kvm-devel

[kvm-devel] [PATCH] MSR_IA32_PERF_STATUS support

[Alexander G. <al...@cs...>]

This patch adds support for the MSR MSR_IA32_PERF_STATUS. This is
required to boot Darwin.

I only implemented dummy support, as the real value would be of no
meaning to any OS I know so far anyway and is only implemented on very
recent Intel CPUs.

[kvm-devel] [resend] [PATCH] MSR_IA32_PERF_STATUS support

[Alexander G. <al...@cs...>]

This is a resend of a mail I sent on 01/10/2008. I did not receive any
response, so I assume it was lost.

The patch is really minor, does not break any compatibility (the MSR is
Intel family 13+) and makes Darwin work.

--- old mail ---

This patch adds support for the MSR MSR_IA32_PERF_STATUS. This is
required to boot Darwin.

I only implemented dummy support, as the real value would be of no
meaning to any OS I know so far anyway and is only implemented on very
recent Intel CPUs.

Re: [kvm-devel] [resend] [PATCH] MSR_IA32_PERF_STATUS support

[Avi K. <av...@qu...>]

Alexander Graf wrote:
> This is a resend of a mail I sent on 01/10/2008. I did not receive any
> response, so I assume it was lost.
>
> The patch is really minor, does not break any compatibility (the MSR is
> Intel family 13+) and makes Darwin work.
>
> --- old mail ---
>
> This patch adds support for the MSR MSR_IA32_PERF_STATUS. This is
> required to boot Darwin.
>
> I only implemented dummy support, as the real value would be of no
> meaning to any OS I know so far anyway and is only implemented on very
> recent Intel CPUs.
>
>   

Looks okay, but please add a signoff line and use the kernel coding 
style (tabs and comment style).



-- 
error compiling committee.c: too many arguments to function

Re: [kvm-devel] [resend] [PATCH] MSR_IA32_PERF_STATUS support

[Alexander G. <al...@cs...>]

Avi Kivity wrote:
> Alexander Graf wrote:
>   
>> This is a resend of a mail I sent on 01/10/2008. I did not receive any
>> response, so I assume it was lost.
>>
>> The patch is really minor, does not break any compatibility (the MSR is
>> Intel family 13+) and makes Darwin work.
>>
>> --- old mail ---
>>
>> This patch adds support for the MSR MSR_IA32_PERF_STATUS. This is
>> required to boot Darwin.
>>
>> I only implemented dummy support, as the real value would be of no
>> meaning to any OS I know so far anyway and is only implemented on very
>> recent Intel CPUs.
>>
>>   
>>     
>
> Looks okay, but please add a signoff line and use the kernel coding 
> style (tabs and comment style).
>
>
>
>   
Ok, should I resend the patch with a signed-off line and proper coding
style?

Re: [kvm-devel] [resend] [PATCH] MSR_IA32_PERF_STATUS support

[Avi K. <av...@qu...>]

Alexander Graf wrote:
> Avi Kivity wrote:
>   
>> Alexander Graf wrote:
>>   
>>     
>>> This is a resend of a mail I sent on 01/10/2008. I did not receive any
>>> response, so I assume it was lost.
>>>
>>> The patch is really minor, does not break any compatibility (the MSR is
>>> Intel family 13+) and makes Darwin work.
>>>
>>> --- old mail ---
>>>
>>> This patch adds support for the MSR MSR_IA32_PERF_STATUS. This is
>>> required to boot Darwin.
>>>
>>> I only implemented dummy support, as the real value would be of no
>>> meaning to any OS I know so far anyway and is only implemented on very
>>> recent Intel CPUs.
>>>
>>>   
>>>     
>>>       
>> Looks okay, but please add a signoff line and use the kernel coding 
>> style (tabs and comment style).
>>
>>
>>
>>   
>>     
> Ok, should I resend the patch with a signed-off line and proper coding
> style?
>   

Yes please.  I'm a little worried that this can confuse other guests, 
but as it's read only and a large variety have run so far without 
complaining, it's probably safe enough.

-- 
error compiling committee.c: too many arguments to function

Re: [kvm-devel] [resend] [PATCH] MSR_IA32_PERF_STATUS support

[Avi K. <av...@qu...>]

Alexander Graf wrote:
> This is a resend of a mail I sent on 01/10/2008. I did not receive any
> response, so I assume it was lost.
>
> The patch is really minor, does not break any compatibility (the MSR is
> Intel family 13+) and makes Darwin work.
>
> --- old mail ---
>
> This patch adds support for the MSR MSR_IA32_PERF_STATUS. This is
> required to boot Darwin.
>
> I only implemented dummy support, as the real value would be of no
> meaning to any OS I know so far anyway and is only implemented on very
> recent Intel CPUs.
>   

I see some documentation of this in 13.1.1, but is there somewhere with 
more information?

-- 
error compiling committee.c: too many arguments to function

Re: [kvm-devel] [resend] [PATCH] MSR_IA32_PERF_STATUS support

[Alexander G. <al...@cs...>]

Avi Kivity wrote:
> Alexander Graf wrote:
>> This is a resend of a mail I sent on 01/10/2008. I did not receive any
>> response, so I assume it was lost.
>>
>> The patch is really minor, does not break any compatibility (the MSR is
>> Intel family 13+) and makes Darwin work.
>>
>> --- old mail ---
>>
>> This patch adds support for the MSR MSR_IA32_PERF_STATUS. This is
>> required to boot Darwin.
>>
>> I only implemented dummy support, as the real value would be of no
>> meaning to any OS I know so far anyway and is only implemented on very
>> recent Intel CPUs.
>>   
>
> I see some documentation of this in 13.1.1, but is there somewhere
> with more information?
>
The only proper source of information I found was the XNU kernel, which
implements a full interpretation of the MSR values. Basically the high
bits are the CPU multiplier, while the low bits represent the number the
TSC advances in one cycle.

There is also a bit which multiplies everything by 2, but maybe the
Intel people on this list can tell us more about it.