blob: 8cc642e35135107db109f14b51f371776b5163bf [file] [log] [blame]
// Copyright 2017 The Dawn Authors
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
#include "dawn_native/metal/TextureMTL.h"
#include "common/Constants.h"
#include "common/Math.h"
#include "common/Platform.h"
#include "dawn_native/DynamicUploader.h"
#include "dawn_native/EnumMaskIterator.h"
#include "dawn_native/metal/DeviceMTL.h"
#include "dawn_native/metal/StagingBufferMTL.h"
#include "dawn_native/metal/UtilsMetal.h"
#include <CoreVideo/CVPixelBuffer.h>
namespace dawn_native { namespace metal {
namespace {
bool UsageNeedsTextureView(wgpu::TextureUsage usage) {
constexpr wgpu::TextureUsage kUsageNeedsTextureView =
wgpu::TextureUsage::Storage | wgpu::TextureUsage::Sampled;
return usage & kUsageNeedsTextureView;
}
MTLTextureUsage MetalTextureUsage(const Format& format,
wgpu::TextureUsage usage,
uint32_t sampleCount) {
MTLTextureUsage result = MTLTextureUsageUnknown; // This is 0
if (usage & (wgpu::TextureUsage::Storage)) {
result |= MTLTextureUsageShaderWrite | MTLTextureUsageShaderRead;
}
if (usage & (wgpu::TextureUsage::Sampled)) {
result |= MTLTextureUsageShaderRead;
// For sampling stencil aspect of combined depth/stencil. See TextureView
// constructor.
if (@available(macOS 10.12, iOS 10.0, *)) {
if (IsSubset(Aspect::Depth | Aspect::Stencil, format.aspects)) {
result |= MTLTextureUsagePixelFormatView;
}
}
}
// MTLTextureUsageRenderTarget is needed to clear multisample textures.
if (usage & (wgpu::TextureUsage::RenderAttachment) || sampleCount > 1) {
result |= MTLTextureUsageRenderTarget;
}
return result;
}
MTLTextureType MetalTextureViewType(wgpu::TextureViewDimension dimension,
unsigned int sampleCount) {
switch (dimension) {
case wgpu::TextureViewDimension::e2D:
return (sampleCount > 1) ? MTLTextureType2DMultisample : MTLTextureType2D;
case wgpu::TextureViewDimension::e2DArray:
return MTLTextureType2DArray;
case wgpu::TextureViewDimension::Cube:
return MTLTextureTypeCube;
case wgpu::TextureViewDimension::CubeArray:
return MTLTextureTypeCubeArray;
case wgpu::TextureViewDimension::e3D:
return MTLTextureType3D;
case wgpu::TextureViewDimension::e1D:
case wgpu::TextureViewDimension::Undefined:
UNREACHABLE();
}
}
bool RequiresCreatingNewTextureView(const TextureBase* texture,
const TextureViewDescriptor* textureViewDescriptor) {
if (texture->GetFormat().format != textureViewDescriptor->format) {
return true;
}
if (texture->GetArrayLayers() != textureViewDescriptor->arrayLayerCount) {
return true;
}
if (texture->GetNumMipLevels() != textureViewDescriptor->mipLevelCount) {
return true;
}
if (IsSubset(Aspect::Depth | Aspect::Stencil, texture->GetFormat().aspects) &&
textureViewDescriptor->aspect == wgpu::TextureAspect::StencilOnly) {
return true;
}
switch (textureViewDescriptor->dimension) {
case wgpu::TextureViewDimension::Cube:
case wgpu::TextureViewDimension::CubeArray:
return true;
default:
break;
}
return false;
}
ResultOrError<wgpu::TextureFormat> GetFormatEquivalentToIOSurfaceFormat(uint32_t format) {
switch (format) {
case kCVPixelFormatType_32RGBA:
return wgpu::TextureFormat::RGBA8Unorm;
case kCVPixelFormatType_32BGRA:
return wgpu::TextureFormat::BGRA8Unorm;
case kCVPixelFormatType_TwoComponent8:
return wgpu::TextureFormat::RG8Unorm;
case kCVPixelFormatType_OneComponent8:
return wgpu::TextureFormat::R8Unorm;
default:
return DAWN_VALIDATION_ERROR("Unsupported IOSurface format");
}
}
#if defined(DAWN_PLATFORM_MACOS)
MTLStorageMode kIOSurfaceStorageMode = MTLStorageModeManaged;
#elif defined(DAWN_PLATFORM_IOS)
MTLStorageMode kIOSurfaceStorageMode = MTLStorageModePrivate;
#else
# error "Unsupported Apple platform."
#endif
}
MTLPixelFormat MetalPixelFormat(wgpu::TextureFormat format) {
switch (format) {
case wgpu::TextureFormat::R8Unorm:
return MTLPixelFormatR8Unorm;
case wgpu::TextureFormat::R8Snorm:
return MTLPixelFormatR8Snorm;
case wgpu::TextureFormat::R8Uint:
return MTLPixelFormatR8Uint;
case wgpu::TextureFormat::R8Sint:
return MTLPixelFormatR8Sint;
case wgpu::TextureFormat::R16Uint:
return MTLPixelFormatR16Uint;
case wgpu::TextureFormat::R16Sint:
return MTLPixelFormatR16Sint;
case wgpu::TextureFormat::R16Float:
return MTLPixelFormatR16Float;
case wgpu::TextureFormat::RG8Unorm:
return MTLPixelFormatRG8Unorm;
case wgpu::TextureFormat::RG8Snorm:
return MTLPixelFormatRG8Snorm;
case wgpu::TextureFormat::RG8Uint:
return MTLPixelFormatRG8Uint;
case wgpu::TextureFormat::RG8Sint:
return MTLPixelFormatRG8Sint;
case wgpu::TextureFormat::R32Uint:
return MTLPixelFormatR32Uint;
case wgpu::TextureFormat::R32Sint:
return MTLPixelFormatR32Sint;
case wgpu::TextureFormat::R32Float:
return MTLPixelFormatR32Float;
case wgpu::TextureFormat::RG16Uint:
return MTLPixelFormatRG16Uint;
case wgpu::TextureFormat::RG16Sint:
return MTLPixelFormatRG16Sint;
case wgpu::TextureFormat::RG16Float:
return MTLPixelFormatRG16Float;
case wgpu::TextureFormat::RGBA8Unorm:
return MTLPixelFormatRGBA8Unorm;
case wgpu::TextureFormat::RGBA8UnormSrgb:
return MTLPixelFormatRGBA8Unorm_sRGB;
case wgpu::TextureFormat::RGBA8Snorm:
return MTLPixelFormatRGBA8Snorm;
case wgpu::TextureFormat::RGBA8Uint:
return MTLPixelFormatRGBA8Uint;
case wgpu::TextureFormat::RGBA8Sint:
return MTLPixelFormatRGBA8Sint;
case wgpu::TextureFormat::BGRA8Unorm:
return MTLPixelFormatBGRA8Unorm;
case wgpu::TextureFormat::BGRA8UnormSrgb:
return MTLPixelFormatBGRA8Unorm_sRGB;
case wgpu::TextureFormat::RGB10A2Unorm:
return MTLPixelFormatRGB10A2Unorm;
case wgpu::TextureFormat::RG11B10Ufloat:
return MTLPixelFormatRG11B10Float;
case wgpu::TextureFormat::RGB9E5Ufloat:
return MTLPixelFormatRGB9E5Float;
case wgpu::TextureFormat::RG32Uint:
return MTLPixelFormatRG32Uint;
case wgpu::TextureFormat::RG32Sint:
return MTLPixelFormatRG32Sint;
case wgpu::TextureFormat::RG32Float:
return MTLPixelFormatRG32Float;
case wgpu::TextureFormat::RGBA16Uint:
return MTLPixelFormatRGBA16Uint;
case wgpu::TextureFormat::RGBA16Sint:
return MTLPixelFormatRGBA16Sint;
case wgpu::TextureFormat::RGBA16Float:
return MTLPixelFormatRGBA16Float;
case wgpu::TextureFormat::RGBA32Uint:
return MTLPixelFormatRGBA32Uint;
case wgpu::TextureFormat::RGBA32Sint:
return MTLPixelFormatRGBA32Sint;
case wgpu::TextureFormat::RGBA32Float:
return MTLPixelFormatRGBA32Float;
case wgpu::TextureFormat::Depth32Float:
return MTLPixelFormatDepth32Float;
case wgpu::TextureFormat::Depth24Plus:
return MTLPixelFormatDepth32Float;
case wgpu::TextureFormat::Depth24PlusStencil8:
return MTLPixelFormatDepth32Float_Stencil8;
#if defined(DAWN_PLATFORM_MACOS)
case wgpu::TextureFormat::BC1RGBAUnorm:
return MTLPixelFormatBC1_RGBA;
case wgpu::TextureFormat::BC1RGBAUnormSrgb:
return MTLPixelFormatBC1_RGBA_sRGB;
case wgpu::TextureFormat::BC2RGBAUnorm:
return MTLPixelFormatBC2_RGBA;
case wgpu::TextureFormat::BC2RGBAUnormSrgb:
return MTLPixelFormatBC2_RGBA_sRGB;
case wgpu::TextureFormat::BC3RGBAUnorm:
return MTLPixelFormatBC3_RGBA;
case wgpu::TextureFormat::BC3RGBAUnormSrgb:
return MTLPixelFormatBC3_RGBA_sRGB;
case wgpu::TextureFormat::BC4RSnorm:
return MTLPixelFormatBC4_RSnorm;
case wgpu::TextureFormat::BC4RUnorm:
return MTLPixelFormatBC4_RUnorm;
case wgpu::TextureFormat::BC5RGSnorm:
return MTLPixelFormatBC5_RGSnorm;
case wgpu::TextureFormat::BC5RGUnorm:
return MTLPixelFormatBC5_RGUnorm;
case wgpu::TextureFormat::BC6HRGBFloat:
return MTLPixelFormatBC6H_RGBFloat;
case wgpu::TextureFormat::BC6HRGBUfloat:
return MTLPixelFormatBC6H_RGBUfloat;
case wgpu::TextureFormat::BC7RGBAUnorm:
return MTLPixelFormatBC7_RGBAUnorm;
case wgpu::TextureFormat::BC7RGBAUnormSrgb:
return MTLPixelFormatBC7_RGBAUnorm_sRGB;
#endif
default:
UNREACHABLE();
}
}
MaybeError ValidateIOSurfaceCanBeWrapped(const DeviceBase*,
const TextureDescriptor* descriptor,
IOSurfaceRef ioSurface,
uint32_t plane) {
// IOSurfaceGetPlaneCount can return 0 for non-planar IOSurfaces but we will treat
// non-planar like it is a single plane.
size_t surfacePlaneCount = std::max(size_t(1), IOSurfaceGetPlaneCount(ioSurface));
if (plane >= surfacePlaneCount) {
return DAWN_VALIDATION_ERROR("IOSurface plane doesn't exist");
}
if (descriptor->dimension != wgpu::TextureDimension::e2D) {
return DAWN_VALIDATION_ERROR("IOSurface texture must be 2D");
}
if (descriptor->mipLevelCount != 1) {
return DAWN_VALIDATION_ERROR("IOSurface mip level count must be 1");
}
if (descriptor->size.depthOrArrayLayers != 1) {
return DAWN_VALIDATION_ERROR("IOSurface array layer count must be 1");
}
if (descriptor->sampleCount != 1) {
return DAWN_VALIDATION_ERROR("IOSurface sample count must be 1");
}
if (descriptor->size.width != IOSurfaceGetWidthOfPlane(ioSurface, plane) ||
descriptor->size.height != IOSurfaceGetHeightOfPlane(ioSurface, plane) ||
descriptor->size.depthOrArrayLayers != 1) {
return DAWN_VALIDATION_ERROR("IOSurface size doesn't match descriptor");
}
wgpu::TextureFormat ioSurfaceFormat;
DAWN_TRY_ASSIGN(ioSurfaceFormat,
GetFormatEquivalentToIOSurfaceFormat(IOSurfaceGetPixelFormat(ioSurface)));
if (descriptor->format != ioSurfaceFormat) {
return DAWN_VALIDATION_ERROR("IOSurface format doesn't match descriptor");
}
return {};
}
NSRef<MTLTextureDescriptor> CreateMetalTextureDescriptor(DeviceBase* device,
const TextureDescriptor* descriptor) {
NSRef<MTLTextureDescriptor> mtlDescRef = AcquireNSRef([MTLTextureDescriptor new]);
MTLTextureDescriptor* mtlDesc = mtlDescRef.Get();
mtlDesc.width = descriptor->size.width;
mtlDesc.height = descriptor->size.height;
mtlDesc.sampleCount = descriptor->sampleCount;
// TODO: add MTLTextureUsagePixelFormatView when needed when we support format
// reinterpretation.
mtlDesc.usage = MetalTextureUsage(device->GetValidInternalFormat(descriptor->format),
descriptor->usage, descriptor->sampleCount);
mtlDesc.pixelFormat = MetalPixelFormat(descriptor->format);
mtlDesc.mipmapLevelCount = descriptor->mipLevelCount;
mtlDesc.storageMode = MTLStorageModePrivate;
// Choose the correct MTLTextureType and paper over differences in how the array layer count
// is specified.
mtlDesc.depth = descriptor->size.depthOrArrayLayers;
mtlDesc.arrayLength = 1;
switch (descriptor->dimension) {
case wgpu::TextureDimension::e2D:
if (mtlDesc.depth > 1) {
ASSERT(mtlDesc.sampleCount == 1);
mtlDesc.textureType = MTLTextureType2DArray;
mtlDesc.arrayLength = mtlDesc.depth;
mtlDesc.depth = 1;
} else {
if (mtlDesc.sampleCount > 1) {
mtlDesc.textureType = MTLTextureType2DMultisample;
} else {
mtlDesc.textureType = MTLTextureType2D;
}
}
break;
case wgpu::TextureDimension::e3D:
ASSERT(mtlDesc.sampleCount == 1);
mtlDesc.textureType = MTLTextureType3D;
break;
case wgpu::TextureDimension::e1D:
UNREACHABLE();
}
return mtlDescRef;
}
// static
ResultOrError<Ref<Texture>> Texture::Create(Device* device,
const TextureDescriptor* descriptor) {
return AcquireRef(new Texture(device, descriptor));
}
Texture::Texture(Device* device, const TextureDescriptor* descriptor)
: TextureBase(device, descriptor, TextureState::OwnedInternal) {
NSRef<MTLTextureDescriptor> mtlDesc = CreateMetalTextureDescriptor(device, descriptor);
mMtlTexture =
AcquireNSPRef([device->GetMTLDevice() newTextureWithDescriptor:mtlDesc.Get()]);
mMtlUsage = [*mtlDesc usage];
if (device->IsToggleEnabled(Toggle::NonzeroClearResourcesOnCreationForTesting)) {
device->ConsumedError(ClearTexture(device->GetPendingCommandContext(),
GetAllSubresources(),
TextureBase::ClearValue::NonZero));
}
}
Texture::Texture(Device* device,
const TextureDescriptor* descriptor,
NSPRef<id<MTLTexture>> mtlTexture)
: TextureBase(device, descriptor, TextureState::OwnedInternal),
mMtlTexture(std::move(mtlTexture)) {
NSRef<MTLTextureDescriptor> mtlDesc = CreateMetalTextureDescriptor(device, descriptor);
mMtlUsage = [*mtlDesc usage];
}
Texture::Texture(Device* device,
const ExternalImageDescriptor* descriptor,
IOSurfaceRef ioSurface,
uint32_t plane)
: TextureBase(device,
reinterpret_cast<const TextureDescriptor*>(descriptor->cTextureDescriptor),
TextureState::OwnedInternal) {
NSRef<MTLTextureDescriptor> mtlDesc = CreateMetalTextureDescriptor(
device, reinterpret_cast<const TextureDescriptor*>(descriptor->cTextureDescriptor));
[*mtlDesc setStorageMode:kIOSurfaceStorageMode];
mMtlTexture = AcquireNSPRef([device->GetMTLDevice() newTextureWithDescriptor:mtlDesc.Get()
iosurface:ioSurface
plane:plane]);
mMtlUsage = [*mtlDesc usage];
SetIsSubresourceContentInitialized(descriptor->isInitialized, GetAllSubresources());
}
Texture::~Texture() {
DestroyInternal();
}
void Texture::DestroyImpl() {
mMtlTexture = nullptr;
}
id<MTLTexture> Texture::GetMTLTexture() {
return mMtlTexture.Get();
}
MaybeError Texture::ClearTexture(CommandRecordingContext* commandContext,
const SubresourceRange& range,
TextureBase::ClearValue clearValue) {
Device* device = ToBackend(GetDevice());
const uint8_t clearColor = (clearValue == TextureBase::ClearValue::Zero) ? 0 : 1;
const double dClearColor = (clearValue == TextureBase::ClearValue::Zero) ? 0.0 : 1.0;
if ((mMtlUsage & MTLTextureUsageRenderTarget) != 0) {
ASSERT(GetFormat().isRenderable);
// End the blit encoder if it is open.
commandContext->EndBlit();
if (GetFormat().HasDepthOrStencil()) {
// Create a render pass to clear each subresource.
for (uint32_t level = range.baseMipLevel;
level < range.baseMipLevel + range.levelCount; ++level) {
for (uint32_t arrayLayer = range.baseArrayLayer;
arrayLayer < range.baseArrayLayer + range.layerCount; arrayLayer++) {
if (clearValue == TextureBase::ClearValue::Zero &&
IsSubresourceContentInitialized(SubresourceRange::SingleMipAndLayer(
level, arrayLayer, range.aspects))) {
// Skip lazy clears if already initialized.
continue;
}
// Note that this creates a descriptor that's autoreleased so we don't use
// AcquireNSRef
NSRef<MTLRenderPassDescriptor> descriptorRef =
[MTLRenderPassDescriptor renderPassDescriptor];
MTLRenderPassDescriptor* descriptor = descriptorRef.Get();
// At least one aspect needs clearing. Iterate the aspects individually to
// determine which to clear.
for (Aspect aspect : IterateEnumMask(range.aspects)) {
if (clearValue == TextureBase::ClearValue::Zero &&
IsSubresourceContentInitialized(SubresourceRange::SingleMipAndLayer(
level, arrayLayer, aspect))) {
// Skip lazy clears if already initialized.
continue;
}
ASSERT(GetDimension() == wgpu::TextureDimension::e2D);
switch (aspect) {
case Aspect::Depth:
descriptor.depthAttachment.texture = GetMTLTexture();
descriptor.depthAttachment.level = level;
descriptor.depthAttachment.slice = arrayLayer;
descriptor.depthAttachment.loadAction = MTLLoadActionClear;
descriptor.depthAttachment.storeAction = MTLStoreActionStore;
descriptor.depthAttachment.clearDepth = dClearColor;
break;
case Aspect::Stencil:
descriptor.stencilAttachment.texture = GetMTLTexture();
descriptor.stencilAttachment.level = level;
descriptor.stencilAttachment.slice = arrayLayer;
descriptor.stencilAttachment.loadAction = MTLLoadActionClear;
descriptor.stencilAttachment.storeAction = MTLStoreActionStore;
descriptor.stencilAttachment.clearStencil =
static_cast<uint32_t>(clearColor);
break;
default:
UNREACHABLE();
}
}
commandContext->BeginRender(descriptor);
commandContext->EndRender();
}
}
} else {
ASSERT(GetFormat().IsColor());
for (uint32_t level = range.baseMipLevel;
level < range.baseMipLevel + range.levelCount; ++level) {
// Create multiple render passes with each subresource as a color attachment to
// clear them all. Only do this for array layers to ensure all attachments have
// the same size.
NSRef<MTLRenderPassDescriptor> descriptor;
uint32_t attachment = 0;
uint32_t numZSlices = GetMipLevelVirtualSize(level).depthOrArrayLayers;
for (uint32_t arrayLayer = range.baseArrayLayer;
arrayLayer < range.baseArrayLayer + range.layerCount; arrayLayer++) {
if (clearValue == TextureBase::ClearValue::Zero &&
IsSubresourceContentInitialized(SubresourceRange::SingleMipAndLayer(
level, arrayLayer, Aspect::Color))) {
// Skip lazy clears if already initialized.
continue;
}
for (uint32_t z = 0; z < numZSlices; ++z) {
if (descriptor == nullptr) {
// Note that this creates a descriptor that's autoreleased so we
// don't use AcquireNSRef
descriptor = [MTLRenderPassDescriptor renderPassDescriptor];
}
[*descriptor colorAttachments][attachment].texture = GetMTLTexture();
[*descriptor colorAttachments][attachment].loadAction =
MTLLoadActionClear;
[*descriptor colorAttachments][attachment].storeAction =
MTLStoreActionStore;
[*descriptor colorAttachments][attachment].clearColor =
MTLClearColorMake(dClearColor, dClearColor, dClearColor,
dClearColor);
[*descriptor colorAttachments][attachment].level = level;
[*descriptor colorAttachments][attachment].slice = arrayLayer;
[*descriptor colorAttachments][attachment].depthPlane = z;
attachment++;
if (attachment == kMaxColorAttachments) {
attachment = 0;
commandContext->BeginRender(descriptor.Get());
commandContext->EndRender();
descriptor = nullptr;
}
}
}
if (descriptor != nullptr) {
commandContext->BeginRender(descriptor.Get());
commandContext->EndRender();
}
}
}
} else {
Extent3D largestMipSize = GetMipLevelVirtualSize(range.baseMipLevel);
// Encode a buffer to texture copy to clear each subresource.
for (Aspect aspect : IterateEnumMask(range.aspects)) {
// Compute the buffer size big enough to fill the largest mip.
const TexelBlockInfo& blockInfo = GetFormat().GetAspectInfo(aspect).block;
// Metal validation layers: sourceBytesPerRow must be at least 64.
uint32_t largestMipBytesPerRow =
std::max((largestMipSize.width / blockInfo.width) * blockInfo.byteSize, 64u);
// Metal validation layers: sourceBytesPerImage must be at least 512.
uint64_t largestMipBytesPerImage =
std::max(static_cast<uint64_t>(largestMipBytesPerRow) *
(largestMipSize.height / blockInfo.height),
512llu);
uint64_t bufferSize = largestMipBytesPerImage * largestMipSize.depthOrArrayLayers;
if (bufferSize > std::numeric_limits<NSUInteger>::max()) {
return DAWN_OUT_OF_MEMORY_ERROR("Unable to allocate buffer.");
}
DynamicUploader* uploader = device->GetDynamicUploader();
UploadHandle uploadHandle;
DAWN_TRY_ASSIGN(uploadHandle,
uploader->Allocate(bufferSize, device->GetPendingCommandSerial(),
blockInfo.byteSize));
memset(uploadHandle.mappedBuffer, clearColor, bufferSize);
id<MTLBuffer> uploadBuffer =
ToBackend(uploadHandle.stagingBuffer)->GetBufferHandle();
for (uint32_t level = range.baseMipLevel;
level < range.baseMipLevel + range.levelCount; ++level) {
Extent3D virtualSize = GetMipLevelVirtualSize(level);
for (uint32_t arrayLayer = range.baseArrayLayer;
arrayLayer < range.baseArrayLayer + range.layerCount; ++arrayLayer) {
if (clearValue == TextureBase::ClearValue::Zero &&
IsSubresourceContentInitialized(
SubresourceRange::SingleMipAndLayer(level, arrayLayer, aspect))) {
// Skip lazy clears if already initialized.
continue;
}
MTLBlitOption blitOption = ComputeMTLBlitOption(GetFormat(), aspect);
[commandContext->EnsureBlit()
copyFromBuffer:uploadBuffer
sourceOffset:uploadHandle.startOffset
sourceBytesPerRow:largestMipBytesPerRow
sourceBytesPerImage:largestMipBytesPerImage
sourceSize:MTLSizeMake(virtualSize.width, virtualSize.height,
virtualSize.depthOrArrayLayers)
toTexture:GetMTLTexture()
destinationSlice:arrayLayer
destinationLevel:level
destinationOrigin:MTLOriginMake(0, 0, 0)
options:blitOption];
}
}
}
}
if (clearValue == TextureBase::ClearValue::Zero) {
SetIsSubresourceContentInitialized(true, range);
device->IncrementLazyClearCountForTesting();
}
return {};
}
void Texture::EnsureSubresourceContentInitialized(CommandRecordingContext* commandContext,
const SubresourceRange& range) {
if (!GetDevice()->IsToggleEnabled(Toggle::LazyClearResourceOnFirstUse)) {
return;
}
if (!IsSubresourceContentInitialized(range)) {
// If subresource has not been initialized, clear it to black as it could
// contain dirty bits from recycled memory
GetDevice()->ConsumedError(
ClearTexture(commandContext, range, TextureBase::ClearValue::Zero));
}
}
// static
ResultOrError<Ref<TextureView>> TextureView::Create(TextureBase* texture,
const TextureViewDescriptor* descriptor) {
return AcquireRef(new TextureView(texture, descriptor));
}
TextureView::TextureView(TextureBase* texture, const TextureViewDescriptor* descriptor)
: TextureViewBase(texture, descriptor) {
id<MTLTexture> mtlTexture = ToBackend(texture)->GetMTLTexture();
if (!UsageNeedsTextureView(texture->GetUsage())) {
mMtlTextureView = nullptr;
} else if (!RequiresCreatingNewTextureView(texture, descriptor)) {
mMtlTextureView = mtlTexture;
} else {
MTLPixelFormat format = MetalPixelFormat(descriptor->format);
if (descriptor->aspect == wgpu::TextureAspect::StencilOnly) {
if (@available(macOS 10.12, iOS 10.0, *)) {
ASSERT(format == MTLPixelFormatDepth32Float_Stencil8);
format = MTLPixelFormatX32_Stencil8;
} else {
// TODO(enga): Add a workaround to back combined depth/stencil textures
// with Sampled usage using two separate textures.
// Or, consider always using the workaround for D32S8.
GetDevice()->ConsumedError(
DAWN_DEVICE_LOST_ERROR("Cannot create stencil-only texture view of "
"combined depth/stencil format."));
}
}
MTLTextureType textureViewType =
MetalTextureViewType(descriptor->dimension, texture->GetSampleCount());
auto mipLevelRange = NSMakeRange(descriptor->baseMipLevel, descriptor->mipLevelCount);
auto arrayLayerRange =
NSMakeRange(descriptor->baseArrayLayer, descriptor->arrayLayerCount);
mMtlTextureView =
AcquireNSPRef([mtlTexture newTextureViewWithPixelFormat:format
textureType:textureViewType
levels:mipLevelRange
slices:arrayLayerRange]);
}
}
id<MTLTexture> TextureView::GetMTLTexture() {
ASSERT(mMtlTextureView != nullptr);
return mMtlTextureView.Get();
}
}} // namespace dawn_native::metal