On Thu, Feb 02, 2017 at 01:04:34AM +0200, Julian Anastasov wrote:
>
> Hello,
>
> On Wed, 1 Feb 2017, Steffen Klassert wrote:
>
> >
> > I think here it is better to go through the whole chain
> > of transformations with
> >
> > child->ops->confirm_neigh(path, daddr);
>
> It may sou
Thu, Feb 02, 2017 at 10:37:23PM CET, f.faine...@gmail.com wrote:
>On 02/02/2017 07:12 AM, Jiri Pirko wrote:
>> From: Jiri Pirko
>>
>> Extend the existing setup_tc ndo call and allow to offload cls_flower
>> rules. Only limited set of dissector keys and actions are supported now.
>> Use previously
From: Jayaprakash Shanmugam
Requirement for VFs to use the VMBus has been removed that's why
removing Hyper-V VF device ID.
Change-ID: I84f0964f443ee0db3e5e444b5ace996eb71b8280
Signed-off-by: Jayaprakash Shanmugam
Tested-by: Andrew Bowers
Signed-off-by: Jeff Kirsher
---
drivers/net/ethernet/
From: Mitch Williams
After refactoring the client open and close code, this is no longer
needed. Remove it.
Change-ID: If8e6e32baa354d857c2fd8b2f19404f1786011c4
Signed-off-by: Mitch Williams
Tested-by: Andrew Bowers
Signed-off-by: Jeff Kirsher
---
drivers/net/ethernet/intel/i40e/i40e.h
This series contains updates to i40e/i40evf only.
Jake fixes up the driver to not call i40e_vsi_kill_vlan() or
i40e_vsi_add_vlan() when the PVID is set or when the VID is less than 1.
Cleaned up a check which really is not needed since there is no real
reason why we cannot just call i40e_del_mac_a
From: Alexander Duyck
This patch does some quick work to pull some of the data off of the stack
and hopefully start storing it in the Tx buffer info section of the Tx
ring. Ideally we should be moving away from having to store much of
anything on the stack and can just maintain it all in the des
From: Jacob Keller
These functions purpose is to add a new MAC filter correctly, whether
we're using VLANs or not. Their goal is to ensure that all active VLANs
get the new MAC filter. Rename them so that their intent is clear. They
function correctly regardless of whether we have any active VLAN
From: Mitch Williams
This device ID was intended for use when running Linux VF drivers under
Hyper-V, but we have determined that it is not necessary. Since it is
unused, and will never be used, remove it.
Change-ID: I74998ab4237db043cd400547bb54a0a5e2a37ea5
Signed-off-by: Mitch Williams
Tested
From: Bimmy Pujari
I40E_MAC_X710 was supposed to be for 10G and I40E_MAC_XL710
was supposed to be for 40G. But function i40e_is_mac_710
sets I40E_MAC_XL710 for all device IDS, I40E_MAC_X710 is not
used at all. As there is nothing to compare there is no need
for this function. Thus deprecating thi
From: Scott Peterson
On packet RX, we perform a DMA sync for CPU before passing the
packet up. Here we limit that sync to the actual length of the
incoming packet, rather than always syncing the entire buffer.
Change-ID: I626aaf6c37275a8ce9e81efcaa773f327b331487
Signed-off-by: Scott Peterson
T
From: Jacob Keller
Use __i40e_del_filter instead of using i40e_del_filter() which will
avoid doing an additional search to delete a filter we already have the
pointer for.
Change-ID: Iea5a7e3cafbf8c682ed9d3b6c69cf5ff53f44daf
Signed-off-by: Jacob Keller
Tested-by: Andrew Bowers
Signed-off-by: J
From: Mitch Williams
The iWarp client cannot continue until this operation has been completed
by the PF driver. Sleep (with timeout) until the reply from the PF
driver has been received.
Change-ID: I5dc41b857bba32d0218b7ce167b5da122dadf349
Signed-off-by: Mitch Williams
Tested-by: Andrew Bowers
From: Tushar Dave
'struct i40e_dma_mem' defined with 'packed' directive causing kernel
unaligned errors on sparc.
e.g.
i40e: Intel(R) Ethernet Connection XL710 Network Driver - version
1.6.16-k
i40e: Copyright (c) 2013 - 2014 Intel Corporation.
Kernel unaligned access at TPC[44894c] dma_4v_alloc
From: Jacob Keller
This function won't be appreciably slower when in VLAN mode, so there is
no real reason to not just call it directly. In either case, we still
must search the full table for a MAC/VLAN pair. We do get to stop
searching a tiny bit early in the case of knowing we are not in VLAN
From: Alan Brady
This patch refactors the macro INTRL_USEC_TO_REG into a static inline
function and fixes a couple subtle bugs caused by the macro.
This patch fixes a bug which was caused by passing a bad register value
to the firmware. If enabling interrupt rate limiting, a non-zero value
for
From: Scott Peterson
This patch reduces the size of struct i40e_rx_buffer by one pointer,
and makes the i40e driver a little more consistent with the igb driver
in terms of packets that span buffers.
We do this by moving the skb field from struct i40e_rx_buffer to
struct i40e_ring. We pass the s
From: Jacob Keller
Fold the check for determining when to call i40e_put_mac_in_vlan directly
into the function so that we don't need to decide which function to use
ahead of time. This allows us to just call i40e_put_mac_in_vlan directly
without having to check ahead of time.
Change-ID: Ifff5269
From: Jacob Keller
Now that we have the separate i40e_(add|rm)_vlan_all_mac functions, we
should not be using the i40e_vsi_kill_vlan or i40e_vsi_add_vlan
functions when PVID is set or when VID is less than 1. This allows us to
remove some checks in i40e_vsi_add_vlan and ensures that callers which
From: Faisal Latif
The function prototype in i40evf_client.h are giving warnings while
compiling i40iwvf module. Move these function prototypes to i40evf.h.
Change-Id: Ie3757f844993aabc27654aaf02ec14fb985ad2c4
Signed-off-by: Faisal Latif
Tested-by: Andrew Bowers
Signed-off-by: Jeff Kirsher
--
From: Alan Brady
Due to the resolution of the register controlling interrupt rate
limiting, setting certain values for the interrupt rate limit make it
appear as though the limiting is not completely accurate. The problem
is that the interrupt rate limit is getting rounded down to the nearest
mu
From: Jacob Keller
We can avoid the minor bit of work by calling check params after we
check for the client instance, since we're about to return early in
cases where we do not have a client.
Change-ID: I56f8ea2ba48d4f571fa331c9ace50819a022fa1c
Signed-off-by: Jacob Keller
Tested-by: Andrew Bowe
From: Jacob Keller
Instead of using i40e_add_filter or i40e_del_filter directly, when
adding a MAC address, we should normally be using i40e_add_mac_filter or
i40e_del_mac_filter. These functions correctly handle the various cases
of VLAN mode or PVID settings. This ensures consistency and avoids
When for instance a mobile Linux device roams from one access point to
another with both APs sharing the same broadcast domain and a
multicast snooping switch in between:
1)(c) <~~~> (AP1) <--[SSW]--> (AP2)
2) (AP1) <--[SSW]--> (AP2) <~~~> (c)
Then currently IPv6 multicast packe
Thu, Feb 02, 2017 at 10:58:15PM CET, t...@herbertland.com wrote:
>I have no idea what a "priority array area manager" is. Googling it
>comes up with nothing and there are no comments in this whole patch
>that either describe what it is or how any of the functions should be
>used. Am I missing somet
Hi Nick,
On Thu, Feb 2, 2017 at 12:59 PM, nick viljoen
wrote:
> Hey Shubham,
>
> I have been doing some similar work-might be worth pooling
> resource if there is interest?
Sure. That sounds great.
>
> We made a presentation at the previous netdev conference about
> what we are doing-you can ch
Thu, Feb 02, 2017 at 11:51:10PM CET, j...@perches.com wrote:
>On Thu, 2017-02-02 at 16:12 +0100, Jiri Pirko wrote:
>> From: Jiri Pirko
>>
>> This introduces a infrastructure for management of linear priority
>> areas. Priority order in an array matters, however order of items inside
>> a priority
Thu, Feb 02, 2017 at 10:40:42PM CET, f.faine...@gmail.com wrote:
>On 02/02/2017 07:12 AM, Jiri Pirko wrote:
>> From: Jiri Pirko
>>
>> This patchset introduces support for offloading TC cls_flower and actions
>> to Spectrum TCAM-base policy engine.
>>
>> The patchset contains patches to allow wor
Thu, Feb 02, 2017 at 11:34:48PM CET, da...@davemloft.net wrote:
>From: Jiri Pirko
>Date: Thu, 2 Feb 2017 16:12:57 +0100
>
>> +static int mlxsw_sp_acl_tcam_region_id_get(struct mlxsw_sp_acl_tcam *tcam,
>> + u16 *p_id)
>> +{
>> +u16 id;
>> +
>> +id = fi
From: Benjamin Serebrin
If the number of virtio queue pairs is not equal to the
number of VCPUs, the virtio guest driver doesn't assign
any CPU affinity for the queue interrupts or the xps
aggregation interrupt.
Google Compute Engine currently provides 1 queue pair for
every VCPU, but limits tha
On Thu, Feb 02, 2017 at 09:31:06PM -0800, William Tu wrote:
>
> Yes, this is auto-generated. We want to use P4 2016 as front end to
> generate ebpf for XDP.
P4 2016 front-end ? is it public? Is there a 2017 version? ;)
just curious.
> >
> > The line 272 is r4 += r2
> > where R4=imm4 and R2=pkt_e
On 2/2/17, 5:50 PM, David Miller wrote:
> I see a lot of "complexity of bridging layer" pushback on these
> changes, and I understand where that is coming from.
>
> But really this isn't even, at a high level, really a bridging change.
>
> What it's doing is making lwtunnel objects more useful.
>
>
On Fri, Feb 03, 2017 at 05:33:45PM +1300, Eric W. Biederman wrote:
>
> The point is that we can make the inode number stable across migration
> and the user space API for namespaces has been designed with that
> possibility in mind.
>
> What you have proposed is the equivalent of reporting a file
Hi,
I am trying to setup macvlans on a centos7 box. The plan is to have
10.1.1.1/24 on the *main* interface, and then a 10.1.1.10/32 on a
macvlan interface. Output below:
2: ens33: mtu 1500 qdisc pfifo_fast
state UP qlen 1000
link/ether 01:0d:26:9d:cd:16 brd ff:ff:ff:ff:ff:ff
inet 10.1.1
On Thu, Feb 2, 2017 at 7:46 PM, Alexei Starovoitov
wrote:
> On Thu, Feb 02, 2017 at 07:26:44PM -0800, William Tu wrote:
>> Thanks. below is my program. The verifier fails at line 272, when
>> writing to ICMP header.
>> -
>> ; ebpf_packetEnd = ((void*)(long)skb->data_end);
>> 206: r2
On Thu, 2017-02-02 at 15:52 -0800, Alexander Duyck wrote:
> On Thu, Feb 2, 2017 at 3:47 PM, Joel Cunningham
> wrote:
> > Hi,
> >
> > I’m studying the synchronization used on different parts of struct
> > net_device and I’m struggling to understand how structure member
> > modifications in dev_i
The function bond_info_query alwarys returns 0. As such, in the function
bond_do_ioctl, it is not necessary to check the returned value. So the
interface type of the function bond_info_query is changed to void. The
redundant check is removed.
Signed-off-by: Zhu Yanjun
---
drivers/net/bonding/bon
From: Eric Dumazet
Date: Wed, 01 Feb 2017 20:47:59 -0800
> From: Eric Dumazet
>
> Debugging issues caused by pfmemalloc is often tedious.
>
> Add a new SNMP counter to more easily diagnose these problems.
>
> Signed-off-by: Eric Dumazet
> Cc: Josef Bacik
Applied.
From: Eric Dumazet
Josef Bacik diagnosed following problem :
I was seeing random disconnects while testing NBD over loopback.
This turned out to be because NBD sets pfmemalloc on it's socket,
however the receiving side is a user space application so does not
have pfmemalloc set on it
Alexei Starovoitov writes:
> On 1/26/17 11:07 AM, Andy Lutomirski wrote:
>> On Thu, Jan 26, 2017 at 10:32 AM, Alexei Starovoitov wrote:
>>> On 1/26/17 10:12 AM, Andy Lutomirski wrote:
On Thu, Jan 26, 2017 at 9:46 AM, Alexei Starovoitov wrote:
>
> On 1/26/17 8:37 AM, Andy Lutom
On Thu, Feb 02, 2017 at 03:21:14PM -0800, John Fastabend wrote:
> Factor out qp assignment.
>
> Signed-off-by: John Fastabend
Acked-by: Michael S. Tsirkin
> ---
> drivers/net/virtio_net.c | 18 +++---
> 1 file changed, 7 insertions(+), 11 deletions(-)
>
> diff --git a/drivers/n
On Thu, Feb 02, 2017 at 03:20:52PM -0800, John Fastabend wrote:
> At this point the do_xdp_prog is mostly if/else branches handling
> the different modes of virtio_net. So remove it and handle running
> the program in the per mode handlers.
>
> Signed-off-by: John Fastabend
This can go in except
On Thu, Feb 02, 2017 at 08:02:55PM -0800, Alexei Starovoitov wrote:
> On Fri, Feb 03, 2017 at 05:42:54AM +0200, Michael S. Tsirkin wrote:
> > On Thu, Feb 02, 2017 at 03:21:57PM -0800, John Fastabend wrote:
> > > Add support for XDP adjust head by allocating a 256B header region
> > > that XDP progr
From: David Ahern
Date: Wed, 1 Feb 2017 18:41:25 -0800
> nothing in devinet.c relies on fib_lookup.h; remove it from the includes
>
> Signed-off-by: David Ahern
Applied.
On Thu, Feb 02, 2017 at 07:16:29PM -0800, John Fastabend wrote:
> Add support for XDP adjust head by allocating a 256B header region
> that XDP programs can grow into. This is only enabled when a XDP
> program is loaded.
>
> In order to ensure that we do not have to unwind queue headroom push
> qu
From: Eric Dumazet
Date: Wed, 01 Feb 2017 18:13:23 -0800
> From: Eric Dumazet
>
> When __alloc_skb() allocates an skb from fast clone cache,
> setting pfmemalloc on the clone is not needed.
>
> Clone will be properly initialized later at skb_clone() time,
> including pfmemalloc field, as it is
On Fri, Feb 03, 2017 at 05:42:54AM +0200, Michael S. Tsirkin wrote:
> On Thu, Feb 02, 2017 at 03:21:57PM -0800, John Fastabend wrote:
> > Add support for XDP adjust head by allocating a 256B header region
> > that XDP programs can grow into. This is only enabled when a XDP
> > program is loaded.
>
On 2/2/17 8:21 PM, Eric W. Biederman wrote:
>
> My brain is mostly elswhere right now so I don't have an implementation
> on how it should be implemented. However Linux fundamentally gets used
> interesting ways, and if we don't implement the option as per mpls exit
> now someone will come along
From: Florian Fainelli
Date: Thu, 2 Feb 2017 19:47:43 -0800
> Le 02/02/17 à 18:54, David Miller a écrit :
>> From: Florian Fainelli
>> Date: Tue, 31 Jan 2017 18:46:43 -0800
>>
>>> From: Mao Wenan
>>>
>>> There is currently no reference count being held on the PHY driver,
>>> which makes it pos
On Thu, 02 Feb 2017 19:14:05 -0800, John Fastabend wrote:
> This series adds adjust head support for virtio. ...
XDP bits look good to me too!
Le 02/02/17 à 18:54, David Miller a écrit :
> From: Florian Fainelli
> Date: Tue, 31 Jan 2017 18:46:43 -0800
>
>> From: Mao Wenan
>>
>> There is currently no reference count being held on the PHY driver,
>> which makes it possible to remove the PHY driver module while the PHY
>> state machine is
On Thu, Feb 02, 2017 at 07:26:44PM -0800, William Tu wrote:
> Thanks. below is my program. The verifier fails at line 272, when
> writing to ICMP header.
> -
> ; ebpf_packetEnd = ((void*)(long)skb->data_end);
> 206: r2 = *(u32 *)(r6 + 4)
> ; ebpf_packetStart = ((void*)(long)skb->data
On Thu, Feb 02, 2017 at 03:21:57PM -0800, John Fastabend wrote:
> Add support for XDP adjust head by allocating a 256B header region
> that XDP programs can grow into. This is only enabled when a XDP
> program is loaded.
>
> In order to ensure that we do not have to unwind queue headroom push
> qu
Shortly after I posted my last series I got access to a more recent
Marvell SDK which had some device tree support for the switch SoCs I'd
been wanting. It was still based on an older kernel but it was a huge
improvement over what came before.
Patch 1/4 is a bit of a cleanup. I did initially strug
Rather than having a separate node for the dfx server add a reg property
to the parent node. This give somes compatibility with the Marvell
supplied SDK.
Signed-off-by: Chris Packham
---
Documentation/devicetree/bindings/net/marvell,prestera.txt | 13 +
arch/arm/boot/dts/armada-xp-98
On 2/2/17 7:14 PM, John Fastabend wrote:
This series adds adjust head support for virtio. The following is my
test setup. I use qemu + virtio as follows,
./x86_64-softmmu/qemu-system-x86_64 \
-hda /var/lib/libvirt/images/Fedora-test0.img \
-m 4096 -enable-kvm -smp 2 -netdev tap,id=hn0,que
Thanks. below is my program. The verifier fails at line 272, when
writing to ICMP header.
-
; ebpf_packetEnd = ((void*)(long)skb->data_end);
206: r2 = *(u32 *)(r6 + 4)
; ebpf_packetStart = ((void*)(long)skb->data);
207: r1 = *(u32 *)(r6 + 0)
...
r10 is "struct hd" at local
Robert Shearman writes:
> On 31/01/17 00:17, Eric W. Biederman wrote:
>> Robert Shearman writes:
>>
>>> It is sometimes desirable to present an MPLS transport network as a
>>> single hop to traffic transiting it because it prevents confusion when
>>> diagnosing failures. An example of where conf
Add support for XDP adjust head by allocating a 256B header region
that XDP programs can grow into. This is only enabled when a XDP
program is loaded.
In order to ensure that we do not have to unwind queue headroom push
queue setup below bpf_prog_add. It reads better to do a prog ref
unwind vs ano
Hi Eric,
[auto build test ERROR on net-next/master]
url:
https://github.com/0day-ci/linux/commits/Eric-Dumazet/net-remove-support-for-per-driver-ndo_busy_poll/20170203-104846
config: x86_64-rhel (attached as .config)
compiler: gcc-6 (Debian 6.2.0-3) 6.2.0 20160901
reproduce:
# save th
For XDP we will need to reset the queues to allow for buffer headroom
to be configured. In order to do this we need to essentially run the
freeze()/restore() code path. Unfortunately the locking requirements
between the freeze/restore and reset paths are different however so
we can not simply reuse
At this point the do_xdp_prog is mostly if/else branches handling
the different modes of virtio_net. So remove it and handle running
the program in the per mode handlers.
Signed-off-by: John Fastabend
---
drivers/net/virtio_net.c | 86 +++---
1 file chan
Factor out qp assignment.
Signed-off-by: John Fastabend
---
drivers/net/virtio_net.c | 20 +++-
1 file changed, 7 insertions(+), 13 deletions(-)
diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
index 3b49363..dba5afb 100644
--- a/drivers/net/virtio_net.c
+++ b/
For XDP use case and to allow ethtool reset tests it is useful to be
able to use reset paths from contexts where rtnl lock is already
held.
This requries updating virtnet_set_queues and free_receive_bufs the
two places where rtnl_lock is taken in virtio_net. To do this we
use the following pattern
This series adds adjust head support for virtio. The following is my
test setup. I use qemu + virtio as follows,
./x86_64-softmmu/qemu-system-x86_64 \
-hda /var/lib/libvirt/images/Fedora-test0.img \
-m 4096 -enable-kvm -smp 2 -netdev tap,id=hn0,queues=4,vhost=on \
-device
virtio-net-pci,ne
From: Andrew Lunn
Date: Thu, 2 Feb 2017 00:46:14 +0100
> Two patches, which have been posted before. Fix simple issues in the
> mv88e6390 support. These don't need to go to stable, since the
> mv88e6390 support in stable is insufficient to be usable.
>
> To apply cleanly, these patches rely on
From: Andrew Lunn
Date: Thu, 2 Feb 2017 00:35:03 +0100
> The 88e1545 PHYs are discrete Marvell PHYs, found in a quad package on
> the zii-devel-b board. Add support for it to the Marvell PHY driver.
>
> Signed-off-by: Andrew Lunn
Applied.
From: Heiner Kallweit
Date: Thu, 2 Feb 2017 08:20:12 +0100
> Most likely a copy & paste error in referenced commit.
> Restore the debug message to what it was before.
>
> Fixes: f573c0b9c4e0 ("stmmac: move stmmac_clk, pclk, clk_ptp_ref and
> stmmac_rst to platform structure")
> Signed-off-by: H
From: Heiner Kallweit
Date: Wed, 1 Feb 2017 22:02:02 +0100
> Chips like Amlogic S905GXBB are supported by this driver but don't
> have support for PTP. Add a separate warning for missing HW support
> to differentiate it from other actual failures.
>
> Signed-off-by: Heiner Kallweit
Applied to
From: Heiner Kallweit
Date: Wed, 1 Feb 2017 20:19:25 +0100
> As documented in Documentation/devicetree/bindings/net/ethernet.txt,
> in RGMII_ID and RGMII_TXID mode the MAC should not add a tx delay.
>
> Signed-off-by: Heiner Kallweit
Applied to net-next, thank you.
From: Andrei Vagin
Date: Wed, 1 Feb 2017 11:00:45 -0800
> This ioctl opens a file to which a socket is bound and
> returns a file descriptor. The caller has to have CAP_NET_ADMIN
> in the socket network namespace.
>
> Currently it is impossible to get a path and a mount point
> for a socket fil
From: Florian Fainelli
Date: Tue, 31 Jan 2017 18:46:43 -0800
> From: Mao Wenan
>
> There is currently no reference count being held on the PHY driver,
> which makes it possible to remove the PHY driver module while the PHY
> state machine is running and polling the PHY. This could cause crashes
On 17-02-02 05:05 PM, Jakub Kicinski wrote:
> On Thu, 02 Feb 2017 15:20:52 -0800, John Fastabend wrote:
>> At this point the do_xdp_prog is mostly if/else branches handling
>> the different modes of virtio_net. So remove it and handle running
>> the program in the per mode handlers.
>>
>> Signed-of
From: Andrew Lunn
Date: Wed, 1 Feb 2017 03:40:04 +0100
> The internal PHYs of the MV88E6390 have a Marvell OUI, but the product
> ID is zero. Work around this by trapping reads to the ID, and if it is
> zero, return the MV88E6390 family ID.
>
> v2: Move the workaround into the central mdio read
From: Nikolay Aleksandrov
Date: Tue, 31 Jan 2017 16:31:58 +0100
> @@ -197,7 +197,8 @@ int br_handle_frame_finish(struct net *net, struct sock
> *sk, struct sk_buff *skb
> if (dst->is_local)
> return br_pass_frame_up(skb);
>
> - dst->used = jiffie
From: Eric Dumazet
We added generic support for busy polling in NAPI layer in linux-4.5
No network driver uses ndo_busy_poll() anymore, we can get rid
of the pointer in struct net_device_ops, and its use in sk_busy_loop()
Saves NETIF_F_BUSY_POLL features bit.
Signed-off-by: Eric Dumazet
---
From: Martin KaFai Lau
Date: Tue, 31 Jan 2017 22:35:31 -0800
> This patchset fixes misc bugs after reinitializing
> queues (e.g. by ethtool -L).
>
> v2:
> * Add another fix to mem leak in tx_ring[t] and tx_cq[t]
> * In mlx4_en_try_alloc_resources(),
> move all xdp_prog logic after calling mlx4
From: Eric Dumazet
In linux-4.5, busy polling was implemented in core
NAPI stack, meaning that all custom implementation can
be removed from drivers.
Not only we remove lot's of tricky code, we also remove
one lock operation in fast path.
Signed-off-by: Eric Dumazet
Cc: Edward Cree
Cc: Bert K
> On Feb 2, 2017, at 5:52 PM, Alexander Duyck wrote:
>
> On Thu, Feb 2, 2017 at 3:47 PM, Joel Cunningham
> wrote:
>> Hi,
>>
>> I’m studying the synchronization used on different parts of struct
>> net_device and I’m struggling to understand how structure member
>> modifications in dev_ioctl
On Fri, 2017-02-03 at 10:18 +0900, Lorenzo Colitti wrote:
> On Fri, Feb 3, 2017 at 9:31 AM, Eric Dumazet wrote:
> >> It should be safe to call sock_net_uid on any type of socket
> >> (including NULL). sk_uid was added to struct sock in 86741ec25462
> >> ("net: core: Add a UID field to struct sock.
Julian, it looks some tweaks are still needed for the IPSEC cases.
And thus there will be at least one more repsin.
Thanks.
I see a lot of "complexity of bridging layer" pushback on these
changes, and I understand where that is coming from.
But really this isn't even, at a high level, really a bridging change.
What it's doing is making lwtunnel objects more useful.
Now that we have lightweight tunnels and netdevs, w
On Fri, Feb 3, 2017 at 9:31 AM, Eric Dumazet wrote:
>> It should be safe to call sock_net_uid on any type of socket
>> (including NULL). sk_uid was added to struct sock in 86741ec25462
>> ("net: core: Add a UID field to struct sock.")
>
> But a request socket or a timewait socket do not have this
From: Eric Dumazet
In linux-4.5, busy polling was implemented in core
NAPI stack, meaning that all custom implementation can
be removed from drivers.
Not only we remove lot's of tricky code, we also remove
one lock operation in fast path.
Signed-off-by: Eric Dumazet
Cc: Edward Cree
Cc: Bert K
Avoid triggering change events for setting conntrack mark or labels
before the conntrack entry has been confirmed. Refactoring on this
patch also makes chenges in later patches easier to review.
Fixes: 182e3042e15d ("openvswitch: Allow matching on conntrack mark")
Fixes: c2ac66735870 ("openvswitc
Add the fields of the conntrack original direction 5-tuple to struct
sw_flow_key. The new fields are initially zeroed, and are populated
whenever a conntrack action is executed and either finds or generates
a conntrack entry. This means that these fields exist for all packets
were not rejected by
Make the array of labels in struct ovs_key_ct_label an union, adding a
u32 array of the same byte size as the existing u8 array. It is
faster to loop through the labels 32 bits at the time, which is also
the alignment of netlink attributes.
Signed-off-by: Jarno Rajahalme
---
include/uapi/linux/
Stateful network admission policy may allow connections to one
direction and reject connections initiated in the other direction.
After policy change it is possible that for a new connection an
overlapping conntrack entry already exist, where the connection
original direction is opposed to the new
We avoid calling into nf_conntrack_in() for expected connections, as
that would remove the expectation that we want to stick around until
we are ready to commit the connection. Instead, we do a lookup in the
expectation table directly. However, after a successful expectation
lookup we have set th
When looking for an existing conntrack entry, the packet 5-tuple
must be inverted if NAT has already been applied, as the current
packet headers do not match any conntrack tuple. For
example, if a packet from private address X to a public address B is
source-NATted to A, the conntrack entry will h
struct sw_flow_key has two 16-bit holes. Move the most matched
conntrack match fields there. In some typical cases this reduces the
size of the key that needs to be hashed into half and into one cache
line.
Signed-off-by: Jarno Rajahalme
---
net/openvswitch/conntrack.c| 42 +
On Thu, 02 Feb 2017 15:20:52 -0800, John Fastabend wrote:
> At this point the do_xdp_prog is mostly if/else branches handling
> the different modes of virtio_net. So remove it and handle running
> the program in the per mode handlers.
>
> Signed-off-by: John Fastabend
I think you rebased away Da
From: Eric Dumazet
In linux-4.5, busy polling was implemented in core
NAPI stack, meaning that all custom implementation can
be removed from drivers.
Not only we remove lot's of code, we also remove one lock
operation in fast path, and allow GRO to do its job.
Signed-off-by: Eric Dumazet
Cc: J
On Thu, Feb 2, 2017 at 4:22 PM, Jakub Kicinski wrote:
> On Thu, 2 Feb 2017 15:40:19 -0800, Michael Chan wrote:
>> On Thu, Feb 2, 2017 at 2:56 PM, Jakub Kicinski wrote:
>> > (b) why
>> > do you encode the two parameters in a single u32? It's the seventh
>> > parameter so it's going on the stack a
On Fri, 2017-02-03 at 09:00 +0900, Lorenzo Colitti wrote:
> On Fri, Feb 3, 2017 at 6:32 AM, Daniel Borkmann wrote:
> >> + sk = skb->sk;
> >> + kuid = sock_net_uid(dev_net(skb->dev), sk);
> >
> >
> > Don't you need to test for fullsock? Do you mean something like below?
>
> It should b
On 02/03/2017 01:00 AM, Lorenzo Colitti wrote:
On Fri, Feb 3, 2017 at 6:32 AM, Daniel Borkmann wrote:
+ sk = skb->sk;
+ kuid = sock_net_uid(dev_net(skb->dev), sk);
Don't you need to test for fullsock? Do you mean something like below?
It should be safe to call sock_net_uid on an
From: Eric Dumazet
In linux-4.5, busy polling was implemented in core
NAPI stack, meaning that all custom implementation can
be removed from drivers.
Not only we remove lot's of code, we also remove one lock
operation in fast path, and allow GRO to do its job.
Signed-off-by: Eric Dumazet
Cc: J
On Thu, 2 Feb 2017 15:40:19 -0800, Michael Chan wrote:
> On Thu, Feb 2, 2017 at 2:56 PM, Jakub Kicinski wrote:
> > On Thu, 2 Feb 2017 11:55:29 -0500, Michael Chan wrote:
> >> @@ -755,8 +757,8 @@ static void bnxt_reuse_rx_agg_bufs(struct bnxt_napi
> >> *bnapi, u16 cp_cons,
> >>
> >> static str
On Thu, Feb 02, 2017 at 12:59:48PM -0800, Chenbo Feng wrote:
> Introduce two eBpf helper function to get the socket cookie and
> socket uid for each packet. The helper function is useful when
> the *sk field inside sk_buff is not empty.
Looks nice.
I'd like to see a sample code/test for it though.
On Fri, Feb 3, 2017 at 6:32 AM, Daniel Borkmann wrote:
>> + sk = skb->sk;
>> + kuid = sock_net_uid(dev_net(skb->dev), sk);
>
>
> Don't you need to test for fullsock? Do you mean something like below?
It should be safe to call sock_net_uid on any type of socket
(including NULL). sk_uid
On Thu, Feb 02, 2017 at 03:08:27PM -0800, Jakub Kicinski wrote:
> > + rcu_read_lock();
> > + act = bpf_prog_run_xdp(xdp_prog, &xdp);
> > + rcu_read_unlock();
> > +
> > + if (orig_data != xdp.data) {
> > + offset = xdp.data - xdp.data_hard_start;
> > + len = xdp.data_end
1 - 100 of 306 matches
Mail list logo
