Hi all The atomic_t r10_bio->remaining starts at 1 in raid10_handle_discard. It means raid10_handle_discard uses it and sets it to 1. So in fact it starts at 0 and sets to 1 when it's used at first time. Then r10_bio->remaining is added by atomic_inc per usage. It decrements the value when leaving raid10_handle_discard and in every callback function. So the count of r10_bio->remaining in this patch is right. Regards Xiao On 03/16/2021 08:29 AM, kernel test robot wrote: > tree: git://git.kernel.org/pub/scm/linux/kernel/git/song/md.git md-next > head: 49c4d345c81f149a29b3db6e521e5191e55f02b6 > commit: f3cf8c2b2caf6ae77b7c786218d3b060faef63a6 [5/6] md/raid10: improve discard request for far layout > config: x86_64-allyesconfig (attached as .config) > compiler: gcc-9 (Debian 9.3.0-22) 9.3.0 > > If you fix the issue, kindly add following tag as appropriate > Reported-by: kernel test robot > > > "coccinelle warnings: (new ones prefixed by >>)" >>> drivers/md/raid10.c:1526:8-27: atomic_dec_and_test variation before object free at line 1532. > drivers/md/raid10.c:1526:8-27: atomic_dec_and_test variation before object free at line 1537. > > vim +1526 drivers/md/raid10.c > > 1520 > 1521 static void raid_end_discard_bio(struct r10bio *r10bio) > 1522 { > 1523 struct r10conf *conf = r10bio->mddev->private; > 1524 struct r10bio *first_r10bio; > 1525 >> 1526 while (atomic_dec_and_test(&r10bio->remaining)) { > 1527 > 1528 allow_barrier(conf); > 1529 > 1530 if (!test_bit(R10BIO_Discard, &r10bio->state)) { > 1531 first_r10bio = (struct r10bio *)r10bio->master_bio; >> 1532 free_r10bio(r10bio); > 1533 r10bio = first_r10bio; > 1534 } else { > 1535 md_write_end(r10bio->mddev); > 1536 bio_endio(r10bio->master_bio); > 1537 free_r10bio(r10bio); > 1538 break; > 1539 } > 1540 } > 1541 } > 1542 > > --- > 0-DAY CI Kernel Test Service, Intel Corporation > https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org