mbox series

[0/4] net/mlx5: prepare to add E-switch rule flags check

Message ID 1542052877-41512-1-git-send-email-viacheslavo@mellanox.com (mailing list archive)
Headers
Series net/mlx5: prepare to add E-switch rule flags check |

Message

Slava Ovsiienko Nov. 12, 2018, 8:01 p.m. UTC 
  The tc flower filter rules are used to control E-switch from
the application side. In order to gain garanteed rule hardware
offload the skip_sw flag should be specified while applying the
rule. Some tc rules is rejected by kernel if skip_sw flag is set
by design. Currently this regards VXLAN tunneling rules only,
which are applied to special VXLAN virtual devices.

Albeit these rules should be applied with no skip_sw flag set
the kernel tries to perform hardware offload. If kernel succeeded
the in_hw flag is set in rule properties and application may check
this flag to get know whether the applied rule is actually hardware
offloaded.

Signed-off-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>

Viacheslav Ovsiienko (4):
  net/mlx5: prepare Netlink communication routine to fix
  net/mlx5: fix Netlink communication routine
  net/mlx5: prepare to add E-switch rule flags check
  net/mlx5: add E-switch rule hardware offload flag check

 drivers/net/mlx5/mlx5_flow.h     |   1 +
 drivers/net/mlx5/mlx5_flow_tcf.c | 316 +++++++++++++++++++++++++++++----------
 2 files changed, 236 insertions(+), 81 deletions(-)

Comments

Shahaf Shuler Nov. 13, 2018, 1:21 p.m. UTC | #1 
Monday, November 12, 2018 10:02 PM, Slava Ovsiienko:
> Subject: [PATCH 0/4] net/mlx5: prepare to add E-switch rule flags check
> 
> The tc flower filter rules are used to control E-switch from the application
> side. In order to gain garanteed rule hardware offload the skip_sw flag
> should be specified while applying the rule. Some tc rules is rejected by
> kernel if skip_sw flag is set by design. Currently this regards VXLAN tunneling
> rules only, which are applied to special VXLAN virtual devices.
> 
> Albeit these rules should be applied with no skip_sw flag set the kernel tries
> to perform hardware offload. If kernel succeeded the in_hw flag is set in rule
> properties and application may check this flag to get know whether the
> applied rule is actually hardware offloaded.
> 
> Signed-off-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
> 

Hi Slava,
In general looks good. Please see few comments.

In addition, please make the signed-of name and the author name match for the next version. 


> Viacheslav Ovsiienko (4):
>   net/mlx5: prepare Netlink communication routine to fix
>   net/mlx5: fix Netlink communication routine
>   net/mlx5: prepare to add E-switch rule flags check
>   net/mlx5: add E-switch rule hardware offload flag check
> 
>  drivers/net/mlx5/mlx5_flow.h     |   1 +
>  drivers/net/mlx5/mlx5_flow_tcf.c | 316
> +++++++++++++++++++++++++++++----------
>  2 files changed, 236 insertions(+), 81 deletions(-)
> 
> --
> 1.8.3.1
Shahaf Shuler Nov. 14, 2018, 2:56 p.m. UTC | #2 
Tuesday, November 13, 2018 3:21 PM, Shahaf Shuler:
> Subject: Re: [dpdk-dev] [PATCH 0/4] net/mlx5: prepare to add E-switch rule
> flags check
> 
> Monday, November 12, 2018 10:02 PM, Slava Ovsiienko:
> > Subject: [PATCH 0/4] net/mlx5: prepare to add E-switch rule flags
> > check
> >
> > The tc flower filter rules are used to control E-switch from the
> > application side. In order to gain garanteed rule hardware offload the
> > skip_sw flag should be specified while applying the rule. Some tc
> > rules is rejected by kernel if skip_sw flag is set by design.
> > Currently this regards VXLAN tunneling rules only, which are applied to
> special VXLAN virtual devices.
> >
> > Albeit these rules should be applied with no skip_sw flag set the
> > kernel tries to perform hardware offload. If kernel succeeded the
> > in_hw flag is set in rule properties and application may check this
> > flag to get know whether the applied rule is actually hardware offloaded.
> >
> > Signed-off-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
> >
> 
> Hi Slava,
> In general looks good. Please see few comments.
> 
> In addition, please make the signed-of name and the author name match for
> the next version.

Series applied to next-net-mlx, with modification of the author name. Thanks.