1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
|
//
// 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 <stdio.h>
#include <string.h>
#include "harness/testHarness.h"
#include "harness/typeWrappers.h"
#include <vector>
#include "procs.h"
#include "utils.h"
#include <time.h>
extern int gWimpyMode;
#ifdef CL_VERSION_2_0
static const char* enqueue_block_first_kernel[] =
{
NL, "void block_fn(uint num, __global int* res)"
NL, "{"
NL, " size_t tid = get_global_id(0);"
NL, ""
NL, " for(int i = 1 ; i < tid ; i++)"
NL, " {"
NL, " for(int j = 0 ; j < num ; j++)"
NL, " atomic_add(res+tid, 1);"
NL, " }"
NL, "}"
NL, ""
NL, "kernel void enqueue_block_first_kernel(uint num, __global int* res)"
NL, "{"
NL, " void (^kernelBlock)(void) = ^{ block_fn(num, res); };"
NL, ""
NL, " ndrange_t ndrange = ndrange_1D(num, 1);"
NL, ""
NL, " int enq_res = enqueue_kernel(get_default_queue(), CLK_ENQUEUE_FLAGS_NO_WAIT, ndrange, kernelBlock);"
NL, " if(enq_res != CLK_SUCCESS) { res[0] = -1; return; }"
NL, ""
NL, "}"
NL
};
static const char* enqueue_block_second_kernel[] =
{
NL, "void block_fn(uint num, __global int* res)"
NL, "{"
NL, " for(int i = 2 ; i < num ; i++)"
NL, " {"
NL, " res[i] = res[i]/num - (i-1);"
NL, " }"
NL, "}"
NL, ""
NL, "kernel void enqueue_block_second_kernel(uint num, __global int* res)"
NL, "{"
NL, " void (^kernelBlock)(void) = ^{ block_fn(num, res); };"
NL, ""
NL, " ndrange_t ndrange = ndrange_1D(1);"
NL, ""
NL, " int enq_res = enqueue_kernel(get_default_queue(), CLK_ENQUEUE_FLAGS_WAIT_KERNEL, ndrange, kernelBlock);"
NL, " if(enq_res != CLK_SUCCESS) { res[0] = -1; return; }"
NL, ""
NL, "}"
NL
};
static int check_kernel_results(cl_int* results, cl_int len)
{
for(cl_int i = 0; i < len; ++i)
{
if(results[i] != 0) return i;
}
return -1;
}
/*
Test checks kernel block execution order in case of two different kernels with enqueue block submitted to one ordered host queue.
*/
int test_host_queue_order(cl_device_id device, cl_context context, cl_command_queue queue, int num_elements)
{
cl_int k, err_ret, res = 0;
clCommandQueueWrapper dev_queue;
cl_int kernel_results[MAX_GWS] = {0};
size_t ret_len;
cl_uint max_queues = 1;
cl_uint maxQueueSize = 0;
err_ret = clGetDeviceInfo(device, CL_DEVICE_QUEUE_ON_DEVICE_MAX_SIZE, sizeof(maxQueueSize), &maxQueueSize, 0);
test_error(err_ret, "clGetDeviceInfo(CL_DEVICE_QUEUE_ON_DEVICE_MAX_SIZE) failed");
err_ret = clGetDeviceInfo(device, CL_DEVICE_MAX_ON_DEVICE_QUEUES, sizeof(max_queues), &max_queues, &ret_len);
test_error(err_ret, "clGetDeviceInfo(CL_DEVICE_MAX_ON_DEVICE_QUEUES) failed");
size_t max_local_size = 1;
err_ret = clGetDeviceInfo(device, CL_DEVICE_MAX_WORK_GROUP_SIZE, sizeof(max_local_size), &max_local_size, &ret_len);
test_error(err_ret, "clGetDeviceInfo(CL_DEVICE_MAX_WORK_GROUP_SIZE) failed");
cl_queue_properties queue_prop_def[] =
{
CL_QUEUE_PROPERTIES, CL_QUEUE_OUT_OF_ORDER_EXEC_MODE_ENABLE|CL_QUEUE_ON_DEVICE|CL_QUEUE_ON_DEVICE_DEFAULT,
CL_QUEUE_SIZE, maxQueueSize,
0
};
dev_queue = clCreateCommandQueueWithProperties(context, device, queue_prop_def, &err_ret);
test_error(err_ret, "clCreateCommandQueueWithProperties(CL_QUEUE_DEVICE|CL_QUEUE_DEFAULT) failed");
cl_int status;
size_t size = 1;
cl_int result[MAX_GWS] = { 0 };
cl_uint num = arr_size(result);
if( gWimpyMode )
{
num = MAX(num / 16, 4);
}
clMemWrapper res_mem;
clProgramWrapper program1, program2;
clKernelWrapper kernel1, kernel2;
cl_event kernel_event;
err_ret = create_single_kernel_helper(
context, &program1, &kernel1, arr_size(enqueue_block_first_kernel),
enqueue_block_first_kernel, "enqueue_block_first_kernel");
if(check_error(err_ret, "Create single kernel failed")) return -1;
err_ret = create_single_kernel_helper(
context, &program2, &kernel2, arr_size(enqueue_block_second_kernel),
enqueue_block_second_kernel, "enqueue_block_second_kernel");
if(check_error(err_ret, "Create single kernel failed")) return -1;
res_mem = clCreateBuffer(context, CL_MEM_READ_WRITE|CL_MEM_COPY_HOST_PTR, sizeof(kernel_results), kernel_results, &err_ret);
test_error(err_ret, "clCreateBuffer() failed");
// Enqueue first kernel
err_ret = clSetKernelArg(kernel1, 0, sizeof(num), &num);
test_error(err_ret, "clSetKernelArg(0) failed");
err_ret = clSetKernelArg(kernel1, 1, sizeof(cl_mem), &res_mem);
test_error(err_ret, "clSetKernelArg(1) failed");
cl_event event1 = clCreateUserEvent(context, &err_ret);
if(check_error(err_ret, "Create user event failed")) return -1;
err_ret = clEnqueueNDRangeKernel(queue, kernel1, 1, NULL, &size, &size, 1, &event1, NULL);
test_error(err_ret, "clEnqueueNDRangeKernel('enqueue_block_first_kernel') failed");
// Enqueue second kernel
err_ret = clSetKernelArg(kernel2, 0, sizeof(num), &num);
test_error(err_ret, "clSetKernelArg(0) failed");
err_ret = clSetKernelArg(kernel2, 1, sizeof(cl_mem), &res_mem);
test_error(err_ret, "clSetKernelArg(1) failed");
err_ret = clEnqueueNDRangeKernel(queue, kernel2, 1, NULL, &size, &size, 0, NULL, &kernel_event);
test_error(err_ret, "clEnqueueNDRangeKernel('enqueue_block_second_kernel') failed");
//Triger execution of first kernel
err_ret = clSetUserEventStatus(event1, CL_COMPLETE);
test_error(err_ret, "clSetUserEventStatus() failed");
// Collect resulsts
err_ret = clEnqueueReadBuffer(queue, res_mem, CL_TRUE, 0, sizeof(result), result, 0, NULL, NULL);
test_error(err_ret, "clEnqueueReadBuffer() failed");
err_ret = clGetEventInfo(kernel_event, CL_EVENT_COMMAND_EXECUTION_STATUS, sizeof(status), &status, &ret_len);
test_error(err_ret, "clGetEventInfo() failed");
if(check_error(status, "Kernel execution status %d", status)) return status;
if((k = check_kernel_results(result, num)) >= 0 && check_error(-1, "'%s' results validation failed: [%d] returned %d expected 0", "test_host_queue_order", k, result[k])) res = -1;
clReleaseEvent(kernel_event);
clReleaseEvent(event1);
return res;
}
#endif
|
