| Message ID | 013ca038-559d-0a56-4d49-366c7ec79ac3@kernel.dk (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | bfq: update internal depth state when queue depth changes | expand |
Resent due to Gmail tricking me into using HTML format. :-( Am Fr., 18. Jan. 2019 um 18:34 Uhr schrieb Jens Axboe <axboe@kernel.dk>: > > A previous commit moved the shallow depth and BFQ depth map calculations > to be done at init time, moving it outside of the hotter IO path. This > potentially causes hangs if the users changes the depth of the scheduler > map, by writing to the 'nr_requests' sysfs file for that device. > > Add a blk-mq-sched hook that allows blk-mq to inform the scheduler if > the depth changes, so that the scheduler can update its internal state. > > Reported-by: Paolo Valente <paolo.valente@linaro.org> > Fixes: f0635b8a416e ("bfq: calculate shallow depths at init time") > Signed-off-by: Jens Axboe <axboe@kernel.dk> Works fine here in 4.18 as far as I could test (running a full backup, some compilations and Gentoo package upgrades, filling memory with web browsers to force swapping). $ grep ^ /sys/block/*/queue/nr_requests /sys/block/bcache0/queue/nr_requests:128 /sys/block/bcache1/queue/nr_requests:128 /sys/block/bcache2/queue/nr_requests:128 /sys/block/bcache3/queue/nr_requests:128 /sys/block/bcache4/queue/nr_requests:128 /sys/block/sda/queue/nr_requests:256 /sys/block/sdb/queue/nr_requests:256 /sys/block/sdc/queue/nr_requests:256 /sys/block/sdd/queue/nr_requests:256 /sys/block/sde/queue/nr_requests:256 /sys/block/sdf/queue/nr_requests:256 Tested-by: Kai Krakow <kai@kaishome.de> > --- > > diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c > index cd307767a134..b09589915667 100644 > --- a/block/bfq-iosched.c > +++ b/block/bfq-iosched.c > @@ -5342,7 +5342,7 @@ static unsigned int bfq_update_depths(struct bfq_data *bfqd, > return min_shallow; > } > > -static int bfq_init_hctx(struct blk_mq_hw_ctx *hctx, unsigned int index) > +static void bfq_depth_updated(struct blk_mq_hw_ctx *hctx) > { > struct bfq_data *bfqd = hctx->queue->elevator->elevator_data; > struct blk_mq_tags *tags = hctx->sched_tags; > @@ -5350,6 +5350,11 @@ static int bfq_init_hctx(struct blk_mq_hw_ctx *hctx, unsigned int index) > > min_shallow = bfq_update_depths(bfqd, &tags->bitmap_tags); > sbitmap_queue_min_shallow_depth(&tags->bitmap_tags, min_shallow); > +} > + > +static int bfq_init_hctx(struct blk_mq_hw_ctx *hctx, unsigned int index) > +{ > + bfq_depth_updated(hctx); > return 0; > } > > @@ -5772,6 +5777,7 @@ static struct elevator_type iosched_bfq_mq = { > .requests_merged = bfq_requests_merged, > .request_merged = bfq_request_merged, > .has_work = bfq_has_work, > + .depth_updated = bfq_depth_updated, > .init_hctx = bfq_init_hctx, > .init_sched = bfq_init_queue, > .exit_sched = bfq_exit_queue, > diff --git a/block/blk-mq.c b/block/blk-mq.c > index 3ba37b9e15e9..a047b297ade5 100644 > --- a/block/blk-mq.c > +++ b/block/blk-mq.c > @@ -3101,6 +3101,8 @@ int blk_mq_update_nr_requests(struct request_queue *q, unsigned int nr) > } > if (ret) > break; > + if (q->elevator && q->elevator->type->ops.depth_updated) > + q->elevator->type->ops.depth_updated(hctx); > } > > if (!ret) > diff --git a/include/linux/elevator.h b/include/linux/elevator.h > index 2e9e2763bf47..6e8bc53740f0 100644 > --- a/include/linux/elevator.h > +++ b/include/linux/elevator.h > @@ -31,6 +31,7 @@ struct elevator_mq_ops { > void (*exit_sched)(struct elevator_queue *); > int (*init_hctx)(struct blk_mq_hw_ctx *, unsigned int); > void (*exit_hctx)(struct blk_mq_hw_ctx *, unsigned int); > + void (*depth_updated)(struct blk_mq_hw_ctx *); > > bool (*allow_merge)(struct request_queue *, struct request *, struct bio *); > bool (*bio_merge)(struct blk_mq_hw_ctx *, struct bio *); > > -- > Jens Axboe >
On 1/19/19 4:13 AM, Kai Krakow wrote: > Am Fr., 18. Jan. 2019 um 18:34 Uhr schrieb Jens Axboe <axboe@kernel.dk>: >> >> A previous commit moved the shallow depth and BFQ depth map calculations >> to be done at init time, moving it outside of the hotter IO path. This >> potentially causes hangs if the users changes the depth of the scheduler >> map, by writing to the 'nr_requests' sysfs file for that device. >> >> Add a blk-mq-sched hook that allows blk-mq to inform the scheduler if >> the depth changes, so that the scheduler can update its internal state. >> >> Reported-by: Paolo Valente <paolo.valente@linaro.org> >> Fixes: f0635b8a416e ("bfq: calculate shallow depths at init time") >> Signed-off-by: Jens Axboe <axboe@kernel.dk> > > Works fine here in 4.18 as far as I could test (running a full backup, > some compilations and Gentoo package upgrades, filling memory with web > browsers to force swapping). > > $ grep ^ /sys/block/*/queue/nr_requests > /sys/block/bcache0/queue/nr_requests:128 > /sys/block/bcache1/queue/nr_requests:128 > /sys/block/bcache2/queue/nr_requests:128 > /sys/block/bcache3/queue/nr_requests:128 > /sys/block/bcache4/queue/nr_requests:128 > /sys/block/sda/queue/nr_requests:256 > /sys/block/sdb/queue/nr_requests:256 > /sys/block/sdc/queue/nr_requests:256 > /sys/block/sdd/queue/nr_requests:256 > /sys/block/sde/queue/nr_requests:256 > /sys/block/sdf/queue/nr_requests:256 > > Tested-by: Kai Krakow <kai@kaishome.de> Great, thanks for testing!
Hi Jens, I'm not finding this fix in either current rc or for-5.2. Am I missing something? Thanks, Paolo > Il giorno 19 gen 2019, alle ore 15:36, Jens Axboe <axboe@kernel.dk> ha scritto: > > On 1/19/19 4:13 AM, Kai Krakow wrote: >> Am Fr., 18. Jan. 2019 um 18:34 Uhr schrieb Jens Axboe <axboe@kernel.dk>: >>> >>> A previous commit moved the shallow depth and BFQ depth map calculations >>> to be done at init time, moving it outside of the hotter IO path. This >>> potentially causes hangs if the users changes the depth of the scheduler >>> map, by writing to the 'nr_requests' sysfs file for that device. >>> >>> Add a blk-mq-sched hook that allows blk-mq to inform the scheduler if >>> the depth changes, so that the scheduler can update its internal state. >>> >>> Reported-by: Paolo Valente <paolo.valente@linaro.org> >>> Fixes: f0635b8a416e ("bfq: calculate shallow depths at init time") >>> Signed-off-by: Jens Axboe <axboe@kernel.dk> >> >> Works fine here in 4.18 as far as I could test (running a full backup, >> some compilations and Gentoo package upgrades, filling memory with web >> browsers to force swapping). >> >> $ grep ^ /sys/block/*/queue/nr_requests >> /sys/block/bcache0/queue/nr_requests:128 >> /sys/block/bcache1/queue/nr_requests:128 >> /sys/block/bcache2/queue/nr_requests:128 >> /sys/block/bcache3/queue/nr_requests:128 >> /sys/block/bcache4/queue/nr_requests:128 >> /sys/block/sda/queue/nr_requests:256 >> /sys/block/sdb/queue/nr_requests:256 >> /sys/block/sdc/queue/nr_requests:256 >> /sys/block/sdd/queue/nr_requests:256 >> /sys/block/sde/queue/nr_requests:256 >> /sys/block/sdf/queue/nr_requests:256 >> >> Tested-by: Kai Krakow <kai@kaishome.de> > > Great, thanks for testing! > > -- > Jens Axboe >
On 4/13/19 9:18 AM, Paolo Valente wrote: > Hi Jens, > I'm not finding this fix in either current rc or for-5.2. Am I missing > something? Looks like that slipped through the cracks, I have applied it for 5.1 just now.
diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c index cd307767a134..b09589915667 100644 --- a/block/bfq-iosched.c +++ b/block/bfq-iosched.c @@ -5342,7 +5342,7 @@ static unsigned int bfq_update_depths(struct bfq_data *bfqd, return min_shallow; } -static int bfq_init_hctx(struct blk_mq_hw_ctx *hctx, unsigned int index) +static void bfq_depth_updated(struct blk_mq_hw_ctx *hctx) { struct bfq_data *bfqd = hctx->queue->elevator->elevator_data; struct blk_mq_tags *tags = hctx->sched_tags; @@ -5350,6 +5350,11 @@ static int bfq_init_hctx(struct blk_mq_hw_ctx *hctx, unsigned int index) min_shallow = bfq_update_depths(bfqd, &tags->bitmap_tags); sbitmap_queue_min_shallow_depth(&tags->bitmap_tags, min_shallow); +} + +static int bfq_init_hctx(struct blk_mq_hw_ctx *hctx, unsigned int index) +{ + bfq_depth_updated(hctx); return 0; } @@ -5772,6 +5777,7 @@ static struct elevator_type iosched_bfq_mq = { .requests_merged = bfq_requests_merged, .request_merged = bfq_request_merged, .has_work = bfq_has_work, + .depth_updated = bfq_depth_updated, .init_hctx = bfq_init_hctx, .init_sched = bfq_init_queue, .exit_sched = bfq_exit_queue, diff --git a/block/blk-mq.c b/block/blk-mq.c index 3ba37b9e15e9..a047b297ade5 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -3101,6 +3101,8 @@ int blk_mq_update_nr_requests(struct request_queue *q, unsigned int nr) } if (ret) break; + if (q->elevator && q->elevator->type->ops.depth_updated) + q->elevator->type->ops.depth_updated(hctx); } if (!ret) diff --git a/include/linux/elevator.h b/include/linux/elevator.h index 2e9e2763bf47..6e8bc53740f0 100644 --- a/include/linux/elevator.h +++ b/include/linux/elevator.h @@ -31,6 +31,7 @@ struct elevator_mq_ops { void (*exit_sched)(struct elevator_queue *); int (*init_hctx)(struct blk_mq_hw_ctx *, unsigned int); void (*exit_hctx)(struct blk_mq_hw_ctx *, unsigned int); + void (*depth_updated)(struct blk_mq_hw_ctx *); bool (*allow_merge)(struct request_queue *, struct request *, struct bio *); bool (*bio_merge)(struct blk_mq_hw_ctx *, struct bio *);
A previous commit moved the shallow depth and BFQ depth map calculations to be done at init time, moving it outside of the hotter IO path. This potentially causes hangs if the users changes the depth of the scheduler map, by writing to the 'nr_requests' sysfs file for that device. Add a blk-mq-sched hook that allows blk-mq to inform the scheduler if the depth changes, so that the scheduler can update its internal state. Reported-by: Paolo Valente <paolo.valente@linaro.org> Fixes: f0635b8a416e ("bfq: calculate shallow depths at init time") Signed-off-by: Jens Axboe <axboe@kernel.dk> ---