Re: [dpdk-dev] ebpf support in dpdk

I have a feeling you are talking about P4 On Fri, Apr 21, 2017 at 1:32 PM, Anupam Kapoor wrote: > hi all, > > i was just wondering if there are plans on adding ebpf support in dpdk ? > imho, it would be pretty useful, among other things, in creating generic > (protocol oblivious) match+actio

[dpdk-dev] FM10420 (RRC) - Problems with PMD matching on virtual functions in ACL table

Hello All. I have an FM10420 card with dual 100G ports, I am running a DPDK application that binds two virtual function ports to it. I tried to set rules to the IES to match on the ports logical id (for the VF's), the rules are configured well but are not working (not matching). If i dont use DPD

[dpdk-dev] Unable to compile DPDK 2.1

Hello all, I was trying to help someone compile DPDK2.1 on Ubuntu 14.04, we were following the instructions on http://www.dpdk.org/doc/quick-start We installed libpcap using "sudo apt-get install libpcap-dev" When starting the 'make' the following error occured: /usr/bin/ld: skipping incompatib

[dpdk-dev] FlowDirector Rules for TX packets

, flow director is just an RX-side packet filtering NIC feature. > > Regards, > Cristian > > > -Original Message- > > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Gal Sagie > > Sent: Friday, March 27, 2015 10:44 AM > > To: > > Subject: [

[dpdk-dev] FlowDirector Rules for TX packets

Hello All, I can define drop filter rules for RX packets using FlowDirector. Support for configuring this from CLI is also available using ethtool with the --config-ntuple option I am wondering, does FlowDirector has any support to drop TX packets? (Meaning that i can define rules which are if ma

[dpdk-dev] Fast Path Query

Your application (which you build on top of DPDK) needs to filter which traffic is control traffic and inject it into the network stack. You can leverage DPDK KNI for that ( http://dpdk.org/doc/guides/prog_guide/kernel_nic_interface.html) Keep in mind you also need to take care of the TX side. On

[dpdk-dev] IPv6 Offload Capabilities

Hello All, I noticed that in version 1.8, there are no flags to indicate IPv6 check sum offloading (only DEV_TX_OFFLOAD_IPV4_CKSUM) which means TSO offloading is also not supported for IPv6. Are there any plans/road map to support this? Thanks Gal.

[dpdk-dev] TSO with chained imbues

Hello all, Does the TSO offload option works for chained mbufs? What i mean is, if i have a chain representing a big TCP buffer, and i only set the first mbuf with TSO flags and tso_segsz (mss) and L4 size, will that work ? (or only applied on the first mbuf) I think that ideally this should work

[dpdk-dev] Request for a feature in PMD

Hello Prashant, You can achieve this behavior, you can look at either the rte_mbuf clone API or look at the reference count field inside the mbuf structure. If you increase this ref count the TX function will not free your mbuf. (there are some API calls to update this variable) Gal. On Sun, Oc

[dpdk-dev] sending and receiving packets

, Oct 18, 2013 at 1:11 PM, Gopi Krishna B wrote: > > > > On Thu, Oct 17, 2013 at 6:27 PM, Gal Sagie wrote: > >> Assigning an IP is a function of the network stack, DPDK and the NIC does >> not need to be aware of that. >> DPDK is just used to poll packets from th

[dpdk-dev] sending and receiving packets

ddr(IFNAME, IF_ADDR, "255.255.255.0") != 0) On Thu, Oct 17, 2013 at 4:26 PM, Gopi Krishna B wrote: > > > On Thu, Oct 17, 2013 at 3:01 PM, Gal Sagie wrote: > >> Rump kernels is a flexible kernel architecture which runs in user space >> and is a very interesting

[dpdk-dev] sending and receiving packets

Rump kernels is a flexible kernel architecture which runs in user space and is a very interesting project, you can read more about it here => http://www.netbsd.org/docs/rump/#rump-about It is currently part of the NetBSD source tree. A project was made to integrate Intel DPDK inside Rump kernel ne

[dpdk-dev] RTE Hash Lib

Hello all, I was wondering if there is a specific reason why the maximum bucket size that can be configured is 16 ? (RTE_HASH_BUCKET_ENTRIES_MAX = 16) Thanks Gal. -- next part -- An HTML attachment was scrubbed... URL:

[dpdk-dev] L3/L4 Filters configuration

nks for your help! On Mon, Jun 10, 2013 at 12:18 PM, Thomas Monjalon wrote: > Hello, > > 09/06/2013 15:29, Gal Sagie : > > Is there any chance to see an example of how to configure L3/L4 filters > per > > port in the NIC ? > > also is there any way to control the

[dpdk-dev] problem with rte_hash

nt is an example of these options. > > I hope this helps. > > Thank you, ++Keith > --- > Keith Wiles > Principal Technologist for Networking > cell 972-213-5533 > Wind River Systems > > > On Jun 9, 2013, at 9:52 AM, Gal Sagie

[dpdk-dev] problem with rte_hash

Hello all, I am trying to do an example of using rte_hash, i am getting the following linker problem: /main.o -lrte_eal -lpthread -lethdev -lrte_malloc -lrte_mbuf -lrte_mempool -lrte_pmd_igb -lrte_pmd_ixgbe -lrte_ring -lrte_timer -lrte_hash /home/gal/dpdk/x86_64-default-linuxapp-gcc/lib/librte_

[dpdk-dev] L3/L4 Filters configuration

Hello all, Reading the poll mode driver section in the programmers manual i came across this line: "Other features such as the L3/L4 5-Tuple packet filtering feature of a port can be configured in the same way" Is there any chance to see an example of how to configure L3/L4 filters per port in t