blob: 5bcf15ab77d430872479ab482b36ad8f1d87497b [file] [log] [blame]
// Copyright 2016 The SwiftShader Authors. All Rights Reserved.
//
// 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 "Assert.hpp"
#include "Coroutine.hpp"
#include "Print.hpp"
#include "Reactor.hpp"
#include "gtest/gtest.h"
#include <array>
#include <cmath>
#include <filesystem>
#include <fstream>
#include <thread>
#include <tuple>
using namespace rr;
using float4 = float[4];
using int4 = int[4];
static std::string testName()
{
auto info = ::testing::UnitTest::GetInstance()->current_test_info();
return std::string{ info->test_suite_name() } + "_" + info->name();
}
int reference(int *p, int y)
{
int x = p[-1];
int z = 4;
for(int i = 0; i < 10; i++)
{
z += (2 << i) - (i / 3);
}
int sum = x + y + z;
return sum;
}
TEST(ReactorUnitTests, Sample)
{
FunctionT<int(int *, int)> function;
{
Pointer<Int> p = function.Arg<0>();
Int x = p[-1];
Int y = function.Arg<1>();
Int z = 4;
For(Int i = 0, i < 10, i++)
{
z += (2 << i) - (i / 3);
}
Float4 v;
v.z = As<Float>(z);
z = As<Int>(Float(Float4(v.xzxx).y));
Int sum = x + y + z;
Return(sum);
}
auto routine = function(testName().c_str());
int one[2] = { 1, 0 };
int result = routine(&one[1], 2);
EXPECT_EQ(result, reference(&one[1], 2));
}
// This test demonstrates the use of a 'trampoline', where a routine calls
// a static function which then generates another routine during the execution
// of the first routine. Also note the code generated for the second routine
// depends on a parameter passed to the first routine.
TEST(ReactorUnitTests, Trampoline)
{
using SecondaryFunc = int(int, int);
static auto generateSecondary = [](int upDown) {
FunctionT<SecondaryFunc> secondary;
{
Int x = secondary.Arg<0>();
Int y = secondary.Arg<1>();
Int r;
if(upDown > 0)
{
r = x + y;
}
else if(upDown < 0)
{
r = x - y;
}
else
{
r = 0;
}
Return(r);
}
static auto routine = secondary((testName() + "_secondary").c_str());
return routine.getEntry();
};
using SecondaryGeneratorFunc = SecondaryFunc *(*)(int);
SecondaryGeneratorFunc secondaryGenerator = (SecondaryGeneratorFunc)generateSecondary;
using PrimaryFunc = int(int, int, int);
FunctionT<PrimaryFunc> primary;
{
Int x = primary.Arg<0>();
Int y = primary.Arg<1>();
Int z = primary.Arg<2>();
Pointer<Byte> secondary = Call(secondaryGenerator, z);
Int r = Call<SecondaryFunc>(secondary, x, y);
Return(r);
}
auto routine = primary((testName() + "_primary").c_str());
int result = routine(100, 20, -3);
EXPECT_EQ(result, 80);
}
TEST(ReactorUnitTests, Uninitialized)
{
#if __has_feature(memory_sanitizer)
// Building the static C++ code with MemorySanitizer enabled does not
// automatically enable MemorySanitizer instrumentation for Reactor
// routines. False positives can also be prevented by unpoisoning all
// memory writes. This Pragma ensures proper instrumentation is enabled.
Pragma(MemorySanitizerInstrumentation, true);
#endif
FunctionT<int()> function;
{
Int a;
Int z = 4;
Int q;
Int c;
Int p;
Bool b;
q += q;
If(b)
{
c = p;
}
Return(a + z + q + c);
}
auto routine = function(testName().c_str());
if(!__has_feature(memory_sanitizer))
{
int result = routine();
EXPECT_EQ(result, result); // Anything is fine, just don't crash
}
else
{
// Optimizations may turn the conditional If() in the Reactor code
// into a conditional move or arithmetic operations, which would not
// trigger a MemorySanitizer error. However, in that case the equals
// operator below should trigger it before the abort is reached.
EXPECT_DEATH(
{
int result = routine();
if(result == 0) abort();
},
"MemorySanitizer: use-of-uninitialized-value");
}
Pragma(MemorySanitizerInstrumentation, false);
}
TEST(ReactorUnitTests, Unreachable)
{
FunctionT<int(int)> function;
{
Int a = function.Arg<0>();
Int z = 4;
Return(a + z);
// Code beyond this point is unreachable but should not cause any
// compilation issues.
z += a;
}
auto routine = function(testName().c_str());
int result = routine(16);
EXPECT_EQ(result, 20);
}
// Stopping in the middle of a `Function<>` is supported and should not affect
// subsequent complete ones.
TEST(ReactorUnitTests, UnfinishedFunction)
{
do
{
FunctionT<int(int)> function;
{
Int a = function.Arg<0>();
Int z = 4;
if((true)) break; // Terminate do-while early.
Return(a + z);
}
} while(true);
FunctionT<int(int)> function;
{
Int a = function.Arg<0>();
Int z = 4;
Return(a - z);
}
auto routine = function(testName().c_str());
int result = routine(16);
EXPECT_EQ(result, 12);
}
// Deriving from `Function<>` and using Reactor variables as members can be a
// convenient way to 'name' function arguments and compose complex functions
// with helper methods. This test checks the interactions between the lifetime
// of the `Function<>` and the variables belonging to the derived class.
struct FunctionMembers : FunctionT<int(int)>
{
FunctionMembers()
: level(Arg<0>())
{
For(Int i = 0, i < 3, i++)
{
pourSomeMore();
}
Return(level);
}
void pourSomeMore()
{
level += 2;
}
Int level;
};
TEST(ReactorUnitTests, FunctionMembers)
{
FunctionMembers function;
auto routine = function(testName().c_str());
int result = routine(3);
EXPECT_EQ(result, 9);
}
// This test excercises modifying the value of a local variable through a
// pointer to it.
TEST(ReactorUnitTests, VariableAddress)
{
FunctionT<int(int)> function;
{
Int a = function.Arg<0>();
Int z = 0;
Pointer<Int> p = &z;
*p = 4;
Return(a + z);
}
auto routine = function(testName().c_str());
int result = routine(16);
EXPECT_EQ(result, 20);
}
// This test exercises taking the address of a local varible at the end of a
// loop and modifying its value through the pointer in the second iteration.
TEST(ReactorUnitTests, LateVariableAddress)
{
FunctionT<int(void)> function;
{
Pointer<Int> p = nullptr;
Int a = 0;
While(a == 0)
{
If(p != Pointer<Int>(nullptr))
{
*p = 1;
}
p = &a;
}
Return(a);
}
auto routine = function(testName().c_str());
int result = routine();
EXPECT_EQ(result, 1);
}
// This test checks that the value of a local variable which has been modified
// though a pointer is correct at the point before its address is (statically)
// obtained.
TEST(ReactorUnitTests, LoadAfterIndirectStore)
{
FunctionT<int(void)> function;
{
Pointer<Int> p = nullptr;
Int a = 0;
Int b = 0;
While(a == 0)
{
If(p != Pointer<Int>(nullptr))
{
*p = 1;
}
// `a` must be loaded from memory here, despite not statically knowing
// yet that its address will be taken below.
b = a + 5;
p = &a;
}
Return(b);
}
auto routine = function(testName().c_str());
int result = routine();
EXPECT_EQ(result, 6);
}
// This test checks that variables statically accessed after a Return statement
// are still loaded, modified, and stored correctly.
TEST(ReactorUnitTests, LoopAfterReturn)
{
FunctionT<int(void)> function;
{
Int min = 100;
Int max = 200;
If(min > max)
{
Return(5);
}
While(min < max)
{
min++;
}
Return(7);
}
auto routine = function(testName().c_str());
int result = routine();
EXPECT_EQ(result, 7);
}
TEST(ReactorUnitTests, ConstantPointer)
{
int c = 44;
FunctionT<int()> function;
{
Int x = *Pointer<Int>(ConstantPointer(&c));
Return(x);
}
auto routine = function(testName().c_str());
int result = routine();
EXPECT_EQ(result, 44);
}
// This test excercises the Optimizer::eliminateLoadsFollowingSingleStore() optimization pass.
// The three load operations for `y` should get eliminated.
TEST(ReactorUnitTests, EliminateLoadsFollowingSingleStore)
{
FunctionT<int(int)> function;
{
Int x = function.Arg<0>();
Int y;
Int z;
// This branch materializes the variables.
If(x != 0) // TODO(b/179922668): Support If(x)
{
y = x;
z = y + y + y;
}
Return(z);
}
Nucleus::setOptimizerCallback([](const Nucleus::OptimizerReport *report) {
EXPECT_EQ(report->allocas, 2);
EXPECT_EQ(report->loads, 2);
EXPECT_EQ(report->stores, 2);
});
auto routine = function(testName().c_str());
int result = routine(11);
EXPECT_EQ(result, 33);
}
// This test excercises the Optimizer::propagateAlloca() optimization pass.
// The pointer variable should not get stored to / loaded from memory.
TEST(ReactorUnitTests, PropagateAlloca)
{
FunctionT<int(int)> function;
{
Int b = function.Arg<0>();
Int a = 22;
Pointer<Int> p;
// This branch materializes both `a` and `p`, and ensures single basic block
// optimizations don't also eliminate the pointer store and load.
If(b != 0) // TODO(b/179922668): Support If(b)
{
p = &a;
}
Return(Int(*p)); // TODO(b/179694472): Support Return(*p)
}
Nucleus::setOptimizerCallback([](const Nucleus::OptimizerReport *report) {
EXPECT_EQ(report->allocas, 1);
EXPECT_EQ(report->loads, 1);
EXPECT_EQ(report->stores, 1);
});
auto routine = function(testName().c_str());
int result = routine(true);
EXPECT_EQ(result, 22);
}
// Corner case for Optimizer::propagateAlloca(). It should not replace loading of `p`
// with the addres of `a`, since it also got the address of `b` assigned.
TEST(ReactorUnitTests, PointerToPointer)
{
FunctionT<int()> function;
{
Int a = 444;
Int b = 555;
Pointer<Int> p = &a;
Pointer<Pointer<Int>> pp = &p;
p = &b;
Return(Int(*Pointer<Int>(*pp))); // TODO(b/179694472): Support **pp
}
auto routine = function(testName().c_str());
int result = routine();
EXPECT_EQ(result, 555);
}
// Corner case for Optimizer::propagateAlloca(). It should not replace loading of `p[i]`
// with any of the addresses of the `a`, `b`, or `c`.
TEST(ReactorUnitTests, ArrayOfPointersToLocals)
{
FunctionT<int(int)> function;
{
Int i = function.Arg<0>();
Int a = 111;
Int b = 222;
Int c = 333;
Array<Pointer<Int>, 3> p;
p[0] = &a;
p[1] = &b;
p[2] = &c;
Return(Int(*Pointer<Int>(p[i]))); // TODO(b/179694472): Support *p[i]
}
auto routine = function(testName().c_str());
int result = routine(1);
EXPECT_EQ(result, 222);
}
TEST(ReactorUnitTests, ModifyLocalThroughPointer)
{
FunctionT<int(void)> function;
{
Int a = 1;
Pointer<Int> p = &a;
Pointer<Pointer<Int>> pp = &p;
Pointer<Int> q = *pp;
*q = 3;
Return(a);
}
auto routine = function(testName().c_str());
int result = routine();
EXPECT_EQ(result, 3);
}
TEST(ReactorUnitTests, ScalarReplacementOfArray)
{
FunctionT<int(void)> function;
{
Array<Int, 2> a;
a[0] = 1;
a[1] = 2;
Return(a[0] + a[1]);
}
auto routine = function(testName().c_str());
int result = routine();
EXPECT_EQ(result, 3);
}
TEST(ReactorUnitTests, CArray)
{
FunctionT<int(void)> function;
{
Int a[2];
a[0] = 1;
a[1] = 2;
auto x = a[0];
a[0] = a[1];
a[1] = x;
Return(a[0] + a[1]);
}
auto routine = function(testName().c_str());
int result = routine();
EXPECT_EQ(result, 3);
}
// SRoA should replace the array elements with scalars, which in turn enables
// eliminating all loads and stores.
TEST(ReactorUnitTests, ReactorArray)
{
FunctionT<int(void)> function;
{
Array<Int, 2> a;
a[0] = 1;
a[1] = 2;
Int x = a[0];
a[0] = a[1];
a[1] = x;
Return(a[0] + a[1]);
}
Nucleus::setOptimizerCallback([](const Nucleus::OptimizerReport *report) {
EXPECT_EQ(report->allocas, 0);
EXPECT_EQ(report->loads, 0);
EXPECT_EQ(report->stores, 0);
});
auto routine = function(testName().c_str());
int result = routine();
EXPECT_EQ(result, 3);
}
// Excercises the optimizeSingleBasicBlockLoadsStores optimization pass.
TEST(ReactorUnitTests, StoresInMultipleBlocks)
{
FunctionT<int(int)> function;
{
Int b = function.Arg<0>();
Int a = 13;
If(b != 0) // TODO(b/179922668): Support If(b)
{
a = 4;
a = a + 3;
}
Else
{
a = 6;
a = a + 5;
}
Return(a);
}
Nucleus::setOptimizerCallback([](const Nucleus::OptimizerReport *report) {
EXPECT_EQ(report->allocas, 1);
EXPECT_EQ(report->loads, 1);
EXPECT_EQ(report->stores, 3);
});
auto routine = function(testName().c_str());
int result = routine(true);
EXPECT_EQ(result, 7);
}
// This is similar to the LoadAfterIndirectStore test except that the indirect
// store is preceded by a direct store. The subsequent load should not be replaced
// by the value written by the direct store.
TEST(ReactorUnitTests, StoreBeforeIndirectStore)
{
FunctionT<int(int)> function;
{
// Int b = function.Arg<0>();
Int b;
Pointer<Int> p = &b;
Int a = 13;
For(Int i = 0, i < 2, i++)
{
a = 10;
*p = 4;
// This load of `a` should not be replaced by the 10 written above, since
// in the second iteration `p` points to `a` and writes 4.
b = a;
p = &a;
}
Return(b);
}
auto routine = function(testName().c_str());
int result = routine(true);
EXPECT_EQ(result, 4);
}
TEST(ReactorUnitTests, AssertTrue)
{
FunctionT<int()> function;
{
Int a = 3;
Int b = 5;
Assert(a < b);
Return(a + b);
}
auto routine = function(testName().c_str());
int result = routine();
EXPECT_EQ(result, 8);
}
TEST(ReactorUnitTests, AssertFalse)
{
FunctionT<int()> function;
{
Int a = 3;
Int b = 5;
Assert(a == b);
Return(a + b);
}
auto routine = function(testName().c_str());
#ifndef NDEBUG
# if !defined(__APPLE__)
const char *stderrRegex = "AssertFalse"; // stderr should contain the assert's expression, file:line, and function
# else
const char *stderrRegex = ""; // TODO(b/156389924): On macOS an stderr redirect can cause googletest to fail the capture
# endif
EXPECT_DEATH(
{
int result = routine();
EXPECT_NE(result, result); // We should never reach this
},
stderrRegex);
#else
int result = routine();
EXPECT_EQ(result, 8);
#endif
}
TEST(ReactorUnitTests, SubVectorLoadStore)
{
FunctionT<int(void *, void *)> function;
{
Pointer<Byte> in = function.Arg<0>();
Pointer<Byte> out = function.Arg<1>();
*Pointer<Int4>(out + 16 * 0) = *Pointer<Int4>(in + 16 * 0);
*Pointer<Short4>(out + 16 * 1) = *Pointer<Short4>(in + 16 * 1);
*Pointer<Byte8>(out + 16 * 2) = *Pointer<Byte8>(in + 16 * 2);
*Pointer<Byte4>(out + 16 * 3) = *Pointer<Byte4>(in + 16 * 3);
*Pointer<Short2>(out + 16 * 4) = *Pointer<Short2>(in + 16 * 4);
Return(0);
}
auto routine = function(testName().c_str());
int8_t in[16 * 5] = { 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16,
17, 18, 19, 20, 21, 22, 23, 24, 0, 0, 0, 0, 0, 0, 0, 0,
25, 26, 27, 28, 29, 30, 31, 32, 0, 0, 0, 0, 0, 0, 0, 0,
33, 34, 35, 36, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
37, 38, 39, 40, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 };
int8_t out[16 * 5] = { -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
-1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
-1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
-1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
-1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1 };
routine(in, out);
for(int row = 0; row < 5; row++)
{
for(int col = 0; col < 16; col++)
{
int i = row * 16 + col;
if(in[i] == 0)
{
EXPECT_EQ(out[i], -1) << "Row " << row << " column " << col << " not left untouched.";
}
else
{
EXPECT_EQ(out[i], in[i]) << "Row " << row << " column " << col << " not equal to input.";
}
}
}
}
TEST(ReactorUnitTests, VectorConstant)
{
FunctionT<int(void *)> function;
{
Pointer<Byte> out = function.Arg<0>();
*Pointer<Int4>(out + 16 * 0) = Int4(0x04030201, 0x08070605, 0x0C0B0A09, 0x100F0E0D);
*Pointer<Short4>(out + 16 * 1) = Short4(0x1211, 0x1413, 0x1615, 0x1817);
*Pointer<Byte8>(out + 16 * 2) = Byte8(0x19, 0x1A, 0x1B, 0x1C, 0x1D, 0x1E, 0x1F, 0x20);
*Pointer<Int2>(out + 16 * 3) = Int2(0x24232221, 0x28272625);
Return(0);
}
auto routine = function(testName().c_str());
int8_t out[16 * 4] = { -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
-1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
-1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
-1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1 };
int8_t exp[16 * 4] = { 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16,
17, 18, 19, 20, 21, 22, 23, 24, -1, -1, -1, -1, -1, -1, -1, -1,
25, 26, 27, 28, 29, 30, 31, 32, -1, -1, -1, -1, -1, -1, -1, -1,
33, 34, 35, 36, 37, 38, 39, 40, -1, -1, -1, -1, -1, -1, -1, -1 };
routine(out);
for(int row = 0; row < 4; row++)
{
for(int col = 0; col < 16; col++)
{
int i = row * 16 + col;
EXPECT_EQ(out[i], exp[i]);
}
}
}
TEST(ReactorUnitTests, Concatenate)
{
FunctionT<int(void *)> function;
{
Pointer<Byte> out = function.Arg<0>();
*Pointer<Int4>(out + 16 * 0) = Int4(Int2(0x04030201, 0x08070605), Int2(0x0C0B0A09, 0x100F0E0D));
*Pointer<Short8>(out + 16 * 1) = Short8(Short4(0x0201, 0x0403, 0x0605, 0x0807), Short4(0x0A09, 0x0C0B, 0x0E0D, 0x100F));
Return(0);
}
auto routine = function(testName().c_str());
int8_t ref[16 * 5] = { 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16,
1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16 };
int8_t out[16 * 5] = { -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1,
-1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1 };
routine(out);
for(int row = 0; row < 2; row++)
{
for(int col = 0; col < 16; col++)
{
int i = row * 16 + col;
EXPECT_EQ(out[i], ref[i]) << "Row " << row << " column " << col << " not equal to reference.";
}
}
}
TEST(ReactorUnitTests, Cast)
{
FunctionT<void(void *)> function;
{
Pointer<Byte> out = function.Arg<0>();
Int4 c = Int4(0x01020304, 0x05060708, 0x09101112, 0x13141516);
*Pointer<Short4>(out + 16 * 0) = Short4(c);
*Pointer<Byte4>(out + 16 * 1 + 0) = Byte4(c);
*Pointer<Byte4>(out + 16 * 1 + 4) = Byte4(As<Byte8>(c));
*Pointer<Byte4>(out + 16 * 1 + 8) = Byte4(As<Short4>(c));
}
auto routine = function(testName().c_str());
int out[2][4];
memset(&out, 0, sizeof(out));
routine(&out);
EXPECT_EQ(out[0][0], 0x07080304);
EXPECT_EQ(out[0][1], 0x15161112);
EXPECT_EQ(out[1][0], 0x16120804);
EXPECT_EQ(out[1][1], 0x01020304);
EXPECT_EQ(out[1][2], 0x06080204);
}
static uint16_t swizzleCode4(int i)
{
auto x = (i >> 0) & 0x03;
auto y = (i >> 2) & 0x03;
auto z = (i >> 4) & 0x03;
auto w = (i >> 6) & 0x03;
return static_cast<uint16_t>((x << 12) | (y << 8) | (z << 4) | (w << 0));
}
TEST(ReactorUnitTests, Swizzle4)
{
FunctionT<void(void *)> function;
{
Pointer<Byte> out = function.Arg<0>();
for(int i = 0; i < 256; i++)
{
*Pointer<Float4>(out + 16 * i) = Swizzle(Float4(1.0f, 2.0f, 3.0f, 4.0f), swizzleCode4(i));
}
for(int i = 0; i < 256; i++)
{
*Pointer<Float4>(out + 16 * (256 + i)) = ShuffleLowHigh(Float4(1.0f, 2.0f, 3.0f, 4.0f), Float4(5.0f, 6.0f, 7.0f, 8.0f), swizzleCode4(i));
}
*Pointer<Float4>(out + 16 * (512 + 0)) = UnpackLow(Float4(1.0f, 2.0f, 3.0f, 4.0f), Float4(5.0f, 6.0f, 7.0f, 8.0f));
*Pointer<Float4>(out + 16 * (512 + 1)) = UnpackHigh(Float4(1.0f, 2.0f, 3.0f, 4.0f), Float4(5.0f, 6.0f, 7.0f, 8.0f));
*Pointer<Int2>(out + 16 * (512 + 2)) = UnpackLow(Short4(1, 2, 3, 4), Short4(5, 6, 7, 8));
*Pointer<Int2>(out + 16 * (512 + 3)) = UnpackHigh(Short4(1, 2, 3, 4), Short4(5, 6, 7, 8));
*Pointer<Short4>(out + 16 * (512 + 4)) = UnpackLow(Byte8(1, 2, 3, 4, 5, 6, 7, 8), Byte8(9, 10, 11, 12, 13, 14, 15, 16));
*Pointer<Short4>(out + 16 * (512 + 5)) = UnpackHigh(Byte8(1, 2, 3, 4, 5, 6, 7, 8), Byte8(9, 10, 11, 12, 13, 14, 15, 16));
for(int i = 0; i < 256; i++)
{
*Pointer<Short4>(out + 16 * (512 + 6) + (8 * i)) =
Swizzle(Short4(1, 2, 3, 4), swizzleCode4(i));
}
for(int i = 0; i < 256; i++)
{
*Pointer<Int4>(out + 16 * (512 + 6 + i) + (8 * 256)) =
Swizzle(Int4(1, 2, 3, 4), swizzleCode4(i));
}
}
auto routine = function(testName().c_str());
struct
{
float f[256 + 256 + 2][4];
int i[388][4];
} out;
memset(&out, 0, sizeof(out));
routine(&out);
for(int i = 0; i < 256; i++)
{
EXPECT_EQ(out.f[i][0], float((i >> 0) & 0x03) + 1.0f);
EXPECT_EQ(out.f[i][1], float((i >> 2) & 0x03) + 1.0f);
EXPECT_EQ(out.f[i][2], float((i >> 4) & 0x03) + 1.0f);
EXPECT_EQ(out.f[i][3], float((i >> 6) & 0x03) + 1.0f);
}
for(int i = 0; i < 256; i++)
{
EXPECT_EQ(out.f[256 + i][0], float((i >> 0) & 0x03) + 1.0f);
EXPECT_EQ(out.f[256 + i][1], float((i >> 2) & 0x03) + 1.0f);
EXPECT_EQ(out.f[256 + i][2], float((i >> 4) & 0x03) + 5.0f);
EXPECT_EQ(out.f[256 + i][3], float((i >> 6) & 0x03) + 5.0f);
}
EXPECT_EQ(out.f[512 + 0][0], 1.0f);
EXPECT_EQ(out.f[512 + 0][1], 5.0f);
EXPECT_EQ(out.f[512 + 0][2], 2.0f);
EXPECT_EQ(out.f[512 + 0][3], 6.0f);
EXPECT_EQ(out.f[512 + 1][0], 3.0f);
EXPECT_EQ(out.f[512 + 1][1], 7.0f);
EXPECT_EQ(out.f[512 + 1][2], 4.0f);
EXPECT_EQ(out.f[512 + 1][3], 8.0f);
EXPECT_EQ(out.i[0][0], 0x00050001);
EXPECT_EQ(out.i[0][1], 0x00060002);
EXPECT_EQ(out.i[0][2], 0x00000000);
EXPECT_EQ(out.i[0][3], 0x00000000);
EXPECT_EQ(out.i[1][0], 0x00070003);
EXPECT_EQ(out.i[1][1], 0x00080004);
EXPECT_EQ(out.i[1][2], 0x00000000);
EXPECT_EQ(out.i[1][3], 0x00000000);
EXPECT_EQ(out.i[2][0], 0x0A020901);
EXPECT_EQ(out.i[2][1], 0x0C040B03);
EXPECT_EQ(out.i[2][2], 0x00000000);
EXPECT_EQ(out.i[2][3], 0x00000000);
EXPECT_EQ(out.i[3][0], 0x0E060D05);
EXPECT_EQ(out.i[3][1], 0x10080F07);
EXPECT_EQ(out.i[3][2], 0x00000000);
EXPECT_EQ(out.i[3][3], 0x00000000);
for(int i = 0; i < 256; i++)
{
EXPECT_EQ(out.i[4 + i / 2][0 + (i % 2) * 2] & 0xFFFF,
((i >> 0) & 0x03) + 1);
EXPECT_EQ(out.i[4 + i / 2][0 + (i % 2) * 2] >> 16,
((i >> 2) & 0x03) + 1);
EXPECT_EQ(out.i[4 + i / 2][1 + (i % 2) * 2] & 0xFFFF,
((i >> 4) & 0x03) + 1);
EXPECT_EQ(out.i[4 + i / 2][1 + (i % 2) * 2] >> 16,
((i >> 6) & 0x03) + 1);
}
for(int i = 0; i < 256; i++)
{
EXPECT_EQ(out.i[132 + i][0], ((i >> 0) & 0x03) + 1);
EXPECT_EQ(out.i[132 + i][1], ((i >> 2) & 0x03) + 1);
EXPECT_EQ(out.i[132 + i][2], ((i >> 4) & 0x03) + 1);
EXPECT_EQ(out.i[132 + i][3], ((i >> 6) & 0x03) + 1);
}
}
TEST(ReactorUnitTests, Swizzle)
{
FunctionT<void(void *)> function;
{
Pointer<Byte> out = function.Arg<0>();
Int4 c = Int4(0x01020304, 0x05060708, 0x09101112, 0x13141516);
*Pointer<Byte16>(out + 16 * 0) = Swizzle(As<Byte16>(c), 0xFEDCBA9876543210ull);
*Pointer<Byte8>(out + 16 * 1) = Swizzle(As<Byte8>(c), 0x76543210u);
*Pointer<UShort8>(out + 16 * 2) = Swizzle(As<UShort8>(c), 0x76543210u);
}
auto routine = function(testName().c_str());
int out[3][4];
memset(&out, 0, sizeof(out));
routine(&out);
EXPECT_EQ(out[0][0], 0x16151413);
EXPECT_EQ(out[0][1], 0x12111009);
EXPECT_EQ(out[0][2], 0x08070605);
EXPECT_EQ(out[0][3], 0x04030201);
EXPECT_EQ(out[1][0], 0x08070605);
EXPECT_EQ(out[1][1], 0x04030201);
EXPECT_EQ(out[2][0], 0x15161314);
EXPECT_EQ(out[2][1], 0x11120910);
EXPECT_EQ(out[2][2], 0x07080506);
EXPECT_EQ(out[2][3], 0x03040102);
}
TEST(ReactorUnitTests, Shuffle)
{
// |select| is [0aaa:0bbb:0ccc:0ddd] where |aaa|, |bbb|, |ccc|
// and |ddd| are 7-bit selection indices. For a total (1 << 12)
// possibilities.
const int kSelectRange = 1 << 12;
// Unfortunately, testing the whole kSelectRange results in a test
// that is far too slow to run, because LLVM spends exponentially more
// time optimizing the function below as the number of test cases
// increases.
//
// To work-around the problem, only test a subset of the range by
// skipping every kRangeIncrement value.
//
// Set this value to 1 if you want to test the whole implementation,
// which will take a little less than 2 minutes on a fast workstation.
//
// The default value here takes about 1390ms, which is a little more than
// what the Swizzle test takes (993 ms) on my machine. A non-power-of-2
// value ensures a better spread over possible values.
const int kRangeIncrement = 11;
auto rangeIndexToSelect = [](int i) {
return static_cast<unsigned short>(
(((i >> 9) & 7) << 0) |
(((i >> 6) & 7) << 4) |
(((i >> 3) & 7) << 8) |
(((i >> 0) & 7) << 12));
};
FunctionT<int(void *)> function;
{
Pointer<Byte> out = function.Arg<0>();
for(int i = 0; i < kSelectRange; i += kRangeIncrement)
{
unsigned short select = rangeIndexToSelect(i);
*Pointer<Float4>(out + 16 * i) = Shuffle(Float4(1.0f, 2.0f, 3.0f, 4.0f),
Float4(5.0f, 6.0f, 7.0f, 8.0f),
select);
*Pointer<Int4>(out + (kSelectRange + i) * 16) = Shuffle(Int4(10, 11, 12, 13),
Int4(14, 15, 16, 17),
select);
*Pointer<UInt4>(out + (2 * kSelectRange + i) * 16) = Shuffle(UInt4(100, 101, 102, 103),
UInt4(104, 105, 106, 107),
select);
}
Return(0);
}
auto routine = function(testName().c_str());
struct
{
float f[kSelectRange][4];
int i[kSelectRange][4];
unsigned u[kSelectRange][4];
} out;
memset(&out, 0, sizeof(out));
routine(&out);
for(int i = 0; i < kSelectRange; i += kRangeIncrement)
{
EXPECT_EQ(out.f[i][0], float(1.0f + (i & 7)));
EXPECT_EQ(out.f[i][1], float(1.0f + ((i >> 3) & 7)));
EXPECT_EQ(out.f[i][2], float(1.0f + ((i >> 6) & 7)));
EXPECT_EQ(out.f[i][3], float(1.0f + ((i >> 9) & 7)));
}
for(int i = 0; i < kSelectRange; i += kRangeIncrement)
{
EXPECT_EQ(out.i[i][0], int(10 + (i & 7)));
EXPECT_EQ(out.i[i][1], int(10 + ((i >> 3) & 7)));
EXPECT_EQ(out.i[i][2], int(10 + ((i >> 6) & 7)));
EXPECT_EQ(out.i[i][3], int(10 + ((i >> 9) & 7)));
}
for(int i = 0; i < kSelectRange; i += kRangeIncrement)
{
EXPECT_EQ(out.u[i][0], unsigned(100 + (i & 7)));
EXPECT_EQ(out.u[i][1], unsigned(100 + ((i >> 3) & 7)));
EXPECT_EQ(out.u[i][2], unsigned(100 + ((i >> 6) & 7)));
EXPECT_EQ(out.u[i][3], unsigned(100 + ((i >> 9) & 7)));
}
}
TEST(ReactorUnitTests, Broadcast)
{
FunctionT<int()> function;
{
Int4 i = 2;
Int j = 3 + i.x;
Int4 k = i * 7;
Return(k.z - j);
}
auto routine = function(testName().c_str());
int result = routine();
EXPECT_EQ(result, 9);
}
TEST(ReactorUnitTests, Branching)
{
FunctionT<int()> function;
{
Int x = 0;
For(Int i = 0, i < 8, i++)
{
If(i < 2)
{
x += 1;
}
Else If(i < 4)
{
x += 10;
}
Else If(i < 6)
{
x += 100;
}
Else
{
x += 1000;
}
For(Int i = 0, i < 5, i++)
x += 10000;
}
For(Int i = 0, i < 10, i++) for(int i = 0; i < 10; i++)
For(Int i = 0, i < 10, i++)
{
x += 1000000;
}
For(Int i = 0, i < 2, i++)
If(x == 1000402222)
{
If(x != 1000402222)
x += 1000000000;
}
Else
x = -5;
Return(x);
}
auto routine = function(testName().c_str());
int result = routine();
EXPECT_EQ(result, 1000402222);
}
TEST(ReactorUnitTests, FMulAdd)
{
Function<Void(Pointer<Float4>, Pointer<Float4>, Pointer<Float4>, Pointer<Float4>)> function;
{
Pointer<Float4> r = function.Arg<0>();
Pointer<Float4> x = function.Arg<1>();
Pointer<Float4> y = function.Arg<2>();
Pointer<Float4> z = function.Arg<3>();
*r = MulAdd(*x, *y, *z);
}
auto routine = function(testName().c_str());
auto callable = (void (*)(float4 *, float4 *, float4 *, float4 *))routine->getEntry();
float x[] = { 0.0f, 2.0f, 4.0f, 1.00000011920929f };
float y[] = { 0.0f, 3.0f, 0.0f, 53400708.0f };
float z[] = { 0.0f, 0.0f, 7.0f, -53400708.0f };
for(size_t i = 0; i < std::size(x); i++)
{
float4 x_in = { x[i], x[i], x[i], x[i] };
float4 y_in = { y[i], y[i], y[i], y[i] };
float4 z_in = { z[i], z[i], z[i], z[i] };
float4 r_out;
callable(&r_out, &x_in, &y_in, &z_in);
// Possible results
float fma = fmaf(x[i], y[i], z[i]);
float mul_add = x[i] * y[i] + z[i];
// If the backend and the CPU support FMA instructions, we assume MulAdd to use
// them. Otherwise it may behave as a multiplication followed by an addition.
if(rr::Caps::fmaIsFast())
{
EXPECT_FLOAT_EQ(r_out[0], fma);
}
else if(r_out[0] != fma)
{
EXPECT_FLOAT_EQ(r_out[0], mul_add);
}
}
}
TEST(ReactorUnitTests, FMA)
{
Function<Void(Pointer<Float4>, Pointer<Float4>, Pointer<Float4>, Pointer<Float4>)> function;
{
Pointer<Float4> r = function.Arg<0>();
Pointer<Float4> x = function.Arg<1>();
Pointer<Float4> y = function.Arg<2>();
Pointer<Float4> z = function.Arg<3>();
*r = FMA(*x, *y, *z);
}
auto routine = function(testName().c_str());
auto callable = (void (*)(float4 *, float4 *, float4 *, float4 *))routine->getEntry();
float x[] = { 0.0f, 2.0f, 4.0f, 1.00000011920929f };
float y[] = { 0.0f, 3.0f, 0.0f, 53400708.0f };
float z[] = { 0.0f, 0.0f, 7.0f, -53400708.0f };
for(size_t i = 0; i < std::size(x); i++)
{
float4 x_in = { x[i], x[i], x[i], x[i] };
float4 y_in = { y[i], y[i], y[i], y[i] };
float4 z_in = { z[i], z[i], z[i], z[i] };
float4 r_out;
callable(&r_out, &x_in, &y_in, &z_in);
float expected = fmaf(x[i], y[i], z[i]);
EXPECT_FLOAT_EQ(r_out[0], expected);
}
}
TEST(ReactorUnitTests, FAbs)
{
Function<Void(Pointer<Float4>, Pointer<Float4>)> function;
{
Pointer<Float4> x = function.Arg<0>();
Pointer<Float4> y = function.Arg<1>();
*y = Abs(*x);
}
auto routine = function(testName().c_str());
auto callable = (void (*)(float4 *, float4 *))routine->getEntry();
float input[] = { 1.0f, -1.0f, -0.0f, 0.0f };
for(float x : input)
{
float4 v_in = { x, x, x, x };
float4 v_out;
callable(&v_in, &v_out);
float expected = fabs(x);
EXPECT_FLOAT_EQ(v_out[0], expected);
}
}
TEST(ReactorUnitTests, Abs)
{
Function<Void(Pointer<Int4>, Pointer<Int4>)> function;
{
Pointer<Int4> x = function.Arg<0>();
Pointer<Int4> y = function.Arg<1>();
*y = Abs(*x);
}
auto routine = function(testName().c_str());
auto callable = (void (*)(int4 *, int4 *))routine->getEntry();
int input[] = { 1, -1, 0, (int)0x80000000 };
for(int x : input)
{
int4 v_in = { x, x, x, x };
int4 v_out;
callable(&v_in, &v_out);
float expected = abs(x);
EXPECT_EQ(v_out[0], expected);
}
}
TEST(ReactorUnitTests, MinMax)
{
FunctionT<int(void *)> function;
{
Pointer<Byte> out = function.Arg<0>();
*Pointer<Float4>(out + 16 * 0) = Min(Float4(1.0f, 0.0f, -0.0f, +0.0f), Float4(0.0f, 1.0f, +0.0f, -0.0f));
*Pointer<Float4>(out + 16 * 1) = Max(Float4(1.0f, 0.0f, -0.0f, +0.0f), Float4(0.0f, 1.0f, +0.0f, -0.0f));
*Pointer<Int4>(out + 16 * 2) = Min(Int4(1, 0, -1, -0), Int4(0, 1, 0, +0));
*Pointer<Int4>(out + 16 * 3) = Max(Int4(1, 0, -1, -0), Int4(0, 1, 0, +0));
*Pointer<UInt4>(out + 16 * 4) = Min(UInt4(1, 0, -1, -0), UInt4(0, 1, 0, +0));
*Pointer<UInt4>(out + 16 * 5) = Max(UInt4(1, 0, -1, -0), UInt4(0, 1, 0, +0));
*Pointer<Short4>(out + 16 * 6) = Min(Short4(1, 0, -1, -0), Short4(0, 1, 0, +0));
*Pointer<Short4>(out + 16 * 7) = Max(Short4(1, 0, -1, -0), Short4(0, 1, 0, +0));
*Pointer<UShort4>(out + 16 * 8) = Min(UShort4(1, 0, -1, -0), UShort4(0, 1, 0, +0));
*Pointer<UShort4>(out + 16 * 9) = Max(UShort4(1, 0, -1, -0), UShort4(0, 1, 0, +0));
Return(0);
}
auto routine = function(testName().c_str());
unsigned int out[10][4];
memset(&out, 0, sizeof(out));
routine(&out);
EXPECT_EQ(out[0][0], 0x00000000u);
EXPECT_EQ(out[0][1], 0x00000000u);
EXPECT_EQ(out[0][2], 0x00000000u);
EXPECT_EQ(out[0][3], 0x80000000u);
EXPECT_EQ(out[1][0], 0x3F800000u);
EXPECT_EQ(out[1][1], 0x3F800000u);
EXPECT_EQ(out[1][2], 0x00000000u);
EXPECT_EQ(out[1][3], 0x80000000u);
EXPECT_EQ(out[2][0], 0x00000000u);
EXPECT_EQ(out[2][1], 0x00000000u);
EXPECT_EQ(out[2][2], 0xFFFFFFFFu);
EXPECT_EQ(out[2][3], 0x00000000u);
EXPECT_EQ(out[3][0], 0x00000001u);
EXPECT_EQ(out[3][1], 0x00000001u);
EXPECT_EQ(out[3][2], 0x00000000u);
EXPECT_EQ(out[3][3], 0x00000000u);
EXPECT_EQ(out[4][0], 0x00000000u);
EXPECT_EQ(out[4][1], 0x00000000u);
EXPECT_EQ(out[4][2], 0x00000000u);
EXPECT_EQ(out[4][3], 0x00000000u);
EXPECT_EQ(out[5][0], 0x00000001u);
EXPECT_EQ(out[5][1], 0x00000001u);
EXPECT_EQ(out[5][2], 0xFFFFFFFFu);
EXPECT_EQ(out[5][3], 0x00000000u);
EXPECT_EQ(out[6][0], 0x00000000u);
EXPECT_EQ(out[6][1], 0x0000FFFFu);
EXPECT_EQ(out[6][2], 0x00000000u);
EXPECT_EQ(out[6][3], 0x00000000u);
EXPECT_EQ(out[7][0], 0x00010001u);
EXPECT_EQ(out[7][1], 0x00000000u);
EXPECT_EQ(out[7][2], 0x00000000u);
EXPECT_EQ(out[7][3], 0x00000000u);
EXPECT_EQ(out[8][0], 0x00000000u);
EXPECT_EQ(out[8][1], 0x00000000u);
EXPECT_EQ(out[8][2], 0x00000000u);
EXPECT_EQ(out[8][3], 0x00000000u);
EXPECT_EQ(out[9][0], 0x00010001u);
EXPECT_EQ(out[9][1], 0x0000FFFFu);
EXPECT_EQ(out[9][2], 0x00000000u);
EXPECT_EQ(out[9][3], 0x00000000u);
}
TEST(ReactorUnitTests, NotNeg)
{
FunctionT<int(void *)> function;
{
Pointer<Byte> out = function.Arg<0>();
*Pointer<Int>(out + 16 * 0) = ~Int(0x55555555);
*Pointer<Short>(out + 16 * 1) = ~Short(0x5555);
*Pointer<Int4>(out + 16 * 2) = ~Int4(0x55555555, 0xAAAAAAAA, 0x00000000, 0xFFFFFFFF);
*Pointer<Short4>(out + 16 * 3) = ~Short4(0x5555, 0xAAAA, 0x0000, 0xFFFF);
*Pointer<Int>(out + 16 * 4) = -Int(0x55555555);
*Pointer<Short>(out + 16 * 5) = -Short(0x5555);
*Pointer<Int4>(out + 16 * 6) = -Int4(0x55555555, 0xAAAAAAAA, 0x00000000, 0xFFFFFFFF);
*Pointer<Short4>(out + 16 * 7) = -Short4(0x5555, 0xAAAA, 0x0000, 0xFFFF);
*Pointer<Float4>(out + 16 * 8) = -Float4(1.0f, -1.0f, 0.0f, -0.0f);
Return(0);
}
auto routine = function(testName().c_str());
unsigned int out[10][4];
memset(&out, 0, sizeof(out));
routine(&out);
EXPECT_EQ(out[0][0], 0xAAAAAAAAu);
EXPECT_EQ(out[0][1], 0x00000000u);
EXPECT_EQ(out[0][2], 0x00000000u);
EXPECT_EQ(out[0][3], 0x00000000u);
EXPECT_EQ(out[1][0], 0x0000AAAAu);
EXPECT_EQ(out[1][1], 0x00000000u);
EXPECT_EQ(out[1][2], 0x00000000u);
EXPECT_EQ(out[1][3], 0x00000000u);
EXPECT_EQ(out[2][0], 0xAAAAAAAAu);
EXPECT_EQ(out[2][1], 0x55555555u);
EXPECT_EQ(out[2][2], 0xFFFFFFFFu);
EXPECT_EQ(out[2][3], 0x00000000u);
EXPECT_EQ(out[3][0], 0x5555AAAAu);
EXPECT_EQ(out[3][1], 0x0000FFFFu);
EXPECT_EQ(out[3][2], 0x00000000u);
EXPECT_EQ(out[3][3], 0x00000000u);
EXPECT_EQ(out[4][0], 0xAAAAAAABu);
EXPECT_EQ(out[4][1], 0x00000000u);
EXPECT_EQ(out[4][2], 0x00000000u);
EXPECT_EQ(out[4][3], 0x00000000u);
EXPECT_EQ(out[5][0], 0x0000AAABu);
EXPECT_EQ(out[5][1], 0x00000000u);
EXPECT_EQ(out[5][2], 0x00000000u);
EXPECT_EQ(out[5][3], 0x00000000u);
EXPECT_EQ(out[6][0], 0xAAAAAAABu);
EXPECT_EQ(out[6][1], 0x55555556u);
EXPECT_EQ(out[6][2], 0x00000000u);
EXPECT_EQ(out[6][3], 0x00000001u);
EXPECT_EQ(out[7][0], 0x5556AAABu);
EXPECT_EQ(out[7][1], 0x00010000u);
EXPECT_EQ(out[7][2], 0x00000000u);
EXPECT_EQ(out[7][3], 0x00000000u);
EXPECT_EQ(out[8][0], 0xBF800000u);
EXPECT_EQ(out[8][1], 0x3F800000u);
EXPECT_EQ(out[8][2], 0x80000000u);
EXPECT_EQ(out[8][3], 0x00000000u);
}
TEST(ReactorUnitTests, RoundInt)
{
FunctionT<int(void *)> function;
{
Pointer<Byte> out = function.Arg<0>();
*Pointer<Int4>(out + 0) = RoundInt(Float4(3.1f, 3.6f, -3.1f, -3.6f));
*Pointer<Int4>(out + 16) = RoundIntClamped(Float4(2147483648.0f, -2147483648.0f, 2147483520, -2147483520));
Return(0);
}
auto routine = function(testName().c_str());
int out[2][4];
memset(&out, 0, sizeof(out));
routine(&out);
EXPECT_EQ(out[0][0], 3);
EXPECT_EQ(out[0][1], 4);
EXPECT_EQ(out[0][2], -3);
EXPECT_EQ(out[0][3], -4);
// x86 returns 0x80000000 for values which cannot be represented in a 32-bit
// integer, but RoundIntClamped() clamps to ensure a positive value for
// positive input. ARM saturates to the largest representable integers.
EXPECT_GE(out[1][0], 2147483520);
EXPECT_LT(out[1][1], -2147483647);
EXPECT_EQ(out[1][2], 2147483520);
EXPECT_EQ(out[1][3], -2147483520);
}
TEST(ReactorUnitTests, FPtoUI)
{
FunctionT<int(void *)> function;
{
Pointer<Byte> out = function.Arg<0>();
*Pointer<UInt>(out + 0) = UInt(Float(0xF0000000u));
*Pointer<UInt>(out + 4) = UInt(Float(0xC0000000u));
*Pointer<UInt>(out + 8) = UInt(Float(0x00000001u));
*Pointer<UInt>(out + 12) = UInt(Float(0xF000F000u));
*Pointer<UInt4>(out + 16) = UInt4(Float4(0xF0000000u, 0x80000000u, 0x00000000u, 0xCCCC0000u));
Return(0);
}
auto routine = function(testName().c_str());
unsigned int out[2][4];
memset(&out, 0, sizeof(out));
routine(&out);
EXPECT_EQ(out[0][0], 0xF0000000u);
EXPECT_EQ(out[0][1], 0xC0000000u);
EXPECT_EQ(out[0][2], 0x00000001u);
EXPECT_EQ(out[0][3], 0xF000F000u);
EXPECT_EQ(out[1][0], 0xF0000000u);
EXPECT_EQ(out[1][1], 0x80000000u);
EXPECT_EQ(out[1][2], 0x00000000u);
EXPECT_EQ(out[1][3], 0xCCCC0000u);
}
TEST(ReactorUnitTests, VectorCompare)
{
FunctionT<int(void *)> function;
{
Pointer<Byte> out = function.Arg<0>();
*Pointer<Int4>(out + 16 * 0) = CmpEQ(Float4(1.0f, 1.0f, -0.0f, +0.0f), Float4(0.0f, 1.0f, +0.0f, -0.0f));
*Pointer<Int4>(out + 16 * 1) = CmpEQ(Int4(1, 0, -1, -0), Int4(0, 1, 0, +0));
*Pointer<Byte8>(out + 16 * 2) = CmpEQ(SByte8(1, 2, 3, 4, 5, 6, 7, 8), SByte8(7, 6, 5, 4, 3, 2, 1, 0));
*Pointer<Int4>(out + 16 * 3) = CmpNLT(Float4(1.0f, 1.0f, -0.0f, +0.0f), Float4(0.0f, 1.0f, +0.0f, -0.0f));
*Pointer<Int4>(out + 16 * 4) = CmpNLT(Int4(1, 0, -1, -0), Int4(0, 1, 0, +0));
*Pointer<Byte8>(out + 16 * 5) = CmpGT(SByte8(1, 2, 3, 4, 5, 6, 7, 8), SByte8(7, 6, 5, 4, 3, 2, 1, 0));
Return(0);
}
auto routine = function(testName().c_str());
unsigned int out[6][4];
memset(&out, 0, sizeof(out));
routine(&out);
EXPECT_EQ(out[0][0], 0x00000000u);
EXPECT_EQ(out[0][1], 0xFFFFFFFFu);
EXPECT_EQ(out[0][2], 0xFFFFFFFFu);
EXPECT_EQ(out[0][3], 0xFFFFFFFFu);
EXPECT_EQ(out[1][0], 0x00000000u);
EXPECT_EQ(out[1][1], 0x00000000u);
EXPECT_EQ(out[1][2], 0x00000000u);
EXPECT_EQ(out[1][3], 0xFFFFFFFFu);
EXPECT_EQ(out[2][0], 0xFF000000u);
EXPECT_EQ(out[2][1], 0x00000000u);
EXPECT_EQ(out[3][0], 0xFFFFFFFFu);
EXPECT_EQ(out[3][1], 0xFFFFFFFFu);
EXPECT_EQ(out[3][2], 0xFFFFFFFFu);
EXPECT_EQ(out[3][3], 0xFFFFFFFFu);
EXPECT_EQ(out[4][0], 0xFFFFFFFFu);
EXPECT_EQ(out[4][1], 0x00000000u);
EXPECT_EQ(out[4][2], 0x00000000u);
EXPECT_EQ(out[4][3], 0xFFFFFFFFu);
EXPECT_EQ(out[5][0], 0x00000000u);
EXPECT_EQ(out[5][1], 0xFFFFFFFFu);
}
TEST(ReactorUnitTests, SaturatedAddAndSubtract)
{
FunctionT<int(void *)> function;
{
Pointer<Byte> out = function.Arg<0>();
*Pointer<Byte8>(out + 8 * 0) =
AddSat(Byte8(1, 2, 3, 4, 5, 6, 7, 8),
Byte8(7, 6, 5, 4, 3, 2, 1, 0));
*Pointer<Byte8>(out + 8 * 1) =
AddSat(Byte8(0xFE, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE),
Byte8(7, 6, 5, 4, 3, 2, 1, 0));
*Pointer<Byte8>(out + 8 * 2) =
SubSat(Byte8(1, 2, 3, 4, 5, 6, 7, 8),
Byte8(7, 6, 5, 4, 3, 2, 1, 0));
*Pointer<SByte8>(out + 8 * 3) =
AddSat(SByte8(1, 2, 3, 4, 5, 6, 7, 8),
SByte8(7, 6, 5, 4, 3, 2, 1, 0));
*Pointer<SByte8>(out + 8 * 4) =
AddSat(SByte8(0x7E, 0x7E, 0x7E, 0x7E, 0x7E, 0x7E, 0x7E, 0x7E),
SByte8(7, 6, 5, 4, 3, 2, 1, 0));
*Pointer<SByte8>(out + 8 * 5) =
AddSat(SByte8(0x81, 0x82, 0x83, 0x84, 0x85, 0x86, 0x87, 0x88),
SByte8(-7, -6, -5, -4, -3, -2, -1, -0));
*Pointer<SByte8>(out + 8 * 6) =
SubSat(SByte8(0x81, 0x82, 0x83, 0x84, 0x85, 0x86, 0x87, 0x88),
SByte8(7, 6, 5, 4, 3, 2, 1, 0));
*Pointer<Short4>(out + 8 * 7) =
AddSat(Short4(1, 2, 3, 4), Short4(3, 2, 1, 0));
*Pointer<Short4>(out + 8 * 8) =
AddSat(Short4(0x7FFE, 0x7FFE, 0x7FFE, 0x7FFE),
Short4(3, 2, 1, 0));
*Pointer<Short4>(out + 8 * 9) =
AddSat(Short4(0x8001, 0x8002, 0x8003, 0x8004),
Short4(-3, -2, -1, -0));
*Pointer<Short4>(out + 8 * 10) =
SubSat(Short4(0x8001, 0x8002, 0x8003, 0x8004),
Short4(3, 2, 1, 0));
*Pointer<UShort4>(out + 8 * 11) =
AddSat(UShort4(1, 2, 3, 4), UShort4(3, 2, 1, 0));
*Pointer<UShort4>(out + 8 * 12) =
AddSat(UShort4(0xFFFE, 0xFFFE, 0xFFFE, 0xFFFE),
UShort4(3, 2, 1, 0));
*Pointer<UShort4>(out + 8 * 13) =
SubSat(UShort4(1, 2, 3, 4), UShort4(3, 2, 1, 0));
Return(0);
}
auto routine = function(testName().c_str());
unsigned int out[14][2];
memset(&out, 0, sizeof(out));
routine(&out);
EXPECT_EQ(out[0][0], 0x08080808u);
EXPECT_EQ(out[0][1], 0x08080808u);
EXPECT_EQ(out[1][0], 0xFFFFFFFFu);
EXPECT_EQ(out[1][1], 0xFEFFFFFFu);
EXPECT_EQ(out[2][0], 0x00000000u);
EXPECT_EQ(out[2][1], 0x08060402u);
EXPECT_EQ(out[3][0], 0x08080808u);
EXPECT_EQ(out[3][1], 0x08080808u);
EXPECT_EQ(out[4][0], 0x7F7F7F7Fu);
EXPECT_EQ(out[4][1], 0x7E7F7F7Fu);
EXPECT_EQ(out[5][0], 0x80808080u);
EXPECT_EQ(out[5][1], 0x88868482u);
EXPECT_EQ(out[6][0], 0x80808080u);
EXPECT_EQ(out[6][1], 0x88868482u);
EXPECT_EQ(out[7][0], 0x00040004u);
EXPECT_EQ(out[7][1], 0x00040004u);
EXPECT_EQ(out[8][0], 0x7FFF7FFFu);
EXPECT_EQ(out[8][1], 0x7FFE7FFFu);
EXPECT_EQ(out[9][0], 0x80008000u);
EXPECT_EQ(out[9][1], 0x80048002u);
EXPECT_EQ(out[10][0], 0x80008000u);
EXPECT_EQ(out[10][1], 0x80048002u);
EXPECT_EQ(out[11][0], 0x00040004u);
EXPECT_EQ(out[11][1], 0x00040004u);
EXPECT_EQ(out[12][0], 0xFFFFFFFFu);
EXPECT_EQ(out[12][1], 0xFFFEFFFFu);
EXPECT_EQ(out[13][0], 0x00000000u);
EXPECT_EQ(out[13][1], 0x00040002u);
}
TEST(ReactorUnitTests, Unpack)
{
FunctionT<int(void *, void *)> function;
{
Pointer<Byte> in = function.Arg<0>();
Pointer<Byte> out = function.Arg<1>();
Byte4 test_byte_a = *Pointer<Byte4>(in + 4 * 0);
Byte4 test_byte_b = *Pointer<Byte4>(in + 4 * 1);
*Pointer<Short4>(out + 8 * 0) =
Unpack(test_byte_a, test_byte_b);
*Pointer<Short4>(out + 8 * 1) = Unpack(test_byte_a);
Return(0);
}
auto routine = function(testName().c_str());
unsigned int in[1][2];
unsigned int out[2][2];
memset(&out, 0, sizeof(out));
in[0][0] = 0xABCDEF12u;
in[0][1] = 0x34567890u;
routine(&in, &out);
EXPECT_EQ(out[0][0], 0x78EF9012u);
EXPECT_EQ(out[0][1], 0x34AB56CDu);
EXPECT_EQ(out[1][0], 0xEFEF1212u);
EXPECT_EQ(out[1][1], 0xABABCDCDu);
}
TEST(ReactorUnitTests, Pack)
{
FunctionT<int(void *)> function;
{
Pointer<Byte> out = function.Arg<0>();
*Pointer<SByte8>(out + 8 * 0) =
PackSigned(Short4(-1, -2, 1, 2),
Short4(3, 4, -3, -4));
*Pointer<Byte8>(out + 8 * 1) =
PackUnsigned(Short4(-1, -2, 1, 2),
Short4(3, 4, -3, -4));
*Pointer<Short8>(out + 8 * 2) =
PackSigned(Int4(-1, -2, 1, 2),
Int4(3, 4, -3, -4));
*Pointer<UShort8>(out + 8 * 4) =
PackUnsigned(Int4(-1, -2, 1, 2),
Int4(3, 4, -3, -4));
Return(0);
}
auto routine = function(testName().c_str());
unsigned int out[6][2];
memset(&out, 0, sizeof(out));
routine(&out);
EXPECT_EQ(out[0][0], 0x0201FEFFu);
EXPECT_EQ(out[0][1], 0xFCFD0403u);
EXPECT_EQ(out[1][0], 0x02010000u);
EXPECT_EQ(out[1][1], 0x00000403u);
EXPECT_EQ(out[2][0], 0xFFFEFFFFu);
EXPECT_EQ(out[2][1], 0x00020001u);
EXPECT_EQ(out[3][0], 0x00040003u);
EXPECT_EQ(out[3][1], 0xFFFCFFFDu);
EXPECT_EQ(out[4][0], 0x00000000u);
EXPECT_EQ(out[4][1], 0x00020001u);
EXPECT_EQ(out[5][0], 0x00040003u);
EXPECT_EQ(out[5][1], 0x00000000u);
}
TEST(ReactorUnitTests, MulHigh)
{
FunctionT<int(void *)> function;
{
Pointer<Byte> out = function.Arg<0>();
*Pointer<Short4>(out + 16 * 0) =
MulHigh(Short4(0x01AA, 0x02DD, 0x03EE, 0xF422),
Short4(0x01BB, 0x02CC, 0x03FF, 0xF411));
*Pointer<UShort4>(out + 16 * 1) =
MulHigh(UShort4(0x01AA, 0x02DD, 0x03EE, 0xF422),
UShort4(0x01BB, 0x02CC, 0x03FF, 0xF411));
*Pointer<Int4>(out + 16 * 2) =
MulHigh(Int4(0x000001AA, 0x000002DD, 0xC8000000, 0xF8000000),
Int4(0x000001BB, 0x84000000, 0x000003EE, 0xD7000000));
*Pointer<UInt4>(out + 16 * 3) =
MulHigh(UInt4(0x000001AAu, 0x000002DDu, 0xC8000000u, 0xD8000000u),
UInt4(0x000001BBu, 0x84000000u, 0x000003EEu, 0xD7000000u));
*Pointer<Int4>(out + 16 * 4) =
MulHigh(Int4(0x7FFFFFFF, 0x7FFFFFFF, 0x80008000, 0xFFFFFFFF),
Int4(0x7FFFFFFF, 0x80000000, 0x80008000, 0xFFFFFFFF));
*Pointer<UInt4>(out + 16 * 5) =
MulHigh(UInt4(0x7FFFFFFFu, 0x7FFFFFFFu, 0x80008000u, 0xFFFFFFFFu),
UInt4(0x7FFFFFFFu, 0x80000000u, 0x80008000u, 0xFFFFFFFFu));
// (U)Short8 variants currently unimplemented.
Return(0);
}
auto routine = function(testName().c_str());
unsigned int out[6][4];
memset(&out, 0, sizeof(out));
routine(&out);
EXPECT_EQ(out[0][0], 0x00080002u);
EXPECT_EQ(out[0][1], 0x008D000Fu);
EXPECT_EQ(out[1][0], 0x00080002u);
EXPECT_EQ(out[1][1], 0xE8C0000Fu);
EXPECT_EQ(out[2][0], 0x00000000u);
EXPECT_EQ(out[2][1], 0xFFFFFE9Cu);
EXPECT_EQ(out[2][2], 0xFFFFFF23u);
EXPECT_EQ(out[2][3], 0x01480000u);
EXPECT_EQ(out[3][0], 0x00000000u);
EXPECT_EQ(out[3][1], 0x00000179u);
EXPECT_EQ(out[3][2], 0x00000311u);
EXPECT_EQ(out[3][3], 0xB5680000u);
EXPECT_EQ(out[4][0], 0x3FFFFFFFu);
EXPECT_EQ(out[4][1], 0xC0000000u);
EXPECT_EQ(out[4][2], 0x3FFF8000u);
EXPECT_EQ(out[4][3], 0x00000000u);
EXPECT_EQ(out[5][0], 0x3FFFFFFFu);
EXPECT_EQ(out[5][1], 0x3FFFFFFFu);
EXPECT_EQ(out[5][2], 0x40008000u);
EXPECT_EQ(out[5][3], 0xFFFFFFFEu);
}
TEST(ReactorUnitTests, MulAdd)
{
FunctionT<int(void *)> function;
{
Pointer<Byte> out = function.Arg<0>();
*Pointer<Int2>(out + 8 * 0) =
MulAdd(Short4(0x1aa, 0x2dd, 0x3ee, 0xF422),
Short4(0x1bb, 0x2cc, 0x3ff, 0xF411));
// (U)Short8 variant is mentioned but unimplemented
Return(0);
}
auto routine = function(testName().c_str());
unsigned int out[1][2];
memset(&out, 0, sizeof(out));
routine(&out);
EXPECT_EQ(out[0][0], 0x000AE34Au);
EXPECT_EQ(out[0][1], 0x009D5254u);
}
TEST(ReactorUnitTests, PointersEqual)
{
FunctionT<int(void *, void *)> function;
{
Pointer<Byte> ptrA = function.Arg<0>();
Pointer<Byte> ptrB = function.Arg<1>();
If(ptrA == ptrB)
{
Return(1);
}
Else
{
Return(0);
}
}
auto routine = function(testName().c_str());
int *a = reinterpret_cast<int *>(uintptr_t(0x0000000000000000));
int *b = reinterpret_cast<int *>(uintptr_t(0x00000000F0000000));
int *c = reinterpret_cast<int *>(uintptr_t(0xF000000000000000));
EXPECT_EQ(routine(&a, &a), 1);
EXPECT_EQ(routine(&b, &b), 1);
EXPECT_EQ(routine(&c, &c), 1);
EXPECT_EQ(routine(&a, &b), 0);
EXPECT_EQ(routine(&b, &a), 0);
EXPECT_EQ(routine(&b, &c), 0);
EXPECT_EQ(routine(&c, &b), 0);
EXPECT_EQ(routine(&c, &a), 0);
EXPECT_EQ(routine(&a, &c), 0);
}
TEST(ReactorUnitTests, Args_2Mixed)
{
// 2 mixed type args
FunctionT<float(int, float)> function;
{
Int a = function.Arg<0>();
Float b = function.Arg<1>();
Return(Float(a) + b);
}
if(auto routine = function(testName().c_str()))
{
float result = routine(1, 2.f);
EXPECT_EQ(result, 3.f);
}
}
TEST(ReactorUnitTests, Args_4Mixed)
{
// 4 mixed type args (max register allocation on Windows)
FunctionT<float(int, float, int, float)> function;
{
Int a = function.Arg<0>();
Float b = function.Arg<1>();
Int c = function.Arg<2>();
Float d = function.Arg<3>();
Return(Float(a) + b + Float(c) + d);
}
if(auto routine = function(testName().c_str()))
{
float result = routine(1, 2.f, 3, 4.f);
EXPECT_EQ(result, 10.f);
}
}
TEST(ReactorUnitTests, Args_5Mixed)
{
// 5 mixed type args (5th spills over to stack on Windows)
FunctionT<float(int, float, int, float, int)> function;
{
Int a = function.Arg<0>();
Float b = function.Arg<1>();
Int c = function.Arg<2>();
Float d = function.Arg<3>();
Int e = function.Arg<4>();
Return(Float(a) + b + Float(c) + d + Float(e));
}
if(auto routine = function(testName().c_str()))
{
float result = routine(1, 2.f, 3, 4.f, 5);
EXPECT_EQ(result, 15.f);
}
}
TEST(ReactorUnitTests, Args_GreaterThan5Mixed)
{
// >5 mixed type args
FunctionT<float(int, float, int, float, int, float, int, float, int, float)> function;
{
Int a = function.Arg<0>();
Float b = function.Arg<1>();
Int c = function.Arg<2>();
Float d = function.Arg<3>();
Int e = function.Arg<4>();
Float f = function.Arg<5>();
Int g = function.Arg<6>();
Float h = function.Arg<7>();
Int i = function.Arg<8>();
Float j = function.Arg<9>();
Return(Float(a) + b + Float(c) + d + Float(e) + f + Float(g) + h + Float(i) + j);
}
if(auto routine = function(testName().c_str()))
{
float result = routine(1, 2.f, 3, 4.f, 5, 6.f, 7, 8.f, 9, 10.f);
EXPECT_EQ(result, 55.f);
}
}
// This test was written because on Windows with Subzero, we would get a crash when executing a function
// with a large number of local variables. The problem was that on Windows, 4K pages are allocated as
// needed for the stack whenever an access is made in a "guard page", at which point the page is committed,
// and the next 4K page becomes the guard page. If a stack access is made that's beyond the guard page,
// a regular page fault occurs. To fix this, Subzero (and any compiler) now emits a call to __chkstk with
// the stack size in EAX, so that it can probe the stack in 4K increments up to that size, committing the
// required pages. See https://docs.microsoft.com/en-us/windows/win32/devnotes/-win32-chkstk.
TEST(ReactorUnitTests, LargeStack)
{
// An empirically large enough value to access outside the guard pages
constexpr int ArrayByteSize = 24 * 1024;
constexpr int ArraySize = ArrayByteSize / sizeof(int32_t);
FunctionT<void(int32_t * v)> function;
{
// Allocate a stack array large enough that writing to the first element will reach beyond
// the guard page.
Array<Int, ArraySize> largeStackArray;
for(int i = 0; i < ArraySize; ++i)
{
largeStackArray[i] = i;
}
Pointer<Int> in = function.Arg<0>();
for(int i = 0; i < ArraySize; ++i)
{
in[i] = largeStackArray[i];
}
}
// LLVM takes very long to generate this routine when O2 optimizations are enabled. Disable for now.
// TODO(b/174031014): Remove this once we fix LLVM taking so long.
ScopedPragma O0(OptimizationLevel, 0);
auto routine = function(testName().c_str());
std::array<int32_t, ArraySize> v;
// Run this in a thread, so that we get the default reserved stack size (8K on Win64).
std::thread t([&] {
routine(v.data());
});
t.join();
for(int i = 0; i < ArraySize; ++i)
{
EXPECT_EQ(v[i], i);
}
}
TEST(ReactorUnitTests, ShlSmallRHSScalar)
{
// TODO(crbug.com/swiftshader/185): Testing a temporary LLVM workaround
if(Caps::backendName().find("LLVM") == std::string::npos) return;
FunctionT<unsigned()> function;
{
auto lhs = UInt(4);
auto rhs = UInt(8);
auto res = lhs << rhs;
Return(res);
}
auto routine = function(testName().c_str());
unsigned res = routine();
EXPECT_EQ(res, 1u << 10u);
}
TEST(ReactorUnitTests, ShlLargeRHSScalar)
{
// TODO(crbug.com/swiftshader/185): Testing a temporary LLVM workaround
if(Caps::backendName().find("LLVM") == std::string::npos) return;
FunctionT<unsigned()> function;
{
auto lhs = UInt(1);
auto rhs = UInt(99);
auto res = lhs << rhs;
Return(res);
}
auto routine = function(testName().c_str());
unsigned res = routine();
EXPECT_EQ(res, 1u << 31u);
}
TEST(ReactorUnitTests, ShrSmallRHSScalar)
{
// TODO(crbug.com/swiftshader/185): Testing a temporary LLVM workaround
if(Caps::backendName().find("LLVM") == std::string::npos) return;
FunctionT<unsigned()> function;
{
auto lhs = UInt(64);
auto rhs = UInt(4);
auto res = lhs >> rhs;
Return(res);
}
auto routine = function(testName().c_str());
unsigned res = routine();
EXPECT_EQ(res, 4u);
}
TEST(ReactorUnitTests, ShrLargeRHSScalar)
{
// TODO(crbug.com/swiftshader/185): Testing a temporary LLVM workaround
if(Caps::backendName().find("LLVM") == std::string::npos) return;
FunctionT<unsigned()> function;
{
auto lhs = UInt(4);
auto rhs = UInt(99);
auto res = lhs >> rhs;
Return(res);
}
auto routine = function(testName().c_str());
unsigned res = routine();
EXPECT_EQ(res, 0u);
}
TEST(ReactorUnitTests, ShlRHSVector)
{
// TODO(crbug.com/swiftshader/185): Testing a temporary LLVM workaround
if(Caps::backendName().find("LLVM") == std::string::npos) return;
FunctionT<void(unsigned *a, unsigned *b, unsigned *c, unsigned *d)> function;
{
Pointer<UInt> a = function.Arg<0>();
Pointer<UInt> b = function.Arg<1>();
Pointer<UInt> c = function.Arg<2>();
Pointer<UInt> d = function.Arg<3>();
auto lhs = UInt4(4, 3, 6, 5);
auto rhs = UInt4(8, 99, 2, 50);
UInt4 res = lhs << rhs;
*a = res.x;
*b = res.y;
*c = res.z;
*d = res.w;
}
auto routine = function(testName().c_str());
unsigned a = 0;
unsigned b = 0;
unsigned c = 0;
unsigned d = 0;
routine(&a, &b, &c, &d);
EXPECT_EQ(a, 1024u);
EXPECT_EQ(b, 0x80000000u);
EXPECT_EQ(c, 24u);
EXPECT_EQ(d, 0x80000000u);
}
TEST(ReactorUnitTests, ShrRHSVector)
{
// TODO(crbug.com/swiftshader/185): Testing a temporary LLVM workaround
if(Caps::backendName().find("LLVM") == std::string::npos) return;
FunctionT<void(unsigned *a, unsigned *b, unsigned *c, unsigned *d)> function;
{
Pointer<UInt> a = function.Arg<0>();
Pointer<UInt> b = function.Arg<1>();
Pointer<UInt> c = function.Arg<2>();
Pointer<UInt> d = function.Arg<3>();
auto lhs = UInt4(745, 23, 234, 54);
auto rhs = UInt4(8, 99, 2, 50);
UInt4 res = lhs >> rhs;
*a = res.x;
*b = res.y;
*c = res.z;
*d = res.w;
}
auto routine = function(testName().c_str());
unsigned a = 0;
unsigned b = 0;
unsigned c = 0;
unsigned d = 0;
routine(&a, &b, &c, &d);
EXPECT_EQ(a, 2u);
EXPECT_EQ(b, 0u);
EXPECT_EQ(c, 58u);
EXPECT_EQ(d, 0u);
}
TEST(ReactorUnitTests, ShrLargeRHSVector)
{
// TODO(crbug.com/swiftshader/185): Testing a temporary LLVM workaround
if(Caps::backendName().find("LLVM") == std::string::npos) return;
FunctionT<unsigned()> function;
{
auto lhs = UInt(4);
auto rhs = UInt(99);
auto res = lhs >> rhs;
Return(res);
}
auto routine = function(testName().c_str());
unsigned res = routine();
EXPECT_EQ(res, 0u);
}
TEST(ReactorUnitTests, Call)
{
struct Class
{
static int Callback(Class *p, int i, float f)
{
p->i = i;
p->f = f;
return i + int(f);
}
int i = 0;
float f = 0.0f;
};
FunctionT<int(void *)> function;
{
Pointer<Byte> c = function.Arg<0>();
auto res = Call(Class::Callback, c, 10, 20.0f);
Return(res);
}
auto routine = function(testName().c_str());
Class c;
int res = routine(&c);
EXPECT_EQ(res, 30);
EXPECT_EQ(c.i, 10);
EXPECT_EQ(c.f, 20.0f);
}
TEST(ReactorUnitTests, CallMemberFunction)
{
struct Class
{
int Callback(int argI, float argF)
{
i = argI;
f = argF;
return i + int(f);
}
int i = 0;
float f = 0.0f;
};
Class c;
FunctionT<int()> function;
{
auto res = Call(&Class::Callback, &c, 10, 20.0f);
Return(res);
}
auto routine = function(testName().c_str());
int res = routine();
EXPECT_EQ(res, 30);
EXPECT_EQ(c.i, 10);
EXPECT_EQ(c.f, 20.0f);
}
TEST(ReactorUnitTests, CallMemberFunctionIndirect)
{
struct Class
{
int Callback(int argI, float argF)
{
i = argI;
f = argF;
return i + int(f);
}
int i = 0;
float f = 0.0f;
};
FunctionT<int(void *)> function;
{
Pointer<Byte> c = function.Arg<0>();
auto res = Call(&Class::Callback, c, 10, 20.0f);
Return(res);
}
auto routine = function(testName().c_str());
Class c;
int res = routine(&c);
EXPECT_EQ(res, 30);
EXPECT_EQ(c.i, 10);
EXPECT_EQ(c.f, 20.0f);
}
TEST(ReactorUnitTests, CallImplicitCast)
{
struct Class
{
static void Callback(Class *c, const char *s)
{
c->str = s;
}
std::string str;
};
FunctionT<void(Class * c, const char *s)> function;
{
Pointer<Byte> c = function.Arg<0>();
Pointer<Byte> s = function.Arg<1>();
Call(Class::Callback, c, s);
}
auto routine = function(testName().c_str());
Class c;
routine(&c, "hello world");
EXPECT_EQ(c.str, "hello world");
}
TEST(ReactorUnitTests, CallBoolReturnFunction)
{
struct Class
{
static bool IsEven(int a)
{
return a % 2 == 0;
}
};
FunctionT<int(int)> function;
{
Int a = function.Arg<0>();
Bool res = Call(Class::IsEven, a);
If(res)
{
Return(1);
}
Return(0);
}
auto routine = function(testName().c_str());
for(int i = 0; i < 10; ++i)
{
EXPECT_EQ(routine(i), i % 2 == 0);
}
}
TEST(ReactorUnitTests, Call_Args4)
{
struct Class
{
static int Func(int a, int b, int c, int d)
{
return a + b + c + d;
}
};
{
FunctionT<int()> function;
{
auto res = Call(Class::Func, 1, 2, 3, 4);
Return(res);
}
auto routine = function(testName().c_str());
int res = routine();
EXPECT_EQ(res, 1 + 2 + 3 + 4);
}
}
TEST(ReactorUnitTests, Call_Args5)
{
struct Class
{
static int Func(int a, int b, int c, int d, int e)
{
return a + b + c + d + e;
}
};
{
FunctionT<int()> function;
{
auto res = Call(Class::Func, 1, 2, 3, 4, 5);
Return(res);
}
auto routine = function(testName().c_str());
int res = routine();
EXPECT_EQ(res, 1 + 2 + 3 + 4 + 5);
}
}
TEST(ReactorUnitTests, Call_ArgsMany)
{
struct Class
{
static int Func(int a, int b, int c, int d, int e, int f, int g, int h)
{
return a + b + c + d + e + f + g + h;
}
};
{
FunctionT<int()> function;
{
auto res = Call(Class::Func, 1, 2, 3, 4, 5, 6, 7, 8);
Return(res);
}
auto routine = function(testName().c_str());
int res = routine();
EXPECT_EQ(res, 1 + 2 + 3 + 4 + 5 + 6 + 7 + 8);
}
}
TEST(ReactorUnitTests, Call_ArgsMixed)
{
struct Class
{
static int Func(int a, float b, int *c, float *d, int e, float f, int *g, float *h)
{
return a + b + *c + *d + e + f + *g + *h;
}
};
{
FunctionT<int()> function;
{
Int c(3);
Float d(4);
Int g(7);
Float h(8);
auto res = Call(Class::Func, 1, 2.f, &c, &d, 5, 6.f, &g, &h);
Return(res);
}
auto routine = function(testName().c_str());
int res = routine();
EXPECT_EQ(res, 1 + 2 + 3 + 4 + 5 + 6 + 7 + 8);
}
}
TEST(ReactorUnitTests, Call_ArgsPointer)
{
struct Class
{
static int Func(int *a)
{
return *a;
}
};
{
FunctionT<int()> function;
{
Int a(12345);
auto res = Call(Class::Func, &a);
Return(res);
}
auto routine = function(testName().c_str());
int res = routine();
EXPECT_EQ(res, 12345);
}
}
TEST(ReactorUnitTests, CallExternalCallRoutine)
{
// routine1 calls Class::Func, passing it a pointer to routine2, and Class::Func calls routine2
auto routine2 = [] {
FunctionT<float(float, int)> function;
{
Float a = function.Arg<0>();
Int b = function.Arg<1>();
Return(a + Float(b));
}
return function("%s2", testName().c_str());
}();
struct Class
{
static float Func(void *p, float a, int b)
{
auto funcToCall = reinterpret_cast<float (*)(float, int)>(p);
return funcToCall(a, b);
}
};
auto routine1 = [] {
FunctionT<float(void *, float, int)> function;
{
Pointer<Byte> funcToCall = function.Arg<0>();
Float a = function.Arg<1>();
Int b = function.Arg<2>();
Float result = Call(Class::Func, funcToCall, a, b);
Return(result);
}
return function(testName().c_str());
}();
float result = routine1((void *)routine2.getEntry(), 12.f, 13);
EXPECT_EQ(result, 25.f);
}
// Check that a complex generated function which utilizes all 8 or 16 XMM
// registers computes the correct result.
// (Note that due to MSC's lack of support for inline assembly in x64,
// this test does not actually check that the register contents are
// preserved, just that the generated function computes the correct value.
// It's necessary to inspect the registers in a debugger to actually verify.)
TEST(ReactorUnitTests, PreserveXMMRegisters)
{
FunctionT<void(void *, void *)> function;
{
Pointer<Byte> in = function.Arg<0>();
Pointer<Byte> out = function.Arg<1>();
Float4 a = *Pointer<Float4>(in + 16 * 0);
Float4 b = *Pointer<Float4>(in + 16 * 1);
Float4 c = *Pointer<Float4>(in + 16 * 2);
Float4 d = *Pointer<Float4>(in + 16 * 3);
Float4 e = *Pointer<Float4>(in + 16 * 4);
Float4 f = *Pointer<Float4>(in + 16 * 5);
Float4 g = *Pointer<Float4>(in + 16 * 6);
Float4 h = *Pointer<Float4>(in + 16 * 7);
Float4 i = *Pointer<Float4>(in + 16 * 8);
Float4 j = *Pointer<Float4>(in + 16 * 9);
Float4 k = *Pointer<Float4>(in + 16 * 10);
Float4 l = *Pointer<Float4>(in + 16 * 11);
Float4 m = *Pointer<Float4>(in + 16 * 12);
Float4 n = *Pointer<Float4>(in + 16 * 13);
Float4 o = *Pointer<Float4>(in + 16 * 14);
Float4 p = *Pointer<Float4>(in + 16 * 15);
Float4 ab = a + b;
Float4 cd = c + d;
Float4 ef = e + f;
Float4 gh = g + h;
Float4 ij = i + j;
Float4 kl = k + l;
Float4 mn = m + n;
Float4 op = o + p;
Float4 abcd = ab + cd;
Float4 efgh = ef + gh;
Float4 ijkl = ij + kl;
Float4 mnop = mn + op;
Float4 abcdefgh = abcd + efgh;
Float4 ijklmnop = ijkl + mnop;
Float4 sum = abcdefgh + ijklmnop;
*Pointer<Float4>(out) = sum;
Return();
}
auto routine = function(testName().c_str());
assert(routine);
float input[64] = { 1.0f, 0.0f, 0.0f, 0.0f,
-1.0f, 1.0f, -1.0f, 0.0f,
1.0f, 2.0f, -2.0f, 0.0f,
-1.0f, 3.0f, -3.0f, 0.0f,
1.0f, 4.0f, -4.0f, 0.0f,
-1.0f, 5.0f, -5.0f, 0.0f,
1.0f, 6.0f, -6.0f, 0.0f,
-1.0f, 7.0f, -7.0f, 0.0f,
1.0f, 8.0f, -8.0f, 0.0f,
-1.0f, 9.0f, -9.0f, 0.0f,
1.0f, 10.0f, -10.0f, 0.0f,
-1.0f, 11.0f, -11.0f, 0.0f,
1.0f, 12.0f, -12.0f, 0.0f,
-1.0f, 13.0f, -13.0f, 0.0f,
1.0f, 14.0f, -14.0f, 0.0f,
-1.0f, 15.0f, -15.0f, 0.0f };
float result[4];
routine(input, result);
EXPECT_EQ(result[0], 0.0f);
EXPECT_EQ(result[1], 120.0f);
EXPECT_EQ(result[2], -120.0f);
EXPECT_EQ(result[3], 0.0f);
}
template<typename T>
class CToReactorTCastTest : public ::testing::Test
{
public:
using CType = typename std::tuple_element<0, T>::type;
using ReactorType = typename std::tuple_element<1, T>::type;
};
using CToReactorTCastTestTypes = ::testing::Types< // Subset of types that can be used as arguments.
// std::pair<bool, Bool>, FIXME(capn): Not supported as argument type by Subzero.
// std::pair<uint8_t, Byte>, FIXME(capn): Not supported as argument type by Subzero.
// std::pair<int8_t, SByte>, FIXME(capn): Not supported as argument type by Subzero.
// std::pair<int16_t, Short>, FIXME(capn): Not supported as argument type by Subzero.
// std::pair<uint16_t, UShort>, FIXME(capn): Not supported as argument type by Subzero.
std::pair<int, Int>,
std::pair<unsigned int, UInt>,
std::pair<float, Float>>;
TYPED_TEST_SUITE(CToReactorTCastTest, CToReactorTCastTestTypes);
TYPED_TEST(CToReactorTCastTest, Casts)
{
using CType = typename TestFixture::CType;
using ReactorType = typename TestFixture::ReactorType;
std::shared_ptr<Routine> routine;
{
Function<Int(ReactorType)> function;
{
ReactorType a = function.template Arg<0>();
ReactorType b = CType{};
RValue<ReactorType> c = RValue<ReactorType>(CType{});
Bool same = (a == b) && (a == c);
Return(IfThenElse(same, Int(1), Int(0))); // TODO: Ability to use Bools as return values.
}
routine = function(testName().c_str());
auto callable = (int (*)(CType))routine->getEntry();
CType in = {};
EXPECT_EQ(callable(in), 1);
}
}
template<typename T>
class GEPTest : public ::testing::Test
{
public:
using CType = typename std::tuple_element<0, T>::type;
using ReactorType = typename std::tuple_element<1, T>::type;
};
using GEPTestTypes = ::testing::Types<
std::pair<bool, Bool>,
std::pair<int8_t, Byte>,
std::pair<int8_t, SByte>,
std::pair<int8_t[4], Byte4>,
std::pair<int8_t[4], SByte4>,
std::pair<int8_t[8], Byte8>,
std::pair<int8_t[8], SByte8>,
std::pair<int8_t[16], Byte16>,
std::pair<int8_t[16], SByte16>,
std::pair<int16_t, Short>,
std::pair<int16_t, UShort>,
std::pair<int16_t[2], Short2>,
std::pair<int16_t[2], UShort2>,
std::pair<int16_t[4], Short4>,
std::pair<int16_t[4], UShort4>,
std::pair<int16_t[8], Short8>,
std::pair<int16_t[8], UShort8>,
std::pair<int, Int>,
std::pair<int, UInt>,
std::pair<int[2], Int2>,
std::pair<int[2], UInt2>,
std::pair<int[4], Int4>,
std::pair<int[4], UInt4>,
std::pair<int64_t, Long>,
std::pair<int16_t, Half>,
std::pair<float, Float>,
std::pair<float[2], Float2>,
std::pair<float[4], Float4>>;
TYPED_TEST_SUITE(GEPTest, GEPTestTypes);
TYPED_TEST(GEPTest, PtrOffsets)
{
using CType = typename TestFixture::CType;
using ReactorType = typename TestFixture::ReactorType;
std::shared_ptr<Routine> routine;
{
Function<Pointer<ReactorType>(Pointer<ReactorType>, Int)> function;
{
Pointer<ReactorType> pointer = function.template Arg<0>();
Int index = function.template Arg<1>();
Return(&pointer[index]);
}
routine = function(testName().c_str());
auto callable = (CType * (*)(CType *, unsigned int)) routine->getEntry();
union PtrInt
{
CType *p;
size_t i;
};
PtrInt base;
base.i = 0x10000;
for(int i = 0; i < 5; i++)
{
PtrInt reference;
reference.p = &base.p[i];
PtrInt result;
result.p = callable(base.p, i);
auto expect = reference.i - base.i;
auto got = result.i - base.i;
EXPECT_EQ(got, expect) << "i:" << i;
}
}
}
static const std::vector<int> fibonacci = {
0,
1,
1,
2,
3,
5,
8,
13,
21,
34,
55,
89,
144,
233,
377,
610,
987,
1597,
2584,
4181,
6765,
10946,
17711,
28657,
46368,
75025,
121393,
196418,
317811,
};
TEST(ReactorUnitTests, Fibonacci)
{
FunctionT<int(int)> function;
{
Int n = function.Arg<0>();
Int current = 0;
Int next = 1;
For(Int i = 0, i < n, i++)
{
auto tmp = current + next;
current = next;
next = tmp;
}
Return(current);
}
auto routine = function(testName().c_str());
for(size_t i = 0; i < fibonacci.size(); i++)
{
EXPECT_EQ(routine(i), fibonacci[i]);
}
}
TEST(ReactorUnitTests, Coroutines_Fibonacci)
{
if(!rr::Caps::coroutinesSupported())
{
SUCCEED() << "Coroutines not supported";
return;
}
Coroutine<int()> function;
{
Yield(Int(0));
Yield(Int(1));
Int current = 1;
Int next = 1;
While(true)
{
Yield(next);
auto tmp = current + next;
current = next;
next = tmp;
}
}
function.finalize(testName().c_str());
auto coroutine = function();
for(size_t i = 0; i < fibonacci.size(); i++)
{
int out = 0;
EXPECT_EQ(coroutine->await(out), true);
EXPECT_EQ(out, fibonacci[i]);
}
}
TEST(ReactorUnitTests, Coroutines_Parameters)
{
if(!rr::Caps::coroutinesSupported())
{
SUCCEED() << "Coroutines not supported";
return;
}
Coroutine<uint8_t(uint8_t * data, int count)> function;
{
Pointer<Byte> data = function.Arg<0>();
Int count = function.Arg<1>();
For(Int i = 0, i < count, i++)
{
Yield(data[i]);
}
}
function.finalize(testName().c_str());
uint8_t data[] = { 10, 20, 30 };
auto coroutine = function(&data[0], 3);
uint8_t out = 0;
EXPECT_EQ(coroutine->await(out), true);
EXPECT_EQ(out, 10);
out = 0;
EXPECT_EQ(coroutine->await(out), true);
EXPECT_EQ(out, 20);
out = 0;
EXPECT_EQ(coroutine->await(out), true);
EXPECT_EQ(out, 30);
out = 99;
EXPECT_EQ(coroutine->await(out), false);
EXPECT_EQ(out, 99);
EXPECT_EQ(coroutine->await(out), false);
EXPECT_EQ(out, 99);
}
// This test was written because Subzero's handling of vector types
// failed when more than one function is generated, as is the case
// with coroutines.
TEST(ReactorUnitTests, Coroutines_Vectors)
{
if(!rr::Caps::coroutinesSupported())
{
SUCCEED() << "Coroutines not supported";
return;
}
Coroutine<int()> function;
{
Int4 a{ 1, 2, 3, 4 };
Yield(rr::Extract(a, 2));
Int4 b{ 5, 6, 7, 8 };
Yield(rr::Extract(b, 1));
Int4 c{ 9, 10, 11, 12 };
Yield(rr::Extract(c, 1));
}
function.finalize(testName().c_str());
auto coroutine = function();
int out;
coroutine->await(out);
EXPECT_EQ(out, 3);
coroutine->await(out);
EXPECT_EQ(out, 6);
coroutine->await(out);
EXPECT_EQ(out, 10);
}
// This test was written to make sure a coroutine without a Yield()
// works correctly, by executing like a regular function with no
// return (the return type is ignored).
// We also run it twice to ensure per instance and/or global state
// is properly cleaned up in between.
TEST(ReactorUnitTests, Coroutines_NoYield)
{
if(!rr::Caps::coroutinesSupported())
{
SUCCEED() << "Coroutines not supported";
return;
}
for(int i = 0; i < 2; ++i)
{
Coroutine<int()> function;
{
Int a;
a = 4;
}
function.finalize(testName().c_str());
auto coroutine = function();
int out;
EXPECT_EQ(coroutine->await(out), false);
}
}
// Test generating one coroutine, and executing it on multiple threads. This makes
// sure the implementation manages per-call instance data correctly.
TEST(ReactorUnitTests, Coroutines_Parallel)
{
if(!rr::Caps::coroutinesSupported())
{
SUCCEED() << "Coroutines not supported";
return;
}
Coroutine<int()> function;
{
Yield(Int(0));
Yield(Int(1));
Int current = 1;
Int next = 1;
While(true)
{
Yield(next);
auto tmp = current + next;
current = next;
next = tmp;
}
}
// Must call on same thread that creates the coroutine
function.finalize(testName().c_str());
std::vector<std::thread> threads;
const size_t numThreads = 100;
for(size_t t = 0; t < numThreads; ++t)
{
threads.emplace_back([&] {
auto coroutine = function();
for(size_t i = 0; i < fibonacci.size(); i++)
{
int out = 0;
EXPECT_EQ(coroutine->await(out), true);
EXPECT_EQ(out, fibonacci[i]);
}
});
}
for(auto &t : threads)
{
t.join();
}
}
template<typename TestFuncType, typename RefFuncType, typename TestValueType>
struct IntrinsicTestParams
{
std::function<TestFuncType> testFunc; // Function we're testing (Reactor)
std::function<RefFuncType> refFunc; // Reference function to test against (C)
std::vector<TestValueType> testValues; // Values to input to functions
};
using IntrinsicTestParams_Float = IntrinsicTestParams<RValue<Float>(RValue<Float>), float(float), float>;
using IntrinsicTestParams_Float4 = IntrinsicTestParams<RValue<Float4>(RValue<Float4>), float(float), float>;
using IntrinsicTestParams_Float4_Float4 = IntrinsicTestParams<RValue<Float4>(RValue<Float4>, RValue<Float4>), float(float, float), std::pair<float, float>>;
// TODO(b/147818976): Each function has its own precision requirements for Vulkan, sometimes broken down
// by input range. These are currently validated by deqp, but we can improve our own tests as well.
// See https://www.khronos.org/registry/vulkan/specs/1.2-extensions/html/vkspec.html#spirvenv-precision-operation
constexpr double INTRINSIC_PRECISION = 1e-4;
struct IntrinsicTest_Float : public testing::TestWithParam<IntrinsicTestParams_Float>
{
void test()
{
FunctionT<float(float)> function;
{
Return(GetParam().testFunc((Float(function.Arg<0>()))));
}
auto routine = function(testName().c_str());
for(auto &&v : GetParam().testValues)
{
SCOPED_TRACE(v);
EXPECT_NEAR(routine(v), GetParam().refFunc(v), INTRINSIC_PRECISION);
}
}
};
// TODO: Move to Reactor.hpp
template<>
struct rr::CToReactor<int[4]>
{
using type = Int4;
static Int4 cast(float[4]);
};
// Value type wrapper around a <type>[4] (i.e. float4, int4)
template<typename T>
struct type4_value
{
using E = typename std::remove_pointer_t<std::decay_t<T>>;
type4_value() = default;
explicit type4_value(E rep)
: v{ rep, rep, rep, rep }
{}
type4_value(E x, E y, E z, E w)
: v{ x, y, z, w }
{}
bool operator==(const type4_value &rhs) const
{
return std::equal(std::begin(v), std::end(v), rhs.v);
}
// For gtest printing
friend std::ostream &operator<<(std::ostream &os, const type4_value &value)
{
return os << "[" << value.v[0] << ", " << value.v[1] << ", " << value.v[2] << ", " << value.v[3] << "]";
}
T v;
};
using float4_value = type4_value<float4>;
using int4_value = type4_value<int4>;
// Invoke a void(type4_value<T>*) routine on &v.v, returning wrapped result in v
template<typename RoutineType, typename T>
type4_value<T> invokeRoutine(RoutineType &routine, type4_value<T> v)
{
routine(&v.v);
return v;
}
// Invoke a void(type4_value<T>*, type4_value<T>*) routine on &v1.v, &v2.v returning wrapped result in v1
template<typename RoutineType, typename T>
type4_value<T> invokeRoutine(RoutineType &routine, type4_value<T> v1, type4_value<T> v2)
{
routine(&v1.v, &v2.v);
return v1;
}
struct IntrinsicTest_Float4 : public testing::TestWithParam<IntrinsicTestParams_Float4>
{
void test()
{
FunctionT<void(float4 *)> function;
{
Pointer<Float4> a = function.Arg<0>();
*a = GetParam().testFunc(*a);
Return();
}
auto routine = function(testName().c_str());
for(auto &&v : GetParam().testValues)
{
SCOPED_TRACE(v);
float4_value result = invokeRoutine(routine, float4_value{ v });
float4_value expected = float4_value{ GetParam().refFunc(v) };
EXPECT_NEAR(result.v[0], expected.v[0], INTRINSIC_PRECISION);
EXPECT_NEAR(result.v[1], expected.v[1], INTRINSIC_PRECISION);
EXPECT_NEAR(result.v[2], expected.v[2], INTRINSIC_PRECISION);
EXPECT_NEAR(result.v[3], expected.v[3], INTRINSIC_PRECISION);
}
}
};
struct IntrinsicTest_Float4_Float4 : public testing::TestWithParam<IntrinsicTestParams_Float4_Float4>
{
void test()
{
FunctionT<void(float4 *, float4 *)> function;
{
Pointer<Float4> a = function.Arg<0>();
Pointer<Float4> b = function.Arg<1>();
*a = GetParam().testFunc(*a, *b);
Return();
}
auto routine = function(testName().c_str());
for(auto &&v : GetParam().testValues)
{
SCOPED_TRACE(v);
float4_value result = invokeRoutine(routine, float4_value{ v.first }, float4_value{ v.second });
float4_value expected = float4_value{ GetParam().refFunc(v.first, v.second) };
EXPECT_NEAR(result.v[0], expected.v[0], INTRINSIC_PRECISION);
EXPECT_NEAR(result.v[1], expected.v[1], INTRINSIC_PRECISION);
EXPECT_NEAR(result.v[2], expected.v[2], INTRINSIC_PRECISION);
EXPECT_NEAR(result.v[3], expected.v[3], INTRINSIC_PRECISION);
}
}
};
// clang-format off
INSTANTIATE_TEST_SUITE_P(IntrinsicTestParams_Float, IntrinsicTest_Float, testing::Values(
IntrinsicTestParams_Float{ [](Float v) { return rr::Exp2(v); }, exp2f, {0.f, 1.f, 123.f} },
IntrinsicTestParams_Float{ [](Float v) { return rr::Log2(v); }, log2f, {1.f, 123.f} },
IntrinsicTestParams_Float{ [](Float v) { return rr::Sqrt(v); }, sqrtf, {0.f, 1.f, 123.f} }
));
// clang-format on
// TODO(b/149110874) Use coshf/sinhf when we've implemented SpirV versions at the SpirV level
float vulkan_sinhf(float a)
{
return ((expf(a) - expf(-a)) / 2);
}
float vulkan_coshf(float a)
{
return ((expf(a) + expf(-a)) / 2);
}
// clang-format off
constexpr float PI = 3.141592653589793f;
INSTANTIATE_TEST_SUITE_P(IntrinsicTestParams_Float4, IntrinsicTest_Float4, testing::Values(
IntrinsicTestParams_Float4{ [](RValue<Float4> v) { return rr::Sin(v); }, sinf, {0.f, 1.f, PI, 123.f} },
IntrinsicTestParams_Float4{ [](RValue<Float4> v) { return rr::Cos(v); }, cosf, {0.f, 1.f, PI, 123.f} },
IntrinsicTestParams_Float4{ [](RValue<Float4> v) { return rr::Tan(v); }, tanf, {0.f, 1.f, PI, 123.f} },
IntrinsicTestParams_Float4{ [](RValue<Float4> v) { return rr::Asin(v); }, asinf, {0.f, 1.f, -1.f} },
IntrinsicTestParams_Float4{ [](RValue<Float4> v) { return rr::Acos(v); }, acosf, {0.f, 1.f, -1.f} },
IntrinsicTestParams_Float4{ [](RValue<Float4> v) { return rr::Atan(v); }, atanf, {0.f, 1.f, PI, 123.f} },
IntrinsicTestParams_Float4{ [](RValue<Float4> v) { return rr::Sinh(v); }, vulkan_sinhf, {0.f, 1.f, PI} },
IntrinsicTestParams_Float4{ [](RValue<Float4> v) { return rr::Cosh(v); }, vulkan_coshf, {0.f, 1.f, PI} },
IntrinsicTestParams_Float4{ [](RValue<Float4> v) { return rr::Tanh(v); }, tanhf, {0.f, 1.f, PI} },
IntrinsicTestParams_Float4{ [](RValue<Float4> v) { return rr::Asinh(v); }, asinhf, {0.f, 1.f, PI, 123.f} },
IntrinsicTestParams_Float4{ [](RValue<Float4> v) { return rr::Acosh(v); }, acoshf, { 1.f, PI, 123.f} },
IntrinsicTestParams_Float4{ [](RValue<Float4> v) { return rr::Atanh(v); }, atanhf, {0.f, 0.9999f, -0.9999f} },
IntrinsicTestParams_Float4{ [](RValue<Float4> v) { return rr::Exp(v); }, expf, {0.f, 1.f, PI} },
IntrinsicTestParams_Float4{ [](RValue<Float4> v) { return rr::Log(v); }, logf, {1.f, PI, 123.f} },
IntrinsicTestParams_Float4{ [](RValue<Float4> v) { return rr::Exp2(v); }, exp2f, {0.f, 1.f, PI, 123.f} },
IntrinsicTestParams_Float4{ [](RValue<Float4> v) { return rr::Log2(v); }, log2f, {1.f, PI, 123.f} },
IntrinsicTestParams_Float4{ [](RValue<Float4> v) { return rr::Sqrt(v); }, sqrtf, {0.f, 1.f, PI, 123.f} }
));
// clang-format on
// clang-format off
INSTANTIATE_TEST_SUITE_P(IntrinsicTestParams_Float4_Float4, IntrinsicTest_Float4_Float4, testing::Values(
IntrinsicTestParams_Float4_Float4{ [](RValue<Float4> v1, RValue<Float4> v2) { return Atan2(v1, v2); }, atan2f, { {0.f, 0.f}, {0.f, -1.f}, {-1.f, 0.f}, {123.f, 123.f} } },
IntrinsicTestParams_Float4_Float4{ [](RValue<Float4> v1, RValue<Float4> v2) { return Pow(v1, v2); }, powf, { {1.f, 0.f}, {1.f, -1.f}, {-1.f, 0.f} } }
));
// clang-format on
TEST_P(IntrinsicTest_Float, Test)
{
test();
}
TEST_P(IntrinsicTest_Float4, Test)
{
test();
}
TEST_P(IntrinsicTest_Float4_Float4, Test)
{
test();
}
TEST(ReactorUnitTests, Intrinsics_Ctlz)
{
// ctlz: counts number of leading zeros
{
Function<UInt(UInt x)> function;
{
UInt x = function.Arg<0>();
Return(rr::Ctlz(x, false));
}
auto routine = function(testName().c_str());
auto callable = (uint32_t(*)(uint32_t))routine->getEntry();
for(uint32_t i = 0; i < 31; ++i)
{
uint32_t result = callable(1 << i);
EXPECT_EQ(result, 31 - i);
}
// Input 0 should return 32 for isZeroUndef == false
{
uint32_t result = callable(0);
EXPECT_EQ(result, 32u);
}
}
{
Function<Void(Pointer<UInt4>, UInt x)> function;
{
Pointer<UInt4> out = function.Arg<0>();
UInt x = function.Arg<1>();
*out = rr::Ctlz(UInt4(x), false);
}
auto routine = function(testName().c_str());
auto callable = (void (*)(uint32_t *, uint32_t))routine->getEntry();
uint32_t x[4];
for(uint32_t i = 0; i < 31; ++i)
{
callable(x, 1 << i);
EXPECT_EQ(x[0], 31 - i);
EXPECT_EQ(x[1], 31 - i);
EXPECT_EQ(x[2], 31 - i);
EXPECT_EQ(x[3], 31 - i);
}
// Input 0 should return 32 for isZeroUndef == false
{
callable(x, 0);
EXPECT_EQ(x[0], 32u);
EXPECT_EQ(x[1], 32u);
EXPECT_EQ(x[2], 32u);
EXPECT_EQ(x[3], 32u);
}
}
}
TEST(ReactorUnitTests, Intrinsics_Cttz)
{
// cttz: counts number of trailing zeros
{
Function<UInt(UInt x)> function;
{
UInt x = function.Arg<0>();
Return(rr::Cttz(x, false));
}
auto routine = function(testName().c_str());
auto callable = (uint32_t(*)(uint32_t))routine->getEntry();
for(uint32_t i = 0; i < 31; ++i)
{
uint32_t result = callable(1 << i);
EXPECT_EQ(result, i);
}
// Input 0 should return 32 for isZeroUndef == false
{
uint32_t result = callable(0);
EXPECT_EQ(result, 32u);
}
}
{
Function<Void(Pointer<UInt4>, UInt x)> function;
{
Pointer<UInt4> out = function.Arg<0>();
UInt x = function.Arg<1>();
*out = rr::Cttz(UInt4(x), false);
}
auto routine = function(testName().c_str());
auto callable = (void (*)(uint32_t *, uint32_t))routine->getEntry();
uint32_t x[4];
for(uint32_t i = 0; i < 31; ++i)
{
callable(x, 1 << i);
EXPECT_EQ(x[0], i);
EXPECT_EQ(x[1], i);
EXPECT_EQ(x[2], i);
EXPECT_EQ(x[3], i);
}
// Input 0 should return 32 for isZeroUndef == false
{
callable(x, 0);
EXPECT_EQ(x[0], 32u);
EXPECT_EQ(x[1], 32u);
EXPECT_EQ(x[2], 32u);
EXPECT_EQ(x[3], 32u);
}
}
}
TEST(ReactorUnitTests, ExtractFromRValue)
{
Function<Void(Pointer<Int4> values, Pointer<Int4> result)> function;
{
Pointer<Int4> vIn = function.Arg<0>();
Pointer<Int4> resultIn = function.Arg<1>();
RValue<Int4> v = *vIn;
Int4 result(678);
If(Extract(v, 0) == 42)
{
result = Insert(result, 1, 0);
}
If(Extract(v, 1) == 42)
{
result = Insert(result, 1, 1);
}
*resultIn = result;
Return();
}
auto routine = function(testName().c_str());
auto entry = (void (*)(int *, int *))routine->getEntry();
int v[4] = { 42, 42, 42, 42 };
int result[4] = { 99, 99, 99, 99 };
entry(v, result);
EXPECT_EQ(result[0], 1);
EXPECT_EQ(result[1], 1);
EXPECT_EQ(result[2], 678);
EXPECT_EQ(result[3], 678);
}
TEST(ReactorUnitTests, AddAtomic)
{
FunctionT<uint32_t(uint32_t * p, uint32_t a)> function;
{
Pointer<UInt> p = function.Arg<0>();
UInt a = function.Arg<1>();
UInt r = rr::AddAtomic(p, a, std::memory_order_relaxed);
Return(r);
}
auto routine = function(testName().c_str());
uint32_t x = 123;
uint32_t y = 456;
uint32_t prevX = routine(&x, y);
EXPECT_EQ(prevX, 123u);
EXPECT_EQ(x, 579u);
}
TEST(ReactorUnitTests, SubAtomic)
{
FunctionT<uint32_t(uint32_t * p, uint32_t a)> function;
{
Pointer<UInt> p = function.Arg<0>();
UInt a = function.Arg<1>();
UInt r = rr::SubAtomic(p, a, std::memory_order_relaxed);
Return(r);
}
auto routine = function(testName().c_str());
uint32_t x = 456;
uint32_t y = 123;
uint32_t prevX = routine(&x, y);
EXPECT_EQ(prevX, 456u);
EXPECT_EQ(x, 333u);
}
TEST(ReactorUnitTests, AndAtomic)
{
FunctionT<uint32_t(uint32_t * p, uint32_t a)> function;
{
Pointer<UInt> p = function.Arg<0>();
UInt a = function.Arg<1>();
UInt r = rr::AndAtomic(p, a, std::memory_order_relaxed);
Return(r);
}
auto routine = function(testName().c_str());
uint32_t x = 0b1111'0000;
uint32_t y = 0b1010'1100;
uint32_t prevX = routine(&x, y);
EXPECT_EQ(prevX, 0b1111'0000u);
EXPECT_EQ(x, 0b1010'0000u);
}
TEST(ReactorUnitTests, OrAtomic)
{
FunctionT<uint32_t(uint32_t * p, uint32_t a)> function;
{
Pointer<UInt> p = function.Arg<0>();
UInt a = function.Arg<1>();
UInt r = rr::OrAtomic(p, a, std::memory_order_relaxed);
Return(r);
}
auto routine = function(testName().c_str());
uint32_t x = 0b1111'0000;
uint32_t y = 0b1010'1100;
uint32_t prevX = routine(&x, y);
EXPECT_EQ(prevX, 0b1111'0000u);
EXPECT_EQ(x, 0b1111'1100u);
}
TEST(ReactorUnitTests, XorAtomic)
{
FunctionT<uint32_t(uint32_t * p, uint32_t a)> function;
{
Pointer<UInt> p = function.Arg<0>();
UInt a = function.Arg<1>();
UInt r = rr::XorAtomic(p, a, std::memory_order_relaxed);
Return(r);
}
auto routine = function(testName().c_str());
uint32_t x = 0b1111'0000;
uint32_t y = 0b1010'1100;
uint32_t prevX = routine(&x, y);
EXPECT_EQ(prevX, 0b1111'0000u);
EXPECT_EQ(x, 0b0101'1100u);
}
TEST(ReactorUnitTests, MinAtomic)
{
{
FunctionT<uint32_t(uint32_t * p, uint32_t a)> function;
{
Pointer<UInt> p = function.Arg<0>();
UInt a = function.Arg<1>();
UInt r = rr::MinAtomic(p, a, std::memory_order_relaxed);
Return(r);
}
auto routine = function(testName().c_str());
uint32_t x = 123;
uint32_t y = 100;
uint32_t prevX = routine(&x, y);
EXPECT_EQ(prevX, 123u);
EXPECT_EQ(x, 100u);
}
{
FunctionT<int32_t(int32_t * p, int32_t a)> function;
{
Pointer<Int> p = function.Arg<0>();
Int a = function.Arg<1>();
Int r = rr::MinAtomic(p, a, std::memory_order_relaxed);
Return(r);
}
auto routine = function(testName().c_str());
int32_t x = -123;
int32_t y = -200;
int32_t prevX = routine(&x, y);
EXPECT_EQ(prevX, -123);
EXPECT_EQ(x, -200);
}
}
TEST(ReactorUnitTests, MaxAtomic)
{
{
FunctionT<uint32_t(uint32_t * p, uint32_t a)> function;
{
Pointer<UInt> p = function.Arg<0>();
UInt a = function.Arg<1>();
UInt r = rr::MaxAtomic(p, a, std::memory_order_relaxed);
Return(r);
}
auto routine = function(testName().c_str());
uint32_t x = 123;
uint32_t y = 100;
uint32_t prevX = routine(&x, y);
EXPECT_EQ(prevX, 123u);
EXPECT_EQ(x, 123u);
}
{
FunctionT<int32_t(int32_t * p, int32_t a)> function;
{
Pointer<Int> p = function.Arg<0>();
Int a = function.Arg<1>();
Int r = rr::MaxAtomic(p, a, std::memory_order_relaxed);
Return(r);
}
auto routine = function(testName().c_str());
int32_t x = -123;
int32_t y = -200;
int32_t prevX = routine(&x, y);
EXPECT_EQ(prevX, -123);
EXPECT_EQ(x, -123);
}
}
TEST(ReactorUnitTests, ExchangeAtomic)
{
FunctionT<uint32_t(uint32_t * p, uint32_t a)> function;
{
Pointer<UInt> p = function.Arg<0>();
UInt a = function.Arg<1>();
UInt r = rr::ExchangeAtomic(p, a, std::memory_order_relaxed);
Return(r);
}
auto routine = function(testName().c_str());
uint32_t x = 123;
uint32_t y = 456;
uint32_t prevX = routine(&x, y);
EXPECT_EQ(prevX, 123u);
EXPECT_EQ(x, y);
}
TEST(ReactorUnitTests, CompareExchangeAtomic)
{
FunctionT<uint32_t(uint32_t * x, uint32_t y, uint32_t compare)> function;
{
Pointer<UInt> x = function.Arg<0>();
UInt y = function.Arg<1>();
UInt compare = function.Arg<2>();
UInt r = rr::CompareExchangeAtomic(x, y, compare, std::memory_order_relaxed, std::memory_order_relaxed);
Return(r);
}
auto routine = function(testName().c_str());
uint32_t x = 123;
uint32_t y = 456;
uint32_t compare = 123;
uint32_t prevX = routine(&x, y, compare);
EXPECT_EQ(prevX, 123u);
EXPECT_EQ(x, y);
x = 123;
y = 456;
compare = 456;
prevX = routine(&x, y, compare);
EXPECT_EQ(prevX, 123u);
EXPECT_EQ(x, 123u);
}
TEST(ReactorUnitTests, SRem)
{
FunctionT<void(int4 *, int4 *)> function;
{
Pointer<Int4> a = function.Arg<0>();
Pointer<Int4> b = function.Arg<1>();
*a = *a % *b;
}
auto routine = function(testName().c_str());
int4_value result = invokeRoutine(routine, int4_value{ 10, 11, 12, 13 }, int4_value{ 3, 3, 3, 3 });
int4_value expected = int4_value{ 10 % 3, 11 % 3, 12 % 3, 13 % 3 };
EXPECT_FLOAT_EQ(result.v[0], expected.v[0]);
EXPECT_FLOAT_EQ(result.v[1], expected.v[1]);
EXPECT_FLOAT_EQ(result.v[2], expected.v[2]);
EXPECT_FLOAT_EQ(result.v[3], expected.v[3]);
}
TEST(ReactorUnitTests, FRem)
{
FunctionT<void(float4 *, float4 *)> function;
{
Pointer<Float4> a = function.Arg<0>();
Pointer<Float4> b = function.Arg<1>();
*a = *a % *b;
}
auto routine = function(testName().c_str());
float4_value result = invokeRoutine(routine, float4_value{ 10.1f, 11.2f, 12.3f, 13.4f }, float4_value{ 3.f, 3.f, 3.f, 3.f });
float4_value expected = float4_value{ fmodf(10.1f, 3.f), fmodf(11.2f, 3.f), fmodf(12.3f, 3.f), fmodf(13.4f, 3.f) };
EXPECT_FLOAT_EQ(result.v[0], expected.v[0]);
EXPECT_FLOAT_EQ(result.v[1], expected.v[1]);
EXPECT_FLOAT_EQ(result.v[2], expected.v[2]);
EXPECT_FLOAT_EQ(result.v[3], expected.v[3]);
}
// Subzero's load instruction assumes that a Constant ptr value is an offset, rather than an absolute
// pointer, and would fail during codegen. This was fixed by casting the constant to a non-const
// variable, and loading from it instead. This test makes sure this works.
TEST(ReactorUnitTests, LoadFromConstantData)
{
const int value = 123;
FunctionT<int()> function;
{
auto p = Pointer<Int>{ ConstantData(&value, sizeof(value)) };
Int v = *p;
Return(v);
}
const int result = function(testName().c_str())();
EXPECT_EQ(result, value);
}
TEST(ReactorUnitTests, Multithreaded_Function)
{
constexpr int numThreads = 8;
constexpr int numLoops = 16;
auto threads = std::unique_ptr<std::thread[]>(new std::thread[numThreads]);
auto results = std::unique_ptr<int[]>(new int[numThreads * numLoops]);
for(int t = 0; t < numThreads; t++)
{
auto threadFunc = [&](int t) {
for(int l = 0; l < numLoops; l++)
{
FunctionT<int(int, int)> function;
{
Int a = function.Arg<0>();
Int b = function.Arg<1>();
Return((a << 16) | b);
}
auto f = function("%s_thread%d_loop%d", testName().c_str(), t, l);
results[t * numLoops + l] = f(t, l);
}
};
threads[t] = std::thread(threadFunc, t);
}
for(int t = 0; t < numThreads; t++)
{
threads[t].join();
}
for(int t = 0; t < numThreads; t++)
{
for(int l = 0; l < numLoops; l++)
{
auto expect = (t << 16) | l;
auto result = results[t * numLoops + l];
EXPECT_EQ(result, expect);
}
}
}
TEST(ReactorUnitTests, Multithreaded_Coroutine)
{
if(!rr::Caps::coroutinesSupported())
{
SUCCEED() << "Coroutines not supported";
return;
}
constexpr int numThreads = 8;
constexpr int numLoops = 16;
struct Result
{
bool yieldReturns[3];
int yieldValues[3];
};
auto threads = std::unique_ptr<std::thread[]>(new std::thread[numThreads]);
auto results = std::unique_ptr<Result[]>(new Result[numThreads * numLoops]);
for(int t = 0; t < numThreads; t++)
{
auto threadFunc = [&](int t) {
for(int l = 0; l < numLoops; l++)
{
Coroutine<int(int, int)> function;
{
Int a = function.Arg<0>();
Int b = function.Arg<1>();
Yield(a);
Yield(b);
}
function.finalize((testName() + "_thread" + std::to_string(t) + "_loop" + std::to_string(l)).c_str());
auto coroutine = function(t, l);
auto &result = results[t * numLoops + l];
result = {};
result.yieldReturns[0] = coroutine->await(result.yieldValues[0]);
result.yieldReturns[1] = coroutine->await(result.yieldValues[1]);
result.yieldReturns[2] = coroutine->await(result.yieldValues[2]);
}
};
threads[t] = std::thread(threadFunc, t);
}
for(int t = 0; t < numThreads; t++)
{
threads[t].join();
}
for(int t = 0; t < numThreads; t++)
{
for(int l = 0; l < numLoops; l++)
{
const auto &result = results[t * numLoops + l];
EXPECT_EQ(result.yieldReturns[0], true);
EXPECT_EQ(result.yieldValues[0], t);
EXPECT_EQ(result.yieldReturns[1], true);
EXPECT_EQ(result.yieldValues[1], l);
EXPECT_EQ(result.yieldReturns[2], false);
EXPECT_EQ(result.yieldValues[2], 0);
}
}
}
// For gtest printing of pairs
namespace std {
template<typename T, typename U>
std::ostream &operator<<(std::ostream &os, const std::pair<T, U> &value)
{
return os << "{ " << value.first << ", " << value.second << " }";
}
} // namespace std
class StdOutCapture
{
public:
~StdOutCapture()
{
stopIfCapturing();
}
void start()
{
stopIfCapturing();
capturing = true;
testing::internal::CaptureStdout();
}
std::string stop()
{
assert(capturing);
capturing = false;
return testing::internal::GetCapturedStdout();
}
private:
void stopIfCapturing()
{
if(capturing)
{
// This stops the capture
testing::internal::GetCapturedStdout();
}
}
bool capturing = false;
};
std::vector<std::string> split(const std::string &s)
{
std::vector<std::string> result;
std::istringstream iss(s);
for(std::string line; std::getline(iss, line);)
{
result.push_back(line);
}
return result;
}
TEST(ReactorUnitTests, PrintPrimitiveTypes)
{
#if defined(ENABLE_RR_PRINT) && !defined(ENABLE_RR_EMIT_PRINT_LOCATION)
FunctionT<void()> function;
{
bool b(true);
int8_t i8(-1);
uint8_t ui8(1);
int16_t i16(-1);
uint16_t ui16(1);
int32_t i32(-1);
uint32_t ui32(1);
int64_t i64(-1);
uint64_t ui64(1);
float f(1);
double d(2);
const char *cstr = "const char*";
std::string str = "std::string";
int *p = nullptr;
RR_WATCH(b);
RR_WATCH(i8);
RR_WATCH(ui8);
RR_WATCH(i16);
RR_WATCH(ui16);
RR_WATCH(i32);
RR_WATCH(ui32);
RR_WATCH(i64);
RR_WATCH(ui64);
RR_WATCH(f);
RR_WATCH(d);
RR_WATCH(cstr);
RR_WATCH(str);
RR_WATCH(p);
}
auto routine = function(testName().c_str());
char pNullptr[64];
snprintf(pNullptr, sizeof(pNullptr), " p: %p", nullptr);
const char *expected[] = {
" b: true",
" i8: -1",
" ui8: 1",
" i16: -1",
" ui16: 1",
" i32: -1",
" ui32: 1",
" i64: -1",
" ui64: 1",
" f: 1.000000",
" d: 2.000000",
" cstr: const char*",
" str: std::string",
pNullptr,
};
constexpr size_t expectedSize = sizeof(expected) / sizeof(expected[0]);
StdOutCapture capture;
capture.start();
routine();
auto output = split(capture.stop());
for(size_t i = 0, j = 1; i < expectedSize; ++i, j += 2)
{
ASSERT_EQ(expected[i], output[j]);
}
#endif
}
TEST(ReactorUnitTests, PrintReactorTypes)
{
#if defined(ENABLE_RR_PRINT) && !defined(ENABLE_RR_EMIT_PRINT_LOCATION)
FunctionT<void()> function;
{
Bool b(true);
Int i(-1);
Int2 i2(-1, -2);
Int4 i4(-1, -2, -3, -4);
UInt ui(1);
UInt2 ui2(1, 2);
UInt4 ui4(1, 2, 3, 4);
Short s(-1);
Short4 s4(-1, -2, -3, -4);
UShort us(1);
UShort4 us4(1, 2, 3, 4);
Float f(1);
Float4 f4(1, 2, 3, 4);
Long l(i);
Pointer<Int> pi = nullptr;
RValue<Int> rvi = i;
Byte by('a');
Byte4 by4(i4);
RR_WATCH(b);
RR_WATCH(i);
RR_WATCH(i2);
RR_WATCH(i4);
RR_WATCH(ui);
RR_WATCH(ui2);
RR_WATCH(ui4);
RR_WATCH(s);
RR_WATCH(s4);
RR_WATCH(us);
RR_WATCH(us4);
RR_WATCH(f);
RR_WATCH(f4);
RR_WATCH(l);
RR_WATCH(pi);
RR_WATCH(rvi);
RR_WATCH(by);
RR_WATCH(by4);
}
auto routine = function(testName().c_str());
char piNullptr[64];
snprintf(piNullptr, sizeof(piNullptr), " pi: %p", nullptr);
const char *expected[] = {
" b: true",
" i: -1",
" i2: [-1, -2]",
" i4: [-1, -2, -3, -4]",
" ui: 1",
" ui2: [1, 2]",
" ui4: [1, 2, 3, 4]",
" s: -1",
" s4: [-1, -2, -3, -4]",
" us: 1",
" us4: [1, 2, 3, 4]",
" f: 1.000000",
" f4: [1.000000, 2.000000, 3.000000, 4.000000]",
" l: -1",
piNullptr,
" rvi: -1",
" by: 97",
" by4: [255, 254, 253, 252]",
};
constexpr size_t expectedSize = sizeof(expected) / sizeof(expected[0]);
StdOutCapture capture;
capture.start();
routine();
auto output = split(capture.stop());
for(size_t i = 0, j = 1; i < expectedSize; ++i, j += 2)
{
ASSERT_EQ(expected[i], output[j]);
}
#endif
}
// Test constant <op> variable
template<typename T, typename Func>
T Arithmetic_LhsConstArg(T arg1, T arg2, Func f)
{
using ReactorT = CToReactorT<T>;
FunctionT<T(T)> function;
{
ReactorT lhs = arg1;
ReactorT rhs = function.template Arg<0>();
ReactorT result = f(lhs, rhs);
Return(result);
}
auto routine = function(testName().c_str());
return routine(arg2);
}
// Test variable <op> constant
template<typename T, typename Func>
T Arithmetic_RhsConstArg(T arg1, T arg2, Func f)
{
using ReactorT = CToReactorT<T>;
FunctionT<T(T)> function;
{
ReactorT lhs = function.template Arg<0>();
ReactorT rhs = arg2;
ReactorT result = f(lhs, rhs);
Return(result);
}
auto routine = function(testName().c_str());
return routine(arg1);
}
// Test constant <op> constant
template<typename T, typename Func>
T Arithmetic_TwoConstArgs(T arg1, T arg2, Func f)
{
using ReactorT = CToReactorT<T>;
FunctionT<T()> function;
{
ReactorT lhs = arg1;
ReactorT rhs = arg2;
ReactorT result = f(lhs, rhs);
Return(result);
}
auto routine = function(testName().c_str());
return routine();
}
template<typename T, typename Func>
void Arithmetic_ConstArgs(T arg1, T arg2, T expected, Func f)
{
SCOPED_TRACE(std::to_string(arg1) + " <op> " + std::to_string(arg2) + " = " + std::to_string(expected));
T result{};
result = Arithmetic_LhsConstArg(arg1, arg2, std::forward<Func>(f));
EXPECT_EQ(result, expected);
result = Arithmetic_RhsConstArg(arg1, arg2, std::forward<Func>(f));
EXPECT_EQ(result, expected);
result = Arithmetic_TwoConstArgs(arg1, arg2, std::forward<Func>(f));
EXPECT_EQ(result, expected);
}
// Test that we generate valid code for when one or both args to arithmetic operations
// are constant. In particular, we want to validate the case for two const args, as
// often lowered instructions do not support this case.
TEST(ReactorUnitTests, Arithmetic_ConstantArgs)
{
Arithmetic_ConstArgs(2, 3, 5, [](auto c1, auto c2) { return c1 + c2; });
Arithmetic_ConstArgs(5, 3, 2, [](auto c1, auto c2) { return c1 - c2; });
Arithmetic_ConstArgs(2, 3, 6, [](auto c1, auto c2) { return c1 * c2; });
Arithmetic_ConstArgs(6, 3, 2, [](auto c1, auto c2) { return c1 / c2; });
Arithmetic_ConstArgs(0xF0F0, 0xAAAA, 0xA0A0, [](auto c1, auto c2) { return c1 & c2; });
Arithmetic_ConstArgs(0xF0F0, 0xAAAA, 0xFAFA, [](auto c1, auto c2) { return c1 | c2; });
Arithmetic_ConstArgs(0xF0F0, 0xAAAA, 0x5A5A, [](auto c1, auto c2) { return c1 ^ c2; });
Arithmetic_ConstArgs(2.f, 3.f, 5.f, [](auto c1, auto c2) { return c1 + c2; });
Arithmetic_ConstArgs(5.f, 3.f, 2.f, [](auto c1, auto c2) { return c1 - c2; });
Arithmetic_ConstArgs(2.f, 3.f, 6.f, [](auto c1, auto c2) { return c1 * c2; });
Arithmetic_ConstArgs(6.f, 3.f, 2.f, [](auto c1, auto c2) { return c1 / c2; });
}
// Test for Subzero bad code-gen that was fixed in swiftshader-cl/50008
// This tests the case of copying enough arguments to local variables so that the locals
// get spilled to the stack when no more registers remain, and making sure these copies
// are generated correctly. Without the aforementioned fix, this fails 100% on Windows x86.
TEST(ReactorUnitTests, SpillLocalCopiesOfArgs)
{
struct Helpers
{
static bool True() { return true; }
};
const int numLoops = 5; // 2 should be enough, but loop more to make sure
FunctionT<int(int, int, int, int, int, int, int, int, int, int, int, int)> function;
{
Int result = 0;
Int a1 = function.Arg<0>();
Int a2 = function.Arg<1>();
Int a3 = function.Arg<2>();
Int a4 = function.Arg<3>();
Int a5 = function.Arg<4>();
Int a6 = function.Arg<5>();
Int a7 = function.Arg<6>();
Int a8 = function.Arg<7>();
Int a9 = function.Arg<8>();
Int a10 = function.Arg<9>();
Int a11 = function.Arg<10>();
Int a12 = function.Arg<11>();
for(int i = 0; i < numLoops; ++i)
{
// Copy all arguments to locals so that Ice::LocalVariableSplitter::handleSimpleVarAssign
// creates Variable copies of arguments. We loop so that we create enough of these so
// that some spill over to the stack.
Int i1 = a1;
Int i2 = a2;
Int i3 = a3;
Int i4 = a4;
Int i5 = a5;
Int i6 = a6;
Int i7 = a7;
Int i8 = a8;
Int i9 = a9;
Int i10 = a10;
Int i11 = a11;
Int i12 = a12;
// Forcibly materialize all variables so that Ice::Variable instances are created for each
// local; otherwise, Reactor r-value optimizations kick in, and the locals are elided.
Variable::materializeAll();
// We also need to create a separate block that uses the variables declared above
// so that rr::optimize() doesn't optimize them out when attempting to eliminate stores
// followed by a load in the same block.
If(Call(Helpers::True))
{
result += (i1 + i2 + i3 + i4 + i5 + i6 + i7 + i8 + i9 + i10 + i11 + i12);
}
}
Return(result);
}
auto routine = function(testName().c_str());
int result = routine(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12);
int expected = numLoops * (1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9 + 10 + 11 + 12);
EXPECT_EQ(result, expected);
}
#if defined(ENABLE_RR_EMIT_ASM_FILE)
TEST(ReactorUnitTests, EmitAsm)
{
// Only supported by LLVM for now
if(Caps::backendName().find("LLVM") == std::string::npos) return;
namespace fs = std::filesystem;
FunctionT<int(void)> function;
{
Int sum;
For(Int i = 0, i < 10, i++)
{
sum += i;
}
Return(sum);
}
auto routine = function(testName().c_str());
// Returns path to first match of filename in current directory
auto findFile = [](const std::string filename) -> fs::path {
for(auto &p : fs::directory_iterator("."))
{
if(!p.is_regular_file())
continue;
auto currFilename = p.path().filename().string();
auto index = currFilename.find(testName());
if(index != std::string::npos)
{
return p.path();
}
}
return {};
};
fs::path path = findFile(testName());
EXPECT_FALSE(path.empty());
// Make sure an asm file was created
std::ifstream fin(path);
EXPECT_TRUE(fin);
// Make sure address of routine is in the file
auto findAddressInFile = [](std::ifstream &fin, size_t address) {
std::string addressString = [&] {
std::stringstream addressSS;
addressSS << "0x" << std::uppercase << std::hex << address;
return addressSS.str();
}();
std::string token;
while(fin >> token)
{
if(token.find(addressString) != std::string::npos)
return true;
}
return false;
};
size_t address = reinterpret_cast<size_t>(routine.getEntry());
EXPECT_TRUE(findAddressInFile(fin, address));
// Delete the file in case subsequent runs generate one with a different sequence number
fin.close();
std::filesystem::remove(path);
}
#endif
////////////////////////////////
// Trait compile time checks. //
////////////////////////////////
// Assert CToReactorT resolves to expected types.
static_assert(std::is_same<CToReactorT<void>, Void>::value, "");
static_assert(std::is_same<CToReactorT<bool>, Bool>::value, "");
static_assert(std::is_same<CToReactorT<uint8_t>, Byte>::value, "");
static_assert(std::is_same<CToReactorT<int8_t>, SByte>::value, "");
static_assert(std::is_same<CToReactorT<int16_t>, Short>::value, "");
static_assert(std::is_same<CToReactorT<uint16_t>, UShort>::value, "");
static_assert(std::is_same<CToReactorT<int32_t>, Int>::value, "");
static_assert(std::is_same<CToReactorT<uint64_t>, Long>::value, "");
static_assert(std::is_same<CToReactorT<uint32_t>, UInt>::value, "");
static_assert(std::is_same<CToReactorT<float>, Float>::value, "");
// Assert CToReactorT for known pointer types resolves to expected types.
static_assert(std::is_same<CToReactorT<void *>, Pointer<Byte>>::value, "");
static_assert(std::is_same<CToReactorT<bool *>, Pointer<Bool>>::value, "");
static_assert(std::is_same<CToReactorT<uint8_t *>, Pointer<Byte>>::value, "");
static_assert(std::is_same<CToReactorT<int8_t *>, Pointer<SByte>>::value, "");
static_assert(std::is_same<CToReactorT<int16_t *>, Pointer<Short>>::value, "");
static_assert(std::is_same<CToReactorT<uint16_t *>, Pointer<UShort>>::value, "");
static_assert(std::is_same<CToReactorT<int32_t *>, Pointer<Int>>::value, "");
static_assert(std::is_same<CToReactorT<uint64_t *>, Pointer<Long>>::value, "");
static_assert(std::is_same<CToReactorT<uint32_t *>, Pointer<UInt>>::value, "");
static_assert(std::is_same<CToReactorT<float *>, Pointer<Float>>::value, "");
static_assert(std::is_same<CToReactorT<uint16_t **>, Pointer<Pointer<UShort>>>::value, "");
static_assert(std::is_same<CToReactorT<uint16_t ***>, Pointer<Pointer<Pointer<UShort>>>>::value, "");
// Assert CToReactorT for unknown pointer types resolves to Pointer<Byte>.
struct S
{};
static_assert(std::is_same<CToReactorT<S *>, Pointer<Byte>>::value, "");
static_assert(std::is_same<CToReactorT<S **>, Pointer<Pointer<Byte>>>::value, "");
static_assert(std::is_same<CToReactorT<S ***>, Pointer<Pointer<Pointer<Byte>>>>::value, "");
// Assert IsRValue<> resolves true for RValue<> types.
static_assert(IsRValue<RValue<Void>>::value, "");
static_assert(IsRValue<RValue<Bool>>::value, "");
static_assert(IsRValue<RValue<Byte>>::value, "");
static_assert(IsRValue<RValue<SByte>>::value, "");
static_assert(IsRValue<RValue<Short>>::value, "");
static_assert(IsRValue<RValue<UShort>>::value, "");
static_assert(IsRValue<RValue<Int>>::value, "");
static_assert(IsRValue<RValue<Long>>::value, "");
static_assert(IsRValue<RValue<UInt>>::value, "");
static_assert(IsRValue<RValue<Float>>::value, "");
// Assert IsLValue<> resolves true for LValue types.
static_assert(IsLValue<Bool>::value, "");
static_assert(IsLValue<Byte>::value, "");
static_assert(IsLValue<SByte>::value, "");
static_assert(IsLValue<Short>::value, "");
static_assert(IsLValue<UShort>::value, "");
static_assert(IsLValue<Int>::value, "");
static_assert(IsLValue<Long>::value, "");
static_assert(IsLValue<UInt>::value, "");
static_assert(IsLValue<Float>::value, "");
// Assert IsReference<> resolves true for Reference types.
static_assert(IsReference<Reference<Bool>>::value, "");
static_assert(IsReference<Reference<Byte>>::value, "");
static_assert(IsReference<Reference<SByte>>::value, "");
static_assert(IsReference<Reference<Short>>::value, "");
static_assert(IsReference<Reference<UShort>>::value, "");
static_assert(IsReference<Reference<Int>>::value, "");
static_assert(IsReference<Reference<Long>>::value, "");
static_assert(IsReference<Reference<UInt>>::value, "");
static_assert(IsReference<Reference<Float>>::value, "");
// Assert IsRValue<> resolves false for LValue types.
static_assert(!IsRValue<Void>::value, "");
static_assert(!IsRValue<Bool>::value, "");
static_assert(!IsRValue<Byte>::value, "");
static_assert(!IsRValue<SByte>::value, "");
static_assert(!IsRValue<Short>::value, "");
static_assert(!IsRValue<UShort>::value, "");
static_assert(!IsRValue<Int>::value, "");
static_assert(!IsRValue<Long>::value, "");
static_assert(!IsRValue<UInt>::value, "");
static_assert(!IsRValue<Float>::value, "");
// Assert IsRValue<> resolves false for Reference types.
static_assert(!IsRValue<Reference<Void>>::value, "");
static_assert(!IsRValue<Reference<Bool>>::value, "");
static_assert(!IsRValue<Reference<Byte>>::value, "");
static_assert(!IsRValue<Reference<SByte>>::value, "");
static_assert(!IsRValue<Reference<Short>>::value, "");
static_assert(!IsRValue<Reference<UShort>>::value, "");
static_assert(!IsRValue<Reference<Int>>::value, "");
static_assert(!IsRValue<Reference<Long>>::value, "");
static_assert(!IsRValue<Reference<UInt>>::value, "");
static_assert(!IsRValue<Reference<Float>>::value, "");
// Assert IsRValue<> resolves false for C types.
static_assert(!IsRValue<void>::value, "");
static_assert(!IsRValue<bool>::value, "");
static_assert(!IsRValue<uint8_t>::value, "");
static_assert(!IsRValue<int8_t>::value, "");
static_assert(!IsRValue<int16_t>::value, "");
static_assert(!IsRValue<uint16_t>::value, "");
static_assert(!IsRValue<int32_t>::value, "");
static_assert(!IsRValue<uint64_t>::value, "");
static_assert(!IsRValue<uint32_t>::value, "");
static_assert(!IsRValue<float>::value, "");
// Assert IsLValue<> resolves false for RValue<> types.
static_assert(!IsLValue<RValue<Void>>::value, "");
static_assert(!IsLValue<RValue<Bool>>::value, "");
static_assert(!IsLValue<RValue<Byte>>::value, "");
static_assert(!IsLValue<RValue<SByte>>::value, "");
static_assert(!IsLValue<RValue<Short>>::value, "");
static_assert(!IsLValue<RValue<UShort>>::value, "");
static_assert(!IsLValue<RValue<Int>>::value, "");
static_assert(!IsLValue<RValue<Long>>::value, "");
static_assert(!IsLValue<RValue<UInt>>::value, "");
static_assert(!IsLValue<RValue<Float>>::value, "");
// Assert IsLValue<> resolves false for Void type.
static_assert(!IsLValue<Void>::value, "");
// Assert IsLValue<> resolves false for Reference<> types.
static_assert(!IsLValue<Reference<Void>>::value, "");
static_assert(!IsLValue<Reference<Bool>>::value, "");
static_assert(!IsLValue<Reference<Byte>>::value, "");
static_assert(!IsLValue<Reference<SByte>>::value, "");
static_assert(!IsLValue<Reference<Short>>::value, "");
static_assert(!IsLValue<Reference<UShort>>::value, "");
static_assert(!IsLValue<Reference<Int>>::value, "");
static_assert(!IsLValue<Reference<Long>>::value, "");
static_assert(!IsLValue<Reference<UInt>>::value, "");
static_assert(!IsLValue<Reference<Float>>::value, "");
// Assert IsLValue<> resolves false for C types.
static_assert(!IsLValue<void>::value, "");
static_assert(!IsLValue<bool>::value, "");
static_assert(!IsLValue<uint8_t>::value, "");
static_assert(!IsLValue<int8_t>::value, "");
static_assert(!IsLValue<int16_t>::value, "");
static_assert(!IsLValue<uint16_t>::value, "");
static_assert(!IsLValue<int32_t>::value, "");
static_assert(!IsLValue<uint64_t>::value, "");
static_assert(!IsLValue<uint32_t>::value, "");
static_assert(!IsLValue<float>::value, "");
// Assert IsDefined<> resolves true for RValue<> types.
static_assert(IsDefined<RValue<Void>>::value, "");
static_assert(IsDefined<RValue<Bool>>::value, "");
static_assert(IsDefined<RValue<Byte>>::value, "");
static_assert(IsDefined<RValue<SByte>>::value, "");
static_assert(IsDefined<RValue<Short>>::value, "");
static_assert(IsDefined<RValue<UShort>>::value, "");
static_assert(IsDefined<RValue<Int>>::value, "");
static_assert(IsDefined<RValue<Long>>::value, "");
static_assert(IsDefined<RValue<UInt>>::value, "");
static_assert(IsDefined<RValue<Float>>::value, "");
// Assert IsDefined<> resolves true for LValue types.
static_assert(IsDefined<Void>::value, "");
static_assert(IsDefined<Bool>::value, "");
static_assert(IsDefined<Byte>::value, "");
static_assert(IsDefined<SByte>::value, "");
static_assert(IsDefined<Short>::value, "");
static_assert(IsDefined<UShort>::value, "");
static_assert(IsDefined<Int>::value, "");
static_assert(IsDefined<Long>::value, "");
static_assert(IsDefined<UInt>::value, "");
static_assert(IsDefined<Float>::value, "");
// Assert IsDefined<> resolves true for Reference<> types.
static_assert(IsDefined<Reference<Bool>>::value, "");
static_assert(IsDefined<Reference<Byte>>::value, "");
static_assert(IsDefined<Reference<SByte>>::value, "");
static_assert(IsDefined<Reference<Short>>::value, "");
static_assert(IsDefined<Reference<UShort>>::value, "");
static_assert(IsDefined<Reference<Int>>::value, "");
static_assert(IsDefined<Reference<Long>>::value, "");
static_assert(IsDefined<Reference<UInt>>::value, "");
static_assert(IsDefined<Reference<Float>>::value, "");
// Assert IsDefined<> resolves true for C types.
static_assert(IsDefined<void>::value, "");
static_assert(IsDefined<bool>::value, "");
static_assert(IsDefined<uint8_t>::value, "");
static_assert(IsDefined<int8_t>::value, "");
static_assert(IsDefined<int16_t>::value, "");
static_assert(IsDefined<uint16_t>::value, "");
static_assert(IsDefined<int32_t>::value, "");
static_assert(IsDefined<uint64_t>::value, "");
static_assert(IsDefined<uint32_t>::value, "");
static_assert(IsDefined<float>::value, "");