Re: [PATCH] x86/perf/intel/cqm: Get rid of the silly for_each_cpu lookups

[Vikas Shivappa]

On Wed, 17 Feb 2016, Thomas Gleixner wrote:

On Wed, 17 Feb 2016, Vikas Shivappa wrote:

Yes, please resend the rapl one. perf_uncore is a different trainwreck which I
fixed already:

     lkml.kernel.org/r/20160217132903.767990...@linutronix.de

Ok , will resend the rapl separately.

the fix i sent however was a little different in that it uses a static 
tmp_cpumask to avoid the loop in the cpumask_and_any below -
        while ((n = cpumask_next(n, src1p)) < nr_cpu_ids)
                if (cpumask_test_cpu(n, src2p))

But we have an extra static - static to avoid having it in the 
stack..

copy below -

+/*
+ * Temporary cpumask used during hot cpu notificaiton handling. The usage
+ * is serialized by hot cpu locks.
+ */
+static cpumask_t tmp_cpumask;
+

 static void rapl_cpu_init(int cpu)
 {
-       int i, phys_id = topology_physical_package_id(cpu);
-
-       /* check if phys_is is already covered */
-       for_each_cpu(i, &rapl_cpu_mask) {
-               if (phys_id == topology_physical_package_id(i))
-                       return;
-       }
-       /* was not found, so add it */
-       cpumask_set_cpu(cpu, &rapl_cpu_mask);
+       /* check if cpu's package is already covered.If not, add it.*/
+       cpumask_and(&tmp_cpumask, &rapl_cpu_mask, topology_core_cpumask(cpu));
+       if (cpumask_empty(&tmp_cpumask))
+               cpumask_set_cpu(cpu, &rapl_cpu_mask);
 }

Thanks,
Vikas


Thanks,

        tglx