* [PATCH net-next 1/1] - ROSE device usage count
@ 2015-05-20 14:31 Richard Stearn
2015-05-20 15:16 ` walter harms
` (2 more replies)
0 siblings, 3 replies; 6+ messages in thread
From: Richard Stearn @ 2015-05-20 14:31 UTC (permalink / raw)
To: linux-hams
Whilst investigating couple of issues in AX25 I discovered that
whilst the NETROM driver had been updated to manage the device
usage count the ROSE drive had not. The result being that the
ROSE driver post kernel 2.4.x would not unload cleanly and
required a reboot to clean up.
As both NETROM and ROSE drivers are broadly similar in design
and implementation I have updated the socket locking and device
usage count management in the ROSE driver to match the NETROM
driver.
git diff against net-next
---------------------------------------------------------------------
diff --git a/net/rose/af_rose.c b/net/rose/af_rose.c
index 36dbc2d..3b90a2d 100644
--- a/net/rose/af_rose.c
+++ b/net/rose/af_rose.c
@@ -662,23 +662,31 @@ static int rose_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
ax25_uid_assoc *user;
int n;
- if (!sock_flag(sk, SOCK_ZAPPED))
+ lock_sock(sk);
+ if (!sock_flag(sk, SOCK_ZAPPED)) {
+ release_sock(sk);
return -EINVAL;
-
- if (addr_len != sizeof(struct sockaddr_rose) && addr_len != sizeof(struct full_sockaddr_rose))
+ }
+ if (addr_len != sizeof(struct sockaddr_rose) && addr_len != sizeof(struct full_sockaddr_rose)) {
+ release_sock(sk);
return -EINVAL;
-
- if (addr->srose_family != AF_ROSE)
+ }
+ if (addr->srose_family != AF_ROSE) {
+ release_sock(sk);
return -EINVAL;
-
- if (addr_len == sizeof(struct sockaddr_rose) && addr->srose_ndigis > 1)
+ }
+ if (addr_len == sizeof(struct sockaddr_rose) && addr->srose_ndigis > 1) {
+ release_sock(sk);
return -EINVAL;
-
- if ((unsigned int) addr->srose_ndigis > ROSE_MAX_DIGIS)
+ }
+ if ((unsigned int) addr->srose_ndigis > ROSE_MAX_DIGIS) {
+ release_sock(sk);
return -EINVAL;
-
- if ((dev = rose_dev_get(&addr->srose_addr)) == NULL)
+ }
+ if ((dev = rose_dev_get(&addr->srose_addr)) == NULL) {
+ release_sock(sk);
return -EADDRNOTAVAIL;
+ }
source = &addr->srose_call;
@@ -687,8 +695,11 @@ static int rose_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
rose->source_call = user->call;
ax25_uid_put(user);
} else {
- if (ax25_uid_policy && !capable(CAP_NET_BIND_SERVICE))
+ if (ax25_uid_policy && !capable(CAP_NET_BIND_SERVICE)) {
+ release_sock(sk);
+ dev_put(dev);
return -EACCES;
+ }
rose->source_call = *source;
}
@@ -709,6 +720,8 @@ static int rose_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
rose_insert_socket(sk);
sock_reset_flag(sk, SOCK_ZAPPED);
+ dev_put(dev);
+ release_sock(sk);
return 0;
}
@@ -785,6 +798,7 @@ static int rose_connect(struct socket *sock, struct sockaddr *uaddr, int addr_le
user = ax25_findbyuid(current_euid());
if (!user) {
+ dev_put(dev);
err = -EINVAL;
goto out_release;
}
@@ -794,6 +808,7 @@ static int rose_connect(struct socket *sock, struct sockaddr *uaddr, int addr_le
rose->device = dev;
ax25_uid_put(user);
+ dev_put(dev);
rose_insert_socket(sk); /* Finish the bind */
}
rose->dest_addr = addr->srose_addr;
@@ -1607,22 +1622,22 @@ static void __exit rose_exit(void)
rose_rt_free();
- ax25_protocol_release(AX25_P_ROSE);
- ax25_linkfail_release(&rose_linkfail_notifier);
-
if (ax25cmp(&rose_callsign, &null_ax25_address) != 0)
ax25_listen_release(&rose_callsign, NULL);
#ifdef CONFIG_SYSCTL
rose_unregister_sysctl();
#endif
+
+ ax25_linkfail_release(&rose_linkfail_notifier);
+ ax25_protocol_release(AX25_P_ROSE);
+
unregister_netdevice_notifier(&rose_dev_notifier);
sock_unregister(PF_ROSE);
for (i = 0; i < rose_ndevs; i++) {
struct net_device *dev = dev_rose[i];
-
if (dev) {
unregister_netdev(dev);
free_netdev(dev);
@@ -1632,5 +1647,4 @@ static void __exit rose_exit(void)
kfree(dev_rose);
proto_unregister(&rose_proto);
}
-
module_exit(rose_exit);
diff --git a/net/rose/rose_loopback.c b/net/rose/rose_loopback.c
index 3444562..ea48cee 100644
--- a/net/rose/rose_loopback.c
+++ b/net/rose/rose_loopback.c
@@ -102,6 +102,7 @@ static void rose_loopback_timer(unsigned long param)
if ((dev = rose_dev_get(dest)) != NULL) {
if (rose_rx_call_request(skb, dev, rose_loopback_neigh, lci_o) == 0)
kfree_skb(skb);
+ dev_put(dev);
} else {
kfree_skb(skb);
}
diff --git a/net/rose/rose_route.c b/net/rose/rose_route.c
index 40148932..46505317 100644
--- a/net/rose/rose_route.c
+++ b/net/rose/rose_route.c
@@ -599,6 +599,7 @@ static struct net_device *rose_ax25_dev_find(char *devname)
if ((dev->flags & IFF_UP) && dev->type == ARPHRD_AX25)
return dev;
+ dev_put(dev);
return NULL;
}
@@ -615,6 +616,8 @@ struct net_device *rose_dev_first(void)
if (first == NULL || strncmp(dev->name, first->name, 3) < 0)
first = dev;
}
+ if (first)
+ dev_hold(first);
rcu_read_unlock();
return first;
@@ -742,6 +745,7 @@ int rose_rt_ioctl(unsigned int cmd, void __user *arg)
if (rose_route.ndigis > AX25_MAX_DIGIS)
return -EINVAL;
err = rose_add_node(&rose_route, dev);
+ dev_put(dev);
return err;
case SIOCDELRT:
@@ -750,6 +754,7 @@ int rose_rt_ioctl(unsigned int cmd, void __user *arg)
if ((dev = rose_ax25_dev_find(rose_route.device)) == NULL)
return -EINVAL;
err = rose_del_node(&rose_route, dev);
+ dev_put(dev);
return err;
case SIOCRSCLRRT:
---------------------------------------------------------------------
--
Regards
Richard
^ permalink raw reply related [flat|nested] 6+ messages in thread* Re: [PATCH net-next 1/1] - ROSE device usage count
2015-05-20 14:31 [PATCH net-next 1/1] - ROSE device usage count Richard Stearn
@ 2015-05-20 15:16 ` walter harms
2015-06-15 8:42 ` Ralf Baechle DL5RB
[not found] ` <555CC7A8.30108@trinnet.net>
2015-06-15 18:45 ` [PATCH net-next 1/1] - ROSE device usage count - repost with signoff Richard Stearn
2 siblings, 1 reply; 6+ messages in thread
From: walter harms @ 2015-05-20 15:16 UTC (permalink / raw)
To: Richard Stearn; +Cc: linux-hams
Am 20.05.2015 16:31, schrieb Richard Stearn:
> Whilst investigating couple of issues in AX25 I discovered that
> whilst the NETROM driver had been updated to manage the device
> usage count the ROSE drive had not. The result being that the
> ROSE driver post kernel 2.4.x would not unload cleanly and
> required a reboot to clean up.
>
> As both NETROM and ROSE drivers are broadly similar in design
> and implementation I have updated the socket locking and device
> usage count management in the ROSE driver to match the NETROM
> driver.
>
> git diff against net-next
> ---------------------------------------------------------------------
> diff --git a/net/rose/af_rose.c b/net/rose/af_rose.c
> index 36dbc2d..3b90a2d 100644
> --- a/net/rose/af_rose.c
> +++ b/net/rose/af_rose.c
> @@ -662,23 +662,31 @@ static int rose_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
> ax25_uid_assoc *user;
> int n;
>
> - if (!sock_flag(sk, SOCK_ZAPPED))
> + lock_sock(sk);
> + if (!sock_flag(sk, SOCK_ZAPPED)) {
> + release_sock(sk);
> return -EINVAL;
> -
> - if (addr_len != sizeof(struct sockaddr_rose) && addr_len != sizeof(struct full_sockaddr_rose))
> + }
> + if (addr_len != sizeof(struct sockaddr_rose) && addr_len != sizeof(struct full_sockaddr_rose)) {
> + release_sock(sk);
> return -EINVAL;
> -
> - if (addr->srose_family != AF_ROSE)
> + }
> + if (addr->srose_family != AF_ROSE) {
> + release_sock(sk);
> return -EINVAL;
> -
> - if (addr_len == sizeof(struct sockaddr_rose) && addr->srose_ndigis > 1)
> + }
> + if (addr_len == sizeof(struct sockaddr_rose) && addr->srose_ndigis > 1) {
> + release_sock(sk);
> return -EINVAL;
> -
> - if ((unsigned int) addr->srose_ndigis > ROSE_MAX_DIGIS)
> + }
> + if ((unsigned int) addr->srose_ndigis > ROSE_MAX_DIGIS) {
> + release_sock(sk);
> return -EINVAL;
> -
> - if ((dev = rose_dev_get(&addr->srose_addr)) == NULL)
> + }
> + if ((dev = rose_dev_get(&addr->srose_addr)) == NULL) {
> + release_sock(sk);
> return -EADDRNOTAVAIL;
> + }
so far i know it is kernel style one line one command
dev = rose_dev_get(&addr->srose_addr);
if (dev == NULL)
maybe you want to send the code to kernel-janitors@vger.kernel.org
to get more feedback ?
re,
wh
> source = &addr->srose_call;
>
> @@ -687,8 +695,11 @@ static int rose_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
> rose->source_call = user->call;
> ax25_uid_put(user);
> } else {
> - if (ax25_uid_policy && !capable(CAP_NET_BIND_SERVICE))
> + if (ax25_uid_policy && !capable(CAP_NET_BIND_SERVICE)) {
> + release_sock(sk);
> + dev_put(dev);
> return -EACCES;
> + }
> rose->source_call = *source;
> }
>
> @@ -709,6 +720,8 @@ static int rose_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
> rose_insert_socket(sk);
>
> sock_reset_flag(sk, SOCK_ZAPPED);
> + dev_put(dev);
> + release_sock(sk);
>
> return 0;
> }
> @@ -785,6 +798,7 @@ static int rose_connect(struct socket *sock, struct sockaddr *uaddr, int addr_le
>
> user = ax25_findbyuid(current_euid());
> if (!user) {
> + dev_put(dev);
> err = -EINVAL;
> goto out_release;
> }
> @@ -794,6 +808,7 @@ static int rose_connect(struct socket *sock, struct sockaddr *uaddr, int addr_le
> rose->device = dev;
> ax25_uid_put(user);
>
> + dev_put(dev);
> rose_insert_socket(sk); /* Finish the bind */
> }
> rose->dest_addr = addr->srose_addr;
> @@ -1607,22 +1622,22 @@ static void __exit rose_exit(void)
>
> rose_rt_free();
>
> - ax25_protocol_release(AX25_P_ROSE);
> - ax25_linkfail_release(&rose_linkfail_notifier);
> -
> if (ax25cmp(&rose_callsign, &null_ax25_address) != 0)
> ax25_listen_release(&rose_callsign, NULL);
>
> #ifdef CONFIG_SYSCTL
> rose_unregister_sysctl();
> #endif
> +
> + ax25_linkfail_release(&rose_linkfail_notifier);
> + ax25_protocol_release(AX25_P_ROSE);
> +
> unregister_netdevice_notifier(&rose_dev_notifier);
>
> sock_unregister(PF_ROSE);
>
> for (i = 0; i < rose_ndevs; i++) {
> struct net_device *dev = dev_rose[i];
> -
> if (dev) {
> unregister_netdev(dev);
> free_netdev(dev);
> @@ -1632,5 +1647,4 @@ static void __exit rose_exit(void)
> kfree(dev_rose);
> proto_unregister(&rose_proto);
> }
> -
> module_exit(rose_exit);
> diff --git a/net/rose/rose_loopback.c b/net/rose/rose_loopback.c
> index 3444562..ea48cee 100644
> --- a/net/rose/rose_loopback.c
> +++ b/net/rose/rose_loopback.c
> @@ -102,6 +102,7 @@ static void rose_loopback_timer(unsigned long param)
> if ((dev = rose_dev_get(dest)) != NULL) {
> if (rose_rx_call_request(skb, dev, rose_loopback_neigh, lci_o) == 0)
> kfree_skb(skb);
> + dev_put(dev);
> } else {
> kfree_skb(skb);
> }
> diff --git a/net/rose/rose_route.c b/net/rose/rose_route.c
> index 40148932..46505317 100644
> --- a/net/rose/rose_route.c
> +++ b/net/rose/rose_route.c
> @@ -599,6 +599,7 @@ static struct net_device *rose_ax25_dev_find(char *devname)
> if ((dev->flags & IFF_UP) && dev->type == ARPHRD_AX25)
> return dev;
>
> + dev_put(dev);
> return NULL;
> }
>
> @@ -615,6 +616,8 @@ struct net_device *rose_dev_first(void)
> if (first == NULL || strncmp(dev->name, first->name, 3) < 0)
> first = dev;
> }
> + if (first)
> + dev_hold(first);
> rcu_read_unlock();
>
> return first;
> @@ -742,6 +745,7 @@ int rose_rt_ioctl(unsigned int cmd, void __user *arg)
> if (rose_route.ndigis > AX25_MAX_DIGIS)
> return -EINVAL;
> err = rose_add_node(&rose_route, dev);
> + dev_put(dev);
> return err;
>
> case SIOCDELRT:
> @@ -750,6 +754,7 @@ int rose_rt_ioctl(unsigned int cmd, void __user *arg)
> if ((dev = rose_ax25_dev_find(rose_route.device)) == NULL)
> return -EINVAL;
> err = rose_del_node(&rose_route, dev);
> + dev_put(dev);
> return err;
>
> case SIOCRSCLRRT:
> ---------------------------------------------------------------------
>
^ permalink raw reply [flat|nested] 6+ messages in thread* Re: [PATCH net-next 1/1] - ROSE device usage count
2015-05-20 15:16 ` walter harms
@ 2015-06-15 8:42 ` Ralf Baechle DL5RB
0 siblings, 0 replies; 6+ messages in thread
From: Ralf Baechle DL5RB @ 2015-06-15 8:42 UTC (permalink / raw)
To: walter harms; +Cc: Richard Stearn, linux-hams
On Wed, May 20, 2015 at 05:16:58PM +0200, walter harms wrote:
> > + if ((dev = rose_dev_get(&addr->srose_addr)) == NULL) {
> > + release_sock(sk);
> > return -EADDRNOTAVAIL;
> > + }
>
> so far i know it is kernel style one line one command
>
> dev = rose_dev_get(&addr->srose_addr);
> if (dev == NULL)
The construct is acceptable because it was already there and also it's
not considered multiple statements for puposes of code formatting unlike
if (!foo->bar)
return ETOOFROB;
> maybe you want to send the code to kernel-janitors@vger.kernel.org
> to get more feedback ?
One this that is grave problem for this patch howeer (see
Documentation/SubmittingPatches) is the missing Signed-off-by: line.
Richard, wanna resubmit or is it ok for me to add your Signed-off-by?
Thanks,
Ralf
^ permalink raw reply [flat|nested] 6+ messages in thread
[parent not found: <555CC7A8.30108@trinnet.net>]
* [PATCH net-next 1/1] - ROSE device usage count - repost with signoff
2015-05-20 14:31 [PATCH net-next 1/1] - ROSE device usage count Richard Stearn
2015-05-20 15:16 ` walter harms
[not found] ` <555CC7A8.30108@trinnet.net>
@ 2015-06-15 18:45 ` Richard Stearn
2 siblings, 0 replies; 6+ messages in thread
From: Richard Stearn @ 2015-06-15 18:45 UTC (permalink / raw)
To: linux-hams
Whilst investigating couple of issues in AX25 I discovered that
whilst the NETROM driver had been updated to manage the device
usage count the ROSE drive had not. The result being that the
ROSE driver post kernel 2.4.x would not unload cleanly and
required a reboot to clean up.
As both NETROM and ROSE drivers are broadly similar in design
and implementation I have updated the socket locking and device
usage count management in the ROSE driver to match the NETROM
driver.
git diff against net-next
---
Signed-off-by: Richard W. Stearn richard@rns-stearn.demon.co.uk
diff --git a/net/rose/af_rose.c b/net/rose/af_rose.c
index 36dbc2d..3b90a2d 100644
--- a/net/rose/af_rose.c
+++ b/net/rose/af_rose.c
@@ -662,23 +662,31 @@ static int rose_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
ax25_uid_assoc *user;
int n;
- if (!sock_flag(sk, SOCK_ZAPPED))
+ lock_sock(sk);
+ if (!sock_flag(sk, SOCK_ZAPPED)) {
+ release_sock(sk);
return -EINVAL;
-
- if (addr_len != sizeof(struct sockaddr_rose) && addr_len != sizeof(struct full_sockaddr_rose))
+ }
+ if (addr_len != sizeof(struct sockaddr_rose) && addr_len != sizeof(struct full_sockaddr_rose)) {
+ release_sock(sk);
return -EINVAL;
-
- if (addr->srose_family != AF_ROSE)
+ }
+ if (addr->srose_family != AF_ROSE) {
+ release_sock(sk);
return -EINVAL;
-
- if (addr_len == sizeof(struct sockaddr_rose) && addr->srose_ndigis > 1)
+ }
+ if (addr_len == sizeof(struct sockaddr_rose) && addr->srose_ndigis > 1) {
+ release_sock(sk);
return -EINVAL;
-
- if ((unsigned int) addr->srose_ndigis > ROSE_MAX_DIGIS)
+ }
+ if ((unsigned int) addr->srose_ndigis > ROSE_MAX_DIGIS) {
+ release_sock(sk);
return -EINVAL;
-
- if ((dev = rose_dev_get(&addr->srose_addr)) == NULL)
+ }
+ if ((dev = rose_dev_get(&addr->srose_addr)) == NULL) {
+ release_sock(sk);
return -EADDRNOTAVAIL;
+ }
source = &addr->srose_call;
@@ -687,8 +695,11 @@ static int rose_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
rose->source_call = user->call;
ax25_uid_put(user);
} else {
- if (ax25_uid_policy && !capable(CAP_NET_BIND_SERVICE))
+ if (ax25_uid_policy && !capable(CAP_NET_BIND_SERVICE)) {
+ release_sock(sk);
+ dev_put(dev);
return -EACCES;
+ }
rose->source_call = *source;
}
@@ -709,6 +720,8 @@ static int rose_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len)
rose_insert_socket(sk);
sock_reset_flag(sk, SOCK_ZAPPED);
+ dev_put(dev);
+ release_sock(sk);
return 0;
}
@@ -785,6 +798,7 @@ static int rose_connect(struct socket *sock, struct sockaddr *uaddr, int addr_le
user = ax25_findbyuid(current_euid());
if (!user) {
+ dev_put(dev);
err = -EINVAL;
goto out_release;
}
@@ -794,6 +808,7 @@ static int rose_connect(struct socket *sock, struct sockaddr *uaddr, int addr_le
rose->device = dev;
ax25_uid_put(user);
+ dev_put(dev);
rose_insert_socket(sk); /* Finish the bind */
}
rose->dest_addr = addr->srose_addr;
@@ -1607,22 +1622,22 @@ static void __exit rose_exit(void)
rose_rt_free();
- ax25_protocol_release(AX25_P_ROSE);
- ax25_linkfail_release(&rose_linkfail_notifier);
-
if (ax25cmp(&rose_callsign, &null_ax25_address) != 0)
ax25_listen_release(&rose_callsign, NULL);
#ifdef CONFIG_SYSCTL
rose_unregister_sysctl();
#endif
+
+ ax25_linkfail_release(&rose_linkfail_notifier);
+ ax25_protocol_release(AX25_P_ROSE);
+
unregister_netdevice_notifier(&rose_dev_notifier);
sock_unregister(PF_ROSE);
for (i = 0; i < rose_ndevs; i++) {
struct net_device *dev = dev_rose[i];
-
if (dev) {
unregister_netdev(dev);
free_netdev(dev);
@@ -1632,5 +1647,4 @@ static void __exit rose_exit(void)
kfree(dev_rose);
proto_unregister(&rose_proto);
}
-
module_exit(rose_exit);
diff --git a/net/rose/rose_loopback.c b/net/rose/rose_loopback.c
index 3444562..ea48cee 100644
--- a/net/rose/rose_loopback.c
+++ b/net/rose/rose_loopback.c
@@ -102,6 +102,7 @@ static void rose_loopback_timer(unsigned long param)
if ((dev = rose_dev_get(dest)) != NULL) {
if (rose_rx_call_request(skb, dev, rose_loopback_neigh, lci_o) == 0)
kfree_skb(skb);
+ dev_put(dev);
} else {
kfree_skb(skb);
}
diff --git a/net/rose/rose_route.c b/net/rose/rose_route.c
index 40148932..46505317 100644
--- a/net/rose/rose_route.c
+++ b/net/rose/rose_route.c
@@ -599,6 +599,7 @@ static struct net_device *rose_ax25_dev_find(char *devname)
if ((dev->flags & IFF_UP) && dev->type == ARPHRD_AX25)
return dev;
+ dev_put(dev);
return NULL;
}
@@ -615,6 +616,8 @@ struct net_device *rose_dev_first(void)
if (first == NULL || strncmp(dev->name, first->name, 3) < 0)
first = dev;
}
+ if (first)
+ dev_hold(first);
rcu_read_unlock();
return first;
@@ -742,6 +745,7 @@ int rose_rt_ioctl(unsigned int cmd, void __user *arg)
if (rose_route.ndigis > AX25_MAX_DIGIS)
return -EINVAL;
err = rose_add_node(&rose_route, dev);
+ dev_put(dev);
return err;
case SIOCDELRT:
@@ -750,6 +754,7 @@ int rose_rt_ioctl(unsigned int cmd, void __user *arg)
if ((dev = rose_ax25_dev_find(rose_route.device)) == NULL)
return -EINVAL;
err = rose_del_node(&rose_route, dev);
+ dev_put(dev);
return err;
case SIOCRSCLRRT:
---------------------------------------------------------------------
--
Regards
Richard
^ permalink raw reply related [flat|nested] 6+ messages in thread
end of thread, other threads:[~2015-06-17 9:12 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2015-05-20 14:31 [PATCH net-next 1/1] - ROSE device usage count Richard Stearn
2015-05-20 15:16 ` walter harms
2015-06-15 8:42 ` Ralf Baechle DL5RB
[not found] ` <555CC7A8.30108@trinnet.net>
[not found] ` <5576BE7C.7070704@free.fr>
2015-06-09 12:34 ` Richard Stearn
[not found] ` <557759F2.2090406@free.fr>
[not found] ` <55780375.1080504@rns-stearn.demon.co.uk>
[not found] ` <557CA340.3030708@free.fr>
[not found] ` <557CA97D.9050300@free.fr>
[not found] ` <20150615012146.GA23451@linux-mips.org>
[not found] ` <557EAB73.7010104@free.fr>
2015-06-17 9:12 ` kernel crash when device rose set down f6bvp
2015-06-15 18:45 ` [PATCH net-next 1/1] - ROSE device usage count - repost with signoff Richard Stearn
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).