src/gpu/GrDrawingManager.cpp - platform/external/skia - Gitiles

 /*
  * Copyright 2015 Google Inc.
  *
  * Use of this source code is governed by a BSD-style license that can be
  * found in the LICENSE file.
  */

 #include "GrDrawingManager.h"

 #include "GrBackendSemaphore.h"
 #include "GrContext.h"
 #include "GrGpu.h"
 #include "GrOnFlushResourceProvider.h"
 #include "GrRenderTargetContext.h"
 #include "GrPathRenderingRenderTargetContext.h"
 #include "GrRenderTargetProxy.h"
 #include "GrResourceProvider.h"
 #include "GrSoftwarePathRenderer.h"
 #include "GrSurfaceProxyPriv.h"
 #include "GrTextureContext.h"
 #include "GrTextureOpList.h"
 #include "SkSurface_Gpu.h"
 #include "SkTTopoSort.h"

 #include "GrTracing.h"
 #include "text/GrAtlasTextContext.h"
 #include "text/GrStencilAndCoverTextContext.h"

 void GrDrawingManager::cleanup() {
     for (int i = 0; i < fOpLists.count(); ++i) {
         // no opList should receive a new command after this
         fOpLists[i]->makeClosed(*fContext->caps());

         // We shouldn't need to do this, but it turns out some clients still hold onto opLists
         // after a cleanup.
         // MDB TODO: is this still true?
         fOpLists[i]->reset();
     }

     fOpLists.reset();

     delete fPathRendererChain;
     fPathRendererChain = nullptr;
     SkSafeSetNull(fSoftwarePathRenderer);
 }

 GrDrawingManager::~GrDrawingManager() {
     this->cleanup();
 }

 void GrDrawingManager::abandon() {
     fAbandoned = true;
     for (int i = 0; i < fOpLists.count(); ++i) {
         fOpLists[i]->abandonGpuResources();
     }
     this->cleanup();
 }

 void GrDrawingManager::freeGpuResources() {
     // a path renderer may be holding onto resources
     delete fPathRendererChain;
     fPathRendererChain = nullptr;
     SkSafeSetNull(fSoftwarePathRenderer);
     for (int i = 0; i < fOpLists.count(); ++i) {
         fOpLists[i]->freeGpuResources();
     }
 }

 void GrDrawingManager::reset() {
     for (int i = 0; i < fOpLists.count(); ++i) {
         fOpLists[i]->reset();
     }
     fFlushState.reset();
 }

 gr_instanced::OpAllocator* GrDrawingManager::instancingAllocator() {
     if (fInstancingAllocator) {
         return fInstancingAllocator.get();
     }

     fInstancingAllocator = fContext->getGpu()->createInstancedRenderingAllocator();
     return fInstancingAllocator.get();
 }

 // MDB TODO: make use of the 'proxy' parameter.
 GrSemaphoresSubmitted GrDrawingManager::internalFlush(GrSurfaceProxy*,
                                                       GrResourceCache::FlushType type,
                                                       int numSemaphores,
                                                       GrBackendSemaphore backendSemaphores[]) {
     GR_CREATE_TRACE_MARKER_CONTEXT("GrDrawingManager", "internalFlush", fContext);

     if (fFlushing || this->wasAbandoned()) {
         return GrSemaphoresSubmitted::kNo;
     }
     fFlushing = true;
     bool flushed = false;

     for (int i = 0; i < fOpLists.count(); ++i) {
         // Semi-usually the GrOpLists are already closed at this point, but sometimes Ganesh
         // needs to flush mid-draw. In that case, the SkGpuDevice's GrOpLists won't be closed
         // but need to be flushed anyway. Closing such GrOpLists here will mean new
         // GrOpLists will be created to replace them if the SkGpuDevice(s) write to them again.
         fOpLists[i]->makeClosed(*fContext->caps());
     }

 #ifdef SK_DEBUG
     // This block checks for any unnecessary splits in the opLists. If two sequential opLists
     // share the same backing GrSurfaceProxy it means the opList was artificially split.
     if (fOpLists.count()) {
         GrRenderTargetOpList* prevOpList = fOpLists[0]->asRenderTargetOpList();
         for (int i = 1; i < fOpLists.count(); ++i) {
             GrRenderTargetOpList* curOpList = fOpLists[i]->asRenderTargetOpList();

             if (prevOpList && curOpList) {
                 SkASSERT(prevOpList->fTarget.get() != curOpList->fTarget.get());
             }

             prevOpList = curOpList;
         }
     }
 #endif

 #ifdef ENABLE_MDB
     SkDEBUGCODE(bool result =)
                         SkTTopoSort<GrOpList, GrOpList::TopoSortTraits>(&fOpLists);
     SkASSERT(result);
 #endif

     GrOnFlushResourceProvider onFlushProvider(this);

     if (!fOnFlushCBObjects.empty()) {
         // MDB TODO: pre-MDB '1' is the correct pre-allocated size. Post-MDB it will need
         // to be larger.
         SkAutoSTArray<1, uint32_t> opListIds(fOpLists.count());
         for (int i = 0; i < fOpLists.count(); ++i) {
             opListIds[i] = fOpLists[i]->uniqueID();
         }

         SkSTArray<1, sk_sp<GrRenderTargetContext>> renderTargetContexts;
         for (GrOnFlushCallbackObject* onFlushCBObject : fOnFlushCBObjects) {
             onFlushCBObject->preFlush(&onFlushProvider,
                                       opListIds.get(), opListIds.count(),
                                       &renderTargetContexts);
             if (!renderTargetContexts.count()) {
                 continue;       // This is fine. No atlases of this type are required for this flush
             }

             for (int j = 0; j < renderTargetContexts.count(); ++j) {
                 GrOpList* opList = renderTargetContexts[j]->getOpList();
                 if (!opList) {
                     continue;   // Odd - but not a big deal
                 }
                 opList->makeClosed(*fContext->caps());
                 opList->prepare(&fFlushState);
                 if (!opList->execute(&fFlushState)) {
                     continue;         // This is bad
                 }
             }
             renderTargetContexts.reset();
         }
     }

 #if 0
     // Enable this to print out verbose GrOp information
     for (int i = 0; i < fOpLists.count(); ++i) {
         SkDEBUGCODE(fOpLists[i]->dump();)
     }
 #endif

     for (int i = 0; i < fOpLists.count(); ++i) {
         if (!fOpLists[i]->instantiate(fContext->resourceProvider())) {
             SkDebugf("OpList failed to instantiate.\n");
             fOpLists[i] = nullptr;
             continue;
         }

         fOpLists[i]->prepare(&fFlushState);
     }

     // Upload all data to the GPU
     fFlushState.preIssueDraws();

     for (int i = 0; i < fOpLists.count(); ++i) {
         if (!fOpLists[i]) {
             continue;
         }

         if (fOpLists[i]->execute(&fFlushState)) {
             flushed = true;
         }
         fOpLists[i]->reset();
     }
     fOpLists.reset();

     SkASSERT(fFlushState.nextDrawToken() == fFlushState.nextTokenToFlush());

     GrSemaphoresSubmitted result = fContext->getGpu()->finishFlush(numSemaphores,
                                                                    backendSemaphores);

     fFlushState.reset();
     // We always have to notify the cache when it requested a flush so it can reset its state.
     if (flushed || type == GrResourceCache::FlushType::kCacheRequested) {
         fContext->getResourceCache()->notifyFlushOccurred(type);
     }
     for (GrOnFlushCallbackObject* onFlushCBObject : fOnFlushCBObjects) {
         onFlushCBObject->postFlush();
     }
     fFlushing = false;

     return result;
 }

 GrSemaphoresSubmitted GrDrawingManager::prepareSurfaceForExternalIO(
         GrSurfaceProxy* proxy, int numSemaphores, GrBackendSemaphore backendSemaphores[]) {
     if (this->wasAbandoned()) {
         return GrSemaphoresSubmitted::kNo;
     }
     SkASSERT(proxy);

     GrSemaphoresSubmitted result;
     if (proxy->priv().hasPendingIO() || numSemaphores) {
         result = this->flush(proxy, numSemaphores, backendSemaphores);
     }

     if (!proxy->instantiate(fContext->resourceProvider())) {
         return result;
     }

     GrSurface* surface = proxy->priv().peekSurface();

     if (fContext->getGpu() && surface->asRenderTarget()) {
         fContext->getGpu()->resolveRenderTarget(surface->asRenderTarget(), proxy->origin());
     }
     return result;
 }

 void GrDrawingManager::addOnFlushCallbackObject(GrOnFlushCallbackObject* onFlushCBObject) {
     fOnFlushCBObjects.push_back(onFlushCBObject);
 }

 sk_sp<GrRenderTargetOpList> GrDrawingManager::newRTOpList(GrRenderTargetProxy* rtp,
                                                           bool managedOpList) {
     SkASSERT(fContext);

     // This is  a temporary fix for the partial-MDB world. In that world we're not reordering
     // so ops that (in the single opList world) would've just glommed onto the end of the single
     // opList but referred to a far earlier RT need to appear in their own opList.
     if (!fOpLists.empty()) {
         fOpLists.back()->makeClosed(*fContext->caps());
     }

     sk_sp<GrRenderTargetOpList> opList(new GrRenderTargetOpList(rtp,
                                                                 fContext->getGpu(),
                                                                 fContext->getAuditTrail()));
     SkASSERT(rtp->getLastOpList() == opList.get());

     if (managedOpList) {
         fOpLists.push_back() = opList;
     }

     return opList;
 }

 sk_sp<GrTextureOpList> GrDrawingManager::newTextureOpList(GrTextureProxy* textureProxy) {
     SkASSERT(fContext);

     // This is  a temporary fix for the partial-MDB world. In that world we're not reordering
     // so ops that (in the single opList world) would've just glommed onto the end of the single
     // opList but referred to a far earlier RT need to appear in their own opList.
     if (!fOpLists.empty()) {
         fOpLists.back()->makeClosed(*fContext->caps());
     }

     sk_sp<GrTextureOpList> opList(new GrTextureOpList(fContext->resourceProvider(),
                                                       textureProxy,
                                                       fContext->getAuditTrail()));

     SkASSERT(textureProxy->getLastOpList() == opList.get());

     fOpLists.push_back() = opList;

     return opList;
 }

 GrAtlasTextContext* GrDrawingManager::getAtlasTextContext() {
     if (!fAtlasTextContext) {
         fAtlasTextContext.reset(GrAtlasTextContext::Create());
     }

     return fAtlasTextContext.get();
 }

 /*
  * This method finds a path renderer that can draw the specified path on
  * the provided target.
  * Due to its expense, the software path renderer has split out so it can
  * can be individually allowed/disallowed via the "allowSW" boolean.
  */
 GrPathRenderer* GrDrawingManager::getPathRenderer(const GrPathRenderer::CanDrawPathArgs& args,
                                                   bool allowSW,
                                                   GrPathRendererChain::DrawType drawType,
                                                   GrPathRenderer::StencilSupport* stencilSupport) {

     if (!fPathRendererChain) {
         fPathRendererChain = new GrPathRendererChain(fContext, fOptionsForPathRendererChain);
     }

     GrPathRenderer* pr = fPathRendererChain->getPathRenderer(args, drawType, stencilSupport);
     if (!pr && allowSW) {
         if (!fSoftwarePathRenderer) {
             fSoftwarePathRenderer =
                     new GrSoftwarePathRenderer(fContext->resourceProvider(),
                                                fOptionsForPathRendererChain.fAllowPathMaskCaching);
         }
         if (GrPathRenderer::CanDrawPath::kNo != fSoftwarePathRenderer->canDrawPath(args)) {
             pr = fSoftwarePathRenderer;
         }
     }

     return pr;
 }

 sk_sp<GrRenderTargetContext> GrDrawingManager::makeRenderTargetContext(
                                                             sk_sp<GrSurfaceProxy> sProxy,
                                                             sk_sp<SkColorSpace> colorSpace,
                                                             const SkSurfaceProps* surfaceProps,
                                                             bool managedOpList) {
     if (this->wasAbandoned() || !sProxy->asRenderTargetProxy()) {
         return nullptr;
     }

     // SkSurface catches bad color space usage at creation. This check handles anything that slips
     // by, including internal usage. We allow a null color space here, for read/write pixels and
     // other special code paths. If a color space is provided, though, enforce all other rules.
     if (colorSpace && !SkSurface_Gpu::Valid(fContext, sProxy->config(), colorSpace.get())) {
         SkDEBUGFAIL("Invalid config and colorspace combination");
         return nullptr;
     }

     sk_sp<GrRenderTargetProxy> rtp(sk_ref_sp(sProxy->asRenderTargetProxy()));

     bool useDIF = false;
     if (surfaceProps) {
         useDIF = surfaceProps->isUseDeviceIndependentFonts();
     }

     if (useDIF && fContext->caps()->shaderCaps()->pathRenderingSupport() &&
         GrFSAAType::kNone != rtp->fsaaType()) {
         // TODO: defer stencil buffer attachment for PathRenderingDrawContext
         if (!rtp->instantiate(fContext->resourceProvider())) {
             return nullptr;
         }
         GrRenderTarget* rt = rtp->priv().peekRenderTarget();

         GrStencilAttachment* sb = fContext->resourceProvider()->attachStencilAttachment(rt);
         if (sb) {
             return sk_sp<GrRenderTargetContext>(new GrPathRenderingRenderTargetContext(
                                                         fContext, this, std::move(rtp),
                                                         std::move(colorSpace), surfaceProps,
                                                         fContext->getAuditTrail(), fSingleOwner));
         }
     }

     return sk_sp<GrRenderTargetContext>(new GrRenderTargetContext(fContext, this, std::move(rtp),
                                                                   std::move(colorSpace),
                                                                   surfaceProps,
                                                                   fContext->getAuditTrail(),
                                                                   fSingleOwner, managedOpList));
 }

 sk_sp<GrTextureContext> GrDrawingManager::makeTextureContext(sk_sp<GrSurfaceProxy> sProxy,
                                                              sk_sp<SkColorSpace> colorSpace) {
     if (this->wasAbandoned() || !sProxy->asTextureProxy()) {
         return nullptr;
     }

     // SkSurface catches bad color space usage at creation. This check handles anything that slips
     // by, including internal usage. We allow a null color space here, for read/write pixels and
     // other special code paths. If a color space is provided, though, enforce all other rules.
     if (colorSpace && !SkSurface_Gpu::Valid(fContext, sProxy->config(), colorSpace.get())) {
         SkDEBUGFAIL("Invalid config and colorspace combination");
         return nullptr;
     }

     // GrTextureRenderTargets should always be using GrRenderTargetContext
     SkASSERT(!sProxy->asRenderTargetProxy());

     sk_sp<GrTextureProxy> textureProxy(sk_ref_sp(sProxy->asTextureProxy()));

     return sk_sp<GrTextureContext>(new GrTextureContext(fContext, this, std::move(textureProxy),
                                                         std::move(colorSpace),
                                                         fContext->getAuditTrail(),
                                                         fSingleOwner));
 }
	/*
	* Copyright 2015 Google Inc.
	*
	* Use of this source code is governed by a BSD-style license that can be
	* found in the LICENSE file.
	*/

	#include "GrDrawingManager.h"

	#include "GrBackendSemaphore.h"
	#include "GrContext.h"
	#include "GrGpu.h"
	#include "GrOnFlushResourceProvider.h"
	#include "GrRenderTargetContext.h"
	#include "GrPathRenderingRenderTargetContext.h"
	#include "GrRenderTargetProxy.h"
	#include "GrResourceProvider.h"
	#include "GrSoftwarePathRenderer.h"
	#include "GrSurfaceProxyPriv.h"
	#include "GrTextureContext.h"
	#include "GrTextureOpList.h"
	#include "SkSurface_Gpu.h"
	#include "SkTTopoSort.h"

	#include "GrTracing.h"
	#include "text/GrAtlasTextContext.h"
	#include "text/GrStencilAndCoverTextContext.h"

	void GrDrawingManager::cleanup() {
	for (int i = 0; i < fOpLists.count(); ++i) {
	// no opList should receive a new command after this
	fOpLists[i]->makeClosed(*fContext->caps());

	// We shouldn't need to do this, but it turns out some clients still hold onto opLists
	// after a cleanup.
	// MDB TODO: is this still true?
	fOpLists[i]->reset();
	}

	fOpLists.reset();

	delete fPathRendererChain;
	fPathRendererChain = nullptr;
	SkSafeSetNull(fSoftwarePathRenderer);
	}

	GrDrawingManager::~GrDrawingManager() {
	this->cleanup();
	}

	void GrDrawingManager::abandon() {
	fAbandoned = true;
	for (int i = 0; i < fOpLists.count(); ++i) {
	fOpLists[i]->abandonGpuResources();
	}
	this->cleanup();
	}

	void GrDrawingManager::freeGpuResources() {
	// a path renderer may be holding onto resources
	delete fPathRendererChain;
	fPathRendererChain = nullptr;
	SkSafeSetNull(fSoftwarePathRenderer);
	for (int i = 0; i < fOpLists.count(); ++i) {
	fOpLists[i]->freeGpuResources();
	}
	}

	void GrDrawingManager::reset() {
	for (int i = 0; i < fOpLists.count(); ++i) {
	fOpLists[i]->reset();
	}
	fFlushState.reset();
	}

	gr_instanced::OpAllocator* GrDrawingManager::instancingAllocator() {
	if (fInstancingAllocator) {
	return fInstancingAllocator.get();
	}

	fInstancingAllocator = fContext->getGpu()->createInstancedRenderingAllocator();
	return fInstancingAllocator.get();
	}

	// MDB TODO: make use of the 'proxy' parameter.
	GrSemaphoresSubmitted GrDrawingManager::internalFlush(GrSurfaceProxy*,
	GrResourceCache::FlushType type,
	int numSemaphores,
	GrBackendSemaphore backendSemaphores[]) {
	GR_CREATE_TRACE_MARKER_CONTEXT("GrDrawingManager", "internalFlush", fContext);

	if (fFlushing \|\| this->wasAbandoned()) {
	return GrSemaphoresSubmitted::kNo;
	}
	fFlushing = true;
	bool flushed = false;

	for (int i = 0; i < fOpLists.count(); ++i) {
	// Semi-usually the GrOpLists are already closed at this point, but sometimes Ganesh
	// needs to flush mid-draw. In that case, the SkGpuDevice's GrOpLists won't be closed
	// but need to be flushed anyway. Closing such GrOpLists here will mean new
	// GrOpLists will be created to replace them if the SkGpuDevice(s) write to them again.
	fOpLists[i]->makeClosed(*fContext->caps());
	}

	#ifdef SK_DEBUG
	// This block checks for any unnecessary splits in the opLists. If two sequential opLists
	// share the same backing GrSurfaceProxy it means the opList was artificially split.
	if (fOpLists.count()) {
	GrRenderTargetOpList* prevOpList = fOpLists[0]->asRenderTargetOpList();
	for (int i = 1; i < fOpLists.count(); ++i) {
	GrRenderTargetOpList* curOpList = fOpLists[i]->asRenderTargetOpList();

	if (prevOpList && curOpList) {
	SkASSERT(prevOpList->fTarget.get() != curOpList->fTarget.get());
	}

	prevOpList = curOpList;
	}
	}
	#endif

	#ifdef ENABLE_MDB
	SkDEBUGCODE(bool result =)
	SkTTopoSort<GrOpList, GrOpList::TopoSortTraits>(&fOpLists);
	SkASSERT(result);
	#endif

	GrOnFlushResourceProvider onFlushProvider(this);

	if (!fOnFlushCBObjects.empty()) {
	// MDB TODO: pre-MDB '1' is the correct pre-allocated size. Post-MDB it will need
	// to be larger.
	SkAutoSTArray<1, uint32_t> opListIds(fOpLists.count());
	for (int i = 0; i < fOpLists.count(); ++i) {
	opListIds[i] = fOpLists[i]->uniqueID();
	}

	SkSTArray<1, sk_sp<GrRenderTargetContext>> renderTargetContexts;
	for (GrOnFlushCallbackObject* onFlushCBObject : fOnFlushCBObjects) {
	onFlushCBObject->preFlush(&onFlushProvider,
	opListIds.get(), opListIds.count(),
	&renderTargetContexts);
	if (!renderTargetContexts.count()) {
	continue; // This is fine. No atlases of this type are required for this flush
	}

	for (int j = 0; j < renderTargetContexts.count(); ++j) {
	GrOpList* opList = renderTargetContexts[j]->getOpList();
	if (!opList) {
	continue; // Odd - but not a big deal
	}
	opList->makeClosed(*fContext->caps());
	opList->prepare(&fFlushState);
	if (!opList->execute(&fFlushState)) {
	continue; // This is bad
	}
	}
	renderTargetContexts.reset();
	}
	}

	#if 0
	// Enable this to print out verbose GrOp information
	for (int i = 0; i < fOpLists.count(); ++i) {
	SkDEBUGCODE(fOpLists[i]->dump();)
	}
	#endif

	for (int i = 0; i < fOpLists.count(); ++i) {
	if (!fOpLists[i]->instantiate(fContext->resourceProvider())) {
	SkDebugf("OpList failed to instantiate.\n");
	fOpLists[i] = nullptr;
	continue;
	}

	fOpLists[i]->prepare(&fFlushState);
	}

	// Upload all data to the GPU
	fFlushState.preIssueDraws();

	for (int i = 0; i < fOpLists.count(); ++i) {
	if (!fOpLists[i]) {
	continue;
	}

	if (fOpLists[i]->execute(&fFlushState)) {
	flushed = true;
	}
	fOpLists[i]->reset();
	}
	fOpLists.reset();

	SkASSERT(fFlushState.nextDrawToken() == fFlushState.nextTokenToFlush());

	GrSemaphoresSubmitted result = fContext->getGpu()->finishFlush(numSemaphores,
	backendSemaphores);

	fFlushState.reset();
	// We always have to notify the cache when it requested a flush so it can reset its state.
	if (flushed \|\| type == GrResourceCache::FlushType::kCacheRequested) {
	fContext->getResourceCache()->notifyFlushOccurred(type);
	}
	for (GrOnFlushCallbackObject* onFlushCBObject : fOnFlushCBObjects) {
	onFlushCBObject->postFlush();
	}
	fFlushing = false;

	return result;
	}

	GrSemaphoresSubmitted GrDrawingManager::prepareSurfaceForExternalIO(
	GrSurfaceProxy* proxy, int numSemaphores, GrBackendSemaphore backendSemaphores[]) {
	if (this->wasAbandoned()) {
	return GrSemaphoresSubmitted::kNo;
	}
	SkASSERT(proxy);

	GrSemaphoresSubmitted result;
	if (proxy->priv().hasPendingIO() \|\| numSemaphores) {
	result = this->flush(proxy, numSemaphores, backendSemaphores);
	}

	if (!proxy->instantiate(fContext->resourceProvider())) {
	return result;
	}

	GrSurface* surface = proxy->priv().peekSurface();

	if (fContext->getGpu() && surface->asRenderTarget()) {
	fContext->getGpu()->resolveRenderTarget(surface->asRenderTarget(), proxy->origin());
	}
	return result;
	}

	void GrDrawingManager::addOnFlushCallbackObject(GrOnFlushCallbackObject* onFlushCBObject) {
	fOnFlushCBObjects.push_back(onFlushCBObject);
	}

	sk_sp<GrRenderTargetOpList> GrDrawingManager::newRTOpList(GrRenderTargetProxy* rtp,
	bool managedOpList) {
	SkASSERT(fContext);

	// This is a temporary fix for the partial-MDB world. In that world we're not reordering
	// so ops that (in the single opList world) would've just glommed onto the end of the single
	// opList but referred to a far earlier RT need to appear in their own opList.
	if (!fOpLists.empty()) {
	fOpLists.back()->makeClosed(*fContext->caps());
	}

	sk_sp<GrRenderTargetOpList> opList(new GrRenderTargetOpList(rtp,
	fContext->getGpu(),
	fContext->getAuditTrail()));
	SkASSERT(rtp->getLastOpList() == opList.get());

	if (managedOpList) {
	fOpLists.push_back() = opList;
	}

	return opList;
	}

	sk_sp<GrTextureOpList> GrDrawingManager::newTextureOpList(GrTextureProxy* textureProxy) {
	SkASSERT(fContext);

	// This is a temporary fix for the partial-MDB world. In that world we're not reordering
	// so ops that (in the single opList world) would've just glommed onto the end of the single
	// opList but referred to a far earlier RT need to appear in their own opList.
	if (!fOpLists.empty()) {
	fOpLists.back()->makeClosed(*fContext->caps());
	}

	sk_sp<GrTextureOpList> opList(new GrTextureOpList(fContext->resourceProvider(),
	textureProxy,
	fContext->getAuditTrail()));

	SkASSERT(textureProxy->getLastOpList() == opList.get());

	fOpLists.push_back() = opList;

	return opList;
	}

	GrAtlasTextContext* GrDrawingManager::getAtlasTextContext() {
	if (!fAtlasTextContext) {
	fAtlasTextContext.reset(GrAtlasTextContext::Create());
	}

	return fAtlasTextContext.get();
	}

	/*
	* This method finds a path renderer that can draw the specified path on
	* the provided target.
	* Due to its expense, the software path renderer has split out so it can
	* can be individually allowed/disallowed via the "allowSW" boolean.
	*/
	GrPathRenderer* GrDrawingManager::getPathRenderer(const GrPathRenderer::CanDrawPathArgs& args,
	bool allowSW,
	GrPathRendererChain::DrawType drawType,
	GrPathRenderer::StencilSupport* stencilSupport) {

	if (!fPathRendererChain) {
	fPathRendererChain = new GrPathRendererChain(fContext, fOptionsForPathRendererChain);
	}

	GrPathRenderer* pr = fPathRendererChain->getPathRenderer(args, drawType, stencilSupport);
	if (!pr && allowSW) {
	if (!fSoftwarePathRenderer) {
	fSoftwarePathRenderer =
	new GrSoftwarePathRenderer(fContext->resourceProvider(),
	fOptionsForPathRendererChain.fAllowPathMaskCaching);
	}
	if (GrPathRenderer::CanDrawPath::kNo != fSoftwarePathRenderer->canDrawPath(args)) {
	pr = fSoftwarePathRenderer;
	}
	}

	return pr;
	}

	sk_sp<GrRenderTargetContext> GrDrawingManager::makeRenderTargetContext(
	sk_sp<GrSurfaceProxy> sProxy,
	sk_sp<SkColorSpace> colorSpace,
	const SkSurfaceProps* surfaceProps,
	bool managedOpList) {
	if (this->wasAbandoned() \|\| !sProxy->asRenderTargetProxy()) {
	return nullptr;
	}

	// SkSurface catches bad color space usage at creation. This check handles anything that slips
	// by, including internal usage. We allow a null color space here, for read/write pixels and
	// other special code paths. If a color space is provided, though, enforce all other rules.
	if (colorSpace && !SkSurface_Gpu::Valid(fContext, sProxy->config(), colorSpace.get())) {
	SkDEBUGFAIL("Invalid config and colorspace combination");
	return nullptr;
	}

	sk_sp<GrRenderTargetProxy> rtp(sk_ref_sp(sProxy->asRenderTargetProxy()));

	bool useDIF = false;
	if (surfaceProps) {
	useDIF = surfaceProps->isUseDeviceIndependentFonts();
	}

	if (useDIF && fContext->caps()->shaderCaps()->pathRenderingSupport() &&
	GrFSAAType::kNone != rtp->fsaaType()) {
	// TODO: defer stencil buffer attachment for PathRenderingDrawContext
	if (!rtp->instantiate(fContext->resourceProvider())) {
	return nullptr;
	}
	GrRenderTarget* rt = rtp->priv().peekRenderTarget();

	GrStencilAttachment* sb = fContext->resourceProvider()->attachStencilAttachment(rt);
	if (sb) {
	return sk_sp<GrRenderTargetContext>(new GrPathRenderingRenderTargetContext(
	fContext, this, std::move(rtp),
	std::move(colorSpace), surfaceProps,
	fContext->getAuditTrail(), fSingleOwner));
	}
	}

	return sk_sp<GrRenderTargetContext>(new GrRenderTargetContext(fContext, this, std::move(rtp),
	std::move(colorSpace),
	surfaceProps,
	fContext->getAuditTrail(),
	fSingleOwner, managedOpList));
	}

	sk_sp<GrTextureContext> GrDrawingManager::makeTextureContext(sk_sp<GrSurfaceProxy> sProxy,
	sk_sp<SkColorSpace> colorSpace) {
	if (this->wasAbandoned() \|\| !sProxy->asTextureProxy()) {
	return nullptr;
	}

	// SkSurface catches bad color space usage at creation. This check handles anything that slips
	// by, including internal usage. We allow a null color space here, for read/write pixels and
	// other special code paths. If a color space is provided, though, enforce all other rules.
	if (colorSpace && !SkSurface_Gpu::Valid(fContext, sProxy->config(), colorSpace.get())) {
	SkDEBUGFAIL("Invalid config and colorspace combination");
	return nullptr;
	}

	// GrTextureRenderTargets should always be using GrRenderTargetContext
	SkASSERT(!sProxy->asRenderTargetProxy());

	sk_sp<GrTextureProxy> textureProxy(sk_ref_sp(sProxy->asTextureProxy()));

	return sk_sp<GrTextureContext>(new GrTextureContext(fContext, this, std::move(textureProxy),
	std::move(colorSpace),
	fContext->getAuditTrail(),
	fSingleOwner));
	}