On Fri, Apr 1, 2016 at 7:16 PM, David Miller wrote:
> From: Alexander Duyck
> Date: Fri, 1 Apr 2016 12:58:41 -0700
>
>> RFC 6864 is pretty explicit about this, IPv4 ID used only for
>> fragmentation. https://tools.ietf.org/html/rfc6864#section-4.1
>>
>> The goal with this change is to try and ke
On 4/1/16 6:14 PM, Leon Romanovsky wrote:
On Fri, Apr 01, 2016 at 12:47:24PM -0700, santosh shilimkar wrote:
(cc-ing netdev)
On 3/30/2016 7:59 PM, Wengang Wang wrote:
在 2016年03月31日 09:51, Wengang Wang 写道:
在 2016年03月31日 01:16, santosh shilimkar 写道:
Hi Wengang,
On 3/30/2016 9:19 AM, Leon
On Fri, Apr 01, 2016 at 07:08:31PM -0700, Eric Dumazet wrote:
> On Fri, 2016-04-01 at 18:21 -0700, Brenden Blanco wrote:
> > Add support for the BPF_PROG_TYPE_PHYS_DEV hook in mlx4 driver. Since
> > bpf programs require a skb context to navigate the packet, build a
> > percpu fake skb with the min
On Sat, 2016-04-02 at 10:19 +0800, Herbert Xu wrote:
> On Fri, Apr 01, 2016 at 07:15:33PM -0700, Eric Dumazet wrote:
> > On Sat, 2016-04-02 at 09:57 +0800, Herbert Xu wrote:
> > >
> > > We could easily fix that by adding a feature bit to control this,
> > > something like SKB_GSO_TCP_FIXEDID.
> >
Hi,
Sorry for digging up an old patch, but... ;-)
dev_disable_lro() is a leftover from ancient times. If you read commit
27660515a,
there is a hint where it should go. Please, read on if you'd like to
fix this properly.
2015-11-03 3:55 GMT+01:00 Jarod Wilson :
> There are some netdev features, w
On Fri, 2016-04-01 at 22:16 -0400, David Miller wrote:
> From: Alexander Duyck
> Date: Fri, 1 Apr 2016 12:58:41 -0700
>
> > RFC 6864 is pretty explicit about this, IPv4 ID used only for
> > fragmentation. https://tools.ietf.org/html/rfc6864#section-4.1
> >
> > The goal with this change is to tr
On Fri, Apr 01, 2016 at 07:15:33PM -0700, Eric Dumazet wrote:
> On Sat, 2016-04-02 at 09:57 +0800, Herbert Xu wrote:
> >
> > We could easily fix that by adding a feature bit to control this,
> > something like SKB_GSO_TCP_FIXEDID.
>
> I understood the patch allowed to aggregate 4 segments having
From: Alexander Duyck
Date: Fri, 1 Apr 2016 12:58:41 -0700
> RFC 6864 is pretty explicit about this, IPv4 ID used only for
> fragmentation. https://tools.ietf.org/html/rfc6864#section-4.1
>
> The goal with this change is to try and keep most of the existing
> behavior in tact without violating
On Sat, 2016-04-02 at 09:57 +0800, Herbert Xu wrote:
> Eric Dumazet wrote:
> >
> > I do not particularly care, but it is worth mentioning that GRO+TSO
> > would not be idempotent anymore.
>
> We could easily fix that by adding a feature bit to control this,
> something like SKB_GSO_TCP_FIXEDID.
On Fri, 2016-04-01 at 18:21 -0700, Brenden Blanco wrote:
> Add support for the BPF_PROG_TYPE_PHYS_DEV hook in mlx4 driver. Since
> bpf programs require a skb context to navigate the packet, build a
> percpu fake skb with the minimal fields. This avoids the costly
> allocation for packets that end
On Fri, Apr 01, 2016 at 05:21:31PM +0800, Lijun Ou wrote:
> The driver for HiSilicon RoCE is a platform driver.
> The driver will support multiple versions of hardware. Currently only "v1"
> for hip06 SoC is supported.
> The driver includes two parts: common driver and hardware-specific
> operation
Eric Dumazet wrote:
>
> I do not particularly care, but it is worth mentioning that GRO+TSO
> would not be idempotent anymore.
We could easily fix that by adding a feature bit to control this,
something like SKB_GSO_TCP_FIXEDID.
Cheers,
--
Email: Herbert Xu
Home Page: http://gondor.apana.org.a
On Fri, Apr 01, 2016 at 11:34:10PM +0200, Johannes Berg wrote:
>
> I was thinking about that one - it's not obvious to me from the code
> how this "explicitly checking for dups" would be done or let's say how
> rhashtable differentiates. But since it seems to work for Ben until
> hitting a certain
Sets the bpf program represented by fd as an early filter in the rx path
of the netdev. The fd must have been created as BPF_PROG_TYPE_PHYS_DEV.
Providing a negative value as fd clears the program. Getting the fd back
via rtnl is not possible, therefore reading of this value merely
provides a bool
Add support for the BPF_PROG_TYPE_PHYS_DEV hook in mlx4 driver. Since
bpf programs require a skb context to navigate the packet, build a
percpu fake skb with the minimal fields. This avoids the costly
allocation for packets that end up being dropped.
Since mlx4 is so far the only user of this pse
This patch set introduces new infrastructure for programmatically
processing packets in the earliest stages of rx, as part of an effort
others are calling Express Data Path (XDP) [1]. Start this effort by
introducing a new bpf program type for early packet filtering, before even
an skb has been all
Add a sample program that only drops packets at the
BPF_PROG_TYPE_PHYS_DEV hook of a link. With the drop-only program,
observed single core rate is ~14.6Mpps.
Other tests were run, for instance without the dropcnt increment or
without reading from the packet header, the packet rate was mostly
unch
Add a new netdev op for drivers implementing the BPF_PROG_TYPE_PHYS_DEV
filter to get configuration. Since the fd is only used by the driver to
fetch the prog, the netdev should just keep a bit to indicate the
program is valid.
Signed-off-by: Brenden Blanco
---
include/linux/netdevice.h | 8 +++
Add a new bpf prog type that is intended to run in early stages of the
packet rx path. Only minimal packet metadata will be available, hence a new
context type, struct xdp_metadata, is exposed to userspace. So far only
expose the readable packet length, and only in read mode.
The PHYS_DEV name is
On Fri, Apr 01, 2016 at 12:47:24PM -0700, santosh shilimkar wrote:
> (cc-ing netdev)
> On 3/30/2016 7:59 PM, Wengang Wang wrote:
> >
> >
> >在 2016年03月31日 09:51, Wengang Wang 写道:
> >>
> >>
> >>在 2016年03月31日 01:16, santosh shilimkar 写道:
> >>>Hi Wengang,
> >>>
> >>>On 3/30/2016 9:19 AM, Leon Romanovsk
1) Missing device reference in IPSEC input path results in crashes during
device unregistration. From Subash Abhinov Kasiviswanathan.
2) Per-queue ISR register writes not being done properly in macb driver,
from Cyrille Pitchen.
3) Stats accounting bugs in bcmgenet, from Patri Gynther.
4
On 04/01/2016 11:28 PM, Daniel Borkmann wrote:
On 04/01/2016 09:00 PM, David Miller wrote:
From: Daniel Borkmann
Date: Fri, 1 Apr 2016 11:41:03 +0200
Moreover, I noticed that when in the non-error path the __skb_pull()
is done and the original offset to mac header was non-zero, we fixup
from
On Fri, Apr 1, 2016 at 4:13 PM, Cong Wang wrote:
> On Fri, Apr 1, 2016 at 3:56 PM, Martin KaFai Lau wrote:
>> + bh_lock_sock(sk);
>> + if (!sock_owned_by_user(sk))
>> + ip6_datagram_dst_update(sk, false);
>> + bh_unlock_sock(sk);
>
>
> My discussion with Eric shows
On Fri, Apr 1, 2016 at 3:56 PM, Martin KaFai Lau wrote:
> + bh_lock_sock(sk);
> + if (!sock_owned_by_user(sk))
> + ip6_datagram_dst_update(sk, false);
> + bh_unlock_sock(sk);
My discussion with Eric shows that we probably don't need to hold
this sock lock here, an
This patch moves the route lookup and update codes for connected
datagram sk to a newly created function ip6_datagram_dst_update()
It will be reused during the pmtu update in the later patch.
Signed-off-by: Martin KaFai Lau
Cc: Cong Wang
Cc: Eric Dumazet
Cc: Wei Wang
---
net/ipv6/datagram.c
This patch adds a release_cb for UDPv6. It does a route lookup
and updates sk->sk_dst_cache if it is needed. It picks up the
left-over job from ip6_sk_update_pmtu() if the sk is owned
by user during the pmtu update.
Signed-off-by: Martin KaFai Lau
Reported-by: Wei Wang
Cc: Cong Wang
Cc: Eric
Move flowi6 init codes for connected datagram sk to a newly created
function ip6_datagram_flow_key_init().
It will be reused during pmtu update in the later patch.
Signed-off-by: Martin KaFai Lau
Cc: Cong Wang
Cc: Eric Dumazet
Cc: Wei Wang
---
net/ipv6/datagram.c | 50 +++
There is a case in connected UDP socket such that
getsockopt(IPV6_MTU) will return a stale MTU value. The reproducible
sequence could be the following:
1. Create a connected UDP socket
2. Send some datagrams out
3. Receive a ICMPV6_PKT_TOOBIG
4. No new outgoing datagrams to trigger the sk_dst_check
There is a case in connected UDP socket such that
getsockopt(IPV6_MTU) will return a stale MTU value. The reproducible
sequence could be the following:
1. Create a connected UDP socket
2. Send some datagrams out
3. Receive a ICMPV6_PKT_TOOBIG
4. No new outgoing datagrams to trigger the sk_dst_check
On Fri, 2016-04-01 at 11:07 +0800, Yangbo Lu wrote:
> This patchset is used to fix a host version register bug in the T4240-R1.0
> -R2.0
> eSDHC controller. To get the SoC version and revision, it's needed to add
> the
> GUTS driver to access the global utilities registers.
>
> So, the first three
Enable RX Checksum offload feature in the ibmvnic driver.
Signed-off-by: Thomas Falcon
Cc: John Allen
---
v4: this patch included since it is enabled by the previous patch
---
drivers/net/ethernet/ibm/ibmvnic.c | 4
1 file changed, 4 insertions(+)
diff --git a/drivers/net/ethernet/ibm/ibm
Allow the VNIC driver to provide descriptors containing
L2/L3/L4 headers to firmware. This feature is needed
for greater hardware compatibility and enablement of checksum
and TCP offloading features.
A new function is included for the hypervisor call,
H_SEND_SUBCRQ_INDIRECT, allowing a DMA-mapped
On 4/1/2016 1:01 PM, Johannes Berg wrote:
On Fri, 2016-04-01 at 10:51 -0700, santosh shilimkar wrote:
Hi Dave,
I noticed something odd while checking the recent
commits of mine in kernel.org tree made it via netdev.
Don't know if its patchwork tool doing this.
Usual author line in my git objec
On Fri, 2016-04-01 at 08:46 +0800, Herbert Xu wrote:
> On Thu, Mar 31, 2016 at 05:29:59PM +0200, Johannes Berg wrote:
> >
> >
> > Does removing this completely disable the "-EEXIST" error? I can't
> > say
> > I fully understand the elasticity stuff in
> > __rhashtable_insert_fast().
> What EEXIST
On 04/01/2016 09:00 PM, David Miller wrote:
From: Daniel Borkmann
Date: Fri, 1 Apr 2016 11:41:03 +0200
Moreover, I noticed that when in the non-error path the __skb_pull()
is done and the original offset to mac header was non-zero, we fixup
from a wrong skb->data offset in the checksum comple
FW reconfiguration in .ndo_open()/.ndo_stop() should reset/
restore queue state. Since we need IRQs to be disabled when
filling rings on RX path we have to move disable_irq() from
.ndo_open() all the way up to IRQ allocation.
nfp_net_start_vec() becomes trivial now so it's inlined.
Signed-off-by
We want the .ndo_open() to have following structure:
- allocate resources;
- configure HW/FW;
- enable the device from stack perspective.
Therefore filling RX rings needs to be moved to the beginning
of .ndo_open().
Signed-off-by: Jakub Kicinski
---
.../net/ethernet/netronome/nfp/nfp_net_comm
| For transmit we can leave the IP ID code as is. For receive we should not be
| snooping into the IP ID for any frames that have the DF bit set as devices
| that have adopted RFC 6864 on their transmit path will end up causing issues.
Currently, GRO does not coalesce TCP packets originating f
We need to be able to disable the link state interrupt when
the device is brought down. We used to just free the IRQ
at the beginning of .ndo_stop(). As we now move towards
more ordered .ndo_open()/.ndo_stop() paths LSC allocation
should be placed in the "allocate resource" section.
Since the IR
nfp_net_{alloc|free}_rings contained strange mix of allocations
and vector initialization. Remove it, declare vector init as
a separate function and handle allocations explicitly.
Signed-off-by: Jakub Kicinski
---
.../net/ethernet/netronome/nfp/nfp_net_common.c| 126 -
1
nfp_net_[rt]x_ring_{alloc,free} should only allocate or free
ring resources without touching the device. Move setting
parameters in the BAR to separate functions. This will make
it possible to reuse alloc/free functions to allocate new
rings while the device is running.
Signed-off-by: Jakub Kici
Since much of the required changes have already been made for
changing MTU at runtime let's use it for ring size changes as
well.
Signed-off-by: Jakub Kicinski
---
drivers/net/ethernet/netronome/nfp/nfp_net.h | 1 +
.../net/ethernet/netronome/nfp/nfp_net_common.c| 126 +++
To be able to switch rings more easly on config changes allocate
them dynamically, separately from nfp_net structure.
Signed-off-by: Jakub Kicinski
---
drivers/net/ethernet/netronome/nfp/nfp_net.h | 6 ++---
.../net/ethernet/netronome/nfp/nfp_net_common.c| 28 +-
2
Free list buffer size needs to be propagated to few functions
as a parameter and added to struct nfp_net_rx_ring since soon
some of the functions will be reused to manage rings with
buffers of size different than nn->fl_bufsz.
Signed-off-by: Jakub Kicinski
---
drivers/net/ethernet/netronome/nfp/
When changing MTU on running device first allocate new rings
and buffers and once it succeeds proceed with changing MTU.
Allocation of new rings is not really necessary for this
operation - it's done to keep the code simple and because
size of the extra ring memory is quite small compared to
the s
Separate allocation of buffers from giving them to FW,
thanks to this it will be possible to move allocation
earlier on .ndo_open() path and reuse buffers during
runtime reconfiguration.
Similar to TX side clean up the spill of functionality
from flush to freeing the ring. Unlike on TX side,
RX r
Hi!
Sorry it takes me so long to iterate this.
Previous series included some not entirely related patches,
this one is cut down. Main issue I'm trying to solve here
is that .ndo_change_mtu() in nfpvf driver is doing full
close/open to reallocate buffers - which if open fails
can result in device
nfp_net_[rt]x_ring_init functions used to be called from probe
path only and some of their functionality was spilled to the
call site. In order to reuse them for ring reconfiguration
we need them to do all the init.
Signed-off-by: Jakub Kicinski
---
.../net/ethernet/netronome/nfp/nfp_net_common
Divide .ndo_open() and .ndo_stop() into logical, callable
chunks. No functional changes.
Signed-off-by: Jakub Kicinski
---
.../net/ethernet/netronome/nfp/nfp_net_common.c| 218 +
1 file changed, 136 insertions(+), 82 deletions(-)
diff --git a/drivers/net/ethernet/netron
Since we never used flush without freeing the ring later
the functionality of the two operations is mixed.
Rename flush to ring reset and move there all the things
which have to be done after FW ring state is cleared.
While at it do some clean-ups.
Signed-off-by: Jakub Kicinski
---
.../net/ether
Soon ring resize will call this functions with values
different than the current configuration we need to
explicitly pass the ring count as parameter.
Signed-off-by: Jakub Kicinski
---
.../net/ethernet/netronome/nfp/nfp_net_common.c| 23 +-
1 file changed, 14 insertions(+
When calculating the RX buffer length we need to account for
up to 2 VLAN tags and up to 8 MPLS labels. Rounding up to 1k
is an relic of a distant past and can be removed. While at
it also remove trivial print statement.
Signed-off-by: Jakub Kicinski
---
drivers/net/ethernet/netronome/nfp/nfp_
On Fri, 2016-04-01 at 15:58 -0400, Aaron Conole wrote:
> From: Aaron Conole
>
> When signaling that a GRO frame is ready to be processed, the network stack
> correctly checks length and aborts processing when a frame is less than 14
> bytes. However, such a condition is really indicative of a bro
When creating an ip6tnl tunnel with ip tunnel, rtnl_link_ops is not set
before ip6_tnl_create2 is called. When register_netdevice is called, there
is no linkinfo attribute in the NEWLINK message because of that.
Setting rtnl_link_ops before calling register_netdevice fixes that.
Signed-off-by: Th
On Fri, 2016-04-01 at 10:51 -0700, santosh shilimkar wrote:
> Hi Dave,
>
> I noticed something odd while checking the recent
> commits of mine in kernel.org tree made it via netdev.
>
> Don't know if its patchwork tool doing this.
> Usual author line in my git objects :
> Author: Santosh
From: Aaron Conole
When signaling that a GRO frame is ready to be processed, the network stack
correctly checks length and aborts processing when a frame is less than 14
bytes. However, such a condition is really indicative of a broken driver,
and should be loudly signaled, rather than silently d
On Fri, Apr 1, 2016 at 12:24 PM, David Miller wrote:
> From: Eric Dumazet
> Date: Fri, 01 Apr 2016 11:49:03 -0700
>
>> For example, TCP stack tracks per socket ID generation, even if it
>> sends DF=1 frames. Damn useful for tcpdump analysis and drop
>> inference.
>
> Thanks for mentioning this, I
Hello,
On Fri, 1 Apr 2016, David Ahern wrote:
> v3
> - Julian comments: changed use of dead in documentation to failed,
> init state to NUD_REACHABLE which simplifies fib_good_nh, use of
> nh_dev for neighbor lookup, fallback to first entry which is what
> current logic does
>
> v
(cc-ing netdev)
On 3/30/2016 7:59 PM, Wengang Wang wrote:
在 2016年03月31日 09:51, Wengang Wang 写道:
在 2016年03月31日 01:16, santosh shilimkar 写道:
Hi Wengang,
On 3/30/2016 9:19 AM, Leon Romanovsky wrote:
On Wed, Mar 30, 2016 at 05:08:22PM +0800, Wengang Wang wrote:
Problem is found that some amo
The following series adds the ability for a hypervisor to set an MTU on the
guest during feature negotiation phase. This is useful for VM orchestration
when, for instance, tunneling is involved and the MTU of the various systems
should be homogenous.
The first patch adds the feature bit as describ
This patch checks the feature bit for the VIRTIO_NET_F_MTU feature. If it
exists, read the advised MTU and use it.
No proper error handling is provided for the case where a user changes the
negotiated MTU. A future commit will add proper error handling. Instead, a
warning is emitted if the guest c
This commit adds the feature bit and associated mtu device entry for the
virtio network device. Future commits will make use of these bits to support
negotiated MTU.
Signed-off-by: Aaron Conole
---
v2,v3:
* No change
include/uapi/linux/virtio_net.h | 3 +++
1 file changed, 3 insertions(+)
diff
From: Eric Dumazet
Date: Fri, 01 Apr 2016 11:49:03 -0700
> For example, TCP stack tracks per socket ID generation, even if it
> sends DF=1 frames. Damn useful for tcpdump analysis and drop
> inference.
Thanks for mentioning this, I never considered this use case.
> With your change, the resulti
From: Marcin Wojtas
Date: Fri, 1 Apr 2016 15:21:18 +0200
> After enabling per-cpu processing it appeared that under heavy load
> changing MTU can result in blocking all port's interrupts and transmitting
> data is not possible after the change.
>
> This commit fixes above issue by disabling per
On 03/31/2016 08:07 PM, Yangbo Lu wrote:
> drivers/clk/clk-qoriq.c | 3 +--
>
For clk part:
Acked-by: Stephen Boyd
--
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
From: Daniel Borkmann
Date: Fri, 1 Apr 2016 11:41:03 +0200
> Moreover, I noticed that when in the non-error path the __skb_pull()
> is done and the original offset to mac header was non-zero, we fixup
> from a wrong skb->data offset in the checksum complete processing.
>
> So the skb_postpush_r
On Fri, 2016-04-01 at 11:05 -0700, Alexander Duyck wrote:
> RFC 6864 states that the IPv4 ID field MUST NOT be used for purposes other
> than fragmentation and reassembly. Currently we are looking at this field
> as a way of identifying what frames can be aggregated and which cannot for
> GRO. W
From: Jisheng Zhang
Date: Fri, 1 Apr 2016 17:11:05 +0800
> L1_CACHE_BYTES may not be the real cacheline size, use cache_line_size
> to determine the cacheline size in runtime.
>
> Signed-off-by: Jisheng Zhang
> Suggested-by: Marcin Wojtas
Applied.
From: roy.qing...@gmail.com
Date: Fri, 1 Apr 2016 17:26:58 +0800
> From: Li RongQing
>
> 1. nothing of idev is changed, so read lock is enough
> 2. ifp is changed, so used ifp->lock or cmpxchg to protect it
>
> Signed-off-by: Li RongQing
You posted this patch twice and didn't post patch 2/2.
From: Jisheng Zhang
Date: Fri, 1 Apr 2016 17:12:49 +0800
> L1_CACHE_BYTES may not be the real cacheline size, use cache_line_size
> to determine the cacheline size in runtime.
>
> Signed-off-by: Jisheng Zhang
> Suggested-by: Marcin Wojtas
Applied.
From: roy.qing...@gmail.com
Date: Fri, 1 Apr 2016 16:16:10 +0800
> From: Li RongQing
>
> remove br_dev_set_multicast_list which does nothing
>
> Signed-off-by: Li RongQing
This will break SIOCADDMULTI et al. on the bridge, see net/core/dev.c
which checks whether this ndo OP is NULL or not.
From: Giuseppe Cavallaro
Date: Fri, 1 Apr 2016 09:07:13 +0200
> This patch series is to fix the problems below and recently debugged
> in this mailing list:
>
> o to fix a problem for the HW where the normal descriptor
> o to fix the mdio registration according to the different
> platform conf
From: Jisheng Zhang
Date: Thu, 31 Mar 2016 17:01:23 +0800
> This is to fix the following maybe-uninitialized warning:
>
> drivers/net/ethernet/marvell/mvpp2.c:6007:18: warning: 'err' may be
> used uninitialized in this function [-Wmaybe-uninitialized]
>
> Signed-off-by: Jisheng Zhang
Applied.
On 04/01/2016 08:33 PM, David Miller wrote:
From: Daniel Borkmann
Date: Fri, 01 Apr 2016 10:10:11 +0200
Dave, do you need me to resubmit this one w/o changes:
http://patchwork.ozlabs.org/patch/603903/ ?
I'll apply this and queue it up for -stable, thanks.
Ok, thanks!
On 04/01/2016 08:10 PM, Alexei Starovoitov wrote:
On 4/1/16 2:58 AM, Naveen N. Rao wrote:
PPC64 eBPF JIT compiler. Works for both ABIv1 and ABIv2.
Enable with:
echo 1 > /proc/sys/net/core/bpf_jit_enable
or
echo 2 > /proc/sys/net/core/bpf_jit_enable
... to see the generated JIT code. This can f
From: Daniel Borkmann
Date: Fri, 01 Apr 2016 10:10:11 +0200
> Dave, do you need me to resubmit this one w/o changes:
> http://patchwork.ozlabs.org/patch/603903/ ?
I'll apply this and queue it up for -stable, thanks.
>From a5a7a9828511ff6a522cf742109768207ff89929 Mon Sep 17 00:00:00 2001
From: Charles-Antoine Couret
Date: Fri, 1 Apr 2016 16:16:35 +0200
Subject: [PATCH] Marvell phy: add fiber status check for some components
This patch is not tested with all Marvell's phy. The new function was actived
only for
On 03/31/2016 05:46 PM, Herbert Xu wrote:
On Thu, Mar 31, 2016 at 05:29:59PM +0200, Johannes Berg wrote:
Does removing this completely disable the "-EEXIST" error? I can't say
I fully understand the elasticity stuff in __rhashtable_insert_fast().
What EEXIST error are you talking about? The o
On 4/1/16 2:58 AM, Naveen N. Rao wrote:
PPC64 eBPF JIT compiler. Works for both ABIv1 and ABIv2.
Enable with:
echo 1 > /proc/sys/net/core/bpf_jit_enable
or
echo 2 > /proc/sys/net/core/bpf_jit_enable
... to see the generated JIT code. This can further be processed with
tools/net/bpf_jit_disasm.
RFC 6864 states that the IPv4 ID field MUST NOT be used for purposes other
than fragmentation and reassembly. Currently we are looking at this field
as a way of identifying what frames can be aggregated and which cannot for
GRO. While this is valid for frames that do not have DF set, it is inval
This patch fixes an issue I found in which we were dropping frames if we
had enabled checksums on GRE headers that were encapsulated by either FOU
or GUE. Without this patch I was barely able to get 1 Gb/s of throughput.
With this patch applied I am now at least getting around 6 Gb/s.
The issue i
This pair of patches addresses a few issues I have discovered over the last
week or so concerning GRO and GRE tunnels.
The first patch addresses an item I called out as an issue with FOU/GUE
encapsulating GRE, and I finally had a chance to test it and verify that
the code concerning it was broken
On 4/1/16 7:41 AM, Naveen N. Rao wrote:
On 2016/03/31 10:52AM, Alexei Starovoitov wrote:
On 3/31/16 4:25 AM, Naveen N. Rao wrote:
...
+
+#ifdef __powerpc__
+#define BPF_KPROBE_READ_RET_IP(ip, ctx){ (ip) = (ctx)->link; }
+#define BPF_KRETPROBE_READ_RET_IP(ip, ctx) BPF_KPROBE_
On 4/1/16 7:37 AM, Naveen N. Rao wrote:
On 2016/03/31 08:19PM, Daniel Borkmann wrote:
On 03/31/2016 07:46 PM, Alexei Starovoitov wrote:
On 3/31/16 4:25 AM, Naveen N. Rao wrote:
clang $(NOSTDINC_FLAGS) $(LINUXINCLUDE) $(EXTRA_CFLAGS) \
-D__KERNEL__ -D__ASM_SYSREG_H -Wno-unused-v
Hi Dave,
I noticed something odd while checking the recent
commits of mine in kernel.org tree made it via netdev.
Don't know if its patchwork tool doing this.
Usual author line in my git objects :
Author: Santosh Shilimkar
But the commits going via your tree seems to be like below.
Use list_* helpers in sctp_list_dequeue, more readable.
Signed-off-by: Marcelo Ricardo Leitner
---
v2: patch rechecked
include/net/sctp/sctp.h | 6 ++
1 file changed, 2 insertions(+), 4 deletions(-)
diff --git a/include/net/sctp/sctp.h b/include/net/sctp/sctp.h
index
65521cfdcadeee35d61f2
2016-03-31 19:19 GMT-07:00 David Miller :
> From: Michael Ma
> Date: Thu, 31 Mar 2016 16:48:43 -0700
>
>> I didn't really know that multiple qdiscs can be isolated using MQ so
> ...
>
> Please stop top-posting.
Sorry that I wasn't aware of this...
On Fri, Apr 01, 2016 at 06:33:48PM +0200, Charles-Antoine Couret wrote:
> >From a5a7a9828511ff6a522cf742109768207ff89929 Mon Sep 17 00:00:00 2001
> From: Charles-Antoine Couret
> Date: Fri, 1 Apr 2016 16:16:35 +0200
> Subject: [PATCH] Marvell phy: add fiber status check for some components
>
> T
There is no point on delaying the packet if we can't fit a single byte
of data on it anymore. So lets just reduce the threshold by the amount
that a data chunk with 4 bytes (rounding) would use.
v2: based on the right tree
Signed-off-by: Marcelo Ricardo Leitner
---
net/sctp/output.c | 3 ++-
1
2016-04-01 4:30 GMT+09:00 David Miller :
> From: Akinobu Mita
> Date: Thu, 31 Mar 2016 01:38:39 +0900
>
>> + struct sk_buff_head tx_queue;
>
> The way the queueing works in this driver is that it is only possible
> to have one SKB being transmitted at one time.
>
> This is evident by how the d
>From a5a7a9828511ff6a522cf742109768207ff89929 Mon Sep 17 00:00:00 2001
From: Charles-Antoine Couret
Date: Fri, 1 Apr 2016 16:16:35 +0200
Subject: [PATCH] Marvell phy: add fiber status check for some components
This patch is not tested with all Marvell's phy. The new function was actived
only for
On Fri, Apr 1, 2016 at 11:52 AM, Eric Dumazet wrote:
> Attackers like to use SYNFLOOD targeting one 5-tuple, as they
> hit a single RX queue (and cpu) on the victim.
>
> If they use random sequence numbers in their SYN, we detect
> they do not match the expected window and send back an ACK.
>
> Th
Since linux 2.6.29, lookups only use rcu locking.
Signed-off-by: Eric Dumazet
---
include/net/inet_hashtables.h | 7 +--
net/ipv6/inet6_hashtables.c | 2 --
2 files changed, 1 insertion(+), 8 deletions(-)
diff --git a/include/net/inet_hashtables.h b/include/net/inet_hashtables.h
index 50f
We want a generic way to insert an RCU grace period before socket
freeing for cases where RCU_SLAB_DESTROY_BY_RCU is adding too
much overhead.
SLAB_DESTROY_BY_RCU strict rules force us to take a reference
on the socket sk_refcnt, and it is a performance problem for UDP
encapsulation, or TCP synflo
TCP uses per cpu 'sockets' to send some packets :
- RST packets ( tcp_v4_send_reset()) )
- ACK packets for SYN_RECV and TIMEWAIT sockets
By setting SOCK_USE_WRITE_QUEUE flag, we tell sock_wfree()
to not call sk_write_space() since these internal sockets
do not care.
This gives a small performance
First round of patches for linux-4.7
Add a generic facility for sockets to be freed after an RCU grace
period, if they need to.
Then UDP stack is changed to no longer use SLAB_DESTROY_BY_RCU,
in order to speedup rx processing for traffic encapsulated in UDP.
It gives a 17 % speedup for normal UDP
We'll soon no longer take a refcount on listeners,
so reqsk_alloc() can not assume a listener refcount is not
zero. We need to use atomic_inc_not_zero()
Signed-off-by: Eric Dumazet
---
include/net/request_sock.h | 31 +++
1 file changed, 15 insertions(+), 16 deletions
When a SYNFLOOD targets a non SO_REUSEPORT listener, multiple
cpus contend on sk->sk_refcnt and sk->sk_wmem_alloc changes.
By letting listeners use SOCK_RCU_FREE infrastructure,
we can relax TCP_LISTEN lookup rules and avoid touching sk_refcnt
Note that we still use SLAB_DESTROY_BY_RCU rules for
Tom Herbert would like not touching UDP socket refcnt for encapsulated
traffic. For this to happen, we need to use normal RCU rules, with a grace
period before freeing a socket. UDP sockets are not short lived in the
high usage case, so the added cost of call_rcu() should not be a concern.
This ac
RX packet processing holds rcu_read_lock(), so we can remove
pairs of rcu_read_lock()/rcu_read_unlock() in lookup functions
if inet_diag also holds rcu before calling them.
This is needed anyway as __inet_lookup_listener() and
inet6_lookup_listener() will soon no longer increment
refcount on the f
1 - 100 of 167 matches
Mail list logo
