netdev - Re: [net-next 04/10] net/mlx5e: Unify constants for WQE_EMPTY_DS

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <250f08af-b02b-f88b-85c7-12ab6d66c874@nvidia.com>
Date:   Tue, 8 Sep 2020 11:59:26 +0300
From:   Maxim Mikityanskiy <maximmi@...dia.com>
To:     Willem de Bruijn <willemdebruijn.kernel@...il.com>,
        Saeed Mahameed <saeedm@...dia.com>
CC:     "David S. Miller" <davem@...emloft.net>,
        Jakub Kicinski <kuba@...nel.org>,
        Network Development <netdev@...r.kernel.org>,
        Maxim Mikityanskiy <maximmi@...lanox.com>
Subject: Re: [net-next 04/10] net/mlx5e: Unify constants for
 WQE_EMPTY_DS_COUNT

On 2020-09-04 18:05, Willem de Bruijn wrote:
> On Thu, Sep 3, 2020 at 11:01 PM Saeed Mahameed <saeedm@...dia.com> wrote:
>>
>> From: Maxim Mikityanskiy <maximmi@...lanox.com>
>>
>> A constant for the number of DS in an empty WQE (i.e. a WQE without data
>> segments) is needed in multiple places (normal TX data path, MPWQE in
>> XDP), but currently we have a constant for XDP and an inline formula in
>> normal TX. This patch introduces a common constant.
>>
>> Additionally, mlx5e_xdp_mpwqe_session_start is converted to use struct
>> assignment, because the code nearby is touched.
>>
>> Signed-off-by: Maxim Mikityanskiy <maximmi@...lanox.com>
>> Signed-off-by: Saeed Mahameed <saeedm@...dia.com>
>> ---
>>   .../net/ethernet/mellanox/mlx5/core/en/txrx.h |  2 ++
>>   .../net/ethernet/mellanox/mlx5/core/en/xdp.c  | 13 +++++++-----
>>   .../net/ethernet/mellanox/mlx5/core/en/xdp.h  | 21 +++++++------------
>>   .../net/ethernet/mellanox/mlx5/core/en_tx.c   |  2 +-
>>   4 files changed, 19 insertions(+), 19 deletions(-)
>>
>> diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h b/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h
>> index d4ee22789ab0..155b89998891 100644
>> --- a/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h
>> +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h
>> @@ -7,6 +7,8 @@
>>   #include "en.h"
>>   #include <linux/indirect_call_wrapper.h>
>>
>> +#define MLX5E_TX_WQE_EMPTY_DS_COUNT (sizeof(struct mlx5e_tx_wqe) / MLX5_SEND_WQE_DS)
>> +
> 
> Out of curiosity, what is the logic for dividing this struct by 16?

The hardware needs the size of a WQE in DS units (16 bytes). An empty 
WQE takes 2 DS (for the ctrl and eth segments), and this macro is this 
initial size of an empty WQE (2 DS). As data segments are added to the 
WQE, it grows, and its size in DS also grows.

> struct mlx5e_tx_wqe {
>          struct mlx5_wqe_ctrl_seg ctrl;
>          struct mlx5_wqe_eth_seg  eth;
>          struct mlx5_wqe_data_seg data[0];
> };
> 
>>   #define INL_HDR_START_SZ (sizeof(((struct mlx5_wqe_eth_seg *)NULL)->inline_hdr.start))
>>
>>   enum mlx5e_icosq_wqe_type {
>> diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
>> index 7fccd2ea7dc9..81cd9a04bcb0 100644
>> --- a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
>> +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c
>> @@ -196,16 +196,19 @@ static void mlx5e_xdp_mpwqe_session_start(struct mlx5e_xdpsq *sq)
>>   {
>>          struct mlx5e_xdp_mpwqe *session = &sq->mpwqe;
>>          struct mlx5e_xdpsq_stats *stats = sq->stats;
>> +       struct mlx5e_tx_wqe *wqe;
>>          u16 pi;
>>
>>          pi = mlx5e_xdpsq_get_next_pi(sq, MLX5E_XDP_MPW_MAX_WQEBBS);
>> -       session->wqe = MLX5E_TX_FETCH_WQE(sq, pi);
>> -
>> +       wqe = MLX5E_TX_FETCH_WQE(sq, pi);
>>          net_prefetchw(session->wqe->data);
> 
> Is this prefetch still valid?

It should be:

net_prefetchw(wqe->data);

Probably a bad rebase.

> And is the temporary variable wqe still
> needed at all?

We want to prefetch as early as possible (before filling *session).

> 
>> -       session->ds_count  = MLX5E_XDP_TX_EMPTY_DS_COUNT;
>> -       session->pkt_count = 0;
>>
>> -       mlx5e_xdp_update_inline_state(sq);
>> +       *session = (struct mlx5e_xdp_mpwqe) {
>> +               .wqe = wqe,
>> +               .ds_count = MLX5E_TX_WQE_EMPTY_DS_COUNT,
>> +               .pkt_count = 0,
>> +               .inline_on = mlx5e_xdp_get_inline_state(sq, session->inline_on),
>> +       };
>>
>>          stats->mpwqe++;
>>   }