[PATCH v3 0/2] drm/xe/pagefault: Add SRCID to pagefault reporting

[PATCH v3 0/2] drm/xe/pagefault: Add SRCID to pagefault reporting

[Jonathan Cavitt]

Add SRCID to the xe_pagefault struct, which reports the ID of the
faulting hardware unit.  This will be passed on to the
xe_vm_get_property_ioctl for reading per-vm faults and will assist in
diagnosing pagefaults.

v2:
- Readd pad check, as the pad in the ioctl struct was not changed
  (jcavitt)

v3:
- Rebase

Jonathan Cavitt (2):
  drm/xe/pagefault: Add SRCID to pagefault struct
  drm/xe/vm: Add srcid to xe_vm_get_property_ioctl fault report

 drivers/gpu/drm/xe/xe_guc_pagefault.c   | 1 +
 drivers/gpu/drm/xe/xe_pagefault.c       | 6 ++++--
 drivers/gpu/drm/xe/xe_pagefault_types.h | 4 +++-
 drivers/gpu/drm/xe/xe_vm.c              | 8 ++++++++
 drivers/gpu/drm/xe/xe_vm_types.h        | 2 ++
 include/uapi/drm/xe_drm.h               | 4 ++--
 6 files changed, 20 insertions(+), 5 deletions(-)

-- 
2.53.0

[PATCH v3 1/2] drm/xe/pagefault: Add SRCID to pagefault struct

[Jonathan Cavitt]

Add SRCID information to pagefault struct for the purpose of reporting
the hardware unit that resulted in the pagefault.

Signed-off-by: Jonathan Cavitt <jonathan.cavitt@intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: David Airlie <airlied@gmail.com>
Cc: Simona Vetter <simona@ffwll.ch>
---
 drivers/gpu/drm/xe/xe_guc_pagefault.c   | 1 +
 drivers/gpu/drm/xe/xe_pagefault.c       | 6 ++++--
 drivers/gpu/drm/xe/xe_pagefault_types.h | 4 +++-
 3 files changed, 8 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/xe/xe_guc_pagefault.c b/drivers/gpu/drm/xe/xe_guc_pagefault.c
index 607e32392f46..17cb7f359261 100644
--- a/drivers/gpu/drm/xe/xe_guc_pagefault.c
+++ b/drivers/gpu/drm/xe/xe_guc_pagefault.c
@@ -91,6 +91,7 @@ int xe_guc_pagefault_handler(struct xe_guc *guc, u32 *msg, u32 len)
 				   FIELD_GET(PFD_FAULT_TYPE, msg[2]));
 	pf.consumer.engine_class = FIELD_GET(PFD_ENG_CLASS, msg[0]);
 	pf.consumer.engine_instance = FIELD_GET(PFD_ENG_INSTANCE, msg[0]);
+	pf.consumer.srcid = FIELD_GET(PFD_SRC_ID, msg[0]);
 
 	pf.producer.private = guc;
 	pf.producer.ops = &guc_pagefault_ops;
diff --git a/drivers/gpu/drm/xe/xe_pagefault.c b/drivers/gpu/drm/xe/xe_pagefault.c
index dd3c068e1a39..42b682ba5f9b 100644
--- a/drivers/gpu/drm/xe/xe_pagefault.c
+++ b/drivers/gpu/drm/xe/xe_pagefault.c
@@ -248,7 +248,8 @@ static void xe_pagefault_print(struct xe_pagefault *pf)
 		   "\tAccessType: %lu\n"
 		   "\tFaultLevel: %lu\n"
 		   "\tEngineClass: %d %s\n"
-		   "\tEngineInstance: %d\n",
+		   "\tEngineInstance: %d\n"
+		   "\tSRCID: 0x%02x\n",
 		   pf->consumer.asid,
 		   upper_32_bits(pf->consumer.page_addr),
 		   lower_32_bits(pf->consumer.page_addr),
@@ -260,7 +261,8 @@ static void xe_pagefault_print(struct xe_pagefault *pf)
 			     pf->consumer.fault_type_level),
 		   pf->consumer.engine_class,
 		   xe_hw_engine_class_to_str(pf->consumer.engine_class),
-		   pf->consumer.engine_instance);
+		   pf->consumer.engine_instance,
+		   pf->consumer.srcid);
 }
 
 static void xe_pagefault_save_to_vm(struct xe_device *xe, struct xe_pagefault *pf)
diff --git a/drivers/gpu/drm/xe/xe_pagefault_types.h b/drivers/gpu/drm/xe/xe_pagefault_types.h
index c4ee625b93dd..60cc269aeea3 100644
--- a/drivers/gpu/drm/xe/xe_pagefault_types.h
+++ b/drivers/gpu/drm/xe/xe_pagefault_types.h
@@ -86,8 +86,10 @@ struct xe_pagefault {
 		u8 engine_class;
 		/** @consumer.engine_instance: engine instance */
 		u8 engine_instance;
+		/** @consumer.srcid: ID of hardware unit producing fault */
+		u8 srcid;
 		/** @consumer.reserved: reserved bits for future expansion */
-		u64 reserved;
+		u8 reserved[7];
 	} consumer;
 	/**
 	 * @producer: State for the producer (i.e., HW/FW interface). Populated
-- 
2.53.0

[PATCH v3 2/2] drm/xe/vm: Add srcid to xe_vm_get_property_ioctl fault report

[Jonathan Cavitt]

Add the SRCID of the faulting hardware unit to the return of the
xe_vm_get_property_ioctl fault report.

v2:
- Readd pad check, as the pad in the ioctl struct was not changed
  (jcavitt)

Signed-off-by: Jonathan Cavitt <jonathan.cavitt@intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: David Airlie <airlied@gmail.com>
Cc: Simona Vetter <simona@ffwll.ch>
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Maxime Ripard <mripard@kernel.org>
Cc: Thomas Zimmermann <tzimmermann@suse.de>
---
 drivers/gpu/drm/xe/xe_vm.c       | 8 ++++++++
 drivers/gpu/drm/xe/xe_vm_types.h | 2 ++
 include/uapi/drm/xe_drm.h        | 4 ++--
 3 files changed, 12 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/xe/xe_vm.c b/drivers/gpu/drm/xe/xe_vm.c
index 080c2fff0e95..7e084c05f10e 100644
--- a/drivers/gpu/drm/xe/xe_vm.c
+++ b/drivers/gpu/drm/xe/xe_vm.c
@@ -636,6 +636,7 @@ void xe_vm_add_fault_entry_pf(struct xe_vm *vm, struct xe_pagefault *pf)
 				  pf->consumer.fault_type_level);
 	e->fault_level = FIELD_GET(XE_PAGEFAULT_LEVEL_MASK,
 				   pf->consumer.fault_type_level);
+	e->srcid = pf->consumer.srcid;
 
 	list_add_tail(&e->list, &vm->faults.list);
 	vm->faults.len++;
@@ -4108,6 +4109,11 @@ static u8 xe_to_user_fault_level(u8 fault_level)
 	return fault_level;
 }
 
+static u8 xe_to_user_srcid(u8 srcid)
+{
+	return srcid;
+}
+
 static int fill_faults(struct xe_vm *vm,
 		       struct drm_xe_vm_get_property *args)
 {
@@ -4135,6 +4141,8 @@ static int fill_faults(struct xe_vm *vm,
 		fault_entry.fault_type = xe_to_user_fault_type(entry->fault_type);
 		fault_entry.fault_level = xe_to_user_fault_level(entry->fault_level);
 
+		fault_entry.srcid = xe_to_user_srcid(entry->srcid);
+
 		memcpy(&fault_list[i], &fault_entry, entry_size);
 
 		i++;
diff --git a/drivers/gpu/drm/xe/xe_vm_types.h b/drivers/gpu/drm/xe/xe_vm_types.h
index 635ed29b9a69..e8aea75341cc 100644
--- a/drivers/gpu/drm/xe/xe_vm_types.h
+++ b/drivers/gpu/drm/xe/xe_vm_types.h
@@ -196,6 +196,7 @@ struct xe_device;
  * @access_type: type of address access that resulted in fault
  * @fault_type: type of fault reported
  * @fault_level: fault level of the fault
+ * @srcid: ID of the faulting hardware unit
  */
 struct xe_vm_fault_entry {
 	struct list_head list;
@@ -204,6 +205,7 @@ struct xe_vm_fault_entry {
 	u8 access_type;
 	u8 fault_type;
 	u8 fault_level;
+	u8 srcid;
 };
 
 struct xe_vm {
diff --git a/include/uapi/drm/xe_drm.h b/include/uapi/drm/xe_drm.h
index 48e9f1fdb78d..2b45b691a032 100644
--- a/include/uapi/drm/xe_drm.h
+++ b/include/uapi/drm/xe_drm.h
@@ -1294,8 +1294,8 @@ struct xe_vm_fault {
 #define FAULT_LEVEL_PML4	3
 #define FAULT_LEVEL_PML5	4
 	__u8 fault_level;
-	/** @pad: MBZ */
-	__u8 pad;
+	/** @srcid: ID of the faulting hardware unit */
+	__u8 srcid;
 	/** @reserved: MBZ */
 	__u64 reserved[4];
 };
-- 
2.53.0

Re: [PATCH v3 0/2] drm/xe/pagefault: Add SRCID to pagefault reporting

[Matthew Brost]

On Thu, Jun 18, 2026 at 02:18:46AM +0800, Jonathan Cavitt wrote:
> Add SRCID to the xe_pagefault struct, which reports the ID of the
> faulting hardware unit.  This will be passed on to the

So is this the EU ID?

> xe_vm_get_property_ioctl for reading per-vm faults and will assist in
> diagnosing pagefaults.
> 

Is anyone asking for this? I'm wondering as this will conflict with one
of my series that is inflight upstream and believe is merged in some
internal branches.

Matt

> v2:
> - Readd pad check, as the pad in the ioctl struct was not changed
>   (jcavitt)
> 
> v3:
> - Rebase
> 
> Jonathan Cavitt (2):
>   drm/xe/pagefault: Add SRCID to pagefault struct
>   drm/xe/vm: Add srcid to xe_vm_get_property_ioctl fault report
> 
>  drivers/gpu/drm/xe/xe_guc_pagefault.c   | 1 +
>  drivers/gpu/drm/xe/xe_pagefault.c       | 6 ++++--
>  drivers/gpu/drm/xe/xe_pagefault_types.h | 4 +++-
>  drivers/gpu/drm/xe/xe_vm.c              | 8 ++++++++
>  drivers/gpu/drm/xe/xe_vm_types.h        | 2 ++
>  include/uapi/drm/xe_drm.h               | 4 ++--
>  6 files changed, 20 insertions(+), 5 deletions(-)
> 
> -- 
> 2.53.0
>