[xdp-hints] Re: [PATCH net-next v3 3/4] net: stmmac: add Rx HWTS metadata to XDP receive pkt

["Song, Yoong Siang" <yoong.siang.song@intel.com>]

On Friday, April 14, 2023 12:47 AM, Keller, Jacob E <jacob.e.keller@intel.com> wrote:
>On 4/12/2023 6:39 PM, Song, Yoong Siang wrote:
>> On Thursday, April 13, 2023 5:46 AM, Stanislav Fomichev <sdf@google.com>
>wrote:
>>> On Wed, Apr 12, 2023 at 1:56 PM Jacob Keller <jacob.e.keller@intel.com>
>wrote:
>>>>
>>>>
>>>>
>>>> On 4/12/2023 10:00 AM, Stanislav Fomichev wrote:
>>>>> On 04/12, Song Yoong Siang wrote:
>>>>>> Add receive hardware timestamp metadata support via kfunc to XDP
>>>>>> receive packets.
>>>>>>
>>>>>> Suggested-by: Stanislav Fomichev <sdf@google.com>
>>>>>> Signed-off-by: Song Yoong Siang <yoong.siang.song@intel.com>
>>>>>> ---
>>>>>>  drivers/net/ethernet/stmicro/stmmac/stmmac.h  |  3 +++
>>>>>> .../net/ethernet/stmicro/stmmac/stmmac_main.c | 26
>>>>>> ++++++++++++++++++-
>>>>>>  2 files changed, 28 insertions(+), 1 deletion(-)
>>>>>>
>>>>>> diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac.h
>>>>>> b/drivers/net/ethernet/stmicro/stmmac/stmmac.h
>>>>>> index ac8ccf851708..826ac0ec88c6 100644
>>>>>> --- a/drivers/net/ethernet/stmicro/stmmac/stmmac.h
>>>>>> +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac.h
>>>>>> @@ -94,6 +94,9 @@ struct stmmac_rx_buffer {
>>>>>>
>>>>>>  struct stmmac_xdp_buff {
>>>>>>      struct xdp_buff xdp;
>>>>>> +    struct stmmac_priv *priv;
>>>>>> +    struct dma_desc *p;
>>>>>> +    struct dma_desc *np;
>>>>>>  };
>>>>>>
>>>>>>  struct stmmac_rx_queue {
>>>>>> diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
>>>>>> b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
>>>>>> index f7bbdf04d20c..ed660927b628 100644
>>>>>> --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
>>>>>> +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
>>>>>> @@ -5315,10 +5315,15 @@ static int stmmac_rx(struct stmmac_priv
>>>>>> *priv, int limit, u32 queue)
>>>>>>
>>>>>>                      xdp_init_buff(&ctx.xdp, buf_sz, &rx_q->xdp_rxq);
>>>>>>                      xdp_prepare_buff(&ctx.xdp, page_address(buf->page),
>>>>>> -                                     buf->page_offset, buf1_len, false);
>>>>>> +                                     buf->page_offset, buf1_len,
>>>>>> + true);
>>>>>>
>>>>>>                      pre_len = ctx.xdp.data_end - ctx.xdp.data_hard_start -
>>>>>>                                buf->page_offset;
>>>>>> +
>>>>>> +                    ctx.priv = priv;
>>>>>> +                    ctx.p = p;
>>>>>> +                    ctx.np = np;
>>>>>> +
>>>>>>                      skb = stmmac_xdp_run_prog(priv, &ctx.xdp);
>>>>>>                      /* Due xdp_adjust_tail: DMA sync for_device
>>>>>>                       * cover max len CPU touch @@ -7071,6
>>>>>> +7076,23 @@ void stmmac_fpe_handshake(struct stmmac_priv *priv,
>bool enable)
>>>>>>      }
>>>>>>  }
>>>>>>
>>>>>> +static int stmmac_xdp_rx_timestamp(const struct xdp_md *_ctx, u64
>>>>>> +*timestamp) {
>>>>>> +    const struct stmmac_xdp_buff *ctx = (void *)_ctx;
>>>>>> +
>>>>>> +    *timestamp = 0;
>>>>>> +    stmmac_get_rx_hwtstamp(ctx->priv, ctx->p, ctx->np,
>>>>>> + timestamp);
>>>>>> +
>>>>>
>>>>> [..]
>>>>>
>>>>>> +    if (*timestamp)
>>>>>
>>>>> Nit: does it make sense to change stmmac_get_rx_hwtstamp to return
>>>>> bool to indicate success/failure? Then you can do:
>>>>>
>>>>> if (!stmmac_get_rx_hwtstamp())
>>>>>       reutrn -ENODATA;
>>>>
>>>> I would make it return the -ENODATA directly since typically bool
>>>> true/false functions have names like "stmmac_has_rx_hwtstamp" or
>>>> similar name that infers you're answering a true/false question.
>>>>
>>>> That might also let you avoid zeroing the timestamp value first?
>>>
>>> SGTM!
>>
>> stmmac_get_rx_hwtstamp() is used in other places where return value is
>> not needed. Additional if statement checking on return value will add
>> cost, but ignoring return value will hit "unused result" warning.
>>
>
>Isn't unused return values only checked if the function is annotated as
>"__must_check"?
I see. Dint aware that. Thanks for your info.
>
>> I think it will be more make sense if I directly retrieve the
>> timestamp value in stmmac_xdp_rx_timestamp(), instead of reuse
>stmmac_get_rx_hwtstamp().
>>
>
>That makes sense too, the XDP flow is a bit special cased relative to the other
>ones.
Yes, agree.
>
>> Let me send out v4 for review.
>>
>> Thanks & Regards
>> Siang
>>
>>>
>>>> Thanks,
>>>> Jake