[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <2f84a4ee-8e45-460a-8e62-3f9a48da892a@gmail.com>
Date: Sun, 26 Oct 2025 14:53:51 +0200
From: Tariq Toukan <ttoukan.linux@...il.com>
To: Zhu Yanjun <yanjun.zhu@...ux.dev>, Tariq Toukan <tariqt@...dia.com>,
Eric Dumazet <edumazet@...gle.com>, Jakub Kicinski <kuba@...nel.org>,
Paolo Abeni <pabeni@...hat.com>, Andrew Lunn <andrew+netdev@...n.ch>,
"David S. Miller" <davem@...emloft.net>
Cc: Saeed Mahameed <saeedm@...dia.com>, Leon Romanovsky <leon@...nel.org>,
Mark Bloch <mbloch@...dia.com>, netdev@...r.kernel.org,
linux-rdma@...r.kernel.org, linux-kernel@...r.kernel.org,
Gal Pressman <gal@...dia.com>, Moshe Shemesh <moshe@...dia.com>,
Shay Drori <shayd@...dia.com>
Subject: Re: [PATCH net-next 0/5] net/mlx5: Add balance ID support for LAG
multiplane groups
On 26/10/2025 1:59, Zhu Yanjun wrote:
> 在 2025/10/23 2:16, Tariq Toukan 写道:
>> Hi,
>>
>> This series adds balance ID support for MLX5 LAG in multiplane
>> configurations.
>>
>> See detailed description by Mark below [1].
>>
>> Regards,
>> Tariq
>>
>>
>> [1]
>> The problem: In complex multiplane LAG setups, we need finer control
>> over LAG
>> groups. Currently, devices with the same system image GUID are treated
>> identically, but hardware now supports per-multiplane-group balance
>> IDs that
>> let us differentiate between them. On such systems image system guid
>> isn't enough to decide which devices should be part of which LAG.
>>
>> The solution: Extend the system image GUID with a balance ID byte when
>> the
>> hardware supports it. This gives us the granularity we need without
>> breaking
>> existing deployments.
>>
>> What this series does:
>>
>> 1. Add the hardware interface bits (load_balance_id and lag_per_mp_group)
>> 2. Clean up some duplicate code while we're here
>> 3. Rework the system image GUID infrastructure to handle variable lengths
>> 4. Update PTP clock pairing to use the new approach
>> 5. Restructure capability setting to make room for the new feature
>> 6. Actually implement the balance ID support
>>
>> The key insight is in patch 6: we only append the balance ID when both
>
> In the above, patch 6 is the following patch? It should be patch 5?
>
> [PATCH net-next 5/5] net/mlx5: Add balance ID support for LAG multiplane
> groups
>
> Yanjun.Zhu
>
Right.
Indices shifted because we sent the preparation IFC patch a priori:
137d1a635513 net/mlx5: IFC add balance ID and LAG per MP group bits
>> capabilities are present, so older hardware and software continue to work
>> exactly as before. For newer setups, you get the extra byte that enables
>> per-multiplane-group load balancing.
>>
>> This has been tested with both old and new hardware configurations.
>>
>>
>> Mark Bloch (5):
>> net/mlx5: Use common mlx5_same_hw_devs function
>> net/mlx5: Add software system image GUID infrastructure
>> net/mlx5: Refactor PTP clock devcom pairing
>> net/mlx5: Refactor HCA cap 2 setting
>> net/mlx5: Add balance ID support for LAG multiplane groups
>>
>> drivers/net/ethernet/mellanox/mlx5/core/dev.c | 12 ++++---
>> .../ethernet/mellanox/mlx5/core/en/devlink.c | 7 ++--
>> .../ethernet/mellanox/mlx5/core/en/mapping.c | 13 +++++---
>> .../ethernet/mellanox/mlx5/core/en/mapping.h | 3 +-
>> .../mellanox/mlx5/core/en/rep/bridge.c | 6 +---
>> .../mellanox/mlx5/core/en/tc/int_port.c | 8 +++--
>> .../ethernet/mellanox/mlx5/core/en/tc_ct.c | 11 ++++---
>> .../net/ethernet/mellanox/mlx5/core/en_tc.c | 32 ++++++++++---------
>> .../mellanox/mlx5/core/esw/devlink_port.c | 6 +---
>> .../mellanox/mlx5/core/eswitch_offloads.c | 8 +++--
>> .../net/ethernet/mellanox/mlx5/core/lag/lag.c | 4 ++-
>> .../ethernet/mellanox/mlx5/core/lib/clock.c | 19 ++++++-----
>> .../ethernet/mellanox/mlx5/core/lib/devcom.h | 2 ++
>> .../net/ethernet/mellanox/mlx5/core/main.c | 23 +++++++++----
>> .../ethernet/mellanox/mlx5/core/mlx5_core.h | 2 ++
>> .../net/ethernet/mellanox/mlx5/core/vport.c | 19 +++++++++++
>> include/linux/mlx5/driver.h | 3 ++
>> 17 files changed, 112 insertions(+), 66 deletions(-)
>>
>>
>> base-commit: d550d63d0082268a31e93a10c64cbc2476b98b24
>
>
Powered by blists - more mailing lists