skia2/tests/GLProgramsTest.cpp
csmartdalton bde96c6263 Move clip mask generation into GrReducedClip
This is a temporary solution to facilitate window rectangles and make
clip mask generation more accessible for testing. The eventual goal is
to simplify clips and merge GrReducedClip into GrClipStackClip.

BUG=skia:
GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=2263343002

Review-Url: https://codereview.chromium.org/2263343002
2016-08-31 12:54:46 -07:00

453 lines
17 KiB
C++

/*
* Copyright 2011 Google Inc.
*
* Use of this source code is governed by a BSD-style license that can be
* found in the LICENSE file.
*/
// This is a GPU-backend specific test. It relies on static intializers to work
#include "SkTypes.h"
#if SK_SUPPORT_GPU && SK_ALLOW_STATIC_GLOBAL_INITIALIZERS
#include "GrAutoLocaleSetter.h"
#include "GrBatchTest.h"
#include "GrContextFactory.h"
#include "GrContextPriv.h"
#include "GrDrawContextPriv.h"
#include "GrDrawingManager.h"
#include "GrInvariantOutput.h"
#include "GrPipeline.h"
#include "GrResourceProvider.h"
#include "GrTest.h"
#include "GrXferProcessor.h"
#include "SkChecksum.h"
#include "SkRandom.h"
#include "Test.h"
#include "batches/GrDrawBatch.h"
#include "effects/GrConfigConversionEffect.h"
#include "effects/GrPorterDuffXferProcessor.h"
#include "effects/GrXfermodeFragmentProcessor.h"
#include "gl/GrGLGpu.h"
#include "glsl/GrGLSLFragmentProcessor.h"
#include "glsl/GrGLSLFragmentShaderBuilder.h"
#include "glsl/GrGLSLProgramBuilder.h"
/*
* A dummy processor which just tries to insert a massive key and verify that it can retrieve the
* whole thing correctly
*/
static const uint32_t kMaxKeySize = 1024;
class GLBigKeyProcessor : public GrGLSLFragmentProcessor {
public:
void emitCode(EmitArgs& args) override {
// pass through
GrGLSLFragmentBuilder* fragBuilder = args.fFragBuilder;
if (args.fInputColor) {
fragBuilder->codeAppendf("%s = %s;\n", args.fOutputColor, args.fInputColor);
} else {
fragBuilder->codeAppendf("%s = vec4(1.0);\n", args.fOutputColor);
}
}
static void GenKey(const GrProcessor& processor, const GrGLSLCaps&, GrProcessorKeyBuilder* b) {
for (uint32_t i = 0; i < kMaxKeySize; i++) {
b->add32(i);
}
}
private:
typedef GrGLSLFragmentProcessor INHERITED;
};
class BigKeyProcessor : public GrFragmentProcessor {
public:
static sk_sp<GrFragmentProcessor> Make() {
return sk_sp<GrFragmentProcessor>(new BigKeyProcessor);
}
const char* name() const override { return "Big Ole Key"; }
GrGLSLFragmentProcessor* onCreateGLSLInstance() const override {
return new GLBigKeyProcessor;
}
private:
BigKeyProcessor() {
this->initClassID<BigKeyProcessor>();
}
virtual void onGetGLSLProcessorKey(const GrGLSLCaps& caps,
GrProcessorKeyBuilder* b) const override {
GLBigKeyProcessor::GenKey(*this, caps, b);
}
bool onIsEqual(const GrFragmentProcessor&) const override { return true; }
void onComputeInvariantOutput(GrInvariantOutput* inout) const override { }
GR_DECLARE_FRAGMENT_PROCESSOR_TEST;
typedef GrFragmentProcessor INHERITED;
};
GR_DEFINE_FRAGMENT_PROCESSOR_TEST(BigKeyProcessor);
sk_sp<GrFragmentProcessor> BigKeyProcessor::TestCreate(GrProcessorTestData*) {
return BigKeyProcessor::Make();
}
//////////////////////////////////////////////////////////////////////////////
class BlockInputFragmentProcessor : public GrFragmentProcessor {
public:
static sk_sp<GrFragmentProcessor> Make(sk_sp<GrFragmentProcessor> fp) {
return sk_sp<GrFragmentProcessor>(new BlockInputFragmentProcessor(fp));
}
const char* name() const override { return "Block Input"; }
GrGLSLFragmentProcessor* onCreateGLSLInstance() const override { return new GLFP; }
private:
class GLFP : public GrGLSLFragmentProcessor {
public:
void emitCode(EmitArgs& args) override {
this->emitChild(0, nullptr, args);
}
private:
typedef GrGLSLFragmentProcessor INHERITED;
};
BlockInputFragmentProcessor(sk_sp<GrFragmentProcessor> child) {
this->initClassID<BlockInputFragmentProcessor>();
this->registerChildProcessor(std::move(child));
}
void onGetGLSLProcessorKey(const GrGLSLCaps& caps, GrProcessorKeyBuilder* b) const override {}
bool onIsEqual(const GrFragmentProcessor&) const override { return true; }
void onComputeInvariantOutput(GrInvariantOutput* inout) const override {
inout->setToOther(kRGBA_GrColorComponentFlags, GrColor_WHITE,
GrInvariantOutput::kWillNot_ReadInput);
this->childProcessor(0).computeInvariantOutput(inout);
}
typedef GrFragmentProcessor INHERITED;
};
//////////////////////////////////////////////////////////////////////////////
/*
* Begin test code
*/
static const int kRenderTargetHeight = 1;
static const int kRenderTargetWidth = 1;
static sk_sp<GrDrawContext> random_draw_context(GrContext* context,
SkRandom* random,
const GrCaps* caps) {
GrSurfaceOrigin origin = random->nextBool() ? kTopLeft_GrSurfaceOrigin
: kBottomLeft_GrSurfaceOrigin;
int sampleCnt = random->nextBool() ? SkTMin(4, caps->maxSampleCount()) : 0;
sk_sp<GrDrawContext> drawContext(context->makeDrawContext(SkBackingFit::kExact,
kRenderTargetWidth,
kRenderTargetHeight,
kRGBA_8888_GrPixelConfig,
nullptr,
sampleCnt,
origin));
return drawContext;
}
static void set_random_xpf(GrPaint* paint, GrProcessorTestData* d) {
sk_sp<GrXPFactory> xpf(GrProcessorTestFactory<GrXPFactory>::Make(d));
SkASSERT(xpf);
paint->setXPFactory(std::move(xpf));
}
static sk_sp<GrFragmentProcessor> create_random_proc_tree(GrProcessorTestData* d,
int minLevels, int maxLevels) {
SkASSERT(1 <= minLevels);
SkASSERT(minLevels <= maxLevels);
// Return a leaf node if maxLevels is 1 or if we randomly chose to terminate.
// If returning a leaf node, make sure that it doesn't have children (e.g. another
// GrComposeEffect)
const float terminateProbability = 0.3f;
if (1 == minLevels) {
bool terminate = (1 == maxLevels) || (d->fRandom->nextF() < terminateProbability);
if (terminate) {
sk_sp<GrFragmentProcessor> fp;
while (true) {
fp = GrProcessorTestFactory<GrFragmentProcessor>::Make(d);
SkASSERT(fp);
if (0 == fp->numChildProcessors()) {
break;
}
}
return fp;
}
}
// If we didn't terminate, choose either the left or right subtree to fulfill
// the minLevels requirement of this tree; the other child can have as few levels as it wants.
// Also choose a random xfer mode that's supported by CreateFrom2Procs().
if (minLevels > 1) {
--minLevels;
}
sk_sp<GrFragmentProcessor> minLevelsChild(create_random_proc_tree(d, minLevels, maxLevels - 1));
sk_sp<GrFragmentProcessor> otherChild(create_random_proc_tree(d, 1, maxLevels - 1));
SkXfermode::Mode mode = static_cast<SkXfermode::Mode>(d->fRandom->nextRangeU(0,
SkXfermode::kLastCoeffMode));
sk_sp<GrFragmentProcessor> fp;
if (d->fRandom->nextF() < 0.5f) {
fp = GrXfermodeFragmentProcessor::MakeFromTwoProcessors(std::move(minLevelsChild),
std::move(otherChild), mode);
SkASSERT(fp);
} else {
fp = GrXfermodeFragmentProcessor::MakeFromTwoProcessors(std::move(otherChild),
std::move(minLevelsChild), mode);
SkASSERT(fp);
}
return fp;
}
static void set_random_color_coverage_stages(GrPaint* paint,
GrProcessorTestData* d,
int maxStages) {
// Randomly choose to either create a linear pipeline of procs or create one proc tree
const float procTreeProbability = 0.5f;
if (d->fRandom->nextF() < procTreeProbability) {
// A full tree with 5 levels (31 nodes) may exceed the max allowed length of the gl
// processor key; maxTreeLevels should be a number from 1 to 4 inclusive.
const int maxTreeLevels = 4;
sk_sp<GrFragmentProcessor> fp(create_random_proc_tree(d, 2, maxTreeLevels));
paint->addColorFragmentProcessor(std::move(fp));
} else {
int numProcs = d->fRandom->nextULessThan(maxStages + 1);
int numColorProcs = d->fRandom->nextULessThan(numProcs + 1);
for (int s = 0; s < numProcs;) {
sk_sp<GrFragmentProcessor> fp(GrProcessorTestFactory<GrFragmentProcessor>::Make(d));
SkASSERT(fp);
// finally add the stage to the correct pipeline in the drawstate
if (s < numColorProcs) {
paint->addColorFragmentProcessor(std::move(fp));
} else {
paint->addCoverageFragmentProcessor(std::move(fp));
}
++s;
}
}
}
static bool set_random_state(GrPaint* paint, SkRandom* random) {
if (random->nextBool()) {
paint->setDisableOutputConversionToSRGB(true);
}
if (random->nextBool()) {
paint->setAllowSRGBInputs(true);
}
if (random->nextBool()) {
paint->setAntiAlias(true);
}
return random->nextBool();
}
// right now, the only thing we seem to care about in drawState's stencil is 'doesWrite()'
static const GrUserStencilSettings* get_random_stencil(SkRandom* random) {
static constexpr GrUserStencilSettings kDoesWriteStencil(
GrUserStencilSettings::StaticInit<
0xffff,
GrUserStencilTest::kAlways,
0xffff,
GrUserStencilOp::kReplace,
GrUserStencilOp::kReplace,
0xffff>()
);
static constexpr GrUserStencilSettings kDoesNotWriteStencil(
GrUserStencilSettings::StaticInit<
0xffff,
GrUserStencilTest::kNever,
0xffff,
GrUserStencilOp::kKeep,
GrUserStencilOp::kKeep,
0xffff>()
);
if (random->nextBool()) {
return &kDoesWriteStencil;
} else {
return &kDoesNotWriteStencil;
}
}
bool GrDrawingManager::ProgramUnitTest(GrContext* context, int maxStages) {
GrDrawingManager* drawingManager = context->contextPriv().drawingManager();
// setup dummy textures
GrSurfaceDesc dummyDesc;
dummyDesc.fFlags = kRenderTarget_GrSurfaceFlag;
dummyDesc.fConfig = kSkia8888_GrPixelConfig;
dummyDesc.fWidth = 34;
dummyDesc.fHeight = 18;
SkAutoTUnref<GrTexture> dummyTexture1(
context->textureProvider()->createTexture(dummyDesc, SkBudgeted::kNo, nullptr, 0));
dummyDesc.fFlags = kNone_GrSurfaceFlags;
dummyDesc.fConfig = kAlpha_8_GrPixelConfig;
dummyDesc.fWidth = 16;
dummyDesc.fHeight = 22;
SkAutoTUnref<GrTexture> dummyTexture2(
context->textureProvider()->createTexture(dummyDesc, SkBudgeted::kNo, nullptr, 0));
if (!dummyTexture1 || ! dummyTexture2) {
SkDebugf("Could not allocate dummy textures");
return false;
}
GrTexture* dummyTextures[] = {dummyTexture1.get(), dummyTexture2.get()};
// dummy scissor state
GrScissorState scissor;
SkRandom random;
static const int NUM_TESTS = 1024;
for (int t = 0; t < NUM_TESTS; t++) {
// setup random render target(can fail)
sk_sp<GrDrawContext> drawContext(random_draw_context(context, &random, context->caps()));
if (!drawContext) {
SkDebugf("Could not allocate drawContext");
return false;
}
GrPaint grPaint;
SkAutoTUnref<GrDrawBatch> batch(GrRandomDrawBatch(&random, context));
SkASSERT(batch);
GrProcessorTestData ptd(&random, context, context->caps(),
drawContext.get(), dummyTextures);
set_random_color_coverage_stages(&grPaint, &ptd, maxStages);
set_random_xpf(&grPaint, &ptd);
bool snapToCenters = set_random_state(&grPaint, &random);
const GrUserStencilSettings* uss = get_random_stencil(&random);
drawContext->drawContextPriv().testingOnly_drawBatch(grPaint, batch, uss, snapToCenters);
}
// Flush everything, test passes if flush is successful(ie, no asserts are hit, no crashes)
drawingManager->flush();
// Validate that GrFPs work correctly without an input.
sk_sp<GrDrawContext> drawContext(context->makeDrawContext(SkBackingFit::kExact,
kRenderTargetWidth,
kRenderTargetHeight,
kRGBA_8888_GrPixelConfig,
nullptr));
if (!drawContext) {
SkDebugf("Could not allocate a drawContext");
return false;
}
int fpFactoryCnt = GrProcessorTestFactory<GrFragmentProcessor>::Count();
for (int i = 0; i < fpFactoryCnt; ++i) {
// Since FP factories internally randomize, call each 10 times.
for (int j = 0; j < 10; ++j) {
SkAutoTUnref<GrDrawBatch> batch(GrRandomDrawBatch(&random, context));
SkASSERT(batch);
GrProcessorTestData ptd(&random, context, context->caps(),
drawContext.get(), dummyTextures);
GrPaint grPaint;
grPaint.setXPFactory(GrPorterDuffXPFactory::Make(SkXfermode::kSrc_Mode));
sk_sp<GrFragmentProcessor> fp(
GrProcessorTestFactory<GrFragmentProcessor>::MakeIdx(i, &ptd));
sk_sp<GrFragmentProcessor> blockFP(
BlockInputFragmentProcessor::Make(std::move(fp)));
grPaint.addColorFragmentProcessor(std::move(blockFP));
drawContext->drawContextPriv().testingOnly_drawBatch(grPaint, batch);
drawingManager->flush();
}
}
return true;
}
static int get_glprograms_max_stages(GrContext* context) {
GrGLGpu* gpu = static_cast<GrGLGpu*>(context->getGpu());
/*
* For the time being, we only support the test with desktop GL or for android on
* ARM platforms
* TODO When we run ES 3.00 GLSL in more places, test again
*/
if (kGL_GrGLStandard == gpu->glStandard() ||
kARM_GrGLVendor == gpu->ctxInfo().vendor()) {
return 6;
} else if (kTegra3_GrGLRenderer == gpu->ctxInfo().renderer() ||
kOther_GrGLRenderer == gpu->ctxInfo().renderer()) {
return 1;
}
return 0;
}
static void test_glprograms_native(skiatest::Reporter* reporter,
const sk_gpu_test::ContextInfo& ctxInfo) {
int maxStages = get_glprograms_max_stages(ctxInfo.grContext());
if (maxStages == 0) {
return;
}
REPORTER_ASSERT(reporter, GrDrawingManager::ProgramUnitTest(ctxInfo.grContext(), maxStages));
}
static void test_glprograms_other_contexts(
skiatest::Reporter* reporter,
const sk_gpu_test::ContextInfo& ctxInfo) {
int maxStages = get_glprograms_max_stages(ctxInfo.grContext());
#ifdef SK_BUILD_FOR_WIN
// Some long shaders run out of temporary registers in the D3D compiler on ANGLE and
// command buffer.
maxStages = SkTMin(maxStages, 2);
#endif
if (maxStages == 0) {
return;
}
REPORTER_ASSERT(reporter, GrDrawingManager::ProgramUnitTest(ctxInfo.grContext(), maxStages));
}
static bool is_native_gl_context_type(sk_gpu_test::GrContextFactory::ContextType type) {
return type == sk_gpu_test::GrContextFactory::kNativeGL_ContextType;
}
static bool is_other_rendering_gl_context_type(sk_gpu_test::GrContextFactory::ContextType type) {
return !is_native_gl_context_type(type) &&
kOpenGL_GrBackend == sk_gpu_test::GrContextFactory::ContextTypeBackend(type) &&
sk_gpu_test::GrContextFactory::IsRenderingContext(type);
}
DEF_GPUTEST(GLPrograms, reporter, /*factory*/) {
// Set a locale that would cause shader compilation to fail because of , as decimal separator.
// skbug 3330
#ifdef SK_BUILD_FOR_WIN
GrAutoLocaleSetter als("sv-SE");
#else
GrAutoLocaleSetter als("sv_SE.UTF-8");
#endif
// We suppress prints to avoid spew
GrContextOptions opts;
opts.fSuppressPrints = true;
sk_gpu_test::GrContextFactory debugFactory(opts);
skiatest::RunWithGPUTestContexts(test_glprograms_native, &is_native_gl_context_type,
reporter, &debugFactory);
skiatest::RunWithGPUTestContexts(test_glprograms_other_contexts,
&is_other_rendering_gl_context_type, reporter, &debugFactory);
}
#endif