Add the RdcSmiDiagnostic module

Provides a RdcSmiDiagnostic module, which will call rocm_smi_lib.

It will support following diagnostics: Get GPU Topology, Check GPU
parameters and check processes running on the GPUs.

The grpc client and server side diagnostics function is added.

The diag module is added to the rdci.

Change-Id: I10a0cf3c20556a61373ab686f82cae75acaa40dd

protos/rdc.proto

@@ -214,6 +214,18 @@ service RdcAPI {
 
   // rdc_status_t rdc_job_remove_all()
   rpc RemoveAllJob(Empty) returns (RemoveAllJobResponse) {}
+
+  // rdc_status_t rdc_diagnostic_run(
+  //  rdc_gpu_group_t group_id,
+  //  rdc_diag_level_t level,
+  //  rdc_diag_response_t* response);
+  rpc DiagnosticRun(DiagnosticRunRequest) returns (DiagnosticRunResponse) {}
+
+  // rdc_status_t rdc_test_case_run(
+  //   rdc_gpu_group_t group_id,
+  //    rdc_diag_test_cases_t test_case,
+  //    rdc_diag_test_result_t* result);
+  rpc DiagnosticTestCaseRun(DiagnosticTestCaseRunRequest) returns (DiagnosticTestCaseRunResponse) {}
 }
 
 message Empty {
@@ -459,4 +471,66 @@ message RemoveJobResponse {
 
 message RemoveAllJobResponse {
   uint32 status = 1;
+}
+
+message DiagnosticRunRequest {
+  uint32 group_id = 1;
+  uint32 level = 2;
+}
+
+message DiagnosticDetail {
+  string msg = 1;
+  uint32 code = 2;
+}
+
+message DiagnosticPerGpuResult {
+  uint32 gpu_index = 1;
+  DiagnosticDetail gpu_result = 2;
+}
+
+message DiagnosticTestResult {
+  uint32 status = 1;
+  DiagnosticDetail details = 2;
+  enum DiagnosticTestCase {
+    COMPUTE_PROCESS = 0;
+    SDMA_QUEUE = 1;
+    COMPUTE_QUEUE = 2;
+    VRAM_CHECK = 3;
+    SYS_MEM_CHECK = 4;
+    NODE_TOPOLOGY = 5;
+    GPU_PARAMETERS = 6;
+  };
+  DiagnosticTestCase test_case = 3;
+  uint32 per_gpu_result_count = 4;
+  repeated DiagnosticPerGpuResult gpu_results = 5;
+  string info = 6;
+}
+
+message DiagnosticResponse {
+  uint32 results_count = 1;
+  repeated DiagnosticTestResult diag_info = 2;
+}
+
+message DiagnosticRunResponse {
+  uint32 status = 1;
+  DiagnosticResponse response = 2;
+}
+
+message DiagnosticTestCaseRunRequest {
+    uint32 group_id = 1;
+    enum TestCaseType {
+      COMPUTE_PROCESS = 0;
+      SDMA_QUEUE = 1;
+      COMPUTE_QUEUE = 2;
+      VRAM_CHECK = 3;
+      SYS_MEM_CHECK = 4;
+      NODE_TOPOLOGY = 5;
+      GPU_PARAMETERS = 6;
+    };
+    TestCaseType test_case = 2;
+}
+
+message DiagnosticTestCaseRunResponse {
+  uint32 status = 1;
+  DiagnosticTestResult result = 2;
 }