mirror of
https://github.com/shadps4-emu/shadPS4.git
synced 2025-06-02 08:43:16 +00:00
shader_recompiler: Implement data share append and consume operations (#814)
* shader_recompiler: Add more format swap modes * texture_cache: Handle stencil texture reads * emulator: Support loading font library * readme: Add thanks section * shader_recompiler: Constant buffers as integers * shader_recompiler: Typed buffers as integers * shader_recompiler: Separate thread bit scalars * We can assume guest shader never mixes them with normal sgprs. This helps avoid errors where ssa could view an sgpr write dominating a thread bit read, due to how control flow is structurized, even though its not possible in actual control flow * shader_recompiler: Implement data append/consume operations * clang format * buffer_cache: Simplify invalidation scheme * video_core: Remove some invalidation remnants * adjust
This commit is contained in:
TheTurtle
GitHub
parent
649527a235
commit
13743b27fc
34 changed files with 512 additions and 272 deletions
|
|
@ -15,8 +15,9 @@
|
|||
namespace VideoCore {
|
||||
|
||||
static constexpr size_t NumVertexBuffers = 32;
|
||||
static constexpr size_t StagingBufferSize = 512_MB;
|
||||
static constexpr size_t UboStreamBufferSize = 64_MB;
|
||||
static constexpr size_t GdsBufferSize = 64_KB;
|
||||
static constexpr size_t StagingBufferSize = 1_GB;
|
||||
static constexpr size_t UboStreamBufferSize = 128_MB;
|
||||
|
||||
BufferCache::BufferCache(const Vulkan::Instance& instance_, Vulkan::Scheduler& scheduler_,
|
||||
const AmdGpu::Liverpool* liverpool_, TextureCache& texture_cache_,
|
||||
|
|
@ -25,7 +26,10 @@ BufferCache::BufferCache(const Vulkan::Instance& instance_, Vulkan::Scheduler& s
|
|||
texture_cache{texture_cache_}, tracker{tracker_},
|
||||
staging_buffer{instance, scheduler, MemoryUsage::Upload, StagingBufferSize},
|
||||
stream_buffer{instance, scheduler, MemoryUsage::Stream, UboStreamBufferSize},
|
||||
gds_buffer{instance, scheduler, MemoryUsage::Stream, 0, AllFlags, GdsBufferSize},
|
||||
memory_tracker{&tracker} {
|
||||
Vulkan::SetObjectName(instance.GetDevice(), gds_buffer.Handle(), "GDS Buffer");
|
||||
|
||||
// Ensure the first slot is used for the null buffer
|
||||
void(slot_buffers.insert(instance, scheduler, MemoryUsage::DeviceLocal, 0, ReadFlags, 1));
|
||||
}
|
||||
|
|
@ -232,6 +236,27 @@ u32 BufferCache::BindIndexBuffer(bool& is_indexed, u32 index_offset) {
|
|||
return regs.num_indices;
|
||||
}
|
||||
|
||||
void BufferCache::InlineDataToGds(u32 gds_offset, u32 value) {
|
||||
ASSERT_MSG(gds_offset % 4 == 0, "GDS offset must be dword aligned");
|
||||
scheduler.EndRendering();
|
||||
const auto cmdbuf = scheduler.CommandBuffer();
|
||||
const vk::BufferMemoryBarrier2 buf_barrier = {
|
||||
.srcStageMask = vk::PipelineStageFlagBits2::eTransfer,
|
||||
.srcAccessMask = vk::AccessFlagBits2::eTransferWrite,
|
||||
.dstStageMask = vk::PipelineStageFlagBits2::eAllCommands,
|
||||
.dstAccessMask = vk::AccessFlagBits2::eMemoryRead,
|
||||
.buffer = gds_buffer.Handle(),
|
||||
.offset = gds_offset,
|
||||
.size = sizeof(u32),
|
||||
};
|
||||
cmdbuf.pipelineBarrier2(vk::DependencyInfo{
|
||||
.dependencyFlags = vk::DependencyFlagBits::eByRegion,
|
||||
.bufferMemoryBarrierCount = 1,
|
||||
.pBufferMemoryBarriers = &buf_barrier,
|
||||
});
|
||||
cmdbuf.updateBuffer(gds_buffer.Handle(), gds_offset, sizeof(u32), &value);
|
||||
}
|
||||
|
||||
std::pair<Buffer*, u32> BufferCache::ObtainBuffer(VAddr device_addr, u32 size, bool is_written,
|
||||
bool is_texel_buffer) {
|
||||
static constexpr u64 StreamThreshold = CACHING_PAGESIZE;
|
||||
|
|
@ -258,6 +283,7 @@ std::pair<Buffer*, u32> BufferCache::ObtainTempBuffer(VAddr gpu_addr, u32 size)
|
|||
if (buffer_id) {
|
||||
Buffer& buffer = slot_buffers[buffer_id];
|
||||
if (buffer.IsInBounds(gpu_addr, size)) {
|
||||
SynchronizeBuffer(buffer, gpu_addr, size, false);
|
||||
return {&buffer, buffer.Offset(gpu_addr)};
|
||||
}
|
||||
}
|
||||
|
|
@ -541,64 +567,48 @@ void BufferCache::SynchronizeBuffer(Buffer& buffer, VAddr device_addr, u32 size,
|
|||
}
|
||||
|
||||
bool BufferCache::SynchronizeBufferFromImage(Buffer& buffer, VAddr device_addr, u32 size) {
|
||||
boost::container::small_vector<ImageId, 8> image_ids;
|
||||
const u32 inv_size = std::min(size, MaxInvalidateDist);
|
||||
texture_cache.ForEachImageInRegion(device_addr, inv_size, [&](ImageId image_id, Image& image) {
|
||||
// Only consider GPU modified images, i.e render targets or storage images.
|
||||
// Also avoid any CPU modified images as the image data is likely to be stale.
|
||||
if (True(image.flags & ImageFlagBits::CpuModified) ||
|
||||
False(image.flags & ImageFlagBits::GpuModified)) {
|
||||
return;
|
||||
}
|
||||
// Image must fully overlap with the provided buffer range.
|
||||
if (image.cpu_addr < device_addr || image.cpu_addr_end > device_addr + size) {
|
||||
return;
|
||||
}
|
||||
image_ids.push_back(image_id);
|
||||
});
|
||||
if (image_ids.empty()) {
|
||||
static constexpr FindFlags find_flags =
|
||||
FindFlags::NoCreate | FindFlags::RelaxDim | FindFlags::RelaxFmt | FindFlags::RelaxSize;
|
||||
ImageInfo info{};
|
||||
info.guest_address = device_addr;
|
||||
info.guest_size_bytes = size;
|
||||
const ImageId image_id = texture_cache.FindImage(info, find_flags);
|
||||
if (!image_id) {
|
||||
return false;
|
||||
}
|
||||
// Sort images by modification tick. If there are overlaps we want to
|
||||
// copy from least to most recently modified.
|
||||
std::ranges::sort(image_ids, [&](ImageId lhs_id, ImageId rhs_id) {
|
||||
const Image& lhs = texture_cache.GetImage(lhs_id);
|
||||
const Image& rhs = texture_cache.GetImage(rhs_id);
|
||||
return lhs.tick_accessed_last < rhs.tick_accessed_last;
|
||||
});
|
||||
boost::container::small_vector<vk::BufferImageCopy, 8> copies;
|
||||
for (const ImageId image_id : image_ids) {
|
||||
copies.clear();
|
||||
Image& image = texture_cache.GetImage(image_id);
|
||||
u32 offset = buffer.Offset(image.cpu_addr);
|
||||
const u32 num_layers = image.info.resources.layers;
|
||||
for (u32 m = 0; m < image.info.resources.levels; m++) {
|
||||
const u32 width = std::max(image.info.size.width >> m, 1u);
|
||||
const u32 height = std::max(image.info.size.height >> m, 1u);
|
||||
const u32 depth =
|
||||
image.info.props.is_volume ? std::max(image.info.size.depth >> m, 1u) : 1u;
|
||||
const auto& [mip_size, mip_pitch, mip_height, mip_ofs] = image.info.mips_layout[m];
|
||||
copies.push_back({
|
||||
.bufferOffset = offset,
|
||||
.bufferRowLength = static_cast<u32>(mip_pitch),
|
||||
.bufferImageHeight = static_cast<u32>(mip_height),
|
||||
.imageSubresource{
|
||||
.aspectMask = image.aspect_mask & ~vk::ImageAspectFlagBits::eStencil,
|
||||
.mipLevel = m,
|
||||
.baseArrayLayer = 0,
|
||||
.layerCount = num_layers,
|
||||
},
|
||||
.imageOffset = {0, 0, 0},
|
||||
.imageExtent = {width, height, depth},
|
||||
});
|
||||
offset += mip_ofs * num_layers;
|
||||
}
|
||||
scheduler.EndRendering();
|
||||
image.Transit(vk::ImageLayout::eTransferSrcOptimal, vk::AccessFlagBits::eTransferRead);
|
||||
const auto cmdbuf = scheduler.CommandBuffer();
|
||||
cmdbuf.copyImageToBuffer(image.image, vk::ImageLayout::eTransferSrcOptimal, buffer.buffer,
|
||||
copies);
|
||||
Image& image = texture_cache.GetImage(image_id);
|
||||
if (image.info.guest_size_bytes > size) {
|
||||
return false;
|
||||
}
|
||||
boost::container::small_vector<vk::BufferImageCopy, 8> copies;
|
||||
u32 offset = buffer.Offset(image.cpu_addr);
|
||||
const u32 num_layers = image.info.resources.layers;
|
||||
for (u32 m = 0; m < image.info.resources.levels; m++) {
|
||||
const u32 width = std::max(image.info.size.width >> m, 1u);
|
||||
const u32 height = std::max(image.info.size.height >> m, 1u);
|
||||
const u32 depth =
|
||||
image.info.props.is_volume ? std::max(image.info.size.depth >> m, 1u) : 1u;
|
||||
const auto& [mip_size, mip_pitch, mip_height, mip_ofs] = image.info.mips_layout[m];
|
||||
copies.push_back({
|
||||
.bufferOffset = offset,
|
||||
.bufferRowLength = static_cast<u32>(mip_pitch),
|
||||
.bufferImageHeight = static_cast<u32>(mip_height),
|
||||
.imageSubresource{
|
||||
.aspectMask = image.aspect_mask & ~vk::ImageAspectFlagBits::eStencil,
|
||||
.mipLevel = m,
|
||||
.baseArrayLayer = 0,
|
||||
.layerCount = num_layers,
|
||||
},
|
||||
.imageOffset = {0, 0, 0},
|
||||
.imageExtent = {width, height, depth},
|
||||
});
|
||||
offset += mip_ofs * num_layers;
|
||||
}
|
||||
scheduler.EndRendering();
|
||||
image.Transit(vk::ImageLayout::eTransferSrcOptimal, vk::AccessFlagBits::eTransferRead);
|
||||
const auto cmdbuf = scheduler.CommandBuffer();
|
||||
cmdbuf.copyImageToBuffer(image.image, vk::ImageLayout::eTransferSrcOptimal, buffer.buffer,
|
||||
copies);
|
||||
return true;
|
||||
}
|
||||
|
||||
|
|
|
|||
|
|
@ -57,6 +57,11 @@ public:
|
|||
PageManager& tracker);
|
||||
~BufferCache();
|
||||
|
||||
/// Returns a pointer to GDS device local buffer.
|
||||
[[nodiscard]] const Buffer* GetGdsBuffer() const noexcept {
|
||||
return &gds_buffer;
|
||||
}
|
||||
|
||||
/// Invalidates any buffer in the logical page range.
|
||||
void InvalidateMemory(VAddr device_addr, u64 size);
|
||||
|
||||
|
|
@ -66,6 +71,9 @@ public:
|
|||
/// Bind host index buffer for the current draw.
|
||||
u32 BindIndexBuffer(bool& is_indexed, u32 index_offset);
|
||||
|
||||
/// Writes a value to GDS buffer.
|
||||
void InlineDataToGds(u32 gds_offset, u32 value);
|
||||
|
||||
/// Obtains a buffer for the specified region.
|
||||
[[nodiscard]] std::pair<Buffer*, u32> ObtainBuffer(VAddr gpu_addr, u32 size, bool is_written,
|
||||
bool is_texel_buffer = false);
|
||||
|
|
@ -130,6 +138,7 @@ private:
|
|||
PageManager& tracker;
|
||||
StreamBuffer staging_buffer;
|
||||
StreamBuffer stream_buffer;
|
||||
Buffer gds_buffer;
|
||||
std::mutex mutex;
|
||||
Common::SlotVector<Buffer> slot_buffers;
|
||||
MemoryTracker memory_tracker;
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue