* [PATCH] ip : fix truesize mismatch in ip fragmentation
From: Eric Dumazet @ 2010-09-21 14:16 UTC (permalink / raw)
To: Nick Bowler; +Cc: linux-kernel, netdev, David S. Miller
In-Reply-To: <20100921140501.GA21572@elliptictech.com>
Le mardi 21 septembre 2010 à 10:05 -0400, Nick Bowler a écrit :
> This hunk introduces some whitespace damage.
>
> Anyway, I tried this with ESP on both IPv4 and IPv6 and it appears to
> correct the issue. Thanks!
>
Indeed good catch.
Here is an updated patch, I added your Tested-by
Thanks for testing !
[PATCH] ip : fix truesize mismatch in ip fragmentation
We should not set frag->destructor to sock_wkfree() until we are sure we
dont hit slow path in ip_fragment(). Or we risk uncharging
frag->truesize twice, and in the end, having negative socket
sk_wmem_alloc counter, or even freeing socket sooner than expected.
Many thanks to Nick Bowler, who provided a very clean bug report and
test program.
While Nick bisection pointed to commit 2b85a34e911bf483 (net: No more
expensive sock_hold()/sock_put() on each tx), underlying bug is older.
Reported-and-bisected-by: Nick Bowler <nbowler@elliptictech.com>
Tested-by: Nick Bowler <nbowler@elliptictech.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
---
net/ipv4/ip_output.c | 8 ++++----
net/ipv6/ip6_output.c | 8 ++++----
2 files changed, 8 insertions(+), 8 deletions(-)
diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
index 04b6989..126d9b3 100644
--- a/net/ipv4/ip_output.c
+++ b/net/ipv4/ip_output.c
@@ -490,7 +490,6 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
if (skb_has_frags(skb)) {
struct sk_buff *frag;
int first_len = skb_pagelen(skb);
- int truesizes = 0;
if (first_len - hlen > mtu ||
((first_len - hlen) & 7) ||
@@ -510,11 +509,13 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
goto slow_path;
BUG_ON(frag->sk);
- if (skb->sk) {
+ }
+ if (skb->sk) {
+ skb_walk_frags(skb, frag) {
frag->sk = skb->sk;
frag->destructor = sock_wfree;
+ skb->truesize -= frag->truesize;
}
- truesizes += frag->truesize;
}
/* Everything is OK. Generate! */
@@ -524,7 +525,6 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
frag = skb_shinfo(skb)->frag_list;
skb_frag_list_init(skb);
skb->data_len = first_len - skb_headlen(skb);
- skb->truesize -= truesizes;
skb->len = first_len;
iph->tot_len = htons(first_len);
iph->frag_off = htons(IP_MF);
diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index d40b330..633217d 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -639,7 +639,6 @@ static int ip6_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
if (skb_has_frags(skb)) {
int first_len = skb_pagelen(skb);
- int truesizes = 0;
if (first_len - hlen > mtu ||
((first_len - hlen) & 7) ||
@@ -658,10 +657,12 @@ static int ip6_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
goto slow_path;
BUG_ON(frag->sk);
- if (skb->sk) {
+ }
+ if (skb->sk) {
+ skb_walk_frags(skb, frag) {
frag->sk = skb->sk;
frag->destructor = sock_wfree;
- truesizes += frag->truesize;
+ skb->truesize -= frag->truesize;
}
}
@@ -693,7 +694,6 @@ static int ip6_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
first_len = skb_pagelen(skb);
skb->data_len = first_len - skb_headlen(skb);
- skb->truesize -= truesizes;
skb->len = first_len;
ipv6_hdr(skb)->payload_len = htons(first_len -
sizeof(struct ipv6hdr));
^ permalink raw reply related
* Re: Regression, bisected: reference leak with IPSec since ~2.6.31
From: Nick Bowler @ 2010-09-21 14:05 UTC (permalink / raw)
To: Eric Dumazet; +Cc: linux-kernel, netdev, David S. Miller
In-Reply-To: <1285018272.2323.243.camel@edumazet-laptop>
On 2010-09-20 23:31 +0200, Eric Dumazet wrote:
> Could you try following patch ?
>
> [PATCH] ip : fix truesize mismatch in ip fragmentation
[...]
> @@ -658,13 +657,15 @@ static int ip6_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
> goto slow_path;
>
> BUG_ON(frag->sk);
> - if (skb->sk) {
> + }
> + if (skb->sk) {
> + skb_walk_frags(skb, frag) {
> frag->sk = skb->sk;
> frag->destructor = sock_wfree;
> - truesizes += frag->truesize;
> + skb->truesize -= frag->truesize;
> }
> }
> -
> +
This hunk introduces some whitespace damage.
Anyway, I tried this with ESP on both IPv4 and IPv6 and it appears to
correct the issue. Thanks!
--
Nick Bowler, Elliptic Technologies (http://www.elliptictech.com/)
^ permalink raw reply
* Re: [RFC PATCH v9 12/16] Add mp(mediate passthru) device.
From: Michael S. Tsirkin @ 2010-09-21 13:14 UTC (permalink / raw)
To: Xin, Xiaohui
Cc: netdev@vger.kernel.org, kvm@vger.kernel.org,
linux-kernel@vger.kernel.org, mingo@elte.hu, davem@davemloft.net,
herbert@gondor.hengli.com.au, jdike@linux.intel.com
In-Reply-To: <F2E9EB7348B8264F86B6AB8151CE2D792B8CAA0F6A@shsmsx502.ccr.corp.intel.com>
On Tue, Sep 21, 2010 at 09:39:31AM +0800, Xin, Xiaohui wrote:
> >From: Michael S. Tsirkin [mailto:mst@redhat.com]
> >Sent: Monday, September 20, 2010 7:37 PM
> >To: Xin, Xiaohui
> >Cc: netdev@vger.kernel.org; kvm@vger.kernel.org; linux-kernel@vger.kernel.org;
> >mingo@elte.hu; davem@davemloft.net; herbert@gondor.hengli.com.au;
> >jdike@linux.intel.com
> >Subject: Re: [RFC PATCH v9 12/16] Add mp(mediate passthru) device.
> >
> >On Mon, Sep 20, 2010 at 04:08:48PM +0800, xiaohui.xin@intel.com wrote:
> >> From: Xin Xiaohui <xiaohui.xin@intel.com>
> >>
> >> ---
> >> Michael,
> >> I have move the ioctl to configure the locked memory to vhost
> >
> >It's ok to move this to vhost but vhost does not
> >know how much memory is needed by the backend.
>
> I think the backend here you mean is mp device.
> Actually, the memory needed is related to vq->num to run zero-copy
> smoothly.
> That means mp device did not know it but vhost did.
Well, this might be so if you insist on locking
all posted buffers immediately. However, let's assume I have a
very large ring and prepost a ton of RX buffers:
there's no need to lock all of them directly:
if we have buffers A and B, we can lock A, pass it
to hardware, and when A is consumed unlock A, lock B
and pass it to hardware.
It's not really critical. But note we can always have userspace
tell MP device all it wants to know, after all.
> And the rlimt stuff is per process, we use current pointer to set
> and check the rlimit, the operations should be in the same process.
Well no, the ring is handled from the kernel thread: we switch the mm to
point to the owner task so copy from/to user and friends work, but you
can't access the rlimit etc.
> Now the check operations are in vhost process, as mp_recvmsg() or
> mp_sendmsg() are called by vhost.
Hmm, what do you mean by the check operations?
send/recv are data path operations, they shouldn't
do any checks, should they?
> So set operations should be in
> vhost process too, it's natural.
>
> >So I think we'll need another ioctl in the backend
> >to tell userspace how much memory is needed?
> >
> Except vhost tells it to mp device, mp did not know
> how much memory is needed to run zero-copy smoothly.
> Is userspace interested about the memory mp is needed?
Couldn't parse this last question.
I think userspace generally does want control over
how much memory we'll lock. We should not just lock
as much as we can.
--
MST
^ permalink raw reply
* Re: Regression, bisected: reference leak with IPSec since ~2.6.31
From: Jarek Poplawski @ 2010-09-21 12:39 UTC (permalink / raw)
To: Eric Dumazet; +Cc: Nick Bowler, linux-kernel, netdev, David S. Miller
In-Reply-To: <1285070310.2617.359.camel@edumazet-laptop>
On Tue, Sep 21, 2010 at 01:58:30PM +0200, Eric Dumazet wrote:
> Le mardi 21 septembre 2010 ?? 10:48 +0000, Jarek Poplawski a écrit :
>
> > But if you were honest about caches there is no reason to be smart
> > no more. ;-)
> >
> > Below is your patch mostly as an example only that I didn't mean
> > nothing complex (ipv4 only, not compiled).
> >
> > Jarek P.
> > ---
> >
> > diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
> > index 04b6989..049d61f 100644
> > --- a/net/ipv4/ip_output.c
> > +++ b/net/ipv4/ip_output.c
> > @@ -490,7 +490,6 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
> > if (skb_has_frags(skb)) {
> > struct sk_buff *frag;
> > int first_len = skb_pagelen(skb);
> > - int truesizes = 0;
> >
> > if (first_len - hlen > mtu ||
> > ((first_len - hlen) & 7) ||
> > @@ -503,18 +502,18 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
> > if (frag->len > mtu ||
> > ((frag->len & 7) && frag->next) ||
> > skb_headroom(frag) < hlen)
> > - goto slow_path;
> > + goto slow_path_clean;
> >
> > /* Partially cloned skb? */
> > if (skb_shared(frag))
> > - goto slow_path;
> > + goto slow_path_clean;
> >
> > BUG_ON(frag->sk);
> > if (skb->sk) {
> > frag->sk = skb->sk;
> > frag->destructor = sock_wfree;
> > + skb->truesize -= frag->truesize;
> > }
> > - truesizes += frag->truesize;
> > }
> >
> > /* Everything is OK. Generate! */
> > @@ -524,7 +523,6 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
> > frag = skb_shinfo(skb)->frag_list;
> > skb_frag_list_init(skb);
> > skb->data_len = first_len - skb_headlen(skb);
> > - skb->truesize -= truesizes;
> > skb->len = first_len;
> > iph->tot_len = htons(first_len);
> > iph->frag_off = htons(IP_MF);
> > @@ -576,6 +574,17 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
> > }
> > IP_INC_STATS(dev_net(dev), IPSTATS_MIB_FRAGFAILS);
> > return err;
> > +
> > +slow_path_clean:
> > + if (skb->sk) {
> > + skb_walk_frags(skb, frag) {
> > + if (!frag->sk)
> > + break;
> > + frag->sk = NULL;
> > + frag->destructor = NULL;
> > + skb->truesize += frag->truesize;
> > + }
> > + }
> > }
> >
> > slow_path:
>
> Nice try, but this adds a bug unfortunately, or if you prefer makes the
> "BUG_ON(frag->sk);" being not fully covered.
>
> If we hit slow path, and one or more frags had frag->sk set, we silently
> unset it, instead of a nice BUG_ON().
Which exactly BUG_ON() on the slow path do you mean?
>
> If you really want to have a slow_path_clean, you should use a different
> frag2 iterator :
>
> slow_path_clean:
> if (skb->sk) {
> skb_walk_frags(skb, frag2) {
> if (frag2 == frag)
> break;
> frag2->sk = NULL;
> frag2->destructor = NULL;
> skb->truesize += frag2->truesize;
> }
> }
>
> Face it, this is too complex for net-2.6, could we just fix the bug, and
> 'optimize' later ?
I thought it's clear I didn't intend to change anything with this
example, but in any case I admit it: I'm 100% OK with your patch.
Jarek P.
^ permalink raw reply
* Re: [RFC][PATCH 1/4] IRQ: IRQ groups for multiqueue devices
From: Ben Hutchings @ 2010-09-21 12:25 UTC (permalink / raw)
To: Thomas Gleixner
Cc: Tom Herbert, netdev, linux-net-drivers, linux-kernel,
Peter Zijlstra, Ingo Molnar
In-Reply-To: <alpine.LFD.2.00.1009202137580.2416@localhost6.localdomain6>
On Mon, 2010-09-20 at 23:27 +0200, Thomas Gleixner wrote:
> Ben,
>
> On Mon, 20 Sep 2010, Ben Hutchings wrote:
>
> it would have been nice if I'd been cc'ed on that, but of course it's
> my fault that there is no entry in MAINTAINERS. No, it's not.
>
> > When initiating I/O on multiqueue devices, we usually want to select a
>
> "we usually" is pretty useless for people not familiar with the
> problem at hand. A [PATCH 0/4] intro would have been helpful along
> with the users (aka [PATCH 2-4/4]) of the new facility.
There was an intro on the netdev list
<http://article.gmane.org/gmane.linux.network/172427> but it probably
doesn't tell you what you want to know anyway.
> Don't take it personally and I'm sure that you're solving a real world
> problem, but I'm starting to get grumpy that especially networking
> folks (aside of various driver developers) believe that adding random
> stuff to kernel/irq is their private pleasure. Is it that hard to talk
> to me upfront ?
>
> > queue for which the response will be handled on the same or a nearby
> > CPU. IRQ groups hold a mapping of CPU to IRQ which will be updated
> > based on the inverse of IRQ CPU-affinities plus CPU topology
> > information.
>
> Can you please explain, why you need that reverse mapping including
> the below code ? What problem does this solve which can not be deduced
> by the exisiting information/infrastructure ? And why is that reverse
> mapping tied to interrupts and not something which we want to see in
> some generic available (library) code ?
Are you thinking of a per-CPU distance map? That seems like it would be
useful to have. I wonder about the storage requirements on larger
systems.
> More comments inline.
>
> > ---
> > include/linux/irq.h | 52 ++++++++++++++++++
> > kernel/irq/manage.c | 149 +++++++++++++++++++++++++++++++++++++++++++++++++++
> > 2 files changed, 201 insertions(+), 0 deletions(-)
> >
> > diff --git a/include/linux/irq.h b/include/linux/irq.h
> > index c03243a..bbddd5f 100644
> > --- a/include/linux/irq.h
> > +++ b/include/linux/irq.h
> > @@ -196,6 +196,8 @@ struct irq_desc {
> > #ifdef CONFIG_SMP
> > cpumask_var_t affinity;
> > const struct cpumask *affinity_hint;
>
> How about updating the kernel doc above the struct ?
>
> > + struct irq_group *group;
>
> Grr, how does this compile ? That needs at least a forward
> declaration of struct irq_group. RFC is _NOT_ an excuse
Naming a struct in a function parameter declarator is not a valid
forward declaration, but this is.
> > + u16 group_index;
>
> What's group_index doing and what's the point of an u16 here ?
Index of this IRQ within the group.
> > unsigned int node;
> > #ifdef CONFIG_GENERIC_PENDING_IRQ
> > cpumask_var_t pending_mask;
> > @@ -498,6 +500,33 @@ static inline void free_desc_masks(struct irq_desc *old_desc,
> > #endif
> > }
> >
> > +/**
> > + * struct irq_group - IRQ group for multiqueue devices
> > + * @closest: For each CPU, the index and distance to the closest IRQ,
> > + * based on affinity masks
>
> index of what ?
Index within the group of the closest IRQ for this CPU; also the queue
index.
[...]
> > +static inline u16 irq_group_get_index(struct irq_group *group, int cpu)
> > +{
> > + return group->closest[cpu].index;
> > +}
> > +
>
> So you have an accessor function for closest[cpu].index. Are the
> other members meant to be accessible directly by random driver ?
No, they should not be touched at all.
> > #else /* !CONFIG_SMP */
> >
> > static inline bool alloc_desc_masks(struct irq_desc *desc, int node,
> > @@ -519,6 +548,29 @@ static inline void free_desc_masks(struct irq_desc *old_desc,
> > struct irq_desc *new_desc)
> > {
> > }
> > +
> > +struct irq_group {
> > +};
> > +
> > +static inline struct irq_group *alloc_irq_group(unsigned int size, gfp_t flags)
> > +{
> > + static struct irq_group dummy;
>
> That will create one static instance per callsite. Is that on
> purpose? If yes, it needs a damned good comment.
Which uses no storage, or maybe 1 byte. I'm open to suggestions for a
better dummy implementation.
> > + return &dummy;
>
> > #endif /* CONFIG_SMP */
> >
> > #endif /* _LINUX_IRQ_H */
> > diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c
> > index c3003e9..3f2b1a9 100644
> > --- a/kernel/irq/manage.c
> > +++ b/kernel/irq/manage.c
> > @@ -100,6 +100,154 @@ void irq_set_thread_affinity(struct irq_desc *desc)
> > }
> > }
> >
> > +static void irq_group_update_neigh(struct irq_group *group,
> > + const struct cpumask *mask,
> > + u16 index, u16 dist)
> > +{
> > + int cpu;
> > +
> > + for_each_cpu(cpu, mask) {
> > + if (dist < group->closest[cpu].dist) {
> > + group->closest[cpu].index = index;
> > + group->closest[cpu].dist = dist;
> > + }
> > + }
> > +}
>
> I have not the faintest idea how group, index and dist are related
> to each other. And I have no intention to decode the information
> about that piecewise by reverse engineering that obfuscated code.
I'm happy to write more comments.
index is the index of the IRQ being updated within the group. dist is
the maximum distance of that IRQ from the CPUs in the mask.
> > +static bool irq_group_copy_neigh(struct irq_group *group, int cpu,
> > + const struct cpumask *mask, u16 dist)
> > +{
> > + int neigh;
> > +
> > + for_each_cpu(neigh, mask) {
> > + if (group->closest[neigh].dist <= dist) {
> > + group->closest[cpu].index = group->closest[neigh].index;
> > + group->closest[cpu].dist = dist;
> > + return true;
> > + }
> > + }
> > + return false;
>
> What's the reason for copying or not ?
We're trying to find closer IRQs for the CPUs which were removed from
the affinity of the IRQ currently being moved.
[...]
> > + /* We could continue into NUMA node distances, but for now
> > + * we give up. */
>
> What are the consequences of giving up ? Does not happen ? Should
> not happen ? Will break ? Don't care ? ....
Poor choice of closest IRQ. Nothing disastrous.
> > + }
>
> This is called from irq_set_affinity() with desc->lock held and
> interrupts disabled. You're not serious about that, are you ?
>
> Damned, there are two iterations over each online cpu and another
> one over the affinity mask. Did you ever extrapolate how long that
> runs on a really large machine ?
No, and I know this sucks. So, do you want to suggest anything better?
Should I look at building those CPU distance maps (trading space for
time)?
> If CPU affinity of an IRQ changes, then it does not matter if one or
> two interrupts end up in the wrong space. Those changes are not
> happening every other interrupt.
>
> > +}
> > +
> > +/**
> > + * alloc_irq_group - allocate IRQ group
> > + * @size: Size of the group
>
> size of what ? I guess number of interrupts, right ?
Right.
> > + * @flags: Allocation flags e.g. %GFP_KERNEL
> > + */
> > +struct irq_group *alloc_irq_group(unsigned int size, gfp_t flags)
> > +{
> > + struct irq_group *group =
> > + kzalloc(sizeof(*group) + size * sizeof(group->irq[0]), flags);
> > + int cpu;
> > +
> > + if (!group)
> > + return NULL;
> > +
> > + /* Initially assign CPUs to IRQs on a rota */
> > + for (cpu = 0; cpu < NR_CPUS; cpu++) {
> > + group->closest[cpu].index = cpu % size;
>
> So here we randomly assign index with the lower cpu numbers no
> matter whether they are online or possible ?
Doesn't matter, because we're mapping CPUs to IRQs not the other way
round.
> > + group->closest[cpu].dist = IRQ_CPU_DIST_INF;
> > + }
> > +
> > + group->size = size;
> > + return group;
> > +}
> > +EXPORT_SYMBOL(alloc_irq_group);
>
> EXPORT_SYMBOL_GPL if at all. Same for the other exports
>
> > +
> > +/**
> > + * free_irq_group - free IRQ group
> > + * @group: IRQ group allocated with alloc_irq_group(), or %NULL
>
> How is this serialized or sanity checked against free_irq() ?
free_irq() should be called first, for all IRQs in the group. The IRQs
should then be really freed with pci_disable_msix() or similar.
> > + */
> > +void free_irq_group(struct irq_group *group)
> > +{
> > + struct irq_desc *desc;
> > + unsigned int i;
> > +
> > + if (!group)
> > + return;
> > +
> > + /* Remove all descriptors from the group */
> > + for (i = 0; i < group->used; i++) {
> > + desc = group->irq[i];
> > + BUG_ON(desc->group != group || desc->group_index != i);
> > + desc->group = NULL;
> > + }
> > +
> > + kfree(group);
> > +}
> > +EXPORT_SYMBOL(free_irq_group);
> > +
> > +/**
> > + * irq_group_add - add IRQ to a group
> > + * @group: IRQ group allocated with alloc_irq_group()
> > + * @irq: Interrupt to add to group
> > + */
> > +void irq_group_add(struct irq_group *group, unsigned int irq)
> > +{
> > + struct irq_desc *desc = irq_to_desc(irq);
>
> Again, how is this serialized against anything else fiddling with
> irq_desc[irq]?
The driver should call this after allocating IRQs with pci_enable_msix()
or similar and before setting the handlers with request_irq(). Is that
sufficient?
Ben.
> > + BUG_ON(desc->group);
> > + BUG_ON(group->used >= group->size);
> > +
> > + desc->group = group;
> > + desc->group_index = group->used;
> > + group->irq[group->used++] = desc;
> > +}
> > +EXPORT_SYMBOL(irq_group_add);
>
> Thanks,
>
> tglx
--
Ben Hutchings, Senior Software Engineer, Solarflare Communications
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.
^ permalink raw reply
* [BUG,E1000E] first packets after device is reported up are silently dropped
From: Arnaud Ebalard @ 2010-09-21 12:07 UTC (permalink / raw)
To: Jeff Kirsher, Jesse Brandeburg, Bruce Allan, Alex Duyck, PJ
Cc: netdev, Brian Haley, Alexey Kuznetsov, Stefan Rompf, David Miller
Hi,
When the link is reported up again (after plugging a cable) by the
driver (E1000E) the first packets sent immediately after that event
are sometimes *silently* dropped by the hardware.
Before describing the tests, here are some info on the hardware and
software. Don't hesitate to ask if you need more:
Kernel: 2.6.35.4
Hardware: Intel 82567LM (rev3) Gigagbit adapter on a DELL E4300
Here is what ethtool reports for the driver:
driver: e1000e
version: 1.0.2-k4
firmware-version: 1.7-7
bus-info: 0000:00:19.0
Switch: tested with a Cisco Catalyst 2960 (100Mbits/s), Planex
FX08-Mini (100Mbit/s), PLanex 5 ports Gigabit
The setup is pretty simple: two different userland tools (umip and
netplug) monitor netlink NEWLINK events and respectively send an ICMPv6
Router Solicitation packet and an IPv4 DHCP request when they receive
the information the interface is UP and RUNNING:
00:21:70:bd:ef:fc > 33:33:00:00:00:02, ethertype IPv6 (0x86dd), length 62: :: > ff02::2: ICMP6, router solicitation, length 8
00:21:70:bd:ef:fc > ff:ff:ff:ff:ff:ff, ethertype IPv4 (0x0800), length 342: 0.0.0.0.68 > 255.255.255.255.67: BOOTP/DHCP, Request from
The idea is to have address autoconfiguration performed as soon as
possible. Because the reemission occurs only after a few seconds, the
net result is a long delay.
I noticed that sometimes the first few packets emitted by the tools are
not answered. I put a tcpdump on the other side. Nothing arrives. I even
checked the led on the switch. Does not blink.
I first thought adding a few ms of delay beteween the reception of the
NEWLINK and the emission of the packets. It seems the higher the better
but at *550ms* I still managed to have the initial packet dropped from
time to time.
I then spent time in the kernel (net/core/dev.c, net/sched/sch_generic.c
drivers/net/e1000e/netdev.c) following the first packet to see where it
gets dropped. I ended up in e1000_xmit_frame() in which everything seems
to be ok. AFAICT, the packet is delivered to the hardware and then
silently killed for some unknown reason.
I added various debug statements in the code (custom printk(),
calls to e1000e_dump()) to try and understand what can be different in
the driver's state when the first packet is deliver and when it is
not. Nothing interesting.
I am currently out of idea. Is this a known bug? What can be happening?
If you have patches you want me to test to get additional info, don't
hesitate!
Cheers,
a+
^ permalink raw reply
* Re: Regression, bisected: reference leak with IPSec since ~2.6.31
From: Eric Dumazet @ 2010-09-21 11:58 UTC (permalink / raw)
To: Jarek Poplawski; +Cc: Nick Bowler, linux-kernel, netdev, David S. Miller
In-Reply-To: <20100921104821.GA8986@ff.dom.local>
Le mardi 21 septembre 2010 à 10:48 +0000, Jarek Poplawski a écrit :
> But if you were honest about caches there is no reason to be smart
> no more. ;-)
>
> Below is your patch mostly as an example only that I didn't mean
> nothing complex (ipv4 only, not compiled).
>
> Jarek P.
> ---
>
> diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
> index 04b6989..049d61f 100644
> --- a/net/ipv4/ip_output.c
> +++ b/net/ipv4/ip_output.c
> @@ -490,7 +490,6 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
> if (skb_has_frags(skb)) {
> struct sk_buff *frag;
> int first_len = skb_pagelen(skb);
> - int truesizes = 0;
>
> if (first_len - hlen > mtu ||
> ((first_len - hlen) & 7) ||
> @@ -503,18 +502,18 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
> if (frag->len > mtu ||
> ((frag->len & 7) && frag->next) ||
> skb_headroom(frag) < hlen)
> - goto slow_path;
> + goto slow_path_clean;
>
> /* Partially cloned skb? */
> if (skb_shared(frag))
> - goto slow_path;
> + goto slow_path_clean;
>
> BUG_ON(frag->sk);
> if (skb->sk) {
> frag->sk = skb->sk;
> frag->destructor = sock_wfree;
> + skb->truesize -= frag->truesize;
> }
> - truesizes += frag->truesize;
> }
>
> /* Everything is OK. Generate! */
> @@ -524,7 +523,6 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
> frag = skb_shinfo(skb)->frag_list;
> skb_frag_list_init(skb);
> skb->data_len = first_len - skb_headlen(skb);
> - skb->truesize -= truesizes;
> skb->len = first_len;
> iph->tot_len = htons(first_len);
> iph->frag_off = htons(IP_MF);
> @@ -576,6 +574,17 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
> }
> IP_INC_STATS(dev_net(dev), IPSTATS_MIB_FRAGFAILS);
> return err;
> +
> +slow_path_clean:
> + if (skb->sk) {
> + skb_walk_frags(skb, frag) {
> + if (!frag->sk)
> + break;
> + frag->sk = NULL;
> + frag->destructor = NULL;
> + skb->truesize += frag->truesize;
> + }
> + }
> }
>
> slow_path:
Nice try, but this adds a bug unfortunately, or if you prefer makes the
"BUG_ON(frag->sk);" being not fully covered.
If we hit slow path, and one or more frags had frag->sk set, we silently
unset it, instead of a nice BUG_ON().
If you really want to have a slow_path_clean, you should use a different
frag2 iterator :
slow_path_clean:
if (skb->sk) {
skb_walk_frags(skb, frag2) {
if (frag2 == frag)
break;
frag2->sk = NULL;
frag2->destructor = NULL;
skb->truesize += frag2->truesize;
}
}
Face it, this is too complex for net-2.6, could we just fix the bug, and
'optimize' later ?
^ permalink raw reply
* [PATCH v5] Gigabit Ethernet driver of Topcliff PCH
From: Masayuki Ohtake @ 2010-09-21 11:44 UTC (permalink / raw)
To: David S. Miller, Randy Dunlap, John Linn, Ralf Baechle,
Kristoffer Glembo
Cc: LKML, Toshiharu Okada, Masayuki Ohtake, Takahiro Shimizu,
Tomoya Morinaga, MeeGo, Wang, Qi, Wang, Yong Y, Andrew, Intel OTC,
Foster, Margie
The dev_pm_ops was used.
Unnecessary "__attribute__ ((packed))" was deleted.
Signed-off-by: Masayuki Ohtake <masa-korg@dsn.okisemi.com>
---
drivers/net/Kconfig | 12 +
drivers/net/Makefile | 1 +
drivers/net/pch_gbe/Makefile | 4 +
drivers/net/pch_gbe/pch_gbe.h | 658 +++++++++
drivers/net/pch_gbe/pch_gbe_api.c | 245 ++++
drivers/net/pch_gbe/pch_gbe_api.h | 36 +
drivers/net/pch_gbe/pch_gbe_ethtool.c | 584 ++++++++
drivers/net/pch_gbe/pch_gbe_main.c | 2473 +++++++++++++++++++++++++++++++++
drivers/net/pch_gbe/pch_gbe_param.c | 501 +++++++
drivers/net/pch_gbe/pch_gbe_phy.c | 274 ++++
drivers/net/pch_gbe/pch_gbe_phy.h | 37 +
11 files changed, 4825 insertions(+), 0 deletions(-)
create mode 100644 drivers/net/pch_gbe/Makefile
create mode 100644 drivers/net/pch_gbe/pch_gbe.h
create mode 100644 drivers/net/pch_gbe/pch_gbe_api.c
create mode 100644 drivers/net/pch_gbe/pch_gbe_api.h
create mode 100644 drivers/net/pch_gbe/pch_gbe_ethtool.c
create mode 100644 drivers/net/pch_gbe/pch_gbe_main.c
create mode 100644 drivers/net/pch_gbe/pch_gbe_param.c
create mode 100644 drivers/net/pch_gbe/pch_gbe_phy.c
create mode 100644 drivers/net/pch_gbe/pch_gbe_phy.h
diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig
index 2decc59..cf3aff8 100644
--- a/drivers/net/Kconfig
+++ b/drivers/net/Kconfig
@@ -2502,6 +2502,18 @@ config S6GMAC
source "drivers/net/stmmac/Kconfig"
+config PCH_GBE
+ tristate "PCH Gigabit Ethernet"
+ depends on PCI
+ ---help---
+ This is a gigabit ethernet driver for Topcliff PCH.
+ Topcliff PCH is the platform controller hub that is used in Intel's
+ general embedded platform.
+ Topcliff PCH has Gigabit Ethernet interface.
+ Using this interface, it is able to access system devices connected
+ to Gigabit Ethernet.
+ This driver enables Gigabit Ethernet function.
+
endif # NETDEV_1000
#
diff --git a/drivers/net/Makefile b/drivers/net/Makefile
index 0a0512a..3df2467 100644
--- a/drivers/net/Makefile
+++ b/drivers/net/Makefile
@@ -295,3 +295,4 @@ obj-$(CONFIG_WIMAX) += wimax/
obj-$(CONFIG_CAIF) += caif/
obj-$(CONFIG_OCTEON_MGMT_ETHERNET) += octeon/
+obj-$(CONFIG_PCH_GBE) += pch_gbe/
diff --git a/drivers/net/pch_gbe/Makefile b/drivers/net/pch_gbe/Makefile
new file mode 100644
index 0000000..31288d4
--- /dev/null
+++ b/drivers/net/pch_gbe/Makefile
@@ -0,0 +1,4 @@
+obj-$(CONFIG_PCH_GBE) += pch_gbe.o
+
+pch_gbe-y := pch_gbe_phy.o pch_gbe_ethtool.o pch_gbe_param.o
+pch_gbe-y += pch_gbe_api.o pch_gbe_main.o
diff --git a/drivers/net/pch_gbe/pch_gbe.h b/drivers/net/pch_gbe/pch_gbe.h
new file mode 100644
index 0000000..91250cd
--- /dev/null
+++ b/drivers/net/pch_gbe/pch_gbe.h
@@ -0,0 +1,658 @@
+/*
+ * Copyright (C) 1999 - 2010 Intel Corporation.
+ * Copyright (C) 2010 OKI SEMICONDUCTOR Co., LTD.
+ *
+ * This code was derived from the Intel e1000e Linux driver.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * 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 _PCH_GBE_H_
+#define _PCH_GBE_H_
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
+#include <linux/mii.h>
+#include <linux/delay.h>
+#include <linux/pci.h>
+#include <linux/netdevice.h>
+#include <linux/etherdevice.h>
+#include <linux/ethtool.h>
+#include <net/ip.h>
+
+/**
+ * pch_gbe_regs_mac_adr - Structure holding values of mac address registers
+ * @high Denotes the 1st to 4th byte from the initial of MAC address
+ * @low Denotes the 5th to 6th byte from the initial of MAC address
+ */
+struct pch_gbe_regs_mac_adr {
+ u32 high;
+ u32 low;
+};
+/**
+ * pch_udc_regs - Structure holding values of MAC registers
+ */
+struct pch_gbe_regs {
+ u32 INT_ST;
+ u32 INT_EN;
+ u32 MODE;
+ u32 RESET;
+ u32 TCPIP_ACC;
+ u32 EX_LIST;
+ u32 INT_ST_HOLD;
+ u32 PHY_INT_CTRL;
+ u32 MAC_RX_EN;
+ u32 RX_FCTRL;
+ u32 PAUSE_REQ;
+ u32 RX_MODE;
+ u32 TX_MODE;
+ u32 RX_FIFO_ST;
+ u32 TX_FIFO_ST;
+ u32 TX_FID;
+ u32 TX_RESULT;
+ u32 PAUSE_PKT1;
+ u32 PAUSE_PKT2;
+ u32 PAUSE_PKT3;
+ u32 PAUSE_PKT4;
+ u32 PAUSE_PKT5;
+ u32 reserve[2];
+ struct pch_gbe_regs_mac_adr mac_adr[16];
+ u32 ADDR_MASK;
+ u32 MIIM;
+ u32 reserve2;
+ u32 RGMII_ST;
+ u32 RGMII_CTRL;
+ u32 reserve3[3];
+ u32 DMA_CTRL;
+ u32 reserve4[3];
+ u32 RX_DSC_BASE;
+ u32 RX_DSC_SIZE;
+ u32 RX_DSC_HW_P;
+ u32 RX_DSC_HW_P_HLD;
+ u32 RX_DSC_SW_P;
+ u32 reserve5[3];
+ u32 TX_DSC_BASE;
+ u32 TX_DSC_SIZE;
+ u32 TX_DSC_HW_P;
+ u32 TX_DSC_HW_P_HLD;
+ u32 TX_DSC_SW_P;
+ u32 reserve6[3];
+ u32 RX_DMA_ST;
+ u32 TX_DMA_ST;
+ u32 reserve7[2];
+ u32 WOL_ST;
+ u32 WOL_CTRL;
+ u32 WOL_ADDR_MASK;
+};
+
+/* Interrupt Status */
+/* Interrupt Status Hold */
+/* Interrupt Enable */
+#define PCH_GBE_INT_RX_DMA_CMPLT 0x00000001 /* Receive DMA Transfer Complete */
+#define PCH_GBE_INT_RX_VALID 0x00000002 /* MAC Normal Receive Complete */
+#define PCH_GBE_INT_RX_FRAME_ERR 0x00000004 /* Receive frame error */
+#define PCH_GBE_INT_RX_FIFO_ERR 0x00000008 /* Receive FIFO Overflow */
+#define PCH_GBE_INT_RX_DMA_ERR 0x00000010 /* Receive DMA Transfer Error */
+#define PCH_GBE_INT_RX_DSC_EMP 0x00000020 /* Receive Descriptor Empty */
+#define PCH_GBE_INT_TX_CMPLT 0x00000100 /* MAC Transmission Complete */
+#define PCH_GBE_INT_TX_DMA_CMPLT 0x00000200 /* DMA Transfer Complete */
+#define PCH_GBE_INT_TX_FIFO_ERR 0x00000400 /* Transmission FIFO underflow. */
+#define PCH_GBE_INT_TX_DMA_ERR 0x00000800 /* Transmission DMA Error */
+#define PCH_GBE_INT_PAUSE_CMPLT 0x00001000 /* Pause Transmission complete */
+#define PCH_GBE_INT_MIIM_CMPLT 0x00010000 /* MIIM I/F Read completion */
+#define PCH_GBE_INT_PHY_INT 0x00100000 /* Interruption from PHY */
+#define PCH_GBE_INT_WOL_DET 0x01000000 /* Wake On LAN Event detection. */
+#define PCH_GBE_INT_TCPIP_ERR 0x10000000 /* TCP/IP Accelerator Error */
+
+/* Mode */
+#define PCH_GBE_MODE_MII_ETHER 0x00000000 /* GIGA Ethernet Mode [MII] */
+#define PCH_GBE_MODE_GMII_ETHER 0x80000000 /* GIGA Ethernet Mode [GMII] */
+#define PCH_GBE_MODE_HALF_DUPLEX 0x00000000 /* Duplex Mode [half duplex] */
+#define PCH_GBE_MODE_FULL_DUPLEX 0x40000000 /* Duplex Mode [full duplex] */
+#define PCH_GBE_MODE_FR_BST 0x04000000 /* Frame bursting is done */
+
+/* Reset */
+#define PCH_GBE_ALL_RST 0x80000000 /* All reset */
+#define PCH_GBE_TX_RST 0x40000000 /* TX MAC, TX FIFO, TX DMA reset */
+#define PCH_GBE_RX_RST 0x04000000 /* RX MAC, RX FIFO, RX DMA reset */
+
+/* TCP/IP Accelerator Control */
+#define PCH_GBE_EX_LIST_EN 0x00000008 /* External List Enable */
+#define PCH_GBE_RX_TCPIPACC_OFF 0x00000004 /* RX TCP/IP ACC Disabled */
+#define PCH_GBE_TX_TCPIPACC_EN 0x00000002 /* TX TCP/IP ACC Enable */
+#define PCH_GBE_RX_TCPIPACC_EN 0x00000001 /* RX TCP/IP ACC Enable */
+
+/* MAC RX Enable */
+#define PCH_GBE_MRE_MAC_RX_EN 0x00000001 /* MAC Receive Enable */
+
+/* RX Flow Control */
+#define PCH_GBE_FL_CTRL_EN 0x80000000 /* Pause packet is enabled */
+
+/* Pause Packet Request */
+#define PCH_GBE_PS_PKT_RQ 0x80000000 /* Pause packet Request */
+
+/* RX Mode */
+#define PCH_GBE_ADD_FIL_EN 0x80000000 /* Address Filtering Enable */
+/* Multicast Filtering Enable */
+#define PCH_GBE_MLT_FIL_EN 0x40000000
+/* Receive Almost Empty Threshold */
+#define PCH_GBE_RH_ALM_EMP_4 0x00000000 /* 4 words */
+#define PCH_GBE_RH_ALM_EMP_8 0x00004000 /* 8 words */
+#define PCH_GBE_RH_ALM_EMP_16 0x00008000 /* 16 words */
+#define PCH_GBE_RH_ALM_EMP_32 0x0000C000 /* 32 words */
+/* Receive Almost Full Threshold */
+#define PCH_GBE_RH_ALM_FULL_4 0x00000000 /* 4 words */
+#define PCH_GBE_RH_ALM_FULL_8 0x00001000 /* 8 words */
+#define PCH_GBE_RH_ALM_FULL_16 0x00002000 /* 16 words */
+#define PCH_GBE_RH_ALM_FULL_32 0x00003000 /* 32 words */
+/* RX FIFO Read Triger Threshold */
+#define PCH_GBE_RH_RD_TRG_4 0x00000000 /* 4 words */
+#define PCH_GBE_RH_RD_TRG_8 0x00000200 /* 8 words */
+#define PCH_GBE_RH_RD_TRG_16 0x00000400 /* 16 words */
+#define PCH_GBE_RH_RD_TRG_32 0x00000600 /* 32 words */
+#define PCH_GBE_RH_RD_TRG_64 0x00000800 /* 64 words */
+#define PCH_GBE_RH_RD_TRG_128 0x00000A00 /* 128 words */
+#define PCH_GBE_RH_RD_TRG_256 0x00000C00 /* 256 words */
+#define PCH_GBE_RH_RD_TRG_512 0x00000E00 /* 512 words */
+
+/* Receive Descriptor bit definitions */
+#define PCH_GBE_RXD_ACC_STAT_BCAST 0x00000400
+#define PCH_GBE_RXD_ACC_STAT_MCAST 0x00000200
+#define PCH_GBE_RXD_ACC_STAT_UCAST 0x00000100
+#define PCH_GBE_RXD_ACC_STAT_TCPIPOK 0x000000C0
+#define PCH_GBE_RXD_ACC_STAT_IPOK 0x00000080
+#define PCH_GBE_RXD_ACC_STAT_TCPOK 0x00000040
+#define PCH_GBE_RXD_ACC_STAT_IP6ERR 0x00000020
+#define PCH_GBE_RXD_ACC_STAT_OFLIST 0x00000010
+#define PCH_GBE_RXD_ACC_STAT_TYPEIP 0x00000008
+#define PCH_GBE_RXD_ACC_STAT_MACL 0x00000004
+#define PCH_GBE_RXD_ACC_STAT_PPPOE 0x00000002
+#define PCH_GBE_RXD_ACC_STAT_VTAGT 0x00000001
+#define PCH_GBE_RXD_GMAC_STAT_PAUSE 0x0200
+#define PCH_GBE_RXD_GMAC_STAT_MARBR 0x0100
+#define PCH_GBE_RXD_GMAC_STAT_MARMLT 0x0080
+#define PCH_GBE_RXD_GMAC_STAT_MARIND 0x0040
+#define PCH_GBE_RXD_GMAC_STAT_MARNOTMT 0x0020
+#define PCH_GBE_RXD_GMAC_STAT_TLONG 0x0010
+#define PCH_GBE_RXD_GMAC_STAT_TSHRT 0x0008
+#define PCH_GBE_RXD_GMAC_STAT_NOTOCTAL 0x0004
+#define PCH_GBE_RXD_GMAC_STAT_NBLERR 0x0002
+#define PCH_GBE_RXD_GMAC_STAT_CRCERR 0x0001
+
+/* Transmit Descriptor bit definitions */
+#define PCH_GBE_TXD_CTRL_TCPIP_ACC_OFF 0x0008
+#define PCH_GBE_TXD_CTRL_ITAG 0x0004
+#define PCH_GBE_TXD_CTRL_ICRC 0x0002
+#define PCH_GBE_TXD_CTRL_APAD 0x0001
+#define PCH_GBE_TXD_WORDS_SHIFT 2
+#define PCH_GBE_TXD_GMAC_STAT_CMPLT 0x2000
+#define PCH_GBE_TXD_GMAC_STAT_ABT 0x1000
+#define PCH_GBE_TXD_GMAC_STAT_EXCOL 0x0800
+#define PCH_GBE_TXD_GMAC_STAT_SNGCOL 0x0400
+#define PCH_GBE_TXD_GMAC_STAT_MLTCOL 0x0200
+#define PCH_GBE_TXD_GMAC_STAT_CRSER 0x0100
+#define PCH_GBE_TXD_GMAC_STAT_TLNG 0x0080
+#define PCH_GBE_TXD_GMAC_STAT_TSHRT 0x0040
+#define PCH_GBE_TXD_GMAC_STAT_LTCOL 0x0020
+#define PCH_GBE_TXD_GMAC_STAT_TFUNDFLW 0x0010
+#define PCH_GBE_TXD_GMAC_STAT_RTYCNT_MASK 0x000F
+
+/* TX Mode */
+#define PCH_GBE_TM_NO_RTRY 0x80000000 /* No Retransmission */
+#define PCH_GBE_TM_LONG_PKT 0x40000000 /* Long Packt TX Enable */
+#define PCH_GBE_TM_ST_AND_FD 0x20000000 /* Stare and Forward */
+#define PCH_GBE_TM_SHORT_PKT 0x10000000 /* Short Packet TX Enable */
+#define PCH_GBE_TM_LTCOL_RETX 0x08000000 /* Retransmission at Late Collision */
+/* Frame Start Threshold */
+#define PCH_GBE_TM_TH_TX_STRT_4 0x00000000 /* 4 words */
+#define PCH_GBE_TM_TH_TX_STRT_8 0x00004000 /* 8 words */
+#define PCH_GBE_TM_TH_TX_STRT_16 0x00008000 /* 16 words */
+#define PCH_GBE_TM_TH_TX_STRT_32 0x0000C000 /* 32 words */
+/* Transmit Almost Empty Threshold */
+#define PCH_GBE_TM_TH_ALM_EMP_4 0x00000000 /* 4 words */
+#define PCH_GBE_TM_TH_ALM_EMP_8 0x00000800 /* 8 words */
+#define PCH_GBE_TM_TH_ALM_EMP_16 0x00001000 /* 16 words */
+#define PCH_GBE_TM_TH_ALM_EMP_32 0x00001800 /* 32 words */
+#define PCH_GBE_TM_TH_ALM_EMP_64 0x00002000 /* 64 words */
+#define PCH_GBE_TM_TH_ALM_EMP_128 0x00002800 /* 128 words */
+#define PCH_GBE_TM_TH_ALM_EMP_256 0x00003000 /* 256 words */
+#define PCH_GBE_TM_TH_ALM_EMP_512 0x00003800 /* 512 words */
+/* Transmit Almost Full Threshold */
+#define PCH_GBE_TM_TH_ALM_FULL_4 0x00000000 /* 4 words */
+#define PCH_GBE_TM_TH_ALM_FULL_8 0x00000200 /* 8 words */
+#define PCH_GBE_TM_TH_ALM_FULL_16 0x00000400 /* 16 words */
+#define PCH_GBE_TM_TH_ALM_FULL_32 0x00000600 /* 32 words */
+
+/* RX FIFO Status */
+#define PCH_GBE_RF_ALM_FULL 0x80000000 /* RX FIFO is almost full. */
+#define PCH_GBE_RF_ALM_EMP 0x40000000 /* RX FIFO is almost empty. */
+#define PCH_GBE_RF_RD_TRG 0x20000000 /* Become more than RH_RD_TRG. */
+#define PCH_GBE_RF_STRWD 0x1FFE0000 /* The word count of RX FIFO. */
+#define PCH_GBE_RF_RCVING 0x00010000 /* Stored in RX FIFO. */
+
+/* MAC Address Mask */
+#define PCH_GBE_BUSY 0x80000000
+
+/* MIIM */
+#define PCH_GBE_MIIM_OPER_WRITE 0x04000000
+#define PCH_GBE_MIIM_OPER_READ 0x00000000
+#define PCH_GBE_MIIM_OPER_READY 0x04000000
+#define PCH_GBE_MIIM_PHY_ADDR_SHIFT 21
+#define PCH_GBE_MIIM_REG_ADDR_SHIFT 16
+
+/* RGMII Status */
+#define PCH_GBE_LINK_UP 0x80000008
+#define PCH_GBE_RXC_SPEED_MSK 0x00000006
+#define PCH_GBE_RXC_SPEED_2_5M 0x00000000 /* 2.5MHz */
+#define PCH_GBE_RXC_SPEED_25M 0x00000002 /* 25MHz */
+#define PCH_GBE_RXC_SPEED_125M 0x00000004 /* 100MHz */
+#define PCH_GBE_DUPLEX_FULL 0x00000001
+
+/* RGMII Control */
+#define PCH_GBE_CRS_SEL 0x00000010
+#define PCH_GBE_RGMII_RATE_125M 0x00000000
+#define PCH_GBE_RGMII_RATE_25M 0x00000008
+#define PCH_GBE_RGMII_RATE_2_5M 0x0000000C
+#define PCH_GBE_RGMII_MODE_GMII 0x00000000
+#define PCH_GBE_RGMII_MODE_RGMII 0x00000002
+#define PCH_GBE_CHIP_TYPE_EXTERNAL 0x00000000
+#define PCH_GBE_CHIP_TYPE_INTERNAL 0x00000001
+
+/* DMA Control */
+#define PCH_GBE_RX_DMA_EN 0x00000002 /* Enables Receive DMA */
+#define PCH_GBE_TX_DMA_EN 0x00000001 /* Enables Transmission DMA */
+
+/* Wake On LAN Status */
+#define PCH_GBE_WLS_BR 0x00000008 /* Broadcas Address */
+#define PCH_GBE_WLS_MLT 0x00000004 /* Multicast Address */
+
+/* The Frame registered in Address Recognizer */
+#define PCH_GBE_WLS_IND 0x00000002
+#define PCH_GBE_WLS_MP 0x00000001 /* Magic packet Address */
+
+/* Wake On LAN Control */
+#define PCH_GBE_WLC_WOL_MODE 0x00010000
+#define PCH_GBE_WLC_IGN_TLONG 0x00000100
+#define PCH_GBE_WLC_IGN_TSHRT 0x00000080
+#define PCH_GBE_WLC_IGN_OCTER 0x00000040
+#define PCH_GBE_WLC_IGN_NBLER 0x00000020
+#define PCH_GBE_WLC_IGN_CRCER 0x00000010
+#define PCH_GBE_WLC_BR 0x00000008
+#define PCH_GBE_WLC_MLT 0x00000004
+#define PCH_GBE_WLC_IND 0x00000002
+#define PCH_GBE_WLC_MP 0x00000001
+
+/* Wake On LAN Address Mask */
+#define PCH_GBE_WLA_BUSY 0x80000000
+
+
+
+/* TX/RX descriptor defines */
+#define PCH_GBE_MAX_TXD 4096
+#define PCH_GBE_DEFAULT_TXD 256
+#define PCH_GBE_MIN_TXD 8
+#define PCH_GBE_MAX_RXD 4096
+#define PCH_GBE_DEFAULT_RXD 256
+#define PCH_GBE_MIN_RXD 8
+
+/* Number of Transmit and Receive Descriptors must be a multiple of 8 */
+#define PCH_GBE_TX_DESC_MULTIPLE 8
+#define PCH_GBE_RX_DESC_MULTIPLE 8
+
+/* Read/Write operation is done through MII Management IF */
+#define PCH_GBE_HAL_MIIM_READ ((u32)0x00000000)
+#define PCH_GBE_HAL_MIIM_WRITE ((u32)0x04000000)
+
+/* flow control values */
+#define PCH_GBE_FC_NONE 0
+#define PCH_GBE_FC_RX_PAUSE 1
+#define PCH_GBE_FC_TX_PAUSE 2
+#define PCH_GBE_FC_FULL 3
+#define PCH_GBE_FC_DEFAULT PCH_GBE_FC_FULL
+
+
+struct pch_gbe_hw;
+/**
+ * struct pch_gbe_functions - HAL APi function pointer
+ * @get_bus_info: for pch_gbe_hal_get_bus_info
+ * @init_hw: for pch_gbe_hal_init_hw
+ * @read_phy_reg: for pch_gbe_hal_read_phy_reg
+ * @write_phy_reg: for pch_gbe_hal_write_phy_reg
+ * @reset_phy: for pch_gbe_hal_phy_hw_reset
+ * @sw_reset_phy: for pch_gbe_hal_phy_sw_reset
+ * @power_up_phy: for pch_gbe_hal_power_up_phy
+ * @power_down_phy: for pch_gbe_hal_power_down_phy
+ * @read_mac_addr: for pch_gbe_hal_read_mac_addr
+ */
+struct pch_gbe_functions {
+ void (*get_bus_info) (struct pch_gbe_hw *);
+ s32 (*init_hw) (struct pch_gbe_hw *);
+ s32 (*read_phy_reg) (struct pch_gbe_hw *, u32, u16 *);
+ s32 (*write_phy_reg) (struct pch_gbe_hw *, u32, u16);
+ void (*reset_phy) (struct pch_gbe_hw *);
+ void (*sw_reset_phy) (struct pch_gbe_hw *);
+ void (*power_up_phy) (struct pch_gbe_hw *hw);
+ void (*power_down_phy) (struct pch_gbe_hw *hw);
+ s32 (*read_mac_addr) (struct pch_gbe_hw *);
+};
+
+/**
+ * struct pch_gbe_mac_info - MAC infomation
+ * @addr[6]: Store the MAC address
+ * @fc: Mode of flow control
+ * @fc_autoneg: Auto negotiation enable for flow control setting
+ * @tx_fc_enable: Enable flag of Transmit flow control
+ * @max_frame_size: Max transmit frame size
+ * @min_frame_size: Min transmit frame size
+ * @autoneg: Auto negotiation enable
+ * @link_speed: Link speed
+ * @link_duplex: Link duplex
+ */
+struct pch_gbe_mac_info {
+ u8 addr[6];
+ u8 fc;
+ u8 fc_autoneg;
+ u8 tx_fc_enable;
+ u32 max_frame_size;
+ u32 min_frame_size;
+ u8 autoneg;
+ u16 link_speed;
+ u16 link_duplex;
+};
+
+/**
+ * struct pch_gbe_phy_info - PHY infomation
+ * @addr: PHY address
+ * @id: PHY's identifier
+ * @revision: PHY's revision
+ * @reset_delay_us: HW reset delay time[us]
+ * @autoneg_advertised: Autoneg advertised
+ */
+struct pch_gbe_phy_info {
+ u32 addr;
+ u32 id;
+ u32 revision;
+ u32 reset_delay_us;
+ u16 autoneg_advertised;
+};
+
+/*!
+ * @ingroup Gigabit Ether driver Layer
+ * @struct pch_gbe_bus_info
+ * @brief Bus infomation
+ */
+struct pch_gbe_bus_info {
+ u8 type;
+ u8 speed;
+ u8 width;
+};
+
+/*!
+ * @ingroup Gigabit Ether driver Layer
+ * @struct pch_gbe_hw
+ * @brief Hardware infomation
+ */
+struct pch_gbe_hw {
+ void *back;
+
+ struct pch_gbe_regs __iomem *reg;
+ spinlock_t miim_lock;
+
+ const struct pch_gbe_functions *func;
+ struct pch_gbe_mac_info mac;
+ struct pch_gbe_phy_info phy;
+ struct pch_gbe_bus_info bus;
+};
+
+/**
+ * struct pch_gbe_rx_desc - Receive Descriptor
+ * @buffer_addr: RX Frame Buffer Address
+ * @tcp_ip_status: TCP/IP Accelerator Status
+ * @rx_words_eob: RX word count and Byte position
+ * @gbec_status: GMAC Status
+ * @dma_status: DMA Status
+ * @reserved1: Reserved
+ * @reserved2: Reserved
+ */
+struct pch_gbe_rx_desc {
+ u32 buffer_addr;
+ u32 tcp_ip_status;
+ u16 rx_words_eob;
+ u16 gbec_status;
+ u8 dma_status;
+ u8 reserved1;
+ u16 reserved2;
+};
+
+/**
+ * struct pch_gbe_tx_desc - Transmit Descriptor
+ * @buffer_addr: TX Frame Buffer Address
+ * @length: Data buffer length
+ * @reserved1: Reserved
+ * @tx_words_eob: TX word count and Byte position
+ * @tx_frame_ctrl: TX Frame Control
+ * @dma_status: DMA Status
+ * @reserved2: Reserved
+ * @gbec_status: GMAC Status
+ */
+struct pch_gbe_tx_desc {
+ u32 buffer_addr;
+ u16 length;
+ u16 reserved1;
+ u16 tx_words_eob;
+ u16 tx_frame_ctrl;
+ u8 dma_status;
+ u8 reserved2;
+ u16 gbec_status;
+};
+
+
+/**
+ * struct pch_gbe_buffer - Buffer infomation
+ * @skb: pointer to a socket buffer
+ * @dma: DMA address
+ * @time_stamp: time stamp
+ * @length: data size
+ */
+struct pch_gbe_buffer {
+ struct sk_buff *skb;
+ dma_addr_t dma;
+ unsigned long time_stamp;
+ u16 length;
+ bool mapped;
+};
+
+/**
+ * struct pch_gbe_tx_ring - tx ring infomation
+ * @tx_lock: spinlock structs
+ * @desc: pointer to the descriptor ring memory
+ * @dma: physical address of the descriptor ring
+ * @size: length of descriptor ring in bytes
+ * @count: number of descriptors in the ring
+ * @next_to_use: next descriptor to associate a buffer with
+ * @next_to_clean: next descriptor to check for DD status bit
+ * @buffer_info: array of buffer information structs
+ */
+struct pch_gbe_tx_ring {
+ spinlock_t tx_lock;
+ struct pch_gbe_tx_desc *desc;
+ dma_addr_t dma;
+ unsigned int size;
+ unsigned int count;
+ unsigned int next_to_use;
+ unsigned int next_to_clean;
+ struct pch_gbe_buffer *buffer_info;
+};
+
+/**
+ * struct pch_gbe_rx_ring - rx ring infomation
+ * @desc: pointer to the descriptor ring memory
+ * @dma: physical address of the descriptor ring
+ * @size: length of descriptor ring in bytes
+ * @count: number of descriptors in the ring
+ * @next_to_use: next descriptor to associate a buffer with
+ * @next_to_clean: next descriptor to check for DD status bit
+ * @buffer_info: array of buffer information structs
+ */
+struct pch_gbe_rx_ring {
+ struct pch_gbe_rx_desc *desc;
+ dma_addr_t dma;
+ unsigned int size;
+ unsigned int count;
+ unsigned int next_to_use;
+ unsigned int next_to_clean;
+ struct pch_gbe_buffer *buffer_info;
+};
+
+/**
+ * struct pch_gbe_hw_stats - Statistics counters collected by the MAC
+ * @rx_packets: total packets received
+ * @tx_packets: total packets transmitted
+ * @rx_bytes: total bytes received
+ * @tx_bytes: total bytes transmitted
+ * @rx_errors: bad packets received
+ * @tx_errors: packet transmit problems
+ * @rx_dropped: no space in Linux buffers
+ * @tx_dropped: no space available in Linux
+ * @multicast: multicast packets received
+ * @collisions: collisions
+ * @rx_crc_errors: received packet with crc error
+ * @rx_frame_errors: received frame alignment error
+ * @rx_alloc_buff_failed: allocate failure of a receive buffer
+ * @tx_length_errors: transmit length error
+ * @tx_aborted_errors: transmit aborted error
+ * @tx_carrier_errors: transmit carrier error
+ * @tx_timeout_count: Number of transmit timeout
+ * @tx_restart_count: Number of transmit restert
+ * @intr_rx_dsc_empty_count: Interrupt count of receive descriptor empty
+ * @intr_rx_frame_err_count: Interrupt count of receive frame error
+ * @intr_rx_fifo_err_count: Interrupt count of receive FIFO error
+ * @intr_rx_dma_err_count: Interrupt count of receive DMA error
+ * @intr_tx_fifo_err_count: Interrupt count of transmit FIFO error
+ * @intr_tx_dma_err_count: Interrupt count of transmit DMA error
+ * @intr_tcpip_err_count: Interrupt count of TCP/IP Accelerator
+ */
+struct pch_gbe_hw_stats {
+ u32 rx_packets;
+ u32 tx_packets;
+ u32 rx_bytes;
+ u32 tx_bytes;
+ u32 rx_errors;
+ u32 tx_errors;
+ u32 rx_dropped;
+ u32 tx_dropped;
+ u32 multicast;
+ u32 collisions;
+ u32 rx_crc_errors;
+ u32 rx_frame_errors;
+ u32 rx_alloc_buff_failed;
+ u32 tx_length_errors;
+ u32 tx_aborted_errors;
+ u32 tx_carrier_errors;
+ u32 tx_timeout_count;
+ u32 tx_restart_count;
+ u32 intr_rx_dsc_empty_count;
+ u32 intr_rx_frame_err_count;
+ u32 intr_rx_fifo_err_count;
+ u32 intr_rx_dma_err_count;
+ u32 intr_tx_fifo_err_count;
+ u32 intr_tx_dma_err_count;
+ u32 intr_tcpip_err_count;
+};
+
+/**
+ * struct pch_gbe_adapter - board specific private data structure
+ * @stats_lock: Spinlock structure for status
+ * @tx_queue_lock: Spinlock structure for transmit
+ * @ethtool_lock: Spinlock structure for ethtool
+ * @irq_sem: Semaphore for interrupt
+ * @netdev: Pointer of network device structure
+ * @pdev: Pointer of pci device structure
+ * @polling_netdev: Pointer of polling network device structure
+ * @napi: NAPI structure
+ * @hw: Pointer of hardware structure
+ * @stats: Hardware status
+ * @reset_task: Reset task
+ * @mii: MII information structure
+ * @watchdog_timer: Watchdog timer list
+ * @wake_up_evt: Wake up event
+ * @config_space: Configuration space
+ * @msg_enable: Driver message level
+ * @led_status: LED status
+ * @tx_ring: Pointer of Tx descriptor ring structure
+ * @rx_ring: Pointer of Rx descriptor ring structure
+ * @rx_buffer_len: Receive buffer length
+ * @tx_queue_len: Transmit queue length
+ * @rx_csum: Receive TCP/IP checksum enable/disable
+ * @tx_csum: Transmit TCP/IP checksum enable/disable
+ * @have_msi: PCI MSI mode flag
+ */
+
+struct pch_gbe_adapter {
+ spinlock_t stats_lock;
+ spinlock_t tx_queue_lock;
+ spinlock_t ethtool_lock;
+ atomic_t irq_sem;
+ struct net_device *netdev;
+ struct pci_dev *pdev;
+ struct net_device *polling_netdev;
+ struct napi_struct napi;
+ struct pch_gbe_hw hw;
+ struct pch_gbe_hw_stats stats;
+ struct work_struct reset_task;
+ struct mii_if_info mii;
+ struct timer_list watchdog_timer;
+ u32 wake_up_evt;
+ u32 *config_space;
+ unsigned long led_status;
+ struct pch_gbe_tx_ring *tx_ring;
+ struct pch_gbe_rx_ring *rx_ring;
+ unsigned long rx_buffer_len;
+ unsigned long tx_queue_len;
+ bool rx_csum;
+ bool tx_csum;
+ bool have_msi;
+};
+
+extern const char pch_driver_version[];
+
+/* pch_gbe_main.c */
+extern int pch_gbe_up(struct pch_gbe_adapter *adapter);
+extern void pch_gbe_down(struct pch_gbe_adapter *adapter);
+extern void pch_gbe_reinit_locked(struct pch_gbe_adapter *adapter);
+extern void pch_gbe_reset(struct pch_gbe_adapter *adapter);
+extern int pch_gbe_setup_tx_resources(struct pch_gbe_adapter *adapter,
+ struct pch_gbe_tx_ring *txdr);
+extern int pch_gbe_setup_rx_resources(struct pch_gbe_adapter *adapter,
+ struct pch_gbe_rx_ring *rxdr);
+extern void pch_gbe_free_tx_resources(struct pch_gbe_adapter *adapter,
+ struct pch_gbe_tx_ring *tx_ring);
+extern void pch_gbe_free_rx_resources(struct pch_gbe_adapter *adapter,
+ struct pch_gbe_rx_ring *rx_ring);
+extern void pch_gbe_update_stats(struct pch_gbe_adapter *adapter);
+extern int pch_gbe_mdio_read(struct net_device *netdev, int addr, int reg);
+extern void pch_gbe_mdio_write(struct net_device *netdev, int addr, int reg,
+ int data);
+/* pch_gbe_param.c */
+extern void pch_gbe_check_options(struct pch_gbe_adapter *adapter);
+
+/* pch_gbe_ethtool.c */
+extern void pch_gbe_set_ethtool_ops(struct net_device *netdev);
+
+/* pch_gbe_mac.c */
+extern s32 pch_gbe_mac_force_mac_fc(struct pch_gbe_hw *hw);
+extern s32 pch_gbe_mac_read_mac_addr(struct pch_gbe_hw *hw);
+extern u16 pch_gbe_mac_ctrl_miim(struct pch_gbe_hw *hw,
+ u32 addr, u32 dir, u32 reg, u16 data);
+#endif /* _PCH_GBE_H_ */
diff --git a/drivers/net/pch_gbe/pch_gbe_api.c b/drivers/net/pch_gbe/pch_gbe_api.c
new file mode 100644
index 0000000..db53d2a
--- /dev/null
+++ b/drivers/net/pch_gbe/pch_gbe_api.c
@@ -0,0 +1,245 @@
+/*
+ * Copyright (C) 1999 - 2010 Intel Corporation.
+ * Copyright (C) 2010 OKI SEMICONDUCTOR Co., LTD.
+ *
+ * This code was derived from the Intel e1000e Linux driver.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * 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 "pch_gbe.h"
+#include "pch_gbe_phy.h"
+
+/* bus type values */
+#define pch_gbe_bus_type_unknown 0
+#define pch_gbe_bus_type_pci 1
+#define pch_gbe_bus_type_pcix 2
+#define pch_gbe_bus_type_pci_express 3
+#define pch_gbe_bus_type_reserved 4
+
+/* bus speed values */
+#define pch_gbe_bus_speed_unknown 0
+#define pch_gbe_bus_speed_33 1
+#define pch_gbe_bus_speed_66 2
+#define pch_gbe_bus_speed_100 3
+#define pch_gbe_bus_speed_120 4
+#define pch_gbe_bus_speed_133 5
+#define pch_gbe_bus_speed_2500 6
+#define pch_gbe_bus_speed_reserved 7
+
+/* bus width values */
+#define pch_gbe_bus_width_unknown 0
+#define pch_gbe_bus_width_pcie_x1 1
+#define pch_gbe_bus_width_pcie_x2 2
+#define pch_gbe_bus_width_pcie_x4 4
+#define pch_gbe_bus_width_32 5
+#define pch_gbe_bus_width_64 6
+#define pch_gbe_bus_width_reserved 7
+
+/**
+ * pch_gbe_plat_get_bus_info - Obtain bus information for adapter
+ * @hw: Pointer to the HW structure
+ */
+static void pch_gbe_plat_get_bus_info(struct pch_gbe_hw *hw)
+{
+ hw->bus.type = pch_gbe_bus_type_pci_express;
+ hw->bus.speed = pch_gbe_bus_speed_2500;
+ hw->bus.width = pch_gbe_bus_width_pcie_x1;
+}
+
+/**
+ * pch_gbe_plat_init_hw - Initialize hardware
+ * @hw: Pointer to the HW structure
+ * Returns
+ * 0: Successfully
+ * Negative value: Failed-EBUSY
+ */
+static s32 pch_gbe_plat_init_hw(struct pch_gbe_hw *hw)
+{
+ s32 ret_val;
+
+ ret_val = pch_gbe_phy_get_id(hw);
+ if (ret_val) {
+ pr_err("pch_gbe_phy_get_id error\n");
+ return ret_val;
+ }
+ pch_gbe_phy_init_setting(hw);
+ /* Setup Mac interface option RGMII */
+#ifdef PCH_GBE_MAC_IFOP_RGMII
+ pch_gbe_phy_set_rgmii(hw);
+#endif
+ return ret_val;
+}
+
+static const struct pch_gbe_functions pch_gbe_ops = {
+ .get_bus_info = pch_gbe_plat_get_bus_info,
+ .init_hw = pch_gbe_plat_init_hw,
+ .read_phy_reg = pch_gbe_phy_read_reg_miic,
+ .write_phy_reg = pch_gbe_phy_write_reg_miic,
+ .reset_phy = pch_gbe_phy_hw_reset,
+ .sw_reset_phy = pch_gbe_phy_sw_reset,
+ .power_up_phy = pch_gbe_phy_power_up,
+ .power_down_phy = pch_gbe_phy_power_down,
+ .read_mac_addr = pch_gbe_mac_read_mac_addr
+};
+
+/**
+ * pch_gbe_plat_init_function_pointers - Init func ptrs
+ * @hw: Pointer to the HW structure
+ */
+void pch_gbe_plat_init_function_pointers(struct pch_gbe_hw *hw)
+{
+ /* Set PHY parameter */
+ hw->phy.reset_delay_us = PCH_GBE_PHY_RESET_DELAY_US;
+ /* Set function pointers */
+ hw->func = &pch_gbe_ops;
+}
+
+/**
+ * pch_gbe_hal_setup_init_funcs - Initializes function pointers
+ * @hw: Pointer to the HW structure
+ * Returns
+ * 0: Successfully
+ * ENOSYS: Function is not registered
+ */
+inline s32 pch_gbe_hal_setup_init_funcs(struct pch_gbe_hw *hw)
+{
+ if (!hw->reg) {
+ pr_err("ERROR: Registers not mapped\n");
+ return -ENOSYS;
+ }
+ pch_gbe_plat_init_function_pointers(hw);
+ return 0;
+}
+
+/**
+ * pch_gbe_hal_get_bus_info - Obtain bus information for adapter
+ * @hw: Pointer to the HW structure
+ */
+inline void pch_gbe_hal_get_bus_info(struct pch_gbe_hw *hw)
+{
+ if (!hw->func->get_bus_info)
+ pr_err("ERROR: configuration\n");
+ else
+ hw->func->get_bus_info(hw);
+}
+
+/**
+ * pch_gbe_hal_init_hw - Initialize hardware
+ * @hw: Pointer to the HW structure
+ * Returns
+ * 0: Successfully
+ * ENOSYS: Function is not registered
+ */
+inline s32 pch_gbe_hal_init_hw(struct pch_gbe_hw *hw)
+{
+ if (!hw->func->init_hw) {
+ pr_err("ERROR: configuration\n");
+ return -ENOSYS;
+ }
+ return hw->func->init_hw(hw);
+}
+
+/**
+ * pch_gbe_hal_read_phy_reg - Reads PHY register
+ * @hw: Pointer to the HW structure
+ * @offset: The register to read
+ * @data: The buffer to store the 16-bit read.
+ * Returns
+ * 0: Successfully
+ * Negative value: Failed
+ */
+inline s32 pch_gbe_hal_read_phy_reg(struct pch_gbe_hw *hw, u32 offset,
+ u16 *data)
+{
+ if (!hw->func->read_phy_reg)
+ return 0;
+ return hw->func->read_phy_reg(hw, offset, data);
+}
+
+/**
+ * pch_gbe_hal_write_phy_reg - Writes PHY register
+ * @hw: Pointer to the HW structure
+ * @offset: The register to read
+ * @data: The value to write.
+ * Returns
+ * 0: Successfully
+ * Negative value: Failed
+ */
+inline s32 pch_gbe_hal_write_phy_reg(struct pch_gbe_hw *hw, u32 offset,
+ u16 data)
+{
+ if (!hw->func->write_phy_reg)
+ return 0;
+ return hw->func->write_phy_reg(hw, offset, data);
+}
+
+/**
+ * pch_gbe_hal_phy_hw_reset - Hard PHY reset
+ * @hw: Pointer to the HW structure
+ */
+inline void pch_gbe_hal_phy_hw_reset(struct pch_gbe_hw *hw)
+{
+ if (!hw->func->reset_phy)
+ pr_err("ERROR: configuration\n");
+ else
+ hw->func->reset_phy(hw);
+}
+
+/**
+ * pch_gbe_hal_phy_sw_reset - Soft PHY reset
+ * @hw: Pointer to the HW structure
+ */
+inline void pch_gbe_hal_phy_sw_reset(struct pch_gbe_hw *hw)
+{
+ if (!hw->func->sw_reset_phy)
+ pr_err("ERROR: configuration\n");
+ else
+ hw->func->sw_reset_phy(hw);
+}
+
+/**
+ * pch_gbe_hal_read_mac_addr - Reads MAC address
+ * @hw: Pointer to the HW structure
+ * Returns
+ * 0: Successfully
+ * ENOSYS: Function is not registered
+ */
+inline s32 pch_gbe_hal_read_mac_addr(struct pch_gbe_hw *hw)
+{
+ if (!hw->func->read_mac_addr) {
+ pr_err("ERROR: configuration\n");
+ return -ENOSYS;
+ }
+ return hw->func->read_mac_addr(hw);
+}
+
+/**
+ * pch_gbe_hal_power_up_phy - Power up PHY
+ * @hw: Pointer to the HW structure
+ */
+inline void pch_gbe_hal_power_up_phy(struct pch_gbe_hw *hw)
+{
+ if (hw->func->power_up_phy)
+ hw->func->power_up_phy(hw);
+}
+
+/**
+ * pch_gbe_hal_power_down_phy - Power down PHY
+ * @hw: Pointer to the HW structure
+ */
+inline void pch_gbe_hal_power_down_phy(struct pch_gbe_hw *hw)
+{
+ if (hw->func->power_down_phy)
+ hw->func->power_down_phy(hw);
+}
diff --git a/drivers/net/pch_gbe/pch_gbe_api.h b/drivers/net/pch_gbe/pch_gbe_api.h
new file mode 100644
index 0000000..94aaac5
--- /dev/null
+++ b/drivers/net/pch_gbe/pch_gbe_api.h
@@ -0,0 +1,36 @@
+/*
+ * Copyright (C) 1999 - 2010 Intel Corporation.
+ * Copyright (C) 2010 OKI SEMICONDUCTOR Co., LTD.
+ *
+ * This code was derived from the Intel e1000e Linux driver.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * 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 _PCH_GBE_API_H_
+#define _PCH_GBE_API_H_
+
+#include "pch_gbe_phy.h"
+
+s32 pch_gbe_hal_setup_init_funcs(struct pch_gbe_hw *hw);
+void pch_gbe_hal_get_bus_info(struct pch_gbe_hw *hw);
+s32 pch_gbe_hal_init_hw(struct pch_gbe_hw *hw);
+s32 pch_gbe_hal_read_phy_reg(struct pch_gbe_hw *hw, u32 offset, u16 *data);
+s32 pch_gbe_hal_write_phy_reg(struct pch_gbe_hw *hw, u32 offset, u16 data);
+void pch_gbe_hal_phy_hw_reset(struct pch_gbe_hw *hw);
+void pch_gbe_hal_phy_sw_reset(struct pch_gbe_hw *hw);
+s32 pch_gbe_hal_read_mac_addr(struct pch_gbe_hw *hw);
+void pch_gbe_hal_power_up_phy(struct pch_gbe_hw *hw);
+void pch_gbe_hal_power_down_phy(struct pch_gbe_hw *hw);
+
+#endif
diff --git a/drivers/net/pch_gbe/pch_gbe_ethtool.c b/drivers/net/pch_gbe/pch_gbe_ethtool.c
new file mode 100644
index 0000000..e06c6ae
--- /dev/null
+++ b/drivers/net/pch_gbe/pch_gbe_ethtool.c
@@ -0,0 +1,584 @@
+/*
+ * Copyright (C) 1999 - 2010 Intel Corporation.
+ * Copyright (C) 2010 OKI SEMICONDUCTOR Co., LTD.
+ *
+ * This code was derived from the Intel e1000e Linux driver.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * 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 "pch_gbe.h"
+#include "pch_gbe_api.h"
+
+/**
+ * pch_gbe_stats - Stats item infomation
+ */
+struct pch_gbe_stats {
+ char string[ETH_GSTRING_LEN];
+ size_t size;
+ size_t offset;
+};
+
+#define PCH_GBE_STAT(m) \
+{ \
+ .string = #m, \
+ .size = FIELD_SIZEOF(struct pch_gbe_hw_stats, m), \
+ .offset = offsetof(struct pch_gbe_hw_stats, m), \
+}
+
+/**
+ * pch_gbe_gstrings_stats - ethtool information status name list
+ */
+static const struct pch_gbe_stats pch_gbe_gstrings_stats[] = {
+ PCH_GBE_STAT(rx_packets),
+ PCH_GBE_STAT(tx_packets),
+ PCH_GBE_STAT(rx_bytes),
+ PCH_GBE_STAT(tx_bytes),
+ PCH_GBE_STAT(rx_errors),
+ PCH_GBE_STAT(tx_errors),
+ PCH_GBE_STAT(rx_dropped),
+ PCH_GBE_STAT(tx_dropped),
+ PCH_GBE_STAT(multicast),
+ PCH_GBE_STAT(collisions),
+ PCH_GBE_STAT(rx_crc_errors),
+ PCH_GBE_STAT(rx_frame_errors),
+ PCH_GBE_STAT(rx_alloc_buff_failed),
+ PCH_GBE_STAT(tx_length_errors),
+ PCH_GBE_STAT(tx_aborted_errors),
+ PCH_GBE_STAT(tx_carrier_errors),
+ PCH_GBE_STAT(tx_timeout_count),
+ PCH_GBE_STAT(tx_restart_count),
+ PCH_GBE_STAT(intr_rx_dsc_empty_count),
+ PCH_GBE_STAT(intr_rx_frame_err_count),
+ PCH_GBE_STAT(intr_rx_fifo_err_count),
+ PCH_GBE_STAT(intr_rx_dma_err_count),
+ PCH_GBE_STAT(intr_tx_fifo_err_count),
+ PCH_GBE_STAT(intr_tx_dma_err_count),
+ PCH_GBE_STAT(intr_tcpip_err_count)
+};
+
+#define PCH_GBE_QUEUE_STATS_LEN 0
+#define PCH_GBE_GLOBAL_STATS_LEN ARRAY_SIZE(pch_gbe_gstrings_stats)
+#define PCH_GBE_STATS_LEN (PCH_GBE_GLOBAL_STATS_LEN + PCH_GBE_QUEUE_STATS_LEN)
+
+#define PCH_GBE_MAC_REGS_LEN (sizeof(struct pch_gbe_regs) / 4)
+#define PCH_GBE_REGS_LEN (PCH_GBE_MAC_REGS_LEN + PCH_GBE_PHY_REGS_LEN)
+/**
+ * pch_gbe_get_settings - Get device-specific settings
+ * @netdev: Network interface device structure
+ * @ecmd: Ethtool command
+ * Returns
+ * 0: Successful.
+ * Negative value: Failed.
+ */
+static int pch_gbe_get_settings(struct net_device *netdev,
+ struct ethtool_cmd *ecmd)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+ int ret;
+
+ ret = mii_ethtool_gset(&adapter->mii, ecmd);
+ ecmd->supported &= ~(SUPPORTED_TP | SUPPORTED_1000baseT_Half);
+ ecmd->advertising &= ~(ADVERTISED_TP | ADVERTISED_1000baseT_Half);
+
+ if (!netif_carrier_ok(adapter->netdev))
+ ecmd->speed = -1;
+ return ret;
+}
+
+/**
+ * pch_gbe_set_settings - Set device-specific settings
+ * @netdev: Network interface device structure
+ * @ecmd: Ethtool command
+ * Returns
+ * 0: Successful.
+ * Negative value: Failed.
+ */
+static int pch_gbe_set_settings(struct net_device *netdev,
+ struct ethtool_cmd *ecmd)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+ struct pch_gbe_hw *hw = &adapter->hw;
+ int ret;
+
+ pch_gbe_hal_write_phy_reg(hw, MII_BMCR, BMCR_RESET);
+
+ if (ecmd->speed == -1)
+ ecmd->speed = SPEED_1000;
+ ecmd->duplex = DUPLEX_FULL;
+ ret = mii_ethtool_sset(&adapter->mii, ecmd);
+ if (ret) {
+ pr_err("Error: mii_ethtool_sset\n");
+ return ret;
+ }
+ hw->mac.link_speed = ecmd->speed;
+ hw->mac.link_duplex = ecmd->duplex;
+ hw->phy.autoneg_advertised = ecmd->advertising;
+ hw->mac.autoneg = ecmd->autoneg;
+ pch_gbe_hal_phy_sw_reset(hw);
+
+ /* reset the link */
+ if (netif_running(adapter->netdev)) {
+ pch_gbe_down(adapter);
+ ret = pch_gbe_up(adapter);
+ } else {
+ pch_gbe_reset(adapter);
+ }
+ return ret;
+}
+
+/**
+ * pch_gbe_get_regs_len - Report the size of device registers
+ * @netdev: Network interface device structure
+ * Returns: the size of device registers.
+ */
+static int pch_gbe_get_regs_len(struct net_device *netdev)
+{
+ return PCH_GBE_REGS_LEN * (int)sizeof(u32);
+}
+
+/**
+ * pch_gbe_get_drvinfo - Report driver information
+ * @netdev: Network interface device structure
+ * @drvinfo: Driver information structure
+ */
+static void pch_gbe_get_drvinfo(struct net_device *netdev,
+ struct ethtool_drvinfo *drvinfo)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+
+ strcpy(drvinfo->driver, KBUILD_MODNAME);
+ strcpy(drvinfo->version, pch_driver_version);
+ strcpy(drvinfo->fw_version, "N/A");
+ strcpy(drvinfo->bus_info, pci_name(adapter->pdev));
+ drvinfo->regdump_len = pch_gbe_get_regs_len(netdev);
+}
+
+/**
+ * pch_gbe_get_regs - Get device registers
+ * @netdev: Network interface device structure
+ * @regs: Ethtool register structure
+ * @p: Buffer pointer of read device register date
+ */
+static void pch_gbe_get_regs(struct net_device *netdev,
+ struct ethtool_regs *regs, void *p)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+ struct pch_gbe_hw *hw = &adapter->hw;
+ struct pci_dev *pdev = adapter->pdev;
+ u32 *regs_buff = p;
+ u16 i, tmp;
+
+ regs->version = 0x1000000 | (__u32)pdev->revision << 16 | pdev->device;
+ for (i = 0; i < PCH_GBE_MAC_REGS_LEN; i++)
+ *regs_buff++ = ioread32(&hw->reg->INT_ST + i);
+ /* PHY register */
+ for (i = 0; i < PCH_GBE_PHY_REGS_LEN; i++) {
+ pch_gbe_hal_read_phy_reg(&adapter->hw, i, &tmp);
+ *regs_buff++ = tmp;
+ }
+}
+
+/**
+ * pch_gbe_get_wol - Report whether Wake-on-Lan is enabled
+ * @netdev: Network interface device structure
+ * @wol: Wake-on-Lan information
+ */
+static void pch_gbe_get_wol(struct net_device *netdev,
+ struct ethtool_wolinfo *wol)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+
+ wol->supported = WAKE_UCAST | WAKE_MCAST | WAKE_BCAST | WAKE_MAGIC;
+ wol->wolopts = 0;
+
+ if ((adapter->wake_up_evt & PCH_GBE_WLC_IND))
+ wol->wolopts |= WAKE_UCAST;
+ if ((adapter->wake_up_evt & PCH_GBE_WLC_MLT))
+ wol->wolopts |= WAKE_MCAST;
+ if ((adapter->wake_up_evt & PCH_GBE_WLC_BR))
+ wol->wolopts |= WAKE_BCAST;
+ if ((adapter->wake_up_evt & PCH_GBE_WLC_MP))
+ wol->wolopts |= WAKE_MAGIC;
+}
+
+/**
+ * pch_gbe_set_wol - Turn Wake-on-Lan on or off
+ * @netdev: Network interface device structure
+ * @wol: Pointer of wake-on-Lan information straucture
+ * Returns
+ * 0: Successful.
+ * Negative value: Failed.
+ */
+static int pch_gbe_set_wol(struct net_device *netdev,
+ struct ethtool_wolinfo *wol)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+
+ if ((wol->wolopts & (WAKE_PHY | WAKE_ARP | WAKE_MAGICSECURE)))
+ return -EOPNOTSUPP;
+ /* these settings will always override what we currently have */
+ adapter->wake_up_evt = 0;
+
+ if ((wol->wolopts & WAKE_UCAST))
+ adapter->wake_up_evt |= PCH_GBE_WLC_IND;
+ if ((wol->wolopts & WAKE_MCAST))
+ adapter->wake_up_evt |= PCH_GBE_WLC_MLT;
+ if ((wol->wolopts & WAKE_BCAST))
+ adapter->wake_up_evt |= PCH_GBE_WLC_BR;
+ if ((wol->wolopts & WAKE_MAGIC))
+ adapter->wake_up_evt |= PCH_GBE_WLC_MP;
+ return 0;
+}
+
+/**
+ * pch_gbe_nway_reset - Restart autonegotiation
+ * @netdev: Network interface device structure
+ * Returns
+ * 0: Successful.
+ * Negative value: Failed.
+ */
+static int pch_gbe_nway_reset(struct net_device *netdev)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+
+ return mii_nway_restart(&adapter->mii);
+}
+
+/**
+ * pch_gbe_get_ringparam - Report ring sizes
+ * @netdev: Network interface device structure
+ * @ring: Ring param structure
+ */
+static void pch_gbe_get_ringparam(struct net_device *netdev,
+ struct ethtool_ringparam *ring)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+ struct pch_gbe_tx_ring *txdr = adapter->tx_ring;
+ struct pch_gbe_rx_ring *rxdr = adapter->rx_ring;
+
+ ring->rx_max_pending = PCH_GBE_MAX_RXD;
+ ring->tx_max_pending = PCH_GBE_MAX_TXD;
+ ring->rx_mini_max_pending = 0;
+ ring->rx_jumbo_max_pending = 0;
+ ring->rx_pending = rxdr->count;
+ ring->tx_pending = txdr->count;
+ ring->rx_mini_pending = 0;
+ ring->rx_jumbo_pending = 0;
+}
+
+/**
+ * pch_gbe_set_ringparam - Set ring sizes
+ * @netdev: Network interface device structure
+ * @ring: Ring param structure
+ * Returns
+ * 0: Successful.
+ * Negative value: Failed.
+ */
+static int pch_gbe_set_ringparam(struct net_device *netdev,
+ struct ethtool_ringparam *ring)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+ struct pch_gbe_tx_ring *txdr, *tx_old;
+ struct pch_gbe_rx_ring *rxdr, *rx_old;
+ int tx_ring_size, rx_ring_size;
+ int err = 0;
+
+ if ((ring->rx_mini_pending) || (ring->rx_jumbo_pending))
+ return -EINVAL;
+ tx_ring_size = (int)sizeof(struct pch_gbe_tx_ring);
+ rx_ring_size = (int)sizeof(struct pch_gbe_rx_ring);
+
+ if ((netif_running(adapter->netdev)))
+ pch_gbe_down(adapter);
+ tx_old = adapter->tx_ring;
+ rx_old = adapter->rx_ring;
+
+ txdr = kzalloc(tx_ring_size, GFP_KERNEL);
+ if (!txdr) {
+ err = -ENOMEM;
+ goto err_alloc_tx;
+ }
+ rxdr = kzalloc(rx_ring_size, GFP_KERNEL);
+ if (!rxdr) {
+ err = -ENOMEM;
+ goto err_alloc_rx;
+ }
+ adapter->tx_ring = txdr;
+ adapter->rx_ring = rxdr;
+
+ rxdr->count =
+ clamp_val(ring->rx_pending, PCH_GBE_MIN_RXD, PCH_GBE_MAX_RXD);
+ rxdr->count = roundup(rxdr->count, PCH_GBE_RX_DESC_MULTIPLE);
+
+ txdr->count =
+ clamp_val(ring->tx_pending, PCH_GBE_MIN_RXD, PCH_GBE_MAX_RXD);
+ txdr->count = roundup(txdr->count, PCH_GBE_TX_DESC_MULTIPLE);
+
+ if ((netif_running(adapter->netdev))) {
+ /* Try to get new resources before deleting old */
+ err = pch_gbe_setup_rx_resources(adapter, adapter->rx_ring);
+ if (err)
+ goto err_setup_rx;
+ err = pch_gbe_setup_tx_resources(adapter, adapter->tx_ring);
+ if (err)
+ goto err_setup_tx;
+ /* save the new, restore the old in order to free it,
+ * then restore the new back again */
+#ifdef RINGFREE
+ adapter->rx_ring = rx_old;
+ adapter->tx_ring = tx_old;
+ pch_gbe_free_rx_resources(adapter, adapter->rx_ring);
+ pch_gbe_free_tx_resources(adapter, adapter->tx_ring);
+ kfree(tx_old);
+ kfree(rx_old);
+ adapter->rx_ring = rxdr;
+ adapter->tx_ring = txdr;
+#else
+ pch_gbe_free_rx_resources(adapter, rx_old);
+ pch_gbe_free_tx_resources(adapter, tx_old);
+ kfree(tx_old);
+ kfree(rx_old);
+ adapter->rx_ring = rxdr;
+ adapter->tx_ring = txdr;
+#endif
+ err = pch_gbe_up(adapter);
+ }
+ return err;
+
+err_setup_tx:
+ pch_gbe_free_rx_resources(adapter, adapter->rx_ring);
+err_setup_rx:
+ adapter->rx_ring = rx_old;
+ adapter->tx_ring = tx_old;
+ kfree(rxdr);
+err_alloc_rx:
+ kfree(txdr);
+err_alloc_tx:
+ if (netif_running(adapter->netdev))
+ pch_gbe_up(adapter);
+ return err;
+}
+
+/**
+ * pch_gbe_get_pauseparam - Report pause parameters
+ * @netdev: Network interface device structure
+ * @pause: Pause parameters structure
+ */
+static void pch_gbe_get_pauseparam(struct net_device *netdev,
+ struct ethtool_pauseparam *pause)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+ struct pch_gbe_hw *hw = &adapter->hw;
+
+ pause->autoneg =
+ ((hw->mac.fc_autoneg) ? AUTONEG_ENABLE : AUTONEG_DISABLE);
+
+ if (hw->mac.fc == PCH_GBE_FC_RX_PAUSE) {
+ pause->rx_pause = 1;
+ } else if (hw->mac.fc == PCH_GBE_FC_TX_PAUSE) {
+ pause->tx_pause = 1;
+ } else if (hw->mac.fc == PCH_GBE_FC_FULL) {
+ pause->rx_pause = 1;
+ pause->tx_pause = 1;
+ }
+}
+
+/**
+ * pch_gbe_set_pauseparam - Set pause paramters
+ * @netdev: Network interface device structure
+ * @pause: Pause parameters structure
+ * Returns
+ * 0: Successful.
+ * Negative value: Failed.
+ */
+static int pch_gbe_set_pauseparam(struct net_device *netdev,
+ struct ethtool_pauseparam *pause)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+ struct pch_gbe_hw *hw = &adapter->hw;
+ int ret = 0;
+
+ hw->mac.fc_autoneg = pause->autoneg;
+ if ((pause->rx_pause) && (pause->tx_pause))
+ hw->mac.fc = PCH_GBE_FC_FULL;
+ else if ((pause->rx_pause) && (!pause->tx_pause))
+ hw->mac.fc = PCH_GBE_FC_RX_PAUSE;
+ else if ((!pause->rx_pause) && (pause->tx_pause))
+ hw->mac.fc = PCH_GBE_FC_TX_PAUSE;
+ else if ((!pause->rx_pause) && (!pause->tx_pause))
+ hw->mac.fc = PCH_GBE_FC_NONE;
+
+ if (hw->mac.fc_autoneg == AUTONEG_ENABLE) {
+ if ((netif_running(adapter->netdev))) {
+ pch_gbe_down(adapter);
+ ret = pch_gbe_up(adapter);
+ } else {
+ pch_gbe_reset(adapter);
+ }
+ } else {
+ ret = pch_gbe_mac_force_mac_fc(hw);
+ }
+ return ret;
+}
+
+/**
+ * pch_gbe_get_rx_csum - Report whether receive checksums are turned on or off
+ * @netdev: Network interface device structure
+ * Returns
+ * true(1): Checksum On
+ * false(0): Checksum Off
+ */
+static u32 pch_gbe_get_rx_csum(struct net_device *netdev)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+
+ return adapter->rx_csum;
+}
+
+/**
+ * pch_gbe_set_rx_csum - Turn receive checksum on or off
+ * @netdev: Network interface device structure
+ * @data: Checksum On[true] or Off[false]
+ * Returns
+ * 0: Successful.
+ * Negative value: Failed.
+ */
+static int pch_gbe_set_rx_csum(struct net_device *netdev, u32 data)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+
+ adapter->rx_csum = data;
+ if ((netif_running(netdev)))
+ pch_gbe_reinit_locked(adapter);
+ else
+ pch_gbe_reset(adapter);
+
+ return 0;
+}
+
+/**
+ * pch_gbe_get_tx_csum - Report whether transmit checksums are turned on or off
+ * @netdev: Network interface device structure
+ * Returns
+ * true(1): Checksum On
+ * false(0): Checksum Off
+ */
+static u32 pch_gbe_get_tx_csum(struct net_device *netdev)
+{
+ return (netdev->features & NETIF_F_HW_CSUM) != 0;
+}
+
+/**
+ * pch_gbe_set_tx_csum - Turn transmit checksums on or off
+ * @netdev: Network interface device structure
+ * @data: Checksum on[true] or off[false]
+ * Returns
+ * 0: Successful.
+ * Negative value: Failed.
+ */
+static int pch_gbe_set_tx_csum(struct net_device *netdev, u32 data)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+
+ adapter->tx_csum = data;
+ if (data)
+ netdev->features |= NETIF_F_HW_CSUM;
+ else
+ netdev->features &= ~NETIF_F_HW_CSUM;
+ return 0;
+}
+
+/**
+ * pch_gbe_get_strings - Return a set of strings that describe the requested
+ * objects
+ * @netdev: Network interface device structure
+ * @stringset: Select the stringset. [ETH_SS_TEST] [ETH_SS_STATS]
+ * @data: Pointer of read string data.
+ */
+static void pch_gbe_get_strings(struct net_device *netdev, u32 stringset,
+ u8 *data)
+{
+ u8 *p = data;
+ int i;
+
+ switch (stringset) {
+ case (u32) ETH_SS_STATS:
+ for (i = 0; i < PCH_GBE_GLOBAL_STATS_LEN; i++) {
+ memcpy(p, pch_gbe_gstrings_stats[i].string,
+ ETH_GSTRING_LEN);
+ p += ETH_GSTRING_LEN;
+ }
+ break;
+ }
+}
+
+/**
+ * pch_gbe_get_ethtool_stats - Return statistics about the device
+ * @netdev: Network interface device structure
+ * @stats: Ethtool statue structure
+ * @data: Pointer of read status area
+ */
+static void pch_gbe_get_ethtool_stats(struct net_device *netdev,
+ struct ethtool_stats *stats, u64 *data)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+ int i;
+ const struct pch_gbe_stats *gstats = pch_gbe_gstrings_stats;
+ char *hw_stats = (char *)&adapter->stats;
+
+ pch_gbe_update_stats(adapter);
+ for (i = 0; i < PCH_GBE_GLOBAL_STATS_LEN; i++) {
+ char *p = hw_stats + gstats->offset;
+ data[i] = gstats->size == sizeof(u64) ? *(u64 *)p:(*(u32 *)p);
+ gstats++;
+ }
+}
+
+static int pch_gbe_get_sset_count(struct net_device *netdev, int sset)
+{
+ switch (sset) {
+ case ETH_SS_STATS:
+ return PCH_GBE_STATS_LEN;
+ default:
+ return -EOPNOTSUPP;
+ }
+}
+
+static const struct ethtool_ops pch_gbe_ethtool_ops = {
+ .get_settings = pch_gbe_get_settings,
+ .set_settings = pch_gbe_set_settings,
+ .get_drvinfo = pch_gbe_get_drvinfo,
+ .get_regs_len = pch_gbe_get_regs_len,
+ .get_regs = pch_gbe_get_regs,
+ .get_wol = pch_gbe_get_wol,
+ .set_wol = pch_gbe_set_wol,
+ .nway_reset = pch_gbe_nway_reset,
+ .get_link = ethtool_op_get_link,
+ .get_ringparam = pch_gbe_get_ringparam,
+ .set_ringparam = pch_gbe_set_ringparam,
+ .get_pauseparam = pch_gbe_get_pauseparam,
+ .set_pauseparam = pch_gbe_set_pauseparam,
+ .get_rx_csum = pch_gbe_get_rx_csum,
+ .set_rx_csum = pch_gbe_set_rx_csum,
+ .get_tx_csum = pch_gbe_get_tx_csum,
+ .set_tx_csum = pch_gbe_set_tx_csum,
+ .get_strings = pch_gbe_get_strings,
+ .get_ethtool_stats = pch_gbe_get_ethtool_stats,
+ .get_sset_count = pch_gbe_get_sset_count,
+};
+
+void pch_gbe_set_ethtool_ops(struct net_device *netdev)
+{
+ SET_ETHTOOL_OPS(netdev, &pch_gbe_ethtool_ops);
+}
diff --git a/drivers/net/pch_gbe/pch_gbe_main.c b/drivers/net/pch_gbe/pch_gbe_main.c
new file mode 100644
index 0000000..adeadfc
--- /dev/null
+++ b/drivers/net/pch_gbe/pch_gbe_main.c
@@ -0,0 +1,2473 @@
+/*
+ * Copyright (C) 1999 - 2010 Intel Corporation.
+ * Copyright (C) 2010 OKI SEMICONDUCTOR Co., LTD.
+ *
+ * This code was derived from the Intel e1000e Linux driver.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * 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 "pch_gbe.h"
+#include "pch_gbe_api.h"
+
+#define DRV_VERSION "1.00"
+const char pch_driver_version[] = DRV_VERSION;
+
+#define PCI_DEVICE_ID_INTEL_IOH1_GBE 0x8802 /* Pci device ID */
+#define PCH_GBE_MAR_ENTRIES 16
+#define PCH_GBE_SHORT_PKT 64
+#define DSC_INIT16 0xC000
+#define PCH_GBE_DMA_ALIGN 0
+#define PCH_GBE_WATCHDOG_PERIOD (1 * HZ) /* watchdog time */
+#define PCH_GBE_COPYBREAK_DEFAULT 256
+#define PCH_GBE_PCI_BAR 1
+
+#define PCH_GBE_TX_WEIGHT 64
+#define PCH_GBE_RX_WEIGHT 64
+#define PCH_GBE_RX_BUFFER_WRITE 16
+
+/* Initialize the wake-on-LAN settings */
+#define PCH_GBE_WL_INIT_SETTING (PCH_GBE_WLC_MP)
+
+#define PCH_GBE_MAC_RGMII_CTRL_SETTING ( \
+ PCH_GBE_CHIP_TYPE_INTERNAL | \
+ PCH_GBE_RGMII_MODE_RGMII | \
+ PCH_GBE_CRS_SEL \
+ )
+
+/* Ethertype field values */
+#define PCH_GBE_MAX_JUMBO_FRAME_SIZE 10318
+#define PCH_GBE_FRAME_SIZE_2048 2048
+#define PCH_GBE_FRAME_SIZE_4096 4096
+#define PCH_GBE_FRAME_SIZE_8192 8192
+
+#define PCH_GBE_GET_DESC(R, i, type) (&(((struct type *)((R).desc))[i]))
+#define PCH_GBE_RX_DESC(R, i) PCH_GBE_GET_DESC(R, i, pch_gbe_rx_desc)
+#define PCH_GBE_TX_DESC(R, i) PCH_GBE_GET_DESC(R, i, pch_gbe_tx_desc)
+#define PCH_GBE_DESC_UNUSED(R) \
+ ((((R)->next_to_clean > (R)->next_to_use) ? 0 : (R)->count) + \
+ (R)->next_to_clean - (R)->next_to_use - 1)
+
+/* Pause packet value */
+#define PCH_GBE_PAUSE_PKT1_VALUE 0x00C28001
+#define PCH_GBE_PAUSE_PKT2_VALUE 0x00000100
+#define PCH_GBE_PAUSE_PKT4_VALUE 0x01000888
+#define PCH_GBE_PAUSE_PKT5_VALUE 0x0000FFFF
+
+#define PCH_GBE_ETH_ALEN 6
+
+/* This defines the bits that are set in the Interrupt Mask
+ * Set/Read Register. Each bit is documented below:
+ * o RXT0 = Receiver Timer Interrupt (ring 0)
+ * o TXDW = Transmit Descriptor Written Back
+ * o RXDMT0 = Receive Descriptor Minimum Threshold hit (ring 0)
+ * o RXSEQ = Receive Sequence Error
+ * o LSC = Link Status Change
+ */
+#define PCH_GBE_INT_ENABLE_MASK ( \
+ PCH_GBE_INT_RX_DMA_CMPLT | \
+ PCH_GBE_INT_RX_DSC_EMP | \
+ PCH_GBE_INT_WOL_DET | \
+ PCH_GBE_INT_TX_CMPLT \
+ )
+
+
+static unsigned int copybreak __read_mostly = PCH_GBE_COPYBREAK_DEFAULT;
+
+/**
+ * pch_gbe_mac_read_mac_addr - Read MAC address
+ * @hw: Pointer to the HW structure
+ * Returns
+ * 0: Successful.
+ */
+s32 pch_gbe_mac_read_mac_addr(struct pch_gbe_hw *hw)
+{
+ u32 adr1a, adr1b;
+
+ adr1a = ioread32(&hw->reg->mac_adr[0].high);
+ adr1b = ioread32(&hw->reg->mac_adr[0].low);
+
+ hw->mac.addr[0] = (u8)(adr1a & 0xFF);
+ hw->mac.addr[1] = (u8)((adr1a >> 8) & 0xFF);
+ hw->mac.addr[2] = (u8)((adr1a >> 16) & 0xFF);
+ hw->mac.addr[3] = (u8)((adr1a >> 24) & 0xFF);
+ hw->mac.addr[4] = (u8)(adr1b & 0xFF);
+ hw->mac.addr[5] = (u8)((adr1b >> 8) & 0xFF);
+
+ pr_debug("hw->mac.addr : %pM\n", hw->mac.addr);
+ return 0;
+}
+
+/**
+ * pch_gbe_wait_clr_bit - Wait to clear a bit
+ * @reg: Pointer of register
+ * @busy: Busy bit
+ */
+void pch_gbe_wait_clr_bit(void *reg, u32 bit)
+{
+ u32 tmp;
+ /* wait busy */
+ tmp = 1000;
+ while ((ioread32(reg) & bit) && --tmp)
+ cpu_relax();
+ if (!tmp)
+ pr_err("Error: busy bit is not cleared\n");
+}
+/**
+ * pch_gbe_mac_mar_set - Set MAC address register
+ * @hw: Pointer to the HW structure
+ * @addr: Pointer to the MAC address
+ * @index: MAC address array register
+ */
+void pch_gbe_mac_mar_set(struct pch_gbe_hw *hw, u8 * addr, u32 index)
+{
+ u32 mar_low, mar_high, adrmask;
+
+ pr_debug("index : 0x%x\n", index);
+
+ /*
+ * HW expects these in little endian so we reverse the byte order
+ * from network order (big endian) to little endian
+ */
+ mar_high = ((u32) addr[0] | ((u32) addr[1] << 8) |
+ ((u32) addr[2] << 16) | ((u32) addr[3] << 24));
+ mar_low = ((u32) addr[4] | ((u32) addr[5] << 8));
+ /* Stop the MAC Address of index. */
+ adrmask = ioread32(&hw->reg->ADDR_MASK);
+ iowrite32((adrmask | (0x0001 << index)), &hw->reg->ADDR_MASK);
+ /* wait busy */
+ pch_gbe_wait_clr_bit(&hw->reg->ADDR_MASK, PCH_GBE_BUSY);
+ /* Set the MAC address to the MAC address 1A/1B register */
+ iowrite32(mar_high, &hw->reg->mac_adr[index].high);
+ iowrite32(mar_low, &hw->reg->mac_adr[index].low);
+ /* Start the MAC address of index */
+ iowrite32((adrmask & ~(0x0001 << index)), &hw->reg->ADDR_MASK);
+ /* wait busy */
+ pch_gbe_wait_clr_bit(&hw->reg->ADDR_MASK, PCH_GBE_BUSY);
+}
+
+/**
+ * pch_gbe_mac_reset_hw - Reset hardware
+ * @hw: Pointer to the HW structure
+ */
+void pch_gbe_mac_reset_hw(struct pch_gbe_hw *hw)
+{
+ /* Read the MAC address. and store to the private data */
+ pch_gbe_mac_read_mac_addr(hw);
+ iowrite32(PCH_GBE_ALL_RST, &hw->reg->RESET);
+#ifdef PCH_GBE_MAC_IFOP_RGMII
+ iowrite32(PCH_GBE_MODE_GMII_ETHER, &hw->reg->MODE);
+#endif
+ pch_gbe_wait_clr_bit(&hw->reg->RESET, PCH_GBE_ALL_RST);
+ /* Setup the receive address */
+ pch_gbe_mac_mar_set(hw, hw->mac.addr, 0);
+ return;
+}
+
+/**
+ * pch_gbe_mac_init_rx_addrs - Initialize receive address's
+ * @hw: Pointer to the HW structure
+ * @mar_count: Receive address registers
+ */
+void pch_gbe_mac_init_rx_addrs(struct pch_gbe_hw *hw, u16 mar_count)
+{
+ u32 i;
+
+ /* Setup the receive address */
+ pch_gbe_mac_mar_set(hw, hw->mac.addr, 0);
+
+ /* Zero out the other receive addresses */
+ for (i = 1; i < mar_count; i++) {
+ iowrite32(0, &hw->reg->mac_adr[i].high);
+ iowrite32(0, &hw->reg->mac_adr[i].low);
+ }
+ iowrite32(0xFFFE, &hw->reg->ADDR_MASK);
+ /* wait busy */
+ pch_gbe_wait_clr_bit(&hw->reg->ADDR_MASK, PCH_GBE_BUSY);
+}
+
+
+/**
+ * pch_gbe_mac_mc_addr_list_update - Update Multicast addresses
+ * @hw: Pointer to the HW structure
+ * @mc_addr_list: Array of multicast addresses to program
+ * @mc_addr_count: Number of multicast addresses to program
+ * @mar_used_count: The first MAC Address register free to program
+ * @mar_total_num: Total number of supported MAC Address Registers
+ */
+void pch_gbe_mac_mc_addr_list_update(struct pch_gbe_hw *hw,
+ u8 *mc_addr_list, u32 mc_addr_count,
+ u32 mar_used_count, u32 mar_total_num)
+{
+ u32 i, adrmask;
+
+ /* Load the first set of multicast addresses into the exact
+ * filters (RAR). If there are not enough to fill the RAR
+ * array, clear the filters.
+ */
+ for (i = mar_used_count; i < mar_total_num; i++) {
+ if (mc_addr_count) {
+ pch_gbe_mac_mar_set(hw, mc_addr_list, i);
+ mc_addr_count--;
+ mc_addr_list += PCH_GBE_ETH_ALEN;
+ } else {
+ /* Clear MAC address mask */
+ adrmask = ioread32(&hw->reg->ADDR_MASK);
+ iowrite32((adrmask | (0x0001 << i)),
+ &hw->reg->ADDR_MASK);
+ /* wait busy */
+ pch_gbe_wait_clr_bit(&hw->reg->ADDR_MASK, PCH_GBE_BUSY);
+ /* Clear MAC address */
+ iowrite32(0, &hw->reg->mac_adr[i].high);
+ iowrite32(0, &hw->reg->mac_adr[i].low);
+ }
+ }
+}
+
+/**
+ * pch_gbe_mac_force_mac_fc - Force the MAC's flow control settings
+ * @hw: Pointer to the HW structure
+ * Returns
+ * 0: Successful.
+ * Negative value: Failed.
+ */
+s32 pch_gbe_mac_force_mac_fc(struct pch_gbe_hw *hw)
+{
+ struct pch_gbe_mac_info *mac = &hw->mac;
+ u32 rx_fctrl;
+
+ pr_debug("mac->fc = %u\n", mac->fc);
+
+ rx_fctrl = ioread32(&hw->reg->RX_FCTRL);
+
+ switch (mac->fc) {
+ case PCH_GBE_FC_NONE:
+ rx_fctrl &= ~PCH_GBE_FL_CTRL_EN;
+ mac->tx_fc_enable = false;
+ break;
+ case PCH_GBE_FC_RX_PAUSE:
+ rx_fctrl |= PCH_GBE_FL_CTRL_EN;
+ mac->tx_fc_enable = false;
+ break;
+ case PCH_GBE_FC_TX_PAUSE:
+ rx_fctrl &= ~PCH_GBE_FL_CTRL_EN;
+ mac->tx_fc_enable = true;
+ break;
+ case PCH_GBE_FC_FULL:
+ rx_fctrl |= PCH_GBE_FL_CTRL_EN;
+ mac->tx_fc_enable = true;
+ break;
+ default:
+ pr_err("Flow control param set incorrectly\n");
+ return -EINVAL;
+ }
+ if (mac->link_duplex == DUPLEX_HALF)
+ rx_fctrl &= ~PCH_GBE_FL_CTRL_EN;
+ iowrite32(rx_fctrl, &hw->reg->RX_FCTRL);
+ pr_debug("RX_FCTRL reg : 0x%08x mac->tx_fc_enable : %d\n",
+ ioread32(&hw->reg->RX_FCTRL), mac->tx_fc_enable);
+ return 0;
+}
+
+/**
+ * pch_gbe_mac_set_wol_event - Set wake-on-lan event
+ * @hw: Pointer to the HW structure
+ * @wu_evt: Wake up event
+ */
+void pch_gbe_mac_set_wol_event(struct pch_gbe_hw *hw, u32 wu_evt)
+{
+ u32 addr_mask;
+
+ pr_debug("wu_evt : 0x%08x ADDR_MASK reg : 0x%08x\n",
+ wu_evt, ioread32(&hw->reg->ADDR_MASK));
+
+ if (wu_evt) {
+ /* Set Wake-On-Lan address mask */
+ addr_mask = ioread32(&hw->reg->ADDR_MASK);
+ iowrite32(addr_mask, &hw->reg->WOL_ADDR_MASK);
+ /* wait busy */
+ pch_gbe_wait_clr_bit(&hw->reg->WOL_ADDR_MASK, PCH_GBE_WLA_BUSY);
+ iowrite32(0, &hw->reg->WOL_ST);
+ iowrite32((wu_evt | PCH_GBE_WLC_WOL_MODE), &hw->reg->WOL_CTRL);
+ iowrite32(0x02, &hw->reg->TCPIP_ACC);
+ iowrite32(PCH_GBE_INT_ENABLE_MASK, &hw->reg->INT_EN);
+ } else {
+ iowrite32(0, &hw->reg->WOL_CTRL);
+ iowrite32(0, &hw->reg->WOL_ST);
+ }
+ return;
+}
+
+/**
+ * pch_gbe_mac_ctrl_miim - Control MIIM interface
+ * @hw: Pointer to the HW structure
+ * @addr: Address of PHY
+ * @dir: Operetion. (Write or Read)
+ * @reg: Access register of PHY
+ * @data: Write data.
+ *
+ * Returns: Read date.
+ */
+u16 pch_gbe_mac_ctrl_miim(struct pch_gbe_hw *hw, u32 addr, u32 dir, u32 reg,
+ u16 data)
+{
+ u32 data_out = 0;
+ unsigned int i;
+ unsigned long flags;
+
+ spin_lock_irqsave(&hw->miim_lock, flags);
+
+ for (i = 100; i; --i) {
+ if ((ioread32(&hw->reg->MIIM) & PCH_GBE_MIIM_OPER_READY))
+ break;
+ udelay(20);
+ }
+ if (i == 0) {
+ pr_err("pch-gbe.miim won't go Ready\n");
+ spin_unlock_irqrestore(&hw->miim_lock, flags);
+ return 0; /* No way to indicate timeout error */
+ }
+ iowrite32(((reg << PCH_GBE_MIIM_REG_ADDR_SHIFT) |
+ (addr << PCH_GBE_MIIM_PHY_ADDR_SHIFT) |
+ dir | data), &hw->reg->MIIM);
+ for (i = 0; i < 100; i++) {
+ udelay(20);
+ data_out = ioread32(&hw->reg->MIIM);
+ if ((data_out & PCH_GBE_MIIM_OPER_READY))
+ break;
+ }
+ spin_unlock_irqrestore(&hw->miim_lock, flags);
+
+ pr_debug("PHY %s: reg=%d, data=0x%04X\n",
+ dir == PCH_GBE_MIIM_OPER_READ ? "READ" : "WRITE", reg,
+ dir == PCH_GBE_MIIM_OPER_READ ? data_out : data);
+ return (u16) data_out;
+}
+
+/**
+ * pch_gbe_mac_set_pause_packet - Set pause packet
+ * @hw: Pointer to the HW structure
+ */
+void pch_gbe_mac_set_pause_packet(struct pch_gbe_hw *hw)
+{
+ unsigned long tmp2, tmp3;
+
+ /* Set Pause packet */
+ tmp2 = hw->mac.addr[1];
+ tmp2 = (tmp2 << 8) | hw->mac.addr[0];
+ tmp2 = PCH_GBE_PAUSE_PKT2_VALUE | (tmp2 << 16);
+
+ tmp3 = hw->mac.addr[5];
+ tmp3 = (tmp3 << 8) | hw->mac.addr[4];
+ tmp3 = (tmp3 << 8) | hw->mac.addr[3];
+ tmp3 = (tmp3 << 8) | hw->mac.addr[2];
+
+ iowrite32(PCH_GBE_PAUSE_PKT1_VALUE, &hw->reg->PAUSE_PKT1);
+ iowrite32(tmp2, &hw->reg->PAUSE_PKT2);
+ iowrite32(tmp3, &hw->reg->PAUSE_PKT3);
+ iowrite32(PCH_GBE_PAUSE_PKT4_VALUE, &hw->reg->PAUSE_PKT4);
+ iowrite32(PCH_GBE_PAUSE_PKT5_VALUE, &hw->reg->PAUSE_PKT5);
+
+ /* Transmit Pause Packet */
+ iowrite32(PCH_GBE_PS_PKT_RQ, &hw->reg->PAUSE_REQ);
+
+ pr_debug("PAUSE_PKT1-5 reg : 0x%08x 0x%08x 0x%08x 0x%08x 0x%08x\n",
+ ioread32(&hw->reg->PAUSE_PKT1), ioread32(&hw->reg->PAUSE_PKT2),
+ ioread32(&hw->reg->PAUSE_PKT3), ioread32(&hw->reg->PAUSE_PKT4),
+ ioread32(&hw->reg->PAUSE_PKT5));
+
+ return;
+}
+
+
+/**
+ * pch_gbe_alloc_queues - Allocate memory for all rings
+ * @adapter: Board private structure to initialize
+ * Returns
+ * 0: Successfully
+ * Negative value: Failed
+ */
+static int pch_gbe_alloc_queues(struct pch_gbe_adapter *adapter)
+{
+ int size;
+
+ size = (int)sizeof(struct pch_gbe_tx_ring);
+ adapter->tx_ring = kzalloc(size, GFP_KERNEL);
+ if (!adapter->tx_ring)
+ return -ENOMEM;
+ size = (int)sizeof(struct pch_gbe_rx_ring);
+ adapter->rx_ring = kzalloc(size, GFP_KERNEL);
+ if (!adapter->rx_ring) {
+ kfree(adapter->tx_ring);
+ return -ENOMEM;
+ }
+ return 0;
+}
+
+/**
+ * pch_gbe_init_stats - Initialize status
+ * @adapter: Board private structure to initialize
+ */
+static void pch_gbe_init_stats(struct pch_gbe_adapter *adapter)
+{
+ memset(&adapter->stats, 0, sizeof(adapter->stats));
+ return;
+}
+
+/**
+ * pch_gbe_init_phy - Initialize PHY
+ * @adapter: Board private structure to initialize
+ * Returns
+ * 0: Successfully
+ * Negative value: Failed
+ */
+static int pch_gbe_init_phy(struct pch_gbe_adapter *adapter)
+{
+ struct net_device *netdev = adapter->netdev;
+ u32 addr;
+ u16 bmcr, stat;
+
+ /* Discover phy addr by searching addrs in order {1,0,2,..., 31} */
+ for (addr = 0; addr < PCH_GBE_PHY_REGS_LEN; addr++) {
+ adapter->mii.phy_id = (addr == 0) ? 1 : (addr == 1) ? 0 : addr;
+ bmcr = pch_gbe_mdio_read(netdev, adapter->mii.phy_id, MII_BMCR);
+ stat = pch_gbe_mdio_read(netdev, adapter->mii.phy_id, MII_BMSR);
+ stat = pch_gbe_mdio_read(netdev, adapter->mii.phy_id, MII_BMSR);
+ if (!((bmcr == 0xFFFF) || ((stat == 0) && (bmcr == 0))))
+ break;
+ }
+ adapter->hw.phy.addr = adapter->mii.phy_id;
+ pr_debug("phy_addr = %d\n", adapter->mii.phy_id);
+ if (addr == 32)
+ return -EAGAIN;
+ /* Selected the phy and isolate the rest */
+ for (addr = 0; addr < PCH_GBE_PHY_REGS_LEN; addr++) {
+ if (addr != adapter->mii.phy_id) {
+ pch_gbe_mdio_write(netdev, addr, MII_BMCR,
+ BMCR_ISOLATE);
+ } else {
+ bmcr = pch_gbe_mdio_read(netdev, addr, MII_BMCR);
+ pch_gbe_mdio_write(netdev, addr, MII_BMCR,
+ bmcr & ~BMCR_ISOLATE);
+ }
+ }
+
+ /* MII setup */
+ adapter->mii.phy_id_mask = 0x1F;
+ adapter->mii.reg_num_mask = 0x1F;
+ adapter->mii.dev = adapter->netdev;
+ adapter->mii.mdio_read = pch_gbe_mdio_read;
+ adapter->mii.mdio_write = pch_gbe_mdio_write;
+ adapter->mii.supports_gmii = mii_check_gmii_support(&adapter->mii);
+ return 0;
+}
+
+/**
+ * pch_gbe_mdio_read - The read function for mii
+ * @netdev: Network interface device structure
+ * @addr: Phy ID
+ * @reg: Access location
+ * Returns
+ * 0: Successfully
+ * Negative value: Failed
+ */
+int pch_gbe_mdio_read(struct net_device *netdev, int addr, int reg)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+ struct pch_gbe_hw *hw = &adapter->hw;
+
+ return pch_gbe_mac_ctrl_miim(hw, addr, PCH_GBE_HAL_MIIM_READ, reg,
+ (u16) 0);
+}
+
+/**
+ * pch_gbe_mdio_write - The write function for mii
+ * @netdev: Network interface device structure
+ * @addr: Phy ID (not used)
+ * @reg: Access location
+ * @data: Write data
+ */
+void pch_gbe_mdio_write(struct net_device *netdev, int addr, int reg, int data)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+ struct pch_gbe_hw *hw = &adapter->hw;
+
+ pch_gbe_mac_ctrl_miim(hw, addr, PCH_GBE_HAL_MIIM_WRITE, reg, data);
+}
+
+/**
+ * pch_gbe_reset_task - Reset processing at the time of transmission timeout
+ * @work: Pointer of board private structure
+ */
+static void pch_gbe_reset_task(struct work_struct *work)
+{
+ struct pch_gbe_adapter *adapter;
+ adapter = container_of(work, struct pch_gbe_adapter, reset_task);
+
+ pch_gbe_reinit_locked(adapter);
+}
+
+/**
+ * pch_gbe_reinit_locked- Re-initialization
+ * @adapter: Board private structure
+ */
+void pch_gbe_reinit_locked(struct pch_gbe_adapter *adapter)
+{
+ struct net_device *netdev = adapter->netdev;
+
+ rtnl_lock();
+ if (netif_running(netdev)) {
+ pch_gbe_down(adapter);
+ pch_gbe_up(adapter);
+ }
+ rtnl_unlock();
+}
+
+/**
+ * pch_gbe_reset - Reset GbE
+ * @adapter: Board private structure
+ */
+void pch_gbe_reset(struct pch_gbe_adapter *adapter)
+{
+ pch_gbe_mac_reset_hw(&adapter->hw);
+ /* Setup the receive address. */
+ pch_gbe_mac_init_rx_addrs(&adapter->hw, PCH_GBE_MAR_ENTRIES);
+ if (pch_gbe_hal_init_hw(&adapter->hw))
+ pr_err("Hardware Error\n");
+}
+
+/**
+ * pch_gbe_free_irq - Free an interrupt
+ * @adapter: Board private structure
+ */
+static void pch_gbe_free_irq(struct pch_gbe_adapter *adapter)
+{
+ struct net_device *netdev = adapter->netdev;
+
+ free_irq(adapter->pdev->irq, netdev);
+ if (adapter->have_msi) {
+ pci_disable_msi(adapter->pdev);
+ pr_debug("call pci_disable_msi\n");
+ }
+}
+
+/**
+ * pch_gbe_irq_disable - Mask off interrupt generation on the NIC
+ * @adapter: Board private structure
+ */
+static void pch_gbe_irq_disable(struct pch_gbe_adapter *adapter)
+{
+ struct pch_gbe_hw *hw = &adapter->hw;
+
+ atomic_inc(&adapter->irq_sem);
+ iowrite32(0, &hw->reg->INT_EN);
+ ioread32(&hw->reg->INT_ST);
+ synchronize_irq(adapter->pdev->irq);
+
+ pr_debug("INT_EN reg : 0x%08x\n", ioread32(&hw->reg->INT_EN));
+}
+
+/**
+ * pch_gbe_irq_enable - Enable default interrupt generation settings
+ * @adapter: Board private structure
+ */
+static void pch_gbe_irq_enable(struct pch_gbe_adapter *adapter)
+{
+ struct pch_gbe_hw *hw = &adapter->hw;
+
+ if (likely(atomic_dec_and_test(&adapter->irq_sem)))
+ iowrite32(PCH_GBE_INT_ENABLE_MASK, &hw->reg->INT_EN);
+ ioread32(&hw->reg->INT_ST);
+ pr_debug("INT_EN reg : 0x%08x\n", ioread32(&hw->reg->INT_EN));
+}
+
+
+
+/**
+ * pch_gbe_setup_tctl - configure the Transmit control registers
+ * @adapter: Board private structure
+ */
+static void pch_gbe_setup_tctl(struct pch_gbe_adapter *adapter)
+{
+ struct pch_gbe_hw *hw = &adapter->hw;
+ u32 tx_mode, tcpip;
+
+ tx_mode = PCH_GBE_TM_LONG_PKT |
+ PCH_GBE_TM_ST_AND_FD |
+ PCH_GBE_TM_SHORT_PKT |
+ PCH_GBE_TM_TH_TX_STRT_8 |
+ PCH_GBE_TM_TH_ALM_EMP_4 | PCH_GBE_TM_TH_ALM_FULL_8;
+
+ iowrite32(tx_mode, &hw->reg->TX_MODE);
+
+ tcpip = ioread32(&hw->reg->TCPIP_ACC);
+ tcpip |= PCH_GBE_TX_TCPIPACC_EN;
+ iowrite32(tcpip, &hw->reg->TCPIP_ACC);
+ return;
+}
+
+/**
+ * pch_gbe_configure_tx - Configure Transmit Unit after Reset
+ * @adapter: Board private structure
+ */
+static void pch_gbe_configure_tx(struct pch_gbe_adapter *adapter)
+{
+ struct pch_gbe_hw *hw = &adapter->hw;
+ u32 tdba, tdlen, dctrl;
+
+ pr_debug("dma addr = 0x%08llx size = 0x%08x\n",
+ (unsigned long long)adapter->tx_ring->dma,
+ adapter->tx_ring->size);
+
+ /* Setup the HW Tx Head and Tail descriptor pointers */
+ tdba = adapter->tx_ring->dma;
+ tdlen = adapter->tx_ring->size - 0x10;
+ iowrite32(tdba, &hw->reg->TX_DSC_BASE);
+ iowrite32(tdlen, &hw->reg->TX_DSC_SIZE);
+ iowrite32(tdba, &hw->reg->TX_DSC_SW_P);
+
+ /* Enables Transmission DMA */
+ dctrl = ioread32(&hw->reg->DMA_CTRL);
+ dctrl |= PCH_GBE_TX_DMA_EN;
+ iowrite32(dctrl, &hw->reg->DMA_CTRL);
+}
+
+/**
+ * pch_gbe_setup_rctl - Configure the receive control registers
+ * @adapter: Board private structure
+ */
+static void pch_gbe_setup_rctl(struct pch_gbe_adapter *adapter)
+{
+ struct pch_gbe_hw *hw = &adapter->hw;
+ u32 rx_mode, tcpip;
+
+ rx_mode = PCH_GBE_ADD_FIL_EN | PCH_GBE_MLT_FIL_EN |
+ PCH_GBE_RH_ALM_EMP_4 | PCH_GBE_RH_ALM_FULL_4 | PCH_GBE_RH_RD_TRG_8;
+
+ iowrite32(rx_mode, &hw->reg->RX_MODE);
+
+ tcpip = ioread32(&hw->reg->TCPIP_ACC);
+
+ if (adapter->rx_csum) {
+ tcpip &= ~PCH_GBE_RX_TCPIPACC_OFF;
+ tcpip |= PCH_GBE_RX_TCPIPACC_EN;
+ } else {
+ tcpip |= PCH_GBE_RX_TCPIPACC_OFF;
+ tcpip &= ~PCH_GBE_RX_TCPIPACC_EN;
+ }
+ iowrite32(tcpip, &hw->reg->TCPIP_ACC);
+ return;
+}
+
+/**
+ * pch_gbe_configure_rx - Configure Receive Unit after Reset
+ * @adapter: Board private structure
+ */
+static void pch_gbe_configure_rx(struct pch_gbe_adapter *adapter)
+{
+ struct pch_gbe_hw *hw = &adapter->hw;
+ u32 rdba, rdlen, rctl, rxdma;
+
+ pr_debug("dma adr = 0x%08llx size = 0x%08x\n",
+ (unsigned long long)adapter->rx_ring->dma,
+ adapter->rx_ring->size);
+
+ pch_gbe_mac_force_mac_fc(hw);
+
+ /* Disables Receive MAC */
+ rctl = ioread32(&hw->reg->MAC_RX_EN);
+ iowrite32((rctl & ~PCH_GBE_MRE_MAC_RX_EN), &hw->reg->MAC_RX_EN);
+
+ /* Disables Receive DMA */
+ rxdma = ioread32(&hw->reg->DMA_CTRL);
+ rxdma &= ~PCH_GBE_RX_DMA_EN;
+ iowrite32(rxdma, &hw->reg->DMA_CTRL);
+
+ pr_debug("MAC_RX_EN reg = 0x%08x DMA_CTRL reg = 0x%08x\n",
+ ioread32(&hw->reg->MAC_RX_EN),
+ ioread32(&hw->reg->DMA_CTRL));
+
+ /* Setup the HW Rx Head and Tail Descriptor Pointers and
+ * the Base and Length of the Rx Descriptor Ring */
+ rdba = adapter->rx_ring->dma;
+ rdlen = adapter->rx_ring->size - 0x10;
+ iowrite32(rdba, &hw->reg->RX_DSC_BASE);
+ iowrite32(rdlen, &hw->reg->RX_DSC_SIZE);
+ iowrite32((rdba + rdlen), &hw->reg->RX_DSC_SW_P);
+
+ /* Enables Receive DMA */
+ rxdma = ioread32(&hw->reg->DMA_CTRL);
+ rxdma |= PCH_GBE_RX_DMA_EN;
+ iowrite32(rxdma, &hw->reg->DMA_CTRL);
+ /* Enables Receive */
+ iowrite32(PCH_GBE_MRE_MAC_RX_EN, &hw->reg->MAC_RX_EN);
+}
+
+/**
+ * pch_gbe_unmap_and_free_tx_resource - Unmap and free tx socket buffer
+ * @adapter: Board private structure
+ * @buffer_info: Buffer information structure
+ */
+static void pch_gbe_unmap_and_free_tx_resource(
+ struct pch_gbe_adapter *adapter, struct pch_gbe_buffer *buffer_info)
+{
+ if (buffer_info->mapped) {
+ dma_unmap_single(&adapter->pdev->dev, buffer_info->dma,
+ buffer_info->length, DMA_TO_DEVICE);
+ buffer_info->mapped = false;
+ }
+ if (buffer_info->skb) {
+ dev_kfree_skb_any(buffer_info->skb);
+ buffer_info->skb = NULL;
+ }
+}
+
+/**
+ * pch_gbe_unmap_and_free_rx_resource - Unmap and free rx socket buffer
+ * @adapter: Board private structure
+ * @buffer_info: Buffer information structure
+ */
+static void pch_gbe_unmap_and_free_rx_resource(
+ struct pch_gbe_adapter *adapter,
+ struct pch_gbe_buffer *buffer_info)
+{
+ if (buffer_info->mapped) {
+ dma_unmap_single(&adapter->pdev->dev, buffer_info->dma,
+ buffer_info->length, DMA_FROM_DEVICE);
+ buffer_info->mapped = false;
+ }
+ if (buffer_info->skb) {
+ dev_kfree_skb_any(buffer_info->skb);
+ buffer_info->skb = NULL;
+ }
+}
+
+/**
+ * pch_gbe_clean_tx_ring - Free Tx Buffers
+ * @adapter: Board private structure
+ * @tx_ring: Ring to be cleaned
+ */
+static void pch_gbe_clean_tx_ring(struct pch_gbe_adapter *adapter,
+ struct pch_gbe_tx_ring *tx_ring)
+{
+ struct pch_gbe_hw *hw = &adapter->hw;
+ struct pch_gbe_buffer *buffer_info;
+ unsigned long size;
+ unsigned int i;
+
+ /* Free all the Tx ring sk_buffs */
+ for (i = 0; i < tx_ring->count; i++) {
+ buffer_info = &tx_ring->buffer_info[i];
+ pch_gbe_unmap_and_free_tx_resource(adapter, buffer_info);
+ }
+ pr_debug("call pch_gbe_unmap_and_free_tx_resource() %d count\n", i);
+
+ size = (unsigned long)sizeof(struct pch_gbe_buffer) * tx_ring->count;
+ memset(tx_ring->buffer_info, 0, size);
+
+ /* Zero out the descriptor ring */
+ memset(tx_ring->desc, 0, tx_ring->size);
+ tx_ring->next_to_use = 0;
+ tx_ring->next_to_clean = 0;
+ iowrite32(tx_ring->dma, &hw->reg->TX_DSC_HW_P);
+ iowrite32((tx_ring->size - 0x10), &hw->reg->TX_DSC_SIZE);
+}
+
+/**
+ * pch_gbe_clean_rx_ring - Free Rx Buffers
+ * @adapter: Board private structure
+ * @rx_ring: Ring to free buffers from
+ */
+static void
+pch_gbe_clean_rx_ring(struct pch_gbe_adapter *adapter,
+ struct pch_gbe_rx_ring *rx_ring)
+{
+ struct pch_gbe_hw *hw = &adapter->hw;
+ struct pch_gbe_buffer *buffer_info;
+ unsigned long size;
+ unsigned int i;
+
+ /* Free all the Rx ring sk_buffs */
+ for (i = 0; i < rx_ring->count; i++) {
+ buffer_info = &rx_ring->buffer_info[i];
+ pch_gbe_unmap_and_free_rx_resource(adapter, buffer_info);
+ }
+ pr_debug("call pch_gbe_unmap_and_free_rx_resource() %d count\n", i);
+ size = (unsigned long)sizeof(struct pch_gbe_buffer) * rx_ring->count;
+ memset(rx_ring->buffer_info, 0, size);
+
+ /* Zero out the descriptor ring */
+ memset(rx_ring->desc, 0, rx_ring->size);
+ rx_ring->next_to_clean = 0;
+ rx_ring->next_to_use = 0;
+ iowrite32(rx_ring->dma, &hw->reg->RX_DSC_HW_P);
+ iowrite32((rx_ring->size - 0x10), &hw->reg->RX_DSC_SIZE);
+}
+
+static void pch_gbe_set_rgmii_ctrl(struct pch_gbe_adapter *adapter, u16 speed,
+ u16 duplex)
+{
+ struct pch_gbe_hw *hw = &adapter->hw;
+ unsigned long rgmii = 0;
+
+ /* Set the RGMII control. */
+#ifdef PCH_GBE_MAC_IFOP_RGMII
+ switch (speed) {
+ case SPEED_10:
+ rgmii = (PCH_GBE_RGMII_RATE_2_5M |
+ PCH_GBE_MAC_RGMII_CTRL_SETTING);
+ break;
+ case SPEED_100:
+ rgmii = (PCH_GBE_RGMII_RATE_25M |
+ PCH_GBE_MAC_RGMII_CTRL_SETTING);
+ break;
+ case SPEED_1000:
+ rgmii = (PCH_GBE_RGMII_RATE_125M |
+ PCH_GBE_MAC_RGMII_CTRL_SETTING);
+ break;
+ }
+ iowrite32(rgmii, &hw->reg->RGMII_CTRL);
+#else /* GMII */
+ rgmii = 0;
+ iowrite32(rgmii, &hw->reg->RGMII_CTRL);
+#endif
+}
+static void pch_gbe_set_mode(struct pch_gbe_adapter *adapter, u16 speed,
+ u16 duplex)
+{
+ struct net_device *netdev = adapter->netdev;
+ struct pch_gbe_hw *hw = &adapter->hw;
+ unsigned long mode = 0;
+
+ /* Set the communication mode */
+ switch (speed) {
+ case SPEED_10:
+ mode = PCH_GBE_MODE_MII_ETHER;
+ netdev->tx_queue_len = 10;
+ break;
+ case SPEED_100:
+ mode = PCH_GBE_MODE_MII_ETHER;
+ netdev->tx_queue_len = 100;
+ break;
+ case SPEED_1000:
+ mode = PCH_GBE_MODE_GMII_ETHER;
+ break;
+ }
+ if (duplex == DUPLEX_FULL)
+ mode |= PCH_GBE_MODE_FULL_DUPLEX;
+ else
+ mode |= PCH_GBE_MODE_HALF_DUPLEX;
+ iowrite32(mode, &hw->reg->MODE);
+}
+
+/**
+ * pch_gbe_watchdog - Watchdog process
+ * @data: Board private structure
+ */
+static void pch_gbe_watchdog(unsigned long data)
+{
+ struct pch_gbe_adapter *adapter = (struct pch_gbe_adapter *)data;
+ struct net_device *netdev = adapter->netdev;
+ struct pch_gbe_hw *hw = &adapter->hw;
+ struct ethtool_cmd cmd;
+
+ pr_debug("right now = %ld\n", jiffies);
+
+ pch_gbe_update_stats(adapter);
+ if ((mii_link_ok(&adapter->mii)) && (!netif_carrier_ok(netdev))) {
+ netdev->tx_queue_len = adapter->tx_queue_len;
+ /* mii library handles link maintenance tasks */
+ if (mii_ethtool_gset(&adapter->mii, &cmd)) {
+ pr_err("ethtool get setting Error\n");
+ mod_timer(&adapter->watchdog_timer,
+ round_jiffies(jiffies +
+ PCH_GBE_WATCHDOG_PERIOD));
+ return;
+ }
+ hw->mac.link_speed = cmd.speed;
+ hw->mac.link_duplex = cmd.duplex;
+ /* Set the RGMII control. */
+ pch_gbe_set_rgmii_ctrl(adapter, hw->mac.link_speed,
+ hw->mac.link_duplex);
+ /* Set the communication mode */
+ pch_gbe_set_mode(adapter, hw->mac.link_speed,
+ hw->mac.link_duplex);
+ netdev_dbg(netdev,
+ "Link is Up %d Mbps %s-Duplex\n",
+ cmd.speed,
+ cmd.duplex == DUPLEX_FULL ? "Full" : "Half");
+ netif_carrier_on(netdev);
+ netif_wake_queue(netdev);
+ } else if ((!mii_link_ok(&adapter->mii)) &&
+ (netif_carrier_ok(netdev))) {
+ netdev_dbg(netdev, "NIC Link is Down\n");
+ hw->mac.link_speed = SPEED_10;
+ hw->mac.link_duplex = DUPLEX_HALF;
+ netif_carrier_off(netdev);
+ netif_stop_queue(netdev);
+ }
+ mod_timer(&adapter->watchdog_timer,
+ round_jiffies(jiffies + PCH_GBE_WATCHDOG_PERIOD));
+}
+
+/**
+ * pch_gbe_tx_queue - Carry out queuing of the transmission data
+ * @adapter: Board private structure
+ * @tx_ring: Tx descriptor ring structure
+ * @skb: Sockt buffer structure
+ */
+static void pch_gbe_tx_queue(struct pch_gbe_adapter *adapter,
+ struct pch_gbe_tx_ring *tx_ring,
+ struct sk_buff *skb)
+{
+ struct pch_gbe_hw *hw = &adapter->hw;
+ struct pch_gbe_tx_desc *tx_desc;
+ struct pch_gbe_buffer *buffer_info;
+ struct sk_buff *tmp_skb;
+ unsigned int frame_ctrl;
+ unsigned int ring_num;
+ unsigned long flags;
+
+ /*-- Set frame control --*/
+ frame_ctrl = 0;
+ if (unlikely(skb->len < PCH_GBE_SHORT_PKT))
+ frame_ctrl |= PCH_GBE_TXD_CTRL_APAD;
+ if (unlikely(!adapter->tx_csum))
+ frame_ctrl |= PCH_GBE_TXD_CTRL_TCPIP_ACC_OFF;
+
+ /* Performs checksum processing */
+ /*
+ * It is because the hardware accelerator does not support a checksum,
+ * when the received data size is less than 64 bytes.
+ */
+ if ((skb->len < PCH_GBE_SHORT_PKT) && (adapter->tx_csum)) {
+ frame_ctrl |= PCH_GBE_TXD_CTRL_APAD |
+ PCH_GBE_TXD_CTRL_TCPIP_ACC_OFF;
+ if (skb->protocol == htons(ETH_P_IP)) {
+ struct iphdr *iph = ip_hdr(skb);
+ unsigned int offset;
+ iph->check = 0;
+ iph->check = ip_fast_csum((u8 *) iph, iph->ihl);
+ offset = skb_transport_offset(skb);
+ if (iph->protocol == IPPROTO_TCP) {
+ skb->csum = 0;
+ tcp_hdr(skb)->check = 0;
+ skb->csum = skb_checksum(skb, offset,
+ skb->len - offset, 0);
+ tcp_hdr(skb)->check =
+ csum_tcpudp_magic(iph->saddr,
+ iph->daddr,
+ skb->len - offset,
+ IPPROTO_TCP,
+ skb->csum);
+ } else if (iph->protocol == IPPROTO_UDP) {
+ skb->csum = 0;
+ udp_hdr(skb)->check = 0;
+ skb->csum =
+ skb_checksum(skb, offset,
+ skb->len - offset, 0);
+ udp_hdr(skb)->check =
+ csum_tcpudp_magic(iph->saddr,
+ iph->daddr,
+ skb->len - offset,
+ IPPROTO_UDP,
+ skb->csum);
+ }
+ }
+ }
+ spin_lock_irqsave(&tx_ring->tx_lock, flags);
+ ring_num = tx_ring->next_to_use;
+ if (unlikely((ring_num + 1) == tx_ring->count))
+ tx_ring->next_to_use = 0;
+ else
+ tx_ring->next_to_use = ring_num + 1;
+
+ spin_unlock_irqrestore(&tx_ring->tx_lock, flags);
+ buffer_info = &tx_ring->buffer_info[ring_num];
+ tmp_skb = buffer_info->skb;
+
+ /* [Header:14][payload] ---> [Header:14][paddong:2][payload] */
+ memcpy(tmp_skb->data, skb->data, ETH_HLEN);
+ tmp_skb->data[ETH_HLEN] = 0x00;
+ tmp_skb->data[ETH_HLEN + 1] = 0x00;
+ tmp_skb->len = skb->len;
+ memcpy(&tmp_skb->data[ETH_HLEN + 2], &skb->data[ETH_HLEN],
+ (skb->len - ETH_HLEN));
+ /*-- Set Buffer infomation --*/
+ buffer_info->length = tmp_skb->len;
+ buffer_info->dma = dma_map_single(&adapter->pdev->dev, tmp_skb->data,
+ buffer_info->length,
+ DMA_TO_DEVICE);
+ if (dma_mapping_error(&adapter->pdev->dev, buffer_info->dma)) {
+ pr_err("TX DMA map failed\n");
+ buffer_info->dma = 0;
+ buffer_info->time_stamp = 0;
+ tx_ring->next_to_use = ring_num;
+ return;
+ }
+ buffer_info->mapped = true;
+ buffer_info->time_stamp = jiffies;
+
+ /*-- Set Tx descriptor --*/
+ tx_desc = PCH_GBE_TX_DESC(*tx_ring, ring_num);
+ tx_desc->buffer_addr = (buffer_info->dma);
+ tx_desc->length = (tmp_skb->len);
+ tx_desc->tx_words_eob = ((tmp_skb->len + 3));
+ tx_desc->tx_frame_ctrl = (frame_ctrl);
+ tx_desc->gbec_status = (DSC_INIT16);
+
+ if (unlikely(++ring_num == tx_ring->count))
+ ring_num = 0;
+
+ /* Update software pointer of TX descriptor */
+ iowrite32(tx_ring->dma +
+ (int)sizeof(struct pch_gbe_tx_desc) * ring_num,
+ &hw->reg->TX_DSC_SW_P);
+ dev_kfree_skb_any(skb);
+}
+
+/**
+ * pch_gbe_update_stats - Update the board statistics counters
+ * @adapter: Board private structure
+ */
+void pch_gbe_update_stats(struct pch_gbe_adapter *adapter)
+{
+ struct net_device *netdev = adapter->netdev;
+ struct pci_dev *pdev = adapter->pdev;
+ struct pch_gbe_hw_stats *stats = &adapter->stats;
+ unsigned long flags;
+
+ /*
+ * Prevent stats update while adapter is being reset, or if the pci
+ * connection is down.
+ */
+ if ((pdev->error_state) && (pdev->error_state != pci_channel_io_normal))
+ return;
+
+ spin_lock_irqsave(&adapter->stats_lock, flags);
+
+ /* Update device status "adapter->stats" */
+ stats->rx_errors = stats->rx_crc_errors + stats->rx_frame_errors;
+ stats->tx_errors = stats->tx_length_errors +
+ stats->tx_aborted_errors +
+ stats->tx_carrier_errors + stats->tx_timeout_count;
+
+ /* Update network device status "adapter->net_stats" */
+ netdev->stats.rx_packets = stats->rx_packets;
+ netdev->stats.rx_bytes = stats->rx_bytes;
+ netdev->stats.rx_dropped = stats->rx_dropped;
+ netdev->stats.tx_packets = stats->tx_packets;
+ netdev->stats.tx_bytes = stats->tx_bytes;
+ netdev->stats.tx_dropped = stats->tx_dropped;
+ /* Fill out the OS statistics structure */
+ netdev->stats.multicast = stats->multicast;
+ netdev->stats.collisions = stats->collisions;
+ /* Rx Errors */
+ netdev->stats.rx_errors = stats->rx_errors;
+ netdev->stats.rx_crc_errors = stats->rx_crc_errors;
+ netdev->stats.rx_frame_errors = stats->rx_frame_errors;
+ /* Tx Errors */
+ netdev->stats.tx_errors = stats->tx_errors;
+ netdev->stats.tx_aborted_errors = stats->tx_aborted_errors;
+ netdev->stats.tx_carrier_errors = stats->tx_carrier_errors;
+
+ spin_unlock_irqrestore(&adapter->stats_lock, flags);
+}
+
+/**
+ * pch_gbe_intr - Interrupt Handler
+ * @irq: Interrupt number
+ * @data: Pointer to a network interface device structure
+ * Returns
+ * - IRQ_HANDLED: Our interrupt
+ * - IRQ_NONE: Not our interrupt
+ */
+static irqreturn_t pch_gbe_intr(int irq, void *data)
+{
+ struct net_device *netdev = data;
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+ struct pch_gbe_hw *hw = &adapter->hw;
+ u32 int_st;
+ u32 int_en;
+
+ /* Check request status */
+ int_st = ioread32(&hw->reg->INT_ST);
+ int_st = int_st & ioread32(&hw->reg->INT_EN);
+ /* When request status is no interruption factor */
+ if (unlikely(!int_st))
+ return IRQ_NONE; /* Not our interrupt. End processing. */
+ pr_debug("%s occur int_st = 0x%08x\n", __func__, int_st);
+ if (int_st & PCH_GBE_INT_RX_FRAME_ERR)
+ adapter->stats.intr_rx_frame_err_count++;
+ if (int_st & PCH_GBE_INT_RX_FIFO_ERR)
+ adapter->stats.intr_rx_fifo_err_count++;
+ if (int_st & PCH_GBE_INT_RX_DMA_ERR)
+ adapter->stats.intr_rx_dma_err_count++;
+ if (int_st & PCH_GBE_INT_TX_FIFO_ERR)
+ adapter->stats.intr_tx_fifo_err_count++;
+ if (int_st & PCH_GBE_INT_TX_DMA_ERR)
+ adapter->stats.intr_tx_dma_err_count++;
+ if (int_st & PCH_GBE_INT_TCPIP_ERR)
+ adapter->stats.intr_tcpip_err_count++;
+ /* When Rx descriptor is empty */
+ if ((int_st & PCH_GBE_INT_RX_DSC_EMP)) {
+ adapter->stats.intr_rx_dsc_empty_count++;
+ pr_err("Rx descriptor is empty\n");
+ int_en = ioread32(&hw->reg->INT_EN);
+ iowrite32((int_en & ~PCH_GBE_INT_RX_DSC_EMP), &hw->reg->INT_EN);
+ if (hw->mac.tx_fc_enable) {
+ /* Set Pause packet */
+ pch_gbe_mac_set_pause_packet(hw);
+ }
+ if ((int_en & (PCH_GBE_INT_RX_DMA_CMPLT | PCH_GBE_INT_TX_CMPLT))
+ == 0) {
+ return IRQ_HANDLED;
+ }
+ }
+
+ /* When request status is Receive interruption */
+ if ((int_st & (PCH_GBE_INT_RX_DMA_CMPLT | PCH_GBE_INT_TX_CMPLT))) {
+ if (likely(napi_schedule_prep(&adapter->napi))) {
+ /* Enable only Rx Descriptor empty */
+ atomic_inc(&adapter->irq_sem);
+ int_en = ioread32(&hw->reg->INT_EN);
+ int_en &=
+ ~(PCH_GBE_INT_RX_DMA_CMPLT | PCH_GBE_INT_TX_CMPLT);
+ iowrite32(int_en, &hw->reg->INT_EN);
+ /* Start polling for NAPI */
+ __napi_schedule(&adapter->napi);
+ }
+ }
+ pr_debug("return = 0x%08x INT_EN reg = 0x%08x\n",
+ IRQ_HANDLED, ioread32(&hw->reg->INT_EN));
+ return IRQ_HANDLED;
+}
+
+/**
+ * pch_gbe_alloc_rx_buffers - Replace used receive buffers; legacy & extended
+ * @adapter: Board private structure
+ * @rx_ring: Rx descriptor ring
+ * @cleaned_count: Cleaned count
+ */
+static void
+pch_gbe_alloc_rx_buffers(struct pch_gbe_adapter *adapter,
+ struct pch_gbe_rx_ring *rx_ring, int cleaned_count)
+{
+ struct net_device *netdev = adapter->netdev;
+ struct pci_dev *pdev = adapter->pdev;
+ struct pch_gbe_hw *hw = &adapter->hw;
+ struct pch_gbe_rx_desc *rx_desc;
+ struct pch_gbe_buffer *buffer_info;
+ struct sk_buff *skb;
+ unsigned int i;
+ unsigned int bufsz;
+
+ bufsz = adapter->rx_buffer_len + PCH_GBE_DMA_ALIGN;
+ i = rx_ring->next_to_use;
+
+ while ((cleaned_count--)) {
+ buffer_info = &rx_ring->buffer_info[i];
+ skb = buffer_info->skb;
+ if (skb) {
+ skb_trim(skb, 0);
+ } else {
+ skb = netdev_alloc_skb(netdev, bufsz);
+ if (unlikely(!skb)) {
+ /* Better luck next round */
+ adapter->stats.rx_alloc_buff_failed++;
+ break;
+ }
+ /* 64byte align */
+ skb_reserve(skb, PCH_GBE_DMA_ALIGN);
+
+ buffer_info->skb = skb;
+ buffer_info->length = adapter->rx_buffer_len;
+ }
+ buffer_info->dma = dma_map_single(&pdev->dev,
+ skb->data,
+ buffer_info->length,
+ DMA_FROM_DEVICE);
+ if (dma_mapping_error(&adapter->pdev->dev, buffer_info->dma)) {
+ dev_kfree_skb(skb);
+ buffer_info->skb = NULL;
+ buffer_info->dma = 0;
+ adapter->stats.rx_alloc_buff_failed++;
+ break; /* while !buffer_info->skb */
+ }
+ buffer_info->mapped = true;
+ rx_desc = PCH_GBE_RX_DESC(*rx_ring, i);
+ rx_desc->buffer_addr = (buffer_info->dma);
+ rx_desc->gbec_status = DSC_INIT16;
+
+ pr_debug("i = %d buffer_info->dma = 0x08%llx buffer_info->length = 0x%x\n",
+ i, (unsigned long long)buffer_info->dma,
+ buffer_info->length);
+
+ if (unlikely(++i == rx_ring->count))
+ i = 0;
+ }
+ if (likely(rx_ring->next_to_use != i)) {
+ rx_ring->next_to_use = i;
+ if (unlikely(i-- == 0))
+ i = (rx_ring->count - 1);
+ iowrite32(rx_ring->dma +
+ (int)sizeof(struct pch_gbe_rx_desc) * i,
+ &hw->reg->RX_DSC_SW_P);
+ }
+ return;
+}
+
+/**
+ * pch_gbe_alloc_tx_buffers - Allocate transmit buffers
+ * @adapter: Board private structure
+ * @tx_ring: Tx descriptor ring
+ */
+static void pch_gbe_alloc_tx_buffers(struct pch_gbe_adapter *adapter,
+ struct pch_gbe_tx_ring *tx_ring)
+{
+ struct pch_gbe_buffer *buffer_info;
+ struct sk_buff *skb;
+ unsigned int i;
+ unsigned int bufsz;
+ struct pch_gbe_tx_desc *tx_desc;
+
+ bufsz =
+ adapter->hw.mac.max_frame_size + PCH_GBE_DMA_ALIGN + NET_IP_ALIGN;
+
+ for (i = 0; i < tx_ring->count; i++) {
+ buffer_info = &tx_ring->buffer_info[i];
+ skb = netdev_alloc_skb(adapter->netdev, bufsz);
+ skb_reserve(skb, PCH_GBE_DMA_ALIGN);
+ buffer_info->skb = skb;
+ tx_desc = PCH_GBE_TX_DESC(*tx_ring, i);
+ tx_desc->gbec_status = (DSC_INIT16);
+ }
+ return;
+}
+
+/**
+ * pch_gbe_clean_tx - Reclaim resources after transmit completes
+ * @adapter: Board private structure
+ * @tx_ring: Tx descriptor ring
+ * Returns
+ * true: Cleaned the descriptor
+ * false: Not cleaned the descriptor
+ */
+static bool
+pch_gbe_clean_tx(struct pch_gbe_adapter *adapter,
+ struct pch_gbe_tx_ring *tx_ring)
+{
+ struct pch_gbe_tx_desc *tx_desc;
+ struct pch_gbe_buffer *buffer_info;
+ struct sk_buff *skb;
+ unsigned int i;
+ unsigned int cleaned_count = 0;
+ bool cleaned = false;
+
+ pr_debug("next_to_clean : %d\n", tx_ring->next_to_clean);
+
+ i = tx_ring->next_to_clean;
+ tx_desc = PCH_GBE_TX_DESC(*tx_ring, i);
+ pr_debug("gbec_status:0x%04x dma_status:0x%04x\n",
+ tx_desc->gbec_status, tx_desc->dma_status);
+
+ while ((tx_desc->gbec_status & DSC_INIT16) == 0x0000) {
+ pr_debug("gbec_status:0x%04x\n", tx_desc->gbec_status);
+ cleaned = true;
+ buffer_info = &tx_ring->buffer_info[i];
+ skb = buffer_info->skb;
+
+ if ((tx_desc->gbec_status & PCH_GBE_TXD_GMAC_STAT_ABT)) {
+ adapter->stats.tx_aborted_errors++;
+ pr_err("Transfer Abort Error\n");
+ } else if ((tx_desc->gbec_status & PCH_GBE_TXD_GMAC_STAT_CRSER)
+ ) {
+ adapter->stats.tx_carrier_errors++;
+ pr_err("Transfer Carrier Sense Error\n");
+ } else if ((tx_desc->gbec_status & PCH_GBE_TXD_GMAC_STAT_EXCOL)
+ ) {
+ adapter->stats.tx_aborted_errors++;
+ pr_err("Transfer Collision Abort Error\n");
+ } else if ((tx_desc->gbec_status &
+ (PCH_GBE_TXD_GMAC_STAT_SNGCOL |
+ PCH_GBE_TXD_GMAC_STAT_MLTCOL))) {
+ adapter->stats.collisions++;
+ adapter->stats.tx_packets++;
+ adapter->stats.tx_bytes += skb->len;
+ pr_debug("Transfer Collision\n");
+ } else if ((tx_desc->gbec_status & PCH_GBE_TXD_GMAC_STAT_CMPLT)
+ ) {
+ adapter->stats.tx_packets++;
+ adapter->stats.tx_bytes += skb->len;
+ }
+ if (buffer_info->mapped) {
+ pr_debug("unmap buffer_info->dma : %d\n", i);
+ dma_unmap_single(&adapter->pdev->dev, buffer_info->dma,
+ buffer_info->length, DMA_TO_DEVICE);
+ buffer_info->mapped = false;
+ }
+ if (buffer_info->skb) {
+ pr_debug("trim buffer_info->skb : %d\n", i);
+ skb_trim(buffer_info->skb, 0);
+ }
+ tx_desc->gbec_status = DSC_INIT16;
+ if (unlikely(++i == tx_ring->count))
+ i = 0;
+ tx_desc = PCH_GBE_TX_DESC(*tx_ring, i);
+
+ /* weight of a sort for tx, to avoid endless transmit cleanup */
+ if (cleaned_count++ == PCH_GBE_TX_WEIGHT)
+ break;
+ }
+ pr_debug("called pch_gbe_unmap_and_free_tx_resource() %d count\n",
+ cleaned_count);
+ /* Recover from running out of Tx resources in xmit_frame */
+ if (unlikely(cleaned && (netif_queue_stopped(adapter->netdev)))) {
+ netif_wake_queue(adapter->netdev);
+ adapter->stats.tx_restart_count++;
+ pr_debug("Tx wake queue\n");
+ }
+ spin_lock(&adapter->tx_queue_lock);
+ tx_ring->next_to_clean = i;
+ spin_unlock(&adapter->tx_queue_lock);
+ pr_debug("next_to_clean : %d\n", tx_ring->next_to_clean);
+ return cleaned;
+}
+
+/**
+ * pch_gbe_clean_rx - Send received data up the network stack; legacy
+ * @adapter: Board private structure
+ * @rx_ring: Rx descriptor ring
+ * @work_done: Completed count
+ * @work_to_do: Request count
+ * Returns
+ * true: Cleaned the descriptor
+ * false: Not cleaned the descriptor
+ */
+static bool
+pch_gbe_clean_rx(struct pch_gbe_adapter *adapter,
+ struct pch_gbe_rx_ring *rx_ring,
+ int *work_done, int work_to_do)
+{
+ struct net_device *netdev = adapter->netdev;
+ struct pci_dev *pdev = adapter->pdev;
+ struct pch_gbe_buffer *buffer_info;
+ struct pch_gbe_rx_desc *rx_desc;
+ u32 length;
+ unsigned char tmp_packet[ETH_HLEN];
+ unsigned int i;
+ unsigned int cleaned_count = 0;
+ bool cleaned = false;
+ struct sk_buff *skb;
+ u8 dma_status;
+ u16 gbec_status;
+ u32 tcp_ip_status;
+ u8 skb_copy_flag = 0;
+ u8 skb_padding_flag = 0;
+
+ i = rx_ring->next_to_clean;
+
+ while (*work_done < work_to_do) {
+ /* Check Rx descriptor status */
+ rx_desc = PCH_GBE_RX_DESC(*rx_ring, i);
+ if (rx_desc->gbec_status == DSC_INIT16)
+ break;
+ cleaned = true;
+ cleaned_count++;
+
+ dma_status = rx_desc->dma_status;
+ gbec_status = rx_desc->gbec_status;
+ tcp_ip_status = rx_desc->tcp_ip_status;
+ rx_desc->gbec_status = DSC_INIT16;
+ buffer_info = &rx_ring->buffer_info[i];
+ skb = buffer_info->skb;
+
+ /* unmap dma */
+ dma_unmap_single(&pdev->dev, buffer_info->dma,
+ buffer_info->length, DMA_FROM_DEVICE);
+ buffer_info->mapped = false;
+ /* Prefetch the packet */
+ prefetch(skb->data);
+
+ pr_debug("RxDecNo = 0x%04x Status[DMA:0x%02x GBE:0x%04x "
+ "TCP:0x%08x] BufInf = 0x%08x\n",
+ i, dma_status, gbec_status, tcp_ip_status,
+ (u32) (buffer_info));
+ /* Error check */
+ if (unlikely(gbec_status & PCH_GBE_RXD_GMAC_STAT_NOTOCTAL)) {
+ adapter->stats.rx_frame_errors++;
+ pr_err("Receive Not Octal Error\n");
+ } else if (unlikely(gbec_status &
+ PCH_GBE_RXD_GMAC_STAT_NBLERR)) {
+ adapter->stats.rx_frame_errors++;
+ pr_err("Receive Nibble Error\n");
+ } else if (unlikely(gbec_status &
+ PCH_GBE_RXD_GMAC_STAT_CRCERR)) {
+ adapter->stats.rx_crc_errors++;
+ pr_err("Receive CRC Error\n");
+ } else {
+ /* get receive length */
+ /* length convert[-3], padding[-2] */
+ length = (rx_desc->rx_words_eob) - 3 - 2;
+
+ /* Decide the data conversion method */
+ if (!adapter->rx_csum) {
+ /* [Header:14][payload] */
+ skb_padding_flag = 0;
+ skb_copy_flag = 1;
+ } else {
+ /* [Header:14][padding:2][payload] */
+ skb_padding_flag = 1;
+ if (length < copybreak)
+ skb_copy_flag = 1;
+ else
+ skb_copy_flag = 0;
+ }
+
+ /* Data conversion */
+ if (skb_copy_flag) { /* recycle skb */
+ struct sk_buff *new_skb;
+ new_skb =
+ netdev_alloc_skb(netdev,
+ length + NET_IP_ALIGN);
+ if (new_skb) {
+ if (!skb_padding_flag) {
+ skb_reserve(new_skb,
+ NET_IP_ALIGN);
+ }
+ memcpy(new_skb->data, skb->data,
+ length);
+ /* save the skb
+ * in buffer_info as good */
+ skb = new_skb;
+ } else if (!skb_padding_flag) {
+ /* dorrop error */
+ pr_err("New skb allocation Error\n");
+ goto dorrop;
+ }
+ } else {
+ buffer_info->skb = NULL;
+ }
+ if (skb_padding_flag) {
+ memcpy(&tmp_packet[0], &skb->data[0], ETH_HLEN);
+ memcpy(&skb->data[NET_IP_ALIGN], &tmp_packet[0],
+ ETH_HLEN);
+ skb_reserve(skb, NET_IP_ALIGN);
+
+ }
+
+ /* update status of driver */
+ adapter->stats.rx_bytes += length;
+ adapter->stats.rx_packets++;
+ if ((gbec_status & PCH_GBE_RXD_GMAC_STAT_MARMLT))
+ adapter->stats.multicast++;
+ /* Write meta date of skb */
+ skb_put(skb, length);
+ skb->protocol = eth_type_trans(skb, netdev);
+ if ((tcp_ip_status & PCH_GBE_RXD_ACC_STAT_TCPIPOK) ==
+ PCH_GBE_RXD_ACC_STAT_TCPIPOK) {
+ skb->ip_summed = CHECKSUM_UNNECESSARY;
+ } else {
+ skb->ip_summed = CHECKSUM_NONE;
+ }
+ napi_gro_receive(&adapter->napi, skb);
+ (*work_done)++;
+ pr_debug("Receive skb->ip_summed: %d length: %d\n",
+ skb->ip_summed, length);
+ }
+dorrop:
+ /* return some buffers to hardware, one at a time is too slow */
+ if (unlikely(cleaned_count >= PCH_GBE_RX_BUFFER_WRITE)) {
+ pch_gbe_alloc_rx_buffers(adapter, rx_ring,
+ cleaned_count);
+ cleaned_count = 0;
+ }
+ if (++i == rx_ring->count)
+ i = 0;
+ }
+ rx_ring->next_to_clean = i;
+ if (cleaned_count)
+ pch_gbe_alloc_rx_buffers(adapter, rx_ring, cleaned_count);
+ return cleaned;
+}
+
+/**
+ * pch_gbe_setup_tx_resources - Allocate Tx resources (Descriptors)
+ * @adapter: Board private structure
+ * @tx_ring: Tx descriptor ring (for a specific queue) to setup
+ * Returns
+ * 0: Successfully
+ * Negative value: Failed
+ */
+int pch_gbe_setup_tx_resources(struct pch_gbe_adapter *adapter,
+ struct pch_gbe_tx_ring *tx_ring)
+{
+ struct pci_dev *pdev = adapter->pdev;
+ struct pch_gbe_tx_desc *tx_desc;
+ int size;
+ int desNo;
+
+ size = (int)sizeof(struct pch_gbe_buffer) * tx_ring->count;
+ tx_ring->buffer_info = vmalloc(size);
+ if (!tx_ring->buffer_info) {
+ pr_err("Unable to allocate memory for the buffer infomation\n");
+ return -ENOMEM;
+ }
+ memset(tx_ring->buffer_info, 0, size);
+
+ tx_ring->size = tx_ring->count * (int)sizeof(struct pch_gbe_tx_desc);
+
+ tx_ring->desc = dma_alloc_coherent(&pdev->dev, tx_ring->size,
+ &tx_ring->dma, GFP_KERNEL);
+ if (!tx_ring->desc) {
+ vfree(tx_ring->buffer_info);
+ pr_err("Unable to allocate memory for the transmit descriptor ring\n");
+ return -ENOMEM;
+ }
+ memset(tx_ring->desc, 0, tx_ring->size);
+
+ tx_ring->next_to_use = 0;
+ tx_ring->next_to_clean = 0;
+ spin_lock_init(&tx_ring->tx_lock);
+
+ for (desNo = 0; desNo < tx_ring->count; desNo++) {
+ tx_desc = PCH_GBE_TX_DESC(*tx_ring, desNo);
+ tx_desc->gbec_status = DSC_INIT16;
+ }
+ pr_debug("tx_ring->desc = 0x%08x tx_ring->dma = 0x%08llx\n"
+ "next_to_clean = 0x%08x next_to_use = 0x%08x\n",
+ (u32) tx_ring->desc, (unsigned long long)tx_ring->dma,
+ tx_ring->next_to_clean, tx_ring->next_to_use);
+ return 0;
+}
+
+/**
+ * pch_gbe_setup_rx_resources - Allocate Rx resources (Descriptors)
+ * @adapter: Board private structure
+ * @rx_ring: Rx descriptor ring (for a specific queue) to setup
+ * Returns
+ * 0: Successfully
+ * Negative value: Failed
+ */
+int pch_gbe_setup_rx_resources(struct pch_gbe_adapter *adapter,
+ struct pch_gbe_rx_ring *rx_ring)
+{
+ struct pci_dev *pdev = adapter->pdev;
+ struct pch_gbe_rx_desc *rx_desc;
+ int size;
+ int desNo;
+
+ size = (int)sizeof(struct pch_gbe_buffer) * rx_ring->count;
+ rx_ring->buffer_info = vmalloc(size);
+ if (!rx_ring->buffer_info) {
+ pr_err("Unable to allocate memory for the receive descriptor ring\n");
+ return -ENOMEM;
+ }
+ memset(rx_ring->buffer_info, 0, size);
+ rx_ring->size = rx_ring->count * (int)sizeof(struct pch_gbe_rx_desc);
+ rx_ring->desc = dma_alloc_coherent(&pdev->dev, rx_ring->size,
+ &rx_ring->dma, GFP_KERNEL);
+
+ if (!rx_ring->desc) {
+ pr_err("Unable to allocate memory for the receive descriptor ring\n");
+ vfree(rx_ring->buffer_info);
+ return -ENOMEM;
+ }
+ memset(rx_ring->desc, 0, rx_ring->size);
+ rx_ring->next_to_clean = 0;
+ rx_ring->next_to_use = 0;
+ for (desNo = 0; desNo < rx_ring->count; desNo++) {
+ rx_desc = PCH_GBE_RX_DESC(*rx_ring, desNo);
+ rx_desc->gbec_status = DSC_INIT16;
+ }
+ pr_debug("rx_ring->desc = 0x%08x rx_ring->dma = 0x%08llx "
+ "next_to_clean = 0x%08x next_to_use = 0x%08x\n",
+ (u32) rx_ring->desc, (unsigned long long)rx_ring->dma,
+ rx_ring->next_to_clean, rx_ring->next_to_use);
+ return 0;
+}
+
+/**
+ * pch_gbe_free_tx_resources - Free Tx Resources
+ * @adapter: Board private structure
+ * @tx_ring: Tx descriptor ring for a specific queue
+ */
+void pch_gbe_free_tx_resources(struct pch_gbe_adapter *adapter,
+ struct pch_gbe_tx_ring *tx_ring)
+{
+ struct pci_dev *pdev = adapter->pdev;
+
+ pch_gbe_clean_tx_ring(adapter, tx_ring);
+ vfree(tx_ring->buffer_info);
+ tx_ring->buffer_info = NULL;
+ pci_free_consistent(pdev, tx_ring->size, tx_ring->desc, tx_ring->dma);
+ tx_ring->desc = NULL;
+}
+
+/**
+ * pch_gbe_free_rx_resources - Free Rx Resources
+ * @adapter: Board private structure
+ * @rx_ring: Ring to clean the resources from
+ */
+void pch_gbe_free_rx_resources(struct pch_gbe_adapter *adapter,
+ struct pch_gbe_rx_ring *rx_ring)
+{
+ struct pci_dev *pdev = adapter->pdev;
+
+ pch_gbe_clean_rx_ring(adapter, rx_ring);
+ vfree(rx_ring->buffer_info);
+ rx_ring->buffer_info = NULL;
+ pci_free_consistent(pdev, rx_ring->size, rx_ring->desc, rx_ring->dma);
+ rx_ring->desc = NULL;
+}
+
+/**
+ * pch_gbe_request_irq - Allocate an interrupt line
+ * @adapter: Board private structure
+ * Returns
+ * 0: Successfully
+ * Negative value: Failed
+ */
+static int pch_gbe_request_irq(struct pch_gbe_adapter *adapter)
+{
+ struct net_device *netdev = adapter->netdev;
+ int err;
+ int flags;
+
+ flags = IRQF_SHARED;
+ adapter->have_msi = false;
+ err = pci_enable_msi(adapter->pdev);
+ pr_debug("call pci_enable_msi\n");
+ if (err) {
+ pr_debug("call pci_enable_msi - Error: %d\n", err);
+ } else {
+ flags = 0;
+ adapter->have_msi = true;
+ }
+ err = request_irq(adapter->pdev->irq, &pch_gbe_intr,
+ flags, netdev->name, netdev);
+ if (err)
+ pr_err("Unable to allocate interrupt Error: %d\n", err);
+ pr_debug("adapter->have_msi : %d flags : 0x%04x return : 0x%04x\n",
+ adapter->have_msi, flags, err);
+ return err;
+}
+
+
+static void pch_gbe_set_multi(struct net_device *netdev);
+/**
+ * pch_gbe_up - Up GbE network device
+ * @adapter: Board private structure
+ * Returns
+ * 0: Successfully
+ * Negative value: Failed
+ */
+int pch_gbe_up(struct pch_gbe_adapter *adapter)
+{
+ struct net_device *netdev = adapter->netdev;
+ struct pch_gbe_tx_ring *tx_ring = adapter->tx_ring;
+ struct pch_gbe_rx_ring *rx_ring = adapter->rx_ring;
+ int err;
+
+ /* hardware has been reset, we need to reload some things */
+ pch_gbe_set_multi(netdev);
+
+ pch_gbe_setup_tctl(adapter);
+ pch_gbe_configure_tx(adapter);
+ pch_gbe_setup_rctl(adapter);
+ pch_gbe_configure_rx(adapter);
+
+ err = pch_gbe_request_irq(adapter);
+ if (err) {
+ pr_err("Error: can't bring device up\n");
+ return err;
+ }
+ pch_gbe_alloc_tx_buffers(adapter, tx_ring);
+ pch_gbe_alloc_rx_buffers(adapter, rx_ring, rx_ring->count);
+ adapter->tx_queue_len = netdev->tx_queue_len;
+
+ mod_timer(&adapter->watchdog_timer, jiffies);
+
+ napi_enable(&adapter->napi);
+ pch_gbe_irq_enable(adapter);
+ netif_start_queue(adapter->netdev);
+
+ return 0;
+}
+
+/**
+ * pch_gbe_down - Down GbE network device
+ * @adapter: Board private structure
+ */
+void pch_gbe_down(struct pch_gbe_adapter *adapter)
+{
+ struct net_device *netdev = adapter->netdev;
+
+ /* signal that we're down so the interrupt handler does not
+ * reschedule our watchdog timer */
+ napi_disable(&adapter->napi);
+ atomic_set(&adapter->irq_sem, 0);
+
+ pch_gbe_irq_disable(adapter);
+ pch_gbe_free_irq(adapter);
+
+ del_timer_sync(&adapter->watchdog_timer);
+
+ netdev->tx_queue_len = adapter->tx_queue_len;
+ netif_carrier_off(netdev);
+ netif_stop_queue(netdev);
+
+ pch_gbe_reset(adapter);
+ pch_gbe_clean_tx_ring(adapter, adapter->tx_ring);
+ pch_gbe_clean_rx_ring(adapter, adapter->rx_ring);
+}
+
+/**
+ * pch_gbe_sw_init - Initialize general software structures (struct pch_gbe_adapter)
+ * @adapter: Board private structure to initialize
+ * Returns
+ * 0: Successfully
+ * Negative value: Failed
+ */
+static int pch_gbe_sw_init(struct pch_gbe_adapter *adapter)
+{
+ struct pch_gbe_hw *hw = &adapter->hw;
+ struct net_device *netdev = adapter->netdev;
+
+ adapter->rx_buffer_len = PCH_GBE_FRAME_SIZE_2048;
+ hw->mac.max_frame_size = netdev->mtu + ETH_HLEN + ETH_FCS_LEN;
+ hw->mac.min_frame_size = ETH_ZLEN + ETH_FCS_LEN;
+
+ /* Initialize the hardware-specific values */
+ if (pch_gbe_hal_setup_init_funcs(hw)) {
+ pr_err("Hardware Initialization Failure\n");
+ return -EIO;
+ }
+ if (pch_gbe_alloc_queues(adapter)) {
+ pr_err("Unable to allocate memory for queues\n");
+ return -ENOMEM;
+ }
+ spin_lock_init(&adapter->hw.miim_lock);
+ spin_lock_init(&adapter->tx_queue_lock);
+ spin_lock_init(&adapter->stats_lock);
+ spin_lock_init(&adapter->ethtool_lock);
+ atomic_set(&adapter->irq_sem, 0);
+ pch_gbe_irq_disable(adapter);
+
+ pch_gbe_init_stats(adapter);
+
+ pr_debug("rx_buffer_len : %d mac.min_frame_size : %d mac.max_frame_size : %d\n",
+ (u32) adapter->rx_buffer_len,
+ hw->mac.min_frame_size, hw->mac.max_frame_size);
+ return 0;
+}
+
+/**
+ * pch_gbe_open - Called when a network interface is made active
+ * @netdev: Network interface device structure
+ * Returns
+ * 0: Successfully
+ * Negative value: Failed
+ */
+static int pch_gbe_open(struct net_device *netdev)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+ struct pch_gbe_hw *hw = &adapter->hw;
+ int err;
+
+ /* allocate transmit descriptors */
+ err = pch_gbe_setup_tx_resources(adapter, adapter->tx_ring);
+ if (err)
+ goto err_setup_tx;
+ /* allocate receive descriptors */
+ err = pch_gbe_setup_rx_resources(adapter, adapter->rx_ring);
+ if (err)
+ goto err_setup_rx;
+ pch_gbe_hal_power_up_phy(hw);
+ err = pch_gbe_up(adapter);
+ if (err)
+ goto err_up;
+ pr_debug("Success End\n");
+ return 0;
+
+err_up:
+ if (!adapter->wake_up_evt)
+ pch_gbe_hal_power_down_phy(hw);
+ pch_gbe_free_rx_resources(adapter, adapter->rx_ring);
+err_setup_rx:
+ pch_gbe_free_tx_resources(adapter, adapter->tx_ring);
+err_setup_tx:
+ pch_gbe_reset(adapter);
+ pr_err("Error End\n");
+ return err;
+}
+
+/**
+ * pch_gbe_stop - Disables a network interface
+ * @netdev: Network interface device structure
+ * Returns
+ * 0: Successfully
+ */
+static int pch_gbe_stop(struct net_device *netdev)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+ struct pch_gbe_hw *hw = &adapter->hw;
+
+ pch_gbe_down(adapter);
+ if (!adapter->wake_up_evt)
+ pch_gbe_hal_power_down_phy(hw);
+ pch_gbe_free_tx_resources(adapter, adapter->tx_ring);
+ pch_gbe_free_rx_resources(adapter, adapter->rx_ring);
+ return 0;
+}
+
+/**
+ * pch_gbe_xmit_frame - Packet transmitting start
+ * @skb: Socket buffer structure
+ * @netdev: Network interface device structure
+ * Returns
+ * - NETDEV_TX_OK: Normal end
+ * - NETDEV_TX_BUSY: Error end
+ */
+static int pch_gbe_xmit_frame(struct sk_buff *skb, struct net_device *netdev)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+ struct pch_gbe_tx_ring *tx_ring = adapter->tx_ring;
+ unsigned long flags;
+
+ if (unlikely(skb->len > (adapter->hw.mac.max_frame_size - 4))) {
+ dev_kfree_skb_any(skb);
+ pr_err("Transfer length Error: skb len: %d > max: %d\n",
+ skb->len, adapter->hw.mac.max_frame_size);
+ adapter->stats.tx_length_errors++;
+ return NETDEV_TX_OK;
+ }
+ if (!spin_trylock_irqsave(&tx_ring->tx_lock, flags)) {
+ /* Collision - tell upper layer to requeue */
+ return NETDEV_TX_LOCKED;
+ }
+ if (unlikely(!PCH_GBE_DESC_UNUSED(tx_ring))) {
+ netif_stop_queue(netdev);
+ spin_unlock_irqrestore(&tx_ring->tx_lock, flags);
+ pr_debug("Return : BUSY next_to use : 0x%08x next_to clean : 0x%08x\n",
+ tx_ring->next_to_use, tx_ring->next_to_clean);
+ return NETDEV_TX_BUSY;
+ }
+ spin_unlock_irqrestore(&tx_ring->tx_lock, flags);
+
+ /* CRC,ITAG no support */
+ pch_gbe_tx_queue(adapter, tx_ring, skb);
+ return NETDEV_TX_OK;
+}
+
+/**
+ * pch_gbe_get_stats - Get System Network Statistics
+ * @netdev: Network interface device structure
+ * Returns: The current stats
+ */
+static struct net_device_stats *pch_gbe_get_stats(struct net_device *netdev)
+{
+ /* only return the current stats */
+ return &netdev->stats;
+}
+
+/**
+ * pch_gbe_set_multi - Multicast and Promiscuous mode set
+ * @netdev: Network interface device structure
+ */
+static void pch_gbe_set_multi(struct net_device *netdev)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+ struct pch_gbe_hw *hw = &adapter->hw;
+ struct netdev_hw_addr *ha;
+ u8 *mta_list;
+ u32 rctl;
+ int i;
+ int mc_count;
+
+ pr_debug("netdev->flags : 0x%08x\n", netdev->flags);
+
+ /* Check for Promiscuous and All Multicast modes */
+ rctl = ioread32(&hw->reg->RX_MODE);
+ mc_count = netdev_mc_count(netdev);
+ if ((netdev->flags & IFF_PROMISC)) {
+ rctl &= ~PCH_GBE_ADD_FIL_EN;
+ rctl &= ~PCH_GBE_MLT_FIL_EN;
+ } else if ((netdev->flags & IFF_ALLMULTI)) {
+ /* all the multicasting receive permissions */
+ rctl |= PCH_GBE_ADD_FIL_EN;
+ rctl &= ~PCH_GBE_MLT_FIL_EN;
+ } else {
+ if (mc_count >= PCH_GBE_MAR_ENTRIES) {
+ /* all the multicasting receive permissions */
+ rctl |= PCH_GBE_ADD_FIL_EN;
+ rctl &= ~PCH_GBE_MLT_FIL_EN;
+ } else {
+ rctl |= (PCH_GBE_ADD_FIL_EN | PCH_GBE_MLT_FIL_EN);
+ }
+ }
+ iowrite32(rctl, &hw->reg->RX_MODE);
+
+ if (mc_count >= PCH_GBE_MAR_ENTRIES)
+ return;
+ mta_list = kmalloc(mc_count * ETH_ALEN, GFP_ATOMIC);
+ if (!mta_list)
+ return;
+
+ /* The shared function expects a packed array of only addresses. */
+ i = 0;
+ netdev_for_each_mc_addr(ha, netdev) {
+ if (i == mc_count)
+ break;
+ memcpy(mta_list + (i++ * ETH_ALEN), &ha->addr, ETH_ALEN);
+ }
+ pch_gbe_mac_mc_addr_list_update(hw, mta_list, i, 1,
+ PCH_GBE_MAR_ENTRIES);
+ kfree(mta_list);
+
+ pr_debug("RX_MODE reg(check bit31,30 ADD,MLT) : 0x%08x netdev->mc_count : 0x%08x\n",
+ ioread32(&hw->reg->RX_MODE), mc_count);
+}
+
+/**
+ * pch_gbe_set_mac - Change the Ethernet Address of the NIC
+ * @netdev: Network interface device structure
+ * @addr: Pointer to an address structure
+ * Returns
+ * 0: Successfully
+ * -EADDRNOTAVAIL: Failed
+ */
+static int pch_gbe_set_mac(struct net_device *netdev, void *addr)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+ struct sockaddr *skaddr = addr;
+ int ret_val;
+
+ if (!is_valid_ether_addr(skaddr->sa_data)) {
+ ret_val = -EADDRNOTAVAIL;
+ } else {
+ memcpy(netdev->dev_addr, skaddr->sa_data, netdev->addr_len);
+ memcpy(adapter->hw.mac.addr, skaddr->sa_data, netdev->addr_len);
+ pch_gbe_mac_mar_set(&adapter->hw, adapter->hw.mac.addr, 0);
+ ret_val = 0;
+ }
+ pr_debug("ret_val : 0x%08x\n", ret_val);
+ pr_debug("dev_addr : %pM\n", netdev->dev_addr);
+ pr_debug("mac_addr : %pM\n", adapter->hw.mac.addr);
+ pr_debug("MAC_ADR1AB reg : 0x%08x 0x%08x\n",
+ ioread32(&adapter->hw.reg->mac_adr[0].high),
+ ioread32(&adapter->hw.reg->mac_adr[0].low));
+ return ret_val;
+}
+
+/**
+ * pch_gbe_change_mtu - Change the Maximum Transfer Unit
+ * @netdev: Network interface device structure
+ * @new_mtu: New value for maximum frame size
+ * Returns
+ * 0: Successfully
+ * -EINVAL: Failed
+ */
+static int pch_gbe_change_mtu(struct net_device *netdev, int new_mtu)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+ int max_frame;
+
+ max_frame = new_mtu + ETH_HLEN + ETH_FCS_LEN;
+ if ((max_frame < ETH_ZLEN + ETH_FCS_LEN) ||
+ (max_frame > PCH_GBE_MAX_JUMBO_FRAME_SIZE)) {
+ pr_err("Invalid MTU setting\n");
+ return -EINVAL;
+ }
+ if (max_frame <= PCH_GBE_FRAME_SIZE_2048)
+ adapter->rx_buffer_len = PCH_GBE_FRAME_SIZE_2048;
+ else if (max_frame <= PCH_GBE_FRAME_SIZE_4096)
+ adapter->rx_buffer_len = PCH_GBE_FRAME_SIZE_4096;
+ else if (max_frame <= PCH_GBE_FRAME_SIZE_8192)
+ adapter->rx_buffer_len = PCH_GBE_FRAME_SIZE_8192;
+ else
+ adapter->rx_buffer_len = PCH_GBE_MAX_JUMBO_FRAME_SIZE;
+ netdev->mtu = new_mtu;
+ adapter->hw.mac.max_frame_size = max_frame;
+
+ if (netif_running(netdev))
+ pch_gbe_reinit_locked(adapter);
+ else
+ pch_gbe_reset(adapter);
+
+ pr_debug("max_frame : %d rx_buffer_len : %d mtu : %d max_frame_size : %d\n",
+ max_frame, (u32) adapter->rx_buffer_len, netdev->mtu,
+ adapter->hw.mac.max_frame_size);
+ return 0;
+}
+
+/**
+ * pch_gbe_ioctl - Controls register through a MII interface
+ * @netdev: Network interface device structure
+ * @ifr: Pointer to ifr structure
+ * @cmd: Control command
+ * Returns
+ * 0: Successfully
+ * Negative value: Failed
+ */
+static int pch_gbe_ioctl(struct net_device *netdev, struct ifreq *ifr, int cmd)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+
+ pr_debug("cmd : 0x%04x\n", cmd);
+
+ return generic_mii_ioctl(&adapter->mii, if_mii(ifr), cmd, NULL);
+}
+
+/**
+ * pch_gbe_tx_timeout - Respond to a Tx Hang
+ * @netdev: Network interface device structure
+ */
+static void pch_gbe_tx_timeout(struct net_device *netdev)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+
+ /* Do the reset outside of interrupt context */
+ adapter->stats.tx_timeout_count++;
+ schedule_work(&adapter->reset_task);
+}
+
+/**
+ * pch_gbe_napi_poll - NAPI receive and transfer polling callback
+ * @napi: Pointer of polling device struct
+ * @budget: The maximum number of a packet
+ * Returns
+ * false: Exit the polling mode
+ * true: Continue the polling mode
+ */
+static int pch_gbe_napi_poll(struct napi_struct *napi, int budget)
+{
+ struct pch_gbe_adapter *adapter =
+ container_of(napi, struct pch_gbe_adapter, napi);
+ struct net_device *netdev = adapter->netdev;
+ int work_done = 0;
+ bool poll_end_flag = false;
+ bool cleaned = false;
+
+ pr_debug("budget : %d\n", budget);
+
+ /* Keep link state information with original netdev */
+ if (!netif_carrier_ok(netdev)) {
+ poll_end_flag = true;
+ } else {
+ cleaned = pch_gbe_clean_tx(adapter, adapter->tx_ring);
+ pch_gbe_clean_rx(adapter, adapter->rx_ring, &work_done, budget);
+
+ if (cleaned)
+ work_done = budget;
+ /* If no Tx and not enough Rx work done,
+ * exit the polling mode
+ */
+ if ((work_done < budget) || !netif_running(netdev))
+ poll_end_flag = true;
+ }
+
+ if (poll_end_flag) {
+ napi_complete(napi);
+ pch_gbe_irq_enable(adapter);
+ }
+
+ pr_debug("poll_end_flag : %d work_done : %d budget : %d\n",
+ poll_end_flag, work_done, budget);
+
+ return work_done;
+}
+
+#ifdef CONFIG_NET_POLL_CONTROLLER
+/**
+ * pch_gbe_netpoll - Used by things like netconsole to send skbs
+ * @netdev: Network interface device structure
+ */
+static void pch_gbe_netpoll(struct net_device *netdev)
+{
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+
+ disable_irq(adapter->pdev->irq);
+ pch_gbe_intr(adapter->pdev->irq, netdev);
+ enable_irq(adapter->pdev->irq);
+}
+#else
+#define pch_gbe_netpoll NULL
+#endif
+
+static const struct net_device_ops pch_gbe_netdev_ops = {
+ .ndo_open = pch_gbe_open,
+ .ndo_stop = pch_gbe_stop,
+ .ndo_start_xmit = pch_gbe_xmit_frame,
+ .ndo_get_stats = pch_gbe_get_stats,
+ .ndo_set_mac_address = pch_gbe_set_mac,
+ .ndo_tx_timeout = pch_gbe_tx_timeout,
+ .ndo_change_mtu = pch_gbe_change_mtu,
+ .ndo_do_ioctl = pch_gbe_ioctl,
+ .ndo_set_multicast_list = &pch_gbe_set_multi,
+ .ndo_poll_controller = pch_gbe_netpoll,
+};
+
+static pci_ers_result_t pch_gbe_io_error_detected(struct pci_dev *pdev,
+ pci_channel_state_t state)
+{
+ struct net_device *netdev = pci_get_drvdata(pdev);
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+
+ netif_device_detach(netdev);
+ if (netif_running(netdev))
+ pch_gbe_down(adapter);
+ pci_disable_device(pdev);
+ /* Request a slot slot reset. */
+ return PCI_ERS_RESULT_NEED_RESET;
+}
+
+static pci_ers_result_t pch_gbe_io_slot_reset(struct pci_dev *pdev)
+{
+ struct net_device *netdev = pci_get_drvdata(pdev);
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+ struct pch_gbe_hw *hw = &adapter->hw;
+
+ if (pci_enable_device(pdev)) {
+ pr_err("Cannot re-enable PCI device after reset\n");
+ return PCI_ERS_RESULT_DISCONNECT;
+ }
+ pci_set_master(pdev);
+ pci_enable_wake(pdev, PCI_D0, 0);
+ pch_gbe_hal_power_up_phy(hw);
+ pch_gbe_reset(adapter);
+ /* Clear wake up status */
+ pch_gbe_mac_set_wol_event(hw, 0);
+
+ return PCI_ERS_RESULT_RECOVERED;
+}
+
+static void pch_gbe_io_resume(struct pci_dev *pdev)
+{
+ struct net_device *netdev = pci_get_drvdata(pdev);
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+
+ if (netif_running(netdev)) {
+ if (pch_gbe_up(adapter)) {
+ pr_debug("can't bring device back up after reset\n");
+ return;
+ }
+ }
+ netif_device_attach(netdev);
+}
+
+static int __pch_gbe_suspend(struct pci_dev *pdev)
+{
+ struct net_device *netdev = pci_get_drvdata(pdev);
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+ struct pch_gbe_hw *hw = &adapter->hw;
+ u32 wufc = adapter->wake_up_evt;
+ int retval = 0;
+
+ netif_device_detach(netdev);
+ if (netif_running(netdev))
+ pch_gbe_down(adapter);
+ if (wufc) {
+ pch_gbe_set_multi(netdev);
+ pch_gbe_setup_rctl(adapter);
+ pch_gbe_configure_rx(adapter);
+ pch_gbe_set_rgmii_ctrl(adapter, hw->mac.link_speed,
+ hw->mac.link_duplex);
+ pch_gbe_set_mode(adapter, hw->mac.link_speed,
+ hw->mac.link_duplex);
+ pch_gbe_mac_set_wol_event(hw, wufc);
+ pci_disable_device(pdev);
+ } else {
+ pch_gbe_hal_power_down_phy(hw);
+ pch_gbe_mac_set_wol_event(hw, wufc);
+ pci_disable_device(pdev);
+ }
+ return retval;
+}
+
+static int pch_gbe_suspend(struct device *device)
+{
+ struct pci_dev *pdev = to_pci_dev(device);
+
+ return __pch_gbe_suspend(pdev);
+}
+
+#ifdef CONFIG_PM
+static int pch_gbe_resume(struct device *device)
+{
+ struct pci_dev *pdev = to_pci_dev(device);
+ struct net_device *netdev = pci_get_drvdata(pdev);
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+ struct pch_gbe_hw *hw = &adapter->hw;
+ u32 err;
+
+ err = pci_enable_device(pdev);
+ if (err) {
+ pr_err("Cannot enable PCI device from suspend\n");
+ return err;
+ }
+ pci_set_master(pdev);
+ pch_gbe_hal_power_up_phy(hw);
+ pch_gbe_reset(adapter);
+ /* Clear wake on lan control and status */
+ pch_gbe_mac_set_wol_event(hw, 0);
+
+ if (netif_running(netdev))
+ pch_gbe_up(adapter);
+ netif_device_attach(netdev);
+
+ return 0;
+}
+#endif /* CONFIG_PM */
+
+static void pch_gbe_shutdown(struct pci_dev *pdev)
+{
+ __pch_gbe_suspend(pdev);
+ if (system_state == SYSTEM_POWER_OFF) {
+ pci_wake_from_d3(pdev, true);
+ pci_set_power_state(pdev, PCI_D3hot);
+ }
+}
+
+static void pch_gbe_remove(struct pci_dev *pdev)
+{
+ struct net_device *netdev = pci_get_drvdata(pdev);
+ struct pch_gbe_adapter *adapter = netdev_priv(netdev);
+
+ flush_scheduled_work();
+ unregister_netdev(netdev);
+
+ pch_gbe_hal_phy_hw_reset(&adapter->hw);
+
+ kfree(adapter->tx_ring);
+ kfree(adapter->rx_ring);
+
+ iounmap(adapter->hw.reg);
+ pci_release_regions(pdev);
+ free_netdev(netdev);
+ pci_disable_device(pdev);
+}
+
+static int pch_gbe_probe(struct pci_dev *pdev,
+ const struct pci_device_id *pci_id)
+{
+ struct net_device *netdev;
+ struct pch_gbe_adapter *adapter;
+ int ret;
+
+ ret = pci_enable_device(pdev);
+ if (ret)
+ return ret;
+
+ if (pci_set_dma_mask(pdev, DMA_BIT_MASK(64))
+ || pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(64))) {
+ ret = pci_set_dma_mask(pdev, DMA_BIT_MASK(32));
+ if (ret) {
+ ret = pci_set_consistent_dma_mask(pdev,
+ DMA_BIT_MASK(32));
+ if (ret) {
+ dev_err(&pdev->dev, "ERR: No usable DMA "
+ "configuration, aborting\n");
+ goto err_disable_device;
+ }
+ }
+ }
+
+ ret = pci_request_regions(pdev, KBUILD_MODNAME);
+ if (ret) {
+ dev_err(&pdev->dev,
+ "ERR: Can't reserve PCI I/O and memory resources\n");
+ goto err_disable_device;
+ }
+ pci_set_master(pdev);
+
+ netdev = alloc_etherdev((int)sizeof(struct pch_gbe_adapter));
+ if (!netdev) {
+ ret = -ENOMEM;
+ dev_err(&pdev->dev,
+ "ERR: Can't allocate and set up an Ethernet device\n");
+ goto err_release_pci;
+ }
+ SET_NETDEV_DEV(netdev, &pdev->dev);
+
+ pci_set_drvdata(pdev, netdev);
+ adapter = netdev_priv(netdev);
+ adapter->netdev = netdev;
+ adapter->pdev = pdev;
+ adapter->hw.back = adapter;
+ adapter->hw.reg = pci_iomap(pdev, PCH_GBE_PCI_BAR, 0);
+ if (!adapter->hw.reg) {
+ ret = -EIO;
+ dev_err(&pdev->dev, "Can't ioremap\n");
+ goto err_free_netdev;
+ }
+
+ netdev->netdev_ops = &pch_gbe_netdev_ops;
+ netdev->watchdog_timeo = PCH_GBE_WATCHDOG_PERIOD;
+ netif_napi_add(netdev, &adapter->napi,
+ pch_gbe_napi_poll, PCH_GBE_RX_WEIGHT);
+ netdev->features = NETIF_F_HW_CSUM | NETIF_F_GRO;
+ pch_gbe_set_ethtool_ops(netdev);
+
+ pch_gbe_mac_reset_hw(&adapter->hw);
+
+ /* setup the private structure */
+ ret = pch_gbe_sw_init(adapter);
+ if (ret)
+ goto err_iounmap;
+
+ /* Initialize PHY */
+ ret = pch_gbe_init_phy(adapter);
+ if (ret) {
+ dev_err(&pdev->dev, "PHY initialize error\n");
+ goto err_free_adapter;
+ }
+ pch_gbe_hal_get_bus_info(&adapter->hw);
+
+ /* Read the MAC address. and store to the private data */
+ ret = pch_gbe_hal_read_mac_addr(&adapter->hw);
+ if (ret) {
+ dev_err(&pdev->dev, "MAC address Read Error\n");
+ goto err_free_adapter;
+ }
+
+ memcpy(netdev->dev_addr, adapter->hw.mac.addr, netdev->addr_len);
+ if (!is_valid_ether_addr(netdev->dev_addr)) {
+ dev_err(&pdev->dev, "Invalid MAC Address\n");
+ ret = -EIO;
+ goto err_free_adapter;
+ }
+ setup_timer(&adapter->watchdog_timer, pch_gbe_watchdog,
+ (unsigned long)adapter);
+
+ INIT_WORK(&adapter->reset_task, pch_gbe_reset_task);
+
+ pch_gbe_check_options(adapter);
+
+ if (adapter->tx_csum)
+ netdev->features |= NETIF_F_HW_CSUM;
+ else
+ netdev->features &= ~NETIF_F_HW_CSUM;
+
+ /* initialize the wol settings based on the eeprom settings */
+ adapter->wake_up_evt = PCH_GBE_WL_INIT_SETTING;
+ dev_info(&pdev->dev, "MAC address : %pM\n", netdev->dev_addr);
+
+ /* reset the hardware with the new settings */
+ pch_gbe_reset(adapter);
+
+ ret = register_netdev(netdev);
+ if (ret)
+ goto err_free_adapter;
+ /* tell the stack to leave us alone until pch_gbe_open() is called */
+ netif_carrier_off(netdev);
+ netif_stop_queue(netdev);
+
+ dev_dbg(&pdev->dev, "OKIsemi(R) PCH Network Connection\n");
+
+ device_set_wakeup_enable(&pdev->dev, 1);
+ return 0;
+
+err_free_adapter:
+ pch_gbe_hal_phy_hw_reset(&adapter->hw);
+ kfree(adapter->tx_ring);
+ kfree(adapter->rx_ring);
+err_iounmap:
+ iounmap(adapter->hw.reg);
+err_free_netdev:
+ free_netdev(netdev);
+err_release_pci:
+ pci_release_regions(pdev);
+err_disable_device:
+ pci_disable_device(pdev);
+ return ret;
+}
+
+static const struct pci_device_id pch_gbe_pcidev_id[] = {
+ {.vendor = PCI_VENDOR_ID_INTEL,
+ .device = PCI_DEVICE_ID_INTEL_IOH1_GBE,
+ .subvendor = PCI_ANY_ID,
+ .subdevice = PCI_ANY_ID,
+ .class = (PCI_CLASS_NETWORK_ETHERNET << 8),
+ .class_mask = (0xFFFF00)
+ },
+ /* required last entry */
+ {0}
+};
+
+#ifdef CONFIG_PM
+static const struct dev_pm_ops pch_gbe_pm_ops = {
+ .suspend = pch_gbe_suspend,
+ .resume = pch_gbe_resume,
+ .freeze = pch_gbe_suspend,
+ .thaw = pch_gbe_resume,
+ .poweroff = pch_gbe_suspend,
+ .restore = pch_gbe_resume,
+};
+#endif
+
+static struct pci_error_handlers pch_gbe_err_handler = {
+ .error_detected = pch_gbe_io_error_detected,
+ .slot_reset = pch_gbe_io_slot_reset,
+ .resume = pch_gbe_io_resume
+};
+
+static struct pci_driver pch_gbe_pcidev = {
+ .name = KBUILD_MODNAME,
+ .id_table = pch_gbe_pcidev_id,
+ .probe = pch_gbe_probe,
+ .remove = pch_gbe_remove,
+#ifdef CONFIG_PM_OPS
+ .driver.pm = &pch_gbe_pm_ops,
+#endif
+ .shutdown = pch_gbe_shutdown,
+ .err_handler = &pch_gbe_err_handler
+};
+
+
+static int __init pch_gbe_init_module(void)
+{
+ int ret;
+
+ ret = pci_register_driver(&pch_gbe_pcidev);
+ if (copybreak != PCH_GBE_COPYBREAK_DEFAULT) {
+ if (copybreak == 0) {
+ pr_info("copybreak disabled\n");
+ } else {
+ pr_info("copybreak enabled for packets <= %u bytes\n",
+ copybreak);
+ }
+ }
+ return ret;
+}
+
+static void __exit pch_gbe_exit_module(void)
+{
+ pci_unregister_driver(&pch_gbe_pcidev);
+}
+
+module_init(pch_gbe_init_module);
+module_exit(pch_gbe_exit_module);
+
+MODULE_DESCRIPTION("OKI semiconductor PCH Gigabit ethernet Driver");
+MODULE_AUTHOR("OKI semiconductor, <masa-korg@dsn.okisemi.com>");
+MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_VERSION);
+MODULE_DEVICE_TABLE(pci, pch_gbe_pcidev_id);
+
+module_param(copybreak, uint, 0644);
+MODULE_PARM_DESC(copybreak,
+ "Maximum size of packet that is copied to a new buffer on receive");
+
+/* pch_gbe_main.c */
diff --git a/drivers/net/pch_gbe/pch_gbe_param.c b/drivers/net/pch_gbe/pch_gbe_param.c
new file mode 100644
index 0000000..9894a3c
--- /dev/null
+++ b/drivers/net/pch_gbe/pch_gbe_param.c
@@ -0,0 +1,501 @@
+/*
+ * Copyright (C) 1999 - 2010 Intel Corporation.
+ * Copyright (C) 2010 OKI SEMICONDUCTOR Co., LTD.
+ *
+ * This code was derived from the Intel e1000e Linux driver.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * 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 "pch_gbe.h"
+
+#define OPTION_UNSET -1
+#define OPTION_DISABLED 0
+#define OPTION_ENABLED 1
+
+/**
+ * TxDescriptors - Transmit Descriptor Count
+ * @Valid Range: PCH_GBE_MIN_TXD - PCH_GBE_MAX_TXD
+ * @Default Value: PCH_GBE_DEFAULT_TXD
+ */
+static int TxDescriptors = OPTION_UNSET;
+module_param(TxDescriptors, int, 0);
+MODULE_PARM_DESC(TxDescriptors, "Number of transmit descriptors");
+
+/**
+ * RxDescriptors -Receive Descriptor Count
+ * @Valid Range: PCH_GBE_MIN_RXD - PCH_GBE_MAX_RXD
+ * @Default Value: PCH_GBE_DEFAULT_RXD
+ */
+static int RxDescriptors = OPTION_UNSET;
+module_param(RxDescriptors, int, 0);
+MODULE_PARM_DESC(RxDescriptors, "Number of receive descriptors");
+
+/**
+ * Speed - User Specified Speed Override
+ * @Valid Range: 0, 10, 100, 1000
+ * - 0: auto-negotiate at all supported speeds
+ * - 10: only link at 10 Mbps
+ * - 100: only link at 100 Mbps
+ * - 1000: only link at 1000 Mbps
+ * @Default Value: 0
+ */
+static int Speed = OPTION_UNSET;
+module_param(Speed, int, 0);
+MODULE_PARM_DESC(Speed, "Speed setting");
+
+/**
+ * Duplex - User Specified Duplex Override
+ * @Valid Range: 0-2
+ * - 0: auto-negotiate for duplex
+ * - 1: only link at half duplex
+ * - 2: only link at full duplex
+ * @Default Value: 0
+ */
+static int Duplex = OPTION_UNSET;
+module_param(Duplex, int, 0);
+MODULE_PARM_DESC(Duplex, "Duplex setting");
+
+#define HALF_DUPLEX 1
+#define FULL_DUPLEX 2
+
+/**
+ * AutoNeg - Auto-negotiation Advertisement Override
+ * @Valid Range: 0x01-0x0F, 0x20-0x2F
+ *
+ * The AutoNeg value is a bit mask describing which speed and duplex
+ * combinations should be advertised during auto-negotiation.
+ * The supported speed and duplex modes are listed below
+ *
+ * Bit 7 6 5 4 3 2 1 0
+ * Speed (Mbps) N/A N/A 1000 N/A 100 100 10 10
+ * Duplex Full Full Half Full Half
+ *
+ * @Default Value: 0x2F (copper)
+ */
+static int AutoNeg = OPTION_UNSET;
+module_param(AutoNeg, int, 0);
+MODULE_PARM_DESC(AutoNeg, "Advertised auto-negotiation setting");
+
+#define PHY_ADVERTISE_10_HALF 0x0001
+#define PHY_ADVERTISE_10_FULL 0x0002
+#define PHY_ADVERTISE_100_HALF 0x0004
+#define PHY_ADVERTISE_100_FULL 0x0008
+#define PHY_ADVERTISE_1000_HALF 0x0010 /* Not used, just FYI */
+#define PHY_ADVERTISE_1000_FULL 0x0020
+#define PCH_AUTONEG_ADVERTISE_DEFAULT 0x2F
+
+/**
+ * FlowControl - User Specified Flow Control Override
+ * @Valid Range: 0-3
+ * - 0: No Flow Control
+ * - 1: Rx only, respond to PAUSE frames but do not generate them
+ * - 2: Tx only, generate PAUSE frames but ignore them on receive
+ * - 3: Full Flow Control Support
+ * @Default Value: Read flow control settings from the EEPROM
+ */
+static int FlowControl = OPTION_UNSET;
+module_param(FlowControl, int, 0);
+MODULE_PARM_DESC(FlowControl, "Flow Control setting");
+
+/*
+ * XsumRX - Receive Checksum Offload Enable/Disable
+ * @Valid Range: 0, 1
+ * - 0: disables all checksum offload
+ * - 1: enables receive IP/TCP/UDP checksum offload
+ * @Default Value: PCH_GBE_DEFAULT_RX_CSUM
+ */
+static int XsumRX = OPTION_UNSET;
+module_param(XsumRX, int, 0);
+MODULE_PARM_DESC(XsumRX, "Disable or enable Receive Checksum offload");
+
+#define PCH_GBE_DEFAULT_RX_CSUM true /* trueorfalse */
+
+/*
+ * XsumTX - Transmit Checksum Offload Enable/Disable
+ * @Valid Range: 0, 1
+ * - 0: disables all checksum offload
+ * - 1: enables transmit IP/TCP/UDP checksum offload
+ * @Default Value: PCH_GBE_DEFAULT_TX_CSUM
+ */
+static int XsumTX = OPTION_UNSET;
+module_param(XsumTX, int, 0);
+MODULE_PARM_DESC(XsumTX, "Disable or enable Transmit Checksum offload");
+
+#define PCH_GBE_DEFAULT_TX_CSUM true /* trueorfalse */
+
+/**
+ * pch_gbe_option - Force the MAC's flow control settings
+ * @hw: Pointer to the HW structure
+ * Returns
+ * 0: Successful.
+ * Negative value: Failed.
+ */
+struct pch_gbe_option {
+ enum { enable_option, range_option, list_option } type;
+ char *name;
+ char *err;
+ int def;
+ union {
+ struct { /* range_option info */
+ int min;
+ int max;
+ } r;
+ struct { /* list_option info */
+ int nr;
+ const struct pch_gbe_opt_list { int i; char *str; } *p;
+ } l;
+ } arg;
+};
+
+static const struct pch_gbe_opt_list speed_list[] = {
+ { 0, "" },
+ { SPEED_10, "" },
+ { SPEED_100, "" },
+ { SPEED_1000, "" }
+};
+
+static const struct pch_gbe_opt_list dplx_list[] = {
+ { 0, "" },
+ { HALF_DUPLEX, "" },
+ { FULL_DUPLEX, "" }
+};
+
+static const struct pch_gbe_opt_list an_list[] =
+ #define AA "AutoNeg advertising "
+ {{ 0x01, AA "10/HD" },
+ { 0x02, AA "10/FD" },
+ { 0x03, AA "10/FD, 10/HD" },
+ { 0x04, AA "100/HD" },
+ { 0x05, AA "100/HD, 10/HD" },
+ { 0x06, AA "100/HD, 10/FD" },
+ { 0x07, AA "100/HD, 10/FD, 10/HD" },
+ { 0x08, AA "100/FD" },
+ { 0x09, AA "100/FD, 10/HD" },
+ { 0x0a, AA "100/FD, 10/FD" },
+ { 0x0b, AA "100/FD, 10/FD, 10/HD" },
+ { 0x0c, AA "100/FD, 100/HD" },
+ { 0x0d, AA "100/FD, 100/HD, 10/HD" },
+ { 0x0e, AA "100/FD, 100/HD, 10/FD" },
+ { 0x0f, AA "100/FD, 100/HD, 10/FD, 10/HD" },
+ { 0x20, AA "1000/FD" },
+ { 0x21, AA "1000/FD, 10/HD" },
+ { 0x22, AA "1000/FD, 10/FD" },
+ { 0x23, AA "1000/FD, 10/FD, 10/HD" },
+ { 0x24, AA "1000/FD, 100/HD" },
+ { 0x25, AA "1000/FD, 100/HD, 10/HD" },
+ { 0x26, AA "1000/FD, 100/HD, 10/FD" },
+ { 0x27, AA "1000/FD, 100/HD, 10/FD, 10/HD" },
+ { 0x28, AA "1000/FD, 100/FD" },
+ { 0x29, AA "1000/FD, 100/FD, 10/HD" },
+ { 0x2a, AA "1000/FD, 100/FD, 10/FD" },
+ { 0x2b, AA "1000/FD, 100/FD, 10/FD, 10/HD" },
+ { 0x2c, AA "1000/FD, 100/FD, 100/HD" },
+ { 0x2d, AA "1000/FD, 100/FD, 100/HD, 10/HD" },
+ { 0x2e, AA "1000/FD, 100/FD, 100/HD, 10/FD" },
+ { 0x2f, AA "1000/FD, 100/FD, 100/HD, 10/FD, 10/HD" }
+};
+
+static const struct pch_gbe_opt_list fc_list[] = {
+ { PCH_GBE_FC_NONE, "Flow Control Disabled" },
+ { PCH_GBE_FC_RX_PAUSE, "Flow Control Receive Only" },
+ { PCH_GBE_FC_TX_PAUSE, "Flow Control Transmit Only" },
+ { PCH_GBE_FC_FULL, "Flow Control Enabled" }
+};
+
+/**
+ * pch_gbe_validate_option - Validate option
+ * @value: value
+ * @opt: option
+ * @adapter: Board private structure
+ * Returns
+ * 0: Successful.
+ * Negative value: Failed.
+ */
+static int pch_gbe_validate_option(int *value,
+ const struct pch_gbe_option *opt,
+ struct pch_gbe_adapter *adapter)
+{
+ if (*value == OPTION_UNSET) {
+ *value = opt->def;
+ return 0;
+ }
+
+ switch (opt->type) {
+ case enable_option:
+ switch (*value) {
+ case OPTION_ENABLED:
+ pr_debug("%s Enabled\n", opt->name);
+ return 0;
+ case OPTION_DISABLED:
+ pr_debug("%s Disabled\n", opt->name);
+ return 0;
+ }
+ break;
+ case range_option:
+ if (*value >= opt->arg.r.min && *value <= opt->arg.r.max) {
+ pr_debug("%s set to %i\n", opt->name, *value);
+ return 0;
+ }
+ break;
+ case list_option: {
+ int i;
+ const struct pch_gbe_opt_list *ent;
+
+ for (i = 0; i < opt->arg.l.nr; i++) {
+ ent = &opt->arg.l.p[i];
+ if (*value == ent->i) {
+ if (ent->str[0] != '\0')
+ pr_debug("%s\n", ent->str);
+ return 0;
+ }
+ }
+ }
+ break;
+ default:
+ BUG();
+ }
+
+ pr_debug("Invalid %s value specified (%i) %s\n",
+ opt->name, *value, opt->err);
+ *value = opt->def;
+ return -1;
+}
+
+/**
+ * pch_gbe_check_copper_options - Range Checking for Link Options, Copper Version
+ * @adapter: Board private structure
+ */
+static void pch_gbe_check_copper_options(struct pch_gbe_adapter *adapter)
+{
+ struct pch_gbe_hw *hw = &adapter->hw;
+ int speed, dplx;
+
+ { /* Speed */
+ static const struct pch_gbe_option opt = {
+ .type = list_option,
+ .name = "Speed",
+ .err = "parameter ignored",
+ .def = 0,
+ .arg = { .l = { .nr = (int)ARRAY_SIZE(speed_list),
+ .p = speed_list } }
+ };
+ speed = Speed;
+ pch_gbe_validate_option(&speed, &opt, adapter);
+ }
+ { /* Duplex */
+ static const struct pch_gbe_option opt = {
+ .type = list_option,
+ .name = "Duplex",
+ .err = "parameter ignored",
+ .def = 0,
+ .arg = { .l = { .nr = (int)ARRAY_SIZE(dplx_list),
+ .p = dplx_list } }
+ };
+ dplx = Duplex;
+ pch_gbe_validate_option(&dplx, &opt, adapter);
+ }
+
+ { /* Autoneg */
+ static const struct pch_gbe_option opt = {
+ .type = list_option,
+ .name = "AutoNeg",
+ .err = "parameter ignored",
+ .def = PCH_AUTONEG_ADVERTISE_DEFAULT,
+ .arg = { .l = { .nr = (int)ARRAY_SIZE(an_list),
+ .p = an_list} }
+ };
+ if (speed || dplx) {
+ pr_debug("AutoNeg specified along with Speed or Duplex, AutoNeg parameter ignored\n");
+ hw->phy.autoneg_advertised = opt.def;
+ } else {
+ hw->phy.autoneg_advertised = AutoNeg;
+ pch_gbe_validate_option(
+ (int *)(&hw->phy.autoneg_advertised),
+ &opt, adapter);
+ }
+ }
+
+ switch (speed + dplx) {
+ case 0:
+ hw->mac.autoneg = hw->mac.fc_autoneg = 1;
+ if ((speed || dplx))
+ pr_debug("Speed and duplex autonegotiation enabled\n");
+ hw->mac.link_speed = SPEED_10;
+ hw->mac.link_duplex = DUPLEX_HALF;
+ break;
+ case HALF_DUPLEX:
+ pr_debug("Half Duplex specified without Speed\n");
+ pr_debug("Using Autonegotiation at Half Duplex only\n");
+ hw->mac.autoneg = hw->mac.fc_autoneg = 1;
+ hw->phy.autoneg_advertised = PHY_ADVERTISE_10_HALF |
+ PHY_ADVERTISE_100_HALF;
+ hw->mac.link_speed = SPEED_10;
+ hw->mac.link_duplex = DUPLEX_HALF;
+ break;
+ case FULL_DUPLEX:
+ pr_debug("Full Duplex specified without Speed\n");
+ pr_debug("Using Autonegotiation at Full Duplex only\n");
+ hw->mac.autoneg = hw->mac.fc_autoneg = 1;
+ hw->phy.autoneg_advertised = PHY_ADVERTISE_10_FULL |
+ PHY_ADVERTISE_100_FULL |
+ PHY_ADVERTISE_1000_FULL;
+ hw->mac.link_speed = SPEED_10;
+ hw->mac.link_duplex = DUPLEX_FULL;
+ break;
+ case SPEED_10:
+ pr_debug("10 Mbps Speed specified without Duplex\n");
+ pr_debug("Using Autonegotiation at 10 Mbps only\n");
+ hw->mac.autoneg = hw->mac.fc_autoneg = 1;
+ hw->phy.autoneg_advertised = PHY_ADVERTISE_10_HALF |
+ PHY_ADVERTISE_10_FULL;
+ hw->mac.link_speed = SPEED_10;
+ hw->mac.link_duplex = DUPLEX_HALF;
+ break;
+ case SPEED_10 + HALF_DUPLEX:
+ pr_debug("Forcing to 10 Mbps Half Duplex\n");
+ hw->mac.autoneg = hw->mac.fc_autoneg = 0;
+ hw->phy.autoneg_advertised = 0;
+ hw->mac.link_speed = SPEED_10;
+ hw->mac.link_duplex = DUPLEX_HALF;
+ break;
+ case SPEED_10 + FULL_DUPLEX:
+ pr_debug("Forcing to 10 Mbps Full Duplex\n");
+ hw->mac.autoneg = hw->mac.fc_autoneg = 0;
+ hw->phy.autoneg_advertised = 0;
+ hw->mac.link_speed = SPEED_10;
+ hw->mac.link_duplex = DUPLEX_FULL;
+ break;
+ case SPEED_100:
+ pr_debug("100 Mbps Speed specified without Duplex\n");
+ pr_debug("Using Autonegotiation at 100 Mbps only\n");
+ hw->mac.autoneg = hw->mac.fc_autoneg = 1;
+ hw->phy.autoneg_advertised = PHY_ADVERTISE_100_HALF |
+ PHY_ADVERTISE_100_FULL;
+ hw->mac.link_speed = SPEED_100;
+ hw->mac.link_duplex = DUPLEX_HALF;
+ break;
+ case SPEED_100 + HALF_DUPLEX:
+ pr_debug("Forcing to 100 Mbps Half Duplex\n");
+ hw->mac.autoneg = hw->mac.fc_autoneg = 0;
+ hw->phy.autoneg_advertised = 0;
+ hw->mac.link_speed = SPEED_100;
+ hw->mac.link_duplex = DUPLEX_HALF;
+ break;
+ case SPEED_100 + FULL_DUPLEX:
+ pr_debug("Forcing to 100 Mbps Full Duplex\n");
+ hw->mac.autoneg = hw->mac.fc_autoneg = 0;
+ hw->phy.autoneg_advertised = 0;
+ hw->mac.link_speed = SPEED_100;
+ hw->mac.link_duplex = DUPLEX_FULL;
+ break;
+ case SPEED_1000:
+ pr_debug("1000 Mbps Speed specified without Duplex\n");
+ goto full_duplex_only;
+ case SPEED_1000 + HALF_DUPLEX:
+ pr_debug("Half Duplex is not supported at 1000 Mbps\n");
+ /* fall through */
+ case SPEED_1000 + FULL_DUPLEX:
+full_duplex_only:
+ pr_debug("Using Autonegotiation at 1000 Mbps Full Duplex only\n");
+ hw->mac.autoneg = hw->mac.fc_autoneg = 1;
+ hw->phy.autoneg_advertised = PHY_ADVERTISE_1000_FULL;
+ hw->mac.link_speed = SPEED_1000;
+ hw->mac.link_duplex = DUPLEX_FULL;
+ break;
+ default:
+ BUG();
+ }
+}
+
+/**
+ * pch_gbe_check_options - Range Checking for Command Line Parameters
+ * @adapter: Board private structure
+ */
+void pch_gbe_check_options(struct pch_gbe_adapter *adapter)
+{
+ struct pch_gbe_hw *hw = &adapter->hw;
+
+ { /* Transmit Descriptor Count */
+ static const struct pch_gbe_option opt = {
+ .type = range_option,
+ .name = "Transmit Descriptors",
+ .err = "using default of "
+ __MODULE_STRING(PCH_GBE_DEFAULT_TXD),
+ .def = PCH_GBE_DEFAULT_TXD,
+ .arg = { .r = { .min = PCH_GBE_MIN_TXD } },
+ .arg = { .r = { .max = PCH_GBE_MAX_TXD } }
+ };
+ struct pch_gbe_tx_ring *tx_ring = adapter->tx_ring;
+ tx_ring->count = TxDescriptors;
+ pch_gbe_validate_option(&tx_ring->count, &opt, adapter);
+ tx_ring->count = roundup(tx_ring->count,
+ PCH_GBE_TX_DESC_MULTIPLE);
+ }
+ { /* Receive Descriptor Count */
+ static const struct pch_gbe_option opt = {
+ .type = range_option,
+ .name = "Receive Descriptors",
+ .err = "using default of "
+ __MODULE_STRING(PCH_GBE_DEFAULT_RXD),
+ .def = PCH_GBE_DEFAULT_RXD,
+ .arg = { .r = { .min = PCH_GBE_MIN_RXD } },
+ .arg = { .r = { .max = PCH_GBE_MAX_RXD } }
+ };
+ struct pch_gbe_rx_ring *rx_ring = adapter->rx_ring;
+ rx_ring->count = RxDescriptors;
+ pch_gbe_validate_option(&rx_ring->count, &opt, adapter);
+ rx_ring->count = roundup(rx_ring->count,
+ PCH_GBE_RX_DESC_MULTIPLE);
+ }
+ { /* Checksum Offload Enable/Disable */
+ static const struct pch_gbe_option opt = {
+ .type = enable_option,
+ .name = "Checksum Offload",
+ .err = "defaulting to Enabled",
+ .def = PCH_GBE_DEFAULT_RX_CSUM
+ };
+ adapter->rx_csum = XsumRX;
+ pch_gbe_validate_option((int *)(&adapter->rx_csum),
+ &opt, adapter);
+ }
+ { /* Checksum Offload Enable/Disable */
+ static const struct pch_gbe_option opt = {
+ .type = enable_option,
+ .name = "Checksum Offload",
+ .err = "defaulting to Enabled",
+ .def = PCH_GBE_DEFAULT_TX_CSUM
+ };
+ adapter->tx_csum = XsumTX;
+ pch_gbe_validate_option((int *)(&adapter->tx_csum),
+ &opt, adapter);
+ }
+ { /* Flow Control */
+ static const struct pch_gbe_option opt = {
+ .type = list_option,
+ .name = "Flow Control",
+ .err = "reading default settings from EEPROM",
+ .def = PCH_GBE_FC_DEFAULT,
+ .arg = { .l = { .nr = (int)ARRAY_SIZE(fc_list),
+ .p = fc_list } }
+ };
+ hw->mac.fc = FlowControl;
+ pch_gbe_validate_option((int *)(&hw->mac.fc),
+ &opt, adapter);
+ }
+
+ pch_gbe_check_copper_options(adapter);
+}
+
+
diff --git a/drivers/net/pch_gbe/pch_gbe_phy.c b/drivers/net/pch_gbe/pch_gbe_phy.c
new file mode 100644
index 0000000..923a687
--- /dev/null
+++ b/drivers/net/pch_gbe/pch_gbe_phy.c
@@ -0,0 +1,274 @@
+/*
+ * Copyright (C) 1999 - 2010 Intel Corporation.
+ * Copyright (C) 2010 OKI SEMICONDUCTOR Co., LTD.
+ *
+ * This code was derived from the Intel e1000e Linux driver.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * 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 "pch_gbe.h"
+#include "pch_gbe_phy.h"
+
+#define PHY_MAX_REG_ADDRESS 0x1F /* 5 bit address bus (0-0x1F) */
+
+/* PHY 1000 MII Register/Bit Definitions */
+/* PHY Registers defined by IEEE */
+#define PHY_CONTROL 0x00 /* Control Register */
+#define PHY_STATUS 0x01 /* Status Regiser */
+#define PHY_ID1 0x02 /* Phy Id Register (word 1) */
+#define PHY_ID2 0x03 /* Phy Id Register (word 2) */
+#define PHY_AUTONEG_ADV 0x04 /* Autoneg Advertisement */
+#define PHY_LP_ABILITY 0x05 /* Link Partner Ability (Base Page) */
+#define PHY_AUTONEG_EXP 0x06 /* Autoneg Expansion Register */
+#define PHY_NEXT_PAGE_TX 0x07 /* Next Page TX */
+#define PHY_LP_NEXT_PAGE 0x08 /* Link Partner Next Page */
+#define PHY_1000T_CTRL 0x09 /* 1000Base-T Control Register */
+#define PHY_1000T_STATUS 0x0A /* 1000Base-T Status Register */
+#define PHY_EXT_STATUS 0x0F /* Extended Status Register */
+#define PHY_PHYSP_CONTROL 0x10 /* PHY Specific Control Register */
+#define PHY_EXT_PHYSP_CONTROL 0x14 /* Extended PHY Specific Control Register */
+#define PHY_LED_CONTROL 0x18 /* LED Control Register */
+#define PHY_EXT_PHYSP_STATUS 0x1B /* Extended PHY Specific Status Register */
+
+/* PHY Control Register */
+#define MII_CR_SPEED_SELECT_MSB 0x0040 /* bits 6,13: 10=1000, 01=100, 00=10 */
+#define MII_CR_COLL_TEST_ENABLE 0x0080 /* Collision test enable */
+#define MII_CR_FULL_DUPLEX 0x0100 /* FDX =1, half duplex =0 */
+#define MII_CR_RESTART_AUTO_NEG 0x0200 /* Restart auto negotiation */
+#define MII_CR_ISOLATE 0x0400 /* Isolate PHY from MII */
+#define MII_CR_POWER_DOWN 0x0800 /* Power down */
+#define MII_CR_AUTO_NEG_EN 0x1000 /* Auto Neg Enable */
+#define MII_CR_SPEED_SELECT_LSB 0x2000 /* bits 6,13: 10=1000, 01=100, 00=10 */
+#define MII_CR_LOOPBACK 0x4000 /* 0 = normal, 1 = loopback */
+#define MII_CR_RESET 0x8000 /* 0 = normal, 1 = PHY reset */
+#define MII_CR_SPEED_1000 0x0040
+#define MII_CR_SPEED_100 0x2000
+#define MII_CR_SPEED_10 0x0000
+
+/* PHY Status Register */
+#define MII_SR_EXTENDED_CAPS 0x0001 /* Extended register capabilities */
+#define MII_SR_JABBER_DETECT 0x0002 /* Jabber Detected */
+#define MII_SR_LINK_STATUS 0x0004 /* Link Status 1 = link */
+#define MII_SR_AUTONEG_CAPS 0x0008 /* Auto Neg Capable */
+#define MII_SR_REMOTE_FAULT 0x0010 /* Remote Fault Detect */
+#define MII_SR_AUTONEG_COMPLETE 0x0020 /* Auto Neg Complete */
+#define MII_SR_PREAMBLE_SUPPRESS 0x0040 /* Preamble may be suppressed */
+#define MII_SR_EXTENDED_STATUS 0x0100 /* Ext. status info in Reg 0x0F */
+#define MII_SR_100T2_HD_CAPS 0x0200 /* 100T2 Half Duplex Capable */
+#define MII_SR_100T2_FD_CAPS 0x0400 /* 100T2 Full Duplex Capable */
+#define MII_SR_10T_HD_CAPS 0x0800 /* 10T Half Duplex Capable */
+#define MII_SR_10T_FD_CAPS 0x1000 /* 10T Full Duplex Capable */
+#define MII_SR_100X_HD_CAPS 0x2000 /* 100X Half Duplex Capable */
+#define MII_SR_100X_FD_CAPS 0x4000 /* 100X Full Duplex Capable */
+#define MII_SR_100T4_CAPS 0x8000 /* 100T4 Capable */
+
+/* Phy Id Register (word 2) */
+#define PHY_REVISION_MASK 0x000F
+
+/* PHY Specific Control Register */
+#define PHYSP_CTRL_ASSERT_CRS_TX 0x0800
+
+
+/* Default value of PHY register */
+#define PHY_CONTROL_DEFAULT 0x1140 /* Control Register */
+#define PHY_AUTONEG_ADV_DEFAULT 0x01e0 /* Autoneg Advertisement */
+#define PHY_NEXT_PAGE_TX_DEFAULT 0x2001 /* Next Page TX */
+#define PHY_1000T_CTRL_DEFAULT 0x0300 /* 1000Base-T Control Register */
+#define PHY_PHYSP_CONTROL_DEFAULT 0x01EE /* PHY Specific Control Register */
+
+/**
+ * pch_gbe_phy_get_id - Retrieve the PHY ID and revision
+ * @hw: Pointer to the HW structure
+ * Returns
+ * 0: Successful.
+ * Negative value: Failed.
+ */
+s32 pch_gbe_phy_get_id(struct pch_gbe_hw *hw)
+{
+ struct pch_gbe_phy_info *phy = &hw->phy;
+ s32 ret;
+ u16 phy_id1;
+ u16 phy_id2;
+
+ ret = pch_gbe_phy_read_reg_miic(hw, PHY_ID1, &phy_id1);
+ if (ret)
+ return ret;
+ ret = pch_gbe_phy_read_reg_miic(hw, PHY_ID2, &phy_id2);
+ if (ret)
+ return ret;
+ /*
+ * PHY_ID1: [bit15-0:ID(21-6)]
+ * PHY_ID2: [bit15-10:ID(5-0)][bit9-4:Model][bit3-0:revision]
+ */
+ phy->id = (u32)phy_id1;
+ phy->id = ((phy->id << 6) | ((phy_id2 & 0xFC00) >> 10));
+ phy->revision = (u32) (phy_id2 & 0x000F);
+ pr_debug("phy->id : 0x%08x phy->revision : 0x%08x\n",
+ phy->id, phy->revision);
+ return 0;
+}
+
+/**
+ * pch_gbe_phy_read_reg_miic - Read MII control register
+ * @hw: Pointer to the HW structure
+ * @offset: Register offset to be read
+ * @data: Pointer to the read data
+ * Returns
+ * 0: Successful.
+ * -EINVAL: Invalid argument.
+ */
+s32 pch_gbe_phy_read_reg_miic(struct pch_gbe_hw *hw, u32 offset, u16 *data)
+{
+ struct pch_gbe_phy_info *phy = &hw->phy;
+
+ if (offset > PHY_MAX_REG_ADDRESS) {
+ pr_err("PHY Address %d is out of range\n", offset);
+ return -EINVAL;
+ }
+ *data = pch_gbe_mac_ctrl_miim(hw, phy->addr, PCH_GBE_HAL_MIIM_READ,
+ offset, (u16)0);
+ return 0;
+}
+
+/**
+ * pch_gbe_phy_write_reg_miic - Write MII control register
+ * @hw: Pointer to the HW structure
+ * @offset: Register offset to be read
+ * @data: data to write to register at offset
+ * Returns
+ * 0: Successful.
+ * -EINVAL: Invalid argument.
+ */
+s32 pch_gbe_phy_write_reg_miic(struct pch_gbe_hw *hw, u32 offset, u16 data)
+{
+ struct pch_gbe_phy_info *phy = &hw->phy;
+
+ if (offset > PHY_MAX_REG_ADDRESS) {
+ pr_err("PHY Address %d is out of range\n", offset);
+ return -EINVAL;
+ }
+ pch_gbe_mac_ctrl_miim(hw, phy->addr, PCH_GBE_HAL_MIIM_WRITE,
+ offset, data);
+ return 0;
+}
+
+/**
+ * pch_gbe_phy_sw_reset - PHY software reset
+ * @hw: Pointer to the HW structure
+ */
+void pch_gbe_phy_sw_reset(struct pch_gbe_hw *hw)
+{
+ u16 phy_ctrl;
+
+ pch_gbe_phy_read_reg_miic(hw, PHY_CONTROL, &phy_ctrl);
+ phy_ctrl |= MII_CR_RESET;
+ pch_gbe_phy_write_reg_miic(hw, PHY_CONTROL, phy_ctrl);
+ udelay(1);
+}
+
+/**
+ * pch_gbe_phy_hw_reset - PHY hardware reset
+ * @hw: Pointer to the HW structure
+ */
+void pch_gbe_phy_hw_reset(struct pch_gbe_hw *hw)
+{
+ pch_gbe_phy_write_reg_miic(hw, PHY_CONTROL, PHY_CONTROL_DEFAULT);
+ pch_gbe_phy_write_reg_miic(hw, PHY_AUTONEG_ADV,
+ PHY_AUTONEG_ADV_DEFAULT);
+ pch_gbe_phy_write_reg_miic(hw, PHY_NEXT_PAGE_TX,
+ PHY_NEXT_PAGE_TX_DEFAULT);
+ pch_gbe_phy_write_reg_miic(hw, PHY_1000T_CTRL, PHY_1000T_CTRL_DEFAULT);
+ pch_gbe_phy_write_reg_miic(hw, PHY_PHYSP_CONTROL,
+ PHY_PHYSP_CONTROL_DEFAULT);
+}
+
+/**
+ * pch_gbe_phy_power_up - restore link in case the phy was powered down
+ * @hw: Pointer to the HW structure
+ */
+void pch_gbe_phy_power_up(struct pch_gbe_hw *hw)
+{
+ u16 mii_reg;
+
+ mii_reg = 0;
+ /* Just clear the power down bit to wake the phy back up */
+ /* according to the manual, the phy will retain its
+ * settings across a power-down/up cycle */
+ pch_gbe_phy_read_reg_miic(hw, PHY_CONTROL, &mii_reg);
+ mii_reg &= ~MII_CR_POWER_DOWN;
+ pch_gbe_phy_write_reg_miic(hw, PHY_CONTROL, mii_reg);
+}
+
+/**
+ * pch_gbe_phy_power_down - Power down PHY
+ * @hw: Pointer to the HW structure
+ */
+void pch_gbe_phy_power_down(struct pch_gbe_hw *hw)
+{
+ u16 mii_reg;
+
+ mii_reg = 0;
+ /* Power down the PHY so no link is implied when interface is down *
+ * The PHY cannot be powered down if any of the following is TRUE *
+ * (a) WoL is enabled
+ * (b) AMT is active
+ */
+ pch_gbe_phy_read_reg_miic(hw, PHY_CONTROL, &mii_reg);
+ mii_reg |= MII_CR_POWER_DOWN;
+ pch_gbe_phy_write_reg_miic(hw, PHY_CONTROL, mii_reg);
+ mdelay(1);
+}
+
+/**
+ * pch_gbe_phy_set_rgmii - RGMII interface setting
+ * @hw: Pointer to the HW structure
+ */
+inline void pch_gbe_phy_set_rgmii(struct pch_gbe_hw *hw)
+{
+ pch_gbe_phy_sw_reset(hw);
+}
+
+/**
+ * pch_gbe_phy_init_setting - PHY initial setting
+ * @hw: Pointer to the HW structure
+ */
+void pch_gbe_phy_init_setting(struct pch_gbe_hw *hw)
+{
+ struct pch_gbe_adapter *adapter;
+ struct ethtool_cmd cmd;
+ int ret;
+ u16 mii_reg;
+
+ adapter = container_of(hw, struct pch_gbe_adapter, hw);
+ ret = mii_ethtool_gset(&adapter->mii, &cmd);
+ if (ret)
+ pr_err("Error: mii_ethtool_gset\n");
+
+ cmd.speed = hw->mac.link_speed;
+ cmd.duplex = hw->mac.link_duplex;
+ cmd.advertising = hw->phy.autoneg_advertised;
+ cmd.autoneg = hw->mac.autoneg;
+ pch_gbe_phy_write_reg_miic(hw, MII_BMCR, BMCR_RESET);
+ ret = mii_ethtool_sset(&adapter->mii, &cmd);
+ if (ret)
+ pr_err("Error: mii_ethtool_sset\n");
+
+ pch_gbe_phy_sw_reset(hw);
+
+ pch_gbe_phy_read_reg_miic(hw, PHY_PHYSP_CONTROL, &mii_reg);
+ mii_reg |= PHYSP_CTRL_ASSERT_CRS_TX;
+ pch_gbe_phy_write_reg_miic(hw, PHY_PHYSP_CONTROL, mii_reg);
+
+}
diff --git a/drivers/net/pch_gbe/pch_gbe_phy.h b/drivers/net/pch_gbe/pch_gbe_phy.h
new file mode 100644
index 0000000..03264dc
--- /dev/null
+++ b/drivers/net/pch_gbe/pch_gbe_phy.h
@@ -0,0 +1,37 @@
+/*
+ * Copyright (C) 1999 - 2010 Intel Corporation.
+ * Copyright (C) 2010 OKI SEMICONDUCTOR Co., LTD.
+ *
+ * This code was derived from the Intel e1000e Linux driver.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * 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 _PCH_GBE_PHY_H_
+#define _PCH_GBE_PHY_H_
+
+#define PCH_GBE_PHY_REGS_LEN 32
+#define PCH_GBE_PHY_RESET_DELAY_US 10
+#define PCH_GBE_MAC_IFOP_RGMII
+
+s32 pch_gbe_phy_get_id(struct pch_gbe_hw *hw);
+s32 pch_gbe_phy_read_reg_miic(struct pch_gbe_hw *hw, u32 offset, u16 *data);
+s32 pch_gbe_phy_write_reg_miic(struct pch_gbe_hw *hw, u32 offset, u16 data);
+void pch_gbe_phy_sw_reset(struct pch_gbe_hw *hw);
+void pch_gbe_phy_hw_reset(struct pch_gbe_hw *hw);
+void pch_gbe_phy_power_up(struct pch_gbe_hw *hw);
+void pch_gbe_phy_power_down(struct pch_gbe_hw *hw);
+void pch_gbe_phy_set_rgmii(struct pch_gbe_hw *hw);
+void pch_gbe_phy_init_setting(struct pch_gbe_hw *hw);
+
+#endif /* _PCH_GBE_PHY_H_ */
--
1.6.2.5
^ permalink raw reply related
* Re: E1000E/82567LM-3: link reported up too soon
From: Arnaud Ebalard @ 2010-09-21 11:03 UTC (permalink / raw)
To: David Miller; +Cc: brian.haley, netdev
In-Reply-To: <20100920.152352.193726412.davem@davemloft.net>
Hi,
David Miller <davem@davemloft.net> writes:
> From: arno@natisbad.org (Arnaud Ebalard)
> Date: Mon, 20 Sep 2010 23:28:57 +0200
>
>> I added printk() to see where/if those first packets get dropped in
>> e1000e code. I followed those first packets to e1000_xmit_frame() and
>> don't see anything obvious happening there, i.e. they are passed to the
>> device as expected.
>
> I suspect the packet is sucessfully given to the chip and the PHY simply
> doesn't put it onto the wire for whatever reason.
I spent additional time on it, putting printk() at various locations in
e1000e driver code and also using driver's debug functions to see what
could be different in the internal state when the first packet is
delivered and when it is not. I just found nothing, so AFAICT your guess
above is backed up by those tests.
I will try and summarize that for Intel ethernet drivers maintainers.
Cheers,
a+
^ permalink raw reply
* Re: Regression, bisected: reference leak with IPSec since ~2.6.31
From: Jarek Poplawski @ 2010-09-21 10:48 UTC (permalink / raw)
To: Eric Dumazet; +Cc: Nick Bowler, linux-kernel, netdev, David S. Miller
In-Reply-To: <1285063657.2617.226.camel@edumazet-laptop>
On Tue, Sep 21, 2010 at 12:07:37PM +0200, Eric Dumazet wrote:
> Le mardi 21 septembre 2010 ?? 11:55 +0200, Eric Dumazet a écrit :
> > Le mardi 21 septembre 2010 ?? 09:38 +0000, Jarek Poplawski a écrit :
> > > On Tue, Sep 21, 2010 at 11:21:00AM +0200, Eric Dumazet wrote:
> >
> > > I hope you're right with this.
> > >
> > > >
> > > > I liked the :
> > > >
> > > > <check everything without changing state>
> > > > if something wrong
> > > > goto slow_path
> > > > else
> > > > <OK, lets do destructive things>
> > > >
> > >
> > > But it's an optimization of the "unlikely" case btw. ;-)
> > >
> >
> > This is a bug fix, in a complex function.
> >
> > in -next branch, we can try to be smart, adding more code in slow_path
> > to revert what was done in the hope fast path would be taken.
> > And pray we dont add another bug :)
But if you were honest about caches there is no reason to be smart
no more. ;-)
Below is your patch mostly as an example only that I didn't mean
nothing complex (ipv4 only, not compiled).
Jarek P.
---
diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
index 04b6989..049d61f 100644
--- a/net/ipv4/ip_output.c
+++ b/net/ipv4/ip_output.c
@@ -490,7 +490,6 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
if (skb_has_frags(skb)) {
struct sk_buff *frag;
int first_len = skb_pagelen(skb);
- int truesizes = 0;
if (first_len - hlen > mtu ||
((first_len - hlen) & 7) ||
@@ -503,18 +502,18 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
if (frag->len > mtu ||
((frag->len & 7) && frag->next) ||
skb_headroom(frag) < hlen)
- goto slow_path;
+ goto slow_path_clean;
/* Partially cloned skb? */
if (skb_shared(frag))
- goto slow_path;
+ goto slow_path_clean;
BUG_ON(frag->sk);
if (skb->sk) {
frag->sk = skb->sk;
frag->destructor = sock_wfree;
+ skb->truesize -= frag->truesize;
}
- truesizes += frag->truesize;
}
/* Everything is OK. Generate! */
@@ -524,7 +523,6 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
frag = skb_shinfo(skb)->frag_list;
skb_frag_list_init(skb);
skb->data_len = first_len - skb_headlen(skb);
- skb->truesize -= truesizes;
skb->len = first_len;
iph->tot_len = htons(first_len);
iph->frag_off = htons(IP_MF);
@@ -576,6 +574,17 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
}
IP_INC_STATS(dev_net(dev), IPSTATS_MIB_FRAGFAILS);
return err;
+
+slow_path_clean:
+ if (skb->sk) {
+ skb_walk_frags(skb, frag) {
+ if (!frag->sk)
+ break;
+ frag->sk = NULL;
+ frag->destructor = NULL;
+ skb->truesize += frag->truesize;
+ }
+ }
}
slow_path:
^ permalink raw reply related
* [PATCH 1/5] dccp: Add packet type information to CCID-specific option parsing
From: Gerrit Renker @ 2010-09-21 10:28 UTC (permalink / raw)
To: davem; +Cc: dccp, netdev, Gerrit Renker
In-Reply-To: <1285064938-5933-1-git-send-email-gerrit@erg.abdn.ac.uk>
This
1. adds packet type information to ccid_hc_{rx,tx}_parse_options(). This is
necessary, since table 3 in RFC 4340, 5.8 leaves it to the CCIDs to state
which options may (not) appear on what packet type.
2. adds such a check for CCID-3's {Loss Event, Receive} Rate as specified in
RFC 4340 8.3 ("Receive Rate options MUST NOT be sent on DCCP-Data packets")
and 8.5 ("Loss Event Rate options MUST NOT be sent on DCCP-Data packets").
3. removes an unused argument `idx' from ccid_hc_{rx,tx}_parse_options(). This
is also no longer necessary, since the CCID-specific option-parsing routines
are passed every single parameter of the type-length-value option encoding.
Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
---
net/dccp/ccid.h | 46 +++++++++++++++++++++++-----------------------
net/dccp/ccids/ccid3.c | 14 ++++++++------
net/dccp/options.c | 16 ++++------------
3 files changed, 35 insertions(+), 41 deletions(-)
--- a/net/dccp/ccid.h
+++ b/net/dccp/ccid.h
@@ -62,18 +62,14 @@ struct ccid_operations {
void (*ccid_hc_tx_exit)(struct sock *sk);
void (*ccid_hc_rx_packet_recv)(struct sock *sk,
struct sk_buff *skb);
- int (*ccid_hc_rx_parse_options)(struct sock *sk,
- unsigned char option,
- unsigned char len, u16 idx,
- unsigned char* value);
+ int (*ccid_hc_rx_parse_options)(struct sock *sk, u8 pkt,
+ u8 opt, u8 *val, u8 len);
int (*ccid_hc_rx_insert_options)(struct sock *sk,
struct sk_buff *skb);
void (*ccid_hc_tx_packet_recv)(struct sock *sk,
struct sk_buff *skb);
- int (*ccid_hc_tx_parse_options)(struct sock *sk,
- unsigned char option,
- unsigned char len, u16 idx,
- unsigned char* value);
+ int (*ccid_hc_tx_parse_options)(struct sock *sk, u8 pkt,
+ u8 opt, u8 *val, u8 len);
int (*ccid_hc_tx_send_packet)(struct sock *sk,
struct sk_buff *skb);
void (*ccid_hc_tx_packet_sent)(struct sock *sk,
@@ -168,27 +164,31 @@ static inline void ccid_hc_tx_packet_recv(struct ccid *ccid, struct sock *sk,
ccid->ccid_ops->ccid_hc_tx_packet_recv(sk, skb);
}
+/**
+ * ccid_hc_tx_parse_options - Parse CCID-specific options sent by the receiver
+ * @pkt: type of packet that @opt appears on (RFC 4340, 5.1)
+ * @opt: the CCID-specific option type (RFC 4340, 5.8 and 10.3)
+ * @val: value of @opt
+ * @len: length of @val in bytes
+ */
static inline int ccid_hc_tx_parse_options(struct ccid *ccid, struct sock *sk,
- unsigned char option,
- unsigned char len, u16 idx,
- unsigned char* value)
+ u8 pkt, u8 opt, u8 *val, u8 len)
{
- int rc = 0;
- if (ccid->ccid_ops->ccid_hc_tx_parse_options != NULL)
- rc = ccid->ccid_ops->ccid_hc_tx_parse_options(sk, option, len, idx,
- value);
- return rc;
+ if (ccid->ccid_ops->ccid_hc_tx_parse_options == NULL)
+ return 0;
+ return ccid->ccid_ops->ccid_hc_tx_parse_options(sk, pkt, opt, val, len);
}
+/**
+ * ccid_hc_rx_parse_options - Parse CCID-specific options sent by the sender
+ * Arguments are analogous to ccid_hc_tx_parse_options()
+ */
static inline int ccid_hc_rx_parse_options(struct ccid *ccid, struct sock *sk,
- unsigned char option,
- unsigned char len, u16 idx,
- unsigned char* value)
+ u8 pkt, u8 opt, u8 *val, u8 len)
{
- int rc = 0;
- if (ccid->ccid_ops->ccid_hc_rx_parse_options != NULL)
- rc = ccid->ccid_ops->ccid_hc_rx_parse_options(sk, option, len, idx, value);
- return rc;
+ if (ccid->ccid_ops->ccid_hc_rx_parse_options == NULL)
+ return 0;
+ return ccid->ccid_ops->ccid_hc_rx_parse_options(sk, pkt, opt, val, len);
}
static inline int ccid_hc_rx_insert_options(struct ccid *ccid, struct sock *sk,
--- a/net/dccp/ccids/ccid3.c
+++ b/net/dccp/ccids/ccid3.c
@@ -481,9 +481,8 @@ done_computing_x:
jiffies + usecs_to_jiffies(t_nfb));
}
-static int ccid3_hc_tx_parse_options(struct sock *sk, unsigned char option,
- unsigned char len, u16 idx,
- unsigned char *value)
+static int ccid3_hc_tx_parse_options(struct sock *sk, u8 packet_type,
+ u8 option, u8 *optval, u8 optlen)
{
struct ccid3_hc_tx_sock *hc = ccid3_hc_tx_sk(sk);
struct ccid3_options_received *opt_recv = &hc->tx_options_received;
@@ -492,12 +491,15 @@ static int ccid3_hc_tx_parse_options(struct sock *sk, unsigned char option,
switch (option) {
case TFRC_OPT_RECEIVE_RATE:
case TFRC_OPT_LOSS_EVENT_RATE:
- if (unlikely(len != 4)) {
+ /* Must be ignored on Data packets, cf. RFC 4342 8.3 and 8.5 */
+ if (packet_type == DCCP_PKT_DATA)
+ break;
+ if (unlikely(optlen != 4)) {
DCCP_WARN("%s(%p), invalid len %d for %u\n",
- dccp_role(sk), sk, len, option);
+ dccp_role(sk), sk, optlen, option);
return -EINVAL;
}
- opt_val = ntohl(get_unaligned((__be32 *)value));
+ opt_val = ntohl(get_unaligned((__be32 *)optval));
if (option == TFRC_OPT_RECEIVE_RATE) {
opt_recv->ccid3or_receive_rate = opt_val;
--- a/net/dccp/options.c
+++ b/net/dccp/options.c
@@ -226,23 +226,15 @@ int dccp_parse_options(struct sock *sk, struct dccp_request_sock *dreq,
dccp_pr_debug("%s rx opt: ELAPSED_TIME=%d\n",
dccp_role(sk), elapsed_time);
break;
- case 128 ... 191: {
- const u16 idx = value - options;
-
+ case 128 ... 191:
if (ccid_hc_rx_parse_options(dp->dccps_hc_rx_ccid, sk,
- opt, len, idx,
- value) != 0)
+ pkt_type, opt, value, len))
goto out_invalid_option;
- }
break;
- case 192 ... 255: {
- const u16 idx = value - options;
-
+ case 192 ... 255:
if (ccid_hc_tx_parse_options(dp->dccps_hc_tx_ccid, sk,
- opt, len, idx,
- value) != 0)
+ pkt_type, opt, value, len))
goto out_invalid_option;
- }
break;
default:
DCCP_CRIT("DCCP(%p): option %d(len=%d) not "
^ permalink raw reply
* [PATCH 3/5] dccp ccid-3: remove dead states
From: Gerrit Renker @ 2010-09-21 10:28 UTC (permalink / raw)
To: davem; +Cc: dccp, netdev, Gerrit Renker
In-Reply-To: <1285064938-5933-3-git-send-email-gerrit@erg.abdn.ac.uk>
This patch is thanks to an investigation by Leandro Sales de Melo and his
colleagues. They worked out two state diagrams which highlight the fact that
the xxx_TERM states in CCID-3/4 are in fact not necessary.
And this can be confirmed by in turn looking at the code: the xxx_TERM states
are only ever set in ccid3_hc_{rx,tx}_exit(): when CCID-3 sets the state
to xxx_TERM, it is at a time where no more processing should be going on,
hence it is not necessary to introduce a dedicated exit state - this is already
implied by unloading the CCID.
Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
---
net/dccp/ccids/ccid3.h | 2 --
net/dccp/ccids/ccid3.c | 37 +++++++++----------------------------
2 files changed, 9 insertions(+), 30 deletions(-)
--- a/net/dccp/ccids/ccid3.h
+++ b/net/dccp/ccids/ccid3.h
@@ -77,7 +77,6 @@ enum ccid3_hc_tx_states {
TFRC_SSTATE_NO_SENT = 1,
TFRC_SSTATE_NO_FBACK,
TFRC_SSTATE_FBACK,
- TFRC_SSTATE_TERM,
};
/**
@@ -130,7 +129,6 @@ static inline struct ccid3_hc_tx_sock *ccid3_hc_tx_sk(const struct sock *sk)
enum ccid3_hc_rx_states {
TFRC_RSTATE_NO_DATA = 1,
TFRC_RSTATE_DATA,
- TFRC_RSTATE_TERM = 127,
};
/**
--- a/net/dccp/ccids/ccid3.c
+++ b/net/dccp/ccids/ccid3.c
@@ -54,7 +54,6 @@ static const char *ccid3_tx_state_name(enum ccid3_hc_tx_states state)
[TFRC_SSTATE_NO_SENT] = "NO_SENT",
[TFRC_SSTATE_NO_FBACK] = "NO_FBACK",
[TFRC_SSTATE_FBACK] = "FBACK",
- [TFRC_SSTATE_TERM] = "TERM",
};
return ccid3_state_names[state];
@@ -208,10 +207,13 @@ static void ccid3_hc_tx_no_feedback_timer(unsigned long data)
ccid3_pr_debug("%s(%p, state=%s) - entry\n", dccp_role(sk), sk,
ccid3_tx_state_name(hc->tx_state));
+ /* Ignore and do not restart after leaving the established state */
+ if ((1 << sk->sk_state) & ~(DCCPF_OPEN | DCCPF_PARTOPEN))
+ goto out;
+
+ /* Reset feedback state to "no feedback received" */
if (hc->tx_state == TFRC_SSTATE_FBACK)
ccid3_hc_tx_set_state(sk, TFRC_SSTATE_NO_FBACK);
- else if (hc->tx_state != TFRC_SSTATE_NO_FBACK)
- goto out;
/*
* Determine new allowed sending rate X as per draft rfc3448bis-00, 4.4
@@ -287,8 +289,7 @@ static int ccid3_hc_tx_send_packet(struct sock *sk, struct sk_buff *skb)
if (unlikely(skb->len == 0))
return -EBADMSG;
- switch (hc->tx_state) {
- case TFRC_SSTATE_NO_SENT:
+ if (hc->tx_state == TFRC_SSTATE_NO_SENT) {
sk_reset_timer(sk, &hc->tx_no_feedback_timer, (jiffies +
usecs_to_jiffies(TFRC_INITIAL_TIMEOUT)));
hc->tx_last_win_count = 0;
@@ -323,9 +324,8 @@ static int ccid3_hc_tx_send_packet(struct sock *sk, struct sk_buff *skb)
ccid3_update_send_interval(hc);
ccid3_hc_tx_set_state(sk, TFRC_SSTATE_NO_FBACK);
- break;
- case TFRC_SSTATE_NO_FBACK:
- case TFRC_SSTATE_FBACK:
+
+ } else {
delay = ktime_us_delta(hc->tx_t_nom, now);
ccid3_pr_debug("delay=%ld\n", (long)delay);
/*
@@ -340,10 +340,6 @@ static int ccid3_hc_tx_send_packet(struct sock *sk, struct sk_buff *skb)
return (u32)delay / USEC_PER_MSEC;
ccid3_hc_tx_update_win_count(hc, now);
- break;
- case TFRC_SSTATE_TERM:
- DCCP_BUG("%s(%p) - Illegal state TERM", dccp_role(sk), sk);
- return -EINVAL;
}
/* prepare to send now (add options etc.) */
@@ -379,11 +375,6 @@ static void ccid3_hc_tx_packet_recv(struct sock *sk, struct sk_buff *skb)
if (!(DCCP_SKB_CB(skb)->dccpd_type == DCCP_PKT_ACK ||
DCCP_SKB_CB(skb)->dccpd_type == DCCP_PKT_DATAACK))
return;
- /* ... and only in the established state */
- if (hc->tx_state != TFRC_SSTATE_FBACK &&
- hc->tx_state != TFRC_SSTATE_NO_FBACK)
- return;
-
/*
* Locate the acknowledged packet in the TX history.
*
@@ -529,9 +520,7 @@ static void ccid3_hc_tx_exit(struct sock *sk)
{
struct ccid3_hc_tx_sock *hc = ccid3_hc_tx_sk(sk);
- ccid3_hc_tx_set_state(sk, TFRC_SSTATE_TERM);
sk_stop_timer(sk, &hc->tx_no_feedback_timer);
-
tfrc_tx_hist_purge(&hc->tx_hist);
}
@@ -590,7 +579,6 @@ static const char *ccid3_rx_state_name(enum ccid3_hc_rx_states state)
static const char *const ccid3_rx_state_names[] = {
[TFRC_RSTATE_NO_DATA] = "NO_DATA",
[TFRC_RSTATE_DATA] = "DATA",
- [TFRC_RSTATE_TERM] = "TERM",
};
return ccid3_rx_state_names[state];
@@ -616,14 +604,9 @@ static void ccid3_hc_rx_send_feedback(struct sock *sk,
{
struct ccid3_hc_rx_sock *hc = ccid3_hc_rx_sk(sk);
struct dccp_sock *dp = dccp_sk(sk);
- ktime_t now;
+ ktime_t now = ktime_get_real();
s64 delta = 0;
- if (unlikely(hc->rx_state == TFRC_RSTATE_TERM))
- return;
-
- now = ktime_get_real();
-
switch (fbtype) {
case CCID3_FBACK_INITIAL:
hc->rx_x_recv = 0;
@@ -827,8 +810,6 @@ static void ccid3_hc_rx_exit(struct sock *sk)
{
struct ccid3_hc_rx_sock *hc = ccid3_hc_rx_sk(sk);
- ccid3_hc_rx_set_state(sk, TFRC_RSTATE_TERM);
-
tfrc_rx_hist_purge(&hc->rx_hist);
tfrc_lh_cleanup(&hc->rx_li_hist);
}
^ permalink raw reply
* [PATCH 5/5] dccp ccid-3: Remove redundant 'options_received' struct
From: Gerrit Renker @ 2010-09-21 10:28 UTC (permalink / raw)
To: davem; +Cc: dccp, netdev, Gerrit Renker
In-Reply-To: <1285064938-5933-5-git-send-email-gerrit@erg.abdn.ac.uk>
The `options_received' struct is redundant, since it re-duplicates the existing
`p' and `x_recv' fields. This patch removes the sub-struct and migrates the
format conversion operations to ccid3_hc_tx_parse_options().
Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
---
net/dccp/ccids/ccid3.h | 7 -------
net/dccp/ccids/ccid3.c | 24 ++++++++----------------
2 files changed, 8 insertions(+), 23 deletions(-)
--- a/net/dccp/ccids/ccid3.c
+++ b/net/dccp/ccids/ccid3.c
@@ -365,11 +365,10 @@ static void ccid3_hc_tx_packet_sent(struct sock *sk, int more,
static void ccid3_hc_tx_packet_recv(struct sock *sk, struct sk_buff *skb)
{
struct ccid3_hc_tx_sock *hc = ccid3_hc_tx_sk(sk);
- struct ccid3_options_received *opt_recv = &hc->tx_options_received;
struct tfrc_tx_hist_entry *acked;
ktime_t now;
unsigned long t_nfb;
- u32 pinv, r_sample;
+ u32 r_sample;
/* we are only interested in ACKs */
if (!(DCCP_SKB_CB(skb)->dccpd_type == DCCP_PKT_ACK ||
@@ -394,17 +393,6 @@ static void ccid3_hc_tx_packet_recv(struct sock *sk, struct sk_buff *skb)
r_sample = dccp_sample_rtt(sk, ktime_us_delta(now, acked->stamp));
hc->tx_rtt = tfrc_ewma(hc->tx_rtt, r_sample, 9);
- /* Update receive rate in units of 64 * bytes/second */
- hc->tx_x_recv = opt_recv->ccid3or_receive_rate;
- hc->tx_x_recv <<= 6;
-
- /* Update loss event rate (which is scaled by 1e6) */
- pinv = opt_recv->ccid3or_loss_event_rate;
- if (pinv == 0)
- hc->tx_p = 0;
- else
- hc->tx_p = tfrc_invert_loss_event_rate(pinv);
-
/*
* Update allowed sending rate X as per draft rfc3448bis-00, 4.2/3
*/
@@ -476,7 +464,6 @@ static int ccid3_hc_tx_parse_options(struct sock *sk, u8 packet_type,
u8 option, u8 *optval, u8 optlen)
{
struct ccid3_hc_tx_sock *hc = ccid3_hc_tx_sk(sk);
- struct ccid3_options_received *opt_recv = &hc->tx_options_received;
__be32 opt_val;
switch (option) {
@@ -493,11 +480,16 @@ static int ccid3_hc_tx_parse_options(struct sock *sk, u8 packet_type,
opt_val = ntohl(get_unaligned((__be32 *)optval));
if (option == TFRC_OPT_RECEIVE_RATE) {
- opt_recv->ccid3or_receive_rate = opt_val;
+ /* Receive Rate is kept in units of 64 bytes/second */
+ hc->tx_x_recv = opt_val;
+ hc->tx_x_recv <<= 6;
+
ccid3_pr_debug("%s(%p), RECEIVE_RATE=%u\n",
dccp_role(sk), sk, opt_val);
} else {
- opt_recv->ccid3or_loss_event_rate = opt_val;
+ /* Update the fixpoint Loss Event Rate fraction */
+ hc->tx_p = tfrc_invert_loss_event_rate(opt_val);
+
ccid3_pr_debug("%s(%p), LOSS_EVENT_RATE=%u\n",
dccp_role(sk), sk, opt_val);
}
--- a/net/dccp/ccids/ccid3.h
+++ b/net/dccp/ccids/ccid3.h
@@ -67,11 +67,6 @@ enum ccid3_options {
TFRC_OPT_RECEIVE_RATE = 194,
};
-struct ccid3_options_received {
- u32 ccid3or_loss_event_rate;
- u32 ccid3or_receive_rate;
-};
-
/* TFRC sender states */
enum ccid3_hc_tx_states {
TFRC_SSTATE_NO_SENT = 1,
@@ -97,7 +92,6 @@ enum ccid3_hc_tx_states {
* @tx_t_ld: Time last doubled during slow start
* @tx_t_nom: Nominal send time of next packet
* @tx_hist: Packet history
- * @tx_options_received: Parsed set of retrieved options
*/
struct ccid3_hc_tx_sock {
u64 tx_x;
@@ -115,7 +109,6 @@ struct ccid3_hc_tx_sock {
ktime_t tx_t_ld;
ktime_t tx_t_nom;
struct tfrc_tx_hist_entry *tx_hist;
- struct ccid3_options_received tx_options_received;
};
static inline struct ccid3_hc_tx_sock *ccid3_hc_tx_sk(const struct sock *sk)
^ permalink raw reply
* [PATCH 4/5] dccp tfrc/ccid-3: computing the loss rate from the Loss Event Rate
From: Gerrit Renker @ 2010-09-21 10:28 UTC (permalink / raw)
To: davem; +Cc: dccp, netdev, Gerrit Renker
In-Reply-To: <1285064938-5933-4-git-send-email-gerrit@erg.abdn.ac.uk>
This adds a function to take care of the following, separate cases occurring in
the computation of the Loss Rate p:
* 1/(2^32-1) is mapped into 0% as per RFC 4342, 8.5;
* 1/0 is mapped into 100%, the maximum;
* to avoid that p = 1/x is rounded down to 0 when x is very large, since this
means accidentally re-entering slow-start indicated by p == 0, the minimum
resolution value of p is now returned instead;
* a bug in ccid3_hc_rx_getsockopt is fixed: 1/0 was mapped into ~0U.
Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
---
net/dccp/ccids/ccid3.c | 9 ++++-----
net/dccp/ccids/lib/tfrc.h | 1 +
net/dccp/ccids/lib/tfrc_equation.c | 14 ++++++++++++++
3 files changed, 19 insertions(+), 5 deletions(-)
--- a/net/dccp/ccids/ccid3.c
+++ b/net/dccp/ccids/ccid3.c
@@ -400,10 +400,10 @@ static void ccid3_hc_tx_packet_recv(struct sock *sk, struct sk_buff *skb)
/* Update loss event rate (which is scaled by 1e6) */
pinv = opt_recv->ccid3or_loss_event_rate;
- if (pinv == ~0U || pinv == 0) /* see RFC 4342, 8.5 */
+ if (pinv == 0)
hc->tx_p = 0;
- else /* can not exceed 100% */
- hc->tx_p = scaled_div(1, pinv);
+ else
+ hc->tx_p = tfrc_invert_loss_event_rate(pinv);
/*
* Update allowed sending rate X as per draft rfc3448bis-00, 4.2/3
@@ -834,8 +834,7 @@ static int ccid3_hc_rx_getsockopt(struct sock *sk, const int optname, int len,
return -EINVAL;
rx_info.tfrcrx_x_recv = hc->rx_x_recv;
rx_info.tfrcrx_rtt = hc->rx_rtt;
- rx_info.tfrcrx_p = hc->rx_pinv == 0 ? ~0U :
- scaled_div(1, hc->rx_pinv);
+ rx_info.tfrcrx_p = tfrc_invert_loss_event_rate(hc->rx_pinv);
len = sizeof(rx_info);
val = &rx_info;
break;
--- a/net/dccp/ccids/lib/tfrc.h
+++ b/net/dccp/ccids/lib/tfrc.h
@@ -57,6 +57,7 @@ static inline u32 tfrc_ewma(const u32 avg, const u32 newval, const u8 weight)
extern u32 tfrc_calc_x(u16 s, u32 R, u32 p);
extern u32 tfrc_calc_x_reverse_lookup(u32 fvalue);
+extern u32 tfrc_invert_loss_event_rate(u32 loss_event_rate);
extern int tfrc_tx_packet_history_init(void);
extern void tfrc_tx_packet_history_exit(void);
--- a/net/dccp/ccids/lib/tfrc_equation.c
+++ b/net/dccp/ccids/lib/tfrc_equation.c
@@ -687,3 +687,17 @@ u32 tfrc_calc_x_reverse_lookup(u32 fvalue)
index = tfrc_binsearch(fvalue, 0);
return (index + 1) * 1000000 / TFRC_CALC_X_ARRSIZE;
}
+
+/**
+ * tfrc_invert_loss_event_rate - Compute p so that 10^6 corresponds to 100%
+ * When @loss_event_rate is large, there is a chance that p is truncated to 0.
+ * To avoid re-entering slow-start in that case, we set p = TFRC_SMALLEST_P > 0.
+ */
+u32 tfrc_invert_loss_event_rate(u32 loss_event_rate)
+{
+ if (loss_event_rate == UINT_MAX) /* see RFC 4342, 8.5 */
+ return 0;
+ if (unlikely(loss_event_rate == 0)) /* map 1/0 into 100% */
+ return 1000000;
+ return max_t(u32, scaled_div(1, loss_event_rate), TFRC_SMALLEST_P);
+}
^ permalink raw reply
* [PATCH 2/5] dccp: Replace magic CCID-specific numbers by symbolic constants
From: Gerrit Renker @ 2010-09-21 10:28 UTC (permalink / raw)
To: davem; +Cc: dccp, netdev, Gerrit Renker
In-Reply-To: <1285064938-5933-2-git-send-email-gerrit@erg.abdn.ac.uk>
The constants DCCPO_{MIN,MAX}_CCID_SPECIFIC are nowhere used in the code, but
instead for the CCID-specific options numbers are used.
This patch unifies the use of CCID-specific option numbers, by adding symbolic
names reflecting the definitions in RFC 4340, 10.3.
Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
---
include/linux/dccp.h | 6 ++++--
net/dccp/options.c | 13 +++----------
2 files changed, 7 insertions(+), 12 deletions(-)
--- a/include/linux/dccp.h
+++ b/include/linux/dccp.h
@@ -165,8 +165,10 @@ enum {
DCCPO_TIMESTAMP_ECHO = 42,
DCCPO_ELAPSED_TIME = 43,
DCCPO_MAX = 45,
- DCCPO_MIN_CCID_SPECIFIC = 128,
- DCCPO_MAX_CCID_SPECIFIC = 255,
+ DCCPO_MIN_RX_CCID_SPECIFIC = 128, /* from sender to receiver */
+ DCCPO_MAX_RX_CCID_SPECIFIC = 191,
+ DCCPO_MIN_TX_CCID_SPECIFIC = 192, /* from receiver to sender */
+ DCCPO_MAX_TX_CCID_SPECIFIC = 255,
};
/* maximum size of a single TLV-encoded DCCP option (sans type/len bytes) */
#define DCCP_SINGLE_OPT_MAXLEN 253
--- a/net/dccp/options.c
+++ b/net/dccp/options.c
@@ -96,18 +96,11 @@ int dccp_parse_options(struct sock *sk, struct dccp_request_sock *dreq,
}
/*
- * CCID-Specific Options (from RFC 4340, sec. 10.3):
- *
- * Option numbers 128 through 191 are for options sent from the
- * HC-Sender to the HC-Receiver; option numbers 192 through 255
- * are for options sent from the HC-Receiver to the HC-Sender.
- *
* CCID-specific options are ignored during connection setup, as
* negotiation may still be in progress (see RFC 4340, 10.3).
* The same applies to Ack Vectors, as these depend on the CCID.
- *
*/
- if (dreq != NULL && (opt >= 128 ||
+ if (dreq != NULL && (opt >= DCCPO_MIN_RX_CCID_SPECIFIC ||
opt == DCCPO_ACK_VECTOR_0 || opt == DCCPO_ACK_VECTOR_1))
goto ignore_option;
@@ -226,12 +219,12 @@ int dccp_parse_options(struct sock *sk, struct dccp_request_sock *dreq,
dccp_pr_debug("%s rx opt: ELAPSED_TIME=%d\n",
dccp_role(sk), elapsed_time);
break;
- case 128 ... 191:
+ case DCCPO_MIN_RX_CCID_SPECIFIC ... DCCPO_MAX_RX_CCID_SPECIFIC:
if (ccid_hc_rx_parse_options(dp->dccps_hc_rx_ccid, sk,
pkt_type, opt, value, len))
goto out_invalid_option;
break;
- case 192 ... 255:
+ case DCCPO_MIN_TX_CCID_SPECIFIC ... DCCPO_MAX_TX_CCID_SPECIFIC:
if (ccid_hc_tx_parse_options(dp->dccps_hc_tx_ccid, sk,
pkt_type, opt, value, len))
goto out_invalid_option;
^ permalink raw reply
* net-next-2.6 [PATCH 0/5] dccp: miscellaneous CCID fixes and clean-up
From: Gerrit Renker @ 2010-09-21 10:28 UTC (permalink / raw)
To: davem; +Cc: dccp, netdev
In-Reply-To: <miscellaneous_dccp_ccid_fixes>
Dear Dave,
please find attached a tested set of 2 + 3 cleanup patches;
* the first two affect the DCCP CCID interface in general,
* the last three conclude the first set of CCID-3 patches.
Patch #1: allows the CCID option-parsing routines to see the packet type.
Patch #2: replaces magic CCID-specific numbers with symbolic constants.
Patch #3: removes a redundant CCID-3 state (thanks to Leandro Sales de Melo).
Patch #4: fixes several issues in computing the CCID-3 loss rate.
Patch #5: removes redundant CCID-3 TX 'options_received' struct.
The set has also been placed into a fresh (today's) copy of net-next-2.6, on
git://eden-feed.erg.abdn.ac.uk/net-next-2.6 [subtree 'dccp']
All patches have been tested and compile independently.
^ permalink raw reply
* Re: Fw: [Bug 18822] New: TCP Communications gets blocked, then resetted
From: Ilpo Järvinen @ 2010-09-21 10:18 UTC (permalink / raw)
To: dc6iq; +Cc: Stephen Hemminger, Netdev
In-Reply-To: <alpine.DEB.2.00.1009211216160.26447@wel-95.cs.helsinki.fi>
[-- Attachment #1: Type: TEXT/PLAIN, Size: 4295 bytes --]
Now with the original reported too among the receivers.
On Tue, 21 Sep 2010, Ilpo Järvinen wrote:
> On Mon, 20 Sep 2010, Stephen Hemminger wrote:
>
> > Begin forwarded message:
> >
> > Date: Mon, 20 Sep 2010 10:39:47 GMT
> > From: bugzilla-daemon@bugzilla.kernel.org
> > To: shemminger@linux-foundation.org
> > Subject: [Bug 18822] New: TCP Communications gets blocked, then resetted
> >
> >
> > https://bugzilla.kernel.org/show_bug.cgi?id=18822
> >
> > Summary: TCP Communications gets blocked, then resetted
> > Product: Networking
> > Version: 2.5
> > Kernel Version: 2.6.32-24-generic #43-Ubuntu
> > Platform: All
> > OS/Version: Linux
> > Tree: Mainline
> > Status: NEW
> > Severity: normal
> > Priority: P1
> > Component: IPV4
> > AssignedTo: shemminger@linux-foundation.org
> > ReportedBy: dc6iq@gmx.de
> > Regression: No
> >
> >
> > Created an attachment (id=30682)
> > --> (https://bugzilla.kernel.org/attachment.cgi?id=30682)
> > both machine dumps
> >
> > having a freshly installed bacula server i could never get a backup from my
> > working machine done to the bacula server.
> >
> > The Connection transmits approximately 3 GiB, then locks up and resets the
> > connection.
> >
> > Transmission gets stuck at a certain point, then the bacula server does not
> > reply to retransmitted packets on IPV4 stack. Retry Count on disks machine (my
> > working machine) raises up to 13, then the Connection is gone. bacula server
> > tries to send a Push packet (after KeepAlive timer runs out), and get the final
> > RST packet from disks, because the connection is gone.
> >
> > In the Attachment you will find the tcpdumps from both machines, actually the
> > sending machine dropped some packets in the dump.
>
> If you didn't already, try with -w directly into a binary file and then
> post process to textual input with -r.
>
> > It might be a possible help: disks is running an 64 bit kernel whereas bacula
> > is running 32 bit. I haven't looked into the option bits very well but it looks
> > like there is a problem hidden:
> >
> > last ack being ok:
> > 09:32:56.142876 IP bacula.elkenet.bacula-sd > disks.elkenet.50766: Flags [.],
> > ack 2005754588, win 9582, options [nop,nop,TS val 21825875 ecr 4530083], length
> > 0
> > next ack packet:
> > 09:32:56.144763 IP bacula.elkenet.bacula-sd > disks.elkenet.50766: Flags [.],
> > ack 2005773412, win 9308, options [nop,nop,TS val 21825876 ecr
> > 4530083,nop,nop,sack 1 {2005774860:2005776308}], length 0
>
> ...I fail to understand to what problem you're trying to point here with
> these two ACKs. Could you elaborate please (if you had something specific
> in mind)?
>
> I went throught the logs... I cannot go through all the checking done
> because tcpdump without enough -v's hides the sequence numbers for pure
> ACKs (09:32:56.150182 shows only the ack seqno, not the other seqno which
> also is used by the validator), I think you need at least -vvv to show
> them nowadays. The last new data segment at 09:32:56.150132 was still
> received as it is reported in SACK, only the retransmissions that
> follow are discarded.
>
> > root@disks:~# uname -a
> > Linux disks 2.6.32-24-generic #43-Ubuntu SMP Thu Sep 16 14:58:24 UTC 2010
> > x86_64 GNU/Linux
> >
> > root@bacula:~# uname -a
> > Linux bacula 2.6.32-24-generic-pae #43-Ubuntu SMP Thu Sep 16 15:30:27 UTC 2010
> > i686 GNU/Linux
> >
> > Doing a 20GB backup on a debian server works fine
> >
> > server:~# uname -a
> > Linux server 2.6.32-5-486 #1 Sat Sep 18 01:43:00 UTC 2010 i686 GNU/Linux
> >
> > Doing a 26 GB backup from a 32 bit Ubuntu works fine as well. Maybe its a 64
> > bit issue...
> >
> > root@elke:~# uname -a
> > Linux elke 2.6.32-24-generic #43-Ubuntu SMP Thu Sep 16 14:17:33 UTC 2010 i686
> > GNU/Linux
>
> Hmm, some ubuntu magic in these kernels.
>
> > If any further input is required, just let me know...
>
> MIBs might immediately tell what caused the segments between
> 09:32:56.150310 and 09:46:30.970140 to be discarded (e.g., take before
> and after snapshots of /proc/net/netstat and /proc/net/snmp and see what
> did increase).
>
> Any TSO enabled?
>
>
--
i.
^ permalink raw reply
* Re: [2.6.36-rc5] INET?: possible irq lock inversion dependency
From: Eric Dumazet @ 2010-09-21 10:13 UTC (permalink / raw)
To: Tetsuo Handa; +Cc: linux-fsdevel, netdev
In-Reply-To: <1285061757.2617.176.camel@edumazet-laptop>
Le mardi 21 septembre 2010 à 11:36 +0200, Eric Dumazet a écrit :
> Le mardi 21 septembre 2010 à 18:10 +0900, Tetsuo Handa a écrit :
> > Eric Dumazet wrote:
> > > I would say someone calls xs_tcp_state_change() while softirqs are not
> > > masked.
> > >
> > > If its expected, then following patch is needed ?
> >
> > After applying your patch, I got the same warning on different place.
> > I got this when doing "mount 127.0.0.1:/usr/src/ /mnt/".
>
> Hmm, but what are the changes in your tree ?
>
(Because I dont have these lockdep splats here with 2.6.36-rc5)
^ permalink raw reply
* Re: Regression, bisected: reference leak with IPSec since ~2.6.31
From: Eric Dumazet @ 2010-09-21 10:07 UTC (permalink / raw)
To: Jarek Poplawski; +Cc: Nick Bowler, linux-kernel, netdev, David S. Miller
In-Reply-To: <1285062948.2617.204.camel@edumazet-laptop>
Le mardi 21 septembre 2010 à 11:55 +0200, Eric Dumazet a écrit :
> Le mardi 21 septembre 2010 à 09:38 +0000, Jarek Poplawski a écrit :
> > On Tue, Sep 21, 2010 at 11:21:00AM +0200, Eric Dumazet wrote:
>
> > I hope you're right with this.
> >
> > >
> > > I liked the :
> > >
> > > <check everything without changing state>
> > > if something wrong
> > > goto slow_path
> > > else
> > > <OK, lets do destructive things>
> > >
> >
> > But it's an optimization of the "unlikely" case btw. ;-)
> >
>
> This is a bug fix, in a complex function.
>
> in -next branch, we can try to be smart, adding more code in slow_path
> to revert what was done in the hope fast path would be taken.
> And pray we dont add another bug :)
>
My initial patch was somehing like :
<loop to check everything is ok>
compute truesizes = sum (frag->truesize)
not modifying frag->sk/frag->destructor
if problem -> goto slow_path
else {
skb->truesize -= truesizes;
send(skb)
for each frag
set frag->sk = original_sk
set frag->destructor = sock_wfree
send(frag)
Problem was if we got an error while sending skb or one frag :
It was leaking some truesize accounting.
(an extra loop was needed to set frag->sk/frag->destructor for remaining
frags before freeing them)
So I added the extra loop to :
for each frag {
set frag->sk = skb->sk
set frag->destructor = sock_wfree
skb->truesize -= frag->truesize
}
It removed the need to compute sum(frag->truesize) in a temp variable
(truesizes)
^ permalink raw reply
* Re: Regression, bisected: reference leak with IPSec since ~2.6.31
From: Eric Dumazet @ 2010-09-21 9:55 UTC (permalink / raw)
To: Jarek Poplawski; +Cc: Nick Bowler, linux-kernel, netdev, David S. Miller
In-Reply-To: <20100921093853.GA8664@ff.dom.local>
Le mardi 21 septembre 2010 à 09:38 +0000, Jarek Poplawski a écrit :
> On Tue, Sep 21, 2010 at 11:21:00AM +0200, Eric Dumazet wrote:
> I hope you're right with this.
>
> >
> > I liked the :
> >
> > <check everything without changing state>
> > if something wrong
> > goto slow_path
> > else
> > <OK, lets do destructive things>
> >
>
> But it's an optimization of the "unlikely" case btw. ;-)
>
This is a bug fix, in a complex function.
in -next branch, we can try to be smart, adding more code in slow_path
to revert what was done in the hope fast path would be taken.
And pray we dont add another bug :)
^ permalink raw reply
* Re: Fw: [Bug 18822] New: TCP Communications gets blocked, then resetted
From: Ilpo Järvinen @ 2010-09-21 9:54 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: Netdev
In-Reply-To: <20100920090451.4ef28f91@nehalam>
On Mon, 20 Sep 2010, Stephen Hemminger wrote:
> Begin forwarded message:
>
> Date: Mon, 20 Sep 2010 10:39:47 GMT
> From: bugzilla-daemon@bugzilla.kernel.org
> To: shemminger@linux-foundation.org
> Subject: [Bug 18822] New: TCP Communications gets blocked, then resetted
>
>
> https://bugzilla.kernel.org/show_bug.cgi?id=18822
>
> Summary: TCP Communications gets blocked, then resetted
> Product: Networking
> Version: 2.5
> Kernel Version: 2.6.32-24-generic #43-Ubuntu
> Platform: All
> OS/Version: Linux
> Tree: Mainline
> Status: NEW
> Severity: normal
> Priority: P1
> Component: IPV4
> AssignedTo: shemminger@linux-foundation.org
> ReportedBy: dc6iq@gmx.de
> Regression: No
>
>
> Created an attachment (id=30682)
> --> (https://bugzilla.kernel.org/attachment.cgi?id=30682)
> both machine dumps
>
> having a freshly installed bacula server i could never get a backup from my
> working machine done to the bacula server.
>
> The Connection transmits approximately 3 GiB, then locks up and resets the
> connection.
>
> Transmission gets stuck at a certain point, then the bacula server does not
> reply to retransmitted packets on IPV4 stack. Retry Count on disks machine (my
> working machine) raises up to 13, then the Connection is gone. bacula server
> tries to send a Push packet (after KeepAlive timer runs out), and get the final
> RST packet from disks, because the connection is gone.
>
> In the Attachment you will find the tcpdumps from both machines, actually the
> sending machine dropped some packets in the dump.
If you didn't already, try with -w directly into a binary file and then
post process to textual input with -r.
> It might be a possible help: disks is running an 64 bit kernel whereas bacula
> is running 32 bit. I haven't looked into the option bits very well but it looks
> like there is a problem hidden:
>
> last ack being ok:
> 09:32:56.142876 IP bacula.elkenet.bacula-sd > disks.elkenet.50766: Flags [.],
> ack 2005754588, win 9582, options [nop,nop,TS val 21825875 ecr 4530083], length
> 0
> next ack packet:
> 09:32:56.144763 IP bacula.elkenet.bacula-sd > disks.elkenet.50766: Flags [.],
> ack 2005773412, win 9308, options [nop,nop,TS val 21825876 ecr
> 4530083,nop,nop,sack 1 {2005774860:2005776308}], length 0
...I fail to understand to what problem you're trying to point here with
these two ACKs. Could you elaborate please (if you had something specific
in mind)?
I went throught the logs... I cannot go through all the checking done
because tcpdump without enough -v's hides the sequence numbers for pure
ACKs (09:32:56.150182 shows only the ack seqno, not the other seqno which
also is used by the validator), I think you need at least -vvv to show
them nowadays. The last new data segment at 09:32:56.150132 was still
received as it is reported in SACK, only the retransmissions that
follow are discarded.
> root@disks:~# uname -a
> Linux disks 2.6.32-24-generic #43-Ubuntu SMP Thu Sep 16 14:58:24 UTC 2010
> x86_64 GNU/Linux
>
> root@bacula:~# uname -a
> Linux bacula 2.6.32-24-generic-pae #43-Ubuntu SMP Thu Sep 16 15:30:27 UTC 2010
> i686 GNU/Linux
>
> Doing a 20GB backup on a debian server works fine
>
> server:~# uname -a
> Linux server 2.6.32-5-486 #1 Sat Sep 18 01:43:00 UTC 2010 i686 GNU/Linux
>
> Doing a 26 GB backup from a 32 bit Ubuntu works fine as well. Maybe its a 64
> bit issue...
>
> root@elke:~# uname -a
> Linux elke 2.6.32-24-generic #43-Ubuntu SMP Thu Sep 16 14:17:33 UTC 2010 i686
> GNU/Linux
Hmm, some ubuntu magic in these kernels.
> If any further input is required, just let me know...
MIBs might immediately tell what caused the segments between
09:32:56.150310 and 09:46:30.970140 to be discarded (e.g., take before
and after snapshots of /proc/net/netstat and /proc/net/snmp and see what
did increase).
Any TSO enabled?
--
i.
^ permalink raw reply
* Re: Regression, bisected: reference leak with IPSec since ~2.6.31
From: Jarek Poplawski @ 2010-09-21 9:38 UTC (permalink / raw)
To: Eric Dumazet; +Cc: Nick Bowler, linux-kernel, netdev, David S. Miller
In-Reply-To: <1285060860.2617.158.camel@edumazet-laptop>
On Tue, Sep 21, 2010 at 11:21:00AM +0200, Eric Dumazet wrote:
> Le mardi 21 septembre 2010 ?? 09:12 +0000, Jarek Poplawski a écrit :
> > On 2010-09-20 23:31, Eric Dumazet wrote:
> > ...
> > > @@ -510,11 +509,13 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
> > > goto slow_path;
> > >
> > > BUG_ON(frag->sk);
> > > - if (skb->sk) {
> > > + }
> > > + if (skb->sk) {
> > > + skb_walk_frags(skb, frag) {
> > > frag->sk = skb->sk;
> > > frag->destructor = sock_wfree;
> >
> > Nice catch, but it seems doing it in the first loop as now, and
> > reverting changes before goto slow_path might be more optimal here.
> >
>
> I thought of this, but found this function already very complex.
>
> Once everything is in cpu caches, the added loop is very cheap.
I hope you're right with this.
>
> I liked the :
>
> <check everything without changing state>
> if something wrong
> goto slow_path
> else
> <OK, lets do destructive things>
>
But it's an optimization of the "unlikely" case btw. ;-)
Jarek P.
^ permalink raw reply
* Re: [2.6.36-rc5] INET?: possible irq lock inversion dependency
From: Eric Dumazet @ 2010-09-21 9:35 UTC (permalink / raw)
To: Tetsuo Handa; +Cc: linux-fsdevel, netdev
In-Reply-To: <201009210910.o8L9Anaf071504@www262.sakura.ne.jp>
Le mardi 21 septembre 2010 à 18:10 +0900, Tetsuo Handa a écrit :
> Eric Dumazet wrote:
> > I would say someone calls xs_tcp_state_change() while softirqs are not
> > masked.
> >
> > If its expected, then following patch is needed ?
>
> After applying your patch, I got the same warning on different place.
> I got this when doing "mount 127.0.0.1:/usr/src/ /mnt/".
Hmm, but what are the changes in your tree ?
--
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply
* RE: [PATCH] qlcnic: dont assume NET_IP_ALIGN is 2
From: Eric Dumazet @ 2010-09-21 9:23 UTC (permalink / raw)
To: Amit Salecha
Cc: David Miller, netdev@vger.kernel.org, Ameen Rahman,
Anirban Chakraborty
In-Reply-To: <99737F4847ED0A48AECC9F4A1974A4B80F86F80278@MNEXMB2.qlogic.org>
Le mardi 21 septembre 2010 à 03:41 -0500, Amit Salecha a écrit :
> > Amit, if you believe this is a problem, you should address it for all
> > NICS, not only qlcnic.
> >
> > Qlcnic was lying to stack, because it consumed 2Kbytes blocs and
> > pretended they were consuming skb->len bytes.
> > (assuming MTU=1500, problem is worse if MTU is bigger)
> >
> > So in order to improve "throughput", you were allowing for memory
> > exhaust and freeze of the _machine_ ?
> >
> This won't lead to such problem. truesize is used for accounting only.
You must have big machines in your lab and never hit OOM ?
You really should take a look on various files in net/core, net/ipv4
trees. And files like "/proc/sys/net/tcp_mem", "/proc/sys/net/udp_mem"
In fact, truesize is _underestimated_ : (we dont account for struct
skb_shared_info) and kmalloc() rounding
We probably should use this patch (without having to check all possible
net drivers !)
Problem is this would slow down alloc_skb(), so this patch is not for
inclusion.
cheap alternative would be to use
size + sizeof(struct sk_buff) + SKB_DATA_ALIGN(sizeof(struct skb_shared_info))
If you think about it, when 128bit arches come, truesize will grow anyway.
If some tuning is needed in our stack, we'll do it.
(socket api SO_RCVBUF/ SO_SNDBUF is the problem, because
applications are not aware of packetization or kernel internals)
SOCK_MIN_RCVBUF is way too small, since sizeof(struct sk_buff)
is already close to 256. I guess we cannot even receive a single frame.
include/net/sock.h | 2 +-
net/core/skbuff.c | 2 +-
net/core/sock.c | 8 ++++----
3 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/include/net/sock.h b/include/net/sock.h
index 8ae97c4..348fc9e 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -1558,7 +1558,7 @@ static inline void sk_wake_async(struct sock *sk, int how, int band)
}
#define SOCK_MIN_SNDBUF 2048
-#define SOCK_MIN_RCVBUF 256
+#define SOCK_MIN_RCVBUF 1024
static inline void sk_stream_moderate_sndbuf(struct sock *sk)
{
diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index 752c197..5ab2e8e 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -196,7 +196,7 @@ struct sk_buff *__alloc_skb(unsigned int size, gfp_t gfp_mask,
* the tail pointer in struct sk_buff!
*/
memset(skb, 0, offsetof(struct sk_buff, tail));
- skb->truesize = size + sizeof(struct sk_buff);
+ skb->truesize = ksize(data) + sizeof(struct sk_buff);
atomic_set(&skb->users, 1);
skb->head = data;
skb->data = data;
diff --git a/net/core/sock.c b/net/core/sock.c
index f3a06c4..803e041 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -535,10 +535,10 @@ int sock_setsockopt(struct socket *sock, int level, int optname,
val = sysctl_wmem_max;
set_sndbuf:
sk->sk_userlocks |= SOCK_SNDBUF_LOCK;
- if ((val * 2) < SOCK_MIN_SNDBUF)
+ if ((val * 4) < SOCK_MIN_SNDBUF)
sk->sk_sndbuf = SOCK_MIN_SNDBUF;
else
- sk->sk_sndbuf = val * 2;
+ sk->sk_sndbuf = val * 4;
/*
* Wake up sending tasks if we
@@ -579,10 +579,10 @@ set_rcvbuf:
* returning the value we actually used in getsockopt
* is the most desirable behavior.
*/
- if ((val * 2) < SOCK_MIN_RCVBUF)
+ if ((val * 4) < SOCK_MIN_RCVBUF)
sk->sk_rcvbuf = SOCK_MIN_RCVBUF;
else
- sk->sk_rcvbuf = val * 2;
+ sk->sk_rcvbuf = val * 4;
break;
case SO_RCVBUFFORCE:
^ permalink raw reply related
* Re: Regression, bisected: reference leak with IPSec since ~2.6.31
From: Eric Dumazet @ 2010-09-21 9:21 UTC (permalink / raw)
To: Jarek Poplawski; +Cc: Nick Bowler, linux-kernel, netdev, David S. Miller
In-Reply-To: <20100921091248.GA8424@ff.dom.local>
Le mardi 21 septembre 2010 à 09:12 +0000, Jarek Poplawski a écrit :
> On 2010-09-20 23:31, Eric Dumazet wrote:
> ...
> > [PATCH] ip : fix truesize mismatch in ip fragmentation
> >
> > We should not set frag->destructor to sock_wkfree() until we are sure we
> > dont hit slow path in ip_fragment(). Or we risk uncharging
> > frag->truesize twice, and in the end, having negative socket
> > sk_wmem_alloc counter, or even freeing socket sooner than expected.
> >
> > Many thanks to Nick Bowler, who provided a very clean bug report and
> > test programs.
> >
> > While Nick bisection pointed to commit 2b85a34e911bf483 (net: No more
> > expensive sock_hold()/sock_put() on each tx), underlying bug is older.
> >
> > Reported-and-bisected-by: Nick Bowler <nbowler@elliptictech.com>
> > Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
> > ---
> > net/ipv4/ip_output.c | 8 ++++----
> > net/ipv6/ip6_output.c | 10 +++++-----
> > 2 files changed, 9 insertions(+), 9 deletions(-)
> >
> > diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
> > index 04b6989..126d9b3 100644
> > --- a/net/ipv4/ip_output.c
> > +++ b/net/ipv4/ip_output.c
> > @@ -490,7 +490,6 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
> > if (skb_has_frags(skb)) {
> > struct sk_buff *frag;
> > int first_len = skb_pagelen(skb);
> > - int truesizes = 0;
> >
> > if (first_len - hlen > mtu ||
> > ((first_len - hlen) & 7) ||
> > @@ -510,11 +509,13 @@ int ip_fragment(struct sk_buff *skb, int (*output)(struct sk_buff *))
> > goto slow_path;
> >
> > BUG_ON(frag->sk);
> > - if (skb->sk) {
> > + }
> > + if (skb->sk) {
> > + skb_walk_frags(skb, frag) {
> > frag->sk = skb->sk;
> > frag->destructor = sock_wfree;
>
> Nice catch, but it seems doing it in the first loop as now, and
> reverting changes before goto slow_path might be more optimal here.
>
I thought of this, but found this function already very complex.
Once everything is in cpu caches, the added loop is very cheap.
I liked the :
<check everything without changing state>
if something wrong
goto slow_path
else
<OK, lets do destructive things>
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox