Merge tag 'kvmarm-fixes-6.12-1' of git://git.kernel.org/pub/scm/linux/kernel/git/kvmarm/kvmarm into HEAD

KVM/arm64 fixes for 6.12, take #1

- Fix pKVM error path on init, making sure we do not change critical
  system registers as we're about to fail

- Make sure that the host's vector length is at capped by a value
  common to all CPUs

- Fix kvm_has_feat*() handling of "negative" features, as the current
  code is pretty broken

- Promote Joey to the status of official reviewer, while James steps
  down -- hopefully only temporarly

1 files changed, 111 insertions, 167 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c
index 8d75061f9f38..9d4f5352a62c 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_5.c
@@ -52,6 +52,42 @@
 #define RDECODE_MSG_CREATE							0x00000000
 #define RDECODE_MESSAGE_CREATE						0x00000001
 
+static const struct amdgpu_hwip_reg_entry vcn_reg_list_4_0_5[] = {
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_POWER_STATUS),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_STATUS),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_CONTEXT_ID),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_CONTEXT_ID2),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_GPCOM_VCPU_DATA0),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_GPCOM_VCPU_DATA1),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_GPCOM_VCPU_CMD),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_RB_BASE_HI),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_RB_BASE_LO),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_RB_BASE_HI2),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_RB_BASE_LO2),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_RB_BASE_HI3),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_RB_BASE_LO3),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_RB_BASE_HI4),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_RB_BASE_LO4),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_RB_RPTR),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_RB_WPTR),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_RB_RPTR2),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_RB_WPTR2),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_RB_RPTR3),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_RB_WPTR3),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_RB_RPTR4),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_RB_WPTR4),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_RB_SIZE),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_RB_SIZE2),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_RB_SIZE3),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_RB_SIZE4),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_PGFSM_CONFIG),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_PGFSM_STATUS),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_DPG_LMA_CTL),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_DPG_LMA_DATA),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_DPG_LMA_MASK),
+	SOC15_REG_ENTRY_STR(VCN, 0, regUVD_DPG_PAUSE)
+};
+
 static int amdgpu_ih_clientid_vcns[] = {
 	SOC15_IH_CLIENTID_VCN,
 	SOC15_IH_CLIENTID_VCN1
@@ -97,6 +133,8 @@ static int vcn_v4_0_5_sw_init(void *handle)
 	struct amdgpu_ring *ring;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 	int i, r;
+	uint32_t reg_count = ARRAY_SIZE(vcn_reg_list_4_0_5);
+	uint32_t *ptr;
 
 	r = amdgpu_vcn_sw_init(adev);
 	if (r)
@@ -168,6 +206,14 @@ static int vcn_v4_0_5_sw_init(void *handle)
 	if (adev->pg_flags & AMD_PG_SUPPORT_VCN_DPG)
 		adev->vcn.pause_dpg_mode = vcn_v4_0_5_pause_dpg_mode;
 
+	/* Allocate memory for VCN IP Dump buffer */
+	ptr = kcalloc(adev->vcn.num_vcn_inst * reg_count, sizeof(uint32_t), GFP_KERNEL);
+	if (!ptr) {
+		DRM_ERROR("Failed to allocate memory for VCN IP Dump\n");
+		adev->vcn.ip_dump = NULL;
+	} else {
+		adev->vcn.ip_dump = ptr;
+	}
 	return 0;
 }
 
@@ -207,6 +253,8 @@ static int vcn_v4_0_5_sw_fini(void *handle)
 
 	r = amdgpu_vcn_sw_fini(adev);
 
+	kfree(adev->vcn.ip_dump);
+
 	return r;
 }
 
@@ -1347,170 +1395,6 @@ static void vcn_v4_0_5_unified_ring_set_wptr(struct amdgpu_ring *ring)
 	}
 }
 
-static int vcn_v4_0_5_limit_sched(struct amdgpu_cs_parser *p,
-				struct amdgpu_job *job)
-{
-	struct drm_gpu_scheduler **scheds;
-
-	/* The create msg must be in the first IB submitted */
-	if (atomic_read(&job->base.entity->fence_seq))
-		return -EINVAL;
-
-	/* if VCN0 is harvested, we can't support AV1 */
-	if (p->adev->vcn.harvest_config & AMDGPU_VCN_HARVEST_VCN0)
-		return -EINVAL;
-
-	scheds = p->adev->gpu_sched[AMDGPU_HW_IP_VCN_ENC]
-		[AMDGPU_RING_PRIO_0].sched;
-	drm_sched_entity_modify_sched(job->base.entity, scheds, 1);
-	return 0;
-}
-
-static int vcn_v4_0_5_dec_msg(struct amdgpu_cs_parser *p, struct amdgpu_job *job,
-			    uint64_t addr)
-{
-	struct ttm_operation_ctx ctx = { false, false };
-	struct amdgpu_bo_va_mapping *map;
-	uint32_t *msg, num_buffers;
-	struct amdgpu_bo *bo;
-	uint64_t start, end;
-	unsigned int i;
-	void *ptr;
-	int r;
-
-	addr &= AMDGPU_GMC_HOLE_MASK;
-	r = amdgpu_cs_find_mapping(p, addr, &bo, &map);
-	if (r) {
-		DRM_ERROR("Can't find BO for addr 0x%08llx\n", addr);
-		return r;
-	}
-
-	start = map->start * AMDGPU_GPU_PAGE_SIZE;
-	end = (map->last + 1) * AMDGPU_GPU_PAGE_SIZE;
-	if (addr & 0x7) {
-		DRM_ERROR("VCN messages must be 8 byte aligned!\n");
-		return -EINVAL;
-	}
-
-	bo->flags |= AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED;
-	amdgpu_bo_placement_from_domain(bo, bo->allowed_domains);
-	r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
-	if (r) {
-		DRM_ERROR("Failed validating the VCN message BO (%d)!\n", r);
-		return r;
-	}
-
-	r = amdgpu_bo_kmap(bo, &ptr);
-	if (r) {
-		DRM_ERROR("Failed mapping the VCN message (%d)!\n", r);
-		return r;
-	}
-
-	msg = ptr + addr - start;
-
-	/* Check length */
-	if (msg[1] > end - addr) {
-		r = -EINVAL;
-		goto out;
-	}
-
-	if (msg[3] != RDECODE_MSG_CREATE)
-		goto out;
-
-	num_buffers = msg[2];
-	for (i = 0, msg = &msg[6]; i < num_buffers; ++i, msg += 4) {
-		uint32_t offset, size, *create;
-
-		if (msg[0] != RDECODE_MESSAGE_CREATE)
-			continue;
-
-		offset = msg[1];
-		size = msg[2];
-
-		if (offset + size > end) {
-			r = -EINVAL;
-			goto out;
-		}
-
-		create = ptr + addr + offset - start;
-
-		/* H264, HEVC and VP9 can run on any instance */
-		if (create[0] == 0x7 || create[0] == 0x10 || create[0] == 0x11)
-			continue;
-
-		r = vcn_v4_0_5_limit_sched(p, job);
-		if (r)
-			goto out;
-	}
-
-out:
-	amdgpu_bo_kunmap(bo);
-	return r;
-}
-
-#define RADEON_VCN_ENGINE_TYPE_ENCODE			(0x00000002)
-#define RADEON_VCN_ENGINE_TYPE_DECODE			(0x00000003)
-
-#define RADEON_VCN_ENGINE_INFO				(0x30000001)
-#define RADEON_VCN_ENGINE_INFO_MAX_OFFSET		16
-
-#define RENCODE_ENCODE_STANDARD_AV1			2
-#define RENCODE_IB_PARAM_SESSION_INIT			0x00000003
-#define RENCODE_IB_PARAM_SESSION_INIT_MAX_OFFSET	64
-
-/* return the offset in ib if id is found, -1 otherwise
- * to speed up the searching we only search upto max_offset
- */
-static int vcn_v4_0_5_enc_find_ib_param(struct amdgpu_ib *ib, uint32_t id, int max_offset)
-{
-	int i;
-
-	for (i = 0; i < ib->length_dw && i < max_offset && ib->ptr[i] >= 8; i += ib->ptr[i]/4) {
-		if (ib->ptr[i + 1] == id)
-			return i;
-	}
-	return -1;
-}
-
-static int vcn_v4_0_5_ring_patch_cs_in_place(struct amdgpu_cs_parser *p,
-					   struct amdgpu_job *job,
-					   struct amdgpu_ib *ib)
-{
-	struct amdgpu_ring *ring = amdgpu_job_ring(job);
-	struct amdgpu_vcn_decode_buffer *decode_buffer;
-	uint64_t addr;
-	uint32_t val;
-	int idx;
-
-	/* The first instance can decode anything */
-	if (!ring->me)
-		return 0;
-
-	/* RADEON_VCN_ENGINE_INFO is at the top of ib block */
-	idx = vcn_v4_0_5_enc_find_ib_param(ib, RADEON_VCN_ENGINE_INFO,
-			RADEON_VCN_ENGINE_INFO_MAX_OFFSET);
-	if (idx < 0) /* engine info is missing */
-		return 0;
-
-	val = amdgpu_ib_get_value(ib, idx + 2); /* RADEON_VCN_ENGINE_TYPE */
-	if (val == RADEON_VCN_ENGINE_TYPE_DECODE) {
-		decode_buffer = (struct amdgpu_vcn_decode_buffer *)&ib->ptr[idx + 6];
-
-		if (!(decode_buffer->valid_buf_flag  & 0x1))
-			return 0;
-
-		addr = ((u64)decode_buffer->msg_buffer_address_hi) << 32 |
-			decode_buffer->msg_buffer_address_lo;
-		return vcn_v4_0_5_dec_msg(p, job, addr);
-	} else if (val == RADEON_VCN_ENGINE_TYPE_ENCODE) {
-		idx = vcn_v4_0_5_enc_find_ib_param(ib, RENCODE_IB_PARAM_SESSION_INIT,
-			RENCODE_IB_PARAM_SESSION_INIT_MAX_OFFSET);
-		if (idx >= 0 && ib->ptr[idx + 2] == RENCODE_ENCODE_STANDARD_AV1)
-			return vcn_v4_0_5_limit_sched(p, job);
-	}
-	return 0;
-}
-
 static const struct amdgpu_ring_funcs vcn_v4_0_5_unified_ring_vm_funcs = {
 	.type = AMDGPU_RING_TYPE_VCN_ENC,
 	.align_mask = 0x3f,
@@ -1518,7 +1402,6 @@ static const struct amdgpu_ring_funcs vcn_v4_0_5_unified_ring_vm_funcs = {
 	.get_rptr = vcn_v4_0_5_unified_ring_get_rptr,
 	.get_wptr = vcn_v4_0_5_unified_ring_get_wptr,
 	.set_wptr = vcn_v4_0_5_unified_ring_set_wptr,
-	.patch_cs_in_place = vcn_v4_0_5_ring_patch_cs_in_place,
 	.emit_frame_size =
 		SOC15_FLUSH_GPU_TLB_NUM_WREG * 3 +
 		SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 4 +
@@ -1733,6 +1616,67 @@ static void vcn_v4_0_5_set_irq_funcs(struct amdgpu_device *adev)
 	}
 }
 
+static void vcn_v4_0_5_print_ip_state(void *handle, struct drm_printer *p)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	int i, j;
+	uint32_t reg_count = ARRAY_SIZE(vcn_reg_list_4_0_5);
+	uint32_t inst_off, is_powered;
+
+	if (!adev->vcn.ip_dump)
+		return;
+
+	drm_printf(p, "num_instances:%d\n", adev->vcn.num_vcn_inst);
+	for (i = 0; i < adev->vcn.num_vcn_inst; i++) {
+		if (adev->vcn.harvest_config & (1 << i)) {
+			drm_printf(p, "\nHarvested Instance:VCN%d Skipping dump\n", i);
+			continue;
+		}
+
+		inst_off = i * reg_count;
+		is_powered = (adev->vcn.ip_dump[inst_off] &
+				UVD_POWER_STATUS__UVD_POWER_STATUS_MASK) != 1;
+
+		if (is_powered) {
+			drm_printf(p, "\nActive Instance:VCN%d\n", i);
+			for (j = 0; j < reg_count; j++)
+				drm_printf(p, "%-50s \t 0x%08x\n", vcn_reg_list_4_0_5[j].reg_name,
+					   adev->vcn.ip_dump[inst_off + j]);
+		} else {
+			drm_printf(p, "\nInactive Instance:VCN%d\n", i);
+		}
+	}
+}
+
+static void vcn_v4_0_5_dump_ip_state(void *handle)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	int i, j;
+	bool is_powered;
+	uint32_t inst_off;
+	uint32_t reg_count = ARRAY_SIZE(vcn_reg_list_4_0_5);
+
+	if (!adev->vcn.ip_dump)
+		return;
+
+	for (i = 0; i < adev->vcn.num_vcn_inst; i++) {
+		if (adev->vcn.harvest_config & (1 << i))
+			continue;
+
+		inst_off = i * reg_count;
+		/* mmUVD_POWER_STATUS is always readable and is first element of the array */
+		adev->vcn.ip_dump[inst_off] = RREG32_SOC15(VCN, i, regUVD_POWER_STATUS);
+		is_powered = (adev->vcn.ip_dump[inst_off] &
+				UVD_POWER_STATUS__UVD_POWER_STATUS_MASK) != 1;
+
+		if (is_powered)
+			for (j = 1; j < reg_count; j++)
+				adev->vcn.ip_dump[inst_off + j] =
+					RREG32(SOC15_REG_ENTRY_OFFSET_INST(vcn_reg_list_4_0_5[j],
+									   i));
+	}
+}
+
 static const struct amd_ip_funcs vcn_v4_0_5_ip_funcs = {
 	.name = "vcn_v4_0_5",
 	.early_init = vcn_v4_0_5_early_init,
@@ -1751,8 +1695,8 @@ static const struct amd_ip_funcs vcn_v4_0_5_ip_funcs = {
 	.post_soft_reset = NULL,
 	.set_clockgating_state = vcn_v4_0_5_set_clockgating_state,
 	.set_powergating_state = vcn_v4_0_5_set_powergating_state,
-	.dump_ip_state = NULL,
-	.print_ip_state = NULL,
+	.dump_ip_state = vcn_v4_0_5_dump_ip_state,
+	.print_ip_state = vcn_v4_0_5_print_ip_state,
 };
 
 const struct amdgpu_ip_block_version vcn_v4_0_5_ip_block = {