Merge tag 'amd-drm-next-6.18-2025-09-09' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-next-6.18-2025-09-09:

amdgpu:
- Add CRIU support for gem objects
- SI UVD fix
- SI DPM fixes
- Misc code cleanups
- RAS updates
- GPUVM debugfs fixes
- Cyan Skillfish updates
- UserQ updates
- OEM i2c fix
- SMU 13.0.x updates
- DPCD probe quirk fix
- Make vbios build number available in sysfs
- HDCP updates
- Brightness curve fixes
- eDP updates
- Vblank fixes
- DCN 3.5 PG fix
- PBN calcution fix

amdkfd:
- Add CRIU support for gem objects
- Flexible array fix
- P2P topology fix
- APU memlimit fixes
- Misc code cleanups

UAPI:
- Add CRIU support for gem objects
  Proposed userspace: checkpoint-restore/criu#2613

radeon:
- Use dev_warn_once() in CS parsers

Signed-off-by: Dave Airlie <airlied@redhat.com>

From: Alex Deucher <alexander.deucher@amd.com>
Link: https://lore.kernel.org/r/20250909161928.942785-1-alexander.deucher@amd.com

Author: airlied

drivers/gpu/drm/amd/amdgpu/Makefile

@@ -84,7 +84,8 @@ amdgpu-y += \
 	vega20_reg_init.o nbio_v7_4.o nbio_v2_3.o nv.o arct_reg_init.o mxgpu_nv.o \
 	nbio_v7_2.o hdp_v4_0.o hdp_v5_0.o aldebaran_reg_init.o aldebaran.o soc21.o soc24.o \
 	sienna_cichlid.o smu_v13_0_10.o nbio_v4_3.o hdp_v6_0.o nbio_v7_7.o hdp_v5_2.o lsdma_v6_0.o \
-	nbio_v7_9.o aqua_vanjaram.o nbio_v7_11.o lsdma_v7_0.o hdp_v7_0.o nbif_v6_3_1.o
+	nbio_v7_9.o aqua_vanjaram.o nbio_v7_11.o lsdma_v7_0.o hdp_v7_0.o nbif_v6_3_1.o \
+	cyan_skillfish_reg_init.o
 
 # add DF block
 amdgpu-y += \

drivers/gpu/drm/amd/amdgpu/amdgpu.h

@@ -1149,9 +1149,6 @@ struct amdgpu_device {
 	/* for userq and VM fences */
 	struct amdgpu_seq64		seq64;
 
-	/* KFD */
-	struct amdgpu_kfd_dev		kfd;
-
 	/* UMC */
 	struct amdgpu_umc		umc;
 
@@ -1314,6 +1311,11 @@ struct amdgpu_device {
 	struct mutex                    userq_mutex;
 	bool                            userq_halt_for_enforce_isolation;
 	struct amdgpu_uid *uid_info;
+
+	/* KFD
+	 * Must be last --ends in a flexible-array member.
+	 */
+	struct amdgpu_kfd_dev		kfd;
 };
 
 static inline uint32_t amdgpu_ip_version(const struct amdgpu_device *adev,

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h

@@ -107,11 +107,13 @@ struct amdgpu_kfd_dev {
 	bool init_complete;
 	struct work_struct reset_work;
 
-	/* HMM page migration MEMORY_DEVICE_PRIVATE mapping */
-	struct dev_pagemap pgmap;
-
 	/* Client for KFD BO GEM handle allocations */
 	struct drm_client_dev client;
+
+	/* HMM page migration MEMORY_DEVICE_PRIVATE mapping
+	 * Must be last --ends in a flexible-array member.
+	 */
+	struct dev_pagemap pgmap;
 };
 
 enum kgd_engine_type {

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c

@@ -213,19 +213,35 @@ int amdgpu_amdkfd_reserve_mem_limit(struct amdgpu_device *adev,
 	spin_lock(&kfd_mem_limit.mem_limit_lock);
 
 	if (kfd_mem_limit.system_mem_used + system_mem_needed >
-	    kfd_mem_limit.max_system_mem_limit)
+	    kfd_mem_limit.max_system_mem_limit) {
 		pr_debug("Set no_system_mem_limit=1 if using shared memory\n");
+		if (!no_system_mem_limit) {
+			ret = -ENOMEM;
+			goto release;
+		}
+	}
 
-	if ((kfd_mem_limit.system_mem_used + system_mem_needed >
-	     kfd_mem_limit.max_system_mem_limit && !no_system_mem_limit) ||
-	    (kfd_mem_limit.ttm_mem_used + ttm_mem_needed >
-	     kfd_mem_limit.max_ttm_mem_limit) ||
-	    (adev && xcp_id >= 0 && adev->kfd.vram_used[xcp_id] + vram_needed >
-	     vram_size - reserved_for_pt - reserved_for_ras - atomic64_read(&adev->vram_pin_size))) {
+	if (kfd_mem_limit.ttm_mem_used + ttm_mem_needed >
+		kfd_mem_limit.max_ttm_mem_limit) {
 		ret = -ENOMEM;
 		goto release;
 	}
 
+	/*if is_app_apu is false and apu_prefer_gtt is true, it is an APU with
+	 * carve out < gtt. In that case, VRAM allocation will go to gtt domain, skip
+	 * VRAM check since ttm_mem_limit check already cover this allocation
+	 */
+
+	if (adev && xcp_id >= 0 && (!adev->apu_prefer_gtt || adev->gmc.is_app_apu)) {
+		uint64_t vram_available =
+			vram_size - reserved_for_pt - reserved_for_ras -
+			atomic64_read(&adev->vram_pin_size);
+		if (adev->kfd.vram_used[xcp_id] + vram_needed > vram_available) {
+			ret = -ENOMEM;
+			goto release;
+		}
+	}
+
 	/* Update memory accounting by decreasing available system
 	 * memory, TTM memory and GPU memory as computed above
 	 */
@@ -1627,11 +1643,15 @@ size_t amdgpu_amdkfd_get_available_memory(struct amdgpu_device *adev,
 	uint64_t vram_available, system_mem_available, ttm_mem_available;
 
 	spin_lock(&kfd_mem_limit.mem_limit_lock);
-	vram_available = KFD_XCP_MEMORY_SIZE(adev, xcp_id)
-		- adev->kfd.vram_used_aligned[xcp_id]
-		- atomic64_read(&adev->vram_pin_size)
-		- reserved_for_pt
-		- reserved_for_ras;
+	if (adev->apu_prefer_gtt && !adev->gmc.is_app_apu)
+		vram_available = KFD_XCP_MEMORY_SIZE(adev, xcp_id)
+			- adev->kfd.vram_used_aligned[xcp_id];
+	else
+		vram_available = KFD_XCP_MEMORY_SIZE(adev, xcp_id)
+			- adev->kfd.vram_used_aligned[xcp_id]
+			- atomic64_read(&adev->vram_pin_size)
+			- reserved_for_pt
+			- reserved_for_ras;
 
 	if (adev->apu_prefer_gtt) {
 		system_mem_available = no_system_mem_limit ?

drivers/gpu/drm/amd/amdgpu/amdgpu_atombios.c

@@ -1816,16 +1816,43 @@ static ssize_t amdgpu_atombios_get_vbios_version(struct device *dev,
 	return sysfs_emit(buf, "%s\n", ctx->vbios_pn);
 }
 
+static ssize_t amdgpu_atombios_get_vbios_build(struct device *dev,
+					       struct device_attribute *attr,
+					       char *buf)
+{
+	struct drm_device *ddev = dev_get_drvdata(dev);
+	struct amdgpu_device *adev = drm_to_adev(ddev);
+	struct atom_context *ctx = adev->mode_info.atom_context;
+
+	return sysfs_emit(buf, "%s\n", ctx->build_num);
+}
+
 static DEVICE_ATTR(vbios_version, 0444, amdgpu_atombios_get_vbios_version,
 		   NULL);
+static DEVICE_ATTR(vbios_build, 0444, amdgpu_atombios_get_vbios_build, NULL);
 
 static struct attribute *amdgpu_vbios_version_attrs[] = {
-	&dev_attr_vbios_version.attr,
-	NULL
+	&dev_attr_vbios_version.attr, &dev_attr_vbios_build.attr, NULL
 };
 
+static umode_t amdgpu_vbios_version_attrs_is_visible(struct kobject *kobj,
+						     struct attribute *attr,
+						     int index)
+{
+	struct device *dev = kobj_to_dev(kobj);
+	struct drm_device *ddev = dev_get_drvdata(dev);
+	struct amdgpu_device *adev = drm_to_adev(ddev);
+	struct atom_context *ctx = adev->mode_info.atom_context;
+
+	if (attr == &dev_attr_vbios_build.attr && !strlen(ctx->build_num))
+		return 0;
+
+	return attr->mode;
+}
+
 const struct attribute_group amdgpu_vbios_version_attr_group = {
-	.attrs = amdgpu_vbios_version_attrs
+	.attrs = amdgpu_vbios_version_attrs,
+	.is_visible = amdgpu_vbios_version_attrs_is_visible,
 };
 
 int amdgpu_atombios_sysfs_init(struct amdgpu_device *adev)

drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c

@@ -1196,7 +1196,10 @@ static void amdgpu_connector_dvi_force(struct drm_connector *connector)
 }
 
 /**
- * Returns the maximum supported HDMI (TMDS) pixel clock in KHz.
+ * amdgpu_max_hdmi_pixel_clock - Return max supported HDMI (TMDS) pixel clock
+ * @adev: pointer to amdgpu_device
+ *
+ * Return: maximum supported HDMI (TMDS) pixel clock in KHz.
  */
 static int amdgpu_max_hdmi_pixel_clock(const struct amdgpu_device *adev)
 {
@@ -1209,8 +1212,14 @@ static int amdgpu_max_hdmi_pixel_clock(const struct amdgpu_device *adev)
 }
 
 /**
- * Validates the given display mode on DVI and HDMI connectors,
- * including analog signals on DVI-I.
+ * amdgpu_connector_dvi_mode_valid - Validate a mode on DVI/HDMI connectors
+ * @connector: DRM connector to validate the mode on
+ * @mode: display mode to validate
+ *
+ * Validate the given display mode on DVI and HDMI connectors, including
+ * analog signals on DVI-I.
+ *
+ * Return: drm_mode_status indicating whether the mode is valid.
  */
 static enum drm_mode_status amdgpu_connector_dvi_mode_valid(struct drm_connector *connector,
 					    const struct drm_display_mode *mode)

drivers/gpu/drm/amd/amdgpu/amdgpu_cper.c

@@ -68,7 +68,6 @@ void amdgpu_cper_entry_fill_hdr(struct amdgpu_device *adev,
 	hdr->error_severity		= sev;
 
 	hdr->valid_bits.platform_id	= 1;
-	hdr->valid_bits.partition_id	= 1;
 	hdr->valid_bits.timestamp	= 1;
 
 	amdgpu_cper_get_timestamp(&hdr->timestamp);
@@ -174,7 +173,7 @@ int amdgpu_cper_entry_fill_runtime_section(struct amdgpu_device *adev,
 	struct cper_sec_nonstd_err *section;
 	bool poison;
 
-	poison = (sev == CPER_SEV_NON_FATAL_CORRECTED) ? false : true;
+	poison = sev != CPER_SEV_NON_FATAL_CORRECTED;
 	section_desc = (struct cper_sec_desc *)((uint8_t *)hdr + SEC_DESC_OFFSET(idx));
 	section = (struct cper_sec_nonstd_err *)((uint8_t *)hdr +
 		   NONSTD_SEC_OFFSET(hdr->sec_cnt, idx));
@@ -220,7 +219,10 @@ int amdgpu_cper_entry_fill_bad_page_threshold_section(struct amdgpu_device *adev
 	section->hdr.valid_bits.err_context_cnt = 1;
 
 	section->info.error_type = RUNTIME;
+	section->info.valid_bits.ms_chk = 1;
 	section->info.ms_chk_bits.err_type_valid = 1;
+	section->info.ms_chk_bits.err_type = 1;
+	section->info.ms_chk_bits.pcc = 1;
 	section->ctx.reg_ctx_type = CPER_CTX_TYPE_CRASH;
 	section->ctx.reg_arr_size = sizeof(section->ctx.reg_dump);
 

drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c

@@ -2136,12 +2136,14 @@ static int amdgpu_pt_info_read(struct seq_file *m, void *unused)
 	struct drm_file *file;
 	struct amdgpu_fpriv *fpriv;
 	struct amdgpu_bo *root_bo;
+	struct amdgpu_device *adev;
 	int r;
 
 	file = m->private;
 	if (!file)
 		return -EINVAL;
 
+	adev = drm_to_adev(file->minor->dev);
 	fpriv = file->driver_priv;
 	if (!fpriv || !fpriv->vm.root.bo)
 		return -ENODEV;
@@ -2153,7 +2155,11 @@ static int amdgpu_pt_info_read(struct seq_file *m, void *unused)
 		return -EINVAL;
 	}
 
-	seq_printf(m, "gpu_address: 0x%llx\n", amdgpu_bo_gpu_offset(fpriv->vm.root.bo));
+	seq_printf(m, "pd_address: 0x%llx\n", amdgpu_gmc_pd_addr(fpriv->vm.root.bo));
+	seq_printf(m, "max_pfn: 0x%llx\n", adev->vm_manager.max_pfn);
+	seq_printf(m, "num_level: 0x%x\n", adev->vm_manager.num_level);
+	seq_printf(m, "block_size: 0x%x\n", adev->vm_manager.block_size);
+	seq_printf(m, "fragment_size: 0x%x\n", adev->vm_manager.fragment_size);
 
 	amdgpu_bo_unreserve(root_bo);
 	amdgpu_bo_unref(&root_bo);

drivers/gpu/drm/amd/amdgpu/amdgpu_device.c

@@ -95,6 +95,7 @@ MODULE_FIRMWARE("amdgpu/picasso_gpu_info.bin");
 MODULE_FIRMWARE("amdgpu/raven2_gpu_info.bin");
 MODULE_FIRMWARE("amdgpu/arcturus_gpu_info.bin");
 MODULE_FIRMWARE("amdgpu/navi12_gpu_info.bin");
+MODULE_FIRMWARE("amdgpu/cyan_skillfish_gpu_info.bin");
 
 #define AMDGPU_RESUME_MS		2000
 #define AMDGPU_MAX_RETRY_LIMIT		2
@@ -2629,6 +2630,9 @@ static int amdgpu_device_parse_gpu_info_fw(struct amdgpu_device *adev)
 			return 0;
 		chip_name = "navi12";
 		break;
+	case CHIP_CYAN_SKILLFISH:
+		chip_name = "cyan_skillfish";
+		break;
 	}
 
 	err = amdgpu_ucode_request(adev, &adev->firmware.gpu_info_fw,

drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c

@@ -2124,14 +2124,17 @@ static int amdgpu_discovery_set_smu_ip_blocks(struct amdgpu_device *adev)
 	case IP_VERSION(11, 0, 5):
 	case IP_VERSION(11, 0, 9):
 	case IP_VERSION(11, 0, 7):
-	case IP_VERSION(11, 0, 8):
 	case IP_VERSION(11, 0, 11):
 	case IP_VERSION(11, 0, 12):
 	case IP_VERSION(11, 0, 13):
 	case IP_VERSION(11, 5, 0):
 	case IP_VERSION(11, 5, 2):
 		amdgpu_device_ip_block_add(adev, &smu_v11_0_ip_block);
 		break;
+	case IP_VERSION(11, 0, 8):
+		if (adev->apu_flags & AMD_APU_IS_CYAN_SKILLFISH2)
+			amdgpu_device_ip_block_add(adev, &smu_v11_0_ip_block);
+		break;
 	case IP_VERSION(12, 0, 0):
 	case IP_VERSION(12, 0, 1):
 		amdgpu_device_ip_block_add(adev, &smu_v12_0_ip_block);
@@ -2746,6 +2749,36 @@ int amdgpu_discovery_set_ip_blocks(struct amdgpu_device *adev)
 		adev->ip_versions[UVD_HWIP][1] = IP_VERSION(2, 6, 0);
 		adev->ip_versions[XGMI_HWIP][0] = IP_VERSION(6, 1, 0);
 		break;
+	case CHIP_CYAN_SKILLFISH:
+		if (adev->apu_flags & AMD_APU_IS_CYAN_SKILLFISH2) {
+			r = amdgpu_discovery_reg_base_init(adev);
+			if (r)
+				return -EINVAL;
+
+			amdgpu_discovery_harvest_ip(adev);
+			amdgpu_discovery_get_gfx_info(adev);
+			amdgpu_discovery_get_mall_info(adev);
+			amdgpu_discovery_get_vcn_info(adev);
+		} else {
+			cyan_skillfish_reg_base_init(adev);
+			adev->sdma.num_instances = 2;
+			adev->ip_versions[MMHUB_HWIP][0] = IP_VERSION(2, 0, 3);
+			adev->ip_versions[ATHUB_HWIP][0] = IP_VERSION(2, 0, 3);
+			adev->ip_versions[OSSSYS_HWIP][0] = IP_VERSION(5, 0, 1);
+			adev->ip_versions[HDP_HWIP][0] = IP_VERSION(5, 0, 1);
+			adev->ip_versions[SDMA0_HWIP][0] = IP_VERSION(5, 0, 1);
+			adev->ip_versions[SDMA1_HWIP][1] = IP_VERSION(5, 0, 1);
+			adev->ip_versions[DF_HWIP][0] = IP_VERSION(3, 5, 0);
+			adev->ip_versions[NBIO_HWIP][0] = IP_VERSION(2, 1, 1);
+			adev->ip_versions[UMC_HWIP][0] = IP_VERSION(8, 1, 1);
+			adev->ip_versions[MP0_HWIP][0] = IP_VERSION(11, 0, 8);
+			adev->ip_versions[MP1_HWIP][0] = IP_VERSION(11, 0, 8);
+			adev->ip_versions[THM_HWIP][0] = IP_VERSION(11, 0, 1);
+			adev->ip_versions[SMUIO_HWIP][0] = IP_VERSION(11, 0, 8);
+			adev->ip_versions[GC_HWIP][0] = IP_VERSION(10, 1, 3);
+			adev->ip_versions[UVD_HWIP][0] = IP_VERSION(2, 0, 3);
+		}
+		break;
 	default:
 		r = amdgpu_discovery_reg_base_init(adev);
 		if (r) {