* Re: [PATCH net] net/mlx5e: Fix ethtool -N flow-type ip4 to RSS context
2025-03-19 12:45 [PATCH net] net/mlx5e: Fix ethtool -N flow-type ip4 to RSS context Maxim Mikityanskiy
@ 2025-03-19 14:24 ` Daniel Borkmann
2025-03-19 15:46 ` Joe Damato
` (3 subsequent siblings)
4 siblings, 0 replies; 12+ messages in thread
From: Daniel Borkmann @ 2025-03-19 14:24 UTC (permalink / raw)
To: Maxim Mikityanskiy, Saeed Mahameed, Tariq Toukan
Cc: Leon Romanovsky, Andrew Lunn, David S. Miller, Eric Dumazet,
Jakub Kicinski, Paolo Abeni, netdev, Maxim Mikityanskiy
On 3/19/25 1:45 PM, Maxim Mikityanskiy wrote:
> There commands can be used to add an RSS context and steer some traffic
> into it:
>
> # ethtool -X eth0 context new
> New RSS context is 1
> # ethtool -N eth0 flow-type ip4 dst-ip 1.1.1.1 context 1
> Added rule with ID 1023
>
> However, the second command fails with EINVAL on mlx5e:
>
> # ethtool -N eth0 flow-type ip4 dst-ip 1.1.1.1 context 1
> rmgr: Cannot insert RX class rule: Invalid argument
> Cannot insert classification rule
>
> It happens when flow_get_tirn calls flow_type_to_traffic_type with
> flow_type = IP_USER_FLOW or IPV6_USER_FLOW. That function only handles
> IPV4_FLOW and IPV6_FLOW cases, but unlike all other cases which are
> common for hash and spec, IPv4 and IPv6 defines different contants for
> hash and for spec:
>
> #define TCP_V4_FLOW 0x01 /* hash or spec (tcp_ip4_spec) */
> #define UDP_V4_FLOW 0x02 /* hash or spec (udp_ip4_spec) */
> ...
> #define IPV4_USER_FLOW 0x0d /* spec only (usr_ip4_spec) */
> #define IP_USER_FLOW IPV4_USER_FLOW
> #define IPV6_USER_FLOW 0x0e /* spec only (usr_ip6_spec; nfc only) */
> #define IPV4_FLOW 0x10 /* hash only */
> #define IPV6_FLOW 0x11 /* hash only */
>
> Extend the switch in flow_type_to_traffic_type to support both, which
> fixes the failing ethtool -N command with flow-type ip4 or ip6.
>
> Fixes: 248d3b4c9a39 ("net/mlx5e: Support flow classification into RSS contexts")
> Signed-off-by: Maxim Mikityanskiy <maxim@isovalent.com>
Awesome, thanks Max!
Tested-by: Daniel Borkmann <daniel@iogearbox.net>
^ permalink raw reply [flat|nested] 12+ messages in thread* Re: [PATCH net] net/mlx5e: Fix ethtool -N flow-type ip4 to RSS context
2025-03-19 12:45 [PATCH net] net/mlx5e: Fix ethtool -N flow-type ip4 to RSS context Maxim Mikityanskiy
2025-03-19 14:24 ` Daniel Borkmann
@ 2025-03-19 15:46 ` Joe Damato
2025-03-20 7:32 ` Tariq Toukan
` (2 subsequent siblings)
4 siblings, 0 replies; 12+ messages in thread
From: Joe Damato @ 2025-03-19 15:46 UTC (permalink / raw)
To: Maxim Mikityanskiy
Cc: Saeed Mahameed, Tariq Toukan, Leon Romanovsky, Andrew Lunn,
David S. Miller, Eric Dumazet, Jakub Kicinski, Paolo Abeni,
netdev, Maxim Mikityanskiy
On Wed, Mar 19, 2025 at 02:45:08PM +0200, Maxim Mikityanskiy wrote:
> There commands can be used to add an RSS context and steer some traffic
> into it:
>
> # ethtool -X eth0 context new
> New RSS context is 1
> # ethtool -N eth0 flow-type ip4 dst-ip 1.1.1.1 context 1
> Added rule with ID 1023
>
> However, the second command fails with EINVAL on mlx5e:
>
> # ethtool -N eth0 flow-type ip4 dst-ip 1.1.1.1 context 1
> rmgr: Cannot insert RX class rule: Invalid argument
> Cannot insert classification rule
>
> It happens when flow_get_tirn calls flow_type_to_traffic_type with
> flow_type = IP_USER_FLOW or IPV6_USER_FLOW. That function only handles
> IPV4_FLOW and IPV6_FLOW cases, but unlike all other cases which are
> common for hash and spec, IPv4 and IPv6 defines different contants for
> hash and for spec:
>
> #define TCP_V4_FLOW 0x01 /* hash or spec (tcp_ip4_spec) */
> #define UDP_V4_FLOW 0x02 /* hash or spec (udp_ip4_spec) */
> ...
> #define IPV4_USER_FLOW 0x0d /* spec only (usr_ip4_spec) */
> #define IP_USER_FLOW IPV4_USER_FLOW
> #define IPV6_USER_FLOW 0x0e /* spec only (usr_ip6_spec; nfc only) */
> #define IPV4_FLOW 0x10 /* hash only */
> #define IPV6_FLOW 0x11 /* hash only */
>
> Extend the switch in flow_type_to_traffic_type to support both, which
> fixes the failing ethtool -N command with flow-type ip4 or ip6.
>
> Fixes: 248d3b4c9a39 ("net/mlx5e: Support flow classification into RSS contexts")
> Signed-off-by: Maxim Mikityanskiy <maxim@isovalent.com>
> ---
> drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
> index 773624bb2c5d..d68230a7b9f4 100644
> --- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
> +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
> @@ -884,8 +884,10 @@ static int flow_type_to_traffic_type(u32 flow_type)
> case ESP_V6_FLOW:
> return MLX5_TT_IPV6_IPSEC_ESP;
> case IPV4_FLOW:
> + case IP_USER_FLOW:
> return MLX5_TT_IPV4;
> case IPV6_FLOW:
> + case IPV6_USER_FLOW:
> return MLX5_TT_IPV6;
> default:
> return -EINVAL;
Good catch.
Reviewed-by: Joe Damato <jdamato@fastly.com>
^ permalink raw reply [flat|nested] 12+ messages in thread* Re: [PATCH net] net/mlx5e: Fix ethtool -N flow-type ip4 to RSS context
2025-03-19 12:45 [PATCH net] net/mlx5e: Fix ethtool -N flow-type ip4 to RSS context Maxim Mikityanskiy
2025-03-19 14:24 ` Daniel Borkmann
2025-03-19 15:46 ` Joe Damato
@ 2025-03-20 7:32 ` Tariq Toukan
2025-03-20 7:43 ` Maxim Mikityanskiy
2025-03-20 8:25 ` Gal Pressman
2025-03-24 20:50 ` patchwork-bot+netdevbpf
4 siblings, 1 reply; 12+ messages in thread
From: Tariq Toukan @ 2025-03-20 7:32 UTC (permalink / raw)
To: Maxim Mikityanskiy, Saeed Mahameed
Cc: Leon Romanovsky, Andrew Lunn, David S. Miller, Eric Dumazet,
Jakub Kicinski, Paolo Abeni, netdev, Maxim Mikityanskiy,
Tariq Toukan
On 19/03/2025 14:45, Maxim Mikityanskiy wrote:
> There commands can be used to add an RSS context and steer some traffic
> into it:
>
> # ethtool -X eth0 context new
> New RSS context is 1
> # ethtool -N eth0 flow-type ip4 dst-ip 1.1.1.1 context 1
> Added rule with ID 1023
>
> However, the second command fails with EINVAL on mlx5e:
>
> # ethtool -N eth0 flow-type ip4 dst-ip 1.1.1.1 context 1
> rmgr: Cannot insert RX class rule: Invalid argument
> Cannot insert classification rule
>
> It happens when flow_get_tirn calls flow_type_to_traffic_type with
> flow_type = IP_USER_FLOW or IPV6_USER_FLOW. That function only handles
> IPV4_FLOW and IPV6_FLOW cases, but unlike all other cases which are
> common for hash and spec, IPv4 and IPv6 defines different contants for
> hash and for spec:
>
> #define TCP_V4_FLOW 0x01 /* hash or spec (tcp_ip4_spec) */
> #define UDP_V4_FLOW 0x02 /* hash or spec (udp_ip4_spec) */
> ...
> #define IPV4_USER_FLOW 0x0d /* spec only (usr_ip4_spec) */
> #define IP_USER_FLOW IPV4_USER_FLOW
> #define IPV6_USER_FLOW 0x0e /* spec only (usr_ip6_spec; nfc only) */
> #define IPV4_FLOW 0x10 /* hash only */
> #define IPV6_FLOW 0x11 /* hash only */
>
> Extend the switch in flow_type_to_traffic_type to support both, which
> fixes the failing ethtool -N command with flow-type ip4 or ip6.
>
Hi Maxim,
Thanks for your patch!
> Fixes: 248d3b4c9a39 ("net/mlx5e: Support flow classification into RSS contexts")
Seems that the issue originates in commit 756c41603a18 ("net/mlx5e:
ethtool, Support user configuration for RX hash fields"), when directly
classifying into an RQ, before the multi RSS context support.
> Signed-off-by: Maxim Mikityanskiy <maxim@isovalent.com>
> ---
> drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
> index 773624bb2c5d..d68230a7b9f4 100644
> --- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
> +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
> @@ -884,8 +884,10 @@ static int flow_type_to_traffic_type(u32 flow_type)
> case ESP_V6_FLOW:
> return MLX5_TT_IPV6_IPSEC_ESP;
> case IPV4_FLOW:
> + case IP_USER_FLOW:
> return MLX5_TT_IPV4;
> case IPV6_FLOW:
> + case IPV6_USER_FLOW:
> return MLX5_TT_IPV6;
> default:
> return -EINVAL;
^ permalink raw reply [flat|nested] 12+ messages in thread* Re: [PATCH net] net/mlx5e: Fix ethtool -N flow-type ip4 to RSS context
2025-03-20 7:32 ` Tariq Toukan
@ 2025-03-20 7:43 ` Maxim Mikityanskiy
2025-03-20 8:09 ` Tariq Toukan
0 siblings, 1 reply; 12+ messages in thread
From: Maxim Mikityanskiy @ 2025-03-20 7:43 UTC (permalink / raw)
To: Tariq Toukan
Cc: Saeed Mahameed, Leon Romanovsky, Andrew Lunn, David S. Miller,
Eric Dumazet, Jakub Kicinski, Paolo Abeni, netdev,
Maxim Mikityanskiy, Tariq Toukan
On Thu, 20 Mar 2025 at 09:32, Tariq Toukan <ttoukan.linux@gmail.com> wrote:
>
>
>
> On 19/03/2025 14:45, Maxim Mikityanskiy wrote:
> > There commands can be used to add an RSS context and steer some traffic
> > into it:
> >
> > # ethtool -X eth0 context new
> > New RSS context is 1
> > # ethtool -N eth0 flow-type ip4 dst-ip 1.1.1.1 context 1
> > Added rule with ID 1023
> >
> > However, the second command fails with EINVAL on mlx5e:
> >
> > # ethtool -N eth0 flow-type ip4 dst-ip 1.1.1.1 context 1
> > rmgr: Cannot insert RX class rule: Invalid argument
> > Cannot insert classification rule
> >
> > It happens when flow_get_tirn calls flow_type_to_traffic_type with
> > flow_type = IP_USER_FLOW or IPV6_USER_FLOW. That function only handles
> > IPV4_FLOW and IPV6_FLOW cases, but unlike all other cases which are
> > common for hash and spec, IPv4 and IPv6 defines different contants for
> > hash and for spec:
> >
> > #define TCP_V4_FLOW 0x01 /* hash or spec (tcp_ip4_spec) */
> > #define UDP_V4_FLOW 0x02 /* hash or spec (udp_ip4_spec) */
> > ...
> > #define IPV4_USER_FLOW 0x0d /* spec only (usr_ip4_spec) */
> > #define IP_USER_FLOW IPV4_USER_FLOW
> > #define IPV6_USER_FLOW 0x0e /* spec only (usr_ip6_spec; nfc only) */
> > #define IPV4_FLOW 0x10 /* hash only */
> > #define IPV6_FLOW 0x11 /* hash only */
> >
> > Extend the switch in flow_type_to_traffic_type to support both, which
> > fixes the failing ethtool -N command with flow-type ip4 or ip6.
> >
>
> Hi Maxim,
> Thanks for your patch!
>
> > Fixes: 248d3b4c9a39 ("net/mlx5e: Support flow classification into RSS contexts")
>
> Seems that the issue originates in commit 756c41603a18 ("net/mlx5e:
> ethtool, Support user configuration for RX hash fields"),
Not really; commit 756c41603a18 configures the hash (not flow
direction), and IPV4_FLOW/IPV6_FLOW are already correct constants for
IP-based hashes. Moreover, we don't support them anyway, see
mlx5e_set_rss_hash_opt:
/* RSS does not support anything other than hashing to queues
* on src IP, dest IP, TCP/UDP src port and TCP/UDP dest
* port.
*/
if (flow_type != TCP_V4_FLOW &&
flow_type != TCP_V6_FLOW &&
flow_type != UDP_V4_FLOW &&
flow_type != UDP_V6_FLOW)
return -EOPNOTSUPP;
> when directly
> classifying into an RQ, before the multi RSS context support.
Direct classification into an RQ actually works before my fix, because
it goes to another branch in flow_get_tirn, that doesn't call
flow_type_to_traffic_type. It's only steering to an RSS context that
was broken for flow-type ip4/ip6.
> > Signed-off-by: Maxim Mikityanskiy <maxim@isovalent.com>
> > ---
> > drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c | 2 ++
> > 1 file changed, 2 insertions(+)
> >
> > diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
> > index 773624bb2c5d..d68230a7b9f4 100644
> > --- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
> > +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
> > @@ -884,8 +884,10 @@ static int flow_type_to_traffic_type(u32 flow_type)
> > case ESP_V6_FLOW:
> > return MLX5_TT_IPV6_IPSEC_ESP;
> > case IPV4_FLOW:
> > + case IP_USER_FLOW:
> > return MLX5_TT_IPV4;
> > case IPV6_FLOW:
> > + case IPV6_USER_FLOW:
> > return MLX5_TT_IPV6;
> > default:
> > return -EINVAL;
>
^ permalink raw reply [flat|nested] 12+ messages in thread* Re: [PATCH net] net/mlx5e: Fix ethtool -N flow-type ip4 to RSS context
2025-03-20 7:43 ` Maxim Mikityanskiy
@ 2025-03-20 8:09 ` Tariq Toukan
0 siblings, 0 replies; 12+ messages in thread
From: Tariq Toukan @ 2025-03-20 8:09 UTC (permalink / raw)
To: Maxim Mikityanskiy
Cc: Saeed Mahameed, Leon Romanovsky, Andrew Lunn, David S. Miller,
Eric Dumazet, Jakub Kicinski, Paolo Abeni, netdev,
Maxim Mikityanskiy, Tariq Toukan
On 20/03/2025 9:43, Maxim Mikityanskiy wrote:
> On Thu, 20 Mar 2025 at 09:32, Tariq Toukan <ttoukan.linux@gmail.com> wrote:
>>
>>
>>
>> On 19/03/2025 14:45, Maxim Mikityanskiy wrote:
>>> There commands can be used to add an RSS context and steer some traffic
>>> into it:
>>>
>>> # ethtool -X eth0 context new
>>> New RSS context is 1
>>> # ethtool -N eth0 flow-type ip4 dst-ip 1.1.1.1 context 1
>>> Added rule with ID 1023
>>>
>>> However, the second command fails with EINVAL on mlx5e:
>>>
>>> # ethtool -N eth0 flow-type ip4 dst-ip 1.1.1.1 context 1
>>> rmgr: Cannot insert RX class rule: Invalid argument
>>> Cannot insert classification rule
>>>
>>> It happens when flow_get_tirn calls flow_type_to_traffic_type with
>>> flow_type = IP_USER_FLOW or IPV6_USER_FLOW. That function only handles
>>> IPV4_FLOW and IPV6_FLOW cases, but unlike all other cases which are
>>> common for hash and spec, IPv4 and IPv6 defines different contants for
>>> hash and for spec:
>>>
>>> #define TCP_V4_FLOW 0x01 /* hash or spec (tcp_ip4_spec) */
>>> #define UDP_V4_FLOW 0x02 /* hash or spec (udp_ip4_spec) */
>>> ...
>>> #define IPV4_USER_FLOW 0x0d /* spec only (usr_ip4_spec) */
>>> #define IP_USER_FLOW IPV4_USER_FLOW
>>> #define IPV6_USER_FLOW 0x0e /* spec only (usr_ip6_spec; nfc only) */
>>> #define IPV4_FLOW 0x10 /* hash only */
>>> #define IPV6_FLOW 0x11 /* hash only */
>>>
>>> Extend the switch in flow_type_to_traffic_type to support both, which
>>> fixes the failing ethtool -N command with flow-type ip4 or ip6.
>>>
>>
>> Hi Maxim,
>> Thanks for your patch!
>>
>>> Fixes: 248d3b4c9a39 ("net/mlx5e: Support flow classification into RSS contexts")
>>
>> Seems that the issue originates in commit 756c41603a18 ("net/mlx5e:
>> ethtool, Support user configuration for RX hash fields"),
>
> Not really; commit 756c41603a18 configures the hash (not flow
> direction), and IPV4_FLOW/IPV6_FLOW are already correct constants for
> IP-based hashes. Moreover, we don't support them anyway, see
> mlx5e_set_rss_hash_opt:
>
> /* RSS does not support anything other than hashing to queues
> * on src IP, dest IP, TCP/UDP src port and TCP/UDP dest
> * port.
> */
> if (flow_type != TCP_V4_FLOW &&
> flow_type != TCP_V6_FLOW &&
> flow_type != UDP_V4_FLOW &&
> flow_type != UDP_V6_FLOW)
> return -EOPNOTSUPP;
>
>> when directly
>> classifying into an RQ, before the multi RSS context support.
>
> Direct classification into an RQ actually works before my fix, because
> it goes to another branch in flow_get_tirn, that doesn't call
> flow_type_to_traffic_type. It's only steering to an RSS context that
> was broken for flow-type ip4/ip6.
>
Thanks for your patch.
LGTM.
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Regards,
Tariq
>>> Signed-off-by: Maxim Mikityanskiy <maxim@isovalent.com>
>>> ---
>>> drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c | 2 ++
>>> 1 file changed, 2 insertions(+)
>>>
>>> diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
>>> index 773624bb2c5d..d68230a7b9f4 100644
>>> --- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
>>> +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
>>> @@ -884,8 +884,10 @@ static int flow_type_to_traffic_type(u32 flow_type)
>>> case ESP_V6_FLOW:
>>> return MLX5_TT_IPV6_IPSEC_ESP;
>>> case IPV4_FLOW:
>>> + case IP_USER_FLOW:
>>> return MLX5_TT_IPV4;
>>> case IPV6_FLOW:
>>> + case IPV6_USER_FLOW:
>>> return MLX5_TT_IPV6;
>>> default:
>>> return -EINVAL;
>>
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [PATCH net] net/mlx5e: Fix ethtool -N flow-type ip4 to RSS context
2025-03-19 12:45 [PATCH net] net/mlx5e: Fix ethtool -N flow-type ip4 to RSS context Maxim Mikityanskiy
` (2 preceding siblings ...)
2025-03-20 7:32 ` Tariq Toukan
@ 2025-03-20 8:25 ` Gal Pressman
2025-03-20 8:28 ` Maxim Mikityanskiy
2025-03-24 20:50 ` patchwork-bot+netdevbpf
4 siblings, 1 reply; 12+ messages in thread
From: Gal Pressman @ 2025-03-20 8:25 UTC (permalink / raw)
To: Maxim Mikityanskiy, Saeed Mahameed, Tariq Toukan
Cc: Leon Romanovsky, Andrew Lunn, David S. Miller, Eric Dumazet,
Jakub Kicinski, Paolo Abeni, netdev, Maxim Mikityanskiy
Hey Maxim!
On 19/03/2025 14:45, Maxim Mikityanskiy wrote:
> diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
> index 773624bb2c5d..d68230a7b9f4 100644
> --- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
> +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
> @@ -884,8 +884,10 @@ static int flow_type_to_traffic_type(u32 flow_type)
> case ESP_V6_FLOW:
> return MLX5_TT_IPV6_IPSEC_ESP;
> case IPV4_FLOW:
> + case IP_USER_FLOW:
They're the same, but I think IPV4_USER_FLOW is the "modern" define that
should be used.
> return MLX5_TT_IPV4;
> case IPV6_FLOW:
> + case IPV6_USER_FLOW:
> return MLX5_TT_IPV6;
> default:
> return -EINVAL;
^ permalink raw reply [flat|nested] 12+ messages in thread* Re: [PATCH net] net/mlx5e: Fix ethtool -N flow-type ip4 to RSS context
2025-03-20 8:25 ` Gal Pressman
@ 2025-03-20 8:28 ` Maxim Mikityanskiy
2025-03-20 8:44 ` Gal Pressman
0 siblings, 1 reply; 12+ messages in thread
From: Maxim Mikityanskiy @ 2025-03-20 8:28 UTC (permalink / raw)
To: Gal Pressman
Cc: Saeed Mahameed, Tariq Toukan, Leon Romanovsky, Andrew Lunn,
David S. Miller, Eric Dumazet, Jakub Kicinski, Paolo Abeni,
netdev, Maxim Mikityanskiy
On Thu, 20 Mar 2025 at 10:25, Gal Pressman <gal@nvidia.com> wrote:
>
> Hey Maxim!
>
> On 19/03/2025 14:45, Maxim Mikityanskiy wrote:
> > diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
> > index 773624bb2c5d..d68230a7b9f4 100644
> > --- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
> > +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
> > @@ -884,8 +884,10 @@ static int flow_type_to_traffic_type(u32 flow_type)
> > case ESP_V6_FLOW:
> > return MLX5_TT_IPV6_IPSEC_ESP;
> > case IPV4_FLOW:
> > + case IP_USER_FLOW:
>
> They're the same, but I think IPV4_USER_FLOW is the "modern" define that
> should be used.
Yeah, I used IP_USER_FLOW for consistency with other places in this
file. If you prefer that, I can resubmit with IPV4_USER_FLOW.
>
> > return MLX5_TT_IPV4;
> > case IPV6_FLOW:
> > + case IPV6_USER_FLOW:
> > return MLX5_TT_IPV6;
> > default:
> > return -EINVAL;
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [PATCH net] net/mlx5e: Fix ethtool -N flow-type ip4 to RSS context
2025-03-20 8:28 ` Maxim Mikityanskiy
@ 2025-03-20 8:44 ` Gal Pressman
2025-03-20 8:58 ` Tariq Toukan
0 siblings, 1 reply; 12+ messages in thread
From: Gal Pressman @ 2025-03-20 8:44 UTC (permalink / raw)
To: Maxim Mikityanskiy
Cc: Saeed Mahameed, Tariq Toukan, Leon Romanovsky, Andrew Lunn,
David S. Miller, Eric Dumazet, Jakub Kicinski, Paolo Abeni,
netdev, Maxim Mikityanskiy
On 20/03/2025 10:28, Maxim Mikityanskiy wrote:
> On Thu, 20 Mar 2025 at 10:25, Gal Pressman <gal@nvidia.com> wrote:
>>
>> Hey Maxim!
>>
>> On 19/03/2025 14:45, Maxim Mikityanskiy wrote:
>>> diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
>>> index 773624bb2c5d..d68230a7b9f4 100644
>>> --- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
>>> +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
>>> @@ -884,8 +884,10 @@ static int flow_type_to_traffic_type(u32 flow_type)
>>> case ESP_V6_FLOW:
>>> return MLX5_TT_IPV6_IPSEC_ESP;
>>> case IPV4_FLOW:
>>> + case IP_USER_FLOW:
>>
>> They're the same, but I think IPV4_USER_FLOW is the "modern" define that
>> should be used.
>
> Yeah, I used IP_USER_FLOW for consistency with other places in this
> file. If you prefer that, I can resubmit with IPV4_USER_FLOW.
I don't mind, up to Tariq.
We can followup with a patch that converts all usages.
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [PATCH net] net/mlx5e: Fix ethtool -N flow-type ip4 to RSS context
2025-03-20 8:44 ` Gal Pressman
@ 2025-03-20 8:58 ` Tariq Toukan
2025-03-20 9:20 ` Maxim Mikityanskiy
0 siblings, 1 reply; 12+ messages in thread
From: Tariq Toukan @ 2025-03-20 8:58 UTC (permalink / raw)
To: Gal Pressman, Maxim Mikityanskiy
Cc: Saeed Mahameed, Tariq Toukan, Leon Romanovsky, Andrew Lunn,
David S. Miller, Eric Dumazet, Jakub Kicinski, Paolo Abeni,
netdev, Maxim Mikityanskiy
On 20/03/2025 10:44, Gal Pressman wrote:
> On 20/03/2025 10:28, Maxim Mikityanskiy wrote:
>> On Thu, 20 Mar 2025 at 10:25, Gal Pressman <gal@nvidia.com> wrote:
>>>
>>> Hey Maxim!
>>>
>>> On 19/03/2025 14:45, Maxim Mikityanskiy wrote:
>>>> diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
>>>> index 773624bb2c5d..d68230a7b9f4 100644
>>>> --- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
>>>> +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
>>>> @@ -884,8 +884,10 @@ static int flow_type_to_traffic_type(u32 flow_type)
>>>> case ESP_V6_FLOW:
>>>> return MLX5_TT_IPV6_IPSEC_ESP;
>>>> case IPV4_FLOW:
>>>> + case IP_USER_FLOW:
>>>
>>> They're the same, but I think IPV4_USER_FLOW is the "modern" define that
>>> should be used.
>>
>> Yeah, I used IP_USER_FLOW for consistency with other places in this
>> file. If you prefer that, I can resubmit with IPV4_USER_FLOW.
>
> I don't mind, up to Tariq.
> We can followup with a patch that converts all usages.
>
Please keep using IP_USER_FLOW for consistency with existing code.
We may converts them all together later.
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [PATCH net] net/mlx5e: Fix ethtool -N flow-type ip4 to RSS context
2025-03-20 8:58 ` Tariq Toukan
@ 2025-03-20 9:20 ` Maxim Mikityanskiy
0 siblings, 0 replies; 12+ messages in thread
From: Maxim Mikityanskiy @ 2025-03-20 9:20 UTC (permalink / raw)
To: Tariq Toukan
Cc: Gal Pressman, Saeed Mahameed, Tariq Toukan, Leon Romanovsky,
Andrew Lunn, David S. Miller, Eric Dumazet, Jakub Kicinski,
Paolo Abeni, netdev, Maxim Mikityanskiy
On Thu, 20 Mar 2025 at 10:58, Tariq Toukan <ttoukan.linux@gmail.com> wrote:
>
>
>
> On 20/03/2025 10:44, Gal Pressman wrote:
> > On 20/03/2025 10:28, Maxim Mikityanskiy wrote:
> >> On Thu, 20 Mar 2025 at 10:25, Gal Pressman <gal@nvidia.com> wrote:
> >>>
> >>> Hey Maxim!
> >>>
> >>> On 19/03/2025 14:45, Maxim Mikityanskiy wrote:
> >>>> diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
> >>>> index 773624bb2c5d..d68230a7b9f4 100644
> >>>> --- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
> >>>> +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c
> >>>> @@ -884,8 +884,10 @@ static int flow_type_to_traffic_type(u32 flow_type)
> >>>> case ESP_V6_FLOW:
> >>>> return MLX5_TT_IPV6_IPSEC_ESP;
> >>>> case IPV4_FLOW:
> >>>> + case IP_USER_FLOW:
> >>>
> >>> They're the same, but I think IPV4_USER_FLOW is the "modern" define that
> >>> should be used.
> >>
> >> Yeah, I used IP_USER_FLOW for consistency with other places in this
> >> file. If you prefer that, I can resubmit with IPV4_USER_FLOW.
> >
> > I don't mind, up to Tariq.
> > We can followup with a patch that converts all usages.
> >
>
> Please keep using IP_USER_FLOW for consistency with existing code.
> We may converts them all together later.
OK, sounds good to me, keeping as is.
Thanks for the reviews!
^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [PATCH net] net/mlx5e: Fix ethtool -N flow-type ip4 to RSS context
2025-03-19 12:45 [PATCH net] net/mlx5e: Fix ethtool -N flow-type ip4 to RSS context Maxim Mikityanskiy
` (3 preceding siblings ...)
2025-03-20 8:25 ` Gal Pressman
@ 2025-03-24 20:50 ` patchwork-bot+netdevbpf
4 siblings, 0 replies; 12+ messages in thread
From: patchwork-bot+netdevbpf @ 2025-03-24 20:50 UTC (permalink / raw)
To: Maxim Mikityanskiy
Cc: saeedm, tariqt, leon, andrew+netdev, davem, edumazet, kuba,
pabeni, netdev, maxim
Hello:
This patch was applied to netdev/net.git (main)
by Jakub Kicinski <kuba@kernel.org>:
On Wed, 19 Mar 2025 14:45:08 +0200 you wrote:
> There commands can be used to add an RSS context and steer some traffic
> into it:
>
> # ethtool -X eth0 context new
> New RSS context is 1
> # ethtool -N eth0 flow-type ip4 dst-ip 1.1.1.1 context 1
> Added rule with ID 1023
>
> [...]
Here is the summary with links:
- [net] net/mlx5e: Fix ethtool -N flow-type ip4 to RSS context
https://git.kernel.org/netdev/net/c/3865bec60683
You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html
^ permalink raw reply [flat|nested] 12+ messages in thread