drm/amd/pm: Get xgmi link status for XGMI_v_6_4_0

Get XGMI_v_6_4_0 link status and populate it to metrics v1_7 for
SMU_v_13_0_6

v2: Get link status register value for each soc from separate
function (Lijo)

Signed-off-by: Asad Kamal <asad.kamal@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

3 files changed, 46 insertions, 1 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
index b47422b0b5b1..74b4349e345a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
@@ -40,6 +40,11 @@
 #define smnPCS_GOPX1_PCS_ERROR_STATUS    0x12200210
 #define smnPCS_GOPX1_PCS_ERROR_NONCORRECTABLE_MASK      0x12200218
 
+#define XGMI_STATE_DISABLE                      0xD1
+#define XGMI_STATE_LS0                          0x81
+#define XGMI_LINK_ACTIVE			1
+#define XGMI_LINK_INACTIVE			0
+
 static DEFINE_MUTEX(xgmi_mutex);
 
 #define AMDGPU_MAX_XGMI_DEVICE_PER_HIVE		4
@@ -289,6 +294,42 @@ static const struct amdgpu_pcs_ras_field xgmi3x16_pcs_ras_fields[] = {
 	 SOC15_REG_FIELD(PCS_XGMI3X16_PCS_ERROR_STATUS, RxCMDPktErr)},
 };
 
+static u32 xgmi_v6_4_get_link_status(struct amdgpu_device *adev, int global_link_num)
+{
+	const u32 smnpcs_xgmi3x16_pcs_state_hist1 = 0x11a00070;
+	const int xgmi_inst = 2;
+	u32 link_inst;
+	u64 addr;
+
+	link_inst = global_link_num % xgmi_inst;
+
+	addr = (smnpcs_xgmi3x16_pcs_state_hist1 | (link_inst << 20)) +
+		adev->asic_funcs->encode_ext_smn_addressing(global_link_num / xgmi_inst);
+
+	return RREG32_PCIE_EXT(addr);
+}
+
+int amdgpu_get_xgmi_link_status(struct amdgpu_device *adev, int global_link_num)
+{
+	u32 xgmi_state_reg_val;
+
+	switch (amdgpu_ip_version(adev, XGMI_HWIP, 0)) {
+	case IP_VERSION(6, 4, 0):
+		xgmi_state_reg_val = xgmi_v6_4_get_link_status(adev, global_link_num);
+		break;
+	default:
+		return -EOPNOTSUPP;
+	}
+
+	if ((xgmi_state_reg_val & 0xFF) == XGMI_STATE_DISABLE)
+		return -ENOLINK;
+
+	if ((xgmi_state_reg_val & 0xFF) == XGMI_STATE_LS0)
+		return XGMI_LINK_ACTIVE;
+
+	return XGMI_LINK_INACTIVE;
+}
+
 /**
  * DOC: AMDGPU XGMI Support
  *
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h
index 8cc7ab38db7c..d1282b4c6348 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h
@@ -84,5 +84,7 @@ int amdgpu_xgmi_reset_on_init(struct amdgpu_device *adev);
 int amdgpu_xgmi_request_nps_change(struct amdgpu_device *adev,
 				   struct amdgpu_hive_info *hive,
 				   int req_nps_mode);
+int amdgpu_get_xgmi_link_status(struct amdgpu_device *adev,
+				int global_link_num);
 
 #endif
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c
index 11ecaa62f419..ab3c93ddce46 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c
@@ -96,7 +96,6 @@ MODULE_FIRMWARE("amdgpu/smu_13_0_14.bin");
 #define PCIE_LC_SPEED_CNTL__LC_CURRENT_DATA_RATE_MASK 0xE0
 #define PCIE_LC_SPEED_CNTL__LC_CURRENT_DATA_RATE__SHIFT 0x5
 #define LINK_SPEED_MAX				4
-
 #define SMU_13_0_6_DSCLK_THRESHOLD 140
 
 #define MCA_BANK_IPID(_ip, _hwid, _type) \
@@ -2448,6 +2447,9 @@ static ssize_t smu_v13_0_6_get_gpu_metrics(struct smu_context *smu, void **table
 			SMUQ10_ROUND(GET_METRIC_FIELD(XgmiReadDataSizeAcc, flag)[i]);
 		gpu_metrics->xgmi_write_data_acc[i] =
 			SMUQ10_ROUND(GET_METRIC_FIELD(XgmiWriteDataSizeAcc, flag)[i]);
+		ret = amdgpu_get_xgmi_link_status(adev, i);
+		if (ret >= 0)
+			gpu_metrics->xgmi_link_status[i] = ret;
 	}
 
 	gpu_metrics->num_partition = adev->xcp_mgr->num_xcps;