Re: [PATCH 3/9] sched: add sched_numa_find_nth_cpu()

From: Yury Norov
Date: Thu Feb 16 2023 - 20:39:17 EST


Hi Jakub,

Can you please fold-in the following patch?

Thanks,
Yury

From: Yury Norov <yury.norov@xxxxxxxxx>
Date: Thu, 16 Feb 2023 17:03:30 -0800
Subject: [PATCH] sched/topology: fix KASAN warning in hop_cmp()

Despite that prev_hop is used conditionally on curr_hop is not the
first hop, it's initialized unconditionally.

Because initialization implies dereferencing, it might happen that
the code dereferences uninitialized memory, which has been spotted by
KASAN. Fix it by reorganizing hop_cmp() logic.

Reported-by: Bruno Goncalves <bgoncalv@xxxxxxxxxx>
Signed-off-by: Yury Norov <yury.norov@xxxxxxxxx>
---
kernel/sched/topology.c | 11 ++++++++---
1 file changed, 8 insertions(+), 3 deletions(-)

diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c
index 48838a05c008..c21b8b1f3537 100644
--- a/kernel/sched/topology.c
+++ b/kernel/sched/topology.c
@@ -2081,14 +2081,19 @@ struct __cmp_key {

static int hop_cmp(const void *a, const void *b)
{
- struct cpumask **prev_hop = *((struct cpumask ***)b - 1);
- struct cpumask **cur_hop = *(struct cpumask ***)b;
+ struct cpumask **prev_hop, **cur_hop = *(struct cpumask ***)b;
struct __cmp_key *k = (struct __cmp_key *)a;

if (cpumask_weight_and(k->cpus, cur_hop[k->node]) <= k->cpu)
return 1;

- k->w = (b == k->masks) ? 0 : cpumask_weight_and(k->cpus, prev_hop[k->node]);
+ if (b == k->masks) {
+ k->w = 0;
+ return 0;
+ }
+
+ prev_hop = *((struct cpumask ***)b - 1);
+ k->w = cpumask_weight_and(k->cpus, prev_hop[k->node]);
if (k->w <= k->cpu)
return 0;

--
2.34.1