CTCaer/RetroArch
1
0
Fork 0
mirror of https://github.com/CTCaer/RetroArch.git synced 2025-01-25 04:35:12 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

Scaler: Fix SSE2 path for argb/bgra -> bgr24.

Browse Source 
Somehow a epi32 was added, but apparently this was never caught for some
reason. Fixes non-GPU screenshots for 32-bit color cores.
This commit is contained in:
Themaister 2019-05-13 20:09:42 +02:00 committed by bparker06
parent ff410f515e
commit a672cc13a9
1 changed files with 31 additions and 18 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

49
libretro-common/gfx/scaler/pixconv.c
View File

@ -480,7 +480,7 @@ void conv_rgba4444_rgb565(void *output_, const void *input_,
#if defined(__SSE2__)
/* :( TODO: Make this saner. */
static INLINE void store_bgr24_sse2(void *output, __m128i a,
__m128i b, __m128i c, __m128i *d)
__m128i b, __m128i c, __m128i d)
{
const __m128i mask_0 = _mm_set_epi32(0, 0, 0, 0x00ffffff);
const __m128i mask_1 = _mm_set_epi32(0, 0, 0x00ffffff, 0);
@ -503,10 +503,10 @@ static INLINE void store_bgr24_sse2(void *output, __m128i a,
__m128i c0 = _mm_srli_si128(_mm_and_si128(c, mask_2), 10);
__m128i c1 = _mm_srli_si128(_mm_and_si128(c, mask_3), 11);
__m128i c2 = _mm_slli_si128(_mm_and_si128(*d, mask_0), 4);
__m128i c3 = _mm_slli_si128(_mm_and_si128(*d, mask_1), 3);
__m128i c4 = _mm_slli_si128(_mm_and_si128(*d, mask_2), 2);
__m128i c5 = _mm_slli_si128(_mm_and_si128(*d, mask_3), 1);
__m128i c2 = _mm_slli_si128(_mm_and_si128(d, mask_0), 4);
__m128i c3 = _mm_slli_si128(_mm_and_si128(d, mask_1), 3);
__m128i c4 = _mm_slli_si128(_mm_and_si128(d, mask_2), 2);
__m128i c5 = _mm_slli_si128(_mm_and_si128(d, mask_3), 1);
__m128i *out = (__m128i*)output;
@ -589,7 +589,7 @@ void conv_0rgb1555_bgr24(void *output_, const void *input_,
_mm_slli_si128(res_hi_ra1, 2));
/* Non-POT pixel sizes for the loss */
store_bgr24_sse2(out, res_lo0, res_hi0, res_lo1, &res_hi1);
store_bgr24_sse2(out, res_lo0, res_hi0, res_lo1, res_hi1);
}
#endif
@ -674,7 +674,7 @@ void conv_rgb565_bgr24(void *output_, const void *input_,
res_hi1 = _mm_or_si128(res_hi_bg1,
_mm_slli_si128(res_hi_ra1, 2));
store_bgr24_sse2(out, res_lo0, res_hi0, res_lo1, &res_hi1);
store_bgr24_sse2(out, res_lo0, res_hi0, res_lo1, res_hi1);
}
#endif
@ -785,11 +785,7 @@ void conv_argb8888_bgr24(void *output_, const void *input_,
__m128i l1 = _mm_loadu_si128((const __m128i*)(input + w + 4));
__m128i l2 = _mm_loadu_si128((const __m128i*)(input + w + 8));
__m128i l3 = _mm_loadu_si128((const __m128i*)(input + w + 12));
l0 = _mm_shuffle_epi32(l0, _MM_SHUFFLE(3, 0, 1, 2));
l1 = _mm_shuffle_epi32(l1, _MM_SHUFFLE(3, 0, 1, 2));
l2 = _mm_shuffle_epi32(l2, _MM_SHUFFLE(3, 0, 1, 2));
l3 = _mm_shuffle_epi32(l3, _MM_SHUFFLE(3, 0, 1, 2));
store_bgr24_sse2(out, l0, l1, l2, &l3);
store_bgr24_sse2(out, l0, l1, l2, l3);
}
#endif
@ -803,6 +799,21 @@ void conv_argb8888_bgr24(void *output_, const void *input_,
}
}
#if defined(__SSE2__)
static INLINE __m128i conv_shuffle_rb_epi32(__m128i c)
{
/* SSSE3 plz */
const __m128i b_mask = _mm_set1_epi32(0x000000ff);
const __m128i g_mask = _mm_set1_epi32(0x0000ff00);
const __m128i r_mask = _mm_set1_epi32(0x00ff0000);
__m128i sl = _mm_and_si128(_mm_slli_epi32(c, 16), r_mask);
__m128i sr = _mm_and_si128(_mm_srli_epi32(c, 16), b_mask);
__m128i g = _mm_and_si128(c, g_mask);
__m128i rb = _mm_or_si128(sl, sr);
return _mm_or_si128(g, rb);
}
#endif
void conv_abgr8888_bgr24(void *output_, const void *input_,
int width, int height,
int out_stride, int in_stride)
@ -823,12 +834,15 @@ void conv_abgr8888_bgr24(void *output_, const void *input_,
#if defined(__SSE2__)
for (; w < max_width; w += 16, out += 48)
{
__m128i a = _mm_loadu_si128((const __m128i*)(input + w + 0));
__m128i b = _mm_loadu_si128((const __m128i*)(input + w + 4));
__m128i c = _mm_loadu_si128((const __m128i*)(input + w + 8));
__m128i d = _mm_loadu_si128((const __m128i*)(input + w + 12));
store_bgr24_sse2(out,
_mm_loadu_si128((const __m128i*)(input + w + 0)),
_mm_loadu_si128((const __m128i*)(input + w + 4)),
_mm_loadu_si128((const __m128i*)(input + w + 8)), &d
);
a = conv_shuffle_rb_epi32(a);
b = conv_shuffle_rb_epi32(b);
c = conv_shuffle_rb_epi32(c);
d = conv_shuffle_rb_epi32(d);
store_bgr24_sse2(out, a, b, c, d);
}
#endif
@ -842,7 +856,6 @@ void conv_abgr8888_bgr24(void *output_, const void *input_,
}
}
void conv_argb8888_abgr8888(void *output_, const void *input_,
int width, int height,
int out_stride, int in_stride)