/* * 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 "GrContextFactory.h" #include "GrInvariantOutput.h" #include "GrPipeline.h" #include "GrTest.h" #include "GrXferProcessor.h" #include "SkChecksum.h" #include "SkRandom.h" #include "Test.h" #include "effects/GrConfigConversionEffect.h" #include "effects/GrPorterDuffXferProcessor.h" #include "gl/GrGLGpu.h" #include "gl/GrGLPathRendering.h" #include "gl/builders/GrGLProgramBuilder.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 GrGLFragmentProcessor { public: GLBigKeyProcessor(const GrProcessor&) {} virtual void emitCode(GrGLFPBuilder* builder, const GrFragmentProcessor& fp, const char* outputColor, const char* inputColor, const TransformedCoordsArray&, const TextureSamplerArray&) {} static void GenKey(const GrProcessor& processor, const GrGLCaps&, GrProcessorKeyBuilder* b) { for (uint32_t i = 0; i < kMaxKeySize; i++) { b->add32(i); } } private: typedef GrGLFragmentProcessor INHERITED; }; class BigKeyProcessor : public GrFragmentProcessor { public: static GrFragmentProcessor* Create() { GR_CREATE_STATIC_PROCESSOR(gBigKeyProcessor, BigKeyProcessor, ()) return SkRef(gBigKeyProcessor); } const char* name() const SK_OVERRIDE { return "Big Ole Key"; } virtual void getGLProcessorKey(const GrGLCaps& caps, GrProcessorKeyBuilder* b) const SK_OVERRIDE { GLBigKeyProcessor::GenKey(*this, caps, b); } GrGLFragmentProcessor* createGLInstance() const SK_OVERRIDE { return SkNEW_ARGS(GLBigKeyProcessor, (*this)); } private: BigKeyProcessor() { this->initClassID(); } bool onIsEqual(const GrFragmentProcessor&) const SK_OVERRIDE { return true; } void onComputeInvariantOutput(GrInvariantOutput* inout) const SK_OVERRIDE { } GR_DECLARE_FRAGMENT_PROCESSOR_TEST; typedef GrFragmentProcessor INHERITED; }; GR_DEFINE_FRAGMENT_PROCESSOR_TEST(BigKeyProcessor); GrFragmentProcessor* BigKeyProcessor::TestCreate(SkRandom*, GrContext*, const GrDrawTargetCaps&, GrTexture*[]) { return BigKeyProcessor::Create(); } /* * Begin test code */ static const int kRenderTargetHeight = 1; static const int kRenderTargetWidth = 1; static GrRenderTarget* random_render_target(GrContext* context, SkRandom* random) { // setup render target GrTextureParams params; GrSurfaceDesc texDesc; texDesc.fWidth = kRenderTargetWidth; texDesc.fHeight = kRenderTargetHeight; texDesc.fFlags = kRenderTarget_GrSurfaceFlag; texDesc.fConfig = kRGBA_8888_GrPixelConfig; texDesc.fOrigin = random->nextBool() == true ? kTopLeft_GrSurfaceOrigin : kBottomLeft_GrSurfaceOrigin; GrContentKey key; static const GrContentKey::Domain kDomain = GrContentKey::GenerateDomain(); GrContentKey::Builder builder(&key, kDomain, 1); builder[0] = texDesc.fOrigin; builder.finish(); GrTexture* texture = context->findAndRefCachedTexture(key); if (!texture) { texture = context->createTexture(texDesc, true); if (texture) { SkAssertResult(context->addResourceToCache(key, texture)); } } return texture ? texture->asRenderTarget() : NULL; } static void set_random_xpf(GrContext* context, const GrDrawTargetCaps& caps, GrPipelineBuilder* pipelineBuilder, SkRandom* random, GrTexture* dummyTextures[]) { SkAutoTUnref xpf( GrProcessorTestFactory::CreateStage(random, context, caps, dummyTextures)); SkASSERT(xpf); pipelineBuilder->setXPFactory(xpf.get()); } static const GrGeometryProcessor* get_random_gp(GrContext* context, const GrDrawTargetCaps& caps, SkRandom* random, GrTexture* dummyTextures[]) { return GrProcessorTestFactory::CreateStage(random, context, caps, dummyTextures); } static void set_random_color_coverage_stages(GrGLGpu* gpu, GrPipelineBuilder* pipelineBuilder, int maxStages, bool usePathRendering, SkRandom* random, GrTexture* dummyTextures[]) { int numProcs = random->nextULessThan(maxStages + 1); int numColorProcs = random->nextULessThan(numProcs + 1); int currTextureCoordSet = 0; for (int s = 0; s < numProcs;) { SkAutoTUnref fp( GrProcessorTestFactory::CreateStage(random, gpu->getContext(), *gpu->caps(), dummyTextures)); SkASSERT(fp); // If adding this effect would exceed the max texture coord set count then generate a // new random effect. if (usePathRendering && gpu->glPathRendering()->texturingMode() == GrGLPathRendering::FixedFunction_TexturingMode) {; int numTransforms = fp->numTransforms(); if (currTextureCoordSet + numTransforms > gpu->glCaps().maxFixedFunctionTextureCoords()) { continue; } currTextureCoordSet += numTransforms; } // finally add the stage to the correct pipeline in the drawstate if (s < numColorProcs) { pipelineBuilder->addColorProcessor(fp); } else { pipelineBuilder->addCoverageProcessor(fp); } ++s; } } static void set_random_state(GrPipelineBuilder* pipelineBuilder, SkRandom* random) { int state = 0; for (int i = 1; i <= GrPipelineBuilder::kLast_StateBit; i <<= 1) { state |= random->nextBool() * i; } pipelineBuilder->enableState(state); } // right now, the only thing we seem to care about in drawState's stencil is 'doesWrite()' static void set_random_stencil(GrPipelineBuilder* pipelineBuilder, SkRandom* random) { GR_STATIC_CONST_SAME_STENCIL(kDoesWriteStencil, kReplace_StencilOp, kReplace_StencilOp, kAlways_StencilFunc, 0xffff, 0xffff, 0xffff); GR_STATIC_CONST_SAME_STENCIL(kDoesNotWriteStencil, kKeep_StencilOp, kKeep_StencilOp, kNever_StencilFunc, 0xffff, 0xffff, 0xffff); if (random->nextBool()) { pipelineBuilder->setStencil(kDoesWriteStencil); } else { pipelineBuilder->setStencil(kDoesNotWriteStencil); } } bool GrDrawTarget::programUnitTest(int maxStages) { GrGLGpu* gpu = static_cast(fContext->getGpu()); // setup dummy textures GrSurfaceDesc dummyDesc; dummyDesc.fFlags = kRenderTarget_GrSurfaceFlag; dummyDesc.fConfig = kSkia8888_GrPixelConfig; dummyDesc.fWidth = 34; dummyDesc.fHeight = 18; SkAutoTUnref dummyTexture1(gpu->createTexture(dummyDesc, false, NULL, 0)); dummyDesc.fFlags = kNone_GrSurfaceFlags; dummyDesc.fConfig = kAlpha_8_GrPixelConfig; dummyDesc.fWidth = 16; dummyDesc.fHeight = 22; SkAutoTUnref dummyTexture2(gpu->createTexture(dummyDesc, false, NULL, 0)); if (!dummyTexture1 || ! dummyTexture2) { SkDebugf("Could not allocate dummy textures"); return false; } GrTexture* dummyTextures[] = {dummyTexture1.get(), dummyTexture2.get()}; // dummy scissor state GrScissorState scissor; // setup clip SkRect screen = SkRect::MakeWH(SkIntToScalar(kRenderTargetWidth), SkIntToScalar(kRenderTargetHeight)); SkClipStack stack; stack.clipDevRect(screen, SkRegion::kReplace_Op, false); // wrap the SkClipStack in a GrClipData GrClipData clipData; clipData.fClipStack.reset(SkRef(&stack)); this->setClip(&clipData); SkRandom random; static const int NUM_TESTS = 512; for (int t = 0; t < NUM_TESTS;) { // setup random render target(can fail) SkAutoTUnref rt(random_render_target(fContext, &random)); if (!rt.get()) { SkDebugf("Could not allocate render target"); return false; } GrPipelineBuilder pipelineBuilder; pipelineBuilder.setRenderTarget(rt.get()); // if path rendering we have to setup a couple of things like the draw type bool usePathRendering = gpu->glCaps().pathRenderingSupport() && random.nextBool(); // twiddle drawstate knobs randomly bool hasGeometryProcessor = !usePathRendering; SkAutoTUnref gp; SkAutoTUnref pathProc; if (hasGeometryProcessor) { gp.reset(get_random_gp(fContext, gpu->glCaps(), &random, dummyTextures)); } else { pathProc.reset(GrPathProcessor::Create(GrColor_WHITE)); } set_random_color_coverage_stages(gpu, &pipelineBuilder, maxStages - hasGeometryProcessor, usePathRendering, &random, dummyTextures); // creates a random xfer processor factory on the draw state set_random_xpf(fContext, gpu->glCaps(), &pipelineBuilder, &random, dummyTextures); set_random_state(&pipelineBuilder, &random); set_random_stencil(&pipelineBuilder, &random); GrDeviceCoordTexture dstCopy; const GrPrimitiveProcessor* primProc; if (hasGeometryProcessor) { primProc = gp.get(); } else { primProc = pathProc.get(); } const GrProcOptInfo& colorPOI = pipelineBuilder.colorProcInfo(primProc); const GrProcOptInfo& coveragePOI = pipelineBuilder.coverageProcInfo(primProc); if (!this->setupDstReadIfNecessary(pipelineBuilder, colorPOI, coveragePOI, &dstCopy, NULL)) { SkDebugf("Couldn't setup dst read texture"); return false; } // create optimized draw state, setup readDst texture if required, and build a descriptor // and program. ODS creation can fail, so we have to check GrPipeline pipeline(pipelineBuilder, colorPOI, coveragePOI, *gpu->caps(), scissor, &dstCopy); if (pipeline.mustSkip()) { continue; } GrBatchTracker bt; primProc->initBatchTracker(&bt, pipeline.getInitBatchTracker()); GrProgramDesc desc; gpu->buildProgramDesc(&desc, *primProc, pipeline, bt); GrGpu::DrawArgs args(primProc, &pipeline, &desc, &bt); SkAutoTUnref program(GrGLProgramBuilder::CreateProgram(args, gpu)); if (NULL == program.get()) { SkDebugf("Failed to create program!"); return false; } // because occasionally optimized drawstate creation will fail for valid reasons, we only // want to increment on success ++t; } return true; } DEF_GPUTEST(GLPrograms, reporter, factory) { for (int type = 0; type < GrContextFactory::kLastGLContextType; ++type) { GrContext* context = factory->get(static_cast(type)); if (context) { GrGLGpu* gpu = static_cast(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 */ int maxStages; if (kGL_GrGLStandard == gpu->glStandard() || kARM_GrGLVendor == gpu->ctxInfo().vendor()) { maxStages = 6; } else if (kTegra3_GrGLRenderer == gpu->ctxInfo().renderer() || kOther_GrGLRenderer == gpu->ctxInfo().renderer()) { maxStages = 1; } else { return; } #if SK_ANGLE // Some long shaders run out of temporary registers in the D3D compiler on ANGLE. if (type == GrContextFactory::kANGLE_GLContextType) { maxStages = 3; } #endif GrTestTarget target; context->getTestTarget(&target); REPORTER_ASSERT(reporter, target.target()->programUnitTest(maxStages)); } } } #endif