Skip to contentSkip to conversation

drm/xe/svm: Add MMU notifier-based madvise autoreset on munmap

RFC New Arvind Yadav 3 participants 17 messages Jun 4 – Jun 5

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

v3 cover drm/xe/svm: Add MMU notifier-based madvise autoreset on munmap

This series implements automatic reset of MADVISE-managed GPU memory attributes when userspace unmaps memory. Non-default attributes are reset when CPU mappings are removed, preventing them from persisting across later mmap() operations.

MMU interval notifiers detect munmap on CPU-only VMAs (not yet GPU-faulted) and queue a worker to reset attributes. GPU-touched VMAs continue using the existing SVM path.

Key aspects:
- cpu_autoreset_active tracks CPU-only vs GPU-touched state.
- MMU notifier callbacks do not allocate. work_struct is embedded in xe_madvise_notifier and initialized at ioctl time.
- Unmap events are coalesced in the callback and processed by a worker.
- teardown_rwsem serializes VM teardown against notifier callbacks.
- maple_tree tracks notifier ranges and supports deduplication.
- MMU notifier registration is performed outside vm->lock.

Patch organization:
1. Track cpu_autoreset_active in xe_vma.
2. Preserve state across GPUVA operations.
3. Clear state on first GPU fault.
4. Add MMU notifier and worker infrastructure.
5. Disable madvise notifier after GPU touch.
6. Wire notifiers into VM lifecycle.
7. Fix partial unmap attribute reset.

Changes in v2:
- Move runtime state to xe_vma bool cpu_autoreset_active. (Matt)
- Embed work_struct in xe_madvise_notifier. (Thomas)
- Coalesce overlapping munmap events.
- Replace closing state with teardown_rwsem. (Matt)
- Use maple_tree for notifier tracking. (Matt)
- Register notifiers outside vm->lock.
- Add xe_vma_effective_create_flags() and fix REMAP split handling.
- Move GPU-touch handling to the successful pagefault path. (Matt)
- Use plain bool active in notifier. (Matt)
- Add lockdep assertions for vm->lock.

Changes in v3:
- Add kernel-doc and lockdep_assert_held to xe_vma_has_cpu_autoreset_active(). (Matt)
- Guard cpu_autoreset_active assignment to CPU_ADDR_MIRROR VMAs only. (Matt)
- Drop lockdep_assert_held_write from xe_svm_handle_pagefault. (Matt)
- Rework GPU-touch handling after page faults. (Matt)
- Rework notifier lifetime handling and teardown.
- Keep exact-match notifier deactivation for split VMA handling.
- Move notifier_ranges allocation before VMA mutation.
- Keep a VM-owned notifier list for teardown.
- Coalesce pending unmap ranges and reset only CPU holes in the worker.
- Clear MADV_AUTORESET on notifier registration failure.
- Coalesce partial_unmap range via min/max instead of blind overwrite. (Matt)
- Restore range_debug() in xe_svm_garbage_collector_add_range(). (Matt)
- Update partial_unmap under garbage_collector.lock.
- Continue processing queued ranges on non-fatal set_default_attr failures.
- Annotate lockless notifier->active access with READ_ONCE/WRITE_ONCE.

Arvind Yadav (6):
drm/xe/vm: Track CPU_AUTORESET state in xe_vma drm/xe/vm: Preserve cpu_autoreset_active across GPUVA operations drm/xe/svm: Clear CPU_AUTORESET_ACTIVE on first GPU fault drm/xe/vm: Add madvise autoreset notifier worker drm/xe/vm: Disable madvise notifier on GPU touch drm/xe/vm: Wire MADVISE_AUTORESET notifiers into VM lifecycle

Himal Prasad Ghimiray (1):
drm/xe/svm: Correct memory attribute reset for partial unmap

 drivers/gpu/drm/xe/xe_pagefault.c  |  13 +-
 drivers/gpu/drm/xe/xe_svm.c        | 102 ++++-
 drivers/gpu/drm/xe/xe_svm.h        |  10 +
 drivers/gpu/drm/xe/xe_vm.c         |  58 ++-
 drivers/gpu/drm/xe/xe_vm.h         |  28 ++
 drivers/gpu/drm/xe/xe_vm_madvise.c | 651 ++++++++++++++++++++++++++++-
 drivers/gpu/drm/xe/xe_vm_madvise.h |   9 +
 drivers/gpu/drm/xe/xe_vm_types.h   |  82 ++++
 8 files changed, 928 insertions(+), 25 deletions(-)
Signature · 1 line
2.43.0
#1 lore raw

RFC v3 1/7 drm/xe/vm: Track CPU_AUTORESET state in xe_vma

Track MADVISE_AUTORESET CPU-only state in struct xe_vma with a dedicated cpu_autoreset_active bool.

Add XE_VMA_CPU_AUTORESET_ACTIVE as a pipeline-only bit to carry this state through MAP/REMAP operations. Runtime state lives in xe_vma.cpu_autoreset_active rather than vma->gpuva.flags.

The state is set at bind time and cleared on the first successful GPU fault.

v2:
- Move runtime state from gpuva.flags to xe_vma bool. (Matt)
- Keep XE_VMA_CPU_AUTORESET_ACTIVE pipeline-only. (Matt)

v3:
- Add kernel-doc and lockdep_assert_held to
xe_vma_has_cpu_autoreset_active(). (Matt)

Cc Matthew Brost <[email protected]>
Cc Thomas Hellström <[email protected]>
Cc Himal Prasad Ghimiray <[email protected]>
---
 drivers/gpu/drm/xe/xe_vm.h       | 14 ++++++++++++++
 drivers/gpu/drm/xe/xe_vm_types.h | 13 +++++++++++++
 2 files changed, 27 insertions(+)
drivers/gpu/drm/xe/xe_vm.h +14 -0
index c5b900f38ded..0dd201cce968 100644
@@ -181,6 +181,20 @@ static inline bool xe_vma_is_userptr(struct xe_vma *vma)
 		!xe_vma_is_cpu_addr_mirror(vma);
 }
 
+/**
+ * xe_vma_has_cpu_autoreset_active() - Check if VMA is still CPU-only
+ * @vma: VMA to check
+ *
+ * The caller must hold vm->lock (read or write).
+ *
+ * Return: true if CPU-only autoreset is armed, false after the first GPU touch.
+ */
+static inline bool xe_vma_has_cpu_autoreset_active(struct xe_vma *vma)
+{
+	lockdep_assert_held(&xe_vma_vm(vma)->lock);
+	return vma->cpu_autoreset_active;
+}
+
 struct xe_vma *xe_vm_find_vma_by_addr(struct xe_vm *vm, u64 page_addr);
 
 int xe_vma_need_vram_for_atomic(struct xe_device *xe, struct xe_vma *vma, bool is_atomic);
drivers/gpu/drm/xe/xe_vm_types.h +13 -0
index 635ed29b9a69..e6380458272e 100644
@@ -52,6 +52,11 @@ struct xe_vm_pgtable_update_op;
 #define XE_VMA_DUMPABLE		(DRM_GPUVA_USERBITS << 8)
 #define XE_VMA_SYSTEM_ALLOCATOR	(DRM_GPUVA_USERBITS << 9)
 #define XE_VMA_MADV_AUTORESET	(DRM_GPUVA_USERBITS << 10)
+/*
+ * Pipeline-only bit used to carry cpu_autoreset_active through MAP/REMAP.
+ * It must not be stored persistently in vma->gpuva.flags.
+ */
+#define XE_VMA_CPU_AUTORESET_ACTIVE	(DRM_GPUVA_USERBITS << 11)
 
 /**
  * struct xe_vma_mem_attr - memory attributes associated with vma
@@ -157,6 +162,14 @@ struct xe_vma {
 	/** @tile_staged: bind is staged for this VMA */
 	u8 tile_staged;
 
+	/**
+	 * @cpu_autoreset_active: CPU mirror VMA has not been GPU-faulted yet.
+	 *
+	 * Set for MADVISE_AUTORESET CPU mirror VMAs and cleared on the first
+	 * successful GPU fault. Protected by vm->lock.
+	 */
+	bool cpu_autoreset_active;
+
 	/**
 	 * @skip_invalidation: Used in madvise to avoid invalidation
 	 * if mem attributes doesn't change
Signature · 1 line
2.43.0
#2 lore raw

RFC v3 2/7 drm/xe/vm: Preserve cpu_autoreset_active across GPUVA operations

In reply to Arvind Yadav

GPUVA split and remap operations rebuild VMAs from create flags. These flags preserve persistent VMA state, but not runtime-only state such as cpu_autoreset_active.

Forward XE_VMA_CPU_AUTORESET_ACTIVE explicitly through the MAP/REMAP pipeline so xe_vma_create() can restore cpu_autoreset_active in the new VMA. The bit remains pipeline-only and is stripped before storing vma->gpuva.flags.

Also relax the UNMAP attribute warning for MADVISE_AUTORESET VMAs, since they may legitimately carry non-default madvise-managed attributes.

v2:
- Move runtime state to xe_vma bool and keep XE_VMA_CPU_AUTORESET_ACTIVE as pipeline-only. (Matt)
- Add xe_vma_effective_create_flags() to centralise flag handling.

v3:
- Guard cpu_autoreset_active assignment to CPU_ADDR_MIRROR VMAs only (Matt)

Cc Matthew Brost <[email protected]>
Cc Thomas Hellström <[email protected]>
Cc Himal Prasad Ghimiray <[email protected]>
---
 drivers/gpu/drm/xe/xe_vm.c | 44 ++++++++++++++++++++++++++++++++------
 1 file changed, 38 insertions(+), 6 deletions(-)
drivers/gpu/drm/xe/xe_vm.c +38 -6
index 080c2fff0e95..dec5279f08a2 100644
@@ -1105,7 +1105,11 @@ static struct xe_vma *xe_vma_create(struct xe_vm *vm,
 	vma->gpuva.vm = &vm->gpuvm;
 	vma->gpuva.va.addr = start;
 	vma->gpuva.va.range = end - start + 1;
-	vma->gpuva.flags = flags;
+	/* Pipeline-only, do not store in gpuva.flags. */
+	vma->gpuva.flags = flags & ~XE_VMA_CPU_AUTORESET_ACTIVE;
+	vma->cpu_autoreset_active =
+		(flags & XE_VMA_SYSTEM_ALLOCATOR) &&
+		(flags & XE_VMA_CPU_AUTORESET_ACTIVE);
 
 	for_each_tile(tile, vm->xe, id)
 		vma->tile_mask |= 0x1 << id;
@@ -2479,8 +2483,10 @@ vm_bind_ioctl_ops_create(struct xe_vm *vm, struct xe_vma_ops *vops,
 				op->map.vma_flags |= XE_VMA_SYSTEM_ALLOCATOR;
 			if (flags & DRM_XE_VM_BIND_FLAG_DUMPABLE)
 				op->map.vma_flags |= XE_VMA_DUMPABLE;
-			if (flags & DRM_XE_VM_BIND_FLAG_MADVISE_AUTORESET)
+			if (flags & DRM_XE_VM_BIND_FLAG_MADVISE_AUTORESET) {
 				op->map.vma_flags |= XE_VMA_MADV_AUTORESET;
+				op->map.vma_flags |= XE_VMA_CPU_AUTORESET_ACTIVE;
+			}
 			op->map.request_decompress = flags & DRM_XE_VM_BIND_FLAG_DECOMPRESS;
 			op->map.pat_index = pat_index;
 			op->map.invalidate_on_bind =
@@ -2800,6 +2806,9 @@ static int vm_bind_ioctl_ops_parse(struct xe_vm *vm, struct drm_gpuva_ops *ops,
 			};
 
 			flags |= op->map.vma_flags & XE_VMA_CREATE_MASK;
+			/* Forward pipeline-only state. */
+			if (op->map.vma_flags & XE_VMA_CPU_AUTORESET_ACTIVE)
+				flags |= XE_VMA_CPU_AUTORESET_ACTIVE;
 
 			vma = new_vma(vm, &op->base.map, &default_attr,
 				      flags);
@@ -2842,6 +2851,10 @@ static int vm_bind_ioctl_ops_parse(struct xe_vm *vm, struct drm_gpuva_ops *ops,
 			op->remap.old_range = op->remap.range;
 
 			flags |= op->base.remap.unmap->va->flags & XE_VMA_CREATE_MASK;
+			/* Forward pipeline-only state. */
+			if (xe_vma_has_cpu_autoreset_active(old))
+				flags |= XE_VMA_CPU_AUTORESET_ACTIVE;
+
 			if (op->base.remap.prev) {
 				vma = new_vma(vm, op->base.remap.prev,
 					      &old->attr, flags);
@@ -4693,6 +4706,17 @@ int xe_vma_need_vram_for_atomic(struct xe_device *xe, struct xe_vma *vma, bool i
 	}
 }
 
+/* Add pipeline-only autoreset state when rebuilding a VMA. */
+static unsigned int xe_vma_effective_create_flags(struct xe_vma *vma)
+{
+	unsigned int flags = vma->gpuva.flags;
+
+	if (xe_vma_has_cpu_autoreset_active(vma))
+		flags |= XE_VMA_CPU_AUTORESET_ACTIVE;
+
+	return flags;
+}
+
 static int xe_vm_alloc_vma(struct xe_vm *vm,
 			   struct drm_gpuvm_map_req *map_req,
 			   bool is_madvise)
@@ -4728,19 +4752,24 @@ static int xe_vm_alloc_vma(struct xe_vm *vm,
 		if (!is_madvise) {
 			if (__op->op == DRM_GPUVA_OP_UNMAP) {
 				vma = gpuva_to_vma(op->base.unmap.va);
-				XE_WARN_ON(!xe_vma_has_default_mem_attrs(vma));
+				/* AUTORESET VMAs may carry madvise-managed attrs. */
+				XE_WARN_ON(!xe_vma_has_default_mem_attrs(vma) &&
+					   !(vma->gpuva.flags & XE_VMA_MADV_AUTORESET));
 				default_pat = vma->attr.default_pat_index;
-				vma_flags = vma->gpuva.flags;
+				vma_flags = xe_vma_effective_create_flags(vma);
 			}
 
 			if (__op->op == DRM_GPUVA_OP_REMAP) {
 				vma = gpuva_to_vma(op->base.remap.unmap->va);
 				default_pat = vma->attr.default_pat_index;
-				vma_flags = vma->gpuva.flags;
+				vma_flags = xe_vma_effective_create_flags(vma);
 			}
 
 			if (__op->op == DRM_GPUVA_OP_MAP) {
 				op->map.vma_flags |= vma_flags & XE_VMA_CREATE_MASK;
+				/* Forward pipeline-only state. */
+				if (vma_flags & XE_VMA_CPU_AUTORESET_ACTIVE)
+					op->map.vma_flags |= XE_VMA_CPU_AUTORESET_ACTIVE;
 				op->map.pat_index = default_pat;
 			}
 		} else {
@@ -4749,7 +4778,7 @@ static int xe_vm_alloc_vma(struct xe_vm *vm,
 				xe_assert(vm->xe, !remap_op);
 				xe_assert(vm->xe, xe_vma_has_no_bo(vma));
 				remap_op = true;
-				vma_flags = vma->gpuva.flags;
+				vma_flags = xe_vma_effective_create_flags(vma);
 			}
 
 			if (__op->op == DRM_GPUVA_OP_MAP) {
@@ -4762,6 +4791,9 @@ static int xe_vm_alloc_vma(struct xe_vm *vm,
 				 * unmapping.
 				 */
 				op->map.vma_flags |= vma_flags & XE_VMA_CREATE_MASK;
+				/* Forward pipeline-only state. */
+				if (vma_flags & XE_VMA_CPU_AUTORESET_ACTIVE)
+					op->map.vma_flags |= XE_VMA_CPU_AUTORESET_ACTIVE;
 			}
 		}
 		print_op(vm->xe, __op);
Signature · 1 line
2.43.0
#3 lore raw

Re: [RFC v3 2/7] drm/xe/vm: Preserve cpu_autoreset_active across GPUVA operations

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Pre-existing issues:
- [Critical] Merging adjacent VMAs during CPU mirror pagefaults silently clobbers creation flags (e.g., READ_ONLY, MADV_AUTORESET), permanently altering memory properties and potentially leading to privilege escalation. --

commit 1d97ca671d925d0481b9d913f989dc8216596164
Author: Arvind Yadav <[email protected]>

drm/xe/vm: Preserve cpu_autoreset_active across GPUVA operations

GPUVA split and remap operations rebuild VMAs from create flags. This commit explicitly forwards XE_VMA_CPU_AUTORESET_ACTIVE through the MAP/REMAP pipeline so xe_vma_create() can restore cpu_autoreset_active. It also relaxes the UNMAP attribute warning for MADVISE_AUTORESET VMAs.

Quoted 3 lines
drivers/gpu/drm/xe/xe_vm.c +0 -0

[ ... ]

Quoted 12 lines
drivers/gpu/drm/xe/xe_vm.c +4 -2
@@ -4728,19 +4752,24 @@ static int xe_vm_alloc_vma(struct xe_vm *vm,
 		if (!is_madvise) {
 			if (__op->op == DRM_GPUVA_OP_UNMAP) {
 				vma = gpuva_to_vma(op->base.unmap.va);
-				XE_WARN_ON(!xe_vma_has_default_mem_attrs(vma));
+				/* AUTORESET VMAs may carry madvise-managed attrs. */
+				XE_WARN_ON(!xe_vma_has_default_mem_attrs(vma) &&
+					   !(vma->gpuva.flags & XE_VMA_MADV_AUTORESET));
 				default_pat = vma->attr.default_pat_index;
-				vma_flags = vma->gpuva.flags;
+				vma_flags = xe_vma_effective_create_flags(vma);
 			}

drivers/gpu/drm/xe/xe_vm.c static int xe_vm_alloc_vma(struct xe_vm *vm, · line 4769

Quoted 10 lines
drivers/gpu/drm/xe/xe_vm.c +1 -1
 
 			if (__op->op == DRM_GPUVA_OP_REMAP) {
 				vma = gpuva_to_vma(op->base.remap.unmap->va);
 				default_pat = vma->attr.default_pat_index;
-				vma_flags = vma->gpuva.flags;
+				vma_flags = xe_vma_effective_create_flags(vma);
 			}
 
 			if (__op->op == DRM_GPUVA_OP_MAP) {
 				op->map.vma_flags |= vma_flags & XE_VMA_CREATE_MASK;

[Severity: Critical]
This is a pre-existing issue, but does this loop silently clobber creation flags like XE_VMA_READ_ONLY when merging adjacent VMAs during CPU mirror pagefaults?

When xe_svm_handle_pagefault() attempts to expand a VMA mapping, it calls xe_vm_find_cpu_addr_mirror_vma_range(). That function coalesces adjacent VMAs using is_cpu_addr_vma_with_default_attr():

drivers/gpu/drm/xe/xe_vm.c:is_cpu_addr_vma_with_default_attr() {
	return vma && xe_vma_is_cpu_addr_mirror(vma) &&
	       xe_vma_has_default_mem_attrs(vma);
}

This check verifies the system allocator flag and default memory attributes, but it doesn't ensure that vma->gpuva.flags (such as XE_VMA_READ_ONLY or XE_VMA_DUMPABLE) match between the adjacent VMAs.

As a result, xe_vm_alloc_vma() creates a single MAP operation for the merged range. In the drm_gpuva_for_each_op() loop here, the local vma_flags variable is continuously overwritten by each UNMAP and REMAP operation.

When the final MAP operation processes this line, it inherits flags exclusively from the last processed VMA. If a read-write VMA is merged with an adjacent read-only VMA, is the XE_VMA_READ_ONLY flag completely lost for the coalesced region, potentially granting writable access to a mapping originally requested as read-only?

Quoted 6 lines
drivers/gpu/drm/xe/xe_vm.c +3 -0
+				/* Forward pipeline-only state. */
+				if (vma_flags & XE_VMA_CPU_AUTORESET_ACTIVE)
+					op->map.vma_flags |= XE_VMA_CPU_AUTORESET_ACTIVE;
 				op->map.pat_index = default_pat;
 			}
 		} else {
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=2
#4 lore raw

RFC v3 3/7 drm/xe/svm: Clear CPU_AUTORESET_ACTIVE on first GPU fault

In reply to Arvind Yadav

CPU address mirror VMAs start with cpu_autoreset_active set, indicating they are still CPU-only.

Clear cpu_autoreset_active after successful fault handling. A successful return means the fault address already has a valid GPU mapping or PTEs were installed. Prefetch faults that do not establish a mapping return an error and do not transition the state.

v2:
- Move xe_vma_gpu_touch() to the success path in xe_svm_handle_pagefault() so prefetch faults that find no range do not transition the state. (Matt)
- Add xe_vma_gpu_touch() helper in xe_vm.h and use

    vma->cpu_autoreset_active instead of vma->gpuva.flags. (Matt)

v3:
- Drop lockdep_assert_held_write from xe_svm_handle_pagefault. (Matt)
- Re-fetch the VMA in xe_pagefault_service() before clearing state.
- Move gpu_touch from xe_svm_handle_pagefault to xe_pagefault_service

    (Matt)
Cc Matthew Brost <[email protected]>
Cc Thomas Hellström <[email protected]>
Cc Himal Prasad Ghimiray <[email protected]>
---
 drivers/gpu/drm/xe/xe_pagefault.c | 10 +++++++++-
 drivers/gpu/drm/xe/xe_vm.h        | 13 +++++++++++++
 2 files changed, 22 insertions(+), 1 deletion(-)
drivers/gpu/drm/xe/xe_pagefault.c +9 -1
index dd3c068e1a39..f64d3df08261 100644
@@ -215,8 +215,16 @@ static int xe_pagefault_service(struct xe_pagefault *pf)
 		err = xe_pagefault_handle_vma(gt, vma, atomic);
 
 unlock_vm:
-	if (!err)
+	if (!err) {
+		/* Re-fetch, fault handling may have replaced the VMA. */
+		vma = xe_vm_find_vma_by_addr(vm, pf->consumer.page_addr);
 		vm->usm.last_fault_vma = vma;
+
+		/* First successful GPU fault ends CPU-only state. */
+		if (vma && xe_vma_is_cpu_addr_mirror(vma) &&
+		    xe_vma_has_cpu_autoreset_active(vma))
+			xe_vma_gpu_touch(vma);
+	}
 	up_write(&vm->lock);
 	xe_vm_put(vm);
 
drivers/gpu/drm/xe/xe_vm.h +13 -0
index 0dd201cce968..0ad704b05687 100644
@@ -450,4 +450,17 @@ static inline struct drm_exec *xe_vm_validation_exec(struct xe_vm *vm)
 	((READ_ONCE(tile_present) & ~READ_ONCE(tile_invalidated)) & BIT((tile)->id))
 
 void xe_vma_mem_attr_copy(struct xe_vma_mem_attr *to, struct xe_vma_mem_attr *from);
+
+/**
+ * xe_vma_gpu_touch() - Mark a VMA as no longer CPU-only
+ * @vma: VMA to update
+ *
+ * Clear cpu_autoreset_active after the first successful GPU fault-in.
+ * Caller must hold vm->lock in write mode.
+ */
+static inline void xe_vma_gpu_touch(struct xe_vma *vma)
+{
+	lockdep_assert_held_write(&xe_vma_vm(vma)->lock);
+	vma->cpu_autoreset_active = false;
+}
 #endif
Signature · 1 line
2.43.0
#5 lore raw

RFC v3 4/7 drm/xe/vm: Add madvise autoreset notifier worker

In reply to Arvind Yadav

Reset madvise-managed VMA attributes when userspace unmaps CPU-only mirror ranges.

The MMU notifier callback cannot take vm->lock and does not allocate. It records a pending range in the preallocated notifier and queues the embedded worker.

If multiple unmap events arrive before the worker runs, the pending range is widened. The worker then walks the CPU mm and resets only real holes, leaving ranges still covered by a CPU VMA unchanged.

Use a maple tree for range lookup and a VM-owned list for notifier lifetime, so teardown does not depend on maple-tree entries.

v2:
- Replace closing state with teardown_rwsem. (Matt)
- Use maple_tree for notifier tracking. (Matt)
- Embed work_struct in notifier; no allocation in callback. (Thomas)
- Coalesce overlapping munmap events via min/max.
- Run notifier removal and workqueue drain outside teardown_rwsem. (Matt)

v3:
- Keep a VM-owned notifier list for teardown.
- Widen pending ranges in the callback and reset only CPU holes in the worker.

Cc Matthew Brost <[email protected]>
Cc Thomas Hellström <[email protected]>
Cc Himal Prasad Ghimiray <[email protected]>
---
 drivers/gpu/drm/xe/xe_vm_madvise.c | 504 +++++++++++++++++++++++++++++
 drivers/gpu/drm/xe/xe_vm_madvise.h |   8 +
 drivers/gpu/drm/xe/xe_vm_types.h   |  61 ++++
 3 files changed, 573 insertions(+)
drivers/gpu/drm/xe/xe_vm_madvise.c +504 -0
index c4fb29004195..c2abe712598a 100644
@@ -6,6 +6,8 @@
 #include "xe_vm_madvise.h"
 
 #include <linux/nospec.h>
+#include <linux/maple_tree.h>
+#include <linux/workqueue.h>
 #include <drm/xe_drm.h>
 
 #include "xe_bo.h"
@@ -14,6 +16,10 @@
 #include "xe_svm.h"
 #include "xe_tlb_inval.h"
 #include "xe_vm.h"
+#include "xe_macros.h"
+
+/* Lockdep class for teardown_rwsem */
+static struct lock_class_key xe_madvise_teardown_key;
 
 struct xe_vmas_in_madvise_range {
 	u64 addr;
@@ -732,3 +738,501 @@ int xe_vm_madvise_ioctl(struct drm_device *dev, void *data, struct drm_file *fil
 	xe_vm_put(vm);
 	return err;
 }
+
+/**
+ * xe_vma_reset_to_default_attrs - Reset madvise attrs to defaults
+ * @vma: VMA to reset
+ */
+static void xe_vma_reset_to_default_attrs(struct xe_vma *vma)
+{
+	struct xe_vma_mem_attr default_attr = {
+		.preferred_loc.devmem_fd = DRM_XE_PREFERRED_LOC_DEFAULT_DEVICE,
+		.preferred_loc.migration_policy = DRM_XE_MIGRATE_ALL_PAGES,
+		.default_pat_index = vma->attr.default_pat_index,
+		.pat_index = vma->attr.default_pat_index,
+		.atomic_access = DRM_XE_ATOMIC_UNDEFINED,
+		.purgeable_state = XE_MADV_PURGEABLE_WILLNEED,
+	};
+
+	xe_vma_mem_attr_copy(&vma->attr, &default_attr);
+}
+
+/**
+ * xe_vm_madvise_process_unmap - Reset attrs for a GPUVA range
+ * @vm: VM
+ * @start: start of range
+ * @end: end of range
+ *
+ * Process CPU-only VMAs overlapping [@start, @end).
+ *
+ * Return: 0 on success, negative error otherwise.
+ */
+static int xe_vm_madvise_process_unmap(struct xe_vm *vm, u64 start, u64 end)
+{
+	u64 addr = start;
+	int err;
+
+	lockdep_assert_held_write(&vm->lock);
+
+	if (xe_vm_is_closed_or_banned(vm))
+		return 0;
+
+	while (addr < end) {
+		struct xe_vma *vma;
+		u64 seg_start, seg_end;
+		bool has_default_attr;
+
+		vma = xe_vm_find_overlapping_vma(vm, addr, end - addr);
+		if (!vma)
+			break;
+
+		/* GPU-touched VMAs are handled by SVM. */
+		if (!xe_vma_has_cpu_autoreset_active(vma)) {
+			addr = xe_vma_end(vma);
+			continue;
+		}
+
+		has_default_attr = xe_vma_has_default_mem_attrs(vma);
+		seg_start = max(addr, xe_vma_start(vma));
+		seg_end = min(end, xe_vma_end(vma));
+
+		/* Merge adjacent default-attr VMAs when possible. */
+		if (has_default_attr &&
+		    xe_vma_start(vma) >= start &&
+		    xe_vma_end(vma) <= end) {
+			seg_start = xe_vma_start(vma);
+			seg_end = xe_vma_end(vma);
+			xe_vm_find_cpu_addr_mirror_vma_range(vm, &seg_start, &seg_end);
+			if (xe_vma_start(vma) == seg_start && xe_vma_end(vma) == seg_end) {
+				/* Nothing to merge. */
+				addr = seg_end;
+				continue;
+			}
+		} else if (xe_vma_start(vma) == seg_start && xe_vma_end(vma) == seg_end) {
+			/* Exact VMA match, reset in place. */
+			xe_vma_reset_to_default_attrs(vma);
+			addr = seg_end;
+			continue;
+		}
+
+		err = xe_vm_alloc_cpu_addr_mirror_vma(vm, seg_start, seg_end - seg_start);
+		if (err) {
+			if (err == -ENOENT) {
+				/* VMA was removed before the worker ran. */
+				addr = seg_end;
+				continue;
+			}
+			return err;
+		}
+
+		addr = seg_end;
+	}
+
+	return 0;
+}
+
+/**
+ * xe_vm_madvise_process_unmap_holes - Reset attrs for CPU holes
+ * @vm: VM
+ * @mm: mm backing the CPU mirror
+ * @start: start of the pending interval
+ * @end: end of the pending interval
+ *
+ * Walk [@start, @end) and process only ranges not covered by a CPU VMA.
+ * Mapped ranges are skipped so partial-unmap siblings keep their attrs.
+ *
+ * Caller must hold vm->lock for write and mmap_read_lock(@mm).
+ *
+ * Return: 0 on success, negative error otherwise.
+ */
+static int xe_vm_madvise_process_unmap_holes(struct xe_vm *vm,
+					     struct mm_struct *mm,
+					     u64 start, u64 end)
+{
+	u64 addr = start;
+
+	lockdep_assert_held_write(&vm->lock);
+	mmap_assert_locked(mm);
+
+	while (addr < end) {
+		struct vm_area_struct *cpu_vma;
+		u64 hole_start, hole_end;
+		int err;
+
+		cpu_vma = find_vma(mm, addr);
+
+		if (cpu_vma && cpu_vma->vm_start <= addr) {
+			addr = min_t(u64, cpu_vma->vm_end, end);
+			continue;
+		}
+
+		hole_start = addr;
+		hole_end = cpu_vma ? min_t(u64, cpu_vma->vm_start, end) : end;
+
+		err = xe_vm_madvise_process_unmap(vm, hole_start, hole_end);
+		if (err)
+			return err;
+
+		addr = hole_end;
+	}
+
+	return 0;
+}
+
+/**
+ * xe_madvise_work_func - Worker to process pending unmap events
+ * @w: work_struct embedded in xe_madvise_notifier
+ *
+ * Drains pending intervals recorded by the callback. The worker loops so
+ * events queued while it is running are not lost.
+ */
+static void xe_madvise_work_func(struct work_struct *w)
+{
+	struct xe_madvise_notifier *notifier =
+		container_of(w, struct xe_madvise_notifier, work);
+	struct xe_vm *vm = notifier->vm;
+
+	for (;;) {
+		struct mm_struct *mm;
+		u64 start, end;
+		int err;
+
+		spin_lock(&notifier->work_lock);
+		if (!notifier->work_pending) {
+			spin_unlock(&notifier->work_lock);
+			break;
+		}
+		start = notifier->work_start;
+		end = notifier->work_end;
+		notifier->work_pending = false;
+		spin_unlock(&notifier->work_lock);
+
+		/* The mm is going away, teardown will clean up. */
+		mm = vm->svm.gpusvm.mm;
+		if (!mm || !mmget_not_zero(mm))
+			break;
+
+		down_write(&vm->lock);
+		mmap_read_lock(mm);
+
+		err = xe_vm_madvise_process_unmap_holes(vm, mm, start, end);
+
+		mmap_read_unlock(mm);
+		up_write(&vm->lock);
+		mmput(mm);
+
+		if (err)
+			drm_warn(&vm->xe->drm,
+				 "madvise autoreset failed [%#llx-%#llx]: %d\n",
+				 start, end, err);
+	}
+}
+
+/**
+ * xe_madvise_notifier_callback - MMU notifier callback for CPU munmap
+ * @mni: mmu_interval_notifier
+ * @range: mmu_notifier_range
+ * @cur_seq: current sequence number
+ *
+ * Records one pending interval without allocating. Later events widen it.
+ * The worker checks the CPU mm before resetting attributes.
+ *
+ * Return: false for non-blockable invalidations, true otherwise.
+ */
+static bool xe_madvise_notifier_callback(struct mmu_interval_notifier *mni,
+					 const struct mmu_notifier_range *range,
+					 unsigned long cur_seq)
+{
+	struct xe_madvise_notifier *notifier =
+		container_of(mni, struct xe_madvise_notifier, mmu_notifier);
+	struct xe_vm *vm = notifier->vm;
+	u64 adj_start, adj_end;
+
+	if (range->event != MMU_NOTIFY_UNMAP)
+		return true;
+
+	if (!mmu_notifier_range_blockable(range))
+		return false;
+
+	if (xe_vm_is_closed(vm))
+		return true;
+
+	mmu_interval_set_seq(mni, cur_seq);
+
+	/* Clamp to notifier boundaries and ignore non-overlap. */
+	adj_start = max_t(u64, range->start, notifier->vma_start);
+	adj_end = min_t(u64, range->end, notifier->vma_end);
+
+	if (adj_start >= adj_end)
+		return true;
+
+	/* Bail if teardown started; trylock fails once fini holds write. */
+	if (!down_read_trylock(&vm->svm.madvise_work.teardown_rwsem))
+		return true;
+
+	/* fini may have NULLed wq before we got here; check under read lock. */
+	if (!vm->svm.madvise_work.wq)
+		goto out;
+
+	spin_lock(&notifier->work_lock);
+	if (notifier->work_pending) {
+		/*
+		 * Widen pending work. The worker only resets CPU holes,
+		 * so mapped siblings are left untouched.
+		 */
+		notifier->work_start = min(notifier->work_start, adj_start);
+		notifier->work_end = max(notifier->work_end, adj_end);
+	} else {
+		notifier->work_start = adj_start;
+		notifier->work_end = adj_end;
+		notifier->work_pending = true;
+	}
+	spin_unlock(&notifier->work_lock);
+
+	queue_work(vm->svm.madvise_work.wq, &notifier->work);
+
+out:
+	up_read(&vm->svm.madvise_work.teardown_rwsem);
+	return true;
+}
+
+static const struct mmu_interval_notifier_ops xe_madvise_notifier_ops = {
+	.invalidate = xe_madvise_notifier_callback,
+};
+
+/**
+ * xe_vm_madvise_init - Initialize madvise notifier infrastructure
+ * @vm: VM
+ *
+ * Sets up workqueue for async munmap processing.
+ *
+ * Return: 0 on success, -ENOMEM on failure
+ */
+int xe_vm_madvise_init(struct xe_vm *vm)
+{
+	/* Already initialized. */
+	if (vm->svm.madvise_work.wq)
+		return 0;
+
+	mt_init(&vm->svm.madvise_notifiers);
+	INIT_LIST_HEAD(&vm->svm.madvise_notifier_list);
+
+	/* Separate class for notifier teardown. */
+	__init_rwsem(&vm->svm.madvise_work.teardown_rwsem,
+		     "xe_madvise_teardown", &xe_madvise_teardown_key);
+
+	/* Not used from reclaim paths. */
+	vm->svm.madvise_work.wq = alloc_workqueue("xe_madvise", WQ_UNBOUND, 0);
+	if (!vm->svm.madvise_work.wq) {
+		mtree_destroy(&vm->svm.madvise_notifiers);
+		return -ENOMEM;
+	}
+
+	return 0;
+}
+
+static void xe_madvise_notifier_free(struct xe_madvise_notifier *notifier)
+{
+	xe_vm_put(notifier->vm);
+	kfree(notifier);
+}
+
+static void xe_madvise_notifier_remove_and_free(struct xe_madvise_notifier *notifier)
+{
+	mmu_interval_notifier_remove(&notifier->mmu_notifier);
+	cancel_work_sync(&notifier->work);
+	xe_madvise_notifier_free(notifier);
+}
+
+static struct xe_madvise_notifier *
+xe_madvise_notifier_alloc(struct xe_vm *vm, u64 start, u64 end)
+{
+	struct xe_madvise_notifier *notifier;
+
+	notifier = kzalloc_obj(*notifier, GFP_KERNEL);
+	if (!notifier)
+		return NULL;
+
+	notifier->vm = xe_vm_get(vm);
+	notifier->vma_start = start;
+	notifier->vma_end = end;
+	INIT_LIST_HEAD(&notifier->link);
+	spin_lock_init(&notifier->work_lock);
+	notifier->work_pending = false;
+	INIT_WORK(&notifier->work, xe_madvise_work_func);
+
+	return notifier;
+}
+
+static bool xe_madvise_notifier_exact(const struct xe_madvise_notifier *notifier,
+				      u64 start, u64 end)
+{
+	return notifier->vma_start == start && notifier->vma_end == end;
+}
+
+static bool xe_madvise_notifier_fully_covered(const struct xe_madvise_notifier *notifier,
+					      u64 start, u64 end)
+{
+	/*
+	 * Broader notifiers may still cover split siblings, so only remove
+	 * notifiers fully covered by the new range.
+	 */
+	return notifier->vma_start >= start && notifier->vma_end <= end;
+}
+
+/**
+ * xe_vm_madvise_fini - Cleanup all madvise notifiers
+ * @vm: VM
+ *
+ * Tears down notifiers and drains workqueue. Safe if init partially failed.
+ */
+void xe_vm_madvise_fini(struct xe_vm *vm)
+{
+	struct xe_madvise_notifier *notifier, *next;
+	struct workqueue_struct *wq;
+	LIST_HEAD(tmp);
+
+	/* Nothing to do if init never ran. */
+	if (!vm->svm.madvise_work.wq)
+		return;
+
+	/* Block new callbacks. */
+	down_write(&vm->svm.madvise_work.teardown_rwsem);
+
+	/* Stage all owned notifiers from the VM list. */
+	list_for_each_entry_safe(notifier, next,
+				 &vm->svm.madvise_notifier_list, link) {
+		list_del_init(&notifier->link);
+		list_add_tail(&notifier->link, &tmp);
+	}
+
+	/* VM is closed; safe to destroy the tree. */
+	mtree_destroy(&vm->svm.madvise_notifiers);
+
+	/* NULL wq so late callbacks bail. */
+	wq = vm->svm.madvise_work.wq;
+	vm->svm.madvise_work.wq = NULL;
+
+	up_write(&vm->svm.madvise_work.teardown_rwsem);
+
+	/*
+	 * Remove notifiers outside rwsem; remove() may block on mmap_lock.
+	 */
+	list_for_each_entry(notifier, &tmp, link)
+		mmu_interval_notifier_remove(&notifier->mmu_notifier);
+
+	/* Drain work before freeing; workers reference notifier via container_of. */
+	if (wq) {
+		drain_workqueue(wq);
+		destroy_workqueue(wq);
+	}
+
+	/* Safe to free now: no callbacks can fire, no workers are running. */
+	list_for_each_entry_safe(notifier, next, &tmp, link) {
+		list_del(&notifier->link);
+		xe_madvise_notifier_free(notifier);
+	}
+}
+
+/**
+ * xe_vm_madvise_register_notifier_range - Register MMU notifier for address range
+ * @vm: VM
+ * @start: Start address (page-aligned)
+ * @end: End address (page-aligned)
+ *
+ * Registers interval notifier for munmap tracking. Uses addresses (not VMA pointers)
+ * to avoid UAF after dropping vm->lock. Deduplicates by range.
+ *
+ * Return: 0 on success, negative error code on failure
+ */
+int xe_vm_madvise_register_notifier_range(struct xe_vm *vm, u64 start, u64 end)
+{
+	struct xe_madvise_notifier *notifier, *old, *tmp;
+	struct xe_madvise_notifier *existing;
+	LIST_HEAD(displaced);
+	int err;
+
+	if (!IS_ALIGNED(start, PAGE_SIZE) || !IS_ALIGNED(end, PAGE_SIZE))
+		return -EINVAL;
+
+	if (WARN_ON_ONCE(end <= start))
+		return -EINVAL;
+
+	if (!vm->svm.gpusvm.mm)
+		return -EINVAL;
+
+	notifier = xe_madvise_notifier_alloc(vm, start, end);
+	if (!notifier)
+		return -ENOMEM;
+
+	/* Insert before vm->lock, this may take mmap_lock. */
+	err = mmu_interval_notifier_insert(&notifier->mmu_notifier,
+					   vm->svm.gpusvm.mm,
+					   start, end - start,
+					   &xe_madvise_notifier_ops);
+	if (err) {
+		xe_madvise_notifier_free(notifier);
+		return err;
+	}
+
+	/* Dedup and store under vm->lock. */
+	down_write(&vm->lock);
+
+	if (xe_vm_is_closed_or_banned(vm)) {
+		err = -ENOENT;
+		goto unlock_remove_new;
+	}
+
+	/* Dedup by stored range; tree slots can be fragmented by partial overlap. */
+	list_for_each_entry(existing, &vm->svm.madvise_notifier_list, link) {
+		if (xe_madvise_notifier_exact(existing, start, end)) {
+			err = 0;
+			goto unlock_remove_new;
+		}
+	}
+
+	/*
+	 * Store first. The VM list owns notifier lifetime, so there is
+	 * nothing to restore on failure.
+	 */
+	err = mtree_store_range(&vm->svm.madvise_notifiers, start, end - 1,
+				notifier, GFP_KERNEL);
+	if (err)
+		goto unlock_remove_new;
+
+	/* Keep the new notifier reachable for teardown. */
+	list_add_tail(&notifier->link, &vm->svm.madvise_notifier_list);
+
+	/*
+	 * Drop fully covered old notifiers. Broader notifiers may still cover
+	 * split siblings, so leave them alive.
+	 */
+	list_for_each_entry_safe(old, tmp, &vm->svm.madvise_notifier_list, link) {
+		if (old == notifier)
+			continue;
+		if (!xe_madvise_notifier_fully_covered(old, start, end))
+			continue;
+
+		list_del_init(&old->link);
+		list_add(&old->link, &displaced);
+	}
+
+	up_write(&vm->lock);
+
+	/*
+	 * Remove outside vm->lock. remove_and_free() drains callbacks and
+	 * work before freeing the notifier.
+	 */
+	list_for_each_entry_safe(old, tmp, &displaced, link) {
+		list_del_init(&old->link);
+		xe_madvise_notifier_remove_and_free(old);
+	}
+
+	return 0;
+
+unlock_remove_new:
+	up_write(&vm->lock);
+	xe_madvise_notifier_remove_and_free(notifier);
+
+	return err;
+}
drivers/gpu/drm/xe/xe_vm_madvise.h +8 -0
index a3078f634c7e..e2013605e190 100644
@@ -6,11 +6,19 @@
 #ifndef _XE_VM_MADVISE_H_
 #define _XE_VM_MADVISE_H_
 
+#include <linux/types.h>
+
 struct drm_device;
 struct drm_file;
 struct xe_bo;
+struct xe_vm;
+struct xe_vma;
 
 int xe_vm_madvise_ioctl(struct drm_device *dev, void *data,
 			struct drm_file *file);
 
+int xe_vm_madvise_init(struct xe_vm *vm);
+void xe_vm_madvise_fini(struct xe_vm *vm);
+int xe_vm_madvise_register_notifier_range(struct xe_vm *vm, u64 start, u64 end);
+
 #endif
drivers/gpu/drm/xe/xe_vm_types.h +61 -0

79 lines not on this page · raw message

Signature · 1 line
2.43.0
#6 lore raw

Re: [RFC v3 4/7] drm/xe/vm: Add madvise autoreset notifier worker

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider: - [High] Returning false for non-blockable invalidations incorrectly prevents the OOM reaper from unmapping…

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [High] Returning false for non-blockable invalidations incorrectly prevents the OOM reaper from unmapping memory.
- [Medium] Redundant VMA splitting for partial overlaps when the VMA already has default attributes. --

commit fe0b22bc527cb1c9caa55068f36be5982435fd31
Author: Arvind Yadav <[email protected]>

drm/xe/vm: Add madvise autoreset notifier worker

This commit adds a worker to reset madvise-managed VMA attributes when userspace unmaps CPU-only mirror ranges. It uses an MMU notifier callback to record pending unmap events and widens the range if multiple unmaps arrive before the worker runs. The worker then resets attributes for ranges not covered by a CPU VMA.

Quoted 4 lines
drivers/gpu/drm/xe/xe_vm_madvise.c +0 -0
index c4fb290041956..c2abe712598a1 100644

[ ... ]

Quoted 1 line
drivers/gpu/drm/xe/xe_vm_madvise.c +0 -0
@@ -732,3 +738,501 @@ int xe_vm_madvise_ioctl(struct drm_device *dev, void *data, struct drm_file *fil

[ ... ]

Quoted 2 lines
drivers/gpu/drm/xe/xe_vm_madvise.c +2 -0
+static int xe_vm_madvise_process_unmap(struct xe_vm *vm, u64 start, u64 end)
+{

[ ... ]

Quoted 14 lines
drivers/gpu/drm/xe/xe_vm_madvise.c +14 -0
+		has_default_attr = xe_vma_has_default_mem_attrs(vma);
+		seg_start = max(addr, xe_vma_start(vma));
+		seg_end = min(end, xe_vma_end(vma));
+
+		/* Merge adjacent default-attr VMAs when possible. */
+		if (has_default_attr &&
+		    xe_vma_start(vma) >= start &&
+		    xe_vma_end(vma) <= end) {
+			seg_start = xe_vma_start(vma);
+			seg_end = xe_vma_end(vma);
+			xe_vm_find_cpu_addr_mirror_vma_range(vm, &seg_start, &seg_end);
+			if (xe_vma_start(vma) == seg_start && xe_vma_end(vma) == seg_end) {
+				/* Nothing to merge. */
+				addr = seg_end;

drivers/gpu/drm/xe/xe_vm_madvise.c int xe_vm_madvise_ioctl(struct drm_device *dev, void *data, struct drm_file *fil · line 818

Quoted 10 lines
drivers/gpu/drm/xe/xe_vm_madvise.c +10 -0
+				continue;
+			}
+		} else if (xe_vma_start(vma) == seg_start && xe_vma_end(vma) == seg_end) {
+			/* Exact VMA match, reset in place. */
+			xe_vma_reset_to_default_attrs(vma);
+			addr = seg_end;
+			continue;
+		}
+
+		err = xe_vm_alloc_cpu_addr_mirror_vma(vm, seg_start, seg_end - seg_start);

[Severity: Medium]
When a CPU hole partially overlaps a GPU VMA that already has default attributes, will this fall through to xe_vm_alloc_cpu_addr_mirror_vma() and cause redundant VMA splitting?

If the overlap is strictly partial, both the if and else if conditions above fail. The fall-through maps the overlapping portion with default attributes, forcing the GPUVM infrastructure to split the existing VMA.

Since the existing VMA already possessed default attributes, this split appears to be redundant and might needlessly fragment a single large VMA into smaller identical pieces. Repeated partial unmaps from userspace could artificially fragment the VMA tree and leak xe_vma objects.

[ ... ]

Quoted 4 lines
drivers/gpu/drm/xe/xe_vm_madvise.c +4 -0
+static bool xe_madvise_notifier_callback(struct mmu_interval_notifier *mni,
+					 const struct mmu_notifier_range *range,
+					 unsigned long cur_seq)
+{

drivers/gpu/drm/xe/xe_vm_madvise.c int xe_vm_madvise_ioctl(struct drm_device *dev, void *data, struct drm_file *fil · line 955

Quoted 10 lines
drivers/gpu/drm/xe/xe_vm_madvise.c +10 -0
+	struct xe_madvise_notifier *notifier =
+		container_of(mni, struct xe_madvise_notifier, mmu_notifier);
+	struct xe_vm *vm = notifier->vm;
+	u64 adj_start, adj_end;
+
+	if (range->event != MMU_NOTIFY_UNMAP)
+		return true;
+
+	if (!mmu_notifier_range_blockable(range))
+		return false;

[Severity: High]
Does returning false for non-blockable invalidations incorrectly prevent the OOM reaper from unmapping memory?

Because the callback implementation is entirely non-blocking (it relies on down_read_trylock(), spin_lock(), and queue_work()), there seems to be no technical limitation requiring it to block.

Returning false from an MMU interval notifier's invalidate callback signals that the invalidation could not be completed without blocking, causing __mmu_notifier_invalidate_range_start() to return -EAGAIN.

Since the OOM reaper operates in a non-blockable context, receiving -EAGAIN means it cannot retry indefinitely and will skip the VMA, failing to reclaim the memory and defeating the purpose of the OOM killer. Could this safely process the event and unconditionally return true for unmap events instead?

Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=4
#7 lore raw

RFC v3 5/7 drm/xe/vm: Disable madvise notifier on GPU touch

In reply to Arvind Yadav

MADVISE_AUTORESET notifiers are only needed while the VMA is CPU-only. After the first GPU fault, later unmaps follow the normal SVM path.

Add an active flag so GPU-touched VMAs stop queueing autoreset work. Broader notifiers from split VMAs may still remain active, but the worker rechecks cpu_autoreset_active under vm->lock before resetting attrs.

v2:
- Use plain bool active; callback reads it lockless. (Matt)
- Use wq check and mt_for_each in deactivate path. (Matt)
- Add lockdep_assert_held_write(&vm->lock) to xe_vm_madvise_gpu_touch().

v3:
- Deactivate only exact-match notifiers to preserve split siblings.
- Use READ_ONCE/WRITE_ONCE for notifier->active since the MMU notifier callback reads it locklessly.

Cc Matthew Brost <[email protected]>
Cc Thomas Hellström <[email protected]>
Cc Himal Prasad Ghimiray <[email protected]>
---
 drivers/gpu/drm/xe/xe_pagefault.c  |  5 +++-
 drivers/gpu/drm/xe/xe_svm.c        |  1 +
 drivers/gpu/drm/xe/xe_vm_madvise.c | 47 +++++++++++++++++++++++++++++-
 drivers/gpu/drm/xe/xe_vm_madvise.h |  1 +
 drivers/gpu/drm/xe/xe_vm_types.h   |  7 +++++
 5 files changed, 59 insertions(+), 2 deletions(-)
drivers/gpu/drm/xe/xe_pagefault.c +4 -1
index f64d3df08261..442a4fcad0c0 100644
@@ -19,6 +19,7 @@
 #include "xe_svm.h"
 #include "xe_trace_bo.h"
 #include "xe_vm.h"
+#include "xe_vm_madvise.h"
 
 /**
  * DOC: Xe page faults
@@ -222,8 +223,10 @@ static int xe_pagefault_service(struct xe_pagefault *pf)
 
 		/* First successful GPU fault ends CPU-only state. */
 		if (vma && xe_vma_is_cpu_addr_mirror(vma) &&
-		    xe_vma_has_cpu_autoreset_active(vma))
+		    xe_vma_has_cpu_autoreset_active(vma)) {
 			xe_vma_gpu_touch(vma);
+			xe_vm_madvise_gpu_touch(vm, vma);
+		}
 	}
 	up_write(&vm->lock);
 	xe_vm_put(vm);
drivers/gpu/drm/xe/xe_svm.c +1 -0
index e1651e70c8f0..b58857668d48 100644
@@ -24,6 +24,7 @@
 #include "xe_vm.h"
 #include "xe_vm_types.h"
 #include "xe_vram_types.h"
+#include "xe_vm_madvise.h"
 
 /* Identifies subclasses of struct drm_pagemap_peer */
 #define XE_PEER_PAGEMAP ((void *)0ul)
drivers/gpu/drm/xe/xe_vm_madvise.c +46 -1

68 lines not on this page · raw message

drivers/gpu/drm/xe/xe_vm_madvise.h +1 -0
index e2013605e190..d7ea6ff6b0c5 100644
@@ -20,5 +20,6 @@ int xe_vm_madvise_ioctl(struct drm_device *dev, void *data,
 int xe_vm_madvise_init(struct xe_vm *vm);
 void xe_vm_madvise_fini(struct xe_vm *vm);
 int xe_vm_madvise_register_notifier_range(struct xe_vm *vm, u64 start, u64 end);
+void xe_vm_madvise_gpu_touch(struct xe_vm *vm, struct xe_vma *vma);
 
 #endif
drivers/gpu/drm/xe/xe_vm_types.h +7 -0
index d47bc338628d..4cb3fd619acf 100644
@@ -50,6 +50,13 @@ struct xe_madvise_notifier {
 	u64 vma_end;
 	/** @link: Entry on vm->svm.madvise_notifier_list. */
 	struct list_head link;
+	/**
+	 * @active: Fast-path callback gate.
+	 *
+	 * Read locklessly by the MMU notifier callback. The worker still checks
+	 * cpu_autoreset_active under vm->lock before resetting attrs.
+	 */
+	bool active;
 	/** @work_lock: Serialises pending interval state. */
 	spinlock_t work_lock;
 	/** @work_pending: Pending interval is available for the worker. */
Signature · 1 line
2.43.0
#8 lore raw

Re: [RFC v3 5/7] drm/xe/vm: Disable madvise notifier on GPU touch

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Pre-existing issues:
- [High] Unbounded accumulation of xe_madvise_notifier structs leading to memory leak and Denial of Service (OOM). --

commit 01cdfed4cb38d71774f4ae69b625419325962b05
Author: Arvind Yadav <[email protected]>
drm/xe/vm: Disable madvise notifier on GPU touch

This commit adds an active flag to xe_madvise_notifier to stop queueing autoreset work once a CPU-only VMA has been GPU-faulted. It modifies the pagefault handler to touch the VMA and deactivate exact-match notifiers in the maple tree.

Quoted 4 lines
drivers/gpu/drm/xe/xe_vm_madvise.c +0 -0
index c2abe712598a1..6c42ce8e3f525 100644

[ ... ]

Quoted 18 lines
drivers/gpu/drm/xe/xe_vm_madvise.c +14 -0
@@ -1236,3 +1238,46 @@ int xe_vm_madvise_register_notifier_range(struct xe_vm *vm, u64 start, u64 end)
 
 	return err;
 }
+
+/**
+ * xe_vm_deactivate_madvise_notifier_for_range - Disable callbacks for a VMA
+ * @vm: VM
+ * @start: VMA start
+ * @end: VMA end
+ */
+static void xe_vm_deactivate_madvise_notifier_for_range(struct xe_vm *vm, u64 start, u64 end)
+{
+	struct xe_madvise_notifier *notifier;
+	unsigned long index = start;
+
+	lockdep_assert_held_write(&vm->lock);
+

drivers/gpu/drm/xe/xe_vm_madvise.c int xe_vm_madvise_register_notifier_range(struct xe_vm *vm, u64 start, u64 end) · line 1264

Quoted 10 lines
drivers/gpu/drm/xe/xe_vm_madvise.c +10 -0
+	if (!vm->svm.madvise_work.wq)
+		return;
+
+	/*
+	 * Only exact-match notifiers are disabled. Broader notifiers may still
+	 * cover CPU-only split siblings.
+	 */
+	mt_for_each(&vm->svm.madvise_notifiers, notifier, index, end - 1)
+		if (notifier->vma_start == start && notifier->vma_end == end)
+			WRITE_ONCE(notifier->active, false);

[Severity: High]
This is a pre-existing issue, but does leaving these deactivated notifiers in memory without freeing them contribute to an unbounded accumulation?

Looking at xe_vm_madvise_register_notifier_range(), notifiers are allocated and linked to the VM indefinitely:

xe_vm_madvise_register_notifier_range() {
    notifier = xe_madvise_notifier_alloc(vm, start, end);
    if (!notifier)
        return -ENOMEM;
    ...
    list_add_tail(&notifier->link, &vm->svm.madvise_notifier_list);
}

Since older notifiers are only freed if they are fully covered by a new range or during VM teardown, could userspace intentionally map, bind with AUTORESET, madvise, and unbind at different addresses across the virtual address space?

This sequence seems like it would bypass the displacement logic and permanently leak the notifier objects, potentially consuming kernel memory and causing a denial of service.

Should xe_vm_deactivate_madvise_notifier_for_range() or the unbind path actively clean up these deactivated objects?

Quoted 19 lines
drivers/gpu/drm/xe/xe_vm_madvise.c +19 -0
+}
+
+/**
+ * xe_vm_madvise_gpu_touch - Disable madvise notifier after GPU touch
+ * @vm: VM
+ * @vma: GPU-touched VMA
+ */
+void xe_vm_madvise_gpu_touch(struct xe_vm *vm, struct xe_vma *vma)
+{
+	lockdep_assert_held_write(&vm->lock);
+
+	/* Only AUTORESET VMAs have madvise notifiers. */
+	if (!(vma->gpuva.flags & XE_VMA_MADV_AUTORESET))
+		return;
+
+	xe_vm_deactivate_madvise_notifier_for_range(vm,
+						xe_vma_start(vma),
+						xe_vma_end(vma));
+}
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=5
#9 lore raw

RFC v3 6/7 drm/xe/vm: Wire MADVISE_AUTORESET notifiers into VM lifecycle

In reply to Arvind Yadav

Initialise the MADVISE_AUTORESET notifier infrastructure for fault-mode VMs and tear it down during VM close.

VM close drops vm->lock around xe_vm_madvise_fini(), since the madvise worker takes vm->lock while draining pending unmap work. The close path then retakes vm->lock before xe_svm_fini().

The MADVISE ioctl collects CPU mirror VMA ranges under vm->lock and registers interval notifiers after dropping it. Registration is best-effort because mmu_interval_notifier_insert() may take mmap_lock.

If registration fails, disable MADV_AUTORESET for the exact VMA range but keep cpu_autoreset_active set, since the VMA is still CPU-only until the first successful GPU fault.

Also skip SVM PTE zapping for CPU mirror VMAs while cpu_autoreset_active is set, since no GPU PTEs should exist yet.

v2:
- Register notifiers outside vm->lock; take vm->lock only for dedup and mtree_store_range. (Matt)
- Collect VMA ranges under vm->lock and register notifiers after unlock. (Matt)

v3:
- Allocate notifier ranges before mutating VMA state.
- Fix notifier-lock unwind when the madvise type is invalid.
- On notifier registration failure, clear MADV_AUTORESET but keep cpu_autoreset_active.

Cc Matthew Brost <[email protected]>
Cc Thomas Hellström <[email protected]>
Cc Himal Prasad Ghimiray <[email protected]>
---
 drivers/gpu/drm/xe/xe_svm.c        |   9 +++
 drivers/gpu/drm/xe/xe_vm.c         |  14 ++++
 drivers/gpu/drm/xe/xe_vm_madvise.c | 105 ++++++++++++++++++++++++++++-
 3 files changed, 125 insertions(+), 3 deletions(-)
drivers/gpu/drm/xe/xe_svm.c +9 -0
index b58857668d48..4169cfae7b51 100644
@@ -914,6 +914,15 @@ int xe_svm_init(struct xe_vm *vm)
 			drm_pagemap_release_owner(&vm->svm.peer);
 			return err;
 		}
+
+		/* Initialize after gpusvm. */
+		err = xe_vm_madvise_init(vm);
+		if (err) {
+			drm_gpusvm_fini(&vm->svm.gpusvm);
+			xe_svm_put_pagemaps(vm);
+			drm_pagemap_release_owner(&vm->svm.peer);
+			return err;
+		}
 	} else {
 		err = drm_gpusvm_init(&vm->svm.gpusvm, "Xe SVM (simple)",
 				      &vm->xe->drm, NULL, 0, 0, 0, NULL,
drivers/gpu/drm/xe/xe_vm.c +14 -0
index dec5279f08a2..e93d715d3da9 100644
@@ -1815,6 +1815,8 @@ struct xe_vm *xe_vm_create(struct xe_device *xe, u32 flags, struct xe_file *xef)
 err_svm_fini:
 	if (flags & XE_VM_FLAG_FAULT_MODE) {
 		vm->size = 0; /* close the vm */
+		/* Safe if madvise init did not run. */
+		xe_vm_madvise_fini(vm);
 		xe_svm_fini(vm);
 	}
 err_no_resv:
@@ -1959,6 +1961,18 @@ void xe_vm_close_and_put(struct xe_vm *vm)
 		xe_vma_destroy_unlocked(vma);
 	}
 
+	/*
+	 * The VM is closed here, so no new ioctl should enter. Drop vm->lock
+	 * while draining madvise workers because they take vm->lock.
+	 */
+	xe_assert(vm->xe, xe_vm_is_closed(vm));
+	up_write(&vm->lock);
+
+	if (vm->flags & XE_VM_FLAG_FAULT_MODE)
+		xe_vm_madvise_fini(vm);
+
+	down_write(&vm->lock);
+
 	xe_svm_fini(vm);
 
 	up_write(&vm->lock);
drivers/gpu/drm/xe/xe_vm_madvise.c +102 -3

162 lines not on this page · raw message

Signature · 1 line
2.43.0
#10 lore raw

Re: [RFC v3 6/7] drm/xe/vm: Wire MADVISE_AUTORESET notifiers into VM lifecycle

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [High] Unbounded kernel memory leak of `xe_madvise_notifier` structures due to missing cleanup on VMA…

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] Unbounded kernel memory leak of xe_madvise_notifier structures due to missing cleanup on VMA unbind. --

commit 842ee67ad0fcd26bd2723ff015ee6cb68da543ef
Author: Arvind Yadav <[email protected]>

drm/xe/vm: Wire MADVISE_AUTORESET notifiers into VM lifecycle

This commit initializes the MADVISE_AUTORESET notifier infrastructure for fault-mode VMs and handles tear down during VM close. It also collects CPU mirror VMA ranges to register interval notifiers after dropping the VM lock.

Quoted 3 lines
drivers/gpu/drm/xe/xe_vm_madvise.c +0 -0

[ ... ]

drivers/gpu/drm/xe/xe_vm_madvise.c int xe_vm_madvise_ioctl(struct drm_device *dev, void *data, struct drm_file *fil · line 813

Quoted 9 lines
drivers/gpu/drm/xe/xe_vm_madvise.c +5 -0
@@ -731,6 +806,30 @@ int xe_vm_madvise_ioctl(struct drm_device *dev, void *data, struct drm_file *fil
 unlock_vm:
 	up_write(&vm->lock);
 
+	if (!err) {
+		for (int i = 0; i < num_notifier_ranges; i++) {
+			int ret = xe_vm_madvise_register_notifier_range(vm,
+							notifier_ranges[i].start,
+							notifier_ranges[i].end);

[Severity: High]
Does this code leak xe_madvise_notifier structures if the VMA is subsequently unbound?

When userspace calls the MADVISE ioctl on a CPU mirror VMA, xe_vm_madvise_register_notifier_range() allocates a new notifier and adds it to the VM madvise_notifier_list.

However, there does not appear to be a hook in the VMA teardown path, such as xe_vma_destroy(), to remove notifiers associated with a destroyed GPU VMA. The notifiers are only freed when the entire VM is destroyed in xe_vm_madvise_fini(), or if they are fully covered by a newly registered notifier.

Could a userspace application repeatedly bind a VMA, call the MADVISE ioctl, and unbind the VMA, shifting the address range slightly each time, to indefinitely accumulate xe_madvise_notifier objects and mmu interval notifiers until it triggers an out of memory condition?

Quoted 5 lines
drivers/gpu/drm/xe/xe_vm_madvise.c +5 -0
+			if (ret) {
+				drm_warn(&vm->xe->drm,
+					 "Failed to register madvise notifier [%#llx-%#llx]: %d\n",
+					 notifier_ranges[i].start,
+					 notifier_ranges[i].end, ret);

[ ... ]

Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=6
#11 lore raw

RFC v3 7/7 drm/xe/svm: Correct memory attribute reset for partial unmap

In reply to Arvind Yadav

From: Himal Prasad Ghimiray <[email protected]>

When performing a partial unmap of an SVM range, the memory attributes were being reset for the entire range instead of just the portion being unmapped. This could lead to unintended side effects and behaviour.

Fix this by restricting the attribute reset to only the affected subrange that is being unmapped.

v3:
- Coalesce partial_unmap range via min/max instead of blind overwrite (Matt)
- Restore range_debug() in xe_svm_garbage_collector_add_range() (Matt)
- Update partial_unmap under garbage_collector.lock to avoid races with the worker.
- Clear consumed partial_unmap state before dropping the GC lock.
- Continue processing remaining queued ranges on non-fatal set_default_attr failure instead of stranding them.

---
 drivers/gpu/drm/xe/xe_svm.c | 87 ++++++++++++++++++++++++++++++-------
 drivers/gpu/drm/xe/xe_svm.h | 10 +++++
 2 files changed, 82 insertions(+), 15 deletions(-)
drivers/gpu/drm/xe/xe_svm.c +72 -15
index 4169cfae7b51..beabae3db2dc 100644
@@ -58,6 +58,8 @@ void *xe_svm_private_page_owner(struct xe_vm *vm, bool force_smem)
 	return force_smem ? NULL : vm->svm.peer.owner;
 }
 
+#define XE_SVM_ATTR_RETRY_MAX 3
+
 static bool xe_svm_range_in_vram(struct xe_svm_range *range)
 {
 	/*
@@ -127,6 +129,34 @@ static void xe_svm_range_free(struct drm_gpusvm_range *range)
 	kfree(range);
 }
 
+static void xe_svm_range_set_unmapped(struct xe_svm_range *range,
+				      const struct mmu_notifier_range *mmu_range)
+{
+	u64 new_start = max_t(u64, xe_svm_range_start(range), mmu_range->start);
+	u64 new_end = min_t(u64, xe_svm_range_end(range), mmu_range->end);
+
+	drm_gpusvm_range_set_unmapped(&range->base, mmu_range);
+	if (range->base.pages.flags.partial_unmap) {
+		if (range->partial_unmap.start || range->partial_unmap.end) {
+			/*
+			 * Coalesce overlapping or adjacent intervals.
+			 * Disjoint ranges cannot be tracked in one interval
+			 * without covering still-mapped memory.
+			 */
+			if (new_start <= range->partial_unmap.end &&
+			    new_end >= range->partial_unmap.start) {
+				range->partial_unmap.start =
+					min(range->partial_unmap.start, new_start);
+				range->partial_unmap.end =
+					max(range->partial_unmap.end, new_end);
+			}
+		} else {
+			range->partial_unmap.start = new_start;
+			range->partial_unmap.end = new_end;
+		}
+	}
+}
+
 static void
 xe_svm_garbage_collector_add_range(struct xe_vm *vm, struct xe_svm_range *range,
 				   const struct mmu_notifier_range *mmu_range)
@@ -135,12 +165,14 @@ xe_svm_garbage_collector_add_range(struct xe_vm *vm, struct xe_svm_range *range,
 
 	range_debug(range, "GARBAGE COLLECTOR ADD");
 
-	drm_gpusvm_range_set_unmapped(&range->base, mmu_range);
-
 	spin_lock(&vm->svm.garbage_collector.lock);
+
+	xe_svm_range_set_unmapped(range, mmu_range);
+
 	if (list_empty(&range->garbage_collector_link))
 		list_add_tail(&range->garbage_collector_link,
 			      &vm->svm.garbage_collector.range_list);
+
 	spin_unlock(&vm->svm.garbage_collector.lock);
 
 	queue_work(xe->usm.pf_wq, &vm->svm.garbage_collector.work);
@@ -380,9 +412,10 @@ static int xe_svm_range_set_default_attr(struct xe_vm *vm, u64 start, u64 end)
 static int xe_svm_garbage_collector(struct xe_vm *vm)
 {
 	struct xe_svm_range *range;
-	u64 range_start;
-	u64 range_end;
-	int err, ret = 0;
+	u64 unmap_start;
+	u64 unmap_end;
+	int err;
+	int retry_count;
 
 	lockdep_assert_held_write(&vm->lock);
 
@@ -397,8 +430,18 @@ static int xe_svm_garbage_collector(struct xe_vm *vm)
 		if (!range)
 			break;
 
-		range_start = xe_svm_range_start(range);
-		range_end = xe_svm_range_end(range);
+		if (range->base.pages.flags.partial_unmap &&
+		    range->partial_unmap.start && range->partial_unmap.end) {
+			unmap_start = range->partial_unmap.start;
+			unmap_end = range->partial_unmap.end;
+		} else {
+			unmap_start = xe_svm_range_start(range);
+			unmap_end = xe_svm_range_end(range);
+		}
+
+		/* Clear consumed state before dropping lock. */
+		range->partial_unmap.start = 0;
+		range->partial_unmap.end = 0;
 
 		list_del(&range->garbage_collector_link);
 		spin_unlock(&vm->svm.garbage_collector.lock);
@@ -412,17 +455,31 @@ static int xe_svm_garbage_collector(struct xe_vm *vm)
 			return err;
 		}
 
-		err = xe_svm_range_set_default_attr(vm, range_start, range_end);
-		if (err) {
-			if (err == -EAGAIN)
-				ret = -EAGAIN;
-			else
-				return err;
-		}
+		/*
+		 * Retry if the VMA was recreated while rebuilding attrs.
+		 * Repeated failures mean the topology is not converging.
+		 */
+		retry_count = 0;
+
+		do {
+			err = xe_svm_range_set_default_attr(vm, unmap_start, unmap_end);
+			if (err == -EAGAIN && ++retry_count > XE_SVM_ATTR_RETRY_MAX) {
+				drm_err(&vm->xe->drm,
+					"SET_ATTR retry limit exceeded for [0x%llx-0x%llx]\n",
+					unmap_start, unmap_end);
+				xe_vm_kill(vm, true);
+				return -EIO;
+			}
+		} while (err == -EAGAIN);
+
+		if (err)
+			drm_warn(&vm->xe->drm,
+				 "SET_ATTR failed for [0x%llx-0x%llx]: %pe\n",
+				 unmap_start, unmap_end, ERR_PTR(err));
 	}
 	spin_unlock(&vm->svm.garbage_collector.lock);
 
-	return ret;
+	return 0;
 }
 
 static void xe_svm_garbage_collector_work_func(struct work_struct *w)
drivers/gpu/drm/xe/xe_svm.h +10 -0
index b7b8eeacf196..4651e044cf53 100644
@@ -46,6 +46,16 @@ struct xe_svm_range {
 	 * range. Protected by GPU SVM notifier lock.
 	 */
 	u8 tile_invalidated;
+	/**
+	 * @partial_unmap: Structure to hold partial unmap range info.
+	 * Valid only if partial unmap is in effect.
+	 */
+	struct {
+		/** @start: Start address of the partial unmap range */
+		u64 start;
+		/** @end: End address of the partial unmap range */
+		u64 end;
+	} partial_unmap;
 };
 
 /**
Signature · 1 line
2.43.0
#12 lore raw

Re: [RFC v3 7/7] drm/xe/svm: Correct memory attribute reset for partial unmap

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider: - [High] Concurrent MMU notifiers fail to queue partial unmap range updates due to list poisoning, causing…

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [High] Concurrent MMU notifiers fail to queue partial unmap range updates due to list poisoning, causing unmapped regions to retain stale attributes.
- [High] Disjoint partial unmaps are silently ignored during coalescing, preventing their VMA memory attributes from being reset. --

commit 63782c32791ae66c7e1c022f8db4a4b168c0d897
Author: Himal Prasad Ghimiray <[email protected]>

drm/xe/svm: Correct memory attribute reset for partial unmap

When performing a partial unmap of an SVM range, the memory attributes were being reset for the entire range instead of just the portion being unmapped. This could lead to unintended side effects and behaviour.

Quoted 4 lines
drivers/gpu/drm/xe/xe_svm.c +0 -0
@@ -127,6 +129,34 @@ static void xe_svm_range_free(struct drm_gpusvm_range *range)

[ ... ]

Quoted 11 lines
drivers/gpu/drm/xe/xe_svm.c +11 -0
+static void xe_svm_range_set_unmapped(struct xe_svm_range *range,
+				      const struct mmu_notifier_range *mmu_range)
+{
+	u64 new_start = max_t(u64, xe_svm_range_start(range), mmu_range->start);
+	u64 new_end = min_t(u64, xe_svm_range_end(range), mmu_range->end);
+
+	drm_gpusvm_range_set_unmapped(&range->base, mmu_range);
+	if (range->base.pages.flags.partial_unmap) {
+		if (range->partial_unmap.start || range->partial_unmap.end) {
+			/*
+			 * Coalesce overlapping or adjacent intervals.

drivers/gpu/drm/xe/xe_svm.c static void xe_svm_range_free(struct drm_gpusvm_range *range) · line 152

Quoted 10 lines
drivers/gpu/drm/xe/xe_svm.c +10 -0
+			 * Disjoint ranges cannot be tracked in one interval
+			 * without covering still-mapped memory.
+			 */
+			if (new_start <= range->partial_unmap.end &&
+			    new_end >= range->partial_unmap.start) {
+				range->partial_unmap.start =
+					min(range->partial_unmap.start, new_start);
+				range->partial_unmap.end =
+					max(range->partial_unmap.end, new_end);
+			}

[Severity: High]
If the new partial unmap is disjoint from the tracked interval, the overlap condition evaluates to false. Because there is no else block here, does this mean the disjoint interval is silently discarded?

If so, the garbage collector might only reset the VMA attributes for the initial subrange, leaving the disjoint unmapped region with stale memory attributes.

Quoted 6 lines
drivers/gpu/drm/xe/xe_svm.c +6 -0
+		} else {
+			range->partial_unmap.start = new_start;
+			range->partial_unmap.end = new_end;
+		}
+	}
+}

[ ... ]

Quoted 5 lines
drivers/gpu/drm/xe/xe_svm.c +0 -1
@@ -135,12 +165,14 @@ xe_svm_garbage_collector_add_range(struct xe_vm *vm, struct xe_svm_range *range,
 
 	range_debug(range, "GARBAGE COLLECTOR ADD");
 
-	drm_gpusvm_range_set_unmapped(&range->base, mmu_range);

drivers/gpu/drm/xe/xe_svm.c xe_svm_garbage_collector_add_range(struct xe_vm *vm, struct xe_svm_range *range, · line 176

Quoted 10 lines
drivers/gpu/drm/xe/xe_svm.c +4 -1
-
 	spin_lock(&vm->svm.garbage_collector.lock);
+
+	xe_svm_range_set_unmapped(range, mmu_range);
+
 	if (list_empty(&range->garbage_collector_link))
 		list_add_tail(&range->garbage_collector_link,
 			      &vm->svm.garbage_collector.range_list);
+
 	spin_unlock(&vm->svm.garbage_collector.lock);

[Severity: High]
Is it possible to lose concurrent MMU notifier updates here due to list poisoning?

In xe_svm_garbage_collector(), the worker thread removes the item using list_del(), which poisons the list node pointers, and then the lock is dropped:

xe_svm_garbage_collector() {
    ...
    list_del(&range->garbage_collector_link);
    spin_unlock(&vm->svm.garbage_collector.lock);
    ...
}

If a concurrent MMU notifier fires for the same range during this unlocked window, xe_svm_garbage_collector_add_range() acquires the lock, updates the partial unmap bounds, and evaluates the list_empty() condition above.

Because the pointers are poisoned, list_empty() evaluates to false, and the range is not re-queued. Would this cause the VMA attributes for this concurrent unmap to never be reset?

Note that switching to list_del_init() would allow it to be re-queued, but could that introduce a use-after-free since the worker then destroys the range?

Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=7
#13 lore raw

✗ CI.checkpatch: warning for drm/xe/svm: Add MMU notifier-based madvise autoreset on munmap (rev3)

In reply to Arvind Yadav
== Series Details ==

== Series Details ==

Series: drm/xe/svm: Add MMU notifier-based madvise autoreset on munmap (rev3)
URL : https://patchwork.freedesktop.org/series/161815/
State : warning

== Summary ==

+ KERNEL=/kernel
+ git clone https://gitlab.freedesktop.org/drm/maintainer-tools mt Cloning into 'mt'...
warning: redirecting to https://gitlab.freedesktop.org/drm/maintainer-tools.git/
+ git -C mt rev-list -n1 origin/master 061140b9bc586ae7f40abc1249c97e1cc72d1b9d
+ cd /kernel
+ git config --global --add safe.directory /kernel
+ git log -n1
commit b896a89fd0c92f4b48c8953bed8abaedfcad63a1

Author: Himal Prasad Ghimiray <[email protected]>
Date:   Thu Jun 4 15:20:12 2026 +0530

drm/xe/svm: Correct memory attribute reset for partial unmap

When performing a partial unmap of an SVM range, the memory attributes were being reset for the entire range instead of just the portion being unmapped. This could lead to unintended side effects and behaviour.

Fix this by restricting the attribute reset to only the affected subrange that is being unmapped.

v3:
- Coalesce partial_unmap range via min/max instead of blind overwrite (Matt)
- Restore range_debug() in xe_svm_garbage_collector_add_range() (Matt)
- Update partial_unmap under garbage_collector.lock to avoid races with the worker.
- Clear consumed partial_unmap state before dropping the GC lock.
- Continue processing remaining queued ranges on non-fatal set_default_attr failure instead of stranding them.

+ /mt/dim checkpatch 70c7b363fab8198549bb9b82134bda0f6bda9452 drm-intel
76568c4603b0 drm/xe/vm: Track CPU_AUTORESET state in xe_vma
38b7fb1e6539 drm/xe/vm: Preserve cpu_autoreset_active across GPUVA operations
-:28: WARNING:COMMIT_LOG_LONG_LINE: Prefer a maximum 75 chars per line (possible unwrapped commit description?) #28:
- Guard cpu_autoreset_active assignment to CPU_ADDR_MIRROR VMAs only (Matt)

total: 0 errors, 1 warnings, 0 checks, 103 lines checked
5b4cd11bc093 drm/xe/svm: Clear CPU_AUTORESET_ACTIVE on first GPU fault
f9f659d075d9 drm/xe/vm: Add madvise autoreset notifier worker
-:32: WARNING:COMMIT_LOG_LONG_LINE: Prefer a maximum 75 chars per line (possible unwrapped commit description?) #32:
- Widen pending ranges in the callback and reset only CPU holes in the worker.

total: 0 errors, 1 warnings, 0 checks, 617 lines checked f5710d0bc501 drm/xe/vm: Disable madvise notifier on GPU touch -:142: CHECK:PARENTHESIS_ALIGNMENT: Alignment should match open parenthesis #142: FILE: drivers/gpu/drm/xe/xe_vm_madvise.c:1281:
+ xe_vm_deactivate_madvise_notifier_for_range(vm,
+ xe_vma_start(vma),

total: 0 errors, 0 warnings, 1 checks, 112 lines checked
931ebcf6f5c9 drm/xe/vm: Wire MADVISE_AUTORESET notifiers into VM lifecycle
b896a89fd0c9 drm/xe/svm: Correct memory attribute reset for partial unmap
-:17: WARNING:COMMIT_LOG_LONG_LINE: Prefer a maximum 75 chars per line (possible unwrapped commit description?) #17:
- Coalesce partial_unmap range via min/max instead of blind overwrite (Matt)

total: 0 errors, 1 warnings, 0 checks, 146 lines checked

#14 lore raw

✓ CI.KUnit: success for drm/xe/svm: Add MMU notifier-based madvise autoreset on munmap (rev3)

In reply to Arvind Yadav
== Series Details ==

== Series Details ==

Series: drm/xe/svm: Add MMU notifier-based madvise autoreset on munmap (rev3)
URL : https://patchwork.freedesktop.org/series/161815/
State : success

== Summary ==

+ trap cleanup EXIT
+ /kernel/tools/testing/kunit/kunit.py run --kunitconfig /kernel/drivers/gpu/drm/xe/.kunitconfig [10:32:44] Configuring KUnit Kernel ...
Generating .config ...
Populating config with:

$ make ARCH=um O=.kunit olddefconfig
[10:32:48] Building KUnit Kernel ...
Populating config with:
$ make ARCH=um O=.kunit olddefconfig
Building with:
$ make all compile_commands.json scripts_gdb ARCH=um O=.kunit --jobs=48
[10:33:19] Starting KUnit Kernel (1/1)...
[10:33:19] ============================================================
Running tests with:
$ .kunit/linux kunit.enable=1 mem=1G console=tty kunit_shutdown=halt
[10:33:19] ================== guc_buf (11 subtests) ===================
[10:33:19] [PASSED] test_smallest
[10:33:19] [PASSED] test_largest
[10:33:19] [PASSED] test_granular
[10:33:19] [PASSED] test_unique
[10:33:19] [PASSED] test_overlap
[10:33:19] [PASSED] test_reusable
[10:33:19] [PASSED] test_too_big
[10:33:19] [PASSED] test_flush
[10:33:19] [PASSED] test_lookup
[10:33:19] [PASSED] test_data
[10:33:19] [PASSED] test_class
[10:33:19] ===================== [PASSED] guc_buf =====================
[10:33:19] =================== guc_dbm (7 subtests) ===================
[10:33:19] [PASSED] test_empty
[10:33:19] [PASSED] test_default
[10:33:19] ======================== test_size  ========================
[10:33:19] [PASSED] 4
[10:33:19] [PASSED] 8
[10:33:19] [PASSED] 32
[10:33:19] [PASSED] 256
[10:33:19] ==================== [PASSED] test_size ====================
[10:33:19] ======================= test_reuse  ========================
[10:33:19] [PASSED] 4
[10:33:19] [PASSED] 8
[10:33:19] [PASSED] 32
[10:33:19] [PASSED] 256
[10:33:19] =================== [PASSED] test_reuse ====================
[10:33:19] =================== test_range_overlap  ====================
[10:33:19] [PASSED] 4
[10:33:19] [PASSED] 8
[10:33:19] [PASSED] 32
[10:33:19] [PASSED] 256
[10:33:19] =============== [PASSED] test_range_overlap ================
[10:33:19] =================== test_range_compact  ====================
[10:33:19] [PASSED] 4
[10:33:19] [PASSED] 8
[10:33:19] [PASSED] 32
[10:33:19] [PASSED] 256
[10:33:19] =============== [PASSED] test_range_compact ================
[10:33:19] ==================== test_range_spare  =====================
[10:33:19] [PASSED] 4
[10:33:19] [PASSED] 8
[10:33:19] [PASSED] 32
[10:33:19] [PASSED] 256
[10:33:19] ================ [PASSED] test_range_spare =================
[10:33:19] ===================== [PASSED] guc_dbm =====================
[10:33:19] =================== guc_idm (6 subtests) ===================
[10:33:19] [PASSED] bad_init
[10:33:19] [PASSED] no_init
[10:33:19] [PASSED] init_fini
[10:33:19] [PASSED] check_used
[10:33:19] [PASSED] check_quota
[10:33:19] [PASSED] check_all
[10:33:19] ===================== [PASSED] guc_idm =====================
[10:33:19] ================== no_relay (3 subtests) ===================
[10:33:19] [PASSED] xe_drops_guc2pf_if_not_ready
[10:33:19] [PASSED] xe_drops_guc2vf_if_not_ready
[10:33:19] [PASSED] xe_rejects_send_if_not_ready
[10:33:19] ==================== [PASSED] no_relay =====================
[10:33:19] ================== pf_relay (14 subtests) ==================
[10:33:19] [PASSED] pf_rejects_guc2pf_too_short
[10:33:19] [PASSED] pf_rejects_guc2pf_too_long
[10:33:19] [PASSED] pf_rejects_guc2pf_no_payload
[10:33:19] [PASSED] pf_fails_no_payload
[10:33:19] [PASSED] pf_fails_bad_origin
[10:33:19] [PASSED] pf_fails_bad_type
[10:33:19] [PASSED] pf_txn_reports_error
[10:33:19] [PASSED] pf_txn_sends_pf2guc
[10:33:19] [PASSED] pf_sends_pf2guc
[10:33:19] [SKIPPED] pf_loopback_nop
[10:33:19] [SKIPPED] pf_loopback_echo
[10:33:19] [SKIPPED] pf_loopback_fail
[10:33:19] [SKIPPED] pf_loopback_busy
[10:33:19] [SKIPPED] pf_loopback_retry
[10:33:19] ==================== [PASSED] pf_relay =====================
[10:33:19] ================== vf_relay (3 subtests) ===================
[10:33:19] [PASSED] vf_rejects_guc2vf_too_short
[10:33:19] [PASSED] vf_rejects_guc2vf_too_long
[10:33:19] [PASSED] vf_rejects_guc2vf_no_payload
[10:33:19] ==================== [PASSED] vf_relay =====================
[10:33:19] ================ pf_gt_config (9 subtests) =================
[10:33:19] [PASSED] fair_contexts_1vf
[10:33:19] [PASSED] fair_doorbells_1vf
[10:33:19] [PASSED] fair_ggtt_1vf
[10:33:19] ====================== fair_vram_1vf  ======================
[10:33:19] [PASSED] 3.50 GiB
[10:33:19] [PASSED] 11.5 GiB
[10:33:19] [PASSED] 15.5 GiB
[10:33:19] [PASSED] 31.5 GiB
[10:33:19] [PASSED] 63.5 GiB
[10:33:19] [PASSED] 1.91 GiB
[10:33:19] ================== [PASSED] fair_vram_1vf ==================
[10:33:19] ================ fair_vram_1vf_admin_only  =================
[10:33:19] [PASSED] 3.50 GiB
[10:33:19] [PASSED] 11.5 GiB
[10:33:19] [PASSED] 15.5 GiB
[10:33:19] [PASSED] 31.5 GiB
[10:33:19] [PASSED] 63.5 GiB
[10:33:19] [PASSED] 1.91 GiB
[10:33:19] ============ [PASSED] fair_vram_1vf_admin_only =============
[10:33:19] ====================== fair_contexts  ======================
[10:33:19] [PASSED] 1 VF
[10:33:19] [PASSED] 2 VFs
[10:33:19] [PASSED] 3 VFs
[10:33:19] [PASSED] 4 VFs
[10:33:19] [PASSED] 5 VFs
[10:33:19] [PASSED] 6 VFs
[10:33:19] [PASSED] 7 VFs
[10:33:19] [PASSED] 8 VFs
[10:33:19] [PASSED] 9 VFs
[10:33:19] [PASSED] 10 VFs
[10:33:19] [PASSED] 11 VFs
[10:33:19] [PASSED] 12 VFs
[10:33:19] [PASSED] 13 VFs
[10:33:19] [PASSED] 14 VFs
[10:33:19] [PASSED] 15 VFs
[10:33:19] [PASSED] 16 VFs
[10:33:19] [PASSED] 17 VFs
[10:33:19] [PASSED] 18 VFs
[10:33:19] [PASSED] 19 VFs
[10:33:19] [PASSED] 20 VFs
[10:33:19] [PASSED] 21 VFs
[10:33:19] [PASSED] 22 VFs
[10:33:19] [PASSED] 23 VFs
[10:33:19] [PASSED] 24 VFs
[10:33:19] [PASSED] 25 VFs
[10:33:19] [PASSED] 26 VFs
[10:33:19] [PASSED] 27 VFs
[10:33:19] [PASSED] 28 VFs
[10:33:19] [PASSED] 29 VFs
[10:33:19] [PASSED] 30 VFs
[10:33:19] [PASSED] 31 VFs
[10:33:19] [PASSED] 32 VFs
[10:33:19] [PASSED] 33 VFs
[10:33:19] [PASSED] 34 VFs
[10:33:19] [PASSED] 35 VFs
[10:33:19] [PASSED] 36 VFs
[10:33:19] [PASSED] 37 VFs
[10:33:19] [PASSED] 38 VFs
[10:33:19] [PASSED] 39 VFs
[10:33:19] [PASSED] 40 VFs
[10:33:19] [PASSED] 41 VFs
[10:33:19] [PASSED] 42 VFs
[10:33:19] [PASSED] 43 VFs
[10:33:19] [PASSED] 44 VFs
[10:33:19] [PASSED] 45 VFs
[10:33:19] [PASSED] 46 VFs
[10:33:19] [PASSED] 47 VFs
[10:33:19] [PASSED] 48 VFs
[10:33:19] [PASSED] 49 VFs
[10:33:19] [PASSED] 50 VFs
[10:33:19] [PASSED] 51 VFs
[10:33:19] [PASSED] 52 VFs
[10:33:19] [PASSED] 53 VFs
[10:33:19] [PASSED] 54 VFs
[10:33:19] [PASSED] 55 VFs
[10:33:19] [PASSED] 56 VFs
[10:33:19] [PASSED] 57 VFs
[10:33:19] [PASSED] 58 VFs
[10:33:19] [PASSED] 59 VFs
[10:33:19] [PASSED] 60 VFs
[10:33:19] [PASSED] 61 VFs
[10:33:19] [PASSED] 62 VFs
[10:33:19] [PASSED] 63 VFs
[10:33:19] ================== [PASSED] fair_contexts ==================
[10:33:19] ===================== fair_doorbells  ======================
[10:33:19] [PASSED] 1 VF
[10:33:19] [PASSED] 2 VFs
[10:33:19] [PASSED] 3 VFs
[10:33:19] [PASSED] 4 VFs
[10:33:19] [PASSED] 5 VFs
[10:33:19] [PASSED] 6 VFs
[10:33:19] [PASSED] 7 VFs
[10:33:19] [PASSED] 8 VFs
[10:33:19] [PASSED] 9 VFs
[10:33:19] [PASSED] 10 VFs
[10:33:19] [PASSED] 11 VFs
[10:33:19] [PASSED] 12 VFs
[10:33:19] [PASSED] 13 VFs
[10:33:19] [PASSED] 14 VFs
[10:33:19] [PASSED] 15 VFs
[10:33:19] [PASSED] 16 VFs
[10:33:19] [PASSED] 17 VFs
[10:33:19] [PASSED] 18 VFs
[10:33:19] [PASSED] 19 VFs
[10:33:19] [PASSED] 20 VFs
[10:33:19] [PASSED] 21 VFs
[10:33:19] [PASSED] 22 VFs
[10:33:19] [PASSED] 23 VFs
[10:33:19] [PASSED] 24 VFs
[10:33:19] [PASSED] 25 VFs
[10:33:19] [PASSED] 26 VFs
[10:33:19] [PASSED] 27 VFs
[10:33:19] [PASSED] 28 VFs
[10:33:19] [PASSED] 29 VFs
[10:33:19] [PASSED] 30 VFs
[10:33:19] [PASSED] 31 VFs
[10:33:19] [PASSED] 32 VFs
[10:33:19] [PASSED] 33 VFs
[10:33:19] [PASSED] 34 VFs
[10:33:19] [PASSED] 35 VFs
[10:33:19] [PASSED] 36 VFs
[10:33:19] [PASSED] 37 VFs
[10:33:19] [PASSED] 38 VFs
[10:33:19] [PASSED] 39 VFs
[10:33:19] [PASSED] 40 VFs
[10:33:19] [PASSED] 41 VFs
[10:33:19] [PASSED] 42 VFs
[10:33:19] [PASSED] 43 VFs
[10:33:19] [PASSED] 44 VFs
[10:33:19] [PASSED] 45 VFs
[10:33:19] [PASSED] 46 VFs
[10:33:19] [PASSED] 47 VFs
[10:33:19] [PASSED] 48 VFs
[10:33:19] [PASSED] 49 VFs
[10:33:19] [PASSED] 50 VFs
[10:33:19] [PASSED] 51 VFs
[10:33:19] [PASSED] 52 VFs
[10:33:19] [PASSED] 53 VFs
[10:33:19] [PASSED] 54 VFs
[10:33:19] [PASSED] 55 VFs
[10:33:19] [PASSED] 56 VFs
[10:33:19] [PASSED] 57 VFs
[10:33:19] [PASSED] 58 VFs
[10:33:19] [PASSED] 59 VFs
[10:33:19] [PASSED] 60 VFs
[10:33:19] [PASSED] 61 VFs
[10:33:19] [PASSED] 62 VFs
[10:33:19] [PASSED] 63 VFs
[10:33:19] ================= [PASSED] fair_doorbells ==================
[10:33:19] ======================== fair_ggtt  ========================
[10:33:19] [PASSED] 1 VF
[10:33:19] [PASSED] 2 VFs
[10:33:19] [PASSED] 3 VFs
[10:33:19] [PASSED] 4 VFs
[10:33:19] [PASSED] 5 VFs
[10:33:19] [PASSED] 6 VFs
[10:33:19] [PASSED] 7 VFs
[10:33:19] [PASSED] 8 VFs
[10:33:19] [PASSED] 9 VFs
[10:33:19] [PASSED] 10 VFs
[10:33:19] [PASSED] 11 VFs
[10:33:19] [PASSED] 12 VFs
[10:33:19] [PASSED] 13 VFs
[10:33:19] [PASSED] 14 VFs
[10:33:19] [PASSED] 15 VFs
[10:33:19] [PASSED] 16 VFs
[10:33:19] [PASSED] 17 VFs
[10:33:19] [PASSED] 18 VFs
[10:33:19] [PASSED] 19 VFs
[10:33:19] [PASSED] 20 VFs
[10:33:19] [PASSED] 21 VFs
[10:33:19] [PASSED] 22 VFs
[10:33:19] [PASSED] 23 VFs
[10:33:19] [PASSED] 24 VFs
[10:33:19] [PASSED] 25 VFs
[10:33:19] [PASSED] 26 VFs
[10:33:19] [PASSED] 27 VFs
[10:33:19] [PASSED] 28 VFs
[10:33:19] [PASSED] 29 VFs
[10:33:19] [PASSED] 30 VFs
[10:33:19] [PASSED] 31 VFs
[10:33:19] [PASSED] 32 VFs
[10:33:19] [PASSED] 33 VFs
[10:33:19] [PASSED] 34 VFs
[10:33:19] [PASSED] 35 VFs
[10:33:19] [PASSED] 36 VFs
[10:33:19] [PASSED] 37 VFs
[10:33:19] [PASSED] 38 VFs
[10:33:19] [PASSED] 39 VFs
[10:33:19] [PASSED] 40 VFs
[10:33:19] [PASSED] 41 VFs
[10:33:19] [PASSED] 42 VFs
[10:33:19] [PASSED] 43 VFs
[10:33:19] [PASSED] 44 VFs
[10:33:19] [PASSED] 45 VFs
[10:33:19] [PASSED] 46 VFs
[10:33:19] [PASSED] 47 VFs
[10:33:19] [PASSED] 48 VFs
[10:33:19] [PASSED] 49 VFs
[10:33:19] [PASSED] 50 VFs
[10:33:19] [PASSED] 51 VFs
[10:33:19] [PASSED] 52 VFs
[10:33:19] [PASSED] 53 VFs
[10:33:19] [PASSED] 54 VFs
[10:33:19] [PASSED] 55 VFs
[10:33:19] [PASSED] 56 VFs
[10:33:19] [PASSED] 57 VFs
[10:33:19] [PASSED] 58 VFs
[10:33:19] [PASSED] 59 VFs
[10:33:19] [PASSED] 60 VFs
[10:33:20] [PASSED] 61 VFs
[10:33:20] [PASSED] 62 VFs
[10:33:20] [PASSED] 63 VFs
[10:33:20] ==================== [PASSED] fair_ggtt ====================
[10:33:20] ======================== fair_vram  ========================
[10:33:20] [PASSED] 1 VF
[10:33:20] [PASSED] 2 VFs
[10:33:20] [PASSED] 3 VFs
[10:33:20] [PASSED] 4 VFs
[10:33:20] [PASSED] 5 VFs
[10:33:20] [PASSED] 6 VFs
[10:33:20] [PASSED] 7 VFs
[10:33:20] [PASSED] 8 VFs
[10:33:20] [PASSED] 9 VFs
[10:33:20] [PASSED] 10 VFs
[10:33:20] [PASSED] 11 VFs
[10:33:20] [PASSED] 12 VFs
[10:33:20] [PASSED] 13 VFs
[10:33:20] [PASSED] 14 VFs
[10:33:20] [PASSED] 15 VFs
[10:33:20] [PASSED] 16 VFs
[10:33:20] [PASSED] 17 VFs
[10:33:20] [PASSED] 18 VFs
[10:33:20] [PASSED] 19 VFs
[10:33:20] [PASSED] 20 VFs
[10:33:20] [PASSED] 21 VFs
[10:33:20] [PASSED] 22 VFs
[10:33:20] [PASSED] 23 VFs
[10:33:20] [PASSED] 24 VFs
[10:33:20] [PASSED] 25 VFs
[10:33:20] [PASSED] 26 VFs
[10:33:20] [PASSED] 27 VFs
[10:33:20] [PASSED] 28 VFs
[10:33:20] [PASSED] 29 VFs
[10:33:20] [PASSED] 30 VFs
[10:33:20] [PASSED] 31 VFs
[10:33:20] [PASSED] 32 VFs
[10:33:20] [PASSED] 33 VFs
[10:33:20] [PASSED] 34 VFs
[10:33:20] [PASSED] 35 VFs
[10:33:20] [PASSED] 36 VFs
[10:33:20] [PASSED] 37 VFs
[10:33:20] [PASSED] 38 VFs
[10:33:20] [PASSED] 39 VFs
[10:33:20] [PASSED] 40 VFs
[10:33:20] [PASSED] 41 VFs
[10:33:20] [PASSED] 42 VFs
[10:33:20] [PASSED] 43 VFs
[10:33:20] [PASSED] 44 VFs
[10:33:20] [PASSED] 45 VFs
[10:33:20] [PASSED] 46 VFs
[10:33:20] [PASSED] 47 VFs
[10:33:20] [PASSED] 48 VFs
[10:33:20] [PASSED] 49 VFs
[10:33:20] [PASSED] 50 VFs
[10:33:20] [PASSED] 51 VFs
[10:33:20] [PASSED] 52 VFs
[10:33:20] [PASSED] 53 VFs
[10:33:20] [PASSED] 54 VFs
[10:33:20] [PASSED] 55 VFs
[10:33:20] [PASSED] 56 VFs
[10:33:20] [PASSED] 57 VFs
[10:33:20] [PASSED] 58 VFs
[10:33:20] [PASSED] 59 VFs
[10:33:20] [PASSED] 60 VFs
[10:33:20] [PASSED] 61 VFs
[10:33:20] [PASSED] 62 VFs
[10:33:20] [PASSED] 63 VFs
[10:33:20] ==================== [PASSED] fair_vram ====================
[10:33:20] ================== [PASSED] pf_gt_config ===================
[10:33:20] ===================== lmtt (1 subtest) =====================
[10:33:20] ======================== test_ops  =========================
[10:33:20] [PASSED] 2-level
[10:33:20] [PASSED] multi-level
[10:33:20] ==================== [PASSED] test_ops =====================
[10:33:20] ====================== [PASSED] lmtt =======================
[10:33:20] ================= pf_service (11 subtests) =================
[10:33:20] [PASSED] pf_negotiate_any
[10:33:20] [PASSED] pf_negotiate_base_match
[10:33:20] [PASSED] pf_negotiate_base_newer
[10:33:20] [PASSED] pf_negotiate_base_next
[10:33:20] [SKIPPED] pf_negotiate_base_older
[10:33:20] [PASSED] pf_negotiate_base_prev
[10:33:20] [PASSED] pf_negotiate_latest_match
[10:33:20] [PASSED] pf_negotiate_latest_newer
[10:33:20] [PASSED] pf_negotiate_latest_next
[10:33:20] [SKIPPED] pf_negotiate_latest_older
[10:33:20] [SKIPPED] pf_negotiate_latest_prev
[10:33:20] =================== [PASSED] pf_service ====================
[10:33:20] ================= xe_guc_g2g (2 subtests) ==================
[10:33:20] ============== xe_live_guc_g2g_kunit_default  ==============
[10:33:20] ========= [SKIPPED] xe_live_guc_g2g_kunit_default ==========
[10:33:20] ============== xe_live_guc_g2g_kunit_allmem  ===============
[10:33:20] ========== [SKIPPED] xe_live_guc_g2g_kunit_allmem ==========
[10:33:20] =================== [SKIPPED] xe_guc_g2g ===================
[10:33:20] =================== xe_mocs (2 subtests) ===================
[10:33:20] ================ xe_live_mocs_kernel_kunit  ================
[10:33:20] =========== [SKIPPED] xe_live_mocs_kernel_kunit ============
[10:33:20] ================ xe_live_mocs_reset_kunit  =================
[10:33:20] ============ [SKIPPED] xe_live_mocs_reset_kunit ============
[10:33:20] ==================== [SKIPPED] xe_mocs =====================
[10:33:20] ================= xe_migrate (2 subtests) ==================
[10:33:20] ================= xe_migrate_sanity_kunit  =================
[10:33:20] ============ [SKIPPED] xe_migrate_sanity_kunit =============
[10:33:20] ================== xe_validate_ccs_kunit  ==================
[10:33:20] ============= [SKIPPED] xe_validate_ccs_kunit ==============
[10:33:20] =================== [SKIPPED] xe_migrate ===================
[10:33:20] ================== xe_dma_buf (1 subtest) ==================
[10:33:20] ==================== xe_dma_buf_kunit  =====================
[10:33:20] ================ [SKIPPED] xe_dma_buf_kunit ================
[10:33:20] =================== [SKIPPED] xe_dma_buf ===================
[10:33:20] ================= xe_bo_shrink (1 subtest) =================
[10:33:20] =================== xe_bo_shrink_kunit  ====================
[10:33:20] =============== [SKIPPED] xe_bo_shrink_kunit ===============
[10:33:20] ================== [SKIPPED] xe_bo_shrink ==================
[10:33:20] ==================== xe_bo (2 subtests) ====================
[10:33:20] ================== xe_ccs_migrate_kunit  ===================
[10:33:20] ============== [SKIPPED] xe_ccs_migrate_kunit ==============
[10:33:20] ==================== xe_bo_evict_kunit  ====================
[10:33:20] =============== [SKIPPED] xe_bo_evict_kunit ================
[10:33:20] ===================== [SKIPPED] xe_bo ======================
[10:33:20] ==================== args (13 subtests) ====================
[10:33:20] [PASSED] count_args_test
[10:33:20] [PASSED] call_args_example
[10:33:20] [PASSED] call_args_test
[10:33:20] [PASSED] drop_first_arg_example
[10:33:20] [PASSED] drop_first_arg_test
[10:33:20] [PASSED] first_arg_example
[10:33:20] [PASSED] first_arg_test
[10:33:20] [PASSED] last_arg_example
[10:33:20] [PASSED] last_arg_test
[10:33:20] [PASSED] pick_arg_example
[10:33:20] [PASSED] if_args_example
[10:33:20] [PASSED] if_args_test
[10:33:20] [PASSED] sep_comma_example
[10:33:20] ====================== [PASSED] args =======================
[10:33:20] =================== xe_pci (3 subtests) ====================
[10:33:20] ==================== check_graphics_ip  ====================
[10:33:20] [PASSED] 12.00 Xe_LP
[10:33:20] [PASSED] 12.10 Xe_LP+
[10:33:20] [PASSED] 12.55 Xe_HPG
[10:33:20] [PASSED] 12.60 Xe_HPC
[10:33:20] [PASSED] 12.70 Xe_LPG
[10:33:20] [PASSED] 12.71 Xe_LPG
[10:33:20] [PASSED] 12.74 Xe_LPG+
[10:33:20] [PASSED] 20.01 Xe2_HPG
[10:33:20] [PASSED] 20.02 Xe2_HPG
[10:33:20] [PASSED] 20.04 Xe2_LPG
[10:33:20] [PASSED] 30.00 Xe3_LPG
[10:33:20] [PASSED] 30.01 Xe3_LPG
[10:33:20] [PASSED] 30.03 Xe3_LPG
[10:33:20] [PASSED] 30.04 Xe3_LPG
[10:33:20] [PASSED] 30.05 Xe3_LPG
[10:33:20] [PASSED] 35.10 Xe3p_LPG
[10:33:20] [PASSED] 35.11 Xe3p_XPC
[10:33:20] ================ [PASSED] check_graphics_ip ================
[10:33:20] ===================== check_media_ip  ======================
[10:33:20] [PASSED] 12.00 Xe_M
[10:33:20] [PASSED] 12.55 Xe_HPM
[10:33:20] [PASSED] 13.00 Xe_LPM+
[10:33:20] [PASSED] 13.01 Xe2_HPM
[10:33:20] [PASSED] 20.00 Xe2_LPM
[10:33:20] [PASSED] 30.00 Xe3_LPM
[10:33:20] [PASSED] 30.02 Xe3_LPM
[10:33:20] [PASSED] 35.00 Xe3p_LPM
[10:33:20] [PASSED] 35.03 Xe3p_HPM
[10:33:20] ================= [PASSED] check_media_ip ==================
[10:33:20] =================== check_platform_desc  ===================
[10:33:20] [PASSED] 0x9A60 (TIGERLAKE)
[10:33:20] [PASSED] 0x9A68 (TIGERLAKE)
[10:33:20] [PASSED] 0x9A70 (TIGERLAKE)
[10:33:20] [PASSED] 0x9A40 (TIGERLAKE)
[10:33:20] [PASSED] 0x9A49 (TIGERLAKE)
[10:33:20] [PASSED] 0x9A59 (TIGERLAKE)
[10:33:20] [PASSED] 0x9A78 (TIGERLAKE)
[10:33:20] [PASSED] 0x9AC0 (TIGERLAKE)
[10:33:20] [PASSED] 0x9AC9 (TIGERLAKE)
[10:33:20] [PASSED] 0x9AD9 (TIGERLAKE)
[10:33:20] [PASSED] 0x9AF8 (TIGERLAKE)
[10:33:20] [PASSED] 0x4C80 (ROCKETLAKE)
[10:33:20] [PASSED] 0x4C8A (ROCKETLAKE)
[10:33:20] [PASSED] 0x4C8B (ROCKETLAKE)
[10:33:20] [PASSED] 0x4C8C (ROCKETLAKE)
[10:33:20] [PASSED] 0x4C90 (ROCKETLAKE)
[10:33:20] [PASSED] 0x4C9A (ROCKETLAKE)
[10:33:20] [PASSED] 0x4680 (ALDERLAKE_S)
[10:33:20] [PASSED] 0x4682 (ALDERLAKE_S)
[10:33:20] [PASSED] 0x4688 (ALDERLAKE_S)
[10:33:20] [PASSED] 0x468A (ALDERLAKE_S)
[10:33:20] [PASSED] 0x468B (ALDERLAKE_S)
[10:33:20] [PASSED] 0x4690 (ALDERLAKE_S)
[10:33:20] [PASSED] 0x4692 (ALDERLAKE_S)
[10:33:20] [PASSED] 0x4693 (ALDERLAKE_S)
[10:33:20] [PASSED] 0x46A0 (ALDERLAKE_P)
[10:33:20] [PASSED] 0x46A1 (ALDERLAKE_P)
[10:33:20] [PASSED] 0x46A2 (ALDERLAKE_P)
[10:33:20] [PASSED] 0x46A3 (ALDERLAKE_P)
[10:33:20] [PASSED] 0x46A6 (ALDERLAKE_P)
[10:33:20] [PASSED] 0x46A8 (ALDERLAKE_P)
[10:33:20] [PASSED] 0x46AA (ALDERLAKE_P)
[10:33:20] [PASSED] 0x462A (ALDERLAKE_P)
[10:33:20] [PASSED] 0x4626 (ALDERLAKE_P)
[10:33:20] [PASSED] 0x4628 (ALDERLAKE_P)
[10:33:20] [PASSED] 0x46B0 (ALDERLAKE_P)
[10:33:20] [PASSED] 0x46B1 (ALDERLAKE_P)
[10:33:20] [PASSED] 0x46B2 (ALDERLAKE_P)
[10:33:20] [PASSED] 0x46B3 (ALDERLAKE_P)
[10:33:20] [PASSED] 0x46C0 (ALDERLAKE_P)
[10:33:20] [PASSED] 0x46C1 (ALDERLAKE_P)
[10:33:20] [PASSED] 0x46C2 (ALDERLAKE_P)
[10:33:20] [PASSED] 0x46C3 (ALDERLAKE_P)
[10:33:20] [PASSED] 0x46D0 (ALDERLAKE_N)
[10:33:20] [PASSED] 0x46D1 (ALDERLAKE_N)
[10:33:20] [PASSED] 0x46D2 (ALDERLAKE_N)
[10:33:20] [PASSED] 0x46D3 (ALDERLAKE_N)
[10:33:20] [PASSED] 0x46D4 (ALDERLAKE_N)
[10:33:20] [PASSED] 0xA721 (ALDERLAKE_P)
[10:33:20] [PASSED] 0xA7A1 (ALDERLAKE_P)
[10:33:20] [PASSED] 0xA7A9 (ALDERLAKE_P)
[10:33:20] [PASSED] 0xA7AC (ALDERLAKE_P)
[10:33:20] [PASSED] 0xA7AD (ALDERLAKE_P)
[10:33:20] [PASSED] 0xA720 (ALDERLAKE_P)
[10:33:20] [PASSED] 0xA7A0 (ALDERLAKE_P)
[10:33:20] [PASSED] 0xA7A8 (ALDERLAKE_P)
[10:33:20] [PASSED] 0xA7AA (ALDERLAKE_P)
[10:33:20] [PASSED] 0xA7AB (ALDERLAKE_P)
[10:33:20] [PASSED] 0xA780 (ALDERLAKE_S)
[10:33:20] [PASSED] 0xA781 (ALDERLAKE_S)
[10:33:20] [PASSED] 0xA782 (ALDERLAKE_S)
[10:33:20] [PASSED] 0xA783 (ALDERLAKE_S)
[10:33:20] [PASSED] 0xA788 (ALDERLAKE_S)
[10:33:20] [PASSED] 0xA789 (ALDERLAKE_S)
[10:33:20] [PASSED] 0xA78A (ALDERLAKE_S)
[10:33:20] [PASSED] 0xA78B (ALDERLAKE_S)
[10:33:20] [PASSED] 0x4905 (DG1)
[10:33:20] [PASSED] 0x4906 (DG1)
[10:33:20] [PASSED] 0x4907 (DG1)
[10:33:20] [PASSED] 0x4908 (DG1)
[10:33:20] [PASSED] 0x4909 (DG1)
[10:33:20] [PASSED] 0x56C0 (DG2)
[10:33:20] [PASSED] 0x56C2 (DG2)
[10:33:20] [PASSED] 0x56C1 (DG2)
[10:33:20] [PASSED] 0x7D51 (METEORLAKE)
[10:33:20] [PASSED] 0x7DD1 (METEORLAKE)
[10:33:20] [PASSED] 0x7D41 (METEORLAKE)
[10:33:20] [PASSED] 0x7D67 (METEORLAKE)
[10:33:20] [PASSED] 0xB640 (METEORLAKE)
[10:33:20] [PASSED] 0x56A0 (DG2)
[10:33:20] [PASSED] 0x56A1 (DG2)
[10:33:20] [PASSED] 0x56A2 (DG2)
[10:33:20] [PASSED] 0x56BE (DG2)
[10:33:20] [PASSED] 0x56BF (DG2)
[10:33:20] [PASSED] 0x5690 (DG2)
[10:33:20] [PASSED] 0x5691 (DG2)
[10:33:20] [PASSED] 0x5692 (DG2)
[10:33:20] [PASSED] 0x56A5 (DG2)
[10:33:20] [PASSED] 0x56A6 (DG2)
[10:33:20] [PASSED] 0x56B0 (DG2)
[10:33:20] [PASSED] 0x56B1 (DG2)
[10:33:20] [PASSED] 0x56BA (DG2)
[10:33:20] [PASSED] 0x56BB (DG2)
[10:33:20] [PASSED] 0x56BC (DG2)
[10:33:20] [PASSED] 0x56BD (DG2)
[10:33:20] [PASSED] 0x5693 (DG2)
[10:33:20] [PASSED] 0x5694 (DG2)
[10:33:20] [PASSED] 0x5695 (DG2)
[10:33:20] [PASSED] 0x56A3 (DG2)
[10:33:20] [PASSED] 0x56A4 (DG2)
[10:33:20] [PASSED] 0x56B2 (DG2)
[10:33:20] [PASSED] 0x56B3 (DG2)
[10:33:20] [PASSED] 0x5696 (DG2)
[10:33:20] [PASSED] 0x5697 (DG2)
[10:33:20] [PASSED] 0xB69 (PVC)
[10:33:20] [PASSED] 0xB6E (PVC)
[10:33:20] [PASSED] 0xBD4 (PVC)
[10:33:20] [PASSED] 0xBD5 (PVC)
[10:33:20] [PASSED] 0xBD6 (PVC)
[10:33:20] [PASSED] 0xBD7 (PVC)
[10:33:20] [PASSED] 0xBD8 (PVC)
[10:33:20] [PASSED] 0xBD9 (PVC)
[10:33:20] [PASSED] 0xBDA (PVC)
[10:33:20] [PASSED] 0xBDB (PVC)
[10:33:20] [PASSED] 0xBE0 (PVC)
[10:33:20] [PASSED] 0xBE1 (PVC)
[10:33:20] [PASSED] 0xBE5 (PVC)
[10:33:20] [PASSED] 0x7D40 (METEORLAKE)
[10:33:20] [PASSED] 0x7D45 (METEORLAKE)
[10:33:20] [PASSED] 0x7D55 (METEORLAKE)
[10:33:20] [PASSED] 0x7D60 (METEORLAKE)
[10:33:20] [PASSED] 0x7DD5 (METEORLAKE)
[10:33:20] [PASSED] 0x6420 (LUNARLAKE)
[10:33:20] [PASSED] 0x64A0 (LUNARLAKE)
[10:33:20] [PASSED] 0x64B0 (LUNARLAKE)
[10:33:20] [PASSED] 0xE202 (BATTLEMAGE)
[10:33:20] [PASSED] 0xE209 (BATTLEMAGE)
[10:33:20] [PASSED] 0xE20B (BATTLEMAGE)
[10:33:20] [PASSED] 0xE20C (BATTLEMAGE)
[10:33:20] [PASSED] 0xE20D (BATTLEMAGE)
[10:33:20] [PASSED] 0xE210 (BATTLEMAGE)
[10:33:20] [PASSED] 0xE211 (BATTLEMAGE)
[10:33:20] [PASSED] 0xE212 (BATTLEMAGE)
[10:33:20] [PASSED] 0xE216 (BATTLEMAGE)
[10:33:20] [PASSED] 0xE220 (BATTLEMAGE)
[10:33:20] [PASSED] 0xE221 (BATTLEMAGE)
[10:33:20] [PASSED] 0xE222 (BATTLEMAGE)
[10:33:20] [PASSED] 0xE223 (BATTLEMAGE)
[10:33:20] [PASSED] 0xB080 (PANTHERLAKE)
[10:33:20] [PASSED] 0xB081 (PANTHERLAKE)
[10:33:20] [PASSED] 0xB082 (PANTHERLAKE)
[10:33:20] [PASSED] 0xB083 (PANTHERLAKE)
[10:33:20] [PASSED] 0xB084 (PANTHERLAKE)
[10:33:20] [PASSED] 0xB085 (PANTHERLAKE)
[10:33:20] [PASSED] 0xB086 (PANTHERLAKE)
[10:33:20] [PASSED] 0xB087 (PANTHERLAKE)
[10:33:20] [PASSED] 0xB08F (PANTHERLAKE)
[10:33:20] [PASSED] 0xB090 (PANTHERLAKE)
[10:33:20] [PASSED] 0xB0A0 (PANTHERLAKE)
[10:33:20] [PASSED] 0xB0B0 (PANTHERLAKE)
[10:33:20] [PASSED] 0xFD80 (PANTHERLAKE)
[10:33:20] [PASSED] 0xFD81 (PANTHERLAKE)
[10:33:20] [PASSED] 0xD740 (NOVALAKE_S)
[10:33:20] [PASSED] 0xD741 (NOVALAKE_S)
[10:33:20] [PASSED] 0xD742 (NOVALAKE_S)
[10:33:20] [PASSED] 0xD743 (NOVALAKE_S)
[10:33:20] [PASSED] 0xD744 (NOVALAKE_S)
[10:33:20] [PASSED] 0xD745 (NOVALAKE_S)
[10:33:20] [PASSED] 0x674C (CRESCENTISLAND)
[10:33:20] [PASSED] 0x674D (CRESCENTISLAND)
[10:33:20] [PASSED] 0x674E (CRESCENTISLAND)
[10:33:20] [PASSED] 0x674F (CRESCENTISLAND)
[10:33:20] [PASSED] 0x6750 (CRESCENTISLAND)
[10:33:20] [PASSED] 0xD750 (NOVALAKE_P)
[10:33:20] [PASSED] 0xD751 (NOVALAKE_P)
[10:33:20] [PASSED] 0xD752 (NOVALAKE_P)
[10:33:20] [PASSED] 0xD753 (NOVALAKE_P)
[10:33:20] [PASSED] 0xD754 (NOVALAKE_P)
[10:33:20] [PASSED] 0xD755 (NOVALAKE_P)
[10:33:20] [PASSED] 0xD756 (NOVALAKE_P)
[10:33:20] [PASSED] 0xD757 (NOVALAKE_P)
[10:33:20] [PASSED] 0xD75F (NOVALAKE_P)
[10:33:20] =============== [PASSED] check_platform_desc ===============
[10:33:20] ===================== [PASSED] xe_pci ======================
[10:33:20] ============= xe_rtp_tables_test (4 subtests) ==============
[10:33:20] ================== xe_rtp_table_gt_test  ===================
[10:33:20] [PASSED] gt_was/14011060649
[10:33:20] [PASSED] gt_was/14011059788
[10:33:20] [PASSED] gt_was/14015795083
[10:33:20] [PASSED] gt_was/16021867713
[10:33:20] [PASSED] gt_was/14019449301
[10:33:20] [PASSED] gt_was/16028005424
[10:33:20] [PASSED] gt_was/14026578760
[10:33:20] [PASSED] gt_was/1409420604
[10:33:20] [PASSED] gt_was/1408615072
[10:33:20] [PASSED] gt_was/22010523718
[10:33:20] [PASSED] gt_was/14011006942
[10:33:20] [PASSED] gt_was/14014830051
[10:33:20] [PASSED] gt_was/18018781329
[10:33:20] [PASSED] gt_was/1509235366
[10:33:20] [PASSED] gt_was/18018781329
[10:33:20] [PASSED] gt_was/16016694945
[10:33:20] [PASSED] gt_was/14018575942
[10:33:20] [PASSED] gt_was/22016670082
[10:33:20] [PASSED] gt_was/22016670082
[10:33:20] [PASSED] gt_was/14017421178
[10:33:20] [PASSED] gt_was/16025250150
[10:33:20] [PASSED] gt_was/14021871409
[10:33:20] [PASSED] gt_was/16021865536
[10:33:20] [PASSED] gt_was/14021486841
[10:33:20] [PASSED] gt_was/14025160223
[10:33:20] [PASSED] gt_was/14026144927, 16029437861
[10:33:20] [PASSED] gt_was/14025635424
[10:33:20] [PASSED] gt_was/16028005424
[10:33:20] ============== [PASSED] xe_rtp_table_gt_test ===============
[10:33:20] ================== xe_rtp_table_gt_test  ===================
[10:33:20] [PASSED] gt_tunings/Tuning: Blend Fill Caching Optimization Disable
[10:33:20] [PASSED] gt_tunings/Tuning: 32B Access Enable
[10:33:20] [PASSED] gt_tunings/Tuning: L3 cache
[10:33:20] [PASSED] gt_tunings/Tuning: L3 cache - media
[10:33:20] [PASSED] gt_tunings/Tuning: Compression Overfetch
[10:33:20] [PASSED] gt_tunings/Tuning: Compression Overfetch - media
[10:33:20] [PASSED] gt_tunings/Tuning: Enable compressible partial write overfetch in L3
[10:33:20] [PASSED] gt_tunings/Tuning: Enable compressible partial write overfetch in L3 - media
[10:33:20] [PASSED] gt_tunings/Tuning: L2 Overfetch Compressible Only
[10:33:20] [PASSED] gt_tunings/Tuning: L2 Overfetch Compressible Only - media
[10:33:20] [PASSED] gt_tunings/Tuning: Stateless compression control
[10:33:20] [PASSED] gt_tunings/Tuning: Stateless compression control - media
[10:33:20] [PASSED] gt_tunings/Tuning: L3 RW flush all Cache
[10:33:20] [PASSED] gt_tunings/Tuning: L3 RW flush all cache - media
[10:33:20] [PASSED] gt_tunings/Tuning: Set STLB Bank Hash Mode to 4KB
[10:33:20] ============== [PASSED] xe_rtp_table_gt_test ===============
[10:33:20] ================== xe_rtp_table_oob_test  ==================
[10:33:20] [PASSED] oob_was/1607983814
[10:33:20] [PASSED] oob_was/16010904313
[10:33:20] [PASSED] oob_was/18022495364
[10:33:20] [PASSED] oob_was/22012773006
[10:33:20] [PASSED] oob_was/14014475959
[10:33:20] [PASSED] oob_was/22011391025
[10:33:20] [PASSED] oob_was/22012727170
[10:33:20] [PASSED] oob_was/22012727685
[10:33:20] [PASSED] oob_was/22016596838
[10:33:20] [PASSED] oob_was/18020744125
[10:33:20] [PASSED] oob_was/1409600907
[10:33:20] [PASSED] oob_was/22014953428
[10:33:20] [PASSED] oob_was/16017236439
[10:33:20] [PASSED] oob_was/14019821291
[10:33:20] [PASSED] oob_was/14015076503
[10:33:20] [PASSED] oob_was/14018913170
[10:33:20] [PASSED] oob_was/14018094691
[10:33:20] [PASSED] oob_was/18024947630
[10:33:20] [PASSED] oob_was/16022287689
[10:33:20] [PASSED] oob_was/13011645652
[10:33:20] [PASSED] oob_was/14022293748
[10:33:20] [PASSED] oob_was/22019794406
[10:33:20] [PASSED] oob_was/22019338487
[10:33:20] [PASSED] oob_was/16023588340
[10:33:20] [PASSED] oob_was/14019789679
[10:33:20] [PASSED] oob_was/14022866841
[10:33:20] [PASSED] oob_was/16021333562
[10:33:20] [PASSED] oob_was/14016712196
[10:33:20] [PASSED] oob_was/14015568240
[10:33:20] [PASSED] oob_was/18013179988
[10:33:20] [PASSED] oob_was/1508761755
[10:33:20] [PASSED] oob_was/16023105232
[10:33:20] [PASSED] oob_was/16026508708
[10:33:20] [PASSED] oob_was/14020001231
[10:33:20] [PASSED] oob_was/16023683509
[10:33:20] [PASSED] oob_was/14025515070
[10:33:20] [PASSED] oob_was/15015404425_disable
[10:33:20] [PASSED] oob_was/16026007364
[10:33:20] [PASSED] oob_was/14020316580
[10:33:20] [PASSED] oob_was/14025883347
[10:33:20] ============== [PASSED] xe_rtp_table_oob_test ==============
[10:33:20] ================ xe_rtp_table_dev_oob_test  ================
[10:33:20] [PASSED] device_oob_was/22010954014
[10:33:20] [PASSED] device_oob_was/15015404425
[10:33:20] [PASSED] device_oob_was/22019338487_display
[10:33:20] [PASSED] device_oob_was/14022085890
[10:33:20] [PASSED] device_oob_was/14026539277
[10:33:20] [PASSED] device_oob_was/14026633728
[10:33:20] [PASSED] device_oob_was/14026746987
[10:33:20] [PASSED] device_oob_was/14026779378
[10:33:20] ============ [PASSED] xe_rtp_table_dev_oob_test ============
[10:33:20] =============== [PASSED] xe_rtp_tables_test ================
[10:33:20] =================== xe_rtp (3 subtests) ====================
[10:33:20] =================== xe_rtp_rules_tests  ====================
[10:33:20] [PASSED] no
[10:33:20] [PASSED] yes
[10:33:20] [PASSED] no-and-no
[10:33:20] [PASSED] no-and-yes
[10:33:20] [PASSED] yes-and-no
[10:33:20] [PASSED] yes-and-yes
[10:33:20] [PASSED] no-or-no
[10:33:20] [PASSED] no-or-yes
[10:33:20] [PASSED] yes-or-no
[10:33:20] [PASSED] yes-or-yes
[10:33:20] [PASSED] no-yes-or-yes-no
[10:33:20] [PASSED] no-yes-or-yes-yes
[10:33:20] [PASSED] yes-yes-or-no-yes
[10:33:20] [PASSED] yes-yes-or-yes-yes
[10:33:20] [PASSED] no-no-or-yes-or-no
[10:33:20] [PASSED] or
[10:33:20] [PASSED] or-yes
[10:33:20] [PASSED] or-no
[10:33:20] [PASSED] yes-or
[10:33:20] [PASSED] no-or
[10:33:20] [PASSED] no-or-or-yes
[10:33:20] [PASSED] yes-or-or-no
[10:33:20] [PASSED] no-or-or-no
[10:33:20] [PASSED] missing-context-engine-class
[10:33:20] [PASSED] missing-context-engine-class-or-yes
[10:33:20] [PASSED] missing-context-engine-class-or-or-yes
[10:33:20] =============== [PASSED] xe_rtp_rules_tests ================
[10:33:20] =============== xe_rtp_process_to_sr_tests  ================
[10:33:20] [PASSED] coalesce-same-reg
[10:33:20] [PASSED] no-match-no-add
[10:33:20] [PASSED] two-regs-two-entries
[10:33:20] [PASSED] clr-one-set-other
[10:33:20] [PASSED] set-field
[10:33:20] [PASSED] conflict-duplicate
[10:33:20] [PASSED] conflict-not-disjoint
[10:33:20] [PASSED] conflict-reg-type
[10:33:20] [PASSED] bad-mcr-reg-forced-to-regular
[10:33:20] [PASSED] bad-regular-reg-forced-to-mcr
[10:33:20] =========== [PASSED] xe_rtp_process_to_sr_tests ============
[10:33:20] ================== xe_rtp_process_tests  ===================
[10:33:20] [PASSED] active1
[10:33:20] [PASSED] active2
[10:33:20] [PASSED] active-inactive
[10:33:20] [PASSED] inactive-active
[10:33:20] [PASSED] inactive-active-inactive
[10:33:20] [PASSED] inactive-inactive-inactive
[10:33:20] ============== [PASSED] xe_rtp_process_tests ===============
[10:33:20] ===================== [PASSED] xe_rtp ======================
[10:33:20] ==================== xe_wa (1 subtest) =====================
[10:33:20] ======================== xe_wa_gt  =========================
[10:33:20] [PASSED] TIGERLAKE B0
[10:33:20] [PASSED] DG1 A0
[10:33:20] [PASSED] DG1 B0
[10:33:20] [PASSED] ALDERLAKE_S A0
[10:33:20] [PASSED] ALDERLAKE_S B0
[10:33:20] [PASSED] ALDERLAKE_S C0
[10:33:20] [PASSED] ALDERLAKE_S D0
[10:33:20] [PASSED] ALDERLAKE_P A0
[10:33:20] [PASSED] ALDERLAKE_P B0
[10:33:20] [PASSED] ALDERLAKE_P C0
[10:33:20] [PASSED] ALDERLAKE_S RPLS D0
[10:33:20] [PASSED] ALDERLAKE_P RPLU E0
[10:33:20] [PASSED] DG2 G10 C0
[10:33:20] [PASSED] DG2 G11 B1
[10:33:20] [PASSED] DG2 G12 A1
[10:33:20] [PASSED] METEORLAKE 12.70(Xe_LPG) A0 13.00(Xe_LPM+) A0
[10:33:20] [PASSED] METEORLAKE 12.71(Xe_LPG) A0 13.00(Xe_LPM+) A0
[10:33:20] [PASSED] METEORLAKE 12.74(Xe_LPG+) A0 13.00(Xe_LPM+) A0
[10:33:20] [PASSED] LUNARLAKE 20.04(Xe2_LPG) A0 20.00(Xe2_LPM) A0
[10:33:20] [PASSED] LUNARLAKE 20.04(Xe2_LPG) B0 20.00(Xe2_LPM) A0
[10:33:20] [PASSED] BATTLEMAGE 20.01(Xe2_HPG) A0 13.01(Xe2_HPM) A1
[10:33:20] [PASSED] PANTHERLAKE 30.00(Xe3_LPG) A0 30.00(Xe3_LPM) A0
[10:33:20] ==================== [PASSED] xe_wa_gt =====================
[10:33:20] ====================== [PASSED] xe_wa ======================
[10:33:20] ============================================================
[10:33:20] Testing complete. Ran 715 tests: passed: 697, skipped: 18
[10:33:20] Elapsed time: 36.134s total, 4.276s configuring, 31.142s building, 0.687s running

+ /kernel/tools/testing/kunit/kunit.py run --kunitconfig /kernel/drivers/gpu/drm/tests/.kunitconfig [10:33:20] Configuring KUnit Kernel ...
Regenerating .config ...
Populating config with:

$ make ARCH=um O=.kunit olddefconfig
[10:33:22] Building KUnit Kernel ...
Populating config with:
$ make ARCH=um O=.kunit olddefconfig
Building with:
$ make all compile_commands.json scripts_gdb ARCH=um O=.kunit --jobs=48
[10:33:46] Starting KUnit Kernel (1/1)...
[10:33:46] ============================================================
Running tests with:
$ .kunit/linux kunit.enable=1 mem=1G console=tty kunit_shutdown=halt
[10:33:46] ============ drm_test_pick_cmdline (2 subtests) ============
[10:33:46] [PASSED] drm_test_pick_cmdline_res_1920_1080_60
[10:33:46] =============== drm_test_pick_cmdline_named  ===============
[10:33:46] [PASSED] NTSC
[10:33:46] [PASSED] NTSC-J
[10:33:46] [PASSED] PAL
[10:33:46] [PASSED] PAL-M
[10:33:46] =========== [PASSED] drm_test_pick_cmdline_named ===========
[10:33:46] ============== [PASSED] drm_test_pick_cmdline ==============
[10:33:46] == drm_test_atomic_get_connector_for_encoder (1 subtest) ===
[10:33:46] [PASSED] drm_test_drm_atomic_get_connector_for_encoder
[10:33:46] ==== [PASSED] drm_test_atomic_get_connector_for_encoder ====
[10:33:46] =========== drm_validate_clone_mode (2 subtests) ===========
[10:33:46] ============== drm_test_check_in_clone_mode  ===============
[10:33:46] [PASSED] in_clone_mode
[10:33:46] [PASSED] not_in_clone_mode
[10:33:46] ========== [PASSED] drm_test_check_in_clone_mode ===========
[10:33:46] =============== drm_test_check_valid_clones  ===============
[10:33:46] [PASSED] not_in_clone_mode
[10:33:46] [PASSED] valid_clone
[10:33:46] [PASSED] invalid_clone
[10:33:46] =========== [PASSED] drm_test_check_valid_clones ===========
[10:33:46] ============= [PASSED] drm_validate_clone_mode =============
[10:33:46] ============= drm_validate_modeset (1 subtest) =============
[10:33:46] [PASSED] drm_test_check_connector_changed_modeset
[10:33:46] ============== [PASSED] drm_validate_modeset ===============
[10:33:46] ====== drm_test_bridge_get_current_state (2 subtests) ======
[10:33:46] [PASSED] drm_test_drm_bridge_get_current_state_atomic
[10:33:46] [PASSED] drm_test_drm_bridge_get_current_state_legacy
[10:33:46] ======== [PASSED] drm_test_bridge_get_current_state ========
[10:33:46] ====== drm_test_bridge_helper_reset_crtc (3 subtests) ======
[10:33:46] [PASSED] drm_test_drm_bridge_helper_reset_crtc_atomic
[10:33:46] [PASSED] drm_test_drm_bridge_helper_reset_crtc_atomic_disabled
[10:33:46] [PASSED] drm_test_drm_bridge_helper_reset_crtc_legacy
[10:33:46] ======== [PASSED] drm_test_bridge_helper_reset_crtc ========
[10:33:46] ============== drm_bridge_alloc (2 subtests) ===============
[10:33:46] [PASSED] drm_test_drm_bridge_alloc_basic
[10:33:46] [PASSED] drm_test_drm_bridge_alloc_get_put
[10:33:46] ================ [PASSED] drm_bridge_alloc =================
[10:33:46] ============= drm_cmdline_parser (40 subtests) =============
[10:33:46] [PASSED] drm_test_cmdline_force_d_only
[10:33:46] [PASSED] drm_test_cmdline_force_D_only_dvi
[10:33:46] [PASSED] drm_test_cmdline_force_D_only_hdmi
[10:33:46] [PASSED] drm_test_cmdline_force_D_only_not_digital
[10:33:46] [PASSED] drm_test_cmdline_force_e_only
[10:33:46] [PASSED] drm_test_cmdline_res
[10:33:46] [PASSED] drm_test_cmdline_res_vesa
[10:33:46] [PASSED] drm_test_cmdline_res_vesa_rblank
[10:33:46] [PASSED] drm_test_cmdline_res_rblank
[10:33:46] [PASSED] drm_test_cmdline_res_bpp
[10:33:46] [PASSED] drm_test_cmdline_res_refresh
[10:33:46] [PASSED] drm_test_cmdline_res_bpp_refresh
[10:33:46] [PASSED] drm_test_cmdline_res_bpp_refresh_interlaced
[10:33:46] [PASSED] drm_test_cmdline_res_bpp_refresh_margins
[10:33:46] [PASSED] drm_test_cmdline_res_bpp_refresh_force_off
[10:33:46] [PASSED] drm_test_cmdline_res_bpp_refresh_force_on
[10:33:46] [PASSED] drm_test_cmdline_res_bpp_refresh_force_on_analog
[10:33:46] [PASSED] drm_test_cmdline_res_bpp_refresh_force_on_digital
[10:33:46] [PASSED] drm_test_cmdline_res_bpp_refresh_interlaced_margins_force_on
[10:33:46] [PASSED] drm_test_cmdline_res_margins_force_on
[10:33:46] [PASSED] drm_test_cmdline_res_vesa_margins
[10:33:46] [PASSED] drm_test_cmdline_name
[10:33:46] [PASSED] drm_test_cmdline_name_bpp
[10:33:46] [PASSED] drm_test_cmdline_name_option
[10:33:46] [PASSED] drm_test_cmdline_name_bpp_option
[10:33:46] [PASSED] drm_test_cmdline_rotate_0
[10:33:46] [PASSED] drm_test_cmdline_rotate_90
[10:33:46] [PASSED] drm_test_cmdline_rotate_180
[10:33:46] [PASSED] drm_test_cmdline_rotate_270
[10:33:46] [PASSED] drm_test_cmdline_hmirror
[10:33:46] [PASSED] drm_test_cmdline_vmirror
[10:33:46] [PASSED] drm_test_cmdline_margin_options
[10:33:46] [PASSED] drm_test_cmdline_multiple_options
[10:33:46] [PASSED] drm_test_cmdline_bpp_extra_and_option
[10:33:46] [PASSED] drm_test_cmdline_extra_and_option
[10:33:46] [PASSED] drm_test_cmdline_freestanding_options
[10:33:46] [PASSED] drm_test_cmdline_freestanding_force_e_and_options
[10:33:46] [PASSED] drm_test_cmdline_panel_orientation
[10:33:46] ================ drm_test_cmdline_invalid  =================
[10:33:46] [PASSED] margin_only
[10:33:46] [PASSED] interlace_only
[10:33:46] [PASSED] res_missing_x
[10:33:46] [PASSED] res_missing_y
[10:33:46] [PASSED] res_bad_y
[10:33:46] [PASSED] res_missing_y_bpp
[10:33:46] [PASSED] res_bad_bpp
[10:33:46] [PASSED] res_bad_refresh
[10:33:46] [PASSED] res_bpp_refresh_force_on_off
[10:33:46] [PASSED] res_invalid_mode
[10:33:46] [PASSED] res_bpp_wrong_place_mode
[10:33:46] [PASSED] name_bpp_refresh
[10:33:46] [PASSED] name_refresh
[10:33:46] [PASSED] name_refresh_wrong_mode
[10:33:46] [PASSED] name_refresh_invalid_mode
[10:33:46] [PASSED] rotate_multiple
[10:33:46] [PASSED] rotate_invalid_val
[10:33:46] [PASSED] rotate_truncated
[10:33:46] [PASSED] invalid_option
[10:33:46] [PASSED] invalid_tv_option
[10:33:46] [PASSED] truncated_tv_option
[10:33:46] ============ [PASSED] drm_test_cmdline_invalid =============
[10:33:46] =============== drm_test_cmdline_tv_options  ===============
[10:33:46] [PASSED] NTSC
[10:33:46] [PASSED] NTSC_443
[10:33:46] [PASSED] NTSC_J
[10:33:46] [PASSED] PAL
[10:33:46] [PASSED] PAL_M
[10:33:46] [PASSED] PAL_N
[10:33:46] [PASSED] SECAM
[10:33:46] [PASSED] MONO_525
[10:33:46] [PASSED] MONO_625
[10:33:46] =========== [PASSED] drm_test_cmdline_tv_options ===========
[10:33:46] =============== [PASSED] drm_cmdline_parser ================
[10:33:46] ========== drmm_connector_hdmi_init (20 subtests) ==========
[10:33:46] [PASSED] drm_test_connector_hdmi_init_valid
[10:33:46] [PASSED] drm_test_connector_hdmi_init_bpc_8
[10:33:46] [PASSED] drm_test_connector_hdmi_init_bpc_10
[10:33:46] [PASSED] drm_test_connector_hdmi_init_bpc_12
[10:33:46] [PASSED] drm_test_connector_hdmi_init_bpc_invalid
[10:33:46] [PASSED] drm_test_connector_hdmi_init_bpc_null
[10:33:46] [PASSED] drm_test_connector_hdmi_init_formats_empty
[10:33:46] [PASSED] drm_test_connector_hdmi_init_formats_no_rgb
[10:33:46] === drm_test_connector_hdmi_init_formats_yuv420_allowed  ===
[10:33:46] [PASSED] supported_formats=0x9 yuv420_allowed=1
[10:33:46] [PASSED] supported_formats=0x9 yuv420_allowed=0
[10:33:46] [PASSED] supported_formats=0x5 yuv420_allowed=1
[10:33:46] [PASSED] supported_formats=0x5 yuv420_allowed=0
[10:33:46] === [PASSED] drm_test_connector_hdmi_init_formats_yuv420_allowed ===
[10:33:46] [PASSED] drm_test_connector_hdmi_init_null_ddc
[10:33:46] [PASSED] drm_test_connector_hdmi_init_null_product
[10:33:46] [PASSED] drm_test_connector_hdmi_init_null_vendor
[10:33:46] [PASSED] drm_test_connector_hdmi_init_product_length_exact
[10:33:46] [PASSED] drm_test_connector_hdmi_init_product_length_too_long
[10:33:46] [PASSED] drm_test_connector_hdmi_init_product_valid
[10:33:46] [PASSED] drm_test_connector_hdmi_init_vendor_length_exact
[10:33:46] [PASSED] drm_test_connector_hdmi_init_vendor_length_too_long
[10:33:46] [PASSED] drm_test_connector_hdmi_init_vendor_valid
[10:33:46] ========= drm_test_connector_hdmi_init_type_valid  =========
[10:33:46] [PASSED] HDMI-A
[10:33:46] [PASSED] HDMI-B
[10:33:46] ===== [PASSED] drm_test_connector_hdmi_init_type_valid =====
[10:33:46] ======== drm_test_connector_hdmi_init_type_invalid  ========
[10:33:46] [PASSED] Unknown
[10:33:46] [PASSED] VGA
[10:33:46] [PASSED] DVI-I
[10:33:46] [PASSED] DVI-D
[10:33:46] [PASSED] DVI-A
[10:33:46] [PASSED] Composite
[10:33:46] [PASSED] SVIDEO
[10:33:46] [PASSED] LVDS
[10:33:46] [PASSED] Component
[10:33:46] [PASSED] DIN
[10:33:46] [PASSED] DP
[10:33:46] [PASSED] TV
[10:33:46] [PASSED] eDP
[10:33:46] [PASSED] Virtual
[10:33:46] [PASSED] DSI
[10:33:46] [PASSED] DPI
[10:33:46] [PASSED] Writeback
[10:33:46] [PASSED] SPI
[10:33:46] [PASSED] USB
[10:33:46] ==== [PASSED] drm_test_connector_hdmi_init_type_invalid ====
[10:33:46] ============ [PASSED] drmm_connector_hdmi_init =============
[10:33:46] ============= drmm_connector_init (3 subtests) =============
[10:33:46] [PASSED] drm_test_drmm_connector_init
[10:33:46] [PASSED] drm_test_drmm_connector_init_null_ddc
[10:33:46] ========= drm_test_drmm_connector_init_type_valid  =========
[10:33:46] [PASSED] Unknown
[10:33:46] [PASSED] VGA
[10:33:46] [PASSED] DVI-I
[10:33:46] [PASSED] DVI-D
[10:33:46] [PASSED] DVI-A
[10:33:46] [PASSED] Composite
[10:33:46] [PASSED] SVIDEO
[10:33:46] [PASSED] LVDS
[10:33:46] [PASSED] Component
[10:33:46] [PASSED] DIN
[10:33:46] [PASSED] DP
[10:33:46] [PASSED] HDMI-A
[10:33:46] [PASSED] HDMI-B
[10:33:46] [PASSED] TV
[10:33:46] [PASSED] eDP
[10:33:46] [PASSED] Virtual
[10:33:46] [PASSED] DSI
[10:33:46] [PASSED] DPI
[10:33:46] [PASSED] Writeback
[10:33:46] [PASSED] SPI
[10:33:46] [PASSED] USB
[10:33:46] ===== [PASSED] drm_test_drmm_connector_init_type_valid =====
[10:33:46] =============== [PASSED] drmm_connector_init ===============
[10:33:46] ========= drm_connector_dynamic_init (6 subtests) ==========
[10:33:46] [PASSED] drm_test_drm_connector_dynamic_init
[10:33:46] [PASSED] drm_test_drm_connector_dynamic_init_null_ddc
[10:33:46] [PASSED] drm_test_drm_connector_dynamic_init_not_added
[10:33:46] [PASSED] drm_test_drm_connector_dynamic_init_properties
[10:33:46] ===== drm_test_drm_connector_dynamic_init_type_valid  ======
[10:33:46] [PASSED] Unknown
[10:33:46] [PASSED] VGA
[10:33:46] [PASSED] DVI-I
[10:33:46] [PASSED] DVI-D
[10:33:46] [PASSED] DVI-A
[10:33:46] [PASSED] Composite
[10:33:46] [PASSED] SVIDEO
[10:33:46] [PASSED] LVDS
[10:33:46] [PASSED] Component
[10:33:46] [PASSED] DIN
[10:33:46] [PASSED] DP
[10:33:46] [PASSED] HDMI-A
[10:33:46] [PASSED] HDMI-B
[10:33:46] [PASSED] TV
[10:33:46] [PASSED] eDP
[10:33:46] [PASSED] Virtual
[10:33:46] [PASSED] DSI
[10:33:46] [PASSED] DPI
[10:33:46] [PASSED] Writeback
[10:33:46] [PASSED] SPI
[10:33:46] [PASSED] USB
[10:33:46] = [PASSED] drm_test_drm_connector_dynamic_init_type_valid ==
[10:33:46] ======== drm_test_drm_connector_dynamic_init_name  =========
[10:33:46] [PASSED] Unknown
[10:33:46] [PASSED] VGA
[10:33:46] [PASSED] DVI-I
[10:33:46] [PASSED] DVI-D
[10:33:46] [PASSED] DVI-A
[10:33:46] [PASSED] Composite
[10:33:46] [PASSED] SVIDEO
[10:33:46] [PASSED] LVDS
[10:33:46] [PASSED] Component
[10:33:46] [PASSED] DIN
[10:33:46] [PASSED] DP
[10:33:46] [PASSED] HDMI-A
[10:33:46] [PASSED] HDMI-B
[10:33:46] [PASSED] TV
[10:33:46] [PASSED] eDP
[10:33:46] [PASSED] Virtual
[10:33:46] [PASSED] DSI
[10:33:46] [PASSED] DPI
[10:33:46] [PASSED] Writeback
[10:33:46] [PASSED] SPI
[10:33:46] [PASSED] USB
[10:33:46] ==== [PASSED] drm_test_drm_connector_dynamic_init_name =====
[10:33:46] =========== [PASSED] drm_connector_dynamic_init ============
[10:33:46] ==== drm_connector_dynamic_register_early (4 subtests) =====
[10:33:46] [PASSED] drm_test_drm_connector_dynamic_register_early_on_list
[10:33:46] [PASSED] drm_test_drm_connector_dynamic_register_early_defer
[10:33:46] [PASSED] drm_test_drm_connector_dynamic_register_early_no_init
[10:33:46] [PASSED] drm_test_drm_connector_dynamic_register_early_no_mode_object
[10:33:46] ====== [PASSED] drm_connector_dynamic_register_early =======
[10:33:46] ======= drm_connector_dynamic_register (7 subtests) ========
[10:33:46] [PASSED] drm_test_drm_connector_dynamic_register_on_list
[10:33:46] [PASSED] drm_test_drm_connector_dynamic_register_no_defer
[10:33:46] [PASSED] drm_test_drm_connector_dynamic_register_no_init
[10:33:46] [PASSED] drm_test_drm_connector_dynamic_register_mode_object
[10:33:46] [PASSED] drm_test_drm_connector_dynamic_register_sysfs
[10:33:46] [PASSED] drm_test_drm_connector_dynamic_register_sysfs_name
[10:33:46] [PASSED] drm_test_drm_connector_dynamic_register_debugfs
[10:33:46] ========= [PASSED] drm_connector_dynamic_register ==========
[10:33:46] = drm_connector_attach_broadcast_rgb_property (2 subtests) =
[10:33:46] [PASSED] drm_test_drm_connector_attach_broadcast_rgb_property
[10:33:46] [PASSED] drm_test_drm_connector_attach_broadcast_rgb_property_hdmi_connector
[10:33:46] === [PASSED] drm_connector_attach_broadcast_rgb_property ===
[10:33:46] ========== drm_get_tv_mode_from_name (2 subtests) ==========
[10:33:46] ========== drm_test_get_tv_mode_from_name_valid  ===========
[10:33:46] [PASSED] NTSC
[10:33:46] [PASSED] NTSC-443
[10:33:46] [PASSED] NTSC-J
[10:33:46] [PASSED] PAL
[10:33:46] [PASSED] PAL-M
[10:33:46] [PASSED] PAL-N
[10:33:46] [PASSED] SECAM
[10:33:46] [PASSED] Mono
[10:33:46] ====== [PASSED] drm_test_get_tv_mode_from_name_valid =======
[10:33:46] [PASSED] drm_test_get_tv_mode_from_name_truncated
[10:33:46] ============ [PASSED] drm_get_tv_mode_from_name ============
[10:33:46] = drm_test_connector_hdmi_compute_mode_clock (12 subtests) =
[10:33:46] [PASSED] drm_test_drm_hdmi_compute_mode_clock_rgb
[10:33:46] [PASSED] drm_test_drm_hdmi_compute_mode_clock_rgb_10bpc
[10:33:46] [PASSED] drm_test_drm_hdmi_compute_mode_clock_rgb_10bpc_vic_1
[10:33:46] [PASSED] drm_test_drm_hdmi_compute_mode_clock_rgb_12bpc
[10:33:46] [PASSED] drm_test_drm_hdmi_compute_mode_clock_rgb_12bpc_vic_1
[10:33:46] [PASSED] drm_test_drm_hdmi_compute_mode_clock_rgb_double
[10:33:46] = drm_test_connector_hdmi_compute_mode_clock_yuv420_valid  =
[10:33:46] [PASSED] VIC 96
[10:33:46] [PASSED] VIC 97
[10:33:46] [PASSED] VIC 101
[10:33:46] [PASSED] VIC 102
[10:33:46] [PASSED] VIC 106
[10:33:46] [PASSED] VIC 107
[10:33:46] === [PASSED] drm_test_connector_hdmi_compute_mode_clock_yuv420_valid ===
[10:33:46] [PASSED] drm_test_connector_hdmi_compute_mode_clock_yuv420_10_bpc
[10:33:46] [PASSED] drm_test_connector_hdmi_compute_mode_clock_yuv420_12_bpc
[10:33:46] [PASSED] drm_test_connector_hdmi_compute_mode_clock_yuv422_8_bpc
[10:33:46] [PASSED] drm_test_connector_hdmi_compute_mode_clock_yuv422_10_bpc
[10:33:46] [PASSED] drm_test_connector_hdmi_compute_mode_clock_yuv422_12_bpc
[10:33:46] === [PASSED] drm_test_connector_hdmi_compute_mode_clock ====
[10:33:46] == drm_hdmi_connector_get_broadcast_rgb_name (2 subtests) ==
[10:33:46] === drm_test_drm_hdmi_connector_get_broadcast_rgb_name  ====
[10:33:46] [PASSED] Automatic
[10:33:46] [PASSED] Full
[10:33:46] [PASSED] Limited 16:235
[10:33:46] === [PASSED] drm_test_drm_hdmi_connector_get_broadcast_rgb_name ===
[10:33:46] [PASSED] drm_test_drm_hdmi_connector_get_broadcast_rgb_name_invalid
[10:33:46] ==== [PASSED] drm_hdmi_connector_get_broadcast_rgb_name ====
[10:33:46] == drm_hdmi_connector_get_output_format_name (2 subtests) ==
[10:33:46] === drm_test_drm_hdmi_connector_get_output_format_name  ====
[10:33:46] [PASSED] RGB
[10:33:46] [PASSED] YUV 4:2:0
[10:33:46] [PASSED] YUV 4:2:2
[10:33:46] [PASSED] YUV 4:4:4
[10:33:46] === [PASSED] drm_test_drm_hdmi_connector_get_output_format_name ===
[10:33:46] [PASSED] drm_test_drm_hdmi_connector_get_output_format_name_invalid
[10:33:46] ==== [PASSED] drm_hdmi_connector_get_output_format_name ====
[10:33:46] ============= drm_damage_helper (21 subtests) ==============
[10:33:46] [PASSED] drm_test_damage_iter_no_damage
[10:33:46] [PASSED] drm_test_damage_iter_no_damage_fractional_src
[10:33:46] [PASSED] drm_test_damage_iter_no_damage_src_moved
[10:33:46] [PASSED] drm_test_damage_iter_no_damage_fractional_src_moved
[10:33:46] [PASSED] drm_test_damage_iter_no_damage_not_visible
[10:33:46] [PASSED] drm_test_damage_iter_no_damage_no_crtc
[10:33:46] [PASSED] drm_test_damage_iter_no_damage_no_fb
[10:33:46] [PASSED] drm_test_damage_iter_simple_damage
[10:33:46] [PASSED] drm_test_damage_iter_single_damage
[10:33:46] [PASSED] drm_test_damage_iter_single_damage_intersect_src
[10:33:46] [PASSED] drm_test_damage_iter_single_damage_outside_src
[10:33:46] [PASSED] drm_test_damage_iter_single_damage_fractional_src
[10:33:46] [PASSED] drm_test_damage_iter_single_damage_intersect_fractional_src
[10:33:46] [PASSED] drm_test_damage_iter_single_damage_outside_fractional_src
[10:33:46] [PASSED] drm_test_damage_iter_single_damage_src_moved
[10:33:46] [PASSED] drm_test_damage_iter_single_damage_fractional_src_moved
[10:33:46] [PASSED] drm_test_damage_iter_damage
[10:33:46] [PASSED] drm_test_damage_iter_damage_one_intersect
[10:33:46] [PASSED] drm_test_damage_iter_damage_one_outside
[10:33:46] [PASSED] drm_test_damage_iter_damage_src_moved
[10:33:46] [PASSED] drm_test_damage_iter_damage_not_visible
[10:33:46] ================ [PASSED] drm_damage_helper ================
[10:33:46] ============== drm_dp_mst_helper (3 subtests) ==============
[10:33:46] ============== drm_test_dp_mst_calc_pbn_mode  ==============
[10:33:46] [PASSED] Clock 154000 BPP 30 DSC disabled
[10:33:46] [PASSED] Clock 234000 BPP 30 DSC disabled
[10:33:46] [PASSED] Clock 297000 BPP 24 DSC disabled
[10:33:46] [PASSED] Clock 332880 BPP 24 DSC enabled
[10:33:46] [PASSED] Clock 324540 BPP 24 DSC enabled
[10:33:46] ========== [PASSED] drm_test_dp_mst_calc_pbn_mode ==========
[10:33:46] ============== drm_test_dp_mst_calc_pbn_div  ===============
[10:33:46] [PASSED] Link rate 2000000 lane count 4
[10:33:46] [PASSED] Link rate 2000000 lane count 2
[10:33:46] [PASSED] Link rate 2000000 lane count 1
[10:33:46] [PASSED] Link rate 1350000 lane count 4
[10:33:46] [PASSED] Link rate 1350000 lane count 2
[10:33:46] [PASSED] Link rate 1350000 lane count 1
[10:33:46] [PASSED] Link rate 1000000 lane count 4
[10:33:46] [PASSED] Link rate 1000000 lane count 2
[10:33:46] [PASSED] Link rate 1000000 lane count 1
[10:33:46] [PASSED] Link rate 810000 lane count 4
[10:33:46] [PASSED] Link rate 810000 lane count 2
[10:33:46] [PASSED] Link rate 810000 lane count 1
[10:33:46] [PASSED] Link rate 540000 lane count 4
[10:33:46] [PASSED] Link rate 540000 lane count 2
[10:33:46] [PASSED] Link rate 540000 lane count 1
[10:33:46] [PASSED] Link rate 270000 lane count 4
[10:33:46] [PASSED] Link rate 270000 lane count 2
[10:33:46] [PASSED] Link rate 270000 lane count 1
[10:33:46] [PASSED] Link rate 162000 lane count 4
[10:33:46] [PASSED] Link rate 162000 lane count 2
[10:33:46] [PASSED] Link rate 162000 lane count 1
[10:33:46] ========== [PASSED] drm_test_dp_mst_calc_pbn_div ===========
[10:33:46] ========= drm_test_dp_mst_sideband_msg_req_decode  =========
[10:33:46] [PASSED] DP_ENUM_PATH_RESOURCES with port number
[10:33:46] [PASSED] DP_POWER_UP_PHY with port number
[10:33:46] [PASSED] DP_POWER_DOWN_PHY with port number
[10:33:46] [PASSED] DP_ALLOCATE_PAYLOAD with SDP stream sinks
[10:33:46] [PASSED] DP_ALLOCATE_PAYLOAD with port number
[10:33:46] [PASSED] DP_ALLOCATE_PAYLOAD with VCPI
[10:33:46] [PASSED] DP_ALLOCATE_PAYLOAD with PBN
[10:33:46] [PASSED] DP_QUERY_PAYLOAD with port number
[10:33:46] [PASSED] DP_QUERY_PAYLOAD with VCPI
[10:33:46] [PASSED] DP_REMOTE_DPCD_READ with port number
[10:33:46] [PASSED] DP_REMOTE_DPCD_READ with DPCD address
[10:33:46] [PASSED] DP_REMOTE_DPCD_READ with max number of bytes
[10:33:46] [PASSED] DP_REMOTE_DPCD_WRITE with port number
[10:33:46] [PASSED] DP_REMOTE_DPCD_WRITE with DPCD address
[10:33:46] [PASSED] DP_REMOTE_DPCD_WRITE with data array
[10:33:46] [PASSED] DP_REMOTE_I2C_READ with port number
[10:33:46] [PASSED] DP_REMOTE_I2C_READ with I2C device ID
[10:33:46] [PASSED] DP_REMOTE_I2C_READ with transactions array
[10:33:46] [PASSED] DP_REMOTE_I2C_WRITE with port number
[10:33:46] [PASSED] DP_REMOTE_I2C_WRITE with I2C device ID
[10:33:46] [PASSED] DP_REMOTE_I2C_WRITE with data array
[10:33:46] [PASSED] DP_QUERY_STREAM_ENC_STATUS with stream ID
[10:33:46] [PASSED] DP_QUERY_STREAM_ENC_STATUS with client ID
[10:33:46] [PASSED] DP_QUERY_STREAM_ENC_STATUS with stream event
[10:33:46] [PASSED] DP_QUERY_STREAM_ENC_STATUS with valid stream event
[10:33:46] [PASSED] DP_QUERY_STREAM_ENC_STATUS with stream behavior
[10:33:46] [PASSED] DP_QUERY_STREAM_ENC_STATUS with a valid stream behavior
[10:33:46] ===== [PASSED] drm_test_dp_mst_sideband_msg_req_decode =====
[10:33:46] ================ [PASSED] drm_dp_mst_helper ================
[10:33:46] ================== drm_exec (7 subtests) ===================
[10:33:46] [PASSED] sanitycheck
[10:33:46] [PASSED] test_lock
[10:33:46] [PASSED] test_lock_unlock
[10:33:46] [PASSED] test_duplicates
[10:33:46] [PASSED] test_prepare
[10:33:46] [PASSED] test_prepare_array
[10:33:46] [PASSED] test_multiple_loops
[10:33:46] ==================== [PASSED] drm_exec =====================
[10:33:46] =========== drm_format_helper_test (17 subtests) ===========
[10:33:46] ============== drm_test_fb_xrgb8888_to_gray8  ==============
[10:33:46] [PASSED] single_pixel_source_buffer
[10:33:46] [PASSED] single_pixel_clip_rectangle
[10:33:46] [PASSED] well_known_colors
[10:33:46] [PASSED] destination_pitch
[10:33:46] ========== [PASSED] drm_test_fb_xrgb8888_to_gray8 ==========
[10:33:46] ============= drm_test_fb_xrgb8888_to_rgb332  ==============
[10:33:46] [PASSED] single_pixel_source_buffer
[10:33:46] [PASSED] single_pixel_clip_rectangle
[10:33:46] [PASSED] well_known_colors
[10:33:46] [PASSED] destination_pitch
[10:33:46] ========= [PASSED] drm_test_fb_xrgb8888_to_rgb332 ==========
[10:33:46] ============= drm_test_fb_xrgb8888_to_rgb565  ==============
[10:33:46] [PASSED] single_pixel_source_buffer
[10:33:46] [PASSED] single_pixel_clip_rectangle
[10:33:46] [PASSED] well_known_colors
[10:33:46] [PASSED] destination_pitch
[10:33:46] ========= [PASSED] drm_test_fb_xrgb8888_to_rgb565 ==========
[10:33:46] ============ drm_test_fb_xrgb8888_to_xrgb1555  =============
[10:33:46] [PASSED] single_pixel_source_buffer
[10:33:46] [PASSED] single_pixel_clip_rectangle
[10:33:46] [PASSED] well_known_colors
[10:33:46] [PASSED] destination_pitch
[10:33:46] ======== [PASSED] drm_test_fb_xrgb8888_to_xrgb1555 =========
[10:33:46] ============ drm_test_fb_xrgb8888_to_argb1555  =============
[10:33:46] [PASSED] single_pixel_source_buffer
[10:33:46] [PASSED] single_pixel_clip_rectangle
[10:33:46] [PASSED] well_known_colors
[10:33:46] [PASSED] destination_pitch
[10:33:46] ======== [PASSED] drm_test_fb_xrgb8888_to_argb1555 =========
[10:33:46] ============ drm_test_fb_xrgb8888_to_rgba5551  =============
[10:33:46] [PASSED] single_pixel_source_buffer
[10:33:46] [PASSED] single_pixel_clip_rectangle
[10:33:46] [PASSED] well_known_colors
[10:33:46] [PASSED] destination_pitch
[10:33:46] ======== [PASSED] drm_test_fb_xrgb8888_to_rgba5551 =========
[10:33:46] ============= drm_test_fb_xrgb8888_to_rgb888  ==============
[10:33:46] [PASSED] single_pixel_source_buffer
[10:33:46] [PASSED] single_pixel_clip_rectangle
[10:33:46] [PASSED] well_known_colors
[10:33:46] [PASSED] destination_pitch
[10:33:46] ========= [PASSED] drm_test_fb_xrgb8888_to_rgb888 ==========
[10:33:46] ============= drm_test_fb_xrgb8888_to_bgr888  ==============
[10:33:46] [PASSED] single_pixel_source_buffer
[10:33:46] [PASSED] single_pixel_clip_rectangle
[10:33:46] [PASSED] well_known_colors
[10:33:46] [PASSED] destination_pitch
[10:33:46] ========= [PASSED] drm_test_fb_xrgb8888_to_bgr888 ==========
[10:33:46] ============ drm_test_fb_xrgb8888_to_argb8888  =============
[10:33:46] [PASSED] single_pixel_source_buffer
[10:33:46] [PASSED] single_pixel_clip_rectangle
[10:33:46] [PASSED] well_known_colors
[10:33:46] [PASSED] destination_pitch
[10:33:46] ======== [PASSED] drm_test_fb_xrgb8888_to_argb8888 =========
[10:33:46] =========== drm_test_fb_xrgb8888_to_xrgb2101010  ===========
[10:33:46] [PASSED] single_pixel_source_buffer
[10:33:46] [PASSED] single_pixel_clip_rectangle
[10:33:46] [PASSED] well_known_colors
[10:33:46] [PASSED] destination_pitch
[10:33:46] ======= [PASSED] drm_test_fb_xrgb8888_to_xrgb2101010 =======
[10:33:46] =========== drm_test_fb_xrgb8888_to_argb2101010  ===========
[10:33:46] [PASSED] single_pixel_source_buffer
[10:33:46] [PASSED] single_pixel_clip_rectangle
[10:33:46] [PASSED] well_known_colors
[10:33:46] [PASSED] destination_pitch
[10:33:46] ======= [PASSED] drm_test_fb_xrgb8888_to_argb2101010 =======
[10:33:46] ============== drm_test_fb_xrgb8888_to_mono  ===============
[10:33:46] [PASSED] single_pixel_source_buffer
[10:33:46] [PASSED] single_pixel_clip_rectangle
[10:33:46] [PASSED] well_known_colors
[10:33:46] [PASSED] destination_pitch
[10:33:46] ========== [PASSED] drm_test_fb_xrgb8888_to_mono ===========
[10:33:46] ==================== drm_test_fb_swab  =====================
[10:33:46] [PASSED] single_pixel_source_buffer
[10:33:46] [PASSED] single_pixel_clip_rectangle
[10:33:46] [PASSED] well_known_colors
[10:33:46] [PASSED] destination_pitch
[10:33:46] ================ [PASSED] drm_test_fb_swab =================
[10:33:46] ============ drm_test_fb_xrgb8888_to_xbgr8888  =============
[10:33:46] [PASSED] single_pixel_source_buffer
[10:33:46] [PASSED] single_pixel_clip_rectangle
[10:33:46] [PASSED] well_known_colors
[10:33:46] [PASSED] destination_pitch
[10:33:46] ======== [PASSED] drm_test_fb_xrgb8888_to_xbgr8888 =========
[10:33:46] ============ drm_test_fb_xrgb8888_to_abgr8888  =============
[10:33:46] [PASSED] single_pixel_source_buffer
[10:33:46] [PASSED] single_pixel_clip_rectangle
[10:33:46] [PASSED] well_known_colors
[10:33:46] [PASSED] destination_pitch
[10:33:46] ======== [PASSED] drm_test_fb_xrgb8888_to_abgr8888 =========
[10:33:46] ================= drm_test_fb_clip_offset  =================
[10:33:46] [PASSED] pass through
[10:33:46] [PASSED] horizontal offset
[10:33:46] [PASSED] vertical offset
[10:33:46] [PASSED] horizontal and vertical offset
[10:33:46] [PASSED] horizontal offset (custom pitch)
[10:33:46] [PASSED] vertical offset (custom pitch)
[10:33:46] [PASSED] horizontal and vertical offset (custom pitch)
[10:33:46] ============= [PASSED] drm_test_fb_clip_offset =============
[10:33:46] =================== drm_test_fb_memcpy  ====================
[10:33:46] [PASSED] single_pixel_source_buffer: XR24 little-endian (0x34325258)
[10:33:46] [PASSED] single_pixel_source_buffer: XRA8 little-endian (0x38415258)
[10:33:46] [PASSED] single_pixel_source_buffer: YU24 little-endian (0x34325559)
[10:33:46] [PASSED] single_pixel_clip_rectangle: XB24 little-endian (0x34324258)
[10:33:46] [PASSED] single_pixel_clip_rectangle: XRA8 little-endian (0x38415258)
[10:33:46] [PASSED] single_pixel_clip_rectangle: YU24 little-endian (0x34325559)
[10:33:46] [PASSED] well_known_colors: XB24 little-endian (0x34324258)
[10:33:46] [PASSED] well_known_colors: XRA8 little-endian (0x38415258)
[10:33:46] [PASSED] well_known_colors: YU24 little-endian (0x34325559)
[10:33:46] [PASSED] destination_pitch: XB24 little-endian (0x34324258)
[10:33:46] [PASSED] destination_pitch: XRA8 little-endian (0x38415258)
[10:33:46] [PASSED] destination_pitch: YU24 little-endian (0x34325559)
[10:33:46] =============== [PASSED] drm_test_fb_memcpy ================
[10:33:46] ============= [PASSED] drm_format_helper_test ==============
[10:33:46] ================= drm_format (18 subtests) =================
[10:33:46] [PASSED] drm_test_format_block_width_invalid
[10:33:46] [PASSED] drm_test_format_block_width_one_plane
[10:33:46] [PASSED] drm_test_format_block_width_two_plane
[10:33:46] [PASSED] drm_test_format_block_width_three_plane
[10:33:46] [PASSED] drm_test_format_block_width_tiled
[10:33:46] [PASSED] drm_test_format_block_height_invalid
[10:33:46] [PASSED] drm_test_format_block_height_one_plane
[10:33:46] [PASSED] drm_test_format_block_height_two_plane
[10:33:46] [PASSED] drm_test_format_block_height_three_plane
[10:33:46] [PASSED] drm_test_format_block_height_tiled
[10:33:46] [PASSED] drm_test_format_min_pitch_invalid
[10:33:46] [PASSED] drm_test_format_min_pitch_one_plane_8bpp
[10:33:46] [PASSED] drm_test_format_min_pitch_one_plane_16bpp
[10:33:46] [PASSED] drm_test_format_min_pitch_one_plane_24bpp
[10:33:46] [PASSED] drm_test_format_min_pitch_one_plane_32bpp
[10:33:46] [PASSED] drm_test_format_min_pitch_two_plane
[10:33:46] [PASSED] drm_test_format_min_pitch_three_plane_8bpp
[10:33:46] [PASSED] drm_test_format_min_pitch_tiled
[10:33:46] =================== [PASSED] drm_format ====================
[10:33:46] ============== drm_framebuffer (10 subtests) ===============
[10:33:46] ========== drm_test_framebuffer_check_src_coords  ==========
[10:33:46] [PASSED] Success: source fits into fb
[10:33:46] [PASSED] Fail: overflowing fb with x-axis coordinate
[10:33:46] [PASSED] Fail: overflowing fb with y-axis coordinate
[10:33:46] [PASSED] Fail: overflowing fb with source width
[10:33:46] [PASSED] Fail: overflowing fb with source height
[10:33:46] ====== [PASSED] drm_test_framebuffer_check_src_coords ======
[10:33:46] [PASSED] drm_test_framebuffer_cleanup
[10:33:46] =============== drm_test_framebuffer_create  ===============
[10:33:46] [PASSED] ABGR8888 normal sizes
[10:33:46] [PASSED] ABGR8888 max sizes
[10:33:46] [PASSED] ABGR8888 pitch greater than min required
[10:33:46] [PASSED] ABGR8888 pitch less than min required
[10:33:46] [PASSED] ABGR8888 Invalid width
[10:33:46] [PASSED] ABGR8888 Invalid buffer handle
[10:33:46] [PASSED] No pixel format
[10:33:46] [PASSED] ABGR8888 Width 0
[10:33:46] [PASSED] ABGR8888 Height 0
[10:33:46] [PASSED] ABGR8888 Out of bound height * pitch combination
[10:33:46] [PASSED] ABGR8888 Large buffer offset
[10:33:46] [PASSED] ABGR8888 Buffer offset for inexistent plane
[10:33:46] [PASSED] ABGR8888 Invalid flag
[10:33:46] [PASSED] ABGR8888 Set DRM_MODE_FB_MODIFIERS without modifiers
[10:33:46] [PASSED] ABGR8888 Valid buffer modifier
[10:33:46] [PASSED] ABGR8888 Invalid buffer modifier(DRM_FORMAT_MOD_SAMSUNG_64_32_TILE)
[10:33:46] [PASSED] ABGR8888 Extra pitches without DRM_MODE_FB_MODIFIERS
[10:33:46] [PASSED] ABGR8888 Extra pitches with DRM_MODE_FB_MODIFIERS
[10:33:46] [PASSED] NV12 Normal sizes
[10:33:46] [PASSED] NV12 Max sizes
[10:33:46] [PASSED] NV12 Invalid pitch
[10:33:46] [PASSED] NV12 Invalid modifier/missing DRM_MODE_FB_MODIFIERS flag
[10:33:46] [PASSED] NV12 different  modifier per-plane
[10:33:46] [PASSED] NV12 with DRM_FORMAT_MOD_SAMSUNG_64_32_TILE
[10:33:46] [PASSED] NV12 Valid modifiers without DRM_MODE_FB_MODIFIERS
[10:33:46] [PASSED] NV12 Modifier for inexistent plane
[10:33:46] [PASSED] NV12 Handle for inexistent plane
[10:33:46] [PASSED] NV12 Handle for inexistent plane without DRM_MODE_FB_MODIFIERS
[10:33:46] [PASSED] YVU420 DRM_MODE_FB_MODIFIERS set without modifier
[10:33:46] [PASSED] YVU420 Normal sizes
[10:33:46] [PASSED] YVU420 Max sizes
[10:33:46] [PASSED] YVU420 Invalid pitch
[10:33:46] [PASSED] YVU420 Different pitches
[10:33:46] [PASSED] YVU420 Different buffer offsets/pitches
[10:33:46] [PASSED] YVU420 Modifier set just for plane 0, without DRM_MODE_FB_MODIFIERS
[10:33:46] [PASSED] YVU420 Modifier set just for planes 0, 1, without DRM_MODE_FB_MODIFIERS
[10:33:46] [PASSED] YVU420 Modifier set just for plane 0, 1, with DRM_MODE_FB_MODIFIERS
[10:33:46] [PASSED] YVU420 Valid modifier
[10:33:46] [PASSED] YVU420 Different modifiers per plane
[10:33:46] [PASSED] YVU420 Modifier for inexistent plane
[10:33:46] [PASSED] YUV420_10BIT Invalid modifier(DRM_FORMAT_MOD_LINEAR)
[10:33:46] [PASSED] X0L2 Normal sizes
[10:33:46] [PASSED] X0L2 Max sizes
[10:33:46] [PASSED] X0L2 Invalid pitch
[10:33:46] [PASSED] X0L2 Pitch greater than minimum required
[10:33:46] [PASSED] X0L2 Handle for inexistent plane
[10:33:46] [PASSED] X0L2 Offset for inexistent plane, without DRM_MODE_FB_MODIFIERS set
[10:33:46] [PASSED] X0L2 Modifier without DRM_MODE_FB_MODIFIERS set
[10:33:46] [PASSED] X0L2 Valid modifier
[10:33:46] [PASSED] X0L2 Modifier for inexistent plane
[10:33:46] =========== [PASSED] drm_test_framebuffer_create ===========
[10:33:46] [PASSED] drm_test_framebuffer_free
[10:33:46] [PASSED] drm_test_framebuffer_init
[10:33:46] [PASSED] drm_test_framebuffer_init_bad_format
[10:33:46] [PASSED] drm_test_framebuffer_init_dev_mismatch
[10:33:46] [PASSED] drm_test_framebuffer_lookup
[10:33:46] [PASSED] drm_test_framebuffer_lookup_inexistent
[10:33:46] [PASSED] drm_test_framebuffer_modifiers_not_supported
[10:33:46] ================= [PASSED] drm_framebuffer =================
[10:33:46] ================ drm_gem_shmem (8 subtests) ================
[10:33:46] [PASSED] drm_gem_shmem_test_obj_create
[10:33:46] [PASSED] drm_gem_shmem_test_obj_create_private
[10:33:46] [PASSED] drm_gem_shmem_test_pin_pages
[10:33:46] [PASSED] drm_gem_shmem_test_vmap
[10:33:46] [PASSED] drm_gem_shmem_test_get_sg_table
[10:33:46] [PASSED] drm_gem_shmem_test_get_pages_sgt
[10:33:46] [PASSED] drm_gem_shmem_test_madvise
[10:33:46] [PASSED] drm_gem_shmem_test_purge
[10:33:46] ================== [PASSED] drm_gem_shmem ==================
[10:33:46] === drm_atomic_helper_connector_hdmi_check (27 subtests) ===
[10:33:46] [PASSED] drm_test_check_broadcast_rgb_auto_cea_mode
[10:33:46] [PASSED] drm_test_check_broadcast_rgb_auto_cea_mode_vic_1
[10:33:46] [PASSED] drm_test_check_broadcast_rgb_full_cea_mode
[10:33:46] [PASSED] drm_test_check_broadcast_rgb_full_cea_mode_vic_1
[10:33:46] [PASSED] drm_test_check_broadcast_rgb_limited_cea_mode
[10:33:46] [PASSED] drm_test_check_broadcast_rgb_limited_cea_mode_vic_1
[10:33:46] ====== drm_test_check_broadcast_rgb_cea_mode_yuv420  =======
[10:33:46] [PASSED] Automatic
[10:33:46] [PASSED] Full
[10:33:46] [PASSED] Limited 16:235
[10:33:46] == [PASSED] drm_test_check_broadcast_rgb_cea_mode_yuv420 ===
[10:33:46] [PASSED] drm_test_check_broadcast_rgb_crtc_mode_changed
[10:33:46] [PASSED] drm_test_check_broadcast_rgb_crtc_mode_not_changed
[10:33:46] [PASSED] drm_test_check_disable_connector
[10:33:46] [PASSED] drm_test_check_hdmi_funcs_reject_rate
[10:33:46] [PASSED] drm_test_check_max_tmds_rate_bpc_fallback_rgb
[10:33:46] [PASSED] drm_test_check_max_tmds_rate_bpc_fallback_yuv420
[10:33:46] [PASSED] drm_test_check_max_tmds_rate_bpc_fallback_ignore_yuv422
[10:33:46] [PASSED] drm_test_check_max_tmds_rate_bpc_fallback_ignore_yuv420
[10:33:46] [PASSED] drm_test_check_driver_unsupported_fallback_yuv420
[10:33:46] [PASSED] drm_test_check_output_bpc_crtc_mode_changed
[10:33:46] [PASSED] drm_test_check_output_bpc_crtc_mode_not_changed
[10:33:46] [PASSED] drm_test_check_output_bpc_dvi
[10:33:46] [PASSED] drm_test_check_output_bpc_format_vic_1
[10:33:46] [PASSED] drm_test_check_output_bpc_format_display_8bpc_only
[10:33:46] [PASSED] drm_test_check_output_bpc_format_display_rgb_only
[10:33:46] [PASSED] drm_test_check_output_bpc_format_driver_8bpc_only
[10:33:46] [PASSED] drm_test_check_output_bpc_format_driver_rgb_only
[10:33:46] [PASSED] drm_test_check_tmds_char_rate_rgb_8bpc
[10:33:46] [PASSED] drm_test_check_tmds_char_rate_rgb_10bpc
[10:33:46] [PASSED] drm_test_check_tmds_char_rate_rgb_12bpc
[10:33:46] ===== [PASSED] drm_atomic_helper_connector_hdmi_check ======
[10:33:46] === drm_atomic_helper_connector_hdmi_reset (6 subtests) ====
[10:33:46] [PASSED] drm_test_check_broadcast_rgb_value
[10:33:46] [PASSED] drm_test_check_bpc_8_value
[10:33:46] [PASSED] drm_test_check_bpc_10_value
[10:33:46] [PASSED] drm_test_check_bpc_12_value
[10:33:46] [PASSED] drm_test_check_format_value
[10:33:46] [PASSED] drm_test_check_tmds_char_value
[10:33:46] ===== [PASSED] drm_atomic_helper_connector_hdmi_reset ======
[10:33:46] = drm_atomic_helper_connector_hdmi_mode_valid (4 subtests) =
[10:33:46] [PASSED] drm_test_check_mode_valid
[10:33:46] [PASSED] drm_test_check_mode_valid_reject
[10:33:46] [PASSED] drm_test_check_mode_valid_reject_rate
[10:33:46] [PASSED] drm_test_check_mode_valid_reject_max_clock
[10:33:46] === [PASSED] drm_atomic_helper_connector_hdmi_mode_valid ===
[10:33:46] = drm_atomic_helper_connector_hdmi_infoframes (5 subtests) =
[10:33:46] [PASSED] drm_test_check_infoframes
[10:33:46] [PASSED] drm_test_check_reject_avi_infoframe
[10:33:46] [PASSED] drm_test_check_reject_hdr_infoframe_bpc_8
[10:33:46] [PASSED] drm_test_check_reject_hdr_infoframe_bpc_10
[10:33:46] [PASSED] drm_test_check_reject_audio_infoframe
[10:33:46] === [PASSED] drm_atomic_helper_connector_hdmi_infoframes ===
[10:33:46] ================= drm_managed (2 subtests) =================
[10:33:46] [PASSED] drm_test_managed_release_action
[10:33:46] [PASSED] drm_test_managed_run_action
[10:33:46] =================== [PASSED] drm_managed ===================
[10:33:46] =================== drm_mm (6 subtests) ====================
[10:33:46] [PASSED] drm_test_mm_init
[10:33:46] [PASSED] drm_test_mm_debug
[10:33:46] [PASSED] drm_test_mm_align32
[10:33:46] [PASSED] drm_test_mm_align64
[10:33:46] [PASSED] drm_test_mm_lowest
[10:33:46] [PASSED] drm_test_mm_highest
[10:33:46] ===================== [PASSED] drm_mm ======================
[10:33:46] ============= drm_modes_analog_tv (5 subtests) =============
[10:33:46] [PASSED] drm_test_modes_analog_tv_mono_576i
[10:33:46] [PASSED] drm_test_modes_analog_tv_ntsc_480i
[10:33:46] [PASSED] drm_test_modes_analog_tv_ntsc_480i_inlined
[10:33:46] [PASSED] drm_test_modes_analog_tv_pal_576i
[10:33:46] [PASSED] drm_test_modes_analog_tv_pal_576i_inlined
[10:33:46] =============== [PASSED] drm_modes_analog_tv ===============
[10:33:46] ============== drm_plane_helper (2 subtests) ===============
[10:33:46] =============== drm_test_check_plane_state  ================
[10:33:46] [PASSED] clipping_simple
[10:33:46] [PASSED] clipping_rotate_reflect
[10:33:46] [PASSED] positioning_simple
[10:33:46] [PASSED] upscaling
[10:33:46] [PASSED] downscaling
[10:33:46] [PASSED] rounding1
[10:33:46] [PASSED] rounding2
[10:33:46] [PASSED] rounding3
[10:33:46] [PASSED] rounding4
[10:33:46] =========== [PASSED] drm_test_check_plane_state ============
[10:33:46] =========== drm_test_check_invalid_plane_state  ============
[10:33:46] [PASSED] positioning_invalid
[10:33:46] [PASSED] upscaling_invalid
[10:33:46] [PASSED] downscaling_invalid
[10:33:46] ======= [PASSED] drm_test_check_invalid_plane_state ========
[10:33:46] ================ [PASSED] drm_plane_helper =================
[10:33:46] ====== drm_connector_helper_tv_get_modes (1 subtest) =======
[10:33:46] ====== drm_test_connector_helper_tv_get_modes_check  =======
[10:33:46] [PASSED] None
[10:33:46] [PASSED] PAL
[10:33:46] [PASSED] NTSC
[10:33:46] [PASSED] Both, NTSC Default
[10:33:46] [PASSED] Both, PAL Default
[10:33:46] [PASSED] Both, NTSC Default, with PAL on command-line
[10:33:46] [PASSED] Both, PAL Default, with NTSC on command-line
[10:33:46] == [PASSED] drm_test_connector_helper_tv_get_modes_check ===
[10:33:46] ======== [PASSED] drm_connector_helper_tv_get_modes ========
[10:33:46] ================== drm_rect (9 subtests) ===================
[10:33:46] [PASSED] drm_test_rect_clip_scaled_div_by_zero
[10:33:46] [PASSED] drm_test_rect_clip_scaled_not_clipped
[10:33:46] [PASSED] drm_test_rect_clip_scaled_clipped
[10:33:46] [PASSED] drm_test_rect_clip_scaled_signed_vs_unsigned
[10:33:46] ================= drm_test_rect_intersect  =================
[10:33:46] [PASSED] top-left x bottom-right: 2x2+1+1 x 2x2+0+0
[10:33:46] [PASSED] top-right x bottom-left: 2x2+0+0 x 2x2+1-1
[10:33:46] [PASSED] bottom-left x top-right: 2x2+1-1 x 2x2+0+0
[10:33:46] [PASSED] bottom-right x top-left: 2x2+0+0 x 2x2+1+1
[10:33:46] [PASSED] right x left: 2x1+0+0 x 3x1+1+0
[10:33:46] [PASSED] left x right: 3x1+1+0 x 2x1+0+0
[10:33:46] [PASSED] up x bottom: 1x2+0+0 x 1x3+0-1
[10:33:46] [PASSED] bottom x up: 1x3+0-1 x 1x2+0+0
[10:33:46] [PASSED] touching corner: 1x1+0+0 x 2x2+1+1
[10:33:46] [PASSED] touching side: 1x1+0+0 x 1x1+1+0
[10:33:46] [PASSED] equal rects: 2x2+0+0 x 2x2+0+0
[10:33:46] [PASSED] inside another: 2x2+0+0 x 1x1+1+1
[10:33:46] [PASSED] far away: 1x1+0+0 x 1x1+3+6
[10:33:46] [PASSED] points intersecting: 0x0+5+10 x 0x0+5+10
[10:33:46] [PASSED] points not intersecting: 0x0+0+0 x 0x0+5+10
[10:33:46] ============= [PASSED] drm_test_rect_intersect =============
[10:33:46] ================ drm_test_rect_calc_hscale  ================
[10:33:46] [PASSED] normal use
[10:33:46] [PASSED] out of max range
[10:33:46] [PASSED] out of min range
[10:33:46] [PASSED] zero dst
[10:33:46] [PASSED] negative src
[10:33:46] [PASSED] negative dst
[10:33:46] ============ [PASSED] drm_test_rect_calc_hscale ============
[10:33:46] ================ drm_test_rect_calc_vscale  ================
[10:33:46] [PASSED] normal use
[10:33:46] [PASSED] out of max range
[10:33:46] [PASSED] out of min range
[10:33:46] [PASSED] zero dst
[10:33:46] [PASSED] negative src
[10:33:46] [PASSED] negative dst
[10:33:46] ============ [PASSED] drm_test_rect_calc_vscale ============
[10:33:46] ================== drm_test_rect_rotate  ===================
[10:33:46] [PASSED] reflect-x
[10:33:46] [PASSED] reflect-y
[10:33:46] [PASSED] rotate-0
[10:33:46] [PASSED] rotate-90
[10:33:46] [PASSED] rotate-180
[10:33:46] [PASSED] rotate-270
[10:33:46] ============== [PASSED] drm_test_rect_rotate ===============
[10:33:46] ================ drm_test_rect_rotate_inv  =================
[10:33:46] [PASSED] reflect-x
[10:33:46] [PASSED] reflect-y
[10:33:46] [PASSED] rotate-0
[10:33:46] [PASSED] rotate-90
[10:33:46] [PASSED] rotate-180
[10:33:46] [PASSED] rotate-270
[10:33:46] ============ [PASSED] drm_test_rect_rotate_inv =============
[10:33:46] ==================== [PASSED] drm_rect =====================
[10:33:46] ============ drm_sysfb_modeset_test (1 subtest) ============
[10:33:46] ============ drm_test_sysfb_build_fourcc_list  =============
[10:33:46] [PASSED] no native formats
[10:33:46] [PASSED] XRGB8888 as native format
[10:33:46] [PASSED] remove duplicates
[10:33:46] [PASSED] convert alpha formats
[10:33:46] [PASSED] random formats
[10:33:46] ======== [PASSED] drm_test_sysfb_build_fourcc_list =========
[10:33:46] ============= [PASSED] drm_sysfb_modeset_test ==============
[10:33:46] ================== drm_fixp (2 subtests) ===================
[10:33:46] [PASSED] drm_test_int2fixp
[10:33:46] [PASSED] drm_test_sm2fixp
[10:33:46] ==================== [PASSED] drm_fixp =====================
[10:33:46] ============================================================
[10:33:46] Testing complete. Ran 621 tests: passed: 621
[10:33:46] Elapsed time: 25.962s total, 1.756s configuring, 24.041s building, 0.139s running

+ /kernel/tools/testing/kunit/kunit.py run --kunitconfig /kernel/drivers/gpu/drm/ttm/tests/.kunitconfig [10:33:46] Configuring KUnit Kernel ...
Regenerating .config ...
Populating config with:

$ make ARCH=um O=.kunit olddefconfig
[10:33:48] Building KUnit Kernel ...
Populating config with:
$ make ARCH=um O=.kunit olddefconfig
Building with:
$ make all compile_commands.json scripts_gdb ARCH=um O=.kunit --jobs=48
[10:33:57] Starting KUnit Kernel (1/1)...
[10:33:57] ============================================================
Running tests with:
$ .kunit/linux kunit.enable=1 mem=1G console=tty kunit_shutdown=halt
[10:33:57] ================= ttm_device (5 subtests) ==================
[10:33:57] [PASSED] ttm_device_init_basic
[10:33:57] [PASSED] ttm_device_init_multiple
[10:33:57] [PASSED] ttm_device_fini_basic
[10:33:57] [PASSED] ttm_device_init_no_vma_man
[10:33:57] ================== ttm_device_init_pools  ==================
[10:33:57] [PASSED] No DMA allocations, no DMA32 required
[10:33:57] [PASSED] DMA allocations, DMA32 required
[10:33:57] [PASSED] No DMA allocations, DMA32 required
[10:33:57] [PASSED] DMA allocations, no DMA32 required
[10:33:57] ============== [PASSED] ttm_device_init_pools ==============
[10:33:57] =================== [PASSED] ttm_device ====================
[10:33:57] ================== ttm_pool (8 subtests) ===================
[10:33:57] ================== ttm_pool_alloc_basic  ===================
[10:33:57] [PASSED] One page
[10:33:57] [PASSED] More than one page
[10:33:57] [PASSED] Above the allocation limit
[10:33:57] [PASSED] One page, with coherent DMA mappings enabled
[10:33:57] [PASSED] Above the allocation limit, with coherent DMA mappings enabled
[10:33:57] ============== [PASSED] ttm_pool_alloc_basic ===============
[10:33:57] ============== ttm_pool_alloc_basic_dma_addr  ==============
[10:33:57] [PASSED] One page
[10:33:57] [PASSED] More than one page
[10:33:57] [PASSED] Above the allocation limit
[10:33:57] [PASSED] One page, with coherent DMA mappings enabled
[10:33:57] [PASSED] Above the allocation limit, with coherent DMA mappings enabled
[10:33:57] ========== [PASSED] ttm_pool_alloc_basic_dma_addr ==========
[10:33:57] [PASSED] ttm_pool_alloc_order_caching_match
[10:33:57] [PASSED] ttm_pool_alloc_caching_mismatch
[10:33:57] [PASSED] ttm_pool_alloc_order_mismatch
[10:33:57] [PASSED] ttm_pool_free_dma_alloc
[10:33:57] [PASSED] ttm_pool_free_no_dma_alloc
[10:33:57] [PASSED] ttm_pool_fini_basic
[10:33:57] ==================== [PASSED] ttm_pool =====================
[10:33:57] ================ ttm_resource (8 subtests) =================
[10:33:57] ================= ttm_resource_init_basic  =================
[10:33:57] [PASSED] Init resource in TTM_PL_SYSTEM
[10:33:57] [PASSED] Init resource in TTM_PL_VRAM
[10:33:57] [PASSED] Init resource in a private placement
[10:33:57] [PASSED] Init resource in TTM_PL_SYSTEM, set placement flags
[10:33:57] ============= [PASSED] ttm_resource_init_basic =============
[10:33:57] [PASSED] ttm_resource_init_pinned
[10:33:57] [PASSED] ttm_resource_fini_basic
[10:33:57] [PASSED] ttm_resource_manager_init_basic
[10:33:57] [PASSED] ttm_resource_manager_usage_basic
[10:33:57] [PASSED] ttm_resource_manager_set_used_basic
[10:33:57] [PASSED] ttm_sys_man_alloc_basic
[10:33:57] [PASSED] ttm_sys_man_free_basic
[10:33:57] ================== [PASSED] ttm_resource ===================
[10:33:57] =================== ttm_tt (15 subtests) ===================
[10:33:57] ==================== ttm_tt_init_basic  ====================
[10:33:57] [PASSED] Page-aligned size
[10:33:57] [PASSED] Extra pages requested
[10:33:57] ================ [PASSED] ttm_tt_init_basic ================
[10:33:57] [PASSED] ttm_tt_init_misaligned
[10:33:57] [PASSED] ttm_tt_fini_basic
[10:33:57] [PASSED] ttm_tt_fini_sg
[10:33:57] [PASSED] ttm_tt_fini_shmem
[10:33:57] [PASSED] ttm_tt_create_basic
[10:33:57] [PASSED] ttm_tt_create_invalid_bo_type
[10:33:57] [PASSED] ttm_tt_create_ttm_exists
[10:33:57] [PASSED] ttm_tt_create_failed
[10:33:57] [PASSED] ttm_tt_destroy_basic
[10:33:57] [PASSED] ttm_tt_populate_null_ttm
[10:33:57] [PASSED] ttm_tt_populate_populated_ttm
[10:33:57] [PASSED] ttm_tt_unpopulate_basic
[10:33:57] [PASSED] ttm_tt_unpopulate_empty_ttm
[10:33:57] [PASSED] ttm_tt_swapin_basic
[10:33:57] ===================== [PASSED] ttm_tt ======================
[10:33:57] =================== ttm_bo (14 subtests) ===================
[10:33:57] =========== ttm_bo_reserve_optimistic_no_ticket  ===========
[10:33:57] [PASSED] Cannot be interrupted and sleeps
[10:33:57] [PASSED] Cannot be interrupted, locks straight away
[10:33:57] [PASSED] Can be interrupted, sleeps
[10:33:57] ======= [PASSED] ttm_bo_reserve_optimistic_no_ticket =======
[10:33:57] [PASSED] ttm_bo_reserve_locked_no_sleep
[10:33:57] [PASSED] ttm_bo_reserve_no_wait_ticket
[10:33:57] [PASSED] ttm_bo_reserve_double_resv
[10:33:57] [PASSED] ttm_bo_reserve_interrupted
[10:33:57] [PASSED] ttm_bo_reserve_deadlock
[10:33:57] [PASSED] ttm_bo_unreserve_basic
[10:33:57] [PASSED] ttm_bo_unreserve_pinned
[10:33:57] [PASSED] ttm_bo_unreserve_bulk
[10:33:57] [PASSED] ttm_bo_fini_basic
[10:33:57] [PASSED] ttm_bo_fini_shared_resv
[10:33:57] [PASSED] ttm_bo_pin_basic
[10:33:57] [PASSED] ttm_bo_pin_unpin_resource
[10:33:57] [PASSED] ttm_bo_multiple_pin_one_unpin
[10:33:57] ===================== [PASSED] ttm_bo ======================
[10:33:57] ============== ttm_bo_validate (22 subtests) ===============
[10:33:57] ============== ttm_bo_init_reserved_sys_man  ===============
[10:33:57] [PASSED] Buffer object for userspace
[10:33:57] [PASSED] Kernel buffer object
[10:33:57] [PASSED] Shared buffer object
[10:33:57] ========== [PASSED] ttm_bo_init_reserved_sys_man ===========
[10:33:57] ============== ttm_bo_init_reserved_mock_man  ==============
[10:33:57] [PASSED] Buffer object for userspace
[10:33:57] [PASSED] Kernel buffer object
[10:33:57] [PASSED] Shared buffer object
[10:33:57] ========== [PASSED] ttm_bo_init_reserved_mock_man ==========
[10:33:57] [PASSED] ttm_bo_init_reserved_resv
[10:33:57] ================== ttm_bo_validate_basic  ==================
[10:33:57] [PASSED] Buffer object for userspace
[10:33:57] [PASSED] Kernel buffer object
[10:33:57] [PASSED] Shared buffer object
[10:33:57] ============== [PASSED] ttm_bo_validate_basic ==============
[10:33:57] [PASSED] ttm_bo_validate_invalid_placement
[10:33:57] ============= ttm_bo_validate_same_placement  ==============
[10:33:57] [PASSED] System manager
[10:33:57] [PASSED] VRAM manager
[10:33:57] ========= [PASSED] ttm_bo_validate_same_placement ==========
[10:33:57] [PASSED] ttm_bo_validate_failed_alloc
[10:33:57] [PASSED] ttm_bo_validate_pinned
[10:33:57] [PASSED] ttm_bo_validate_busy_placement
[10:33:57] ================ ttm_bo_validate_multihop  =================
[10:33:57] [PASSED] Buffer object for userspace
[10:33:57] [PASSED] Kernel buffer object
[10:33:57] [PASSED] Shared buffer object
[10:33:57] ============ [PASSED] ttm_bo_validate_multihop =============
[10:33:57] ========== ttm_bo_validate_no_placement_signaled  ==========
[10:33:57] [PASSED] Buffer object in system domain, no page vector
[10:33:57] [PASSED] Buffer object in system domain with an existing page vector
[10:33:57] ====== [PASSED] ttm_bo_validate_no_placement_signaled ======
[10:33:57] ======== ttm_bo_validate_no_placement_not_signaled  ========
[10:33:57] [PASSED] Buffer object for userspace
[10:33:57] [PASSED] Kernel buffer object
[10:33:57] [PASSED] Shared buffer object
[10:33:57] ==== [PASSED] ttm_bo_validate_no_placement_not_signaled ====
[10:33:57] [PASSED] ttm_bo_validate_move_fence_signaled
[10:33:57] ========= ttm_bo_validate_move_fence_not_signaled  =========
[10:33:57] [PASSED] Waits for GPU
[10:33:57] [PASSED] Tries to lock straight away
[10:33:57] ===== [PASSED] ttm_bo_validate_move_fence_not_signaled =====
[10:33:57] [PASSED] ttm_bo_validate_swapout
[10:33:57] [PASSED] ttm_bo_validate_happy_evict
[10:33:57] [PASSED] ttm_bo_validate_all_pinned_evict
[10:33:57] [PASSED] ttm_bo_validate_allowed_only_evict
[10:33:57] [PASSED] ttm_bo_validate_deleted_evict
[10:33:57] [PASSED] ttm_bo_validate_busy_domain_evict
[10:33:57] [PASSED] ttm_bo_validate_evict_gutting
[10:33:57] [PASSED] ttm_bo_validate_recrusive_evict
[10:33:57] ================= [PASSED] ttm_bo_validate =================
[10:33:57] ============================================================
[10:33:57] Testing complete. Ran 102 tests: passed: 102
[10:33:57] Elapsed time: 11.432s total, 1.654s configuring, 9.563s building, 0.180s running

+ cleanup
++ stat -c %u:%g /kernel
+ chown -R 1003:1003 /kernel

#15 lore raw

✓ Xe.CI.BAT: success for drm/xe/svm: Add MMU notifier-based madvise autoreset on munmap (rev3)

In reply to Arvind Yadav
== Series Details ==

== Series Details ==

Series: drm/xe/svm: Add MMU notifier-based madvise autoreset on munmap (rev3)
URL : https://patchwork.freedesktop.org/series/161815/
State : success

== Summary ==

CI Bug Log - changes from xe-5201-70c7b363fab8198549bb9b82134bda0f6bda9452_BAT -> xe-pw-161815v3_BAT
====================================================

Summary
-------

**SUCCESS**

No regressions found.

Participating hosts (13 -> 13)
------------------------------

No changes in participating hosts

Changes
-------

No changes found

Build changes
-------------

* Linux: xe-5201-70c7b363fab8198549bb9b82134bda0f6bda9452 -> xe-pw-161815v3

IGT_8948: 8948
xe-5201-70c7b363fab8198549bb9b82134bda0f6bda9452: 70c7b363fab8198549bb9b82134bda0f6bda9452
xe-pw-161815v3: 161815v3

== Logs ==

For more details see: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-161815v3/index.html

#16 lore raw

✗ Xe.CI.FULL: failure for drm/xe/svm: Add MMU notifier-based madvise autoreset on munmap (rev3)

In reply to Arvind Yadav
== Series Details ==

== Series Details ==

Series: drm/xe/svm: Add MMU notifier-based madvise autoreset on munmap (rev3)
URL : https://patchwork.freedesktop.org/series/161815/
State : failure

== Summary ==

CI Bug Log - changes from xe-5201-70c7b363fab8198549bb9b82134bda0f6bda9452_FULL -> xe-pw-161815v3_FULL
====================================================

Summary
-------

**FAILURE**

Serious unknown changes coming with xe-pw-161815v3_FULL absolutely need to be verified manually.

If you think the reported changes have nothing to do with the changes
introduced in xe-pw-161815v3_FULL, please notify your bug team ([email protected]) to allow them to document this new failure mode, which will reduce false positives in CI.

Participating hosts (2 -> 2)
------------------------------

No changes in participating hosts

Possible new issues
-------------------

Here are the unknown changes that may have been introduced in xe-pw-161815v3_FULL:

### IGT changes ###

#### Possible regressions ####

* igt@sriov_basic@enable-vfs-bind-unbind-each-numvfs-all:
- shard-bmg: [PASS][1] -> [FAIL][2]
[1]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-5201-70c7b363fab8198549bb9b82134bda0f6bda9452/shard-bmg-6/igt@[email protected]
[2]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-161815v3/shard-bmg-6/igt@[email protected]

Known issues
------------

Here are the changes found in xe-pw-161815v3_FULL that come from known issues:

### IGT changes ###

#### Issues hit ####

* igt@kms_hdr@invalid-metadata-sizes@pipe-a-hdmi-a-3-xrgb16161616f:
- shard-bmg: [PASS][3] -> [SKIP][4] ([Intel XE#7915]) +1 other test skip
[3]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-5201-70c7b363fab8198549bb9b82134bda0f6bda9452/shard-bmg-9/igt@kms_hdr@[email protected]
[4]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-161815v3/shard-bmg-1/igt@kms_hdr@[email protected]

* igt@kms_vrr@seamless-rr-switch-virtual@pipe-a-edp-1:
- shard-lnl: [PASS][5] -> [FAIL][6] ([Intel XE#2142]) +1 other test fail
[5]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-5201-70c7b363fab8198549bb9b82134bda0f6bda9452/shard-lnl-2/igt@kms_vrr@[email protected]
[6]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-161815v3/shard-lnl-8/igt@kms_vrr@[email protected]

* igt@xe_survivability@runtime-survivability:
- shard-bmg: [PASS][7] -> [DMESG-WARN][8] ([Intel XE#6627] / [Intel XE#7419])
[7]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-5201-70c7b363fab8198549bb9b82134bda0f6bda9452/shard-bmg-4/igt@[email protected]
[8]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-161815v3/shard-bmg-8/igt@[email protected]

#### Possible fixes ####

* igt@kms_cursor_legacy@flip-vs-cursor-atomic:
- shard-bmg: [FAIL][9] ([Intel XE#7571]) -> [PASS][10]
[9]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-5201-70c7b363fab8198549bb9b82134bda0f6bda9452/shard-bmg-9/igt@[email protected]
[10]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-161815v3/shard-bmg-1/igt@[email protected]

* igt@kms_hdr@static-swap@pipe-a-hdmi-a-3-xrgb2101010:
- shard-bmg: [SKIP][11] ([Intel XE#7915]) -> [PASS][12] +1 other test pass
[11]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-5201-70c7b363fab8198549bb9b82134bda0f6bda9452/shard-bmg-8/igt@kms_hdr@[email protected]
[12]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-161815v3/shard-bmg-9/igt@kms_hdr@[email protected]

* igt@kms_pm_backlight@fade-with-suspend:
- shard-lnl: [DMESG-WARN][13] -> [PASS][14] +1 other test pass
[13]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-5201-70c7b363fab8198549bb9b82134bda0f6bda9452/shard-lnl-8/igt@[email protected]
[14]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-161815v3/shard-lnl-7/igt@[email protected]

#### Warnings ####

* igt@kms_hdr@brightness-with-hdr:
- shard-bmg: [SKIP][15] ([Intel XE#3544] / [Intel XE#7915] / [Intel XE#7916]) -> [SKIP][16] ([Intel XE#3544] / [Intel XE#7916])
[15]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-5201-70c7b363fab8198549bb9b82134bda0f6bda9452/shard-bmg-1/igt@[email protected]
[16]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-161815v3/shard-bmg-5/igt@[email protected]

* igt@kms_hdr@brightness-with-hdr@pipe-a-hdmi-a-3-xrgb16161616f:
- shard-bmg: [SKIP][17] ([Intel XE#7915]) -> [SKIP][18] ([Intel XE#7916]) +1 other test skip
[17]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-5201-70c7b363fab8198549bb9b82134bda0f6bda9452/shard-bmg-1/igt@kms_hdr@[email protected]
[18]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-161815v3/shard-bmg-5/igt@kms_hdr@[email protected]

* igt@kms_tiled_display@basic-test-pattern-with-chamelium:
- shard-bmg: [SKIP][19] ([Intel XE#2509] / [Intel XE#7437]) -> [SKIP][20] ([Intel XE#2426] / [Intel XE#5848])
[19]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-5201-70c7b363fab8198549bb9b82134bda0f6bda9452/shard-bmg-5/igt@[email protected]
[20]: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-161815v3/shard-bmg-7/igt@[email protected]

[Intel XE#2142]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2142
[Intel XE#2426]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2426
[Intel XE#2509]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/2509
[Intel XE#3544]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/3544
[Intel XE#5848]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/5848
[Intel XE#6627]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/6627
[Intel XE#7419]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7419
[Intel XE#7437]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7437
[Intel XE#7571]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7571
[Intel XE#7915]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7915
[Intel XE#7916]: https://gitlab.freedesktop.org/drm/xe/kernel/issues/7916

Build changes
-------------

* Linux: xe-5201-70c7b363fab8198549bb9b82134bda0f6bda9452 -> xe-pw-161815v3

IGT_8948: 8948
xe-5201-70c7b363fab8198549bb9b82134bda0f6bda9452: 70c7b363fab8198549bb9b82134bda0f6bda9452
xe-pw-161815v3: 161815v3

== Logs ==

For more details see: https://intel-gfx-ci.01.org/tree/intel-xe/xe-pw-161815v3/index.html