dolphin/Source/Plugins/Plugin_VideoOGL/Src/ProgramShaderCache.cpp
degasus fd06342a97 set hint GL_STREAM_READ
it's wrong, but so we are guaranteed to get pinned memory.
it's slower for rendering, but faster for mapping.
2013-01-28 13:03:31 +01:00

303 lines
8.8 KiB
C++

// Copyright (C) 2003 Dolphin Project.
// This program is free software: you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation, version 2.0.
// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU General Public License 2.0 for more details.
// A copy of the GPL 2.0 should have been included with the program.
// If not, see http://www.gnu.org/licenses/
// Official SVN repository and contact information can be found at
// http://code.google.com/p/dolphin-emu/
#include "ProgramShaderCache.h"
#include "MathUtil.h"
namespace OGL
{
static const u32 UBO_LENGTH = 1024*1024;
GLuint ProgramShaderCache::CurrentProgram = 0;
ProgramShaderCache::PCache ProgramShaderCache::pshaders;
GLuint ProgramShaderCache::s_ps_vs_ubo;
u32 ProgramShaderCache::s_ubo_iterator;
GLintptr ProgramShaderCache::s_vs_data_offset;
float *ProgramShaderCache::s_ubo_buffer;
u32 ProgramShaderCache::s_ubo_buffer_size;
bool ProgramShaderCache::s_ubo_dirty;
LinearDiskCache<ProgramShaderCache::ShaderUID, u8> g_program_disk_cache;
GLenum ProgramFormat;
GLuint ProgramShaderCache::PCacheEntry::prog_format = 0;
std::pair<u32, u32> ProgramShaderCache::CurrentShaderProgram;
const char *UniformNames[NUM_UNIFORMS] =
{
// PIXEL SHADER UNIFORMS
I_COLORS,
I_KCOLORS,
I_ALPHA,
I_TEXDIMS,
I_ZBIAS ,
I_INDTEXSCALE ,
I_INDTEXMTX,
I_FOG,
I_PLIGHTS,
I_PMATERIALS,
// VERTEX SHADER UNIFORMS
I_POSNORMALMATRIX,
I_PROJECTION ,
I_MATERIALS,
I_LIGHTS,
I_TEXMATRICES,
I_TRANSFORMMATRICES ,
I_NORMALMATRICES ,
I_POSTTRANSFORMMATRICES,
I_DEPTHPARAMS,
};
void ProgramShaderCache::SetProgramVariables(PCacheEntry &entry)
{
// Bind UBO
if (g_ActiveConfig.backend_info.bSupportsGLSLUBO)
{
GLint PSBlock_id = glGetUniformBlockIndex(entry.prog_id, "PSBlock");
GLint VSBlock_id = glGetUniformBlockIndex(entry.prog_id, "VSBlock");
if(PSBlock_id != -1)
glUniformBlockBinding(entry.prog_id, PSBlock_id, 1);
if(VSBlock_id != -1)
glUniformBlockBinding(entry.prog_id, VSBlock_id, 2);
}
// We cache our uniform locations for now
// Once we move up to a newer version of GLSL, ~1.30
// We can remove this
// (Sonicadvance): For some reason this fails on my hardware
//glGetUniformIndices(entry.prog_id, NUM_UNIFORMS, UniformNames, entry.UniformLocations);
// Got to do it this crappy way.
if (!g_ActiveConfig.backend_info.bSupportsGLSLUBO)
for (int a = 0; a < NUM_UNIFORMS; ++a)
entry.UniformLocations[a] = glGetUniformLocation(entry.prog_id, UniformNames[a]);
// Bind Texture Sampler
for (int a = 0; a <= 9; ++a)
{
char name[8];
snprintf(name, 8, "samp%d", a);
// Still need to get sampler locations since we aren't binding them statically in the shaders
int loc = glGetUniformLocation(entry.prog_id, name);
if (loc != -1)
glUniform1i(loc, a);
}
}
void ProgramShaderCache::SetProgramBindings ( ProgramShaderCache::PCacheEntry& entry )
{
if (g_ActiveConfig.backend_info.bSupportsDualSourceBlend)
{
// So we do support extended blending
// So we need to set a few more things here.
// Bind our out locations
glBindFragDataLocationIndexed(entry.prog_id, 0, 0, "ocol0");
glBindFragDataLocationIndexed(entry.prog_id, 0, 1, "ocol1");
}
else
glBindFragDataLocation(entry.prog_id, 0, "ocol0");
// Need to set some attribute locations
glBindAttribLocation(entry.prog_id, SHADER_POSITION_ATTRIB, "rawpos");
glBindAttribLocation(entry.prog_id, SHADER_POSMTX_ATTRIB, "fposmtx");
glBindAttribLocation(entry.prog_id, SHADER_COLOR0_ATTRIB, "color0");
glBindAttribLocation(entry.prog_id, SHADER_COLOR1_ATTRIB, "color1");
glBindAttribLocation(entry.prog_id, SHADER_NORM0_ATTRIB, "rawnorm0");
glBindAttribLocation(entry.prog_id, SHADER_NORM1_ATTRIB, "rawnorm1");
glBindAttribLocation(entry.prog_id, SHADER_NORM2_ATTRIB, "rawnorm2");
for(int i=0; i<8; i++) {
char attrib_name[8];
snprintf(attrib_name, 8, "tex%d", i);
glBindAttribLocation(entry.prog_id, SHADER_TEXTURE0_ATTRIB+i, attrib_name);
}
}
void ProgramShaderCache::SetBothShaders(GLuint PS, GLuint VS)
{
if(!PS || !VS) {
ERROR_LOG(VIDEO, "tried to bind a zero shader");
return;
}
std::pair<u32, u32> ShaderPair = std::make_pair(PS, VS);
// program is already bound
if(ShaderPair == CurrentShaderProgram) return;
PCache::iterator iter = pshaders.find(ShaderPair);
if (iter != pshaders.end())
{
PCacheEntry &entry = iter->second;
glUseProgram(entry.prog_id);
CurrentProgram = entry.prog_id;
CurrentShaderProgram = ShaderPair;
return;
}
PCacheEntry entry;
entry.Create(PS, VS);
// Right, the program is created now
// Let's attach everything
glAttachShader(entry.prog_id, entry.vsid);
glAttachShader(entry.prog_id, entry.psid);
if (g_ActiveConfig.backend_info.bSupportsGLSLCache)
glProgramParameteri(entry.prog_id, GL_PROGRAM_BINARY_RETRIEVABLE_HINT, GL_TRUE);
SetProgramBindings(entry);
glLinkProgram(entry.prog_id);
glUseProgram(entry.prog_id);
SetProgramVariables(entry);
pshaders[ShaderPair] = entry;
CurrentShaderProgram = ShaderPair;
CurrentProgram = entry.prog_id;
}
void ProgramShaderCache::SetMultiPSConstant4fv(unsigned int offset, const float *f, unsigned int count)
{
s_ubo_dirty = true;
memcpy(s_ubo_buffer+(offset*4), f, count*4*sizeof(float));
}
void ProgramShaderCache::SetMultiVSConstant4fv(unsigned int offset, const float *f, unsigned int count)
{
s_ubo_dirty = true;
memcpy(s_ubo_buffer+(offset*4)+s_vs_data_offset/sizeof(float), f, count*4*sizeof(float));
}
void ProgramShaderCache::UploadConstants()
{
if(s_ubo_dirty) {
if(s_ubo_iterator + s_ubo_buffer_size >= UBO_LENGTH) {
glBufferData(GL_UNIFORM_BUFFER, UBO_LENGTH, NULL, GL_STREAM_READ);
s_ubo_iterator = 0;
}
void *ubo = glMapBufferRange(GL_UNIFORM_BUFFER, s_ubo_iterator, s_ubo_buffer_size, GL_MAP_WRITE_BIT | GL_MAP_UNSYNCHRONIZED_BIT | GL_MAP_INVALIDATE_RANGE_BIT);
memcpy(ubo, s_ubo_buffer, s_ubo_buffer_size);
glUnmapBuffer(GL_UNIFORM_BUFFER);
glBindBufferRange(GL_UNIFORM_BUFFER, 1, s_ps_vs_ubo, s_ubo_iterator, s_vs_data_offset);
glBindBufferRange(GL_UNIFORM_BUFFER, 2, s_ps_vs_ubo, s_ubo_iterator + s_vs_data_offset, s_ubo_buffer_size - s_vs_data_offset);
s_ubo_iterator += s_ubo_buffer_size;
}
s_ubo_dirty = false;
}
GLuint ProgramShaderCache::GetCurrentProgram(void)
{
return CurrentProgram;
}
ProgramShaderCache::PCacheEntry ProgramShaderCache::GetShaderProgram(void)
{
return pshaders[CurrentShaderProgram];
}
void ProgramShaderCache::Init(void)
{
// We have to get the UBO alignment here because
// if we generate a buffer that isn't aligned
// then the UBO will fail.
if (g_ActiveConfig.backend_info.bSupportsGLSLUBO)
{
GLint Align;
glGetIntegerv(GL_UNIFORM_BUFFER_OFFSET_ALIGNMENT, &Align);
GLintptr const ps_data_size = ROUND_UP(C_PENVCONST_END * sizeof(float) * 4, Align);
GLintptr const vs_data_size = ROUND_UP(C_VENVCONST_END * sizeof(float) * 4, Align);
s_vs_data_offset = ps_data_size;
s_ubo_buffer_size = ps_data_size + vs_data_size;
// We multiply by *4*4 because we need to get down to basic machine units.
// So multiply by four to get how many floats we have from vec4s
// Then once more to get bytes
glGenBuffers(1, &s_ps_vs_ubo);
glBindBuffer(GL_UNIFORM_BUFFER, s_ps_vs_ubo);
glBufferData(GL_UNIFORM_BUFFER, UBO_LENGTH, NULL, GL_STREAM_READ);
s_ubo_iterator = 0;
s_ubo_buffer = new float[s_ubo_buffer_size/sizeof(float)];
memset(s_ubo_buffer, 0, s_ubo_buffer_size);
s_ubo_dirty = true;
}
// Read our shader cache, only if supported
if (g_ActiveConfig.backend_info.bSupportsGLSLCache)
{
PCacheEntry::prog_format = PCacheEntry::SetProgramFormat();
char cache_filename[MAX_PATH];
sprintf(cache_filename, "%sogl-%s-shaders.cache", File::GetUserPath(D_SHADERCACHE_IDX).c_str(),
SConfig::GetInstance().m_LocalCoreStartupParameter.m_strUniqueID.c_str());
ProgramShaderCacheInserter inserter;
g_program_disk_cache.OpenAndRead(cache_filename, inserter);
}
CurrentProgram = 0;
CurrentShaderProgram = std::pair<u32,u32>(0,0);
}
void ProgramShaderCache::Shutdown(void)
{
if (g_ActiveConfig.backend_info.bSupportsGLSLCache)
{
PCache::iterator iter = pshaders.begin();
for (; iter != pshaders.end(); ++iter)
{
g_program_disk_cache.Append(iter->second.uid, iter->second.GetProgram(), iter->second.Size());
iter->second.FreeProgram();
}
g_program_disk_cache.Sync();
g_program_disk_cache.Close();
}
glUseProgram(0);
PCache::iterator iter = pshaders.begin();
for (; iter != pshaders.end(); ++iter)
iter->second.Destroy();
pshaders.clear();
if (g_ActiveConfig.backend_info.bSupportsGLSLUBO)
{
glBindBuffer(GL_UNIFORM_BUFFER, 0);
glDeleteBuffers(1, &s_ps_vs_ubo);
s_ps_vs_ubo = 0;
delete [] s_ubo_buffer;
s_ubo_buffer = 0;
}
}
} // namespace OGL