* [PATCH] atm: clip causes unregister hang
From: Stephen Hemminger @ 2006-04-12 19:45 UTC (permalink / raw)
To: Herbert Xu; +Cc: davem, chas, linux-atm-general, netdev, stable
In-Reply-To: <E1FTjlM-00050V-00@gondolin.me.apana.org.au>
If Classical IP over ATM module is loaded, its neighbor table gets
populated when permanent neighbor entries are created; but these entries
are not flushed when the device is removed. Since the entry never gets
flushed the unregister of the network device never completes.
Bug-reference: http://bugzilla.kernel.org/show_bug.cgi?id=6295
Signed-off-by: Stephen Hemminger <shemminger@osdl.org>
--- linux-2.6.16.2.orig/net/atm/clip.c 2006-04-12 10:10:43.000000000 -0700
+++ linux-2.6.16.2/net/atm/clip.c 2006-04-12 11:22:47.000000000 -0700
@@ -613,12 +613,19 @@
static int clip_device_event(struct notifier_block *this,unsigned long event,
- void *dev)
+ void *arg)
{
+ struct net_device *dev = arg;
+
+ if (event == NETDEV_UNREGISTER) {
+ neigh_ifdown(&clip_tbl, dev);
+ return NOTIFY_DONE;
+ }
+
/* ignore non-CLIP devices */
- if (((struct net_device *) dev)->type != ARPHRD_ATM ||
- ((struct net_device *) dev)->hard_start_xmit != clip_start_xmit)
+ if (dev->type != ARPHRD_ATM || dev->hard_start_xmit != clip_start_xmit)
return NOTIFY_DONE;
+
switch (event) {
case NETDEV_UP:
DPRINTK("clip_device_event NETDEV_UP\n");
@@ -688,8 +695,7 @@
DPRINTK("atmarpd_close\n");
atmarpd = NULL; /* assumed to be atomic */
barrier();
- unregister_inetaddr_notifier(&clip_inet_notifier);
- unregister_netdevice_notifier(&clip_dev_notifier);
+
if (skb_peek(&sk_atm(vcc)->sk_receive_queue))
printk(KERN_ERR "atmarpd_close: closing with requests "
"pending\n");
@@ -731,10 +737,6 @@
vcc->push = NULL;
vcc->pop = NULL; /* crash */
vcc->push_oam = NULL; /* crash */
- if (register_netdevice_notifier(&clip_dev_notifier))
- printk(KERN_ERR "register_netdevice_notifier failed\n");
- if (register_inetaddr_notifier(&clip_inet_notifier))
- printk(KERN_ERR "register_inetaddr_notifier failed\n");
return 0;
}
@@ -992,6 +994,8 @@
clip_tbl_hook = &clip_tbl;
register_atm_ioctl(&clip_ioctl_ops);
+ register_netdevice_notifier(&clip_dev_notifier);
+ register_inetaddr_notifier(&clip_inet_notifier);
#ifdef CONFIG_PROC_FS
{
@@ -1012,6 +1016,9 @@
remove_proc_entry("arp", atm_proc_root);
+ unregister_inetaddr_notifier(&clip_inet_notifier);
+ unregister_netdevice_notifier(&clip_dev_notifier);
+
deregister_atm_ioctl(&clip_ioctl_ops);
/* First, stop the idle timer, so it stops banging
^ permalink raw reply
* Re: [PATCH] atm: clip causes unregister hang
From: Herbert Xu @ 2006-04-12 20:00 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: davem, chas, linux-atm-general, netdev, stable
In-Reply-To: <20060412124533.14e0c4ff@localhost.localdomain>
Hi Stephen:
On Wed, Apr 12, 2006 at 12:45:33PM -0700, Stephen Hemminger wrote:
>
> /* ignore non-CLIP devices */
> - if (((struct net_device *) dev)->type != ARPHRD_ATM ||
> - ((struct net_device *) dev)->hard_start_xmit != clip_start_xmit)
> + if (dev->type != ARPHRD_ATM || dev->hard_start_xmit != clip_start_xmit)
> return NOTIFY_DONE;
I think we need to check whether atm_init_atmarp has been done before
passing this point. We also need to make sure that it doesn't get pulled
out from under us while we're doing this.
Cheers,
--
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <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
* Re: [RFD][PATCH] typhoon and core sample for folding away VLAN stuff
From: Ben Greear @ 2006-04-12 20:10 UTC (permalink / raw)
To: Ingo Oeser
Cc: Ingo Oeser, Denis Vlasenko, Dave Dillow, netdev, David S. Miller,
linux-kernel, jgarzik
In-Reply-To: <200604122132.46113.ioe-lkml@rameria.de>
What is the reasoning for this change? Is the compiler
able to optomize the right-hand-side to a constant with your
change in place?
> - if (veth->h_vlan_proto != __constant_htons(ETH_P_8021Q)) {
> + if (veth->h_vlan_proto != htons(ETH_P_8021Q)) {
> return -EINVAL;
> }
--
Ben Greear <greearb@candelatech.com>
Candela Technologies Inc http://www.candelatech.com
^ permalink raw reply
* Re: [PATCH] atm: clip causes unregister hang
From: Stephen Hemminger @ 2006-04-12 20:15 UTC (permalink / raw)
To: Herbert Xu; +Cc: davem, chas, linux-atm-general, netdev, stable
In-Reply-To: <20060412200015.GA19878@gondor.apana.org.au>
On Thu, 13 Apr 2006 06:00:15 +1000
Herbert Xu <herbert@gondor.apana.org.au> wrote:
> Hi Stephen:
>
> On Wed, Apr 12, 2006 at 12:45:33PM -0700, Stephen Hemminger wrote:
> >
> > /* ignore non-CLIP devices */
> > - if (((struct net_device *) dev)->type != ARPHRD_ATM ||
> > - ((struct net_device *) dev)->hard_start_xmit != clip_start_xmit)
> > + if (dev->type != ARPHRD_ATM || dev->hard_start_xmit != clip_start_xmit)
> > return NOTIFY_DONE;
>
> I think we need to check whether atm_init_atmarp has been done before
> passing this point. We also need to make sure that it doesn't get pulled
> out from under us while we're doing this.
>
> Cheers,
If atm_init_atmarp has not been done, then to_atamarpd is a nop because
atmarpd == NULL.
^ permalink raw reply
* [2.6 patch] net/ipv4/: possible cleanups
From: Adrian Bunk @ 2006-04-12 20:22 UTC (permalink / raw)
To: David S. Miller; +Cc: netdev
In-Reply-To: <20060406.133857.40478787.davem@davemloft.net>
On Thu, Apr 06, 2006 at 01:38:57PM -0700, David S. Miller wrote:
> From: Adrian Bunk <bunk@stusta.de>
> Date: Thu, 6 Apr 2006 10:49:09 +0200
>
> > On Wed, Apr 05, 2006 at 10:25:43PM -0700, David S. Miller wrote:
> > > > - remove the following unused EXPORT_SYMBO's:
> > > > - devinet.c: devinet_ioctl
> > >
> > > Used by wan drivers, can't remove.
> >
> > Used only by drivers that:
> > - are marked as BROKEN since at least 2.6.0
> > - are removed in a maintainer-approved patch I sent before this one
>
> Then get the WAN drivers removed, once they are you can remove the
> export, but until then you can't.
Done.
cu
Adrian
<-- snip -->
This patch contains the following possible cleanups:
- make the following needlessly global function static:
- arp.c: arp_rcv()
- remove the following unused EXPORT_SYMBOL's:
- devinet.c: devinet_ioctl
- fib_frontend.c: ip_rt_ioctl
- inet_hashtables.c: inet_bind_bucket_create
- inet_hashtables.c: inet_bind_hash
- tcp_input.c: sysctl_tcp_abc
- tcp_ipv4.c: sysctl_tcp_tw_reuse
- tcp_output.c: sysctl_tcp_mtu_probing
- tcp_output.c: sysctl_tcp_base_mss
Signed-off-by: Adrian Bunk <bunk@stusta.de>
---
include/net/arp.h | 2 --
net/ipv4/arp.c | 4 ++--
net/ipv4/devinet.c | 1 -
net/ipv4/fib_frontend.c | 1 -
net/ipv4/inet_hashtables.c | 4 ----
net/ipv4/tcp_input.c | 1 -
net/ipv4/tcp_ipv4.c | 1 -
net/ipv4/tcp_output.c | 3 ---
8 files changed, 2 insertions(+), 15 deletions(-)
--- linux-2.6.16-mm2-full/include/net/arp.h.old 2006-04-04 00:17:55.000000000 +0200
+++ linux-2.6.16-mm2-full/include/net/arp.h 2006-04-04 00:18:01.000000000 +0200
@@ -10,8 +10,6 @@
extern struct neigh_table arp_tbl;
extern void arp_init(void);
-extern int arp_rcv(struct sk_buff *skb, struct net_device *dev,
- struct packet_type *pt, struct net_device *orig_dev);
extern int arp_find(unsigned char *haddr, struct sk_buff *skb);
extern int arp_ioctl(unsigned int cmd, void __user *arg);
extern void arp_send(int type, int ptype, u32 dest_ip,
--- linux-2.6.16-mm2-full/net/ipv4/arp.c.old 2006-04-04 00:18:10.000000000 +0200
+++ linux-2.6.16-mm2-full/net/ipv4/arp.c 2006-04-04 00:18:31.000000000 +0200
@@ -928,7 +928,8 @@
* Receive an arp request from the device layer.
*/
-int arp_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt, struct net_device *orig_dev)
+static int arp_rcv(struct sk_buff *skb, struct net_device *dev,
+ struct packet_type *pt, struct net_device *orig_dev)
{
struct arphdr *arp;
@@ -1417,7 +1418,6 @@
EXPORT_SYMBOL(arp_broken_ops);
EXPORT_SYMBOL(arp_find);
-EXPORT_SYMBOL(arp_rcv);
EXPORT_SYMBOL(arp_create);
EXPORT_SYMBOL(arp_xmit);
EXPORT_SYMBOL(arp_send);
--- linux-2.6.16-mm2-full/net/ipv4/devinet.c.old 2006-04-04 00:46:22.000000000 +0200
+++ linux-2.6.16-mm2-full/net/ipv4/devinet.c 2006-04-04 00:46:31.000000000 +0200
@@ -1556,7 +1556,6 @@
#endif
}
-EXPORT_SYMBOL(devinet_ioctl);
EXPORT_SYMBOL(in_dev_finish_destroy);
EXPORT_SYMBOL(inet_select_addr);
EXPORT_SYMBOL(inetdev_by_index);
--- linux-2.6.16-mm2-full/net/ipv4/fib_frontend.c.old 2006-04-04 00:46:57.000000000 +0200
+++ linux-2.6.16-mm2-full/net/ipv4/fib_frontend.c 2006-04-04 00:47:05.000000000 +0200
@@ -667,4 +667,3 @@
EXPORT_SYMBOL(inet_addr_type);
EXPORT_SYMBOL(ip_dev_find);
-EXPORT_SYMBOL(ip_rt_ioctl);
--- linux-2.6.16-mm2-full/net/ipv4/inet_hashtables.c.old 2006-04-04 00:48:35.000000000 +0200
+++ linux-2.6.16-mm2-full/net/ipv4/inet_hashtables.c 2006-04-04 00:49:43.000000000 +0200
@@ -43,8 +43,6 @@
return tb;
}
-EXPORT_SYMBOL(inet_bind_bucket_create);
-
/*
* Caller must hold hashbucket lock for this tb with local BH disabled
*/
@@ -64,8 +62,6 @@
inet_csk(sk)->icsk_bind_hash = tb;
}
-EXPORT_SYMBOL(inet_bind_hash);
-
/*
* Get rid of any references to a local port held by the given sock.
*/
--- linux-2.6.16-mm2-full/net/ipv4/tcp_input.c.old 2006-04-04 00:50:56.000000000 +0200
+++ linux-2.6.16-mm2-full/net/ipv4/tcp_input.c 2006-04-04 00:51:03.000000000 +0200
@@ -4559,7 +4559,6 @@
EXPORT_SYMBOL(sysctl_tcp_ecn);
EXPORT_SYMBOL(sysctl_tcp_reordering);
-EXPORT_SYMBOL(sysctl_tcp_abc);
EXPORT_SYMBOL(tcp_parse_options);
EXPORT_SYMBOL(tcp_rcv_established);
EXPORT_SYMBOL(tcp_rcv_state_process);
--- linux-2.6.16-mm2-full/net/ipv4/tcp_ipv4.c.old 2006-04-04 00:51:38.000000000 +0200
+++ linux-2.6.16-mm2-full/net/ipv4/tcp_ipv4.c 2006-04-04 00:51:46.000000000 +0200
@@ -1858,5 +1858,4 @@
#endif
EXPORT_SYMBOL(sysctl_local_port_range);
EXPORT_SYMBOL(sysctl_tcp_low_latency);
-EXPORT_SYMBOL(sysctl_tcp_tw_reuse);
--- linux-2.6.16-mm2-full/net/ipv4/tcp_output.c.old 2006-04-04 00:52:13.000000000 +0200
+++ linux-2.6.16-mm2-full/net/ipv4/tcp_output.c 2006-04-04 00:52:32.000000000 +0200
@@ -59,9 +59,6 @@
int sysctl_tcp_mtu_probing = 0;
int sysctl_tcp_base_mss = 512;
-EXPORT_SYMBOL(sysctl_tcp_mtu_probing);
-EXPORT_SYMBOL(sysctl_tcp_base_mss);
-
static void update_send_head(struct sock *sk, struct tcp_sock *tp,
struct sk_buff *skb)
{
^ permalink raw reply
* Re: [RFD][PATCH] typhoon and core sample for folding away VLAN stuff
From: Stephen Hemminger @ 2006-04-12 20:23 UTC (permalink / raw)
To: Ben Greear
Cc: Ingo Oeser, Ingo Oeser, Denis Vlasenko, Dave Dillow, netdev,
David S. Miller, linux-kernel, jgarzik
In-Reply-To: <443D5E9E.80002@candelatech.com>
On Wed, 12 Apr 2006 13:10:06 -0700
Ben Greear <greearb@candelatech.com> wrote:
> What is the reasoning for this change? Is the compiler
> able to optomize the right-hand-side to a constant with your
> change in place?
>
> > - if (veth->h_vlan_proto != __constant_htons(ETH_P_8021Q)) {
> > + if (veth->h_vlan_proto != htons(ETH_P_8021Q)) {
> > return -EINVAL;
> > }
Read the source, the macro handles it.
For i386:
htons maps to __cpu_to_be16
cpu_to_be16 maps to swab16 which is defined in swab.h
# define __swab16(x) \
(__builtin_constant_p((__u16)(x)) ? \
___swab16((x)) : \
__fswab16((x)))
^ permalink raw reply
* Re: [PATCH] atm: clip causes unregister hang
From: Herbert Xu @ 2006-04-12 20:25 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: davem, chas, linux-atm-general, netdev, stable
In-Reply-To: <20060412131527.71f42d58@localhost.localdomain>
On Wed, Apr 12, 2006 at 01:15:27PM -0700, Stephen Hemminger wrote:
>
> If atm_init_atmarp has not been done, then to_atamarpd is a nop because
> atmarpd == NULL.
Good point. But I don't see anything that will keep atmarpd from
closing the socket. Previously the unregister_notifier calls served
as sort of a barrier which prevented the close from proceeding until
we're out of the event handlers.
Now that thw unregister_notifier calls have been moved we need some
other way of ensuring that. Holding the RTNL should be enough.
Cheers,
--
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <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
* Re: [RFD][PATCH] typhoon and core sample for folding away VLAN stuff
From: David S. Miller @ 2006-04-12 20:51 UTC (permalink / raw)
To: greearb; +Cc: ioe-lkml, netdev, vda, dave, netdev, linux-kernel, jgarzik
In-Reply-To: <443D5E9E.80002@candelatech.com>
From: Ben Greear <greearb@candelatech.com>
Date: Wed, 12 Apr 2006 13:10:06 -0700
> What is the reasoning for this change? Is the compiler
> able to optomize the right-hand-side to a constant with your
> change in place?
>
> > - if (veth->h_vlan_proto != __constant_htons(ETH_P_8021Q)) {
> > + if (veth->h_vlan_proto != htons(ETH_P_8021Q)) {
> > return -EINVAL;
> > }
As a policy, Ben, we only use __constant_htons() in compile
time initializers of data structures. Otherwise we use the
normal htons().
That's why.
^ permalink raw reply
* Re: [2.6 patch] net/ipv4/: possible cleanups
From: David S. Miller @ 2006-04-12 20:54 UTC (permalink / raw)
To: bunk; +Cc: netdev
In-Reply-To: <20060412202225.GD6517@stusta.de>
From: Adrian Bunk <bunk@stusta.de>
Date: Wed, 12 Apr 2006 22:22:25 +0200
> Done.
I was eagerly awaiting this after seeing the wan drivers disappear.
Thanks Bunk-bot. :-)
^ permalink raw reply
* [2.6 patch] net/xfrm/xfrm_state: unexport xfrm_state_mtu
From: Adrian Bunk @ 2006-04-12 20:54 UTC (permalink / raw)
To: netdev
This patch removes the unused EXPORT_SYMBOL(xfrm_state_mtu).
Signed-off-by: Adrian Bunk <bunk@stusta.de>
--- linux-2.6.17-rc1-mm2-full/net/xfrm/xfrm_state.c.old 2006-04-12 22:38:57.000000000 +0200
+++ linux-2.6.17-rc1-mm2-full/net/xfrm/xfrm_state.c 2006-04-12 22:39:03.000000000 +0200
@@ -1152,8 +1152,6 @@
return res;
}
-EXPORT_SYMBOL(xfrm_state_mtu);
-
int xfrm_init_state(struct xfrm_state *x)
{
struct xfrm_state_afinfo *afinfo;
^ permalink raw reply
* Re: [PATCH] acxsm: Reduce the number of ACX_PACKED instructions
From: John W. Linville @ 2006-04-12 20:58 UTC (permalink / raw)
To: Carlos Martin; +Cc: netdev, Denis Vlasenko, acx100-devel
In-Reply-To: <11433168763687-git-send-email-carlos@cmartin.tk>
On Sat, Mar 25, 2006 at 09:01:16PM +0100, Carlos Martin wrote:
> Up to now, we were using ACX_PACKED after every field. I've finally
> found out how to use only one at the end of each struct whilst
> maintaining the typedef where it is now.
>
> This should also apply to acx with a bit of fuzz, but I consider it to
> be in maintenance mode, so this doesn't qualify for it.
>
> Signed-off-by: Carlos Martin <carlos@cmartin.tk>
>
> ---
>
> acx_struct.h | 874 +++++++++++++++++++++++++++++-----------------------------
> common.c | 26 +-
> ioctl.c | 10 -
> usb.c | 8 -
> 4 files changed, 457 insertions(+), 461 deletions(-)
>
> 5fb2fdfd8c028a40921bbf9ef7ec4c53c03fcab4
> diff --git a/acx_struct.h b/acx_struct.h
Denis, have you got this patch? Or do I need to apply it?
Carlos, please make your diffs from the root of the kernel tree.
Thanks,
John
--
John W. Linville
linville@tuxdriver.com
^ permalink raw reply
* Re: [patch] net/ieee80211: report hidden ESSIDs as zero-length, not "<hidden>"
From: John W. Linville @ 2006-04-12 21:04 UTC (permalink / raw)
To: Dan Williams; +Cc: netdev, Zhu Yi, James Ketrenos
In-Reply-To: <1143215843.3417.16.camel@localhost.localdomain>
On Fri, Mar 24, 2006 at 10:57:22AM -0500, Dan Williams wrote:
> <hidden> this is something that ieee80211 does that's completely wrong.
> Drivers need to report the _exact_ ESSID from the air in their scan
> results. It's up to the user space app to deal with ESSID length of 0.
What is the outcome of this? Did we get agreement on a patch?
John
--
John W. Linville
linville@tuxdriver.com
^ permalink raw reply
* Re: Window shrinking (was Linux v2.6.16-rc6)
From: Roberto Nibali @ 2006-04-12 21:24 UTC (permalink / raw)
To: Mark Butler; +Cc: andy.furniss, David S. Miller, michal.k.k.piotrowski, netdev
In-Reply-To: <443ABFE1.6030601@middle.net>
>>>> Thanks Mark, I guess packeteer closes window down properly, I
>>>> thought Dave's reply meant that doing that was Treason.
>>>
>>> Packeteer is almost certainly being cavalier about the way it reduces
>>> windows. It could be a serious problem, depending on the way it
>>> treats traffic on the return path. The "treason" thing is a joke.
>>> It is like a bank extending you a credit line one day, and revoking
>>> it the next.
>>
>> I don't use or know of anyone who uses Packeteer - or have you tested?
I had the distinct pleasure of partly get involved with debugging
network stalls related to Linux clients (2.6.x kernel) and a Packeteer.
>> to mean that it was illegal to close down a window at all - you
>> cleared that up - ie it is legal if you close it by <= the amount of
>> data that has just been acked. I assume this won't cause the Treason
>> messaage so don't really understand why it is cavalier - or do you
>> just mean the whole idea of window manipulation to shape may be dodgey
>> but legal?
>
> I thought that Packeteer was causing the error messages. If not then no
> problem. The "treason" messages will not occur if the window is reduced
> normally. The window is there for a reason - namely flow control. A
> zero rwnd means "I can't handle any more data right now". That is
> perfectly legal as long as previously granted transmit credit is not
> withdrawn. Generally speaking the rwnd always drops to zero when the
> receive buffer is full.
Regarding Packeteer traffic shaper and Linux TCP stacks:
A customer of ours has had significant problems with the packeteer
traffic shaper in the past. Unfortunately my consulting contract only
lasted so long as to point out the problem with the shaper in
conjunction with Linux clients, thus I cannot provide you with more
detailed information.
The setup at the customer side (ISP) was like follows: they had
installed a squid-proxy farm for their clients and used the Packeteer to
do some sort of micro-billing, shaping and general QoS. The problem of
window shrinking by the shaper affecting the client's performance
manifested itself most of the time when their customers were accessing a
site via that proxy-farm, which delivered its site-pictures using
content caching services like Akamai (a really horrible example for
testing squid's patience is, among others, http://www.pro-sieben.de).
This caused quite a burst of quick TCP connection setups and teardowns,
eventually resulting in a complete stall for 10-15s.
Disabling the Packeteer traffic shaper completely solved this issue and
customers were not experiencing any stalls anymore when surfing the net.
Updating the firmware of the shaper did help somewhat, so I suspect
their TCP window handling is also error-prone to some extend. So I went
on and blamed the Packeteer traffic shaper device. It was not until I
tested the whole setup with their pilot squid-farm based on Solaris
(SunOS 2.5.1) when I had to rethink my blaming, since routing their
customers over the Solaris squid proxies did not exhibit the problem
when enabling the traffic shaper for the same troublesome websites. So,
this again showed an indication towards an issue with Linux clients and
the Packeteer traffic shaper. The squid-farm is running on a SuSE 9.3
based kernel. Due to performance constraints we had to go with the Linux
solution and disable the traffic shaper.
As soon as I get some more consulting days and if the customer desires,
I'll be debugging this issue in greater detail. I will send a debug
report to netdev in this case. Chances are slim though, since they only
have one Packeteer so far and no test network to perform test conducts,
so erroneous tests would cause major downtime for a lot of this ISP's
customers.
My 2 cents,
Roberto Nibali, ratz
--
echo
'[q]sa[ln0=aln256%Pln256/snlbx]sb3135071790101768542287578439snlbxq' | dc
^ permalink raw reply
* Re: [PATCH] shutting down an interface should remove ipv4 addresses
From: Roberto Nibali @ 2006-04-12 21:38 UTC (permalink / raw)
To: David S. Miller; +Cc: koszik, netdev
In-Reply-To: <20060411.013503.04535700.davem@davemloft.net>
David S. Miller wrote:
> From: Matyas Koszik <koszik@atw.hu>
> Date: Tue, 11 Apr 2006 10:08:01 +0200 (CEST)
>
>> Then it maybe shouldn't affect the flow of packets while the
>> interface is down - or is it also something people depend on?
>
> Yes, people probably do depend upon it.
Not that my voice weights in too much, but I should like to note that
our network and firewall setup depends on this behaviour since the early
2.2.x kernel releases.
2-stage network setup (address assignment and link state) is very much
to our liking :). I would not be too thrilled to change that again.
Best regards,
Roberto Nibali, ratz
--
echo
'[q]sa[ln0=aln256%Pln256/snlbx]sb3135071790101768542287578439snlbxq' | dc
^ permalink raw reply
* kernel panic (on DHCP discover?) in sky2 driver of 2.6.17-rc1
From: Guenther Thomsen @ 2006-04-12 21:42 UTC (permalink / raw)
To: shemminger, John W. Linville; +Cc: netdev
I'm happy to report, that the version of the sky2 driver in 2.6.17-rc1
yields line rate at low CPU utilization (as determined using ttcp).
Unfortunately, it's not quite bug-free yet ;-}
When enabling the second interface (of the same network controller) the
kernel panics (perhaps during DHCP discovery?):
--8<--
[root@penguin1 ~]# ifup eth1
Determining IP information for eth1...Unable to handle kernel paging
request at ffffc20000014000 RIP:
<ffffffff811a3329>{sky2_mac_init+522}
PGD 13fc49067 PUD 13fc4a067 PMD 13fc4b067 PTE 0
Oops: 0000 [1] SMP
CPU 3os linked in: autofs4 sr_mod cdrom dm_mod button usb_storage
uhci_hcd 11BladeRunner_sk98lin #1
RIP: 0010:[<ffffffff811a3329>] <ffffffff811a68 RCX: 000000000000001e
RDX: 0000000000004008 RSI: ffffc20000010000 11 0000000000000fe0 R12:
0000000000001000
R13: ffff81013fae61a8 R14:
S: 010 DS: 0000 ES: 0000 CR0:
000000008005003b
CR2: ffffc20000014000fe40)
Stack: 0000000000001000 ffff81013fae6000 ffff81013fae6500
00000f1013fae6000
Call Trace: <ffffffff811a3e0c>{sky2_up+334} <ffffffff81
<ffffffff81144cf4>{sprintf+144}
<ffffffff8123b616>{inet_ioctl{s_ioctl+44}
<ffffffff81085703>{sys_ioctl+107}
<ffffffff81009a2ac_init+522} RSP <ffff81013487fd28>
CR2: ffffc20000014000
<0>Kerne
-->8--
or (2nd try):
--8<--
[root@penguin1 ~]# Unable to handle kernel paging request at
ffffc20000014000 RIP:
<ffffffff811a3329>{sky2_mac_init+522}
PGD 13fc49067 PUD 13fc4a067 PMD 13fc4b067 PTE 0
Oops: 0000 [1] SMP
CPU 2
Modules linked in: autofs4 sr_mod cdrom dm_mod button usb_storage
uhci_hcd ehci_hcd e752x_edac edac_mc shpcR: 00:[<ffffffff811a3329>]
<ffffffff811a3329>{sky2_mac_init+522}
RDX: 0000000000004008 RSI: ffffc20000010000 RDI: 0000000000000000
R1 0000000001000
R13: ffff81013f0511a8 R14: 0000000000000001 R15: 0000S0000 CR0:
000000008005003b
CR2: ffffc20000014000 CR3: 000000013425b000000000001000 ffff81013f051000
ffff81013f051500 0000000000000000
Call Trace: <ffffffff811a3e0c>{sky2_up+334}
<ffffffff811fea04>{dev_op844cf4>{sprintf+144}
<ffffffff8123b616>{inet_ioctl+74}
<fffffffff81085703>{ys_ioctl+107}
<ffffffff81009ac8>{tracesys+209}
+} RSP <ffff81013534fd28>
CR2: ffffc20000014000
<0>Kernel panic - n
-->8--
The kernel is vanilla 2.6.17-rc1, the sky2 driver was compiled into the
kernel. OS is RedHat Fedora Core 4. The kernel was compiled using
gcc32.
The system is a Blade of a BladeRunner 4130 of Penguincomputing, it
contains two Xeon CPU (+ HT enabled) and an on-board 8062 network
controller of Marvell (88E8062 is stamped on the chip).
The hardware seems to work fine using 2.6.15(.7) with the sk98lin driver
version 8.31 of Syskonnect (skd.de).
Please let me know, if I can provide further information or assist in
any other way.
best regards
Guenther
^ permalink raw reply
* Re: kernel panic (on DHCP discover?) in sky2 driver of 2.6.17-rc1
From: Stephen Hemminger @ 2006-04-12 21:48 UTC (permalink / raw)
To: Guenther Thomsen; +Cc: John W. Linville, netdev
In-Reply-To: <200604121442.56750.gthomsen@bluearc.com>
You need this patch, which Jeff hasn't applied yet.
-----
Subject: sky2: crash when bringing up second port
Sky2 driver will oops referencing bad memory if used on
a dual port card. The problem is accessing past end of
MIB counter space.
Signed-off-by: Stephen Hemminger <shemminger@osdl.org>
--- test-2.6.orig/drivers/net/sky2.c
+++ test-2.6/drivers/net/sky2.c
@@ -579,8 +579,8 @@ static void sky2_mac_init(struct sky2_hw
reg = gma_read16(hw, port, GM_PHY_ADDR);
gma_write16(hw, port, GM_PHY_ADDR, reg | GM_PAR_MIB_CLR);
- for (i = 0; i < GM_MIB_CNT_SIZE; i++)
- gma_read16(hw, port, GM_MIB_CNT_BASE + 8 * i);
+ for (i = GM_MIB_CNT_BASE; i <= GM_MIB_CNT_END; i += 4)
+ gma_read16(hw, port, i);
gma_write16(hw, port, GM_PHY_ADDR, reg);
/* transmit control */
--- test-2.6.orig/drivers/net/sky2.h
+++ test-2.6/drivers/net/sky2.h
@@ -1375,7 +1375,7 @@ enum {
GM_PHY_ADDR = 0x0088, /* 16 bit r/w GPHY Address Register */
/* MIB Counters */
GM_MIB_CNT_BASE = 0x0100, /* Base Address of MIB Counters */
- GM_MIB_CNT_SIZE = 256,
+ GM_MIB_CNT_END = 0x025C, /* Last MIB counter */
};
^ permalink raw reply
* Re: [PATCH] atm: clip causes unregister hang
From: Stephen Hemminger @ 2006-04-12 21:52 UTC (permalink / raw)
To: Herbert Xu, davem; +Cc: chas, linux-atm-general, netdev, stable
In-Reply-To: <20060412202551.GA20085@gondor.apana.org.au>
If Classical IP over ATM module is loaded, its neighbor table gets
populated when permanent neighbor entries are created; but these entries
are not flushed when the device is removed. Since the entry never gets
flushed the unregister of the network device never completes.
This version of the patch also adds locking around the reference to
the atm arp daemon to avoid races with events and daemon state changes.
(Note: barrier() was never really safe)
Bug-reference: http://bugzilla.kernel.org/show_bug.cgi?id=6295
Signed-off-by: Stephen Hemminger <shemminger@osdl.org>
--- clip.orig/net/atm/clip.c 2006-04-12 14:10:45.000000000 -0700
+++ clip/net/atm/clip.c 2006-04-12 14:21:41.000000000 -0700
@@ -613,12 +613,19 @@
static int clip_device_event(struct notifier_block *this,unsigned long event,
- void *dev)
+ void *arg)
{
+ struct net_device *dev = arg;
+
+ if (event == NETDEV_UNREGISTER) {
+ neigh_ifdown(&clip_tbl, dev);
+ return NOTIFY_DONE;
+ }
+
/* ignore non-CLIP devices */
- if (((struct net_device *) dev)->type != ARPHRD_ATM ||
- ((struct net_device *) dev)->hard_start_xmit != clip_start_xmit)
+ if (dev->type != ARPHRD_ATM || dev->hard_start_xmit != clip_start_xmit)
return NOTIFY_DONE;
+
switch (event) {
case NETDEV_UP:
DPRINTK("clip_device_event NETDEV_UP\n");
@@ -686,14 +693,12 @@
static void atmarpd_close(struct atm_vcc *vcc)
{
DPRINTK("atmarpd_close\n");
- atmarpd = NULL; /* assumed to be atomic */
- barrier();
- unregister_inetaddr_notifier(&clip_inet_notifier);
- unregister_netdevice_notifier(&clip_dev_notifier);
- if (skb_peek(&sk_atm(vcc)->sk_receive_queue))
- printk(KERN_ERR "atmarpd_close: closing with requests "
- "pending\n");
+
+ rtnl_lock();
+ atmarpd = NULL;
skb_queue_purge(&sk_atm(vcc)->sk_receive_queue);
+ rtnl_unlock();
+
DPRINTK("(done)\n");
module_put(THIS_MODULE);
}
@@ -714,7 +719,12 @@
static int atm_init_atmarp(struct atm_vcc *vcc)
{
- if (atmarpd) return -EADDRINUSE;
+ rtnl_lock();
+ if (atmarpd) {
+ rtnl_unlock();
+ return -EADDRINUSE;
+ }
+
if (start_timer) {
start_timer = 0;
init_timer(&idle_timer);
@@ -731,10 +741,7 @@
vcc->push = NULL;
vcc->pop = NULL; /* crash */
vcc->push_oam = NULL; /* crash */
- if (register_netdevice_notifier(&clip_dev_notifier))
- printk(KERN_ERR "register_netdevice_notifier failed\n");
- if (register_inetaddr_notifier(&clip_inet_notifier))
- printk(KERN_ERR "register_inetaddr_notifier failed\n");
+ rtnl_unlock();
return 0;
}
@@ -992,6 +999,8 @@
clip_tbl_hook = &clip_tbl;
register_atm_ioctl(&clip_ioctl_ops);
+ register_netdevice_notifier(&clip_dev_notifier);
+ register_inetaddr_notifier(&clip_inet_notifier);
#ifdef CONFIG_PROC_FS
{
@@ -1012,6 +1021,9 @@
remove_proc_entry("arp", atm_proc_root);
+ unregister_inetaddr_notifier(&clip_inet_notifier);
+ unregister_netdevice_notifier(&clip_dev_notifier);
+
deregister_atm_ioctl(&clip_ioctl_ops);
/* First, stop the idle timer, so it stops banging
^ permalink raw reply
* [git patches] net driver fixes
From: Jeff Garzik @ 2006-04-12 22:14 UTC (permalink / raw)
To: Andrew Morton, Linus Torvalds; +Cc: netdev, linux-kernel
Please pull from 'upstream-linus' branch of
master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6.git
to receive the following updates:
drivers/net/b44.c | 64 ++++++---------
drivers/net/bnx2.c | 2
drivers/net/hydra.h | 177 ------------------------------------------
drivers/net/ixgb/ixgb_main.c | 13 ++-
drivers/net/mv643xx_eth.c | 19 +++-
drivers/net/natsemi.c | 2
drivers/net/pcmcia/axnet_cs.c | 2
drivers/net/skge.c | 2
drivers/net/sky2.c | 6 -
drivers/net/sky2.h | 2
drivers/net/starfire.c | 2
drivers/net/typhoon.c | 2
drivers/net/via-rhine.c | 7 -
13 files changed, 67 insertions(+), 233 deletions(-)
Adrian Bunk:
drivers/net/via-rhine.c: make a function static
remove drivers/net/hydra.h
Andreas Schwab:
Use pci_set_consistent_dma_mask in ixgb driver
Brent Cook:
mv643xx_eth: Always free completed tx descs on tx interrupt
Dale Farnsworth:
mv643xx_eth: Fix tx_timeout to only conditionally wake tx queue
Gary Zambrano:
b44: disable default tx pause
b44: increase version to 1.00
Jeff Garzik:
[netdrvr b44] trim trailing whitespace
Komuro:
network: axnet_cs.c: add missing 'PRIV' in ei_rx_overrun
Randy Dunlap:
net drivers: fix section attributes for gcc
Roger Luethi:
via-rhine: execute bounce buffers code on Rhine-I only
Stephen Hemminger:
dlink pci cards using wrong driver
sky2: bad memory reference on dual port cards
diff --git a/drivers/net/b44.c b/drivers/net/b44.c
index c4e12b5..3d30668 100644
--- a/drivers/net/b44.c
+++ b/drivers/net/b44.c
@@ -2,6 +2,7 @@
*
* Copyright (C) 2002 David S. Miller (davem@redhat.com)
* Fixed by Pekka Pietikainen (pp@ee.oulu.fi)
+ * Copyright (C) 2006 Broadcom Corporation.
*
* Distribute under GPL.
*/
@@ -28,8 +29,8 @@
#define DRV_MODULE_NAME "b44"
#define PFX DRV_MODULE_NAME ": "
-#define DRV_MODULE_VERSION "0.97"
-#define DRV_MODULE_RELDATE "Nov 30, 2005"
+#define DRV_MODULE_VERSION "1.00"
+#define DRV_MODULE_RELDATE "Apr 7, 2006"
#define B44_DEF_MSG_ENABLE \
(NETIF_MSG_DRV | \
@@ -136,7 +137,7 @@ static inline unsigned long br32(const s
return readl(bp->regs + reg);
}
-static inline void bw32(const struct b44 *bp,
+static inline void bw32(const struct b44 *bp,
unsigned long reg, unsigned long val)
{
writel(val, bp->regs + reg);
@@ -286,13 +287,13 @@ static void __b44_cam_write(struct b44 *
val |= ((u32) data[4]) << 8;
val |= ((u32) data[5]) << 0;
bw32(bp, B44_CAM_DATA_LO, val);
- val = (CAM_DATA_HI_VALID |
+ val = (CAM_DATA_HI_VALID |
(((u32) data[0]) << 8) |
(((u32) data[1]) << 0));
bw32(bp, B44_CAM_DATA_HI, val);
bw32(bp, B44_CAM_CTRL, (CAM_CTRL_WRITE |
(index << CAM_CTRL_INDEX_SHIFT)));
- b44_wait_bit(bp, B44_CAM_CTRL, CAM_CTRL_BUSY, 100, 1);
+ b44_wait_bit(bp, B44_CAM_CTRL, CAM_CTRL_BUSY, 100, 1);
}
static inline void __b44_disable_ints(struct b44 *bp)
@@ -410,25 +411,18 @@ static void __b44_set_flow_ctrl(struct b
static void b44_set_flow_ctrl(struct b44 *bp, u32 local, u32 remote)
{
- u32 pause_enab = bp->flags & (B44_FLAG_TX_PAUSE |
- B44_FLAG_RX_PAUSE);
+ u32 pause_enab = 0;
- if (local & ADVERTISE_PAUSE_CAP) {
- if (local & ADVERTISE_PAUSE_ASYM) {
- if (remote & LPA_PAUSE_CAP)
- pause_enab |= (B44_FLAG_TX_PAUSE |
- B44_FLAG_RX_PAUSE);
- else if (remote & LPA_PAUSE_ASYM)
- pause_enab |= B44_FLAG_RX_PAUSE;
- } else {
- if (remote & LPA_PAUSE_CAP)
- pause_enab |= (B44_FLAG_TX_PAUSE |
- B44_FLAG_RX_PAUSE);
- }
- } else if (local & ADVERTISE_PAUSE_ASYM) {
- if ((remote & LPA_PAUSE_CAP) &&
- (remote & LPA_PAUSE_ASYM))
- pause_enab |= B44_FLAG_TX_PAUSE;
+ /* The driver supports only rx pause by default because
+ the b44 mac tx pause mechanism generates excessive
+ pause frames.
+ Use ethtool to turn on b44 tx pause if necessary.
+ */
+ if ((local & ADVERTISE_PAUSE_CAP) &&
+ (local & ADVERTISE_PAUSE_ASYM)){
+ if ((remote & LPA_PAUSE_ASYM) &&
+ !(remote & LPA_PAUSE_CAP))
+ pause_enab |= B44_FLAG_RX_PAUSE;
}
__b44_set_flow_ctrl(bp, pause_enab);
@@ -1063,7 +1057,7 @@ static int b44_change_mtu(struct net_dev
spin_unlock_irq(&bp->lock);
b44_enable_ints(bp);
-
+
return 0;
}
@@ -1381,7 +1375,7 @@ static void b44_init_hw(struct b44 *bp)
bw32(bp, B44_DMARX_ADDR, bp->rx_ring_dma + bp->dma_offset);
bw32(bp, B44_DMARX_PTR, bp->rx_pending);
- bp->rx_prod = bp->rx_pending;
+ bp->rx_prod = bp->rx_pending;
bw32(bp, B44_MIB_CTRL, MIB_CTRL_CLR_ON_READ);
@@ -1553,9 +1547,9 @@ static void __b44_set_rx_mode(struct net
val |= RXCONFIG_ALLMULTI;
else
i = __b44_load_mcast(bp, dev);
-
+
for (; i < 64; i++) {
- __b44_cam_write(bp, zero, i);
+ __b44_cam_write(bp, zero, i);
}
bw32(bp, B44_RXCONFIG, val);
val = br32(bp, B44_CAM_CTRL);
@@ -1737,7 +1731,7 @@ static int b44_set_ringparam(struct net_
spin_unlock_irq(&bp->lock);
b44_enable_ints(bp);
-
+
return 0;
}
@@ -1782,7 +1776,7 @@ static int b44_set_pauseparam(struct net
spin_unlock_irq(&bp->lock);
b44_enable_ints(bp);
-
+
return 0;
}
@@ -1898,7 +1892,7 @@ static int __devinit b44_get_invariants(
bp->core_unit = ssb_core_unit(bp);
bp->dma_offset = SB_PCI_DMA;
- /* XXX - really required?
+ /* XXX - really required?
bp->flags |= B44_FLAG_BUGGY_TXPTR;
*/
out:
@@ -1946,7 +1940,7 @@ static int __devinit b44_init_one(struct
"aborting.\n");
goto err_out_free_res;
}
-
+
err = pci_set_consistent_dma_mask(pdev, (u64) B44_DMA_MASK);
if (err) {
printk(KERN_ERR PFX "No usable DMA configuration, "
@@ -2041,9 +2035,9 @@ static int __devinit b44_init_one(struct
pci_save_state(bp->pdev);
- /* Chip reset provides power to the b44 MAC & PCI cores, which
+ /* Chip reset provides power to the b44 MAC & PCI cores, which
* is necessary for MAC register access.
- */
+ */
b44_chip_reset(bp);
printk(KERN_INFO "%s: Broadcom 4400 10/100BaseT Ethernet ", dev->name);
@@ -2091,10 +2085,10 @@ static int b44_suspend(struct pci_dev *p
del_timer_sync(&bp->timer);
- spin_lock_irq(&bp->lock);
+ spin_lock_irq(&bp->lock);
b44_halt(bp);
- netif_carrier_off(bp->dev);
+ netif_carrier_off(bp->dev);
netif_device_detach(bp->dev);
b44_free_rings(bp);
diff --git a/drivers/net/bnx2.c b/drivers/net/bnx2.c
index 2671da2..5ca99e2 100644
--- a/drivers/net/bnx2.c
+++ b/drivers/net/bnx2.c
@@ -63,7 +63,7 @@
/* Time in jiffies before concluding the transmitter is hung. */
#define TX_TIMEOUT (5*HZ)
-static char version[] __devinitdata =
+static const char version[] __devinitdata =
"Broadcom NetXtreme II Gigabit Ethernet Driver " DRV_MODULE_NAME " v" DRV_MODULE_VERSION " (" DRV_MODULE_RELDATE ")\n";
MODULE_AUTHOR("Michael Chan <mchan@broadcom.com>");
diff --git a/drivers/net/hydra.h b/drivers/net/hydra.h
deleted file mode 100644
index 3741414..0000000
--- a/drivers/net/hydra.h
+++ /dev/null
@@ -1,177 +0,0 @@
-/* $Linux: hydra.h,v 1.0 1994/10/26 02:03:47 cgd Exp $ */
-
-/*
- * Copyright (c) 1994 Timo Rossi
- * All rights reserved.
- *
- * Redistribution and use in source and binary forms, with or without
- * modification, are permitted provided that the following conditions
- * are met:
- * 1. Redistributions of source code must retain the above copyright
- * notice, this list of conditions and the following disclaimer.
- * 2. Redistributions in binary form must reproduce the above copyright
- * notice, this list of conditions and the following disclaimer in the
- * documentation and/or other materials provided with the distribution.
- * 3. All advertising materials mentioning features or use of this software
- * must display the following acknowledgement:
- * This product includes software developed by Timo Rossi
- * 4. The name of the author may not be used to endorse or promote products
- * derived from this software without specific prior written permission
- *
- * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
- * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
- * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
- * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
- * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
- * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
- * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
- * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
- * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
- */
-
-/*
- * The Hydra Systems card uses the National Semiconductor
- * 8390 NIC (Network Interface Controller) chip, located
- * at card base address + 0xffe1. NIC registers are accessible
- * only at odd byte addresses, so the register offsets must
- * be multiplied by two.
- *
- * Card address PROM is located at card base + 0xffc0 (even byte addresses)
- *
- * RAM starts at the card base address, and is 16K or 64K.
- * The current Amiga NetBSD hydra driver is hardwired for 16K.
- * It seems that the RAM should be accessed as words or longwords only.
- *
- */
-
-/* adapted for Linux by Topi Kanerva 03/29/95
- with original author's permission */
-
-#define HYDRA_NIC_BASE 0xffe1
-
-/* Page0 registers */
-
-#define NIC_CR 0 /* Command register */
-#define NIC_PSTART (1*2) /* Page start (write) */
-#define NIC_PSTOP (2*2) /* Page stop (write) */
-#define NIC_BNDRY (3*2) /* Boundary pointer */
-#define NIC_TSR (4*2) /* Transmit status (read) */
-#define NIC_TPSR (4*2) /* Transmit page start (write) */
-#define NIC_NCR (5*2) /* Number of collisions, read */
-#define NIC_TBCR0 (5*2) /* Transmit byte count low (write) */
-#define NIC_FIFO (6*2) /* FIFO reg. (read) */
-#define NIC_TBCR1 (6*2) /* Transmit byte count high (write) */
-#define NIC_ISR (7*2) /* Interrupt status register */
-#define NIC_RBCR0 (0xa*2) /* Remote byte count low (write) */
-#define NIC_RBCR1 (0xb*2) /* Remote byte count high (write) */
-#define NIC_RSR (0xc*2) /* Receive status (read) */
-#define NIC_RCR (0xc*2) /* Receive config (write) */
-#define NIC_CNTR0 (0xd*2) /* Frame alignment error count (read) */
-#define NIC_TCR (0xd*2) /* Transmit config (write) */
-#define NIC_CNTR1 (0xe*2) /* CRC error counter (read) */
-#define NIC_DCR (0xe*2) /* Data config (write) */
-#define NIC_CNTR2 (0xf*2) /* missed packet counter (read) */
-#define NIC_IMR (0xf*2) /* Interrupt mask reg. (write) */
-
-/* Page1 registers */
-
-#define NIC_PAR0 (1*2) /* Physical address */
-#define NIC_PAR1 (2*2)
-#define NIC_PAR2 (3*2)
-#define NIC_PAR3 (4*2)
-#define NIC_PAR4 (5*2)
-#define NIC_PAR5 (6*2)
-#define NIC_CURR (7*2) /* Current RX ring-buffer page */
-#define NIC_MAR0 (8*2) /* Multicast address */
-#define NIC_MAR1 (9*2)
-#define NIC_MAR2 (0xa*2)
-#define NIC_MAR3 (0xb*2)
-#define NIC_MAR4 (0xc*2)
-#define NIC_MAR5 (0xd*2)
-#define NIC_MAR6 (0xe*2)
-#define NIC_MAR7 (0xf*2)
-
-/* Command register definitions */
-
-#define CR_STOP 0x01 /* Stop -- software reset command */
-#define CR_START 0x02 /* Start */
-#define CR_TXP 0x04 /* Transmit packet */
-
-#define CR_RD0 0x08 /* Remote DMA cmd */
-#define CR_RD1 0x10
-#define CR_RD2 0x20
-
-#define CR_NODMA CR_RD2
-
-#define CR_PS0 0x40 /* Page select */
-#define CR_PS1 0x80
-
-#define CR_PAGE0 0
-#define CR_PAGE1 CR_PS0
-#define CR_PAGE2 CR_PS1
-
-/* Interrupt status reg. definitions */
-
-#define ISR_PRX 0x01 /* Packet received without errors */
-#define ISR_PTX 0x02 /* Packet transmitted without errors */
-#define ISR_RXE 0x04 /* Receive error */
-#define ISR_TXE 0x08 /* Transmit error */
-#define ISR_OVW 0x10 /* Ring buffer overrun */
-#define ISR_CNT 0x20 /* Counter overflow */
-#define ISR_RDC 0x40 /* Remote DMA compile */
-#define ISR_RST 0x80 /* Reset status */
-
-/* Data config reg. definitions */
-
-#define DCR_WTS 0x01 /* Word transfer select */
-#define DCR_BOS 0x02 /* Byte order select */
-#define DCR_LAS 0x04 /* Long address select */
-#define DCR_LS 0x08 /* Loopback select */
-#define DCR_AR 0x10 /* Auto-init remote */
-#define DCR_FT0 0x20 /* FIFO threshold select */
-#define DCR_FT1 0x40
-
-/* Transmit config reg. definitions */
-
-#define TCR_CRC 0x01 /* Inhibit CRC */
-#define TCR_LB0 0x02 /* Loopback control */
-#define TCR_LB1 0x04
-#define TCR_ATD 0x08 /* Auto transmit disable */
-#define TCR_OFST 0x10 /* Collision offset enable */
-
-/* Transmit status reg. definitions */
-
-#define TSR_PTX 0x01 /* Packet transmitted */
-#define TSR_COL 0x04 /* Transmit collided */
-#define TSR_ABT 0x08 /* Transmit aborted */
-#define TSR_CRS 0x10 /* Carrier sense lost */
-#define TSR_FU 0x20 /* FIFO underrun */
-#define TSR_CDH 0x40 /* CD Heartbeat */
-#define TSR_OWC 0x80 /* Out of Window Collision */
-
-/* Receiver config register definitions */
-
-#define RCR_SEP 0x01 /* Save errored packets */
-#define RCR_AR 0x02 /* Accept runt packets */
-#define RCR_AB 0x04 /* Accept broadcast */
-#define RCR_AM 0x08 /* Accept multicast */
-#define RCR_PRO 0x10 /* Promiscuous mode */
-#define RCR_MON 0x20 /* Monitor mode */
-
-/* Receiver status register definitions */
-
-#define RSR_PRX 0x01 /* Packet received without error */
-#define RSR_CRC 0x02 /* CRC error */
-#define RSR_FAE 0x04 /* Frame alignment error */
-#define RSR_FO 0x08 /* FIFO overrun */
-#define RSR_MPA 0x10 /* Missed packet */
-#define RSR_PHY 0x20 /* Physical address */
-#define RSR_DIS 0x40 /* Received disabled */
-#define RSR_DFR 0x80 /* Deferring (jabber) */
-
-/* Hydra System card address PROM offset */
-
-#define HYDRA_ADDRPROM 0xffc0
-
-
diff --git a/drivers/net/ixgb/ixgb_main.c b/drivers/net/ixgb/ixgb_main.c
index f9f77e4..cfd67d8 100644
--- a/drivers/net/ixgb/ixgb_main.c
+++ b/drivers/net/ixgb/ixgb_main.c
@@ -357,18 +357,20 @@ ixgb_probe(struct pci_dev *pdev,
if((err = pci_enable_device(pdev)))
return err;
- if(!(err = pci_set_dma_mask(pdev, DMA_64BIT_MASK))) {
+ if(!(err = pci_set_dma_mask(pdev, DMA_64BIT_MASK)) &&
+ !(err = pci_set_consistent_dma_mask(pdev, DMA_64BIT_MASK))) {
pci_using_dac = 1;
} else {
- if((err = pci_set_dma_mask(pdev, DMA_32BIT_MASK))) {
+ if((err = pci_set_dma_mask(pdev, DMA_32BIT_MASK)) ||
+ (err = pci_set_consistent_dma_mask(pdev, DMA_32BIT_MASK))) {
IXGB_ERR("No usable DMA configuration, aborting\n");
- return err;
+ goto err_dma_mask;
}
pci_using_dac = 0;
}
if((err = pci_request_regions(pdev, ixgb_driver_name)))
- return err;
+ goto err_request_regions;
pci_set_master(pdev);
@@ -502,6 +504,9 @@ err_ioremap:
free_netdev(netdev);
err_alloc_etherdev:
pci_release_regions(pdev);
+err_request_regions:
+err_dma_mask:
+ pci_disable_device(pdev);
return err;
}
diff --git a/drivers/net/mv643xx_eth.c b/drivers/net/mv643xx_eth.c
index 9f26613..ea62a3e 100644
--- a/drivers/net/mv643xx_eth.c
+++ b/drivers/net/mv643xx_eth.c
@@ -281,10 +281,16 @@ static void mv643xx_eth_tx_timeout_task(
{
struct mv643xx_private *mp = netdev_priv(dev);
- netif_device_detach(dev);
+ if (!netif_running(dev))
+ return;
+
+ netif_stop_queue(dev);
+
eth_port_reset(mp->port_num);
eth_port_start(dev);
- netif_device_attach(dev);
+
+ if (mp->tx_ring_size - mp->tx_desc_count >= MAX_DESCS_PER_SKB)
+ netif_wake_queue(dev);
}
/**
@@ -552,9 +558,9 @@ static irqreturn_t mv643xx_eth_int_handl
#else
if (eth_int_cause & ETH_INT_CAUSE_RX)
mv643xx_eth_receive_queue(dev, INT_MAX);
+#endif
if (eth_int_cause_ext & ETH_INT_CAUSE_TX)
mv643xx_eth_free_completed_tx_descs(dev);
-#endif
/*
* If no real interrupt occured, exit.
@@ -1186,7 +1192,12 @@ static int mv643xx_eth_start_xmit(struct
BUG_ON(netif_queue_stopped(dev));
BUG_ON(skb == NULL);
- BUG_ON(mp->tx_ring_size - mp->tx_desc_count < MAX_DESCS_PER_SKB);
+
+ if (mp->tx_ring_size - mp->tx_desc_count < MAX_DESCS_PER_SKB) {
+ printk(KERN_ERR "%s: transmit with queue full\n", dev->name);
+ netif_stop_queue(dev);
+ return 1;
+ }
if (has_tiny_unaligned_frags(skb)) {
if ((skb_linearize(skb, GFP_ATOMIC) != 0)) {
diff --git a/drivers/net/natsemi.c b/drivers/net/natsemi.c
index 7826afb..9062775 100644
--- a/drivers/net/natsemi.c
+++ b/drivers/net/natsemi.c
@@ -238,7 +238,7 @@ static int full_duplex[MAX_UNITS];
#define NATSEMI_RX_LIMIT 2046 /* maximum supported by hardware */
/* These identify the driver base version and may not be removed. */
-static char version[] __devinitdata =
+static const char version[] __devinitdata =
KERN_INFO DRV_NAME " dp8381x driver, version "
DRV_VERSION ", " DRV_RELDATE "\n"
KERN_INFO " originally by Donald Becker <becker@scyld.com>\n"
diff --git a/drivers/net/pcmcia/axnet_cs.c b/drivers/net/pcmcia/axnet_cs.c
index 56233af..448a094 100644
--- a/drivers/net/pcmcia/axnet_cs.c
+++ b/drivers/net/pcmcia/axnet_cs.c
@@ -1560,7 +1560,7 @@ static void ei_receive(struct net_device
static void ei_rx_overrun(struct net_device *dev)
{
- axnet_dev_t *info = (axnet_dev_t *)dev;
+ axnet_dev_t *info = PRIV(dev);
long e8390_base = dev->base_addr;
unsigned char was_txing, must_resend = 0;
struct ei_device *ei_local = (struct ei_device *) netdev_priv(dev);
diff --git a/drivers/net/skge.c b/drivers/net/skge.c
index 35dbf05..a70c2b0 100644
--- a/drivers/net/skge.c
+++ b/drivers/net/skge.c
@@ -78,6 +78,8 @@ static const struct pci_device_id skge_i
{ PCI_DEVICE(PCI_VENDOR_ID_SYSKONNECT, PCI_DEVICE_ID_SYSKONNECT_GE) },
{ PCI_DEVICE(PCI_VENDOR_ID_SYSKONNECT, PCI_DEVICE_ID_SYSKONNECT_YU) },
{ PCI_DEVICE(PCI_VENDOR_ID_DLINK, PCI_DEVICE_ID_DLINK_DGE510T), },
+ { PCI_DEVICE(PCI_VENDOR_ID_DLINK, 0x4b00) },
+ { PCI_DEVICE(PCI_VENDOR_ID_DLINK, 0x4b01) },
{ PCI_DEVICE(PCI_VENDOR_ID_MARVELL, 0x4320) },
{ PCI_DEVICE(PCI_VENDOR_ID_MARVELL, 0x5005) }, /* Belkin */
{ PCI_DEVICE(PCI_VENDOR_ID_CNET, PCI_DEVICE_ID_CNET_GIGACARD) },
diff --git a/drivers/net/sky2.c b/drivers/net/sky2.c
index 68f9c20..67b0eab 100644
--- a/drivers/net/sky2.c
+++ b/drivers/net/sky2.c
@@ -99,8 +99,6 @@ MODULE_PARM_DESC(disable_msi, "Disable M
static const struct pci_device_id sky2_id_table[] = {
{ PCI_DEVICE(PCI_VENDOR_ID_SYSKONNECT, 0x9000) },
{ PCI_DEVICE(PCI_VENDOR_ID_SYSKONNECT, 0x9E00) },
- { PCI_DEVICE(PCI_VENDOR_ID_DLINK, 0x4b00) },
- { PCI_DEVICE(PCI_VENDOR_ID_DLINK, 0x4b01) },
{ PCI_DEVICE(PCI_VENDOR_ID_MARVELL, 0x4340) },
{ PCI_DEVICE(PCI_VENDOR_ID_MARVELL, 0x4341) },
{ PCI_DEVICE(PCI_VENDOR_ID_MARVELL, 0x4342) },
@@ -579,8 +577,8 @@ static void sky2_mac_init(struct sky2_hw
reg = gma_read16(hw, port, GM_PHY_ADDR);
gma_write16(hw, port, GM_PHY_ADDR, reg | GM_PAR_MIB_CLR);
- for (i = 0; i < GM_MIB_CNT_SIZE; i++)
- gma_read16(hw, port, GM_MIB_CNT_BASE + 8 * i);
+ for (i = GM_MIB_CNT_BASE; i <= GM_MIB_CNT_END; i += 4)
+ gma_read16(hw, port, i);
gma_write16(hw, port, GM_PHY_ADDR, reg);
/* transmit control */
diff --git a/drivers/net/sky2.h b/drivers/net/sky2.h
index 62532b4..89dd18c 100644
--- a/drivers/net/sky2.h
+++ b/drivers/net/sky2.h
@@ -1375,7 +1375,7 @@ enum {
GM_PHY_ADDR = 0x0088, /* 16 bit r/w GPHY Address Register */
/* MIB Counters */
GM_MIB_CNT_BASE = 0x0100, /* Base Address of MIB Counters */
- GM_MIB_CNT_SIZE = 256,
+ GM_MIB_CNT_END = 0x025C, /* Last MIB counter */
};
diff --git a/drivers/net/starfire.c b/drivers/net/starfire.c
index 45ad036..9b7805b 100644
--- a/drivers/net/starfire.c
+++ b/drivers/net/starfire.c
@@ -335,7 +335,7 @@ do { \
/* These identify the driver base version and may not be removed. */
-static char version[] __devinitdata =
+static const char version[] __devinitdata =
KERN_INFO "starfire.c:v1.03 7/26/2000 Written by Donald Becker <becker@scyld.com>\n"
KERN_INFO " (unofficial 2.2/2.4 kernel port, version " DRV_VERSION ", " DRV_RELDATE ")\n";
diff --git a/drivers/net/typhoon.c b/drivers/net/typhoon.c
index c1ce87a..d9258d4 100644
--- a/drivers/net/typhoon.c
+++ b/drivers/net/typhoon.c
@@ -134,7 +134,7 @@ static const int multicast_filter_limit
#include "typhoon.h"
#include "typhoon-firmware.h"
-static char version[] __devinitdata =
+static const char version[] __devinitdata =
"typhoon.c: version " DRV_MODULE_VERSION " (" DRV_MODULE_RELDATE ")\n";
MODULE_AUTHOR("David Dillow <dave@thedillows.org>");
diff --git a/drivers/net/via-rhine.c b/drivers/net/via-rhine.c
index a9b2150..6a23964 100644
--- a/drivers/net/via-rhine.c
+++ b/drivers/net/via-rhine.c
@@ -469,7 +469,7 @@ struct rhine_private {
struct sk_buff *tx_skbuff[TX_RING_SIZE];
dma_addr_t tx_skbuff_dma[TX_RING_SIZE];
- /* Tx bounce buffers */
+ /* Tx bounce buffers (Rhine-I only) */
unsigned char *tx_buf[TX_RING_SIZE];
unsigned char *tx_bufs;
dma_addr_t tx_bufs_dma;
@@ -1043,7 +1043,8 @@ static void alloc_tbufs(struct net_devic
rp->tx_ring[i].desc_length = cpu_to_le32(TXDESC);
next += sizeof(struct tx_desc);
rp->tx_ring[i].next_desc = cpu_to_le32(next);
- rp->tx_buf[i] = &rp->tx_bufs[i * PKT_BUF_SZ];
+ if (rp->quirks & rqRhineI)
+ rp->tx_buf[i] = &rp->tx_bufs[i * PKT_BUF_SZ];
}
rp->tx_ring[i-1].next_desc = cpu_to_le32(rp->tx_ring_dma);
@@ -1091,7 +1092,7 @@ static void rhine_check_media(struct net
}
/* Called after status of force_media possibly changed */
-void rhine_set_carrier(struct mii_if_info *mii)
+static void rhine_set_carrier(struct mii_if_info *mii)
{
if (mii->force_media) {
/* autoneg is off: Link is always assumed to be up */
^ permalink raw reply related
* Re: kernel panic (on DHCP discover?) in sky2 driver of 2.6.17-rc1
From: Guenther Thomsen @ 2006-04-12 22:26 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: John W. Linville, netdev
In-Reply-To: <20060412144812.6b679204@localhost.localdomain>
On Wednesday 12 April 2006 14:48, Stephen Hemminger wrote:
> You need this patch, which Jeff hasn't applied yet.
> -----
> Subject: sky2: crash when bringing up second port
>
> Sky2 driver will oops referencing bad memory if used on
> a dual port card. The problem is accessing past end of
> MIB counter space.
>
> Signed-off-by: Stephen Hemminger <shemminger@osdl.org>
>
>
> --- test-2.6.orig/drivers/net/sky2.c
> +++ test-2.6/drivers/net/sky2.c
> @@ -579,8 +579,8 @@ static void sky2_mac_init(struct sky2_hw
> reg = gma_read16(hw, port, GM_PHY_ADDR);
> gma_write16(hw, port, GM_PHY_ADDR, reg | GM_PAR_MIB_CLR);
>
> - for (i = 0; i < GM_MIB_CNT_SIZE; i++)
> - gma_read16(hw, port, GM_MIB_CNT_BASE + 8 * i);
> + for (i = GM_MIB_CNT_BASE; i <= GM_MIB_CNT_END; i += 4)
> + gma_read16(hw, port, i);
> gma_write16(hw, port, GM_PHY_ADDR, reg);
>
> /* transmit control */
> --- test-2.6.orig/drivers/net/sky2.h
> +++ test-2.6/drivers/net/sky2.h
> @@ -1375,7 +1375,7 @@ enum {
> GM_PHY_ADDR = 0x0088, /* 16 bit r/w GPHY Address Register */
> /* MIB Counters */
> GM_MIB_CNT_BASE = 0x0100, /* Base Address of MIB Counters */
> - GM_MIB_CNT_SIZE = 256,
> + GM_MIB_CNT_END = 0x025C, /* Last MIB counter */
> };
Thanks for the very quick response. The patch indeed prevents the panic
when bringing up the second interface, but now the host doesn't receive
any packets anymore. It still sends packets (ARP requests, naturally).
If I inject the Ethernet address of a second host into the arp table of
the test subject, ICMP Echo requests are sent, but then sendmsg's
buffer space is exhausted (?):
--8<--
[root@penguin1 ~]# arp -s 192.168.65.67 00:A0:D1:E1:F3:2C
[root@penguin1 ~]# ping 192.168.65.67
PING 192.168.65.67 (192.168.65.67) 56(84) bytes of data.
ping: sendmsg: No buffer space available
ping: sendmsg: No buffer space available
ping: sendmsg: No buffer space available
ping: sendmsg: No buffer space available
ping: sendmsg: No buffer space available
ping: sendmsg: No buffer space available
--- 192.168.65.67 ping statistics ---
19 packets transmitted, 0 received, 100% packet loss, time 37012ms
-->8--
There is no hint of a malfunction to be found in the kernel's message
buffer.
best regards
Guenther
^ permalink raw reply
* [PATCH] atm: clip timer race
From: Stephen Hemminger @ 2006-04-12 22:42 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: Herbert Xu, davem, chas, linux-atm-general, netdev
In-Reply-To: <20060412145254.4dd21be6@localhost.localdomain>
By inspection, the clip idle timer code is racy on SMP.
Here is a safe version of timer management.
Untested, I don't have ATM hardware.
Signed-off-by: Stephen Hemminger <shemminger@osdl.org>
--- clip.orig/net/atm/clip.c 2006-04-12 14:24:10.000000000 -0700
+++ clip/net/atm/clip.c 2006-04-12 14:40:01.000000000 -0700
@@ -54,8 +54,6 @@
static struct atm_vcc *atmarpd;
static struct neigh_table clip_tbl;
static struct timer_list idle_timer;
-static int start_timer = 1;
-
static int to_atmarpd(enum atmarp_ctrl_type type,int itf,unsigned long ip)
{
@@ -725,13 +723,8 @@
return -EADDRINUSE;
}
- if (start_timer) {
- start_timer = 0;
- init_timer(&idle_timer);
- idle_timer.expires = jiffies+CLIP_CHECK_INTERVAL*HZ;
- idle_timer.function = idle_timer_check;
- add_timer(&idle_timer);
- }
+ mod_timer(&idle_timer, jiffies+CLIP_CHECK_INTERVAL*HZ);
+
atmarpd = vcc;
set_bit(ATM_VF_META,&vcc->flags);
set_bit(ATM_VF_READY,&vcc->flags);
@@ -1002,6 +995,8 @@
register_netdevice_notifier(&clip_dev_notifier);
register_inetaddr_notifier(&clip_inet_notifier);
+ setup_timer(&idle_timer, idle_timer_check, 0);
+
#ifdef CONFIG_PROC_FS
{
struct proc_dir_entry *p;
@@ -1029,8 +1024,7 @@
/* First, stop the idle timer, so it stops banging
* on the table.
*/
- if (start_timer == 0)
- del_timer(&idle_timer);
+ del_timer_sync(&idle_timer);
/* Next, purge the table, so that the device
* unregister loop below does not hang due to
^ permalink raw reply
* Re: [git patches] net driver fixes
From: Kumar Gala @ 2006-04-12 22:43 UTC (permalink / raw)
To: Jeff Garzik; +Cc: Andrew Morton, Linus Torvalds, netdev, linux-kernel
In-Reply-To: <20060412221437.GA20899@havoc.gtf.org>
Jeff,
Noticed Andy's gianfar fixes aren't in here. I'd be good to see if
we can get them in for 2.6.17
- kumar
On Apr 12, 2006, at 5:14 PM, Jeff Garzik wrote:
>
> Please pull from 'upstream-linus' branch of
> master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6.git
>
> to receive the following updates:
>
> drivers/net/b44.c | 64 ++++++---------
> drivers/net/bnx2.c | 2
> drivers/net/hydra.h | 177
> ------------------------------------------
> drivers/net/ixgb/ixgb_main.c | 13 ++-
> drivers/net/mv643xx_eth.c | 19 +++-
> drivers/net/natsemi.c | 2
> drivers/net/pcmcia/axnet_cs.c | 2
> drivers/net/skge.c | 2
> drivers/net/sky2.c | 6 -
> drivers/net/sky2.h | 2
> drivers/net/starfire.c | 2
> drivers/net/typhoon.c | 2
> drivers/net/via-rhine.c | 7 -
> 13 files changed, 67 insertions(+), 233 deletions(-)
>
> Adrian Bunk:
> drivers/net/via-rhine.c: make a function static
> remove drivers/net/hydra.h
>
> Andreas Schwab:
> Use pci_set_consistent_dma_mask in ixgb driver
>
> Brent Cook:
> mv643xx_eth: Always free completed tx descs on tx interrupt
>
> Dale Farnsworth:
> mv643xx_eth: Fix tx_timeout to only conditionally wake tx queue
>
> Gary Zambrano:
> b44: disable default tx pause
> b44: increase version to 1.00
>
> Jeff Garzik:
> [netdrvr b44] trim trailing whitespace
>
> Komuro:
> network: axnet_cs.c: add missing 'PRIV' in ei_rx_overrun
>
> Randy Dunlap:
> net drivers: fix section attributes for gcc
>
> Roger Luethi:
> via-rhine: execute bounce buffers code on Rhine-I only
>
> Stephen Hemminger:
> dlink pci cards using wrong driver
> sky2: bad memory reference on dual port cards
^ permalink raw reply
* Re: [git patches] net driver fixes
From: Jeff Garzik @ 2006-04-12 22:45 UTC (permalink / raw)
To: Kumar Gala; +Cc: Andrew Morton, Linus Torvalds, netdev, linux-kernel
In-Reply-To: <53E69375-551D-41FC-9D09-E3D39EC80A19@kernel.crashing.org>
Kumar Gala wrote:
> Jeff,
>
> Noticed Andy's gianfar fixes aren't in here. I'd be good to see if we
> can get them in for 2.6.17
Never saw them...
^ permalink raw reply
* Re: [git patches] net driver fixes
From: Kumar Gala @ 2006-04-12 22:47 UTC (permalink / raw)
To: Jeff Garzik; +Cc: Andrew Morton, Linus Torvalds, netdev, linux-kernel
In-Reply-To: <443D8310.5020304@garzik.org>
On Apr 12, 2006, at 5:45 PM, Jeff Garzik wrote:
> Kumar Gala wrote:
>> Jeff,
>> Noticed Andy's gianfar fixes aren't in here. I'd be good to see
>> if we can get them in for 2.6.17
>
> Never saw them...
Odd, posted to netdev, Andy may not have copied you on them.
http://marc.theaimsgroup.com/?l=linux-netdev&m=114437381506340&w=2
^ permalink raw reply
* Re: [git patches] net driver fixes
From: Jeff Garzik @ 2006-04-12 23:04 UTC (permalink / raw)
To: Kumar Gala; +Cc: Andrew Morton, Linus Torvalds, netdev, linux-kernel
In-Reply-To: <1BF241D7-89AE-4519-A56F-9241CE5D57CF@kernel.crashing.org>
Kumar Gala wrote:
>
> On Apr 12, 2006, at 5:45 PM, Jeff Garzik wrote:
>
>> Kumar Gala wrote:
>>> Jeff,
>>> Noticed Andy's gianfar fixes aren't in here. I'd be good to see if
>>> we can get them in for 2.6.17
>>
>> Never saw them...
>
> Odd, posted to netdev, Andy may not have copied you on them.
>
> http://marc.theaimsgroup.com/?l=linux-netdev&m=114437381506340&w=2
That's fine, but I don't have it. Please resend, git-applymbox doesn't
work on URLs.
Jeff
^ permalink raw reply
* [PATCH] Fix locking in gianfar
From: Andy Fleming @ 2006-04-07 1:36 UTC (permalink / raw)
To: Netdev
This patch fixes several bugs in the gianfar driver, including a
major one
where spinlocks were horribly broken:
* Split gianfar locks into two types: TX and RX
* Made it so gfar_start() now clears RHALT
* Fixed a bug where calling gfar_start_xmit() with interrupts off would
corrupt the interrupt state
* Fixed a bug where a frame could potentially arrive, and never be
handled
(if no more frames arrived
* Fixed a bug where the rx_work_limit would never be observed by the rx
completion code
* Fixed a bug where the interrupt handlers were not actually
protected by
their spinlocks
Signed-off-by: Andy Fleming <afleming@freescale.com>
---
drivers/net/gianfar.c | 56 ++++++++++++++++
+-----------------
drivers/net/gianfar.h | 67 +++++++++++++++++++++++++++
+-------------
drivers/net/gianfar_ethtool.c | 20 +++++++++---
drivers/net/gianfar_sysfs.c | 24 +++++++--------
4 files changed, 100 insertions(+), 67 deletions(-)
5b638b01fefa46929a284b48e51ae36ad0c63009
diff --git a/drivers/net/gianfar.c b/drivers/net/gianfar.c
index 771e25d..218d317 100644
--- a/drivers/net/gianfar.c
+++ b/drivers/net/gianfar.c
@@ -210,7 +210,8 @@ static int gfar_probe(struct platform_de
goto regs_fail;
}
- spin_lock_init(&priv->lock);
+ spin_lock_init(&priv->txlock);
+ spin_lock_init(&priv->rxlock);
platform_set_drvdata(pdev, dev);
@@ -515,11 +516,13 @@ void stop_gfar(struct net_device *dev)
phy_stop(priv->phydev);
/* Lock it down */
- spin_lock_irqsave(&priv->lock, flags);
+ spin_lock_irqsave(&priv->txlock, flags);
+ spin_lock(&priv->rxlock);
gfar_halt(dev);
- spin_unlock_irqrestore(&priv->lock, flags);
+ spin_unlock(&priv->rxlock);
+ spin_unlock_irqrestore(&priv->txlock, flags);
/* Free the IRQs */
if (priv->einfo->device_flags & FSL_GIANFAR_DEV_HAS_MULTI_INTR) {
@@ -605,14 +608,15 @@ void gfar_start(struct net_device *dev)
tempval |= DMACTRL_INIT_SETTINGS;
gfar_write(&priv->regs->dmactrl, tempval);
- /* Clear THLT, so that the DMA starts polling now */
- gfar_write(®s->tstat, TSTAT_CLEAR_THALT);
-
/* Make sure we aren't stopped */
tempval = gfar_read(&priv->regs->dmactrl);
tempval &= ~(DMACTRL_GRS | DMACTRL_GTS);
gfar_write(&priv->regs->dmactrl, tempval);
+ /* Clear THLT/RHLT, so that the DMA starts polling now */
+ gfar_write(®s->tstat, TSTAT_CLEAR_THALT);
+ gfar_write(®s->rstat, RSTAT_CLEAR_RHALT);
+
/* Unmask the interrupts we look for */
gfar_write(®s->imask, IMASK_DEFAULT);
}
@@ -928,12 +932,13 @@ static int gfar_start_xmit(struct sk_buf
struct txfcb *fcb = NULL;
struct txbd8 *txbdp;
u16 status;
+ unsigned long flags;
/* Update transmit stats */
priv->stats.tx_bytes += skb->len;
/* Lock priv now */
- spin_lock_irq(&priv->lock);
+ spin_lock_irqsave(&priv->txlock, flags);
/* Point at the first free tx descriptor */
txbdp = priv->cur_tx;
@@ -1004,7 +1009,7 @@ static int gfar_start_xmit(struct sk_buf
gfar_write(&priv->regs->tstat, TSTAT_CLEAR_THALT);
/* Unlock priv */
- spin_unlock_irq(&priv->lock);
+ spin_unlock_irqrestore(&priv->txlock, flags);
return 0;
}
@@ -1049,7 +1054,7 @@ static void gfar_vlan_rx_register(struct
unsigned long flags;
u32 tempval;
- spin_lock_irqsave(&priv->lock, flags);
+ spin_lock_irqsave(&priv->rxlock, flags);
priv->vlgrp = grp;
@@ -1076,7 +1081,7 @@ static void gfar_vlan_rx_register(struct
gfar_write(&priv->regs->rctrl, tempval);
}
- spin_unlock_irqrestore(&priv->lock, flags);
+ spin_unlock_irqrestore(&priv->rxlock, flags);
}
@@ -1085,12 +1090,12 @@ static void gfar_vlan_rx_kill_vid(struct
struct gfar_private *priv = netdev_priv(dev);
unsigned long flags;
- spin_lock_irqsave(&priv->lock, flags);
+ spin_lock_irqsave(&priv->rxlock, flags);
if (priv->vlgrp)
priv->vlgrp->vlan_devices[vid] = NULL;
- spin_unlock_irqrestore(&priv->lock, flags);
+ spin_unlock_irqrestore(&priv->rxlock, flags);
}
@@ -1179,7 +1184,7 @@ static irqreturn_t gfar_transmit(int irq
gfar_write(&priv->regs->ievent, IEVENT_TX_MASK);
/* Lock priv */
- spin_lock(&priv->lock);
+ spin_lock(&priv->txlock);
bdp = priv->dirty_tx;
while ((bdp->status & TXBD_READY) == 0) {
/* If dirty_tx and cur_tx are the same, then either the */
@@ -1224,7 +1229,7 @@ static irqreturn_t gfar_transmit(int irq
else
gfar_write(&priv->regs->txic, 0);
- spin_unlock(&priv->lock);
+ spin_unlock(&priv->txlock);
return IRQ_HANDLED;
}
@@ -1305,9 +1310,10 @@ irqreturn_t gfar_receive(int irq, void *
{
struct net_device *dev = (struct net_device *) dev_id;
struct gfar_private *priv = netdev_priv(dev);
-
#ifdef CONFIG_GFAR_NAPI
u32 tempval;
+#else
+ unsigned long flags;
#endif
/* Clear IEVENT, so rx interrupt isn't called again
@@ -1330,7 +1336,7 @@ irqreturn_t gfar_receive(int irq, void *
}
#else
- spin_lock(&priv->lock);
+ spin_lock_irqsave(&priv->rxlock, flags);
gfar_clean_rx_ring(dev, priv->rx_ring_size);
/* If we are coalescing interrupts, update the timer */
@@ -1341,7 +1347,7 @@ irqreturn_t gfar_receive(int irq, void *
else
gfar_write(&priv->regs->rxic, 0);
- spin_unlock(&priv->lock);
+ spin_unlock_irqrestore(&priv->rxlock, flags);
#endif
return IRQ_HANDLED;
@@ -1490,13 +1496,6 @@ int gfar_clean_rx_ring(struct net_device
/* Update the current rxbd pointer to be the next one */
priv->cur_rx = bdp;
- /* If no packets have arrived since the
- * last one we processed, clear the IEVENT RX and
- * BSY bits so that another interrupt won't be
- * generated when we set IMASK */
- if (bdp->status & RXBD_EMPTY)
- gfar_write(&priv->regs->ievent, IEVENT_RX_MASK);
-
return howmany;
}
@@ -1516,7 +1515,7 @@ static int gfar_poll(struct net_device *
rx_work_limit -= howmany;
*budget -= howmany;
- if (rx_work_limit >= 0) {
+ if (rx_work_limit > 0) {
netif_rx_complete(dev);
/* Clear the halt bit in RSTAT */
@@ -1533,7 +1532,8 @@ static int gfar_poll(struct net_device *
gfar_write(&priv->regs->rxic, 0);
}
- return (rx_work_limit < 0) ? 1 : 0;
+ /* Return 1 if there's more work to do */
+ return (rx_work_limit > 0) ? 0 : 1;
}
#endif
@@ -1629,7 +1629,7 @@ static void adjust_link(struct net_devic
struct phy_device *phydev = priv->phydev;
int new_state = 0;
- spin_lock_irqsave(&priv->lock, flags);
+ spin_lock_irqsave(&priv->txlock, flags);
if (phydev->link) {
u32 tempval = gfar_read(®s->maccfg2);
u32 ecntrl = gfar_read(®s->ecntrl);
@@ -1694,7 +1694,7 @@ static void adjust_link(struct net_devic
if (new_state && netif_msg_link(priv))
phy_print_status(phydev);
- spin_unlock_irqrestore(&priv->lock, flags);
+ spin_unlock_irqrestore(&priv->txlock, flags);
}
/* Update the hash table based on the current list of multicast
diff --git a/drivers/net/gianfar.h b/drivers/net/gianfar.h
index d37d540..127c98c 100644
--- a/drivers/net/gianfar.h
+++ b/drivers/net/gianfar.h
@@ -656,43 +656,62 @@ struct gfar {
* the buffer descriptor determines the actual condition.
*/
struct gfar_private {
- /* pointers to arrays of skbuffs for tx and rx */
+ /* Fields controlled by TX lock */
+ spinlock_t txlock;
+
+ /* Pointer to the array of skbuffs */
struct sk_buff ** tx_skbuff;
- struct sk_buff ** rx_skbuff;
- /* indices pointing to the next free sbk in skb arrays */
+ /* next free skb in the array */
u16 skb_curtx;
- u16 skb_currx;
- /* index of the first skb which hasn't been transmitted
- * yet. */
+ /* First skb in line to be transmitted */
u16 skb_dirtytx;
/* Configuration info for the coalescing features */
unsigned char txcoalescing;
unsigned short txcount;
unsigned short txtime;
+
+ /* Buffer descriptor pointers */
+ struct txbd8 *tx_bd_base; /* First tx buffer descriptor */
+ struct txbd8 *cur_tx; /* Next free ring entry */
+ struct txbd8 *dirty_tx; /* First buffer in line
+ to be transmitted */
+ unsigned int tx_ring_size;
+
+ /* RX Locked fields */
+ spinlock_t rxlock;
+
+ /* skb array and index */
+ struct sk_buff ** rx_skbuff;
+ u16 skb_currx;
+
+ /* RX Coalescing values */
unsigned char rxcoalescing;
unsigned short rxcount;
unsigned short rxtime;
- /* GFAR addresses */
- struct rxbd8 *rx_bd_base; /* Base addresses of Rx and Tx Buffers */
- struct txbd8 *tx_bd_base;
+ struct rxbd8 *rx_bd_base; /* First Rx buffers */
struct rxbd8 *cur_rx; /* Next free rx ring entry */
- struct txbd8 *cur_tx; /* Next free ring entry */
- struct txbd8 *dirty_tx; /* The Ring entry to be freed. */
- struct gfar __iomem *regs; /* Pointer to the GFAR memory mapped
Registers */
- u32 __iomem *hash_regs[16];
- int hash_width;
- struct net_device_stats stats; /* linux network statistics */
- struct gfar_extra_stats extra_stats;
- spinlock_t lock;
+
+ /* RX parameters */
+ unsigned int rx_ring_size;
unsigned int rx_buffer_size;
unsigned int rx_stash_size;
unsigned int rx_stash_index;
- unsigned int tx_ring_size;
- unsigned int rx_ring_size;
+
+ struct vlan_group *vlgrp;
+
+ /* Unprotected fields */
+ /* Pointer to the GFAR memory mapped Registers */
+ struct gfar __iomem *regs;
+
+ /* Hash registers and their width */
+ u32 __iomem *hash_regs[16];
+ int hash_width;
+
+ /* global parameters */
unsigned int fifo_threshold;
unsigned int fifo_starve;
unsigned int fifo_starve_off;
@@ -702,13 +721,15 @@ struct gfar_private {
extended_hash:1,
bd_stash_en:1;
unsigned short padding;
- struct vlan_group *vlgrp;
- /* Info structure initialized by board setup code */
+
unsigned int interruptTransmit;
unsigned int interruptReceive;
unsigned int interruptError;
+
+ /* info structure initialized by platform code */
struct gianfar_platform_data *einfo;
+ /* PHY stuff */
struct phy_device *phydev;
struct mii_bus *mii_bus;
int oldspeed;
@@ -716,6 +737,10 @@ struct gfar_private {
int oldlink;
uint32_t msg_enable;
+
+ /* Network Statistics */
+ struct net_device_stats stats;
+ struct gfar_extra_stats extra_stats;
};
static inline u32 gfar_read(volatile unsigned __iomem *addr)
diff --git a/drivers/net/gianfar_ethtool.c b/drivers/net/
gianfar_ethtool.c
index 5de7b2e..d69698c 100644
--- a/drivers/net/gianfar_ethtool.c
+++ b/drivers/net/gianfar_ethtool.c
@@ -455,10 +455,14 @@ static int gfar_sringparam(struct net_de
/* Halt TX and RX, and process the frames which
* have already been received */
- spin_lock_irqsave(&priv->lock, flags);
+ spin_lock_irqsave(&priv->txlock, flags);
+ spin_lock(&priv->rxlock);
+
gfar_halt(dev);
gfar_clean_rx_ring(dev, priv->rx_ring_size);
- spin_unlock_irqrestore(&priv->lock, flags);
+
+ spin_unlock(&priv->rxlock);
+ spin_unlock_irqrestore(&priv->txlock, flags);
/* Now we take down the rings to rebuild them */
stop_gfar(dev);
@@ -488,10 +492,14 @@ static int gfar_set_rx_csum(struct net_d
/* Halt TX and RX, and process the frames which
* have already been received */
- spin_lock_irqsave(&priv->lock, flags);
+ spin_lock_irqsave(&priv->txlock, flags);
+ spin_lock(&priv->rxlock);
+
gfar_halt(dev);
gfar_clean_rx_ring(dev, priv->rx_ring_size);
- spin_unlock_irqrestore(&priv->lock, flags);
+
+ spin_unlock(&priv->rxlock);
+ spin_unlock_irqrestore(&priv->txlock, flags);
/* Now we take down the rings to rebuild them */
stop_gfar(dev);
@@ -523,7 +531,7 @@ static int gfar_set_tx_csum(struct net_d
if (!(priv->einfo->device_flags & FSL_GIANFAR_DEV_HAS_CSUM))
return -EOPNOTSUPP;
- spin_lock_irqsave(&priv->lock, flags);
+ spin_lock_irqsave(&priv->txlock, flags);
gfar_halt(dev);
if (data)
@@ -532,7 +540,7 @@ static int gfar_set_tx_csum(struct net_d
dev->features &= ~NETIF_F_IP_CSUM;
gfar_start(dev);
- spin_unlock_irqrestore(&priv->lock, flags);
+ spin_unlock_irqrestore(&priv->txlock, flags);
return 0;
}
diff --git a/drivers/net/gianfar_sysfs.c b/drivers/net/gianfar_sysfs.c
index 51ef181..a6d5c43 100644
--- a/drivers/net/gianfar_sysfs.c
+++ b/drivers/net/gianfar_sysfs.c
@@ -82,7 +82,7 @@ static ssize_t gfar_set_bd_stash(struct
else
return count;
- spin_lock_irqsave(&priv->lock, flags);
+ spin_lock_irqsave(&priv->rxlock, flags);
/* Set the new stashing value */
priv->bd_stash_en = new_setting;
@@ -96,7 +96,7 @@ static ssize_t gfar_set_bd_stash(struct
gfar_write(&priv->regs->attr, temp);
- spin_unlock_irqrestore(&priv->lock, flags);
+ spin_unlock_irqrestore(&priv->rxlock, flags);
return count;
}
@@ -118,7 +118,7 @@ static ssize_t gfar_set_rx_stash_size(st
u32 temp;
unsigned long flags;
- spin_lock_irqsave(&priv->lock, flags);
+ spin_lock_irqsave(&priv->rxlock, flags);
if (length > priv->rx_buffer_size)
return count;
@@ -142,7 +142,7 @@ static ssize_t gfar_set_rx_stash_size(st
gfar_write(&priv->regs->attr, temp);
- spin_unlock_irqrestore(&priv->lock, flags);
+ spin_unlock_irqrestore(&priv->rxlock, flags);
return count;
}
@@ -166,7 +166,7 @@ static ssize_t gfar_set_rx_stash_index(s
u32 temp;
unsigned long flags;
- spin_lock_irqsave(&priv->lock, flags);
+ spin_lock_irqsave(&priv->rxlock, flags);
if (index > priv->rx_stash_size)
return count;
@@ -180,7 +180,7 @@ static ssize_t gfar_set_rx_stash_index(s
temp |= ATTRELI_EI(index);
gfar_write(&priv->regs->attreli, flags);
- spin_unlock_irqrestore(&priv->lock, flags);
+ spin_unlock_irqrestore(&priv->rxlock, flags);
return count;
}
@@ -205,7 +205,7 @@ static ssize_t gfar_set_fifo_threshold(s
if (length > GFAR_MAX_FIFO_THRESHOLD)
return count;
- spin_lock_irqsave(&priv->lock, flags);
+ spin_lock_irqsave(&priv->txlock, flags);
priv->fifo_threshold = length;
@@ -214,7 +214,7 @@ static ssize_t gfar_set_fifo_threshold(s
temp |= length;
gfar_write(&priv->regs->fifo_tx_thr, temp);
- spin_unlock_irqrestore(&priv->lock, flags);
+ spin_unlock_irqrestore(&priv->txlock, flags);
return count;
}
@@ -240,7 +240,7 @@ static ssize_t gfar_set_fifo_starve(stru
if (num > GFAR_MAX_FIFO_STARVE)
return count;
- spin_lock_irqsave(&priv->lock, flags);
+ spin_lock_irqsave(&priv->txlock, flags);
priv->fifo_starve = num;
@@ -249,7 +249,7 @@ static ssize_t gfar_set_fifo_starve(stru
temp |= num;
gfar_write(&priv->regs->fifo_tx_starve, temp);
- spin_unlock_irqrestore(&priv->lock, flags);
+ spin_unlock_irqrestore(&priv->txlock, flags);
return count;
}
@@ -274,7 +274,7 @@ static ssize_t gfar_set_fifo_starve_off(
if (num > GFAR_MAX_FIFO_STARVE_OFF)
return count;
- spin_lock_irqsave(&priv->lock, flags);
+ spin_lock_irqsave(&priv->txlock, flags);
priv->fifo_starve_off = num;
@@ -283,7 +283,7 @@ static ssize_t gfar_set_fifo_starve_off(
temp |= num;
gfar_write(&priv->regs->fifo_tx_starve_shutoff, temp);
- spin_unlock_irqrestore(&priv->lock, flags);
+ spin_unlock_irqrestore(&priv->txlock, flags);
return count;
}
--
1.2.4
-
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 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