diff options
Diffstat (limited to 'block')
38 files changed, 1478 insertions, 1381 deletions
diff --git a/block/Kconfig b/block/Kconfig index c6ce41a5e5b2..d5d4197b7ed2 100644 --- a/block/Kconfig +++ b/block/Kconfig @@ -35,6 +35,9 @@ config BLK_CGROUP_RWSTAT config BLK_DEV_BSG_COMMON tristate +config BLK_ICQ + bool + config BLK_DEV_BSGLIB bool "Block layer SG support v4 helper lib" select BLK_DEV_BSG_COMMON diff --git a/block/Kconfig.iosched b/block/Kconfig.iosched index 885fee86dfca..615516146086 100644 --- a/block/Kconfig.iosched +++ b/block/Kconfig.iosched @@ -18,6 +18,7 @@ config MQ_IOSCHED_KYBER config IOSCHED_BFQ tristate "BFQ I/O scheduler" + select BLK_ICQ help BFQ I/O scheduler for BLK-MQ. BFQ distributes the bandwidth of of the device among all processes according to their weights, diff --git a/block/Makefile b/block/Makefile index 44df57e562bf..f38eaa612929 100644 --- a/block/Makefile +++ b/block/Makefile @@ -5,7 +5,7 @@ obj-y := bdev.o fops.o bio.o elevator.o blk-core.o blk-sysfs.o \ blk-flush.o blk-settings.o blk-ioc.o blk-map.o \ - blk-exec.o blk-merge.o blk-timeout.o \ + blk-merge.o blk-timeout.o \ blk-lib.o blk-mq.o blk-mq-tag.o blk-stat.o \ blk-mq-sysfs.o blk-mq-cpumap.o blk-mq-sched.o ioctl.o \ genhd.o ioprio.o badblocks.o partitions/ blk-rq-qos.o \ diff --git a/block/bdev.c b/block/bdev.c index b4dab2fb6a74..8bf93a19041b 100644 --- a/block/bdev.c +++ b/block/bdev.c @@ -665,7 +665,7 @@ static void blkdev_flush_mapping(struct block_device *bdev) static int blkdev_get_whole(struct block_device *bdev, fmode_t mode) { struct gendisk *disk = bdev->bd_disk; - int ret = 0; + int ret; if (disk->fops->open) { ret = disk->fops->open(bdev, mode); @@ -750,21 +750,11 @@ struct block_device *blkdev_get_no_open(dev_t dev) if (!kobject_get_unless_zero(&bdev->bd_device.kobj)) bdev = NULL; iput(inode); - - if (!bdev) - return NULL; - if ((bdev->bd_disk->flags & GENHD_FL_HIDDEN) || - !try_module_get(bdev->bd_disk->fops->owner)) { - put_device(&bdev->bd_device); - return NULL; - } - return bdev; } void blkdev_put_no_open(struct block_device *bdev) { - module_put(bdev->bd_disk->fops->owner); put_device(&bdev->bd_device); } @@ -820,12 +810,14 @@ struct block_device *blkdev_get_by_dev(dev_t dev, fmode_t mode, void *holder) ret = -ENXIO; if (!disk_live(disk)) goto abort_claiming; + if (!try_module_get(disk->fops->owner)) + goto abort_claiming; if (bdev_is_partition(bdev)) ret = blkdev_get_part(bdev, mode); else ret = blkdev_get_whole(bdev, mode); if (ret) - goto abort_claiming; + goto put_module; if (mode & FMODE_EXCL) { bd_finish_claiming(bdev, holder); @@ -837,7 +829,7 @@ struct block_device *blkdev_get_by_dev(dev_t dev, fmode_t mode, void *holder) * used in blkdev_get/put(). */ if ((mode & FMODE_WRITE) && !bdev->bd_write_holder && - (disk->flags & GENHD_FL_BLOCK_EVENTS_ON_EXCL_WRITE)) { + (disk->event_flags & DISK_EVENT_FLAG_BLOCK_ON_EXCL_WRITE)) { bdev->bd_write_holder = true; unblock_events = false; } @@ -847,7 +839,8 @@ struct block_device *blkdev_get_by_dev(dev_t dev, fmode_t mode, void *holder) if (unblock_events) disk_unblock_events(disk); return bdev; - +put_module: + module_put(disk->fops->owner); abort_claiming: if (mode & FMODE_EXCL) bd_abort_claiming(bdev, holder); @@ -956,20 +949,21 @@ void blkdev_put(struct block_device *bdev, fmode_t mode) blkdev_put_whole(bdev, mode); mutex_unlock(&disk->open_mutex); + module_put(disk->fops->owner); blkdev_put_no_open(bdev); } EXPORT_SYMBOL(blkdev_put); /** - * lookup_bdev - lookup a struct block_device by name - * @pathname: special file representing the block device - * @dev: return value of the block device's dev_t + * lookup_bdev() - Look up a struct block_device by name. + * @pathname: Name of the block device in the filesystem. + * @dev: Pointer to the block device's dev_t, if found. * * Lookup the block device's dev_t at @pathname in the current - * namespace if possible and return it by @dev. + * namespace if possible and return it in @dev. * - * RETURNS: - * 0 if succeeded, errno otherwise. + * Context: May sleep. + * Return: 0 if succeeded, negative errno otherwise. */ int lookup_bdev(const char *pathname, dev_t *dev) { diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c index fec18118dc30..0c612a911696 100644 --- a/block/bfq-iosched.c +++ b/block/bfq-iosched.c @@ -433,26 +433,21 @@ static struct bfq_io_cq *icq_to_bic(struct io_cq *icq) /** * bfq_bic_lookup - search into @ioc a bic associated to @bfqd. - * @bfqd: the lookup key. - * @ioc: the io_context of the process doing I/O. * @q: the request queue. */ -static struct bfq_io_cq *bfq_bic_lookup(struct bfq_data *bfqd, - struct io_context *ioc, - struct request_queue *q) +static struct bfq_io_cq *bfq_bic_lookup(struct request_queue *q) { - if (ioc) { - unsigned long flags; - struct bfq_io_cq *icq; + struct bfq_io_cq *icq; + unsigned long flags; - spin_lock_irqsave(&q->queue_lock, flags); - icq = icq_to_bic(ioc_lookup_icq(ioc, q)); - spin_unlock_irqrestore(&q->queue_lock, flags); + if (!current->io_context) + return NULL; - return icq; - } + spin_lock_irqsave(&q->queue_lock, flags); + icq = icq_to_bic(ioc_lookup_icq(q)); + spin_unlock_irqrestore(&q->queue_lock, flags); - return NULL; + return icq; } /* @@ -565,26 +560,134 @@ static struct request *bfq_choose_req(struct bfq_data *bfqd, } } +#define BFQ_LIMIT_INLINE_DEPTH 16 + +#ifdef CONFIG_BFQ_GROUP_IOSCHED +static bool bfqq_request_over_limit(struct bfq_queue *bfqq, int limit) +{ + struct bfq_data *bfqd = bfqq->bfqd; + struct bfq_entity *entity = &bfqq->entity; + struct bfq_entity *inline_entities[BFQ_LIMIT_INLINE_DEPTH]; + struct bfq_entity **entities = inline_entities; + int depth, level; + int class_idx = bfqq->ioprio_class - 1; + struct bfq_sched_data *sched_data; + unsigned long wsum; + bool ret = false; + + if (!entity->on_st_or_in_serv) + return false; + + /* +1 for bfqq entity, root cgroup not included */ + depth = bfqg_to_blkg(bfqq_group(bfqq))->blkcg->css.cgroup->level + 1; + if (depth > BFQ_LIMIT_INLINE_DEPTH) { + entities = kmalloc_array(depth, sizeof(*entities), GFP_NOIO); + if (!entities) + return false; + } + + spin_lock_irq(&bfqd->lock); + sched_data = entity->sched_data; + /* Gather our ancestors as we need to traverse them in reverse order */ + level = 0; + for_each_entity(entity) { + /* + * If at some level entity is not even active, allow request + * queueing so that BFQ knows there's work to do and activate + * entities. + */ + if (!entity->on_st_or_in_serv) + goto out; + /* Uh, more parents than cgroup subsystem thinks? */ + if (WARN_ON_ONCE(level >= depth)) + break; + entities[level++] = entity; + } + WARN_ON_ONCE(level != depth); + for (level--; level >= 0; level--) { + entity = entities[level]; + if (level > 0) { + wsum = bfq_entity_service_tree(entity)->wsum; + } else { + int i; + /* + * For bfqq itself we take into account service trees + * of all higher priority classes and multiply their + * weights so that low prio queue from higher class + * gets more requests than high prio queue from lower + * class. + */ + wsum = 0; + for (i = 0; i <= class_idx; i++) { + wsum = wsum * IOPRIO_BE_NR + + sched_data->service_tree[i].wsum; + } + } + limit = DIV_ROUND_CLOSEST(limit * entity->weight, wsum); + if (entity->allocated >= limit) { + bfq_log_bfqq(bfqq->bfqd, bfqq, + "too many requests: allocated %d limit %d level %d", + entity->allocated, limit, level); + ret = true; + break; + } + } +out: + spin_unlock_irq(&bfqd->lock); + if (entities != inline_entities) + kfree(entities); + return ret; +} +#else +static bool bfqq_request_over_limit(struct bfq_queue *bfqq, int limit) +{ + return false; +} +#endif + /* * Async I/O can easily starve sync I/O (both sync reads and sync * writes), by consuming all tags. Similarly, storms of sync writes, * such as those that sync(2) may trigger, can starve sync reads. * Limit depths of async I/O and sync writes so as to counter both * problems. + * + * Also if a bfq queue or its parent cgroup consume more tags than would be + * appropriate for their weight, we trim the available tag depth to 1. This + * avoids a situation where one cgroup can starve another cgroup from tags and + * thus block service differentiation among cgroups. Note that because the + * queue / cgroup already has many requests allocated and queued, this does not + * significantly affect service guarantees coming from the BFQ scheduling + * algorithm. */ static void bfq_limit_depth(unsigned int op, struct blk_mq_alloc_data *data) { struct bfq_data *bfqd = data->q->elevator->elevator_data; + struct bfq_io_cq *bic = bfq_bic_lookup(data->q); + struct bfq_queue *bfqq = bic ? bic_to_bfqq(bic, op_is_sync(op)) : NULL; + int depth; + unsigned limit = data->q->nr_requests; + + /* Sync reads have full depth available */ + if (op_is_sync(op) && !op_is_write(op)) { + depth = 0; + } else { + depth = bfqd->word_depths[!!bfqd->wr_busy_queues][op_is_sync(op)]; + limit = (limit * depth) >> bfqd->full_depth_shift; + } - if (op_is_sync(op) && !op_is_write(op)) - return; - - data->shallow_depth = - bfqd->word_depths[!!bfqd->wr_busy_queues][op_is_sync(op)]; + /* + * Does queue (or any parent entity) exceed number of requests that + * should be available to it? Heavily limit depth so that it cannot + * consume more available requests and thus starve other entities. + */ + if (bfqq && bfqq_request_over_limit(bfqq, limit)) + depth = 1; bfq_log(bfqd, "[%s] wr_busy %d sync %d depth %u", - __func__, bfqd->wr_busy_queues, op_is_sync(op), - data->shallow_depth); + __func__, bfqd->wr_busy_queues, op_is_sync(op), depth); + if (depth) + data->shallow_depth = depth; } static struct bfq_queue * @@ -1113,7 +1216,8 @@ bfq_bfqq_resume_state(struct bfq_queue *bfqq, struct bfq_data *bfqd, static int bfqq_process_refs(struct bfq_queue *bfqq) { - return bfqq->ref - bfqq->allocated - bfqq->entity.on_st_or_in_serv - + return bfqq->ref - bfqq->entity.allocated - + bfqq->entity.on_st_or_in_serv - (bfqq->weight_counter != NULL) - bfqq->stable_ref; } @@ -1982,20 +2086,19 @@ static void bfq_update_io_intensity(struct bfq_queue *bfqq, u64 now_ns) * aspect, see the comments on the choice of the queue for injection * in bfq_select_queue(). * - * Turning back to the detection of a waker queue, a queue Q is deemed - * as a waker queue for bfqq if, for three consecutive times, bfqq - * happens to become non empty right after a request of Q has been - * completed. In this respect, even if bfqq is empty, we do not check - * for a waker if it still has some in-flight I/O. In fact, in this - * case bfqq is actually still being served by the drive, and may - * receive new I/O on the completion of some of the in-flight - * requests. In particular, on the first time, Q is tentatively set as - * a candidate waker queue, while on the third consecutive time that Q - * is detected, the field waker_bfqq is set to Q, to confirm that Q is - * a waker queue for bfqq. These detection steps are performed only if - * bfqq has a long think time, so as to make it more likely that - * bfqq's I/O is actually being blocked by a synchronization. This - * last filter, plus the above three-times requirement, make false + * Turning back to the detection of a waker queue, a queue Q is deemed as a + * waker queue for bfqq if, for three consecutive times, bfqq happens to become + * non empty right after a request of Q has been completed within given + * timeout. In this respect, even if bfqq is empty, we do not check for a waker + * if it still has some in-flight I/O. In fact, in this case bfqq is actually + * still being served by the drive, and may receive new I/O on the completion + * of some of the in-flight requests. In particular, on the first time, Q is + * tentatively set as a candidate waker queue, while on the third consecutive + * time that Q is detected, the field waker_bfqq is set to Q, to confirm that Q + * is a waker queue for bfqq. These detection steps are performed only if bfqq + * has a long think time, so as to make it more likely that bfqq's I/O is + * actually being blocked by a synchronization. This last filter, plus the + * above three-times requirement and time limit for detection, make false * positives less likely. * * NOTE @@ -2019,6 +2122,8 @@ static void bfq_update_io_intensity(struct bfq_queue *bfqq, u64 now_ns) static void bfq_check_waker(struct bfq_data *bfqd, struct bfq_queue *bfqq, u64 now_ns) { + char waker_name[MAX_BFQQ_NAME_LENGTH]; + if (!bfqd->last_completed_rq_bfqq || bfqd->last_completed_rq_bfqq == bfqq || bfq_bfqq_has_short_ttime(bfqq) || @@ -2027,8 +2132,16 @@ static void bfq_check_waker(struct bfq_data *bfqd, struct bfq_queue *bfqq, bfqd->last_completed_rq_bfqq == bfqq->waker_bfqq) return; + /* + * We reset waker detection logic also if too much time has passed + * since the first detection. If wakeups are rare, pointless idling + * doesn't hurt throughput that much. The condition below makes sure + * we do not uselessly idle blocking waker in more than 1/64 cases. + */ if (bfqd->last_completed_rq_bfqq != - bfqq->tentative_waker_bfqq) { + bfqq->tentative_waker_bfqq || + now_ns > bfqq->waker_detection_started + + 128 * (u64)bfqd->bfq_slice_idle) { /* * First synchronization detected with a * candidate waker queue, or with a different @@ -2037,12 +2150,19 @@ static void bfq_check_waker(struct bfq_data *bfqd, struct bfq_queue *bfqq, bfqq->tentative_waker_bfqq = bfqd->last_completed_rq_bfqq; bfqq->num_waker_detections = 1; + bfqq->waker_detection_started = now_ns; + bfq_bfqq_name(bfqq->tentative_waker_bfqq, waker_name, + MAX_BFQQ_NAME_LENGTH); + bfq_log_bfqq(bfqd, bfqq, "set tenative waker %s", waker_name); } else /* Same tentative waker queue detected again */ bfqq->num_waker_detections++; if (bfqq->num_waker_detections == 3) { bfqq->waker_bfqq = bfqd->last_completed_rq_bfqq; bfqq->tentative_waker_bfqq = NULL; + bfq_bfqq_name(bfqq->waker_bfqq, waker_name, + MAX_BFQQ_NAME_LENGTH); + bfq_log_bfqq(bfqd, bfqq, "set waker %s", waker_name); /* * If the waker queue disappears, then @@ -2332,7 +2452,7 @@ static bool bfq_bio_merge(struct request_queue *q, struct bio *bio, * returned by bfq_bic_lookup does not go away before * bfqd->lock is taken. */ - struct bfq_io_cq *bic = bfq_bic_lookup(bfqd, current->io_context, q); + struct bfq_io_cq *bic = bfq_bic_lookup(q); bool ret; spin_lock_irq(&bfqd->lock); @@ -5878,6 +5998,22 @@ static void bfq_rq_enqueued(struct bfq_data *bfqd, struct bfq_queue *bfqq, } } +static void bfqq_request_allocated(struct bfq_queue *bfqq) +{ + struct bfq_entity *entity = &bfqq->entity; + + for_each_entity(entity) + entity->allocated++; +} + +static void bfqq_request_freed(struct bfq_queue *bfqq) +{ + struct bfq_entity *entity = &bfqq->entity; + + for_each_entity(entity) + entity->allocated--; +} + /* returns true if it causes the idle timer to be disabled */ static bool __bfq_insert_request(struct bfq_data *bfqd, struct request *rq) { @@ -5891,8 +6027,8 @@ static bool __bfq_insert_request(struct bfq_data *bfqd, struct request *rq) * Release the request's reference to the old bfqq * and make sure one is taken to the shared queue. */ - new_bfqq->allocated++; - bfqq->allocated--; + bfqq_request_allocated(new_bfqq); + bfqq_request_freed(bfqq); new_bfqq->ref++; /* * If the bic associated with the process @@ -5991,48 +6127,7 @@ static void bfq_insert_request(struct blk_mq_hw_ctx *hctx, struct request *rq, spin_lock_irq(&bfqd->lock); bfqq = bfq_init_rq(rq); - - /* - * Reqs with at_head or passthrough flags set are to be put - * directly into dispatch list. Additional case for putting rq - * directly into the dispatch queue: the only active - * bfq_queues are bfqq and either its waker bfq_queue or one - * of its woken bfq_queues. The rationale behind this - * additional condition is as follows: - * - consider a bfq_queue, say Q1, detected as a waker of - * another bfq_queue, say Q2 - * - by definition of a waker, Q1 blocks the I/O of Q2, i.e., - * some I/O of Q1 needs to be completed for new I/O of Q2 - * to arrive. A notable example of waker is journald - * - so, Q1 and Q2 are in any respect the queues of two - * cooperating processes (or of two cooperating sets of - * processes): the goal of Q1's I/O is doing what needs to - * be done so that new Q2's I/O can finally be - * issued. Therefore, if the service of Q1's I/O is delayed, - * then Q2's I/O is delayed too. Conversely, if Q2's I/O is - * delayed, the goal of Q1's I/O is hindered. - * - as a consequence, if some I/O of Q1/Q2 arrives while - * Q2/Q1 is the only queue in service, there is absolutely - * no point in delaying the service of such an I/O. The - * only possible result is a throughput loss - * - so, when the above condition holds, the best option is to - * have the new I/O dispatched as soon as possible - * - the most effective and efficient way to attain the above - * goal is to put the new I/O directly in the dispatch - * list - * - as an additional restriction, Q1 and Q2 must be the only - * busy queues for this commit to put the I/O of Q2/Q1 in - * the dispatch list. This is necessary, because, if also - * other queues are waiting for service, then putting new - * I/O directly in the dispatch list may evidently cause a - * violation of service guarantees for the other queues - */ - if (!bfqq || - (bfqq != bfqd->in_service_queue && - bfqd->in_service_queue != NULL && - bfq_tot_busy_queues(bfqd) == 1 + bfq_bfqq_busy(bfqq) && - (bfqq->waker_bfqq == bfqd->in_service_queue || - bfqd->in_service_queue->waker_bfqq == bfqq)) || at_head) { + if (!bfqq || at_head) { if (at_head) list_add(&rq->queuelist, &bfqd->dispatch); else @@ -6059,7 +6154,6 @@ static void bfq_insert_request(struct blk_mq_hw_ctx *hctx, struct request *rq, * merge). */ cmd_flags = rq->cmd_flags; - spin_unlock_irq(&bfqd->lock); bfq_update_insert_stats(q, bfqq, idle_timer_disabled, @@ -6251,8 +6345,7 @@ static void bfq_completed_request(struct bfq_queue *bfqq, struct bfq_data *bfqd) static void bfq_finish_requeue_request_body(struct bfq_queue *bfqq) { - bfqq->allocated--; - + bfqq_request_freed(bfqq); bfq_put_queue(bfqq); } @@ -6476,6 +6569,16 @@ static void bfq_finish_requeue_request(struct request *rq) rq->elv.priv[1] = NULL; } +static void bfq_finish_request(struct request *rq) +{ + bfq_finish_requeue_request(rq); + + if (rq->elv.icq) { + put_io_context(rq->elv.icq->ioc); + rq->elv.icq = NULL; + } +} + /* * Removes the association between the current task and bfqq, assuming * that bic points to the bfq iocontext of the task. @@ -6573,6 +6676,8 @@ static struct bfq_queue *bfq_get_bfqq_handle_split(struct bfq_data *bfqd, */ static void bfq_prepare_request(struct request *rq) { + rq->elv.icq = ioc_find_get_icq(rq->q); + /* * Regardless of whether we have an icq attached, we have to * clear the scheduler pointers, as they might point to @@ -6672,7 +6777,7 @@ static struct bfq_queue *bfq_init_rq(struct request *rq) } } - bfqq->allocated++; + bfqq_request_allocated(bfqq); bfqq->ref++; bfq_log_bfqq(bfqd, bfqq, "get_request %p: bfqq %p, %d", rq, bfqq, bfqq->ref); @@ -6835,11 +6940,11 @@ void bfq_put_async_queues(struct bfq_data *bfqd, struct bfq_group *bfqg) * See the comments on bfq_limit_depth for the purpose of * the depths set in the function. Return minimum shallow depth we'll use. */ -static unsigned int bfq_update_depths(struct bfq_data *bfqd, - struct sbitmap_queue *bt) +static void bfq_update_depths(struct bfq_data *bfqd, struct sbitmap_queue *bt) { - unsigned int i, j, min_shallow = UINT_MAX; + unsigned int depth = 1U << bt->sb.shift; + bfqd->full_depth_shift = bt->sb.shift; /* * In-word depths if no bfq_queue is being weight-raised: * leaving 25% of tags only for sync reads. @@ -6851,13 +6956,13 @@ static unsigned int bfq_update_depths(struct bfq_data *bfqd, * limit 'something'. */ /* no more than 50% of tags for async I/O */ - bfqd->word_depths[0][0] = max((1U << bt->sb.shift) >> 1, 1U); + bfqd->word_depths[0][0] = max(depth >> 1, 1U); /* * no more than 75% of tags for sync writes (25% extra tags * w.r.t. async I/O, to prevent async I/O from starving sync * writes) */ - bfqd->word_depths[0][1] = max(((1U << bt->sb.shift) * 3) >> 2, 1U); + bfqd->word_depths[0][1] = max((depth * 3) >> 2, 1U); /* * In-word depths in case some bfq_queue is being weight- @@ -6867,25 +6972,18 @@ static unsigned int bfq_update_depths(struct bfq_data *bfqd, * shortage. */ /* no more than ~18% of tags for async I/O */ - bfqd->word_depths[1][0] = max(((1U << bt->sb.shift) * 3) >> 4, 1U); + bfqd->word_depths[1][0] = max((depth * 3) >> 4, 1U); /* no more than ~37% of tags for sync writes (~20% extra tags) */ - bfqd->word_depths[1][1] = max(((1U << bt->sb.shift) * 6) >> 4, 1U); - - for (i = 0; i < 2; i++) - for (j = 0; j < 2; j++) - min_shallow = min(min_shallow, bfqd->word_depths[i][j]); - - return min_shallow; + bfqd->word_depths[1][1] = max((depth * 6) >> 4, 1U); } 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; - unsigned int min_shallow; - min_shallow = bfq_update_depths(bfqd, &tags->bitmap_tags); - sbitmap_queue_min_shallow_depth(&tags->bitmap_tags, min_shallow); + bfq_update_depths(bfqd, &tags->bitmap_tags); + sbitmap_queue_min_shallow_depth(&tags->bitmap_tags, 1); } static int bfq_init_hctx(struct blk_mq_hw_ctx *hctx, unsigned int index) @@ -7300,7 +7398,7 @@ static struct elevator_type iosched_bfq_mq = { .limit_depth = bfq_limit_depth, .prepare_request = bfq_prepare_request, .requeue_request = bfq_finish_requeue_request, - .finish_request = bfq_finish_requeue_request, + .finish_request = bfq_finish_request, .exit_icq = bfq_exit_icq, .insert_requests = bfq_insert_requests, .dispatch_request = bfq_dispatch_request, diff --git a/block/bfq-iosched.h b/block/bfq-iosched.h index a73488eec8a4..07288b9da389 100644 --- a/block/bfq-iosched.h +++ b/block/bfq-iosched.h @@ -25,7 +25,7 @@ #define BFQ_DEFAULT_GRP_IOPRIO 0 #define BFQ_DEFAULT_GRP_CLASS IOPRIO_CLASS_BE -#define MAX_PID_STR_LENGTH 12 +#define MAX_BFQQ_NAME_LENGTH 16 /* * Soft real-time applications are extremely more latency sensitive @@ -170,6 +170,9 @@ struct bfq_entity { /* budget, used also to calculate F_i: F_i = S_i + @budget / @weight */ int budget; + /* Number of requests allocated in the subtree of this entity */ + int allocated; + /* device weight, if non-zero, it overrides the default weight of * bfq_group_data */ int dev_weight; @@ -266,8 +269,6 @@ struct bfq_queue { struct request *next_rq; /* number of sync and async requests queued */ int queued[2]; - /* number of requests currently allocated */ - int allocated; /* number of pending metadata requests */ int meta_pending; /* fifo list of requests in sort_list */ @@ -387,6 +388,8 @@ struct bfq_queue { struct bfq_queue *tentative_waker_bfqq; /* number of times the same tentative waker has been detected */ unsigned int num_waker_detections; + /* time when we started considering this waker */ + u64 waker_detection_started; /* node for woken_list, see below */ struct hlist_node woken_list_node; @@ -768,6 +771,7 @@ struct bfq_data { * function) */ unsigned int word_depths[2][2]; + unsigned int full_depth_shift; }; enum bfqq_state_flags { @@ -1079,26 +1083,27 @@ void bfq_add_bfqq_busy(struct bfq_data *bfqd, struct bfq_queue *bfqq); /* --------------- end of interface of B-WF2Q+ ---------------- */ /* Logging facilities. */ -static inline void bfq_pid_to_str(int pid, char *str, int len) +static inline void bfq_bfqq_name(struct bfq_queue *bfqq, char *str, int len) { - if (pid != -1) - snprintf(str, len, "%d", pid); + char type = bfq_bfqq_sync(bfqq) ? 'S' : 'A'; + + if (bfqq->pid != -1) + snprintf(str, len, "bfq%d%c", bfqq->pid, type); else - snprintf(str, len, "SHARED-"); + snprintf(str, len, "bfqSHARED-%c", type); } #ifdef CONFIG_BFQ_GROUP_IOSCHED struct bfq_group *bfqq_group(struct bfq_queue *bfqq); #define bfq_log_bfqq(bfqd, bfqq, fmt, args...) do { \ - char pid_str[MAX_PID_STR_LENGTH]; \ + char pid_str[MAX_BFQQ_NAME_LENGTH]; \ if (likely(!blk_trace_note_message_enabled((bfqd)->queue))) \ break; \ - bfq_pid_to_str((bfqq)->pid, pid_str, MAX_PID_STR_LENGTH); \ + bfq_bfqq_name((bfqq), pid_str, MAX_BFQQ_NAME_LENGTH); \ blk_add_cgroup_trace_msg((bfqd)->queue, \ bfqg_to_blkg(bfqq_group(bfqq))->blkcg, \ - "bfq%s%c " fmt, pid_str, \ - bfq_bfqq_sync((bfqq)) ? 'S' : 'A', ##args); \ + "%s " fmt, pid_str, ##args); \ } while (0) #define bfq_log_bfqg(bfqd, bfqg, fmt, args...) do { \ @@ -1109,13 +1114,11 @@ struct bfq_group *bfqq_group(struct bfq_queue *bfqq); #else /* CONFIG_BFQ_GROUP_IOSCHED */ #define bfq_log_bfqq(bfqd, bfqq, fmt, args...) do { \ - char pid_str[MAX_PID_STR_LENGTH]; \ + char pid_str[MAX_BFQQ_NAME_LENGTH]; \ if (likely(!blk_trace_note_message_enabled((bfqd)->queue))) \ break; \ - bfq_pid_to_str((bfqq)->pid, pid_str, MAX_PID_STR_LENGTH); \ - blk_add_trace_msg((bfqd)->queue, "bfq%s%c " fmt, pid_str, \ - bfq_bfqq_sync((bfqq)) ? 'S' : 'A', \ - ##args); \ + bfq_bfqq_name((bfqq), pid_str, MAX_BFQQ_NAME_LENGTH); \ + blk_add_trace_msg((bfqd)->queue, "%s " fmt, pid_str, ##args); \ } while (0) #define bfq_log_bfqg(bfqd, bfqg, fmt, args...) do {} while (0) diff --git a/block/bio.c b/block/bio.c index 15ab0d6d1c06..0d400ba2dbd1 100644 --- a/block/bio.c +++ b/block/bio.c @@ -26,7 +26,7 @@ #include "blk-rq-qos.h" struct bio_alloc_cache { - struct bio_list free_list; + struct bio *free_list; unsigned int nr; }; @@ -630,7 +630,8 @@ static void bio_alloc_cache_prune(struct bio_alloc_cache *cache, unsigned int i = 0; struct bio *bio; - while ((bio = bio_list_pop(&cache->free_list)) != NULL) { + while ((bio = cache->free_list) != NULL) { + cache->free_list = bio->bi_next; cache->nr--; bio_free(bio); if (++i == nr) @@ -689,7 +690,8 @@ void bio_put(struct bio *bio) bio_uninit(bio); cache = per_cpu_ptr(bio->bi_pool->cache, get_cpu()); - bio_list_add_head(&cache->free_list, bio); + bio->bi_next = cache->free_list; + cache->free_list = bio; if (++cache->nr > ALLOC_CACHE_MAX + ALLOC_CACHE_SLACK) bio_alloc_cache_prune(cache, ALLOC_CACHE_SLACK); put_cpu(); @@ -1033,6 +1035,28 @@ int bio_add_page(struct bio *bio, struct page *page, } EXPORT_SYMBOL(bio_add_page); +/** + * bio_add_folio - Attempt to add part of a folio to a bio. + * @bio: BIO to add to. + * @folio: Folio to add. + * @len: How many bytes from the folio to add. + * @off: First byte in this folio to add. + * + * Filesystems that use folios can call this function instead of calling + * bio_add_page() for each page in the folio. If @off is bigger than + * PAGE_SIZE, this function can create a bio_vec that starts in a page + * after the bv_page. BIOs do not support folios that are 4GiB or larger. + * + * Return: Whether the addition was successful. + */ +bool bio_add_folio(struct bio *bio, struct folio *folio, size_t len, + size_t off) +{ + if (len > UINT_MAX || off > UINT_MAX) + return 0; + return bio_add_page(bio, &folio->page, len, off) > 0; +} + void __bio_release_pages(struct bio *bio, bool mark_dirty) { struct bvec_iter_all iter_all; @@ -1704,8 +1728,9 @@ struct bio *bio_alloc_kiocb(struct kiocb *kiocb, unsigned short nr_vecs, return bio_alloc_bioset(GFP_KERNEL, nr_vecs, bs); cache = per_cpu_ptr(bs->cache, get_cpu()); - bio = bio_list_pop(&cache->free_list); - if (bio) { + if (cache->free_list) { + bio = cache->free_list; + cache->free_list = bio->bi_next; cache->nr--; put_cpu(); bio_init(bio, nr_vecs ? bio->bi_inline_vecs : NULL, nr_vecs); diff --git a/block/blk-cgroup.c b/block/blk-cgroup.c index 88b1fce90520..650f7e27989f 100644 --- a/block/blk-cgroup.c +++ b/block/blk-cgroup.c @@ -30,6 +30,7 @@ #include <linux/blk-cgroup.h> #include <linux/tracehook.h> #include <linux/psi.h> +#include <linux/part_stat.h> #include "blk.h" #include "blk-ioprio.h" #include "blk-throttle.h" @@ -640,7 +641,7 @@ int blkg_conf_prep(struct blkcg *blkcg, const struct blkcg_policy *pol, */ ret = blk_queue_enter(q, 0); if (ret) - return ret; + goto fail; rcu_read_lock(); spin_lock_irq(&q->queue_lock); @@ -676,13 +677,13 @@ int blkg_conf_prep(struct blkcg *blkcg, const struct blkcg_policy *pol, new_blkg = blkg_alloc(pos, q, GFP_KERNEL); if (unlikely(!new_blkg)) { ret = -ENOMEM; - goto fail; + goto fail_exit_queue; } if (radix_tree_preload(GFP_KERNEL)) { blkg_free(new_blkg); ret = -ENOMEM; - goto fail; + goto fail_exit_queue; } rcu_read_lock(); @@ -722,9 +723,10 @@ fail_preloaded: fail_unlock: spin_unlock_irq(&q->queue_lock); rcu_read_unlock(); +fail_exit_queue: + blk_queue_exit(q); fail: blkdev_put_no_open(bdev); - blk_queue_exit(q); /* * If queue was bypassing, we should retry. Do so after a * short msleep(). It isn't strictly necessary but queue diff --git a/block/blk-core.c b/block/blk-core.c index 9ee32f85d74e..97f8bc8d3a79 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -16,7 +16,6 @@ #include <linux/module.h> #include <linux/bio.h> #include <linux/blkdev.h> -#include <linux/blk-mq.h> #include <linux/blk-pm.h> #include <linux/blk-integrity.h> #include <linux/highmem.h> @@ -40,6 +39,7 @@ #include <linux/debugfs.h> #include <linux/bpf.h> #include <linux/psi.h> +#include <linux/part_stat.h> #include <linux/sched/sysctl.h> #include <linux/blk-crypto.h> @@ -47,7 +47,6 @@ #include <trace/events/block.h> #include "blk.h" -#include "blk-mq.h" #include "blk-mq-sched.h" #include "blk-pm.h" #include "blk-throttle.h" @@ -67,6 +66,7 @@ DEFINE_IDA(blk_queue_ida); * For queue allocation */ struct kmem_cache *blk_requestq_cachep; +struct kmem_cache *blk_requestq_srcu_cachep; /* * Controlling structure to kblockd @@ -109,23 +109,6 @@ bool blk_queue_flag_test_and_set(unsigned int flag, struct request_queue *q) } EXPORT_SYMBOL_GPL(blk_queue_flag_test_and_set); -void blk_rq_init(struct request_queue *q, struct request *rq) -{ - memset(rq, 0, sizeof(*rq)); - - INIT_LIST_HEAD(&rq->queuelist); - rq->q = q; - rq->__sector = (sector_t) -1; - INIT_HLIST_NODE(&rq->hash); - RB_CLEAR_NODE(&rq->rb_node); - rq->tag = BLK_MQ_NO_TAG; - rq->internal_tag = BLK_MQ_NO_TAG; - rq->start_time_ns = ktime_get_ns(); - rq->part = NULL; - blk_crypto_rq_set_defaults(rq); -} -EXPORT_SYMBOL(blk_rq_init); - #define REQ_OP_NAME(name) [REQ_OP_##name] = #name static const char *const blk_op_name[] = { REQ_OP_NAME(READ), @@ -216,38 +199,15 @@ int blk_status_to_errno(blk_status_t status) } EXPORT_SYMBOL_GPL(blk_status_to_errno); -void blk_print_req_error(struct request *req, blk_status_t status) +const char *blk_status_to_str(blk_status_t status) { int idx = (__force int)status; if (WARN_ON_ONCE(idx >= ARRAY_SIZE(blk_errors))) - return; - - printk_ratelimited(KERN_ERR - "%s error, dev %s, sector %llu op 0x%x:(%s) flags 0x%x " - "phys_seg %u prio class %u\n", - blk_errors[idx].name, - req->rq_disk ? req->rq_disk->disk_name : "?", - blk_rq_pos(req), req_op(req), blk_op_str(req_op(req)), - req->cmd_flags & ~REQ_OP_MASK, - req->nr_phys_segments, - IOPRIO_PRIO_CLASS(req->ioprio)); + return "<null>"; + return blk_errors[idx].name; } -void blk_dump_rq_flags(struct request *rq, char *msg) -{ - printk(KERN_INFO "%s: dev %s: flags=%llx\n", msg, - rq->rq_disk ? rq->rq_disk->disk_name : "?", - (unsigned long long) rq->cmd_flags); - - printk(KERN_INFO " sector %llu, nr/cnr %u/%u\n", - (unsigned long long)blk_rq_pos(rq), - blk_rq_sectors(rq), blk_rq_cur_sectors(rq)); - printk(KERN_INFO " bio %p, biotail %p, len %u\n", - rq->bio, rq->biotail, blk_rq_bytes(rq)); -} -EXPORT_SYMBOL(blk_dump_rq_flags); - /** * blk_sync_queue - cancel any pending callbacks on a queue * @q: the queue @@ -363,8 +323,10 @@ void blk_cleanup_queue(struct request_queue *q) blk_queue_flag_set(QUEUE_FLAG_DEAD, q); blk_sync_queue(q); - if (queue_is_mq(q)) + if (queue_is_mq(q)) { + blk_mq_cancel_work_sync(q); blk_mq_exit_queue(q); + } /* * In theory, request pool of sched_tags belongs to request queue. @@ -476,21 +438,27 @@ static void blk_timeout_work(struct work_struct *work) { } -struct request_queue *blk_alloc_queue(int node_id) +struct request_queue *blk_alloc_queue(int node_id, bool alloc_srcu) { struct request_queue *q; int ret; - q = kmem_cache_alloc_node(blk_requestq_cachep, - GFP_KERNEL | __GFP_ZERO, node_id); + q = kmem_cache_alloc_node(blk_get_queue_kmem_cache(alloc_srcu), + GFP_KERNEL | __GFP_ZERO, node_id); if (!q) return NULL; + if (alloc_srcu) { + blk_queue_flag_set(QUEUE_FLAG_HAS_SRCU, q); + if (init_srcu_struct(q->srcu) != 0) + goto fail_q; + } + q->last_merge = NULL; q->id = ida_simple_get(&blk_queue_ida, 0, 0, GFP_KERNEL); if (q->id < 0) - goto fail_q; + goto fail_srcu; ret = bioset_init(&q->bio_split, BIO_POOL_SIZE, 0, 0); if (ret) @@ -547,8 +515,11 @@ fail_split: bioset_exit(&q->bio_split); fail_id: ida_simple_remove(&blk_queue_ida, q->id); +fail_srcu: + if (alloc_srcu) + cleanup_srcu_struct(q->srcu); fail_q: - kmem_cache_free(blk_requestq_cachep, q); + kmem_cache_free(blk_get_queue_kmem_cache(alloc_srcu), q); return NULL; } @@ -592,7 +563,7 @@ static int __init setup_fail_make_request(char *str) } __setup("fail_make_request=", setup_fail_make_request); -static bool should_fail_request(struct block_device *part, unsigned int bytes) +bool should_fail_request(struct block_device *part, unsigned int bytes) { return part->bd_make_it_fail && should_fail(&fail_make_request, bytes); } @@ -606,15 +577,6 @@ static int __init fail_make_request_debugfs(void) } late_initcall(fail_make_request_debugfs); - -#else /* CONFIG_FAIL_MAKE_REQUEST */ - -static inline bool should_fail_request(struct block_device *part, - unsigned int bytes) -{ - return false; -} - #endif /* CONFIG_FAIL_MAKE_REQUEST */ static inline bool bio_check_ro(struct bio *bio) @@ -800,15 +762,6 @@ noinline_for_stack bool submit_bio_checks(struct bio *bio) break; } - /* - * Various block parts want %current->io_context, so allocate it up - * front rather than dealing with lots of pain to allocate it only - * where needed. This may fail and the block layer knows how to live - * with it. - */ - if (unlikely(!current->io_context)) - create_task_io_context(current, GFP_ATOMIC, q->node); - if (blk_throtl_bio(bio)) return false; @@ -834,17 +787,21 @@ end_io: static void __submit_bio_fops(struct gendisk *disk, struct bio *bio) { - if (unlikely(bio_queue_enter(bio) != 0)) - return; - if (submit_bio_checks(bio) && blk_crypto_bio_prep(&bio)) - disk->fops->submit_bio(bio); - blk_queue_exit(disk->queue); + if (blk_crypto_bio_prep(&bio)) { + if (likely(bio_queue_enter(bio) == 0)) { + disk->fops->submit_bio(bio); + blk_queue_exit(disk->queue); + } + } } static void __submit_bio(struct bio *bio) { struct gendisk *disk = bio->bi_bdev->bd_disk; + if (unlikely(!submit_bio_checks(bio))) + return; + if (!disk->fops->submit_bio) blk_mq_submit_bio(bio); else @@ -1015,6 +972,7 @@ EXPORT_SYMBOL(submit_bio); /** * bio_poll - poll for BIO completions * @bio: bio to poll for + * @iob: batches of IO * @flags: BLK_POLL_* flags that control the behavior * * Poll for completions on queue associated with the bio. Returns number of @@ -1087,135 +1045,7 @@ int iocb_bio_iopoll(struct kiocb *kiocb, struct io_comp_batch *iob, } EXPORT_SYMBOL_GPL(iocb_bio_iopoll); -/** - * blk_cloned_rq_check_limits - Helper function to check a cloned request - * for the new queue limits - * @q: the queue - * @rq: the request being checked - * - * Description: - * @rq may have been made based on weaker limitations of upper-level queues - * in request stacking drivers, and it may violate the limitation of @q. - * Since the block layer and the underlying device driver trust @rq - * after it is inserted to @q, it should be checked against @q before - * the insertion using this generic function. - * - * Request stacking drivers like request-based dm may change the queue - * limits when retrying requests on other queues. Those requests need - * to be checked against the new queue limits again during dispatch. - */ -static blk_status_t blk_cloned_rq_check_limits(struct request_queue *q, - struct request *rq) -{ - unsigned int max_sectors = blk_queue_get_max_sectors(q, req_op(rq)); - - if (blk_rq_sectors(rq) > max_sectors) { - /* - * SCSI device does not have a good way to return if - * Write Same/Zero is actually supported. If a device rejects - * a non-read/write command (discard, write same,etc.) the - * low-level device driver will set the relevant queue limit to - * 0 to prevent blk-lib from issuing more of the offending - * operations. Commands queued prior to the queue limit being - * reset need to be completed with BLK_STS_NOTSUPP to avoid I/O - * errors being propagated to upper layers. - */ - if (max_sectors == 0) - return BLK_STS_NOTSUPP; - - printk(KERN_ERR "%s: over max size limit. (%u > %u)\n", - __func__, blk_rq_sectors(rq), max_sectors); - return BLK_STS_IOERR; - } - - /* - * The queue settings related to segment counting may differ from the - * original queue. - */ - rq->nr_phys_segments = blk_recalc_rq_segments(rq); - if (rq->nr_phys_segments > queue_max_segments(q)) { - printk(KERN_ERR "%s: over max segments limit. (%hu > %hu)\n", - __func__, rq->nr_phys_segments, queue_max_segments(q)); - return BLK_STS_IOERR; - } - - return BLK_STS_OK; -} - -/** - * blk_insert_cloned_request - Helper for stacking drivers to submit a request - * @q: the queue to submit the request - * @rq: the request being queued - */ -blk_status_t blk_insert_cloned_request(struct request_queue *q, struct request *rq) -{ - blk_status_t ret; - - ret = blk_cloned_rq_check_limits(q, rq); - if (ret != BLK_STS_OK) - return ret; - - if (rq->rq_disk && - should_fail_request(rq->rq_disk->part0, blk_rq_bytes(rq))) - return BLK_STS_IOERR; - - if (blk_crypto_insert_cloned_request(rq)) - return BLK_STS_IOERR; - - blk_account_io_start(rq); - - /* - * Since we have a scheduler attached on the top device, - * bypass a potential scheduler on the bottom device for - * insert. - */ - return blk_mq_request_issue_directly(rq, true); -} -EXPORT_SYMBOL_GPL(blk_insert_cloned_request); - -/** - * blk_rq_err_bytes - determine number of bytes till the next failure boundary - * @rq: request to examine - * - * Description: - * A request could be merge of IOs which require different failure - * handling. This function determines the number of bytes which - * can be failed from the beginning of the request without - * crossing into area which need to be retried further. - * - * Return: - * The number of bytes to fail. - */ -unsigned int blk_rq_err_bytes(const struct request *rq) -{ - unsigned int ff = rq->cmd_flags & REQ_FAILFAST_MASK; - unsigned int bytes = 0; - struct bio *bio; - - if (!(rq->rq_flags & RQF_MIXED_MERGE)) - return blk_rq_bytes(rq); - - /* - * Currently the only 'mixing' which can happen is between - * different fastfail types. We can safely fail portions - * which have all the failfast bits that the first one has - - * the ones which are at least as eager to fail as the first - * one. - */ - for (bio = rq->bio; bio; bio = bio->bi_next) { - if ((bio->bi_opf & ff) != ff) - break; - bytes += bio->bi_iter.bi_size; - } - - /* this could lead to infinite loop */ - BUG_ON(blk_rq_bytes(rq) && !bytes); - return bytes; -} -EXPORT_SYMBOL_GPL(blk_rq_err_bytes); - -static void update_io_ticks(struct block_device *part, unsigned long now, - bool end) +void update_io_ticks(struct block_device *part, unsigned long now, bool end) { unsigned long stamp; again: @@ -1230,30 +1060,6 @@ again: } } -void __blk_account_io_done(struct request *req, u64 now) -{ - const int sgrp = op_stat_group(req_op(req)); - - part_stat_lock(); - update_io_ticks(req->part, jiffies, true); - part_stat_inc(req->part, ios[sgrp]); - part_stat_add(req->part, nsecs[sgrp], now - req->start_time_ns); - part_stat_unlock(); -} - -void __blk_account_io_start(struct request *rq) -{ - /* passthrough requests can hold bios that do not have ->bi_bdev set */ - if (rq->bio && rq->bio->bi_bdev) - rq->part = rq->bio->bi_bdev; - else - rq->part = rq->rq_disk->part0; - - part_stat_lock(); - update_io_ticks(rq->part, jiffies, false); - part_stat_unlock(); -} - static unsigned long __part_start_io_acct(struct block_device *part, unsigned int sectors, unsigned int op) { @@ -1317,46 +1123,6 @@ void disk_end_io_acct(struct gendisk *disk, unsigned int op, } EXPORT_SYMBOL(disk_end_io_acct); -/* - * Steal bios from a request and add them to a bio list. - * The request must not have been partially completed before. - */ -void blk_steal_bios(struct bio_list *list, struct request *rq) -{ - if (rq->bio) { - if (list->tail) - list->tail->bi_next = rq->bio; - else - list->head = rq->bio; - list->tail = rq->biotail; - - rq->bio = NULL; - rq->biotail = NULL; - } - - rq->__data_len = 0; -} -EXPORT_SYMBOL_GPL(blk_steal_bios); - -#if ARCH_IMPLEMENTS_FLUSH_DCACHE_PAGE -/** - * rq_flush_dcache_pages - Helper function to flush all pages in a request - * @rq: the request to be flushed - * - * Description: - * Flush all pages in @rq. - */ -void rq_flush_dcache_pages(struct request *rq) -{ - struct req_iterator iter; - struct bio_vec bvec; - - rq_for_each_segment(bvec, rq, iter) - flush_dcache_page(bvec.bv_page); -} -EXPORT_SYMBOL_GPL(rq_flush_dcache_pages); -#endif - /** * blk_lld_busy - Check if underlying low-level drivers of a device are busy * @q : the queue of the device being checked @@ -1385,93 +1151,6 @@ int blk_lld_busy(struct request_queue *q) } EXPORT_SYMBOL_GPL(blk_lld_busy); -/** - * blk_rq_unprep_clone - Helper function to free all bios in a cloned request - * @rq: the clone request to be cleaned up - * - * Description: - * Free all bios in @rq for a cloned request. - */ -void blk_rq_unprep_clone(struct request *rq) -{ - struct bio *bio; - - while ((bio = rq->bio) != NULL) { - rq->bio = bio->bi_next; - - bio_put(bio); - } -} -EXPORT_SYMBOL_GPL(blk_rq_unprep_clone); - -/** - * blk_rq_prep_clone - Helper function to setup clone request - * @rq: the request to be setup - * @rq_src: original request to be cloned - * @bs: bio_set that bios for clone are allocated from - * @gfp_mask: memory allocation mask for bio - * @bio_ctr: setup function to be called for each clone bio. - * Returns %0 for success, non %0 for failure. - * @data: private data to be passed to @bio_ctr - * - * Description: - * Clones bios in @rq_src to @rq, and copies attributes of @rq_src to @rq. - * Also, pages which the original bios are pointing to are not copied - * and the cloned bios just point same pages. - * So cloned bios must be completed before original bios, which means - * the caller must complete @rq before @rq_src. - */ -int blk_rq_prep_clone(struct request *rq, struct request *rq_src, - struct bio_set *bs, gfp_t gfp_mask, - int (*bio_ctr)(struct bio *, struct bio *, void *), - void *data) -{ - struct bio *bio, *bio_src; - - if (!bs) - bs = &fs_bio_set; - - __rq_for_each_bio(bio_src, rq_src) { - bio = bio_clone_fast(bio_src, gfp_mask, bs); - if (!bio) - goto free_and_out; - - if (bio_ctr && bio_ctr(bio, bio_src, data)) - goto free_and_out; - - if (rq->bio) { - rq->biotail->bi_next = bio; - rq->biotail = bio; - } else { - rq->bio = rq->biotail = bio; - } - bio = NULL; - } - - /* Copy attributes of the original request to the clone request. */ - rq->__sector = blk_rq_pos(rq_src); - rq->__data_len = blk_rq_bytes(rq_src); - if (rq_src->rq_flags & RQF_SPECIAL_PAYLOAD) { - rq->rq_flags |= RQF_SPECIAL_PAYLOAD; - rq->special_vec = rq_src->special_vec; - } - rq->nr_phys_segments = rq_src->nr_phys_segments; - rq->ioprio = rq_src->ioprio; - - if (rq->bio && blk_crypto_rq_bio_prep(rq, rq->bio, gfp_mask) < 0) - goto free_and_out; - - return 0; - -free_and_out: - if (bio) - bio_put(bio); - blk_rq_unprep_clone(rq); - - return -ENOMEM; -} -EXPORT_SYMBOL_GPL(blk_rq_prep_clone); - int kblockd_schedule_work(struct work_struct *work) { return queue_work(kblockd_workqueue, work); @@ -1636,6 +1315,9 @@ int __init blk_dev_init(void) sizeof_field(struct request, cmd_flags)); BUILD_BUG_ON(REQ_OP_BITS + REQ_FLAG_BITS > 8 * sizeof_field(struct bio, bi_opf)); + BUILD_BUG_ON(ALIGN(offsetof(struct request_queue, srcu), + __alignof__(struct request_queue)) != + sizeof(struct request_queue)); /* used for unplugging and affects IO latency/throughput - HIGHPRI */ kblockd_workqueue = alloc_workqueue("kblockd", @@ -1646,6 +1328,10 @@ int __init blk_dev_init(void) blk_requestq_cachep = kmem_cache_create("request_queue", sizeof(struct request_queue), 0, SLAB_PANIC, NULL); + blk_requestq_srcu_cachep = kmem_cache_create("request_queue_srcu", + sizeof(struct request_queue) + + sizeof(struct srcu_struct), 0, SLAB_PANIC, NULL); + blk_debugfs_root = debugfs_create_dir("block", NULL); return 0; diff --git a/block/blk-crypto-profile.c b/block/blk-crypto-profile.c index 605ba0626a5c..96c511967386 100644 --- a/block/blk-crypto-profile.c +++ b/block/blk-crypto-profile.c @@ -463,11 +463,6 @@ bool blk_crypto_register(struct blk_crypto_profile *profile, } EXPORT_SYMBOL_GPL(blk_crypto_register); -void blk_crypto_unregister(struct request_queue *q) -{ - q->crypto_profile = NULL; -} - /** * blk_crypto_intersect_capabilities() - restrict supported crypto capabilities * by child device diff --git a/block/blk-exec.c b/block/blk-exec.c deleted file mode 100644 index 1b8b47f6e79b..000000000000 --- a/block/blk-exec.c +++ /dev/null @@ -1,116 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -/* - * Functions related to setting various queue properties from drivers - */ -#include <linux/kernel.h> -#include <linux/module.h> -#include <linux/bio.h> -#include <linux/blkdev.h> -#include <linux/blk-mq.h> -#include <linux/sched/sysctl.h> - -#include "blk.h" -#include "blk-mq-sched.h" - -/** - * blk_end_sync_rq - executes a completion event on a request - * @rq: request to complete - * @error: end I/O status of the request - */ -static void blk_end_sync_rq(struct request *rq, blk_status_t error) -{ - struct completion *waiting = rq->end_io_data; - - rq->end_io_data = (void *)(uintptr_t)error; - - /* - * complete last, if this is a stack request the process (and thus - * the rq pointer) could be invalid right after this complete() - */ - complete(waiting); -} - -/** - * blk_execute_rq_nowait - insert a request to I/O scheduler for execution - * @bd_disk: matching gendisk - * @rq: request to insert - * @at_head: insert request at head or tail of queue - * @done: I/O completion handler - * - * Description: - * Insert a fully prepared request at the back of the I/O scheduler queue - * for execution. Don't wait for completion. - * - * Note: - * This function will invoke @done directly if the queue is dead. - */ -void blk_execute_rq_nowait(struct gendisk *bd_disk, struct request *rq, - int at_head, rq_end_io_fn *done) -{ - WARN_ON(irqs_disabled()); - WARN_ON(!blk_rq_is_passthrough(rq)); - - rq->rq_disk = bd_disk; - rq->end_io = done; - - blk_account_io_start(rq); - - /* - * don't check dying flag for MQ because the request won't - * be reused after dying flag is set - */ - blk_mq_sched_insert_request(rq, at_head, true, false); -} -EXPORT_SYMBOL_GPL(blk_execute_rq_nowait); - -static bool blk_rq_is_poll(struct request *rq) -{ - if (!rq->mq_hctx) - return false; - if (rq->mq_hctx->type != HCTX_TYPE_POLL) - return false; - if (WARN_ON_ONCE(!rq->bio)) - return false; - return true; -} - -static void blk_rq_poll_completion(struct request *rq, struct completion *wait) -{ - do { - bio_poll(rq->bio, NULL, 0); - cond_resched(); - } while (!completion_done(wait)); -} - -/** - * blk_execute_rq - insert a request into queue for execution - * @bd_disk: matching gendisk - * @rq: request to insert - * @at_head: insert request at head or tail of queue - * - * Description: - * Insert a fully prepared request at the back of the I/O scheduler queue - * for execution and wait for completion. - * Return: The blk_status_t result provided to blk_mq_end_request(). - */ -blk_status_t blk_execute_rq(struct gendisk *bd_disk, struct request *rq, int at_head) -{ - DECLARE_COMPLETION_ONSTACK(wait); - unsigned long hang_check; - - rq->end_io_data = &wait; - blk_execute_rq_nowait(bd_disk, rq, at_head, blk_end_sync_rq); - - /* Prevent hang_check timer from firing at us during very long I/O */ - hang_check = sysctl_hung_task_timeout_secs; - - if (blk_rq_is_poll(rq)) - blk_rq_poll_completion(rq, &wait); - else if (hang_check) - while (!wait_for_completion_io_timeout(&wait, hang_check * (HZ/2))); - else - wait_for_completion_io(&wait); - - return (blk_status_t)(uintptr_t)rq->end_io_data; -} -EXPORT_SYMBOL(blk_execute_rq); diff --git a/block/blk-flush.c b/block/blk-flush.c index 8e364bda5166..e4df894189ce 100644 --- a/block/blk-flush.c +++ b/block/blk-flush.c @@ -69,6 +69,7 @@ #include <linux/blkdev.h> #include <linux/gfp.h> #include <linux/blk-mq.h> +#include <linux/part_stat.h> #include "blk.h" #include "blk-mq.h" @@ -95,6 +96,12 @@ enum { static void blk_kick_flush(struct request_queue *q, struct blk_flush_queue *fq, unsigned int flags); +static inline struct blk_flush_queue * +blk_get_flush_queue(struct request_queue *q, struct blk_mq_ctx *ctx) +{ + return blk_mq_map_queue(q, REQ_OP_FLUSH, ctx)->fq; +} + static unsigned int blk_flush_policy(unsigned long fflags, struct request *rq) { unsigned int policy = 0; @@ -138,7 +145,7 @@ static void blk_flush_queue_rq(struct request *rq, bool add_front) static void blk_account_io_flush(struct request *rq) { - struct block_device *part = rq->rq_disk->part0; + struct block_device *part = rq->q->disk->part0; part_stat_lock(); part_stat_inc(part, ios[STAT_FLUSH]); @@ -222,7 +229,7 @@ static void flush_end_io(struct request *flush_rq, blk_status_t error) /* release the tag's ownership to the req cloned from */ spin_lock_irqsave(&fq->mq_flush_lock, flags); - if (!refcount_dec_and_test(&flush_rq->ref)) { + if (!req_ref_put_and_test(flush_rq)) { fq->rq_status = error; spin_unlock_irqrestore(&fq->mq_flush_lock, flags); return; @@ -235,8 +242,10 @@ static void flush_end_io(struct request *flush_rq, blk_status_t error) * avoiding use-after-free. */ WRITE_ONCE(flush_rq->state, MQ_RQ_IDLE); - if (fq->rq_status != BLK_STS_OK) + if (fq->rq_status != BLK_STS_OK) { error = fq->rq_status; + fq->rq_status = BLK_STS_OK; + } if (!q->elevator) { flush_rq->tag = BLK_MQ_NO_TAG; @@ -332,7 +341,6 @@ static void blk_kick_flush(struct request_queue *q, struct blk_flush_queue *fq, flush_rq->cmd_flags = REQ_OP_FLUSH | REQ_PREFLUSH; flush_rq->cmd_flags |= (flags & REQ_DRV) | (flags & REQ_FAILFAST_MASK); flush_rq->rq_flags |= RQF_FLUSH_SEQ; - flush_rq->rq_disk = first_rq->rq_disk; flush_rq->end_io = flush_end_io; /* * Order WRITE ->end_io and WRITE rq->ref, and its pair is the one @@ -341,7 +349,7 @@ static void blk_kick_flush(struct request_queue *q, struct blk_flush_queue *fq, * and READ flush_rq->end_io */ smp_wmb(); - refcount_set(&flush_rq->ref, 1); + req_ref_set(flush_rq, 1); blk_flush_queue_rq(flush_rq, false); } @@ -379,7 +387,7 @@ static void mq_flush_data_end_io(struct request *rq, blk_status_t error) * @rq is being submitted. Analyze what needs to be done and put it on the * right queue. */ -bool blk_insert_flush(struct request *rq) +void blk_insert_flush(struct request *rq) { struct request_queue *q = rq->q; unsigned long fflags = q->queue_flags; /* may change, cache */ @@ -409,7 +417,7 @@ bool blk_insert_flush(struct request *rq) */ if (!policy) { blk_mq_end_request(rq, 0); - return true; + return; } BUG_ON(rq->bio != rq->biotail); /*assumes zero or single bio rq */ @@ -420,8 +428,10 @@ bool blk_insert_flush(struct request *rq) * for normal execution. */ if ((policy & REQ_FSEQ_DATA) && - !(policy & (REQ_FSEQ_PREFLUSH | REQ_FSEQ_POSTFLUSH))) - return false; + !(policy & (REQ_FSEQ_PREFLUSH | REQ_FSEQ_POSTFLUSH))) { + blk_mq_request_bypass_insert(rq, false, true); + return; + } /* * @rq should go through flush machinery. Mark it part of flush @@ -437,8 +447,6 @@ bool blk_insert_flush(struct request *rq) spin_lock_irq(&fq->mq_flush_lock); blk_flush_complete_seq(rq, fq, REQ_FSEQ_ACTIONS & ~policy, 0); spin_unlock_irq(&fq->mq_flush_lock); - - return true; } /** diff --git a/block/blk-integrity.c b/block/blk-integrity.c index d670d54e5f7a..69eed260a823 100644 --- a/block/blk-integrity.c +++ b/block/blk-integrity.c @@ -411,7 +411,7 @@ void blk_integrity_register(struct gendisk *disk, struct blk_integrity *template #ifdef CONFIG_BLK_INLINE_ENCRYPTION if (disk->queue->crypto_profile) { pr_warn("blk-integrity: Integrity and hardware inline encryption are not supported together. Disabling hardware inline encryption.\n"); - blk_crypto_unregister(disk->queue); + disk->queue->crypto_profile = NULL; } #endif } diff --git a/block/blk-ioc.c b/block/blk-ioc.c index 57299f860d41..11f49f78db32 100644 --- a/block/blk-ioc.c +++ b/block/blk-ioc.c @@ -8,22 +8,25 @@ #include <linux/bio.h> #include <linux/blkdev.h> #include <linux/slab.h> +#include <linux/security.h> #include <linux/sched/task.h> #include "blk.h" +#include "blk-mq-sched.h" /* * For io context allocations */ static struct kmem_cache *iocontext_cachep; +#ifdef CONFIG_BLK_ICQ /** * get_io_context - increment reference count to io_context * @ioc: io_context to get * * Increment reference count to @ioc. */ -void get_io_context(struct io_context *ioc) +static void get_io_context(struct io_context *ioc) { BUG_ON(atomic_long_read(&ioc->refcount) <= 0); atomic_long_inc(&ioc->refcount); @@ -53,6 +56,16 @@ static void ioc_exit_icq(struct io_cq *icq) icq->flags |= ICQ_EXITED; } +static void ioc_exit_icqs(struct io_context *ioc) +{ + struct io_cq *icq; + + spin_lock_irq(&ioc->lock); + hlist_for_each_entry(icq, &ioc->icq_list, ioc_node) + ioc_exit_icq(icq); + spin_unlock_irq(&ioc->lock); +} + /* * Release an icq. Called with ioc locked for blk-mq, and with both ioc * and queue locked for legacy. @@ -132,68 +145,74 @@ static void ioc_release_fn(struct work_struct *work) kmem_cache_free(iocontext_cachep, ioc); } -/** - * put_io_context - put a reference of io_context - * @ioc: io_context to put - * - * Decrement reference count of @ioc and release it if the count reaches - * zero. +/* + * Releasing icqs requires reverse order double locking and we may already be + * holding a queue_lock. Do it asynchronously from a workqueue. */ -void put_io_context(struct io_context *ioc) +static bool ioc_delay_free(struct io_context *ioc) { unsigned long flags; - bool free_ioc = false; - if (ioc == NULL) - return; - - BUG_ON(atomic_long_read(&ioc->refcount) <= 0); - - /* - * Releasing ioc requires reverse order double locking and we may - * already be holding a queue_lock. Do it asynchronously from wq. - */ - if (atomic_long_dec_and_test(&ioc->refcount)) { - spin_lock_irqsave(&ioc->lock, flags); - if (!hlist_empty(&ioc->icq_list)) - queue_work(system_power_efficient_wq, - &ioc->release_work); - else - free_ioc = true; + spin_lock_irqsave(&ioc->lock, flags); + if (!hlist_empty(&ioc->icq_list)) { + queue_work(system_power_efficient_wq, &ioc->release_work); spin_unlock_irqrestore(&ioc->lock, flags); + return true; } - - if (free_ioc) - kmem_cache_free(iocontext_cachep, ioc); + spin_unlock_irqrestore(&ioc->lock, flags); + return false; } /** - * put_io_context_active - put active reference on ioc - * @ioc: ioc of interest + * ioc_clear_queue - break any ioc association with the specified queue + * @q: request_queue being cleared * - * Undo get_io_context_active(). If active reference reaches zero after - * put, @ioc can never issue further IOs and ioscheds are notified. + * Walk @q->icq_list and exit all io_cq's. */ -void put_io_context_active(struct io_context *ioc) +void ioc_clear_queue(struct request_queue *q) { - struct io_cq *icq; + LIST_HEAD(icq_list); - if (!atomic_dec_and_test(&ioc->active_ref)) { - put_io_context(ioc); - return; - } + spin_lock_irq(&q->queue_lock); + list_splice_init(&q->icq_list, &icq_list); + spin_unlock_irq(&q->queue_lock); - spin_lock_irq(&ioc->lock); - hlist_for_each_entry(icq, &ioc->icq_list, ioc_node) { - if (icq->flags & ICQ_EXITED) - continue; + rcu_read_lock(); + while (!list_empty(&icq_list)) { + struct io_cq *icq = + list_entry(icq_list.next, struct io_cq, q_node); - ioc_exit_icq(icq); + spin_lock_irq(&icq->ioc->lock); + if (!(icq->flags & ICQ_DESTROYED)) + ioc_destroy_icq(icq); + spin_unlock_irq(&icq->ioc->lock); } - spin_unlock_irq(&ioc->lock); + rcu_read_unlock(); +} +#else /* CONFIG_BLK_ICQ */ +static inline void ioc_exit_icqs(struct io_context *ioc) +{ +} +static inline bool ioc_delay_free(struct io_context *ioc) +{ + return false; +} +#endif /* CONFIG_BLK_ICQ */ - put_io_context(ioc); +/** + * put_io_context - put a reference of io_context + * @ioc: io_context to put + * + * Decrement reference count of @ioc and release it if the count reaches + * zero. + */ +void put_io_context(struct io_context *ioc) +{ + BUG_ON(atomic_long_read(&ioc->refcount) <= 0); + if (atomic_long_dec_and_test(&ioc->refcount) && !ioc_delay_free(ioc)) + kmem_cache_free(iocontext_cachep, ioc); } +EXPORT_SYMBOL_GPL(put_io_context); /* Called by the exiting task */ void exit_io_context(struct task_struct *task) @@ -205,132 +224,109 @@ void exit_io_context(struct task_struct *task) task->io_context = NULL; task_unlock(task); - atomic_dec(&ioc->nr_tasks); - put_io_context_active(ioc); -} - -static void __ioc_clear_queue(struct list_head *icq_list) -{ - unsigned long flags; - - rcu_read_lock(); - while (!list_empty(icq_list)) { - struct io_cq *icq = list_entry(icq_list->next, - struct io_cq, q_node); - struct io_context *ioc = icq->ioc; - - spin_lock_irqsave(&ioc->lock, flags); - if (icq->flags & ICQ_DESTROYED) { - spin_unlock_irqrestore(&ioc->lock, flags); - continue; - } - ioc_destroy_icq(icq); - spin_unlock_irqrestore(&ioc->lock, flags); + if (atomic_dec_and_test(&ioc->active_ref)) { + ioc_exit_icqs(ioc); + put_io_context(ioc); } - rcu_read_unlock(); } -/** - * ioc_clear_queue - break any ioc association with the specified queue - * @q: request_queue being cleared - * - * Walk @q->icq_list and exit all io_cq's. - */ -void ioc_clear_queue(struct request_queue *q) -{ - LIST_HEAD(icq_list); - - spin_lock_irq(&q->queue_lock); - list_splice_init(&q->icq_list, &icq_list); - spin_unlock_irq(&q->queue_lock); - - __ioc_clear_queue(&icq_list); -} - -int create_task_io_context(struct task_struct *task, gfp_t gfp_flags, int node) +static struct io_context *alloc_io_context(gfp_t gfp_flags, int node) { struct io_context *ioc; - int ret; ioc = kmem_cache_alloc_node(iocontext_cachep, gfp_flags | __GFP_ZERO, node); if (unlikely(!ioc)) - return -ENOMEM; + return NULL; - /* initialize */ atomic_long_set(&ioc->refcount, 1); - atomic_set(&ioc->nr_tasks, 1); atomic_set(&ioc->active_ref, 1); +#ifdef CONFIG_BLK_ICQ spin_lock_init(&ioc->lock); INIT_RADIX_TREE(&ioc->icq_tree, GFP_ATOMIC); INIT_HLIST_HEAD(&ioc->icq_list); INIT_WORK(&ioc->release_work, ioc_release_fn); +#endif + return ioc; +} - /* - * Try to install. ioc shouldn't be installed if someone else - * already did or @task, which isn't %current, is exiting. Note - * that we need to allow ioc creation on exiting %current as exit - * path may issue IOs from e.g. exit_files(). The exit path is - * responsible for not issuing IO after exit_io_context(). - */ - task_lock(task); - if (!task->io_context && - (task == current || !(task->flags & PF_EXITING))) - task->io_context = ioc; - else - kmem_cache_free(iocontext_cachep, ioc); +int set_task_ioprio(struct task_struct *task, int ioprio) +{ + int err; + const struct cred *cred = current_cred(), *tcred; - ret = task->io_context ? 0 : -EBUSY; + rcu_read_lock(); + tcred = __task_cred(task); + if (!uid_eq(tcred->uid, cred->euid) && + !uid_eq(tcred->uid, cred->uid) && !capable(CAP_SYS_NICE)) { + rcu_read_unlock(); + return -EPERM; + } + rcu_read_unlock(); - task_unlock(task); + err = security_task_setioprio(task, ioprio); + if (err) + return err; - return ret; -} + task_lock(task); + if (unlikely(!task->io_context)) { + struct io_context *ioc; -/** - * get_task_io_context - get io_context of a task - * @task: task of interest - * @gfp_flags: allocation flags, used if allocation is necessary - * @node: allocation node, used if allocation is necessary - * - * Return io_context of @task. If it doesn't exist, it is created with - * @gfp_flags and @node. The returned io_context has its reference count - * incremented. - * - * This function always goes through task_lock() and it's better to use - * %current->io_context + get_io_context() for %current. - */ -struct io_context *get_task_io_context(struct task_struct *task, - gfp_t gfp_flags, int node) -{ - struct io_context *ioc; + task_unlock(task); - might_sleep_if(gfpflags_allow_blocking(gfp_flags)); + ioc = alloc_io_context(GFP_ATOMIC, NUMA_NO_NODE); + if (!ioc) + return -ENOMEM; - do { task_lock(task); - ioc = task->io_context; - if (likely(ioc)) { - get_io_context(ioc); - task_unlock(task); - return ioc; + if (task->flags & PF_EXITING) { + err = -ESRCH; + kmem_cache_free(iocontext_cachep, ioc); + goto out; } - task_unlock(task); - } while (!create_task_io_context(task, gfp_flags, node)); + if (task->io_context) + kmem_cache_free(iocontext_cachep, ioc); + else + task->io_context = ioc; + } + task->io_context->ioprio = ioprio; +out: + task_unlock(task); + return err; +} +EXPORT_SYMBOL_GPL(set_task_ioprio); - return NULL; +int __copy_io(unsigned long clone_flags, struct task_struct *tsk) +{ + struct io_context *ioc = current->io_context; + + /* + * Share io context with parent, if CLONE_IO is set + */ + if (clone_flags & CLONE_IO) { + atomic_inc(&ioc->active_ref); + tsk->io_context = ioc; + } else if (ioprio_valid(ioc->ioprio)) { + tsk->io_context = alloc_io_context(GFP_KERNEL, NUMA_NO_NODE); + if (!tsk->io_context) + return -ENOMEM; + tsk->io_context->ioprio = ioc->ioprio; + } + + return 0; } +#ifdef CONFIG_BLK_ICQ /** * ioc_lookup_icq - lookup io_cq from ioc - * @ioc: the associated io_context * @q: the associated request_queue * * Look up io_cq associated with @ioc - @q pair from @ioc. Must be called * with @q->queue_lock held. */ -struct io_cq *ioc_lookup_icq(struct io_context *ioc, struct request_queue *q) +struct io_cq *ioc_lookup_icq(struct request_queue *q) { + struct io_context *ioc = current->io_context; struct io_cq *icq; lockdep_assert_held(&q->queue_lock); @@ -359,9 +355,7 @@ EXPORT_SYMBOL(ioc_lookup_icq); /** * ioc_create_icq - create and link io_cq - * @ioc: io_context of interest * @q: request_queue of interest - * @gfp_mask: allocation mask * * Make sure io_cq linking @ioc and @q exists. If icq doesn't exist, they * will be created using @gfp_mask. @@ -369,19 +363,19 @@ EXPORT_SYMBOL(ioc_lookup_icq); * The caller is responsible for ensuring @ioc won't go away and @q is * alive and will stay alive until this function returns. */ -struct io_cq *ioc_create_icq(struct io_context *ioc, struct request_queue *q, - gfp_t gfp_mask) +static struct io_cq *ioc_create_icq(struct request_queue *q) { + struct io_context *ioc = current->io_context; struct elevator_type *et = q->elevator->type; struct io_cq *icq; /* allocate stuff */ - icq = kmem_cache_alloc_node(et->icq_cache, gfp_mask | __GFP_ZERO, + icq = kmem_cache_alloc_node(et->icq_cache, GFP_ATOMIC | __GFP_ZERO, q->node); if (!icq) return NULL; - if (radix_tree_maybe_preload(gfp_mask) < 0) { + if (radix_tree_maybe_preload(GFP_ATOMIC) < 0) { kmem_cache_free(et->icq_cache, icq); return NULL; } @@ -402,7 +396,7 @@ struct io_cq *ioc_create_icq(struct io_context *ioc, struct request_queue *q, et->ops.init_icq(icq); } else { kmem_cache_free(et->icq_cache, icq); - icq = ioc_lookup_icq(ioc, q); + icq = ioc_lookup_icq(q); if (!icq) printk(KERN_ERR "cfq: icq link failed!\n"); } @@ -413,6 +407,46 @@ struct io_cq *ioc_create_icq(struct io_context *ioc, struct request_queue *q, return icq; } +struct io_cq *ioc_find_get_icq(struct request_queue *q) +{ + struct io_context *ioc = current->io_context; + struct io_cq *icq = NULL; + + if (unlikely(!ioc)) { + ioc = alloc_io_context(GFP_ATOMIC, q->node); + if (!ioc) + return NULL; + + task_lock(current); + if (current->io_context) { + kmem_cache_free(iocontext_cachep, ioc); + ioc = current->io_context; + } else { + current->io_context = ioc; + } + + get_io_context(ioc); + task_unlock(current); + } else { + get_io_context(ioc); + + spin_lock_irq(&q->queue_lock); + icq = ioc_lookup_icq(q); + spin_unlock_irq(&q->queue_lock); + } + + if (!icq) { + icq = ioc_create_icq(q); + if (!icq) { + put_io_context(ioc); + return NULL; + } + } + return icq; +} +EXPORT_SYMBOL_GPL(ioc_find_get_icq); +#endif /* CONFIG_BLK_ICQ */ + static int __init blk_ioc_init(void) { iocontext_cachep = kmem_cache_create("blkdev_ioc", diff --git a/block/blk-iocost.c b/block/blk-iocost.c index a5b37cc65b17..769b64394298 100644 --- a/block/blk-iocost.c +++ b/block/blk-iocost.c @@ -2311,7 +2311,14 @@ static void ioc_timer_fn(struct timer_list *timer) hwm = current_hweight_max(iocg); new_hwi = hweight_after_donation(iocg, old_hwi, hwm, usage, &now); - if (new_hwi < hwm) { + /* + * Donation calculation assumes hweight_after_donation + * to be positive, a condition that a donor w/ hwa < 2 + * can't meet. Don't bother with donation if hwa is + * below 2. It's not gonna make a meaningful difference + * anyway. + */ + if (new_hwi < hwm && hwa >= 2) { iocg->hweight_donating = hwa; iocg->hweight_after_donation = new_hwi; list_add(&iocg->surplus_list, &surpluses); diff --git a/block/blk-ioprio.c b/block/blk-ioprio.c index 332a07761bf8..2e7f10e1c03f 100644 --- a/block/blk-ioprio.c +++ b/block/blk-ioprio.c @@ -62,6 +62,7 @@ struct ioprio_blkg { struct ioprio_blkcg { struct blkcg_policy_data cpd; enum prio_policy prio_policy; + bool prio_set; }; static inline struct ioprio_blkg *pd_to_ioprio(struct blkg_policy_data *pd) @@ -112,7 +113,7 @@ static ssize_t ioprio_set_prio_policy(struct kernfs_open_file *of, char *buf, if (ret < 0) return ret; blkcg->prio_policy = ret; - + blkcg->prio_set = true; return nbytes; } @@ -190,6 +191,10 @@ static void blkcg_ioprio_track(struct rq_qos *rqos, struct request *rq, struct bio *bio) { struct ioprio_blkcg *blkcg = ioprio_blkcg_from_bio(bio); + u16 prio; + + if (!blkcg->prio_set) + return; /* * Except for IOPRIO_CLASS_NONE, higher I/O priority numbers @@ -199,8 +204,10 @@ static void blkcg_ioprio_track(struct rq_qos *rqos, struct request *rq, * bio I/O priority is not modified. If the bio I/O priority equals * IOPRIO_CLASS_NONE, the cgroup I/O priority is assigned to the bio. */ - bio->bi_ioprio = max_t(u16, bio->bi_ioprio, - IOPRIO_PRIO_VALUE(blkcg->prio_policy, 0)); + prio = max_t(u16, bio->bi_ioprio, + IOPRIO_PRIO_VALUE(blkcg->prio_policy, 0)); + if (prio > bio->bi_ioprio) + bio->bi_ioprio = prio; } static void blkcg_ioprio_exit(struct rq_qos *rqos) diff --git a/block/blk-merge.c b/block/blk-merge.c index 893c1a60b701..4de34a332c9f 100644 --- a/block/blk-merge.c +++ b/block/blk-merge.c @@ -8,10 +8,12 @@ #include <linux/blkdev.h> #include <linux/blk-integrity.h> #include <linux/scatterlist.h> +#include <linux/part_stat.h> #include <trace/events/block.h> #include "blk.h" +#include "blk-mq-sched.h" #include "blk-rq-qos.h" #include "blk-throttle.h" @@ -775,8 +777,7 @@ static struct request *attempt_merge(struct request_queue *q, if (req_op(req) != req_op(next)) return NULL; - if (rq_data_dir(req) != rq_data_dir(next) - || req->rq_disk != next->rq_disk) + if (rq_data_dir(req) != rq_data_dir(next)) return NULL; if (req_op(req) == REQ_OP_WRITE_SAME && @@ -903,10 +904,6 @@ bool blk_rq_merge_ok(struct request *rq, struct bio *bio) if (bio_data_dir(bio) != rq_data_dir(rq)) return false; - /* must be same device */ - if (rq->rq_disk != bio->bi_bdev->bd_disk) - return false; - /* only merge integrity protected bio into ditto rq */ if (blk_integrity_merge_bio(rq->q, rq, bio) == false) return false; @@ -1067,7 +1064,6 @@ static enum bio_merge_status blk_attempt_bio_merge(struct request_queue *q, * @q: request_queue new bio is being queued at * @bio: new bio being queued * @nr_segs: number of segments in @bio - * @same_queue_rq: output value, will be true if there's an existing request * from the passed in @q already in the plug list * * Determine whether @bio being queued on @q can be merged with the previous @@ -1084,7 +1080,7 @@ static enum bio_merge_status blk_attempt_bio_merge(struct request_queue *q, * Caller must ensure !blk_queue_nomerges(q) beforehand. */ bool blk_attempt_plug_merge(struct request_queue *q, struct bio *bio, - unsigned int nr_segs, bool *same_queue_rq) + unsigned int nr_segs) { struct blk_plug *plug; struct request *rq; @@ -1096,12 +1092,6 @@ bool blk_attempt_plug_merge(struct request_queue *q, struct bio *bio, /* check the previously added entry for a quick merge attempt */ rq = rq_list_peek(&plug->mq_list); if (rq->q == q) { - /* - * Only blk-mq multiple hardware queues case checks the rq in - * the same queue, there should be only one such rq in a queue - */ - *same_queue_rq = true; - if (blk_attempt_bio_merge(q, rq, bio, nr_segs, false) == BIO_MERGE_OK) return true; diff --git a/block/blk-mq-debugfs.c b/block/blk-mq-debugfs.c index 4f2cf8399f3d..3a790eb4995c 100644 --- a/block/blk-mq-debugfs.c +++ b/block/blk-mq-debugfs.c @@ -11,6 +11,7 @@ #include "blk.h" #include "blk-mq.h" #include "blk-mq-debugfs.h" +#include "blk-mq-sched.h" #include "blk-mq-tag.h" #include "blk-rq-qos.h" @@ -29,6 +30,9 @@ static int queue_poll_stat_show(void *data, struct seq_file *m) struct request_queue *q = data; int bucket; + if (!q->poll_stat) + return 0; + for (bucket = 0; bucket < (BLK_MQ_POLL_STATS_BKTS / 2); bucket++) { seq_printf(m, "read (%d Bytes): ", 1 << (9 + bucket)); print_stat(m, &q->poll_stat[2 * bucket]); @@ -122,7 +126,6 @@ static const char *const blk_queue_flag_name[] = { QUEUE_FLAG_NAME(FUA), QUEUE_FLAG_NAME(DAX), QUEUE_FLAG_NAME(STATS), - QUEUE_FLAG_NAME(POLL_STATS), QUEUE_FLAG_NAME(REGISTERED), QUEUE_FLAG_NAME(QUIESCED), QUEUE_FLAG_NAME(PCI_P2PDMA), diff --git a/block/blk-mq-sched.c b/block/blk-mq-sched.c index ba21449439cc..55488ba97823 100644 --- a/block/blk-mq-sched.c +++ b/block/blk-mq-sched.c @@ -18,32 +18,6 @@ #include "blk-mq-tag.h" #include "blk-wbt.h" -void blk_mq_sched_assign_ioc(struct request *rq) -{ - struct request_queue *q = rq->q; - struct io_context *ioc; - struct io_cq *icq; - - /* - * May not have an IO context if it's a passthrough request - */ - ioc = current->io_context; - if (!ioc) - return; - - spin_lock_irq(&q->queue_lock); - icq = ioc_lookup_icq(ioc, q); - spin_unlock_irq(&q->queue_lock); - - if (!icq) { - icq = ioc_create_icq(ioc, q, GFP_ATOMIC); - if (!icq) - return; - } - get_io_context(icq->ioc); - rq->elv.icq = icq; -} - /* * Mark a hardware queue as needing a restart. For shared queues, maintain * a count of how many hardware queues are marked for restart. @@ -501,7 +475,8 @@ void blk_mq_sched_insert_requests(struct blk_mq_hw_ctx *hctx, * us one extra enqueue & dequeue to sw queue. */ if (!hctx->dispatch_busy && !run_queue_async) { - blk_mq_try_issue_list_directly(hctx, list); + blk_mq_run_dispatch_ops(hctx->queue, + blk_mq_try_issue_list_directly(hctx, list)); if (list_empty(list)) goto out; } diff --git a/block/blk-mq-sched.h b/block/blk-mq-sched.h index 25d1034952b6..025013972453 100644 --- a/block/blk-mq-sched.h +++ b/block/blk-mq-sched.h @@ -8,8 +8,6 @@ #define MAX_SCHED_RQ (16 * BLKDEV_DEFAULT_RQ) -void blk_mq_sched_assign_ioc(struct request *rq); - bool blk_mq_sched_try_merge(struct request_queue *q, struct bio *bio, unsigned int nr_segs, struct request **merged_request); bool blk_mq_sched_bio_merge(struct request_queue *q, struct bio *bio, diff --git a/block/blk-mq-sysfs.c b/block/blk-mq-sysfs.c index 253c857cba47..674786574075 100644 --- a/block/blk-mq-sysfs.c +++ b/block/blk-mq-sysfs.c @@ -36,8 +36,6 @@ static void blk_mq_hw_sysfs_release(struct kobject *kobj) struct blk_mq_hw_ctx *hctx = container_of(kobj, struct blk_mq_hw_ctx, kobj); - if (hctx->flags & BLK_MQ_F_BLOCKING) - cleanup_srcu_struct(hctx->srcu); blk_free_flush_queue(hctx->fq); sbitmap_free(&hctx->ctx_map); free_cpumask_var(hctx->cpumask); diff --git a/block/blk-mq-tag.c b/block/blk-mq-tag.c index 995336abee33..e55a6834c9a6 100644 --- a/block/blk-mq-tag.c +++ b/block/blk-mq-tag.c @@ -215,7 +215,8 @@ void blk_mq_put_tags(struct blk_mq_tags *tags, int *tag_array, int nr_tags) struct bt_iter_data { struct blk_mq_hw_ctx *hctx; - busy_iter_fn *fn; + struct request_queue *q; + busy_tag_iter_fn *fn; void *data; bool reserved; }; @@ -228,7 +229,7 @@ static struct request *blk_mq_find_and_get_req(struct blk_mq_tags *tags, spin_lock_irqsave(&tags->lock, flags); rq = tags->rqs[bitnr]; - if (!rq || rq->tag != bitnr || !refcount_inc_not_zero(&rq->ref)) + if (!rq || rq->tag != bitnr || !req_ref_inc_not_zero(rq)) rq = NULL; spin_unlock_irqrestore(&tags->lock, flags); return rq; @@ -238,11 +239,18 @@ static bool bt_iter(struct sbitmap *bitmap, unsigned int bitnr, void *data) { struct bt_iter_data *iter_data = data; struct blk_mq_hw_ctx *hctx = iter_data->hctx; - struct blk_mq_tags *tags = hctx->tags; + struct request_queue *q = iter_data->q; + struct blk_mq_tag_set *set = q->tag_set; bool reserved = iter_data->reserved; + struct blk_mq_tags *tags; struct request *rq; bool ret = true; + if (blk_mq_is_shared_tags(set->flags)) + tags = set->shared_tags; + else + tags = hctx->tags; + if (!reserved) bitnr += tags->nr_reserved_tags; /* @@ -253,8 +261,8 @@ static bool bt_iter(struct sbitmap *bitmap, unsigned int bitnr, void *data) if (!rq) return true; - if (rq->q == hctx->queue && rq->mq_hctx == hctx) - ret = iter_data->fn(hctx, rq, iter_data->data, reserved); + if (rq->q == q && (!hctx || rq->mq_hctx == hctx)) + ret = iter_data->fn(rq, iter_data->data, reserved); blk_mq_put_rq_ref(rq); return ret; } @@ -262,6 +270,7 @@ static bool bt_iter(struct sbitmap *bitmap, unsigned int bitnr, void *data) /** * bt_for_each - iterate over the requests associated with a hardware queue * @hctx: Hardware queue to examine. + * @q: Request queue to examine. * @bt: sbitmap to examine. This is either the breserved_tags member * or the bitmap_tags member of struct blk_mq_tags. * @fn: Pointer to the function that will be called for each request @@ -273,14 +282,16 @@ static bool bt_iter(struct sbitmap *bitmap, unsigned int bitnr, void *data) * @reserved: Indicates whether @bt is the breserved_tags member or the * bitmap_tags member of struct blk_mq_tags. */ -static void bt_for_each(struct blk_mq_hw_ctx *hctx, struct sbitmap_queue *bt, - busy_iter_fn *fn, void *data, bool reserved) +static void bt_for_each(struct blk_mq_hw_ctx *hctx, struct request_queue *q, + struct sbitmap_queue *bt, busy_tag_iter_fn *fn, + void *data, bool reserved) { struct bt_iter_data iter_data = { .hctx = hctx, .fn = fn, .data = data, .reserved = reserved, + .q = q, }; sbitmap_for_each_set(&bt->sb, bt_iter, &iter_data); @@ -457,12 +468,9 @@ EXPORT_SYMBOL(blk_mq_tagset_wait_completed_request); * called for all requests on all queues that share that tag set and not only * for requests associated with @q. */ -void blk_mq_queue_tag_busy_iter(struct request_queue *q, busy_iter_fn *fn, +void blk_mq_queue_tag_busy_iter(struct request_queue *q, busy_tag_iter_fn *fn, void *priv) { - struct blk_mq_hw_ctx *hctx; - int i; - /* * __blk_mq_update_nr_hw_queues() updates nr_hw_queues and queue_hw_ctx * while the queue is frozen. So we can use q_usage_counter to avoid @@ -471,19 +479,34 @@ void blk_mq_queue_tag_busy_iter(struct request_queue *q, busy_iter_fn *fn, if (!percpu_ref_tryget(&q->q_usage_counter)) return; - queue_for_each_hw_ctx(q, hctx, i) { - struct blk_mq_tags *tags = hctx->tags; - - /* - * If no software queues are currently mapped to this - * hardware queue, there's nothing to check - */ - if (!blk_mq_hw_queue_mapped(hctx)) - continue; + if (blk_mq_is_shared_tags(q->tag_set->flags)) { + struct blk_mq_tags *tags = q->tag_set->shared_tags; + struct sbitmap_queue *bresv = &tags->breserved_tags; + struct sbitmap_queue *btags = &tags->bitmap_tags; if (tags->nr_reserved_tags) - bt_for_each(hctx, &tags->breserved_tags, fn, priv, true); - bt_for_each(hctx, &tags->bitmap_tags, fn, priv, false); + bt_for_each(NULL, q, bresv, fn, priv, true); + bt_for_each(NULL, q, btags, fn, priv, false); + } else { + struct blk_mq_hw_ctx *hctx; + int i; + + queue_for_each_hw_ctx(q, hctx, i) { + struct blk_mq_tags *tags = hctx->tags; + struct sbitmap_queue *bresv = &tags->breserved_tags; + struct sbitmap_queue *btags = &tags->bitmap_tags; + + /* + * If no software queues are currently mapped to this + * hardware queue, there's nothing to check + */ + if (!blk_mq_hw_queue_mapped(hctx)) + continue; + + if (tags->nr_reserved_tags) + bt_for_each(hctx, q, bresv, fn, priv, true); + bt_for_each(hctx, q, btags, fn, priv, false); + } } blk_queue_exit(q); } diff --git a/block/blk-mq-tag.h b/block/blk-mq-tag.h index df787b5a23bd..5668e28be0b7 100644 --- a/block/blk-mq-tag.h +++ b/block/blk-mq-tag.h @@ -28,7 +28,7 @@ extern void blk_mq_tag_resize_shared_tags(struct blk_mq_tag_set *set, extern void blk_mq_tag_update_sched_shared_tags(struct request_queue *q); extern void blk_mq_tag_wakeup_all(struct blk_mq_tags *tags, bool); -void blk_mq_queue_tag_busy_iter(struct request_queue *q, busy_iter_fn *fn, +void blk_mq_queue_tag_busy_iter(struct request_queue *q, busy_tag_iter_fn *fn, void *priv); void blk_mq_all_tag_iter(struct blk_mq_tags *tags, busy_tag_iter_fn *fn, void *priv); diff --git a/block/blk-mq.c b/block/blk-mq.c index 3ab34c4f20da..a6d4780580fc 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -28,6 +28,7 @@ #include <linux/crash_dump.h> #include <linux/prefetch.h> #include <linux/blk-crypto.h> +#include <linux/part_stat.h> #include <trace/events/block.h> @@ -126,8 +127,7 @@ struct mq_inflight { unsigned int inflight[2]; }; -static bool blk_mq_check_inflight(struct blk_mq_hw_ctx *hctx, - struct request *rq, void *priv, +static bool blk_mq_check_inflight(struct request *rq, void *priv, bool reserved) { struct mq_inflight *mi = priv; @@ -259,17 +259,9 @@ EXPORT_SYMBOL_GPL(blk_mq_quiesce_queue_nowait); */ void blk_mq_wait_quiesce_done(struct request_queue *q) { - struct blk_mq_hw_ctx *hctx; - unsigned int i; - bool rcu = false; - - queue_for_each_hw_ctx(q, hctx, i) { - if (hctx->flags & BLK_MQ_F_BLOCKING) - synchronize_srcu(hctx->srcu); - else - rcu = true; - } - if (rcu) + if (blk_queue_has_srcu(q)) + synchronize_srcu(q->srcu); + else synchronize_rcu(); } EXPORT_SYMBOL_GPL(blk_mq_wait_quiesce_done); @@ -327,6 +319,23 @@ void blk_mq_wake_waiters(struct request_queue *q) blk_mq_tag_wakeup_all(hctx->tags, true); } +void blk_rq_init(struct request_queue *q, struct request *rq) +{ + memset(rq, 0, sizeof(*rq)); + + INIT_LIST_HEAD(&rq->queuelist); + rq->q = q; + rq->__sector = (sector_t) -1; + INIT_HLIST_NODE(&rq->hash); + RB_CLEAR_NODE(&rq->rb_node); + rq->tag = BLK_MQ_NO_TAG; + rq->internal_tag = BLK_MQ_NO_TAG; + rq->start_time_ns = ktime_get_ns(); + rq->part = NULL; + blk_crypto_rq_set_defaults(rq); +} +EXPORT_SYMBOL(blk_rq_init); + static struct request *blk_mq_rq_ctx_init(struct blk_mq_alloc_data *data, struct blk_mq_tags *tags, unsigned int tag, u64 alloc_time_ns) { @@ -359,7 +368,6 @@ static struct request *blk_mq_rq_ctx_init(struct blk_mq_alloc_data *data, rq->start_time_ns = ktime_get_ns(); else rq->start_time_ns = 0; - rq->rq_disk = NULL; rq->part = NULL; #ifdef CONFIG_BLK_RQ_ALLOC_TIME rq->alloc_time_ns = alloc_time_ns; @@ -377,20 +385,16 @@ static struct request *blk_mq_rq_ctx_init(struct blk_mq_alloc_data *data, INIT_LIST_HEAD(&rq->queuelist); /* tag was already set */ WRITE_ONCE(rq->deadline, 0); - refcount_set(&rq->ref, 1); + req_ref_set(rq, 1); if (rq->rq_flags & RQF_ELV) { struct elevator_queue *e = data->q->elevator; - rq->elv.icq = NULL; INIT_HLIST_NODE(&rq->hash); RB_CLEAR_NODE(&rq->rb_node); if (!op_is_flush(data->cmd_flags) && e->type->ops.prepare_request) { - if (e->type->icq_cache) - blk_mq_sched_assign_ioc(rq); - e->type->ops.prepare_request(rq); rq->rq_flags |= RQF_ELVPRIV; } @@ -616,16 +620,9 @@ void blk_mq_free_request(struct request *rq) struct request_queue *q = rq->q; struct blk_mq_hw_ctx *hctx = rq->mq_hctx; - if (rq->rq_flags & RQF_ELVPRIV) { - struct elevator_queue *e = q->elevator; - - if (e->type->ops.finish_request) - e->type->ops.finish_request(rq); - if (rq->elv.icq) { - put_io_context(rq->elv.icq->ioc); - rq->elv.icq = NULL; - } - } + if ((rq->rq_flags & RQF_ELVPRIV) && + q->elevator->type->ops.finish_request) + q->elevator->type->ops.finish_request(rq); if (rq->rq_flags & RQF_MQ_INFLIGHT) __blk_mq_dec_active_requests(hctx); @@ -636,7 +633,7 @@ void blk_mq_free_request(struct request *rq) rq_qos_done(q, rq); WRITE_ONCE(rq->state, MQ_RQ_IDLE); - if (refcount_dec_and_test(&rq->ref)) + if (req_ref_put_and_test(rq)) __blk_mq_free_request(rq); } EXPORT_SYMBOL_GPL(blk_mq_free_request); @@ -649,6 +646,20 @@ void blk_mq_free_plug_rqs(struct blk_plug *plug) blk_mq_free_request(rq); } +void blk_dump_rq_flags(struct request *rq, char *msg) +{ + printk(KERN_INFO "%s: dev %s: flags=%llx\n", msg, + rq->q->disk ? rq->q->disk->disk_name : "?", + (unsigned long long) rq->cmd_flags); + + printk(KERN_INFO " sector %llu, nr/cnr %u/%u\n", + (unsigned long long)blk_rq_pos(rq), + blk_rq_sectors(rq), blk_rq_cur_sectors(rq)); + printk(KERN_INFO " bio %p, biotail %p, len %u\n", + rq->bio, rq->biotail, blk_rq_bytes(rq)); +} +EXPORT_SYMBOL(blk_dump_rq_flags); + static void req_bio_endio(struct request *rq, struct bio *bio, unsigned int nbytes, blk_status_t error) { @@ -685,6 +696,60 @@ static void blk_account_io_completion(struct request *req, unsigned int bytes) } } +static void blk_print_req_error(struct request *req, blk_status_t status) +{ + printk_ratelimited(KERN_ERR + "%s error, dev %s, sector %llu op 0x%x:(%s) flags 0x%x " + "phys_seg %u prio class %u\n", + blk_status_to_str(status), + req->q->disk ? req->q->disk->disk_name : "?", + blk_rq_pos(req), req_op(req), blk_op_str(req_op(req)), + req->cmd_flags & ~REQ_OP_MASK, + req->nr_phys_segments, + IOPRIO_PRIO_CLASS(req->ioprio)); +} + +/* + * Fully end IO on a request. Does not support partial completions, or + * errors. + */ +static void blk_complete_request(struct request *req) +{ + const bool is_flush = (req->rq_flags & RQF_FLUSH_SEQ) != 0; + int total_bytes = blk_rq_bytes(req); + struct bio *bio = req->bio; + + trace_block_rq_complete(req, BLK_STS_OK, total_bytes); + + if (!bio) + return; + +#ifdef CONFIG_BLK_DEV_INTEGRITY + if (blk_integrity_rq(req) && req_op(req) == REQ_OP_READ) + req->q->integrity.profile->complete_fn(req, total_bytes); +#endif + + blk_account_io_completion(req, total_bytes); + + do { + struct bio *next = bio->bi_next; + + /* Completion has already been traced */ + bio_clear_flag(bio, BIO_TRACE_COMPLETION); + if (!is_flush) + bio_endio(bio); + bio = next; + } while (bio); + + /* + * Reset counters so that the request stacking driver + * can find how many bytes remain in the request + * later. + */ + req->bio = NULL; + req->__data_len = 0; +} + /** * blk_update_request - Complete multiple bytes without completing the request * @req: the request being processed @@ -791,6 +856,48 @@ bool blk_update_request(struct request *req, blk_status_t error, } EXPORT_SYMBOL_GPL(blk_update_request); +static void __blk_account_io_done(struct request *req, u64 now) +{ + const int sgrp = op_stat_group(req_op(req)); + + part_stat_lock(); + update_io_ticks(req->part, jiffies, true); + part_stat_inc(req->part, ios[sgrp]); + part_stat_add(req->part, nsecs[sgrp], now - req->start_time_ns); + part_stat_unlock(); +} + +static inline void blk_account_io_done(struct request *req, u64 now) +{ + /* + * Account IO completion. flush_rq isn't accounted as a + * normal IO on queueing nor completion. Accounting the + * containing request is enough. + */ + if (blk_do_io_stat(req) && req->part && + !(req->rq_flags & RQF_FLUSH_SEQ)) + __blk_account_io_done(req, now); +} + +static void __blk_account_io_start(struct request *rq) +{ + /* passthrough requests can hold bios that do not have ->bi_bdev set */ + if (rq->bio && rq->bio->bi_bdev) + rq->part = rq->bio->bi_bdev; + else if (rq->q->disk) + rq->part = rq->q->disk->part0; + + part_stat_lock(); + update_io_ticks(rq->part, jiffies, false); + part_stat_unlock(); +} + +static inline void blk_account_io_start(struct request *req) +{ + if (blk_do_io_stat(req)) + __blk_account_io_start(req); +} + static inline void __blk_mq_end_request_acct(struct request *rq, u64 now) { if (rq->rq_flags & RQF_STATS) { @@ -856,17 +963,18 @@ void blk_mq_end_request_batch(struct io_comp_batch *iob) prefetch(rq->bio); prefetch(rq->rq_next); - blk_update_request(rq, BLK_STS_OK, blk_rq_bytes(rq)); + blk_complete_request(rq); if (iob->need_ts) __blk_mq_end_request_acct(rq, now); + rq_qos_done(rq->q, rq); + WRITE_ONCE(rq->state, MQ_RQ_IDLE); - if (!refcount_dec_and_test(&rq->ref)) + if (!req_ref_put_and_test(rq)) continue; blk_crypto_free_request(rq); blk_pm_mark_last_busy(rq); - rq_qos_done(rq->q, rq); if (nr_tags == TAG_COMP_BATCH || cur_hctx != rq->mq_hctx) { if (cur_hctx) @@ -995,26 +1103,6 @@ void blk_mq_complete_request(struct request *rq) } EXPORT_SYMBOL(blk_mq_complete_request); -static void hctx_unlock(struct blk_mq_hw_ctx *hctx, int srcu_idx) - __releases(hctx->srcu) -{ - if (!(hctx->flags & BLK_MQ_F_BLOCKING)) - rcu_read_unlock(); - else - srcu_read_unlock(hctx->srcu, srcu_idx); -} - -static void hctx_lock(struct blk_mq_hw_ctx *hctx, int *srcu_idx) - __acquires(hctx->srcu) -{ - if (!(hctx->flags & BLK_MQ_F_BLOCKING)) { - /* shut up gcc false positive */ - *srcu_idx = 0; - rcu_read_lock(); - } else - *srcu_idx = srcu_read_lock(hctx->srcu); -} - /** * blk_mq_start_request - Start processing a request * @rq: Pointer to request to be started @@ -1057,6 +1145,107 @@ void blk_mq_start_request(struct request *rq) } EXPORT_SYMBOL(blk_mq_start_request); +/** + * blk_end_sync_rq - executes a completion event on a request + * @rq: request to complete + * @error: end I/O status of the request + */ +static void blk_end_sync_rq(struct request *rq, blk_status_t error) +{ + struct completion *waiting = rq->end_io_data; + + rq->end_io_data = (void *)(uintptr_t)error; + + /* + * complete last, if this is a stack request the process (and thus + * the rq pointer) could be invalid right after this complete() + */ + complete(waiting); +} + +/** + * blk_execute_rq_nowait - insert a request to I/O scheduler for execution + * @rq: request to insert + * @at_head: insert request at head or tail of queue + * @done: I/O completion handler + * + * Description: + * Insert a fully prepared request at the back of the I/O scheduler queue + * for execution. Don't wait for completion. + * + * Note: + * This function will invoke @done directly if the queue is dead. + */ +void blk_execute_rq_nowait(struct request *rq, bool at_head, rq_end_io_fn *done) +{ + WARN_ON(irqs_disabled()); + WARN_ON(!blk_rq_is_passthrough(rq)); + + rq->end_io = done; + + blk_account_io_start(rq); + + /* + * don't check dying flag for MQ because the request won't + * be reused after dying flag is set + */ + blk_mq_sched_insert_request(rq, at_head, true, false); +} +EXPORT_SYMBOL_GPL(blk_execute_rq_nowait); + +static bool blk_rq_is_poll(struct request *rq) +{ + if (!rq->mq_hctx) + return false; + if (rq->mq_hctx->type != HCTX_TYPE_POLL) + return false; + if (WARN_ON_ONCE(!rq->bio)) + return false; + return true; +} + +static void blk_rq_poll_completion(struct request *rq, struct completion *wait) +{ + do { + bio_poll(rq->bio, NULL, 0); + cond_resched(); + } while (!completion_done(wait)); +} + +/** + * blk_execute_rq - insert a request into queue for execution + * @rq: request to insert + * @at_head: insert request at head or tail of queue + * + * Description: + * Insert a fully prepared request at the back of the I/O scheduler queue + * for execution and wait for completion. + * Return: The blk_status_t result provided to blk_mq_end_request(). + */ +blk_status_t blk_execute_rq(struct request *rq, bool at_head) +{ + DECLARE_COMPLETION_ONSTACK(wait); + unsigned long hang_check; + + rq->end_io_data = &wait; + blk_execute_rq_nowait(rq, at_head, blk_end_sync_rq); + + /* Prevent hang_check timer from firing at us during very long I/O */ + hang_check = sysctl_hung_task_timeout_secs; + + if (blk_rq_is_poll(rq)) + blk_rq_poll_completion(rq, &wait); + else if (hang_check) + while (!wait_for_completion_io_timeout(&wait, + hang_check * (HZ/2))) + ; + else + wait_for_completion_io(&wait); + + return (blk_status_t)(uintptr_t)rq->end_io_data; +} +EXPORT_SYMBOL(blk_execute_rq); + static void __blk_mq_requeue_request(struct request *rq) { struct request_queue *q = rq->q; @@ -1159,14 +1348,15 @@ void blk_mq_delay_kick_requeue_list(struct request_queue *q, } EXPORT_SYMBOL(blk_mq_delay_kick_requeue_list); -static bool blk_mq_rq_inflight(struct blk_mq_hw_ctx *hctx, struct request *rq, - void *priv, bool reserved) +static bool blk_mq_rq_inflight(struct request *rq, void *priv, + bool reserved) { /* - * If we find a request that isn't idle and the queue matches, - * we know the queue is busy. Return false to stop the iteration. + * 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. */ - if (blk_mq_request_started(rq) && rq->q == hctx->queue) { + if (blk_mq_request_started(rq)) { bool *busy = priv; *busy = true; @@ -1224,12 +1414,11 @@ void blk_mq_put_rq_ref(struct request *rq) { if (is_flush_rq(rq)) rq->end_io(rq, 0); - else if (refcount_dec_and_test(&rq->ref)) + else if (req_ref_put_and_test(rq)) __blk_mq_free_request(rq); } -static bool blk_mq_check_expired(struct blk_mq_hw_ctx *hctx, - struct request *rq, void *priv, bool reserved) +static bool blk_mq_check_expired(struct request *rq, void *priv, bool reserved) { unsigned long *next = priv; @@ -1770,19 +1959,14 @@ out: */ static void __blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx) { - int srcu_idx; - /* * We can't run the queue inline with ints disabled. Ensure that * we catch bad users of this early. */ WARN_ON_ONCE(in_interrupt()); - might_sleep_if(hctx->flags & BLK_MQ_F_BLOCKING); - - hctx_lock(hctx, &srcu_idx); - blk_mq_sched_dispatch_requests(hctx); - hctx_unlock(hctx, srcu_idx); + blk_mq_run_dispatch_ops(hctx->queue, + blk_mq_sched_dispatch_requests(hctx)); } static inline int blk_mq_first_mapped_cpu(struct blk_mq_hw_ctx *hctx) @@ -1894,7 +2078,6 @@ EXPORT_SYMBOL(blk_mq_delay_run_hw_queue); */ void blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async) { - int srcu_idx; bool need_run; /* @@ -1905,10 +2088,9 @@ void blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async) * And queue will be rerun in blk_mq_unquiesce_queue() if it is * quiesced. */ - hctx_lock(hctx, &srcu_idx); - need_run = !blk_queue_quiesced(hctx->queue) && - blk_mq_hctx_has_pending(hctx); - hctx_unlock(hctx, srcu_idx); + __blk_mq_run_dispatch_ops(hctx->queue, false, + need_run = !blk_queue_quiesced(hctx->queue) && + blk_mq_hctx_has_pending(hctx)); if (need_run) __blk_mq_delay_run_hw_queue(hctx, async, 0); @@ -2201,98 +2383,6 @@ static void blk_mq_commit_rqs(struct blk_mq_hw_ctx *hctx, int *queued, *queued = 0; } -static void blk_mq_plug_issue_direct(struct blk_plug *plug, bool from_schedule) -{ - struct blk_mq_hw_ctx *hctx = NULL; - struct request *rq; - int queued = 0; - int errors = 0; - - while ((rq = rq_list_pop(&plug->mq_list))) { - bool last = rq_list_empty(plug->mq_list); - blk_status_t ret; - - if (hctx != rq->mq_hctx) { - if (hctx) - blk_mq_commit_rqs(hctx, &queued, from_schedule); - hctx = rq->mq_hctx; - } - - ret = blk_mq_request_issue_directly(rq, last); - switch (ret) { - case BLK_STS_OK: - queued++; - break; - case BLK_STS_RESOURCE: - case BLK_STS_DEV_RESOURCE: - blk_mq_request_bypass_insert(rq, false, last); - blk_mq_commit_rqs(hctx, &queued, from_schedule); - return; - default: - blk_mq_end_request(rq, ret); - errors++; - break; - } - } - - /* - * If we didn't flush the entire list, we could have told the driver - * there was more coming, but that turned out to be a lie. - */ - if (errors) - blk_mq_commit_rqs(hctx, &queued, from_schedule); -} - -void blk_mq_flush_plug_list(struct blk_plug *plug, bool from_schedule) -{ - struct blk_mq_hw_ctx *this_hctx; - struct blk_mq_ctx *this_ctx; - unsigned int depth; - LIST_HEAD(list); - - if (rq_list_empty(plug->mq_list)) - return; - plug->rq_count = 0; - - if (!plug->multiple_queues && !plug->has_elevator && !from_schedule) { - blk_mq_plug_issue_direct(plug, false); - if (rq_list_empty(plug->mq_list)) - return; - } - - this_hctx = NULL; - this_ctx = NULL; - depth = 0; - do { - struct request *rq; - - rq = rq_list_pop(&plug->mq_list); - - if (!this_hctx) { - this_hctx = rq->mq_hctx; - this_ctx = rq->mq_ctx; - } else if (this_hctx != rq->mq_hctx || this_ctx != rq->mq_ctx) { - trace_block_unplug(this_hctx->queue, depth, - !from_schedule); - blk_mq_sched_insert_requests(this_hctx, this_ctx, - &list, from_schedule); - depth = 0; - this_hctx = rq->mq_hctx; - this_ctx = rq->mq_ctx; - - } - - list_add(&rq->queuelist, &list); - depth++; - } while (!rq_list_empty(plug->mq_list)); - - if (!list_empty(&list)) { - trace_block_unplug(this_hctx->queue, depth, !from_schedule); - blk_mq_sched_insert_requests(this_hctx, this_ctx, &list, - from_schedule); - } -} - static void blk_mq_bio_to_request(struct request *rq, struct bio *bio, unsigned int nr_segs) { @@ -2403,33 +2493,141 @@ insert: static void blk_mq_try_issue_directly(struct blk_mq_hw_ctx *hctx, struct request *rq) { - blk_status_t ret; - int srcu_idx; - - might_sleep_if(hctx->flags & BLK_MQ_F_BLOCKING); - - hctx_lock(hctx, &srcu_idx); + blk_status_t ret = + __blk_mq_try_issue_directly(hctx, rq, false, true); - ret = __blk_mq_try_issue_directly(hctx, rq, false, true); if (ret == BLK_STS_RESOURCE || ret == BLK_STS_DEV_RESOURCE) blk_mq_request_bypass_insert(rq, false, true); else if (ret != BLK_STS_OK) blk_mq_end_request(rq, ret); +} - hctx_unlock(hctx, srcu_idx); +static blk_status_t blk_mq_request_issue_directly(struct request *rq, bool last) +{ + return __blk_mq_try_issue_directly(rq->mq_hctx, rq, true, last); } -blk_status_t blk_mq_request_issue_directly(struct request *rq, bool last) +static void blk_mq_plug_issue_direct(struct blk_plug *plug, bool from_schedule) { - blk_status_t ret; - int srcu_idx; - struct blk_mq_hw_ctx *hctx = rq->mq_hctx; + struct blk_mq_hw_ctx *hctx = NULL; + struct request *rq; + int queued = 0; + int errors = 0; - hctx_lock(hctx, &srcu_idx); - ret = __blk_mq_try_issue_directly(hctx, rq, true, last); - hctx_unlock(hctx, srcu_idx); + while ((rq = rq_list_pop(&plug->mq_list))) { + bool last = rq_list_empty(plug->mq_list); + blk_status_t ret; - return ret; + if (hctx != rq->mq_hctx) { + if (hctx) + blk_mq_commit_rqs(hctx, &queued, from_schedule); + hctx = rq->mq_hctx; + } + + ret = blk_mq_request_issue_directly(rq, last); + switch (ret) { + case BLK_STS_OK: + queued++; + break; + case BLK_STS_RESOURCE: + case BLK_STS_DEV_RESOURCE: + blk_mq_request_bypass_insert(rq, false, last); + blk_mq_commit_rqs(hctx, &queued, from_schedule); + return; + default: + blk_mq_end_request(rq, ret); + errors++; + break; + } + } + + /* + * If we didn't flush the entire list, we could have told the driver + * there was more coming, but that turned out to be a lie. + */ + if (errors) + blk_mq_commit_rqs(hctx, &queued, from_schedule); +} + +static void __blk_mq_flush_plug_list(struct request_queue *q, + struct blk_plug *plug) +{ + if (blk_queue_quiesced(q)) + return; + q->mq_ops->queue_rqs(&plug->mq_list); +} + +void blk_mq_flush_plug_list(struct blk_plug *plug, bool from_schedule) +{ + struct blk_mq_hw_ctx *this_hctx; + struct blk_mq_ctx *this_ctx; + struct request *rq; + unsigned int depth; + LIST_HEAD(list); + + if (rq_list_empty(plug->mq_list)) + return; + plug->rq_count = 0; + + if (!plug->multiple_queues && !plug->has_elevator && !from_schedule) { + struct request_queue *q; + + rq = rq_list_peek(&plug->mq_list); + q = rq->q; + + /* + * Peek first request and see if we have a ->queue_rqs() hook. + * If we do, we can dispatch the whole plug list in one go. We + * already know at this point that all requests belong to the + * same queue, caller must ensure that's the case. + * + * Since we pass off the full list to the driver at this point, + * we do not increment the active request count for the queue. + * Bypass shared tags for now because of that. + */ + if (q->mq_ops->queue_rqs && + !(rq->mq_hctx->flags & BLK_MQ_F_TAG_QUEUE_SHARED)) { + blk_mq_run_dispatch_ops(q, + __blk_mq_flush_plug_list(q, plug)); + if (rq_list_empty(plug->mq_list)) + return; + } + + blk_mq_run_dispatch_ops(q, + blk_mq_plug_issue_direct(plug, false)); + if (rq_list_empty(plug->mq_list)) + return; + } + + this_hctx = NULL; + this_ctx = NULL; + depth = 0; + do { + rq = rq_list_pop(&plug->mq_list); + + if (!this_hctx) { + this_hctx = rq->mq_hctx; + this_ctx = rq->mq_ctx; + } else if (this_hctx != rq->mq_hctx || this_ctx != rq->mq_ctx) { + trace_block_unplug(this_hctx->queue, depth, + !from_schedule); + blk_mq_sched_insert_requests(this_hctx, this_ctx, + &list, from_schedule); + depth = 0; + this_hctx = rq->mq_hctx; + this_ctx = rq->mq_ctx; + + } + + list_add(&rq->queuelist, &list); + depth++; + } while (!rq_list_empty(plug->mq_list)); + + if (!list_empty(&list)) { + trace_block_unplug(this_hctx->queue, depth, !from_schedule); + blk_mq_sched_insert_requests(this_hctx, this_ctx, &list, + from_schedule); + } } void blk_mq_try_issue_list_directly(struct blk_mq_hw_ctx *hctx, @@ -2468,21 +2666,6 @@ void blk_mq_try_issue_list_directly(struct blk_mq_hw_ctx *hctx, hctx->queue->mq_ops->commit_rqs(hctx); } -static void blk_add_rq_to_plug(struct blk_plug *plug, struct request *rq) -{ - if (!plug->multiple_queues) { - struct request *nxt = rq_list_peek(&plug->mq_list); - - if (nxt && nxt->q != rq->q) - plug->multiple_queues = true; - } - if (!plug->has_elevator && (rq->rq_flags & RQF_ELV)) - plug->has_elevator = true; - rq->rq_next = NULL; - rq_list_add(&plug->mq_list, rq); - plug->rq_count++; -} - /* * Allow 2x BLK_MAX_REQUEST_COUNT requests on plug queue for multiple * queues. This is important for md arrays to benefit from merging @@ -2495,12 +2678,33 @@ static inline unsigned short blk_plug_max_rq_count(struct blk_plug *plug) return BLK_MAX_REQUEST_COUNT; } +static void blk_add_rq_to_plug(struct blk_plug *plug, struct request *rq) +{ + struct request *last = rq_list_peek(&plug->mq_list); + + if (!plug->rq_count) { + trace_block_plug(rq->q); + } else if (plug->rq_count >= blk_plug_max_rq_count(plug) || + (!blk_queue_nomerges(rq->q) && + blk_rq_bytes(last) >= BLK_PLUG_FLUSH_SIZE)) { + blk_mq_flush_plug_list(plug, false); + trace_block_plug(rq->q); + } + + if (!plug->multiple_queues && last && last->q != rq->q) + plug->multiple_queues = true; + if (!plug->has_elevator && (rq->rq_flags & RQF_ELV)) + plug->has_elevator = true; + rq->rq_next = NULL; + rq_list_add(&plug->mq_list, rq); + plug->rq_count++; +} + static bool blk_mq_attempt_bio_merge(struct request_queue *q, - struct bio *bio, unsigned int nr_segs, - bool *same_queue_rq) + struct bio *bio, unsigned int nr_segs) { if (!blk_queue_nomerges(q) && bio_mergeable(bio)) { - if (blk_attempt_plug_merge(q, bio, nr_segs, same_queue_rq)) + if (blk_attempt_plug_merge(q, bio, nr_segs)) return true; if (blk_mq_sched_bio_merge(q, bio, nr_segs)) return true; @@ -2510,9 +2714,7 @@ static bool blk_mq_attempt_bio_merge(struct request_queue *q, static struct request *blk_mq_get_new_requests(struct request_queue *q, struct blk_plug *plug, - struct bio *bio, - unsigned int nsegs, - bool *same_queue_rq) + struct bio *bio) { struct blk_mq_alloc_data data = { .q = q, @@ -2521,11 +2723,9 @@ static struct request *blk_mq_get_new_requests(struct request_queue *q, }; struct request *rq; - if (blk_mq_attempt_bio_merge(q, bio, nsegs, same_queue_rq)) + if (unlikely(bio_queue_enter(bio))) return NULL; - rq_qos_throttle(q, bio); - if (plug) { data.nr_tags = plug->nr_ios; plug->nr_ios = 1; @@ -2535,63 +2735,32 @@ static struct request *blk_mq_get_new_requests(struct request_queue *q, rq = __blk_mq_alloc_requests(&data); if (rq) return rq; - rq_qos_cleanup(q, bio); if (bio->bi_opf & REQ_NOWAIT) bio_wouldblock_error(bio); - + blk_queue_exit(q); return NULL; } -static inline bool blk_mq_can_use_cached_rq(struct request *rq, - struct bio *bio) -{ - if (blk_mq_get_hctx_type(bio->bi_opf) != rq->mq_hctx->type) - return false; - - if (op_is_flush(rq->cmd_flags) != op_is_flush(bio->bi_opf)) - return false; - - return true; -} - -static inline struct request *blk_mq_get_request(struct request_queue *q, - struct blk_plug *plug, - struct bio *bio, - unsigned int nsegs, - bool *same_queue_rq) +static inline struct request *blk_mq_get_cached_request(struct request_queue *q, + struct blk_plug *plug, struct bio *bio) { struct request *rq; - bool checked = false; - - if (plug) { - rq = rq_list_peek(&plug->cached_rq); - if (rq && rq->q == q) { - if (unlikely(!submit_bio_checks(bio))) - return NULL; - if (blk_mq_attempt_bio_merge(q, bio, nsegs, - same_queue_rq)) - return NULL; - checked = true; - if (!blk_mq_can_use_cached_rq(rq, bio)) - goto fallback; - rq->cmd_flags = bio->bi_opf; - plug->cached_rq = rq_list_next(rq); - INIT_LIST_HEAD(&rq->queuelist); - rq_qos_throttle(q, bio); - return rq; - } - } + if (!plug) + return NULL; + rq = rq_list_peek(&plug->cached_rq); + if (!rq || rq->q != q) + return NULL; -fallback: - if (unlikely(bio_queue_enter(bio))) + if (blk_mq_get_hctx_type(bio->bi_opf) != rq->mq_hctx->type) return NULL; - if (!checked && !submit_bio_checks(bio)) + if (op_is_flush(rq->cmd_flags) != op_is_flush(bio->bi_opf)) return NULL; - rq = blk_mq_get_new_requests(q, plug, bio, nsegs, same_queue_rq); - if (!rq) - blk_queue_exit(q); + + rq->cmd_flags = bio->bi_opf; + plug->cached_rq = rq_list_next(rq); + INIT_LIST_HEAD(&rq->queuelist); return rq; } @@ -2611,10 +2780,9 @@ fallback: void blk_mq_submit_bio(struct bio *bio) { struct request_queue *q = bdev_get_queue(bio->bi_bdev); + struct blk_plug *plug = blk_mq_plug(q, bio); const int is_sync = op_is_sync(bio->bi_opf); struct request *rq; - struct blk_plug *plug; - bool same_queue_rq = false; unsigned int nr_segs = 1; blk_status_t ret; @@ -2628,11 +2796,18 @@ void blk_mq_submit_bio(struct bio *bio) if (!bio_integrity_prep(bio)) return; - plug = blk_mq_plug(q, bio); - rq = blk_mq_get_request(q, plug, bio, nr_segs, &same_queue_rq); - if (unlikely(!rq)) + if (blk_mq_attempt_bio_merge(q, bio, nr_segs)) return; + rq_qos_throttle(q, bio); + + rq = blk_mq_get_cached_request(q, plug, bio); + if (!rq) { + rq = blk_mq_get_new_requests(q, plug, bio); + if (unlikely(!rq)) + return; + } + trace_block_getrq(bio); rq_qos_track(q, rq, bio); @@ -2647,72 +2822,217 @@ void blk_mq_submit_bio(struct bio *bio) return; } - if (op_is_flush(bio->bi_opf) && blk_insert_flush(rq)) + if (op_is_flush(bio->bi_opf)) { + blk_insert_flush(rq); return; + } - if (plug && (q->nr_hw_queues == 1 || - blk_mq_is_shared_tags(rq->mq_hctx->flags) || - q->mq_ops->commit_rqs || !blk_queue_nonrot(q))) { - /* - * Use plugging if we have a ->commit_rqs() hook as well, as - * we know the driver uses bd->last in a smart fashion. - * - * Use normal plugging if this disk is slow HDD, as sequential - * IO may benefit a lot from plug merging. - */ - unsigned int request_count = plug->rq_count; - struct request *last = NULL; - - if (!request_count) { - trace_block_plug(q); - } else if (!blk_queue_nomerges(q)) { - last = rq_list_peek(&plug->mq_list); - if (blk_rq_bytes(last) < BLK_PLUG_FLUSH_SIZE) - last = NULL; - } - - if (request_count >= blk_plug_max_rq_count(plug) || last) { - blk_mq_flush_plug_list(plug, false); - trace_block_plug(q); - } - + if (plug) blk_add_rq_to_plug(plug, rq); - } else if (rq->rq_flags & RQF_ELV) { - /* Insert the request at the IO scheduler queue */ + else if ((rq->rq_flags & RQF_ELV) || + (rq->mq_hctx->dispatch_busy && + (q->nr_hw_queues == 1 || !is_sync))) blk_mq_sched_insert_request(rq, false, true, true); - } else if (plug && !blk_queue_nomerges(q)) { - struct request *next_rq = NULL; + else + blk_mq_run_dispatch_ops(rq->q, + blk_mq_try_issue_directly(rq->mq_hctx, rq)); +} + +/** + * blk_cloned_rq_check_limits - Helper function to check a cloned request + * for the new queue limits + * @q: the queue + * @rq: the request being checked + * + * Description: + * @rq may have been made based on weaker limitations of upper-level queues + * in request stacking drivers, and it may violate the limitation of @q. + * Since the block layer and the underlying device driver trust @rq + * after it is inserted to @q, it should be checked against @q before + * the insertion using this generic function. + * + * Request stacking drivers like request-based dm may change the queue + * limits when retrying requests on other queues. Those requests need + * to be checked against the new queue limits again during dispatch. + */ +static blk_status_t blk_cloned_rq_check_limits(struct request_queue *q, + struct request *rq) +{ + unsigned int max_sectors = blk_queue_get_max_sectors(q, req_op(rq)); + if (blk_rq_sectors(rq) > max_sectors) { /* - * We do limited plugging. If the bio can be merged, do that. - * Otherwise the existing request in the plug list will be - * issued. So the plug list will have one request at most - * The plug list might get flushed before this. If that happens, - * the plug list is empty, and same_queue_rq is invalid. + * SCSI device does not have a good way to return if + * Write Same/Zero is actually supported. If a device rejects + * a non-read/write command (discard, write same,etc.) the + * low-level device driver will set the relevant queue limit to + * 0 to prevent blk-lib from issuing more of the offending + * operations. Commands queued prior to the queue limit being + * reset need to be completed with BLK_STS_NOTSUPP to avoid I/O + * errors being propagated to upper layers. */ - if (same_queue_rq) { - next_rq = rq_list_pop(&plug->mq_list); - plug->rq_count--; - } - blk_add_rq_to_plug(plug, rq); - trace_block_plug(q); + if (max_sectors == 0) + return BLK_STS_NOTSUPP; + + printk(KERN_ERR "%s: over max size limit. (%u > %u)\n", + __func__, blk_rq_sectors(rq), max_sectors); + return BLK_STS_IOERR; + } + + /* + * The queue settings related to segment counting may differ from the + * original queue. + */ + rq->nr_phys_segments = blk_recalc_rq_segments(rq); + if (rq->nr_phys_segments > queue_max_segments(q)) { + printk(KERN_ERR "%s: over max segments limit. (%hu > %hu)\n", + __func__, rq->nr_phys_segments, queue_max_segments(q)); + return BLK_STS_IOERR; + } + + return BLK_STS_OK; +} + +/** + * blk_insert_cloned_request - Helper for stacking drivers to submit a request + * @q: the queue to submit the request + * @rq: the request being queued + */ +blk_status_t blk_insert_cloned_request(struct request_queue *q, struct request *rq) +{ + blk_status_t ret; + + ret = blk_cloned_rq_check_limits(q, rq); + if (ret != BLK_STS_OK) + return ret; + + if (rq->q->disk && + should_fail_request(rq->q->disk->part0, blk_rq_bytes(rq))) + return BLK_STS_IOERR; + + if (blk_crypto_insert_cloned_request(rq)) + return BLK_STS_IOERR; + + blk_account_io_start(rq); + + /* + * Since we have a scheduler attached on the top device, + * bypass a potential scheduler on the bottom device for + * insert. + */ + blk_mq_run_dispatch_ops(rq->q, + ret = blk_mq_request_issue_directly(rq, true)); + return ret; +} +EXPORT_SYMBOL_GPL(blk_insert_cloned_request); + +/** + * blk_rq_unprep_clone - Helper function to free all bios in a cloned request + * @rq: the clone request to be cleaned up + * + * Description: + * Free all bios in @rq for a cloned request. + */ +void blk_rq_unprep_clone(struct request *rq) +{ + struct bio *bio; + + while ((bio = rq->bio) != NULL) { + rq->bio = bio->bi_next; + + bio_put(bio); + } +} +EXPORT_SYMBOL_GPL(blk_rq_unprep_clone); + +/** + * blk_rq_prep_clone - Helper function to setup clone request + * @rq: the request to be setup + * @rq_src: original request to be cloned + * @bs: bio_set that bios for clone are allocated from + * @gfp_mask: memory allocation mask for bio + * @bio_ctr: setup function to be called for each clone bio. + * Returns %0 for success, non %0 for failure. + * @data: private data to be passed to @bio_ctr + * + * Description: + * Clones bios in @rq_src to @rq, and copies attributes of @rq_src to @rq. + * Also, pages which the original bios are pointing to are not copied + * and the cloned bios just point same pages. + * So cloned bios must be completed before original bios, which means + * the caller must complete @rq before @rq_src. + */ +int blk_rq_prep_clone(struct request *rq, struct request *rq_src, + struct bio_set *bs, gfp_t gfp_mask, + int (*bio_ctr)(struct bio *, struct bio *, void *), + void *data) +{ + struct bio *bio, *bio_src; + + if (!bs) + bs = &fs_bio_set; + + __rq_for_each_bio(bio_src, rq_src) { + bio = bio_clone_fast(bio_src, gfp_mask, bs); + if (!bio) + goto free_and_out; - if (next_rq) { - trace_block_unplug(q, 1, true); - blk_mq_try_issue_directly(next_rq->mq_hctx, next_rq); + if (bio_ctr && bio_ctr(bio, bio_src, data)) + goto free_and_out; + + if (rq->bio) { + rq->biotail->bi_next = bio; + rq->biotail = bio; + } else { + rq->bio = rq->biotail = bio; } - } else if ((q->nr_hw_queues > 1 && is_sync) || - !rq->mq_hctx->dispatch_busy) { - /* - * There is no scheduler and we can try to send directly - * to the hardware. - */ - blk_mq_try_issue_directly(rq->mq_hctx, rq); - } else { - /* Default case. */ - blk_mq_sched_insert_request(rq, false, true, true); + bio = NULL; + } + + /* Copy attributes of the original request to the clone request. */ + rq->__sector = blk_rq_pos(rq_src); + rq->__data_len = blk_rq_bytes(rq_src); + if (rq_src->rq_flags & RQF_SPECIAL_PAYLOAD) { + rq->rq_flags |= RQF_SPECIAL_PAYLOAD; + rq->special_vec = rq_src->special_vec; + } + rq->nr_phys_segments = rq_src->nr_phys_segments; + rq->ioprio = rq_src->ioprio; + + if (rq->bio && blk_crypto_rq_bio_prep(rq, rq->bio, gfp_mask) < 0) + goto free_and_out; + + return 0; + +free_and_out: + if (bio) + bio_put(bio); + blk_rq_unprep_clone(rq); + + return -ENOMEM; +} +EXPORT_SYMBOL_GPL(blk_rq_prep_clone); + +/* + * Steal bios from a request and add them to a bio list. + * The request must not have been partially completed before. + */ +void blk_steal_bios(struct bio_list *list, struct request *rq) +{ + if (rq->bio) { + if (list->tail) + list->tail->bi_next = rq->bio; + else + list->head = rq->bio; + list->tail = rq->biotail; + + rq->bio = NULL; + rq->biotail = NULL; } + + rq->__data_len = 0; } +EXPORT_SYMBOL_GPL(blk_steal_bios); static size_t order_to_size(unsigned int order) { @@ -2740,7 +3060,7 @@ static void blk_mq_clear_rq_mapping(struct blk_mq_tags *drv_tags, unsigned long rq_addr = (unsigned long)rq; if (rq_addr >= start && rq_addr < end) { - WARN_ON_ONCE(refcount_read(&rq->ref) != 0); + WARN_ON_ONCE(req_ref_read(rq) != 0); cmpxchg(&drv_tags->rqs[i], rq, NULL); } } @@ -3074,7 +3394,7 @@ static void blk_mq_clear_flush_rq_mapping(struct blk_mq_tags *tags, if (!tags) return; - WARN_ON_ONCE(refcount_read(&flush_rq->ref) != 0); + WARN_ON_ONCE(req_ref_read(flush_rq) != 0); for (i = 0; i < queue_depth; i++) cmpxchg(&tags->rqs[i], flush_rq, NULL); @@ -3128,20 +3448,6 @@ static void blk_mq_exit_hw_queues(struct request_queue *q, } } -static int blk_mq_hw_ctx_size(struct blk_mq_tag_set *tag_set) -{ - int hw_ctx_size = sizeof(struct blk_mq_hw_ctx); - - BUILD_BUG_ON(ALIGN(offsetof(struct blk_mq_hw_ctx, srcu), - __alignof__(struct blk_mq_hw_ctx)) != - sizeof(struct blk_mq_hw_ctx)); - - if (tag_set->flags & BLK_MQ_F_BLOCKING) - hw_ctx_size += sizeof(struct srcu_struct); - - return hw_ctx_size; -} - static int blk_mq_init_hctx(struct request_queue *q, struct blk_mq_tag_set *set, struct blk_mq_hw_ctx *hctx, unsigned hctx_idx) @@ -3179,7 +3485,7 @@ blk_mq_alloc_hctx(struct request_queue *q, struct blk_mq_tag_set *set, struct blk_mq_hw_ctx *hctx; gfp_t gfp = GFP_NOIO | __GFP_NOWARN | __GFP_NORETRY; - hctx = kzalloc_node(blk_mq_hw_ctx_size(set), gfp, node); + hctx = kzalloc_node(sizeof(struct blk_mq_hw_ctx), gfp, node); if (!hctx) goto fail_alloc_hctx; @@ -3221,8 +3527,6 @@ blk_mq_alloc_hctx(struct request_queue *q, struct blk_mq_tag_set *set, if (!hctx->fq) goto free_bitmap; - if (hctx->flags & BLK_MQ_F_BLOCKING) - init_srcu_struct(hctx->srcu); blk_mq_hctx_kobj_init(hctx); return hctx; @@ -3558,7 +3862,7 @@ static struct request_queue *blk_mq_init_queue_data(struct blk_mq_tag_set *set, struct request_queue *q; int ret; - q = blk_alloc_queue(set->numa_node); + q = blk_alloc_queue(set->numa_node, set->flags & BLK_MQ_F_BLOCKING); if (!q) return ERR_PTR(-ENOMEM); q->queuedata = queuedata; @@ -3707,6 +4011,9 @@ static void blk_mq_realloc_hw_ctxs(struct blk_mq_tag_set *set, int blk_mq_init_allocated_queue(struct blk_mq_tag_set *set, struct request_queue *q) { + WARN_ON_ONCE(blk_queue_has_srcu(q) != + !!(set->flags & BLK_MQ_F_BLOCKING)); + /* mark the queue as mq asap */ q->mq_ops = set->ops; @@ -4243,11 +4550,10 @@ EXPORT_SYMBOL_GPL(blk_mq_update_nr_hw_queues); /* Enable polling stats and return whether they were already enabled. */ static bool blk_poll_stats_enable(struct request_queue *q) { - if (test_bit(QUEUE_FLAG_POLL_STATS, &q->queue_flags) || - blk_queue_flag_test_and_set(QUEUE_FLAG_POLL_STATS, q)) + if (q->poll_stat) return true; - blk_stat_add_callback(q, q->poll_cb); - return false; + + return blk_stats_alloc_enable(q); } static void blk_mq_poll_stats_start(struct request_queue *q) @@ -4256,8 +4562,7 @@ static void blk_mq_poll_stats_start(struct request_queue *q) * We don't arm the callback if polling stats are not enabled or the * callback is already active. */ - if (!test_bit(QUEUE_FLAG_POLL_STATS, &q->queue_flags) || - blk_stat_is_active(q->poll_cb)) + if (!q->poll_stat || blk_stat_is_active(q->poll_cb)) return; blk_stat_activate_msecs(q->poll_cb, 100); @@ -4417,6 +4722,19 @@ unsigned int blk_mq_rq_cpu(struct request *rq) } EXPORT_SYMBOL(blk_mq_rq_cpu); +void blk_mq_cancel_work_sync(struct request_queue *q) +{ + if (queue_is_mq(q)) { + struct blk_mq_hw_ctx *hctx; + int i; + + cancel_delayed_work_sync(&q->requeue_work); + + queue_for_each_hw_ctx(q, hctx, i) + cancel_delayed_work_sync(&hctx->run_work); + } +} + static int __init blk_mq_init(void) { int i; diff --git a/block/blk-mq.h b/block/blk-mq.h index 8acfa650f575..948791ea2a3e 100644 --- a/block/blk-mq.h +++ b/block/blk-mq.h @@ -65,9 +65,6 @@ void blk_mq_request_bypass_insert(struct request *rq, bool at_head, bool run_queue); void blk_mq_insert_requests(struct blk_mq_hw_ctx *hctx, struct blk_mq_ctx *ctx, struct list_head *list); - -/* Used by blk_insert_cloned_request() to issue request directly */ -blk_status_t blk_mq_request_issue_directly(struct request *rq, bool last); void blk_mq_try_issue_list_directly(struct blk_mq_hw_ctx *hctx, struct list_head *list); @@ -128,6 +125,8 @@ extern void blk_mq_hctx_kobj_init(struct blk_mq_hw_ctx *hctx); void blk_mq_free_plug_rqs(struct blk_plug *plug); void blk_mq_flush_plug_list(struct blk_plug *plug, bool from_schedule); +void blk_mq_cancel_work_sync(struct request_queue *q); + void blk_mq_release(struct request_queue *q); static inline struct blk_mq_ctx *__blk_mq_get_ctx(struct request_queue *q, @@ -375,5 +374,24 @@ static inline bool hctx_may_queue(struct blk_mq_hw_ctx *hctx, return __blk_mq_active_requests(hctx) < depth; } +/* run the code block in @dispatch_ops with rcu/srcu read lock held */ +#define __blk_mq_run_dispatch_ops(q, check_sleep, dispatch_ops) \ +do { \ + if (!blk_queue_has_srcu(q)) { \ + rcu_read_lock(); \ + (dispatch_ops); \ + rcu_read_unlock(); \ + } else { \ + int srcu_idx; \ + \ + might_sleep_if(check_sleep); \ + srcu_idx = srcu_read_lock((q)->srcu); \ + (dispatch_ops); \ + srcu_read_unlock((q)->srcu, srcu_idx); \ + } \ +} while (0) + +#define blk_mq_run_dispatch_ops(q, dispatch_ops) \ + __blk_mq_run_dispatch_ops(q, true, dispatch_ops) \ #endif diff --git a/block/blk-stat.c b/block/blk-stat.c index ae3dd1fb8e61..2ea01b5c1aca 100644 --- a/block/blk-stat.c +++ b/block/blk-stat.c @@ -15,7 +15,7 @@ struct blk_queue_stats { struct list_head callbacks; spinlock_t lock; - bool enable_accounting; + int accounting; }; void blk_rq_stat_init(struct blk_rq_stat *stat) @@ -161,7 +161,7 @@ void blk_stat_remove_callback(struct request_queue *q, spin_lock_irqsave(&q->stats->lock, flags); list_del_rcu(&cb->list); - if (list_empty(&q->stats->callbacks) && !q->stats->enable_accounting) + if (list_empty(&q->stats->callbacks) && !q->stats->accounting) blk_queue_flag_clear(QUEUE_FLAG_STATS, q); spin_unlock_irqrestore(&q->stats->lock, flags); @@ -184,13 +184,24 @@ void blk_stat_free_callback(struct blk_stat_callback *cb) call_rcu(&cb->rcu, blk_stat_free_callback_rcu); } +void blk_stat_disable_accounting(struct request_queue *q) +{ + unsigned long flags; + + spin_lock_irqsave(&q->stats->lock, flags); + if (!--q->stats->accounting) + blk_queue_flag_clear(QUEUE_FLAG_STATS, q); + spin_unlock_irqrestore(&q->stats->lock, flags); +} +EXPORT_SYMBOL_GPL(blk_stat_disable_accounting); + void blk_stat_enable_accounting(struct request_queue *q) { unsigned long flags; spin_lock_irqsave(&q->stats->lock, flags); - q->stats->enable_accounting = true; - blk_queue_flag_set(QUEUE_FLAG_STATS, q); + if (!q->stats->accounting++) + blk_queue_flag_set(QUEUE_FLAG_STATS, q); spin_unlock_irqrestore(&q->stats->lock, flags); } EXPORT_SYMBOL_GPL(blk_stat_enable_accounting); @@ -205,7 +216,7 @@ struct blk_queue_stats *blk_alloc_queue_stats(void) INIT_LIST_HEAD(&stats->callbacks); spin_lock_init(&stats->lock); - stats->enable_accounting = false; + stats->accounting = 0; return stats; } @@ -219,3 +230,21 @@ void blk_free_queue_stats(struct blk_queue_stats *stats) kfree(stats); } + +bool blk_stats_alloc_enable(struct request_queue *q) +{ + struct blk_rq_stat *poll_stat; + + poll_stat = kcalloc(BLK_MQ_POLL_STATS_BKTS, sizeof(*poll_stat), + GFP_ATOMIC); + if (!poll_stat) + return false; + + if (cmpxchg(&q->poll_stat, NULL, poll_stat) != NULL) { + kfree(poll_stat); + return true; + } + + blk_stat_add_callback(q, q->poll_cb); + return false; +} diff --git a/block/blk-stat.h b/block/blk-stat.h index 17b47a86eefb..17e1eb4ec7e2 100644 --- a/block/blk-stat.h +++ b/block/blk-stat.h @@ -64,11 +64,13 @@ struct blk_stat_callback { struct blk_queue_stats *blk_alloc_queue_stats(void); void blk_free_queue_stats(struct blk_queue_stats *); +bool blk_stats_alloc_enable(struct request_queue *q); void blk_stat_add(struct request *rq, u64 now); /* record time/size info in request but not add a callback */ void blk_stat_enable_accounting(struct request_queue *q); +void blk_stat_disable_accounting(struct request_queue *q); /** * blk_stat_alloc_callback() - Allocate a block statistics callback. diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c index cef1f713370b..e20eadfcf5c8 100644 --- a/block/blk-sysfs.c +++ b/block/blk-sysfs.c @@ -16,6 +16,7 @@ #include "blk.h" #include "blk-mq.h" #include "blk-mq-debugfs.h" +#include "blk-mq-sched.h" #include "blk-wbt.h" #include "blk-throttle.h" @@ -734,7 +735,8 @@ static void blk_free_queue_rcu(struct rcu_head *rcu_head) { struct request_queue *q = container_of(rcu_head, struct request_queue, rcu_head); - kmem_cache_free(blk_requestq_cachep, q); + + kmem_cache_free(blk_get_queue_kmem_cache(blk_queue_has_srcu(q)), q); } /* Unconfigure the I/O scheduler and dissociate from the cgroup controller. */ @@ -747,7 +749,7 @@ static void blk_exit_queue(struct request_queue *q) */ if (q->elevator) { ioc_clear_queue(q); - __elevator_exit(q, q->elevator); + elevator_exit(q); } /* @@ -785,24 +787,15 @@ static void blk_release_queue(struct kobject *kobj) might_sleep(); - if (test_bit(QUEUE_FLAG_POLL_STATS, &q->queue_flags)) + if (q->poll_stat) blk_stat_remove_callback(q, q->poll_cb); blk_stat_free_callback(q->poll_cb); - blk_free_queue_stats(q->stats); - - if (queue_is_mq(q)) { - struct blk_mq_hw_ctx *hctx; - int i; - - cancel_delayed_work_sync(&q->requeue_work); - - queue_for_each_hw_ctx(q, hctx, i) - cancel_delayed_work_sync(&hctx->run_work); - } - blk_exit_queue(q); + blk_free_queue_stats(q->stats); + kfree(q->poll_stat); + blk_queue_free_zone_bitmaps(q); if (queue_is_mq(q)) diff --git a/block/blk-throttle.c b/block/blk-throttle.c index 39bb6e68a9a2..7c462c006b26 100644 --- a/block/blk-throttle.c +++ b/block/blk-throttle.c @@ -13,6 +13,7 @@ #include <linux/blk-cgroup.h> #include "blk.h" #include "blk-cgroup-rwstat.h" +#include "blk-stat.h" #include "blk-throttle.h" /* Max dispatch from a group in 1 round */ diff --git a/block/blk.h b/block/blk.h index b4fed2033e48..8bd43b3ad33d 100644 --- a/block/blk.h +++ b/block/blk.h @@ -2,15 +2,10 @@ #ifndef BLK_INTERNAL_H #define BLK_INTERNAL_H -#include <linux/idr.h> -#include <linux/blk-mq.h> -#include <linux/part_stat.h> #include <linux/blk-crypto.h> #include <linux/memblock.h> /* for max_pfn/max_low_pfn */ #include <xen/xen.h> #include "blk-crypto-internal.h" -#include "blk-mq.h" -#include "blk-mq-sched.h" struct elevator_type; @@ -32,15 +27,10 @@ struct blk_flush_queue { }; extern struct kmem_cache *blk_requestq_cachep; +extern struct kmem_cache *blk_requestq_srcu_cachep; extern struct kobj_type blk_queue_ktype; extern struct ida blk_queue_ida; -static inline struct blk_flush_queue * -blk_get_flush_queue(struct request_queue *q, struct blk_mq_ctx *ctx) -{ - return blk_mq_map_queue(q, REQ_OP_FLUSH, ctx)->fq; -} - static inline void __blk_get_queue(struct request_queue *q) { kobject_get(&q->kobj); @@ -250,16 +240,13 @@ static inline void blk_integrity_del(struct gendisk *disk) unsigned long blk_rq_timeout(unsigned long timeout); void blk_add_timer(struct request *req); -void blk_print_req_error(struct request *req, blk_status_t status); +const char *blk_status_to_str(blk_status_t status); bool blk_attempt_plug_merge(struct request_queue *q, struct bio *bio, - unsigned int nr_segs, bool *same_queue_rq); + unsigned int nr_segs); bool blk_bio_list_merge(struct request_queue *q, struct list_head *list, struct bio *bio, unsigned int nr_segs); -void __blk_account_io_start(struct request *req); -void __blk_account_io_done(struct request *req, u64 now); - /* * Plug flush limits */ @@ -271,23 +258,14 @@ void __blk_account_io_done(struct request *req, u64 now); */ #define ELV_ON_HASH(rq) ((rq)->rq_flags & RQF_HASHED) -bool blk_insert_flush(struct request *rq); +void blk_insert_flush(struct request *rq); int elevator_switch_mq(struct request_queue *q, struct elevator_type *new_e); -void __elevator_exit(struct request_queue *, struct elevator_queue *); +void elevator_exit(struct request_queue *q); int elv_register_queue(struct request_queue *q, bool uevent); void elv_unregister_queue(struct request_queue *q); -static inline void elevator_exit(struct request_queue *q, - struct elevator_queue *e) -{ - lockdep_assert_held(&q->sysfs_lock); - - blk_mq_sched_free_rqs(q); - __elevator_exit(q, e); -} - ssize_t part_size_show(struct device *dev, struct device_attribute *attr, char *buf); ssize_t part_stat_show(struct device *dev, struct device_attribute *attr, @@ -347,26 +325,10 @@ int blk_dev_init(void); */ static inline bool blk_do_io_stat(struct request *rq) { - return (rq->rq_flags & RQF_IO_STAT) && rq->rq_disk; -} - -static inline void blk_account_io_done(struct request *req, u64 now) -{ - /* - * Account IO completion. flush_rq isn't accounted as a - * normal IO on queueing nor completion. Accounting the - * containing request is enough. - */ - if (blk_do_io_stat(req) && req->part && - !(req->rq_flags & RQF_FLUSH_SEQ)) - __blk_account_io_done(req, now); + return (rq->rq_flags & RQF_IO_STAT) && rq->q->disk; } -static inline void blk_account_io_start(struct request *req) -{ - if (blk_do_io_stat(req)) - __blk_account_io_start(req); -} +void update_io_ticks(struct block_device *part, unsigned long now, bool end); static inline void req_set_nomerge(struct request_queue *q, struct request *req) { @@ -402,13 +364,15 @@ static inline unsigned int bio_aligned_discard_max_sectors( /* * Internal io_context interface */ -void get_io_context(struct io_context *ioc); -struct io_cq *ioc_lookup_icq(struct io_context *ioc, struct request_queue *q); -struct io_cq *ioc_create_icq(struct io_context *ioc, struct request_queue *q, - gfp_t gfp_mask); +struct io_cq *ioc_find_get_icq(struct request_queue *q); +struct io_cq *ioc_lookup_icq(struct request_queue *q); +#ifdef CONFIG_BLK_ICQ void ioc_clear_queue(struct request_queue *q); - -int create_task_io_context(struct task_struct *task, gfp_t gfp_mask, int node); +#else +static inline void ioc_clear_queue(struct request_queue *q) +{ +} +#endif /* CONFIG_BLK_ICQ */ #ifdef CONFIG_BLK_DEV_THROTTLING_LOW extern ssize_t blk_throtl_sample_time_show(struct request_queue *q, char *page); @@ -467,7 +431,15 @@ int bio_add_hw_page(struct request_queue *q, struct bio *bio, struct page *page, unsigned int len, unsigned int offset, unsigned int max_sectors, bool *same_page); -struct request_queue *blk_alloc_queue(int node_id); +static inline struct kmem_cache *blk_get_queue_kmem_cache(bool srcu) +{ + if (srcu) + return blk_requestq_srcu_cachep; + return blk_requestq_cachep; +} +struct request_queue *blk_alloc_queue(int node_id, bool alloc_srcu); + +int disk_scan_partitions(struct gendisk *disk, fmode_t mode); int disk_alloc_events(struct gendisk *disk); void disk_add_events(struct gendisk *disk); @@ -493,4 +465,45 @@ int disk_register_independent_access_ranges(struct gendisk *disk, struct blk_independent_access_ranges *new_iars); void disk_unregister_independent_access_ranges(struct gendisk *disk); +#ifdef CONFIG_FAIL_MAKE_REQUEST +bool should_fail_request(struct block_device *part, unsigned int bytes); +#else /* CONFIG_FAIL_MAKE_REQUEST */ +static inline bool should_fail_request(struct block_device *part, + unsigned int bytes) +{ + return false; +} +#endif /* CONFIG_FAIL_MAKE_REQUEST */ + +/* + * Optimized request reference counting. Ideally we'd make timeouts be more + * clever, as that's the only reason we need references at all... But until + * this happens, this is faster than using refcount_t. Also see: + * + * abc54d634334 ("io_uring: switch to atomic_t for io_kiocb reference count") + */ +#define req_ref_zero_or_close_to_overflow(req) \ + ((unsigned int) atomic_read(&(req->ref)) + 127u <= 127u) + +static inline bool req_ref_inc_not_zero(struct request *req) +{ + return atomic_inc_not_zero(&req->ref); +} + +static inline bool req_ref_put_and_test(struct request *req) +{ + WARN_ON_ONCE(req_ref_zero_or_close_to_overflow(req)); + return atomic_dec_and_test(&req->ref); +} + +static inline void req_ref_set(struct request *req, int value) +{ + atomic_set(&req->ref, value); +} + +static inline int req_ref_read(struct request *req) +{ + return atomic_read(&req->ref); +} + #endif /* BLK_INTERNAL_H */ diff --git a/block/bsg-lib.c b/block/bsg-lib.c index 10aa378702fa..acfe1357bf6c 100644 --- a/block/bsg-lib.c +++ b/block/bsg-lib.c @@ -92,7 +92,7 @@ static int bsg_transport_sg_io_fn(struct request_queue *q, struct sg_io_v4 *hdr, goto out_unmap_bidi_rq; bio = rq->bio; - blk_execute_rq(NULL, rq, !(hdr->flags & BSG_FLAG_Q_AT_TAIL)); + blk_execute_rq(rq, !(hdr->flags & BSG_FLAG_Q_AT_TAIL)); /* * The assignments below don't make much sense, but are kept for diff --git a/block/elevator.c b/block/elevator.c index 1f39f6e8ebb9..ec98aed39c4f 100644 --- a/block/elevator.c +++ b/block/elevator.c @@ -188,8 +188,10 @@ static void elevator_release(struct kobject *kobj) kfree(e); } -void __elevator_exit(struct request_queue *q, struct elevator_queue *e) +void elevator_exit(struct request_queue *q) { + struct elevator_queue *e = q->elevator; + mutex_lock(&e->sysfs_lock); blk_mq_exit_sched(q, e); mutex_unlock(&e->sysfs_lock); @@ -595,7 +597,8 @@ int elevator_switch_mq(struct request_queue *q, elv_unregister_queue(q); ioc_clear_queue(q); - elevator_exit(q, q->elevator); + blk_mq_sched_free_rqs(q); + elevator_exit(q); } ret = blk_mq_init_sched(q, new_e); @@ -605,7 +608,8 @@ int elevator_switch_mq(struct request_queue *q, if (new_e) { ret = elv_register_queue(q, true); if (ret) { - elevator_exit(q, q->elevator); + blk_mq_sched_free_rqs(q); + elevator_exit(q); goto out; } } @@ -694,12 +698,18 @@ void elevator_init_mq(struct request_queue *q) if (!e) return; + /* + * We are called before adding disk, when there isn't any FS I/O, + * so freezing queue plus canceling dispatch work is enough to + * drain any dispatch activities originated from passthrough + * requests, then no need to quiesce queue which may add long boot + * latency, especially when lots of disks are involved. + */ blk_mq_freeze_queue(q); - blk_mq_quiesce_queue(q); + blk_mq_cancel_work_sync(q); err = blk_mq_init_sched(q, e); - blk_mq_unquiesce_queue(q); blk_mq_unfreeze_queue(q); if (err) { diff --git a/block/fops.c b/block/fops.c index ad732a36f9b3..26bf15c770d2 100644 --- a/block/fops.c +++ b/block/fops.c @@ -15,6 +15,7 @@ #include <linux/falloc.h> #include <linux/suspend.h> #include <linux/fs.h> +#include <linux/module.h> #include "blk.h" static inline struct inode *bdev_file_inode(struct file *file) @@ -340,8 +341,7 @@ static ssize_t __blkdev_direct_IO_async(struct kiocb *iocb, } else { ret = bio_iov_iter_get_pages(bio, iter); if (unlikely(ret)) { - bio->bi_status = BLK_STS_IOERR; - bio_endio(bio); + bio_put(bio); return ret; } } @@ -566,21 +566,48 @@ static ssize_t blkdev_read_iter(struct kiocb *iocb, struct iov_iter *to) { struct block_device *bdev = iocb->ki_filp->private_data; loff_t size = bdev_nr_bytes(bdev); + size_t count = iov_iter_count(to); loff_t pos = iocb->ki_pos; size_t shorted = 0; - ssize_t ret; + ssize_t ret = 0; - if (unlikely(pos + iov_iter_count(to) > size)) { + if (unlikely(pos + count > size)) { if (pos >= size) return 0; size -= pos; - if (iov_iter_count(to) > size) { - shorted = iov_iter_count(to) - size; + if (count > size) { + shorted = count - size; iov_iter_truncate(to, size); } } - ret = generic_file_read_iter(iocb, to); + if (iocb->ki_flags & IOCB_DIRECT) { + struct address_space *mapping = iocb->ki_filp->f_mapping; + + if (iocb->ki_flags & IOCB_NOWAIT) { + if (filemap_range_needs_writeback(mapping, iocb->ki_pos, + iocb->ki_pos + count - 1)) + return -EAGAIN; + } else { + ret = filemap_write_and_wait_range(mapping, + iocb->ki_pos, + iocb->ki_pos + count - 1); + if (ret < 0) + return ret; + } + + file_accessed(iocb->ki_filp); + + ret = blkdev_direct_IO(iocb, to); + if (ret >= 0) { + iocb->ki_pos += ret; + count -= ret; + } + if (ret < 0 || !count) + return ret; + } + + ret = filemap_read(iocb, to, ret); if (unlikely(shorted)) iov_iter_reexpand(to, iov_iter_count(to) + shorted); diff --git a/block/genhd.c b/block/genhd.c index c5392cc24d37..626c8406f21a 100644 --- a/block/genhd.c +++ b/block/genhd.c @@ -25,8 +25,10 @@ #include <linux/log2.h> #include <linux/pm_runtime.h> #include <linux/badblocks.h> +#include <linux/part_stat.h> #include "blk.h" +#include "blk-mq-sched.h" #include "blk-rq-qos.h" static struct kobject *block_depr; @@ -372,17 +374,21 @@ void disk_uevent(struct gendisk *disk, enum kobject_action action) } EXPORT_SYMBOL_GPL(disk_uevent); -static void disk_scan_partitions(struct gendisk *disk) +int disk_scan_partitions(struct gendisk *disk, fmode_t mode) { struct block_device *bdev; - if (!get_capacity(disk) || !disk_part_scan_enabled(disk)) - return; + if (disk->flags & (GENHD_FL_NO_PART | GENHD_FL_HIDDEN)) + return -EINVAL; + if (disk->open_partitions) + return -EBUSY; set_bit(GD_NEED_PART_SCAN, &disk->state); - bdev = blkdev_get_by_dev(disk_devt(disk), FMODE_READ, NULL); - if (!IS_ERR(bdev)) - blkdev_put(bdev, FMODE_READ); + bdev = blkdev_get_by_dev(disk_devt(disk), mode, NULL); + if (IS_ERR(bdev)) + return PTR_ERR(bdev); + blkdev_put(bdev, mode); + return 0; } /** @@ -425,6 +431,8 @@ int __must_check device_add_disk(struct device *parent, struct gendisk *disk, DISK_MAX_PARTS); disk->minors = DISK_MAX_PARTS; } + if (disk->first_minor + disk->minors > MINORMASK + 1) + return -EINVAL; } else { if (WARN_ON(disk->minors)) return -EINVAL; @@ -434,13 +442,8 @@ int __must_check device_add_disk(struct device *parent, struct gendisk *disk, return ret; disk->major = BLOCK_EXT_MAJOR; disk->first_minor = ret; - disk->flags |= GENHD_FL_EXT_DEVT; } - ret = disk_alloc_events(disk); - if (ret) - goto out_free_ext_minor; - /* delay uevents, until we scanned partition table */ dev_set_uevent_suppress(ddev, 1); @@ -451,7 +454,12 @@ int __must_check device_add_disk(struct device *parent, struct gendisk *disk, ddev->devt = MKDEV(disk->major, disk->first_minor); ret = device_add(ddev); if (ret) - goto out_disk_release_events; + goto out_free_ext_minor; + + ret = disk_alloc_events(disk); + if (ret) + goto out_device_del; + if (!sysfs_deprecated) { ret = sysfs_create_link(block_depr, &ddev->kobj, kobject_name(&ddev->kobj)); @@ -490,14 +498,7 @@ int __must_check device_add_disk(struct device *parent, struct gendisk *disk, if (ret) goto out_put_slave_dir; - if (disk->flags & GENHD_FL_HIDDEN) { - /* - * Don't let hidden disks show up in /proc/partitions, - * and don't bother scanning for partitions either. - */ - disk->flags |= GENHD_FL_SUPPRESS_PARTITION_INFO; - disk->flags |= GENHD_FL_NO_PART_SCAN; - } else { + if (!(disk->flags & GENHD_FL_HIDDEN)) { ret = bdi_register(disk->bdi, "%u:%u", disk->major, disk->first_minor); if (ret) @@ -509,7 +510,8 @@ int __must_check device_add_disk(struct device *parent, struct gendisk *disk, goto out_unregister_bdi; bdev_add(disk->part0, ddev->devt); - disk_scan_partitions(disk); + if (get_capacity(disk)) + disk_scan_partitions(disk, FMODE_READ); /* * Announce the disk and partitions after all partitions are @@ -539,8 +541,6 @@ out_del_block_link: sysfs_remove_link(block_depr, dev_name(ddev)); out_device_del: device_del(ddev); -out_disk_release_events: - disk_release_events(disk); out_free_ext_minor: if (disk->major == BLOCK_EXT_MAJOR) blk_free_ext_minor(disk->first_minor); @@ -720,8 +720,7 @@ void __init printk_all_partitions(void) * Don't show empty devices or things that have been * suppressed */ - if (get_capacity(disk) == 0 || - (disk->flags & GENHD_FL_SUPPRESS_PARTITION_INFO)) + if (get_capacity(disk) == 0 || (disk->flags & GENHD_FL_HIDDEN)) continue; /* @@ -814,11 +813,7 @@ static int show_partition(struct seq_file *seqf, void *v) struct block_device *part; unsigned long idx; - /* Don't show non-partitionable removeable devices or empty devices */ - if (!get_capacity(sgp) || (!disk_max_parts(sgp) && - (sgp->flags & GENHD_FL_REMOVABLE))) - return 0; - if (sgp->flags & GENHD_FL_SUPPRESS_PARTITION_INFO) + if (!get_capacity(sgp) || (sgp->flags & GENHD_FL_HIDDEN)) return 0; rcu_read_lock(); @@ -874,7 +869,8 @@ static ssize_t disk_ext_range_show(struct device *dev, { struct gendisk *disk = dev_to_disk(dev); - return sprintf(buf, "%d\n", disk_max_parts(disk)); + return sprintf(buf, "%d\n", + (disk->flags & GENHD_FL_NO_PART) ? 1 : DISK_MAX_PARTS); } static ssize_t disk_removable_show(struct device *dev, @@ -1111,6 +1107,8 @@ static void disk_release(struct device *dev) might_sleep(); WARN_ON_ONCE(disk_live(disk)); + blk_mq_cancel_work_sync(disk->queue); + disk_release_events(disk); kfree(disk->random); xa_destroy(&disk->part_tbl); @@ -1341,7 +1339,7 @@ struct gendisk *__blk_alloc_disk(int node, struct lock_class_key *lkclass) struct request_queue *q; struct gendisk *disk; - q = blk_alloc_queue(node); + q = blk_alloc_queue(node, false); if (!q) return NULL; diff --git a/block/ioctl.c b/block/ioctl.c index 0a1d10ac2e1a..4a86340133e4 100644 --- a/block/ioctl.c +++ b/block/ioctl.c @@ -82,31 +82,6 @@ static int compat_blkpg_ioctl(struct block_device *bdev, } #endif -static int blkdev_reread_part(struct block_device *bdev, fmode_t mode) -{ - struct block_device *tmp; - - if (!disk_part_scan_enabled(bdev->bd_disk) || bdev_is_partition(bdev)) - return -EINVAL; - if (!capable(CAP_SYS_ADMIN)) - return -EACCES; - if (bdev->bd_disk->open_partitions) - return -EBUSY; - - /* - * Reopen the device to revalidate the driver state and force a - * partition rescan. - */ - mode &= ~FMODE_EXCL; - set_bit(GD_NEED_PART_SCAN, &bdev->bd_disk->state); - - tmp = blkdev_get_by_dev(bdev->bd_dev, mode, NULL); - if (IS_ERR(tmp)) - return PTR_ERR(tmp); - blkdev_put(tmp, mode); - return 0; -} - static int blk_ioctl_discard(struct block_device *bdev, fmode_t mode, unsigned long arg, unsigned long flags) { @@ -522,7 +497,11 @@ static int blkdev_common_ioctl(struct block_device *bdev, fmode_t mode, bdev->bd_disk->bdi->ra_pages = (arg * 512) / PAGE_SIZE; return 0; case BLKRRPART: - return blkdev_reread_part(bdev, mode); + if (!capable(CAP_SYS_ADMIN)) + return -EACCES; + if (bdev_is_partition(bdev)) + return -EINVAL; + return disk_scan_partitions(bdev->bd_disk, mode & ~FMODE_EXCL); case BLKTRACESTART: case BLKTRACESTOP: case BLKTRACETEARDOWN: diff --git a/block/ioprio.c b/block/ioprio.c index 0e4ff245f2bf..2fe068fcaad5 100644 --- a/block/ioprio.c +++ b/block/ioprio.c @@ -22,46 +22,14 @@ */ #include <linux/gfp.h> #include <linux/kernel.h> -#include <linux/export.h> #include <linux/ioprio.h> #include <linux/cred.h> #include <linux/blkdev.h> #include <linux/capability.h> -#include <linux/sched/user.h> -#include <linux/sched/task.h> #include <linux/syscalls.h> #include <linux/security.h> #include <linux/pid_namespace.h> -int set_task_ioprio(struct task_struct *task, int ioprio) -{ - int err; - struct io_context *ioc; - const struct cred *cred = current_cred(), *tcred; - - rcu_read_lock(); - tcred = __task_cred(task); - if (!uid_eq(tcred->uid, cred->euid) && - !uid_eq(tcred->uid, cred->uid) && !capable(CAP_SYS_NICE)) { - rcu_read_unlock(); - return -EPERM; - } - rcu_read_unlock(); - - err = security_task_setioprio(task, ioprio); - if (err) - return err; - - ioc = get_task_io_context(task, GFP_ATOMIC, NUMA_NO_NODE); - if (ioc) { - ioc->ioprio = ioprio; - put_io_context(ioc); - } - - return err; -} -EXPORT_SYMBOL_GPL(set_task_ioprio); - int ioprio_check_cap(int ioprio) { int class = IOPRIO_PRIO_CLASS(ioprio); @@ -69,7 +37,14 @@ int ioprio_check_cap(int ioprio) switch (class) { case IOPRIO_CLASS_RT: - if (!capable(CAP_SYS_NICE) && !capable(CAP_SYS_ADMIN)) + /* + * Originally this only checked for CAP_SYS_ADMIN, + * which was implicitly allowed for pid 0 by security + * modules such as SELinux. Make sure we check + * CAP_SYS_ADMIN first to avoid a denial/avc for + * possibly missing CAP_SYS_NICE permission. + */ + if (!capable(CAP_SYS_ADMIN) && !capable(CAP_SYS_NICE)) return -EPERM; fallthrough; /* rt has prio field too */ @@ -213,6 +188,7 @@ SYSCALL_DEFINE2(ioprio_get, int, which, int, who) pgrp = task_pgrp(current); else pgrp = find_vpid(who); + read_lock(&tasklist_lock); do_each_pid_thread(pgrp, PIDTYPE_PGID, p) { tmpio = get_task_ioprio(p); if (tmpio < 0) @@ -222,6 +198,8 @@ SYSCALL_DEFINE2(ioprio_get, int, which, int, who) else ret = ioprio_best(ret, tmpio); } while_each_pid_thread(pgrp, PIDTYPE_PGID, p); + read_unlock(&tasklist_lock); + break; case IOPRIO_WHO_USER: uid = make_kuid(current_user_ns(), who); diff --git a/block/kyber-iosched.c b/block/kyber-iosched.c index fdd74a4df56f..70ff2a599ef6 100644 --- a/block/kyber-iosched.c +++ b/block/kyber-iosched.c @@ -433,6 +433,7 @@ static void kyber_exit_sched(struct elevator_queue *e) int i; del_timer_sync(&kqd->timer); + blk_stat_disable_accounting(kqd->q); for (i = 0; i < KYBER_NUM_DOMAINS; i++) sbitmap_queue_free(&kqd->domain_tokens[i]); diff --git a/block/partitions/core.c b/block/partitions/core.c index 334b72ef1d73..c2a1635922b1 100644 --- a/block/partitions/core.c +++ b/block/partitions/core.c @@ -98,13 +98,12 @@ static void bdev_set_nr_sectors(struct block_device *bdev, sector_t sectors) static struct parsed_partitions *allocate_partitions(struct gendisk *hd) { struct parsed_partitions *state; - int nr; + int nr = DISK_MAX_PARTS; state = kzalloc(sizeof(*state), GFP_KERNEL); if (!state) return NULL; - nr = disk_max_parts(hd); state->parts = vzalloc(array_size(nr, sizeof(state->parts[0]))); if (!state->parts) { kfree(state); @@ -326,7 +325,7 @@ static struct block_device *add_partition(struct gendisk *disk, int partno, lockdep_assert_held(&disk->open_mutex); - if (partno >= disk_max_parts(disk)) + if (partno >= DISK_MAX_PARTS) return ERR_PTR(-EINVAL); /* @@ -527,18 +526,15 @@ out_unlock: static bool disk_unlock_native_capacity(struct gendisk *disk) { - const struct block_device_operations *bdops = disk->fops; - - if (bdops->unlock_native_capacity && - !(disk->flags & GENHD_FL_NATIVE_CAPACITY)) { - printk(KERN_CONT "enabling native capacity\n"); - bdops->unlock_native_capacity(disk); - disk->flags |= GENHD_FL_NATIVE_CAPACITY; - return true; - } else { + if (!disk->fops->unlock_native_capacity || + test_and_set_bit(GD_NATIVE_CAPACITY, &disk->state)) { printk(KERN_CONT "truncated\n"); return false; } + + printk(KERN_CONT "enabling native capacity\n"); + disk->fops->unlock_native_capacity(disk); + return true; } void blk_drop_partitions(struct gendisk *disk) @@ -607,7 +603,7 @@ static int blk_add_partitions(struct gendisk *disk) struct parsed_partitions *state; int ret = -EAGAIN, p; - if (!disk_part_scan_enabled(disk)) + if (disk->flags & GENHD_FL_NO_PART) return 0; state = check_partition(disk); @@ -690,7 +686,7 @@ rescan: * userspace for this particular setup. */ if (invalidate) { - if (disk_part_scan_enabled(disk) || + if (!(disk->flags & GENHD_FL_NO_PART) || !(disk->flags & GENHD_FL_REMOVABLE)) set_capacity(disk, 0); } |