Re: [patch net-next v3 06/14] rocker: introduce worlds infrastructure

[John Fastabend <john.fastabend@gmail.com>]

On 15-10-06 11:14 PM, Jiri Pirko wrote:
> Wed, Oct 07, 2015 at 03:50:08AM CEST, sfeldma@gmail.com wrote:
> 
> <snip>
> 
>>
>>> Also I wonder how this works when a pkt ingresses a port in mode A and
>>> egresses a port in mode B? What fib/fdb tables does it cross when this
>>> happens? It seems easier to just have two switch devices not a
>>> hybrid. If this per port implementation maps to some hardware that
>>> would be really interesting though.
>>
>> In retrospect, I regret adding the port mode feature to rocker.  I
>> like the world idea, so we can have a device with different
>> pipeline/resources, but we should have locked all ports on a switch to
>> one mode, or even as you hinted at earlier, use a unique sub-device ID
>> for a switch with all ports in a particular mode.  If you want to
>> ports with different worlds, just instantiate a switch in each world.
>> Instantiating new devices is easy.
>>
>> But, now Jiri has locked on to the dynamic port mode idea with pit
>> bull zeal, to the point of being able to switch a port mode at any
>> time from one mode to another from the host.  I just don't see that as
>> a real-world use-case.  Life is too short and we need to be focusing
>> on switchdev features, not refactoring or adding cool but useless
>> features.
> 
> Can can still change this if you want. We can make
> ROCKER_TLV_CMD_PORT_SETTINGS_MODE read-only in hw (As it is in fact now
> as we have only one world).
> 
> Then we add another property:
> static Property rocker_properties[] = {
> 	    DEFINE_PROP_STRING("name", Rocker, name),
> 	    DEFINE_PROP_STRING("world", Rocker, world),
> 		....
> 
> and we use this value in pci_rocker_init instead of r->world_dflt
> 
> Looks straightforward.
> 
> 

I think that would map better to something real hardware would do. In
this case it looks like you reprogrammed the device or pushed a
microcode update at it and loaded a new world. Reporting the string
in ethtool or something might be a nice touch as well.

Just a couple editorial comments, hopefully I'm not sounding critical
of the multiworld stuff I really want to use this! And two we have
devices that can change characteristics at runtime such as the parser,
table type/sizes/layouts, even supported actions so I do want the
sort of dynamic knobs originally proposed just not as strings. But I
think that is our "worlds" vs "profiles" debate.

Thanks!
John