stress-ng: add --permute N option, run permutations of stressors

stress-ng --with cpu,nop,vm --permute 8 will work through all the
permutation of the 3 stressors running 8 instances of each stressor
in each run. This cannot be used with --seq or --all option.

Signed-off-by: Colin Ian King <colin.i.king@gmail.com>

core-bitops.h

@@ -126,6 +126,24 @@ static inline uint32_t OPTIMIZE3 stress_bitreverse32(const uint32_t val)
 #endif
 }
 
+/*
+ *  stress_popcount64()
+ *	population count (count number of 1 bits) in a uint64_t value
+ */
+static inline uint32_t OPTIMIZE3 stress_popcount64(const uint64_t val)
+{
+#if defined(HAVE_BUILTIN_POPCOUNT)
+	return  (uint32_t)__builtin_popcountll(val);
+#else
+	/* Brian Kernighan's count bits */
+	register uint64_t j, v = val;
+
+	for (j = 0; v; j++)
+		v &= v - 1;
+	return j;
+#endif
+}
+
 /*
  *  stress_popcount32()
  *	population count (count number of 1 bits) in a uint32_t value

stress-ng.1

@@ -407,6 +407,12 @@ with Linux 4.7 one needs to have CAP_SYS_ADMIN capabilities for this
 option to work, or adjust  /proc/sys/kernel/perf_event_paranoid to below
 2 to use this without CAP_SYS_ADMIN.
 .TP
+.B \-\-permute N
+run all permutations of the selected stressors with N instances of the
+permutated stressors per run.  If N is less than zero, then the number
+of CPUs online is used for the number of instances. If N is zero, then
+the number of configured CPUs in the system is used.
+.TP
 .B \-q, \-\-quiet
 do not show any output.
 .TP