Netdev List
 help / color / mirror / Atom feed
* [PATCH 6/12] drivers/net: Add missing unlock
From: Julia Lawall @ 2010-03-29 15:35 UTC (permalink / raw)
  To: netdev, linux-kernel, kernel-janitors

From: Julia Lawall <julia@diku.dk>

Unlock the lock before leaving the function.

A simplified version of the semantic patch that finds this problem is as
follows: (http://coccinelle.lip6.fr/)

// <smpl>
@r exists@
expression E1;
identifier f;
@@

f (...) { <+...
* spin_lock_irqsave (E1,...);
... when != E1
* return ...;
...+> }
// </smpl>

Signed-off-by: Julia Lawall <julia@diku.dk>

---
 drivers/net/sgiseeq.c               |    4 +++-
 1 files changed, 3 insertions(+), 1 deletions(-)

diff --git a/drivers/net/sgiseeq.c b/drivers/net/sgiseeq.c
index ed999d3..37f6a00 100644
--- a/drivers/net/sgiseeq.c
+++ b/drivers/net/sgiseeq.c
@@ -592,8 +592,10 @@ static int sgiseeq_start_xmit(struct sk_buff *skb, struct net_device *dev)
 	/* Setup... */
 	len = skb->len;
 	if (len < ETH_ZLEN) {
-		if (skb_padto(skb, ETH_ZLEN))
+		if (skb_padto(skb, ETH_ZLEN)) {
+			spin_unlock_irqrestore(&sp->tx_lock, flags);
 			return NETDEV_TX_OK;
+		}
 		len = ETH_ZLEN;
 	}
 

^ permalink raw reply related

* Re: [PATCH net-next-2.6] ipv6: Use __fls() instead of fls() in __ipv6_addr_diff().
From: YOSHIFUJI Hideaki @ 2010-03-29 15:49 UTC (permalink / raw)
  To: davem; +Cc: YOSHIFUJI Hideaki, netdev
In-Reply-To: <201003260801.o2Q8134t023352@94.43.138.210.xn.2iij.net>

Sorry, I withdraw this because this is incorrect;
__fls() returns fls() - 1.

--yoshfuji

(2010/03/26 17:01), YOSHIFUJI Hideaki wrote:
> Because we have ensured that the argument is non-zero,
> it is better to use __fls() and generate better code.
> 
> Signed-off-by: YOSHIFUJI Hideaki<yoshfuji@linux-ipv6.org>
> ---
>   include/net/ipv6.h |    2 +-
>   1 files changed, 1 insertions(+), 1 deletions(-)
> 
> diff --git a/include/net/ipv6.h b/include/net/ipv6.h
> index e72fb10..619ab34 100644
> --- a/include/net/ipv6.h
> +++ b/include/net/ipv6.h
> @@ -422,7 +422,7 @@ static inline int __ipv6_addr_diff(const void *token1, const void *token2, int a
>   	for (i = 0; i<  addrlen; i++) {
>   		__be32 xb = a1[i] ^ a2[i];
>   		if (xb)
> -			return i * 32 + 32 - fls(ntohl(xb));
> +			return i * 32 + 32 - __fls(ntohl(xb));
>   	}
> 
>   	/*


^ permalink raw reply

* Re: [PATCH] NETLABEL: Fix an RCU warning
From: Paul E. McKenney @ 2010-03-29 15:58 UTC (permalink / raw)
  To: Paul Moore; +Cc: Eric Dumazet, David Howells, netdev
In-Reply-To: <201003291130.10752.paul.moore@hp.com>

On Mon, Mar 29, 2010 at 11:30:10AM -0400, Paul Moore wrote:
> On Monday 29 March 2010 11:24:53 am Paul E. McKenney wrote:
> > On Thu, Mar 25, 2010 at 12:28:04PM +0100, Eric Dumazet wrote:
> > > Le jeudi 25 mars 2010 à 11:06 +0000, David Howells a écrit :
> > > > Fix an RCU warning in the netlabel code due to missing rcu read locking
> > > > around an rcu_dereference() in netlbl_unlhsh_hash() when called from
> > > > netlbl_unlhsh_netdev_handler():
> > > > 
> > > > ===================================================
> > > > [ INFO: suspicious rcu_dereference_check() usage. ]
> > > > ---------------------------------------------------
> > > > net/netlabel/netlabel_unlabeled.c:246 invoked rcu_dereference_check()
> > > > without protection!
> > > > 
> > > > other info that might help us debug this:
> > > > 
> > > > 
> > > > rcu_scheduler_active = 1, debug_locks = 0
> > > > 
> > > > 2 locks held by ip/5108:
> > > >  #0:  (rtnl_mutex){+.+.+.}, at: [<ffffffff812c4a36>]
> > > >  rtnl_lock+0x12/0x14 #1:  (netlbl_unlhsh_lock){+.+...}, at:
> > > >  [<ffffffff8134daa4>] netlbl_unlhsh_netdev_handler+0x1e/0x86
> > > > 
> > > > stack backtrace:
> > > > Pid: 5108, comm: ip Not tainted 2.6.34-rc2-cachefs #114
> > > > 
> > > > Call Trace:
> > > >  [<ffffffff8105121f>] lockdep_rcu_dereference+0xaa/0xb2
> > > >  [<ffffffff8134d781>] netlbl_unlhsh_hash+0x3e/0x50
> > > >  [<ffffffff8134d7a1>] netlbl_unlhsh_search_iface+0xe/0x84
> > > >  [<ffffffff8134daaf>] netlbl_unlhsh_netdev_handler+0x29/0x86
> > > >  [<ffffffff81048362>] notifier_call_chain+0x32/0x5e
> > > >  [<ffffffff810483fe>] raw_notifier_call_chain+0xf/0x11
> > > >  [<ffffffff812ba924>] call_netdevice_notifiers+0x16/0x18
> > > >  [<ffffffff812bac22>] __dev_notify_flags+0x37/0x5b
> > > >  [<ffffffff812bac8c>] dev_change_flags+0x46/0x52
> > > >  [<ffffffff812c41af>] do_setlink+0x250/0x3cd
> > > >  [<ffffffff812c4ee8>] rtnl_newlink+0x2b6/0x49d
> > > >  [<ffffffff812c4cdd>] ? rtnl_newlink+0xab/0x49d
> > > >  [<ffffffff812c4c17>] rtnetlink_rcv_msg+0x1b7/0x1d2
> > > >  [<ffffffff812c4a60>] ? rtnetlink_rcv_msg+0x0/0x1d2
> > > >  [<ffffffff812cc1dc>] netlink_rcv_skb+0x3e/0x8f
> > > >  [<ffffffff812c4a59>] rtnetlink_rcv+0x21/0x28
> > > >  [<ffffffff812cbefe>] netlink_unicast+0x218/0x28f
> > > >  [<ffffffff812cc76e>] netlink_sendmsg+0x26b/0x27a
> > > >  [<ffffffff812a9f1d>] sock_sendmsg+0xd4/0xf5
> > > >  [<ffffffff81096dca>] ? might_fault+0x4e/0x9e
> > > >  [<ffffffff81096dca>] ? might_fault+0x4e/0x9e
> > > >  [<ffffffff81096e13>] ? might_fault+0x97/0x9e
> > > >  [<ffffffff81096dca>] ? might_fault+0x4e/0x9e
> > > >  [<ffffffff812b4122>] ? verify_iovec+0x59/0x97
> > > >  [<ffffffff812aa1d9>] sys_sendmsg+0x209/0x273
> > > >  [<ffffffff810976dc>] ? __do_fault+0x395/0x3cd
> > > >  [<ffffffff8109928f>] ? handle_mm_fault+0x324/0x69d
> > > >  [<ffffffff81051e4e>] ? trace_hardirqs_on_caller+0x10c/0x130
> > > >  [<ffffffff81074972>] ? audit_syscall_entry+0x17d/0x1b0
> > > >  [<ffffffff81364154>] ? trace_hardirqs_on_thunk+0x3a/0x3f
> > > >  [<ffffffff81001eeb>] system_call_fastpath+0x16/0x1b
> > > > 
> > > > Signed-off-by: David Howells <dhowells@redhat.com>
> > > > ---
> > > > 
> > > >  net/netlabel/netlabel_unlabeled.c |    2 ++
> > > >  1 files changed, 2 insertions(+), 0 deletions(-)
> > > > 
> > > > diff --git a/net/netlabel/netlabel_unlabeled.c
> > > > b/net/netlabel/netlabel_unlabeled.c index 852d9d7..7ea64e4 100644
> > > > --- a/net/netlabel/netlabel_unlabeled.c
> > > > +++ b/net/netlabel/netlabel_unlabeled.c
> > > > @@ -799,6 +799,7 @@ static int netlbl_unlhsh_netdev_handler(struct
> > > > notifier_block *this,
> > > > 
> > > >  	/* XXX - should this be a check for NETDEV_DOWN or _UNREGISTER? */
> > > >  	if (event == NETDEV_DOWN) {
> > > > 
> > > > +		rcu_read_lock();
> > > > 
> > > >  		spin_lock(&netlbl_unlhsh_lock);
> > > >  		iface = netlbl_unlhsh_search_iface(dev->ifindex);
> > > >  		if (iface != NULL && iface->valid) {
> > > > 
> > > > @@ -807,6 +808,7 @@ static int netlbl_unlhsh_netdev_handler(struct
> > > > notifier_block *this,
> > > > 
> > > >  		} else
> > > >  		
> > > >  			iface = NULL;
> > > >  		
> > > >  		spin_unlock(&netlbl_unlhsh_lock);
> > > > 
> > > > +		rcu_read_unlock();
> > > > 
> > > >  	}
> > > >  	
> > > >  	if (iface != NULL)
> > > > 
> > > > --
> > > 
> > > Sorry this is not the right fix.
> > > 
> > > Fix is to change the dereference check to take into account the lock
> > > owned here.
> > 
> > So we need the rcu_dereference() in netlbl_unlhsh_search_iface()
> > to become someething like the following?
> > 
> > 	bkt_list = &rcu_dereference_check(netlbl_unlhsh,
> > 					  rcu_read_lock_held() ||
> > 					  lockdep_is_held(&netlbl_unlhsh_lock))->tbl[bkt];
> > 
> > Or is this the wrong lock?
> 
> As Eric pointed out in response to the message above, I believe the solution 
> is to simply remove the rcu_dereference() call in the netlbl_unlhsh_hash() 
> function.

It would be at the moment, but this will break once Arnd Bergmann gets
his sparse-based checks done.  With these checks, we decorate RCU-protected
pointers, and then sparse yells if you access such a pointer without the
proper rcu_dereference() invocation.

							Thanx, Paul

^ permalink raw reply

* [PATCH] r8169: offical fix for CVE-2009-4537 (overlength frame DMAs)
From: Neil Horman @ 2010-03-29 16:03 UTC (permalink / raw)
  To: linux-kernel, netdev
  Cc: michael.s.gilbert, davem, nhorman, romieu, eric.dumazet

Official patch to fix the r8169 frame length check error.

Based on this initial thread:
http://marc.info/?l=linux-netdev&m=126202972828626&w=1
This is the official patch to fix the frame length problems in the r8169
driver.  As noted in the previous thread, while this patch incurs a performance
hit on the driver, its possible to improve performance dynamically by updating
the mtu and rx_copybreak values at runtime to return performance to what it was
for those NICS which are unaffected by the ideosyncracy (if there are any).

Summary:

    A while back Eric submitted a patch for r8169 in which the proper
allocated frame size was written to RXMaxSize to prevent the NIC from dmaing too
much data.  This was done in commit fdd7b4c3302c93f6833e338903ea77245eb510b4.  A
long time prior to that however, Francois posted
126fa4b9ca5d9d7cb7d46f779ad3bd3631ca387c, which expiclitly disabled the MaxSize
setting due to the fact that the hardware behaved in odd ways when overlong
frames were received on NIC's supported by this driver.  This was mentioned in a
security conference recently:
http://events.ccc.de/congress/2009/Fahrplan//events/3596.en.html

It seems that if we can't enable frame size filtering, then, as Eric correctly
noticed, we can find ourselves DMA-ing too much data to a buffer, causing
corruption.  As a result is seems that we are forced to allocate a frame which
is ready to handle a maximally sized receive.

This obviously has performance issues with it, so to mitigate that issue, this
patch does two things:

1) Raises the copybreak value to the frame allocation size, which should force
appropriately sized packets to get allocated on rx, rather than a full new 16k
buffer.

2) This patch only disables frame filtering initially (i.e., during the NIC
open), changing the MTU results in ring buffer allocation of a size in relation
to the new mtu (along with a warning indicating that this is dangerous).

Because of item (2), individuals who can't cope with the performance hit (or can
otherwise filter frames to prevent the bug), or who have hardware they are sure
is unaffected by this issue, can manually lower the copybreak and reset the mtu
such that performance is restored easily.

Signed-off-by: Neil Horman <nhorman@redhat.com>


 r8169.c |   29 ++++++++++++++++++++++++-----
 1 file changed, 24 insertions(+), 5 deletions(-)

diff --git a/drivers/net/r8169.c b/drivers/net/r8169.c
index 964305c..1db95d4 100644
--- a/drivers/net/r8169.c
+++ b/drivers/net/r8169.c
@@ -187,7 +187,12 @@ static DEFINE_PCI_DEVICE_TABLE(rtl8169_pci_tbl) = {
 
 MODULE_DEVICE_TABLE(pci, rtl8169_pci_tbl);
 
-static int rx_copybreak = 200;
+/*
+ * we set our copybreak very high so that we don't have
+ * to allocate 16k frames all the time (see note in
+ * rtl8169_open()
+ */
+static int rx_copybreak = 16383;
 static int use_dac = -1;
 static struct {
 	u32 msg_enable;
@@ -3254,9 +3259,13 @@ static void __devexit rtl8169_remove_one(struct pci_dev *pdev)
 }
 
 static void rtl8169_set_rxbufsize(struct rtl8169_private *tp,
-				  struct net_device *dev)
+				  unsigned int mtu)
 {
-	unsigned int max_frame = dev->mtu + VLAN_ETH_HLEN + ETH_FCS_LEN;
+	unsigned int max_frame = mtu + VLAN_ETH_HLEN + ETH_FCS_LEN;
+
+	if (max_frame != 16383)
+		printk(KERN_WARNING "WARNING! Changing of MTU on this NIC"
+			"May lead to frame reception errors!\n");
 
 	tp->rx_buf_sz = (max_frame > RX_BUF_SIZE) ? max_frame : RX_BUF_SIZE;
 }
@@ -3269,7 +3278,17 @@ static int rtl8169_open(struct net_device *dev)
 
 	pm_runtime_get_sync(&pdev->dev);
 
-	rtl8169_set_rxbufsize(tp, dev);
+	/*
+	 * Note that we use a magic value here, its wierd I know
+	 * its done because, some subset of rtl8169 hardware suffers from
+	 * a problem in which frames received that are longer than
+	 * the size set in RxMaxSize register return garbage sizes
+	 * when received.  To avoid this we need to turn off filtering,
+	 * which is done by setting a value of 16383 in the RxMaxSize register
+	 * and allocating 16k frames to handle the largest possible rx value
+	 * thats what the magic math below does.
+	 */
+	rtl8169_set_rxbufsize(tp, 16383 - VLAN_ETH_HLEN - ETH_FCS_LEN);
 
 	/*
 	 * Rx and Tx desscriptors needs 256 bytes alignment.
@@ -3929,7 +3948,7 @@ static int rtl8169_change_mtu(struct net_device *dev, int new_mtu)
 
 	rtl8169_down(dev);
 
-	rtl8169_set_rxbufsize(tp, dev);
+	rtl8169_set_rxbufsize(tp, dev->mtu);
 
 	ret = rtl8169_init_ring(dev);
 	if (ret < 0)

^ permalink raw reply related

* [PATCH net-next-2.6 (TAKE 2)] ipv6: Use __fls() instead of fls() in __ipv6_addr_diff().
From: YOSHIFUJI Hideaki @ 2010-03-29 16:00 UTC (permalink / raw)
  To: davem; +Cc: yoshfuji, netdev

Because we have ensured that the argument is non-zero,
it is better to use __fls() and generate better code.

Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
---
 include/net/ipv6.h |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/include/net/ipv6.h b/include/net/ipv6.h
index e72fb10..033ddd4 100644
--- a/include/net/ipv6.h
+++ b/include/net/ipv6.h
@@ -422,7 +422,7 @@ static inline int __ipv6_addr_diff(const void *token1, const void *token2, int a
 	for (i = 0; i < addrlen; i++) {
 		__be32 xb = a1[i] ^ a2[i];
 		if (xb)
-			return i * 32 + 32 - fls(ntohl(xb));
+			return i * 32 + 31 - __fls(ntohl(xb));
 	}
 
 	/*
-- 
1.5.6.5


^ permalink raw reply related

* Re: [patch] iwlwifi: range checking issue
From: reinette chatre @ 2010-03-29 16:13 UTC (permalink / raw)
  To: Zhu, Yi
  Cc: Dan Carpenter, Intel Linux Wireless, John W. Linville,
	Kolekar, Abhijeet, Berg, Johannes, Guy, Wey-Yi W,
	linux-wireless@vger.kernel.org, netdev@vger.kernel.org,
	linux-kernel@vger.kernel.org, kernel-janitors@vger.kernel.org
In-Reply-To: <1269828072.4043.247.camel@debian>

On Sun, 2010-03-28 at 19:01 -0700, Zhu, Yi wrote:
> On Sun, 2010-03-28 at 19:55 +0800, Dan Carpenter wrote:
> > IWL_RATE_COUNT is 13 and IWL_RATE_COUNT_LEGACY is 12.
> > 
> > IWL_RATE_COUNT_LEGACY is the right one here because iwl3945_rates
> > doesn't support 60M and also that's how "rates" is defined in
> > iwlcore_init_geos() from drivers/net/wireless/iwlwifi/iwl-core.c.
> > 
> >         rates = kzalloc((sizeof(struct ieee80211_rate) * IWL_RATE_COUNT_LEGACY),
> >                         GFP_KERNEL);
> > 
> > Signed-off-by: Dan Carpenter <error27@gmail.com>
> 
> Acked-by: Zhu Yi <yi.zhu@intel.com>

Great catch. Since this is a fix for a buffer overflow ... could you
please pass it on to stable also?

Thank you

Reinette

^ permalink raw reply

* Re: behavior of recvmmsg() on blocking sockets
From: Chris Friesen @ 2010-03-29 16:18 UTC (permalink / raw)
  To: Brandon Black; +Cc: Arnaldo Carvalho de Melo, linux-kernel, netdev
In-Reply-To: <84621a61003270619p6b4fe81bi24bb1961aba77ffb@mail.gmail.com>

On 03/27/2010 07:19 AM, Brandon Black wrote:

> I've been playing with the timeout argument to recvmmsg as well now,
> and I'm struggling to see how one would ever use it correctly with the
> current implementation.

I'd probably do something like this:

prev = current time
loop forever
	cur = current time
	timeout = max_latency - (cur - prev)
	recvmmsg(timeout)
	process all received messages
	prev = cur


Basically you determine the max latency you're willing to wait for a
packet to be handled, then subtract the amount of time you spent
processing messages from that and pass it into the recvmmsg() call as
the timeout.  That way no messages will be delayed for longer than the
max latency. (Not considering scheduling delays.)

Chris

^ permalink raw reply

* Re: [PATCH net-next-2.6 2/3] bridge br_multicast: Make functions less ipv4 dependent.
From: Stephen Hemminger @ 2010-03-29 16:12 UTC (permalink / raw)
  To: YOSHIFUJI Hideaki, davem; +Cc: netdev
In-Reply-To: <201003291101.o2TB1R0W006599@94.43.138.210.xn.2iij.net>

On Mon, 29 Mar 2010 20:01:27 +0900
YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org> wrote:

> Introduce struct br_ip{} to store ip address and protocol
> and make functions more generic so that we can support
> both IPv4 and IPv6 with less pain.
> 
> Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
> ---
>  net/bridge/br_multicast.c |  181 +++++++++++++++++++++++++++++++--------------
>  net/bridge/br_private.h   |   12 +++-
>  2 files changed, 135 insertions(+), 58 deletions(-)
> 
> diff --git a/net/bridge/br_multicast.c b/net/bridge/br_multicast.c
> index 9f0c4f0..5338574 100644
> --- a/net/bridge/br_multicast.c
> +++ b/net/bridge/br_multicast.c
> @@ -27,48 +27,86 @@
>  
>  #include "br_private.h"
>  
> -static inline int br_ip_hash(struct net_bridge_mdb_htable *mdb, __be32 ip)
> +#define IP4_BRIP(ip)	(struct br_ip) {			\
> +				.u.ip4 = ip,			\
> +				.proto = htons(ETH_P_IP),	\
> +			}
> +

Too messy with macros, I will fix.


-- 

^ permalink raw reply

* Re: eth1: Detected Hardware Unit Hang
From: Allan, Bruce W @ 2010-03-29 16:41 UTC (permalink / raw)
  To: Paweł Staszewski, Linux Network Development list
  Cc: e1000-devel@lists.sourceforge.net
In-Reply-To: <4BB0C853.2080607@itcare.pl>

[adding e1000-devel]

Please provide more information:
* what NIC/LOM is this on (preferably send full output from lspci -vvv)
* what type of networking workload is running at the time the hang occurred
* a dump of the NIC/LOM statistics might also help (ethtool -S eth1)

Have you tried the latest standalone e1000e driver on e1000.sf.net?  Does it reproduce the issue?

If we cannot reproduce the hang in-house, would you be able/willing to run a debug driver to gather more information?

Thanks,
Bruce.

-----Original Message-----
From: netdev-owner@vger.kernel.org [mailto:netdev-owner@vger.kernel.org] On Behalf Of Pawel Staszewski
Sent: Monday, March 29, 2010 8:34 AM
To: Linux Network Development list
Subject: eth1: Detected Hardware Unit Hang

After update to kernel from 2.6.29.1 to 2.6.33.1 i have this info in dmesg:

0000:05:00.0: eth1: Detected Hardware Unit Hang:
   TDH <1e>
   TDT <a>
   next_to_use <a>
   next_to_clean <1d>
buffer_info[next_to_clean]:
   time_stamp <33bae15>
   next_to_watch <20>
   jiffies <33bafaf>
   next_to_watch.status <0>
MAC Status <80080783>
PHY Status <796d>
PHY 1000BASE-T Status <3800>
PHY Extended Status <3000>
PCI Status <10>
0000:05:00.0: eth1: Detected Hardware Unit Hang:
   TDH <1e>
   TDT <a>
   next_to_use <a>
   next_to_clean <1d>
buffer_info[next_to_clean]:
   time_stamp <33bae15>
   next_to_watch <20>
   jiffies <33bb1a3>
   next_to_watch.status <0>
MAC Status <80080783>
PHY Status <796d>
PHY 1000BASE-T Status <3800>
PHY Extended Status <3000>
PCI Status <10>
0000:05:00.0: eth1: Detected Hardware Unit Hang:
   TDH <1e>
   TDT <a>
   next_to_use <a>
   next_to_clean <1d>
buffer_info[next_to_clean]:
   time_stamp <33bae15>
   next_to_watch <20>
   jiffies <33bb397>
   next_to_watch.status <0>
MAC Status <80080783>
PHY Status <796d>
PHY 1000BASE-T Status <3800>
PHY Extended Status <3000>
PCI Status <10>
------------[ cut here ]------------
WARNING: at net/sched/sch_generic.c:255 dev_watchdog+0x118/0x19c()
Hardware name: X7DCT
NETDEV WATCHDOG: eth1 (e1000e): transmit queue 0 timed out
Modules linked in:
Pid: 0, comm: swapper Not tainted 2.6.33.1 #2
Call Trace:
  [<c1024e3d>] ? warn_slowpath_common+0x52/0x71
  [<c1024e49>] ? warn_slowpath_common+0x5e/0x71
  [<c1024e8e>] ? warn_slowpath_fmt+0x26/0x2a
  [<c1261f54>] ? dev_watchdog+0x118/0x19c
  [<c102135c>] ? __wake_up+0x29/0x39
  [<c10320c6>] ? insert_work+0x40/0x44
  [<c1261e3c>] ? dev_watchdog+0x0/0x19c
  [<c102cc15>] ? run_timer_softirq+0x11a/0x173
  [<c1028e5b>] ? __do_softirq+0x74/0xdf
  [<c1028ee9>] ? do_softirq+0x23/0x27
  [<c10290be>] ? irq_exit+0x26/0x58
  [<c10102d7>] ? smp_apic_timer_interrupt+0x6c/0x76
  [<c12c5f9a>] ? apic_timer_interrupt+0x2a/0x30
  [<c1007e06>] ? mwait_idle+0x49/0x4e
  [<c10017e8>] ? cpu_idle+0x41/0x5a
---[ end trace bcca9926a046332c ]---


With kernel 2.6.29.1 all was ok.
--
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

------------------------------------------------------------------------------
Download Intel&#174; Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
E1000-devel mailing list
E1000-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/e1000-devel
To learn more about Intel&#174; Ethernet, visit http://communities.intel.com/community/wired

^ permalink raw reply

* Re: Network performance - iperf
From: Rick Jones @ 2010-03-29 16:47 UTC (permalink / raw)
  To: michal.simek
  Cc: LKML, John Williams, netdev, Grant Likely, John Linn,
	Steven J. Magnani, Arnd Bergmann, akpm
In-Reply-To: <4BB09021.6020202@petalogix.com>

I don't know how to set fixed socket buffer sizes in iperf, if you were running 
netperf though I would suggest fixing the socket buffer sizes with the 
test-specific -s (affects local) and -S (affects remote) options:

netperf -t TCP_STREAM -H <remote> -l 30 -- -s 32K -S 32K -m 32K

to test the hypothesis that the autotuning of the socket buffers/window size is 
allowing the windows to grow in the larger memory cases beyond what the TLB in 
your processor is comfortable with.

Particularly if you didn't see much degredation as RAM is increased on something 
like:

netperf -t TCP_RR -H <remote> -l 30 -- -r 1

which is a simple request/response test that will never try to have more than 
one packet in flight at a time, regardless of how large the window gets.

happy benchmarking,

rick jones
http://www.netperf.org/

^ permalink raw reply

* Re: Network performance - iperf
From: Rick Jones @ 2010-03-29 16:57 UTC (permalink / raw)
  To: michal.simek
  Cc: LKML, John Williams, netdev, Grant Likely, John Linn,
	Steven J. Magnani, Arnd Bergmann, akpm
In-Reply-To: <4BB0D9A1.3090107@hp.com>

Rick Jones wrote:
> I don't know how to set fixed socket buffer sizes in iperf, if you were 
> running netperf though I would suggest fixing the socket buffer sizes 
> with the test-specific -s (affects local) and -S (affects remote) options:
> 
> netperf -t TCP_STREAM -H <remote> -l 30 -- -s 32K -S 32K -m 32K
> 
> to test the hypothesis that the autotuning of the socket buffers/window 
> size is allowing the windows to grow in the larger memory cases beyond 
> what the TLB in your processor is comfortable with.

BTW, by default, netperf will allocate a "ring" of send buffers - the number 
allocated will be one more than the socket buffer size divided by the send size 
- so in the example above, there will be two 32KB buffers allocated in netperf's 
send ring.  A similar calculation may happen on the receive side.

That can be controlled via the global (before the "--") -W option.

     -W send,recv      Set the number of send,recv buffers

So, you might make the netperf command:

netperf -t TCP_STREAM -H <remote> -l 30 -W 1,1 -- -s 32K -S 32K -m 32K

happy benchmarking,

rick jones

^ permalink raw reply

* [PATCH 0/3] add support for Janz MODULbus devices
From: Ira W. Snyder @ 2010-03-29 16:58 UTC (permalink / raw)
  To: linux-kernel-u79uwXL29TY76Z2rM5mHXA
  Cc: socketcan-core-0fE9KPoRgkgATYTw5x5z8w,
	netdev-u79uwXL29TY76Z2rM5mHXA, sameo-VuQAYsv1563Yd54FQh9/CA

This patch series adds support for the Janz CMOD-IO carrier board, as well
as the Janz VMOD-ICAN3 Intelligent CAN controller and the Janz VMOD-TTL
Digital IO controller. The CMOD-IO carrier board is a PCI to MODULbus
bridge, into which plug MODULbus daughterboards.  I only have access to two
types of daughtercards, the VMOD-ICAN3 and VMOD-TTL boards mentioned above.

The CAN driver has been tested under high loads. I am able to generate ~60%
bus utilization. With two VMOD-ICAN3 boards looped back to each other,
neither one loses any packets when only a single board is generating
packets at maximum speed. Once both boards start generating packets, one
board will sometimes loose arbitration, and cause some lost packets.

After much review from the last posting of the series, I believe that this
should be ready for mainline. Many thanks to everyone that helped!

RFCv4 -> RFCv5:
- removed some dev_dbg() statements
- change default MODULbus modules to "none"
- accept both "empty" and "" in the MODULbus modules parameter
- print an error message when no MODULbus modules are specified
- add sysfs support to get/set CAN bus termination
- fix error-active and error-passive state reporting
- add support for netlink bus-error messages
- fixed sparse warnings
- used mfd-core API for multifunction device probing

RFCv3 -> RFCv4:
- addressed many review comments
- switch to NAPI
- add TX flow control
- mark functions with __devinit and __devexit
- add sysfs readout of MODULbus number (hex switch)
- implement GPIO driver for VMOD-TTL

RFCv2 -> RFCv3:
- addressed many review comments
- correct CAN bus error handling
- use struct device to track subdevices
- use structures for register layout
- add lots of #defines for register values
- use better function prefixes

RFCv1 -> RFCv2:
- converted to a multi-driver model
- addressed many review comments
- added CAN bus error handling
- use a work function only instead of work + NAPI
- use SJA1000 bittiming calculation code

Thanks,
Ira

^ permalink raw reply

* [PATCH 1/3] mfd: add support for Janz CMOD-IO PCI MODULbus Carrier Board
From: Ira W. Snyder @ 2010-03-29 16:58 UTC (permalink / raw)
  To: linux-kernel-u79uwXL29TY76Z2rM5mHXA
  Cc: socketcan-core-0fE9KPoRgkgATYTw5x5z8w,
	netdev-u79uwXL29TY76Z2rM5mHXA, sameo-VuQAYsv1563Yd54FQh9/CA

The Janz CMOD-IO PCI MODULbus carrier board is a PCI to MODULbus bridge,
which may host many different types of MODULbus daughterboards, including
CAN and GPIO controllers.

Signed-off-by: Ira W. Snyder <iws-lulEs6mt1IksTUYHLfqkUA@public.gmane.org>
Reviewed-by: Wolfgang Grandegger <wg-5Yr1BZd7O62+XT7JhA+gdA@public.gmane.org>
---
 drivers/mfd/Kconfig       |   10 ++
 drivers/mfd/Makefile      |    1 +
 drivers/mfd/janz-cmodio.c |  302 +++++++++++++++++++++++++++++++++++++++++++++
 include/linux/mfd/janz.h  |   54 ++++++++
 include/linux/pci_ids.h   |    2 +
 5 files changed, 369 insertions(+), 0 deletions(-)
 create mode 100644 drivers/mfd/janz-cmodio.c
 create mode 100644 include/linux/mfd/janz.h

diff --git a/drivers/mfd/Kconfig b/drivers/mfd/Kconfig
index 02fcd09..41329b4 100644
--- a/drivers/mfd/Kconfig
+++ b/drivers/mfd/Kconfig
@@ -408,6 +408,16 @@ config MFD_RDC321X
 	  southbridge which provides access to GPIOs and Watchdog using the
 	  southbridge PCI device configuration space.
 
+config MFD_JANZ_CMODIO
+	tristate "Support for Janz CMOD-IO PCI MODULbus Carrier Board"
+	select MFD_CORE
+	depends on PCI
+	help
+	  This is the core driver for the Janz CMOD-IO PCI MODULbus
+	  carrier board. This device is a PCI to MODULbus bridge which may
+	  host many different types of MODULbus daughterboards, including
+	  CAN and GPIO controllers.
+
 endmenu
 
 menu "Multimedia Capabilities Port drivers"
diff --git a/drivers/mfd/Makefile b/drivers/mfd/Makefile
index f5daffe..42a35e4 100644
--- a/drivers/mfd/Makefile
+++ b/drivers/mfd/Makefile
@@ -64,3 +64,4 @@ obj-$(CONFIG_MFD_TIMBERDALE)    += timberdale.o
 obj-$(CONFIG_PMIC_ADP5520)	+= adp5520.o
 obj-$(CONFIG_LPC_SCH)		+= lpc_sch.o
 obj-$(CONFIG_MFD_RDC321X)	+= rdc321x-southbridge.o
+obj-$(CONFIG_MFD_JANZ_CMODIO)	+= janz-cmodio.o
diff --git a/drivers/mfd/janz-cmodio.c b/drivers/mfd/janz-cmodio.c
new file mode 100644
index 0000000..ac7d59b
--- /dev/null
+++ b/drivers/mfd/janz-cmodio.c
@@ -0,0 +1,302 @@
+/*
+ * Janz CMOD-IO MODULbus Carrier Board PCI Driver
+ *
+ * Copyright (c) 2010 Ira W. Snyder <iws-lulEs6mt1IksTUYHLfqkUA@public.gmane.org>
+ *
+ * Lots of inspiration and code was copied from drivers/mfd/sm501.c
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2 of the License, or (at your
+ * option) any later version.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/pci.h>
+#include <linux/interrupt.h>
+#include <linux/delay.h>
+#include <linux/platform_device.h>
+#include <linux/mfd/core.h>
+
+#include <linux/mfd/janz.h>
+
+#define DRV_NAME "janz-cmodio"
+
+/* Size of each MODULbus module in PCI BAR4 */
+#define CMODIO_MODULBUS_SIZE	0x200
+
+/* Maximum number of MODULbus modules on a CMOD-IO carrier board */
+#define CMODIO_MAX_MODULES	4
+
+/* Module Parameters */
+static unsigned int num_modules = CMODIO_MAX_MODULES;
+static unsigned char *modules[CMODIO_MAX_MODULES] = {
+	"empty", "empty", "empty", "empty",
+};
+
+module_param_array(modules, charp, &num_modules, S_IRUGO);
+MODULE_PARM_DESC(modules, "MODULbus modules attached to the carrier board");
+
+/* Unique Device Id */
+static unsigned int cmodio_id;
+
+struct cmodio_device {
+	/* Parent PCI device */
+	struct pci_dev *pdev;
+
+	/* PLX control registers */
+	struct janz_cmodio_onboard_regs __iomem *ctrl;
+
+	/* hex switch position */
+	u8 hex;
+
+	/* mfd-core API */
+	struct mfd_cell cells[CMODIO_MAX_MODULES];
+	struct resource resources[3 * CMODIO_MAX_MODULES];
+	struct janz_platform_data pdata[CMODIO_MAX_MODULES];
+};
+
+/*
+ * Subdevices using the mfd-core API
+ */
+
+static int __devinit cmodio_setup_subdevice(struct cmodio_device *priv,
+					    char *name, unsigned int devno,
+					    unsigned int modno)
+{
+	struct janz_platform_data *pdata;
+	struct mfd_cell *cell;
+	struct resource *res;
+	struct pci_dev *pci;
+
+	pci = priv->pdev;
+	cell = &priv->cells[devno];
+	res = &priv->resources[devno * 3];
+	pdata = &priv->pdata[devno];
+
+	cell->name = name;
+	cell->resources = res;
+	cell->num_resources = 3;
+
+	/* Setup the subdevice ID -- must be unique */
+	cell->id = cmodio_id++;
+
+	/* Add platform data */
+	pdata->modno = modno;
+	cell->platform_data = pdata;
+	cell->data_size = sizeof(*pdata);
+
+	/* MODULbus registers -- PCI BAR3 is big-endian MODULbus access */
+	res->flags = IORESOURCE_MEM;
+	res->parent = &pci->resource[3];
+	res->start = pci->resource[3].start + (CMODIO_MODULBUS_SIZE * modno);
+	res->end = res->start + CMODIO_MODULBUS_SIZE - 1;
+	res++;
+
+	/* PLX Control Registers -- PCI BAR4 is interrupt and other registers */
+	res->flags = IORESOURCE_MEM;
+	res->parent = &pci->resource[4];
+	res->start = pci->resource[4].start;
+	res->end = pci->resource[4].end;
+	res++;
+
+	/*
+	 * IRQ
+	 *
+	 * The start and end fields are used as an offset to the irq_base
+	 * parameter passed into the mfd_add_devices() function call. All
+	 * devices share the same IRQ.
+	 */
+	res->flags = IORESOURCE_IRQ;
+	res->parent = NULL;
+	res->start = 0;
+	res->end = 0;
+	res++;
+
+	return 0;
+}
+
+/* Probe each submodule using kernel parameters */
+static int __devinit cmodio_probe_submodules(struct cmodio_device *priv)
+{
+	struct pci_dev *pdev = priv->pdev;
+	unsigned int num_probed = 0;
+	char *name;
+	int i;
+
+	for (i = 0; i < num_modules; i++) {
+		name = modules[i];
+		if (!strcmp(name, "") || !strcmp(name, "empty"))
+			continue;
+
+		dev_dbg(&priv->pdev->dev, "MODULbus %d: name %s\n", i, name);
+		cmodio_setup_subdevice(priv, name, num_probed, i);
+		num_probed++;
+	}
+
+	/* print an error message if no modules were probed */
+	if (num_probed == 0) {
+		dev_err(&priv->pdev->dev, "no MODULbus modules specified, "
+					  "please set the ``modules'' kernel "
+					  "parameter according to your "
+					  "hardware configuration\n");
+		return -ENODEV;
+	}
+
+	return mfd_add_devices(&pdev->dev, 0, priv->cells,
+			       num_probed, NULL, pdev->irq);
+}
+
+/*
+ * SYSFS Attributes
+ */
+
+static ssize_t mbus_show(struct device *dev, struct device_attribute *attr,
+			 char *buf)
+{
+	struct cmodio_device *priv = dev_get_drvdata(dev);
+
+	return snprintf(buf, PAGE_SIZE, "%x\n", priv->hex);
+}
+
+static DEVICE_ATTR(modulbus_number, S_IRUGO, mbus_show, NULL);
+
+static struct attribute *cmodio_sysfs_attrs[] = {
+	&dev_attr_modulbus_number.attr,
+	NULL,
+};
+
+static const struct attribute_group cmodio_sysfs_attr_group = {
+	.attrs = cmodio_sysfs_attrs,
+};
+
+/*
+ * PCI Driver
+ */
+
+static int __devinit cmodio_pci_probe(struct pci_dev *dev,
+				      const struct pci_device_id *id)
+{
+	struct cmodio_device *priv;
+	int ret;
+
+	priv = kzalloc(sizeof(*priv), GFP_KERNEL);
+	if (!priv) {
+		dev_err(&dev->dev, "unable to allocate private data\n");
+		ret = -ENOMEM;
+		goto out_return;
+	}
+
+	pci_set_drvdata(dev, priv);
+	priv->pdev = dev;
+
+	/* Hardware Initialization */
+	ret = pci_enable_device(dev);
+	if (ret) {
+		dev_err(&dev->dev, "unable to enable device\n");
+		goto out_free_priv;
+	}
+
+	pci_set_master(dev);
+	ret = pci_request_regions(dev, DRV_NAME);
+	if (ret) {
+		dev_err(&dev->dev, "unable to request regions\n");
+		goto out_pci_disable_device;
+	}
+
+	/* Onboard configuration registers */
+	priv->ctrl = pci_ioremap_bar(dev, 4);
+	if (!priv->ctrl) {
+		dev_err(&dev->dev, "unable to remap onboard regs\n");
+		ret = -ENOMEM;
+		goto out_pci_release_regions;
+	}
+
+	/* Read the hex switch on the carrier board */
+	priv->hex = ioread8(&priv->ctrl->int_enable);
+
+	/* Add the MODULbus number (hex switch value) to the device's sysfs */
+	ret = sysfs_create_group(&dev->dev.kobj, &cmodio_sysfs_attr_group);
+	if (ret) {
+		dev_err(&dev->dev, "unable to create sysfs attributes\n");
+		goto out_unmap_ctrl;
+	}
+
+	/*
+	 * Disable all interrupt lines, each submodule will enable its
+	 * own interrupt line if needed
+	 */
+	iowrite8(0xf, &priv->ctrl->int_disable);
+
+	/* Register drivers for all submodules */
+	ret = cmodio_probe_submodules(priv);
+	if (ret) {
+		dev_err(&dev->dev, "unable to probe submodules\n");
+		goto out_sysfs_remove_group;
+	}
+
+	return 0;
+
+out_sysfs_remove_group:
+	sysfs_remove_group(&dev->dev.kobj, &cmodio_sysfs_attr_group);
+out_unmap_ctrl:
+	iounmap(priv->ctrl);
+out_pci_release_regions:
+	pci_release_regions(dev);
+out_pci_disable_device:
+	pci_disable_device(dev);
+out_free_priv:
+	kfree(priv);
+out_return:
+	return ret;
+}
+
+static void __devexit cmodio_pci_remove(struct pci_dev *dev)
+{
+	struct cmodio_device *priv = pci_get_drvdata(dev);
+
+	mfd_remove_devices(&dev->dev);
+	sysfs_remove_group(&dev->dev.kobj, &cmodio_sysfs_attr_group);
+	iounmap(priv->ctrl);
+	pci_release_regions(dev);
+	pci_disable_device(dev);
+	kfree(priv);
+}
+
+/* The list of devices that this module will support */
+static DEFINE_PCI_DEVICE_TABLE(cmodio_pci_ids) = {
+	{ PCI_VENDOR_ID_PLX, PCI_DEVICE_ID_PLX_9030, PCI_VENDOR_ID_JANZ, 0x0101 },
+	{ PCI_VENDOR_ID_PLX, PCI_DEVICE_ID_PLX_9050, PCI_VENDOR_ID_JANZ, 0x0100 },
+	{ 0, }
+};
+MODULE_DEVICE_TABLE(pci, cmodio_pci_ids);
+
+static struct pci_driver cmodio_pci_driver = {
+	.name     = DRV_NAME,
+	.id_table = cmodio_pci_ids,
+	.probe    = cmodio_pci_probe,
+	.remove   = __devexit_p(cmodio_pci_remove),
+};
+
+/*
+ * Module Init / Exit
+ */
+
+static int __init cmodio_init(void)
+{
+	return pci_register_driver(&cmodio_pci_driver);
+}
+
+static void __exit cmodio_exit(void)
+{
+	pci_unregister_driver(&cmodio_pci_driver);
+}
+
+MODULE_AUTHOR("Ira W. Snyder <iws-lulEs6mt1IksTUYHLfqkUA@public.gmane.org>");
+MODULE_DESCRIPTION("Janz CMOD-IO PCI MODULbus Carrier Board Driver");
+MODULE_LICENSE("GPL");
+
+module_init(cmodio_init);
+module_exit(cmodio_exit);
diff --git a/include/linux/mfd/janz.h b/include/linux/mfd/janz.h
new file mode 100644
index 0000000..e9994c4
--- /dev/null
+++ b/include/linux/mfd/janz.h
@@ -0,0 +1,54 @@
+/*
+ * Common Definitions for Janz MODULbus devices
+ *
+ * Copyright (c) 2010 Ira W. Snyder <iws-lulEs6mt1IksTUYHLfqkUA@public.gmane.org>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2 of the License, or (at your
+ * option) any later version.
+ */
+
+#ifndef JANZ_H
+#define JANZ_H
+
+struct janz_platform_data {
+	/* MODULbus Module Number */
+	unsigned int modno;
+};
+
+/* PLX bridge chip onboard registers */
+struct janz_cmodio_onboard_regs {
+	u8 unused1;
+
+	/*
+	 * Read access: interrupt status
+	 * Write access: interrupt disable
+	 */
+	u8 int_disable;
+	u8 unused2;
+
+	/*
+	 * Read access: MODULbus number (hex switch)
+	 * Write access: interrupt enable
+	 */
+	u8 int_enable;
+	u8 unused3;
+
+	/* write-only */
+	u8 reset_assert;
+	u8 unused4;
+
+	/* write-only */
+	u8 reset_deassert;
+	u8 unused5;
+
+	/* read-write access to serial EEPROM */
+	u8 eep;
+	u8 unused6;
+
+	/* write-only access to EEPROM chip select */
+	u8 enid;
+};
+
+#endif /* JANZ_H */
diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h
index 9f688d2..5b42bb2 100644
--- a/include/linux/pci_ids.h
+++ b/include/linux/pci_ids.h
@@ -2718,3 +2718,5 @@
 #define PCI_DEVICE_ID_RME_DIGI32	0x9896
 #define PCI_DEVICE_ID_RME_DIGI32_PRO	0x9897
 #define PCI_DEVICE_ID_RME_DIGI32_8	0x9898
+
+#define PCI_VENDOR_ID_JANZ		0x13c3
-- 
1.5.4.3

^ permalink raw reply related

* [PATCH 2/3] can: add support for Janz VMOD-ICAN3 Intelligent CAN module
From: Ira W. Snyder @ 2010-03-29 16:58 UTC (permalink / raw)
  To: linux-kernel-u79uwXL29TY76Z2rM5mHXA
  Cc: socketcan-core-0fE9KPoRgkgATYTw5x5z8w,
	netdev-u79uwXL29TY76Z2rM5mHXA, sameo-VuQAYsv1563Yd54FQh9/CA

The Janz VMOD-ICAN3 is a MODULbus daughterboard which fits onto any
MODULbus carrier board. It is an intelligent CAN controller with a
microcontroller and associated firmware.

Signed-off-by: Ira W. Snyder <iws-lulEs6mt1IksTUYHLfqkUA@public.gmane.org>
---
 drivers/net/can/Kconfig      |   10 +
 drivers/net/can/Makefile     |    1 +
 drivers/net/can/janz-ican3.c | 1830 ++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 1841 insertions(+), 0 deletions(-)
 create mode 100644 drivers/net/can/janz-ican3.c

diff --git a/drivers/net/can/Kconfig b/drivers/net/can/Kconfig
index 05b7517..2c5227c 100644
--- a/drivers/net/can/Kconfig
+++ b/drivers/net/can/Kconfig
@@ -63,6 +63,16 @@ config CAN_BFIN
 	  To compile this driver as a module, choose M here: the
 	  module will be called bfin_can.
 
+config CAN_JANZ_ICAN3
+	tristate "Janz VMOD-ICAN3 Intelligent CAN controller"
+	depends on CAN_DEV && MFD_JANZ_CMODIO
+	---help---
+	  Driver for Janz VMOD-ICAN3 Intelligent CAN controller module, which
+	  connects to a MODULbus carrier board.
+
+	  This driver can also be built as a module. If so, the module will be
+	  called janz-ican3.ko.
+
 source "drivers/net/can/mscan/Kconfig"
 
 source "drivers/net/can/sja1000/Kconfig"
diff --git a/drivers/net/can/Makefile b/drivers/net/can/Makefile
index 7a702f2..9047cd0 100644
--- a/drivers/net/can/Makefile
+++ b/drivers/net/can/Makefile
@@ -15,5 +15,6 @@ obj-$(CONFIG_CAN_AT91)		+= at91_can.o
 obj-$(CONFIG_CAN_TI_HECC)	+= ti_hecc.o
 obj-$(CONFIG_CAN_MCP251X)	+= mcp251x.o
 obj-$(CONFIG_CAN_BFIN)		+= bfin_can.o
+obj-$(CONFIG_CAN_JANZ_ICAN3)	+= janz-ican3.o
 
 ccflags-$(CONFIG_CAN_DEBUG_DEVICES) := -DDEBUG
diff --git a/drivers/net/can/janz-ican3.c b/drivers/net/can/janz-ican3.c
new file mode 100644
index 0000000..6e533dc
--- /dev/null
+++ b/drivers/net/can/janz-ican3.c
@@ -0,0 +1,1830 @@
+/*
+ * Janz MODULbus VMOD-ICAN3 CAN Interface Driver
+ *
+ * Copyright (c) 2010 Ira W. Snyder <iws-lulEs6mt1IksTUYHLfqkUA@public.gmane.org>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2 of the License, or (at your
+ * option) any later version.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/delay.h>
+#include <linux/platform_device.h>
+
+#include <linux/netdevice.h>
+#include <linux/can.h>
+#include <linux/can/dev.h>
+#include <linux/can/error.h>
+
+#include <linux/mfd/janz.h>
+
+/* the DPM has 64k of memory, organized into 256x 256 byte pages */
+#define DPM_NUM_PAGES		256
+#define DPM_PAGE_SIZE		256
+#define DPM_PAGE_ADDR(p)	((p) * DPM_PAGE_SIZE)
+
+/* JANZ ICAN3 "old-style" host interface queue page numbers */
+#define QUEUE_OLD_CONTROL	0
+#define QUEUE_OLD_RB0		1
+#define QUEUE_OLD_RB1		2
+#define QUEUE_OLD_WB0		3
+#define QUEUE_OLD_WB1		4
+
+/* Janz ICAN3 "old-style" host interface control registers */
+#define MSYNC_PEER		0x00		/* ICAN only */
+#define MSYNC_LOCL		0x01		/* host only */
+#define TARGET_RUNNING		0x02
+
+#define MSYNC_RB0		0x01
+#define MSYNC_RB1		0x02
+#define MSYNC_RBLW		0x04
+#define MSYNC_RB_MASK		(MSYNC_RB0 | MSYNC_RB1)
+
+#define MSYNC_WB0		0x10
+#define MSYNC_WB1		0x20
+#define MSYNC_WBLW		0x40
+#define MSYNC_WB_MASK		(MSYNC_WB0 | MSYNC_WB1)
+
+/* Janz ICAN3 "new-style" host interface queue page numbers */
+#define QUEUE_TOHOST		5
+#define QUEUE_FROMHOST_MID	6
+#define QUEUE_FROMHOST_HIGH	7
+#define QUEUE_FROMHOST_LOW	8
+
+/* The first free page in the DPM is #9 */
+#define DPM_FREE_START		9
+
+/* Janz ICAN3 "new-style" and "fast" host interface descriptor flags */
+#define DESC_VALID		0x80
+#define DESC_WRAP		0x40
+#define DESC_INTERRUPT		0x20
+#define DESC_IVALID		0x10
+#define DESC_LEN(len)		(len)
+
+/* Janz ICAN3 Firmware Messages */
+#define MSG_CONNECTI		0x02
+#define MSG_DISCONNECT		0x03
+#define MSG_IDVERS		0x04
+#define MSG_MSGLOST		0x05
+#define MSG_NEWHOSTIF		0x08
+#define MSG_INQUIRY		0x0a
+#define MSG_SETAFILMASK		0x10
+#define MSG_INITFDPMQUEUE	0x11
+#define MSG_HWCONF		0x12
+#define MSG_FMSGLOST		0x15
+#define MSG_CEVTIND		0x37
+#define MSG_CBTRREQ		0x41
+#define MSG_COFFREQ		0x42
+#define MSG_CONREQ		0x43
+#define MSG_CCONFREQ		0x47
+
+/*
+ * Janz ICAN3 CAN Inquiry Message Types
+ *
+ * NOTE: there appears to be a firmware bug here. You must send
+ * NOTE: INQUIRY_STATUS and expect to receive an INQUIRY_EXTENDED
+ * NOTE: response. The controller never responds to a message with
+ * NOTE: the INQUIRY_EXTENDED subspec :(
+ */
+#define INQUIRY_STATUS		0x00
+#define INQUIRY_TERMINATION	0x01
+#define INQUIRY_EXTENDED	0x04
+
+/* Janz ICAN3 CAN Set Acceptance Filter Mask Message Types */
+#define SETAFILMASK_REJECT	0x00
+#define SETAFILMASK_FASTIF	0x02
+
+/* Janz ICAN3 CAN Hardware Configuration Message Types */
+#define HWCONF_TERMINATE_ON	0x01
+#define HWCONF_TERMINATE_OFF	0x00
+
+/* Janz ICAN3 CAN Event Indication Message Types */
+#define CEVTIND_EI		0x01
+#define CEVTIND_DOI		0x02
+#define CEVTIND_LOST		0x04
+#define CEVTIND_FULL		0x08
+#define CEVTIND_BEI		0x10
+
+#define CEVTIND_CHIP_SJA1000	0x02
+
+#define ICAN3_BUSERR_QUOTA_MAX	255
+
+/* Janz ICAN3 CAN Frame Conversion */
+#define ICAN3_ECHO	0x10
+#define ICAN3_EFF_RTR	0x40
+#define ICAN3_SFF_RTR	0x10
+#define ICAN3_EFF	0x80
+
+#define ICAN3_CAN_TYPE_MASK	0x0f
+#define ICAN3_CAN_TYPE_SFF	0x00
+#define ICAN3_CAN_TYPE_EFF	0x01
+
+#define ICAN3_CAN_DLC_MASK	0x0f
+
+/*
+ * SJA1000 Status and Error Register Definitions
+ *
+ * Copied from drivers/net/can/sja1000/sja1000.h
+ */
+
+/* status register content */
+#define SR_BS		0x80
+#define SR_ES		0x40
+#define SR_TS		0x20
+#define SR_RS		0x10
+#define SR_TCS		0x08
+#define SR_TBS		0x04
+#define SR_DOS		0x02
+#define SR_RBS		0x01
+
+#define SR_CRIT (SR_BS|SR_ES)
+
+/* ECC register */
+#define ECC_SEG		0x1F
+#define ECC_DIR		0x20
+#define ECC_ERR		6
+#define ECC_BIT		0x00
+#define ECC_FORM	0x40
+#define ECC_STUFF	0x80
+#define ECC_MASK	0xc0
+
+/* Number of buffers for use in the "new-style" host interface */
+#define ICAN3_NEW_BUFFERS	16
+
+/* Number of buffers for use in the "fast" host interface */
+#define ICAN3_TX_BUFFERS	512
+#define ICAN3_RX_BUFFERS	1024
+
+/* SJA1000 Clock Input */
+#define ICAN3_CAN_CLOCK		8000000
+
+/* Driver Name */
+#define DRV_NAME "janz-ican3"
+
+/* DPM Control Registers -- starts at offset 0x100 in the MODULbus registers */
+struct ican3_dpm_control {
+	/* window address register */
+	u8 window_address;
+	u8 unused1;
+
+	/*
+	 * Read access: clear interrupt from microcontroller
+	 * Write access: send interrupt to microcontroller
+	 */
+	u8 interrupt;
+	u8 unused2;
+
+	/* write-only: reset all hardware on the module */
+	u8 hwreset;
+	u8 unused3;
+
+	/* write-only: generate an interrupt to the TPU */
+	u8 tpuinterrupt;
+};
+
+struct ican3_dev {
+
+	/* must be the first member */
+	struct can_priv can;
+
+	/* CAN network device */
+	struct net_device *ndev;
+	struct napi_struct napi;
+
+	/* Device for printing */
+	struct device *dev;
+
+	/* module number */
+	unsigned int num;
+
+	/* base address of registers and IRQ */
+	struct janz_cmodio_onboard_regs __iomem *ctrl;
+	struct ican3_dpm_control __iomem *dpmctrl;
+	void __iomem *dpm;
+	int irq;
+
+	/* CAN bus termination status */
+	struct completion termination_comp;
+	bool termination_enabled;
+
+	/* CAN bus error status registers */
+	struct completion buserror_comp;
+	struct can_berr_counter bec;
+
+	/* old and new style host interface */
+	unsigned int iftype;
+
+	/*
+	 * Any function which changes the current DPM page must hold this
+	 * lock while it is performing data accesses. This ensures that the
+	 * function will not be preempted and end up reading data from a
+	 * different DPM page than it expects.
+	 */
+	spinlock_t lock;
+
+	/* new host interface */
+	unsigned int rx_int;
+	unsigned int rx_num;
+	unsigned int tx_num;
+
+	/* fast host interface */
+	unsigned int fastrx_start;
+	unsigned int fastrx_int;
+	unsigned int fastrx_num;
+	unsigned int fasttx_start;
+	unsigned int fasttx_num;
+
+	/* first free DPM page */
+	unsigned int free_page;
+};
+
+struct ican3_msg {
+	u8 control;
+	u8 spec;
+	__le16 len;
+	u8 data[252];
+};
+
+struct ican3_new_desc {
+	u8 control;
+	u8 pointer;
+};
+
+struct ican3_fast_desc {
+	u8 control;
+	u8 command;
+	u8 data[14];
+};
+
+/* write to the window basic address register */
+static inline void ican3_set_page(struct ican3_dev *mod, unsigned int page)
+{
+	BUG_ON(page >= DPM_NUM_PAGES);
+	iowrite8(page, &mod->dpmctrl->window_address);
+}
+
+/*
+ * ICAN3 "old-style" host interface
+ */
+
+/*
+ * Recieve a message from the ICAN3 "old-style" firmware interface
+ *
+ * LOCKING: must hold mod->lock
+ *
+ * returns 0 on success, -ENOMEM when no message exists
+ */
+static int ican3_old_recv_msg(struct ican3_dev *mod, struct ican3_msg *msg)
+{
+	unsigned int mbox, mbox_page;
+	u8 locl, peer, xord;
+
+	/* get the MSYNC registers */
+	ican3_set_page(mod, QUEUE_OLD_CONTROL);
+	peer = ioread8(mod->dpm + MSYNC_PEER);
+	locl = ioread8(mod->dpm + MSYNC_LOCL);
+	xord = locl ^ peer;
+
+	if ((xord & MSYNC_RB_MASK) == 0x00) {
+		dev_dbg(mod->dev, "no mbox for reading\n");
+		return -ENOMEM;
+	}
+
+	/* find the first free mbox to read */
+	if ((xord & MSYNC_RB_MASK) == MSYNC_RB_MASK)
+		mbox = (xord & MSYNC_RBLW) ? MSYNC_RB0 : MSYNC_RB1;
+	else
+		mbox = (xord & MSYNC_RB0) ? MSYNC_RB0 : MSYNC_RB1;
+
+	/* copy the message */
+	mbox_page = (mbox == MSYNC_RB0) ? QUEUE_OLD_RB0 : QUEUE_OLD_RB1;
+	ican3_set_page(mod, mbox_page);
+	memcpy_fromio(msg, mod->dpm, sizeof(*msg));
+
+	/*
+	 * notify the firmware that the read buffer is available
+	 * for it to fill again
+	 */
+	locl ^= mbox;
+
+	ican3_set_page(mod, QUEUE_OLD_CONTROL);
+	iowrite8(locl, mod->dpm + MSYNC_LOCL);
+	return 0;
+}
+
+/*
+ * Send a message through the "old-style" firmware interface
+ *
+ * LOCKING: must hold mod->lock
+ *
+ * returns 0 on success, -ENOMEM when no free space exists
+ */
+static int ican3_old_send_msg(struct ican3_dev *mod, struct ican3_msg *msg)
+{
+	unsigned int mbox, mbox_page;
+	u8 locl, peer, xord;
+
+	/* get the MSYNC registers */
+	ican3_set_page(mod, QUEUE_OLD_CONTROL);
+	peer = ioread8(mod->dpm + MSYNC_PEER);
+	locl = ioread8(mod->dpm + MSYNC_LOCL);
+	xord = locl ^ peer;
+
+	if ((xord & MSYNC_WB_MASK) == MSYNC_WB_MASK) {
+		dev_err(mod->dev, "no mbox for writing\n");
+		return -ENOMEM;
+	}
+
+	/* calculate a free mbox to use */
+	mbox = (xord & MSYNC_WB0) ? MSYNC_WB1 : MSYNC_WB0;
+
+	/* copy the message to the DPM */
+	mbox_page = (mbox == MSYNC_WB0) ? QUEUE_OLD_WB0 : QUEUE_OLD_WB1;
+	ican3_set_page(mod, mbox_page);
+	memcpy_toio(mod->dpm, msg, sizeof(*msg));
+
+	locl ^= mbox;
+	if (mbox == MSYNC_WB1)
+		locl |= MSYNC_WBLW;
+
+	ican3_set_page(mod, QUEUE_OLD_CONTROL);
+	iowrite8(locl, mod->dpm + MSYNC_LOCL);
+	return 0;
+}
+
+/*
+ * ICAN3 "new-style" Host Interface Setup
+ */
+
+static void __devinit ican3_init_new_host_interface(struct ican3_dev *mod)
+{
+	struct ican3_new_desc desc;
+	unsigned long flags;
+	void __iomem *dst;
+	int i;
+
+	spin_lock_irqsave(&mod->lock, flags);
+
+	/* setup the internal datastructures for RX */
+	mod->rx_num = 0;
+	mod->rx_int = 0;
+
+	/* tohost queue descriptors are in page 5 */
+	ican3_set_page(mod, QUEUE_TOHOST);
+	dst = mod->dpm;
+
+	/* initialize the tohost (rx) queue descriptors: pages 9-24 */
+	for (i = 0; i < ICAN3_NEW_BUFFERS; i++) {
+		desc.control = DESC_INTERRUPT | DESC_LEN(1); /* I L=1 */
+		desc.pointer = mod->free_page;
+
+		/* set wrap flag on last buffer */
+		if (i == ICAN3_NEW_BUFFERS - 1)
+			desc.control |= DESC_WRAP;
+
+		memcpy_toio(dst, &desc, sizeof(desc));
+		dst += sizeof(desc);
+		mod->free_page++;
+	}
+
+	/* fromhost (tx) mid queue descriptors are in page 6 */
+	ican3_set_page(mod, QUEUE_FROMHOST_MID);
+	dst = mod->dpm;
+
+	/* setup the internal datastructures for TX */
+	mod->tx_num = 0;
+
+	/* initialize the fromhost mid queue descriptors: pages 25-40 */
+	for (i = 0; i < ICAN3_NEW_BUFFERS; i++) {
+		desc.control = DESC_VALID | DESC_LEN(1); /* V L=1 */
+		desc.pointer = mod->free_page;
+
+		/* set wrap flag on last buffer */
+		if (i == ICAN3_NEW_BUFFERS - 1)
+			desc.control |= DESC_WRAP;
+
+		memcpy_toio(dst, &desc, sizeof(desc));
+		dst += sizeof(desc);
+		mod->free_page++;
+	}
+
+	/* fromhost hi queue descriptors are in page 7 */
+	ican3_set_page(mod, QUEUE_FROMHOST_HIGH);
+	dst = mod->dpm;
+
+	/* initialize only a single buffer in the fromhost hi queue (unused) */
+	desc.control = DESC_VALID | DESC_WRAP | DESC_LEN(1); /* VW L=1 */
+	desc.pointer = mod->free_page;
+	memcpy_toio(dst, &desc, sizeof(desc));
+	mod->free_page++;
+
+	/* fromhost low queue descriptors are in page 8 */
+	ican3_set_page(mod, QUEUE_FROMHOST_LOW);
+	dst = mod->dpm;
+
+	/* initialize only a single buffer in the fromhost low queue (unused) */
+	desc.control = DESC_VALID | DESC_WRAP | DESC_LEN(1); /* VW L=1 */
+	desc.pointer = mod->free_page;
+	memcpy_toio(dst, &desc, sizeof(desc));
+	mod->free_page++;
+
+	spin_unlock_irqrestore(&mod->lock, flags);
+}
+
+/*
+ * ICAN3 Fast Host Interface Setup
+ */
+
+static void __devinit ican3_init_fast_host_interface(struct ican3_dev *mod)
+{
+	struct ican3_fast_desc desc;
+	unsigned long flags;
+	unsigned int addr;
+	void __iomem *dst;
+	int i;
+
+	spin_lock_irqsave(&mod->lock, flags);
+
+	/* save the start recv page */
+	mod->fastrx_start = mod->free_page;
+	mod->fastrx_num = 0;
+	mod->fastrx_int = 0;
+
+	/* build a single fast tohost queue descriptor */
+	memset(&desc, 0, sizeof(desc));
+	desc.control = 0x00;
+	desc.command = 1;
+
+	/* build the tohost queue descriptor ring in memory */
+	addr = 0;
+	for (i = 0; i < ICAN3_RX_BUFFERS; i++) {
+
+		/* set the wrap bit on the last buffer */
+		if (i == ICAN3_RX_BUFFERS - 1)
+			desc.control |= DESC_WRAP;
+
+		/* switch to the correct page */
+		ican3_set_page(mod, mod->free_page);
+
+		/* copy the descriptor to the DPM */
+		dst = mod->dpm + addr;
+		memcpy_toio(dst, &desc, sizeof(desc));
+		addr += sizeof(desc);
+
+		/* move to the next page if necessary */
+		if (addr >= DPM_PAGE_SIZE) {
+			addr = 0;
+			mod->free_page++;
+		}
+	}
+
+	/* make sure we page-align the next queue */
+	if (addr != 0)
+		mod->free_page++;
+
+	/* save the start xmit page */
+	mod->fasttx_start = mod->free_page;
+	mod->fasttx_num = 0;
+
+	/* build a single fast fromhost queue descriptor */
+	memset(&desc, 0, sizeof(desc));
+	desc.control = DESC_VALID;
+	desc.command = 1;
+
+	/* build the fromhost queue descriptor ring in memory */
+	addr = 0;
+	for (i = 0; i < ICAN3_TX_BUFFERS; i++) {
+
+		/* set the wrap bit on the last buffer */
+		if (i == ICAN3_TX_BUFFERS - 1)
+			desc.control |= DESC_WRAP;
+
+		/* switch to the correct page */
+		ican3_set_page(mod, mod->free_page);
+
+		/* copy the descriptor to the DPM */
+		dst = mod->dpm + addr;
+		memcpy_toio(dst, &desc, sizeof(desc));
+		addr += sizeof(desc);
+
+		/* move to the next page if necessary */
+		if (addr >= DPM_PAGE_SIZE) {
+			addr = 0;
+			mod->free_page++;
+		}
+	}
+
+	spin_unlock_irqrestore(&mod->lock, flags);
+}
+
+/*
+ * ICAN3 "new-style" Host Interface Message Helpers
+ */
+
+/*
+ * LOCKING: must hold mod->lock
+ */
+static int ican3_new_send_msg(struct ican3_dev *mod, struct ican3_msg *msg)
+{
+	struct ican3_new_desc desc;
+	void __iomem *desc_addr = mod->dpm + (mod->tx_num * sizeof(desc));
+
+	/* switch to the fromhost mid queue, and read the buffer descriptor */
+	ican3_set_page(mod, QUEUE_FROMHOST_MID);
+	memcpy_fromio(&desc, desc_addr, sizeof(desc));
+
+	if (!(desc.control & DESC_VALID)) {
+		dev_dbg(mod->dev, "%s: no free buffers\n", __func__);
+		return -ENOMEM;
+	}
+
+	/* switch to the data page, copy the data */
+	ican3_set_page(mod, desc.pointer);
+	memcpy_toio(mod->dpm, msg, sizeof(*msg));
+
+	/* switch back to the descriptor, set the valid bit, write it back */
+	ican3_set_page(mod, QUEUE_FROMHOST_MID);
+	desc.control ^= DESC_VALID;
+	memcpy_toio(desc_addr, &desc, sizeof(desc));
+
+	/* update the tx number */
+	mod->tx_num = (desc.control & DESC_WRAP) ? 0 : (mod->tx_num + 1);
+	return 0;
+}
+
+/*
+ * LOCKING: must hold mod->lock
+ */
+static int ican3_new_recv_msg(struct ican3_dev *mod, struct ican3_msg *msg)
+{
+	struct ican3_new_desc desc;
+	void __iomem *desc_addr = mod->dpm + (mod->rx_num * sizeof(desc));
+
+	/* switch to the tohost queue, and read the buffer descriptor */
+	ican3_set_page(mod, QUEUE_TOHOST);
+	memcpy_fromio(&desc, desc_addr, sizeof(desc));
+
+	if (!(desc.control & DESC_VALID)) {
+		dev_dbg(mod->dev, "%s: no buffers to recv\n", __func__);
+		return -ENOMEM;
+	}
+
+	/* switch to the data page, copy the data */
+	ican3_set_page(mod, desc.pointer);
+	memcpy_fromio(msg, mod->dpm, sizeof(*msg));
+
+	/* switch back to the descriptor, toggle the valid bit, write it back */
+	ican3_set_page(mod, QUEUE_TOHOST);
+	desc.control ^= DESC_VALID;
+	memcpy_toio(desc_addr, &desc, sizeof(desc));
+
+	/* update the rx number */
+	mod->rx_num = (desc.control & DESC_WRAP) ? 0 : (mod->rx_num + 1);
+	return 0;
+}
+
+/*
+ * Message Send / Recv Helpers
+ */
+
+static int ican3_send_msg(struct ican3_dev *mod, struct ican3_msg *msg)
+{
+	unsigned long flags;
+	int ret;
+
+	spin_lock_irqsave(&mod->lock, flags);
+
+	if (mod->iftype == 0)
+		ret = ican3_old_send_msg(mod, msg);
+	else
+		ret = ican3_new_send_msg(mod, msg);
+
+	spin_unlock_irqrestore(&mod->lock, flags);
+	return ret;
+}
+
+static int ican3_recv_msg(struct ican3_dev *mod, struct ican3_msg *msg)
+{
+	unsigned long flags;
+	int ret;
+
+	spin_lock_irqsave(&mod->lock, flags);
+
+	if (mod->iftype == 0)
+		ret = ican3_old_recv_msg(mod, msg);
+	else
+		ret = ican3_new_recv_msg(mod, msg);
+
+	spin_unlock_irqrestore(&mod->lock, flags);
+	return ret;
+}
+
+/*
+ * Quick Pre-constructed Messages
+ */
+
+static int __devinit ican3_msg_connect(struct ican3_dev *mod)
+{
+	struct ican3_msg msg;
+
+	memset(&msg, 0, sizeof(msg));
+	msg.spec = MSG_CONNECTI;
+	msg.len = cpu_to_le16(0);
+
+	return ican3_send_msg(mod, &msg);
+}
+
+static int __devexit ican3_msg_disconnect(struct ican3_dev *mod)
+{
+	struct ican3_msg msg;
+
+	memset(&msg, 0, sizeof(msg));
+	msg.spec = MSG_DISCONNECT;
+	msg.len = cpu_to_le16(0);
+
+	return ican3_send_msg(mod, &msg);
+}
+
+static int __devinit ican3_msg_newhostif(struct ican3_dev *mod)
+{
+	struct ican3_msg msg;
+	int ret;
+
+	memset(&msg, 0, sizeof(msg));
+	msg.spec = MSG_NEWHOSTIF;
+	msg.len = cpu_to_le16(0);
+
+	/* If we're not using the old interface, switching seems bogus */
+	WARN_ON(mod->iftype != 0);
+
+	ret = ican3_send_msg(mod, &msg);
+	if (ret)
+		return ret;
+
+	/* mark the module as using the new host interface */
+	mod->iftype = 1;
+	return 0;
+}
+
+static int __devinit ican3_msg_fasthostif(struct ican3_dev *mod)
+{
+	struct ican3_msg msg;
+	unsigned int addr;
+
+	memset(&msg, 0, sizeof(msg));
+	msg.spec = MSG_INITFDPMQUEUE;
+	msg.len = cpu_to_le16(8);
+
+	/* write the tohost queue start address */
+	addr = DPM_PAGE_ADDR(mod->fastrx_start);
+	msg.data[0] = addr & 0xff;
+	msg.data[1] = (addr >> 8) & 0xff;
+	msg.data[2] = (addr >> 16) & 0xff;
+	msg.data[3] = (addr >> 24) & 0xff;
+
+	/* write the fromhost queue start address */
+	addr = DPM_PAGE_ADDR(mod->fasttx_start);
+	msg.data[4] = addr & 0xff;
+	msg.data[5] = (addr >> 8) & 0xff;
+	msg.data[6] = (addr >> 16) & 0xff;
+	msg.data[7] = (addr >> 24) & 0xff;
+
+	/* If we're not using the new interface yet, we cannot do this */
+	WARN_ON(mod->iftype != 1);
+
+	return ican3_send_msg(mod, &msg);
+}
+
+/*
+ * Setup the CAN filter to either accept or reject all
+ * messages from the CAN bus.
+ */
+static int __devinit ican3_set_id_filter(struct ican3_dev *mod, bool accept)
+{
+	struct ican3_msg msg;
+	int ret;
+
+	/* Standard Frame Format */
+	memset(&msg, 0, sizeof(msg));
+	msg.spec = MSG_SETAFILMASK;
+	msg.len = cpu_to_le16(5);
+	msg.data[0] = 0x00; /* IDLo LSB */
+	msg.data[1] = 0x00; /* IDLo MSB */
+	msg.data[2] = 0xff; /* IDHi LSB */
+	msg.data[3] = 0x07; /* IDHi MSB */
+
+	/* accept all frames for fast host if, or reject all frames */
+	msg.data[4] = accept ? SETAFILMASK_FASTIF : SETAFILMASK_REJECT;
+
+	ret = ican3_send_msg(mod, &msg);
+	if (ret)
+		return ret;
+
+	/* Extended Frame Format */
+	memset(&msg, 0, sizeof(msg));
+	msg.spec = MSG_SETAFILMASK;
+	msg.len = cpu_to_le16(13);
+	msg.data[0] = 0;    /* MUX = 0 */
+	msg.data[1] = 0x00; /* IDLo LSB */
+	msg.data[2] = 0x00;
+	msg.data[3] = 0x00;
+	msg.data[4] = 0x20; /* IDLo MSB */
+	msg.data[5] = 0xff; /* IDHi LSB */
+	msg.data[6] = 0xff;
+	msg.data[7] = 0xff;
+	msg.data[8] = 0x3f; /* IDHi MSB */
+
+	/* accept all frames for fast host if, or reject all frames */
+	msg.data[9] = accept ? SETAFILMASK_FASTIF : SETAFILMASK_REJECT;
+
+	return ican3_send_msg(mod, &msg);
+}
+
+/*
+ * Bring the CAN bus online or offline
+ */
+static int ican3_set_bus_state(struct ican3_dev *mod, bool on)
+{
+	struct ican3_msg msg;
+
+	memset(&msg, 0, sizeof(msg));
+	msg.spec = on ? MSG_CONREQ : MSG_COFFREQ;
+	msg.len = cpu_to_le16(0);
+
+	return ican3_send_msg(mod, &msg);
+}
+
+static int ican3_set_termination(struct ican3_dev *mod, bool on)
+{
+	struct ican3_msg msg;
+
+	memset(&msg, 0, sizeof(msg));
+	msg.spec = MSG_HWCONF;
+	msg.len = cpu_to_le16(2);
+	msg.data[0] = 0x00;
+	msg.data[1] = on ? HWCONF_TERMINATE_ON : HWCONF_TERMINATE_OFF;
+
+	return ican3_send_msg(mod, &msg);
+}
+
+static int ican3_send_inquiry(struct ican3_dev *mod, u8 subspec)
+{
+	struct ican3_msg msg;
+
+	memset(&msg, 0, sizeof(msg));
+	msg.spec = MSG_INQUIRY;
+	msg.len = cpu_to_le16(2);
+	msg.data[0] = subspec;
+	msg.data[1] = 0x00;
+
+	return ican3_send_msg(mod, &msg);
+}
+
+static int ican3_set_buserror(struct ican3_dev *mod, u8 quota)
+{
+	struct ican3_msg msg;
+
+	memset(&msg, 0, sizeof(msg));
+	msg.spec = MSG_CCONFREQ;
+	msg.len = cpu_to_le16(2);
+	msg.data[0] = 0x00;
+	msg.data[1] = quota;
+
+	return ican3_send_msg(mod, &msg);
+}
+
+/*
+ * ICAN3 to Linux CAN Frame Conversion
+ */
+
+static void ican3_to_can_frame(struct ican3_dev *mod,
+			       struct ican3_fast_desc *desc,
+			       struct can_frame *cf)
+{
+	if ((desc->command & ICAN3_CAN_TYPE_MASK) == ICAN3_CAN_TYPE_SFF) {
+		if (desc->data[1] & ICAN3_SFF_RTR)
+			cf->can_id |= CAN_RTR_FLAG;
+
+		cf->can_id |= desc->data[0] << 3;
+		cf->can_id |= (desc->data[1] & 0xe0) >> 5;
+		cf->can_dlc = desc->data[1] & ICAN3_CAN_DLC_MASK;
+		memcpy(cf->data, &desc->data[2], sizeof(cf->data));
+	} else {
+		cf->can_dlc = desc->data[0] & ICAN3_CAN_DLC_MASK;
+		if (desc->data[0] & ICAN3_EFF_RTR)
+			cf->can_id |= CAN_RTR_FLAG;
+
+		if (desc->data[0] & ICAN3_EFF) {
+			cf->can_id |= CAN_EFF_FLAG;
+			cf->can_id |= desc->data[2] << 21; /* 28-21 */
+			cf->can_id |= desc->data[3] << 13; /* 20-13 */
+			cf->can_id |= desc->data[4] << 5;  /* 12-5  */
+			cf->can_id |= (desc->data[5] & 0xf8) >> 3;
+		} else {
+			cf->can_id |= desc->data[2] << 3;  /* 10-3  */
+			cf->can_id |= desc->data[3] >> 5;  /* 2-0   */
+		}
+
+		memcpy(cf->data, &desc->data[6], sizeof(cf->data));
+	}
+}
+
+static void can_frame_to_ican3(struct ican3_dev *mod,
+			       struct can_frame *cf,
+			       struct ican3_fast_desc *desc)
+{
+	/* clear out any stale data in the descriptor */
+	memset(desc->data, 0, sizeof(desc->data));
+
+	/* we always use the extended format, with the ECHO flag set */
+	desc->command = ICAN3_CAN_TYPE_EFF;
+	desc->data[0] |= cf->can_dlc;
+	desc->data[1] |= ICAN3_ECHO;
+
+	if (cf->can_id & CAN_RTR_FLAG)
+		desc->data[0] |= ICAN3_EFF_RTR;
+
+	/* pack the id into the correct places */
+	if (cf->can_id & CAN_EFF_FLAG) {
+		desc->data[0] |= ICAN3_EFF;
+		desc->data[2] = (cf->can_id & 0x1fe00000) >> 21; /* 28-21 */
+		desc->data[3] = (cf->can_id & 0x001fe000) >> 13; /* 20-13 */
+		desc->data[4] = (cf->can_id & 0x00001fe0) >> 5;  /* 12-5  */
+		desc->data[5] = (cf->can_id & 0x0000001f) << 3;  /* 4-0   */
+	} else {
+		desc->data[2] = (cf->can_id & 0x7F8) >> 3; /* bits 10-3 */
+		desc->data[3] = (cf->can_id & 0x007) << 5; /* bits 2-0  */
+	}
+
+	/* copy the data bits into the descriptor */
+	memcpy(&desc->data[6], cf->data, sizeof(cf->data));
+}
+
+/*
+ * Interrupt Handling
+ */
+
+/*
+ * Handle an ID + Version message response from the firmware. We never generate
+ * this message in production code, but it is very useful when debugging to be
+ * able to display this message.
+ */
+static void ican3_handle_idvers(struct ican3_dev *mod, struct ican3_msg *msg)
+{
+	dev_dbg(mod->dev, "IDVERS response: %s\n", msg->data);
+}
+
+static void ican3_handle_msglost(struct ican3_dev *mod, struct ican3_msg *msg)
+{
+	struct net_device *dev = mod->ndev;
+	struct net_device_stats *stats = &dev->stats;
+	struct can_frame *cf;
+	struct sk_buff *skb;
+
+	/*
+	 * Report that communication messages with the microcontroller firmware
+	 * are being lost. These are never CAN frames, so we do not generate an
+	 * error frame for userspace
+	 */
+	if (msg->spec == MSG_MSGLOST) {
+		dev_err(mod->dev, "lost %d control messages\n", msg->data[0]);
+		return;
+	}
+
+	/*
+	 * Oops, this indicates that we have lost messages in the fast queue,
+	 * which are exclusively CAN messages. Our driver isn't reading CAN
+	 * frames fast enough.
+	 *
+	 * We'll pretend that the SJA1000 told us that it ran out of buffer
+	 * space, because there is not a better message for this.
+	 */
+	skb = alloc_can_err_skb(dev, &cf);
+	if (skb) {
+		cf->can_id |= CAN_ERR_CRTL;
+		cf->data[1] = CAN_ERR_CRTL_RX_OVERFLOW;
+		stats->rx_errors++;
+		stats->rx_bytes += cf->can_dlc;
+		netif_rx(skb);
+	}
+}
+
+/*
+ * Handle CAN Event Indication Messages from the firmware
+ *
+ * The ICAN3 firmware provides the values of some SJA1000 registers when it
+ * generates this message. The code below is largely copied from the
+ * drivers/net/can/sja1000/sja1000.c file, and adapted as necessary
+ */
+static int ican3_handle_cevtind(struct ican3_dev *mod, struct ican3_msg *msg)
+{
+	struct net_device *dev = mod->ndev;
+	struct net_device_stats *stats = &dev->stats;
+	enum can_state state = mod->can.state;
+	u8 status, isrc, rxerr, txerr;
+	struct can_frame *cf;
+	struct sk_buff *skb;
+
+	/* we can only handle the SJA1000 part */
+	if (msg->data[1] != CEVTIND_CHIP_SJA1000) {
+		dev_err(mod->dev, "unable to handle errors on non-SJA1000\n");
+		return -ENODEV;
+	}
+
+	/* check the message length for sanity */
+	if (le16_to_cpu(msg->len) < 6) {
+		dev_err(mod->dev, "error message too short\n");
+		return -EINVAL;
+	}
+
+	skb = alloc_can_err_skb(dev, &cf);
+	if (skb == NULL)
+		return -ENOMEM;
+
+	isrc = msg->data[0];
+	status = msg->data[3];
+	rxerr = msg->data[4];
+	txerr = msg->data[5];
+
+	/* data overrun interrupt */
+	if (isrc == CEVTIND_DOI || isrc == CEVTIND_LOST) {
+		dev_dbg(mod->dev, "data overrun interrupt\n");
+		cf->can_id |= CAN_ERR_CRTL;
+		cf->data[1] = CAN_ERR_CRTL_RX_OVERFLOW;
+		stats->rx_over_errors++;
+		stats->rx_errors++;
+	}
+
+	/* error warning + passive interrupt */
+	if (isrc == CEVTIND_EI) {
+		dev_dbg(mod->dev, "error warning + passive interrupt\n");
+		if (status & SR_BS) {
+			state = CAN_STATE_BUS_OFF;
+			cf->can_id |= CAN_ERR_BUSOFF;
+			can_bus_off(dev);
+		} else if (status & SR_ES) {
+			if (rxerr >= 128 || txerr >= 128)
+				state = CAN_STATE_ERROR_PASSIVE;
+			else
+				state = CAN_STATE_ERROR_WARNING;
+		} else {
+			state = CAN_STATE_ERROR_ACTIVE;
+		}
+	}
+
+	/* bus error interrupt */
+	if (isrc == CEVTIND_BEI) {
+		u8 ecc = msg->data[2];
+
+		dev_dbg(mod->dev, "bus error interrupt\n");
+		mod->can.can_stats.bus_error++;
+		stats->rx_errors++;
+		cf->can_id |= CAN_ERR_PROT | CAN_ERR_BUSERROR;
+
+		switch (ecc & ECC_MASK) {
+		case ECC_BIT:
+			cf->data[2] |= CAN_ERR_PROT_BIT;
+			break;
+		case ECC_FORM:
+			cf->data[2] |= CAN_ERR_PROT_FORM;
+			break;
+		case ECC_STUFF:
+			cf->data[2] |= CAN_ERR_PROT_STUFF;
+			break;
+		default:
+			cf->data[2] |= CAN_ERR_PROT_UNSPEC;
+			cf->data[3] = ecc & ECC_SEG;
+			break;
+		}
+
+		if ((ecc & ECC_DIR) == 0)
+			cf->data[2] |= CAN_ERR_PROT_TX;
+
+		cf->data[6] = txerr;
+		cf->data[7] = rxerr;
+	}
+
+	if (state != mod->can.state && (state == CAN_STATE_ERROR_WARNING ||
+					state == CAN_STATE_ERROR_PASSIVE)) {
+		cf->can_id |= CAN_ERR_CRTL;
+		if (state == CAN_STATE_ERROR_WARNING) {
+			mod->can.can_stats.error_warning++;
+			cf->data[1] = (txerr > rxerr) ?
+				CAN_ERR_CRTL_TX_WARNING :
+				CAN_ERR_CRTL_RX_WARNING;
+		} else {
+			mod->can.can_stats.error_passive++;
+			cf->data[1] = (txerr > rxerr) ?
+				CAN_ERR_CRTL_TX_PASSIVE :
+				CAN_ERR_CRTL_RX_PASSIVE;
+		}
+
+		cf->data[6] = txerr;
+		cf->data[7] = rxerr;
+	}
+
+	mod->can.state = state;
+	stats->rx_errors++;
+	stats->rx_bytes += cf->can_dlc;
+	netif_rx(skb);
+	return 0;
+}
+
+static void ican3_handle_inquiry(struct ican3_dev *mod, struct ican3_msg *msg)
+{
+	switch (msg->data[0]) {
+	case INQUIRY_STATUS:
+	case INQUIRY_EXTENDED:
+		mod->bec.rxerr = msg->data[5];
+		mod->bec.txerr = msg->data[6];
+		complete(&mod->buserror_comp);
+		break;
+	case INQUIRY_TERMINATION:
+		mod->termination_enabled = msg->data[6] & HWCONF_TERMINATE_ON;
+		complete(&mod->termination_comp);
+		break;
+	default:
+		dev_err(mod->dev, "recieved an unknown inquiry response\n");
+		break;
+	}
+}
+
+static void ican3_handle_unknown_message(struct ican3_dev *mod,
+					struct ican3_msg *msg)
+{
+	dev_warn(mod->dev, "recieved unknown message: spec 0x%.2x length %d\n",
+			   msg->spec, le16_to_cpu(msg->len));
+}
+
+/*
+ * Handle a control message from the firmware
+ */
+static void ican3_handle_message(struct ican3_dev *mod, struct ican3_msg *msg)
+{
+	dev_dbg(mod->dev, "%s: modno %d spec 0x%.2x len %d bytes\n", __func__,
+			   mod->num, msg->spec, le16_to_cpu(msg->len));
+
+	switch (msg->spec) {
+	case MSG_IDVERS:
+		ican3_handle_idvers(mod, msg);
+		break;
+	case MSG_MSGLOST:
+	case MSG_FMSGLOST:
+		ican3_handle_msglost(mod, msg);
+		break;
+	case MSG_CEVTIND:
+		ican3_handle_cevtind(mod, msg);
+		break;
+	case MSG_INQUIRY:
+		ican3_handle_inquiry(mod, msg);
+		break;
+	default:
+		ican3_handle_unknown_message(mod, msg);
+		break;
+	}
+}
+
+/*
+ * Check that there is room in the TX ring to transmit another skb
+ *
+ * LOCKING: must hold mod->lock
+ */
+static bool ican3_txok(struct ican3_dev *mod)
+{
+	struct ican3_fast_desc __iomem *desc;
+	u8 control;
+
+	/* copy the control bits of the descriptor */
+	ican3_set_page(mod, mod->fasttx_start + (mod->fasttx_num / 16));
+	desc = mod->dpm + ((mod->fasttx_num % 16) * sizeof(*desc));
+	control = ioread8(&desc->control);
+
+	/* if the control bits are not valid, then we have no more space */
+	if (!(control & DESC_VALID))
+		return false;
+
+	return true;
+}
+
+/*
+ * Recieve one CAN frame from the hardware
+ *
+ * This works like the core of a NAPI function, but is intended to be called
+ * from workqueue context instead. This driver already needs a workqueue to
+ * process control messages, so we use the workqueue instead of using NAPI.
+ * This was done to simplify locking.
+ *
+ * CONTEXT: must be called from user context
+ */
+static int ican3_recv_skb(struct ican3_dev *mod)
+{
+	struct net_device *ndev = mod->ndev;
+	struct net_device_stats *stats = &ndev->stats;
+	struct ican3_fast_desc desc;
+	void __iomem *desc_addr;
+	struct can_frame *cf;
+	struct sk_buff *skb;
+	unsigned long flags;
+
+	spin_lock_irqsave(&mod->lock, flags);
+
+	/* copy the whole descriptor */
+	ican3_set_page(mod, mod->fastrx_start + (mod->fastrx_num / 16));
+	desc_addr = mod->dpm + ((mod->fastrx_num % 16) * sizeof(desc));
+	memcpy_fromio(&desc, desc_addr, sizeof(desc));
+
+	spin_unlock_irqrestore(&mod->lock, flags);
+
+	/* check that we actually have a CAN frame */
+	if (!(desc.control & DESC_VALID))
+		return -ENOBUFS;
+
+	/* allocate an skb */
+	skb = alloc_can_skb(ndev, &cf);
+	if (unlikely(skb == NULL)) {
+		stats->rx_dropped++;
+		goto err_noalloc;
+	}
+
+	/* convert the ICAN3 frame into Linux CAN format */
+	ican3_to_can_frame(mod, &desc, cf);
+
+	/* receive the skb, update statistics */
+	netif_receive_skb(skb);
+	stats->rx_packets++;
+	stats->rx_bytes += cf->can_dlc;
+
+err_noalloc:
+	/* toggle the valid bit and return the descriptor to the ring */
+	desc.control ^= DESC_VALID;
+
+	spin_lock_irqsave(&mod->lock, flags);
+
+	ican3_set_page(mod, mod->fastrx_start + (mod->fastrx_num / 16));
+	memcpy_toio(desc_addr, &desc, 1);
+
+	/* update the next buffer pointer */
+	mod->fastrx_num = (desc.control & DESC_WRAP) ? 0
+						     : (mod->fastrx_num + 1);
+
+	/* there are still more buffers to process */
+	spin_unlock_irqrestore(&mod->lock, flags);
+	return 0;
+}
+
+static int ican3_napi(struct napi_struct *napi, int budget)
+{
+	struct ican3_dev *mod = container_of(napi, struct ican3_dev, napi);
+	struct ican3_msg msg;
+	unsigned long flags;
+	int received = 0;
+	int ret;
+
+	/* process all communication messages */
+	while (true) {
+		ret = ican3_recv_msg(mod, &msg);
+		if (ret)
+			break;
+
+		ican3_handle_message(mod, &msg);
+	}
+
+	/* process all CAN frames from the fast interface */
+	while (received < budget) {
+		ret = ican3_recv_skb(mod);
+		if (ret)
+			break;
+
+		received++;
+	}
+
+	/* We have processed all packets that the adapter had, but it
+	 * was less than our budget, stop polling */
+	if (received < budget)
+		napi_complete(napi);
+
+	spin_lock_irqsave(&mod->lock, flags);
+
+	/* Wake up the transmit queue if necessary */
+	if (netif_queue_stopped(mod->ndev) && ican3_txok(mod))
+		netif_wake_queue(mod->ndev);
+
+	spin_unlock_irqrestore(&mod->lock, flags);
+
+	/* re-enable interrupt generation */
+	iowrite8(1 << mod->num, &mod->ctrl->int_enable);
+	return received;
+}
+
+static irqreturn_t ican3_irq(int irq, void *dev_id)
+{
+	struct ican3_dev *mod = dev_id;
+	u8 stat;
+
+	/*
+	 * The interrupt status register on this device reports interrupts
+	 * as zeroes instead of using ones like most other devices
+	 */
+	stat = ioread8(&mod->ctrl->int_disable) & (1 << mod->num);
+	if (stat == (1 << mod->num))
+		return IRQ_NONE;
+
+	/* clear the MODULbus interrupt from the microcontroller */
+	ioread8(&mod->dpmctrl->interrupt);
+
+	/* disable interrupt generation, schedule the NAPI poller */
+	iowrite8(1 << mod->num, &mod->ctrl->int_disable);
+	napi_schedule(&mod->napi);
+	return IRQ_HANDLED;
+}
+
+/*
+ * Firmware reset, startup, and shutdown
+ */
+
+/*
+ * Reset an ICAN module to its power-on state
+ *
+ * CONTEXT: no network device registered
+ * LOCKING: work function disabled
+ */
+static int ican3_reset_module(struct ican3_dev *mod)
+{
+	u8 val = 1 << mod->num;
+	unsigned long start;
+	u8 runold, runnew;
+
+	/* disable interrupts so no more work is scheduled */
+	iowrite8(1 << mod->num, &mod->ctrl->int_disable);
+
+	/* flush any pending work */
+	flush_scheduled_work();
+
+	/* the first unallocated page in the DPM is #9 */
+	mod->free_page = DPM_FREE_START;
+
+	ican3_set_page(mod, QUEUE_OLD_CONTROL);
+	runold = ioread8(mod->dpm + TARGET_RUNNING);
+
+	/* reset the module */
+	iowrite8(val, &mod->ctrl->reset_assert);
+	iowrite8(val, &mod->ctrl->reset_deassert);
+
+	/* wait until the module has finished resetting and is running */
+	start = jiffies;
+	do {
+		ican3_set_page(mod, QUEUE_OLD_CONTROL);
+		runnew = ioread8(mod->dpm + TARGET_RUNNING);
+		if (runnew == (runold ^ 0xff))
+			return 0;
+
+		msleep(10);
+	} while (time_before(jiffies, start + HZ / 4));
+
+	dev_err(mod->dev, "failed to reset CAN module\n");
+	return -ETIMEDOUT;
+}
+
+static void __devexit ican3_shutdown_module(struct ican3_dev *mod)
+{
+	ican3_msg_disconnect(mod);
+	ican3_reset_module(mod);
+}
+
+/*
+ * Startup an ICAN module, bringing it into fast mode
+ */
+static int __devinit ican3_startup_module(struct ican3_dev *mod)
+{
+	int ret;
+
+	ret = ican3_reset_module(mod);
+	if (ret) {
+		dev_err(mod->dev, "unable to reset module\n");
+		return ret;
+	}
+
+	/* re-enable interrupts so we can send messages */
+	iowrite8(1 << mod->num, &mod->ctrl->int_enable);
+
+	ret = ican3_msg_connect(mod);
+	if (ret) {
+		dev_err(mod->dev, "unable to connect to module\n");
+		return ret;
+	}
+
+	ican3_init_new_host_interface(mod);
+	ret = ican3_msg_newhostif(mod);
+	if (ret) {
+		dev_err(mod->dev, "unable to switch to new-style interface\n");
+		return ret;
+	}
+
+	/* default to "termination on" */
+	ret = ican3_set_termination(mod, true);
+	if (ret) {
+		dev_err(mod->dev, "unable to enable termination\n");
+		return ret;
+	}
+
+	/* default to "bus errors enabled" */
+	ret = ican3_set_buserror(mod, ICAN3_BUSERR_QUOTA_MAX);
+	if (ret) {
+		dev_err(mod->dev, "unable to set bus-error\n");
+		return ret;
+	}
+
+	ican3_init_fast_host_interface(mod);
+	ret = ican3_msg_fasthostif(mod);
+	if (ret) {
+		dev_err(mod->dev, "unable to switch to fast host interface\n");
+		return ret;
+	}
+
+	ret = ican3_set_id_filter(mod, true);
+	if (ret) {
+		dev_err(mod->dev, "unable to set acceptance filter\n");
+		return ret;
+	}
+
+	return 0;
+}
+
+/*
+ * CAN Network Device
+ */
+
+static int ican3_open(struct net_device *ndev)
+{
+	struct ican3_dev *mod = netdev_priv(ndev);
+	u8 quota;
+	int ret;
+
+	/* open the CAN layer */
+	ret = open_candev(ndev);
+	if (ret) {
+		dev_err(mod->dev, "unable to start CAN layer\n");
+		return ret;
+	}
+
+	/* set the bus error generation state appropriately */
+	if (mod->can.ctrlmode & CAN_CTRLMODE_BERR_REPORTING)
+		quota = ICAN3_BUSERR_QUOTA_MAX;
+	else
+		quota = 0;
+
+	ret = ican3_set_buserror(mod, quota);
+	if (ret) {
+		dev_err(mod->dev, "unable to set bus-error\n");
+		close_candev(ndev);
+		return ret;
+	}
+
+	/* bring the bus online */
+	ret = ican3_set_bus_state(mod, true);
+	if (ret) {
+		dev_err(mod->dev, "unable to set bus-on\n");
+		close_candev(ndev);
+		return ret;
+	}
+
+	/* start up the network device */
+	mod->can.state = CAN_STATE_ERROR_ACTIVE;
+	netif_start_queue(ndev);
+
+	return 0;
+}
+
+static int ican3_stop(struct net_device *ndev)
+{
+	struct ican3_dev *mod = netdev_priv(ndev);
+	int ret;
+
+	/* stop the network device xmit routine */
+	netif_stop_queue(ndev);
+	mod->can.state = CAN_STATE_STOPPED;
+
+	/* bring the bus offline, stop receiving packets */
+	ret = ican3_set_bus_state(mod, false);
+	if (ret) {
+		dev_err(mod->dev, "unable to set bus-off\n");
+		return ret;
+	}
+
+	/* close the CAN layer */
+	close_candev(ndev);
+	return 0;
+}
+
+static int ican3_xmit(struct sk_buff *skb, struct net_device *ndev)
+{
+	struct ican3_dev *mod = netdev_priv(ndev);
+	struct net_device_stats *stats = &ndev->stats;
+	struct can_frame *cf = (struct can_frame *)skb->data;
+	struct ican3_fast_desc desc;
+	void __iomem *desc_addr;
+	unsigned long flags;
+
+	spin_lock_irqsave(&mod->lock, flags);
+
+	/* check that we can actually transmit */
+	if (!ican3_txok(mod)) {
+		dev_err(mod->dev, "no free descriptors, stopping queue\n");
+		netif_stop_queue(ndev);
+		spin_unlock_irqrestore(&mod->lock, flags);
+		return NETDEV_TX_BUSY;
+	}
+
+	/* copy the control bits of the descriptor */
+	ican3_set_page(mod, mod->fasttx_start + (mod->fasttx_num / 16));
+	desc_addr = mod->dpm + ((mod->fasttx_num % 16) * sizeof(desc));
+	memset(&desc, 0, sizeof(desc));
+	memcpy_fromio(&desc, desc_addr, 1);
+
+	/* convert the Linux CAN frame into ICAN3 format */
+	can_frame_to_ican3(mod, cf, &desc);
+
+	/*
+	 * the programming manual says that you must set the IVALID bit, then
+	 * interrupt, then set the valid bit. Quite weird, but it seems to be
+	 * required for this to work
+	 */
+	desc.control |= DESC_IVALID;
+	memcpy_toio(desc_addr, &desc, sizeof(desc));
+
+	/* generate a MODULbus interrupt to the microcontroller */
+	iowrite8(0x01, &mod->dpmctrl->interrupt);
+
+	desc.control ^= DESC_VALID;
+	memcpy_toio(desc_addr, &desc, sizeof(desc));
+
+	/* update the next buffer pointer */
+	mod->fasttx_num = (desc.control & DESC_WRAP) ? 0
+						     : (mod->fasttx_num + 1);
+
+	/* update statistics */
+	stats->tx_packets++;
+	stats->tx_bytes += cf->can_dlc;
+	kfree_skb(skb);
+
+	/*
+	 * This hardware doesn't have TX-done notifications, so we'll try and
+	 * emulate it the best we can using ECHO skbs. Get the next TX
+	 * descriptor, and see if we have room to send. If not, stop the queue.
+	 * It will be woken when the ECHO skb for the current packet is recv'd.
+	 */
+
+	/* copy the control bits of the descriptor */
+	if (!ican3_txok(mod))
+		netif_stop_queue(ndev);
+
+	spin_unlock_irqrestore(&mod->lock, flags);
+	return NETDEV_TX_OK;
+}
+
+static const struct net_device_ops ican3_netdev_ops = {
+	.ndo_open	= ican3_open,
+	.ndo_stop	= ican3_stop,
+	.ndo_start_xmit	= ican3_xmit,
+};
+
+/*
+ * Low-level CAN Device
+ */
+
+/* This structure was stolen from drivers/net/can/sja1000/sja1000.c */
+static struct can_bittiming_const ican3_bittiming_const = {
+	.name = DRV_NAME,
+	.tseg1_min = 1,
+	.tseg1_max = 16,
+	.tseg2_min = 1,
+	.tseg2_max = 8,
+	.sjw_max = 4,
+	.brp_min = 1,
+	.brp_max = 64,
+	.brp_inc = 1,
+};
+
+/*
+ * This routine was stolen from drivers/net/can/sja1000/sja1000.c
+ *
+ * The bittiming register command for the ICAN3 just sets the bit timing
+ * registers on the SJA1000 chip directly
+ */
+static int ican3_set_bittiming(struct net_device *ndev)
+{
+	struct ican3_dev *mod = netdev_priv(ndev);
+	struct can_bittiming *bt = &mod->can.bittiming;
+	struct ican3_msg msg;
+	u8 btr0, btr1;
+
+	btr0 = ((bt->brp - 1) & 0x3f) | (((bt->sjw - 1) & 0x3) << 6);
+	btr1 = ((bt->prop_seg + bt->phase_seg1 - 1) & 0xf) |
+		(((bt->phase_seg2 - 1) & 0x7) << 4);
+	if (mod->can.ctrlmode & CAN_CTRLMODE_3_SAMPLES)
+		btr1 |= 0x80;
+
+	memset(&msg, 0, sizeof(msg));
+	msg.spec = MSG_CBTRREQ;
+	msg.len = cpu_to_le16(4);
+	msg.data[0] = 0x00;
+	msg.data[1] = 0x00;
+	msg.data[2] = btr0;
+	msg.data[3] = btr1;
+
+	return ican3_send_msg(mod, &msg);
+}
+
+static int ican3_set_mode(struct net_device *ndev, enum can_mode mode)
+{
+	struct ican3_dev *mod = netdev_priv(ndev);
+	int ret;
+
+	if (mode != CAN_MODE_START)
+		return -ENOTSUPP;
+
+	/* bring the bus online */
+	ret = ican3_set_bus_state(mod, true);
+	if (ret) {
+		dev_err(mod->dev, "unable to set bus-on\n");
+		return ret;
+	}
+
+	/* start up the network device */
+	mod->can.state = CAN_STATE_ERROR_ACTIVE;
+
+	if (netif_queue_stopped(ndev))
+		netif_wake_queue(ndev);
+
+	return 0;
+}
+
+static int ican3_get_berr_counter(const struct net_device *ndev,
+				  struct can_berr_counter *bec)
+{
+	struct ican3_dev *mod = netdev_priv(ndev);
+	int ret;
+
+	ret = ican3_send_inquiry(mod, INQUIRY_STATUS);
+	if (ret)
+		return ret;
+
+	ret = wait_for_completion_timeout(&mod->buserror_comp, HZ);
+	if (ret <= 0) {
+		dev_info(mod->dev, "%s timed out\n", __func__);
+		return -ETIMEDOUT;
+	}
+
+	bec->rxerr = mod->bec.rxerr;
+	bec->txerr = mod->bec.txerr;
+	return 0;
+}
+
+/*
+ * Sysfs Attributes
+ */
+
+static ssize_t ican3_sysfs_show_term(struct device *dev,
+				     struct device_attribute *attr,
+				     char *buf)
+{
+	struct ican3_dev *mod = netdev_priv(to_net_dev(dev));
+	int ret;
+
+	ret = ican3_send_inquiry(mod, INQUIRY_TERMINATION);
+	if (ret)
+		return ret;
+
+	ret = wait_for_completion_timeout(&mod->termination_comp, HZ);
+	if (ret <= 0) {
+		dev_info(mod->dev, "%s timed out\n", __func__);
+		return -ETIMEDOUT;
+	}
+
+	return snprintf(buf, PAGE_SIZE, "%u\n", mod->termination_enabled);
+}
+
+static ssize_t ican3_sysfs_set_term(struct device *dev,
+				    struct device_attribute *attr,
+				    const char *buf, size_t count)
+{
+	struct ican3_dev *mod = netdev_priv(to_net_dev(dev));
+	unsigned long enable;
+	int ret;
+
+	if (strict_strtoul(buf, 0, &enable))
+		return -EINVAL;
+
+	ret = ican3_set_termination(mod, enable);
+	if (ret)
+		return ret;
+
+	return count;
+}
+
+static DEVICE_ATTR(termination, S_IWUGO | S_IRUGO, ican3_sysfs_show_term,
+						   ican3_sysfs_set_term);
+
+static struct attribute *ican3_sysfs_attrs[] = {
+	&dev_attr_termination.attr,
+	NULL,
+};
+
+static struct attribute_group ican3_sysfs_attr_group = {
+	.attrs = ican3_sysfs_attrs,
+};
+
+/*
+ * PCI Subsystem
+ */
+
+static int __devinit ican3_probe(struct platform_device *pdev)
+{
+	struct janz_platform_data *pdata;
+	struct net_device *ndev;
+	struct ican3_dev *mod;
+	struct resource *res;
+	struct device *dev;
+	int ret;
+
+	pdata = pdev->dev.platform_data;
+	if (!pdata)
+		return -ENXIO;
+
+	dev_dbg(&pdev->dev, "probe: module number %d\n", pdata->modno);
+
+	/* save the struct device for printing */
+	dev = &pdev->dev;
+
+	/* allocate the CAN device and private data */
+	ndev = alloc_candev(sizeof(*mod), 0);
+	if (!ndev) {
+		dev_err(dev, "unable to allocate CANdev\n");
+		ret = -ENOMEM;
+		goto out_return;
+	}
+
+	platform_set_drvdata(pdev, ndev);
+	mod = netdev_priv(ndev);
+	mod->ndev = ndev;
+	mod->dev = &pdev->dev;
+	mod->num = pdata->modno;
+	netif_napi_add(ndev, &mod->napi, ican3_napi, ICAN3_RX_BUFFERS);
+	spin_lock_init(&mod->lock);
+	init_completion(&mod->termination_comp);
+	init_completion(&mod->buserror_comp);
+
+	/* setup device-specific sysfs attributes */
+	ndev->sysfs_groups[0] = &ican3_sysfs_attr_group;
+
+	/* the first unallocated page in the DPM is 9 */
+	mod->free_page = DPM_FREE_START;
+
+	ndev->netdev_ops = &ican3_netdev_ops;
+	ndev->flags |= IFF_ECHO;
+	SET_NETDEV_DEV(ndev, &pdev->dev);
+
+	mod->can.clock.freq = ICAN3_CAN_CLOCK;
+	mod->can.bittiming_const = &ican3_bittiming_const;
+	mod->can.do_set_bittiming = ican3_set_bittiming;
+	mod->can.do_set_mode = ican3_set_mode;
+	mod->can.do_get_berr_counter = ican3_get_berr_counter;
+	mod->can.ctrlmode_supported = CAN_CTRLMODE_3_SAMPLES
+				    | CAN_CTRLMODE_BERR_REPORTING;
+
+	/* find our IRQ number */
+	mod->irq = platform_get_irq(pdev, 0);
+	if (mod->irq < 0) {
+		dev_err(dev, "IRQ line not found\n");
+		ret = -ENODEV;
+		goto out_free_ndev;
+	}
+
+	ndev->irq = mod->irq;
+
+	/* get access to the MODULbus registers for this module */
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!res) {
+		dev_err(dev, "MODULbus registers not found\n");
+		ret = -ENODEV;
+		goto out_free_ndev;
+	}
+
+	mod->dpm = ioremap(res->start, resource_size(res));
+	if (!mod->dpm) {
+		dev_err(dev, "MODULbus registers not ioremap\n");
+		ret = -ENOMEM;
+		goto out_free_ndev;
+	}
+
+	mod->dpmctrl = mod->dpm + DPM_PAGE_SIZE;
+
+	/* get access to the control registers for this module */
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 1);
+	if (!res) {
+		dev_err(dev, "CONTROL registers not found\n");
+		ret = -ENODEV;
+		goto out_iounmap_dpm;
+	}
+
+	mod->ctrl = ioremap(res->start, resource_size(res));
+	if (!mod->ctrl) {
+		dev_err(dev, "CONTROL registers not ioremap\n");
+		ret = -ENOMEM;
+		goto out_iounmap_dpm;
+	}
+
+	/* disable our IRQ, then hookup the IRQ handler */
+	iowrite8(1 << mod->num, &mod->ctrl->int_disable);
+	ret = request_irq(mod->irq, ican3_irq, IRQF_SHARED, DRV_NAME, mod);
+	if (ret) {
+		dev_err(dev, "unable to request IRQ\n");
+		goto out_iounmap_ctrl;
+	}
+
+	/* reset and initialize the CAN controller into fast mode */
+	napi_enable(&mod->napi);
+	ret = ican3_startup_module(mod);
+	if (ret) {
+		dev_err(dev, "%s: unable to start CANdev\n", __func__);
+		goto out_free_irq;
+	}
+
+	/* register with the Linux CAN layer */
+	ret = register_candev(ndev);
+	if (ret) {
+		dev_err(dev, "%s: unable to register CANdev\n", __func__);
+		goto out_free_irq;
+	}
+
+	dev_info(dev, "module %d: registered CAN device\n", pdata->modno);
+	return 0;
+
+out_free_irq:
+	napi_disable(&mod->napi);
+	iowrite8(1 << mod->num, &mod->ctrl->int_disable);
+	free_irq(mod->irq, mod);
+out_iounmap_ctrl:
+	iounmap(mod->ctrl);
+out_iounmap_dpm:
+	iounmap(mod->dpm);
+out_free_ndev:
+	free_candev(ndev);
+out_return:
+	return ret;
+}
+
+static int __devexit ican3_remove(struct platform_device *pdev)
+{
+	struct net_device *ndev = platform_get_drvdata(pdev);
+	struct ican3_dev *mod = netdev_priv(ndev);
+
+	/* unregister the netdevice, stop interrupts */
+	unregister_netdev(ndev);
+	napi_disable(&mod->napi);
+	iowrite8(1 << mod->num, &mod->ctrl->int_disable);
+	free_irq(mod->irq, mod);
+
+	/* put the module into reset */
+	ican3_shutdown_module(mod);
+
+	/* unmap all registers */
+	iounmap(mod->ctrl);
+	iounmap(mod->dpm);
+
+	free_candev(ndev);
+
+	return 0;
+}
+
+static struct platform_driver ican3_driver = {
+	.driver		= {
+		.name	= DRV_NAME,
+		.owner	= THIS_MODULE,
+	},
+	.probe		= ican3_probe,
+	.remove		= __devexit_p(ican3_remove),
+};
+
+static int __init ican3_init(void)
+{
+	return platform_driver_register(&ican3_driver);
+}
+
+static void __exit ican3_exit(void)
+{
+	platform_driver_unregister(&ican3_driver);
+}
+
+MODULE_AUTHOR("Ira W. Snyder <iws-lulEs6mt1IksTUYHLfqkUA@public.gmane.org>");
+MODULE_DESCRIPTION("Janz MODULbus VMOD-ICAN3 Driver");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:janz-ican3");
+
+module_init(ican3_init);
+module_exit(ican3_exit);
-- 
1.5.4.3

^ permalink raw reply related

* [PATCH 3/3] gpio: add support for Janz VMOD-TTL Digital IO module
From: Ira W. Snyder @ 2010-03-29 16:58 UTC (permalink / raw)
  To: linux-kernel-u79uwXL29TY76Z2rM5mHXA
  Cc: socketcan-core-0fE9KPoRgkgATYTw5x5z8w,
	netdev-u79uwXL29TY76Z2rM5mHXA, sameo-VuQAYsv1563Yd54FQh9/CA

The Janz VMOD-TTL is a MODULbus daughterboard which fits onto any MODULbus
carrier board. It essentially consists of some various logic and a Zilog
Z8536 CIO Counter/Timer and Parallel IO Unit.

The board must be physically configured with jumpers to enable a user to
drive output signals. I am only interested in outputs, so I have made this
driver as simple as possible. It only supports a very minimal subset of the
features provided by the Z8536 chip.

Signed-off-by: Ira W. Snyder <iws-lulEs6mt1IksTUYHLfqkUA@public.gmane.org>
---
 drivers/gpio/Kconfig    |   10 ++
 drivers/gpio/Makefile   |    1 +
 drivers/gpio/janz-ttl.c |  257 +++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 268 insertions(+), 0 deletions(-)
 create mode 100644 drivers/gpio/janz-ttl.c

diff --git a/drivers/gpio/Kconfig b/drivers/gpio/Kconfig
index dd10eb8..1f38ff3 100644
--- a/drivers/gpio/Kconfig
+++ b/drivers/gpio/Kconfig
@@ -318,4 +318,14 @@ config GPIO_UCB1400
 	  To compile this driver as a module, choose M here: the
 	  module will be called ucb1400_gpio.
 
+comment "MODULbus GPIO expanders:"
+
+config GPIO_JANZ_TTL
+	tristate "Janz VMOD-TTL Digital IO Module"
+	depends on MFD_JANZ_CMODIO
+	help
+	  This enables support for the Janz VMOD-TTL Digital IO module.
+	  This driver provides support for driving the pins in output
+	  mode only. Input mode is not supported.
+
 endif
diff --git a/drivers/gpio/Makefile b/drivers/gpio/Makefile
index d3226d3..94a96c5 100644
--- a/drivers/gpio/Makefile
+++ b/drivers/gpio/Makefile
@@ -29,3 +29,4 @@ obj-$(CONFIG_GPIO_WM8350)	+= wm8350-gpiolib.o
 obj-$(CONFIG_GPIO_WM8994)	+= wm8994-gpio.o
 obj-$(CONFIG_GPIO_SCH)		+= sch_gpio.o
 obj-$(CONFIG_GPIO_RDC321X)	+= rdc321x-gpio.o
+obj-$(CONFIG_GPIO_JANZ_TTL)	+= janz-ttl.o
diff --git a/drivers/gpio/janz-ttl.c b/drivers/gpio/janz-ttl.c
new file mode 100644
index 0000000..d97eeda
--- /dev/null
+++ b/drivers/gpio/janz-ttl.c
@@ -0,0 +1,257 @@
+/*
+ * Janz MODULbus VMOD-TTL GPIO Driver
+ *
+ * Copyright (c) 2010 Ira W. Snyder <iws-lulEs6mt1IksTUYHLfqkUA@public.gmane.org>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2 of the License, or (at your
+ * option) any later version.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/delay.h>
+#include <linux/platform_device.h>
+#include <linux/io.h>
+#include <linux/gpio.h>
+
+#include <linux/mfd/janz.h>
+
+#define DRV_NAME "janz-ttl"
+
+#define PORTA_DIRECTION		0x23
+#define PORTB_DIRECTION		0x2B
+#define PORTC_DIRECTION		0x06
+#define PORTA_IOCTL		0x24
+#define PORTB_IOCTL		0x2C
+#define PORTC_IOCTL		0x07
+
+#define MASTER_INT_CTL		0x00
+#define MASTER_CONF_CTL		0x01
+
+#define CONF_PAE		(1 << 2)
+#define CONF_PBE		(1 << 7)
+#define CONF_PCE		(1 << 4)
+
+struct ttl_control_regs {
+	__be16 portc;
+	__be16 portb;
+	__be16 porta;
+	__be16 control;
+};
+
+struct ttl_module {
+	struct gpio_chip gpio;
+
+	/* base address of registers */
+	struct ttl_control_regs __iomem *regs;
+
+	u8 portc_shadow;
+	u8 portb_shadow;
+	u8 porta_shadow;
+
+	spinlock_t lock;
+};
+
+static int ttl_get_value(struct gpio_chip *gpio, unsigned offset)
+{
+	struct ttl_module *mod = dev_get_drvdata(gpio->dev);
+	u8 *shadow;
+	int ret;
+
+	if (offset < 8) {
+		shadow = &mod->porta_shadow;
+	} else if (offset < 16) {
+		shadow = &mod->portb_shadow;
+		offset -= 8;
+	} else {
+		shadow = &mod->portc_shadow;
+		offset -= 16;
+	}
+
+	spin_lock(&mod->lock);
+	ret = *shadow & (1 << offset);
+	spin_unlock(&mod->lock);
+	return ret;
+}
+
+static void ttl_set_value(struct gpio_chip *gpio, unsigned offset, int value)
+{
+	struct ttl_module *mod = dev_get_drvdata(gpio->dev);
+	void __iomem *port;
+	u8 *shadow;
+
+	if (offset < 8) {
+		port = &mod->regs->porta;
+		shadow = &mod->porta_shadow;
+	} else if (offset < 16) {
+		port = &mod->regs->portb;
+		shadow = &mod->portb_shadow;
+		offset -= 8;
+	} else {
+		port = &mod->regs->portc;
+		shadow = &mod->portc_shadow;
+		offset -= 16;
+	}
+
+	spin_lock(&mod->lock);
+	if (value)
+		*shadow |= (1 << offset);
+	else
+		*shadow &= ~(1 << offset);
+
+	iowrite16be(*shadow, port);
+	spin_unlock(&mod->lock);
+}
+
+static void __devinit ttl_write_reg(struct ttl_module *mod, u8 reg, u16 val)
+{
+	iowrite16be(reg, &mod->regs->control);
+	iowrite16be(val, &mod->regs->control);
+}
+
+static void __devinit ttl_setup_device(struct ttl_module *mod)
+{
+	/* reset the device to a known state */
+	iowrite16be(0x0000, &mod->regs->control);
+	iowrite16be(0x0001, &mod->regs->control);
+	iowrite16be(0x0000, &mod->regs->control);
+
+	/* put all ports in open-drain mode */
+	ttl_write_reg(mod, PORTA_IOCTL, 0x00ff);
+	ttl_write_reg(mod, PORTB_IOCTL, 0x00ff);
+	ttl_write_reg(mod, PORTC_IOCTL, 0x000f);
+
+	/* set all ports as outputs */
+	ttl_write_reg(mod, PORTA_DIRECTION, 0x0000);
+	ttl_write_reg(mod, PORTB_DIRECTION, 0x0000);
+	ttl_write_reg(mod, PORTC_DIRECTION, 0x0000);
+
+	/* set all ports to drive zeroes */
+	iowrite16be(0x0000, &mod->regs->porta);
+	iowrite16be(0x0000, &mod->regs->portb);
+	iowrite16be(0x0000, &mod->regs->portc);
+
+	/* enable all ports */
+	ttl_write_reg(mod, MASTER_CONF_CTL, CONF_PAE | CONF_PBE | CONF_PCE);
+}
+
+static int __devinit ttl_probe(struct platform_device *pdev)
+{
+	struct janz_platform_data *pdata;
+	struct device *dev = &pdev->dev;
+	struct ttl_module *mod;
+	struct gpio_chip *gpio;
+	struct resource *res;
+	int ret;
+
+	pdata = pdev->dev.platform_data;
+	if (!pdata) {
+		dev_err(dev, "no platform data\n");
+		ret = -ENXIO;
+		goto out_return;
+	}
+
+	mod = kzalloc(sizeof(*mod), GFP_KERNEL);
+	if (!mod) {
+		dev_err(dev, "unable to allocate private data\n");
+		ret = -ENOMEM;
+		goto out_return;
+	}
+
+	platform_set_drvdata(pdev, mod);
+	spin_lock_init(&mod->lock);
+
+	/* get access to the MODULbus registers for this module */
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!res) {
+		dev_err(dev, "MODULbus registers not found\n");
+		ret = -ENODEV;
+		goto out_free_mod;
+	}
+
+	mod->regs = ioremap(res->start, resource_size(res));
+	if (!mod->regs) {
+		dev_err(dev, "MODULbus registers not ioremap\n");
+		ret = -ENOMEM;
+		goto out_free_mod;
+	}
+
+	ttl_setup_device(mod);
+
+	/* Initialize the GPIO data structures */
+	gpio = &mod->gpio;
+	gpio->dev = &pdev->dev;
+	gpio->label = pdev->name;
+	gpio->get = ttl_get_value;
+	gpio->set = ttl_set_value;
+	gpio->owner = THIS_MODULE;
+
+	/* request dynamic allocation */
+	gpio->base = -1;
+	gpio->ngpio = 20;
+
+	ret = gpiochip_add(gpio);
+	if (ret) {
+		dev_err(dev, "unable to add GPIO chip\n");
+		goto out_iounmap_regs;
+	}
+
+	dev_info(&pdev->dev, "module %d: registered GPIO device\n",
+			     pdata->modno);
+	return 0;
+
+out_iounmap_regs:
+	iounmap(mod->regs);
+out_free_mod:
+	kfree(mod);
+out_return:
+	return ret;
+}
+
+static int __devexit ttl_remove(struct platform_device *pdev)
+{
+	struct ttl_module *mod = platform_get_drvdata(pdev);
+	struct device *dev = &pdev->dev;
+	int ret;
+
+	ret = gpiochip_remove(&mod->gpio);
+	if (ret) {
+		dev_err(dev, "unable to remove GPIO chip\n");
+		return ret;
+	}
+
+	iounmap(mod->regs);
+	kfree(mod);
+	return 0;
+}
+
+static struct platform_driver ttl_driver = {
+	.driver		= {
+		.name	= DRV_NAME,
+		.owner	= THIS_MODULE,
+	},
+	.probe		= ttl_probe,
+	.remove		= __devexit_p(ttl_remove),
+};
+
+static int __init ttl_init(void)
+{
+	return platform_driver_register(&ttl_driver);
+}
+
+static void __exit ttl_exit(void)
+{
+	platform_driver_unregister(&ttl_driver);
+}
+
+MODULE_AUTHOR("Ira W. Snyder <iws-lulEs6mt1IksTUYHLfqkUA@public.gmane.org>");
+MODULE_DESCRIPTION("Janz MODULbus VMOD-TTL Driver");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:janz-ttl");
+
+module_init(ttl_init);
+module_exit(ttl_exit);
-- 
1.5.4.3

^ permalink raw reply related

* Re: UDP path MTU discovery
From: Rick Jones @ 2010-03-29 17:01 UTC (permalink / raw)
  To: Andi Kleen; +Cc: Glen Turner, netdev
In-Reply-To: <877how25kx.fsf@basil.nowhere.org>

> In theory one could probably add some hack in the the kernel UDP code
> to hold one packet and retransmit it immediately with fragments when
> the ICMP comes in. However that would be quite far in behaviour from
> traditional UDP and be considered very ugly. It could also mess up
> congestion avoidance schemes done by the application. 
> 
> Still might be preferable over rewriting zillions of applications?

But which of the last N datagrams sent by the application should be retained for 
retransmission?  It could be scores if not hundreds of datagrams depending on 
the behaviour of the application and the latency to the narrow part of the network.

That the IPv6 specification was heavily "influenced" by "the router guys" seems 
increasingly clear...

rick jones

^ permalink raw reply

* Re: behavior of recvmmsg() on blocking sockets
From: Brandon Black @ 2010-03-29 17:24 UTC (permalink / raw)
  To: Chris Friesen; +Cc: Arnaldo Carvalho de Melo, linux-kernel, netdev
In-Reply-To: <4BB0D2CF.9040903@nortel.com>

On Mon, Mar 29, 2010 at 11:18 AM, Chris Friesen <cfriesen@nortel.com> wrote:
>
> prev = current time
> loop forever
>        cur = current time
>        timeout = max_latency - (cur - prev)
>        recvmmsg(timeout)
>        process all received messages
>        prev = cur
>
>
> Basically you determine the max latency you're willing to wait for a
> packet to be handled, then subtract the amount of time you spent
> processing messages from that and pass it into the recvmmsg() call as
> the timeout.  That way no messages will be delayed for longer than the
> max latency. (Not considering scheduling delays.)

With a blocking socket, you'd also need to set SO_RCVTIMEO on the
underlying socket to some value that makes sense and is below your max
latency, because recvmmsg()'s timeout argument only applies in-between
underlying recvmsg() calls, not during them.  You're going to spend a
lot of time spinning if max_latency is low and there are any gaps in
the input stream though.  I guess for some uses this must makes sense.

The other potential usage is with non-blocking sockets, in which case
the timeout argument is putting on upper boundary on how long
recvmmsg() can spend fetching packets from the queue before it must
return, even if more packets are available.  Seems like for a given
kernel and hardware you could accomplish the same by tuning the vlen
argument.  In either case though, it seems like if you're running into
your hard latency limit on a non-blocking packet fetch and there are
already more packets waiting, you're probably (at least) verging on
being unable to meet the latency requirement for (at least) some of
your packets due to a hard lack of CPU horsepower for the workload.

-- Brandon

^ permalink raw reply

* Re: eth1: Detected Hardware Unit Hang
From: Paweł Staszewski @ 2010-03-29 17:29 UTC (permalink / raw)
  To: Allan, Bruce W
  Cc: Linux Network Development list, e1000-devel@lists.sourceforge.net
In-Reply-To: <8DD2590731AB5D4C9DBF71A877482A9061BB3254@orsmsx509.amr.corp.intel.com>

[-- Attachment #1: Type: text/plain, Size: 4089 bytes --]

lspci -vvv + ethtool -S in attached files.

Network traffic when i get this info:
eth1:    RX:    157.22 Mb/s    TX:    379.27 Mb/s

ethtool -i eth1
driver: e1000e
version: 1.0.2-k2
firmware-version: 0.5-7
bus-info: 0000:05:00.0
This is: Intel Corporation 82573L Gigabit Ethernet Controller


But in this server i have another gigabit interface:
Intel Corporation 82573E Gigabit Ethernet Controller
this interface has two times more traffic than eth0 (82573L)
ethtool -i eth0
driver: e1000e
version: 1.0.2-k2
firmware-version: 0.15-5
bus-info: 0000:04:00.0

And also this server was working 4months without problems on 2.6.29.1 kernel

Drivers that I use for e1000e are from kernel (standard kernel build-in 
e1000e driver).
I don't tried other drivers.

This is production server so I can't make too much tests.


W dniu 2010-03-29 18:41, Allan, Bruce W pisze:
> [adding e1000-devel]
>
> Please provide more information:
> * what NIC/LOM is this on (preferably send full output from lspci -vvv)
> * what type of networking workload is running at the time the hang occurred
> * a dump of the NIC/LOM statistics might also help (ethtool -S eth1)
>
> Have you tried the latest standalone e1000e driver on e1000.sf.net?  Does it reproduce the issue?
>
> If we cannot reproduce the hang in-house, would you be able/willing to run a debug driver to gather more information?
>
> Thanks,
> Bruce.
>
> -----Original Message-----
> From: netdev-owner@vger.kernel.org [mailto:netdev-owner@vger.kernel.org] On Behalf Of Pawel Staszewski
> Sent: Monday, March 29, 2010 8:34 AM
> To: Linux Network Development list
> Subject: eth1: Detected Hardware Unit Hang
>
> After update to kernel from 2.6.29.1 to 2.6.33.1 i have this info in dmesg:
>
> 0000:05:00.0: eth1: Detected Hardware Unit Hang:
>     TDH<1e>
>     TDT<a>
>     next_to_use<a>
>     next_to_clean<1d>
> buffer_info[next_to_clean]:
>     time_stamp<33bae15>
>     next_to_watch<20>
>     jiffies<33bafaf>
>     next_to_watch.status<0>
> MAC Status<80080783>
> PHY Status<796d>
> PHY 1000BASE-T Status<3800>
> PHY Extended Status<3000>
> PCI Status<10>
> 0000:05:00.0: eth1: Detected Hardware Unit Hang:
>     TDH<1e>
>     TDT<a>
>     next_to_use<a>
>     next_to_clean<1d>
> buffer_info[next_to_clean]:
>     time_stamp<33bae15>
>     next_to_watch<20>
>     jiffies<33bb1a3>
>     next_to_watch.status<0>
> MAC Status<80080783>
> PHY Status<796d>
> PHY 1000BASE-T Status<3800>
> PHY Extended Status<3000>
> PCI Status<10>
> 0000:05:00.0: eth1: Detected Hardware Unit Hang:
>     TDH<1e>
>     TDT<a>
>     next_to_use<a>
>     next_to_clean<1d>
> buffer_info[next_to_clean]:
>     time_stamp<33bae15>
>     next_to_watch<20>
>     jiffies<33bb397>
>     next_to_watch.status<0>
> MAC Status<80080783>
> PHY Status<796d>
> PHY 1000BASE-T Status<3800>
> PHY Extended Status<3000>
> PCI Status<10>
> ------------[ cut here ]------------
> WARNING: at net/sched/sch_generic.c:255 dev_watchdog+0x118/0x19c()
> Hardware name: X7DCT
> NETDEV WATCHDOG: eth1 (e1000e): transmit queue 0 timed out
> Modules linked in:
> Pid: 0, comm: swapper Not tainted 2.6.33.1 #2
> Call Trace:
>    [<c1024e3d>] ? warn_slowpath_common+0x52/0x71
>    [<c1024e49>] ? warn_slowpath_common+0x5e/0x71
>    [<c1024e8e>] ? warn_slowpath_fmt+0x26/0x2a
>    [<c1261f54>] ? dev_watchdog+0x118/0x19c
>    [<c102135c>] ? __wake_up+0x29/0x39
>    [<c10320c6>] ? insert_work+0x40/0x44
>    [<c1261e3c>] ? dev_watchdog+0x0/0x19c
>    [<c102cc15>] ? run_timer_softirq+0x11a/0x173
>    [<c1028e5b>] ? __do_softirq+0x74/0xdf
>    [<c1028ee9>] ? do_softirq+0x23/0x27
>    [<c10290be>] ? irq_exit+0x26/0x58
>    [<c10102d7>] ? smp_apic_timer_interrupt+0x6c/0x76
>    [<c12c5f9a>] ? apic_timer_interrupt+0x2a/0x30
>    [<c1007e06>] ? mwait_idle+0x49/0x4e
>    [<c10017e8>] ? cpu_idle+0x41/0x5a
> ---[ end trace bcca9926a046332c ]---
>
>
> With kernel 2.6.29.1 all was ok.
> --
> 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
>
>
>    


[-- Attachment #2: lspci.txt --]
[-- Type: text/plain, Size: 30803 bytes --]

00:00.0 Host bridge: Intel Corporation 5100 Chipset Memory Controller Hub (rev 90)
	Subsystem: Super Micro Computer Inc Device de80
	Control: I/O- Mem- BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx-
	Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 0, Cache Line Size: 32 bytes
	Capabilities: [50] Power Management version 2
		Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0+,D1-,D2-,D3hot+,D3cold+)
		Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-
	Capabilities: [58] MSI: Enable- Count=1/1 Maskable- 64bit-
		Address: fee00000  Data: 0000
	Capabilities: [6c] Express (v1) Root Port (Slot-), MSI 00
		DevCap:	MaxPayload 256 bytes, PhantFunc 0, Latency L0s unlimited, L1 unlimited
			ExtTag- RBE- FLReset-
		DevCtl:	Report errors: Correctable- Non-Fatal- Fatal- Unsupported-
			RlxdOrd- ExtTag- PhantFunc- AuxPwr- NoSnoop+
			MaxPayload 128 bytes, MaxReadReq 128 bytes
		DevSta:	CorrErr- UncorrErr- FatalErr- UnsuppReq- AuxPwr- TransPend-
		LnkCap:	Port #0, Speed 2.5GT/s, Width x4, ASPM L0s, Latency L0 unlimited, L1 unlimited
			ClockPM- Surprise+ LLActRep+ BwNot-
		LnkCtl:	ASPM Disabled; RCB 64 bytes Disabled- Retrain- CommClk-
			ExtSynch- ClockPM- AutWidDis- BWInt- AutBWInt-
		LnkSta:	Speed 2.5GT/s, Width x4, TrErr- Train- SlotClk+ DLActive+ BWMgmt- ABWMgmt-
		RootCtl: ErrCorrectable- ErrNon-Fatal- ErrFatal- PMEIntEna- CRSVisible-
		RootCap: CRSVisible-
		RootSta: PME ReqID 0000, PMEStatus- PMEPending-

00:02.0 PCI bridge: Intel Corporation 5100 Chipset PCI Express x8 Port 2-3 (rev 90) (prog-if 00 [Normal decode])
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx+
	Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 0, Cache Line Size: 32 bytes
	Bus: primary=00, secondary=01, subordinate=01, sec-latency=0
	I/O behind bridge: 00002000-00002fff
	Memory behind bridge: f8000000-f80fffff
	Prefetchable memory behind bridge: 00000000fff00000-00000000000fffff
	Secondary status: 66MHz- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- <SERR- <PERR-
	BridgeCtl: Parity- SERR- NoISA+ VGA- MAbort- >Reset- FastB2B-
		PriDiscTmr- SecDiscTmr- DiscTmrStat- DiscTmrSERREn-
	Capabilities: [50] Power Management version 2
		Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0+,D1-,D2-,D3hot+,D3cold+)
		Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-
	Capabilities: [58] MSI: Enable+ Count=1/1 Maskable- 64bit-
		Address: feeff00c  Data: 4181
	Capabilities: [6c] Express (v1) Root Port (Slot+), MSI 00
		DevCap:	MaxPayload 256 bytes, PhantFunc 0, Latency L0s unlimited, L1 unlimited
			ExtTag- RBE- FLReset-
		DevCtl:	Report errors: Correctable- Non-Fatal- Fatal- Unsupported-
			RlxdOrd- ExtTag- PhantFunc- AuxPwr- NoSnoop+
			MaxPayload 128 bytes, MaxReadReq 128 bytes
		DevSta:	CorrErr- UncorrErr- FatalErr- UnsuppReq- AuxPwr- TransPend-
		LnkCap:	Port #2, Speed 2.5GT/s, Width x8, ASPM L0s, Latency L0 unlimited, L1 unlimited
			ClockPM- Surprise+ LLActRep+ BwNot-
		LnkCtl:	ASPM Disabled; RCB 64 bytes Disabled- Retrain- CommClk+
			ExtSynch- ClockPM- AutWidDis- BWInt- AutBWInt-
		LnkSta:	Speed 2.5GT/s, Width x8, TrErr- Train- SlotClk+ DLActive+ BWMgmt- ABWMgmt-
		SltCap:	AttnBtn- PwrCtrl- MRL- AttnInd- PwrInd- HotPlug- Surpise-
			Slot # 2e, PowerLimit 25.000000; Interlock- NoCompl-
		SltCtl:	Enable: AttnBtn+ PwrFlt+ MRL+ PresDet+ CmdCplt- HPIrq+ LinkChg-
			Control: AttnInd Off, PwrInd On, Power- Interlock-
		SltSta:	Status: AttnBtn- PowerFlt- MRL+ CmdCplt- PresDet+ Interlock-
			Changed: MRL- PresDet- LinkState-
		RootCtl: ErrCorrectable- ErrNon-Fatal- ErrFatal- PMEIntEna- CRSVisible-
		RootCap: CRSVisible-
		RootSta: PME ReqID 0000, PMEStatus- PMEPending-
	Kernel driver in use: pcieport

00:04.0 PCI bridge: Intel Corporation 5100 Chipset PCI Express x16 Port 4-7 (rev 90) (prog-if 00 [Normal decode])
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx+
	Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 0, Cache Line Size: 32 bytes
	Bus: primary=00, secondary=02, subordinate=02, sec-latency=0
	I/O behind bridge: 00006000-00006fff
	Memory behind bridge: f8400000-f85fffff
	Prefetchable memory behind bridge: 00000000f8700000-00000000f88fffff
	Secondary status: 66MHz- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- <SERR- <PERR-
	BridgeCtl: Parity- SERR- NoISA+ VGA- MAbort- >Reset- FastB2B-
		PriDiscTmr- SecDiscTmr- DiscTmrStat- DiscTmrSERREn-
	Capabilities: [50] Power Management version 2
		Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0+,D1-,D2-,D3hot+,D3cold+)
		Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-
	Capabilities: [58] MSI: Enable+ Count=1/1 Maskable- 64bit-
		Address: feeff00c  Data: 4189
	Capabilities: [6c] Express (v1) Root Port (Slot+), MSI 00
		DevCap:	MaxPayload 256 bytes, PhantFunc 0, Latency L0s unlimited, L1 unlimited
			ExtTag- RBE- FLReset-
		DevCtl:	Report errors: Correctable- Non-Fatal- Fatal- Unsupported-
			RlxdOrd- ExtTag- PhantFunc- AuxPwr- NoSnoop+
			MaxPayload 128 bytes, MaxReadReq 128 bytes
		DevSta:	CorrErr- UncorrErr- FatalErr- UnsuppReq- AuxPwr- TransPend-
		LnkCap:	Port #4, Speed 2.5GT/s, Width x16, ASPM L0s, Latency L0 unlimited, L1 unlimited
			ClockPM- Surprise+ LLActRep+ BwNot-
		LnkCtl:	ASPM Disabled; RCB 64 bytes Disabled- Retrain- CommClk-
			ExtSynch- ClockPM- AutWidDis- BWInt- AutBWInt-
		LnkSta:	Speed 2.5GT/s, Width x4, TrErr- Train- SlotClk+ DLActive- BWMgmt- ABWMgmt-
		SltCap:	AttnBtn+ PwrCtrl+ MRL+ AttnInd+ PwrInd+ HotPlug+ Surpise-
			Slot # 30, PowerLimit 25.000000; Interlock- NoCompl-
		SltCtl:	Enable: AttnBtn+ PwrFlt+ MRL+ PresDet+ CmdCplt- HPIrq+ LinkChg-
			Control: AttnInd Off, PwrInd Off, Power+ Interlock-
		SltSta:	Status: AttnBtn- PowerFlt- MRL+ CmdCplt- PresDet- Interlock-
			Changed: MRL- PresDet- LinkState-
		RootCtl: ErrCorrectable- ErrNon-Fatal- ErrFatal- PMEIntEna- CRSVisible-
		RootCap: CRSVisible-
		RootSta: PME ReqID 0000, PMEStatus- PMEPending-
	Kernel driver in use: pcieport

00:08.0 System peripheral: Intel Corporation 5100 Chipset DMA Engine (rev 90)
	Subsystem: Super Micro Computer Inc Device de80
	Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx+
	Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 0
	Interrupt: pin A routed to IRQ 40
	Region 0: Memory at fe700000 (64-bit, non-prefetchable) [size=1K]
	Capabilities: [50] Power Management version 2
		Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0+,D1-,D2-,D3hot+,D3cold+)
		Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-
	Capabilities: [58] MSI: Enable+ Count=1/1 Maskable- 64bit-
		Address: feeff00c  Data: 4152
	Capabilities: [6c] Express (v1) Root Complex Integrated Endpoint, MSI 00
		DevCap:	MaxPayload 128 bytes, PhantFunc 0, Latency L0s <64ns, L1 <1us
			ExtTag- RBE- FLReset-
		DevCtl:	Report errors: Correctable- Non-Fatal- Fatal+ Unsupported-
			RlxdOrd- ExtTag- PhantFunc- AuxPwr- NoSnoop+
			MaxPayload 128 bytes, MaxReadReq 128 bytes
		DevSta:	CorrErr- UncorrErr- FatalErr- UnsuppReq- AuxPwr- TransPend-
		LnkCap:	Port #0, Speed unknown, Width x0, ASPM unknown, Latency L0 <64ns, L1 <1us
			ClockPM- Surprise- LLActRep- BwNot-
		LnkCtl:	ASPM Disabled; Disabled- Retrain- CommClk-
			ExtSynch- ClockPM- AutWidDis- BWInt- AutBWInt-
		LnkSta:	Speed unknown, Width x0, TrErr- Train- SlotClk- DLActive- BWMgmt- ABWMgmt-
	Kernel driver in use: ioatdma

00:10.0 Host bridge: Intel Corporation 5100 Chipset FSB Registers (rev 90)
	Subsystem: Super Micro Computer Inc Device de80
	Control: I/O- Mem- BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
	Status: Cap- 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-

00:10.1 Host bridge: Intel Corporation 5100 Chipset FSB Registers (rev 90)
	Subsystem: Super Micro Computer Inc Device de80
	Control: I/O- Mem- BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
	Status: Cap- 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-

00:10.2 Host bridge: Intel Corporation 5100 Chipset FSB Registers (rev 90)
	Subsystem: Super Micro Computer Inc Device de80
	Control: I/O- Mem- BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
	Status: Cap- 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-

00:11.0 Host bridge: Intel Corporation 5100 Chipset Reserved Registers (rev 90)
	Subsystem: Super Micro Computer Inc Device de80
	Control: I/O- Mem- BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
	Status: Cap- 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-

00:13.0 Host bridge: Intel Corporation 5100 Chipset Reserved Registers (rev 90)
	Subsystem: Super Micro Computer Inc Device de80
	Control: I/O- Mem- BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
	Status: Cap- 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-

00:15.0 Host bridge: Intel Corporation 5100 Chipset DDR Channel 0 Registers (rev 90)
	Subsystem: Super Micro Computer Inc Device de80
	Control: I/O- Mem- BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
	Status: Cap- 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-

00:16.0 Host bridge: Intel Corporation 5100 Chipset DDR Channel 1 Registers (rev 90)
	Subsystem: Super Micro Computer Inc Device de80
	Control: I/O- Mem- BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
	Status: Cap- 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-

00:1a.0 USB Controller: Intel Corporation 82801I (ICH9 Family) USB UHCI Controller #4 (rev 02) (prog-if 00 [UHCI])
	Subsystem: Super Micro Computer Inc Device de80
	Control: I/O+ Mem- BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
	Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 0
	Interrupt: pin A routed to IRQ 5
	Region 4: I/O ports at 1820 [size=32]
	Capabilities: [50] PCI Advanced Features
		AFCap: TP+ FLR+
		AFCtrl: FLR-
		AFStatus: TP-

00:1a.7 USB Controller: Intel Corporation 82801I (ICH9 Family) USB2 EHCI Controller #2 (rev 02) (prog-if 20 [EHCI])
	Subsystem: Super Micro Computer Inc Device de80
	Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx-
	Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 0
	Interrupt: pin C routed to IRQ 11
	Region 0: Memory at f8600000 (32-bit, non-prefetchable) [size=1K]
	Capabilities: [50] Power Management version 2
		Flags: PMEClk- DSI- D1- D2- AuxCurrent=375mA PME(D0+,D1-,D2-,D3hot+,D3cold+)
		Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-
	Capabilities: [58] Debug port: BAR=1 offset=00a0
	Capabilities: [98] PCI Advanced Features
		AFCap: TP+ FLR+
		AFCtrl: FLR-
		AFStatus: TP-

00:1c.0 PCI bridge: Intel Corporation 82801I (ICH9 Family) PCI Express Port 1 (rev 02) (prog-if 00 [Normal decode])
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx+
	Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 0, Cache Line Size: 32 bytes
	Bus: primary=00, secondary=03, subordinate=03, sec-latency=0
	I/O behind bridge: 00007000-00007fff
	Memory behind bridge: f8900000-f8afffff
	Prefetchable memory behind bridge: 00000000f8b00000-00000000f8cfffff
	Secondary status: 66MHz- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- <SERR- <PERR-
	BridgeCtl: Parity- SERR- NoISA+ VGA- MAbort- >Reset+ FastB2B-
		PriDiscTmr- SecDiscTmr- DiscTmrStat- DiscTmrSERREn-
	Capabilities: [40] Express (v1) Root Port (Slot+), MSI 00
		DevCap:	MaxPayload 128 bytes, PhantFunc 0, Latency L0s <64ns, L1 <1us
			ExtTag- RBE+ FLReset-
		DevCtl:	Report errors: Correctable- Non-Fatal- Fatal- Unsupported-
			RlxdOrd- ExtTag- PhantFunc- AuxPwr- NoSnoop-
			MaxPayload 128 bytes, MaxReadReq 128 bytes
		DevSta:	CorrErr- UncorrErr- FatalErr- UnsuppReq- AuxPwr+ TransPend-
		LnkCap:	Port #1, Speed 2.5GT/s, Width x1, ASPM L0s L1, Latency L0 <256ns, L1 <4us
			ClockPM- Surprise- LLActRep+ BwNot-
		LnkCtl:	ASPM Disabled; RCB 64 bytes Disabled+ Retrain- CommClk+
			ExtSynch- ClockPM- AutWidDis- BWInt- AutBWInt-
		LnkSta:	Speed 2.5GT/s, Width x0, TrErr- Train- SlotClk+ DLActive- BWMgmt- ABWMgmt-
		SltCap:	AttnBtn- PwrCtrl- MRL- AttnInd- PwrInd- HotPlug+ Surpise+
			Slot #  0, PowerLimit 0.000000; Interlock- NoCompl-
		SltCtl:	Enable: AttnBtn- PwrFlt- MRL- PresDet- CmdCplt- HPIrq- LinkChg-
			Control: AttnInd Unknown, PwrInd Unknown, Power- Interlock-
		SltSta:	Status: AttnBtn- PowerFlt- MRL- CmdCplt- PresDet- Interlock-
			Changed: MRL- PresDet- LinkState-
		RootCtl: ErrCorrectable- ErrNon-Fatal- ErrFatal- PMEIntEna- CRSVisible-
		RootCap: CRSVisible-
		RootSta: PME ReqID 0000, PMEStatus- PMEPending-
	Capabilities: [80] MSI: Enable+ Count=1/1 Maskable- 64bit-
		Address: feeff00c  Data: 4191
	Capabilities: [90] Subsystem: Super Micro Computer Inc Device de80
	Capabilities: [a0] Power Management version 2
		Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0+,D1-,D2-,D3hot+,D3cold+)
		Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-
	Kernel driver in use: pcieport

00:1c.4 PCI bridge: Intel Corporation 82801I (ICH9 Family) PCI Express Port 5 (rev 02) (prog-if 00 [Normal decode])
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx+
	Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 0, Cache Line Size: 32 bytes
	Bus: primary=00, secondary=04, subordinate=04, sec-latency=0
	I/O behind bridge: 00003000-00003fff
	Memory behind bridge: f8100000-f81fffff
	Prefetchable memory behind bridge: 00000000f8d00000-00000000f8efffff
	Secondary status: 66MHz- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- <SERR- <PERR-
	BridgeCtl: Parity- SERR- NoISA+ VGA- MAbort- >Reset- FastB2B-
		PriDiscTmr- SecDiscTmr- DiscTmrStat- DiscTmrSERREn-
	Capabilities: [40] Express (v1) Root Port (Slot+), MSI 00
		DevCap:	MaxPayload 128 bytes, PhantFunc 0, Latency L0s <64ns, L1 <1us
			ExtTag- RBE+ FLReset-
		DevCtl:	Report errors: Correctable- Non-Fatal- Fatal- Unsupported-
			RlxdOrd- ExtTag- PhantFunc- AuxPwr- NoSnoop-
			MaxPayload 128 bytes, MaxReadReq 128 bytes
		DevSta:	CorrErr- UncorrErr- FatalErr- UnsuppReq- AuxPwr+ TransPend-
		LnkCap:	Port #5, Speed 2.5GT/s, Width x1, ASPM L0s L1, Latency L0 <256ns, L1 <4us
			ClockPM- Surprise- LLActRep+ BwNot-
		LnkCtl:	ASPM Disabled; RCB 64 bytes Disabled- Retrain- CommClk+
			ExtSynch- ClockPM- AutWidDis- BWInt- AutBWInt-
		LnkSta:	Speed 2.5GT/s, Width x1, TrErr- Train- SlotClk+ DLActive+ BWMgmt- ABWMgmt-
		SltCap:	AttnBtn- PwrCtrl- MRL- AttnInd- PwrInd- HotPlug+ Surpise+
			Slot #  5, PowerLimit 10.000000; Interlock- NoCompl-
		SltCtl:	Enable: AttnBtn- PwrFlt- MRL- PresDet- CmdCplt- HPIrq- LinkChg-
			Control: AttnInd Unknown, PwrInd Unknown, Power- Interlock-
		SltSta:	Status: AttnBtn- PowerFlt- MRL- CmdCplt- PresDet+ Interlock-
			Changed: MRL- PresDet+ LinkState+
		RootCtl: ErrCorrectable- ErrNon-Fatal- ErrFatal- PMEIntEna- CRSVisible-
		RootCap: CRSVisible-
		RootSta: PME ReqID 0000, PMEStatus- PMEPending-
	Capabilities: [80] MSI: Enable+ Count=1/1 Maskable- 64bit-
		Address: feeff00c  Data: 4199
	Capabilities: [90] Subsystem: Super Micro Computer Inc Device de80
	Capabilities: [a0] Power Management version 2
		Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0+,D1-,D2-,D3hot+,D3cold+)
		Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-
	Kernel driver in use: pcieport

00:1c.5 PCI bridge: Intel Corporation 82801I (ICH9 Family) PCI Express Port 6 (rev 02) (prog-if 00 [Normal decode])
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx+
	Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 0, Cache Line Size: 32 bytes
	Bus: primary=00, secondary=05, subordinate=05, sec-latency=0
	I/O behind bridge: 00004000-00004fff
	Memory behind bridge: f8200000-f82fffff
	Prefetchable memory behind bridge: 00000000f8f00000-00000000f90fffff
	Secondary status: 66MHz- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- <SERR- <PERR-
	BridgeCtl: Parity- SERR- NoISA+ VGA- MAbort- >Reset- FastB2B-
		PriDiscTmr- SecDiscTmr- DiscTmrStat- DiscTmrSERREn-
	Capabilities: [40] Express (v1) Root Port (Slot+), MSI 00
		DevCap:	MaxPayload 128 bytes, PhantFunc 0, Latency L0s <64ns, L1 <1us
			ExtTag- RBE+ FLReset-
		DevCtl:	Report errors: Correctable- Non-Fatal- Fatal- Unsupported-
			RlxdOrd- ExtTag- PhantFunc- AuxPwr- NoSnoop-
			MaxPayload 128 bytes, MaxReadReq 128 bytes
		DevSta:	CorrErr- UncorrErr- FatalErr- UnsuppReq- AuxPwr+ TransPend-
		LnkCap:	Port #6, Speed 2.5GT/s, Width x1, ASPM L0s L1, Latency L0 <256ns, L1 <4us
			ClockPM- Surprise- LLActRep+ BwNot-
		LnkCtl:	ASPM Disabled; RCB 64 bytes Disabled- Retrain- CommClk+
			ExtSynch- ClockPM- AutWidDis- BWInt- AutBWInt-
		LnkSta:	Speed 2.5GT/s, Width x1, TrErr- Train- SlotClk+ DLActive+ BWMgmt- ABWMgmt-
		SltCap:	AttnBtn- PwrCtrl- MRL- AttnInd- PwrInd- HotPlug+ Surpise+
			Slot #  6, PowerLimit 10.000000; Interlock- NoCompl-
		SltCtl:	Enable: AttnBtn- PwrFlt- MRL- PresDet- CmdCplt- HPIrq- LinkChg-
			Control: AttnInd Unknown, PwrInd Unknown, Power- Interlock-
		SltSta:	Status: AttnBtn- PowerFlt- MRL- CmdCplt- PresDet+ Interlock-
			Changed: MRL- PresDet+ LinkState+
		RootCtl: ErrCorrectable- ErrNon-Fatal- ErrFatal- PMEIntEna- CRSVisible-
		RootCap: CRSVisible-
		RootSta: PME ReqID 0000, PMEStatus- PMEPending-
	Capabilities: [80] MSI: Enable+ Count=1/1 Maskable- 64bit-
		Address: feeff00c  Data: 41a1
	Capabilities: [90] Subsystem: Super Micro Computer Inc Device de80
	Capabilities: [a0] Power Management version 2
		Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0+,D1-,D2-,D3hot+,D3cold+)
		Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-
	Kernel driver in use: pcieport

00:1d.0 USB Controller: Intel Corporation 82801I (ICH9 Family) USB UHCI Controller #1 (rev 02) (prog-if 00 [UHCI])
	Subsystem: Super Micro Computer Inc Device de80
	Control: I/O+ Mem- BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
	Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 0
	Interrupt: pin A routed to IRQ 7
	Region 4: I/O ports at 1840 [size=32]
	Capabilities: [50] PCI Advanced Features
		AFCap: TP+ FLR+
		AFCtrl: FLR-
		AFStatus: TP-

00:1d.1 USB Controller: Intel Corporation 82801I (ICH9 Family) USB UHCI Controller #2 (rev 02) (prog-if 00 [UHCI])
	Subsystem: Super Micro Computer Inc Device de80
	Control: I/O+ Mem- BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
	Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 0
	Interrupt: pin B routed to IRQ 11
	Region 4: I/O ports at 1860 [size=32]
	Capabilities: [50] PCI Advanced Features
		AFCap: TP+ FLR+
		AFCtrl: FLR-
		AFStatus: TP-

00:1d.2 USB Controller: Intel Corporation 82801I (ICH9 Family) USB UHCI Controller #3 (rev 02) (prog-if 00 [UHCI])
	Subsystem: Super Micro Computer Inc Device de80
	Control: I/O+ Mem- BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
	Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 0
	Interrupt: pin C routed to IRQ 10
	Region 4: I/O ports at 1880 [size=32]
	Capabilities: [50] PCI Advanced Features
		AFCap: TP+ FLR+
		AFCtrl: FLR-
		AFStatus: TP-

00:1d.7 USB Controller: Intel Corporation 82801I (ICH9 Family) USB2 EHCI Controller #1 (rev 02) (prog-if 20 [EHCI])
	Subsystem: Super Micro Computer Inc Device de80
	Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx-
	Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 0
	Interrupt: pin A routed to IRQ 7
	Region 0: Memory at f8600400 (32-bit, non-prefetchable) [size=1K]
	Capabilities: [50] Power Management version 2
		Flags: PMEClk- DSI- D1- D2- AuxCurrent=375mA PME(D0+,D1-,D2-,D3hot+,D3cold+)
		Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-
	Capabilities: [58] Debug port: BAR=1 offset=00a0
	Capabilities: [98] PCI Advanced Features
		AFCap: TP+ FLR+
		AFCtrl: FLR-
		AFStatus: TP-

00:1e.0 PCI bridge: Intel Corporation 82801 PCI Bridge (rev 92) (prog-if 01 [Subtractive decode])
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx-
	Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 0
	Bus: primary=00, secondary=06, subordinate=06, sec-latency=64
	I/O behind bridge: 00005000-00005fff
	Memory behind bridge: f8300000-f83fffff
	Prefetchable memory behind bridge: 00000000f0000000-00000000f7ffffff
	Secondary status: 66MHz- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort+ <SERR- <PERR-
	BridgeCtl: Parity- SERR- NoISA+ VGA+ MAbort- >Reset- FastB2B-
		PriDiscTmr- SecDiscTmr- DiscTmrStat- DiscTmrSERREn-
	Capabilities: [50] Subsystem: Super Micro Computer Inc Device de80

00:1f.0 ISA bridge: Intel Corporation 82801IR (ICH9R) LPC Interface Controller (rev 02)
	Subsystem: Super Micro Computer Inc Device de80
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
	Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 0
	Capabilities: [e0] Vendor Specific Information <?>

00:1f.2 IDE interface: Intel Corporation 82801IR/IO/IH (ICH9R/DO/DH) 4 port SATA IDE Controller (rev 02) (prog-if 8a [Master SecP PriP])
	Subsystem: Super Micro Computer Inc Device de80
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
	Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 0
	Interrupt: pin B routed to IRQ 17
	Region 0: I/O ports at 01f0 [size=8]
	Region 1: I/O ports at 03f4 [size=1]
	Region 2: I/O ports at 0170 [size=8]
	Region 3: I/O ports at 0374 [size=1]
	Region 4: I/O ports at 18d0 [size=16]
	Region 5: I/O ports at 18c0 [size=16]
	Capabilities: [70] Power Management version 3
		Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
		Status: D0 NoSoftRst+ PME-Enable- DSel=0 DScale=0 PME-
	Capabilities: [b0] PCI Advanced Features
		AFCap: TP+ FLR+
		AFCtrl: FLR-
		AFStatus: TP-
	Kernel driver in use: ata_piix

00:1f.3 SMBus: Intel Corporation 82801I (ICH9 Family) SMBus Controller (rev 02)
	Subsystem: Super Micro Computer Inc Device de80
	Control: I/O+ Mem+ BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx-
	Status: Cap- 66MHz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Interrupt: pin B routed to IRQ 10
	Region 0: Memory at f8600800 (64-bit, non-prefetchable) [size=256]
	Region 4: I/O ports at 1100 [size=32]

00:1f.5 IDE interface: Intel Corporation 82801I (ICH9 Family) 2 port SATA IDE Controller (rev 02) (prog-if 85 [Master SecO PriO])
	Subsystem: Super Micro Computer Inc Device de80
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
	Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 0
	Interrupt: pin C routed to IRQ 18
	Region 0: I/O ports at 1c28 [size=8]
	Region 1: I/O ports at 1c1c [size=4]
	Region 2: I/O ports at 1c20 [size=8]
	Region 3: I/O ports at 1c18 [size=4]
	Region 4: I/O ports at 18f0 [size=16]
	Region 5: I/O ports at 18e0 [size=16]
	Capabilities: [70] Power Management version 3
		Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
		Status: D0 NoSoftRst+ PME-Enable- DSel=0 DScale=0 PME-
	Capabilities: [b0] PCI Advanced Features
		AFCap: TP+ FLR+
		AFCtrl: FLR-
		AFStatus: TP-
	Kernel driver in use: ata_piix

01:00.0 Ethernet controller: Intel Corporation 82598EB 10-Gigabit AT CX4 Network Connection (rev 01)
	Subsystem: Super Micro Computer Inc Device af80
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx+
	Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 0, Cache Line Size: 32 bytes
	Interrupt: pin A routed to IRQ 16
	Region 0: Memory at f8040000 (32-bit, non-prefetchable) [size=128K]
	Region 1: Memory at f8000000 (32-bit, non-prefetchable) [size=256K]
	Region 2: I/O ports at 2000 [size=32]
	Region 3: Memory at f8060000 (32-bit, non-prefetchable) [size=16K]
	Capabilities: [40] Power Management version 3
		Flags: PMEClk- DSI+ D1- D2- AuxCurrent=0mA PME(D0+,D1-,D2-,D3hot+,D3cold-)
		Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=1 PME-
	Capabilities: [50] MSI: Enable- Count=1/1 Maskable- 64bit+
		Address: 0000000000000000  Data: 0000
	Capabilities: [60] MSI-X: Enable+ Count=18 Masked-
		Vector table: BAR=3 offset=00000000
		PBA: BAR=3 offset=00002000
	Capabilities: [a0] Express (v2) Endpoint, MSI 00
		DevCap:	MaxPayload 256 bytes, PhantFunc 0, Latency L0s <512ns, L1 <64us
			ExtTag- AttnBtn- AttnInd- PwrInd- RBE+ FLReset-
		DevCtl:	Report errors: Correctable- Non-Fatal- Fatal- Unsupported-
			RlxdOrd+ ExtTag- PhantFunc- AuxPwr- NoSnoop+
			MaxPayload 128 bytes, MaxReadReq 512 bytes
		DevSta:	CorrErr- UncorrErr- FatalErr- UnsuppReq- AuxPwr- TransPend-
		LnkCap:	Port #2, Speed 2.5GT/s, Width x8, ASPM L0s L1, Latency L0 <4us, L1 <64us
			ClockPM- Surprise- LLActRep- BwNot-
		LnkCtl:	ASPM Disabled; RCB 64 bytes Disabled- Retrain- CommClk+
			ExtSynch- ClockPM- AutWidDis- BWInt- AutBWInt-
		LnkSta:	Speed 2.5GT/s, Width x8, TrErr- Train- SlotClk+ DLActive- BWMgmt- ABWMgmt-
		DevCap2: Completion Timeout: Range ABCD, TimeoutDis+
		DevCtl2: Completion Timeout: 16ms to 55ms, TimeoutDis-
		LnkCtl2: Target Link Speed: 2.5GT/s, EnterCompliance- SpeedDis-, Selectable De-emphasis: -6dB
			 Transmit Margin: Normal Operating Range, EnterModifiedCompliance- ComplianceSOS-
			 Compliance De-emphasis: -6dB
		LnkSta2: Current De-emphasis Level: -6dB
	Kernel driver in use: ixgbe

04:00.0 Ethernet controller: Intel Corporation 82573E Gigabit Ethernet Controller (Copper) (rev 03)
	Subsystem: Super Micro Computer Inc Device 108c
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx+
	Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 0, Cache Line Size: 64 bytes
	Interrupt: pin A routed to IRQ 29
	Region 0: Memory at f8100000 (32-bit, non-prefetchable) [size=128K]
	Region 2: I/O ports at 3000 [size=32]
	Capabilities: [c8] Power Management version 2
		Flags: PMEClk- DSI+ D1- D2- AuxCurrent=0mA PME(D0+,D1-,D2-,D3hot+,D3cold+)
		Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=1 PME-
	Capabilities: [d0] MSI: Enable+ Count=1/1 Maskable- 64bit+
		Address: 00000000fee0300c  Data: 4162
	Capabilities: [e0] Express (v1) Endpoint, MSI 00
		DevCap:	MaxPayload 256 bytes, PhantFunc 0, Latency L0s <512ns, L1 <64us
			ExtTag- AttnBtn- AttnInd- PwrInd- RBE- FLReset-
		DevCtl:	Report errors: Correctable- Non-Fatal- Fatal- Unsupported-
			RlxdOrd+ ExtTag- PhantFunc- AuxPwr- NoSnoop+
			MaxPayload 128 bytes, MaxReadReq 512 bytes
		DevSta:	CorrErr- UncorrErr+ FatalErr- UnsuppReq+ AuxPwr+ TransPend+
		LnkCap:	Port #0, Speed 2.5GT/s, Width x1, ASPM unknown, Latency L0 <128ns, L1 <64us
			ClockPM- Surprise- LLActRep- BwNot-
		LnkCtl:	ASPM Disabled; RCB 64 bytes Disabled- Retrain- CommClk-
			ExtSynch- ClockPM- AutWidDis- BWInt- AutBWInt-
		LnkSta:	Speed 2.5GT/s, Width x1, TrErr- Train- SlotClk+ DLActive- BWMgmt- ABWMgmt-
	Kernel driver in use: e1000e

05:00.0 Ethernet controller: Intel Corporation 82573L Gigabit Ethernet Controller
	Subsystem: Super Micro Computer Inc Device 109a
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx+
	Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 0, Cache Line Size: 64 bytes
	Interrupt: pin A routed to IRQ 30
	Region 0: Memory at f8200000 (32-bit, non-prefetchable) [size=128K]
	Region 2: I/O ports at 4000 [size=32]
	Capabilities: [c8] Power Management version 2
		Flags: PMEClk- DSI+ D1- D2- AuxCurrent=0mA PME(D0+,D1-,D2-,D3hot+,D3cold+)
		Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=1 PME-
	Capabilities: [d0] MSI: Enable+ Count=1/1 Maskable- 64bit+
		Address: 00000000fee0c00c  Data: 4172
	Capabilities: [e0] Express (v1) Endpoint, MSI 00
		DevCap:	MaxPayload 256 bytes, PhantFunc 0, Latency L0s <512ns, L1 <64us
			ExtTag- AttnBtn- AttnInd- PwrInd- RBE- FLReset-
		DevCtl:	Report errors: Correctable- Non-Fatal- Fatal- Unsupported-
			RlxdOrd+ ExtTag- PhantFunc- AuxPwr- NoSnoop+
			MaxPayload 128 bytes, MaxReadReq 512 bytes
		DevSta:	CorrErr- UncorrErr+ FatalErr- UnsuppReq+ AuxPwr+ TransPend-
		LnkCap:	Port #0, Speed 2.5GT/s, Width x1, ASPM unknown, Latency L0 <128ns, L1 <64us
			ClockPM+ Surprise- LLActRep- BwNot-
		LnkCtl:	ASPM Disabled; RCB 64 bytes Disabled- Retrain- CommClk-
			ExtSynch- ClockPM- AutWidDis- BWInt- AutBWInt-
		LnkSta:	Speed 2.5GT/s, Width x1, TrErr- Train- SlotClk+ DLActive- BWMgmt- ABWMgmt-
	Kernel driver in use: e1000e

06:01.0 VGA compatible controller: ATI Technologies Inc ES1000 (rev 02) (prog-if 00 [VGA controller])
	Subsystem: Super Micro Computer Inc Device de80
	Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping+ SERR+ FastB2B+ DisINTx-
	Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
	Latency: 66 (2000ns min), Cache Line Size: 32 bytes
	Interrupt: pin A routed to IRQ 11
	Region 0: Memory at f0000000 (32-bit, prefetchable) [size=128M]
	Region 1: I/O ports at 5000 [size=256]
	Region 2: Memory at f8300000 (32-bit, non-prefetchable) [size=64K]
	[virtual] Expansion ROM at f8320000 [disabled] [size=128K]
	Capabilities: [50] Power Management version 2
		Flags: PMEClk- DSI- D1+ D2+ AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
		Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME-


[-- Attachment #3: ethtool-S-eth1.txt --]
[-- Type: text/plain, Size: 1257 bytes --]

NIC statistics:
     rx_packets: 4196289220
     tx_packets: 4485418409
     rx_bytes: 3072804779742
     tx_bytes: 4684778074029
     rx_broadcast: 15782
     tx_broadcast: 0
     rx_multicast: 0
     tx_multicast: 0
     rx_errors: 0
     tx_errors: 0
     tx_dropped: 0
     multicast: 0
     collisions: 0
     rx_length_errors: 0
     rx_over_errors: 0
     rx_crc_errors: 0
     rx_frame_errors: 0
     rx_no_buffer_count: 0
     rx_missed_errors: 7
     tx_aborted_errors: 0
     tx_carrier_errors: 0
     tx_fifo_errors: 0
     tx_heartbeat_errors: 0
     tx_window_errors: 0
     tx_abort_late_coll: 0
     tx_deferred_ok: 0
     tx_single_coll_ok: 0
     tx_multi_coll_ok: 0
     tx_timeout_count: 5
     tx_restart_queue: 114
     rx_long_length_errors: 0
     rx_short_length_errors: 0
     rx_align_errors: 0
     tx_tcp_seg_good: 3341840
     tx_tcp_seg_failed: 0
     rx_flow_control_xon: 0
     rx_flow_control_xoff: 0
     tx_flow_control_xon: 0
     tx_flow_control_xoff: 0
     rx_long_byte_count: 3072804779742
     rx_csum_offload_good: 4178709356
     rx_csum_offload_errors: 57534
     rx_header_split: 0
     alloc_rx_buff_failed: 0
     tx_smbus: 0
     rx_smbus: 0
     dropped_smbus: 0
     rx_dma_failed: 0
     tx_dma_failed: 0

^ permalink raw reply

* Re: eth1: Detected Hardware Unit Hang
From: Paweł Staszewski @ 2010-03-29 17:36 UTC (permalink / raw)
  To: Allan, Bruce W
  Cc: Linux Network Development list, e1000-devel@lists.sourceforge.net
In-Reply-To: <4BB0E394.2060908@itcare.pl>

W dniu 2010-03-29 19:29, Paweł Staszewski pisze:
> lspci -vvv + ethtool -S in attached files.
>
> Network traffic when i get this info:
> eth1:    RX:    157.22 Mb/s    TX:    379.27 Mb/s
>
> ethtool -i eth1
> driver: e1000e
> version: 1.0.2-k2
> firmware-version: 0.5-7
> bus-info: 0000:05:00.0
> This is: Intel Corporation 82573L Gigabit Ethernet Controller
>
>
> But in this server i have another gigabit interface:
> Intel Corporation 82573E Gigabit Ethernet Controller
> this interface has two times more traffic than eth0 (82573L)
> ethtool -i eth0
> driver: e1000e
> version: 1.0.2-k2
> firmware-version: 0.15-5
> bus-info: 0000:04:00.0
>
I forgot to add that i have no problems with (eth0) 82573E


> And also this server was working 4months without problems on 2.6.29.1 
> kernel
>
> Drivers that I use for e1000e are from kernel (standard kernel 
> build-in e1000e driver).
> I don't tried other drivers.
>
> This is production server so I can't make too much tests.
>
>
> W dniu 2010-03-29 18:41, Allan, Bruce W pisze:
>> [adding e1000-devel]
>>
>> Please provide more information:
>> * what NIC/LOM is this on (preferably send full output from lspci -vvv)
>> * what type of networking workload is running at the time the hang 
>> occurred
>> * a dump of the NIC/LOM statistics might also help (ethtool -S eth1)
>>
>> Have you tried the latest standalone e1000e driver on e1000.sf.net?  
>> Does it reproduce the issue?
>>
>> If we cannot reproduce the hang in-house, would you be able/willing 
>> to run a debug driver to gather more information?
>>
>> Thanks,
>> Bruce.
>>
>> -----Original Message-----
>> From: netdev-owner@vger.kernel.org 
>> [mailto:netdev-owner@vger.kernel.org] On Behalf Of Pawel Staszewski
>> Sent: Monday, March 29, 2010 8:34 AM
>> To: Linux Network Development list
>> Subject: eth1: Detected Hardware Unit Hang
>>
>> After update to kernel from 2.6.29.1 to 2.6.33.1 i have this info in 
>> dmesg:
>>
>> 0000:05:00.0: eth1: Detected Hardware Unit Hang:
>>     TDH<1e>
>>     TDT<a>
>>     next_to_use<a>
>>     next_to_clean<1d>
>> buffer_info[next_to_clean]:
>>     time_stamp<33bae15>
>>     next_to_watch<20>
>>     jiffies<33bafaf>
>>     next_to_watch.status<0>
>> MAC Status<80080783>
>> PHY Status<796d>
>> PHY 1000BASE-T Status<3800>
>> PHY Extended Status<3000>
>> PCI Status<10>
>> 0000:05:00.0: eth1: Detected Hardware Unit Hang:
>>     TDH<1e>
>>     TDT<a>
>>     next_to_use<a>
>>     next_to_clean<1d>
>> buffer_info[next_to_clean]:
>>     time_stamp<33bae15>
>>     next_to_watch<20>
>>     jiffies<33bb1a3>
>>     next_to_watch.status<0>
>> MAC Status<80080783>
>> PHY Status<796d>
>> PHY 1000BASE-T Status<3800>
>> PHY Extended Status<3000>
>> PCI Status<10>
>> 0000:05:00.0: eth1: Detected Hardware Unit Hang:
>>     TDH<1e>
>>     TDT<a>
>>     next_to_use<a>
>>     next_to_clean<1d>
>> buffer_info[next_to_clean]:
>>     time_stamp<33bae15>
>>     next_to_watch<20>
>>     jiffies<33bb397>
>>     next_to_watch.status<0>
>> MAC Status<80080783>
>> PHY Status<796d>
>> PHY 1000BASE-T Status<3800>
>> PHY Extended Status<3000>
>> PCI Status<10>
>> ------------[ cut here ]------------
>> WARNING: at net/sched/sch_generic.c:255 dev_watchdog+0x118/0x19c()
>> Hardware name: X7DCT
>> NETDEV WATCHDOG: eth1 (e1000e): transmit queue 0 timed out
>> Modules linked in:
>> Pid: 0, comm: swapper Not tainted 2.6.33.1 #2
>> Call Trace:
>>    [<c1024e3d>] ? warn_slowpath_common+0x52/0x71
>>    [<c1024e49>] ? warn_slowpath_common+0x5e/0x71
>>    [<c1024e8e>] ? warn_slowpath_fmt+0x26/0x2a
>>    [<c1261f54>] ? dev_watchdog+0x118/0x19c
>>    [<c102135c>] ? __wake_up+0x29/0x39
>>    [<c10320c6>] ? insert_work+0x40/0x44
>>    [<c1261e3c>] ? dev_watchdog+0x0/0x19c
>>    [<c102cc15>] ? run_timer_softirq+0x11a/0x173
>>    [<c1028e5b>] ? __do_softirq+0x74/0xdf
>>    [<c1028ee9>] ? do_softirq+0x23/0x27
>>    [<c10290be>] ? irq_exit+0x26/0x58
>>    [<c10102d7>] ? smp_apic_timer_interrupt+0x6c/0x76
>>    [<c12c5f9a>] ? apic_timer_interrupt+0x2a/0x30
>>    [<c1007e06>] ? mwait_idle+0x49/0x4e
>>    [<c10017e8>] ? cpu_idle+0x41/0x5a
>> ---[ end trace bcca9926a046332c ]---
>>
>>
>> With kernel 2.6.29.1 all was ok.
>> -- 
>> 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 0/3] sky2 minor driver updates
From: Stephen Hemminger @ 2010-03-29 17:36 UTC (permalink / raw)
  To: David S. Miller; +Cc: netdev

These are minor changes related to chip versions as described
in current Marvell driver.
-- 


^ permalink raw reply

* [PATCH 1/3] sky2: support Yukon EC_U rev B1 and later
From: Stephen Hemminger @ 2010-03-29 17:36 UTC (permalink / raw)
  To: David S. Miller; +Cc: netdev
In-Reply-To: <20100329173617.765470658@vyatta.com>

[-- Attachment #1: sky2-ec-u-b1.patch --]
[-- Type: text/plain, Size: 2155 bytes --]

Need to change logic to support later versions of Yukon 2 EC_U chip.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>

--- a/drivers/net/sky2.c	2010-03-29 08:48:16.490429444 -0700
+++ b/drivers/net/sky2.c	2010-03-29 09:27:36.485594685 -0700
@@ -876,6 +876,10 @@ static void sky2_mac_init(struct sky2_hw
 
 	if (hw->dev[port]->mtu > ETH_DATA_LEN)
 		reg |= GM_SMOD_JUMBO_ENA;
+
+	if (hw->chip_id == CHIP_ID_YUKON_EC_U &&
+	    hw->chip_rev == CHIP_REV_YU_EC_U_B1)
+		reg |= GM_NEW_FLOW_CTRL;
 
 	gma_write16(hw, port, GM_SERIAL_MODE, reg);
 
@@ -1413,8 +1417,7 @@ static void sky2_rx_start(struct sky2_po
 	/* These chips have no ram buffer?
 	 * MAC Rx RAM Read is controlled by hardware */
 	if (hw->chip_id == CHIP_ID_YUKON_EC_U &&
-	    (hw->chip_rev == CHIP_REV_YU_EC_U_A1 ||
-	     hw->chip_rev == CHIP_REV_YU_EC_U_B0))
+	    hw->chip_rev > CHIP_REV_YU_EC_U_A0)
 		sky2_write32(hw, Q_ADDR(rxq, Q_TEST), F_M_RX_RAM_DIS);
 
 	sky2_prefetch_init(hw, rxq, sky2->rx_le_map, RX_LE_SIZE - 1);
--- a/drivers/net/sky2.h	2010-03-29 08:52:02.480429332 -0700
+++ b/drivers/net/sky2.h	2010-03-29 09:25:32.122900699 -0700
@@ -557,6 +557,7 @@ enum yukon_ec_u_rev {
 	CHIP_REV_YU_EC_U_A0  = 1,
 	CHIP_REV_YU_EC_U_A1  = 2,
 	CHIP_REV_YU_EC_U_B0  = 3,
+	CHIP_REV_YU_EC_U_B1  = 5,
 };
 enum yukon_fe_rev {
 	CHIP_REV_YU_FE_A1    = 1,
@@ -1775,10 +1776,13 @@ enum {
 /*	GM_SERIAL_MODE			16 bit r/w	Serial Mode Register */
 enum {
 	GM_SMOD_DATABL_MSK	= 0x1f<<11, /* Bit 15..11:	Data Blinder (r/o) */
-	GM_SMOD_LIMIT_4		= 1<<10, /* Bit 10:	4 consecutive Tx trials */
-	GM_SMOD_VLAN_ENA	= 1<<9,	/* Bit  9:	Enable VLAN  (Max. Frame Len) */
-	GM_SMOD_JUMBO_ENA	= 1<<8,	/* Bit  8:	Enable Jumbo (Max. Frame Len) */
-	 GM_SMOD_IPG_MSK	= 0x1f	/* Bit 4..0:	Inter-Packet Gap (IPG) */
+	GM_SMOD_LIMIT_4		= 1<<10, /* 4 consecutive Tx trials */
+	GM_SMOD_VLAN_ENA	= 1<<9,	 /* Enable VLAN  (Max. Frame Len) */
+	GM_SMOD_JUMBO_ENA	= 1<<8,	 /* Enable Jumbo (Max. Frame Len) */
+
+	GM_NEW_FLOW_CTRL	= 1<<6,	 /* Enable New Flow-Control */
+
+	GM_SMOD_IPG_MSK		= 0x1f	 /* Bit 4..0:	Inter-Packet Gap (IPG) */
 };
 
 #define DATA_BLIND_VAL(x)	(((x)<<11) & GM_SMOD_DATABL_MSK)

-- 


^ permalink raw reply

* [PATCH 2/3] sky2: add XL revisions
From: Stephen Hemminger @ 2010-03-29 17:36 UTC (permalink / raw)
  To: David S. Miller; +Cc: netdev
In-Reply-To: <20100329173617.765470658@vyatta.com>

[-- Attachment #1: sky2-xl-rev.patch --]
[-- Type: text/plain, Size: 2288 bytes --]

Add definitions for Yukon XL revisions.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>

--- a/drivers/net/sky2.h	2010-03-29 09:25:32.122900699 -0700
+++ b/drivers/net/sky2.h	2010-03-29 09:27:47.582049092 -0700
@@ -548,6 +548,14 @@ enum {
 	CHIP_ID_YUKON_UL_2 = 0xba, /* YUKON-2 Ultra 2 */
 	CHIP_ID_YUKON_OPT  = 0xbc, /* YUKON-2 Optima */
 };
+
+enum yukon_xl_rev {
+	CHIP_REV_YU_XL_A0  = 0,
+	CHIP_REV_YU_XL_A1  = 1,
+	CHIP_REV_YU_XL_A2  = 2,
+	CHIP_REV_YU_XL_A3  = 3,
+};
+
 enum yukon_ec_rev {
 	CHIP_REV_YU_EC_A1    = 0,  /* Chip Rev. for Yukon-EC A1/A0 */
 	CHIP_REV_YU_EC_A2    = 1,  /* Chip Rev. for Yukon-EC A2 */
--- a/drivers/net/sky2.c	2010-03-29 09:27:44.992064516 -0700
+++ b/drivers/net/sky2.c	2010-03-29 09:27:47.582049092 -0700
@@ -226,7 +226,7 @@ static void sky2_power_on(struct sky2_hw
 	/* disable Core Clock Division, */
 	sky2_write32(hw, B2_Y2_CLK_CTRL, Y2_CLK_DIV_DIS);
 
-	if (hw->chip_id == CHIP_ID_YUKON_XL && hw->chip_rev > 1)
+	if (hw->chip_id == CHIP_ID_YUKON_XL && hw->chip_rev > CHIP_REV_YU_XL_A1)
 		/* enable bits are inverted */
 		sky2_write8(hw, B2_Y2_CLK_GATE,
 			    Y2_PCI_CLK_LNK1_DIS | Y2_COR_CLK_LNK1_DIS |
@@ -268,7 +268,7 @@ static void sky2_power_on(struct sky2_hw
 
 static void sky2_power_aux(struct sky2_hw *hw)
 {
-	if (hw->chip_id == CHIP_ID_YUKON_XL && hw->chip_rev > 1)
+	if (hw->chip_id == CHIP_ID_YUKON_XL && hw->chip_rev > CHIP_REV_YU_XL_A1)
 		sky2_write8(hw, B2_Y2_CLK_GATE, 0);
 	else
 		/* enable bits are inverted */
@@ -651,7 +651,7 @@ static void sky2_phy_power_up(struct sky
 	reg1 = sky2_pci_read32(hw, PCI_DEV_REG1);
 	reg1 &= ~phy_power[port];
 
-	if (hw->chip_id == CHIP_ID_YUKON_XL && hw->chip_rev > 1)
+	if (hw->chip_id == CHIP_ID_YUKON_XL && hw->chip_rev > CHIP_REV_YU_XL_A1)
 		reg1 |= coma_mode[port];
 
 	sky2_pci_write32(hw, PCI_DEV_REG1, reg1);
@@ -823,7 +823,9 @@ static void sky2_mac_init(struct sky2_hw
 
 	sky2_write8(hw, SK_REG(port, GMAC_CTRL), GMC_RST_CLR);
 
-	if (hw->chip_id == CHIP_ID_YUKON_XL && hw->chip_rev == 0 && port == 1) {
+	if (hw->chip_id == CHIP_ID_YUKON_XL &&
+	    hw->chip_rev == CHIP_REV_YU_XL_A0 &&
+	    port == 1) {
 		/* WA DEV_472 -- looks like crossed wires on port 2 */
 		/* clear GMAC 1 Control reset */
 		sky2_write8(hw, SK_REG(0, GMAC_CTRL), GMC_RST_CLR);

-- 


^ permalink raw reply

* [PATCH 3/3] sky2: avoid duplicate link up on Optima chip
From: Stephen Hemminger @ 2010-03-29 17:36 UTC (permalink / raw)
  To: David S. Miller; +Cc: netdev
In-Reply-To: <20100329173617.765470658@vyatta.com>

[-- Attachment #1: sky2-ul-link-up.patch --]
[-- Type: text/plain, Size: 597 bytes --]

The Optima version has feature to detect link quickly without PHY interrupt,
but it causes duplicate link up events.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>

--- a/drivers/net/sky2.c	2010-03-29 10:29:17.530956219 -0700
+++ b/drivers/net/sky2.c	2010-03-29 10:29:18.261894374 -0700
@@ -2146,7 +2146,8 @@ static void sky2_phy_intr(struct sky2_hw
 		   istatus, phystat);
 
 	if (istatus & PHY_M_IS_AN_COMPL) {
-		if (sky2_autoneg_done(sky2, phystat) == 0)
+		if (sky2_autoneg_done(sky2, phystat) == 0 &&
+		    !netif_carrier_ok(dev))
 			sky2_link_up(sky2);
 		goto out;
 	}

-- 


^ permalink raw reply

* Re: Network performance - iperf
From: Eric Dumazet @ 2010-03-29 17:45 UTC (permalink / raw)
  To: michal.simek
  Cc: LKML, John Williams, netdev, Grant Likely, John Linn,
	Steven J. Magnani, Arnd Bergmann, akpm
In-Reply-To: <4BB0C6E1.2090606@petalogix.com>

Le lundi 29 mars 2010 à 17:27 +0200, Michal Simek a écrit :
> Michal Simek wrote:
> > Eric Dumazet wrote:
> >> Le lundi 29 mars 2010 à 13:33 +0200, Michal Simek a écrit :
> >>
> >>> Do you have any idea howto improve TCP/UDP performance in general?
> >>> Or tests which can point me on weak places.
> >>
> >> Could you post "netstat -s" on your receiver, after fresh boot and your
> >> iperf session, for 32 MB and 256 MB ram case ?
> >>
> > 
> > I am not sure if is helpful but look below.
> > 
> Sorry I forget to c&p that second part. :-(
> 

Sorry, your netstat is not up2date.

If you cannot correct it to last version
[ net-tools 1.60 , netstat 1.42 ], please send

cat /proc/net/snmp
cat /proc/net/netstat

^ permalink raw reply

* Re: behavior of recvmmsg() on blocking sockets
From: Chris Friesen @ 2010-03-29 17:48 UTC (permalink / raw)
  To: Brandon Black; +Cc: Arnaldo Carvalho de Melo, linux-kernel, netdev
In-Reply-To: <84621a61003291024r38121763o546e0f09e2d63bc3@mail.gmail.com>

On 03/29/2010 11:24 AM, Brandon Black wrote:
> On Mon, Mar 29, 2010 at 11:18 AM, Chris Friesen <cfriesen@nortel.com> wrote:
>>
>> prev = current time
>> loop forever
>>        cur = current time
>>        timeout = max_latency - (cur - prev)
>>        recvmmsg(timeout)
>>        process all received messages
>>        prev = cur
>>
>>
>> Basically you determine the max latency you're willing to wait for a
>> packet to be handled, then subtract the amount of time you spent
>> processing messages from that and pass it into the recvmmsg() call as
>> the timeout.  That way no messages will be delayed for longer than the
>> max latency. (Not considering scheduling delays.)
> 
> With a blocking socket, you'd also need to set SO_RCVTIMEO on the
> underlying socket to some value that makes sense and is below your max
> latency, because recvmmsg()'s timeout argument only applies in-between
> underlying recvmsg() calls, not during them.

Hmm...that's a good point.  For some reason I had been under the
impression that the timeout affected the underlying recvmsg() calls as
well.  It think it would make more sense for the kernel to abort a
blocking recvmsg() call once the timeout expires.

As for spending a lot of time spinning if there are gaps in the input
stream...in the cases where the time-based usage makes sense the normal
situation is that there are a lot of packets coming in.  A 10gig
ethernet pipe can theoretically receive something like 19 packets per
usec.  Doesn't take much of a delay before you probably have packets
waiting.

Chris

^ permalink raw reply


This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox