From a347c7ad8edf4c5685154f3fdc3c12fc1db800ba Mon Sep 17 00:00:00 2001 From: Roman Pen Date: Sun, 10 Jun 2018 22:38:24 +0200 Subject: blk-mq: reinit q->tag_set_list entry only after grace period It is not allowed to reinit q->tag_set_list list entry while RCU grace period has not completed yet, otherwise the following soft lockup in blk_mq_sched_restart() happens: [ 1064.252652] watchdog: BUG: soft lockup - CPU#12 stuck for 23s! [fio:9270] [ 1064.254445] task: ffff99b912e8b900 task.stack: ffffa6d54c758000 [ 1064.254613] RIP: 0010:blk_mq_sched_restart+0x96/0x150 [ 1064.256510] Call Trace: [ 1064.256664] [ 1064.256824] blk_mq_free_request+0xea/0x100 [ 1064.256987] msg_io_conf+0x59/0xd0 [ibnbd_client] [ 1064.257175] complete_rdma_req+0xf2/0x230 [ibtrs_client] [ 1064.257340] ? ibtrs_post_recv_empty+0x4d/0x70 [ibtrs_core] [ 1064.257502] ibtrs_clt_rdma_done+0xd1/0x1e0 [ibtrs_client] [ 1064.257669] ib_create_qp+0x321/0x380 [ib_core] [ 1064.257841] ib_process_cq_direct+0xbd/0x120 [ib_core] [ 1064.258007] irq_poll_softirq+0xb7/0xe0 [ 1064.258165] __do_softirq+0x106/0x2a2 [ 1064.258328] irq_exit+0x92/0xa0 [ 1064.258509] do_IRQ+0x4a/0xd0 [ 1064.258660] common_interrupt+0x7a/0x7a [ 1064.258818] Meanwhile another context frees other queue but with the same set of shared tags: [ 1288.201183] INFO: task bash:5910 blocked for more than 180 seconds. [ 1288.201833] bash D 0 5910 5820 0x00000000 [ 1288.202016] Call Trace: [ 1288.202315] schedule+0x32/0x80 [ 1288.202462] schedule_timeout+0x1e5/0x380 [ 1288.203838] wait_for_completion+0xb0/0x120 [ 1288.204137] __wait_rcu_gp+0x125/0x160 [ 1288.204287] synchronize_sched+0x6e/0x80 [ 1288.204770] blk_mq_free_queue+0x74/0xe0 [ 1288.204922] blk_cleanup_queue+0xc7/0x110 [ 1288.205073] ibnbd_clt_unmap_device+0x1bc/0x280 [ibnbd_client] [ 1288.205389] ibnbd_clt_unmap_dev_store+0x169/0x1f0 [ibnbd_client] [ 1288.205548] kernfs_fop_write+0x109/0x180 [ 1288.206328] vfs_write+0xb3/0x1a0 [ 1288.206476] SyS_write+0x52/0xc0 [ 1288.206624] do_syscall_64+0x68/0x1d0 [ 1288.206774] entry_SYSCALL_64_after_hwframe+0x3d/0xa2 What happened is the following: 1. There are several MQ queues with shared tags. 2. One queue is about to be freed and now task is in blk_mq_del_queue_tag_set(). 3. Other CPU is in blk_mq_sched_restart() and loops over all queues in tag list in order to find hctx to restart. Because linked list entry was modified in blk_mq_del_queue_tag_set() without proper waiting for a grace period, blk_mq_sched_restart() never ends, spining in list_for_each_entry_rcu_rr(), thus soft lockup. Fix is simple: reinit list entry after an RCU grace period elapsed. Fixes: Fixes: 705cda97ee3a ("blk-mq: Make it safe to use RCU to iterate over blk_mq_tag_set.tag_list") Cc: stable@vger.kernel.org Cc: Sagi Grimberg Cc: linux-block@vger.kernel.org Reviewed-by: Christoph Hellwig Reviewed-by: Ming Lei Reviewed-by: Bart Van Assche Signed-off-by: Roman Pen Signed-off-by: Jens Axboe --- block/blk-mq.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'block/blk-mq.c') diff --git a/block/blk-mq.c b/block/blk-mq.c index d2de0a719ab8..2be78cc30ec5 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -2349,7 +2349,6 @@ static void blk_mq_del_queue_tag_set(struct request_queue *q) mutex_lock(&set->tag_list_lock); list_del_rcu(&q->tag_set_list); - INIT_LIST_HEAD(&q->tag_set_list); if (list_is_singular(&set->tag_list)) { /* just transitioned to unshared */ set->flags &= ~BLK_MQ_F_TAG_SHARED; @@ -2357,8 +2356,8 @@ static void blk_mq_del_queue_tag_set(struct request_queue *q) blk_mq_update_tag_set_depth(set, false); } mutex_unlock(&set->tag_list_lock); - synchronize_rcu(); + INIT_LIST_HEAD(&q->tag_set_list); } static void blk_mq_add_queue_tag_set(struct blk_mq_tag_set *set, -- cgit v1.2.3-70-g09d2 From da661267398869a553b7f67d739d360aaa1361b6 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 14 Jun 2018 13:58:45 +0200 Subject: blk-mq: don't time out requests again that are in the timeout handler We can currently call the timeout handler again on a request that has already been handed over to the timeout handler. Prevent that with a new flag. Fixes: 12f5b931 ("blk-mq: Remove generation seqeunce") Reported-by: Andrew Randrianasulu Tested-by: Andrew Randrianasulu Signed-off-by: Christoph Hellwig Signed-off-by: Jens Axboe --- block/blk-mq.c | 5 +++++ include/linux/blkdev.h | 2 ++ 2 files changed, 7 insertions(+) (limited to 'block/blk-mq.c') diff --git a/block/blk-mq.c b/block/blk-mq.c index 2be78cc30ec5..8e57b84e50e9 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -671,6 +671,7 @@ static void __blk_mq_requeue_request(struct request *rq) if (blk_mq_request_started(rq)) { WRITE_ONCE(rq->state, MQ_RQ_IDLE); + rq->rq_flags &= ~RQF_TIMED_OUT; if (q->dma_drain_size && blk_rq_bytes(rq)) rq->nr_phys_segments--; } @@ -770,6 +771,7 @@ EXPORT_SYMBOL(blk_mq_tag_to_rq); static void blk_mq_rq_timed_out(struct request *req, bool reserved) { + req->rq_flags |= RQF_TIMED_OUT; if (req->q->mq_ops->timeout) { enum blk_eh_timer_return ret; @@ -779,6 +781,7 @@ static void blk_mq_rq_timed_out(struct request *req, bool reserved) WARN_ON_ONCE(ret != BLK_EH_RESET_TIMER); } + req->rq_flags &= ~RQF_TIMED_OUT; blk_add_timer(req); } @@ -788,6 +791,8 @@ static bool blk_mq_req_expired(struct request *rq, unsigned long *next) if (blk_mq_rq_state(rq) != MQ_RQ_IN_FLIGHT) return false; + if (rq->rq_flags & RQF_TIMED_OUT) + return false; deadline = blk_rq_deadline(rq); if (time_after_eq(jiffies, deadline)) diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index bca3a92eb55f..fa6f11751430 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -127,6 +127,8 @@ typedef __u32 __bitwise req_flags_t; #define RQF_ZONE_WRITE_LOCKED ((__force req_flags_t)(1 << 19)) /* already slept for hybrid poll */ #define RQF_MQ_POLL_SLEPT ((__force req_flags_t)(1 << 20)) +/* ->timeout has been called, don't expire again */ +#define RQF_TIMED_OUT ((__force req_flags_t)(1 << 21)) /* flags that prevent us from merging requests: */ #define RQF_NOMERGE_FLAGS \ -- cgit v1.2.3-70-g09d2