[dpdk-dev] [PATCH] common/cnxk: consider adjust value for tm burst calculation
Jerin Jacob
jerinjacobk at gmail.com
Wed Nov 3 16:13:55 CET 2021
On Mon, Nov 1, 2021 at 12:59 PM <skoteshwar at marvell.com> wrote:
>
> From: Satha Rao <skoteshwar at marvell.com>
>
> To support lower pps in packet mode we are changing adjust value,
> same needs to be consider for burst size calculations.
>
> When both peak and committed rates requested, then peak rate should
> be larger than committed rate.
>
> Signed-off-by: Satha Rao <skoteshwar at marvell.com>
Acked-by: Jerin Jacob <jerinj at marvell.com>
Applied to dpdk-next-net-mrvl/for-next-net. Thanks
> ---
> drivers/common/cnxk/roc_nix_tm_ops.c | 9 +++++++++
> 1 file changed, 9 insertions(+)
>
> diff --git a/drivers/common/cnxk/roc_nix_tm_ops.c b/drivers/common/cnxk/roc_nix_tm_ops.c
> index eee80d5..b80c77c 100644
> --- a/drivers/common/cnxk/roc_nix_tm_ops.c
> +++ b/drivers/common/cnxk/roc_nix_tm_ops.c
> @@ -124,6 +124,11 @@
> profile->pkt_mode_adj += adjust;
> profile->commit.rate += (adjust * profile->commit.rate);
> profile->peak.rate += (adjust * profile->peak.rate);
> + /* Number of tokens freed after scheduling was proportional
> + * to adjust value
> + */
> + profile->commit.size *= adjust;
> + profile->peak.size *= adjust;
> }
>
> return 0;
> @@ -181,6 +186,10 @@
> return NIX_ERR_TM_INVALID_PEAK_RATE;
> }
>
> + /* If PIR and CIR are requested, PIR should always be larger than CIR */
> + if (peak_rate && commit_rate && (commit_rate > peak_rate))
> + return NIX_ERR_TM_INVALID_PEAK_RATE;
> +
> if (!skip_ins)
> TAILQ_INSERT_TAIL(&nix->shaper_profile_list, profile, shaper);
>
> --
> 1.8.3.1
>
More information about the dev
mailing list