Merge pull request #1908 from asalmanp/prop_mulit_coop

[HIP] add hip specific properties for cooperative kernel multi device

hipamd/include/hip/hip_runtime_api.h

@@ -128,6 +128,14 @@ typedef struct hipDeviceProp_t {
     int kernelExecTimeoutEnabled;    ///<Run time limit for kernels executed on the device
     int ECCEnabled;                  ///<Device has ECC support enabled
     int tccDriver;                   ///< 1:If device is Tesla device using TCC driver, else 0
+    int cooperativeMultiDeviceUnmatchedFunc;        ///< HIP device supports cooperative launch on multiple
+                                                    ///devices with unmatched functions
+    int cooperativeMultiDeviceUnmatchedGridDim;     ///< HIP device supports cooperative launch on multiple
+                                                    ///devices with unmatched grid dimensions
+    int cooperativeMultiDeviceUnmatchedBlockDim;    ///< HIP device supports cooperative launch on multiple
+                                                    ///devices with unmatched block dimensions
+    int cooperativeMultiDeviceUnmatchedSharedMem;   ///< HIP device supports cooperative launch on multiple
+                                                    ///devices with unmatched shared memories
 
 } hipDeviceProp_t;
 
@@ -313,7 +321,6 @@ typedef enum hipDeviceAttribute_t {
     hipDeviceAttributeIntegrated,                        ///< iGPU
     hipDeviceAttributeCooperativeLaunch,                 ///< Support cooperative launch
     hipDeviceAttributeCooperativeMultiDeviceLaunch,      ///< Support cooperative launch on multiple devices
-
     hipDeviceAttributeMaxTexture1DWidth,    ///< Maximum number of elements in 1D images
     hipDeviceAttributeMaxTexture2DWidth,    ///< Maximum dimension width of 2D images in image elements
     hipDeviceAttributeMaxTexture2DHeight,   ///< Maximum dimension height of 2D images in image elements
@@ -329,7 +336,16 @@ typedef enum hipDeviceAttribute_t {
     hipDeviceAttributeTexturePitchAlignment, ///<Pitch alignment requirement for 2D texture references bound to pitched memory;
     hipDeviceAttributeKernelExecTimeout,    ///<Run time limit for kernels executed on the device
     hipDeviceAttributeCanMapHostMemory,     ///<Device can map host memory into device address space
-    hipDeviceAttributeEccEnabled            ///<Device has ECC support enabled
+    hipDeviceAttributeEccEnabled,            ///<Device has ECC support enabled
+
+    hipDeviceAttributeCooperativeMultiDeviceUnmatchedFunc,        ///< Supports cooperative launch on multiple
+                                                                  ///devices with unmatched functions
+    hipDeviceAttributeCooperativeMultiDeviceUnmatchedGridDim,     ///< Supports cooperative launch on multiple
+                                                                  ///devices with unmatched grid dimensions
+    hipDeviceAttributeCooperativeMultiDeviceUnmatchedBlockDim,    ///< Supports cooperative launch on multiple
+                                                                  ///devices with unmatched block dimensions
+    hipDeviceAttributeCooperativeMultiDeviceUnmatchedSharedMem,   ///< Supports cooperative launch on multiple
+                                                                  ///devices with unmatched shared memories
 
 } hipDeviceAttribute_t;
 

hipamd/include/hip/nvcc_detail/hip_runtime_api.h

@@ -1134,6 +1134,10 @@ inline static hipError_t hipGetDeviceProperties(hipDeviceProp_t* p_prop, int dev
     p_prop->integrated = cdprop.integrated;
     p_prop->cooperativeLaunch = cdprop.cooperativeLaunch;
     p_prop->cooperativeMultiDeviceLaunch = cdprop.cooperativeMultiDeviceLaunch;
+    p_prop->cooperativeMultiDeviceUnmatchedFunc = 0;
+    p_prop->cooperativeMultiDeviceUnmatchedGridDim = 0;
+    p_prop->cooperativeMultiDeviceUnmatchedBlockDim = 0;
+    p_prop->cooperativeMultiDeviceUnmatchedSharedMem = 0;
 
     p_prop->maxTexture1D    = cdprop.maxTexture1D;
     p_prop->maxTexture2D[0] = cdprop.maxTexture2D[0];

hipamd/src/hip_device.cpp

@@ -310,6 +310,18 @@ hipError_t ihipDeviceGetAttribute(int* pi, hipDeviceAttribute_t attr, int device
             case hipDeviceAttributeCooperativeMultiDeviceLaunch:
                 *pi = prop->cooperativeMultiDeviceLaunch;
                 break;
+            case hipDeviceAttributeCooperativeMultiDeviceUnmatchedFunc:
+                *pi = prop->cooperativeMultiDeviceUnmatchedFunc;
+                break;
+            case hipDeviceAttributeCooperativeMultiDeviceUnmatchedGridDim:
+                *pi = prop->cooperativeMultiDeviceUnmatchedGridDim;
+                break;
+            case hipDeviceAttributeCooperativeMultiDeviceUnmatchedBlockDim:
+                *pi = prop->cooperativeMultiDeviceUnmatchedBlockDim;
+                break;
+            case hipDeviceAttributeCooperativeMultiDeviceUnmatchedSharedMem:
+                *pi = prop->cooperativeMultiDeviceUnmatchedSharedMem;
+                break;
             case hipDeviceAttributeMaxPitch:
                 *pi = prop->memPitch;
                 break;

hipamd/src/hip_hcc.cpp

@@ -901,6 +901,11 @@ hipError_t ihipDevice_t::initProperties(hipDeviceProp_t* prop) {
     prop->cooperativeLaunch = (prop->gcnArch < 900) ? 0 : 1;
     prop->cooperativeMultiDeviceLaunch = (prop->gcnArch < 900) ? 0 : 1;
 
+    prop->cooperativeMultiDeviceUnmatchedFunc = prop->cooperativeMultiDeviceLaunch;
+    prop->cooperativeMultiDeviceUnmatchedGridDim = prop->cooperativeMultiDeviceLaunch;
+    prop->cooperativeMultiDeviceUnmatchedBlockDim = prop->cooperativeMultiDeviceLaunch;
+    prop->cooperativeMultiDeviceUnmatchedSharedMem = prop->cooperativeMultiDeviceLaunch;
+
     err = hsa_agent_get_info(_hsaAgent, (hsa_agent_info_t)HSA_EXT_AGENT_INFO_IMAGE_1D_MAX_ELEMENTS,
           &prop->maxTexture1D);
     DeviceErrorCheck(err);