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

Simplify the SSSE3 code for better understanding. 

The end result is unchanged.

git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6815 8ced0084-cf51-0410-be5f-012b33b47a6e
This commit is contained in:
xsacha 2011-01-11 07:25:36 +00:00
parent 9c09500a13
commit 299f163d38
1 changed files with 5 additions and 28 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

33
Source/Core/VideoCommon/Src/DataReader.h
View File

@ -74,40 +74,17 @@ typedef void (*DataReadU32xNfunc)(u32 *buf);
extern DataReadU32xNfunc DataReadU32xFuncs[16]; extern DataReadU32xNfunc DataReadU32xFuncs[16];
#if _M_SSE >= 0x301 #if _M_SSE >= 0x301
const __m128i bs_mask = _mm_set_epi32(0x0C0D0E0FL, 0x08090A0BL, 0x04050607L, 0x00010203L);
const __m128i mask1 = _mm_set_epi8(15,14,13,12,11,10,9,8,7,6,5,4,0,1,2,3);
const __m128i mask2 = _mm_set_epi8(15,14,13,12,11,10,9,8,4,5,6,7,0,1,2,3);
const __m128i mask3 = _mm_set_epi8(15,14,13,12,8,9,10,11,4,5,6,7,0,1,2,3);
const __m128i mask4 = _mm_set_epi8(12,13,14,15,8,9,10,11,4,5,6,7,0,1,2,3);
template<unsigned int N> template<unsigned int N>
void DataReadU32xN_SSSE3(u32 *bufx16) void DataReadU32xN_SSSE3(u32 *bufx16)
{ {
memcpy(bufx16, g_pVideoData, sizeof(u32) * N); memcpy(bufx16, g_pVideoData, sizeof(u32) * N);
__m128i* buf = (__m128i *)bufx16; __m128i* buf = (__m128i *)bufx16;
switch(N) if (N>12) { _mm_store_si128(buf, _mm_shuffle_epi8(_mm_load_si128(buf), bs_mask)); buf++; }
{ if (N>8) { _mm_store_si128(buf, _mm_shuffle_epi8(_mm_load_si128(buf), bs_mask)); buf++; }
case 13: _mm_store_si128(buf, _mm_shuffle_epi8(_mm_load_si128(buf), mask4)); buf++; if (N>4) { _mm_store_si128(buf, _mm_shuffle_epi8(_mm_load_si128(buf), bs_mask)); buf++; }
case 9: _mm_store_si128(buf, _mm_shuffle_epi8(_mm_load_si128(buf), mask4)); buf++; _mm_store_si128(buf, _mm_shuffle_epi8(_mm_load_si128(buf), bs_mask));
case 5: _mm_store_si128(buf, _mm_shuffle_epi8(_mm_load_si128(buf), mask4)); buf++;
case 1: _mm_store_si128(buf, _mm_shuffle_epi8(_mm_load_si128(buf), mask1));
break;
case 14: _mm_store_si128(buf, _mm_shuffle_epi8(_mm_load_si128(buf), mask4)); buf++;
case 10: _mm_store_si128(buf, _mm_shuffle_epi8(_mm_load_si128(buf), mask4)); buf++;
case 6: _mm_store_si128(buf, _mm_shuffle_epi8(_mm_load_si128(buf), mask4)); buf++;
case 2: _mm_store_si128(buf, _mm_shuffle_epi8(_mm_load_si128(buf), mask2));
break;
case 15: _mm_store_si128(buf, _mm_shuffle_epi8(_mm_load_si128(buf), mask4)); buf++;
case 11: _mm_store_si128(buf, _mm_shuffle_epi8(_mm_load_si128(buf), mask4)); buf++;
case 7: _mm_store_si128(buf, _mm_shuffle_epi8(_mm_load_si128(buf), mask4)); buf++;
case 3: _mm_store_si128(buf, _mm_shuffle_epi8(_mm_load_si128(buf), mask3));
break;
case 16: _mm_store_si128(buf, _mm_shuffle_epi8(_mm_load_si128(buf), mask4)); buf++;
case 12: _mm_store_si128(buf, _mm_shuffle_epi8(_mm_load_si128(buf), mask4)); buf++;
case 8: _mm_store_si128(buf, _mm_shuffle_epi8(_mm_load_si128(buf), mask4)); buf++;
case 4: _mm_store_si128(buf, _mm_shuffle_epi8(_mm_load_si128(buf), mask4));
break;
}
g_pVideoData += (sizeof(u32) * N); g_pVideoData += (sizeof(u32) * N);
} }