From 53e11d245c34656af56625bb06d59b9934428aba Mon Sep 17 00:00:00 2001
From: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
Date: Thu, 3 Apr 2025 20:03:01 +0100
Subject: [PATCH 01/16] drm/xe: Adjust ringbuf emission for maximum possible
 size

MAX_JOB_SIZE_DW seems to be undersized. For the worst case emission from
 __emit_job_gen12_render_compute I hand count 57 dwords so lets bump this
to an even 58.

Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
Reviewed-by: Francois Dugast <francois.dugast@intel.com>
Link: https://lore.kernel.org/r/20250403190317.6064-2-tvrtko.ursulin@igalia.com
Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
---
 drivers/gpu/drm/xe/xe_ring_ops_types.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/xe/xe_ring_ops_types.h b/drivers/gpu/drm/xe/xe_ring_ops_types.h
index 1ae56e2ee7b4..d7e3e150a9a5 100644
--- a/drivers/gpu/drm/xe/xe_ring_ops_types.h
+++ b/drivers/gpu/drm/xe/xe_ring_ops_types.h
@@ -8,7 +8,7 @@
 
 struct xe_sched_job;
 
-#define MAX_JOB_SIZE_DW 48
+#define MAX_JOB_SIZE_DW 58
 #define MAX_JOB_SIZE_BYTES (MAX_JOB_SIZE_DW * 4)
 
 /**
-- 
2.51.0


From bd7c0cb695e87c0e43247be8196b4919edbe0e85 Mon Sep 17 00:00:00 2001
From: Matthew Auld <matthew.auld@intel.com>
Date: Mon, 14 Apr 2025 14:25:40 +0100
Subject: [PATCH 02/16] drm/xe/userptr: fix notifier vs folio deadlock
MIME-Version: 1.0
Content-Type: text/plain; charset=utf8
Content-Transfer-Encoding: 8bit

User is reporting what smells like notifier vs folio deadlock, where
migrate_pages_batch() on core kernel side is holding folio lock(s) and
then interacting with the mappings of it, however those mappings are
tied to some userptr, which means calling into the notifier callback and
grabbing the notifier lock. With perfect timing it looks possible that
the pages we pulled from the hmm fault can get sniped by
migrate_pages_batch() at the same time that we are holding the notifier
lock to mark the pages as accessed/dirty, but at this point we also want
to grab the folio locks(s) to mark them as dirty, but if they are
contended from notifier/migrate_pages_batch side then we deadlock since
folio lock won't be dropped until we drop the notifier lock.

Fortunately the mark_page_accessed/dirty is not really needed in the
first place it seems and should have already been done by hmm fault, so
just remove it.

Link: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/4765
Fixes: 0a98219bcc96 ("drm/xe/hmm: Don't dereference struct page pointers without notifier lock")
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Cc: Thomas Hellström <thomas.hellstrom@intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: <stable@vger.kernel.org> # v6.10+
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Link: https://lore.kernel.org/r/20250414132539.26654-2-matthew.auld@intel.com
---
 drivers/gpu/drm/xe/xe_hmm.c | 24 ------------------------
 1 file changed, 24 deletions(-)

diff --git a/drivers/gpu/drm/xe/xe_hmm.c b/drivers/gpu/drm/xe/xe_hmm.c
index c3cc0fa105e8..57b71956ddf4 100644
--- a/drivers/gpu/drm/xe/xe_hmm.c
+++ b/drivers/gpu/drm/xe/xe_hmm.c
@@ -19,29 +19,6 @@ static u64 xe_npages_in_range(unsigned long start, unsigned long end)
 	return (end - start) >> PAGE_SHIFT;
 }
 
-/**
- * xe_mark_range_accessed() - mark a range is accessed, so core mm
- * have such information for memory eviction or write back to
- * hard disk
- * @range: the range to mark
- * @write: if write to this range, we mark pages in this range
- * as dirty
- */
-static void xe_mark_range_accessed(struct hmm_range *range, bool write)
-{
-	struct page *page;
-	u64 i, npages;
-
-	npages = xe_npages_in_range(range->start, range->end);
-	for (i = 0; i < npages; i++) {
-		page = hmm_pfn_to_page(range->hmm_pfns[i]);
-		if (write)
-			set_page_dirty_lock(page);
-
-		mark_page_accessed(page);
-	}
-}
-
 static int xe_alloc_sg(struct xe_device *xe, struct sg_table *st,
 		       struct hmm_range *range, struct rw_semaphore *notifier_sem)
 {
@@ -331,7 +308,6 @@ int xe_hmm_userptr_populate_range(struct xe_userptr_vma *uvma,
 	if (ret)
 		goto out_unlock;
 
-	xe_mark_range_accessed(&hmm_range, write);
 	userptr->sg = &userptr->sgt;
 	xe_hmm_userptr_set_mapped(uvma);
 	userptr->notifier_seq = hmm_range.notifier_seq;
-- 
2.51.0


From d755887f8e5a2a18e15e6632a5193e5feea18499 Mon Sep 17 00:00:00 2001
From: Matthew Auld <matthew.auld@intel.com>
Date: Thu, 10 Apr 2025 17:27:17 +0100
Subject: [PATCH 03/16] drm/xe/dma_buf: stop relying on placement in unmap
MIME-Version: 1.0
Content-Type: text/plain; charset=utf8
Content-Transfer-Encoding: 8bit

The is_vram() is checking the current placement, however if we consider
exported VRAM with dynamic dma-buf, it looks possible for the xe driver
to async evict the memory, notifying the importer, however importer does
not have to call unmap_attachment() immediately, but rather just as
"soon as possible", like when the dma-resv idles. Following from this we
would then pipeline the move, attaching the fence to the manager, and
then update the current placement. But when the unmap_attachment() runs
at some later point we might see that is_vram() is now false, and take
the complete wrong path when dma-unmapping the sg, leading to
explosions.

To fix this check if the sgl was mapping a struct page.

v2:
  - The attachment can be mapped multiple times it seems, so we can't
    really rely on encoding something in the attachment->priv. Instead
    see if the page_link has an encoded struct page. For vram we expect
    this to be NULL.

Link: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/4563
Fixes: dd08ebf6c352 ("drm/xe: Introduce a new DRM driver for Intel GPUs")
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Cc: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: <stable@vger.kernel.org> # v6.8+
Acked-by: Christian König <christian.koenig@amd.com>
Link: https://lore.kernel.org/r/20250410162716.159403-2-matthew.auld@intel.com
---
 drivers/gpu/drm/xe/xe_dma_buf.c | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/xe/xe_dma_buf.c b/drivers/gpu/drm/xe/xe_dma_buf.c
index f67803e15a0e..f7a20264ea33 100644
--- a/drivers/gpu/drm/xe/xe_dma_buf.c
+++ b/drivers/gpu/drm/xe/xe_dma_buf.c
@@ -145,10 +145,7 @@ static void xe_dma_buf_unmap(struct dma_buf_attachment *attach,
 			     struct sg_table *sgt,
 			     enum dma_data_direction dir)
 {
-	struct dma_buf *dma_buf = attach->dmabuf;
-	struct xe_bo *bo = gem_to_xe_bo(dma_buf->priv);
-
-	if (!xe_bo_is_vram(bo)) {
+	if (sg_page(sgt->sgl)) {
 		dma_unmap_sgtable(attach->dev, sgt, dir, 0);
 		sg_free_table(sgt);
 		kfree(sgt);
-- 
2.51.0


From ba1f62a0cac84757ca35f4217e3cd3a2654233ae Mon Sep 17 00:00:00 2001
From: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Date: Wed, 16 Apr 2025 13:16:22 -0700
Subject: [PATCH 04/16] drm/xe/pxp: do not queue unneeded terminations from
 debugfs

The PXP terminate debugfs currently unconditionally simulates a
termination, no matter what the HW status is. This is unneeded if PXP is
not in use and can cause errors if the HW init hasn't completed yet.
To solve these issues, we can simply limit the terminations to the cases
where PXP is fully initialized and in use.

v2: s/pxp_status/ready/ to avoid confusion with pxp->status (John)

Fixes: 385a8015b214 ("drm/xe/pxp: Add PXP debugfs support")
Closes: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/4749
Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: John Harrison <John.C.Harrison@Intel.com>
Link: https://lore.kernel.org/r/20250416201622.1295369-1-daniele.ceraolospurio@intel.com
---
 drivers/gpu/drm/xe/xe_pxp_debugfs.c | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/xe/xe_pxp_debugfs.c b/drivers/gpu/drm/xe/xe_pxp_debugfs.c
index ccfbacf08efc..525a2f6bb076 100644
--- a/drivers/gpu/drm/xe/xe_pxp_debugfs.c
+++ b/drivers/gpu/drm/xe/xe_pxp_debugfs.c
@@ -66,9 +66,18 @@ static int pxp_terminate(struct seq_file *m, void *data)
 {
 	struct xe_pxp *pxp = node_to_pxp(m->private);
 	struct drm_printer p = drm_seq_file_printer(m);
+	int ready = xe_pxp_get_readiness_status(pxp);
 
-	if (!xe_pxp_is_enabled(pxp))
-		return -ENODEV;
+	if (ready < 0)
+		return ready; /* disabled or error occurred */
+	else if (!ready)
+		return -EBUSY; /* init still in progress */
+
+	/* no need for a termination if PXP is not active */
+	if (pxp->status != XE_PXP_ACTIVE) {
+		drm_printf(&p, "PXP not active\n");
+		return 0;
+	}
 
 	/* simulate a termination interrupt */
 	spin_lock_irq(&pxp->xe->irq.lock);
-- 
2.51.0


From 8f9d815a89f5d08566c52ae1d18f05fe976b2c59 Mon Sep 17 00:00:00 2001
From: Michal Wajdeczko <michal.wajdeczko@intel.com>
Date: Mon, 14 Apr 2025 22:23:46 +0200
Subject: [PATCH 05/16] drm/xe/guc: Fix out-of-bound while enabling engine
 activity stats

In the PF mode we allocate array of struct engine_activity_group
that holds activity data split for the PF and all potential VFs.
But while preparing data for use by VFs we ended with bad index.

 [ ] BUG: KASAN: slab-out-of-bounds in xe_guc_engine_activity_function_stats+0x41e/0x4f0 [xe]
 [ ] Call Trace:
 [ ]  <TASK>
 [ ]  dump_stack_lvl+0x91/0xf0
 [ ]  print_report+0xd1/0x680
 [ ]  ? __virt_addr_valid+0x23a/0x440
 [ ]  ? kasan_addr_to_slab+0xd/0xb0
 [ ]  kasan_report+0xe7/0x130
 [ ]  ? xe_guc_engine_activity_function_stats+0x41e/0x4f0 [xe]
 [ ]  ? xe_guc_engine_activity_function_stats+0x41e/0x4f0 [xe]
 [ ]  __asan_report_store8_noabort+0x17/0x30
 [ ]  xe_guc_engine_activity_function_stats+0x41e/0x4f0 [xe]
 [ ]  pf_engine_activity_stats+0x1b6/0x7f0 [xe]
 [ ]  ? kobject_put+0x5f/0x470
 [ ]  xe_pci_sriov_configure+0x28c9/0x3270 [xe]
 [ ]  ? __pfx_dev_attr_store+0x10/0x10
 [ ]  ? kstrtoull+0x3b/0x70
 [ ]  ? __pfx___lock_acquire+0x10/0x10
 [ ]  ? kstrtou16+0x65/0xf0
 [ ]  sriov_numvfs_store+0x20c/0x400
 [ ]  ? __pfx_sriov_numvfs_store+0x10/0x10
 [ ]  ? __pfx__copy_from_iter+0x10/0x10
 [ ]  ? __pfx_dev_attr_store+0x10/0x10
 [ ]  dev_attr_store+0x3b/0x80
 [ ]  ? sysfs_file_ops+0x135/0x190

Fixes: 2de3f38fbf89 ("drm/xe: Add support for per-function engine activity")
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
Cc: Riana Tauro <riana.tauro@intel.com>
Reviewed-by: Riana Tauro <riana.tauro@intel.com>
Link: https://lore.kernel.org/r/20250414202347.1909-1-michal.wajdeczko@intel.com
---
 drivers/gpu/drm/xe/xe_guc_engine_activity.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/xe/xe_guc_engine_activity.c b/drivers/gpu/drm/xe/xe_guc_engine_activity.c
index b96fea78df8b..0fb48f8f05d8 100644
--- a/drivers/gpu/drm/xe/xe_guc_engine_activity.c
+++ b/drivers/gpu/drm/xe/xe_guc_engine_activity.c
@@ -304,6 +304,8 @@ static void engine_activity_set_cpu_ts(struct xe_guc *guc, unsigned int index)
 	struct engine_activity_group *eag = &engine_activity->eag[index];
 	int i, j;
 
+	xe_gt_assert(guc_to_gt(guc), index < engine_activity->num_activity_group);
+
 	for (i = 0; i < GUC_MAX_ENGINE_CLASSES; i++)
 		for (j = 0; j < GUC_MAX_INSTANCES_PER_CLASS; j++)
 			eag->engine[i][j].last_cpu_ts = ktime_get();
@@ -374,8 +376,9 @@ static int engine_activity_enable_function_stats(struct xe_guc *guc, int num_vfs
 		return ret;
 	}
 
-	for (i = 0; i < engine_activity->num_functions; i++)
-		engine_activity_set_cpu_ts(guc, i + 1);
+	/* skip PF as it was already setup */
+	for (i = 1; i < engine_activity->num_functions; i++)
+		engine_activity_set_cpu_ts(guc, i);
 
 	return 0;
 }
-- 
2.51.0


From 73bf722bbb618df6d7936f97e1d68298fc893718 Mon Sep 17 00:00:00 2001
From: Michal Wajdeczko <michal.wajdeczko@intel.com>
Date: Mon, 14 Apr 2025 22:23:47 +0200
Subject: [PATCH 06/16] drm/xe: Use GT oriented message to report engine
 activity error

We are enabling/disabling engine activity on per-GT basis, so any
errors should be also reported per GT, like:

 [ ] xe 0000:00:02.0: [drm] GT0: PF: Failed to enable engine activity function stats (-ENOSPC)
 [ ] xe 0000:00:02.0: [drm] GT1: PF: Failed to enable engine activity function stats (-ENOSPC)

Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
Cc: Riana Tauro <riana.tauro@intel.com>
Reviewed-by: Riana Tauro <riana.tauro@intel.com>
Link: https://lore.kernel.org/r/20250414202347.1909-2-michal.wajdeczko@intel.com
---
 drivers/gpu/drm/xe/xe_pci_sriov.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/xe/xe_pci_sriov.c b/drivers/gpu/drm/xe/xe_pci_sriov.c
index d69b6b2a3061..8813efdcafbb 100644
--- a/drivers/gpu/drm/xe/xe_pci_sriov.c
+++ b/drivers/gpu/drm/xe/xe_pci_sriov.c
@@ -7,6 +7,7 @@
 #include "xe_device.h"
 #include "xe_gt_sriov_pf_config.h"
 #include "xe_gt_sriov_pf_control.h"
+#include "xe_gt_sriov_printk.h"
 #include "xe_guc_engine_activity.h"
 #include "xe_pci_sriov.h"
 #include "xe_pm.h"
@@ -121,8 +122,8 @@ static void pf_engine_activity_stats(struct xe_device *xe, unsigned int num_vfs,
 	for_each_gt(gt, xe, id) {
 		ret = xe_guc_engine_activity_function_stats(&gt->uc.guc, num_vfs, enable);
 		if (ret)
-			xe_sriov_info(xe, "Failed to %s engine activity function stats (%pe)\n",
-				      str_enable_disable(enable), ERR_PTR(ret));
+			xe_gt_sriov_info(gt, "Failed to %s engine activity function stats (%pe)\n",
+					 str_enable_disable(enable), ERR_PTR(ret));
 	}
 }
 
-- 
2.51.0


From e9dea328e8392d01b21544587cbcf313e4fdbe26 Mon Sep 17 00:00:00 2001
From: Satyanarayana K V P <satyanarayana.k.v.p@intel.com>
Date: Thu, 3 Apr 2025 17:36:40 +0530
Subject: [PATCH 07/16] drm/xe: Introduce fault injection for guc mmio
 send/recv.
MIME-Version: 1.0
Content-Type: text/plain; charset=utf8
Content-Transfer-Encoding: 8bit

Fault can be injected with below steps.

FAILTYPE=fail_function
FAILFUNC=xe_guc_mmio_send_recv

echo > /sys/kernel/debug/$FAILTYPE/inject
echo $FAILFUNC > /sys/kernel/debug/$FAILTYPE/inject
printf %#x -5 > /sys/kernel/debug/$FAILTYPE/$FAILFUNC/retval

echo N > /sys/kernel/debug/$FAILTYPE/task-filter
echo 10 > /sys/kernel/debug/$FAILTYPE/probability
echo 0 > /sys/kernel/debug/$FAILTYPE/interval
echo -1 > /sys/kernel/debug/$FAILTYPE/times
echo 0 > /sys/kernel/debug/$FAILTYPE/space
echo 1 > /sys/kernel/debug/$FAILTYPE/verbose

Signed-off-by: Satyanarayana K V P <satyanarayana.k.v.p@intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Michał Wajdeczko <michal.wajdeczko@intel.com>
Reviewed-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Link: https://lore.kernel.org/r/20250403120641.7258-2-satyanarayana.k.v.p@intel.com
---
 drivers/gpu/drm/xe/xe_guc.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/xe/xe_guc.c b/drivers/gpu/drm/xe/xe_guc.c
index 38866135c019..c5aace59b62c 100644
--- a/drivers/gpu/drm/xe/xe_guc.c
+++ b/drivers/gpu/drm/xe/xe_guc.c
@@ -1394,6 +1394,7 @@ proto:
 	/* Use data from the GuC response as our return value */
 	return FIELD_GET(GUC_HXG_RESPONSE_MSG_0_DATA0, header);
 }
+ALLOW_ERROR_INJECTION(xe_guc_mmio_send_recv, ERRNO);
 
 int xe_guc_mmio_send(struct xe_guc *guc, const u32 *request, u32 len)
 {
-- 
2.51.0


From 104080e33937aad54b4fbbe847bba750847abfdb Mon Sep 17 00:00:00 2001
From: Satyanarayana K V P <satyanarayana.k.v.p@intel.com>
Date: Thu, 3 Apr 2025 17:36:41 +0530
Subject: [PATCH 08/16] drm/xe: Introduce fault injection for guc CTB send/recv
MIME-Version: 1.0
Content-Type: text/plain; charset=utf8
Content-Transfer-Encoding: 8bit

Fault can be injected with below steps.

FAILTYPE=fail_function
FAILFUNC=xe_guc_ct_send_recv

echo > /sys/kernel/debug/$FAILTYPE/inject
echo $FAILFUNC > /sys/kernel/debug/$FAILTYPE/inject
printf %#x -19 > /sys/kernel/debug/$FAILTYPE/$FAILFUNC/retval

echo N > /sys/kernel/debug/$FAILTYPE/task-filter
echo 10 > /sys/kernel/debug/$FAILTYPE/probability
echo 0 > /sys/kernel/debug/$FAILTYPE/interval
echo -1 > /sys/kernel/debug/$FAILTYPE/times
echo 0 > /sys/kernel/debug/$FAILTYPE/space
echo 1 > /sys/kernel/debug/$FAILTYPE/verbose

Signed-off-by: Satyanarayana K V P <satyanarayana.k.v.p@intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Michał Wajdeczko <michal.wajdeczko@intel.com>
Reviewed-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Link: https://lore.kernel.org/r/20250403120641.7258-3-satyanarayana.k.v.p@intel.com
---
 drivers/gpu/drm/xe/xe_guc_ct.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/xe/xe_guc_ct.c b/drivers/gpu/drm/xe/xe_guc_ct.c
index 0a4fef7d7225..2447de0ebedf 100644
--- a/drivers/gpu/drm/xe/xe_guc_ct.c
+++ b/drivers/gpu/drm/xe/xe_guc_ct.c
@@ -1089,6 +1089,7 @@ int xe_guc_ct_send_recv(struct xe_guc_ct *ct, const u32 *action, u32 len,
 	KUNIT_STATIC_STUB_REDIRECT(xe_guc_ct_send_recv, ct, action, len, response_buffer);
 	return guc_ct_send_recv(ct, action, len, response_buffer, false);
 }
+ALLOW_ERROR_INJECTION(xe_guc_ct_send_recv, ERRNO);
 
 int xe_guc_ct_send_recv_no_fail(struct xe_guc_ct *ct, const u32 *action,
 				u32 len, u32 *response_buffer)
-- 
2.51.0


From 7a0322122cfdd9a6f10fc7701023d75c98eb3d22 Mon Sep 17 00:00:00 2001
From: Harshit Mogalapalli <harshit.m.mogalapalli@oracle.com>
Date: Sun, 23 Mar 2025 05:49:06 -0700
Subject: [PATCH 09/16] drm/xe/svm: fix dereferencing error pointer in
 drm_gpusvm_range_alloc()

xe_svm_range_alloc() returns ERR_PTR(-ENOMEM) on failure and there is a
dereference of "range" after that:

	-->     range->gpusvm = gpusvm;

In xe_svm_range_alloc(), when memory allocation fails return NULL
instead to handle this situation.

Fixes: 99624bdff867 ("drm/gpusvm: Add support for GPU Shared Virtual Memory")
Reported-by: Dan Carpenter <dan.carpenter@linaro.org>
Closes: https://lore.kernel.org/all/adaef4dd-5866-48ca-bc22-4a1ddef20381@stanley.mountain/
Signed-off-by: Harshit Mogalapalli <harshit.m.mogalapalli@oracle.com>
Reviewed-by: Matthew Brost <matthew.brost@intel.com>
Signed-off-by: Matthew Brost <matthew.brost@intel.com>
Link: https://lore.kernel.org/r/20250323124907.3946370-1-harshit.m.mogalapalli@oracle.com
---
 drivers/gpu/drm/xe/xe_svm.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/xe/xe_svm.c b/drivers/gpu/drm/xe/xe_svm.c
index c7424c824a14..d927f50f2a68 100644
--- a/drivers/gpu/drm/xe/xe_svm.c
+++ b/drivers/gpu/drm/xe/xe_svm.c
@@ -80,7 +80,7 @@ xe_svm_range_alloc(struct drm_gpusvm *gpusvm)
 
 	range = kzalloc(sizeof(*range), GFP_KERNEL);
 	if (!range)
-		return ERR_PTR(-ENOMEM);
+		return NULL;
 
 	INIT_LIST_HEAD(&range->garbage_collector_link);
 	xe_vm_get(gpusvm_to_vm(gpusvm));
-- 
2.51.0


From 532da44b54a10d50ebad14a8a02bd0b78ec23e8b Mon Sep 17 00:00:00 2001
From: John Harrison <John.C.Harrison@Intel.com>
Date: Thu, 17 Apr 2025 12:52:12 -0700
Subject: [PATCH 10/16] drm/xe/guc: Fix capture of steering registers
MIME-Version: 1.0
Content-Type: text/plain; charset=utf8
Content-Transfer-Encoding: 8bit

The list of registers to capture on a GPU hang includes some that
require steering. Unfortunately, the flag to say this was being wiped
to due a missing OR on the assignment of the next flag field.

Fix that.

Fixes: b170d696c1e2 ("drm/xe/guc: Add XE_LP steered register lists")
Cc: Zhanjun Dong <zhanjun.dong@intel.com>
Cc: Alan Previn <alan.previn.teres.alexis@intel.com>
Cc: Matt Roper <matthew.d.roper@intel.com>
Cc: Lucas De Marchi <lucas.demarchi@intel.com>
Cc: "Thomas Hellström" <thomas.hellstrom@linux.intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: intel-xe@lists.freedesktop.org
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Zhanjun Dong <zhanjun.dong@intel.com>
Link: https://lore.kernel.org/r/20250417195215.3002210-2-John.C.Harrison@Intel.com
---
 drivers/gpu/drm/xe/xe_guc_capture.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/xe/xe_guc_capture.c b/drivers/gpu/drm/xe/xe_guc_capture.c
index f6d523e4c5fe..9095618648bc 100644
--- a/drivers/gpu/drm/xe/xe_guc_capture.c
+++ b/drivers/gpu/drm/xe/xe_guc_capture.c
@@ -359,7 +359,7 @@ static void __fill_ext_reg(struct __guc_mmio_reg_descr *ext,
 
 	ext->reg = XE_REG(extlist->reg.__reg.addr);
 	ext->flags = FIELD_PREP(GUC_REGSET_STEERING_NEEDED, 1);
-	ext->flags = FIELD_PREP(GUC_REGSET_STEERING_GROUP, slice_id);
+	ext->flags |= FIELD_PREP(GUC_REGSET_STEERING_GROUP, slice_id);
 	ext->flags |= FIELD_PREP(GUC_REGSET_STEERING_INSTANCE, subslice_id);
 	ext->regname = extlist->name;
 }
-- 
2.51.0


From 8393f3e155d902bd66c3033b073ec10d1409b2ee Mon Sep 17 00:00:00 2001
From: John Harrison <John.C.Harrison@Intel.com>
Date: Thu, 17 Apr 2025 12:52:13 -0700
Subject: [PATCH 11/16] drm/xe/guc: Use the steering flag when printing
 registers

The printing code was doing a test on which list a register was in to
decide whether it is steered or not. That might be valid at this
moment but there may be other reasons for extended lists in the
future. Plus, there is a flag specifically for identifying steered
registers. So, just use that instead - it is simpler and safer.

Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://lore.kernel.org/r/20250417195215.3002210-3-John.C.Harrison@Intel.com
---
 drivers/gpu/drm/xe/xe_guc_capture.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/xe/xe_guc_capture.c b/drivers/gpu/drm/xe/xe_guc_capture.c
index 9095618648bc..f4b08136e472 100644
--- a/drivers/gpu/drm/xe/xe_guc_capture.c
+++ b/drivers/gpu/drm/xe/xe_guc_capture.c
@@ -1672,18 +1672,16 @@ snapshot_print_by_list_order(struct xe_hw_engine_snapshot *snapshot, struct drm_
 {
 	struct xe_gt *gt = snapshot->hwe->gt;
 	struct xe_device *xe = gt_to_xe(gt);
-	struct xe_guc *guc = &gt->uc.guc;
 	struct xe_devcoredump *devcoredump = &xe->devcoredump;
 	struct xe_devcoredump_snapshot *devcore_snapshot = &devcoredump->snapshot;
 	struct gcap_reg_list_info *reginfo = NULL;
 	u32 i, last_value = 0;
-	bool is_ext, low32_ready = false;
+	bool low32_ready = false;
 
 	if (!list || !list->list || list->num_regs == 0)
 		return;
 	XE_WARN_ON(!devcore_snapshot->matched_node);
 
-	is_ext = list == guc->capture->extlists;
 	reginfo = &devcore_snapshot->matched_node->reginfo[type];
 
 	/*
@@ -1749,7 +1747,7 @@ snapshot_print_by_list_order(struct xe_hw_engine_snapshot *snapshot, struct drm_
 			 */
 			XE_WARN_ON(low32_ready);
 
-			if (is_ext) {
+			if (FIELD_GET(GUC_REGSET_STEERING_NEEDED, reg_desc->flags)) {
 				int dss, group, instance;
 
 				group = FIELD_GET(GUC_REGSET_STEERING_GROUP, reg_desc->flags);
-- 
2.51.0


From fa597710be6e6625b875d95c717f66b7ab83b986 Mon Sep 17 00:00:00 2001
From: John Harrison <John.C.Harrison@Intel.com>
Date: Thu, 17 Apr 2025 14:33:03 -0700
Subject: [PATCH 12/16] drm/xe/guc: Cache DSS info when creating capture
 register list

Calculating the DSS id (index of a steered register) currently
requires reading state from the hwconfig table and that currently
requires dynamically allocating memory. The GuC based register capture
(for dev core dumps) includes this index as part of the register name
in the dump. However, it was calculating said index at the time of the
dump for every dump. That is wasteful. It also breaks anyone trying to
do the dump at a time when memory allocations are not allowed.

So rather than calculating on every print, just calculate at start of
day when creating the register list in the first place.

Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://lore.kernel.org/r/20250417213303.3021243-1-John.C.Harrison@Intel.com
---
 drivers/gpu/drm/xe/xe_guc_capture.c       | 96 +++++++++++------------
 drivers/gpu/drm/xe/xe_guc_capture_types.h |  2 +
 2 files changed, 48 insertions(+), 50 deletions(-)

diff --git a/drivers/gpu/drm/xe/xe_guc_capture.c b/drivers/gpu/drm/xe/xe_guc_capture.c
index f4b08136e472..859a3ba91be5 100644
--- a/drivers/gpu/drm/xe/xe_guc_capture.c
+++ b/drivers/gpu/drm/xe/xe_guc_capture.c
@@ -105,49 +105,49 @@ struct __guc_capture_parsed_output {
  *     3. Incorrect order will trigger XE_WARN.
  */
 #define COMMON_XELP_BASE_GLOBAL \
-	{ FORCEWAKE_GT,			REG_32BIT,	0,	0,	"FORCEWAKE_GT"}
+	{ FORCEWAKE_GT,			REG_32BIT,	0,	0,	0,	"FORCEWAKE_GT"}
 
 #define COMMON_BASE_ENGINE_INSTANCE \
-	{ RING_HWSTAM(0),		REG_32BIT,	0,	0,	"HWSTAM"}, \
-	{ RING_HWS_PGA(0),		REG_32BIT,	0,	0,	"RING_HWS_PGA"}, \
-	{ RING_HEAD(0),			REG_32BIT,	0,	0,	"RING_HEAD"}, \
-	{ RING_TAIL(0),			REG_32BIT,	0,	0,	"RING_TAIL"}, \
-	{ RING_CTL(0),			REG_32BIT,	0,	0,	"RING_CTL"}, \
-	{ RING_MI_MODE(0),		REG_32BIT,	0,	0,	"RING_MI_MODE"}, \
-	{ RING_MODE(0),			REG_32BIT,	0,	0,	"RING_MODE"}, \
-	{ RING_ESR(0),			REG_32BIT,	0,	0,	"RING_ESR"}, \
-	{ RING_EMR(0),			REG_32BIT,	0,	0,	"RING_EMR"}, \
-	{ RING_EIR(0),			REG_32BIT,	0,	0,	"RING_EIR"}, \
-	{ RING_IMR(0),			REG_32BIT,	0,	0,	"RING_IMR"}, \
-	{ RING_IPEHR(0),		REG_32BIT,	0,	0,	"IPEHR"}, \
-	{ RING_INSTDONE(0),		REG_32BIT,	0,	0,	"RING_INSTDONE"}, \
-	{ INDIRECT_RING_STATE(0),	REG_32BIT,	0,	0,	"INDIRECT_RING_STATE"}, \
-	{ RING_ACTHD(0),		REG_64BIT_LOW_DW, 0,	0,	NULL}, \
-	{ RING_ACTHD_UDW(0),		REG_64BIT_HI_DW, 0,	0,	"ACTHD"}, \
-	{ RING_BBADDR(0),		REG_64BIT_LOW_DW, 0,	0,	NULL}, \
-	{ RING_BBADDR_UDW(0),		REG_64BIT_HI_DW, 0,	0,	"RING_BBADDR"}, \
-	{ RING_START(0),		REG_64BIT_LOW_DW, 0,	0,	NULL}, \
-	{ RING_START_UDW(0),		REG_64BIT_HI_DW, 0,	0,	"RING_START"}, \
-	{ RING_DMA_FADD(0),		REG_64BIT_LOW_DW, 0,	0,	NULL}, \
-	{ RING_DMA_FADD_UDW(0),		REG_64BIT_HI_DW, 0,	0,	"RING_DMA_FADD"}, \
-	{ RING_EXECLIST_STATUS_LO(0),	REG_64BIT_LOW_DW, 0,	0,	NULL}, \
-	{ RING_EXECLIST_STATUS_HI(0),	REG_64BIT_HI_DW, 0,	0,	"RING_EXECLIST_STATUS"}, \
-	{ RING_EXECLIST_SQ_CONTENTS_LO(0), REG_64BIT_LOW_DW, 0,	0,	NULL}, \
-	{ RING_EXECLIST_SQ_CONTENTS_HI(0), REG_64BIT_HI_DW, 0,	0,	"RING_EXECLIST_SQ_CONTENTS"}
+	{ RING_HWSTAM(0),		REG_32BIT,	0,	0,	0,	"HWSTAM"}, \
+	{ RING_HWS_PGA(0),		REG_32BIT,	0,	0,	0,	"RING_HWS_PGA"}, \
+	{ RING_HEAD(0),			REG_32BIT,	0,	0,	0,	"RING_HEAD"}, \
+	{ RING_TAIL(0),			REG_32BIT,	0,	0,	0,	"RING_TAIL"}, \
+	{ RING_CTL(0),			REG_32BIT,	0,	0,	0,	"RING_CTL"}, \
+	{ RING_MI_MODE(0),		REG_32BIT,	0,	0,	0,	"RING_MI_MODE"}, \
+	{ RING_MODE(0),			REG_32BIT,	0,	0,	0,	"RING_MODE"}, \
+	{ RING_ESR(0),			REG_32BIT,	0,	0,	0,	"RING_ESR"}, \
+	{ RING_EMR(0),			REG_32BIT,	0,	0,	0,	"RING_EMR"}, \
+	{ RING_EIR(0),			REG_32BIT,	0,	0,	0,	"RING_EIR"}, \
+	{ RING_IMR(0),			REG_32BIT,	0,	0,	0,	"RING_IMR"}, \
+	{ RING_IPEHR(0),		REG_32BIT,	0,	0,	0,	"IPEHR"}, \
+	{ RING_INSTDONE(0),		REG_32BIT,	0,	0,	0,	"RING_INSTDONE"}, \
+	{ INDIRECT_RING_STATE(0),	REG_32BIT,	0,	0,	0,	"INDIRECT_RING_STATE"}, \
+	{ RING_ACTHD(0),		REG_64BIT_LOW_DW, 0,	0,	0,	NULL}, \
+	{ RING_ACTHD_UDW(0),		REG_64BIT_HI_DW, 0,	0,	0,	"ACTHD"}, \
+	{ RING_BBADDR(0),		REG_64BIT_LOW_DW, 0,	0,	0,	NULL}, \
+	{ RING_BBADDR_UDW(0),		REG_64BIT_HI_DW, 0,	0,	0,	"RING_BBADDR"}, \
+	{ RING_START(0),		REG_64BIT_LOW_DW, 0,	0,	0,	NULL}, \
+	{ RING_START_UDW(0),		REG_64BIT_HI_DW, 0,	0,	0,	"RING_START"}, \
+	{ RING_DMA_FADD(0),		REG_64BIT_LOW_DW, 0,	0,	0,	NULL}, \
+	{ RING_DMA_FADD_UDW(0),		REG_64BIT_HI_DW, 0,	0,	0,	"RING_DMA_FADD"}, \
+	{ RING_EXECLIST_STATUS_LO(0),	REG_64BIT_LOW_DW, 0,	0,	0,	NULL}, \
+	{ RING_EXECLIST_STATUS_HI(0),	REG_64BIT_HI_DW, 0,	0,	0,	"RING_EXECLIST_STATUS"}, \
+	{ RING_EXECLIST_SQ_CONTENTS_LO(0), REG_64BIT_LOW_DW, 0,	0,	0,	NULL}, \
+	{ RING_EXECLIST_SQ_CONTENTS_HI(0), REG_64BIT_HI_DW, 0,	0,	0,	"RING_EXECLIST_SQ_CONTENTS"}
 
 #define COMMON_XELP_RC_CLASS \
-	{ RCU_MODE,			REG_32BIT,	0,	0,	"RCU_MODE"}
+	{ RCU_MODE,			REG_32BIT,	0,	0,	0,	"RCU_MODE"}
 
 #define COMMON_XELP_RC_CLASS_INSTDONE \
-	{ SC_INSTDONE,			REG_32BIT,	0,	0,	"SC_INSTDONE"}, \
-	{ SC_INSTDONE_EXTRA,		REG_32BIT,	0,	0,	"SC_INSTDONE_EXTRA"}, \
-	{ SC_INSTDONE_EXTRA2,		REG_32BIT,	0,	0,	"SC_INSTDONE_EXTRA2"}
+	{ SC_INSTDONE,			REG_32BIT,	0,	0,	0,	"SC_INSTDONE"}, \
+	{ SC_INSTDONE_EXTRA,		REG_32BIT,	0,	0,	0,	"SC_INSTDONE_EXTRA"}, \
+	{ SC_INSTDONE_EXTRA2,		REG_32BIT,	0,	0,	0,	"SC_INSTDONE_EXTRA2"}
 
 #define XELP_VEC_CLASS_REGS \
-	{ SFC_DONE(0),			0,	0,	0,	"SFC_DONE[0]"}, \
-	{ SFC_DONE(1),			0,	0,	0,	"SFC_DONE[1]"}, \
-	{ SFC_DONE(2),			0,	0,	0,	"SFC_DONE[2]"}, \
-	{ SFC_DONE(3),			0,	0,	0,	"SFC_DONE[3]"}
+	{ SFC_DONE(0),			0,	0,	0,	0,	"SFC_DONE[0]"}, \
+	{ SFC_DONE(1),			0,	0,	0,	0,	"SFC_DONE[1]"}, \
+	{ SFC_DONE(2),			0,	0,	0,	0,	"SFC_DONE[2]"}, \
+	{ SFC_DONE(3),			0,	0,	0,	0,	"SFC_DONE[3]"}
 
 /* XE_LP Global */
 static const struct __guc_mmio_reg_descr xe_lp_global_regs[] = {
@@ -352,7 +352,7 @@ static const struct __ext_steer_reg xehpg_extregs[] = {
 
 static void __fill_ext_reg(struct __guc_mmio_reg_descr *ext,
 			   const struct __ext_steer_reg *extlist,
-			   int slice_id, int subslice_id)
+			   u32 dss_id, u16 slice_id, u16 subslice_id)
 {
 	if (!ext || !extlist)
 		return;
@@ -361,6 +361,7 @@ static void __fill_ext_reg(struct __guc_mmio_reg_descr *ext,
 	ext->flags = FIELD_PREP(GUC_REGSET_STEERING_NEEDED, 1);
 	ext->flags |= FIELD_PREP(GUC_REGSET_STEERING_GROUP, slice_id);
 	ext->flags |= FIELD_PREP(GUC_REGSET_STEERING_INSTANCE, subslice_id);
+	ext->dss_id = dss_id;
 	ext->regname = extlist->name;
 }
 
@@ -397,7 +398,7 @@ static void guc_capture_alloc_steered_lists(struct xe_guc *guc)
 {
 	struct xe_gt *gt = guc_to_gt(guc);
 	u16 slice, subslice;
-	int iter, i, total = 0;
+	int dss, i, total = 0;
 	const struct __guc_mmio_reg_descr_group *lists = guc->capture->reglists;
 	const struct __guc_mmio_reg_descr_group *list;
 	struct __guc_mmio_reg_descr_group *extlists;
@@ -454,15 +455,15 @@ static void guc_capture_alloc_steered_lists(struct xe_guc *guc)
 
 	/* For steering registers, the list is generated at run-time */
 	extarray = (struct __guc_mmio_reg_descr *)extlists[0].list;
-	for_each_dss_steering(iter, gt, slice, subslice) {
+	for_each_dss_steering(dss, gt, slice, subslice) {
 		for (i = 0; i < ARRAY_SIZE(xe_extregs); ++i) {
-			__fill_ext_reg(extarray, &xe_extregs[i], slice, subslice);
+			__fill_ext_reg(extarray, &xe_extregs[i], dss, slice, subslice);
 			++extarray;
 		}
 
 		if (has_xehpg_extregs)
 			for (i = 0; i < ARRAY_SIZE(xehpg_extregs); ++i) {
-				__fill_ext_reg(extarray, &xehpg_extregs[i], slice, subslice);
+				__fill_ext_reg(extarray, &xehpg_extregs[i], dss, slice, subslice);
 				++extarray;
 			}
 	}
@@ -1747,17 +1748,12 @@ snapshot_print_by_list_order(struct xe_hw_engine_snapshot *snapshot, struct drm_
 			 */
 			XE_WARN_ON(low32_ready);
 
-			if (FIELD_GET(GUC_REGSET_STEERING_NEEDED, reg_desc->flags)) {
-				int dss, group, instance;
-
-				group = FIELD_GET(GUC_REGSET_STEERING_GROUP, reg_desc->flags);
-				instance = FIELD_GET(GUC_REGSET_STEERING_INSTANCE, reg_desc->flags);
-				dss = xe_gt_mcr_steering_info_to_dss_id(gt, group, instance);
-
-				drm_printf(p, "\t%s[%u]: 0x%08x\n", reg_desc->regname, dss, value);
-			} else {
+			if (FIELD_GET(GUC_REGSET_STEERING_NEEDED, reg_desc->flags))
+				drm_printf(p, "\t%s[%u]: 0x%08x\n", reg_desc->regname,
+					   reg_desc->dss_id, value);
+			else
 				drm_printf(p, "\t%s: 0x%08x\n", reg_desc->regname, value);
-			}
+
 			break;
 		}
 	}
diff --git a/drivers/gpu/drm/xe/xe_guc_capture_types.h b/drivers/gpu/drm/xe/xe_guc_capture_types.h
index ca2d390ccbee..6cb439115597 100644
--- a/drivers/gpu/drm/xe/xe_guc_capture_types.h
+++ b/drivers/gpu/drm/xe/xe_guc_capture_types.h
@@ -39,6 +39,8 @@ struct __guc_mmio_reg_descr {
 	u32 flags;
 	/** @mask: The mask to apply */
 	u32 mask;
+	/** @dss_id: Cached index for steered registers */
+	u32 dss_id;
 	/** @regname: Name of the register */
 	const char *regname;
 };
-- 
2.51.0


From c6a4d46ec1d714b8055d1f4121197f15a4a02c68 Mon Sep 17 00:00:00 2001
From: Matthew Auld <matthew.auld@intel.com>
Date: Wed, 16 Apr 2025 16:09:15 +0100
Subject: [PATCH 13/16] drm/xe: evict user memory in PM notifier
MIME-Version: 1.0
Content-Type: text/plain; charset=utf8
Content-Transfer-Encoding: 8bit

In the case of VRAM we might need to allocate large amounts of
GFP_KERNEL memory on suspend, however doing that directly in the driver
.suspend()/.prepare() callback is not advisable (no swap for example).

To improve on this we can instead hook up to the PM notifier framework
which is invoked at an earlier stage. We effectively call the evict
routine twice, where the notifier will have hopefully have cleared out
most if not everything by the time we call it a second time when
entering the .suspend() callback. For s4 we also get the added benefit
of allocating the system pages before the hibernation image size is
calculated, which looks more sensible.

Note that the .suspend() hook is still responsible for dealing with all
the pinned memory. Improving that is left to another patch.

Link: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/1181
Link: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/4288
Link: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/4566
Suggested-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Link: https://lore.kernel.org/r/20250416150913.434369-6-matthew.auld@intel.com
---
 drivers/gpu/drm/xe/xe_bo_evict.c     | 45 ++++++++++++++++-------
 drivers/gpu/drm/xe/xe_bo_evict.h     |  1 +
 drivers/gpu/drm/xe/xe_device_types.h |  3 ++
 drivers/gpu/drm/xe/xe_pci.c          |  2 +-
 drivers/gpu/drm/xe/xe_pm.c           | 55 ++++++++++++++++++++++++----
 drivers/gpu/drm/xe/xe_pm.h           |  2 +-
 6 files changed, 84 insertions(+), 24 deletions(-)

diff --git a/drivers/gpu/drm/xe/xe_bo_evict.c b/drivers/gpu/drm/xe/xe_bo_evict.c
index 2bf74eb7f281..748360fd2439 100644
--- a/drivers/gpu/drm/xe/xe_bo_evict.c
+++ b/drivers/gpu/drm/xe/xe_bo_evict.c
@@ -47,25 +47,17 @@ static int xe_bo_apply_to_pinned(struct xe_device *xe,
 }
 
 /**
- * xe_bo_evict_all - evict all BOs from VRAM
- *
+ * xe_bo_evict_all_user - evict all non-pinned user BOs from VRAM
  * @xe: xe device
  *
- * Evict non-pinned user BOs first (via GPU), evict pinned external BOs next
- * (via GPU), wait for evictions, and finally evict pinned kernel BOs via CPU.
- * All eviction magic done via TTM calls.
+ * Evict non-pinned user BOs (via GPU).
  *
  * Evict == move VRAM BOs to temporary (typically system) memory.
- *
- * This function should be called before the device goes into a suspend state
- * where the VRAM loses power.
  */
-int xe_bo_evict_all(struct xe_device *xe)
+int xe_bo_evict_all_user(struct xe_device *xe)
 {
 	struct ttm_device *bdev = &xe->ttm;
-	struct xe_tile *tile;
 	u32 mem_type;
-	u8 id;
 	int ret;
 
 	/* User memory */
@@ -91,9 +83,34 @@ int xe_bo_evict_all(struct xe_device *xe)
 		}
 	}
 
-	ret = xe_bo_apply_to_pinned(xe, &xe->pinned.late.external,
-				    &xe->pinned.late.external,
-				    xe_bo_evict_pinned);
+	return 0;
+}
+
+/**
+ * xe_bo_evict_all - evict all BOs from VRAM
+ * @xe: xe device
+ *
+ * Evict non-pinned user BOs first (via GPU), evict pinned external BOs next
+ * (via GPU), wait for evictions, and finally evict pinned kernel BOs via CPU.
+ * All eviction magic done via TTM calls.
+ *
+ * Evict == move VRAM BOs to temporary (typically system) memory.
+ *
+ * This function should be called before the device goes into a suspend state
+ * where the VRAM loses power.
+ */
+int xe_bo_evict_all(struct xe_device *xe)
+{
+	struct xe_tile *tile;
+	u8 id;
+	int ret;
+
+	ret = xe_bo_evict_all_user(xe);
+	if (ret)
+		return ret;
+
+	ret = xe_bo_apply_to_pinned(xe, &xe->pinned.late.kernel_bo_present,
+				    &xe->pinned.late.evicted, xe_bo_evict_pinned);
 
 	if (!ret)
 		ret = xe_bo_apply_to_pinned(xe, &xe->pinned.late.kernel_bo_present,
diff --git a/drivers/gpu/drm/xe/xe_bo_evict.h b/drivers/gpu/drm/xe/xe_bo_evict.h
index d63eb3fc5cc9..e7f048634b32 100644
--- a/drivers/gpu/drm/xe/xe_bo_evict.h
+++ b/drivers/gpu/drm/xe/xe_bo_evict.h
@@ -9,6 +9,7 @@
 struct xe_device;
 
 int xe_bo_evict_all(struct xe_device *xe);
+int xe_bo_evict_all_user(struct xe_device *xe);
 int xe_bo_restore_early(struct xe_device *xe);
 int xe_bo_restore_late(struct xe_device *xe);
 
diff --git a/drivers/gpu/drm/xe/xe_device_types.h b/drivers/gpu/drm/xe/xe_device_types.h
index a42cb26e7d6d..3745389ead0d 100644
--- a/drivers/gpu/drm/xe/xe_device_types.h
+++ b/drivers/gpu/drm/xe/xe_device_types.h
@@ -522,6 +522,9 @@ struct xe_device {
 		struct mutex lock;
 	} d3cold;
 
+	/** @pm_notifier: Our PM notifier to perform actions in response to various PM events. */
+	struct notifier_block pm_notifier;
+
 	/** @pmt: Support the PMT driver callback interface */
 	struct {
 		/** @pmt.lock: protect access for telemetry data */
diff --git a/drivers/gpu/drm/xe/xe_pci.c b/drivers/gpu/drm/xe/xe_pci.c
index 07fe994f2a80..882398e09b7e 100644
--- a/drivers/gpu/drm/xe/xe_pci.c
+++ b/drivers/gpu/drm/xe/xe_pci.c
@@ -742,7 +742,7 @@ static void xe_pci_remove(struct pci_dev *pdev)
 		return;
 
 	xe_device_remove(xe);
-	xe_pm_runtime_fini(xe);
+	xe_pm_fini(xe);
 }
 
 /*
diff --git a/drivers/gpu/drm/xe/xe_pm.c b/drivers/gpu/drm/xe/xe_pm.c
index 4e112fbacada..d8a411d3ee96 100644
--- a/drivers/gpu/drm/xe/xe_pm.c
+++ b/drivers/gpu/drm/xe/xe_pm.c
@@ -286,6 +286,29 @@ static u32 vram_threshold_value(struct xe_device *xe)
 	return DEFAULT_VRAM_THRESHOLD;
 }
 
+static int xe_pm_notifier_callback(struct notifier_block *nb,
+				   unsigned long action, void *data)
+{
+	struct xe_device *xe = container_of(nb, struct xe_device, pm_notifier);
+	int err = 0;
+
+	switch (action) {
+	case PM_HIBERNATION_PREPARE:
+	case PM_SUSPEND_PREPARE:
+		xe_pm_runtime_get(xe);
+		err = xe_bo_evict_all_user(xe);
+		xe_pm_runtime_put(xe);
+		if (err)
+			drm_dbg(&xe->drm, "Notifier evict user failed (%d)\n", err);
+		break;
+	}
+
+	if (err)
+		return NOTIFY_BAD;
+
+	return NOTIFY_DONE;
+}
+
 /**
  * xe_pm_init - Initialize Xe Power Management
  * @xe: xe device instance
@@ -299,6 +322,11 @@ int xe_pm_init(struct xe_device *xe)
 	u32 vram_threshold;
 	int err;
 
+	xe->pm_notifier.notifier_call = xe_pm_notifier_callback;
+	err = register_pm_notifier(&xe->pm_notifier);
+	if (err)
+		return err;
+
 	/* For now suspend/resume is only allowed with GuC */
 	if (!xe_device_uc_enabled(xe))
 		return 0;
@@ -308,24 +336,23 @@ int xe_pm_init(struct xe_device *xe)
 	if (xe->d3cold.capable) {
 		err = xe_device_sysfs_init(xe);
 		if (err)
-			return err;
+			goto err_unregister;
 
 		vram_threshold = vram_threshold_value(xe);
 		err = xe_pm_set_vram_threshold(xe, vram_threshold);
 		if (err)
-			return err;
+			goto err_unregister;
 	}
 
 	xe_pm_runtime_init(xe);
-
 	return 0;
+
+err_unregister:
+	unregister_pm_notifier(&xe->pm_notifier);
+	return err;
 }
 
-/**
- * xe_pm_runtime_fini - Finalize Runtime PM
- * @xe: xe device instance
- */
-void xe_pm_runtime_fini(struct xe_device *xe)
+static void xe_pm_runtime_fini(struct xe_device *xe)
 {
 	struct device *dev = xe->drm.dev;
 
@@ -333,6 +360,18 @@ void xe_pm_runtime_fini(struct xe_device *xe)
 	pm_runtime_forbid(dev);
 }
 
+/**
+ * xe_pm_fini - Finalize PM
+ * @xe: xe device instance
+ */
+void xe_pm_fini(struct xe_device *xe)
+{
+	if (xe_device_uc_enabled(xe))
+		xe_pm_runtime_fini(xe);
+
+	unregister_pm_notifier(&xe->pm_notifier);
+}
+
 static void xe_pm_write_callback_task(struct xe_device *xe,
 				      struct task_struct *task)
 {
diff --git a/drivers/gpu/drm/xe/xe_pm.h b/drivers/gpu/drm/xe/xe_pm.h
index 998d1ed64556..59678b310e55 100644
--- a/drivers/gpu/drm/xe/xe_pm.h
+++ b/drivers/gpu/drm/xe/xe_pm.h
@@ -17,7 +17,7 @@ int xe_pm_resume(struct xe_device *xe);
 
 int xe_pm_init_early(struct xe_device *xe);
 int xe_pm_init(struct xe_device *xe);
-void xe_pm_runtime_fini(struct xe_device *xe);
+void xe_pm_fini(struct xe_device *xe);
 bool xe_pm_runtime_suspended(struct xe_device *xe);
 int xe_pm_runtime_suspend(struct xe_device *xe);
 int xe_pm_runtime_resume(struct xe_device *xe);
-- 
2.51.0


From e28647b677789b51b2de3e7934bb86dbd8e67fdf Mon Sep 17 00:00:00 2001
From: Matthew Auld <matthew.auld@intel.com>
Date: Wed, 16 Apr 2025 16:09:16 +0100
Subject: [PATCH 14/16] drm/xe: share bo dma-resv with backup object
MIME-Version: 1.0
Content-Type: text/plain; charset=utf8
Content-Transfer-Encoding: 8bit

We end up needing to grab both locks together anyway and keep them held
until we complete the copy or add the fence. Plus the backup_obj is
short lived and tied to the parent object, so seems reasonable to share
the same dma-resv. This will simplify the locking here, and in follow
up patches.

v2:
  - Hold reference to the parent bo to be sure the shared dma-resv can't
    go out of scope too soon. (Thomas)

Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Cc: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Link: https://lore.kernel.org/r/20250416150913.434369-7-matthew.auld@intel.com
---
 drivers/gpu/drm/xe/xe_bo.c       | 28 +++++++++++++---------------
 drivers/gpu/drm/xe/xe_bo_types.h |  2 ++
 2 files changed, 15 insertions(+), 15 deletions(-)

diff --git a/drivers/gpu/drm/xe/xe_bo.c b/drivers/gpu/drm/xe/xe_bo.c
index c337790c81ae..79adaee5a0e9 100644
--- a/drivers/gpu/drm/xe/xe_bo.c
+++ b/drivers/gpu/drm/xe/xe_bo.c
@@ -1120,13 +1120,15 @@ int xe_bo_evict_pinned(struct xe_bo *bo)
 	if (bo->flags & XE_BO_FLAG_PINNED_NORESTORE)
 		goto out_unlock_bo;
 
-	backup = xe_bo_create_locked(xe, NULL, NULL, bo->size, ttm_bo_type_kernel,
-				     XE_BO_FLAG_SYSTEM | XE_BO_FLAG_NEEDS_CPU_ACCESS |
-				     XE_BO_FLAG_PINNED);
+	backup = ___xe_bo_create_locked(xe, NULL, NULL, bo->ttm.base.resv, NULL, bo->size,
+					DRM_XE_GEM_CPU_CACHING_WB, ttm_bo_type_kernel,
+					XE_BO_FLAG_SYSTEM | XE_BO_FLAG_NEEDS_CPU_ACCESS |
+					XE_BO_FLAG_PINNED);
 	if (IS_ERR(backup)) {
 		ret = PTR_ERR(backup);
 		goto out_unlock_bo;
 	}
+	backup->parent_obj = xe_bo_get(bo); /* Released by bo_destroy */
 
 	if (xe_bo_is_user(bo) || (bo->flags & XE_BO_FLAG_PINNED_LATE_RESTORE)) {
 		struct xe_migrate *migrate;
@@ -1177,7 +1179,6 @@ int xe_bo_evict_pinned(struct xe_bo *bo)
 
 out_backup:
 	xe_bo_vunmap(backup);
-	xe_bo_unlock(backup);
 	if (ret)
 		xe_bo_put(backup);
 out_unlock_bo:
@@ -1212,17 +1213,12 @@ int xe_bo_restore_pinned(struct xe_bo *bo)
 	if (!backup)
 		return 0;
 
-	xe_bo_lock(backup, false);
+	xe_bo_lock(bo, false);
 
 	ret = ttm_bo_validate(&backup->ttm, &backup->placement, &ctx);
 	if (ret)
 		goto out_backup;
 
-	if (WARN_ON(!dma_resv_trylock(bo->ttm.base.resv))) {
-		ret = -EBUSY;
-		goto out_backup;
-	}
-
 	if (xe_bo_is_user(bo) || (bo->flags & XE_BO_FLAG_PINNED_LATE_RESTORE)) {
 		struct xe_migrate *migrate;
 		struct dma_fence *fence;
@@ -1271,15 +1267,14 @@ int xe_bo_restore_pinned(struct xe_bo *bo)
 
 	bo->backup_obj = NULL;
 
-out_unlock_bo:
-	if (unmap)
-		xe_bo_vunmap(bo);
-	xe_bo_unlock(bo);
 out_backup:
 	xe_bo_vunmap(backup);
-	xe_bo_unlock(backup);
 	if (!bo->backup_obj)
 		xe_bo_put(backup);
+out_unlock_bo:
+	if (unmap)
+		xe_bo_vunmap(bo);
+	xe_bo_unlock(bo);
 	return ret;
 }
 
@@ -1532,6 +1527,9 @@ static void xe_ttm_bo_destroy(struct ttm_buffer_object *ttm_bo)
 	if (bo->vm && xe_bo_is_user(bo))
 		xe_vm_put(bo->vm);
 
+	if (bo->parent_obj)
+		xe_bo_put(bo->parent_obj);
+
 	mutex_lock(&xe->mem_access.vram_userfault.lock);
 	if (!list_empty(&bo->vram_userfault_link))
 		list_del(&bo->vram_userfault_link);
diff --git a/drivers/gpu/drm/xe/xe_bo_types.h b/drivers/gpu/drm/xe/xe_bo_types.h
index 81396181aaea..eb5e83c5f233 100644
--- a/drivers/gpu/drm/xe/xe_bo_types.h
+++ b/drivers/gpu/drm/xe/xe_bo_types.h
@@ -30,6 +30,8 @@ struct xe_bo {
 	struct ttm_buffer_object ttm;
 	/** @backup_obj: The backup object when pinned and suspended (vram only) */
 	struct xe_bo *backup_obj;
+	/** @parent_obj: Ref to parent bo if this a backup_obj */
+	struct xe_bo *parent_obj;
 	/** @size: Size of this buffer object */
 	size_t size;
 	/** @flags: flags for this buffer object */
-- 
2.51.0


From 7e3f4a352337f1ce09c1795b58be23dfab29e610 Mon Sep 17 00:00:00 2001
From: Matthew Auld <matthew.auld@intel.com>
Date: Wed, 16 Apr 2025 16:09:17 +0100
Subject: [PATCH 15/16] drm/xe: handle pinned memory in PM notifier
MIME-Version: 1.0
Content-Type: text/plain; charset=utf8
Content-Transfer-Encoding: 8bit

Userspace is still alive and kicking at this point so actually moving
pinned stuff here is tricky. However, we can instead pre-allocate the
backup storage upfront from the notifier, such that we scoop up as much
as we can, and then leave the final .suspend() to do the actual copy (or
allocate anything that we missed). That way the bulk of our allocations
will hopefully be done outside the more restrictive .suspend().

We do need to be extra careful though, since the pinned handling can now
race with PM notifier, like something becoming unpinned after we prepare
it from the notifier.

v2 (Thomas):
  - Fix kernel doc and drop the pin as soon as we are done with the
    restore, instead of deferring to later.

Suggested-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Link: https://lore.kernel.org/r/20250416150913.434369-8-matthew.auld@intel.com
---
 drivers/gpu/drm/xe/xe_bo.c       | 123 +++++++++++++++++++++++++++----
 drivers/gpu/drm/xe/xe_bo.h       |   2 +
 drivers/gpu/drm/xe/xe_bo_evict.c |  51 ++++++++++++-
 drivers/gpu/drm/xe/xe_bo_evict.h |   2 +
 drivers/gpu/drm/xe/xe_pm.c       |  17 ++++-
 5 files changed, 176 insertions(+), 19 deletions(-)

diff --git a/drivers/gpu/drm/xe/xe_bo.c b/drivers/gpu/drm/xe/xe_bo.c
index 79adaee5a0e9..61420f863ac5 100644
--- a/drivers/gpu/drm/xe/xe_bo.c
+++ b/drivers/gpu/drm/xe/xe_bo.c
@@ -1084,6 +1084,80 @@ out_unref:
 	return lret;
 }
 
+/**
+ * xe_bo_notifier_prepare_pinned() - Prepare a pinned VRAM object to be backed
+ * up in system memory.
+ * @bo: The buffer object to prepare.
+ *
+ * On successful completion, the object backup pages are allocated. Expectation
+ * is that this is called from the PM notifier, prior to suspend/hibernation.
+ *
+ * Return: 0 on success. Negative error code on failure.
+ */
+int xe_bo_notifier_prepare_pinned(struct xe_bo *bo)
+{
+	struct xe_device *xe = ttm_to_xe_device(bo->ttm.bdev);
+	struct xe_bo *backup;
+	int ret = 0;
+
+	xe_bo_lock(bo, false);
+
+	xe_assert(xe, !bo->backup_obj);
+
+	/*
+	 * Since this is called from the PM notifier we might have raced with
+	 * someone unpinning this after we dropped the pinned list lock and
+	 * grabbing the above bo lock.
+	 */
+	if (!xe_bo_is_pinned(bo))
+		goto out_unlock_bo;
+
+	if (!xe_bo_is_vram(bo))
+		goto out_unlock_bo;
+
+	if (bo->flags & XE_BO_FLAG_PINNED_NORESTORE)
+		goto out_unlock_bo;
+
+	backup = ___xe_bo_create_locked(xe, NULL, NULL, bo->ttm.base.resv, NULL, bo->size,
+					DRM_XE_GEM_CPU_CACHING_WB, ttm_bo_type_kernel,
+					XE_BO_FLAG_SYSTEM | XE_BO_FLAG_NEEDS_CPU_ACCESS |
+					XE_BO_FLAG_PINNED);
+	if (IS_ERR(backup)) {
+		ret = PTR_ERR(backup);
+		goto out_unlock_bo;
+	}
+
+	backup->parent_obj = xe_bo_get(bo); /* Released by bo_destroy */
+	ttm_bo_pin(&backup->ttm);
+	bo->backup_obj = backup;
+
+out_unlock_bo:
+	xe_bo_unlock(bo);
+	return ret;
+}
+
+/**
+ * xe_bo_notifier_unprepare_pinned() - Undo the previous prepare operation.
+ * @bo: The buffer object to undo the prepare for.
+ *
+ * Always returns 0. The backup object is removed, if still present. Expectation
+ * it that this called from the PM notifier when undoing the prepare step.
+ *
+ * Return: Always returns 0.
+ */
+int xe_bo_notifier_unprepare_pinned(struct xe_bo *bo)
+{
+	xe_bo_lock(bo, false);
+	if (bo->backup_obj) {
+		ttm_bo_unpin(&bo->backup_obj->ttm);
+		xe_bo_put(bo->backup_obj);
+		bo->backup_obj = NULL;
+	}
+	xe_bo_unlock(bo);
+
+	return 0;
+}
+
 /**
  * xe_bo_evict_pinned() - Evict a pinned VRAM object to system memory
  * @bo: The buffer object to move.
@@ -1098,7 +1172,8 @@ out_unref:
 int xe_bo_evict_pinned(struct xe_bo *bo)
 {
 	struct xe_device *xe = ttm_to_xe_device(bo->ttm.bdev);
-	struct xe_bo *backup;
+	struct xe_bo *backup = bo->backup_obj;
+	bool backup_created = false;
 	bool unmap = false;
 	int ret = 0;
 
@@ -1120,15 +1195,18 @@ int xe_bo_evict_pinned(struct xe_bo *bo)
 	if (bo->flags & XE_BO_FLAG_PINNED_NORESTORE)
 		goto out_unlock_bo;
 
-	backup = ___xe_bo_create_locked(xe, NULL, NULL, bo->ttm.base.resv, NULL, bo->size,
-					DRM_XE_GEM_CPU_CACHING_WB, ttm_bo_type_kernel,
-					XE_BO_FLAG_SYSTEM | XE_BO_FLAG_NEEDS_CPU_ACCESS |
-					XE_BO_FLAG_PINNED);
-	if (IS_ERR(backup)) {
-		ret = PTR_ERR(backup);
-		goto out_unlock_bo;
+	if (!backup) {
+		backup = ___xe_bo_create_locked(xe, NULL, NULL, bo->ttm.base.resv, NULL, bo->size,
+						DRM_XE_GEM_CPU_CACHING_WB, ttm_bo_type_kernel,
+						XE_BO_FLAG_SYSTEM | XE_BO_FLAG_NEEDS_CPU_ACCESS |
+						XE_BO_FLAG_PINNED);
+		if (IS_ERR(backup)) {
+			ret = PTR_ERR(backup);
+			goto out_unlock_bo;
+		}
+		backup->parent_obj = xe_bo_get(bo); /* Released by bo_destroy */
+		backup_created = true;
 	}
-	backup->parent_obj = xe_bo_get(bo); /* Released by bo_destroy */
 
 	if (xe_bo_is_user(bo) || (bo->flags & XE_BO_FLAG_PINNED_LATE_RESTORE)) {
 		struct xe_migrate *migrate;
@@ -1175,11 +1253,12 @@ int xe_bo_evict_pinned(struct xe_bo *bo)
 				   bo->size);
 	}
 
-	bo->backup_obj = backup;
+	if (!bo->backup_obj)
+		bo->backup_obj = backup;
 
 out_backup:
 	xe_bo_vunmap(backup);
-	if (ret)
+	if (ret && backup_created)
 		xe_bo_put(backup);
 out_unlock_bo:
 	if (unmap)
@@ -1215,9 +1294,11 @@ int xe_bo_restore_pinned(struct xe_bo *bo)
 
 	xe_bo_lock(bo, false);
 
-	ret = ttm_bo_validate(&backup->ttm, &backup->placement, &ctx);
-	if (ret)
-		goto out_backup;
+	if (!xe_bo_is_pinned(backup)) {
+		ret = ttm_bo_validate(&backup->ttm, &backup->placement, &ctx);
+		if (ret)
+			goto out_unlock_bo;
+	}
 
 	if (xe_bo_is_user(bo) || (bo->flags & XE_BO_FLAG_PINNED_LATE_RESTORE)) {
 		struct xe_migrate *migrate;
@@ -1257,7 +1338,7 @@ int xe_bo_restore_pinned(struct xe_bo *bo)
 		if (iosys_map_is_null(&bo->vmap)) {
 			ret = xe_bo_vmap(bo);
 			if (ret)
-				goto out_unlock_bo;
+				goto out_backup;
 			unmap = true;
 		}
 
@@ -1269,8 +1350,11 @@ int xe_bo_restore_pinned(struct xe_bo *bo)
 
 out_backup:
 	xe_bo_vunmap(backup);
-	if (!bo->backup_obj)
+	if (!bo->backup_obj) {
+		if (xe_bo_is_pinned(backup))
+			ttm_bo_unpin(&backup->ttm);
 		xe_bo_put(backup);
+	}
 out_unlock_bo:
 	if (unmap)
 		xe_bo_vunmap(bo);
@@ -2304,6 +2388,13 @@ void xe_bo_unpin(struct xe_bo *bo)
 		xe_assert(xe, !list_empty(&bo->pinned_link));
 		list_del_init(&bo->pinned_link);
 		spin_unlock(&xe->pinned.lock);
+
+		if (bo->backup_obj) {
+			if (xe_bo_is_pinned(bo->backup_obj))
+				ttm_bo_unpin(&bo->backup_obj->ttm);
+			xe_bo_put(bo->backup_obj);
+			bo->backup_obj = NULL;
+		}
 	}
 	ttm_bo_unpin(&bo->ttm);
 	if (bo->ttm.ttm && ttm_tt_is_populated(bo->ttm.ttm))
diff --git a/drivers/gpu/drm/xe/xe_bo.h b/drivers/gpu/drm/xe/xe_bo.h
index 0a19b50045b2..8bc449c78cc7 100644
--- a/drivers/gpu/drm/xe/xe_bo.h
+++ b/drivers/gpu/drm/xe/xe_bo.h
@@ -277,6 +277,8 @@ int xe_bo_migrate(struct xe_bo *bo, u32 mem_type);
 int xe_bo_evict(struct xe_bo *bo, bool force_alloc);
 
 int xe_bo_evict_pinned(struct xe_bo *bo);
+int xe_bo_notifier_prepare_pinned(struct xe_bo *bo);
+int xe_bo_notifier_unprepare_pinned(struct xe_bo *bo);
 int xe_bo_restore_pinned(struct xe_bo *bo);
 
 int xe_bo_dma_unmap_pinned(struct xe_bo *bo);
diff --git a/drivers/gpu/drm/xe/xe_bo_evict.c b/drivers/gpu/drm/xe/xe_bo_evict.c
index 748360fd2439..ed3746d32b27 100644
--- a/drivers/gpu/drm/xe/xe_bo_evict.c
+++ b/drivers/gpu/drm/xe/xe_bo_evict.c
@@ -34,7 +34,13 @@ static int xe_bo_apply_to_pinned(struct xe_device *xe,
 		ret = pinned_fn(bo);
 		if (ret && pinned_list != new_list) {
 			spin_lock(&xe->pinned.lock);
-			list_move(&bo->pinned_link, pinned_list);
+			/*
+			 * We might no longer be pinned, since PM notifier can
+			 * call this. If the pinned link is now empty, keep it
+			 * that way.
+			 */
+			if (!list_empty(&bo->pinned_link))
+				list_move(&bo->pinned_link, pinned_list);
 			spin_unlock(&xe->pinned.lock);
 		}
 		xe_bo_put(bo);
@@ -46,6 +52,49 @@ static int xe_bo_apply_to_pinned(struct xe_device *xe,
 	return ret;
 }
 
+/**
+ * xe_bo_notifier_prepare_all_pinned() - Pre-allocate the backing pages for all
+ * pinned VRAM objects which need to be saved.
+ * @xe: xe device
+ *
+ * Should be called from PM notifier when preparing for s3/s4.
+ *
+ * Return: 0 on success, negative error code on error.
+ */
+int xe_bo_notifier_prepare_all_pinned(struct xe_device *xe)
+{
+	int ret;
+
+	ret = xe_bo_apply_to_pinned(xe, &xe->pinned.early.kernel_bo_present,
+				    &xe->pinned.early.kernel_bo_present,
+				    xe_bo_notifier_prepare_pinned);
+	if (!ret)
+		ret = xe_bo_apply_to_pinned(xe, &xe->pinned.late.kernel_bo_present,
+					    &xe->pinned.late.kernel_bo_present,
+					    xe_bo_notifier_prepare_pinned);
+
+	return ret;
+}
+
+/**
+ * xe_bo_notifier_unprepare_all_pinned() - Remove the backing pages for all
+ * pinned VRAM objects which have been restored.
+ * @xe: xe device
+ *
+ * Should be called from PM notifier after exiting s3/s4 (either on success or
+ * failure).
+ */
+void xe_bo_notifier_unprepare_all_pinned(struct xe_device *xe)
+{
+	(void)xe_bo_apply_to_pinned(xe, &xe->pinned.early.kernel_bo_present,
+				    &xe->pinned.early.kernel_bo_present,
+				    xe_bo_notifier_unprepare_pinned);
+
+	(void)xe_bo_apply_to_pinned(xe, &xe->pinned.late.kernel_bo_present,
+				    &xe->pinned.late.kernel_bo_present,
+				    xe_bo_notifier_unprepare_pinned);
+}
+
 /**
  * xe_bo_evict_all_user - evict all non-pinned user BOs from VRAM
  * @xe: xe device
diff --git a/drivers/gpu/drm/xe/xe_bo_evict.h b/drivers/gpu/drm/xe/xe_bo_evict.h
index e7f048634b32..e8385cb7f5e9 100644
--- a/drivers/gpu/drm/xe/xe_bo_evict.h
+++ b/drivers/gpu/drm/xe/xe_bo_evict.h
@@ -10,6 +10,8 @@ struct xe_device;
 
 int xe_bo_evict_all(struct xe_device *xe);
 int xe_bo_evict_all_user(struct xe_device *xe);
+int xe_bo_notifier_prepare_all_pinned(struct xe_device *xe);
+void xe_bo_notifier_unprepare_all_pinned(struct xe_device *xe);
 int xe_bo_restore_early(struct xe_device *xe);
 int xe_bo_restore_late(struct xe_device *xe);
 
diff --git a/drivers/gpu/drm/xe/xe_pm.c b/drivers/gpu/drm/xe/xe_pm.c
index d8a411d3ee96..38514cef817e 100644
--- a/drivers/gpu/drm/xe/xe_pm.c
+++ b/drivers/gpu/drm/xe/xe_pm.c
@@ -297,9 +297,22 @@ static int xe_pm_notifier_callback(struct notifier_block *nb,
 	case PM_SUSPEND_PREPARE:
 		xe_pm_runtime_get(xe);
 		err = xe_bo_evict_all_user(xe);
-		xe_pm_runtime_put(xe);
-		if (err)
+		if (err) {
 			drm_dbg(&xe->drm, "Notifier evict user failed (%d)\n", err);
+			xe_pm_runtime_put(xe);
+			break;
+		}
+
+		err = xe_bo_notifier_prepare_all_pinned(xe);
+		if (err) {
+			drm_dbg(&xe->drm, "Notifier prepare pin failed (%d)\n", err);
+			xe_pm_runtime_put(xe);
+		}
+		break;
+	case PM_POST_HIBERNATION:
+	case PM_POST_SUSPEND:
+		xe_bo_notifier_unprepare_all_pinned(xe);
+		xe_pm_runtime_put(xe);
 		break;
 	}
 
-- 
2.51.0


From 4ea512714c42c69828b4a2647d206bf404043ad5 Mon Sep 17 00:00:00 2001
From: Jeevaka Prabu Badrappan <jeevaka.badrappan@intel.com>
Date: Tue, 22 Apr 2025 17:18:52 +0000
Subject: [PATCH 16/16] drm/xe: Fix CFI violation when accessing sysfs files

When an attribute group is created with sysfs_create_group() or
sysfs_create_files() the ->sysfs_ops() callback is set to
kobj_sysfs_ops, which sets the ->show() callback to kobj_attr_show().
kobj_attr_show() uses container_of() to get the ->show() callback
from the attribute it was passed, meaning the ->show() callback needs
to be the same type as the ->show() callback in 'struct kobj_attribute'.

However, cur_freq_show() has the type of the ->show() callback in
'struct device_attribute', which causes a CFI violation when opening the
'id' sysfs node under gtidle/freq/throttle. This happens to work because
the layout of 'struct kobj_attribute' and 'struct device_attribute' are
the same, so the container_of() cast happens to allow the ->show()
callback to still work.

Changed the type of cur_freq_show() and few more functions to match the
->show() callback in 'struct kobj_attributes' to resolve the CFI
violation.

CFI failure seen while accessing sysfs files under
/sys/class/drm/card0/device/tile0/gt*/gtidle/*
/sys/class/drm/card0/device/tile0/gt*/freq0/*
/sys/class/drm/card0/device/tile0/gt*/freq0/throttle/*

[ 2599.618075] RIP: 0010:__cfi_cur_freq_show+0xd/0x10 [xe]
[ 2599.624452] Code: 44 c1 44 89 fa e8 03 95 39 f2 48 98 5b 41 5e 41 5f 5d c3 c9
[ 2599.646638] RSP: 0018:ffffbe438ead7d10 EFLAGS: 00010286
[ 2599.652823] RAX: ffff9f7d8b3845d8 RBX: ffff9f7dee8c95d8 RCX: 0000000000000000
[ 2599.661246] RDX: ffff9f7e6f439000 RSI: ffffffffc13ada30 RDI: ffff9f7d975d4b00
[ 2599.669669] RBP: ffffbe438ead7d18 R08: 0000000000001000 R09: ffff9f7e6f439000
[ 2599.678092] R10: 00000000e07304a6 R11: ffffffffc1241ca0 R12: ffffffffb4836ea0
[ 2599.688435] R13: ffff9f7e45fb1180 R14: ffff9f7d975d4b00 R15: ffff9f7e6f439000
[ 2599.696860] FS: 000076b02b66cfc0(0000) GS:ffff9f80ef400000(0000) knlGS:00000
[ 2599.706412] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 2599.713196] CR2: 00005f80d94641a9 CR3: 00000001e44ec006 CR4: 0000000100f72ef0
[ 2599.721618] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 2599.730041] DR3: 0000000000000000 DR6: 00000000ffff07f0 DR7: 0000000000000400
[ 2599.738464] PKRU: 55555554
[ 2599.741655] Call Trace:
[ 2599.744541] <TASK>
[ 2599.747017] ? __die_body+0x69/0xb0
[ 2599.751151] ? die+0xa9/0xd0
[ 2599.754548] ? do_trap+0x89/0x160
[ 2599.758476] ? __cfi_cur_freq_show+0xd/0x10 [xe b37985c94829727668bd7c5b33c1]
[ 2599.768315] ? handle_invalid_op+0x69/0x90
[ 2599.773167] ? __cfi_cur_freq_show+0xd/0x10 [xe b37985c94829727668bd7c5b33c1]
[ 2599.783010] ? exc_invalid_op+0x36/0x60
[ 2599.787552] ? fred_hwexc+0x123/0x1a0
[ 2599.791873] ? fred_entry_from_kernel+0x7b/0xd0
[ 2599.797219] ? asm_fred_entrypoint_kernel+0x45/0x70
[ 2599.802976] ? act_freq_show+0x70/0x70 [xe b37985c94829727668bd7c5b33c1d9998]
[ 2599.812301] ? __cfi_cur_freq_show+0xd/0x10 [xe b37985c94829727668bd7c5b33c1]
[ 2599.822137] ? __kmalloc_node_noprof+0x1f3/0x420
[ 2599.827594] ? __kvmalloc_node_noprof+0xcb/0x180
[ 2599.833045] ? kobj_attr_show+0x22/0x40
[ 2599.837571] sysfs_kf_seq_show+0xa8/0x110
[ 2599.842302] kernfs_seq_show+0x38/0x50

Signed-off-by: Jeevaka Prabu Badrappan <jeevaka.badrappan@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://lore.kernel.org/r/20250422171852.85558-1-jeevaka.badrappan@intel.com
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
---
 drivers/gpu/drm/xe/xe_gt_freq.c     | 82 ++++++++++++++------------
 drivers/gpu/drm/xe/xe_gt_idle.c     | 28 +++++----
 drivers/gpu/drm/xe/xe_gt_throttle.c | 90 ++++++++++++++---------------
 3 files changed, 107 insertions(+), 93 deletions(-)

diff --git a/drivers/gpu/drm/xe/xe_gt_freq.c b/drivers/gpu/drm/xe/xe_gt_freq.c
index 604bdc7c8173..868a5d2c1a52 100644
--- a/drivers/gpu/drm/xe/xe_gt_freq.c
+++ b/drivers/gpu/drm/xe/xe_gt_freq.c
@@ -56,9 +56,10 @@ dev_to_xe(struct device *dev)
 	return gt_to_xe(kobj_to_gt(dev->kobj.parent));
 }
 
-static ssize_t act_freq_show(struct device *dev,
-			     struct device_attribute *attr, char *buf)
+static ssize_t act_freq_show(struct kobject *kobj,
+			     struct kobj_attribute *attr, char *buf)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_guc_pc *pc = dev_to_pc(dev);
 	u32 freq;
 
@@ -68,11 +69,12 @@ static ssize_t act_freq_show(struct device *dev,
 
 	return sysfs_emit(buf, "%d\n", freq);
 }
-static DEVICE_ATTR_RO(act_freq);
+static struct kobj_attribute attr_act_freq = __ATTR_RO(act_freq);
 
-static ssize_t cur_freq_show(struct device *dev,
-			     struct device_attribute *attr, char *buf)
+static ssize_t cur_freq_show(struct kobject *kobj,
+			     struct kobj_attribute *attr, char *buf)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_guc_pc *pc = dev_to_pc(dev);
 	u32 freq;
 	ssize_t ret;
@@ -85,11 +87,12 @@ static ssize_t cur_freq_show(struct device *dev,
 
 	return sysfs_emit(buf, "%d\n", freq);
 }
-static DEVICE_ATTR_RO(cur_freq);
+static struct kobj_attribute attr_cur_freq = __ATTR_RO(cur_freq);
 
-static ssize_t rp0_freq_show(struct device *dev,
-			     struct device_attribute *attr, char *buf)
+static ssize_t rp0_freq_show(struct kobject *kobj,
+			     struct kobj_attribute *attr, char *buf)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_guc_pc *pc = dev_to_pc(dev);
 	u32 freq;
 
@@ -99,11 +102,12 @@ static ssize_t rp0_freq_show(struct device *dev,
 
 	return sysfs_emit(buf, "%d\n", freq);
 }
-static DEVICE_ATTR_RO(rp0_freq);
+static struct kobj_attribute attr_rp0_freq = __ATTR_RO(rp0_freq);
 
-static ssize_t rpe_freq_show(struct device *dev,
-			     struct device_attribute *attr, char *buf)
+static ssize_t rpe_freq_show(struct kobject *kobj,
+			     struct kobj_attribute *attr, char *buf)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_guc_pc *pc = dev_to_pc(dev);
 	u32 freq;
 
@@ -113,11 +117,12 @@ static ssize_t rpe_freq_show(struct device *dev,
 
 	return sysfs_emit(buf, "%d\n", freq);
 }
-static DEVICE_ATTR_RO(rpe_freq);
+static struct kobj_attribute attr_rpe_freq = __ATTR_RO(rpe_freq);
 
-static ssize_t rpa_freq_show(struct device *dev,
-			     struct device_attribute *attr, char *buf)
+static ssize_t rpa_freq_show(struct kobject *kobj,
+			     struct kobj_attribute *attr, char *buf)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_guc_pc *pc = dev_to_pc(dev);
 	u32 freq;
 
@@ -127,20 +132,22 @@ static ssize_t rpa_freq_show(struct device *dev,
 
 	return sysfs_emit(buf, "%d\n", freq);
 }
-static DEVICE_ATTR_RO(rpa_freq);
+static struct kobj_attribute attr_rpa_freq = __ATTR_RO(rpa_freq);
 
-static ssize_t rpn_freq_show(struct device *dev,
-			     struct device_attribute *attr, char *buf)
+static ssize_t rpn_freq_show(struct kobject *kobj,
+			     struct kobj_attribute *attr, char *buf)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_guc_pc *pc = dev_to_pc(dev);
 
 	return sysfs_emit(buf, "%d\n", xe_guc_pc_get_rpn_freq(pc));
 }
-static DEVICE_ATTR_RO(rpn_freq);
+static struct kobj_attribute attr_rpn_freq = __ATTR_RO(rpn_freq);
 
-static ssize_t min_freq_show(struct device *dev,
-			     struct device_attribute *attr, char *buf)
+static ssize_t min_freq_show(struct kobject *kobj,
+			     struct kobj_attribute *attr, char *buf)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_guc_pc *pc = dev_to_pc(dev);
 	u32 freq;
 	ssize_t ret;
@@ -154,9 +161,10 @@ static ssize_t min_freq_show(struct device *dev,
 	return sysfs_emit(buf, "%d\n", freq);
 }
 
-static ssize_t min_freq_store(struct device *dev, struct device_attribute *attr,
-			      const char *buff, size_t count)
+static ssize_t min_freq_store(struct kobject *kobj,
+			      struct kobj_attribute *attr, const char *buff, size_t count)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_guc_pc *pc = dev_to_pc(dev);
 	u32 freq;
 	ssize_t ret;
@@ -173,11 +181,12 @@ static ssize_t min_freq_store(struct device *dev, struct device_attribute *attr,
 
 	return count;
 }
-static DEVICE_ATTR_RW(min_freq);
+static struct kobj_attribute attr_min_freq = __ATTR_RW(min_freq);
 
-static ssize_t max_freq_show(struct device *dev,
-			     struct device_attribute *attr, char *buf)
+static ssize_t max_freq_show(struct kobject *kobj,
+			     struct kobj_attribute *attr, char *buf)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_guc_pc *pc = dev_to_pc(dev);
 	u32 freq;
 	ssize_t ret;
@@ -191,9 +200,10 @@ static ssize_t max_freq_show(struct device *dev,
 	return sysfs_emit(buf, "%d\n", freq);
 }
 
-static ssize_t max_freq_store(struct device *dev, struct device_attribute *attr,
-			      const char *buff, size_t count)
+static ssize_t max_freq_store(struct kobject *kobj,
+			      struct kobj_attribute *attr, const char *buff, size_t count)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_guc_pc *pc = dev_to_pc(dev);
 	u32 freq;
 	ssize_t ret;
@@ -210,17 +220,17 @@ static ssize_t max_freq_store(struct device *dev, struct device_attribute *attr,
 
 	return count;
 }
-static DEVICE_ATTR_RW(max_freq);
+static struct kobj_attribute attr_max_freq = __ATTR_RW(max_freq);
 
 static const struct attribute *freq_attrs[] = {
-	&dev_attr_act_freq.attr,
-	&dev_attr_cur_freq.attr,
-	&dev_attr_rp0_freq.attr,
-	&dev_attr_rpa_freq.attr,
-	&dev_attr_rpe_freq.attr,
-	&dev_attr_rpn_freq.attr,
-	&dev_attr_min_freq.attr,
-	&dev_attr_max_freq.attr,
+	&attr_act_freq.attr,
+	&attr_cur_freq.attr,
+	&attr_rp0_freq.attr,
+	&attr_rpa_freq.attr,
+	&attr_rpe_freq.attr,
+	&attr_rpn_freq.attr,
+	&attr_min_freq.attr,
+	&attr_max_freq.attr,
 	NULL
 };
 
diff --git a/drivers/gpu/drm/xe/xe_gt_idle.c b/drivers/gpu/drm/xe/xe_gt_idle.c
index fbbace7b0b12..c11206410a4d 100644
--- a/drivers/gpu/drm/xe/xe_gt_idle.c
+++ b/drivers/gpu/drm/xe/xe_gt_idle.c
@@ -249,9 +249,10 @@ int xe_gt_idle_pg_print(struct xe_gt *gt, struct drm_printer *p)
 	return 0;
 }
 
-static ssize_t name_show(struct device *dev,
-			 struct device_attribute *attr, char *buff)
+static ssize_t name_show(struct kobject *kobj,
+			 struct kobj_attribute *attr, char *buff)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_gt_idle *gtidle = dev_to_gtidle(dev);
 	struct xe_guc_pc *pc = gtidle_to_pc(gtidle);
 	ssize_t ret;
@@ -262,11 +263,12 @@ static ssize_t name_show(struct device *dev,
 
 	return ret;
 }
-static DEVICE_ATTR_RO(name);
+static struct kobj_attribute name_attr = __ATTR_RO(name);
 
-static ssize_t idle_status_show(struct device *dev,
-				struct device_attribute *attr, char *buff)
+static ssize_t idle_status_show(struct kobject *kobj,
+				struct kobj_attribute *attr, char *buff)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_gt_idle *gtidle = dev_to_gtidle(dev);
 	struct xe_guc_pc *pc = gtidle_to_pc(gtidle);
 	enum xe_gt_idle_state state;
@@ -277,6 +279,7 @@ static ssize_t idle_status_show(struct device *dev,
 
 	return sysfs_emit(buff, "%s\n", gt_idle_state_to_string(state));
 }
+static struct kobj_attribute idle_status_attr = __ATTR_RO(idle_status);
 
 u64 xe_gt_idle_residency_msec(struct xe_gt_idle *gtidle)
 {
@@ -291,10 +294,11 @@ u64 xe_gt_idle_residency_msec(struct xe_gt_idle *gtidle)
 	return residency;
 }
 
-static DEVICE_ATTR_RO(idle_status);
-static ssize_t idle_residency_ms_show(struct device *dev,
-				      struct device_attribute *attr, char *buff)
+
+static ssize_t idle_residency_ms_show(struct kobject *kobj,
+				      struct kobj_attribute *attr, char *buff)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_gt_idle *gtidle = dev_to_gtidle(dev);
 	struct xe_guc_pc *pc = gtidle_to_pc(gtidle);
 	u64 residency;
@@ -305,12 +309,12 @@ static ssize_t idle_residency_ms_show(struct device *dev,
 
 	return sysfs_emit(buff, "%llu\n", residency);
 }
-static DEVICE_ATTR_RO(idle_residency_ms);
+static struct kobj_attribute idle_residency_attr = __ATTR_RO(idle_residency_ms);
 
 static const struct attribute *gt_idle_attrs[] = {
-	&dev_attr_name.attr,
-	&dev_attr_idle_status.attr,
-	&dev_attr_idle_residency_ms.attr,
+	&name_attr.attr,
+	&idle_status_attr.attr,
+	&idle_residency_attr.attr,
 	NULL,
 };
 
diff --git a/drivers/gpu/drm/xe/xe_gt_throttle.c b/drivers/gpu/drm/xe/xe_gt_throttle.c
index 8db78d616b6f..aa962c783cdf 100644
--- a/drivers/gpu/drm/xe/xe_gt_throttle.c
+++ b/drivers/gpu/drm/xe/xe_gt_throttle.c
@@ -114,115 +114,115 @@ static u32 read_reason_vr_tdc(struct xe_gt *gt)
 	return tdc;
 }
 
-static ssize_t status_show(struct device *dev,
-			   struct device_attribute *attr,
-			   char *buff)
+static ssize_t status_show(struct kobject *kobj,
+			   struct kobj_attribute *attr, char *buff)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_gt *gt = dev_to_gt(dev);
 	bool status = !!read_status(gt);
 
 	return sysfs_emit(buff, "%u\n", status);
 }
-static DEVICE_ATTR_RO(status);
+static struct kobj_attribute attr_status = __ATTR_RO(status);
 
-static ssize_t reason_pl1_show(struct device *dev,
-			       struct device_attribute *attr,
-			       char *buff)
+static ssize_t reason_pl1_show(struct kobject *kobj,
+			       struct kobj_attribute *attr, char *buff)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_gt *gt = dev_to_gt(dev);
 	bool pl1 = !!read_reason_pl1(gt);
 
 	return sysfs_emit(buff, "%u\n", pl1);
 }
-static DEVICE_ATTR_RO(reason_pl1);
+static struct kobj_attribute attr_reason_pl1 = __ATTR_RO(reason_pl1);
 
-static ssize_t reason_pl2_show(struct device *dev,
-			       struct device_attribute *attr,
-			       char *buff)
+static ssize_t reason_pl2_show(struct kobject *kobj,
+			       struct kobj_attribute *attr, char *buff)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_gt *gt = dev_to_gt(dev);
 	bool pl2 = !!read_reason_pl2(gt);
 
 	return sysfs_emit(buff, "%u\n", pl2);
 }
-static DEVICE_ATTR_RO(reason_pl2);
+static struct kobj_attribute attr_reason_pl2 = __ATTR_RO(reason_pl2);
 
-static ssize_t reason_pl4_show(struct device *dev,
-			       struct device_attribute *attr,
-			       char *buff)
+static ssize_t reason_pl4_show(struct kobject *kobj,
+			       struct kobj_attribute *attr, char *buff)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_gt *gt = dev_to_gt(dev);
 	bool pl4 = !!read_reason_pl4(gt);
 
 	return sysfs_emit(buff, "%u\n", pl4);
 }
-static DEVICE_ATTR_RO(reason_pl4);
+static struct kobj_attribute attr_reason_pl4 = __ATTR_RO(reason_pl4);
 
-static ssize_t reason_thermal_show(struct device *dev,
-				   struct device_attribute *attr,
-				   char *buff)
+static ssize_t reason_thermal_show(struct kobject *kobj,
+				   struct kobj_attribute *attr, char *buff)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_gt *gt = dev_to_gt(dev);
 	bool thermal = !!read_reason_thermal(gt);
 
 	return sysfs_emit(buff, "%u\n", thermal);
 }
-static DEVICE_ATTR_RO(reason_thermal);
+static struct kobj_attribute attr_reason_thermal = __ATTR_RO(reason_thermal);
 
-static ssize_t reason_prochot_show(struct device *dev,
-				   struct device_attribute *attr,
-				   char *buff)
+static ssize_t reason_prochot_show(struct kobject *kobj,
+				   struct kobj_attribute *attr, char *buff)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_gt *gt = dev_to_gt(dev);
 	bool prochot = !!read_reason_prochot(gt);
 
 	return sysfs_emit(buff, "%u\n", prochot);
 }
-static DEVICE_ATTR_RO(reason_prochot);
+static struct kobj_attribute attr_reason_prochot = __ATTR_RO(reason_prochot);
 
-static ssize_t reason_ratl_show(struct device *dev,
-				struct device_attribute *attr,
-				char *buff)
+static ssize_t reason_ratl_show(struct kobject *kobj,
+				struct kobj_attribute *attr, char *buff)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_gt *gt = dev_to_gt(dev);
 	bool ratl = !!read_reason_ratl(gt);
 
 	return sysfs_emit(buff, "%u\n", ratl);
 }
-static DEVICE_ATTR_RO(reason_ratl);
+static struct kobj_attribute attr_reason_ratl = __ATTR_RO(reason_ratl);
 
-static ssize_t reason_vr_thermalert_show(struct device *dev,
-					 struct device_attribute *attr,
-					 char *buff)
+static ssize_t reason_vr_thermalert_show(struct kobject *kobj,
+					 struct kobj_attribute *attr, char *buff)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_gt *gt = dev_to_gt(dev);
 	bool thermalert = !!read_reason_vr_thermalert(gt);
 
 	return sysfs_emit(buff, "%u\n", thermalert);
 }
-static DEVICE_ATTR_RO(reason_vr_thermalert);
+static struct kobj_attribute attr_reason_vr_thermalert = __ATTR_RO(reason_vr_thermalert);
 
-static ssize_t reason_vr_tdc_show(struct device *dev,
-				  struct device_attribute *attr,
-				  char *buff)
+static ssize_t reason_vr_tdc_show(struct kobject *kobj,
+				  struct kobj_attribute *attr, char *buff)
 {
+	struct device *dev = kobj_to_dev(kobj);
 	struct xe_gt *gt = dev_to_gt(dev);
 	bool tdc = !!read_reason_vr_tdc(gt);
 
 	return sysfs_emit(buff, "%u\n", tdc);
 }
-static DEVICE_ATTR_RO(reason_vr_tdc);
+static struct kobj_attribute attr_reason_vr_tdc = __ATTR_RO(reason_vr_tdc);
 
 static struct attribute *throttle_attrs[] = {
-	&dev_attr_status.attr,
-	&dev_attr_reason_pl1.attr,
-	&dev_attr_reason_pl2.attr,
-	&dev_attr_reason_pl4.attr,
-	&dev_attr_reason_thermal.attr,
-	&dev_attr_reason_prochot.attr,
-	&dev_attr_reason_ratl.attr,
-	&dev_attr_reason_vr_thermalert.attr,
-	&dev_attr_reason_vr_tdc.attr,
+	&attr_status.attr,
+	&attr_reason_pl1.attr,
+	&attr_reason_pl2.attr,
+	&attr_reason_pl4.attr,
+	&attr_reason_thermal.attr,
+	&attr_reason_prochot.attr,
+	&attr_reason_ratl.attr,
+	&attr_reason_vr_thermalert.attr,
+	&attr_reason_vr_tdc.attr,
 	NULL
 };
 
-- 
2.51.0