* Re: [PATCH RFC]: napi_struct V4
From: David Miller @ 2007-07-26 0:37 UTC (permalink / raw)
To: shemminger; +Cc: netdev, jgarzik, hadi, rusty
In-Reply-To: <20070725095654.38a10abc@oldman.hamilton.local>
From: Stephen Hemminger <shemminger@linux-foundation.org>
Date: Wed, 25 Jul 2007 09:56:54 +0100
> The usage of NAPI on 8139cp and 8139too seems dodgy;
> these drivers expect this to work:
>
> local_irq_save(flags);
> cpw16_f(IntrMask, cp_intr_mask);
> __netif_rx_complete(dev);
> local_irq_restore(flags);
>
> It works on SMP only because if poll races with IRQ,
> the IRQ is not masked or cleared so the IRQ will get restarted.
>
> Better would be to change it to:
> spin_lock_irqsave(&cp->lock, flags);
> cpw16_f(IntrMask, cp_intr_mask);
> __netif_rx_complete(dev);
> spin_unlock_irqrestore(&cp->lock, flags);
>
> Which actually is same code on UP.
I've made these fixes to my tree, thanks Stephen.
^ permalink raw reply
* [PATCH 2.6.22 2/3]S2io: Fix crash when resetting adapter
From: Ramkrishna Vepa @ 2007-07-26 2:40 UTC (permalink / raw)
To: netdev; +Cc: jeff, ram.vepa
- Removed the call to pci_set_power_state to reset the adapter as it was resulting
in system crash on some platforms.
(Resending; Removed HTML sections in the patch)
Signed-off-by: Santosh Rastapur <santosh.rastapur@neterion.com>
---
diff -Nurp 2.0.24.1P1/drivers/net/s2io.c 2.0.24.1P2/drivers/net/s2io.c
--- 2.0.24.1P1/drivers/net/s2io.c 2007-07-20 16:16:01.000000000 -0700
+++ 2.0.24.1P2/drivers/net/s2io.c 2007-07-20 16:26:13.000000000 -0700
@@ -3382,23 +3382,8 @@ static void s2io_reset(struct s2io_nic *
/* Back up the PCI-X CMD reg, dont want to lose MMRBC, OST settings */
pci_read_config_word(sp->pdev, PCIX_COMMAND_REGISTER, &(pci_cmd));
- if (sp->device_type == XFRAME_II_DEVICE) {
- int ret;
- ret = pci_set_power_state(sp->pdev, 3);
- if (!ret)
- ret = pci_set_power_state(sp->pdev, 0);
- else {
- DBG_PRINT(ERR_DBG,"%s PME based SW_Reset failed!\n",
- __FUNCTION__);
- goto old_way;
- }
- msleep(20);
- goto new_way;
- }
-old_way:
val64 = SW_RESET_ALL;
writeq(val64, &bar0->sw_reset);
-new_way:
if (strstr(sp->product_name, "CX4")) {
msleep(750);
}
^ permalink raw reply
* Re: [PATCH][v2] Netfilter Kconfig: Expose IPv4/6 connection tracking options by selecting NF_CONNTRACK_ENABLED
From: Patrick McHardy @ 2007-07-26 0:46 UTC (permalink / raw)
To: Al Boldi; +Cc: netfilter-devel, netdev, Sam Ravnborg
In-Reply-To: <200707251802.22200.a1426z@gawab.com>
[Removed a few CCs]
Al Boldi wrote:
> Make NF_CONNTRACK_IPV4 and NF_CONNTRACK_IPV6 select NF_CONNTRACK_ENABLED.
One thought that occured to me after the last of many false bugreports
that were actually caused by failure to configure the new options
properly. Most people know they want NF_CONNTRACK (and its selected by
default with old configs), what they're missing is that they now also
need to select IPv4 connection tracking. So what would really make sense
is to make NF_CONNTRACK_IPV4 default to "m" (and really *everyone*
using conntrack wants this). But with your proposed change this would
default to selecting NF_CONNTRACK by default, which I'm not so sure
is a good idea. So I'm leaning towards just using "m" as default for
IPv4 conntrack to save people trouble and myself some bugreports, but
I also like your simplification ...
Maybe we can do something to have the NF_CONNTRACK_ENABLED option select
NF_CONNTRACK_IPV4 (which really is what we actually want) and combine
that with automatic selection of NF_CONNTRACK? I believe the only case
with negative impact would be people that currently use only IPv6
connection tracking, which is most likely nobody.
^ permalink raw reply
* [PATCH 2.6.22 3/3]S2io: Increment received packet count correctly
From: Ramkrishna Vepa @ 2007-07-26 2:43 UTC (permalink / raw)
To: netdev; +Cc: jeff
- Fix to increment the received packet count correctly.
(Resending; Removed HTML sections in the patch)
Signed-off-by: Santosh Rastapur <santosh.rastapur@neterion.com>
---
diff -Nurp 2.0.24.1P2/drivers/net/s2io.c 2.0.24.1P3/drivers/net/s2io.c
--- 2.0.24.1P2/drivers/net/s2io.c 2007-07-20 16:26:13.000000000 -0700
+++ 2.0.24.1P3/drivers/net/s2io.c 2007-07-20 16:40:02.000000000 -0700
@@ -84,7 +84,7 @@
#include "s2io.h"
#include "s2io-regs.h"
-#define DRV_VERSION "2.0.24.1"
+#define DRV_VERSION "2.0.25.1"
/* S2io Driver name & version. */
static char s2io_driver_name[] = "Neterion";
@@ -6734,6 +6734,7 @@ static int rx_osm_handler(struct ring_in
}
/* Updating statistics */
+ sp->stats.rx_packets++;
rxdp->Host_Control = 0;
if (sp->rxd_mode == RXD_MODE_1) {
int len = RXD_GET_BUFFER0_SIZE_1(rxdp->Control_2);
^ permalink raw reply
* [PATCH 2.6.22 1/3]S2io: Mask spurious interrupts
From: Ramkrishna Vepa @ 2007-07-26 2:35 UTC (permalink / raw)
To: netdev; +Cc: jeff, ram.vepa
- Mask single and double bit ETQ ecc errors to inhibit spurious interrupts.
(Resending; Removed HTML sections in the patch)
Signed-off-by: Santosh Rastapur <santosh.rastapur@neterion.com>
---
diff -Nurp 2.0.24.1/drivers/net/s2io-regs.h 2.0.24.1P1/drivers/net/s2io-regs.h
--- 2.0.24.1/drivers/net/s2io-regs.h 2007-07-20 16:13:29.000000000 -0700
+++ 2.0.24.1P1/drivers/net/s2io-regs.h 2007-07-20 16:18:27.000000000 -0700
@@ -747,10 +747,9 @@ struct XENA_dev_config {
#define MC_ERR_REG_MIRI_CRI_ERR_1 BIT(23)
#define MC_ERR_REG_SM_ERR BIT(31)
#define MC_ERR_REG_ECC_ALL_SNG (BIT(2) | BIT(3) | BIT(4) | BIT(5) |\
- BIT(6) | BIT(7) | BIT(17) | BIT(19))
+ BIT(17) | BIT(19))
#define MC_ERR_REG_ECC_ALL_DBL (BIT(10) | BIT(11) | BIT(12) |\
- BIT(13) | BIT(14) | BIT(15) |\
- BIT(18) | BIT(20))
+ BIT(13) | BIT(18) | BIT(20))
u64 mc_err_mask;
u64 mc_err_alarm;
^ permalink raw reply
* Re: [PATCH] IPv6: ipv6_addr_type() doesn't know about RFC4193 addresses
From: Dave Johnson @ 2007-07-26 1:11 UTC (permalink / raw)
To: David Miller; +Cc: netdev, linux-kernel, lksctp-developers, sakkiped
In-Reply-To: <20070725.171203.102545287.davem@davemloft.net>
David Miller writes:
> Contrarily, there may be ipv6_addr_type() call sites that really
> do want to reject rfc4193 addresses.
A quick look through the callers and only these functions should be
effected, they check either RESERVED or UNICAST from ipv6_addr_type():
net/ipv6/addrconf.c: ipv6_dev_get_saddr()
net/ipv6/exthdrs.c: ipv6_dest_hao()
net/ipv6/ip6_tunnel.c: ip6_tnl_set_cap()
net/ipv6/netfilter/ip6t_REJECT.c: send_reset()
net/ipv6/route.c: ip6_route_add()
net/ipv6/route.c: ip6_pkt_drop()
net/sctp/ipv6.c: sctp_v6_available()
net/sctp/ipv6.c: sctp_v6_addr_valid()
--
Dave Johnson
Starent Networks
^ permalink raw reply
* Re: [PATCH][v2] Netfilter Kconfig: Expose IPv4/6 connection tracking options by selecting NF_CONNTRACK_ENABLED
From: Yasuyuki KOZAKAI @ 2007-07-26 1:18 UTC (permalink / raw)
To: kaber; +Cc: netdev, a1426z, netfilter-devel, sam
In-Reply-To: <46A7EECD.9070801@trash.net>
From: Patrick McHardy <kaber@trash.net>
Date: Thu, 26 Jul 2007 02:46:05 +0200
> [Removed a few CCs]
>
> Al Boldi wrote:
> > Make NF_CONNTRACK_IPV4 and NF_CONNTRACK_IPV6 select NF_CONNTRACK_ENABLED.
>
>
> One thought that occured to me after the last of many false bugreports
> that were actually caused by failure to configure the new options
> properly. Most people know they want NF_CONNTRACK (and its selected by
> default with old configs), what they're missing is that they now also
> need to select IPv4 connection tracking. So what would really make sense
> is to make NF_CONNTRACK_IPV4 default to "m" (and really *everyone*
> using conntrack wants this). But with your proposed change this would
> default to selecting NF_CONNTRACK by default, which I'm not so sure
> is a good idea. So I'm leaning towards just using "m" as default for
> IPv4 conntrack to save people trouble and myself some bugreports, but
> I also like your simplification ...
>
> Maybe we can do something to have the NF_CONNTRACK_ENABLED option select
> NF_CONNTRACK_IPV4 (which really is what we actually want) and combine
> that with automatic selection of NF_CONNTRACK? I believe the only case
> with negative impact would be people that currently use only IPv6
> connection tracking, which is most likely nobody.
I agree. I've not heard trouble with NF_CONNTRACK_IPV6. I think that is
because it is purely new feature.
BTW, it's too late to restore IP_NF_CONNTRACK in stable and current tree
for a while ?
-- Yasuyuki Kozakai
^ permalink raw reply
* Re: [PATCH][v2] Netfilter Kconfig: Expose IPv4/6 connection tracking options by selecting NF_CONNTRACK_ENABLED
From: Yasuyuki KOZAKAI @ 2007-07-26 1:18 UTC (permalink / raw)
To: kaber; +Cc: a1426z, netdev, netfilter-devel, sam
In-Reply-To: <46A7EECD.9070801@trash.net>
From: Patrick McHardy <kaber@trash.net>
Date: Thu, 26 Jul 2007 02:46:05 +0200
> [Removed a few CCs]
>
> Al Boldi wrote:
> > Make NF_CONNTRACK_IPV4 and NF_CONNTRACK_IPV6 select NF_CONNTRACK_ENABLED.
>
>
> One thought that occured to me after the last of many false bugreports
> that were actually caused by failure to configure the new options
> properly. Most people know they want NF_CONNTRACK (and its selected by
> default with old configs), what they're missing is that they now also
> need to select IPv4 connection tracking. So what would really make sense
> is to make NF_CONNTRACK_IPV4 default to "m" (and really *everyone*
> using conntrack wants this). But with your proposed change this would
> default to selecting NF_CONNTRACK by default, which I'm not so sure
> is a good idea. So I'm leaning towards just using "m" as default for
> IPv4 conntrack to save people trouble and myself some bugreports, but
> I also like your simplification ...
>
> Maybe we can do something to have the NF_CONNTRACK_ENABLED option select
> NF_CONNTRACK_IPV4 (which really is what we actually want) and combine
> that with automatic selection of NF_CONNTRACK? I believe the only case
> with negative impact would be people that currently use only IPv6
> connection tracking, which is most likely nobody.
I agree. I've not heard trouble with NF_CONNTRACK_IPV6. I think that is
because it is purely new feature.
BTW, it's too late to restore IP_NF_CONNTRACK in stable and current tree
for a while ?
-- Yasuyuki Kozakai
^ permalink raw reply
* Re: [PATCH RFC]: napi_struct V4
From: Jeff Garzik @ 2007-07-26 1:55 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: David Miller, netdev, hadi, rusty
In-Reply-To: <20070725095654.38a10abc@oldman.hamilton.local>
Stephen Hemminger wrote:
> The usage of NAPI on 8139cp and 8139too seems dodgy;
> these drivers expect this to work:
>
> local_irq_save(flags);
> cpw16_f(IntrMask, cp_intr_mask);
> __netif_rx_complete(dev);
> local_irq_restore(flags);
>
> It works on SMP only because if poll races with IRQ,
> the IRQ is not masked or cleared so the IRQ will get restarted.
Expect? This _does_ work on both UP and SMP.
> Better would be to change it to:
> spin_lock_irqsave(&cp->lock, flags);
> cpw16_f(IntrMask, cp_intr_mask);
> __netif_rx_complete(dev);
> spin_unlock_irqrestore(&cp->lock, flags);
>
> Which actually is same code on UP.
I'm missing your point?
You claim local_irq_save() works only on SMP, then suggest a solution
that you agree produces the same result?
I don't see any logic to your request, only added overhead for no reason.
Jeff
^ permalink raw reply
* Re: [PATCH RFC]: napi_struct V4
From: David Miller @ 2007-07-26 1:56 UTC (permalink / raw)
To: jgarzik; +Cc: shemminger, netdev, hadi, rusty
In-Reply-To: <46A7FEFC.4040005@pobox.com>
From: Jeff Garzik <jgarzik@pobox.com>
Date: Wed, 25 Jul 2007 21:55:08 -0400
> I don't see any logic to your request, only added overhead for no reason.
There may be some flawed logic in what Stephen stated, but
the change really is needed.
It must be atomic to execute the:
enable_interrupts();
netif_rx_complete();
sequence wrt. the same code path in the interrupt handler.
^ permalink raw reply
* Re: [PATCH RFC]: napi_struct V4
From: Jeff Garzik @ 2007-07-26 2:00 UTC (permalink / raw)
To: David Miller; +Cc: shemminger, netdev, hadi, rusty
In-Reply-To: <20070725.185632.116383625.davem@davemloft.net>
David Miller wrote:
> From: Jeff Garzik <jgarzik@pobox.com>
> Date: Wed, 25 Jul 2007 21:55:08 -0400
>
>> I don't see any logic to your request, only added overhead for no reason.
>
> There may be some flawed logic in what Stephen stated, but
> the change really is needed.
>
> It must be atomic to execute the:
>
> enable_interrupts();
> netif_rx_complete();
>
> sequence wrt. the same code path in the interrupt handler.
Sure. And how did the existing code fail to achieve that?
Jeff
^ permalink raw reply
* Re: [PATCH RFC]: napi_struct V4
From: David Miller @ 2007-07-26 2:02 UTC (permalink / raw)
To: jeff; +Cc: shemminger, netdev, hadi, rusty
In-Reply-To: <46A8003F.1050208@garzik.org>
From: Jeff Garzik <jeff@garzik.org>
Date: Wed, 25 Jul 2007 22:00:31 -0400
> David Miller wrote:
> > From: Jeff Garzik <jgarzik@pobox.com>
> > Date: Wed, 25 Jul 2007 21:55:08 -0400
> >
> >> I don't see any logic to your request, only added overhead for no reason.
> >
> > There may be some flawed logic in what Stephen stated, but
> > the change really is needed.
> >
> > It must be atomic to execute the:
> >
> > enable_interrupts();
> > netif_rx_complete();
> >
> > sequence wrt. the same code path in the interrupt handler.
>
> Sure. And how did the existing code fail to achieve that?
The interrupt handler can run on another cpu in betwen those two
statements, running the NAPI test-and-do-something operations in
parallel with the netif_rx_complete() which causes problems as Rusty
and I discussed yesterday.
^ permalink raw reply
* Re: [DRIVER SUBMISSION] DRBD wants to go mainline
From: Kyle Moffett @ 2007-07-26 3:43 UTC (permalink / raw)
To: david
Cc: Satyam Sharma, Lars Ellenberg, Jens Axboe, Andrew Morton,
LKML Kernel, netdev
In-Reply-To: <Pine.LNX.4.64.0707251900410.21378@asgard.lang.hm>
On Jul 25, 2007, at 22:03:37, david@lang.hm wrote:
> On Wed, 25 Jul 2007, Satyam Sharma wrote:
>> On 7/25/07, Lars Ellenberg <lars.ellenberg@linbit.com> wrote:
>>> On Wed, Jul 25, 2007 at 04:41:53AM +0530, Satyam Sharma wrote:
>>>> [...]
>>>> But where does the "send" come into the picture over here -- a
>>>> send won't block forever, so I don't foresee any issues
>>>> whatsoever w.r.t. kthreads conversion for that. [ BTW I hope
>>>> you're *not* using any signals-based interface for your kernel
>>>> thread _at all_. Kthreads disallow (ignore) all signals by
>>>> default, as they should, and you really shouldn't need to write
>>>> any logic to handle or > do-certain-things-on-seeing a signal
>>>> in a well designed kernel thread. ] and the sending latency is
>>>> crucial to performance, while the recv will not timeout for the
>>>> next few seconds. Again, I don't see what sending latency has
>>>> to do with a kernel_thread to kthread conversion. Or with
>>>> signals, for that matter. Anyway, as Kyle Moffett mentioned
>>>> elsewhere, you could probably look at other examples (say
>>>> cifs_demultiplexer_thread() in fs/cifs/connect.c).
>>>
>>> the basic problem, and what we use signals for, is: it is
>>> waiting in recv, waiting for the peer to say something. but I
>>> want it to stop recv, and go send something "right now".
>>
>> That's ... weird. Most (all?) communication between any two
>> parties would follow a protocol where someone recv's stuff, does
>> something with it, and sends it back ... what would you send
>> "right now" if you didn't receive anything?
>
> becouse even though you didn't receive anything you now have
> something important to send.
>
> remember that both sides can be sitting in receive mode. this puts
> them both in a position to respond to the other if the other has
> something to say.
Why not just have 2 threads, one for "sending" and one for
"receiving". When your receiving thread gets data it takes
appropriate locks and processes it, then releases the locks and goes
back to waiting for packets. Your sending thread would take
appropriate locks, generate data to send, release locks, and transmit
packets. You don't have to interrupt the receive thread to send
packets, so where's the latency problem, exactly?
If I were writing that in userspace I would have:
(A) The pool of IO-generating threads (IE: What would ordinarily be
userspace)
(B) One or a small number of data-reception threads.
(C) One or a small number of data-transmission threads.
When you get packets to process in your network-reception thread(s),
you queue appropriate disk IOs and any appropriate responses with
your transmission thread(s). You can basically just sit in a loop on
tcp_recvmsg=>demultiplex=>do-stuff. When your IO-generators actually
make stuff to send you queue such data for disk IO, then packetize it
and hand it off to your data-transmission threads.
If you made all your sockets and inter-thread pipes nonblocking then
in userspace you would just epoll_wait() on the sockets and pipes and
be easily able to react to any IO from anywhere.
In kernel space there are similar nonblocking interfaces, although it
would probably be easier just to use a couple threads.
Cheers,
Kyle Moffett
^ permalink raw reply
* Re: [PATCH][v2] Netfilter Kconfig: Expose IPv4/6 connection tracking options by selecting NF_CONNTRACK_ENABLED
From: Al Boldi @ 2007-07-26 3:53 UTC (permalink / raw)
To: Patrick McHardy; +Cc: netfilter-devel, netdev, Sam Ravnborg
In-Reply-To: <46A7EECD.9070801@trash.net>
Patrick McHardy wrote:
> Al Boldi wrote:
> > Make NF_CONNTRACK_IPV4 and NF_CONNTRACK_IPV6 select
> > NF_CONNTRACK_ENABLED.
>
> One thought that occured to me after the last of many false bugreports
> that were actually caused by failure to configure the new options
> properly. Most people know they want NF_CONNTRACK (and its selected by
> default with old configs), what they're missing is that they now also
> need to select IPv4 connection tracking. So what would really make sense
> is to make NF_CONNTRACK_IPV4 default to "m" (and really *everyone*
> using conntrack wants this). But with your proposed change this would
> default to selecting NF_CONNTRACK by default, which I'm not so sure
> is a good idea.
Making NF_CONNTRACK_IPV4 default to "m" would select NF_CONNTRACK to "m" if
it hasn't been selected by the user to be "y", which seems reasonable.
> So I'm leaning towards just using "m" as default for
> IPv4 conntrack to save people trouble and myself some bugreports, but
> I also like your simplification ...
I was also planning to submit another patch to make all netfilter
childoptions options default to their parent, i.e: NF_CONNTRACK_FTP would
default NF_CONNTRACK. This could be one big Kconfig time-saver.
> Maybe we can do something to have the NF_CONNTRACK_ENABLED option select
> NF_CONNTRACK_IPV4 (which really is what we actually want) and combine
> that with automatic selection of NF_CONNTRACK? I believe the only case
> with negative impact would be people that currently use only IPv6
> connection tracking, which is most likely nobody.
I think that wouldn't be advisable, as this would add an unnecessary
dependency. But of course, it's your call...
Thanks!
--
Al
^ permalink raw reply
* Re: [PATCH RFC]: napi_struct V4
From: David Miller @ 2007-07-26 4:09 UTC (permalink / raw)
To: netdev; +Cc: shemminger, jgarzik, hadi, rusty
In-Reply-To: <20070725.013154.34764933.davem@davemloft.net>
From: David Miller <davem@davemloft.net>
Date: Wed, 25 Jul 2007 01:31:54 -0700 (PDT)
> Besides that the only major issue is netpoll and I have some
> ideas on how to handle that, which I'll try to implement
> tonight and tomorrow.
Ok, here is how I'm trying to crack this nut.
The way I'll solve this is to just change the problem space
and therefore simplify things enough that it's actually sane
and implementable.
None of the receive packet handling of netpoll is even needed,
so we rip all of it out. Every use was out of tree, and now
subsumed by another facility. Basically that was kdump, which
is now done via kexec.
As a result there is no more messing around with fake NAPI polls and
all that other crap, instead ->poll_controller() merely has to try to
process TX queue acks to free up space and wake up the transmit
queue(s) of the device.
The downside is that TX ack processing has to run with the
netif_tx_lock() for this to work out. Basically, netpoll carefully
grabs the TX lock first by disabling interrupts and using
netif_tx_trylock(), if successful it does dev->poll_controller() and
then tries to transmit the packet.
Non-NAPI driver should be able to continue functioning as-is,
they just run their interrupt handler which doesn't schedule
polls and will do the TX queue acking.
NAPI drivers do need to change their ->poll_controller() to do TX
queue acking directly.
I wanted to do this in a way that didn't require touching non-NAPI
drivers as this change is getting big enough as it is and the
change is incomplete if it makes netconsole totally broken in the
process.
Anyways, here are snippets of patch from my current tree that should
show generally what is going on, with bnx2 and tg3 as two converted
NAPI ->poll_controler() examples. I'm about to boot this up and play
with netconsole to test out that it works properly.
diff --git a/drivers/net/bnx2.c b/drivers/net/bnx2.c
index a729da0..5187aaf 100644
--- a/drivers/net/bnx2.c
+++ b/drivers/net/bnx2.c
@@ -428,7 +428,7 @@ bnx2_netif_stop(struct bnx2 *bp)
{
bnx2_disable_int_sync(bp);
if (netif_running(bp->dev)) {
- netif_poll_disable(bp->dev);
+ napi_disable(&bp->napi);
netif_tx_disable(bp->dev);
bp->dev->trans_start = jiffies; /* prevent tx timeout */
}
@@ -440,7 +440,7 @@ bnx2_netif_start(struct bnx2 *bp)
if (atomic_dec_and_test(&bp->intr_sem)) {
if (netif_running(bp->dev)) {
netif_wake_queue(bp->dev);
- netif_poll_enable(bp->dev);
+ napi_enable(&bp->napi);
bnx2_enable_int(bp);
}
}
@@ -2268,7 +2268,7 @@ bnx2_phy_int(struct bnx2 *bp)
}
static void
-bnx2_tx_int(struct bnx2 *bp)
+__bnx2_tx_int(struct bnx2 *bp)
{
struct status_block *sblk = bp->status_blk;
u16 hw_cons, sw_cons, sw_ring_cons;
@@ -2347,11 +2347,9 @@ bnx2_tx_int(struct bnx2 *bp)
if (unlikely(netif_queue_stopped(bp->dev)) &&
(bnx2_tx_avail(bp) > bp->tx_wake_thresh)) {
- netif_tx_lock(bp->dev);
if ((netif_queue_stopped(bp->dev)) &&
(bnx2_tx_avail(bp) > bp->tx_wake_thresh))
netif_wake_queue(bp->dev);
- netif_tx_unlock(bp->dev);
}
}
@@ -2551,7 +2549,7 @@ bnx2_msi(int irq, void *dev_instance)
if (unlikely(atomic_read(&bp->intr_sem) != 0))
return IRQ_HANDLED;
- netif_rx_schedule(dev);
+ netif_rx_schedule(dev, &bp->napi);
return IRQ_HANDLED;
}
@@ -2568,7 +2566,7 @@ bnx2_msi_1shot(int irq, void *dev_instance)
if (unlikely(atomic_read(&bp->intr_sem) != 0))
return IRQ_HANDLED;
- netif_rx_schedule(dev);
+ netif_rx_schedule(dev, &bp->napi);
return IRQ_HANDLED;
}
@@ -2604,9 +2602,9 @@ bnx2_interrupt(int irq, void *dev_instance)
if (unlikely(atomic_read(&bp->intr_sem) != 0))
return IRQ_HANDLED;
- if (netif_rx_schedule_prep(dev)) {
+ if (netif_rx_schedule_prep(dev, &bp->napi)) {
bp->last_status_idx = sblk->status_idx;
- __netif_rx_schedule(dev);
+ __netif_rx_schedule(dev, &bp->napi);
}
return IRQ_HANDLED;
@@ -2632,12 +2630,14 @@ bnx2_has_work(struct bnx2 *bp)
}
static int
-bnx2_poll(struct net_device *dev, int *budget)
+bnx2_poll(struct napi_struct *napi, int budget)
{
- struct bnx2 *bp = netdev_priv(dev);
+ struct bnx2 *bp = container_of(napi, struct bnx2, napi);
+ struct net_device *dev = bp->dev;
struct status_block *sblk = bp->status_blk;
u32 status_attn_bits = sblk->status_attn_bits;
u32 status_attn_bits_ack = sblk->status_attn_bits_ack;
+ int work_done = 0;
if ((status_attn_bits & STATUS_ATTN_EVENTS) !=
(status_attn_bits_ack & STATUS_ATTN_EVENTS)) {
@@ -2652,26 +2652,20 @@ bnx2_poll(struct net_device *dev, int *budget)
REG_RD(bp, BNX2_HC_COMMAND);
}
- if (bp->status_blk->status_tx_quick_consumer_index0 != bp->hw_tx_cons)
- bnx2_tx_int(bp);
-
- if (bp->status_blk->status_rx_quick_consumer_index0 != bp->hw_rx_cons) {
- int orig_budget = *budget;
- int work_done;
-
- if (orig_budget > dev->quota)
- orig_budget = dev->quota;
-
- work_done = bnx2_rx_int(bp, orig_budget);
- *budget -= work_done;
- dev->quota -= work_done;
+ if (bp->status_blk->status_tx_quick_consumer_index0 != bp->hw_tx_cons) {
+ netif_tx_lock(dev);
+ __bnx2_tx_int(bp);
+ netif_tx_unlock(dev);
}
+ if (bp->status_blk->status_rx_quick_consumer_index0 != bp->hw_rx_cons)
+ work_done = bnx2_rx_int(bp, budget);
+
bp->last_status_idx = bp->status_blk->status_idx;
rmb();
if (!bnx2_has_work(bp)) {
- netif_rx_complete(dev);
+ netif_rx_complete(dev, napi);
if (likely(bp->flags & USING_MSI_FLAG)) {
REG_WR(bp, BNX2_PCICFG_INT_ACK_CMD,
BNX2_PCICFG_INT_ACK_CMD_INDEX_VALID |
@@ -2686,10 +2680,9 @@ bnx2_poll(struct net_device *dev, int *budget)
REG_WR(bp, BNX2_PCICFG_INT_ACK_CMD,
BNX2_PCICFG_INT_ACK_CMD_INDEX_VALID |
bp->last_status_idx);
- return 0;
}
- return 1;
+ return work_done;
}
/* Called with rtnl_lock from vlan functions and also netif_tx_lock
@@ -5146,10 +5139,7 @@ bnx2_vlan_rx_register(struct net_device *dev, struct vlan_group *vlgrp)
}
#endif
-/* Called with netif_tx_lock.
- * bnx2_tx_int() runs without netif_tx_lock unless it needs to call
- * netif_wake_queue().
- */
+/* Called with netif_tx_lock. */
static int
bnx2_start_xmit(struct sk_buff *skb, struct net_device *dev)
{
@@ -5292,6 +5282,8 @@ bnx2_close(struct net_device *dev)
struct bnx2 *bp = netdev_priv(dev);
u32 reset_code;
+ napi_disable(&bp->napi);
+
/* Calling flush_scheduled_work() may deadlock because
* linkwatch_event() may be on the workqueue and it will try to get
* the rtnl_lock which we are holding.
@@ -6370,9 +6362,7 @@ poll_bnx2(struct net_device *dev)
{
struct bnx2 *bp = netdev_priv(dev);
- disable_irq(bp->pdev->irq);
- bnx2_interrupt(bp->pdev->irq, dev);
- enable_irq(bp->pdev->irq);
+ __bnx2_tx_int(bp);
}
#endif
@@ -6856,11 +6846,10 @@ bnx2_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
#ifdef BCM_VLAN
dev->vlan_rx_register = bnx2_vlan_rx_register;
#endif
- dev->poll = bnx2_poll;
dev->ethtool_ops = &bnx2_ethtool_ops;
- dev->weight = 64;
bp = netdev_priv(dev);
+ netif_napi_init(&bp->napi, bnx2_poll, 64);
#if defined(HAVE_POLL_CONTROLLER) || defined(CONFIG_NET_POLL_CONTROLLER)
dev->poll_controller = poll_bnx2;
diff --git a/drivers/net/bnx2.h b/drivers/net/bnx2.h
index 102adfe..fbae439 100644
--- a/drivers/net/bnx2.h
+++ b/drivers/net/bnx2.h
@@ -6473,6 +6473,8 @@ struct bnx2 {
struct net_device *dev;
struct pci_dev *pdev;
+ struct napi_struct napi;
+
atomic_t intr_sem;
struct status_block *status_blk;
diff --git a/drivers/net/tg3.c b/drivers/net/tg3.c
index 887b9a5..efaa79e 100644
--- a/drivers/net/tg3.c
+++ b/drivers/net/tg3.c
@@ -574,7 +574,7 @@ static void tg3_restart_ints(struct tg3 *tp)
static inline void tg3_netif_stop(struct tg3 *tp)
{
tp->dev->trans_start = jiffies; /* prevent tx timeout */
- netif_poll_disable(tp->dev);
+ napi_disable(&tp->napi);
netif_tx_disable(tp->dev);
}
@@ -585,7 +585,7 @@ static inline void tg3_netif_start(struct tg3 *tp)
* so long as all callers are assured to have free tx slots
* (such as after tg3_init_hw)
*/
- netif_poll_enable(tp->dev);
+ napi_enable(&tp->napi);
tp->hw_status->status |= SD_STATUS_UPDATED;
tg3_enable_ints(tp);
}
@@ -3107,7 +3107,7 @@ static inline u32 tg3_tx_avail(struct tg3 *tp)
* need special logic to handle SKBs that have not had all
* of their frags sent yet, like SunGEM does.
*/
-static void tg3_tx(struct tg3 *tp)
+static void __tg3_tx(struct tg3 *tp)
{
u32 hw_idx = tp->hw_status->idx[0].tx_consumer;
u32 sw_idx = tp->tx_cons;
@@ -3163,11 +3163,9 @@ static void tg3_tx(struct tg3 *tp)
if (unlikely(netif_queue_stopped(tp->dev) &&
(tg3_tx_avail(tp) > TG3_TX_WAKEUP_THRESH(tp)))) {
- netif_tx_lock(tp->dev);
if (netif_queue_stopped(tp->dev) &&
(tg3_tx_avail(tp) > TG3_TX_WAKEUP_THRESH(tp)))
netif_wake_queue(tp->dev);
- netif_tx_unlock(tp->dev);
}
}
@@ -3471,11 +3469,12 @@ next_pkt_nopost:
return received;
}
-static int tg3_poll(struct net_device *netdev, int *budget)
+static int tg3_poll(struct napi_struct *napi, int budget)
{
- struct tg3 *tp = netdev_priv(netdev);
+ struct tg3 *tp = container_of(napi, struct tg3, napi);
+ struct net_device *netdev = tp->dev;
struct tg3_hw_status *sblk = tp->hw_status;
- int done;
+ int work_done = 0;
/* handle link change and other phy events */
if (!(tp->tg3_flags &
@@ -3492,9 +3491,11 @@ static int tg3_poll(struct net_device *netdev, int *budget)
/* run TX completion thread */
if (sblk->idx[0].tx_consumer != tp->tx_cons) {
- tg3_tx(tp);
+ netif_tx_lock(netdev);
+ __tg3_tx(tp);
+ netif_tx_unlock(netdev);
if (unlikely(tp->tg3_flags & TG3_FLAG_TX_RECOVERY_PENDING)) {
- netif_rx_complete(netdev);
+ netif_rx_complete(netdev, napi);
schedule_work(&tp->reset_task);
return 0;
}
@@ -3502,20 +3503,10 @@ static int tg3_poll(struct net_device *netdev, int *budget)
/* run RX thread, within the bounds set by NAPI.
* All RX "locking" is done by ensuring outside
- * code synchronizes with dev->poll()
+ * code synchronizes with tg3->napi.poll()
*/
- if (sblk->idx[0].rx_producer != tp->rx_rcb_ptr) {
- int orig_budget = *budget;
- int work_done;
-
- if (orig_budget > netdev->quota)
- orig_budget = netdev->quota;
-
- work_done = tg3_rx(tp, orig_budget);
-
- *budget -= work_done;
- netdev->quota -= work_done;
- }
+ if (sblk->idx[0].rx_producer != tp->rx_rcb_ptr)
+ work_done = tg3_rx(tp, budget);
if (tp->tg3_flags & TG3_FLAG_TAGGED_STATUS) {
tp->last_tag = sblk->status_tag;
@@ -3524,13 +3515,12 @@ static int tg3_poll(struct net_device *netdev, int *budget)
sblk->status &= ~SD_STATUS_UPDATED;
/* if no more work, tell net stack and NIC we're done */
- done = !tg3_has_work(tp);
- if (done) {
- netif_rx_complete(netdev);
+ if (!tg3_has_work(tp)) {
+ netif_rx_complete(netdev, napi);
tg3_restart_ints(tp);
}
- return (done ? 0 : 1);
+ return work_done;
}
static void tg3_irq_quiesce(struct tg3 *tp)
@@ -3577,7 +3567,7 @@ static irqreturn_t tg3_msi_1shot(int irq, void *dev_id)
prefetch(&tp->rx_rcb[tp->rx_rcb_ptr]);
if (likely(!tg3_irq_sync(tp)))
- netif_rx_schedule(dev); /* schedule NAPI poll */
+ netif_rx_schedule(dev, &tp->napi);
return IRQ_HANDLED;
}
@@ -3602,7 +3592,7 @@ static irqreturn_t tg3_msi(int irq, void *dev_id)
*/
tw32_mailbox(MAILBOX_INTERRUPT_0 + TG3_64BIT_REG_LOW, 0x00000001);
if (likely(!tg3_irq_sync(tp)))
- netif_rx_schedule(dev); /* schedule NAPI poll */
+ netif_rx_schedule(dev, &tp->napi);
return IRQ_RETVAL(1);
}
@@ -3644,7 +3634,7 @@ static irqreturn_t tg3_interrupt(int irq, void *dev_id)
sblk->status &= ~SD_STATUS_UPDATED;
if (likely(tg3_has_work(tp))) {
prefetch(&tp->rx_rcb[tp->rx_rcb_ptr]);
- netif_rx_schedule(dev); /* schedule NAPI poll */
+ netif_rx_schedule(dev, &tp->napi);
} else {
/* No work, shared interrupt perhaps? re-enable
* interrupts, and flush that PCI write
@@ -3690,7 +3680,7 @@ static irqreturn_t tg3_interrupt_tagged(int irq, void *dev_id)
tw32_mailbox_f(MAILBOX_INTERRUPT_0 + TG3_64BIT_REG_LOW, 0x00000001);
if (tg3_irq_sync(tp))
goto out;
- if (netif_rx_schedule_prep(dev)) {
+ if (netif_rx_schedule_prep(dev, &tp->napi)) {
prefetch(&tp->rx_rcb[tp->rx_rcb_ptr]);
/* Update last_tag to mark that this status has been
* seen. Because interrupt may be shared, we may be
@@ -3698,7 +3688,7 @@ static irqreturn_t tg3_interrupt_tagged(int irq, void *dev_id)
* if tg3_poll() is not scheduled.
*/
tp->last_tag = sblk->status_tag;
- __netif_rx_schedule(dev);
+ __netif_rx_schedule(dev, &tp->napi);
}
out:
return IRQ_RETVAL(handled);
@@ -3737,7 +3727,7 @@ static int tg3_restart_hw(struct tg3 *tp, int reset_phy)
tg3_full_unlock(tp);
del_timer_sync(&tp->timer);
tp->irq_sync = 0;
- netif_poll_enable(tp->dev);
+ napi_enable(&tp->napi);
dev_close(tp->dev);
tg3_full_lock(tp, 0);
}
@@ -3749,7 +3739,7 @@ static void tg3_poll_controller(struct net_device *dev)
{
struct tg3 *tp = netdev_priv(dev);
- tg3_interrupt(tp->pdev->irq, dev);
+ __tg3_tx(tp);
}
#endif
@@ -3932,7 +3922,7 @@ static int tg3_start_xmit(struct sk_buff *skb, struct net_device *dev)
len = skb_headlen(skb);
/* We are running in BH disabled context with netif_tx_lock
- * and TX reclaim runs via tp->poll inside of a software
+ * and TX reclaim runs via tp->napi.poll inside of a software
* interrupt. Furthermore, IRQ processing runs lockless so we have
* no IRQ context deadlocks to worry about either. Rejoice!
*/
@@ -4087,7 +4077,7 @@ static int tg3_start_xmit_dma_bug(struct sk_buff *skb, struct net_device *dev)
len = skb_headlen(skb);
/* We are running in BH disabled context with netif_tx_lock
- * and TX reclaim runs via tp->poll inside of a software
+ * and TX reclaim runs via tp->napi.poll inside of a software
* interrupt. Furthermore, IRQ processing runs lockless so we have
* no IRQ context deadlocks to worry about either. Rejoice!
*/
@@ -7456,6 +7446,7 @@ static int tg3_close(struct net_device *dev)
{
struct tg3 *tp = netdev_priv(dev);
+ napi_disable(&tp->napi);
cancel_work_sync(&tp->reset_task);
netif_stop_queue(dev);
@@ -11897,9 +11888,8 @@ static int __devinit tg3_init_one(struct pci_dev *pdev,
dev->set_mac_address = tg3_set_mac_addr;
dev->do_ioctl = tg3_ioctl;
dev->tx_timeout = tg3_tx_timeout;
- dev->poll = tg3_poll;
+ netif_napi_init(&tp->napi, tg3_poll, 64);
dev->ethtool_ops = &tg3_ethtool_ops;
- dev->weight = 64;
dev->watchdog_timeo = TG3_TX_TIMEOUT;
dev->change_mtu = tg3_change_mtu;
dev->irq = pdev->irq;
diff --git a/drivers/net/tg3.h b/drivers/net/tg3.h
index 5c21f49..a6a23bb 100644
--- a/drivers/net/tg3.h
+++ b/drivers/net/tg3.h
@@ -2176,6 +2176,7 @@ struct tg3 {
dma_addr_t tx_desc_mapping;
/* begin "rx thread" cacheline section */
+ struct napi_struct napi;
void (*write32_rx_mbox) (struct tg3 *, u32,
u32);
u32 rx_rcb_ptr;
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 4a616d7..90c628e 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -31,6 +31,7 @@
#ifdef __KERNEL__
#include <linux/timer.h>
+#include <linux/delay.h>
#include <asm/atomic.h>
#include <asm/cache.h>
#include <asm/byteorder.h>
@@ -38,6 +39,7 @@
#include <linux/device.h>
#include <linux/percpu.h>
#include <linux/dmaengine.h>
+#include <linux/workqueue.h>
struct vlan_group;
struct ethtool_ops;
@@ -258,7 +260,6 @@ enum netdev_state_t
__LINK_STATE_PRESENT,
__LINK_STATE_SCHED,
__LINK_STATE_NOCARRIER,
- __LINK_STATE_RX_SCHED,
__LINK_STATE_LINKWATCH_PENDING,
__LINK_STATE_DORMANT,
__LINK_STATE_QDISC_RUNNING,
@@ -278,6 +279,63 @@ struct netdev_boot_setup {
extern int __init netdev_boot_setup(char *str);
/*
+ * Structure for NAPI scheduling similar to tasklet but with weighting
+ */
+struct napi_struct {
+ struct list_head poll_list;
+ unsigned long state;
+ int weight;
+ int quota;
+ int (*poll)(struct napi_struct *, int);
+};
+
+enum
+{
+ NAPI_STATE_SCHED, /* Poll is scheduled */
+};
+
+static inline void netif_napi_init(struct napi_struct *napi,
+ int (*poll)(struct napi_struct *, int),
+ int weight)
+{
+ napi->poll = poll;
+ napi->weight = weight;
+}
+
+extern void FASTCALL(__napi_schedule(struct napi_struct *n));
+
+static inline int napi_schedule_prep(struct napi_struct *n)
+{
+ return !test_and_set_bit(NAPI_STATE_SCHED, &n->state);
+}
+
+static inline void napi_schedule(struct napi_struct *n)
+{
+ if (napi_schedule_prep(n))
+ __napi_schedule(n);
+}
+
+static inline void napi_complete(struct napi_struct *n)
+{
+ BUG_ON(!test_bit(NAPI_STATE_SCHED, &n->state));
+ smp_mb__before_clear_bit();
+ clear_bit(NAPI_STATE_SCHED, &n->state);
+}
+
+static inline void napi_disable(struct napi_struct *n)
+{
+ while (test_and_set_bit(NAPI_STATE_SCHED, &n->state))
+ msleep_interruptible(1);
+}
+
+static inline void napi_enable(struct napi_struct *n)
+{
+ BUG_ON(!test_bit(NAPI_STATE_SCHED, &n->state));
+ smp_mb__before_clear_bit();
+ clear_bit(NAPI_STATE_SCHED, &n->state);
+}
+
+/*
* The DEVICE structure.
* Actually, this whole structure is a big mistake. It mixes I/O
* data with strictly "high-level" data, and it has to know about
@@ -430,12 +488,6 @@ struct net_device
/*
* Cache line mostly used on receive path (including eth_type_trans())
*/
- struct list_head poll_list ____cacheline_aligned_in_smp;
- /* Link to poll list */
-
- int (*poll) (struct net_device *dev, int *quota);
- int quota;
- int weight;
unsigned long last_rx; /* Time of last Rx */
/* Interface address info used in eth_type_trans() */
unsigned char dev_addr[MAX_ADDR_LEN]; /* hw address, (before bcast
@@ -663,9 +715,6 @@ extern int call_netdevice_notifiers(unsigned long val, void *v);
extern struct net_device *dev_get_by_index(int ifindex);
extern struct net_device *__dev_get_by_index(int ifindex);
extern int dev_restart(struct net_device *dev);
-#ifdef CONFIG_NETPOLL_TRAP
-extern int netpoll_trap(void);
-#endif
typedef int gifconf_func_t(struct net_device * dev, char __user * bufptr, int len);
extern int register_gifconf(unsigned int family, gifconf_func_t * gifconf);
@@ -678,7 +727,6 @@ static inline int unregister_gifconf(unsigned int family)
* Incoming packets are placed on per-cpu queues so that
* no locking is needed.
*/
-
struct softnet_data
{
struct net_device *output_queue;
@@ -686,7 +734,7 @@ struct softnet_data
struct list_head poll_list;
struct sk_buff *completion_queue;
- struct net_device backlog_dev; /* Sorry. 8) */
+ struct napi_struct backlog;
#ifdef CONFIG_NET_DMA
struct dma_chan *net_dma;
#endif
@@ -711,12 +759,6 @@ static inline void netif_start_queue(struct net_device *dev)
static inline void netif_wake_queue(struct net_device *dev)
{
-#ifdef CONFIG_NETPOLL_TRAP
- if (netpoll_trap()) {
- clear_bit(__LINK_STATE_XOFF, &dev->state);
- return;
- }
-#endif
if (test_and_clear_bit(__LINK_STATE_XOFF, &dev->state))
__netif_schedule(dev);
}
@@ -752,10 +794,6 @@ static inline void netif_start_subqueue(struct net_device *dev, u16 queue_index)
static inline void netif_stop_subqueue(struct net_device *dev, u16 queue_index)
{
#ifdef CONFIG_NETDEVICES_MULTIQUEUE
-#ifdef CONFIG_NETPOLL_TRAP
- if (netpoll_trap())
- return;
-#endif
set_bit(__LINK_STATE_XOFF, &dev->egress_subqueue[queue_index].state);
#endif
}
@@ -774,10 +812,6 @@ static inline int netif_subqueue_stopped(const struct net_device *dev,
static inline void netif_wake_subqueue(struct net_device *dev, u16 queue_index)
{
#ifdef CONFIG_NETDEVICES_MULTIQUEUE
-#ifdef CONFIG_NETPOLL_TRAP
- if (netpoll_trap())
- return;
-#endif
if (test_and_clear_bit(__LINK_STATE_XOFF,
&dev->egress_subqueue[queue_index].state))
__netif_schedule(dev);
@@ -796,20 +830,7 @@ static inline int netif_is_multiqueue(const struct net_device *dev)
/* Use this variant when it is known for sure that it
* is executing from interrupt context.
*/
-static inline void dev_kfree_skb_irq(struct sk_buff *skb)
-{
- if (atomic_dec_and_test(&skb->users)) {
- struct softnet_data *sd;
- unsigned long flags;
-
- local_irq_save(flags);
- sd = &__get_cpu_var(softnet_data);
- skb->next = sd->completion_queue;
- sd->completion_queue = skb;
- raise_softirq_irqoff(NET_TX_SOFTIRQ);
- local_irq_restore(flags);
- }
-}
+extern void dev_kfree_skb_irq(struct sk_buff *skb);
/* Use this variant in places where it could be invoked
* either from interrupt or non-interrupt context.
@@ -955,60 +976,41 @@ static inline u32 netif_msg_init(int debug_value, int default_msg_enable_bits)
return (1 << debug_value) - 1;
}
-/* Test if receive needs to be scheduled */
-static inline int __netif_rx_schedule_prep(struct net_device *dev)
-{
- return !test_and_set_bit(__LINK_STATE_RX_SCHED, &dev->state);
-}
/* Test if receive needs to be scheduled but only if up */
-static inline int netif_rx_schedule_prep(struct net_device *dev)
+static inline int netif_rx_schedule_prep(struct net_device *dev,
+ struct napi_struct *napi)
{
- return netif_running(dev) && __netif_rx_schedule_prep(dev);
+ return netif_running(dev) && napi_schedule_prep(napi);
}
/* Add interface to tail of rx poll list. This assumes that _prep has
* already been called and returned 1.
*/
-
-extern void __netif_rx_schedule(struct net_device *dev);
-
-/* Try to reschedule poll. Called by irq handler. */
-
-static inline void netif_rx_schedule(struct net_device *dev)
+static inline void __netif_rx_schedule(struct net_device *dev,
+ struct napi_struct *napi)
{
- if (netif_rx_schedule_prep(dev))
- __netif_rx_schedule(dev);
+ dev_hold(dev);
+ __napi_schedule(napi);
}
-/* Try to reschedule poll. Called by dev->poll() after netif_rx_complete().
- * Do not inline this?
- */
-static inline int netif_rx_reschedule(struct net_device *dev, int undo)
-{
- if (netif_rx_schedule_prep(dev)) {
- unsigned long flags;
-
- dev->quota += undo;
+/* Try to reschedule poll. Called by irq handler. */
- local_irq_save(flags);
- list_add_tail(&dev->poll_list, &__get_cpu_var(softnet_data).poll_list);
- __raise_softirq_irqoff(NET_RX_SOFTIRQ);
- local_irq_restore(flags);
- return 1;
- }
- return 0;
+static inline void netif_rx_schedule(struct net_device *dev,
+ struct napi_struct *napi)
+{
+ if (netif_rx_schedule_prep(dev, napi))
+ __netif_rx_schedule(dev, napi);
}
/* same as netif_rx_complete, except that local_irq_save(flags)
* has already been issued
*/
-static inline void __netif_rx_complete(struct net_device *dev)
+static inline void __netif_rx_complete(struct net_device *dev,
+ struct napi_struct *napi)
{
- BUG_ON(!test_bit(__LINK_STATE_RX_SCHED, &dev->state));
- list_del(&dev->poll_list);
- smp_mb__before_clear_bit();
- clear_bit(__LINK_STATE_RX_SCHED, &dev->state);
+ napi_complete(napi);
+ dev_put(dev);
}
/* Remove interface from poll list: it must be in the poll list
@@ -1016,28 +1018,16 @@ static inline void __netif_rx_complete(struct net_device *dev)
* it completes the work. The device cannot be out of poll list at this
* moment, it is BUG().
*/
-static inline void netif_rx_complete(struct net_device *dev)
+static inline void netif_rx_complete(struct net_device *dev,
+ struct napi_struct *napi)
{
unsigned long flags;
local_irq_save(flags);
- __netif_rx_complete(dev);
+ __netif_rx_complete(dev, napi);
local_irq_restore(flags);
}
-static inline void netif_poll_disable(struct net_device *dev)
-{
- while (test_and_set_bit(__LINK_STATE_RX_SCHED, &dev->state))
- /* No hurry. */
- schedule_timeout_interruptible(1);
-}
-
-static inline void netif_poll_enable(struct net_device *dev)
-{
- smp_mb__before_clear_bit();
- clear_bit(__LINK_STATE_RX_SCHED, &dev->state);
-}
-
static inline void netif_tx_lock(struct net_device *dev)
{
spin_lock(&dev->_xmit_lock);
diff --git a/include/linux/netpoll.h b/include/linux/netpoll.h
index 29930b7..ddae7c1 100644
--- a/include/linux/netpoll.h
+++ b/include/linux/netpoll.h
@@ -16,7 +16,6 @@ struct netpoll {
struct net_device *dev;
char dev_name[IFNAMSIZ];
const char *name;
- void (*rx_hook)(struct netpoll *, int, char *, int);
u32 local_ip, remote_ip;
u16 local_port, remote_port;
@@ -25,71 +24,13 @@ struct netpoll {
struct netpoll_info {
atomic_t refcnt;
- spinlock_t poll_lock;
- int poll_owner;
- int rx_flags;
- spinlock_t rx_lock;
- struct netpoll *rx_np; /* netpoll that registered an rx_hook */
- struct sk_buff_head arp_tx; /* list of arp requests to reply to */
struct sk_buff_head txq;
struct delayed_work tx_work;
};
-void netpoll_poll(struct netpoll *np);
void netpoll_send_udp(struct netpoll *np, const char *msg, int len);
int netpoll_parse_options(struct netpoll *np, char *opt);
int netpoll_setup(struct netpoll *np);
-int netpoll_trap(void);
-void netpoll_set_trap(int trap);
void netpoll_cleanup(struct netpoll *np);
-int __netpoll_rx(struct sk_buff *skb);
-
-
-#ifdef CONFIG_NETPOLL
-static inline int netpoll_rx(struct sk_buff *skb)
-{
- struct netpoll_info *npinfo = skb->dev->npinfo;
- unsigned long flags;
- int ret = 0;
-
- if (!npinfo || (!npinfo->rx_np && !npinfo->rx_flags))
- return 0;
-
- spin_lock_irqsave(&npinfo->rx_lock, flags);
- /* check rx_flags again with the lock held */
- if (npinfo->rx_flags && __netpoll_rx(skb))
- ret = 1;
- spin_unlock_irqrestore(&npinfo->rx_lock, flags);
-
- return ret;
-}
-
-static inline void *netpoll_poll_lock(struct net_device *dev)
-{
- rcu_read_lock(); /* deal with race on ->npinfo */
- if (dev->npinfo) {
- spin_lock(&dev->npinfo->poll_lock);
- dev->npinfo->poll_owner = smp_processor_id();
- return dev->npinfo;
- }
- return NULL;
-}
-
-static inline void netpoll_poll_unlock(void *have)
-{
- struct netpoll_info *npi = have;
-
- if (npi) {
- npi->poll_owner = -1;
- spin_unlock(&npi->poll_lock);
- }
- rcu_read_unlock();
-}
-
-#else
-#define netpoll_rx(a) 0
-#define netpoll_poll_lock(a) NULL
-#define netpoll_poll_unlock(a)
-#endif
#endif
diff --git a/net/core/dev.c b/net/core/dev.c
index ee40355..591cccf 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -220,7 +220,8 @@ static RAW_NOTIFIER_HEAD(netdev_chain);
* Device drivers call our routines to queue packets here. We empty the
* queue in the local softnet handler.
*/
-DEFINE_PER_CPU(struct softnet_data, softnet_data) = { NULL };
+
+DEFINE_PER_CPU(struct softnet_data, softnet_data);
#ifdef CONFIG_SYSFS
extern int netdev_sysfs_init(void);
@@ -996,16 +997,12 @@ int dev_close(struct net_device *dev)
clear_bit(__LINK_STATE_START, &dev->state);
/* Synchronize to scheduled poll. We cannot touch poll list,
- * it can be even on different cpu. So just clear netif_running(),
- * and wait when poll really will happen. Actually, the best place
- * for this is inside dev->stop() after device stopped its irq
- * engine, but this requires more changes in devices. */
-
+ * it can be even on different cpu. So just clear netif_running().
+ *
+ * dev->stop() will invoke napi_disable() on all of it's
+ * napi_struct instances on this device.
+ */
smp_mb__after_clear_bit(); /* Commit netif_running(). */
- while (test_bit(__LINK_STATE_RX_SCHED, &dev->state)) {
- /* No hurry. */
- msleep(1);
- }
/*
* Call the device specific close. This cannot fail.
@@ -1188,21 +1185,21 @@ void __netif_schedule(struct net_device *dev)
}
EXPORT_SYMBOL(__netif_schedule);
-void __netif_rx_schedule(struct net_device *dev)
+void dev_kfree_skb_irq(struct sk_buff *skb)
{
- unsigned long flags;
+ if (atomic_dec_and_test(&skb->users)) {
+ struct softnet_data *sd;
+ unsigned long flags;
- local_irq_save(flags);
- dev_hold(dev);
- list_add_tail(&dev->poll_list, &__get_cpu_var(softnet_data).poll_list);
- if (dev->quota < 0)
- dev->quota += dev->weight;
- else
- dev->quota = dev->weight;
- __raise_softirq_irqoff(NET_RX_SOFTIRQ);
- local_irq_restore(flags);
+ local_irq_save(flags);
+ sd = &__get_cpu_var(softnet_data);
+ skb->next = sd->completion_queue;
+ sd->completion_queue = skb;
+ raise_softirq_irqoff(NET_TX_SOFTIRQ);
+ local_irq_restore(flags);
+ }
}
-EXPORT_SYMBOL(__netif_rx_schedule);
+EXPORT_SYMBOL(dev_kfree_skb_irq);
void dev_kfree_skb_any(struct sk_buff *skb)
{
@@ -1661,10 +1658,6 @@ int netif_rx(struct sk_buff *skb)
struct softnet_data *queue;
unsigned long flags;
- /* if netpoll wants it, pretend we never saw it */
- if (netpoll_rx(skb))
- return NET_RX_DROP;
-
if (!skb->tstamp.tv64)
net_timestamp(skb);
@@ -1685,7 +1678,7 @@ enqueue:
return NET_RX_SUCCESS;
}
- netif_rx_schedule(&queue->backlog_dev);
+ napi_schedule(&queue->backlog);
goto enqueue;
}
@@ -1726,6 +1719,7 @@ static inline struct net_device *skb_bond(struct sk_buff *skb)
return dev;
}
+
static void net_tx_action(struct softirq_action *h)
{
struct softnet_data *sd = &__get_cpu_var(softnet_data);
@@ -1881,10 +1875,6 @@ int netif_receive_skb(struct sk_buff *skb)
int ret = NET_RX_DROP;
__be16 type;
- /* if we've gotten here through NAPI, check netpoll */
- if (skb->dev->poll && netpoll_rx(skb))
- return NET_RX_DROP;
-
if (!skb->tstamp.tv64)
net_timestamp(skb);
@@ -1972,90 +1962,96 @@ out:
return ret;
}
-static int process_backlog(struct net_device *backlog_dev, int *budget)
+static int process_backlog(struct napi_struct *napi, int quota)
{
int work = 0;
- int quota = min(backlog_dev->quota, *budget);
struct softnet_data *queue = &__get_cpu_var(softnet_data);
unsigned long start_time = jiffies;
- backlog_dev->weight = weight_p;
- for (;;) {
+ napi->weight = weight_p;
+ do {
struct sk_buff *skb;
struct net_device *dev;
local_irq_disable();
skb = __skb_dequeue(&queue->input_pkt_queue);
- if (!skb)
- goto job_done;
local_irq_enable();
-
+ if (!skb) {
+ napi_complete(napi);
+ break;
+ }
+
dev = skb->dev;
netif_receive_skb(skb);
dev_put(dev);
+ } while (++work < quota && jiffies == start_time);
- work++;
-
- if (work >= quota || jiffies - start_time > 1)
- break;
-
- }
-
- backlog_dev->quota -= work;
- *budget -= work;
- return -1;
+ return work;
+}
-job_done:
- backlog_dev->quota -= work;
- *budget -= work;
+/**
+ * __napi_schedule - schedule for receive
+ * @napi: entry to schedule
+ *
+ * The entry's receive function will be scheduled to run
+ */
+void fastcall __napi_schedule(struct napi_struct *n)
+{
+ unsigned long flags;
- list_del(&backlog_dev->poll_list);
- smp_mb__before_clear_bit();
- netif_poll_enable(backlog_dev);
+ if (n->quota < 0)
+ n->quota += n->weight;
+ else
+ n->quota = n->weight;
- local_irq_enable();
- return 0;
+ local_irq_save(flags);
+ list_add_tail(&n->poll_list, &__get_cpu_var(softnet_data).poll_list);
+ __raise_softirq_irqoff(NET_RX_SOFTIRQ);
+ local_irq_restore(flags);
}
+EXPORT_SYMBOL(__napi_schedule);
+
static void net_rx_action(struct softirq_action *h)
{
- struct softnet_data *queue = &__get_cpu_var(softnet_data);
+ struct list_head list;
unsigned long start_time = jiffies;
int budget = netdev_budget;
- void *have;
local_irq_disable();
+ list_replace_init(&__get_cpu_var(softnet_data).poll_list, &list);
+ local_irq_enable();
- while (!list_empty(&queue->poll_list)) {
- struct net_device *dev;
+ while (!list_empty(&list)) {
+ struct napi_struct *n;
- if (budget <= 0 || jiffies - start_time > 1)
- goto softnet_break;
+ /* if softirq window is exhuasted then punt */
+ if (unlikely(budget <= 0 || jiffies != start_time)) {
+ local_irq_disable();
+ list_splice(&list, &__get_cpu_var(softnet_data).poll_list);
+ __raise_softirq_irqoff(NET_RX_SOFTIRQ);
+ local_irq_enable();
+ break;
+ }
- local_irq_enable();
+ n = list_entry(list.next, struct napi_struct, poll_list);
- dev = list_entry(queue->poll_list.next,
- struct net_device, poll_list);
- have = netpoll_poll_lock(dev);
+ list_del(&n->poll_list);
- if (dev->quota <= 0 || dev->poll(dev, &budget)) {
- netpoll_poll_unlock(have);
- local_irq_disable();
- list_move_tail(&dev->poll_list, &queue->poll_list);
- if (dev->quota < 0)
- dev->quota += dev->weight;
- else
- dev->quota = dev->weight;
- } else {
- netpoll_poll_unlock(have);
- dev_put(dev);
- local_irq_disable();
+ /* if quota not exhausted process work */
+ if (likely(n->quota > 0)) {
+ int work = n->poll(n, min(budget, n->quota));
+
+ budget -= work;
+ n->quota -= work;
}
+
+ /* if napi_complete not called, reschedule */
+ if (test_bit(NAPI_STATE_SCHED, &n->state))
+ __napi_schedule(n);
}
-out:
- local_irq_enable();
#ifdef CONFIG_NET_DMA
/*
* There may not be any more sk_buffs coming right now, so push
@@ -2070,12 +2066,6 @@ out:
}
}
#endif
- return;
-
-softnet_break:
- __get_cpu_var(netdev_rx_stat).time_squeeze++;
- __raise_softirq_irqoff(NET_RX_SOFTIRQ);
- goto out;
}
static gifconf_func_t * gifconf_list [NPROTO];
@@ -3980,10 +3970,8 @@ static int __init net_dev_init(void)
skb_queue_head_init(&queue->input_pkt_queue);
queue->completion_queue = NULL;
INIT_LIST_HEAD(&queue->poll_list);
- set_bit(__LINK_STATE_START, &queue->backlog_dev.state);
- queue->backlog_dev.weight = weight_p;
- queue->backlog_dev.poll = process_backlog;
- atomic_set(&queue->backlog_dev.refcnt, 1);
+
+ netif_napi_init(&queue->backlog, process_backlog, weight_p);
}
netdev_dma_register();
diff --git a/net/core/netpoll.c b/net/core/netpoll.c
index de1b26a..6737aba 100644
--- a/net/core/netpoll.c
+++ b/net/core/netpoll.c
@@ -32,23 +32,15 @@
#define MAX_UDP_CHUNK 1460
#define MAX_SKBS 32
-#define MAX_QUEUE_DEPTH (MAX_SKBS / 2)
static struct sk_buff_head skb_pool;
-static atomic_t trapped;
-
#define USEC_PER_POLL 50
-#define NETPOLL_RX_ENABLED 1
-#define NETPOLL_RX_DROP 2
#define MAX_SKB_SIZE \
(MAX_UDP_CHUNK + sizeof(struct udphdr) + \
sizeof(struct iphdr) + sizeof(struct ethhdr))
-static void zap_completion_queue(void);
-static void arp_reply(struct sk_buff *skb);
-
static void queue_process(struct work_struct *work)
{
struct netpoll_info *npinfo =
@@ -81,90 +73,6 @@ static void queue_process(struct work_struct *work)
}
}
-static __sum16 checksum_udp(struct sk_buff *skb, struct udphdr *uh,
- unsigned short ulen, __be32 saddr, __be32 daddr)
-{
- __wsum psum;
-
- if (uh->check == 0 || skb_csum_unnecessary(skb))
- return 0;
-
- psum = csum_tcpudp_nofold(saddr, daddr, ulen, IPPROTO_UDP, 0);
-
- if (skb->ip_summed == CHECKSUM_COMPLETE &&
- !csum_fold(csum_add(psum, skb->csum)))
- return 0;
-
- skb->csum = psum;
-
- return __skb_checksum_complete(skb);
-}
-
-/*
- * Check whether delayed processing was scheduled for our NIC. If so,
- * we attempt to grab the poll lock and use ->poll() to pump the card.
- * If this fails, either we've recursed in ->poll() or it's already
- * running on another CPU.
- *
- * Note: we don't mask interrupts with this lock because we're using
- * trylock here and interrupts are already disabled in the softirq
- * case. Further, we test the poll_owner to avoid recursion on UP
- * systems where the lock doesn't exist.
- *
- * In cases where there is bi-directional communications, reading only
- * one message at a time can lead to packets being dropped by the
- * network adapter, forcing superfluous retries and possibly timeouts.
- * Thus, we set our budget to greater than 1.
- */
-static void poll_napi(struct netpoll *np)
-{
- struct netpoll_info *npinfo = np->dev->npinfo;
- int budget = 16;
-
- if (test_bit(__LINK_STATE_RX_SCHED, &np->dev->state) &&
- npinfo->poll_owner != smp_processor_id() &&
- spin_trylock(&npinfo->poll_lock)) {
- npinfo->rx_flags |= NETPOLL_RX_DROP;
- atomic_inc(&trapped);
-
- np->dev->poll(np->dev, &budget);
-
- atomic_dec(&trapped);
- npinfo->rx_flags &= ~NETPOLL_RX_DROP;
- spin_unlock(&npinfo->poll_lock);
- }
-}
-
-static void service_arp_queue(struct netpoll_info *npi)
-{
- struct sk_buff *skb;
-
- if (unlikely(!npi))
- return;
-
- skb = skb_dequeue(&npi->arp_tx);
-
- while (skb != NULL) {
- arp_reply(skb);
- skb = skb_dequeue(&npi->arp_tx);
- }
-}
-
-void netpoll_poll(struct netpoll *np)
-{
- if (!np->dev || !netif_running(np->dev) || !np->dev->poll_controller)
- return;
-
- /* Process pending work on NIC */
- np->dev->poll_controller(np->dev);
- if (np->dev->poll)
- poll_napi(np);
-
- service_arp_queue(np->dev->npinfo);
-
- zap_completion_queue();
-}
-
static void refill_skbs(void)
{
struct sk_buff *skb;
@@ -181,52 +89,18 @@ static void refill_skbs(void)
spin_unlock_irqrestore(&skb_pool.lock, flags);
}
-static void zap_completion_queue(void)
-{
- unsigned long flags;
- struct softnet_data *sd = &get_cpu_var(softnet_data);
-
- if (sd->completion_queue) {
- struct sk_buff *clist;
-
- local_irq_save(flags);
- clist = sd->completion_queue;
- sd->completion_queue = NULL;
- local_irq_restore(flags);
-
- while (clist != NULL) {
- struct sk_buff *skb = clist;
- clist = clist->next;
- if (skb->destructor)
- dev_kfree_skb_any(skb); /* put this one back */
- else
- __kfree_skb(skb);
- }
- }
-
- put_cpu_var(softnet_data);
-}
-
static struct sk_buff *find_skb(struct netpoll *np, int len, int reserve)
{
- int count = 0;
struct sk_buff *skb;
- zap_completion_queue();
refill_skbs();
-repeat:
skb = alloc_skb(len, GFP_ATOMIC);
if (!skb)
skb = skb_dequeue(&skb_pool);
- if (!skb) {
- if (++count < 10) {
- netpoll_poll(np);
- goto repeat;
- }
+ if (!skb)
return NULL;
- }
atomic_set(&skb->users, 1);
skb_reserve(skb, reserve);
@@ -246,8 +120,7 @@ static void netpoll_send_skb(struct netpoll *np, struct sk_buff *skb)
}
/* don't get messages out of order, and no recursion */
- if (skb_queue_len(&npinfo->txq) == 0 &&
- npinfo->poll_owner != smp_processor_id()) {
+ if (skb_queue_len(&npinfo->txq) == 0) {
unsigned long flags;
local_irq_save(flags);
@@ -255,6 +128,7 @@ static void netpoll_send_skb(struct netpoll *np, struct sk_buff *skb)
for (tries = jiffies_to_usecs(1)/USEC_PER_POLL;
tries > 0; --tries) {
if (netif_tx_trylock(dev)) {
+ dev->poll_controller(dev);
if (!netif_queue_stopped(dev) &&
!netif_subqueue_stopped(dev, skb->queue_mapping))
status = dev->hard_start_xmit(skb, dev);
@@ -265,9 +139,6 @@ static void netpoll_send_skb(struct netpoll *np, struct sk_buff *skb)
}
- /* tickle device maybe there is some cleanup */
- netpoll_poll(np);
-
udelay(USEC_PER_POLL);
}
local_irq_restore(flags);
@@ -340,185 +211,6 @@ void netpoll_send_udp(struct netpoll *np, const char *msg, int len)
netpoll_send_skb(np, skb);
}
-static void arp_reply(struct sk_buff *skb)
-{
- struct netpoll_info *npinfo = skb->dev->npinfo;
- struct arphdr *arp;
- unsigned char *arp_ptr;
- int size, type = ARPOP_REPLY, ptype = ETH_P_ARP;
- __be32 sip, tip;
- unsigned char *sha;
- struct sk_buff *send_skb;
- struct netpoll *np = NULL;
-
- if (npinfo->rx_np && npinfo->rx_np->dev == skb->dev)
- np = npinfo->rx_np;
- if (!np)
- return;
-
- /* No arp on this interface */
- if (skb->dev->flags & IFF_NOARP)
- return;
-
- if (!pskb_may_pull(skb, (sizeof(struct arphdr) +
- (2 * skb->dev->addr_len) +
- (2 * sizeof(u32)))))
- return;
-
- skb_reset_network_header(skb);
- skb_reset_transport_header(skb);
- arp = arp_hdr(skb);
-
- if ((arp->ar_hrd != htons(ARPHRD_ETHER) &&
- arp->ar_hrd != htons(ARPHRD_IEEE802)) ||
- arp->ar_pro != htons(ETH_P_IP) ||
- arp->ar_op != htons(ARPOP_REQUEST))
- return;
-
- arp_ptr = (unsigned char *)(arp+1);
- /* save the location of the src hw addr */
- sha = arp_ptr;
- arp_ptr += skb->dev->addr_len;
- memcpy(&sip, arp_ptr, 4);
- arp_ptr += 4;
- /* if we actually cared about dst hw addr, it would get copied here */
- arp_ptr += skb->dev->addr_len;
- memcpy(&tip, arp_ptr, 4);
-
- /* Should we ignore arp? */
- if (tip != htonl(np->local_ip) || LOOPBACK(tip) || MULTICAST(tip))
- return;
-
- size = sizeof(struct arphdr) + 2 * (skb->dev->addr_len + 4);
- send_skb = find_skb(np, size + LL_RESERVED_SPACE(np->dev),
- LL_RESERVED_SPACE(np->dev));
-
- if (!send_skb)
- return;
-
- skb_reset_network_header(send_skb);
- arp = (struct arphdr *) skb_put(send_skb, size);
- send_skb->dev = skb->dev;
- send_skb->protocol = htons(ETH_P_ARP);
-
- /* Fill the device header for the ARP frame */
-
- if (np->dev->hard_header &&
- np->dev->hard_header(send_skb, skb->dev, ptype,
- sha, np->local_mac,
- send_skb->len) < 0) {
- kfree_skb(send_skb);
- return;
- }
-
- /*
- * Fill out the arp protocol part.
- *
- * we only support ethernet device type,
- * which (according to RFC 1390) should always equal 1 (Ethernet).
- */
-
- arp->ar_hrd = htons(np->dev->type);
- arp->ar_pro = htons(ETH_P_IP);
- arp->ar_hln = np->dev->addr_len;
- arp->ar_pln = 4;
- arp->ar_op = htons(type);
-
- arp_ptr=(unsigned char *)(arp + 1);
- memcpy(arp_ptr, np->dev->dev_addr, np->dev->addr_len);
- arp_ptr += np->dev->addr_len;
- memcpy(arp_ptr, &tip, 4);
- arp_ptr += 4;
- memcpy(arp_ptr, sha, np->dev->addr_len);
- arp_ptr += np->dev->addr_len;
- memcpy(arp_ptr, &sip, 4);
-
- netpoll_send_skb(np, send_skb);
-}
-
-int __netpoll_rx(struct sk_buff *skb)
-{
- int proto, len, ulen;
- struct iphdr *iph;
- struct udphdr *uh;
- struct netpoll_info *npi = skb->dev->npinfo;
- struct netpoll *np = npi->rx_np;
-
- if (!np)
- goto out;
- if (skb->dev->type != ARPHRD_ETHER)
- goto out;
-
- /* check if netpoll clients need ARP */
- if (skb->protocol == htons(ETH_P_ARP) &&
- atomic_read(&trapped)) {
- skb_queue_tail(&npi->arp_tx, skb);
- return 1;
- }
-
- proto = ntohs(eth_hdr(skb)->h_proto);
- if (proto != ETH_P_IP)
- goto out;
- if (skb->pkt_type == PACKET_OTHERHOST)
- goto out;
- if (skb_shared(skb))
- goto out;
-
- iph = (struct iphdr *)skb->data;
- if (!pskb_may_pull(skb, sizeof(struct iphdr)))
- goto out;
- if (iph->ihl < 5 || iph->version != 4)
- goto out;
- if (!pskb_may_pull(skb, iph->ihl*4))
- goto out;
- if (ip_fast_csum((u8 *)iph, iph->ihl) != 0)
- goto out;
-
- len = ntohs(iph->tot_len);
- if (skb->len < len || len < iph->ihl*4)
- goto out;
-
- /*
- * Our transport medium may have padded the buffer out.
- * Now We trim to the true length of the frame.
- */
- if (pskb_trim_rcsum(skb, len))
- goto out;
-
- if (iph->protocol != IPPROTO_UDP)
- goto out;
-
- len -= iph->ihl*4;
- uh = (struct udphdr *)(((char *)iph) + iph->ihl*4);
- ulen = ntohs(uh->len);
-
- if (ulen != len)
- goto out;
- if (checksum_udp(skb, uh, ulen, iph->saddr, iph->daddr))
- goto out;
- if (np->local_ip && np->local_ip != ntohl(iph->daddr))
- goto out;
- if (np->remote_ip && np->remote_ip != ntohl(iph->saddr))
- goto out;
- if (np->local_port && np->local_port != ntohs(uh->dest))
- goto out;
-
- np->rx_hook(np, ntohs(uh->source),
- (char *)(uh+1),
- ulen - sizeof(struct udphdr));
-
- kfree_skb(skb);
- return 1;
-
-out:
- if (atomic_read(&trapped)) {
- kfree_skb(skb);
- return 1;
- }
-
- return 0;
-}
-
int netpoll_parse_options(struct netpoll *np, char *opt)
{
char *cur=opt, *delim;
@@ -631,7 +323,6 @@ int netpoll_setup(struct netpoll *np)
struct net_device *ndev = NULL;
struct in_device *in_dev;
struct netpoll_info *npinfo;
- unsigned long flags;
int err;
if (np->dev_name)
@@ -650,13 +341,6 @@ int netpoll_setup(struct netpoll *np)
goto release;
}
- npinfo->rx_flags = 0;
- npinfo->rx_np = NULL;
- spin_lock_init(&npinfo->poll_lock);
- npinfo->poll_owner = -1;
-
- spin_lock_init(&npinfo->rx_lock);
- skb_queue_head_init(&npinfo->arp_tx);
skb_queue_head_init(&npinfo->txq);
INIT_DELAYED_WORK(&npinfo->tx_work, queue_process);
@@ -735,13 +419,6 @@ int netpoll_setup(struct netpoll *np)
np->name, HIPQUAD(np->local_ip));
}
- if (np->rx_hook) {
- spin_lock_irqsave(&npinfo->rx_lock, flags);
- npinfo->rx_flags |= NETPOLL_RX_ENABLED;
- npinfo->rx_np = np;
- spin_unlock_irqrestore(&npinfo->rx_lock, flags);
- }
-
/* fill up the skb queue */
refill_skbs();
@@ -771,20 +448,11 @@ core_initcall(netpoll_init);
void netpoll_cleanup(struct netpoll *np)
{
struct netpoll_info *npinfo;
- unsigned long flags;
if (np->dev) {
npinfo = np->dev->npinfo;
if (npinfo) {
- if (npinfo->rx_np == np) {
- spin_lock_irqsave(&npinfo->rx_lock, flags);
- npinfo->rx_np = NULL;
- npinfo->rx_flags &= ~NETPOLL_RX_ENABLED;
- spin_unlock_irqrestore(&npinfo->rx_lock, flags);
- }
-
if (atomic_dec_and_test(&npinfo->refcnt)) {
- skb_queue_purge(&npinfo->arp_tx);
skb_queue_purge(&npinfo->txq);
cancel_rearming_delayed_work(&npinfo->tx_work);
@@ -805,23 +473,7 @@ void netpoll_cleanup(struct netpoll *np)
np->dev = NULL;
}
-int netpoll_trap(void)
-{
- return atomic_read(&trapped);
-}
-
-void netpoll_set_trap(int trap)
-{
- if (trap)
- atomic_inc(&trapped);
- else
- atomic_dec(&trapped);
-}
-
-EXPORT_SYMBOL(netpoll_set_trap);
-EXPORT_SYMBOL(netpoll_trap);
EXPORT_SYMBOL(netpoll_parse_options);
EXPORT_SYMBOL(netpoll_setup);
EXPORT_SYMBOL(netpoll_cleanup);
EXPORT_SYMBOL(netpoll_send_udp);
-EXPORT_SYMBOL(netpoll_poll);
^ permalink raw reply related
* [PATCH 0/4 -rev1] Age Entry For IPv4 & IPv6 Route Table
From: Varun Chandramohan @ 2007-07-26 5:14 UTC (permalink / raw)
To: davem; +Cc: netdev, kaber, socketcan, shemminger, krkumar2
Hi,
This is the rev 1 of the patch set i sent out earlier. A few changes suggested were added.
Orginial Message:
According to the RFC 4292 (IP Forwarding Table MIB) there is a need for an age entry for all the routes in the routing table. The entry in the RFC is inetCidrRouteAge and oid is inetCidrRouteAge.1.10.
Many snmp application require this age entry. So iam adding the age field in the routing table for ipv4 and ipv6 and providing the interface for this value netlink.
Signed-off-by: Varun Chandramohan <varunc@linux.vnet.ibm.com>
---
^ permalink raw reply
* [PATCH 1/4 -rev1] New attribute RTA_AGE
From: Varun Chandramohan @ 2007-07-26 5:14 UTC (permalink / raw)
To: davem; +Cc: netdev, kaber, socketcan, shemminger, krkumar2
A new attribute RTA_AGE is added for the age value to be exported to userlevel using netlink
Signed-off-by: Varun Chandramohan <varunc@linux.vnet.ibm.com>
---
include/linux/rtnetlink.h | 1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h
index 6127858..884f507 100644
--- a/include/linux/rtnetlink.h
+++ b/include/linux/rtnetlink.h
@@ -263,6 +263,7 @@ enum rtattr_type_t
RTA_SESSION,
RTA_MP_ALGO, /* no longer used */
RTA_TABLE,
+ RTA_AGE,
__RTA_MAX
};
--
1.4.3.4
^ permalink raw reply related
* [PATCH 2/4 -rev1] Add new timeval_to_sec function
From: Varun Chandramohan @ 2007-07-26 5:15 UTC (permalink / raw)
To: davem; +Cc: netdev, kaber, socketcan, shemminger, krkumar2
A new function for converting timeval to time_t is added in time.h. Its a common function used in different
places. The timeout is now rounded up as per the suggestion.
Signed-off-by: Varun Chandramohan <varunc@linux.vnet.ibm.com>
---
include/linux/time.h | 11 +++++++++++
1 files changed, 11 insertions(+), 0 deletions(-)
diff --git a/include/linux/time.h b/include/linux/time.h
index dda9be6..908329a 100644
--- a/include/linux/time.h
+++ b/include/linux/time.h
@@ -147,6 +147,17 @@ static inline s64 timeval_to_ns(const st
}
/**
+ * timeval_to_sec - Convert timeval to seconds
+ * @tv: pointer to the timeval variable to be converted
+ *
+ * Returns the seconds representation of timeval parameter.
+ */
+static inline time_t timeval_to_sec(const struct timeval *tv)
+{
+ return (tv->tv_sec + (tv->tv_usec + 999999)/1000000);
+}
+
+/**
* ns_to_timespec - Convert nanoseconds to timespec
* @nsec: the nanoseconds value to be converted
*
--
1.4.3.4
^ permalink raw reply related
* [PATCH 3/4 -rev1] Initilize and populate age field
From: Varun Chandramohan @ 2007-07-26 5:15 UTC (permalink / raw)
To: davem; +Cc: netdev, kaber, socketcan, shemminger, krkumar2
The age field is filled with the current time at the time of creation of the route. When the routes are dumped
then the age value stored in the route structure is subtracted from the current time value and the difference is the age expressed in secs.
Signed-off-by: Varun Chandramohan <varunc@linux.vnet.ibm.com>
---
net/ipv4/fib_hash.c | 3 +++
net/ipv4/fib_lookup.h | 3 ++-
net/ipv4/fib_semantics.c | 16 +++++++++++++---
net/ipv4/fib_trie.c | 1 +
4 files changed, 19 insertions(+), 4 deletions(-)
diff --git a/net/ipv4/fib_hash.c b/net/ipv4/fib_hash.c
index 07e843a..faa7364 100644
--- a/net/ipv4/fib_hash.c
+++ b/net/ipv4/fib_hash.c
@@ -448,6 +448,7 @@ static int fn_hash_insert(struct fib_tab
fa->fa_info = fi;
fa->fa_type = cfg->fc_type;
fa->fa_scope = cfg->fc_scope;
+ fa->fa_age = 0;
state = fa->fa_state;
fa->fa_state &= ~FA_S_ACCESSED;
fib_hash_genid++;
@@ -507,6 +508,7 @@ static int fn_hash_insert(struct fib_tab
new_fa->fa_type = cfg->fc_type;
new_fa->fa_scope = cfg->fc_scope;
new_fa->fa_state = 0;
+ new_fa->fa_age = 0;
/*
* Insert new entry to the list.
@@ -697,6 +699,7 @@ fn_hash_dump_bucket(struct sk_buff *skb,
f->fn_key,
fz->fz_order,
fa->fa_tos,
+ &fa->fa_age,
fa->fa_info,
NLM_F_MULTI) < 0) {
cb->args[4] = i;
diff --git a/net/ipv4/fib_lookup.h b/net/ipv4/fib_lookup.h
index eef9eec..c9145b5 100644
--- a/net/ipv4/fib_lookup.h
+++ b/net/ipv4/fib_lookup.h
@@ -13,6 +13,7 @@ struct fib_alias {
u8 fa_type;
u8 fa_scope;
u8 fa_state;
+ time_t fa_age;
};
#define FA_S_ACCESSED 0x01
@@ -27,7 +28,7 @@ extern struct fib_info *fib_create_info(
extern int fib_nh_match(struct fib_config *cfg, struct fib_info *fi);
extern int fib_dump_info(struct sk_buff *skb, u32 pid, u32 seq, int event,
u32 tb_id, u8 type, u8 scope, __be32 dst,
- int dst_len, u8 tos, struct fib_info *fi,
+ int dst_len, u8 tos, time_t *age, struct fib_info *fi,
unsigned int);
extern void rtmsg_fib(int event, __be32 key, struct fib_alias *fa,
int dst_len, u32 tb_id, struct nl_info *info,
diff --git a/net/ipv4/fib_semantics.c b/net/ipv4/fib_semantics.c
index c434119..1822d92 100644
--- a/net/ipv4/fib_semantics.c
+++ b/net/ipv4/fib_semantics.c
@@ -278,7 +278,8 @@ static inline size_t fib_nlmsg_size(stru
+ nla_total_size(4) /* RTA_TABLE */
+ nla_total_size(4) /* RTA_DST */
+ nla_total_size(4) /* RTA_PRIORITY */
- + nla_total_size(4); /* RTA_PREFSRC */
+ + nla_total_size(4) /* RTA_PREFSRC */
+ + nla_total_size(4); /*RTA_AGE*/
/* space for nested metrics */
payload += nla_total_size((RTAX_MAX * nla_total_size(4)));
@@ -313,7 +314,7 @@ void rtmsg_fib(int event, __be32 key, st
err = fib_dump_info(skb, info->pid, seq, event, tb_id,
fa->fa_type, fa->fa_scope, key, dst_len,
- fa->fa_tos, fa->fa_info, nlm_flags);
+ fa->fa_tos, &fa->fa_age, fa->fa_info, nlm_flags);
if (err < 0) {
/* -EMSGSIZE implies BUG in fib_nlmsg_size() */
WARN_ON(err == -EMSGSIZE);
@@ -940,11 +941,12 @@ __be32 __fib_res_prefsrc(struct fib_resu
}
int fib_dump_info(struct sk_buff *skb, u32 pid, u32 seq, int event,
- u32 tb_id, u8 type, u8 scope, __be32 dst, int dst_len, u8 tos,
+ u32 tb_id, u8 type, u8 scope, __be32 dst, int dst_len, u8 tos, time_t *age,
struct fib_info *fi, unsigned int flags)
{
struct nlmsghdr *nlh;
struct rtmsg *rtm;
+ struct timeval tv;
nlh = nlmsg_put(skb, pid, seq, event, sizeof(*rtm), flags);
if (nlh == NULL)
@@ -985,6 +987,14 @@ int fib_dump_info(struct sk_buff *skb, u
NLA_PUT_U32(skb, RTA_FLOW, fi->fib_nh[0].nh_tclassid);
#endif
}
+
+ do_gettimeofday(&tv);
+ if (!*age) {
+ *age = timeval_to_sec(&tv);
+ NLA_PUT_U32(skb, RTA_AGE, *age);
+ } else {
+ NLA_PUT_U32(skb, RTA_AGE, timeval_to_sec(&tv) - *age);
+ }
#ifdef CONFIG_IP_ROUTE_MULTIPATH
if (fi->fib_nhs > 1) {
struct rtnexthop *rtnh;
diff --git a/net/ipv4/fib_trie.c b/net/ipv4/fib_trie.c
index 30e332a..be2d1d6 100644
--- a/net/ipv4/fib_trie.c
+++ b/net/ipv4/fib_trie.c
@@ -1884,6 +1884,7 @@ static int fn_trie_dump_fa(t_key key, in
xkey,
plen,
fa->fa_tos,
+ &fa->fa_age,
fa->fa_info, 0) < 0) {
cb->args[4] = i;
return -1;
--
1.4.3.4
^ permalink raw reply related
* [PATCH 4/4 -rev1] Initialize and fill IPv6 route age
From: Varun Chandramohan @ 2007-07-26 5:16 UTC (permalink / raw)
To: davem; +Cc: netdev, kaber, socketcan, shemminger, krkumar2
The age field of the ipv6 route structures are initilized with the current timeval at the time of route
creation. When the route dump is called the route age value stored in the structure is subtracted from the
present timeval and the difference is passed on as the route age. The dumpflg clarity is added as per
suggestion.
Signed-off-by: Varun Chandramohan <varunc@linux.vnet.ibm.com>
---
include/net/ip6_fib.h | 1 +
include/net/ip6_route.h | 3 +++
net/ipv6/addrconf.c | 5 +++++
net/ipv6/route.c | 23 +++++++++++++++++++----
4 files changed, 28 insertions(+), 4 deletions(-)
diff --git a/include/net/ip6_fib.h b/include/net/ip6_fib.h
index c48ea87..e30a1cf 100644
--- a/include/net/ip6_fib.h
+++ b/include/net/ip6_fib.h
@@ -98,6 +98,7 @@ struct rt6_info
u32 rt6i_flags;
u32 rt6i_metric;
+ time_t rt6i_age;
atomic_t rt6i_ref;
struct fib6_table *rt6i_table;
diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h
index 5456fdd..fc9716c 100644
--- a/include/net/ip6_route.h
+++ b/include/net/ip6_route.h
@@ -36,6 +36,9 @@ struct route_info {
#define RT6_LOOKUP_F_REACHABLE 0x2
#define RT6_LOOKUP_F_HAS_SADDR 0x4
+#define RT6_SET_ROUTE_INFO 0x0
+#define RT6_GET_ROUTE_INFO 0x1
+
extern struct rt6_info ip6_null_entry;
#ifdef CONFIG_IPV6_MULTIPLE_TABLES
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 5a5f8bd..715c766 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -4187,6 +4187,7 @@ EXPORT_SYMBOL(unregister_inet6addr_notif
int __init addrconf_init(void)
{
+ struct timeval tv;
int err = 0;
/* The addrconf netdev notifier requires that loopback_dev
@@ -4214,10 +4215,14 @@ int __init addrconf_init(void)
if (err)
return err;
+ do_gettimeofday(&tv);
ip6_null_entry.rt6i_idev = in6_dev_get(&loopback_dev);
+ ip6_null_entry.rt6i_age = timeval_to_sec(&tv);
#ifdef CONFIG_IPV6_MULTIPLE_TABLES
ip6_prohibit_entry.rt6i_idev = in6_dev_get(&loopback_dev);
+ ip6_prohibit_entry.rt6i_age = timeval_to_sec(&tv);
ip6_blk_hole_entry.rt6i_idev = in6_dev_get(&loopback_dev);
+ ip6_blk_hole_entry.rt6i_age = timeval_to_sec(&tv);
#endif
register_netdevice_notifier(&ipv6_dev_notf);
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index fe8d983..686566f 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -600,7 +600,14 @@ static int __ip6_ins_rt(struct rt6_info
{
int err;
struct fib6_table *table;
+ struct timeval tv;
+ do_gettimeofday(&tv);
+ /* Update the timeval for new routes
+ * We add it here to make it common irrespective
+ * of how the new route is added.
+ */
+ rt->rt6i_age = timeval_to_sec(&tv);
table = rt->rt6i_table;
write_lock_bh(&table->tb6_lock);
err = fib6_add(&table->tb6_root, rt, info);
@@ -2111,6 +2118,7 @@ static inline size_t rt6_nlmsg_size(void
+ nla_total_size(4) /* RTA_IIF */
+ nla_total_size(4) /* RTA_OIF */
+ nla_total_size(4) /* RTA_PRIORITY */
+ + nla_total_size(4) /*RTA_AGE*/
+ RTAX_MAX * nla_total_size(4) /* RTA_METRICS */
+ nla_total_size(sizeof(struct rta_cacheinfo));
}
@@ -2118,10 +2126,11 @@ static inline size_t rt6_nlmsg_size(void
static int rt6_fill_node(struct sk_buff *skb, struct rt6_info *rt,
struct in6_addr *dst, struct in6_addr *src,
int iif, int type, u32 pid, u32 seq,
- int prefix, unsigned int flags)
+ int prefix, unsigned int flags, int dumpflg)
{
struct rtmsg *rtm;
struct nlmsghdr *nlh;
+ struct timeval tv;
long expires;
u32 table;
@@ -2185,6 +2194,12 @@ static int rt6_fill_node(struct sk_buff
if (ipv6_get_saddr(&rt->u.dst, dst, &saddr_buf) == 0)
NLA_PUT(skb, RTA_PREFSRC, 16, &saddr_buf);
}
+
+ do_gettimeofday(&tv);
+ if (dumpflg == RT6_GET_ROUTE_INFO)
+ NLA_PUT_U32(skb, RTA_AGE, timeval_to_sec(&tv) - rt->rt6i_age);
+ else
+ NLA_PUT_U32(skb, RTA_AGE, rt->rt6i_age);
if (rtnetlink_put_metrics(skb, rt->u.dst.metrics) < 0)
goto nla_put_failure;
@@ -2222,7 +2237,7 @@ int rt6_dump_route(struct rt6_info *rt,
return rt6_fill_node(arg->skb, rt, NULL, NULL, 0, RTM_NEWROUTE,
NETLINK_CB(arg->cb->skb).pid, arg->cb->nlh->nlmsg_seq,
- prefix, NLM_F_MULTI);
+ prefix, NLM_F_MULTI, RT6_GET_ROUTE_INFO);
}
static int inet6_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void *arg)
@@ -2287,7 +2302,7 @@ static int inet6_rtm_getroute(struct sk_
err = rt6_fill_node(skb, rt, &fl.fl6_dst, &fl.fl6_src, iif,
RTM_NEWROUTE, NETLINK_CB(in_skb).pid,
- nlh->nlmsg_seq, 0, 0);
+ nlh->nlmsg_seq, 0, 0, RT6_GET_ROUTE_INFO);
if (err < 0) {
kfree_skb(skb);
goto errout;
@@ -2316,7 +2331,7 @@ void inet6_rt_notify(int event, struct r
if (skb == NULL)
goto errout;
- err = rt6_fill_node(skb, rt, NULL, NULL, 0, event, pid, seq, 0, 0);
+ err = rt6_fill_node(skb, rt, NULL, NULL, 0, event, pid, seq, 0, 0, RT6_SET_ROUTE_INFO);
if (err < 0) {
/* -EMSGSIZE implies BUG in rt6_nlmsg_size() */
WARN_ON(err == -EMSGSIZE);
--
1.4.3.4
^ permalink raw reply related
* [PATCH] gfar: Fix modpost warning
From: Kumar Gala @ 2007-07-26 5:52 UTC (permalink / raw)
To: Jeff Garzik; +Cc: netdev, linuxppc-dev
Fix the following modpost warning:
WARNING: vmlinux.o(.init.text+0x1aa6c): Section mismatch: reference to .exit.text:gfar_mdio_exit (between 'gfar_init' and 'gfar_mdio_init')
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
---
drivers/net/gianfar_mii.c | 2 +-
drivers/net/gianfar_mii.h | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/gianfar_mii.c b/drivers/net/gianfar_mii.c
index ac3596f..100bf41 100644
--- a/drivers/net/gianfar_mii.c
+++ b/drivers/net/gianfar_mii.c
@@ -245,7 +245,7 @@ int __init gfar_mdio_init(void)
return driver_register(&gianfar_mdio_driver);
}
-void __exit gfar_mdio_exit(void)
+void gfar_mdio_exit(void)
{
driver_unregister(&gianfar_mdio_driver);
}
diff --git a/drivers/net/gianfar_mii.h b/drivers/net/gianfar_mii.h
index 5d34004..b373091 100644
--- a/drivers/net/gianfar_mii.h
+++ b/drivers/net/gianfar_mii.h
@@ -42,5 +42,5 @@ struct gfar_mii {
int gfar_mdio_read(struct mii_bus *bus, int mii_id, int regnum);
int gfar_mdio_write(struct mii_bus *bus, int mii_id, int regnum, u16 value);
int __init gfar_mdio_init(void);
-void __exit gfar_mdio_exit(void);
+void gfar_mdio_exit(void);
#endif /* GIANFAR_PHY_H */
--
1.5.2.2
^ permalink raw reply related
* Re: [PATCH 2/4 -rev1] Add new timeval_to_sec function
From: Eric Dumazet @ 2007-07-26 5:58 UTC (permalink / raw)
To: Varun Chandramohan; +Cc: davem, netdev, kaber, socketcan, shemminger, krkumar2
In-Reply-To: <20070726104526.0348e746.varunc@linux.vnet.ibm.com>
Varun Chandramohan a écrit :
> A new function for converting timeval to time_t is added in time.h. Its a common function used in different
> places. The timeout is now rounded up as per the suggestion.
>
> Signed-off-by: Varun Chandramohan <varunc@linux.vnet.ibm.com>
> ---
> include/linux/time.h | 11 +++++++++++
> 1 files changed, 11 insertions(+), 0 deletions(-)
>
> diff --git a/include/linux/time.h b/include/linux/time.h
> index dda9be6..908329a 100644
> --- a/include/linux/time.h
> +++ b/include/linux/time.h
> @@ -147,6 +147,17 @@ static inline s64 timeval_to_ns(const st
> }
>
> /**
> + * timeval_to_sec - Convert timeval to seconds
> + * @tv: pointer to the timeval variable to be converted
> + *
> + * Returns the seconds representation of timeval parameter.
> + */
> +static inline time_t timeval_to_sec(const struct timeval *tv)
> +{
> + return (tv->tv_sec + (tv->tv_usec + 999999)/1000000);
> +}
> +
> +/**
return tv->tv_sec + (tv->tv_usec ? 1 : 0);
is much faster
Also, the comment is not accurate, as it doesnt mention the roundup at all.
^ permalink raw reply
* Re: [PATCH 3/4 -rev1] Initilize and populate age field
From: Eric Dumazet @ 2007-07-26 6:03 UTC (permalink / raw)
To: Varun Chandramohan; +Cc: davem, netdev, kaber, socketcan, shemminger, krkumar2
In-Reply-To: <20070726104554.cc95eedd.varunc@linux.vnet.ibm.com>
Varun Chandramohan a écrit :
> The age field is filled with the current time at the time of creation of the route. When the routes are dumped
> then the age value stored in the route structure is subtracted from the current time value and the difference is the age expressed in secs.
>
> Signed-off-by: Varun Chandramohan <varunc@linux.vnet.ibm.com>
> ---
> net/ipv4/fib_hash.c | 3 +++
> net/ipv4/fib_lookup.h | 3 ++-
> net/ipv4/fib_semantics.c | 16 +++++++++++++---
> net/ipv4/fib_trie.c | 1 +
> 4 files changed, 19 insertions(+), 4 deletions(-)
>
> diff --git a/net/ipv4/fib_hash.c b/net/ipv4/fib_hash.c
> index 07e843a..faa7364 100644
> --- a/net/ipv4/fib_hash.c
> +++ b/net/ipv4/fib_hash.c
> @@ -448,6 +448,7 @@ static int fn_hash_insert(struct fib_tab
> fa->fa_info = fi;
> fa->fa_type = cfg->fc_type;
> fa->fa_scope = cfg->fc_scope;
> + fa->fa_age = 0;
> state = fa->fa_state;
> fa->fa_state &= ~FA_S_ACCESSED;
> fib_hash_genid++;
> @@ -507,6 +508,7 @@ static int fn_hash_insert(struct fib_tab
> new_fa->fa_type = cfg->fc_type;
> new_fa->fa_scope = cfg->fc_scope;
> new_fa->fa_state = 0;
> + new_fa->fa_age = 0;
>
> /*
> * Insert new entry to the list.
> @@ -697,6 +699,7 @@ fn_hash_dump_bucket(struct sk_buff *skb,
> f->fn_key,
> fz->fz_order,
> fa->fa_tos,
> + &fa->fa_age,
> fa->fa_info,
> NLM_F_MULTI) < 0) {
> cb->args[4] = i;
> diff --git a/net/ipv4/fib_lookup.h b/net/ipv4/fib_lookup.h
> index eef9eec..c9145b5 100644
> --- a/net/ipv4/fib_lookup.h
> +++ b/net/ipv4/fib_lookup.h
> @@ -13,6 +13,7 @@ struct fib_alias {
> u8 fa_type;
> u8 fa_scope;
> u8 fa_state;
> + time_t fa_age;
> };
>
> #define FA_S_ACCESSED 0x01
> @@ -27,7 +28,7 @@ extern struct fib_info *fib_create_info(
> extern int fib_nh_match(struct fib_config *cfg, struct fib_info *fi);
> extern int fib_dump_info(struct sk_buff *skb, u32 pid, u32 seq, int event,
> u32 tb_id, u8 type, u8 scope, __be32 dst,
> - int dst_len, u8 tos, struct fib_info *fi,
> + int dst_len, u8 tos, time_t *age, struct fib_info *fi,
> unsigned int);
> extern void rtmsg_fib(int event, __be32 key, struct fib_alias *fa,
> int dst_len, u32 tb_id, struct nl_info *info,
> diff --git a/net/ipv4/fib_semantics.c b/net/ipv4/fib_semantics.c
> index c434119..1822d92 100644
> --- a/net/ipv4/fib_semantics.c
> +++ b/net/ipv4/fib_semantics.c
> @@ -278,7 +278,8 @@ static inline size_t fib_nlmsg_size(stru
> + nla_total_size(4) /* RTA_TABLE */
> + nla_total_size(4) /* RTA_DST */
> + nla_total_size(4) /* RTA_PRIORITY */
> - + nla_total_size(4); /* RTA_PREFSRC */
> + + nla_total_size(4) /* RTA_PREFSRC */
> + + nla_total_size(4); /*RTA_AGE*/
Not clear to me what happens on platforms where sizeof(time_t) is 8
>
> /* space for nested metrics */
> payload += nla_total_size((RTAX_MAX * nla_total_size(4)));
> @@ -313,7 +314,7 @@ void rtmsg_fib(int event, __be32 key, st
>
> err = fib_dump_info(skb, info->pid, seq, event, tb_id,
> fa->fa_type, fa->fa_scope, key, dst_len,
> - fa->fa_tos, fa->fa_info, nlm_flags);
> + fa->fa_tos, &fa->fa_age, fa->fa_info, nlm_flags);
> if (err < 0) {
> /* -EMSGSIZE implies BUG in fib_nlmsg_size() */
> WARN_ON(err == -EMSGSIZE);
> @@ -940,11 +941,12 @@ __be32 __fib_res_prefsrc(struct fib_resu
> }
>
> int fib_dump_info(struct sk_buff *skb, u32 pid, u32 seq, int event,
> - u32 tb_id, u8 type, u8 scope, __be32 dst, int dst_len, u8 tos,
> + u32 tb_id, u8 type, u8 scope, __be32 dst, int dst_len, u8 tos, time_t *age,
> struct fib_info *fi, unsigned int flags)
> {
> struct nlmsghdr *nlh;
> struct rtmsg *rtm;
> + struct timeval tv;
>
> nlh = nlmsg_put(skb, pid, seq, event, sizeof(*rtm), flags);
> if (nlh == NULL)
> @@ -985,6 +987,14 @@ int fib_dump_info(struct sk_buff *skb, u
> NLA_PUT_U32(skb, RTA_FLOW, fi->fib_nh[0].nh_tclassid);
> #endif
> }
> +
> + do_gettimeofday(&tv);
> + if (!*age) {
> + *age = timeval_to_sec(&tv);
> + NLA_PUT_U32(skb, RTA_AGE, *age);
here, what happens if sizeof(time_t) is not 4 ?
> + } else {
> + NLA_PUT_U32(skb, RTA_AGE, timeval_to_sec(&tv) - *age);
> + }
> #ifdef CONFIG_IP_ROUTE_MULTIPATH
> if (fi->fib_nhs > 1) {
> struct rtnexthop *rtnh;
> diff --git a/net/ipv4/fib_trie.c b/net/ipv4/fib_trie.c
> index 30e332a..be2d1d6 100644
> --- a/net/ipv4/fib_trie.c
> +++ b/net/ipv4/fib_trie.c
> @@ -1884,6 +1884,7 @@ static int fn_trie_dump_fa(t_key key, in
> xkey,
> plen,
> fa->fa_tos,
> + &fa->fa_age,
> fa->fa_info, 0) < 0) {
> cb->args[4] = i;
> return -1;
^ permalink raw reply
* Re: commit 7e92b4fc34 - x86, serial: convert legacy COM ports to platform devices - broke my serial console
From: H. Peter Anvin @ 2007-07-26 6:09 UTC (permalink / raw)
To: Jeff Garzik
Cc: Adrian Bunk, Andrew Morton, Alexey Dobriyan, linux-kernel, netdev
In-Reply-To: <alpine.LFD.0.999.0707241113180.3607@woody.linux-foundation.org>
Jeff Garzik wrote:
>
> On Tue, 24 Jul 2007, Adrian Bunk wrote:
>> buffered_rmqueue() and prep_new_page() are static functions with only
>> one caller each, and for the normal non-debug case it's a really nice
>> optimization to have them inlined automatically.
>
> I'm not at all sure I agree.
>
> Inlining big functions doesn't actually tend to generally generate any
> better code, so if gcc's logic is "single callsite - always inline", then
> that logic is likely not right.
>
Only up to a threshold, as far as I know.
-hpa
^ permalink raw reply
* Re: [PATCH 2/4 -rev1] Add new timeval_to_sec function
From: Oliver Hartkopp @ 2007-07-26 6:17 UTC (permalink / raw)
To: Eric Dumazet
Cc: Varun Chandramohan, davem, netdev, kaber, socketcan, shemminger,
krkumar2
In-Reply-To: <46A837FE.5060900@cosmosbay.com>
Eric Dumazet wrote:
> Varun Chandramohan a écrit :
>>
>> + return (tv->tv_sec + (tv->tv_usec + 999999)/1000000);
>
> return tv->tv_sec + (tv->tv_usec ? 1 : 0);
>
> is much faster
It additionally wins the design award in opposite to my suggestion ;-)
Thanks!
^ 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