* [PATCH] ax25_cb refcounting & waitqueue usage
@ 2003-07-06 17:27 Jeroen Vreeken
2003-07-07 13:37 ` Arnaldo Carvalho de Melo
0 siblings, 1 reply; 6+ messages in thread
From: Jeroen Vreeken @ 2003-07-06 17:27 UTC (permalink / raw)
To: linux-hams; +Cc: ralf
[-- Attachment #1: Type: text/plain, Size: 478 bytes --]
Hi,
Second patch for this weekend...
This patch does two things, it adds reference counting to the ax25_cb
structure (until now only the ax25_cb list was protected, not the parts
using it) and it fixes an oops on interrupted socket syscalls. (e.g.
pressing Ctrl-C while in the connect function) The waitqueue was not
removed from the list and caused an oops in the socket release.
I have had only a quick look at Steven's patch but I don't think the two
will clash...
Jeroen
[-- Attachment #2: ax25-2.5.74.rxq.diff --]
[-- Type: application/octet-stream, Size: 6320 bytes --]
diff -ru linux-2.5.74/net/ax25/af_ax25.c linux-2.5.74.rxq/net/ax25/af_ax25.c
--- linux-2.5.74/net/ax25/af_ax25.c 2003-07-06 01:05:13.000000000 +0200
+++ linux-2.5.74.rxq/net/ax25/af_ax25.c 2003-07-06 18:43:45.000000000 +0200
@@ -54,13 +54,20 @@
ax25_cb *ax25_list;
spinlock_t ax25_list_lock = SPIN_LOCK_UNLOCKED;
+spinlock_t ax25_cb_lock = SPIN_LOCK_UNLOCKED;
+
static struct proto_ops ax25_proto_ops;
/*
* Free an allocated ax25 control block. This is done to centralise
* the MOD count code.
*/
-void ax25_free_cb(ax25_cb *ax25)
+/*
+ * Changed name to include undescores. Nobody should call this without
+ * going through the ref counter anymore
+ * - PE1RXQ
+ */
+void __ax25_free_cb(ax25_cb *ax25)
{
if (ax25->digipeat != NULL) {
kfree(ax25->digipeat);
@@ -70,9 +77,25 @@
kfree(ax25);
}
+void ax25_up_cb(ax25_cb *ax25)
+{
+ spin_lock_bh(&ax25_cb_lock);
+ ax25->refcount++;
+ spin_unlock_bh(&ax25_cb_lock);
+}
+
+void ax25_down_cb(ax25_cb *ax25)
+{
+ spin_lock_bh(&ax25_cb_lock);
+ ax25->refcount--;
+ if (!ax25->refcount)
+ __ax25_free_cb(ax25);
+ spin_unlock_bh(&ax25_cb_lock);
+}
+
static void ax25_free_sock(struct sock *sk)
{
- ax25_free_cb(ax25_sk(sk));
+ ax25_down_cb(ax25_sk(sk));
}
/*
@@ -85,6 +108,7 @@
spin_lock_bh(&ax25_list_lock);
if ((s = ax25_list) == ax25) {
ax25_list = s->next;
+ ax25_down_cb(ax25);
spin_unlock_bh(&ax25_list_lock);
return;
}
@@ -92,6 +116,7 @@
while (s != NULL && s->next != NULL) {
if (s->next == ax25) {
s->next = ax25->next;
+ ax25_down_cb(ax25);
spin_unlock_bh(&ax25_list_lock);
return;
}
@@ -156,6 +181,7 @@
void ax25_insert_socket(ax25_cb *ax25)
{
spin_lock_bh(&ax25_list_lock);
+ ax25_up_cb(ax25);
ax25->next = ax25_list;
ax25_list = ax25;
spin_unlock_bh(&ax25_list_lock);
@@ -240,6 +266,7 @@
if (s->digipeat != NULL && s->digipeat->ndigi != 0)
continue;
}
+ ax25_up_cb(s);
spin_unlock_bh(&ax25_list_lock);
return s;
@@ -352,7 +379,7 @@
sk_free(ax25->sk);
}
} else {
- ax25_free_cb(ax25);
+ ax25_down_cb(ax25);
}
}
@@ -507,6 +534,7 @@
return NULL;
memset(ax25, 0x00, sizeof(*ax25));
+ ax25_up_cb(ax25);
skb_queue_head_init(&ax25->write_queue);
skb_queue_head_init(&ax25->frag_queue);
@@ -877,7 +905,7 @@
break;
default:
sk_free(sk);
- ax25_free_cb(ax25);
+ ax25_down_cb(ax25);
return NULL;
}
@@ -1092,7 +1120,7 @@
int addr_len, int flags)
{
struct sock *sk = sock->sk;
- ax25_cb *ax25 = ax25_sk(sk);
+ ax25_cb *ax25 = ax25_sk(sk), *ax25t;
struct full_sockaddr_ax25 *fsa = (struct full_sockaddr_ax25 *)uaddr;
ax25_digi *digi = NULL;
int ct = 0, err = 0;
@@ -1207,11 +1235,12 @@
}
if (sk->sk_type == SOCK_SEQPACKET &&
- ax25_find_cb(&ax25->source_addr, &fsa->fsa_ax25.sax25_call, digi,
- ax25->ax25_dev->dev)) {
+ (ax25t=ax25_find_cb(&ax25->source_addr, &fsa->fsa_ax25.sax25_call, digi,
+ ax25->ax25_dev->dev))) {
if (digi != NULL)
kfree(digi);
err = -EADDRINUSE; /* Already such a connection */
+ ax25_down_cb(ax25t);
goto out;
}
@@ -1272,6 +1301,8 @@
lock_sock(sk);
continue;
}
+ current->state = TASK_RUNNING;
+ remove_wait_queue(sk->sk_sleep, &wait);
return -ERESTARTSYS;
}
current->state = TASK_RUNNING;
@@ -1339,6 +1370,8 @@
lock_sock(sk);
continue;
}
+ current->state = TASK_RUNNING;
+ remove_wait_queue(sk->sk_sleep, &wait);
return -ERESTARTSYS;
}
current->state = TASK_RUNNING;
@@ -1955,6 +1988,8 @@
EXPORT_SYMBOL(ax25_rebuild_header);
EXPORT_SYMBOL(ax25_findbyuid);
EXPORT_SYMBOL(ax25_find_cb);
+EXPORT_SYMBOL(ax25_up_cb);
+EXPORT_SYMBOL(ax25_down_cb);
EXPORT_SYMBOL(ax25_linkfail_register);
EXPORT_SYMBOL(ax25_linkfail_release);
EXPORT_SYMBOL(ax25_listen_register);
diff -ru linux-2.5.74/net/ax25/ax25_in.c linux-2.5.74.rxq/net/ax25/ax25_in.c
--- linux-2.5.74/net/ax25/ax25_in.c 2003-07-06 01:05:12.000000000 +0200
+++ linux-2.5.74.rxq/net/ax25/ax25_in.c 2003-07-06 18:44:42.000000000 +0200
@@ -329,6 +329,7 @@
if (ax25_process_rx_frame(ax25, skb, type, dama) == 0)
kfree_skb(skb);
+ ax25_down_cb(ax25);
return 0;
}
diff -ru linux-2.5.74/net/ax25/ax25_ip.c linux-2.5.74.rxq/net/ax25/ax25_ip.c
--- linux-2.5.74/net/ax25/ax25_ip.c 2003-07-06 01:05:13.000000000 +0200
+++ linux-2.5.74.rxq/net/ax25/ax25_ip.c 2003-07-06 18:46:14.000000000 +0200
@@ -107,6 +107,7 @@
ax25_address *src, *dst;
ax25_dev *ax25_dev;
ax25_route _route, *route = &_route;
+ ax25_cb *ax25;
dst = (ax25_address *)(bp + 1);
src = (ax25_address *)(bp + 8);
@@ -167,9 +168,14 @@
skb_pull(ourskb, AX25_HEADER_LEN - 1); /* Keep PID */
ourskb->nh.raw = ourskb->data;
- ax25_send_frame(ourskb, ax25_dev->values[AX25_VALUES_PACLEN], &src_c,
-&dst_c, route->digipeat, dev);
-
+ ax25=ax25_send_frame(
+ ourskb,
+ ax25_dev->values[AX25_VALUES_PACLEN],
+ &src_c,
+ &dst_c, route->digipeat, dev);
+ if (ax25) {
+ ax25_down_cb(ax25);
+ }
goto put;
}
}
diff -ru linux-2.5.74/net/ax25/ax25_out.c linux-2.5.74.rxq/net/ax25/ax25_out.c
--- linux-2.5.74/net/ax25/ax25_out.c 2003-07-06 01:05:13.000000000 +0200
+++ linux-2.5.74.rxq/net/ax25/ax25_out.c 2003-07-06 18:46:53.000000000 +0200
@@ -71,7 +71,7 @@
if (digi != NULL) {
if ((ax25->digipeat = kmalloc(sizeof(ax25_digi), GFP_ATOMIC)) == NULL) {
- ax25_free_cb(ax25);
+ ax25_down_cb(ax25);
return NULL;
}
memcpy(ax25->digipeat, digi, sizeof(ax25_digi));
diff -ru linux-2.5.74/include/net/ax25.h linux-2.5.74.rxq/include/net/ax25.h
--- linux-2.5.74/include/net/ax25.h 2003-07-06 01:12:50.000000000 +0200
+++ linux-2.5.74.rxq/include/net/ax25.h 2003-07-05 23:19:25.000000000 +0200
@@ -199,6 +199,7 @@
unsigned char window;
struct timer_list timer;
struct sock *sk; /* Backlink to socket */
+ int refcount;
} ax25_cb;
#define ax25_sk(__sk) ((ax25_cb *)(__sk)->sk_protinfo)
@@ -206,7 +207,8 @@
/* af_ax25.c */
extern ax25_cb *ax25_list;
extern spinlock_t ax25_list_lock;
-extern void ax25_free_cb(ax25_cb *);
+extern void ax25_up_cb(ax25_cb *);
+extern void ax25_down_cb(ax25_cb *);
extern void ax25_insert_socket(ax25_cb *);
struct sock *ax25_find_listener(ax25_address *, int, struct net_device *, int);
struct sock *ax25_get_socket(ax25_address *, ax25_address *, int);
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] ax25_cb refcounting & waitqueue usage
2003-07-06 17:27 [PATCH] ax25_cb refcounting & waitqueue usage Jeroen Vreeken
@ 2003-07-07 13:37 ` Arnaldo Carvalho de Melo
2003-07-07 17:23 ` Jeroen Vreeken
0 siblings, 1 reply; 6+ messages in thread
From: Arnaldo Carvalho de Melo @ 2003-07-07 13:37 UTC (permalink / raw)
To: Jeroen Vreeken; +Cc: linux-hams, ralf
Em Sun, Jul 06, 2003 at 07:27:34PM +0200, Jeroen Vreeken escreveu:
> Hi,
>
> Second patch for this weekend...
> This patch does two things, it adds reference counting to the ax25_cb
> structure (until now only the ax25_cb list was protected, not the parts
> using it) and it fixes an oops on interrupted socket syscalls. (e.g.
> pressing Ctrl-C while in the connect function) The waitqueue was not
> removed from the list and caused an oops in the socket release.
>
> I have had only a quick look at Steven's patch but I don't think the two
> will clash...
Hey, why not use sk->refcnt by means of sock_hold/sock_put? Current sources
even do that implicitely when you do a sk_add_node (sock_hold) and a
sk_del_node{_init} (sock_put)... Now it is just a matter of doing a sock_hold
when searching in some list and then dropping the refcnt when done using the
sock returned by the search function.
- Arnaldo
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] ax25_cb refcounting & waitqueue usage
2003-07-07 13:37 ` Arnaldo Carvalho de Melo
@ 2003-07-07 17:23 ` Jeroen Vreeken
2003-07-07 17:25 ` Arnaldo Carvalho de Melo
0 siblings, 1 reply; 6+ messages in thread
From: Jeroen Vreeken @ 2003-07-07 17:23 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo; +Cc: linux-hams, ralf
On 2003.07.07 15:37:10 +0200 Arnaldo Carvalho de Melo wrote:
> Em Sun, Jul 06, 2003 at 07:27:34PM +0200, Jeroen Vreeken escreveu:
> > Hi,
> >
> > Second patch for this weekend...
> > This patch does two things, it adds reference counting to the ax25_cb
> > structure (until now only the ax25_cb list was protected, not the parts
> > using it) and it fixes an oops on interrupted socket syscalls. (e.g.
> > pressing Ctrl-C while in the connect function) The waitqueue was not
> > removed from the list and caused an oops in the socket release.
> >
> > I have had only a quick look at Steven's patch but I don't think the
> two
> > will clash...
>
> Hey, why not use sk->refcnt by means of sock_hold/sock_put? Current
> sources
> even do that implicitely when you do a sk_add_node (sock_hold) and a
> sk_del_node{_init} (sock_put)... Now it is just a matter of doing a
> sock_hold
> when searching in some list and then dropping the refcnt when done using
> the
> sock returned by the search function.
Unfortunatly the ax25_cbs are not directly related to a sock structure...
Sometimes they don't even have them, e.g. when you do ip in connected mode
or netrom and rose interlinks. In those cases you can even share an ax25_cb
between a socket and those protocols in which case you absolutly need the
refcounting....
Consider the following:
- ip datagram is send to ax25 layer.
- ax25 layer uses a lookup in the ax25_cb list to find an existing
connection
- list code locks, searches, unlocks, returns
***** preempt/interrupt/other processor
- application closes socket
- ax25_cb is removed from the list (again with proper list locking)
- ax25_cb is freed
***** preempt/interrupt/other processor back to ip handling routine
- ip handling routine uses ax25_cb
- BOOM!
Jeroen
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] ax25_cb refcounting & waitqueue usage
2003-07-07 17:23 ` Jeroen Vreeken
@ 2003-07-07 17:25 ` Arnaldo Carvalho de Melo
2003-07-08 22:52 ` Jeroen Vreeken
0 siblings, 1 reply; 6+ messages in thread
From: Arnaldo Carvalho de Melo @ 2003-07-07 17:25 UTC (permalink / raw)
To: Jeroen Vreeken; +Cc: linux-hams, ralf
Em Mon, Jul 07, 2003 at 07:23:40PM +0200, Jeroen Vreeken escreveu:
> On 2003.07.07 15:37:10 +0200 Arnaldo Carvalho de Melo wrote:
> > Em Sun, Jul 06, 2003 at 07:27:34PM +0200, Jeroen Vreeken escreveu:
> > > Hi,
> > >
> > > Second patch for this weekend...
> > > This patch does two things, it adds reference counting to the ax25_cb
> > > structure (until now only the ax25_cb list was protected, not the parts
> > > using it) and it fixes an oops on interrupted socket syscalls. (e.g.
> > > pressing Ctrl-C while in the connect function) The waitqueue was not
> > > removed from the list and caused an oops in the socket release.
> > >
> > > I have had only a quick look at Steven's patch but I don't think the
> > two
> > > will clash...
> >
> > Hey, why not use sk->refcnt by means of sock_hold/sock_put? Current
> > sources
> > even do that implicitely when you do a sk_add_node (sock_hold) and a
> > sk_del_node{_init} (sock_put)... Now it is just a matter of doing a
> > sock_hold
> > when searching in some list and then dropping the refcnt when done using
> > the
> > sock returned by the search function.
>
> Unfortunatly the ax25_cbs are not directly related to a sock structure...
> Sometimes they don't even have them, e.g. when you do ip in connected mode
> or netrom and rose interlinks. In those cases you can even share an ax25_cb
> between a socket and those protocols in which case you absolutly need the
> refcounting....
> Consider the following:
>
> - ip datagram is send to ax25 layer.
> - ax25 layer uses a lookup in the ax25_cb list to find an existing
> connection
> - list code locks, searches, unlocks, returns
> ***** preempt/interrupt/other processor
> - application closes socket
> - ax25_cb is removed from the list (again with proper list locking)
> - ax25_cb is freed
> ***** preempt/interrupt/other processor back to ip handling routine
> - ip handling routine uses ax25_cb
> - BOOM!
Got it, i.e. the fact that ax25_cb structs are not directly always associated
with a struct sock, the scenario above is the same for struct sock refcounting,
so may I suggest that you use the same arquitecture of list handling in struct
sock? i.e. use hlist and have a ax25_cb_add, ax25_cb_del{_init} that does the
refcounting, etc, so that it "looks like" what we have for struct sock and that
way a casual ax25 hacker can take advantage of what he learnt from studying
struct sock handling? I plan to attack the list searching problem in a general
way, i.e. abstracting more the search in a bsearch(3) way, using inlines and
macros so that all the protocols have a more consistent implementation.
- Arnaldo
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] ax25_cb refcounting & waitqueue usage
2003-07-07 17:25 ` Arnaldo Carvalho de Melo
@ 2003-07-08 22:52 ` Jeroen Vreeken
2003-07-09 17:06 ` Jeroen Vreeken
0 siblings, 1 reply; 6+ messages in thread
From: Jeroen Vreeken @ 2003-07-08 22:52 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo; +Cc: linux-hams, ralf
On 2003.07.07 19:25:04 +0200 Arnaldo Carvalho de Melo wrote:
> so may I suggest that you use the same arquitecture of list handling in
> struct
> sock? i.e. use hlist and have a ax25_cb_add, ax25_cb_del{_init} that does
> the
> refcounting, etc, so that it "looks like" what we have for struct sock
> and that
> way a casual ax25 hacker can take advantage of what he learnt from
> studying
> struct sock handling? I plan to attack the list searching problem in a
> general
> way, i.e. abstracting more the search in a bsearch(3) way, using inlines
> and
> macros so that all the protocols have a more consistent implementation.
Good idea...
I now have the first ax25 stack with hlists running on my test system...
It needs some cleaning up (some old junk commented out at the moment and I
need to rename some functions to be more descriptive) which will probably
happen in the next few days.
I also removed the bpq header length stuff in the skb allocation, it now
uses the hard_header_length of the network device... The stack should not
have to now which devices there might possibly be.
Jeroen
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH] ax25_cb refcounting & waitqueue usage
2003-07-08 22:52 ` Jeroen Vreeken
@ 2003-07-09 17:06 ` Jeroen Vreeken
0 siblings, 0 replies; 6+ messages in thread
From: Jeroen Vreeken @ 2003-07-09 17:06 UTC (permalink / raw)
To: ralf; +Cc: Arnaldo Carvalho de Melo, linux-hams
[-- Attachment #1: Type: text/plain, Size: 595 bytes --]
Hi,
This is a new patch to replace my first one...
(Anybody on this list should know by now that I always send at least two
versions of each patch ;)
It does the following things:
-fixes waitqueue oops on interrupted socket call
-Adds refcounting to the ax25_cb struct with two new functions:
ax25_cb_hold() and ax25_cb_put()
-skb allocation uses dev->hard_header_len instead of some predicted worst
case.
-The ax25_cb list uses hlists
The biggest change is the hlist one combined with the refcounting.
This makes ax25 look somewhat like the other network stack as Arnaldo
sugested.
Jeroen
[-- Attachment #2: ax25-2.5.74.rxq.diff --]
[-- Type: application/octet-stream, Size: 14158 bytes --]
diff -ru linux-2.5.74/net/ax25/af_ax25.c linux-2.5.74.rxq/net/ax25/af_ax25.c
--- linux-2.5.74/net/ax25/af_ax25.c 2003-07-06 01:05:13.000000000 +0200
+++ linux-2.5.74.rxq/net/ax25/af_ax25.c 2003-07-09 18:22:17.000000000 +0200
@@ -51,16 +51,23 @@
-ax25_cb *ax25_list;
+HLIST_HEAD(ax25_list);
spinlock_t ax25_list_lock = SPIN_LOCK_UNLOCKED;
+spinlock_t ax25_cb_lock = SPIN_LOCK_UNLOCKED;
+
static struct proto_ops ax25_proto_ops;
/*
* Free an allocated ax25 control block. This is done to centralise
* the MOD count code.
*/
-void ax25_free_cb(ax25_cb *ax25)
+/*
+ * Changed name to include undescores. Nobody should call this without
+ * going through the ref counter anymore
+ * - PE1RXQ
+ */
+void __ax25_free_cb(ax25_cb *ax25)
{
if (ax25->digipeat != NULL) {
kfree(ax25->digipeat);
@@ -70,35 +77,36 @@
kfree(ax25);
}
+void ax25_cb_hold(ax25_cb *ax25)
+{
+ spin_lock_bh(&ax25_cb_lock);
+ ax25->refcount++;
+ spin_unlock_bh(&ax25_cb_lock);
+}
+
+void ax25_cb_put(ax25_cb *ax25)
+{
+ spin_lock_bh(&ax25_cb_lock);
+ ax25->refcount--;
+ if (!ax25->refcount)
+ __ax25_free_cb(ax25);
+ spin_unlock_bh(&ax25_cb_lock);
+}
+
static void ax25_free_sock(struct sock *sk)
{
- ax25_free_cb(ax25_sk(sk));
+ ax25_cb_put(ax25_sk(sk));
}
/*
* Socket removal during an interrupt is now safe.
*/
-static void ax25_remove_socket(ax25_cb *ax25)
+static void ax25_cb_del(ax25_cb *ax25)
{
- ax25_cb *s;
-
spin_lock_bh(&ax25_list_lock);
- if ((s = ax25_list) == ax25) {
- ax25_list = s->next;
- spin_unlock_bh(&ax25_list_lock);
- return;
- }
-
- while (s != NULL && s->next != NULL) {
- if (s->next == ax25) {
- s->next = ax25->next;
- spin_unlock_bh(&ax25_list_lock);
- return;
- }
-
- s = s->next;
- }
+ hlist_del_init(&ax25->ax25_node);
spin_unlock_bh(&ax25_list_lock);
+ ax25_cb_put(ax25);
}
/*
@@ -108,17 +116,17 @@
{
ax25_dev *ax25_dev;
ax25_cb *s;
+ struct hlist_node *node;
if ((ax25_dev = ax25_dev_ax25dev(dev)) == NULL)
return;
spin_lock_bh(&ax25_list_lock);
- for (s = ax25_list; s != NULL; s = s->next) {
+ ax25_for_each(s, node, &ax25_list)
if (s->ax25_dev == ax25_dev) {
s->ax25_dev = NULL;
ax25_disconnect(s, ENETUNREACH);
}
- }
spin_unlock_bh(&ax25_list_lock);
}
@@ -153,11 +161,11 @@
/*
* Add a socket to the bound sockets list.
*/
-void ax25_insert_socket(ax25_cb *ax25)
+void ax25_cb_add(ax25_cb *ax25)
{
spin_lock_bh(&ax25_list_lock);
- ax25->next = ax25_list;
- ax25_list = ax25;
+ ax25_cb_hold(ax25);
+ hlist_add_head(&ax25->ax25_node, &ax25_list);
spin_unlock_bh(&ax25_list_lock);
}
@@ -169,9 +177,10 @@
struct net_device *dev, int type)
{
ax25_cb *s;
+ struct hlist_node *node;
spin_lock_bh(&ax25_list_lock);
- for (s = ax25_list; s != NULL; s = s->next) {
+ ax25_for_each(s, node, &ax25_list) {
if ((s->iamdigi && !digi) || (!s->iamdigi && digi))
continue;
if (s->sk && !ax25cmp(&s->source_addr, addr) &&
@@ -197,9 +206,10 @@
{
struct sock *sk = NULL;
ax25_cb *s;
+ struct hlist_node *node;
spin_lock_bh(&ax25_list_lock);
- for (s = ax25_list; s != NULL; s = s->next) {
+ ax25_for_each(s, node, &ax25_list) {
if (s->sk && !ax25cmp(&s->source_addr, my_addr) &&
!ax25cmp(&s->dest_addr, dest_addr) &&
s->sk->sk_type == type) {
@@ -223,9 +233,10 @@
ax25_digi *digi, struct net_device *dev)
{
ax25_cb *s;
+ struct hlist_node *node;
spin_lock_bh(&ax25_list_lock);
- for (s = ax25_list; s != NULL; s = s->next) {
+ ax25_for_each(s, node, &ax25_list) {
if (s->sk && s->sk->sk_type != SOCK_SEQPACKET)
continue;
if (s->ax25_dev == NULL)
@@ -240,6 +251,7 @@
if (s->digipeat != NULL && s->digipeat->ndigi != 0)
continue;
}
+ ax25_cb_hold(s);
spin_unlock_bh(&ax25_list_lock);
return s;
@@ -257,16 +269,17 @@
{
struct sock *sk = NULL;
ax25_cb *s;
+ struct hlist_node *node;
spin_lock_bh(&ax25_list_lock);
- for (s = ax25_list; s != NULL; s = s->next) {
+ ax25_for_each(s, node, &ax25_list)
if (s->sk != NULL && ax25cmp(&s->source_addr, addr) == 0 &&
s->sk->sk_type == SOCK_RAW) {
sk = s->sk;
lock_sock(sk);
break;
}
- }
+
spin_unlock_bh(&ax25_list_lock);
return sk;
@@ -312,7 +325,7 @@
{
struct sk_buff *skb;
- ax25_remove_socket(ax25);
+ ax25_cb_del(ax25);
ax25_stop_heartbeat(ax25);
ax25_stop_t1timer(ax25);
@@ -352,7 +365,7 @@
sk_free(ax25->sk);
}
} else {
- ax25_free_cb(ax25);
+ ax25_cb_put(ax25);
}
}
@@ -507,6 +520,7 @@
return NULL;
memset(ax25, 0x00, sizeof(*ax25));
+ ax25_cb_hold(ax25);
skb_queue_head_init(&ax25->write_queue);
skb_queue_head_init(&ax25->frag_queue);
@@ -877,7 +891,7 @@
break;
default:
sk_free(sk);
- ax25_free_cb(ax25);
+ ax25_cb_put(ax25);
return NULL;
}
@@ -1076,7 +1090,7 @@
ax25_fillin_cb(ax25, ax25_dev);
done:
- ax25_insert_socket(ax25);
+ ax25_cb_add(ax25);
sk->sk_zapped = 0;
out:
@@ -1092,7 +1106,7 @@
int addr_len, int flags)
{
struct sock *sk = sock->sk;
- ax25_cb *ax25 = ax25_sk(sk);
+ ax25_cb *ax25 = ax25_sk(sk), *ax25t;
struct full_sockaddr_ax25 *fsa = (struct full_sockaddr_ax25 *)uaddr;
ax25_digi *digi = NULL;
int ct = 0, err = 0;
@@ -1198,7 +1212,7 @@
goto out;
ax25_fillin_cb(ax25, ax25->ax25_dev);
- ax25_insert_socket(ax25);
+ ax25_cb_add(ax25);
} else {
if (ax25->ax25_dev == NULL) {
err = -EHOSTUNREACH;
@@ -1207,11 +1221,12 @@
}
if (sk->sk_type == SOCK_SEQPACKET &&
- ax25_find_cb(&ax25->source_addr, &fsa->fsa_ax25.sax25_call, digi,
- ax25->ax25_dev->dev)) {
+ (ax25t=ax25_find_cb(&ax25->source_addr, &fsa->fsa_ax25.sax25_call, digi,
+ ax25->ax25_dev->dev))) {
if (digi != NULL)
kfree(digi);
err = -EADDRINUSE; /* Already such a connection */
+ ax25_cb_put(ax25t);
goto out;
}
@@ -1272,6 +1287,8 @@
lock_sock(sk);
continue;
}
+ current->state = TASK_RUNNING;
+ remove_wait_queue(sk->sk_sleep, &wait);
return -ERESTARTSYS;
}
current->state = TASK_RUNNING;
@@ -1339,6 +1356,8 @@
lock_sock(sk);
continue;
}
+ current->state = TASK_RUNNING;
+ remove_wait_queue(sk->sk_sleep, &wait);
return -ERESTARTSYS;
}
current->state = TASK_RUNNING;
@@ -1518,7 +1537,7 @@
SOCK_DEBUG(sk, "AX.25: sendto: building packet.\n");
/* Assume the worst case */
- size = len + 3 + ax25_addr_size(dp) + AX25_BPQ_HEADER_LEN;
+ size = len + ax25->ax25_dev->dev->hard_header_len;
skb = sock_alloc_send_skb(sk, size, msg->msg_flags&MSG_DONTWAIT, &err);
if (skb == NULL)
@@ -1844,6 +1863,7 @@
int len = 0;
off_t pos = 0;
off_t begin = 0;
+ struct hlist_node *node;
spin_lock_bh(&ax25_list_lock);
@@ -1852,7 +1872,7 @@
* magic dev src_addr dest_addr,digi1,digi2,.. st vs vr va t1 t1 t2 t2 t3 t3 idle idle n2 n2 rtt window paclen Snd-Q Rcv-Q inode
*/
- for (ax25 = ax25_list; ax25 != NULL; ax25 = ax25->next) {
+ ax25_for_each(ax25, node, &ax25_list) {
len += sprintf(buffer+len, "%8.8lx %s %s%s ",
(long) ax25,
ax25->ax25_dev == NULL? "???" : ax25->ax25_dev->dev->name,
@@ -1955,6 +1975,8 @@
EXPORT_SYMBOL(ax25_rebuild_header);
EXPORT_SYMBOL(ax25_findbyuid);
EXPORT_SYMBOL(ax25_find_cb);
+EXPORT_SYMBOL(ax25_cb_hold);
+EXPORT_SYMBOL(ax25_cb_put);
EXPORT_SYMBOL(ax25_linkfail_register);
EXPORT_SYMBOL(ax25_linkfail_release);
EXPORT_SYMBOL(ax25_listen_register);
diff -ru linux-2.5.74/net/ax25/ax25_ds_subr.c linux-2.5.74.rxq/net/ax25/ax25_ds_subr.c
--- linux-2.5.74/net/ax25/ax25_ds_subr.c 2003-07-06 01:05:13.000000000 +0200
+++ linux-2.5.74.rxq/net/ax25/ax25_ds_subr.c 2003-07-09 18:18:00.000000000 +0200
@@ -40,6 +40,7 @@
void ax25_ds_enquiry_response(ax25_cb *ax25)
{
ax25_cb *ax25o;
+ struct hlist_node *node;
/* Please note that neither DK4EG´s nor DG2FEF´s
* DAMA spec mention the following behaviour as seen
@@ -80,7 +81,7 @@
ax25_ds_set_timer(ax25->ax25_dev);
spin_lock_bh(&ax25_list_lock);
- for (ax25o = ax25_list; ax25o != NULL; ax25o = ax25o->next) {
+ ax25_for_each(ax25o, node, &ax25_list) {
if (ax25o == ax25)
continue;
@@ -160,9 +161,10 @@
{
ax25_cb *ax25;
int res = 0;
+ struct hlist_node *node;
spin_lock_bh(&ax25_list_lock);
- for (ax25 = ax25_list; ax25 != NULL ; ax25 = ax25->next)
+ ax25_for_each(ax25, node, &ax25_list)
if (ax25->ax25_dev == ax25_dev && (ax25->condition & AX25_COND_DAMA_MODE) && ax25->state > AX25_STATE_1) {
res = 1;
break;
diff -ru linux-2.5.74/net/ax25/ax25_ds_timer.c linux-2.5.74.rxq/net/ax25/ax25_ds_timer.c
--- linux-2.5.74/net/ax25/ax25_ds_timer.c 2003-07-06 01:05:13.000000000 +0200
+++ linux-2.5.74.rxq/net/ax25/ax25_ds_timer.c 2003-07-09 18:18:21.000000000 +0200
@@ -74,6 +74,7 @@
{
ax25_dev *ax25_dev = (struct ax25_dev *) arg;
ax25_cb *ax25;
+ struct hlist_node *node;
if (ax25_dev == NULL || !ax25_dev->dama.slave)
return; /* Yikes! */
@@ -84,7 +85,7 @@
}
spin_lock_bh(&ax25_list_lock);
- for (ax25=ax25_list; ax25 != NULL; ax25 = ax25->next) {
+ ax25_for_each(ax25, node, &ax25_list) {
if (ax25->ax25_dev != ax25_dev || !(ax25->condition & AX25_COND_DAMA_MODE))
continue;
diff -ru linux-2.5.74/net/ax25/ax25_in.c linux-2.5.74.rxq/net/ax25/ax25_in.c
--- linux-2.5.74/net/ax25/ax25_in.c 2003-07-06 01:05:12.000000000 +0200
+++ linux-2.5.74.rxq/net/ax25/ax25_in.c 2003-07-09 18:15:47.000000000 +0200
@@ -329,6 +329,7 @@
if (ax25_process_rx_frame(ax25, skb, type, dama) == 0)
kfree_skb(skb);
+ ax25_cb_put(ax25);
return 0;
}
@@ -429,7 +430,7 @@
ax25->state = AX25_STATE_3;
- ax25_insert_socket(ax25);
+ ax25_cb_add(ax25);
ax25_start_heartbeat(ax25);
ax25_start_t3timer(ax25);
diff -ru linux-2.5.74/net/ax25/ax25_ip.c linux-2.5.74.rxq/net/ax25/ax25_ip.c
--- linux-2.5.74/net/ax25/ax25_ip.c 2003-07-06 01:05:13.000000000 +0200
+++ linux-2.5.74.rxq/net/ax25/ax25_ip.c 2003-07-09 18:16:33.000000000 +0200
@@ -107,6 +107,7 @@
ax25_address *src, *dst;
ax25_dev *ax25_dev;
ax25_route _route, *route = &_route;
+ ax25_cb *ax25;
dst = (ax25_address *)(bp + 1);
src = (ax25_address *)(bp + 8);
@@ -167,9 +168,14 @@
skb_pull(ourskb, AX25_HEADER_LEN - 1); /* Keep PID */
ourskb->nh.raw = ourskb->data;
- ax25_send_frame(ourskb, ax25_dev->values[AX25_VALUES_PACLEN], &src_c,
-&dst_c, route->digipeat, dev);
-
+ ax25=ax25_send_frame(
+ ourskb,
+ ax25_dev->values[AX25_VALUES_PACLEN],
+ &src_c,
+ &dst_c, route->digipeat, dev);
+ if (ax25) {
+ ax25_cb_put(ax25);
+ }
goto put;
}
}
diff -ru linux-2.5.74/net/ax25/ax25_out.c linux-2.5.74.rxq/net/ax25/ax25_out.c
--- linux-2.5.74/net/ax25/ax25_out.c 2003-07-06 01:05:13.000000000 +0200
+++ linux-2.5.74.rxq/net/ax25/ax25_out.c 2003-07-09 18:17:26.000000000 +0200
@@ -71,7 +71,7 @@
if (digi != NULL) {
if ((ax25->digipeat = kmalloc(sizeof(ax25_digi), GFP_ATOMIC)) == NULL) {
- ax25_free_cb(ax25);
+ ax25_cb_put(ax25);
return NULL;
}
memcpy(ax25->digipeat, digi, sizeof(ax25_digi));
@@ -93,7 +93,7 @@
#endif
}
- ax25_insert_socket(ax25);
+ ax25_cb_add(ax25);
ax25->state = AX25_STATE_1;
diff -ru linux-2.5.74/net/ax25/ax25_subr.c linux-2.5.74.rxq/net/ax25/ax25_subr.c
--- linux-2.5.74/net/ax25/ax25_subr.c 2003-07-06 01:05:13.000000000 +0200
+++ linux-2.5.74.rxq/net/ax25/ax25_subr.c 2003-07-07 22:30:04.000000000 +0200
@@ -158,10 +158,10 @@
struct sk_buff *skb;
unsigned char *dptr;
- if ((skb = alloc_skb(AX25_BPQ_HEADER_LEN + ax25_addr_size(ax25->digipeat) + 2, GFP_ATOMIC)) == NULL)
+ if ((skb = alloc_skb(ax25->ax25_dev->dev->hard_header_len + 2, GFP_ATOMIC)) == NULL)
return;
- skb_reserve(skb, AX25_BPQ_HEADER_LEN + ax25_addr_size(ax25->digipeat));
+ skb_reserve(skb, ax25->ax25_dev->dev->hard_header_len);
skb->nh.raw = skb->data;
@@ -202,10 +202,10 @@
if (dev == NULL)
return;
- if ((skb = alloc_skb(AX25_BPQ_HEADER_LEN + ax25_addr_size(digi) + 1, GFP_ATOMIC)) == NULL)
+ if ((skb = alloc_skb(dev->hard_header_len + 1, GFP_ATOMIC)) == NULL)
return; /* Next SABM will get DM'd */
- skb_reserve(skb, AX25_BPQ_HEADER_LEN + ax25_addr_size(digi));
+ skb_reserve(skb, dev->hard_header_len);
skb->nh.raw = skb->data;
ax25_digi_invert(digi, &retdigi);
diff -ru linux-2.5.74/include/net/ax25.h linux-2.5.74.rxq/include/net/ax25.h
--- linux-2.5.74/include/net/ax25.h 2003-07-06 01:12:50.000000000 +0200
+++ linux-2.5.74.rxq/include/net/ax25.h 2003-07-09 18:09:38.000000000 +0200
@@ -10,6 +10,7 @@
#include <linux/ax25.h>
#include <linux/spinlock.h>
#include <linux/timer.h>
+#include <linux/list.h>
#include <asm/atomic.h>
#define AX25_T1CLAMPLO 1
@@ -180,7 +181,7 @@
} ax25_dev;
typedef struct ax25_cb {
- struct ax25_cb *next;
+ struct hlist_node ax25_node;
ax25_address source_addr, dest_addr;
ax25_digi *digipeat;
ax25_dev *ax25_dev;
@@ -199,15 +200,20 @@
unsigned char window;
struct timer_list timer;
struct sock *sk; /* Backlink to socket */
+ int refcount;
} ax25_cb;
#define ax25_sk(__sk) ((ax25_cb *)(__sk)->sk_protinfo)
+#define ax25_for_each(__ax25, node, list) \
+ hlist_for_each_entry(__ax25, node, list, ax25_node)
+
/* af_ax25.c */
-extern ax25_cb *ax25_list;
+extern struct hlist_head ax25_list;
extern spinlock_t ax25_list_lock;
-extern void ax25_free_cb(ax25_cb *);
-extern void ax25_insert_socket(ax25_cb *);
+extern void ax25_cb_hold(ax25_cb *);
+extern void ax25_cb_put(ax25_cb *);
+extern void ax25_cb_add(ax25_cb *);
struct sock *ax25_find_listener(ax25_address *, int, struct net_device *, int);
struct sock *ax25_get_socket(ax25_address *, ax25_address *, int);
extern ax25_cb *ax25_find_cb(ax25_address *, ax25_address *, ax25_digi *, struct net_device *);
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2003-07-09 17:06 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2003-07-06 17:27 [PATCH] ax25_cb refcounting & waitqueue usage Jeroen Vreeken
2003-07-07 13:37 ` Arnaldo Carvalho de Melo
2003-07-07 17:23 ` Jeroen Vreeken
2003-07-07 17:25 ` Arnaldo Carvalho de Melo
2003-07-08 22:52 ` Jeroen Vreeken
2003-07-09 17:06 ` Jeroen Vreeken
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox