Signed-off-by: Kees Cook <keescook@chromium.org> Signed-off-by: David Windsor <dwindsor@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2017-06-30 10:08:01 +00:00
sk, __func__, refcount_read(&sk->sk_refcnt));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
printk(KERN_DEBUG "%d LLC sockets are still alive\n",
atomic_read(&llc_sock_nr));
} else {
atomic_dec(&llc_sock_nr);
printk(KERN_DEBUG "LLC socket %p released in %s, %d are still alive\n", sk,
net: replace remaining __FUNCTION__ occurrences __FUNCTION__ is gcc-specific, use __func__ Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-03-06 04:47:47 +00:00
__func__, atomic_read(&llc_sock_nr));
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
#endif
sock_put(sk);
}
/**
* llc_sk_reset - resets a connection
* @sk: LLC socket to reset
*
* Resets a connection to the out of service state. Stops its timers
* and frees any frames in the queues of the connection.
*/
void llc_sk_reset(struct sock *sk)
{
struct llc_sock *llc = llc_sk(sk);
llc_conn_ac_stop_all_timers(sk, NULL);
skb_queue_purge(&sk->sk_write_queue);
skb_queue_purge(&llc->pdu_unack_q);
llc->remote_busy_flag = 0;
llc->cause_flag = 0;
llc->retry_count = 0;
llc_conn_set_p_flag(sk, 0);
llc->f_flag = 0;
llc->s_flag = 0;
llc->ack_pf = 0;
llc->first_pdu_Ns = 0;
llc->ack_must_be_send = 0;
llc->dec_step = 1;
llc->inc_cntr = 2;
llc->dec_cntr = 2;
llc->X = 0;
llc->failed_data_req = 0 ;
llc->last_nr = 0;
}
Reference in New Issue Copy Permalink