* [PATCH net-2.6] gre: fix improper error handling
From: Dmitry Kozlov @ 2011-07-22 9:26 UTC (permalink / raw)
To: davem; +Cc: netdev
Fix improper protocol err_handler, current implementation is fully
unapplicable and may cause kernel crash due to double kfree_skb.
Signed-off-by: Dmitry Kozlov <xeb@mail.ru>
--
net/ipv4/gre.c | 23 +++++++----------------
1 files changed, 7 insertions(+), 16 deletions(-)
diff --git a/net/ipv4/gre.c b/net/ipv4/gre.c
index c6933f2..a8126fe 100644
--- a/net/ipv4/gre.c
+++ b/net/ipv4/gre.c
@@ -15,6 +15,7 @@
#include <linux/kmod.h>
#include <linux/skbuff.h>
#include <linux/in.h>
+#include <linux/ip.h>
#include <linux/netdevice.h>
#include <linux/version.h>
#include <linux/spinlock.h>
@@ -97,27 +98,17 @@ drop:
static void gre_err(struct sk_buff *skb, u32 info)
{
const struct gre_protocol *proto;
- u8 ver;
-
- if (!pskb_may_pull(skb, 12))
- goto drop;
-
- ver = skb->data[1]&0x7f;
+ const struct iphdr *iph = (const struct iphdr *)skb->data;
+ u8 ver = skb->data[(iph->ihl<<2) + 1]&0x7f;
+
if (ver >= GREPROTO_MAX)
- goto drop;
+ return;
rcu_read_lock();
proto = rcu_dereference(gre_proto[ver]);
- if (!proto || !proto->err_handler)
- goto drop_unlock;
- proto->err_handler(skb, info);
- rcu_read_unlock();
- return;
-
-drop_unlock:
+ if (proto && proto->err_handler)
+ proto->err_handler(skb, info);
rcu_read_unlock();
-drop:
- kfree_skb(skb);
}
static const struct net_protocol net_gre_protocol = {
^ permalink raw reply related
* Re: IPv6: autoconfiguration and suspend/resume or link down/up
From: Herbert Xu @ 2011-07-22 9:21 UTC (permalink / raw)
To: David Miller; +Cc: jbohac, netdev, shemminger
In-Reply-To: <20110722.010628.1678943945721626312.davem@davemloft.net>
On Fri, Jul 22, 2011 at 01:06:28AM -0700, David Miller wrote:
>
> Suspend is even more important because while we were suspended we
> could be on the same network but the routers present and available
> might have changed completely.
Unfortunately virtual machine live migration also uses the suspend
& resume mechanism. In that case we don't wish to renegotiate
everything, since the goal is to minimise the outage window.
This would suggest that putting the policy in user-space may be the
best option.
Cheers,
--
Email: Herbert Xu <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
^ permalink raw reply
* Re: [PATCH net-2.6] jme: Fix unmap error (Causing system freeze)
From: Guo-Fu Tseng @ 2011-07-22 8:49 UTC (permalink / raw)
To: Michał Mirosław
Cc: Jason Lamb, linux-netdev, Jason Long, stable, Marcus Becker,
Aries Lee, Devinchiu, Marc Schiffbauer, David Miller
In-Reply-To: <CAHXqBF+1aJ0O9WxrZTKfYMm57emysfPUnA-=M9DZN5kUGnhW8g@mail.gmail.com>
On Fri, 22 Jul 2011 09:45:42 +0200, Michał Mirosław wrote
> 2011/7/21 <cooldavid@cooldavid.org>:
> > From: Guo-Fu Tseng <cooldavid@cooldavid.org>
> >
> > This patch add the missing dma_unmap().
> > Which solved the critical issue of system freeze on heavy load.
> >
> > Michal Miroslaw's rejected patch:
> > [PATCH v2 10/46] net: jme: convert to generic DMA API
> > Pointed out the issue also, thank you Michal.
> > But the fix was incorrect. It would unmap needed address
> > when low memory.
>
> My patch also fixed another issue - from correctness point of view,
> you should not use virt_to_page+dma_map_page here, but dma_map_single.
>
> Best Regards,
> Michał Mirosław
Hi Michał:
Yes, I did noticed that. :)
And I agreed that using {dma|pci}_map_single is better than
calling virt_to_page in driver.
It's just that I'm not sure if it is really preferred to use
dma_map_single over pci_map_single for PCI device driver?
You are very welcome to send another patch for converting the
DMA API using.
Guo-Fu Tseng
_______________________________________________
stable mailing list
stable@linux.kernel.org
http://linux.kernel.org/mailman/listinfo/stable
^ permalink raw reply
* tcp/udp checksum on loopback interface
From: Pierre Louis Aublin @ 2011-07-22 8:01 UTC (permalink / raw)
To: netdev
Hello everybody
I am interested in the reliability of TCP and UDP using the loopback
interface.
I found that there is no checksum verification on the body of packets
transmitted through the loopback interface :
http://kerneltrap.org/mailarchive/linux-netdev/2007/10/2/327241
I ran a simple test using iperf and wireshark and I indeed found the
validation of the checksum is disabled at packets reception.
At the reception, I guess the piece of code responsible for this
behaviour is (line 1196 of net/ipv4/udp.c, in udp_recvmsg() for UDP) :
if (skb_csum_unnecessary(skb))
err = skb_copy_datagram_iovec(skb, sizeof(struct udphdr),
msg->msg_iov, len);
Is this the responsible piece of code? If yes, where does the
CHECKSUM_UNNECESSARY flag is set?
However, I am not sure whether the checksum is computed on the whole
packet at emission.
In __ip_append_data (net/ipv4/ip_output.c), I found the flag csummode is
at the beginning at CHECKSUM_NONE.
Then, it is augmented using the dev->features flag of the network device
(e.g. the loopback interface).
As the loopback interface has the NETIF_F_NO_CSUM flag, csummode stays
at CHECKSUM_NONE and the checksum is not computed on the body of the
message.
Can you confirm the checksum is computed only on the header of the
packet when sending it through the loopback interface?
Finally, why this behaviour? Is it because you assume message can not
get corrupted while staying on the same machine?
Thank you in advance
Pierre Louis Aublin
^ permalink raw reply
* Re: IPv6: autoconfiguration and suspend/resume or link down/up
From: David Miller @ 2011-07-22 8:06 UTC (permalink / raw)
To: jbohac; +Cc: netdev, herbert, shemminger
In-Reply-To: <20110719180222.GA7509@midget.suse.cz>
From: Jiri Bohac <jbohac@suse.cz>
Date: Tue, 19 Jul 2011 20:02:53 +0200
> When the cable is unplugged and plugged in again, we already get
> notified through linkwatch -> netdev_state_change ->
> -> call_netdevice_notifiers(NETDEV_CHANGE, ...)
> However, if the device has already been autoconfigured,
> addrconf_notify() only handles this event by printing a
> message.
>
> So my idea was to:
> - handle link up/down in addrconf_notify() similarly to
> NETDEV_UP/NETDEV_DOWN
>
> - on suspend, faking a link down event; on resume, faking a link up event
> (or better, having a special event type for suspend/resume)
>
> This would cause autoconfiguration to be restarted on resume as
> well as cable plug/unplug, solving both the above problems.
>
> Or do we want to completely rely on userspace tools
> (networkmanager/ifplug) and expect them to do NETDEV_DOWN on
> unplug/suspend and NETDEV_UP on plug/resume?
>
> Any thoughts?
This is an oft-reocurring discussion, what to do on link up/down
events wrt. all sorts of autoconfiguration.
My gut instinct is that on any link state change (physical link down,
suspend) we should be prepared to renegotiate everything and anything
since as you state we could be on a different physical network.
Suspend is even more important because while we were suspended we
could be on the same network but the routers present and available
might have changed completely.
In userspace I've noticed that we've grown an ecosystem of stupid
tools and facilities, none (or very few) of which monitor link status
in order to do handle things intelligently. DHCP servers are a great
example. DHCP servers spit out broadcast discover packets before we
even have a link up.
Filling this void is NetworkManager, which does listen on a netlink
socket for device state changes, hotplug, etc. And in response it
explicitly tells various facilities to reprobe the network.
This is why I'm reluctant to give NetworkManager a hard time, because
whilst it's a huge beast, it is at least trying to do the right thing.
:-)
^ permalink raw reply
* Re: [net-next 10/10] ixgbe: convert to ndo_fix_features
From: Michal Miroslaw @ 2011-07-22 8:02 UTC (permalink / raw)
To: Jeff Kirsher; +Cc: davem, Don Skidmore, netdev, gospo, sassmann
In-Reply-To: <1311314951-11142-11-git-send-email-jeffrey.t.kirsher@intel.com>
On Thu, Jul 21, 2011 at 11:09:11PM -0700, Jeff Kirsher wrote:
> From: Don Skidmore <donald.c.skidmore@intel.com>
>
> Private rx_csum flags are now duplicate of netdev->features &
> NETIF_F_RXCSUM. We remove those duplicates and now use the net_device_ops
> ndo_set_features. This was based on the original patch submitted by
> Michal Miroslaw <mirq-linux@rere.qmqm.pl>. I also removed the special
> case not requiring a reset for X540 hardware. It is needed just as it is
> in 82599 hardware.
Looks mostly good now. Minor hints below.
[...]
> +static u32 ixgbe_fix_features(struct net_device *netdev, u32 data)
> +{
[...]
> + /* Turn off LRO if not RSC capable or invalid ITR settings */
> + if (!(adapter->flags2 & IXGBE_FLAG2_RSC_CAPABLE)) {
> + data &= ~NETIF_F_LRO;
> + } else if (!(adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED) &&
> + (adapter->rx_itr_setting != 1 &&
> + adapter->rx_itr_setting > IXGBE_MAX_RSC_INT_RATE)) {
> + data &= ~NETIF_F_LRO;
> + e_info(probe, "rx-usecs set too low, not enabling RSC\n");
> + }
Better:
... else if (data & NETIF_F_LRO && adapter->rx_itr_setting != 1 && adapter->rx_itr_setting > IXGBE_MAX_RSC_INT_RATE) {
e_info(...)
data &= ~NETIF_F_LRO;
}
> +
> + return data;
> +}
> +
> +static int ixgbe_set_features(struct net_device *netdev, u32 data)
> +{
> + struct ixgbe_adapter *adapter = netdev_priv(netdev);
> + bool need_reset = false;
> +
> + /* If Rx checksum is disabled, then RSC/LRO should also be disabled */
> + if (!(data & NETIF_F_RXCSUM))
> + adapter->flags &= ~IXGBE_FLAG_RX_CSUM_ENABLED;
> + else
> + adapter->flags |= IXGBE_FLAG_RX_CSUM_ENABLED;
This exactly mirrors NETIF_F_RXCSUM. Waiting for later cleanup?
[...]
> + /*
> + * Check if Flow Director n-tuple support was enabled or disabled. If
> + * the state changed, we need to reset.
> + */
> + if (!(adapter->flags & IXGBE_FLAG_FDIR_PERFECT_CAPABLE)) {
> + /* turn off ATR, enable perfect filters and reset */
> + if (data & NETIF_F_NTUPLE) {
> + adapter->flags &= ~IXGBE_FLAG_FDIR_HASH_CAPABLE;
> + adapter->flags |= IXGBE_FLAG_FDIR_PERFECT_CAPABLE;
> + need_reset = true;
> + }
> + } else if (!(data & NETIF_F_NTUPLE)) {
> + /* turn off Flow Director, set ATR and reset */
> + adapter->flags &= ~IXGBE_FLAG_FDIR_PERFECT_CAPABLE;
> + if ((adapter->flags & IXGBE_FLAG_RSS_ENABLED) &&
> + !(adapter->flags & IXGBE_FLAG_DCB_ENABLED))
> + adapter->flags |= IXGBE_FLAG_FDIR_HASH_CAPABLE;
> + need_reset = true;
> + }
You could make this more readable:
old = adapter->flags;
if (data & NETIF_F_NTUPLE) {
adapter->flags |= IXGBE_FLAG_FDIR_PERFECT_CAPABLE;
adapter->flags &= ~IXGBE_FLAG_FDIR_HASH_CAPABLE;
} else {
adapter->flags &= ~IXGBE_FLAG_FDIR_PERFECT_CAPABLE;
if ((adapter->flags & IXGBE_FLAG_RSS_ENABLED) &&
!(adapter->flags & IXGBE_FLAG_DCB_ENABLED))
adapter->flags |= IXGBE_FLAG_FDIR_HASH_CAPABLE;
}
if (old != adapter->flags)
need_reset = true;
> +
> + if (need_reset)
> + ixgbe_do_reset(netdev);
> +
> + return 0;
> +
> +}
> +
> static const struct net_device_ops ixgbe_netdev_ops = {
> .ndo_open = ixgbe_open,
> .ndo_stop = ixgbe_close,
> @@ -7153,6 +7228,8 @@ static const struct net_device_ops ixgbe_netdev_ops = {
> .ndo_fcoe_disable = ixgbe_fcoe_disable,
> .ndo_fcoe_get_wwn = ixgbe_fcoe_get_wwn,
> #endif /* IXGBE_FCOE */
> + .ndo_set_features = ixgbe_set_features,
> + .ndo_fix_features = ixgbe_fix_features,
> };
>
> static void __devinit ixgbe_probe_vf(struct ixgbe_adapter *adapter,
> @@ -7420,20 +7497,24 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev,
>
> netdev->features = NETIF_F_SG |
> NETIF_F_IP_CSUM |
> + NETIF_F_IPV6_CSUM |
> NETIF_F_HW_VLAN_TX |
> NETIF_F_HW_VLAN_RX |
> - NETIF_F_HW_VLAN_FILTER;
> + NETIF_F_HW_VLAN_FILTER |
> + NETIF_F_TSO |
> + NETIF_F_TSO6 |
> + NETIF_F_GRO |
> + NETIF_F_RXHASH |
> + NETIF_F_RXCSUM;
Drop NETIF_F_GRO here, as its always set by network core now.
> - netdev->features |= NETIF_F_IPV6_CSUM;
> - netdev->features |= NETIF_F_TSO;
> - netdev->features |= NETIF_F_TSO6;
> - netdev->features |= NETIF_F_GRO;
> - netdev->features |= NETIF_F_RXHASH;
> + netdev->hw_features = netdev->features;
>
> switch (adapter->hw.mac.type) {
> case ixgbe_mac_82599EB:
> case ixgbe_mac_X540:
> netdev->features |= NETIF_F_SCTP_CSUM;
> + netdev->hw_features |= NETIF_F_SCTP_CSUM |
> + NETIF_F_NTUPLE;
NTUPLE disabled by default. That's the idea?
Best Regards,
Michał Mirosław
^ permalink raw reply
* Re: icmp_route_lookup returns wrong source address information
From: David Miller @ 2011-07-22 7:59 UTC (permalink / raw)
To: fw; +Cc: netdev
In-Reply-To: <20110720192258.GA21528@Chamillionaire.breakpoint.cc>
From: Florian Westphal <fw@strlen.de>
Date: Wed, 20 Jul 2011 21:22:58 +0200
> With linux-2.6 and net-next trees, whenever an icmp error message is
> sent in response to a to-be-forwarded packet, the destination address
> of the original packet is used as the IP header source address.
>
> Example:
> $ traceroute breakpoint.cc
> traceroute to breakpoint.cc (85.10.199.196), 30 hops max, 40 byte packets
> 1 chamillionaire.breakpoint.cc (85.10.199.196) 0.476 ms 0.468 ms 0.793 ms
>
> But the expected 1st hop is 192.168.20.7 in my setup.
>
> I bisected this down to 77968b78242ee25e2a4d759f0fca8dd52df6d479
> ("ipv4: Pass flow keys down into datagram packet building engine.")
>
> Specifically, it is caused by this hunk:
...
> Problem is that xfrm_decode_session_reverse() may rebuild fl4 from scratch.
>
> In my setup, "goto relookup_failed" will be hit a bit later on, and fl4->saddr will
> be set to iph->daddr...
Thanks for the detailed report Florian, I'll look into this.
^ permalink raw reply
* Re: [PATCH net-2.6] jme: Fix unmap error (Causing system freeze)
From: Michał Mirosław @ 2011-07-22 7:45 UTC (permalink / raw)
To: cooldavid
Cc: David Miller, linux-netdev, Aries Lee, Devinchiu, Jason Lamb,
Marcus Becker, Marc Schiffbauer, Jason Long, stable
In-Reply-To: <1311217056-9073-1-git-send-email-cooldavid@cooldavid.org>
2011/7/21 <cooldavid@cooldavid.org>:
> From: Guo-Fu Tseng <cooldavid@cooldavid.org>
>
> This patch add the missing dma_unmap().
> Which solved the critical issue of system freeze on heavy load.
>
> Michal Miroslaw's rejected patch:
> [PATCH v2 10/46] net: jme: convert to generic DMA API
> Pointed out the issue also, thank you Michal.
> But the fix was incorrect. It would unmap needed address
> when low memory.
My patch also fixed another issue - from correctness point of view,
you should not use virt_to_page+dma_map_page here, but dma_map_single.
Best Regards,
Michał Mirosław
^ permalink raw reply
* Re: linux-next: manual merge of the trivial tree with the net tree
From: Francois Romieu @ 2011-07-22 6:42 UTC (permalink / raw)
To: Stephen Rothwell
Cc: Jiri Kosina, linux-next, linux-kernel, Phil Carmody, David Miller,
netdev
In-Reply-To: <20110722141842.963b38d45ebf8cb5d0a25fcf@canb.auug.org.au>
Stephen Rothwell <sfr@canb.auug.org.au> :
[...]
> diff --cc drivers/net/r8169.c
> index 40bcb82,6f36306..0000000
> --- a/drivers/net/r8169.c
> +++ b/drivers/net/r8169.c
> @@@ -682,19 -666,8 +682,19 @@@ struct rtl8169_private
> struct rtl8169_counters counters;
> u32 saved_wolopts;
>
> - const struct firmware *fw;
> + struct rtl_fw {
> + const struct firmware *fw;
> +
> +#define RTL_VER_SIZE 32
> +
> + char version[RTL_VER_SIZE];
> +
> + struct rtl_fw_phy_action {
> + __le32 *code;
> + size_t size;
> + } phy_action;
> + } *rtl_fw;
> - #define RTL_FIRMWARE_UNKNOWN ERR_PTR(-EAGAIN);
> + #define RTL_FIRMWARE_UNKNOWN ERR_PTR(-EAGAIN)
There should not be further conflict in this part of the code until
things get merged with Linus's tree. Do what suits best.
Thanks.
--
Ueimor
^ permalink raw reply
* Re: [PULL net (try 2)] vhost-net: zercopy mode fixes
From: David Miller @ 2011-07-22 6:37 UTC (permalink / raw)
To: mst; +Cc: kvm, virtualization, netdev, linux-kernel, mashirle, xma
In-Reply-To: <20110722063238.GB9637@redhat.com>
From: "Michael S. Tsirkin" <mst@redhat.com>
Date: Fri, 22 Jul 2011 09:32:38 +0300
> Fixing a corrupted pull request sent earlier.
> Sorry about the noise!
>
> The following includes vhost-net fixes - both in the
> experimental zero copy mode.
> Please pull for 3.1.
Pulled, thanks!
^ permalink raw reply
* [PULL net (try 2)] vhost-net: zercopy mode fixes
From: Michael S. Tsirkin @ 2011-07-22 6:32 UTC (permalink / raw)
To: David Miller; +Cc: kvm, virtualization, netdev, linux-kernel, mashirle, xma
In-Reply-To: <20110722060046.GA8876@redhat.com>
Fixing a corrupted pull request sent earlier.
Sorry about the noise!
The following includes vhost-net fixes - both in the
experimental zero copy mode.
Please pull for 3.1.
Thanks!
The following changes since commit fd99beb9375b44309d5ed9c2b62063e23151610b:
Merge branch 'vhost-net-next' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost (2011-07-19 10:12:44 -0700)
are available in the git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost.git vhost-net-next
Michael S. Tsirkin (1):
vhost-net: update used ring on backend change
Shirley Ma (1):
vhost: handle wrap around in # of bufs math
drivers/vhost/net.c | 18 ++++++++++++++----
1 files changed, 14 insertions(+), 4 deletions(-)
^ permalink raw reply
* Re: linux-next: build warning after merge of the final tree
From: David Miller @ 2011-07-22 6:26 UTC (permalink / raw)
To: sfr; +Cc: netdev, linux-next, linux-kernel, matt
In-Reply-To: <20110722162433.a280b3b71cff9c08fe18956f@canb.auug.org.au>
From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Fri, 22 Jul 2011 16:24:33 +1000
> warning: (PPC) selects HAVE_BPF_JIT which has unmet direct dependencies (NET)
>
> Introduced by commit 0ca87f05ba8b ("net: filter: BPF 'JIT' compiler for
> PPC64").
>
> I suspect that the fix is to move the
>
> config HAVE_BPF_JIT
> bool
>
> lines to outside the "if NET" in net/Kconfig.
I'm going to fix this the way it's dealt with in the x86_64 case
for now, thanks for the report:
--------------------
net: Fix ppc64 BPF JIT dependencies.
Signed-off-by: David S. Miller <davem@davemloft.net>
---
arch/powerpc/Kconfig | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
index 39860fc..cdf7a0a 100644
--- a/arch/powerpc/Kconfig
+++ b/arch/powerpc/Kconfig
@@ -134,7 +134,7 @@ config PPC
select GENERIC_IRQ_SHOW_LEVEL
select HAVE_RCU_TABLE_FREE if SMP
select HAVE_SYSCALL_TRACEPOINTS
- select HAVE_BPF_JIT if PPC64
+ select HAVE_BPF_JIT if (PPC64 && NET)
config EARLY_PRINTK
bool
--
1.7.6
^ permalink raw reply related
* Re: [PATCH] acenic: include NET_SKB_PAD headroom to incoming skbs
From: David Miller @ 2011-07-22 6:25 UTC (permalink / raw)
To: eric.dumazet; +Cc: netdev
In-Reply-To: <1311312533.3886.2.camel@edumazet-laptop>
From: Eric Dumazet <eric.dumazet@gmail.com>
Date: Fri, 22 Jul 2011 07:28:53 +0200
> Some workloads need some headroom (NET_SKB_PAD) to avoid expensive
> reallocations.
>
> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Applied, thanks Eric.
^ permalink raw reply
* linux-next: build warning after merge of the final tree
From: Stephen Rothwell @ 2011-07-22 6:24 UTC (permalink / raw)
To: David Miller, netdev; +Cc: linux-next, linux-kernel, Matt Evans
[-- Attachment #1: Type: text/plain, Size: 517 bytes --]
Hi all,
After merging the final tree, today's linux-next build (powerpc64
allnoconfig) produced this warning:
warning: (PPC) selects HAVE_BPF_JIT which has unmet direct dependencies (NET)
Introduced by commit 0ca87f05ba8b ("net: filter: BPF 'JIT' compiler for
PPC64").
I suspect that the fix is to move the
config HAVE_BPF_JIT
bool
lines to outside the "if NET" in net/Kconfig.
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/
[-- Attachment #2: Type: application/pgp-signature, Size: 490 bytes --]
^ permalink raw reply
* Re: [PULL net] vhost-net: zercopy mode fixes
From: David Miller @ 2011-07-22 6:24 UTC (permalink / raw)
To: mst; +Cc: kvm, virtualization, netdev, linux-kernel, mashirle, xma
In-Reply-To: <20110722060046.GA8876@redhat.com>
From: "Michael S. Tsirkin" <mst@redhat.com>
Date: Fri, 22 Jul 2011 09:00:46 +0300
> The following includes vhost-net fixes - both in the
> experimental zero copy mode.
> Please pull for 3.1.
> Thanks!
>
Where is this "the following"? I don't see any GIT url to pull
from or anything :-)
^ permalink raw reply
* Re: [net-next 00/10 v2][pull request] Intel Wired LAN Driver Update
From: David Miller @ 2011-07-22 6:23 UTC (permalink / raw)
To: jeffrey.t.kirsher; +Cc: netdev, gospo, sassmann
In-Reply-To: <1311314951-11142-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Thu, 21 Jul 2011 23:09:01 -0700
> The following series contains updates to e1000, igb and ixgbe.
>
> v2: fix my fat finger mess up while reviewing a patch.
>
> The following are changes since commit 61463a30f65225e19e68f59dbd7b888bb308ec99:
> can: make function can_get_bittiming static
> and are available in the git repository at:
> master.kernel.org:/pub/scm/linux/kernel/git/jkirsher/net-next-2.6 master
Looks a lot better, pulled, thanks!
^ permalink raw reply
* RE: Bridging behavior apparently changed around the Fedora 14 time
From: Greg Scott @ 2011-07-22 6:20 UTC (permalink / raw)
To: netdev; +Cc: Lynn Hanson, Joe Whalen, Graham Parenteau, David Lamparter
In-Reply-To: <925A849792280C4E80C5461017A4B8A2A04149@mail733.InfraSupportEtc.com>
Hang on a second - David, I think you're trying to tell me I'm the
problem. Maybe so...
I send a frame in on eth1/br0. It traverses a bunch of ebtables and
iptables DNAT and SNAT rules and then wants to come back out on
eth1/br0. But this is both a bridge and a router, and trying to
remember my bridge basics - as I recall, bridges might not want to
forward frames back out the same physical interface they came in on.
After running this frame/packet through a gauntlet of ebtables/iptables
rules and NATing the snot out of the IP packet inside the frame, I
wonder if it's still the same layer 2 frame when it gets ready to come
back out on eth1? Even though NAT changed the layer 3 routed packet IP
info, I wonder if it's still the same layer 2 datalink frame? So the
bridge says, "No way Jose, that's the same physical interface you came
in on" and drops it.
And then turning on promisc mode makes the bridge less picky and it
happily forwards all frames everywhere.
Is this what you've been trying to tell me and I'm too thickheaded to
get it?
If so, to work around the problem, I was reading about some ebtables
broute ACCEPT/DROP targets that are apparently named badly - DROP
apparently tells the bridge to route the packet instead of bridging the
frame, while ACCEPT tells the bridge to bridge it. Should I be looking
in this direction?
But thinking about this some more - there are cases, such as cascaded
switches, or hubs behind a switch, where you want a bridge to forward
frames back out the same interface. Well . . . maybe. There's still
something about forwarding out the same interface a frame comes in on
that's bugging me.
- Greg
^ permalink raw reply
* [net-next 10/10] ixgbe: convert to ndo_fix_features
From: Jeff Kirsher @ 2011-07-22 6:09 UTC (permalink / raw)
To: davem; +Cc: Don Skidmore, netdev, gospo, sassmann, Michal Miroslaw,
Jeff Kirsher
In-Reply-To: <1311314951-11142-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Don Skidmore <donald.c.skidmore@intel.com>
Private rx_csum flags are now duplicate of netdev->features &
NETIF_F_RXCSUM. We remove those duplicates and now use the net_device_ops
ndo_set_features. This was based on the original patch submitted by
Michal Miroslaw <mirq-linux@rere.qmqm.pl>. I also removed the special
case not requiring a reset for X540 hardware. It is needed just as it is
in 82599 hardware.
Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com>
Cc: Michal Miroslaw <mirq-linux@rere.qmqm.pl>
Tested-by: Phil Schmitt <phillip.j.schmitt@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe.h | 5 +-
drivers/net/ixgbe/ixgbe_ethtool.c | 188 -------------------------------------
drivers/net/ixgbe/ixgbe_main.c | 135 +++++++++++++++++++++-----
3 files changed, 111 insertions(+), 217 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe.h b/drivers/net/ixgbe/ixgbe.h
index 0d610c7..e04a8e4 100644
--- a/drivers/net/ixgbe/ixgbe.h
+++ b/drivers/net/ixgbe/ixgbe.h
@@ -587,13 +587,10 @@ extern s32 ixgbe_fdir_erase_perfect_filter_82599(struct ixgbe_hw *hw,
u16 soft_id);
extern void ixgbe_atr_compute_perfect_hash_82599(union ixgbe_atr_input *input,
union ixgbe_atr_input *mask);
-extern void ixgbe_configure_rscctl(struct ixgbe_adapter *adapter,
- struct ixgbe_ring *ring);
-extern void ixgbe_clear_rscctl(struct ixgbe_adapter *adapter,
- struct ixgbe_ring *ring);
extern void ixgbe_set_rx_mode(struct net_device *netdev);
extern int ixgbe_setup_tc(struct net_device *dev, u8 tc);
extern void ixgbe_tx_ctxtdesc(struct ixgbe_ring *, u32, u32, u32, u32);
+extern void ixgbe_do_reset(struct net_device *netdev);
#ifdef IXGBE_FCOE
extern void ixgbe_configure_fcoe(struct ixgbe_adapter *adapter);
extern int ixgbe_fso(struct ixgbe_ring *tx_ring, struct sk_buff *skb,
diff --git a/drivers/net/ixgbe/ixgbe_ethtool.c b/drivers/net/ixgbe/ixgbe_ethtool.c
index 6035cfa..dc64955 100644
--- a/drivers/net/ixgbe/ixgbe_ethtool.c
+++ b/drivers/net/ixgbe/ixgbe_ethtool.c
@@ -442,109 +442,6 @@ static int ixgbe_set_pauseparam(struct net_device *netdev,
return 0;
}
-static void ixgbe_do_reset(struct net_device *netdev)
-{
- struct ixgbe_adapter *adapter = netdev_priv(netdev);
-
- if (netif_running(netdev))
- ixgbe_reinit_locked(adapter);
- else
- ixgbe_reset(adapter);
-}
-
-static u32 ixgbe_get_rx_csum(struct net_device *netdev)
-{
- struct ixgbe_adapter *adapter = netdev_priv(netdev);
- return adapter->flags & IXGBE_FLAG_RX_CSUM_ENABLED;
-}
-
-static void ixgbe_set_rsc(struct ixgbe_adapter *adapter)
-{
- int i;
-
- for (i = 0; i < adapter->num_rx_queues; i++) {
- struct ixgbe_ring *ring = adapter->rx_ring[i];
- if (adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED) {
- set_ring_rsc_enabled(ring);
- ixgbe_configure_rscctl(adapter, ring);
- } else {
- ixgbe_clear_rscctl(adapter, ring);
- }
- }
-}
-
-static int ixgbe_set_rx_csum(struct net_device *netdev, u32 data)
-{
- struct ixgbe_adapter *adapter = netdev_priv(netdev);
- bool need_reset = false;
-
- if (data) {
- adapter->flags |= IXGBE_FLAG_RX_CSUM_ENABLED;
- } else {
- adapter->flags &= ~IXGBE_FLAG_RX_CSUM_ENABLED;
-
- if (adapter->flags2 & IXGBE_FLAG2_RSC_CAPABLE) {
- adapter->flags2 &= ~IXGBE_FLAG2_RSC_ENABLED;
- netdev->features &= ~NETIF_F_LRO;
- }
-
- switch (adapter->hw.mac.type) {
- case ixgbe_mac_X540:
- ixgbe_set_rsc(adapter);
- break;
- case ixgbe_mac_82599EB:
- need_reset = true;
- break;
- default:
- break;
- }
- }
-
- if (need_reset)
- ixgbe_do_reset(netdev);
-
- return 0;
-}
-
-static u32 ixgbe_get_tx_csum(struct net_device *netdev)
-{
- return (netdev->features & NETIF_F_IP_CSUM) != 0;
-}
-
-static int ixgbe_set_tx_csum(struct net_device *netdev, u32 data)
-{
- struct ixgbe_adapter *adapter = netdev_priv(netdev);
- u32 feature_list;
-
- feature_list = (NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM);
- switch (adapter->hw.mac.type) {
- case ixgbe_mac_82599EB:
- case ixgbe_mac_X540:
- feature_list |= NETIF_F_SCTP_CSUM;
- break;
- default:
- break;
- }
- if (data)
- netdev->features |= feature_list;
- else
- netdev->features &= ~feature_list;
-
- return 0;
-}
-
-static int ixgbe_set_tso(struct net_device *netdev, u32 data)
-{
- if (data) {
- netdev->features |= NETIF_F_TSO;
- netdev->features |= NETIF_F_TSO6;
- } else {
- netdev->features &= ~NETIF_F_TSO;
- netdev->features &= ~NETIF_F_TSO6;
- }
- return 0;
-}
-
static u32 ixgbe_get_msglevel(struct net_device *netdev)
{
struct ixgbe_adapter *adapter = netdev_priv(netdev);
@@ -2289,81 +2186,6 @@ static int ixgbe_set_coalesce(struct net_device *netdev,
return 0;
}
-static int ixgbe_set_flags(struct net_device *netdev, u32 data)
-{
- struct ixgbe_adapter *adapter = netdev_priv(netdev);
- bool need_reset = false;
- int rc;
-
-#ifdef CONFIG_IXGBE_DCB
- if ((adapter->flags & IXGBE_FLAG_DCB_ENABLED) &&
- !(data & ETH_FLAG_RXVLAN))
- return -EINVAL;
-#endif
-
- need_reset = (data & ETH_FLAG_RXVLAN) !=
- (netdev->features & NETIF_F_HW_VLAN_RX);
-
- if ((data & ETH_FLAG_RXHASH) &&
- !(adapter->flags & IXGBE_FLAG_RSS_ENABLED))
- return -EOPNOTSUPP;
-
- rc = ethtool_op_set_flags(netdev, data, ETH_FLAG_LRO | ETH_FLAG_NTUPLE |
- ETH_FLAG_RXVLAN | ETH_FLAG_TXVLAN |
- ETH_FLAG_RXHASH);
- if (rc)
- return rc;
-
- /* if state changes we need to update adapter->flags and reset */
- if ((adapter->flags2 & IXGBE_FLAG2_RSC_CAPABLE) &&
- (!!(data & ETH_FLAG_LRO) !=
- !!(adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED))) {
- if ((data & ETH_FLAG_LRO) &&
- (!adapter->rx_itr_setting ||
- (adapter->rx_itr_setting > IXGBE_MAX_RSC_INT_RATE))) {
- e_info(probe, "rx-usecs set too low, "
- "not enabling RSC.\n");
- } else {
- adapter->flags2 ^= IXGBE_FLAG2_RSC_ENABLED;
- switch (adapter->hw.mac.type) {
- case ixgbe_mac_X540:
- ixgbe_set_rsc(adapter);
- break;
- case ixgbe_mac_82599EB:
- need_reset = true;
- break;
- default:
- break;
- }
- }
- }
-
- /*
- * Check if Flow Director n-tuple support was enabled or disabled. If
- * the state changed, we need to reset.
- */
- if (!(adapter->flags & IXGBE_FLAG_FDIR_PERFECT_CAPABLE)) {
- /* turn off ATR, enable perfect filters and reset */
- if (data & ETH_FLAG_NTUPLE) {
- adapter->flags &= ~IXGBE_FLAG_FDIR_HASH_CAPABLE;
- adapter->flags |= IXGBE_FLAG_FDIR_PERFECT_CAPABLE;
- need_reset = true;
- }
- } else if (!(data & ETH_FLAG_NTUPLE)) {
- /* turn off Flow Director, set ATR and reset */
- adapter->flags &= ~IXGBE_FLAG_FDIR_PERFECT_CAPABLE;
- if ((adapter->flags & IXGBE_FLAG_RSS_ENABLED) &&
- !(adapter->flags & IXGBE_FLAG_DCB_ENABLED))
- adapter->flags |= IXGBE_FLAG_FDIR_HASH_CAPABLE;
- need_reset = true;
- }
-
- if (need_reset)
- ixgbe_do_reset(netdev);
-
- return 0;
-}
-
static int ixgbe_get_ethtool_fdir_entry(struct ixgbe_adapter *adapter,
struct ethtool_rxnfc *cmd)
{
@@ -2746,16 +2568,8 @@ static const struct ethtool_ops ixgbe_ethtool_ops = {
.set_ringparam = ixgbe_set_ringparam,
.get_pauseparam = ixgbe_get_pauseparam,
.set_pauseparam = ixgbe_set_pauseparam,
- .get_rx_csum = ixgbe_get_rx_csum,
- .set_rx_csum = ixgbe_set_rx_csum,
- .get_tx_csum = ixgbe_get_tx_csum,
- .set_tx_csum = ixgbe_set_tx_csum,
- .get_sg = ethtool_op_get_sg,
- .set_sg = ethtool_op_set_sg,
.get_msglevel = ixgbe_get_msglevel,
.set_msglevel = ixgbe_set_msglevel,
- .get_tso = ethtool_op_get_tso,
- .set_tso = ixgbe_set_tso,
.self_test = ixgbe_diag_test,
.get_strings = ixgbe_get_strings,
.set_phys_id = ixgbe_set_phys_id,
@@ -2763,8 +2577,6 @@ static const struct ethtool_ops ixgbe_ethtool_ops = {
.get_ethtool_stats = ixgbe_get_ethtool_stats,
.get_coalesce = ixgbe_get_coalesce,
.set_coalesce = ixgbe_set_coalesce,
- .get_flags = ethtool_op_get_flags,
- .set_flags = ixgbe_set_flags,
.get_rxnfc = ixgbe_get_rxnfc,
.set_rxnfc = ixgbe_set_rxnfc,
};
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index c6fa9c7..1be6175 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -2862,28 +2862,11 @@ static void ixgbe_setup_mrqc(struct ixgbe_adapter *adapter)
}
/**
- * ixgbe_clear_rscctl - disable RSC for the indicated ring
- * @adapter: address of board private structure
- * @ring: structure containing ring specific data
- **/
-void ixgbe_clear_rscctl(struct ixgbe_adapter *adapter,
- struct ixgbe_ring *ring)
-{
- struct ixgbe_hw *hw = &adapter->hw;
- u32 rscctrl;
- u8 reg_idx = ring->reg_idx;
-
- rscctrl = IXGBE_READ_REG(hw, IXGBE_RSCCTL(reg_idx));
- rscctrl &= ~IXGBE_RSCCTL_RSCEN;
- IXGBE_WRITE_REG(hw, IXGBE_RSCCTL(reg_idx), rscctrl);
-}
-
-/**
* ixgbe_configure_rscctl - enable RSC for the indicated ring
* @adapter: address of board private structure
* @index: index of ring to set
**/
-void ixgbe_configure_rscctl(struct ixgbe_adapter *adapter,
+static void ixgbe_configure_rscctl(struct ixgbe_adapter *adapter,
struct ixgbe_ring *ring)
{
struct ixgbe_hw *hw = &adapter->hw;
@@ -7122,6 +7105,98 @@ int ixgbe_setup_tc(struct net_device *dev, u8 tc)
return 0;
}
+void ixgbe_do_reset(struct net_device *netdev)
+{
+ struct ixgbe_adapter *adapter = netdev_priv(netdev);
+
+ if (netif_running(netdev))
+ ixgbe_reinit_locked(adapter);
+ else
+ ixgbe_reset(adapter);
+}
+
+static u32 ixgbe_fix_features(struct net_device *netdev, u32 data)
+{
+ struct ixgbe_adapter *adapter = netdev_priv(netdev);
+
+#ifdef CONFIG_DCB
+ if (adapter->flags & IXGBE_FLAG_DCB_ENABLED)
+ data &= ~NETIF_F_HW_VLAN_RX;
+#endif
+
+ /* return error if RXHASH is being enabled when RSS is not supported */
+ if (!(adapter->flags & IXGBE_FLAG_RSS_ENABLED))
+ data &= ~NETIF_F_RXHASH;
+
+ /* If Rx checksum is disabled, then RSC/LRO should also be disabled */
+ if (!(data & NETIF_F_RXCSUM))
+ data &= ~NETIF_F_LRO;
+
+ /* Turn off LRO if not RSC capable or invalid ITR settings */
+ if (!(adapter->flags2 & IXGBE_FLAG2_RSC_CAPABLE)) {
+ data &= ~NETIF_F_LRO;
+ } else if (!(adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED) &&
+ (adapter->rx_itr_setting != 1 &&
+ adapter->rx_itr_setting > IXGBE_MAX_RSC_INT_RATE)) {
+ data &= ~NETIF_F_LRO;
+ e_info(probe, "rx-usecs set too low, not enabling RSC\n");
+ }
+
+ return data;
+}
+
+static int ixgbe_set_features(struct net_device *netdev, u32 data)
+{
+ struct ixgbe_adapter *adapter = netdev_priv(netdev);
+ bool need_reset = false;
+
+ /* If Rx checksum is disabled, then RSC/LRO should also be disabled */
+ if (!(data & NETIF_F_RXCSUM))
+ adapter->flags &= ~IXGBE_FLAG_RX_CSUM_ENABLED;
+ else
+ adapter->flags |= IXGBE_FLAG_RX_CSUM_ENABLED;
+
+ /* Make sure RSC matches LRO, reset if change */
+ if (!!(data & NETIF_F_LRO) !=
+ !!(adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED)) {
+ adapter->flags2 ^= IXGBE_FLAG2_RSC_ENABLED;
+ switch (adapter->hw.mac.type) {
+ case ixgbe_mac_X540:
+ case ixgbe_mac_82599EB:
+ need_reset = true;
+ break;
+ default:
+ break;
+ }
+ }
+
+ /*
+ * Check if Flow Director n-tuple support was enabled or disabled. If
+ * the state changed, we need to reset.
+ */
+ if (!(adapter->flags & IXGBE_FLAG_FDIR_PERFECT_CAPABLE)) {
+ /* turn off ATR, enable perfect filters and reset */
+ if (data & NETIF_F_NTUPLE) {
+ adapter->flags &= ~IXGBE_FLAG_FDIR_HASH_CAPABLE;
+ adapter->flags |= IXGBE_FLAG_FDIR_PERFECT_CAPABLE;
+ need_reset = true;
+ }
+ } else if (!(data & NETIF_F_NTUPLE)) {
+ /* turn off Flow Director, set ATR and reset */
+ adapter->flags &= ~IXGBE_FLAG_FDIR_PERFECT_CAPABLE;
+ if ((adapter->flags & IXGBE_FLAG_RSS_ENABLED) &&
+ !(adapter->flags & IXGBE_FLAG_DCB_ENABLED))
+ adapter->flags |= IXGBE_FLAG_FDIR_HASH_CAPABLE;
+ need_reset = true;
+ }
+
+ if (need_reset)
+ ixgbe_do_reset(netdev);
+
+ return 0;
+
+}
+
static const struct net_device_ops ixgbe_netdev_ops = {
.ndo_open = ixgbe_open,
.ndo_stop = ixgbe_close,
@@ -7153,6 +7228,8 @@ static const struct net_device_ops ixgbe_netdev_ops = {
.ndo_fcoe_disable = ixgbe_fcoe_disable,
.ndo_fcoe_get_wwn = ixgbe_fcoe_get_wwn,
#endif /* IXGBE_FCOE */
+ .ndo_set_features = ixgbe_set_features,
+ .ndo_fix_features = ixgbe_fix_features,
};
static void __devinit ixgbe_probe_vf(struct ixgbe_adapter *adapter,
@@ -7420,20 +7497,24 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev,
netdev->features = NETIF_F_SG |
NETIF_F_IP_CSUM |
+ NETIF_F_IPV6_CSUM |
NETIF_F_HW_VLAN_TX |
NETIF_F_HW_VLAN_RX |
- NETIF_F_HW_VLAN_FILTER;
+ NETIF_F_HW_VLAN_FILTER |
+ NETIF_F_TSO |
+ NETIF_F_TSO6 |
+ NETIF_F_GRO |
+ NETIF_F_RXHASH |
+ NETIF_F_RXCSUM;
- netdev->features |= NETIF_F_IPV6_CSUM;
- netdev->features |= NETIF_F_TSO;
- netdev->features |= NETIF_F_TSO6;
- netdev->features |= NETIF_F_GRO;
- netdev->features |= NETIF_F_RXHASH;
+ netdev->hw_features = netdev->features;
switch (adapter->hw.mac.type) {
case ixgbe_mac_82599EB:
case ixgbe_mac_X540:
netdev->features |= NETIF_F_SCTP_CSUM;
+ netdev->hw_features |= NETIF_F_SCTP_CSUM |
+ NETIF_F_NTUPLE;
break;
default:
break;
@@ -7472,6 +7553,8 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev,
netdev->vlan_features |= NETIF_F_HIGHDMA;
}
+ if (adapter->flags2 & IXGBE_FLAG2_RSC_CAPABLE)
+ netdev->hw_features |= NETIF_F_LRO;
if (adapter->flags2 & IXGBE_FLAG2_RSC_ENABLED)
netdev->features |= NETIF_F_LRO;
@@ -7508,8 +7591,10 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev,
if (err)
goto err_sw_init;
- if (!(adapter->flags & IXGBE_FLAG_RSS_ENABLED))
+ if (!(adapter->flags & IXGBE_FLAG_RSS_ENABLED)) {
+ netdev->hw_features &= ~NETIF_F_RXHASH;
netdev->features &= ~NETIF_F_RXHASH;
+ }
switch (pdev->device) {
case IXGBE_DEV_ID_82599_SFP:
--
1.7.6
^ permalink raw reply related
* [net-next 09/10] ixgbe: only enable WoL for magic packet by default
From: Jeff Kirsher @ 2011-07-22 6:09 UTC (permalink / raw)
To: davem; +Cc: Andy Gospodarek, netdev, gospo, sassmann, Martin Wilck,
Jeff Kirsher
In-Reply-To: <1311314951-11142-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Andy Gospodarek <andy@greyhouse.net>
Martin Wilck <martin.wilck@ts.fujitsu.com> reported that systems using
the ixgbe-driver that were capable of WoL were rebooting almost as soon
as they were shut down. This is because the default WoL settings
enabled magic packet, broadcast, unicast, and multicast.
Other Intel devices seem to use the stored eeprom value for initial WoL
capabilities. The 82578DM (e1000e) and 82576 (igb) the devices I looked
at had only the magic packet enabled in the eeprom, so that seems
appropriate on ixgbe-based devices as well. I set the WoL options on my
82578DM to be the same default as the ixgbe devices (umbg) and saw the
same as Martin -- almost as soon as my box shutdown, it booted again.
This patch changes the default to only be the magic packet. This is the
same as the default for most Intel and non-Intel hardware currently
upstream.
Signed-off-by: Andy Gospodarek <andy@greyhouse.net>
CC: Martin Wilck <martin.wilck@ts.fujitsu.com>
Tested-by: Phil Schmitt <phillip.j.schmitt@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_main.c | 9 +++------
1 files changed, 3 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index 358f6c9..c6fa9c7 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -7515,18 +7515,15 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev,
case IXGBE_DEV_ID_82599_SFP:
/* Only this subdevice supports WOL */
if (pdev->subsystem_device == IXGBE_SUBDEV_ID_82599_SFP)
- adapter->wol = (IXGBE_WUFC_MAG | IXGBE_WUFC_EX |
- IXGBE_WUFC_MC | IXGBE_WUFC_BC);
+ adapter->wol = IXGBE_WUFC_MAG;
break;
case IXGBE_DEV_ID_82599_COMBO_BACKPLANE:
/* All except this subdevice support WOL */
if (pdev->subsystem_device != IXGBE_SUBDEV_ID_82599_KX4_KR_MEZZ)
- adapter->wol = (IXGBE_WUFC_MAG | IXGBE_WUFC_EX |
- IXGBE_WUFC_MC | IXGBE_WUFC_BC);
+ adapter->wol = IXGBE_WUFC_MAG;
break;
case IXGBE_DEV_ID_82599_KX4:
- adapter->wol = (IXGBE_WUFC_MAG | IXGBE_WUFC_EX |
- IXGBE_WUFC_MC | IXGBE_WUFC_BC);
+ adapter->wol = IXGBE_WUFC_MAG;
break;
default:
adapter->wol = 0;
--
1.7.6
^ permalink raw reply related
* [net-next 08/10] ixgbe: remove ifdef check for non-existent define
From: Jeff Kirsher @ 2011-07-22 6:09 UTC (permalink / raw)
To: davem; +Cc: Emil Tantilov, netdev, gospo, sassmann, Jeff Kirsher
In-Reply-To: <1311314951-11142-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Emil Tantilov <emil.s.tantilov@intel.com>
Signed-off-by: Emil Tantilov <emil.s.tantilov@intel.com>
Tested-by: Phil Schmitt <phillip.j.schmitt@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_main.c | 2 --
1 files changed, 0 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index a222af3..358f6c9 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -6021,9 +6021,7 @@ static void ixgbe_watchdog_link_is_up(struct ixgbe_adapter *adapter)
(flow_tx ? "TX" : "None"))));
netif_carrier_on(netdev);
-#ifdef HAVE_IPLINK_VF_CONFIG
ixgbe_check_vf_rate_limit(adapter);
-#endif /* HAVE_IPLINK_VF_CONFIG */
}
/**
--
1.7.6
^ permalink raw reply related
* [net-next 06/10] ixgbe: Move interrupt related values out of ring and into q_vector
From: Jeff Kirsher @ 2011-07-22 6:09 UTC (permalink / raw)
To: davem; +Cc: Alexander Duyck, netdev, gospo, sassmann, Jeff Kirsher
In-Reply-To: <1311314951-11142-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Alexander Duyck <alexander.h.duyck@intel.com>
This change moves work_limit, total_packets, and total_bytes into the ring
container struct of the q_vector. The advantage of this is that it should
reduce the size of memory used in the event of multiple rings being
assigned to a single q_vector. In addition it should help to reduce the
total workload for calculating itr since now total_packets and total_bytes
will be the total work done of the interrupt instead of for the ring.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Ross Brattain <ross.b.brattain@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe.h | 24 +++---
drivers/net/ixgbe/ixgbe_ethtool.c | 6 +-
drivers/net/ixgbe/ixgbe_main.c | 170 ++++++++++---------------------------
3 files changed, 62 insertions(+), 138 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe.h b/drivers/net/ixgbe/ixgbe.h
index a5ed9fd..fbae735 100644
--- a/drivers/net/ixgbe/ixgbe.h
+++ b/drivers/net/ixgbe/ixgbe.h
@@ -214,12 +214,10 @@ struct ixgbe_ring {
struct ixgbe_rx_buffer *rx_buffer_info;
};
unsigned long state;
- u8 atr_sample_rate;
- u8 atr_count;
+ u8 __iomem *tail;
+
u16 count; /* amount of descriptors */
u16 rx_buf_len;
- u16 next_to_use;
- u16 next_to_clean;
u8 queue_index; /* needed for multiqueue queue management */
u8 reg_idx; /* holds the special value that gets
@@ -227,15 +225,13 @@ struct ixgbe_ring {
* associated with this ring, which is
* different for DCB and RSS modes
*/
- u8 dcb_tc;
-
- u16 work_limit; /* max work per interrupt */
-
- u8 __iomem *tail;
+ u8 atr_sample_rate;
+ u8 atr_count;
- unsigned int total_bytes;
- unsigned int total_packets;
+ u16 next_to_use;
+ u16 next_to_clean;
+ u8 dcb_tc;
struct ixgbe_queue_stats stats;
struct u64_stats_sync syncp;
union {
@@ -283,6 +279,9 @@ struct ixgbe_ring_container {
#else
DECLARE_BITMAP(idx, MAX_TX_QUEUES);
#endif
+ unsigned int total_bytes; /* total bytes processed this int */
+ unsigned int total_packets; /* total packets processed this int */
+ u16 work_limit; /* total work allowed per interrupt */
u8 count; /* total number of rings in vector */
u8 itr; /* current ITR setting for ring */
};
@@ -417,6 +416,9 @@ struct ixgbe_adapter {
u16 eitr_low;
u16 eitr_high;
+ /* Work limits */
+ u16 tx_work_limit;
+
/* TX */
struct ixgbe_ring *tx_ring[MAX_TX_QUEUES] ____cacheline_aligned_in_smp;
int num_tx_queues;
diff --git a/drivers/net/ixgbe/ixgbe_ethtool.c b/drivers/net/ixgbe/ixgbe_ethtool.c
index cd842f9..6035cfa 100644
--- a/drivers/net/ixgbe/ixgbe_ethtool.c
+++ b/drivers/net/ixgbe/ixgbe_ethtool.c
@@ -2103,7 +2103,7 @@ static int ixgbe_get_coalesce(struct net_device *netdev,
{
struct ixgbe_adapter *adapter = netdev_priv(netdev);
- ec->tx_max_coalesced_frames_irq = adapter->tx_ring[0]->work_limit;
+ ec->tx_max_coalesced_frames_irq = adapter->tx_work_limit;
/* only valid if in constant ITR mode */
switch (adapter->rx_itr_setting) {
@@ -2192,7 +2192,7 @@ static int ixgbe_set_coalesce(struct net_device *netdev,
return -EINVAL;
if (ec->tx_max_coalesced_frames_irq)
- adapter->tx_ring[0]->work_limit = ec->tx_max_coalesced_frames_irq;
+ adapter->tx_work_limit = ec->tx_max_coalesced_frames_irq;
if (ec->rx_coalesce_usecs > 1) {
/* check the limits */
@@ -2267,12 +2267,14 @@ static int ixgbe_set_coalesce(struct net_device *netdev,
else
/* rx only or mixed */
q_vector->eitr = adapter->rx_eitr_param;
+ q_vector->tx.work_limit = adapter->tx_work_limit;
ixgbe_write_eitr(q_vector);
}
/* Legacy Interrupt Mode */
} else {
q_vector = adapter->q_vector[0];
q_vector->eitr = adapter->rx_eitr_param;
+ q_vector->tx.work_limit = adapter->tx_work_limit;
ixgbe_write_eitr(q_vector);
}
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index bd1fd8f..298c95b 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -805,7 +805,7 @@ static bool ixgbe_clean_tx_irq(struct ixgbe_q_vector *q_vector,
eop_desc = IXGBE_TX_DESC_ADV(tx_ring, eop);
while ((eop_desc->wb.status & cpu_to_le32(IXGBE_TXD_STAT_DD)) &&
- (count < tx_ring->work_limit)) {
+ (count < q_vector->tx.work_limit)) {
bool cleaned = false;
rmb(); /* read buffer_info after eop_desc */
for ( ; !cleaned; count++) {
@@ -834,11 +834,11 @@ static bool ixgbe_clean_tx_irq(struct ixgbe_q_vector *q_vector,
}
tx_ring->next_to_clean = i;
- tx_ring->total_bytes += total_bytes;
- tx_ring->total_packets += total_packets;
- u64_stats_update_begin(&tx_ring->syncp);
- tx_ring->stats.packets += total_packets;
tx_ring->stats.bytes += total_bytes;
+ tx_ring->stats.packets += total_packets;
+ u64_stats_update_begin(&tx_ring->syncp);
+ q_vector->tx.total_bytes += total_bytes;
+ q_vector->tx.total_packets += total_packets;
u64_stats_update_end(&tx_ring->syncp);
if (check_for_tx_hang(tx_ring) && ixgbe_check_tx_hang(tx_ring)) {
@@ -886,7 +886,7 @@ static bool ixgbe_clean_tx_irq(struct ixgbe_q_vector *q_vector,
}
}
- return count < tx_ring->work_limit;
+ return count < q_vector->tx.work_limit;
}
#ifdef CONFIG_IXGBE_DCA
@@ -1486,12 +1486,12 @@ next_desc:
}
#endif /* IXGBE_FCOE */
- rx_ring->total_packets += total_rx_packets;
- rx_ring->total_bytes += total_rx_bytes;
u64_stats_update_begin(&rx_ring->syncp);
rx_ring->stats.packets += total_rx_packets;
rx_ring->stats.bytes += total_rx_bytes;
u64_stats_update_end(&rx_ring->syncp);
+ q_vector->rx.total_packets += total_rx_packets;
+ q_vector->rx.total_bytes += total_rx_bytes;
}
static int ixgbe_clean_rxonly(struct napi_struct *, int);
@@ -1597,11 +1597,8 @@ enum latency_range {
/**
* ixgbe_update_itr - update the dynamic ITR value based on statistics
- * @adapter: pointer to adapter
- * @eitr: eitr setting (ints per sec) to give last timeslice
- * @itr_setting: current throttle rate in ints/second
- * @packets: the number of packets during this measurement interval
- * @bytes: the number of bytes during this measurement interval
+ * @q_vector: structure containing interrupt and ring information
+ * @ring_container: structure containing ring performance data
*
* Stores a new ITR value based on packets and byte
* counts during the last interrupt. The advantage of per interrupt
@@ -1613,17 +1610,18 @@ enum latency_range {
* this functionality is controlled by the InterruptThrottleRate module
* parameter (see ixgbe_param.c)
**/
-static u8 ixgbe_update_itr(struct ixgbe_adapter *adapter,
- u32 eitr, u8 itr_setting,
- int packets, int bytes)
+static void ixgbe_update_itr(struct ixgbe_q_vector *q_vector,
+ struct ixgbe_ring_container *ring_container)
{
- unsigned int retval = itr_setting;
- u32 timepassed_us;
u64 bytes_perint;
+ struct ixgbe_adapter *adapter = q_vector->adapter;
+ int bytes = ring_container->total_bytes;
+ int packets = ring_container->total_packets;
+ u32 timepassed_us;
+ u8 itr_setting = ring_container->itr;
if (packets == 0)
- goto update_itr_done;
-
+ return;
/* simple throttlerate management
* 0-20MB/s lowest (100000 ints/s)
@@ -1631,28 +1629,32 @@ static u8 ixgbe_update_itr(struct ixgbe_adapter *adapter,
* 100-1249MB/s bulk (8000 ints/s)
*/
/* what was last interrupt timeslice? */
- timepassed_us = 1000000/eitr;
+ timepassed_us = 1000000/q_vector->eitr;
bytes_perint = bytes / timepassed_us; /* bytes/usec */
switch (itr_setting) {
case lowest_latency:
if (bytes_perint > adapter->eitr_low)
- retval = low_latency;
+ itr_setting = low_latency;
break;
case low_latency:
if (bytes_perint > adapter->eitr_high)
- retval = bulk_latency;
+ itr_setting = bulk_latency;
else if (bytes_perint <= adapter->eitr_low)
- retval = lowest_latency;
+ itr_setting = lowest_latency;
break;
case bulk_latency:
if (bytes_perint <= adapter->eitr_high)
- retval = low_latency;
+ itr_setting = low_latency;
break;
}
-update_itr_done:
- return retval;
+ /* clear work counters since we have the values we need */
+ ring_container->total_bytes = 0;
+ ring_container->total_packets = 0;
+
+ /* write updated itr to ring container */
+ ring_container->itr = itr_setting;
}
/**
@@ -1698,42 +1700,13 @@ void ixgbe_write_eitr(struct ixgbe_q_vector *q_vector)
IXGBE_WRITE_REG(hw, IXGBE_EITR(v_idx), itr_reg);
}
-static void ixgbe_set_itr_msix(struct ixgbe_q_vector *q_vector)
+static void ixgbe_set_itr(struct ixgbe_q_vector *q_vector)
{
- struct ixgbe_adapter *adapter = q_vector->adapter;
- int i, r_idx;
- u32 new_itr;
- u8 current_itr, ret_itr;
-
- r_idx = find_first_bit(q_vector->tx.idx, adapter->num_tx_queues);
- for (i = 0; i < q_vector->tx.count; i++) {
- struct ixgbe_ring *tx_ring = adapter->tx_ring[r_idx];
- ret_itr = ixgbe_update_itr(adapter, q_vector->eitr,
- q_vector->tx.itr,
- tx_ring->total_packets,
- tx_ring->total_bytes);
- /* if the result for this queue would decrease interrupt
- * rate for this vector then use that result */
- q_vector->tx.itr = ((q_vector->tx.itr > ret_itr) ?
- q_vector->tx.itr - 1 : ret_itr);
- r_idx = find_next_bit(q_vector->tx.idx, adapter->num_tx_queues,
- r_idx + 1);
- }
+ u32 new_itr = q_vector->eitr;
+ u8 current_itr;
- r_idx = find_first_bit(q_vector->rx.idx, adapter->num_rx_queues);
- for (i = 0; i < q_vector->rx.count; i++) {
- struct ixgbe_ring *rx_ring = adapter->rx_ring[r_idx];
- ret_itr = ixgbe_update_itr(adapter, q_vector->eitr,
- q_vector->rx.itr,
- rx_ring->total_packets,
- rx_ring->total_bytes);
- /* if the result for this queue would decrease interrupt
- * rate for this vector then use that result */
- q_vector->rx.itr = ((q_vector->rx.itr > ret_itr) ?
- q_vector->rx.itr - 1 : ret_itr);
- r_idx = find_next_bit(q_vector->rx.idx, adapter->num_rx_queues,
- r_idx + 1);
- }
+ ixgbe_update_itr(q_vector, &q_vector->tx);
+ ixgbe_update_itr(q_vector, &q_vector->rx);
current_itr = max(q_vector->rx.itr, q_vector->tx.itr);
@@ -1746,16 +1719,17 @@ static void ixgbe_set_itr_msix(struct ixgbe_q_vector *q_vector)
new_itr = 20000; /* aka hwitr = ~200 */
break;
case bulk_latency:
- default:
new_itr = 8000;
break;
+ default:
+ break;
}
if (new_itr != q_vector->eitr) {
/* do an exponential smoothing */
new_itr = ((q_vector->eitr * 9) + new_itr)/10;
- /* save the algorithm value here, not the smoothed one */
+ /* save the algorithm value here */
q_vector->eitr = new_itr;
ixgbe_write_eitr(q_vector);
@@ -2001,8 +1975,6 @@ static irqreturn_t ixgbe_msix_clean_tx(int irq, void *data)
r_idx = find_first_bit(q_vector->tx.idx, adapter->num_tx_queues);
for (i = 0; i < q_vector->tx.count; i++) {
tx_ring = adapter->tx_ring[r_idx];
- tx_ring->total_bytes = 0;
- tx_ring->total_packets = 0;
r_idx = find_next_bit(q_vector->tx.idx, adapter->num_tx_queues,
r_idx + 1);
}
@@ -2034,8 +2006,6 @@ static irqreturn_t ixgbe_msix_clean_rx(int irq, void *data)
r_idx = find_first_bit(q_vector->rx.idx, adapter->num_rx_queues);
for (i = 0; i < q_vector->rx.count; i++) {
rx_ring = adapter->rx_ring[r_idx];
- rx_ring->total_bytes = 0;
- rx_ring->total_packets = 0;
r_idx = find_next_bit(q_vector->rx.idx, adapter->num_rx_queues,
r_idx + 1);
}
@@ -2063,8 +2033,6 @@ static irqreturn_t ixgbe_msix_clean_many(int irq, void *data)
r_idx = find_first_bit(q_vector->tx.idx, adapter->num_tx_queues);
for (i = 0; i < q_vector->tx.count; i++) {
ring = adapter->tx_ring[r_idx];
- ring->total_bytes = 0;
- ring->total_packets = 0;
r_idx = find_next_bit(q_vector->tx.idx, adapter->num_tx_queues,
r_idx + 1);
}
@@ -2072,8 +2040,6 @@ static irqreturn_t ixgbe_msix_clean_many(int irq, void *data)
r_idx = find_first_bit(q_vector->rx.idx, adapter->num_rx_queues);
for (i = 0; i < q_vector->rx.count; i++) {
ring = adapter->rx_ring[r_idx];
- ring->total_bytes = 0;
- ring->total_packets = 0;
r_idx = find_next_bit(q_vector->rx.idx, adapter->num_rx_queues,
r_idx + 1);
}
@@ -2115,7 +2081,7 @@ static int ixgbe_clean_rxonly(struct napi_struct *napi, int budget)
if (work_done < budget) {
napi_complete(napi);
if (adapter->rx_itr_setting & 1)
- ixgbe_set_itr_msix(q_vector);
+ ixgbe_set_itr(q_vector);
if (!test_bit(__IXGBE_DOWN, &adapter->state))
ixgbe_irq_enable_queues(adapter,
((u64)1 << q_vector->v_idx));
@@ -2173,7 +2139,7 @@ static int ixgbe_clean_rxtx_many(struct napi_struct *napi, int budget)
if (work_done < budget) {
napi_complete(napi);
if (adapter->rx_itr_setting & 1)
- ixgbe_set_itr_msix(q_vector);
+ ixgbe_set_itr(q_vector);
if (!test_bit(__IXGBE_DOWN, &adapter->state))
ixgbe_irq_enable_queues(adapter,
((u64)1 << q_vector->v_idx));
@@ -2215,7 +2181,7 @@ static int ixgbe_clean_txonly(struct napi_struct *napi, int budget)
if (work_done < budget) {
napi_complete(napi);
if (adapter->tx_itr_setting & 1)
- ixgbe_set_itr_msix(q_vector);
+ ixgbe_set_itr(q_vector);
if (!test_bit(__IXGBE_DOWN, &adapter->state))
ixgbe_irq_enable_queues(adapter,
((u64)1 << q_vector->v_idx));
@@ -2244,6 +2210,7 @@ static inline void map_vector_to_txq(struct ixgbe_adapter *a, int v_idx,
set_bit(t_idx, q_vector->tx.idx);
q_vector->tx.count++;
tx_ring->q_vector = q_vector;
+ q_vector->tx.work_limit = a->tx_work_limit;
}
/**
@@ -2386,51 +2353,6 @@ free_queue_irqs:
return err;
}
-static void ixgbe_set_itr(struct ixgbe_adapter *adapter)
-{
- struct ixgbe_q_vector *q_vector = adapter->q_vector[0];
- struct ixgbe_ring *rx_ring = adapter->rx_ring[0];
- struct ixgbe_ring *tx_ring = adapter->tx_ring[0];
- u32 new_itr = q_vector->eitr;
- u8 current_itr;
-
- q_vector->tx.itr = ixgbe_update_itr(adapter, new_itr,
- q_vector->tx.itr,
- tx_ring->total_packets,
- tx_ring->total_bytes);
- q_vector->rx.itr = ixgbe_update_itr(adapter, new_itr,
- q_vector->rx.itr,
- rx_ring->total_packets,
- rx_ring->total_bytes);
-
- current_itr = max(q_vector->rx.itr, q_vector->tx.itr);
-
- switch (current_itr) {
- /* counts and packets in update_itr are dependent on these numbers */
- case lowest_latency:
- new_itr = 100000;
- break;
- case low_latency:
- new_itr = 20000; /* aka hwitr = ~200 */
- break;
- case bulk_latency:
- new_itr = 8000;
- break;
- default:
- break;
- }
-
- if (new_itr != q_vector->eitr) {
- /* do an exponential smoothing */
- new_itr = ((q_vector->eitr * 9) + new_itr)/10;
-
- /* save the algorithm value here */
- q_vector->eitr = new_itr;
-
- ixgbe_write_eitr(q_vector);
- }
-}
-
/**
* ixgbe_irq_enable - Enable default interrupt generation settings
* @adapter: board private structure
@@ -2528,10 +2450,6 @@ static irqreturn_t ixgbe_intr(int irq, void *data)
ixgbe_check_fan_failure(adapter, eicr);
if (napi_schedule_prep(&(q_vector->napi))) {
- adapter->tx_ring[0]->total_packets = 0;
- adapter->tx_ring[0]->total_bytes = 0;
- adapter->rx_ring[0]->total_packets = 0;
- adapter->rx_ring[0]->total_bytes = 0;
/* would disable interrupts here but EIAM disabled it */
__napi_schedule(&(q_vector->napi));
}
@@ -4299,7 +4217,7 @@ static int ixgbe_poll(struct napi_struct *napi, int budget)
if (work_done < budget) {
napi_complete(napi);
if (adapter->rx_itr_setting & 1)
- ixgbe_set_itr(adapter);
+ ixgbe_set_itr(q_vector);
if (!test_bit(__IXGBE_DOWN, &adapter->state))
ixgbe_irq_enable_queues(adapter, IXGBE_EIMS_RTX_QUEUE);
}
@@ -5224,6 +5142,9 @@ static int __devinit ixgbe_sw_init(struct ixgbe_adapter *adapter)
adapter->tx_ring_count = IXGBE_DEFAULT_TXD;
adapter->rx_ring_count = IXGBE_DEFAULT_RXD;
+ /* set default work limits */
+ adapter->tx_work_limit = adapter->tx_ring_count;
+
/* initialize eeprom parameters */
if (ixgbe_init_eeprom_params_generic(hw)) {
e_dev_err("EEPROM initialization failed\n");
@@ -5270,7 +5191,6 @@ int ixgbe_setup_tx_resources(struct ixgbe_ring *tx_ring)
tx_ring->next_to_use = 0;
tx_ring->next_to_clean = 0;
- tx_ring->work_limit = tx_ring->count;
return 0;
err:
--
1.7.6
^ permalink raw reply related
* [net-next 07/10 v2] ixgbe: Pass staterr instead of re-reading status and error bits from descriptor
From: Jeff Kirsher @ 2011-07-22 6:09 UTC (permalink / raw)
To: davem; +Cc: Alexander Duyck, netdev, gospo, sassmann, Jeff Kirsher
In-Reply-To: <1311314951-11142-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Alexander Duyck <alexander.h.duyck@intel.com>
This change is meant to address possible race conditions from the status
and error bits on the RX descriptors being re-read by multiple functions in
the RX cleanup path. To resolve this I have added code that will pass the
staterr value to those functions.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Ross Brattain <ross.b.brattain@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe.h | 5 +++--
drivers/net/ixgbe/ixgbe_fcoe.c | 35 ++++++-----------------------------
drivers/net/ixgbe/ixgbe_main.c | 39 ++++++++++++++++++++++++++++-----------
3 files changed, 37 insertions(+), 42 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe.h b/drivers/net/ixgbe/ixgbe.h
index fbae735..0d610c7 100644
--- a/drivers/net/ixgbe/ixgbe.h
+++ b/drivers/net/ixgbe/ixgbe.h
@@ -600,8 +600,9 @@ extern int ixgbe_fso(struct ixgbe_ring *tx_ring, struct sk_buff *skb,
u32 tx_flags, u8 *hdr_len);
extern void ixgbe_cleanup_fcoe(struct ixgbe_adapter *adapter);
extern int ixgbe_fcoe_ddp(struct ixgbe_adapter *adapter,
- union ixgbe_adv_rx_desc *rx_desc,
- struct sk_buff *skb);
+ union ixgbe_adv_rx_desc *rx_desc,
+ struct sk_buff *skb,
+ u32 staterr);
extern int ixgbe_fcoe_ddp_get(struct net_device *netdev, u16 xid,
struct scatterlist *sgl, unsigned int sgc);
extern int ixgbe_fcoe_ddp_target(struct net_device *netdev, u16 xid,
diff --git a/drivers/net/ixgbe/ixgbe_fcoe.c b/drivers/net/ixgbe/ixgbe_fcoe.c
index f0c1018..824edae 100644
--- a/drivers/net/ixgbe/ixgbe_fcoe.c
+++ b/drivers/net/ixgbe/ixgbe_fcoe.c
@@ -37,25 +37,6 @@
#include <scsi/libfcoe.h>
/**
- * ixgbe_rx_is_fcoe - check the rx desc for incoming pkt type
- * @rx_desc: advanced rx descriptor
- *
- * Returns : true if it is FCoE pkt
- */
-static inline bool ixgbe_rx_is_fcoe(union ixgbe_adv_rx_desc *rx_desc)
-{
- u16 p;
-
- p = le16_to_cpu(rx_desc->wb.lower.lo_dword.hs_rss.pkt_info);
- if (p & IXGBE_RXDADV_PKTTYPE_ETQF) {
- p &= IXGBE_RXDADV_PKTTYPE_ETQF_MASK;
- p >>= IXGBE_RXDADV_PKTTYPE_ETQF_SHIFT;
- return p == IXGBE_ETQF_FILTER_FCOE;
- }
- return false;
-}
-
-/**
* ixgbe_fcoe_clear_ddp - clear the given ddp context
* @ddp - ptr to the ixgbe_fcoe_ddp
*
@@ -136,7 +117,6 @@ out_ddp_put:
return len;
}
-
/**
* ixgbe_fcoe_ddp_setup - called to set up ddp context
* @netdev: the corresponding net_device
@@ -380,23 +360,20 @@ int ixgbe_fcoe_ddp_target(struct net_device *netdev, u16 xid,
*/
int ixgbe_fcoe_ddp(struct ixgbe_adapter *adapter,
union ixgbe_adv_rx_desc *rx_desc,
- struct sk_buff *skb)
+ struct sk_buff *skb,
+ u32 staterr)
{
u16 xid;
u32 fctl;
- u32 sterr, fceofe, fcerr, fcstat;
+ u32 fceofe, fcerr, fcstat;
int rc = -EINVAL;
struct ixgbe_fcoe *fcoe;
struct ixgbe_fcoe_ddp *ddp;
struct fc_frame_header *fh;
struct fcoe_crc_eof *crc;
- if (!ixgbe_rx_is_fcoe(rx_desc))
- goto ddp_out;
-
- sterr = le32_to_cpu(rx_desc->wb.upper.status_error);
- fcerr = (sterr & IXGBE_RXDADV_ERR_FCERR);
- fceofe = (sterr & IXGBE_RXDADV_ERR_FCEOFE);
+ fcerr = (staterr & IXGBE_RXDADV_ERR_FCERR);
+ fceofe = (staterr & IXGBE_RXDADV_ERR_FCEOFE);
if (fcerr == IXGBE_FCERR_BADCRC)
skb_checksum_none_assert(skb);
else
@@ -425,7 +402,7 @@ int ixgbe_fcoe_ddp(struct ixgbe_adapter *adapter,
if (fcerr | fceofe)
goto ddp_out;
- fcstat = (sterr & IXGBE_RXDADV_STAT_FCSTAT);
+ fcstat = (staterr & IXGBE_RXDADV_STAT_FCSTAT);
if (fcstat) {
/* update length of DDPed data */
ddp->len = le32_to_cpu(rx_desc->wb.lower.hi_dword.rss);
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index 298c95b..a222af3 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -1039,6 +1039,24 @@ static inline void ixgbe_rx_hash(union ixgbe_adv_rx_desc *rx_desc,
}
/**
+ * ixgbe_rx_is_fcoe - check the rx desc for incoming pkt type
+ * @adapter: address of board private structure
+ * @rx_desc: advanced rx descriptor
+ *
+ * Returns : true if it is FCoE pkt
+ */
+static inline bool ixgbe_rx_is_fcoe(struct ixgbe_adapter *adapter,
+ union ixgbe_adv_rx_desc *rx_desc)
+{
+ __le16 pkt_info = rx_desc->wb.lower.lo_dword.hs_rss.pkt_info;
+
+ return (adapter->flags & IXGBE_FLAG_FCOE_ENABLED) &&
+ ((pkt_info & cpu_to_le16(IXGBE_RXDADV_PKTTYPE_ETQF_MASK)) ==
+ (cpu_to_le16(IXGBE_ETQF_FILTER_FCOE <<
+ IXGBE_RXDADV_PKTTYPE_ETQF_SHIFT)));
+}
+
+/**
* ixgbe_receive_skb - Send a completed packet up the stack
* @adapter: board private structure
* @skb: packet to send up
@@ -1070,14 +1088,14 @@ static void ixgbe_receive_skb(struct ixgbe_q_vector *q_vector,
* @adapter: address of board private structure
* @status_err: hardware indication of status of receive
* @skb: skb currently being received and modified
+ * @status_err: status error value of last descriptor in packet
**/
static inline void ixgbe_rx_checksum(struct ixgbe_adapter *adapter,
union ixgbe_adv_rx_desc *rx_desc,
- struct sk_buff *skb)
+ struct sk_buff *skb,
+ u32 status_err)
{
- u32 status_err = le32_to_cpu(rx_desc->wb.upper.status_error);
-
- skb_checksum_none_assert(skb);
+ skb->ip_summed = CHECKSUM_NONE;
/* Rx csum disabled */
if (!(adapter->flags & IXGBE_FLAG_RX_CSUM_ENABLED))
@@ -1421,14 +1439,12 @@ static void ixgbe_clean_rx_irq(struct ixgbe_q_vector *q_vector,
}
/* ERR_MASK will only have valid bits if EOP set */
- if (staterr & IXGBE_RXDADV_ERR_FRAME_ERR_MASK) {
- /* trim packet back to size 0 and recycle it */
- __pskb_trim(skb, 0);
- rx_buffer_info->skb = skb;
+ if (unlikely(staterr & IXGBE_RXDADV_ERR_FRAME_ERR_MASK)) {
+ dev_kfree_skb_any(skb);
goto next_desc;
}
- ixgbe_rx_checksum(adapter, rx_desc, skb);
+ ixgbe_rx_checksum(adapter, rx_desc, skb, staterr);
if (adapter->netdev->features & NETIF_F_RXHASH)
ixgbe_rx_hash(rx_desc, skb);
@@ -1439,8 +1455,9 @@ static void ixgbe_clean_rx_irq(struct ixgbe_q_vector *q_vector,
skb->protocol = eth_type_trans(skb, rx_ring->netdev);
#ifdef IXGBE_FCOE
/* if ddp, not passing to ULD unless for FCP_RSP or error */
- if (adapter->flags & IXGBE_FLAG_FCOE_ENABLED) {
- ddp_bytes = ixgbe_fcoe_ddp(adapter, rx_desc, skb);
+ if (ixgbe_rx_is_fcoe(adapter, rx_desc)) {
+ ddp_bytes = ixgbe_fcoe_ddp(adapter, rx_desc, skb,
+ staterr);
if (!ddp_bytes)
goto next_desc;
}
--
1.7.6
^ permalink raw reply related
* [net-next 05/10] ixgbe: add structure for containing RX/TX rings to q_vector
From: Jeff Kirsher @ 2011-07-22 6:09 UTC (permalink / raw)
To: davem; +Cc: Alexander Duyck, netdev, gospo, sassmann, Jeff Kirsher
In-Reply-To: <1311314951-11142-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Alexander Duyck <alexander.h.duyck@intel.com>
This patch adds support for a ring container structure to be used within
the q_vector. The basic idea is to provide a means of separating the RX
and TX rings while maintaining a common structure for their containment.
The advantage to this is that later we should be able to pass this
structure to the update_itr functions without needing to pass individual
rings.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Ross Brattain <ross.b.brattain@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe.h | 16 +++--
drivers/net/ixgbe/ixgbe_ethtool.c | 6 +-
drivers/net/ixgbe/ixgbe_main.c | 150 ++++++++++++++++++------------------
3 files changed, 88 insertions(+), 84 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe.h b/drivers/net/ixgbe/ixgbe.h
index 744b641..a5ed9fd 100644
--- a/drivers/net/ixgbe/ixgbe.h
+++ b/drivers/net/ixgbe/ixgbe.h
@@ -277,6 +277,15 @@ struct ixgbe_ring_feature {
int mask;
} ____cacheline_internodealigned_in_smp;
+struct ixgbe_ring_container {
+#if MAX_RX_QUEUES > MAX_TX_QUEUES
+ DECLARE_BITMAP(idx, MAX_RX_QUEUES);
+#else
+ DECLARE_BITMAP(idx, MAX_TX_QUEUES);
+#endif
+ u8 count; /* total number of rings in vector */
+ u8 itr; /* current ITR setting for ring */
+};
#define MAX_RX_PACKET_BUFFERS ((adapter->flags & IXGBE_FLAG_DCB_ENABLED) \
? 8 : 1)
@@ -294,12 +303,7 @@ struct ixgbe_q_vector {
int cpu; /* CPU for DCA */
#endif
struct napi_struct napi;
- DECLARE_BITMAP(rxr_idx, MAX_RX_QUEUES); /* Rx ring indices */
- DECLARE_BITMAP(txr_idx, MAX_TX_QUEUES); /* Tx ring indices */
- u8 rxr_count; /* Rx ring count assigned to this vector */
- u8 txr_count; /* Tx ring count assigned to this vector */
- u8 tx_itr;
- u8 rx_itr;
+ struct ixgbe_ring_container rx, tx;
u32 eitr;
cpumask_var_t affinity_mask;
char name[IFNAMSIZ + 9];
diff --git a/drivers/net/ixgbe/ixgbe_ethtool.c b/drivers/net/ixgbe/ixgbe_ethtool.c
index 074e9ba..cd842f9 100644
--- a/drivers/net/ixgbe/ixgbe_ethtool.c
+++ b/drivers/net/ixgbe/ixgbe_ethtool.c
@@ -2122,7 +2122,7 @@ static int ixgbe_get_coalesce(struct net_device *netdev,
}
/* if in mixed tx/rx queues per vector mode, report only rx settings */
- if (adapter->q_vector[0]->txr_count && adapter->q_vector[0]->rxr_count)
+ if (adapter->q_vector[0]->tx.count && adapter->q_vector[0]->rx.count)
return 0;
/* only valid if in constant ITR mode */
@@ -2187,7 +2187,7 @@ static int ixgbe_set_coalesce(struct net_device *netdev,
bool need_reset = false;
/* don't accept tx specific changes if we've got mixed RxTx vectors */
- if (adapter->q_vector[0]->txr_count && adapter->q_vector[0]->rxr_count
+ if (adapter->q_vector[0]->tx.count && adapter->q_vector[0]->rx.count
&& ec->tx_coalesce_usecs)
return -EINVAL;
@@ -2261,7 +2261,7 @@ static int ixgbe_set_coalesce(struct net_device *netdev,
int num_vectors = adapter->num_msix_vectors - NON_Q_VECTORS;
for (i = 0; i < num_vectors; i++) {
q_vector = adapter->q_vector[i];
- if (q_vector->txr_count && !q_vector->rxr_count)
+ if (q_vector->tx.count && !q_vector->rx.count)
/* tx only */
q_vector->eitr = adapter->tx_eitr_param;
else
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index 7275c5c..bd1fd8f 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -959,17 +959,17 @@ static void ixgbe_update_dca(struct ixgbe_q_vector *q_vector)
if (q_vector->cpu == cpu)
goto out_no_update;
- r_idx = find_first_bit(q_vector->txr_idx, adapter->num_tx_queues);
- for (i = 0; i < q_vector->txr_count; i++) {
+ r_idx = find_first_bit(q_vector->tx.idx, adapter->num_tx_queues);
+ for (i = 0; i < q_vector->tx.count; i++) {
ixgbe_update_tx_dca(adapter, adapter->tx_ring[r_idx], cpu);
- r_idx = find_next_bit(q_vector->txr_idx, adapter->num_tx_queues,
+ r_idx = find_next_bit(q_vector->tx.idx, adapter->num_tx_queues,
r_idx + 1);
}
- r_idx = find_first_bit(q_vector->rxr_idx, adapter->num_rx_queues);
- for (i = 0; i < q_vector->rxr_count; i++) {
+ r_idx = find_first_bit(q_vector->rx.idx, adapter->num_rx_queues);
+ for (i = 0; i < q_vector->rx.count; i++) {
ixgbe_update_rx_dca(adapter, adapter->rx_ring[r_idx], cpu);
- r_idx = find_next_bit(q_vector->rxr_idx, adapter->num_rx_queues,
+ r_idx = find_next_bit(q_vector->rx.idx, adapter->num_rx_queues,
r_idx + 1);
}
@@ -1517,31 +1517,31 @@ static void ixgbe_configure_msix(struct ixgbe_adapter *adapter)
for (v_idx = 0; v_idx < q_vectors; v_idx++) {
q_vector = adapter->q_vector[v_idx];
/* XXX for_each_set_bit(...) */
- r_idx = find_first_bit(q_vector->rxr_idx,
+ r_idx = find_first_bit(q_vector->rx.idx,
adapter->num_rx_queues);
- for (i = 0; i < q_vector->rxr_count; i++) {
+ for (i = 0; i < q_vector->rx.count; i++) {
u8 reg_idx = adapter->rx_ring[r_idx]->reg_idx;
ixgbe_set_ivar(adapter, 0, reg_idx, v_idx);
- r_idx = find_next_bit(q_vector->rxr_idx,
+ r_idx = find_next_bit(q_vector->rx.idx,
adapter->num_rx_queues,
r_idx + 1);
}
- r_idx = find_first_bit(q_vector->txr_idx,
+ r_idx = find_first_bit(q_vector->tx.idx,
adapter->num_tx_queues);
- for (i = 0; i < q_vector->txr_count; i++) {
+ for (i = 0; i < q_vector->tx.count; i++) {
u8 reg_idx = adapter->tx_ring[r_idx]->reg_idx;
ixgbe_set_ivar(adapter, 1, reg_idx, v_idx);
- r_idx = find_next_bit(q_vector->txr_idx,
+ r_idx = find_next_bit(q_vector->tx.idx,
adapter->num_tx_queues,
r_idx + 1);
}
- if (q_vector->txr_count && !q_vector->rxr_count)
+ if (q_vector->tx.count && !q_vector->rx.count)
/* tx only */
q_vector->eitr = adapter->tx_eitr_param;
- else if (q_vector->rxr_count)
+ else if (q_vector->rx.count)
/* rx or mixed */
q_vector->eitr = adapter->rx_eitr_param;
@@ -1705,37 +1705,37 @@ static void ixgbe_set_itr_msix(struct ixgbe_q_vector *q_vector)
u32 new_itr;
u8 current_itr, ret_itr;
- r_idx = find_first_bit(q_vector->txr_idx, adapter->num_tx_queues);
- for (i = 0; i < q_vector->txr_count; i++) {
+ r_idx = find_first_bit(q_vector->tx.idx, adapter->num_tx_queues);
+ for (i = 0; i < q_vector->tx.count; i++) {
struct ixgbe_ring *tx_ring = adapter->tx_ring[r_idx];
ret_itr = ixgbe_update_itr(adapter, q_vector->eitr,
- q_vector->tx_itr,
+ q_vector->tx.itr,
tx_ring->total_packets,
tx_ring->total_bytes);
/* if the result for this queue would decrease interrupt
* rate for this vector then use that result */
- q_vector->tx_itr = ((q_vector->tx_itr > ret_itr) ?
- q_vector->tx_itr - 1 : ret_itr);
- r_idx = find_next_bit(q_vector->txr_idx, adapter->num_tx_queues,
+ q_vector->tx.itr = ((q_vector->tx.itr > ret_itr) ?
+ q_vector->tx.itr - 1 : ret_itr);
+ r_idx = find_next_bit(q_vector->tx.idx, adapter->num_tx_queues,
r_idx + 1);
}
- r_idx = find_first_bit(q_vector->rxr_idx, adapter->num_rx_queues);
- for (i = 0; i < q_vector->rxr_count; i++) {
+ r_idx = find_first_bit(q_vector->rx.idx, adapter->num_rx_queues);
+ for (i = 0; i < q_vector->rx.count; i++) {
struct ixgbe_ring *rx_ring = adapter->rx_ring[r_idx];
ret_itr = ixgbe_update_itr(adapter, q_vector->eitr,
- q_vector->rx_itr,
+ q_vector->rx.itr,
rx_ring->total_packets,
rx_ring->total_bytes);
/* if the result for this queue would decrease interrupt
* rate for this vector then use that result */
- q_vector->rx_itr = ((q_vector->rx_itr > ret_itr) ?
- q_vector->rx_itr - 1 : ret_itr);
- r_idx = find_next_bit(q_vector->rxr_idx, adapter->num_rx_queues,
+ q_vector->rx.itr = ((q_vector->rx.itr > ret_itr) ?
+ q_vector->rx.itr - 1 : ret_itr);
+ r_idx = find_next_bit(q_vector->rx.idx, adapter->num_rx_queues,
r_idx + 1);
}
- current_itr = max(q_vector->rx_itr, q_vector->tx_itr);
+ current_itr = max(q_vector->rx.itr, q_vector->tx.itr);
switch (current_itr) {
/* counts and packets in update_itr are dependent on these numbers */
@@ -1995,15 +1995,15 @@ static irqreturn_t ixgbe_msix_clean_tx(int irq, void *data)
struct ixgbe_ring *tx_ring;
int i, r_idx;
- if (!q_vector->txr_count)
+ if (!q_vector->tx.count)
return IRQ_HANDLED;
- r_idx = find_first_bit(q_vector->txr_idx, adapter->num_tx_queues);
- for (i = 0; i < q_vector->txr_count; i++) {
+ r_idx = find_first_bit(q_vector->tx.idx, adapter->num_tx_queues);
+ for (i = 0; i < q_vector->tx.count; i++) {
tx_ring = adapter->tx_ring[r_idx];
tx_ring->total_bytes = 0;
tx_ring->total_packets = 0;
- r_idx = find_next_bit(q_vector->txr_idx, adapter->num_tx_queues,
+ r_idx = find_next_bit(q_vector->tx.idx, adapter->num_tx_queues,
r_idx + 1);
}
@@ -2031,16 +2031,16 @@ static irqreturn_t ixgbe_msix_clean_rx(int irq, void *data)
ixgbe_update_dca(q_vector);
#endif
- r_idx = find_first_bit(q_vector->rxr_idx, adapter->num_rx_queues);
- for (i = 0; i < q_vector->rxr_count; i++) {
+ r_idx = find_first_bit(q_vector->rx.idx, adapter->num_rx_queues);
+ for (i = 0; i < q_vector->rx.count; i++) {
rx_ring = adapter->rx_ring[r_idx];
rx_ring->total_bytes = 0;
rx_ring->total_packets = 0;
- r_idx = find_next_bit(q_vector->rxr_idx, adapter->num_rx_queues,
+ r_idx = find_next_bit(q_vector->rx.idx, adapter->num_rx_queues,
r_idx + 1);
}
- if (!q_vector->rxr_count)
+ if (!q_vector->rx.count)
return IRQ_HANDLED;
/* EIAM disabled interrupts (on this vector) for us */
@@ -2057,24 +2057,24 @@ static irqreturn_t ixgbe_msix_clean_many(int irq, void *data)
int r_idx;
int i;
- if (!q_vector->txr_count && !q_vector->rxr_count)
+ if (!q_vector->tx.count && !q_vector->rx.count)
return IRQ_HANDLED;
- r_idx = find_first_bit(q_vector->txr_idx, adapter->num_tx_queues);
- for (i = 0; i < q_vector->txr_count; i++) {
+ r_idx = find_first_bit(q_vector->tx.idx, adapter->num_tx_queues);
+ for (i = 0; i < q_vector->tx.count; i++) {
ring = adapter->tx_ring[r_idx];
ring->total_bytes = 0;
ring->total_packets = 0;
- r_idx = find_next_bit(q_vector->txr_idx, adapter->num_tx_queues,
+ r_idx = find_next_bit(q_vector->tx.idx, adapter->num_tx_queues,
r_idx + 1);
}
- r_idx = find_first_bit(q_vector->rxr_idx, adapter->num_rx_queues);
- for (i = 0; i < q_vector->rxr_count; i++) {
+ r_idx = find_first_bit(q_vector->rx.idx, adapter->num_rx_queues);
+ for (i = 0; i < q_vector->rx.count; i++) {
ring = adapter->rx_ring[r_idx];
ring->total_bytes = 0;
ring->total_packets = 0;
- r_idx = find_next_bit(q_vector->rxr_idx, adapter->num_rx_queues,
+ r_idx = find_next_bit(q_vector->rx.idx, adapter->num_rx_queues,
r_idx + 1);
}
@@ -2106,7 +2106,7 @@ static int ixgbe_clean_rxonly(struct napi_struct *napi, int budget)
ixgbe_update_dca(q_vector);
#endif
- r_idx = find_first_bit(q_vector->rxr_idx, adapter->num_rx_queues);
+ r_idx = find_first_bit(q_vector->rx.idx, adapter->num_rx_queues);
rx_ring = adapter->rx_ring[r_idx];
ixgbe_clean_rx_irq(q_vector, rx_ring, &work_done, budget);
@@ -2147,27 +2147,27 @@ static int ixgbe_clean_rxtx_many(struct napi_struct *napi, int budget)
ixgbe_update_dca(q_vector);
#endif
- r_idx = find_first_bit(q_vector->txr_idx, adapter->num_tx_queues);
- for (i = 0; i < q_vector->txr_count; i++) {
+ r_idx = find_first_bit(q_vector->tx.idx, adapter->num_tx_queues);
+ for (i = 0; i < q_vector->tx.count; i++) {
ring = adapter->tx_ring[r_idx];
tx_clean_complete &= ixgbe_clean_tx_irq(q_vector, ring);
- r_idx = find_next_bit(q_vector->txr_idx, adapter->num_tx_queues,
+ r_idx = find_next_bit(q_vector->tx.idx, adapter->num_tx_queues,
r_idx + 1);
}
/* attempt to distribute budget to each queue fairly, but don't allow
* the budget to go below 1 because we'll exit polling */
- budget /= (q_vector->rxr_count ?: 1);
+ budget /= (q_vector->rx.count ?: 1);
budget = max(budget, 1);
- r_idx = find_first_bit(q_vector->rxr_idx, adapter->num_rx_queues);
- for (i = 0; i < q_vector->rxr_count; i++) {
+ r_idx = find_first_bit(q_vector->rx.idx, adapter->num_rx_queues);
+ for (i = 0; i < q_vector->rx.count; i++) {
ring = adapter->rx_ring[r_idx];
ixgbe_clean_rx_irq(q_vector, ring, &work_done, budget);
- r_idx = find_next_bit(q_vector->rxr_idx, adapter->num_rx_queues,
+ r_idx = find_next_bit(q_vector->rx.idx, adapter->num_rx_queues,
r_idx + 1);
}
- r_idx = find_first_bit(q_vector->rxr_idx, adapter->num_rx_queues);
+ r_idx = find_first_bit(q_vector->rx.idx, adapter->num_rx_queues);
ring = adapter->rx_ring[r_idx];
/* If all Rx work done, exit the polling mode */
if (work_done < budget) {
@@ -2205,7 +2205,7 @@ static int ixgbe_clean_txonly(struct napi_struct *napi, int budget)
ixgbe_update_dca(q_vector);
#endif
- r_idx = find_first_bit(q_vector->txr_idx, adapter->num_tx_queues);
+ r_idx = find_first_bit(q_vector->tx.idx, adapter->num_tx_queues);
tx_ring = adapter->tx_ring[r_idx];
if (!ixgbe_clean_tx_irq(q_vector, tx_ring))
@@ -2230,8 +2230,8 @@ static inline void map_vector_to_rxq(struct ixgbe_adapter *a, int v_idx,
struct ixgbe_q_vector *q_vector = a->q_vector[v_idx];
struct ixgbe_ring *rx_ring = a->rx_ring[r_idx];
- set_bit(r_idx, q_vector->rxr_idx);
- q_vector->rxr_count++;
+ set_bit(r_idx, q_vector->rx.idx);
+ q_vector->rx.count++;
rx_ring->q_vector = q_vector;
}
@@ -2241,8 +2241,8 @@ static inline void map_vector_to_txq(struct ixgbe_adapter *a, int v_idx,
struct ixgbe_q_vector *q_vector = a->q_vector[v_idx];
struct ixgbe_ring *tx_ring = a->tx_ring[t_idx];
- set_bit(t_idx, q_vector->txr_idx);
- q_vector->txr_count++;
+ set_bit(t_idx, q_vector->tx.idx);
+ q_vector->tx.count++;
tx_ring->q_vector = q_vector;
}
@@ -2332,10 +2332,10 @@ static int ixgbe_request_msix_irqs(struct ixgbe_adapter *adapter)
if (err)
return err;
-#define SET_HANDLER(_v) (((_v)->rxr_count && (_v)->txr_count) \
+#define SET_HANDLER(_v) (((_v)->rx.count && (_v)->tx.count) \
? &ixgbe_msix_clean_many : \
- (_v)->rxr_count ? &ixgbe_msix_clean_rx : \
- (_v)->txr_count ? &ixgbe_msix_clean_tx : \
+ (_v)->rx.count ? &ixgbe_msix_clean_rx : \
+ (_v)->tx.count ? &ixgbe_msix_clean_tx : \
NULL)
for (vector = 0; vector < q_vectors; vector++) {
struct ixgbe_q_vector *q_vector = adapter->q_vector[vector];
@@ -2394,16 +2394,16 @@ static void ixgbe_set_itr(struct ixgbe_adapter *adapter)
u32 new_itr = q_vector->eitr;
u8 current_itr;
- q_vector->tx_itr = ixgbe_update_itr(adapter, new_itr,
- q_vector->tx_itr,
+ q_vector->tx.itr = ixgbe_update_itr(adapter, new_itr,
+ q_vector->tx.itr,
tx_ring->total_packets,
tx_ring->total_bytes);
- q_vector->rx_itr = ixgbe_update_itr(adapter, new_itr,
- q_vector->rx_itr,
+ q_vector->rx.itr = ixgbe_update_itr(adapter, new_itr,
+ q_vector->rx.itr,
rx_ring->total_packets,
rx_ring->total_bytes);
- current_itr = max(q_vector->rx_itr, q_vector->tx_itr);
+ current_itr = max(q_vector->rx.itr, q_vector->tx.itr);
switch (current_itr) {
/* counts and packets in update_itr are dependent on these numbers */
@@ -2553,10 +2553,10 @@ static inline void ixgbe_reset_q_vectors(struct ixgbe_adapter *adapter)
for (i = 0; i < q_vectors; i++) {
struct ixgbe_q_vector *q_vector = adapter->q_vector[i];
- bitmap_zero(q_vector->rxr_idx, MAX_RX_QUEUES);
- bitmap_zero(q_vector->txr_idx, MAX_TX_QUEUES);
- q_vector->rxr_count = 0;
- q_vector->txr_count = 0;
+ bitmap_zero(q_vector->rx.idx, MAX_RX_QUEUES);
+ bitmap_zero(q_vector->tx.idx, MAX_TX_QUEUES);
+ q_vector->rx.count = 0;
+ q_vector->tx.count = 0;
}
}
@@ -2601,8 +2601,8 @@ static void ixgbe_free_irq(struct ixgbe_adapter *adapter)
i--;
for (; i >= 0; i--) {
/* free only the irqs that were actually requested */
- if (!adapter->q_vector[i]->rxr_count &&
- !adapter->q_vector[i]->txr_count)
+ if (!adapter->q_vector[i]->rx.count &&
+ !adapter->q_vector[i]->tx.count)
continue;
free_irq(adapter->msix_entries[i].vector,
@@ -3616,10 +3616,10 @@ static void ixgbe_napi_enable_all(struct ixgbe_adapter *adapter)
q_vector = adapter->q_vector[q_idx];
napi = &q_vector->napi;
if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED) {
- if (!q_vector->rxr_count || !q_vector->txr_count) {
- if (q_vector->txr_count == 1)
+ if (!q_vector->rx.count || !q_vector->tx.count) {
+ if (q_vector->tx.count == 1)
napi->poll = &ixgbe_clean_txonly;
- else if (q_vector->rxr_count == 1)
+ else if (q_vector->rx.count == 1)
napi->poll = &ixgbe_clean_rxonly;
}
}
@@ -4965,7 +4965,7 @@ static int ixgbe_alloc_q_vectors(struct ixgbe_adapter *adapter)
if (!q_vector)
goto err_out;
q_vector->adapter = adapter;
- if (q_vector->txr_count && !q_vector->rxr_count)
+ if (q_vector->tx.count && !q_vector->rx.count)
q_vector->eitr = adapter->tx_eitr_param;
else
q_vector->eitr = adapter->rx_eitr_param;
@@ -5979,7 +5979,7 @@ static void ixgbe_check_hang_subtask(struct ixgbe_adapter *adapter)
/* get one bit for every active tx/rx interrupt vector */
for (i = 0; i < adapter->num_msix_vectors - NON_Q_VECTORS; i++) {
struct ixgbe_q_vector *qv = adapter->q_vector[i];
- if (qv->rxr_count || qv->txr_count)
+ if (qv->rx.count || qv->tx.count)
eics |= ((u64)1 << i);
}
}
--
1.7.6
^ permalink raw reply related
* [net-next 04/10] ixgbe: inline the ixgbe_maybe_stop_tx function
From: Jeff Kirsher @ 2011-07-22 6:09 UTC (permalink / raw)
To: davem; +Cc: Alexander Duyck, netdev, gospo, sassmann, Jeff Kirsher
In-Reply-To: <1311314951-11142-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Alexander Duyck <alexander.h.duyck@intel.com>
The ixgbe_maybe_stop_tx function is only a few lines long and is called
multiple times through the xmit hotpath. In order to streamline things it
makes sense to just inline it.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Ross Brattain <ross.b.brattain@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_main.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index f05ddf3..7275c5c 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -6785,7 +6785,7 @@ static int __ixgbe_maybe_stop_tx(struct ixgbe_ring *tx_ring, u16 size)
return 0;
}
-static int ixgbe_maybe_stop_tx(struct ixgbe_ring *tx_ring, u16 size)
+static inline int ixgbe_maybe_stop_tx(struct ixgbe_ring *tx_ring, u16 size)
{
if (likely(ixgbe_desc_unused(tx_ring) >= size))
return 0;
--
1.7.6
^ permalink raw reply related
* [net-next 03/10] ixgbe: Update ATR to use recorded TX queues instead of CPU for routing
From: Jeff Kirsher @ 2011-07-22 6:09 UTC (permalink / raw)
To: davem; +Cc: Alexander Duyck, netdev, gospo, sassmann, Jeff Kirsher
In-Reply-To: <1311314951-11142-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Alexander Duyck <alexander.h.duyck@intel.com>
This change is meant to update ATR so that it will use the recorded RX
queue instead of the CPU in the case of routing. This change is meant to
help ixgbe default behavior to more closely match that of the kernel.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Ross Brattain <ross.b.brattain@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
---
drivers/net/ixgbe/ixgbe_main.c | 7 +++----
1 files changed, 3 insertions(+), 4 deletions(-)
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index de30796..f05ddf3 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -6795,11 +6795,10 @@ static int ixgbe_maybe_stop_tx(struct ixgbe_ring *tx_ring, u16 size)
static u16 ixgbe_select_queue(struct net_device *dev, struct sk_buff *skb)
{
struct ixgbe_adapter *adapter = netdev_priv(dev);
- int txq = smp_processor_id();
+ int txq = skb_rx_queue_recorded(skb) ? skb_get_rx_queue(skb) :
+ smp_processor_id();
#ifdef IXGBE_FCOE
- __be16 protocol;
-
- protocol = vlan_get_protocol(skb);
+ __be16 protocol = vlan_get_protocol(skb);
if (((protocol == htons(ETH_P_FCOE)) ||
(protocol == htons(ETH_P_FIP))) &&
--
1.7.6
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox