mirror of
https://github.com/dolphin-emu/dolphin.git
synced 2025-07-31 10:09:36 -06:00
Remove the global namespace a bit and remove some dead code.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@7043 8ced0084-cf51-0410-be5f-012b33b47a6e
This commit is contained in:
@ -25,7 +25,6 @@
|
||||
#include "HW/Memmap.h"
|
||||
#include "ConfigManager.h"
|
||||
|
||||
bool textureChanged[8];
|
||||
const bool renderFog = false;
|
||||
namespace BPFunctions
|
||||
{
|
||||
|
@ -27,7 +27,6 @@
|
||||
#include "HW/Memmap.h"
|
||||
|
||||
volatile bool g_bSkipCurrentFrame = false;
|
||||
volatile bool g_EFBAccessRequested = false;
|
||||
extern u8* g_pVideoData;
|
||||
|
||||
namespace
|
||||
@ -79,20 +78,9 @@ void Fifo_SetRendering(bool enabled)
|
||||
g_bSkipCurrentFrame = !enabled;
|
||||
}
|
||||
|
||||
// Executed from another thread, no the graphics thread!
|
||||
// Basically, all it does is set a flag so that the loop will eventually exit, then
|
||||
// waits for the event to be set, which happens when the loop does exit.
|
||||
// If we look stuck in here, then the video thread is stuck in something and won't exit
|
||||
// the loop. Switch to the video thread and investigate.
|
||||
void Fifo_ExitLoop()
|
||||
{
|
||||
Fifo_ExitLoopNonBlocking();
|
||||
EmuRunning = true;
|
||||
}
|
||||
|
||||
// May be executed from any thread, even the graphics thread.
|
||||
// Created to allow for self shutdown.
|
||||
void Fifo_ExitLoopNonBlocking()
|
||||
void Fifo_ExitLoop()
|
||||
{
|
||||
// This should break the wait loop in CPU thread
|
||||
CommandProcessor::fifo.bFF_GPReadEnable = false;
|
||||
|
@ -36,7 +36,6 @@ void Fifo_SendFifoData(u8* _uData, u32 len);
|
||||
// These two are for dual core mode only.
|
||||
void Fifo_EnterLoop();
|
||||
void Fifo_ExitLoop();
|
||||
void Fifo_ExitLoopNonBlocking();
|
||||
void Fifo_RunLoop(bool run);
|
||||
bool AtBreakpoint();
|
||||
void Fifo_DoState(PointerWrap &f);
|
||||
|
@ -213,7 +213,7 @@ bool FifoCommandRunnable()
|
||||
"* Some other sort of bug\n\n"
|
||||
"Dolphin will now likely crash or hang. Enjoy." , cmd_byte);
|
||||
Host_SysMessage(szTemp);
|
||||
Core::Callback_VideoLog(szTemp, true);
|
||||
Core::Callback_VideoLog(szTemp);
|
||||
{
|
||||
SCPFifoStruct &fifo = CommandProcessor::fifo;
|
||||
|
||||
@ -238,7 +238,7 @@ bool FifoCommandRunnable()
|
||||
,fifo.bFF_Breakpoint ? "true" : "false");
|
||||
|
||||
Host_SysMessage(szTmp);
|
||||
Core::Callback_VideoLog(szTmp, true);
|
||||
Core::Callback_VideoLog(szTmp);
|
||||
}
|
||||
}
|
||||
break;
|
||||
|
@ -39,7 +39,6 @@ files = [
|
||||
'VertexShaderManager.cpp',
|
||||
'VideoConfig.cpp',
|
||||
'VideoState.cpp',
|
||||
'XFBConvert.cpp',
|
||||
'XFMemory.cpp',
|
||||
'XFStructs.cpp',
|
||||
'memcpy_amd.cpp',
|
||||
|
@ -181,14 +181,14 @@ int TexDecoder_GetPaletteSize(int format)
|
||||
}
|
||||
}
|
||||
|
||||
inline u32 decodeIA8(u16 val)
|
||||
static inline u32 decodeIA8(u16 val)
|
||||
{
|
||||
int a = val >> 8;
|
||||
int i = val & 0xFF;
|
||||
return (a << 24) | (i << 16) | (i << 8) | i;
|
||||
}
|
||||
|
||||
inline u32 decode5A3(u16 val)
|
||||
static inline u32 decode5A3(u16 val)
|
||||
{
|
||||
int r,g,b,a;
|
||||
if ((val & 0x8000))
|
||||
@ -208,7 +208,7 @@ inline u32 decode5A3(u16 val)
|
||||
return (a << 24) | (r << 16) | (g << 8) | b;
|
||||
}
|
||||
|
||||
inline u32 decode5A3RGBA(u16 val)
|
||||
static inline u32 decode5A3RGBA(u16 val)
|
||||
{
|
||||
int r,g,b,a;
|
||||
if ((val&0x8000))
|
||||
@ -228,7 +228,7 @@ inline u32 decode5A3RGBA(u16 val)
|
||||
return r | (g<<8) | (b << 16) | (a << 24);
|
||||
}
|
||||
|
||||
inline u32 decode565RGBA(u16 val)
|
||||
static inline u32 decode565RGBA(u16 val)
|
||||
{
|
||||
int r,g,b,a;
|
||||
r=Convert5To8((val>>11) & 0x1f);
|
||||
@ -238,7 +238,7 @@ inline u32 decode565RGBA(u16 val)
|
||||
return r | (g<<8) | (b << 16) | (a << 24);
|
||||
}
|
||||
|
||||
inline u32 decodeIA8Swapped(u16 val)
|
||||
static inline u32 decodeIA8Swapped(u16 val)
|
||||
{
|
||||
int a = val & 0xFF;
|
||||
int i = val >> 8;
|
||||
|
@ -1,292 +0,0 @@
|
||||
// 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/
|
||||
|
||||
#if _WIN32
|
||||
#include <intrin.h>
|
||||
#endif
|
||||
|
||||
#include "OpenCL.h"
|
||||
|
||||
#include <xmmintrin.h>
|
||||
|
||||
#include "XFBConvert.h"
|
||||
#include "Common.h"
|
||||
|
||||
namespace {
|
||||
|
||||
const __m128i _bias1 = _mm_set_epi32(128/2 << 16, 0, 128/2 << 16, 16 << 16);
|
||||
const __m128i _bias2 = _mm_set_epi32(128/2 << 16, 16 << 16, 128/2 << 16, 0);
|
||||
|
||||
__m128i _y[256];
|
||||
__m128i _u[256];
|
||||
__m128i _v[256];
|
||||
|
||||
__m128i _r1[256];
|
||||
__m128i _r2[256];
|
||||
__m128i _g1[256];
|
||||
__m128i _g2[256];
|
||||
__m128i _b1[256];
|
||||
__m128i _b2[256];
|
||||
|
||||
} // namespace
|
||||
bool Inited = false;
|
||||
|
||||
cl_kernel To_kernel;
|
||||
cl_program To_program;
|
||||
cl_kernel From_kernel;
|
||||
cl_program From_program;
|
||||
|
||||
const char *__ConvertFromXFB = "int bound(int i) \n \
|
||||
{ \n \
|
||||
return (i>255)?255:((i<0)?0:i); \n \
|
||||
} \n \
|
||||
\n \
|
||||
void yuv2rgb(int y, int u, int v, int &r, int &g, int &b) \n \
|
||||
{ \n \
|
||||
b = bound((76283*(y - 16) + 132252*(u - 128))>>16); \n \
|
||||
g = bound((76283*(y - 16) - 53281 *(v - 128) - 25624*(u - 128))>>16); //last one u? \n \
|
||||
r = bound((76283*(y - 16) + 104595*(v - 128))>>16); \n \
|
||||
} \n \
|
||||
\n \
|
||||
void ConvertFromXFB(u32 *dst, const u8* _pXFB) \n \
|
||||
{ \n \
|
||||
const unsigned char *src = _pXFB; \n \
|
||||
int id = get_global_id(0); \n \
|
||||
int srcOffset = id * 4; \n \
|
||||
int dstOffset = id; \n \
|
||||
u32 numBlocks = (width * height) / 2; \n \
|
||||
\n \
|
||||
int Y1 = src[srcOffset]; \n \
|
||||
int U = src[srcOffset + 1]; \n \
|
||||
int Y2 = src[srcOffset + 2]; \n \
|
||||
int V = src[srcOffset + 3]; \n \
|
||||
\n \
|
||||
int r, g, b; \n \
|
||||
yuv2rgb(Y1,U,V, r,g,b); \n \
|
||||
dst[dstOffset] = 0xFF000000 | (r<<16) | (g<<8) | (b); \n \
|
||||
yuv2rgb(Y2,U,V, r,g,b); \n \
|
||||
dst[dstOffset + 1] = 0xFF000000 | (r<<16) | (g<<8) | (b); \n \
|
||||
} \n";
|
||||
|
||||
const char *__ConvertToXFB = "__kernel void ConvertToXFB(__global unsigned int *dst, __global const unsigned char* _pEFB) \n \
|
||||
{ \n \
|
||||
const unsigned char *src = _pEFB;\n \
|
||||
int id = get_global_id(0);\n \
|
||||
int srcOffset = id * 8; \n \
|
||||
\n \
|
||||
int y1 = (((16843 * src[srcOffset]) + (33030 * src[srcOffset + 1]) + (6423 * src[srcOffset + 2])) >> 16) + 16; \n \
|
||||
int u1 = ((-(9699 * src[srcOffset]) - (19071 * src[srcOffset + 1]) + (28770 * src[srcOffset + 2])) >> 16) + 128;\n \
|
||||
srcOffset += 4;\n \
|
||||
\n \
|
||||
int y2 = (((16843 * src[srcOffset]) + (33030 * src[srcOffset + 1]) + (6423 * src[srcOffset + 2])) >> 16) + 16;\n \
|
||||
int v2 = (((28770 * src[srcOffset]) - (24117 * src[srcOffset + 1]) - (4653 * src[srcOffset + 2])) >> 16) + 128;\n \
|
||||
\n \
|
||||
dst[id] = (v2 << 24) | (y2 << 16) | (u1 << 8) | (y1); \n \
|
||||
} \n ";
|
||||
|
||||
void InitKernels()
|
||||
{
|
||||
|
||||
|
||||
From_program = OpenCL::CompileProgram(__ConvertFromXFB);
|
||||
From_kernel = OpenCL::CompileKernel(From_program, "ConvertFromXFB");
|
||||
|
||||
To_program = OpenCL::CompileProgram(__ConvertToXFB);
|
||||
To_kernel = OpenCL::CompileKernel(To_program, "ConvertToXFB");
|
||||
Inited = true;
|
||||
}
|
||||
|
||||
void InitXFBConvTables()
|
||||
{
|
||||
for (int i = 0; i < 256; i++)
|
||||
{
|
||||
_y[i] = _mm_set_epi32(0xFFFFFFF, 76283*(i - 16), 76283*(i - 16), 76283*(i - 16));
|
||||
_u[i] = _mm_set_epi32( 0, 0, -25624 * (i - 128), 132252 * (i - 128));
|
||||
_v[i] = _mm_set_epi32( 0, 104595 * (i - 128), -53281 * (i - 128), 0);
|
||||
|
||||
_r1[i] = _mm_add_epi32(_mm_set_epi32( 28770 * i / 2, 0, -9699 * i / 2, 16843 * i),
|
||||
_bias1);
|
||||
_g1[i] = _mm_set_epi32(-24117 * i / 2, 0, -19071 * i / 2, 33030 * i);
|
||||
_b1[i] = _mm_set_epi32( -4653 * i / 2, 0, 28770 * i / 2, 6423 * i);
|
||||
|
||||
_r2[i] = _mm_add_epi32(_mm_set_epi32( 28770 * i / 2, 16843 * i, -9699 * i / 2, 0),
|
||||
_bias2);
|
||||
_g2[i] = _mm_set_epi32(-24117 * i / 2, 33030 * i, -19071 * i / 2, 0);
|
||||
_b2[i] = _mm_set_epi32( -4653 * i / 2, 6423 * i, 28770 * i / 2, 0);
|
||||
}
|
||||
}
|
||||
|
||||
void ConvertFromXFB(u32 *dst, const u8* _pXFB, int width, int height)
|
||||
{
|
||||
if (((size_t)dst & 0xF) != 0) {
|
||||
PanicAlert("ConvertFromXFB - unaligned destination");
|
||||
}
|
||||
const unsigned char *src = _pXFB;
|
||||
u32 numBlocks = ((width * height) / 2) / 2;
|
||||
if(!Inited)
|
||||
InitKernels();
|
||||
int err;
|
||||
|
||||
size_t global = 0; // global domain size for our calculation
|
||||
size_t local = 0; // local domain size for our calculation
|
||||
printf("width %d, height %d\n", width, height);
|
||||
// Create the input and output arrays in device memory for our calculation
|
||||
//
|
||||
cl_mem _dst = clCreateBuffer(OpenCL::g_context, CL_MEM_WRITE_ONLY, sizeof(unsigned int) * numBlocks, NULL, NULL);
|
||||
if (!dst)
|
||||
{
|
||||
printf("Error: Failed to allocate device memory!\n");
|
||||
exit(1);
|
||||
}
|
||||
cl_mem _src = clCreateBuffer(OpenCL::g_context, CL_MEM_READ_ONLY | CL_MEM_COPY_HOST_PTR, sizeof(unsigned char) * width * height, (void*)_pXFB, NULL);
|
||||
if (!src)
|
||||
{
|
||||
printf("Error: Failed to allocate device memory!\n");
|
||||
exit(1);
|
||||
}
|
||||
// Set the arguments to our compute kernel
|
||||
//
|
||||
err = 0;
|
||||
err = clSetKernelArg(From_kernel, 0, sizeof(cl_mem), &_dst);
|
||||
err |= clSetKernelArg(From_kernel, 1, sizeof(cl_mem), &_src);
|
||||
if (err != CL_SUCCESS)
|
||||
{
|
||||
printf("Error: Failed to set kernel arguments! %d\n", err);
|
||||
exit(1);
|
||||
}
|
||||
|
||||
// Get the maximum work group size for executing the kernel on the device
|
||||
//
|
||||
err = clGetKernelWorkGroupInfo(From_kernel, OpenCL::device_id, CL_KERNEL_WORK_GROUP_SIZE, sizeof(size_t), &local, NULL);
|
||||
if (err != CL_SUCCESS)
|
||||
{
|
||||
printf("Error: Failed to retrieve kernel work group info! %d\n", err);
|
||||
local = 64;
|
||||
}
|
||||
|
||||
// Execute the kernel over the entire range of our 1d input data set
|
||||
// using the maximum number of work group items for this device
|
||||
//
|
||||
global = numBlocks;
|
||||
if(global < local)
|
||||
{
|
||||
// Global can't be less than local
|
||||
}
|
||||
err = clEnqueueNDRangeKernel(OpenCL::g_cmdq, From_kernel, 1, NULL, &global, &local, 0, NULL, NULL);
|
||||
if (err != CL_SUCCESS)
|
||||
{
|
||||
printf("Error: Failed to execute kernel! %d\n", err);
|
||||
return;
|
||||
}
|
||||
|
||||
// Wait for the command commands to get serviced before reading back results
|
||||
//
|
||||
clFinish(OpenCL::g_cmdq);
|
||||
|
||||
// Read back the results from the device to verify the output
|
||||
//
|
||||
err = clEnqueueReadBuffer( OpenCL::g_cmdq, _dst, CL_TRUE, 0, sizeof(unsigned int) * numBlocks, dst, 0, NULL, NULL );
|
||||
if (err != CL_SUCCESS)
|
||||
{
|
||||
printf("Error: Failed to read output array! %d\n", err);
|
||||
exit(1);
|
||||
}
|
||||
clReleaseMemObject(_dst);
|
||||
clReleaseMemObject(_src);
|
||||
}
|
||||
|
||||
|
||||
void ConvertToXFB(u32 *dst, const u8* _pEFB, int width, int height)
|
||||
{
|
||||
const unsigned char *src = _pEFB;
|
||||
|
||||
u32 numBlocks = ((width * height) / 2) / 4;
|
||||
if (((size_t)dst & 0xF) != 0) {
|
||||
PanicAlert("ConvertToXFB - unaligned XFB");
|
||||
}
|
||||
if(!Inited)
|
||||
InitKernels();
|
||||
|
||||
int err;
|
||||
|
||||
size_t global = 0; // global domain size for our calculation
|
||||
size_t local = 0; // local domain size for our calculation
|
||||
printf("width %d, height %d\n", width, height);
|
||||
// Create the input and output arrays in device memory for our calculation
|
||||
//
|
||||
cl_mem _dst = clCreateBuffer(OpenCL::g_context, CL_MEM_WRITE_ONLY, sizeof(unsigned int) * numBlocks, NULL, NULL);
|
||||
if (!dst)
|
||||
{
|
||||
printf("Error: Failed to allocate device memory!\n");
|
||||
exit(1);
|
||||
}
|
||||
cl_mem _src = clCreateBuffer(OpenCL::g_context, CL_MEM_READ_ONLY | CL_MEM_COPY_HOST_PTR, sizeof(unsigned char) * width * height, (void*)_pEFB, NULL);
|
||||
if (!src)
|
||||
{
|
||||
printf("Error: Failed to allocate device memory!\n");
|
||||
exit(1);
|
||||
}
|
||||
// Set the arguments to our compute kernel
|
||||
//
|
||||
err = 0;
|
||||
err = clSetKernelArg(To_kernel, 0, sizeof(cl_mem), &_dst);
|
||||
err |= clSetKernelArg(To_kernel, 1, sizeof(cl_mem), &_src);
|
||||
if (err != CL_SUCCESS)
|
||||
{
|
||||
printf("Error: Failed to set kernel arguments! %d\n", err);
|
||||
exit(1);
|
||||
}
|
||||
|
||||
// Get the maximum work group size for executing the kernel on the device
|
||||
//
|
||||
err = clGetKernelWorkGroupInfo(To_kernel, OpenCL::device_id, CL_KERNEL_WORK_GROUP_SIZE, sizeof(size_t), &local, NULL);
|
||||
if (err != CL_SUCCESS)
|
||||
{
|
||||
printf("Error: Failed to retrieve kernel work group info! %d\n", err);
|
||||
local = 64;
|
||||
}
|
||||
|
||||
// Execute the kernel over the entire range of our 1d input data set
|
||||
// using the maximum number of work group items for this device
|
||||
//
|
||||
global = numBlocks;
|
||||
if(global < local)
|
||||
{
|
||||
// Global can't be less than local
|
||||
}
|
||||
err = clEnqueueNDRangeKernel(OpenCL::g_cmdq, To_kernel, 1, NULL, &global, &local, 0, NULL, NULL);
|
||||
if (err != CL_SUCCESS)
|
||||
{
|
||||
printf("Error: Failed to execute kernel! %d\n", err);
|
||||
return;
|
||||
}
|
||||
|
||||
// Wait for the command commands to get serviced before reading back results
|
||||
//
|
||||
clFinish(OpenCL::g_cmdq);
|
||||
|
||||
// Read back the results from the device to verify the output
|
||||
//
|
||||
err = clEnqueueReadBuffer( OpenCL::g_cmdq, _dst, CL_TRUE, 0, sizeof(unsigned int) * numBlocks, dst, 0, NULL, NULL );
|
||||
if (err != CL_SUCCESS)
|
||||
{
|
||||
printf("Error: Failed to read output array! %d\n", err);
|
||||
exit(1);
|
||||
}
|
||||
clReleaseMemObject(_dst);
|
||||
clReleaseMemObject(_src);
|
||||
}
|
@ -1,39 +0,0 @@
|
||||
// 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/
|
||||
|
||||
#ifndef _XFB_CONVERT
|
||||
#define _XFB_CONVERT
|
||||
|
||||
#include "Common.h"
|
||||
|
||||
// This must be called once before calling the two conversion functions
|
||||
// below.
|
||||
void InitXFBConvTables();
|
||||
|
||||
// These implementations could likely be made considerably faster by
|
||||
// reducing precision so that intermediate calculations are done in
|
||||
// 15-bit precision instead of 32-bit. However, this would complicate
|
||||
// the code and since we have a GPU implementation too, there really
|
||||
// isn't much point.
|
||||
|
||||
// Converts 4:2:2 YUV (YUYV) data to 32-bit RGBA data.
|
||||
void ConvertFromXFB(u32 *dst, const u8* _pXFB, int width, int height);
|
||||
|
||||
// Converts 32-bit RGBA data to 4:2:2 YUV (YUYV) data.
|
||||
void ConvertToXFB(u32 *dst, const u8* _pEFB, int width, int height);
|
||||
|
||||
#endif
|
Reference in New Issue
Block a user