tools/sched_ext: scx_flatcg: fix potential stack overflow from VLA in fcg_read_stats

fcg_read_stats() had a VLA allocating 21 * nr_cpus * 8 bytes on the
stack, risking stack overflow on large CPU counts (nr_cpus can be up
to 512).

Fix by using a single heap allocation with the correct size, reusing
it across all stat indices, and freeing it at the end.

Signed-off-by: David Carlier <devnexen@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>

Author: devnexen

tools/sched_ext/scx_flatcg.c

@@ -102,22 +102,27 @@ static float read_cpu_util(__u64 *last_sum, __u64 *last_idle)
 
 static void fcg_read_stats(struct scx_flatcg *skel, __u64 *stats)
 {
-	__u64 cnts[FCG_NR_STATS][skel->rodata->nr_cpus];
+	__u64 *cnts;
 	__u32 idx;
 
+	cnts = calloc(skel->rodata->nr_cpus, sizeof(__u64));
+	if (!cnts)
+		return;
+
 	memset(stats, 0, sizeof(stats[0]) * FCG_NR_STATS);
-	memset(cnts, 0, sizeof(cnts));
 
 	for (idx = 0; idx < FCG_NR_STATS; idx++) {
 		int ret, cpu;
 
 		ret = bpf_map_lookup_elem(bpf_map__fd(skel->maps.stats),
-					  &idx, cnts[idx]);
+					  &idx, cnts);
 		if (ret < 0)
 			continue;
 		for (cpu = 0; cpu < skel->rodata->nr_cpus; cpu++)
-			stats[idx] += cnts[idx][cpu];
+			stats[idx] += cnts[cpu];
 	}
+
+	free(cnts);
 }
 
 int main(int argc, char **argv)