mirror of https://github.com/PCSX2/pcsx2.git
gsdx-ogl: texture management
Improve arb_buffer_storage implementation Try harder to align the texture buffer Strangely arb_buffer_storage is 3 times slower on my PC (nvidia) Tester are welcome! Open the ini file "ogl_texture_storage = 1" <= enable the extension "ogl_texture_storage = 0" <= disable the extension Note: you need an opengl 4.4 driver or one that support arb_buffer_storage (i.e. not catalyst)
This commit is contained in:
Gregory Hainaut
parent
41091f8ebf
commit
403518e852
|
|
@ -29,52 +29,63 @@ namespace PboPool {
|
|||
|
||||
GLuint m_pool[PBO_POOL_SIZE];
|
||||
uint32 m_offset[PBO_POOL_SIZE];
|
||||
uint32 m_initial_offset[PBO_POOL_SIZE]; // work around silly driver
|
||||
char* m_map[PBO_POOL_SIZE];
|
||||
uint32 m_current_pbo = 0;
|
||||
uint32 m_size;
|
||||
const uint32 m_pbo_size = (640*480*16) << 2;
|
||||
bool m_buffer_storage = false;
|
||||
|
||||
// Option for buffer storage
|
||||
// Note there is a barrier (but maybe coherent is faster)
|
||||
const GLbitfield map_flags = GL_MAP_WRITE_BIT | GL_MAP_PERSISTENT_BIT | GL_MAP_COHERENT_BIT;
|
||||
// FIXME do I need GL_DYNAMIC_STORAGE_BIT to allow write?
|
||||
const GLbitfield create_flags = map_flags | GL_DYNAMIC_STORAGE_BIT;
|
||||
|
||||
// Normally driver must aligned the map....
|
||||
void* align_map(void* ptr) {
|
||||
void* aligned_map = (char*)(((uptr)ptr + 63) & ~0x3F);
|
||||
|
||||
m_initial_offset[m_current_pbo] = (uptr)aligned_map-(uptr)ptr;
|
||||
if (m_initial_offset[m_current_pbo])
|
||||
fprintf(stderr, "Buggy driver detected!!! Buffer alignment is not 64B as the spec request it\n");
|
||||
|
||||
return aligned_map;
|
||||
}
|
||||
|
||||
void Init() {
|
||||
gl_GenBuffers(countof(m_pool), m_pool);
|
||||
m_buffer_storage = (theApp.GetConfig("ogl_texture_storage", 0) == 1) && GLLoader::found_GL_ARB_buffer_storage;
|
||||
|
||||
for (size_t i = 0; i < countof(m_pool); i++) {
|
||||
BindPbo();
|
||||
|
||||
if (GLLoader::found_GL_ARB_buffer_storage) {
|
||||
gl_BufferStorage(GL_PIXEL_UNPACK_BUFFER, m_pbo_size, NULL, GL_MAP_WRITE_BIT | GL_MAP_PERSISTENT_BIT | GL_DYNAMIC_STORAGE_BIT | GL_CLIENT_STORAGE_BIT);
|
||||
// Note the +64 gives additional room to realign the buffer (buggy driver....)
|
||||
if (m_buffer_storage) {
|
||||
gl_BufferStorage(GL_PIXEL_UNPACK_BUFFER, m_pbo_size+64, NULL, create_flags);
|
||||
m_map[m_current_pbo] = (char*)align_map(gl_MapBufferRange(GL_PIXEL_UNPACK_BUFFER, 0, m_pbo_size+64, map_flags));
|
||||
// Workaround silly driver. (would be 0 otherwise)
|
||||
m_offset[m_current_pbo] = m_initial_offset[m_current_pbo];
|
||||
} else {
|
||||
gl_BufferData(GL_PIXEL_UNPACK_BUFFER, m_pbo_size, NULL, GL_STREAM_COPY);
|
||||
}
|
||||
|
||||
m_offset[m_current_pbo] = 0;
|
||||
gl_BufferData(GL_PIXEL_UNPACK_BUFFER, m_pbo_size+64, NULL, GL_STREAM_COPY);
|
||||
m_map[m_current_pbo] = NULL;
|
||||
|
||||
NextPbo();
|
||||
}
|
||||
UnbindPbo();
|
||||
m_offset[m_current_pbo] = 0;
|
||||
}
|
||||
|
||||
void MapAll() {
|
||||
if (m_map[m_current_pbo] != NULL) return;
|
||||
|
||||
// FIXME I'm not sure it is allowed to map another buffer after we get a pointer
|
||||
GLbitfield flags = GL_MAP_WRITE_BIT | GL_MAP_UNSYNCHRONIZED_BIT | GL_MAP_INVALIDATE_BUFFER_BIT | GL_MAP_PERSISTENT_BIT;
|
||||
for (size_t i = 0; i < countof(m_pool); i++) {
|
||||
BindPbo();
|
||||
m_map[m_current_pbo] = (char*)gl_MapBufferRange(GL_PIXEL_UNPACK_BUFFER, 0, m_pbo_size, flags);
|
||||
NextPbo();
|
||||
}
|
||||
UnbindPbo();
|
||||
}
|
||||
|
||||
char* Map(uint32 size) {
|
||||
char* map;
|
||||
m_size = size;
|
||||
|
||||
if (m_size >= m_pbo_size) {
|
||||
fprintf(stderr, "BUG: PBO too small %d but need %d\n", m_pbo_size, m_size);
|
||||
}
|
||||
|
||||
if (!GLLoader::found_GL_ARB_buffer_storage) {
|
||||
if (!m_buffer_storage) {
|
||||
GLbitfield flags = GL_MAP_WRITE_BIT | GL_MAP_UNSYNCHRONIZED_BIT | GL_MAP_INVALIDATE_RANGE_BIT;
|
||||
|
||||
if (m_offset[m_current_pbo] + m_size >= m_pbo_size) {
|
||||
|
|
@ -90,26 +101,28 @@ namespace PboPool {
|
|||
// Pbo ready let's get a pointer
|
||||
BindPbo();
|
||||
|
||||
return (char*)gl_MapBufferRange(GL_PIXEL_UNPACK_BUFFER, m_offset[m_current_pbo], m_size, flags);
|
||||
} else {
|
||||
MapAll();
|
||||
// Be sure the map is aligned
|
||||
map = (char*)gl_MapBufferRange(GL_PIXEL_UNPACK_BUFFER, m_offset[m_current_pbo], m_size, flags);
|
||||
|
||||
} else {
|
||||
if (m_offset[m_current_pbo] + m_size >= m_pbo_size) {
|
||||
NextPbo();
|
||||
|
||||
// Mark current pbo free
|
||||
m_offset[m_current_pbo] = 0;
|
||||
// Mark current pbo free. Will be 0 if driver aligned properly the buffer
|
||||
m_offset[m_current_pbo] = m_initial_offset[m_current_pbo];
|
||||
}
|
||||
|
||||
// Note: it still need it because texsubimage will access currently bound buffer
|
||||
// Note: texsubimage will access currently bound buffer
|
||||
// Pbo ready let's get a pointer
|
||||
BindPbo();
|
||||
|
||||
return m_map[m_current_pbo] + m_offset[m_current_pbo];
|
||||
}
|
||||
map = m_map[m_current_pbo] + m_offset[m_current_pbo];
|
||||
}
|
||||
|
||||
// FIXME: unmap buffer when the context is dettached (not sure it is required actually)
|
||||
return map;
|
||||
}
|
||||
|
||||
// Used to unmap the buffer when context was detached.
|
||||
void UnmapAll() {
|
||||
if (m_map[m_current_pbo] == NULL) return;
|
||||
|
||||
|
|
@ -123,8 +136,8 @@ namespace PboPool {
|
|||
}
|
||||
|
||||
void Unmap() {
|
||||
if (GLLoader::found_GL_ARB_buffer_storage) {
|
||||
gl_MemoryBarrier(GL_CLIENT_MAPPED_BUFFER_BARRIER_BIT);
|
||||
if (m_buffer_storage) {
|
||||
//gl_MemoryBarrier(GL_CLIENT_MAPPED_BUFFER_BARRIER_BIT);
|
||||
} else {
|
||||
gl_UnmapBuffer(GL_PIXEL_UNPACK_BUFFER);
|
||||
}
|
||||
|
|
@ -135,6 +148,8 @@ namespace PboPool {
|
|||
}
|
||||
|
||||
void Destroy() {
|
||||
if (m_buffer_storage)
|
||||
UnmapAll();
|
||||
gl_DeleteBuffers(countof(m_pool), m_pool);
|
||||
}
|
||||
|
||||
|
|
@ -152,7 +167,7 @@ namespace PboPool {
|
|||
|
||||
void EndTransfer() {
|
||||
// Note: keep offset aligned for SSE/AVX
|
||||
m_offset[m_current_pbo] = (m_offset[m_current_pbo] + m_size + 31) & ~0x1F;
|
||||
m_offset[m_current_pbo] = (m_offset[m_current_pbo] + m_size + 63) & ~0x3F;
|
||||
}
|
||||
}
|
||||
|
||||
|
|
@ -319,11 +334,11 @@ bool GSTextureOGL::Update(const GSVector4i& r, const void* data, int pitch)
|
|||
char* map = PboPool::Map(r.height() * line_size);
|
||||
|
||||
for (uint32 h = r.height(); h > 0; h--) {
|
||||
// avoid a crash if map is not aligned
|
||||
if ((uint32)map & 0x1F)
|
||||
if ((uptr)map & 0x3F) {
|
||||
memcpy(map, src, line_size);
|
||||
else
|
||||
} else {
|
||||
GSVector4i::storent(map, src, line_size);
|
||||
}
|
||||
src += pitch;
|
||||
map += line_size;
|
||||
}
|
||||
|
|
@ -333,7 +348,6 @@ bool GSTextureOGL::Update(const GSVector4i& r, const void* data, int pitch)
|
|||
glTexSubImage2D(GL_TEXTURE_2D, 0, r.x, r.y, r.width(), r.height(), m_int_format, m_int_type, (const void*)PboPool::Offset());
|
||||
|
||||
// FIXME OGL4: investigate, only 1 unpack buffer always bound
|
||||
//if (!GLLoader::found_GL_ARB_buffer_storage)
|
||||
PboPool::UnbindPbo();
|
||||
|
||||
PboPool::EndTransfer();
|
||||
|
|
|
|||
|
|
@ -60,6 +60,11 @@ typedef unsigned int uint32;
|
|||
typedef signed int int32;
|
||||
typedef unsigned long long uint64;
|
||||
typedef signed long long int64;
|
||||
#ifdef __x86_64__
|
||||
typedef uint64 uptr;
|
||||
#else
|
||||
typedef uint32 uptr;
|
||||
#endif
|
||||
|
||||
// stdc
|
||||
|
||||
|
|
|
|||
Loading…
Reference in New Issue