Message-ID: <20180529185524.GA23487@vader>
Date:   Tue, 29 May 2018 11:55:24 -0700
From:   Omar Sandoval <osandov@...ndov.com>
To:     Jianchao Wang <jianchao.w.wang@...cle.com>
Cc:     axboe@...nel.dk, holger@...lied-asynchrony.com,
        linux-block@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH V2 2/2] block: kyber: make kyber more friendly with
 merging

On Wed, May 23, 2018 at 02:33:22PM +0800, Jianchao Wang wrote:
> Currently, kyber is very unfriendly with merging. kyber depends
> on ctx rq_list to do merging, however, most of time, it will not
> leave any requests in ctx rq_list. This is because even if tokens
> of one domain is used up, kyber will try to dispatch requests
> from other domain and flush the rq_list there.
> 
> To improve this, we setup kyber_ctx_queue (kcq) which is similar
> with ctx, but it has rq_lists for different domain and build same
> mapping between kcq and khd as the ctx & hctx. Then we could merge,
> insert and dispatch for different domains separately. At the same
> time, only flush the rq_list of kcq when get domain token successfully.
> Then if one domain token is used up, the requests could be left in
> the rq_list of that domain and maybe merged with following io.
> 
> Following is my test result on machine with 8 cores and NVMe card
> INTEL SSDPEKKR128G7
> 
> fio size=256m ioengine=libaio iodepth=64 direct=1 numjobs=8
> seq/random
> +------+---------------------------------------------------------------+
> |patch?| bw(MB/s) |   iops    | slat(usec) |    clat(usec)   |  merge  |
> +----------------------------------------------------------------------+
> | w/o  |  606/612 | 151k/153k |  6.89/7.03 | 3349.21/3305.40 |   0/0   |
> +----------------------------------------------------------------------+
> | w/   | 1083/616 | 277k/154k |  4.93/6.95 | 1830.62/3279.95 | 223k/3k |
> +----------------------------------------------------------------------+
> When set numjobs to 16, the bw and iops could reach 1662MB/s and 425k
> on my platform.
> 
> Signed-off-by: Jianchao Wang <jianchao.w.wang@...cle.com>
> Tested-by: Holger Hoffstätte <holger@...lied-asynchrony.com>

Thanks, the per-hctx kqcs make much more sense. This is pretty cool! A
few nitpicks below.

> ---
>  block/kyber-iosched.c | 197 +++++++++++++++++++++++++++++++++++++++++---------
>  1 file changed, 162 insertions(+), 35 deletions(-)
> 
> diff --git a/block/kyber-iosched.c b/block/kyber-iosched.c
> index 0d6d25e3..7ca1364 100644
> --- a/block/kyber-iosched.c
> +++ b/block/kyber-iosched.c
> @@ -72,6 +72,15 @@ static const unsigned int kyber_batch_size[] = {
>  	[KYBER_OTHER] = 8,
>  };
>  
> +struct kyber_ctx_queue {

I'm not crazy about this name, but I can't really think of anything
better, so this will have to do.

> +	/*
> +	 * Copied from blk_mq_ctx->index_hw
> +	 */
> +	unsigned int index;

We can just calculate this as kcq - khd->kcqs in the one place we use
it.

> +	spinlock_t lock;
> +	struct list_head rq_list[KYBER_NUM_DOMAINS];
> +} ____cacheline_aligned_in_smp;
> +
>  struct kyber_queue_data {
>  	struct request_queue *q;
>  
> @@ -99,6 +108,8 @@ struct kyber_hctx_data {
>  	struct list_head rqs[KYBER_NUM_DOMAINS];
>  	unsigned int cur_domain;
>  	unsigned int batching;
> +	struct kyber_ctx_queue *kcqs;
> +	struct sbitmap kcq_map[KYBER_NUM_DOMAINS];
>  	wait_queue_entry_t domain_wait[KYBER_NUM_DOMAINS];
>  	struct sbq_wait_state *domain_ws[KYBER_NUM_DOMAINS];
>  	atomic_t wait_index[KYBER_NUM_DOMAINS];
> @@ -107,10 +118,8 @@ struct kyber_hctx_data {
>  static int kyber_domain_wake(wait_queue_entry_t *wait, unsigned mode, int flags,
>  			     void *key);
>  
> -static int rq_sched_domain(const struct request *rq)
> +static int kyber_sched_domain(unsigned int op)

Please make this return an unsigned int since that's what we use for
sched_domain everywhere except the stats bucket function.

>  {
> -	unsigned int op = rq->cmd_flags;
> -
>  	if ((op & REQ_OP_MASK) == REQ_OP_READ)
>  		return KYBER_READ;
>  	else if ((op & REQ_OP_MASK) == REQ_OP_WRITE && op_is_sync(op))
> @@ -284,6 +293,11 @@ static unsigned int kyber_sched_tags_shift(struct kyber_queue_data *kqd)
>  	return kqd->q->queue_hw_ctx[0]->sched_tags->bitmap_tags.sb.shift;
>  }
>  
> +static int kyber_bucket_fn(const struct request *rq)
> +{
> +	return kyber_sched_domain(rq->cmd_flags);
> +}
> +
>  static struct kyber_queue_data *kyber_queue_data_alloc(struct request_queue *q)
>  {
>  	struct kyber_queue_data *kqd;
> @@ -297,11 +311,10 @@ static struct kyber_queue_data *kyber_queue_data_alloc(struct request_queue *q)
>  		goto err;
>  	kqd->q = q;
>  
> -	kqd->cb = blk_stat_alloc_callback(kyber_stat_timer_fn, rq_sched_domain,
> +	kqd->cb = blk_stat_alloc_callback(kyber_stat_timer_fn, kyber_bucket_fn,
>  					  KYBER_NUM_DOMAINS, kqd);
>  	if (!kqd->cb)
>  		goto err_kqd;
> -

Unrelated whitespace change.

>  	/*
>  	 * The maximum number of tokens for any scheduling domain is at least
>  	 * the queue depth of a single hardware queue. If the hardware doesn't
> @@ -376,15 +389,45 @@ static void kyber_exit_sched(struct elevator_queue *e)
>  	kfree(kqd);
>  }
>  
> +static void kyber_ctx_queue_init(struct kyber_ctx_queue *kcq)
> +{
> +	unsigned int i;
> +
> +	spin_lock_init(&kcq->lock);
> +	for (i = 0; i < KYBER_NUM_DOMAINS; i++)
> +		INIT_LIST_HEAD(&kcq->rq_list[i]);
> +}
> +
>  static int kyber_init_hctx(struct blk_mq_hw_ctx *hctx, unsigned int hctx_idx)
>  {
>  	struct kyber_hctx_data *khd;
> -	int i;
> +	int i, sd;
>  
>  	khd = kmalloc_node(sizeof(*khd), GFP_KERNEL, hctx->numa_node);
>  	if (!khd)
>  		return -ENOMEM;
>  
> +	khd->kcqs = kmalloc_array_node(hctx->nr_ctx,
> +			sizeof(struct kyber_ctx_queue),
> +			GFP_KERNEL, hctx->numa_node);

Argument whitespace alignment is still wrong here. I guess you have to
pass --strict to checkpatch.pl to get it to warn about that. Usually I
wouldn't be nitpicky about this but I made sure to keep this file nice
and neat, and we need a v3 anyways :)

> +	if (!khd->kcqs)
> +		goto err_khd;
> +	/*
> +	 * clone the mapping between hctx and ctx to khd and kcq
> +	 */
> +	for (i = 0; i < hctx->nr_ctx; i++) {
> +		kyber_ctx_queue_init(&khd->kcqs[i]);
> +		khd->kcqs[i].index = i;
> +	}
> +
> +	sd = 0;
> +	for (i = 0; i < KYBER_NUM_DOMAINS; i++) {
> +		if (sbitmap_init_node(&khd->kcq_map[i], hctx->nr_ctx,
> +				ilog2(8), GFP_KERNEL, hctx->numa_node))

Whitespace again.

> +			goto err_kcq_map;

In this error case, I'd prefer if we did something similar to
kyber_queue_data_alloc() instead of having this obscurely-named sd
variable:

			while (--i >= 0)
				sbitmap_free(&khd->kcq_map[i]);
			goto err_kcqs;

Where err_kcqs is just the kfree(khd->kcqs).

> +		sd++;
> +	}
> +
>  	spin_lock_init(&khd->lock);
>  
>  	for (i = 0; i < KYBER_NUM_DOMAINS; i++) {
> @@ -402,10 +445,24 @@ static int kyber_init_hctx(struct blk_mq_hw_ctx *hctx, unsigned int hctx_idx)
>  	hctx->sched_data = khd;
>  
>  	return 0;
> +
> +err_kcq_map:
> +	for (i = 0; i < sd; i++)
> +		sbitmap_free(&khd->kcq_map[i]);
> +	kfree(khd->kcqs);
> +err_khd:
> +	kfree(khd);
> +	return -ENOMEM;
>  }
>  
>  static void kyber_exit_hctx(struct blk_mq_hw_ctx *hctx, unsigned int hctx_idx)
>  {
> +	struct kyber_hctx_data *khd = hctx->sched_data;
> +	int i;
> +
> +	for (i = 0; i < KYBER_NUM_DOMAINS; i++)
> +		sbitmap_free(&khd->kcq_map[i]);
> +	kfree(khd->kcqs);
>  	kfree(hctx->sched_data);
>  }
>  
> @@ -427,7 +484,7 @@ static void rq_clear_domain_token(struct kyber_queue_data *kqd,
>  
>  	nr = rq_get_domain_token(rq);
>  	if (nr != -1) {
> -		sched_domain = rq_sched_domain(rq);
> +		sched_domain = kyber_sched_domain(rq->cmd_flags);
>  		sbitmap_queue_clear(&kqd->domain_tokens[sched_domain], nr,
>  				    rq->mq_ctx->cpu);
>  	}
> @@ -446,11 +503,51 @@ static void kyber_limit_depth(unsigned int op, struct blk_mq_alloc_data *data)
>  	}
>  }
>  
> +static bool kyber_bio_merge(struct blk_mq_hw_ctx *hctx, struct bio *bio)
> +{
> +	struct kyber_hctx_data *khd = hctx->sched_data;
> +	struct blk_mq_ctx *ctx = blk_mq_get_ctx(hctx->queue);
> +	struct kyber_ctx_queue *kcq = &khd->kcqs[ctx->index_hw];
> +	struct list_head *rq_list = &kcq->rq_list[kyber_sched_domain(bio->bi_opf)];
> +	bool merged;
> +
> +	spin_lock(&kcq->lock);
> +	merged = blk_mq_bio_list_merge(hctx->queue, rq_list, bio);
> +	spin_unlock(&kcq->lock);
> +	blk_mq_put_ctx(ctx);
> +
> +	return merged;
> +}
> +
>  static void kyber_prepare_request(struct request *rq, struct bio *bio)
>  {
>  	rq_set_domain_token(rq, -1);
>  }
>  
> +static void kyber_insert_requests(struct blk_mq_hw_ctx *hctx,
> +		struct list_head *rq_list, bool at_head)

Whitespace.

> +{
> +	struct kyber_hctx_data *khd = hctx->sched_data;
> +	struct kyber_ctx_queue *kcq;
> +	struct request *rq, *next;
> +	struct list_head *head;
> +	unsigned int sched_domain;
> +
> +	list_for_each_entry_safe(rq, next, rq_list, queuelist) {

Please declare the variables which are local to the loop iteration
inside of the loop body. I.e.,

		unsigned int sched_domain;
		struct kyber_ctx_queue *kcq;
		struct list_head *head;

> +		sched_domain = kyber_sched_domain(rq->cmd_flags);
> +		kcq = &khd->kcqs[rq->mq_ctx->index_hw];
> +		head = &kcq->rq_list[sched_domain];
> +		spin_lock(&kcq->lock);
> +		if (at_head)
> +			list_move(&rq->queuelist, head);
> +		else
> +			list_move_tail(&rq->queuelist, head);
> +		sbitmap_set_bit(&khd->kcq_map[sched_domain], kcq->index);
> +		blk_mq_sched_request_inserted(rq);
> +		spin_unlock(&kcq->lock);
> +	}
> +}
> +
>  static void kyber_finish_request(struct request *rq)
>  {
>  	struct kyber_queue_data *kqd = rq->q->elevator->elevator_data;
> @@ -469,7 +566,7 @@ static void kyber_completed_request(struct request *rq)
>  	 * Check if this request met our latency goal. If not, quickly gather
>  	 * some statistics and start throttling.
>  	 */
> -	sched_domain = rq_sched_domain(rq);
> +	sched_domain = kyber_sched_domain(rq->cmd_flags);
>  	switch (sched_domain) {
>  	case KYBER_READ:
>  		target = kqd->read_lat_nsec;
> @@ -495,19 +592,36 @@ static void kyber_completed_request(struct request *rq)
>  		blk_stat_activate_msecs(kqd->cb, 10);
>  }
>  
> -static void kyber_flush_busy_ctxs(struct kyber_hctx_data *khd,
> -				  struct blk_mq_hw_ctx *hctx)
> +struct flush_kcq_data {
> +	struct kyber_hctx_data *khd;
> +	unsigned int sched_domain;
> +	struct list_head *list;
> +};
> +
> +static bool flush_busy_kcq(struct sbitmap *sb, unsigned int bitnr, void *data)
>  {
> -	LIST_HEAD(rq_list);
> -	struct request *rq, *next;
> +	struct flush_kcq_data *flush_data = data;
> +	struct kyber_ctx_queue *kcq = &flush_data->khd->kcqs[bitnr];
>  
> -	blk_mq_flush_busy_ctxs(hctx, &rq_list);
> -	list_for_each_entry_safe(rq, next, &rq_list, queuelist) {
> -		unsigned int sched_domain;
> +	spin_lock(&kcq->lock);
> +	list_splice_tail_init(&kcq->rq_list[flush_data->sched_domain],
> +			flush_data->list);
> +	sbitmap_clear_bit(sb, bitnr);
> +	spin_unlock(&kcq->lock);
>  
> -		sched_domain = rq_sched_domain(rq);
> -		list_move_tail(&rq->queuelist, &khd->rqs[sched_domain]);
> -	}
> +	return true;
> +}
> +
> +static void kyber_flush_busy_kcqs(struct kyber_hctx_data *khd,
> +		unsigned int sched_domain, struct list_head *list)
> +{
> +	struct flush_kcq_data data = {
> +		.khd = khd,
> +		.sched_domain = sched_domain,
> +		.list = list,
> +	};
> +
> +	sbitmap_for_each_set(&khd->kcq_map[sched_domain], flush_busy_kcq, &data);
>  }
>  
>  static int kyber_domain_wake(wait_queue_entry_t *wait, unsigned mode, int flags,
> @@ -570,26 +684,19 @@ static int kyber_get_domain_token(struct kyber_queue_data *kqd,
>  static struct request *
>  kyber_dispatch_cur_domain(struct kyber_queue_data *kqd,
>  			  struct kyber_hctx_data *khd,
> -			  struct blk_mq_hw_ctx *hctx,
> -			  bool *flushed)
> +			  struct blk_mq_hw_ctx *hctx)
>  {
>  	struct list_head *rqs;
>  	struct request *rq;
>  	int nr;
>  
>  	rqs = &khd->rqs[khd->cur_domain];
> -	rq = list_first_entry_or_null(rqs, struct request, queuelist);
>  
>  	/*
> -	 * If there wasn't already a pending request and we haven't flushed the
> -	 * software queues yet, flush the software queues and check again.
> +	 * If we do have cur_domain rqs on khd or kcq list, then try to require
> +	 * the token
>  	 */
> -	if (!rq && !*flushed) {
> -		kyber_flush_busy_ctxs(khd, hctx);
> -		*flushed = true;
> -		rq = list_first_entry_or_null(rqs, struct request, queuelist);
> -	}
> -
> +	rq = list_first_entry_or_null(rqs, struct request, queuelist);
>  	if (rq) {
>  		nr = kyber_get_domain_token(kqd, khd, hctx);
>  		if (nr >= 0) {
> @@ -598,8 +705,25 @@ kyber_dispatch_cur_domain(struct kyber_queue_data *kqd,
>  			list_del_init(&rq->queuelist);
>  			return rq;
>  		}
> +	} else if (sbitmap_any_bit_set(&khd->kcq_map[khd->cur_domain])) {
> +		nr = kyber_get_domain_token(kqd, khd, hctx);
> +		if (nr >= 0) {
> +			kyber_flush_busy_kcqs(khd, khd->cur_domain, rqs);
> +			rq = list_first_entry_or_null(rqs, struct request, queuelist);

This should just be list_first_entry() if we're so sure that we will
always get a request.

> +			/*
> +			 * khd->lock and kcq->lock will ensure that, if kcq_map[cur_domain]
> +			 * is set, we must be able to get requests from the kcq
> +			 */
> +			khd->batching++;
> +			rq_set_domain_token(rq, nr);
> +			list_del_init(&rq->queuelist);
> +			return rq;
> +		}
> +		/*
> +		 * if not get domain token, the rqs could be left on kcqs to merged
> +		 * with following ios.
> +		 */
>  	}
> -
>  	/* There were either no pending requests or no tokens. */
>  	return NULL;
>  }

Clever logic in this function :) I think it needs a more complete
explanation rather than the little pieces sprinkled about, how about
this (applied on top of your patch):

diff --git a/block/kyber-iosched.c b/block/kyber-iosched.c
index f60bb4ce1fe6..f4f17527b7be 100644
--- a/block/kyber-iosched.c
+++ b/block/kyber-iosched.c
@@ -696,8 +696,12 @@ kyber_dispatch_cur_domain(struct kyber_queue_data *kqd,
 	rqs = &khd->rqs[khd->cur_domain];
 
 	/*
-	 * If we do have cur_domain rqs on khd or kcq list, then try to require
-	 * the token
+	 * If we already have a flushed request, then we just need to get a
+	 * token for it. Otherwise, if there are pending requests in the kcqs,
+	 * flush the kcqs, but only if we can get a token. If not, we should
+	 * leave the requests in the kcqs so that they can be merged. Note that
+	 * khd->lock serializes the flushes, so if we observed any bit set in
+	 * the kcq_map, we will always get a request.
 	 */
 	rq = list_first_entry_or_null(rqs, struct request, queuelist);
 	if (rq) {
@@ -712,20 +716,12 @@ kyber_dispatch_cur_domain(struct kyber_queue_data *kqd,
 		nr = kyber_get_domain_token(kqd, khd, hctx);
 		if (nr >= 0) {
 			kyber_flush_busy_kcqs(khd, khd->cur_domain, rqs);
-			rq = list_first_entry_or_null(rqs, struct request, queuelist);
-			/*
-			 * khd->lock and kcq->lock will ensure that, if kcq_map[cur_domain]
-			 * is set, we must be able to get requests from the kcq
-			 */
+			rq = list_first_entry(rqs, struct request, queuelist);
 			khd->batching++;
 			rq_set_domain_token(rq, nr);
 			list_del_init(&rq->queuelist);
 			return rq;
 		}
-		/*
-		 * if not get domain token, the rqs could be left on kcqs to merged
-		 * with following ios.
-		 */
 	}
 	/* There were either no pending requests or no tokens. */
 	return NULL;

> @@ -608,7 +732,6 @@ static struct request *kyber_dispatch_request(struct blk_mq_hw_ctx *hctx)
>  {
>  	struct kyber_queue_data *kqd = hctx->queue->elevator->elevator_data;
>  	struct kyber_hctx_data *khd = hctx->sched_data;
> -	bool flushed = false;
>  	struct request *rq;
>  	int i;
>  
> @@ -619,7 +742,7 @@ static struct request *kyber_dispatch_request(struct blk_mq_hw_ctx *hctx)
>  	 * from the batch.
>  	 */
>  	if (khd->batching < kyber_batch_size[khd->cur_domain]) {
> -		rq = kyber_dispatch_cur_domain(kqd, khd, hctx, &flushed);
> +		rq = kyber_dispatch_cur_domain(kqd, khd, hctx);
>  		if (rq)
>  			goto out;
>  	}
> @@ -640,7 +763,7 @@ static struct request *kyber_dispatch_request(struct blk_mq_hw_ctx *hctx)
>  		else
>  			khd->cur_domain++;
>  
> -		rq = kyber_dispatch_cur_domain(kqd, khd, hctx, &flushed);
> +		rq = kyber_dispatch_cur_domain(kqd, khd, hctx);
>  		if (rq)
>  			goto out;
>  	}
> @@ -657,10 +780,12 @@ static bool kyber_has_work(struct blk_mq_hw_ctx *hctx)
>  	int i;
>  
>  	for (i = 0; i < KYBER_NUM_DOMAINS; i++) {
> -		if (!list_empty_careful(&khd->rqs[i]))
> +		if (!list_empty_careful(&khd->rqs[i]) ||
> +				sbitmap_any_bit_set(&khd->kcq_map[i]))

Whitespace.

>  			return true;
>  	}
> -	return sbitmap_any_bit_set(&hctx->ctx_map);
> +
> +	return false;
>  }
>  
>  #define KYBER_LAT_SHOW_STORE(op)					\
> @@ -831,7 +956,9 @@ static struct elevator_type kyber_sched = {
>  		.init_hctx = kyber_init_hctx,
>  		.exit_hctx = kyber_exit_hctx,
>  		.limit_depth = kyber_limit_depth,
> +		.bio_merge = kyber_bio_merge,
>  		.prepare_request = kyber_prepare_request,
> +		.insert_requests = kyber_insert_requests,
>  		.finish_request = kyber_finish_request,
>  		.requeue_request = kyber_finish_request,
>  		.completed_request = kyber_completed_request,
> -- 
> 2.7.4
> 

Powered by blists - more mailing lists