Skip to contentSkip to conversation

drm/i915: Move LMEM (VRAM) management over to TTM

PATCH Superseded Thomas Hellström 7 participants 62 messages May 18, 2021 – May 19, 2021

Subsystems (8)

v2 of 11 Superseded Latest is v11 →

lore
Share & tools
lore

Thomas Hellström

author reviewer
#0 lore raw

v2 cover drm/i915: Move LMEM (VRAM) management over to TTM

This is an initial patch series to move discrete memory management over to TTM. It will be followed up shortly with adding more functionality.

The buddy allocator is temporarily removed along with its selftests and It is replaced with the TTM range manager and some selftests are adjusted to account for introduced fragmentation. Work is ongoing to reintroduce the buddy allocator as a TTM resource manager.

A new memcpy ttm move is introduced that uses kmap_local() functionality rather than vmap(). Among other things stated in the patch commit message it helps us deal with page-pased LMEM memory. It is generic enough to replace the ttm memcpy move with some additional work if so desired. On x86 it also enables prefetching reads from write-combined memory.

Finally the old i915 gem object LMEM backend is replaced with a i915 gem object TTM backend and some additional i915 gem object ops are introduced to support the added functionality.
Currently it is used only to support management and eviction of the LMEM region, but work is underway to extend the support to system memory. In this way we use TTM the way it was originally intended, having the GPU binding taken care of by driver code.

Intention is to follow up with
- System memory support
- Pipelined accelerated moves / migration
- Re-added buddy allocator in the TTM framework

v2:
- Add patches to move pagefaulting over to TTM
- Break out TTM changes to separate patches
- Address various review comments as detailed in the affected patches

Cc Christian König <[email protected]>

Maarten Lankhorst (4):
drm/i915: Disable mmap ioctl for gen12+
drm/ttm: Add BO and offset arguments for vm_access and vm_fault ttm
handlers.
drm/i915: Use ttm mmap handling for ttm bo's. drm/i915/ttm: Add io sgt caching to i915_ttm_io_mem_pfn

Thomas Hellström (11):
drm/i915: Untangle the vma pages_mutex
drm/i915: Don't free shared locks while shared
drm/i915: Fix i915_sg_page_sizes to record dma segments rather than
physical pages
drm/ttm: Export functions to initialize and finalize the ttm range
manager standalone
drm/i915/ttm Initialize the ttm device and memory managers drm/i915/ttm: Embed a ttm buffer object in the i915 gem object drm/ttm: Export ttm_bo_tt_destroy()
drm/i915/ttm Add a generic TTM memcpy move for page-based iomem drm/ttm, drm/amdgpu: Allow the driver some control over swapping drm/i915/ttm: Introduce a TTM i915 gem object backend drm/i915/lmem: Verify checks for lmem residency

 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c       |   8 +-
 drivers/gpu/drm/i915/Kconfig                  |   1 +
 drivers/gpu/drm/i915/Makefile                 |   4 +-
 drivers/gpu/drm/i915/display/intel_display.c  |   2 +-
 drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c    |   2 +-
 drivers/gpu/drm/i915/gem/i915_gem_lmem.c      |  71 +-
 drivers/gpu/drm/i915/gem/i915_gem_lmem.h      |   5 -
 drivers/gpu/drm/i915/gem/i915_gem_mman.c      |  24 +-
 drivers/gpu/drm/i915/gem/i915_gem_mman.h      |   2 +
 drivers/gpu/drm/i915/gem/i915_gem_object.c    | 149 +++-
 drivers/gpu/drm/i915/gem/i915_gem_object.h    |  19 +-
 .../gpu/drm/i915/gem/i915_gem_object_types.h  |  39 +-
 drivers/gpu/drm/i915/gem/i915_gem_pages.c     |   6 +-
 drivers/gpu/drm/i915/gem/i915_gem_phys.c      |   2 +-
 drivers/gpu/drm/i915/gem/i915_gem_region.c    | 126 +--
 drivers/gpu/drm/i915/gem/i915_gem_region.h    |   4 -
 drivers/gpu/drm/i915/gem/i915_gem_shmem.c     |   4 +-
 drivers/gpu/drm/i915/gem/i915_gem_stolen.c    |  10 +-
 drivers/gpu/drm/i915/gem/i915_gem_stolen.h    |   9 +-
 drivers/gpu/drm/i915/gem/i915_gem_ttm.c       | 626 ++++++++++++++
 drivers/gpu/drm/i915/gem/i915_gem_ttm.h       |  48 ++
 .../gpu/drm/i915/gem/i915_gem_ttm_bo_util.c   | 194 +++++
 .../gpu/drm/i915/gem/i915_gem_ttm_bo_util.h   | 107 +++
 drivers/gpu/drm/i915/gem/i915_gem_userptr.c   |   2 +-
 drivers/gpu/drm/i915/gt/intel_ggtt.c          |  19 +-
 drivers/gpu/drm/i915/gt/intel_gt.c            |   2 -
 drivers/gpu/drm/i915/gt/intel_gtt.c           |  45 +-
 drivers/gpu/drm/i915/gt/intel_gtt.h           |  30 +-
 drivers/gpu/drm/i915/gt/intel_ppgtt.c         |   2 +-
 drivers/gpu/drm/i915/gt/intel_region_lmem.c   |  30 +-
 drivers/gpu/drm/i915/i915_buddy.c             | 435 ----------
 drivers/gpu/drm/i915/i915_buddy.h             | 131 ---
 drivers/gpu/drm/i915/i915_drv.c               |  13 +
 drivers/gpu/drm/i915/i915_drv.h               |   7 +-
 drivers/gpu/drm/i915/i915_gem.c               |   6 +-
 drivers/gpu/drm/i915/i915_globals.c           |   1 -
 drivers/gpu/drm/i915/i915_globals.h           |   1 -
 drivers/gpu/drm/i915/i915_scatterlist.c       |  70 ++
 drivers/gpu/drm/i915/i915_scatterlist.h       |  20 +-
 drivers/gpu/drm/i915/i915_vma.c               |  33 +-
 drivers/gpu/drm/i915/intel_memory_region.c    | 181 ++--
 drivers/gpu/drm/i915/intel_memory_region.h    |  45 +-
 drivers/gpu/drm/i915/intel_region_ttm.c       | 246 ++++++
 drivers/gpu/drm/i915/intel_region_ttm.h       |  32 +
 drivers/gpu/drm/i915/selftests/i915_buddy.c   | 789 ------------------
 .../drm/i915/selftests/i915_mock_selftests.h  |   1 -
 .../drm/i915/selftests/intel_memory_region.c  | 133 +--
 drivers/gpu/drm/i915/selftests/mock_region.c  |  50 +-
 drivers/gpu/drm/nouveau/nouveau_ttm.c         |   4 +-
 drivers/gpu/drm/radeon/radeon_ttm.c           |   4 +-
 drivers/gpu/drm/ttm/ttm_bo.c                  |  42 +-
 drivers/gpu/drm/ttm/ttm_bo_vm.c               |  84 +-
 drivers/gpu/drm/ttm/ttm_range_manager.c       |  55 +-
 drivers/gpu/drm/ttm/ttm_tt.c                  |   4 +
 drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c    |   8 +-
 include/drm/ttm/ttm_bo_api.h                  |   9 +-
 include/drm/ttm/ttm_bo_driver.h               |  23 +
 57 files changed, 2068 insertions(+), 1951 deletions(-)
 create mode 100644 drivers/gpu/drm/i915/gem/i915_gem_ttm.c
 create mode 100644 drivers/gpu/drm/i915/gem/i915_gem_ttm.h
 create mode 100644 drivers/gpu/drm/i915/gem/i915_gem_ttm_bo_util.c
 create mode 100644 drivers/gpu/drm/i915/gem/i915_gem_ttm_bo_util.h
 delete mode 100644 drivers/gpu/drm/i915/i915_buddy.c
 delete mode 100644 drivers/gpu/drm/i915/i915_buddy.h
 create mode 100644 drivers/gpu/drm/i915/intel_region_ttm.c
 create mode 100644 drivers/gpu/drm/i915/intel_region_ttm.h
 delete mode 100644 drivers/gpu/drm/i915/selftests/i915_buddy.c
Signature · 6 lines
2.31.1

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
author reviewer
#1 lore raw

PATCH v2 1/15 drm/i915: Untangle the vma pages_mutex

From: Thomas Hellström <[email protected]>

Any sleeping dma_resv lock taken while the vma pages_mutex is held will cause a lockdep splat.
Move the i915_gem_object_pin_pages() call out of the pages_mutex critical section.

---
 drivers/gpu/drm/i915/i915_vma.c | 33 +++++++++++++++++++--------------
 1 file changed, 19 insertions(+), 14 deletions(-)
drivers/gpu/drm/i915/i915_vma.c +19 -14
index a6cd0fa62847..7b1c0f4e60d7 100644
@@ -800,32 +800,37 @@ static bool try_qad_pin(struct i915_vma *vma, unsigned int flags)
 static int vma_get_pages(struct i915_vma *vma)
 {
 	int err = 0;
+	bool pinned_pages = false;
 
 	if (atomic_add_unless(&vma->pages_count, 1, 0))
 		return 0;
 
+	if (vma->obj) {
+		err = i915_gem_object_pin_pages(vma->obj);
+		if (err)
+			return err;
+		pinned_pages = true;
+	}
+
 	/* Allocations ahoy! */
-	if (mutex_lock_interruptible(&vma->pages_mutex))
-		return -EINTR;
+	if (mutex_lock_interruptible(&vma->pages_mutex)) {
+		err = -EINTR;
+		goto unpin;
+	}
 
 	if (!atomic_read(&vma->pages_count)) {
-		if (vma->obj) {
-			err = i915_gem_object_pin_pages(vma->obj);
-			if (err)
-				goto unlock;
-		}
-
 		err = vma->ops->set_pages(vma);
-		if (err) {
-			if (vma->obj)
-				i915_gem_object_unpin_pages(vma->obj);
+		if (err)
 			goto unlock;
-		}
+		pinned_pages = false;
 	}
 	atomic_inc(&vma->pages_count);
 
 unlock:
 	mutex_unlock(&vma->pages_mutex);
+unpin:
+	if (pinned_pages)
+		__i915_gem_object_unpin_pages(vma->obj);
 
 	return err;
 }
@@ -838,10 +843,10 @@ static void __vma_put_pages(struct i915_vma *vma, unsigned int count)
 	if (atomic_sub_return(count, &vma->pages_count) == 0) {
 		vma->ops->clear_pages(vma);
 		GEM_BUG_ON(vma->pages);
-		if (vma->obj)
-			i915_gem_object_unpin_pages(vma->obj);
 	}
 	mutex_unlock(&vma->pages_mutex);
+	if (vma->obj)
+		i915_gem_object_unpin_pages(vma->obj);
 }
 
 static void vma_put_pages(struct i915_vma *vma)
Signature · 1 line
2.31.1
maintainer Reviewed-by on v2 1/15
#2 lore raw

Re: [PATCH v2 01/15] drm/i915: Untangle the vma pages_mutex

Hey,

This needs a small fix, otherwise looks good.

Op 18-05-2021 om 10:26 schreef Thomas Hellström:

Quoted 67 lines

From: Thomas Hellström <[email protected]>

Any sleeping dma_resv lock taken while the vma pages_mutex is held will cause a lockdep splat.
Move the i915_gem_object_pin_pages() call out of the pages_mutex critical section.

---
 drivers/gpu/drm/i915/i915_vma.c | 33 +++++++++++++++++++--------------
 1 file changed, 19 insertions(+), 14 deletions(-)
drivers/gpu/drm/i915/i915_vma.c +17 -12
index a6cd0fa62847..7b1c0f4e60d7 100644
@@ -800,32 +800,37 @@ static bool try_qad_pin(struct i915_vma *vma, unsigned int flags)
 static int vma_get_pages(struct i915_vma *vma)
 {
 	int err = 0;
+	bool pinned_pages = false;
 
 	if (atomic_add_unless(&vma->pages_count, 1, 0))
 		return 0;
 
+	if (vma->obj) {
+		err = i915_gem_object_pin_pages(vma->obj);
+		if (err)
+			return err;
+		pinned_pages = true;
+	}
+
 	/* Allocations ahoy! */
-	if (mutex_lock_interruptible(&vma->pages_mutex))
-		return -EINTR;
+	if (mutex_lock_interruptible(&vma->pages_mutex)) {
+		err = -EINTR;
+		goto unpin;
+	}
 
 	if (!atomic_read(&vma->pages_count)) {
-		if (vma->obj) {
-			err = i915_gem_object_pin_pages(vma->obj);
-			if (err)
-				goto unlock;
-		}
-
 		err = vma->ops->set_pages(vma);
-		if (err) {
-			if (vma->obj)
-				i915_gem_object_unpin_pages(vma->obj);
+		if (err)
 			goto unlock;
-		}
+		pinned_pages = false;
 	}
 	atomic_inc(&vma->pages_count);
 
 unlock:
 	mutex_unlock(&vma->pages_mutex);
+unpin:
+	if (pinned_pages)
+		__i915_gem_object_unpin_pages(vma->obj);
 
 	return err;
 }

drivers/gpu/drm/i915/i915_vma.c static void __vma_put_pages(struct i915_vma *vma, unsigned int count) · line 849

Quoted 9 lines
drivers/gpu/drm/i915/i915_vma.c +2 -2
@@ -838,10 +843,10 @@ static void __vma_put_pages(struct i915_vma *vma, unsigned int count)
 	if (atomic_sub_return(count, &vma->pages_count) == 0) {
 		vma->ops->clear_pages(vma);
 		GEM_BUG_ON(vma->pages);
-		if (vma->obj)
-			i915_gem_object_unpin_pages(vma->obj);
 	}
 	mutex_unlock(&vma->pages_mutex);
+	if (vma->obj)
+		i915_gem_object_unpin_pages(vma->obj);

You're unconditionally unpinning pages here, if pages_count wasn't dropped to 0, we will go negative.

With that fixed:

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

author reviewer
#3 lore raw

Re: [PATCH v2 01/15] drm/i915: Untangle the vma pages_mutex

Quoted 83 lines · Maarten Lankhorst

Hey,

This needs a small fix, otherwise looks good.

Op 18-05-2021 om 10:26 schreef Thomas Hellström:

Quoted 67 lines

From: Thomas Hellström <[email protected]>

Any sleeping dma_resv lock taken while the vma pages_mutex is held will cause a lockdep splat.
Move the i915_gem_object_pin_pages() call out of the pages_mutex critical section.

---
  drivers/gpu/drm/i915/i915_vma.c | 33 +++++++++++++++++++--------------
  1 file changed, 19 insertions(+), 14 deletions(-)
drivers/gpu/drm/i915/i915_vma.c +17 -12
index a6cd0fa62847..7b1c0f4e60d7 100644
@@ -800,32 +800,37 @@ static bool try_qad_pin(struct i915_vma *vma, unsigned int flags)
  static int vma_get_pages(struct i915_vma *vma)
  {
  	int err = 0;
+	bool pinned_pages = false;
 
  	if (atomic_add_unless(&vma->pages_count, 1, 0))
  		return 0;
 
+	if (vma->obj) {
+		err = i915_gem_object_pin_pages(vma->obj);
+		if (err)
+			return err;
+		pinned_pages = true;
+	}
+
  	/* Allocations ahoy! */
-	if (mutex_lock_interruptible(&vma->pages_mutex))
-		return -EINTR;
+	if (mutex_lock_interruptible(&vma->pages_mutex)) {
+		err = -EINTR;
+		goto unpin;
+	}
 
  	if (!atomic_read(&vma->pages_count)) {
-		if (vma->obj) {
-			err = i915_gem_object_pin_pages(vma->obj);
-			if (err)
-				goto unlock;
-		}
-
  		err = vma->ops->set_pages(vma);
-		if (err) {
-			if (vma->obj)
-				i915_gem_object_unpin_pages(vma->obj);
+		if (err)
  			goto unlock;
-		}
+		pinned_pages = false;
  	}
  	atomic_inc(&vma->pages_count);
 
  unlock:
  	mutex_unlock(&vma->pages_mutex);
+unpin:
+	if (pinned_pages)
+		__i915_gem_object_unpin_pages(vma->obj);
 
  	return err;
  }

drivers/gpu/drm/i915/i915_vma.c static void __vma_put_pages(struct i915_vma *vma, unsigned int count) · line 849

Quoted 9 lines
drivers/gpu/drm/i915/i915_vma.c +2 -2
@@ -838,10 +843,10 @@ static void __vma_put_pages(struct i915_vma *vma, unsigned int count)
  	if (atomic_sub_return(count, &vma->pages_count) == 0) {
  		vma->ops->clear_pages(vma);
  		GEM_BUG_ON(vma->pages);
-		if (vma->obj)
-			i915_gem_object_unpin_pages(vma->obj);
  	}
  	mutex_unlock(&vma->pages_mutex);
+	if (vma->obj)
+		i915_gem_object_unpin_pages(vma->obj);

You're unconditionally unpinning pages here, if pages_count wasn't dropped to 0, we will go negative.

Quoted 3 lines

With that fixed:

Ah yes, thanks. That was a leftover from an earlier version...

/Thomas

author reviewer
#4 lore raw

PATCH v2 2/15 drm/i915: Don't free shared locks while shared

In reply to Thomas Hellström

We are currently sharing the VM reservation locks across a number of gem objects with page-table memory. Since TTM will individiualize the reservation locks when freeing objects, including accessing the shared locks, make sure that the shared locks are not freed until that is done. For PPGTT we add an additional refcount, for GGTT we take additional measures to make sure objects sharing the GGTT reservation lock are freed at GGTT takedown

---
v2: Try harder to make sure objects sharing the GGTT reservation lock are freed at GGTT takedown.

---
 drivers/gpu/drm/i915/gem/i915_gem_object.c    |  3 ++
 .../gpu/drm/i915/gem/i915_gem_object_types.h  |  1 +
 drivers/gpu/drm/i915/gt/intel_ggtt.c          | 19 ++++++--
 drivers/gpu/drm/i915/gt/intel_gtt.c           | 45 +++++++++++++++----
 drivers/gpu/drm/i915/gt/intel_gtt.h           | 30 ++++++++++++-
 drivers/gpu/drm/i915/gt/intel_ppgtt.c         |  2 +-
 drivers/gpu/drm/i915/i915_drv.c               |  5 +++
 7 files changed, 92 insertions(+), 13 deletions(-)
drivers/gpu/drm/i915/gem/i915_gem_object.c +3 -0
index 28144410df86..abadf0994ad0 100644
@@ -252,6 +252,9 @@ static void __i915_gem_free_objects(struct drm_i915_private *i915,
 		if (obj->mm.n_placements > 1)
 			kfree(obj->mm.placements);
 
+		if (obj->resv_shared_from)
+			i915_vm_resv_put(obj->resv_shared_from);
+
 		/* But keep the pointer alive for RCU-protected lookups */
 		call_rcu(&obj->rcu, __i915_gem_free_object_rcu);
 		cond_resched();
drivers/gpu/drm/i915/gem/i915_gem_object_types.h +1 -0
index 0727d0c76aa0..450340a73186 100644
@@ -149,6 +149,7 @@ struct drm_i915_gem_object {
 	 * when i915_gem_ww_ctx_backoff() or i915_gem_ww_ctx_fini() are called.
 	 */
 	struct list_head obj_link;
+	struct dma_resv *resv_shared_from;
 
 	union {
 		struct rcu_head rcu;
drivers/gpu/drm/i915/gt/intel_ggtt.c +16 -3
index 35069ca5d7de..10c23a749a95 100644
@@ -746,7 +746,6 @@ static void ggtt_cleanup_hw(struct i915_ggtt *ggtt)
 
 	mutex_unlock(&ggtt->vm.mutex);
 	i915_address_space_fini(&ggtt->vm);
-	dma_resv_fini(&ggtt->vm.resv);
 
 	arch_phys_wc_del(ggtt->mtrr);
 
@@ -768,6 +767,19 @@ void i915_ggtt_driver_release(struct drm_i915_private *i915)
 	ggtt_cleanup_hw(ggtt);
 }
 
+/**
+ * i915_ggtt_driver_late_release - Cleanup of GGTT that needs to be done after
+ * all free objects have been drained.
+ * @i915: i915 device
+ */
+void i915_ggtt_driver_late_release(struct drm_i915_private *i915)
+{
+	struct i915_ggtt *ggtt = &i915->ggtt;
+
+	GEM_WARN_ON(kref_read(&ggtt->vm.resv_ref) != 1);
+	dma_resv_fini(&ggtt->vm._resv);
+}
+
 static unsigned int gen6_get_total_gtt_size(u16 snb_gmch_ctl)
 {
 	snb_gmch_ctl >>= SNB_GMCH_GGMS_SHIFT;
@@ -829,6 +841,7 @@ static int ggtt_probe_common(struct i915_ggtt *ggtt, u64 size)
 		return -ENOMEM;
 	}
 
+	kref_init(&ggtt->vm.resv_ref);
 	ret = setup_scratch_page(&ggtt->vm);
 	if (ret) {
 		drm_err(&i915->drm, "Scratch setup failed\n");
@@ -1135,7 +1148,7 @@ static int ggtt_probe_hw(struct i915_ggtt *ggtt, struct intel_gt *gt)
 	ggtt->vm.gt = gt;
 	ggtt->vm.i915 = i915;
 	ggtt->vm.dma = i915->drm.dev;
-	dma_resv_init(&ggtt->vm.resv);
+	dma_resv_init(&ggtt->vm._resv);
 
 	if (INTEL_GEN(i915) <= 5)
 		ret = i915_gmch_probe(ggtt);
@@ -1144,7 +1157,7 @@ static int ggtt_probe_hw(struct i915_ggtt *ggtt, struct intel_gt *gt)
 	else
 		ret = gen8_gmch_probe(ggtt);
 	if (ret) {
-		dma_resv_fini(&ggtt->vm.resv);
+		dma_resv_fini(&ggtt->vm._resv);
 		return ret;
 	}
 
drivers/gpu/drm/i915/gt/intel_gtt.c +37 -8

88 lines not on this page · raw message

drivers/gpu/drm/i915/gt/intel_gtt.h +29 -1
index ca00b45827b7..b1b687143144 100644
@@ -245,7 +245,9 @@ struct i915_address_space {
 	atomic_t open;
 
 	struct mutex mutex; /* protects vma and our lists */
-	struct dma_resv resv; /* reservation lock for all pd objects, and buffer pool */
+
+	struct kref resv_ref; /* kref to keep the reservation lock alive. */
+	struct dma_resv _resv; /* reservation lock for all pd objects, and buffer pool */
 #define VM_CLASS_GGTT 0
 #define VM_CLASS_PPGTT 1
 #define VM_CLASS_DPT 2
@@ -404,13 +406,38 @@ i915_vm_get(struct i915_address_space *vm)
 	return vm;
 }
 
+/**
+ * i915_vm_resv_get - Obtain a reference on the vm's reservation lock
+ * @vm: The vm whose reservation lock we want to share.
+ *
+ * Return: A pointer to the vm's reservation lock.
+ */
+static inline struct dma_resv *i915_vm_resv_get(struct i915_address_space *vm)
+{
+	kref_get(&vm->resv_ref);
+	return &vm->_resv;
+}
+
 void i915_vm_release(struct kref *kref);
 
+void i915_vm_resv_release(struct kref *kref);
+
 static inline void i915_vm_put(struct i915_address_space *vm)
 {
 	kref_put(&vm->ref, i915_vm_release);
 }
 
+/**
+ * i915_vm_resv_put - Release a reference on the vm's reservation lock
+ * @resv: Pointer to a reservation lock obtained from i915_vm_resv_get()
+ */
+static inline void i915_vm_resv_put(struct dma_resv *resv)
+{
+	struct i915_address_space *vm = container_of(resv, typeof(*vm), _resv);
+
+	kref_put(&vm->resv_ref, i915_vm_resv_release);
+}
+
 static inline struct i915_address_space *
 i915_vm_open(struct i915_address_space *vm)
 {
@@ -506,6 +533,7 @@ void i915_ggtt_enable_guc(struct i915_ggtt *ggtt);
 void i915_ggtt_disable_guc(struct i915_ggtt *ggtt);
 int i915_init_ggtt(struct drm_i915_private *i915);
 void i915_ggtt_driver_release(struct drm_i915_private *i915);
+void i915_ggtt_driver_late_release(struct drm_i915_private *i915);
 
 static inline bool i915_ggtt_has_aperture(const struct i915_ggtt *ggtt)
 {
drivers/gpu/drm/i915/gt/intel_ppgtt.c +1 -1
index 4e3d80c2295c..aee3a8929245 100644
@@ -307,7 +307,7 @@ void ppgtt_init(struct i915_ppgtt *ppgtt, struct intel_gt *gt)
 	ppgtt->vm.dma = i915->drm.dev;
 	ppgtt->vm.total = BIT_ULL(INTEL_INFO(i915)->ppgtt_size);
 
-	dma_resv_init(&ppgtt->vm.resv);
+	dma_resv_init(&ppgtt->vm._resv);
 	i915_address_space_init(&ppgtt->vm, VM_CLASS_PPGTT);
 
 	ppgtt->vm.vma_ops.bind_vma    = ppgtt_bind_vma;
drivers/gpu/drm/i915/i915_drv.c +5 -0
index 5118dc8386b2..92bccc5623a8 100644
@@ -631,6 +631,8 @@ static int i915_driver_hw_probe(struct drm_i915_private *dev_priv)
 	intel_memory_regions_driver_release(dev_priv);
 err_ggtt:
 	i915_ggtt_driver_release(dev_priv);
+	i915_gem_drain_freed_objects(dev_priv);
+	i915_ggtt_driver_late_release(dev_priv);
 err_perf:
 	i915_perf_fini(dev_priv);
 	return ret;
@@ -880,6 +882,8 @@ int i915_driver_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	i915_driver_hw_remove(i915);
 	intel_memory_regions_driver_release(i915);
 	i915_ggtt_driver_release(i915);
+	i915_gem_drain_freed_objects(i915);
+	i915_ggtt_driver_late_release(i915);
 out_cleanup_mmio:
 	i915_driver_mmio_release(i915);
 out_runtime_pm_put:
@@ -936,6 +940,7 @@ static void i915_driver_release(struct drm_device *dev)
 	intel_memory_regions_driver_release(dev_priv);
 	i915_ggtt_driver_release(dev_priv);
 	i915_gem_drain_freed_objects(dev_priv);
+	i915_ggtt_driver_late_release(dev_priv);
 
 	i915_driver_mmio_release(dev_priv);
 
Signature · 6 lines
2.31.1

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
maintainer Reviewed-by on v2 2/15
#5 lore raw

Re: [PATCH v2 02/15] drm/i915: Don't free shared locks while shared

Op 18-05-2021 om 10:26 schreef Thomas Hellström:

Quoted 41 lines

We are currently sharing the VM reservation locks across a number of gem objects with page-table memory. Since TTM will individiualize the reservation locks when freeing objects, including accessing the shared locks, make sure that the shared locks are not freed until that is done. For PPGTT we add an additional refcount, for GGTT we take additional measures to make sure objects sharing the GGTT reservation lock are freed at GGTT takedown

---
v2: Try harder to make sure objects sharing the GGTT reservation lock are freed at GGTT takedown.

---
 drivers/gpu/drm/i915/gem/i915_gem_object.c    |  3 ++
 .../gpu/drm/i915/gem/i915_gem_object_types.h  |  1 +
 drivers/gpu/drm/i915/gt/intel_ggtt.c          | 19 ++++++--
 drivers/gpu/drm/i915/gt/intel_gtt.c           | 45 +++++++++++++++----
 drivers/gpu/drm/i915/gt/intel_gtt.h           | 30 ++++++++++++-
 drivers/gpu/drm/i915/gt/intel_ppgtt.c         |  2 +-
 drivers/gpu/drm/i915/i915_drv.c               |  5 +++
 7 files changed, 92 insertions(+), 13 deletions(-)
drivers/gpu/drm/i915/gem/i915_gem_object.c +3 -0
index 28144410df86..abadf0994ad0 100644
@@ -252,6 +252,9 @@ static void __i915_gem_free_objects(struct drm_i915_private *i915,
 		if (obj->mm.n_placements > 1)
 			kfree(obj->mm.placements);
 
+		if (obj->resv_shared_from)
+			i915_vm_resv_put(obj->resv_shared_from);
+
 		/* But keep the pointer alive for RCU-protected lookups */
 		call_rcu(&obj->rcu, __i915_gem_free_object_rcu);
 		cond_resched();
drivers/gpu/drm/i915/gem/i915_gem_object_types.h +0 -0
index 0727d0c76aa0..450340a73186 100644

drivers/gpu/drm/i915/gem/i915_gem_object_types.h struct drm_i915_gem_object { · line 152

Quoted 4 lines
drivers/gpu/drm/i915/gem/i915_gem_object_types.h +1 -0
@@ -149,6 +149,7 @@ struct drm_i915_gem_object {
 	 * when i915_gem_ww_ctx_backoff() or i915_gem_ww_ctx_fini() are called.
 	 */
 	struct list_head obj_link;
+	struct dma_resv *resv_shared_from;

Since this can only be a vm object, would it make sense to make this a pointer to the vm address space, so we can call vm_resv_put on it directly?

The current pointer type and name makes it look generic, but if you try to use it with anything but an address space, it will blow up.

Otherwise looks good. I guess we cannot force all bo's to be deleted before the vm is freed. :-)

So with that fixed

~Maarten

Quoted 264 lines
drivers/gpu/drm/i915/gem/i915_gem_object_types.h +0 -0
 	union {
 		struct rcu_head rcu;
drivers/gpu/drm/i915/gt/intel_ggtt.c +16 -3
index 35069ca5d7de..10c23a749a95 100644
@@ -746,7 +746,6 @@ static void ggtt_cleanup_hw(struct i915_ggtt *ggtt)
 
 	mutex_unlock(&ggtt->vm.mutex);
 	i915_address_space_fini(&ggtt->vm);
-	dma_resv_fini(&ggtt->vm.resv);
 
 	arch_phys_wc_del(ggtt->mtrr);
 
@@ -768,6 +767,19 @@ void i915_ggtt_driver_release(struct drm_i915_private *i915)
 	ggtt_cleanup_hw(ggtt);
 }
 
+/**
+ * i915_ggtt_driver_late_release - Cleanup of GGTT that needs to be done after
+ * all free objects have been drained.
+ * @i915: i915 device
+ */
+void i915_ggtt_driver_late_release(struct drm_i915_private *i915)
+{
+	struct i915_ggtt *ggtt = &i915->ggtt;
+
+	GEM_WARN_ON(kref_read(&ggtt->vm.resv_ref) != 1);
+	dma_resv_fini(&ggtt->vm._resv);
+}
+
 static unsigned int gen6_get_total_gtt_size(u16 snb_gmch_ctl)
 {
 	snb_gmch_ctl >>= SNB_GMCH_GGMS_SHIFT;
@@ -829,6 +841,7 @@ static int ggtt_probe_common(struct i915_ggtt *ggtt, u64 size)
 		return -ENOMEM;
 	}
 
+	kref_init(&ggtt->vm.resv_ref);
 	ret = setup_scratch_page(&ggtt->vm);
 	if (ret) {
 		drm_err(&i915->drm, "Scratch setup failed\n");
@@ -1135,7 +1148,7 @@ static int ggtt_probe_hw(struct i915_ggtt *ggtt, struct intel_gt *gt)
 	ggtt->vm.gt = gt;
 	ggtt->vm.i915 = i915;
 	ggtt->vm.dma = i915->drm.dev;
-	dma_resv_init(&ggtt->vm.resv);
+	dma_resv_init(&ggtt->vm._resv);
 
 	if (INTEL_GEN(i915) <= 5)
 		ret = i915_gmch_probe(ggtt);
@@ -1144,7 +1157,7 @@ static int ggtt_probe_hw(struct i915_ggtt *ggtt, struct intel_gt *gt)
 	else
 		ret = gen8_gmch_probe(ggtt);
 	if (ret) {
-		dma_resv_fini(&ggtt->vm.resv);
+		dma_resv_fini(&ggtt->vm._resv);
 		return ret;
 	}
 
drivers/gpu/drm/i915/gt/intel_gtt.c +37 -8
index 9b98f9d9faa3..695b22b17644 100644
@@ -22,8 +22,11 @@ struct drm_i915_gem_object *alloc_pt_lmem(struct i915_address_space *vm, int sz)
 	 * object underneath, with the idea that one object_lock() will lock
 	 * them all at once.
 	 */
-	if (!IS_ERR(obj))
-		obj->base.resv = &vm->resv;
+	if (!IS_ERR(obj)) {
+		obj->base.resv = i915_vm_resv_get(vm);
+		obj->resv_shared_from = obj->base.resv;
+	}
+
 	return obj;
 }
 
@@ -40,8 +43,11 @@ struct drm_i915_gem_object *alloc_pt_dma(struct i915_address_space *vm, int sz)
 	 * object underneath, with the idea that one object_lock() will lock
 	 * them all at once.
 	 */
-	if (!IS_ERR(obj))
-		obj->base.resv = &vm->resv;
+	if (!IS_ERR(obj)) {
+		obj->base.resv = i915_vm_resv_get(vm);
+		obj->resv_shared_from = obj->base.resv;
+	}
+
 	return obj;
 }
 
@@ -102,7 +108,7 @@ void __i915_vm_close(struct i915_address_space *vm)
 int i915_vm_lock_objects(struct i915_address_space *vm,
 			 struct i915_gem_ww_ctx *ww)
 {
-	if (vm->scratch[0]->base.resv == &vm->resv) {
+	if (vm->scratch[0]->base.resv == &vm->_resv) {
 		return i915_gem_object_lock(vm->scratch[0], ww);
 	} else {
 		struct i915_ppgtt *ppgtt = i915_vm_to_ppgtt(vm);
@@ -118,6 +124,22 @@ void i915_address_space_fini(struct i915_address_space *vm)
 	mutex_destroy(&vm->mutex);
 }
 
+/**
+ * i915_vm_resv_release - Final struct i915_address_space destructor
+ * @kref: Pointer to the &i915_address_space.resv_ref member.
+ *
+ * This function is called when the last lock sharer no longer shares the
+ * &i915_address_space._resv lock.
+ */
+void i915_vm_resv_release(struct kref *kref)
+{
+	struct i915_address_space *vm =
+		container_of(kref, typeof(*vm), resv_ref);
+
+	dma_resv_fini(&vm->_resv);
+	kfree(vm);
+}
+
 static void __i915_vm_release(struct work_struct *work)
 {
 	struct i915_address_space *vm =
@@ -125,9 +147,8 @@ static void __i915_vm_release(struct work_struct *work)
 
 	vm->cleanup(vm);
 	i915_address_space_fini(vm);
-	dma_resv_fini(&vm->resv);
 
-	kfree(vm);
+	i915_vm_resv_put(&vm->_resv);
 }
 
 void i915_vm_release(struct kref *kref)
@@ -144,6 +165,14 @@ void i915_vm_release(struct kref *kref)
 void i915_address_space_init(struct i915_address_space *vm, int subclass)
 {
 	kref_init(&vm->ref);
+
+	/*
+	 * Special case for GGTT that has already done an early
+	 * kref_init here.
+	 */
+	if (!kref_read(&vm->resv_ref))
+		kref_init(&vm->resv_ref);
+
 	INIT_RCU_WORK(&vm->rcu, __i915_vm_release);
 	atomic_set(&vm->open, 1);
 
@@ -170,7 +199,7 @@ void i915_address_space_init(struct i915_address_space *vm, int subclass)
 		might_alloc(GFP_KERNEL);
 		mutex_release(&vm->mutex.dep_map, _THIS_IP_);
 	}
-	dma_resv_init(&vm->resv);
+	dma_resv_init(&vm->_resv);
 
 	GEM_BUG_ON(!vm->total);
 	drm_mm_init(&vm->mm, 0, vm->total);
drivers/gpu/drm/i915/gt/intel_gtt.h +29 -1
index ca00b45827b7..b1b687143144 100644
@@ -245,7 +245,9 @@ struct i915_address_space {
 	atomic_t open;
 
 	struct mutex mutex; /* protects vma and our lists */
-	struct dma_resv resv; /* reservation lock for all pd objects, and buffer pool */
+
+	struct kref resv_ref; /* kref to keep the reservation lock alive. */
+	struct dma_resv _resv; /* reservation lock for all pd objects, and buffer pool */
 #define VM_CLASS_GGTT 0
 #define VM_CLASS_PPGTT 1
 #define VM_CLASS_DPT 2
@@ -404,13 +406,38 @@ i915_vm_get(struct i915_address_space *vm)
 	return vm;
 }
 
+/**
+ * i915_vm_resv_get - Obtain a reference on the vm's reservation lock
+ * @vm: The vm whose reservation lock we want to share.
+ *
+ * Return: A pointer to the vm's reservation lock.
+ */
+static inline struct dma_resv *i915_vm_resv_get(struct i915_address_space *vm)
+{
+	kref_get(&vm->resv_ref);
+	return &vm->_resv;
+}
+
 void i915_vm_release(struct kref *kref);
 
+void i915_vm_resv_release(struct kref *kref);
+
 static inline void i915_vm_put(struct i915_address_space *vm)
 {
 	kref_put(&vm->ref, i915_vm_release);
 }
 
+/**
+ * i915_vm_resv_put - Release a reference on the vm's reservation lock
+ * @resv: Pointer to a reservation lock obtained from i915_vm_resv_get()
+ */
+static inline void i915_vm_resv_put(struct dma_resv *resv)
+{
+	struct i915_address_space *vm = container_of(resv, typeof(*vm), _resv);
+
+	kref_put(&vm->resv_ref, i915_vm_resv_release);
+}
+
 static inline struct i915_address_space *
 i915_vm_open(struct i915_address_space *vm)
 {
@@ -506,6 +533,7 @@ void i915_ggtt_enable_guc(struct i915_ggtt *ggtt);
 void i915_ggtt_disable_guc(struct i915_ggtt *ggtt);
 int i915_init_ggtt(struct drm_i915_private *i915);
 void i915_ggtt_driver_release(struct drm_i915_private *i915);
+void i915_ggtt_driver_late_release(struct drm_i915_private *i915);
 
 static inline bool i915_ggtt_has_aperture(const struct i915_ggtt *ggtt)
 {
drivers/gpu/drm/i915/gt/intel_ppgtt.c +1 -1
index 4e3d80c2295c..aee3a8929245 100644
@@ -307,7 +307,7 @@ void ppgtt_init(struct i915_ppgtt *ppgtt, struct intel_gt *gt)
 	ppgtt->vm.dma = i915->drm.dev;
 	ppgtt->vm.total = BIT_ULL(INTEL_INFO(i915)->ppgtt_size);
 
-	dma_resv_init(&ppgtt->vm.resv);
+	dma_resv_init(&ppgtt->vm._resv);
 	i915_address_space_init(&ppgtt->vm, VM_CLASS_PPGTT);
 
 	ppgtt->vm.vma_ops.bind_vma    = ppgtt_bind_vma;
drivers/gpu/drm/i915/i915_drv.c +5 -0
index 5118dc8386b2..92bccc5623a8 100644
@@ -631,6 +631,8 @@ static int i915_driver_hw_probe(struct drm_i915_private *dev_priv)
 	intel_memory_regions_driver_release(dev_priv);
 err_ggtt:
 	i915_ggtt_driver_release(dev_priv);
+	i915_gem_drain_freed_objects(dev_priv);
+	i915_ggtt_driver_late_release(dev_priv);
 err_perf:
 	i915_perf_fini(dev_priv);
 	return ret;
@@ -880,6 +882,8 @@ int i915_driver_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	i915_driver_hw_remove(i915);
 	intel_memory_regions_driver_release(i915);
 	i915_ggtt_driver_release(i915);
+	i915_gem_drain_freed_objects(i915);
+	i915_ggtt_driver_late_release(i915);
 out_cleanup_mmio:
 	i915_driver_mmio_release(i915);
 out_runtime_pm_put:
@@ -936,6 +940,7 @@ static void i915_driver_release(struct drm_device *dev)
 	intel_memory_regions_driver_release(dev_priv);
 	i915_ggtt_driver_release(dev_priv);
 	i915_gem_drain_freed_objects(dev_priv);
+	i915_ggtt_driver_late_release(dev_priv);
 
 	i915_driver_mmio_release(dev_priv);
 
#6 lore raw

Re: [Intel-gfx] [PATCH v2 02/15] drm/i915: Don't free shared locks while shared

Quoted 48 lines · Maarten Lankhorst

Op 18-05-2021 om 10:26 schreef Thomas Hellström:

Quoted 41 lines

We are currently sharing the VM reservation locks across a number of gem objects with page-table memory. Since TTM will individiualize the reservation locks when freeing objects, including accessing the shared locks, make sure that the shared locks are not freed until that is done. For PPGTT we add an additional refcount, for GGTT we take additional measures to make sure objects sharing the GGTT reservation lock are freed at GGTT takedown

---
v2: Try harder to make sure objects sharing the GGTT reservation lock are freed at GGTT takedown.

---
  drivers/gpu/drm/i915/gem/i915_gem_object.c    |  3 ++
  .../gpu/drm/i915/gem/i915_gem_object_types.h  |  1 +
  drivers/gpu/drm/i915/gt/intel_ggtt.c          | 19 ++++++--
  drivers/gpu/drm/i915/gt/intel_gtt.c           | 45 +++++++++++++++----
  drivers/gpu/drm/i915/gt/intel_gtt.h           | 30 ++++++++++++-
  drivers/gpu/drm/i915/gt/intel_ppgtt.c         |  2 +-
  drivers/gpu/drm/i915/i915_drv.c               |  5 +++
  7 files changed, 92 insertions(+), 13 deletions(-)
drivers/gpu/drm/i915/gem/i915_gem_object.c +3 -0
index 28144410df86..abadf0994ad0 100644
@@ -252,6 +252,9 @@ static void __i915_gem_free_objects(struct drm_i915_private *i915,
  		if (obj->mm.n_placements > 1)
  			kfree(obj->mm.placements);
 
+		if (obj->resv_shared_from)
+			i915_vm_resv_put(obj->resv_shared_from);
+
  		/* But keep the pointer alive for RCU-protected lookups */
  		call_rcu(&obj->rcu, __i915_gem_free_object_rcu);
  		cond_resched();
drivers/gpu/drm/i915/gem/i915_gem_object_types.h +0 -0
index 0727d0c76aa0..450340a73186 100644

drivers/gpu/drm/i915/gem/i915_gem_object_types.h struct drm_i915_gem_object { · line 152

Quoted 4 lines
drivers/gpu/drm/i915/gem/i915_gem_object_types.h +1 -0
@@ -149,6 +149,7 @@ struct drm_i915_gem_object {
  	 * when i915_gem_ww_ctx_backoff() or i915_gem_ww_ctx_fini() are called.
  	 */
  	struct list_head obj_link;
+	struct dma_resv *resv_shared_from;

Since this can only be a vm object, would it make sense to make this a pointer to the vm address space, so we can call vm_resv_put on it directly?

Quoted 7 lines

The current pointer type and name makes it look generic, but if you try to use it with anything but an address space, it will blow up.

Otherwise looks good. I guess we cannot force all bo's to be deleted before the vm is freed. :-)

So with that fixed

Thanks, I'll take a look at that.

Thomas

author reviewer
#7 lore raw

PATCH v2 3/15 drm/i915: Fix i915_sg_page_sizes to record dma segments rather than physical pages

In reply to Thomas Hellström

All users of this function actually want the dma segment sizes, but that's not what's calculated. Fix that and rename the function to i915_sg_dma_sizes to reflect what's calculated.

---
 drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c  |  2 +-
 drivers/gpu/drm/i915/gem/i915_gem_phys.c    |  2 +-
 drivers/gpu/drm/i915/gem/i915_gem_userptr.c |  2 +-
 drivers/gpu/drm/i915/i915_scatterlist.h     | 16 ++++++++++++----
 4 files changed, 15 insertions(+), 7 deletions(-)
drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c +1 -1
index ccede73c6465..616c3a2f1baf 100644
@@ -209,7 +209,7 @@ static int i915_gem_object_get_pages_dmabuf(struct drm_i915_gem_object *obj)
 	if (IS_ERR(pages))
 		return PTR_ERR(pages);
 
-	sg_page_sizes = i915_sg_page_sizes(pages->sgl);
+	sg_page_sizes = i915_sg_dma_sizes(pages->sgl);
 
 	__i915_gem_object_set_pages(obj, pages, sg_page_sizes);
 
drivers/gpu/drm/i915/gem/i915_gem_phys.c +1 -1
index 81dc2bf59bc3..36f373dc493c 100644
@@ -208,7 +208,7 @@ static int i915_gem_object_shmem_to_phys(struct drm_i915_gem_object *obj)
 
 err_xfer:
 	if (!IS_ERR_OR_NULL(pages)) {
-		unsigned int sg_page_sizes = i915_sg_page_sizes(pages->sgl);
+		unsigned int sg_page_sizes = i915_sg_dma_sizes(pages->sgl);
 
 		__i915_gem_object_set_pages(obj, pages, sg_page_sizes);
 	}
drivers/gpu/drm/i915/gem/i915_gem_userptr.c +1 -1
index a657b99ec760..602f0ed983ec 100644
@@ -173,7 +173,7 @@ static int i915_gem_userptr_get_pages(struct drm_i915_gem_object *obj)
 		goto err;
 	}
 
-	sg_page_sizes = i915_sg_page_sizes(st->sgl);
+	sg_page_sizes = i915_sg_dma_sizes(st->sgl);
 
 	__i915_gem_object_set_pages(obj, st, sg_page_sizes);
 
drivers/gpu/drm/i915/i915_scatterlist.h +12 -4
index 9cb26a224034..b96baad66a3a 100644
@@ -101,15 +101,23 @@ static inline struct scatterlist *__sg_next(struct scatterlist *sg)
 	     (((__iter).curr += PAGE_SIZE) >= (__iter).max) ?		\
 	     (__iter) = __sgt_iter(__sg_next((__iter).sgp), false), 0 : 0)
 
-static inline unsigned int i915_sg_page_sizes(struct scatterlist *sg)
+/**
+ * i915_sg_dma_sizes - Record the dma segment sizes of a scatterlist
+ * @sg: The scatterlist
+ *
+ * Return: An unsigned int with segment sizes logically or'ed together.
+ * A caller can use this information to determine what hardware page table
+ * entry sizes can be used to map the memory represented by the scatterlist.
+ */
+static inline unsigned int i915_sg_dma_sizes(struct scatterlist *sg)
 {
 	unsigned int page_sizes;
 
 	page_sizes = 0;
-	while (sg) {
+	while (sg && sg_dma_len(sg)) {
 		GEM_BUG_ON(sg->offset);
-		GEM_BUG_ON(!IS_ALIGNED(sg->length, PAGE_SIZE));
-		page_sizes |= sg->length;
+		GEM_BUG_ON(!IS_ALIGNED(sg_dma_len(sg), PAGE_SIZE));
+		page_sizes |= sg_dma_len(sg);
 		sg = __sg_next(sg);
 	}
 
Signature · 6 lines
2.31.1

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Reviewed-by on v2 3/15
#8 lore raw

Re: [Intel-gfx] [PATCH v2 03/15] drm/i915: Fix i915_sg_page_sizes to record dma segments rather than physical pages

Quoted 6 lines · Thomas Hellström

All users of this function actually want the dma segment sizes, but that's not what's calculated. Fix that and rename the function to i915_sg_dma_sizes to reflect what's calculated.

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

author reviewer
#9 lore raw

PATCH v2 4/15 drm/ttm: Export functions to initialize and finalize the ttm range manager standalone

In reply to Thomas Hellström

i915 mock selftests are run without the device set up. In order to be able to run the region related mock selftests, export functions in order for the TTM range manager to be set up without a device to attach it to.

---
 drivers/gpu/drm/ttm/ttm_range_manager.c | 55 +++++++++++++++++--------
 include/drm/ttm/ttm_bo_driver.h         | 23 +++++++++++
 2 files changed, 61 insertions(+), 17 deletions(-)
drivers/gpu/drm/ttm/ttm_range_manager.c +38 -17
index b9d5da6e6a81..6957dfb0cf5a 100644
@@ -125,55 +125,76 @@ static const struct ttm_resource_manager_func ttm_range_manager_func = {
 	.debug = ttm_range_man_debug
 };
 
-int ttm_range_man_init(struct ttm_device *bdev,
-		       unsigned type, bool use_tt,
-		       unsigned long p_size)
+struct ttm_resource_manager *
+ttm_range_man_init_standalone(unsigned long size, bool use_tt)
 {
 	struct ttm_resource_manager *man;
 	struct ttm_range_manager *rman;
 
 	rman = kzalloc(sizeof(*rman), GFP_KERNEL);
 	if (!rman)
-		return -ENOMEM;
+		return ERR_PTR(-ENOMEM);
 
 	man = &rman->manager;
 	man->use_tt = use_tt;
 
 	man->func = &ttm_range_manager_func;
 
-	ttm_resource_manager_init(man, p_size);
+	ttm_resource_manager_init(man, size);
 
-	drm_mm_init(&rman->mm, 0, p_size);
+	drm_mm_init(&rman->mm, 0, size);
 	spin_lock_init(&rman->lock);
 
-	ttm_set_driver_manager(bdev, type, &rman->manager);
+	return man;
+}
+EXPORT_SYMBOL(ttm_range_man_init_standalone);
+
+int ttm_range_man_init(struct ttm_device *bdev,
+		       unsigned int type, bool use_tt,
+		       unsigned long p_size)
+{
+	struct ttm_resource_manager *man;
+
+	man = ttm_range_man_init_standalone(p_size, use_tt);
+	if (IS_ERR(man))
+		return PTR_ERR(man);
+
 	ttm_resource_manager_set_used(man, true);
+	ttm_set_driver_manager(bdev, type, man);
+
 	return 0;
 }
 EXPORT_SYMBOL(ttm_range_man_init);
 
+void ttm_range_man_fini_standalone(struct ttm_resource_manager *man)
+{
+	struct ttm_range_manager *rman = to_range_manager(man);
+	struct drm_mm *mm = &rman->mm;
+
+	spin_lock(&rman->lock);
+	drm_mm_clean(mm);
+	drm_mm_takedown(mm);
+	spin_unlock(&rman->lock);
+
+	ttm_resource_manager_cleanup(man);
+	kfree(rman);
+}
+EXPORT_SYMBOL(ttm_range_man_fini_standalone);
+
 int ttm_range_man_fini(struct ttm_device *bdev,
 		       unsigned type)
 {
 	struct ttm_resource_manager *man = ttm_manager_type(bdev, type);
-	struct ttm_range_manager *rman = to_range_manager(man);
-	struct drm_mm *mm = &rman->mm;
 	int ret;
 
 	ttm_resource_manager_set_used(man, false);
-
 	ret = ttm_resource_manager_evict_all(bdev, man);
 	if (ret)
 		return ret;
 
-	spin_lock(&rman->lock);
-	drm_mm_clean(mm);
-	drm_mm_takedown(mm);
-	spin_unlock(&rman->lock);
-
-	ttm_resource_manager_cleanup(man);
 	ttm_set_driver_manager(bdev, type, NULL);
-	kfree(rman);
+	ttm_range_man_fini_standalone(man);
+
 	return 0;
 }
 EXPORT_SYMBOL(ttm_range_man_fini);
include/drm/ttm/ttm_bo_driver.h +23 -0
index dbccac957f8f..734b1712ea72 100644
@@ -321,6 +321,20 @@ int ttm_range_man_init(struct ttm_device *bdev,
 		       unsigned type, bool use_tt,
 		       unsigned long p_size);
 
+/**
+ * ttm_range_man_init_standalone - Initialize a ttm range manager without
+ * device interaction.
+ * @size: Size of the area to be managed in pages.
+ * @use_tt: The memory type requires tt backing.
+ *
+ * This function is intended for selftests. It initializes a range manager
+ * without any device interaction.
+ *
+ * Return: pointer to a range manager on success. Error pointer on failure.
+ */
+struct ttm_resource_manager *
+ttm_range_man_init_standalone(unsigned long size, bool use_tt);
+
 /**
  * ttm_range_man_fini
  *
@@ -332,4 +346,13 @@ int ttm_range_man_init(struct ttm_device *bdev,
 int ttm_range_man_fini(struct ttm_device *bdev,
 		       unsigned type);
 
+/**
+ * ttm_range_man_fini_standalone
+ * @man: The range manager
+ *
+ * Tear down a range manager initialized with
+ * ttm_range_manager_init_standalone().
+ */
+void ttm_range_man_fini_standalone(struct ttm_resource_manager *man);
+
 #endif
Signature · 1 line
2.31.1
#10 lore raw

Re: [PATCH v2 04/15] drm/ttm: Export functions to initialize and finalize the ttm range manager standalone

Quoted 118 lines · Thomas Hellström

i915 mock selftests are run without the device set up. In order to be able to run the region related mock selftests, export functions in order for the TTM range manager to be set up without a device to attach it to.

---
 drivers/gpu/drm/ttm/ttm_range_manager.c | 55 +++++++++++++++++--------
 include/drm/ttm/ttm_bo_driver.h         | 23 +++++++++++
 2 files changed, 61 insertions(+), 17 deletions(-)
drivers/gpu/drm/ttm/ttm_range_manager.c +38 -17
index b9d5da6e6a81..6957dfb0cf5a 100644
@@ -125,55 +125,76 @@ static const struct ttm_resource_manager_func ttm_range_manager_func = {
 	.debug = ttm_range_man_debug
 };
 
-int ttm_range_man_init(struct ttm_device *bdev,
-		       unsigned type, bool use_tt,
-		       unsigned long p_size)
+struct ttm_resource_manager *
+ttm_range_man_init_standalone(unsigned long size, bool use_tt)
 {
 	struct ttm_resource_manager *man;
 	struct ttm_range_manager *rman;
 
 	rman = kzalloc(sizeof(*rman), GFP_KERNEL);
 	if (!rman)
-		return -ENOMEM;
+		return ERR_PTR(-ENOMEM);
 
 	man = &rman->manager;
 	man->use_tt = use_tt;
 
 	man->func = &ttm_range_manager_func;
 
-	ttm_resource_manager_init(man, p_size);
+	ttm_resource_manager_init(man, size);
 
-	drm_mm_init(&rman->mm, 0, p_size);
+	drm_mm_init(&rman->mm, 0, size);
 	spin_lock_init(&rman->lock);
 
-	ttm_set_driver_manager(bdev, type, &rman->manager);
+	return man;
+}
+EXPORT_SYMBOL(ttm_range_man_init_standalone);
+
+int ttm_range_man_init(struct ttm_device *bdev,
+		       unsigned int type, bool use_tt,
+		       unsigned long p_size)
+{
+	struct ttm_resource_manager *man;
+
+	man = ttm_range_man_init_standalone(p_size, use_tt);
+	if (IS_ERR(man))
+		return PTR_ERR(man);
+
 	ttm_resource_manager_set_used(man, true);
+	ttm_set_driver_manager(bdev, type, man);
+
 	return 0;
 }
 EXPORT_SYMBOL(ttm_range_man_init);
 
+void ttm_range_man_fini_standalone(struct ttm_resource_manager *man)
+{
+	struct ttm_range_manager *rman = to_range_manager(man);
+	struct drm_mm *mm = &rman->mm;
+
+	spin_lock(&rman->lock);
+	drm_mm_clean(mm);
+	drm_mm_takedown(mm);
+	spin_unlock(&rman->lock);
+
+	ttm_resource_manager_cleanup(man);
+	kfree(rman);
+}
+EXPORT_SYMBOL(ttm_range_man_fini_standalone);
+
 int ttm_range_man_fini(struct ttm_device *bdev,
 		       unsigned type)
 {
 	struct ttm_resource_manager *man = ttm_manager_type(bdev, type);
-	struct ttm_range_manager *rman = to_range_manager(man);
-	struct drm_mm *mm = &rman->mm;
 	int ret;
 
 	ttm_resource_manager_set_used(man, false);
-
 	ret = ttm_resource_manager_evict_all(bdev, man);
 	if (ret)
 		return ret;
 
-	spin_lock(&rman->lock);
-	drm_mm_clean(mm);
-	drm_mm_takedown(mm);
-	spin_unlock(&rman->lock);
-
-	ttm_resource_manager_cleanup(man);
 	ttm_set_driver_manager(bdev, type, NULL);
-	kfree(rman);
+	ttm_range_man_fini_standalone(man);
+
 	return 0;
 }
 EXPORT_SYMBOL(ttm_range_man_fini);
include/drm/ttm/ttm_bo_driver.h +1 -0
index dbccac957f8f..734b1712ea72 100644
@@ -321,6 +321,20 @@ int ttm_range_man_init(struct ttm_device *bdev,
 		       unsigned type, bool use_tt,
 		       unsigned long p_size);
 
+/**

include/drm/ttm/ttm_bo_driver.h int ttm_range_man_init(struct ttm_device *bdev, · line 334

Quoted 10 lines
include/drm/ttm/ttm_bo_driver.h +10 -0
+ * ttm_range_man_init_standalone - Initialize a ttm range manager without
+ * device interaction.
+ * @size: Size of the area to be managed in pages.
+ * @use_tt: The memory type requires tt backing.
+ *
+ * This function is intended for selftests. It initializes a range manager
+ * without any device interaction.
+ *
+ * Return: pointer to a range manager on success. Error pointer on failure.
+ */

Kerneldoc is great and I'm happy you're updating them (Christian's not so much good for this), but I think would be good to go one step further with a prep patch:

- Make sure ttm_bo_driver.h is appropriately included in Documentation/gpu/drm-mm.rst.

- Fix up any kerneldoc fallout. Specifically I think common usage at least is that for non-inline functions, the kerneldoc is in the .c file, not in the headers.

But also this might be way too much work since ttm hasn't been properly kerneldoc-ified, so maybe later.
-Daniel

Quoted 23 lines
include/drm/ttm/ttm_bo_driver.h +12 -0
+struct ttm_resource_manager *
+ttm_range_man_init_standalone(unsigned long size, bool use_tt);
+
 /**
  * ttm_range_man_fini
  *
@@ -332,4 +346,13 @@ int ttm_range_man_init(struct ttm_device *bdev,
 int ttm_range_man_fini(struct ttm_device *bdev,
 		       unsigned type);
 
+/**
+ * ttm_range_man_fini_standalone
+ * @man: The range manager
+ *
+ * Tear down a range manager initialized with
+ * ttm_range_manager_init_standalone().
+ */
+void ttm_range_man_fini_standalone(struct ttm_resource_manager *man);
+
 #endif
Signature · 1 line
2.31.1
Signature · 3 lines
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
maintainer
#11 lore raw

Re: [PATCH v2 04/15] drm/ttm: Export functions to initialize and finalize the ttm range manager standalone

In reply to Thomas Hellström

Am 18.05.21 um 10:26 schrieb Thomas Hellström:

Quoted 3 lines

i915 mock selftests are run without the device set up. In order to be able to run the region related mock selftests, export functions in order for the TTM range manager to be set up without a device to attach it to.

From the code it looks good, but to be honest I don't think that this makes much sense from the organizational point of view.

If a self test exercises internals of TTM it should be moved into TTM as well.

Christian.

Quoted 145 lines
---
  drivers/gpu/drm/ttm/ttm_range_manager.c | 55 +++++++++++++++++--------
  include/drm/ttm/ttm_bo_driver.h         | 23 +++++++++++
  2 files changed, 61 insertions(+), 17 deletions(-)
drivers/gpu/drm/ttm/ttm_range_manager.c +38 -17
index b9d5da6e6a81..6957dfb0cf5a 100644
@@ -125,55 +125,76 @@ static const struct ttm_resource_manager_func ttm_range_manager_func = {
  	.debug = ttm_range_man_debug
  };
 
-int ttm_range_man_init(struct ttm_device *bdev,
-		       unsigned type, bool use_tt,
-		       unsigned long p_size)
+struct ttm_resource_manager *
+ttm_range_man_init_standalone(unsigned long size, bool use_tt)
  {
  	struct ttm_resource_manager *man;
  	struct ttm_range_manager *rman;
 
  	rman = kzalloc(sizeof(*rman), GFP_KERNEL);
  	if (!rman)
-		return -ENOMEM;
+		return ERR_PTR(-ENOMEM);
 
  	man = &rman->manager;
  	man->use_tt = use_tt;
 
  	man->func = &ttm_range_manager_func;
 
-	ttm_resource_manager_init(man, p_size);
+	ttm_resource_manager_init(man, size);
 
-	drm_mm_init(&rman->mm, 0, p_size);
+	drm_mm_init(&rman->mm, 0, size);
  	spin_lock_init(&rman->lock);
 
-	ttm_set_driver_manager(bdev, type, &rman->manager);
+	return man;
+}
+EXPORT_SYMBOL(ttm_range_man_init_standalone);
+
+int ttm_range_man_init(struct ttm_device *bdev,
+		       unsigned int type, bool use_tt,
+		       unsigned long p_size)
+{
+	struct ttm_resource_manager *man;
+
+	man = ttm_range_man_init_standalone(p_size, use_tt);
+	if (IS_ERR(man))
+		return PTR_ERR(man);
+
  	ttm_resource_manager_set_used(man, true);
+	ttm_set_driver_manager(bdev, type, man);
+
  	return 0;
  }
  EXPORT_SYMBOL(ttm_range_man_init);
 
+void ttm_range_man_fini_standalone(struct ttm_resource_manager *man)
+{
+	struct ttm_range_manager *rman = to_range_manager(man);
+	struct drm_mm *mm = &rman->mm;
+
+	spin_lock(&rman->lock);
+	drm_mm_clean(mm);
+	drm_mm_takedown(mm);
+	spin_unlock(&rman->lock);
+
+	ttm_resource_manager_cleanup(man);
+	kfree(rman);
+}
+EXPORT_SYMBOL(ttm_range_man_fini_standalone);
+
  int ttm_range_man_fini(struct ttm_device *bdev,
  		       unsigned type)
  {
  	struct ttm_resource_manager *man = ttm_manager_type(bdev, type);
-	struct ttm_range_manager *rman = to_range_manager(man);
-	struct drm_mm *mm = &rman->mm;
  	int ret;
 
  	ttm_resource_manager_set_used(man, false);
-
  	ret = ttm_resource_manager_evict_all(bdev, man);
  	if (ret)
  		return ret;
 
-	spin_lock(&rman->lock);
-	drm_mm_clean(mm);
-	drm_mm_takedown(mm);
-	spin_unlock(&rman->lock);
-
-	ttm_resource_manager_cleanup(man);
  	ttm_set_driver_manager(bdev, type, NULL);
-	kfree(rman);
+	ttm_range_man_fini_standalone(man);
+
  	return 0;
  }
  EXPORT_SYMBOL(ttm_range_man_fini);
include/drm/ttm/ttm_bo_driver.h +23 -0
index dbccac957f8f..734b1712ea72 100644
@@ -321,6 +321,20 @@ int ttm_range_man_init(struct ttm_device *bdev,
  		       unsigned type, bool use_tt,
  		       unsigned long p_size);
 
+/**
+ * ttm_range_man_init_standalone - Initialize a ttm range manager without
+ * device interaction.
+ * @size: Size of the area to be managed in pages.
+ * @use_tt: The memory type requires tt backing.
+ *
+ * This function is intended for selftests. It initializes a range manager
+ * without any device interaction.
+ *
+ * Return: pointer to a range manager on success. Error pointer on failure.
+ */
+struct ttm_resource_manager *
+ttm_range_man_init_standalone(unsigned long size, bool use_tt);
+
  /**
   * ttm_range_man_fini
   *
@@ -332,4 +346,13 @@ int ttm_range_man_init(struct ttm_device *bdev,
  int ttm_range_man_fini(struct ttm_device *bdev,
  		       unsigned type);
 
+/**
+ * ttm_range_man_fini_standalone
+ * @man: The range manager
+ *
+ * Tear down a range manager initialized with
+ * ttm_range_manager_init_standalone().
+ */
+void ttm_range_man_fini_standalone(struct ttm_resource_manager *man);
+
  #endif
author reviewer
#12 lore raw

Re: [PATCH v2 04/15] drm/ttm: Export functions to initialize and finalize the ttm range manager standalone

Quoted 8 lines · Christian König

Am 18.05.21 um 10:26 schrieb Thomas Hellström:

Quoted 3 lines

i915 mock selftests are run without the device set up. In order to be able to run the region related mock selftests, export functions in order for the TTM range manager to be set up without a device to attach it to.

From the code it looks good, but to be honest I don't think that this makes much sense from the organizational point of view.

If a self test exercises internals of TTM it should be moved into TTM as well.

This particular selftest actually exercises i915 memory regions which is a level above TTM, but the memory regions are backed by TTM. Since they are mock selftests they don't have a TTM device. For the buddy allocator the situation would be the same, but there we have selftests that exercise the allocator standalone, and those would probably fit best into a TTM selftest infrastructure.

Although in this particular case, we could of course add a mock TTM device and be done. Pls let me know what you think.

/Thomas

maintainer
#13 lore raw

Re: [PATCH v2 04/15] drm/ttm: Export functions to initialize and finalize the ttm range manager standalone

Am 18.05.21 um 15:06 schrieb Thomas Hellström:

Quoted 7 lines · Christian König
Quoted 5 lines

Am 18.05.21 um 10:26 schrieb Thomas Hellström:

Quoted 3 lines

i915 mock selftests are run without the device set up. In order to be able to run the region related mock selftests, export functions in order for the TTM range manager to be set up without a device to attach it to.

Quoted 7 lines
Quoted 3 lines

From the code it looks good, but to be honest I don't think that this makes much sense from the organizational point of view.

If a self test exercises internals of TTM it should be moved into TTM as well.

This particular selftest actually exercises i915 memory regions which is a level above TTM, but the memory regions are backed by TTM. Since they are mock selftests they don't have a TTM device. For the buddy allocator the situation would be the same, but there we have selftests that exercise the allocator standalone, and those would probably fit best into a TTM selftest infrastructure.

Although in this particular case, we could of course add a mock TTM device and be done. Pls let me know what you think.

Well if you can do that with a dummy device then that would certainly be the preferred option.

Christian.

Quoted 4 lines

/Thomas

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

author reviewer
#14 lore raw

PATCH v2 5/15 drm/i915/ttm Initialize the ttm device and memory managers

In reply to Thomas Hellström

Temporarily remove the buddy allocator and related selftests and hook up the TTM range manager for i915 regions.

Also modify the mock region selftests somewhat to account for a fragmenting manager.

---
v2:
- Fix an error unwind in lmem_get_pages() (Reported by Matthew Auld)
- Break out and modify usage of i915_sg_dma_sizes() (Reported by Mattew Auld)
- Break out TTM changes to a separate patch (Reported by Christian König)

---
 drivers/gpu/drm/i915/Kconfig                  |   1 +
 drivers/gpu/drm/i915/Makefile                 |   2 +-
 drivers/gpu/drm/i915/gem/i915_gem_lmem.c      |  59 +-
 .../gpu/drm/i915/gem/i915_gem_object_types.h  |   6 +-
 drivers/gpu/drm/i915/gem/i915_gem_pages.c     |   3 +-
 drivers/gpu/drm/i915/gem/i915_gem_region.c    | 120 ---
 drivers/gpu/drm/i915/gem/i915_gem_region.h    |   4 -
 drivers/gpu/drm/i915/gem/i915_gem_shmem.c     |   4 +-
 drivers/gpu/drm/i915/gem/i915_gem_stolen.c    |  10 +-
 drivers/gpu/drm/i915/gem/i915_gem_stolen.h    |   9 +-
 drivers/gpu/drm/i915/gt/intel_gt.c            |   2 -
 drivers/gpu/drm/i915/gt/intel_region_lmem.c   |  27 +-
 drivers/gpu/drm/i915/i915_buddy.c             | 435 ----------
 drivers/gpu/drm/i915/i915_buddy.h             | 131 ---
 drivers/gpu/drm/i915/i915_drv.c               |   8 +
 drivers/gpu/drm/i915/i915_drv.h               |   7 +-
 drivers/gpu/drm/i915/i915_gem.c               |   1 +
 drivers/gpu/drm/i915/i915_globals.c           |   1 -
 drivers/gpu/drm/i915/i915_globals.h           |   1 -
 drivers/gpu/drm/i915/i915_scatterlist.c       |  70 ++
 drivers/gpu/drm/i915/i915_scatterlist.h       |   4 +
 drivers/gpu/drm/i915/intel_memory_region.c    | 180 ++--
 drivers/gpu/drm/i915/intel_memory_region.h    |  44 +-
 drivers/gpu/drm/i915/intel_region_ttm.c       | 245 ++++++
 drivers/gpu/drm/i915/intel_region_ttm.h       |  29 +
 drivers/gpu/drm/i915/selftests/i915_buddy.c   | 789 ------------------
 .../drm/i915/selftests/i915_mock_selftests.h  |   1 -
 .../drm/i915/selftests/intel_memory_region.c  | 133 +--
 drivers/gpu/drm/i915/selftests/mock_region.c  |  50 +-
 29 files changed, 622 insertions(+), 1754 deletions(-)
 delete mode 100644 drivers/gpu/drm/i915/i915_buddy.c
 delete mode 100644 drivers/gpu/drm/i915/i915_buddy.h
 create mode 100644 drivers/gpu/drm/i915/intel_region_ttm.c
 create mode 100644 drivers/gpu/drm/i915/intel_region_ttm.h
 delete mode 100644 drivers/gpu/drm/i915/selftests/i915_buddy.c
drivers/gpu/drm/i915/Kconfig +1 -0
index 1e1cb245fca7..b63d374dff23 100644
@@ -26,6 +26,7 @@ config DRM_I915
 	select SND_HDA_I915 if SND_HDA_CORE
 	select CEC_CORE if CEC_NOTIFIER
 	select VMAP_PFN
+	select DRM_TTM
 	help
 	  Choose this option if you have a system that has "Intel Graphics
 	  Media Accelerator" or "HD Graphics" integrated graphics,
drivers/gpu/drm/i915/Makefile +1 -1
index d0d936d9137b..cb8823570996 100644
@@ -50,6 +50,7 @@ i915-y += i915_drv.o \
 	  intel_memory_region.o \
 	  intel_pch.o \
 	  intel_pm.o \
+	  intel_region_ttm.o \
 	  intel_runtime_pm.o \
 	  intel_sideband.o \
 	  intel_step.o \
@@ -160,7 +161,6 @@ gem-y += \
 i915-y += \
 	  $(gem-y) \
 	  i915_active.o \
-	  i915_buddy.o \
 	  i915_cmd_parser.o \
 	  i915_gem_evict.o \
 	  i915_gem_gtt.o \
drivers/gpu/drm/i915/gem/i915_gem_lmem.c +57 -2

73 lines not on this page · raw message

drivers/gpu/drm/i915/gem/i915_gem_object_types.h +4 -2
index 450340a73186..dbd7fffe956e 100644
@@ -232,10 +232,12 @@ struct drm_i915_gem_object {
 		 * Memory region for this object.
 		 */
 		struct intel_memory_region *region;
+
 		/**
-		 * List of memory region blocks allocated for this object.
+		 * Memory manager node allocated for this object.
 		 */
-		struct list_head blocks;
+		void *st_mm_node;
+
 		/**
 		 * Element within memory_region->objects or region->purgeable
 		 * if the object is marked as DONTNEED. Access is protected by
drivers/gpu/drm/i915/gem/i915_gem_pages.c +2 -1
index aed8a37ccdc9..62ee2185a41b 100644
@@ -473,7 +473,8 @@ __i915_gem_object_get_sg(struct drm_i915_gem_object *obj,
 
 	might_sleep();
 	GEM_BUG_ON(n >= obj->base.size >> PAGE_SHIFT);
-	GEM_BUG_ON(!i915_gem_object_has_pinned_pages(obj));
+	if (!i915_gem_object_has_pinned_pages(obj))
+		assert_object_held(obj);
 
 	/* As we iterate forward through the sg, we record each entry in a
 	 * radixtree for quick repeated (backwards) lookups. If we have seen
drivers/gpu/drm/i915/gem/i915_gem_region.c +0 -120

129 lines not on this page · raw message

drivers/gpu/drm/i915/gem/i915_gem_region.h +0 -4
index ebddc86d78f7..84fcb3297400 100644
@@ -12,10 +12,6 @@ struct intel_memory_region;
 struct drm_i915_gem_object;
 struct sg_table;
 
-int i915_gem_object_get_pages_buddy(struct drm_i915_gem_object *obj);
-void i915_gem_object_put_pages_buddy(struct drm_i915_gem_object *obj,
-				     struct sg_table *pages);
-
 void i915_gem_object_init_memory_region(struct drm_i915_gem_object *obj,
 					struct intel_memory_region *mem);
 void i915_gem_object_release_memory_region(struct drm_i915_gem_object *obj);
drivers/gpu/drm/i915/gem/i915_gem_shmem.c +3 -1
index a9bfa66c8da1..5d16c4462fda 100644
@@ -628,11 +628,13 @@ static const struct intel_memory_region_ops shmem_region_ops = {
 	.init_object = shmem_object_init,
 };
 
-struct intel_memory_region *i915_gem_shmem_setup(struct drm_i915_private *i915)
+struct intel_memory_region *i915_gem_shmem_setup(struct drm_i915_private *i915,
+						 u16 type, u16 instance)
 {
 	return intel_memory_region_create(i915, 0,
 					  totalram_pages() << PAGE_SHIFT,
 					  PAGE_SIZE, 0,
+					  type, instance,
 					  &shmem_region_ops);
 }
 
drivers/gpu/drm/i915/gem/i915_gem_stolen.c +6 -4
index 293f640faa0a..c42abee206da 100644
@@ -770,7 +770,8 @@ static const struct intel_memory_region_ops i915_region_stolen_lmem_ops = {
 };
 
 struct intel_memory_region *
-i915_gem_stolen_lmem_setup(struct drm_i915_private *i915)
+i915_gem_stolen_lmem_setup(struct drm_i915_private *i915, u16 type,
+			   u16 instance)
 {
 	struct intel_uncore *uncore = &i915->uncore;
 	struct pci_dev *pdev = to_pci_dev(i915->drm.dev);
@@ -788,6 +789,7 @@ i915_gem_stolen_lmem_setup(struct drm_i915_private *i915)
 
 	mem = intel_memory_region_create(i915, lmem_base, lmem_size,
 					 I915_GTT_PAGE_SIZE_4K, io_start,
+					 type, instance,
 					 &i915_region_stolen_lmem_ops);
 	if (IS_ERR(mem))
 		return mem;
@@ -809,14 +811,15 @@ i915_gem_stolen_lmem_setup(struct drm_i915_private *i915)
 }
 
 struct intel_memory_region*
-i915_gem_stolen_smem_setup(struct drm_i915_private *i915)
+i915_gem_stolen_smem_setup(struct drm_i915_private *i915, u16 type,
+			   u16 instance)
 {
 	struct intel_memory_region *mem;
 
 	mem = intel_memory_region_create(i915,
 					 intel_graphics_stolen_res.start,
 					 resource_size(&intel_graphics_stolen_res),
-					 PAGE_SIZE, 0,
+					 PAGE_SIZE, 0, type, instance,
 					 &i915_region_stolen_smem_ops);
 	if (IS_ERR(mem))
 		return mem;
@@ -824,7 +827,6 @@ i915_gem_stolen_smem_setup(struct drm_i915_private *i915)
 	intel_memory_region_set_name(mem, "stolen-system");
 
 	mem->private = true;
-
 	return mem;
 }
 
drivers/gpu/drm/i915/gem/i915_gem_stolen.h +7 -2
index 2bec6c367b9c..ccdf7befc571 100644
@@ -21,8 +21,13 @@ int i915_gem_stolen_insert_node_in_range(struct drm_i915_private *dev_priv,
 					 u64 end);
 void i915_gem_stolen_remove_node(struct drm_i915_private *dev_priv,
 				 struct drm_mm_node *node);
-struct intel_memory_region *i915_gem_stolen_smem_setup(struct drm_i915_private *i915);
-struct intel_memory_region *i915_gem_stolen_lmem_setup(struct drm_i915_private *i915);
+struct intel_memory_region *
+i915_gem_stolen_smem_setup(struct drm_i915_private *i915, u16 type,
+			   u16 instance);
+struct intel_memory_region *
+i915_gem_stolen_lmem_setup(struct drm_i915_private *i915, u16 type,
+			   u16 instance);
+
 struct drm_i915_gem_object *
 i915_gem_object_create_stolen(struct drm_i915_private *dev_priv,
 			      resource_size_t size);
drivers/gpu/drm/i915/gt/intel_gt.c +0 -2
index 8d77dcbad059..3f88ecdee031 100644
@@ -68,8 +68,6 @@ int intel_gt_probe_lmem(struct intel_gt *gt)
 	id = INTEL_REGION_LMEM;
 
 	mem->id = id;
-	mem->type = INTEL_MEMORY_LOCAL;
-	mem->instance = 0;
 
 	intel_memory_region_set_name(mem, "local%u", mem->instance);
 
drivers/gpu/drm/i915/gt/intel_region_lmem.c +21 -6
index 73fceb0c25fc..f7366b054f8e 100644
@@ -5,6 +5,8 @@
 
 #include "i915_drv.h"
 #include "intel_memory_region.h"
+#include "intel_region_lmem.h"
+#include "intel_region_ttm.h"
 #include "gem/i915_gem_lmem.h"
 #include "gem/i915_gem_region.h"
 #include "intel_region_lmem.h"
@@ -66,9 +68,9 @@ static void release_fake_lmem_bar(struct intel_memory_region *mem)
 static void
 region_lmem_release(struct intel_memory_region *mem)
 {
-	release_fake_lmem_bar(mem);
+	intel_region_ttm_fini(mem);
 	io_mapping_fini(&mem->iomap);
-	intel_memory_region_release_buddy(mem);
+	release_fake_lmem_bar(mem);
 }
 
 static int
@@ -83,12 +85,21 @@ region_lmem_init(struct intel_memory_region *mem)
 
 	if (!io_mapping_init_wc(&mem->iomap,
 				mem->io_start,
-				resource_size(&mem->region)))
-		return -EIO;
+				resource_size(&mem->region))) {
+		ret = -EIO;
+		goto out_no_io;
+	}
 
-	ret = intel_memory_region_init_buddy(mem);
+	ret = intel_region_ttm_init(mem);
 	if (ret)
-		io_mapping_fini(&mem->iomap);
+		goto out_no_buddy;
+
+	return 0;
+
+out_no_buddy:
+	io_mapping_fini(&mem->iomap);
+out_no_io:
+	release_fake_lmem_bar(mem);
 
 	return ret;
 }
@@ -127,6 +138,8 @@ intel_gt_setup_fake_lmem(struct intel_gt *gt)
 					 mappable_end,
 					 PAGE_SIZE,
 					 io_start,
+					 INTEL_MEMORY_LOCAL,
+					 0,
 					 &intel_region_lmem_ops);
 	if (!IS_ERR(mem)) {
 		drm_info(&i915->drm, "Intel graphics fake LMEM: %pR\n",
@@ -198,6 +211,8 @@ static struct intel_memory_region *setup_lmem(struct intel_gt *gt)
 					 lmem_size,
 					 I915_GTT_PAGE_SIZE_4K,
 					 io_start,
+					 INTEL_MEMORY_LOCAL,
+					 0,
 					 &intel_region_lmem_ops);
 	if (IS_ERR(mem))
 		return mem;
drivers/gpu/drm/i915/i915_buddy.c deleted +0 -435

435 lines not on this page · raw message

drivers/gpu/drm/i915/i915_buddy.h deleted +0 -131

131 lines not on this page · raw message

drivers/gpu/drm/i915/i915_drv.c +8 -0
index 92bccc5623a8..122dd297b6af 100644
@@ -84,6 +84,7 @@
 #include "intel_gvt.h"
 #include "intel_memory_region.h"
 #include "intel_pm.h"
+#include "intel_region_ttm.h"
 #include "intel_sideband.h"
 #include "vlv_suspend.h"
 
@@ -335,6 +336,10 @@ static int i915_driver_early_probe(struct drm_i915_private *dev_priv)
 	if (ret < 0)
 		goto err_workqueues;
 
+	ret = intel_region_ttm_device_init(dev_priv);
+	if (ret)
+		goto err_ttm;
+
 	intel_wopcm_init_early(&dev_priv->wopcm);
 
 	intel_gt_init_early(&dev_priv->gt, dev_priv);
@@ -359,6 +364,8 @@ static int i915_driver_early_probe(struct drm_i915_private *dev_priv)
 err_gem:
 	i915_gem_cleanup_early(dev_priv);
 	intel_gt_driver_late_release(&dev_priv->gt);
+	intel_region_ttm_device_fini(dev_priv);
+err_ttm:
 	vlv_suspend_cleanup(dev_priv);
 err_workqueues:
 	i915_workqueues_cleanup(dev_priv);
@@ -376,6 +383,7 @@ static void i915_driver_late_release(struct drm_i915_private *dev_priv)
 	intel_power_domains_cleanup(dev_priv);
 	i915_gem_cleanup_early(dev_priv);
 	intel_gt_driver_late_release(&dev_priv->gt);
+	intel_region_ttm_device_fini(dev_priv);
 	vlv_suspend_cleanup(dev_priv);
 	i915_workqueues_cleanup(dev_priv);
 
drivers/gpu/drm/i915/i915_drv.h +6 -1
index 128198e8b4d0..1276a2f60740 100644
@@ -60,6 +60,7 @@
 #include <drm/drm_atomic.h>
 #include <drm/drm_connector.h>
 #include <drm/i915_mei_hdcp_interface.h>
+#include <drm/ttm/ttm_device.h>
 
 #include "i915_params.h"
 #include "i915_reg.h"
@@ -1166,6 +1167,9 @@ struct drm_i915_private {
 	/* Mutex to protect the above hdcp component related values. */
 	struct mutex hdcp_comp_mutex;
 
+	/* The TTM device structure. */
+	struct ttm_device bdev;
+
 	I915_SELFTEST_DECLARE(struct i915_selftest_stash selftest;)
 
 	/*
@@ -1751,7 +1755,8 @@ void i915_gem_cleanup_userptr(struct drm_i915_private *dev_priv);
 void i915_gem_init_early(struct drm_i915_private *dev_priv);
 void i915_gem_cleanup_early(struct drm_i915_private *dev_priv);
 
-struct intel_memory_region *i915_gem_shmem_setup(struct drm_i915_private *i915);
+struct intel_memory_region *i915_gem_shmem_setup(struct drm_i915_private *i915,
+						 u16 type, u16 instance);
 
 static inline void i915_gem_drain_freed_objects(struct drm_i915_private *i915)
 {
drivers/gpu/drm/i915/i915_gem.c +1 -0
index d0018c5f88bd..180f6e9107d4 100644
@@ -1109,6 +1109,7 @@ int i915_gem_init(struct drm_i915_private *dev_priv)
 	}
 
 	i915_gem_drain_freed_objects(dev_priv);
+
 	return ret;
 }
 
drivers/gpu/drm/i915/i915_globals.c +0 -1
index 3aa213684293..77f1911c463b 100644
@@ -87,7 +87,6 @@ static void __i915_globals_cleanup(void)
 
 static __initconst int (* const initfn[])(void) = {
 	i915_global_active_init,
-	i915_global_buddy_init,
 	i915_global_context_init,
 	i915_global_gem_context_init,
 	i915_global_objects_init,
drivers/gpu/drm/i915/i915_globals.h +0 -1
index b2f5cd9b9b1a..2d199f411a4a 100644
@@ -27,7 +27,6 @@ void i915_globals_exit(void);
 
 /* constructors */
 int i915_global_active_init(void);
-int i915_global_buddy_init(void);
 int i915_global_context_init(void);
 int i915_global_gem_context_init(void);
 int i915_global_objects_init(void);
drivers/gpu/drm/i915/i915_scatterlist.c +70 -0

82 lines not on this page · raw message

drivers/gpu/drm/i915/i915_scatterlist.h +4 -0
index b96baad66a3a..5acca45ea981 100644
@@ -13,6 +13,8 @@
 
 #include "i915_gem.h"
 
+struct drm_mm_node;
+
 /*
  * Optimised SGL iterator for GEM objects
  */
@@ -141,4 +143,6 @@ static inline unsigned int i915_sg_segment_size(void)
 
 bool i915_sg_trim(struct sg_table *orig_st);
 
+struct sg_table *i915_sg_from_mm_node(const struct drm_mm_node *node,
+				      u64 region_start);
 #endif
drivers/gpu/drm/i915/intel_memory_region.c +51 -129

245 lines not on this page · raw message

drivers/gpu/drm/i915/intel_memory_region.h +23 -21

94 lines not on this page · raw message

drivers/gpu/drm/i915/intel_region_ttm.c new file +245 -0

245 lines not on this page · raw message

drivers/gpu/drm/i915/intel_region_ttm.h new file +29 -0
new file mode 100644
index 000000000000..40129f7e0317
@@ -0,0 +1,29 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2021 Intel Corporation
+ */
+#ifndef _INTEL_REGION_TTM_H_
+#define _INTEL_REGION_TTM_H_
+
+#include <linux/types.h>
+
+struct drm_i915_private;
+struct intel_memory_region;
+
+int intel_region_ttm_device_init(struct drm_i915_private *dev_priv);
+
+void intel_region_ttm_device_fini(struct drm_i915_private *dev_priv);
+
+int intel_region_ttm_init(struct intel_memory_region *mem);
+
+void intel_region_ttm_fini(struct intel_memory_region *mem);
+
+struct sg_table *intel_region_ttm_node_to_st(struct intel_memory_region *mem,
+					     void *node);
+
+void *intel_region_ttm_node_alloc(struct intel_memory_region *mem,
+				  resource_size_t size,
+				  unsigned int flags);
+void intel_region_ttm_node_free(struct intel_memory_region *mem,
+				void *node);
+#endif /* _INTEL_REGION_TTM_H_ */
drivers/gpu/drm/i915/selftests/i915_buddy.c deleted +0 -789

789 lines not on this page · raw message

drivers/gpu/drm/i915/selftests/i915_mock_selftests.h +0 -1
index 3db34d3eea58..34e5caf38093 100644
@@ -33,5 +33,4 @@ selftest(evict, i915_gem_evict_mock_selftests)
 selftest(gtt, i915_gem_gtt_mock_selftests)
 selftest(hugepages, i915_gem_huge_page_mock_selftests)
 selftest(contexts, i915_gem_context_mock_selftests)
-selftest(buddy, i915_buddy_mock_selftests)
 selftest(memory_region, intel_memory_region_mock_selftests)
drivers/gpu/drm/i915/selftests/intel_memory_region.c +40 -93

213 lines not on this page · raw message

drivers/gpu/drm/i915/selftests/mock_region.c +43 -7

80 lines not on this page · raw message

Signature · 6 lines
2.31.1

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Reviewed-by on v2 5/15
#15 lore raw

Re: [Intel-gfx] [PATCH v2 05/15] drm/i915/ttm Initialize the ttm device and memory managers

Quoted 14 lines · Thomas Hellström

Temporarily remove the buddy allocator and related selftests and hook up the TTM range manager for i915 regions.

Also modify the mock region selftests somewhat to account for a fragmenting manager.

---
v2:
- Fix an error unwind in lmem_get_pages() (Reported by Matthew Auld)
- Break out and modify usage of i915_sg_dma_sizes() (Reported by Mattew Auld)
- Break out TTM changes to a separate patch (Reported by Christian König)
---

<snip>

Quoted 7 lines
drivers/gpu/drm/i915/i915_scatterlist.c +6 -0
+/**
+ * i915_sg_from_mm_node - Create an sg_table from a struct drm_mm_node
+ * @node: The drm_mm_node.
+ * @region_start: An offset to add to the dma addresses of the sg list.
+ *
+ * Create a struct sg_table, initializing it from a struct drm_mm_node,

drivers/gpu/drm/i915/i915_scatterlist.c bool i915_sg_trim(struct sg_table *orig_st) · line 56

Quoted 10 lines
drivers/gpu/drm/i915/i915_scatterlist.c +10 -0
+ * taking a maximum segment length into account, splitting into segments
+ * if necessary.
+ *
+ * Return: A pointer to a kmalloced struct sg_table on success, negative
+ * error code cast to an error pointer on failure.
+ */
+struct sg_table *i915_sg_from_mm_node(const struct drm_mm_node *node,
+                                     u64 region_start)
+{
+       const u64 max_segment = SZ_1G; /* Do we have a limit on this? */

For lmem just INT_MAX I think, which is the limit of the sg, but really doesn't matter for now, this should be totally fine for now.

Assuming CI is happy,

Also we could maybe fling this series with the HAX autoprobing patch for DG1 at trybot, just to see where we are?

#16 lore raw

Re: [Intel-gfx] [PATCH v2 05/15] drm/i915/ttm Initialize the ttm device and memory managers

In reply to Thomas Hellström
Quoted 14 lines · Thomas Hellström

Temporarily remove the buddy allocator and related selftests and hook up the TTM range manager for i915 regions.

Also modify the mock region selftests somewhat to account for a fragmenting manager.

---
v2:
- Fix an error unwind in lmem_get_pages() (Reported by Matthew Auld)
- Break out and modify usage of i915_sg_dma_sizes() (Reported by Mattew Auld)
- Break out TTM changes to a separate patch (Reported by Christian König)
---

<snip>

Quoted 9 lines
drivers/gpu/drm/i915/selftests/mock_region.c +9 -0
+
+static int mock_region_get_pages(struct drm_i915_gem_object *obj)
+{
+       unsigned int flags;
+       struct sg_table *pages;
+
+       flags = I915_ALLOC_MIN_PAGE_SIZE;
+       if (obj->flags & I915_BO_ALLOC_CONTIGUOUS)
+               flags |= I915_ALLOC_CONTIGUOUS;

drivers/gpu/drm/i915/selftests/mock_region.c line 39

Quoted 10 lines
drivers/gpu/drm/i915/selftests/mock_region.c +10 -0
+
+       obj->mm.st_mm_node = intel_region_ttm_node_alloc(obj->mm.region,
+                                                        obj->base.size,
+                                                        flags);
+       if (IS_ERR(obj->mm.st_mm_node))
+               return PTR_ERR(obj->mm.st_mm_node);
+
+       pages = intel_region_ttm_node_to_st(obj->mm.region, obj->mm.st_mm_node);
+       if (IS_ERR(pages))
+               return PTR_ERR(pages);

Needs some onion?

author reviewer
#17 lore raw

Re: [Intel-gfx] [PATCH v2 05/15] drm/i915/ttm Initialize the ttm device and memory managers

Quoted 29 lines · Matthew Auld
Quoted 13 lines · Thomas Hellström

Temporarily remove the buddy allocator and related selftests and hook up the TTM range manager for i915 regions.

Also modify the mock region selftests somewhat to account for a fragmenting manager.

---
v2:
- Fix an error unwind in lmem_get_pages() (Reported by Matthew Auld)
- Break out and modify usage of i915_sg_dma_sizes() (Reported by Mattew Auld)
- Break out TTM changes to a separate patch (Reported by Christian König)
---

<snip>

Quoted 12 lines
drivers/gpu/drm/i915/selftests/mock_region.c +12 -0
+
+static int mock_region_get_pages(struct drm_i915_gem_object *obj)
+{
+       unsigned int flags;
+       struct sg_table *pages;
+
+       flags = I915_ALLOC_MIN_PAGE_SIZE;
+       if (obj->flags & I915_BO_ALLOC_CONTIGUOUS)
+               flags |= I915_ALLOC_CONTIGUOUS;
+
+       obj->mm.st_mm_node = intel_region_ttm_node_alloc(obj->mm.region,
+                                                        obj->base.size,
Quoted 8 lines

drivers/gpu/drm/i915/selftests/mock_region.c line 39

Quoted 7 lines
drivers/gpu/drm/i915/selftests/mock_region.c +7 -0
+                                                        flags);
+       if (IS_ERR(obj->mm.st_mm_node))
+               return PTR_ERR(obj->mm.st_mm_node);
+
+       pages = intel_region_ttm_node_to_st(obj->mm.region, obj->mm.st_mm_node);
+       if (IS_ERR(pages))
+               return PTR_ERR(pages);

Needs some onion?

Ah, yes.

/Thomas

author reviewer
#18 lore raw

PATCH v2 6/15 drm/i915/ttm: Embed a ttm buffer object in the i915 gem object

In reply to Thomas Hellström

Embed a struct ttm_buffer_object into the i915 gem object, making sure we alias the gem object part. It's a bit unfortunate that the struct ttm_buffer_ojbect embeds a gem object since we otherwise could make the TTM part private to the TTM backend, and use the usual i915 gem object for the other backends.
To make this a bit more storage efficient for the other backends, we'd have to use a pointer for the gem object which would require a lot of changes in the driver. We postpone that for later.

---
 drivers/gpu/drm/i915/gem/i915_gem_object.c       |  7 +++++++
 drivers/gpu/drm/i915/gem/i915_gem_object_types.h | 12 +++++++++++-
 2 files changed, 18 insertions(+), 1 deletion(-)
drivers/gpu/drm/i915/gem/i915_gem_object.c +7 -0
index abadf0994ad0..c8953e3f5c70 100644
@@ -62,6 +62,13 @@ void i915_gem_object_init(struct drm_i915_gem_object *obj,
 			  const struct drm_i915_gem_object_ops *ops,
 			  struct lock_class_key *key, unsigned flags)
 {
+	/*
+	 * A gem object is embedded both in a struct ttm_buffer_object :/ and
+	 * in a drm_i915_gem_object. Make sure they are aliased.
+	 */
+	BUILD_BUG_ON(offsetof(typeof(*obj), base) !=
+		     offsetof(typeof(*obj), __do_not_access.base));
+
 	spin_lock_init(&obj->vma.lock);
 	INIT_LIST_HEAD(&obj->vma.list);
 
drivers/gpu/drm/i915/gem/i915_gem_object_types.h +11 -1
index dbd7fffe956e..98f69d8fd37d 100644
@@ -10,6 +10,7 @@
 #include <linux/mmu_notifier.h>
 
 #include <drm/drm_gem.h>
+#include <drm/ttm/ttm_bo_api.h>
 #include <uapi/drm/i915_drm.h>
 
 #include "i915_active.h"
@@ -99,7 +100,16 @@ struct i915_gem_object_page_iter {
 };
 
 struct drm_i915_gem_object {
-	struct drm_gem_object base;
+	/*
+	 * We might have reason to revisit the below since it wastes
+	 * a lot of space for non-ttm gem objects.
+	 * In any case, always use the accessors for the ttm_buffer_object
+	 * when accessing it.
+	 */
+	union {
+		struct drm_gem_object base;
+		struct ttm_buffer_object __do_not_access;
+	};
 
 	const struct drm_i915_gem_object_ops *ops;
 
Signature · 6 lines
2.31.1

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
maintainer Acked-by on v2 6/15
#19 lore raw

Re: [PATCH v2 06/15] drm/i915/ttm: Embed a ttm buffer object in the i915 gem object

Op 18-05-2021 om 10:26 schreef Thomas Hellström:

Quoted 53 lines

Embed a struct ttm_buffer_object into the i915 gem object, making sure we alias the gem object part. It's a bit unfortunate that the struct ttm_buffer_ojbect embeds a gem object since we otherwise could make the TTM part private to the TTM backend, and use the usual i915 gem object for the other backends.
To make this a bit more storage efficient for the other backends, we'd have to use a pointer for the gem object which would require a lot of changes in the driver. We postpone that for later.

---
 drivers/gpu/drm/i915/gem/i915_gem_object.c       |  7 +++++++
 drivers/gpu/drm/i915/gem/i915_gem_object_types.h | 12 +++++++++++-
 2 files changed, 18 insertions(+), 1 deletion(-)
drivers/gpu/drm/i915/gem/i915_gem_object.c +7 -0
index abadf0994ad0..c8953e3f5c70 100644
@@ -62,6 +62,13 @@ void i915_gem_object_init(struct drm_i915_gem_object *obj,
 			  const struct drm_i915_gem_object_ops *ops,
 			  struct lock_class_key *key, unsigned flags)
 {
+	/*
+	 * A gem object is embedded both in a struct ttm_buffer_object :/ and
+	 * in a drm_i915_gem_object. Make sure they are aliased.
+	 */
+	BUILD_BUG_ON(offsetof(typeof(*obj), base) !=
+		     offsetof(typeof(*obj), __do_not_access.base));
+
 	spin_lock_init(&obj->vma.lock);
 	INIT_LIST_HEAD(&obj->vma.list);
 
drivers/gpu/drm/i915/gem/i915_gem_object_types.h +4 -1
index dbd7fffe956e..98f69d8fd37d 100644
@@ -10,6 +10,7 @@
 #include <linux/mmu_notifier.h>
 
 #include <drm/drm_gem.h>
+#include <drm/ttm/ttm_bo_api.h>
 #include <uapi/drm/i915_drm.h>
 
 #include "i915_active.h"
@@ -99,7 +100,16 @@ struct i915_gem_object_page_iter {
 };
 
 struct drm_i915_gem_object {
-	struct drm_gem_object base;
+	/*
+	 * We might have reason to revisit the below since it wastes
+	 * a lot of space for non-ttm gem objects.

drivers/gpu/drm/i915/gem/i915_gem_object_types.h struct i915_gem_object_page_iter { · line 115

Quoted 10 lines
drivers/gpu/drm/i915/gem/i915_gem_object_types.h +7 -0
+	 * In any case, always use the accessors for the ttm_buffer_object
+	 * when accessing it.
+	 */
+	union {
+		struct drm_gem_object base;
+		struct ttm_buffer_object __do_not_access;
+	};
 
 	const struct drm_i915_gem_object_ops *ops;
 

Considering Dave did roughly the same in his patches, I don't think there's a better way to do this.

Although he just wrapped base.base using sizes. This works too. It probably needs someone else's r-b too, to ensure this is allowed.

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

author reviewer
#20 lore raw

PATCH v2 7/15 drm/ttm: Export ttm_bo_tt_destroy()

In reply to Thomas Hellström

For the upcoming kmapping i915 memcpy_move, export ttm_bo_tt_destroy(). A future change might be to move the new memcpy_move into ttm, replacing the old ioremapping one.

---
 drivers/gpu/drm/ttm/ttm_bo.c | 1 +
 1 file changed, 1 insertion(+)
drivers/gpu/drm/ttm/ttm_bo.c +1 -0
index ca1b098b6a56..4479c55aaa1d 100644
@@ -1221,3 +1221,4 @@ void ttm_bo_tt_destroy(struct ttm_buffer_object *bo)
 	ttm_tt_destroy(bo->bdev, bo->ttm);
 	bo->ttm = NULL;
 }
+EXPORT_SYMBOL(ttm_bo_tt_destroy);
Signature · 1 line
2.31.1
maintainer Reviewed-by on v2 7/15
#21 lore raw

Re: [PATCH v2 07/15] drm/ttm: Export ttm_bo_tt_destroy()

Op 18-05-2021 om 10:26 schreef Thomas Hellström:

Quoted 15 lines

For the upcoming kmapping i915 memcpy_move, export ttm_bo_tt_destroy(). A future change might be to move the new memcpy_move into ttm, replacing the old ioremapping one.

---
 drivers/gpu/drm/ttm/ttm_bo.c | 1 +
 1 file changed, 1 insertion(+)
drivers/gpu/drm/ttm/ttm_bo.c +0 -0
index ca1b098b6a56..4479c55aaa1d 100644

drivers/gpu/drm/ttm/ttm_bo.c void ttm_bo_tt_destroy(struct ttm_buffer_object *bo) · line 1224

Quoted 4 lines
drivers/gpu/drm/ttm/ttm_bo.c +1 -0
@@ -1221,3 +1221,4 @@ void ttm_bo_tt_destroy(struct ttm_buffer_object *bo)
 	ttm_tt_destroy(bo->bdev, bo->ttm);
 	bo->ttm = NULL;
 }
+EXPORT_SYMBOL(ttm_bo_tt_destroy);

Looks sane, could we reorder the patches to put all ttm changes first?

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

maintainer
#22 lore raw

Re: [PATCH v2 07/15] drm/ttm: Export ttm_bo_tt_destroy()

In reply to Thomas Hellström

Am 18.05.21 um 10:26 schrieb Thomas Hellström:

Quoted 3 lines

For the upcoming kmapping i915 memcpy_move, export ttm_bo_tt_destroy(). A future change might be to move the new memcpy_move into ttm, replacing the old ioremapping one.

Well this is an upfront NAK for that approach.

I've worked quite hard to not export those functions any more since this is not something drivers should be allowed to mess with.

If you need this we should probably move the functionality into TTM instead.

Regards,
Christian.

Quoted 12 lines
---
  drivers/gpu/drm/ttm/ttm_bo.c | 1 +
  1 file changed, 1 insertion(+)
drivers/gpu/drm/ttm/ttm_bo.c +0 -0
index ca1b098b6a56..4479c55aaa1d 100644

drivers/gpu/drm/ttm/ttm_bo.c void ttm_bo_tt_destroy(struct ttm_buffer_object *bo) · line 1224

Quoted 4 lines
drivers/gpu/drm/ttm/ttm_bo.c +1 -0
@@ -1221,3 +1221,4 @@ void ttm_bo_tt_destroy(struct ttm_buffer_object *bo)
  	ttm_tt_destroy(bo->bdev, bo->ttm);
  	bo->ttm = NULL;
  }
+EXPORT_SYMBOL(ttm_bo_tt_destroy);

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

author reviewer
#23 lore raw

PATCH v2 8/15 drm/i915/ttm Add a generic TTM memcpy move for page-based iomem

In reply to Thomas Hellström

The internal ttm_bo_util memcpy uses vmap functionality, and while it probably might be possible to use it for copying in- and out of sglist represented io memory, using io_mem_reserve() / io_mem_free() callbacks, that would cause problems with fault().
Instead, implement a method mapping page-by-page using kmap_local() semantics. As an additional benefit we then avoid the occasional global TLB flushes of vmap() and consuming vmap space, elimination of a critical point of failure and with a slight change of semantics we could also push the memcpy out async for testing and async driver develpment purposes. Pushing out async can be done since there is no memory allocation going on that could violate the dma_fence lockdep rules.

For copies from iomem, use the WC prefetching memcpy variant for additional speed.

Note that drivers that don't want to use struct io_mapping but relies on memremap functionality, and that don't want to use scatterlists for VRAM may well define specialized (hopefully reusable) iterators for their particular environment.

---
v2:
- Move new TTM exports to a separate commit. (Reported by Christian König)
- Avoid having the iterator init functions inline. (Reported by Jani Nikula)
- Remove a stray comment.

---
 drivers/gpu/drm/i915/Makefile                 |   1 +
 .../gpu/drm/i915/gem/i915_gem_ttm_bo_util.c   | 194 ++++++++++++++++++
 .../gpu/drm/i915/gem/i915_gem_ttm_bo_util.h   | 107 ++++++++++
 3 files changed, 302 insertions(+)
 create mode 100644 drivers/gpu/drm/i915/gem/i915_gem_ttm_bo_util.c
 create mode 100644 drivers/gpu/drm/i915/gem/i915_gem_ttm_bo_util.h
drivers/gpu/drm/i915/Makefile +1 -0
index cb8823570996..958ccc1edfed 100644
@@ -155,6 +155,7 @@ gem-y += \
 	gem/i915_gem_stolen.o \
 	gem/i915_gem_throttle.o \
 	gem/i915_gem_tiling.o \
+	gem/i915_gem_ttm_bo_util.o \
 	gem/i915_gem_userptr.o \
 	gem/i915_gem_wait.o \
 	gem/i915_gemfs.o
drivers/gpu/drm/i915/gem/i915_gem_ttm_bo_util.c new file +194 -0

194 lines not on this page · raw message

drivers/gpu/drm/i915/gem/i915_gem_ttm_bo_util.h new file +107 -0

107 lines not on this page · raw message

Signature · 6 lines
2.31.1

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
maintainer
#24 lore raw

Re: [PATCH v2 08/15] drm/i915/ttm Add a generic TTM memcpy move for page-based iomem

Am 18.05.21 um 10:26 schrieb Thomas Hellström:

Quoted 12 lines

The internal ttm_bo_util memcpy uses vmap functionality, and while it probably might be possible to use it for copying in- and out of sglist represented io memory, using io_mem_reserve() / io_mem_free() callbacks, that would cause problems with fault().
Instead, implement a method mapping page-by-page using kmap_local() semantics. As an additional benefit we then avoid the occasional global TLB flushes of vmap() and consuming vmap space, elimination of a critical point of failure and with a slight change of semantics we could also push the memcpy out async for testing and async driver develpment purposes. Pushing out async can be done since there is no memory allocation going on that could violate the dma_fence lockdep rules.

Quoted 7 lines

For copies from iomem, use the WC prefetching memcpy variant for additional speed.

Note that drivers that don't want to use struct io_mapping but relies on memremap functionality, and that don't want to use scatterlists for VRAM may well define specialized (hopefully reusable) iterators for their particular environment.

In general yes please since I have that as TODO for TTM for a very long time.

But I would prefer to fix the implementation in TTM instead and give it proper cursor handling.

Amdgpu is also using page based iomem and we are having similar workarounds in place there as well.

I think it makes sense to unify this inside TTM and remove the old memcpy util function when done.

Regards,
Christian.

Quoted 341 lines

---
v2:
- Move new TTM exports to a separate commit. (Reported by Christian König)
- Avoid having the iterator init functions inline. (Reported by Jani Nikula)
- Remove a stray comment.

---
  drivers/gpu/drm/i915/Makefile                 |   1 +
  .../gpu/drm/i915/gem/i915_gem_ttm_bo_util.c   | 194 ++++++++++++++++++
  .../gpu/drm/i915/gem/i915_gem_ttm_bo_util.h   | 107 ++++++++++
  3 files changed, 302 insertions(+)

create mode 100644 drivers/gpu/drm/i915/gem/i915_gem_ttm_bo_util.c create mode 100644 drivers/gpu/drm/i915/gem/i915_gem_ttm_bo_util.h

drivers/gpu/drm/i915/Makefile +1 -0
index cb8823570996..958ccc1edfed 100644
@@ -155,6 +155,7 @@ gem-y += \
  	gem/i915_gem_stolen.o \
  	gem/i915_gem_throttle.o \
  	gem/i915_gem_tiling.o \
+	gem/i915_gem_ttm_bo_util.o \
  	gem/i915_gem_userptr.o \
  	gem/i915_gem_wait.o \
  	gem/i915_gemfs.o
drivers/gpu/drm/i915/gem/i915_gem_ttm_bo_util.c new file +194 -0
new file mode 100644
index 000000000000..5f347a85bf44
@@ -0,0 +1,194 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2021 Intel Corporation
+ */
+
+/**
+ * DOC: Usage and intentions.
+ *
+ * This file contains functionality that we might want to move into
+ * ttm_bo_util.c if there is a common interest.
+ * Currently a kmap_local only memcpy with support for page-based iomem regions,
+ * and fast memcpy from write-combined memory.
+ */
+
+#include <linux/dma-buf-map.h>
+#include <linux/highmem.h>
+#include <linux/io-mapping.h>
+#include <linux/scatterlist.h>
+
+#include "i915_memcpy.h"
+
+#include "gem/i915_gem_ttm_bo_util.h"
+
+static void i915_ttm_kmap_iter_tt_kmap_local(struct i915_ttm_kmap_iter *iter,
+					     struct dma_buf_map *dmap,
+					     pgoff_t i)
+{
+	struct i915_ttm_kmap_iter_tt *iter_tt =
+		container_of(iter, typeof(*iter_tt), base);
+
+	dma_buf_map_set_vaddr(dmap, kmap_local_page(iter_tt->tt->pages[i]));
+}
+
+static void i915_ttm_kmap_iter_iomap_kmap_local(struct i915_ttm_kmap_iter *iter,
+						struct dma_buf_map *dmap,
+						pgoff_t i)
+{
+	struct i915_ttm_kmap_iter_iomap *iter_io =
+		container_of(iter, typeof(*iter_io), base);
+	void __iomem *addr;
+
+retry:
+	while (i >= iter_io->cache.end) {
+		iter_io->cache.sg = iter_io->cache.sg ?
+			sg_next(iter_io->cache.sg) : iter_io->st->sgl;
+		iter_io->cache.i = iter_io->cache.end;
+		iter_io->cache.end += sg_dma_len(iter_io->cache.sg) >>
+			PAGE_SHIFT;
+		iter_io->cache.offs = sg_dma_address(iter_io->cache.sg) -
+			iter_io->start;
+	}
+
+	if (i < iter_io->cache.i) {
+		iter_io->cache.end = 0;
+		iter_io->cache.sg = NULL;
+		goto retry;
+	}
+
+	addr = io_mapping_map_local_wc(iter_io->iomap, iter_io->cache.offs +
+				       (((resource_size_t)i - iter_io->cache.i)
+					<< PAGE_SHIFT));
+	dma_buf_map_set_vaddr_iomem(dmap, addr);
+}
+
+static const struct i915_ttm_kmap_iter_ops i915_ttm_kmap_iter_tt_ops = {
+	.kmap_local = i915_ttm_kmap_iter_tt_kmap_local
+};
+
+static const struct i915_ttm_kmap_iter_ops i915_ttm_kmap_iter_io_ops = {
+	.kmap_local =  i915_ttm_kmap_iter_iomap_kmap_local
+};
+
+static void kunmap_local_dma_buf_map(struct dma_buf_map *map)
+{
+	if (map->is_iomem)
+		io_mapping_unmap_local(map->vaddr_iomem);
+	else
+		kunmap_local(map->vaddr);
+}
+
+/**
+ * i915_ttm_move_memcpy - Helper to perform a memcpy ttm move operation.
+ * @bo: The struct ttm_buffer_object.
+ * @new_mem: The struct ttm_resource we're moving to (copy destination).
+ * @new_kmap: A struct i915_ttm_kmap_iter representing the destination resource.
+ * @old_kmap: A struct i915_ttm_kmap_iter representing the source resource.
+ */
+void i915_ttm_move_memcpy(struct ttm_buffer_object *bo,
+			  struct ttm_resource *new_mem,
+			  struct i915_ttm_kmap_iter *new_kmap,
+			  struct i915_ttm_kmap_iter *old_kmap)
+{
+	struct ttm_device *bdev = bo->bdev;
+	struct ttm_resource_manager *man = ttm_manager_type(bdev, new_mem->mem_type);
+	struct ttm_tt *ttm = bo->ttm;
+	struct ttm_resource *old_mem = &bo->mem;
+	struct ttm_resource old_copy = *old_mem;
+	struct ttm_resource_manager *old_man = ttm_manager_type(bdev, old_mem->mem_type);
+	struct dma_buf_map old_map, new_map;
+	pgoff_t i;
+
+	/* Single TTM move. NOP */
+	if (old_man->use_tt && man->use_tt)
+		goto done;
+
+	/* Don't move nonexistent data. Clear destination instead. */
+	if (old_man->use_tt && !man->use_tt &&
+	    (!ttm || !ttm_tt_is_populated(ttm))) {
+		if (ttm && !(ttm->page_flags & TTM_PAGE_FLAG_ZERO_ALLOC))
+			goto done;
+
+		for (i = 0; i < new_mem->num_pages; ++i) {
+			new_kmap->ops->kmap_local(new_kmap, &new_map, i);
+			memset_io(new_map.vaddr_iomem, 0, PAGE_SIZE);
+			kunmap_local_dma_buf_map(&new_map);
+		}
+		goto done;
+	}
+
+	for (i = 0; i < new_mem->num_pages; ++i) {
+		new_kmap->ops->kmap_local(new_kmap, &new_map, i);
+		old_kmap->ops->kmap_local(old_kmap, &old_map, i);
+		if (!old_map.is_iomem ||
+		    !i915_memcpy_from_wc(new_map.vaddr, old_map.vaddr, PAGE_SIZE)) {
+			if (!old_map.is_iomem) {
+				dma_buf_map_memcpy_to(&new_map, old_map.vaddr,
+						      PAGE_SIZE);
+			} else if (!new_map.is_iomem) {
+				memcpy_fromio(new_map.vaddr, old_map.vaddr_iomem,
+					      PAGE_SIZE);
+			} else {
+				pgoff_t j;
+				u32 __iomem *src = old_map.vaddr_iomem;
+				u32 __iomem *dst = new_map.vaddr_iomem;
+
+				for (j = 0; j < (PAGE_SIZE >> 2); ++j)
+					iowrite32(ioread32(src++), dst++);
+			}
+		}
+		kunmap_local_dma_buf_map(&old_map);
+		kunmap_local_dma_buf_map(&new_map);
+	}
+
+done:
+	old_copy = *old_mem;
+
+	ttm_bo_assign_mem(bo, new_mem);
+
+	if (!man->use_tt)
+		ttm_bo_tt_destroy(bo);
+
+	ttm_resource_free(bo, &old_copy);
+}
+
+/**
+ * i915_ttm_kmap_iter_iomap_init - Initialize a struct i915_ttm_kmap_iter_iomap
+ * @iter_io: The struct i915_ttm_kmap_iter_iomap to initialize.
+ * @iomap: The struct io_mapping representing the underlying linear io_memory.
+ * @st: sg_table into @iomap, representing the memory of the struct
+ * ttm_resource.
+ * @start: Offset that needs to be subtracted from @st to make
+ * sg_dma_address(st->sgl) - @start == 0 for @iomap start.
+ *
+ * Return: Pointer to the embedded struct i915_ttm_kmap_iter.
+ */
+struct i915_ttm_kmap_iter *
+i915_ttm_kmap_iter_iomap_init(struct i915_ttm_kmap_iter_iomap *iter_io,
+			      struct io_mapping *iomap,
+			      struct sg_table *st,
+			      resource_size_t start)
+{
+	iter_io->base.ops = &i915_ttm_kmap_iter_io_ops;
+	iter_io->iomap = iomap;
+	iter_io->st = st;
+	iter_io->start = start;
+	memset(&iter_io->cache, 0, sizeof(iter_io->cache));
+	return &iter_io->base;
+}
+
+/**
+ * ttm_kmap_iter_tt_init - Initialize a struct i915_ttm_kmap_iter_tt
+ * @iter_tt: The struct i915_ttm_kmap_iter_tt to initialize.
+ * @tt: Struct ttm_tt holding page pointers of the struct ttm_resource.
+ *
+ * Return: Pointer to the embedded struct i915_ttm_kmap_iter.
+ */
+struct i915_ttm_kmap_iter *
+i915_ttm_kmap_iter_tt_init(struct i915_ttm_kmap_iter_tt *iter_tt,
+			   struct ttm_tt *tt)
+{
+	iter_tt->base.ops = &i915_ttm_kmap_iter_tt_ops;
+	iter_tt->tt = tt;
+	return &iter_tt->base;
+}
drivers/gpu/drm/i915/gem/i915_gem_ttm_bo_util.h new file +107 -0
new file mode 100644
index 000000000000..f1c8c7ae7a42
@@ -0,0 +1,107 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2021 Intel Corporation
+ */
+
+/*
+ * This files contains functionality that we might want to move into
+ * ttm_bo_util.c if there is a common interest.
+ */
+#ifndef _I915_GEM_TTM_BO_UTIL_H_
+#define _I915_GEM_TTM_BO_UTIL_H_
+
+#include <drm/ttm/ttm_bo_driver.h>
+struct dma_buf_map;
+struct io_mapping;
+struct sg_table;
+struct scatterlist;
+
+struct ttm_tt;
+struct i915_ttm_kmap_iter;
+
+/**
+ * struct i915_ttm_kmap_iter_ops - Ops structure for a struct
+ * i915_ttm_kmap_iter.
+ */
+struct i915_ttm_kmap_iter_ops {
+	/**
+	 * kmap_local - Map a PAGE_SIZE part of the resource using
+	 * kmap_local semantics.
+	 * @res_kmap: Pointer to the struct i915_ttm_kmap_iter representing
+	 * the resource.
+	 * @dmap: The struct dma_buf_map holding the virtual address after
+	 * the operation.
+	 * @i: The location within the resource to map. PAGE_SIZE granularity.
+	 */
+	void (*kmap_local)(struct i915_ttm_kmap_iter *res_kmap,
+			   struct dma_buf_map *dmap, pgoff_t i);
+};
+
+/**
+ * struct i915_ttm_kmap_iter - Iterator for kmap_local type operations on a
+ * resource.
+ * @ops: Pointer to the operations struct.
+ *
+ * This struct is intended to be embedded in a resource-specific specialization
+ * implementing operations for the resource.
+ *
+ * Nothing stops us from extending the operations to vmap, vmap_pfn etc,
+ * replacing some or parts of the ttm_bo_util. cpu-map functionality.
+ */
+struct i915_ttm_kmap_iter {
+	const struct i915_ttm_kmap_iter_ops *ops;
+};
+
+/**
+ * struct i915_ttm_kmap_iter_tt - Specialization for a tt (page) backed struct
+ * ttm_resource.
+ * @base: Embedded struct i915_ttm_kmap_iter providing the usage interface
+ * @tt: Cached struct ttm_tt.
+ */
+struct i915_ttm_kmap_iter_tt {
+	struct i915_ttm_kmap_iter base;
+	struct ttm_tt *tt;
+};
+
+/**
+ * struct i915_ttm_kmap_iter_iomap - Specialization for a struct io_mapping +
+ * struct sg_table backed struct ttm_resource.
+ * @base: Embedded struct i915_ttm_kmap_iter providing the usage interface.
+ * @iomap: struct io_mapping representing the underlying linear io_memory.
+ * @st: sg_table into @iomap, representing the memory of the struct ttm_resource.
+ * @start: Offset that needs to be subtracted from @st to make
+ * sg_dma_address(st->sgl) - @start == 0 for @iomap start.
+ * @cache: Scatterlist traversal cache for fast lookups.
+ * @cache.sg: Pointer to the currently cached scatterlist segment.
+ * @cache.i: First index of @sg. PAGE_SIZE granularity.
+ * @cache.end: Last index + 1 of @sg. PAGE_SIZE granularity.
+ * @cache.offs: First offset into @iomap of @sg. PAGE_SIZE granularity.
+ */
+struct i915_ttm_kmap_iter_iomap {
+	struct i915_ttm_kmap_iter base;
+	struct io_mapping *iomap;
+	struct sg_table *st;
+	resource_size_t start;
+	struct {
+		struct scatterlist *sg;
+		pgoff_t i;
+		pgoff_t end;
+		pgoff_t offs;
+	} cache;
+};
+
+void i915_ttm_move_memcpy(struct ttm_buffer_object *bo,
+			  struct ttm_resource *new_mem,
+			  struct i915_ttm_kmap_iter *new_iter,
+			  struct i915_ttm_kmap_iter *old_iter);
+
+struct i915_ttm_kmap_iter *
+i915_ttm_kmap_iter_tt_init(struct i915_ttm_kmap_iter_tt *iter_tt,
+			   struct ttm_tt *tt);
+
+struct i915_ttm_kmap_iter *
+i915_ttm_kmap_iter_iomap_init(struct i915_ttm_kmap_iter_iomap *iter_io,
+			      struct io_mapping *iomap,
+			      struct sg_table *st,
+			      resource_size_t start);
+#endif
author reviewer
#25 lore raw

Re: [PATCH v2 08/15] drm/i915/ttm Add a generic TTM memcpy move for page-based iomem

Quoted 25 lines · Christian König

Am 18.05.21 um 10:26 schrieb Thomas Hellström:

Quoted 12 lines

The internal ttm_bo_util memcpy uses vmap functionality, and while it probably might be possible to use it for copying in- and out of sglist represented io memory, using io_mem_reserve() / io_mem_free() callbacks, that would cause problems with fault().
Instead, implement a method mapping page-by-page using kmap_local() semantics. As an additional benefit we then avoid the occasional global TLB flushes of vmap() and consuming vmap space, elimination of a critical point of failure and with a slight change of semantics we could also push the memcpy out async for testing and async driver develpment purposes. Pushing out async can be done since there is no memory allocation going on that could violate the dma_fence lockdep rules.

Quoted 7 lines

For copies from iomem, use the WC prefetching memcpy variant for additional speed.

Note that drivers that don't want to use struct io_mapping but relies on memremap functionality, and that don't want to use scatterlists for VRAM may well define specialized (hopefully reusable) iterators for their particular environment.

In general yes please since I have that as TODO for TTM for a very long time.

Quoted 8 lines

But I would prefer to fix the implementation in TTM instead and give it proper cursor handling.

Amdgpu is also using page based iomem and we are having similar workarounds in place there as well.

I think it makes sense to unify this inside TTM and remove the old memcpy util function when done.

Regards,
Christian.

Christian,

I was thinking when we replace the bo.mem with a pointer (and perhaps have a driver callback to allocate the bo->mem,
we could perhaps embed a struct ttm_kmap_iter and use it for all mapping in one way or another). That would mean perhaps land this is i915 now and sort out the unification once the struct ttm_resource, struct ttm_buffer_object separation has landed?

/Thomas

maintainer
#26 lore raw

Re: [PATCH v2 08/15] drm/i915/ttm Add a generic TTM memcpy move for page-based iomem

Am 18.05.21 um 14:04 schrieb Thomas Hellström:

Quoted 33 lines · Christian König
Quoted 31 lines

Am 18.05.21 um 10:26 schrieb Thomas Hellström:

Quoted 12 lines

The internal ttm_bo_util memcpy uses vmap functionality, and while it probably might be possible to use it for copying in- and out of sglist represented io memory, using io_mem_reserve() / io_mem_free() callbacks, that would cause problems with fault().
Instead, implement a method mapping page-by-page using kmap_local() semantics. As an additional benefit we then avoid the occasional global TLB flushes of vmap() and consuming vmap space, elimination of a critical point of failure and with a slight change of semantics we could also push the memcpy out async for testing and async driver develpment purposes. Pushing out async can be done since there is no memory allocation going on that could violate the dma_fence lockdep rules.

Quoted 7 lines

For copies from iomem, use the WC prefetching memcpy variant for additional speed.

Note that drivers that don't want to use struct io_mapping but relies on memremap functionality, and that don't want to use scatterlists for VRAM may well define specialized (hopefully reusable) iterators for their particular environment.

In general yes please since I have that as TODO for TTM for a very long time.

But I would prefer to fix the implementation in TTM instead and give it proper cursor handling.

Amdgpu is also using page based iomem and we are having similar workarounds in place there as well.

I think it makes sense to unify this inside TTM and remove the old memcpy util function when done.

Quoted 7 lines
Quoted 2 lines

Regards,
Christian.

Christian,

I was thinking when we replace the bo.mem with a pointer (and perhaps have a driver callback to allocate the bo->mem,
we could perhaps embed a struct ttm_kmap_iter and use it for all mapping in one way or another). That would mean perhaps land this is i915 now and sort out the unification once the struct ttm_resource, struct ttm_buffer_object separation has landed?

That stuff is ready, reviewed and I'm just waiting for some amdgpu changes to land in drm-misc-next to push it.

But yes in general an iterator for the resource object sounds like the right plan to me as well.

Christian.

Quoted 4 lines

/Thomas

author reviewer
#27 lore raw

Re: [PATCH v2 08/15] drm/i915/ttm Add a generic TTM memcpy move for page-based iomem

Quoted 39 lines · Christian König

Am 18.05.21 um 14:04 schrieb Thomas Hellström:

Quoted 38 lines · Christian König
Quoted 25 lines

Am 18.05.21 um 10:26 schrieb Thomas Hellström:

Quoted 12 lines

The internal ttm_bo_util memcpy uses vmap functionality, and while it probably might be possible to use it for copying in- and out of sglist represented io memory, using io_mem_reserve() / io_mem_free() callbacks, that would cause problems with fault().
Instead, implement a method mapping page-by-page using kmap_local() semantics. As an additional benefit we then avoid the occasional global TLB flushes of vmap() and consuming vmap space, elimination of a critical point of failure and with a slight change of semantics we could also push the memcpy out async for testing and async driver develpment purposes. Pushing out async can be done since there is no memory allocation going on that could violate the dma_fence lockdep rules.

Quoted 7 lines

For copies from iomem, use the WC prefetching memcpy variant for additional speed.

Note that drivers that don't want to use struct io_mapping but relies on memremap functionality, and that don't want to use scatterlists for VRAM may well define specialized (hopefully reusable) iterators for their particular environment.

In general yes please since I have that as TODO for TTM for a very long time.

Quoted 8 lines

But I would prefer to fix the implementation in TTM instead and give it proper cursor handling.

Amdgpu is also using page based iomem and we are having similar workarounds in place there as well.

I think it makes sense to unify this inside TTM and remove the old memcpy util function when done.

Regards,
Christian.

Christian,

Quoted 8 lines
Quoted 2 lines

I was thinking when we replace the bo.mem with a pointer (and perhaps have a driver callback to allocate the bo->mem,
we could perhaps embed a struct ttm_kmap_iter and use it for all mapping in one way or another). That would mean perhaps land this is i915 now and sort out the unification once the struct ttm_resource, struct ttm_buffer_object separation has landed?

That stuff is ready, reviewed and I'm just waiting for some amdgpu changes to land in drm-misc-next to push it.

But yes in general an iterator for the resource object sounds like the right plan to me as well.

Christian.

OK, so then are you OK with landing this in i915 for now? That would also ofc mean the export you NAK'd but strictly for this memcpy use until we merge it with TTM?

/Thomas

Quoted 6 lines
Quoted 5 lines

/Thomas

maintainer
#28 lore raw

Re: [PATCH v2 08/15] drm/i915/ttm Add a generic TTM memcpy move for page-based iomem

Am 18.05.21 um 14:52 schrieb Thomas Hellström:

Quoted 44 lines · Christian König
Quoted 42 lines

Am 18.05.21 um 14:04 schrieb Thomas Hellström:

Quoted 40 lines · Christian König
Quoted 25 lines

Am 18.05.21 um 10:26 schrieb Thomas Hellström:

Quoted 12 lines

The internal ttm_bo_util memcpy uses vmap functionality, and while it probably might be possible to use it for copying in- and out of sglist represented io memory, using io_mem_reserve() / io_mem_free() callbacks, that would cause problems with fault().
Instead, implement a method mapping page-by-page using kmap_local() semantics. As an additional benefit we then avoid the occasional global TLB flushes of vmap() and consuming vmap space, elimination of a critical point of failure and with a slight change of semantics we could also push the memcpy out async for testing and async driver develpment purposes. Pushing out async can be done since there is no memory allocation going on that could violate the dma_fence lockdep rules.

Quoted 7 lines

For copies from iomem, use the WC prefetching memcpy variant for additional speed.

Note that drivers that don't want to use struct io_mapping but relies on memremap functionality, and that don't want to use scatterlists for VRAM may well define specialized (hopefully reusable) iterators for their particular environment.

In general yes please since I have that as TODO for TTM for a very long time.

Quoted 8 lines

But I would prefer to fix the implementation in TTM instead and give it proper cursor handling.

Amdgpu is also using page based iomem and we are having similar workarounds in place there as well.

I think it makes sense to unify this inside TTM and remove the old memcpy util function when done.

Regards,
Christian.

Christian,

I was thinking when we replace the bo.mem with a pointer (and perhaps have a driver callback to allocate the bo->mem,
we could perhaps embed a struct ttm_kmap_iter and use it for all mapping in one way or another). That would mean perhaps land this is i915 now and sort out the unification once the struct ttm_resource, struct ttm_buffer_object separation has landed?

Quoted 7 lines
Quoted 5 lines

That stuff is ready, reviewed and I'm just waiting for some amdgpu changes to land in drm-misc-next to push it.

But yes in general an iterator for the resource object sounds like the right plan to me as well.

Christian.

OK, so then are you OK with landing this in i915 for now? That would also ofc mean the export you NAK'd but strictly for this memcpy use until we merge it with TTM?

Well you can of course prototype that in i915, but I really don't want to export the TT functions upstream.

Can we cleanly move that functionality into TTM instead?

Christian.

Quoted 10 lines

/Thomas

Quoted 6 lines
Quoted 5 lines

/Thomas

author reviewer
#29 lore raw

Re: [PATCH v2 08/15] drm/i915/ttm Add a generic TTM memcpy move for page-based iomem

Quoted 47 lines · Christian König

Am 18.05.21 um 14:52 schrieb Thomas Hellström:

Quoted 46 lines · Christian König
Quoted 44 lines

Am 18.05.21 um 14:04 schrieb Thomas Hellström:

Quoted 33 lines · Christian König
Quoted 31 lines

Am 18.05.21 um 10:26 schrieb Thomas Hellström:

Quoted 12 lines

The internal ttm_bo_util memcpy uses vmap functionality, and while it probably might be possible to use it for copying in- and out of sglist represented io memory, using io_mem_reserve() / io_mem_free() callbacks, that would cause problems with fault().
Instead, implement a method mapping page-by-page using kmap_local() semantics. As an additional benefit we then avoid the occasional global TLB flushes of vmap() and consuming vmap space, elimination of a critical point of failure and with a slight change of semantics we could also push the memcpy out async for testing and async driver develpment purposes. Pushing out async can be done since there is no memory allocation going on that could violate the dma_fence lockdep rules.

Quoted 7 lines

For copies from iomem, use the WC prefetching memcpy variant for additional speed.

Note that drivers that don't want to use struct io_mapping but relies on memremap functionality, and that don't want to use scatterlists for VRAM may well define specialized (hopefully reusable) iterators for their particular environment.

In general yes please since I have that as TODO for TTM for a very long time.

But I would prefer to fix the implementation in TTM instead and give it proper cursor handling.

Amdgpu is also using page based iomem and we are having similar workarounds in place there as well.

I think it makes sense to unify this inside TTM and remove the old memcpy util function when done.

Quoted 7 lines
Quoted 2 lines

Regards,
Christian.

Christian,

I was thinking when we replace the bo.mem with a pointer (and perhaps have a driver callback to allocate the bo->mem,
we could perhaps embed a struct ttm_kmap_iter and use it for all mapping in one way or another). That would mean perhaps land this is i915 now and sort out the unification once the struct ttm_resource, struct ttm_buffer_object separation has landed?

That stuff is ready, reviewed and I'm just waiting for some amdgpu changes to land in drm-misc-next to push it.

Quoted 7 lines
Quoted 5 lines
Quoted 3 lines

But yes in general an iterator for the resource object sounds like the right plan to me as well.

Christian.

OK, so then are you OK with landing this in i915 for now? That would also ofc mean the export you NAK'd but strictly for this memcpy use until we merge it with TTM?

Well you can of course prototype that in i915, but I really don't want to export the TT functions upstream.

I understand, I once had the same thoughts trying to avoid that as far as possible, so this function was actually then added to the ttm_bo interface, (hence the awkward naming) as a helper for drivers implementing move(), essentially a very special case of ttm_bo_move_accel_cleanup(), but anyway, see below:

Quoted 2 lines

Can we cleanly move that functionality into TTM instead?

I'll take a look at that, but I think we'd initially be having iterators mimicing the current move_memcpy() for the linear iomem !WC cases, hope that's OK.

/Thomas

Quoted 14 lines

Christian.

Quoted 10 lines

/Thomas

Quoted 6 lines
Quoted 5 lines

/Thomas

maintainer
#30 lore raw

Re: [PATCH v2 08/15] drm/i915/ttm Add a generic TTM memcpy move for page-based iomem

Am 18.05.21 um 15:24 schrieb Thomas Hellström:

Quoted 57 lines · Christian König
Quoted 47 lines

Am 18.05.21 um 14:52 schrieb Thomas Hellström:

Quoted 46 lines · Christian König
Quoted 44 lines

Am 18.05.21 um 14:04 schrieb Thomas Hellström:

Quoted 33 lines · Christian König
Quoted 31 lines

Am 18.05.21 um 10:26 schrieb Thomas Hellström:

Quoted 12 lines

The internal ttm_bo_util memcpy uses vmap functionality, and while it probably might be possible to use it for copying in- and out of sglist represented io memory, using io_mem_reserve() / io_mem_free() callbacks, that would cause problems with fault().
Instead, implement a method mapping page-by-page using kmap_local() semantics. As an additional benefit we then avoid the occasional global TLB flushes of vmap() and consuming vmap space, elimination of a critical point of failure and with a slight change of semantics we could also push the memcpy out async for testing and async driver develpment purposes. Pushing out async can be done since there is no memory allocation going on that could violate the dma_fence lockdep rules.

Quoted 7 lines

For copies from iomem, use the WC prefetching memcpy variant for additional speed.

Note that drivers that don't want to use struct io_mapping but relies on memremap functionality, and that don't want to use scatterlists for VRAM may well define specialized (hopefully reusable) iterators for their particular environment.

In general yes please since I have that as TODO for TTM for a very long time.

But I would prefer to fix the implementation in TTM instead and give it proper cursor handling.

Amdgpu is also using page based iomem and we are having similar workarounds in place there as well.

I think it makes sense to unify this inside TTM and remove the old memcpy util function when done.

Quoted 7 lines
Quoted 2 lines

Regards,
Christian.

Christian,

I was thinking when we replace the bo.mem with a pointer (and perhaps have a driver callback to allocate the bo->mem,
we could perhaps embed a struct ttm_kmap_iter and use it for all mapping in one way or another). That would mean perhaps land this is i915 now and sort out the unification once the struct ttm_resource, struct ttm_buffer_object separation has landed?

That stuff is ready, reviewed and I'm just waiting for some amdgpu changes to land in drm-misc-next to push it.

Quoted 7 lines
Quoted 5 lines
Quoted 3 lines

But yes in general an iterator for the resource object sounds like the right plan to me as well.

Christian.

OK, so then are you OK with landing this in i915 for now? That would also ofc mean the export you NAK'd but strictly for this memcpy use until we merge it with TTM?

Well you can of course prototype that in i915, but I really don't want to export the TT functions upstream.

Quoted 7 lines

I understand, I once had the same thoughts trying to avoid that as far as possible, so this function was actually then added to the ttm_bo interface, (hence the awkward naming) as a helper for drivers implementing move(), essentially a very special case of ttm_bo_move_accel_cleanup(), but anyway, see below:

Quoted 2 lines

Can we cleanly move that functionality into TTM instead?

I'll take a look at that, but I think we'd initially be having iterators mimicing the current move_memcpy() for the linear iomem !WC cases, hope that's OK.

Yeah, that's peefectly fine with me. I can tackle cleaning up all drivers and move over to the new implementation when that is fully complete.

As I said we already have the same problem in amdgpu and only solved it by avoiding memcpy all together.

Christian.

Quoted 18 lines

/Thomas

Quoted 14 lines

Christian.

Quoted 10 lines

/Thomas

Quoted 6 lines
Quoted 5 lines

/Thomas

author reviewer
#31 lore raw

PATCH v2 9/15 drm/ttm, drm/amdgpu: Allow the driver some control over swapping

In reply to Thomas Hellström

We are calling the eviction_valuable driver callback at eviction time to determine whether we actually can evict a buffer object.
The upcoming i915 TTM backend needs the same functionality for swapout, and that might actually be beneficial to other drivers as well.

Add an eviction_valuable call also in the swapout path. Try to keep the current behaviour for all drivers by returning true if the buffer object is already in the TTM_PL_SYSTEM placement. We change behaviour for the case where a buffer object is in a TT backed placement when swapped out, in which case the drivers normal eviction_valuable path is run.

Finally export ttm_tt_unpopulate() and don't swap out bos that are not populated. This allows a driver to purge a bo at swapout time if its content is no longer valuable rather than to have TTM swap the contents out.

---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c |  4 +++
 drivers/gpu/drm/ttm/ttm_bo.c            | 41 +++++++++++++++----------
 drivers/gpu/drm/ttm/ttm_tt.c            |  4 +++
 3 files changed, 33 insertions(+), 16 deletions(-)
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +4 -0
index 8c7ec09eb1a4..d5a9d7a88315 100644
@@ -1399,6 +1399,10 @@ static bool amdgpu_ttm_bo_eviction_valuable(struct ttm_buffer_object *bo,
 	struct dma_fence *f;
 	int i;
 
+	/* Swapout? */
+	if (bo->mem.mem_type == TTM_PL_SYSTEM)
+		return true;
+
 	if (bo->type == ttm_bo_type_kernel &&
 	    !amdgpu_vm_evictable(ttm_to_amdgpu_bo(bo)))
 		return false;
drivers/gpu/drm/ttm/ttm_bo.c +25 -16

86 lines not on this page · raw message

drivers/gpu/drm/ttm/ttm_tt.c +4 -0
index 539e0232cb3b..7878ca4876c5 100644
@@ -258,6 +258,9 @@ int ttm_tt_swapout(struct ttm_device *bdev, struct ttm_tt *ttm,
 	struct page *to_page;
 	int i, ret;
 
+	if (!ttm_tt_is_populated(ttm))
+		return 0;
+
 	swap_storage = shmem_file_setup("ttm swap", size, 0);
 	if (IS_ERR(swap_storage)) {
 		pr_err("Failed allocating swap storage\n");
@@ -399,6 +402,7 @@ void ttm_tt_unpopulate(struct ttm_device *bdev, struct ttm_tt *ttm)
 
 	ttm->page_flags &= ~TTM_PAGE_FLAG_PRIV_POPULATED;
 }
+EXPORT_SYMBOL(ttm_tt_unpopulate);
 
 #ifdef CONFIG_DEBUG_FS
 
Signature · 6 lines
2.31.1

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
maintainer Reviewed-by on v2 9/15
#32 lore raw

Re: [Intel-gfx] [PATCH v2 09/15] drm/ttm, drm/amdgpu: Allow the driver some control over swapping

Op 18-05-2021 om 10:26 schreef Thomas Hellström:

Quoted 66 lines

We are calling the eviction_valuable driver callback at eviction time to determine whether we actually can evict a buffer object.
The upcoming i915 TTM backend needs the same functionality for swapout, and that might actually be beneficial to other drivers as well.

Add an eviction_valuable call also in the swapout path. Try to keep the current behaviour for all drivers by returning true if the buffer object is already in the TTM_PL_SYSTEM placement. We change behaviour for the case where a buffer object is in a TT backed placement when swapped out, in which case the drivers normal eviction_valuable path is run.

Finally export ttm_tt_unpopulate() and don't swap out bos that are not populated. This allows a driver to purge a bo at swapout time if its content is no longer valuable rather than to have TTM swap the contents out.

---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c |  4 +++
 drivers/gpu/drm/ttm/ttm_bo.c            | 41 +++++++++++++++----------
 drivers/gpu/drm/ttm/ttm_tt.c            |  4 +++
 3 files changed, 33 insertions(+), 16 deletions(-)
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +4 -0
index 8c7ec09eb1a4..d5a9d7a88315 100644
@@ -1399,6 +1399,10 @@ static bool amdgpu_ttm_bo_eviction_valuable(struct ttm_buffer_object *bo,
 	struct dma_fence *f;
 	int i;
 
+	/* Swapout? */
+	if (bo->mem.mem_type == TTM_PL_SYSTEM)
+		return true;
+
 	if (bo->type == ttm_bo_type_kernel &&
 	    !amdgpu_vm_evictable(ttm_to_amdgpu_bo(bo)))
 		return false;
drivers/gpu/drm/ttm/ttm_bo.c +7 -1
index 4479c55aaa1d..6a3f3112f62a 100644
@@ -531,6 +531,10 @@ static int ttm_bo_evict(struct ttm_buffer_object *bo,
 bool ttm_bo_eviction_valuable(struct ttm_buffer_object *bo,
 			      const struct ttm_place *place)
 {
+	dma_resv_assert_held(bo->base.resv);
+	if (bo->mem.mem_type == TTM_PL_SYSTEM)
+		return true;
+
 	/* Don't evict this BO if it's outside of the
 	 * requested placement range
 	 */
@@ -553,7 +557,9 @@ EXPORT_SYMBOL(ttm_bo_eviction_valuable);
  * b. Otherwise, trylock it.
  */
 static bool ttm_bo_evict_swapout_allowable(struct ttm_buffer_object *bo,
-			struct ttm_operation_ctx *ctx, bool *locked, bool *busy)
+					   struct ttm_operation_ctx *ctx,
+					   const struct ttm_place *place,
+					   bool *locked, bool *busy)
 {
 	bool ret = false;
 

drivers/gpu/drm/ttm/ttm_bo.c static bool ttm_bo_evict_swapout_allowable(struct ttm_buffer_object *bo, · line 584

Quoted 8 lines
drivers/gpu/drm/ttm/ttm_bo.c +5 -0
@@ -571,6 +577,12 @@ static bool ttm_bo_evict_swapout_allowable(struct ttm_buffer_object *bo,
 			*busy = !ret;
 	}
 
+	if (ret && place && !bo->bdev->funcs->eviction_valuable(bo, place)) {
+		ret = false;
+		if (locked)
+			dma_resv_unlock(bo->base.resv);
+	}

Probably meant to check and clear *locked here?

With that fixed:

Quoted 83 lines
drivers/gpu/drm/ttm/ttm_bo.c +13 -15
+
 	return ret;
 }
 
@@ -625,20 +637,14 @@ int ttm_mem_evict_first(struct ttm_device *bdev,
 		list_for_each_entry(bo, &man->lru[i], lru) {
 			bool busy;
 
-			if (!ttm_bo_evict_swapout_allowable(bo, ctx, &locked,
-							    &busy)) {
+			if (!ttm_bo_evict_swapout_allowable(bo, ctx, place,
+							    &locked, &busy)) {
 				if (busy && !busy_bo && ticket !=
 				    dma_resv_locking_ctx(bo->base.resv))
 					busy_bo = bo;
 				continue;
 			}
 
-			if (place && !bdev->funcs->eviction_valuable(bo,
-								      place)) {
-				if (locked)
-					dma_resv_unlock(bo->base.resv);
-				continue;
-			}
 			if (!ttm_bo_get_unless_zero(bo)) {
 				if (locked)
 					dma_resv_unlock(bo->base.resv);
@@ -1138,10 +1144,18 @@ EXPORT_SYMBOL(ttm_bo_wait);
 int ttm_bo_swapout(struct ttm_buffer_object *bo, struct ttm_operation_ctx *ctx,
 		   gfp_t gfp_flags)
 {
+	struct ttm_place place = {};
 	bool locked;
 	int ret;
 
-	if (!ttm_bo_evict_swapout_allowable(bo, ctx, &locked, NULL))
+	/*
+	 * While the bo may already reside in SYSTEM placement, set
+	 * SYSTEM as new placement to cover also the move further below.
+	 * The driver may use the fact that we're moving from SYSTEM
+	 * as an indication that we're about to swap out.
+	 */
+	place.mem_type = TTM_PL_SYSTEM;
+	if (!ttm_bo_evict_swapout_allowable(bo, ctx, &place, &locked, NULL))
 		return -EBUSY;
 
 	if (!ttm_bo_get_unless_zero(bo)) {
@@ -1166,12 +1180,7 @@ int ttm_bo_swapout(struct ttm_buffer_object *bo, struct ttm_operation_ctx *ctx,
 	if (bo->mem.mem_type != TTM_PL_SYSTEM) {
 		struct ttm_operation_ctx ctx = { false, false };
 		struct ttm_resource evict_mem;
-		struct ttm_place place, hop;
-
-		memset(&place, 0, sizeof(place));
-		memset(&hop, 0, sizeof(hop));
-
-		place.mem_type = TTM_PL_SYSTEM;
+		struct ttm_place hop = {};
 
 		ret = ttm_resource_alloc(bo, &place, &evict_mem);
 		if (unlikely(ret))
drivers/gpu/drm/ttm/ttm_tt.c +4 -0
index 539e0232cb3b..7878ca4876c5 100644
@@ -258,6 +258,9 @@ int ttm_tt_swapout(struct ttm_device *bdev, struct ttm_tt *ttm,
 	struct page *to_page;
 	int i, ret;
 
+	if (!ttm_tt_is_populated(ttm))
+		return 0;
+
 	swap_storage = shmem_file_setup("ttm swap", size, 0);
 	if (IS_ERR(swap_storage)) {
 		pr_err("Failed allocating swap storage\n");
@@ -399,6 +402,7 @@ void ttm_tt_unpopulate(struct ttm_device *bdev, struct ttm_tt *ttm)
 
 	ttm->page_flags &= ~TTM_PAGE_FLAG_PRIV_POPULATED;
 }
+EXPORT_SYMBOL(ttm_tt_unpopulate);
 
 #ifdef CONFIG_DEBUG_FS
 
author reviewer
#33 lore raw

Re: [PATCH v2 09/15] drm/ttm, drm/amdgpu: Allow the driver some control over swapping

In reply to Thomas Hellström
Quoted 11 lines · Thomas Hellström

We are calling the eviction_valuable driver callback at eviction time to determine whether we actually can evict a buffer object.
The upcoming i915 TTM backend needs the same functionality for swapout, and that might actually be beneficial to other drivers as well.

Add an eviction_valuable call also in the swapout path. Try to keep the current behaviour for all drivers by returning true if the buffer object is already in the TTM_PL_SYSTEM placement. We change behaviour for the case where a buffer object is in a TT backed placement when swapped out, in which case the drivers normal eviction_valuable path is run.

Quoted 7 lines

Finally export ttm_tt_unpopulate() and don't swap out bos that are not populated. This allows a driver to purge a bo at swapout time if its content is no longer valuable rather than to have TTM swap the contents out.

Christian,

Here we have a ttm_tt_unpopulate() export as well at the end. I figure you will push back on that one. What we really need is a functionality to just drop the bo contents and end up in system memory unpopulated. Should I perhaps add a utility function to do that instead? like ttm_bo_purge()?

Thanks,

Thomas

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

maintainer
#34 lore raw

Re: [PATCH v2 09/15] drm/ttm, drm/amdgpu: Allow the driver some control over swapping

Am 18.05.21 um 17:15 schrieb Thomas Hellström:

Quoted 18 lines · Thomas Hellström
Quoted 16 lines

We are calling the eviction_valuable driver callback at eviction time to determine whether we actually can evict a buffer object.
The upcoming i915 TTM backend needs the same functionality for swapout, and that might actually be beneficial to other drivers as well.

Add an eviction_valuable call also in the swapout path. Try to keep the current behaviour for all drivers by returning true if the buffer object is already in the TTM_PL_SYSTEM placement. We change behaviour for the case where a buffer object is in a TT backed placement when swapped out, in which case the drivers normal eviction_valuable path is run.

Finally export ttm_tt_unpopulate() and don't swap out bos that are not populated. This allows a driver to purge a bo at swapout time if its content is no longer valuable rather than to have TTM swap the contents out.

Quoted 6 lines
Quoted 2 lines

Christian,

Here we have a ttm_tt_unpopulate() export as well at the end. I figure you will push back on that one. What we really need is a functionality to just drop the bo contents and end up in system memory unpopulated. Should I perhaps add a utility function to do that instead? like ttm_bo_purge()?

We already have that. Just call ttm_bo_validate() without any place to put the buffer.

See how ttm_bo_pipeline_gutting() is used.

Christian.

Quoted 6 lines

Thanks,

Thomas

author reviewer
#35 lore raw

Re: [PATCH v2 09/15] drm/ttm, drm/amdgpu: Allow the driver some control over swapping

Quoted 26 lines · Christian König

Am 18.05.21 um 17:15 schrieb Thomas Hellström:

Quoted 23 lines · Thomas Hellström
Quoted 11 lines

We are calling the eviction_valuable driver callback at eviction time to determine whether we actually can evict a buffer object.
The upcoming i915 TTM backend needs the same functionality for swapout, and that might actually be beneficial to other drivers as well.

Add an eviction_valuable call also in the swapout path. Try to keep the current behaviour for all drivers by returning true if the buffer object is already in the TTM_PL_SYSTEM placement. We change behaviour for the case where a buffer object is in a TT backed placement when swapped out, in which case the drivers normal eviction_valuable path is run.

Quoted 7 lines

Finally export ttm_tt_unpopulate() and don't swap out bos that are not populated. This allows a driver to purge a bo at swapout time if its content is no longer valuable rather than to have TTM swap the contents out.

Christian,

Quoted 7 lines
Quoted 1 line

Here we have a ttm_tt_unpopulate() export as well at the end. I figure you will push back on that one. What we really need is a functionality to just drop the bo contents and end up in system memory unpopulated. Should I perhaps add a utility function to do that instead? like ttm_bo_purge()?

We already have that. Just call ttm_bo_validate() without any place to put the buffer.

See how ttm_bo_pipeline_gutting() is used.

Christian.

OK, so is that reentrant from the move() or swap_notify() callback.

/Thomas

Quoted 8 lines
Quoted 7 lines

Thanks,

Thomas

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

maintainer
#36 lore raw

Re: [PATCH v2 09/15] drm/ttm, drm/amdgpu: Allow the driver some control over swapping

Am 18.05.21 um 17:20 schrieb Thomas Hellström:

Quoted 30 lines · Christian König
Quoted 28 lines

Am 18.05.21 um 17:15 schrieb Thomas Hellström:

Quoted 24 lines · Thomas Hellström
Quoted 11 lines

We are calling the eviction_valuable driver callback at eviction time to determine whether we actually can evict a buffer object.
The upcoming i915 TTM backend needs the same functionality for swapout, and that might actually be beneficial to other drivers as well.

Add an eviction_valuable call also in the swapout path. Try to keep the current behaviour for all drivers by returning true if the buffer object is already in the TTM_PL_SYSTEM placement. We change behaviour for the case where a buffer object is in a TT backed placement when swapped out, in which case the drivers normal eviction_valuable path is run.

Quoted 7 lines

Finally export ttm_tt_unpopulate() and don't swap out bos that are not populated. This allows a driver to purge a bo at swapout time if its content is no longer valuable rather than to have TTM swap the contents out.

Christian,

Here we have a ttm_tt_unpopulate() export as well at the end. I figure you will push back on that one. What we really need is a functionality to just drop the bo contents and end up in system memory unpopulated. Should I perhaps add a utility function to do that instead? like ttm_bo_purge()?

Quoted 7 lines
Quoted 5 lines

We already have that. Just call ttm_bo_validate() without any place to put the buffer.

See how ttm_bo_pipeline_gutting() is used.

Christian.

OK, so is that reentrant from the move() or swap_notify() callback.

That sounds like a design bug to me since you should never need to do this.

When you want to destroy the backing store of a buffer during eviction you should just do this by returning an empty placement from the evict_flags callback.

It is TTMs job to deal with the buffer placement and drivers are no longer allowed to mess with that.

Regards,
Christian.

Quoted 13 lines

/Thomas

Quoted 8 lines
Quoted 7 lines

Thanks,

Thomas

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

author reviewer
#37 lore raw

Re: [PATCH v2 09/15] drm/ttm, drm/amdgpu: Allow the driver some control over swapping

Quoted 35 lines · Christian König

Am 18.05.21 um 17:20 schrieb Thomas Hellström:

Quoted 34 lines · Christian König
Quoted 32 lines

Am 18.05.21 um 17:15 schrieb Thomas Hellström:

Quoted 18 lines · Thomas Hellström
Quoted 16 lines

We are calling the eviction_valuable driver callback at eviction time to determine whether we actually can evict a buffer object.
The upcoming i915 TTM backend needs the same functionality for swapout, and that might actually be beneficial to other drivers as well.

Add an eviction_valuable call also in the swapout path. Try to keep the current behaviour for all drivers by returning true if the buffer object is already in the TTM_PL_SYSTEM placement. We change behaviour for the case where a buffer object is in a TT backed placement when swapped out, in which case the drivers normal eviction_valuable path is run.

Finally export ttm_tt_unpopulate() and don't swap out bos that are not populated. This allows a driver to purge a bo at swapout time if its content is no longer valuable rather than to have TTM swap the contents out.

Quoted 6 lines
Quoted 2 lines

Christian,

Here we have a ttm_tt_unpopulate() export as well at the end. I figure you will push back on that one. What we really need is a functionality to just drop the bo contents and end up in system memory unpopulated. Should I perhaps add a utility function to do that instead? like ttm_bo_purge()?

We already have that. Just call ttm_bo_validate() without any place to put the buffer.

See how ttm_bo_pipeline_gutting() is used.

Quoted 7 lines
Quoted 3 lines
Quoted 1 line

Christian.

OK, so is that reentrant from the move() or swap_notify() callback.

That sounds like a design bug to me since you should never need to do this.

When you want to destroy the backing store of a buffer during eviction you should just do this by returning an empty placement from the evict_flags callback.

So this is for the functionality where the user has indicated that the contents is no longer of value, but the buffer itself
is cached until evicted or swapped out for performance reasons. So the above would work for eviction, but what about swapout. Could we add some similar functionality there?

/Thomas

Quoted 18 lines

Regards,
Christian.

Quoted 13 lines

/Thomas

Quoted 8 lines
Quoted 7 lines

Thanks,

Thomas

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

maintainer
#38 lore raw

Re: [PATCH v2 09/15] drm/ttm, drm/amdgpu: Allow the driver some control over swapping

Am 18.05.21 um 17:38 schrieb Thomas Hellström:

Quoted 39 lines · Christian König
Quoted 37 lines

Am 18.05.21 um 17:20 schrieb Thomas Hellström:

Quoted 36 lines · Christian König
Quoted 33 lines

Am 18.05.21 um 17:15 schrieb Thomas Hellström:

Quoted 18 lines · Thomas Hellström
Quoted 16 lines

We are calling the eviction_valuable driver callback at eviction time to determine whether we actually can evict a buffer object.
The upcoming i915 TTM backend needs the same functionality for swapout, and that might actually be beneficial to other drivers as well.

Add an eviction_valuable call also in the swapout path. Try to keep the current behaviour for all drivers by returning true if the buffer object is already in the TTM_PL_SYSTEM placement. We change behaviour for the case where a buffer object is in a TT backed placement when swapped out, in which case the drivers normal eviction_valuable path is run.

Finally export ttm_tt_unpopulate() and don't swap out bos that are not populated. This allows a driver to purge a bo at swapout time if its content is no longer valuable rather than to have TTM swap the contents out.

Quoted 6 lines
Quoted 2 lines

Christian,

Here we have a ttm_tt_unpopulate() export as well at the end. I figure you will push back on that one. What we really need is a functionality to just drop the bo contents and end up in system memory unpopulated. Should I perhaps add a utility function to do that instead? like ttm_bo_purge()?

We already have that. Just call ttm_bo_validate() without any place to put the buffer.

See how ttm_bo_pipeline_gutting() is used.

Christian.

Quoted 8 lines
Quoted 5 lines
Quoted 1 line

OK, so is that reentrant from the move() or swap_notify() callback.

That sounds like a design bug to me since you should never need to do this.

When you want to destroy the backing store of a buffer during eviction you should just do this by returning an empty placement from the evict_flags callback.

So this is for the functionality where the user has indicated that the contents is no longer of value, but the buffer itself
is cached until evicted or swapped out for performance reasons. So the above would work for eviction, but what about swapout. Could we add some similar functionality there?

Amdgpu has the same functionality and you don't need to handle swap at all.

Just return from the evict_flags that you want to drop the backing store as soon as the BO leaves the GTT domain.

Christian.

Quoted 21 lines

/Thomas

Quoted 18 lines

Regards,
Christian.

Quoted 13 lines

/Thomas

Quoted 8 lines
Quoted 7 lines

Thanks,

Thomas

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

author reviewer
#39 lore raw

Re: [PATCH v2 09/15] drm/ttm, drm/amdgpu: Allow the driver some control over swapping

Quoted 44 lines · Christian König

Am 18.05.21 um 17:38 schrieb Thomas Hellström:

Quoted 43 lines · Christian König
Quoted 41 lines

Am 18.05.21 um 17:20 schrieb Thomas Hellström:

Quoted 30 lines · Christian König
Quoted 28 lines

Am 18.05.21 um 17:15 schrieb Thomas Hellström:

Quoted 24 lines · Thomas Hellström
Quoted 11 lines

We are calling the eviction_valuable driver callback at eviction time to determine whether we actually can evict a buffer object.
The upcoming i915 TTM backend needs the same functionality for swapout, and that might actually be beneficial to other drivers as well.

Add an eviction_valuable call also in the swapout path. Try to keep the current behaviour for all drivers by returning true if the buffer object is already in the TTM_PL_SYSTEM placement. We change behaviour for the case where a buffer object is in a TT backed placement when swapped out, in which case the drivers normal eviction_valuable path is run.

Quoted 7 lines

Finally export ttm_tt_unpopulate() and don't swap out bos that are not populated. This allows a driver to purge a bo at swapout time if its content is no longer valuable rather than to have TTM swap the contents out.

Christian,

Here we have a ttm_tt_unpopulate() export as well at the end. I figure you will push back on that one. What we really need is a functionality to just drop the bo contents and end up in system memory unpopulated. Should I perhaps add a utility function to do that instead? like ttm_bo_purge()?

Quoted 7 lines
Quoted 5 lines

We already have that. Just call ttm_bo_validate() without any place to put the buffer.

See how ttm_bo_pipeline_gutting() is used.

Christian.

OK, so is that reentrant from the move() or swap_notify() callback.

That sounds like a design bug to me since you should never need to do this.

Quoted 8 lines
Quoted 4 lines
Quoted 1 line

When you want to destroy the backing store of a buffer during eviction you should just do this by returning an empty placement from the evict_flags callback.

So this is for the functionality where the user has indicated that the contents is no longer of value, but the buffer itself
is cached until evicted or swapped out for performance reasons. So the above would work for eviction, but what about swapout. Could we add some similar functionality there?

Amdgpu has the same functionality and you don't need to handle swap at all.

Just return from the evict_flags that you want to drop the backing store as soon as the BO leaves the GTT domain.

Hmm, the pipeline_gutting function seems ok, but overly complex if the bo is already idle, Am I allowed to optimize it slightly for the latter case?

/Thomas

Quoted 25 lines

Christian.

Quoted 21 lines

/Thomas

Quoted 18 lines

Regards,
Christian.

Quoted 13 lines

/Thomas

Quoted 8 lines
Quoted 7 lines

Thanks,

Thomas

maintainer
#40 lore raw

Re: [PATCH v2 09/15] drm/ttm, drm/amdgpu: Allow the driver some control over swapping

Am 18.05.21 um 18:07 schrieb Thomas Hellström:

Quoted 48 lines · Christian König
Quoted 46 lines

Am 18.05.21 um 17:38 schrieb Thomas Hellström:

Quoted 45 lines · Christian König
Quoted 42 lines

Am 18.05.21 um 17:20 schrieb Thomas Hellström:

Quoted 30 lines · Christian König
Quoted 28 lines

Am 18.05.21 um 17:15 schrieb Thomas Hellström:

Quoted 24 lines · Thomas Hellström
Quoted 11 lines

We are calling the eviction_valuable driver callback at eviction time to determine whether we actually can evict a buffer object.
The upcoming i915 TTM backend needs the same functionality for swapout, and that might actually be beneficial to other drivers as well.

Add an eviction_valuable call also in the swapout path. Try to keep the current behaviour for all drivers by returning true if the buffer object is already in the TTM_PL_SYSTEM placement. We change behaviour for the case where a buffer object is in a TT backed placement when swapped out, in which case the drivers normal eviction_valuable path is run.

Quoted 7 lines

Finally export ttm_tt_unpopulate() and don't swap out bos that are not populated. This allows a driver to purge a bo at swapout time if its content is no longer valuable rather than to have TTM swap the contents out.

Christian,

Here we have a ttm_tt_unpopulate() export as well at the end. I figure you will push back on that one. What we really need is a functionality to just drop the bo contents and end up in system memory unpopulated. Should I perhaps add a utility function to do that instead? like ttm_bo_purge()?

Quoted 7 lines
Quoted 5 lines

We already have that. Just call ttm_bo_validate() without any place to put the buffer.

See how ttm_bo_pipeline_gutting() is used.

Christian.

OK, so is that reentrant from the move() or swap_notify() callback.

That sounds like a design bug to me since you should never need to do this.

When you want to destroy the backing store of a buffer during eviction you should just do this by returning an empty placement from the evict_flags callback.

Quoted 8 lines
Quoted 6 lines
Quoted 2 lines

So this is for the functionality where the user has indicated that the contents is no longer of value, but the buffer itself
is cached until evicted or swapped out for performance reasons. So the above would work for eviction, but what about swapout. Could we add some similar functionality there?

Amdgpu has the same functionality and you don't need to handle swap at all.

Just return from the evict_flags that you want to drop the backing store as soon as the BO leaves the GTT domain.

Hmm, the pipeline_gutting function seems ok, but overly complex if the bo is already idle, Am I allowed to optimize it slightly for the latter case?

Yeah, sure. We just never hat that use case so far.

Christian.

Quoted 29 lines

/Thomas

Quoted 25 lines

Christian.

Quoted 21 lines

/Thomas

Quoted 18 lines

Regards,
Christian.

Quoted 13 lines

/Thomas

Quoted 8 lines
Quoted 7 lines

Thanks,

Thomas

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

author reviewer
#41 lore raw

Re: [PATCH v2 09/15] drm/ttm, drm/amdgpu: Allow the driver some control over swapping

Quoted 52 lines · Christian König

Am 18.05.21 um 18:07 schrieb Thomas Hellström:

Quoted 51 lines · Christian König
Quoted 49 lines

Am 18.05.21 um 17:38 schrieb Thomas Hellström:

Quoted 47 lines · Christian König
Quoted 35 lines

Am 18.05.21 um 17:20 schrieb Thomas Hellström:

Quoted 34 lines · Christian König
Quoted 32 lines

Am 18.05.21 um 17:15 schrieb Thomas Hellström:

Quoted 18 lines · Thomas Hellström
Quoted 16 lines

We are calling the eviction_valuable driver callback at eviction time to determine whether we actually can evict a buffer object.
The upcoming i915 TTM backend needs the same functionality for swapout, and that might actually be beneficial to other drivers as well.

Add an eviction_valuable call also in the swapout path. Try to keep the current behaviour for all drivers by returning true if the buffer object is already in the TTM_PL_SYSTEM placement. We change behaviour for the case where a buffer object is in a TT backed placement when swapped out, in which case the drivers normal eviction_valuable path is run.

Finally export ttm_tt_unpopulate() and don't swap out bos that are not populated. This allows a driver to purge a bo at swapout time if its content is no longer valuable rather than to have TTM swap the contents out.

Quoted 6 lines
Quoted 2 lines

Christian,

Here we have a ttm_tt_unpopulate() export as well at the end. I figure you will push back on that one. What we really need is a functionality to just drop the bo contents and end up in system memory unpopulated. Should I perhaps add a utility function to do that instead? like ttm_bo_purge()?

We already have that. Just call ttm_bo_validate() without any place to put the buffer.

See how ttm_bo_pipeline_gutting() is used.

Quoted 7 lines
Quoted 3 lines
Quoted 1 line

Christian.

OK, so is that reentrant from the move() or swap_notify() callback.

That sounds like a design bug to me since you should never need to do this.

When you want to destroy the backing store of a buffer during eviction you should just do this by returning an empty placement from the evict_flags callback.

So this is for the functionality where the user has indicated that the contents is no longer of value, but the buffer itself
is cached until evicted or swapped out for performance reasons. So the above would work for eviction, but what about swapout. Could we add some similar functionality there?

Quoted 7 lines
Quoted 5 lines
Quoted 3 lines

Amdgpu has the same functionality and you don't need to handle swap at all.

Just return from the evict_flags that you want to drop the backing store as soon as the BO leaves the GTT domain.

Hmm, the pipeline_gutting function seems ok, but overly complex if the bo is already idle, Am I allowed to optimize it slightly for the latter case?

Yeah, sure. We just never hat that use case so far.

One thing about the code here that makes me worried is that the "destination" ttm_tt is allocated *after* pipeline_gutting. We're not really allowed to fail here because that would leave the BO in a state where codepaths (fault for example) try to access a NULL ttm_tt. While the idle case can get away with ttm_tt_unpopulate, for the async case, ttm_tt really needs to be pre-allocated, so that we can leave the bo in a consistent state.

/Thomas

Quoted 33 lines

Christian.

Quoted 29 lines

/Thomas

Quoted 25 lines

Christian.

Quoted 21 lines

/Thomas

Quoted 18 lines

Regards,
Christian.

Quoted 13 lines

/Thomas

Quoted 8 lines
Quoted 7 lines

Thanks,

Thomas

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

maintainer
#42 lore raw

Re: [PATCH v2 09/15] drm/ttm, drm/amdgpu: Allow the driver some control over swapping

Am 19.05.21 um 08:27 schrieb Thomas Hellström:

Quoted 56 lines · Christian König
Quoted 54 lines

Am 18.05.21 um 18:07 schrieb Thomas Hellström:

Quoted 53 lines · Christian König
Quoted 51 lines

Am 18.05.21 um 17:38 schrieb Thomas Hellström:

Quoted 39 lines · Christian König
Quoted 37 lines

Am 18.05.21 um 17:20 schrieb Thomas Hellström:

Quoted 36 lines · Christian König
Quoted 33 lines

Am 18.05.21 um 17:15 schrieb Thomas Hellström:

Quoted 18 lines · Thomas Hellström
Quoted 16 lines

We are calling the eviction_valuable driver callback at eviction time to determine whether we actually can evict a buffer object.
The upcoming i915 TTM backend needs the same functionality for swapout, and that might actually be beneficial to other drivers as well.

Add an eviction_valuable call also in the swapout path. Try to keep the current behaviour for all drivers by returning true if the buffer object is already in the TTM_PL_SYSTEM placement. We change behaviour for the case where a buffer object is in a TT backed placement when swapped out, in which case the drivers normal eviction_valuable path is run.

Finally export ttm_tt_unpopulate() and don't swap out bos that are not populated. This allows a driver to purge a bo at swapout time if its content is no longer valuable rather than to have TTM swap the contents out.

Quoted 6 lines
Quoted 2 lines

Christian,

Here we have a ttm_tt_unpopulate() export as well at the end. I figure you will push back on that one. What we really need is a functionality to just drop the bo contents and end up in system memory unpopulated. Should I perhaps add a utility function to do that instead? like ttm_bo_purge()?

We already have that. Just call ttm_bo_validate() without any place to put the buffer.

See how ttm_bo_pipeline_gutting() is used.

Christian.

Quoted 8 lines
Quoted 5 lines
Quoted 1 line

OK, so is that reentrant from the move() or swap_notify() callback.

That sounds like a design bug to me since you should never need to do this.

When you want to destroy the backing store of a buffer during eviction you should just do this by returning an empty placement from the evict_flags callback.

So this is for the functionality where the user has indicated that the contents is no longer of value, but the buffer itself
is cached until evicted or swapped out for performance reasons. So the above would work for eviction, but what about swapout. Could we add some similar functionality there?

Amdgpu has the same functionality and you don't need to handle swap at all.

Quoted 7 lines
Quoted 5 lines
Quoted 3 lines
Quoted 1 line

Just return from the evict_flags that you want to drop the backing store as soon as the BO leaves the GTT domain.

Hmm, the pipeline_gutting function seems ok, but overly complex if the bo is already idle, Am I allowed to optimize it slightly for the latter case?

Yeah, sure. We just never hat that use case so far.

One thing about the code here that makes me worried is that the "destination" ttm_tt is allocated *after* pipeline_gutting. We're not really allowed to fail here because that would leave the BO in a state where codepaths (fault for example) try to access a NULL ttm_tt. While the idle case can get away with ttm_tt_unpopulate, for the async case, ttm_tt really needs to be pre-allocated, so that we can leave the bo in a consistent state.

Well the original plan was to make tt allocation purely optional.

But I didn't had the time so far to actually fix that.

Christian.

Quoted 37 lines

/Thomas

Quoted 33 lines

Christian.

Quoted 29 lines

/Thomas

Quoted 25 lines

Christian.

Quoted 21 lines

/Thomas

Quoted 18 lines

Regards,
Christian.

Quoted 13 lines

/Thomas

Quoted 8 lines
Quoted 7 lines

Thanks,

Thomas

author reviewer
#43 lore raw

PATCH v2 10/15 drm/i915/ttm: Introduce a TTM i915 gem object backend

In reply to Thomas Hellström

Most logical place to introduce TTM buffer objects is as an i915 gem object backend. We need to add some ops to account for added functionality like delayed delete and LRU list manipulation.

Initially we support only LMEM and SYSTEM memory, but SYSTEM (which in this case means evicted LMEM objects) is not visible to i915 GEM yet. The plan is to move the i915 gem system region over to the TTM system memory type in upcoming patches.

We set up GPU bindings directly both from LMEM and from the system region, as there is no need to use the legacy TTM_TT memory type. We reserve that for future porting of GGTT bindings to TTM.

Remove the old lmem backend.

---
v2:
- Break out needed TTM functionality to a separate patch (Reported by Christian König).
- Fix an unhandled error (Reported by Matthew Auld and Maarten Lankhorst)
- Remove a stray leftover sg_table allocation (Reported by Matthew Auld)
- Use ttm_tt_unpopulate() rather than ttm_tt_destroy() in the purge path as some TTM functionality relies on having a ttm_tt present for !is_iomem.

---
 drivers/gpu/drm/i915/Makefile                 |   1 +
 drivers/gpu/drm/i915/gem/i915_gem_lmem.c      |  84 ---
 drivers/gpu/drm/i915/gem/i915_gem_lmem.h      |   5 -
 drivers/gpu/drm/i915/gem/i915_gem_object.c    | 125 +++--
 drivers/gpu/drm/i915/gem/i915_gem_object.h    |   9 +
 .../gpu/drm/i915/gem/i915_gem_object_types.h  |  18 +
 drivers/gpu/drm/i915/gem/i915_gem_region.c    |   6 +-
 drivers/gpu/drm/i915/gem/i915_gem_ttm.c       | 519 ++++++++++++++++++
 drivers/gpu/drm/i915/gem/i915_gem_ttm.h       |  48 ++
 drivers/gpu/drm/i915/gt/intel_region_lmem.c   |   3 +-
 drivers/gpu/drm/i915/i915_gem.c               |   5 +-
 drivers/gpu/drm/i915/intel_memory_region.c    |   1 -
 drivers/gpu/drm/i915/intel_memory_region.h    |   1 -
 drivers/gpu/drm/i915/intel_region_ttm.c       |   5 +-
 drivers/gpu/drm/i915/intel_region_ttm.h       |   7 +-
 15 files changed, 696 insertions(+), 141 deletions(-)
 create mode 100644 drivers/gpu/drm/i915/gem/i915_gem_ttm.c
 create mode 100644 drivers/gpu/drm/i915/gem/i915_gem_ttm.h
drivers/gpu/drm/i915/Makefile +1 -0
index 958ccc1edfed..ef0d884a9e2d 100644
@@ -155,6 +155,7 @@ gem-y += \
 	gem/i915_gem_stolen.o \
 	gem/i915_gem_throttle.o \
 	gem/i915_gem_tiling.o \
+	gem/i915_gem_ttm.o \
 	gem/i915_gem_ttm_bo_util.o \
 	gem/i915_gem_userptr.o \
 	gem/i915_gem_wait.o \
drivers/gpu/drm/i915/gem/i915_gem_lmem.c +0 -84

97 lines not on this page · raw message

drivers/gpu/drm/i915/gem/i915_gem_lmem.h +0 -5
index fac6bc5a5ebb..ea76fd11ccb0 100644
@@ -26,9 +26,4 @@ i915_gem_object_create_lmem(struct drm_i915_private *i915,
 			    resource_size_t size,
 			    unsigned int flags);
 
-int __i915_gem_lmem_object_init(struct intel_memory_region *mem,
-				struct drm_i915_gem_object *obj,
-				resource_size_t size,
-				unsigned int flags);
-
 #endif /* !__I915_GEM_LMEM_H */
drivers/gpu/drm/i915/gem/i915_gem_object.c +86 -39

164 lines not on this page · raw message

drivers/gpu/drm/i915/gem/i915_gem_object.h +9 -0
index 2ebd79537aea..ae5930e307d5 100644
@@ -200,6 +200,9 @@ static inline bool i915_gem_object_trylock(struct drm_i915_gem_object *obj)
 
 static inline void i915_gem_object_unlock(struct drm_i915_gem_object *obj)
 {
+	if (obj->ops->adjust_lru)
+		obj->ops->adjust_lru(obj);
+
 	dma_resv_unlock(obj->base.resv);
 }
 
@@ -587,6 +590,12 @@ int i915_gem_object_read_from_page(struct drm_i915_gem_object *obj, u64 offset,
 
 bool i915_gem_object_is_shmem(const struct drm_i915_gem_object *obj);
 
+void __i915_gem_free_object_rcu(struct rcu_head *head);
+
+void __i915_gem_free_object(struct drm_i915_gem_object *obj);
+
+bool i915_gem_object_evictable(struct drm_i915_gem_object *obj);
+
 #ifdef CONFIG_MMU_NOTIFIER
 static inline bool
 i915_gem_object_is_userptr(struct drm_i915_gem_object *obj)
drivers/gpu/drm/i915/gem/i915_gem_object_types.h +18 -0
index 98f69d8fd37d..b350765e1935 100644
@@ -63,6 +63,20 @@ struct drm_i915_gem_object_ops {
 		      const struct drm_i915_gem_pwrite *arg);
 
 	int (*dmabuf_export)(struct drm_i915_gem_object *obj);
+
+	/**
+	 * adjust_lru - notify that the madvise value was updated
+	 * @obj: The gem object
+	 *
+	 * The madvise value may have been updated, or object was recently
+	 * referenced so act accordingly (Perhaps changing an LRU list etc).
+	 */
+	void (*adjust_lru)(struct drm_i915_gem_object *obj);
+
+	/**
+	 * delayed_free - Override the default delayed free implementation
+	 */
+	void (*delayed_free)(struct drm_i915_gem_object *obj);
 	void (*release)(struct drm_i915_gem_object *obj);
 
 	const char *name; /* friendly name for debug, e.g. lockdep classes */
@@ -307,6 +321,10 @@ struct drm_i915_gem_object {
 		bool dirty:1;
 	} mm;
 
+	struct {
+		struct sg_table *cached_io_st;
+	} ttm;
+
 	/** Record of address bit 17 of each page at last unbind. */
 	unsigned long *bit_17;
 
drivers/gpu/drm/i915/gem/i915_gem_region.c +1 -5
index f25e6646c5b7..d1f1840540dd 100644
@@ -18,11 +18,7 @@ void i915_gem_object_init_memory_region(struct drm_i915_gem_object *obj,
 
 	mutex_lock(&mem->objects.lock);
 
-	if (obj->flags & I915_BO_ALLOC_VOLATILE)
-		list_add(&obj->mm.region_link, &mem->objects.purgeable);
-	else
-		list_add(&obj->mm.region_link, &mem->objects.list);
-
+	list_add(&obj->mm.region_link, &mem->objects.list);
 	mutex_unlock(&mem->objects.lock);
 }
 
drivers/gpu/drm/i915/gem/i915_gem_ttm.c new file +519 -0

519 lines not on this page · raw message

drivers/gpu/drm/i915/gem/i915_gem_ttm.h new file +48 -0

48 lines not on this page · raw message

drivers/gpu/drm/i915/gt/intel_region_lmem.c +2 -1
index f7366b054f8e..4ae1f717a94c 100644
@@ -9,6 +9,7 @@
 #include "intel_region_ttm.h"
 #include "gem/i915_gem_lmem.h"
 #include "gem/i915_gem_region.h"
+#include "gem/i915_gem_ttm.h"
 #include "intel_region_lmem.h"
 
 static int init_fake_lmem_bar(struct intel_memory_region *mem)
@@ -107,7 +108,7 @@ region_lmem_init(struct intel_memory_region *mem)
 static const struct intel_memory_region_ops intel_region_lmem_ops = {
 	.init = region_lmem_init,
 	.release = region_lmem_release,
-	.init_object = __i915_gem_lmem_object_init,
+	.init_object = __i915_gem_ttm_object_init,
 };
 
 struct intel_memory_region *
drivers/gpu/drm/i915/i915_gem.c +4 -1
index 180f6e9107d4..350283ab9a83 100644
@@ -1005,8 +1005,11 @@ i915_gem_madvise_ioctl(struct drm_device *dev, void *data,
 		}
 	}
 
-	if (obj->mm.madv != __I915_MADV_PURGED)
+	if (obj->mm.madv != __I915_MADV_PURGED) {
 		obj->mm.madv = args->madv;
+		if (obj->ops->adjust_lru)
+			obj->ops->adjust_lru(obj);
+	}
 
 	if (i915_gem_object_has_pages(obj)) {
 		struct list_head *list;
drivers/gpu/drm/i915/intel_memory_region.c +0 -1
index 4092cc987679..bd27e897d4d0 100644
@@ -149,7 +149,6 @@ intel_memory_region_create(struct drm_i915_private *i915,
 
 	mutex_init(&mem->objects.lock);
 	INIT_LIST_HEAD(&mem->objects.list);
-	INIT_LIST_HEAD(&mem->objects.purgeable);
 	INIT_LIST_HEAD(&mem->reserved);
 
 	mutex_init(&mem->mm_lock);
drivers/gpu/drm/i915/intel_memory_region.h +0 -1
index e69cde13daf2..7b5fa97c0b59 100644
@@ -100,7 +100,6 @@ struct intel_memory_region {
 	struct {
 		struct mutex lock; /* Protects access to objects */
 		struct list_head list;
-		struct list_head purgeable;
 	} objects;
 
 	size_t chunk_size;
drivers/gpu/drm/i915/intel_region_ttm.c +3 -2
index 9fb447e19ebd..e429c9a38371 100644
@@ -10,8 +10,7 @@
 
 #include "intel_region_ttm.h"
 
-/* A Zero-initialized driver for now. We don't have a TTM backend yet. */
-static struct ttm_device_funcs i915_ttm_bo_driver;
+extern struct ttm_device_funcs i915_ttm_bo_driver;
 
 /**
  * DOC: TTM support structure
@@ -197,6 +196,7 @@ struct sg_table *intel_region_ttm_node_to_st(struct intel_memory_region *mem,
 	return i915_sg_from_mm_node(node, mem->region.start);
 }
 
+#ifdef CONFIG_DRM_I915_SELFTEST
 /**
  * intel_region_ttm_node_alloc - Allocate memory resources from a region
  * @mem: The memory region,
@@ -243,3 +243,4 @@ void *intel_region_ttm_node_alloc(struct intel_memory_region *mem,
 		ret = -ENXIO;
 	return ret ? ERR_PTR(ret) : res.mm_node;
 }
+#endif
drivers/gpu/drm/i915/intel_region_ttm.h +5 -2
index 40129f7e0317..9a5b0437d73f 100644
@@ -21,9 +21,12 @@ void intel_region_ttm_fini(struct intel_memory_region *mem);
 struct sg_table *intel_region_ttm_node_to_st(struct intel_memory_region *mem,
 					     void *node);
 
+void intel_region_ttm_node_free(struct intel_memory_region *mem,
+				void *node);
+
+#ifdef CONFIG_DRM_I915_SELFTEST
 void *intel_region_ttm_node_alloc(struct intel_memory_region *mem,
 				  resource_size_t size,
 				  unsigned int flags);
-void intel_region_ttm_node_free(struct intel_memory_region *mem,
-				void *node);
+#endif
 #endif /* _INTEL_REGION_TTM_H_ */
Signature · 6 lines
2.31.1

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Reviewed-by on v2 10/15
#44 lore raw

Re: [Intel-gfx] [PATCH v2 10/15] drm/i915/ttm: Introduce a TTM i915 gem object backend

Quoted 326 lines · Thomas Hellström

Most logical place to introduce TTM buffer objects is as an i915 gem object backend. We need to add some ops to account for added functionality like delayed delete and LRU list manipulation.

Initially we support only LMEM and SYSTEM memory, but SYSTEM (which in this case means evicted LMEM objects) is not visible to i915 GEM yet. The plan is to move the i915 gem system region over to the TTM system memory type in upcoming patches.

We set up GPU bindings directly both from LMEM and from the system region, as there is no need to use the legacy TTM_TT memory type. We reserve that for future porting of GGTT bindings to TTM.

Remove the old lmem backend.

---
v2:
- Break out needed TTM functionality to a separate patch (Reported by Christian König).
- Fix an unhandled error (Reported by Matthew Auld and Maarten Lankhorst)
- Remove a stray leftover sg_table allocation (Reported by Matthew Auld)
- Use ttm_tt_unpopulate() rather than ttm_tt_destroy() in the purge path as some TTM functionality relies on having a ttm_tt present for !is_iomem.

---
 drivers/gpu/drm/i915/Makefile                 |   1 +
 drivers/gpu/drm/i915/gem/i915_gem_lmem.c      |  84 ---
 drivers/gpu/drm/i915/gem/i915_gem_lmem.h      |   5 -
 drivers/gpu/drm/i915/gem/i915_gem_object.c    | 125 +++--
 drivers/gpu/drm/i915/gem/i915_gem_object.h    |   9 +
 .../gpu/drm/i915/gem/i915_gem_object_types.h  |  18 +
 drivers/gpu/drm/i915/gem/i915_gem_region.c    |   6 +-
 drivers/gpu/drm/i915/gem/i915_gem_ttm.c       | 519 ++++++++++++++++++
 drivers/gpu/drm/i915/gem/i915_gem_ttm.h       |  48 ++
 drivers/gpu/drm/i915/gt/intel_region_lmem.c   |   3 +-
 drivers/gpu/drm/i915/i915_gem.c               |   5 +-
 drivers/gpu/drm/i915/intel_memory_region.c    |   1 -
 drivers/gpu/drm/i915/intel_memory_region.h    |   1 -
 drivers/gpu/drm/i915/intel_region_ttm.c       |   5 +-
 drivers/gpu/drm/i915/intel_region_ttm.h       |   7 +-
 15 files changed, 696 insertions(+), 141 deletions(-)
 create mode 100644 drivers/gpu/drm/i915/gem/i915_gem_ttm.c
 create mode 100644 drivers/gpu/drm/i915/gem/i915_gem_ttm.h
drivers/gpu/drm/i915/Makefile +1 -0
index 958ccc1edfed..ef0d884a9e2d 100644
@@ -155,6 +155,7 @@ gem-y += \
        gem/i915_gem_stolen.o \
        gem/i915_gem_throttle.o \
        gem/i915_gem_tiling.o \
+       gem/i915_gem_ttm.o \
        gem/i915_gem_ttm_bo_util.o \
        gem/i915_gem_userptr.o \
        gem/i915_gem_wait.o \
drivers/gpu/drm/i915/gem/i915_gem_lmem.c +0 -84
index 3b4aa28a076d..2b8cd15de1d9 100644
@@ -4,74 +4,10 @@
  */
 
 #include "intel_memory_region.h"
-#include "intel_region_ttm.h"
 #include "gem/i915_gem_region.h"
 #include "gem/i915_gem_lmem.h"
 #include "i915_drv.h"
 
-static void lmem_put_pages(struct drm_i915_gem_object *obj,
-                          struct sg_table *pages)
-{
-       intel_region_ttm_node_free(obj->mm.region, obj->mm.st_mm_node);
-       obj->mm.dirty = false;
-       sg_free_table(pages);
-       kfree(pages);
-}
-
-static int lmem_get_pages(struct drm_i915_gem_object *obj)
-{
-       unsigned int flags;
-       struct sg_table *pages;
-
-       flags = I915_ALLOC_MIN_PAGE_SIZE;
-       if (obj->flags & I915_BO_ALLOC_CONTIGUOUS)
-               flags |= I915_ALLOC_CONTIGUOUS;
-
-       obj->mm.st_mm_node = intel_region_ttm_node_alloc(obj->mm.region,
-                                                        obj->base.size,
-                                                        flags);
-       if (IS_ERR(obj->mm.st_mm_node))
-               return PTR_ERR(obj->mm.st_mm_node);
-
-       /* Range manager is always contigous */
-       if (obj->mm.region->is_range_manager)
-               obj->flags |= I915_BO_ALLOC_CONTIGUOUS;
-       pages = intel_region_ttm_node_to_st(obj->mm.region, obj->mm.st_mm_node);
-       if (IS_ERR(pages)) {
-               intel_region_ttm_node_free(obj->mm.region, obj->mm.st_mm_node);
-               return PTR_ERR(pages);
-       }
-
-       __i915_gem_object_set_pages(obj, pages, i915_sg_dma_sizes(pages->sgl));
-
-       if (obj->flags & I915_BO_ALLOC_CPU_CLEAR) {
-               void __iomem *vaddr =
-                       i915_gem_object_lmem_io_map(obj, 0, obj->base.size);
-
-               if (!vaddr) {
-                       struct sg_table *pages =
-                               __i915_gem_object_unset_pages(obj);
-
-                       if (!IS_ERR_OR_NULL(pages))
-                               lmem_put_pages(obj, pages);
-               }
-
-               memset_io(vaddr, 0, obj->base.size);
-               io_mapping_unmap(vaddr);
-       }
-
-       return 0;
-}
-
-const struct drm_i915_gem_object_ops i915_gem_lmem_obj_ops = {
-       .name = "i915_gem_object_lmem",
-       .flags = I915_GEM_OBJECT_HAS_IOMEM,
-
-       .get_pages = lmem_get_pages,
-       .put_pages = lmem_put_pages,
-       .release = i915_gem_object_release_memory_region,
-};
-
 void __iomem *
 i915_gem_object_lmem_io_map(struct drm_i915_gem_object *obj,
                            unsigned long n,
@@ -103,23 +39,3 @@ i915_gem_object_create_lmem(struct drm_i915_private *i915,
        return i915_gem_object_create_region(i915->mm.regions[INTEL_REGION_LMEM],
                                             size, flags);
 }
-
-int __i915_gem_lmem_object_init(struct intel_memory_region *mem,
-                               struct drm_i915_gem_object *obj,
-                               resource_size_t size,
-                               unsigned int flags)
-{
-       static struct lock_class_key lock_class;
-       struct drm_i915_private *i915 = mem->i915;
-
-       drm_gem_private_object_init(&i915->drm, &obj->base, size);
-       i915_gem_object_init(obj, &i915_gem_lmem_obj_ops, &lock_class, flags);
-
-       obj->read_domains = I915_GEM_DOMAIN_WC | I915_GEM_DOMAIN_GTT;
-
-       i915_gem_object_set_cache_coherency(obj, I915_CACHE_NONE);
-
-       i915_gem_object_init_memory_region(obj, mem);
-
-       return 0;
-}
drivers/gpu/drm/i915/gem/i915_gem_lmem.h +0 -5
index fac6bc5a5ebb..ea76fd11ccb0 100644
@@ -26,9 +26,4 @@ i915_gem_object_create_lmem(struct drm_i915_private *i915,
                            resource_size_t size,
                            unsigned int flags);
 
-int __i915_gem_lmem_object_init(struct intel_memory_region *mem,
-                               struct drm_i915_gem_object *obj,
-                               resource_size_t size,
-                               unsigned int flags);
-
 #endif /* !__I915_GEM_LMEM_H */
drivers/gpu/drm/i915/gem/i915_gem_object.c +69 -39
index c8953e3f5c70..8580996107ce 100644
@@ -172,7 +172,7 @@ static void i915_gem_close_object(struct drm_gem_object *gem, struct drm_file *f
        }
 }
 
-static void __i915_gem_free_object_rcu(struct rcu_head *head)
+void __i915_gem_free_object_rcu(struct rcu_head *head)
 {
        struct drm_i915_gem_object *obj =
                container_of(head, typeof(*obj), rcu);
@@ -208,59 +208,69 @@ static void __i915_gem_object_free_mmaps(struct drm_i915_gem_object *obj)
        }
 }
 
-static void __i915_gem_free_objects(struct drm_i915_private *i915,
-                                   struct llist_node *freed)
+void __i915_gem_free_object(struct drm_i915_gem_object *obj)
 {
-       struct drm_i915_gem_object *obj, *on;
+       trace_i915_gem_object_destroy(obj);
 
-       llist_for_each_entry_safe(obj, on, freed, freed) {
-               trace_i915_gem_object_destroy(obj);
+       if (!list_empty(&obj->vma.list)) {
+               struct i915_vma *vma;
+
+               /*
+                * Note that the vma keeps an object reference while
+                * it is active, so it *should* not sleep while we
+                * destroy it. Our debug code errs insits it *might*.
+                * For the moment, play along.
+                */
+               spin_lock(&obj->vma.lock);
+               while ((vma = list_first_entry_or_null(&obj->vma.list,
+                                                      struct i915_vma,
+                                                      obj_link))) {
+                       GEM_BUG_ON(vma->obj != obj);
+                       spin_unlock(&obj->vma.lock);
 
-               if (!list_empty(&obj->vma.list)) {
-                       struct i915_vma *vma;
+                       __i915_vma_put(vma);
 
-                       /*
-                        * Note that the vma keeps an object reference while
-                        * it is active, so it *should* not sleep while we
-                        * destroy it. Our debug code errs insits it *might*.
-                        * For the moment, play along.
-                        */
                        spin_lock(&obj->vma.lock);
-                       while ((vma = list_first_entry_or_null(&obj->vma.list,
-                                                              struct i915_vma,
-                                                              obj_link))) {
-                               GEM_BUG_ON(vma->obj != obj);
-                               spin_unlock(&obj->vma.lock);
+               }
+               spin_unlock(&obj->vma.lock);
+       }
 
-                               __i915_vma_put(vma);
+       __i915_gem_object_free_mmaps(obj);
 
-                               spin_lock(&obj->vma.lock);
-                       }
-                       spin_unlock(&obj->vma.lock);
-               }
+       GEM_BUG_ON(!list_empty(&obj->lut_list));
 
-               __i915_gem_object_free_mmaps(obj);
+       atomic_set(&obj->mm.pages_pin_count, 0);
+       __i915_gem_object_put_pages(obj);
+       GEM_BUG_ON(i915_gem_object_has_pages(obj));
+       bitmap_free(obj->bit_17);
 
-               GEM_BUG_ON(!list_empty(&obj->lut_list));
+       if (obj->base.import_attach)
+               drm_prime_gem_destroy(&obj->base, NULL);
 
-               atomic_set(&obj->mm.pages_pin_count, 0);
-               __i915_gem_object_put_pages(obj);
-               GEM_BUG_ON(i915_gem_object_has_pages(obj));
-               bitmap_free(obj->bit_17);
+       drm_gem_free_mmap_offset(&obj->base);
 
-               if (obj->base.import_attach)
-                       drm_prime_gem_destroy(&obj->base, NULL);
+       if (obj->ops->release)
+               obj->ops->release(obj);
 
-               drm_gem_free_mmap_offset(&obj->base);
+       if (obj->mm.n_placements > 1)
+               kfree(obj->mm.placements);
 
-               if (obj->ops->release)
-                       obj->ops->release(obj);
+       if (obj->resv_shared_from)
+               i915_vm_resv_put(obj->resv_shared_from);
+}
 
-               if (obj->mm.n_placements > 1)
-                       kfree(obj->mm.placements);
+static void __i915_gem_free_objects(struct drm_i915_private *i915,
+                                   struct llist_node *freed)
+{
+       struct drm_i915_gem_object *obj, *on;
 
-               if (obj->resv_shared_from)
-                       i915_vm_resv_put(obj->resv_shared_from);
+       llist_for_each_entry_safe(obj, on, freed, freed) {
+               might_sleep();
+               if (obj->ops->delayed_free) {
+                       obj->ops->delayed_free(obj);
+                       continue;
+               }
+               __i915_gem_free_object(obj);
 
                /* But keep the pointer alive for RCU-protected lookups */
                call_rcu(&obj->rcu, __i915_gem_free_object_rcu);
@@ -318,6 +328,7 @@ static void i915_gem_free_object(struct drm_gem_object *gem_obj)
         * worker and performing frees directly from subsequent allocations for
         * crude but effective memory throttling.
         */
+
        if (llist_add(&obj->freed, &i915->mm.free_list))
                queue_work(i915->wq, &i915->mm.free_work);
 }
@@ -410,6 +421,42 @@ int i915_gem_object_read_from_page(struct drm_i915_gem_object *obj, u64 offset,
        return 0;
 }
 
+/**
+ * i915_gem_object_evictable - Whether object is likely evictable after unbind.
+ * @obj: The object to check
+ *
+ * This function checks whether the object is likely unvictable after unbind.
+ * If the object is not locked when checking, the result is only advisory.
+ * If the object is locked when checking, and the function returns true,
+ * then an eviction should indeed be possible. But since unlocked vma
+ * unpinning and unbinding is currently possible, the object can actually
+ * become evictable even if this function returns false.
+ *
+ * Return: true if the object may be evictable. False otherwise.
+ */
+bool i915_gem_object_evictable(struct drm_i915_gem_object *obj)
+{
+       struct i915_vma *vma;
+       int pin_count = atomic_read(&obj->mm.pages_pin_count);
+
+       if (!pin_count)

drivers/gpu/drm/i915/gem/i915_gem_object.c int i915_gem_object_read_from_page(struct drm_i915_gem_object *obj, u64 offset, · line 452

Quoted 10 lines
drivers/gpu/drm/i915/gem/i915_gem_object.c +10 -0
+               return true;
+
+       spin_lock(&obj->vma.lock);
+       list_for_each_entry(vma, &obj->vma.list, obj_link) {
+               if (i915_vma_is_pinned(vma)) {
+                       spin_unlock(&obj->vma.lock);
+                       return false;
+               }
+               if (atomic_read(&vma->pages_count))
+                       pin_count--;

Can't pages_count be > 1, which would also be reflected in pages_pin_count? The vma_pin path looks very complex.

Quoted 15 lines
drivers/gpu/drm/i915/gem/i915_gem_object.c +7 -0
+       }
+       spin_unlock(&obj->vma.lock);
+       GEM_WARN_ON(pin_count < 0);
+
+       return pin_count == 0;
+}
+
 void i915_gem_init__objects(struct drm_i915_private *i915)
 {
        INIT_WORK(&i915->mm.free_work, __i915_gem_free_work);
drivers/gpu/drm/i915/gem/i915_gem_object.h +0 -0
index 2ebd79537aea..ae5930e307d5 100644

drivers/gpu/drm/i915/gem/i915_gem_object.h static inline bool i915_gem_object_trylock(struct drm_i915_gem_object *obj) · line 204

Quoted 5 lines
drivers/gpu/drm/i915/gem/i915_gem_object.h +2 -0
@@ -200,6 +200,9 @@ static inline bool i915_gem_object_trylock(struct drm_i915_gem_object *obj)
 
 static inline void i915_gem_object_unlock(struct drm_i915_gem_object *obj)
 {
+       if (obj->ops->adjust_lru)
+               obj->ops->adjust_lru(obj);

Interesting, so we bump the lru even when we just drop the lock?

Quoted 105 lines
drivers/gpu/drm/i915/gem/i915_gem_object.h +7 -0
+
        dma_resv_unlock(obj->base.resv);
 }
 
@@ -587,6 +590,12 @@ int i915_gem_object_read_from_page(struct drm_i915_gem_object *obj, u64 offset,
 
 bool i915_gem_object_is_shmem(const struct drm_i915_gem_object *obj);
 
+void __i915_gem_free_object_rcu(struct rcu_head *head);
+
+void __i915_gem_free_object(struct drm_i915_gem_object *obj);
+
+bool i915_gem_object_evictable(struct drm_i915_gem_object *obj);
+
 #ifdef CONFIG_MMU_NOTIFIER
 static inline bool
 i915_gem_object_is_userptr(struct drm_i915_gem_object *obj)
drivers/gpu/drm/i915/gem/i915_gem_object_types.h +18 -0
index 98f69d8fd37d..b350765e1935 100644
@@ -63,6 +63,20 @@ struct drm_i915_gem_object_ops {
                      const struct drm_i915_gem_pwrite *arg);
 
        int (*dmabuf_export)(struct drm_i915_gem_object *obj);
+
+       /**
+        * adjust_lru - notify that the madvise value was updated
+        * @obj: The gem object
+        *
+        * The madvise value may have been updated, or object was recently
+        * referenced so act accordingly (Perhaps changing an LRU list etc).
+        */
+       void (*adjust_lru)(struct drm_i915_gem_object *obj);
+
+       /**
+        * delayed_free - Override the default delayed free implementation
+        */
+       void (*delayed_free)(struct drm_i915_gem_object *obj);
        void (*release)(struct drm_i915_gem_object *obj);
 
        const char *name; /* friendly name for debug, e.g. lockdep classes */
@@ -307,6 +321,10 @@ struct drm_i915_gem_object {
                bool dirty:1;
        } mm;
 
+       struct {
+               struct sg_table *cached_io_st;
+       } ttm;
+
        /** Record of address bit 17 of each page at last unbind. */
        unsigned long *bit_17;
 
drivers/gpu/drm/i915/gem/i915_gem_region.c +1 -5
index f25e6646c5b7..d1f1840540dd 100644
@@ -18,11 +18,7 @@ void i915_gem_object_init_memory_region(struct drm_i915_gem_object *obj,
 
        mutex_lock(&mem->objects.lock);
 
-       if (obj->flags & I915_BO_ALLOC_VOLATILE)
-               list_add(&obj->mm.region_link, &mem->objects.purgeable);
-       else
-               list_add(&obj->mm.region_link, &mem->objects.list);
-
+       list_add(&obj->mm.region_link, &mem->objects.list);
        mutex_unlock(&mem->objects.lock);
 }
 
drivers/gpu/drm/i915/gem/i915_gem_ttm.c new file +29 -0
new file mode 100644
index 000000000000..790f5ec45c4d
@@ -0,0 +1,519 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2021 Intel Corporation
+ */
+
+#include <drm/ttm/ttm_bo_driver.h>
+#include <drm/ttm/ttm_placement.h>
+
+#include "i915_drv.h"
+#include "intel_memory_region.h"
+#include "intel_region_ttm.h"
+
+#include "gem/i915_gem_object.h"
+#include "gem/i915_gem_region.h"
+#include "gem/i915_gem_ttm.h"
+#include "gem/i915_gem_ttm_bo_util.h"
+
+#define I915_PL_LMEM0 TTM_PL_PRIV
+#define I915_PL_SYSTEM TTM_PL_SYSTEM
+#define I915_PL_STOLEN TTM_PL_VRAM
+#define I915_PL_GGTT TTM_PL_TT
+
+#define I915_TTM_PRIO_PURGE     0
+#define I915_TTM_PRIO_NO_PAGES  1
+#define I915_TTM_PRIO_HAS_PAGES 2
+
+/**
+ * struct i915_ttm_tt - TTM page vector with additional private information
+ * @ttm: The base TTM page vector.

drivers/gpu/drm/i915/gem/i915_gem_ttm.c line 39

Quoted 10 lines
drivers/gpu/drm/i915/gem/i915_gem_ttm.c new file +10 -0
+ * @dev: The struct device used for dma mapping and unmapping.
+ * @cached_st: The cached scatter-gather table.
+ *
+ * Note that DMA may be going on right up to the point where the page-
+ * vector is unpopulated in delayed destroy. Hence keep the
+ * scatter-gather table mapped and cached up to that point. This is
+ * different from the cached gem object io scatter-gather table which
+ * doesn't have an associated dma mapping.
+ */
+struct i915_ttm_tt {

What is the _tt here btw? Translation table? We also have use_tt elsewhere. ttm_tt looks like it just holds an array of pages, and associated data? ttm_pv?

Quoted 99 lines
drivers/gpu/drm/i915/gem/i915_gem_ttm.c new file +99 -0
+       struct ttm_tt ttm;
+       struct device *dev;
+       struct sg_table *cached_st;
+};
+
+static const struct ttm_place lmem0_sys_placement_flags[] = {
+       {
+               .fpfn = 0,
+               .lpfn = 0,
+               .mem_type = I915_PL_LMEM0,
+               .flags = 0,
+       }, {
+               .fpfn = 0,
+               .lpfn = 0,
+               .mem_type = I915_PL_SYSTEM,
+               .flags = 0,
+       }
+};
+
+struct ttm_placement i915_lmem0_placement = {
+       .num_placement = 1,
+       .placement = &lmem0_sys_placement_flags[0],
+       .num_busy_placement = 1,
+       .busy_placement = &lmem0_sys_placement_flags[0],
+};
+
+struct ttm_placement i915_lmem0_sys_placement = {
+       .num_placement = 1,
+       .placement = &lmem0_sys_placement_flags[0],
+       .num_busy_placement = 2,
+       .busy_placement = &lmem0_sys_placement_flags[0],
+};
+
+struct ttm_placement i915_sys_placement = {
+       .num_placement = 1,
+       .placement = &lmem0_sys_placement_flags[1],
+       .num_busy_placement = 1,
+       .busy_placement = &lmem0_sys_placement_flags[1],
+};
+
+static void i915_ttm_adjust_lru(struct drm_i915_gem_object *obj);
+
+static struct ttm_tt *i915_ttm_tt_create(struct ttm_buffer_object *bo,
+                                        uint32_t page_flags)
+{
+       struct ttm_resource_manager *man =
+               ttm_manager_type(bo->bdev, bo->mem.mem_type);
+       struct drm_i915_gem_object *obj = i915_ttm_to_gem(bo);
+       struct i915_ttm_tt *i915_tt;
+       int ret;
+
+       i915_tt = kzalloc(sizeof(*i915_tt), GFP_KERNEL);
+       if (!i915_tt)
+               return NULL;
+
+       if (obj->flags & I915_BO_ALLOC_CPU_CLEAR &&
+           man->use_tt)
+               page_flags |= TTM_PAGE_FLAG_ZERO_ALLOC;
+
+       ret = ttm_tt_init(&i915_tt->ttm, bo, page_flags, ttm_cached);
+       if (ret) {
+               kfree(i915_tt);
+               return NULL;
+       }
+
+       i915_tt->dev = obj->base.dev->dev;
+
+       return &i915_tt->ttm;
+}
+
+static void i915_ttm_tt_unpopulate(struct ttm_device *bdev, struct ttm_tt *ttm)
+{
+       struct i915_ttm_tt *i915_tt = container_of(ttm, typeof(*i915_tt), ttm);
+
+       if (i915_tt->cached_st) {
+               dma_unmap_sgtable(i915_tt->dev, i915_tt->cached_st,
+                                 DMA_BIDIRECTIONAL, 0);
+               sg_free_table(i915_tt->cached_st);
+               kfree(i915_tt->cached_st);
+               i915_tt->cached_st = NULL;
+       }
+       ttm_pool_free(&bdev->pool, ttm);
+}
+
+static void i915_ttm_tt_destroy(struct ttm_device *bdev, struct ttm_tt *ttm)
+{
+       struct i915_ttm_tt *i915_tt = container_of(ttm, typeof(*i915_tt), ttm);
+
+       ttm_tt_destroy_common(bdev, ttm);
+       kfree(i915_tt);
+}
+
+static bool i915_ttm_eviction_valuable(struct ttm_buffer_object *bo,
+                                      const struct ttm_place *place)
+{
+       struct drm_i915_gem_object *obj = i915_ttm_to_gem(bo);
+
+       /* Will do for now. Our pinned objects are still on TTM's LRU lists */
+       if (!i915_gem_object_evictable(obj))

drivers/gpu/drm/i915/gem/i915_gem_ttm.c line 148

Quoted 10 lines
drivers/gpu/drm/i915/gem/i915_gem_ttm.c new file +10 -0
+               return false;
+
+       /* This isn't valid with a buddy allocator */
+       return ttm_bo_eviction_valuable(bo, place);
+}
+
+static void i915_ttm_evict_flags(struct ttm_buffer_object *bo,
+                                struct ttm_placement *placement)
+{
+       *placement = i915_sys_placement;

What's the story here?

Quoted 67 lines
drivers/gpu/drm/i915/gem/i915_gem_ttm.c new file +67 -0
+}
+
+static int i915_ttm_move_notify(struct ttm_buffer_object *bo)
+{
+       struct drm_i915_gem_object *obj = i915_ttm_to_gem(bo);
+       int ret;
+
+       ret = i915_gem_object_unbind(obj, I915_GEM_OBJECT_UNBIND_ACTIVE);
+       if (ret)
+               return ret;
+
+       ret = __i915_gem_object_put_pages(obj);
+       if (ret)
+               return ret;
+
+       return 0;
+}
+
+static void i915_ttm_free_cached_io_st(struct drm_i915_gem_object *obj)
+{
+       if (obj->ttm.cached_io_st) {
+               sg_free_table(obj->ttm.cached_io_st);
+               kfree(obj->ttm.cached_io_st);
+               obj->ttm.cached_io_st = NULL;
+       }
+}
+
+static void i915_ttm_purge(struct drm_i915_gem_object *obj)
+{
+       struct ttm_buffer_object *bo = i915_gem_to_ttm(obj);
+
+       if (obj->mm.madv == __I915_MADV_PURGED)
+               return;
+
+       i915_ttm_free_cached_io_st(obj);
+
+       ttm_resource_free(bo, &bo->mem);
+       if (bo->ttm) {
+               ttm_tt_unpopulate(bo->bdev, bo->ttm);
+               bo->ttm->page_flags |= TTM_PAGE_FLAG_ZERO_ALLOC;
+       }
+
+       obj->mm.madv = __I915_MADV_PURGED;
+}
+
+static void i915_ttm_swap_notify(struct ttm_buffer_object *bo)
+{
+       struct drm_i915_gem_object *obj = i915_ttm_to_gem(bo);
+       int ret = i915_ttm_move_notify(bo);
+
+       GEM_WARN_ON(ret);
+       GEM_WARN_ON(obj->ttm.cached_io_st);
+       if (!ret && obj->mm.madv != I915_MADV_WILLNEED)
+               i915_ttm_purge(obj);
+}
+
+static void i915_ttm_delete_mem_notify(struct ttm_buffer_object *bo)
+{
+       struct drm_i915_gem_object *obj = i915_ttm_to_gem(bo);
+
+       if (likely(obj)) {
+               /* This releases all gem object bindings to the backend. */
+               __i915_gem_free_object(obj);
+       }
+}
+
+static struct intel_memory_region *

drivers/gpu/drm/i915/gem/i915_gem_ttm.c line 225

Quoted 10 lines
drivers/gpu/drm/i915/gem/i915_gem_ttm.c new file +10 -0
+i915_ttm_region(struct ttm_device *bdev, int ttm_mem_type)
+{
+       struct drm_i915_private *i915 = container_of(bdev, typeof(*i915), bdev);
+
+       /* There's some room for optimization here... */
+       GEM_BUG_ON(ttm_mem_type != I915_PL_SYSTEM &&
+                  ttm_mem_type < I915_PL_LMEM0);
+       if (ttm_mem_type == I915_PL_SYSTEM)
+               return intel_memory_region_lookup(i915, INTEL_MEMORY_SYSTEM,
+                                                 0);

So at the moment we just have lmem using the new ttm backend, right? Stolen and system are unchanged?

Quoted 280 lines
drivers/gpu/drm/i915/gem/i915_gem_ttm.c new file +280 -0
+
+       return intel_memory_region_lookup(i915, INTEL_MEMORY_LOCAL,
+                                         ttm_mem_type - I915_PL_LMEM0);
+}
+
+static struct sg_table *i915_ttm_tt_get_st(struct ttm_tt *ttm)
+{
+       struct i915_ttm_tt *i915_tt = container_of(ttm, typeof(*i915_tt), ttm);
+       struct scatterlist *sg;
+       struct sg_table *st;
+       int ret;
+
+       if (i915_tt->cached_st)
+               return i915_tt->cached_st;
+
+       st = kzalloc(sizeof(*st), GFP_KERNEL);
+       if (!st)
+               return ERR_PTR(-ENOMEM);
+
+       sg = __sg_alloc_table_from_pages
+               (st, ttm->pages, ttm->num_pages, 0,
+                (unsigned long)ttm->num_pages << PAGE_SHIFT,
+                i915_sg_segment_size(), NULL, 0, GFP_KERNEL);
+       if (IS_ERR(sg)) {
+               kfree(st);
+               return ERR_CAST(sg);
+       }
+
+       ret = dma_map_sgtable(i915_tt->dev, st, DMA_BIDIRECTIONAL, 0);
+       if (ret) {
+               sg_free_table(st);
+               kfree(st);
+               return ERR_PTR(ret);
+       }
+
+       i915_tt->cached_st = st;
+       return st;
+}
+
+static struct sg_table *
+i915_ttm_resource_get_st(struct drm_i915_gem_object *obj,
+                        struct ttm_resource *res)
+{
+       struct ttm_buffer_object *bo = i915_gem_to_ttm(obj);
+       struct ttm_resource_manager *man =
+               ttm_manager_type(bo->bdev, res->mem_type);
+
+       if (man->use_tt)
+               return i915_ttm_tt_get_st(bo->ttm);
+
+       return intel_region_ttm_node_to_st(obj->mm.region, res->mm_node);
+}
+
+static int i915_ttm_move(struct ttm_buffer_object *bo, bool evict,
+                        struct ttm_operation_ctx *ctx,
+                        struct ttm_resource *new_mem,
+                        struct ttm_place *hop)
+{
+       struct drm_i915_gem_object *obj = i915_ttm_to_gem(bo);
+       struct ttm_resource_manager *new_man =
+               ttm_manager_type(bo->bdev, new_mem->mem_type);
+       struct ttm_resource_manager *old_man =
+               ttm_manager_type(bo->bdev, bo->mem.mem_type);
+       struct intel_memory_region *new_reg, *old_reg;
+       union {
+               struct i915_ttm_kmap_iter_tt tt;
+               struct i915_ttm_kmap_iter_iomap io;
+       } _new_iter, _old_iter;
+       struct i915_ttm_kmap_iter *new_iter, *old_iter;
+       struct sg_table *new_st;
+       int ret;
+
+       new_reg = i915_ttm_region(bo->bdev, new_mem->mem_type);
+       old_reg = i915_ttm_region(bo->bdev, bo->mem.mem_type);
+       GEM_BUG_ON(!new_reg || !old_reg);
+
+       /* Sync for now. We could do the actual copy async. */
+       ret = ttm_bo_wait_ctx(bo, ctx);
+       if (ret)
+               return ret;
+
+       ret = i915_ttm_move_notify(bo);
+       if (ret)
+               return ret;
+
+       if (obj->mm.madv != I915_MADV_WILLNEED) {
+               i915_ttm_purge(obj);
+               ttm_resource_free(bo, new_mem);
+               return 0;
+       }
+
+       /* Populate ttm with pages if needed. Typically system memory. */
+       if (new_man->use_tt && bo->ttm) {
+               ret = ttm_tt_populate(bo->bdev, bo->ttm, ctx);
+               if (ret)
+                       return ret;
+       }
+
+       new_st = i915_ttm_resource_get_st(obj, new_mem);
+       if (IS_ERR(new_st))
+               return PTR_ERR(new_st);
+
+       new_iter = new_man->use_tt ?
+               i915_ttm_kmap_iter_tt_init(&_new_iter.tt, bo->ttm) :
+               i915_ttm_kmap_iter_iomap_init(&_new_iter.io, &new_reg->iomap,
+                                             new_st, new_reg->region.start);
+
+       old_iter = old_man->use_tt ?
+               i915_ttm_kmap_iter_tt_init(&_old_iter.tt, bo->ttm) :
+               i915_ttm_kmap_iter_iomap_init(&_old_iter.io, &old_reg->iomap,
+                                             obj->ttm.cached_io_st,
+                                             old_reg->region.start);
+
+       i915_ttm_move_memcpy(bo, new_mem, new_iter, old_iter);
+       i915_ttm_free_cached_io_st(obj);
+
+       if (!new_man->use_tt)
+               obj->ttm.cached_io_st = new_st;
+
+       return 0;
+}
+
+struct ttm_device_funcs i915_ttm_bo_driver = {
+       .ttm_tt_create = i915_ttm_tt_create,
+       .ttm_tt_unpopulate = i915_ttm_tt_unpopulate,
+       .ttm_tt_destroy = i915_ttm_tt_destroy,
+       .eviction_valuable = i915_ttm_eviction_valuable,
+       .evict_flags = i915_ttm_evict_flags,
+       .move = i915_ttm_move,
+       .verify_access = NULL,
+       .swap_notify = i915_ttm_swap_notify,
+       .delete_mem_notify = i915_ttm_delete_mem_notify,
+};
+
+static int i915_ttm_get_pages(struct drm_i915_gem_object *obj)
+{
+       struct ttm_buffer_object *bo = i915_gem_to_ttm(obj);
+       struct ttm_operation_ctx ctx = {
+               .interruptible = true,
+               .no_wait_gpu = false,
+       };
+       struct sg_table *st;
+       int ret;
+
+       /* Swap in. */
+       if (bo->ttm && (bo->ttm->page_flags & TTM_PAGE_FLAG_SWAPPED)) {
+               ret = ttm_tt_populate(bo->bdev, bo->ttm, &ctx);
+               if (ret)
+                       return ret;
+       }
+
+       /* Move to the requested placement. */
+       ret = ttm_bo_validate(bo, &i915_lmem0_placement, &ctx);
+       if (ret)
+               return ret == -ENOSPC ? -ENXIO : ret;
+
+       /* Object either has a page vector or is an iomem object */
+       st = bo->ttm ? i915_ttm_tt_get_st(bo->ttm) : obj->ttm.cached_io_st;
+       if (IS_ERR(st))
+               return PTR_ERR(st);
+
+       __i915_gem_object_set_pages(obj, st, i915_sg_dma_sizes(st->sgl));
+
+       i915_ttm_adjust_lru(obj);
+
+       return ret;
+}
+
+static void i915_ttm_put_pages(struct drm_i915_gem_object *obj,
+                              struct sg_table *st)
+{
+       /*
+        * We're currently not called from a shrinker, so put_pages()
+        * typically means the object is about to destroyed, or called
+        * from move_notify(). So just avoid doing much for now.
+        * If the object is not destroyed next, The TTM eviction logic
+        * and shrinkers will move it out if needed.
+        */
+
+       i915_ttm_adjust_lru(obj);
+}
+
+static void i915_ttm_adjust_lru(struct drm_i915_gem_object *obj)
+{
+       struct ttm_buffer_object *bo = i915_gem_to_ttm(obj);
+
+       /*
+        * Don't manipulate the TTM LRUs while in TTM bo destruction.
+        * We're called through i915_ttm_delete_mem_notify().
+        */
+       if (!kref_read(&bo->kref))
+               return;
+
+       /*
+        * Put on the correct LRU list depending on the MADV status
+        */
+       spin_lock(&bo->bdev->lru_lock);
+       if (obj->mm.madv != I915_MADV_WILLNEED) {
+               bo->priority = I915_TTM_PRIO_PURGE;
+       } else if (!i915_gem_object_has_pages(obj)) {
+               if (bo->priority < I915_TTM_PRIO_HAS_PAGES)
+                       bo->priority = I915_TTM_PRIO_HAS_PAGES;
+       } else {
+               if (bo->priority > I915_TTM_PRIO_NO_PAGES)
+                       bo->priority = I915_TTM_PRIO_NO_PAGES;
+       }
+
+       ttm_bo_move_to_lru_tail(bo, &bo->mem, NULL);
+       spin_unlock(&bo->bdev->lru_lock);
+}
+
+/*
+ * TTM-backed gem object destruction requires some clarification.
+ * Basically we have two possibilities here. We can either rely on the
+ * i915 delayed destruction and put the TTM object when the object
+ * is idle. This would be detected by TTM which would bypass the
+ * TTM delayed destroy handling. The other approach is to put the TTM
+ * object early and rely on the TTM destroyed handling, and then free
+ * the leftover parts of the GEM object once TTM's destroyed list handling is
+ * complete. For now, we rely on the latter for two reasons:
+ * a) TTM can evict an object even when it's on the delayed destroy list,
+ * which in theory allows for complete eviction.
+ * b) There is work going on in TTM to allow freeing an object even when
+ * it's not idle, and using the TTM destroyed list handling could help us
+ * benefit from that.
+ */
+static void i915_ttm_delayed_free(struct drm_i915_gem_object *obj)
+{
+       ttm_bo_put(i915_gem_to_ttm(obj));
+}
+
+static const struct drm_i915_gem_object_ops i915_gem_ttm_obj_ops = {
+       .name = "i915_gem_object_ttm",
+       .flags = I915_GEM_OBJECT_HAS_IOMEM,
+
+       .get_pages = i915_ttm_get_pages,
+       .put_pages = i915_ttm_put_pages,
+       .truncate = i915_ttm_purge,
+       .adjust_lru = i915_ttm_adjust_lru,
+       .delayed_free = i915_ttm_delayed_free,
+};
+
+void i915_ttm_bo_destroy(struct ttm_buffer_object *bo)
+{
+       struct drm_i915_gem_object *obj = i915_ttm_to_gem(bo);
+
+       i915_gem_object_release_memory_region(obj);
+       call_rcu(&obj->rcu, __i915_gem_free_object_rcu);
+}
+
+/**
+ * __i915_gem_ttm_object_init - Initialize a ttm-backed i915 gem object
+ * @mem: The initial memory region for the object.
+ * @obj: The gem object.
+ * @size: Object size in bytes.
+ * @flags: gem object flags.
+ *
+ * Return: 0 on success, negative error code on failure.
+ */
+int __i915_gem_ttm_object_init(struct intel_memory_region *mem,
+                              struct drm_i915_gem_object *obj,
+                              resource_size_t size,
+                              unsigned int flags)
+{
+       static struct lock_class_key lock_class;
+       struct drm_i915_private *i915 = mem->i915;
+       size_t alignment = 0;
+       int ret;
+
+       /* Adjust alignment to GPU- and CPU huge page sizes. */
+
+       if (mem->is_range_manager) {
+               if (size >= SZ_1G)
+                       alignment = SZ_1G >> PAGE_SHIFT;
+               else if (size >= SZ_2M)
+                       alignment = SZ_2M >> PAGE_SHIFT;
+               else if (size >= SZ_64K)
+                       alignment = SZ_64K >> PAGE_SHIFT;
+       }
+

drivers/gpu/drm/i915/gem/i915_gem_ttm.c line 515

Quoted 10 lines
drivers/gpu/drm/i915/gem/i915_gem_ttm.c new file +10 -0
+       drm_gem_private_object_init(&i915->drm, &obj->base, size);
+       i915_gem_object_init(obj, &i915_gem_ttm_obj_ops, &lock_class, flags);
+       i915_gem_object_init_memory_region(obj, mem);
+       i915_gem_object_make_unshrinkable(obj);
+       obj->read_domains = I915_GEM_DOMAIN_WC | I915_GEM_DOMAIN_GTT;
+       i915_gem_object_set_cache_coherency(obj, I915_CACHE_NONE);
+
+       ret = ttm_bo_init(&i915->bdev, i915_gem_to_ttm(obj), size,
+                         ttm_bo_type_kernel, &i915_sys_placement, alignment,
+                         true, NULL, NULL, i915_ttm_bo_destroy);

Move this further up, or add some onion?

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

author reviewer
#45 lore raw

Re: [Intel-gfx] [PATCH v2 10/15] drm/i915/ttm: Introduce a TTM i915 gem object backend

Thanks a lot for reviewing, Matthew!

Quoted 42 lines · Matthew Auld
Quoted 40 lines · Thomas Hellström

Most logical place to introduce TTM buffer objects is as an i915 gem object backend. We need to add some ops to account for added functionality like delayed delete and LRU list manipulation.

Initially we support only LMEM and SYSTEM memory, but SYSTEM (which in this case means evicted LMEM objects) is not visible to i915 GEM yet. The plan is to move the i915 gem system region over to the TTM system memory type in upcoming patches.

We set up GPU bindings directly both from LMEM and from the system region, as there is no need to use the legacy TTM_TT memory type. We reserve that for future porting of GGTT bindings to TTM.

Remove the old lmem backend.

drivers/gpu/drm/i915/gem/i915_gem_object.c +23 -0
+/**
+ * i915_gem_object_evictable - Whether object is likely evictable after unbind.
+ * @obj: The object to check
+ *
+ * This function checks whether the object is likely unvictable after unbind.
+ * If the object is not locked when checking, the result is only advisory.
+ * If the object is locked when checking, and the function returns true,
+ * then an eviction should indeed be possible. But since unlocked vma
+ * unpinning and unbinding is currently possible, the object can actually
+ * become evictable even if this function returns false.
+ *
+ * Return: true if the object may be evictable. False otherwise.
+ */
+bool i915_gem_object_evictable(struct drm_i915_gem_object *obj)
+{
+       struct i915_vma *vma;
+       int pin_count = atomic_read(&obj->mm.pages_pin_count);
+
+       if (!pin_count)
+               return true;
+
+       spin_lock(&obj->vma.lock);
+       list_for_each_entry(vma, &obj->vma.list, obj_link) {
Quoted 8 lines

drivers/gpu/drm/i915/gem/i915_gem_object.c int i915_gem_object_read_from_page(struct drm_i915_gem_object *obj, u64 offset, · line 452

Quoted 6 lines
drivers/gpu/drm/i915/gem/i915_gem_object.c +6 -0
+               if (i915_vma_is_pinned(vma)) {
+                       spin_unlock(&obj->vma.lock);
+                       return false;
+               }
+               if (atomic_read(&vma->pages_count))
+                       pin_count--;

Can't pages_count be > 1, which would also be reflected in pages_pin_count? The vma_pin path looks very complex.

Yes, and Maarten pointed out a bug in it as well. We now only take a pages_pin_count when vma->pages_count transitions from 0->1 so the above code should be correct, I think.

Quoted 17 lines
Quoted 17 lines
drivers/gpu/drm/i915/gem/i915_gem_object.c +7 -0
+       }
+       spin_unlock(&obj->vma.lock);
+       GEM_WARN_ON(pin_count < 0);
+
+       return pin_count == 0;
+}
+
  void i915_gem_init__objects(struct drm_i915_private *i915)
  {
         INIT_WORK(&i915->mm.free_work, __i915_gem_free_work);
drivers/gpu/drm/i915/gem/i915_gem_object.h +0 -0
index 2ebd79537aea..ae5930e307d5 100644
@@ -200,6 +200,9 @@ static inline bool i915_gem_object_trylock(struct drm_i915_gem_object *obj)
 
Quoted 5 lines

drivers/gpu/drm/i915/gem/i915_gem_object.h static inline bool i915_gem_object_trylock(struct drm_i915_gem_object *obj) · line 204

Quoted 4 lines
drivers/gpu/drm/i915/gem/i915_gem_object.h +2 -0
  static inline void i915_gem_object_unlock(struct drm_i915_gem_object *obj)
  {
+       if (obj->ops->adjust_lru)
+               obj->ops->adjust_lru(obj);

Interesting, so we bump the lru even when we just drop the lock?

Yes, as an initial approximation. TTM has historically done this and it may not be the best choice, but dropping the lock means we've just used the object for something, typically CS, so hence we bump the LRU.

Quoted 110 lines
Quoted 110 lines
drivers/gpu/drm/i915/gem/i915_gem_object.h +7 -0
+
         dma_resv_unlock(obj->base.resv);
  }
 
@@ -587,6 +590,12 @@ int i915_gem_object_read_from_page(struct drm_i915_gem_object *obj, u64 offset,
 
  bool i915_gem_object_is_shmem(const struct drm_i915_gem_object *obj);
 
+void __i915_gem_free_object_rcu(struct rcu_head *head);
+
+void __i915_gem_free_object(struct drm_i915_gem_object *obj);
+
+bool i915_gem_object_evictable(struct drm_i915_gem_object *obj);
+
  #ifdef CONFIG_MMU_NOTIFIER
  static inline bool
  i915_gem_object_is_userptr(struct drm_i915_gem_object *obj)
drivers/gpu/drm/i915/gem/i915_gem_object_types.h +18 -0
index 98f69d8fd37d..b350765e1935 100644
@@ -63,6 +63,20 @@ struct drm_i915_gem_object_ops {
                       const struct drm_i915_gem_pwrite *arg);
 
         int (*dmabuf_export)(struct drm_i915_gem_object *obj);
+
+       /**
+        * adjust_lru - notify that the madvise value was updated
+        * @obj: The gem object
+        *
+        * The madvise value may have been updated, or object was recently
+        * referenced so act accordingly (Perhaps changing an LRU list etc).
+        */
+       void (*adjust_lru)(struct drm_i915_gem_object *obj);
+
+       /**
+        * delayed_free - Override the default delayed free implementation
+        */
+       void (*delayed_free)(struct drm_i915_gem_object *obj);
         void (*release)(struct drm_i915_gem_object *obj);
 
         const char *name; /* friendly name for debug, e.g. lockdep classes */
@@ -307,6 +321,10 @@ struct drm_i915_gem_object {
                 bool dirty:1;
         } mm;
 
+       struct {
+               struct sg_table *cached_io_st;
+       } ttm;
+
         /** Record of address bit 17 of each page at last unbind. */
         unsigned long *bit_17;
 
drivers/gpu/drm/i915/gem/i915_gem_region.c +1 -5
index f25e6646c5b7..d1f1840540dd 100644
@@ -18,11 +18,7 @@ void i915_gem_object_init_memory_region(struct drm_i915_gem_object *obj,
 
         mutex_lock(&mem->objects.lock);
 
-       if (obj->flags & I915_BO_ALLOC_VOLATILE)
-               list_add(&obj->mm.region_link, &mem->objects.purgeable);
-       else
-               list_add(&obj->mm.region_link, &mem->objects.list);
-
+       list_add(&obj->mm.region_link, &mem->objects.list);
         mutex_unlock(&mem->objects.lock);
  }
 
drivers/gpu/drm/i915/gem/i915_gem_ttm.c new file +34 -0
new file mode 100644
index 000000000000..790f5ec45c4d
@@ -0,0 +1,519 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2021 Intel Corporation
+ */
+
+#include <drm/ttm/ttm_bo_driver.h>
+#include <drm/ttm/ttm_placement.h>
+
+#include "i915_drv.h"
+#include "intel_memory_region.h"
+#include "intel_region_ttm.h"
+
+#include "gem/i915_gem_object.h"
+#include "gem/i915_gem_region.h"
+#include "gem/i915_gem_ttm.h"
+#include "gem/i915_gem_ttm_bo_util.h"
+
+#define I915_PL_LMEM0 TTM_PL_PRIV
+#define I915_PL_SYSTEM TTM_PL_SYSTEM
+#define I915_PL_STOLEN TTM_PL_VRAM
+#define I915_PL_GGTT TTM_PL_TT
+
+#define I915_TTM_PRIO_PURGE     0
+#define I915_TTM_PRIO_NO_PAGES  1
+#define I915_TTM_PRIO_HAS_PAGES 2
+
+/**
+ * struct i915_ttm_tt - TTM page vector with additional private information
+ * @ttm: The base TTM page vector.
+ * @dev: The struct device used for dma mapping and unmapping.
+ * @cached_st: The cached scatter-gather table.
+ *
+ * Note that DMA may be going on right up to the point where the page-
+ * vector is unpopulated in delayed destroy. Hence keep the
Quoted 8 lines

drivers/gpu/drm/i915/gem/i915_gem_ttm.c line 39

Quoted 5 lines
drivers/gpu/drm/i915/gem/i915_gem_ttm.c new file +5 -0
+ * scatter-gather table mapped and cached up to that point. This is
+ * different from the cached gem object io scatter-gather table which
+ * doesn't have an associated dma mapping.
+ */
+struct i915_ttm_tt {

What is the _tt here btw? Translation table? We also have use_tt elsewhere. ttm_tt looks like it just holds an array of pages, and associated data? ttm_pv?

It's just bad naming. i915_ttm_pv would be better, I agree, but other drivers consistently use tt.

Quoted 103 lines
Quoted 103 lines
drivers/gpu/drm/i915/gem/i915_gem_ttm.c new file +102 -0
+       struct ttm_tt ttm;
+       struct device *dev;
+       struct sg_table *cached_st;
+};
+
+static const struct ttm_place lmem0_sys_placement_flags[] = {
+       {
+               .fpfn = 0,
+               .lpfn = 0,
+               .mem_type = I915_PL_LMEM0,
+               .flags = 0,
+       }, {
+               .fpfn = 0,
+               .lpfn = 0,
+               .mem_type = I915_PL_SYSTEM,
+               .flags = 0,
+       }
+};
+
+struct ttm_placement i915_lmem0_placement = {
+       .num_placement = 1,
+       .placement = &lmem0_sys_placement_flags[0],
+       .num_busy_placement = 1,
+       .busy_placement = &lmem0_sys_placement_flags[0],
+};
+
+struct ttm_placement i915_lmem0_sys_placement = {
+       .num_placement = 1,
+       .placement = &lmem0_sys_placement_flags[0],
+       .num_busy_placement = 2,
+       .busy_placement = &lmem0_sys_placement_flags[0],
+};
+
+struct ttm_placement i915_sys_placement = {
+       .num_placement = 1,
+       .placement = &lmem0_sys_placement_flags[1],
+       .num_busy_placement = 1,
+       .busy_placement = &lmem0_sys_placement_flags[1],
+};
+
+static void i915_ttm_adjust_lru(struct drm_i915_gem_object *obj);
+
+static struct ttm_tt *i915_ttm_tt_create(struct ttm_buffer_object *bo,
+                                        uint32_t page_flags)
+{
+       struct ttm_resource_manager *man =
+               ttm_manager_type(bo->bdev, bo->mem.mem_type);
+       struct drm_i915_gem_object *obj = i915_ttm_to_gem(bo);
+       struct i915_ttm_tt *i915_tt;
+       int ret;
+
+       i915_tt = kzalloc(sizeof(*i915_tt), GFP_KERNEL);
+       if (!i915_tt)
+               return NULL;
+
+       if (obj->flags & I915_BO_ALLOC_CPU_CLEAR &&
+           man->use_tt)
+               page_flags |= TTM_PAGE_FLAG_ZERO_ALLOC;
+
+       ret = ttm_tt_init(&i915_tt->ttm, bo, page_flags, ttm_cached);
+       if (ret) {
+               kfree(i915_tt);
+               return NULL;
+       }
+
+       i915_tt->dev = obj->base.dev->dev;
+
+       return &i915_tt->ttm;
+}
+
+static void i915_ttm_tt_unpopulate(struct ttm_device *bdev, struct ttm_tt *ttm)
+{
+       struct i915_ttm_tt *i915_tt = container_of(ttm, typeof(*i915_tt), ttm);
+
+       if (i915_tt->cached_st) {
+               dma_unmap_sgtable(i915_tt->dev, i915_tt->cached_st,
+                                 DMA_BIDIRECTIONAL, 0);
+               sg_free_table(i915_tt->cached_st);
+               kfree(i915_tt->cached_st);
+               i915_tt->cached_st = NULL;
+       }
+       ttm_pool_free(&bdev->pool, ttm);
+}
+
+static void i915_ttm_tt_destroy(struct ttm_device *bdev, struct ttm_tt *ttm)
+{
+       struct i915_ttm_tt *i915_tt = container_of(ttm, typeof(*i915_tt), ttm);
+
+       ttm_tt_destroy_common(bdev, ttm);
+       kfree(i915_tt);
+}
+
+static bool i915_ttm_eviction_valuable(struct ttm_buffer_object *bo,
+                                      const struct ttm_place *place)
+{
+       struct drm_i915_gem_object *obj = i915_ttm_to_gem(bo);
+
+       /* Will do for now. Our pinned objects are still on TTM's LRU lists */
+       if (!i915_gem_object_evictable(obj))
+               return false;
+
+       /* This isn't valid with a buddy allocator */
Quoted 8 lines

drivers/gpu/drm/i915/gem/i915_gem_ttm.c line 148

Quoted 7 lines
drivers/gpu/drm/i915/gem/i915_gem_ttm.c new file +7 -0
+       return ttm_bo_eviction_valuable(bo, place);
+}
+
+static void i915_ttm_evict_flags(struct ttm_buffer_object *bo,
+                                struct ttm_placement *placement)
+{
+       *placement = i915_sys_placement;

What's the story here?

Here it's indicated that the bo is about to be evicted and TTM asks us where to place it. We place it in system memory, but that system memory is not yet visible to GEM.

Quoted 72 lines
Quoted 72 lines
drivers/gpu/drm/i915/gem/i915_gem_ttm.c new file +71 -0
+}
+
+static int i915_ttm_move_notify(struct ttm_buffer_object *bo)
+{
+       struct drm_i915_gem_object *obj = i915_ttm_to_gem(bo);
+       int ret;
+
+       ret = i915_gem_object_unbind(obj, I915_GEM_OBJECT_UNBIND_ACTIVE);
+       if (ret)
+               return ret;
+
+       ret = __i915_gem_object_put_pages(obj);
+       if (ret)
+               return ret;
+
+       return 0;
+}
+
+static void i915_ttm_free_cached_io_st(struct drm_i915_gem_object *obj)
+{
+       if (obj->ttm.cached_io_st) {
+               sg_free_table(obj->ttm.cached_io_st);
+               kfree(obj->ttm.cached_io_st);
+               obj->ttm.cached_io_st = NULL;
+       }
+}
+
+static void i915_ttm_purge(struct drm_i915_gem_object *obj)
+{
+       struct ttm_buffer_object *bo = i915_gem_to_ttm(obj);
+
+       if (obj->mm.madv == __I915_MADV_PURGED)
+               return;
+
+       i915_ttm_free_cached_io_st(obj);
+
+       ttm_resource_free(bo, &bo->mem);
+       if (bo->ttm) {
+               ttm_tt_unpopulate(bo->bdev, bo->ttm);
+               bo->ttm->page_flags |= TTM_PAGE_FLAG_ZERO_ALLOC;
+       }
+
+       obj->mm.madv = __I915_MADV_PURGED;
+}
+
+static void i915_ttm_swap_notify(struct ttm_buffer_object *bo)
+{
+       struct drm_i915_gem_object *obj = i915_ttm_to_gem(bo);
+       int ret = i915_ttm_move_notify(bo);
+
+       GEM_WARN_ON(ret);
+       GEM_WARN_ON(obj->ttm.cached_io_st);
+       if (!ret && obj->mm.madv != I915_MADV_WILLNEED)
+               i915_ttm_purge(obj);
+}
+
+static void i915_ttm_delete_mem_notify(struct ttm_buffer_object *bo)
+{
+       struct drm_i915_gem_object *obj = i915_ttm_to_gem(bo);
+
+       if (likely(obj)) {
+               /* This releases all gem object bindings to the backend. */
+               __i915_gem_free_object(obj);
+       }
+}
+
+static struct intel_memory_region *
+i915_ttm_region(struct ttm_device *bdev, int ttm_mem_type)
+{
+       struct drm_i915_private *i915 = container_of(bdev, typeof(*i915), bdev);
+
Quoted 8 lines

drivers/gpu/drm/i915/gem/i915_gem_ttm.c line 225

Quoted 6 lines
drivers/gpu/drm/i915/gem/i915_gem_ttm.c new file +6 -0
+       /* There's some room for optimization here... */
+       GEM_BUG_ON(ttm_mem_type != I915_PL_SYSTEM &&
+                  ttm_mem_type < I915_PL_LMEM0);
+       if (ttm_mem_type == I915_PL_SYSTEM)
+               return intel_memory_region_lookup(i915, INTEL_MEMORY_SYSTEM,
+                                                 0);

So at the moment we just have lmem using the new ttm backend, right? Stolen and system are unchanged?

ATM yes.

Quoted 284 lines
Quoted 284 lines
drivers/gpu/drm/i915/gem/i915_gem_ttm.c new file +283 -0
+
+       return intel_memory_region_lookup(i915, INTEL_MEMORY_LOCAL,
+                                         ttm_mem_type - I915_PL_LMEM0);
+}
+
+static struct sg_table *i915_ttm_tt_get_st(struct ttm_tt *ttm)
+{
+       struct i915_ttm_tt *i915_tt = container_of(ttm, typeof(*i915_tt), ttm);
+       struct scatterlist *sg;
+       struct sg_table *st;
+       int ret;
+
+       if (i915_tt->cached_st)
+               return i915_tt->cached_st;
+
+       st = kzalloc(sizeof(*st), GFP_KERNEL);
+       if (!st)
+               return ERR_PTR(-ENOMEM);
+
+       sg = __sg_alloc_table_from_pages
+               (st, ttm->pages, ttm->num_pages, 0,
+                (unsigned long)ttm->num_pages << PAGE_SHIFT,
+                i915_sg_segment_size(), NULL, 0, GFP_KERNEL);
+       if (IS_ERR(sg)) {
+               kfree(st);
+               return ERR_CAST(sg);
+       }
+
+       ret = dma_map_sgtable(i915_tt->dev, st, DMA_BIDIRECTIONAL, 0);
+       if (ret) {
+               sg_free_table(st);
+               kfree(st);
+               return ERR_PTR(ret);
+       }
+
+       i915_tt->cached_st = st;
+       return st;
+}
+
+static struct sg_table *
+i915_ttm_resource_get_st(struct drm_i915_gem_object *obj,
+                        struct ttm_resource *res)
+{
+       struct ttm_buffer_object *bo = i915_gem_to_ttm(obj);
+       struct ttm_resource_manager *man =
+               ttm_manager_type(bo->bdev, res->mem_type);
+
+       if (man->use_tt)
+               return i915_ttm_tt_get_st(bo->ttm);
+
+       return intel_region_ttm_node_to_st(obj->mm.region, res->mm_node);
+}
+
+static int i915_ttm_move(struct ttm_buffer_object *bo, bool evict,
+                        struct ttm_operation_ctx *ctx,
+                        struct ttm_resource *new_mem,
+                        struct ttm_place *hop)
+{
+       struct drm_i915_gem_object *obj = i915_ttm_to_gem(bo);
+       struct ttm_resource_manager *new_man =
+               ttm_manager_type(bo->bdev, new_mem->mem_type);
+       struct ttm_resource_manager *old_man =
+               ttm_manager_type(bo->bdev, bo->mem.mem_type);
+       struct intel_memory_region *new_reg, *old_reg;
+       union {
+               struct i915_ttm_kmap_iter_tt tt;
+               struct i915_ttm_kmap_iter_iomap io;
+       } _new_iter, _old_iter;
+       struct i915_ttm_kmap_iter *new_iter, *old_iter;
+       struct sg_table *new_st;
+       int ret;
+
+       new_reg = i915_ttm_region(bo->bdev, new_mem->mem_type);
+       old_reg = i915_ttm_region(bo->bdev, bo->mem.mem_type);
+       GEM_BUG_ON(!new_reg || !old_reg);
+
+       /* Sync for now. We could do the actual copy async. */
+       ret = ttm_bo_wait_ctx(bo, ctx);
+       if (ret)
+               return ret;
+
+       ret = i915_ttm_move_notify(bo);
+       if (ret)
+               return ret;
+
+       if (obj->mm.madv != I915_MADV_WILLNEED) {
+               i915_ttm_purge(obj);
+               ttm_resource_free(bo, new_mem);
+               return 0;
+       }
+
+       /* Populate ttm with pages if needed. Typically system memory. */
+       if (new_man->use_tt && bo->ttm) {
+               ret = ttm_tt_populate(bo->bdev, bo->ttm, ctx);
+               if (ret)
+                       return ret;
+       }
+
+       new_st = i915_ttm_resource_get_st(obj, new_mem);
+       if (IS_ERR(new_st))
+               return PTR_ERR(new_st);
+
+       new_iter = new_man->use_tt ?
+               i915_ttm_kmap_iter_tt_init(&_new_iter.tt, bo->ttm) :
+               i915_ttm_kmap_iter_iomap_init(&_new_iter.io, &new_reg->iomap,
+                                             new_st, new_reg->region.start);
+
+       old_iter = old_man->use_tt ?
+               i915_ttm_kmap_iter_tt_init(&_old_iter.tt, bo->ttm) :
+               i915_ttm_kmap_iter_iomap_init(&_old_iter.io, &old_reg->iomap,
+                                             obj->ttm.cached_io_st,
+                                             old_reg->region.start);
+
+       i915_ttm_move_memcpy(bo, new_mem, new_iter, old_iter);
+       i915_ttm_free_cached_io_st(obj);
+
+       if (!new_man->use_tt)
+               obj->ttm.cached_io_st = new_st;
+
+       return 0;
+}
+
+struct ttm_device_funcs i915_ttm_bo_driver = {
+       .ttm_tt_create = i915_ttm_tt_create,
+       .ttm_tt_unpopulate = i915_ttm_tt_unpopulate,
+       .ttm_tt_destroy = i915_ttm_tt_destroy,
+       .eviction_valuable = i915_ttm_eviction_valuable,
+       .evict_flags = i915_ttm_evict_flags,
+       .move = i915_ttm_move,
+       .verify_access = NULL,
+       .swap_notify = i915_ttm_swap_notify,
+       .delete_mem_notify = i915_ttm_delete_mem_notify,
+};
+
+static int i915_ttm_get_pages(struct drm_i915_gem_object *obj)
+{
+       struct ttm_buffer_object *bo = i915_gem_to_ttm(obj);
+       struct ttm_operation_ctx ctx = {
+               .interruptible = true,
+               .no_wait_gpu = false,
+       };
+       struct sg_table *st;
+       int ret;
+
+       /* Swap in. */
+       if (bo->ttm && (bo->ttm->page_flags & TTM_PAGE_FLAG_SWAPPED)) {
+               ret = ttm_tt_populate(bo->bdev, bo->ttm, &ctx);
+               if (ret)
+                       return ret;
+       }
+
+       /* Move to the requested placement. */
+       ret = ttm_bo_validate(bo, &i915_lmem0_placement, &ctx);
+       if (ret)
+               return ret == -ENOSPC ? -ENXIO : ret;
+
+       /* Object either has a page vector or is an iomem object */
+       st = bo->ttm ? i915_ttm_tt_get_st(bo->ttm) : obj->ttm.cached_io_st;
+       if (IS_ERR(st))
+               return PTR_ERR(st);
+
+       __i915_gem_object_set_pages(obj, st, i915_sg_dma_sizes(st->sgl));
+
+       i915_ttm_adjust_lru(obj);
+
+       return ret;
+}
+
+static void i915_ttm_put_pages(struct drm_i915_gem_object *obj,
+                              struct sg_table *st)
+{
+       /*
+        * We're currently not called from a shrinker, so put_pages()
+        * typically means the object is about to destroyed, or called
+        * from move_notify(). So just avoid doing much for now.
+        * If the object is not destroyed next, The TTM eviction logic
+        * and shrinkers will move it out if needed.
+        */
+
+       i915_ttm_adjust_lru(obj);
+}
+
+static void i915_ttm_adjust_lru(struct drm_i915_gem_object *obj)
+{
+       struct ttm_buffer_object *bo = i915_gem_to_ttm(obj);
+
+       /*
+        * Don't manipulate the TTM LRUs while in TTM bo destruction.
+        * We're called through i915_ttm_delete_mem_notify().
+        */
+       if (!kref_read(&bo->kref))
+               return;
+
+       /*
+        * Put on the correct LRU list depending on the MADV status
+        */
+       spin_lock(&bo->bdev->lru_lock);
+       if (obj->mm.madv != I915_MADV_WILLNEED) {
+               bo->priority = I915_TTM_PRIO_PURGE;
+       } else if (!i915_gem_object_has_pages(obj)) {
+               if (bo->priority < I915_TTM_PRIO_HAS_PAGES)
+                       bo->priority = I915_TTM_PRIO_HAS_PAGES;
+       } else {
+               if (bo->priority > I915_TTM_PRIO_NO_PAGES)
+                       bo->priority = I915_TTM_PRIO_NO_PAGES;
+       }
+
+       ttm_bo_move_to_lru_tail(bo, &bo->mem, NULL);
+       spin_unlock(&bo->bdev->lru_lock);
+}
+
+/*
+ * TTM-backed gem object destruction requires some clarification.
+ * Basically we have two possibilities here. We can either rely on the
+ * i915 delayed destruction and put the TTM object when the object
+ * is idle. This would be detected by TTM which would bypass the
+ * TTM delayed destroy handling. The other approach is to put the TTM
+ * object early and rely on the TTM destroyed handling, and then free
+ * the leftover parts of the GEM object once TTM's destroyed list handling is
+ * complete. For now, we rely on the latter for two reasons:
+ * a) TTM can evict an object even when it's on the delayed destroy list,
+ * which in theory allows for complete eviction.
+ * b) There is work going on in TTM to allow freeing an object even when
+ * it's not idle, and using the TTM destroyed list handling could help us
+ * benefit from that.
+ */
+static void i915_ttm_delayed_free(struct drm_i915_gem_object *obj)
+{
+       ttm_bo_put(i915_gem_to_ttm(obj));
+}
+
+static const struct drm_i915_gem_object_ops i915_gem_ttm_obj_ops = {
+       .name = "i915_gem_object_ttm",
+       .flags = I915_GEM_OBJECT_HAS_IOMEM,
+
+       .get_pages = i915_ttm_get_pages,
+       .put_pages = i915_ttm_put_pages,
+       .truncate = i915_ttm_purge,
+       .adjust_lru = i915_ttm_adjust_lru,
+       .delayed_free = i915_ttm_delayed_free,
+};
+
+void i915_ttm_bo_destroy(struct ttm_buffer_object *bo)
+{
+       struct drm_i915_gem_object *obj = i915_ttm_to_gem(bo);
+
+       i915_gem_object_release_memory_region(obj);
+       call_rcu(&obj->rcu, __i915_gem_free_object_rcu);
+}
+
+/**
+ * __i915_gem_ttm_object_init - Initialize a ttm-backed i915 gem object
+ * @mem: The initial memory region for the object.
+ * @obj: The gem object.
+ * @size: Object size in bytes.
+ * @flags: gem object flags.
+ *
+ * Return: 0 on success, negative error code on failure.
+ */
+int __i915_gem_ttm_object_init(struct intel_memory_region *mem,
+                              struct drm_i915_gem_object *obj,
+                              resource_size_t size,
+                              unsigned int flags)
+{
+       static struct lock_class_key lock_class;
+       struct drm_i915_private *i915 = mem->i915;
+       size_t alignment = 0;
+       int ret;
+
+       /* Adjust alignment to GPU- and CPU huge page sizes. */
+
+       if (mem->is_range_manager) {
+               if (size >= SZ_1G)
+                       alignment = SZ_1G >> PAGE_SHIFT;
+               else if (size >= SZ_2M)
+                       alignment = SZ_2M >> PAGE_SHIFT;
+               else if (size >= SZ_64K)
+                       alignment = SZ_64K >> PAGE_SHIFT;
+       }
+
+       drm_gem_private_object_init(&i915->drm, &obj->base, size);
+       i915_gem_object_init(obj, &i915_gem_ttm_obj_ops, &lock_class, flags);
+       i915_gem_object_init_memory_region(obj, mem);
Quoted 8 lines

drivers/gpu/drm/i915/gem/i915_gem_ttm.c line 515

Quoted 7 lines
drivers/gpu/drm/i915/gem/i915_gem_ttm.c new file +7 -0
+       i915_gem_object_make_unshrinkable(obj);
+       obj->read_domains = I915_GEM_DOMAIN_WC | I915_GEM_DOMAIN_GTT;
+       i915_gem_object_set_cache_coherency(obj, I915_CACHE_NONE);
+
+       ret = ttm_bo_init(&i915->bdev, i915_gem_to_ttm(obj), size,
+                         ttm_bo_type_kernel, &i915_sys_placement, alignment,
+                         true, NULL, NULL, i915_ttm_bo_destroy);

Move this further up, or add some onion?

Might need to revisit that again to figure out what destructors are called autmatically on failure there. Thing is once ttm_bo_init has succeeded, it's published on LRU lists and available for mmap, so we'd better have everything initialized.

Quoted 2 lines

/Thomas

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

author reviewer
#46 lore raw

PATCH v2 11/15 drm/i915/lmem: Verify checks for lmem residency

In reply to Thomas Hellström

Since objects can be migrated or evicted when not pinned or locked, update the checks for lmem residency or future residency so that the value returned is not immediately stale.

---
v2: Simplify i915_gem_object_migratable() (Reported by Mattew Auld)

---
 drivers/gpu/drm/i915/display/intel_display.c |  2 +-
 drivers/gpu/drm/i915/gem/i915_gem_lmem.c     | 42 +++++++++++++++++++-
 drivers/gpu/drm/i915/gem/i915_gem_object.c   | 18 +++++++++
 drivers/gpu/drm/i915/gem/i915_gem_object.h   |  4 ++
 4 files changed, 64 insertions(+), 2 deletions(-)
drivers/gpu/drm/i915/display/intel_display.c +1 -1
index de1f13d203b5..b95def2d5af3 100644
@@ -11615,7 +11615,7 @@ intel_user_framebuffer_create(struct drm_device *dev,
 
 	/* object is backed with LMEM for discrete */
 	i915 = to_i915(obj->base.dev);
-	if (HAS_LMEM(i915) && !i915_gem_object_is_lmem(obj)) {
+	if (HAS_LMEM(i915) && !i915_gem_object_validates_to_lmem(obj)) {
 		/* object is "remote", not in local memory */
 		i915_gem_object_put(obj);
 		return ERR_PTR(-EREMOTE);
drivers/gpu/drm/i915/gem/i915_gem_lmem.c +41 -1

51 lines not on this page · raw message

drivers/gpu/drm/i915/gem/i915_gem_object.c +18 -0
index 8580996107ce..8484d8940531 100644
@@ -457,6 +457,24 @@ bool i915_gem_object_evictable(struct drm_i915_gem_object *obj)
 	return pin_count == 0;
 }
 
+/**
+ * i915_gem_object_migratable - Whether the object is migratable out of the
+ * current region.
+ * @obj: Pointer to the object.
+ *
+ * Return: Whether the object is allowed to be resident in other
+ * regions than the current while pages are present.
+ */
+bool i915_gem_object_migratable(struct drm_i915_gem_object *obj)
+{
+	struct intel_memory_region *mr = READ_ONCE(obj->mm.region);
+
+	if (!mr)
+		return false;
+
+	return obj->mm.n_placements > 1;
+}
+
 void i915_gem_init__objects(struct drm_i915_private *i915)
 {
 	INIT_WORK(&i915->mm.free_work, __i915_gem_free_work);
drivers/gpu/drm/i915/gem/i915_gem_object.h +4 -0
index ae5930e307d5..a3ad8cf4eefd 100644
@@ -596,6 +596,10 @@ void __i915_gem_free_object(struct drm_i915_gem_object *obj);
 
 bool i915_gem_object_evictable(struct drm_i915_gem_object *obj);
 
+bool i915_gem_object_migratable(struct drm_i915_gem_object *obj);
+
+bool i915_gem_object_validates_to_lmem(struct drm_i915_gem_object *obj);
+
 #ifdef CONFIG_MMU_NOTIFIER
 static inline bool
 i915_gem_object_is_userptr(struct drm_i915_gem_object *obj)
Signature · 6 lines
2.31.1

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
author reviewer
#48 lore raw

PATCH v2 12/15 drm/i915: Disable mmap ioctl for gen12+

In reply to Thomas Hellström

From: Maarten Lankhorst <[email protected]>

The paltform should exclusively use mmap_offset, one less path to worry about for discrete.

---
 drivers/gpu/drm/i915/gem/i915_gem_mman.c | 7 +++++++
 1 file changed, 7 insertions(+)
drivers/gpu/drm/i915/gem/i915_gem_mman.c +7 -0
index 8598a1c78a4c..65db290efd16 100644
@@ -56,10 +56,17 @@ int
 i915_gem_mmap_ioctl(struct drm_device *dev, void *data,
 		    struct drm_file *file)
 {
+	struct drm_i915_private *i915 = to_i915(dev);
 	struct drm_i915_gem_mmap *args = data;
 	struct drm_i915_gem_object *obj;
 	unsigned long addr;
 
+	/* mmap ioctl is disallowed for all platforms after TGL-LP.  This also
+	 * covers all platforms with local memory.
+	 */
+	if (INTEL_GEN(i915) >= 12 && !IS_TIGERLAKE(i915))
+		return -EOPNOTSUPP;
+
 	if (args->flags & ~(I915_MMAP_WC))
 		return -EINVAL;
 
Signature · 6 lines
2.31.1

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
author reviewer Reviewed-by on v2 12/15
#49 lore raw

Re: [Intel-gfx] [PATCH v2 12/15] drm/i915: Disable mmap ioctl for gen12+

Quoted 4 lines · Thomas Hellström

From: Maarten Lankhorst <[email protected]>

The paltform should exclusively use mmap_offset, one less path to worry about for discrete.

s/paltform/platform/

Quoted 2 lines

Otherwise,

Quoted 16 lines
---
  drivers/gpu/drm/i915/gem/i915_gem_mman.c | 7 +++++++
  1 file changed, 7 insertions(+)
drivers/gpu/drm/i915/gem/i915_gem_mman.c +1 -0
index 8598a1c78a4c..65db290efd16 100644
@@ -56,10 +56,17 @@ int
  i915_gem_mmap_ioctl(struct drm_device *dev, void *data,
  		    struct drm_file *file)
  {
+	struct drm_i915_private *i915 = to_i915(dev);
  	struct drm_i915_gem_mmap *args = data;
  	struct drm_i915_gem_object *obj;
  	unsigned long addr;

drivers/gpu/drm/i915/gem/i915_gem_mman.c int · line 72

Quoted 10 lines
drivers/gpu/drm/i915/gem/i915_gem_mman.c +6 -0
 
+	/* mmap ioctl is disallowed for all platforms after TGL-LP.  This also
+	 * covers all platforms with local memory.
+	 */
+	if (INTEL_GEN(i915) >= 12 && !IS_TIGERLAKE(i915))
+		return -EOPNOTSUPP;
+
  	if (args->flags & ~(I915_MMAP_WC))
  		return -EINVAL;
 

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

author reviewer
#50 lore raw

PATCH v2 13/15 drm/ttm: Add BO and offset arguments for vm_access and vm_fault ttm handlers.

In reply to Thomas Hellström

From: Maarten Lankhorst <[email protected]>

This allows other drivers that may not setup the vma in the same way to use the ttm bo helpers.

Also clarify the documentation a bit, especially related to VM_FAULT_RETRY.

---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c    |  4 +-
 drivers/gpu/drm/nouveau/nouveau_ttm.c      |  4 +-
 drivers/gpu/drm/radeon/radeon_ttm.c        |  4 +-
 drivers/gpu/drm/ttm/ttm_bo_vm.c            | 84 +++++++++++++---------
 drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c |  8 ++-
 include/drm/ttm/ttm_bo_api.h               |  9 ++-
 6 files changed, 75 insertions(+), 38 deletions(-)
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +3 -1
index d5a9d7a88315..89dafe14f828 100644
@@ -1919,7 +1919,9 @@ static vm_fault_t amdgpu_ttm_fault(struct vm_fault *vmf)
 	if (ret)
 		goto unlock;
 
-	ret = ttm_bo_vm_fault_reserved(vmf, vmf->vma->vm_page_prot,
+	ret = ttm_bo_vm_fault_reserved(bo, vmf,
+				       drm_vma_node_start(&bo->base.vma_node),
+				       vmf->vma->vm_page_prot,
 				       TTM_BO_VM_NUM_PREFAULT, 1);
 	if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))
 		return ret;
drivers/gpu/drm/nouveau/nouveau_ttm.c +3 -1
index b81ae90b8449..555fb6d8be8b 100644
@@ -144,7 +144,9 @@ static vm_fault_t nouveau_ttm_fault(struct vm_fault *vmf)
 
 	nouveau_bo_del_io_reserve_lru(bo);
 	prot = vm_get_page_prot(vma->vm_flags);
-	ret = ttm_bo_vm_fault_reserved(vmf, prot, TTM_BO_VM_NUM_PREFAULT, 1);
+	ret = ttm_bo_vm_fault_reserved(bo, vmf,
+				       drm_vma_node_start(&bo->base.vma_node),
+				       prot, TTM_BO_VM_NUM_PREFAULT, 1);
 	nouveau_bo_add_io_reserve_lru(bo);
 	if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))
 		return ret;
drivers/gpu/drm/radeon/radeon_ttm.c +3 -1
index 3361d11769a2..ba48a2acdef0 100644
@@ -816,7 +816,9 @@ static vm_fault_t radeon_ttm_fault(struct vm_fault *vmf)
 	if (ret)
 		goto unlock_resv;
 
-	ret = ttm_bo_vm_fault_reserved(vmf, vmf->vma->vm_page_prot,
+	ret = ttm_bo_vm_fault_reserved(bo, vmf,
+				       drm_vma_node_start(&bo->base.vma_node),
+				       vmf->vma->vm_page_prot,
 				       TTM_BO_VM_NUM_PREFAULT, 1);
 	if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))
 		goto unlock_mclk;
drivers/gpu/drm/ttm/ttm_bo_vm.c +52 -32

169 lines not on this page · raw message

drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c +6 -2
index 45c9c6a7f1d6..56ecace0cf5c 100644
@@ -477,7 +477,9 @@ vm_fault_t vmw_bo_vm_fault(struct vm_fault *vmf)
 	else
 		prot = vm_get_page_prot(vma->vm_flags);
 
-	ret = ttm_bo_vm_fault_reserved(vmf, prot, num_prefault, 1);
+	ret = ttm_bo_vm_fault_reserved(bo, vmf,
+				       drm_vma_node_start(&bo->base.vma_node),
+				       prot, num_prefault, 1);
 	if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))
 		return ret;
 
@@ -546,7 +548,9 @@ vm_fault_t vmw_bo_vm_huge_fault(struct vm_fault *vmf,
 		prot = vm_get_page_prot(vma->vm_flags);
 	}
 
-	ret = ttm_bo_vm_fault_reserved(vmf, prot, 1, fault_page_size);
+	ret = ttm_bo_vm_fault_reserved(bo, vmf,
+				       drm_vma_node_start(&bo->base.vma_node),
+				       prot, 1, fault_page_size);
 	if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))
 		return ret;
 
include/drm/ttm/ttm_bo_api.h +8 -1
index 639521880c29..434f91f1fdbf 100644
@@ -605,7 +605,9 @@ int ttm_mem_evict_first(struct ttm_device *bdev,
 vm_fault_t ttm_bo_vm_reserve(struct ttm_buffer_object *bo,
 			     struct vm_fault *vmf);
 
-vm_fault_t ttm_bo_vm_fault_reserved(struct vm_fault *vmf,
+vm_fault_t ttm_bo_vm_fault_reserved(struct ttm_buffer_object *bo,
+				    struct vm_fault *vmf,
+				    unsigned long mmap_base,
 				    pgprot_t prot,
 				    pgoff_t num_prefault,
 				    pgoff_t fault_page_size);
@@ -616,6 +618,11 @@ void ttm_bo_vm_open(struct vm_area_struct *vma);
 
 void ttm_bo_vm_close(struct vm_area_struct *vma);
 
+int ttm_bo_vm_access_reserved(struct ttm_buffer_object *bo,
+			      struct vm_area_struct *vma,
+			      unsigned long offset,
+			      void *buf, int len, int write);
+
 int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
 		     void *buf, int len, int write);
 bool ttm_bo_delayed_delete(struct ttm_device *bdev, bool remove_all);
Signature · 1 line
2.31.1
author reviewer
#51 lore raw

Re: [Intel-gfx] [PATCH v2 13/15] drm/ttm: Add BO and offset arguments for vm_access and vm_fault ttm handlers.

+ Christian König

Quoted 8 lines · Thomas Hellström

From: Maarten Lankhorst <[email protected]>

This allows other drivers that may not setup the vma in the same way to use the ttm bo helpers.

Also clarify the documentation a bit, especially related to VM_FAULT_RETRY.

Lgtm. Reviewed-by: Thomas Hellström <[email protected]>

Quoted 290 lines
---
  drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c    |  4 +-
  drivers/gpu/drm/nouveau/nouveau_ttm.c      |  4 +-
  drivers/gpu/drm/radeon/radeon_ttm.c        |  4 +-
  drivers/gpu/drm/ttm/ttm_bo_vm.c            | 84 +++++++++++++---------
  drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c |  8 ++-
  include/drm/ttm/ttm_bo_api.h               |  9 ++-
  6 files changed, 75 insertions(+), 38 deletions(-)
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +3 -1
index d5a9d7a88315..89dafe14f828 100644
@@ -1919,7 +1919,9 @@ static vm_fault_t amdgpu_ttm_fault(struct vm_fault *vmf)
  	if (ret)
  		goto unlock;
 
-	ret = ttm_bo_vm_fault_reserved(vmf, vmf->vma->vm_page_prot,
+	ret = ttm_bo_vm_fault_reserved(bo, vmf,
+				       drm_vma_node_start(&bo->base.vma_node),
+				       vmf->vma->vm_page_prot,
  				       TTM_BO_VM_NUM_PREFAULT, 1);
  	if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))
  		return ret;
drivers/gpu/drm/nouveau/nouveau_ttm.c +3 -1
index b81ae90b8449..555fb6d8be8b 100644
@@ -144,7 +144,9 @@ static vm_fault_t nouveau_ttm_fault(struct vm_fault *vmf)
 
  	nouveau_bo_del_io_reserve_lru(bo);
  	prot = vm_get_page_prot(vma->vm_flags);
-	ret = ttm_bo_vm_fault_reserved(vmf, prot, TTM_BO_VM_NUM_PREFAULT, 1);
+	ret = ttm_bo_vm_fault_reserved(bo, vmf,
+				       drm_vma_node_start(&bo->base.vma_node),
+				       prot, TTM_BO_VM_NUM_PREFAULT, 1);
  	nouveau_bo_add_io_reserve_lru(bo);
  	if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))
  		return ret;
drivers/gpu/drm/radeon/radeon_ttm.c +3 -1
index 3361d11769a2..ba48a2acdef0 100644
@@ -816,7 +816,9 @@ static vm_fault_t radeon_ttm_fault(struct vm_fault *vmf)
  	if (ret)
  		goto unlock_resv;
 
-	ret = ttm_bo_vm_fault_reserved(vmf, vmf->vma->vm_page_prot,
+	ret = ttm_bo_vm_fault_reserved(bo, vmf,
+				       drm_vma_node_start(&bo->base.vma_node),
+				       vmf->vma->vm_page_prot,
  				       TTM_BO_VM_NUM_PREFAULT, 1);
  	if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))
  		goto unlock_mclk;
drivers/gpu/drm/ttm/ttm_bo_vm.c +52 -32
index b31b18058965..ed00ccf1376e 100644
@@ -42,7 +42,7 @@
  #include <linux/mem_encrypt.h>
 
  static vm_fault_t ttm_bo_vm_fault_idle(struct ttm_buffer_object *bo,
-				struct vm_fault *vmf)
+				       struct vm_fault *vmf)
  {
  	vm_fault_t ret = 0;
  	int err = 0;
@@ -122,7 +122,8 @@ static unsigned long ttm_bo_io_mem_pfn(struct ttm_buffer_object *bo,
   * Return:
   *    0 on success and the bo was reserved.
   *    VM_FAULT_RETRY if blocking wait.
- *    VM_FAULT_NOPAGE if blocking wait and retrying was not allowed.
+ *    VM_FAULT_NOPAGE if blocking wait and retrying was not allowed, or wait interrupted.
+ *    VM_FAULT_SIGBUS if wait on bo->moving failed for reason other than a signal.
   */
  vm_fault_t ttm_bo_vm_reserve(struct ttm_buffer_object *bo,
  			     struct vm_fault *vmf)
@@ -254,7 +255,9 @@ static vm_fault_t ttm_bo_vm_insert_huge(struct vm_fault *vmf,
 
  /**
   * ttm_bo_vm_fault_reserved - TTM fault helper
+ * @bo: The buffer object
   * @vmf: The struct vm_fault given as argument to the fault callback
+ * @mmap_base: The base of the mmap, to which the @vmf fault is relative to.
   * @prot: The page protection to be used for this memory area.
   * @num_prefault: Maximum number of prefault pages. The caller may want to
   * specify this based on madvice settings and the size of the GPU object
@@ -265,19 +268,28 @@ static vm_fault_t ttm_bo_vm_insert_huge(struct vm_fault *vmf,
   * memory backing the buffer object, and then returns a return code
   * instructing the caller to retry the page access.
   *
+ * This function ensures any pipelined wait is finished.
+ *
+ * WARNING:
+ * On VM_FAULT_RETRY, the bo will be unlocked by this function when
+ * #FAULT_FLAG_RETRY_NOWAIT is not set inside @vmf->flags. In this
+ * case, the caller should not unlock the @bo.
+ *
   * Return:
- *   VM_FAULT_NOPAGE on success or pending signal
+ *   0 on success.
+ *   VM_FAULT_NOPAGE on pending signal
   *   VM_FAULT_SIGBUS on unspecified error
   *   VM_FAULT_OOM on out-of-memory
- *   VM_FAULT_RETRY if retryable wait
+ *   VM_FAULT_RETRY if retryable wait, see WARNING above.
   */
-vm_fault_t ttm_bo_vm_fault_reserved(struct vm_fault *vmf,
+vm_fault_t ttm_bo_vm_fault_reserved(struct ttm_buffer_object *bo,
+				    struct vm_fault *vmf,
+				    unsigned long mmap_base,
  				    pgprot_t prot,
  				    pgoff_t num_prefault,
  				    pgoff_t fault_page_size)
  {
  	struct vm_area_struct *vma = vmf->vma;
-	struct ttm_buffer_object *bo = vma->vm_private_data;
  	struct ttm_device *bdev = bo->bdev;
  	unsigned long page_offset;
  	unsigned long page_last;
@@ -286,15 +298,11 @@ vm_fault_t ttm_bo_vm_fault_reserved(struct vm_fault *vmf,
  	struct page *page;
  	int err;
  	pgoff_t i;
-	vm_fault_t ret = VM_FAULT_NOPAGE;
+	vm_fault_t ret;
  	unsigned long address = vmf->address;
 
-	/*
-	 * Wait for buffer data in transit, due to a pipelined
-	 * move.
-	 */
  	ret = ttm_bo_vm_fault_idle(bo, vmf);
-	if (unlikely(ret != 0))
+	if (ret)
  		return ret;
 
  	err = ttm_mem_io_reserve(bdev, &bo->mem);
@@ -302,9 +310,8 @@ vm_fault_t ttm_bo_vm_fault_reserved(struct vm_fault *vmf,
  		return VM_FAULT_SIGBUS;
 
  	page_offset = ((address - vma->vm_start) >> PAGE_SHIFT) +
-		vma->vm_pgoff - drm_vma_node_start(&bo->base.vma_node);
-	page_last = vma_pages(vma) + vma->vm_pgoff -
-		drm_vma_node_start(&bo->base.vma_node);
+		vma->vm_pgoff - mmap_base;
+	page_last = vma_pages(vma) + vma->vm_pgoff - mmap_base;
 
  	if (unlikely(page_offset >= bo->mem.num_pages))
  		return VM_FAULT_SIGBUS;
@@ -344,8 +351,7 @@ vm_fault_t ttm_bo_vm_fault_reserved(struct vm_fault *vmf,
  			} else if (unlikely(!page)) {
  				break;
  			}
-			page->index = drm_vma_node_start(&bo->base.vma_node) +
-				page_offset;
+			page->index = mmap_base + page_offset;
  			pfn = page_to_pfn(page);
  		}
 
@@ -392,7 +398,10 @@ vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf)
  		return ret;
 
  	prot = vma->vm_page_prot;
-	ret = ttm_bo_vm_fault_reserved(vmf, prot, TTM_BO_VM_NUM_PREFAULT, 1);
+	ret = ttm_bo_vm_fault_reserved(bo, vmf,
+				       drm_vma_node_start(&bo->base.vma_node),
+				       prot, TTM_BO_VM_NUM_PREFAULT, 1);
+
  	if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))
  		return ret;
 
@@ -460,22 +469,16 @@ static int ttm_bo_vm_access_kmap(struct ttm_buffer_object *bo,
  	return len;
  }
 
-int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
-		     void *buf, int len, int write)
+int ttm_bo_vm_access_reserved(struct ttm_buffer_object *bo,
+			      struct vm_area_struct *vma,
+			      unsigned long offset,
+			      void *buf, int len, int write)
  {
-	struct ttm_buffer_object *bo = vma->vm_private_data;
-	unsigned long offset = (addr) - vma->vm_start +
-		((vma->vm_pgoff - drm_vma_node_start(&bo->base.vma_node))
-		 << PAGE_SHIFT);
  	int ret;
 
  	if (len < 1 || (offset + len) >> PAGE_SHIFT > bo->mem.num_pages)
  		return -EIO;
 
-	ret = ttm_bo_reserve(bo, true, false, NULL);
-	if (ret)
-		return ret;
-
  	switch (bo->mem.mem_type) {
  	case TTM_PL_SYSTEM:
  		if (unlikely(bo->ttm->page_flags & TTM_PAGE_FLAG_SWAPPED)) {
@@ -485,16 +488,33 @@ int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
  		}
  		fallthrough;
  	case TTM_PL_TT:
-		ret = ttm_bo_vm_access_kmap(bo, offset, buf, len, write);
-		break;
+		return ttm_bo_vm_access_kmap(bo, offset, buf, len, write);
  	default:
  		if (bo->bdev->funcs->access_memory)
-			ret = bo->bdev->funcs->access_memory(
+			return bo->bdev->funcs->access_memory(
  				bo, offset, buf, len, write);
  		else
-			ret = -EIO;
+			return -EIO;
  	}
 
+	return ret;
+}
+EXPORT_SYMBOL(ttm_bo_vm_access_reserved);
+
+int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
+		     void *buf, int len, int write)
+{
+	struct ttm_buffer_object *bo = vma->vm_private_data;
+	unsigned long offset = (addr) - vma->vm_start +
+		((vma->vm_pgoff - drm_vma_node_start(&bo->base.vma_node))
+		 << PAGE_SHIFT);
+	int ret;
+
+	ret = ttm_bo_reserve(bo, true, false, NULL);
+	if (ret)
+		return ret;
+
+	ret = ttm_bo_vm_access_reserved(bo, vma, offset, buf, len, write);
  	ttm_bo_unreserve(bo);
 
  	return ret;
drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c +6 -2
index 45c9c6a7f1d6..56ecace0cf5c 100644
@@ -477,7 +477,9 @@ vm_fault_t vmw_bo_vm_fault(struct vm_fault *vmf)
  	else
  		prot = vm_get_page_prot(vma->vm_flags);
 
-	ret = ttm_bo_vm_fault_reserved(vmf, prot, num_prefault, 1);
+	ret = ttm_bo_vm_fault_reserved(bo, vmf,
+				       drm_vma_node_start(&bo->base.vma_node),
+				       prot, num_prefault, 1);
  	if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))
  		return ret;
 
@@ -546,7 +548,9 @@ vm_fault_t vmw_bo_vm_huge_fault(struct vm_fault *vmf,
  		prot = vm_get_page_prot(vma->vm_flags);
  	}
 
-	ret = ttm_bo_vm_fault_reserved(vmf, prot, 1, fault_page_size);
+	ret = ttm_bo_vm_fault_reserved(bo, vmf,
+				       drm_vma_node_start(&bo->base.vma_node),
+				       prot, 1, fault_page_size);
  	if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))
  		return ret;
 
include/drm/ttm/ttm_bo_api.h +8 -1
index 639521880c29..434f91f1fdbf 100644
@@ -605,7 +605,9 @@ int ttm_mem_evict_first(struct ttm_device *bdev,
  vm_fault_t ttm_bo_vm_reserve(struct ttm_buffer_object *bo,
  			     struct vm_fault *vmf);
 
-vm_fault_t ttm_bo_vm_fault_reserved(struct vm_fault *vmf,
+vm_fault_t ttm_bo_vm_fault_reserved(struct ttm_buffer_object *bo,
+				    struct vm_fault *vmf,
+				    unsigned long mmap_base,
  				    pgprot_t prot,
  				    pgoff_t num_prefault,
  				    pgoff_t fault_page_size);
@@ -616,6 +618,11 @@ void ttm_bo_vm_open(struct vm_area_struct *vma);
 
  void ttm_bo_vm_close(struct vm_area_struct *vma);
 
+int ttm_bo_vm_access_reserved(struct ttm_buffer_object *bo,
+			      struct vm_area_struct *vma,
+			      unsigned long offset,
+			      void *buf, int len, int write);
+
  int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
  		     void *buf, int len, int write);
  bool ttm_bo_delayed_delete(struct ttm_device *bdev, bool remove_all);
maintainer
#52 lore raw

Re: [Intel-gfx] [PATCH v2 13/15] drm/ttm: Add BO and offset arguments for vm_access and vm_fault ttm handlers.

Can you send me the patch directly and not just on CC?

Thanks,
Christian.

Am 18.05.21 um 10:59 schrieb Thomas Hellström:

Quoted 295 lines

+ Christian König

Quoted 8 lines · Thomas Hellström

From: Maarten Lankhorst <[email protected]>

This allows other drivers that may not setup the vma in the same way to use the ttm bo helpers.

Also clarify the documentation a bit, especially related to VM_FAULT_RETRY.

Lgtm. Reviewed-by: Thomas Hellström <[email protected]>

Quoted 281 lines

---
  drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c    |  4 +- drivers/gpu/drm/nouveau/nouveau_ttm.c      |  4 +- drivers/gpu/drm/radeon/radeon_ttm.c        |  4 +- drivers/gpu/drm/ttm/ttm_bo_vm.c            | 84 +++++++++++++--------- drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c |  8 ++- include/drm/ttm/ttm_bo_api.h               |  9 ++-

  6 files changed, 75 insertions(+), 38 deletions(-)
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +3 -0
index d5a9d7a88315..89dafe14f828 100644
@@ -1919,7 +1919,9 @@ static vm_fault_t amdgpu_ttm_fault(struct 
 vm_fault *vmf)
       if (ret)
           goto unlock;
   -    ret = ttm_bo_vm_fault_reserved(vmf, vmf->vma->vm_page_prot,
+    ret = ttm_bo_vm_fault_reserved(bo, vmf,
+ drm_vma_node_start(&bo->base.vma_node),
+                       vmf->vma->vm_page_prot,
                          TTM_BO_VM_NUM_PREFAULT, 1);
       if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))

          return ret;

drivers/gpu/drm/nouveau/nouveau_ttm.c +3 -1
index b81ae90b8449..555fb6d8be8b 100644
@@ -144,7 +144,9 @@ static vm_fault_t nouveau_ttm_fault(struct 
 vm_fault *vmf)
         nouveau_bo_del_io_reserve_lru(bo);
       prot = vm_get_page_prot(vma->vm_flags);
-    ret = ttm_bo_vm_fault_reserved(vmf, prot, 
 TTM_BO_VM_NUM_PREFAULT, 1);
+    ret = ttm_bo_vm_fault_reserved(bo, vmf,
+ drm_vma_node_start(&bo->base.vma_node),
+                       prot, TTM_BO_VM_NUM_PREFAULT, 1);
       nouveau_bo_add_io_reserve_lru(bo);
       if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))

          return ret;

drivers/gpu/drm/radeon/radeon_ttm.c +3 -0
index 3361d11769a2..ba48a2acdef0 100644
@@ -816,7 +816,9 @@ static vm_fault_t radeon_ttm_fault(struct 
 vm_fault *vmf)
       if (ret)
           goto unlock_resv;
   -    ret = ttm_bo_vm_fault_reserved(vmf, vmf->vma->vm_page_prot,
+    ret = ttm_bo_vm_fault_reserved(bo, vmf,
+ drm_vma_node_start(&bo->base.vma_node),
+                       vmf->vma->vm_page_prot,
                          TTM_BO_VM_NUM_PREFAULT, 1);
       if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))

          goto unlock_mclk;

drivers/gpu/drm/ttm/ttm_bo_vm.c +2 -2
index b31b18058965..ed00ccf1376e 100644
@@ -42,7 +42,7 @@
   #include <linux/mem_encrypt.h>
     static vm_fault_t ttm_bo_vm_fault_idle(struct ttm_buffer_object *bo,
-                struct vm_fault *vmf)
+                       struct vm_fault *vmf)
   {
       vm_fault_t ret = 0;
       int err = 0;
@@ -122,7 +122,8 @@ static unsigned long ttm_bo_io_mem_pfn(struct 
 ttm_buffer_object *bo,
    * Return:
    *    0 on success and the bo was reserved.
    *    VM_FAULT_RETRY if blocking wait.
- *    VM_FAULT_NOPAGE if blocking wait and retrying was not allowed.
+ *    VM_FAULT_NOPAGE if blocking wait and retrying was not allowed, 

or wait interrupted.
+ *    VM_FAULT_SIGBUS if wait on bo->moving failed for reason other than a signal.
   */
  vm_fault_t ttm_bo_vm_reserve(struct ttm_buffer_object *bo, struct vm_fault *vmf)

drivers/gpu/drm/ttm/ttm_bo_vm.c +2 -0
@@ -254,7 +255,9 @@ static vm_fault_t ttm_bo_vm_insert_huge(struct 
 vm_fault *vmf,
     /**
    * ttm_bo_vm_fault_reserved - TTM fault helper
+ * @bo: The buffer object
    * @vmf: The struct vm_fault given as argument to the fault callback
+ * @mmap_base: The base of the mmap, to which the @vmf fault is 

relative to.
   * @prot: The page protection to be used for this memory area.
   * @num_prefault: Maximum number of prefault pages. The caller may want to
   * specify this based on madvice settings and the size of the GPU object

drivers/gpu/drm/ttm/ttm_bo_vm.c +13 -4
@@ -265,19 +268,28 @@ static vm_fault_t ttm_bo_vm_insert_huge(struct 
 vm_fault *vmf,
    * memory backing the buffer object, and then returns a return code
    * instructing the caller to retry the page access.
    *
+ * This function ensures any pipelined wait is finished.
+ *
+ * WARNING:
+ * On VM_FAULT_RETRY, the bo will be unlocked by this function when
+ * #FAULT_FLAG_RETRY_NOWAIT is not set inside @vmf->flags. In this
+ * case, the caller should not unlock the @bo.
+ *
    * Return:
- *   VM_FAULT_NOPAGE on success or pending signal
+ *   0 on success.
+ *   VM_FAULT_NOPAGE on pending signal
    *   VM_FAULT_SIGBUS on unspecified error
    *   VM_FAULT_OOM on out-of-memory
- *   VM_FAULT_RETRY if retryable wait
+ *   VM_FAULT_RETRY if retryable wait, see WARNING above.
    */
-vm_fault_t ttm_bo_vm_fault_reserved(struct vm_fault *vmf,
+vm_fault_t ttm_bo_vm_fault_reserved(struct ttm_buffer_object *bo,
+                    struct vm_fault *vmf,
+                    unsigned long mmap_base,
                       pgprot_t prot,
                       pgoff_t num_prefault,
                       pgoff_t fault_page_size)
   {
       struct vm_area_struct *vma = vmf->vma;
-    struct ttm_buffer_object *bo = vma->vm_private_data;
       struct ttm_device *bdev = bo->bdev;
       unsigned long page_offset;

      unsigned long page_last;

drivers/gpu/drm/ttm/ttm_bo_vm.c +1 -1
@@ -286,15 +298,11 @@ vm_fault_t ttm_bo_vm_fault_reserved(struct 
 vm_fault *vmf,
       struct page *page;
       int err;
       pgoff_t i;
-    vm_fault_t ret = VM_FAULT_NOPAGE;
+    vm_fault_t ret;
       unsigned long address = vmf->address;

  -    /*
-     * Wait for buffer data in transit, due to a pipelined -     * move.
-     */
      ret = ttm_bo_vm_fault_idle(bo, vmf);
-    if (unlikely(ret != 0))
+    if (ret)
          return ret;
        err = ttm_mem_io_reserve(bdev, &bo->mem);

drivers/gpu/drm/ttm/ttm_bo_vm.c +0 -0
@@ -302,9 +310,8 @@ vm_fault_t ttm_bo_vm_fault_reserved(struct 
 vm_fault *vmf,
           return VM_FAULT_SIGBUS;

        page_offset = ((address - vma->vm_start) >> PAGE_SHIFT) + -        vma->vm_pgoff - drm_vma_node_start(&bo->base.vma_node); -    page_last = vma_pages(vma) + vma->vm_pgoff -
-        drm_vma_node_start(&bo->base.vma_node);
+        vma->vm_pgoff - mmap_base;
+    page_last = vma_pages(vma) + vma->vm_pgoff - mmap_base; if (unlikely(page_offset >= bo->mem.num_pages)) return VM_FAULT_SIGBUS;

drivers/gpu/drm/ttm/ttm_bo_vm.c +1 -2
@@ -344,8 +351,7 @@ vm_fault_t ttm_bo_vm_fault_reserved(struct 
 vm_fault *vmf,
               } else if (unlikely(!page)) {
                   break;
               }
-            page->index = drm_vma_node_start(&bo->base.vma_node) +
-                page_offset;
+            page->index = mmap_base + page_offset;
               pfn = page_to_pfn(page);
           }

  @@ -392,7 +398,10 @@ vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf) return ret;
        prot = vma->vm_page_prot;
-    ret = ttm_bo_vm_fault_reserved(vmf, prot,
TTM_BO_VM_NUM_PREFAULT, 1);
+    ret = ttm_bo_vm_fault_reserved(bo, vmf,
+ drm_vma_node_start(&bo->base.vma_node),
+                       prot, TTM_BO_VM_NUM_PREFAULT, 1);
+
      if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT)) return ret;
  @@ -460,22 +469,16 @@ static int ttm_bo_vm_access_kmap(struct ttm_buffer_object *bo, return len;
  }
  -int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr, -             void *buf, int len, int write)
+int ttm_bo_vm_access_reserved(struct ttm_buffer_object *bo,
+                  struct vm_area_struct *vma,
+                  unsigned long offset,
+                  void *buf, int len, int write)
  {
-    struct ttm_buffer_object *bo = vma->vm_private_data;
-    unsigned long offset = (addr) - vma->vm_start +
-        ((vma->vm_pgoff - drm_vma_node_start(&bo->base.vma_node))
-         << PAGE_SHIFT);
      int ret;
        if (len < 1 || (offset + len) >> PAGE_SHIFT > bo->mem.num_pages) return -EIO;
  -    ret = ttm_bo_reserve(bo, true, false, NULL);
-    if (ret)
-        return ret;
-
      switch (bo->mem.mem_type) {
      case TTM_PL_SYSTEM:
          if (unlikely(bo->ttm->page_flags & TTM_PAGE_FLAG_SWAPPED)) {

drivers/gpu/drm/ttm/ttm_bo_vm.c +0 -0
@@ -485,16 +488,33 @@ int ttm_bo_vm_access(struct vm_area_struct 
 *vma, unsigned long addr,
           }
           fallthrough;

      case TTM_PL_TT:
-        ret = ttm_bo_vm_access_kmap(bo, offset, buf, len, write); -        break;
+        return ttm_bo_vm_access_kmap(bo, offset, buf, len, write); default:
          if (bo->bdev->funcs->access_memory)
-            ret = bo->bdev->funcs->access_memory(
+            return bo->bdev->funcs->access_memory(
                  bo, offset, buf, len, write);
          else
-            ret = -EIO;
+            return -EIO;
      }
  +    return ret;

drivers/gpu/drm/ttm/ttm_bo_vm.c +17 -0
+}
+EXPORT_SYMBOL(ttm_bo_vm_access_reserved);
+
+int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
+             void *buf, int len, int write)
+{
+    struct ttm_buffer_object *bo = vma->vm_private_data;
+    unsigned long offset = (addr) - vma->vm_start +
+        ((vma->vm_pgoff - drm_vma_node_start(&bo->base.vma_node))
+         << PAGE_SHIFT);
+    int ret;
+
+    ret = ttm_bo_reserve(bo, true, false, NULL);
+    if (ret)
+        return ret;
+
+    ret = ttm_bo_vm_access_reserved(bo, vma, offset, buf, len, write);

      ttm_bo_unreserve(bo);
        return ret;

drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c +0 -0
index 45c9c6a7f1d6..56ecace0cf5c 100644
@@ -477,7 +477,9 @@ vm_fault_t vmw_bo_vm_fault(struct vm_fault *vmf)

      else
          prot = vm_get_page_prot(vma->vm_flags);
  -    ret = ttm_bo_vm_fault_reserved(vmf, prot, num_prefault, 1); +    ret = ttm_bo_vm_fault_reserved(bo, vmf,
+ drm_vma_node_start(&bo->base.vma_node),
+                       prot, num_prefault, 1);
      if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT)) return ret;
  @@ -546,7 +548,9 @@ vm_fault_t vmw_bo_vm_huge_fault(struct vm_fault *vmf, prot = vm_get_page_prot(vma->vm_flags);
      }
  -    ret = ttm_bo_vm_fault_reserved(vmf, prot, 1, fault_page_size); +    ret = ttm_bo_vm_fault_reserved(bo, vmf,
+ drm_vma_node_start(&bo->base.vma_node),
+                       prot, 1, fault_page_size);
      if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT)) return ret;
  diff --git a/include/drm/ttm/ttm_bo_api.h b/include/drm/ttm/ttm_bo_api.h index 639521880c29..434f91f1fdbf 100644

include/drm/ttm/ttm_bo_api.h +7 -0
@@ -605,7 +605,9 @@ int ttm_mem_evict_first(struct ttm_device *bdev,
   vm_fault_t ttm_bo_vm_reserve(struct ttm_buffer_object *bo,
                    struct vm_fault *vmf);
   -vm_fault_t ttm_bo_vm_fault_reserved(struct vm_fault *vmf,
+vm_fault_t ttm_bo_vm_fault_reserved(struct ttm_buffer_object *bo,
+                    struct vm_fault *vmf,
+                    unsigned long mmap_base,
                       pgprot_t prot,
                       pgoff_t num_prefault,
                       pgoff_t fault_page_size);
@@ -616,6 +618,11 @@ void ttm_bo_vm_open(struct vm_area_struct *vma);
     void ttm_bo_vm_close(struct vm_area_struct *vma);
   +int ttm_bo_vm_access_reserved(struct ttm_buffer_object *bo,
+                  struct vm_area_struct *vma,
+                  unsigned long offset,
+                  void *buf, int len, int write);
+
   int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
                void *buf, int len, int write);
   bool ttm_bo_delayed_delete(struct ttm_device *bdev, bool remove_all);
author reviewer
#53 lore raw

Re: [Intel-gfx] [PATCH v2 13/15] drm/ttm: Add BO and offset arguments for vm_access and vm_fault ttm handlers.

Quoted 4 lines · Christian König

Can you send me the patch directly and not just on CC?

Thanks,
Christian.

Original patch sent. Pls remember to CC lists on reply, though.

The reason we need this is because of i915's strange mmap functionality which allows a bo to be mapped at multiple offsets and  the vma->private is not a bo...

Thanks,

Thomas

Quoted 290 lines

Am 18.05.21 um 10:59 schrieb Thomas Hellström:

Quoted 288 lines

+ Christian König

Quoted 8 lines · Thomas Hellström

From: Maarten Lankhorst <[email protected]>

This allows other drivers that may not setup the vma in the same way to use the ttm bo helpers.

Also clarify the documentation a bit, especially related to VM_FAULT_RETRY.

Lgtm. Reviewed-by: Thomas Hellström <[email protected]>

Quoted 274 lines

---
  drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c    |  4 +- drivers/gpu/drm/nouveau/nouveau_ttm.c      |  4 +- drivers/gpu/drm/radeon/radeon_ttm.c        |  4 +- drivers/gpu/drm/ttm/ttm_bo_vm.c            | 84
+++++++++++++---------
  drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c |  8 ++- include/drm/ttm/ttm_bo_api.h               |  9 ++-

  6 files changed, 75 insertions(+), 38 deletions(-)
drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +3 -0
index d5a9d7a88315..89dafe14f828 100644
@@ -1919,7 +1919,9 @@ static vm_fault_t amdgpu_ttm_fault(struct 
 vm_fault *vmf)
       if (ret)
           goto unlock;
   -    ret = ttm_bo_vm_fault_reserved(vmf, vmf->vma->vm_page_prot,
+    ret = ttm_bo_vm_fault_reserved(bo, vmf,
+ drm_vma_node_start(&bo->base.vma_node),
+                       vmf->vma->vm_page_prot,
                          TTM_BO_VM_NUM_PREFAULT, 1);
       if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))

          return ret;

drivers/gpu/drm/nouveau/nouveau_ttm.c +3 -1
index b81ae90b8449..555fb6d8be8b 100644
@@ -144,7 +144,9 @@ static vm_fault_t nouveau_ttm_fault(struct 
 vm_fault *vmf)
         nouveau_bo_del_io_reserve_lru(bo);
       prot = vm_get_page_prot(vma->vm_flags);
-    ret = ttm_bo_vm_fault_reserved(vmf, prot, 
 TTM_BO_VM_NUM_PREFAULT, 1);
+    ret = ttm_bo_vm_fault_reserved(bo, vmf,
+ drm_vma_node_start(&bo->base.vma_node),
+                       prot, TTM_BO_VM_NUM_PREFAULT, 1);
       nouveau_bo_add_io_reserve_lru(bo);
       if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))

          return ret;

drivers/gpu/drm/radeon/radeon_ttm.c +3 -0
index 3361d11769a2..ba48a2acdef0 100644
@@ -816,7 +816,9 @@ static vm_fault_t radeon_ttm_fault(struct 
 vm_fault *vmf)
       if (ret)
           goto unlock_resv;
   -    ret = ttm_bo_vm_fault_reserved(vmf, vmf->vma->vm_page_prot,
+    ret = ttm_bo_vm_fault_reserved(bo, vmf,
+ drm_vma_node_start(&bo->base.vma_node),
+                       vmf->vma->vm_page_prot,
                          TTM_BO_VM_NUM_PREFAULT, 1);
       if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))

          goto unlock_mclk;

drivers/gpu/drm/ttm/ttm_bo_vm.c +2 -2
index b31b18058965..ed00ccf1376e 100644
@@ -42,7 +42,7 @@
   #include <linux/mem_encrypt.h>
     static vm_fault_t ttm_bo_vm_fault_idle(struct ttm_buffer_object *bo,
-                struct vm_fault *vmf)
+                       struct vm_fault *vmf)
   {
       vm_fault_t ret = 0;
       int err = 0;
@@ -122,7 +122,8 @@ static unsigned long ttm_bo_io_mem_pfn(struct 
 ttm_buffer_object *bo,
    * Return:
    *    0 on success and the bo was reserved.
    *    VM_FAULT_RETRY if blocking wait.
- *    VM_FAULT_NOPAGE if blocking wait and retrying was not allowed.
+ *    VM_FAULT_NOPAGE if blocking wait and retrying was not 

allowed, or wait interrupted.
+ *    VM_FAULT_SIGBUS if wait on bo->moving failed for reason other than a signal.
   */
  vm_fault_t ttm_bo_vm_reserve(struct ttm_buffer_object *bo, struct vm_fault *vmf)

drivers/gpu/drm/ttm/ttm_bo_vm.c +2 -0
@@ -254,7 +255,9 @@ static vm_fault_t ttm_bo_vm_insert_huge(struct 
 vm_fault *vmf,
     /**
    * ttm_bo_vm_fault_reserved - TTM fault helper
+ * @bo: The buffer object
    * @vmf: The struct vm_fault given as argument to the fault callback
+ * @mmap_base: The base of the mmap, to which the @vmf fault is 

relative to.
   * @prot: The page protection to be used for this memory area.
   * @num_prefault: Maximum number of prefault pages. The caller may want to
   * specify this based on madvice settings and the size of the GPU object

drivers/gpu/drm/ttm/ttm_bo_vm.c +13 -4
@@ -265,19 +268,28 @@ static vm_fault_t ttm_bo_vm_insert_huge(struct 
 vm_fault *vmf,
    * memory backing the buffer object, and then returns a return code
    * instructing the caller to retry the page access.
    *
+ * This function ensures any pipelined wait is finished.
+ *
+ * WARNING:
+ * On VM_FAULT_RETRY, the bo will be unlocked by this function when
+ * #FAULT_FLAG_RETRY_NOWAIT is not set inside @vmf->flags. In this
+ * case, the caller should not unlock the @bo.
+ *
    * Return:
- *   VM_FAULT_NOPAGE on success or pending signal
+ *   0 on success.
+ *   VM_FAULT_NOPAGE on pending signal
    *   VM_FAULT_SIGBUS on unspecified error
    *   VM_FAULT_OOM on out-of-memory
- *   VM_FAULT_RETRY if retryable wait
+ *   VM_FAULT_RETRY if retryable wait, see WARNING above.
    */
-vm_fault_t ttm_bo_vm_fault_reserved(struct vm_fault *vmf,
+vm_fault_t ttm_bo_vm_fault_reserved(struct ttm_buffer_object *bo,
+                    struct vm_fault *vmf,
+                    unsigned long mmap_base,
                       pgprot_t prot,
                       pgoff_t num_prefault,
                       pgoff_t fault_page_size)
   {
       struct vm_area_struct *vma = vmf->vma;
-    struct ttm_buffer_object *bo = vma->vm_private_data;
       struct ttm_device *bdev = bo->bdev;
       unsigned long page_offset;

      unsigned long page_last;

drivers/gpu/drm/ttm/ttm_bo_vm.c +1 -1
@@ -286,15 +298,11 @@ vm_fault_t ttm_bo_vm_fault_reserved(struct 
 vm_fault *vmf,
       struct page *page;
       int err;
       pgoff_t i;
-    vm_fault_t ret = VM_FAULT_NOPAGE;
+    vm_fault_t ret;
       unsigned long address = vmf->address;

  -    /*
-     * Wait for buffer data in transit, due to a pipelined -     * move.
-     */
      ret = ttm_bo_vm_fault_idle(bo, vmf);
-    if (unlikely(ret != 0))
+    if (ret)
          return ret;
        err = ttm_mem_io_reserve(bdev, &bo->mem);

drivers/gpu/drm/ttm/ttm_bo_vm.c +0 -0
@@ -302,9 +310,8 @@ vm_fault_t ttm_bo_vm_fault_reserved(struct 
 vm_fault *vmf,
           return VM_FAULT_SIGBUS;

        page_offset = ((address - vma->vm_start) >> PAGE_SHIFT) + -        vma->vm_pgoff - drm_vma_node_start(&bo->base.vma_node); -    page_last = vma_pages(vma) + vma->vm_pgoff -
-        drm_vma_node_start(&bo->base.vma_node);
+        vma->vm_pgoff - mmap_base;
+    page_last = vma_pages(vma) + vma->vm_pgoff - mmap_base; if (unlikely(page_offset >= bo->mem.num_pages)) return VM_FAULT_SIGBUS;

drivers/gpu/drm/ttm/ttm_bo_vm.c +1 -2
@@ -344,8 +351,7 @@ vm_fault_t ttm_bo_vm_fault_reserved(struct 
 vm_fault *vmf,
               } else if (unlikely(!page)) {
                   break;
               }
-            page->index = drm_vma_node_start(&bo->base.vma_node) +
-                page_offset;
+            page->index = mmap_base + page_offset;
               pfn = page_to_pfn(page);
           }

  @@ -392,7 +398,10 @@ vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf) return ret;
        prot = vma->vm_page_prot;
-    ret = ttm_bo_vm_fault_reserved(vmf, prot,
TTM_BO_VM_NUM_PREFAULT, 1);
+    ret = ttm_bo_vm_fault_reserved(bo, vmf,
+ drm_vma_node_start(&bo->base.vma_node),
+                       prot, TTM_BO_VM_NUM_PREFAULT, 1);
+
      if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT)) return ret;
  @@ -460,22 +469,16 @@ static int ttm_bo_vm_access_kmap(struct ttm_buffer_object *bo, return len;
  }
  -int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr, -             void *buf, int len, int write)
+int ttm_bo_vm_access_reserved(struct ttm_buffer_object *bo,
+                  struct vm_area_struct *vma,
+                  unsigned long offset,
+                  void *buf, int len, int write)
  {
-    struct ttm_buffer_object *bo = vma->vm_private_data;
-    unsigned long offset = (addr) - vma->vm_start +
-        ((vma->vm_pgoff - drm_vma_node_start(&bo->base.vma_node))
-         << PAGE_SHIFT);
      int ret;
        if (len < 1 || (offset + len) >> PAGE_SHIFT > bo->mem.num_pages) return -EIO;
  -    ret = ttm_bo_reserve(bo, true, false, NULL);
-    if (ret)
-        return ret;
-
      switch (bo->mem.mem_type) {
      case TTM_PL_SYSTEM:
          if (unlikely(bo->ttm->page_flags & TTM_PAGE_FLAG_SWAPPED)) {

drivers/gpu/drm/ttm/ttm_bo_vm.c +0 -0
@@ -485,16 +488,33 @@ int ttm_bo_vm_access(struct vm_area_struct 
 *vma, unsigned long addr,
           }
           fallthrough;

      case TTM_PL_TT:
-        ret = ttm_bo_vm_access_kmap(bo, offset, buf, len, write); -        break;
+        return ttm_bo_vm_access_kmap(bo, offset, buf, len, write); default:
          if (bo->bdev->funcs->access_memory)
-            ret = bo->bdev->funcs->access_memory(
+            return bo->bdev->funcs->access_memory(
                  bo, offset, buf, len, write);
          else
-            ret = -EIO;
+            return -EIO;
      }
  +    return ret;

drivers/gpu/drm/ttm/ttm_bo_vm.c +17 -0
+}
+EXPORT_SYMBOL(ttm_bo_vm_access_reserved);
+
+int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
+             void *buf, int len, int write)
+{
+    struct ttm_buffer_object *bo = vma->vm_private_data;
+    unsigned long offset = (addr) - vma->vm_start +
+        ((vma->vm_pgoff - drm_vma_node_start(&bo->base.vma_node))
+         << PAGE_SHIFT);
+    int ret;
+
+    ret = ttm_bo_reserve(bo, true, false, NULL);
+    if (ret)
+        return ret;
+
+    ret = ttm_bo_vm_access_reserved(bo, vma, offset, buf, len, write);

      ttm_bo_unreserve(bo);
        return ret;

drivers/gpu/drm/vmwgfx/vmwgfx_page_dirty.c +0 -0
index 45c9c6a7f1d6..56ecace0cf5c 100644
@@ -477,7 +477,9 @@ vm_fault_t vmw_bo_vm_fault(struct vm_fault *vmf)

      else
          prot = vm_get_page_prot(vma->vm_flags);
  -    ret = ttm_bo_vm_fault_reserved(vmf, prot, num_prefault, 1); +    ret = ttm_bo_vm_fault_reserved(bo, vmf,
+ drm_vma_node_start(&bo->base.vma_node),
+                       prot, num_prefault, 1);
      if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT)) return ret;
  @@ -546,7 +548,9 @@ vm_fault_t vmw_bo_vm_huge_fault(struct vm_fault *vmf, prot = vm_get_page_prot(vma->vm_flags);
      }
  -    ret = ttm_bo_vm_fault_reserved(vmf, prot, 1, fault_page_size); +    ret = ttm_bo_vm_fault_reserved(bo, vmf,
+ drm_vma_node_start(&bo->base.vma_node),
+                       prot, 1, fault_page_size);
      if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT)) return ret;
  diff --git a/include/drm/ttm/ttm_bo_api.h b/include/drm/ttm/ttm_bo_api.h index 639521880c29..434f91f1fdbf 100644

include/drm/ttm/ttm_bo_api.h +3 -0
@@ -605,7 +605,9 @@ int ttm_mem_evict_first(struct ttm_device *bdev,
   vm_fault_t ttm_bo_vm_reserve(struct ttm_buffer_object *bo,
                    struct vm_fault *vmf);
   -vm_fault_t ttm_bo_vm_fault_reserved(struct vm_fault *vmf,
+vm_fault_t ttm_bo_vm_fault_reserved(struct ttm_buffer_object *bo,
+                    struct vm_fault *vmf,
+                    unsigned long mmap_base,
                       pgprot_t prot,
                       pgoff_t num_prefault,
                       pgoff_t fault_page_size);
@@ -616,6 +618,11 @@ void ttm_bo_vm_open(struct vm_area_struct *vma);
     void ttm_bo_vm_close(struct vm_area_struct *vma);
Quoted 9 lines
Quoted 8 lines

include/drm/ttm/ttm_bo_api.h void ttm_bo_vm_open(struct vm_area_struct *vma); · line 628

Quoted 8 lines
include/drm/ttm/ttm_bo_api.h +4 -0
   +int ttm_bo_vm_access_reserved(struct ttm_buffer_object *bo,
+                  struct vm_area_struct *vma,
+                  unsigned long offset,
+                  void *buf, int len, int write);
+
   int ttm_bo_vm_access(struct vm_area_struct *vma, unsigned long addr,
                void *buf, int len, int write);
   bool ttm_bo_delayed_delete(struct ttm_device *bdev, bool remove_all);

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

author reviewer
#54 lore raw

PATCH v2 14/15 drm/i915: Use ttm mmap handling for ttm bo's.

In reply to Thomas Hellström

From: Maarten Lankhorst <[email protected]>

Use the ttm handlers for servicing page faults, and vm_access.

---
 drivers/gpu/drm/i915/gem/i915_gem_mman.c      |  17 ++-
 drivers/gpu/drm/i915/gem/i915_gem_mman.h      |   2 +
 .../gpu/drm/i915/gem/i915_gem_object_types.h  |   1 +
 drivers/gpu/drm/i915/gem/i915_gem_ttm.c       | 105 +++++++++++++++++-
 4 files changed, 118 insertions(+), 7 deletions(-)
drivers/gpu/drm/i915/gem/i915_gem_mman.c +11 -6
index 65db290efd16..2bf89349dde9 100644
@@ -19,6 +19,7 @@
 #include "i915_gem_mman.h"
 #include "i915_trace.h"
 #include "i915_user_extensions.h"
+#include "i915_gem_ttm.h"
 #include "i915_vma.h"
 
 static inline bool
@@ -789,7 +790,7 @@ i915_gem_mmap_offset_ioctl(struct drm_device *dev, void *data,
 	return __assign_mmap_offset(file, args->handle, type, &args->offset);
 }
 
-static void vm_open(struct vm_area_struct *vma)
+void i915_gem_mmap_vm_open(struct vm_area_struct *vma)
 {
 	struct i915_mmap_offset *mmo = vma->vm_private_data;
 	struct drm_i915_gem_object *obj = mmo->obj;
@@ -798,7 +799,7 @@ static void vm_open(struct vm_area_struct *vma)
 	i915_gem_object_get(obj);
 }
 
-static void vm_close(struct vm_area_struct *vma)
+void i915_gem_mmap_vm_close(struct vm_area_struct *vma)
 {
 	struct i915_mmap_offset *mmo = vma->vm_private_data;
 	struct drm_i915_gem_object *obj = mmo->obj;
@@ -810,15 +811,15 @@ static void vm_close(struct vm_area_struct *vma)
 static const struct vm_operations_struct vm_ops_gtt = {
 	.fault = vm_fault_gtt,
 	.access = vm_access,
-	.open = vm_open,
-	.close = vm_close,
+	.open = i915_gem_mmap_vm_open,
+	.close = i915_gem_mmap_vm_close,
 };
 
 static const struct vm_operations_struct vm_ops_cpu = {
 	.fault = vm_fault_cpu,
 	.access = vm_access,
-	.open = vm_open,
-	.close = vm_close,
+	.open = i915_gem_mmap_vm_open,
+	.close = i915_gem_mmap_vm_close,
 };
 
 static int singleton_release(struct inode *inode, struct file *file)
@@ -953,6 +954,10 @@ int i915_gem_mmap(struct file *filp, struct vm_area_struct *vma)
 	}
 	vma->vm_page_prot = pgprot_decrypted(vma->vm_page_prot);
 
+	/* override ops per-object if desired */
+	if (obj->ops->mmap_ops)
+		vma->vm_ops = obj->ops->mmap_ops;
+
 	return 0;
 }
 
drivers/gpu/drm/i915/gem/i915_gem_mman.h +2 -0
index efee9e0d2508..e5bd02a6db12 100644
@@ -28,5 +28,7 @@ void __i915_gem_object_release_mmap_gtt(struct drm_i915_gem_object *obj);
 void i915_gem_object_release_mmap_gtt(struct drm_i915_gem_object *obj);
 
 void i915_gem_object_release_mmap_offset(struct drm_i915_gem_object *obj);
+void i915_gem_mmap_vm_open(struct vm_area_struct *vma);
+void i915_gem_mmap_vm_close(struct vm_area_struct *vma);
 
 #endif
drivers/gpu/drm/i915/gem/i915_gem_object_types.h +1 -0
index b350765e1935..31d828e91cf4 100644
@@ -79,6 +79,7 @@ struct drm_i915_gem_object_ops {
 	void (*delayed_free)(struct drm_i915_gem_object *obj);
 	void (*release)(struct drm_i915_gem_object *obj);
 
+	const struct vm_operations_struct *mmap_ops;
 	const char *name; /* friendly name for debug, e.g. lockdep classes */
 };
 
drivers/gpu/drm/i915/gem/i915_gem_ttm.c +104 -1

135 lines not on this page · raw message

Signature · 6 lines
2.31.1

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
author reviewer Reviewed-by on v2 14/15
#55 lore raw

Re: [Intel-gfx] [PATCH v2 14/15] drm/i915: Use ttm mmap handling for ttm bo's.

Quoted 5 lines · Thomas Hellström

From: Maarten Lankhorst <[email protected]>

Use the ttm handlers for servicing page faults, and vm_access.

LGTM. Just need to make sure we don't forget about the caching.

Quoted 229 lines
---
  drivers/gpu/drm/i915/gem/i915_gem_mman.c      |  17 ++-
  drivers/gpu/drm/i915/gem/i915_gem_mman.h      |   2 +
  .../gpu/drm/i915/gem/i915_gem_object_types.h  |   1 +
  drivers/gpu/drm/i915/gem/i915_gem_ttm.c       | 105 +++++++++++++++++-
  4 files changed, 118 insertions(+), 7 deletions(-)
drivers/gpu/drm/i915/gem/i915_gem_mman.c +11 -6
index 65db290efd16..2bf89349dde9 100644
@@ -19,6 +19,7 @@
  #include "i915_gem_mman.h"
  #include "i915_trace.h"
  #include "i915_user_extensions.h"
+#include "i915_gem_ttm.h"
  #include "i915_vma.h"
 
  static inline bool
@@ -789,7 +790,7 @@ i915_gem_mmap_offset_ioctl(struct drm_device *dev, void *data,
  	return __assign_mmap_offset(file, args->handle, type, &args->offset);
  }
 
-static void vm_open(struct vm_area_struct *vma)
+void i915_gem_mmap_vm_open(struct vm_area_struct *vma)
  {
  	struct i915_mmap_offset *mmo = vma->vm_private_data;
  	struct drm_i915_gem_object *obj = mmo->obj;
@@ -798,7 +799,7 @@ static void vm_open(struct vm_area_struct *vma)
  	i915_gem_object_get(obj);
  }
 
-static void vm_close(struct vm_area_struct *vma)
+void i915_gem_mmap_vm_close(struct vm_area_struct *vma)
  {
  	struct i915_mmap_offset *mmo = vma->vm_private_data;
  	struct drm_i915_gem_object *obj = mmo->obj;
@@ -810,15 +811,15 @@ static void vm_close(struct vm_area_struct *vma)
  static const struct vm_operations_struct vm_ops_gtt = {
  	.fault = vm_fault_gtt,
  	.access = vm_access,
-	.open = vm_open,
-	.close = vm_close,
+	.open = i915_gem_mmap_vm_open,
+	.close = i915_gem_mmap_vm_close,
  };
 
  static const struct vm_operations_struct vm_ops_cpu = {
  	.fault = vm_fault_cpu,
  	.access = vm_access,
-	.open = vm_open,
-	.close = vm_close,
+	.open = i915_gem_mmap_vm_open,
+	.close = i915_gem_mmap_vm_close,
  };
 
  static int singleton_release(struct inode *inode, struct file *file)
@@ -953,6 +954,10 @@ int i915_gem_mmap(struct file *filp, struct vm_area_struct *vma)
  	}
  	vma->vm_page_prot = pgprot_decrypted(vma->vm_page_prot);
 
+	/* override ops per-object if desired */
+	if (obj->ops->mmap_ops)
+		vma->vm_ops = obj->ops->mmap_ops;
+
  	return 0;
  }
 
drivers/gpu/drm/i915/gem/i915_gem_mman.h +2 -0
index efee9e0d2508..e5bd02a6db12 100644
@@ -28,5 +28,7 @@ void __i915_gem_object_release_mmap_gtt(struct drm_i915_gem_object *obj);
  void i915_gem_object_release_mmap_gtt(struct drm_i915_gem_object *obj);
 
  void i915_gem_object_release_mmap_offset(struct drm_i915_gem_object *obj);
+void i915_gem_mmap_vm_open(struct vm_area_struct *vma);
+void i915_gem_mmap_vm_close(struct vm_area_struct *vma);
 
  #endif
drivers/gpu/drm/i915/gem/i915_gem_object_types.h +1 -0
index b350765e1935..31d828e91cf4 100644
@@ -79,6 +79,7 @@ struct drm_i915_gem_object_ops {
  	void (*delayed_free)(struct drm_i915_gem_object *obj);
  	void (*release)(struct drm_i915_gem_object *obj);
 
+	const struct vm_operations_struct *mmap_ops;
  	const char *name; /* friendly name for debug, e.g. lockdep classes */
  };
 
drivers/gpu/drm/i915/gem/i915_gem_ttm.c +103 -1
index 790f5ec45c4d..fe9ac50b2470 100644
@@ -14,6 +14,7 @@
  #include "gem/i915_gem_region.h"
  #include "gem/i915_gem_ttm.h"
  #include "gem/i915_gem_ttm_bo_util.h"
+#include "gem/i915_gem_mman.h"
 
  #define I915_PL_LMEM0 TTM_PL_PRIV
  #define I915_PL_SYSTEM TTM_PL_SYSTEM
@@ -345,6 +346,44 @@ static int i915_ttm_move(struct ttm_buffer_object *bo, bool evict,
  	return 0;
  }
 
+static int i915_ttm_io_mem_reserve(struct ttm_device *bdev, struct ttm_resource *mem)
+{
+	if (mem->mem_type < I915_PL_LMEM0)
+		return 0;
+
+	/* We may need to revisit this later, but this allows all caching to be used in mmap */
+	mem->bus.caching = ttm_cached;
+	mem->bus.is_iomem = true;
+
+	return 0;
+}
+
+static unsigned long i915_ttm_io_mem_pfn(struct ttm_buffer_object *bo,
+					 unsigned long page_offset)
+{
+	struct drm_i915_gem_object *obj = i915_ttm_to_gem(bo);
+	struct sg_table *sgt = obj->ttm.cached_io_st;
+	struct scatterlist *sg;
+	unsigned int i;
+
+	GEM_WARN_ON(bo->ttm);
+
+	for_each_sgtable_dma_sg(sgt, sg, i) {
+		unsigned long sg_max = sg->length >> PAGE_SHIFT;
+
+		if (page_offset < sg_max) {
+			unsigned long base =
+				obj->mm.region->iomap.base - obj->mm.region->region.start;
+
+			return ((base + sg_dma_address(sg)) >> PAGE_SHIFT) + page_offset;
+		}
+
+		page_offset -= sg_max;
+	}
+	GEM_BUG_ON(1);
+	return 0;
+}
+
  struct ttm_device_funcs i915_ttm_bo_driver = {
  	.ttm_tt_create = i915_ttm_tt_create,
  	.ttm_tt_unpopulate = i915_ttm_tt_unpopulate,
@@ -355,6 +394,8 @@ struct ttm_device_funcs i915_ttm_bo_driver = {
  	.verify_access = NULL,
  	.swap_notify = i915_ttm_swap_notify,
  	.delete_mem_notify = i915_ttm_delete_mem_notify,
+	.io_mem_reserve = i915_ttm_io_mem_reserve,
+	.io_mem_pfn = i915_ttm_io_mem_pfn,
  };
 
  static int i915_ttm_get_pages(struct drm_i915_gem_object *obj)
@@ -454,7 +495,68 @@ static void i915_ttm_delayed_free(struct drm_i915_gem_object *obj)
  	ttm_bo_put(i915_gem_to_ttm(obj));
  }
 
-static const struct drm_i915_gem_object_ops i915_gem_ttm_obj_ops = {
+static vm_fault_t vm_fault_ttm(struct vm_fault *vmf)
+{
+	struct vm_area_struct *area = vmf->vma;
+	struct i915_mmap_offset *mmo = area->vm_private_data;
+	struct drm_i915_gem_object *obj = mmo->obj;
+	vm_fault_t ret;
+
+	/* Sanity check that we allow writing into this object */
+	if (unlikely(i915_gem_object_is_readonly(obj) &&
+		     area->vm_flags & VM_WRITE))
+		return VM_FAULT_SIGBUS;
+
+	ret = ttm_bo_vm_reserve(i915_gem_to_ttm(obj), vmf);
+	if (ret)
+		return ret;
+
+	ret = ttm_bo_vm_fault_reserved(i915_gem_to_ttm(obj), vmf,
+				       drm_vma_node_start(&mmo->vma_node),
+				       vmf->vma->vm_page_prot,
+				       TTM_BO_VM_NUM_PREFAULT, 1);
+	if (ret == VM_FAULT_RETRY && !(vmf->flags & FAULT_FLAG_RETRY_NOWAIT))
+		return ret;
+
+	dma_resv_unlock(obj->base.resv);
+
+	return ret;
+}
+
+static int
+vm_access_ttm(struct vm_area_struct *area, unsigned long addr,
+	  void *buf, int len, int write)
+{
+	struct i915_mmap_offset *mmo = area->vm_private_data;
+	struct drm_i915_gem_object *obj = mmo->obj;
+	int err = 0;
+
+	if (i915_gem_object_is_readonly(obj) && write)
+		return -EACCES;
+
+	addr -= area->vm_start;
+	if (addr >= obj->base.size)
+		return -EINVAL;
+
+	err = i915_gem_object_lock_interruptible(obj, NULL);
+	if (err)
+		return err;
+
+	len = ttm_bo_vm_access_reserved(i915_gem_to_ttm(obj), area,
+					addr, buf, len, write);
+	i915_gem_object_unlock(obj);
+
+	return len;
+}
+
+static const struct vm_operations_struct vm_ops_ttm = {
+	.fault = vm_fault_ttm,
+	.access = vm_access_ttm,
+	.open = i915_gem_mmap_vm_open,
+	.close = i915_gem_mmap_vm_close,
+};
+
+const struct drm_i915_gem_object_ops i915_gem_ttm_obj_ops = {
  	.name = "i915_gem_object_ttm",
  	.flags = I915_GEM_OBJECT_HAS_IOMEM,
 

drivers/gpu/drm/i915/gem/i915_gem_ttm.c static const struct drm_i915_gem_object_ops i915_gem_ttm_obj_ops = { · line 571

Quoted 7 lines
drivers/gpu/drm/i915/gem/i915_gem_ttm.c +1 -0
@@ -463,6 +565,7 @@ static const struct drm_i915_gem_object_ops i915_gem_ttm_obj_ops = {
  	.truncate = i915_ttm_purge,
  	.adjust_lru = i915_ttm_adjust_lru,
  	.delayed_free = i915_ttm_delayed_free,
+	.mmap_ops = &vm_ops_ttm,
  };
 
  void i915_ttm_bo_destroy(struct ttm_buffer_object *bo)

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

author reviewer
#56 lore raw

PATCH v2 15/15 drm/i915/ttm: Add io sgt caching to i915_ttm_io_mem_pfn

In reply to Thomas Hellström

From: Maarten Lankhorst <[email protected]>

Instead of walking the sg table manually, use our caching helpers to do the sgt caching. To prevent lifetime issues of ttm_bo vs i915_gem_object, we will use a separate member, instead of re-using the dma page member.

---
 drivers/gpu/drm/i915/gem/i915_gem_object.h    |  6 +--
 .../gpu/drm/i915/gem/i915_gem_object_types.h  |  1 +
 drivers/gpu/drm/i915/gem/i915_gem_pages.c     |  3 +-
 drivers/gpu/drm/i915/gem/i915_gem_ttm.c       | 46 ++++++++++---------
 4 files changed, 30 insertions(+), 26 deletions(-)
drivers/gpu/drm/i915/gem/i915_gem_object.h +3 -3
index a3ad8cf4eefd..ff59e6c640e6 100644
@@ -342,14 +342,14 @@ struct scatterlist *
 __i915_gem_object_get_sg(struct drm_i915_gem_object *obj,
 			 struct i915_gem_object_page_iter *iter,
 			 unsigned int n,
-			 unsigned int *offset, bool allow_alloc);
+			 unsigned int *offset, bool allow_alloc, bool dma);
 
 static inline struct scatterlist *
 i915_gem_object_get_sg(struct drm_i915_gem_object *obj,
 		       unsigned int n,
 		       unsigned int *offset, bool allow_alloc)
 {
-	return __i915_gem_object_get_sg(obj, &obj->mm.get_page, n, offset, allow_alloc);
+	return __i915_gem_object_get_sg(obj, &obj->mm.get_page, n, offset, allow_alloc, false);
 }
 
 static inline struct scatterlist *
@@ -357,7 +357,7 @@ i915_gem_object_get_sg_dma(struct drm_i915_gem_object *obj,
 			   unsigned int n,
 			   unsigned int *offset, bool allow_alloc)
 {
-	return __i915_gem_object_get_sg(obj, &obj->mm.get_dma_page, n, offset, allow_alloc);
+	return __i915_gem_object_get_sg(obj, &obj->mm.get_dma_page, n, offset, allow_alloc, true);
 }
 
 struct page *
drivers/gpu/drm/i915/gem/i915_gem_object_types.h +1 -0
index 31d828e91cf4..828310802b9f 100644
@@ -324,6 +324,7 @@ struct drm_i915_gem_object {
 
 	struct {
 		struct sg_table *cached_io_st;
+		struct i915_gem_object_page_iter get_io_page;
 	} ttm;
 
 	/** Record of address bit 17 of each page at last unbind. */
drivers/gpu/drm/i915/gem/i915_gem_pages.c +1 -2
index 62ee2185a41b..577352b4f2f6 100644
@@ -465,9 +465,8 @@ __i915_gem_object_get_sg(struct drm_i915_gem_object *obj,
 			 struct i915_gem_object_page_iter *iter,
 			 unsigned int n,
 			 unsigned int *offset,
-			 bool allow_alloc)
+			 bool allow_alloc, bool dma)
 {
-	const bool dma = iter == &obj->mm.get_dma_page;
 	struct scatterlist *sg;
 	unsigned int idx, count;
 
drivers/gpu/drm/i915/gem/i915_gem_ttm.c +25 -21

76 lines not on this page · raw message

Signature · 1 line
2.31.1
author reviewer Reviewed-by on v2 15/15
#57 lore raw

Re: [Intel-gfx] [PATCH v2 15/15] drm/i915/ttm: Add io sgt caching to i915_ttm_io_mem_pfn

Quoted 147 lines · Thomas Hellström

From: Maarten Lankhorst <[email protected]>

Instead of walking the sg table manually, use our caching helpers to do the sgt caching. To prevent lifetime issues of ttm_bo vs i915_gem_object, we will use a separate member, instead of re-using the dma page member.

---
  drivers/gpu/drm/i915/gem/i915_gem_object.h    |  6 +--
  .../gpu/drm/i915/gem/i915_gem_object_types.h  |  1 +
  drivers/gpu/drm/i915/gem/i915_gem_pages.c     |  3 +-
  drivers/gpu/drm/i915/gem/i915_gem_ttm.c       | 46 ++++++++++---------
  4 files changed, 30 insertions(+), 26 deletions(-)
drivers/gpu/drm/i915/gem/i915_gem_object.h +3 -3
index a3ad8cf4eefd..ff59e6c640e6 100644
@@ -342,14 +342,14 @@ struct scatterlist *
  __i915_gem_object_get_sg(struct drm_i915_gem_object *obj,
  			 struct i915_gem_object_page_iter *iter,
  			 unsigned int n,
-			 unsigned int *offset, bool allow_alloc);
+			 unsigned int *offset, bool allow_alloc, bool dma);
 
  static inline struct scatterlist *
  i915_gem_object_get_sg(struct drm_i915_gem_object *obj,
  		       unsigned int n,
  		       unsigned int *offset, bool allow_alloc)
  {
-	return __i915_gem_object_get_sg(obj, &obj->mm.get_page, n, offset, allow_alloc);
+	return __i915_gem_object_get_sg(obj, &obj->mm.get_page, n, offset, allow_alloc, false);
  }
 
  static inline struct scatterlist *
@@ -357,7 +357,7 @@ i915_gem_object_get_sg_dma(struct drm_i915_gem_object *obj,
  			   unsigned int n,
  			   unsigned int *offset, bool allow_alloc)
  {
-	return __i915_gem_object_get_sg(obj, &obj->mm.get_dma_page, n, offset, allow_alloc);
+	return __i915_gem_object_get_sg(obj, &obj->mm.get_dma_page, n, offset, allow_alloc, true);
  }
 
  struct page *
drivers/gpu/drm/i915/gem/i915_gem_object_types.h +1 -0
index 31d828e91cf4..828310802b9f 100644
@@ -324,6 +324,7 @@ struct drm_i915_gem_object {
 
  	struct {
  		struct sg_table *cached_io_st;
+		struct i915_gem_object_page_iter get_io_page;
  	} ttm;
 
  	/** Record of address bit 17 of each page at last unbind. */
drivers/gpu/drm/i915/gem/i915_gem_pages.c +1 -2
index 62ee2185a41b..577352b4f2f6 100644
@@ -465,9 +465,8 @@ __i915_gem_object_get_sg(struct drm_i915_gem_object *obj,
  			 struct i915_gem_object_page_iter *iter,
  			 unsigned int n,
  			 unsigned int *offset,
-			 bool allow_alloc)
+			 bool allow_alloc, bool dma)
  {
-	const bool dma = iter == &obj->mm.get_dma_page;
  	struct scatterlist *sg;
  	unsigned int idx, count;
 
drivers/gpu/drm/i915/gem/i915_gem_ttm.c +22 -21
index fe9ac50b2470..1eaefb89e859 100644
@@ -167,11 +167,20 @@ static int i915_ttm_move_notify(struct ttm_buffer_object *bo)
 
  static void i915_ttm_free_cached_io_st(struct drm_i915_gem_object *obj)
  {
-	if (obj->ttm.cached_io_st) {
-		sg_free_table(obj->ttm.cached_io_st);
-		kfree(obj->ttm.cached_io_st);
-		obj->ttm.cached_io_st = NULL;
-	}
+	struct radix_tree_iter iter;
+	void __rcu **slot;
+
+	if (!obj->ttm.cached_io_st)
+		return;
+
+	rcu_read_lock();
+	radix_tree_for_each_slot(slot, &obj->ttm.get_io_page.radix, &iter, 0)
+		radix_tree_delete(&obj->ttm.get_io_page.radix, iter.index);
+	rcu_read_unlock();
+
+	sg_free_table(obj->ttm.cached_io_st);
+	kfree(obj->ttm.cached_io_st);
+	obj->ttm.cached_io_st = NULL;
  }
 
  static void i915_ttm_purge(struct drm_i915_gem_object *obj)
@@ -340,8 +349,11 @@ static int i915_ttm_move(struct ttm_buffer_object *bo, bool evict,
  	i915_ttm_move_memcpy(bo, new_mem, new_iter, old_iter);
  	i915_ttm_free_cached_io_st(obj);
 
-	if (!new_man->use_tt)
+	if (!new_man->use_tt) {
  		obj->ttm.cached_io_st = new_st;
+		obj->ttm.get_io_page.sg_pos = new_st->sgl;
+		obj->ttm.get_io_page.sg_idx = 0;
+	}
 
  	return 0;
  }
@@ -362,26 +374,15 @@ static unsigned long i915_ttm_io_mem_pfn(struct ttm_buffer_object *bo,
  					 unsigned long page_offset)
  {
  	struct drm_i915_gem_object *obj = i915_ttm_to_gem(bo);
-	struct sg_table *sgt = obj->ttm.cached_io_st;
+	unsigned long base = obj->mm.region->iomap.base - obj->mm.region->region.start;
  	struct scatterlist *sg;
-	unsigned int i;
+	unsigned int ofs;
 
  	GEM_WARN_ON(bo->ttm);
 
-	for_each_sgtable_dma_sg(sgt, sg, i) {
-		unsigned long sg_max = sg->length >> PAGE_SHIFT;
+	sg = __i915_gem_object_get_sg(obj, &obj->ttm.get_io_page, page_offset, &ofs, true, true);
 
-		if (page_offset < sg_max) {
-			unsigned long base =
-				obj->mm.region->iomap.base - obj->mm.region->region.start;
-
-			return ((base + sg_dma_address(sg)) >> PAGE_SHIFT) + page_offset;
-		}
-
-		page_offset -= sg_max;
-	}
-	GEM_BUG_ON(1);
-	return 0;
+	return ((base + sg_dma_address(sg)) >> PAGE_SHIFT) + ofs;
  }
 
  struct ttm_device_funcs i915_ttm_bo_driver = {

drivers/gpu/drm/i915/gem/i915_gem_ttm.c int __i915_gem_ttm_object_init(struct intel_memory_region *mem, · line 619

Quoted 6 lines
drivers/gpu/drm/i915/gem/i915_gem_ttm.c +3 -0
@@ -613,6 +614,9 @@ int __i915_gem_ttm_object_init(struct intel_memory_region *mem,
  	obj->read_domains = I915_GEM_DOMAIN_WC | I915_GEM_DOMAIN_GTT;
  	i915_gem_object_set_cache_coherency(obj, I915_CACHE_NONE);
 
+	INIT_RADIX_TREE(&obj->ttm.get_io_page.radix, GFP_KERNEL | __GFP_NOWARN);
+	mutex_init(&obj->ttm.get_io_page.lock);
+

Perhaps a mutex_destroy() in error path and destructor for full debugging purposes?

Otherwise LGTM,

Quoted 3 lines
drivers/gpu/drm/i915/gem/i915_gem_ttm.c +0 -0
  	ret = ttm_bo_init(&i915->bdev, i915_gem_to_ttm(obj), size,
  			  ttm_bo_type_kernel, &i915_sys_placement, alignment,
  			  true, NULL, NULL, i915_ttm_bo_destroy);
#58 lore raw

[Intel-gfx] ✗ Fi.CI.CHECKPATCH: warning for drm/i915: Move LMEM (VRAM) management over to TTM (rev2)

In reply to Thomas Hellström
== Series Details ==

== Series Details ==

Series: drm/i915: Move LMEM (VRAM) management over to TTM (rev2)
URL : https://patchwork.freedesktop.org/series/90022/
State : warning

== Summary ==

$ dim checkpatch origin/drm-tip
f95eea0001a4 drm/i915: Untangle the vma pages_mutex
425af44621a2 drm/i915: Don't free shared locks while shared
a1fb9cdc093f drm/i915: Fix i915_sg_page_sizes to record dma segments rather than physical pages
1b455d3307a9 drm/ttm: Export functions to initialize and finalize the ttm range manager standalone
01c0540773c7 drm/i915/ttm Initialize the ttm device and memory managers
-:479: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating?
#479: 
deleted file mode 100644

total: 0 errors, 1 warnings, 0 checks, 1478 lines checked
a3796d8feda9 drm/i915/ttm: Embed a ttm buffer object in the i915 gem object 32e0235ecbbf drm/ttm: Export ttm_bo_tt_destroy()
3ed6aaedd277 drm/i915/ttm Add a generic TTM memcpy move for page-based iomem -:46: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating? #46:
new file mode 100644

total: 0 errors, 1 warnings, 0 checks, 308 lines checked
75550f88727b drm/ttm, drm/amdgpu: Allow the driver some control over swapping 36729bdb371d drm/i915/ttm: Introduce a TTM i915 gem object backend
-:408: WARNING:FILE_PATH_CHANGES: added, moved or deleted file(s), does MAINTAINERS need updating? #408:
new file mode 100644

total: 0 errors, 1 warnings, 0 checks, 962 lines checked
f897aee184bd drm/i915/lmem: Verify checks for lmem residency
d85d5bfe5022 drm/i915: Disable mmap ioctl for gen12+
7199fd0658f1 drm/ttm: Add BO and offset arguments for vm_access and vm_fault ttm handlers. -:214: CHECK:OPEN_ENDED_LINE: Lines should not end with a '('
#214: FILE: drivers/gpu/drm/ttm/ttm_bo_vm.c:494:
+ return bo->bdev->funcs->access_memory(

total: 0 errors, 0 warnings, 1 checks, 240 lines checked bea92372457b drm/i915: Use ttm mmap handling for ttm bo's. -:196: CHECK:PARENTHESIS_ALIGNMENT: Alignment should match open parenthesis #196: FILE: drivers/gpu/drm/i915/gem/i915_gem_ttm.c:528:
+vm_access_ttm(struct vm_area_struct *area, unsigned long addr,
+ void *buf, int len, int write)

total: 0 errors, 0 warnings, 1 checks, 201 lines checked 17af87b6b030 drm/i915/ttm: Add io sgt caching to i915_ttm_io_mem_pfn

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#59 lore raw

[Intel-gfx] ✗ Fi.CI.SPARSE: warning for drm/i915: Move LMEM (VRAM) management over to TTM (rev2)

In reply to Thomas Hellström
== Series Details ==

== Series Details ==

Series: drm/i915: Move LMEM (VRAM) management over to TTM (rev2)
URL : https://patchwork.freedesktop.org/series/90022/
State : warning

== Summary ==

$ dim sparse --fast origin/drm-tip
Sparse version: v0.6.2
Fast mode used, each commit won't be checked separately.
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+./drivers/gpu/drm/amd/amdgpu/../amdgpu/amdgv_sriovmsg.h:274:49: error: static assertion failed: "amd_sriov_msg_pf2vf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c:1345:25: error: incompatible types in comparison expression (different address spaces):
+drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c:1345:25:    struct dma_fence *
+drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c:1345:25:    struct dma_fence [noderef] __rcu *
+drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c:1346:17: error: incompatible types in comparison expression (different address spaces):
+drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c:1346:17:    struct dma_fence *
+drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c:1346:17:    struct dma_fence [noderef] __rcu *
+drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c:1405:17: error: incompatible types in comparison expression (different address spaces):
+drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c:1405:17:    struct dma_fence *
+drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c:1405:17:    struct dma_fence [noderef] __rcu *
+drivers/gpu/drm/amd/amdgpu/amdgpu_device.c:287:16: error: incompatible types in comparison expression (different type sizes):
+drivers/gpu/drm/amd/amdgpu/amdgpu_device.c:287:16:    unsigned long *
+drivers/gpu/drm/amd/amdgpu/amdgpu_device.c:287:16:    unsigned long long *
+drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c:274:25: error: incompatible types in comparison expression (different address spaces):
+drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c:274:25:    struct dma_fence *
+drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c:274:25:    struct dma_fence [noderef] __rcu *
+drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c:275:17: error: incompatible types in comparison expression (different address spaces):
+drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c:275:17:    struct dma_fence *
+drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c:275:17:    struct dma_fence [noderef] __rcu *
+drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c:329:17: error: incompatible types in comparison expression (different address spaces):
+drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c:329:17:    struct dma_fence *
+drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c:329:17:    struct dma_fence [noderef] __rcu *
+drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.h:90:56: error: marked inline, but without a definition
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_info must be 1 KB"
+drivers/gpu/drm/amd/amdgpu/amdgv_sriovmsg.h:270:49: error: static assertion failed: "amd_sriov_msg_vf2pf_in

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#60 lore raw

[Intel-gfx] ✗ Fi.CI.BAT: failure for drm/i915: Move LMEM (VRAM) management over to TTM (rev2)

In reply to Thomas Hellström
== Series Details ==

== Series Details ==

Series: drm/i915: Move LMEM (VRAM) management over to TTM (rev2)
URL : https://patchwork.freedesktop.org/series/90022/
State : failure

== Summary ==

CI Bug Log - changes from CI_DRM_10094 -> Patchwork_20142
====================================================

Summary
-------

**FAILURE**

Serious unknown changes coming with Patchwork_20142 absolutely need to be verified manually.

If you think the reported changes have nothing to do with the changes introduced in Patchwork_20142, please notify your bug team to allow them to document this new failure mode, which will reduce false positives in CI.

External URL: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/index.html

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

Here are the unknown changes that may have been introduced in Patchwork_20142:

### IGT changes ###

#### Possible regressions ####

* igt@runner@aborted:
- fi-bdw-5557u: NOTRUN -> [FAIL][1]
[1]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-bdw-5557u/igt@[email protected]

#### Warnings ####

* igt@i915_selftest@live@execlists:
- fi-bsw-kefka: [INCOMPLETE][2] ([i915#2782] / [i915#2940]) -> [DMESG-FAIL][3]
[2]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/fi-bsw-kefka/igt@i915_selftest@[email protected]
[3]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-bsw-kefka/igt@i915_selftest@[email protected]

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

Here are the changes found in Patchwork_20142 that come from known issues:

### IGT changes ###

#### Issues hit ####

* igt@amdgpu/amd_basic@cs-gfx:
- fi-hsw-4770: NOTRUN -> [SKIP][4] ([fdo#109271] / [fdo#109315]) +17 similar issues
[4]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-hsw-4770/igt@amdgpu/[email protected]

* igt@amdgpu/amd_cs_nop@nop-compute0:
- fi-ilk-650: NOTRUN -> [SKIP][5] ([fdo#109271]) +30 similar issues
[5]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-ilk-650/igt@amdgpu/[email protected]

* igt@core_hotunplug@unbind-rebind:
- fi-hsw-4770: NOTRUN -> [WARN][6] ([i915#2283])
[6]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-hsw-4770/igt@[email protected]
- fi-bdw-5557u: NOTRUN -> [WARN][7] ([i915#2283])
[7]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-bdw-5557u/igt@[email protected]

* igt@gem_busy@busy@all:
- fi-elk-e7500: [PASS][8] -> [FAIL][9] ([i915#3457])
[8]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/fi-elk-e7500/igt@gem_busy@[email protected]
[9]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-elk-e7500/igt@gem_busy@[email protected]

* igt@gem_exec_fence@basic-await@vecs0:
- fi-bsw-kefka: [PASS][10] -> [FAIL][11] ([i915#3457]) +2 similar issues
[10]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/fi-bsw-kefka/igt@gem_exec_fence@[email protected]
[11]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-bsw-kefka/igt@gem_exec_fence@[email protected]

* igt@gem_exec_fence@basic-busy@bcs0:
- fi-cml-u2: NOTRUN -> [SKIP][12] ([i915#1208]) +1 similar issue
[12]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cml-u2/igt@gem_exec_fence@[email protected]

* igt@gem_exec_fence@nb-await@vcs0:
- fi-glk-dsi: NOTRUN -> [FAIL][13] ([i915#3457]) +6 similar issues
[13]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-glk-dsi/igt@gem_exec_fence@[email protected]

* igt@gem_exec_fence@nb-await@vecs0:
- fi-bsw-n3050: [PASS][14] -> [FAIL][15] ([i915#3457])
[14]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/fi-bsw-n3050/igt@gem_exec_fence@[email protected]
[15]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-bsw-n3050/igt@gem_exec_fence@[email protected]

* igt@gem_huc_copy@huc-copy:
- fi-cml-u2: NOTRUN -> [SKIP][16] ([i915#2190])
[16]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cml-u2/igt@[email protected]
- fi-hsw-4770: NOTRUN -> [SKIP][17] ([fdo#109271]) +4 similar issues
[17]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-hsw-4770/igt@[email protected]
- fi-skl-6700k2: NOTRUN -> [SKIP][18] ([fdo#109271] / [i915#2190])
[18]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-skl-6700k2/igt@[email protected]
- fi-glk-dsi: NOTRUN -> [SKIP][19] ([fdo#109271] / [i915#2190])
[19]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-glk-dsi/igt@[email protected]
- fi-kbl-7500u: NOTRUN -> [SKIP][20] ([fdo#109271] / [i915#2190])
[20]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-kbl-7500u/igt@[email protected]
- fi-kbl-7567u: NOTRUN -> [SKIP][21] ([fdo#109271] / [i915#2190])
[21]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-kbl-7567u/igt@[email protected]
- fi-cfl-8109u: NOTRUN -> [SKIP][22] ([fdo#109271] / [i915#2190])
[22]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cfl-8109u/igt@[email protected]
- fi-cfl-8700k: NOTRUN -> [SKIP][23] ([fdo#109271] / [i915#2190])
[23]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cfl-8700k/igt@[email protected]
- fi-icl-u2: NOTRUN -> [SKIP][24] ([i915#2190])
[24]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-icl-u2/igt@[email protected]

* igt@gem_wait@busy@all:
- fi-bwr-2160: [PASS][25] -> [FAIL][26] ([i915#3457])
[25]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/fi-bwr-2160/igt@gem_wait@[email protected]
[26]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-bwr-2160/igt@gem_wait@[email protected]

* igt@i915_module_load@reload:
- fi-icl-u2: NOTRUN -> [DMESG-WARN][27] ([i915#3457]) +1 similar issue
[27]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-icl-u2/igt@[email protected]
- fi-cml-u2: NOTRUN -> [DMESG-WARN][28] ([i915#3457]) +1 similar issue
[28]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cml-u2/igt@[email protected]
- fi-cfl-8700k: NOTRUN -> [DMESG-WARN][29] ([i915#3457]) +1 similar issue
[29]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cfl-8700k/igt@[email protected]
- fi-hsw-4770: NOTRUN -> [DMESG-WARN][30] ([i915#3457]) +1 similar issue
[30]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-hsw-4770/igt@[email protected]
- fi-ilk-650: NOTRUN -> [DMESG-FAIL][31] ([i915#3457]) +1 similar issue
[31]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-ilk-650/igt@[email protected]
- fi-skl-6700k2: NOTRUN -> [DMESG-WARN][32] ([i915#3457]) +1 similar issue
[32]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-skl-6700k2/igt@[email protected]
- fi-glk-dsi: NOTRUN -> [DMESG-FAIL][33] ([i915#3457])
[33]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-glk-dsi/igt@[email protected]

* igt@i915_pm_backlight@basic-brightness:
- fi-hsw-4770: NOTRUN -> [SKIP][34] ([fdo#109271] / [i915#3012])
[34]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-hsw-4770/igt@[email protected]

* igt@i915_selftest@live@execlists:
- fi-cfl-8109u: NOTRUN -> [DMESG-FAIL][35] ([i915#3462])
[35]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cfl-8109u/igt@i915_selftest@[email protected]
- fi-glk-dsi: NOTRUN -> [DMESG-FAIL][36] ([i915#3462])
[36]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-glk-dsi/igt@i915_selftest@[email protected]
- fi-icl-u2: NOTRUN -> [DMESG-FAIL][37] ([i915#3462])
[37]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-icl-u2/igt@i915_selftest@[email protected]
- fi-cfl-8700k: NOTRUN -> [DMESG-FAIL][38] ([i915#3462])
[38]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cfl-8700k/igt@i915_selftest@[email protected]
- fi-kbl-7567u: NOTRUN -> [INCOMPLETE][39] ([i915#2782] / [i915#3462] / [i915#794])
[39]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-kbl-7567u/igt@i915_selftest@[email protected]
- fi-skl-6700k2: NOTRUN -> [DMESG-FAIL][40] ([i915#3462])
[40]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-skl-6700k2/igt@i915_selftest@[email protected]
- fi-cml-u2: NOTRUN -> [DMESG-FAIL][41] ([i915#3462])
[41]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cml-u2/igt@i915_selftest@[email protected]
- fi-bdw-5557u: NOTRUN -> [DMESG-FAIL][42] ([i915#3462])
[42]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-bdw-5557u/igt@i915_selftest@[email protected]
- fi-kbl-7500u: NOTRUN -> [DMESG-FAIL][43] ([i915#3462])
[43]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-kbl-7500u/igt@i915_selftest@[email protected]

* igt@i915_selftest@live@mman:
- fi-bdw-5557u: NOTRUN -> [DMESG-WARN][44] ([i915#3457]) +1 similar issue
[44]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-bdw-5557u/igt@i915_selftest@[email protected]
- fi-cfl-8109u: NOTRUN -> [DMESG-WARN][45] ([i915#3457]) +1 similar issue
[45]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cfl-8109u/igt@i915_selftest@[email protected]
- fi-snb-2600: NOTRUN -> [DMESG-WARN][46] ([i915#3457]) +1 similar issue
[46]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-snb-2600/igt@i915_selftest@[email protected]
- fi-kbl-7567u: NOTRUN -> [DMESG-WARN][47] ([i915#3457]) +1 similar issue
[47]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-kbl-7567u/igt@i915_selftest@[email protected]
- fi-glk-dsi: NOTRUN -> [DMESG-WARN][48] ([i915#3457])
[48]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-glk-dsi/igt@i915_selftest@[email protected]
- fi-kbl-7500u: NOTRUN -> [DMESG-WARN][49] ([i915#3457]) +1 similar issue
[49]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-kbl-7500u/igt@i915_selftest@[email protected]

* igt@kms_addfb_basic@addfb25-y-tiled-small-legacy:
- fi-snb-2600: NOTRUN -> [SKIP][50] ([fdo#109271]) +37 similar issues
[50]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-snb-2600/igt@[email protected]

* igt@kms_busy@basic@modeset:
- fi-ilk-650: NOTRUN -> [INCOMPLETE][51] ([i915#3457])
[51]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-ilk-650/igt@kms_busy@[email protected]

* igt@kms_chamelium@common-hpd-after-suspend:
- fi-hsw-4770: NOTRUN -> [SKIP][52] ([fdo#109271] / [fdo#111827]) +8 similar issues
[52]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-hsw-4770/igt@[email protected]

* igt@kms_chamelium@dp-crc-fast:
- fi-kbl-7567u: NOTRUN -> [SKIP][53] ([fdo#109271] / [fdo#111827]) +8 similar issues
[53]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-kbl-7567u/igt@[email protected]
- fi-bdw-5557u: NOTRUN -> [SKIP][54] ([fdo#109271] / [fdo#111827]) +8 similar issues
[54]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-bdw-5557u/igt@[email protected]

* igt@kms_chamelium@dp-hpd-fast:
- fi-ilk-650: NOTRUN -> [SKIP][55] ([fdo#109271] / [fdo#111827]) +8 similar issues
[55]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-ilk-650/igt@[email protected]
- fi-skl-6700k2: NOTRUN -> [SKIP][56] ([fdo#109271]) +13 similar issues
[56]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-skl-6700k2/igt@[email protected]
- fi-cml-u2: NOTRUN -> [SKIP][57] ([fdo#109284] / [fdo#111827]) +8 similar issues
[57]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cml-u2/igt@[email protected]

* igt@kms_chamelium@hdmi-crc-fast:
- fi-snb-2600: NOTRUN -> [SKIP][58] ([fdo#109271] / [fdo#111827]) +8 similar issues
[58]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-snb-2600/igt@[email protected]

* igt@kms_chamelium@hdmi-hpd-fast:
- fi-glk-dsi: NOTRUN -> [SKIP][59] ([fdo#109271] / [fdo#111827]) +8 similar issues
[59]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-glk-dsi/igt@[email protected]
- fi-cfl-8109u: NOTRUN -> [SKIP][60] ([fdo#109271] / [fdo#111827]) +8 similar issues
[60]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cfl-8109u/igt@[email protected]

* igt@kms_chamelium@vga-edid-read:
- fi-cfl-8700k: NOTRUN -> [SKIP][61] ([fdo#109271] / [fdo#111827]) +8 similar issues
[61]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cfl-8700k/igt@[email protected]
- fi-icl-u2: NOTRUN -> [SKIP][62] ([fdo#109309]) +1 similar issue
[62]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-icl-u2/igt@[email protected]

* igt@kms_cursor_legacy@basic-busy-flip-before-cursor-atomic:
- fi-ilk-650: NOTRUN -> [FAIL][63] ([i915#3457]) +1 similar issue
[63]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-ilk-650/igt@[email protected]
- fi-kbl-soraka: [PASS][64] -> [DMESG-WARN][65] ([i915#1982] / [i915#3457])
[64]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/fi-kbl-soraka/igt@[email protected]
[65]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-kbl-soraka/igt@[email protected]

* igt@kms_force_connector_basic@force-load-detect:
- fi-cml-u2: NOTRUN -> [SKIP][66] ([fdo#109285])
[66]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cml-u2/igt@[email protected]
- fi-icl-u2: NOTRUN -> [SKIP][67] ([fdo#109285])
[67]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-icl-u2/igt@[email protected]

* igt@kms_pipe_crc_basic@compare-crc-sanitycheck-pipe-d:
- fi-hsw-4770: NOTRUN -> [SKIP][68] ([fdo#109271] / [i915#533])
[68]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-hsw-4770/igt@[email protected]
- fi-kbl-7500u: NOTRUN -> [SKIP][69] ([fdo#109271] / [i915#533])
[69]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-kbl-7500u/igt@[email protected]
- fi-cml-u2: NOTRUN -> [SKIP][70] ([fdo#109278] / [i915#533])
[70]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cml-u2/igt@[email protected]
- fi-cfl-8700k: NOTRUN -> [SKIP][71] ([fdo#109271] / [i915#533])
[71]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cfl-8700k/igt@[email protected]
- fi-icl-u2: NOTRUN -> [SKIP][72] ([fdo#109278])
[72]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-icl-u2/igt@[email protected]
- fi-skl-6700k2: NOTRUN -> [SKIP][73] ([fdo#109271] / [i915#533])
[73]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-skl-6700k2/igt@[email protected]
- fi-cfl-8109u: NOTRUN -> [SKIP][74] ([fdo#109271] / [i915#533])
[74]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cfl-8109u/igt@[email protected]
- fi-glk-dsi: NOTRUN -> [SKIP][75] ([fdo#109271] / [i915#533])
[75]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-glk-dsi/igt@[email protected]
- fi-kbl-7567u: NOTRUN -> [SKIP][76] ([fdo#109271] / [i915#533])
[76]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-kbl-7567u/igt@[email protected]

* igt@kms_pipe_crc_basic@hang-read-crc-pipe-a:
- fi-bwr-2160: [PASS][77] -> [FAIL][78] ([i915#53])
[77]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/fi-bwr-2160/igt@[email protected]
[78]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-bwr-2160/igt@[email protected]

* igt@kms_pipe_crc_basic@read-crc-pipe-b:
- fi-elk-e7500: [PASS][79] -> [FAIL][80] ([i915#53]) +1 similar issue
[79]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/fi-elk-e7500/igt@[email protected]
[80]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-elk-e7500/igt@[email protected]

* igt@kms_pipe_crc_basic@read-crc-pipe-c:
- fi-kbl-7567u: NOTRUN -> [SKIP][81] ([fdo#109271]) +41 similar issues
[81]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-kbl-7567u/igt@[email protected]

* igt@kms_psr@cursor_plane_move:
- fi-bdw-5557u: NOTRUN -> [SKIP][82] ([fdo#109271]) +9 similar issues
[82]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-bdw-5557u/igt@kms_psr@cursor_plane_move.html

* igt@kms_psr@primary_mmap_gtt:
- fi-cfl-8109u: NOTRUN -> [SKIP][83] ([fdo#109271]) +8 similar issues
[83]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cfl-8109u/igt@kms_psr@primary_mmap_gtt.html
- fi-kbl-7500u: NOTRUN -> [SKIP][84] ([fdo#109271]) +10 similar issues
[84]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-kbl-7500u/igt@kms_psr@primary_mmap_gtt.html
- fi-hsw-4770: NOTRUN -> [SKIP][85] ([fdo#109271] / [i915#1072]) +3 similar issues
[85]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-hsw-4770/igt@kms_psr@primary_mmap_gtt.html

* igt@kms_psr@primary_page_flip:
- fi-glk-dsi: NOTRUN -> [SKIP][86] ([fdo#109271]) +10 similar issues
[86]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-glk-dsi/igt@kms_psr@primary_page_flip.html

* igt@prime_vgem@basic-userptr:
- fi-cfl-8700k: NOTRUN -> [SKIP][87] ([fdo#109271]) +8 similar issues
[87]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cfl-8700k/igt@[email protected]
- fi-icl-u2: NOTRUN -> [SKIP][88] ([i915#3301])
[88]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-icl-u2/igt@[email protected]
- fi-cml-u2: NOTRUN -> [SKIP][89] ([i915#3301])
[89]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cml-u2/igt@[email protected]

* igt@runner@aborted:
- fi-cfl-8700k: NOTRUN -> [FAIL][90] ([i915#3363])
[90]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cfl-8700k/igt@[email protected]
- fi-cfl-8109u: NOTRUN -> [FAIL][91] ([i915#2426] / [i915#3363])
[91]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cfl-8109u/igt@[email protected]
- fi-icl-u2: NOTRUN -> [FAIL][92] ([i915#2426] / [i915#2782] / [i915#3363])
[92]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-icl-u2/igt@[email protected]
- fi-glk-dsi: NOTRUN -> [FAIL][93] ([i915#3363] / [k.org#202321])
[93]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-glk-dsi/igt@[email protected]
- fi-kbl-7500u: NOTRUN -> [FAIL][94] ([i915#1436] / [i915#3363])
[94]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-kbl-7500u/igt@[email protected]
- fi-cml-u2: NOTRUN -> [FAIL][95] ([i915#2082] / [i915#2426] / [i915#3363])
[95]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-cml-u2/igt@[email protected]
- fi-kbl-7567u: NOTRUN -> [FAIL][96] ([i915#1436] / [i915#2426] / [i915#3363])
[96]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-kbl-7567u/igt@[email protected]
- fi-skl-6700k2: NOTRUN -> [FAIL][97] ([i915#1436] / [i915#3363])
[97]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-skl-6700k2/igt@[email protected]

#### Possible fixes ####

* igt@gem_exec_fence@nb-await@bcs0:
- fi-bsw-kefka: [FAIL][98] ([i915#3457]) -> [PASS][99]
[98]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/fi-bsw-kefka/igt@gem_exec_fence@[email protected]
[99]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-bsw-kefka/igt@gem_exec_fence@[email protected]
- fi-bsw-n3050: [FAIL][100] ([i915#3457]) -> [PASS][101] +1 similar issue
[100]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/fi-bsw-n3050/igt@gem_exec_fence@[email protected]
[101]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-bsw-n3050/igt@gem_exec_fence@[email protected]

* igt@kms_busy@basic@flip:
- fi-ilk-650: [INCOMPLETE][102] ([i915#3457]) -> [PASS][103]
[102]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/fi-ilk-650/igt@kms_busy@[email protected]
[103]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-ilk-650/igt@kms_busy@[email protected]

* igt@kms_pipe_crc_basic@nonblocking-crc-pipe-a:
- fi-elk-e7500: [FAIL][104] ([i915#53]) -> [PASS][105] +1 similar issue
[104]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/fi-elk-e7500/igt@[email protected]
[105]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-elk-e7500/igt@[email protected]

* igt@kms_pipe_crc_basic@read-crc-pipe-a-frame-sequence:
- fi-bsw-kefka: [FAIL][106] ([i915#53]) -> [PASS][107]
[106]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/fi-bsw-kefka/igt@[email protected]
[107]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-bsw-kefka/igt@[email protected]

#### Warnings ####

* igt@gem_exec_gttfill@basic:
- fi-pnv-d510: [FAIL][108] ([i915#3472]) -> [FAIL][109] ([i915#3457] / [i915#3472])
[108]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/fi-pnv-d510/igt@[email protected]
[109]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-pnv-d510/igt@[email protected]

* igt@i915_module_load@reload:
- fi-elk-e7500: [DMESG-FAIL][110] ([i915#3457]) -> [DMESG-WARN][111] ([i915#3457])
[110]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/fi-elk-e7500/igt@[email protected]
[111]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-elk-e7500/igt@[email protected]

* igt@i915_selftest@live@mman:
- fi-bwr-2160: [DMESG-WARN][112] ([i915#3457]) -> [DMESG-FAIL][113] ([i915#3457])
[112]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/fi-bwr-2160/igt@i915_selftest@[email protected]
[113]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/fi-bwr-2160/igt@i915_selftest@[email protected]

{name}: This element is suppressed. This means it is ignored when computing
the status of the difference (SUCCESS, WARNING, or FAILURE).

[fdo#109271]: https://bugs.freedesktop.org/show_bug.cgi?id=109271
[fdo#109278]: https://bugs.freedesktop.org/show_bug.cgi?id=109278
[fdo#109284]: https://bugs.freedesktop.org/show_bug.cgi?id=109284
[fdo#109285]: https://bugs.freedesktop.org/show_bug.cgi?id=109285
[fdo#109309]: https://bugs.freedesktop.org/show_bug.cgi?id=109309
[fdo#109315]: https://bugs.freedesktop.org/show_bug.cgi?id=109315
[fdo#111827]: https://bugs.freedesktop.org/show_bug.cgi?id=111827
[i915#1021]: https://gitlab.freedesktop.org/drm/intel/issues/1021
[i915#1072]: https://gitlab.freedesktop.org/drm/intel/issues/1072
[i915#1208]: https://gitlab.freedesktop.org/drm/intel/issues/1208
[i915#1222]: https://gitlab.freedesktop.org/drm/intel/issues/1222
[i915#1436]: https://gitlab.freedesktop.org/drm/intel/issues/1436
[i915#1982]: https://gitlab.freedesktop.org/drm/intel/issues/1982
[i915#2082]: https://gitlab.freedesktop.org/drm/intel/issues/2082
[i915#2190]: https://gitlab.freedesktop.org/drm/intel/issues/2190
[i915#2283]: https://gitlab.freedesktop.org/drm/intel/issues/2283
[i915#2426]: https://gitlab.freedesktop.org/drm/intel/issues/2426
[i915#2505]: https://gitlab.freedesktop.org/drm/intel/issues/2505
[i915#2782]: https://gitlab.freedesktop.org/drm/intel/issues/2782
[i915#2932]: https://gitlab.freedesktop.org/drm/intel/issues/2932
[i915#2940]: https://gitlab.freedesktop.org/drm/intel/issues/2940
[i915#2966]: https://gitlab.freedesktop.org/drm/intel/issues/2966
[i915#3012]: https://gitlab.freedesktop.org/drm/intel/issues/3012
[i915#3301]: https://gitlab.freedesktop.org/drm/intel/issues/3301
[i915#3363]: https://gitlab.freedesktop.org/drm/intel/issues/3363
[i915#3457]: https://gitlab.freedesktop.org/drm/intel/issues/3457
[i915#3462]: https://gitlab.freedesktop.org/drm/intel/issues/3462
[i915#3472]: https://gitlab.freedesktop.org/drm/intel/issues/3472
[i915#53]: https://gitlab.freedesktop.org/drm/intel/issues/53
[i915#533]: https://gitlab.freedesktop.org/drm/intel/issues/533
[i915#794]: https://gitlab.freedesktop.org/drm/intel/issues/794
[k.org#202321]: https://bugzilla.kernel.org/show_bug.cgi?id=202321

Participating hosts (24 -> 35)
------------------------------

Additional (13): fi-kbl-7567u fi-cml-u2 fi-jsl-1 fi-ehl-1 fi-bdw-5557u fi-glk-dsi fi-icl-u2 fi-cfl-8700k fi-kbl-7500u fi-hsw-4770 fi-cfl-8109u fi-skl-6700k2 fi-snb-2600
Missing (2): fi-bsw-cyan fi-bsw-nick

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

* Linux: CI_DRM_10094 -> Patchwork_20142

CI-20190529: 20190529
CI_DRM_10094: 8c6b35ef6a4accca881b1142b3a73ba3725797c7 @ git://anongit.freedesktop.org/gfx-ci/linux
IGT_6085: aacb1384378131b87121219a6a7ccd8b8f1e5d8d @ git://anongit.freedesktop.org/xorg/app/intel-gpu-tools
Patchwork_20142: 17af87b6b030fd638c3912e4b284577c4871111d @ git://anongit.freedesktop.org/gfx-ci/linux

== Linux commits ==

17af87b6b030 drm/i915/ttm: Add io sgt caching to i915_ttm_io_mem_pfn
bea92372457b drm/i915: Use ttm mmap handling for ttm bo's.
7199fd0658f1 drm/ttm: Add BO and offset arguments for vm_access and vm_fault ttm handlers. d85d5bfe5022 drm/i915: Disable mmap ioctl for gen12+
f897aee184bd drm/i915/lmem: Verify checks for lmem residency
36729bdb371d drm/i915/ttm: Introduce a TTM i915 gem object backend
75550f88727b drm/ttm, drm/amdgpu: Allow the driver some control over swapping 3ed6aaedd277 drm/i915/ttm Add a generic TTM memcpy move for page-based iomem 32e0235ecbbf drm/ttm: Export ttm_bo_tt_destroy()
a3796d8feda9 drm/i915/ttm: Embed a ttm buffer object in the i915 gem object 01c0540773c7 drm/i915/ttm Initialize the ttm device and memory managers 1b455d3307a9 drm/ttm: Export functions to initialize and finalize the ttm range manager standalone a1fb9cdc093f drm/i915: Fix i915_sg_page_sizes to record dma segments rather than physical pages 425af44621a2 drm/i915: Don't free shared locks while shared
f95eea0001a4 drm/i915: Untangle the vma pages_mutex

== Logs ==

For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/index.html

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

#61 lore raw

[Intel-gfx] ✓ Fi.CI.IGT: success for drm/i915: Move LMEM (VRAM) management over to TTM (rev2)

In reply to Thomas Hellström
== Series Details ==

== Series Details ==

Series: drm/i915: Move LMEM (VRAM) management over to TTM (rev2)
URL : https://patchwork.freedesktop.org/series/90022/
State : success

== Summary ==

CI Bug Log - changes from CI_DRM_10094_full -> Patchwork_20142_full
====================================================

Summary
-------

**WARNING**

Minor unknown changes coming with Patchwork_20142_full need to be verified manually.

If you think the reported changes have nothing to do with the changes introduced in Patchwork_20142_full, please notify your bug team to allow them to document this new failure mode, which will reduce false positives in CI.

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

Here are the unknown changes that may have been introduced in Patchwork_20142_full:

### IGT changes ###

#### Warnings ####

* igt@kms_vblank@pipe-a-ts-continuation-suspend:
- shard-kbl: [DMESG-WARN][1] ([i915#180] / [i915#295]) -> [INCOMPLETE][2]
[1]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-kbl7/igt@[email protected]
[2]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-kbl4/igt@[email protected]

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

Here are the changes found in Patchwork_20142_full that come from known issues:

### IGT changes ###

#### Issues hit ####

* igt@api_intel_allocator@execbuf-with-allocator:
- shard-kbl: NOTRUN -> [DMESG-WARN][3] ([i915#3457])
[3]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-kbl2/igt@[email protected]

* igt@api_intel_bb@intel-bb-blit-y:
- shard-glk: NOTRUN -> [FAIL][4] ([i915#3471])
[4]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-glk5/igt@[email protected]

* igt@gem_create@create-clear:
- shard-glk: [PASS][5] -> [FAIL][6] ([i915#1888] / [i915#3160])
[5]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-glk4/igt@[email protected]
[6]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-glk1/igt@[email protected]

* igt@gem_ctx_engines@invalid-engines:
- shard-kbl: NOTRUN -> [INCOMPLETE][7] ([i915#3468])
[7]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-kbl7/igt@[email protected]

* igt@gem_ctx_persistence@legacy-engines-cleanup:
- shard-snb: NOTRUN -> [SKIP][8] ([fdo#109271] / [i915#1099]) +1 similar issue
[8]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-snb2/igt@[email protected]

* igt@gem_ctx_ringsize@active@bcs0:
- shard-skl: [PASS][9] -> [INCOMPLETE][10] ([i915#3316] / [i915#3457])
[9]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-skl3/igt@gem_ctx_ringsize@[email protected]
[10]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-skl2/igt@gem_ctx_ringsize@[email protected]

* igt@gem_exec_fair@basic-pace@bcs0:
- shard-tglb: [PASS][11] -> [FAIL][12] ([i915#2842] / [i915#3457]) +1 similar issue
[11]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-tglb5/igt@gem_exec_fair@[email protected]
[12]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb8/igt@gem_exec_fair@[email protected]

* igt@gem_exec_flush@basic-batch-kernel-default-cmd:
- shard-snb: NOTRUN -> [SKIP][13] ([fdo#109271]) +185 similar issues
[13]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-snb5/igt@[email protected]

* igt@gem_exec_reloc@basic-wide-active@bcs0:
- shard-apl: NOTRUN -> [FAIL][14] ([i915#2389] / [i915#3457]) +3 similar issues
[14]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-apl6/igt@gem_exec_reloc@[email protected]

* igt@gem_exec_schedule@independent@rcs0:
- shard-glk: NOTRUN -> [FAIL][15] ([i915#3457]) +1 similar issue
[15]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-glk5/igt@gem_exec_schedule@[email protected]

* igt@gem_exec_schedule@preempt-hang@vcs0:
- shard-glk: [PASS][16] -> [FAIL][17] ([i915#3457]) +33 similar issues
[16]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-glk7/igt@gem_exec_schedule@[email protected]
[17]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-glk6/igt@gem_exec_schedule@[email protected]

* igt@gem_exec_schedule@u-submit-golden-slice@vecs0:
- shard-apl: NOTRUN -> [FAIL][18] ([i915#3457]) +10 similar issues
[18]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-apl7/igt@gem_exec_schedule@[email protected]

* igt@gem_exec_schedule@u-submit-late-slice@vecs0:
- shard-apl: [PASS][19] -> [FAIL][20] ([i915#3457])
[19]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-apl7/igt@gem_exec_schedule@[email protected]
[20]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-apl7/igt@gem_exec_schedule@[email protected]

* igt@gem_mmap_gtt@cpuset-basic-small-copy-xy:
- shard-snb: NOTRUN -> [INCOMPLETE][21] ([i915#3468])
[21]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-snb5/igt@[email protected]

* igt@gem_mmap_gtt@cpuset-medium-copy-xy:
- shard-tglb: [PASS][22] -> [INCOMPLETE][23] ([i915#3468] / [i915#750])
[22]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-tglb3/igt@[email protected]
[23]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb3/igt@[email protected]
- shard-iclb: [PASS][24] -> [INCOMPLETE][25] ([i915#3468]) +1 similar issue
[24]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-iclb3/igt@[email protected]
[25]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-iclb1/igt@[email protected]

* igt@gem_render_copy@mixed-tiled-to-yf-tiled-ccs:
- shard-apl: NOTRUN -> [INCOMPLETE][26] ([i915#3468]) +2 similar issues
[26]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-apl6/igt@[email protected]

* igt@gem_render_copy@y-tiled-mc-ccs-to-yf-tiled-ccs:
- shard-tglb: NOTRUN -> [INCOMPLETE][27] ([i915#3468])
[27]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb1/igt@[email protected]

* igt@gem_render_copy@yf-tiled-ccs-to-linear:
- shard-skl: NOTRUN -> [INCOMPLETE][28] ([i915#198] / [i915#3468])
[28]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-skl1/igt@[email protected]

* igt@gem_render_copy@yf-tiled-ccs-to-y-tiled-ccs:
- shard-skl: NOTRUN -> [INCOMPLETE][29] ([i915#3468]) +1 similar issue
[29]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-skl2/igt@[email protected]

* igt@gem_userptr_blits@set-cache-level:
- shard-snb: NOTRUN -> [FAIL][30] ([i915#3324])
[30]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-snb5/igt@[email protected]

* igt@gem_userptr_blits@vma-merge:
- shard-apl: NOTRUN -> [FAIL][31] ([i915#3318] / [i915#3457])
[31]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-apl7/igt@[email protected]
- shard-tglb: NOTRUN -> [FAIL][32] ([i915#3318] / [i915#3457])
[32]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb1/igt@[email protected]

* igt@gen7_exec_parse@oacontrol-tracking:
- shard-tglb: NOTRUN -> [SKIP][33] ([fdo#109289])
[33]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb1/igt@[email protected]

* igt@gen9_exec_parse@bb-large:
- shard-kbl: NOTRUN -> [FAIL][34] ([i915#3296])
[34]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-kbl7/igt@[email protected]

* igt@gen9_exec_parse@cmd-crossing-page:
- shard-tglb: NOTRUN -> [SKIP][35] ([fdo#112306])
[35]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb1/igt@[email protected]

* igt@i915_hangman@error-state-capture@bcs0:
- shard-snb: NOTRUN -> [DMESG-WARN][36] ([i915#3457]) +3 similar issues
[36]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-snb2/igt@i915_hangman@[email protected]

* igt@i915_module_load@reload-with-fault-injection:
- shard-snb: [PASS][37] -> [DMESG-WARN][38] ([i915#3389])
[37]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-snb6/igt@[email protected]
[38]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-snb5/igt@[email protected]

* igt@i915_pm_lpsp@kms-lpsp@kms-lpsp-dp:
- shard-apl: NOTRUN -> [SKIP][39] ([fdo#109271] / [i915#1937])
[39]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-apl7/igt@i915_pm_lpsp@[email protected]

* igt@i915_pm_rpm@cursor:
- shard-kbl: [PASS][40] -> [DMESG-WARN][41] ([i915#3457])
[40]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-kbl4/igt@[email protected]
[41]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-kbl4/igt@[email protected]
- shard-glk: [PASS][42] -> [DMESG-WARN][43] ([i915#3457]) +2 similar issues
[42]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-glk3/igt@[email protected]
[43]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-glk4/igt@[email protected]

* igt@i915_pm_rpm@cursor-dpms:
- shard-iclb: [PASS][44] -> [DMESG-WARN][45] ([i915#3457])
[44]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-iclb6/igt@[email protected]
[45]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-iclb3/igt@[email protected]

* igt@i915_pm_rpm@gem-mmap-type@uc:
- shard-glk: [PASS][46] -> [DMESG-WARN][47] ([i915#3475])
[46]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-glk6/igt@i915_pm_rpm@[email protected]
[47]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-glk5/igt@i915_pm_rpm@[email protected]

* igt@i915_pm_rpm@modeset-pc8-residency-stress:
- shard-apl: NOTRUN -> [SKIP][48] ([fdo#109271]) +191 similar issues
[48]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-apl7/igt@[email protected]
- shard-tglb: NOTRUN -> [SKIP][49] ([fdo#109506] / [i915#2411])
[49]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb1/igt@[email protected]

* igt@i915_pm_rpm@system-suspend:
- shard-skl: [PASS][50] -> [INCOMPLETE][51] ([i915#151])
[50]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-skl8/igt@[email protected]
[51]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-skl2/igt@[email protected]

* igt@i915_pm_rps@min-max-config-loaded:
- shard-tglb: NOTRUN -> [DMESG-WARN][52] ([i915#3457])
[52]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb1/igt@[email protected]

* igt@i915_selftest@live@execlists:
- shard-skl: NOTRUN -> [INCOMPLETE][53] ([i915#2782] / [i915#3462])
[53]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-skl1/igt@i915_selftest@[email protected]
- shard-apl: NOTRUN -> [DMESG-FAIL][54] ([i915#3462])
[54]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-apl7/igt@i915_selftest@[email protected]

* igt@i915_selftest@live@gt_pm:
- shard-skl: NOTRUN -> [DMESG-FAIL][55] ([i915#1886] / [i915#2291])
[55]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-skl1/igt@i915_selftest@live@gt_pm.html

* igt@i915_selftest@live@mman:
- shard-apl: NOTRUN -> [DMESG-WARN][56] ([i915#3457]) +5 similar issues
[56]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-apl7/igt@i915_selftest@[email protected]
- shard-skl: NOTRUN -> [DMESG-WARN][57] ([i915#3457]) +1 similar issue
[57]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-skl1/igt@i915_selftest@[email protected]

* igt@i915_suspend@forcewake:
- shard-kbl: [PASS][58] -> [DMESG-WARN][59] ([i915#180]) +3 similar issues
[58]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-kbl1/igt@[email protected]
[59]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-kbl6/igt@[email protected]

* igt@kms_big_fb@y-tiled-8bpp-rotate-90:
- shard-tglb: NOTRUN -> [SKIP][60] ([fdo#111614]) +1 similar issue
[60]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb1/igt@[email protected]

* igt@kms_big_joiner@basic:
- shard-skl: NOTRUN -> [SKIP][61] ([fdo#109271] / [i915#2705])
[61]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-skl2/igt@[email protected]

* igt@kms_ccs@pipe-c-ccs-on-another-bo:
- shard-glk: NOTRUN -> [SKIP][62] ([fdo#109271]) +15 similar issues
[62]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-glk5/igt@[email protected]

* igt@kms_chamelium@hdmi-edid-read:
- shard-tglb: NOTRUN -> [SKIP][63] ([fdo#109284] / [fdo#111827]) +4 similar issues
[63]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb1/igt@[email protected]

* igt@kms_chamelium@hdmi-hpd-with-enabled-mode:
- shard-kbl: NOTRUN -> [SKIP][64] ([fdo#109271] / [fdo#111827]) +4 similar issues
[64]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-kbl2/igt@[email protected]

* igt@kms_color_chamelium@pipe-a-ctm-limited-range:
- shard-apl: NOTRUN -> [SKIP][65] ([fdo#109271] / [fdo#111827]) +20 similar issues
[65]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-apl7/igt@[email protected]

* igt@kms_color_chamelium@pipe-c-gamma:
- shard-glk: NOTRUN -> [SKIP][66] ([fdo#109271] / [fdo#111827]) +1 similar issue
[66]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-glk5/igt@[email protected]

* igt@kms_color_chamelium@pipe-d-degamma:
- shard-skl: NOTRUN -> [SKIP][67] ([fdo#109271] / [fdo#111827]) +5 similar issues
[67]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-skl3/igt@[email protected]

* igt@kms_color_chamelium@pipe-invalid-ctm-matrix-sizes:
- shard-snb: NOTRUN -> [SKIP][68] ([fdo#109271] / [fdo#111827]) +9 similar issues
[68]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-snb5/igt@[email protected]

* igt@kms_cursor_crc@pipe-a-cursor-512x512-rapid-movement:
- shard-tglb: NOTRUN -> [SKIP][69] ([fdo#109279] / [i915#3359] / [i915#3457])
[69]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb1/igt@[email protected]

* igt@kms_cursor_crc@pipe-a-cursor-64x21-random:
- shard-kbl: NOTRUN -> [FAIL][70] ([i915#3444] / [i915#3457]) +5 similar issues
[70]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-kbl3/igt@[email protected]

* igt@kms_cursor_crc@pipe-a-cursor-dpms:
- shard-skl: [PASS][71] -> [FAIL][72] ([i915#3444] / [i915#3457])
[71]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-skl2/igt@[email protected]
[72]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-skl3/igt@[email protected]

* igt@kms_cursor_crc@pipe-b-cursor-32x10-onscreen:
- shard-tglb: NOTRUN -> [SKIP][73] ([i915#3359] / [i915#3457])
[73]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb1/igt@[email protected]

* igt@kms_cursor_crc@pipe-b-cursor-alpha-opaque:
- shard-apl: NOTRUN -> [FAIL][74] ([i915#3444] / [i915#3457]) +8 similar issues
[74]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-apl6/igt@[email protected]

* igt@kms_cursor_crc@pipe-b-cursor-suspend:
- shard-skl: NOTRUN -> [FAIL][75] ([i915#3444] / [i915#3457]) +6 similar issues
[75]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-skl3/igt@[email protected]

* igt@kms_cursor_crc@pipe-c-cursor-128x128-sliding:
- shard-glk: NOTRUN -> [FAIL][76] ([i915#3444] / [i915#3457])
[76]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-glk5/igt@[email protected]

* igt@kms_cursor_crc@pipe-c-cursor-256x85-random:
- shard-snb: NOTRUN -> [SKIP][77] ([fdo#109271] / [i915#3457]) +27 similar issues
[77]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-snb2/igt@[email protected]

* igt@kms_cursor_crc@pipe-c-cursor-32x32-onscreen:
- shard-glk: NOTRUN -> [SKIP][78] ([fdo#109271] / [i915#3457]) +4 similar issues
[78]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-glk5/igt@[email protected]

* igt@kms_cursor_crc@pipe-c-cursor-512x170-rapid-movement:
- shard-kbl: NOTRUN -> [SKIP][79] ([fdo#109271] / [i915#3457]) +9 similar issues
[79]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-kbl3/igt@[email protected]

* igt@kms_cursor_crc@pipe-c-cursor-dpms:
- shard-iclb: [PASS][80] -> [FAIL][81] ([i915#3457])
[80]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-iclb5/igt@[email protected]
[81]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-iclb1/igt@[email protected]

* igt@kms_cursor_crc@pipe-d-cursor-128x42-onscreen:
- shard-tglb: NOTRUN -> [FAIL][82] ([i915#2124] / [i915#3457]) +3 similar issues
[82]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb1/igt@[email protected]

* igt@kms_cursor_crc@pipe-d-cursor-256x85-offscreen:
- shard-tglb: [PASS][83] -> [FAIL][84] ([i915#2124] / [i915#3457]) +1 similar issue
[83]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-tglb3/igt@[email protected]
[84]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb5/igt@[email protected]

* igt@kms_cursor_crc@pipe-d-cursor-32x32-offscreen:
- shard-tglb: NOTRUN -> [SKIP][85] ([i915#3319] / [i915#3457])
[85]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb1/igt@[email protected]

* igt@kms_cursor_crc@pipe-d-cursor-64x64-sliding:
- shard-apl: NOTRUN -> [SKIP][86] ([fdo#109271] / [i915#3457]) +27 similar issues
[86]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-apl7/igt@[email protected]

* igt@kms_cursor_crc@pipe-d-cursor-alpha-opaque:
- shard-skl: NOTRUN -> [SKIP][87] ([fdo#109271] / [i915#3457]) +6 similar issues
[87]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-skl1/igt@[email protected]

* igt@kms_fbcon_fbt@fbc-suspend:
- shard-kbl: NOTRUN -> [INCOMPLETE][88] ([i915#155] / [i915#180] / [i915#636])
[88]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-kbl7/igt@[email protected]

* igt@kms_frontbuffer_tracking@fbc-2p-primscrn-spr-indfb-draw-blt:
- shard-kbl: NOTRUN -> [SKIP][89] ([fdo#109271]) +72 similar issues
[89]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-kbl3/igt@[email protected]

* igt@kms_frontbuffer_tracking@fbc-2p-scndscrn-spr-indfb-move:
- shard-tglb: NOTRUN -> [SKIP][90] ([fdo#111825]) +9 similar issues
[90]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb1/igt@[email protected]

* igt@kms_pipe_crc_basic@suspend-read-crc-pipe-d:
- shard-kbl: NOTRUN -> [SKIP][91] ([fdo#109271] / [i915#533]) +1 similar issue
[91]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-kbl3/igt@[email protected]
- shard-apl: NOTRUN -> [SKIP][92] ([fdo#109271] / [i915#533])
[92]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-apl7/igt@[email protected]

* igt@kms_plane_alpha_blend@pipe-b-alpha-opaque-fb:
- shard-skl: NOTRUN -> [FAIL][93] ([fdo#108145] / [i915#265])
[93]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-skl2/igt@[email protected]

* igt@kms_plane_alpha_blend@pipe-b-alpha-transparent-fb:
- shard-apl: NOTRUN -> [FAIL][94] ([i915#265]) +1 similar issue
[94]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-apl6/igt@[email protected]
- shard-skl: NOTRUN -> [FAIL][95] ([i915#265])
[95]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-skl3/igt@[email protected]

* igt@kms_plane_alpha_blend@pipe-c-alpha-basic:
- shard-kbl: NOTRUN -> [FAIL][96] ([fdo#108145] / [i915#265]) +2 similar issues
[96]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-kbl3/igt@[email protected]
- shard-apl: NOTRUN -> [FAIL][97] ([fdo#108145] / [i915#265]) +2 similar issues
[97]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-apl7/igt@[email protected]

* igt@kms_plane_alpha_blend@pipe-c-alpha-transparent-fb:
- shard-glk: NOTRUN -> [FAIL][98] ([i915#265])
[98]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-glk5/igt@[email protected]

* igt@kms_plane_cursor@pipe-a-primary-size-64:
- shard-snb: NOTRUN -> [FAIL][99] ([i915#2657] / [i915#3457])
[99]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-snb7/igt@[email protected]

* igt@kms_plane_cursor@pipe-a-viewport-size-64:
- shard-apl: NOTRUN -> [FAIL][100] ([i915#2657] / [i915#3457])
[100]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-apl7/igt@[email protected]

* igt@kms_plane_cursor@pipe-b-overlay-size-128:
- shard-snb: NOTRUN -> [FAIL][101] ([i915#2657])
[101]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-snb2/igt@[email protected]

* igt@kms_plane_cursor@pipe-b-overlay-size-256:
- shard-glk: [PASS][102] -> [FAIL][103] ([i915#2657])
[102]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-glk3/igt@[email protected]
[103]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-glk4/igt@[email protected]

* igt@kms_plane_cursor@pipe-b-overlay-size-64:
- shard-kbl: NOTRUN -> [FAIL][104] ([i915#2657] / [i915#3457]) +2 similar issues
[104]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-kbl3/igt@[email protected]

* igt@kms_plane_cursor@pipe-b-primary-size-128:
- shard-tglb: NOTRUN -> [FAIL][105] ([i915#3461])
[105]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb1/igt@[email protected]

* igt@kms_plane_cursor@pipe-b-viewport-size-128:
- shard-glk: NOTRUN -> [FAIL][106] ([i915#2657]) +1 similar issue
[106]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-glk5/igt@[email protected]

* igt@kms_plane_cursor@pipe-b-viewport-size-256:
- shard-skl: NOTRUN -> [FAIL][107] ([i915#2657]) +1 similar issue
[107]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-skl3/igt@[email protected]

* igt@kms_plane_cursor@pipe-c-overlay-size-128:
- shard-kbl: NOTRUN -> [FAIL][108] ([i915#2657])
[108]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-kbl3/igt@[email protected]

* igt@kms_plane_lowres@pipe-d-tiling-yf:
- shard-tglb: NOTRUN -> [SKIP][109] ([fdo#112054])
[109]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb1/igt@[email protected]

* igt@kms_plane_multiple@atomic-pipe-a-tiling-yf:
- shard-tglb: NOTRUN -> [SKIP][110] ([fdo#111615])
[110]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb1/igt@[email protected]

* igt@kms_psr2_sf@overlay-plane-update-sf-dmg-area-2:
- shard-apl: NOTRUN -> [SKIP][111] ([fdo#109271] / [i915#658]) +5 similar issues
[111]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-apl7/igt@[email protected]
- shard-tglb: NOTRUN -> [SKIP][112] ([i915#2920])
[112]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb1/igt@[email protected]

* igt@kms_psr2_sf@overlay-plane-update-sf-dmg-area-3:
- shard-skl: NOTRUN -> [SKIP][113] ([fdo#109271] / [i915#658]) +1 similar issue
[113]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-skl3/igt@[email protected]

* igt@kms_psr2_sf@overlay-primary-update-sf-dmg-area-3:
- shard-kbl: NOTRUN -> [SKIP][114] ([fdo#109271] / [i915#658]) +2 similar issues
[114]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-kbl2/igt@[email protected]

* igt@kms_psr@psr2_cursor_render:
- shard-tglb: NOTRUN -> [FAIL][115] ([i915#132] / [i915#3457])
[115]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb1/igt@kms_psr@psr2_cursor_render.html

* igt@kms_writeback@writeback-invalid-parameters:
- shard-apl: NOTRUN -> [SKIP][116] ([fdo#109271] / [i915#2437])
[116]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-apl7/igt@[email protected]

* igt@kms_writeback@writeback-pixel-formats:
- shard-skl: NOTRUN -> [SKIP][117] ([fdo#109271] / [i915#2437]) +1 similar issue
[117]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-skl2/igt@[email protected]

* igt@nouveau_crc@pipe-c-source-outp-complete:
- shard-tglb: NOTRUN -> [SKIP][118] ([i915#2530]) +1 similar issue
[118]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb1/igt@[email protected]

* igt@perf@polling-parameterized:
- shard-glk: [PASS][119] -> [FAIL][120] ([i915#1542])
[119]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-glk1/igt@[email protected]
[120]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-glk7/igt@[email protected]

* igt@perf_pmu@busy-idle-no-semaphores@rcs0:
- shard-glk: [PASS][121] -> [WARN][122] ([i915#3457])
[121]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-glk1/igt@perf_pmu@[email protected]
[122]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-glk7/igt@perf_pmu@[email protected]

* igt@perf_pmu@module-unload:
- shard-skl: [PASS][123] -> [DMESG-WARN][124] ([i915#1982] / [i915#262])
[123]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-skl10/igt@[email protected]
[124]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-skl3/igt@[email protected]

* igt@prime_nv_api@i915_self_import:
- shard-skl: NOTRUN -> [SKIP][125] ([fdo#109271]) +48 similar issues
[125]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-skl2/igt@prime_nv_api@i915_self_import.html

* igt@prime_nv_test@i915_import_cpu_mmap:
- shard-tglb: NOTRUN -> [SKIP][126] ([fdo#109291])
[126]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-tglb1/igt@prime_nv_test@i915_import_cpu_mmap.html

* igt@sysfs_clients@sema-50:
- shard-apl: NOTRUN -> [SKIP][127] ([fdo#109271] / [i915#2994]) +2 similar issues
[127]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-apl7/igt@[email protected]

* igt@sysfs_clients@split-50:
- shard-kbl: NOTRUN -> [SKIP][128] ([fdo#109271] / [i915#2994]) +2 similar issues
[128]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-kbl7/igt@[email protected]

#### Possible fixes ####

* igt@gem_exec_fair@basic-pace@vcs1:
- shard-kbl: [FAIL][129] ([i915#2842] / [i915#3457]) -> [PASS][130]
[129]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_10094/shard-kbl1/igt@gem_exec_fair@[email protected]
[130]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/shard-kbl1/igt@gem_exec_fair@[email protected]

* igt@gem_exec_fence@parallel@vcs0:
- shard-glk: [FAIL][

== Logs ==

For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_20142/index.html

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx