Re: [PATCH net-next 12/18] IB/mlx5: Add kernel offload flow-tag

2016-06-21 Thread Or Gerlitz
On 6/21/2016 6:18 PM, Eric Dumazet wrote: One solution would be to setup a special netdev used only for sniffers (No IP address on it) -> Only changes would happen in the driver, to set skb->dev to this 'debug' device. Eric, Yep, that was an option too, but when we realized that libpcap has t

Re: [PATCH net-next 12/18] IB/mlx5: Add kernel offload flow-tag

2016-06-21 Thread Eric Dumazet
On Tue, Jun 21, 2016 at 6:04 AM, Saeed Mahameed wrote: > > Alexei , I don't understand your concern. > We already have a full/complete working dpdk bypass solution in > userspace nothing extra is required from the kernel. > > We just want to see this traffic and any other rdma traffic in tcpdump

Re: [PATCH net-next 12/18] IB/mlx5: Add kernel offload flow-tag

2016-06-21 Thread Saeed Mahameed
On Tue, Jun 21, 2016 at 5:18 AM, Alexei Starovoitov wrote: > On Sat, Jun 18, 2016 at 01:31:26AM +0300, Saeed Mahameed wrote: >> >> We simply want to selectively be able to see RoCE/RDMA ETH standard >> traffic in tcpdump, for diagnostic purposes. >> so in order to not overwhelm the kernel TCP/IP s

Re: [PATCH net-next 12/18] IB/mlx5: Add kernel offload flow-tag

2016-06-20 Thread Alexei Starovoitov
On Sat, Jun 18, 2016 at 01:31:26AM +0300, Saeed Mahameed wrote: > > We simply want to selectively be able to see RoCE/RDMA ETH standard > traffic in tcpdump, for diagnostic purposes. > so in order to not overwhelm the kernel TCP/IP stack with this > traffic, this patch in particular > configures C

Re: [PATCH net-next 12/18] IB/mlx5: Add kernel offload flow-tag

2016-06-19 Thread Saeed Mahameed
On Sat, Jun 18, 2016 at 2:34 AM, Eric Dumazet wrote: > On Fri, Jun 17, 2016 at 3:31 PM, Saeed Mahameed > wrote: >> >> Today there are some bad usages and abuse to skb->protocol where some >> device drivers set skb->protocol = 0xff to skip the kernel TCP/IP >> processing for the same diagnosti

Re: [PATCH net-next 12/18] IB/mlx5: Add kernel offload flow-tag

2016-06-17 Thread Eric Dumazet
On Fri, Jun 17, 2016 at 3:31 PM, Saeed Mahameed wrote: > On Fri, Jun 17, 2016 at 7:00 PM, Alexei Starovoitov > wrote: >> On Fri, Jun 17, 2016 at 05:43:53PM +0300, Saeed Mahameed wrote: >>> From: Maor Gottlieb >>> >>> Add kernel offload flow tag for packets that will bypass the kernel >>> stack,

Re: [PATCH net-next 12/18] IB/mlx5: Add kernel offload flow-tag

2016-06-17 Thread Saeed Mahameed
On Fri, Jun 17, 2016 at 7:00 PM, Alexei Starovoitov wrote: > On Fri, Jun 17, 2016 at 05:43:53PM +0300, Saeed Mahameed wrote: >> From: Maor Gottlieb >> >> Add kernel offload flow tag for packets that will bypass the kernel >> stack, e.g (RoCE/RDMA/RAW ETH (DPDK), etc ..). > > so the whole series i

Re: [PATCH net-next 12/18] IB/mlx5: Add kernel offload flow-tag

2016-06-17 Thread John Fastabend
On 16-06-17 09:00 AM, Alexei Starovoitov wrote: > On Fri, Jun 17, 2016 at 05:43:53PM +0300, Saeed Mahameed wrote: >> From: Maor Gottlieb >> >> Add kernel offload flow tag for packets that will bypass the kernel >> stack, e.g (RoCE/RDMA/RAW ETH (DPDK), etc ..). > > so the whole series is an elabor

Re: [PATCH net-next 12/18] IB/mlx5: Add kernel offload flow-tag

2016-06-17 Thread Alexei Starovoitov
On Fri, Jun 17, 2016 at 05:43:53PM +0300, Saeed Mahameed wrote: > From: Maor Gottlieb > > Add kernel offload flow tag for packets that will bypass the kernel > stack, e.g (RoCE/RDMA/RAW ETH (DPDK), etc ..). so the whole series is an elaborate way to enable dpdk? how nice. NACK.

[PATCH net-next 12/18] IB/mlx5: Add kernel offload flow-tag

2016-06-17 Thread Saeed Mahameed
From: Maor Gottlieb Add kernel offload flow tag for packets that will bypass the kernel stack, e.g (RoCE/RDMA/RAW ETH (DPDK), etc ..). User leftover FTEs are shared with sniffer, therefore leftover rules should be added with the bypass flow-tag. Signed-off-by: Maor Gottlieb Signed-off-by: Saee