Re: [PATCH] asm-generic: architecture independent readq/writeq for32bit environment

[Hitoshi Mitake]

On Thu, Feb 16, 2012 at 14:59, hpanvin@xxxxxxxxx <hpa@xxxxxxxxx> wrote:
> The point is that the type for all readq/writeq should be u64.

So do you mean that readq/writeq should be defined without
bulid_mmio_read/write?
(in the case of x86)
bulid_mmio_read/write define the parameter type of read/write[bwlq] as void *.

And are there some situations that void * typed parameter of
readq/writeq causes problems?

>
> Hitoshi Mitake <h.mitake@xxxxxxxxx> wrote:
>
>>On Thu, Feb 9, 2012 at 23:45, H. Peter Anvin <hpa@xxxxxxxxx> wrote:
>>> On 02/09/2012 05:37 AM, Hitoshi Mitake wrote:
>>>>
>>>> On Tue, Feb 7, 2012 at 12:47, hpanvin@xxxxxxxxx<hpa@xxxxxxxxx>
>> wrote:
>>>>>
>>>>> Should be volatile u64 * not volatile void *...
>>>>
>>>>
>>>> Is this the type of parameters?
>>>> The parameters of atomic readq/writeq defined in
>>arch/x86/include/asm/io.h
>>>> are defined as void *. I think that atomic readq/writeq and
>>non-atomic
>>>> readq/writeq
>>>> should have same typed parameters and return values.
>>>>
>>>
>>> That sounds like a bug.
>>>
>>>        -hpa
>>>
>>
>>Sorry for my bad writing, I didn't mention:
>>the parameter type == the return value type
>>
>>My intention is:
>>parameter type of atomic readq/writeq == parameter type of non-atomic
>>readq/writeq
>>&&
>>return value type of atomic readq/writeq == return value type of
>>non-atomic readq/writeq
>>
>># == means "should be equal"
>>
>>--
>>Hitoshi Mitake
>>h.mitake@xxxxxxxxx
>
> --
> Sent from my Android phone with K-9 Mail. Please excuse my brevity.



-- 
Hitoshi Mitake
h.mitake@xxxxxxxxx
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/