netdev - Re: [PATCH net-next v2 6/7] net: ethernet: ti: cpsw_ale: add helper to setup classifier defaults

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date: Mon, 1 Jul 2024 13:32:08 +0300
From: Roger Quadros <rogerq@...nel.org>
To: Simon Horman <horms@...nel.org>
Cc: "David S. Miller" <davem@...emloft.net>,
 Eric Dumazet <edumazet@...gle.com>, Jakub Kicinski <kuba@...nel.org>,
 Paolo Abeni <pabeni@...hat.com>, Siddharth Vadapalli <s-vadapalli@...com>,
 Julien Panis <jpanis@...libre.com>, Andrew Lunn <andrew@...n.ch>,
 srk@...com, vigneshr@...com, danishanwar@...com, pekka Varis
 <p-varis@...com>, netdev@...r.kernel.org, linux-kernel@...r.kernel.org,
 linux-omap@...r.kernel.org
Subject: Re: [PATCH net-next v2 6/7] net: ethernet: ti: cpsw_ale: add helper
 to setup classifier defaults


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@...nel.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