ShuriZma
/
dolphin
mirror of https://github.com/dolphin-emu/dolphin.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #3193 from Tilka/swizzle 

VertexLoaderX64: optimize 4444 color conversion
This commit is contained in:
flacs 2015-10-23 07:51:40 +02:00
parent 5228758383 33784456a5
commit ece2e91446
3 changed files with 12 additions and 22 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
Source/Core/VideoCommon/VertexLoaderBase.cpp
View File

@ -150,9 +150,9 @@ public:
else
{
ERROR_LOG(VIDEO, "Can't compare vertex loaders that expect different vertex formats!");
ERROR_LOG(VIDEO, "a: m_VertexSize %d, m_native_components 0x%08x, stride %d\n",
ERROR_LOG(VIDEO, "a: m_VertexSize %d, m_native_components 0x%08x, stride %d",
a->m_VertexSize, a->m_native_components, a->m_native_vtx_decl.stride);
ERROR_LOG(VIDEO, "b: m_VertexSize %d, m_native_components 0x%08x, stride %d\n",
ERROR_LOG(VIDEO, "b: m_VertexSize %d, m_native_components 0x%08x, stride %d",
b->m_VertexSize, b->m_native_components, b->m_native_vtx_decl.stride);
}
}

2
Source/Core/VideoCommon/VertexLoaderManager.cpp
View File

@ -71,7 +71,7 @@ void UpdateVertexArrayPointers()
for (int i = 0; i < 12; i++)
{
// Only update the array base if the vertex description states we are going to use it.
if (g_main_cp_state.vtx_desc.GetVertexArrayStatus(i) >= 0x2)
if (g_main_cp_state.vtx_desc.GetVertexArrayStatus(i) & MASK_INDEXED)
cached_arraybases[i] = Memory::GetPointer(g_main_cp_state.array_bases[i]);
}

28
Source/Core/VideoCommon/VertexLoaderX64.cpp
View File

@ -314,34 +314,24 @@ void VertexLoaderX64::ReadColor(OpArg data, u64 attribute, int format)
LoadAndSwap(16, scratch1, data);
if (cpu_info.bBMI2)
{
MOV(32, R(scratch3), Imm32(0x0F0F0F0F));
PDEP(32, scratch2, scratch1, R(scratch3));
MOV(32, R(scratch3), Imm32(0xF0F0F0F0));
PDEP(32, scratch1, scratch1, R(scratch3));
MOV(32, R(scratch2), Imm32(0x0F0F0F0F));
PDEP(32, scratch1, scratch1, R(scratch2));
}
else
{
MOV(32, R(scratch3), R(scratch1));
SHL(32, R(scratch1), Imm8(12));
AND(32, R(scratch1), Imm32(0x0F000000));
MOV(32, R(scratch2), R(scratch1));
MOV(32, R(scratch1), R(scratch3));
SHL(32, R(scratch1), Imm8(8));
AND(32, R(scratch1), Imm32(0x000F0000));
OR(32, R(scratch2), R(scratch1));
OR(32, R(scratch1), R(scratch2));
AND(32, R(scratch1), Imm32(0x00FF00FF));
MOV(32, R(scratch1), R(scratch3));
MOV(32, R(scratch2), R(scratch1));
SHL(32, R(scratch1), Imm8(4));
AND(32, R(scratch1), Imm32(0x00000F00));
OR(32, R(scratch2), R(scratch1));
OR(32, R(scratch1), R(scratch2));
AND(32, R(scratch1), Imm32(0x0F0F0F0F));
AND(32, R(scratch3), Imm8(0x0F));
OR(32, R(scratch2), R(scratch3));
MOV(32, R(scratch1), R(scratch2));
SHL(32, R(scratch1), Imm8(4));
}
MOV(32, R(scratch2), R(scratch1));
SHL(32, R(scratch1), Imm8(4));
OR(32, R(scratch1), R(scratch2));
SwapAndStore(32, MDisp(dst_reg, m_dst_ofs), scratch1);
load_bytes = 2;