Re: [PATCH 2/3] x86/sev: Add defines for GHCB version 2 MSR protocol requests

[Brijesh Singh]

On 6/23/21 4:32 AM, Borislav Petkov wrote:
> On Wed, Jun 23, 2021 at 08:40:00AM +0200, Joerg Roedel wrote:
>> From: Brijesh Singh <brijesh.singh@xxxxxxx>
>>
>> Add the necessary defines for supporting the GHCB version 2 protocol.
>> This includes defines for:
>>
>> 	- MSR-based AP hlt request/response
>> 	- Hypervisor Feature request/response
>>
>> This is the bare minimum of requests that need to be supported by a GHCB
>> version 2 implementation. There are more requests in the specification,
>> but those depend on Secure Nested Paging support being available.
>>
>> These defines are shared between SEV host and guest support, so they are
>> submitted as an individual patch without users yet to avoid merge
>> conflicts in the future.
>>
>> Signed-off-by: Brijesh Singh <brijesh.singh@xxxxxxx>
>> Signed-off-by: Tom Lendacky <thomas.lendacky@xxxxxxx>
>> Co-developed-by: Tom Lendacky <thomas.lendacky@xxxxxxx>
>> Signed-off-by: Joerg Roedel <jroedel@xxxxxxx>
>> ---
>>  arch/x86/include/asm/sev-common.h | 15 +++++++++++++++
>>  1 file changed, 15 insertions(+)
>>
>> diff --git a/arch/x86/include/asm/sev-common.h b/arch/x86/include/asm/sev-common.h
>> index 1cc9e7dd8107..9aa2f29b4c97 100644
>> --- a/arch/x86/include/asm/sev-common.h
>> +++ b/arch/x86/include/asm/sev-common.h
>> @@ -47,6 +47,21 @@
>>  		(((unsigned long)reg & GHCB_MSR_CPUID_REG_MASK) << GHCB_MSR_CPUID_REG_POS) | \
>>  		(((unsigned long)fn) << GHCB_MSR_CPUID_FUNC_POS))
>>  
>> +/* AP Reset Hold */
>> +#define GHCB_MSR_AP_RESET_HOLD_REQ		0x006
>> +#define GHCB_MSR_AP_RESET_HOLD_RESP		0x007
>> +#define GHCB_MSR_AP_RESET_HOLD_RESULT_POS	12
>> +#define GHCB_MSR_AP_RESET_HOLD_RESULT_MASK	GENMASK_ULL(51, 0)
>> +
>> +/* GHCB Hypervisor Feature Request/Response */
>> +#define GHCB_MSR_HV_FT_REQ			0x080
>> +#define GHCB_MSR_HV_FT_RESP			0x081
>> +#define GHCB_MSR_HV_FT_POS			12
>> +#define GHCB_MSR_HV_FT_MASK			GENMASK_ULL(51, 0)
>> +
>> +#define GHCB_MSR_HV_FT_RESP_VAL(v)		\
>> +	((unsigned long)((v) >> GHCB_MSR_HV_FT_POS) & GHCB_MSR_HV_FT_MASK)
>> +
> Ok, so I took a critical look at this and it doesn't make sense to have
> a differently named define each time you need the [63:12] slice of
> GHCBData. So you can simply use GHCB_DATA(msr_value) instead, see below.
>
> Complaints?

Looks good to me.