src/gpu/GrSoftwarePathRenderer.cpp

/*
 * Copyright 2012 Google Inc.
 *
 * Use of this source code is governed by a BSD-style license that can be
 * found in the LICENSE file.
 */

#include "GrSoftwarePathRenderer.h"
#include "GrAuditTrail.h"
#include "GrClip.h"
#include "GrContextPriv.h"
#include "GrGpuResourcePriv.h"
#include "GrOpFlushState.h"
#include "GrOpList.h"
#include "GrResourceProvider.h"
#include "GrSWMaskHelper.h"
#include "SkMakeUnique.h"
#include "SkSemaphore.h"
#include "SkTaskGroup.h"
#include "SkTraceEvent.h"
#include "ops/GrDrawOp.h"
#include "ops/GrRectOpFactory.h"

////////////////////////////////////////////////////////////////////////////////
GrPathRenderer::CanDrawPath
GrSoftwarePathRenderer::onCanDrawPath(const CanDrawPathArgs& args) const {
    // Pass on any style that applies. The caller will apply the style if a suitable renderer is
    // not found and try again with the new GrShape.
    if (!args.fShape->style().applies() && SkToBool(fResourceProvider) &&
        (args.fAAType == GrAAType::kCoverage || args.fAAType == GrAAType::kNone)) {
        // This is the fallback renderer for when a path is too complicated for the GPU ones.
        return CanDrawPath::kAsBackup;
    }
    return CanDrawPath::kNo;
}

////////////////////////////////////////////////////////////////////////////////
static bool get_unclipped_shape_dev_bounds(const GrShape& shape, const SkMatrix& matrix,
                                           SkIRect* devBounds) {
    SkRect shapeBounds = shape.styledBounds();
    if (shapeBounds.isEmpty()) {
        return false;
    }
    SkRect shapeDevBounds;
    matrix.mapRect(&shapeDevBounds, shapeBounds);
    // Even though these are "unclipped" bounds we still clip to the int32_t range.
    // This is the largest int32_t that is representable exactly as a float. The next 63 larger ints
    // would round down to this value when cast to a float, but who really cares.
    // INT32_MIN is exactly representable.
    static constexpr int32_t kMaxInt = 2147483520;
    if (!shapeDevBounds.intersect(SkRect::MakeLTRB(INT32_MIN, INT32_MIN, kMaxInt, kMaxInt))) {
        return false;
    }
    shapeDevBounds.roundOut(devBounds);
    return true;
}

// Gets the shape bounds, the clip bounds, and the intersection (if any). Returns false if there
// is no intersection.
static bool get_shape_and_clip_bounds(int width, int height,
                                      const GrClip& clip,
                                      const GrShape& shape,
                                      const SkMatrix& matrix,
                                      SkIRect* unclippedDevShapeBounds,
                                      SkIRect* clippedDevShapeBounds,
                                      SkIRect* devClipBounds) {
    // compute bounds as intersection of rt size, clip, and path
    clip.getConservativeBounds(width, height, devClipBounds);

    if (!get_unclipped_shape_dev_bounds(shape, matrix, unclippedDevShapeBounds)) {
        *unclippedDevShapeBounds = SkIRect::EmptyIRect();
        *clippedDevShapeBounds = SkIRect::EmptyIRect();
        return false;
    }
    if (!clippedDevShapeBounds->intersect(*devClipBounds, *unclippedDevShapeBounds)) {
        *clippedDevShapeBounds = SkIRect::EmptyIRect();
        return false;
    }
    return true;
}

////////////////////////////////////////////////////////////////////////////////

void GrSoftwarePathRenderer::DrawNonAARect(GrRenderTargetContext* renderTargetContext,
                                           GrPaint&& paint,
                                           const GrUserStencilSettings& userStencilSettings,
                                           const GrClip& clip,
                                           const SkMatrix& viewMatrix,
                                           const SkRect& rect,
                                           const SkMatrix& localMatrix) {
    renderTargetContext->addDrawOp(clip,
                                   GrRectOpFactory::MakeNonAAFillWithLocalMatrix(
                                           std::move(paint), viewMatrix, localMatrix, rect,
                                           GrAAType::kNone, &userStencilSettings));
}

void GrSoftwarePathRenderer::DrawAroundInvPath(GrRenderTargetContext* renderTargetContext,
                                               GrPaint&& paint,
                                               const GrUserStencilSettings& userStencilSettings,
                                               const GrClip& clip,
                                               const SkMatrix& viewMatrix,
                                               const SkIRect& devClipBounds,
                                               const SkIRect& devPathBounds) {
    SkMatrix invert;
    if (!viewMatrix.invert(&invert)) {
        return;
    }

    SkRect rect;
    if (devClipBounds.fTop < devPathBounds.fTop) {
        rect.iset(devClipBounds.fLeft, devClipBounds.fTop,
                  devClipBounds.fRight, devPathBounds.fTop);
        DrawNonAARect(renderTargetContext, GrPaint::Clone(paint), userStencilSettings, clip,
                      SkMatrix::I(), rect, invert);
    }
    if (devClipBounds.fLeft < devPathBounds.fLeft) {
        rect.iset(devClipBounds.fLeft, devPathBounds.fTop,
                  devPathBounds.fLeft, devPathBounds.fBottom);
        DrawNonAARect(renderTargetContext, GrPaint::Clone(paint), userStencilSettings, clip,
                      SkMatrix::I(), rect, invert);
    }
    if (devClipBounds.fRight > devPathBounds.fRight) {
        rect.iset(devPathBounds.fRight, devPathBounds.fTop,
                  devClipBounds.fRight, devPathBounds.fBottom);
        DrawNonAARect(renderTargetContext, GrPaint::Clone(paint), userStencilSettings, clip,
                      SkMatrix::I(), rect, invert);
    }
    if (devClipBounds.fBottom > devPathBounds.fBottom) {
        rect.iset(devClipBounds.fLeft, devPathBounds.fBottom,
                  devClipBounds.fRight, devClipBounds.fBottom);
        DrawNonAARect(renderTargetContext, std::move(paint), userStencilSettings, clip,
                      SkMatrix::I(), rect, invert);
    }
}

void GrSoftwarePathRenderer::DrawToTargetWithShapeMask(
        sk_sp<GrTextureProxy> proxy,
        GrRenderTargetContext* renderTargetContext,
        GrPaint&& paint,
        const GrUserStencilSettings& userStencilSettings,
        const GrClip& clip,
        const SkMatrix& viewMatrix,
        const SkIPoint& textureOriginInDeviceSpace,
        const SkIRect& deviceSpaceRectToDraw) {
    SkMatrix invert;
    if (!viewMatrix.invert(&invert)) {
        return;
    }

    SkRect dstRect = SkRect::Make(deviceSpaceRectToDraw);

    // We use device coords to compute the texture coordinates. We take the device coords and apply
    // a translation so that the top-left of the device bounds maps to 0,0, and then a scaling
    // matrix to normalized coords.
    SkMatrix maskMatrix = SkMatrix::MakeTrans(SkIntToScalar(-textureOriginInDeviceSpace.fX),
                                              SkIntToScalar(-textureOriginInDeviceSpace.fY));
    maskMatrix.preConcat(viewMatrix);
    paint.addCoverageFragmentProcessor(GrSimpleTextureEffect::Make(
            std::move(proxy), nullptr, maskMatrix, GrSamplerState::Filter::kNearest));
    DrawNonAARect(renderTargetContext, std::move(paint), userStencilSettings, clip, SkMatrix::I(),
                  dstRect, invert);
}

static sk_sp<GrTextureProxy> make_deferred_mask_texture_proxy(GrContext* context, SkBackingFit fit,
                                                              int width, int height) {
    GrSurfaceDesc desc;
    desc.fOrigin = kTopLeft_GrSurfaceOrigin;
    desc.fWidth = width;
    desc.fHeight = height;
    desc.fConfig = kAlpha_8_GrPixelConfig;

    sk_sp<GrSurfaceContext> sContext =
            context->contextPriv().makeDeferredSurfaceContext(desc, fit, SkBudgeted::kYes);
    if (!sContext || !sContext->asTextureProxy()) {
        return nullptr;
    }
    return sContext->asTextureProxyRef();
}

namespace {

class GrMaskUploaderPrepareCallback : public GrPrepareCallback {
public:
    GrMaskUploaderPrepareCallback(sk_sp<GrTextureProxy> proxy, const SkIRect& maskBounds,
                                  const SkMatrix& viewMatrix, const GrShape& shape, GrAA aa)
            : fProxy(std::move(proxy))
            , fMaskBounds(maskBounds)
            , fViewMatrix(viewMatrix)
            , fShape(shape)
            , fAA(aa)
            , fWaited(false) {}

    ~GrMaskUploaderPrepareCallback() override {
        if (!fWaited) {
            // This can happen if our owning op list fails to instantiate (so it never prepares)
            fPixelsReady.wait();
        }
    }

    void operator()(GrOpFlushState* flushState) override {
        TRACE_EVENT0("skia", "Mask Uploader Pre Flush Callback");
        auto uploadMask = [this](GrDrawOp::WritePixelsFn& writePixelsFn) {
            TRACE_EVENT0("skia", "Mask Upload");
            this->fPixelsReady.wait();
            this->fWaited = true;
            // If the worker thread was unable to allocate pixels, this check will fail, and we'll
            // end up drawing with an uninitialized mask texture, but at least we won't crash.
            if (this->fPixels.addr()) {
                writePixelsFn(this->fProxy.get(), 0, 0,
                              this->fPixels.width(), this->fPixels.height(),
                              kAlpha_8_GrPixelConfig,
                              this->fPixels.addr(), this->fPixels.rowBytes());
                // Free this memory immediately, so it can be recycled. This avoids memory pressure
                // when there is a large amount of threaded work still running during flush.
                this->fPixels.reset();
            }
        };
        flushState->addASAPUpload(std::move(uploadMask));
    }

    SkAutoPixmapStorage* getPixels() { return &fPixels; }
    SkSemaphore* getSemaphore() { return &fPixelsReady; }
    const SkIRect& getMaskBounds() const { return fMaskBounds; }
    const SkMatrix* getViewMatrix() const { return &fViewMatrix; }
    const GrShape& getShape() const { return fShape; }
    GrAA getAA() const { return fAA; }

private:
    // NOTE: This ref cnt isn't thread safe!
    sk_sp<GrTextureProxy> fProxy;
    SkAutoPixmapStorage fPixels;
    SkSemaphore fPixelsReady;

    SkIRect fMaskBounds;
    SkMatrix fViewMatrix;
    GrShape fShape;
    GrAA fAA;
    bool fWaited;
};

}

////////////////////////////////////////////////////////////////////////////////
// return true on success; false on failure
bool GrSoftwarePathRenderer::onDrawPath(const DrawPathArgs& args) {
    GR_AUDIT_TRAIL_AUTO_FRAME(args.fRenderTargetContext->auditTrail(),
                              "GrSoftwarePathRenderer::onDrawPath");
    if (!fResourceProvider) {
        return false;
    }

    // We really need to know if the shape will be inverse filled or not
    bool inverseFilled = false;
    SkTLazy<GrShape> tmpShape;
    SkASSERT(!args.fShape->style().applies());
    // If the path is hairline, ignore inverse fill.
    inverseFilled = args.fShape->inverseFilled() &&
                    !IsStrokeHairlineOrEquivalent(args.fShape->style(), *args.fViewMatrix, nullptr);

    SkIRect unclippedDevShapeBounds, clippedDevShapeBounds, devClipBounds;
    // To prevent overloading the cache with entries during animations we limit the cache of masks
    // to cases where the matrix preserves axis alignment.
    bool useCache = fAllowCaching && !inverseFilled && args.fViewMatrix->preservesAxisAlignment() &&
                    args.fShape->hasUnstyledKey() && GrAAType::kCoverage == args.fAAType;

    if (!get_shape_and_clip_bounds(args.fRenderTargetContext->width(),
                                   args.fRenderTargetContext->height(),
                                   *args.fClip, *args.fShape,
                                   *args.fViewMatrix, &unclippedDevShapeBounds,
                                   &clippedDevShapeBounds,
                                   &devClipBounds)) {
        if (inverseFilled) {
            DrawAroundInvPath(args.fRenderTargetContext, std::move(args.fPaint),
                              *args.fUserStencilSettings, *args.fClip, *args.fViewMatrix,
                              devClipBounds, unclippedDevShapeBounds);
        }
        return true;
    }

    const SkIRect* boundsForMask = &clippedDevShapeBounds;
    if (useCache) {
        // Use the cache only if >50% of the path is visible.
        int unclippedWidth = unclippedDevShapeBounds.width();
        int unclippedHeight = unclippedDevShapeBounds.height();
        int unclippedArea = unclippedWidth * unclippedHeight;
        int clippedArea = clippedDevShapeBounds.width() * clippedDevShapeBounds.height();
        int maxTextureSize = args.fRenderTargetContext->caps()->maxTextureSize();
        if (unclippedArea > 2 * clippedArea || unclippedWidth > maxTextureSize ||
            unclippedHeight > maxTextureSize) {
            useCache = false;
        } else {
            boundsForMask = &unclippedDevShapeBounds;
        }
    }

    GrUniqueKey maskKey;
    if (useCache) {
        // We require the upper left 2x2 of the matrix to match exactly for a cache hit.
        SkScalar sx = args.fViewMatrix->get(SkMatrix::kMScaleX);
        SkScalar sy = args.fViewMatrix->get(SkMatrix::kMScaleY);
        SkScalar kx = args.fViewMatrix->get(SkMatrix::kMSkewX);
        SkScalar ky = args.fViewMatrix->get(SkMatrix::kMSkewY);
        static const GrUniqueKey::Domain kDomain = GrUniqueKey::GenerateDomain();
#ifdef SK_BUILD_FOR_ANDROID_FRAMEWORK
        // Fractional translate does not affect caching on Android. This is done for better cache
        // hit ratio and speed, but it is matching HWUI behavior, which doesn't consider the matrix
        // at all when caching paths.
        GrUniqueKey::Builder builder(&maskKey, kDomain, 4 + args.fShape->unstyledKeySize());
#else
        SkScalar tx = args.fViewMatrix->get(SkMatrix::kMTransX);
        SkScalar ty = args.fViewMatrix->get(SkMatrix::kMTransY);
        // Allow 8 bits each in x and y of subpixel positioning.
        SkFixed fracX = SkScalarToFixed(SkScalarFraction(tx)) & 0x0000FF00;
        SkFixed fracY = SkScalarToFixed(SkScalarFraction(ty)) & 0x0000FF00;
        GrUniqueKey::Builder builder(&maskKey, kDomain, 5 + args.fShape->unstyledKeySize());
#endif
        builder[0] = SkFloat2Bits(sx);
        builder[1] = SkFloat2Bits(sy);
        builder[2] = SkFloat2Bits(kx);
        builder[3] = SkFloat2Bits(ky);
#ifdef SK_BUILD_FOR_ANDROID_FRAMEWORK
        args.fShape->writeUnstyledKey(&builder[4]);
#else
        builder[4] = fracX | (fracY >> 8);
        args.fShape->writeUnstyledKey(&builder[5]);
#endif
    }

    sk_sp<GrTextureProxy> proxy;
    if (useCache) {
        proxy = fResourceProvider->findProxyByUniqueKey(maskKey, kTopLeft_GrSurfaceOrigin);
    }
    if (!proxy) {
        SkBackingFit fit = useCache ? SkBackingFit::kExact : SkBackingFit::kApprox;
        GrAA aa = GrAAType::kCoverage == args.fAAType ? GrAA::kYes : GrAA::kNo;

        SkTaskGroup* taskGroup = args.fContext->contextPriv().getTaskGroup();
        if (taskGroup) {
            proxy = make_deferred_mask_texture_proxy(args.fContext, fit,
                                                     boundsForMask->width(),
                                                     boundsForMask->height());
            if (!proxy) {
                return false;
            }

            auto uploader = skstd::make_unique<GrMaskUploaderPrepareCallback>(
                    proxy, *boundsForMask, *args.fViewMatrix, *args.fShape, aa);
            GrMaskUploaderPrepareCallback* uploaderRaw = uploader.get();

            auto drawAndUploadMask = [uploaderRaw] {
                TRACE_EVENT0("skia", "Threaded SW Mask Render");
                GrSWMaskHelper helper(uploaderRaw->getPixels());
                if (helper.init(uploaderRaw->getMaskBounds())) {
                    helper.drawShape(uploaderRaw->getShape(), *uploaderRaw->getViewMatrix(),
                                     SkRegion::kReplace_Op, uploaderRaw->getAA(), 0xFF);
                } else {
                    SkDEBUGFAIL("Unable to allocate SW mask.");
                }
                uploaderRaw->getSemaphore()->signal();
            };
            taskGroup->add(std::move(drawAndUploadMask));
            args.fRenderTargetContext->getOpList()->addPrepareCallback(std::move(uploader));
        } else {
            GrSWMaskHelper helper;
            if (!helper.init(*boundsForMask)) {
                return false;
            }
            helper.drawShape(*args.fShape, *args.fViewMatrix, SkRegion::kReplace_Op, aa, 0xFF);
            proxy = helper.toTextureProxy(args.fContext, fit);
        }

        if (!proxy) {
            return false;
        }
        if (useCache) {
            SkASSERT(proxy->origin() == kTopLeft_GrSurfaceOrigin);
            fResourceProvider->assignUniqueKeyToProxy(maskKey, proxy.get());
        }
    }
    if (inverseFilled) {
        DrawAroundInvPath(args.fRenderTargetContext, GrPaint::Clone(args.fPaint),
                          *args.fUserStencilSettings, *args.fClip, *args.fViewMatrix, devClipBounds,
                          unclippedDevShapeBounds);
    }
    DrawToTargetWithShapeMask(
            std::move(proxy), args.fRenderTargetContext, std::move(args.fPaint),
            *args.fUserStencilSettings, *args.fClip, *args.fViewMatrix,
            SkIPoint{boundsForMask->fLeft, boundsForMask->fTop}, *boundsForMask);

    return true;
}