On 12/01/2026 10:47 am, Jan Beulich wrote:
> On 12.01.2026 11:40, Andrew Cooper wrote:
>> This only exists to have it's type taken, despite there being a perfectly
>> good
>> concrete type to use.
>>
>> No functional change.
>>
>> Signed-off-by: Andrew Cooper <[email protected]>
> Acked-by: Jan Beulich <[email protected]>
Thanks.
>
>> --- a/xen/include/xen/nodemask.h
>> +++ b/xen/include/xen/nodemask.h
>> @@ -67,8 +67,6 @@ typedef struct { DECLARE_BITMAP(bits, MAX_NUMNODES); }
>> nodemask_t;
>>
>> #define nodemask_bits(src) ((src)->bits)
>>
>> -extern nodemask_t _unused_nodemask_arg_;
>> -
>> #define node_set(node, dst) __node_set((node), &(dst))
>> static inline void __node_set(int node, volatile nodemask_t *dstp)
>> {
>> @@ -215,7 +213,7 @@ static inline int __last_node(const nodemask_t *srcp,
>> int nbits)
>>
>> #define nodemask_of_node(node)
>> \
>> ({ \
>> - typeof(_unused_nodemask_arg_) m; \
>> + nodemask_t m; \
>> if (sizeof(m) == sizeof(unsigned long)) { \
>> m.bits[0] = 1UL<<(node); \
>> } else { \
> Hard to see why Linux would have introduced that either. (It still has it,
> btw.)
Yeah, it is a bizarre construct. I only noticed it when trying to do
some more typeof cleanup.
~Andrew
