Re: [Xen-devel] [PATCHv4 2/5] arm: provide add_sized()

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 11.05.15 at 16:45, <david.vrabel@xxxxxxxxxx> wrote:
> On 11/05/15 15:37, David Vrabel wrote:
>> add_sized(ptr, inc) adds inc to the value at ptr using only the correct
>> size of loads and stores for the type of *ptr.  The add is /not/ atomic.
>> 
>> This is needed for ticket locks to ensure the increment of the head ticket
>> does not affect the tail ticket.
> [...]
>>  
>> +#define add_sized(p, x) ({                                              \
>> +    typeof(*p) __x = (x);                                               \
>> +    switch ( sizeof(*p) )                                               \
>> +    {                                                                   \
>> +    case 1: add_u8_sized((uint8_t *)p, (uint8_t)__x); break;            \
>> +    case 2: add_u16_sized((uint16_t *)p, (uint16_t)__x); break;         \
>> +    case 4: add_u32_sized((uint32_t *)p, (uint32_t)__x); break;         \
>> +    default: __bad_atomic_size(); break;                                \
>> +    }                                                                   \
>> +    __x;                                                                \
> 
> I should have cleaned up this arm implementation in the same way I did
> x86.  Please drop this "__x;" before committing, or I can resubmit if it
> is preferred.

No problem. But I suppose the casts on __x inside the individual
case statements aren't needed either.

Also it looks a little odd that x86 handles 8-byte operands here,
but ARM (and namely ARM64) doesn't.

Jan


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