[PATCH v2] blk-mq: add tracepoint block_rq_tag_wait

[PATCH v2] blk-mq: add tracepoint block_rq_tag_wait

[Aaron Tomlin]

In high-performance storage environments, particularly when utilising
RAID controllers with shared tag sets (BLK_MQ_F_TAG_HCTX_SHARED), severe
latency spikes can occur when fast devices (SSDs) are starved of hardware
tags when sharing the same blk_mq_tag_set.

Currently, diagnosing this specific hardware queue contention is
difficult. When a CPU thread exhausts the tag pool, blk_mq_get_tag()
forces the current thread to block uninterruptible via io_schedule().
While this can be inferred via sched:sched_switch or dynamically
traced by attaching a kprobe to blk_mq_mark_tag_wait(), there is no
dedicated, out-of-the-box observability for this event.

This patch introduces the block_rq_tag_wait static trace point in the
tag allocation slow-path. It triggers immediately before the thread
yields the CPU, exposing the exact hardware context (hctx) that is
starved, the specific pool experiencing starvation (hardware or software
scheduler), and the total pool depth.

This provides storage engineers and performance monitoring agents
with a zero-configuration, low-overhead mechanism to definitively
identify shared-tag bottlenecks and tune I/O schedulers or cgroup
throttling accordingly.

Signed-off-by: Aaron Tomlin <atomlin@atomlin.com>
---
Changes in v1 [1]:
 - Improved the description of the trace point (Damien Le Moal)
 - Removed the redundant "active requests" (Laurence Oberman)
 - Introduced pool-specific starvation tracking

[1]: https://lore.kernel.org/lkml/20260317182835.258183-1-atomlin@atomlin.com/

 block/blk-mq-tag.c           |  4 ++++
 include/trace/events/block.h | 43 ++++++++++++++++++++++++++++++++++++
 2 files changed, 47 insertions(+)

diff --git a/block/blk-mq-tag.c b/block/blk-mq-tag.c
index 33946cdb5716..a6691a4fe7a7 100644
--- a/block/blk-mq-tag.c
+++ b/block/blk-mq-tag.c
@@ -13,6 +13,7 @@
 #include <linux/kmemleak.h>
 
 #include <linux/delay.h>
+#include <trace/events/block.h>
 #include "blk.h"
 #include "blk-mq.h"
 #include "blk-mq-sched.h"
@@ -187,6 +188,9 @@ unsigned int blk_mq_get_tag(struct blk_mq_alloc_data *data)
 		if (tag != BLK_MQ_NO_TAG)
 			break;
 
+		trace_block_rq_tag_wait(data->q, data->hctx,
+					!!(data->rq_flags & RQF_SCHED_TAGS));
+
 		bt_prev = bt;
 		io_schedule();
 
diff --git a/include/trace/events/block.h b/include/trace/events/block.h
index 6aa79e2d799c..f7708d0d7a0c 100644
--- a/include/trace/events/block.h
+++ b/include/trace/events/block.h
@@ -226,6 +226,49 @@ DECLARE_EVENT_CLASS(block_rq,
 		  IOPRIO_PRIO_LEVEL(__entry->ioprio), __entry->comm)
 );
 
+/**
+ * block_rq_tag_wait - triggered when a request is starved of a tag
+ * @q: request queue of the target device
+ * @hctx: hardware context of the request experiencing starvation
+ * @is_sched_tag: indicates whether the starved pool is the software scheduler
+ *
+ * Called immediately before the submitting context is forced to block due
+ * to the exhaustion of available tags (i.e., physical hardware driver tags
+ * or software scheduler tags). This trace point indicates that the context
+ * will be placed into an uninterruptible state via io_schedule() until an
+ * active request completes and relinquishes its assigned tag.
+ */
+TRACE_EVENT(block_rq_tag_wait,
+
+	TP_PROTO(struct request_queue *q, struct blk_mq_hw_ctx *hctx, bool is_sched_tag),
+
+	TP_ARGS(q, hctx, is_sched_tag),
+
+	TP_STRUCT__entry(
+		__field( dev_t,		dev			)
+		__field( u32,		hctx_id			)
+		__field( u32,		nr_tags			)
+		__field( bool,		is_sched_tag		)
+	),
+
+	TP_fast_assign(
+		__entry->dev		= disk_devt(q->disk);
+		__entry->hctx_id	= hctx->queue_num;
+		__entry->is_sched_tag	= is_sched_tag;
+
+		if (__entry->is_sched_tag)
+			__entry->nr_tags = hctx->sched_tags->nr_tags;
+		else
+			__entry->nr_tags = hctx->tags->nr_tags;
+	),
+
+	TP_printk("%d,%d hctx=%u starved on %s tags (depth=%u)",
+		  MAJOR(__entry->dev), MINOR(__entry->dev),
+		  __entry->hctx_id,
+		  __entry->is_sched_tag ? "scheduler" : "hardware",
+		  __entry->nr_tags)
+);
+
 /**
  * block_rq_insert - insert block operation request into queue
  * @rq: block IO operation request
-- 
2.51.0

Re: [PATCH v2] blk-mq: add tracepoint block_rq_tag_wait

[Chaitanya Kulkarni]

On 3/18/26 18:53, Aaron Tomlin wrote:
> In high-performance storage environments, particularly when utilising
> RAID controllers with shared tag sets (BLK_MQ_F_TAG_HCTX_SHARED), severe
> latency spikes can occur when fast devices (SSDs) are starved of hardware
> tags when sharing the same blk_mq_tag_set.
>
> Currently, diagnosing this specific hardware queue contention is
> difficult. When a CPU thread exhausts the tag pool, blk_mq_get_tag()
> forces the current thread to block uninterruptible via io_schedule().
> While this can be inferred viasched:sched_switch or dynamically
> traced by attaching a kprobe to blk_mq_mark_tag_wait(), there is no
> dedicated, out-of-the-box observability for this event.
>
> This patch introduces the block_rq_tag_wait static trace point in the
> tag allocation slow-path. It triggers immediately before the thread
> yields the CPU, exposing the exact hardware context (hctx) that is
> starved, the specific pool experiencing starvation (hardware or software
> scheduler), and the total pool depth.
>
> This provides storage engineers and performance monitoring agents
> with a zero-configuration, low-overhead mechanism to definitively
> identify shared-tag bottlenecks and tune I/O schedulers or cgroup
> throttling accordingly.
>
> Signed-off-by: Aaron Tomlin<atomlin@atomlin.com>
> ---
> Changes in v1 [1]:
>   - Improved the description of the trace point (Damien Le Moal)
>   - Removed the redundant "active requests" (Laurence Oberman)
>   - Introduced pool-specific starvation tracking
>
> [1]:https://lore.kernel.org/lkml/20260317182835.258183-1-atomlin@atomlin.com/


LGTM.

Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>

-ck

Re: [PATCH v2] blk-mq: add tracepoint block_rq_tag_wait

[Damien Le Moal]

On 3/19/26 10:53, Aaron Tomlin wrote:
> In high-performance storage environments, particularly when utilising
> RAID controllers with shared tag sets (BLK_MQ_F_TAG_HCTX_SHARED), severe
> latency spikes can occur when fast devices (SSDs) are starved of hardware
> tags when sharing the same blk_mq_tag_set.
> 
> Currently, diagnosing this specific hardware queue contention is
> difficult. When a CPU thread exhausts the tag pool, blk_mq_get_tag()
> forces the current thread to block uninterruptible via io_schedule().
> While this can be inferred via sched:sched_switch or dynamically
> traced by attaching a kprobe to blk_mq_mark_tag_wait(), there is no
> dedicated, out-of-the-box observability for this event.
> 
> This patch introduces the block_rq_tag_wait static trace point in the
> tag allocation slow-path. It triggers immediately before the thread
> yields the CPU, exposing the exact hardware context (hctx) that is
> starved, the specific pool experiencing starvation (hardware or software
> scheduler), and the total pool depth.
> 
> This provides storage engineers and performance monitoring agents
> with a zero-configuration, low-overhead mechanism to definitively
> identify shared-tag bottlenecks and tune I/O schedulers or cgroup
> throttling accordingly.
> 
> Signed-off-by: Aaron Tomlin <atomlin@atomlin.com>
> ---
> Changes in v1 [1]:
>  - Improved the description of the trace point (Damien Le Moal)
>  - Removed the redundant "active requests" (Laurence Oberman)
>  - Introduced pool-specific starvation tracking
> 
> [1]: https://lore.kernel.org/lkml/20260317182835.258183-1-atomlin@atomlin.com/
> 
>  block/blk-mq-tag.c           |  4 ++++
>  include/trace/events/block.h | 43 ++++++++++++++++++++++++++++++++++++
>  2 files changed, 47 insertions(+)
> 
> diff --git a/block/blk-mq-tag.c b/block/blk-mq-tag.c
> index 33946cdb5716..a6691a4fe7a7 100644
> --- a/block/blk-mq-tag.c
> +++ b/block/blk-mq-tag.c
> @@ -13,6 +13,7 @@
>  #include <linux/kmemleak.h>
>  
>  #include <linux/delay.h>
> +#include <trace/events/block.h>
>  #include "blk.h"
>  #include "blk-mq.h"
>  #include "blk-mq-sched.h"
> @@ -187,6 +188,9 @@ unsigned int blk_mq_get_tag(struct blk_mq_alloc_data *data)
>  		if (tag != BLK_MQ_NO_TAG)
>  			break;
>  
> +		trace_block_rq_tag_wait(data->q, data->hctx,
> +					!!(data->rq_flags & RQF_SCHED_TAGS));

I do not think that the "!!" is needed here.

Other than this, this looks OK to me.

Reviewed-by: Damien Le Moal <dlemoal@kernel.org>

> +
>  		bt_prev = bt;
>  		io_schedule();
>  
> diff --git a/include/trace/events/block.h b/include/trace/events/block.h
> index 6aa79e2d799c..f7708d0d7a0c 100644
> --- a/include/trace/events/block.h
> +++ b/include/trace/events/block.h
> @@ -226,6 +226,49 @@ DECLARE_EVENT_CLASS(block_rq,
>  		  IOPRIO_PRIO_LEVEL(__entry->ioprio), __entry->comm)
>  );
>  
> +/**
> + * block_rq_tag_wait - triggered when a request is starved of a tag
> + * @q: request queue of the target device
> + * @hctx: hardware context of the request experiencing starvation
> + * @is_sched_tag: indicates whether the starved pool is the software scheduler
> + *
> + * Called immediately before the submitting context is forced to block due
> + * to the exhaustion of available tags (i.e., physical hardware driver tags
> + * or software scheduler tags). This trace point indicates that the context
> + * will be placed into an uninterruptible state via io_schedule() until an
> + * active request completes and relinquishes its assigned tag.
> + */
> +TRACE_EVENT(block_rq_tag_wait,
> +
> +	TP_PROTO(struct request_queue *q, struct blk_mq_hw_ctx *hctx, bool is_sched_tag),
> +
> +	TP_ARGS(q, hctx, is_sched_tag),
> +
> +	TP_STRUCT__entry(
> +		__field( dev_t,		dev			)
> +		__field( u32,		hctx_id			)
> +		__field( u32,		nr_tags			)
> +		__field( bool,		is_sched_tag		)
> +	),
> +
> +	TP_fast_assign(
> +		__entry->dev		= disk_devt(q->disk);
> +		__entry->hctx_id	= hctx->queue_num;
> +		__entry->is_sched_tag	= is_sched_tag;
> +
> +		if (__entry->is_sched_tag)
> +			__entry->nr_tags = hctx->sched_tags->nr_tags;
> +		else
> +			__entry->nr_tags = hctx->tags->nr_tags;
> +	),
> +
> +	TP_printk("%d,%d hctx=%u starved on %s tags (depth=%u)",
> +		  MAJOR(__entry->dev), MINOR(__entry->dev),
> +		  __entry->hctx_id,
> +		  __entry->is_sched_tag ? "scheduler" : "hardware",
> +		  __entry->nr_tags)
> +);
> +
>  /**
>   * block_rq_insert - insert block operation request into queue
>   * @rq: block IO operation request


-- 
Damien Le Moal
Western Digital Research

Re: [PATCH v2] blk-mq: add tracepoint block_rq_tag_wait

[Johannes Thumshirn]

Looks good,

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>

Re: [PATCH v2] blk-mq: add tracepoint block_rq_tag_wait

[Laurence Oberman]

On Thu, 2026-03-19 at 07:32 +0000, Johannes Thumshirn wrote:
> Looks good,
> 
> Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>

Looks good now.

Reviewed-by: Laurence Oberman <loberman@redhat.com>
Tested-by:   Laurence Oberman <loberman@redhat.com>

Re: [PATCH v2] blk-mq: add tracepoint block_rq_tag_wait

[Steven Rostedt]

On Wed, 18 Mar 2026 21:53:00 -0400
Aaron Tomlin <atomlin@atomlin.com> wrote:

> +	TP_fast_assign(
> +		__entry->dev		= disk_devt(q->disk);
> +		__entry->hctx_id	= hctx->queue_num;
> +		__entry->is_sched_tag	= is_sched_tag;
> +
> +		if (__entry->is_sched_tag)

Nit, but why use __entry->is_sched_tag instead of is_sched_tag.

Not sure if the compiler will optimize it (likely it will), but it seems
cleaner to use the variable directly and not the one assigned.

Perhaps the compiler is smart enough to use one register for both updates.

-- Steve


> +			__entry->nr_tags = hctx->sched_tags->nr_tags;
> +		else
> +			__entry->nr_tags = hctx->tags->nr_tags;
> +	),
> +

Re: [PATCH v2] blk-mq: add tracepoint block_rq_tag_wait

[Aaron Tomlin]

[-- Attachment #1: Type: text/plain, Size: 783 bytes --]

On Thu, Mar 19, 2026 at 05:52:49PM -0400, Steven Rostedt wrote:
> On Wed, 18 Mar 2026 21:53:00 -0400
> Aaron Tomlin <atomlin@atomlin.com> wrote:
> 
> > +	TP_fast_assign(
> > +		__entry->dev		= disk_devt(q->disk);
> > +		__entry->hctx_id	= hctx->queue_num;
> > +		__entry->is_sched_tag	= is_sched_tag;
> > +
> > +		if (__entry->is_sched_tag)
> 
> Nit, but why use __entry->is_sched_tag instead of is_sched_tag.
> 
> Not sure if the compiler will optimize it (likely it will), but it seems
> cleaner to use the variable directly and not the one assigned.
> 
> Perhaps the compiler is smart enough to use one register for both updates.
> 
Hi Steve,

Thank you for your feedback.

That was an oversight - I'll correct it now.


Kind regards,
-- 
Aaron Tomlin

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]