DPDK patches and discussions
 help / color / mirror / Atom feed
From: Hemant Agrawal <hemant.agrawal@oss.nxp.com>
To: Gagandeep Singh <g.singh@nxp.com>,
	dev@dpdk.org, Hemant Agrawal <hemant.agrawal@nxp.com>,
	Sachin Saxena <sachin.saxena@nxp.com>
Cc: Jun Yang <jun.yang@nxp.com>
Subject: Re: [v1 5/7] net/dpaa2: support FLC stashing API
Date: Tue, 6 Aug 2024 15:27:03 +0530	[thread overview]
Message-ID: <c7ee6018-6938-eea5-74d1-c3607162e1e8@oss.nxp.com> (raw)
In-Reply-To: <20240806084136.3212615-6-g.singh@nxp.com>

Hi Gagan,


I think this patch shall come before the 4th patch:   [v1 4/7] 
crypto/dpaa2_sec: enhance pdcp FLC handling


- Hemant


On 06-08-2024 14:11, Gagandeep Singh wrote:
> From: Jun Yang <jun.yang@nxp.com>
>
> Configure flow steering action with FLC enabled to align stashing
> setting with RSS configuration.
>
> Signed-off-by: Jun Yang <jun.yang@nxp.com>
> Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
> ---
>   drivers/bus/fslmc/portal/dpaa2_hw_pvt.h | 44 +++++++++++++++++++++++++
>   drivers/net/dpaa2/dpaa2_ethdev.c        | 25 +++++++-------
>   2 files changed, 58 insertions(+), 11 deletions(-)
>
> diff --git a/drivers/bus/fslmc/portal/dpaa2_hw_pvt.h b/drivers/bus/fslmc/portal/dpaa2_hw_pvt.h
> index 169c7917ea..4c30e6db18 100644
> --- a/drivers/bus/fslmc/portal/dpaa2_hw_pvt.h
> +++ b/drivers/bus/fslmc/portal/dpaa2_hw_pvt.h
> @@ -179,6 +179,7 @@ struct __rte_cache_aligned dpaa2_queue {
>   	uint16_t resv;
>   	uint64_t offloads;
>   	uint64_t lpbk_cntx;
> +	uint8_t data_stashing_off;
>   };
>   
>   struct swp_active_dqs {
> @@ -463,6 +464,49 @@ struct qbman_result *get_swp_active_dqs(uint16_t dpio_index)
>   	return rte_global_active_dqs_list[dpio_index].global_active_dqs;
>   }
>   
> +/* 00 00 00 - last 6 bit represent data, annotation,
> + * context stashing setting 01 01 00 (0x14)
> + * (in following order ->DS AS CS)
> + * to enable 1 line data, 1 line annotation.
> + * For LX2, this setting should be 01 00 00 (0x10)
> + */
> +#define DPAA2_FLC_STASHING_MAX_BIT_SIZE 2
> +#define DPAA2_FLC_STASHING_MAX_CACHE_LINE \
> +	((1ULL << DPAA2_FLC_STASHING_MAX_BIT_SIZE) - 1)
> +
> +enum dpaa2_flc_stashing_type {
> +	DPAA2_FLC_CNTX_STASHING = 0,
> +	DPAA2_FLC_ANNO_STASHING =
> +		DPAA2_FLC_CNTX_STASHING + DPAA2_FLC_STASHING_MAX_BIT_SIZE,
> +	DPAA2_FLC_DATA_STASHING =
> +		DPAA2_FLC_ANNO_STASHING + DPAA2_FLC_STASHING_MAX_BIT_SIZE,
> +	DPAA2_FLC_END_STASHING =
> +		DPAA2_FLC_DATA_STASHING + DPAA2_FLC_STASHING_MAX_BIT_SIZE
> +};
> +
> +#define DPAA2_STASHING_ALIGN_SIZE (1 << DPAA2_FLC_END_STASHING)
> +
> +static inline void
> +dpaa2_flc_stashing_set(enum dpaa2_flc_stashing_type type,
> +	uint8_t cache_line, uint64_t *flc)
> +{
> +	RTE_ASSERT(cache_line <= DPAA2_FLC_STASHING_MAX_CACHE_LINE);
> +	RTE_ASSERT(type == DPAA2_FLC_CNTX_STASHING ||
> +		type == DPAA2_FLC_ANNO_STASHING ||
> +		type == DPAA2_FLC_DATA_STASHING);
> +
> +	(*flc) &= ~(DPAA2_FLC_STASHING_MAX_CACHE_LINE << type);
> +	(*flc) |= (cache_line << type);
> +}
> +
> +static inline void
> +dpaa2_flc_stashing_clear_all(uint64_t *flc)
> +{
> +	dpaa2_flc_stashing_set(DPAA2_FLC_CNTX_STASHING, 0, flc);
> +	dpaa2_flc_stashing_set(DPAA2_FLC_ANNO_STASHING, 0, flc);
> +	dpaa2_flc_stashing_set(DPAA2_FLC_DATA_STASHING, 0, flc);
> +}
> +
>   static inline
>   void set_swp_active_dqs(uint16_t dpio_index, struct qbman_result *dqs)
>   {
> diff --git a/drivers/net/dpaa2/dpaa2_ethdev.c b/drivers/net/dpaa2/dpaa2_ethdev.c
> index 449bbda7ca..726bc0cf3e 100644
> --- a/drivers/net/dpaa2/dpaa2_ethdev.c
> +++ b/drivers/net/dpaa2/dpaa2_ethdev.c
> @@ -786,17 +786,20 @@ dpaa2_dev_rx_queue_setup(struct rte_eth_dev *dev,
>   	if ((dpaa2_svr_family & 0xffff0000) != SVR_LS2080A) {
>   		options |= DPNI_QUEUE_OPT_FLC;
>   		cfg.flc.stash_control = true;
> -		cfg.flc.value &= 0xFFFFFFFFFFFFFFC0;
> -		/* 00 00 00 - last 6 bit represent annotation, context stashing,
> -		 * data stashing setting 01 01 00 (0x14)
> -		 * (in following order ->DS AS CS)
> -		 * to enable 1 line data, 1 line annotation.
> -		 * For LX2, this setting should be 01 00 00 (0x10)
> -		 */
> -		if ((dpaa2_svr_family & 0xffff0000) == SVR_LX2160A)
> -			cfg.flc.value |= 0x10;
> -		else
> -			cfg.flc.value |= 0x14;
> +		dpaa2_flc_stashing_clear_all(&cfg.flc.value);
> +		if (getenv("DPAA2_DATA_STASHING_OFF")) {
> +			dpaa2_flc_stashing_set(DPAA2_FLC_DATA_STASHING, 0,
> +				&cfg.flc.value);
> +			dpaa2_q->data_stashing_off = 1;
> +		} else {
> +			dpaa2_flc_stashing_set(DPAA2_FLC_DATA_STASHING, 1,
> +				&cfg.flc.value);
> +			dpaa2_q->data_stashing_off = 0;
> +		}
> +		if ((dpaa2_svr_family & 0xffff0000) != SVR_LX2160A) {
> +			dpaa2_flc_stashing_set(DPAA2_FLC_ANNO_STASHING, 1,
> +				&cfg.flc.value);
> +		}
>   	}
>   	ret = dpni_set_queue(dpni, CMD_PRI_LOW, priv->token, DPNI_QUEUE_RX,
>   			     dpaa2_q->tc_index, flow_id, options, &cfg);

  reply	other threads:[~2024-08-06  9:57 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-08-06  8:41 [v1 0/7] DPAA2 crypto changes Gagandeep Singh
2024-08-06  8:41 ` [v1 1/7] crypto/dpaa2_sec: fix memory leak Gagandeep Singh
2024-08-06  8:41 ` [v1 2/7] common/dpaax: caamflib: fix PDCP SNOW-ZUC wdog DECO err Gagandeep Singh
2024-08-06  8:41 ` [v1 3/7] crypto/dpaa2_sec: enhance IPsec RFLC handling Gagandeep Singh
2024-08-06  8:41 ` [v1 4/7] crypto/dpaa2_sec: enhance pdcp FLC handling Gagandeep Singh
2024-08-06  8:41 ` [v1 5/7] net/dpaa2: support FLC stashing API Gagandeep Singh
2024-08-06  9:57   ` Hemant Agrawal [this message]
2024-08-06  8:41 ` [v1 6/7] crypto/dpaa2_sec: remove prefetch code in event mode Gagandeep Singh
2024-08-06  8:41 ` [v1 7/7] crypto/dpaa2_sec: rework debug code Gagandeep Singh
2024-08-06  8:47   ` Hemant Agrawal
2024-08-06 10:27 ` [v2 0/7] DPAA2 crypto changes Gagandeep Singh
2024-08-06 10:27   ` [v2 1/7] crypto/dpaa2_sec: fix memory leak Gagandeep Singh
2024-08-06 10:27   ` [v2 2/7] common/dpaax: caamflib: fix PDCP SNOW-ZUC wdog DECO err Gagandeep Singh
2024-08-06 10:27   ` [v2 3/7] net/dpaa2: support FLC stashing API Gagandeep Singh
2024-08-06 10:27   ` [v2 4/7] crypto/dpaa2_sec: enhance IPsec RFLC handling Gagandeep Singh
2024-08-06 10:27   ` [v2 5/7] crypto/dpaa2_sec: enhance pdcp FLC handling Gagandeep Singh
2024-08-06 10:27   ` [v2 6/7] crypto/dpaa2_sec: remove prefetch code in event mode Gagandeep Singh
2024-08-06 10:27   ` [v2 7/7] crypto/dpaa2_sec: rework debug code Gagandeep Singh
2024-10-04 13:36     ` David Marchand
2024-10-04 16:00       ` [EXTERNAL] " Akhil Goyal
2024-10-07  8:46         ` Gagandeep Singh
2024-10-07  8:49           ` David Marchand
2024-09-18  5:35   ` [EXTERNAL] [v2 0/7] DPAA2 crypto changes Akhil Goyal

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=c7ee6018-6938-eea5-74d1-c3607162e1e8@oss.nxp.com \
    --to=hemant.agrawal@oss.nxp.com \
    --cc=dev@dpdk.org \
    --cc=g.singh@nxp.com \
    --cc=hemant.agrawal@nxp.com \
    --cc=jun.yang@nxp.com \
    --cc=sachin.saxena@nxp.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).