[dpdk-dev] [RFC PATCH v2 4/5] lib/bpf: use wait event scheme for Rx/Tx iteration

[Ananyev, Konstantin]

> 
> First, fix the bug that keyword const of func arg should be after "*".

I believe there is no bug here.

> This is because const before "*" means the value of "cbi" should not be
> changed. 

Exactly, it says that the function itself will not change the value of "cbi".
It just waits for the value to be changed by someone else.
So please keep parameter list intact.

> But we should monitor that cbi->use changed and then we can
> jump out of loop.
> 
> Second, instead of polling for cbi->use to be updated, use
> wait event scheme.
> 
> Fixes: a93ff62a8938 ("bpf: introduce basic Rx/Tx filters")
> Cc: konstantin.ananyev at intel.com
> Cc: stable at dpdk.org
> 
> Signed-off-by: Feifei Wang <feifei.wang2 at arm.com>
> Reviewed-by: Ruifeng Wang <ruifeng.wang at arm.com>
> ---
>  lib/bpf/bpf_pkt.c | 11 ++++-------
>  1 file changed, 4 insertions(+), 7 deletions(-)
> 
> diff --git a/lib/bpf/bpf_pkt.c b/lib/bpf/bpf_pkt.c
> index 6e8248f0d6..08ed8ff68c 100644
> --- a/lib/bpf/bpf_pkt.c
> +++ b/lib/bpf/bpf_pkt.c
> @@ -111,9 +111,9 @@ bpf_eth_cbi_unuse(struct bpf_eth_cbi *cbi)
>   * Waits till datapath finished using given callback.
>   */
>  static void
> -bpf_eth_cbi_wait(const struct bpf_eth_cbi *cbi)
> +bpf_eth_cbi_wait(struct bpf_eth_cbi *const cbi)
>  {
> -	uint32_t nuse, puse;
> +	uint32_t puse;
> 
>  	/* make sure all previous loads and stores are completed */
>  	rte_smp_mb();
> @@ -122,11 +122,8 @@ bpf_eth_cbi_wait(const struct bpf_eth_cbi *cbi)
> 
>  	/* in use, busy wait till current RX/TX iteration is finished */
>  	if ((puse & BPF_ETH_CBI_INUSE) != 0) {
> -		do {
> -			rte_pause();
> -			rte_compiler_barrier();
> -			nuse = cbi->use;
> -		} while (nuse == puse);
> +		rte_compiler_barrier();
> +		rte_wait_event_32(&cbi->use, UINT_MAX, puse, ==, __ATOMIC_RELAXED);
>  	}
>  }
> 
> --
> 2.25.1