Re: [PATCH] PNP: make the resource type an unsigned long

[Rene Herman <rene.herman@keyaccess.nl>]

On 09-08-08 07:25, H. Peter Anvin wrote:

>>> Seems a bit pointless ... either one of those flags is >= 32 bits, in 
>>> which case we need u64, or it's not, in which case there is no reason 
>>> to burden the output with bits we don't need.
>>
>> Yes, it's a not a functional patch -- only a type-consistency one. 
>> Right now we're mixing ints (signed ones even) and unsigned longs and 
>> while in this case that's not a functional problem it's messy and 
>> inconsistent.
>>
>> I agree (as Andrew said earlier as well) that the struct resource 
>> flags member should probably just be a u32 but it's not. Changing that 
>> would be a bigger change than just a simple conistency thing.
>>
> 
> You're going in the wrong direction for consistency.  long is different 
> on 32 and 64 bits, and really should be avoided unless that is intended.

I know and fair enough but changing struct resource is just a bit too 
central for my tastes.

<shrug>

Rene.