1 //
2 // Copyright (c) 2017 The Khronos Group Inc.
3 //
4 // Licensed under the Apache License, Version 2.0 (the "License");
5 // you may not use this file except in compliance with the License.
6 // You may obtain a copy of the License at
7 //
8 //    http://www.apache.org/licenses/LICENSE-2.0
9 //
10 // Unless required by applicable law or agreed to in writing, software
11 // distributed under the License is distributed on an "AS IS" BASIS,
12 // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13 // See the License for the specific language governing permissions and
14 // limitations under the License.
15 //
16 
17 #include "function_list.h"
18 #include "test_functions.h"
19 #include "utility.h"
20 
21 #include <cstring>
22 
BuildKernel(const char * name,int vectorSize,cl_kernel * k,cl_program * p,bool relaxedMode)23 static int BuildKernel(const char *name, int vectorSize, cl_kernel *k,
24                        cl_program *p, bool relaxedMode)
25 {
26     const char *c[] = { "#pragma OPENCL EXTENSION cl_khr_fp64 : enable\n",
27                         "__kernel void math_kernel",
28                         sizeNames[vectorSize],
29                         "( __global double",
30                         sizeNames[vectorSize],
31                         "* out, __global double",
32                         sizeNames[vectorSize],
33                         "* in1, __global double",
34                         sizeNames[vectorSize],
35                         "* in2,  __global double",
36                         sizeNames[vectorSize],
37                         "* in3 )\n"
38                         "{\n"
39                         "   size_t i = get_global_id(0);\n"
40                         "   out[i] = ",
41                         name,
42                         "( in1[i], in2[i], in3[i] );\n"
43                         "}\n" };
44 
45     const char *c3[] = {
46         "#pragma OPENCL EXTENSION cl_khr_fp64 : enable\n",
47         "__kernel void math_kernel",
48         sizeNames[vectorSize],
49         "( __global double* out, __global double* in, __global double* in2, "
50         "__global double* in3)\n"
51         "{\n"
52         "   size_t i = get_global_id(0);\n"
53         "   if( i + 1 < get_global_size(0) )\n"
54         "   {\n"
55         "       double3 d0 = vload3( 0, in + 3 * i );\n"
56         "       double3 d1 = vload3( 0, in2 + 3 * i );\n"
57         "       double3 d2 = vload3( 0, in3 + 3 * i );\n"
58         "       d0 = ",
59         name,
60         "( d0, d1, d2 );\n"
61         "       vstore3( d0, 0, out + 3*i );\n"
62         "   }\n"
63         "   else\n"
64         "   {\n"
65         "       size_t parity = i & 1;   // Figure out how many elements are "
66         "left over after BUFFER_SIZE % (3*sizeof(float)). Assume power of two "
67         "buffer size \n"
68         "       double3 d0;\n"
69         "       double3 d1;\n"
70         "       double3 d2;\n"
71         "       switch( parity )\n"
72         "       {\n"
73         "           case 1:\n"
74         "               d0 = (double3)( in[3*i], NAN, NAN ); \n"
75         "               d1 = (double3)( in2[3*i], NAN, NAN ); \n"
76         "               d2 = (double3)( in3[3*i], NAN, NAN ); \n"
77         "               break;\n"
78         "           case 0:\n"
79         "               d0 = (double3)( in[3*i], in[3*i+1], NAN ); \n"
80         "               d1 = (double3)( in2[3*i], in2[3*i+1], NAN ); \n"
81         "               d2 = (double3)( in3[3*i], in3[3*i+1], NAN ); \n"
82         "               break;\n"
83         "       }\n"
84         "       d0 = ",
85         name,
86         "( d0, d1, d2 );\n"
87         "       switch( parity )\n"
88         "       {\n"
89         "           case 0:\n"
90         "               out[3*i+1] = d0.y; \n"
91         "               // fall through\n"
92         "           case 1:\n"
93         "               out[3*i] = d0.x; \n"
94         "               break;\n"
95         "       }\n"
96         "   }\n"
97         "}\n"
98     };
99 
100     const char **kern = c;
101     size_t kernSize = sizeof(c) / sizeof(c[0]);
102 
103     if (sizeValues[vectorSize] == 3)
104     {
105         kern = c3;
106         kernSize = sizeof(c3) / sizeof(c3[0]);
107     }
108 
109     char testName[32];
110     snprintf(testName, sizeof(testName) - 1, "math_kernel%s",
111              sizeNames[vectorSize]);
112 
113     return MakeKernel(kern, (cl_uint)kernSize, testName, k, p, relaxedMode);
114 }
115 
116 typedef struct BuildKernelInfo
117 {
118     cl_uint offset; // the first vector size to build
119     cl_kernel *kernels;
120     cl_program *programs;
121     const char *nameInCode;
122     bool relaxedMode; // Whether to build with -cl-fast-relaxed-math.
123 } BuildKernelInfo;
124 
BuildKernelFn(cl_uint job_id,cl_uint thread_id UNUSED,void * p)125 static cl_int BuildKernelFn(cl_uint job_id, cl_uint thread_id UNUSED, void *p)
126 {
127     BuildKernelInfo *info = (BuildKernelInfo *)p;
128     cl_uint i = info->offset + job_id;
129     return BuildKernel(info->nameInCode, i, info->kernels + i,
130                        info->programs + i, info->relaxedMode);
131 }
132 
TestFunc_mad_Double(const Func * f,MTdata d,bool relaxedMode)133 int TestFunc_mad_Double(const Func *f, MTdata d, bool relaxedMode)
134 {
135     int error;
136     cl_program programs[VECTOR_SIZE_COUNT];
137     cl_kernel kernels[VECTOR_SIZE_COUNT];
138     float maxError = 0.0f;
139     double maxErrorVal = 0.0f;
140     double maxErrorVal2 = 0.0f;
141     double maxErrorVal3 = 0.0f;
142     uint64_t step = getTestStep(sizeof(double), BUFFER_SIZE);
143 
144     logFunctionInfo(f->name, sizeof(cl_double), relaxedMode);
145 
146     // Init the kernels
147     {
148         BuildKernelInfo build_info = { gMinVectorSizeIndex, kernels, programs,
149                                        f->nameInCode, relaxedMode };
150         if ((error = ThreadPool_Do(BuildKernelFn,
151                                    gMaxVectorSizeIndex - gMinVectorSizeIndex,
152                                    &build_info)))
153             return error;
154     }
155 
156     for (uint64_t i = 0; i < (1ULL << 32); i += step)
157     {
158         // Init input array
159         double *p = (double *)gIn;
160         double *p2 = (double *)gIn2;
161         double *p3 = (double *)gIn3;
162         for (size_t j = 0; j < BUFFER_SIZE / sizeof(double); j++)
163         {
164             p[j] = DoubleFromUInt32(genrand_int32(d));
165             p2[j] = DoubleFromUInt32(genrand_int32(d));
166             p3[j] = DoubleFromUInt32(genrand_int32(d));
167         }
168 
169         if ((error = clEnqueueWriteBuffer(gQueue, gInBuffer, CL_FALSE, 0,
170                                           BUFFER_SIZE, gIn, 0, NULL, NULL)))
171         {
172             vlog_error("\n*** Error %d in clEnqueueWriteBuffer ***\n", error);
173             return error;
174         }
175 
176         if ((error = clEnqueueWriteBuffer(gQueue, gInBuffer2, CL_FALSE, 0,
177                                           BUFFER_SIZE, gIn2, 0, NULL, NULL)))
178         {
179             vlog_error("\n*** Error %d in clEnqueueWriteBuffer2 ***\n", error);
180             return error;
181         }
182 
183         if ((error = clEnqueueWriteBuffer(gQueue, gInBuffer3, CL_FALSE, 0,
184                                           BUFFER_SIZE, gIn3, 0, NULL, NULL)))
185         {
186             vlog_error("\n*** Error %d in clEnqueueWriteBuffer3 ***\n", error);
187             return error;
188         }
189 
190         // write garbage into output arrays
191         for (auto j = gMinVectorSizeIndex; j < gMaxVectorSizeIndex; j++)
192         {
193             uint32_t pattern = 0xffffdead;
194             memset_pattern4(gOut[j], &pattern, BUFFER_SIZE);
195             if ((error =
196                      clEnqueueWriteBuffer(gQueue, gOutBuffer[j], CL_FALSE, 0,
197                                           BUFFER_SIZE, gOut[j], 0, NULL, NULL)))
198             {
199                 vlog_error("\n*** Error %d in clEnqueueWriteBuffer2(%d) ***\n",
200                            error, j);
201                 goto exit;
202             }
203         }
204 
205         // Run the kernels
206         for (auto j = gMinVectorSizeIndex; j < gMaxVectorSizeIndex; j++)
207         {
208             size_t vectorSize = sizeof(cl_double) * sizeValues[j];
209             size_t localCount = (BUFFER_SIZE + vectorSize - 1)
210                 / vectorSize; // BUFFER_SIZE / vectorSize  rounded up
211             if ((error = clSetKernelArg(kernels[j], 0, sizeof(gOutBuffer[j]),
212                                         &gOutBuffer[j])))
213             {
214                 LogBuildError(programs[j]);
215                 goto exit;
216             }
217             if ((error = clSetKernelArg(kernels[j], 1, sizeof(gInBuffer),
218                                         &gInBuffer)))
219             {
220                 LogBuildError(programs[j]);
221                 goto exit;
222             }
223             if ((error = clSetKernelArg(kernels[j], 2, sizeof(gInBuffer2),
224                                         &gInBuffer2)))
225             {
226                 LogBuildError(programs[j]);
227                 goto exit;
228             }
229             if ((error = clSetKernelArg(kernels[j], 3, sizeof(gInBuffer3),
230                                         &gInBuffer3)))
231             {
232                 LogBuildError(programs[j]);
233                 goto exit;
234             }
235 
236             if ((error =
237                      clEnqueueNDRangeKernel(gQueue, kernels[j], 1, NULL,
238                                             &localCount, NULL, 0, NULL, NULL)))
239             {
240                 vlog_error("FAILED -- could not execute kernel\n");
241                 goto exit;
242             }
243         }
244 
245         // Get that moving
246         if ((error = clFlush(gQueue))) vlog("clFlush failed\n");
247 
248         // Calculate the correctly rounded reference result
249         double *r = (double *)gOut_Ref;
250         double *s = (double *)gIn;
251         double *s2 = (double *)gIn2;
252         double *s3 = (double *)gIn3;
253         for (size_t j = 0; j < BUFFER_SIZE / sizeof(double); j++)
254             r[j] = (double)f->dfunc.f_fff(s[j], s2[j], s3[j]);
255 
256         // Read the data back
257         for (auto j = gMinVectorSizeIndex; j < gMaxVectorSizeIndex; j++)
258         {
259             if ((error =
260                      clEnqueueReadBuffer(gQueue, gOutBuffer[j], CL_TRUE, 0,
261                                          BUFFER_SIZE, gOut[j], 0, NULL, NULL)))
262             {
263                 vlog_error("ReadArray failed %d\n", error);
264                 goto exit;
265             }
266         }
267 
268         if (gSkipCorrectnessTesting) break;
269 
270         // Verify data -- No verification possible.
271         // MAD is a random number generator.
272         if (0 == (i & 0x0fffffff))
273         {
274             vlog(".");
275             fflush(stdout);
276         }
277     }
278 
279     if (!gSkipCorrectnessTesting)
280     {
281         if (gWimpyMode)
282             vlog("Wimp pass");
283         else
284             vlog("passed");
285 
286         vlog("\t%8.2f @ {%a, %a, %a}", maxError, maxErrorVal, maxErrorVal2,
287              maxErrorVal3);
288     }
289 
290     vlog("\n");
291 
292 exit:
293     // Release
294     for (auto k = gMinVectorSizeIndex; k < gMaxVectorSizeIndex; k++)
295     {
296         clReleaseKernel(kernels[k]);
297         clReleaseProgram(programs[k]);
298     }
299 
300     return error;
301 }
302