netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Roger Quadros <rogerq@kernel.org>
To: Simon Horman <horms@kernel.org>
Cc: "David S. Miller" <davem@davemloft.net>,
	Eric Dumazet <edumazet@google.com>,
	Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
	Siddharth Vadapalli <s-vadapalli@ti.com>,
	Julien Panis <jpanis@baylibre.com>, Andrew Lunn <andrew@lunn.ch>,
	srk@ti.com, vigneshr@ti.com, danishanwar@ti.com,
	pekka Varis <p-varis@ti.com>,
	netdev@vger.kernel.org, linux-kernel@vger.kernel.org,
	linux-omap@vger.kernel.org
Subject: Re: [PATCH net-next v2 6/7] net: ethernet: ti: cpsw_ale: add helper to setup classifier defaults
Date: Mon, 1 Jul 2024 13:32:08 +0300	[thread overview]
Message-ID: <4124798a-cda5-47fe-a67b-e84d72f3ecf8@kernel.org> (raw)
In-Reply-To: <20240701073505.GI17134@kernel.org>


On 01/07/2024 10:35, Simon Horman wrote:
> On Fri, Jun 28, 2024 at 03:01:55PM +0300, Roger Quadros wrote:
>> Default behaviour is to have 8 classifiers to map 8 DSCP/PCP
>> priorities to N receive threads (flows). N depends on number of
>> RX channels enabled for the port.
>>
>> Signed-off-by: Roger Quadros <rogerq@kernel.org>
>> ---
>>  drivers/net/ethernet/ti/cpsw_ale.c | 57 ++++++++++++++++++++++++++++++++++++++
>>  drivers/net/ethernet/ti/cpsw_ale.h |  1 +
>>  2 files changed, 58 insertions(+)
>>
>> diff --git a/drivers/net/ethernet/ti/cpsw_ale.c b/drivers/net/ethernet/ti/cpsw_ale.c
>> index 75a17184d34c..51da527388df 100644
>> --- a/drivers/net/ethernet/ti/cpsw_ale.c
>> +++ b/drivers/net/ethernet/ti/cpsw_ale.c
>> @@ -1650,3 +1650,60 @@ static void cpsw_ale_policer_thread_idx_enable(struct cpsw_ale *ale, u32 idx,
>>  	regmap_field_write(ale->fields[ALE_THREAD_VALUE], thread_id);
>>  	regmap_field_write(ale->fields[ALE_THREAD_ENABLE], enable ? 1 : 0);
>>  }
>> +
>> +/* Disable all policer entries and thread mappings */
>> +static void cpsw_ale_policer_reset(struct cpsw_ale *ale)
>> +{
>> +	int i;
>> +
>> +	for (i = 0; i < ale->params.num_policers ; i++) {
>> +		cpsw_ale_policer_read_idx(ale, i);
>> +		regmap_field_write(ale->fields[POL_PORT_MEN], 0);
>> +		regmap_field_write(ale->fields[POL_PRI_MEN], 0);
>> +		regmap_field_write(ale->fields[POL_OUI_MEN], 0);
>> +		regmap_field_write(ale->fields[POL_DST_MEN], 0);
>> +		regmap_field_write(ale->fields[POL_SRC_MEN], 0);
>> +		regmap_field_write(ale->fields[POL_OVLAN_MEN], 0);
>> +		regmap_field_write(ale->fields[POL_IVLAN_MEN], 0);
>> +		regmap_field_write(ale->fields[POL_ETHERTYPE_MEN], 0);
>> +		regmap_field_write(ale->fields[POL_IPSRC_MEN], 0);
>> +		regmap_field_write(ale->fields[POL_IPDST_MEN], 0);
>> +		regmap_field_write(ale->fields[POL_EN], 0);
>> +		regmap_field_write(ale->fields[POL_RED_DROP_EN], 0);
>> +		regmap_field_write(ale->fields[POL_YELLOW_DROP_EN], 0);
>> +		regmap_field_write(ale->fields[POL_PRIORITY_THREAD_EN], 0);
>> +
>> +		cpsw_ale_policer_thread_idx_enable(ale, i, 0, 0);
>> +	}
>> +}
>> +
>> +/* Default classifer is to map 8 user priorities to N receive channels */
>> +void cpsw_ale_classifier_setup_default(struct cpsw_ale *ale, int num_rx_ch)
>> +{
>> +	int pri, idx;
>> +	int pri_thread_map[8][9] = {	{ 0, 0, 0, 0, 0, 0, 0, 0, },
>> +					{ 0, 0, 0, 0, 1, 1, 1, 1, },
>> +					{ 0, 0, 0, 0, 1, 1, 2, 2, },
>> +					{ 1, 0, 0, 1, 2, 2, 3, 3, },
>> +					{ 1, 0, 0, 1, 2, 3, 4, 4, },
>> +					{ 1, 0, 0, 2, 3, 4, 5, 5, },
>> +					{ 1, 0, 0, 2, 3, 4, 5, 6, },
>> +					{ 2, 0, 1, 3, 4, 5, 6, 7, } };
> 
> Hi Roger,
> 
> Perhaps it is obvious, but I'm wondering if it is appropriate
> to add a comment explaining the layout of the table, especially
> the latter rows where the mapping of priority to receive channel
> seems somewhat non-trivial.

Sure. I took the table straight off from the All new switch book. [1]

Priorities 3 to 7 are straight forward. Priorities 0 to 2 are listed like so in
decreasing order of priority

0 (default)	Best Effort
2 		Spare (undefined)
1 (lowest)	Background

[1] Table 13-2 IEEE 802.1p Recommended Priority Mappings to Class of Service.

> 
>> +
>> +	cpsw_ale_policer_reset(ale);
>> +
>> +	/* use first 8 classifiers to map 8 (DSCP/PCP) priorities to channels */
>> +	for (pri = 0; pri < 8; pri++) {
>> +		idx = pri;
>> +
>> +		/* Classifier 'idx' match on priority 'pri' */
>> +		cpsw_ale_policer_read_idx(ale, idx);
>> +		regmap_field_write(ale->fields[POL_PRI_VAL], pri);
>> +		regmap_field_write(ale->fields[POL_PRI_MEN], 1);
>> +		cpsw_ale_policer_write_idx(ale, idx);
>> +
>> +		/* Map Classifier 'idx' to thread provided by the map */
>> +		cpsw_ale_policer_thread_idx_enable(ale, idx,
>> +						   pri_thread_map[num_rx_ch - 1][pri], 1);
>> +	}
>> +}
>> diff --git a/drivers/net/ethernet/ti/cpsw_ale.h b/drivers/net/ethernet/ti/cpsw_ale.h
>> index 2cb76acc6d16..1e4e9a3dd234 100644
>> --- a/drivers/net/ethernet/ti/cpsw_ale.h
>> +++ b/drivers/net/ethernet/ti/cpsw_ale.h
>> @@ -193,5 +193,6 @@ int cpsw_ale_vlan_add_modify(struct cpsw_ale *ale, u16 vid, int port_mask,
>>  int cpsw_ale_vlan_del_modify(struct cpsw_ale *ale, u16 vid, int port_mask);
>>  void cpsw_ale_set_unreg_mcast(struct cpsw_ale *ale, int unreg_mcast_mask,
>>  			      bool add);
>> +void cpsw_ale_classifier_setup_default(struct cpsw_ale *ale, int num_rx_ch);
>>  
>>  #endif
>>
>> -- 
>> 2.34.1
>>
>>

-- 
cheers,
-roger

  reply	other threads:[~2024-07-01 10:32 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-06-28 12:01 [PATCH net-next v2 0/7] net: ethernet: ti: am65-cpsw: Add multi queue RX support Roger Quadros
2024-06-28 12:01 ` [PATCH net-next v2 1/7] net: ethernet: ti: am65-cpsw: Introduce multi queue Rx Roger Quadros
2024-07-01  7:35   ` Simon Horman
2024-07-01  9:44     ` Roger Quadros
2024-06-28 12:01 ` [PATCH net-next v2 2/7] net: ethernet: ti: cpsw_ale: use regfields for ALE registers Roger Quadros
2024-07-01  7:36   ` Simon Horman
2024-06-28 12:01 ` [PATCH net-next v2 3/7] net: ethernet: ti: cpsw_ale: use regfields for number of Entries and Policers Roger Quadros
2024-07-01  7:36   ` Simon Horman
2024-06-28 12:01 ` [PATCH net-next v2 4/7] net: ethernet: ti: cpsw_ale: add Policer and Thread control register fields Roger Quadros
2024-07-01  7:36   ` Simon Horman
2024-06-28 12:01 ` [PATCH net-next v2 5/7] net: ethernet: ti: cpsw_ale: add policer/classifier helpers Roger Quadros
2024-07-01  7:35   ` Simon Horman
2024-07-01  9:46     ` Roger Quadros
2024-06-28 12:01 ` [PATCH net-next v2 6/7] net: ethernet: ti: cpsw_ale: add helper to setup classifier defaults Roger Quadros
2024-07-01  7:35   ` Simon Horman
2024-07-01 10:32     ` Roger Quadros [this message]
2024-07-02  9:24       ` Simon Horman
2024-06-28 12:01 ` [PATCH net-next v2 7/7] net: ethernet: ti: am65-cpsw: setup priority to flow mapping Roger Quadros
2024-07-01  7:36   ` Simon Horman

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=4124798a-cda5-47fe-a67b-e84d72f3ecf8@kernel.org \
    --to=rogerq@kernel.org \
    --cc=andrew@lunn.ch \
    --cc=danishanwar@ti.com \
    --cc=davem@davemloft.net \
    --cc=edumazet@google.com \
    --cc=horms@kernel.org \
    --cc=jpanis@baylibre.com \
    --cc=kuba@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-omap@vger.kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=p-varis@ti.com \
    --cc=pabeni@redhat.com \
    --cc=s-vadapalli@ti.com \
    --cc=srk@ti.com \
    --cc=vigneshr@ti.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).