blob: 7f22af2a1cfe083081f5e988847ff038217a1cf9 [file] [log] [blame]
//
// 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 "function_list.h"
#include "test_functions.h"
#include "utility.h"
#include <climits>
#include <cstring>
static int BuildKernelDouble(const char *name, int vectorSize, cl_kernel *k,
cl_program *p, bool relaxedMode)
{
const char *c[] = { "#pragma OPENCL EXTENSION cl_khr_fp64 : enable\n",
"__kernel void math_kernel",
sizeNames[vectorSize],
"( __global double",
sizeNames[vectorSize],
"* out, __global int",
sizeNames[vectorSize],
"* out2, __global double",
sizeNames[vectorSize],
"* in1, __global double",
sizeNames[vectorSize],
"* in2 )\n"
"{\n"
" size_t i = get_global_id(0);\n"
" out[i] = ",
name,
"( in1[i], in2[i], out2 + i );\n"
"}\n" };
const char *c3[] = {
"#pragma OPENCL EXTENSION cl_khr_fp64 : enable\n",
"__kernel void math_kernel",
sizeNames[vectorSize],
"( __global double* out, __global int* out2, __global double* in, "
"__global double* in2)\n"
"{\n"
" size_t i = get_global_id(0);\n"
" if( i + 1 < get_global_size(0) )\n"
" {\n"
" double3 d0 = vload3( 0, in + 3 * i );\n"
" double3 d1 = vload3( 0, in2 + 3 * i );\n"
" int3 i0 = 0xdeaddead;\n"
" d0 = ",
name,
"( d0, d1, &i0 );\n"
" vstore3( d0, 0, out + 3*i );\n"
" vstore3( i0, 0, out2 + 3*i );\n"
" }\n"
" else\n"
" {\n"
" size_t parity = i & 1; // Figure out how many elements are "
"left over after BUFFER_SIZE % (3*sizeof(float)). Assume power of two "
"buffer size \n"
" double3 d0;\n"
" double3 d1;\n"
" int3 i0 = 0xdeaddead;\n"
" switch( parity )\n"
" {\n"
" case 1:\n"
" d0 = (double3)( in[3*i], NAN, NAN ); \n"
" d1 = (double3)( in2[3*i], NAN, NAN ); \n"
" break;\n"
" case 0:\n"
" d0 = (double3)( in[3*i], in[3*i+1], NAN ); \n"
" d1 = (double3)( in2[3*i], in2[3*i+1], NAN ); \n"
" break;\n"
" }\n"
" d0 = ",
name,
"( d0, d1, &i0 );\n"
" switch( parity )\n"
" {\n"
" case 0:\n"
" out[3*i+1] = d0.y; \n"
" out2[3*i+1] = i0.y; \n"
" // fall through\n"
" case 1:\n"
" out[3*i] = d0.x; \n"
" out2[3*i] = i0.x; \n"
" break;\n"
" }\n"
" }\n"
"}\n"
};
const char **kern = c;
size_t kernSize = sizeof(c) / sizeof(c[0]);
if (sizeValues[vectorSize] == 3)
{
kern = c3;
kernSize = sizeof(c3) / sizeof(c3[0]);
}
char testName[32];
snprintf(testName, sizeof(testName) - 1, "math_kernel%s",
sizeNames[vectorSize]);
return MakeKernel(kern, (cl_uint)kernSize, testName, k, p, relaxedMode);
}
typedef struct BuildKernelInfo
{
cl_uint offset; // the first vector size to build
cl_kernel *kernels;
cl_program *programs;
const char *nameInCode;
bool relaxedMode; // Whether to build with -cl-fast-relaxed-math.
} BuildKernelInfo;
static cl_int BuildKernel_DoubleFn(cl_uint job_id, cl_uint thread_id UNUSED,
void *p)
{
BuildKernelInfo *info = (BuildKernelInfo *)p;
cl_uint i = info->offset + job_id;
return BuildKernelDouble(info->nameInCode, i, info->kernels + i,
info->programs + i, info->relaxedMode);
}
typedef struct ComputeReferenceInfoD_
{
const double *x;
const double *y;
double *r;
int *i;
long double (*f_ffpI)(long double, long double, int *);
cl_uint lim;
cl_uint count;
} ComputeReferenceInfoD;
static cl_int ReferenceD(cl_uint jid, cl_uint tid, void *userInfo)
{
ComputeReferenceInfoD *cri = (ComputeReferenceInfoD *)userInfo;
cl_uint lim = cri->lim;
cl_uint count = cri->count;
cl_uint off = jid * count;
const double *x = cri->x + off;
const double *y = cri->y + off;
double *r = cri->r + off;
int *i = cri->i + off;
long double (*f)(long double, long double, int *) = cri->f_ffpI;
cl_uint j;
if (off + count > lim) count = lim - off;
Force64BitFPUPrecision();
for (j = 0; j < count; ++j)
r[j] = (double)f((long double)x[j], (long double)y[j], i + j);
return CL_SUCCESS;
}
int TestFunc_DoubleI_Double_Double(const Func *f, MTdata d, bool relaxedMode)
{
uint64_t i;
uint32_t j, k;
int error;
cl_program programs[VECTOR_SIZE_COUNT];
cl_kernel kernels[VECTOR_SIZE_COUNT];
float maxError = 0.0f;
int64_t maxError2 = 0;
int ftz = f->ftz || gForceFTZ;
double maxErrorVal = 0.0f;
double maxErrorVal2 = 0.0f;
size_t bufferSize = (gWimpyMode) ? gWimpyBufferSize : BUFFER_SIZE;
uint64_t step = getTestStep(sizeof(double), bufferSize);
logFunctionInfo(f->name, sizeof(cl_double), relaxedMode);
cl_uint threadCount = GetThreadCount();
Force64BitFPUPrecision();
int testingRemquo = !strcmp(f->name, "remquo");
// Init the kernels
{
BuildKernelInfo build_info = { gMinVectorSizeIndex, kernels, programs,
f->nameInCode, relaxedMode };
if ((error = ThreadPool_Do(BuildKernel_DoubleFn,
gMaxVectorSizeIndex - gMinVectorSizeIndex,
&build_info)))
return error;
}
for (i = 0; i < (1ULL << 32); i += step)
{
// Init input array
double *p = (double *)gIn;
double *p2 = (double *)gIn2;
for (j = 0; j < bufferSize / sizeof(double); j++)
{
p[j] = DoubleFromUInt32(genrand_int32(d));
p2[j] = DoubleFromUInt32(genrand_int32(d));
}
if ((error = clEnqueueWriteBuffer(gQueue, gInBuffer, CL_TRUE, 0,
bufferSize, gIn, 0, NULL, NULL)))
{
vlog_error("\n*** Error %d in clEnqueueWriteBuffer ***\n", error);
return error;
}
if ((error = clEnqueueWriteBuffer(gQueue, gInBuffer2, CL_TRUE, 0,
bufferSize, gIn2, 0, NULL, NULL)))
{
vlog_error("\n*** Error %d in clEnqueueWriteBuffer2 ***\n", error);
return error;
}
// write garbage into output arrays
for (j = gMinVectorSizeIndex; j < gMaxVectorSizeIndex; j++)
{
uint32_t pattern = 0xffffdead;
memset_pattern4(gOut[j], &pattern, bufferSize);
if ((error =
clEnqueueWriteBuffer(gQueue, gOutBuffer[j], CL_TRUE, 0,
bufferSize, gOut[j], 0, NULL, NULL)))
{
vlog_error("\n*** Error %d in clEnqueueWriteBuffer2(%d) ***\n",
error, j);
goto exit;
}
memset_pattern4(gOut2[j], &pattern, bufferSize);
if ((error =
clEnqueueWriteBuffer(gQueue, gOutBuffer2[j], CL_TRUE, 0,
bufferSize, gOut2[j], 0, NULL, NULL)))
{
vlog_error("\n*** Error %d in clEnqueueWriteBuffer2b(%d) ***\n",
error, j);
goto exit;
}
}
// Run the kernels
for (j = gMinVectorSizeIndex; j < gMaxVectorSizeIndex; j++)
{
size_t vectorSize = sizeof(cl_double) * sizeValues[j];
size_t localCount = (bufferSize + vectorSize - 1)
/ vectorSize; // bufferSize / vectorSize rounded up
if ((error = clSetKernelArg(kernels[j], 0, sizeof(gOutBuffer[j]),
&gOutBuffer[j])))
{
LogBuildError(programs[j]);
goto exit;
}
if ((error = clSetKernelArg(kernels[j], 1, sizeof(gOutBuffer2[j]),
&gOutBuffer2[j])))
{
LogBuildError(programs[j]);
goto exit;
}
if ((error = clSetKernelArg(kernels[j], 2, sizeof(gInBuffer),
&gInBuffer)))
{
LogBuildError(programs[j]);
goto exit;
}
if ((error = clSetKernelArg(kernels[j], 3, sizeof(gInBuffer2),
&gInBuffer2)))
{
LogBuildError(programs[j]);
goto exit;
}
if ((error =
clEnqueueNDRangeKernel(gQueue, kernels[j], 1, NULL,
&localCount, NULL, 0, NULL, NULL)))
{
vlog_error("FAILED -- could not execute kernel\n");
goto exit;
}
}
// Get that moving
if ((error = clFlush(gQueue))) vlog("clFlush failed\n");
// Calculate the correctly rounded reference result
double *s = (double *)gIn;
double *s2 = (double *)gIn2;
if (threadCount > 1)
{
ComputeReferenceInfoD cri;
cri.x = s;
cri.y = s2;
cri.r = (double *)gOut_Ref;
cri.i = (int *)gOut_Ref2;
cri.f_ffpI = f->dfunc.f_ffpI;
cri.lim = bufferSize / sizeof(double);
cri.count = (cri.lim + threadCount - 1) / threadCount;
ThreadPool_Do(ReferenceD, threadCount, &cri);
}
else
{
double *r = (double *)gOut_Ref;
int *r2 = (int *)gOut_Ref2;
for (j = 0; j < bufferSize / sizeof(double); j++)
r[j] = (double)f->dfunc.f_ffpI(s[j], s2[j], r2 + j);
}
// Read the data back
for (j = gMinVectorSizeIndex; j < gMaxVectorSizeIndex; j++)
{
if ((error =
clEnqueueReadBuffer(gQueue, gOutBuffer[j], CL_TRUE, 0,
bufferSize, gOut[j], 0, NULL, NULL)))
{
vlog_error("ReadArray failed %d\n", error);
goto exit;
}
if ((error =
clEnqueueReadBuffer(gQueue, gOutBuffer2[j], CL_TRUE, 0,
bufferSize, gOut2[j], 0, NULL, NULL)))
{
vlog_error("ReadArray2 failed %d\n", error);
goto exit;
}
}
if (gSkipCorrectnessTesting) break;
// Verify data
uint64_t *t = (uint64_t *)gOut_Ref;
int32_t *t2 = (int32_t *)gOut_Ref2;
for (j = 0; j < bufferSize / sizeof(double); j++)
{
for (k = gMinVectorSizeIndex; k < gMaxVectorSizeIndex; k++)
{
uint64_t *q = (uint64_t *)gOut[k];
int32_t *q2 = (int32_t *)gOut2[k];
// Check for exact match to correctly rounded result
if (t[j] == q[j] && t2[j] == q2[j]) continue;
// Check for paired NaNs
if ((t[j] & 0x7fffffffffffffffUL) > 0x7ff0000000000000UL
&& (q[j] & 0x7fffffffffffffffUL) > 0x7ff0000000000000UL
&& t2[j] == q2[j])
continue;
double test = ((double *)q)[j];
int correct2 = INT_MIN;
long double correct = f->dfunc.f_ffpI(s[j], s2[j], &correct2);
float err = Bruteforce_Ulp_Error_Double(test, correct);
int64_t iErr;
// in case of remquo, we only care about the sign and last
// seven bits of integer as per the spec.
if (testingRemquo)
iErr = (long long)(q2[j] & 0x0000007f)
- (long long)(correct2 & 0x0000007f);
else
iErr = (long long)q2[j] - (long long)correct2;
// For remquo, if y = 0, x is infinite, or either is NaN
// then the standard either neglects to say what is returned
// in iptr or leaves it undefined or implementation defined.
int iptrUndefined = fabs(((double *)gIn)[j]) == INFINITY
|| ((double *)gIn2)[j] == 0.0 || isnan(((double *)gIn2)[j])
|| isnan(((double *)gIn)[j]);
if (iptrUndefined) iErr = 0;
int fail = !(fabsf(err) <= f->double_ulps && iErr == 0);
if (ftz && fail)
{
// retry per section 6.5.3.2
if (IsDoubleResultSubnormal(correct, f->double_ulps))
{
fail = fail && !(test == 0.0f && iErr == 0);
if (!fail) err = 0.0f;
}
// retry per section 6.5.3.3
if (IsDoubleSubnormal(s[j]))
{
int correct3i, correct4i;
long double correct3 =
f->dfunc.f_ffpI(0.0, s2[j], &correct3i);
long double correct4 =
f->dfunc.f_ffpI(-0.0, s2[j], &correct4i);
float err2 =
Bruteforce_Ulp_Error_Double(test, correct3);
float err3 =
Bruteforce_Ulp_Error_Double(test, correct4);
int64_t iErr3 = (long long)q2[j] - (long long)correct3i;
int64_t iErr4 = (long long)q2[j] - (long long)correct4i;
fail = fail
&& ((!(fabsf(err2) <= f->double_ulps && iErr3 == 0))
&& (!(fabsf(err3) <= f->double_ulps
&& iErr4 == 0)));
if (fabsf(err2) < fabsf(err)) err = err2;
if (fabsf(err3) < fabsf(err)) err = err3;
if (llabs(iErr3) < llabs(iErr)) iErr = iErr3;
if (llabs(iErr4) < llabs(iErr)) iErr = iErr4;
// retry per section 6.5.3.4
if (IsDoubleResultSubnormal(correct2, f->double_ulps)
|| IsDoubleResultSubnormal(correct3,
f->double_ulps))
{
fail = fail
&& !(test == 0.0f
&& (iErr3 == 0 || iErr4 == 0));
if (!fail) err = 0.0f;
}
// try with both args as zero
if (IsDoubleSubnormal(s2[j]))
{
int correct7i, correct8i;
correct3 = f->dfunc.f_ffpI(0.0, 0.0, &correct3i);
correct4 = f->dfunc.f_ffpI(-0.0, 0.0, &correct4i);
long double correct7 =
f->dfunc.f_ffpI(0.0, -0.0, &correct7i);
long double correct8 =
f->dfunc.f_ffpI(-0.0, -0.0, &correct8i);
err2 = Bruteforce_Ulp_Error_Double(test, correct3);
err3 = Bruteforce_Ulp_Error_Double(test, correct4);
float err4 =
Bruteforce_Ulp_Error_Double(test, correct7);
float err5 =
Bruteforce_Ulp_Error_Double(test, correct8);
iErr3 = (long long)q2[j] - (long long)correct3i;
iErr4 = (long long)q2[j] - (long long)correct4i;
int64_t iErr7 =
(long long)q2[j] - (long long)correct7i;
int64_t iErr8 =
(long long)q2[j] - (long long)correct8i;
fail = fail
&& ((!(fabsf(err2) <= f->double_ulps
&& iErr3 == 0))
&& (!(fabsf(err3) <= f->double_ulps
&& iErr4 == 0))
&& (!(fabsf(err4) <= f->double_ulps
&& iErr7 == 0))
&& (!(fabsf(err5) <= f->double_ulps
&& iErr8 == 0)));
if (fabsf(err2) < fabsf(err)) err = err2;
if (fabsf(err3) < fabsf(err)) err = err3;
if (fabsf(err4) < fabsf(err)) err = err4;
if (fabsf(err5) < fabsf(err)) err = err5;
if (llabs(iErr3) < llabs(iErr)) iErr = iErr3;
if (llabs(iErr4) < llabs(iErr)) iErr = iErr4;
if (llabs(iErr7) < llabs(iErr)) iErr = iErr7;
if (llabs(iErr8) < llabs(iErr)) iErr = iErr8;
// retry per section 6.5.3.4
if (IsDoubleResultSubnormal(correct3,
f->double_ulps)
|| IsDoubleResultSubnormal(correct4,
f->double_ulps)
|| IsDoubleResultSubnormal(correct7,
f->double_ulps)
|| IsDoubleResultSubnormal(correct8,
f->double_ulps))
{
fail = fail
&& !(test == 0.0f
&& (iErr3 == 0 || iErr4 == 0
|| iErr7 == 0 || iErr8 == 0));
if (!fail) err = 0.0f;
}
}
}
else if (IsDoubleSubnormal(s2[j]))
{
int correct3i, correct4i;
long double correct3 =
f->dfunc.f_ffpI(s[j], 0.0, &correct3i);
long double correct4 =
f->dfunc.f_ffpI(s[j], -0.0, &correct4i);
float err2 =
Bruteforce_Ulp_Error_Double(test, correct3);
float err3 =
Bruteforce_Ulp_Error_Double(test, correct4);
int64_t iErr3 = (long long)q2[j] - (long long)correct3i;
int64_t iErr4 = (long long)q2[j] - (long long)correct4i;
fail = fail
&& ((!(fabsf(err2) <= f->double_ulps && iErr3 == 0))
&& (!(fabsf(err3) <= f->double_ulps
&& iErr4 == 0)));
if (fabsf(err2) < fabsf(err)) err = err2;
if (fabsf(err3) < fabsf(err)) err = err3;
if (llabs(iErr3) < llabs(iErr)) iErr = iErr3;
if (llabs(iErr4) < llabs(iErr)) iErr = iErr4;
// retry per section 6.5.3.4
if (IsDoubleResultSubnormal(correct2, f->double_ulps)
|| IsDoubleResultSubnormal(correct3,
f->double_ulps))
{
fail = fail
&& !(test == 0.0f
&& (iErr3 == 0 || iErr4 == 0));
if (!fail) err = 0.0f;
}
}
}
if (fabsf(err) > maxError)
{
maxError = fabsf(err);
maxErrorVal = s[j];
}
if (llabs(iErr) > maxError2)
{
maxError2 = llabs(iErr);
maxErrorVal2 = s[j];
}
if (fail)
{
vlog_error(
"\nERROR: %sD%s: {%f, %lld} ulp error at {%.13la, "
"%.13la} ({ 0x%16.16llx, 0x%16.16llx}): *{%.13la, "
"%d} ({ 0x%16.16llx, 0x%8.8x}) vs. {%.13la, %d} ({ "
"0x%16.16llx, 0x%8.8x})\n",
f->name, sizeNames[k], err, iErr, ((double *)gIn)[j],
((double *)gIn2)[j], ((cl_ulong *)gIn)[j],
((cl_ulong *)gIn2)[j], ((double *)gOut_Ref)[j],
((int *)gOut_Ref2)[j], ((cl_ulong *)gOut_Ref)[j],
((cl_uint *)gOut_Ref2)[j], test, q2[j],
((cl_ulong *)q)[j], ((cl_uint *)q2)[j]);
error = -1;
goto exit;
}
}
}
if (0 == (i & 0x0fffffff))
{
if (gVerboseBruteForce)
{
vlog("base:%14u step:%10zu bufferSize:%10zd \n", i, step,
bufferSize);
}
else
{
vlog(".");
}
fflush(stdout);
}
}
if (!gSkipCorrectnessTesting)
{
if (gWimpyMode)
vlog("Wimp pass");
else
vlog("passed");
}
if (gMeasureTimes)
{
// Init input array
double *p = (double *)gIn;
for (j = 0; j < bufferSize / sizeof(cl_double); j++)
p[j] = DoubleFromUInt32(genrand_int32(d));
if ((error = clEnqueueWriteBuffer(gQueue, gInBuffer, CL_TRUE, 0,
bufferSize, gIn, 0, NULL, NULL)))
{
vlog_error("\n*** Error %d in clEnqueueWriteBuffer ***\n", error);
return error;
}
if ((error = clEnqueueWriteBuffer(gQueue, gInBuffer2, CL_TRUE, 0,
bufferSize, gIn2, 0, NULL, NULL)))
{
vlog_error("\n*** Error %d in clEnqueueWriteBuffer2 ***\n", error);
return error;
}
// Run the kernels
for (j = gMinVectorSizeIndex; j < gMaxVectorSizeIndex; j++)
{
size_t vectorSize = sizeof(cl_double) * sizeValues[j];
size_t localCount = (bufferSize + vectorSize - 1)
/ vectorSize; // bufferSize / vectorSize rounded up
if ((error = clSetKernelArg(kernels[j], 0, sizeof(gOutBuffer[j]),
&gOutBuffer[j])))
{
LogBuildError(programs[j]);
goto exit;
}
if ((error = clSetKernelArg(kernels[j], 1, sizeof(gOutBuffer2[j]),
&gOutBuffer2[j])))
{
LogBuildError(programs[j]);
goto exit;
}
if ((error = clSetKernelArg(kernels[j], 2, sizeof(gInBuffer),
&gInBuffer)))
{
LogBuildError(programs[j]);
goto exit;
}
if ((error = clSetKernelArg(kernels[j], 3, sizeof(gInBuffer2),
&gInBuffer2)))
{
LogBuildError(programs[j]);
goto exit;
}
double sum = 0.0;
double bestTime = INFINITY;
for (k = 0; k < PERF_LOOP_COUNT; k++)
{
uint64_t startTime = GetTime();
if ((error = clEnqueueNDRangeKernel(gQueue, kernels[j], 1, NULL,
&localCount, NULL, 0, NULL,
NULL)))
{
vlog_error("FAILED -- could not execute kernel\n");
goto exit;
}
// Make sure OpenCL is done
if ((error = clFinish(gQueue)))
{
vlog_error("Error %d at clFinish\n", error);
goto exit;
}
uint64_t endTime = GetTime();
double time = SubtractTime(endTime, startTime);
sum += time;
if (time < bestTime) bestTime = time;
}
if (gReportAverageTimes) bestTime = sum / PERF_LOOP_COUNT;
double clocksPerOp = bestTime * (double)gDeviceFrequency
* gComputeDevices * gSimdSize * 1e6
/ (bufferSize / sizeof(double));
vlog_perf(clocksPerOp, LOWER_IS_BETTER, "clocks / element", "%sD%s",
f->name, sizeNames[j]);
}
for (; j < gMaxVectorSizeIndex; j++) vlog("\t -- ");
}
if (!gSkipCorrectnessTesting)
vlog("\t{%8.2f, %lld} @ %a", maxError, maxError2, maxErrorVal);
vlog("\n");
exit:
// Release
for (k = gMinVectorSizeIndex; k < gMaxVectorSizeIndex; k++)
{
clReleaseKernel(kernels[k]);
clReleaseProgram(programs[k]);
}
return error;
}