| /* |
| * Copyright 2010 Google Inc. |
| * |
| * Use of this source code is governed by a BSD-style license that can be |
| * found in the LICENSE file. |
| */ |
| |
| #include "SkGr.h" |
| #include "SkColorFilter.h" |
| #include "SkConfig8888.h" |
| #include "SkData.h" |
| #include "SkMessageBus.h" |
| #include "SkPixelRef.h" |
| #include "SkTextureCompressor.h" |
| #include "GrResourceCache.h" |
| #include "GrGpu.h" |
| #include "effects/GrDitherEffect.h" |
| #include "GrDrawTargetCaps.h" |
| #include "effects/GrYUVtoRGBEffect.h" |
| |
| #ifndef SK_IGNORE_ETC1_SUPPORT |
| # include "ktx.h" |
| # include "etc1.h" |
| #endif |
| |
| /* Fill out buffer with the compressed format Ganesh expects from a colortable |
| based bitmap. [palette (colortable) + indices]. |
| |
| At the moment Ganesh only supports 8bit version. If Ganesh allowed we others |
| we could detect that the colortable.count is <= 16, and then repack the |
| indices as nibbles to save RAM, but it would take more time (i.e. a lot |
| slower than memcpy), so skipping that for now. |
| |
| Ganesh wants a full 256 palette entry, even though Skia's ctable is only as big |
| as the colortable.count says it is. |
| */ |
| static void build_index8_data(void* buffer, const SkBitmap& bitmap) { |
| SkASSERT(kIndex_8_SkColorType == bitmap.colorType()); |
| |
| SkAutoLockPixels alp(bitmap); |
| if (!bitmap.readyToDraw()) { |
| SkDEBUGFAIL("bitmap not ready to draw!"); |
| return; |
| } |
| |
| SkColorTable* ctable = bitmap.getColorTable(); |
| char* dst = (char*)buffer; |
| |
| const int count = ctable->count(); |
| |
| SkDstPixelInfo dstPI; |
| dstPI.fColorType = kRGBA_8888_SkColorType; |
| dstPI.fAlphaType = kPremul_SkAlphaType; |
| dstPI.fPixels = buffer; |
| dstPI.fRowBytes = count * sizeof(SkPMColor); |
| |
| SkSrcPixelInfo srcPI; |
| srcPI.fColorType = kN32_SkColorType; |
| srcPI.fAlphaType = kPremul_SkAlphaType; |
| srcPI.fPixels = ctable->lockColors(); |
| srcPI.fRowBytes = count * sizeof(SkPMColor); |
| |
| srcPI.convertPixelsTo(&dstPI, count, 1); |
| |
| ctable->unlockColors(); |
| |
| // always skip a full 256 number of entries, even if we memcpy'd fewer |
| dst += 256 * sizeof(GrColor); |
| |
| if ((unsigned)bitmap.width() == bitmap.rowBytes()) { |
| memcpy(dst, bitmap.getPixels(), bitmap.getSize()); |
| } else { |
| // need to trim off the extra bytes per row |
| size_t width = bitmap.width(); |
| size_t rowBytes = bitmap.rowBytes(); |
| const char* src = (const char*)bitmap.getPixels(); |
| for (int y = 0; y < bitmap.height(); y++) { |
| memcpy(dst, src, width); |
| src += rowBytes; |
| dst += width; |
| } |
| } |
| } |
| |
| //////////////////////////////////////////////////////////////////////////////// |
| |
| static void generate_bitmap_cache_id(const SkBitmap& bitmap, GrCacheID* id) { |
| // Our id includes the offset, width, and height so that bitmaps created by extractSubset() |
| // are unique. |
| uint32_t genID = bitmap.getGenerationID(); |
| SkIPoint origin = bitmap.pixelRefOrigin(); |
| int16_t width = SkToS16(bitmap.width()); |
| int16_t height = SkToS16(bitmap.height()); |
| |
| GrCacheID::Key key; |
| memcpy(key.fData8 + 0, &genID, 4); |
| memcpy(key.fData8 + 4, &origin.fX, 4); |
| memcpy(key.fData8 + 8, &origin.fY, 4); |
| memcpy(key.fData8 + 12, &width, 2); |
| memcpy(key.fData8 + 14, &height, 2); |
| static const size_t kKeyDataSize = 16; |
| memset(key.fData8 + kKeyDataSize, 0, sizeof(key) - kKeyDataSize); |
| GR_STATIC_ASSERT(sizeof(key) >= kKeyDataSize); |
| static const GrCacheID::Domain gBitmapTextureDomain = GrCacheID::GenerateDomain(); |
| id->reset(gBitmapTextureDomain, key); |
| } |
| |
| static void generate_bitmap_texture_desc(const SkBitmap& bitmap, GrSurfaceDesc* desc) { |
| desc->fFlags = kNone_GrSurfaceFlags; |
| desc->fWidth = bitmap.width(); |
| desc->fHeight = bitmap.height(); |
| desc->fConfig = SkImageInfo2GrPixelConfig(bitmap.info()); |
| desc->fSampleCnt = 0; |
| } |
| |
| namespace { |
| |
| // When the SkPixelRef genID changes, invalidate a corresponding GrResource described by key. |
| class GrResourceInvalidator : public SkPixelRef::GenIDChangeListener { |
| public: |
| explicit GrResourceInvalidator(GrResourceKey key) : fKey(key) {} |
| private: |
| GrResourceKey fKey; |
| |
| virtual void onChange() SK_OVERRIDE { |
| const GrResourceInvalidatedMessage message = { fKey }; |
| SkMessageBus<GrResourceInvalidatedMessage>::Post(message); |
| } |
| }; |
| |
| } // namespace |
| |
| static void add_genID_listener(GrResourceKey key, SkPixelRef* pixelRef) { |
| SkASSERT(pixelRef); |
| pixelRef->addGenIDChangeListener(SkNEW_ARGS(GrResourceInvalidator, (key))); |
| } |
| |
| static GrTexture* sk_gr_allocate_texture(GrContext* ctx, |
| bool cache, |
| const GrTextureParams* params, |
| const SkBitmap& bm, |
| GrSurfaceDesc desc, |
| const void* pixels, |
| size_t rowBytes) { |
| GrTexture* result; |
| if (cache) { |
| // This texture is likely to be used again so leave it in the cache |
| GrCacheID cacheID; |
| generate_bitmap_cache_id(bm, &cacheID); |
| |
| GrResourceKey key; |
| result = ctx->createTexture(params, desc, cacheID, pixels, rowBytes, &key); |
| if (result) { |
| add_genID_listener(key, bm.pixelRef()); |
| } |
| } else { |
| // This texture is unlikely to be used again (in its present form) so |
| // just use a scratch texture. This will remove the texture from the |
| // cache so no one else can find it. Additionally, once unlocked, the |
| // scratch texture will go to the end of the list for purging so will |
| // likely be available for this volatile bitmap the next time around. |
| result = ctx->refScratchTexture(desc, GrContext::kExact_ScratchTexMatch); |
| if (pixels) { |
| result->writePixels(0, 0, bm.width(), bm.height(), desc.fConfig, pixels, rowBytes); |
| } |
| } |
| return result; |
| } |
| |
| #ifndef SK_IGNORE_ETC1_SUPPORT |
| static GrTexture *load_etc1_texture(GrContext* ctx, bool cache, |
| const GrTextureParams* params, |
| const SkBitmap &bm, GrSurfaceDesc desc) { |
| SkAutoTUnref<SkData> data(bm.pixelRef()->refEncodedData()); |
| |
| // Is this even encoded data? |
| if (NULL == data) { |
| return NULL; |
| } |
| |
| // Is this a valid PKM encoded data? |
| const uint8_t *bytes = data->bytes(); |
| if (etc1_pkm_is_valid(bytes)) { |
| uint32_t encodedWidth = etc1_pkm_get_width(bytes); |
| uint32_t encodedHeight = etc1_pkm_get_height(bytes); |
| |
| // Does the data match the dimensions of the bitmap? If not, |
| // then we don't know how to scale the image to match it... |
| if (encodedWidth != static_cast<uint32_t>(bm.width()) || |
| encodedHeight != static_cast<uint32_t>(bm.height())) { |
| return NULL; |
| } |
| |
| // Everything seems good... skip ahead to the data. |
| bytes += ETC_PKM_HEADER_SIZE; |
| desc.fConfig = kETC1_GrPixelConfig; |
| } else if (SkKTXFile::is_ktx(bytes)) { |
| SkKTXFile ktx(data); |
| |
| // Is it actually an ETC1 texture? |
| if (!ktx.isCompressedFormat(SkTextureCompressor::kETC1_Format)) { |
| return NULL; |
| } |
| |
| // Does the data match the dimensions of the bitmap? If not, |
| // then we don't know how to scale the image to match it... |
| if (ktx.width() != bm.width() || ktx.height() != bm.height()) { |
| return NULL; |
| } |
| |
| bytes = ktx.pixelData(); |
| desc.fConfig = kETC1_GrPixelConfig; |
| } else { |
| return NULL; |
| } |
| |
| return sk_gr_allocate_texture(ctx, cache, params, bm, desc, bytes, 0); |
| } |
| #endif // SK_IGNORE_ETC1_SUPPORT |
| |
| static GrTexture *load_yuv_texture(GrContext* ctx, bool cache, const GrTextureParams* params, |
| const SkBitmap& bm, const GrSurfaceDesc& desc) { |
| // Subsets are not supported, the whole pixelRef is loaded when using YUV decoding |
| if ((bm.pixelRef()->info().width() != bm.info().width()) || |
| (bm.pixelRef()->info().height() != bm.info().height())) { |
| return NULL; |
| } |
| |
| SkPixelRef* pixelRef = bm.pixelRef(); |
| SkISize yuvSizes[3]; |
| if ((NULL == pixelRef) || !pixelRef->getYUV8Planes(yuvSizes, NULL, NULL, NULL)) { |
| return NULL; |
| } |
| |
| // Allocate the memory for YUV |
| size_t totalSize(0); |
| size_t sizes[3], rowBytes[3]; |
| for (int i = 0; i < 3; ++i) { |
| rowBytes[i] = yuvSizes[i].fWidth; |
| totalSize += sizes[i] = rowBytes[i] * yuvSizes[i].fHeight; |
| } |
| SkAutoMalloc storage(totalSize); |
| void* planes[3]; |
| planes[0] = storage.get(); |
| planes[1] = (uint8_t*)planes[0] + sizes[0]; |
| planes[2] = (uint8_t*)planes[1] + sizes[1]; |
| |
| SkYUVColorSpace colorSpace; |
| |
| // Get the YUV planes |
| if (!pixelRef->getYUV8Planes(yuvSizes, planes, rowBytes, &colorSpace)) { |
| return NULL; |
| } |
| |
| GrSurfaceDesc yuvDesc; |
| yuvDesc.fConfig = kAlpha_8_GrPixelConfig; |
| SkAutoTUnref<GrTexture> yuvTextures[3]; |
| for (int i = 0; i < 3; ++i) { |
| yuvDesc.fWidth = yuvSizes[i].fWidth; |
| yuvDesc.fHeight = yuvSizes[i].fHeight; |
| yuvTextures[i].reset( |
| ctx->refScratchTexture(yuvDesc, GrContext::kApprox_ScratchTexMatch)); |
| if (!yuvTextures[i] || |
| !yuvTextures[i]->writePixels(0, 0, yuvDesc.fWidth, yuvDesc.fHeight, |
| yuvDesc.fConfig, planes[i], rowBytes[i])) { |
| return NULL; |
| } |
| } |
| |
| GrSurfaceDesc rtDesc = desc; |
| rtDesc.fFlags = rtDesc.fFlags | |
| kRenderTarget_GrSurfaceFlag | |
| kNoStencil_GrSurfaceFlag; |
| |
| GrTexture* result = sk_gr_allocate_texture(ctx, cache, params, bm, rtDesc, NULL, 0); |
| |
| GrRenderTarget* renderTarget = result ? result->asRenderTarget() : NULL; |
| if (renderTarget) { |
| SkAutoTUnref<GrFragmentProcessor> yuvToRgbProcessor( |
| GrYUVtoRGBEffect::Create(yuvTextures[0], yuvTextures[1], yuvTextures[2], colorSpace)); |
| GrPaint paint; |
| paint.addColorProcessor(yuvToRgbProcessor); |
| SkRect r = SkRect::MakeWH(SkIntToScalar(yuvSizes[0].fWidth), |
| SkIntToScalar(yuvSizes[0].fHeight)); |
| GrContext::AutoRenderTarget autoRT(ctx, renderTarget); |
| GrContext::AutoMatrix am; |
| am.setIdentity(ctx); |
| GrContext::AutoClip ac(ctx, GrContext::AutoClip::kWideOpen_InitialClip); |
| ctx->drawRect(paint, r); |
| } else { |
| SkSafeSetNull(result); |
| } |
| |
| return result; |
| } |
| |
| static GrTexture* sk_gr_create_bitmap_texture(GrContext* ctx, |
| bool cache, |
| const GrTextureParams* params, |
| const SkBitmap& origBitmap) { |
| SkBitmap tmpBitmap; |
| |
| const SkBitmap* bitmap = &origBitmap; |
| |
| GrSurfaceDesc desc; |
| generate_bitmap_texture_desc(*bitmap, &desc); |
| |
| if (kIndex_8_SkColorType == bitmap->colorType()) { |
| // build_compressed_data doesn't do npot->pot expansion |
| // and paletted textures can't be sub-updated |
| if (cache && ctx->supportsIndex8PixelConfig(params, bitmap->width(), bitmap->height())) { |
| size_t imageSize = GrCompressedFormatDataSize(kIndex_8_GrPixelConfig, |
| bitmap->width(), bitmap->height()); |
| SkAutoMalloc storage(imageSize); |
| build_index8_data(storage.get(), origBitmap); |
| |
| // our compressed data will be trimmed, so pass width() for its |
| // "rowBytes", since they are the same now. |
| return sk_gr_allocate_texture(ctx, cache, params, origBitmap, |
| desc, storage.get(), bitmap->width()); |
| } else { |
| origBitmap.copyTo(&tmpBitmap, kN32_SkColorType); |
| // now bitmap points to our temp, which has been promoted to 32bits |
| bitmap = &tmpBitmap; |
| desc.fConfig = SkImageInfo2GrPixelConfig(bitmap->info()); |
| } |
| } |
| |
| // Is this an ETC1 encoded texture? |
| #ifndef SK_IGNORE_ETC1_SUPPORT |
| else if ( |
| // We do not support scratch ETC1 textures, hence they should all be at least |
| // trying to go to the cache. |
| cache |
| // Make sure that the underlying device supports ETC1 textures before we go ahead |
| // and check the data. |
| && ctx->getGpu()->caps()->isConfigTexturable(kETC1_GrPixelConfig) |
| // If the bitmap had compressed data and was then uncompressed, it'll still return |
| // compressed data on 'refEncodedData' and upload it. Probably not good, since if |
| // the bitmap has available pixels, then they might not be what the decompressed |
| // data is. |
| && !(bitmap->readyToDraw())) { |
| GrTexture *texture = load_etc1_texture(ctx, cache, params, *bitmap, desc); |
| if (texture) { |
| return texture; |
| } |
| } |
| #endif // SK_IGNORE_ETC1_SUPPORT |
| |
| else { |
| GrTexture *texture = load_yuv_texture(ctx, cache, params, *bitmap, desc); |
| if (texture) { |
| return texture; |
| } |
| } |
| SkAutoLockPixels alp(*bitmap); |
| if (!bitmap->readyToDraw()) { |
| return NULL; |
| } |
| |
| return sk_gr_allocate_texture(ctx, cache, params, origBitmap, desc, |
| bitmap->getPixels(), bitmap->rowBytes()); |
| } |
| |
| bool GrIsBitmapInCache(const GrContext* ctx, |
| const SkBitmap& bitmap, |
| const GrTextureParams* params) { |
| GrCacheID cacheID; |
| generate_bitmap_cache_id(bitmap, &cacheID); |
| |
| GrSurfaceDesc desc; |
| generate_bitmap_texture_desc(bitmap, &desc); |
| return ctx->isTextureInCache(desc, cacheID, params); |
| } |
| |
| GrTexture* GrRefCachedBitmapTexture(GrContext* ctx, |
| const SkBitmap& bitmap, |
| const GrTextureParams* params) { |
| GrTexture* result = NULL; |
| |
| bool cache = !bitmap.isVolatile(); |
| |
| if (cache) { |
| // If the bitmap isn't changing try to find a cached copy first. |
| |
| GrCacheID cacheID; |
| generate_bitmap_cache_id(bitmap, &cacheID); |
| |
| GrSurfaceDesc desc; |
| generate_bitmap_texture_desc(bitmap, &desc); |
| |
| result = ctx->findAndRefTexture(desc, cacheID, params); |
| } |
| if (NULL == result) { |
| result = sk_gr_create_bitmap_texture(ctx, cache, params, bitmap); |
| } |
| if (NULL == result) { |
| SkDebugf("---- failed to create texture for cache [%d %d]\n", |
| bitmap.width(), bitmap.height()); |
| } |
| return result; |
| } |
| |
| /////////////////////////////////////////////////////////////////////////////// |
| |
| // alphatype is ignore for now, but if GrPixelConfig is expanded to encompass |
| // alpha info, that will be considered. |
| GrPixelConfig SkImageInfo2GrPixelConfig(SkColorType ct, SkAlphaType) { |
| switch (ct) { |
| case kUnknown_SkColorType: |
| return kUnknown_GrPixelConfig; |
| case kAlpha_8_SkColorType: |
| return kAlpha_8_GrPixelConfig; |
| case kRGB_565_SkColorType: |
| return kRGB_565_GrPixelConfig; |
| case kARGB_4444_SkColorType: |
| return kRGBA_4444_GrPixelConfig; |
| case kRGBA_8888_SkColorType: |
| return kRGBA_8888_GrPixelConfig; |
| case kBGRA_8888_SkColorType: |
| return kBGRA_8888_GrPixelConfig; |
| case kIndex_8_SkColorType: |
| return kIndex_8_GrPixelConfig; |
| } |
| SkASSERT(0); // shouldn't get here |
| return kUnknown_GrPixelConfig; |
| } |
| |
| bool GrPixelConfig2ColorType(GrPixelConfig config, SkColorType* ctOut) { |
| SkColorType ct; |
| switch (config) { |
| case kAlpha_8_GrPixelConfig: |
| ct = kAlpha_8_SkColorType; |
| break; |
| case kIndex_8_GrPixelConfig: |
| ct = kIndex_8_SkColorType; |
| break; |
| case kRGB_565_GrPixelConfig: |
| ct = kRGB_565_SkColorType; |
| break; |
| case kRGBA_4444_GrPixelConfig: |
| ct = kARGB_4444_SkColorType; |
| break; |
| case kRGBA_8888_GrPixelConfig: |
| ct = kRGBA_8888_SkColorType; |
| break; |
| case kBGRA_8888_GrPixelConfig: |
| ct = kBGRA_8888_SkColorType; |
| break; |
| default: |
| return false; |
| } |
| if (ctOut) { |
| *ctOut = ct; |
| } |
| return true; |
| } |
| |
| /////////////////////////////////////////////////////////////////////////////// |
| |
| void SkPaint2GrPaintNoShader(GrContext* context, const SkPaint& skPaint, GrColor paintColor, |
| bool constantColor, GrPaint* grPaint) { |
| |
| grPaint->setDither(skPaint.isDither()); |
| grPaint->setAntiAlias(skPaint.isAntiAlias()); |
| |
| SkXfermode::Coeff sm; |
| SkXfermode::Coeff dm; |
| |
| SkXfermode* mode = skPaint.getXfermode(); |
| GrFragmentProcessor* xferProcessor = NULL; |
| if (SkXfermode::asFragmentProcessorOrCoeff(mode, &xferProcessor, &sm, &dm)) { |
| if (xferProcessor) { |
| grPaint->addColorProcessor(xferProcessor)->unref(); |
| sm = SkXfermode::kOne_Coeff; |
| dm = SkXfermode::kZero_Coeff; |
| } |
| } else { |
| //SkDEBUGCODE(SkDebugf("Unsupported xfer mode.\n");) |
| // Fall back to src-over |
| sm = SkXfermode::kOne_Coeff; |
| dm = SkXfermode::kISA_Coeff; |
| } |
| grPaint->setBlendFunc(sk_blend_to_grblend(sm), sk_blend_to_grblend(dm)); |
| |
| //set the color of the paint to the one of the parameter |
| grPaint->setColor(paintColor); |
| |
| SkColorFilter* colorFilter = skPaint.getColorFilter(); |
| if (colorFilter) { |
| // if the source color is a constant then apply the filter here once rather than per pixel |
| // in a shader. |
| if (constantColor) { |
| SkColor filtered = colorFilter->filterColor(skPaint.getColor()); |
| grPaint->setColor(SkColor2GrColor(filtered)); |
| } else { |
| SkAutoTUnref<GrFragmentProcessor> fp(colorFilter->asFragmentProcessor(context)); |
| if (fp.get()) { |
| grPaint->addColorProcessor(fp); |
| } |
| } |
| } |
| |
| #ifndef SK_IGNORE_GPU_DITHER |
| // If the dither flag is set, then we need to see if the underlying context |
| // supports it. If not, then install a dither effect. |
| if (skPaint.isDither() && grPaint->numColorStages() > 0) { |
| // What are we rendering into? |
| const GrRenderTarget *target = context->getRenderTarget(); |
| SkASSERT(target); |
| |
| // Suspect the dithering flag has no effect on these configs, otherwise |
| // fall back on setting the appropriate state. |
| if (target->config() == kRGBA_8888_GrPixelConfig || |
| target->config() == kBGRA_8888_GrPixelConfig) { |
| // The dither flag is set and the target is likely |
| // not going to be dithered by the GPU. |
| SkAutoTUnref<GrFragmentProcessor> fp(GrDitherEffect::Create()); |
| if (fp.get()) { |
| grPaint->addColorProcessor(fp); |
| grPaint->setDither(false); |
| } |
| } |
| } |
| #endif |
| } |
| |
| /** |
| * Unlike GrContext::AutoMatrix, this doesn't require setting a new matrix. GrContext::AutoMatrix |
| * likes to set the new matrix in its constructor because it is usually necessary to simulataneously |
| * update a GrPaint. This AutoMatrix is used while initially setting up GrPaint, however. |
| */ |
| class AutoMatrix { |
| public: |
| AutoMatrix(GrContext* context) { |
| fMatrix = context->getMatrix(); |
| fContext = context; |
| } |
| ~AutoMatrix() { |
| SkASSERT(fContext); |
| fContext->setMatrix(fMatrix); |
| } |
| private: |
| GrContext* fContext; |
| SkMatrix fMatrix; |
| }; |
| |
| void SkPaint2GrPaintShader(GrContext* context, const SkPaint& skPaint, |
| bool constantColor, GrPaint* grPaint) { |
| SkShader* shader = skPaint.getShader(); |
| if (NULL == shader) { |
| SkPaint2GrPaintNoShader(context, skPaint, SkColor2GrColor(skPaint.getColor()), |
| constantColor, grPaint); |
| return; |
| } |
| |
| GrColor paintColor = SkColor2GrColor(skPaint.getColor()); |
| |
| // Start a new block here in order to preserve our context state after calling |
| // asFragmentProcessor(). Since these calls get passed back to the client, we don't really |
| // want them messing around with the context. |
| { |
| // SkShader::asFragmentProcessor() may do offscreen rendering. Save off the current RT, |
| // clip, and matrix. We don't reset the matrix on the context because |
| // SkShader::asFragmentProcessor may use GrContext::getMatrix() to know the transformation |
| // from local coords to device space. |
| GrContext::AutoRenderTarget art(context, NULL); |
| GrContext::AutoClip ac(context, GrContext::AutoClip::kWideOpen_InitialClip); |
| AutoMatrix am(context); |
| |
| // Allow the shader to modify paintColor and also create an effect to be installed as |
| // the first color effect on the GrPaint. |
| GrFragmentProcessor* fp = NULL; |
| if (shader->asFragmentProcessor(context, skPaint, NULL, &paintColor, &fp) && fp) { |
| grPaint->addColorProcessor(fp)->unref(); |
| constantColor = false; |
| } |
| } |
| |
| // The grcolor is automatically set when calling asFragmentProcessor. |
| // If the shader can be seen as an effect it returns true and adds its effect to the grpaint. |
| SkPaint2GrPaintNoShader(context, skPaint, paintColor, constantColor, grPaint); |
| } |