archive/dolphin
1
0
Fork 0
mirror of https://github.com/dolphin-emu/dolphin.git synced 2024-09-20 19:31:53 +02:00
Code Issues Projects Releases Packages Wiki Activity

Fix for bit reduction regression in GX_TF_RGB565 textures from previous commit.

Browse source 
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@6726 8ced0084-cf51-0410-be5f-012b33b47a6e
This commit is contained in:
james.jdunne 2011-01-03 01:24:35 +00:00
parent ecf92f5c3c
commit b7f7a248c5
1 changed files with 12 additions and 13 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

25
Source/Core/VideoCommon/Src/TextureDecoder.cpp
View file

@ -1311,27 +1311,26 @@ PC_TexFormat TexDecoder_Decode_RGBA(u32 * dst, const u8 * src, int width, int he
// where hg, fe, ba, and dc are 16-bit colors in big-endian order
const __m128i rgb565x4 = _mm_loadl_epi64(dxtsrc);
// The big-endian 16-bit colors `ba` and `dc` look like 0b_RRRrrGGG_gggBBBbb in a little endian xmm register
// Unpack `hgfe dcba` to `hhgg ffee ddcc bbaa`, where each 32-bit word is now 0b_RRRrrGGG_gggBBBbb_RRRrrGGG_gggBBBbb
const __m128i c0 = _mm_unpacklo_epi8(rgb565x4, rgb565x4);
// The big-endian 16-bit colors `ba` and `dc` look like 0b_gggBBBbb_RRRrrGGg in a little endian xmm register
// Unpack `hgfe dcba` to `hhgg ffee ddcc bbaa`, where each 32-bit word is now 0b_gggBBBbb_RRRrrGGg_gggBBBbb_RRRrrGGg
const __m128i c0 = _mm_unpacklo_epi16(rgb565x4, rgb565x4);
// swizzle 0b_RRRrrGGg_gggBBBbb_RRRrrGGg_gggBBBbb
// swizzle 0b_gggBBBbb_RRRrrGGg_gggBBBbb_RRRrrGGg
// to 0b_11111111_BBBbbBBB_GGggggGG_RRRrrRRR
// 0b_RRRrrGGg_gggBBBbb_RRRrrGGg_gggBBBbb >> 8 [32] =
// 0b_00000000_RRRrrGGg_gggBBBbb_RRRrrGGg &
// 0b_gggBBBbb_RRRrrGGg_gggBBBbb_RRRrrGGg &
// 0b_00000000_00000000_00000000_11111000 =
// 0b_00000000_00000000_00000000_RRRrr000
const __m128i r0 = _mm_and_si128(_mm_srli_epi32(c0, 8), kMaskR0);
const __m128i r0 = _mm_and_si128(c0, kMaskR0);
// 0b_00000000_00000000_00000000_RRRrr000 >> 5 [32] =
// 0b_00000000_00000000_00000000_00000RRR
const __m128i r1 = _mm_srli_epi32(r0, 5);
// 0b_RRRrrGGg_gggBBBbb_RRRrrGGg_gggBBBbb << 5 [32] =
// 0b_GGggggBB_BbbRRRrr_GGggggBB_Bbb00000 &
// 0b_gggBBBbb_RRRrrGGg_gggBBBbb_RRRrrGGg >> 3 [32] =
// 0b_000gggBB_BbbRRRrr_GGggggBB_BbbRRRrr &
// 0b_00000000_00000000_11111100_00000000 =
// 0b_00000000_00000000_GGgggg00_00000000
const __m128i gtmp = _mm_slli_epi32(c0, 5);
const __m128i gtmp = _mm_srli_epi32(c0, 3);
const __m128i g0 = _mm_and_si128(gtmp, kMaskG0);
// 0b_GGggggBB_BbbRRRrr_GGggggBB_Bbb00000 >> 6 [32] =
// 0b_000000GG_ggggBBBb_bRRRrrGG_ggggBBBb &
@ -1339,11 +1338,11 @@ PC_TexFormat TexDecoder_Decode_RGBA(u32 * dst, const u8 * src, int width, int he
// 0b_00000000_00000000_000000GG_00000000 =
const __m128i g1 = _mm_and_si128(_mm_srli_epi32(gtmp, 6), kMaskG1);
// 0b_RRRrrGGg_gggBBBbb_RRRrrGGg_gggBBBbb << 3 [32] =
// 0b_rrGGgggg_BBBbbRRR_rrGGgggg_BBBbb000 &
// 0b_gggBBBbb_RRRrrGGg_gggBBBbb_RRRrrGGg >> 5 [32] =
// 0b_00000ggg_BBBbbRRR_rrGGgggg_BBBbbRRR &
// 0b_00000000_11111000_00000000_00000000 =
// 0b_00000000_BBBbb000_00000000_00000000
const __m128i b0 = _mm_and_si128(_mm_slli_epi32(c0, 3), kMaskB0);
const __m128i b0 = _mm_and_si128(_mm_srli_epi32(c0, 5), kMaskB0);
// 0b_00000000_BBBbb000_00000000_00000000 >> 5 [16] =
// 0b_00000000_00000BBB_00000000_00000000
const __m128i b1 = _mm_srli_epi16(b0, 5);