On Thu, Jul 10, 2014 at 04:32:49PM -0400, John W. Linville wrote:
> This is a Linux-specific virtual PMD driver backed by an AF_PACKET
> socket. This implementation uses mmap'ed ring buffers to limit copying
> and user/kernel transitions. The PACKET_FANOUT_HASH behavior of
> AF_PACKET is used for
Per netif_receive_skb function description, it may only be called from
interrupt contex, but KNI is run on kthread that like as user-space
contex. It may occur deallock, if netif_receive_skb called from kthread,
so it should be repleaced by netif_rx or adding local_bh_disable/enable
around netif_re
Thanks Jeff, it works as expected, like below command line:
./l2fwd/build/l2fwd -c 0x3 -n 4 --vdev=eth_packet0,iface=p786p1
--vdev=eth_packet1,iface=p786p2 -- -p 0x3
> -Original Message-
> From: Shaw, Jeffrey B
> Sent: Saturday, July 12, 2014 4:32 AM
> To: Zhou, Danny; John W. Linville
>
Danny, can you specify multiple --vdev parameters?
"--vdev=eth_packet0,iface=eth0 --vdev=eth_packet1,iface=eth1"
-Original Message-
From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Zhou, Danny
Sent: Friday, July 11, 2014 1:27 PM
To: John W. Linville
Cc: dev at dpdk.org
Subject: Re: [d
I want to run a common DPDK L2 or L3 forward benchmark for bi-direction
traffics, so at least two ports are required. Just like how to measure Linux
bridge or OVS performance, you need add at least two ports into a bridge.
> -Original Message-
> From: John W. Linville [mailto:linville at
Does it support specifying multiple NIC interfaces using command line option
like "--vdev=eth_packet0,iface=..."? Say "iface=eth0,eth1,eth2...", tried but
it doesn't work.
> -Original Message-
> From: Zhou, Danny
> Sent: Saturday, July 12, 2014 2:01 AM
> To: 'John W. Linville'
> Cc: dev
2014-07-11 11:30, John W. Linville:
> On Fri, Jul 11, 2014 at 05:04:04PM +0200, Thomas Monjalon wrote:
> > 2014-07-11 10:51, John W. Linville:
> > > On Fri, Jul 11, 2014 at 03:26:39PM +0200, Thomas Monjalon wrote:
> > > > Thank you for this nice work.
> > > >
> > > > I think it would be well suite
Tried on 3.12, both of them are undefined. Anyway, will comment them out and
see what performance it could achieve.
> -Original Message-
> From: John W. Linville [mailto:linville at tuxdriver.com]
> Sent: Saturday, July 12, 2014 1:41 AM
> To: Zhou, Danny
> Cc: dev at dpdk.org
> Subject: R
Hello.
I've noticed that DPDK 1.7.0 version has been released a few days ago.
The DPDK vSwitch package, however, seems supporting old version, 1.6.x,
only.
Could the DPDK vSwitch be compiled by the latest DPDK?
Or Do I need to use 1.6.x version of DPDK?
Thanks in advance.
Best regards
Byeong
Hi,
I tried another workaround and works fine:
It similar to other kcompat function, add __kc_ prefix to function
definition
and #define skb_set_hash(a,b,c) __kc_skb_set_hash(a,b,c).
Thanks,
2014-07-06 0:27 GMT+09:00 Wiles, Roger Keith :
> Hi All,
>
> I got a build failure on :
>
> [10:20][ke
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Thomas Monjalon
> Sent: Friday, July 11, 2014 9:48 AM
> To: John W. Linville
> Cc: dev at dpdk.org
> Subject: Re: [dpdk-dev] [PATCH] librte_pmd_packet: add PMD for AF_PACKET-
> based virtual devices
>
> 2014-07-
Looks like you used a pretty new kernel version with new socket options that
old kernel like my 3.12 does not support. When I tried this patch, it just
cannot build, and compiler complains like below. Which Linux distribution does
this patch work for? How to ensure it works for old kernels?
/ho
2014-07-11 10:51, John W. Linville:
> On Fri, Jul 11, 2014 at 03:26:39PM +0200, Thomas Monjalon wrote:
> > 2014-07-10 16:32, John W. Linville:
> > > This is a Linux-specific virtual PMD driver backed by an AF_PACKET
> > > socket. This implementation uses mmap'ed ring buffers to limit copying
> > >
Ah, yes...sorry, I misunderstood what you wanted to do. The syntax
below is what I would expect to use.
John
On Fri, Jul 11, 2014 at 08:35:01PM +, Zhou, Danny wrote:
> Thanks Jeff, it works as expected, like below command line:
>
> ./l2fwd/build/l2fwd -c 0x3 -n 4 --vdev=eth_packet0,iface=p7
On Fri, Jul 11, 2014 at 03:29:17PM +, Venkatesan, Venky wrote:
> > -Original Message-
> > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of John W.
> > Linville
> > Sent: Friday, July 11, 2014 7:49 AM
> > To: Stephen Hemminger
> > Cc: dev at dpdk.org
> > Subject: Re: [dpdk-dev] [
> -Original Message-
> From: Stephen Hemminger [mailto:stephen at networkplumber.org]
> Sent: Friday, July 11, 2014 8:16 AM
> To: Richardson, Bruce
> Cc: John W. Linville; dev at dpdk.org
> Subject: Re: [dpdk-dev] [PATCH] librte_pmd_packet: add PMD for AF_PACKET-
> based virtual devices
>
I'm not sure that would make any sense -- the AF_PACKET sockets are
mapped to specific interfaces.
What are you trying to do with a syntax like that?
John
On Fri, Jul 11, 2014 at 07:04:19PM +, Zhou, Danny wrote:
> Does it support specifying multiple NIC interfaces using command line option
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of John W. Linville
> Sent: Friday, July 11, 2014 7:49 AM
> To: Stephen Hemminger
> Cc: dev at dpdk.org
> Subject: Re: [dpdk-dev] [PATCH] librte_pmd_packet: add PMD for
> AF_PACKET- based virtual devices
>
> On Fr
2014-07-10 16:32, John W. Linville:
> This is a Linux-specific virtual PMD driver backed by an AF_PACKET
> socket. This implementation uses mmap'ed ring buffers to limit copying
> and user/kernel transitions. The PACKET_FANOUT_HASH behavior of
> AF_PACKET is used for frame reception. In the curr
Hi Sujith,
It makes sens, using VFIO makes a far cleaner implementation.
You worked on the performance, could you share some measurements ? Did you
introduce vectorization functions as Intel did a while ago? There are allways
tradeoffs between pps and latency, do you include documentation to co
Hi Stephen,
2014-07-11 06:03, Stephen Hemminger:
> On Thu, 19 Jun 2014 23:02:30 -0700
>
> Stephen Hemminger wrote:
> > This set of patches adds new functionality to VMXNET3 and
> > removes some dead code, and fixes a memory leak.
> >
> > These patches assume that the earlier VMXNET3 cleanup pat
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of John W. Linville
> Sent: Friday, July 11, 2014 7:49 AM
> To: Stephen Hemminger
> Cc: dev at dpdk.org
> Subject: Re: [dpdk-dev] [PATCH] librte_pmd_packet: add PMD for AF_PACKET-
> based virtual devices
>
> On Fri
Not sure what the issue might be, PACKET_FANOUT_FLAG_ROLLOVER is
defined in include/uapi/linux/if_packet.h in the v3.12 tree.
On Fri, Jul 11, 2014 at 06:01:27PM +, Zhou, Danny wrote:
> Tried on 3.12, both of them are undefined. Anyway, will comment them out and
> see what performance it could
Hi Matt,
I normal only compile the core DPDK code and then compile my Pktgen traffic
generator application
(git clone https://github.com/Pktgen/Pktgen-DPDK) which uses a fair amount of
functions from DPDK.
Thanks
++Keith
Keith Wiles, Principal Technologist with CTO office, Wind River
mobile 97
> -Original Message-
> From: Zachary.Jen at cas-well.com [mailto:Zachary.Jen at cas-well.com]
> Sent: Friday, July 11, 2014 4:05 AM
> To: dev at dpdk.org
> Cc: Richardson, Bruce; Alan.Yu at cas-well.com
> Subject: Re: [dpdk-dev] DPDK Performance issue with l2fwd
>
> Hi Bruce:
>
> Thanks
Hi,
I was playing with vhost example to measure VM to VM throughput. With
default SW mode I saw 9Gbps. While trying HW mode, traffic does not
work.
Looking at the code I saw a snippet in
lib/librte_pmd_ixgbe/ixgbe_rxtx.c:ixgbe_vmdq_rx_hw_configure -
/* PFDMA Tx General Switch Control Enab
On Fri, Jul 11, 2014 at 05:38:17PM +, Richardson, Bruce wrote:
> > -Original Message-
> > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Thomas Monjalon
> > Sent: Friday, July 11, 2014 9:48 AM
> > To: John W. Linville
> > Cc: dev at dpdk.org
> > Subject: Re: [dpdk-dev] [PATCH]
On Fri, Jul 11, 2014 at 05:20:42PM +, Zhou, Danny wrote:
> Looks like you used a pretty new kernel version with new socket options that
> old kernel like my 3.12 does not support. When I tried this patch, it just
> cannot build, and compiler complains like below. Which Linux distribution
> d
On Fri, Jul 11, 2014 at 03:29:17PM +, Venkatesan, Venky wrote:
> > -Original Message-
> > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of John W. Linville
> > Sent: Friday, July 11, 2014 7:49 AM
> > To: Stephen Hemminger
> > Cc: dev at dpdk.org
> > Subject: Re: [dpdk-dev] [PATCH
On Fri, Jul 11, 2014 at 05:04:04PM +0200, Thomas Monjalon wrote:
> 2014-07-11 10:51, John W. Linville:
> > On Fri, Jul 11, 2014 at 03:26:39PM +0200, Thomas Monjalon wrote:
> > > 2014-07-10 16:32, John W. Linville:
> > > > This is a Linux-specific virtual PMD driver backed by an AF_PACKET
> > > > so
Hi Bruce:
Thanks for your suggestion.
I have tried to use testpmd to test 16 ports today.
The result is so Interested. It can work, although some ports get low
performance (only get 80%).
Besides, I also do another test in l2fwd.
I tried to use 82580 * 16 ports in the same platform with l2fwd tes
On Fri, Jul 11, 2014 at 03:26:39PM +0200, Thomas Monjalon wrote:
> 2014-07-10 16:32, John W. Linville:
> > This is a Linux-specific virtual PMD driver backed by an AF_PACKET
> > socket. This implementation uses mmap'ed ring buffers to limit copying
> > and user/kernel transitions. The PACKET_FANO
On Fri, Jul 11, 2014 at 06:11:47AM -0700, Stephen Hemminger wrote:
> On Thu, 10 Jul 2014 16:32:49 -0400
> "John W. Linville" wrote:
>
> > This is a Linux-specific virtual PMD driver backed by an AF_PACKET
> > socket. This implementation uses mmap'ed ring buffers to limit copying
> > and user/ker
Hi,
Thank you for sending this patch.
Please, could you re-send it with a more detailed explanation and a
Signed-off-by line?
There is a how-to for patch contributions here:
http://dpdk.org/dev#send
2014-07-11 01:04, Yao-Po Wang:
> ---
> lib/librte_eal/linuxapp/kni/kni_net.c | 2 +-
>
Thanks Roger,
We saw similar issues with regard to kcompat.h. Can I ask if you've done
anything beyond the example applications under 14.04?
--
Matt Laswell
infinite io
On Thu, Jul 10, 2014 at 7:07 PM, Wiles, Roger Keith <
keith.wiles at windriver.com> wrote:
> The one problem I had with 14.
On Fri, 11 Jul 2014 15:06:25 +
"Richardson, Bruce" wrote:
> > -Original Message-
> > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of John W. Linville
> > Sent: Friday, July 11, 2014 7:49 AM
> > To: Stephen Hemminger
> > Cc: dev at dpdk.org
> > Subject: Re: [dpdk-dev] [PATCH] l
Hi Waterman
I'm seeing the same problem, here is my Ubuntu version, I'm using ubunto
desktop.
Linux schoi5-desktop 3.13.0-30-generic #55-Ubuntu SMP Fri Jul 4 21:40:53 UTC
2014 x86_64 x86_64 x86_64 GNU/Linux
Regards,
Choi, Sy Jong
Platform Application Engineer
-Original Message-
From:
Hi Hobywan,
Thanks for the email !
We?ve been working on performance benchmarking. Also, we felt that it would be
better to push the driver after Anatoly?s patch (vfio-pci) got in so that we
could make the necessary modification before submission.
Now that 1.7.0 is out, we are hopeful of submit
On Thu, 10 Jul 2014 16:32:49 -0400
"John W. Linville" wrote:
> This is a Linux-specific virtual PMD driver backed by an AF_PACKET
> socket. This implementation uses mmap'ed ring buffers to limit copying
> and user/kernel transitions. The PACKET_FANOUT_HASH behavior of
> AF_PACKET is used for fr
On Thu, 19 Jun 2014 23:02:30 -0700
Stephen Hemminger wrote:
> This set of patches adds new functionality to VMXNET3 and
> removes some dead code, and fixes a memory leak.
>
> These patches assume that the earlier VMXNET3 cleanup patches
> sent to the mailing list are already merged (they haven't
> > This patch set enables clang compilation on FreeBSD and Linux. It
> > includes patches to fix a number of compilation errors thrown up by
> > clang, and then adds in the appropriate toolchain makefiles and
> > compile-time configurations.
> >
> > This set has been tested with clang v3.3 on Free
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Richardson, Bruce
> Sent: Friday, July 11, 2014 8:13 AM
> To: dev at dpdk.org
> Subject: [dpdk-dev] [PATCH] nic_uio: add device ids for i40e devices
>
> The FreeBSD nic_uio driver was missing the #defines to i
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Olson, Matt Lyle
> Sent: Thursday, July 10, 2014 4:12 AM
> To: dev at dpdk.org
> Subject: [dpdk-dev] Running kni with low amount of cores
>
> Hello,
>
> I have two NIC devices and a quad core system that I'm
Chris,
Here is what I verified...
>> $ uname -r
3.12.6
>> $ grep __fentry__ /proc/kallsyms
Nothing found here.
>> $ grep CONFIG_FTRACE /lib/modules/$(uname -r)/build/.config
CONFIG_FTRACE=y
CONFIG_FTRACE_SYSCALLS=y
CONFIG_FTRACE_MCOUNT_RECORD=y
# CONFIG_FTRACE_STARTUP_TEST is not set
-Or
---
lib/librte_eal/linuxapp/kni/kni_net.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/librte_eal/linuxapp/kni/kni_net.c
b/lib/librte_eal/linuxapp/kni/kni_net.c
index d3c0190..28cc5ab 100644
--- a/lib/librte_eal/linuxapp/kni/kni_net.c
+++ b/lib/librte_eal/linuxapp/kni/k
The FreeBSD nic_uio driver was missing the #defines to include the device ids
for devices using the i40e driver. This change adds in the missing defines.
Signed-off-by: Bruce Richardson
---
lib/librte_eal/bsdapp/nic_uio/nic_uio.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/lib/librte_e
The one problem I had with 14.04 was the kcompat.h file. It looks like a hash
routine has changed its arguments. I edited the kcompat.h file and was about to
change the code to allow DPDK to build. It is not affix but it worked for me.
lib/librte_eal/linuxapp/kni/ethtool/igb/kcompat.h
/* C
47 matches
Mail list logo
