Re: [Xen-devel] [PATCH v2 08/25] x86/cpuid: Introduce named feature bitfields

[Andrew Cooper <andrew.cooper3@xxxxxxxxxx>]

On 09/01/17 16:31, Jan Beulich wrote:
>>>> On 09.01.17 at 12:03, <andrew.cooper3@xxxxxxxxxx> wrote:
>> --- a/xen/include/asm-x86/cpuid.h
>> +++ b/xen/include/asm-x86/cpuid.h
>> @@ -92,6 +92,10 @@ struct cpuid_policy
>>       * Everything else should be considered inaccurate, and not necesserily 
>> 0.
>>       */
>>  
>> +#define DECL_BITFIELD(word) _DECL_BITFIELD(FEATURESET_ ## word)
>> +#define _DECL_BITFIELD(x)   __DECL_BITFIELD(x)
>> +#define __DECL_BITFIELD(x)  NAMED_BITFIELD_ ## x
> If this was a C file, I'd be fine with this naming choice, but in a header
> I think it would be better to reduce the chance of name space issues
> and use CPUID_BITFIELD_ ## x instead here (and - obviously - in the
> script).

Ok.

>
>> @@ -291,6 +292,28 @@ def crunch_numbers(state):
>>      for k, v in state.deep_deps.iteritems():
>>          state.deep_deps[k] = featureset_to_uint32s(v, nr_entries)
>>  
>> +    # Calculate the bitfield name declarations
>> +    for word in xrange(nr_entries):
>> +
>> +        names = []
>> +        for bit in xrange(32):
>> +
>> +            name = state.names.get(word * 32 + bit, "")
>> +
>> +            # Prepend an underscore if the name starts with a digit.
>> +            if name and name[0] in "0123456789":
>> +                name = "_" + name
>> +
>> +            # Don't generate names for the duplicate features, or ones
>> +            # fast-forwarded from other state
>> +            if (name.startswith("E1D_") or
>> +                name in ("APIC", "OSXSAVE", "OSPKE")):
>> +                name = ""
>> +
>> +            names.append(name.lower())
>> +
>> +        state.bitfields.append("bool " + ":1, ".join(names) + ":1")
> It's probably my rudimentary Python knowledge that makes me
> confused by the two :1 here. I trust the result is what it needs
> to be.

"separator".join(array) is implode(array, "separator") in other languages.

The final entry in the array still needs a bitfield suffix to be correct C.

>  So with the naming aspect addressed
> Reviewed-by: Jan Beulich <jbeulich@xxxxxxxx>

Thanks.

~Andrew

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
https://lists.xen.org/xen-devel