[PATCH v4 26/28] cxl/mem: Trace Dynamic capacity Event Record

Ira Weiny posted 28 patches 1 month, 2 weeks ago
There is a newer version of this series
[PATCH v4 26/28] cxl/mem: Trace Dynamic capacity Event Record
Posted by ira.weiny@intel.com 1 month, 2 weeks ago
From: Navneet Singh <navneet.singh@intel.com>

CXL rev 3.1 section 8.2.9.2.1 adds the Dynamic Capacity Event Records.
User space can use trace events for debugging of DC capacity changes.

Add DC trace points to the trace log.

Signed-off-by: Navneet Singh <navneet.singh@intel.com>
Signed-off-by: Ira Weiny <ira.weiny@intel.com>

---
Changes:
[djiang: Use 3.1 spec reference]
---
 drivers/cxl/core/mbox.c  |  4 +++
 drivers/cxl/core/trace.h | 65 ++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 69 insertions(+)

diff --git a/drivers/cxl/core/mbox.c b/drivers/cxl/core/mbox.c
index 6b25d15403a3..816e28cc5a40 100644
--- a/drivers/cxl/core/mbox.c
+++ b/drivers/cxl/core/mbox.c
@@ -994,6 +994,10 @@ static void __cxl_event_trace_record(const struct cxl_memdev *cxlmd,
 		ev_type = CXL_CPER_EVENT_DRAM;
 	else if (uuid_equal(uuid, &CXL_EVENT_MEM_MODULE_UUID))
 		ev_type = CXL_CPER_EVENT_MEM_MODULE;
+	else if (uuid_equal(uuid, &CXL_EVENT_DC_EVENT_UUID)) {
+		trace_cxl_dynamic_capacity(cxlmd, type, &record->event.dcd);
+		return;
+	}
 
 	cxl_event_trace_record(cxlmd, type, ev_type, uuid, &record->event);
 }
diff --git a/drivers/cxl/core/trace.h b/drivers/cxl/core/trace.h
index 9167cfba7f59..1303024b5239 100644
--- a/drivers/cxl/core/trace.h
+++ b/drivers/cxl/core/trace.h
@@ -731,6 +731,71 @@ TRACE_EVENT(cxl_poison,
 	)
 );
 
+/*
+ * Dynamic Capacity Event Record - DER
+ *
+ * CXL rev 3.1 section 8.2.9.2.1.6 Table 8-50
+ */
+
+#define CXL_DC_ADD_CAPACITY			0x00
+#define CXL_DC_REL_CAPACITY			0x01
+#define CXL_DC_FORCED_REL_CAPACITY		0x02
+#define CXL_DC_REG_CONF_UPDATED			0x03
+#define show_dc_evt_type(type)	__print_symbolic(type,		\
+	{ CXL_DC_ADD_CAPACITY,	"Add capacity"},		\
+	{ CXL_DC_REL_CAPACITY,	"Release capacity"},		\
+	{ CXL_DC_FORCED_REL_CAPACITY,	"Forced capacity release"},	\
+	{ CXL_DC_REG_CONF_UPDATED,	"Region Configuration Updated"	} \
+)
+
+TRACE_EVENT(cxl_dynamic_capacity,
+
+	TP_PROTO(const struct cxl_memdev *cxlmd, enum cxl_event_log_type log,
+		 struct cxl_event_dcd *rec),
+
+	TP_ARGS(cxlmd, log, rec),
+
+	TP_STRUCT__entry(
+		CXL_EVT_TP_entry
+
+		/* Dynamic capacity Event */
+		__field(u8, event_type)
+		__field(u16, hostid)
+		__field(u8, region_id)
+		__field(u64, dpa_start)
+		__field(u64, length)
+		__array(u8, tag, CXL_EXTENT_TAG_LEN)
+		__field(u16, sh_extent_seq)
+	),
+
+	TP_fast_assign(
+		CXL_EVT_TP_fast_assign(cxlmd, log, rec->hdr);
+
+		/* Dynamic_capacity Event */
+		__entry->event_type = rec->event_type;
+
+		/* DCD event record data */
+		__entry->hostid = le16_to_cpu(rec->host_id);
+		__entry->region_id = rec->region_index;
+		__entry->dpa_start = le64_to_cpu(rec->extent.start_dpa);
+		__entry->length = le64_to_cpu(rec->extent.length);
+		memcpy(__entry->tag, &rec->extent.tag, CXL_EXTENT_TAG_LEN);
+		__entry->sh_extent_seq = le16_to_cpu(rec->extent.shared_extn_seq);
+	),
+
+	CXL_EVT_TP_printk("event_type='%s' host_id='%d' region_id='%d' " \
+		"starting_dpa=%llx length=%llx tag=%s " \
+		"shared_extent_sequence=%d",
+		show_dc_evt_type(__entry->event_type),
+		__entry->hostid,
+		__entry->region_id,
+		__entry->dpa_start,
+		__entry->length,
+		__print_hex(__entry->tag, CXL_EXTENT_TAG_LEN),
+		__entry->sh_extent_seq
+	)
+);
+
 #endif /* _CXL_EVENTS_H */
 
 #define TRACE_INCLUDE_FILE trace

-- 
2.46.0
Re: [PATCH v4 26/28] cxl/mem: Trace Dynamic capacity Event Record
Posted by Jonathan Cameron 1 month, 2 weeks ago
On Mon, 07 Oct 2024 18:16:32 -0500
ira.weiny@intel.com wrote:

> From: Navneet Singh <navneet.singh@intel.com>
> 
> CXL rev 3.1 section 8.2.9.2.1 adds the Dynamic Capacity Event Records.
> User space can use trace events for debugging of DC capacity changes.
> 
> Add DC trace points to the trace log.
> 
> Signed-off-by: Navneet Singh <navneet.singh@intel.com>
> Signed-off-by: Ira Weiny <ira.weiny@intel.com>
Minor comment inline about tag formatting.

Either way
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>

> 
> ---
> Changes:
> [djiang: Use 3.1 spec reference]
> ---
>  drivers/cxl/core/mbox.c  |  4 +++
>  drivers/cxl/core/trace.h | 65 ++++++++++++++++++++++++++++++++++++++++++++++++
>  2 files changed, 69 insertions(+)
> 
> diff --git a/drivers/cxl/core/mbox.c b/drivers/cxl/core/mbox.c
> index 6b25d15403a3..816e28cc5a40 100644
> --- a/drivers/cxl/core/mbox.c
> +++ b/drivers/cxl/core/mbox.c
> @@ -994,6 +994,10 @@ static void __cxl_event_trace_record(const struct cxl_memdev *cxlmd,
>  		ev_type = CXL_CPER_EVENT_DRAM;
>  	else if (uuid_equal(uuid, &CXL_EVENT_MEM_MODULE_UUID))
>  		ev_type = CXL_CPER_EVENT_MEM_MODULE;
> +	else if (uuid_equal(uuid, &CXL_EVENT_DC_EVENT_UUID)) {
> +		trace_cxl_dynamic_capacity(cxlmd, type, &record->event.dcd);
> +		return;
> +	}
>  
>  	cxl_event_trace_record(cxlmd, type, ev_type, uuid, &record->event);
>  }
> diff --git a/drivers/cxl/core/trace.h b/drivers/cxl/core/trace.h
> index 9167cfba7f59..1303024b5239 100644
> --- a/drivers/cxl/core/trace.h
> +++ b/drivers/cxl/core/trace.h

> +TRACE_EVENT(cxl_dynamic_capacity,
> +
> +	TP_PROTO(const struct cxl_memdev *cxlmd, enum cxl_event_log_type log,
> +		 struct cxl_event_dcd *rec),
> +
> +	TP_ARGS(cxlmd, log, rec),
> +
> +	TP_STRUCT__entry(
> +		CXL_EVT_TP_entry
> +
> +		/* Dynamic capacity Event */
> +		__field(u8, event_type)
> +		__field(u16, hostid)
> +		__field(u8, region_id)
> +		__field(u64, dpa_start)
> +		__field(u64, length)
> +		__array(u8, tag, CXL_EXTENT_TAG_LEN)
> +		__field(u16, sh_extent_seq)
> +	),
> +
> +	TP_fast_assign(
> +		CXL_EVT_TP_fast_assign(cxlmd, log, rec->hdr);
> +
> +		/* Dynamic_capacity Event */
> +		__entry->event_type = rec->event_type;
> +
> +		/* DCD event record data */
> +		__entry->hostid = le16_to_cpu(rec->host_id);
> +		__entry->region_id = rec->region_index;
> +		__entry->dpa_start = le64_to_cpu(rec->extent.start_dpa);
> +		__entry->length = le64_to_cpu(rec->extent.length);
> +		memcpy(__entry->tag, &rec->extent.tag, CXL_EXTENT_TAG_LEN);
> +		__entry->sh_extent_seq = le16_to_cpu(rec->extent.shared_extn_seq);
> +	),
> +
> +	CXL_EVT_TP_printk("event_type='%s' host_id='%d' region_id='%d' " \
> +		"starting_dpa=%llx length=%llx tag=%s " \
> +		"shared_extent_sequence=%d",
> +		show_dc_evt_type(__entry->event_type),
> +		__entry->hostid,
> +		__entry->region_id,
> +		__entry->dpa_start,
> +		__entry->length,
> +		__print_hex(__entry->tag, CXL_EXTENT_TAG_LEN),

%pU maybe?
https://elixir.bootlin.com/linux/v6.11.2/source/include/ras/ras_event.h#L248
uses it for the GUIDs in CPER etc.

I guess it depends on how strongly we want to push John's vision of these
always being UUIDs! (I'm in favor and here is just formatting a debug print
so that shouldn't be a problem even for those who want for some odd reason
to use something else for tags :)



> +		__entry->sh_extent_seq
> +	)
> +);
> +
>  #endif /* _CXL_EVENTS_H */
>  
>  #define TRACE_INCLUDE_FILE trace
>
Re: [PATCH v4 26/28] cxl/mem: Trace Dynamic capacity Event Record
Posted by Ira Weiny 1 month ago
Jonathan Cameron wrote:
> On Mon, 07 Oct 2024 18:16:32 -0500
> ira.weiny@intel.com wrote:
> 
> > From: Navneet Singh <navneet.singh@intel.com>
> > 
> > CXL rev 3.1 section 8.2.9.2.1 adds the Dynamic Capacity Event Records.
> > User space can use trace events for debugging of DC capacity changes.
> > 
> > Add DC trace points to the trace log.
> > 
> > Signed-off-by: Navneet Singh <navneet.singh@intel.com>
> > Signed-off-by: Ira Weiny <ira.weiny@intel.com>
> Minor comment inline about tag formatting.
> 
> Either way
> Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
> 

[snip]

> > +
> > +	CXL_EVT_TP_printk("event_type='%s' host_id='%d' region_id='%d' " \
> > +		"starting_dpa=%llx length=%llx tag=%s " \
> > +		"shared_extent_sequence=%d",
> > +		show_dc_evt_type(__entry->event_type),
> > +		__entry->hostid,
> > +		__entry->region_id,
> > +		__entry->dpa_start,
> > +		__entry->length,
> > +		__print_hex(__entry->tag, CXL_EXTENT_TAG_LEN),
> 
> %pU maybe?
> https://elixir.bootlin.com/linux/v6.11.2/source/include/ras/ras_event.h#L248
> uses it for the GUIDs in CPER etc.
> 
> I guess it depends on how strongly we want to push John's vision of these
> always being UUIDs! (I'm in favor and here is just formatting a debug print
> so that shouldn't be a problem even for those who want for some odd reason
> to use something else for tags :)

I'm not pushing against the UUID idea any more.  Just forgot this print.

Changed.

Ira


[snip]