* Re: via-rhine: Problem with lost link after a while
From: Bjarke Istrup Pedersen @ 2012-04-13 6:16 UTC (permalink / raw)
To: Francois Romieu
Cc: linux-kernel, netdev, Svenning Sørensen, Andreas Mohr
In-Reply-To: <20120410225534.GA28480@electric-eye.fr.zoreil.com>
11. apr. 2012 00.55 skrev Francois Romieu <romieu@fr.zoreil.com>:
> Bjarke Istrup Pedersen <gurligebis@gentoo.org> :
>> 10. apr. 2012 22.42 skrev Francois Romieu <romieu@fr.zoreil.com>:
> [...]
>> > http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git;h=3f8c91a7398b9266fbe7abcbe4bd5dffef907643
> [...]
>> Great, I'll try a 3.4-rc2 kernel, and see how it runs.
>>
>> The thread I was talking about earlier is here:
>> http://lists.soekris.com/pipermail/soekris-tech/2012-April/018318.html
>> Is there any of the changes he has there, that makes sense in the new
>> driver you wrote ?
>
> (I did not write a new driver)
>
> Regarding Svenning's patch:
> - the wmb in alloc_rbufs may help rhine_reset_task().
> - one should probably add one in rhine_rx() as well.
> - rhine_start_tx() is supposed to stop queueing when there is no room left.
> I'm curious to know if the "Tx descriptor busy" test triggered.
> - the rmb() in rhine_tx() will not make a difference for a single core but
> it's a good reminder that I should not have forgotten to propagate the
> xmit / Tx completion fix back from the r8169 driver to the via-rhine one
> (sigh)
>
> mmiowb is probably missing. I doubt it hits hard right now.
>
> I have not checked if MMIO flushes are missing. Actually I need some sleep.
I've been testing v3.4-rc2 with this patch applied for the past few
days, with alot of trafic across, both incoming and outgoing, with VPN
which is normally what can trigger problems, and it seems rock solid
:o)
Do you need me to do any specific testing of this patch?
/Bjarke
> diff --git a/drivers/net/ethernet/via/via-rhine.c b/drivers/net/ethernet/via/via-rhine.c
> index fcfa01f..dfa9fc0 100644
> --- a/drivers/net/ethernet/via/via-rhine.c
> +++ b/drivers/net/ethernet/via/via-rhine.c
> @@ -1163,6 +1163,7 @@ static void alloc_rbufs(struct net_device *dev)
> PCI_DMA_FROMDEVICE);
>
> rp->rx_ring[i].addr = cpu_to_le32(rp->rx_skbuff_dma[i]);
> + wmb();
> rp->rx_ring[i].rx_status = cpu_to_le32(DescOwn);
> }
> rp->dirty_rx = (unsigned int)(i - RX_RING_SIZE);
> @@ -1709,8 +1710,13 @@ static netdev_tx_t rhine_start_tx(struct sk_buff *skb,
> ioaddr + ChipCmd1);
> IOSYNC;
>
> - if (rp->cur_tx == rp->dirty_tx + TX_QUEUE_LEN)
> + if (rp->cur_tx == rp->dirty_tx + TX_QUEUE_LEN) {
> + smp_wmb();
> netif_stop_queue(dev);
> + smp_mb();
> + if (rp->cur_tx != rp->dirty_tx + TX_QUEUE_LEN)
> + netif_wake_queue(dev);
> + }
>
> netif_dbg(rp, tx_queued, dev, "Transmit frame #%d queued in slot %d\n",
> rp->cur_tx - 1, entry);
> @@ -1759,6 +1765,7 @@ static void rhine_tx(struct net_device *dev)
> struct rhine_private *rp = netdev_priv(dev);
> int txstatus = 0, entry = rp->dirty_tx % TX_RING_SIZE;
>
> + smp_rmb();
> /* find and cleanup dirty tx descriptors */
> while (rp->dirty_tx != rp->cur_tx) {
> txstatus = le32_to_cpu(rp->tx_ring[entry].tx_status);
> @@ -1806,8 +1813,12 @@ static void rhine_tx(struct net_device *dev)
> rp->tx_skbuff[entry] = NULL;
> entry = (++rp->dirty_tx) % TX_RING_SIZE;
> }
> - if ((rp->cur_tx - rp->dirty_tx) < TX_QUEUE_LEN - 4)
> +
> + smp_mb();
> + if (netif_queue_stopped(dev) &&
> + (rp->cur_tx - rp->dirty_tx) < TX_QUEUE_LEN - 4) {
> netif_wake_queue(dev);
> + }
> }
>
> /**
> @@ -1947,6 +1958,7 @@ static int rhine_rx(struct net_device *dev, int limit)
> PCI_DMA_FROMDEVICE);
> rp->rx_ring[entry].addr = cpu_to_le32(rp->rx_skbuff_dma[entry]);
> }
> + wmb();
> rp->rx_ring[entry].rx_status = cpu_to_le32(DescOwn);
> }
>
> --
> Ueimor
>
> Will code drivers for food.
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
^ permalink raw reply
* Re: [PATCH] netfilter: ipvs: Verify that IP_VS protocol has been registered
From: Simon Horman @ 2012-04-13 6:22 UTC (permalink / raw)
To: Sasha Levin
Cc: Pablo Neira Ayuso, wensong, ja, kaber, davem, davej, netdev,
lvs-devel, netfilter-devel, linux-kernel
In-Reply-To: <20120413041150.GA26149@verge.net.au>
On Fri, Apr 13, 2012 at 01:11:50PM +0900, Simon Horman wrote:
> On Fri, Apr 13, 2012 at 02:54:13AM +0200, Sasha Levin wrote:
> > On Thu, Apr 12, 2012 at 1:46 AM, Pablo Neira Ayuso <pablo@netfilter.org> wrote:
> > >> If you return here, I think you'll leave things in inconsistent state,
> > >> ie. the tcp protocol is registered. You have to unregister it before
> > >> leaving.
> >
> > I thought that the cleanup callback is getting called for failed init
> > calls, if that's not the case then we can probably call it ourselves
> > if any of these failed.
>
> Good point. In any case, I think that I have found a new problem.
[sni@]
I have confirmed that the cleanup code is called on error, so I think that
your existing patch is fine.
Acked-by: Simon Horman <horms@verge.net.au>
Pablo, please consider taking this change. The other problem I raised,
though related, can be handled separately as it won't occur if the
ftp helper isn't initialised.
^ permalink raw reply
* Re: [net-next 3/4 (V3)] stmmac: add the Energy Efficient Ethernet support
From: Giuseppe CAVALLARO @ 2012-04-13 6:35 UTC (permalink / raw)
To: David Miller; +Cc: netdev, bhutchings, rayagond
In-Reply-To: <20120412.163258.602029261264172098.davem@davemloft.net>
On 4/12/2012 10:32 PM, David Miller wrote:
> From: Giuseppe CAVALLARO <peppe.cavallaro@st.com>
> Date: Fri, 6 Apr 2012 11:29:17 +0200
>
>> +static int stmmac_ethtool_op_set_eee(struct net_device *dev,
>> + struct ethtool_value *eee)
>> +{
>> + struct stmmac_priv *priv = netdev_priv(dev);
>> +
>> + if ((!eee->data) && (priv->eee_enabled)) {
>> + stmmac_disable_eee_mode(priv);
>> + priv->eee_enabled = eee->data;
>> + } else if ((eee->data) && (!priv->eee_enabled))
>> + /* We are asking for enabling the EEE but this
>> + * has to be verified by invoking the eee_init function.
>> + * For this reason we cannot set eee_enabled to
>> + * eee->data, directly. */
>> + priv->eee_enabled = stmmac_eee_init(priv);
>> +
>> + return 0;
>
> If stmmac_eee_init() determines that it cannot enable eee, you should
> return an appropriate error code here so that ethtool can report that
> fact.
Ok, thanks
peppe
^ permalink raw reply
* Re: [PATCH net-next v2] tcp: bind() use stronger condition for bind_conflict
From: Eric Dumazet @ 2012-04-13 7:30 UTC (permalink / raw)
To: Alexandru Copot
Cc: davem, fbl, kuznet, jmorris, yoshfuji, kaber, netdev,
Daniel Baluta
In-Reply-To: <1333698444-24144-1-git-send-email-alex.mihai.c@gmail.com>
Le vendredi 06 avril 2012 à 10:47 +0300, Alexandru Copot a écrit :
> From: Alex Copot <alex.mihai.c@gmail.com>
>
> We must try harder to get unique (addr, port) pairs when
> doing port autoselection for sockets with SO_REUSEADDR
> option set.
>
...
Excellent patch Alex !
Two minor coding style problems in this part :
> + if (!relax && reuse && sk2->sk_reuse &&
> + sk2->sk_state != TCP_LISTEN) {
> + const __be32 sk2_rcv_saddr = sk_rcv_saddr(sk2);
> + if (!sk2_rcv_saddr || !sk_rcv_saddr(sk) ||
> + sk2_rcv_saddr == sk_rcv_saddr(sk))
> + break;
> + }
> }
Correct indentation, and one empty line after variable declaration :
if (!relax && reuse && sk2->sk_reuse &&
sk2->sk_state != TCP_LISTEN) {
const __be32 sk2_rcv_saddr = sk_rcv_saddr(sk2);
if (!sk2_rcv_saddr || !sk_rcv_saddr(sk) ||
sk2_rcv_saddr == sk_rcv_saddr(sk))
break;
}
if (condition_one ||
condition_two) {
foo();
instead of
if (condition_on ||
condition_two)
foo();
^ permalink raw reply
* Re: [PATCH] netfilter: ipvs: Verify that IP_VS protocol has been registered
From: Hans Schillstrom @ 2012-04-13 8:03 UTC (permalink / raw)
To: Simon Horman
Cc: Sasha Levin, Pablo Neira Ayuso, wensong@linux-vs.org, ja@ssi.bg,
kaber@trash.net, davem@davemloft.net, davej@redhat.com,
netdev@vger.kernel.org, lvs-devel@vger.kernel.org,
netfilter-devel@vger.kernel.org, linux-kernel@vger.kernel.org
In-Reply-To: <20120413041150.GA26149@verge.net.au>
Hello Simon and Sasha
Sorry for not helping been quite busy for a while
On Friday 13 April 2012 06:11:50 Simon Horman wrote:
> On Fri, Apr 13, 2012 at 02:54:13AM +0200, Sasha Levin wrote:
> > On Thu, Apr 12, 2012 at 1:46 AM, Pablo Neira Ayuso <pablo@netfilter.org> wrote:
> > >> If you return here, I think you'll leave things in inconsistent state,
> > >> ie. the tcp protocol is registered. You have to unregister it before
> > >> leaving.
> >
> > I thought that the cleanup callback is getting called for failed init
> > calls, if that's not the case then we can probably call it ourselves
> > if any of these failed.
>
> Good point. In any case, I think that I have found a new problem.
>
> With your proposed patch in place I see a panic in ftp helper registration
> in the case where protocol registration fails. I have not had any
> luck resolving this so far. Perhaps ipvs->net needs to be reset to NULL
> if initialisation fails and that can be checked when modules register
> themselves?
I think Pablo was touching the root cause,
There is a generic fault in the patch, you must take care of un_register of the protocol
i.e. my sugguestion is do like this:
#ifdef CONFIG_IP_VS_PROTO_TCP
if (register_ip_vs_proto_netns(net, &ip_vs_protocol_tcp) < 0)
goto unreg;
#endif
#ifdef CONFIG_IP_VS_PROTO_UDP
if (register_ip_vs_proto_netns(net, &ip_vs_protocol_udp) < 0)
goto unreg;
#endif
...
unreg:
ip_vs_protocol_net_cleanup(net);
return -ENOMEM;
}
and in register_ip_vs_proto_netns() kzalloc() could be changed to GFP_KERNEL
since this will allways be called from process context
>
> IPVS: Registered protocols (TCP, UDP, SCTP, AH, ESP)
> IPVS: Connection hash table configured (size=4096, memory=64Kbytes)
> IPVS: Each connection entry needs 264 bytes at least
> input: AT Translated Set 2 keyboard as /devices/platform/i8042/serio0/input/input1
> IPVS: [rr] scheduler registered.
> IPVS: [wrr] scheduler registered.
> IPVS: [lc] scheduler registered.
> IPVS: [wlc] scheduler registered.
> IPVS: [lblc] scheduler registered.
> IPVS: [lblcr] scheduler registered.
> IPVS: [dh] scheduler registered.
> IPVS: [sh] scheduler registered.
> IPVS: [sed] scheduler registered.
> IPVS: [nq] scheduler registered.
> general protection fault: 0000 [#1] SMP
> CPU 0
> Modules linked in:
>
> Pid: 1, comm: swapper/0 Not tainted 3.3.0-03812-gf51f739 #219 Bochs Bochs
> RIP: 0010:[<ffffffff811e21ba>] [<ffffffff811e21ba>] register_ip_vs_app+0x3a/0x70
> RSP: 0018:ffff880002039df0 EFLAGS: 00010246
> RAX: 002a002900280027 RBX: ffff8800020ce680 RCX: 0000000000000000
> RDX: 0000000000000000 RSI: ffff8800020ce680 RDI: ffffffff8162ab80
> RBP: ffff880002039e00 R08: 0000000000000000 R09: 0000000000000000
> R10: 0000000000000000 R11: 0000000000000001 R12: ffff880003916800
> R13: 00000000fffffff4 R14: ffffffff816e5f00 R15: ffff880003916800
> FS: 0000000000000000(0000) GS:ffff880002600000(0000) knlGS:0000000000000000
> CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
> CR2: 00000000004697e0 CR3: 0000000001605000 CR4: 00000000000006b0
> DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
> Process swapper/0 (pid: 1, threadinfo ffff880002038000, task ffff8800020375a0)
> Stack:
> ffff8800020ce680 0000000000000000 ffff880002039e50 ffffffff81675610
> ffffffff81614060 00ffffff00000006 ffffffff8149dc9d ffffffff8162b8a0
> 0000000000000000 ffffffff81675699 0000000000000000 0000000000000000
> Call Trace:
> [<ffffffff81675610>] __ip_vs_ftp_init+0x6b/0xf4
> [<ffffffff81675699>] ? __ip_vs_ftp_init+0xf4/0xf4
> [<ffffffff811a6011>] ops_init.constprop.11+0x91/0x110
> [<ffffffff81675699>] ? __ip_vs_ftp_init+0xf4/0xf4
> [<ffffffff811a60fc>] register_pernet_operations.isra.8+0x6c/0xc0
> [<ffffffff811a61e0>] register_pernet_subsys+0x20/0x40
> [<ffffffff81675593>] ? ip_vs_sed_init+0x12/0x12
> [<ffffffff816756a9>] ip_vs_ftp_init+0x10/0x12
> [<ffffffff810001ca>] do_one_initcall+0x3a/0x160
> [<ffffffff81654bae>] kernel_init+0x9b/0x115
> [<ffffffff81298bf4>] kernel_thread_helper+0x4/0x10
> [<ffffffff81654b13>] ? start_kernel+0x31d/0x31d
> [<ffffffff81298bf0>] ? gs_change+0xb/0xb
> Code: f8 48 89 f3 4c 8b a7 40 09 00 00 e8 e1 a6 ff ff 48 c7 c7 80 ab 62 81 e8 25 30 0b 00 49 8b 84 24 08 01 00 00 48 c7 c7 80 ab 62 81 <48> 89 58 08 48 89 03 49 8d 84 24 08 01 00 00 48 89 43 08 49 89
> RIP [<ffffffff811e21ba>] register_ip_vs_app+0x3a/0x70
> RSP <ffff880002039df0>
> ---[ end trace 3e5d9bede957f8b4 ]---
>
> --
> 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
>
--
Regards
Hans Schillstrom <hans.schillstrom@ericsson.com>
^ permalink raw reply
* Re: [PATCH net-next v2] tcp: bind() use stronger condition for bind_conflict
From: Alexandru Copot @ 2012-04-13 8:12 UTC (permalink / raw)
To: Eric Dumazet
Cc: davem, fbl, kuznet, jmorris, yoshfuji, kaber, netdev,
Daniel Baluta
In-Reply-To: <1334302258.9375.5.camel@edumazet-laptop>
On Fri, Apr 13, 2012 at 10:30 AM, Eric Dumazet <eric.dumazet@gmail.com> wrote:
> Excellent patch Alex !
Thanks, Eric !
> Two minor coding style problems in this part :
> Correct indentation, and one empty line after variable declaration :
I will fix those. That empty line is missing because that's the way it
was in the code above my changes, but I will add it.
Should I send a v3 or reply here with the new patch ?
Alex Copot
^ permalink raw reply
* [PATCH net-next] inet: makes syn_ack_timeout mandatory
From: Eric Dumazet @ 2012-04-13 8:16 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Gerrit Renker, dccp
From: : Eric Dumazet <edumazet@google.com>
There are two struct request_sock_ops providers, tcp and dccp.
inet_csk_reqsk_queue_prune() can avoid testing syn_ack_timeout being
NULL if we make it non NULL like syn_ack_timeout
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Cc: dccp@vger.kernel.org
---
include/linux/dccp.h | 2 ++
net/dccp/ipv4.c | 6 ++++++
net/dccp/ipv6.c | 1 +
net/ipv4/inet_connection_sock.c | 3 +--
4 files changed, 10 insertions(+), 2 deletions(-)
diff --git a/include/linux/dccp.h b/include/linux/dccp.h
index eaf95a0..d16294e 100644
--- a/include/linux/dccp.h
+++ b/include/linux/dccp.h
@@ -549,6 +549,8 @@ static inline const char *dccp_role(const struct sock *sk)
return NULL;
}
+extern void dccp_syn_ack_timeout(struct sock *sk, struct request_sock *req);
+
#endif /* __KERNEL__ */
#endif /* _LINUX_DCCP_H */
diff --git a/net/dccp/ipv4.c b/net/dccp/ipv4.c
index caf6e17..059cda9 100644
--- a/net/dccp/ipv4.c
+++ b/net/dccp/ipv4.c
@@ -574,6 +574,11 @@ static void dccp_v4_reqsk_destructor(struct request_sock *req)
kfree(inet_rsk(req)->opt);
}
+void dccp_syn_ack_timeout(struct sock *sk, struct request_sock *req)
+{
+}
+EXPORT_SYMBOL(dccp_syn_ack_timeout);
+
static struct request_sock_ops dccp_request_sock_ops __read_mostly = {
.family = PF_INET,
.obj_size = sizeof(struct dccp_request_sock),
@@ -581,6 +586,7 @@ static struct request_sock_ops dccp_request_sock_ops __read_mostly = {
.send_ack = dccp_reqsk_send_ack,
.destructor = dccp_v4_reqsk_destructor,
.send_reset = dccp_v4_ctl_send_reset,
+ .syn_ack_timeout = dccp_syn_ack_timeout,
};
int dccp_v4_conn_request(struct sock *sk, struct sk_buff *skb)
diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c
index 4dc588f..bd539b2 100644
--- a/net/dccp/ipv6.c
+++ b/net/dccp/ipv6.c
@@ -343,6 +343,7 @@ static struct request_sock_ops dccp6_request_sock_ops = {
.send_ack = dccp_reqsk_send_ack,
.destructor = dccp_v6_reqsk_destructor,
.send_reset = dccp_v6_ctl_send_reset,
+ .syn_ack_timeout = dccp_syn_ack_timeout,
};
static struct sock *dccp_v6_hnd_req(struct sock *sk,struct sk_buff *skb)
diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c
index 19d66ce..664e9fc 100644
--- a/net/ipv4/inet_connection_sock.c
+++ b/net/ipv4/inet_connection_sock.c
@@ -555,8 +555,7 @@ void inet_csk_reqsk_queue_prune(struct sock *parent,
syn_ack_recalc(req, thresh, max_retries,
queue->rskq_defer_accept,
&expire, &resend);
- if (req->rsk_ops->syn_ack_timeout)
- req->rsk_ops->syn_ack_timeout(parent, req);
+ req->rsk_ops->syn_ack_timeout(parent, req);
if (!expire &&
(!resend ||
!req->rsk_ops->rtx_syn_ack(parent, req, NULL) ||
^ permalink raw reply related
* [PATCH net-next v3] tcp: bind() use stronger condition for bind_conflict
From: Alexandru Copot @ 2012-04-13 8:21 UTC (permalink / raw)
To: davem, eric.dumazet
Cc: fbl, kuznet, jmorris, yoshfuji, kaber, netdev, Alex Copot,
Daniel Baluta
From: Alex Copot <alex.mihai.c@gmail.com>
We must try harder to get unique (addr, port) pairs when
doing port autoselection for sockets with SO_REUSEADDR
option set.
We achieve this by adding a relaxation parameter to
inet_csk_bind_conflict. When 'relax' parameter is off
we return a conflict whenever the current searched
pair (addr, port) is not unique.
This tries to address the problems reported in patch:
8d238b25b1ec22a73b1c2206f111df2faaff8285
Revert "tcp: bind() fix when many ports are bound"
Tests where ran for creating and binding(0) many sockets
on 100 IPs. The results are, on average:
* 60000 sockets, 600 ports / IP:
* 0.210 s, 620 (IP, port) duplicates without patch
* 0.219 s, no duplicates with patch
* 100000 sockets, 1000 ports / IP:
* 0.371 s, 1720 duplicates without patch
* 0.373 s, no duplicates with patch
* 200000 sockets, 2000 ports / IP:
* 0.766 s, 6900 duplicates without patch
* 0.768 s, no duplicates with patch
* 500000 sockets, 5000 ports / IP:
* 2.227 s, 41500 duplicates without patch
* 2.284 s, no duplicates with patch
Signed-off-by: Alex Copot <alex.mihai.c@gmail.com>
Signed-off-by: Daniel Baluta <dbaluta@ixiacom.com>
---
Changes from V2:
- Fixed coding style issues
Changes from V1:
- Fixed coding style
- Replaced int flag with bool
- Added test results
---
include/net/inet6_connection_sock.h | 2 +-
include/net/inet_connection_sock.h | 4 ++--
net/ipv4/inet_connection_sock.c | 18 ++++++++++++++----
net/ipv6/inet6_connection_sock.c | 2 +-
4 files changed, 18 insertions(+), 8 deletions(-)
diff --git a/include/net/inet6_connection_sock.h b/include/net/inet6_connection_sock.h
index 3207e58..1866a67 100644
--- a/include/net/inet6_connection_sock.h
+++ b/include/net/inet6_connection_sock.h
@@ -23,7 +23,7 @@ struct sock;
struct sockaddr;
extern int inet6_csk_bind_conflict(const struct sock *sk,
- const struct inet_bind_bucket *tb);
+ const struct inet_bind_bucket *tb, bool relax);
extern struct dst_entry* inet6_csk_route_req(struct sock *sk,
const struct request_sock *req);
diff --git a/include/net/inet_connection_sock.h b/include/net/inet_connection_sock.h
index dbf9aab..46c9e2c 100644
--- a/include/net/inet_connection_sock.h
+++ b/include/net/inet_connection_sock.h
@@ -60,7 +60,7 @@ struct inet_connection_sock_af_ops {
#endif
void (*addr2sockaddr)(struct sock *sk, struct sockaddr *);
int (*bind_conflict)(const struct sock *sk,
- const struct inet_bind_bucket *tb);
+ const struct inet_bind_bucket *tb, bool relax);
};
/** inet_connection_sock - INET connection oriented sock
@@ -245,7 +245,7 @@ extern struct request_sock *inet_csk_search_req(const struct sock *sk,
const __be32 raddr,
const __be32 laddr);
extern int inet_csk_bind_conflict(const struct sock *sk,
- const struct inet_bind_bucket *tb);
+ const struct inet_bind_bucket *tb, bool relax);
extern int inet_csk_get_port(struct sock *sk, unsigned short snum);
extern struct dst_entry* inet_csk_route_req(struct sock *sk,
diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c
index 19d66ce..9a3e88d 100644
--- a/net/ipv4/inet_connection_sock.c
+++ b/net/ipv4/inet_connection_sock.c
@@ -53,7 +53,7 @@ void inet_get_local_port_range(int *low, int *high)
EXPORT_SYMBOL(inet_get_local_port_range);
int inet_csk_bind_conflict(const struct sock *sk,
- const struct inet_bind_bucket *tb)
+ const struct inet_bind_bucket *tb, bool relax)
{
struct sock *sk2;
struct hlist_node *node;
@@ -79,6 +79,14 @@ int inet_csk_bind_conflict(const struct sock *sk,
sk2_rcv_saddr == sk_rcv_saddr(sk))
break;
}
+ if (!relax && reuse && sk2->sk_reuse &&
+ sk2->sk_state != TCP_LISTEN) {
+ const __be32 sk2_rcv_saddr = sk_rcv_saddr(sk2);
+
+ if (!sk2_rcv_saddr || !sk_rcv_saddr(sk) ||
+ sk2_rcv_saddr == sk_rcv_saddr(sk))
+ break;
+ }
}
}
return node != NULL;
@@ -122,12 +130,13 @@ again:
(tb->num_owners < smallest_size || smallest_size == -1)) {
smallest_size = tb->num_owners;
smallest_rover = rover;
- if (atomic_read(&hashinfo->bsockets) > (high - low) + 1) {
+ if (atomic_read(&hashinfo->bsockets) > (high - low) + 1 &&
+ !inet_csk(sk)->icsk_af_ops->bind_conflict(sk, tb, false)) {
snum = smallest_rover;
goto tb_found;
}
}
- if (!inet_csk(sk)->icsk_af_ops->bind_conflict(sk, tb)) {
+ if (!inet_csk(sk)->icsk_af_ops->bind_conflict(sk, tb, false)) {
snum = rover;
goto tb_found;
}
@@ -178,12 +187,13 @@ tb_found:
goto success;
} else {
ret = 1;
- if (inet_csk(sk)->icsk_af_ops->bind_conflict(sk, tb)) {
+ if (inet_csk(sk)->icsk_af_ops->bind_conflict(sk, tb, true)) {
if (sk->sk_reuse && sk->sk_state != TCP_LISTEN &&
smallest_size != -1 && --attempts >= 0) {
spin_unlock(&head->lock);
goto again;
}
+
goto fail_unlock;
}
}
diff --git a/net/ipv6/inet6_connection_sock.c b/net/ipv6/inet6_connection_sock.c
index 02dd203..e6cee52 100644
--- a/net/ipv6/inet6_connection_sock.c
+++ b/net/ipv6/inet6_connection_sock.c
@@ -28,7 +28,7 @@
#include <net/inet6_connection_sock.h>
int inet6_csk_bind_conflict(const struct sock *sk,
- const struct inet_bind_bucket *tb)
+ const struct inet_bind_bucket *tb, bool relax)
{
const struct sock *sk2;
const struct hlist_node *node;
--
1.7.10
^ permalink raw reply related
* Re: [PATCH] netfilter: ipvs: Verify that IP_VS protocol has been registered
From: Julian Anastasov @ 2012-04-13 8:35 UTC (permalink / raw)
To: Simon Horman
Cc: Sasha Levin, Pablo Neira Ayuso, wensong, kaber, davem, davej,
netdev, lvs-devel, netfilter-devel, linux-kernel
In-Reply-To: <20120413062250.GA6163@verge.net.au>
Hello,
On Fri, 13 Apr 2012, Simon Horman wrote:
> On Fri, Apr 13, 2012 at 01:11:50PM +0900, Simon Horman wrote:
> > On Fri, Apr 13, 2012 at 02:54:13AM +0200, Sasha Levin wrote:
> > > On Thu, Apr 12, 2012 at 1:46 AM, Pablo Neira Ayuso <pablo@netfilter.org> wrote:
> > > >> If you return here, I think you'll leave things in inconsistent state,
> > > >> ie. the tcp protocol is registered. You have to unregister it before
> > > >> leaving.
> > >
> > > I thought that the cleanup callback is getting called for failed init
> > > calls, if that's not the case then we can probably call it ourselves
> > > if any of these failed.
> >
> > Good point. In any case, I think that I have found a new problem.
>
> [sni@]
>
> I have confirmed that the cleanup code is called on error, so I think that
> your existing patch is fine.
>
> Acked-by: Simon Horman <horms@verge.net.au>
>
> Pablo, please consider taking this change. The other problem I raised,
> though related, can be handled separately as it won't occur if the
> ftp helper isn't initialised.
Simon, I'm sending 2 patches for this ftp problem
in separate thread. Please, test if the problem is solved.
Regards
--
Julian Anastasov <ja@ssi.bg>
^ permalink raw reply
* Re: [PATCH net-next v3] tcp: bind() use stronger condition for bind_conflict
From: Eric Dumazet @ 2012-04-13 8:53 UTC (permalink / raw)
To: Alexandru Copot
Cc: davem, fbl, kuznet, jmorris, yoshfuji, kaber, netdev,
Daniel Baluta
In-Reply-To: <1334305305-19952-1-git-send-email-alex.mihai.c@gmail.com>
On Fri, 2012-04-13 at 11:21 +0300, Alexandru Copot wrote:
> From: Alex Copot <alex.mihai.c@gmail.com>
>
> We must try harder to get unique (addr, port) pairs when
> doing port autoselection for sockets with SO_REUSEADDR
> option set.
...
> }
> + if (!relax && reuse && sk2->sk_reuse &&
> + sk2->sk_state != TCP_LISTEN) {
> + const __be32 sk2_rcv_saddr = sk_rcv_saddr(sk2);
> +
Oh well, missing space here :(
I said :
if (condition_one ||
condition_two) {
not
if (condition_one ||
condition_two) {
Signed-off-by: Eric Dumazet <edumazet@google.com>
Thanks
^ permalink raw reply
* Re: [PATCH 6/6] tilegx network driver: initial support
From: Arnd Bergmann @ 2012-04-13 10:34 UTC (permalink / raw)
To: Chris Metcalf; +Cc: linux-kernel, netdev
In-Reply-To: <4F8763E1.1020605@tilera.com>
On Thursday 12 April 2012, Chris Metcalf wrote:
> On 4/10/2012 6:42 AM, Arnd Bergmann wrote:
> > Ok, but please remove tile_net_devs then.
> > I think a better abstraction for tile_net_devs_for_channel would be
> > some interface that lets you add private data to a channel so when
> > you get data from a channel, you can extract that pointer from the driver
> > using the channel.
> I think what would be clearer is to document how and why we are using this
> additional data structure. We do access via both arrays where it is
> efficient to do so, so getting rid of either of them doesn't seem right.
> Let's keep the "normal" tile_net_devs[] as is, indexed by devno, and make
> the tile_net_devs_for_channel[] more abstracted by using the following code:
The tile_net_devs still feels dirty. You basically only
use it in tile_net_handle_egress_timer(), but there you don't
actually take the mutex that protects addition and removal from
the array, so it's racy in case of hotplug.
A more conservative way to do this is to have the timer per
device (or by channel, if you like), so it does not have to
iterate the array.
> /*
> * Provide support for efficiently mapping a channel to the device
> * that is using that channel, or NULL if none. The pointers in this
> * array are only non-NULL when pointing to active tilegx net_device
> * structures, and they are cleared before the struture itself is
> * released.
> *
> * HACK: We use "32", not TILE_NET_CHANNELS, because it is fairly
> * subtle that the 5 bit "idesc.channel" field never exceeds
> * TILE_NET_CHANNELS.
> */
> static struct net_device *channel_to_dev[32];
>
> static void bychannel_add(struct net_device *dev)
> {
> struct tile_net_priv *priv = netdev_priv(dev);
> BUG_ON(channel_to_dev[priv->channel] != NULL);
> channel_to_dev[priv->channel] = dev;
> }
>
> static void bychannel_delete(struct net_device *dev)
> {
> struct tile_net_priv *priv = netdev_priv(dev);
> channel_to_dev[priv->channel] = NULL;
> }
>
> static inline struct net_device *bychannel_lookup(int channel)
> {
> return channel_to_dev[channel];
> }
>
>
> We then call bychannel_add() in the open method, and bychannel_delete() in
> the close method, so it's clear that the pointers have appropriate lifetimes.
Ok.
Arnd
^ permalink raw reply
* Re: suspicious RCU usage warnings in 3.3.0
From: mroos @ 2012-04-13 11:55 UTC (permalink / raw)
To: Paul E. McKenney; +Cc: David Miller, shemminger, Linux Kernel list, netdev
In-Reply-To: <20120412045428.GB2497@linux.vnet.ibm.com>
> sparc64: Eliminate obsolete __handle_softirq() function
>
> The invocation of softirq is now handled by irq_exit(), so there is no
> need for sparc64 to invoke it on the trap-return path. In fact, doing so
> is a bug because if the trap occurred in the idle loop, this invocation
> can result in lockdep-RCU failures. The problem is that RCU ignores idle
> CPUs, and the sparc64 trap-return path to the softirq handlers fails to
> tell RCU that the CPU must be considered non-idle while those handlers
> are executing. This means that RCU is ignoring any RCU read-side critical
> sections in those handlers, which in turn means that RCU-protected data
> can be yanked out from under those read-side critical sections.
>
> The shiny new lockdep-RCU ability to detect RCU read-side critical sections
> that RCU is ignoring located this problem.
>
> The fix is straightforward: Make sparc64 stop manually invoking the
> softirq handlers.
>
> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
It works for me on Sun Fire V100 - no more RCU warnings under ping
flood.
Tested-by: Meelis Roos <mroos@linux.ee>
--
Meelis Roos (mroos@linux.ee)
^ permalink raw reply
* Netlink, route monitoring, RTM_DELROUTE not issued for ppp peer address. Bug or feature?
From: Denys Fedoryshchenko @ 2012-04-13 12:10 UTC (permalink / raw)
To: netdev
Hi
I am trying to implement my own routing daemon, and using
RTMGRP_IPV4_ROUTE to monitor routes.
I notice, that on PPPoE, when interface disappearing, RTM_DELROUTE will
be sent to listener only for local address, but nothing about peer
address attached to this interface.
CBSS_PPPoE ~ # ifconfig ppp0
ppp0 Link encap:Point-to-Point Protocol
inet addr:172.16.146.1 P-t-P:172.16.146.10
Mask:255.255.255.255
UP POINTOPOINT RUNNING NOARP MULTICAST MTU:1200 Metric:1
RX packets:40439 errors:0 dropped:0 overruns:0 frame:0
TX packets:63211 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:3
RX bytes:3584507 (3.4 MiB) TX bytes:47180929 (44.9 MiB)
CBSS_PPPoE ~ # pppkill tek234@visp.net.lb (this is killing pppd
process for ppp0 interface)
ip monitor route will show only RTM_DELROUTE for RTN_LOCAL, but nothing
about RTN_UNICAST for 172.16.146.10.
"Deleted local 172.16.146.1 dev ppp0 table local proto kernel scope
host src 172.16.146.1"
My program also receive same.
Is it a bug, that no RTM_DELROUTE issued for peer ip address? Or should
i monitor also RTMGRP_IPV4_IFADDR for link disappearance?
Thanks for help!
^ permalink raw reply
* Re: linux-next: Tree for Apr 12
From: Alan Cox @ 2012-04-13 12:11 UTC (permalink / raw)
To: Andrew Morton; +Cc: Stephen Rothwell, linux-next, LKML, netdev
In-Reply-To: <20120412142415.564710b8.akpm@linux-foundation.org>
On Thu, 12 Apr 2012 14:24:15 -0700
Andrew Morton <akpm@linux-foundation.org> wrote:
> On Thu, 12 Apr 2012 14:59:31 +1000
> Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>
> > I have created today's linux-next tree at
> > git://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git
>
> This isn't working for me. Some time between April 3 and April 12
> someone merged something into the non-mm part of linux-next which broke
> ssh.
You are getting further than me - on 32bit -next of the 12th explodes
with random userspace crashes in init or similar early code.
Alan
^ permalink raw reply
* [0/3] bridge: Do not send multicast queries by default
From: Herbert Xu @ 2012-04-13 12:36 UTC (permalink / raw)
To: David S. Miller, netdev
Hi:
This series of patches is aimed to change the default multicast
snooping behaviour to one that is safer to deploy in the wild.
There have been numerous reports of switches misbehaving with
our current behaviour of sending general queries, presumably
because we're using a zero source IP address which is unavoidable
as using anything else would interfere with multicast querier
elections (incidentally, I noticed that our IPv6 code has been
"fixed" to not use zero source addresses, which is wrong as we
may end up being THE MLD querier in a network).
Since our queries aren't actually required for multicast snooping
to function, but is merely an optimisation mostly for faster
start-up convergence, I think we should disable this by default.
Cheers,
--
Email: Herbert Xu <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
^ permalink raw reply
* [PATCH 2/3] bridge: Restart queries when last querier expires
From: Herbert Xu @ 2012-04-13 12:37 UTC (permalink / raw)
To: David S. Miller, netdev
In-Reply-To: <20120413123641.GA758@gondor.apana.org.au>
bridge: Restart queries when last querier expires
As it stands when we discover that a real querier (one that queries
with a non-zero source address) we stop querying. However, even
after said querier has fallen off the edge of the earth, we will
never restart querying (unless the bridge itself is restarted).
This patch fixes this by kicking our own querier into gear when
the timer for other queriers expire.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
net/bridge/br_multicast.c | 19 ++++++++++++++++++-
1 file changed, 18 insertions(+), 1 deletion(-)
diff --git a/net/bridge/br_multicast.c b/net/bridge/br_multicast.c
index dca0c66..b0796d7 100644
--- a/net/bridge/br_multicast.c
+++ b/net/bridge/br_multicast.c
@@ -36,6 +36,8 @@
#define mlock_dereference(X, br) \
rcu_dereference_protected(X, lockdep_is_held(&br->multicast_lock))
+static void br_multicast_start_querier(struct net_bridge *br);
+
#if IS_ENABLED(CONFIG_IPV6)
static inline int ipv6_is_transient_multicast(const struct in6_addr *addr)
{
@@ -740,6 +742,21 @@ static void br_multicast_local_router_expired(unsigned long data)
{
}
+static void br_multicast_querier_expired(unsigned long data)
+{
+ struct net_bridge_port *port = (void *)data;
+ struct net_bridge *br = port->br;
+
+ spin_lock(&br->multicast_lock);
+ if (!netif_running(br->dev) || br->multicast_disabled)
+ goto out;
+
+ br_multicast_start_querier(br);
+
+out:
+ spin_unlock(&br->multicast_lock);
+}
+
static void __br_multicast_send_query(struct net_bridge *br,
struct net_bridge_port *port,
struct br_ip *ip)
@@ -1562,7 +1579,7 @@ void br_multicast_init(struct net_bridge *br)
setup_timer(&br->multicast_router_timer,
br_multicast_local_router_expired, 0);
setup_timer(&br->multicast_querier_timer,
- br_multicast_local_router_expired, 0);
+ br_multicast_querier_expired, 0);
setup_timer(&br->multicast_query_timer, br_multicast_query_expired,
(unsigned long)br);
}
^ permalink raw reply related
* [PATCH 1/3] bridge: Add br_multicast_start_querier
From: Herbert Xu @ 2012-04-13 12:37 UTC (permalink / raw)
To: David S. Miller, netdev
In-Reply-To: <20120413123641.GA758@gondor.apana.org.au>
bridge: Add br_multicast_start_querier
This patch adds the helper br_multicast_start_querier so that
the code which starts the queriers in br_multicast_toggle can
be reused elsewhere.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
net/bridge/br_multicast.c | 25 ++++++++++++++++---------
1 file changed, 16 insertions(+), 9 deletions(-)
diff --git a/net/bridge/br_multicast.c b/net/bridge/br_multicast.c
index 27ca25e..dca0c66 100644
--- a/net/bridge/br_multicast.c
+++ b/net/bridge/br_multicast.c
@@ -1689,9 +1689,23 @@ unlock:
return err;
}
-int br_multicast_toggle(struct net_bridge *br, unsigned long val)
+static void br_multicast_start_querier(struct net_bridge *br)
{
struct net_bridge_port *port;
+
+ br_multicast_open(br);
+
+ list_for_each_entry(port, &br->port_list, list) {
+ if (port->state == BR_STATE_DISABLED ||
+ port->state == BR_STATE_BLOCKING)
+ continue;
+
+ __br_multicast_enable_port(port);
+ }
+}
+
+int br_multicast_toggle(struct net_bridge *br, unsigned long val)
+{
int err = 0;
struct net_bridge_mdb_htable *mdb;
@@ -1721,14 +1735,7 @@ rollback:
goto rollback;
}
- br_multicast_open(br);
- list_for_each_entry(port, &br->port_list, list) {
- if (port->state == BR_STATE_DISABLED ||
- port->state == BR_STATE_BLOCKING)
- continue;
-
- __br_multicast_enable_port(port);
- }
+ br_multicast_start_querier(br);
unlock:
spin_unlock_bh(&br->multicast_lock);
^ permalink raw reply related
* [PATCH 3/3] bridge: Add multicast_querier toggle and disable queries by default
From: Herbert Xu @ 2012-04-13 12:37 UTC (permalink / raw)
To: David S. Miller, netdev
In-Reply-To: <20120413123641.GA758@gondor.apana.org.au>
bridge: Add multicast_querier toggle and disable queries by default
Sending general queries was implemented as an optimisation to speed
up convergence on start-up. In order to prevent interference with
multicast routers a zero source address has to be used.
Unfortunately these packets appear to cause some multicast-aware
switches to misbehave, e.g., by disrupting multicast packets to us.
Since the multicast snooping feature still functions without sending
our own queries, this patch will change the default to not send
queries.
For those that need queries in order to speed up convergence on start-up,
a toggle is provided to restore the previous behaviour.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
---
net/bridge/br_multicast.c | 20 ++++++++++++++++++++
net/bridge/br_private.h | 2 ++
net/bridge/br_sysfs_br.c | 18 ++++++++++++++++++
3 files changed, 40 insertions(+)
diff --git a/net/bridge/br_multicast.c b/net/bridge/br_multicast.c
index b0796d7..26df642 100644
--- a/net/bridge/br_multicast.c
+++ b/net/bridge/br_multicast.c
@@ -783,6 +783,7 @@ static void br_multicast_send_query(struct net_bridge *br,
struct br_ip br_group;
if (!netif_running(br->dev) || br->multicast_disabled ||
+ !br->multicast_querier ||
timer_pending(&br->multicast_querier_timer))
return;
@@ -1565,6 +1566,7 @@ void br_multicast_init(struct net_bridge *br)
br->hash_max = 512;
br->multicast_router = 1;
+ br->multicast_querier = 0;
br->multicast_last_member_count = 2;
br->multicast_startup_query_count = 2;
@@ -1760,6 +1762,24 @@ unlock:
return err;
}
+int br_multicast_set_querier(struct net_bridge *br, unsigned long val)
+{
+ val = !!val;
+
+ spin_lock_bh(&br->multicast_lock);
+ if (br->multicast_querier == val)
+ goto unlock;
+
+ br->multicast_querier = val;
+ if (val)
+ br_multicast_start_querier(br);
+
+unlock:
+ spin_unlock_bh(&br->multicast_lock);
+
+ return 0;
+}
+
int br_multicast_set_hash_max(struct net_bridge *br, unsigned long val)
{
int err = -ENOENT;
diff --git a/net/bridge/br_private.h b/net/bridge/br_private.h
index e1d8822..f8ffd8c 100644
--- a/net/bridge/br_private.h
+++ b/net/bridge/br_private.h
@@ -224,6 +224,7 @@ struct net_bridge
unsigned char multicast_router;
u8 multicast_disabled:1;
+ u8 multicast_querier:1;
u32 hash_elasticity;
u32 hash_max;
@@ -417,6 +418,7 @@ extern int br_multicast_set_router(struct net_bridge *br, unsigned long val);
extern int br_multicast_set_port_router(struct net_bridge_port *p,
unsigned long val);
extern int br_multicast_toggle(struct net_bridge *br, unsigned long val);
+extern int br_multicast_set_querier(struct net_bridge *br, unsigned long val);
extern int br_multicast_set_hash_max(struct net_bridge *br, unsigned long val);
static inline bool br_multicast_is_router(struct net_bridge *br)
diff --git a/net/bridge/br_sysfs_br.c b/net/bridge/br_sysfs_br.c
index c236c0e..a27243e 100644
--- a/net/bridge/br_sysfs_br.c
+++ b/net/bridge/br_sysfs_br.c
@@ -379,6 +379,23 @@ static ssize_t store_multicast_snooping(struct device *d,
static DEVICE_ATTR(multicast_snooping, S_IRUGO | S_IWUSR,
show_multicast_snooping, store_multicast_snooping);
+static ssize_t show_multicast_querier(struct device *d,
+ struct device_attribute *attr,
+ char *buf)
+{
+ struct net_bridge *br = to_bridge(d);
+ return sprintf(buf, "%d\n", br->multicast_querier);
+}
+
+static ssize_t store_multicast_querier(struct device *d,
+ struct device_attribute *attr,
+ const char *buf, size_t len)
+{
+ return store_bridge_parm(d, buf, len, br_multicast_set_querier);
+}
+static DEVICE_ATTR(multicast_querier, S_IRUGO | S_IWUSR,
+ show_multicast_querier, store_multicast_querier);
+
static ssize_t show_hash_elasticity(struct device *d,
struct device_attribute *attr, char *buf)
{
@@ -702,6 +719,7 @@ static struct attribute *bridge_attrs[] = {
#ifdef CONFIG_BRIDGE_IGMP_SNOOPING
&dev_attr_multicast_router.attr,
&dev_attr_multicast_snooping.attr,
+ &dev_attr_multicast_querier.attr,
&dev_attr_hash_elasticity.attr,
&dev_attr_hash_max.attr,
&dev_attr_multicast_last_member_count.attr,
^ permalink raw reply related
* Re: suspicious RCU usage warnings in 3.3.0
From: Paul E. McKenney @ 2012-04-13 13:35 UTC (permalink / raw)
To: mroos; +Cc: David Miller, shemminger, Linux Kernel list, netdev
In-Reply-To: <alpine.SOC.1.00.1204131444150.2736@math.ut.ee>
On Fri, Apr 13, 2012 at 02:55:12PM +0300, mroos@linux.ee wrote:
> > sparc64: Eliminate obsolete __handle_softirq() function
> >
> > The invocation of softirq is now handled by irq_exit(), so there is no
> > need for sparc64 to invoke it on the trap-return path. In fact, doing so
> > is a bug because if the trap occurred in the idle loop, this invocation
> > can result in lockdep-RCU failures. The problem is that RCU ignores idle
> > CPUs, and the sparc64 trap-return path to the softirq handlers fails to
> > tell RCU that the CPU must be considered non-idle while those handlers
> > are executing. This means that RCU is ignoring any RCU read-side critical
> > sections in those handlers, which in turn means that RCU-protected data
> > can be yanked out from under those read-side critical sections.
> >
> > The shiny new lockdep-RCU ability to detect RCU read-side critical sections
> > that RCU is ignoring located this problem.
> >
> > The fix is straightforward: Make sparc64 stop manually invoking the
> > softirq handlers.
> >
> > Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
>
> It works for me on Sun Fire V100 - no more RCU warnings under ping
> flood.
>
> Tested-by: Meelis Roos <mroos@linux.ee>
OK, if this thing is going to actually work, I guess I need to update
the changelog to give credit where it is due, please see below.
My main concern about my patch is my removal of this line:
bne,pn %icc, __handle_softirq
It is quite possible that this should instead change to look as follows:
bne,pn %icc, __handle_preemption
This code is under #ifndef CONFIG_SMP, so Meelis's testing would not
reach it.
Anyway, patch with updated changelog below.
Thanx, Paul
------------------------------------------------------------------------
sparc64: Eliminate obsolete __handle_softirq() function
The invocation of softirq is now handled by irq_exit(), so there is no
need for sparc64 to invoke it on the trap-return path. In fact, doing so
is a bug because if the trap occurred in the idle loop, this invocation
can result in lockdep-RCU failures. The problem is that RCU ignores idle
CPUs, and the sparc64 trap-return path to the softirq handlers fails to
tell RCU that the CPU must be considered non-idle while those handlers
are executing. This means that RCU is ignoring any RCU read-side critical
sections in those handlers, which in turn means that RCU-protected data
can be yanked out from under those read-side critical sections.
The shiny new lockdep-RCU ability to detect RCU read-side critical sections
that RCU is ignoring located this problem.
The fix is straightforward: Make sparc64 stop manually invoking the
softirq handlers.
Reported-by: Meelis Roos <mroos@linux.ee>
Suggested-by: David Miller <davem@davemloft.net>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Tested-by: Meelis Roos <mroos@linux.ee>
diff --git a/arch/sparc/kernel/rtrap_64.S b/arch/sparc/kernel/rtrap_64.S
index 77f1b95..9171fc2 100644
--- a/arch/sparc/kernel/rtrap_64.S
+++ b/arch/sparc/kernel/rtrap_64.S
@@ -20,11 +20,6 @@
.text
.align 32
-__handle_softirq:
- call do_softirq
- nop
- ba,a,pt %xcc, __handle_softirq_continue
- nop
__handle_preemption:
call schedule
wrpr %g0, RTRAP_PSTATE, %pstate
@@ -89,9 +84,7 @@ rtrap:
cmp %l1, 0
/* mm/ultra.S:xcall_report_regs KNOWS about this load. */
- bne,pn %icc, __handle_softirq
ldx [%sp + PTREGS_OFF + PT_V9_TSTATE], %l1
-__handle_softirq_continue:
rtrap_xcall:
sethi %hi(0xf << 20), %l4
and %l1, %l4, %l4
^ permalink raw reply related
* SV: Netlink, route monitoring, RTM_DELROUTE not issued for ppp peer address. Bug or feature?
From: Stian Skjelstad @ 2012-04-13 13:59 UTC (permalink / raw)
To: 'Denys Fedoryshchenko', netdev
In-Reply-To: <8986ed37b55c756ecdeec0315f73e63a@visp.net.lb>
I've seen the same issue, and got no response so far to that very same question. Have you tested this on older kernels? I have worked around this issue by testing my local copy of routing entries that use 'via'. If they are not routable when after the RTM_DELROUTE event, I autoremove them from my local copy.
Stian Skjelstad
-----Opprinnelig melding-----
Fra: netdev-owner@vger.kernel.org [mailto:netdev-owner@vger.kernel.org] På vegne av Denys Fedoryshchenko
Sendt: 13. april 2012 14:11
Til: netdev@vger.kernel.org
Emne: Netlink, route monitoring, RTM_DELROUTE not issued for ppp peer address. Bug or feature?
Hi
I am trying to implement my own routing daemon, and using RTMGRP_IPV4_ROUTE to monitor routes.
I notice, that on PPPoE, when interface disappearing, RTM_DELROUTE will be sent to listener only for local address, but nothing about peer address attached to this interface.
CBSS_PPPoE ~ # ifconfig ppp0
ppp0 Link encap:Point-to-Point Protocol
inet addr:172.16.146.1 P-t-P:172.16.146.10
Mask:255.255.255.255
UP POINTOPOINT RUNNING NOARP MULTICAST MTU:1200 Metric:1
RX packets:40439 errors:0 dropped:0 overruns:0 frame:0
TX packets:63211 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:3
RX bytes:3584507 (3.4 MiB) TX bytes:47180929 (44.9 MiB)
CBSS_PPPoE ~ # pppkill tek234@visp.net.lb (this is killing pppd process for ppp0 interface)
ip monitor route will show only RTM_DELROUTE for RTN_LOCAL, but nothing about RTN_UNICAST for 172.16.146.10.
"Deleted local 172.16.146.1 dev ppp0 table local proto kernel scope host src 172.16.146.1"
My program also receive same.
Is it a bug, that no RTM_DELROUTE issued for peer ip address? Or should i monitor also RTMGRP_IPV4_IFADDR for link disappearance?
Thanks for help!
--
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
* [PATCH] [trivial] net: Fix spelling typo in net
From: Masanari Iida @ 2012-04-13 14:33 UTC (permalink / raw)
To: netdev, trivial, davem, linux-wireless; +Cc: linux-kernel, Masanari Iida
Correct spelling typo within drivers/net.
Signed-off-by: Masanari Iida <standby24x7@gmail.com>
---
drivers/net/ethernet/mellanox/mlx4/en_ethtool.c | 2 +-
drivers/net/ethernet/qlogic/qlge/qlge_main.c | 2 +-
drivers/net/ethernet/tehuti/tehuti.c | 2 +-
drivers/net/ethernet/ti/tlan.c | 2 +-
drivers/net/irda/sh_irda.c | 2 +-
drivers/net/irda/sh_sir.c | 2 +-
drivers/net/wan/lmc/lmc_main.c | 2 +-
drivers/net/wimax/i2400m/usb.c | 2 +-
drivers/net/wireless/ipw2x00/ipw2100.c | 4 ++--
drivers/net/wireless/ipw2x00/ipw2200.c | 2 +-
drivers/net/wireless/rtlwifi/rtl8192de/hw.c | 2 +-
11 files changed, 12 insertions(+), 12 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c
index 70346fd..e15a1fb 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c
@@ -83,7 +83,7 @@ static const char main_strings[][ETH_GSTRING_LEN] = {
#define NUM_ALL_STATS (NUM_MAIN_STATS + NUM_PORT_STATS + NUM_PKT_STATS + NUM_PERF_STATS)
static const char mlx4_en_test_names[][ETH_GSTRING_LEN]= {
- "Interupt Test",
+ "Interrupt Test",
"Link Test",
"Speed Test",
"Register Test",
diff --git a/drivers/net/ethernet/qlogic/qlge/qlge_main.c b/drivers/net/ethernet/qlogic/qlge/qlge_main.c
index 49343ec..09d8d33 100644
--- a/drivers/net/ethernet/qlogic/qlge/qlge_main.c
+++ b/drivers/net/ethernet/qlogic/qlge/qlge_main.c
@@ -3845,7 +3845,7 @@ static int ql_wol(struct ql_adapter *qdev)
if (qdev->wol & (WAKE_ARP | WAKE_MAGICSECURE | WAKE_PHY | WAKE_UCAST |
WAKE_MCAST | WAKE_BCAST)) {
netif_err(qdev, ifdown, qdev->ndev,
- "Unsupported WOL paramter. qdev->wol = 0x%x.\n",
+ "Unsupported WOL parameter. qdev->wol = 0x%x.\n",
qdev->wol);
return -EINVAL;
}
diff --git a/drivers/net/ethernet/tehuti/tehuti.c b/drivers/net/ethernet/tehuti/tehuti.c
index ad973ff..4179966 100644
--- a/drivers/net/ethernet/tehuti/tehuti.c
+++ b/drivers/net/ethernet/tehuti/tehuti.c
@@ -1317,7 +1317,7 @@ static void print_rxdd(struct rxd_desc *rxdd, u32 rxd_val1, u16 len,
static void print_rxfd(struct rxf_desc *rxfd)
{
- DBG("=== RxF desc CHIP ORDER/ENDIANESS =============\n"
+ DBG("=== RxF desc CHIP ORDER/ENDIANNESS =============\n"
"info 0x%x va_lo %u pa_lo 0x%x pa_hi 0x%x len 0x%x\n",
rxfd->info, rxfd->va_lo, rxfd->pa_lo, rxfd->pa_hi, rxfd->len);
}
diff --git a/drivers/net/ethernet/ti/tlan.c b/drivers/net/ethernet/ti/tlan.c
index 817ad3b..bb8b802 100644
--- a/drivers/net/ethernet/ti/tlan.c
+++ b/drivers/net/ethernet/ti/tlan.c
@@ -2545,7 +2545,7 @@ static void tlan_phy_reset(struct net_device *dev)
phy = priv->phy[priv->phy_num];
- TLAN_DBG(TLAN_DEBUG_GNRL, "%s: Reseting PHY.\n", dev->name);
+ TLAN_DBG(TLAN_DEBUG_GNRL, "%s: Resetting PHY.\n", dev->name);
tlan_mii_sync(dev->base_addr);
value = MII_GC_LOOPBK | MII_GC_RESET;
tlan_mii_write_reg(dev, phy, MII_GEN_CTL, value);
diff --git a/drivers/net/irda/sh_irda.c b/drivers/net/irda/sh_irda.c
index 725d6b3..eb315b8 100644
--- a/drivers/net/irda/sh_irda.c
+++ b/drivers/net/irda/sh_irda.c
@@ -737,7 +737,7 @@ static int sh_irda_stop(struct net_device *ndev)
netif_stop_queue(ndev);
pm_runtime_put_sync(&self->pdev->dev);
- dev_info(&ndev->dev, "stoped\n");
+ dev_info(&ndev->dev, "stopped\n");
return 0;
}
diff --git a/drivers/net/irda/sh_sir.c b/drivers/net/irda/sh_sir.c
index e6661b5..256eddf 100644
--- a/drivers/net/irda/sh_sir.c
+++ b/drivers/net/irda/sh_sir.c
@@ -685,7 +685,7 @@ static int sh_sir_stop(struct net_device *ndev)
netif_stop_queue(ndev);
- dev_info(&ndev->dev, "stoped\n");
+ dev_info(&ndev->dev, "stopped\n");
return 0;
}
diff --git a/drivers/net/wan/lmc/lmc_main.c b/drivers/net/wan/lmc/lmc_main.c
index 76a8a4a..e38626d 100644
--- a/drivers/net/wan/lmc/lmc_main.c
+++ b/drivers/net/wan/lmc/lmc_main.c
@@ -1120,7 +1120,7 @@ static void lmc_running_reset (struct net_device *dev) /*fold00*/
{
lmc_softc_t *sc = dev_to_sc(dev);
- lmc_trace(dev, "lmc_runnig_reset in");
+ lmc_trace(dev, "lmc_running_reset in");
/* stop interrupts */
/* Clear the interrupt mask */
diff --git a/drivers/net/wimax/i2400m/usb.c b/drivers/net/wimax/i2400m/usb.c
index 29b1e03..713d033 100644
--- a/drivers/net/wimax/i2400m/usb.c
+++ b/drivers/net/wimax/i2400m/usb.c
@@ -695,7 +695,7 @@ int i2400mu_resume(struct usb_interface *iface)
d_fnstart(3, dev, "(iface %p)\n", iface);
rmb(); /* see i2400m->updown's documentation */
if (i2400m->updown == 0) {
- d_printf(1, dev, "fw was down, no resume neeed\n");
+ d_printf(1, dev, "fw was down, no resume needed\n");
goto out;
}
d_printf(1, dev, "fw was up, resuming\n");
diff --git a/drivers/net/wireless/ipw2x00/ipw2100.c b/drivers/net/wireless/ipw2x00/ipw2100.c
index f0551f8..ab075ef 100644
--- a/drivers/net/wireless/ipw2x00/ipw2100.c
+++ b/drivers/net/wireless/ipw2x00/ipw2100.c
@@ -3773,7 +3773,7 @@ IPW2100_ORD(STAT_TX_HOST_REQUESTS, "requested Host Tx's (MSDU)"),
IPW2100_ORD(COUNTRY_CODE,
"IEEE country code as recv'd from beacon"),
IPW2100_ORD(COUNTRY_CHANNELS,
- "channels suported by country"),
+ "channels supported by country"),
IPW2100_ORD(RESET_CNT, "adapter resets (warm)"),
IPW2100_ORD(BEACON_INTERVAL, "Beacon interval"),
IPW2100_ORD(ANTENNA_DIVERSITY,
@@ -4062,7 +4062,7 @@ static int ipw2100_switch_mode(struct ipw2100_priv *priv, u32 mode)
ipw2100_firmware.version = 0;
#endif
- printk(KERN_INFO "%s: Reseting on mode change.\n", priv->net_dev->name);
+ printk(KERN_INFO "%s: Resetting on mode change.\n", priv->net_dev->name);
priv->reset_backoff = 0;
schedule_reset(priv);
diff --git a/drivers/net/wireless/ipw2x00/ipw2200.c b/drivers/net/wireless/ipw2x00/ipw2200.c
index 2b02257..bfbe662 100644
--- a/drivers/net/wireless/ipw2x00/ipw2200.c
+++ b/drivers/net/wireless/ipw2x00/ipw2200.c
@@ -7024,7 +7024,7 @@ static int ipw_qos_activate(struct ipw_priv *priv,
cpu_to_le16(burst_duration);
} else if (priv->ieee->iw_mode == IW_MODE_ADHOC) {
if (type == IEEE_B) {
- IPW_DEBUG_QOS("QoS activate IBSS nework mode %d\n",
+ IPW_DEBUG_QOS("QoS activate IBSS network mode %d\n",
type);
if (priv->qos_data.qos_enable == 0)
active_one = &def_parameters_CCK;
diff --git a/drivers/net/wireless/rtlwifi/rtl8192de/hw.c b/drivers/net/wireless/rtlwifi/rtl8192de/hw.c
index 509f5af..7ac9b8c 100644
--- a/drivers/net/wireless/rtlwifi/rtl8192de/hw.c
+++ b/drivers/net/wireless/rtlwifi/rtl8192de/hw.c
@@ -1745,7 +1745,7 @@ static void _rtl92de_efuse_update_chip_version(struct ieee80211_hw *hw)
break;
default:
chipver |= CHIP_92D_D_CUT;
- RT_TRACE(rtlpriv, COMP_INIT, DBG_EMERG, "Unkown CUT!\n");
+ RT_TRACE(rtlpriv, COMP_INIT, DBG_EMERG, "Unknown CUT!\n");
break;
}
rtlpriv->rtlhal.version = chipver;
--
1.7.10.128.g7945c
^ permalink raw reply related
* [PATCH v2] net/ethernet: ks8851_mll fix rx frame buffer overflow
From: Davide Ciminaghi @ 2012-04-13 14:48 UTC (permalink / raw)
To: David S. Miller, Alexey Dobriyan, Thomas Meyer, Wan ZongShun,
Lucas De Marchi, netdev
Cc: raffaele.recalcati
At the beginning of ks_rcv(), a for loop retrieves the
header information relevant to all the frames stored
in the mac's internal buffers. The number of pending
frames is stored as an 8 bits field in KS_RXFCTR.
If interrupts are disabled long enough to allow for more than
32 frames to accumulate in the MAC's internal buffers, a buffer
overflow occurs.
This patch fixes the problem by making the
driver's frame_head_info buffer big enough.
Well actually, since the chip appears to have 12K of
internal rx buffers and the shortest ethernet frame should
be 64 bytes long, maybe the limit could be set to
12*1024/64 = 192 frames, but 255 should be safer.
Signed-off-by: Davide Ciminaghi <ciminaghi@gnudd.com>
Signed-off-by: Raffaele Recalcati <raffaele.recalcati@bticino.it>
---
The buffer overflow triggers weird random oopses.
The following setup helped reproducing the bug rather frequently
on our omap3 platform with 2.6.37 kernel:
- nfs rootfs
- host $ sudo ping -f $TARGET_IP_ADDRESS
- host $ scp -r BIG_DIR user@$TARGET_IP_ADDRESS:
drivers/net/ethernet/micrel/ks8851_mll.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ethernet/micrel/ks8851_mll.c b/drivers/net/ethernet/micrel/ks8851_mll.c
index 2784bc7..927dbd2 100644
--- a/drivers/net/ethernet/micrel/ks8851_mll.c
+++ b/drivers/net/ethernet/micrel/ks8851_mll.c
@@ -40,7 +40,7 @@
#define DRV_NAME "ks8851_mll"
static u8 KS_DEFAULT_MAC_ADDRESS[] = { 0x00, 0x10, 0xA1, 0x86, 0x95, 0x11 };
-#define MAX_RECV_FRAMES 32
+#define MAX_RECV_FRAMES 255
#define MAX_BUF_SIZE 2048
#define TX_BUF_SIZE 2000
#define RX_BUF_SIZE 2000
--
1.7.0.4
^ permalink raw reply related
* Re: [0/3] bridge: Do not send multicast queries by default
From: David Miller @ 2012-04-13 14:53 UTC (permalink / raw)
To: herbert; +Cc: netdev
In-Reply-To: <20120413123641.GA758@gondor.apana.org.au>
From: Herbert Xu <herbert@gondor.hengli.com.au>
Date: Fri, 13 Apr 2012 20:36:41 +0800
> (incidentally, I noticed that our IPv6 code has been "fixed" to not
> use zero source addresses, which is wrong as we may end up being THE
> MLD querier in a network).
I seem to recall it was explicitly changed to be this way and that
there was a good reason for this, see the history.
^ permalink raw reply
* Re: suspicious RCU usage warnings in 3.3.0
From: David Miller @ 2012-04-13 14:55 UTC (permalink / raw)
To: paulmck; +Cc: mroos, shemminger, linux-kernel, netdev
In-Reply-To: <20120413133513.GB2457@linux.vnet.ibm.com>
From: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Date: Fri, 13 Apr 2012 06:35:13 -0700
DANGER WILL ROBINSON, McKenney is hacking sparc64 assembler!
:-))))
> My main concern about my patch is my removal of this line:
>
> bne,pn %icc, __handle_softirq
>
> It is quite possible that this should instead change to look as follows:
>
> bne,pn %icc, __handle_preemption
>
I would say it should not, because the __handle_softirq code path,
if taken, unconditionally branches to the very next instruction
after this branch and it's delay slot.
> Anyway, patch with updated changelog below.
Great work, I'll apply this after I test it out a bit myself.
Thanks a lot Paul!
^ permalink raw reply
* Re: [PATCH 1/2] net: In unregister_netdevice_notifier unregister the netdevices.
From: David Miller @ 2012-04-13 15:05 UTC (permalink / raw)
To: ebiederm
Cc: eric.dumazet, ericvh, davej, linux-kernel, netdev, levinsasha928,
remi.denis-courmont
In-Reply-To: <m1aa2otjf4.fsf_-_@fess.ebiederm.org>
From: ebiederm@xmission.com (Eric W. Biederman)
Date: Fri, 06 Apr 2012 18:33:35 -0700
>>From a51fda849355b2ffb27cdb5d66fa1a96e7f47335 Mon Sep 17 00:00:00 2001
> From: Eric W. Biederman <ebiederm@xmission.com>
> Date: Sun, 25 Dec 2011 19:39:52 -0800
> Subject:
Yikes, I had to edit this turd out :-) Otherwise with this empty
Subject, GIT uses the first line of your commit log body as the
first line.
> We already synthesize events in register_netdevice_notifier and synthesizing
> events in unregister_netdevice_notifier allows to us remove the need for
> special case cleanup code.
>
> This change should be safe as it adds no new cases for existing callers
> of unregiser_netdevice_notifier to handle.
>
> Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Applied.
^ permalink raw reply
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