From mboxrd@z Thu Jan 1 00:00:00 1970 From: Coly Li Subject: Re: [PATCH 07/13] block: lift setting the readahead size into the block layer Date: Tue, 22 Sep 2020 17:51:40 +0800 Message-ID: References: <20200921080734.452759-1-hch@lst.de> <20200921080734.452759-8-hch@lst.de> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20200921080734.452759-8-hch@lst.de> Content-Language: en-US To: Christoph Hellwig , Jens Axboe Cc: Song Liu , Hans de Goede , Richard Weinberger , Minchan Kim , Johannes Thumshirn , Justin Sanders , linux-mtd@lists.infradead.org, dm-devel@redhat.com, linux-block@vger.kernel.org, linux-bcache@vger.kernel.org, linux-kernel@vger.kernel.org, drbd-dev@lists.linbit.com, linux-raid@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-mm@kvack.org, cgroups@vger.kernel.org List-Id: dm-devel.ids On 2020/9/21 16:07, Christoph Hellwig wrote: > Drivers shouldn't really mess with the readahead size, as that is a VM > concept. Instead set it based on the optimal I/O size by lifting the > algorithm from the md driver when registering the disk. Also set > bdi->io_pages there as well by applying the same scheme based on > max_sectors. > > Signed-off-by: Christoph Hellwig > Reviewed-by: Johannes Thumshirn For the bcache part, Acked-by: Coly Li Thanks. Coly Li > --- > block/blk-settings.c | 5 ++--- > block/blk-sysfs.c | 9 +++++++++ > drivers/block/aoe/aoeblk.c | 1 - > drivers/block/drbd/drbd_nl.c | 12 +----------- > drivers/md/bcache/super.c | 3 --- > drivers/md/raid0.c | 16 ---------------- > drivers/md/raid10.c | 24 +----------------------- > drivers/md/raid5.c | 13 +------------ > 8 files changed, 14 insertions(+), 69 deletions(-) > > diff --git a/block/blk-settings.c b/block/blk-settings.c > index 76a7e03bcd6cac..01049e9b998f1d 100644 > --- a/block/blk-settings.c > +++ b/block/blk-settings.c > @@ -452,6 +452,8 @@ EXPORT_SYMBOL(blk_limits_io_opt); > void blk_queue_io_opt(struct request_queue *q, unsigned int opt) > { > blk_limits_io_opt(&q->limits, opt); > + q->backing_dev_info->ra_pages = > + max(queue_io_opt(q) * 2 / PAGE_SIZE, VM_READAHEAD_PAGES); > } > EXPORT_SYMBOL(blk_queue_io_opt); > > @@ -628,9 +630,6 @@ void disk_stack_limits(struct gendisk *disk, struct block_device *bdev, > printk(KERN_NOTICE "%s: Warning: Device %s is misaligned\n", > top, bottom); > } > - > - t->backing_dev_info->io_pages = > - t->limits.max_sectors >> (PAGE_SHIFT - 9); > } > EXPORT_SYMBOL(disk_stack_limits); > > diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c > index 81722cdcf0cb21..83915b4a1fc3ad 100644 > --- a/block/blk-sysfs.c > +++ b/block/blk-sysfs.c > @@ -854,6 +854,15 @@ int blk_register_queue(struct gendisk *disk) > percpu_ref_switch_to_percpu(&q->q_usage_counter); > } > > + /* > + * For read-ahead of large files to be effective, we need to read ahead > + * at least twice the optimal I/O size. > + */ > + q->backing_dev_info->ra_pages = > + max(queue_io_opt(q) * 2 / PAGE_SIZE, VM_READAHEAD_PAGES); > + q->backing_dev_info->io_pages = > + queue_max_sectors(q) >> (PAGE_SHIFT - 9); > + > ret = blk_trace_init_sysfs(dev); > if (ret) > return ret; > diff --git a/drivers/block/aoe/aoeblk.c b/drivers/block/aoe/aoeblk.c > index d8cfc233e64b93..c34e71b0c4a98c 100644 > --- a/drivers/block/aoe/aoeblk.c > +++ b/drivers/block/aoe/aoeblk.c > @@ -406,7 +406,6 @@ aoeblk_gdalloc(void *vp) > WARN_ON(d->gd); > WARN_ON(d->flags & DEVFL_UP); > blk_queue_max_hw_sectors(q, BLK_DEF_MAX_SECTORS); > - q->backing_dev_info->ra_pages = SZ_2M / PAGE_SIZE; > blk_queue_io_opt(q, SZ_2M); > d->bufpool = mp; > d->blkq = gd->queue = q; > diff --git a/drivers/block/drbd/drbd_nl.c b/drivers/block/drbd/drbd_nl.c > index aaff5bde391506..f8fb1c9b1bb6c1 100644 > --- a/drivers/block/drbd/drbd_nl.c > +++ b/drivers/block/drbd/drbd_nl.c > @@ -1360,18 +1360,8 @@ static void drbd_setup_queue_param(struct drbd_device *device, struct drbd_backi > decide_on_discard_support(device, q, b, discard_zeroes_if_aligned); > decide_on_write_same_support(device, q, b, o, disable_write_same); > > - if (b) { > + if (b) > blk_stack_limits(&q->limits, &b->limits, 0); > - > - if (q->backing_dev_info->ra_pages != > - b->backing_dev_info->ra_pages) { > - drbd_info(device, "Adjusting my ra_pages to backing device's (%lu -> %lu)\n", > - q->backing_dev_info->ra_pages, > - b->backing_dev_info->ra_pages); > - q->backing_dev_info->ra_pages = > - b->backing_dev_info->ra_pages; > - } > - } > fixup_discard_if_not_supported(q); > fixup_write_zeroes(device, q); > } > diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c > index 48113005ed86ad..6bfa771673623e 100644 > --- a/drivers/md/bcache/super.c > +++ b/drivers/md/bcache/super.c > @@ -1427,9 +1427,6 @@ static int cached_dev_init(struct cached_dev *dc, unsigned int block_size) > if (ret) > return ret; > > - dc->disk.disk->queue->backing_dev_info->ra_pages = > - max(dc->disk.disk->queue->backing_dev_info->ra_pages, > - q->backing_dev_info->ra_pages); > blk_queue_io_opt(dc->disk.disk->queue, > max(queue_io_opt(dc->disk.disk->queue), queue_io_opt(q))); > > diff --git a/drivers/md/raid0.c b/drivers/md/raid0.c > index f54a449f97aa79..aa2d7279176880 100644 > --- a/drivers/md/raid0.c > +++ b/drivers/md/raid0.c > @@ -410,22 +410,6 @@ static int raid0_run(struct mddev *mddev) > mdname(mddev), > (unsigned long long)mddev->array_sectors); > > - if (mddev->queue) { > - /* calculate the max read-ahead size. > - * For read-ahead of large files to be effective, we need to > - * readahead at least twice a whole stripe. i.e. number of devices > - * multiplied by chunk size times 2. > - * If an individual device has an ra_pages greater than the > - * chunk size, then we will not drive that device as hard as it > - * wants. We consider this a configuration error: a larger > - * chunksize should be used in that case. > - */ > - int stripe = mddev->raid_disks * > - (mddev->chunk_sectors << 9) / PAGE_SIZE; > - if (mddev->queue->backing_dev_info->ra_pages < 2* stripe) > - mddev->queue->backing_dev_info->ra_pages = 2* stripe; > - } > - > dump_zones(mddev); > > ret = md_integrity_register(mddev); > diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c > index 9956a04ac13bd6..5d1bdee313ec33 100644 > --- a/drivers/md/raid10.c > +++ b/drivers/md/raid10.c > @@ -3873,19 +3873,6 @@ static int raid10_run(struct mddev *mddev) > mddev->resync_max_sectors = size; > set_bit(MD_FAILFAST_SUPPORTED, &mddev->flags); > > - if (mddev->queue) { > - int stripe = conf->geo.raid_disks * > - ((mddev->chunk_sectors << 9) / PAGE_SIZE); > - > - /* Calculate max read-ahead size. > - * We need to readahead at least twice a whole stripe.... > - * maybe... > - */ > - stripe /= conf->geo.near_copies; > - if (mddev->queue->backing_dev_info->ra_pages < 2 * stripe) > - mddev->queue->backing_dev_info->ra_pages = 2 * stripe; > - } > - > if (md_integrity_register(mddev)) > goto out_free_conf; > > @@ -4723,17 +4710,8 @@ static void end_reshape(struct r10conf *conf) > conf->reshape_safe = MaxSector; > spin_unlock_irq(&conf->device_lock); > > - /* read-ahead size must cover two whole stripes, which is > - * 2 * (datadisks) * chunksize where 'n' is the number of raid devices > - */ > - if (conf->mddev->queue) { > - int stripe = conf->geo.raid_disks * > - ((conf->mddev->chunk_sectors << 9) / PAGE_SIZE); > - stripe /= conf->geo.near_copies; > - if (conf->mddev->queue->backing_dev_info->ra_pages < 2 * stripe) > - conf->mddev->queue->backing_dev_info->ra_pages = 2 * stripe; > + if (conf->mddev->queue) > raid10_set_io_opt(conf); > - } > conf->fullsync = 0; > } > > diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c > index 9a7d1250894ef1..7ace1f76b14736 100644 > --- a/drivers/md/raid5.c > +++ b/drivers/md/raid5.c > @@ -7522,8 +7522,6 @@ static int raid5_run(struct mddev *mddev) > int data_disks = conf->previous_raid_disks - conf->max_degraded; > int stripe = data_disks * > ((mddev->chunk_sectors << 9) / PAGE_SIZE); > - if (mddev->queue->backing_dev_info->ra_pages < 2 * stripe) > - mddev->queue->backing_dev_info->ra_pages = 2 * stripe; > > chunk_size = mddev->chunk_sectors << 9; > blk_queue_io_min(mddev->queue, chunk_size); > @@ -8111,17 +8109,8 @@ static void end_reshape(struct r5conf *conf) > spin_unlock_irq(&conf->device_lock); > wake_up(&conf->wait_for_overlap); > > - /* read-ahead size must cover two whole stripes, which is > - * 2 * (datadisks) * chunksize where 'n' is the number of raid devices > - */ > - if (conf->mddev->queue) { > - int data_disks = conf->raid_disks - conf->max_degraded; > - int stripe = data_disks * ((conf->chunk_sectors << 9) > - / PAGE_SIZE); > - if (conf->mddev->queue->backing_dev_info->ra_pages < 2 * stripe) > - conf->mddev->queue->backing_dev_info->ra_pages = 2 * stripe; > + if (conf->mddev->queue) > raid5_set_io_opt(conf); > - } > } > } > >