RPCSX/rpcsx
1
0
Fork 0
mirror of https://github.com/RPCSX/rpcsx.git synced 2026-04-20 22:05:12 +00:00
Code Issues Projects Releases Packages Wiki Activity

gl: Deswizzle textures on the GPU

Browse source
This commit is contained in:
kd-11 2022-05-27 23:42:57 +03:00 • committed by kd-11
parent 532563e861
commit 60a2a39e88
17 changed files with 413 additions and 222 deletions
Download patch file Download diff file Expand all files Collapse all files

96
rpcs3/Emu/RSX/GL/GLTexture.cpp
View file

@ -21,13 +21,13 @@ namespace gl
std::pair<buffer*, buffer*> prepare_compute_resources(usz staging_data_length)
{
if (g_upload_transfer_buffer.size() < staging_data_length)
if (g_upload_transfer_buffer.size() < static_cast<GLsizeiptr>(staging_data_length))
{
g_upload_transfer_buffer.remove();
g_upload_transfer_buffer.create(staging_data_length, nullptr, buffer::memory_type::host_visible, GL_STREAM_DRAW);
}
if (g_compute_decode_buffer.size() < staging_data_length * 3)
if (g_compute_decode_buffer.size() < static_cast<GLsizeiptr>(staging_data_length) * 3)
{
g_compute_decode_buffer.remove();
g_compute_decode_buffer.create(std::max<GLsizeiptr>(512, staging_data_length * 3), nullptr, buffer::memory_type::local, GL_STATIC_COPY);
@ -43,6 +43,31 @@ namespace gl
g_compute_decode_buffer.remove();
}
template <typename WordType, bool SwapBytes>
void do_deswizzle_transformation(gl::command_context& cmd, u32 block_size, buffer* dst, buffer* src, u32 data_length, u16 width, u16 height, u16 depth)
{
switch (block_size)
{
case 4:
gl::get_compute_task<gl::cs_deswizzle_3d<u32, WordType, SwapBytes>>()->run(
cmd, dst, 0, src, 0,
data_length, width, height, depth, 1);
break;
case 8:
gl::get_compute_task<gl::cs_deswizzle_3d<u64, WordType, SwapBytes>>()->run(
cmd, dst, 0, src, 0,
data_length, width, height, depth, 1);
break;
case 16:
gl::get_compute_task<gl::cs_deswizzle_3d<u128, WordType, SwapBytes>>()->run(
cmd, dst, 0, src, 0,
data_length, width, height, depth, 1);
break;
default:
fmt::throw_exception("Unreachable");
}
}
GLenum get_target(rsx::texture_dimension_extended type)
{
switch (type)
@ -623,11 +648,12 @@ namespace gl
const std::vector<rsx::subresource_layout> &input_layouts,
bool is_swizzled, GLenum gl_format, GLenum gl_type, std::vector<std::byte>& staging_buffer)
{
const auto driver_caps = gl::get_driver_caps();
rsx::texture_uploader_capabilities caps
{
.supports_byteswap = true,
.supports_vtc_decoding = false,
.supports_hw_deswizzle = false,
.supports_hw_deswizzle = driver_caps.ARB_compute_shader_supported,
.supports_zero_copy = false,
.alignment = 4
};
@ -635,9 +661,12 @@ namespace gl
pixel_unpack_settings unpack_settings;
unpack_settings.row_length(0).alignment(4);
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, GL_NONE);
glBindBuffer(GL_PIXEL_PACK_BUFFER, GL_NONE);
if (rsx::is_compressed_host_format(format)) [[likely]]
{
caps.supports_vtc_decoding = gl::get_driver_caps().vendor_NVIDIA;
caps.supports_vtc_decoding = driver_caps.vendor_NVIDIA;
unpack_settings.apply();
glBindTexture(static_cast<GLenum>(dst->get_target()), dst->id());
@ -688,7 +717,7 @@ namespace gl
else
{
bool apply_settings = true;
bool use_compute_transform = false;
bool use_compute_transform = is_swizzled;
buffer *upload_scratch_mem = nullptr, *compute_scratch_mem = nullptr;
image_memory_requirements mem_info;
pixel_buffer_layout mem_layout;
@ -698,6 +727,8 @@ namespace gl
u8 block_size_in_bytes = rsx::get_format_block_size_in_bytes(format);
u64 image_linear_size;
gl::buffer deswizzle_buf;
switch (gl_type)
{
case GL_BYTE:
@ -710,8 +741,6 @@ namespace gl
case GL_FLOAT:
case GL_UNSIGNED_INT_24_8:
case GL_FLOAT_32_UNSIGNED_INT_24_8_REV:
mem_layout.format = gl_format;
mem_layout.type = gl_type;
mem_layout.swap_bytes = true;
mem_layout.size = 4;
use_compute_transform = true;
@ -747,14 +776,61 @@ namespace gl
if (use_compute_transform)
{
// 0. Preconf
mem_layout.swap_bytes = op.require_swap;
mem_layout.format = gl_format;
mem_layout.type = gl_type;
// 1. Unmap buffer
upload_scratch_mem->unmap();
// 2. Upload memory to GPU
upload_scratch_mem->copy_to(compute_scratch_mem, 0, 0, image_linear_size);
if (!op.require_deswizzle)
{
upload_scratch_mem->copy_to(compute_scratch_mem, 0, 0, image_linear_size);
}
else
{
// 2.1 Copy data to deswizzle buf
if (deswizzle_buf.size() < image_linear_size)
{
deswizzle_buf.remove();
deswizzle_buf.create(gl::buffer::target::ssbo, image_linear_size, nullptr, gl::buffer::memory_type::local);
}
upload_scratch_mem->copy_to(&deswizzle_buf, 0, 0, image_linear_size);
// 2.2 Apply compute transform to deswizzle input and dump it in compute_scratch_mem
ensure(op.element_size == 2 || op.element_size == 4);
const auto block_size = op.element_size * op.block_length;
if (op.require_swap)
{
mem_layout.swap_bytes = false;
if (op.element_size == 4) [[ likely ]]
{
do_deswizzle_transformation<u32, true>(cmd, block_size, compute_scratch_mem, &deswizzle_buf, image_linear_size, layout.width_in_texel, layout.height_in_texel, layout.depth);
}
else
{
do_deswizzle_transformation<u16, true>(cmd, block_size, compute_scratch_mem, &deswizzle_buf, image_linear_size, layout.width_in_texel, layout.height_in_texel, layout.depth);
}
}
else
{
if (op.element_size == 4) [[ likely ]]
{
do_deswizzle_transformation<u32, false>(cmd, block_size, compute_scratch_mem, &deswizzle_buf, image_linear_size, layout.width_in_texel, layout.height_in_texel, layout.depth);
}
else
{
do_deswizzle_transformation<u16, false>(cmd, block_size, compute_scratch_mem, &deswizzle_buf, image_linear_size, layout.width_in_texel, layout.height_in_texel, layout.depth);
}
}
}
// 3. Update configuration
mem_layout.swap_bytes = op.require_swap;
mem_info.image_size_in_texels = image_linear_size / block_size_in_bytes;
mem_info.image_size_in_bytes = image_linear_size;
mem_info.memory_required = 0;
@ -773,6 +849,8 @@ namespace gl
dst->copy_from(out_pointer, static_cast<texture::format>(gl_format), static_cast<texture::type>(gl_type), layout.level, region, unpack_settings);
}
}
deswizzle_buf.remove();
}
}