[SLOF] [PATCH v2 01/20] Add a TPM driver implementation

[Stefan Berger]

On 11/26/2015 12:39 AM, Alexey Kardashevskiy wrote:
> On 11/19/2015 10:58 PM, Thomas Huth wrote:
>> On 19/11/15 12:50, Stefan Berger wrote:
>>> On 11/18/2015 08:07 AM, Thomas Huth wrote:
>>>> On 17/11/15 18:02, Stefan Berger wrote:
>>>>> From: Stefan Berger <stefanb at linux.vnet.ibm.com>
>>>>>
>>>>> This patch adds a TPM driver for the CRQ interface as used by
>>>>> the QEMU PAPR implementation.
>>>>>
>>>>> Signed-off-by: Stefan Berger <stefanb at linux.vnet.ibm.com>
>> [...]
>>>>> +/**** driver structures ****/
>>>>> +
>>>>> +struct tpm_driver tpm_drivers[TPM_NUM_DRIVERS] = {
>>>>> +    [PAPR_DRIVER_IDX] = {
>>>>> +        .setdurations      = spapr_vtpm_set_durations,
>>>>> +        .probe             = spapr_vtpm_probe,
>>>>> +        .init              = spapr_vtpm_init,
>>>>> +        .activate          = spapr_vtpm_activate,
>>>>> +        .ready             = spapr_vtpm_endcycle,
>>>>> +        .senddata          = spapr_vtpm_senddata,
>>>>> +        .transfer          = spapr_vtpm_transfer,
>>>>> +        .waitresponseready = spapr_vtpm_waitresponseready,
>>>>> +        .readresponse      = spapr_vtpm_readresponse,
>>>>> +        .sha1threshold     = 100 * 1024,
>>>>> +        .getbuffersize     = spapr_vtpm_get_buffersize,
>>>>> +        .getstate          = spapr_vtpm_get_state,
>>>>> +        .geterror          = spapr_vtpm_get_error,
>>>>> +    },
>>>>> +};
>>>> Do you plan other TPM drivers in the near future? If not, this struct
>>>> tpm_driver interface with all those function pointers sounds a little
>>>> bit over-engineered to me right now.
>>>
>>> I think it's a nice separation and allows for other drivers if it ever
>>> became necessary.
>>
>> Well, function pointers are always more difficult to read, and to debug
>> later!
>> And since I don't see any other TPM driver implementation coming around
>> in the near future, I really don't think that this is necessary here.
>
> The very first version of this did have 2 TPM drivers so it is 
> possible, I just do not know how actual it is for pseries guests

Recent JS2x blades (other than JS21) do not have the TPM, otherwise we 
would have a driver for it ...

     Stefan