Merge branch 'main' into fp16_basic_int2fp

CMakeLists.txt

@@ -105,7 +105,6 @@ if(CMAKE_COMPILER_IS_GNUCC OR "${CMAKE_CXX_COMPILER_ID}" MATCHES "(Apple)?Clang"
         add_cxx_flag_if_supported(-Wall)
         # Suppress warnings that currently trigger on the code base.
         # This list should shrink over time when warnings are fixed.
-        add_cxx_flag_if_supported(-Wno-sometimes-uninitialized)
         add_cxx_flag_if_supported(-Wno-sign-compare)
     endif()
     add_cxx_flag_if_supported(-Wno-narrowing)

presubmit.sh

@@ -77,7 +77,6 @@ cmake .. -G Ninja \
       -DBUILD_WSI_XLIB_SUPPORT=OFF \
       -DBUILD_WSI_XCB_SUPPORT=OFF \
       -DBUILD_WSI_WAYLAND_SUPPORT=OFF \
-      -DUSE_GAS=OFF \
       -C helper.cmake ..
 cmake --build . -j2
 

test_conformance/SVM/CMakeLists.txt

@@ -17,4 +17,6 @@ set(${MODULE_NAME}_SOURCES
     test_migrate.cpp
 )
 
+set_gnulike_module_compile_flags("-Wno-sometimes-uninitialized")
+
 include(../CMakeCommon.txt)

test_conformance/basic/test_async_copy.cpp

@@ -20,8 +20,7 @@
 #include <string.h>
 #include <sys/types.h>
 #include <sys/stat.h>
-
-
+#include <vector>
 
 #include "procs.h"
 #include "harness/conversions.h"
@@ -86,8 +85,7 @@ int test_copy(cl_device_id deviceID, cl_context context, cl_command_queue queue,
     clKernelWrapper kernel;
     clMemWrapper streams[ 2 ];
     size_t threads[ 1 ], localThreads[ 1 ];
-    void *inBuffer, *outBuffer;
-    MTdata d;
+    MTdataHolder d(gRandomSeed);
     char vecNameString[64]; vecNameString[0] = 0;
     if (vecSize == 1)
         sprintf(vecNameString, "%s", get_explicit_type_name(vecType));
@@ -109,9 +107,15 @@ int test_copy(cl_device_id deviceID, cl_context context, cl_command_queue queue,
     char programSource[4096]; programSource[0]=0;
     char *programPtr;
 
-    sprintf(programSource, kernelCode,
-            vecType == kDouble ? "#pragma OPENCL EXTENSION cl_khr_fp64 : enable" : "",
-            vecNameString, vecNameString, vecNameString, vecNameString, get_explicit_type_name(vecType), vecNameString, vecNameString);
+    std::string extStr = "";
+    if (vecType == kDouble)
+        extStr = "#pragma OPENCL EXTENSION cl_khr_fp64 : enable";
+    else if (vecType == kHalf)
+        extStr = "#pragma OPENCL EXTENSION cl_khr_fp16 : enable";
+
+    sprintf(programSource, kernelCode, extStr.c_str(), vecNameString,
+            vecNameString, vecNameString, vecNameString,
+            get_explicit_type_name(vecType), vecNameString, vecNameString);
     //log_info("program: %s\n", programSource);
     programPtr = programSource;
 
@@ -150,9 +154,10 @@ int test_copy(cl_device_id deviceID, cl_context context, cl_command_queue queue,
     size_t globalBufferSize = numberOfLocalWorkgroups*localBufferSize;
     size_t globalWorkgroupSize = numberOfLocalWorkgroups*localWorkgroupSize;
 
-    inBuffer = (void*)malloc(globalBufferSize);
-    outBuffer = (void*)malloc(globalBufferSize);
-    memset(outBuffer, 0, globalBufferSize);
+    std::vector<unsigned char> inBuffer(globalBufferSize);
+    std::vector<unsigned char> outBuffer(globalBufferSize);
+
+    outBuffer.assign(globalBufferSize, 0);
 
     cl_int copiesPerWorkItemInt, copiesPerWorkgroup;
     copiesPerWorkItemInt = (int)numberOfCopiesPerWorkitem;
@@ -164,13 +169,15 @@ int test_copy(cl_device_id deviceID, cl_context context, cl_command_queue queue,
     threads[0] = globalWorkgroupSize;
     localThreads[0] = localWorkgroupSize;
 
-    d = init_genrand( gRandomSeed );
-    generate_random_data( vecType, globalBufferSize/get_explicit_type_size(vecType), d, inBuffer );
-    free_mtdata(d); d = NULL;
+    generate_random_data(vecType,
+                         globalBufferSize / get_explicit_type_size(vecType), d,
+                         &inBuffer.front());
 
-    streams[ 0 ] = clCreateBuffer( context, CL_MEM_COPY_HOST_PTR, globalBufferSize, inBuffer, &error );
+    streams[0] = clCreateBuffer(context, CL_MEM_COPY_HOST_PTR, globalBufferSize,
+                                &inBuffer.front(), &error);
     test_error( error, "Unable to create input buffer" );
-    streams[ 1 ] = clCreateBuffer( context, CL_MEM_COPY_HOST_PTR, globalBufferSize, outBuffer, &error );
+    streams[1] = clCreateBuffer(context, CL_MEM_COPY_HOST_PTR, globalBufferSize,
+                                &outBuffer.front(), &error);
     test_error( error, "Unable to create output buffer" );
 
     error = clSetKernelArg( kernel, 0, sizeof( streams[ 0 ] ), &streams[ 0 ] );
@@ -189,16 +196,18 @@ int test_copy(cl_device_id deviceID, cl_context context, cl_command_queue queue,
     test_error( error, "Unable to queue kernel" );
 
     // Read
-    error = clEnqueueReadBuffer( queue, streams[ 1 ], CL_TRUE, 0, globalBufferSize, outBuffer, 0, NULL, NULL );
+    error = clEnqueueReadBuffer(queue, streams[1], CL_TRUE, 0, globalBufferSize,
+                                &outBuffer.front(), 0, NULL, NULL);
     test_error( error, "Unable to read results" );
 
     // Verify
     int failuresPrinted = 0;
-    if( memcmp( inBuffer, outBuffer, globalBufferSize ) != 0 )
+    if (memcmp(&inBuffer.front(), &outBuffer.front(), globalBufferSize) != 0)
     {
         size_t typeSize = get_explicit_type_size(vecType)* vecSize;
-        unsigned char * inchar = (unsigned char*)inBuffer;
-        unsigned char * outchar = (unsigned char*)outBuffer;
+        unsigned char *inchar = static_cast<unsigned char *>(&inBuffer.front());
+        unsigned char *outchar =
+            static_cast<unsigned char *>(&outBuffer.front());
         for (int i=0; i< (int)globalBufferSize; i+=(int)elementSize) {
             if (memcmp( ((char *)inchar)+i, ((char *)outchar)+i, typeSize) != 0 )
             {
@@ -226,26 +235,29 @@ int test_copy(cl_device_id deviceID, cl_context context, cl_command_queue queue,
         }
     }
 
-    free(inBuffer);
-    free(outBuffer);
-
     return failuresPrinted ? -1 : 0;
 }
 
 int test_copy_all_types(cl_device_id deviceID, cl_context context, cl_command_queue queue, const char *kernelCode) {
-    ExplicitType vecType[] = { kChar, kUChar, kShort, kUShort, kInt, kUInt, kLong, kULong, kFloat, kDouble, kNumExplicitTypes };
+    const std::vector<ExplicitType> vecType = { kChar,  kUChar, kShort, kUShort,
+                                                kInt,   kUInt,  kLong,  kULong,
+                                                kFloat, kHalf,  kDouble };
     unsigned int vecSizes[] = { 1, 2, 3, 4, 8, 16, 0 };
     unsigned int size, typeIndex;
 
     int errors = 0;
 
-    for( typeIndex = 0; vecType[ typeIndex ] != kNumExplicitTypes; typeIndex++ )
-    {
-        if( vecType[ typeIndex ] == kDouble && !is_extension_available( deviceID, "cl_khr_fp64" ) )
-            continue;
+    bool fp16Support = is_extension_available(deviceID, "cl_khr_fp16");
+    bool fp64Support = is_extension_available(deviceID, "cl_khr_fp64");
 
+    for (typeIndex = 0; typeIndex < vecType.size(); typeIndex++)
+    {
         if (( vecType[ typeIndex ] == kLong || vecType[ typeIndex ] == kULong ) && !gHasLong )
             continue;
+        else if (vecType[typeIndex] == kDouble && !fp64Support)
+            continue;
+        else if (vecType[typeIndex] == kHalf && !fp16Support)
+            continue;
 
         for( size = 0; vecSizes[ size ] != 0; size++ )
         {
@@ -259,9 +271,6 @@ int test_copy_all_types(cl_device_id deviceID, cl_context context, cl_command_qu
     return 0;
 }
 
-
-
-
 int test_async_copy_global_to_local(cl_device_id deviceID, cl_context context, cl_command_queue queue, int num_elements)
 {
     return test_copy_all_types( deviceID, context, queue, async_global_to_local_kernel );

test_conformance/basic/test_async_strided_copy.cpp

@@ -211,9 +211,9 @@ int test_strided_copy(cl_device_id deviceID, cl_context context, cl_command_queu
         if (memcmp(&inBuffer.at(i), &outBuffer.at(i), typeSize) != 0)
         {
             unsigned char *inchar =
-                static_cast<unsigned char *>(inBuffer.data());
+                static_cast<unsigned char *>(&inBuffer.at(i));
             unsigned char *outchar =
-                static_cast<unsigned char *>(outBuffer.data());
+                static_cast<unsigned char *>(&outBuffer.at(i));
             char values[4096];
             values[0] = 0;
 

test_conformance/mem_host_flags/checker.h

@@ -219,7 +219,7 @@ cl_int cBuffer_checker<T>::SetupASSubBuffer(cl_mem_flags parent_buffer_flag)
         err = CL_SUCCESS;
     }
 
-    cl_mem_flags f;
+    cl_mem_flags f = 0;
     if (parent_buffer_flag & CL_MEM_HOST_READ_ONLY)
         f = CL_MEM_HOST_READ_ONLY;
     else if (parent_buffer_flag & CL_MEM_HOST_WRITE_ONLY)

test_conformance/non_uniform_work_group/CMakeLists.txt

@@ -10,8 +10,6 @@ set(${MODULE_NAME}_SOURCES
     tools.cpp
 )
 
-set_gnulike_module_compile_flags("-Wno-unused-but-set-variable")
-
 include(../CMakeCommon.txt)
 
 # end of file #

test_conformance/non_uniform_work_group/TestNonUniformWorkGroup.cpp

@@ -448,13 +448,8 @@ void TestNonUniformWorkGroup::verifyData (DataContainerAttrib * reference, DataC
 }
 
 void TestNonUniformWorkGroup::calculateExpectedValues () {
-  size_t nonRemainderGlobalSize[MAX_DIMS];
   size_t numberOfPossibleRegions[MAX_DIMS];
 
-  nonRemainderGlobalSize[0] = _globalSize[0] - (_globalSize[0] % _enqueuedLocalSize[0]);
-  nonRemainderGlobalSize[1] = _globalSize[1] - (_globalSize[1] % _enqueuedLocalSize[1]);
-  nonRemainderGlobalSize[2] = _globalSize[2] - (_globalSize[2] % _enqueuedLocalSize[2]);
-
   numberOfPossibleRegions[0] = (_globalSize[0]>1)?2:1;
   numberOfPossibleRegions[1] = (_globalSize[1]>1)?2:1;
   numberOfPossibleRegions[2] = (_globalSize[2]>1)?2:1;
@@ -502,6 +497,11 @@ size_t TestNonUniformWorkGroup::getMaxLocalWorkgroupSize (const cl_device_id &de
   if (TestNonUniformWorkGroup::_maxLocalWorkgroupSize == 0) {
     err = clGetDeviceInfo(device, CL_DEVICE_MAX_WORK_GROUP_SIZE,
       sizeof(TestNonUniformWorkGroup::_maxLocalWorkgroupSize), &TestNonUniformWorkGroup::_maxLocalWorkgroupSize, NULL);
+    if (err)
+    {
+        log_error("clGetDeviceInfo failed\n");
+        return 0;
+    }
   }
 
   return TestNonUniformWorkGroup::_maxLocalWorkgroupSize;

test_conformance/relationals/test_comparisons_fp.cpp

@@ -14,6 +14,8 @@
 // limitations under the License.
 //
 
+#include <cstdint>
+#include <functional>
 #include <iostream>
 #include <map>
 #include <memory>