On 10/20/2017 03:00 PM, Cornelia Huck wrote: > On Thu, 19 Oct 2017 18:39:37 +0200 > Halil Pasic <pa...@linux.vnet.ibm.com> wrote: > >> On 09/15/2017 09:27 AM, Cornelia Huck wrote: >>>>> >>>>>> + DEFINE_PROP_UINT8("chpid_type", CcwTesterDevice, chpid_type, >>>>>> + 0x98), >>> This might also need re-evaluation - we should not really need a new >>> chpid type. >>> >> >> I'm back at ccw-tester again (for v2). I've realized we did not agree >> on what to use here (chpid_type). Shall I use 0x25 (Fibre Channel) or >> EMULATED_CCW_3270_CHPID_TYPE, or even 0x32 (virtio-ccw) as a default >> value? And should EMULATED_CCW_3270_CHPID_TYPE be called like that >> (is it really supposed to be specific to 3270? >> >> Sorry I did not notice sooner. > > It might make sense to pick whatever z/VM commonly uses for emulated > devices. (Or ask them if they reserved something explicitly for testing > -- that would be an even better match.) >
OK, I will approach the z/VM guys. About EMULATED_CCW_3270_CHPID_TYPE, does it really make sense to have a separate chpid type for 3270? (You have missed that question, so I'm asking it again). Halil