1
0
Fork 0

initial sloppy texture dumping implementation (opengl only)

This commit is contained in:
khang06 2019-04-03 21:42:21 -04:00 committed by James Rowe
parent 94b3c63bf9
commit deff865ac9
1 changed files with 76 additions and 14 deletions

View File

@ -14,6 +14,7 @@
#include <vector>
#include <boost/range/iterator_range.hpp>
#include <glad/glad.h>
#include <lodepng.h>
#include "common/alignment.h"
#include "common/bit_field.h"
#include "common/color.h"
@ -22,8 +23,11 @@
#include "common/microprofile.h"
#include "common/scope_exit.h"
#include "common/vector_math.h"
#include "core/core.h"
#include "core/frontend/emu_window.h"
#include "core/hle/kernel/process.h"
#include "core/memory.h"
#include "core/settings.h"
#include "video_core/pica_state.h"
#include "video_core/renderer_base.h"
#include "video_core/renderer_opengl/gl_rasterizer_cache.h"
@ -860,6 +864,27 @@ void CachedSurface::UploadGLTexture(const Common::Rectangle<u32>& rect, GLuint r
ASSERT(gl_buffer_size == width * height * GetGLBytesPerPixel(pixel_format));
// Decode and dump texture if texture dumping is enabled
bool should_dump = false;
std::string dump_path;
if (Settings::values.dump_textures) {
dump_path = fmt::format("{}/textures", FileUtil::GetUserPath(FileUtil::UserPath::DumpDir));
if (!FileUtil::IsDirectory(dump_path))
FileUtil::CreateDir(dump_path);
dump_path += fmt::format(
"/{:016X}",
Core::System::GetInstance().Kernel().GetCurrentProcess()->codeset->program_id);
if (!FileUtil::IsDirectory(dump_path))
FileUtil::CreateDir(dump_path);
// Hash the encoded texture
const u64 tex_hash = Common::ComputeHash64(gl_buffer.get(), gl_buffer_size);
dump_path += fmt::format("/tex1_{}x{}_{:016X}_{}.png", width, height, tex_hash,
static_cast<u32>(pixel_format));
if (!FileUtil::Exists(dump_path)) {
should_dump = true;
}
}
// Load data from memory to the surface
GLint x0 = static_cast<GLint>(rect.left);
GLint y0 = static_cast<GLint>(rect.bottom);
@ -896,6 +921,37 @@ void CachedSurface::UploadGLTexture(const Common::Rectangle<u32>& rect, GLuint r
&gl_buffer[buffer_offset]);
glPixelStorei(GL_UNPACK_ROW_LENGTH, 0);
if (should_dump) {
// Dump texture to RGBA8 and encode as PNG
LOG_INFO(Render_OpenGL, "Dumping texture to {}", dump_path);
std::vector<u8> decoded_texture;
std::vector<u8> png;
decoded_texture.resize(width * height * 4);
glBindTexture(GL_TEXTURE_2D, target_tex);
glGetTexImage(GL_TEXTURE_2D, 0, GL_RGBA, GL_UNSIGNED_BYTE, &decoded_texture[0]);
glBindTexture(GL_TEXTURE_2D, 0);
// Flip the RGBA8 texture
assert(rgba8_tex.size() = width * height * 4);
const u64 line_size = width * 4;
// Thanks MSVC for not being able to make variable length arrays
u8* temp_row = new u8[line_size];
u32 offset_1;
u32 offset_2;
for (u64 line = 0; line < height / 2; line++) {
offset_1 = line * line_size;
offset_2 = (height - line - 1) * line_size;
// Swap lines
std::memcpy(temp_row, &decoded_texture[offset_1], line_size);
std::memcpy(&decoded_texture[offset_1], &decoded_texture[offset_2], line_size);
std::memcpy(&decoded_texture[offset_2], temp_row, line_size);
}
delete temp_row;
u32 png_error = lodepng::encode(dump_path, decoded_texture, width, height);
if (png_error) {
LOG_CRITICAL(Render_OpenGL, "Failed to save decoded texture! {}",
lodepng_error_text(png_error));
}
}
cur_state.texture_units[0].texture_2d = old_tex;
cur_state.Apply();
@ -1045,7 +1101,8 @@ Surface FindMatch(const SurfaceCache& surface_cache, const SurfaceParams& params
surface->type != SurfaceType::Fill)
return;
// Found a match, update only if this is better than the previous one
// Found a match, update only if this is better than the previous
// one
auto UpdateMatch = [&] {
match_surface = surface;
match_valid = is_valid;
@ -1244,8 +1301,8 @@ Surface RasterizerCacheOpenGL::GetSurface(const SurfaceParams& params, ScaleMatc
if (surface == nullptr) {
u16 target_res_scale = params.res_scale;
if (match_res_scale != ScaleMatch::Exact) {
// This surface may have a subrect of another surface with a higher res_scale, find it
// to adjust our params
// This surface may have a subrect of another surface with a higher
// res_scale, find it to adjust our params
SurfaceParams find_params = params;
Surface expandable = FindMatch<MatchFlags::Expand | MatchFlags::Invalid>(
surface_cache, find_params, match_res_scale);
@ -1346,7 +1403,8 @@ SurfaceRect_Tuple RasterizerCacheOpenGL::GetSurfaceSubRect(const SurfaceParams&
// Can't have gaps in a surface
new_params.width = aligned_params.stride;
new_params.UpdateParams();
// GetSurface will create the new surface and possibly adjust res_scale if necessary
// GetSurface will create the new surface and possibly adjust res_scale if
// necessary
surface = GetSurface(new_params, match_res_scale, load_if_create);
} else if (load_if_create) {
ValidateSurface(surface, aligned_params.addr, aligned_params.size);
@ -1503,9 +1561,10 @@ const CachedTextureCube& RasterizerCacheOpenGL::GetTextureCube(const TextureCube
if (surface) {
face.watcher = surface->CreateWatcher();
} else {
// Can occur when texture address is invalid. We mark the watcher with nullptr in
// this case and the content of the face wouldn't get updated. These are usually
// leftover setup in the texture unit and games are not supposed to draw using them.
// Can occur when texture address is invalid. We mark the watcher
// with nullptr in this case and the content of the face wouldn't
// get updated. These are usually leftover setup in the texture unit
// and games are not supposed to draw using them.
face.watcher = nullptr;
}
}
@ -1605,7 +1664,8 @@ SurfaceSurfaceRect_Tuple RasterizerCacheOpenGL::GetFramebufferSurfaces(
auto color_vp_interval = color_params.GetSubRectInterval(viewport_clamped);
auto depth_vp_interval = depth_params.GetSubRectInterval(viewport_clamped);
// Make sure that framebuffers don't overlap if both color and depth are being used
// Make sure that framebuffers don't overlap if both color and depth are being
// used
if (using_color_fb && using_depth_fb &&
boost::icl::length(color_vp_interval & depth_vp_interval)) {
LOG_CRITICAL(Render_OpenGL, "Color and depth framebuffer memory regions overlap; "
@ -1793,9 +1853,10 @@ void RasterizerCacheOpenGL::FlushRegion(PAddr addr, u32 size, Surface flush_surf
SurfaceRegions flushed_intervals;
for (auto& pair : RangeFromInterval(dirty_regions, flush_interval)) {
// small sizes imply that this most likely comes from the cpu, flush the entire region
// the point is to avoid thousands of small writes every frame if the cpu decides to access
// that region, anything higher than 8 you're guaranteed it comes from a service
// small sizes imply that this most likely comes from the cpu, flush the
// entire region the point is to avoid thousands of small writes every frame
// if the cpu decides to access that region, anything higher than 8 you're
// guaranteed it comes from a service
const auto interval = size <= 8 ? pair.first : pair.first & flush_interval;
auto& surface = pair.second;
@ -1852,7 +1913,8 @@ void RasterizerCacheOpenGL::InvalidateRegion(PAddr addr, u32 size, const Surface
cached_surface->invalid_regions.insert(interval);
cached_surface->InvalidateAllWatcher();
// Remove only "empty" fill surfaces to avoid destroying and recreating OGL textures
// Remove only "empty" fill surfaces to avoid destroying and recreating
// OGL textures
if (cached_surface->type == SurfaceType::Fill &&
cached_surface->IsSurfaceFullyInvalid()) {
remove_surfaces.emplace(cached_surface);
@ -1921,8 +1983,8 @@ void RasterizerCacheOpenGL::UpdatePagesCachedCount(PAddr addr, u32 size, int del
const u32 page_start = addr >> Memory::PAGE_BITS;
const u32 page_end = page_start + num_pages;
// Interval maps will erase segments if count reaches 0, so if delta is negative we have to
// subtract after iterating
// Interval maps will erase segments if count reaches 0, so if delta is negative
// we have to subtract after iterating
const auto pages_interval = PageMap::interval_type::right_open(page_start, page_end);
if (delta > 0)
cached_pages.add({pages_interval, delta});