* Re: [PATCH] ifb: add multi-queue support
From: Changli Gao @ 2009-11-13 9:38 UTC (permalink / raw)
To: Jarek Poplawski
Cc: Eric Dumazet, David S. Miller, Stephen Hemminger, Patrick McHardy,
Tom Herbert, netdev
In-Reply-To: <20091113091825.GA7449@ff.dom.local>
On Fri, Nov 13, 2009 at 5:18 PM, Jarek Poplawski <jarkao2@gmail.com> wrote:
> On Fri, Nov 13, 2009 at 04:54:50PM +0800, Changli Gao wrote:
>> On Fri, Nov 13, 2009 at 3:45 PM, Jarek Poplawski <jarkao2@gmail.com> wrote:
>>
>> I have done a simple test. I run a simple program on computer A, which
>> sends SYN packets with random source ports to Computer B's 80 port (No
>> socket listens on that port, so tcp reset packets will be sent) in
>> 90kpps. On computer B, I redirect the traffic to IFB. At the same
>> time, I ping from B to A to get the RTT between them. I can't see any
>> difference between the original IFB and my MQ version. They are both:
>>
>> CPU idle: 50%
>> Latency: 0.3-0.4ms, burst 2ms.
>>
>
> I'm mostly concerned with routers doing forwarding with 1Gb or 10Gb
> NICs (including multiqueue). Alas/happily I don't have such a problem,
> but can't help you with testing either.
>
Oh, :) . I know more than one companies use kernel threads to forward
packets, and there isn't explicit extra overhead at all. And as you
know, as throughput increases, NAPI will bind the NIC to a CPU, and
softirqd will be waked up to do the work, which should be done in
SoftIRQ context. At that time, there isn't any difference between my
approach and the current kernel's.
--
Regards，
Changli Gao(xiaosuo@gmail.com)
^ permalink raw reply
* [net-next-2.6 PATCH] ixgbe: Make queue pairs on single MSI-X interrupts
From: Jeff Kirsher @ 2009-11-13 9:50 UTC (permalink / raw)
To: davem; +Cc: netdev, gospo, Peter P Waskiewicz Jr, Jeff Kirsher
From: PJ Waskiewicz <peter.p.waskiewicz.jr@intel.com>
This patch pairs similar-numbered Rx and Tx queues onto a single
MSI-X vector. For example, Tx queue 0 and Rx queue 0's interrupt
with be ethX-RxTx-0. This allows for more efficient cleanup, since
fewer interrupts will be firing during device operation. It also
helps with a cleaner CPU affinity for IRQ affinity.
Signed-off-by: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_main.c | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index 0489286..884152d 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -3626,10 +3626,10 @@ static int ixgbe_set_interrupt_capability(struct ixgbe_adapter *adapter)
* It's easy to be greedy for MSI-X vectors, but it really
* doesn't do us much good if we have a lot more vectors
* than CPU's. So let's be conservative and only ask for
- * (roughly) twice the number of vectors as there are CPU's.
+ * (roughly) the same number of vectors as there are CPU's.
*/
v_budget = min(adapter->num_rx_queues + adapter->num_tx_queues,
- (int)(num_online_cpus() * 2)) + NON_Q_VECTORS;
+ (int)num_online_cpus()) + NON_Q_VECTORS;
/*
* At the same time, hardware can only support a maximum of
^ permalink raw reply related
* Re: [net-next-2.6 PATCH] net: fast consecutive name allocation
From: Octavian Purdila @ 2009-11-13 9:51 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: Eric Dumazet, netdev
In-Reply-To: <20091112222608.79d90e9e@nehalam>
On Friday 13 November 2009 08:26:08 you wrote:
> On Fri, 13 Nov 2009 07:12:35 +0100
>
> Eric Dumazet <eric.dumazet@gmail.com> wrote:
> > Octavian Purdila a écrit :
> > > On Friday 13 November 2009 07:01:14 you wrote:
> > >> This patch speeds up the network device name allocation for the case
> > >> where a significant number of devices of the same type are created
> > >> consecutively.
> > >>
> > >> Tests performed on a PPC750 @ 800Mhz machine with per device sysctl
> > >> and sysfs entries disabled:
> > >>
> > >> Without the patch With the patch
> > >>
> > >> real 0m 43.43s real 0m 0.49s
> > >> user 0m 0.00s user 0m 0.00s
> > >> sys 0m 43.43s sys 0m 0.48s
>
> No one has give a reasonable use case for this network device name
> explosion, what is the benchmark doing this nosense, and how do I
> get paid to do it...
>
For us the usecase is creating interfaces that get used by applications that
generate all sorts of traffic. This allows us to simulate realistic end user
traffic (e.g. coming from a full blown stack). That sounds reasonable to us :)
Also, I've seen other people reporting here to use more then 8000 interfaces.
> But I have to say no for another reason. You cause the kernel to choose
> a different name for the case where a device is deleted or renamed.
> The old code would find and fill the hole when a new device was added.
>
> Since this is a semantic ABI change, the kind that drives users nuts.
>
The intent was to keep the old behavior. When the device is deleted we stop
fast allocation and we resume it only after we go through the old code once
again. Did I miss something?
^ permalink raw reply
* Re: [PATCH] ifb: add multi-queue support
From: Jarek Poplawski @ 2009-11-13 9:57 UTC (permalink / raw)
To: Changli Gao
Cc: Eric Dumazet, David S. Miller, Stephen Hemminger, Patrick McHardy,
Tom Herbert, netdev
In-Reply-To: <412e6f7f0911130138td181935w36cab3119972753e@mail.gmail.com>
On Fri, Nov 13, 2009 at 05:38:56PM +0800, Changli Gao wrote:
> On Fri, Nov 13, 2009 at 5:18 PM, Jarek Poplawski <jarkao2@gmail.com> wrote:
> > I'm mostly concerned with routers doing forwarding with 1Gb or 10Gb
> > NICs (including multiqueue). Alas/happily I don't have such a problem,
> > but can't help you with testing either.
> >
>
> Oh, :) . I know more than one companies use kernel threads to forward
> packets, and there isn't explicit extra overhead at all. And as you
> know, as throughput increases, NAPI will bind the NIC to a CPU, and
> softirqd will be waked up to do the work, which should be done in
> SoftIRQ context. At that time, there isn't any difference between my
> approach and the current kernel's.
I'm not against your solution at all. It only needs more proof... You
seem to forget the main networking paths now are just softirq, and
it's probably for some reason. If kernel threads are good enough, it
seems we should do more such changes.
Jarek P.
^ permalink raw reply
* Re: [net-next-2.6 PATCH] net: fast consecutive name allocation
From: Octavian Purdila @ 2009-11-13 9:55 UTC (permalink / raw)
To: Eric Dumazet; +Cc: netdev
In-Reply-To: <4AFCF8D3.6090905@gmail.com>
On Friday 13 November 2009 08:12:35 you wrote:
> Octavian Purdila a écrit :
> > On Friday 13 November 2009 07:01:14 you wrote:
> >> This patch speeds up the network device name allocation for the case
> >> where a significant number of devices of the same type are created
> >> consecutively.
> >>
> >> Tests performed on a PPC750 @ 800Mhz machine with per device sysctl
> >> and sysfs entries disabled:
> >>
> >> Without the patch With the patch
> >>
> >> real 0m 43.43s real 0m 0.49s
> >> user 0m 0.00s user 0m 0.00s
> >> sys 0m 43.43s sys 0m 0.48s
> >
> > Oops, pasting root prompts (e.g. # modprobe ....) directly into the git
> > commit message is not a good idea :) Here it is again, with the full
> > commit message.
> >
> > [net-next-2.6 PATCH] net: fast consecutive name allocation
> >
> > This patch speeds up the network device name allocation for the case
> > where a significant number of devices of the same type are created
> > consecutively.
> >
> > Tests performed on a PPC750 @ 800Mhz machine with per device sysctl
> > and sysfs entries disabled:
> >
> > $ time insmod /lib/modules/dummy.ko numdummies=8000
> >
> > Without the patch With the patch
> >
> > real 0m 43.43s real 0m 0.49s
> > user 0m 0.00s user 0m 0.00s
> > sys 0m 43.43s sys 0m 0.48s
> >
> > Signed-off-by: Octavian Purdila <opurdila@ixiacom.com>
> > ---
>
> Honestly I dont like this bloat.
>
> Changing dummy.c is trivial, and you can allocate 100.000.000 dummies if
> you want now :)
>
Yep we can do that - actually we are doing exactly this in our drivers. But in
that way, you get to "bloat" every driver which needs this.
^ permalink raw reply
* [net-next-2.6 PATCH] allow access to sysfs_groups member
From: Kurt Van Dijck @ 2009-11-13 10:51 UTC (permalink / raw)
To: Stephen Hemminger, Oliver Hartkopp, Wolfgang Grandegger; +Cc: netdev
In a recent post, I asked for way to hold the uevent that adds a
net_device until a driver had added some extra sysfs files.
The use-case for such thing is a (PCMCIA) CAN card, that has 2 (CAN)
network chips. The driver could add a 'channel' sysfs file that
indicates the channel on the card, but this sysfs file is not present
yet when the uevent is generated.
This patch applied to a 2.6.30 kernel did allow to use non-standard
sysfs properties in udev rules (together with a match on device/driver).
I believe the above use-case is not limited to CAN, but any other
network type.
This patch allows adding sysfs attribute groups during netdevice registration.
The idea is that before the register_netdev call, several calls to
netdev_sysfs_add_group can be done.
The existing sysfs groups that are added in netdev_register_kobject use the
same function too.
I did not touch the number of possible groups (currently still 3).
Signed-off-by: Kurt Van Dijck <kurt.van.dijck@eia.be>
---
include/linux/netdevice.h | 2 ++
net/core/net-sysfs.c | 28 +++++++++++++++++++++++-----
2 files changed, 25 insertions(+), 5 deletions(-)
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 8380009..9d16687 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -1115,6 +1115,8 @@ extern int dev_open(struct net_device *dev);
extern int dev_close(struct net_device *dev);
extern void dev_disable_lro(struct net_device *dev);
extern int dev_queue_xmit(struct sk_buff *skb);
+extern int netdev_sysfs_add_group(struct net_device *net,
+ struct attribute_group *grp);
extern int register_netdevice(struct net_device *dev);
extern void unregister_netdevice(struct net_device *dev);
extern void free_netdev(struct net_device *dev);
diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c
index 753c420..2938fdc 100644
--- a/net/core/net-sysfs.c
+++ b/net/core/net-sysfs.c
@@ -527,27 +527,45 @@ void netdev_unregister_kobject(struct net_device * net)
device_del(dev);
}
+/* Add a sysfs group to the netdev groups */
+int netdev_sysfs_add_group(struct net_device *net,
+ struct attribute_group *grp)
+{
+ struct attribute_group **groups = net->sysfs_groups;
+ struct attribute_group **end;
+
+ /* end pointer, with room for null terminator */
+ end = &net->sysfs_groups[ARRAY_SIZE(net->sysfs_groups) - 1];
+ for (; groups < end; ++groups) {
+ if (!*groups) {
+ *groups = grp;
+ return 0;
+ }
+ }
+ return -ENOSPC;
+}
+EXPORT_SYMBOL(netdev_sysfs_add_group);
+
/* Create sysfs entries for network device. */
int netdev_register_kobject(struct net_device *net)
{
struct device *dev = &(net->dev);
- const struct attribute_group **groups = net->sysfs_groups;
dev->class = &net_class;
dev->platform_data = net;
- dev->groups = groups;
+ dev->groups = net->sysfs_groups;
dev_set_name(dev, "%s", net->name);
#ifdef CONFIG_SYSFS
- *groups++ = &netstat_group;
+ netdev_sysfs_add_group(net, &netstat_group);
#ifdef CONFIG_WIRELESS_EXT_SYSFS
if (net->ieee80211_ptr)
- *groups++ = &wireless_group;
+ netdev_sysfs_add_group(net, &wireless_group);
#ifdef CONFIG_WIRELESS_EXT
else if (net->wireless_handlers)
- *groups++ = &wireless_group;
+ netdev_sysfs_add_group(net, &wireless_group);
#endif
#endif
#endif /* CONFIG_SYSFS */
^ permalink raw reply related
* Re: Problem with VLANs and via-velocity driver
From: Séguier Régis @ 2009-11-13 10:36 UTC (permalink / raw)
To: Kevin Shanahan; +Cc: netdev
In-Reply-To: <4AFCFF67.3060802@trash.net>
Patrick McHardy a écrit :
> Kevin Shanahan wrote:
>
>> Hi,
>>
>> I've had some problems with getting a fairly simple (I thought) VLAN
>> configuration working with the on board Via NICs on my Via M700
>> board. Looks like as soon as a tagged VLAN interface is added, the
>> underlying "raw" (untagged) interface stops responding.
>>
>> ...
>>
>> A bit of searching found a few references to similar problems going
>> back a few years (2005, 2007). Sounded like there were some driver
>> issues, but it wasn't clear from the messages I found whether they
>> were believed to be fixed or not. I tried the same test using a
>> differnt NIC with the tg3 driver and there were no problems, so it
>> looks to me like it's still a via-velocity issue. Unfortunately I
>> don't have room to add NICs to this machine and need to use the on
>> board Via hardware.
>>
>
> There's some special-casing for VID 0 in velocity_init_cam_filter().
> Does "ip link add link eth0 type vlan id 0" make any difference?
> If not, does "ip link set eth0 promisc on"?
>
When we patch the driver to support multiple vlan, we decide to use the
vlan 0 to desactivate the vlan filtering.
via-velocity: fix vlan receipt
- vlans were using a single CAM register (see mac_set_vlan_cam)
- setting the address filtering registers for vlans is not
needed when there is no vlan
The non-tagged interface is filtered out as soon as a tagged
(!= 0) interface is created. Its traffic appears again when an
zero-tagged interface is created.
Tested on Via Epia SN (VT6130 chipset) with several vlans whose
tag was above or beyond 255.
Signed-off-by: Séguier Régis <rseguier@e-teleport.net>
Acked-by: Francois Romieu <romieu@fr.zoreil.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
--
Régis Séguier
^ permalink raw reply
* Re: Problem with VLANs and via-velocity driver
From: Patrick McHardy @ 2009-11-13 11:11 UTC (permalink / raw)
To: technique; +Cc: Kevin Shanahan, netdev
In-Reply-To: <4AFD36C0.3090302@e-teleport.net>
Séguier Régis wrote:
> Patrick McHardy a écrit :
>> Kevin Shanahan wrote:
>>
>>> I've had some problems with getting a fairly simple (I thought) VLAN
>>> configuration working with the on board Via NICs on my Via M700
>>> board. Looks like as soon as a tagged VLAN interface is added, the
>>> underlying "raw" (untagged) interface stops responding.
>>>
>>>
>> There's some special-casing for VID 0 in velocity_init_cam_filter().
>> Does "ip link add link eth0 type vlan id 0" make any difference?
>> If not, does "ip link set eth0 promisc on"?
>>
> When we patch the driver to support multiple vlan, we decide to use the
> vlan 0 to desactivate the vlan filtering.
Why? The usual behaviour is to receive both tagged and untagged
frames.
> via-velocity: fix vlan receipt
>
> - vlans were using a single CAM register (see mac_set_vlan_cam)
> - setting the address filtering registers for vlans is not
> needed when there is no vlan
>
> The non-tagged interface is filtered out as soon as a tagged
> (!= 0) interface is created. Its traffic appears again when an
> zero-tagged interface is created.
VID 0 doesn't mean untagged, it means the tag contains only a priority
value.
^ permalink raw reply
* Re: [PATCH] ifb: add multi-queue support
From: Changli Gao @ 2009-11-13 11:25 UTC (permalink / raw)
To: Jarek Poplawski
Cc: Eric Dumazet, David S. Miller, Stephen Hemminger, Patrick McHardy,
Tom Herbert, netdev
In-Reply-To: <20091113095731.GA7749@ff.dom.local>
On Fri, Nov 13, 2009 at 5:57 PM, Jarek Poplawski <jarkao2@gmail.com> wrote:
>
> I'm not against your solution at all. It only needs more proof...
I know. :)
> You seem to forget the main networking paths now are just softirq, and
> it's probably for some reason. If kernel threads are good enough, it
> seems we should do more such changes.
>
I find there is still a softirqd kernel thread for each online CPU,
and these threads will be waked up if there are more SoftIRQs need to
be done after restarting the SoftIRQ processing many times, to keep
the whole system responsible. Did I miss sth.? And at the other side,
real time branch just wants to make all the activities based on kernel
threads, even ISR.
--
Regards，
Changli Gao(xiaosuo@gmail.com)
^ permalink raw reply
* Re: Problem with VLANs and via-velocity driver
From: Séguier Régis @ 2009-11-13 11:40 UTC (permalink / raw)
To: Patrick McHardy; +Cc: Kevin Shanahan, netdev
In-Reply-To: <4AFD3ED6.2040606@trash.net>
Patrick McHardy a écrit :
>
>>> There's some special-casing for VID 0 in velocity_init_cam_filter().
>>> Does "ip link add link eth0 type vlan id 0" make any difference?
>>> If not, does "ip link set eth0 promisc on"?
>>>
>>>
>> When we patch the driver to support multiple vlan, we decide to use the
>> vlan 0 to desactivate the vlan filtering.
>>
>
> Why? The usual behaviour is to receive both tagged and untagged
> frames.
>
We decide this with françois as an alternative to desactivate the NIC
vlan filtering instead of creating a modul parameter.
The vlan id 0 need to be created but not necessary activate.
In september 2008, we had a discution about exactly the same problem.
You speak about the possibility of an ALLVLAN device flag.
I think i proposed a patch to desactivate NIC vlan filtering when we are
in promisc mode but not used.
--
Régis Séguier
^ permalink raw reply
* Re: [patch] Fix: 'return -ENOMEM' instead of 'return ENOMEM'
From: Joel Becker @ 2009-11-13 11:56 UTC (permalink / raw)
To: Ingo Molnar
Cc: Randy Dunlap, Stephen M. Cameron, Mike Christie, David Airlie,
James Bottomley, Jens Axboe, Evgeniy Polyakov, iss_storagedev,
Eric Dumazet, Andy Whitcroft, Dave Airlie, Hannes Eder, dri-devel,
Alexey Dobriyan, Mike Miller, Mark Fasheh, Karsten Keil, rostedt,
Karen Xie, James E.J. Bottomley, Hannes Reinecke, Andreas
In-Reply-To: <20091113075306.GB2775@elte.hu>
On Fri, Nov 13, 2009 at 08:53:06AM +0100, Ingo Molnar wrote:
> > Sorry, I read the code wrong. This function is just a handler. The
> > caller, dlm_send_begin_reco_message(), expects the positive EAGAIN as
> > a non-error case.
>
> Well, at minimum the error code usage is very confused. The
> dlm_begin_reco_handler gets registered via o2net_register_handler().
> Here's where dlm_begin_reco_handler gets registered, followed by
> dlm_finalize_reco_handler right afterwards:
<snip>
> A negative error code right there. The other event handlers are seem to
> be similar - dlm_begin_reco_handler() is the odd one out.
The usage of the handlers - 'status' is the return code of the
handler - is pretty straightforward. Once you're in the mindset of the
o2net code, of course. This is old stuff, from the earliest days of
ocfs2 development, so we haven't had to spelunk in here for a while ;-)
Sunil and I both thought while looking at this code that we
could probably return -EAGAIN just as easily, though. We'll probably
clean it that way at some point.
Joel
--
There are morethings in heaven and earth, Horatio,
Than are dreamt of in your philosophy.
Joel Becker
Principal Software Developer
Oracle
E-mail: joel.becker@oracle.com
Phone: (650) 506-8127
^ permalink raw reply
* Re: [PATCH 11/13] pcmcia: pcmcia_request_window() doesn't need a pointer to a pointer
From: Karsten Keil @ 2009-11-13 12:21 UTC (permalink / raw)
To: Dominik Brodowski
Cc: linux-pcmcia, netdev, linux-wireless, linux-scsi, Jiri Kosina
In-Reply-To: <1257971899-9274-11-git-send-email-linux@dominikbrodowski.net>
On Mittwoch, 11. November 2009 21:38:17 Dominik Brodowski wrote:
> pcmcia_request_window() only needs a pointer to struct pcmcia_device, not
> a pointer to a pointer.
>
> CC: netdev@vger.kernel.org
> CC: linux-wireless@vger.kernel.org
> CC: linux-scsi@vger.kernel.org
> CC: Jiri Kosina <jkosina@suse.cz>
> CC: Karsten Keil <isdn@linux-pingi.de>
> Signed-off-by: Dominik Brodowski <linux@dominikbrodowski.net>
> ---
> drivers/char/pcmcia/ipwireless/main.c | 4 ++--
> drivers/isdn/hisax/sedlbauer_cs.c | 2 +-
For ISDN
Acked-by: Karsten Keil <keil@b1-systems.de>
^ permalink raw reply
* Re: [PATCH 06/13] pcmcia: Pass struct pcmcia_device to pcmcia_map_mem_page()
From: Karsten Keil @ 2009-11-13 12:22 UTC (permalink / raw)
To: Dominik Brodowski
Cc: linux-pcmcia, Magnus Damm, netdev, linux-wireless, linux-scsi,
Jiri Kosina
In-Reply-To: <1257971899-9274-6-git-send-email-linux@dominikbrodowski.net>
On Mittwoch, 11. November 2009 21:38:12 Dominik Brodowski wrote:
> From: Magnus Damm <damm@opensource.se>
>
> No logic changes, just pass struct pcmcia_device to pcmcia_map_mem_page()
>
> [linux@dominikbrodowski.net: update to 2.6.31]
> CC: netdev@vger.kernel.org
> CC: linux-wireless@vger.kernel.org
> CC: linux-scsi@vger.kernel.org
> CC: Jiri Kosina <jkosina@suse.cz>
> CC: Karsten Keil <isdn@linux-pingi.de>
> Signed-off-by: Magnus Damm <damm@opensource.se>
> Signed-off-by: Dominik Brodowski <linux@dominikbrodowski.net>
> ---
> drivers/char/pcmcia/ipwireless/main.c | 4 ++--
> drivers/isdn/hisax/sedlbauer_cs.c | 2 +-
For ISDN
Acked-by: Karsten Keil <keil@b1-systems.de>
^ permalink raw reply
* [PATCH] forcedeth: mac address fix
From: Stanislav O. Bezzubtsev @ 2009-11-13 12:22 UTC (permalink / raw)
Cc: davem, aabdulla, yinghai, netdev, linux-kernel,
Stanislav O. Bezzubtsev
Set second bit of randomly generated mac. That marks MAC
as locally assigned. IEEE802.3, section one, 3.2.3.
Signed-off-by: Stanislav O. Bezzubtsev <stas@lvk.cs.msu.su>
---
drivers/net/forcedeth.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/net/forcedeth.c b/drivers/net/forcedeth.c
index e1da466..b404c7a 100644
--- a/drivers/net/forcedeth.c
+++ b/drivers/net/forcedeth.c
@@ -5821,7 +5821,7 @@ static int __devinit nv_probe(struct pci_dev *pci_dev, const struct pci_device_i
dev->dev_addr);
dev_printk(KERN_ERR, &pci_dev->dev,
"Please complain to your hardware vendor. Switching to a random MAC.\n");
- dev->dev_addr[0] = 0x00;
+ dev->dev_addr[0] = 0x02; /* set local assignment bit */
dev->dev_addr[1] = 0x00;
dev->dev_addr[2] = 0x6c;
get_random_bytes(&dev->dev_addr[3], 3);
--
1.6.5
^ permalink raw reply related
* Re: [PATCH] ifb: add multi-queue support
From: Jarek Poplawski @ 2009-11-13 12:32 UTC (permalink / raw)
To: Changli Gao
Cc: Eric Dumazet, David S. Miller, Stephen Hemminger, Patrick McHardy,
Tom Herbert, netdev
In-Reply-To: <412e6f7f0911130325r20e2fbbbg21e2e384de1f7c0f@mail.gmail.com>
On Fri, Nov 13, 2009 at 07:25:33PM +0800, Changli Gao wrote:
> I find there is still a softirqd kernel thread for each online CPU,
> and these threads will be waked up if there are more SoftIRQs need to
> be done after restarting the SoftIRQ processing many times, to keep
> the whole system responsible. Did I miss sth.? And at the other side,
> real time branch just wants to make all the activities based on kernel
> threads, even ISR.
I wish them well! But again: for some strange reason network admins,
who sometimes care to complain here, rarely (very rarely) mention this
RT branch... (Actually, I recall only one case, and it probably wasn't
even real admin ;-)
Jarek P.
^ permalink raw reply
* Re: Enable syn cookies by default
From: Olaf van der Spek @ 2009-11-13 12:42 UTC (permalink / raw)
To: David Miller; +Cc: netdev
In-Reply-To: <20091021.060443.193726483.davem@davemloft.net>
On Wed, Oct 21, 2009 at 2:04 PM, David Miller <davem@davemloft.net> wrote:
>> Anybody?
>
> Would please you be patient?
>
> In case you haven't fucking noticed, all of the major kernel
> developers are in Japan at the annual kernel summit and the Japan
> Linux Symposium since late last week.
>
> So nobody has the time to look into anything requiring real
> long thinking like this issue does.
Hi David,
Have you had a chance to do some real long thinking? ;)
Greetings,
Olaf
^ permalink raw reply
* Re: [PATCH] ifb: add multi-queue support
From: Eric Dumazet @ 2009-11-13 13:10 UTC (permalink / raw)
To: Changli Gao
Cc: Jarek Poplawski, David S. Miller, Stephen Hemminger,
Patrick McHardy, Tom Herbert, netdev
In-Reply-To: <412e6f7f0911130325r20e2fbbbg21e2e384de1f7c0f@mail.gmail.com>
Changli Gao a écrit :
> On Fri, Nov 13, 2009 at 5:57 PM, Jarek Poplawski <jarkao2@gmail.com> wrote:
>> I'm not against your solution at all. It only needs more proof...
>
> I know. :)
>
>> You seem to forget the main networking paths now are just softirq, and
>> it's probably for some reason. If kernel threads are good enough, it
>> seems we should do more such changes.
>>
>
> I find there is still a softirqd kernel thread for each online CPU,
> and these threads will be waked up if there are more SoftIRQs need to
> be done after restarting the SoftIRQ processing many times, to keep
> the whole system responsible. Did I miss sth.? And at the other side,
> real time branch just wants to make all the activities based on kernel
> threads, even ISR.
>
>
You seem to focus on stress loads, where we enter ksoftirqd more to get
more throughput (and cpu fairness), at latency expense.
But in normal situations, ksoftirq is not started at all.
^ permalink raw reply
* Re: [PATCH] ifb: add multi-queue support
From: Eric Dumazet @ 2009-11-13 13:55 UTC (permalink / raw)
To: Changli Gao
Cc: Jarek Poplawski, David S. Miller, Stephen Hemminger,
Patrick McHardy, Tom Herbert, netdev
In-Reply-To: <412e6f7f0911130054i7a508a6ah16368f11bdc7353d@mail.gmail.com>
Changli Gao a écrit :
> I have done a simple test. I run a simple program on computer A, which
> sends SYN packets with random source ports to Computer B's 80 port (No
> socket listens on that port, so tcp reset packets will be sent) in
> 90kpps. On computer B, I redirect the traffic to IFB. At the same
> time, I ping from B to A to get the RTT between them. I can't see any
> difference between the original IFB and my MQ version. They are both:
>
> CPU idle: 50%
> Latency: 0.3-0.4ms, burst 2ms.
Yes, but redo your test with normal load (including several user threads
fighting for scheduler/cpu use) :)
softirq preempts user threads instantly (they are interrupts after all :) ),
while workqueue do compete with other workqueues/threads.
Tom Herbert RPS solution is still using softirqs.
I like to test workqueue based solution with IFB, but some people might still
keep previous pure softirq solution for production environment, not benchmarks...
^ permalink raw reply
* Re: [net-next-2.6 PATCH v6 5/7 RFC] TCPCT part 1e: implement socket option TCP_COOKIE_TRANSACTIONS
From: Andi Kleen @ 2009-11-13 14:11 UTC (permalink / raw)
To: William Allen Simpson; +Cc: Linux Kernel Network Developers
In-Reply-To: <4AFCEA59.3060707@gmail.com>
William Allen Simpson <william.allen.simpson@gmail.com> writes:
> This is a straightforward re-implementation of an earlier (year-old)
> patch that no longer applies cleanly, with permission of the original
> author (Adam Langley):
>
> http://thread.gmane.org/gmane.linux.network/102586
>
> The principle difference is using a TCP option to carry the cookie nonce,
> instead of a user configured offset in the data.
>
> Allocations have been rearranged to avoid requiring GFP_ATOMIC.
>
> Requires:
> net: TCP_MSS_DEFAULT, TCP_MSS_DESIRED
> TCPCT part 1c: sysctl_tcp_cookie_size, socket option TCP_COOKIE_TRANSACTIONS
> TCPCT part 1d: define TCP cookie option, extend existing struct's
BTW if you add new socket options you should always submit a manpage
patch to linux-man@vger.kernel.org too
-Andi
--
ak@linux.intel.com -- Speaking for myself only.
^ permalink raw reply
* Re: [net-next-2.6 PATCH v6 2/7 RFC] TCPCT part 1b: generate Responder Cookie
From: William Allen Simpson @ 2009-11-13 14:35 UTC (permalink / raw)
To: Eric Dumazet; +Cc: Linux Kernel Network Developers
In-Reply-To: <4AFCFAD0.4010701@gmail.com>
Eric Dumazet wrote:
> Small point :
>
> + if (unlikely(time_after_eq(jiffy, tcp_secret_generating->expires))) {
> + spin_lock_bh(&tcp_secret_locker);
> + if (!time_after_eq(jiffy, tcp_secret_generating->expires)) {
> + /* refreshed by another */
> + spin_unlock_bh(&tcp_secret_locker);
> + memcpy(bakery,
> + &tcp_secret_generating->secrets[0],
> + sizeof(tcp_secret_generating->secrets));
>
> Technically speaking, you should perform the memcpy() before spin_unlock_bh()
>
> + if (unlikely(time_after_eq(jiffy, tcp_secret_generating->expires))) {
> + spin_lock_bh(&tcp_secret_locker);
> + if (!time_after_eq(jiffy, tcp_secret_generating->expires)) {
> + /* refreshed by another */
> + memcpy(bakery,
> + &tcp_secret_generating->secrets[0],
> + sizeof(tcp_secret_generating->secrets));
> + spin_unlock_bh(&tcp_secret_locker);
>
Technically speaking is exactly what I need! I'd thought from your
earlier description of memory barriers that it should be after the unlock.
Done.
^ permalink raw reply
* Re: [net-next-2.6 PATCH v6 2/7 RFC] TCPCT part 1b: generate Responder Cookie
From: William Allen Simpson @ 2009-11-13 14:51 UTC (permalink / raw)
To: Joe Perches
Cc: Linux Kernel Network Developers, Eric Dumazet, Paul E. McKenney
In-Reply-To: <1258093594.16857.43.camel@Joe-Laptop.home>
[-- Attachment #1: Type: text/plain, Size: 411 bytes --]
Joe Perches wrote:
> Rather than using stack, could you use
> u32 *secrets = &tcp_secret_secondary->secrets[0];
>
That was because I thought it needed to be copied after the unlock.
Since Eric says it can be inside the locked path, I can do away with
that entirely, and just use the result (bakery) pointer itself. Saves a
memcpy, too.
Here's my revised attempt (untested). Any other technical corrections?
[-- Attachment #2: TCPCT+1b6+.patch --]
[-- Type: text/plain, Size: 6856 bytes --]
diff --git a/include/linux/cryptohash.h b/include/linux/cryptohash.h
index c118b2a..ec78a4b 100644
--- a/include/linux/cryptohash.h
+++ b/include/linux/cryptohash.h
@@ -2,6 +2,7 @@
#define __CRYPTOHASH_H
#define SHA_DIGEST_WORDS 5
+#define SHA_MESSAGE_BYTES (512 /*bits*/ / 8)
#define SHA_WORKSPACE_WORDS 80
void sha_init(__u32 *buf);
diff --git a/include/net/tcp.h b/include/net/tcp.h
index ec183fd..4a99a8e 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -1478,6 +1478,14 @@ struct tcp_request_sock_ops {
#endif
};
+/* Using SHA1 for now, define some constants.
+ */
+#define COOKIE_DIGEST_WORDS (SHA_DIGEST_WORDS)
+#define COOKIE_MESSAGE_WORDS (SHA_MESSAGE_BYTES / 4)
+#define COOKIE_WORKSPACE_WORDS (COOKIE_DIGEST_WORDS + COOKIE_MESSAGE_WORDS)
+
+extern int tcp_cookie_generator(u32 *bakery);
+
extern void tcp_v4_init(void);
extern void tcp_init(void);
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index e0cfa63..3ae01bf 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -264,6 +264,7 @@
#include <linux/cache.h>
#include <linux/err.h>
#include <linux/crypto.h>
+#include <linux/time.h>
#include <net/icmp.h>
#include <net/tcp.h>
@@ -2842,6 +2843,136 @@ EXPORT_SYMBOL(tcp_md5_hash_key);
#endif
+/**
+ * Each Responder maintains up to two secret values concurrently for
+ * efficient secret rollover. Each secret value has 4 states:
+ *
+ * Generating. (tcp_secret_generating != tcp_secret_primary)
+ * Generates new Responder-Cookies, but not yet used for primary
+ * verification. This is a short-term state, typically lasting only
+ * one round trip time (RTT).
+ *
+ * Primary. (tcp_secret_generating == tcp_secret_primary)
+ * Used both for generation and primary verification.
+ *
+ * Retiring. (tcp_secret_retiring != tcp_secret_secondary)
+ * Used for verification, until the first failure that can be
+ * verified by the newer Generating secret. At that time, this
+ * cookie's state is changed to Secondary, and the Generating
+ * cookie's state is changed to Primary. This is a short-term state,
+ * typically lasting only one round trip time (RTT).
+ *
+ * Secondary. (tcp_secret_retiring == tcp_secret_secondary)
+ * Used for secondary verification, after primary verification
+ * failures. This state lasts no more than twice the Maximum Segment
+ * Lifetime (2MSL). Then, the secret is discarded.
+ */
+struct tcp_cookie_secret {
+ /* The secret is divided into two parts. The digest part is the
+ * equivalent of previously hashing a secret and saving the state,
+ * and serves as an initialization vector (IV). The message part
+ * serves as the trailing secret.
+ */
+ u32 secrets[COOKIE_WORKSPACE_WORDS];
+ unsigned long expires;
+};
+
+#define TCP_SECRET_1MSL (HZ * TCP_PAWS_MSL)
+#define TCP_SECRET_2MSL (HZ * TCP_PAWS_MSL * 2)
+#define TCP_SECRET_LIFE (HZ * 600)
+
+static struct tcp_cookie_secret tcp_secret_one;
+static struct tcp_cookie_secret tcp_secret_two;
+
+/* Essentially a circular list, without dynamic allocation. */
+static struct tcp_cookie_secret *tcp_secret_generating;
+static struct tcp_cookie_secret *tcp_secret_primary;
+static struct tcp_cookie_secret *tcp_secret_retiring;
+static struct tcp_cookie_secret *tcp_secret_secondary;
+
+static DEFINE_SPINLOCK(tcp_secret_locker);
+
+/* Select a pseudo-random word in the cookie workspace.
+ */
+static inline u32 tcp_cookie_work(const u32 *ws, const int n)
+{
+ return ws[COOKIE_DIGEST_WORDS + ((COOKIE_MESSAGE_WORDS-1) & ws[n])];
+}
+
+/* Fill bakery[COOKIE_WORKSPACE_WORDS] with generator, updating as needed.
+ * Called in softirq context.
+ * Returns: 0 for success.
+ */
+int tcp_cookie_generator(u32 *bakery)
+{
+ unsigned long jiffy = jiffies;
+
+ if (unlikely(time_after_eq(jiffy, tcp_secret_generating->expires))) {
+ spin_lock_bh(&tcp_secret_locker);
+ if (!time_after_eq(jiffy, tcp_secret_generating->expires)) {
+ /* refreshed by another */
+ memcpy(bakery,
+ &tcp_secret_generating->secrets[0],
+ COOKIE_WORKSPACE_WORDS);
+ spin_unlock_bh(&tcp_secret_locker);
+ } else {
+ /* still needs refreshing */
+ get_random_bytes(bakery, COOKIE_WORKSPACE_WORDS);
+
+ /* The first time, paranoia assumes that the
+ * randomization function isn't as strong. But,
+ * this secret initialization is delayed until
+ * the last possible moment (packet arrival).
+ * Although that time is observable, it is
+ * unpredictably variable. Mash in the most
+ * volatile clock bits available, and expire the
+ * secret extra quickly.
+ */
+ if (unlikely(tcp_secret_primary->expires ==
+ tcp_secret_secondary->expires)) {
+ struct timespec tv;
+
+ getnstimeofday(&tv);
+ bakery[COOKIE_DIGEST_WORDS+0] ^=
+ (u32)tv.tv_nsec;
+
+ tcp_secret_secondary->expires = jiffy
+ + TCP_SECRET_1MSL
+ + (0x0f & tcp_cookie_work(bakery, 0));
+ } else {
+ tcp_secret_secondary->expires = jiffy
+ + TCP_SECRET_LIFE
+ + (0xff & tcp_cookie_work(bakery, 1));
+ tcp_secret_primary->expires = jiffy
+ + TCP_SECRET_2MSL
+ + (0x1f & tcp_cookie_work(bakery, 2));
+ }
+ memcpy(&tcp_secret_generating->secrets[0],
+ bakery, COOKIE_WORKSPACE_WORDS);
+
+ rcu_assign_pointer(tcp_secret_generating,
+ tcp_secret_secondary);
+ rcu_assign_pointer(tcp_secret_retiring,
+ tcp_secret_primary);
+ spin_unlock_bh(&tcp_secret_locker);
+ /*
+ * Neither call_rcu() or synchronize_rcu() are needed.
+ * Retiring data is not freed. It is replaced after
+ * further (locked) pointer updates, and a quiet time
+ * (minimum 1MSL, maximum LIFE - 2MSL).
+ */
+ }
+ } else {
+ rcu_read_lock_bh();
+ memcpy(bakery,
+ &rcu_dereference(tcp_secret_generating)->secrets[0],
+ COOKIE_WORKSPACE_WORDS);
+ rcu_read_unlock_bh();
+ }
+ return 0;
+}
+EXPORT_SYMBOL(tcp_cookie_generator);
+
void tcp_done(struct sock *sk)
{
if (sk->sk_state == TCP_SYN_SENT || sk->sk_state == TCP_SYN_RECV)
@@ -2876,6 +3012,7 @@ void __init tcp_init(void)
struct sk_buff *skb = NULL;
unsigned long nr_pages, limit;
int order, i, max_share;
+ unsigned long jiffy = jiffies;
BUILD_BUG_ON(sizeof(struct tcp_skb_cb) > sizeof(skb->cb));
@@ -2969,6 +3106,15 @@ void __init tcp_init(void)
tcp_hashinfo.ehash_mask + 1, tcp_hashinfo.bhash_size);
tcp_register_congestion_control(&tcp_reno);
+
+ memset(&tcp_secret_one.secrets[0], 0, sizeof(tcp_secret_one.secrets));
+ memset(&tcp_secret_two.secrets[0], 0, sizeof(tcp_secret_two.secrets));
+ tcp_secret_one.expires = jiffy; /* past due */
+ tcp_secret_two.expires = jiffy; /* past due */
+ tcp_secret_generating = &tcp_secret_one;
+ tcp_secret_primary = &tcp_secret_one;
+ tcp_secret_retiring = &tcp_secret_two;
+ tcp_secret_secondary = &tcp_secret_two;
}
EXPORT_SYMBOL(tcp_close);
--
1.6.3.3
^ permalink raw reply related
* [PATCH net-next 1/2] Phonet: put protocols array under RCU
From: Rémi Denis-Courmont @ 2009-11-13 15:01 UTC (permalink / raw)
To: netdev; +Cc: Rémi Denis-Courmont
From: Rémi Denis-Courmont <remi.denis-courmont@nokia.com>
Signed-off-by: Rémi Denis-Courmont <remi.denis-courmont@nokia.com>
---
net/phonet/af_phonet.c | 20 +++++++++++---------
1 files changed, 11 insertions(+), 9 deletions(-)
diff --git a/net/phonet/af_phonet.c b/net/phonet/af_phonet.c
index 8d3a55b..ed65da2 100644
--- a/net/phonet/af_phonet.c
+++ b/net/phonet/af_phonet.c
@@ -35,7 +35,6 @@
/* Transport protocol registration */
static struct phonet_protocol *proto_tab[PHONET_NPROTO] __read_mostly;
-static DEFINE_SPINLOCK(proto_tab_lock);
static struct phonet_protocol *phonet_proto_get(int protocol)
{
@@ -44,11 +43,11 @@ static struct phonet_protocol *phonet_proto_get(int protocol)
if (protocol >= PHONET_NPROTO)
return NULL;
- spin_lock(&proto_tab_lock);
+ rcu_read_lock();
pp = proto_tab[protocol];
if (pp && !try_module_get(pp->prot->owner))
pp = NULL;
- spin_unlock(&proto_tab_lock);
+ rcu_read_unlock();
return pp;
}
@@ -439,6 +438,8 @@ static struct packet_type phonet_packet_type __read_mostly = {
.func = phonet_rcv,
};
+static DEFINE_MUTEX(proto_tab_lock);
+
int __init_or_module phonet_proto_register(int protocol,
struct phonet_protocol *pp)
{
@@ -451,12 +452,12 @@ int __init_or_module phonet_proto_register(int protocol,
if (err)
return err;
- spin_lock(&proto_tab_lock);
+ mutex_lock(&proto_tab_lock);
if (proto_tab[protocol])
err = -EBUSY;
else
- proto_tab[protocol] = pp;
- spin_unlock(&proto_tab_lock);
+ rcu_assign_pointer(proto_tab[protocol], pp);
+ mutex_unlock(&proto_tab_lock);
return err;
}
@@ -464,10 +465,11 @@ EXPORT_SYMBOL(phonet_proto_register);
void phonet_proto_unregister(int protocol, struct phonet_protocol *pp)
{
- spin_lock(&proto_tab_lock);
+ mutex_lock(&proto_tab_lock);
BUG_ON(proto_tab[protocol] != pp);
- proto_tab[protocol] = NULL;
- spin_unlock(&proto_tab_lock);
+ rcu_assign_pointer(proto_tab[protocol], NULL);
+ mutex_unlock(&proto_tab_lock);
+ synchronize_rcu();
proto_unregister(pp->prot);
}
EXPORT_SYMBOL(phonet_proto_unregister);
--
1.6.3.3
^ permalink raw reply related
* [PATCH net-next 2/2] Phonet: convert routing table to RCU
From: Rémi Denis-Courmont @ 2009-11-13 15:01 UTC (permalink / raw)
To: netdev; +Cc: Rémi Denis-Courmont
In-Reply-To: <1258124479-1167-1-git-send-email-remi@remlab.net>
From: Rémi Denis-Courmont <remi.denis-courmont@nokia.com>
Signed-off-by: Rémi Denis-Courmont <remi.denis-courmont@nokia.com>
---
net/phonet/pn_dev.c | 59 +++++++++++++++++++++++++++++---------------------
1 files changed, 34 insertions(+), 25 deletions(-)
diff --git a/net/phonet/pn_dev.c b/net/phonet/pn_dev.c
index 6d64fda..3287f8f 100644
--- a/net/phonet/pn_dev.c
+++ b/net/phonet/pn_dev.c
@@ -34,7 +34,7 @@
#include <net/phonet/pn_dev.h>
struct phonet_routes {
- spinlock_t lock;
+ struct mutex lock;
struct net_device *table[64];
};
@@ -248,17 +248,22 @@ static void phonet_route_autodel(struct net_device *dev)
/* Remove left-over Phonet routes */
bitmap_zero(deleted, 64);
- spin_lock_bh(&pnn->routes.lock);
+ mutex_lock(&pnn->routes.lock);
for (i = 0; i < 64; i++)
if (dev == pnn->routes.table[i]) {
+ rcu_assign_pointer(pnn->routes.table[i], NULL);
set_bit(i, deleted);
- pnn->routes.table[i] = NULL;
- dev_put(dev);
}
- spin_unlock_bh(&pnn->routes.lock);
+ mutex_unlock(&pnn->routes.lock);
+
+ if (bitmap_empty(deleted, 64))
+ return; /* short-circuit RCU */
+ synchronize_rcu();
for (i = find_first_bit(deleted, 64); i < 64;
- i = find_next_bit(deleted, 64, i + 1))
+ i = find_next_bit(deleted, 64, i + 1)) {
rtm_phonet_notify(RTM_DELROUTE, dev, i);
+ dev_put(dev);
+ }
}
/* notify Phonet of device events */
@@ -300,7 +305,7 @@ static int phonet_init_net(struct net *net)
INIT_LIST_HEAD(&pnn->pndevs.list);
spin_lock_init(&pnn->pndevs.lock);
- spin_lock_init(&pnn->routes.lock);
+ mutex_init(&pnn->routes.lock);
net_assign_generic(net, phonet_net_id, pnn);
return 0;
}
@@ -361,13 +366,13 @@ int phonet_route_add(struct net_device *dev, u8 daddr)
int err = -EEXIST;
daddr = daddr >> 2;
- spin_lock_bh(&routes->lock);
+ mutex_lock(&routes->lock);
if (routes->table[daddr] == NULL) {
- routes->table[daddr] = dev;
+ rcu_assign_pointer(routes->table[daddr], dev);
dev_hold(dev);
err = 0;
}
- spin_unlock_bh(&routes->lock);
+ mutex_unlock(&routes->lock);
return err;
}
@@ -375,17 +380,20 @@ int phonet_route_del(struct net_device *dev, u8 daddr)
{
struct phonet_net *pnn = net_generic(dev_net(dev), phonet_net_id);
struct phonet_routes *routes = &pnn->routes;
- int err = -ENOENT;
daddr = daddr >> 2;
- spin_lock_bh(&routes->lock);
- if (dev == routes->table[daddr]) {
- routes->table[daddr] = NULL;
- dev_put(dev);
- err = 0;
- }
- spin_unlock_bh(&routes->lock);
- return err;
+ mutex_lock(&routes->lock);
+ if (dev == routes->table[daddr])
+ rcu_assign_pointer(routes->table[daddr], NULL);
+ else
+ dev = NULL;
+ mutex_unlock(&routes->lock);
+
+ if (!dev)
+ return -ENOENT;
+ synchronize_rcu();
+ dev_put(dev);
+ return 0;
}
struct net_device *phonet_route_get(struct net *net, u8 daddr)
@@ -397,9 +405,9 @@ struct net_device *phonet_route_get(struct net *net, u8 daddr)
ASSERT_RTNL(); /* no need to hold the device */
daddr >>= 2;
- spin_lock_bh(&routes->lock);
- dev = routes->table[daddr];
- spin_unlock_bh(&routes->lock);
+ rcu_read_lock();
+ dev = rcu_dereference(routes->table[daddr]);
+ rcu_read_unlock();
return dev;
}
@@ -409,11 +417,12 @@ struct net_device *phonet_route_output(struct net *net, u8 daddr)
struct phonet_routes *routes = &pnn->routes;
struct net_device *dev;
- spin_lock_bh(&routes->lock);
- dev = routes->table[daddr >> 2];
+ daddr >>= 2;
+ rcu_read_lock();
+ dev = rcu_dereference(routes->table[daddr]);
if (dev)
dev_hold(dev);
- spin_unlock_bh(&routes->lock);
+ rcu_read_unlock();
if (!dev)
dev = phonet_device_get(net); /* Default route */
--
1.6.3.3
^ permalink raw reply related
* Re: [net-next-2.6 PATCH v6 4/7 RFC] TCPCT part 1d: define TCP cookie option, extend existing struct's
From: William Allen Simpson @ 2009-11-13 16:06 UTC (permalink / raw)
To: Eric Dumazet; +Cc: Linux Kernel Network Developers, Joe Perches
In-Reply-To: <4AFCFD86.6020504@gmail.com>
[-- Attachment #1: Type: text/plain, Size: 1190 bytes --]
Eric Dumazet wrote:
> +/**
> + * A tcp_sock contains a pointer to the current value, and this is cloned to
> + * the tcp_timewait_sock.
> + *
> + * @cookie_pair: variable data from the option exchange.
> + *
> + * @cookie_desired: user specified tcpct_cookie_desired. Zero
> + * indicates default (sysctl_tcp_cookie_size).
> + * After cookie sent, remembers size of cookie.
> + * Range 0, TCP_COOKIE_MIN to TCP_COOKIE_MAX.
> + *
> + * @s_data_desired: user specified tcpct_s_data_desired. When the
> + * constant payload is specified (@s_data_constant),
> + * holds its length instead.
> + * Range 0 to TCP_MSS_DESIRED.
> + *
> + * @s_data_payload: constant data that is to be included in the
> + * payload of SYN or SYNACK segments when the
> + * cookie option is present.
> + */
>
> Thanks for this kerneldoc William ;)
>
> But header should be :
> /**
> * struct tcp_cookie_values - Some description...
>
Botheration. Just unlucky in following another example (again).
A private message suggests including the initialization code here
(instead of part 1f) to avoid some page flipping. Done.
Here's my revised attempt (untested). Any technical corrections?
[-- Attachment #2: TCPCT+1d6+.patch --]
[-- Type: text/plain, Size: 10956 bytes --]
diff --git a/include/linux/tcp.h b/include/linux/tcp.h
index eaa3113..6c5ff66 100644
--- a/include/linux/tcp.h
+++ b/include/linux/tcp.h
@@ -247,31 +247,38 @@ struct tcp_options_received {
sack_ok : 4, /* SACK seen on SYN packet */
snd_wscale : 4, /* Window scaling received from sender */
rcv_wscale : 4; /* Window scaling to send to receiver */
-/* SACKs data */
+ u8 cookie_plus:6, /* bytes in authenticator/cookie option */
+ cookie_out_never:1,
+ cookie_in_always:1;
u8 num_sacks; /* Number of SACK blocks */
- u16 user_mss; /* mss requested by user in ioctl */
+ u16 user_mss; /* mss requested by user in ioctl */
u16 mss_clamp; /* Maximal mss, negotiated at connection setup */
};
static inline void tcp_clear_options(struct tcp_options_received *rx_opt)
{
- rx_opt->tstamp_ok = rx_opt->sack_ok = rx_opt->wscale_ok = rx_opt->snd_wscale = 0;
+ rx_opt->tstamp_ok = rx_opt->sack_ok = 0;
+ rx_opt->wscale_ok = rx_opt->snd_wscale = 0;
+ rx_opt->cookie_plus = 0;
}
/* This is the max number of SACKS that we'll generate and process. It's safe
- * to increse this, although since:
+ * to increase this, although since:
* size = TCPOLEN_SACK_BASE_ALIGNED (4) + n * TCPOLEN_SACK_PERBLOCK (8)
* only four options will fit in a standard TCP header */
#define TCP_NUM_SACKS 4
+struct tcp_cookie_values;
+struct tcp_request_sock_ops;
+
struct tcp_request_sock {
struct inet_request_sock req;
#ifdef CONFIG_TCP_MD5SIG
/* Only used by TCP MD5 Signature so far. */
const struct tcp_request_sock_ops *af_specific;
#endif
- u32 rcv_isn;
- u32 snt_isn;
+ u32 rcv_isn;
+ u32 snt_isn;
};
static inline struct tcp_request_sock *tcp_rsk(const struct request_sock *req)
@@ -441,6 +448,12 @@ struct tcp_sock {
/* TCP MD5 Signature Option information */
struct tcp_md5sig_info *md5sig_info;
#endif
+
+ /* When the cookie options are generated and exchanged, then this
+ * object holds a reference to them (cookie_values->kref). Also
+ * contains related tcp_cookie_transactions fields.
+ */
+ struct tcp_cookie_values *cookie_values;
};
static inline struct tcp_sock *tcp_sk(const struct sock *sk)
@@ -459,6 +472,10 @@ struct tcp_timewait_sock {
u16 tw_md5_keylen;
u8 tw_md5_key[TCP_MD5SIG_MAXKEYLEN];
#endif
+ /* Few sockets in timewait have cookies; in that case, then this
+ * object holds a reference to them (tw_cookie_values->kref).
+ */
+ struct tcp_cookie_values *tw_cookie_values;
};
static inline struct tcp_timewait_sock *tcp_twsk(const struct sock *sk)
diff --git a/include/net/tcp.h b/include/net/tcp.h
index 738b65f..3a4c840 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -30,6 +30,7 @@
#include <linux/dmaengine.h>
#include <linux/crypto.h>
#include <linux/cryptohash.h>
+#include <linux/kref.h>
#include <net/inet_connection_sock.h>
#include <net/inet_timewait_sock.h>
@@ -164,6 +165,7 @@ extern void tcp_time_wait(struct sock *sk, int state, int timeo);
#define TCPOPT_SACK 5 /* SACK Block */
#define TCPOPT_TIMESTAMP 8 /* Better RTT estimations/PAWS */
#define TCPOPT_MD5SIG 19 /* MD5 Signature (RFC2385) */
+#define TCPOPT_COOKIE 253 /* Cookie extension (experimental) */
/*
* TCP option lengths
@@ -174,6 +176,10 @@ extern void tcp_time_wait(struct sock *sk, int state, int timeo);
#define TCPOLEN_SACK_PERM 2
#define TCPOLEN_TIMESTAMP 10
#define TCPOLEN_MD5SIG 18
+#define TCPOLEN_COOKIE_BASE 2 /* Cookie-less header extension */
+#define TCPOLEN_COOKIE_PAIR 3 /* Cookie pair header extension */
+#define TCPOLEN_COOKIE_MIN (TCPOLEN_COOKIE_BASE+TCP_COOKIE_MIN)
+#define TCPOLEN_COOKIE_MAX (TCPOLEN_COOKIE_BASE+TCP_COOKIE_MAX)
/* But this is what stacks really send out. */
#define TCPOLEN_TSTAMP_ALIGNED 12
@@ -1482,6 +1488,83 @@ struct tcp_request_sock_ops {
extern int tcp_cookie_generator(u32 *bakery);
+/**
+ * struct tcp_cookie_values - each socket needs extra space for the
+ * cookies, together with (optional) space for any SYN data.
+ *
+ * A tcp_sock contains a pointer to the current value, and this is
+ * cloned to the tcp_timewait_sock.
+ *
+ * @cookie_pair: variable data from the option exchange.
+ *
+ * @cookie_desired: user specified tcpct_cookie_desired. Zero
+ * indicates default (sysctl_tcp_cookie_size).
+ * After cookie sent, remembers size of cookie.
+ * Range 0, TCP_COOKIE_MIN to TCP_COOKIE_MAX.
+ *
+ * @s_data_desired: user specified tcpct_s_data_desired. When the
+ * constant payload is specified (@s_data_constant),
+ * holds its length instead.
+ * Range 0 to TCP_MSS_DESIRED.
+ *
+ * @s_data_payload: constant data that is to be included in the
+ * payload of SYN or SYNACK segments when the
+ * cookie option is present.
+ */
+struct tcp_cookie_values {
+ struct kref kref;
+ u8 cookie_pair[TCP_COOKIE_PAIR_SIZE];
+ u8 cookie_pair_size;
+ u8 cookie_desired;
+ u16 s_data_desired:11,
+ s_data_constant:1,
+ s_data_in:1,
+ s_data_out:1,
+ s_data_unused:2;
+ u8 s_data_payload[0];
+};
+
+static inline void tcp_cookie_values_release(struct kref *kref)
+{
+ kfree(container_of(kref, struct tcp_cookie_values, kref));
+}
+
+/* The length of constant payload data. Note that s_data_desired is
+ * overloaded, depending on s_data_constant: either the length of constant
+ * data (returned here) or the limit on variable data.
+ */
+static inline int tcp_s_data_size(const struct tcp_sock *tp)
+{
+ return (tp->cookie_values != NULL && tp->cookie_values->s_data_constant)
+ ? tp->cookie_values->s_data_desired
+ : 0;
+}
+
+/**
+ * struct tcp_extend_values - tcp_ipv?.c to tcp_output.c workspace.
+ *
+ * As tcp_request_sock has already been extended in other places, the
+ * only remaining method is to pass stack values along as function
+ * parameters. These parameters are not needed after sending SYNACK.
+ *
+ * @cookie_bakery: cryptographic secret and message workspace.
+ *
+ * @cookie_plus: bytes in authenticator/cookie option, copied from
+ * struct tcp_options_received (above).
+ */
+struct tcp_extend_values {
+ struct request_values rv;
+ u32 cookie_bakery[COOKIE_WORKSPACE_WORDS];
+ u8 cookie_plus:6,
+ cookie_out_never:1,
+ cookie_in_always:1;
+};
+
+static inline struct tcp_extend_values *tcp_xv(struct request_values *rvp)
+{
+ return (struct tcp_extend_values *)rvp;
+}
+
extern void tcp_v4_init(void);
extern void tcp_init(void);
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 094231b..2ae1985 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -1834,6 +1834,19 @@ static int tcp_v4_init_sock(struct sock *sk)
tp->af_specific = &tcp_sock_ipv4_specific;
#endif
+ /* TCP Cookie Transactions */
+ if (sysctl_tcp_cookie_size > 0) {
+ /* Default, cookies without s_data_payload. */
+ tp->cookie_values =
+ kzalloc(sizeof(*tp->cookie_values),
+ sk->sk_allocation);
+ if (tp->cookie_values != NULL)
+ kref_init(&tp->cookie_values->kref);
+ }
+ /* Presumed zeroed, in order of appearance:
+ * cookie_in_always, cookie_out_never,
+ * s_data_constant, s_data_in, s_data_out
+ */
sk->sk_sndbuf = sysctl_tcp_wmem[1];
sk->sk_rcvbuf = sysctl_tcp_rmem[1];
@@ -1887,6 +1900,13 @@ void tcp_v4_destroy_sock(struct sock *sk)
sk->sk_sndmsg_page = NULL;
}
+ /* TCP Cookie Transactions */
+ if (tp->cookie_values != NULL) {
+ kref_put(&tp->cookie_values->kref,
+ tcp_cookie_values_release);
+ tp->cookie_values = NULL;
+ }
+
percpu_counter_dec(&tcp_sockets_allocated);
}
diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c
index 7a42990..2b50da8 100644
--- a/net/ipv4/tcp_minisocks.c
+++ b/net/ipv4/tcp_minisocks.c
@@ -389,14 +389,42 @@ struct sock *tcp_create_openreq_child(struct sock *sk, struct request_sock *req,
const struct inet_request_sock *ireq = inet_rsk(req);
struct tcp_request_sock *treq = tcp_rsk(req);
struct inet_connection_sock *newicsk = inet_csk(newsk);
- struct tcp_sock *newtp;
+ struct tcp_sock *newtp = tcp_sk(newsk);
+ struct tcp_sock *oldtp = tcp_sk(sk);
+ struct tcp_cookie_values *oldcvp = oldtp->cookie_values;
+
+ /* TCP Cookie Transactions require space for the cookie pair,
+ * as it differs for each connection. There is no need to
+ * copy any s_data_payload stored at the original socket.
+ * Failure will prevent resuming the connection.
+ *
+ * Presumed copied, in order of appearance:
+ * cookie_in_always, cookie_out_never
+ */
+ if (oldcvp != NULL) {
+ struct tcp_cookie_values *newcvp =
+ kzalloc(sizeof(*newtp->cookie_values),
+ GFP_ATOMIC);
+
+ if (newcvp != NULL) {
+ kref_init(&newcvp->kref);
+ newcvp->cookie_desired =
+ oldcvp->cookie_desired;
+ newtp->cookie_values = newcvp;
+ } else {
+ /* Not Yet Implemented */
+ newtp->cookie_values = NULL;
+ }
+ }
/* Now setup tcp_sock */
- newtp = tcp_sk(newsk);
newtp->pred_flags = 0;
- newtp->rcv_wup = newtp->copied_seq = newtp->rcv_nxt = treq->rcv_isn + 1;
- newtp->snd_sml = newtp->snd_una = newtp->snd_nxt = treq->snt_isn + 1;
- newtp->snd_up = treq->snt_isn + 1;
+
+ newtp->rcv_wup = newtp->copied_seq =
+ newtp->rcv_nxt = treq->rcv_isn + 1;
+
+ newtp->snd_sml = newtp->snd_una = newtp->snd_nxt =
+ newtp->snd_up = treq->snt_isn + 1 + tcp_s_data_size(oldtp);
tcp_prequeue_init(newtp);
@@ -429,8 +457,8 @@ struct sock *tcp_create_openreq_child(struct sock *sk, struct request_sock *req,
tcp_set_ca_state(newsk, TCP_CA_Open);
tcp_init_xmit_timers(newsk);
skb_queue_head_init(&newtp->out_of_order_queue);
- newtp->write_seq = treq->snt_isn + 1;
- newtp->pushed_seq = newtp->write_seq;
+ newtp->write_seq = newtp->pushed_seq =
+ treq->snt_isn + 1 + tcp_s_data_size(oldtp);
newtp->rx_opt.saw_tstamp = 0;
@@ -596,7 +624,8 @@ struct sock *tcp_check_req(struct sock *sk, struct sk_buff *skb,
* Invalid ACK: reset will be sent by listening socket
*/
if ((flg & TCP_FLAG_ACK) &&
- (TCP_SKB_CB(skb)->ack_seq != tcp_rsk(req)->snt_isn + 1))
+ (TCP_SKB_CB(skb)->ack_seq != tcp_rsk(req)->snt_isn + 1 +
+ tcp_s_data_size(tcp_sk(sk))))
return sk;
/* Also, it would be not so bad idea to check rcv_tsecr, which
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index 3e327bc..2b16f9a 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -1865,6 +1865,19 @@ static int tcp_v6_init_sock(struct sock *sk)
tp->af_specific = &tcp_sock_ipv6_specific;
#endif
+ /* TCP Cookie Transactions */
+ if (sysctl_tcp_cookie_size > 0) {
+ /* Default, cookies without s_data_payload. */
+ tp->cookie_values =
+ kzalloc(sizeof(*tp->cookie_values),
+ sk->sk_allocation);
+ if (tp->cookie_values != NULL)
+ kref_init(&tp->cookie_values->kref);
+ }
+ /* Presumed zeroed, in order of appearance:
+ * cookie_in_always, cookie_out_never,
+ * s_data_constant, s_data_in, s_data_out
+ */
sk->sk_sndbuf = sysctl_tcp_wmem[1];
sk->sk_rcvbuf = sysctl_tcp_rmem[1];
--
1.6.3.3
^ permalink raw reply related
* [PATCH] net/can: add driver for mscan family & mpc52xx_mscan
From: Wolfram Sang @ 2009-11-13 16:14 UTC (permalink / raw)
To: netdev-u79uwXL29TY76Z2rM5mHXA
Cc: Grant Likely, socketcan-core-0fE9KPoRgkgATYTw5x5z8w,
linuxppc-dev-mnsaURCQ41sdnm+yROfE0A, David Miller,
Wolfgang Grandegger
Taken from socketcan-svn, fixed remaining todos, cleaned up, tested with a
phyCORE-MPC5200B-IO and a custom board.
Signed-off-by: Wolfram Sang <w.sang-bIcnvbaLZ9MEGnE8C9+IrQ@public.gmane.org>
Cc: Wolfgang Grandegger <wg-5Yr1BZd7O62+XT7JhA+gdA@public.gmane.org>
Cc: Grant Likely <grant.likely-s3s/WqlpOiPyB63q8FvJNQ@public.gmane.org>
Cc: David Miller <davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>
---
This patch is based on net-next as of yesterday.
To make the review easier for those who are already familiar with earlier
versions of this driver (especially for Wolfgang), I put my development branch
online, so you can check my changes incrementally:
http://git.pengutronix.de/?p=wsa/linux-2.6.git;a=summary
Documentation/powerpc/dts-bindings/fsl/mpc5200.txt | 9 +
drivers/net/can/Kconfig | 19 +
drivers/net/can/Makefile | 1 +
drivers/net/can/mscan/Makefile | 5 +
drivers/net/can/mscan/mpc52xx_can.c | 279 ++++++++
drivers/net/can/mscan/mscan.c | 699 ++++++++++++++++++++
drivers/net/can/mscan/mscan.h | 262 ++++++++
7 files changed, 1274 insertions(+), 0 deletions(-)
create mode 100644 drivers/net/can/mscan/Makefile
create mode 100644 drivers/net/can/mscan/mpc52xx_can.c
create mode 100644 drivers/net/can/mscan/mscan.c
create mode 100644 drivers/net/can/mscan/mscan.h
diff --git a/Documentation/powerpc/dts-bindings/fsl/mpc5200.txt b/Documentation/powerpc/dts-bindings/fsl/mpc5200.txt
index 8447fd7..b151fb1 100644
--- a/Documentation/powerpc/dts-bindings/fsl/mpc5200.txt
+++ b/Documentation/powerpc/dts-bindings/fsl/mpc5200.txt
@@ -178,3 +178,12 @@ External interrupts:
external irq3: interrupts = <1 3 n>;
'n' is sense (0: level high, 1: edge rising, 2: edge falling 3: level low)
+fsl,mpc5200-mscan nodes
+-----------------------
+In addition to the required compatible-, reg- and interrupt-properites, you can
+also specify which clock shall be used for the bus:
+
+- fsl,mscan-clk-src - a string describing the clock source. Valid values
+ are "ip" for IP_CLK and "sys" for SYS_XTAL.
+ "sys" is the default in case the property is not
+ present.
diff --git a/drivers/net/can/Kconfig b/drivers/net/can/Kconfig
index b819cc2..c16e6ff 100644
--- a/drivers/net/can/Kconfig
+++ b/drivers/net/can/Kconfig
@@ -108,6 +108,25 @@ config CAN_MCP251X
---help---
Driver for the Microchip MCP251x SPI CAN controllers.
+config CAN_MSCAN
+ depends on CAN_DEV && (PPC || M68K || M68KNOMMU)
+ tristate "Support for Freescale MSCAN based chips"
+ ---help---
+ The Motorola Scalable Controller Area Network (MSCAN) definition
+ is based on the MSCAN12 definition which is the specific
+ implementation of the Motorola Scalable CAN concept targeted for
+ the Motorola MC68HC12 Microcontroller Family.
+
+config CAN_MPC52XX
+ tristate "Freescale MPC5xxx onboard CAN controller"
+ depends on CAN_MSCAN && PPC_MPC52xx
+ ---help---
+ If you say yes here you get support for Freescale's MPC52xx
+ onboard dualCAN controller.
+
+ This driver can also be built as a module. If so, the module
+ will be called mpc5xxx_can.
+
config CAN_DEBUG_DEVICES
bool "CAN devices debugging messages"
depends on CAN
diff --git a/drivers/net/can/Makefile b/drivers/net/can/Makefile
index 1489181..56899fe 100644
--- a/drivers/net/can/Makefile
+++ b/drivers/net/can/Makefile
@@ -10,6 +10,7 @@ can-dev-y := dev.o
obj-y += usb/
obj-$(CONFIG_CAN_SJA1000) += sja1000/
+obj-$(CONFIG_CAN_MSCAN) += mscan/
obj-$(CONFIG_CAN_AT91) += at91_can.o
obj-$(CONFIG_CAN_TI_HECC) += ti_hecc.o
obj-$(CONFIG_CAN_MCP251X) += mcp251x.o
diff --git a/drivers/net/can/mscan/Makefile b/drivers/net/can/mscan/Makefile
new file mode 100644
index 0000000..2bd9f04
--- /dev/null
+++ b/drivers/net/can/mscan/Makefile
@@ -0,0 +1,5 @@
+
+obj-$(CONFIG_CAN_MPC52XX) += mscan-mpc52xx.o
+mscan-mpc52xx-objs := mscan.o mpc52xx_can.o
+
+ccflags-$(CONFIG_CAN_DEBUG_DEVICES) := -DDEBUG
diff --git a/drivers/net/can/mscan/mpc52xx_can.c b/drivers/net/can/mscan/mpc52xx_can.c
new file mode 100644
index 0000000..4707a82
--- /dev/null
+++ b/drivers/net/can/mscan/mpc52xx_can.c
@@ -0,0 +1,279 @@
+/*
+ * CAN bus driver for the Freescale MPC5xxx embedded CPU.
+ *
+ * Copyright (C) 2004-2005 Andrey Volkov <avolkov-ppI4tVfbJvJWk0Htik3J/w@public.gmane.org>,
+ * Varma Electronics Oy
+ * Copyright (C) 2008-2009 Wolfgang Grandegger <wg-5Yr1BZd7O62+XT7JhA+gdA@public.gmane.org>
+ * Copyright (C) 2009 Wolfram Sang, Pengutronix <w.sang-bIcnvbaLZ9MEGnE8C9+IrQ@public.gmane.org>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the version 2 of the GNU General Public License
+ * as published by the Free Software Foundation
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/interrupt.h>
+#include <linux/platform_device.h>
+#include <linux/netdevice.h>
+#include <linux/can.h>
+#include <linux/can/dev.h>
+#include <linux/of_platform.h>
+#include <sysdev/fsl_soc.h>
+#include <linux/io.h>
+#include <asm/mpc52xx.h>
+
+#include "mscan.h"
+
+
+#define DRV_NAME "mpc5xxx_can"
+
+static struct of_device_id mpc52xx_cdm_ids[] __devinitdata = {
+ { .compatible = "fsl,mpc5200-cdm", },
+ { .compatible = "fsl,mpc5200b-cdm", },
+ {}
+};
+
+/*
+ * Get the frequency of the external oscillator clock connected
+ * to the SYS_XTAL_IN pin, or return 0 if it cannot be determined.
+ */
+static unsigned int __devinit mpc52xx_can_xtal_freq(struct of_device *of)
+{
+ struct mpc52xx_cdm __iomem *cdm;
+ struct device_node *np_cdm;
+ unsigned int freq;
+ u32 val;
+
+ freq = mpc5xxx_get_bus_frequency(of->node);
+ if (!freq)
+ return 0;
+
+ /*
+ * Determine SYS_XTAL_IN frequency from the clock domain settings
+ */
+ np_cdm = of_find_matching_node(NULL, mpc52xx_cdm_ids);
+ if (!np_cdm) {
+ dev_err(&of->dev, "can't get clock node!\n");
+ return 0;
+ }
+ cdm = of_iomap(np_cdm, 0);
+ of_node_put(np_cdm);
+
+ if (in_8(&cdm->ipb_clk_sel) & 0x1)
+ freq *= 2;
+ val = in_be32(&cdm->rstcfg);
+ if (val & (1 << 5))
+ freq *= 8;
+ else
+ freq *= 4;
+ if (val & (1 << 6))
+ freq /= 12;
+ else
+ freq /= 16;
+
+ iounmap(cdm);
+
+ return freq;
+}
+
+/*
+ * Get frequency of the MSCAN clock source
+ *
+ * Either the oscillator clock (SYS_XTAL_IN) or the IP bus clock (IP_CLK)
+ * can be selected. According to the MPC5200 user's manual, the oscillator
+ * clock is the better choice as it has less jitter but due to a hardware
+ * bug, it can not be selected for the old MPC5200 Rev. A chips.
+ */
+
+static unsigned int __devinit mpc52xx_can_clock_freq(struct of_device *of,
+ int clock_src)
+{
+ unsigned int pvr;
+
+ pvr = mfspr(SPRN_PVR);
+
+ if (clock_src == MSCAN_CLKSRC_BUS || pvr == 0x80822011)
+ return mpc5xxx_get_bus_frequency(of->node);
+
+ return mpc52xx_can_xtal_freq(of);
+}
+
+static int __devinit mpc5xxx_can_probe(struct of_device *ofdev,
+ const struct of_device_id *id)
+{
+ struct device_node *np = ofdev->node;
+ struct net_device *dev;
+ struct mscan_priv *priv;
+ void __iomem *base;
+ const char *clk_src;
+ int err, irq, clock_src;
+
+ base = of_iomap(ofdev->node, 0);
+ if (!base) {
+ dev_err(&ofdev->dev, "couldn't ioremap\n");
+ err = -ENOMEM;
+ goto exit_release_mem;
+ }
+
+ irq = irq_of_parse_and_map(np, 0);
+ if (!irq) {
+ dev_err(&ofdev->dev, "no irq found\n");
+ err = -ENODEV;
+ goto exit_unmap_mem;
+ }
+
+ dev = alloc_mscandev();
+ if (!dev) {
+ err = -ENOMEM;
+ goto exit_dispose_irq;
+ }
+
+ priv = netdev_priv(dev);
+ priv->reg_base = base;
+ dev->irq = irq;
+
+ /*
+ * Either the oscillator clock (SYS_XTAL_IN) or the IP bus clock
+ * (IP_CLK) can be selected as MSCAN clock source. According to
+ * the MPC5200 user's manual, the oscillator clock is the better
+ * choice as it has less jitter. For this reason, it is selected
+ * by default.
+ */
+ clk_src = of_get_property(np, "fsl,mscan-clk-src", NULL);
+ if (clk_src && strcmp(clk_src, "ip") == 0)
+ clock_src = MSCAN_CLKSRC_BUS;
+ else
+ clock_src = MSCAN_CLKSRC_XTAL;
+ priv->can.clock.freq = mpc52xx_can_clock_freq(ofdev, clock_src);
+ if (!priv->can.clock.freq) {
+ dev_err(&ofdev->dev, "couldn't get MSCAN clock frequency\n");
+ err = -ENODEV;
+ goto exit_free_mscan;
+ }
+
+ SET_NETDEV_DEV(dev, &ofdev->dev);
+
+ err = register_mscandev(dev, clock_src);
+ if (err) {
+ dev_err(&ofdev->dev, "registering %s failed (err=%d)\n",
+ DRV_NAME, err);
+ goto exit_free_mscan;
+ }
+
+ dev_set_drvdata(&ofdev->dev, dev);
+
+ dev_info(&ofdev->dev, "MSCAN at 0x%p, irq %d, clock %d Hz\n",
+ priv->reg_base, dev->irq, priv->can.clock.freq);
+
+ return 0;
+
+exit_free_mscan:
+ free_candev(dev);
+exit_dispose_irq:
+ irq_dispose_mapping(irq);
+exit_unmap_mem:
+ iounmap(base);
+exit_release_mem:
+ return err;
+}
+
+static int __devexit mpc5xxx_can_remove(struct of_device *ofdev)
+{
+ struct net_device *dev = dev_get_drvdata(&ofdev->dev);
+ struct mscan_priv *priv = netdev_priv(dev);
+
+ dev_set_drvdata(&ofdev->dev, NULL);
+
+ unregister_mscandev(dev);
+ iounmap(priv->reg_base);
+ irq_dispose_mapping(dev->irq);
+ free_candev(dev);
+
+ return 0;
+}
+
+#ifdef CONFIG_PM
+static struct mscan_regs saved_regs;
+static int mpc5xxx_can_suspend(struct of_device *ofdev, pm_message_t state)
+{
+ struct net_device *dev = dev_get_drvdata(&ofdev->dev);
+ struct mscan_priv *priv = netdev_priv(dev);
+ struct mscan_regs *regs = (struct mscan_regs *)priv->reg_base;
+
+ _memcpy_fromio(&saved_regs, regs, sizeof(*regs));
+
+ return 0;
+}
+
+static int mpc5xxx_can_resume(struct of_device *ofdev)
+{
+ struct net_device *dev = dev_get_drvdata(&ofdev->dev);
+ struct mscan_priv *priv = netdev_priv(dev);
+ struct mscan_regs *regs = (struct mscan_regs *)priv->reg_base;
+
+ regs->canctl0 |= MSCAN_INITRQ;
+ while ((regs->canctl1 & MSCAN_INITAK) == 0)
+ udelay(10);
+
+ regs->canctl1 = saved_regs.canctl1;
+ regs->canbtr0 = saved_regs.canbtr0;
+ regs->canbtr1 = saved_regs.canbtr1;
+ regs->canidac = saved_regs.canidac;
+
+ /* restore masks, buffers etc. */
+ _memcpy_toio(®s->canidar1_0, (void *)&saved_regs.canidar1_0,
+ sizeof(*regs) - offsetof(struct mscan_regs, canidar1_0));
+
+ regs->canctl0 &= ~MSCAN_INITRQ;
+ regs->cantbsel = saved_regs.cantbsel;
+ regs->canrier = saved_regs.canrier;
+ regs->cantier = saved_regs.cantier;
+ regs->canctl0 = saved_regs.canctl0;
+
+ return 0;
+}
+#endif
+
+static struct of_device_id __devinitdata mpc5xxx_can_table[] = {
+ {.compatible = "fsl,mpc5200-mscan"},
+ {.compatible = "fsl,mpc5200b-mscan"},
+ {},
+};
+
+static struct of_platform_driver mpc5xxx_can_driver = {
+ .owner = THIS_MODULE,
+ .name = "mpc5xxx_can",
+ .probe = mpc5xxx_can_probe,
+ .remove = __devexit_p(mpc5xxx_can_remove),
+#ifdef CONFIG_PM
+ .suspend = mpc5xxx_can_suspend,
+ .resume = mpc5xxx_can_resume,
+#endif
+ .match_table = mpc5xxx_can_table,
+};
+
+static int __init mpc5xxx_can_init(void)
+{
+ return of_register_platform_driver(&mpc5xxx_can_driver);
+}
+module_init(mpc5xxx_can_init);
+
+static void __exit mpc5xxx_can_exit(void)
+{
+ return of_unregister_platform_driver(&mpc5xxx_can_driver);
+};
+module_exit(mpc5xxx_can_exit);
+
+MODULE_AUTHOR("Wolfgang Grandegger <wg-5Yr1BZd7O62+XT7JhA+gdA@public.gmane.org>");
+MODULE_DESCRIPTION("Freescale MPC5200 CAN driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/net/can/mscan/mscan.c b/drivers/net/can/mscan/mscan.c
new file mode 100644
index 0000000..49542ca
--- /dev/null
+++ b/drivers/net/can/mscan/mscan.c
@@ -0,0 +1,699 @@
+/*
+ * CAN bus driver for the alone generic (as possible as) MSCAN controller.
+ *
+ * Copyright (C) 2005-2006 Andrey Volkov <avolkov-ppI4tVfbJvJWk0Htik3J/w@public.gmane.org>,
+ * Varma Electronics Oy
+ * Copyright (C) 2008-2009 Wolfgang Grandegger <wg-5Yr1BZd7O62+XT7JhA+gdA@public.gmane.org>
+ * Copytight (C) 2008-2009 Pengutronix <kernel-bIcnvbaLZ9MEGnE8C9+IrQ@public.gmane.org>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the version 2 of the GNU General Public License
+ * as published by the Free Software Foundation
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/interrupt.h>
+#include <linux/delay.h>
+#include <linux/netdevice.h>
+#include <linux/if_arp.h>
+#include <linux/if_ether.h>
+#include <linux/list.h>
+#include <linux/can.h>
+#include <linux/can/dev.h>
+#include <linux/can/error.h>
+#include <linux/io.h>
+
+#include "mscan.h"
+
+#define MSCAN_NORMAL_MODE 0
+#define MSCAN_SLEEP_MODE MSCAN_SLPRQ
+#define MSCAN_INIT_MODE (MSCAN_INITRQ | MSCAN_SLPRQ)
+#define MSCAN_POWEROFF_MODE (MSCAN_CSWAI | MSCAN_SLPRQ)
+#define MSCAN_SET_MODE_RETRIES 255
+#define MSCAN_ECHO_SKB_MAX 3
+
+#define BTR0_BRP_MASK 0x3f
+#define BTR0_SJW_SHIFT 6
+#define BTR0_SJW_MASK (0x3 << BTR0_SJW_SHIFT)
+
+#define BTR1_TSEG1_MASK 0xf
+#define BTR1_TSEG2_SHIFT 4
+#define BTR1_TSEG2_MASK (0x7 << BTR1_TSEG2_SHIFT)
+#define BTR1_SAM_SHIFT 7
+
+#define BTR0_SET_BRP(brp) (((brp) - 1) & BTR0_BRP_MASK)
+#define BTR0_SET_SJW(sjw) ((((sjw) - 1) << BTR0_SJW_SHIFT) & \
+ BTR0_SJW_MASK)
+
+#define BTR1_SET_TSEG1(tseg1) (((tseg1) - 1) & BTR1_TSEG1_MASK)
+#define BTR1_SET_TSEG2(tseg2) ((((tseg2) - 1) << BTR1_TSEG2_SHIFT) & \
+ BTR1_TSEG2_MASK)
+#define BTR1_SET_SAM(sam) ((sam) ? 1 << BTR1_SAM_SHIFT : 0)
+
+static struct can_bittiming_const mscan_bittiming_const = {
+ .name = "mscan",
+ .tseg1_min = 4,
+ .tseg1_max = 16,
+ .tseg2_min = 2,
+ .tseg2_max = 8,
+ .sjw_max = 4,
+ .brp_min = 1,
+ .brp_max = 64,
+ .brp_inc = 1,
+};
+
+struct mscan_state {
+ u8 mode;
+ u8 canrier;
+ u8 cantier;
+};
+
+#define F_RX_PROGRESS 0
+#define F_TX_PROGRESS 1
+#define F_TX_WAIT_ALL 2
+
+static enum can_state state_map[] = {
+ CAN_STATE_ERROR_ACTIVE,
+ CAN_STATE_ERROR_WARNING,
+ CAN_STATE_ERROR_PASSIVE,
+ CAN_STATE_BUS_OFF
+};
+
+static int mscan_set_mode(struct net_device *dev, u8 mode)
+{
+ struct mscan_priv *priv = netdev_priv(dev);
+ struct mscan_regs *regs = (struct mscan_regs *)priv->reg_base;
+ int ret = 0;
+ int i;
+ u8 canctl1;
+
+ if (mode != MSCAN_NORMAL_MODE) {
+
+ if (priv->tx_active) {
+ /* Abort transfers before going to sleep */#
+ out_8(®s->cantarq, priv->tx_active);
+ /* Suppress TX done interrupts */
+ out_8(®s->cantier, 0);
+ }
+
+ canctl1 = in_8(®s->canctl1);
+ if ((mode & MSCAN_SLPRQ) && (canctl1 & MSCAN_SLPAK) == 0) {
+ out_8(®s->canctl0,
+ in_8(®s->canctl0) | MSCAN_SLPRQ);
+ for (i = 0; i < MSCAN_SET_MODE_RETRIES; i++) {
+ if (in_8(®s->canctl1) & MSCAN_SLPAK)
+ break;
+ udelay(100);
+ }
+ /*
+ * The mscan controller will fail to enter sleep mode,
+ * while there are irregular activities on bus, like
+ * somebody keeps retransmitting. This behavior is
+ * undocumented and seems to differ between mscan built
+ * in mpc5200b and mpc5200. We proceed in that case,
+ * since otherwise the slprq will be kept set and the
+ * controller will get stuck. NOTE: INITRQ or CSWAI
+ * will abort all active transmit actions, if still
+ * any, at once.
+ */
+ if (i >= MSCAN_SET_MODE_RETRIES)
+ dev_dbg(dev->dev.parent,
+ "device failed to enter sleep mode. "
+ "We proceed anyhow.\n");
+ else
+ priv->can.state = CAN_STATE_SLEEPING;
+ }
+
+ if ((mode & MSCAN_INITRQ) && (canctl1 & MSCAN_INITAK) == 0) {
+ out_8(®s->canctl0,
+ in_8(®s->canctl0) | MSCAN_INITRQ);
+ for (i = 0; i < MSCAN_SET_MODE_RETRIES; i++) {
+ if (in_8(®s->canctl1) & MSCAN_INITAK)
+ break;
+ }
+ if (i >= MSCAN_SET_MODE_RETRIES)
+ ret = -ENODEV;
+ }
+ if (!ret)
+ priv->can.state = CAN_STATE_STOPPED;
+
+ if (mode & MSCAN_CSWAI)
+ out_8(®s->canctl0,
+ in_8(®s->canctl0) | MSCAN_CSWAI);
+
+ } else {
+ canctl1 = in_8(®s->canctl1);
+ if (canctl1 & (MSCAN_SLPAK | MSCAN_INITAK)) {
+ out_8(®s->canctl0, in_8(®s->canctl0) &
+ ~(MSCAN_SLPRQ | MSCAN_INITRQ));
+ for (i = 0; i < MSCAN_SET_MODE_RETRIES; i++) {
+ canctl1 = in_8(®s->canctl1);
+ if (!(canctl1 & (MSCAN_INITAK | MSCAN_SLPAK)))
+ break;
+ }
+ if (i >= MSCAN_SET_MODE_RETRIES)
+ ret = -ENODEV;
+ else
+ priv->can.state = CAN_STATE_ERROR_ACTIVE;
+ }
+ }
+ return ret;
+}
+
+static int mscan_start(struct net_device *dev)
+{
+ struct mscan_priv *priv = netdev_priv(dev);
+ struct mscan_regs *regs = (struct mscan_regs *)priv->reg_base;
+ u8 canrflg;
+ int err;
+
+ out_8(®s->canrier, 0);
+
+ INIT_LIST_HEAD(&priv->tx_head);
+ priv->prev_buf_id = 0;
+ priv->cur_pri = 0;
+ priv->tx_active = 0;
+ priv->shadow_canrier = 0;
+ priv->flags = 0;
+
+ err = mscan_set_mode(dev, MSCAN_NORMAL_MODE);
+ if (err)
+ return err;
+
+ canrflg = in_8(®s->canrflg);
+ priv->shadow_statflg = canrflg & MSCAN_STAT_MSK;
+ priv->can.state = state_map[max(MSCAN_STATE_RX(canrflg),
+ MSCAN_STATE_TX(canrflg))];
+ out_8(®s->cantier, 0);
+
+ /* Enable receive interrupts. */
+ out_8(®s->canrier, MSCAN_OVRIE | MSCAN_RXFIE | MSCAN_CSCIE |
+ MSCAN_RSTATE1 | MSCAN_RSTATE0 | MSCAN_TSTATE1 | MSCAN_TSTATE0);
+
+ return 0;
+}
+
+static netdev_tx_t mscan_start_xmit(struct sk_buff *skb, struct net_device *dev)
+{
+ struct can_frame *frame = (struct can_frame *)skb->data;
+ struct mscan_priv *priv = netdev_priv(dev);
+ struct mscan_regs *regs = (struct mscan_regs *)priv->reg_base;
+ int i, rtr, buf_id;
+ u32 can_id;
+
+ if (frame->can_dlc > 8)
+ return -EINVAL;
+
+ out_8(®s->cantier, 0);
+
+ i = ~priv->tx_active & MSCAN_TXE;
+ buf_id = ffs(i) - 1;
+ switch (hweight8(i)) {
+ case 0:
+ netif_stop_queue(dev);
+ dev_err(dev->dev.parent, "Tx Ring full when queue awake!\n");
+ return NETDEV_TX_BUSY;
+ case 1:
+ /*
+ * if buf_id < 3, then current frame will be send out of order,
+ * since buffer with lower id have higher priority (hell..)
+ */
+ netif_stop_queue(dev);
+ case 2:
+ if (buf_id < priv->prev_buf_id) {
+ priv->cur_pri++;
+ if (priv->cur_pri == 0xff) {
+ set_bit(F_TX_WAIT_ALL, &priv->flags);
+ netif_stop_queue(dev);
+ }
+ }
+ set_bit(F_TX_PROGRESS, &priv->flags);
+ break;
+ }
+ priv->prev_buf_id = buf_id;
+ out_8(®s->cantbsel, i);
+
+ rtr = frame->can_id & CAN_RTR_FLAG;
+
+ if (frame->can_id & CAN_EFF_FLAG) {
+ can_id = (frame->can_id & CAN_EFF_MASK) << 1;
+ if (rtr)
+ can_id |= 1;
+ out_be16(®s->tx.idr3_2, can_id);
+
+ can_id >>= 16;
+ can_id = (can_id & 0x7) | ((can_id << 2) & 0xffe0) | (3 << 3);
+ } else {
+ can_id = (frame->can_id & CAN_SFF_MASK) << 5;
+ if (rtr)
+ can_id |= 1 << 4;
+ }
+ out_be16(®s->tx.idr1_0, can_id);
+
+ if (!rtr) {
+ void __iomem *data = ®s->tx.dsr1_0;
+ u16 *payload = (u16 *) frame->data;
+ /* It is safe to write into dsr[dlc+1] */
+ for (i = 0; i < (frame->can_dlc + 1) / 2; i++) {
+ out_be16(data, *payload++);
+ data += 2 + _MSCAN_RESERVED_DSR_SIZE;
+ }
+ }
+
+ out_8(®s->tx.dlr, frame->can_dlc);
+ out_8(®s->tx.tbpr, priv->cur_pri);
+
+ /* Start transmission. */
+ out_8(®s->cantflg, 1 << buf_id);
+
+ if (!test_bit(F_TX_PROGRESS, &priv->flags))
+ dev->trans_start = jiffies;
+
+ list_add_tail(&priv->tx_queue[buf_id].list, &priv->tx_head);
+
+ can_put_echo_skb(skb, dev, buf_id);
+
+ /* Enable interrupt. */
+ priv->tx_active |= 1 << buf_id;
+ out_8(®s->cantier, priv->tx_active);
+
+ return NETDEV_TX_OK;
+}
+
+/* This function returns the old state to see where we came from */
+static enum can_state check_set_state(struct net_device *dev, u8 canrflg)
+{
+ struct mscan_priv *priv = netdev_priv(dev);
+ enum can_state state, old_state = priv->can.state;
+
+ if (canrflg & MSCAN_CSCIF && old_state <= CAN_STATE_BUS_OFF) {
+ state = state_map[max(MSCAN_STATE_RX(canrflg),
+ MSCAN_STATE_TX(canrflg))];
+ priv->can.state = state;
+ }
+ return old_state;
+}
+
+static void mscan_get_rx_frame(struct net_device *dev, struct can_frame *frame)
+{
+ struct mscan_priv *priv = netdev_priv(dev);
+ struct mscan_regs *regs = (struct mscan_regs *)priv->reg_base;
+ u32 can_id;
+ int i;
+
+ can_id = in_be16(®s->rx.idr1_0);
+ if (can_id & (1 << 3)) {
+ frame->can_id = CAN_EFF_FLAG;
+ can_id = ((can_id << 16) | in_be16(®s->rx.idr3_2));
+ can_id = ((can_id & 0xffe00000) |
+ ((can_id & 0x7ffff) << 2)) >> 2;
+ } else {
+ can_id >>= 4;
+ frame->can_id = 0;
+ }
+
+ frame->can_id |= can_id >> 1;
+ if (can_id & 1)
+ frame->can_id |= CAN_RTR_FLAG;
+ frame->can_dlc = in_8(®s->rx.dlr) & 0xf;
+
+ if (!(frame->can_id & CAN_RTR_FLAG)) {
+ void __iomem *data = ®s->rx.dsr1_0;
+ u16 *payload = (u16 *) frame->data;
+ for (i = 0; i < (frame->can_dlc + 1) / 2; i++) {
+ *payload++ = in_be16(data);
+ data += 2 + _MSCAN_RESERVED_DSR_SIZE;
+ }
+ }
+
+ out_8(®s->canrflg, MSCAN_RXF);
+}
+
+static void mscan_get_err_frame(struct net_device *dev, struct can_frame *frame,
+ u8 canrflg)
+{
+ struct mscan_priv *priv = netdev_priv(dev);
+ struct mscan_regs *regs = (struct mscan_regs *)priv->reg_base;
+ struct net_device_stats *stats = &dev->stats;
+ enum can_state old_state;
+
+ dev_dbg(dev->dev.parent, "error interrupt (canrflg=%#x)\n", canrflg);
+ frame->can_id = CAN_ERR_FLAG;
+
+ if (canrflg & MSCAN_OVRIF) {
+ frame->can_id |= CAN_ERR_CRTL;
+ frame->data[1] = CAN_ERR_CRTL_RX_OVERFLOW;
+ stats->rx_over_errors++;
+ stats->rx_errors++;
+ } else
+ frame->data[1] = 0;
+
+ old_state = check_set_state(dev, canrflg);
+ /* State changed */
+ if (old_state != priv->can.state) {
+ switch (priv->can.state) {
+ case CAN_STATE_ERROR_WARNING:
+ frame->can_id |= CAN_ERR_CRTL;
+ priv->can.can_stats.error_warning++;
+ if ((priv->shadow_statflg & MSCAN_RSTAT_MSK) <
+ (canrflg & MSCAN_RSTAT_MSK))
+ frame->data[1] |= CAN_ERR_CRTL_RX_WARNING;
+
+ if ((priv->shadow_statflg & MSCAN_TSTAT_MSK) <
+ (canrflg & MSCAN_TSTAT_MSK))
+ frame->data[1] |= CAN_ERR_CRTL_TX_WARNING;
+ break;
+ case CAN_STATE_ERROR_PASSIVE:
+ frame->can_id |= CAN_ERR_CRTL;
+ priv->can.can_stats.error_passive++;
+ frame->data[1] |= CAN_ERR_CRTL_RX_PASSIVE;
+ break;
+ case CAN_STATE_BUS_OFF:
+ frame->can_id |= CAN_ERR_BUSOFF;
+ /*
+ * The MSCAN on the MPC5200 does recover from bus-off
+ * automatically. To avoid that we stop the chip doing
+ * a light-weight stop (we are in irq-context).
+ */
+ out_8(®s->cantier, 0);
+ out_8(®s->canrier, 0);
+ out_8(®s->canctl0, in_8(®s->canctl0) |
+ MSCAN_SLPRQ | MSCAN_INITRQ);
+ can_bus_off(dev);
+ break;
+ default:
+ break;
+ }
+ }
+ priv->shadow_statflg = canrflg & MSCAN_STAT_MSK;
+ frame->can_dlc = CAN_ERR_DLC;
+ out_8(®s->canrflg, MSCAN_ERR_IF);
+}
+
+static int mscan_rx_poll(struct napi_struct *napi, int quota)
+{
+ struct mscan_priv *priv = container_of(napi, struct mscan_priv, napi);
+ struct net_device *dev = napi->dev;
+ struct mscan_regs *regs = (struct mscan_regs *)priv->reg_base;
+ struct net_device_stats *stats = &dev->stats;
+ int npackets = 0;
+ int ret = 1;
+ struct sk_buff *skb;
+ struct can_frame *frame;
+ u8 canrflg;
+
+ while (npackets < quota && ((canrflg = in_8(®s->canrflg)) &
+ (MSCAN_RXF | MSCAN_ERR_IF))) {
+
+ skb = alloc_can_skb(dev, &frame);
+ if (!skb) {
+ if (printk_ratelimit())
+ dev_notice(dev->dev.parent, "packet dropped\n");
+ stats->rx_dropped++;
+ out_8(®s->canrflg, canrflg);
+ continue;
+ }
+
+ if (canrflg & MSCAN_RXF)
+ mscan_get_rx_frame(dev, frame);
+ else if (canrflg & MSCAN_ERR_IF)
+ mscan_get_err_frame(dev, frame, canrflg);
+
+ stats->rx_packets++;
+ stats->rx_bytes += frame->can_dlc;
+ npackets++;
+ netif_receive_skb(skb);
+ }
+
+ if (!(in_8(®s->canrflg) & (MSCAN_RXF | MSCAN_ERR_IF))) {
+ napi_complete(&priv->napi);
+ clear_bit(F_RX_PROGRESS, &priv->flags);
+ if (priv->can.state < CAN_STATE_BUS_OFF)
+ out_8(®s->canrier, priv->shadow_canrier);
+ ret = 0;
+ }
+ return ret;
+}
+
+static irqreturn_t mscan_isr(int irq, void *dev_id)
+{
+ struct net_device *dev = (struct net_device *)dev_id;
+ struct mscan_priv *priv = netdev_priv(dev);
+ struct mscan_regs *regs = (struct mscan_regs *)priv->reg_base;
+ struct net_device_stats *stats = &dev->stats;
+ u8 cantier, cantflg, canrflg;
+ irqreturn_t ret = IRQ_NONE;
+
+ cantier = in_8(®s->cantier) & MSCAN_TXE;
+ cantflg = in_8(®s->cantflg) & cantier;
+
+ if (cantier && cantflg) {
+
+ struct list_head *tmp, *pos;
+
+ list_for_each_safe(pos, tmp, &priv->tx_head) {
+ struct tx_queue_entry *entry =
+ list_entry(pos, struct tx_queue_entry, list);
+ u8 mask = entry->mask;
+
+ if (!(cantflg & mask))
+ continue;
+
+ out_8(®s->cantbsel, mask);
+ stats->tx_bytes += in_8(®s->tx.dlr);
+ stats->tx_packets++;
+ can_get_echo_skb(dev, entry->id);
+ priv->tx_active &= ~mask;
+ list_del(pos);
+ }
+
+ if (list_empty(&priv->tx_head)) {
+ clear_bit(F_TX_WAIT_ALL, &priv->flags);
+ clear_bit(F_TX_PROGRESS, &priv->flags);
+ priv->cur_pri = 0;
+ } else
+ dev->trans_start = jiffies;
+
+ if (!test_bit(F_TX_WAIT_ALL, &priv->flags))
+ netif_wake_queue(dev);
+
+ out_8(®s->cantier, priv->tx_active);
+ ret = IRQ_HANDLED;
+ }
+
+ canrflg = in_8(®s->canrflg);
+ if ((canrflg & ~MSCAN_STAT_MSK) &&
+ !test_and_set_bit(F_RX_PROGRESS, &priv->flags)) {
+ if (canrflg & ~MSCAN_STAT_MSK) {
+ priv->shadow_canrier = in_8(®s->canrier);
+ out_8(®s->canrier, 0);
+ napi_schedule(&priv->napi);
+ ret = IRQ_HANDLED;
+ } else
+ clear_bit(F_RX_PROGRESS, &priv->flags);
+ }
+ return ret;
+}
+
+static int mscan_do_set_mode(struct net_device *dev, enum can_mode mode)
+{
+
+ struct mscan_priv *priv = netdev_priv(dev);
+ int ret = 0;
+
+ if (!priv->open_time)
+ return -EINVAL;
+
+ switch (mode) {
+ case CAN_MODE_SLEEP:
+ case CAN_MODE_STOP:
+ netif_stop_queue(dev);
+ mscan_set_mode(dev,
+ (mode ==
+ CAN_MODE_STOP) ? MSCAN_INIT_MODE :
+ MSCAN_SLEEP_MODE);
+ break;
+ case CAN_MODE_START:
+ if (priv->can.state <= CAN_STATE_BUS_OFF)
+ mscan_set_mode(dev, MSCAN_INIT_MODE);
+ ret = mscan_start(dev);
+ if (ret)
+ break;
+ if (netif_queue_stopped(dev))
+ netif_wake_queue(dev);
+ break;
+
+ default:
+ ret = -EOPNOTSUPP;
+ break;
+ }
+ return ret;
+}
+
+static int mscan_do_set_bittiming(struct net_device *dev)
+{
+ struct mscan_priv *priv = netdev_priv(dev);
+ struct mscan_regs *regs = (struct mscan_regs *)priv->reg_base;
+ struct can_bittiming *bt = &priv->can.bittiming;
+ u8 btr0, btr1;
+
+ btr0 = BTR0_SET_BRP(bt->brp) | BTR0_SET_SJW(bt->sjw);
+ btr1 = (BTR1_SET_TSEG1(bt->prop_seg + bt->phase_seg1) |
+ BTR1_SET_TSEG2(bt->phase_seg2) |
+ BTR1_SET_SAM(priv->can.ctrlmode & CAN_CTRLMODE_3_SAMPLES));
+
+ dev_info(dev->dev.parent, "setting BTR0=0x%02x BTR1=0x%02x\n",
+ btr0, btr1);
+
+ out_8(®s->canbtr0, btr0);
+ out_8(®s->canbtr1, btr1);
+
+ return 0;
+}
+
+static int mscan_open(struct net_device *dev)
+{
+ int ret;
+ struct mscan_priv *priv = netdev_priv(dev);
+ struct mscan_regs *regs = (struct mscan_regs *)priv->reg_base;
+
+ /* common open */
+ ret = open_candev(dev);
+ if (ret)
+ return ret;
+
+ napi_enable(&priv->napi);
+
+ ret = request_irq(dev->irq, mscan_isr, 0, dev->name, dev);
+ if (ret < 0) {
+ napi_disable(&priv->napi);
+ printk(KERN_ERR "%s - failed to attach interrupt\n",
+ dev->name);
+ return ret;
+ }
+
+ priv->open_time = jiffies;
+
+ out_8(®s->canctl1, in_8(®s->canctl1) & ~MSCAN_LISTEN);
+
+ ret = mscan_start(dev);
+ if (ret)
+ return ret;
+
+ netif_start_queue(dev);
+
+ return 0;
+}
+
+static int mscan_close(struct net_device *dev)
+{
+ struct mscan_priv *priv = netdev_priv(dev);
+ struct mscan_regs *regs = (struct mscan_regs *)priv->reg_base;
+
+ netif_stop_queue(dev);
+ napi_disable(&priv->napi);
+
+ out_8(®s->cantier, 0);
+ out_8(®s->canrier, 0);
+ mscan_set_mode(dev, MSCAN_INIT_MODE);
+ close_candev(dev);
+ free_irq(dev->irq, dev);
+ priv->open_time = 0;
+
+ return 0;
+}
+
+static const struct net_device_ops mscan_netdev_ops = {
+ .ndo_open = mscan_open,
+ .ndo_stop = mscan_close,
+ .ndo_start_xmit = mscan_start_xmit,
+};
+
+int register_mscandev(struct net_device *dev, int clock_src)
+{
+ struct mscan_priv *priv = netdev_priv(dev);
+ struct mscan_regs *regs = (struct mscan_regs *)priv->reg_base;
+ u8 ctl1;
+
+ ctl1 = in_8(®s->canctl1);
+ if (clock_src)
+ ctl1 |= MSCAN_CLKSRC;
+ else
+ ctl1 &= ~MSCAN_CLKSRC;
+
+ ctl1 |= MSCAN_CANE;
+ out_8(®s->canctl1, ctl1);
+ udelay(100);
+
+ /* acceptance mask/acceptance code (accept everything) */
+ out_be16(®s->canidar1_0, 0);
+ out_be16(®s->canidar3_2, 0);
+ out_be16(®s->canidar5_4, 0);
+ out_be16(®s->canidar7_6, 0);
+
+ out_be16(®s->canidmr1_0, 0xffff);
+ out_be16(®s->canidmr3_2, 0xffff);
+ out_be16(®s->canidmr5_4, 0xffff);
+ out_be16(®s->canidmr7_6, 0xffff);
+ /* Two 32 bit Acceptance Filters */
+ out_8(®s->canidac, MSCAN_AF_32BIT);
+
+ mscan_set_mode(dev, MSCAN_INIT_MODE);
+
+ return register_candev(dev);
+}
+EXPORT_SYMBOL_GPL(register_mscandev);
+
+void unregister_mscandev(struct net_device *dev)
+{
+ struct mscan_priv *priv = netdev_priv(dev);
+ struct mscan_regs *regs = (struct mscan_regs *)priv->reg_base;
+ mscan_set_mode(dev, MSCAN_INIT_MODE);
+ out_8(®s->canctl1, in_8(®s->canctl1) & ~MSCAN_CANE);
+ unregister_candev(dev);
+}
+EXPORT_SYMBOL_GPL(unregister_mscandev);
+
+struct net_device *alloc_mscandev(void)
+{
+ struct net_device *dev;
+ struct mscan_priv *priv;
+ int i;
+
+ dev = alloc_candev(sizeof(struct mscan_priv), MSCAN_ECHO_SKB_MAX);
+ if (!dev)
+ return NULL;
+ priv = netdev_priv(dev);
+
+ dev->netdev_ops = &mscan_netdev_ops;
+
+ dev->flags |= IFF_ECHO; /* we support local echo */
+
+ netif_napi_add(dev, &priv->napi, mscan_rx_poll, 8);
+
+ priv->can.bittiming_const = &mscan_bittiming_const;
+ priv->can.do_set_bittiming = mscan_do_set_bittiming;
+ priv->can.do_set_mode = mscan_do_set_mode;
+
+ for (i = 0; i < TX_QUEUE_SIZE; i++) {
+ priv->tx_queue[i].id = i;
+ priv->tx_queue[i].mask = 1 << i;
+ }
+
+ return dev;
+}
+EXPORT_SYMBOL_GPL(alloc_mscandev);
+
+MODULE_AUTHOR("Andrey Volkov <avolkov-ppI4tVfbJvJWk0Htik3J/w@public.gmane.org>");
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("CAN port driver for a MSCAN based chips");
diff --git a/drivers/net/can/mscan/mscan.h b/drivers/net/can/mscan/mscan.h
new file mode 100644
index 0000000..57820f5
--- /dev/null
+++ b/drivers/net/can/mscan/mscan.h
@@ -0,0 +1,262 @@
+/*
+ * Definitions of consts/structs to drive the Freescale MSCAN.
+ *
+ * Copyright (C) 2005-2006 Andrey Volkov <avolkov-ppI4tVfbJvJWk0Htik3J/w@public.gmane.org>,
+ * Varma Electronics Oy
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the version 2 of the GNU General Public License
+ * as published by the Free Software Foundation
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ */
+
+#ifndef __MSCAN_H__
+#define __MSCAN_H__
+
+#include <linux/types.h>
+
+/* MSCAN control register 0 (CANCTL0) bits */
+#define MSCAN_RXFRM 0x80
+#define MSCAN_RXACT 0x40
+#define MSCAN_CSWAI 0x20
+#define MSCAN_SYNCH 0x10
+#define MSCAN_TIME 0x08
+#define MSCAN_WUPE 0x04
+#define MSCAN_SLPRQ 0x02
+#define MSCAN_INITRQ 0x01
+
+/* MSCAN control register 1 (CANCTL1) bits */
+#define MSCAN_CANE 0x80
+#define MSCAN_CLKSRC 0x40
+#define MSCAN_LOOPB 0x20
+#define MSCAN_LISTEN 0x10
+#define MSCAN_WUPM 0x04
+#define MSCAN_SLPAK 0x02
+#define MSCAN_INITAK 0x01
+
+/* Use the MPC5200 MSCAN variant? */
+#ifdef CONFIG_PPC
+#define MSCAN_FOR_MPC5200
+#endif
+
+#ifdef MSCAN_FOR_MPC5200
+#define MSCAN_CLKSRC_BUS 0
+#define MSCAN_CLKSRC_XTAL MSCAN_CLKSRC
+#else
+#define MSCAN_CLKSRC_BUS MSCAN_CLKSRC
+#define MSCAN_CLKSRC_XTAL 0
+#endif
+
+/* MSCAN receiver flag register (CANRFLG) bits */
+#define MSCAN_WUPIF 0x80
+#define MSCAN_CSCIF 0x40
+#define MSCAN_RSTAT1 0x20
+#define MSCAN_RSTAT0 0x10
+#define MSCAN_TSTAT1 0x08
+#define MSCAN_TSTAT0 0x04
+#define MSCAN_OVRIF 0x02
+#define MSCAN_RXF 0x01
+#define MSCAN_ERR_IF (MSCAN_OVRIF | MSCAN_CSCIF)
+#define MSCAN_RSTAT_MSK (MSCAN_RSTAT1 | MSCAN_RSTAT0)
+#define MSCAN_TSTAT_MSK (MSCAN_TSTAT1 | MSCAN_TSTAT0)
+#define MSCAN_STAT_MSK (MSCAN_RSTAT_MSK | MSCAN_TSTAT_MSK)
+
+#define MSCAN_STATE_BUS_OFF (MSCAN_RSTAT1 | MSCAN_RSTAT0 | \
+ MSCAN_TSTAT1 | MSCAN_TSTAT0)
+#define MSCAN_STATE_TX(canrflg) (((canrflg)&MSCAN_TSTAT_MSK)>>2)
+#define MSCAN_STATE_RX(canrflg) (((canrflg)&MSCAN_RSTAT_MSK)>>4)
+#define MSCAN_STATE_ACTIVE 0
+#define MSCAN_STATE_WARNING 1
+#define MSCAN_STATE_PASSIVE 2
+#define MSCAN_STATE_BUSOFF 3
+
+/* MSCAN receiver interrupt enable register (CANRIER) bits */
+#define MSCAN_WUPIE 0x80
+#define MSCAN_CSCIE 0x40
+#define MSCAN_RSTATE1 0x20
+#define MSCAN_RSTATE0 0x10
+#define MSCAN_TSTATE1 0x08
+#define MSCAN_TSTATE0 0x04
+#define MSCAN_OVRIE 0x02
+#define MSCAN_RXFIE 0x01
+
+/* MSCAN transmitter flag register (CANTFLG) bits */
+#define MSCAN_TXE2 0x04
+#define MSCAN_TXE1 0x02
+#define MSCAN_TXE0 0x01
+#define MSCAN_TXE (MSCAN_TXE2 | MSCAN_TXE1 | MSCAN_TXE0)
+
+/* MSCAN transmitter interrupt enable register (CANTIER) bits */
+#define MSCAN_TXIE2 0x04
+#define MSCAN_TXIE1 0x02
+#define MSCAN_TXIE0 0x01
+#define MSCAN_TXIE (MSCAN_TXIE2 | MSCAN_TXIE1 | MSCAN_TXIE0)
+
+/* MSCAN transmitter message abort request (CANTARQ) bits */
+#define MSCAN_ABTRQ2 0x04
+#define MSCAN_ABTRQ1 0x02
+#define MSCAN_ABTRQ0 0x01
+
+/* MSCAN transmitter message abort ack (CANTAAK) bits */
+#define MSCAN_ABTAK2 0x04
+#define MSCAN_ABTAK1 0x02
+#define MSCAN_ABTAK0 0x01
+
+/* MSCAN transmit buffer selection (CANTBSEL) bits */
+#define MSCAN_TX2 0x04
+#define MSCAN_TX1 0x02
+#define MSCAN_TX0 0x01
+
+/* MSCAN ID acceptance control register (CANIDAC) bits */
+#define MSCAN_IDAM1 0x20
+#define MSCAN_IDAM0 0x10
+#define MSCAN_IDHIT2 0x04
+#define MSCAN_IDHIT1 0x02
+#define MSCAN_IDHIT0 0x01
+
+#define MSCAN_AF_32BIT 0x00
+#define MSCAN_AF_16BIT MSCAN_IDAM0
+#define MSCAN_AF_8BIT MSCAN_IDAM1
+#define MSCAN_AF_CLOSED (MSCAN_IDAM0|MSCAN_IDAM1)
+#define MSCAN_AF_MASK (~(MSCAN_IDAM0|MSCAN_IDAM1))
+
+/* MSCAN Miscellaneous Register (CANMISC) bits */
+#define MSCAN_BOHOLD 0x01
+
+#ifdef MSCAN_FOR_MPC5200
+#define _MSCAN_RESERVED_(n, num) u8 _res##n[num]
+#define _MSCAN_RESERVED_DSR_SIZE 2
+#else
+#define _MSCAN_RESERVED_(n, num)
+#define _MSCAN_RESERVED_DSR_SIZE 0
+#endif
+
+/* Structure of the hardware registers */
+struct mscan_regs {
+ /* (see doc S12MSCANV3/D) MPC5200 MSCAN */
+ u8 canctl0; /* + 0x00 0x00 */
+ u8 canctl1; /* + 0x01 0x01 */
+ _MSCAN_RESERVED_(1, 2); /* + 0x02 */
+ u8 canbtr0; /* + 0x04 0x02 */
+ u8 canbtr1; /* + 0x05 0x03 */
+ _MSCAN_RESERVED_(2, 2); /* + 0x06 */
+ u8 canrflg; /* + 0x08 0x04 */
+ u8 canrier; /* + 0x09 0x05 */
+ _MSCAN_RESERVED_(3, 2); /* + 0x0a */
+ u8 cantflg; /* + 0x0c 0x06 */
+ u8 cantier; /* + 0x0d 0x07 */
+ _MSCAN_RESERVED_(4, 2); /* + 0x0e */
+ u8 cantarq; /* + 0x10 0x08 */
+ u8 cantaak; /* + 0x11 0x09 */
+ _MSCAN_RESERVED_(5, 2); /* + 0x12 */
+ u8 cantbsel; /* + 0x14 0x0a */
+ u8 canidac; /* + 0x15 0x0b */
+ u8 reserved; /* + 0x16 0x0c */
+ _MSCAN_RESERVED_(6, 5); /* + 0x17 */
+#ifndef MSCAN_FOR_MPC5200
+ u8 canmisc; /* 0x0d */
+#endif
+ u8 canrxerr; /* + 0x1c 0x0e */
+ u8 cantxerr; /* + 0x1d 0x0f */
+ _MSCAN_RESERVED_(7, 2); /* + 0x1e */
+ u16 canidar1_0; /* + 0x20 0x10 */
+ _MSCAN_RESERVED_(8, 2); /* + 0x22 */
+ u16 canidar3_2; /* + 0x24 0x12 */
+ _MSCAN_RESERVED_(9, 2); /* + 0x26 */
+ u16 canidmr1_0; /* + 0x28 0x14 */
+ _MSCAN_RESERVED_(10, 2); /* + 0x2a */
+ u16 canidmr3_2; /* + 0x2c 0x16 */
+ _MSCAN_RESERVED_(11, 2); /* + 0x2e */
+ u16 canidar5_4; /* + 0x30 0x18 */
+ _MSCAN_RESERVED_(12, 2); /* + 0x32 */
+ u16 canidar7_6; /* + 0x34 0x1a */
+ _MSCAN_RESERVED_(13, 2); /* + 0x36 */
+ u16 canidmr5_4; /* + 0x38 0x1c */
+ _MSCAN_RESERVED_(14, 2); /* + 0x3a */
+ u16 canidmr7_6; /* + 0x3c 0x1e */
+ _MSCAN_RESERVED_(15, 2); /* + 0x3e */
+ struct {
+ u16 idr1_0; /* + 0x40 0x20 */
+ _MSCAN_RESERVED_(16, 2); /* + 0x42 */
+ u16 idr3_2; /* + 0x44 0x22 */
+ _MSCAN_RESERVED_(17, 2); /* + 0x46 */
+ u16 dsr1_0; /* + 0x48 0x24 */
+ _MSCAN_RESERVED_(18, 2); /* + 0x4a */
+ u16 dsr3_2; /* + 0x4c 0x26 */
+ _MSCAN_RESERVED_(19, 2); /* + 0x4e */
+ u16 dsr5_4; /* + 0x50 0x28 */
+ _MSCAN_RESERVED_(20, 2); /* + 0x52 */
+ u16 dsr7_6; /* + 0x54 0x2a */
+ _MSCAN_RESERVED_(21, 2); /* + 0x56 */
+ u8 dlr; /* + 0x58 0x2c */
+ u8:8; /* + 0x59 0x2d */
+ _MSCAN_RESERVED_(22, 2); /* + 0x5a */
+ u16 time; /* + 0x5c 0x2e */
+ } rx;
+ _MSCAN_RESERVED_(23, 2); /* + 0x5e */
+ struct {
+ u16 idr1_0; /* + 0x60 0x30 */
+ _MSCAN_RESERVED_(24, 2); /* + 0x62 */
+ u16 idr3_2; /* + 0x64 0x32 */
+ _MSCAN_RESERVED_(25, 2); /* + 0x66 */
+ u16 dsr1_0; /* + 0x68 0x34 */
+ _MSCAN_RESERVED_(26, 2); /* + 0x6a */
+ u16 dsr3_2; /* + 0x6c 0x36 */
+ _MSCAN_RESERVED_(27, 2); /* + 0x6e */
+ u16 dsr5_4; /* + 0x70 0x38 */
+ _MSCAN_RESERVED_(28, 2); /* + 0x72 */
+ u16 dsr7_6; /* + 0x74 0x3a */
+ _MSCAN_RESERVED_(29, 2); /* + 0x76 */
+ u8 dlr; /* + 0x78 0x3c */
+ u8 tbpr; /* + 0x79 0x3d */
+ _MSCAN_RESERVED_(30, 2); /* + 0x7a */
+ u16 time; /* + 0x7c 0x3e */
+ } tx;
+ _MSCAN_RESERVED_(31, 2); /* + 0x7e */
+} __attribute__ ((packed));
+
+#undef _MSCAN_RESERVED_
+#define MSCAN_REGION sizeof(struct mscan)
+
+#define TX_QUEUE_SIZE 3
+
+struct tx_queue_entry {
+ struct list_head list;
+ u8 mask;
+ u8 id;
+};
+
+struct mscan_priv {
+ struct can_priv can; /* must be the first member */
+ long open_time;
+ unsigned long flags;
+ void __iomem *reg_base; /* ioremap'ed address to registers */
+ u8 shadow_statflg;
+ u8 shadow_canrier;
+ u8 cur_pri;
+ u8 prev_buf_id;
+ u8 tx_active;
+
+ struct list_head tx_head;
+ struct tx_queue_entry tx_queue[TX_QUEUE_SIZE];
+ struct napi_struct napi;
+};
+
+struct net_device *alloc_mscandev(void);
+/*
+ * clock_src:
+ * 1 = The MSCAN clock source is the onchip Bus Clock.
+ * 0 = The MSCAN clock source is the chip Oscillator Clock.
+ */
+extern int register_mscandev(struct net_device *dev, int clock_src);
+extern void unregister_mscandev(struct net_device *dev);
+
+#endif /* __MSCAN_H__ */
--
1.6.3.3
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox