* Re: [Drbd-dev] [PATCH 2/2 v2] netlink: kill eff_cap from struct netlink_skb_parms
From: Patrick McHardy @ 2011-03-08 14:50 UTC (permalink / raw)
To: Evgeniy Polyakov
Cc: Chris Wright, David Miller, linux-fbdev, netdev,
linux-security-module, dm-devel, drbd-dev
In-Reply-To: <20110304012956.GA13573@ioremap.net>
Am 04.03.2011 02:29, schrieb Evgeniy Polyakov:
> Hi.
>
> On Thu, Mar 03, 2011 at 11:37:46PM +0100, Lars Ellenberg (lars.ellenberg@linbit.com) wrote:
>> If so, then this change introduces the possibility for normal users to
>> send privileged commands to connector based subsystems, even if they
>> may not be able to bind() to suitable sockets to receive any replies.
>>
>> Am I missing something?
>
> Yup, connector is very async at that place, but I wonder why the hell I
> ever made that decision. I believe we can replace it with pure sync call
> of the registered connector callback, since netlink is synchronous and
> no one has any problem with it.
>
Are you going to do this or do you want me to take care of it?
^ permalink raw reply
* Re: [Lxc-users] Bad checksums and lost packets with macvlan on dummy
From: Patrick McHardy @ 2011-03-08 14:41 UTC (permalink / raw)
To: Daniel Lezcano; +Cc: Eric Dumazet, Andrian Nord, lxc-users, Linux Netdev List
In-Reply-To: <4D6E8D80.3090907@free.fr>
Am 02.03.2011 19:33, schrieb Daniel Lezcano:
> On 03/02/2011 07:03 PM, Patrick McHardy wrote:
>> Am 02.03.2011 17:03, schrieb Daniel Lezcano:
>>> On 03/02/2011 12:03 PM, Patrick McHardy wrote:
>>>> Am 01.03.2011 21:04, schrieb Daniel Lezcano:
>>>>> On 03/01/2011 05:51 PM, Patrick McHardy wrote:
>>>>>>> Patrick, do you have any suggestions to fix this ?
>>>>>> Since the frames are only looped back locally, I suppose the easiest
>>>>>> fix would be to mark them with CHECKSUM_UNNECESSARY. Alternatively
>>>>>> we need to complete the checksum manually, similar to what
>>>>>> dev_hard_start_xmit() does.
>>>>> That sounds very simple to fix, maybe too much simple :)
>>>>>
>>>>> I did the following change:
>>>>>
>>>>> --- linux-next.orig/drivers/net/macvlan.c
>>>>> +++ linux-next/drivers/net/macvlan.c
>>>>> @@ -222,6 +222,7 @@ static int macvlan_queue_xmit(struct sk_
>>>>>
>>>>> if (vlan->mode == MACVLAN_MODE_BRIDGE) {
>>>>> const struct ethhdr *eth = (void *)skb->data;
>>>>> + skb->ip_summed = CHECKSUM_UNNECESSARY;
>>>>>
>>>>> /* send to other bridge ports directly */
>>>>> if (is_multicast_ether_addr(eth->h_dest)) {
>>>>>
>>>>>
>>>>> and that fixed the problem. Do you think it is acceptable ?
>>>> The only problem I see is if the packets are bridged to a
>>>> different networking device (or redirected using the mirred
>>>> action), in this case the checksum will not be completed.
>>>> This would be a very strange setup though and probably wouldn't
>>>> be using dummy as lower device, so I'm not sure we have to
>>>> worry about this case.
>>> I am not sure to get it, do you say the patch is correct ?
>> Its correct with a short-coming that doesn't seem to matter.
>>
>>> If my understanding is correct, the packet will be flagged
>>> CHECKSUM_UNNECESSARY only for the macvlan devices, right ?
>> Only for packets bridged between macvlan devices. A setup like
>> the following would cause problems:
>>
>> br0
>> |
>> .----------.
>> | |
>> macvlan0 macvlan1 eth0
>> | |
>> -------.-------
>> dummy0
>>
>> In this case packets sent from macvlan0 will show up on
>> eth0 with incorrect setups. However this setup doesn't
>> seem realistic to me, you would simply use eth0 instead
>> of dummy0.
>
> Ok, I understand. thanks for the clarification.
>
>>> By the way, this problem occurs for any lower device with offloading
>>> capabilities with a macvlan port in bridge mode.
>> True. This doesn't affect outgoing packets since their checksum
>> will be completed in dev_hard_start_xmit(), but it affects
>> packets bridged between macvlans.
>
> One last question. In the case of broadcast packets with maclvan in
> bridge mode.
> We will have the packets going through each macvlan port and also to the
> lower-device, right ?
> For the latter, don't we have a problem if the packet is flagged
> CHECKSUM_UNNECESSARY ?
>
> Shouldn't we restore the ip_summed field before sending through
> dev_queue_xmit ?
Yes, that seems correct in order to have dev_hard_start_xmit() complete
the checksum if necessary.
^ permalink raw reply
* [PATCH 7/8] Phonet: support active connection without pipe controller on modem
From: Rémi Denis-Courmont @ 2011-03-08 13:57 UTC (permalink / raw)
To: netdev
In-Reply-To: <201103081556.18877.remi.denis-courmont@nokia.com>
This provides support for newer ISI modems with no need for the
earlier experimental compile-time alternative choice. With this,
we can now use the same kernel and userspace with both types of
modems.
This also avoids confusing two different and incompatible state
machines, actively connected vs accepted sockets, and adds
connection response error handling (processing "SYN/RST" of sorts).
Signed-off-by: Rémi Denis-Courmont <remi.denis-courmont@nokia.com>
---
Documentation/networking/phonet.txt | 53 +++++------
net/phonet/pep.c | 172 ++++++++++++++++++++--------------
net/phonet/socket.c | 102 ++++++++-------------
3 files changed, 165 insertions(+), 162 deletions(-)
diff --git a/Documentation/networking/phonet.txt b/Documentation/networking/phonet.txt
index cacac96..3d12779 100644
--- a/Documentation/networking/phonet.txt
+++ b/Documentation/networking/phonet.txt
@@ -154,9 +154,28 @@ connections, one per accept()'d socket.
write(cfd, msg, msglen);
}
-Connections are established between two endpoints by a "third party"
-application. This means that both endpoints are passive; so connect()
-is not possible.
+Connections are traditionally established between two endpoints by a
+"third party" application. This means that both endpoints are passive.
+
+
+As of Linux kernel version 2.6.39, it is also possible to connect
+two endpoints directly, using connect() on the active side. This is
+intended to support the newer Nokia Wireless Modem API, as found in
+e.g. the Nokia Slim Modem in the ST-Ericsson U8500 platform:
+
+ struct sockaddr_spn spn;
+ int fd;
+
+ fd = socket(PF_PHONET, SOCK_SEQPACKET, PN_PROTO_PIPE);
+ memset(&spn, 0, sizeof(spn));
+ spn.spn_family = AF_PHONET;
+ spn.spn_obj = ...;
+ spn.spn_dev = ...;
+ spn.spn_resource = 0xD9;
+ connect(fd, (struct sockaddr *)&spn, sizeof(spn));
+ /* normal I/O here ... */
+ close(fd);
+
WARNING:
When polling a connected pipe socket for writability, there is an
@@ -189,17 +208,8 @@ The pipe protocol provides two socket options at the SOL_PNPIPE level:
Phonet Pipe-controller Implementation
-------------------------------------
-Phonet Pipe-controller is enabled by selecting the CONFIG_PHONET_PIPECTRLR Kconfig
-option. It is useful when communicating with those Nokia Modems which do not
-implement Pipe controller in them e.g. Nokia Slim Modem used in ST-Ericsson
-U8500 platform.
-
-The implementation is based on the Data Connection Establishment Sequence
-depicted in 'Nokia Wireless Modem API - Wireless_modem_user_guide.pdf'
-document.
-
-It allows a phonet sequenced socket (host-pep) to initiate a Pipe connection
-between itself and a remote pipe-end point (e.g. modem).
+Phonet Pipe-controller is enabled by selecting the CONFIG_PHONET_PIPECTRLR
+Kconfig option.
The implementation adds socket options at SOL_PNPIPE level:
@@ -207,21 +217,6 @@ The implementation adds socket options at SOL_PNPIPE level:
is disabled. If the value is non-zero, the pipe is enabled. If the pipe
is not (yet) connected, ENOTCONN is error is returned.
-The implementation also adds socket 'connect'. On calling the 'connect', pipe
-will be created between the source socket and the destination, and the pipe
-state will be set to PIPE_DISABLED.
-
-After a pipe has been created and enabled successfully, the Pipe data can be
-exchanged between the host-pep and remote-pep (modem).
-
-User-space would typically follow below sequence with Pipe controller:-
--socket
--bind
--setsockopt for PNPIPE_PIPE_HANDLE
--connect
--setsockopt for PNPIPE_ENCAP_IP
--setsockopt for PNPIPE_ENABLE
-
Authors
-------
diff --git a/net/phonet/pep.c b/net/phonet/pep.c
index abfb795..671effb 100644
--- a/net/phonet/pep.c
+++ b/net/phonet/pep.c
@@ -136,7 +136,6 @@ static int pep_indicate(struct sock *sk, u8 id, u8 code,
#define PAD 0x00
-#ifdef CONFIG_PHONET_PIPECTRLR
static int pipe_handler_request(struct sock *sk, u8 id, u8 code,
const void *data, int len)
{
@@ -168,11 +167,7 @@ static int pipe_handler_send_created_ind(struct sock *sk)
data, 4, GFP_ATOMIC);
}
-static int pipe_handler_send_ind(struct sock *sk, u8 id)
-{
- return pep_indicate(sk, id, PAD, NULL, 0, GFP_ATOMIC);
-}
-
+#ifdef CONFIG_PHONET_PIPECTRLR
static int pipe_handler_enable_pipe(struct sock *sk, int enable)
{
u8 id = enable ? PNS_PEP_ENABLE_REQ : PNS_PEP_DISABLE_REQ;
@@ -376,32 +371,11 @@ static int pipe_do_rcv(struct sock *sk, struct sk_buff *skb)
sk->sk_state_change(sk);
break;
-#ifdef CONFIG_PHONET_PIPECTRLR
- case PNS_PEP_DISCONNECT_RESP:
- sk->sk_state = TCP_CLOSE;
- break;
-#endif
-
case PNS_PEP_ENABLE_REQ:
/* Wait for PNS_PIPE_(ENABLED|REDIRECTED)_IND */
pep_reply(sk, skb, PN_PIPE_NO_ERROR, NULL, 0, GFP_ATOMIC);
break;
-#ifdef CONFIG_PHONET_PIPECTRLR
- case PNS_PEP_ENABLE_RESP:
- pipe_handler_send_ind(sk, PNS_PIPE_ENABLED_IND);
-
- if (!pn_flow_safe(pn->tx_fc)) {
- atomic_set(&pn->tx_credits, 1);
- sk->sk_write_space(sk);
- }
- if (sk->sk_state == TCP_ESTABLISHED)
- break; /* Nothing to do */
- sk->sk_state = TCP_ESTABLISHED;
- pipe_grant_credits(sk, GFP_ATOMIC);
- break;
-#endif
-
case PNS_PEP_RESET_REQ:
switch (hdr->state_after_reset) {
case PN_PIPE_DISABLE:
@@ -420,15 +394,6 @@ static int pipe_do_rcv(struct sock *sk, struct sk_buff *skb)
pep_reply(sk, skb, PN_PIPE_NO_ERROR, NULL, 0, GFP_ATOMIC);
break;
-#ifdef CONFIG_PHONET_PIPECTRLR
- case PNS_PEP_DISABLE_RESP:
- atomic_set(&pn->tx_credits, 0);
- pipe_handler_send_ind(sk, PNS_PIPE_DISABLED_IND);
- sk->sk_state = TCP_SYN_RECV;
- pn->rx_credits = 0;
- break;
-#endif
-
case PNS_PEP_CTRL_REQ:
if (skb_queue_len(&pn->ctrlreq_queue) >= PNPIPE_CTRLREQ_MAX) {
atomic_inc(&sk->sk_drops);
@@ -521,7 +486,6 @@ static void pipe_destruct(struct sock *sk)
skb_queue_purge(&pn->ctrlreq_queue);
}
-#ifdef CONFIG_PHONET_PIPECTRLR
static u8 pipe_negotiate_fc(const u8 *fcs, unsigned n)
{
unsigned i;
@@ -546,6 +510,8 @@ static int pep_connresp_rcv(struct sock *sk, struct sk_buff *skb)
return -EINVAL;
hdr = pnp_hdr(skb);
+ if (hdr->error_code != PN_PIPE_NO_ERROR)
+ return -ECONNREFUSED;
/* Parse sub-blocks */
n_sb = hdr->data[4];
@@ -573,14 +539,74 @@ static int pep_connresp_rcv(struct sock *sk, struct sk_buff *skb)
n_sb--;
}
- sk->sk_state = TCP_SYN_RECV;
- sk->sk_backlog_rcv = pipe_do_rcv;
- pn->rx_credits = 0;
- sk->sk_state_change(sk);
-
return pipe_handler_send_created_ind(sk);
}
-#endif
+
+/* Queue an skb to an actively connected sock.
+ * Socket lock must be held. */
+static int pipe_handler_do_rcv(struct sock *sk, struct sk_buff *skb)
+{
+ struct pep_sock *pn = pep_sk(sk);
+ struct pnpipehdr *hdr = pnp_hdr(skb);
+ int err = NET_RX_SUCCESS;
+
+ switch (hdr->message_id) {
+ case PNS_PIPE_ALIGNED_DATA:
+ __skb_pull(skb, 1);
+ /* fall through */
+ case PNS_PIPE_DATA:
+ __skb_pull(skb, 3); /* Pipe data header */
+ if (!pn_flow_safe(pn->rx_fc)) {
+ err = sock_queue_rcv_skb(sk, skb);
+ if (!err)
+ return NET_RX_SUCCESS;
+ err = NET_RX_DROP;
+ break;
+ }
+
+ if (pn->rx_credits == 0) {
+ atomic_inc(&sk->sk_drops);
+ err = NET_RX_DROP;
+ break;
+ }
+ pn->rx_credits--;
+ skb->dev = NULL;
+ skb_set_owner_r(skb, sk);
+ err = skb->len;
+ skb_queue_tail(&sk->sk_receive_queue, skb);
+ if (!sock_flag(sk, SOCK_DEAD))
+ sk->sk_data_ready(sk, err);
+ return NET_RX_SUCCESS;
+
+ case PNS_PEP_CONNECT_RESP:
+ if (sk->sk_state != TCP_SYN_SENT)
+ break;
+ if (!sock_flag(sk, SOCK_DEAD))
+ sk->sk_state_change(sk);
+ if (pep_connresp_rcv(sk, skb)) {
+ sk->sk_state = TCP_CLOSE_WAIT;
+ break;
+ }
+
+ sk->sk_state = TCP_ESTABLISHED;
+ if (!pn_flow_safe(pn->tx_fc)) {
+ atomic_set(&pn->tx_credits, 1);
+ sk->sk_write_space(sk);
+ }
+ pipe_grant_credits(sk, GFP_ATOMIC);
+ break;
+
+ case PNS_PEP_DISCONNECT_RESP:
+ /* sock should already be dead, nothing to do */
+ break;
+
+ case PNS_PEP_STATUS_IND:
+ pipe_rcv_status(sk, skb);
+ break;
+ }
+ kfree_skb(skb);
+ return err;
+}
/* Listening sock must be locked */
static struct sock *pep_find_pipe(const struct hlist_head *hlist,
@@ -649,12 +675,6 @@ static int pep_do_rcv(struct sock *sk, struct sk_buff *skb)
sk->sk_data_ready(sk, 0);
return NET_RX_SUCCESS;
-#ifdef CONFIG_PHONET_PIPECTRLR
- case PNS_PEP_CONNECT_RESP:
- pep_connresp_rcv(sk, skb);
- break;
-#endif
-
case PNS_PEP_DISCONNECT_REQ:
pep_reply(sk, skb, PN_PIPE_NO_ERROR, NULL, 0, GFP_ATOMIC);
break;
@@ -667,15 +687,19 @@ static int pep_do_rcv(struct sock *sk, struct sk_buff *skb)
case PNS_PEP_ENABLE_REQ:
case PNS_PEP_DISABLE_REQ:
/* invalid handle is not even allowed here! */
- default:
break;
+
+ default:
+ if ((1 << sk->sk_state)
+ & ~(TCPF_CLOSE|TCPF_LISTEN|TCPF_CLOSE_WAIT))
+ /* actively connected socket */
+ return pipe_handler_do_rcv(sk, skb);
}
drop:
kfree_skb(skb);
return NET_RX_SUCCESS;
}
-#ifndef CONFIG_PHONET_PIPECTRLR
static int pipe_do_remove(struct sock *sk)
{
struct pep_sock *pn = pep_sk(sk);
@@ -693,7 +717,6 @@ static int pipe_do_remove(struct sock *sk)
ph->data[0] = PAD;
return pn_skb_send(sk, skb, NULL);
}
-#endif
/* associated socket ceases to exist */
static void pep_sock_close(struct sock *sk, long timeout)
@@ -706,13 +729,12 @@ static void pep_sock_close(struct sock *sk, long timeout)
lock_sock(sk);
if ((1 << sk->sk_state) & (TCPF_SYN_RECV|TCPF_ESTABLISHED)) {
-#ifndef CONFIG_PHONET_PIPECTRLR
- /* Forcefully remove dangling Phonet pipe */
- pipe_do_remove(sk);
-#else
- /* send pep disconnect request */
- pipe_handler_request(sk, PNS_PEP_DISCONNECT_REQ, PAD, NULL, 0);
-#endif
+ if (sk->sk_backlog_rcv == pipe_do_rcv)
+ /* Forcefully remove dangling Phonet pipe */
+ pipe_do_remove(sk);
+ else
+ pipe_handler_request(sk, PNS_PEP_DISCONNECT_REQ, PAD,
+ NULL, 0);
}
sk->sk_state = TCP_CLOSE;
@@ -844,20 +866,22 @@ drop:
return newsk;
}
-#ifdef CONFIG_PHONET_PIPECTRLR
static int pep_sock_connect(struct sock *sk, struct sockaddr *addr, int len)
{
struct pep_sock *pn = pep_sk(sk);
- const struct sockaddr_pn *spn = (struct sockaddr_pn *)addr;
+ int err;
u8 data[4] = { 0 /* sub-blocks */, PAD, PAD, PAD };
- pn->pn_sk.dobject = pn_sockaddr_get_object(spn);
- pn->pn_sk.resource = pn_sockaddr_get_resource(spn);
pn->pipe_handle = 1; /* anything but INVALID_HANDLE */
- return pipe_handler_request(sk, PNS_PEP_CONNECT_REQ,
- PN_PIPE_DISABLE, data, 4);
+ err = pipe_handler_request(sk, PNS_PEP_CONNECT_REQ,
+ PN_PIPE_ENABLE, data, 4);
+ if (err) {
+ pn->pipe_handle = PN_PIPE_INVALID_HANDLE;
+ return err;
+ }
+ sk->sk_state = TCP_SYN_SENT;
+ return 0;
}
-#endif
static int pep_ioctl(struct sock *sk, int cmd, unsigned long arg)
{
@@ -890,8 +914,16 @@ static int pep_init(struct sock *sk)
sk->sk_destruct = pipe_destruct;
INIT_HLIST_HEAD(&pn->hlist);
+ pn->listener = NULL;
skb_queue_head_init(&pn->ctrlreq_queue);
+ atomic_set(&pn->tx_credits, 0);
+ pn->ifindex = 0;
+ pn->peer_type = 0;
pn->pipe_handle = PN_PIPE_INVALID_HANDLE;
+ pn->rx_credits = 0;
+ pn->rx_fc = pn->tx_fc = PN_LEGACY_FLOW_CONTROL;
+ pn->init_enable = 1;
+ pn->aligned = 0;
return 0;
}
@@ -1219,9 +1251,9 @@ static void pep_sock_unhash(struct sock *sk)
lock_sock(sk);
-#ifndef CONFIG_PHONET_PIPECTRLR
- if ((1 << sk->sk_state) & ~(TCPF_CLOSE|TCPF_LISTEN)) {
+ if (pn->listener != NULL) {
skparent = pn->listener;
+ pn->listener = NULL;
release_sock(sk);
pn = pep_sk(skparent);
@@ -1229,7 +1261,7 @@ static void pep_sock_unhash(struct sock *sk)
sk_del_node_init(sk);
sk = skparent;
}
-#endif
+
/* Unhash a listening sock only when it is closed
* and all of its active connected pipes are closed. */
if (hlist_empty(&pn->hlist))
@@ -1243,9 +1275,7 @@ static void pep_sock_unhash(struct sock *sk)
static struct proto pep_proto = {
.close = pep_sock_close,
.accept = pep_sock_accept,
-#ifdef CONFIG_PHONET_PIPECTRLR
.connect = pep_sock_connect,
-#endif
.ioctl = pep_ioctl,
.init = pep_init,
.setsockopt = pep_setsockopt,
diff --git a/net/phonet/socket.c b/net/phonet/socket.c
index 1eccfc3..b1adafa 100644
--- a/net/phonet/socket.c
+++ b/net/phonet/socket.c
@@ -225,15 +225,18 @@ static int pn_socket_autobind(struct socket *sock)
return 0; /* socket was already bound */
}
-#ifdef CONFIG_PHONET_PIPECTRLR
static int pn_socket_connect(struct socket *sock, struct sockaddr *addr,
int len, int flags)
{
struct sock *sk = sock->sk;
+ struct pn_sock *pn = pn_sk(sk);
struct sockaddr_pn *spn = (struct sockaddr_pn *)addr;
- long timeo;
+ struct task_struct *tsk = current;
+ long timeo = sock_rcvtimeo(sk, flags & O_NONBLOCK);
int err;
+ if (pn_socket_autobind(sock))
+ return -ENOBUFS;
if (len < sizeof(struct sockaddr_pn))
return -EINVAL;
if (spn->spn_family != AF_PHONET)
@@ -243,82 +246,61 @@ static int pn_socket_connect(struct socket *sock, struct sockaddr *addr,
switch (sock->state) {
case SS_UNCONNECTED:
- sk->sk_state = TCP_CLOSE;
- break;
- case SS_CONNECTING:
- switch (sk->sk_state) {
- case TCP_SYN_RECV:
- sock->state = SS_CONNECTED;
- err = -EISCONN;
- goto out;
- case TCP_CLOSE:
- err = -EALREADY;
- if (flags & O_NONBLOCK)
- goto out;
- goto wait_connect;
- }
- break;
- case SS_CONNECTED:
- switch (sk->sk_state) {
- case TCP_SYN_RECV:
+ if (sk->sk_state != TCP_CLOSE) {
err = -EISCONN;
goto out;
- case TCP_CLOSE:
- sock->state = SS_UNCONNECTED;
- break;
}
break;
- case SS_DISCONNECTING:
- case SS_FREE:
- break;
+ case SS_CONNECTING:
+ err = -EALREADY;
+ goto out;
+ default:
+ err = -EISCONN;
+ goto out;
}
- sk->sk_state = TCP_CLOSE;
- sk_stream_kill_queues(sk);
+ pn->dobject = pn_sockaddr_get_object(spn);
+ pn->resource = pn_sockaddr_get_resource(spn);
sock->state = SS_CONNECTING;
+
err = sk->sk_prot->connect(sk, addr, len);
- if (err < 0) {
+ if (err) {
sock->state = SS_UNCONNECTED;
- sk->sk_state = TCP_CLOSE;
+ pn->dobject = 0;
goto out;
}
- err = -EINPROGRESS;
-wait_connect:
- if (sk->sk_state != TCP_SYN_RECV && (flags & O_NONBLOCK))
- goto out;
-
- timeo = sock_sndtimeo(sk, flags & O_NONBLOCK);
- release_sock(sk);
+ while (sk->sk_state == TCP_SYN_SENT) {
+ DEFINE_WAIT(wait);
- err = -ERESTARTSYS;
- timeo = wait_event_interruptible_timeout(*sk_sleep(sk),
- sk->sk_state != TCP_CLOSE,
- timeo);
-
- lock_sock(sk);
- if (timeo < 0)
- goto out; /* -ERESTARTSYS */
-
- err = -ETIMEDOUT;
- if (timeo == 0 && sk->sk_state != TCP_SYN_RECV)
- goto out;
+ if (!timeo) {
+ err = -EINPROGRESS;
+ goto out;
+ }
+ if (signal_pending(tsk)) {
+ err = sock_intr_errno(timeo);
+ goto out;
+ }
- if (sk->sk_state != TCP_SYN_RECV) {
- sock->state = SS_UNCONNECTED;
- err = sock_error(sk);
- if (!err)
- err = -ECONNREFUSED;
- goto out;
+ prepare_to_wait_exclusive(sk_sleep(sk), &wait,
+ TASK_INTERRUPTIBLE);
+ release_sock(sk);
+ timeo = schedule_timeout(timeo);
+ lock_sock(sk);
+ finish_wait(sk_sleep(sk), &wait);
}
- sock->state = SS_CONNECTED;
- err = 0;
+ if ((1 << sk->sk_state) & (TCPF_SYN_RECV|TCPF_ESTABLISHED))
+ err = 0;
+ else if (sk->sk_state == TCP_CLOSE_WAIT)
+ err = -ECONNRESET;
+ else
+ err = -ECONNREFUSED;
+ sock->state = err ? SS_UNCONNECTED : SS_CONNECTED;
out:
release_sock(sk);
return err;
}
-#endif
static int pn_socket_accept(struct socket *sock, struct socket *newsock,
int flags)
@@ -486,11 +468,7 @@ const struct proto_ops phonet_stream_ops = {
.owner = THIS_MODULE,
.release = pn_socket_release,
.bind = pn_socket_bind,
-#ifdef CONFIG_PHONET_PIPECTRLR
.connect = pn_socket_connect,
-#else
- .connect = sock_no_connect,
-#endif
.socketpair = sock_no_socketpair,
.accept = pn_socket_accept,
.getname = pn_socket_getname,
--
1.7.1
^ permalink raw reply related
* [PATCH 8/8] Phonet: kill the ST-Ericsson pipe controller Kconfig
From: Rémi Denis-Courmont @ 2011-03-08 13:57 UTC (permalink / raw)
To: netdev
In-Reply-To: <201103081556.18877.remi.denis-courmont@nokia.com>
This is now a run-time choice so that a single kernel can support both
old and new generation ISI modems.
Signed-off-by: Rémi Denis-Courmont <remi.denis-courmont@nokia.com>
---
Documentation/networking/phonet.txt | 13 -------------
include/linux/phonet.h | 2 --
net/phonet/Kconfig | 12 ------------
net/phonet/pep.c | 25 -------------------------
4 files changed, 0 insertions(+), 52 deletions(-)
diff --git a/Documentation/networking/phonet.txt b/Documentation/networking/phonet.txt
index 3d12779..8100358 100644
--- a/Documentation/networking/phonet.txt
+++ b/Documentation/networking/phonet.txt
@@ -205,19 +205,6 @@ The pipe protocol provides two socket options at the SOL_PNPIPE level:
socket descriptors that are already connected or being connected.
-Phonet Pipe-controller Implementation
--------------------------------------
-
-Phonet Pipe-controller is enabled by selecting the CONFIG_PHONET_PIPECTRLR
-Kconfig option.
-
-The implementation adds socket options at SOL_PNPIPE level:
-
- PNPIPE_ENABLE accepts one integer value (int). If set to zero, the pipe
- is disabled. If the value is non-zero, the pipe is enabled. If the pipe
- is not (yet) connected, ENOTCONN is error is returned.
-
-
Authors
-------
diff --git a/include/linux/phonet.h b/include/linux/phonet.h
index 32a0965..6fb1384 100644
--- a/include/linux/phonet.h
+++ b/include/linux/phonet.h
@@ -37,8 +37,6 @@
#define PNPIPE_ENCAP 1
#define PNPIPE_IFINDEX 2
#define PNPIPE_HANDLE 3
-#define PNPIPE_ENABLE 4
-/* unused slot */
#define PNADDR_ANY 0
#define PNADDR_BROADCAST 0xFC
diff --git a/net/phonet/Kconfig b/net/phonet/Kconfig
index 0d9b8a2..6ec7d55 100644
--- a/net/phonet/Kconfig
+++ b/net/phonet/Kconfig
@@ -14,15 +14,3 @@ config PHONET
To compile this driver as a module, choose M here: the module
will be called phonet. If unsure, say N.
-
-config PHONET_PIPECTRLR
- bool "Phonet Pipe Controller (EXPERIMENTAL)"
- depends on PHONET && EXPERIMENTAL
- default N
- help
- The Pipe Controller implementation in Phonet stack to support Pipe
- data with Nokia Slim modems like WG2.5 used on ST-Ericsson U8500
- platform.
-
- This option is incompatible with older Nokia modems.
- Say N here unless you really know what you are doing.
diff --git a/net/phonet/pep.c b/net/phonet/pep.c
index 671effb..68e635f 100644
--- a/net/phonet/pep.c
+++ b/net/phonet/pep.c
@@ -167,15 +167,6 @@ static int pipe_handler_send_created_ind(struct sock *sk)
data, 4, GFP_ATOMIC);
}
-#ifdef CONFIG_PHONET_PIPECTRLR
-static int pipe_handler_enable_pipe(struct sock *sk, int enable)
-{
- u8 id = enable ? PNS_PEP_ENABLE_REQ : PNS_PEP_DISABLE_REQ;
-
- return pipe_handler_request(sk, id, PAD, NULL, 0);
-}
-#endif
-
static int pep_accept_conn(struct sock *sk, struct sk_buff *skb)
{
static const u8 data[20] = {
@@ -968,16 +959,6 @@ static int pep_setsockopt(struct sock *sk, int level, int optname,
}
goto out_norel;
-#ifdef CONFIG_PHONET_PIPECTRLR
- case PNPIPE_ENABLE:
- if ((1 << sk->sk_state) & ~(TCPF_SYN_RECV|TCPF_ESTABLISHED)) {
- err = -ENOTCONN;
- break;
- }
- err = pipe_handler_enable_pipe(sk, val);
- break;
-#endif
-
default:
err = -ENOPROTOOPT;
}
@@ -1013,12 +994,6 @@ static int pep_getsockopt(struct sock *sk, int level, int optname,
return -EINVAL;
break;
-#ifdef CONFIG_PHONET_PIPECTRLR
- case PNPIPE_ENABLE:
- val = sk->sk_state == TCP_ESTABLISHED;
- break;
-#endif
-
default:
return -ENOPROTOOPT;
}
--
1.7.1
^ permalink raw reply related
* [PATCH 3/8] Phonet: correct pipe backlog callback return values
From: Rémi Denis-Courmont @ 2011-03-08 13:57 UTC (permalink / raw)
To: netdev
In-Reply-To: <201103081556.18877.remi.denis-courmont@nokia.com>
In some cases, the Phonet pipe backlog callbacks returned negative
errno instead. In some other cases, NET_RX_DROP was returned even
though there was no buffering issue.
Signed-off-by: Rémi Denis-Courmont <remi.denis-courmont@nokia.com>
---
net/phonet/pep.c | 25 +++++++++++--------------
1 files changed, 11 insertions(+), 14 deletions(-)
diff --git a/net/phonet/pep.c b/net/phonet/pep.c
index 875e86c..40952c7 100644
--- a/net/phonet/pep.c
+++ b/net/phonet/pep.c
@@ -522,7 +522,8 @@ static int pipe_do_rcv(struct sock *sk, struct sk_buff *skb)
if (!pn_flow_safe(pn->rx_fc)) {
err = sock_queue_rcv_skb(sk, skb);
if (!err)
- return 0;
+ return NET_RX_SUCCESS;
+ err = -ENOBUFS;
break;
}
@@ -575,7 +576,7 @@ static int pipe_do_rcv(struct sock *sk, struct sk_buff *skb)
}
out:
kfree_skb(skb);
- return err;
+ return (err == -ENOBUFS) ? NET_RX_DROP : NET_RX_SUCCESS;
queue:
skb->dev = NULL;
@@ -584,7 +585,7 @@ queue:
skb_queue_tail(queue, skb);
if (!sock_flag(sk, SOCK_DEAD))
sk->sk_data_ready(sk, err);
- return 0;
+ return NET_RX_SUCCESS;
}
/* Destroy connected sock. */
@@ -686,11 +687,6 @@ static int pep_connreq_rcv(struct sock *sk, struct sk_buff *skb)
}
peer_type = hdr->other_pep_type << 8;
- if (unlikely(sk->sk_state != TCP_LISTEN) || sk_acceptq_is_full(sk)) {
- pep_reject_conn(sk, skb, PN_PIPE_ERR_PEP_IN_USE);
- return -ENOBUFS;
- }
-
/* Parse sub-blocks (options) */
n_sb = hdr->data[4];
while (n_sb > 0) {
@@ -790,7 +786,6 @@ static int pep_do_rcv(struct sock *sk, struct sk_buff *skb)
struct sock *sknode;
struct pnpipehdr *hdr;
struct sockaddr_pn dst;
- int err = NET_RX_SUCCESS;
u8 pipe_handle;
if (!pskb_may_pull(skb, sizeof(*hdr)))
@@ -814,18 +809,20 @@ static int pep_do_rcv(struct sock *sk, struct sk_buff *skb)
sock_put(sknode);
if (net_ratelimit())
printk(KERN_WARNING"Phonet unconnected PEP ignored");
- err = NET_RX_DROP;
goto drop;
}
switch (hdr->message_id) {
case PNS_PEP_CONNECT_REQ:
- err = pep_connreq_rcv(sk, skb);
+ if (sk->sk_state == TCP_LISTEN && !sk_acceptq_is_full(sk))
+ pep_connreq_rcv(sk, skb);
+ else
+ pep_reject_conn(sk, skb, PN_PIPE_ERR_PEP_IN_USE);
break;
#ifdef CONFIG_PHONET_PIPECTRLR
case PNS_PEP_CONNECT_RESP:
- err = pep_connresp_rcv(sk, skb);
+ pep_connresp_rcv(sk, skb);
break;
#endif
@@ -842,11 +839,11 @@ static int pep_do_rcv(struct sock *sk, struct sk_buff *skb)
case PNS_PEP_DISABLE_REQ:
/* invalid handle is not even allowed here! */
default:
- err = NET_RX_DROP;
+ break;
}
drop:
kfree_skb(skb);
- return err;
+ return NET_RX_SUCCESS;
}
#ifndef CONFIG_PHONET_PIPECTRLR
--
1.7.1
^ permalink raw reply related
* [PATCH 4/8] Phonet: factor common code to send control messages
From: Rémi Denis-Courmont @ 2011-03-08 13:57 UTC (permalink / raw)
To: netdev
In-Reply-To: <201103081556.18877.remi.denis-courmont@nokia.com>
Signed-off-by: Rémi Denis-Courmont <remi.denis-courmont@nokia.com>
---
net/phonet/pep.c | 225 ++++++++++++++++++------------------------------------
1 files changed, 73 insertions(+), 152 deletions(-)
diff --git a/net/phonet/pep.c b/net/phonet/pep.c
index 40952c7..610794a 100644
--- a/net/phonet/pep.c
+++ b/net/phonet/pep.c
@@ -77,24 +77,34 @@ static unsigned char *pep_get_sb(struct sk_buff *skb, u8 *ptype, u8 *plen,
return data;
}
-static int pep_reply(struct sock *sk, struct sk_buff *oskb,
- u8 code, const void *data, int len, gfp_t priority)
+static struct sk_buff *pep_alloc_skb(struct sock *sk, const void *payload,
+ int len, gfp_t priority)
+{
+ struct sk_buff *skb = alloc_skb(MAX_PNPIPE_HEADER + len, priority);
+ if (!skb)
+ return NULL;
+ skb_set_owner_w(skb, sk);
+
+ skb_reserve(skb, MAX_PNPIPE_HEADER);
+ __skb_put(skb, len);
+ skb_copy_to_linear_data(skb, payload, len);
+ __skb_push(skb, sizeof(struct pnpipehdr));
+ skb_reset_transport_header(skb);
+ return skb;
+}
+
+static int pep_reply(struct sock *sk, struct sk_buff *oskb, u8 code,
+ const void *data, int len, gfp_t priority)
{
const struct pnpipehdr *oph = pnp_hdr(oskb);
struct pnpipehdr *ph;
struct sk_buff *skb;
struct sockaddr_pn peer;
- skb = alloc_skb(MAX_PNPIPE_HEADER + len, priority);
+ skb = pep_alloc_skb(sk, data, len, priority);
if (!skb)
return -ENOMEM;
- skb_set_owner_w(skb, sk);
- skb_reserve(skb, MAX_PNPIPE_HEADER);
- __skb_put(skb, len);
- skb_copy_to_linear_data(skb, data, len);
- __skb_push(skb, sizeof(*ph));
- skb_reset_transport_header(skb);
ph = pnp_hdr(skb);
ph->utid = oph->utid;
ph->message_id = oph->message_id + 1; /* REQ -> RESP */
@@ -105,135 +115,69 @@ static int pep_reply(struct sock *sk, struct sk_buff *oskb,
return pn_skb_send(sk, skb, &peer);
}
-#define PAD 0x00
-
-#ifdef CONFIG_PHONET_PIPECTRLR
-static int pipe_handler_send_req(struct sock *sk, u8 msg_id, gfp_t priority)
+static int pep_indicate(struct sock *sk, u8 id, u8 code,
+ const void *data, int len, gfp_t priority)
{
- int len;
+ struct pep_sock *pn = pep_sk(sk);
struct pnpipehdr *ph;
struct sk_buff *skb;
- struct pep_sock *pn = pep_sk(sk);
-
- static const u8 data[4] = {
- PAD, PAD, PAD, PAD,
- };
- switch (msg_id) {
- case PNS_PEP_CONNECT_REQ:
- len = sizeof(data);
- break;
-
- case PNS_PEP_DISCONNECT_REQ:
- case PNS_PEP_ENABLE_REQ:
- case PNS_PEP_DISABLE_REQ:
- len = 0;
- break;
-
- default:
- return -EINVAL;
- }
-
- skb = alloc_skb(MAX_PNPIPE_HEADER + len, priority);
+ skb = pep_alloc_skb(sk, data, len, priority);
if (!skb)
return -ENOMEM;
- skb_set_owner_w(skb, sk);
- skb_reserve(skb, MAX_PNPIPE_HEADER);
- if (len) {
- __skb_put(skb, len);
- skb_copy_to_linear_data(skb, data, len);
- }
- __skb_push(skb, sizeof(*ph));
- skb_reset_transport_header(skb);
ph = pnp_hdr(skb);
- ph->utid = msg_id; /* whatever */
- ph->message_id = msg_id;
+ ph->utid = 0;
+ ph->message_id = id;
ph->pipe_handle = pn->pipe_handle;
- ph->error_code = PN_PIPE_NO_ERROR;
-
+ ph->data[0] = code;
return pn_skb_send(sk, skb, NULL);
}
-static int pipe_handler_send_created_ind(struct sock *sk, u8 msg_id)
+#define PAD 0x00
+
+#ifdef CONFIG_PHONET_PIPECTRLR
+static int pipe_handler_request(struct sock *sk, u8 id, u8 code,
+ const void *data, int len)
{
- int err_code;
+ struct pep_sock *pn = pep_sk(sk);
struct pnpipehdr *ph;
struct sk_buff *skb;
- struct pep_sock *pn = pep_sk(sk);
- static u8 data[4] = {
- 0x03, 0x04,
- };
- data[2] = pn->tx_fc;
- data[3] = pn->rx_fc;
-
- /*
- * actually, below is number of sub-blocks and not error code.
- * Pipe_created_ind message format does not have any
- * error code field. However, the Phonet stack will always send
- * an error code as part of pnpipehdr. So, use that err_code to
- * specify the number of sub-blocks.
- */
- err_code = 0x01;
-
- skb = alloc_skb(MAX_PNPIPE_HEADER + sizeof(data), GFP_ATOMIC);
+ skb = pep_alloc_skb(sk, data, len, GFP_KERNEL);
if (!skb)
return -ENOMEM;
- skb_set_owner_w(skb, sk);
- skb_reserve(skb, MAX_PNPIPE_HEADER);
- __skb_put(skb, sizeof(data));
- skb_copy_to_linear_data(skb, data, sizeof(data));
- __skb_push(skb, sizeof(*ph));
- skb_reset_transport_header(skb);
ph = pnp_hdr(skb);
- ph->utid = 0;
- ph->message_id = msg_id;
+ ph->utid = id; /* whatever */
+ ph->message_id = id;
ph->pipe_handle = pn->pipe_handle;
- ph->error_code = err_code;
-
+ ph->data[0] = code;
return pn_skb_send(sk, skb, NULL);
}
-static int pipe_handler_send_ind(struct sock *sk, u8 msg_id)
+static int pipe_handler_send_created_ind(struct sock *sk)
{
- int err_code;
- struct pnpipehdr *ph;
- struct sk_buff *skb;
struct pep_sock *pn = pep_sk(sk);
+ u8 data[4] = {
+ PN_PIPE_SB_NEGOTIATED_FC, pep_sb_size(2),
+ pn->tx_fc, pn->rx_fc,
+ };
- /*
- * actually, below is a filler.
- * Pipe_enabled/disabled_ind message format does not have any
- * error code field. However, the Phonet stack will always send
- * an error code as part of pnpipehdr. So, use that err_code to
- * specify the filler value.
- */
- err_code = 0x0;
-
- skb = alloc_skb(MAX_PNPIPE_HEADER, GFP_ATOMIC);
- if (!skb)
- return -ENOMEM;
- skb_set_owner_w(skb, sk);
-
- skb_reserve(skb, MAX_PNPIPE_HEADER);
- __skb_push(skb, sizeof(*ph));
- skb_reset_transport_header(skb);
- ph = pnp_hdr(skb);
- ph->utid = 0;
- ph->message_id = msg_id;
- ph->pipe_handle = pn->pipe_handle;
- ph->error_code = err_code;
+ return pep_indicate(sk, PNS_PIPE_CREATED_IND, 1 /* sub-blocks */,
+ data, 4, GFP_ATOMIC);
+}
- return pn_skb_send(sk, skb, NULL);
+static int pipe_handler_send_ind(struct sock *sk, u8 id)
+{
+ return pep_indicate(sk, id, PAD, NULL, 0, GFP_ATOMIC);
}
static int pipe_handler_enable_pipe(struct sock *sk, int enable)
{
u8 id = enable ? PNS_PEP_ENABLE_REQ : PNS_PEP_DISABLE_REQ;
- return pipe_handler_send_req(sk, id, GFP_KERNEL);
+ return pipe_handler_request(sk, id, PAD, NULL, 0);
}
#endif
@@ -274,23 +218,21 @@ static int pep_ctrlreq_error(struct sock *sk, struct sk_buff *oskb, u8 code,
struct sk_buff *skb;
struct pnpipehdr *ph;
struct sockaddr_pn dst;
+ u8 data[4] = {
+ oph->data[0], /* PEP type */
+ code, /* error code, at an unusual offset */
+ PAD, PAD,
+ };
- skb = alloc_skb(MAX_PNPIPE_HEADER + 4, priority);
+ skb = pep_alloc_skb(sk, data, 4, priority);
if (!skb)
return -ENOMEM;
- skb_set_owner_w(skb, sk);
-
- skb_reserve(skb, MAX_PHONET_HEADER);
- ph = (struct pnpipehdr *)skb_put(skb, sizeof(*ph) + 4);
+ ph = pnp_hdr(skb);
ph->utid = oph->utid;
ph->message_id = PNS_PEP_CTRL_RESP;
ph->pipe_handle = oph->pipe_handle;
ph->data[0] = oph->data[1]; /* CTRL id */
- ph->data[1] = oph->data[0]; /* PEP type */
- ph->data[2] = code; /* error code, at an usual offset */
- ph->data[3] = PAD;
- ph->data[4] = PAD;
pn_skb_get_src_sockaddr(oskb, &dst);
return pn_skb_send(sk, skb, &dst);
@@ -298,34 +240,15 @@ static int pep_ctrlreq_error(struct sock *sk, struct sk_buff *oskb, u8 code,
static int pipe_snd_status(struct sock *sk, u8 type, u8 status, gfp_t priority)
{
- struct pep_sock *pn = pep_sk(sk);
- struct pnpipehdr *ph;
- struct sk_buff *skb;
+ u8 data[4] = { type, PAD, PAD, status };
- skb = alloc_skb(MAX_PNPIPE_HEADER + 4, priority);
- if (!skb)
- return -ENOMEM;
- skb_set_owner_w(skb, sk);
-
- skb_reserve(skb, MAX_PNPIPE_HEADER + 4);
- __skb_push(skb, sizeof(*ph) + 4);
- skb_reset_transport_header(skb);
- ph = pnp_hdr(skb);
- ph->utid = 0;
- ph->message_id = PNS_PEP_STATUS_IND;
- ph->pipe_handle = pn->pipe_handle;
- ph->pep_type = PN_PEP_TYPE_COMMON;
- ph->data[1] = type;
- ph->data[2] = PAD;
- ph->data[3] = PAD;
- ph->data[4] = status;
-
- return pn_skb_send(sk, skb, NULL);
+ return pep_indicate(sk, PNS_PEP_STATUS_IND, PN_PEP_TYPE_COMMON,
+ data, 4, priority);
}
/* Send our RX flow control information to the sender.
* Socket must be locked. */
-static void pipe_grant_credits(struct sock *sk)
+static void pipe_grant_credits(struct sock *sk, gfp_t priority)
{
struct pep_sock *pn = pep_sk(sk);
@@ -335,16 +258,16 @@ static void pipe_grant_credits(struct sock *sk)
case PN_LEGACY_FLOW_CONTROL: /* TODO */
break;
case PN_ONE_CREDIT_FLOW_CONTROL:
- pipe_snd_status(sk, PN_PEP_IND_FLOW_CONTROL,
- PEP_IND_READY, GFP_ATOMIC);
- pn->rx_credits = 1;
+ if (pipe_snd_status(sk, PN_PEP_IND_FLOW_CONTROL,
+ PEP_IND_READY, priority) == 0)
+ pn->rx_credits = 1;
break;
case PN_MULTI_CREDIT_FLOW_CONTROL:
if ((pn->rx_credits + CREDITS_THR) > CREDITS_MAX)
break;
if (pipe_snd_status(sk, PN_PEP_IND_ID_MCFC_GRANT_CREDITS,
CREDITS_MAX - pn->rx_credits,
- GFP_ATOMIC) == 0)
+ priority) == 0)
pn->rx_credits = CREDITS_MAX;
break;
}
@@ -474,7 +397,7 @@ static int pipe_do_rcv(struct sock *sk, struct sk_buff *skb)
if (sk->sk_state == TCP_ESTABLISHED)
break; /* Nothing to do */
sk->sk_state = TCP_ESTABLISHED;
- pipe_grant_credits(sk);
+ pipe_grant_credits(sk, GFP_ATOMIC);
break;
#endif
@@ -561,7 +484,7 @@ static int pipe_do_rcv(struct sock *sk, struct sk_buff *skb)
if (sk->sk_state == TCP_ESTABLISHED)
break; /* Nothing to do */
sk->sk_state = TCP_ESTABLISHED;
- pipe_grant_credits(sk);
+ pipe_grant_credits(sk, GFP_ATOMIC);
break;
case PNS_PIPE_DISABLED_IND:
@@ -655,7 +578,7 @@ static int pep_connresp_rcv(struct sock *sk, struct sk_buff *skb)
pn->rx_credits = 0;
sk->sk_state_change(sk);
- return pipe_handler_send_created_ind(sk, PNS_PIPE_CREATED_IND);
+ return pipe_handler_send_created_ind(sk);
}
#endif
@@ -853,19 +776,15 @@ static int pipe_do_remove(struct sock *sk)
struct pnpipehdr *ph;
struct sk_buff *skb;
- skb = alloc_skb(MAX_PNPIPE_HEADER, GFP_KERNEL);
+ skb = pep_alloc_skb(sk, NULL, 0, GFP_KERNEL);
if (!skb)
return -ENOMEM;
- skb_reserve(skb, MAX_PNPIPE_HEADER);
- __skb_push(skb, sizeof(*ph));
- skb_reset_transport_header(skb);
ph = pnp_hdr(skb);
ph->utid = 0;
ph->message_id = PNS_PIPE_REMOVE_REQ;
ph->pipe_handle = pn->pipe_handle;
ph->data[0] = PAD;
-
return pn_skb_send(sk, skb, NULL);
}
#endif
@@ -894,7 +813,7 @@ static void pep_sock_close(struct sock *sk, long timeout)
pipe_do_remove(sk);
#else
/* send pep disconnect request */
- pipe_handler_send_req(sk, PNS_PEP_DISCONNECT_REQ, GFP_KERNEL);
+ pipe_handler_request(sk, PNS_PEP_DISCONNECT_REQ, PAD, NULL, 0);
sk->sk_state = TCP_CLOSE;
#endif
}
@@ -980,10 +899,12 @@ static int pep_sock_connect(struct sock *sk, struct sockaddr *addr, int len)
{
struct pep_sock *pn = pep_sk(sk);
const struct sockaddr_pn *spn = (struct sockaddr_pn *)addr;
+ u8 data[4] = { 0 /* sub-blocks */, PAD, PAD, PAD };
pn->pn_sk.dobject = pn_sockaddr_get_object(spn);
pn->pn_sk.resource = pn_sockaddr_get_resource(spn);
- return pipe_handler_send_req(sk, PNS_PEP_CONNECT_REQ, GFP_KERNEL);
+ return pipe_handler_request(sk, PNS_PEP_CONNECT_REQ,
+ PN_PIPE_DISABLE, data, 4);
}
#endif
@@ -1280,7 +1201,7 @@ struct sk_buff *pep_read(struct sock *sk)
struct sk_buff *skb = skb_dequeue(&sk->sk_receive_queue);
if (sk->sk_state == TCP_ESTABLISHED)
- pipe_grant_credits(sk);
+ pipe_grant_credits(sk, GFP_ATOMIC);
return skb;
}
@@ -1325,7 +1246,7 @@ static int pep_recvmsg(struct kiocb *iocb, struct sock *sk,
}
if (sk->sk_state == TCP_ESTABLISHED)
- pipe_grant_credits(sk);
+ pipe_grant_credits(sk, GFP_KERNEL);
release_sock(sk);
copy:
msg->msg_flags |= MSG_EOR;
--
1.7.1
^ permalink raw reply related
* [PATCH 6/8] Phonet: provide pipe socket option to retrieve the pipe identifier
From: Rémi Denis-Courmont @ 2011-03-08 13:57 UTC (permalink / raw)
To: netdev
In-Reply-To: <201103081556.18877.remi.denis-courmont@nokia.com>
User-space sometimes needs this information.
This replaces the buggy code in CONFIG_PHONET_PIPECTRLR case:
this was never used and error cases not handled correctly.
Signed-off-by: Rémi Denis-Courmont <remi.denis-courmont@nokia.com>
---
Documentation/networking/phonet.txt | 7 ++++---
include/linux/phonet.h | 2 +-
net/phonet/pep.c | 15 +++++++--------
3 files changed, 12 insertions(+), 12 deletions(-)
diff --git a/Documentation/networking/phonet.txt b/Documentation/networking/phonet.txt
index 24ad2ad..cacac96 100644
--- a/Documentation/networking/phonet.txt
+++ b/Documentation/networking/phonet.txt
@@ -181,6 +181,10 @@ The pipe protocol provides two socket options at the SOL_PNPIPE level:
interface index of the network interface created by PNPIPE_ENCAP,
or zero if encapsulation is off.
+ PNPIPE_HANDLE is a read-only integer value. It contains the underlying
+ identifier ("pipe handle") of the pipe. This is only defined for
+ socket descriptors that are already connected or being connected.
+
Phonet Pipe-controller Implementation
-------------------------------------
@@ -199,9 +203,6 @@ between itself and a remote pipe-end point (e.g. modem).
The implementation adds socket options at SOL_PNPIPE level:
- PNPIPE_PIPE_HANDLE
- It accepts an integer argument for setting value of pipe handle.
-
PNPIPE_ENABLE accepts one integer value (int). If set to zero, the pipe
is disabled. If the value is non-zero, the pipe is enabled. If the pipe
is not (yet) connected, ENOTCONN is error is returned.
diff --git a/include/linux/phonet.h b/include/linux/phonet.h
index 26c8df7..32a0965 100644
--- a/include/linux/phonet.h
+++ b/include/linux/phonet.h
@@ -36,7 +36,7 @@
/* Socket options for SOL_PNPIPE level */
#define PNPIPE_ENCAP 1
#define PNPIPE_IFINDEX 2
-#define PNPIPE_PIPE_HANDLE 3
+#define PNPIPE_HANDLE 3
#define PNPIPE_ENABLE 4
/* unused slot */
diff --git a/net/phonet/pep.c b/net/phonet/pep.c
index c0fab4c..abfb795 100644
--- a/net/phonet/pep.c
+++ b/net/phonet/pep.c
@@ -853,6 +853,7 @@ static int pep_sock_connect(struct sock *sk, struct sockaddr *addr, int len)
pn->pn_sk.dobject = pn_sockaddr_get_object(spn);
pn->pn_sk.resource = pn_sockaddr_get_resource(spn);
+ pn->pipe_handle = 1; /* anything but INVALID_HANDLE */
return pipe_handler_request(sk, PNS_PEP_CONNECT_REQ,
PN_PIPE_DISABLE, data, 4);
}
@@ -909,14 +910,6 @@ static int pep_setsockopt(struct sock *sk, int level, int optname,
lock_sock(sk);
switch (optname) {
-#ifdef CONFIG_PHONET_PIPECTRLR
- case PNPIPE_PIPE_HANDLE:
- if (val) {
- pn->pipe_handle = val;
- break;
- }
-#endif
-
case PNPIPE_ENCAP:
if (val && val != PNPIPE_ENCAP_IP) {
err = -EINVAL;
@@ -982,6 +975,12 @@ static int pep_getsockopt(struct sock *sk, int level, int optname,
val = pn->ifindex;
break;
+ case PNPIPE_HANDLE:
+ val = pn->pipe_handle;
+ if (val == PN_PIPE_INVALID_HANDLE)
+ return -EINVAL;
+ break;
+
#ifdef CONFIG_PHONET_PIPECTRLR
case PNPIPE_ENABLE:
val = sk->sk_state == TCP_ESTABLISHED;
--
1.7.1
^ permalink raw reply related
* [PATCH 5/8] Phonet: allocate sock from accept syscall rather than soft IRQ
From: Rémi Denis-Courmont @ 2011-03-08 13:57 UTC (permalink / raw)
To: netdev
In-Reply-To: <201103081556.18877.remi.denis-courmont@nokia.com>
This moves most of the accept logic to process context, and simplifies
the code a bit at the same time.
Signed-off-by: Rémi Denis-Courmont <remi.denis-courmont@nokia.com>
---
include/net/phonet/pep.h | 1 -
net/phonet/pep.c | 284 +++++++++++++++++++---------------------------
net/phonet/socket.c | 10 +-
3 files changed, 121 insertions(+), 174 deletions(-)
diff --git a/include/net/phonet/pep.h b/include/net/phonet/pep.h
index 38eed1b..b669fe6 100644
--- a/include/net/phonet/pep.h
+++ b/include/net/phonet/pep.h
@@ -28,7 +28,6 @@ struct pep_sock {
/* XXX: union-ify listening vs connected stuff ? */
/* Listening socket stuff: */
- struct hlist_head ackq;
struct hlist_head hlist;
/* Connected socket stuff: */
diff --git a/net/phonet/pep.c b/net/phonet/pep.c
index 610794a..c0fab4c 100644
--- a/net/phonet/pep.c
+++ b/net/phonet/pep.c
@@ -42,7 +42,7 @@
* TCP_ESTABLISHED connected pipe in enabled state
*
* pep_sock locking:
- * - sk_state, ackq, hlist: sock lock needed
+ * - sk_state, hlist: sock lock needed
* - listener: read only
* - pipe_handle: read only
*/
@@ -202,11 +202,12 @@ static int pep_accept_conn(struct sock *sk, struct sk_buff *skb)
GFP_KERNEL);
}
-static int pep_reject_conn(struct sock *sk, struct sk_buff *skb, u8 code)
+static int pep_reject_conn(struct sock *sk, struct sk_buff *skb, u8 code,
+ gfp_t priority)
{
static const u8 data[4] = { PAD, PAD, PAD, 0 /* sub-blocks */ };
WARN_ON(code == PN_PIPE_NO_ERROR);
- return pep_reply(sk, skb, code, data, sizeof(data), GFP_ATOMIC);
+ return pep_reply(sk, skb, code, data, sizeof(data), priority);
}
/* Control requests are not sent by the pipe service and have a specific
@@ -365,7 +366,7 @@ static int pipe_do_rcv(struct sock *sk, struct sk_buff *skb)
switch (hdr->message_id) {
case PNS_PEP_CONNECT_REQ:
- pep_reject_conn(sk, skb, PN_PIPE_ERR_PEP_IN_USE);
+ pep_reject_conn(sk, skb, PN_PIPE_ERR_PEP_IN_USE, GFP_ATOMIC);
break;
case PNS_PEP_DISCONNECT_REQ:
@@ -574,7 +575,6 @@ static int pep_connresp_rcv(struct sock *sk, struct sk_buff *skb)
sk->sk_state = TCP_SYN_RECV;
sk->sk_backlog_rcv = pipe_do_rcv;
- sk->sk_destruct = pipe_destruct;
pn->rx_credits = 0;
sk->sk_state_change(sk);
@@ -582,96 +582,6 @@ static int pep_connresp_rcv(struct sock *sk, struct sk_buff *skb)
}
#endif
-static int pep_connreq_rcv(struct sock *sk, struct sk_buff *skb)
-{
- struct sock *newsk;
- struct pep_sock *newpn, *pn = pep_sk(sk);
- struct pnpipehdr *hdr;
- struct sockaddr_pn dst, src;
- u16 peer_type;
- u8 pipe_handle, enabled, n_sb;
- u8 aligned = 0;
-
- if (!pskb_pull(skb, sizeof(*hdr) + 4))
- return -EINVAL;
-
- hdr = pnp_hdr(skb);
- pipe_handle = hdr->pipe_handle;
- switch (hdr->state_after_connect) {
- case PN_PIPE_DISABLE:
- enabled = 0;
- break;
- case PN_PIPE_ENABLE:
- enabled = 1;
- break;
- default:
- pep_reject_conn(sk, skb, PN_PIPE_ERR_INVALID_PARAM);
- return -EINVAL;
- }
- peer_type = hdr->other_pep_type << 8;
-
- /* Parse sub-blocks (options) */
- n_sb = hdr->data[4];
- while (n_sb > 0) {
- u8 type, buf[1], len = sizeof(buf);
- const u8 *data = pep_get_sb(skb, &type, &len, buf);
-
- if (data == NULL)
- return -EINVAL;
- switch (type) {
- case PN_PIPE_SB_CONNECT_REQ_PEP_SUB_TYPE:
- if (len < 1)
- return -EINVAL;
- peer_type = (peer_type & 0xff00) | data[0];
- break;
- case PN_PIPE_SB_ALIGNED_DATA:
- aligned = data[0] != 0;
- break;
- }
- n_sb--;
- }
-
- skb = skb_clone(skb, GFP_ATOMIC);
- if (!skb)
- return -ENOMEM;
-
- /* Create a new to-be-accepted sock */
- newsk = sk_alloc(sock_net(sk), PF_PHONET, GFP_ATOMIC, sk->sk_prot);
- if (!newsk) {
- kfree_skb(skb);
- return -ENOMEM;
- }
- sock_init_data(NULL, newsk);
- newsk->sk_state = TCP_SYN_RECV;
- newsk->sk_backlog_rcv = pipe_do_rcv;
- newsk->sk_protocol = sk->sk_protocol;
- newsk->sk_destruct = pipe_destruct;
-
- newpn = pep_sk(newsk);
- pn_skb_get_dst_sockaddr(skb, &dst);
- pn_skb_get_src_sockaddr(skb, &src);
- newpn->pn_sk.sobject = pn_sockaddr_get_object(&dst);
- newpn->pn_sk.dobject = pn_sockaddr_get_object(&src);
- newpn->pn_sk.resource = pn_sockaddr_get_resource(&dst);
- skb_queue_head_init(&newpn->ctrlreq_queue);
- newpn->pipe_handle = pipe_handle;
- atomic_set(&newpn->tx_credits, 0);
- newpn->peer_type = peer_type;
- newpn->rx_credits = 0;
- newpn->rx_fc = newpn->tx_fc = PN_LEGACY_FLOW_CONTROL;
- newpn->init_enable = enabled;
- newpn->aligned = aligned;
-
- BUG_ON(!skb_queue_empty(&newsk->sk_receive_queue));
- skb_queue_head(&newsk->sk_receive_queue, skb);
- if (!sock_flag(sk, SOCK_DEAD))
- sk->sk_data_ready(sk, 0);
-
- sk_acceptq_added(sk);
- sk_add_node(newsk, &pn->ackq);
- return 0;
-}
-
/* Listening sock must be locked */
static struct sock *pep_find_pipe(const struct hlist_head *hlist,
const struct sockaddr_pn *dst,
@@ -726,22 +636,18 @@ static int pep_do_rcv(struct sock *sk, struct sk_buff *skb)
if (sknode)
return sk_receive_skb(sknode, skb, 1);
- /* Look for a pipe handle pending accept */
- sknode = pep_find_pipe(&pn->ackq, &dst, pipe_handle);
- if (sknode) {
- sock_put(sknode);
- if (net_ratelimit())
- printk(KERN_WARNING"Phonet unconnected PEP ignored");
- goto drop;
- }
-
switch (hdr->message_id) {
case PNS_PEP_CONNECT_REQ:
- if (sk->sk_state == TCP_LISTEN && !sk_acceptq_is_full(sk))
- pep_connreq_rcv(sk, skb);
- else
- pep_reject_conn(sk, skb, PN_PIPE_ERR_PEP_IN_USE);
- break;
+ if (sk->sk_state != TCP_LISTEN || sk_acceptq_is_full(sk)) {
+ pep_reject_conn(sk, skb, PN_PIPE_ERR_PEP_IN_USE,
+ GFP_ATOMIC);
+ break;
+ }
+ skb_queue_head(&sk->sk_receive_queue, skb);
+ sk_acceptq_added(sk);
+ if (!sock_flag(sk, SOCK_DEAD))
+ sk->sk_data_ready(sk, 0);
+ return NET_RX_SUCCESS;
#ifdef CONFIG_PHONET_PIPECTRLR
case PNS_PEP_CONNECT_RESP:
@@ -799,24 +705,16 @@ static void pep_sock_close(struct sock *sk, long timeout)
sk_common_release(sk);
lock_sock(sk);
- if (sk->sk_state == TCP_LISTEN) {
- /* Destroy the listen queue */
- struct sock *sknode;
- struct hlist_node *p, *n;
-
- sk_for_each_safe(sknode, p, n, &pn->ackq)
- sk_del_node_init(sknode);
- sk->sk_state = TCP_CLOSE;
- } else if ((1 << sk->sk_state) & (TCPF_SYN_RECV|TCPF_ESTABLISHED)) {
+ if ((1 << sk->sk_state) & (TCPF_SYN_RECV|TCPF_ESTABLISHED)) {
#ifndef CONFIG_PHONET_PIPECTRLR
/* Forcefully remove dangling Phonet pipe */
pipe_do_remove(sk);
#else
/* send pep disconnect request */
pipe_handler_request(sk, PNS_PEP_DISCONNECT_REQ, PAD, NULL, 0);
- sk->sk_state = TCP_CLOSE;
#endif
}
+ sk->sk_state = TCP_CLOSE;
ifindex = pn->ifindex;
pn->ifindex = 0;
@@ -827,69 +725,121 @@ static void pep_sock_close(struct sock *sk, long timeout)
sock_put(sk);
}
-static int pep_wait_connreq(struct sock *sk, int noblock)
+static struct sock *pep_sock_accept(struct sock *sk, int flags, int *errp)
{
- struct task_struct *tsk = current;
- struct pep_sock *pn = pep_sk(sk);
- long timeo = sock_rcvtimeo(sk, noblock);
-
- for (;;) {
- DEFINE_WAIT(wait);
+ struct pep_sock *pn = pep_sk(sk), *newpn;
+ struct sock *newsk = NULL;
+ struct sk_buff *skb;
+ struct pnpipehdr *hdr;
+ struct sockaddr_pn dst, src;
+ int err;
+ u16 peer_type;
+ u8 pipe_handle, enabled, n_sb;
+ u8 aligned = 0;
- if (sk->sk_state != TCP_LISTEN)
- return -EINVAL;
- if (!hlist_empty(&pn->ackq))
- break;
- if (!timeo)
- return -EWOULDBLOCK;
- if (signal_pending(tsk))
- return sock_intr_errno(timeo);
+ skb = skb_recv_datagram(sk, 0, flags & O_NONBLOCK, errp);
+ if (!skb)
+ return NULL;
- prepare_to_wait_exclusive(sk_sleep(sk), &wait,
- TASK_INTERRUPTIBLE);
- release_sock(sk);
- timeo = schedule_timeout(timeo);
- lock_sock(sk);
- finish_wait(sk_sleep(sk), &wait);
+ lock_sock(sk);
+ if (sk->sk_state != TCP_LISTEN) {
+ err = -EINVAL;
+ goto drop;
}
+ sk_acceptq_removed(sk);
- return 0;
-}
+ err = -EPROTO;
+ if (!pskb_may_pull(skb, sizeof(*hdr) + 4))
+ goto drop;
-static struct sock *pep_sock_accept(struct sock *sk, int flags, int *errp)
-{
- struct pep_sock *pn = pep_sk(sk);
- struct sock *newsk = NULL;
- struct sk_buff *oskb;
- int err;
+ hdr = pnp_hdr(skb);
+ pipe_handle = hdr->pipe_handle;
+ switch (hdr->state_after_connect) {
+ case PN_PIPE_DISABLE:
+ enabled = 0;
+ break;
+ case PN_PIPE_ENABLE:
+ enabled = 1;
+ break;
+ default:
+ pep_reject_conn(sk, skb, PN_PIPE_ERR_INVALID_PARAM,
+ GFP_KERNEL);
+ goto drop;
+ }
+ peer_type = hdr->other_pep_type << 8;
- lock_sock(sk);
- err = pep_wait_connreq(sk, flags & O_NONBLOCK);
- if (err)
- goto out;
+ /* Parse sub-blocks (options) */
+ n_sb = hdr->data[4];
+ while (n_sb > 0) {
+ u8 type, buf[1], len = sizeof(buf);
+ const u8 *data = pep_get_sb(skb, &type, &len, buf);
- newsk = __sk_head(&pn->ackq);
+ if (data == NULL)
+ goto drop;
+ switch (type) {
+ case PN_PIPE_SB_CONNECT_REQ_PEP_SUB_TYPE:
+ if (len < 1)
+ goto drop;
+ peer_type = (peer_type & 0xff00) | data[0];
+ break;
+ case PN_PIPE_SB_ALIGNED_DATA:
+ aligned = data[0] != 0;
+ break;
+ }
+ n_sb--;
+ }
- oskb = skb_dequeue(&newsk->sk_receive_queue);
- err = pep_accept_conn(newsk, oskb);
- if (err) {
- skb_queue_head(&newsk->sk_receive_queue, oskb);
+ /* Check for duplicate pipe handle */
+ newsk = pep_find_pipe(&pn->hlist, &dst, pipe_handle);
+ if (unlikely(newsk)) {
+ __sock_put(newsk);
newsk = NULL;
- goto out;
+ pep_reject_conn(sk, skb, PN_PIPE_ERR_PEP_IN_USE, GFP_KERNEL);
+ goto drop;
+ }
+
+ /* Create a new to-be-accepted sock */
+ newsk = sk_alloc(sock_net(sk), PF_PHONET, GFP_KERNEL, sk->sk_prot);
+ if (!newsk) {
+ pep_reject_conn(sk, skb, PN_PIPE_ERR_OVERLOAD, GFP_KERNEL);
+ err = -ENOBUFS;
+ goto drop;
}
- kfree_skb(oskb);
+ sock_init_data(NULL, newsk);
+ newsk->sk_state = TCP_SYN_RECV;
+ newsk->sk_backlog_rcv = pipe_do_rcv;
+ newsk->sk_protocol = sk->sk_protocol;
+ newsk->sk_destruct = pipe_destruct;
+
+ newpn = pep_sk(newsk);
+ pn_skb_get_dst_sockaddr(skb, &dst);
+ pn_skb_get_src_sockaddr(skb, &src);
+ newpn->pn_sk.sobject = pn_sockaddr_get_object(&dst);
+ newpn->pn_sk.dobject = pn_sockaddr_get_object(&src);
+ newpn->pn_sk.resource = pn_sockaddr_get_resource(&dst);
sock_hold(sk);
- pep_sk(newsk)->listener = sk;
+ newpn->listener = sk;
+ skb_queue_head_init(&newpn->ctrlreq_queue);
+ newpn->pipe_handle = pipe_handle;
+ atomic_set(&newpn->tx_credits, 0);
+ newpn->ifindex = 0;
+ newpn->peer_type = peer_type;
+ newpn->rx_credits = 0;
+ newpn->rx_fc = newpn->tx_fc = PN_LEGACY_FLOW_CONTROL;
+ newpn->init_enable = enabled;
+ newpn->aligned = aligned;
- sock_hold(newsk);
- sk_del_node_init(newsk);
- sk_acceptq_removed(sk);
+ err = pep_accept_conn(newsk, skb);
+ if (err) {
+ sock_put(newsk);
+ newsk = NULL;
+ goto drop;
+ }
sk_add_node(newsk, &pn->hlist);
- __sock_put(newsk);
-
-out:
+drop:
release_sock(sk);
+ kfree_skb(skb);
*errp = err;
return newsk;
}
@@ -937,7 +887,7 @@ static int pep_init(struct sock *sk)
{
struct pep_sock *pn = pep_sk(sk);
- INIT_HLIST_HEAD(&pn->ackq);
+ sk->sk_destruct = pipe_destruct;
INIT_HLIST_HEAD(&pn->hlist);
skb_queue_head_init(&pn->ctrlreq_queue);
pn->pipe_handle = PN_PIPE_INVALID_HANDLE;
diff --git a/net/phonet/socket.c b/net/phonet/socket.c
index 65a0333..1eccfc3 100644
--- a/net/phonet/socket.c
+++ b/net/phonet/socket.c
@@ -327,6 +327,9 @@ static int pn_socket_accept(struct socket *sock, struct socket *newsock,
struct sock *newsk;
int err;
+ if (unlikely(sk->sk_state != TCP_LISTEN))
+ return -EINVAL;
+
newsk = sk->sk_prot->accept(sk, flags, &err);
if (!newsk)
return err;
@@ -363,13 +366,8 @@ static unsigned int pn_socket_poll(struct file *file, struct socket *sock,
poll_wait(file, sk_sleep(sk), wait);
- switch (sk->sk_state) {
- case TCP_LISTEN:
- return hlist_empty(&pn->ackq) ? 0 : POLLIN;
- case TCP_CLOSE:
+ if (sk->sk_state == TCP_CLOSE)
return POLLERR;
- }
-
if (!skb_queue_empty(&sk->sk_receive_queue))
mask |= POLLIN | POLLRDNORM;
if (!skb_queue_empty(&pn->ctrlreq_queue))
--
1.7.1
^ permalink raw reply related
* [PATCH 2/8] Phonet: return an error when skb TX fails
From: Rémi Denis-Courmont @ 2011-03-08 13:57 UTC (permalink / raw)
To: netdev
In-Reply-To: <201103081556.18877.remi.denis-courmont@nokia.com>
Signed-off-by: Rémi Denis-Courmont <remi.denis-courmont@nokia.com>
---
net/phonet/af_phonet.c | 8 +++-----
1 files changed, 3 insertions(+), 5 deletions(-)
diff --git a/net/phonet/af_phonet.c b/net/phonet/af_phonet.c
index 4706b77..c6fffd9 100644
--- a/net/phonet/af_phonet.c
+++ b/net/phonet/af_phonet.c
@@ -195,11 +195,7 @@ static int pn_send(struct sk_buff *skb, struct net_device *dev,
if (skb->pkt_type == PACKET_LOOPBACK) {
skb_reset_mac_header(skb);
skb_orphan(skb);
- if (irq)
- netif_rx(skb);
- else
- netif_rx_ni(skb);
- err = 0;
+ err = (irq ? netif_rx(skb) : netif_rx_ni(skb)) ? -ENOBUFS : 0;
} else {
err = dev_hard_header(skb, dev, ntohs(skb->protocol),
NULL, NULL, skb->len);
@@ -208,6 +204,8 @@ static int pn_send(struct sk_buff *skb, struct net_device *dev,
goto drop;
}
err = dev_queue_xmit(skb);
+ if (unlikely(err > 0))
+ err = net_xmit_errno(err);
}
return err;
--
1.7.1
^ permalink raw reply related
* [PATCH 1/8] Phonet: fix NULL-deref in a8059512b120362b15424f152b2548fe8b11bd0c
From: Rémi Denis-Courmont @ 2011-03-08 13:56 UTC (permalink / raw)
To: netdev
In-Reply-To: <201103081556.18877.remi.denis-courmont@nokia.com>
Signed-off-by: Rémi Denis-Courmont <remi.denis-courmont@nokia.com>
---
net/phonet/af_phonet.c | 5 ++---
1 files changed, 2 insertions(+), 3 deletions(-)
diff --git a/net/phonet/af_phonet.c b/net/phonet/af_phonet.c
index 30cc676..4706b77 100644
--- a/net/phonet/af_phonet.c
+++ b/net/phonet/af_phonet.c
@@ -262,10 +262,9 @@ int pn_skb_send(struct sock *sk, struct sk_buff *skb,
else if (phonet_address_lookup(net, daddr) == 0) {
dev = phonet_device_get(net);
skb->pkt_type = PACKET_LOOPBACK;
- } else if (pn_sockaddr_get_object(target) == 0) {
+ } else if (dst == 0) {
/* Resource routing (small race until phonet_rcv()) */
- struct sock *sk = pn_find_sock_by_res(net,
- target->spn_resource);
+ struct sock *sk = pn_find_sock_by_res(net, res);
if (sk) {
sock_put(sk);
dev = phonet_device_get(net);
--
1.7.1
^ permalink raw reply related
* [RFCv3] [PATCH 0/8] net-next: Phonet fixes and cleanup
From: Rémi Denis-Courmont @ 2011-03-08 13:56 UTC (permalink / raw)
To: netdev
Changes are available in the git repository at:
git://git.remlab.net/linux-phonet.git master
Rémi Denis-Courmont (8):
Phonet: fix NULL-deref in a8059512b120362b15424f152b2548fe8b11bd0c
Phonet: return an error when skb TX fails
Phonet: correct pipe backlog callback return values
Phonet: factor common code to send control messages
Phonet: allocate sock from accept syscall rather than soft IRQ
Phonet: provide pipe socket option to retrieve the pipe identifier
Phonet: support active connection without pipe controller on modem
Phonet: kill the ST-Ericsson pipe controller Kconfig
Documentation/networking/phonet.txt | 67 ++---
include/linux/phonet.h | 4 +-
include/net/phonet/pep.h | 1 -
net/phonet/Kconfig | 12 -
net/phonet/af_phonet.c | 13 +-
net/phonet/pep.c | 682 ++++++++++++++---------------------
net/phonet/socket.c | 112 +++----
7 files changed, 352 insertions(+), 539 deletions(-)
--
Rémi Denis-Courmont
http://www.remlab.net/
^ permalink raw reply
* Re: [Patch V2] bonding: fix netpoll in active-backup mode
From: Neil Horman @ 2011-03-08 13:43 UTC (permalink / raw)
To: Amerigo Wang; +Cc: linux-kernel, Jay Vosburgh, netdev
In-Reply-To: <1299578336-5888-1-git-send-email-amwang@redhat.com>
On Tue, Mar 08, 2011 at 05:58:56PM +0800, Amerigo Wang wrote:
> V2: avoid calling slave_diable_netpoll() with write_lock_bh() held.
>
> netconsole doesn't work in active-backup mode, because we don't do anything
> for nic failover in active-backup mode. We should disable netpoll on the
> failing slave when it is detected down and enable netpoll when it becomes
> the active slave.
>
You still haven't explained why it needs to be this way. what exactly is the
shortcomming with leaving netpoll enabled on all slaves, regardless of state?
It should be sufficient if, during a failover in a mode where the inactive slave
should not tx data, that you simply clear the slave __LINK_STATE_START bit. That will
prevent higher layers from sending any queued data without you needing to muck
about with netpoll state.
Neil
> Tested by ifdown the current active slave and ifup it again for several times,
> netconsole works well.
>
> Signed-off-by: WANG Cong <amwang@redhat.com>
> Cc: Neil Horman <nhorman@tuxdriver.com>
>
> ---
>
> drivers/net/bonding/bond_main.c | 236 +++++++++++++++++++++------------------
> 1 files changed, 125 insertions(+), 111 deletions(-)
>
> diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
> index 0592e6d..102a558 100644
> --- a/drivers/net/bonding/bond_main.c
> +++ b/drivers/net/bonding/bond_main.c
> @@ -907,6 +907,120 @@ static void bond_mc_list_flush(struct net_device *bond_dev,
> }
> }
>
> +/*--------------------------- Netpoll code ---------------------------*/
> +#ifdef CONFIG_NET_POLL_CONTROLLER
> +static inline int slave_enable_netpoll(struct slave *slave)
> +{
> + struct netpoll *np;
> + int err = 0;
> +
> + if (slave->np)
> + return 0;
> +
> + np = kzalloc(sizeof(*np), GFP_KERNEL);
> + err = -ENOMEM;
> + if (!np)
> + goto out;
> +
> + np->dev = slave->dev;
> + err = __netpoll_setup(np);
> + if (err) {
> + kfree(np);
> + goto out;
> + }
> + slave->np = np;
> +out:
> + return err;
> +}
> +static inline void slave_disable_netpoll(struct slave *slave)
> +{
> + struct netpoll *np = slave->np;
> +
> + if (!np)
> + return;
> +
> + slave->np = NULL;
> + synchronize_rcu_bh();
> + __netpoll_cleanup(np);
> + kfree(np);
> +}
> +static inline bool slave_dev_support_netpoll(struct net_device *slave_dev)
> +{
> + if (slave_dev->priv_flags & IFF_DISABLE_NETPOLL)
> + return false;
> + if (!slave_dev->netdev_ops->ndo_poll_controller)
> + return false;
> + return true;
> +}
> +
> +static void bond_poll_controller(struct net_device *bond_dev)
> +{
> +}
> +
> +static void __bond_netpoll_cleanup(struct bonding *bond)
> +{
> + struct slave *slave;
> + int i;
> +
> + bond_for_each_slave(bond, slave, i)
> + if (IS_UP(slave->dev))
> + slave_disable_netpoll(slave);
> +}
> +static void bond_netpoll_cleanup(struct net_device *bond_dev)
> +{
> + struct bonding *bond = netdev_priv(bond_dev);
> +
> + read_lock(&bond->lock);
> + __bond_netpoll_cleanup(bond);
> + read_unlock(&bond->lock);
> +}
> +
> +static int bond_netpoll_setup(struct net_device *dev, struct netpoll_info *ni)
> +{
> + struct bonding *bond = netdev_priv(dev);
> + struct slave *slave;
> + int i, err = 0;
> +
> + read_lock(&bond->lock);
> + bond_for_each_slave(bond, slave, i) {
> + if (!IS_UP(slave->dev))
> + continue;
> + err = slave_enable_netpoll(slave);
> + if (err) {
> + __bond_netpoll_cleanup(bond);
> + break;
> + }
> + }
> + read_unlock(&bond->lock);
> + return err;
> +}
> +
> +static struct netpoll_info *bond_netpoll_info(struct bonding *bond)
> +{
> + return bond->dev->npinfo;
> +}
> +
> +#else
> +static inline int slave_enable_netpoll(struct slave *slave)
> +{
> + return 0;
> +}
> +static inline void slave_disable_netpoll(struct slave *slave)
> +{
> +}
> +static void bond_netpoll_cleanup(struct net_device *bond_dev)
> +{
> +}
> +static int bond_netpoll_setup(struct net_device *dev, struct netpoll_info *ni)
> +{
> + return 0;
> +}
> +static struct netpoll_info *bond_netpoll_info(struct bonding *bond)
> +{
> + return NULL;
> +}
> +#endif
> +
> /*--------------------------- Active slave change ---------------------------*/
>
> /*
> @@ -1159,6 +1273,7 @@ void bond_change_active_slave(struct bonding *bond, struct slave *new_active)
> bond_set_slave_inactive_flags(old_active);
>
> if (new_active) {
> + struct netpoll_info *ni;
> bond_set_slave_active_flags(new_active);
>
> if (bond->params.fail_over_mac)
> @@ -1174,6 +1289,13 @@ void bond_change_active_slave(struct bonding *bond, struct slave *new_active)
> }
>
> write_unlock_bh(&bond->curr_slave_lock);
> +
> + ni = bond_netpoll_info(bond);
> + if (ni) {
> + new_active->dev->npinfo = ni;
> + slave_enable_netpoll(new_active);
> + }
> +
> read_unlock(&bond->lock);
>
> netdev_bonding_change(bond->dev, NETDEV_BONDING_FAILOVER);
> @@ -1280,116 +1402,6 @@ static void bond_detach_slave(struct bonding *bond, struct slave *slave)
> bond->slave_cnt--;
> }
>
> -#ifdef CONFIG_NET_POLL_CONTROLLER
> -static inline int slave_enable_netpoll(struct slave *slave)
> -{
> - struct netpoll *np;
> - int err = 0;
> -
> - np = kzalloc(sizeof(*np), GFP_KERNEL);
> - err = -ENOMEM;
> - if (!np)
> - goto out;
> -
> - np->dev = slave->dev;
> - err = __netpoll_setup(np);
> - if (err) {
> - kfree(np);
> - goto out;
> - }
> - slave->np = np;
> -out:
> - return err;
> -}
> -static inline void slave_disable_netpoll(struct slave *slave)
> -{
> - struct netpoll *np = slave->np;
> -
> - if (!np)
> - return;
> -
> - slave->np = NULL;
> - synchronize_rcu_bh();
> - __netpoll_cleanup(np);
> - kfree(np);
> -}
> -static inline bool slave_dev_support_netpoll(struct net_device *slave_dev)
> -{
> - if (slave_dev->priv_flags & IFF_DISABLE_NETPOLL)
> - return false;
> - if (!slave_dev->netdev_ops->ndo_poll_controller)
> - return false;
> - return true;
> -}
> -
> -static void bond_poll_controller(struct net_device *bond_dev)
> -{
> -}
> -
> -static void __bond_netpoll_cleanup(struct bonding *bond)
> -{
> - struct slave *slave;
> - int i;
> -
> - bond_for_each_slave(bond, slave, i)
> - if (IS_UP(slave->dev))
> - slave_disable_netpoll(slave);
> -}
> -static void bond_netpoll_cleanup(struct net_device *bond_dev)
> -{
> - struct bonding *bond = netdev_priv(bond_dev);
> -
> - read_lock(&bond->lock);
> - __bond_netpoll_cleanup(bond);
> - read_unlock(&bond->lock);
> -}
> -
> -static int bond_netpoll_setup(struct net_device *dev, struct netpoll_info *ni)
> -{
> - struct bonding *bond = netdev_priv(dev);
> - struct slave *slave;
> - int i, err = 0;
> -
> - read_lock(&bond->lock);
> - bond_for_each_slave(bond, slave, i) {
> - if (!IS_UP(slave->dev))
> - continue;
> - err = slave_enable_netpoll(slave);
> - if (err) {
> - __bond_netpoll_cleanup(bond);
> - break;
> - }
> - }
> - read_unlock(&bond->lock);
> - return err;
> -}
> -
> -static struct netpoll_info *bond_netpoll_info(struct bonding *bond)
> -{
> - return bond->dev->npinfo;
> -}
> -
> -#else
> -static inline int slave_enable_netpoll(struct slave *slave)
> -{
> - return 0;
> -}
> -static inline void slave_disable_netpoll(struct slave *slave)
> -{
> -}
> -static void bond_netpoll_cleanup(struct net_device *bond_dev)
> -{
> -}
> -static int bond_netpoll_setup(struct net_device *dev, struct netpoll_info *ni)
> -{
> - return 0;
> -}
> -static struct netpoll_info *bond_netpoll_info(struct bonding *bond)
> -{
> - return NULL;
> -}
> -#endif
> -
> /*---------------------------------- IOCTL ----------------------------------*/
>
> static int bond_sethwaddr(struct net_device *bond_dev,
> @@ -2532,8 +2544,10 @@ static void bond_miimon_commit(struct bonding *bond)
> bond_alb_handle_link_change(bond, slave,
> BOND_LINK_DOWN);
>
> - if (slave == bond->curr_active_slave)
> + if (slave == bond->curr_active_slave) {
> + slave_disable_netpoll(slave);
> goto do_failover;
> + }
>
> continue;
>
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
^ permalink raw reply
* Re: [patch net-next-2.6] net: reinject arps into bonding slave instead of master
From: Andy Gospodarek @ 2011-03-08 13:42 UTC (permalink / raw)
To: Jiri Pirko
Cc: Andy Gospodarek, Nicolas de Pesloüan, netdev, davem,
shemminger, kaber, fubar, eric.dumazet
In-Reply-To: <20110308071350.GA2826@psychotron.redhat.com>
On Tue, Mar 08, 2011 at 08:13:51AM +0100, Jiri Pirko wrote:
> Mon, Mar 07, 2011 at 11:43:38PM CET, andy@greyhouse.net wrote:
[...]
> >
> >This patch doesn't work.
> >
> >My setup has bond0.100 -> bond0 -> eth2 and eth3. ARP monitoring is
> >enabled as is arp_valiate.
> >
> >The initial problem was just that just before vlan_on_bond_hook is
> >called, skb->dev = bond0.100 and orig_dev = eth2. (This is after
> >running goto another_route and having been called back through
> >__netif_receive_skb since vlan_hwaccel_do_receive it true.)
> >
> >Now vlan_on_bond_hook is called and we have 2 skbs.
> >
> >The original skb still have skb->dev = bond0.100 and orig_dev = eth2.
> >Since bond_arp_rcv is registered for traffic only to bond0, the handler
> >is not hit and the frame is dropped (or processed by another handler).
> >
> >The cloned skb has skb->dev = bond0 and is put back on the receive queue
> >and comes back through __netif_receive_skb. This frame will match the
> >ptype entry for bond_arp_rcv, but since orig_dev = bond0 in this case,
> >the code in bond_arp_rcv will not handle the frame.
> >
> >If we truly want to track the original interface that received the
> >frame, the following is a better option. With the recursive nature of
> >__netif_receive_skb at this point, we should really consider setting
> >orig_dev from skb_iif rather than just from skb->dev.
> >
> >diff --git a/net/core/dev.c b/net/core/dev.c
> >index 30440e7..500fdbc 100644
> >--- a/net/core/dev.c
> >+++ b/net/core/dev.c
> >@@ -3135,7 +3135,6 @@ static int __netif_receive_skb(struct sk_buff *skb)
> >
> > if (!skb->skb_iif)
> > skb->skb_iif = skb->dev->ifindex;
> >- orig_dev = skb->dev;
> >
> > skb_reset_network_header(skb);
> > skb_reset_transport_header(skb);
> >@@ -3145,6 +3144,7 @@ static int __netif_receive_skb(struct sk_buff *skb)
> >
> > rcu_read_lock();
> >
> >+ orig_dev = dev_get_by_index_rcu(dev_net(skb->dev),skb->skb_iif);
> > another_round:
> >
> > __this_cpu_inc(softnet_data.processed);
> >
>
> This was proposed earlier. people did not like this very much :(
>
Interesting. I'll have to look back at the discussion.
> I forgot to include crucial part of "goto another_round for vlan".
> Following patch should work (will test it once I get to work):
>
> Subject: [patch net-next 2.6] net: reinject arps into bonding slave instead of master
>
> Recent patch "bonding: move processing of recv handlers into
> handle_frame()" caused a regression on following net scheme:
>
> eth0 - bond0 - bond0.5
>
> where arp monitoring is happening over vlan. This patch fixes it by
> reinjecting the arp packet into bonding slave device so the bonding
> rx_handler can pickup and process it.
>
> also instead of calling __netif_receive_skb recursively, "goto another
> round" does this recursion. The point is the orig_dev variable remains
> intact.
>
> Signed-off-by: Jiri Pirko <jpirko@redhat.com>
> ---
> net/core/dev.c | 11 +++++------
> 1 files changed, 5 insertions(+), 6 deletions(-)
>
> diff --git a/net/core/dev.c b/net/core/dev.c
> index c71bd18..ec330e1 100644
> --- a/net/core/dev.c
> +++ b/net/core/dev.c
> @@ -3094,12 +3094,12 @@ void netdev_rx_handler_unregister(struct net_device *dev)
> }
> EXPORT_SYMBOL_GPL(netdev_rx_handler_unregister);
>
> -static void vlan_on_bond_hook(struct sk_buff *skb)
> +static void vlan_on_bond_hook(struct sk_buff *skb, struct net_device *orig_dev)
> {
> /*
> * Make sure ARP frames received on VLAN interfaces stacked on
> * bonding interfaces still make their way to any base bonding
> - * device that may have registered for a specific ptype.
> + * device by reinjecting the frame into bonding slave (orig_dev)
> */
> if (skb->dev->priv_flags & IFF_802_1Q_VLAN &&
> vlan_dev_real_dev(skb->dev)->priv_flags & IFF_BONDING &&
> @@ -3108,7 +3108,7 @@ static void vlan_on_bond_hook(struct sk_buff *skb)
>
> if (!skb2)
> return;
> - skb2->dev = vlan_dev_real_dev(skb->dev);
> + skb2->dev = orig_dev;
> netif_rx(skb2);
> }
>
I'm pretty sure this patch will have the same catastrophic problem your
last one did. By cloning and setting skb2->dev = orig_dev you just
inserted a frame identical to the one we received right back into the
stack. It only took a few minutes for my box to melt as one frame on
the wire will cause an infinite number of frames to be received by the
stack.
> @@ -3196,13 +3196,12 @@ ncls:
> pt_prev = NULL;
> }
> if (vlan_hwaccel_do_receive(&skb)) {
> - ret = __netif_receive_skb(skb);
> - goto out;
> + goto another_round;
> } else if (unlikely(!skb))
> goto out;
> }
>
> - vlan_on_bond_hook(skb);
> + vlan_on_bond_hook(skb, orig_dev);
>
> /* deliver only exact match when indicated */
> null_or_dev = deliver_exact ? skb->dev : NULL;
^ permalink raw reply
* Re: [Patch] bonding: fix netpoll in active-backup mode
From: Neil Horman @ 2011-03-08 13:26 UTC (permalink / raw)
To: Cong Wang
Cc: linux-kernel, Jay Vosburgh, David S. Miller, Herbert Xu,
Paul E. McKenney, John W. Linville, Eric Dumazet, netdev
In-Reply-To: <4D75AD50.7060400@redhat.com>
On Tue, Mar 08, 2011 at 12:15:12PM +0800, Cong Wang wrote:
> 于 2011年03月08日 02:50, Neil Horman 写道:
> >On Mon, Mar 07, 2011 at 10:11:50PM +0800, Amerigo Wang wrote:
> >>netconsole doesn't work in active-backup mode, because we don't do anything
> >>for nic failover in active-backup mode. This patch fixes the problem by:
> >>
> >>1) make slave_enable_netpoll() and slave_disable_netpoll() callable in softirq
> >> context, that is, moving code after synchronize_rcu_bh() into call_rcu_bh()
> >> callback function, teaching kzalloc() to use GFP_ATOMIC.
> >>
> >>2) disable netpoll on old slave and enable netpoll on the new slave.
> >>
> >>Tested by ifdown the current active slave and ifup it again for several times,
> >>netconsole works well.
> >>
> >>Signed-off-by: WANG Cong<amwang@redhat.com>
> >>
> >I may be missing soething but this seems way over-complicated to me. I presume
> >the problem is that in active backup mode a failover results in the new active
> >slave not having netpoll setup on it? If thats the case, why not just setup
> >netpoll on all slaves when ndo_netpoll_setup is called on the bonding interface?
> >I don't see anything immeidately catastrophic that would happen as a result.
>
>
> But we still need to clean up the netpoll on the failing slave, which still
> needs to call slave_disable_netpoll() in monitor code, I see no big differences
> with the solution I take.
>
Why? I understand you want to free up that memory, but I don't see any special
state codified in that structure that can't wait until you disable netpoll on
the bond as a whole. Save yourself the time and trouble, enable netpoll on both
slaves when its enabled on the bond, and tear it down when its torn down on the
bond. Do worry about doing anything during a failover.
Neil
>
> >And then you wouldn't have to worry about disabling/enabling anything on a
> >failover (or during a panic for that matter). As for the rcu bits? Why are
> >they needed? One would presume that wouldn't (or at least shouldn't) be able to
> >teardown our netpoll setup until such time as all the pending frames for that
> >netpoll client have been transmitted. If we're not blocknig on that RCU isn't
> >really going to help. Seems like the proper fix is take a reference to the
> >appropriate npinfo struct in netpoll_send_skb, and drop it from the skbs
> >destructor or some such.
>
> I saw a "scheduling while in atomic" warning without touching the rcu bits.
>
> Thanks!
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
^ permalink raw reply
* Re: bonding can't change to another slave if you ifdown the active slave
From: WANG Cong @ 2011-03-08 12:51 UTC (permalink / raw)
To: netdev
In-Reply-To: <4D75D244.9070409@gmail.com>
On Tue, 08 Mar 2011 14:52:52 +0800, Weiping Pan wrote:
> ok, I use "Host-only mode" and get
>
>
>
> an conclusion, that if the first enslaved nic is pulled out, bonding
> can't handle well.
>
> First test.
> I first enslave eth6, then pull it out, bonding doesn't work. on host,
...
>
>
> Second test
> I first enslave eth6, then pull eth7 out, bonding works well. on guest,
Can you show me your /proc/net/bonding/bond0 before and after pulling down
eth6 or eth7? And what does `ip link show` say?
^ permalink raw reply
* Re: [PATCH v3] tcp: ioctl type SIOCOUTQNSD returns amount of data not sent
From: Steffen Sledz @ 2011-03-08 11:57 UTC (permalink / raw)
To: netdev, Alan Cox
Cc: linux-kernel, David S. Miller, Alexey Kuznetsov,
Pekka Savola (ipv6), James Morris, Hideaki YOSHIFUJI,
Patrick McHardy, Mario Schuknecht
In-Reply-To: <1299441595-18455-1-git-send-email-sledz@dresearch.de>
Am 06.03.2011 20:59, schrieb Steffen Sledz:
> From: Mario Schuknecht <m.schuknecht@dresearch.de>
>
> In contrast to SIOCOUTQ which returns the amount of data sent
> but not yet acknowledged plus data not yet sent this patch only
> returns the data not sent.
>
> For various methods of live streaming bitrate control it may
> be helpful to know how much data are in the tcp outqueue are
> not sent yet.
>
> Signed-off-by: Mario Schuknecht <m.schuknecht@dresearch.de>
> Signed-off-by: Steffen Sledz <sledz@dresearch.de>
> ---
> include/linux/sockios.h | 4 +++-
> net/ipv4/tcp.c | 9 +++++++++
> 2 files changed, 12 insertions(+), 1 deletions(-)
This is the patch version modified according to the comments from Alan Cox.
Anything else to do?
Steffen
--
DResearch Fahrzeugelektronik GmbH
Otto-Schmirgal-Str. 3, 10319 Berlin, Germany
Tel: +49 30 515932-237 mailto:sledz@DResearch.de
Fax: +49 30 515932-299
Geschäftsführer: Dr. Michael Weber, Werner Mögle;
Amtsgericht Berlin Charlottenburg; HRB 130120 B;
Ust.-IDNr. DE273952058
^ permalink raw reply
* [PATCHv2 13/17] mlx4: Add blue flame support for kernel consumers
From: Eli Cohen @ 2011-03-08 11:25 UTC (permalink / raw)
To: davem, roland; +Cc: netdev, yevgenyp
Using blue flame can improve latency by allowing the HW to more efficiently
access the WQE. This patch presents two functions that are used to allocate or
release HW resources for using blue flame; the caller need to supply a struct
mlx4_bf object when allocating resources. Consumers that make use of this API
should post doorbells to the UAR object pointed by the initialized struct
mlx4_bf.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
---
Changes v2: fix debug message
drivers/net/mlx4/main.c | 31 ++++++++++++++
drivers/net/mlx4/mlx4.h | 3 +
drivers/net/mlx4/pd.c | 94 +++++++++++++++++++++++++++++++++++++++++++
include/linux/mlx4/device.h | 13 ++++++
include/linux/mlx4/qp.h | 1 +
5 files changed, 142 insertions(+), 0 deletions(-)
diff --git a/drivers/net/mlx4/main.c b/drivers/net/mlx4/main.c
index 2765a3c..1403377 100644
--- a/drivers/net/mlx4/main.c
+++ b/drivers/net/mlx4/main.c
@@ -39,6 +39,7 @@
#include <linux/pci.h>
#include <linux/dma-mapping.h>
#include <linux/slab.h>
+#include <linux/io-mapping.h>
#include <linux/mlx4/device.h>
#include <linux/mlx4/doorbell.h>
@@ -718,8 +719,31 @@ static void mlx4_free_icms(struct mlx4_dev *dev)
mlx4_free_icm(dev, priv->fw.aux_icm, 0);
}
+static int map_bf_area(struct mlx4_dev *dev)
+{
+ struct mlx4_priv *priv = mlx4_priv(dev);
+ resource_size_t bf_start;
+ resource_size_t bf_len;
+ int err = 0;
+
+ bf_start = pci_resource_start(dev->pdev, 2) + (dev->caps.num_uars << PAGE_SHIFT);
+ bf_len = pci_resource_len(dev->pdev, 2) - (dev->caps.num_uars << PAGE_SHIFT);
+ priv->bf_mapping = io_mapping_create_wc(bf_start, bf_len);
+ if (!priv->bf_mapping)
+ err = -ENOMEM;
+
+ return err;
+}
+
+static void unmap_bf_area(struct mlx4_dev *dev)
+{
+ if (mlx4_priv(dev)->bf_mapping)
+ io_mapping_free(mlx4_priv(dev)->bf_mapping);
+}
+
static void mlx4_close_hca(struct mlx4_dev *dev)
{
+ unmap_bf_area(dev);
mlx4_CLOSE_HCA(dev, 0);
mlx4_free_icms(dev);
mlx4_UNMAP_FA(dev);
@@ -772,6 +796,9 @@ static int mlx4_init_hca(struct mlx4_dev *dev)
goto err_stop_fw;
}
+ if (map_bf_area(dev))
+ mlx4_dbg(dev, "Failed to map blue flame area\n");
+
init_hca.log_uar_sz = ilog2(dev->caps.num_uars);
err = mlx4_init_icm(dev, &dev_cap, &init_hca, icm_size);
@@ -802,6 +829,7 @@ err_free_icm:
mlx4_free_icms(dev);
err_stop_fw:
+ unmap_bf_area(dev);
mlx4_UNMAP_FA(dev);
mlx4_free_icm(dev, priv->fw.fw_icm, 0);
@@ -1127,6 +1155,9 @@ static int __mlx4_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
INIT_LIST_HEAD(&priv->pgdir_list);
mutex_init(&priv->pgdir_mutex);
+ INIT_LIST_HEAD(&priv->bf_list);
+ mutex_init(&priv->bf_mutex);
+
/*
* Now reset the HCA before we touch the PCI capabilities or
* attempt a firmware command, since a boot ROM may have left
diff --git a/drivers/net/mlx4/mlx4.h b/drivers/net/mlx4/mlx4.h
index 0da5bb7..c5f5d36 100644
--- a/drivers/net/mlx4/mlx4.h
+++ b/drivers/net/mlx4/mlx4.h
@@ -313,6 +313,9 @@ struct mlx4_priv {
struct mlx4_port_info port[MLX4_MAX_PORTS + 1];
struct mlx4_sense sense;
struct mutex port_mutex;
+ struct list_head bf_list;
+ struct mutex bf_mutex;
+ struct io_mapping *bf_mapping;
};
static inline struct mlx4_priv *mlx4_priv(struct mlx4_dev *dev)
diff --git a/drivers/net/mlx4/pd.c b/drivers/net/mlx4/pd.c
index c4988d6..5210a0f 100644
--- a/drivers/net/mlx4/pd.c
+++ b/drivers/net/mlx4/pd.c
@@ -32,6 +32,7 @@
*/
#include <linux/errno.h>
+#include <linux/io-mapping.h>
#include <asm/page.h>
@@ -77,6 +78,7 @@ int mlx4_uar_alloc(struct mlx4_dev *dev, struct mlx4_uar *uar)
return -ENOMEM;
uar->pfn = (pci_resource_start(dev->pdev, 2) >> PAGE_SHIFT) + uar->index;
+ uar->map = NULL;
return 0;
}
@@ -88,6 +90,98 @@ void mlx4_uar_free(struct mlx4_dev *dev, struct mlx4_uar *uar)
}
EXPORT_SYMBOL_GPL(mlx4_uar_free);
+int mlx4_bf_alloc(struct mlx4_dev *dev, struct mlx4_bf *bf)
+{
+ struct mlx4_priv *priv = mlx4_priv(dev);
+ struct mlx4_uar *uar;
+ int err = 0;
+ int idx;
+
+ if (!priv->bf_mapping)
+ return -ENOMEM;
+
+ mutex_lock(&priv->bf_mutex);
+ if (!list_empty(&priv->bf_list))
+ uar = list_entry(priv->bf_list.next, struct mlx4_uar, bf_list);
+ else {
+ uar = kmalloc(sizeof *uar, GFP_KERNEL);
+ if (!uar) {
+ err = -ENOMEM;
+ goto out;
+ }
+ err = mlx4_uar_alloc(dev, uar);
+ if (err)
+ goto free_kmalloc;
+
+ uar->map = ioremap(uar->pfn << PAGE_SHIFT, PAGE_SIZE);
+ if (!uar->map) {
+ err = -ENOMEM;
+ goto free_uar;
+ }
+
+ uar->bf_map = io_mapping_map_wc(priv->bf_mapping, uar->index << PAGE_SHIFT);
+ if (!uar->bf_map) {
+ err = -ENOMEM;
+ goto unamp_uar;
+ }
+ uar->free_bf_bmap = 0;
+ list_add(&uar->bf_list, &priv->bf_list);
+ }
+
+ bf->uar = uar;
+ idx = ffz(uar->free_bf_bmap);
+ uar->free_bf_bmap |= 1 << idx;
+ bf->uar = uar;
+ bf->offset = 0;
+ bf->buf_size = dev->caps.bf_reg_size / 2;
+ bf->reg = uar->bf_map + idx * dev->caps.bf_reg_size;
+ if (uar->free_bf_bmap == (1 << dev->caps.bf_regs_per_page) - 1)
+ list_del_init(&uar->bf_list);
+
+ goto out;
+
+unamp_uar:
+ bf->uar = NULL;
+ iounmap(uar->map);
+
+free_uar:
+ mlx4_uar_free(dev, uar);
+
+free_kmalloc:
+ kfree(uar);
+
+out:
+ mutex_unlock(&priv->bf_mutex);
+ return err;
+}
+EXPORT_SYMBOL_GPL(mlx4_bf_alloc);
+
+void mlx4_bf_free(struct mlx4_dev *dev, struct mlx4_bf *bf)
+{
+ struct mlx4_priv *priv = mlx4_priv(dev);
+ int idx;
+
+ if (!bf->uar || !bf->uar->bf_map)
+ return;
+
+ mutex_lock(&priv->bf_mutex);
+ idx = (bf->reg - bf->uar->bf_map) / dev->caps.bf_reg_size;
+ bf->uar->free_bf_bmap &= ~(1 << idx);
+ if (!bf->uar->free_bf_bmap) {
+ if (!list_empty(&bf->uar->bf_list))
+ list_del(&bf->uar->bf_list);
+
+ io_mapping_unmap(bf->uar->bf_map);
+ iounmap(bf->uar->map);
+ mlx4_uar_free(dev, bf->uar);
+ kfree(bf->uar);
+ } else if (list_empty(&bf->uar->bf_list))
+ list_add(&bf->uar->bf_list, &priv->bf_list);
+
+ mutex_unlock(&priv->bf_mutex);
+}
+EXPORT_SYMBOL_GPL(mlx4_bf_free);
+
int mlx4_init_uar_table(struct mlx4_dev *dev)
{
if (dev->caps.num_uars <= 128) {
diff --git a/include/linux/mlx4/device.h b/include/linux/mlx4/device.h
index 0492146..65364f1 100644
--- a/include/linux/mlx4/device.h
+++ b/include/linux/mlx4/device.h
@@ -334,6 +334,17 @@ struct mlx4_fmr {
struct mlx4_uar {
unsigned long pfn;
int index;
+ struct list_head bf_list;
+ unsigned free_bf_bmap;
+ void __iomem *map;
+ void __iomem *bf_map;
+};
+
+struct mlx4_bf {
+ unsigned long offset;
+ int buf_size;
+ struct mlx4_uar *uar;
+ void __iomem *reg;
};
struct mlx4_cq {
@@ -461,6 +472,8 @@ void mlx4_pd_free(struct mlx4_dev *dev, u32 pdn);
int mlx4_uar_alloc(struct mlx4_dev *dev, struct mlx4_uar *uar);
void mlx4_uar_free(struct mlx4_dev *dev, struct mlx4_uar *uar);
+int mlx4_bf_alloc(struct mlx4_dev *dev, struct mlx4_bf *bf);
+void mlx4_bf_free(struct mlx4_dev *dev, struct mlx4_bf *bf);
int mlx4_mtt_init(struct mlx4_dev *dev, int npages, int page_shift,
struct mlx4_mtt *mtt);
diff --git a/include/linux/mlx4/qp.h b/include/linux/mlx4/qp.h
index 0eeb2a1..9e9eb21 100644
--- a/include/linux/mlx4/qp.h
+++ b/include/linux/mlx4/qp.h
@@ -303,6 +303,7 @@ struct mlx4_wqe_data_seg {
enum {
MLX4_INLINE_ALIGN = 64,
+ MLX4_INLINE_SEG = 1 << 31,
};
struct mlx4_wqe_inline_seg {
--
1.6.0.2
^ permalink raw reply related
* Re: Bug inkvm_set_irq
From: Michael S. Tsirkin @ 2011-03-08 11:13 UTC (permalink / raw)
To: Jean-Philippe Menil; +Cc: netdev, kvm, virtualization
In-Reply-To: <4D70B339.3080001@univ-nantes.fr>
On Fri, Mar 04, 2011 at 10:39:05AM +0100, Jean-Philippe Menil wrote:
> Yes, it's a 2.6.37.2 kernel.
OK, here's a debugging patch.
Please run with slab debugging as previously until you see
'eventfd bug detected!' in dmesg or until there is a crash.
It might be also useful to enable timestampts on printk with
Symbol: PRINTK_TIME [=y]
│ Type : boolean
│ Prompt: Show timing information on printks
once you see the error, please upload the
full dmesg output somewhere to we can track what
goes on.
Hopefully there won't be an oops this time which
should make it easier for you to test (no need to
reboot).
diff --git a/virt/kvm/eventfd.c b/virt/kvm/eventfd.c
index c1f1e3c..3cb679b 100644
--- a/virt/kvm/eventfd.c
+++ b/virt/kvm/eventfd.c
@@ -32,6 +32,7 @@
#include <linux/eventfd.h>
#include <linux/kernel.h>
#include <linux/slab.h>
+#include <linux/nmi.h>
#include "iodev.h"
@@ -43,6 +44,8 @@
* --------------------------------------------------------------------
*/
+#define KVM_BAD_PTR ((void*)(long)(0x6b6b6b6b6b6b6b6bull))
+
struct _irqfd {
struct kvm *kvm;
struct eventfd_ctx *eventfd;
@@ -61,6 +64,13 @@ irqfd_inject(struct work_struct *work)
{
struct _irqfd *irqfd = container_of(work, struct _irqfd, inject);
struct kvm *kvm = irqfd->kvm;
+ if (kvm == KVM_BAD_PTR) {
+ printk(KERN_ERR "Eventfd bug detected!\n");
+ printk(KERN_ERR "%s(work=%p,irqfd=%p,kvm=%p,gsi=%d)\n", __func__,
+ work, irqfd, kvm, irqfd->gsi);
+ trigger_all_cpu_backtrace();
+ return;
+ }
kvm_set_irq(kvm, KVM_USERSPACE_IRQ_SOURCE_ID, irqfd->gsi, 1);
kvm_set_irq(kvm, KVM_USERSPACE_IRQ_SOURCE_ID, irqfd->gsi, 0);
@@ -75,6 +85,8 @@ irqfd_shutdown(struct work_struct *work)
struct _irqfd *irqfd = container_of(work, struct _irqfd, shutdown);
u64 cnt;
+ printk(KERN_ERR "%s(work=%p,irqfd=%p,kvm=%p, gsi=%d)\n", __func__,
+ work, irqfd, irqfd->kvm, irqfd->gsi);
/*
* Synchronize with the wait-queue and unhook ourselves to prevent
* further events.
@@ -91,6 +103,8 @@ irqfd_shutdown(struct work_struct *work)
* It is now safe to release the object's resources
*/
eventfd_ctx_put(irqfd->eventfd);
+ printk(KERN_ERR "kfree at %s(work=%p,irqfd=%p)\n", __func__,
+ work, irqfd);
kfree(irqfd);
}
@@ -111,6 +125,8 @@ static void
irqfd_deactivate(struct _irqfd *irqfd)
{
BUG_ON(!irqfd_is_active(irqfd));
+ printk(KERN_ERR "%s(irqfd=%p,kvm=%p, gsi=%d)\n", __func__,
+ irqfd, irqfd->kvm, irqfd->gsi);
list_del_init(&irqfd->list);
@@ -178,6 +194,8 @@ kvm_irqfd_assign(struct kvm *kvm, int fd, int gsi)
irqfd->kvm = kvm;
irqfd->gsi = gsi;
+ printk(KERN_ERR "%s(irqfd=%p,kvm=%p, gsi=%d)\n", __func__,
+ irqfd, kvm, gsi);
INIT_LIST_HEAD(&irqfd->list);
INIT_WORK(&irqfd->inject, irqfd_inject);
INIT_WORK(&irqfd->shutdown, irqfd_shutdown);
@@ -264,6 +282,8 @@ kvm_irqfd_deassign(struct kvm *kvm, int fd, int gsi)
struct _irqfd *irqfd, *tmp;
struct eventfd_ctx *eventfd;
+ printk(KERN_ERR "%s(kvm=%p, gsi=%d)\n", __func__,
+ kvm, gsi);
eventfd = eventfd_ctx_fdget(fd);
if (IS_ERR(eventfd))
return PTR_ERR(eventfd);
@@ -305,6 +325,7 @@ void
kvm_irqfd_release(struct kvm *kvm)
{
struct _irqfd *irqfd, *tmp;
+ printk(KERN_ERR "%s(kvm=%p)\n", __func__, kvm);
spin_lock_irq(&kvm->irqfds.lock);
^ permalink raw reply related
* [PATCH] Make CUBIC Hystart more robust to RTT variations
From: Lucas Nussbaum @ 2011-03-08 11:10 UTC (permalink / raw)
To: WANG Cong; +Cc: netdev
In-Reply-To: <il4vur$3ka$1@dough.gmane.org>
CUBIC Hystart uses two heuristics to exit slow start earlier, before
losses start to occur. Unfortunately, it tends to exit slow start far
too early, causing poor performance since convergence to the optimal
cwnd is then very slow. This was reported in
http://permalink.gmane.org/gmane.linux.network/188169 and
https://partner-bugzilla.redhat.com/show_bug.cgi?id=616985
I am using an experimental testbed (http://www.grid5000.fr/) with two
machines connected using Gigabit ethernet to a dedicated 10-Gb backbone.
RTT between both machines is 11.3ms. Using TCP CUBIC without Hystart,
cwnd grows to ~2200. With Hystart enabled, CUBIC exits slow start with
cwnd lower than 100, and often lower than 20, which leads to the poor
performance that I reported.
After instrumenting TCP CUBIC, I found out that the segment-to-ack RTT
tends to vary quite a lot even when the network is not congested, due to
several factors including the fact that TCP sends packet in burst (so
the packets are queued locally before being sent, increasing their RTT),
and delayed ACKs on the destination host.
The patch below increases the thresholds used by the two Hystart
heuristics. First, the length of an ACK train needs to reach 2*minRTT.
Second, the max RTT of a group of packets also needs to reach 2*minRTT.
In my setup, this causes Hystart to exit slow start when cwnd is in the
1900-2000 range using the ACK train heuristics, and sometimes to exit in
the 700-900 range using the delay increase heuristic, dramatically
improving performance.
I could provide access to my testbed if someone wants to do further
experiments.
Signed-off-by: Lucas Nussbaum <lucas.nussbaum@loria.fr>
--
| Lucas Nussbaum MCF Université Nancy 2 |
| lucas.nussbaum@loria.fr LORIA / AlGorille |
| http://www.loria.fr/~lnussbau/ +33 3 54 95 86 19 |
---
diff --git a/net/ipv4/tcp_cubic.c b/net/ipv4/tcp_cubic.c
index 71d5f2f..e404de4 100644
--- a/net/ipv4/tcp_cubic.c
+++ b/net/ipv4/tcp_cubic.c
@@ -344,7 +344,7 @@ static void hystart_update(struct sock *sk, u32 delay)
/* first detection parameter - ack-train detection */
if (curr_jiffies - ca->last_jiffies <= msecs_to_jiffies(2)) {
ca->last_jiffies = curr_jiffies;
- if (curr_jiffies - ca->round_start >= ca->delay_min>>4)
+ if (curr_jiffies - ca->round_start >= ca->delay_min>>2)
ca->found |= HYSTART_ACK_TRAIN;
}
@@ -355,8 +355,7 @@ static void hystart_update(struct sock *sk, u32 delay)
ca->sample_cnt++;
} else {
- if (ca->curr_rtt > ca->delay_min +
- HYSTART_DELAY_THRESH(ca->delay_min>>4))
+ if (ca->curr_rtt > ca->delay_min<<1)
ca->found |= HYSTART_DELAY;
}
/*
^ permalink raw reply related
* Re: [Bugme-new] [Bug 30092] New: smsc911x.c drops long packets with VLAN tags
From: Steve.Glendinning @ 2011-03-08 10:27 UTC (permalink / raw)
To: David Miller
Cc: akpm, bugme-daemon, bugzilla-daemon, netdev, weinholt,
Ian.Saturley
In-Reply-To: <20110307.123103.115933638.davem@davemloft.net>
> > Subject: [PATCH] net/smsc911x.c: Set the VLAN1 register to fix VLAN
MTU problem
> > From: Göran Weinholt <weinholt@csbnet.se>
> >
> > The smsc911x driver would drop frames longer than 1518 bytes, which is
a
> > problem for networks with VLAN tagging. The VLAN1 tag register is used
> > to increase the legal frame size to 1522 when a VLAN tag is
identified.
> >
> > Signed-off-by: Göran Weinholt <weinholt@csbnet.se>
>
> Applied, thanks.
Thanks David, I was just about to ack this but you beat me to it :-)
As Göran said, the datasheet is a little vague. I've checked here, and
there aren't any bad implications of always turning this option on.
Acked-by: Steve Glendinning <steve.glendinning@smsc.com>
^ permalink raw reply
* Re: [PATCH] Make CUBIC Hystart more robust to RTT variations
From: WANG Cong @ 2011-03-08 10:21 UTC (permalink / raw)
To: netdev
In-Reply-To: <20110308093215.GA23842@xanadu.blop.info>
On Tue, 08 Mar 2011 10:32:15 +0100, Lucas Nussbaum wrote:
> + {
> +// printk("hystart_update: cwnd=%u found=%d
> delay_min=%u cur_jif=%u round_start=%u curr_rtt=%u\n", tp->snd_cwnd,
> ca->found, ca
Please remove this line from your patch.
^ permalink raw reply
* [RFC PATCH] iproute2: Add IPsec extended sequence number support
From: Steffen Klassert @ 2011-03-08 10:12 UTC (permalink / raw)
To: Herbert Xu, David Miller
Cc: Alex Badea, Andreas Gruenbacher, netdev, linux-crypto
In-Reply-To: <20110308100407.GB31402@secunet.com>
---
The patch is based on branch 'net-next' of
git://git.kernel.org/pub/scm/linux/kernel/git/shemminger/iproute2.git
include/linux/xfrm.h | 12 ++++++++++++
ip/ipxfrm.c | 8 +++++++-
ip/xfrm_state.c | 35 ++++++++++++++++++++++++++++-------
3 files changed, 47 insertions(+), 8 deletions(-)
diff --git a/include/linux/xfrm.h b/include/linux/xfrm.h
index b405b70..fcda53c 100644
--- a/include/linux/xfrm.h
+++ b/include/linux/xfrm.h
@@ -84,6 +84,16 @@ struct xfrm_replay_state {
__u32 bitmap;
};
+struct xfrm_replay_state_esn {
+ unsigned int bmp_len;
+ __u32 oseq;
+ __u32 seq;
+ __u32 oseq_hi;
+ __u32 seq_hi;
+ __u32 replay_window;
+ __u32 bmp[0];
+};
+
struct xfrm_algo {
char alg_name[64];
unsigned int alg_key_len; /* in bits */
@@ -284,6 +294,7 @@ enum xfrm_attr_type_t {
XFRMA_ALG_AUTH_TRUNC, /* struct xfrm_algo_auth */
XFRMA_MARK, /* struct xfrm_mark */
XFRMA_TFCPAD, /* __u32 */
+ XFRMA_REPLAY_ESN_VAL, /* struct xfrm_replay_esn */
__XFRMA_MAX
#define XFRMA_MAX (__XFRMA_MAX - 1)
@@ -351,6 +362,7 @@ struct xfrm_usersa_info {
#define XFRM_STATE_ICMP 16
#define XFRM_STATE_AF_UNSPEC 32
#define XFRM_STATE_ALIGN4 64
+#define XFRM_STATE_ESN 128
};
struct xfrm_usersa_id {
diff --git a/ip/ipxfrm.c b/ip/ipxfrm.c
index 0c7aaad..1dc496e 100644
--- a/ip/ipxfrm.c
+++ b/ip/ipxfrm.c
@@ -691,6 +691,12 @@ done:
void xfrm_xfrma_print(struct rtattr *tb[], __u16 family,
FILE *fp, const char *prefix)
{
+
+ if (tb[XFRMA_REPLAY_ESN_VAL]) {
+ struct rtattr *rta = tb[XFRMA_REPLAY_ESN_VAL];
+ struct xfrm_replay_state_esn *repl = (struct xfrm_replay_state_esn *) RTA_DATA(rta);
+ fprintf(fp, "\treplay-window %u\n", repl->replay_window);
+ }
if (tb[XFRMA_MARK]) {
struct rtattr *rta = tb[XFRMA_MARK];
struct xfrm_mark *m = (struct xfrm_mark *) RTA_DATA(rta);
@@ -841,7 +847,6 @@ void xfrm_state_info_print(struct xfrm_usersa_info *xsinfo,
STRBUF_CAT(buf, "\t");
fputs(buf, fp);
- fprintf(fp, "replay-window %u ", xsinfo->replay_window);
if (show_stats > 0)
fprintf(fp, "seq 0x%08u ", xsinfo->seq);
if (show_stats > 0 || xsinfo->flags) {
@@ -855,6 +860,7 @@ void xfrm_state_info_print(struct xfrm_usersa_info *xsinfo,
XFRM_FLAG_PRINT(fp, flags, XFRM_STATE_ICMP, "icmp");
XFRM_FLAG_PRINT(fp, flags, XFRM_STATE_AF_UNSPEC, "af-unspec");
XFRM_FLAG_PRINT(fp, flags, XFRM_STATE_ALIGN4, "align4");
+ XFRM_FLAG_PRINT(fp, flags, XFRM_STATE_ESN, "replay-esn");
if (flags)
fprintf(fp, "%x", flags);
}
diff --git a/ip/xfrm_state.c b/ip/xfrm_state.c
index 8ac3437..9a25ad3 100644
--- a/ip/xfrm_state.c
+++ b/ip/xfrm_state.c
@@ -51,6 +51,7 @@
#define RTA_BUF_SIZE 2048
#define XFRM_ALGO_KEY_BUF_SIZE 512
#define CTX_BUF_SIZE 256
+#define XFRM_REPLAY_BMP_SIZE_U32 16
static void usage(void) __attribute__((noreturn));
@@ -59,7 +60,7 @@ static void usage(void)
fprintf(stderr, "Usage: ip xfrm state { add | update } ID [ XFRM_OPT ] [ ctx SEC_CTX ] [ mode MODE ]\n");
fprintf(stderr, " [ reqid REQID ] [ seq SEQ ] [ replay-window SIZE ] [ flag FLAG-LIST ]\n");
fprintf(stderr, " [ encap ENCAP ] [ sel SELECTOR ] [ replay-seq SEQ ]\n");
- fprintf(stderr, " [ replay-oseq SEQ ] [ LIMIT-LIST ]\n");
+ fprintf(stderr, " [ replay-oseq SEQ ] [ replay-seqhi SEQ ] [ replay-oseqhi SEQ ] [ LIMIT-LIST ]\n");
fprintf(stderr, "Usage: ip xfrm state allocspi ID [ mode MODE ] [ reqid REQID ] [ seq SEQ ]\n");
fprintf(stderr, " [ min SPI max SPI ]\n");
fprintf(stderr, "Usage: ip xfrm state { delete | get } ID\n");
@@ -218,6 +219,8 @@ static int xfrm_state_flag_parse(__u8 *flags, int *argcp, char ***argvp)
*flags |= XFRM_STATE_AF_UNSPEC;
else if (strcmp(*argv, "align4") == 0)
*flags |= XFRM_STATE_ALIGN4;
+ else if (strcmp(*argv, "replay-esn") == 0)
+ *flags |= XFRM_STATE_ESN;
else {
PREV_ARG(); /* back track */
break;
@@ -243,7 +246,11 @@ static int xfrm_state_modify(int cmd, unsigned flags, int argc, char **argv)
struct xfrm_usersa_info xsinfo;
char buf[RTA_BUF_SIZE];
} req;
- struct xfrm_replay_state replay;
+ struct {
+ struct xfrm_replay_state_esn state;
+ __u32 bmp[XFRM_REPLAY_BMP_SIZE_U32];
+ } replay;
+
char *idp = NULL;
char *aeadop = NULL;
char *ealgop = NULL;
@@ -258,9 +265,12 @@ static int xfrm_state_modify(int cmd, unsigned flags, int argc, char **argv)
} ctx;
memset(&req, 0, sizeof(req));
+
memset(&replay, 0, sizeof(replay));
memset(&ctx, 0, sizeof(ctx));
+ replay.state.bmp_len = XFRM_REPLAY_BMP_SIZE_U32;
+
req.n.nlmsg_len = NLMSG_LENGTH(sizeof(req.xsinfo));
req.n.nlmsg_flags = NLM_F_REQUEST|flags;
req.n.nlmsg_type = cmd;
@@ -285,16 +295,24 @@ static int xfrm_state_modify(int cmd, unsigned flags, int argc, char **argv)
xfrm_seq_parse(&req.xsinfo.seq, &argc, &argv);
} else if (strcmp(*argv, "replay-window") == 0) {
NEXT_ARG();
- if (get_u8(&req.xsinfo.replay_window, *argv, 0))
+ if (get_u32(&replay.state.replay_window, *argv, 0))
invarg("\"replay-window\" value is invalid", *argv);
} else if (strcmp(*argv, "replay-seq") == 0) {
NEXT_ARG();
- if (get_u32(&replay.seq, *argv, 0))
+ if (get_u32(&replay.state.seq, *argv, 0))
invarg("\"replay-seq\" value is invalid", *argv);
} else if (strcmp(*argv, "replay-oseq") == 0) {
NEXT_ARG();
- if (get_u32(&replay.oseq, *argv, 0))
+ if (get_u32(&replay.state.oseq, *argv, 0))
invarg("\"replay-oseq\" value is invalid", *argv);
+ } else if (strcmp(*argv, "replay-seqhi") == 0) {
+ NEXT_ARG();
+ if (get_u32(&replay.state.seq_hi, *argv, 0))
+ invarg("\"replay-seqhi\" value is invalid", *argv);
+ } else if (strcmp(*argv, "replay-oseqhi") == 0) {
+ NEXT_ARG();
+ if (get_u32(&replay.state.oseq_hi, *argv, 0))
+ invarg("\"replay-oseqhi\" value is invalid", *argv);
} else if (strcmp(*argv, "flag") == 0) {
NEXT_ARG();
xfrm_state_flag_parse(&req.xsinfo.flags, &argc, &argv);
@@ -472,9 +490,12 @@ static int xfrm_state_modify(int cmd, unsigned flags, int argc, char **argv)
argc--; argv++;
}
- if (replay.seq || replay.oseq)
- addattr_l(&req.n, sizeof(req.buf), XFRMA_REPLAY_VAL,
+ if (replay.state.replay_window || replay.state.seq || replay.state.oseq ||
+ replay.state.seq_hi || replay.state.oseq_hi ) {
+
+ addattr_l(&req.n, sizeof(req.buf), XFRMA_REPLAY_ESN_VAL,
(void *)&replay, sizeof(replay));
+ }
if (!idp) {
fprintf(stderr, "Not enough information: \"ID\" is required\n");
--
1.7.0.4
^ permalink raw reply related
* [PATCH] Fix "possible circular locking dependency" in rfcomm
From: Ferraton, Jean RegisX @ 2011-03-08 10:10 UTC (permalink / raw)
To: marcel@holtmann.org, davem@davemloft.net, eric.dumazet@gmail.com,
xiaosuo@gmail.com
Changes in sock.c in function __rfcomm_sock_close, to release sock before calling rfcomm_dlc_close to avoid possible deadlock when calling rfcomm_lock (in rfcomm_dlc_close).
Signed-off-by: Jean Regis Ferraton <jrferraton@gmail.com>
---
net/bluetooth/rfcomm/sock.c | 3 +++
1 files changed, 3 insertions(+), 0 deletions(-)
diff --git a/net/bluetooth/rfcomm/sock.c b/net/bluetooth/rfcomm/sock.c
index 66cc1f0..3b5ec09 100644
--- a/net/bluetooth/rfcomm/sock.c
+++ b/net/bluetooth/rfcomm/sock.c
@@ -234,7 +234,10 @@ static void __rfcomm_sock_close(struct sock *sk)
case BT_CONNECT2:
case BT_CONFIG:
case BT_CONNECTED:
+ /* release_sock to avoid deadlock when calling rfcomm_lock() */
+ release_sock(sk);
rfcomm_dlc_close(d, 0);
+ lock_sock(sk);
default:
sock_set_flag(sk, SOCK_ZAPPED);
--
1.6.2.5
---------------------------------------------------------------------
Intel Corporation SAS (French simplified joint stock company)
Registered headquarters: "Les Montalets"- 2, rue de Paris,
92196 Meudon Cedex, France
Registration Number: 302 456 199 R.C.S. NANTERRE
Capital: 4,572,000 Euros
This e-mail and any attachments may contain confidential material for
the sole use of the intended recipient(s). Any review or distribution
by others is strictly prohibited. If you are not the intended
recipient, please contact the sender and delete all copies.
^ permalink raw reply related
* [RFC v2 PATCH 9/9] xfrm: Add user interface for esn and big anti-replay windows
From: Steffen Klassert @ 2011-03-08 10:10 UTC (permalink / raw)
To: Herbert Xu, David Miller
Cc: Alex Badea, Andreas Gruenbacher, netdev, linux-crypto
In-Reply-To: <20110308100407.GB31402@secunet.com>
This patch adds a netlink based user interface to configure
esn and big anti-replay windows. The new netlink attribute
XFRMA_REPLAY_ESN_VAL is used to configure the new implementation.
If the XFRM_STATE_ESN flag is set, we use esn and support for big
anti-replay windows for the configured state. If this flag is not
set we use the new implementation with 32 bit sequence numbers.
A big anti-replay window can be configured in this case anyway.
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
---
net/xfrm/xfrm_state.c | 2 +
net/xfrm/xfrm_user.c | 99 ++++++++++++++++++++++++++++++++++++++++++-------
2 files changed, 87 insertions(+), 14 deletions(-)
diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c
index 104cddf..47e37bf 100644
--- a/net/xfrm/xfrm_state.c
+++ b/net/xfrm/xfrm_state.c
@@ -356,6 +356,8 @@ static void xfrm_state_gc_destroy(struct xfrm_state *x)
kfree(x->calg);
kfree(x->encap);
kfree(x->coaddr);
+ kfree(x->replay_esn);
+ kfree(x->preplay_esn);
if (x->inner_mode)
xfrm_put_mode(x->inner_mode);
if (x->inner_mode_iaf)
diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c
index f7b3c85..706385a 100644
--- a/net/xfrm/xfrm_user.c
+++ b/net/xfrm/xfrm_user.c
@@ -119,6 +119,19 @@ static inline int verify_sec_ctx_len(struct nlattr **attrs)
return 0;
}
+static inline int verify_replay(struct xfrm_usersa_info *p,
+ struct nlattr **attrs)
+{
+ struct nlattr *rt = attrs[XFRMA_REPLAY_ESN_VAL];
+
+ if (!rt)
+ return 0;
+
+ if (p->replay_window != 0)
+ return -EINVAL;
+
+ return 0;
+}
static int verify_newsa_info(struct xfrm_usersa_info *p,
struct nlattr **attrs)
@@ -214,6 +227,8 @@ static int verify_newsa_info(struct xfrm_usersa_info *p,
goto out;
if ((err = verify_sec_ctx_len(attrs)))
goto out;
+ if ((err = verify_replay(p, attrs)))
+ goto out;
err = -EINVAL;
switch (p->mode) {
@@ -345,6 +360,33 @@ static int attach_aead(struct xfrm_algo_aead **algpp, u8 *props,
return 0;
}
+static int xfrm_alloc_replay_state_esn(struct xfrm_replay_state_esn **replay_esn,
+ struct xfrm_replay_state_esn **preplay_esn,
+ struct nlattr *rta)
+{
+ struct xfrm_replay_state_esn *p, *pp, *up;
+
+ if (!rta)
+ return 0;
+
+ up = nla_data(rta);
+
+ p = kmemdup(up, xfrm_replay_state_esn_len(up), GFP_KERNEL);
+ if (!p)
+ return -ENOMEM;
+
+ pp = kmemdup(up, xfrm_replay_state_esn_len(up), GFP_KERNEL);
+ if (!pp) {
+ kfree(p);
+ return -ENOMEM;
+ }
+
+ *replay_esn = p;
+ *preplay_esn = pp;
+
+ return 0;
+}
+
static inline int xfrm_user_sec_ctx_size(struct xfrm_sec_ctx *xfrm_ctx)
{
int len = 0;
@@ -380,10 +422,20 @@ static void copy_from_user_state(struct xfrm_state *x, struct xfrm_usersa_info *
static void xfrm_update_ae_params(struct xfrm_state *x, struct nlattr **attrs)
{
struct nlattr *rp = attrs[XFRMA_REPLAY_VAL];
+ struct nlattr *re = attrs[XFRMA_REPLAY_ESN_VAL];
struct nlattr *lt = attrs[XFRMA_LTIME_VAL];
struct nlattr *et = attrs[XFRMA_ETIMER_THRESH];
struct nlattr *rt = attrs[XFRMA_REPLAY_THRESH];
+ if (re) {
+ struct xfrm_replay_state_esn *replay_esn;
+ replay_esn = nla_data(re);
+ memcpy(x->replay_esn, replay_esn,
+ xfrm_replay_state_esn_len(replay_esn));
+ memcpy(x->preplay_esn, replay_esn,
+ xfrm_replay_state_esn_len(replay_esn));
+ }
+
if (rp) {
struct xfrm_replay_state *replay;
replay = nla_data(rp);
@@ -467,13 +519,14 @@ static struct xfrm_state *xfrm_state_construct(struct net *net,
security_xfrm_state_alloc(x, nla_data(attrs[XFRMA_SEC_CTX])))
goto error;
+ if ((err = xfrm_alloc_replay_state_esn(&x->replay_esn, &x->preplay_esn,
+ attrs[XFRMA_REPLAY_ESN_VAL])))
+ goto error;
+
x->km.seq = p->seq;
x->replay_maxdiff = net->xfrm.sysctl_aevent_rseqth;
/* sysctl_xfrm_aevent_etime is in 100ms units */
x->replay_maxage = (net->xfrm.sysctl_aevent_etime*HZ)/XFRM_AE_ETH_M;
- x->preplay.bitmap = 0;
- x->preplay.seq = x->replay.seq+x->replay_maxdiff;
- x->preplay.oseq = x->replay.oseq +x->replay_maxdiff;
if ((err = xfrm_init_replay(x)))
goto error;
@@ -709,6 +762,10 @@ static int copy_to_user_state_extra(struct xfrm_state *x,
if (xfrm_mark_put(skb, &x->mark))
goto nla_put_failure;
+ if (x->replay_esn)
+ NLA_PUT(skb, XFRMA_REPLAY_ESN_VAL,
+ xfrm_replay_state_esn_len(x->replay_esn), x->replay_esn);
+
if (x->security && copy_sec_ctx(x->security, skb) < 0)
goto nla_put_failure;
@@ -1578,10 +1635,14 @@ static int xfrm_flush_sa(struct sk_buff *skb, struct nlmsghdr *nlh,
return 0;
}
-static inline size_t xfrm_aevent_msgsize(void)
+static inline size_t xfrm_aevent_msgsize(struct xfrm_state *x)
{
+ size_t replay_size = x->replay_esn ?
+ xfrm_replay_state_esn_len(x->replay_esn) :
+ sizeof(struct xfrm_replay_state);
+
return NLMSG_ALIGN(sizeof(struct xfrm_aevent_id))
- + nla_total_size(sizeof(struct xfrm_replay_state))
+ + nla_total_size(replay_size)
+ nla_total_size(sizeof(struct xfrm_lifetime_cur))
+ nla_total_size(sizeof(struct xfrm_mark))
+ nla_total_size(4) /* XFRM_AE_RTHR */
@@ -1606,7 +1667,13 @@ static int build_aevent(struct sk_buff *skb, struct xfrm_state *x, const struct
id->reqid = x->props.reqid;
id->flags = c->data.aevent;
- NLA_PUT(skb, XFRMA_REPLAY_VAL, sizeof(x->replay), &x->replay);
+ if (x->replay_esn)
+ NLA_PUT(skb, XFRMA_REPLAY_ESN_VAL,
+ xfrm_replay_state_esn_len(x->replay_esn),
+ x->replay_esn);
+ else
+ NLA_PUT(skb, XFRMA_REPLAY_VAL, sizeof(x->replay), &x->replay);
+
NLA_PUT(skb, XFRMA_LTIME_VAL, sizeof(x->curlft), &x->curlft);
if (id->flags & XFRM_AE_RTHR)
@@ -1639,16 +1706,16 @@ static int xfrm_get_ae(struct sk_buff *skb, struct nlmsghdr *nlh,
struct xfrm_aevent_id *p = nlmsg_data(nlh);
struct xfrm_usersa_id *id = &p->sa_id;
- r_skb = nlmsg_new(xfrm_aevent_msgsize(), GFP_ATOMIC);
- if (r_skb == NULL)
- return -ENOMEM;
-
mark = xfrm_mark_get(attrs, &m);
x = xfrm_state_lookup(net, mark, &id->daddr, id->spi, id->proto, id->family);
- if (x == NULL) {
- kfree_skb(r_skb);
+ if (x == NULL)
return -ESRCH;
+
+ r_skb = nlmsg_new(xfrm_aevent_msgsize(x), GFP_ATOMIC);
+ if (r_skb == NULL) {
+ xfrm_state_put(x);
+ return -ENOMEM;
}
/*
@@ -1680,9 +1747,10 @@ static int xfrm_new_ae(struct sk_buff *skb, struct nlmsghdr *nlh,
struct xfrm_mark m;
struct xfrm_aevent_id *p = nlmsg_data(nlh);
struct nlattr *rp = attrs[XFRMA_REPLAY_VAL];
+ struct nlattr *re = attrs[XFRMA_REPLAY_ESN_VAL];
struct nlattr *lt = attrs[XFRMA_LTIME_VAL];
- if (!lt && !rp)
+ if (!lt && !rp && !re)
return err;
/* pedantic mode - thou shalt sayeth replaceth */
@@ -2147,6 +2215,7 @@ static const struct nla_policy xfrma_policy[XFRMA_MAX+1] = {
[XFRMA_KMADDRESS] = { .len = sizeof(struct xfrm_user_kmaddress) },
[XFRMA_MARK] = { .len = sizeof(struct xfrm_mark) },
[XFRMA_TFCPAD] = { .type = NLA_U32 },
+ [XFRMA_REPLAY_ESN_VAL] = { .len = sizeof(struct xfrm_replay_state_esn) },
};
static struct xfrm_link {
@@ -2274,7 +2343,7 @@ static int xfrm_aevent_state_notify(struct xfrm_state *x, const struct km_event
struct net *net = xs_net(x);
struct sk_buff *skb;
- skb = nlmsg_new(xfrm_aevent_msgsize(), GFP_ATOMIC);
+ skb = nlmsg_new(xfrm_aevent_msgsize(x), GFP_ATOMIC);
if (skb == NULL)
return -ENOMEM;
@@ -2328,6 +2397,8 @@ static inline size_t xfrm_sa_len(struct xfrm_state *x)
l += nla_total_size(sizeof(*x->encap));
if (x->tfcpad)
l += nla_total_size(sizeof(x->tfcpad));
+ if (x->replay_esn)
+ l += nla_total_size(xfrm_replay_state_esn_len(x->replay_esn));
if (x->security)
l += nla_total_size(sizeof(struct xfrm_user_sec_ctx) +
x->security->ctx_len);
--
1.7.0.4
^ permalink raw reply related
* [RFC v2 PATCH 8/9] xfrm: Add support for IPsec extended sequence numbers
From: Steffen Klassert @ 2011-03-08 10:09 UTC (permalink / raw)
To: Herbert Xu, David Miller
Cc: Alex Badea, Andreas Gruenbacher, netdev, linux-crypto
In-Reply-To: <20110308100407.GB31402@secunet.com>
This patch adds support for IPsec extended sequence numbers (esn)
as defined in RFC 4303. The bits to manage the anti-replay window
are based on a patch from Alex Badea.
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
---
include/net/xfrm.h | 1 +
net/xfrm/xfrm_input.c | 4 +
net/xfrm/xfrm_replay.c | 190 +++++++++++++++++++++++++++++++++++++++++++++++-
3 files changed, 194 insertions(+), 1 deletions(-)
diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 570eb77..a556f56 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -1427,6 +1427,7 @@ extern int xfrm_state_delete(struct xfrm_state *x);
extern int xfrm_state_flush(struct net *net, u8 proto, struct xfrm_audit *audit_info);
extern void xfrm_sad_getinfo(struct net *net, struct xfrmk_sadinfo *si);
extern void xfrm_spd_getinfo(struct net *net, struct xfrmk_spdinfo *si);
+extern u32 xfrm_replay_seqhi(struct xfrm_state *x, __be32 net_seq);
extern int xfrm_init_replay(struct xfrm_state *x);
extern int xfrm_state_mtu(struct xfrm_state *x, int mtu);
extern int xfrm_init_state(struct xfrm_state *x);
diff --git a/net/xfrm/xfrm_input.c b/net/xfrm/xfrm_input.c
index 55d5f5c..872065c 100644
--- a/net/xfrm/xfrm_input.c
+++ b/net/xfrm/xfrm_input.c
@@ -107,6 +107,7 @@ int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
struct net *net = dev_net(skb->dev);
int err;
__be32 seq;
+ __be32 seq_hi;
struct xfrm_state *x;
xfrm_address_t *daddr;
struct xfrm_mode *inner_mode;
@@ -184,7 +185,10 @@ int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
spin_unlock(&x->lock);
+ seq_hi = htonl(xfrm_replay_seqhi(x, seq));
+
XFRM_SKB_CB(skb)->seq.input.low = seq;
+ XFRM_SKB_CB(skb)->seq.input.hi = seq_hi;
nexthdr = x->type->input(x, skb);
diff --git a/net/xfrm/xfrm_replay.c b/net/xfrm/xfrm_replay.c
index 50589ea..2f5be5b 100644
--- a/net/xfrm/xfrm_replay.c
+++ b/net/xfrm/xfrm_replay.c
@@ -20,6 +20,31 @@
#include <net/xfrm.h>
+u32 xfrm_replay_seqhi(struct xfrm_state *x, __be32 net_seq)
+{
+ u32 seq, seq_hi, bottom;
+ struct xfrm_replay_state_esn *replay_esn = x->replay_esn;
+
+ if (!(x->props.flags & XFRM_STATE_ESN))
+ return 0;
+
+ seq = ntohl(net_seq);
+ seq_hi = replay_esn->seq_hi;
+ bottom = replay_esn->seq - replay_esn->replay_window + 1;
+
+ if (likely(replay_esn->seq >= replay_esn->replay_window - 1)) {
+ /* A. same subspace */
+ if (unlikely(seq < bottom))
+ seq_hi++;
+ } else {
+ /* B. window spans two subspaces */
+ if (unlikely(seq >= bottom))
+ seq_hi--;
+ }
+
+ return seq_hi;
+}
+
static void xfrm_replay_notify(struct xfrm_state *x, int event)
{
struct km_event c;
@@ -313,6 +338,160 @@ static void xfrm_replay_notify_bmp(struct xfrm_state *x, int event)
x->xflags &= ~XFRM_TIME_DEFER;
}
+static int xfrm_replay_overflow_esn(struct xfrm_state *x, struct sk_buff *skb)
+{
+ int err = 0;
+ struct xfrm_replay_state_esn *replay_esn = x->replay_esn;
+ struct net *net = xs_net(x);
+
+ if (x->type->flags & XFRM_TYPE_REPLAY_PROT) {
+ XFRM_SKB_CB(skb)->seq.output.low = ++replay_esn->oseq;
+ XFRM_SKB_CB(skb)->seq.output.hi = replay_esn->oseq_hi;
+
+ if (unlikely(replay_esn->oseq == 0)) {
+ XFRM_SKB_CB(skb)->seq.output.hi = ++replay_esn->oseq_hi;
+
+ if (replay_esn->oseq_hi == 0) {
+ replay_esn->oseq--;
+ replay_esn->oseq_hi--;
+ xfrm_audit_state_replay_overflow(x, skb);
+ err = -EOVERFLOW;
+
+ return err;
+ }
+ }
+ if (xfrm_aevent_is_on(net))
+ x->repl->notify(x, XFRM_REPLAY_UPDATE);
+ }
+
+ return err;
+}
+
+static int xfrm_replay_check_esn(struct xfrm_state *x,
+ struct sk_buff *skb, __be32 net_seq)
+{
+ unsigned int bitnr, nr;
+ u32 diff;
+ struct xfrm_replay_state_esn *replay_esn = x->replay_esn;
+ u32 seq = ntohl(net_seq);
+ u32 pos = (replay_esn->seq - 1) % replay_esn->replay_window;
+ u32 wsize = replay_esn->replay_window;
+ u32 top = replay_esn->seq;
+ u32 bottom = top - wsize + 1;
+
+ if (unlikely(seq == 0 && replay_esn->seq_hi == 0 &&
+ (replay_esn->seq < replay_esn->replay_window - 1)))
+ goto err;
+
+ diff = top - seq;
+
+ if (likely(top >= wsize - 1)) {
+ /* A. same subspace */
+ if (likely(seq > top) || seq < bottom)
+ return 0;
+ } else {
+ /* B. window spans two subspaces */
+ if (likely(seq > top && seq < bottom))
+ return 0;
+ if (seq >= bottom)
+ diff = ~seq + top + 1;
+ }
+
+ if (diff >= replay_esn->replay_window) {
+ x->stats.replay_window++;
+ goto err;
+ }
+
+ if (pos >= diff) {
+ bitnr = (pos - diff) % replay_esn->replay_window;
+ nr = bitnr >> 5;
+ bitnr = bitnr & 0x1F;
+ if (replay_esn->bmp[nr] & (1U << bitnr))
+ goto err_replay;
+ } else {
+ bitnr = replay_esn->replay_window - (diff - pos);
+ nr = bitnr >> 5;
+ bitnr = bitnr & 0x1F;
+ if (replay_esn->bmp[nr] & (1U << bitnr))
+ goto err_replay;
+ }
+ return 0;
+
+err_replay:
+ x->stats.replay++;
+err:
+ xfrm_audit_state_replay(x, skb, net_seq);
+ return -EINVAL;
+}
+
+static void xfrm_replay_advance_esn(struct xfrm_state *x, __be32 net_seq)
+{
+ unsigned int bitnr, nr, i;
+ int wrap;
+ u32 diff, pos, seq, seq_hi;
+ struct xfrm_replay_state_esn *replay_esn = x->replay_esn;
+
+ if (!replay_esn->replay_window)
+ return;
+
+ seq = ntohl(net_seq);
+ pos = (replay_esn->seq - 1) % replay_esn->replay_window;
+ seq_hi = xfrm_replay_seqhi(x, net_seq);
+ wrap = seq_hi - replay_esn->seq_hi;
+
+ if ((!wrap && seq > replay_esn->seq) || wrap > 0) {
+ if (likely(!wrap))
+ diff = seq - replay_esn->seq;
+ else
+ diff = ~replay_esn->seq + seq + 1;
+
+ if (diff < replay_esn->replay_window) {
+ for (i = 1; i < diff; i++) {
+ bitnr = (pos + i) % replay_esn->replay_window;
+ nr = bitnr >> 5;
+ bitnr = bitnr & 0x1F;
+ replay_esn->bmp[nr] &= ~(1U << bitnr);
+ }
+
+ bitnr = (pos + diff) % replay_esn->replay_window;
+ nr = bitnr >> 5;
+ bitnr = bitnr & 0x1F;
+ replay_esn->bmp[nr] |= (1U << bitnr);
+ } else {
+ nr = replay_esn->replay_window >> 5;
+ for (i = 0; i <= nr; i++)
+ replay_esn->bmp[i] = 0;
+
+ bitnr = (pos + diff) % replay_esn->replay_window;
+ nr = bitnr >> 5;
+ bitnr = bitnr & 0x1F;
+ replay_esn->bmp[nr] |= (1U << bitnr);
+ }
+
+ replay_esn->seq = seq;
+
+ if (unlikely(wrap > 0))
+ replay_esn->seq_hi++;
+ } else {
+ diff = replay_esn->seq - seq;
+
+ if (pos >= diff) {
+ bitnr = (pos - diff) % replay_esn->replay_window;
+ nr = bitnr >> 5;
+ bitnr = bitnr & 0x1F;
+ replay_esn->bmp[nr] |= (1U << bitnr);
+ } else {
+ bitnr = replay_esn->replay_window - (diff - pos);
+ nr = bitnr >> 5;
+ bitnr = bitnr & 0x1F;
+ replay_esn->bmp[nr] |= (1U << bitnr);
+ }
+ }
+
+ if (xfrm_aevent_is_on(xs_net(x)))
+ xfrm_replay_notify(x, XFRM_REPLAY_UPDATE);
+}
+
static struct xfrm_replay xfrm_replay_legacy = {
.advance = xfrm_replay_advance,
.check = xfrm_replay_check,
@@ -327,6 +506,13 @@ static struct xfrm_replay xfrm_replay_bmp = {
.overflow = xfrm_replay_overflow_bmp,
};
+static struct xfrm_replay xfrm_replay_esn = {
+ .advance = xfrm_replay_advance_esn,
+ .check = xfrm_replay_check_esn,
+ .notify = xfrm_replay_notify_bmp,
+ .overflow = xfrm_replay_overflow_esn,
+};
+
int xfrm_init_replay(struct xfrm_state *x)
{
struct xfrm_replay_state_esn *replay_esn = x->replay_esn;
@@ -336,11 +522,13 @@ int xfrm_init_replay(struct xfrm_state *x)
replay_esn->bmp_len * sizeof(__u32))
return -EINVAL;
+ if ((x->props.flags & XFRM_STATE_ESN) && x->replay_esn)
+ x->repl = &xfrm_replay_esn;
+ else
x->repl = &xfrm_replay_bmp;
} else
x->repl = &xfrm_replay_legacy;
-
return 0;
}
EXPORT_SYMBOL(xfrm_init_replay);
--
1.7.0.4
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox