Merge tag 'for-5.17/drivers-2022-01-11' of git://git.kernel.dk/linux-block

Pull block driver updates from Jens Axboe: - mtip32xx pci cleanups (Bjorn) - mtip32xx conversion to generic power management (Vaibhav) - rsxx pci powermanagement cleanups (Bjorn) - Remove the rsxx driver. This hardware never saw much adoption, and it's been end of lifed for a while. (Christoph) - MD pull request from Song: - REQ_NOWAIT support (Vishal Verma) - raid6 benchmark optimization (Dirk Müller) - Fix for acct bioset (Xiao Ni) - Clean up max_queued_requests (Mariusz Tkaczyk) - PREEMPT_RT optimization (Davidlohr Bueso) - Use default_groups in kobj_type (Greg Kroah-Hartman) - Use attribute groups in pktcdvd and rnbd (Greg) - NVMe pull request from Christoph: - increment request genctr on completion (Keith Busch, Geliang Tang) - add a 'iopolicy' module parameter (Hannes Reinecke) - print out valid arguments when reading from /dev/nvme-fabrics (Hannes Reinecke) - Use struct_group() in drbd (Kees) - null_blk fixes (Ming) - Get rid of congestion logic in pktcdvd (Neil) - Floppy ejection hang fix (Tasos) - Floppy max user request size fix (Xiongwei) - Loop locking fix (Tetsuo) * tag 'for-5.17/drivers-2022-01-11' of git://git.kernel.dk/linux-block: (32 commits) md: use default_groups in kobj_type md: Move alloc/free acct bioset in to personality lib/raid6: Use strict priority ranking for pq gen() benchmarking lib/raid6: skip benchmark of non-chosen xor_syndrome functions md: fix spelling of "its" md: raid456 add nowait support md: raid10 add nowait support md: raid1 add nowait support md: add support for REQ_NOWAIT md: drop queue limitation for RAID1 and RAID10 md/raid5: play nice with PREEMPT_RT block/rnbd-clt-sysfs: use default_groups in kobj_type pktcdvd: convert to use attribute groups block: null_blk: only set set->nr_maps as 3 if active poll_queues is > 0 nvme: add 'iopolicy' module parameter nvme: drop unused variable ctrl in nvme_setup_cmd nvme: increment request genctr on completion nvme-fabrics: print out valid arguments when reading from /dev/nvme-fabrics block: remove the rsxx driver rsxx: Drop PCI legacy power management ...

Linus Torvalds committed 4 years ago

Commit c9193f42 parents d3c8108 d85bd8240 files · +612 −4501Browse files

Showing the changes to drivers/md (9 of 40 files). Show all

Unified · Split

9 files

drivers/md/md-cluster.c +1 −1
drivers/md/md.c +41 −12
drivers/md/md.h +2 −0
drivers/md/raid0.c +28 −10
drivers/md/raid1-10.c +0 −6
drivers/md/raid1.c +56 −27
drivers/md/raid10.c +67 −40
drivers/md/raid5.c +51 −16
drivers/md/raid5.h +3 −1

drivers/md/md-cluster.c+1−1

@@ -574,7 +574,7 @@ static int process_recvd_msg(struct mddev mddev, struct cluster_msg msg) · 573 unmodified lines
574	int ret = 0;	574	int ret = 0;
575		575
576	if (WARN(mddev->cluster_info->slot_number - 1 == le32_to_cpu(msg->slot),	576	if (WARN(mddev->cluster_info->slot_number - 1 == le32_to_cpu(msg->slot),
577	"node %d received it's own msg\n", le32_to_cpu(msg->slot)))	577	"node %d received its own msg\n", le32_to_cpu(msg->slot)))
578	return -1;	578	return -1;
579	switch (le32_to_cpu(msg->type)) {	579	switch (le32_to_cpu(msg->type)) {
580	case METADATA_UPDATED:	580	case METADATA_UPDATED:

drivers/md/md.c+41−12

 	rcu_read_lock();
 	if (is_suspended(mddev, bio)) {
 		DEFINE_WAIT(__wait);
+		/* Bail out if REQ_NOWAIT is set for the bio */
+		if (bio->bi_opf & REQ_NOWAIT) {
+			rcu_read_unlock();
+			bio_wouldblock_error(bio);
+			return;
+		}
 		for (;;) {
 			prepare_to_wait(&mddev->sb_wait, &__wait,
 					TASK_UNINTERRUPTIBLE);
 	&rdev_ppl_size.attr,
 	NULL,
 };
+ATTRIBUTE_GROUPS(rdev_default);
 static ssize_t
 rdev_attr_show(struct kobject *kobj, struct attribute *attr, char *page)
 {
 static struct kobj_type rdev_ktype = {
 	.release	= rdev_free,
 	.sysfs_ops	= &rdev_sysfs_ops,
-	.default_attrs	= rdev_default_attrs,
+	.default_groups	= rdev_default_groups,
 };
 int md_rdev_init(struct md_rdev *rdev)
 	int err;
 	struct md_rdev *rdev;
 	struct md_personality *pers;
+	bool nowait = true;
 	if (list_empty(&mddev->disks))
 		/* cannot run an array with no devices.. */
 			}
 		}
 		sysfs_notify_dirent_safe(rdev->sysfs_state);
+		nowait = nowait && blk_queue_nowait(bdev_get_queue(rdev->bdev));
 	}
+	/* Set the NOWAIT flags if all underlying devices support it */
+	if (nowait)
+		blk_queue_flag_set(QUEUE_FLAG_NOWAIT, mddev->queue);
 	if (!bioset_initialized(&mddev->bio_set)) {
 		err = bioset_init(&mddev->bio_set, BIO_POOL_SIZE, 0, BIOSET_NEED_BVECS);
 		if (err)
 		if (err)
 			goto exit_bio_set;
 	}
-	if (mddev->level != 1 && mddev->level != 10 &&
-	    !bioset_initialized(&mddev->io_acct_set)) {
-		err = bioset_init(&mddev->io_acct_set, BIO_POOL_SIZE,
-				  offsetof(struct md_io_acct, bio_clone), 0);
-		if (err)
-			goto exit_sync_set;
-	}
 	spin_lock(&pers_lock);
 	pers = find_pers(mddev->level, mddev->clevel);
 	module_put(pers->owner);
 	md_bitmap_destroy(mddev);
 abort:
-	if (mddev->level != 1 && mddev->level != 10)
-		bioset_exit(&mddev->io_acct_set);
-exit_sync_set:
 	bioset_exit(&mddev->sync_set);
 exit_bio_set:
 	bioset_exit(&mddev->bio_set);
 	set_bit(MD_SB_CHANGE_DEVS, &mddev->sb_flags);
 	if (!mddev->thread)
 		md_update_sb(mddev, 1);
+	/*
+	 * If the new disk does not support REQ_NOWAIT,
+	 * disable on the whole MD.
+	 */
+	if (!blk_queue_nowait(bdev_get_queue(rdev->bdev))) {
+		pr_info("%s: Disabling nowait because %s does not support nowait\n",
+			mdname(mddev), bdevname(rdev->bdev, b));
+		blk_queue_flag_clear(QUEUE_FLAG_NOWAIT, mddev->queue);
+	}
 	/*
 	 * Kick recovery, maybe this spare has to be added to the
 	 * array immediately.
 	spin_lock(&pers_lock);
 	/* ensure module won't be unloaded */
 	if (!md_cluster_ops || !try_module_get(md_cluster_mod)) {
-		pr_warn("can't find md-cluster module or get it's reference.\n");
+		pr_warn("can't find md-cluster module or get its reference.\n");
 		spin_unlock(&pers_lock);
 		return -ENOENT;
 	}
 }
 EXPORT_SYMBOL_GPL(md_submit_discard_bio);
+int acct_bioset_init(struct mddev *mddev)
+{
+	int err = 0;
+	if (!bioset_initialized(&mddev->io_acct_set))
+		err = bioset_init(&mddev->io_acct_set, BIO_POOL_SIZE,
+			offsetof(struct md_io_acct, bio_clone), 0);
+	return err;
+}
+EXPORT_SYMBOL_GPL(acct_bioset_init);
+void acct_bioset_exit(struct mddev *mddev)
+{
+	bioset_exit(&mddev->io_acct_set);
+}
+EXPORT_SYMBOL_GPL(acct_bioset_exit);
 static void md_end_io_acct(struct bio *bio)
 {
 	struct md_io_acct *md_io_acct = bio->bi_private;

drivers/md/md.h+2−0

@@ -721,6 +721,8 @@ extern void md_error(struct mddev mddev, struct md_rdev rdev); · 720 unmodified lines
721	extern void md_finish_reshape(struct mddev *mddev);	721	extern void md_finish_reshape(struct mddev *mddev);
722	void md_submit_discard_bio(struct mddev mddev, struct md_rdev rdev,	722	void md_submit_discard_bio(struct mddev mddev, struct md_rdev rdev,
723	struct bio *bio, sector_t start, sector_t size);	723	struct bio *bio, sector_t start, sector_t size);
		724	int acct_bioset_init(struct mddev *mddev);
		725	void acct_bioset_exit(struct mddev *mddev);
724	void md_account_bio(struct mddev mddev, struct bio *bio);	726	void md_account_bio(struct mddev mddev, struct bio *bio);
725		727
726	extern bool __must_check md_flush_request(struct mddev mddev, struct bio bio);	728	extern bool __must_check md_flush_request(struct mddev mddev, struct bio bio);

drivers/md/raid0.c+28−10

@@ -356,7 +356,21 @@ static sector_t raid0_size(struct mddev *mddev, sector_t sectors, int raid_disks · 355 unmodified lines
356	return array_sectors;	356	return array_sectors;
357	}	357	}
358		358
359	static void raid0_free(struct mddev mddev, void priv);	359	static void free_conf(struct mddev mddev, struct r0conf conf)
		360	{
		361	kfree(conf->strip_zone);
		362	kfree(conf->devlist);
		363	kfree(conf);
		364	mddev->private = NULL;
		365	}
		366
		367	static void raid0_free(struct mddev mddev, void priv)
		368	{
		369	struct r0conf *conf = priv;
		370
		371	free_conf(mddev, conf);
		372	acct_bioset_exit(mddev);
		373	}
360		374
361	static int raid0_run(struct mddev *mddev)	375	static int raid0_run(struct mddev *mddev)
362	{	376	{
@@ -370,11 +384,16 @@ static int raid0_run(struct mddev *mddev) · 7 unmodified lines
370	if (md_check_no_bitmap(mddev))	384	if (md_check_no_bitmap(mddev))
371	return -EINVAL;	385	return -EINVAL;
372		386
		387	if (acct_bioset_init(mddev)) {
		388	pr_err("md/raid0:%s: alloc acct bioset failed.\n", mdname(mddev));
		389	return -ENOMEM;
		390	}
		391
373	/* if private is not null, we are here after takeover */	392	/* if private is not null, we are here after takeover */
374	if (mddev->private == NULL) {	393	if (mddev->private == NULL) {
375	ret = create_strip_zones(mddev, &conf);	394	ret = create_strip_zones(mddev, &conf);
376	if (ret < 0)	395	if (ret < 0)
377	return ret;	396	goto exit_acct_set;
378	mddev->private = conf;	397	mddev->private = conf;
379	}	398	}
380	conf = mddev->private;	399	conf = mddev->private;
@@ -413,17 +432,16 @@ static int raid0_run(struct mddev *mddev) · 32 unmodified lines
413	dump_zones(mddev);	432	dump_zones(mddev);
414		433
415	ret = md_integrity_register(mddev);	434	ret = md_integrity_register(mddev);
		435	if (ret)
		436	goto free;
416		437
417	return ret;	438	return ret;
418	}
419		439
420	static void raid0_free(struct mddev mddev, void priv)	440	free:
421	{	441	free_conf(mddev, conf);
422	struct r0conf *conf = priv;	442	exit_acct_set:
423		443	acct_bioset_exit(mddev);
424	kfree(conf->strip_zone);	444	return ret;
425	kfree(conf->devlist);
426	kfree(conf);
427	}	445	}
428		446
429	static void raid0_handle_discard(struct mddev mddev, struct bio bio)	447	static void raid0_handle_discard(struct mddev mddev, struct bio bio)

drivers/md/raid1-10.c+0−6

 #define BIO_SPECIAL(bio) ((unsigned long)bio <= 2)
-/* When there are this many requests queue to be written by
- * the raid thread, we become 'congested' to provide back-pressure
- * for writeback.
- */
-static int max_queued_requests = 1024;
 /* for managing resync I/O pages */
 struct resync_pages {
 	void		*raid_bio;

drivers/md/raid1.c+56−27

@@ -929,8 +929,10 @@ static void lower_barrier(struct r1conf *conf, sector_t sector_nr) · 928 unmodified lines
929	wake_up(&conf->wait_barrier);	929	wake_up(&conf->wait_barrier);
930	}	930	}
931		931
932	static void _wait_barrier(struct r1conf *conf, int idx)	932	static bool _wait_barrier(struct r1conf *conf, int idx, bool nowait)
933	{	933	{
		934	bool ret = true;
		935
934	/*	936	/*
935	* We need to increase conf->nr_pending[idx] very early here,	937	* We need to increase conf->nr_pending[idx] very early here,
936	* then raise_barrier() can be blocked when it waits for	938	* then raise_barrier() can be blocked when it waits for
@@ -961,7 +963,7 @@ static void _wait_barrier(struct r1conf *conf, int idx) · 24 unmodified lines
961	*/	963	*/
962	if (!READ_ONCE(conf->array_frozen) &&	964	if (!READ_ONCE(conf->array_frozen) &&
963	!atomic_read(&conf->barrier[idx]))	965	!atomic_read(&conf->barrier[idx]))
964	return;	966	return ret;
965		967
966	/*	968	/*
967	* After holding conf->resync_lock, conf->nr_pending[idx]	969	* After holding conf->resync_lock, conf->nr_pending[idx]
@@ -979,18 +981,27 @@ static void _wait_barrier(struct r1conf *conf, int idx) · 11 unmodified lines
979	*/	981	*/
980	wake_up(&conf->wait_barrier);	982	wake_up(&conf->wait_barrier);
981	/* Wait for the barrier in same barrier unit bucket to drop. */	983	/* Wait for the barrier in same barrier unit bucket to drop. */
982	wait_event_lock_irq(conf->wait_barrier,	984
983	!conf->array_frozen &&	985	/* Return false when nowait flag is set */
984	!atomic_read(&conf->barrier[idx]),	986	if (nowait) {
985	conf->resync_lock);	987	ret = false;
986	atomic_inc(&conf->nr_pending[idx]);	988	} else {
		989	wait_event_lock_irq(conf->wait_barrier,
		990	!conf->array_frozen &&
		991	!atomic_read(&conf->barrier[idx]),
		992	conf->resync_lock);
		993	atomic_inc(&conf->nr_pending[idx]);
		994	}
		995
987	atomic_dec(&conf->nr_waiting[idx]);	996	atomic_dec(&conf->nr_waiting[idx]);
988	spin_unlock_irq(&conf->resync_lock);	997	spin_unlock_irq(&conf->resync_lock);
		998	return ret;
989	}	999	}
990		1000
991	static void wait_read_barrier(struct r1conf *conf, sector_t sector_nr)	1001	static bool wait_read_barrier(struct r1conf *conf, sector_t sector_nr, bool nowait)
992	{	1002	{
993	int idx = sector_to_idx(sector_nr);	1003	int idx = sector_to_idx(sector_nr);
		1004	bool ret = true;
994		1005
995	/*	1006	/*
996	* Very similar to _wait_barrier(). The difference is, for read	1007	* Very similar to _wait_barrier(). The difference is, for read
@@ -1002,7 +1013,7 @@ static void wait_read_barrier(struct r1conf *conf, sector_t sector_nr) · 5 unmodified lines
1002	atomic_inc(&conf->nr_pending[idx]);	1013	atomic_inc(&conf->nr_pending[idx]);
1003		1014
1004	if (!READ_ONCE(conf->array_frozen))	1015	if (!READ_ONCE(conf->array_frozen))
1005	return;	1016	return ret;
1006		1017
1007	spin_lock_irq(&conf->resync_lock);	1018	spin_lock_irq(&conf->resync_lock);
1008	atomic_inc(&conf->nr_waiting[idx]);	1019	atomic_inc(&conf->nr_waiting[idx]);
@@ -1013,19 +1024,28 @@ static void wait_read_barrier(struct r1conf *conf, sector_t sector_nr) · 4 unmodified lines
1013	*/	1024	*/
1014	wake_up(&conf->wait_barrier);	1025	wake_up(&conf->wait_barrier);
1015	/* Wait for array to be unfrozen */	1026	/* Wait for array to be unfrozen */
1016	wait_event_lock_irq(conf->wait_barrier,	1027
1017	!conf->array_frozen,	1028	/* Return false when nowait flag is set */
1018	conf->resync_lock);	1029	if (nowait) {
1019	atomic_inc(&conf->nr_pending[idx]);	1030	/* Return false when nowait flag is set */
		1031	ret = false;
		1032	} else {
		1033	wait_event_lock_irq(conf->wait_barrier,
		1034	!conf->array_frozen,
		1035	conf->resync_lock);
		1036	atomic_inc(&conf->nr_pending[idx]);
		1037	}
		1038
1020	atomic_dec(&conf->nr_waiting[idx]);	1039	atomic_dec(&conf->nr_waiting[idx]);
1021	spin_unlock_irq(&conf->resync_lock);	1040	spin_unlock_irq(&conf->resync_lock);
		1041	return ret;
1022	}	1042	}
1023		1043
1024	static void wait_barrier(struct r1conf *conf, sector_t sector_nr)	1044	static bool wait_barrier(struct r1conf *conf, sector_t sector_nr, bool nowait)
1025	{	1045	{
1026	int idx = sector_to_idx(sector_nr);	1046	int idx = sector_to_idx(sector_nr);
1027		1047
1028	_wait_barrier(conf, idx);	1048	return _wait_barrier(conf, idx, nowait);
1029	}	1049	}
1030		1050
1031	static void _allow_barrier(struct r1conf *conf, int idx)	1051	static void _allow_barrier(struct r1conf *conf, int idx)
@@ -1236,7 +1256,11 @@ static void raid1_read_request(struct mddev mddev, struct bio bio, · 204 unmodified lines
1236	* Still need barrier for READ in case that whole	1256	* Still need barrier for READ in case that whole
1237	* array is frozen.	1257	* array is frozen.
1238	*/	1258	*/
1239	wait_read_barrier(conf, bio->bi_iter.bi_sector);	1259	if (!wait_read_barrier(conf, bio->bi_iter.bi_sector,
		1260	bio->bi_opf & REQ_NOWAIT)) {
		1261	bio_wouldblock_error(bio);
		1262	return;
		1263	}
1240		1264
1241	if (!r1_bio)	1265	if (!r1_bio)
1242	r1_bio = alloc_r1bio(mddev, bio);	1266	r1_bio = alloc_r1bio(mddev, bio);
@@ -1336,6 +1360,10 @@ static void raid1_write_request(struct mddev mddev, struct bio bio, · 93 unmodified lines
1336	bio->bi_iter.bi_sector, bio_end_sector(bio))) {	1360	bio->bi_iter.bi_sector, bio_end_sector(bio))) {
1337		1361
1338	DEFINE_WAIT(w);	1362	DEFINE_WAIT(w);
		1363	if (bio->bi_opf & REQ_NOWAIT) {
		1364	bio_wouldblock_error(bio);
		1365	return;
		1366	}
1339	for (;;) {	1367	for (;;) {
1340	prepare_to_wait(&conf->wait_barrier,	1368	prepare_to_wait(&conf->wait_barrier,
1341	&w, TASK_IDLE);	1369	&w, TASK_IDLE);
@@ -1353,17 +1381,15 @@ static void raid1_write_request(struct mddev mddev, struct bio bio, · 11 unmodified lines
1353	* thread has put up a bar for new requests.	1381	* thread has put up a bar for new requests.
1354	* Continue immediately if no resync is active currently.	1382	* Continue immediately if no resync is active currently.
1355	*/	1383	*/
1356	wait_barrier(conf, bio->bi_iter.bi_sector);	1384	if (!wait_barrier(conf, bio->bi_iter.bi_sector,
		1385	bio->bi_opf & REQ_NOWAIT)) {
		1386	bio_wouldblock_error(bio);
		1387	return;
		1388	}
1357		1389
1358	r1_bio = alloc_r1bio(mddev, bio);	1390	r1_bio = alloc_r1bio(mddev, bio);
1359	r1_bio->sectors = max_write_sectors;	1391	r1_bio->sectors = max_write_sectors;
1360		1392
1361	if (conf->pending_count >= max_queued_requests) {
1362	md_wakeup_thread(mddev->thread);
1363	raid1_log(mddev, "wait queued");
1364	wait_event(conf->wait_barrier,
1365	conf->pending_count < max_queued_requests);
1366	}
1367	/* first select target devices under rcu_lock and	1393	/* first select target devices under rcu_lock and
1368	* inc refcount on their rdev. Record them by setting	1394	* inc refcount on their rdev. Record them by setting
1369	* bios[x] to bio	1395	* bios[x] to bio
@@ -1458,9 +1484,14 @@ static void raid1_write_request(struct mddev mddev, struct bio bio, · 88 unmodified lines
1458	rdev_dec_pending(conf->mirrors[j].rdev, mddev);	1484	rdev_dec_pending(conf->mirrors[j].rdev, mddev);
1459	r1_bio->state = 0;	1485	r1_bio->state = 0;
1460	allow_barrier(conf, bio->bi_iter.bi_sector);	1486	allow_barrier(conf, bio->bi_iter.bi_sector);
		1487
		1488	if (bio->bi_opf & REQ_NOWAIT) {
		1489	bio_wouldblock_error(bio);
		1490	return;
		1491	}
1461	raid1_log(mddev, "wait rdev %d blocked", blocked_rdev->raid_disk);	1492	raid1_log(mddev, "wait rdev %d blocked", blocked_rdev->raid_disk);
1462	md_wait_for_blocked_rdev(blocked_rdev, mddev);	1493	md_wait_for_blocked_rdev(blocked_rdev, mddev);
1463	wait_barrier(conf, bio->bi_iter.bi_sector);	1494	wait_barrier(conf, bio->bi_iter.bi_sector, false);
1464	goto retry_write;	1495	goto retry_write;
1465	}	1496	}
1466		1497
@@ -1688,7 +1719,7 @@ static void close_sync(struct r1conf *conf) · 221 unmodified lines
1688	int idx;	1719	int idx;
1689		1720
1690	for (idx = 0; idx < BARRIER_BUCKETS_NR; idx++) {	1721	for (idx = 0; idx < BARRIER_BUCKETS_NR; idx++) {
1691	_wait_barrier(conf, idx);	1722	_wait_barrier(conf, idx, false);
1692	_allow_barrier(conf, idx);	1723	_allow_barrier(conf, idx);
1693	}	1724	}
1694		1725
@@ -3410,5 +3441,3 @@ MODULE_DESCRIPTION("RAID1 (mirroring) personality for MD"); · 1715 unmodified lines
3410	MODULE_ALIAS("md-personality-3"); /* RAID1 */	3441	MODULE_ALIAS("md-personality-3"); /* RAID1 */
3411	MODULE_ALIAS("md-raid1");	3442	MODULE_ALIAS("md-raid1");
3412	MODULE_ALIAS("md-level-1");	3443	MODULE_ALIAS("md-level-1");
3413
3414	module_param(max_queued_requests, int, S_IRUGO\|S_IWUSR);

drivers/md/raid10.c+67−40

 	wake_up(&conf->wait_barrier);
 }
-static void wait_barrier(struct r10conf *conf)
+static bool wait_barrier(struct r10conf *conf, bool nowait)
 {
+	bool ret = true;
 	spin_lock_irq(&conf->resync_lock);
 	if (conf->barrier) {
 		struct bio_list *bio_list = current->bio_list;
 		 * that queue to get the nr_pending
 		 * count down.
 		 */
-		raid10_log(conf->mddev, "wait barrier");
+		/* Return false when nowait flag is set */
-		wait_event_lock_irq(conf->wait_barrier,
+		if (nowait) {
-				    !conf->barrier ||
+			ret = false;
-				    (atomic_read(&conf->nr_pending) &&
+		} else {
-				     bio_list &&
+			raid10_log(conf->mddev, "wait barrier");
-				     (!bio_list_empty(&bio_list[0]) ||
+			wait_event_lock_irq(conf->wait_barrier,
-				      !bio_list_empty(&bio_list[1]))) ||
+					    !conf->barrier ||
-				     /* move on if recovery thread is
+					    (atomic_read(&conf->nr_pending) &&
-				      * blocked by us
+					     bio_list &&
-				      */
+					     (!bio_list_empty(&bio_list[0]) ||
-				     (conf->mddev->thread->tsk == current &&
+					      !bio_list_empty(&bio_list[1]))) ||
-				      test_bit(MD_RECOVERY_RUNNING,
+					     /* move on if recovery thread is
-					       &conf->mddev->recovery) &&
+					      * blocked by us
-				      conf->nr_queued > 0),
+					      */
-				    conf->resync_lock);
+					     (conf->mddev->thread->tsk == current &&
+					      test_bit(MD_RECOVERY_RUNNING,
+						       &conf->mddev->recovery) &&
+					      conf->nr_queued > 0),
+					    conf->resync_lock);
+		}
 		conf->nr_waiting--;
 		if (!conf->nr_waiting)
 			wake_up(&conf->wait_barrier);
 	}
-	atomic_inc(&conf->nr_pending);
+	/* Only increment nr_pending when we wait */
+	if (ret)
+		atomic_inc(&conf->nr_pending);
 	spin_unlock_irq(&conf->resync_lock);
+	return ret;
 }
 static void allow_barrier(struct r10conf *conf)
  * currently.
  * 2. If IO spans the reshape position.  Need to wait for reshape to pass.
  */
-static void regular_request_wait(struct mddev *mddev, struct r10conf *conf,
+static bool regular_request_wait(struct mddev *mddev, struct r10conf *conf,
 				 struct bio *bio, sector_t sectors)
 {
-	wait_barrier(conf);
+	/* Bail out if REQ_NOWAIT is set for the bio */
+	if (!wait_barrier(conf, bio->bi_opf & REQ_NOWAIT)) {
+		bio_wouldblock_error(bio);
+		return false;
+	}
 	while (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery) &&
 	    bio->bi_iter.bi_sector < conf->reshape_progress &&
 	    bio->bi_iter.bi_sector + sectors > conf->reshape_progress) {
-		raid10_log(conf->mddev, "wait reshape");
 		allow_barrier(conf);
+		if (bio->bi_opf & REQ_NOWAIT) {
+			bio_wouldblock_error(bio);
+			return false;
+		}
+		raid10_log(conf->mddev, "wait reshape");
 		wait_event(conf->wait_barrier,
 			   conf->reshape_progress <= bio->bi_iter.bi_sector ||
 			   conf->reshape_progress >= bio->bi_iter.bi_sector +
 			   sectors);
-		wait_barrier(conf);
+		wait_barrier(conf, false);
 	}
+	return true;
 }
 static void raid10_read_request(struct mddev *mddev, struct bio *bio,
 		rcu_read_unlock();
 	}
-	regular_request_wait(mddev, conf, bio, r10_bio->sectors);
+	if (!regular_request_wait(mddev, conf, bio, r10_bio->sectors))
+		return;
 	rdev = read_balance(conf, r10_bio, &max_sectors);
 	if (!rdev) {
 		if (err_rdev) {
 		bio_chain(split, bio);
 		allow_barrier(conf);
 		submit_bio_noacct(bio);
-		wait_barrier(conf);
+		wait_barrier(conf, false);
 		bio = split;
 		r10_bio->master_bio = bio;
 		r10_bio->sectors = max_sectors;
 		raid10_log(conf->mddev, "%s wait rdev %d blocked",
 				__func__, blocked_rdev->raid_disk);
 		md_wait_for_blocked_rdev(blocked_rdev, mddev);
-		wait_barrier(conf);
+		wait_barrier(conf, false);
 		goto retry_wait;
 	}
 }
 					    bio->bi_iter.bi_sector,
 					    bio_end_sector(bio)))) {
 		DEFINE_WAIT(w);
+		/* Bail out if REQ_NOWAIT is set for the bio */
+		if (bio->bi_opf & REQ_NOWAIT) {
+			bio_wouldblock_error(bio);
+			return;
+		}
 		for (;;) {
 			prepare_to_wait(&conf->wait_barrier,
 					&w, TASK_IDLE);
 	}
 	sectors = r10_bio->sectors;
-	regular_request_wait(mddev, conf, bio, sectors);
+	if (!regular_request_wait(mddev, conf, bio, sectors))
+		return;
 	if (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery) &&
 	    (mddev->reshape_backwards
 	     ? (bio->bi_iter.bi_sector < conf->reshape_safe &&
 		set_mask_bits(&mddev->sb_flags, 0,
 			      BIT(MD_SB_CHANGE_DEVS) | BIT(MD_SB_CHANGE_PENDING));
 		md_wakeup_thread(mddev->thread);
+		if (bio->bi_opf & REQ_NOWAIT) {
+			allow_barrier(conf);
+			bio_wouldblock_error(bio);
+			return;
+		}
 		raid10_log(conf->mddev, "wait reshape metadata");
 		wait_event(mddev->sb_wait,
 			   !test_bit(MD_SB_CHANGE_PENDING, &mddev->sb_flags));
 		conf->reshape_safe = mddev->reshape_position;
 	}
-	if (conf->pending_count >= max_queued_requests) {
-		md_wakeup_thread(mddev->thread);
-		raid10_log(mddev, "wait queued");
-		wait_event(conf->wait_barrier,
-			   conf->pending_count < max_queued_requests);
-	}
 	/* first select target devices under rcu_lock and
 	 * inc refcount on their rdev.  Record them by setting
 	 * bios[x] to bio
 		bio_chain(split, bio);
 		allow_barrier(conf);
 		submit_bio_noacct(bio);
-		wait_barrier(conf);
+		wait_barrier(conf, false);
 		bio = split;
 		r10_bio->master_bio = bio;
 	}
 	if (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery))
 		return -EAGAIN;
-	wait_barrier(conf);
+	if (WARN_ON_ONCE(bio->bi_opf & REQ_NOWAIT)) {
+		bio_wouldblock_error(bio);
+		return 0;
+	}
+	wait_barrier(conf, false);
 	/*
 	 * Check reshape again to avoid reshape happens after checking
 		allow_barrier(conf);
 		/* Resend the fist split part */
 		submit_bio_noacct(split);
-		wait_barrier(conf);
+		wait_barrier(conf, false);
 	}
 	div_u64_rem(bio_end, stripe_size, &remainder);
 	if (remainder) {
 		/* Resend the second split part */
 		submit_bio_noacct(bio);
 		bio = split;
-		wait_barrier(conf);
+		wait_barrier(conf, false);
 	}
 	bio_start = bio->bi_iter.bi_sector;
 		end_disk_offset += geo->stride;
 		atomic_inc(&first_r10bio->remaining);
 		raid_end_discard_bio(r10_bio);
-		wait_barrier(conf);
+		wait_barrier(conf, false);
 		goto retry_discard;
 	}
 static void close_sync(struct r10conf *conf)
 {
-	wait_barrier(conf);
+	wait_barrier(conf, false);
 	allow_barrier(conf);
 	mempool_exit(&conf->r10buf_pool);
 	if (need_flush ||
 	    time_after(jiffies, conf->reshape_checkpoint + 10*HZ)) {
 		/* Need to update reshape_position in metadata */
-		wait_barrier(conf);
+		wait_barrier(conf, false);
 		mddev->reshape_position = conf->reshape_progress;
 		if (mddev->reshape_backwards)
 			mddev->curr_resync_completed = raid10_size(mddev, 0, 0)
 MODULE_ALIAS("md-personality-9"); /* RAID10 */
 MODULE_ALIAS("md-raid10");
 MODULE_ALIAS("md-level-10");
-module_param(max_queued_requests, int, S_IRUGO|S_IWUSR);

drivers/md/raid5.c+51−16

 	struct r5conf *conf = sh->raid_conf;
 	int level = conf->level;
 	struct raid5_percpu *percpu;
-	unsigned long cpu;
-	cpu = get_cpu();
+	local_lock(&conf->percpu->lock);
-	percpu = per_cpu_ptr(conf->percpu, cpu);
+	percpu = this_cpu_ptr(conf->percpu);
 	if (test_bit(STRIPE_OP_BIOFILL, &ops_request)) {
 		ops_run_biofill(sh);
 		overlap_clear++;
 			BUG();
 	}
-	if (overlap_clear && !sh->batch_head)
+	if (overlap_clear && !sh->batch_head) {
 		for (i = disks; i--; ) {
 			struct r5dev *dev = &sh->dev[i];
 			if (test_and_clear_bit(R5_Overlap, &dev->flags))
 				wake_up(&sh->raid_conf->wait_for_overlap);
 		}
-	put_cpu();
+	}
+	local_unlock(&conf->percpu->lock);
 }
 static void free_stripe(struct kmem_cache *sc, struct stripe_head *sh)
 	struct stripe_head *sh;
 	int stripe_sectors;
+	/* We need to handle this when io_uring supports discard/trim */
+	if (WARN_ON_ONCE(bi->bi_opf & REQ_NOWAIT))
+		return;
 	if (mddev->reshape_position != MaxSector)
 		/* Skip discard while reshape is happening */
 		return;
 	last_sector = bio_end_sector(bi);
 	bi->bi_next = NULL;
+	/* Bail out if conflicts with reshape and REQ_NOWAIT is set */
+	if ((bi->bi_opf & REQ_NOWAIT) &&
+	    (conf->reshape_progress != MaxSector) &&
+	    (mddev->reshape_backwards
+	    ? (logical_sector > conf->reshape_progress && logical_sector <= conf->reshape_safe)
+	    : (logical_sector >= conf->reshape_safe && logical_sector < conf->reshape_progress))) {
+		bio_wouldblock_error(bi);
+		if (rw == WRITE)
+			md_write_end(mddev);
+		return true;
+	}
 	md_account_bio(mddev, &bi);
 	prepare_to_wait(&conf->wait_for_overlap, &w, TASK_UNINTERRUPTIBLE);
 	for (; logical_sector < last_sector; logical_sector += RAID5_STRIPE_SECTORS(conf)) {
 		return -ENOMEM;
 	}
+	local_lock_init(&percpu->lock);
 	return 0;
 }
 	struct md_rdev *rdev;
 	struct md_rdev *journal_dev = NULL;
 	sector_t reshape_offset = 0;
-	int i;
+	int i, ret = 0;
 	long long min_offset_diff = 0;
 	int first = 1;
-	if (mddev_init_writes_pending(mddev) < 0)
+	if (acct_bioset_init(mddev)) {
+		pr_err("md/raid456:%s: alloc acct bioset failed.\n", mdname(mddev));
 		return -ENOMEM;
+	}
+	if (mddev_init_writes_pending(mddev) < 0) {
+		ret = -ENOMEM;
+		goto exit_acct_set;
+	}
 	if (mddev->recovery_cp != MaxSector)
 		pr_notice("md/raid:%s: not clean -- starting background reconstruction\n",
 	    (mddev->bitmap_info.offset || mddev->bitmap_info.file)) {
 		pr_notice("md/raid:%s: array cannot have both journal and bitmap\n",
 			  mdname(mddev));
-		return -EINVAL;
+		ret = -EINVAL;
+		goto exit_acct_set;
 	}
 	if (mddev->reshape_position != MaxSector) {
 		if (journal_dev) {
 			pr_warn("md/raid:%s: don't support reshape with journal - aborting.\n",
 				mdname(mddev));
-			return -EINVAL;
+			ret = -EINVAL;
+			goto exit_acct_set;
 		}
 		if (mddev->new_level != mddev->level) {
 			pr_warn("md/raid:%s: unsupported reshape required - aborting.\n",
 				mdname(mddev));
-			return -EINVAL;
+			ret = -EINVAL;
+			goto exit_acct_set;
 		}
 		old_disks = mddev->raid_disks - mddev->delta_disks;
 		/* reshape_position must be on a new-stripe boundary, and one
 		if (sector_div(here_new, chunk_sectors * new_data_disks)) {
 			pr_warn("md/raid:%s: reshape_position not on a stripe boundary\n",
 				mdname(mddev));
-			return -EINVAL;
+			ret = -EINVAL;
+			goto exit_acct_set;
 		}
 		reshape_offset = here_new * chunk_sectors;
 		/* here_new is the stripe we will write to */
 			else if (mddev->ro == 0) {
 				pr_warn("md/raid:%s: in-place reshape must be started in read-only mode - aborting\n",
 					mdname(mddev));
-				return -EINVAL;
+				ret = -EINVAL;
+				goto exit_acct_set;
 			}
 		} else if (mddev->reshape_backwards
 		    ? (here_new * chunk_sectors + min_offset_diff <=
 			/* Reading from the same stripe as writing to - bad */
 			pr_warn("md/raid:%s: reshape_position too early for auto-recovery - aborting.\n",
 				mdname(mddev));
-			return -EINVAL;
+			ret = -EINVAL;
+			goto exit_acct_set;
 		}
 		pr_debug("md/raid:%s: reshape will continue\n", mdname(mddev));
 		/* OK, we should be able to continue; */
 	else
 		conf = mddev->private;
-	if (IS_ERR(conf))
+	if (IS_ERR(conf)) {
-		return PTR_ERR(conf);
+		ret = PTR_ERR(conf);
+		goto exit_acct_set;
+	}
 	if (test_bit(MD_HAS_JOURNAL, &mddev->flags)) {
 		if (!journal_dev) {
 	free_conf(conf);
 	mddev->private = NULL;
 	pr_warn("md/raid:%s: failed to run raid set.\n", mdname(mddev));
-	return -EIO;
+	ret = -EIO;
+exit_acct_set:
+	acct_bioset_exit(mddev);
+	return ret;
 }
 static void raid5_free(struct mddev *mddev, void *priv)
 	struct r5conf *conf = priv;
 	free_conf(conf);
+	acct_bioset_exit(mddev);
 	mddev->to_remove = &raid5_attrs_group;
 }

drivers/md/raid5.h+3−1

 #include <linux/raid/xor.h>
 #include <linux/dmaengine.h>
+#include <linux/local_lock.h>
 /*
  *
 					     * lists and performing address
 					     * conversions
 					     */
-		int scribble_obj_size;
+		int             scribble_obj_size;
+		local_lock_t    lock;
 	} __percpu *percpu;
 	int scribble_disks;
 	int scribble_sectors;

Merge tag 'for-5.17/drivers-2022-01-11' of git://git.kernel.dk/linux-block

Keyboard shortcuts

General

Issues and pull requests

Issues, pull requests and Inbox

Repository code