* Re: [PATCH net-next-2.6] net: build_ehash_secret() and rt_bind_peer() cleanups
From: Changli Gao @ 2010-08-19 15:58 UTC (permalink / raw)
To: Eric Dumazet; +Cc: David Miller, netdev, Mathieu Desnoyers
In-Reply-To: <1282232815.2549.61.camel@edumazet-laptop>
On Thu, Aug 19, 2010 at 11:46 PM, Eric Dumazet <eric.dumazet@gmail.com> wrote:
> Now cmpxchg() is available on all arches, we can use it in
> build_ehash_secret() and rt_bind_peer() instead of using spinlocks.
>
> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
> CC: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
> ---
> net/ipv4/af_inet.c | 8 +++-----
> net/ipv4/route.c | 11 +++--------
> 2 files changed, 6 insertions(+), 13 deletions(-)
>
> diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c
> index 6a1100c..f581f77 100644
> --- a/net/ipv4/af_inet.c
> +++ b/net/ipv4/af_inet.c
> @@ -227,18 +227,16 @@ EXPORT_SYMBOL(inet_ehash_secret);
>
> /*
> * inet_ehash_secret must be set exactly once
> - * Instead of using a dedicated spinlock, we (ab)use inetsw_lock
> */
> void build_ehash_secret(void)
> {
> u32 rnd;
> +
> do {
> get_random_bytes(&rnd, sizeof(rnd));
> } while (rnd == 0);
> - spin_lock_bh(&inetsw_lock);
> - if (!inet_ehash_secret)
> - inet_ehash_secret = rnd;
> - spin_unlock_bh(&inetsw_lock);
> +
> + cmpxchg(&inet_ehash_secret, 0, rnd);
> }
> EXPORT_SYMBOL(build_ehash_secret);
>
> diff --git a/net/ipv4/route.c b/net/ipv4/route.c
> index 3f56b6e..ae3dba7 100644
> --- a/net/ipv4/route.c
> +++ b/net/ipv4/route.c
> @@ -1268,18 +1268,13 @@ skip_hashing:
>
> void rt_bind_peer(struct rtable *rt, int create)
> {
> - static DEFINE_SPINLOCK(rt_peer_lock);
> struct inet_peer *peer;
>
> peer = inet_getpeer(rt->rt_dst, create);
>
> - spin_lock_bh(&rt_peer_lock);
> - if (rt->peer == NULL) {
> - rt->peer = peer;
> - peer = NULL;
> - }
> - spin_unlock_bh(&rt_peer_lock);
> - if (peer)
> + peer = cmpxchg(&rt->peer, NULL, peer);
> +
> + if (unlikely(peer))
It isn't correct, and should be
if (unlikely(cmpxchg(&rt->peer, NULL, peer) != NULL))
> inet_putpeer(peer);
> }
--
Regards,
Changli Gao(xiaosuo@gmail.com)
^ permalink raw reply
* Re: [PATCH 1/5] ptp: Added a brand new class driver for ptp clocks.
From: Arnd Bergmann @ 2010-08-19 15:48 UTC (permalink / raw)
To: Ira W. Snyder
Cc: Richard Cochran, Rodolfo Giometti, john stultz,
devicetree-discuss, linux-kernel, netdev, linuxppc-dev,
linux-arm-kernel, Krzysztof Halasa
In-Reply-To: <20100819152301.GB25193@ovro.caltech.edu>
On Thursday 19 August 2010, Ira W. Snyder wrote:
> Perhaps you were thinking of the vhost example (taken from
> drivers/vhost/Kconfig):
>
> config VHOST_NET
> tristate "Host kernel accelerator for virtio net (EXPERIMENTAL)"
> depends on NET && EVENTFD && (TUN || !TUN) && (MACVTAP || !MACVTAP) && EXPERIMENTAL
>
> They have a similar construct with both TUN and MACVTAP there. Perhaps
> the parens are a necessary part of the "X || !X" syntax? Just a random
> guess.
Yes, that's the one I was thinking of.
My mistake was that the effect is slightly different here. VHOST and TUN are
both tristate. What we guarantee here is that if TUN is "m", VHOST cannot be "y",
because its dependency cannot be fulfilled for y.
Arnd
^ permalink raw reply
* [PATCH net-next-2.6] net: build_ehash_secret() and rt_bind_peer() cleanups
From: Eric Dumazet @ 2010-08-19 15:46 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Mathieu Desnoyers
Now cmpxchg() is available on all arches, we can use it in
build_ehash_secret() and rt_bind_peer() instead of using spinlocks.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
CC: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
---
net/ipv4/af_inet.c | 8 +++-----
net/ipv4/route.c | 11 +++--------
2 files changed, 6 insertions(+), 13 deletions(-)
diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c
index 6a1100c..f581f77 100644
--- a/net/ipv4/af_inet.c
+++ b/net/ipv4/af_inet.c
@@ -227,18 +227,16 @@ EXPORT_SYMBOL(inet_ehash_secret);
/*
* inet_ehash_secret must be set exactly once
- * Instead of using a dedicated spinlock, we (ab)use inetsw_lock
*/
void build_ehash_secret(void)
{
u32 rnd;
+
do {
get_random_bytes(&rnd, sizeof(rnd));
} while (rnd == 0);
- spin_lock_bh(&inetsw_lock);
- if (!inet_ehash_secret)
- inet_ehash_secret = rnd;
- spin_unlock_bh(&inetsw_lock);
+
+ cmpxchg(&inet_ehash_secret, 0, rnd);
}
EXPORT_SYMBOL(build_ehash_secret);
diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index 3f56b6e..ae3dba7 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -1268,18 +1268,13 @@ skip_hashing:
void rt_bind_peer(struct rtable *rt, int create)
{
- static DEFINE_SPINLOCK(rt_peer_lock);
struct inet_peer *peer;
peer = inet_getpeer(rt->rt_dst, create);
- spin_lock_bh(&rt_peer_lock);
- if (rt->peer == NULL) {
- rt->peer = peer;
- peer = NULL;
- }
- spin_unlock_bh(&rt_peer_lock);
- if (peer)
+ peer = cmpxchg(&rt->peer, NULL, peer);
+
+ if (unlikely(peer))
inet_putpeer(peer);
}
^ permalink raw reply related
* Re: [PATCH v2] netfilter: save the hash of the tuple in the original direction for latter use
From: Eric Dumazet @ 2010-08-19 15:46 UTC (permalink / raw)
To: Changli Gao
Cc: Mathieu Desnoyers, Patrick McHardy, David S. Miller,
netfilter-devel, netdev
In-Reply-To: <AANLkTim5wn0Y-+081gvm7FkWx3NioRvtRMWy2uF4nji2@mail.gmail.com>
Le jeudi 19 août 2010 à 23:35 +0800, Changli Gao a écrit :
>
> In fact, I don't know the reason clearly. This code is derived from
> the older one. Maybe there isn't enough entropy when initializing.
>
Yes, we wanted to initialize these random data as late as possible, that
is when the first entry is added in conntrack hash.
--
To unsubscribe from this list: send the line "unsubscribe netfilter-devel" 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 v2] netfilter: save the hash of the tuple in the original direction for latter use
From: yao zhao @ 2010-08-19 15:41 UTC (permalink / raw)
To: Changli Gao
Cc: Mathieu Desnoyers, Patrick McHardy, David S. Miller, Eric Dumazet,
netfilter-devel, netdev
In-Reply-To: <AANLkTim5wn0Y-+081gvm7FkWx3NioRvtRMWy2uF4nji2@mail.gmail.com>
On Thu, Aug 19, 2010 at 11:35 AM, Changli Gao <xiaosuo@gmail.com> wrote:
> On Thu, Aug 19, 2010 at 11:21 PM, Mathieu Desnoyers
> <mathieu.desnoyers@polymtl.ca> wrote:
>> * Changli Gao (xiaosuo@gmail.com) wrote:
>>> Since we don't change the tuple in the original direction, we can save it
>>> in ct->tuplehash[IP_CT_DIR_REPLY].hnode.pprev for __nf_conntrack_confirm()
>>> use.
>>>
>>> __hash_conntrack() is split into two steps: ____hash_conntrack() is used
>>> to get the raw hash, and __hash_bucket() is used to get the bucket id.
>>>
>>> In SYN-flood case, early_drop() doesn't need to recompute the hash again.
>>>
>>> Signed-off-by: Changli Gao <xiaosuo@gmail.com>
>>> ---
>>> v2: use cmpxchg() to save 2 variables.
>>> net/netfilter/nf_conntrack_core.c | 114 ++++++++++++++++++++++++++------------
>>> 1 file changed, 79 insertions(+), 35 deletions(-)
>>> diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
>>> index df3eedb..0e02205 100644
>>> --- a/net/netfilter/nf_conntrack_core.c
>>> +++ b/net/netfilter/nf_conntrack_core.c
>>> @@ -65,14 +65,20 @@ EXPORT_SYMBOL_GPL(nf_conntrack_max);
>>> DEFINE_PER_CPU(struct nf_conn, nf_conntrack_untracked);
>>> EXPORT_PER_CPU_SYMBOL(nf_conntrack_untracked);
>>>
>>> -static int nf_conntrack_hash_rnd_initted;
>>> -static unsigned int nf_conntrack_hash_rnd;
>>> -
>>> -static u_int32_t __hash_conntrack(const struct nf_conntrack_tuple *tuple,
>>> - u16 zone, unsigned int size, unsigned int rnd)
>>> +static u32 ____hash_conntrack(const struct nf_conntrack_tuple *tuple, u16 zone)
>>> {
>>> unsigned int n;
>>> u_int32_t h;
>>> + static unsigned int rnd;
>>
>> Why are you putting a hidden static variable here ? It should probably
>> be declared outside of the function scope.
>
> It is only used in this function.
Looks like it is only initialized once to rnd then it should go a "init".
>
>>
>>> +
>>> + if (unlikely(!rnd)) {
>>> + unsigned int rand;
>>> +
>>> + get_random_bytes(&rand, sizeof(rand));
>>> + if (!rand)
>>> + rand = 1;
>>> + cmpxchg(&rnd, 0, rand);
>>
>> This really belongs to a "init()" function, not in a dynamic check in
>> the __hash_conntrack hot path. If you do that a init time, then you
>> don't even need the cmpxchg.
>>
>> Or maybe I completely misunderstand you goal here; maybe you are really
>> trying to re-calculate random bytes. But more explanation is called for,
>> because I really don't see where the value is brought back to 0.
>>
>
> In fact, I don't know the reason clearly. This code is derived from
> the older one. Maybe there isn't enough entropy when initializing.
>
> --
> Regards,
> Changli Gao(xiaosuo@gmail.com)
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
yao
--
To unsubscribe from this list: send the line "unsubscribe netfilter-devel" 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 1/5] ptp: Added a brand new class driver for ptp clocks.
From: Richard Cochran @ 2010-08-19 15:38 UTC (permalink / raw)
To: Arnd Bergmann
Cc: john stultz, netdev, linux-kernel, linuxppc-dev,
devicetree-discuss, linux-arm-kernel, Krzysztof Halasa,
Rodolfo Giometti
In-Reply-To: <201008191428.04546.arnd@arndb.de>
On Thu, Aug 19, 2010 at 02:28:04PM +0200, Arnd Bergmann wrote:
> My point was that a syscall is better than an ioctl based interface here,
> which I definitely still think. Given that John knows much more about
> clocks than I do, we still need to get agreement on the question that
> he raised, which is whether we actually need to expose this clock to the
> user or not.
>
> If we can find a way to sync system time accurate enough with PTP and
> PPS, user applications may not need to see two separate clocks at all.
At the very least, one user application (the PTPd) needs to see the
PTP clock.
> > SYSCALL_DEFINE3(clock_adjtime, const clockid_t, clkid,
> > int, ppb, struct timespec __user *, ts)
> >
> > ppb - desired frequency adjustment in parts per billion
> > ts - desired time step (or jump) in <sec,nsec> to correct
> > a measured offset
> >
> > Arguably, this syscall might be useful for other clocks, too.
>
> This is a mix of adjtime and adjtimex with the addition of
> the clkid parameter, right?
Sort of, but not really. ADJTIME(3) takes an offset and slowly
corrects the clock using a servo in the kernel, over hours.
For this function, the offset passed in the 'ts' parameter will be
immediately corrected, by jumping to the new time. This reflects the
way that PTP works. After the first few samples, the PTPd has an
estimate of the offset to the master and the rate difference. The PTPd
can proceed in one of two ways.
1. If resetting the clock is not desired, then the clock is set to the
maximum adjustment (in the right direction) until the clock time is
close to the master's time.
2. The estimated offset is added to the current time, resulting in a
jump in time.
We need clock_adjtime(id, 0, ts) for the second case.
> Have you considered passing a struct timex instead of ppb and ts?
Yes, but the timex is not suitable, IMHO.
> Is using ppb instead of the timex ppm required to get the accuracy
> you want?
That is one very good reason.
Another is this: can you explain what the 20+ fields mean?
Consider the field, freq. The comment says "frequency offset (scaled
ppm)." To what is it scaled? The only way I know of to find out is to
read the NTP code (which is fairly complex) and see what the unit
really is meant to be. Ditto for the other fields.
The timex structure reveals, AFAICT, the inner workings of the kernel
clock servo. For PTP, we don't need or want the kernel servo. The PTPd
has its own clock servo in user space.
Richard
^ permalink raw reply
* Re: [PATCH v2] netfilter: save the hash of the tuple in the original direction for latter use
From: Changli Gao @ 2010-08-19 15:35 UTC (permalink / raw)
To: Mathieu Desnoyers
Cc: Patrick McHardy, David S. Miller, Eric Dumazet, netfilter-devel,
netdev
In-Reply-To: <20100819152149.GA19744@Krystal>
On Thu, Aug 19, 2010 at 11:21 PM, Mathieu Desnoyers
<mathieu.desnoyers@polymtl.ca> wrote:
> * Changli Gao (xiaosuo@gmail.com) wrote:
>> Since we don't change the tuple in the original direction, we can save it
>> in ct->tuplehash[IP_CT_DIR_REPLY].hnode.pprev for __nf_conntrack_confirm()
>> use.
>>
>> __hash_conntrack() is split into two steps: ____hash_conntrack() is used
>> to get the raw hash, and __hash_bucket() is used to get the bucket id.
>>
>> In SYN-flood case, early_drop() doesn't need to recompute the hash again.
>>
>> Signed-off-by: Changli Gao <xiaosuo@gmail.com>
>> ---
>> v2: use cmpxchg() to save 2 variables.
>> net/netfilter/nf_conntrack_core.c | 114 ++++++++++++++++++++++++++------------
>> 1 file changed, 79 insertions(+), 35 deletions(-)
>> diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
>> index df3eedb..0e02205 100644
>> --- a/net/netfilter/nf_conntrack_core.c
>> +++ b/net/netfilter/nf_conntrack_core.c
>> @@ -65,14 +65,20 @@ EXPORT_SYMBOL_GPL(nf_conntrack_max);
>> DEFINE_PER_CPU(struct nf_conn, nf_conntrack_untracked);
>> EXPORT_PER_CPU_SYMBOL(nf_conntrack_untracked);
>>
>> -static int nf_conntrack_hash_rnd_initted;
>> -static unsigned int nf_conntrack_hash_rnd;
>> -
>> -static u_int32_t __hash_conntrack(const struct nf_conntrack_tuple *tuple,
>> - u16 zone, unsigned int size, unsigned int rnd)
>> +static u32 ____hash_conntrack(const struct nf_conntrack_tuple *tuple, u16 zone)
>> {
>> unsigned int n;
>> u_int32_t h;
>> + static unsigned int rnd;
>
> Why are you putting a hidden static variable here ? It should probably
> be declared outside of the function scope.
It is only used in this function.
>
>> +
>> + if (unlikely(!rnd)) {
>> + unsigned int rand;
>> +
>> + get_random_bytes(&rand, sizeof(rand));
>> + if (!rand)
>> + rand = 1;
>> + cmpxchg(&rnd, 0, rand);
>
> This really belongs to a "init()" function, not in a dynamic check in
> the __hash_conntrack hot path. If you do that a init time, then you
> don't even need the cmpxchg.
>
> Or maybe I completely misunderstand you goal here; maybe you are really
> trying to re-calculate random bytes. But more explanation is called for,
> because I really don't see where the value is brought back to 0.
>
In fact, I don't know the reason clearly. This code is derived from
the older one. Maybe there isn't enough entropy when initializing.
--
Regards,
Changli Gao(xiaosuo@gmail.com)
^ permalink raw reply
* Re: [PATCH 1/5] ptp: Added a brand new class driver for ptp clocks.
From: Ira W. Snyder @ 2010-08-19 15:23 UTC (permalink / raw)
To: Arnd Bergmann
Cc: Richard Cochran, Rodolfo Giometti, john stultz,
devicetree-discuss, linux-kernel, netdev, linuxppc-dev,
linux-arm-kernel, Krzysztof Halasa
In-Reply-To: <201008191429.50008.arnd@arndb.de>
On Thu, Aug 19, 2010 at 02:29:49PM +0200, Arnd Bergmann wrote:
> On Thursday 19 August 2010, Richard Cochran wrote:
> >
> > On Wed, Aug 18, 2010 at 05:02:03PM +0200, Arnd Bergmann wrote:
> > > You might want to use callbacks for these system calls that you
> > > can register to at module load time, like it is done for the
> > > existing syscalls.
> >
> > Can you point me to a specific example?
>
> The struct k_clock contains callback functions to clock source
> specific implementations of the syscalls and other functions.
> Adding a new clock source that is (potentially) implemented
> as a module means you have to define a new k_clock in that module
> that you register using register_posix_clock.
>
> If you need additional syscalls to be implemented by the same
> module, you can put more callback functions into struct k_clock
> that would then only be implemented by the new module.
>
> > > The simpler way (e.g. for testing) is using Kconfig dependencies, like
> > >
> > > config PTP
> > > bool "IEEE 1588 Precision Time Protocol"
> > >
> > > config PPS
> > > tristate "Pulse per Second"
> > > depends on PTP || !PTP
> > >
> > > The depends statement is a way of expressing that when PTP is enabled,
> > > PPS cannot be a module, while it may be a module if PTP is disabled.
> >
> > THis did not work for me.
> >
> > What I got was, in effect, two independent booleans.
>
> Hmm, right. I wonder what I was thinking of then.
>
> You can certainly do
>
> config PTP
> bool "IEEE 1588 Precision Time Protocol"
> depends on PPS != m
>
> but that will hide the PTP option if PPS is set to 'm', which is normally
> not what you want.
>
Arnd,
Perhaps you were thinking of the vhost example (taken from
drivers/vhost/Kconfig):
config VHOST_NET
tristate "Host kernel accelerator for virtio net (EXPERIMENTAL)"
depends on NET && EVENTFD && (TUN || !TUN) && (MACVTAP || !MACVTAP) && EXPERIMENTAL
They have a similar construct with both TUN and MACVTAP there. Perhaps
the parens are a necessary part of the "X || !X" syntax? Just a random
guess.
Ira
^ permalink raw reply
* Re: [PATCH v2] netfilter: save the hash of the tuple in the original direction for latter use
From: Mathieu Desnoyers @ 2010-08-19 15:21 UTC (permalink / raw)
To: Changli Gao
Cc: Patrick McHardy, David S. Miller, Eric Dumazet, netfilter-devel,
netdev
In-Reply-To: <1282228389-3029-1-git-send-email-xiaosuo@gmail.com>
* Changli Gao (xiaosuo@gmail.com) wrote:
> Since we don't change the tuple in the original direction, we can save it
> in ct->tuplehash[IP_CT_DIR_REPLY].hnode.pprev for __nf_conntrack_confirm()
> use.
>
> __hash_conntrack() is split into two steps: ____hash_conntrack() is used
> to get the raw hash, and __hash_bucket() is used to get the bucket id.
>
> In SYN-flood case, early_drop() doesn't need to recompute the hash again.
>
> Signed-off-by: Changli Gao <xiaosuo@gmail.com>
> ---
> v2: use cmpxchg() to save 2 variables.
> net/netfilter/nf_conntrack_core.c | 114 ++++++++++++++++++++++++++------------
> 1 file changed, 79 insertions(+), 35 deletions(-)
> diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
> index df3eedb..0e02205 100644
> --- a/net/netfilter/nf_conntrack_core.c
> +++ b/net/netfilter/nf_conntrack_core.c
> @@ -65,14 +65,20 @@ EXPORT_SYMBOL_GPL(nf_conntrack_max);
> DEFINE_PER_CPU(struct nf_conn, nf_conntrack_untracked);
> EXPORT_PER_CPU_SYMBOL(nf_conntrack_untracked);
>
> -static int nf_conntrack_hash_rnd_initted;
> -static unsigned int nf_conntrack_hash_rnd;
> -
> -static u_int32_t __hash_conntrack(const struct nf_conntrack_tuple *tuple,
> - u16 zone, unsigned int size, unsigned int rnd)
> +static u32 ____hash_conntrack(const struct nf_conntrack_tuple *tuple, u16 zone)
> {
> unsigned int n;
> u_int32_t h;
> + static unsigned int rnd;
Why are you putting a hidden static variable here ? It should probably
be declared outside of the function scope.
> +
> + if (unlikely(!rnd)) {
> + unsigned int rand;
> +
> + get_random_bytes(&rand, sizeof(rand));
> + if (!rand)
> + rand = 1;
> + cmpxchg(&rnd, 0, rand);
This really belongs to a "init()" function, not in a dynamic check in
the __hash_conntrack hot path. If you do that a init time, then you
don't even need the cmpxchg.
Or maybe I completely misunderstand you goal here; maybe you are really
trying to re-calculate random bytes. But more explanation is called for,
because I really don't see where the value is brought back to 0.
Thanks,
Mathieu
> + }
>
> /* The direction must be ignored, so we hash everything up to the
> * destination ports (which is a multiple of 4) and treat the last
> @@ -83,14 +89,29 @@ static u_int32_t __hash_conntrack(const struct nf_conntrack_tuple *tuple,
> zone ^ rnd ^ (((__force __u16)tuple->dst.u.all << 16) |
> tuple->dst.protonum));
>
> - return ((u64)h * size) >> 32;
> + return h;
> +}
> +
> +static u32 __hash_bucket(u32 __hash, unsigned int size)
> +{
> + return ((u64)__hash * size) >> 32;
> +}
> +
> +static u32 hash_bucket(u32 __hash, const struct net *net)
> +{
> + return __hash_bucket(__hash, net->ct.htable_size);
> +}
> +
> +static u_int32_t __hash_conntrack(const struct nf_conntrack_tuple *tuple,
> + u16 zone, unsigned int size)
> +{
> + return __hash_bucket(____hash_conntrack(tuple, zone), size);
> }
>
> static inline u_int32_t hash_conntrack(const struct net *net, u16 zone,
> const struct nf_conntrack_tuple *tuple)
> {
> - return __hash_conntrack(tuple, zone, net->ct.htable_size,
> - nf_conntrack_hash_rnd);
> + return __hash_conntrack(tuple, zone, net->ct.htable_size);
> }
>
> bool
> @@ -292,13 +313,13 @@ static void death_by_timeout(unsigned long ul_conntrack)
> * OR
> * - Caller must lock nf_conntrack_lock before calling this function
> */
> -struct nf_conntrack_tuple_hash *
> -__nf_conntrack_find(struct net *net, u16 zone,
> - const struct nf_conntrack_tuple *tuple)
> +static struct nf_conntrack_tuple_hash *
> +____nf_conntrack_find(struct net *net, u16 zone,
> + const struct nf_conntrack_tuple *tuple, u32 __hash)
> {
> struct nf_conntrack_tuple_hash *h;
> struct hlist_nulls_node *n;
> - unsigned int hash = hash_conntrack(net, zone, tuple);
> + unsigned int hash = hash_bucket(__hash, net);
>
> /* Disable BHs the entire time since we normally need to disable them
> * at least once for the stats anyway.
> @@ -327,19 +348,27 @@ begin:
>
> return NULL;
> }
> +
> +struct nf_conntrack_tuple_hash *
> +__nf_conntrack_find(struct net *net, u16 zone,
> + const struct nf_conntrack_tuple *tuple)
> +{
> + return ____nf_conntrack_find(net, zone, tuple,
> + ____hash_conntrack(tuple, zone));
> +}
> EXPORT_SYMBOL_GPL(__nf_conntrack_find);
>
> /* Find a connection corresponding to a tuple. */
> -struct nf_conntrack_tuple_hash *
> -nf_conntrack_find_get(struct net *net, u16 zone,
> - const struct nf_conntrack_tuple *tuple)
> +static struct nf_conntrack_tuple_hash *
> +__nf_conntrack_find_get(struct net *net, u16 zone,
> + const struct nf_conntrack_tuple *tuple, u32 __hash)
> {
> struct nf_conntrack_tuple_hash *h;
> struct nf_conn *ct;
>
> rcu_read_lock();
> begin:
> - h = __nf_conntrack_find(net, zone, tuple);
> + h = ____nf_conntrack_find(net, zone, tuple, __hash);
> if (h) {
> ct = nf_ct_tuplehash_to_ctrack(h);
> if (unlikely(nf_ct_is_dying(ct) ||
> @@ -357,6 +386,14 @@ begin:
>
> return h;
> }
> +
> +struct nf_conntrack_tuple_hash *
> +nf_conntrack_find_get(struct net *net, u16 zone,
> + const struct nf_conntrack_tuple *tuple)
> +{
> + return __nf_conntrack_find_get(net, zone, tuple,
> + ____hash_conntrack(tuple, zone));
> +}
> EXPORT_SYMBOL_GPL(nf_conntrack_find_get);
>
> static void __nf_conntrack_hash_insert(struct nf_conn *ct,
> @@ -409,7 +446,8 @@ __nf_conntrack_confirm(struct sk_buff *skb)
> return NF_ACCEPT;
>
> zone = nf_ct_zone(ct);
> - hash = hash_conntrack(net, zone, &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple);
> + /* reuse the __hash saved before */
> + hash = hash_bucket(*(unsigned long *)&ct->tuplehash[IP_CT_DIR_REPLY].hnnode.pprev, net);
> repl_hash = hash_conntrack(net, zone, &ct->tuplehash[IP_CT_DIR_REPLY].tuple);
>
> /* We're not in hash table, and we refuse to set up related
> @@ -567,25 +605,20 @@ static noinline int early_drop(struct net *net, unsigned int hash)
> return dropped;
> }
>
> -struct nf_conn *nf_conntrack_alloc(struct net *net, u16 zone,
> - const struct nf_conntrack_tuple *orig,
> - const struct nf_conntrack_tuple *repl,
> - gfp_t gfp)
> +static struct nf_conn *
> +__nf_conntrack_alloc(struct net *net, u16 zone,
> + const struct nf_conntrack_tuple *orig,
> + const struct nf_conntrack_tuple *repl,
> + gfp_t gfp, u32 __hash)
> {
> struct nf_conn *ct;
>
> - if (unlikely(!nf_conntrack_hash_rnd_initted)) {
> - get_random_bytes(&nf_conntrack_hash_rnd,
> - sizeof(nf_conntrack_hash_rnd));
> - nf_conntrack_hash_rnd_initted = 1;
> - }
> -
> /* We don't want any race condition at early drop stage */
> atomic_inc(&net->ct.count);
>
> if (nf_conntrack_max &&
> unlikely(atomic_read(&net->ct.count) > nf_conntrack_max)) {
> - unsigned int hash = hash_conntrack(net, zone, orig);
> + unsigned int hash = hash_bucket(__hash, net);
> if (!early_drop(net, hash)) {
> atomic_dec(&net->ct.count);
> if (net_ratelimit())
> @@ -616,7 +649,8 @@ struct nf_conn *nf_conntrack_alloc(struct net *net, u16 zone,
> ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple = *orig;
> ct->tuplehash[IP_CT_DIR_ORIGINAL].hnnode.pprev = NULL;
> ct->tuplehash[IP_CT_DIR_REPLY].tuple = *repl;
> - ct->tuplehash[IP_CT_DIR_REPLY].hnnode.pprev = NULL;
> + /* save __hash for reusing when confirming */
> + *(unsigned long *)(&ct->tuplehash[IP_CT_DIR_REPLY].hnnode.pprev) = __hash;
> /* Don't set timer yet: wait for confirmation */
> setup_timer(&ct->timeout, death_by_timeout, (unsigned long)ct);
> write_pnet(&ct->ct_net, net);
> @@ -643,6 +677,14 @@ out_free:
> return ERR_PTR(-ENOMEM);
> #endif
> }
> +
> +struct nf_conn *nf_conntrack_alloc(struct net *net, u16 zone,
> + const struct nf_conntrack_tuple *orig,
> + const struct nf_conntrack_tuple *repl,
> + gfp_t gfp)
> +{
> + return __nf_conntrack_alloc(net, zone, orig, repl, gfp, 0);
> +}
> EXPORT_SYMBOL_GPL(nf_conntrack_alloc);
>
> void nf_conntrack_free(struct nf_conn *ct)
> @@ -664,7 +706,7 @@ init_conntrack(struct net *net, struct nf_conn *tmpl,
> struct nf_conntrack_l3proto *l3proto,
> struct nf_conntrack_l4proto *l4proto,
> struct sk_buff *skb,
> - unsigned int dataoff)
> + unsigned int dataoff, u32 __hash)
> {
> struct nf_conn *ct;
> struct nf_conn_help *help;
> @@ -678,7 +720,8 @@ init_conntrack(struct net *net, struct nf_conn *tmpl,
> return NULL;
> }
>
> - ct = nf_conntrack_alloc(net, zone, tuple, &repl_tuple, GFP_ATOMIC);
> + ct = __nf_conntrack_alloc(net, zone, tuple, &repl_tuple, GFP_ATOMIC,
> + __hash);
> if (IS_ERR(ct)) {
> pr_debug("Can't allocate conntrack.\n");
> return (struct nf_conntrack_tuple_hash *)ct;
> @@ -755,6 +798,7 @@ resolve_normal_ct(struct net *net, struct nf_conn *tmpl,
> struct nf_conntrack_tuple_hash *h;
> struct nf_conn *ct;
> u16 zone = tmpl ? nf_ct_zone(tmpl) : NF_CT_DEFAULT_ZONE;
> + u32 __hash;
>
> if (!nf_ct_get_tuple(skb, skb_network_offset(skb),
> dataoff, l3num, protonum, &tuple, l3proto,
> @@ -764,10 +808,11 @@ resolve_normal_ct(struct net *net, struct nf_conn *tmpl,
> }
>
> /* look for tuple match */
> - h = nf_conntrack_find_get(net, zone, &tuple);
> + __hash = ____hash_conntrack(&tuple, zone);
> + h = __nf_conntrack_find_get(net, zone, &tuple, __hash);
> if (!h) {
> h = init_conntrack(net, tmpl, &tuple, l3proto, l4proto,
> - skb, dataoff);
> + skb, dataoff, __hash);
> if (!h)
> return NULL;
> if (IS_ERR(h))
> @@ -1307,8 +1352,7 @@ int nf_conntrack_set_hashsize(const char *val, struct kernel_param *kp)
> ct = nf_ct_tuplehash_to_ctrack(h);
> hlist_nulls_del_rcu(&h->hnnode);
> bucket = __hash_conntrack(&h->tuple, nf_ct_zone(ct),
> - hashsize,
> - nf_conntrack_hash_rnd);
> + hashsize);
> hlist_nulls_add_head_rcu(&h->hnnode, &hash[bucket]);
> }
> }
--
Mathieu Desnoyers
Operating System Efficiency R&D Consultant
EfficiOS Inc.
http://www.efficios.com
^ permalink raw reply
* [PATCH NEXT 10/12] qlcnic: firmware initialization update
From: Amit Kumar Salecha @ 2010-08-19 15:08 UTC (permalink / raw)
To: davem; +Cc: netdev, ameen.rahman, anirban.chakraborty, Sony Chacko
In-Reply-To: <1282230513-25286-1-git-send-email-amit.salecha@qlogic.com>
From: Sony Chacko <sony.chacko@qlogic.com>
Cleanup legacy code which is not valid for Qlogic
CNA adapters.
Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>
Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com>
---
drivers/net/qlcnic/qlcnic.h | 4 +-
drivers/net/qlcnic/qlcnic_hdr.h | 3 +-
drivers/net/qlcnic/qlcnic_init.c | 133 +++++++++-----------------------------
drivers/net/qlcnic/qlcnic_main.c | 38 +++--------
4 files changed, 47 insertions(+), 131 deletions(-)
diff --git a/drivers/net/qlcnic/qlcnic.h b/drivers/net/qlcnic/qlcnic.h
index b58c411..43cb925 100644
--- a/drivers/net/qlcnic/qlcnic.h
+++ b/drivers/net/qlcnic/qlcnic.h
@@ -54,6 +54,8 @@
#define _QLCNIC_LINUX_SUBVERSION 7
#define QLCNIC_LINUX_VERSIONID "5.0.7"
#define QLCNIC_DRV_IDC_VER 0x01
+#define QLCNIC_DRIVER_VERSION ((_QLCNIC_LINUX_MAJOR << 16) |\
+ (_QLCNIC_LINUX_MINOR << 8) | (_QLCNIC_LINUX_SUBVERSION))
#define QLCNIC_VERSION_CODE(a, b, c) (((a) << 24) + ((b) << 16) + (c))
#define _major(v) (((v) >> 24) & 0xff)
@@ -1233,7 +1235,7 @@ void qlcnic_reset_rx_buffers_list(struct qlcnic_adapter *adapter);
void qlcnic_release_rx_buffers(struct qlcnic_adapter *adapter);
void qlcnic_release_tx_buffers(struct qlcnic_adapter *adapter);
-int qlcnic_init_firmware(struct qlcnic_adapter *adapter);
+int qlcnic_check_fw_status(struct qlcnic_adapter *adapter);
void qlcnic_watchdog_task(struct work_struct *work);
void qlcnic_post_rx_buffers(struct qlcnic_adapter *adapter, u32 ringid,
struct qlcnic_host_rds_ring *rds_ring);
diff --git a/drivers/net/qlcnic/qlcnic_hdr.h b/drivers/net/qlcnic/qlcnic_hdr.h
index eae03b5..794f657 100644
--- a/drivers/net/qlcnic/qlcnic_hdr.h
+++ b/drivers/net/qlcnic/qlcnic_hdr.h
@@ -747,7 +747,8 @@ enum {
#define QLCNIC_RESET_TIMEOUT_SECS 10
#define QLCNIC_INIT_TIMEOUT_SECS 30
-
+#define QLCNIC_HEARTBEAT_PERIOD_MSECS 200
+#define QLCNIC_HEARTBEAT_RETRY_COUNT 30
#define ISR_MSI_INT_TRIGGER(FUNC) (QLCNIC_PCIX_PS_REG(PCIX_MSI_F(FUNC)))
#define ISR_LEGACY_INT_TRIGGERED(VAL) (((VAL) & 0x300) == 0x200)
diff --git a/drivers/net/qlcnic/qlcnic_init.c b/drivers/net/qlcnic/qlcnic_init.c
index 9076675..e7a399f 100644
--- a/drivers/net/qlcnic/qlcnic_init.c
+++ b/drivers/net/qlcnic/qlcnic_init.c
@@ -435,11 +435,14 @@ int qlcnic_pinit_from_rom(struct qlcnic_adapter *adapter)
u32 off;
struct pci_dev *pdev = adapter->pdev;
- /* resetall */
+ QLCWR32(adapter, CRB_CMDPEG_STATE, 0);
+ QLCWR32(adapter, CRB_RCVPEG_STATE, 0);
+
qlcnic_rom_lock(adapter);
QLCWR32(adapter, QLCNIC_ROMUSB_GLB_SW_RESET, 0xfeffffff);
qlcnic_rom_unlock(adapter);
+ /* Init HW CRB block */
if (qlcnic_rom_fast_read(adapter, 0, &n) != 0 || (n != 0xcafecafe) ||
qlcnic_rom_fast_read(adapter, 4, &n) != 0) {
dev_err(&pdev->dev, "ERROR Reading crb_init area: val:%x\n", n);
@@ -520,13 +523,10 @@ int qlcnic_pinit_from_rom(struct qlcnic_adapter *adapter)
}
kfree(buf);
- /* p2dn replyCount */
+ /* Initialize protocol process engine */
QLCWR32(adapter, QLCNIC_CRB_PEG_NET_D + 0xec, 0x1e);
- /* disable_peg_cache 0 & 1*/
QLCWR32(adapter, QLCNIC_CRB_PEG_NET_D + 0x4c, 8);
QLCWR32(adapter, QLCNIC_CRB_PEG_NET_I + 0x4c, 8);
-
- /* peg_clr_all */
QLCWR32(adapter, QLCNIC_CRB_PEG_NET_0 + 0x8, 0);
QLCWR32(adapter, QLCNIC_CRB_PEG_NET_0 + 0xc, 0);
QLCWR32(adapter, QLCNIC_CRB_PEG_NET_1 + 0x8, 0);
@@ -535,10 +535,36 @@ int qlcnic_pinit_from_rom(struct qlcnic_adapter *adapter)
QLCWR32(adapter, QLCNIC_CRB_PEG_NET_2 + 0xc, 0);
QLCWR32(adapter, QLCNIC_CRB_PEG_NET_3 + 0x8, 0);
QLCWR32(adapter, QLCNIC_CRB_PEG_NET_3 + 0xc, 0);
+ QLCWR32(adapter, QLCNIC_CRB_PEG_NET_4 + 0x8, 0);
+ QLCWR32(adapter, QLCNIC_CRB_PEG_NET_4 + 0xc, 0);
+ msleep(1);
+ QLCWR32(adapter, QLCNIC_PEG_HALT_STATUS1, 0);
+ QLCWR32(adapter, QLCNIC_PEG_HALT_STATUS2, 0);
return 0;
}
int
+qlcnic_check_fw_status(struct qlcnic_adapter *adapter)
+{
+ u32 heartbit, ret = -EIO;
+ int retries = QLCNIC_HEARTBEAT_RETRY_COUNT;
+
+ adapter->heartbit = QLCRD32(adapter, QLCNIC_PEG_ALIVE_COUNTER);
+ do {
+ msleep(QLCNIC_HEARTBEAT_PERIOD_MSECS);
+ heartbit = QLCRD32(adapter, QLCNIC_PEG_ALIVE_COUNTER);
+ if (heartbit != adapter->heartbit) {
+ /* Complete firmware handshake */
+ QLCWR32(adapter, CRB_CMDPEG_STATE, PHAN_INITIALIZE_ACK);
+ ret = QLCNIC_RCODE_SUCCESS;
+ break;
+ }
+ } while (--retries);
+
+ return ret;
+}
+
+int
qlcnic_setup_idc_param(struct qlcnic_adapter *adapter) {
int timeo;
@@ -905,35 +931,12 @@ qlcnic_get_bios_version(struct qlcnic_adapter *adapter)
int
qlcnic_need_fw_reset(struct qlcnic_adapter *adapter)
{
- u32 count, old_count;
u32 val, version, major, minor, build;
- int i, timeout;
if (adapter->need_fw_reset)
return 1;
- /* last attempt had failed */
- if (QLCRD32(adapter, CRB_CMDPEG_STATE) == PHAN_INITIALIZE_FAILED)
- return 1;
-
- old_count = QLCRD32(adapter, QLCNIC_PEG_ALIVE_COUNTER);
-
- for (i = 0; i < 10; i++) {
-
- timeout = msleep_interruptible(200);
- if (timeout) {
- QLCWR32(adapter, CRB_CMDPEG_STATE,
- PHAN_INITIALIZE_FAILED);
- return -EINTR;
- }
-
- count = QLCRD32(adapter, QLCNIC_PEG_ALIVE_COUNTER);
- if (count != old_count)
- break;
- }
-
- /* firmware is dead */
- if (count == old_count)
+ if (qlcnic_check_fw_status(adapter))
return 1;
/* check if we have got newer or different file firmware */
@@ -1158,78 +1161,6 @@ qlcnic_release_firmware(struct qlcnic_adapter *adapter)
adapter->fw = NULL;
}
-static int qlcnic_cmd_peg_ready(struct qlcnic_adapter *adapter)
-{
- u32 val;
- int retries = 60;
-
- do {
- val = QLCRD32(adapter, CRB_CMDPEG_STATE);
-
- switch (val) {
- case PHAN_INITIALIZE_COMPLETE:
- case PHAN_INITIALIZE_ACK:
- return 0;
- case PHAN_INITIALIZE_FAILED:
- goto out_err;
- default:
- break;
- }
-
- msleep(500);
-
- } while (--retries);
-
- QLCWR32(adapter, CRB_CMDPEG_STATE, PHAN_INITIALIZE_FAILED);
-
-out_err:
- dev_err(&adapter->pdev->dev, "Command Peg initialization not "
- "complete, state: 0x%x.\n", val);
- return -EIO;
-}
-
-static int
-qlcnic_receive_peg_ready(struct qlcnic_adapter *adapter)
-{
- u32 val;
- int retries = 2000;
-
- do {
- val = QLCRD32(adapter, CRB_RCVPEG_STATE);
-
- if (val == PHAN_PEG_RCV_INITIALIZED)
- return 0;
-
- msleep(10);
-
- } while (--retries);
-
- if (!retries) {
- dev_err(&adapter->pdev->dev, "Receive Peg initialization not "
- "complete, state: 0x%x.\n", val);
- return -EIO;
- }
-
- return 0;
-}
-
-int qlcnic_init_firmware(struct qlcnic_adapter *adapter)
-{
- int err;
-
- err = qlcnic_cmd_peg_ready(adapter);
- if (err)
- return err;
-
- err = qlcnic_receive_peg_ready(adapter);
- if (err)
- return err;
-
- QLCWR32(adapter, CRB_CMDPEG_STATE, PHAN_INITIALIZE_ACK);
-
- return err;
-}
-
static void
qlcnic_handle_linkevent(struct qlcnic_adapter *adapter,
struct qlcnic_fw_msg *msg)
diff --git a/drivers/net/qlcnic/qlcnic_main.c b/drivers/net/qlcnic/qlcnic_main.c
index 6e246c8..fa87a96 100644
--- a/drivers/net/qlcnic/qlcnic_main.c
+++ b/drivers/net/qlcnic/qlcnic_main.c
@@ -940,18 +940,13 @@ static int qlcnic_check_npar_opertional(struct qlcnic_adapter *adapter)
static int
qlcnic_start_firmware(struct qlcnic_adapter *adapter)
{
- int val, err, first_boot;
+ int err;
err = qlcnic_can_start_firmware(adapter);
if (err < 0)
return err;
else if (!err)
- goto wait_init;
-
- first_boot = QLCRD32(adapter, QLCNIC_CAM_RAM(0x1fc));
- if (first_boot == 0x55555555)
- /* This is the first boot after power up */
- QLCWR32(adapter, QLCNIC_CAM_RAM(0x1fc), QLCNIC_BDINFO_MAGIC);
+ goto check_fw_status;
if (load_fw_file)
qlcnic_request_firmware(adapter);
@@ -963,21 +958,12 @@ qlcnic_start_firmware(struct qlcnic_adapter *adapter)
}
err = qlcnic_need_fw_reset(adapter);
- if (err < 0)
- goto err_out;
if (err == 0)
- goto wait_init;
-
- if (first_boot != 0x55555555) {
- QLCWR32(adapter, CRB_CMDPEG_STATE, 0);
- QLCWR32(adapter, CRB_RCVPEG_STATE, 0);
- qlcnic_pinit_from_rom(adapter);
- msleep(1);
- }
-
- QLCWR32(adapter, QLCNIC_PEG_HALT_STATUS1, 0);
- QLCWR32(adapter, QLCNIC_PEG_HALT_STATUS2, 0);
+ goto set_dev_ready;
+ err = qlcnic_pinit_from_rom(adapter);
+ if (err)
+ goto err_out;
qlcnic_set_port_mode(adapter);
err = qlcnic_load_firmware(adapter);
@@ -985,18 +971,14 @@ qlcnic_start_firmware(struct qlcnic_adapter *adapter)
goto err_out;
qlcnic_release_firmware(adapter);
+ QLCWR32(adapter, CRB_DRIVER_VERSION, QLCNIC_DRIVER_VERSION);
- val = (_QLCNIC_LINUX_MAJOR << 16)
- | ((_QLCNIC_LINUX_MINOR << 8))
- | (_QLCNIC_LINUX_SUBVERSION);
- QLCWR32(adapter, CRB_DRIVER_VERSION, val);
-
-wait_init:
- /* Handshake with the card before we register the devices. */
- err = qlcnic_init_firmware(adapter);
+check_fw_status:
+ err = qlcnic_check_fw_status(adapter);
if (err)
goto err_out;
+set_dev_ready:
QLCWR32(adapter, QLCNIC_CRB_DEV_STATE, QLCNIC_DEV_READY);
qlcnic_idc_debug_info(adapter, 1);
err = qlcnic_check_npar_opertional(adapter);
--
1.6.0.2
^ permalink raw reply related
* [PATCH NEXT 09/12] qlcnic: fix endiness in eswitch statistics
From: Amit Kumar Salecha @ 2010-08-19 15:08 UTC (permalink / raw)
To: davem; +Cc: netdev, ameen.rahman, anirban.chakraborty
In-Reply-To: <1282230513-25286-1-git-send-email-amit.salecha@qlogic.com>
Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com>
---
drivers/net/qlcnic/qlcnic_ctx.c | 18 ++++++++++++++++--
1 files changed, 16 insertions(+), 2 deletions(-)
diff --git a/drivers/net/qlcnic/qlcnic_ctx.c b/drivers/net/qlcnic/qlcnic_ctx.c
index 74ae3b0..315705b 100644
--- a/drivers/net/qlcnic/qlcnic_ctx.c
+++ b/drivers/net/qlcnic/qlcnic_ctx.c
@@ -946,6 +946,7 @@ int qlcnic_get_port_stats(struct qlcnic_adapter *adapter, const u8 func,
const u8 rx_tx, struct __qlcnic_esw_statistics *esw_stats) {
size_t stats_size = sizeof(struct __qlcnic_esw_statistics);
+ struct __qlcnic_esw_statistics *stats;
dma_addr_t stats_dma_t;
void *stats_addr;
u32 arg1;
@@ -980,8 +981,21 @@ int qlcnic_get_port_stats(struct qlcnic_adapter *adapter, const u8 func,
LSD(stats_dma_t),
QLCNIC_CDRP_CMD_GET_ESWITCH_STATS);
- if (!err)
- memcpy(esw_stats, stats_addr, stats_size);
+ if (!err) {
+ stats = (struct __qlcnic_esw_statistics *)stats_addr;
+ esw_stats->context_id = le16_to_cpu(stats->context_id);
+ esw_stats->version = le16_to_cpu(stats->version);
+ esw_stats->size = le16_to_cpu(stats->size);
+ esw_stats->multicast_frames =
+ le64_to_cpu(stats->multicast_frames);
+ esw_stats->broadcast_frames =
+ le64_to_cpu(stats->broadcast_frames);
+ esw_stats->unicast_frames = le64_to_cpu(stats->unicast_frames);
+ esw_stats->dropped_frames = le64_to_cpu(stats->dropped_frames);
+ esw_stats->local_frames = le64_to_cpu(stats->local_frames);
+ esw_stats->errors = le64_to_cpu(stats->errors);
+ esw_stats->numbytes = le64_to_cpu(stats->numbytes);
+ }
pci_free_consistent(adapter->pdev, stats_size, stats_addr,
stats_dma_t);
--
1.6.0.2
^ permalink raw reply related
* [PATCH NEXT 12/12] qlcnic: update version 5.0.8
From: Amit Kumar Salecha @ 2010-08-19 15:08 UTC (permalink / raw)
To: davem; +Cc: netdev, ameen.rahman, anirban.chakraborty
In-Reply-To: <1282230513-25286-1-git-send-email-amit.salecha@qlogic.com>
Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com>
---
drivers/net/qlcnic/qlcnic.h | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/qlcnic/qlcnic.h b/drivers/net/qlcnic/qlcnic.h
index 43cb925..508d531 100644
--- a/drivers/net/qlcnic/qlcnic.h
+++ b/drivers/net/qlcnic/qlcnic.h
@@ -51,8 +51,8 @@
#define _QLCNIC_LINUX_MAJOR 5
#define _QLCNIC_LINUX_MINOR 0
-#define _QLCNIC_LINUX_SUBVERSION 7
-#define QLCNIC_LINUX_VERSIONID "5.0.7"
+#define _QLCNIC_LINUX_SUBVERSION 8
+#define QLCNIC_LINUX_VERSIONID "5.0.8"
#define QLCNIC_DRV_IDC_VER 0x01
#define QLCNIC_DRIVER_VERSION ((_QLCNIC_LINUX_MAJOR << 16) |\
(_QLCNIC_LINUX_MINOR << 8) | (_QLCNIC_LINUX_SUBVERSION))
--
1.6.0.2
^ permalink raw reply related
* [PATCH NEXT 11/12] qlcnic: rom lock recovery
From: Amit Kumar Salecha @ 2010-08-19 15:08 UTC (permalink / raw)
To: davem; +Cc: netdev, ameen.rahman, anirban.chakraborty, Sucheta Chakraborty
In-Reply-To: <1282230513-25286-1-git-send-email-amit.salecha@qlogic.com>
From: Sucheta Chakraborty <sucheta.chakraborty@qlogic.com>
Fw can get stuck while holding pci semaphore. Driver will not
be able to perform fw initialization, without this lock.
Release semaphore forcefully in that case.
Signed-off-by: Sucheta Chakraborty <sucheta.chakraborty@qlogic.com>
Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com>
---
drivers/net/qlcnic/qlcnic_hw.c | 4 ++--
drivers/net/qlcnic/qlcnic_init.c | 14 ++++++++++++--
2 files changed, 14 insertions(+), 4 deletions(-)
diff --git a/drivers/net/qlcnic/qlcnic_hw.c b/drivers/net/qlcnic/qlcnic_hw.c
index 9d40ce0..5e6f486 100644
--- a/drivers/net/qlcnic/qlcnic_hw.c
+++ b/drivers/net/qlcnic/qlcnic_hw.c
@@ -297,8 +297,8 @@ qlcnic_pcie_sem_lock(struct qlcnic_adapter *adapter, int sem, u32 id_reg)
break;
if (++timeout >= QLCNIC_PCIE_SEM_TIMEOUT) {
dev_err(&adapter->pdev->dev,
- "Failed to acquire sem=%d lock;reg_id=%d\n",
- sem, id_reg);
+ "Failed to acquire sem=%d lock; holdby=%d\n",
+ sem, id_reg ? QLCRD32(adapter, id_reg) : -1);
return -EIO;
}
msleep(1);
diff --git a/drivers/net/qlcnic/qlcnic_init.c b/drivers/net/qlcnic/qlcnic_init.c
index e7a399f..a174521 100644
--- a/drivers/net/qlcnic/qlcnic_init.c
+++ b/drivers/net/qlcnic/qlcnic_init.c
@@ -928,15 +928,25 @@ qlcnic_get_bios_version(struct qlcnic_adapter *adapter)
return (bios_ver << 16) + ((bios_ver >> 8) & 0xff00) + (bios_ver >> 24);
}
+static void qlcnic_rom_lock_recovery(struct qlcnic_adapter *adapter)
+{
+ if (qlcnic_pcie_sem_lock(adapter, 2, QLCNIC_ROM_LOCK_ID))
+ dev_info(&adapter->pdev->dev, "Resetting rom_lock\n");
+
+ qlcnic_pcie_sem_unlock(adapter, 2);
+}
+
int
qlcnic_need_fw_reset(struct qlcnic_adapter *adapter)
{
u32 val, version, major, minor, build;
- if (adapter->need_fw_reset)
+ if (qlcnic_check_fw_status(adapter)) {
+ qlcnic_rom_lock_recovery(adapter);
return 1;
+ }
- if (qlcnic_check_fw_status(adapter))
+ if (adapter->need_fw_reset)
return 1;
/* check if we have got newer or different file firmware */
--
1.6.0.2
^ permalink raw reply related
* [PATCH NEXT 06/12] qlcnic: support anti mac spoofing
From: Amit Kumar Salecha @ 2010-08-19 15:08 UTC (permalink / raw)
To: davem; +Cc: netdev, ameen.rahman, anirban.chakraborty, Sony Chacko
In-Reply-To: <1282230513-25286-1-git-send-email-amit.salecha@qlogic.com>
From: Sony Chacko <sony.chacko@qlogic.com>
Administrator can configure to drop packet in transmit,
if it doesn't match interface mac address, in case of virtual function.
Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>
Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com>
---
drivers/net/qlcnic/qlcnic.h | 1 +
drivers/net/qlcnic/qlcnic_main.c | 11 +++++++++++
2 files changed, 12 insertions(+), 0 deletions(-)
diff --git a/drivers/net/qlcnic/qlcnic.h b/drivers/net/qlcnic/qlcnic.h
index 9433a05..b58c411 100644
--- a/drivers/net/qlcnic/qlcnic.h
+++ b/drivers/net/qlcnic/qlcnic.h
@@ -900,6 +900,7 @@ struct qlcnic_mac_req {
#define QLCNIC_BRIDGE_ENABLED 0X10
#define QLCNIC_DIAG_ENABLED 0x20
#define QLCNIC_ESWITCH_ENABLED 0x40
+#define QLCNIC_MACSPOOF 0x200
#define QLCNIC_IS_MSI_FAMILY(adapter) \
((adapter)->flags & (QLCNIC_MSI_ENABLED | QLCNIC_MSIX_ENABLED))
diff --git a/drivers/net/qlcnic/qlcnic_main.c b/drivers/net/qlcnic/qlcnic_main.c
index aa1f6b3..de4be00 100644
--- a/drivers/net/qlcnic/qlcnic_main.c
+++ b/drivers/net/qlcnic/qlcnic_main.c
@@ -762,6 +762,11 @@ static void
qlcnic_set_eswitch_port_features(struct qlcnic_adapter *adapter,
struct qlcnic_esw_func_cfg *esw_cfg)
{
+ adapter->flags &= ~QLCNIC_MACSPOOF;
+ if (adapter->op_mode == QLCNIC_NON_PRIV_FUNC)
+ if (esw_cfg->mac_anti_spoof)
+ adapter->flags |= QLCNIC_MACSPOOF;
+
qlcnic_set_netdev_features(adapter, esw_cfg);
}
@@ -1912,6 +1917,12 @@ qlcnic_xmit_frame(struct sk_buff *skb, struct net_device *netdev)
return NETDEV_TX_BUSY;
}
+ if (adapter->flags & QLCNIC_MACSPOOF) {
+ if (compare_ether_addr(eth_hdr(skb)->h_source,
+ adapter->mac_addr))
+ goto drop_packet;
+ }
+
frag_count = skb_shinfo(skb)->nr_frags + 1;
/* 4 fragments per cmd des */
--
1.6.0.2
^ permalink raw reply related
* [PATCH NEXT 03/12] qlcnic: replace magic numbers with defines
From: Amit Kumar Salecha @ 2010-08-19 15:08 UTC (permalink / raw)
To: davem; +Cc: netdev, ameen.rahman, anirban.chakraborty, Sony Chacko
In-Reply-To: <1282230513-25286-1-git-send-email-amit.salecha@qlogic.com>
From: Sony Chacko <sony.chacko@qlogic.com>
Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>
---
drivers/net/qlcnic/qlcnic.h | 1 +
drivers/net/qlcnic/qlcnic_hdr.h | 4 ++++
drivers/net/qlcnic/qlcnic_init.c | 4 ++--
drivers/net/qlcnic/qlcnic_main.c | 2 +-
4 files changed, 8 insertions(+), 3 deletions(-)
diff --git a/drivers/net/qlcnic/qlcnic.h b/drivers/net/qlcnic/qlcnic.h
index 5c549c5..d198367 100644
--- a/drivers/net/qlcnic/qlcnic.h
+++ b/drivers/net/qlcnic/qlcnic.h
@@ -148,6 +148,7 @@
#define DEFAULT_RCV_DESCRIPTORS_1G 2048
#define DEFAULT_RCV_DESCRIPTORS_10G 4096
+#define MAX_RDS_RINGS 2
#define get_next_index(index, length) \
(((index) + 1) & ((length) - 1))
diff --git a/drivers/net/qlcnic/qlcnic_hdr.h b/drivers/net/qlcnic/qlcnic_hdr.h
index bd346d9..39db4df 100644
--- a/drivers/net/qlcnic/qlcnic_hdr.h
+++ b/drivers/net/qlcnic/qlcnic_hdr.h
@@ -745,6 +745,10 @@ enum {
#define FW_POLL_DELAY (1 * HZ)
#define FW_FAIL_THRESH 2
+#define QLCNIC_RESET_TIMEOUT_SECS 10
+#define QLCNIC_INIT_TIMEOUT_SECS 30
+
+
#define ISR_MSI_INT_TRIGGER(FUNC) (QLCNIC_PCIX_PS_REG(PCIX_MSI_F(FUNC)))
#define ISR_LEGACY_INT_TRIGGERED(VAL) (((VAL) & 0x300) == 0x200)
diff --git a/drivers/net/qlcnic/qlcnic_init.c b/drivers/net/qlcnic/qlcnic_init.c
index 08da258..9076675 100644
--- a/drivers/net/qlcnic/qlcnic_init.c
+++ b/drivers/net/qlcnic/qlcnic_init.c
@@ -553,12 +553,12 @@ qlcnic_setup_idc_param(struct qlcnic_adapter *adapter) {
}
adapter->physical_port = (val >> 2);
if (qlcnic_rom_fast_read(adapter, QLCNIC_ROM_DEV_INIT_TIMEOUT, &timeo))
- timeo = 30;
+ timeo = QLCNIC_INIT_TIMEOUT_SECS;
adapter->dev_init_timeo = timeo;
if (qlcnic_rom_fast_read(adapter, QLCNIC_ROM_DRV_RESET_TIMEOUT, &timeo))
- timeo = 10;
+ timeo = QLCNIC_RESET_TIMEOUT_SECS;
adapter->reset_ack_timeo = timeo;
diff --git a/drivers/net/qlcnic/qlcnic_main.c b/drivers/net/qlcnic/qlcnic_main.c
index beadf2e..c6f19c9 100644
--- a/drivers/net/qlcnic/qlcnic_main.c
+++ b/drivers/net/qlcnic/qlcnic_main.c
@@ -754,7 +754,7 @@ qlcnic_check_options(struct qlcnic_adapter *adapter)
adapter->num_txd = MAX_CMD_DESCRIPTORS;
- adapter->max_rds_rings = 2;
+ adapter->max_rds_rings = MAX_RDS_RINGS;
}
static int
--
1.6.0.2
^ permalink raw reply related
* [PATCH NEXT 04/12] qlcnic: configure port on eswitch
From: Amit Kumar Salecha @ 2010-08-19 15:08 UTC (permalink / raw)
To: davem; +Cc: netdev, ameen.rahman, anirban.chakraborty, Rajesh Borundia
In-Reply-To: <1282230513-25286-1-git-send-email-amit.salecha@qlogic.com>
From: Rajesh Borundia <rajesh.borundia@qlogic.com>
o Nic partition capable devices has embedded switch, this needs to support
various features like external switch.
Signed-off-by: Rajesh Borundia <rajesh.borundia@qlogic.com>
Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com>
---
drivers/net/qlcnic/qlcnic.h | 22 +++--
drivers/net/qlcnic/qlcnic_ctx.c | 174 ++++++++++++++++++++++++++---------
drivers/net/qlcnic/qlcnic_hdr.h | 7 ++
drivers/net/qlcnic/qlcnic_main.c | 185 +++++++++++++++++++++-----------------
4 files changed, 254 insertions(+), 134 deletions(-)
diff --git a/drivers/net/qlcnic/qlcnic.h b/drivers/net/qlcnic/qlcnic.h
index d198367..9433a05 100644
--- a/drivers/net/qlcnic/qlcnic.h
+++ b/drivers/net/qlcnic/qlcnic.h
@@ -556,6 +556,7 @@ struct qlcnic_recv_context {
#define QLCNIC_CDRP_CMD_GET_ESWITCH_STATUS 0x00000026
#define QLCNIC_CDRP_CMD_SET_PORTMIRRORING 0x00000027
#define QLCNIC_CDRP_CMD_CONFIGURE_ESWITCH 0x00000028
+#define QLCNIC_CDRP_CMD_GET_ESWITCH_PORT_CONFIG 0x00000029
#define QLCNIC_CDRP_CMD_GET_ESWITCH_STATS 0x0000002a
#define QLCNIC_RCODE_SUCCESS 0
@@ -1044,7 +1045,7 @@ struct qlcnic_pci_info {
};
struct qlcnic_npar_info {
- u16 vlan_id;
+ u16 pvid;
u16 min_bw;
u16 max_bw;
u8 phy_port;
@@ -1052,11 +1053,13 @@ struct qlcnic_npar_info {
u8 active;
u8 enable_pm;
u8 dest_npar;
- u8 host_vlan_tag;
- u8 promisc_mode;
u8 discard_tagged;
u8 mac_learning;
+ u8 mac_anti_spoof;
+ u8 promisc_mode;
+ u8 offload_flags;
};
+
struct qlcnic_eswitch {
u8 port;
u8 active_vports;
@@ -1088,7 +1091,6 @@ struct qlcnic_eswitch {
#define IS_VALID_BW(bw) (bw >= MIN_BW && bw <= MAX_BW)
#define IS_VALID_TX_QUEUES(que) (que > 0 && que <= MAX_TX_QUEUES)
#define IS_VALID_RX_QUEUES(que) (que > 0 && que <= MAX_RX_QUEUES)
-#define IS_VALID_MODE(mode) (mode == 0 || mode == 1)
struct qlcnic_pci_func_cfg {
u16 func_type;
@@ -1120,12 +1122,16 @@ struct qlcnic_pm_func_cfg {
struct qlcnic_esw_func_cfg {
u16 vlan_id;
+ u8 op_mode;
+ u8 op_type;
u8 pci_func;
u8 host_vlan_tag;
u8 promisc_mode;
u8 discard_tagged;
u8 mac_learning;
- u8 reserved;
+ u8 mac_anti_spoof;
+ u8 offload_flags;
+ u8 reserved[5];
};
#define QLCNIC_STATS_VERSION 1
@@ -1276,8 +1282,10 @@ int qlcnic_get_eswitch_capabilities(struct qlcnic_adapter *, u8,
int qlcnic_get_eswitch_status(struct qlcnic_adapter *, u8,
struct qlcnic_eswitch *);
int qlcnic_toggle_eswitch(struct qlcnic_adapter *, u8, u8);
-int qlcnic_config_switch_port(struct qlcnic_adapter *, u8, int, u8, u8,
- u8, u8, u16);
+int qlcnic_config_switch_port(struct qlcnic_adapter *,
+ struct qlcnic_esw_func_cfg *);
+int qlcnic_get_eswitch_port_config(struct qlcnic_adapter *,
+ struct qlcnic_esw_func_cfg *);
int qlcnic_config_port_mirroring(struct qlcnic_adapter *, u8, u8, u8);
int qlcnic_get_port_stats(struct qlcnic_adapter *, const u8, const u8,
struct __qlcnic_esw_statistics *);
diff --git a/drivers/net/qlcnic/qlcnic_ctx.c b/drivers/net/qlcnic/qlcnic_ctx.c
index 57c9b09..74ae3b0 100644
--- a/drivers/net/qlcnic/qlcnic_ctx.c
+++ b/drivers/net/qlcnic/qlcnic_ctx.c
@@ -813,9 +813,8 @@ int qlcnic_get_eswitch_capabilities(struct qlcnic_adapter *adapter, u8 port,
arg2 = QLCRD32(adapter, QLCNIC_ARG2_CRB_OFFSET);
eswitch->port = arg1 & 0xf;
- eswitch->active_vports = LSB(arg2);
- eswitch->max_ucast_filters = MSB(arg2);
- eswitch->max_active_vlans = LSB(MSW(arg2));
+ eswitch->max_ucast_filters = LSW(arg2);
+ eswitch->max_active_vlans = MSW(arg2) & 0xfff;
if (arg1 & BIT_6)
eswitch->flags |= QLCNIC_SWITCH_VLAN_FILTERING;
if (arg1 & BIT_7)
@@ -943,47 +942,6 @@ int qlcnic_config_port_mirroring(struct qlcnic_adapter *adapter, u8 id,
return err;
}
-/* Configure eSwitch port */
-int qlcnic_config_switch_port(struct qlcnic_adapter *adapter, u8 id,
- int vlan_tagging, u8 discard_tagged, u8 promsc_mode,
- u8 mac_learn, u8 pci_func, u16 vlan_id)
-{
- int err = -EIO;
- u32 arg1;
- struct qlcnic_eswitch *eswitch;
-
- if (adapter->op_mode != QLCNIC_MGMT_FUNC)
- return err;
-
- eswitch = &adapter->eswitch[id];
- if (!(eswitch->flags & QLCNIC_SWITCH_ENABLE))
- return err;
-
- arg1 = eswitch->port | (discard_tagged ? BIT_4 : 0);
- arg1 |= (promsc_mode ? BIT_6 : 0) | (mac_learn ? BIT_7 : 0);
- arg1 |= pci_func << 8;
- if (vlan_tagging)
- arg1 |= BIT_5 | (vlan_id << 16);
-
- err = qlcnic_issue_cmd(adapter,
- adapter->ahw.pci_func,
- adapter->fw_hal_version,
- arg1,
- 0,
- 0,
- QLCNIC_CDRP_CMD_CONFIGURE_ESWITCH);
-
- if (err != QLCNIC_RCODE_SUCCESS) {
- dev_err(&adapter->pdev->dev,
- "Failed to configure eswitch port%d\n", eswitch->port);
- } else {
- dev_info(&adapter->pdev->dev,
- "Configured eSwitch for port %d\n", eswitch->port);
- }
-
- return err;
-}
-
int qlcnic_get_port_stats(struct qlcnic_adapter *adapter, const u8 func,
const u8 rx_tx, struct __qlcnic_esw_statistics *esw_stats) {
@@ -1108,3 +1066,131 @@ err_ret:
"rx_ctx=%d\n", func_esw, port, rx_tx);
return -EIO;
}
+
+static int
+__qlcnic_get_eswitch_port_config(struct qlcnic_adapter *adapter,
+ u32 *arg1, u32 *arg2)
+{
+ int err = -EIO;
+ u8 pci_func;
+ pci_func = (*arg1 >> 8);
+ err = qlcnic_issue_cmd(adapter,
+ adapter->ahw.pci_func,
+ adapter->fw_hal_version,
+ *arg1,
+ 0,
+ 0,
+ QLCNIC_CDRP_CMD_GET_ESWITCH_PORT_CONFIG);
+
+ if (err == QLCNIC_RCODE_SUCCESS) {
+ *arg1 = QLCRD32(adapter, QLCNIC_ARG1_CRB_OFFSET);
+ *arg2 = QLCRD32(adapter, QLCNIC_ARG2_CRB_OFFSET);
+ dev_info(&adapter->pdev->dev,
+ "eSwitch port config for pci func%d\n", pci_func);
+ } else {
+ dev_err(&adapter->pdev->dev,
+ "Failed to get eswitch port config%d\n", pci_func);
+ }
+ return err;
+}
+/* Configure eSwitch port
+op_mode = 0 for setting default port behavior
+op_mode = 1 for setting vlan id
+op_mode = 2 for deleting vlan id
+op_type = 0 for vlan_id
+op_type = 1 for port vlan_id
+*/
+int qlcnic_config_switch_port(struct qlcnic_adapter *adapter,
+ struct qlcnic_esw_func_cfg *esw_cfg)
+{
+ int err = -EIO;
+ u32 arg1, arg2 = 0;
+ u8 pci_func;
+
+ if (adapter->op_mode != QLCNIC_MGMT_FUNC)
+ return err;
+ pci_func = esw_cfg->pci_func;
+ arg1 = (adapter->npars[pci_func].phy_port & BIT_0);
+ arg1 |= (pci_func << 8);
+
+ if (__qlcnic_get_eswitch_port_config(adapter, &arg1, &arg2))
+ return err;
+ arg1 &= ~(0x0ff << 8);
+ arg1 |= (pci_func << 8);
+ arg1 &= ~(BIT_2 | BIT_3);
+ switch (esw_cfg->op_mode) {
+ case QLCNIC_PORT_DEFAULTS:
+ arg1 |= (BIT_4 | BIT_6 | BIT_7);
+ arg2 |= (BIT_0 | BIT_1);
+ if (adapter->capabilities & QLCNIC_FW_CAPABILITY_TSO)
+ arg2 |= (BIT_2 | BIT_3);
+ if (!(esw_cfg->discard_tagged))
+ arg1 &= ~BIT_4;
+ if (!(esw_cfg->promisc_mode))
+ arg1 &= ~BIT_6;
+ if (!(esw_cfg->mac_learning))
+ arg1 &= ~BIT_7;
+ if (!(esw_cfg->mac_anti_spoof))
+ arg2 &= ~BIT_0;
+ if (!(esw_cfg->offload_flags & BIT_0))
+ arg2 &= ~(BIT_1 | BIT_2 | BIT_3);
+ if (!(esw_cfg->offload_flags & BIT_1))
+ arg2 &= ~BIT_2;
+ if (!(esw_cfg->offload_flags & BIT_2))
+ arg2 &= ~BIT_3;
+ break;
+ case QLCNIC_ADD_VLAN:
+ arg1 |= (BIT_2 | BIT_5);
+ arg1 |= (esw_cfg->vlan_id << 16);
+ break;
+ case QLCNIC_DEL_VLAN:
+ arg1 |= (BIT_3 | BIT_5);
+ arg1 &= ~(0x0ffff << 16);
+ default:
+ return err;
+ }
+
+ err = qlcnic_issue_cmd(adapter,
+ adapter->ahw.pci_func,
+ adapter->fw_hal_version,
+ arg1,
+ arg2,
+ 0,
+ QLCNIC_CDRP_CMD_CONFIGURE_ESWITCH);
+
+ if (err != QLCNIC_RCODE_SUCCESS) {
+ dev_err(&adapter->pdev->dev,
+ "Failed to configure eswitch port%d\n", pci_func);
+ } else {
+ dev_info(&adapter->pdev->dev,
+ "Configured eSwitch for port %d\n", pci_func);
+ }
+
+ return err;
+}
+
+int
+qlcnic_get_eswitch_port_config(struct qlcnic_adapter *adapter,
+ struct qlcnic_esw_func_cfg *esw_cfg)
+{
+ u32 arg1, arg2;
+ u8 phy_port;
+ if (adapter->op_mode == QLCNIC_MGMT_FUNC)
+ phy_port = adapter->npars[esw_cfg->pci_func].phy_port;
+ else
+ phy_port = adapter->physical_port;
+ arg1 = phy_port;
+ arg1 |= (esw_cfg->pci_func << 8);
+ if (__qlcnic_get_eswitch_port_config(adapter, &arg1, &arg2))
+ return -EIO;
+
+ esw_cfg->discard_tagged = !!(arg1 & BIT_4);
+ esw_cfg->host_vlan_tag = !!(arg1 & BIT_5);
+ esw_cfg->promisc_mode = !!(arg1 & BIT_6);
+ esw_cfg->mac_learning = !!(arg1 & BIT_7);
+ esw_cfg->vlan_id = LSW(arg1 >> 16);
+ esw_cfg->mac_anti_spoof = (arg2 & 0x1);
+ esw_cfg->offload_flags = ((arg2 >> 1) & 0x7);
+
+ return 0;
+}
diff --git a/drivers/net/qlcnic/qlcnic_hdr.h b/drivers/net/qlcnic/qlcnic_hdr.h
index 39db4df..eae03b5 100644
--- a/drivers/net/qlcnic/qlcnic_hdr.h
+++ b/drivers/net/qlcnic/qlcnic_hdr.h
@@ -775,6 +775,7 @@ struct qlcnic_legacy_intr_set {
#define QLCNIC_DRV_OP_MODE 0x1b2170
#define QLCNIC_MSIX_BASE 0x132110
#define QLCNIC_MAX_PCI_FUNC 8
+#define QLCNIC_MAX_VLAN_FILTERS 64
/* PCI function operational mode */
enum {
@@ -783,6 +784,12 @@ enum {
QLCNIC_NON_PRIV_FUNC = 2
};
+enum {
+ QLCNIC_PORT_DEFAULTS = 0,
+ QLCNIC_ADD_VLAN = 1,
+ QLCNIC_DEL_VLAN = 2
+};
+
#define QLC_DEV_DRV_DEFAULT 0x11111111
#define LSB(x) ((uint8_t)(x))
diff --git a/drivers/net/qlcnic/qlcnic_main.c b/drivers/net/qlcnic/qlcnic_main.c
index c6f19c9..7bb3285 100644
--- a/drivers/net/qlcnic/qlcnic_main.c
+++ b/drivers/net/qlcnic/qlcnic_main.c
@@ -506,7 +506,6 @@ qlcnic_init_pci_info(struct qlcnic_adapter *adapter)
adapter->npars[pfn].active = pci_info[i].active;
adapter->npars[pfn].type = pci_info[i].type;
adapter->npars[pfn].phy_port = pci_info[i].default_port;
- adapter->npars[pfn].mac_learning = DEFAULT_MAC_LEARN;
adapter->npars[pfn].min_bw = pci_info[i].tx_min_bw;
adapter->npars[pfn].max_bw = pci_info[i].tx_max_bw;
}
@@ -758,47 +757,64 @@ qlcnic_check_options(struct qlcnic_adapter *adapter)
}
static int
+qlcnic_reset_eswitch_config(struct qlcnic_adapter *adapter,
+ struct qlcnic_npar_info *npar, int pci_func)
+{
+ struct qlcnic_esw_func_cfg esw_cfg;
+ esw_cfg.op_mode = QLCNIC_PORT_DEFAULTS;
+ esw_cfg.pci_func = pci_func;
+ esw_cfg.vlan_id = npar->pvid;
+ esw_cfg.mac_learning = npar->mac_learning;
+ esw_cfg.discard_tagged = npar->discard_tagged;
+ esw_cfg.mac_anti_spoof = npar->mac_anti_spoof;
+ esw_cfg.offload_flags = npar->offload_flags;
+ esw_cfg.promisc_mode = npar->promisc_mode;
+ if (qlcnic_config_switch_port(adapter, &esw_cfg))
+ return -EIO;
+
+ esw_cfg.op_mode = QLCNIC_ADD_VLAN;
+ if (qlcnic_config_switch_port(adapter, &esw_cfg))
+ return -EIO;
+
+ return 0;
+}
+
+static int
qlcnic_reset_npar_config(struct qlcnic_adapter *adapter)
{
- int i, err = 0;
+ int i, err;
struct qlcnic_npar_info *npar;
struct qlcnic_info nic_info;
if (!(adapter->flags & QLCNIC_ESWITCH_ENABLED) ||
- !adapter->need_fw_reset)
+ !adapter->need_fw_reset || adapter->op_mode != QLCNIC_MGMT_FUNC)
return 0;
- if (adapter->op_mode == QLCNIC_MGMT_FUNC) {
- /* Set the NPAR config data after FW reset */
- for (i = 0; i < QLCNIC_MAX_PCI_FUNC; i++) {
- npar = &adapter->npars[i];
- if (npar->type != QLCNIC_TYPE_NIC)
- continue;
- err = qlcnic_get_nic_info(adapter, &nic_info, i);
- if (err)
- goto err_out;
- nic_info.min_tx_bw = npar->min_bw;
- nic_info.max_tx_bw = npar->max_bw;
- err = qlcnic_set_nic_info(adapter, &nic_info);
- if (err)
- goto err_out;
-
- if (npar->enable_pm) {
- err = qlcnic_config_port_mirroring(adapter,
- npar->dest_npar, 1, i);
- if (err)
- goto err_out;
+ /* Set the NPAR config data after FW reset */
+ for (i = 0; i < QLCNIC_MAX_PCI_FUNC; i++) {
+ npar = &adapter->npars[i];
+ if (npar->type != QLCNIC_TYPE_NIC)
+ continue;
+ err = qlcnic_get_nic_info(adapter, &nic_info, i);
+ if (err)
+ return err;
+ nic_info.min_tx_bw = npar->min_bw;
+ nic_info.max_tx_bw = npar->max_bw;
+ err = qlcnic_set_nic_info(adapter, &nic_info);
+ if (err)
+ return err;
- }
- npar->mac_learning = DEFAULT_MAC_LEARN;
- npar->host_vlan_tag = 0;
- npar->promisc_mode = 0;
- npar->discard_tagged = 0;
- npar->vlan_id = 0;
+ if (npar->enable_pm) {
+ err = qlcnic_config_port_mirroring(adapter,
+ npar->dest_npar, 1, i);
+ if (err)
+ return err;
}
+ err = qlcnic_reset_eswitch_config(adapter, npar, i);
+ if (err)
+ return err;
}
-err_out:
- return err;
+ return 0;
}
static int
@@ -863,12 +879,10 @@ wait_init:
QLCWR32(adapter, QLCNIC_CRB_DEV_STATE, QLCNIC_DEV_READY);
qlcnic_idc_debug_info(adapter, 1);
-
- qlcnic_check_options(adapter);
if (qlcnic_reset_npar_config(adapter))
goto err_out;
qlcnic_dev_set_npar_ready(adapter);
-
+ qlcnic_check_options(adapter);
adapter->need_fw_reset = 0;
qlcnic_release_firmware(adapter);
@@ -3082,9 +3096,6 @@ validate_pm_config(struct qlcnic_adapter *adapter,
if (adapter->npars[dest_pci_func].type != QLCNIC_TYPE_NIC)
return QL_STATUS_INVALID_PARAM;
- if (!IS_VALID_MODE(pm_cfg[i].action))
- return QL_STATUS_INVALID_PARAM;
-
s_esw_id = adapter->npars[src_pci_func].phy_port;
d_esw_id = adapter->npars[dest_pci_func].phy_port;
@@ -3118,7 +3129,7 @@ qlcnic_sysfs_write_pm_config(struct file *filp, struct kobject *kobj,
return ret;
for (i = 0; i < count; i++) {
pci_func = pm_cfg[i].pci_func;
- action = pm_cfg[i].action;
+ action = !!pm_cfg[i].action;
id = adapter->npars[pci_func].phy_port;
ret = qlcnic_config_port_mirroring(adapter, id,
action, pci_func);
@@ -3129,7 +3140,7 @@ qlcnic_sysfs_write_pm_config(struct file *filp, struct kobject *kobj,
for (i = 0; i < count; i++) {
pci_func = pm_cfg[i].pci_func;
id = adapter->npars[pci_func].phy_port;
- adapter->npars[pci_func].enable_pm = pm_cfg[i].action;
+ adapter->npars[pci_func].enable_pm = !!pm_cfg[i].action;
adapter->npars[pci_func].dest_npar = id;
}
return size;
@@ -3161,30 +3172,38 @@ qlcnic_sysfs_read_pm_config(struct file *filp, struct kobject *kobj,
static int
validate_esw_config(struct qlcnic_adapter *adapter,
- struct qlcnic_esw_func_cfg *esw_cfg, int count)
+ struct qlcnic_esw_func_cfg *esw_cfg, int count)
{
u8 pci_func;
int i;
-
for (i = 0; i < count; i++) {
pci_func = esw_cfg[i].pci_func;
if (pci_func >= QLCNIC_MAX_PCI_FUNC)
return QL_STATUS_INVALID_PARAM;
- if (adapter->npars[i].type != QLCNIC_TYPE_NIC)
- return QL_STATUS_INVALID_PARAM;
+ if (adapter->op_mode == QLCNIC_MGMT_FUNC)
+ if (adapter->npars[pci_func].type != QLCNIC_TYPE_NIC)
+ return QL_STATUS_INVALID_PARAM;
- if (esw_cfg->host_vlan_tag == 1)
+ switch (esw_cfg[i].op_mode) {
+ case QLCNIC_PORT_DEFAULTS:
+ break;
+ case QLCNIC_ADD_VLAN:
if (!IS_VALID_VLAN(esw_cfg[i].vlan_id))
return QL_STATUS_INVALID_PARAM;
-
- if (!IS_VALID_MODE(esw_cfg[i].promisc_mode)
- || !IS_VALID_MODE(esw_cfg[i].host_vlan_tag)
- || !IS_VALID_MODE(esw_cfg[i].mac_learning)
- || !IS_VALID_MODE(esw_cfg[i].discard_tagged))
+ if (!esw_cfg[i].op_type)
+ return QL_STATUS_INVALID_PARAM;
+ break;
+ case QLCNIC_DEL_VLAN:
+ if (!IS_VALID_VLAN(esw_cfg[i].vlan_id))
+ return QL_STATUS_INVALID_PARAM;
+ if (!esw_cfg[i].op_type)
+ return QL_STATUS_INVALID_PARAM;
+ break;
+ default:
return QL_STATUS_INVALID_PARAM;
+ }
}
-
return 0;
}
@@ -3195,8 +3214,9 @@ qlcnic_sysfs_write_esw_config(struct file *file, struct kobject *kobj,
struct device *dev = container_of(kobj, struct device, kobj);
struct qlcnic_adapter *adapter = dev_get_drvdata(dev);
struct qlcnic_esw_func_cfg *esw_cfg;
+ struct qlcnic_npar_info *npar;
int count, rem, i, ret;
- u8 id, pci_func;
+ u8 pci_func;
count = size / sizeof(struct qlcnic_esw_func_cfg);
rem = size % sizeof(struct qlcnic_esw_func_cfg);
@@ -3209,28 +3229,28 @@ qlcnic_sysfs_write_esw_config(struct file *file, struct kobject *kobj,
return ret;
for (i = 0; i < count; i++) {
- pci_func = esw_cfg[i].pci_func;
- id = adapter->npars[pci_func].phy_port;
- ret = qlcnic_config_switch_port(adapter, id,
- esw_cfg[i].host_vlan_tag,
- esw_cfg[i].discard_tagged,
- esw_cfg[i].promisc_mode,
- esw_cfg[i].mac_learning,
- esw_cfg[i].pci_func,
- esw_cfg[i].vlan_id);
- if (ret)
- return ret;
+ if (qlcnic_config_switch_port(adapter, &esw_cfg[i]))
+ return QL_STATUS_INVALID_PARAM;
}
for (i = 0; i < count; i++) {
pci_func = esw_cfg[i].pci_func;
- adapter->npars[pci_func].promisc_mode = esw_cfg[i].promisc_mode;
- adapter->npars[pci_func].mac_learning = esw_cfg[i].mac_learning;
- adapter->npars[pci_func].vlan_id = esw_cfg[i].vlan_id;
- adapter->npars[pci_func].discard_tagged =
- esw_cfg[i].discard_tagged;
- adapter->npars[pci_func].host_vlan_tag =
- esw_cfg[i].host_vlan_tag;
+ npar = &adapter->npars[pci_func];
+ switch (esw_cfg[i].op_mode) {
+ case QLCNIC_PORT_DEFAULTS:
+ npar->promisc_mode = esw_cfg[i].promisc_mode;
+ npar->mac_learning = esw_cfg[i].mac_learning;
+ npar->offload_flags = esw_cfg[i].offload_flags;
+ npar->mac_anti_spoof = esw_cfg[i].mac_anti_spoof;
+ npar->discard_tagged = esw_cfg[i].discard_tagged;
+ break;
+ case QLCNIC_ADD_VLAN:
+ npar->pvid = esw_cfg[i].vlan_id;
+ break;
+ case QLCNIC_DEL_VLAN:
+ npar->pvid = 0;
+ break;
+ }
}
return size;
@@ -3243,7 +3263,7 @@ qlcnic_sysfs_read_esw_config(struct file *file, struct kobject *kobj,
struct device *dev = container_of(kobj, struct device, kobj);
struct qlcnic_adapter *adapter = dev_get_drvdata(dev);
struct qlcnic_esw_func_cfg esw_cfg[QLCNIC_MAX_PCI_FUNC];
- int i;
+ u8 i;
if (size != sizeof(esw_cfg))
return QL_STATUS_INVALID_PARAM;
@@ -3251,12 +3271,9 @@ qlcnic_sysfs_read_esw_config(struct file *file, struct kobject *kobj,
for (i = 0; i < QLCNIC_MAX_PCI_FUNC; i++) {
if (adapter->npars[i].type != QLCNIC_TYPE_NIC)
continue;
-
- esw_cfg[i].host_vlan_tag = adapter->npars[i].host_vlan_tag;
- esw_cfg[i].promisc_mode = adapter->npars[i].promisc_mode;
- esw_cfg[i].discard_tagged = adapter->npars[i].discard_tagged;
- esw_cfg[i].vlan_id = adapter->npars[i].vlan_id;
- esw_cfg[i].mac_learning = adapter->npars[i].mac_learning;
+ esw_cfg[i].pci_func = i;
+ if (qlcnic_get_eswitch_port_config(adapter, &esw_cfg[i]))
+ return QL_STATUS_INVALID_PARAM;
}
memcpy(buf, &esw_cfg, size);
@@ -3580,15 +3597,16 @@ qlcnic_create_diag_entries(struct qlcnic_adapter *adapter)
dev_info(dev, "failed to create crb sysfs entry\n");
if (device_create_bin_file(dev, &bin_attr_mem))
dev_info(dev, "failed to create mem sysfs entry\n");
- if (!(adapter->flags & QLCNIC_ESWITCH_ENABLED) ||
- adapter->op_mode != QLCNIC_MGMT_FUNC)
+ if (!(adapter->flags & QLCNIC_ESWITCH_ENABLED))
+ return;
+ if (device_create_bin_file(dev, &bin_attr_esw_config))
+ dev_info(dev, "failed to create esw config sysfs entry");
+ if (adapter->op_mode != QLCNIC_MGMT_FUNC)
return;
if (device_create_bin_file(dev, &bin_attr_pci_config))
dev_info(dev, "failed to create pci config sysfs entry");
if (device_create_bin_file(dev, &bin_attr_npar_config))
dev_info(dev, "failed to create npar config sysfs entry");
- if (device_create_bin_file(dev, &bin_attr_esw_config))
- dev_info(dev, "failed to create esw config sysfs entry");
if (device_create_bin_file(dev, &bin_attr_pm_config))
dev_info(dev, "failed to create pm config sysfs entry");
if (device_create_bin_file(dev, &bin_attr_esw_stats))
@@ -3607,12 +3625,13 @@ qlcnic_remove_diag_entries(struct qlcnic_adapter *adapter)
device_remove_file(dev, &dev_attr_diag_mode);
device_remove_bin_file(dev, &bin_attr_crb);
device_remove_bin_file(dev, &bin_attr_mem);
- if (!(adapter->flags & QLCNIC_ESWITCH_ENABLED) ||
- adapter->op_mode != QLCNIC_MGMT_FUNC)
+ if (!(adapter->flags & QLCNIC_ESWITCH_ENABLED))
+ return;
+ device_remove_bin_file(dev, &bin_attr_esw_config);
+ if (adapter->op_mode != QLCNIC_MGMT_FUNC)
return;
device_remove_bin_file(dev, &bin_attr_pci_config);
device_remove_bin_file(dev, &bin_attr_npar_config);
- device_remove_bin_file(dev, &bin_attr_esw_config);
device_remove_bin_file(dev, &bin_attr_pm_config);
device_remove_bin_file(dev, &bin_attr_esw_stats);
}
--
1.6.0.2
^ permalink raw reply related
* [PATCH NEXT 07/12] qlcnic: fix npar state
From: Amit Kumar Salecha @ 2010-08-19 15:08 UTC (permalink / raw)
To: davem; +Cc: netdev, ameen.rahman, anirban.chakraborty
In-Reply-To: <1282230513-25286-1-git-send-email-amit.salecha@qlogic.com>
Privilege functions should wait for npar state to be operational.
Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com>
---
drivers/net/qlcnic/qlcnic_main.c | 43 +++++++++++++++++++++++++------------
1 files changed, 29 insertions(+), 14 deletions(-)
diff --git a/drivers/net/qlcnic/qlcnic_main.c b/drivers/net/qlcnic/qlcnic_main.c
index de4be00..6a8e970 100644
--- a/drivers/net/qlcnic/qlcnic_main.c
+++ b/drivers/net/qlcnic/qlcnic_main.c
@@ -916,6 +916,27 @@ qlcnic_reset_npar_config(struct qlcnic_adapter *adapter)
return 0;
}
+static int qlcnic_check_npar_opertional(struct qlcnic_adapter *adapter)
+{
+ u8 npar_opt_timeo = QLCNIC_DEV_NPAR_OPER_TIMEO;
+ u32 npar_state;
+
+ if (adapter->op_mode == QLCNIC_MGMT_FUNC)
+ return 0;
+
+ npar_state = QLCRD32(adapter, QLCNIC_CRB_DEV_NPAR_STATE);
+ while (npar_state != QLCNIC_DEV_NPAR_OPER && --npar_opt_timeo) {
+ msleep(1000);
+ npar_state = QLCRD32(adapter, QLCNIC_CRB_DEV_NPAR_STATE);
+ }
+ if (!npar_opt_timeo) {
+ dev_err(&adapter->pdev->dev,
+ "Waiting for NPAR state to opertional timeout\n");
+ return -EIO;
+ }
+ return 0;
+}
+
static int
qlcnic_start_firmware(struct qlcnic_adapter *adapter)
{
@@ -978,6 +999,11 @@ wait_init:
QLCWR32(adapter, QLCNIC_CRB_DEV_STATE, QLCNIC_DEV_READY);
qlcnic_idc_debug_info(adapter, 1);
+ err = qlcnic_check_npar_opertional(adapter);
+ if (err) {
+ qlcnic_release_firmware(adapter);
+ return err;
+ }
if (qlcnic_set_default_offload_settings(adapter))
goto err_out;
if (qlcnic_reset_npar_config(adapter))
@@ -2919,29 +2945,18 @@ static void qlcnic_io_resume(struct pci_dev *pdev)
FW_POLL_DELAY);
}
-
static int
qlcnicvf_start_firmware(struct qlcnic_adapter *adapter)
{
int err;
- u8 npar_opt_timeo = QLCNIC_DEV_NPAR_OPER_TIMEO;
- u32 npar_state;
err = qlcnic_can_start_firmware(adapter);
if (err)
return err;
- npar_state = QLCRD32(adapter, QLCNIC_CRB_DEV_NPAR_STATE);
- while (npar_state != QLCNIC_DEV_NPAR_OPER && --npar_opt_timeo) {
- msleep(1000);
- npar_state = QLCRD32(adapter, QLCNIC_CRB_DEV_NPAR_STATE);
- }
-
- if (!npar_opt_timeo) {
- dev_err(&adapter->pdev->dev,
- "Waiting for NPAR state to opertional timeout\n");
- return -EIO;
- }
+ err = qlcnic_check_npar_opertional(adapter);
+ if (err)
+ return err;
qlcnic_check_options(adapter);
--
1.6.0.2
^ permalink raw reply related
* [PATCH NEXT 08/12] qlcnic: mark device state as failed
From: Amit Kumar Salecha @ 2010-08-19 15:08 UTC (permalink / raw)
To: davem; +Cc: netdev, ameen.rahman, anirban.chakraborty
In-Reply-To: <1282230513-25286-1-git-send-email-amit.salecha@qlogic.com>
Mark device state failed in error path.
Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com>
---
drivers/net/qlcnic/qlcnic_main.c | 23 +++++++++++++----------
1 files changed, 13 insertions(+), 10 deletions(-)
diff --git a/drivers/net/qlcnic/qlcnic_main.c b/drivers/net/qlcnic/qlcnic_main.c
index 6a8e970..6e246c8 100644
--- a/drivers/net/qlcnic/qlcnic_main.c
+++ b/drivers/net/qlcnic/qlcnic_main.c
@@ -94,7 +94,7 @@ static void qlcnic_create_diag_entries(struct qlcnic_adapter *adapter);
static void qlcnic_remove_diag_entries(struct qlcnic_adapter *adapter);
static void qlcnic_idc_debug_info(struct qlcnic_adapter *adapter, u8 encoding);
-static void qlcnic_clr_all_drv_state(struct qlcnic_adapter *adapter);
+static void qlcnic_clr_all_drv_state(struct qlcnic_adapter *adapter, u8);
static int qlcnic_can_start_firmware(struct qlcnic_adapter *adapter);
static irqreturn_t qlcnic_tmp_intr(int irq, void *data);
@@ -1561,7 +1561,7 @@ err_out_disable_msi:
qlcnic_teardown_intr(adapter);
err_out_decr_ref:
- qlcnic_clr_all_drv_state(adapter);
+ qlcnic_clr_all_drv_state(adapter, 0);
err_out_iounmap:
qlcnic_cleanup_pci_map(adapter);
@@ -1600,7 +1600,7 @@ static void __devexit qlcnic_remove(struct pci_dev *pdev)
if (adapter->eswitch != NULL)
kfree(adapter->eswitch);
- qlcnic_clr_all_drv_state(adapter);
+ qlcnic_clr_all_drv_state(adapter, 0);
clear_bit(__QLCNIC_RESETTING, &adapter->state);
@@ -1632,7 +1632,7 @@ static int __qlcnic_shutdown(struct pci_dev *pdev)
if (netif_running(netdev))
qlcnic_down(adapter, netdev);
- qlcnic_clr_all_drv_state(adapter);
+ qlcnic_clr_all_drv_state(adapter, 0);
clear_bit(__QLCNIC_RESETTING, &adapter->state);
@@ -2379,7 +2379,7 @@ qlcnic_clr_drv_state(struct qlcnic_adapter *adapter)
}
static void
-qlcnic_clr_all_drv_state(struct qlcnic_adapter *adapter)
+qlcnic_clr_all_drv_state(struct qlcnic_adapter *adapter, u8 failed)
{
u32 val;
@@ -2390,7 +2390,11 @@ qlcnic_clr_all_drv_state(struct qlcnic_adapter *adapter)
QLC_DEV_CLR_REF_CNT(val, adapter->portnum);
QLCWR32(adapter, QLCNIC_CRB_DEV_REF_COUNT, val);
- if (!(val & 0x11111111))
+ if (failed) {
+ QLCWR32(adapter, QLCNIC_CRB_DEV_STATE, QLCNIC_DEV_FAILED);
+ dev_info(&adapter->pdev->dev,
+ "Device state set to Failed. Please Reboot\n");
+ } else if (!(val & 0x11111111))
QLCWR32(adapter, QLCNIC_CRB_DEV_STATE, QLCNIC_DEV_COLD);
val = QLCRD32(adapter, QLCNIC_CRB_DRV_STATE);
@@ -2605,7 +2609,7 @@ err_ret:
dev_err(&adapter->pdev->dev, "Fwinit work failed state=%u "
"fw_wait_cnt=%u\n", dev_state, adapter->fw_wait_cnt);
netif_device_attach(adapter->netdev);
- qlcnic_clr_all_drv_state(adapter);
+ qlcnic_clr_all_drv_state(adapter, 0);
}
static void
@@ -2641,8 +2645,7 @@ err_ret:
dev_err(&adapter->pdev->dev, "detach failed; status=%d temp=%d\n",
status, adapter->temp);
netif_device_attach(netdev);
- qlcnic_clr_all_drv_state(adapter);
-
+ qlcnic_clr_all_drv_state(adapter, 1);
}
/*Transit NPAR state to NON Operational */
@@ -2879,7 +2882,7 @@ static int qlcnic_attach_func(struct pci_dev *pdev)
if (netif_running(netdev)) {
err = qlcnic_attach(adapter);
if (err) {
- qlcnic_clr_all_drv_state(adapter);
+ qlcnic_clr_all_drv_state(adapter, 1);
clear_bit(__QLCNIC_AER, &adapter->state);
netif_device_attach(netdev);
return err;
--
1.6.0.2
^ permalink raw reply related
* [PATCH NEXT 02/12] qlcnic: remove unused code
From: Amit Kumar Salecha @ 2010-08-19 15:08 UTC (permalink / raw)
To: davem; +Cc: netdev, ameen.rahman, anirban.chakraborty, Sony Chacko
In-Reply-To: <1282230513-25286-1-git-send-email-amit.salecha@qlogic.com>
From: Sony Chacko <sony.chacko@qlogic.com>
Serial number references are not used in driver.
Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>
---
drivers/net/qlcnic/qlcnic.h | 1 -
drivers/net/qlcnic/qlcnic_main.c | 19 -------------------
2 files changed, 0 insertions(+), 20 deletions(-)
diff --git a/drivers/net/qlcnic/qlcnic.h b/drivers/net/qlcnic/qlcnic.h
index f6b887d..5c549c5 100644
--- a/drivers/net/qlcnic/qlcnic.h
+++ b/drivers/net/qlcnic/qlcnic.h
@@ -933,7 +933,6 @@ struct qlcnic_adapter {
u8 max_rds_rings;
u8 max_sds_rings;
- u8 driver_mismatch;
u8 msix_supported;
u8 rx_csum;
u8 portnum;
diff --git a/drivers/net/qlcnic/qlcnic_main.c b/drivers/net/qlcnic/qlcnic_main.c
index 384cc52..beadf2e 100644
--- a/drivers/net/qlcnic/qlcnic_main.c
+++ b/drivers/net/qlcnic/qlcnic_main.c
@@ -709,24 +709,8 @@ qlcnic_check_options(struct qlcnic_adapter *adapter)
{
u32 fw_major, fw_minor, fw_build;
char brd_name[QLCNIC_MAX_BOARD_NAME_LEN];
- char serial_num[32];
- int i, offset, val;
- int *ptr32;
struct pci_dev *pdev = adapter->pdev;
struct qlcnic_info nic_info;
- adapter->driver_mismatch = 0;
-
- ptr32 = (int *)&serial_num;
- offset = QLCNIC_FW_SERIAL_NUM_OFFSET;
- for (i = 0; i < 8; i++) {
- if (qlcnic_rom_fast_read(adapter, offset, &val) == -1) {
- dev_err(&pdev->dev, "error reading board info\n");
- adapter->driver_mismatch = 1;
- return;
- }
- ptr32[i] = cpu_to_le32(val);
- offset += sizeof(u32);
- }
fw_major = QLCRD32(adapter, QLCNIC_FW_VERSION_MAJOR);
fw_minor = QLCRD32(adapter, QLCNIC_FW_VERSION_MINOR);
@@ -1584,9 +1568,6 @@ static int qlcnic_open(struct net_device *netdev)
struct qlcnic_adapter *adapter = netdev_priv(netdev);
int err;
- if (adapter->driver_mismatch)
- return -EIO;
-
err = qlcnic_attach(adapter);
if (err)
return err;
--
1.6.0.2
^ permalink raw reply related
* [PATCH NEXT 05/12] qlcnic: configure offload setting on eswitch
From: Amit Kumar Salecha @ 2010-08-19 15:08 UTC (permalink / raw)
To: davem; +Cc: netdev, ameen.rahman, anirban.chakraborty, Rajesh Borundia
In-Reply-To: <1282230513-25286-1-git-send-email-amit.salecha@qlogic.com>
From: Rajesh Borundia <rajesh.borundia@qlogic.com>
Device is not capable of enabling/disabling offload setting per
port in case of Nic Partition.So offload settings needs to be
enabled/disabled per eswitch and it will affect all the function
on that eswitch.
Signed-off-by: Rajesh Borundia <rajesh.borundia@qlogic.com>
Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com>
---
drivers/net/qlcnic/qlcnic_ethtool.c | 18 +++++-
drivers/net/qlcnic/qlcnic_main.c | 121 +++++++++++++++++++++++++++++++++--
2 files changed, 133 insertions(+), 6 deletions(-)
diff --git a/drivers/net/qlcnic/qlcnic_ethtool.c b/drivers/net/qlcnic/qlcnic_ethtool.c
index e38fc3d..2805f88 100644
--- a/drivers/net/qlcnic/qlcnic_ethtool.c
+++ b/drivers/net/qlcnic/qlcnic_ethtool.c
@@ -804,6 +804,20 @@ qlcnic_get_ethtool_stats(struct net_device *dev,
}
}
+static int qlcnic_set_tx_csum(struct net_device *dev, u32 data)
+{
+ struct qlcnic_adapter *adapter = netdev_priv(dev);
+
+ if ((adapter->flags & QLCNIC_ESWITCH_ENABLED))
+ return -EOPNOTSUPP;
+ if (data)
+ dev->features |= (NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM);
+ else
+ dev->features &= ~(NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM);
+
+ return 0;
+
+}
static u32 qlcnic_get_tx_csum(struct net_device *dev)
{
return dev->features & NETIF_F_IP_CSUM;
@@ -819,6 +833,8 @@ static int qlcnic_set_rx_csum(struct net_device *dev, u32 data)
{
struct qlcnic_adapter *adapter = netdev_priv(dev);
+ if ((adapter->flags & QLCNIC_ESWITCH_ENABLED))
+ return -EOPNOTSUPP;
if (!!data) {
adapter->rx_csum = !!data;
return 0;
@@ -1070,7 +1086,7 @@ const struct ethtool_ops qlcnic_ethtool_ops = {
.get_pauseparam = qlcnic_get_pauseparam,
.set_pauseparam = qlcnic_set_pauseparam,
.get_tx_csum = qlcnic_get_tx_csum,
- .set_tx_csum = ethtool_op_set_tx_csum,
+ .set_tx_csum = qlcnic_set_tx_csum,
.set_sg = ethtool_op_set_sg,
.get_tso = qlcnic_get_tso,
.set_tso = qlcnic_set_tso,
diff --git a/drivers/net/qlcnic/qlcnic_main.c b/drivers/net/qlcnic/qlcnic_main.c
index 7bb3285..aa1f6b3 100644
--- a/drivers/net/qlcnic/qlcnic_main.c
+++ b/drivers/net/qlcnic/qlcnic_main.c
@@ -110,6 +110,8 @@ static void qlcnic_dev_set_npar_ready(struct qlcnic_adapter *);
static int qlcnicvf_config_led(struct qlcnic_adapter *, u32, u32);
static int qlcnicvf_config_bridged_mode(struct qlcnic_adapter *, u32);
static int qlcnicvf_start_firmware(struct qlcnic_adapter *);
+static void qlcnic_set_netdev_features(struct qlcnic_adapter *,
+ struct qlcnic_esw_func_cfg *);
/* PCI Device ID Table */
#define ENTRY(device) \
{PCI_DEVICE(PCI_VENDOR_ID_QLOGIC, (device)), \
@@ -756,6 +758,98 @@ qlcnic_check_options(struct qlcnic_adapter *adapter)
adapter->max_rds_rings = MAX_RDS_RINGS;
}
+static void
+qlcnic_set_eswitch_port_features(struct qlcnic_adapter *adapter,
+ struct qlcnic_esw_func_cfg *esw_cfg)
+{
+ qlcnic_set_netdev_features(adapter, esw_cfg);
+}
+
+static int
+qlcnic_set_eswitch_port_config(struct qlcnic_adapter *adapter)
+{
+ struct qlcnic_esw_func_cfg esw_cfg;
+
+ if (!(adapter->flags & QLCNIC_ESWITCH_ENABLED))
+ return 0;
+
+ esw_cfg.pci_func = adapter->ahw.pci_func;
+ if (qlcnic_get_eswitch_port_config(adapter, &esw_cfg))
+ return -EIO;
+ qlcnic_set_eswitch_port_features(adapter, &esw_cfg);
+
+ return 0;
+}
+
+static void
+qlcnic_set_netdev_features(struct qlcnic_adapter *adapter,
+ struct qlcnic_esw_func_cfg *esw_cfg)
+{
+ struct net_device *netdev = adapter->netdev;
+ unsigned long features, vlan_features;
+
+ features = (NETIF_F_SG | NETIF_F_IP_CSUM |
+ NETIF_F_IPV6_CSUM | NETIF_F_GRO);
+ vlan_features = (NETIF_F_SG | NETIF_F_IP_CSUM |
+ NETIF_F_IPV6_CSUM);
+
+ if (adapter->capabilities & QLCNIC_FW_CAPABILITY_TSO) {
+ features |= (NETIF_F_TSO | NETIF_F_TSO6);
+ vlan_features |= (NETIF_F_TSO | NETIF_F_TSO6);
+ }
+ if (adapter->capabilities & QLCNIC_FW_CAPABILITY_HW_LRO)
+ features |= NETIF_F_LRO;
+
+ if (esw_cfg->offload_flags & BIT_0) {
+ netdev->features |= features;
+ adapter->rx_csum = 1;
+ if (!(esw_cfg->offload_flags & BIT_1))
+ netdev->features &= ~NETIF_F_TSO;
+ if (!(esw_cfg->offload_flags & BIT_2))
+ netdev->features &= ~NETIF_F_TSO6;
+ } else {
+ netdev->features &= ~features;
+ adapter->rx_csum = 0;
+ }
+
+ netdev->vlan_features = (features & vlan_features);
+}
+
+static int
+qlcnic_set_default_offload_settings(struct qlcnic_adapter *adapter)
+{
+ struct qlcnic_esw_func_cfg esw_cfg;
+ struct qlcnic_npar_info *npar;
+ u8 i;
+
+ if (!(adapter->flags & QLCNIC_ESWITCH_ENABLED) ||
+ adapter->need_fw_reset ||
+ adapter->op_mode != QLCNIC_MGMT_FUNC)
+ return 0;
+
+ for (i = 0; i < QLCNIC_MAX_PCI_FUNC; i++) {
+ if (adapter->npars[i].type != QLCNIC_TYPE_NIC)
+ continue;
+ memset(&esw_cfg, 0, sizeof(struct qlcnic_esw_func_cfg));
+ esw_cfg.pci_func = i;
+ esw_cfg.offload_flags = BIT_0;
+ esw_cfg.mac_learning = BIT_0;
+ if (adapter->capabilities & QLCNIC_FW_CAPABILITY_TSO)
+ esw_cfg.offload_flags |= (BIT_1 | BIT_2);
+ if (qlcnic_config_switch_port(adapter, &esw_cfg))
+ return -EIO;
+ npar = &adapter->npars[i];
+ npar->pvid = esw_cfg.vlan_id;
+ npar->mac_learning = esw_cfg.offload_flags;
+ npar->mac_anti_spoof = esw_cfg.mac_anti_spoof;
+ npar->discard_tagged = esw_cfg.discard_tagged;
+ npar->promisc_mode = esw_cfg.promisc_mode;
+ npar->offload_flags = esw_cfg.offload_flags;
+ }
+
+ return 0;
+}
+
static int
qlcnic_reset_eswitch_config(struct qlcnic_adapter *adapter,
struct qlcnic_npar_info *npar, int pci_func)
@@ -879,6 +973,8 @@ wait_init:
QLCWR32(adapter, QLCNIC_CRB_DEV_STATE, QLCNIC_DEV_READY);
qlcnic_idc_debug_info(adapter, 1);
+ if (qlcnic_set_default_offload_settings(adapter))
+ goto err_out;
if (qlcnic_reset_npar_config(adapter))
goto err_out;
qlcnic_dev_set_npar_ready(adapter);
@@ -974,6 +1070,8 @@ __qlcnic_up(struct qlcnic_adapter *adapter, struct net_device *netdev)
if (test_bit(__QLCNIC_DEV_UP, &adapter->state))
return 0;
+ if (qlcnic_set_eswitch_port_config(adapter))
+ return -EIO;
if (qlcnic_fw_create_ctx(adapter))
return -EIO;
@@ -1291,7 +1389,6 @@ qlcnic_setup_netdev(struct qlcnic_adapter *adapter,
if (adapter->capabilities & QLCNIC_FW_CAPABILITY_HW_LRO)
netdev->features |= NETIF_F_LRO;
-
netdev->irq = adapter->msix_entries[0].vector;
if (qlcnic_read_mac_addr(adapter))
@@ -3216,7 +3313,7 @@ qlcnic_sysfs_write_esw_config(struct file *file, struct kobject *kobj,
struct qlcnic_esw_func_cfg *esw_cfg;
struct qlcnic_npar_info *npar;
int count, rem, i, ret;
- u8 pci_func;
+ u8 pci_func, op_mode = 0;
count = size / sizeof(struct qlcnic_esw_func_cfg);
rem = size % sizeof(struct qlcnic_esw_func_cfg);
@@ -3229,10 +3326,24 @@ qlcnic_sysfs_write_esw_config(struct file *file, struct kobject *kobj,
return ret;
for (i = 0; i < count; i++) {
- if (qlcnic_config_switch_port(adapter, &esw_cfg[i]))
- return QL_STATUS_INVALID_PARAM;
+ if (adapter->op_mode == QLCNIC_MGMT_FUNC)
+ if (qlcnic_config_switch_port(adapter, &esw_cfg[i]))
+ return QL_STATUS_INVALID_PARAM;
+ if (adapter->ahw.pci_func == esw_cfg[i].pci_func)
+ op_mode = esw_cfg[i].op_mode;
+ qlcnic_get_eswitch_port_config(adapter, &esw_cfg[i]);
+ esw_cfg[i].op_mode = op_mode;
+ esw_cfg[i].pci_func = adapter->ahw.pci_func;
+ switch (esw_cfg[i].op_mode) {
+ case QLCNIC_PORT_DEFAULTS:
+ qlcnic_set_eswitch_port_features(adapter,
+ &esw_cfg[i]);
+ break;
+ }
}
+ if (adapter->op_mode != QLCNIC_MGMT_FUNC)
+ goto out;
for (i = 0; i < count; i++) {
pci_func = esw_cfg[i].pci_func;
npar = &adapter->npars[pci_func];
@@ -3252,7 +3363,7 @@ qlcnic_sysfs_write_esw_config(struct file *file, struct kobject *kobj,
break;
}
}
-
+out:
return size;
}
--
1.6.0.2
^ permalink raw reply related
* [PATCH NEXT 01/12] qlcnic: fix inconsistent lock state
From: Amit Kumar Salecha @ 2010-08-19 15:08 UTC (permalink / raw)
To: davem; +Cc: netdev, ameen.rahman, anirban.chakraborty
In-Reply-To: <1282230513-25286-1-git-send-email-amit.salecha@qlogic.com>
Spin_lock(rds_ring->lock) is not required while posting buffers
from qlcnic_open and freeing buffers from qlcnic_down.
Signed-off-by: Amit Kumar Salecha <amit.salecha@qlogic.com>
---
drivers/net/qlcnic/qlcnic_init.c | 7 -------
1 files changed, 0 insertions(+), 7 deletions(-)
diff --git a/drivers/net/qlcnic/qlcnic_init.c b/drivers/net/qlcnic/qlcnic_init.c
index 75ba744..08da258 100644
--- a/drivers/net/qlcnic/qlcnic_init.c
+++ b/drivers/net/qlcnic/qlcnic_init.c
@@ -136,8 +136,6 @@ void qlcnic_reset_rx_buffers_list(struct qlcnic_adapter *adapter)
for (ring = 0; ring < adapter->max_rds_rings; ring++) {
rds_ring = &recv_ctx->rds_rings[ring];
- spin_lock(&rds_ring->lock);
-
INIT_LIST_HEAD(&rds_ring->free_list);
rx_buf = rds_ring->rx_buf_arr;
@@ -146,8 +144,6 @@ void qlcnic_reset_rx_buffers_list(struct qlcnic_adapter *adapter)
&rds_ring->free_list);
rx_buf++;
}
-
- spin_unlock(&rds_ring->lock);
}
}
@@ -1587,8 +1583,6 @@ qlcnic_post_rx_buffers(struct qlcnic_adapter *adapter, u32 ringid,
int producer, count = 0;
struct list_head *head;
- spin_lock(&rds_ring->lock);
-
producer = rds_ring->producer;
head = &rds_ring->free_list;
@@ -1618,7 +1612,6 @@ qlcnic_post_rx_buffers(struct qlcnic_adapter *adapter, u32 ringid,
writel((producer-1) & (rds_ring->num_desc-1),
rds_ring->crb_rcv_producer);
}
- spin_unlock(&rds_ring->lock);
}
static void
--
1.6.0.2
^ permalink raw reply related
* [PATCH NEXT 0/12]qlcnic: driver update
From: Amit Kumar Salecha @ 2010-08-19 15:08 UTC (permalink / raw)
To: davem; +Cc: netdev, ameen.rahman, anirban.chakraborty
Hi
Series of 12 patches to configure eswitch features and driver updates.
-Amit
^ permalink raw reply
* [PATCH v2] netfilter: save the hash of the tuple in the original direction for latter use
From: Changli Gao @ 2010-08-19 14:33 UTC (permalink / raw)
To: Patrick McHardy
Cc: David S. Miller, Eric Dumazet, Mathieu Desnoyers, netfilter-devel,
netdev, Changli Gao
Since we don't change the tuple in the original direction, we can save it
in ct->tuplehash[IP_CT_DIR_REPLY].hnode.pprev for __nf_conntrack_confirm()
use.
__hash_conntrack() is split into two steps: ____hash_conntrack() is used
to get the raw hash, and __hash_bucket() is used to get the bucket id.
In SYN-flood case, early_drop() doesn't need to recompute the hash again.
Signed-off-by: Changli Gao <xiaosuo@gmail.com>
---
v2: use cmpxchg() to save 2 variables.
net/netfilter/nf_conntrack_core.c | 114 ++++++++++++++++++++++++++------------
1 file changed, 79 insertions(+), 35 deletions(-)
diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
index df3eedb..0e02205 100644
--- a/net/netfilter/nf_conntrack_core.c
+++ b/net/netfilter/nf_conntrack_core.c
@@ -65,14 +65,20 @@ EXPORT_SYMBOL_GPL(nf_conntrack_max);
DEFINE_PER_CPU(struct nf_conn, nf_conntrack_untracked);
EXPORT_PER_CPU_SYMBOL(nf_conntrack_untracked);
-static int nf_conntrack_hash_rnd_initted;
-static unsigned int nf_conntrack_hash_rnd;
-
-static u_int32_t __hash_conntrack(const struct nf_conntrack_tuple *tuple,
- u16 zone, unsigned int size, unsigned int rnd)
+static u32 ____hash_conntrack(const struct nf_conntrack_tuple *tuple, u16 zone)
{
unsigned int n;
u_int32_t h;
+ static unsigned int rnd;
+
+ if (unlikely(!rnd)) {
+ unsigned int rand;
+
+ get_random_bytes(&rand, sizeof(rand));
+ if (!rand)
+ rand = 1;
+ cmpxchg(&rnd, 0, rand);
+ }
/* The direction must be ignored, so we hash everything up to the
* destination ports (which is a multiple of 4) and treat the last
@@ -83,14 +89,29 @@ static u_int32_t __hash_conntrack(const struct nf_conntrack_tuple *tuple,
zone ^ rnd ^ (((__force __u16)tuple->dst.u.all << 16) |
tuple->dst.protonum));
- return ((u64)h * size) >> 32;
+ return h;
+}
+
+static u32 __hash_bucket(u32 __hash, unsigned int size)
+{
+ return ((u64)__hash * size) >> 32;
+}
+
+static u32 hash_bucket(u32 __hash, const struct net *net)
+{
+ return __hash_bucket(__hash, net->ct.htable_size);
+}
+
+static u_int32_t __hash_conntrack(const struct nf_conntrack_tuple *tuple,
+ u16 zone, unsigned int size)
+{
+ return __hash_bucket(____hash_conntrack(tuple, zone), size);
}
static inline u_int32_t hash_conntrack(const struct net *net, u16 zone,
const struct nf_conntrack_tuple *tuple)
{
- return __hash_conntrack(tuple, zone, net->ct.htable_size,
- nf_conntrack_hash_rnd);
+ return __hash_conntrack(tuple, zone, net->ct.htable_size);
}
bool
@@ -292,13 +313,13 @@ static void death_by_timeout(unsigned long ul_conntrack)
* OR
* - Caller must lock nf_conntrack_lock before calling this function
*/
-struct nf_conntrack_tuple_hash *
-__nf_conntrack_find(struct net *net, u16 zone,
- const struct nf_conntrack_tuple *tuple)
+static struct nf_conntrack_tuple_hash *
+____nf_conntrack_find(struct net *net, u16 zone,
+ const struct nf_conntrack_tuple *tuple, u32 __hash)
{
struct nf_conntrack_tuple_hash *h;
struct hlist_nulls_node *n;
- unsigned int hash = hash_conntrack(net, zone, tuple);
+ unsigned int hash = hash_bucket(__hash, net);
/* Disable BHs the entire time since we normally need to disable them
* at least once for the stats anyway.
@@ -327,19 +348,27 @@ begin:
return NULL;
}
+
+struct nf_conntrack_tuple_hash *
+__nf_conntrack_find(struct net *net, u16 zone,
+ const struct nf_conntrack_tuple *tuple)
+{
+ return ____nf_conntrack_find(net, zone, tuple,
+ ____hash_conntrack(tuple, zone));
+}
EXPORT_SYMBOL_GPL(__nf_conntrack_find);
/* Find a connection corresponding to a tuple. */
-struct nf_conntrack_tuple_hash *
-nf_conntrack_find_get(struct net *net, u16 zone,
- const struct nf_conntrack_tuple *tuple)
+static struct nf_conntrack_tuple_hash *
+__nf_conntrack_find_get(struct net *net, u16 zone,
+ const struct nf_conntrack_tuple *tuple, u32 __hash)
{
struct nf_conntrack_tuple_hash *h;
struct nf_conn *ct;
rcu_read_lock();
begin:
- h = __nf_conntrack_find(net, zone, tuple);
+ h = ____nf_conntrack_find(net, zone, tuple, __hash);
if (h) {
ct = nf_ct_tuplehash_to_ctrack(h);
if (unlikely(nf_ct_is_dying(ct) ||
@@ -357,6 +386,14 @@ begin:
return h;
}
+
+struct nf_conntrack_tuple_hash *
+nf_conntrack_find_get(struct net *net, u16 zone,
+ const struct nf_conntrack_tuple *tuple)
+{
+ return __nf_conntrack_find_get(net, zone, tuple,
+ ____hash_conntrack(tuple, zone));
+}
EXPORT_SYMBOL_GPL(nf_conntrack_find_get);
static void __nf_conntrack_hash_insert(struct nf_conn *ct,
@@ -409,7 +446,8 @@ __nf_conntrack_confirm(struct sk_buff *skb)
return NF_ACCEPT;
zone = nf_ct_zone(ct);
- hash = hash_conntrack(net, zone, &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple);
+ /* reuse the __hash saved before */
+ hash = hash_bucket(*(unsigned long *)&ct->tuplehash[IP_CT_DIR_REPLY].hnnode.pprev, net);
repl_hash = hash_conntrack(net, zone, &ct->tuplehash[IP_CT_DIR_REPLY].tuple);
/* We're not in hash table, and we refuse to set up related
@@ -567,25 +605,20 @@ static noinline int early_drop(struct net *net, unsigned int hash)
return dropped;
}
-struct nf_conn *nf_conntrack_alloc(struct net *net, u16 zone,
- const struct nf_conntrack_tuple *orig,
- const struct nf_conntrack_tuple *repl,
- gfp_t gfp)
+static struct nf_conn *
+__nf_conntrack_alloc(struct net *net, u16 zone,
+ const struct nf_conntrack_tuple *orig,
+ const struct nf_conntrack_tuple *repl,
+ gfp_t gfp, u32 __hash)
{
struct nf_conn *ct;
- if (unlikely(!nf_conntrack_hash_rnd_initted)) {
- get_random_bytes(&nf_conntrack_hash_rnd,
- sizeof(nf_conntrack_hash_rnd));
- nf_conntrack_hash_rnd_initted = 1;
- }
-
/* We don't want any race condition at early drop stage */
atomic_inc(&net->ct.count);
if (nf_conntrack_max &&
unlikely(atomic_read(&net->ct.count) > nf_conntrack_max)) {
- unsigned int hash = hash_conntrack(net, zone, orig);
+ unsigned int hash = hash_bucket(__hash, net);
if (!early_drop(net, hash)) {
atomic_dec(&net->ct.count);
if (net_ratelimit())
@@ -616,7 +649,8 @@ struct nf_conn *nf_conntrack_alloc(struct net *net, u16 zone,
ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple = *orig;
ct->tuplehash[IP_CT_DIR_ORIGINAL].hnnode.pprev = NULL;
ct->tuplehash[IP_CT_DIR_REPLY].tuple = *repl;
- ct->tuplehash[IP_CT_DIR_REPLY].hnnode.pprev = NULL;
+ /* save __hash for reusing when confirming */
+ *(unsigned long *)(&ct->tuplehash[IP_CT_DIR_REPLY].hnnode.pprev) = __hash;
/* Don't set timer yet: wait for confirmation */
setup_timer(&ct->timeout, death_by_timeout, (unsigned long)ct);
write_pnet(&ct->ct_net, net);
@@ -643,6 +677,14 @@ out_free:
return ERR_PTR(-ENOMEM);
#endif
}
+
+struct nf_conn *nf_conntrack_alloc(struct net *net, u16 zone,
+ const struct nf_conntrack_tuple *orig,
+ const struct nf_conntrack_tuple *repl,
+ gfp_t gfp)
+{
+ return __nf_conntrack_alloc(net, zone, orig, repl, gfp, 0);
+}
EXPORT_SYMBOL_GPL(nf_conntrack_alloc);
void nf_conntrack_free(struct nf_conn *ct)
@@ -664,7 +706,7 @@ init_conntrack(struct net *net, struct nf_conn *tmpl,
struct nf_conntrack_l3proto *l3proto,
struct nf_conntrack_l4proto *l4proto,
struct sk_buff *skb,
- unsigned int dataoff)
+ unsigned int dataoff, u32 __hash)
{
struct nf_conn *ct;
struct nf_conn_help *help;
@@ -678,7 +720,8 @@ init_conntrack(struct net *net, struct nf_conn *tmpl,
return NULL;
}
- ct = nf_conntrack_alloc(net, zone, tuple, &repl_tuple, GFP_ATOMIC);
+ ct = __nf_conntrack_alloc(net, zone, tuple, &repl_tuple, GFP_ATOMIC,
+ __hash);
if (IS_ERR(ct)) {
pr_debug("Can't allocate conntrack.\n");
return (struct nf_conntrack_tuple_hash *)ct;
@@ -755,6 +798,7 @@ resolve_normal_ct(struct net *net, struct nf_conn *tmpl,
struct nf_conntrack_tuple_hash *h;
struct nf_conn *ct;
u16 zone = tmpl ? nf_ct_zone(tmpl) : NF_CT_DEFAULT_ZONE;
+ u32 __hash;
if (!nf_ct_get_tuple(skb, skb_network_offset(skb),
dataoff, l3num, protonum, &tuple, l3proto,
@@ -764,10 +808,11 @@ resolve_normal_ct(struct net *net, struct nf_conn *tmpl,
}
/* look for tuple match */
- h = nf_conntrack_find_get(net, zone, &tuple);
+ __hash = ____hash_conntrack(&tuple, zone);
+ h = __nf_conntrack_find_get(net, zone, &tuple, __hash);
if (!h) {
h = init_conntrack(net, tmpl, &tuple, l3proto, l4proto,
- skb, dataoff);
+ skb, dataoff, __hash);
if (!h)
return NULL;
if (IS_ERR(h))
@@ -1307,8 +1352,7 @@ int nf_conntrack_set_hashsize(const char *val, struct kernel_param *kp)
ct = nf_ct_tuplehash_to_ctrack(h);
hlist_nulls_del_rcu(&h->hnnode);
bucket = __hash_conntrack(&h->tuple, nf_ct_zone(ct),
- hashsize,
- nf_conntrack_hash_rnd);
+ hashsize);
hlist_nulls_add_head_rcu(&h->hnnode, &hash[bucket]);
}
}
^ permalink raw reply related
* Re: [patch 04/20] ibmveth: Add tx_copybreak
From: Brian King @ 2010-08-19 13:34 UTC (permalink / raw)
To: Anton Blanchard; +Cc: Santiago Leon, netdev
In-Reply-To: <20100819072503.937364037@samba.org>
On 08/19/2010 02:25 AM, Anton Blanchard wrote:
> Use the existing bounce buffer if we send a buffer under a certain size.
> This saves the overhead of a TCE map/unmap.
>
> I can't see any reason for the wmb() in the bounce buffer case, if we need
> a barrier it will be before we call h_send_logical_lan but we have
> nothing in the common case. Remove it.
>
> Signed-off-by: Anton Blanchard <anton@samba.org>
> ---
>
> Index: powerpc.git/drivers/net/ibmveth.c
> ===================================================================
> --- powerpc.git.orig/drivers/net/ibmveth.c 2010-08-10 18:35:52.293272470 +1000
> +++ powerpc.git/drivers/net/ibmveth.c 2010-08-10 18:35:55.440741255 +1000
> @@ -117,6 +117,11 @@ MODULE_DESCRIPTION("IBM i/pSeries Virtua
> MODULE_LICENSE("GPL");
> MODULE_VERSION(ibmveth_driver_version);
>
> +static unsigned int tx_copybreak __read_mostly = 128;
Where did the 128 come from? I thought I heard the 2k value discussed in
the past.
Thanks,
Brian
--
Brian King
Linux on Power Virtualization
IBM Linux Technology Center
^ permalink raw reply
* Re: [patch] qlge: pull NULL check ahead of dereference
From: Ron Mercer @ 2010-08-19 12:59 UTC (permalink / raw)
To: Dan Carpenter, Linux Driver, David S. Miller, Breno Leitao,
"netdev@vger.kernel.org"
In-Reply-To: <20100819121942.GA14895@linux-ox1b.qlogic.org>
On Thu, Aug 19, 2010 at 05:19:42AM -0700, Ron Mercer wrote:
> On Thu, Aug 19, 2010 at 03:02:15AM -0700, Dan Carpenter wrote:
> > There was a dereference before NULL check issue introduced in 1e213303d
> > "qlge: Add tx multiqueue support." I've pulled the NULL check of
> > "net_rsp" forward a couple lines to restore the original semantics.
> >
>
> net_rsp is initialized to NULL and will never be NULL if entries are
> processed. It should be fine as it is currently coded.
>
On second thought you are right but the early exit should be above the index
write.
ql_write_cq_idx(rx_ring);
^ 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