On 20 November 2013 09:50, Bernd Oppolzer <[email protected]>wrote:
> KM is known to the machine - if not, this should be S0C1, not S0C6. > > But all function codes are disabled, so function code 18 is not known > to the machine, so the S0C6. Is this correct? > > This is not very nice: there should be mentioned in the PoOp that the > different functions of KM have to be enabled individually, and that > that may be another reason for S0C6. > "Yes we know what KM is, and no you have not enabled the priceless feature that makes me do it for you!" Isn't this exactly as expected? Once enabled, the query will show you which particular cryptic operations are provided for that model. With Linux we learned that even when your hardware can, it depends on the particular libica version whether it will use it. Rob (who considers it unwise that we did not get instrumentation at least to count them)
