kernel/workqueue.c:8289:34: warning: array subscript 1 is above array bounds of 'int[1]'
kernel test robot
build bottree: https://git.kernel.org/pub/scm/linux/kernel/git/tj/wq.git for-7.1 head: 41e3ccca00b374b7f39cf68e818b59a921cd7069 commit: 5920d046f7ae3bf9cf51b9d915c1fff13d299d84 [14/18] workqueue: add WQ_AFFN_CACHE_SHARD affinity scope config: nios2-randconfig-001-20260402 (https://download.01.org/0day-ci/archive/20260402/[email protected]/config) compiler: nios2-linux-gcc (GCC) 8.5.0 reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20260402/[email protected]/reproduce)
If you fix the issue in a separate patch/commit (i.e. not just a new version of the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <[email protected]>
| Closes: https://lore.kernel.org/oe-kbuild-all/[email protected]/
All warnings (new ones prefixed by >>):
kernel/workqueue.c: In function 'workqueue_init_topology':
>> kernel/workqueue.c:8289:34: warning: array subscript 1 is above array bounds of 'int[1]' [-Warray-bounds]
cpu_shard_id[c] = cpu_shard_id[cpumask_first(sibling_cpus)];
~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~vim +8289 kernel/workqueue.c
8251
8252 /**
8253 * llc_populate_cpu_shard_id - populate cpu_shard_id[] for each CPU in an LLC pod
8254 * @pod_cpus: the cpumask of CPUs in the LLC pod
8255 * @smt_pods: the SMT pod type, used to identify sibling groups
8256 * @nr_cores: number of distinct cores in @pod_cpus (from llc_count_cores())
8257 *
8258 * Walks @pod_cpus in order. At each SMT group leader, advances to the next
8259 * shard once the current shard is full. Results are written to cpu_shard_id[].
8260 */
8261 static void __init llc_populate_cpu_shard_id(const struct cpumask *pod_cpus,
8262 struct wq_pod_type *smt_pods,
8263 int nr_cores)
8264 {
8265 struct llc_shard_layout layout = llc_calc_shard_layout(nr_cores);
8266 const struct cpumask *sibling_cpus;
8267 /* Count the number of cores in the current shard_id */
8268 int cores_in_shard = 0;
8269 /* This is a cursor for the shards. Go from zero to nr_shards - 1*/
8270 int shard_id = 0;
8271 int c;
8272
8273 /* Iterate at every CPU for a given LLC pod, and assign it a shard */
8274 for_each_cpu(c, pod_cpus) {
8275 sibling_cpus = smt_pods->pod_cpus[smt_pods->cpu_pod[c]];
8276 if (cpumask_first(sibling_cpus) == c) {
8277 /* This is the CPU leader for the siblings */
8278 if (llc_shard_is_full(cores_in_shard, shard_id, &layout)) {
8279 shard_id++;
8280 cores_in_shard = 0;
8281 }
8282 cores_in_shard++;
8283 cpu_shard_id[c] = shard_id;
8284 } else {
8285 /*
8286 * The siblings' shard MUST be the same as the leader.
8287 * never split threads in the same core.
8288 */
> 8289 cpu_shard_id[c] = cpu_shard_id[cpumask_first(sibling_cpus)]; 8290 }
8291 }
8292
8293 WARN_ON_ONCE(shard_id != (layout.nr_shards - 1));
8294 }
8295
Signature · 2 lines
0-DAY CI Kernel Test Service https://github.com/intel/lkp-tests/wiki
Quoted 12 lines · kernel test robot
tree: https://git.kernel.org/pub/scm/linux/kernel/git/tj/wq.git for-7.1 head: 41e3ccca00b374b7f39cf68e818b59a921cd7069 commit: 5920d046f7ae3bf9cf51b9d915c1fff13d299d84 [14/18] workqueue: add WQ_AFFN_CACHE_SHARD affinity scope config: nios2-randconfig-001-20260402 (https://download.01.org/0day-ci/archive/20260402/[email protected]/config) compiler: nios2-linux-gcc (GCC) 8.5.0 reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20260402/[email protected]/reproduce)If you fix the issue in a separate patch/commit (i.e. not just a new version of the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <[email protected]>
| Closes: https://lore.kernel.org/oe-kbuild-all/[email protected]/
Quoted 6 lines
All warnings (new ones prefixed by >>):
kernel/workqueue.c: In function 'workqueue_init_topology':Quoted 1 line
Quoted 1 line
kernel/workqueue.c:8289:34: warning: array subscript 1 is above array bounds of 'int[1]' [-Warray-bounds]cpu_shard_id[c] = cpu_shard_id[cpumask_first(sibling_cpus)]; ~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~
I can finally reproduce it. This is related to nios2 having NR_CPU=1.
In function ‘llc_populate_cpu_shard_id’,
inlined from ‘precompute_cache_shard_ids’ at /home/leit/Devel/linux-next/kernel/workqueue.c:8321:3,
inlined from ‘workqueue_init_topology’ at /home/leit/Devel/linux-next/kernel/workqueue.c:8354:2:
/home/leit/Devel/linux-next/kernel/workqueue.c:8289:55: warning: array subscript 1 is above array bounds of ‘int[1]’ [-Warray-bounds]
8289 | cpu_shard_id[c] = cpu_shard_id[cpumask_first(sibling_cpus)];
| ~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/home/leit/Devel/linux-next/kernel/workqueue.c: In function ‘workqueue_init_topology’:
/home/leit/Devel/linux-next/kernel/workqueue.c:8171:12: note: while referencing ‘cpu_shard_id’
8171 | static int cpu_shard_id[NR_CPUS] __initdata;
| ^~~~~~~~~~~~
make[1]: Leaving directory '/tmp/wq-nios2'
Working on a fix,
Thanks for the report, dear robot :-)
--breno