On Tue, Aug 17, 2021 at 10:58 PM Xiao Ni <xni@xxxxxxxxxx> wrote:
>
> One warning message is triggered like this:
> [ 695.110751] =============================
> [ 695.131439] WARNING: suspicious RCU usage
> [ 695.151389] 4.18.0-319.el8.x86_64+debug #1 Not tainted
> [ 695.174413] -----------------------------
> [ 695.192603] drivers/md/raid10.c:1776 suspicious
> rcu_dereference_check() usage!
> [ 695.225107] other info that might help us debug this:
> [ 695.260940] rcu_scheduler_active = 2, debug_locks = 1
> [ 695.290157] no locks held by mkfs.xfs/10186.
>
> In the first loop of function raid10_handle_discard. It already
> determines which disk need to handle discard request and add the
> rdev reference count rdev->nr_pending. So the conf->mirrors will
> not change until all bios come back from underlayer disks. It
> doesn't need to use rcu_dereference to get rdev.
>
> Fixes: d30588b2731f ('md/raid10: improve raid10 discard request')
> Signed-off-by: Xiao Ni <xni@xxxxxxxxxx>
> Acked-by: Guoqing Jiang <guoqing.jiang@xxxxxxxxx>
Applied to md-fixes. Thanks!
Song
> ---
> V2: Fix comment style problem
>
> drivers/md/raid10.c | 14 ++++++++++----
> 1 file changed, 10 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
> index 16977e8..d5d9233 100644
> --- a/drivers/md/raid10.c
> +++ b/drivers/md/raid10.c
> @@ -1712,6 +1712,11 @@ static int raid10_handle_discard(struct mddev *mddev, struct bio *bio)
> } else
> r10_bio->master_bio = (struct bio *)first_r10bio;
>
> + /*
> + * first select target devices under rcu_lock and
> + * inc refcount on their rdev. Record them by setting
> + * bios[x] to bio
> + */
> rcu_read_lock();
> for (disk = 0; disk < geo->raid_disks; disk++) {
> struct md_rdev *rdev = rcu_dereference(conf->mirrors[disk].rdev);
> @@ -1743,9 +1748,6 @@ static int raid10_handle_discard(struct mddev *mddev, struct bio *bio)
> for (disk = 0; disk < geo->raid_disks; disk++) {
> sector_t dev_start, dev_end;
> struct bio *mbio, *rbio = NULL;
> - struct md_rdev *rdev = rcu_dereference(conf->mirrors[disk].rdev);
> - struct md_rdev *rrdev = rcu_dereference(
> - conf->mirrors[disk].replacement);
>
> /*
> * Now start to calculate the start and end address for each disk.
> @@ -1775,9 +1777,12 @@ static int raid10_handle_discard(struct mddev *mddev, struct bio *bio)
>
> /*
> * It only handles discard bio which size is >= stripe size, so
> - * dev_end > dev_start all the time
> + * dev_end > dev_start all the time.
> + * It doesn't need to use rcu lock to get rdev here. We already
> + * add rdev->nr_pending in the first loop.
> */
> if (r10_bio->devs[disk].bio) {
> + struct md_rdev *rdev = conf->mirrors[disk].rdev;
> mbio = bio_clone_fast(bio, GFP_NOIO, &mddev->bio_set);
> mbio->bi_end_io = raid10_end_discard_request;
> mbio->bi_private = r10_bio;
> @@ -1790,6 +1795,7 @@ static int raid10_handle_discard(struct mddev *mddev, struct bio *bio)
> bio_endio(mbio);
> }
> if (r10_bio->devs[disk].repl_bio) {
> + struct md_rdev *rrdev = conf->mirrors[disk].replacement;
> rbio = bio_clone_fast(bio, GFP_NOIO, &mddev->bio_set);
> rbio->bi_end_io = raid10_end_discard_request;
> rbio->bi_private = r10_bio;
> --
> 2.7.5
>
