Merge tag 'amd-drm-fixes-6.10-2024-06-26' of...

Merge tag 'amd-drm-fixes-6.10-2024-06-26' of https://gitlab.freedesktop.org/agd5f/linux into drm-fixes

amd-drm-fixes-6.10-2024-06-26:

amdgpu:
- SMU 14.x fix
- vram info parsing fix
- mode1 reset fix
- LTTPR fix
- Virtual display fix
- Avoid spurious error in PSP init
Signed-off-by: Dave Airlie <airlied@redhat.com>

From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240626221408.2019633-1-alexander.deucher@amd.com

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

@@ -400,7 +400,7 @@ amdgpu_atomfirmware_get_vram_info(struct amdgpu_device *adev,
 					mem_channel_number = vram_info->v30.channel_num;
 					mem_channel_width = vram_info->v30.channel_width;
 					if (vram_width)
-						*vram_width = mem_channel_number * (1 << mem_channel_width);
+						*vram_width = mem_channel_number * 16;
 					break;
 				default:
 					return -EINVAL;

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

@@ -5220,11 +5220,14 @@ int amdgpu_device_mode1_reset(struct amdgpu_device *adev)
 
 	dev_info(adev->dev, "GPU mode1 reset\n");
 
+	/* Cache the state before bus master disable. The saved config space
+	 * values are used in other cases like restore after mode-2 reset.
+	 */
+	amdgpu_device_cache_pci_state(adev->pdev);
+
 	/* disable BM */
 	pci_clear_master(adev->pdev);
 
-	amdgpu_device_cache_pci_state(adev->pdev);
-
 	if (amdgpu_dpm_is_mode1_reset_supported(adev)) {
 		dev_info(adev->dev, "GPU smu mode1 reset\n");
 		ret = amdgpu_dpm_mode1_reset(adev);

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

@@ -640,6 +640,20 @@ static const char *psp_gfx_cmd_name(enum psp_gfx_cmd_id cmd_id)
 	}
 }
 
+static bool psp_err_warn(struct psp_context *psp)
+{
+	struct psp_gfx_cmd_resp *cmd = psp->cmd_buf_mem;
+
+	/* This response indicates reg list is already loaded */
+	if (amdgpu_ip_version(psp->adev, MP0_HWIP, 0) == IP_VERSION(13, 0, 2) &&
+	    cmd->cmd_id == GFX_CMD_ID_LOAD_IP_FW &&
+	    cmd->cmd.cmd_load_ip_fw.fw_type == GFX_FW_TYPE_REG_LIST &&
+	    cmd->resp.status == TEE_ERROR_CANCEL)
+		return false;
+
+	return true;
+}
+
 static int
 psp_cmd_submit_buf(struct psp_context *psp,
 		   struct amdgpu_firmware_info *ucode,
@@ -699,9 +713,12 @@ psp_cmd_submit_buf(struct psp_context *psp,
 			dev_warn(psp->adev->dev,
 				 "failed to load ucode %s(0x%X) ",
 				 amdgpu_ucode_name(ucode->ucode_id), ucode->ucode_id);
-		dev_warn(psp->adev->dev,
+		if (psp_err_warn(psp))
+			dev_warn(
+				psp->adev->dev,
 				"psp gfx command %s(0x%X) failed and response status is (0x%X)\n",
-			 psp_gfx_cmd_name(psp->cmd_buf_mem->cmd_id), psp->cmd_buf_mem->cmd_id,
+				psp_gfx_cmd_name(psp->cmd_buf_mem->cmd_id),
+				psp->cmd_buf_mem->cmd_id,
 				psp->cmd_buf_mem->resp.status);
 		/* If any firmware (including CAP) load fails under SRIOV, it should
 		 * return failure to stop the VF from initializing.

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

@@ -3,6 +3,7 @@
 #include <drm/drm_atomic_helper.h>
 #include <drm/drm_edid.h>
 #include <drm/drm_simple_kms_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
 #include <drm/drm_vblank.h>
 
 #include "amdgpu.h"
@@ -314,7 +315,13 @@ static int amdgpu_vkms_prepare_fb(struct drm_plane *plane,
 		return 0;
 	}
 	afb = to_amdgpu_framebuffer(new_state->fb);
-	obj = new_state->fb->obj[0];
+
+	obj = drm_gem_fb_get_obj(new_state->fb, 0);
+	if (!obj) {
+		DRM_ERROR("Failed to get obj from framebuffer\n");
+		return -EINVAL;
+	}
+
 	rbo = gem_to_amdgpu_bo(obj);
 	adev = amdgpu_ttm_adev(rbo->tbo.bdev);
 
@@ -368,12 +375,19 @@ static void amdgpu_vkms_cleanup_fb(struct drm_plane *plane,
 				   struct drm_plane_state *old_state)
 {
 	struct amdgpu_bo *rbo;
+	struct drm_gem_object *obj;
 	int r;
 
 	if (!old_state->fb)
 		return;
 
-	rbo = gem_to_amdgpu_bo(old_state->fb->obj[0]);
+	obj = drm_gem_fb_get_obj(old_state->fb, 0);
+	if (!obj) {
+		DRM_ERROR("Failed to get obj from framebuffer\n");
+		return;
+	}
+
+	rbo = gem_to_amdgpu_bo(obj);
 	r = amdgpu_bo_reserve(rbo, false);
 	if (unlikely(r)) {
 		DRM_ERROR("failed to reserve rbo before unpin\n");

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

@@ -465,6 +465,7 @@ struct psp_gfx_rb_frame
 
 enum tee_error_code {
 	TEE_SUCCESS			= 0x00000000,
+	TEE_ERROR_CANCEL		= 0xFFFF0002,
 	TEE_ERROR_NOT_SUPPORTED		= 0xFFFF000A,
 };
 

drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c

@@ -1590,9 +1590,17 @@ static bool retrieve_link_cap(struct dc_link *link)
 			return false;
 	}
 
-	if (dp_is_lttpr_present(link))
+	if (dp_is_lttpr_present(link)) {
 		configure_lttpr_mode_transparent(link);
 
+		// Echo TOTAL_LTTPR_CNT back downstream
+		core_link_write_dpcd(
+				link,
+				DP_TOTAL_LTTPR_CNT,
+				&link->dpcd_caps.lttpr_caps.phy_repeater_cnt,
+				sizeof(link->dpcd_caps.lttpr_caps.phy_repeater_cnt));
+	}
+
 	/* Read DP tunneling information. */
 	status = dpcd_get_tunneling_device_data(link);
 

drivers/gpu/drm/amd/display/include/dpcd_defs.h

@@ -177,4 +177,9 @@ enum dpcd_psr_sink_states {
 #define DP_SINK_PR_PIXEL_DEVIATION_PER_LINE     0x379
 #define DP_SINK_PR_MAX_NUMBER_OF_DEVIATION_LINE 0x37A
 
+/* Remove once drm_dp_helper.h is updated upstream */
+#ifndef DP_TOTAL_LTTPR_CNT
+#define DP_TOTAL_LTTPR_CNT                                  0xF000A /* 2.1 */
+#endif
+
 #endif /* __DAL_DPCD_DEFS_H__ */

drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c

@@ -324,6 +324,18 @@ static int smu_dpm_set_umsch_mm_enable(struct smu_context *smu,
 	return ret;
 }
 
+static int smu_set_mall_enable(struct smu_context *smu)
+{
+	int ret = 0;
+
+	if (!smu->ppt_funcs->set_mall_enable)
+		return 0;
+
+	ret = smu->ppt_funcs->set_mall_enable(smu);
+
+	return ret;
+}
+
 /**
  * smu_dpm_set_power_gate - power gate/ungate the specific IP block
  *
@@ -1791,6 +1803,7 @@ static int smu_hw_init(void *handle)
 		smu_dpm_set_jpeg_enable(smu, true);
 		smu_dpm_set_vpe_enable(smu, true);
 		smu_dpm_set_umsch_mm_enable(smu, true);
+		smu_set_mall_enable(smu);
 		smu_set_gfx_cgpg(smu, true);
 	}
 

drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h

@@ -1394,6 +1394,11 @@ struct pptable_funcs {
 	 */
 	int (*dpm_set_umsch_mm_enable)(struct smu_context *smu, bool enable);
 
+	/**
+	 * @set_mall_enable: Init MALL power gating control.
+	 */
+	int (*set_mall_enable)(struct smu_context *smu);
+
 	/**
 	 * @notify_rlc_state: Notify RLC power state to SMU.
 	 */

drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu_v14_0_0_ppsmc.h

@@ -106,8 +106,8 @@
 #define PPSMC_MSG_DisableLSdma                  0x35 ///< Disable LSDMA
 #define PPSMC_MSG_SetSoftMaxVpe                 0x36 ///<
 #define PPSMC_MSG_SetSoftMinVpe                 0x37 ///<
-#define PPSMC_MSG_AllocMALLCache                0x38 ///< Allocating MALL Cache
-#define PPSMC_MSG_ReleaseMALLCache              0x39 ///< Releasing MALL Cache
+#define PPSMC_MSG_MALLPowerController           0x38 ///< Set MALL control
+#define PPSMC_MSG_MALLPowerState                0x39 ///< Enter/Exit MALL PG
 #define PPSMC_Message_Count                     0x3A ///< Total number of PPSMC messages
 /** @}*/
 

drivers/gpu/drm/amd/pm/swsmu/inc/smu_types.h

@@ -272,7 +272,9 @@
 	__SMU_DUMMY_MAP(SetSoftMinVpe), \
 	__SMU_DUMMY_MAP(GetMetricsVersion), \
 	__SMU_DUMMY_MAP(EnableUCLKShadow), \
-	__SMU_DUMMY_MAP(RmaDueToBadPageThreshold),
+	__SMU_DUMMY_MAP(RmaDueToBadPageThreshold), \
+	__SMU_DUMMY_MAP(MALLPowerController), \
+	__SMU_DUMMY_MAP(MALLPowerState),
 
 #undef __SMU_DUMMY_MAP
 #define __SMU_DUMMY_MAP(type)	SMU_MSG_##type

drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_0_ppt.c

@@ -52,6 +52,19 @@
 #define mmMP1_SMN_C2PMSG_90			0x029a
 #define mmMP1_SMN_C2PMSG_90_BASE_IDX		    0
 
+/* MALLPowerController message arguments (Defines for the Cache mode control) */
+#define SMU_MALL_PMFW_CONTROL 0
+#define SMU_MALL_DRIVER_CONTROL 1
+
+/*
+ * MALLPowerState message arguments
+ * (Defines for the Allocate/Release Cache mode if in driver mode)
+ */
+#define SMU_MALL_EXIT_PG 0
+#define SMU_MALL_ENTER_PG 1
+
+#define SMU_MALL_PG_CONFIG_DEFAULT SMU_MALL_PG_CONFIG_DRIVER_CONTROL_ALWAYS_ON
+
 #define FEATURE_MASK(feature) (1ULL << feature)
 #define SMC_DPM_FEATURE ( \
 	FEATURE_MASK(FEATURE_CCLK_DPM_BIT) | \
@@ -66,6 +79,12 @@
 	FEATURE_MASK(FEATURE_GFX_DPM_BIT)	| \
 	FEATURE_MASK(FEATURE_VPE_DPM_BIT))
 
+enum smu_mall_pg_config {
+	SMU_MALL_PG_CONFIG_PMFW_CONTROL = 0,
+	SMU_MALL_PG_CONFIG_DRIVER_CONTROL_ALWAYS_ON = 1,
+	SMU_MALL_PG_CONFIG_DRIVER_CONTROL_ALWAYS_OFF = 2,
+};
+
 static struct cmn2asic_msg_mapping smu_v14_0_0_message_map[SMU_MSG_MAX_COUNT] = {
 	MSG_MAP(TestMessage,                    PPSMC_MSG_TestMessage,				1),
 	MSG_MAP(GetSmuVersion,                  PPSMC_MSG_GetPmfwVersion,			1),
@@ -113,6 +132,8 @@ static struct cmn2asic_msg_mapping smu_v14_0_0_message_map[SMU_MSG_MAX_COUNT] =
 	MSG_MAP(PowerDownUmsch,                 PPSMC_MSG_PowerDownUmsch,			1),
 	MSG_MAP(SetSoftMaxVpe,                  PPSMC_MSG_SetSoftMaxVpe,			1),
 	MSG_MAP(SetSoftMinVpe,                  PPSMC_MSG_SetSoftMinVpe,			1),
+	MSG_MAP(MALLPowerController,            PPSMC_MSG_MALLPowerController,		1),
+	MSG_MAP(MALLPowerState,                 PPSMC_MSG_MALLPowerState,			1),
 };
 
 static struct cmn2asic_mapping smu_v14_0_0_feature_mask_map[SMU_FEATURE_COUNT] = {
@@ -1423,6 +1444,57 @@ static int smu_v14_0_common_get_dpm_table(struct smu_context *smu, struct dpm_cl
 	return 0;
 }
 
+static int smu_v14_0_1_init_mall_power_gating(struct smu_context *smu, enum smu_mall_pg_config pg_config)
+{
+	struct amdgpu_device *adev = smu->adev;
+	int ret = 0;
+
+	if (pg_config == SMU_MALL_PG_CONFIG_PMFW_CONTROL) {
+		ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_MALLPowerController,
+								SMU_MALL_PMFW_CONTROL, NULL);
+		if (ret) {
+			dev_err(adev->dev, "Init MALL PMFW CONTROL Failure\n");
+			return ret;
+		}
+	} else {
+		ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_MALLPowerController,
+								SMU_MALL_DRIVER_CONTROL, NULL);
+		if (ret) {
+			dev_err(adev->dev, "Init MALL Driver CONTROL Failure\n");
+			return ret;
+		}
+
+		if (pg_config == SMU_MALL_PG_CONFIG_DRIVER_CONTROL_ALWAYS_ON) {
+			ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_MALLPowerState,
+									SMU_MALL_EXIT_PG, NULL);
+			if (ret) {
+				dev_err(adev->dev, "EXIT MALL PG Failure\n");
+				return ret;
+			}
+		} else if (pg_config == SMU_MALL_PG_CONFIG_DRIVER_CONTROL_ALWAYS_OFF) {
+			ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_MALLPowerState,
+									SMU_MALL_ENTER_PG, NULL);
+			if (ret) {
+				dev_err(adev->dev, "Enter MALL PG Failure\n");
+				return ret;
+			}
+		}
+	}
+
+	return ret;
+}
+
+static int smu_v14_0_common_set_mall_enable(struct smu_context *smu)
+{
+	enum smu_mall_pg_config pg_config = SMU_MALL_PG_CONFIG_DEFAULT;
+	int ret = 0;
+
+	if (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 1))
+		ret = smu_v14_0_1_init_mall_power_gating(smu, pg_config);
+
+	return ret;
+}
+
 static const struct pptable_funcs smu_v14_0_0_ppt_funcs = {
 	.check_fw_status = smu_v14_0_check_fw_status,
 	.check_fw_version = smu_v14_0_check_fw_version,
@@ -1454,6 +1526,7 @@ static const struct pptable_funcs smu_v14_0_0_ppt_funcs = {
 	.dpm_set_vpe_enable = smu_v14_0_0_set_vpe_enable,
 	.dpm_set_umsch_mm_enable = smu_v14_0_0_set_umsch_mm_enable,
 	.get_dpm_clock_table = smu_v14_0_common_get_dpm_table,
+	.set_mall_enable = smu_v14_0_common_set_mall_enable,
 };
 
 static void smu_v14_0_0_set_smu_mailbox_registers(struct smu_context *smu)