Subject: ia64: implement efficient node_to_cpumask_ptr() Now, arch generic node_to_cpumask_ptr() is slightly inefficient. it makes cpumask_t local variable. generic version -------------------------------------- #define node_to_cpumask_ptr(v, node) \ cpumask_t _##v = node_to_cpumask(node), *v = &_##v but x86 specific node_to_cpumask_ptr() is more efficient. x86 version ------------------------------------ #define node_to_cpumask_ptr(v, node) \ cpumask_t *v = &(node_to_cpumask_map[node]) So, if we make ia64 specific node_to_cpusmask_ptr(), we can reduce stack a bit. Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@xxxxxxxxxxxxxx> CC: Mike Travis <travis@xxxxxxx> --- include/asm-ia64/topology.h | 7 +++++++ 1 file changed, 7 insertions(+) Index: b/include/asm-ia64/topology.h =================================================================== --- a/include/asm-ia64/topology.h +++ b/include/asm-ia64/topology.h @@ -37,6 +37,13 @@ */ #define node_to_cpumask(node) (node_to_cpu_mask[node]) +/* Replace default node_to_cpumask_ptr with optimized version */ +#define node_to_cpumask_ptr(v, node) \ + const cpumask_t *v = &node_to_cpumask(node) + +#define node_to_cpumask_ptr_next(v, node) \ + v = &node_to_cpumask(node) + /* * Returns the number of the node containing Node 'nid'. * Not implemented here. Multi-level hierarchies detected with -- To unsubscribe from this list: send the line "unsubscribe linux-ia64" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
|