| Message ID | 20240903081653.65613-3-songmuchun@bytedance.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | Fix some starvation problems in block layer | expand |
On Tue, Sep 03, 2024 at 04:16:52PM +0800, Muchun Song wrote: > Supposing the following scenario. > > CPU0 CPU1 > > blk_mq_insert_request() 1) store blk_mq_unquiesce_queue() > blk_mq_run_hw_queue() blk_queue_flag_clear(QUEUE_FLAG_QUIESCED) 3) store > if (blk_queue_quiesced()) 2) load blk_mq_run_hw_queues() > return blk_mq_run_hw_queue() > blk_mq_sched_dispatch_requests() if (!blk_mq_hctx_has_pending()) 4) load > return > > The full memory barrier should be inserted between 1) and 2), as well as > between 3) and 4) to make sure that either CPU0 sees QUEUE_FLAG_QUIESCED is > cleared or CPU1 sees dispatch list or setting of bitmap of software queue. > Otherwise, either CPU will not re-run the hardware queue causing starvation. > > So the first solution is to 1) add a pair of memory barrier to fix the > problem, another solution is to 2) use hctx->queue->queue_lock to synchronize > QUEUE_FLAG_QUIESCED. Here, we chose 2) to fix it since memory barrier is not > easy to be maintained. > > Fixes: f4560ffe8cec1 ("blk-mq: use QUEUE_FLAG_QUIESCED to quiesce queue") > Cc: stable@vger.kernel.org > Cc: Muchun Song <muchun.song@linux.dev> > Signed-off-by: Muchun Song <songmuchun@bytedance.com> Reviewed-by: Ming Lei <ming.lei@redhat.com> thanks, Ming
On 9/3/24 2:16 AM, Muchun Song wrote: > Supposing the following scenario. > > CPU0 CPU1 > > blk_mq_insert_request() 1) store blk_mq_unquiesce_queue() > blk_mq_run_hw_queue() blk_queue_flag_clear(QUEUE_FLAG_QUIESCED) 3) store > if (blk_queue_quiesced()) 2) load blk_mq_run_hw_queues() > return blk_mq_run_hw_queue() > blk_mq_sched_dispatch_requests() if (!blk_mq_hctx_has_pending()) 4) load > return > > The full memory barrier should be inserted between 1) and 2), as well as > between 3) and 4) to make sure that either CPU0 sees QUEUE_FLAG_QUIESCED is > cleared or CPU1 sees dispatch list or setting of bitmap of software queue. > Otherwise, either CPU will not re-run the hardware queue causing starvation. > > So the first solution is to 1) add a pair of memory barrier to fix the > problem, another solution is to 2) use hctx->queue->queue_lock to synchronize > QUEUE_FLAG_QUIESCED. Here, we chose 2) to fix it since memory barrier is not > easy to be maintained. Same comment here, 72-74 chars wide please. > diff --git a/block/blk-mq.c b/block/blk-mq.c > index b2d0f22de0c7f..ac39f2a346a52 100644 > --- a/block/blk-mq.c > +++ b/block/blk-mq.c > @@ -2202,6 +2202,24 @@ void blk_mq_delay_run_hw_queue(struct blk_mq_hw_ctx *hctx, unsigned long msecs) > } > EXPORT_SYMBOL(blk_mq_delay_run_hw_queue); > > +static inline bool blk_mq_hw_queue_need_run(struct blk_mq_hw_ctx *hctx) > +{ > + bool need_run; > + > + /* > + * When queue is quiesced, we may be switching io scheduler, or > + * updating nr_hw_queues, or other things, and we can't run queue > + * any more, even blk_mq_hctx_has_pending() can't be called safely. > + * > + * And queue will be rerun in blk_mq_unquiesce_queue() if it is > + * quiesced. > + */ > + __blk_mq_run_dispatch_ops(hctx->queue, false, > + need_run = !blk_queue_quiesced(hctx->queue) && > + blk_mq_hctx_has_pending(hctx)); > + return need_run; > +} This __blk_mq_run_dispatch_ops() is also way too wide, why didn't you just break it like where you copied it from? > + > /** > * blk_mq_run_hw_queue - Start to run a hardware queue. > * @hctx: Pointer to the hardware queue to run. > @@ -2222,20 +2240,23 @@ void blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async) > > might_sleep_if(!async && hctx->flags & BLK_MQ_F_BLOCKING); > > - /* > - * When queue is quiesced, we may be switching io scheduler, or > - * updating nr_hw_queues, or other things, and we can't run queue > - * any more, even __blk_mq_hctx_has_pending() can't be called safely. > - * > - * And queue will be rerun in blk_mq_unquiesce_queue() if it is > - * quiesced. > - */ > - __blk_mq_run_dispatch_ops(hctx->queue, false, > - need_run = !blk_queue_quiesced(hctx->queue) && > - blk_mq_hctx_has_pending(hctx)); > + need_run = blk_mq_hw_queue_need_run(hctx); > + if (!need_run) { > + unsigned long flags; > > - if (!need_run) > - return; > + /* > + * synchronize with blk_mq_unquiesce_queue(), becuase we check > + * if hw queue is quiesced locklessly above, we need the use > + * ->queue_lock to make sure we see the up-to-date status to > + * not miss rerunning the hw queue. > + */ > + spin_lock_irqsave(&hctx->queue->queue_lock, flags); > + need_run = blk_mq_hw_queue_need_run(hctx); > + spin_unlock_irqrestore(&hctx->queue->queue_lock, flags); > + > + if (!need_run) > + return; > + } Is this not solvable on the unquiesce side instead? It's rather a shame to add overhead to the fast path to avoid a race with something that's super unlikely, like quisce.
On Tue, Sep 10, 2024 at 07:22:16AM -0600, Jens Axboe wrote: > On 9/3/24 2:16 AM, Muchun Song wrote: > > Supposing the following scenario. > > > > CPU0 CPU1 > > > > blk_mq_insert_request() 1) store blk_mq_unquiesce_queue() > > blk_mq_run_hw_queue() blk_queue_flag_clear(QUEUE_FLAG_QUIESCED) 3) store > > if (blk_queue_quiesced()) 2) load blk_mq_run_hw_queues() > > return blk_mq_run_hw_queue() > > blk_mq_sched_dispatch_requests() if (!blk_mq_hctx_has_pending()) 4) load > > return > > > > The full memory barrier should be inserted between 1) and 2), as well as > > between 3) and 4) to make sure that either CPU0 sees QUEUE_FLAG_QUIESCED is > > cleared or CPU1 sees dispatch list or setting of bitmap of software queue. > > Otherwise, either CPU will not re-run the hardware queue causing starvation. > > > > So the first solution is to 1) add a pair of memory barrier to fix the > > problem, another solution is to 2) use hctx->queue->queue_lock to synchronize > > QUEUE_FLAG_QUIESCED. Here, we chose 2) to fix it since memory barrier is not > > easy to be maintained. > > Same comment here, 72-74 chars wide please. > > > diff --git a/block/blk-mq.c b/block/blk-mq.c > > index b2d0f22de0c7f..ac39f2a346a52 100644 > > --- a/block/blk-mq.c > > +++ b/block/blk-mq.c > > @@ -2202,6 +2202,24 @@ void blk_mq_delay_run_hw_queue(struct blk_mq_hw_ctx *hctx, unsigned long msecs) > > } > > EXPORT_SYMBOL(blk_mq_delay_run_hw_queue); > > > > +static inline bool blk_mq_hw_queue_need_run(struct blk_mq_hw_ctx *hctx) > > +{ > > + bool need_run; > > + > > + /* > > + * When queue is quiesced, we may be switching io scheduler, or > > + * updating nr_hw_queues, or other things, and we can't run queue > > + * any more, even blk_mq_hctx_has_pending() can't be called safely. > > + * > > + * And queue will be rerun in blk_mq_unquiesce_queue() if it is > > + * quiesced. > > + */ > > + __blk_mq_run_dispatch_ops(hctx->queue, false, > > + need_run = !blk_queue_quiesced(hctx->queue) && > > + blk_mq_hctx_has_pending(hctx)); > > + return need_run; > > +} > > This __blk_mq_run_dispatch_ops() is also way too wide, why didn't you > just break it like where you copied it from? > > > + > > /** > > * blk_mq_run_hw_queue - Start to run a hardware queue. > > * @hctx: Pointer to the hardware queue to run. > > @@ -2222,20 +2240,23 @@ void blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async) > > > > might_sleep_if(!async && hctx->flags & BLK_MQ_F_BLOCKING); > > > > - /* > > - * When queue is quiesced, we may be switching io scheduler, or > > - * updating nr_hw_queues, or other things, and we can't run queue > > - * any more, even __blk_mq_hctx_has_pending() can't be called safely. > > - * > > - * And queue will be rerun in blk_mq_unquiesce_queue() if it is > > - * quiesced. > > - */ > > - __blk_mq_run_dispatch_ops(hctx->queue, false, > > - need_run = !blk_queue_quiesced(hctx->queue) && > > - blk_mq_hctx_has_pending(hctx)); > > + need_run = blk_mq_hw_queue_need_run(hctx); > > + if (!need_run) { > > + unsigned long flags; > > > > - if (!need_run) > > - return; > > + /* > > + * synchronize with blk_mq_unquiesce_queue(), becuase we check > > + * if hw queue is quiesced locklessly above, we need the use > > + * ->queue_lock to make sure we see the up-to-date status to > > + * not miss rerunning the hw queue. > > + */ > > + spin_lock_irqsave(&hctx->queue->queue_lock, flags); > > + need_run = blk_mq_hw_queue_need_run(hctx); > > + spin_unlock_irqrestore(&hctx->queue->queue_lock, flags); > > + > > + if (!need_run) > > + return; > > + } > > Is this not solvable on the unquiesce side instead? It's rather a shame > to add overhead to the fast path to avoid a race with something that's > super unlikely, like quisce. Yeah, it can be solved by adding synchronize_rcu()/srcu() in unquiesce side, but SCSI may call it in non-sleepable context via scsi_internal_device_unblock_nowait(). Thanks, Ming
> On Sep 10, 2024, at 21:22, Jens Axboe <axboe@kernel.dk> wrote: > > On 9/3/24 2:16 AM, Muchun Song wrote: >> Supposing the following scenario. >> >> CPU0 CPU1 >> >> blk_mq_insert_request() 1) store blk_mq_unquiesce_queue() >> blk_mq_run_hw_queue() blk_queue_flag_clear(QUEUE_FLAG_QUIESCED) 3) store >> if (blk_queue_quiesced()) 2) load blk_mq_run_hw_queues() >> return blk_mq_run_hw_queue() >> blk_mq_sched_dispatch_requests() if (!blk_mq_hctx_has_pending()) 4) load >> return >> >> The full memory barrier should be inserted between 1) and 2), as well as >> between 3) and 4) to make sure that either CPU0 sees QUEUE_FLAG_QUIESCED is >> cleared or CPU1 sees dispatch list or setting of bitmap of software queue. >> Otherwise, either CPU will not re-run the hardware queue causing starvation. >> >> So the first solution is to 1) add a pair of memory barrier to fix the >> problem, another solution is to 2) use hctx->queue->queue_lock to synchronize >> QUEUE_FLAG_QUIESCED. Here, we chose 2) to fix it since memory barrier is not >> easy to be maintained. > > Same comment here, 72-74 chars wide please. OK. > >> diff --git a/block/blk-mq.c b/block/blk-mq.c >> index b2d0f22de0c7f..ac39f2a346a52 100644 >> --- a/block/blk-mq.c >> +++ b/block/blk-mq.c >> @@ -2202,6 +2202,24 @@ void blk_mq_delay_run_hw_queue(struct blk_mq_hw_ctx *hctx, unsigned long msecs) >> } >> EXPORT_SYMBOL(blk_mq_delay_run_hw_queue); >> >> +static inline bool blk_mq_hw_queue_need_run(struct blk_mq_hw_ctx *hctx) >> +{ >> + bool need_run; >> + >> + /* >> + * When queue is quiesced, we may be switching io scheduler, or >> + * updating nr_hw_queues, or other things, and we can't run queue >> + * any more, even blk_mq_hctx_has_pending() can't be called safely. >> + * >> + * And queue will be rerun in blk_mq_unquiesce_queue() if it is >> + * quiesced. >> + */ >> + __blk_mq_run_dispatch_ops(hctx->queue, false, >> + need_run = !blk_queue_quiesced(hctx->queue) && >> + blk_mq_hctx_has_pending(hctx)); >> + return need_run; >> +} > > This __blk_mq_run_dispatch_ops() is also way too wide, why didn't you > just break it like where you copied it from? I thought the rule allows max 80 chars pre line, so I adjusted the code to let them align with the above "(". Seems you prefer the previous way, I can keep it the same as before. Muchun, Thanks. > >> + >> /** >> * blk_mq_run_hw_queue - Start to run a hardware queue. >> * @hctx: Pointer to the hardware queue to run. >> @@ -2222,20 +2240,23 @@ void blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async) >> >> might_sleep_if(!async && hctx->flags & BLK_MQ_F_BLOCKING); >> >> - /* >> - * When queue is quiesced, we may be switching io scheduler, or >> - * updating nr_hw_queues, or other things, and we can't run queue >> - * any more, even __blk_mq_hctx_has_pending() can't be called safely. >> - * >> - * And queue will be rerun in blk_mq_unquiesce_queue() if it is >> - * quiesced. >> - */ >> - __blk_mq_run_dispatch_ops(hctx->queue, false, >> - need_run = !blk_queue_quiesced(hctx->queue) && >> - blk_mq_hctx_has_pending(hctx)); >> + need_run = blk_mq_hw_queue_need_run(hctx); >> + if (!need_run) { >> + unsigned long flags; >> >> - if (!need_run) >> - return; >> + /* >> + * synchronize with blk_mq_unquiesce_queue(), becuase we check >> + * if hw queue is quiesced locklessly above, we need the use >> + * ->queue_lock to make sure we see the up-to-date status to >> + * not miss rerunning the hw queue. >> + */ >> + spin_lock_irqsave(&hctx->queue->queue_lock, flags); >> + need_run = blk_mq_hw_queue_need_run(hctx); >> + spin_unlock_irqrestore(&hctx->queue->queue_lock, flags); >> + >> + if (!need_run) >> + return; >> + } > > Is this not solvable on the unquiesce side instead? It's rather a shame > to add overhead to the fast path to avoid a race with something that's > super unlikely, like quisce. > > -- > Jens Axboe
> On Sep 11, 2024, at 11:54, Ming Lei <ming.lei@redhat.com> wrote: > > On Tue, Sep 10, 2024 at 07:22:16AM -0600, Jens Axboe wrote: >> On 9/3/24 2:16 AM, Muchun Song wrote: >>> Supposing the following scenario. >>> >>> CPU0 CPU1 >>> >>> blk_mq_insert_request() 1) store blk_mq_unquiesce_queue() >>> blk_mq_run_hw_queue() blk_queue_flag_clear(QUEUE_FLAG_QUIESCED) 3) store >>> if (blk_queue_quiesced()) 2) load blk_mq_run_hw_queues() >>> return blk_mq_run_hw_queue() >>> blk_mq_sched_dispatch_requests() if (!blk_mq_hctx_has_pending()) 4) load >>> return >>> >>> The full memory barrier should be inserted between 1) and 2), as well as >>> between 3) and 4) to make sure that either CPU0 sees QUEUE_FLAG_QUIESCED is >>> cleared or CPU1 sees dispatch list or setting of bitmap of software queue. >>> Otherwise, either CPU will not re-run the hardware queue causing starvation. >>> >>> So the first solution is to 1) add a pair of memory barrier to fix the >>> problem, another solution is to 2) use hctx->queue->queue_lock to synchronize >>> QUEUE_FLAG_QUIESCED. Here, we chose 2) to fix it since memory barrier is not >>> easy to be maintained. >> >> Same comment here, 72-74 chars wide please. >> >>> diff --git a/block/blk-mq.c b/block/blk-mq.c >>> index b2d0f22de0c7f..ac39f2a346a52 100644 >>> --- a/block/blk-mq.c >>> +++ b/block/blk-mq.c >>> @@ -2202,6 +2202,24 @@ void blk_mq_delay_run_hw_queue(struct blk_mq_hw_ctx *hctx, unsigned long msecs) >>> } >>> EXPORT_SYMBOL(blk_mq_delay_run_hw_queue); >>> >>> +static inline bool blk_mq_hw_queue_need_run(struct blk_mq_hw_ctx *hctx) >>> +{ >>> + bool need_run; >>> + >>> + /* >>> + * When queue is quiesced, we may be switching io scheduler, or >>> + * updating nr_hw_queues, or other things, and we can't run queue >>> + * any more, even blk_mq_hctx_has_pending() can't be called safely. >>> + * >>> + * And queue will be rerun in blk_mq_unquiesce_queue() if it is >>> + * quiesced. >>> + */ >>> + __blk_mq_run_dispatch_ops(hctx->queue, false, >>> + need_run = !blk_queue_quiesced(hctx->queue) && >>> + blk_mq_hctx_has_pending(hctx)); >>> + return need_run; >>> +} >> >> This __blk_mq_run_dispatch_ops() is also way too wide, why didn't you >> just break it like where you copied it from? >> >>> + >>> /** >>> * blk_mq_run_hw_queue - Start to run a hardware queue. >>> * @hctx: Pointer to the hardware queue to run. >>> @@ -2222,20 +2240,23 @@ void blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async) >>> >>> might_sleep_if(!async && hctx->flags & BLK_MQ_F_BLOCKING); >>> >>> - /* >>> - * When queue is quiesced, we may be switching io scheduler, or >>> - * updating nr_hw_queues, or other things, and we can't run queue >>> - * any more, even __blk_mq_hctx_has_pending() can't be called safely. >>> - * >>> - * And queue will be rerun in blk_mq_unquiesce_queue() if it is >>> - * quiesced. >>> - */ >>> - __blk_mq_run_dispatch_ops(hctx->queue, false, >>> - need_run = !blk_queue_quiesced(hctx->queue) && >>> - blk_mq_hctx_has_pending(hctx)); >>> + need_run = blk_mq_hw_queue_need_run(hctx); >>> + if (!need_run) { >>> + unsigned long flags; >>> >>> - if (!need_run) >>> - return; >>> + /* >>> + * synchronize with blk_mq_unquiesce_queue(), becuase we check >>> + * if hw queue is quiesced locklessly above, we need the use >>> + * ->queue_lock to make sure we see the up-to-date status to >>> + * not miss rerunning the hw queue. >>> + */ >>> + spin_lock_irqsave(&hctx->queue->queue_lock, flags); >>> + need_run = blk_mq_hw_queue_need_run(hctx); >>> + spin_unlock_irqrestore(&hctx->queue->queue_lock, flags); >>> + >>> + if (!need_run) >>> + return; >>> + } >> >> Is this not solvable on the unquiesce side instead? It's rather a shame >> to add overhead to the fast path to avoid a race with something that's >> super unlikely, like quisce. > > Yeah, it can be solved by adding synchronize_rcu()/srcu() in unquiesce > side, but SCSI may call it in non-sleepable context via scsi_internal_device_unblock_nowait(). Another approach will be like the fix for BLK_MQ_S_STOPPED (in patch 3), we could add a pair of mb into blk_queue_quiesced() and blk_mq_unquiesce_queue(). In which case, the fix will not affect any fast path, only slow path need the barrier overhead. diff --git a/block/blk-mq.c b/block/blk-mq.c index b2d0f22de0c7f..45588ddb08d6b 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -264,6 +264,12 @@ void blk_mq_unquiesce_queue(struct request_queue *q) ; } else if (!--q->quiesce_depth) { blk_queue_flag_clear(QUEUE_FLAG_QUIESCED, q); + /* + * Pairs with the smp_mb() in blk_queue_quiesced() to order the + * clearing of QUEUE_FLAG_QUIESCED above and the checking of + * dispatch list in the subsequent routine. + */ + smp_mb__after_atomic(); run_queue = true; } spin_unlock_irqrestore(&q->queue_lock, flags); diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index b8196e219ac22..7a71462892b66 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -628,7 +628,25 @@ void blk_queue_flag_clear(unsigned int flag, struct request_queue *q); #define blk_noretry_request(rq) \ ((rq)->cmd_flags & (REQ_FAILFAST_DEV|REQ_FAILFAST_TRANSPORT| \ REQ_FAILFAST_DRIVER)) -#define blk_queue_quiesced(q) test_bit(QUEUE_FLAG_QUIESCED, &(q)->queue_flags) + +static inline bool blk_queue_quiesced(struct request_queue *q) +{ + /* Fast path: hardware queue is unquiesced most of the time. */ + if (likely(!test_bit(QUEUE_FLAG_QUIESCED, &q->queue_flags))) + return false; + + /* + * This barrier is used to order adding of dispatch list before and + * the test of QUEUE_FLAG_QUIESCED below. Pairs with the memory barrier + * in blk_mq_unquiesce_queue() so that dispatch code could either see + * QUEUE_FLAG_QUIESCED is cleared or dispatch list is not empty to + * avoid missing dispatching requests. + */ + smp_mb(); + + return test_bit(QUEUE_FLAG_QUIESCED, &q->queue_flags); +} + #define blk_queue_pm_only(q) atomic_read(&(q)->pm_only) #define blk_queue_registered(q) test_bit(QUEUE_FLAG_REGISTERED, &(q)->queue_flags) #define blk_queue_sq_sched(q) test_bit(QUEUE_FLAG_SQ_SCHED, &(q)->queue_flags) Muchun, Thanks. > > > Thanks, > Ming
> On Sep 11, 2024, at 11:59, Muchun Song <muchun.song@linux.dev> wrote: > > > >> On Sep 11, 2024, at 11:54, Ming Lei <ming.lei@redhat.com> wrote: >> >>> On Tue, Sep 10, 2024 at 07:22:16AM -0600, Jens Axboe wrote: >>> On 9/3/24 2:16 AM, Muchun Song wrote: >>>> Supposing the following scenario. >>>> >>>> CPU0 CPU1 >>>> >>>> blk_mq_insert_request() 1) store blk_mq_unquiesce_queue() >>>> blk_mq_run_hw_queue() blk_queue_flag_clear(QUEUE_FLAG_QUIESCED) 3) store >>>> if (blk_queue_quiesced()) 2) load blk_mq_run_hw_queues() >>>> return blk_mq_run_hw_queue() >>>> blk_mq_sched_dispatch_requests() if (!blk_mq_hctx_has_pending()) 4) load >>>> return >>>> >>>> The full memory barrier should be inserted between 1) and 2), as well as >>>> between 3) and 4) to make sure that either CPU0 sees QUEUE_FLAG_QUIESCED is >>>> cleared or CPU1 sees dispatch list or setting of bitmap of software queue. >>>> Otherwise, either CPU will not re-run the hardware queue causing starvation. >>>> >>>> So the first solution is to 1) add a pair of memory barrier to fix the >>>> problem, another solution is to 2) use hctx->queue->queue_lock to synchronize >>>> QUEUE_FLAG_QUIESCED. Here, we chose 2) to fix it since memory barrier is not >>>> easy to be maintained. >>> >>> Same comment here, 72-74 chars wide please. >>> >>>> diff --git a/block/blk-mq.c b/block/blk-mq.c >>>> index b2d0f22de0c7f..ac39f2a346a52 100644 >>>> --- a/block/blk-mq.c >>>> +++ b/block/blk-mq.c >>>> @@ -2202,6 +2202,24 @@ void blk_mq_delay_run_hw_queue(struct blk_mq_hw_ctx *hctx, unsigned long msecs) >>>> } >>>> EXPORT_SYMBOL(blk_mq_delay_run_hw_queue); >>>> >>>> +static inline bool blk_mq_hw_queue_need_run(struct blk_mq_hw_ctx *hctx) >>>> +{ >>>> + bool need_run; >>>> + >>>> + /* >>>> + * When queue is quiesced, we may be switching io scheduler, or >>>> + * updating nr_hw_queues, or other things, and we can't run queue >>>> + * any more, even blk_mq_hctx_has_pending() can't be called safely. >>>> + * >>>> + * And queue will be rerun in blk_mq_unquiesce_queue() if it is >>>> + * quiesced. >>>> + */ >>>> + __blk_mq_run_dispatch_ops(hctx->queue, false, >>>> + need_run = !blk_queue_quiesced(hctx->queue) && >>>> + blk_mq_hctx_has_pending(hctx)); >>>> + return need_run; >>>> +} >>> >>> This __blk_mq_run_dispatch_ops() is also way too wide, why didn't you >>> just break it like where you copied it from? >>> >>>> + >>>> /** >>>> * blk_mq_run_hw_queue - Start to run a hardware queue. >>>> * @hctx: Pointer to the hardware queue to run. >>>> @@ -2222,20 +2240,23 @@ void blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async) >>>> >>>> might_sleep_if(!async && hctx->flags & BLK_MQ_F_BLOCKING); >>>> >>>> - /* >>>> - * When queue is quiesced, we may be switching io scheduler, or >>>> - * updating nr_hw_queues, or other things, and we can't run queue >>>> - * any more, even __blk_mq_hctx_has_pending() can't be called safely. >>>> - * >>>> - * And queue will be rerun in blk_mq_unquiesce_queue() if it is >>>> - * quiesced. >>>> - */ >>>> - __blk_mq_run_dispatch_ops(hctx->queue, false, >>>> - need_run = !blk_queue_quiesced(hctx->queue) && >>>> - blk_mq_hctx_has_pending(hctx)); >>>> + need_run = blk_mq_hw_queue_need_run(hctx); >>>> + if (!need_run) { >>>> + unsigned long flags; >>>> >>>> - if (!need_run) >>>> - return; >>>> + /* >>>> + * synchronize with blk_mq_unquiesce_queue(), becuase we check >>>> + * if hw queue is quiesced locklessly above, we need the use >>>> + * ->queue_lock to make sure we see the up-to-date status to >>>> + * not miss rerunning the hw queue. >>>> + */ >>>> + spin_lock_irqsave(&hctx->queue->queue_lock, flags); >>>> + need_run = blk_mq_hw_queue_need_run(hctx); >>>> + spin_unlock_irqrestore(&hctx->queue->queue_lock, flags); >>>> + >>>> + if (!need_run) >>>> + return; >>>> + } >>> >>> Is this not solvable on the unquiesce side instead? It's rather a shame >>> to add overhead to the fast path to avoid a race with something that's >>> super unlikely, like quisce. >> >> Yeah, it can be solved by adding synchronize_rcu()/srcu() in unquiesce >> side, but SCSI may call it in non-sleepable context via scsi_internal_device_unblock_nowait(). > > Another approach will be like the fix for BLK_MQ_S_STOPPED (in patch 3), > we could add a pair of mb into blk_queue_quiesced() and > blk_mq_unquiesce_queue(). In which case, the fix will not affect any fast > path, only slow path need the barrier overhead. I misunderstood Jens’s question. I think Ming is right. This approach only tries to reduce the overhead as much as possible even for slow path compared to spinlock_based approach. Not solving the problem only from the unquiesce side. Muchun, Thanks. > > diff --git a/block/blk-mq.c b/block/blk-mq.c > index b2d0f22de0c7f..45588ddb08d6b 100644 > --- a/block/blk-mq.c > +++ b/block/blk-mq.c > @@ -264,6 +264,12 @@ void blk_mq_unquiesce_queue(struct request_queue *q) > ; > } else if (!--q->quiesce_depth) { > blk_queue_flag_clear(QUEUE_FLAG_QUIESCED, q); > + /* > + * Pairs with the smp_mb() in blk_queue_quiesced() to order the > + * clearing of QUEUE_FLAG_QUIESCED above and the checking of > + * dispatch list in the subsequent routine. > + */ > + smp_mb__after_atomic(); > run_queue = true; > } > spin_unlock_irqrestore(&q->queue_lock, flags); > diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h > index b8196e219ac22..7a71462892b66 100644 > --- a/include/linux/blkdev.h > +++ b/include/linux/blkdev.h > @@ -628,7 +628,25 @@ void blk_queue_flag_clear(unsigned int flag, struct request_queue *q); > #define blk_noretry_request(rq) \ > ((rq)->cmd_flags & (REQ_FAILFAST_DEV|REQ_FAILFAST_TRANSPORT| \ > REQ_FAILFAST_DRIVER)) > -#define blk_queue_quiesced(q) test_bit(QUEUE_FLAG_QUIESCED, &(q)->queue_flags) > + > +static inline bool blk_queue_quiesced(struct request_queue *q) > +{ > + /* Fast path: hardware queue is unquiesced most of the time. */ > + if (likely(!test_bit(QUEUE_FLAG_QUIESCED, &q->queue_flags))) > + return false; > + > + /* > + * This barrier is used to order adding of dispatch list before and > + * the test of QUEUE_FLAG_QUIESCED below. Pairs with the memory barrier > + * in blk_mq_unquiesce_queue() so that dispatch code could either see > + * QUEUE_FLAG_QUIESCED is cleared or dispatch list is not empty to > + * avoid missing dispatching requests. > + */ > + smp_mb(); > + > + return test_bit(QUEUE_FLAG_QUIESCED, &q->queue_flags); > +} > + > #define blk_queue_pm_only(q) atomic_read(&(q)->pm_only) > #define blk_queue_registered(q) test_bit(QUEUE_FLAG_REGISTERED, &(q)->queue_flags) > #define blk_queue_sq_sched(q) test_bit(QUEUE_FLAG_SQ_SCHED, &(q)->queue_flags) > > Muchun, > Thanks. > >> >> >> Thanks, >> Ming > >
> On Sep 11, 2024, at 11:54, Ming Lei <ming.lei@redhat.com> wrote: > > On Tue, Sep 10, 2024 at 07:22:16AM -0600, Jens Axboe wrote: >> On 9/3/24 2:16 AM, Muchun Song wrote: >>> Supposing the following scenario. >>> >>> CPU0 CPU1 >>> >>> blk_mq_insert_request() 1) store blk_mq_unquiesce_queue() >>> blk_mq_run_hw_queue() blk_queue_flag_clear(QUEUE_FLAG_QUIESCED) 3) store >>> if (blk_queue_quiesced()) 2) load blk_mq_run_hw_queues() >>> return blk_mq_run_hw_queue() >>> blk_mq_sched_dispatch_requests() if (!blk_mq_hctx_has_pending()) 4) load >>> return >>> >>> The full memory barrier should be inserted between 1) and 2), as well as >>> between 3) and 4) to make sure that either CPU0 sees QUEUE_FLAG_QUIESCED is >>> cleared or CPU1 sees dispatch list or setting of bitmap of software queue. >>> Otherwise, either CPU will not re-run the hardware queue causing starvation. >>> >>> So the first solution is to 1) add a pair of memory barrier to fix the >>> problem, another solution is to 2) use hctx->queue->queue_lock to synchronize >>> QUEUE_FLAG_QUIESCED. Here, we chose 2) to fix it since memory barrier is not >>> easy to be maintained. >> >> Same comment here, 72-74 chars wide please. >> >>> diff --git a/block/blk-mq.c b/block/blk-mq.c >>> index b2d0f22de0c7f..ac39f2a346a52 100644 >>> --- a/block/blk-mq.c >>> +++ b/block/blk-mq.c >>> @@ -2202,6 +2202,24 @@ void blk_mq_delay_run_hw_queue(struct blk_mq_hw_ctx *hctx, unsigned long msecs) >>> } >>> EXPORT_SYMBOL(blk_mq_delay_run_hw_queue); >>> >>> +static inline bool blk_mq_hw_queue_need_run(struct blk_mq_hw_ctx *hctx) >>> +{ >>> + bool need_run; >>> + >>> + /* >>> + * When queue is quiesced, we may be switching io scheduler, or >>> + * updating nr_hw_queues, or other things, and we can't run queue >>> + * any more, even blk_mq_hctx_has_pending() can't be called safely. >>> + * >>> + * And queue will be rerun in blk_mq_unquiesce_queue() if it is >>> + * quiesced. >>> + */ >>> + __blk_mq_run_dispatch_ops(hctx->queue, false, >>> + need_run = !blk_queue_quiesced(hctx->queue) && >>> + blk_mq_hctx_has_pending(hctx)); >>> + return need_run; >>> +} >> >> This __blk_mq_run_dispatch_ops() is also way too wide, why didn't you >> just break it like where you copied it from? >> >>> + >>> /** >>> * blk_mq_run_hw_queue - Start to run a hardware queue. >>> * @hctx: Pointer to the hardware queue to run. >>> @@ -2222,20 +2240,23 @@ void blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async) >>> >>> might_sleep_if(!async && hctx->flags & BLK_MQ_F_BLOCKING); >>> >>> - /* >>> - * When queue is quiesced, we may be switching io scheduler, or >>> - * updating nr_hw_queues, or other things, and we can't run queue >>> - * any more, even __blk_mq_hctx_has_pending() can't be called safely. >>> - * >>> - * And queue will be rerun in blk_mq_unquiesce_queue() if it is >>> - * quiesced. >>> - */ >>> - __blk_mq_run_dispatch_ops(hctx->queue, false, >>> - need_run = !blk_queue_quiesced(hctx->queue) && >>> - blk_mq_hctx_has_pending(hctx)); >>> + need_run = blk_mq_hw_queue_need_run(hctx); >>> + if (!need_run) { >>> + unsigned long flags; >>> >>> - if (!need_run) >>> - return; >>> + /* >>> + * synchronize with blk_mq_unquiesce_queue(), becuase we check >>> + * if hw queue is quiesced locklessly above, we need the use >>> + * ->queue_lock to make sure we see the up-to-date status to >>> + * not miss rerunning the hw queue. >>> + */ >>> + spin_lock_irqsave(&hctx->queue->queue_lock, flags); >>> + need_run = blk_mq_hw_queue_need_run(hctx); >>> + spin_unlock_irqrestore(&hctx->queue->queue_lock, flags); >>> + >>> + if (!need_run) >>> + return; >>> + } >> >> Is this not solvable on the unquiesce side instead? It's rather a shame >> to add overhead to the fast path to avoid a race with something that's >> super unlikely, like quisce. > > Yeah, it can be solved by adding synchronize_rcu()/srcu() in unquiesce > side, but SCSI may call it in non-sleepable context via scsi_internal_device_unblock_nowait(). Hi Ming and Jens, I use call_srcu/call_rcu to make it non-sleepable. Does this make sense to you? diff --git a/block/blk-mq.c b/block/blk-mq.c index 12bf38bec1044..86cdff28b2ce6 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -247,6 +247,13 @@ void blk_mq_quiesce_queue(struct request_queue *q) } EXPORT_SYMBOL_GPL(blk_mq_quiesce_queue); +static void blk_mq_run_hw_queues_rcu(struct rcu_head *rh) +{ + struct request_queue *q = container_of(rh, struct request_queue, + rcu_head); + blk_mq_run_hw_queues(q, true); +} + /* * blk_mq_unquiesce_queue() - counterpart of blk_mq_quiesce_queue() * @q: request queue. @@ -269,8 +276,13 @@ void blk_mq_unquiesce_queue(struct request_queue *q) spin_unlock_irqrestore(&q->queue_lock, flags); /* dispatch requests which are inserted during quiescing */ - if (run_queue) - blk_mq_run_hw_queues(q, true); + if (run_queue) { + if (q->tag_set->flags & BLK_MQ_F_BLOCKING) + call_srcu(q->tag_set->srcu, &q->rcu_head, + blk_mq_run_hw_queues_rcu); + else + call_rcu(&q->rcu_head, blk_mq_run_hw_queues_rcu); + } } EXPORT_SYMBOL_GPL(blk_mq_unquiesce_queue); > > > Thanks, > Ming
> On Sep 12, 2024, at 11:27, Muchun Song <muchun.song@linux.dev> wrote: > > > >> On Sep 11, 2024, at 11:54, Ming Lei <ming.lei@redhat.com> wrote: >> >> On Tue, Sep 10, 2024 at 07:22:16AM -0600, Jens Axboe wrote: >>> On 9/3/24 2:16 AM, Muchun Song wrote: >>>> Supposing the following scenario. >>>> >>>> CPU0 CPU1 >>>> >>>> blk_mq_insert_request() 1) store blk_mq_unquiesce_queue() >>>> blk_mq_run_hw_queue() blk_queue_flag_clear(QUEUE_FLAG_QUIESCED) 3) store >>>> if (blk_queue_quiesced()) 2) load blk_mq_run_hw_queues() >>>> return blk_mq_run_hw_queue() >>>> blk_mq_sched_dispatch_requests() if (!blk_mq_hctx_has_pending()) 4) load >>>> return >>>> >>>> The full memory barrier should be inserted between 1) and 2), as well as >>>> between 3) and 4) to make sure that either CPU0 sees QUEUE_FLAG_QUIESCED is >>>> cleared or CPU1 sees dispatch list or setting of bitmap of software queue. >>>> Otherwise, either CPU will not re-run the hardware queue causing starvation. >>>> >>>> So the first solution is to 1) add a pair of memory barrier to fix the >>>> problem, another solution is to 2) use hctx->queue->queue_lock to synchronize >>>> QUEUE_FLAG_QUIESCED. Here, we chose 2) to fix it since memory barrier is not >>>> easy to be maintained. >>> >>> Same comment here, 72-74 chars wide please. >>> >>>> diff --git a/block/blk-mq.c b/block/blk-mq.c >>>> index b2d0f22de0c7f..ac39f2a346a52 100644 >>>> --- a/block/blk-mq.c >>>> +++ b/block/blk-mq.c >>>> @@ -2202,6 +2202,24 @@ void blk_mq_delay_run_hw_queue(struct blk_mq_hw_ctx *hctx, unsigned long msecs) >>>> } >>>> EXPORT_SYMBOL(blk_mq_delay_run_hw_queue); >>>> >>>> +static inline bool blk_mq_hw_queue_need_run(struct blk_mq_hw_ctx *hctx) >>>> +{ >>>> + bool need_run; >>>> + >>>> + /* >>>> + * When queue is quiesced, we may be switching io scheduler, or >>>> + * updating nr_hw_queues, or other things, and we can't run queue >>>> + * any more, even blk_mq_hctx_has_pending() can't be called safely. >>>> + * >>>> + * And queue will be rerun in blk_mq_unquiesce_queue() if it is >>>> + * quiesced. >>>> + */ >>>> + __blk_mq_run_dispatch_ops(hctx->queue, false, >>>> + need_run = !blk_queue_quiesced(hctx->queue) && >>>> + blk_mq_hctx_has_pending(hctx)); >>>> + return need_run; >>>> +} >>> >>> This __blk_mq_run_dispatch_ops() is also way too wide, why didn't you >>> just break it like where you copied it from? >>> >>>> + >>>> /** >>>> * blk_mq_run_hw_queue - Start to run a hardware queue. >>>> * @hctx: Pointer to the hardware queue to run. >>>> @@ -2222,20 +2240,23 @@ void blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async) >>>> >>>> might_sleep_if(!async && hctx->flags & BLK_MQ_F_BLOCKING); >>>> >>>> - /* >>>> - * When queue is quiesced, we may be switching io scheduler, or >>>> - * updating nr_hw_queues, or other things, and we can't run queue >>>> - * any more, even __blk_mq_hctx_has_pending() can't be called safely. >>>> - * >>>> - * And queue will be rerun in blk_mq_unquiesce_queue() if it is >>>> - * quiesced. >>>> - */ >>>> - __blk_mq_run_dispatch_ops(hctx->queue, false, >>>> - need_run = !blk_queue_quiesced(hctx->queue) && >>>> - blk_mq_hctx_has_pending(hctx)); >>>> + need_run = blk_mq_hw_queue_need_run(hctx); >>>> + if (!need_run) { >>>> + unsigned long flags; >>>> >>>> - if (!need_run) >>>> - return; >>>> + /* >>>> + * synchronize with blk_mq_unquiesce_queue(), becuase we check >>>> + * if hw queue is quiesced locklessly above, we need the use >>>> + * ->queue_lock to make sure we see the up-to-date status to >>>> + * not miss rerunning the hw queue. >>>> + */ >>>> + spin_lock_irqsave(&hctx->queue->queue_lock, flags); >>>> + need_run = blk_mq_hw_queue_need_run(hctx); >>>> + spin_unlock_irqrestore(&hctx->queue->queue_lock, flags); >>>> + >>>> + if (!need_run) >>>> + return; >>>> + } >>> >>> Is this not solvable on the unquiesce side instead? It's rather a shame >>> to add overhead to the fast path to avoid a race with something that's >>> super unlikely, like quisce. >> >> Yeah, it can be solved by adding synchronize_rcu()/srcu() in unquiesce >> side, but SCSI may call it in non-sleepable context via scsi_internal_device_unblock_nowait(). > > Hi Ming and Jens, > > I use call_srcu/call_rcu to make it non-sleepable. Does this make sense to you? Sorry for the noise. call_srcu/call_rcu can't be easy to do this. Because call_srcu/call_rcu could be issued twice if users try to unquiesce the queue again before the callback of blk_mq_run_hw_queues_rcu has been executed. Thanks. > > diff --git a/block/blk-mq.c b/block/blk-mq.c > index 12bf38bec1044..86cdff28b2ce6 100644 > --- a/block/blk-mq.c > +++ b/block/blk-mq.c > @@ -247,6 +247,13 @@ void blk_mq_quiesce_queue(struct request_queue *q) > } > EXPORT_SYMBOL_GPL(blk_mq_quiesce_queue); > > +static void blk_mq_run_hw_queues_rcu(struct rcu_head *rh) > +{ > + struct request_queue *q = container_of(rh, struct request_queue, > + rcu_head); > + blk_mq_run_hw_queues(q, true); > +} > + > /* > * blk_mq_unquiesce_queue() - counterpart of blk_mq_quiesce_queue() > * @q: request queue. > @@ -269,8 +276,13 @@ void blk_mq_unquiesce_queue(struct request_queue *q) > spin_unlock_irqrestore(&q->queue_lock, flags); > > /* dispatch requests which are inserted during quiescing */ > - if (run_queue) > - blk_mq_run_hw_queues(q, true); > + if (run_queue) { > + if (q->tag_set->flags & BLK_MQ_F_BLOCKING) > + call_srcu(q->tag_set->srcu, &q->rcu_head, > + blk_mq_run_hw_queues_rcu); > + else > + call_rcu(&q->rcu_head, blk_mq_run_hw_queues_rcu); > + } > } > EXPORT_SYMBOL_GPL(blk_mq_unquiesce_queue); > >> >> >> Thanks, >> Ming
diff --git a/block/blk-mq.c b/block/blk-mq.c index b2d0f22de0c7f..ac39f2a346a52 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -2202,6 +2202,24 @@ void blk_mq_delay_run_hw_queue(struct blk_mq_hw_ctx *hctx, unsigned long msecs) } EXPORT_SYMBOL(blk_mq_delay_run_hw_queue); +static inline bool blk_mq_hw_queue_need_run(struct blk_mq_hw_ctx *hctx) +{ + bool need_run; + + /* + * When queue is quiesced, we may be switching io scheduler, or + * updating nr_hw_queues, or other things, and we can't run queue + * any more, even blk_mq_hctx_has_pending() can't be called safely. + * + * And queue will be rerun in blk_mq_unquiesce_queue() if it is + * quiesced. + */ + __blk_mq_run_dispatch_ops(hctx->queue, false, + need_run = !blk_queue_quiesced(hctx->queue) && + blk_mq_hctx_has_pending(hctx)); + return need_run; +} + /** * blk_mq_run_hw_queue - Start to run a hardware queue. * @hctx: Pointer to the hardware queue to run. @@ -2222,20 +2240,23 @@ void blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx, bool async) might_sleep_if(!async && hctx->flags & BLK_MQ_F_BLOCKING); - /* - * When queue is quiesced, we may be switching io scheduler, or - * updating nr_hw_queues, or other things, and we can't run queue - * any more, even __blk_mq_hctx_has_pending() can't be called safely. - * - * And queue will be rerun in blk_mq_unquiesce_queue() if it is - * quiesced. - */ - __blk_mq_run_dispatch_ops(hctx->queue, false, - need_run = !blk_queue_quiesced(hctx->queue) && - blk_mq_hctx_has_pending(hctx)); + need_run = blk_mq_hw_queue_need_run(hctx); + if (!need_run) { + unsigned long flags; - if (!need_run) - return; + /* + * Synchronize with blk_mq_unquiesce_queue(), becuase we check + * if hw queue is quiesced locklessly above, we need the use + * ->queue_lock to make sure we see the up-to-date status to + * not miss rerunning the hw queue. + */ + spin_lock_irqsave(&hctx->queue->queue_lock, flags); + need_run = blk_mq_hw_queue_need_run(hctx); + spin_unlock_irqrestore(&hctx->queue->queue_lock, flags); + + if (!need_run) + return; + } if (async || !cpumask_test_cpu(raw_smp_processor_id(), hctx->cpumask)) { blk_mq_delay_run_hw_queue(hctx, 0);
Supposing the following scenario. CPU0 CPU1 blk_mq_insert_request() 1) store blk_mq_unquiesce_queue() blk_mq_run_hw_queue() blk_queue_flag_clear(QUEUE_FLAG_QUIESCED) 3) store if (blk_queue_quiesced()) 2) load blk_mq_run_hw_queues() return blk_mq_run_hw_queue() blk_mq_sched_dispatch_requests() if (!blk_mq_hctx_has_pending()) 4) load return The full memory barrier should be inserted between 1) and 2), as well as between 3) and 4) to make sure that either CPU0 sees QUEUE_FLAG_QUIESCED is cleared or CPU1 sees dispatch list or setting of bitmap of software queue. Otherwise, either CPU will not re-run the hardware queue causing starvation. So the first solution is to 1) add a pair of memory barrier to fix the problem, another solution is to 2) use hctx->queue->queue_lock to synchronize QUEUE_FLAG_QUIESCED. Here, we chose 2) to fix it since memory barrier is not easy to be maintained. Fixes: f4560ffe8cec1 ("blk-mq: use QUEUE_FLAG_QUIESCED to quiesce queue") Cc: stable@vger.kernel.org Cc: Muchun Song <muchun.song@linux.dev> Signed-off-by: Muchun Song <songmuchun@bytedance.com> --- block/blk-mq.c | 47 ++++++++++++++++++++++++++++++++++------------- 1 file changed, 34 insertions(+), 13 deletions(-)