Compare commits

..

5 Commits

Author SHA1 Message Date
lizzie 4464c3dbcb Revert "[android] Add Enable GPU Buffer Readback option to Android (#4132)"
This reverts commit d142b5dd6a.
2026-06-29 04:17:07 +00:00
lizzie ebefa6ea19 Revert? 2026-06-29 04:17:07 +00:00
xbzk b6ee847947 [hle/am] make Service::Process move-only to fix #3908 KProcess use-after-free (#4137)
#3908 changed process_creation.*, CreateProcess/CreateApplicationProcess, to return std::optional<Process> instead of std::unique_ptr<Process>, so the AM sites now transfer a Process by value via make_unique<Service::Process>(*std::move(opt)).

The consequence: Process owns a refcounted KProcess* but its user-declared dtor suppressed the implicit move ctor, so that "move" silently shallow-copied and the temporary's dtor Close()/RemoveProcess()'d the shared handle -> use-after-free.

It's seems to be user end based, so whether it crashes may depend on machine, compiler, allocator reuse, refcount slack, and the AM event-observer thread race, idk. It reliably crashed my MSVC build at launching games (cstack: ProcessHolder -> MultiWait -> KSynchronizationObject::Wait -> null) multiple times.

Fix: give Process a move ctor that steals the handle (nulling the source so the moved-from dtor is a no-op) and delete copy/move-assign, making the optional<->unique_ptr transfer safe.

Bonus: explicited delete for the 3 kinds of assignment: copy ctor (the one used in eden), copy assign and move assign (currently unused) to force compile error if they ever come to use.

Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4137
Reviewed-by: Lizzie <lizzie@eden-emu.dev>
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
2026-06-29 02:39:11 +02:00
wildcard 6c16440996 [vulkan] Fix EDS 0-2 (#4117)
EDS3 states were dynamic just because the driver supports them. But in EDS0-EDS2, we does not actually emit the EDS3 dynamic commands for them. So the pipeline builder skipped baking some fixed graphics state into the Vulkan pipeline, but the runtime also did not set that state dynamically. That leaves wrong Vulkan state during rendering causing rendering glitches in XC2.

Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4117
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
2026-06-28 15:07:54 +02:00
lizzie 09c583506b [tests] Require backing base pointer to be nonnull for HostMemory tests (#4138)
A basic requirement, probably more helpful than outright crashing in the middle of a test.

Signed-off-by: lizzie <lizzie@eden-emu.dev>

Reviewed-on: https://git.eden-emu.dev/eden-emu/eden/pulls/4138
Reviewed-by: MaranBr <maranbr@eden-emu.dev>
Reviewed-by: CamilleLaVey <camillelavey99@gmail.com>
2026-06-28 08:01:41 +02:00
12 changed files with 713 additions and 414 deletions
+544 -329
View File
File diff suppressed because it is too large Load Diff
+32 -12
View File
@@ -1,23 +1,43 @@
// SPDX-License-Identifier: MPL-2.0
// Copyright 2022 Skyline Team and Contributors (https://github.com/skyline-emu/)
// Copyright © 2022 Skyline Team and Contributors (https://github.com/skyline-emu/)
#pragma once
#include <cstdint>
namespace bcn {
/// @brief Decodes a BC1 encoded image to R8G8B8A8
void DecodeBc1(const uint8_t *src, uint8_t *dst, size_t x, size_t y, size_t width, size_t height, bool isSigned);
/// @brief Decodes a BC2 encoded image to R8G8B8A8
void DecodeBc2(const uint8_t *src, uint8_t *dst, size_t x, size_t y, size_t width, size_t height, bool isSigned);
//// @brief Decodes a BC3 encoded image to R8G8B8A8
void DecodeBc3(const uint8_t *src, uint8_t *dst, size_t x, size_t y, size_t width, size_t height, bool isSigned);
/// @brief Decodes a BC4 encoded image to R8
/**
* @brief Decodes a BC1 encoded image to R8G8B8A8
*/
void DecodeBc1(const uint8_t *src, uint8_t *dst, size_t x, size_t y, size_t width, size_t height);
/**
* @brief Decodes a BC2 encoded image to R8G8B8A8
*/
void DecodeBc2(const uint8_t *src, uint8_t *dst, size_t x, size_t y, size_t width, size_t height);
/**
* @brief Decodes a BC3 encoded image to R8G8B8A8
*/
void DecodeBc3(const uint8_t *src, uint8_t *dst, size_t x, size_t y, size_t width, size_t height);
/**
* @brief Decodes a BC4 encoded image to R8
*/
void DecodeBc4(const uint8_t *src, uint8_t *dst, size_t x, size_t y, size_t width, size_t height, bool isSigned);
//// @brief Decodes a BC5 encoded image to R8G8
/**
* @brief Decodes a BC5 encoded image to R8G8
*/
void DecodeBc5(const uint8_t *src, uint8_t *dst, size_t x, size_t y, size_t width, size_t height, bool isSigned);
//// @brief Decodes a BC6 encoded image to R16G16B16A16
/**
* @brief Decodes a BC6 encoded image to R16G16B16A16
*/
void DecodeBc6(const uint8_t *src, uint8_t *dst, size_t x, size_t y, size_t width, size_t height, bool isSigned);
/// @brief Decodes a BC7 encoded image to R8G8B8A8
void DecodeBc7(const uint8_t *src, uint8_t *dst, size_t x, size_t y, size_t width, size_t height, bool isSigned);
/**
* @brief Decodes a BC7 encoded image to R8G8B8A8
*/
void DecodeBc7(const uint8_t *src, uint8_t *dst, size_t x, size_t y, size_t width, size_t height);
}
@@ -31,7 +31,6 @@ enum class BooleanSetting(override val key: String) : AbstractBooleanSetting {
RENDERER_REACTIVE_FLUSHING("use_reactive_flushing"),
ENABLE_BUFFER_HISTORY("enable_buffer_history"),
USE_OPTIMIZED_VERTEX_BUFFERS("use_optimized_vertex_buffers"),
ENABLE_GPU_BUFFER_READBACK("enable_gpu_buffer_readback"),
SYNC_MEMORY_OPERATIONS("sync_memory_operations"),
BUFFER_REORDER_DISABLE("disable_buffer_reorder"),
RENDERER_DEBUG("debug"),
@@ -806,13 +806,6 @@ abstract class SettingsItem(
descriptionId = R.string.enable_buffer_history_description
)
)
put(
SwitchSetting(
BooleanSetting.ENABLE_GPU_BUFFER_READBACK,
titleId = R.string.enable_gpu_buffer_readback,
descriptionId = R.string.enable_gpu_buffer_readback_description
)
)
put(
SwitchSetting(
BooleanSetting.USE_OPTIMIZED_VERTEX_BUFFERS,
@@ -292,7 +292,6 @@ class SettingsFragmentPresenter(
add(BooleanSetting.RENDERER_FORCE_MAX_CLOCK.key)
add(BooleanSetting.RENDERER_REACTIVE_FLUSHING.key)
add(BooleanSetting.ENABLE_BUFFER_HISTORY.key)
add(BooleanSetting.ENABLE_GPU_BUFFER_READBACK.key)
add(BooleanSetting.USE_OPTIMIZED_VERTEX_BUFFERS.key)
add(HeaderSetting(R.string.hacks))
@@ -503,8 +503,6 @@
<string name="renderer_reactive_flushing_description">Improves rendering accuracy in some games at the cost of performance.</string>
<string name="enable_buffer_history">Enable buffer history</string>
<string name="enable_buffer_history_description">Enables access to previous buffer states. This option may improve rendering quality and performance consistency in some games.</string>
<string name="enable_gpu_buffer_readback">Enable GPU Buffer Readback</string>
<string name="enable_gpu_buffer_readback_description">Preserves GPU-modified buffer data by reading it back before uploads. Some games require this to render certain effects properly. May cause issues if the hardware cannot handle the additional workload.</string>
<string name="use_optimized_vertex_buffers">Optimized Vertex Buffers</string>
<string name="use_optimized_vertex_buffers_description">Enables optimized vertex buffer binding for improved performance. Requires Mesa 26.0+ Turnip drivers/ QCOM drivers. Will crash on older Turnip drivers (25.3 and below).</string>
+11
View File
@@ -6,6 +6,8 @@
#pragma once
#include <utility>
#include "common/common_types.h"
namespace Core {
@@ -28,6 +30,15 @@ public:
inline explicit Process(Core::System& system) noexcept : m_system(system) {}
inline ~Process() { this->Finalize(); }
Process(const Process&) = delete;
Process& operator=(const Process&) = delete;
Process& operator=(Process&&) = delete;
inline Process(Process&& other) noexcept
: m_system(other.m_system), m_process(std::exchange(other.m_process, nullptr)),
m_main_thread_stack_size(std::exchange(other.m_main_thread_stack_size, 0)),
m_main_thread_priority(std::exchange(other.m_main_thread_priority, 0)),
m_process_started(std::exchange(other.m_process_started, false)) {}
bool Initialize(Loader::AppLoader& loader, Loader::ResultStatus& out_load_result);
void Finalize();
+27 -2
View File
@@ -1,3 +1,6 @@
// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
// SPDX-License-Identifier: GPL-3.0-or-later
// SPDX-FileCopyrightText: Copyright 2021 yuzu Emulator Project
// SPDX-License-Identifier: GPL-2.0-or-later
@@ -15,12 +18,19 @@ static constexpr auto PERMS = Common::MemoryPermission::ReadWrite;
static constexpr auto HEAP = false;
TEST_CASE("HostMemory: Initialize and deinitialize", "[common]") {
{ HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE); }
{ HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE); }
{
HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
REQUIRE(mem.BackingBasePointer() != nullptr);
}
{
HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
REQUIRE(mem.BackingBasePointer() != nullptr);
}
}
TEST_CASE("HostMemory: Simple map", "[common]") {
HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
REQUIRE(mem.BackingBasePointer() != nullptr);
mem.Map(0x5000, 0x8000, 0x1000, PERMS, HEAP);
volatile u8* const data = mem.VirtualBasePointer() + 0x5000;
@@ -30,6 +40,7 @@ TEST_CASE("HostMemory: Simple map", "[common]") {
TEST_CASE("HostMemory: Simple mirror map", "[common]") {
HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
REQUIRE(mem.BackingBasePointer() != nullptr);
mem.Map(0x5000, 0x3000, 0x2000, PERMS, HEAP);
mem.Map(0x8000, 0x4000, 0x1000, PERMS, HEAP);
@@ -41,6 +52,7 @@ TEST_CASE("HostMemory: Simple mirror map", "[common]") {
TEST_CASE("HostMemory: Simple unmap", "[common]") {
HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
REQUIRE(mem.BackingBasePointer() != nullptr);
mem.Map(0x5000, 0x3000, 0x2000, PERMS, HEAP);
volatile u8* const data = mem.VirtualBasePointer() + 0x5000;
@@ -52,6 +64,7 @@ TEST_CASE("HostMemory: Simple unmap", "[common]") {
TEST_CASE("HostMemory: Simple unmap and remap", "[common]") {
HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
REQUIRE(mem.BackingBasePointer() != nullptr);
mem.Map(0x5000, 0x3000, 0x2000, PERMS, HEAP);
volatile u8* const data = mem.VirtualBasePointer() + 0x5000;
@@ -69,6 +82,7 @@ TEST_CASE("HostMemory: Simple unmap and remap", "[common]") {
TEST_CASE("HostMemory: Nieche allocation", "[common]") {
HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
REQUIRE(mem.BackingBasePointer() != nullptr);
mem.Map(0x0000, 0, 0x20000, PERMS, HEAP);
mem.Unmap(0x0000, 0x4000, HEAP);
mem.Map(0x1000, 0, 0x2000, PERMS, HEAP);
@@ -78,6 +92,7 @@ TEST_CASE("HostMemory: Nieche allocation", "[common]") {
TEST_CASE("HostMemory: Full unmap", "[common]") {
HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
REQUIRE(mem.BackingBasePointer() != nullptr);
mem.Map(0x8000, 0, 0x4000, PERMS, HEAP);
mem.Unmap(0x8000, 0x4000, HEAP);
mem.Map(0x6000, 0, 0x16000, PERMS, HEAP);
@@ -85,6 +100,7 @@ TEST_CASE("HostMemory: Full unmap", "[common]") {
TEST_CASE("HostMemory: Right out of bounds unmap", "[common]") {
HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
REQUIRE(mem.BackingBasePointer() != nullptr);
mem.Map(0x0000, 0, 0x4000, PERMS, HEAP);
mem.Unmap(0x2000, 0x4000, HEAP);
mem.Map(0x2000, 0x80000, 0x4000, PERMS, HEAP);
@@ -92,6 +108,7 @@ TEST_CASE("HostMemory: Right out of bounds unmap", "[common]") {
TEST_CASE("HostMemory: Left out of bounds unmap", "[common]") {
HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
mem.Map(0x8000, 0, 0x4000, PERMS, HEAP);
mem.Unmap(0x6000, 0x4000, HEAP);
mem.Map(0x8000, 0, 0x2000, PERMS, HEAP);
@@ -99,6 +116,7 @@ TEST_CASE("HostMemory: Left out of bounds unmap", "[common]") {
TEST_CASE("HostMemory: Multiple placeholder unmap", "[common]") {
HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
REQUIRE(mem.BackingBasePointer() != nullptr);
mem.Map(0x0000, 0, 0x4000, PERMS, HEAP);
mem.Map(0x4000, 0, 0x1b000, PERMS, HEAP);
mem.Unmap(0x3000, 0x1c000, HEAP);
@@ -107,6 +125,7 @@ TEST_CASE("HostMemory: Multiple placeholder unmap", "[common]") {
TEST_CASE("HostMemory: Unmap between placeholders", "[common]") {
HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
REQUIRE(mem.BackingBasePointer() != nullptr);
mem.Map(0x0000, 0, 0x4000, PERMS, HEAP);
mem.Map(0x4000, 0, 0x4000, PERMS, HEAP);
mem.Unmap(0x2000, 0x4000, HEAP);
@@ -115,6 +134,7 @@ TEST_CASE("HostMemory: Unmap between placeholders", "[common]") {
TEST_CASE("HostMemory: Unmap to origin", "[common]") {
HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
REQUIRE(mem.BackingBasePointer() != nullptr);
mem.Map(0x4000, 0, 0x4000, PERMS, HEAP);
mem.Map(0x8000, 0, 0x4000, PERMS, HEAP);
mem.Unmap(0x4000, 0x4000, HEAP);
@@ -124,6 +144,7 @@ TEST_CASE("HostMemory: Unmap to origin", "[common]") {
TEST_CASE("HostMemory: Unmap to right", "[common]") {
HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
REQUIRE(mem.BackingBasePointer() != nullptr);
mem.Map(0x4000, 0, 0x4000, PERMS, HEAP);
mem.Map(0x8000, 0, 0x4000, PERMS, HEAP);
mem.Unmap(0x8000, 0x4000, HEAP);
@@ -132,6 +153,7 @@ TEST_CASE("HostMemory: Unmap to right", "[common]") {
TEST_CASE("HostMemory: Partial right unmap check bindings", "[common]") {
HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
REQUIRE(mem.BackingBasePointer() != nullptr);
mem.Map(0x4000, 0x10000, 0x4000, PERMS, HEAP);
volatile u8* const ptr = mem.VirtualBasePointer() + 0x4000;
@@ -144,6 +166,7 @@ TEST_CASE("HostMemory: Partial right unmap check bindings", "[common]") {
TEST_CASE("HostMemory: Partial left unmap check bindings", "[common]") {
HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
REQUIRE(mem.BackingBasePointer() != nullptr);
mem.Map(0x4000, 0x10000, 0x4000, PERMS, HEAP);
volatile u8* const ptr = mem.VirtualBasePointer() + 0x4000;
@@ -158,6 +181,7 @@ TEST_CASE("HostMemory: Partial left unmap check bindings", "[common]") {
TEST_CASE("HostMemory: Partial middle unmap check bindings", "[common]") {
HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
REQUIRE(mem.BackingBasePointer() != nullptr);
mem.Map(0x4000, 0x10000, 0x4000, PERMS, HEAP);
volatile u8* const ptr = mem.VirtualBasePointer() + 0x4000;
@@ -172,6 +196,7 @@ TEST_CASE("HostMemory: Partial middle unmap check bindings", "[common]") {
TEST_CASE("HostMemory: Partial sparse middle unmap and check bindings", "[common]") {
HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE);
REQUIRE(mem.BackingBasePointer() != nullptr);
mem.Map(0x4000, 0x10000, 0x2000, PERMS, HEAP);
mem.Map(0x6000, 0x20000, 0x2000, PERMS, HEAP);
+26 -24
View File
@@ -1619,35 +1619,36 @@ void BufferCache<P>::TouchBuffer(Buffer& buffer, BufferId buffer_id) noexcept {
template <class P>
bool BufferCache<P>::SynchronizeBuffer(Buffer& buffer, DAddr device_addr, u32 size) {
upload_copies.clear();
u64 total_size_bytes = 0;
u64 staging_offset = 0;
u64 largest_copy = 0;
const DAddr buffer_start = buffer.cpu_addr_cached;
memory_tracker.ForEachUploadRange(device_addr, size, [&](u64 device_addr_out, u64 range_size) {
upload_copies.push_back(BufferCopy{
.src_offset = total_size_bytes,
.dst_offset = device_addr_out - buffer_start,
.size = range_size,
DAddr buffer_start = buffer.CpuAddr();
auto push = [&](u64 start, u64 end) {
if (start >= end) {
return;
}
u64 sz = end - start;
upload_copies.push_back({
.src_offset = staging_offset,
.dst_offset = start - buffer_start,
.size = sz
});
total_size_bytes += range_size;
largest_copy = (std::max)(largest_copy, range_size);
staging_offset += sz;
largest_copy = (std::max)(largest_copy, sz);
};
memory_tracker.ForEachUploadRange(device_addr, size, [&](u64 addr, u64 range_size) {
u64 start = addr;
u64 end = addr + range_size;
gpu_modified_ranges.ForEachInRange(start, range_size, [&](u64 gstart, u64 gsize) {
u64 gend = gstart + gsize;
push(start, gstart);
start = (std::max)(start, gend);
});
push(start, end);
});
if (total_size_bytes == 0) {
if (upload_copies.empty()) {
return true;
}
if (Settings::values.enable_gpu_buffer_readback.GetValue()) {
u64 min_offset = (std::numeric_limits<u64>::max)();
u64 max_offset = 0;
for (const auto& copy : upload_copies) {
min_offset = (std::min)(min_offset, copy.dst_offset);
max_offset = (std::max)(max_offset, copy.dst_offset + copy.size);
}
const DAddr sync_addr = buffer.CpuAddr() + min_offset;
const u64 sync_size = max_offset - min_offset;
DownloadBufferMemory(buffer, sync_addr, sync_size);
}
const std::span<BufferCopy> copies_span(upload_copies.data(), upload_copies.size());
UploadMemory(buffer, total_size_bytes, largest_copy, copies_span);
any_buffer_uploaded = true;
UploadMemory(buffer, staging_offset, largest_copy, std::span(upload_copies));
return false;
}
@@ -1698,6 +1699,7 @@ void BufferCache<P>::MappedUploadMemory([[maybe_unused]] Buffer& buffer,
u8* const src_pointer = staging_pointer.data() + copy.src_offset;
const DAddr device_addr = buffer.CpuAddr() + copy.dst_offset;
device_memory.ReadBlockUnsafe(device_addr, src_pointer, copy.size);
// Apply the staging offset
copy.src_offset += upload_staging.offset;
}
@@ -492,10 +492,14 @@ PipelineCache::PipelineCache(Tegra::MaxwellDeviceMemoryManager& device_memory_,
device.IsExtExtendedDynamicState3BlendingSupported();
dynamic_features.has_extended_dynamic_state_3_enables =
device.IsExtExtendedDynamicState3EnablesSupported();
dynamic_features.has_dynamic_state3_depth_clamp_enable = false;
dynamic_features.has_dynamic_state3_depth_clamp_enable =
dynamic_features.has_extended_dynamic_state_3_enables &&
device.SupportsDynamicState3DepthClampEnable();
dynamic_features.has_dynamic_state3_logic_op_enable =
dynamic_features.has_extended_dynamic_state_3_enables &&
device.SupportsDynamicState3LogicOpEnable();
dynamic_features.has_dynamic_state3_line_stipple_enable =
dynamic_features.has_extended_dynamic_state_3_enables &&
device.SupportsDynamicState3LineStippleEnable();
// VIDS: Independent toggle (not affected by dyna_state levels)
+51 -25
View File
@@ -1,4 +1,4 @@
// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
// SPDX-FileCopyrightText: Copyright 2025 Eden Emulator Project
// SPDX-License-Identifier: GPL-3.0-or-later
// SPDX-FileCopyrightText: Copyright 2020 yuzu Emulator Project
@@ -9,7 +9,6 @@
#include <span>
#include <bc_decoder.h>
#include "common/assert.h"
#include "common/common_types.h"
#include "video_core/texture_cache/decode_bc.h"
@@ -23,8 +22,11 @@ using VideoCore::Surface::PixelFormat;
constexpr bool IsSigned(PixelFormat pixel_format) {
switch (pixel_format) {
case PixelFormat::BC4_SNORM:
case PixelFormat::BC4_UNORM:
case PixelFormat::BC5_SNORM:
case PixelFormat::BC5_UNORM:
case PixelFormat::BC6H_SFLOAT:
case PixelFormat::BC6H_UFLOAT:
return true;
default:
return false;
@@ -60,28 +62,9 @@ u32 ConvertedBytesPerBlock(VideoCore::Surface::PixelFormat pixel_format) {
}
}
void DecompressBCn(std::span<const u8> input, std::span<u8> output, BufferImageCopy& copy, VideoCore::Surface::PixelFormat pixel_format) {
auto const f = [pixel_format]{
switch (pixel_format) {
case PixelFormat::BC1_RGBA_UNORM:
case PixelFormat::BC1_RGBA_SRGB: return &bcn::DecodeBc1;
case PixelFormat::BC2_UNORM:
case PixelFormat::BC2_SRGB: return &bcn::DecodeBc2;
case PixelFormat::BC3_UNORM:
case PixelFormat::BC3_SRGB: return &bcn::DecodeBc3;
case PixelFormat::BC4_SNORM:
case PixelFormat::BC4_UNORM: return &bcn::DecodeBc4;
case PixelFormat::BC5_SNORM:
case PixelFormat::BC5_UNORM: return &bcn::DecodeBc5;
case PixelFormat::BC6H_SFLOAT:
case PixelFormat::BC6H_UFLOAT: return &bcn::DecodeBc6;
case PixelFormat::BC7_SRGB:
case PixelFormat::BC7_UNORM: return &bcn::DecodeBc7;
default:
UNREACHABLE_MSG("Unimplemented BCn decompression {}", pixel_format);
return &bcn::DecodeBc1;
}
}();
template <auto decompress, PixelFormat pixel_format>
void DecompressBlocks(std::span<const u8> input, std::span<u8> output, BufferImageCopy& copy,
bool is_signed = false) {
const u32 out_bpp = ConvertedBytesPerBlock(pixel_format);
const u32 block_size = BlockSize(pixel_format);
const u32 width = copy.image_extent.width;
@@ -99,7 +82,11 @@ void DecompressBCn(std::span<const u8> input, std::span<u8> output, BufferImageC
for (u32 x = 0; x < width; x += block_width) {
const u8* src = input.data() + src_offset;
u8* const dst = output.data() + dst_offset;
f(src, dst, x, y, width, height, IsSigned(pixel_format));
if constexpr (IsSigned(pixel_format)) {
decompress(src, dst, x, y, width, height, is_signed);
} else {
decompress(src, dst, x, y, width, height);
}
src_offset += block_size;
dst_offset += block_width * out_bpp;
}
@@ -109,4 +96,43 @@ void DecompressBCn(std::span<const u8> input, std::span<u8> output, BufferImageC
}
}
void DecompressBCn(std::span<const u8> input, std::span<u8> output, BufferImageCopy& copy,
VideoCore::Surface::PixelFormat pixel_format) {
switch (pixel_format) {
case PixelFormat::BC1_RGBA_UNORM:
case PixelFormat::BC1_RGBA_SRGB:
DecompressBlocks<bcn::DecodeBc1, PixelFormat::BC1_RGBA_UNORM>(input, output, copy);
break;
case PixelFormat::BC2_UNORM:
case PixelFormat::BC2_SRGB:
DecompressBlocks<bcn::DecodeBc2, PixelFormat::BC2_UNORM>(input, output, copy);
break;
case PixelFormat::BC3_UNORM:
case PixelFormat::BC3_SRGB:
DecompressBlocks<bcn::DecodeBc3, PixelFormat::BC3_UNORM>(input, output, copy);
break;
case PixelFormat::BC4_SNORM:
case PixelFormat::BC4_UNORM:
DecompressBlocks<bcn::DecodeBc4, PixelFormat::BC4_UNORM>(
input, output, copy, pixel_format == PixelFormat::BC4_SNORM);
break;
case PixelFormat::BC5_SNORM:
case PixelFormat::BC5_UNORM:
DecompressBlocks<bcn::DecodeBc5, PixelFormat::BC5_UNORM>(
input, output, copy, pixel_format == PixelFormat::BC5_SNORM);
break;
case PixelFormat::BC6H_SFLOAT:
case PixelFormat::BC6H_UFLOAT:
DecompressBlocks<bcn::DecodeBc6, PixelFormat::BC6H_UFLOAT>(
input, output, copy, pixel_format == PixelFormat::BC6H_SFLOAT);
break;
case PixelFormat::BC7_SRGB:
case PixelFormat::BC7_UNORM:
DecompressBlocks<bcn::DecodeBc7, PixelFormat::BC7_UNORM>(input, output, copy);
break;
default:
LOG_WARNING(HW_GPU, "Unimplemented BCn decompression {}", pixel_format);
}
}
} // namespace VideoCommon
+17 -10
View File
@@ -1,4 +1,4 @@
// SPDX-FileCopyrightText: Copyright 2026 Eden Emulator Project
// SPDX-FileCopyrightText: Copyright 2025 Eden Emulator Project
// SPDX-License-Identifier: GPL-3.0-or-later
// SPDX-FileCopyrightText: Copyright 2020 yuzu Emulator Project
@@ -922,7 +922,8 @@ boost::container::small_vector<BufferImageCopy, 16> UnswizzleImage(Tegra::Memory
return copies;
}
void ConvertImage(std::span<const u8> input, const ImageInfo& info, std::span<u8> output, std::span<BufferImageCopy> copies) {
void ConvertImage(std::span<const u8> input, const ImageInfo& info, std::span<u8> output,
std::span<BufferImageCopy> copies) {
u32 output_offset = 0;
Common::ScratchBuffer<u8> decode_scratch;
@@ -954,10 +955,10 @@ void ConvertImage(std::span<const u8> input, const ImageInfo& info, std::span<u8
} else if (astc) {
// BC1 uses 0.5 bytes per texel
// BC3 uses 1 byte per texel
auto const compress = recompression_setting == Settings::AstcRecompression::Bc1
? Tegra::Texture::BCN::CompressBC1
: Tegra::Texture::BCN::CompressBC3;
const auto bpp_div = compress == Tegra::Texture::BCN::CompressBC1 ? 2 : 1;
const auto compress = recompression_setting == Settings::AstcRecompression::Bc1
? Tegra::Texture::BCN::CompressBC1
: Tegra::Texture::BCN::CompressBC3;
const auto bpp_div = recompression_setting == Settings::AstcRecompression::Bc1 ? 2 : 1;
const u32 plane_dim = copy.image_extent.width * copy.image_extent.height;
const u32 level_size = plane_dim * copy.image_extent.depth *
@@ -974,12 +975,18 @@ void ConvertImage(std::span<const u8> input, const ImageInfo& info, std::span<u8
copy.image_subresource.num_layers * copy.image_extent.depth,
output.subspan(output_offset));
const u32 aligned_plane_dim = Common::AlignUp(copy.image_extent.width, 4) * Common::AlignUp(copy.image_extent.height, 4);
copy.buffer_size = (aligned_plane_dim * copy.image_extent.depth * copy.image_subresource.num_layers) / bpp_div;
output_offset += u32(copy.buffer_size);
const u32 aligned_plane_dim = Common::AlignUp(copy.image_extent.width, 4) *
Common::AlignUp(copy.image_extent.height, 4);
copy.buffer_size =
(aligned_plane_dim * copy.image_extent.depth * copy.image_subresource.num_layers) /
bpp_div;
output_offset += static_cast<u32>(copy.buffer_size);
} else {
DecompressBCn(input_offset, output.subspan(output_offset), copy, info.format);
output_offset += copy.image_extent.width * copy.image_extent.height * copy.image_subresource.num_layers * ConvertedBytesPerBlock(info.format);
output_offset += copy.image_extent.width * copy.image_extent.height *
copy.image_subresource.num_layers *
ConvertedBytesPerBlock(info.format);
}
copy.buffer_row_length = mip_size.width;