Re: [net-next-2.6 PATCH v2] net: implement mechanism for HW based QOS

[John Fastabend <john.r.fastabend@intel.com>]

On 12/20/2010 2:48 PM, Ben Hutchings wrote:
> On Fri, 2010-12-17 at 08:56 -0800, John Fastabend wrote:
> [...]
>> diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
>> index cc916c5..c5d7949 100644
>> --- a/include/linux/netdevice.h
>> +++ b/include/linux/netdevice.h
>> @@ -646,6 +646,12 @@ struct xps_dev_maps {
>>      (nr_cpu_ids * sizeof(struct xps_map *)))
>>  #endif /* CONFIG_XPS */
>>  
>> +/* HW offloaded queuing disciplines txq count and offset maps */
>> +struct netdev_tc_txq {
>> +	u16 count;
>> +	u16 offset;
>> +};
>> +
>>  /*
>>   * This structure defines the management hooks for network devices.
>>   * The following hooks can be defined; unless noted otherwise, they are
>> @@ -1146,6 +1152,9 @@ struct net_device {
>>  	/* Data Center Bridging netlink ops */
>>  	const struct dcbnl_rtnl_ops *dcbnl_ops;
>>  #endif
>> +	u8 num_tc;
>> +	struct netdev_tc_txq tc_to_txq[16];
>> +	u8 prio_tc_map[16];
> 
> That seems like a fair amount of data to add to every net device,
> considering that users may create e.g. a lot of VLAN devices and they
> won't use this state at all.  Have you considered putting these in a
> structure that is accessed indirectly?

This was Eric's suggestion per his response saves an indirection and avoids false sharing. 

> 
>>  #if defined(CONFIG_FCOE) || defined(CONFIG_FCOE_MODULE)
>>  	/* max exchange id for FCoE LRO by ddp */
>> @@ -1162,6 +1171,57 @@ struct net_device {
>>  #define	NETDEV_ALIGN		32
>>  
>>  static inline
>> +int netdev_get_prio_tc_map(const struct net_device *dev, u32 prio)
>> +{
>> +	return dev->prio_tc_map[prio & 15];
>> +}
>> +
>> +static inline
>> +int netdev_set_prio_tc_map(struct net_device *dev, u8 prio, u8 tc)
>> +{
>> +	if (tc >= dev->num_tc)
>> +		return -EINVAL;
>> +
>> +	dev->prio_tc_map[prio & 15] = tc & 15;
>> +	return 0;
>> +}
> [...]
> 
> Please name the magic numbers 15 and 16.
> 

TC_MAX_QUEUES and TC_BITMASK should work. Thanks for looking over this.

> Ben.
>