On 10/05/2024 02:40, Damien Le Moal wrote:
Hi Damien,
block/bdev.c:377:17: warning: symbol 'blockdev_mnt' was not declared.
Should it be static?
block/blk-settings.c:263:9: warning: context imbalance in
'queue_limits_commit_update' - wrong count at exit
block/blk-cgroup.c:811:5: warning: context imbalance in
'blkg_conf_prep' - wrong count at exit
block/blk-cgroup.c:941:9: warning: context imbalance in
'blkg_conf_exit' - different lock contexts for basic block
block/blk-iocost.c:732:9: warning: context imbalance in 'iocg_lock' -
wrong count at exit
block/blk-iocost.c:743:28: warning: context imbalance in 'iocg_unlock'
- unexpected unlock
block/blk-zoned.c:576:30: warning: context imbalance in
'disk_get_and_lock_zone_wplug' - wrong count at exit
block/blk-zoned.c: note: in included file (through include/linux/blkdev.h):
./include/linux/bio.h:592:9: warning: context imbalance in
'blk_zone_wplug_handle_write' - unexpected unlock
block/blk-zoned.c:1721:31: warning: context imbalance in
'blk_revalidate_seq_zone' - unexpected unlock
block/bfq-iosched.c:5498:9: warning: context imbalance in
'bfq_exit_icq' - different lock contexts for basic block
Actually most pre-date v6.9 anyway, apart from the zoned stuff. And the
bdev.c static warning is an outstanding patch, which I replied to.
I will have a look at the zone stuff. This is all from the new addition of zone
write plugging, so all my bad (I did run with lockdep but did not compile test
with sparse).
Can you confirm that you looked to solve these zoned device sparse
warnings and they are difficult to solve?
At a glance, some of those pre-v6.9 issues looks non-trivial to fix,
which may be the reason that they have not been fixed...
Yeah. The context imbalance warnings are sometimes hard to address depending on
the code. Let's see.
I am looking the other sparse warnings in block/ now...
John
