Re: [Xen-devel] [PATCH v2] arm/monitor vm-events: Implement guest-request support

[Razvan Cojocaru <rcojocaru@xxxxxxxxxxxxxxx>]

On 02/26/2016 02:20 PM, Corneliu ZUZU wrote:
> On 2/26/2016 2:14 PM, Razvan Cojocaru wrote:
>> On 02/26/2016 02:05 PM, Corneliu ZUZU wrote:
>>> On 2/26/2016 1:56 PM, Jan Beulich wrote:
>>>>>>> On 26.02.16 at 12:07, <czuzu@xxxxxxxxxxxxxxx> wrote:
>>>>> --- a/xen/include/asm-x86/altp2m.h
>>>>> +++ b/xen/include/asm-x86/altp2m.h
>>>>> @@ -15,8 +15,8 @@
>>>>>     * this program; If not, see <http://www.gnu.org/licenses/>.
>>>>>     */
>>>>>    -#ifndef _X86_ALTP2M_H
>>>>> -#define _X86_ALTP2M_H
>>>>> +#ifndef __ASM_X86_ALTP2M_H
>>>>> +#define __ASM_X86_ALTP2M_H
>>>> Unrelated change? (No need to undo, but please don't mix such
>>>> into patches especially when they are quite large already anyway.)
>>> Noted.
>>>
>>>>> @@ -33,5 +33,9 @@ void altp2m_vcpu_initialise(struct vcpu *v);
>>>>>    void altp2m_vcpu_destroy(struct vcpu *v);
>>>>>    void altp2m_vcpu_reset(struct vcpu *v);
>>>>>    -#endif /* _X86_ALTP2M_H */
>>>>> +static inline uint16_t altp2m_vcpu_idx(struct vcpu *v)
>>>> const
>>> 'const', as in:
>>>
>>> +static inline const uint16_t altp2m_vcpu_idx(struct vcpu *v)
>> Since there's no functional difference between returning const uint6_t
>> and plain uint16_t, I assume that Jan meant "const struct vcpu *v".
> 
> I thought the functional difference would be when calling:
> 
> uint16_t idx = altp2m_vcpu_idx(v); // => can subsequently modify idx
> const uint16_t idx = altp2m_vcpu_idx(v); // => cannot subsequently
> modify idx (unless const is casted to non-const)

True, but you can write that code regardless of whether
altp2m_vcpu_idx(v) returns const or non-const uint16_t, so saying "const
uint16_t altp2m_vcpu_idx(struct vcpu *v)" instead of "uint16_t
altp2m_vcpu_idx(struct vcpu *v)" only add syntactic noise with no actual
benefit.


Cheers,
Razvan

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