| // |
| // Copyright (c) 2017 The Khronos Group Inc. |
| // |
| // Licensed under the Apache License, Version 2.0 (the "License"); |
| // you may not use this file except in compliance with the License. |
| // You may obtain a copy of the License at |
| // |
| // http://www.apache.org/licenses/LICENSE-2.0 |
| // |
| // Unless required by applicable law or agreed to in writing, software |
| // distributed under the License is distributed on an "AS IS" BASIS, |
| // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| // See the License for the specific language governing permissions and |
| // limitations under the License. |
| // |
| #include "harness/compat.h" |
| |
| #include <stdio.h> |
| #include <string.h> |
| #include <assert.h> |
| #include <sys/types.h> |
| #include <sys/stat.h> |
| |
| #include "procs.h" |
| #include "harness/errorHelpers.h" |
| |
| #define STRING_LENGTH 1024 |
| |
| void createKernelSourceCode(char *source, int num_pipes) |
| { |
| int i; |
| char str[256]; |
| int str_length; |
| |
| strcpy(source, "__kernel void test_multiple_pipe_write(__global int *src, "); |
| |
| for(i = 0; i < num_pipes; i++) { |
| sprintf(str, "__write_only pipe int pipe%d, ", i); |
| strcat(source, str); |
| } |
| sprintf(str, "int num_pipes ) \n{\n int gid = get_global_id(0);\n reserve_id_t res_id;\n\n"); |
| strcat(source, str); |
| sprintf(str, " if(gid < (get_global_size(0))/num_pipes)\n {\n res_id = reserve_write_pipe(pipe0, 1);\n if(is_valid_reserve_id(res_id))\n {\n"); |
| strcat(source, str); |
| sprintf(str, " write_pipe(pipe0, res_id, 0, &src[gid]);\n commit_write_pipe(pipe0, res_id);\n }\n }\n"); |
| strcat(source, str); |
| for(i = 1; i < num_pipes; i++){ |
| sprintf(str, " else if(gid < (%d*get_global_size(0))/num_pipes)\n {\n res_id = reserve_write_pipe(pipe%d, 1);\n if(is_valid_reserve_id(res_id))\n {\n", i+1, i); |
| strcat(source, str); |
| sprintf(str, " write_pipe(pipe%d, res_id, 0, &src[gid]);\n commit_write_pipe(pipe%d, res_id);\n }\n }\n", i, i); |
| strcat(source, str); |
| } |
| strcat(source, "}\n\n__kernel void test_multiple_pipe_read(__global int *dst, "); |
| |
| for(i = 0; i < num_pipes; i++) { |
| sprintf(str, "__read_only pipe int pipe%d, ", i); |
| strcat(source, str); |
| } |
| sprintf(str, "int num_pipes ) \n{\n int gid = get_global_id(0);\n reserve_id_t res_id;\n\n"); |
| strcat(source, str); |
| sprintf(str, " if(gid < (get_global_size(0))/num_pipes)\n {\n res_id = reserve_read_pipe(pipe0, 1);\n if(is_valid_reserve_id(res_id))\n {\n"); |
| strcat(source, str); |
| sprintf(str, " read_pipe(pipe0, res_id, 0, &dst[gid]);\n commit_read_pipe(pipe0, res_id);\n }\n }\n"); |
| strcat(source, str); |
| for(i = 1; i < num_pipes; i++){ |
| sprintf(str, " else if(gid < (%d*get_global_size(0))/num_pipes)\n {\n res_id = reserve_read_pipe(pipe%d, 1);\n if(is_valid_reserve_id(res_id))\n {\n", i+1, i); |
| strcat(source, str); |
| sprintf(str, " read_pipe(pipe%d, res_id, 0, &dst[gid]);\n commit_read_pipe(pipe%d, res_id);\n }\n }\n", i, i); |
| strcat(source, str); |
| } |
| strcat(source, "}"); |
| |
| str_length = strlen(source); |
| assert(str_length <= STRING_LENGTH*num_pipes); |
| } |
| |
| static int verify_result(void *ptr1, void *ptr2, int n) |
| { |
| int i; |
| int sum_input = 0, sum_output = 0; |
| cl_char *inptr = (cl_char *)ptr1; |
| cl_char *outptr = (cl_char *)ptr2; |
| |
| for(i = 0; i < n; i++) |
| { |
| sum_input += inptr[i]; |
| sum_output += outptr[i]; |
| } |
| if(sum_input != sum_output){ |
| return -1; |
| } |
| return 0; |
| } |
| |
| static int verify_result_int(void *ptr1, void *ptr2, int n) |
| { |
| int i; |
| int sum_input = 0, sum_output = 0; |
| cl_int *inptr = (cl_int *)ptr1; |
| cl_int *outptr = (cl_int *)ptr2; |
| |
| for(i = 0; i < n; i++) |
| { |
| sum_input += inptr[i]; |
| sum_output += outptr[i]; |
| } |
| if(sum_input != sum_output){ |
| return -1; |
| } |
| return 0; |
| } |
| |
| int test_pipe_max_args(cl_device_id deviceID, cl_context context, cl_command_queue queue, int num_elements) |
| { |
| |
| cl_mem pipes[1024]; |
| cl_mem buffers[2]; |
| void *outptr; |
| cl_int *inptr; |
| cl_program program; |
| cl_kernel kernel[2]; |
| size_t global_work_size[3]; |
| cl_int err; |
| cl_int size; |
| int num_pipe_elements = 1024; |
| int i, j; |
| int max_pipe_args; |
| char *source; |
| cl_event producer_sync_event = NULL; |
| cl_event consumer_sync_event = NULL; |
| MTdata d = init_genrand( gRandomSeed ); |
| const char* kernelName[] = {"test_multiple_pipe_write", "test_multiple_pipe_read"}; |
| |
| size_t min_alignment = get_min_alignment(context); |
| |
| err = clGetDeviceInfo(deviceID, CL_DEVICE_MAX_PIPE_ARGS, sizeof(max_pipe_args), (void*)&max_pipe_args, NULL); |
| if(err){ |
| print_error(err, " clGetDeviceInfo failed\n"); |
| return -1; |
| } |
| if(max_pipe_args < 16){ |
| log_error("The device should support minimum 16 pipe objects that could be passed as arguments to the kernel"); |
| return -1; |
| } |
| |
| global_work_size[0] = (cl_uint)num_pipe_elements * max_pipe_args; |
| size = sizeof(int) * num_pipe_elements * max_pipe_args; |
| source = (char *)malloc(STRING_LENGTH * sizeof(char) * max_pipe_args); |
| |
| inptr = (cl_int *)align_malloc(size, min_alignment); |
| |
| for(i = 0; i < num_pipe_elements * max_pipe_args; i++){ |
| inptr[i] = (int)genrand_int32(d); |
| } |
| |
| buffers[0] = clCreateBuffer(context, CL_MEM_COPY_HOST_PTR, size, inptr, &err); |
| if(err){ |
| clReleaseMemObject(buffers[0]); |
| free(source); |
| print_error(err, " clCreateBuffer failed\n"); |
| return -1; |
| } |
| outptr = align_malloc(size, min_alignment); |
| buffers[1] = clCreateBuffer(context, CL_MEM_USE_HOST_PTR, size, outptr, &err); |
| if ( err ){ |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| align_free( outptr ); |
| free(source); |
| print_error(err, " clCreateBuffer failed\n" ); |
| return -1; |
| } |
| |
| for(i = 0; i < max_pipe_args; i++){ |
| pipes[i] = clCreatePipe(context, CL_MEM_HOST_NO_ACCESS, sizeof(int), num_pipe_elements, NULL, &err); |
| if(err){ |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| align_free( outptr ); |
| free(source); |
| for(j = 0; j < i; j++) { |
| clReleaseMemObject(pipes[j]); |
| } |
| print_error(err, " clCreatePipe failed\n"); |
| return -1; |
| } |
| } |
| |
| createKernelSourceCode(source, max_pipe_args); |
| |
| // Create producer kernel |
| err = create_single_kernel_helper_with_build_options(context, &program, &kernel[0], 1, (const char**)&source, kernelName[0], "-cl-std=CL2.0"); |
| if(err){ |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| for(j = 0; j < max_pipe_args; j++) { |
| clReleaseMemObject(pipes[j]); |
| } |
| align_free(outptr); |
| free(source); |
| print_error(err, "Error creating program\n"); |
| return -1; |
| } |
| //Create consumer kernel |
| kernel[1] = clCreateKernel(program, kernelName[1], &err); |
| if( kernel[1] == NULL || err != CL_SUCCESS) |
| { |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| for(j = 0; j < max_pipe_args; j++) { |
| clReleaseMemObject(pipes[j]); |
| } |
| align_free(outptr); |
| free(source); |
| print_error(err, " Error creating kernel\n"); |
| return -1; |
| } |
| |
| err = clSetKernelArg(kernel[0], 0, sizeof(cl_mem), (void*)&buffers[0]); |
| for( i = 0; i < max_pipe_args; i++){ |
| err |= clSetKernelArg(kernel[0], i+1, sizeof(cl_mem), (void*)&pipes[i]); |
| } |
| err |= clSetKernelArg(kernel[0], max_pipe_args + 1, sizeof(int), (void*)&max_pipe_args); |
| err |= clSetKernelArg(kernel[1], 0, sizeof(cl_mem), (void*)&buffers[1]); |
| for( i = 0; i < max_pipe_args; i++){ |
| err |= clSetKernelArg(kernel[1], i+1, sizeof(cl_mem), (void*)&pipes[i]); |
| } |
| err |= clSetKernelArg(kernel[1], max_pipe_args + 1, sizeof(int), (void*)&max_pipe_args); |
| if ( err != CL_SUCCESS ){ |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| for(j = 0; j < max_pipe_args; j++) { |
| clReleaseMemObject(pipes[j]); |
| } |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseProgram(program); |
| align_free(outptr); |
| free(source); |
| print_error(err, " clSetKernelArg failed"); |
| return -1; |
| } |
| |
| // Launch Producer kernel |
| err = clEnqueueNDRangeKernel( queue, kernel[0], 1, NULL, global_work_size, NULL, 0, NULL, &producer_sync_event ); |
| if ( err != CL_SUCCESS ){ |
| print_error( err, " clEnqueueNDRangeKernel failed" ); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| for(j = 0; j < max_pipe_args; j++) { |
| clReleaseMemObject(pipes[j]); |
| } |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseEvent(producer_sync_event); |
| clReleaseEvent(consumer_sync_event); |
| clReleaseProgram(program); |
| align_free(outptr); |
| free(source); |
| return -1; |
| } |
| |
| // Launch Consumer kernel |
| err = clEnqueueNDRangeKernel( queue, kernel[1], 1, NULL, global_work_size, NULL, 1, &producer_sync_event, &consumer_sync_event ); |
| if ( err != CL_SUCCESS ){ |
| print_error( err, " clEnqueueNDRangeKernel failed" ); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| for(j = 0; j < max_pipe_args; j++) { |
| clReleaseMemObject(pipes[j]); |
| } |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseEvent(producer_sync_event); |
| clReleaseEvent(consumer_sync_event); |
| clReleaseProgram(program); |
| align_free(outptr); |
| free(source); |
| return -1; |
| } |
| |
| err = clEnqueueReadBuffer(queue, buffers[1], true, 0, size, outptr, 1, &consumer_sync_event, NULL); |
| if ( err != CL_SUCCESS ){ |
| print_error( err, " clEnqueueReadBuffer failed" ); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| for(j = 0; j < max_pipe_args; j++) { |
| clReleaseMemObject(pipes[j]); |
| } |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseEvent(producer_sync_event); |
| clReleaseEvent(consumer_sync_event); |
| clReleaseProgram(program); |
| align_free(outptr); |
| free(source); |
| return -1; |
| } |
| |
| err = clWaitForEvents(1, &consumer_sync_event); |
| if ( err != CL_SUCCESS ){ |
| print_error( err, " clWaitForEvents failed" ); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| for(j = 0; j < max_pipe_args; j++) { |
| clReleaseMemObject(pipes[j]); |
| } |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseEvent(producer_sync_event); |
| clReleaseEvent(consumer_sync_event); |
| clReleaseProgram(program); |
| align_free(outptr); |
| free(source); |
| return -1; |
| } |
| |
| if( verify_result( inptr, outptr, num_pipe_elements*sizeof(cl_int))){ |
| log_error("test_pipe_max_args failed\n"); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| for(j = 0; j < max_pipe_args; j++) { |
| clReleaseMemObject(pipes[j]); |
| } |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseEvent(producer_sync_event); |
| clReleaseEvent(consumer_sync_event); |
| clReleaseProgram(program); |
| align_free(outptr); |
| free(source); |
| return -1; |
| } |
| else { |
| log_info("test_pipe_max_args passed\n"); |
| } |
| //cleanup |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| for(j = 0; j < max_pipe_args; j++) { |
| clReleaseMemObject(pipes[j]); |
| } |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseEvent(producer_sync_event); |
| clReleaseEvent(consumer_sync_event); |
| clReleaseProgram(program); |
| align_free(outptr); |
| free(source); |
| |
| return 0; |
| } |
| |
| |
| int test_pipe_max_packet_size(cl_device_id deviceID, cl_context context, cl_command_queue queue, int num_elements) |
| { |
| cl_mem pipe; |
| cl_mem buffers[2]; |
| void *outptr; |
| cl_char *inptr; |
| cl_program program; |
| cl_kernel kernel[2]; |
| size_t global_work_size[3]; |
| cl_int err; |
| size_t size; |
| int num_pipe_elements = 1024; |
| int i; |
| cl_uint max_pipe_packet_size; |
| char *source; |
| char str[256]; |
| int str_length; |
| cl_event producer_sync_event = NULL; |
| cl_event consumer_sync_event = NULL; |
| MTdata d = init_genrand( gRandomSeed ); |
| const char* kernelName[] = {"test_pipe_max_packet_size_write", "test_pipe_max_packet_size_read"}; |
| |
| size_t min_alignment = get_min_alignment(context); |
| |
| global_work_size[0] = (cl_uint)num_pipe_elements; |
| |
| source = (char*)malloc(STRING_LENGTH*sizeof(char)); |
| |
| err = clGetDeviceInfo(deviceID, CL_DEVICE_PIPE_MAX_PACKET_SIZE, sizeof(max_pipe_packet_size), (void*)&max_pipe_packet_size, NULL); |
| if(err){ |
| print_error(err, " clGetDeviceInfo failed\n"); |
| return -1; |
| } |
| if(max_pipe_packet_size < 1024){ |
| log_error("The device should support minimum packet size of 1024 bytes"); |
| return -1; |
| } |
| |
| if(max_pipe_packet_size > (32*1024*1024/num_pipe_elements)) |
| { |
| max_pipe_packet_size = 32*1024*1024/num_pipe_elements; |
| } |
| |
| size = max_pipe_packet_size * num_pipe_elements; |
| |
| inptr = (cl_char *)align_malloc(size, min_alignment); |
| |
| for(i = 0; i < size; i++){ |
| inptr[i] = (char)genrand_int32(d); |
| } |
| |
| buffers[0] = clCreateBuffer(context, CL_MEM_COPY_HOST_PTR, size, inptr, &err); |
| if(err){ |
| clReleaseMemObject(buffers[0]); |
| free(source); |
| print_error(err, " clCreateBuffer failed\n"); |
| return -1; |
| } |
| outptr = align_malloc(size, min_alignment); |
| buffers[1] = clCreateBuffer(context, CL_MEM_USE_HOST_PTR, size, outptr, &err); |
| if ( err ){ |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| align_free( outptr ); |
| free(source); |
| print_error(err, " clCreateBuffer failed\n" ); |
| return -1; |
| } |
| |
| pipe = clCreatePipe(context, CL_MEM_HOST_NO_ACCESS, max_pipe_packet_size, num_pipe_elements, NULL, &err); |
| if(err){ |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| align_free( outptr ); |
| free(source); |
| clReleaseMemObject(pipe); |
| print_error(err, " clCreatePipe failed\n"); |
| return -1; |
| } |
| |
| sprintf(str, "typedef struct{\n char a[%d];\n}TestStruct;\n\n__kernel void test_pipe_max_packet_size_write(__global TestStruct *src, __write_only pipe TestStruct out_pipe)\n{\n", max_pipe_packet_size); |
| strcpy(source,str); |
| strcat(source, " int gid = get_global_id(0);\n reserve_id_t res_id;\n\n"); |
| sprintf(str, " res_id = reserve_write_pipe(out_pipe, 1);\n if(is_valid_reserve_id(res_id))\n {\n"); |
| strcat(source, str); |
| sprintf(str, " write_pipe(out_pipe, res_id, 0, &src[gid]);\n commit_write_pipe(out_pipe, res_id);\n }\n}\n\n"); |
| strcat(source, str); |
| sprintf(str, "__kernel void test_pipe_max_packet_size_read(__read_only pipe TestStruct in_pipe, __global TestStruct *dst)\n{\n"); |
| strcat(source, str); |
| strcat(source, " int gid = get_global_id(0);\n reserve_id_t res_id;\n\n"); |
| sprintf(str, " res_id = reserve_read_pipe(in_pipe, 1);\n if(is_valid_reserve_id(res_id))\n {\n"); |
| strcat(source, str); |
| sprintf(str, " read_pipe(in_pipe, res_id, 0, &dst[gid]);\n commit_read_pipe(in_pipe, res_id);\n }\n}\n\n"); |
| strcat(source, str); |
| |
| str_length = strlen(source); |
| assert(str_length <= STRING_LENGTH); |
| |
| // Create producer kernel |
| err = create_single_kernel_helper_with_build_options(context, &program, &kernel[0], 1, (const char**)&source, kernelName[0], "-cl-std=CL2.0"); |
| if(err){ |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| clReleaseMemObject(pipe); |
| align_free(outptr); |
| free(source); |
| print_error(err, "Error creating program\n"); |
| return -1; |
| } |
| //Create consumer kernel |
| kernel[1] = clCreateKernel(program, kernelName[1], &err); |
| if( kernel[1] == NULL || err != CL_SUCCESS) |
| { |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| clReleaseMemObject(pipe); |
| align_free(outptr); |
| free(source); |
| print_error(err, "Error creating kernel\n"); |
| return -1; |
| } |
| |
| err = clSetKernelArg(kernel[0], 0, sizeof(cl_mem), (void*)&buffers[0]); |
| err |= clSetKernelArg(kernel[0], 1, sizeof(cl_mem), (void*)&pipe); |
| err |= clSetKernelArg(kernel[1], 0, sizeof(cl_mem), (void*)&pipe); |
| err |= clSetKernelArg(kernel[1], 1, sizeof(cl_mem), (void*)&buffers[1]); |
| if ( err != CL_SUCCESS ){ |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| clReleaseMemObject(pipe); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseProgram(program); |
| align_free(outptr); |
| free(source); |
| print_error(err, " clSetKernelArg failed"); |
| return -1; |
| } |
| |
| // Launch Producer kernel |
| err = clEnqueueNDRangeKernel( queue, kernel[0], 1, NULL, global_work_size, NULL, 0, NULL, &producer_sync_event ); |
| if ( err != CL_SUCCESS ){ |
| print_error( err, " clEnqueueNDRangeKernel failed" ); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| clReleaseMemObject(pipe); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseEvent(producer_sync_event); |
| clReleaseEvent(consumer_sync_event); |
| clReleaseProgram(program); |
| align_free(outptr); |
| free(source); |
| return -1; |
| } |
| |
| // Launch Consumer kernel |
| err = clEnqueueNDRangeKernel( queue, kernel[1], 1, NULL, global_work_size, NULL, 1, &producer_sync_event, &consumer_sync_event ); |
| if ( err != CL_SUCCESS ){ |
| print_error( err, " clEnqueueNDRangeKernel failed" ); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| clReleaseMemObject(pipe); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseEvent(producer_sync_event); |
| clReleaseEvent(consumer_sync_event); |
| clReleaseProgram(program); |
| align_free(outptr); |
| free(source); |
| return -1; |
| } |
| |
| err = clEnqueueReadBuffer(queue, buffers[1], true, 0, size, outptr, 1, &consumer_sync_event, NULL); |
| if ( err != CL_SUCCESS ){ |
| print_error( err, " clEnqueueReadBuffer failed" ); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| clReleaseMemObject(pipe); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseEvent(producer_sync_event); |
| clReleaseEvent(consumer_sync_event); |
| clReleaseProgram(program); |
| align_free(outptr); |
| free(source); |
| return -1; |
| } |
| |
| if( verify_result( inptr, outptr, size)){ |
| log_error("test_pipe_max_packet_size failed\n"); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| clReleaseMemObject(pipe); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseEvent(producer_sync_event); |
| clReleaseEvent(consumer_sync_event); |
| clReleaseProgram(program); |
| align_free(outptr); |
| free(source); |
| return -1; |
| } |
| else { |
| log_info("test_pipe_max_packet_size passed\n"); |
| } |
| //cleanup |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| clReleaseMemObject(pipe); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseEvent(producer_sync_event); |
| clReleaseEvent(consumer_sync_event); |
| clReleaseProgram(program); |
| align_free(outptr); |
| free(source); |
| |
| return 0; |
| } |
| |
| int test_pipe_max_active_reservations(cl_device_id deviceID, cl_context context, cl_command_queue queue, int num_elements) |
| { |
| cl_mem pipe; |
| cl_mem buffers[2]; |
| cl_mem buf_reservations; |
| cl_mem buf_status; |
| cl_mem buf_reserve_id_t_size; |
| cl_mem buf_reserve_id_t_size_aligned; |
| cl_int *inptr; |
| void *outptr; |
| int size, i; |
| cl_program program; |
| cl_kernel kernel[3]; |
| size_t global_work_size[3]; |
| cl_int err; |
| int status = 0; |
| cl_uint max_active_reservations = 0; |
| cl_ulong max_global_size = 0; |
| int reserve_id_t_size; |
| int temp; |
| char *source; |
| char str[256]; |
| int str_length; |
| cl_event sync_event = NULL; |
| cl_event read_event = NULL; |
| MTdata d = init_genrand( gRandomSeed ); |
| const char* kernelName[3] = {"test_pipe_max_active_reservations_write", "test_pipe_max_active_reservations_read", "pipe_get_reserve_id_t_size"}; |
| |
| size_t min_alignment = get_min_alignment(context); |
| |
| source = (char*)malloc(2*STRING_LENGTH*sizeof(char)); |
| |
| global_work_size[0] = 1; |
| |
| err = clGetDeviceInfo(deviceID, CL_DEVICE_PIPE_MAX_ACTIVE_RESERVATIONS, sizeof(max_active_reservations), (void*)&max_active_reservations, NULL); |
| if(err){ |
| print_error(err, " clGetDeviceInfo failed\n"); |
| return -1; |
| } |
| |
| err = clGetDeviceInfo(deviceID, CL_DEVICE_GLOBAL_MEM_SIZE, sizeof(max_global_size), (void*)&max_global_size, NULL); |
| if(err){ |
| print_error(err, " clGetDeviceInfo failed\n"); |
| return -1; |
| } |
| |
| max_active_reservations = (max_active_reservations > max_global_size) ? 1<<16 : max_active_reservations; |
| |
| if(max_active_reservations < 1){ |
| log_error("The device should support minimum active reservations of 1"); |
| return -1; |
| } |
| |
| // To get reserve_id_t size |
| buf_reserve_id_t_size = clCreateBuffer(context, CL_MEM_HOST_READ_ONLY, sizeof(reserve_id_t_size), NULL, &err); |
| if ( err ){ |
| clReleaseMemObject(buf_reserve_id_t_size); |
| print_error(err, " clCreateBuffer failed\n" ); |
| return -1; |
| } |
| |
| sprintf(str, "__kernel void test_pipe_max_active_reservations_write(__global int *src, __write_only pipe int out_pipe, __global char *reserve_id, __global int *reserve_id_t_size_aligned, __global int *status)\n{\n"); |
| strcpy(source,str); |
| sprintf(str, " __global reserve_id_t *res_id_ptr;\n int reserve_idx;\n int commit_idx;\n"); |
| strcat(source, str); |
| sprintf(str, " for(reserve_idx = 0; reserve_idx < %d; reserve_idx++)\n {\n", max_active_reservations); |
| strcat(source, str); |
| sprintf(str, " res_id_ptr = (__global reserve_id_t*)(reserve_id + reserve_idx*reserve_id_t_size_aligned[0]);\n"); |
| strcat(source, str); |
| sprintf(str, " *res_id_ptr = reserve_write_pipe(out_pipe, 1);\n"); |
| strcat(source, str); |
| sprintf(str, " if(is_valid_reserve_id(res_id_ptr[0]))\n {\n write_pipe(out_pipe, res_id_ptr[0], 0, &src[reserve_idx]);\n }\n"); |
| strcat(source, str); |
| sprintf(str, " else\n {\n *status = -1;\n return;\n }\n }\n"); |
| strcat(source, str); |
| sprintf(str, " for(commit_idx = 0; commit_idx < %d; commit_idx++)\n {\n", max_active_reservations); |
| strcat(source, str); |
| sprintf(str, " res_id_ptr = (__global reserve_id_t*)(reserve_id + commit_idx*reserve_id_t_size_aligned[0]);\n"); |
| strcat(source, str); |
| sprintf(str, " commit_write_pipe(out_pipe, res_id_ptr[0]);\n }\n}\n\n"); |
| strcat(source, str); |
| sprintf(str, "__kernel void test_pipe_max_active_reservations_read(__read_only pipe int in_pipe, __global int *dst, __global char *reserve_id, __global int *reserve_id_t_size_aligned, __global int *status)\n{\n"); |
| strcat(source, str); |
| sprintf(str, " __global reserve_id_t *res_id_ptr;\n int reserve_idx;\n int commit_idx;\n"); |
| strcat(source, str); |
| sprintf(str, " for(reserve_idx = 0; reserve_idx < %d; reserve_idx++)\n {\n", max_active_reservations); |
| strcat(source, str); |
| sprintf(str, " res_id_ptr = (__global reserve_id_t*)(reserve_id + reserve_idx*reserve_id_t_size_aligned[0]);\n"); |
| strcat(source, str); |
| sprintf(str, " *res_id_ptr = reserve_read_pipe(in_pipe, 1);\n"); |
| strcat(source, str); |
| sprintf(str, " if(is_valid_reserve_id(res_id_ptr[0]))\n {\n read_pipe(in_pipe, res_id_ptr[0], 0, &dst[reserve_idx]);\n }\n"); |
| strcat(source, str); |
| sprintf(str, " else\n {\n *status = -1;\n return;\n }\n }\n"); |
| strcat(source, str); |
| sprintf(str, " for(commit_idx = 0; commit_idx < %d; commit_idx++)\n {\n", max_active_reservations); |
| strcat(source, str); |
| sprintf(str, " res_id_ptr = (__global reserve_id_t*)(reserve_id + commit_idx*reserve_id_t_size_aligned[0]);\n"); |
| strcat(source, str); |
| sprintf(str, " commit_read_pipe(in_pipe, res_id_ptr[0]);\n }\n}\n\n"); |
| strcat(source, str); |
| sprintf(str, "__kernel void pipe_get_reserve_id_t_size(__global int *reserve_id_t_size) \n"); |
| strcat(source, str); |
| sprintf(str, "{\n *reserve_id_t_size = sizeof(reserve_id_t);\n}\n"); |
| strcat(source, str); |
| |
| str_length = strlen(source); |
| assert(str_length <= 2*STRING_LENGTH); |
| |
| // Create producer kernel |
| err = create_single_kernel_helper_with_build_options(context, &program, &kernel[0], 1, (const char**)&source, kernelName[0], "-cl-std=CL2.0"); |
| if(err){ |
| clReleaseMemObject(buf_reserve_id_t_size); |
| print_error(err, "Error creating program\n"); |
| return -1; |
| } |
| |
| // Create consumer kernel |
| kernel[1] = clCreateKernel(program, kernelName[1], &err); |
| if( kernel[1] == NULL || err != CL_SUCCESS) |
| { |
| clReleaseMemObject(buf_reserve_id_t_size); |
| print_error(err, "Error creating kernel\n"); |
| return -1; |
| } |
| |
| // Create size query kernel for reserve_id_t |
| kernel[2] = clCreateKernel(program, kernelName[2], &err); |
| if( kernel[2] == NULL || err != CL_SUCCESS) |
| { |
| clReleaseMemObject(buf_reserve_id_t_size); |
| print_error(err, "Error creating kernel\n"); |
| return -1; |
| } |
| err = clSetKernelArg(kernel[2], 0, sizeof(cl_mem), (void*)&buf_reserve_id_t_size); |
| if(err){ |
| clReleaseMemObject(buf_reserve_id_t_size); |
| print_error(err, "Error creating program\n"); |
| return -1; |
| } |
| //Launch size query kernel for reserve_id_t |
| err = clEnqueueNDRangeKernel( queue, kernel[2], 1, NULL, global_work_size, NULL, 0, NULL, &sync_event ); |
| if ( err != CL_SUCCESS ){ |
| print_error( err, " clEnqueueNDRangeKernel failed" ); |
| clReleaseMemObject(buf_reserve_id_t_size); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseKernel(kernel[2]); |
| clReleaseEvent(sync_event); |
| clReleaseProgram(program); |
| return -1; |
| } |
| |
| err = clEnqueueReadBuffer(queue, buf_reserve_id_t_size, true, 0, sizeof(reserve_id_t_size), &reserve_id_t_size, 1, &sync_event, &read_event); |
| if ( err != CL_SUCCESS ){ |
| print_error( err, " clEnqueueReadBuffer failed" ); |
| clReleaseMemObject(buf_reserve_id_t_size); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseKernel(kernel[2]); |
| clReleaseEvent(sync_event); |
| clReleaseProgram(program); |
| return -1; |
| } |
| |
| err = clWaitForEvents(1, &read_event); |
| if ( err != CL_SUCCESS ){ |
| print_error( err, " clWaitForEvents failed" ); |
| clReleaseMemObject(buf_reserve_id_t_size); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseKernel(kernel[2]); |
| clReleaseEvent(sync_event); |
| clReleaseEvent(read_event); |
| clReleaseProgram(program); |
| return -1; |
| } |
| |
| // Round reserve_id_t_size to the nearest power of 2 |
| temp = 1; |
| while(temp < reserve_id_t_size) |
| temp *= 2; |
| reserve_id_t_size = temp; |
| |
| size = sizeof(cl_int) * max_active_reservations; |
| inptr = (cl_int *)align_malloc(size, min_alignment); |
| |
| for(i = 0; i < max_active_reservations; i++){ |
| inptr[i] = (int)genrand_int32(d); |
| } |
| |
| buffers[0] = clCreateBuffer(context, CL_MEM_COPY_HOST_PTR, size, inptr, &err); |
| if ( err ){ |
| clReleaseMemObject(buf_reserve_id_t_size); |
| clReleaseMemObject(buffers[0]); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseKernel(kernel[2]); |
| clReleaseEvent(sync_event); |
| clReleaseEvent(read_event); |
| clReleaseProgram(program); |
| print_error(err, " clCreateBuffer failed\n" ); |
| return -1; |
| } |
| |
| outptr = align_malloc(size, min_alignment); |
| buffers[1] = clCreateBuffer(context, CL_MEM_HOST_READ_ONLY, size, NULL, &err); |
| if ( err ){ |
| clReleaseMemObject(buf_reserve_id_t_size); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseKernel(kernel[2]); |
| clReleaseEvent(sync_event); |
| clReleaseEvent(read_event); |
| clReleaseProgram(program); |
| align_free(outptr); |
| print_error(err, " clCreateBuffer failed\n" ); |
| return -1; |
| } |
| |
| buf_reserve_id_t_size_aligned = clCreateBuffer(context, CL_MEM_COPY_HOST_PTR, sizeof(reserve_id_t_size), &reserve_id_t_size, &err); |
| if ( err ){ |
| clReleaseMemObject(buf_reserve_id_t_size); |
| clReleaseMemObject(buf_reserve_id_t_size_aligned); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseKernel(kernel[2]); |
| clReleaseEvent(sync_event); |
| clReleaseEvent(read_event); |
| clReleaseProgram(program); |
| print_error(err, " clCreateBuffer failed\n" ); |
| return -1; |
| } |
| |
| //For error status |
| buf_status = clCreateBuffer(context, CL_MEM_COPY_HOST_PTR, sizeof(int), &status, &err); |
| if ( err ){ |
| clReleaseMemObject(buf_status); |
| clReleaseMemObject(buf_reserve_id_t_size); |
| clReleaseMemObject(buf_reserve_id_t_size_aligned); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| align_free(outptr); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseKernel(kernel[2]); |
| clReleaseEvent(sync_event); |
| clReleaseEvent(read_event); |
| clReleaseProgram(program); |
| print_error(err, " clCreateBuffer failed\n" ); |
| return -1; |
| } |
| |
| pipe = clCreatePipe(context, CL_MEM_HOST_NO_ACCESS, sizeof(int), max_active_reservations, NULL, &err); |
| if(err){ |
| clReleaseMemObject(buf_status); |
| clReleaseMemObject(buf_reserve_id_t_size); |
| clReleaseMemObject(buf_reserve_id_t_size_aligned); |
| clReleaseMemObject(pipe); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| align_free(outptr); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseKernel(kernel[2]); |
| clReleaseEvent(sync_event); |
| clReleaseEvent(read_event); |
| clReleaseProgram(program); |
| print_error(err, " clCreatePipe failed\n"); |
| return -1; |
| } |
| |
| // Global buffer to hold all active reservation ids |
| buf_reservations = clCreateBuffer(context, CL_MEM_HOST_NO_ACCESS, reserve_id_t_size*max_active_reservations, NULL, &err); |
| if ( err != CL_SUCCESS ){ |
| print_error( err, " clCreateBuffer failed" ); |
| clReleaseMemObject(buf_status); |
| clReleaseMemObject(buf_reserve_id_t_size); |
| clReleaseMemObject(buf_reserve_id_t_size_aligned); |
| clReleaseMemObject(buf_reservations); |
| clReleaseMemObject(pipe); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| align_free(outptr); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseKernel(kernel[2]); |
| clReleaseEvent(sync_event); |
| clReleaseEvent(read_event); |
| clReleaseProgram(program); |
| return -1; |
| } |
| |
| err = clSetKernelArg(kernel[0], 0, sizeof(cl_mem), (void*)&buffers[0]); |
| err |= clSetKernelArg(kernel[0], 1, sizeof(cl_mem), (void*)&pipe); |
| err |= clSetKernelArg(kernel[0], 2, sizeof(cl_mem), (void*)&buf_reservations); |
| err |= clSetKernelArg(kernel[0], 3, sizeof(cl_mem), (void*)&buf_reserve_id_t_size_aligned); |
| err |= clSetKernelArg(kernel[0], 4, sizeof(cl_mem), (void*)&buf_status); |
| if ( err != CL_SUCCESS ){ |
| clReleaseMemObject(buf_status); |
| clReleaseMemObject(buf_reserve_id_t_size); |
| clReleaseMemObject(buf_reserve_id_t_size_aligned); |
| clReleaseMemObject(buf_reservations); |
| clReleaseMemObject(pipe); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| align_free(outptr); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseKernel(kernel[2]); |
| clReleaseProgram(program); |
| print_error(err, " clSetKernelArg failed"); |
| return -1; |
| } |
| |
| err = clSetKernelArg(kernel[1], 0, sizeof(cl_mem), (void*)&pipe); |
| err |= clSetKernelArg(kernel[1], 1, sizeof(cl_mem), (void*)&buffers[1]); |
| err |= clSetKernelArg(kernel[1], 2, sizeof(cl_mem), (void*)&buf_reservations); |
| err |= clSetKernelArg(kernel[1], 3, sizeof(cl_mem), (void*)&buf_reserve_id_t_size_aligned); |
| err |= clSetKernelArg(kernel[1], 4, sizeof(cl_mem), (void*)&buf_status); |
| if ( err != CL_SUCCESS ){ |
| clReleaseMemObject(buf_status); |
| clReleaseMemObject(buf_reserve_id_t_size); |
| clReleaseMemObject(buf_reserve_id_t_size_aligned); |
| clReleaseMemObject(buf_reservations); |
| clReleaseMemObject(pipe); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| align_free(outptr); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseKernel(kernel[2]); |
| clReleaseProgram(program); |
| print_error(err, " clSetKernelArg failed"); |
| return -1; |
| } |
| |
| // Launch Producer kernel |
| err = clEnqueueNDRangeKernel(queue, kernel[0], 1, NULL, global_work_size, NULL, 0, NULL, &sync_event); |
| if ( err != CL_SUCCESS ){ |
| print_error( err, " clEnqueueNDRangeKernel failed" ); |
| clReleaseMemObject(buf_status); |
| clReleaseMemObject(buf_reserve_id_t_size); |
| clReleaseMemObject(buf_reserve_id_t_size_aligned); |
| clReleaseMemObject(buf_reservations); |
| clReleaseMemObject(pipe); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| align_free(outptr); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseKernel(kernel[2]); |
| clReleaseEvent(sync_event); |
| clReleaseProgram(program); |
| return -1; |
| } |
| |
| err = clEnqueueReadBuffer(queue, buf_status, true, 0, sizeof(int), &status, 1, &sync_event, NULL); |
| if ( err != CL_SUCCESS ){ |
| print_error( err, " clEnqueueReadBuffer failed" ); |
| clReleaseMemObject(buf_status); |
| clReleaseMemObject(buf_reserve_id_t_size); |
| clReleaseMemObject(buf_reserve_id_t_size_aligned); |
| clReleaseMemObject(buf_reservations); |
| clReleaseMemObject(pipe); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| align_free(outptr); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseKernel(kernel[2]); |
| clReleaseEvent(sync_event); |
| clReleaseProgram(program); |
| return -1; |
| } |
| |
| if(status != 0) |
| { |
| log_error("test_pipe_max_active_reservations failed\n"); |
| clReleaseMemObject(buf_status); |
| clReleaseMemObject(buf_reserve_id_t_size); |
| clReleaseMemObject(buf_reserve_id_t_size_aligned); |
| clReleaseMemObject(buf_reservations); |
| clReleaseMemObject(pipe); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| align_free(outptr); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseEvent(sync_event); |
| clReleaseProgram(program); |
| return -1; |
| } |
| |
| // Launch Consumer kernel |
| err = clEnqueueNDRangeKernel(queue, kernel[1], 1, NULL, global_work_size, NULL, 0, NULL, &sync_event); |
| if ( err != CL_SUCCESS ){ |
| print_error( err, " clEnqueueNDRangeKernel failed" ); |
| clReleaseMemObject(buf_status); |
| clReleaseMemObject(buf_reserve_id_t_size); |
| clReleaseMemObject(buf_reserve_id_t_size_aligned); |
| clReleaseMemObject(buf_reservations); |
| clReleaseMemObject(pipe); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| align_free(outptr); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseKernel(kernel[2]); |
| clReleaseEvent(sync_event); |
| clReleaseProgram(program); |
| return -1; |
| } |
| |
| err = clEnqueueReadBuffer(queue, buf_status, true, 0, sizeof(int), &status, 1, &sync_event, NULL); |
| if ( err != CL_SUCCESS ){ |
| print_error( err, " clEnqueueReadBuffer failed" ); |
| clReleaseMemObject(buf_status); |
| clReleaseMemObject(buf_reserve_id_t_size); |
| clReleaseMemObject(buf_reserve_id_t_size_aligned); |
| clReleaseMemObject(buf_reservations); |
| clReleaseMemObject(pipe); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| align_free(outptr); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseKernel(kernel[2]); |
| clReleaseEvent(sync_event); |
| clReleaseProgram(program); |
| return -1; |
| } |
| |
| if(status != 0) |
| { |
| log_error("test_pipe_max_active_reservations failed\n"); |
| clReleaseMemObject(buf_status); |
| clReleaseMemObject(buf_reserve_id_t_size); |
| clReleaseMemObject(buf_reserve_id_t_size_aligned); |
| clReleaseMemObject(buf_reservations); |
| clReleaseMemObject(pipe); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| align_free(outptr); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseEvent(sync_event); |
| clReleaseProgram(program); |
| return -1; |
| } |
| |
| err = clEnqueueReadBuffer(queue, buffers[1], true, 0, size, outptr, 1, &sync_event, NULL); |
| if ( err != CL_SUCCESS ){ |
| print_error( err, " clEnqueueReadBuffer failed" ); |
| clReleaseMemObject(buf_status); |
| clReleaseMemObject(buf_reserve_id_t_size); |
| clReleaseMemObject(buf_reserve_id_t_size_aligned); |
| clReleaseMemObject(buf_reservations); |
| clReleaseMemObject(pipe); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| align_free(outptr); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseKernel(kernel[2]); |
| clReleaseEvent(sync_event); |
| clReleaseProgram(program); |
| return -1; |
| } |
| |
| if( verify_result_int( inptr, outptr, max_active_reservations)){ |
| log_error("test_pipe_max_active_reservations failed\n"); |
| clReleaseMemObject(buf_status); |
| clReleaseMemObject(buf_reserve_id_t_size); |
| clReleaseMemObject(buf_reserve_id_t_size_aligned); |
| clReleaseMemObject(buf_reservations); |
| clReleaseMemObject(pipe); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| align_free(outptr); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseKernel(kernel[2]); |
| clReleaseEvent(sync_event); |
| clReleaseProgram(program); |
| return -1; |
| } |
| else { |
| log_info("test_pipe_max_active_reservations passed\n"); |
| } |
| |
| //cleanup |
| clReleaseMemObject(buf_status); |
| clReleaseMemObject(buf_reserve_id_t_size); |
| clReleaseMemObject(buf_reserve_id_t_size_aligned); |
| clReleaseMemObject(buf_reservations); |
| clReleaseMemObject(pipe); |
| clReleaseMemObject(buffers[0]); |
| clReleaseMemObject(buffers[1]); |
| align_free(outptr); |
| clReleaseKernel(kernel[0]); |
| clReleaseKernel(kernel[1]); |
| clReleaseEvent(sync_event); |
| clReleaseProgram(program); |
| return 0; |
| } |
