Message-Id: <1467429991-18469-1-git-send-email-tahsin@google.com>
Date:	Fri,  1 Jul 2016 20:26:31 -0700
From:	Tahsin Erdogan <tahsin@...gle.com>
To:	Jens Axboe <axboe@...nel.dk>
Cc:	linux-block@...r.kernel.org, linux-kernel@...r.kernel.org,
	Tahsin Erdogan <tahsin@...gle.com>
Subject: [PATCH] block: do not merge requests without consulting with io scheduler

Before merging a bio into an existing request, io scheduler is called to
get its approval first. However, the requests that come from a plug
flush may get merged by block layer without consulting with io
scheduler.

In case of CFQ, this can cause fairness problems. For instance, if a
request gets merged into a low weight cgroup's request, high weight cgroup
now will depend on low weight cgroup to get scheduled. If high weigt cgroup
needs that io request to complete before submitting more requests, then it
will also lose its timeslice.

Following script demonstrates the problem. Group g1 has a low weight, g2
and g3 have equal high weights but g2's requests are adjacent to g1's
requests so they are subject to merging. Due to these merges, g2 gets
poor disk time allocation.

cat > cfq-merge-repro.sh << "EOF"
#!/bin/bash
set -e

IO_ROOT=/mnt-cgroup/io

mkdir -p $IO_ROOT

if ! mount | grep -qw $IO_ROOT; then
  mount -t cgroup none -oblkio $IO_ROOT
fi

cd $IO_ROOT

for i in g1 g2 g3; do
  if [ -d $i ]; then
    rmdir $i
  fi
done

mkdir g1 && echo 10 > g1/blkio.weight
mkdir g2 && echo 495 > g2/blkio.weight
mkdir g3 && echo 495 > g3/blkio.weight

RUNTIME=10

(echo $BASHPID > g1/cgroup.procs &&
 fio --readonly --name name1 --filename /dev/sdb \
     --rw read --size 64k --bs 64k --time_based \
     --runtime=$RUNTIME --offset=0k &> /dev/null)&

(echo $BASHPID > g2/cgroup.procs &&
 fio --readonly --name name1 --filename /dev/sdb \
     --rw read --size 64k --bs 64k --time_based \
     --runtime=$RUNTIME --offset=64k &> /dev/null)&

(echo $BASHPID > g3/cgroup.procs &&
 fio --readonly --name name1 --filename /dev/sdb \
     --rw read --size 64k --bs 64k --time_based \
     --runtime=$RUNTIME --offset=256k &> /dev/null)&

sleep $((RUNTIME+1))

for i in g1 g2 g3; do
  echo ---- $i ----
  cat $i/blkio.time
done

EOF
# ./cfq-merge-repro.sh
---- g1 ----
8:16 162
---- g2 ----
8:16 165
---- g3 ----
8:16 686

After applying the patch:

# ./cfq-merge-repro.sh
---- g1 ----
8:16 90
---- g2 ----
8:16 445
---- g3 ----
8:16 471

Signed-off-by: Tahsin Erdogan <tahsin@...gle.com>
---
 block/blk-merge.c        |  6 ++++++
 block/cfq-iosched.c      | 15 +++++++++++----
 block/deadline-iosched.c |  2 +-
 block/elevator.c         | 16 ++++++++--------
 include/linux/elevator.h | 11 ++++++++---
 5 files changed, 34 insertions(+), 16 deletions(-)

diff --git a/block/blk-merge.c b/block/blk-merge.c
index 261353166dcf..624171db0942 100644
--- a/block/blk-merge.c
+++ b/block/blk-merge.c
@@ -743,6 +743,12 @@ int attempt_front_merge(struct request_queue *q, struct request *rq)
 int blk_attempt_req_merge(struct request_queue *q, struct request *rq,
 			  struct request *next)
 {
+	struct elevator_queue *e = q->elevator;
+
+	if (e->type->ops.elevator_allow_rq_merge_fn)
+		if (!e->type->ops.elevator_allow_rq_merge_fn(q, rq, next))
+			return 0;
+
 	return attempt_merge(q, rq, next);
 }
 
diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c
index 4a349787bc62..2209e2bf9636 100644
--- a/block/cfq-iosched.c
+++ b/block/cfq-iosched.c
@@ -2531,7 +2531,7 @@ static int cfq_merge(struct request_queue *q, struct request **req,
 	struct request *__rq;
 
 	__rq = cfq_find_rq_fmerge(cfqd, bio);
-	if (__rq && elv_rq_merge_ok(__rq, bio)) {
+	if (__rq && elv_bio_merge_ok(__rq, bio)) {
 		*req = __rq;
 		return ELEVATOR_FRONT_MERGE;
 	}
@@ -2588,8 +2588,8 @@ cfq_merged_requests(struct request_queue *q, struct request *rq,
 		cfq_del_cfqq_rr(cfqd, cfqq);
 }
 
-static int cfq_allow_merge(struct request_queue *q, struct request *rq,
-			   struct bio *bio)
+static int cfq_allow_bio_merge(struct request_queue *q, struct request *rq,
+			       struct bio *bio)
 {
 	struct cfq_data *cfqd = q->elevator->elevator_data;
 	struct cfq_io_cq *cic;
@@ -2613,6 +2613,12 @@ static int cfq_allow_merge(struct request_queue *q, struct request *rq,
 	return cfqq == RQ_CFQQ(rq);
 }
 
+static int cfq_allow_rq_merge(struct request_queue *q, struct request *rq,
+			      struct request *next)
+{
+	return RQ_CFQQ(rq) == RQ_CFQQ(next);
+}
+
 static inline void cfq_del_timer(struct cfq_data *cfqd, struct cfq_queue *cfqq)
 {
 	del_timer(&cfqd->idle_slice_timer);
@@ -4729,7 +4735,8 @@ static struct elevator_type iosched_cfq = {
 		.elevator_merge_fn = 		cfq_merge,
 		.elevator_merged_fn =		cfq_merged_request,
 		.elevator_merge_req_fn =	cfq_merged_requests,
-		.elevator_allow_merge_fn =	cfq_allow_merge,
+		.elevator_allow_bio_merge_fn =	cfq_allow_bio_merge,
+		.elevator_allow_rq_merge_fn =	cfq_allow_rq_merge,
 		.elevator_bio_merged_fn =	cfq_bio_merged,
 		.elevator_dispatch_fn =		cfq_dispatch_requests,
 		.elevator_add_req_fn =		cfq_insert_request,
diff --git a/block/deadline-iosched.c b/block/deadline-iosched.c
index d0dd7882d8c7..408213844175 100644
--- a/block/deadline-iosched.c
+++ b/block/deadline-iosched.c
@@ -137,7 +137,7 @@ deadline_merge(struct request_queue *q, struct request **req, struct bio *bio)
 		if (__rq) {
 			BUG_ON(sector != blk_rq_pos(__rq));
 
-			if (elv_rq_merge_ok(__rq, bio)) {
+			if (elv_bio_merge_ok(__rq, bio)) {
 				ret = ELEVATOR_FRONT_MERGE;
 				goto out;
 			}
diff --git a/block/elevator.c b/block/elevator.c
index c3555c9c672f..57a42590a48f 100644
--- a/block/elevator.c
+++ b/block/elevator.c
@@ -53,13 +53,13 @@ static LIST_HEAD(elv_list);
  * Query io scheduler to see if the current process issuing bio may be
  * merged with rq.
  */
-static int elv_iosched_allow_merge(struct request *rq, struct bio *bio)
+static int elv_iosched_allow_bio_merge(struct request *rq, struct bio *bio)
 {
 	struct request_queue *q = rq->q;
 	struct elevator_queue *e = q->elevator;
 
-	if (e->type->ops.elevator_allow_merge_fn)
-		return e->type->ops.elevator_allow_merge_fn(q, rq, bio);
+	if (e->type->ops.elevator_allow_bio_merge_fn)
+		return e->type->ops.elevator_allow_bio_merge_fn(q, rq, bio);
 
 	return 1;
 }
@@ -67,17 +67,17 @@ static int elv_iosched_allow_merge(struct request *rq, struct bio *bio)
 /*
  * can we safely merge with this request?
  */
-bool elv_rq_merge_ok(struct request *rq, struct bio *bio)
+bool elv_bio_merge_ok(struct request *rq, struct bio *bio)
 {
 	if (!blk_rq_merge_ok(rq, bio))
 		return 0;
 
-	if (!elv_iosched_allow_merge(rq, bio))
+	if (!elv_iosched_allow_bio_merge(rq, bio))
 		return 0;
 
 	return 1;
 }
-EXPORT_SYMBOL(elv_rq_merge_ok);
+EXPORT_SYMBOL(elv_bio_merge_ok);
 
 static struct elevator_type *elevator_find(const char *name)
 {
@@ -426,7 +426,7 @@ int elv_merge(struct request_queue *q, struct request **req, struct bio *bio)
 	/*
 	 * First try one-hit cache.
 	 */
-	if (q->last_merge && elv_rq_merge_ok(q->last_merge, bio)) {
+	if (q->last_merge && elv_bio_merge_ok(q->last_merge, bio)) {
 		ret = blk_try_merge(q->last_merge, bio);
 		if (ret != ELEVATOR_NO_MERGE) {
 			*req = q->last_merge;
@@ -441,7 +441,7 @@ int elv_merge(struct request_queue *q, struct request **req, struct bio *bio)
 	 * See if our hash lookup can find a potential backmerge.
 	 */
 	__rq = elv_rqhash_find(q, bio->bi_iter.bi_sector);
-	if (__rq && elv_rq_merge_ok(__rq, bio)) {
+	if (__rq && elv_bio_merge_ok(__rq, bio)) {
 		*req = __rq;
 		return ELEVATOR_BACK_MERGE;
 	}
diff --git a/include/linux/elevator.h b/include/linux/elevator.h
index 638b324f0291..29f35f8eed1d 100644
--- a/include/linux/elevator.h
+++ b/include/linux/elevator.h
@@ -16,7 +16,11 @@ typedef void (elevator_merge_req_fn) (struct request_queue *, struct request *,
 
 typedef void (elevator_merged_fn) (struct request_queue *, struct request *, int);
 
-typedef int (elevator_allow_merge_fn) (struct request_queue *, struct request *, struct bio *);
+typedef int (elevator_allow_bio_merge_fn) (struct request_queue *,
+					   struct request *, struct bio *);
+
+typedef int (elevator_allow_rq_merge_fn) (struct request_queue *,
+					  struct request *, struct request *);
 
 typedef void (elevator_bio_merged_fn) (struct request_queue *,
 						struct request *, struct bio *);
@@ -46,7 +50,8 @@ struct elevator_ops
 	elevator_merge_fn *elevator_merge_fn;
 	elevator_merged_fn *elevator_merged_fn;
 	elevator_merge_req_fn *elevator_merge_req_fn;
-	elevator_allow_merge_fn *elevator_allow_merge_fn;
+	elevator_allow_bio_merge_fn *elevator_allow_bio_merge_fn;
+	elevator_allow_rq_merge_fn *elevator_allow_rq_merge_fn;
 	elevator_bio_merged_fn *elevator_bio_merged_fn;
 
 	elevator_dispatch_fn *elevator_dispatch_fn;
@@ -157,7 +162,7 @@ extern ssize_t elv_iosched_store(struct request_queue *, const char *, size_t);
 extern int elevator_init(struct request_queue *, char *);
 extern void elevator_exit(struct elevator_queue *);
 extern int elevator_change(struct request_queue *, const char *);
-extern bool elv_rq_merge_ok(struct request *, struct bio *);
+extern bool elv_bio_merge_ok(struct request *, struct bio *);
 extern struct elevator_queue *elevator_alloc(struct request_queue *,
 					struct elevator_type *);
 
-- 
2.8.0.rc3.226.g39d4020

Powered by blists - more mailing lists