Commit badcb143 authored by Yu Kuai's avatar Yu Kuai Committed by Zheng Zengkai
Browse files

block: update nsecs[] in part_stat_show() and diskstats_show() 

hulk inclusion
category: bugfix
bugzilla: https://gitee.com/openeuler/kernel/issues/I57S8D


CVE: NA

--------------------------------

commit 7ec2ec68 ("block: update io_ticks when io hang") fixed that
%util will be zero for iostat when io is hanged, however, avgqu-sz is
still zero while it represents the number of io that are hunged. On the
other hand, for some slow device, if an io is started before and done
after diskstats is read, the avgqu-sz will be miscalculated.

To fix the problem, update 'nsecs[]' when part_stat_show() or
diskstats_show() is called. In order to do that, add 'stat_time' in
struct hd_struct and 'rq_stat_time' in struct request to record the
time. And during iteration, update 'nsecs[]' for each inflight request.

Signed-off-by: default avatarYu Kuai <yukuai3@huawei.com>
Reviewed-by: default avatarJason Yan <yanaijie@huawei.com>
Signed-off-by: default avatarZheng Zengkai <zhengzengkai@huawei.com>
parent 510d96b0

block/blk-core.c

+13 −2
Original line number Diff line number Diff line
@@ -1292,13 +1292,24 @@ void blk_account_io_done(struct request *req, u64 now) 
	    !(req->rq_flags & RQF_FLUSH_SEQ)) {

		const int sgrp = op_stat_group(req_op(req));

		struct hd_struct *part;

		u64 stat_time;



		part_stat_lock();

		part = req->part;



		update_io_ticks(part, jiffies, true);

		part_stat_inc(part, ios[sgrp]);

		part_stat_add(part, nsecs[sgrp], now - req->start_time_ns);

		stat_time = READ_ONCE(req->stat_time_ns);

		/*

		 * This might fail if 'req->stat_time_ns' is updated

		 * in blk_mq_check_inflight_with_stat().

		 */

		if (likely(cmpxchg64(&req->stat_time_ns, stat_time, now)

			   == stat_time)) {

			u64 duation = stat_time ? now - stat_time :

				now - req->start_time_ns;



			part_stat_add(req->part, nsecs[sgrp], duation);

		}

		part_stat_unlock();



		hd_struct_put(part);

block/blk-mq.c

+45 −0
Original line number Diff line number Diff line
@@ -102,6 +102,50 @@ struct mq_inflight { 
	unsigned int inflight[2];

};



static bool blk_mq_check_inflight_with_stat(struct blk_mq_hw_ctx *hctx,

					    struct request *rq, void *priv,

					    bool reserved)

{

	struct mq_inflight *mi = priv;



	if ((!mi->part->partno || rq->part == mi->part) &&

	    blk_mq_rq_state(rq) == MQ_RQ_IN_FLIGHT) {

		u64 stat_time;



		mi->inflight[rq_data_dir(rq)]++;

		if (!rq->part)

			return true;



		stat_time = READ_ONCE(rq->stat_time_ns);

		/*

		 * This might fail if 'req->stat_time_ns' is updated in

		 * blk_account_io_done().

		 */

		if (likely(cmpxchg64(&rq->stat_time_ns, stat_time,

				   rq->part->stat_time) == stat_time)) {

			int sgrp = op_stat_group(req_op(rq));

			u64 duation = stat_time ?

				rq->part->stat_time - stat_time :

				rq->part->stat_time - rq->start_time_ns;



			part_stat_add(rq->part, nsecs[sgrp], duation);

		}

	}



	return true;

}



unsigned int blk_mq_in_flight_with_stat(struct request_queue *q,

					struct hd_struct *part)

{

	struct mq_inflight mi = { .part = part };



	blk_mq_queue_tag_busy_iter(q, blk_mq_check_inflight_with_stat, &mi);



	return mi.inflight[0] + mi.inflight[1];

}





static bool blk_mq_check_inflight(struct blk_mq_hw_ctx *hctx,

				  struct request *rq, void *priv,

				  bool reserved)
@@ -328,6 +372,7 @@ 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->stat_time_ns = 0;

	rq->io_start_time_ns = 0;

	rq->stats_sectors = 0;

	rq->nr_phys_segments = 0;

block/blk-mq.h

+2 −0
Original line number Diff line number Diff line
@@ -187,6 +187,8 @@ static inline bool blk_mq_hw_queue_mapped(struct blk_mq_hw_ctx *hctx) 
unsigned int blk_mq_in_flight(struct request_queue *q, struct hd_struct *part);

void blk_mq_in_flight_rw(struct request_queue *q, struct hd_struct *part,

			 unsigned int inflight[2]);

unsigned int blk_mq_in_flight_with_stat(struct request_queue *q,

					struct hd_struct *part);



static inline void blk_mq_put_dispatch_budget(struct request_queue *q)

{

block/genhd.c

+39 −22
Original line number Diff line number Diff line
@@ -1293,25 +1293,52 @@ ssize_t part_size_show(struct device *dev, 
		(unsigned long long)part_nr_sects_read(p));

}



static void part_set_stat_time(struct hd_struct *hd)

{

	u64 now = ktime_get_ns();



again:

	hd->stat_time = now;

	if (hd->partno) {

		hd = &part_to_disk(hd)->part0;

		goto again;

	}

}



static void part_get_stat_info(struct hd_struct *hd, struct disk_stats *stat,

			       unsigned int *inflight)

{

	struct request_queue *q = part_to_disk(hd)->queue;



	if (queue_is_mq(q)) {

		part_stat_lock();

		spin_lock(&hd->bd_stat_lock);

		part_set_stat_time(hd);

		*inflight = blk_mq_in_flight_with_stat(q, hd);

		spin_unlock(&hd->bd_stat_lock);

		part_stat_unlock();

	} else {

		*inflight = part_in_flight(hd);

	}



	if (*inflight) {

		part_stat_lock();

		update_io_ticks(hd, jiffies, true);

		part_stat_unlock();

	}



	part_stat_read_all(hd, stat);

}



ssize_t part_stat_show(struct device *dev,

		       struct device_attribute *attr, char *buf)

{

	struct hd_struct *p = dev_to_part(dev);

	struct request_queue *q = part_to_disk(p)->queue;

	struct disk_stats stat;

	unsigned int inflight;



	if (queue_is_mq(q))

		inflight = blk_mq_in_flight(q, p);

	else

		inflight = part_in_flight(p);

	part_get_stat_info(p, &stat, &inflight);



	if (inflight) {

		part_stat_lock();

		update_io_ticks(p, jiffies, true);

		part_stat_unlock();

	}

	part_stat_read_all(p, &stat);

	return sprintf(buf,

		"%8lu %8lu %8llu %8u "

		"%8lu %8lu %8llu %8u "
@@ -1628,17 +1655,7 @@ static int diskstats_show(struct seq_file *seqf, void *v) 


	disk_part_iter_init(&piter, gp, DISK_PITER_INCL_EMPTY_PART0);

	while ((hd = disk_part_iter_next(&piter))) {

		if (queue_is_mq(gp->queue))

			inflight = blk_mq_in_flight(gp->queue, hd);

		else

			inflight = part_in_flight(hd);



		if (inflight) {

			part_stat_lock();

			update_io_ticks(hd, jiffies, true);

			part_stat_unlock();

		}

		part_stat_read_all(hd, &stat);

		part_get_stat_info(hd, &stat, &inflight);

		seq_printf(seqf, "%4d %7d %s "

			   "%lu %lu %lu %u "

			   "%lu %lu %lu %u "

block/partitions/core.c

+2 −0
Original line number Diff line number Diff line
@@ -415,6 +415,8 @@ static struct hd_struct *add_partition(struct gendisk *disk, int partno, 
	p->nr_sects = len;

	p->partno = partno;

	p->read_only = get_disk_ro(disk) | test_bit(partno, disk->user_ro_bitmap);

	p->stat_time = 0;

	spin_lock_init(&p->bd_stat_lock);



	if (info) {

		struct partition_meta_info *pinfo;