From 27b13e209ddca5979847a1b57890e0372c1edcee Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Fri, 17 Nov 2023 10:35:22 +0800 Subject: blk-throttle: fix lockdep warning of "cgroup_mutex or RCU read lock required!" Inside blkg_for_each_descendant_pre(), both css_for_each_descendant_pre() and blkg_lookup() requires RCU read lock, and either cgroup_assert_mutex_or_rcu_locked() or rcu_read_lock_held() is called. Fix the warning by adding rcu read lock. Reported-by: Changhui Zhong Signed-off-by: Ming Lei Link: https://lore.kernel.org/r/20231117023527.3188627-2-ming.lei@redhat.com Signed-off-by: Jens Axboe --- block/blk-throttle.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'block') diff --git a/block/blk-throttle.c b/block/blk-throttle.c index 13e4377a8b28..16f5766620a4 100644 --- a/block/blk-throttle.c +++ b/block/blk-throttle.c @@ -1320,6 +1320,7 @@ static void tg_conf_updated(struct throtl_grp *tg, bool global) tg_bps_limit(tg, READ), tg_bps_limit(tg, WRITE), tg_iops_limit(tg, READ), tg_iops_limit(tg, WRITE)); + rcu_read_lock(); /* * Update has_rules[] flags for the updated tg's subtree. A tg is * considered to have rules if either the tg itself or any of its @@ -1347,6 +1348,7 @@ static void tg_conf_updated(struct throtl_grp *tg, bool global) this_tg->latency_target = max(this_tg->latency_target, parent_tg->latency_target); } + rcu_read_unlock(); /* * We're already holding queue_lock and know @tg is valid. Let's -- cgit v1.2.3 From 35a99d6557cacbc177314735342f77a2dda41872 Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Fri, 17 Nov 2023 10:35:23 +0800 Subject: blk-cgroup: avoid to warn !rcu_read_lock_held() in blkg_lookup() So far, all callers either holds spin lock or rcu read explicitly, and most of the caller has added WARN_ON_ONCE(!rcu_read_lock_held()) or lockdep_assert_held(&disk->queue->queue_lock). Remove WARN_ON_ONCE(!rcu_read_lock_held()) from blkg_lookup() for killing the false positive warning from blkg_conf_prep(). Reported-by: Changhui Zhong Fixes: 83462a6c971c ("blkcg: Drop unnecessary RCU read [un]locks from blkg_conf_prep/finish()") Signed-off-by: Ming Lei Link: https://lore.kernel.org/r/20231117023527.3188627-3-ming.lei@redhat.com Signed-off-by: Jens Axboe --- block/blk-cgroup.h | 2 -- 1 file changed, 2 deletions(-) (limited to 'block') diff --git a/block/blk-cgroup.h b/block/blk-cgroup.h index 624c03c8fe64..fd482439afbc 100644 --- a/block/blk-cgroup.h +++ b/block/blk-cgroup.h @@ -249,8 +249,6 @@ static inline struct blkcg_gq *blkg_lookup(struct blkcg *blkcg, { struct blkcg_gq *blkg; - WARN_ON_ONCE(!rcu_read_lock_held()); - if (blkcg == &blkcg_root) return q->root_blkg; -- cgit v1.2.3 From e63a57303599b17290cd8bc48e6f20b24289a8bc Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Fri, 17 Nov 2023 10:35:24 +0800 Subject: blk-cgroup: bypass blkcg_deactivate_policy after destroying blkcg_deactivate_policy() can be called after blkg_destroy_all() returns, and it isn't necessary since blkg_destroy_all has covered policy deactivation. Signed-off-by: Ming Lei Link: https://lore.kernel.org/r/20231117023527.3188627-4-ming.lei@redhat.com Signed-off-by: Jens Axboe --- block/blk-cgroup.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) (limited to 'block') diff --git a/block/blk-cgroup.c b/block/blk-cgroup.c index 4a42ea2972ad..4b48c2c44098 100644 --- a/block/blk-cgroup.c +++ b/block/blk-cgroup.c @@ -577,6 +577,7 @@ static void blkg_destroy_all(struct gendisk *disk) struct request_queue *q = disk->queue; struct blkcg_gq *blkg, *n; int count = BLKG_DESTROY_BATCH_SIZE; + int i; restart: spin_lock_irq(&q->queue_lock); @@ -602,6 +603,18 @@ restart: } } + /* + * Mark policy deactivated since policy offline has been done, and + * the free is scheduled, so future blkcg_deactivate_policy() can + * be bypassed + */ + for (i = 0; i < BLKCG_MAX_POLS; i++) { + struct blkcg_policy *pol = blkcg_policy[i]; + + if (pol) + __clear_bit(pol->plid, q->blkcg_pols); + } + q->root_blkg = NULL; spin_unlock_irq(&q->queue_lock); } -- cgit v1.2.3 From 1898efcdbed32bb1c67269c985a50bab0dbc9493 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Wed, 25 Oct 2023 16:10:18 +0200 Subject: block: update the stable_writes flag in bdev_add Propagate the per-queue stable_write flags into each bdev inode in bdev_add. This makes sure devices that require stable writes have it set for I/O on the block device node as well. Note that this doesn't cover the case of a flag changing on a live device yet. We should handle that as well, but I plan to cover it as part of a more general rework of how changing runtime paramters on block devices works. Fixes: 1cb039f3dc16 ("bdi: replace BDI_CAP_STABLE_WRITES with a queue and a sb flag") Reported-by: Ilya Dryomov Signed-off-by: Christoph Hellwig Link: https://lore.kernel.org/r/20231025141020.192413-3-hch@lst.de Tested-by: Ilya Dryomov Reviewed-by: Darrick J. Wong Signed-off-by: Christian Brauner --- block/bdev.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'block') diff --git a/block/bdev.c b/block/bdev.c index e4cfb7adb645..750aec178b6a 100644 --- a/block/bdev.c +++ b/block/bdev.c @@ -425,6 +425,8 @@ void bdev_set_nr_sectors(struct block_device *bdev, sector_t sectors) void bdev_add(struct block_device *bdev, dev_t dev) { + if (bdev_stable_writes(bdev)) + mapping_set_stable_writes(bdev->bd_inode->i_mapping); bdev->bd_dev = dev; bdev->bd_inode->i_rdev = dev; bdev->bd_inode->i_ino = dev; -- cgit v1.2.3 From c96b8175522a2c52e297ee0a49827a668f95e1e8 Mon Sep 17 00:00:00 2001 From: Damien Le Moal Date: Mon, 20 Nov 2023 16:06:11 +0900 Subject: block: Remove blk_set_runtime_active() The function blk_set_runtime_active() is called only from blk_post_runtime_resume(), so there is no need for that function to be exported. Open-code this function directly in blk_post_runtime_resume() and remove it. Signed-off-by: Damien Le Moal Reviewed-by: Bart Van Assche Reviewed-by: Johannes Thumshirn Link: https://lore.kernel.org/r/20231120070611.33951-1-dlemoal@kernel.org Signed-off-by: Jens Axboe --- block/blk-pm.c | 33 +++++---------------------------- include/linux/blk-pm.h | 1 - 2 files changed, 5 insertions(+), 29 deletions(-) (limited to 'block') diff --git a/block/blk-pm.c b/block/blk-pm.c index 6b72b2e03fc8..42e842074715 100644 --- a/block/blk-pm.c +++ b/block/blk-pm.c @@ -163,38 +163,15 @@ EXPORT_SYMBOL(blk_pre_runtime_resume); * @q: the queue of the device * * Description: - * For historical reasons, this routine merely calls blk_set_runtime_active() - * to do the real work of restarting the queue. It does this regardless of - * whether the device's runtime-resume succeeded; even if it failed the + * Restart the queue of a runtime suspended device. It does this regardless + * of whether the device's runtime-resume succeeded; even if it failed the * driver or error handler will need to communicate with the device. * * This function should be called near the end of the device's - * runtime_resume callback. + * runtime_resume callback to correct queue runtime PM status and re-enable + * peeking requests from the queue. */ void blk_post_runtime_resume(struct request_queue *q) -{ - blk_set_runtime_active(q); -} -EXPORT_SYMBOL(blk_post_runtime_resume); - -/** - * blk_set_runtime_active - Force runtime status of the queue to be active - * @q: the queue of the device - * - * If the device is left runtime suspended during system suspend the resume - * hook typically resumes the device and corrects runtime status - * accordingly. However, that does not affect the queue runtime PM status - * which is still "suspended". This prevents processing requests from the - * queue. - * - * This function can be used in driver's resume hook to correct queue - * runtime PM status and re-enable peeking requests from the queue. It - * should be called before first request is added to the queue. - * - * This function is also called by blk_post_runtime_resume() for - * runtime resumes. It does everything necessary to restart the queue. - */ -void blk_set_runtime_active(struct request_queue *q) { int old_status; @@ -211,4 +188,4 @@ void blk_set_runtime_active(struct request_queue *q) if (old_status != RPM_ACTIVE) blk_clear_pm_only(q); } -EXPORT_SYMBOL(blk_set_runtime_active); +EXPORT_SYMBOL(blk_post_runtime_resume); diff --git a/include/linux/blk-pm.h b/include/linux/blk-pm.h index 2580e05a8ab6..004b38a538ff 100644 --- a/include/linux/blk-pm.h +++ b/include/linux/blk-pm.h @@ -15,7 +15,6 @@ extern int blk_pre_runtime_suspend(struct request_queue *q); extern void blk_post_runtime_suspend(struct request_queue *q, int err); extern void blk_pre_runtime_resume(struct request_queue *q); extern void blk_post_runtime_resume(struct request_queue *q); -extern void blk_set_runtime_active(struct request_queue *q); #else static inline void blk_pm_runtime_init(struct request_queue *q, struct device *dev) {} -- cgit v1.2.3 From 67d995e069535c32829f5d368d919063492cec6e Mon Sep 17 00:00:00 2001 From: Yu Kuai Date: Tue, 28 Nov 2023 20:30:27 +0800 Subject: block: warn once for each partition in bio_check_ro() Commit 1b0a151c10a6 ("blk-core: use pr_warn_ratelimited() in bio_check_ro()") fix message storm by limit the rate, however, there will still be lots of message in the long term. Fix it better by warn once for each partition. Signed-off-by: Yu Kuai Signed-off-by: Christoph Hellwig Link: https://lore.kernel.org/r/20231128123027.971610-3-yukuai1@huaweicloud.com Signed-off-by: Jens Axboe --- block/blk-core.c | 14 +++++++++++--- include/linux/blk_types.h | 1 + 2 files changed, 12 insertions(+), 3 deletions(-) (limited to 'block') diff --git a/block/blk-core.c b/block/blk-core.c index fdf25b8d6e78..2eca76ccf4ee 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -501,9 +501,17 @@ static inline void bio_check_ro(struct bio *bio) if (op_is_write(bio_op(bio)) && bdev_read_only(bio->bi_bdev)) { if (op_is_flush(bio->bi_opf) && !bio_sectors(bio)) return; - pr_warn_ratelimited("Trying to write to read-only block-device %pg\n", - bio->bi_bdev); - /* Older lvm-tools actually trigger this */ + + if (bio->bi_bdev->bd_ro_warned) + return; + + bio->bi_bdev->bd_ro_warned = true; + /* + * Use ioctl to set underlying disk of raid/dm to read-only + * will trigger this. + */ + pr_warn("Trying to write to read-only block-device %pg\n", + bio->bi_bdev); } } diff --git a/include/linux/blk_types.h b/include/linux/blk_types.h index f7d40692dd94..b29ebd53417d 100644 --- a/include/linux/blk_types.h +++ b/include/linux/blk_types.h @@ -70,6 +70,7 @@ struct block_device { #ifdef CONFIG_FAIL_MAKE_REQUEST bool bd_make_it_fail; #endif + bool bd_ro_warned; /* * keep this out-of-line as it's both big and not needed in the fast * path -- cgit v1.2.3 From 3649ff0a0b152b5f00e8f56a5ce0da0945aae278 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Tue, 28 Nov 2023 11:40:19 -0800 Subject: block: Document the role of the two attribute groups It is nontrivial to derive the role of the two attribute groups in source file block/blk-sysfs.c. Hence add a comment that explains their roles. See also commit 6d85ebf95c44 ("blk-sysfs: add a new attr_group for blk_mq"). Cc: Christoph Hellwig Cc: Yu Kuai Signed-off-by: Bart Van Assche Link: https://lore.kernel.org/r/20231128194019.72762-1-bvanassche@acm.org Signed-off-by: Jens Axboe --- block/blk-sysfs.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'block') diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c index 63e481262336..0b2d04766324 100644 --- a/block/blk-sysfs.c +++ b/block/blk-sysfs.c @@ -615,6 +615,7 @@ static ssize_t queue_wb_lat_store(struct request_queue *q, const char *page, QUEUE_RW_ENTRY(queue_wb_lat, "wbt_lat_usec"); #endif +/* Common attributes for bio-based and request-based queues. */ static struct attribute *queue_attrs[] = { &queue_ra_entry.attr, &queue_max_hw_sectors_entry.attr, @@ -659,6 +660,7 @@ static struct attribute *queue_attrs[] = { NULL, }; +/* Request-based queue attributes that are not relevant for bio-based queues. */ static struct attribute *blk_mq_queue_attrs[] = { &queue_requests_entry.attr, &elv_iosched_entry.attr, -- cgit v1.2.3 From 0e4237ae8d159e3d28f3cd83146a46f576ffb586 Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Fri, 1 Dec 2023 16:56:05 +0800 Subject: blk-mq: don't count completed flush data request as inflight in case of quiesce Request queue quiesce may interrupt flush sequence, and the original request may have been marked as COMPLETE, but can't get finished because of queue quiesce. This way is fine from driver viewpoint, because flush sequence is block layer concept, and it isn't related with driver. However, driver(such as dm-rq) can call blk_mq_queue_inflight() to count & drain inflight requests, then the wait & drain never gets done because the completed & not-finished flush request is counted as inflight. Fix this issue by not counting completed flush data request as inflight in case of quiesce. Cc: Mike Snitzer Cc: David Jeffery Cc: John Pittman Signed-off-by: Ming Lei Link: https://lore.kernel.org/r/20231201085605.577730-1-ming.lei@redhat.com Signed-off-by: Jens Axboe --- block/blk-mq.c | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-) (limited to 'block') diff --git a/block/blk-mq.c b/block/blk-mq.c index 900c1be1fee1..ac18f802c027 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -1512,14 +1512,26 @@ void blk_mq_delay_kick_requeue_list(struct request_queue *q, } EXPORT_SYMBOL(blk_mq_delay_kick_requeue_list); +static bool blk_is_flush_data_rq(struct request *rq) +{ + return (rq->rq_flags & RQF_FLUSH_SEQ) && !is_flush_rq(rq); +} + static bool blk_mq_rq_inflight(struct request *rq, void *priv) { /* * If we find a request that isn't idle we know the queue is busy * as it's checked in the iter. * Return false to stop the iteration. + * + * In case of queue quiesce, if one flush data request is completed, + * don't count it as inflight given the flush sequence is suspended, + * and the original flush data request is invisible to driver, just + * like other pending requests because of quiesce */ - if (blk_mq_request_started(rq)) { + if (blk_mq_request_started(rq) && !(blk_queue_quiesced(rq->q) && + blk_is_flush_data_rq(rq) && + blk_mq_request_completed(rq))) { bool *busy = priv; *busy = true; -- cgit v1.2.3