summaryrefslogtreecommitdiffstats
path: root/drivers/block/null_blk
diff options
context:
space:
mode:
authorDaniel Baumann <daniel.baumann@progress-linux.org>2024-08-07 13:11:22 +0000
committerDaniel Baumann <daniel.baumann@progress-linux.org>2024-08-07 13:11:22 +0000
commitb20732900e4636a467c0183a47f7396700f5f743 (patch)
tree42f079ff82e701ebcb76829974b4caca3e5b6798 /drivers/block/null_blk
parentAdding upstream version 6.8.12. (diff)
downloadlinux-b20732900e4636a467c0183a47f7396700f5f743.tar.xz
linux-b20732900e4636a467c0183a47f7396700f5f743.zip
Adding upstream version 6.9.7.upstream/6.9.7
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to 'drivers/block/null_blk')
-rw-r--r--drivers/block/null_blk/main.c581
-rw-r--r--drivers/block/null_blk/null_blk.h24
-rw-r--r--drivers/block/null_blk/trace.h5
-rw-r--r--drivers/block/null_blk/zoned.c27
4 files changed, 184 insertions, 453 deletions
diff --git a/drivers/block/null_blk/main.c b/drivers/block/null_blk/main.c
index 3584f389b9..620679a0ac 100644
--- a/drivers/block/null_blk/main.c
+++ b/drivers/block/null_blk/main.c
@@ -115,6 +115,18 @@ module_param_string(init_hctx, g_init_hctx_str, sizeof(g_init_hctx_str), 0444);
MODULE_PARM_DESC(init_hctx, "Fault injection to fail hctx init. init_hctx=<interval>,<probability>,<space>,<times>");
#endif
+/*
+ * Historic queue modes.
+ *
+ * These days nothing but NULL_Q_MQ is actually supported, but we keep it the
+ * enum for error reporting.
+ */
+enum {
+ NULL_Q_BIO = 0,
+ NULL_Q_RQ = 1,
+ NULL_Q_MQ = 2,
+};
+
static int g_queue_mode = NULL_Q_MQ;
static int null_param_store_val(const char *str, int *val, int min, int max)
@@ -165,8 +177,8 @@ static bool g_blocking;
module_param_named(blocking, g_blocking, bool, 0444);
MODULE_PARM_DESC(blocking, "Register as a blocking blk-mq driver device");
-static bool shared_tags;
-module_param(shared_tags, bool, 0444);
+static bool g_shared_tags;
+module_param_named(shared_tags, g_shared_tags, bool, 0444);
MODULE_PARM_DESC(shared_tags, "Share tag set between devices for blk-mq");
static bool g_shared_tag_bitmap;
@@ -392,13 +404,25 @@ static int nullb_update_nr_hw_queues(struct nullb_device *dev,
static int nullb_apply_submit_queues(struct nullb_device *dev,
unsigned int submit_queues)
{
- return nullb_update_nr_hw_queues(dev, submit_queues, dev->poll_queues);
+ int ret;
+
+ mutex_lock(&lock);
+ ret = nullb_update_nr_hw_queues(dev, submit_queues, dev->poll_queues);
+ mutex_unlock(&lock);
+
+ return ret;
}
static int nullb_apply_poll_queues(struct nullb_device *dev,
unsigned int poll_queues)
{
- return nullb_update_nr_hw_queues(dev, dev->submit_queues, poll_queues);
+ int ret;
+
+ mutex_lock(&lock);
+ ret = nullb_update_nr_hw_queues(dev, dev->submit_queues, poll_queues);
+ mutex_unlock(&lock);
+
+ return ret;
}
NULLB_DEVICE_ATTR(size, ulong, NULL);
@@ -426,6 +450,7 @@ NULLB_DEVICE_ATTR(zone_max_open, uint, NULL);
NULLB_DEVICE_ATTR(zone_max_active, uint, NULL);
NULLB_DEVICE_ATTR(virt_boundary, bool, NULL);
NULLB_DEVICE_ATTR(no_sched, bool, NULL);
+NULLB_DEVICE_ATTR(shared_tags, bool, NULL);
NULLB_DEVICE_ATTR(shared_tag_bitmap, bool, NULL);
static ssize_t nullb_device_power_show(struct config_item *item, char *page)
@@ -444,28 +469,32 @@ static ssize_t nullb_device_power_store(struct config_item *item,
if (ret < 0)
return ret;
+ ret = count;
+ mutex_lock(&lock);
if (!dev->power && newp) {
if (test_and_set_bit(NULLB_DEV_FL_UP, &dev->flags))
- return count;
+ goto out;
+
ret = null_add_dev(dev);
if (ret) {
clear_bit(NULLB_DEV_FL_UP, &dev->flags);
- return ret;
+ goto out;
}
set_bit(NULLB_DEV_FL_CONFIGURED, &dev->flags);
dev->power = newp;
+ ret = count;
} else if (dev->power && !newp) {
if (test_and_clear_bit(NULLB_DEV_FL_UP, &dev->flags)) {
- mutex_lock(&lock);
dev->power = newp;
null_del_dev(dev->nullb);
- mutex_unlock(&lock);
}
clear_bit(NULLB_DEV_FL_CONFIGURED, &dev->flags);
}
- return count;
+out:
+ mutex_unlock(&lock);
+ return ret;
}
CONFIGFS_ATTR(nullb_device_, power);
@@ -571,6 +600,7 @@ static struct configfs_attribute *nullb_device_attrs[] = {
&nullb_device_attr_zone_offline,
&nullb_device_attr_virt_boundary,
&nullb_device_attr_no_sched,
+ &nullb_device_attr_shared_tags,
&nullb_device_attr_shared_tag_bitmap,
NULL,
};
@@ -653,10 +683,11 @@ static ssize_t memb_group_features_show(struct config_item *item, char *page)
"badblocks,blocking,blocksize,cache_size,"
"completion_nsec,discard,home_node,hw_queue_depth,"
"irqmode,max_sectors,mbps,memory_backed,no_sched,"
- "poll_queues,power,queue_mode,shared_tag_bitmap,size,"
- "submit_queues,use_per_node_hctx,virt_boundary,zoned,"
- "zone_capacity,zone_max_active,zone_max_open,"
- "zone_nr_conv,zone_offline,zone_readonly,zone_size\n");
+ "poll_queues,power,queue_mode,shared_tag_bitmap,"
+ "shared_tags,size,submit_queues,use_per_node_hctx,"
+ "virt_boundary,zoned,zone_capacity,zone_max_active,"
+ "zone_max_open,zone_nr_conv,zone_offline,zone_readonly,"
+ "zone_size\n");
}
CONFIGFS_ATTR_RO(memb_group_, features);
@@ -738,6 +769,7 @@ static struct nullb_device *null_alloc_dev(void)
dev->zone_max_active = g_zone_max_active;
dev->virt_boundary = g_virt_boundary;
dev->no_sched = g_no_sched;
+ dev->shared_tags = g_shared_tags;
dev->shared_tag_bitmap = g_shared_tag_bitmap;
return dev;
}
@@ -752,98 +784,11 @@ static void null_free_dev(struct nullb_device *dev)
kfree(dev);
}
-static void put_tag(struct nullb_queue *nq, unsigned int tag)
-{
- clear_bit_unlock(tag, nq->tag_map);
-
- if (waitqueue_active(&nq->wait))
- wake_up(&nq->wait);
-}
-
-static unsigned int get_tag(struct nullb_queue *nq)
-{
- unsigned int tag;
-
- do {
- tag = find_first_zero_bit(nq->tag_map, nq->queue_depth);
- if (tag >= nq->queue_depth)
- return -1U;
- } while (test_and_set_bit_lock(tag, nq->tag_map));
-
- return tag;
-}
-
-static void free_cmd(struct nullb_cmd *cmd)
-{
- put_tag(cmd->nq, cmd->tag);
-}
-
-static enum hrtimer_restart null_cmd_timer_expired(struct hrtimer *timer);
-
-static struct nullb_cmd *__alloc_cmd(struct nullb_queue *nq)
-{
- struct nullb_cmd *cmd;
- unsigned int tag;
-
- tag = get_tag(nq);
- if (tag != -1U) {
- cmd = &nq->cmds[tag];
- cmd->tag = tag;
- cmd->error = BLK_STS_OK;
- cmd->nq = nq;
- if (nq->dev->irqmode == NULL_IRQ_TIMER) {
- hrtimer_init(&cmd->timer, CLOCK_MONOTONIC,
- HRTIMER_MODE_REL);
- cmd->timer.function = null_cmd_timer_expired;
- }
- return cmd;
- }
-
- return NULL;
-}
-
-static struct nullb_cmd *alloc_cmd(struct nullb_queue *nq, struct bio *bio)
-{
- struct nullb_cmd *cmd;
- DEFINE_WAIT(wait);
-
- do {
- /*
- * This avoids multiple return statements, multiple calls to
- * __alloc_cmd() and a fast path call to prepare_to_wait().
- */
- cmd = __alloc_cmd(nq);
- if (cmd) {
- cmd->bio = bio;
- return cmd;
- }
- prepare_to_wait(&nq->wait, &wait, TASK_UNINTERRUPTIBLE);
- io_schedule();
- finish_wait(&nq->wait, &wait);
- } while (1);
-}
-
-static void end_cmd(struct nullb_cmd *cmd)
-{
- int queue_mode = cmd->nq->dev->queue_mode;
-
- switch (queue_mode) {
- case NULL_Q_MQ:
- blk_mq_end_request(cmd->rq, cmd->error);
- return;
- case NULL_Q_BIO:
- cmd->bio->bi_status = cmd->error;
- bio_endio(cmd->bio);
- break;
- }
-
- free_cmd(cmd);
-}
-
static enum hrtimer_restart null_cmd_timer_expired(struct hrtimer *timer)
{
- end_cmd(container_of(timer, struct nullb_cmd, timer));
+ struct nullb_cmd *cmd = container_of(timer, struct nullb_cmd, timer);
+ blk_mq_end_request(blk_mq_rq_from_pdu(cmd), cmd->error);
return HRTIMER_NORESTART;
}
@@ -856,7 +801,9 @@ static void null_cmd_end_timer(struct nullb_cmd *cmd)
static void null_complete_rq(struct request *rq)
{
- end_cmd(blk_mq_rq_to_pdu(rq));
+ struct nullb_cmd *cmd = blk_mq_rq_to_pdu(rq);
+
+ blk_mq_end_request(rq, cmd->error);
}
static struct nullb_page *null_alloc_page(void)
@@ -1273,7 +1220,7 @@ static int null_transfer(struct nullb *nullb, struct page *page,
static int null_handle_rq(struct nullb_cmd *cmd)
{
- struct request *rq = cmd->rq;
+ struct request *rq = blk_mq_rq_from_pdu(cmd);
struct nullb *nullb = cmd->nq->dev->nullb;
int err;
unsigned int len;
@@ -1298,63 +1245,21 @@ static int null_handle_rq(struct nullb_cmd *cmd)
return 0;
}
-static int null_handle_bio(struct nullb_cmd *cmd)
-{
- struct bio *bio = cmd->bio;
- struct nullb *nullb = cmd->nq->dev->nullb;
- int err;
- unsigned int len;
- sector_t sector = bio->bi_iter.bi_sector;
- struct bio_vec bvec;
- struct bvec_iter iter;
-
- spin_lock_irq(&nullb->lock);
- bio_for_each_segment(bvec, bio, iter) {
- len = bvec.bv_len;
- err = null_transfer(nullb, bvec.bv_page, len, bvec.bv_offset,
- op_is_write(bio_op(bio)), sector,
- bio->bi_opf & REQ_FUA);
- if (err) {
- spin_unlock_irq(&nullb->lock);
- return err;
- }
- sector += len >> SECTOR_SHIFT;
- }
- spin_unlock_irq(&nullb->lock);
- return 0;
-}
-
-static void null_stop_queue(struct nullb *nullb)
-{
- struct request_queue *q = nullb->q;
-
- if (nullb->dev->queue_mode == NULL_Q_MQ)
- blk_mq_stop_hw_queues(q);
-}
-
-static void null_restart_queue_async(struct nullb *nullb)
-{
- struct request_queue *q = nullb->q;
-
- if (nullb->dev->queue_mode == NULL_Q_MQ)
- blk_mq_start_stopped_hw_queues(q, true);
-}
-
static inline blk_status_t null_handle_throttled(struct nullb_cmd *cmd)
{
struct nullb_device *dev = cmd->nq->dev;
struct nullb *nullb = dev->nullb;
blk_status_t sts = BLK_STS_OK;
- struct request *rq = cmd->rq;
+ struct request *rq = blk_mq_rq_from_pdu(cmd);
if (!hrtimer_active(&nullb->bw_timer))
hrtimer_restart(&nullb->bw_timer);
if (atomic_long_sub_return(blk_rq_bytes(rq), &nullb->cur_bytes) < 0) {
- null_stop_queue(nullb);
+ blk_mq_stop_hw_queues(nullb->q);
/* race with timer */
if (atomic_long_read(&nullb->cur_bytes) > 0)
- null_restart_queue_async(nullb);
+ blk_mq_start_stopped_hw_queues(nullb->q, true);
/* requeue request */
sts = BLK_STS_DEV_RESOURCE;
}
@@ -1381,37 +1286,29 @@ static inline blk_status_t null_handle_memory_backed(struct nullb_cmd *cmd,
sector_t nr_sectors)
{
struct nullb_device *dev = cmd->nq->dev;
- int err;
if (op == REQ_OP_DISCARD)
return null_handle_discard(dev, sector, nr_sectors);
+ return errno_to_blk_status(null_handle_rq(cmd));
- if (dev->queue_mode == NULL_Q_BIO)
- err = null_handle_bio(cmd);
- else
- err = null_handle_rq(cmd);
-
- return errno_to_blk_status(err);
}
static void nullb_zero_read_cmd_buffer(struct nullb_cmd *cmd)
{
+ struct request *rq = blk_mq_rq_from_pdu(cmd);
struct nullb_device *dev = cmd->nq->dev;
struct bio *bio;
- if (dev->memory_backed)
- return;
-
- if (dev->queue_mode == NULL_Q_BIO && bio_op(cmd->bio) == REQ_OP_READ) {
- zero_fill_bio(cmd->bio);
- } else if (req_op(cmd->rq) == REQ_OP_READ) {
- __rq_for_each_bio(bio, cmd->rq)
+ if (!dev->memory_backed && req_op(rq) == REQ_OP_READ) {
+ __rq_for_each_bio(bio, rq)
zero_fill_bio(bio);
}
}
static inline void nullb_complete_cmd(struct nullb_cmd *cmd)
{
+ struct request *rq = blk_mq_rq_from_pdu(cmd);
+
/*
* Since root privileges are required to configure the null_blk
* driver, it is fine that this driver does not initialize the
@@ -1425,20 +1322,10 @@ static inline void nullb_complete_cmd(struct nullb_cmd *cmd)
/* Complete IO by inline, softirq or timer */
switch (cmd->nq->dev->irqmode) {
case NULL_IRQ_SOFTIRQ:
- switch (cmd->nq->dev->queue_mode) {
- case NULL_Q_MQ:
- blk_mq_complete_request(cmd->rq);
- break;
- case NULL_Q_BIO:
- /*
- * XXX: no proper submitting cpu information available.
- */
- end_cmd(cmd);
- break;
- }
+ blk_mq_complete_request(rq);
break;
case NULL_IRQ_NONE:
- end_cmd(cmd);
+ blk_mq_end_request(rq, cmd->error);
break;
case NULL_IRQ_TIMER:
null_cmd_end_timer(cmd);
@@ -1499,7 +1386,7 @@ static enum hrtimer_restart nullb_bwtimer_fn(struct hrtimer *timer)
return HRTIMER_NORESTART;
atomic_long_set(&nullb->cur_bytes, mb_per_tick(mbps));
- null_restart_queue_async(nullb);
+ blk_mq_start_stopped_hw_queues(nullb->q, true);
hrtimer_forward_now(&nullb->bw_timer, timer_interval);
@@ -1516,26 +1403,6 @@ static void nullb_setup_bwtimer(struct nullb *nullb)
hrtimer_start(&nullb->bw_timer, timer_interval, HRTIMER_MODE_REL);
}
-static struct nullb_queue *nullb_to_queue(struct nullb *nullb)
-{
- int index = 0;
-
- if (nullb->nr_queues != 1)
- index = raw_smp_processor_id() / ((nr_cpu_ids + nullb->nr_queues - 1) / nullb->nr_queues);
-
- return &nullb->queues[index];
-}
-
-static void null_submit_bio(struct bio *bio)
-{
- sector_t sector = bio->bi_iter.bi_sector;
- sector_t nr_sectors = bio_sectors(bio);
- struct nullb *nullb = bio->bi_bdev->bd_disk->private_data;
- struct nullb_queue *nq = nullb_to_queue(nullb);
-
- null_handle_cmd(alloc_cmd(nq, bio), sector, nr_sectors, bio_op(bio));
-}
-
#ifdef CONFIG_BLK_DEV_NULL_BLK_FAULT_INJECTION
static bool should_timeout_request(struct request *rq)
@@ -1655,7 +1522,7 @@ static int null_poll(struct blk_mq_hw_ctx *hctx, struct io_comp_batch *iob)
blk_rq_sectors(req));
if (!blk_mq_add_to_batch(req, iob, (__force int) cmd->error,
blk_mq_end_request_batch))
- end_cmd(cmd);
+ blk_mq_end_request(req, cmd->error);
nr++;
}
@@ -1711,7 +1578,6 @@ static blk_status_t null_queue_rq(struct blk_mq_hw_ctx *hctx,
hrtimer_init(&cmd->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
cmd->timer.function = null_cmd_timer_expired;
}
- cmd->rq = rq;
cmd->error = BLK_STS_OK;
cmd->nq = nq;
cmd->fake_timeout = should_timeout_request(rq) ||
@@ -1770,34 +1636,8 @@ static void null_queue_rqs(struct request **rqlist)
*rqlist = requeue_list;
}
-static void cleanup_queue(struct nullb_queue *nq)
-{
- bitmap_free(nq->tag_map);
- kfree(nq->cmds);
-}
-
-static void cleanup_queues(struct nullb *nullb)
-{
- int i;
-
- for (i = 0; i < nullb->nr_queues; i++)
- cleanup_queue(&nullb->queues[i]);
-
- kfree(nullb->queues);
-}
-
-static void null_exit_hctx(struct blk_mq_hw_ctx *hctx, unsigned int hctx_idx)
-{
- struct nullb_queue *nq = hctx->driver_data;
- struct nullb *nullb = nq->dev->nullb;
-
- nullb->nr_queues--;
-}
-
static void null_init_queue(struct nullb *nullb, struct nullb_queue *nq)
{
- init_waitqueue_head(&nq->wait);
- nq->queue_depth = nullb->queue_depth;
nq->dev = nullb->dev;
INIT_LIST_HEAD(&nq->poll_list);
spin_lock_init(&nq->poll_lock);
@@ -1815,7 +1655,6 @@ static int null_init_hctx(struct blk_mq_hw_ctx *hctx, void *driver_data,
nq = &nullb->queues[hctx_idx];
hctx->driver_data = nq;
null_init_queue(nullb, nq);
- nullb->nr_queues++;
return 0;
}
@@ -1828,7 +1667,6 @@ static const struct blk_mq_ops null_mq_ops = {
.poll = null_poll,
.map_queues = null_map_queues,
.init_hctx = null_init_hctx,
- .exit_hctx = null_exit_hctx,
};
static void null_del_dev(struct nullb *nullb)
@@ -1849,21 +1687,20 @@ static void null_del_dev(struct nullb *nullb)
if (test_bit(NULLB_DEV_FL_THROTTLED, &nullb->dev->flags)) {
hrtimer_cancel(&nullb->bw_timer);
atomic_long_set(&nullb->cur_bytes, LONG_MAX);
- null_restart_queue_async(nullb);
+ blk_mq_start_stopped_hw_queues(nullb->q, true);
}
put_disk(nullb->disk);
- if (dev->queue_mode == NULL_Q_MQ &&
- nullb->tag_set == &nullb->__tag_set)
+ if (nullb->tag_set == &nullb->__tag_set)
blk_mq_free_tag_set(nullb->tag_set);
- cleanup_queues(nullb);
+ kfree(nullb->queues);
if (null_cache_active(nullb))
null_free_device_storage(nullb->dev, true);
kfree(nullb);
dev->nullb = NULL;
}
-static void null_config_discard(struct nullb *nullb)
+static void null_config_discard(struct nullb *nullb, struct queue_limits *lim)
{
if (nullb->dev->discard == false)
return;
@@ -1880,43 +1717,14 @@ static void null_config_discard(struct nullb *nullb)
return;
}
- blk_queue_max_discard_sectors(nullb->q, UINT_MAX >> 9);
+ lim->max_hw_discard_sectors = UINT_MAX >> 9;
}
-static const struct block_device_operations null_bio_ops = {
- .owner = THIS_MODULE,
- .submit_bio = null_submit_bio,
- .report_zones = null_report_zones,
-};
-
-static const struct block_device_operations null_rq_ops = {
+static const struct block_device_operations null_ops = {
.owner = THIS_MODULE,
.report_zones = null_report_zones,
};
-static int setup_commands(struct nullb_queue *nq)
-{
- struct nullb_cmd *cmd;
- int i;
-
- nq->cmds = kcalloc(nq->queue_depth, sizeof(*cmd), GFP_KERNEL);
- if (!nq->cmds)
- return -ENOMEM;
-
- nq->tag_map = bitmap_zalloc(nq->queue_depth, GFP_KERNEL);
- if (!nq->tag_map) {
- kfree(nq->cmds);
- return -ENOMEM;
- }
-
- for (i = 0; i < nq->queue_depth; i++) {
- cmd = &nq->cmds[i];
- cmd->tag = -1U;
- }
-
- return 0;
-}
-
static int setup_queues(struct nullb *nullb)
{
int nqueues = nr_cpu_ids;
@@ -1929,101 +1737,66 @@ static int setup_queues(struct nullb *nullb)
if (!nullb->queues)
return -ENOMEM;
- nullb->queue_depth = nullb->dev->hw_queue_depth;
return 0;
}
-static int init_driver_queues(struct nullb *nullb)
+static int null_init_tag_set(struct blk_mq_tag_set *set, int poll_queues)
{
- struct nullb_queue *nq;
- int i, ret = 0;
-
- for (i = 0; i < nullb->dev->submit_queues; i++) {
- nq = &nullb->queues[i];
-
- null_init_queue(nullb, nq);
-
- ret = setup_commands(nq);
- if (ret)
- return ret;
- nullb->nr_queues++;
+ set->ops = &null_mq_ops;
+ set->cmd_size = sizeof(struct nullb_cmd);
+ set->timeout = 5 * HZ;
+ set->nr_maps = 1;
+ if (poll_queues) {
+ set->nr_hw_queues += poll_queues;
+ set->nr_maps += 2;
}
- return 0;
+ return blk_mq_alloc_tag_set(set);
}
-static int null_gendisk_register(struct nullb *nullb)
+static int null_init_global_tag_set(void)
{
- sector_t size = ((sector_t)nullb->dev->size * SZ_1M) >> SECTOR_SHIFT;
- struct gendisk *disk = nullb->disk;
-
- set_capacity(disk, size);
+ int error;
- disk->major = null_major;
- disk->first_minor = nullb->index;
- disk->minors = 1;
- if (queue_is_mq(nullb->q))
- disk->fops = &null_rq_ops;
- else
- disk->fops = &null_bio_ops;
- disk->private_data = nullb;
- strscpy_pad(disk->disk_name, nullb->disk_name, DISK_NAME_LEN);
-
- if (nullb->dev->zoned) {
- int ret = null_register_zoned_dev(nullb);
+ if (tag_set.ops)
+ return 0;
- if (ret)
- return ret;
- }
+ tag_set.nr_hw_queues = g_submit_queues;
+ tag_set.queue_depth = g_hw_queue_depth;
+ tag_set.numa_node = g_home_node;
+ tag_set.flags = BLK_MQ_F_SHOULD_MERGE;
+ if (g_no_sched)
+ tag_set.flags |= BLK_MQ_F_NO_SCHED;
+ if (g_shared_tag_bitmap)
+ tag_set.flags |= BLK_MQ_F_TAG_HCTX_SHARED;
+ if (g_blocking)
+ tag_set.flags |= BLK_MQ_F_BLOCKING;
- return add_disk(disk);
+ error = null_init_tag_set(&tag_set, g_poll_queues);
+ if (error)
+ tag_set.ops = NULL;
+ return error;
}
-static int null_init_tag_set(struct nullb *nullb, struct blk_mq_tag_set *set)
+static int null_setup_tagset(struct nullb *nullb)
{
- unsigned int flags = BLK_MQ_F_SHOULD_MERGE;
- int hw_queues, numa_node;
- unsigned int queue_depth;
- int poll_queues;
-
- if (nullb) {
- hw_queues = nullb->dev->submit_queues;
- poll_queues = nullb->dev->poll_queues;
- queue_depth = nullb->dev->hw_queue_depth;
- numa_node = nullb->dev->home_node;
- if (nullb->dev->no_sched)
- flags |= BLK_MQ_F_NO_SCHED;
- if (nullb->dev->shared_tag_bitmap)
- flags |= BLK_MQ_F_TAG_HCTX_SHARED;
- if (nullb->dev->blocking)
- flags |= BLK_MQ_F_BLOCKING;
- } else {
- hw_queues = g_submit_queues;
- poll_queues = g_poll_queues;
- queue_depth = g_hw_queue_depth;
- numa_node = g_home_node;
- if (g_no_sched)
- flags |= BLK_MQ_F_NO_SCHED;
- if (g_shared_tag_bitmap)
- flags |= BLK_MQ_F_TAG_HCTX_SHARED;
- if (g_blocking)
- flags |= BLK_MQ_F_BLOCKING;
+ if (nullb->dev->shared_tags) {
+ nullb->tag_set = &tag_set;
+ return null_init_global_tag_set();
}
- set->ops = &null_mq_ops;
- set->cmd_size = sizeof(struct nullb_cmd);
- set->flags = flags;
- set->driver_data = nullb;
- set->nr_hw_queues = hw_queues;
- set->queue_depth = queue_depth;
- set->numa_node = numa_node;
- if (poll_queues) {
- set->nr_hw_queues += poll_queues;
- set->nr_maps = 3;
- } else {
- set->nr_maps = 1;
- }
-
- return blk_mq_alloc_tag_set(set);
+ nullb->tag_set = &nullb->__tag_set;
+ nullb->tag_set->driver_data = nullb;
+ nullb->tag_set->nr_hw_queues = nullb->dev->submit_queues;
+ nullb->tag_set->queue_depth = nullb->dev->hw_queue_depth;
+ nullb->tag_set->numa_node = nullb->dev->home_node;
+ nullb->tag_set->flags = BLK_MQ_F_SHOULD_MERGE;
+ if (nullb->dev->no_sched)
+ nullb->tag_set->flags |= BLK_MQ_F_NO_SCHED;
+ if (nullb->dev->shared_tag_bitmap)
+ nullb->tag_set->flags |= BLK_MQ_F_TAG_HCTX_SHARED;
+ if (nullb->dev->blocking)
+ nullb->tag_set->flags |= BLK_MQ_F_BLOCKING;
+ return null_init_tag_set(nullb->tag_set, nullb->dev->poll_queues);
}
static int null_validate_conf(struct nullb_device *dev)
@@ -2032,11 +1805,15 @@ static int null_validate_conf(struct nullb_device *dev)
pr_err("legacy IO path is no longer available\n");
return -EINVAL;
}
+ if (dev->queue_mode == NULL_Q_BIO) {
+ pr_err("BIO-based IO path is no longer available, using blk-mq instead.\n");
+ dev->queue_mode = NULL_Q_MQ;
+ }
dev->blocksize = round_down(dev->blocksize, 512);
dev->blocksize = clamp_t(unsigned int, dev->blocksize, 512, 4096);
- if (dev->queue_mode == NULL_Q_MQ && dev->use_per_node_hctx) {
+ if (dev->use_per_node_hctx) {
if (dev->submit_queues != nr_online_nodes)
dev->submit_queues = nr_online_nodes;
} else if (dev->submit_queues > nr_cpu_ids)
@@ -2048,8 +1825,6 @@ static int null_validate_conf(struct nullb_device *dev)
if (dev->poll_queues > g_poll_queues)
dev->poll_queues = g_poll_queues;
dev->prev_poll_queues = dev->poll_queues;
-
- dev->queue_mode = min_t(unsigned int, dev->queue_mode, NULL_Q_MQ);
dev->irqmode = min_t(unsigned int, dev->irqmode, NULL_IRQ_TIMER);
/* Do memory allocation, so set blocking */
@@ -2060,9 +1835,6 @@ static int null_validate_conf(struct nullb_device *dev)
dev->cache_size = min_t(unsigned long, ULONG_MAX / 1024 / 1024,
dev->cache_size);
dev->mbps = min_t(unsigned int, 1024 * 40, dev->mbps);
- /* can not stop a queue */
- if (dev->queue_mode == NULL_Q_BIO)
- dev->mbps = 0;
if (dev->zoned &&
(!dev->zone_size || !is_power_of_2(dev->zone_size))) {
@@ -2102,6 +1874,12 @@ static bool null_setup_fault(void)
static int null_add_dev(struct nullb_device *dev)
{
+ struct queue_limits lim = {
+ .logical_block_size = dev->blocksize,
+ .physical_block_size = dev->blocksize,
+ .max_hw_sectors = dev->max_sectors,
+ };
+
struct nullb *nullb;
int rv;
@@ -2123,36 +1901,25 @@ static int null_add_dev(struct nullb_device *dev)
if (rv)
goto out_free_nullb;
- if (dev->queue_mode == NULL_Q_MQ) {
- if (shared_tags) {
- nullb->tag_set = &tag_set;
- rv = 0;
- } else {
- nullb->tag_set = &nullb->__tag_set;
- rv = null_init_tag_set(nullb, nullb->tag_set);
- }
+ rv = null_setup_tagset(nullb);
+ if (rv)
+ goto out_cleanup_queues;
+ if (dev->virt_boundary)
+ lim.virt_boundary_mask = PAGE_SIZE - 1;
+ null_config_discard(nullb, &lim);
+ if (dev->zoned) {
+ rv = null_init_zoned_dev(dev, &lim);
if (rv)
- goto out_cleanup_queues;
-
- nullb->tag_set->timeout = 5 * HZ;
- nullb->disk = blk_mq_alloc_disk(nullb->tag_set, nullb);
- if (IS_ERR(nullb->disk)) {
- rv = PTR_ERR(nullb->disk);
goto out_cleanup_tags;
- }
- nullb->q = nullb->disk->queue;
- } else if (dev->queue_mode == NULL_Q_BIO) {
- rv = -ENOMEM;
- nullb->disk = blk_alloc_disk(nullb->dev->home_node);
- if (!nullb->disk)
- goto out_cleanup_queues;
+ }
- nullb->q = nullb->disk->queue;
- rv = init_driver_queues(nullb);
- if (rv)
- goto out_cleanup_disk;
+ nullb->disk = blk_mq_alloc_disk(nullb->tag_set, &lim, nullb);
+ if (IS_ERR(nullb->disk)) {
+ rv = PTR_ERR(nullb->disk);
+ goto out_cleanup_zone;
}
+ nullb->q = nullb->disk->queue;
if (dev->mbps) {
set_bit(NULLB_DEV_FL_THROTTLED, &dev->flags);
@@ -2164,34 +1931,15 @@ static int null_add_dev(struct nullb_device *dev)
blk_queue_write_cache(nullb->q, true, true);
}
- if (dev->zoned) {
- rv = null_init_zoned_dev(dev, nullb->q);
- if (rv)
- goto out_cleanup_disk;
- }
-
nullb->q->queuedata = nullb;
blk_queue_flag_set(QUEUE_FLAG_NONROT, nullb->q);
- mutex_lock(&lock);
rv = ida_alloc(&nullb_indexes, GFP_KERNEL);
- if (rv < 0) {
- mutex_unlock(&lock);
- goto out_cleanup_zone;
- }
+ if (rv < 0)
+ goto out_cleanup_disk;
+
nullb->index = rv;
dev->index = rv;
- mutex_unlock(&lock);
-
- blk_queue_logical_block_size(nullb->q, dev->blocksize);
- blk_queue_physical_block_size(nullb->q, dev->blocksize);
- if (dev->max_sectors)
- blk_queue_max_hw_sectors(nullb->q, dev->max_sectors);
-
- if (dev->virt_boundary)
- blk_queue_virt_boundary(nullb->q, PAGE_SIZE - 1);
-
- null_config_discard(nullb);
if (config_item_name(&dev->group.cg_item)) {
/* Use configfs dir name as the device name */
@@ -2201,13 +1949,26 @@ static int null_add_dev(struct nullb_device *dev)
sprintf(nullb->disk_name, "nullb%d", nullb->index);
}
- rv = null_gendisk_register(nullb);
+ set_capacity(nullb->disk,
+ ((sector_t)nullb->dev->size * SZ_1M) >> SECTOR_SHIFT);
+ nullb->disk->major = null_major;
+ nullb->disk->first_minor = nullb->index;
+ nullb->disk->minors = 1;
+ nullb->disk->fops = &null_ops;
+ nullb->disk->private_data = nullb;
+ strscpy_pad(nullb->disk->disk_name, nullb->disk_name, DISK_NAME_LEN);
+
+ if (nullb->dev->zoned) {
+ rv = null_register_zoned_dev(nullb);
+ if (rv)
+ goto out_ida_free;
+ }
+
+ rv = add_disk(nullb->disk);
if (rv)
goto out_ida_free;
- mutex_lock(&lock);
list_add_tail(&nullb->list, &nullb_list);
- mutex_unlock(&lock);
pr_info("disk %s created\n", nullb->disk_name);
@@ -2215,15 +1976,15 @@ static int null_add_dev(struct nullb_device *dev)
out_ida_free:
ida_free(&nullb_indexes, nullb->index);
-out_cleanup_zone:
- null_free_zoned_dev(dev);
out_cleanup_disk:
put_disk(nullb->disk);
+out_cleanup_zone:
+ null_free_zoned_dev(dev);
out_cleanup_tags:
- if (dev->queue_mode == NULL_Q_MQ && nullb->tag_set == &nullb->__tag_set)
+ if (nullb->tag_set == &nullb->__tag_set)
blk_mq_free_tag_set(nullb->tag_set);
out_cleanup_queues:
- cleanup_queues(nullb);
+ kfree(nullb->queues);
out_free_nullb:
kfree(nullb);
dev->nullb = NULL;
@@ -2256,7 +2017,9 @@ static int null_create_dev(void)
if (!dev)
return -ENOMEM;
+ mutex_lock(&lock);
ret = null_add_dev(dev);
+ mutex_unlock(&lock);
if (ret) {
null_free_dev(dev);
return ret;
@@ -2299,7 +2062,7 @@ static int __init null_init(void)
return -EINVAL;
}
- if (g_queue_mode == NULL_Q_MQ && g_use_per_node_hctx) {
+ if (g_use_per_node_hctx) {
if (g_submit_queues != nr_online_nodes) {
pr_warn("submit_queues param is set to %u.\n",
nr_online_nodes);
@@ -2311,18 +2074,12 @@ static int __init null_init(void)
g_submit_queues = 1;
}
- if (g_queue_mode == NULL_Q_MQ && shared_tags) {
- ret = null_init_tag_set(NULL, &tag_set);
- if (ret)
- return ret;
- }
-
config_group_init(&nullb_subsys.su_group);
mutex_init(&nullb_subsys.su_mutex);
ret = configfs_register_subsystem(&nullb_subsys);
if (ret)
- goto err_tagset;
+ return ret;
mutex_init(&lock);
@@ -2349,9 +2106,6 @@ err_dev:
unregister_blkdev(null_major, "nullb");
err_conf:
configfs_unregister_subsystem(&nullb_subsys);
-err_tagset:
- if (g_queue_mode == NULL_Q_MQ && shared_tags)
- blk_mq_free_tag_set(&tag_set);
return ret;
}
@@ -2370,7 +2124,7 @@ static void __exit null_exit(void)
}
mutex_unlock(&lock);
- if (g_queue_mode == NULL_Q_MQ && shared_tags)
+ if (tag_set.ops)
blk_mq_free_tag_set(&tag_set);
mutex_destroy(&lock);
@@ -2380,4 +2134,5 @@ module_init(null_init);
module_exit(null_exit);
MODULE_AUTHOR("Jens Axboe <axboe@kernel.dk>");
+MODULE_DESCRIPTION("multi queue aware block test driver");
MODULE_LICENSE("GPL");
diff --git a/drivers/block/null_blk/null_blk.h b/drivers/block/null_blk/null_blk.h
index 929f659dd2..477b977468 100644
--- a/drivers/block/null_blk/null_blk.h
+++ b/drivers/block/null_blk/null_blk.h
@@ -16,11 +16,6 @@
#include <linux/mutex.h>
struct nullb_cmd {
- union {
- struct request *rq;
- struct bio *bio;
- };
- unsigned int tag;
blk_status_t error;
bool fake_timeout;
struct nullb_queue *nq;
@@ -28,16 +23,11 @@ struct nullb_cmd {
};
struct nullb_queue {
- unsigned long *tag_map;
- wait_queue_head_t wait;
- unsigned int queue_depth;
struct nullb_device *dev;
unsigned int requeue_selection;
struct list_head poll_list;
spinlock_t poll_lock;
-
- struct nullb_cmd *cmds;
};
struct nullb_zone {
@@ -60,13 +50,6 @@ struct nullb_zone {
unsigned int capacity;
};
-/* Queue modes */
-enum {
- NULL_Q_BIO = 0,
- NULL_Q_RQ = 1,
- NULL_Q_MQ = 2,
-};
-
struct nullb_device {
struct nullb *nullb;
struct config_group group;
@@ -119,6 +102,7 @@ struct nullb_device {
bool zoned; /* if device is zoned */
bool virt_boundary; /* virtual boundary on/off for the device */
bool no_sched; /* no IO scheduler for the device */
+ bool shared_tags; /* share tag set between devices for blk-mq */
bool shared_tag_bitmap; /* use hostwide shared tags */
};
@@ -130,14 +114,12 @@ struct nullb {
struct gendisk *disk;
struct blk_mq_tag_set *tag_set;
struct blk_mq_tag_set __tag_set;
- unsigned int queue_depth;
atomic_long_t cur_bytes;
struct hrtimer bw_timer;
unsigned long cache_flush_pos;
spinlock_t lock;
struct nullb_queue *queues;
- unsigned int nr_queues;
char disk_name[DISK_NAME_LEN];
};
@@ -147,7 +129,7 @@ blk_status_t null_process_cmd(struct nullb_cmd *cmd, enum req_op op,
sector_t sector, unsigned int nr_sectors);
#ifdef CONFIG_BLK_DEV_ZONED
-int null_init_zoned_dev(struct nullb_device *dev, struct request_queue *q);
+int null_init_zoned_dev(struct nullb_device *dev, struct queue_limits *lim);
int null_register_zoned_dev(struct nullb *nullb);
void null_free_zoned_dev(struct nullb_device *dev);
int null_report_zones(struct gendisk *disk, sector_t sector,
@@ -160,7 +142,7 @@ ssize_t zone_cond_store(struct nullb_device *dev, const char *page,
size_t count, enum blk_zone_cond cond);
#else
static inline int null_init_zoned_dev(struct nullb_device *dev,
- struct request_queue *q)
+ struct queue_limits *lim)
{
pr_err("CONFIG_BLK_DEV_ZONED not enabled\n");
return -EINVAL;
diff --git a/drivers/block/null_blk/trace.h b/drivers/block/null_blk/trace.h
index 6b2b370e78..ef2d05d5f0 100644
--- a/drivers/block/null_blk/trace.h
+++ b/drivers/block/null_blk/trace.h
@@ -41,10 +41,11 @@ TRACE_EVENT(nullb_zone_op,
__field(unsigned int, zone_cond)
),
TP_fast_assign(
- __entry->op = req_op(cmd->rq);
+ __entry->op = req_op(blk_mq_rq_from_pdu(cmd));
__entry->zone_no = zone_no;
__entry->zone_cond = zone_cond;
- __assign_disk_name(__entry->disk, cmd->rq->q->disk);
+ __assign_disk_name(__entry->disk,
+ blk_mq_rq_from_pdu(cmd)->q->disk);
),
TP_printk("%s req=%-15s zone_no=%u zone_cond=%-10s",
__print_disk_name(__entry->disk),
diff --git a/drivers/block/null_blk/zoned.c b/drivers/block/null_blk/zoned.c
index 6f5e099486..27928deccc 100644
--- a/drivers/block/null_blk/zoned.c
+++ b/drivers/block/null_blk/zoned.c
@@ -58,7 +58,8 @@ static inline void null_unlock_zone(struct nullb_device *dev,
mutex_unlock(&zone->mutex);
}
-int null_init_zoned_dev(struct nullb_device *dev, struct request_queue *q)
+int null_init_zoned_dev(struct nullb_device *dev,
+ struct queue_limits *lim)
{
sector_t dev_capacity_sects, zone_capacity_sects;
struct nullb_zone *zone;
@@ -112,7 +113,7 @@ int null_init_zoned_dev(struct nullb_device *dev, struct request_queue *q)
if (dev->zone_max_active && dev->zone_max_open > dev->zone_max_active) {
dev->zone_max_open = dev->zone_max_active;
pr_info("changed the maximum number of open zones to %u\n",
- dev->nr_zones);
+ dev->zone_max_open);
} else if (dev->zone_max_open >= dev->nr_zones - dev->zone_nr_conv) {
dev->zone_max_open = 0;
pr_info("zone_max_open limit disabled, limit >= zone count\n");
@@ -151,27 +152,22 @@ int null_init_zoned_dev(struct nullb_device *dev, struct request_queue *q)
sector += dev->zone_size_sects;
}
+ lim->zoned = true;
+ lim->chunk_sectors = dev->zone_size_sects;
+ lim->max_zone_append_sectors = dev->zone_size_sects;
+ lim->max_open_zones = dev->zone_max_open;
+ lim->max_active_zones = dev->zone_max_active;
return 0;
}
int null_register_zoned_dev(struct nullb *nullb)
{
- struct nullb_device *dev = nullb->dev;
struct request_queue *q = nullb->q;
- disk_set_zoned(nullb->disk);
blk_queue_flag_set(QUEUE_FLAG_ZONE_RESETALL, q);
blk_queue_required_elevator_features(q, ELEVATOR_F_ZBD_SEQ_WRITE);
- blk_queue_chunk_sectors(q, dev->zone_size_sects);
nullb->disk->nr_zones = bdev_nr_zones(nullb->disk->part0);
- blk_queue_max_zone_append_sectors(q, dev->zone_size_sects);
- disk_set_max_open_zones(nullb->disk, dev->zone_max_open);
- disk_set_max_active_zones(nullb->disk, dev->zone_max_active);
-
- if (queue_is_mq(q))
- return blk_revalidate_disk_zones(nullb->disk, NULL);
-
- return 0;
+ return blk_revalidate_disk_zones(nullb->disk, NULL);
}
void null_free_zoned_dev(struct nullb_device *dev)
@@ -394,10 +390,7 @@ static blk_status_t null_zone_write(struct nullb_cmd *cmd, sector_t sector,
*/
if (append) {
sector = zone->wp;
- if (dev->queue_mode == NULL_Q_MQ)
- cmd->rq->__sector = sector;
- else
- cmd->bio->bi_iter.bi_sector = sector;
+ blk_mq_rq_from_pdu(cmd)->__sector = sector;
} else if (sector != zone->wp) {
ret = BLK_STS_IOERR;
goto unlock;