cedric pushed a commit to branch master.
http://git.enlightenment.org/core/efl.git/commit/?id=657d495aa9607774409206099d809a1a737b736a
commit 657d495aa9607774409206099d809a1a737b736a
Author: Cedric BAIL <ced...@osg.samsung.com>
Date: Wed Apr 22 15:24:13 2015 +0200
evas: implement _op_blend_p_mas_dp_neon and _op_blend_pas_mas_dp_neon in
NEON intrinsics.
Reviewers: raster, cedric
Reviewed By: cedric
Subscribers: cedric
Projects: #efl
Differential Revision: https://phab.enlightenment.org/D2391
Signed-off-by: Cedric BAIL <ced...@osg.samsung.com>
---
.../evas_op_blend/op_blend_pixel_mask_neon.c | 384 ++++++++++++++++++---
1 file changed, 337 insertions(+), 47 deletions(-)
diff --git a/src/lib/evas/common/evas_op_blend/op_blend_pixel_mask_neon.c
b/src/lib/evas/common/evas_op_blend/op_blend_pixel_mask_neon.c
index 0c1029b..317e7d5 100644
--- a/src/lib/evas/common/evas_op_blend/op_blend_pixel_mask_neon.c
+++ b/src/lib/evas/common/evas_op_blend/op_blend_pixel_mask_neon.c
@@ -1,62 +1,352 @@
+#ifdef BUILD_NEON
+#include <arm_neon.h>
+#endif
/* blend pixel x mask --> dst */
-// FIXME: These functions most likely don't perform the correct operation.
-// Test them with masks and images.
-
#ifdef BUILD_NEON
-#if 0
static void
_op_blend_pas_mas_dp_neon(DATA32 *s, DATA8 *m, DATA32 c EINA_UNUSED, DATA32
*d, int l) {
- DATA32 *e;
+ uint16x8_t m_16x8;
+ uint16x8_t ms0_16x8;
+ uint16x8_t ms1_16x8;
+ uint16x8_t temp0_16x8;
+ uint16x8_t temp1_16x8;
+ uint16x8_t x255_16x8;
+ uint32_t m_32;
+ uint32x2_t m_32x2;
+ uint32x4_t a_32x4;
+ uint32x4_t ad_32x4;
+ uint32x4_t cond_32x4;
+ uint32x4_t d_32x4;
+ uint32x4_t m_32x4;
+ uint32x4_t ms_32x4;
+ uint32x4_t s_32x4;
+ uint32x4_t temp_32x4;
+ uint32x4_t x0_32x4;
+ uint32x4_t x1_32x4;
+ uint8x16_t a_8x16;
+ uint8x16_t d_8x16;
+ uint8x16_t m_8x16;
+ uint8x16_t ms_8x16;
+ uint8x16_t s_8x16;
+ uint8x16_t temp_8x16;
+ uint8x16_t x0_8x16;
+ uint8x16_t x1_8x16;
+ uint8x8_t a0_8x8;
+ uint8x8_t a1_8x8;
+ uint8x8_t d0_8x8;
+ uint8x8_t d1_8x8;
+ uint8x8_t m0_8x8;
+ uint8x8_t m1_8x8;
+ uint8x8_t m_8x8;
+ uint8x8_t ms0_8x8;
+ uint8x8_t ms1_8x8;
+ uint8x8_t s0_8x8;
+ uint8x8_t s1_8x8;
+ uint8x8_t temp0_8x8;
+ uint8x8_t temp1_8x8;
+
+ x1_8x16 = vdupq_n_u8(0x1);
+ x1_32x4 = vreinterpretq_u32_u8(x1_8x16);
+ x255_16x8 = vdupq_n_u16(0xff);
+ x0_8x16 = vdupq_n_u8(0x0);
+ x0_32x4 = vreinterpretq_u32_u8(x0_8x16);
+
+ DATA32 *end = d + (l & ~3);
+ while (d < end)
+ {
+ unsigned int k = *((unsigned int *)m);
+ // shortcut if *m==0
+ if (k == 0)
+ {
+ m+=4;
+ d+=4;
+ s+=4;
+ continue;
+ }
+ // shortcut if *m==0xff
+ if (~k == 0)
+ {
+ // load 4 elements from s
+ s_32x4 = vld1q_u32(s);
+ s_8x16 = vreinterpretq_u8_u32(s_32x4);
+
+ // load 4 elements from d
+ d_32x4 = vld1q_u32(d);
+ d_8x16 = vreinterpretq_u8_u32(d_32x4);
+ d0_8x8 = vget_low_u8(d_8x16);
+ d1_8x8 = vget_high_u8(d_8x16);
+
+ // substract 256 - *s
+ a_8x16 = vsubq_u8(x0_8x16, s_8x16);
+ a_32x4 = vreinterpretq_u32_u8(a_8x16);
+
+ // shift alpha>>24 and place it into every 8bit element
+ a_32x4 = vshrq_n_u32(a_32x4, 24);
+ a_32x4 = vmulq_u32(a_32x4, x1_32x4);
+ a_8x16 = vreinterpretq_u8_u32(a_32x4);
+ a0_8x8 = vget_low_u8(a_8x16);
+ a1_8x8 = vget_high_u8(a_8x16);
+
+ // multiply MUL_256(a, *d)
+ temp0_16x8 = vmull_u8(a0_8x8, d0_8x8);
+ temp1_16x8 = vmull_u8(a1_8x8, d1_8x8);
+ temp0_8x8 = vshrn_n_u16(temp0_16x8,8);
+ temp1_8x8 = vshrn_n_u16(temp1_16x8,8);
+ temp_8x16 = vcombine_u8(temp0_8x8, temp1_8x8);
+ temp_32x4 = vreinterpretq_u32_u8(temp_8x16);
+
+ // if alpha is 0, replace a*d with d
+ cond_32x4 = vceqq_u32(a_32x4, x0_32x4);
+ ad_32x4 = vbslq_u32(cond_32x4, d_32x4, temp_32x4);
+
+ // add *s
+ d_32x4 = vaddq_u32(s_32x4, ad_32x4);
+
+ // save result
+ vst1q_u32(d, d_32x4);
+ m+=4;
+ d+=4;
+ s+=4;
+ continue;
+ }
+ // load 4 elements from m
+ m_32 = k;
+ m_32x2 = vset_lane_u32(m_32, m_32x2, 0);
+
+ // load 4 elements from s
+ s_32x4 = vld1q_u32(s);
+ s_8x16 = vreinterpretq_u8_u32(s_32x4);
+ s0_8x8 = vget_low_u8(s_8x16);
+ s1_8x8 = vget_high_u8(s_8x16);
+
+ // load 4 elements from d
+ d_32x4 = vld1q_u32(d);
+ d_8x16 = vreinterpretq_u8_u32(d_32x4);
+ d0_8x8 = vget_low_u8(d_8x16);
+ d1_8x8 = vget_high_u8(d_8x16);
+
+ // make m 32 bit wide
+ m_8x8 = vreinterpret_u8_u32(m_32x2);
+ m_16x8 = vmovl_u8(m_8x8);
+ m_8x16 = vreinterpretq_u8_u16(m_16x8);
+ m_8x8 = vget_low_u8(m_8x16);
+ m_16x8 = vmovl_u8(m_8x8);
+ m_32x4 = vreinterpretq_u32_u16(m_16x8);
+
+ // place m into every 8 bit element of vector
+ m_32x4 = vmulq_u32(m_32x4, x1_32x4);
+ m_8x16 = vreinterpretq_u8_u32(m_32x4);
+ m0_8x8 = vget_low_u8(m_8x16);
+ m1_8x8 = vget_high_u8(m_8x16);
+
+ // multiply MUL_SYM(m, *s);
+ ms0_16x8 = vmull_u8(m0_8x8, s0_8x8);
+ ms1_16x8 = vmull_u8(m1_8x8, s1_8x8);
+ ms0_16x8 = vaddq_u16(ms0_16x8, x255_16x8);
+ ms1_16x8 = vaddq_u16(ms1_16x8, x255_16x8);
+ ms0_8x8 = vshrn_n_u16(ms0_16x8, 8);
+ ms1_8x8 = vshrn_n_u16(ms1_16x8, 8);
+ ms_8x16 = vcombine_u8(ms0_8x8, ms1_8x8);
+
+ // substract 256 - m*s
+ a_8x16 = vsubq_u8(x0_8x16, ms_8x16);
+ a_32x4 = vreinterpretq_u32_u8(a_8x16);
+
+ // shift alpha>>24 and place it into every 8bit element
+ a_32x4 = vshrq_n_u32(a_32x4, 24);
+ a_32x4 = vmulq_u32(a_32x4, x1_32x4);
+ a_8x16 = vreinterpretq_u8_u32(a_32x4);
+ a0_8x8 = vget_low_u8(a_8x16);
+ a1_8x8 = vget_high_u8(a_8x16);
+
+ // multiply MUL_256(a, *d)
+ temp0_16x8 = vmull_u8(a0_8x8, d0_8x8);
+ temp1_16x8 = vmull_u8(a1_8x8, d1_8x8);
+ temp0_8x8 = vshrn_n_u16(temp0_16x8,8);
+ temp1_8x8 = vshrn_n_u16(temp1_16x8,8);
+ temp_8x16 = vcombine_u8(temp0_8x8, temp1_8x8);
+ temp_32x4 = vreinterpretq_u32_u8(temp_8x16);
+
+ // if alpha is 0, replace a*d with d
+ cond_32x4 = vceqq_u32(a_32x4, x0_32x4);
+ ad_32x4 = vbslq_u32(cond_32x4, d_32x4, temp_32x4);
+
+ // add m*s
+ ms_32x4 = vreinterpretq_u32_u8(ms_8x16);
+ d_32x4 = vaddq_u32(ms_32x4, ad_32x4);
+
+ // save result
+ vst1q_u32(d, d_32x4);
+
+ d+=4;
+ s+=4;
+ m+=4;
+ }
+
int alpha;
- UNROLL8_PLD_WHILE(d, l, e,
- {
- alpha = *m;
- switch(alpha)
- {
- case 0:
- break;
- case 255:
- *d = *s;
- break;
- default:
- alpha++;
- *d = INTERP_256(alpha, *s, *d);
- break;
- }
- m++; s++; d++;
- });
+ DATA32 temp;
+
+ end += (l & 3);
+ while (d < end)
+ {
+ alpha = *m;
+ switch(alpha)
+ {
+ case 0:
+ break;
+ case 255:
+ alpha = 256 - (*s >> 24);
+ *d = *s + MUL_256(alpha, *d);
+ break;
+ default:
+ temp = MUL_SYM(alpha, *s);
+ alpha = 256 - (temp >> 24);
+ *d = temp + MUL_256(alpha, *d);
+ break;
+ }
+ m++; s++; d++;
+ }
}
static void
-_op_blend_pan_mas_dp_neon(DATA32 *s, DATA8 *m, DATA32 c EINA_UNUSED, DATA32
*d, int l) {
- DATA32 *e;
+_op_blend_p_mas_dp_neon(DATA32 *s, DATA8 *m, DATA32 c EINA_UNUSED, DATA32 *d,
int l) {
+ uint16x8_t m_16x8;
+ uint16x8_t ms0_16x8;
+ uint16x8_t ms1_16x8;
+ uint16x8_t temp0_16x8;
+ uint16x8_t temp1_16x8;
+ uint16x8_t x255_16x8;
+ uint32x2_t m_32x2;
+ uint32x4_t a_32x4;
+ uint32x4_t ad_32x4;
+ uint32x4_t cond_32x4;
+ uint32x4_t d_32x4;
+ uint32x4_t m_32x4;
+ uint32x4_t ms_32x4;
+ uint32x4_t s_32x4;
+ uint32x4_t temp_32x4;
+ uint32x4_t x0_32x4;
+ uint32x4_t x1_32x4;
+ uint8x16_t a_8x16;
+ uint8x16_t d_8x16;
+ uint8x16_t m_8x16;
+ uint8x16_t ms_8x16;
+ uint8x16_t s_8x16;
+ uint8x16_t temp_8x16;
+ uint8x16_t x0_8x16;
+ uint8x16_t x1_8x16;
+ uint8x8_t a0_8x8;
+ uint8x8_t a1_8x8;
+ uint8x8_t d0_8x8;
+ uint8x8_t d1_8x8;
+ uint8x8_t m0_8x8;
+ uint8x8_t m1_8x8;
+ uint8x8_t m_8x8;
+ uint8x8_t ms0_8x8;
+ uint8x8_t ms1_8x8;
+ uint8x8_t s0_8x8;
+ uint8x8_t s1_8x8;
+ uint8x8_t temp0_8x8;
+ uint8x8_t temp1_8x8;
+
+ x1_8x16 = vdupq_n_u8(0x1);
+ x1_32x4 = vreinterpretq_u32_u8(x1_8x16);
+ x255_16x8 = vdupq_n_u16(0xff);
+ x0_8x16 = vdupq_n_u8(0x0);
+ x0_32x4 = vreinterpretq_u32_u8(x0_8x16);
+
+ DATA32 *end = d + (l & ~3);
+ while (d < end)
+ {
+ // load 4 elements from m
+ m_32x2 = vld1_lane_u32((DATA32*)m, m_32x2, 0);
+
+ // load 4 elements from s
+ s_32x4 = vld1q_u32(s);
+ s_8x16 = vreinterpretq_u8_u32(s_32x4);
+ s0_8x8 = vget_low_u8(s_8x16);
+ s1_8x8 = vget_high_u8(s_8x16);
+
+ // load 4 elements from d
+ d_32x4 = vld1q_u32(d);
+ d_8x16 = vreinterpretq_u8_u32(d_32x4);
+ d0_8x8 = vget_low_u8(d_8x16);
+ d1_8x8 = vget_high_u8(d_8x16);
+
+ // make m 32 bit wide
+ m_8x8 = vreinterpret_u8_u32(m_32x2);
+ m_16x8 = vmovl_u8(m_8x8);
+ m_8x16 = vreinterpretq_u8_u16(m_16x8);
+ m_8x8 = vget_low_u8(m_8x16);
+ m_16x8 = vmovl_u8(m_8x8);
+ m_32x4 = vreinterpretq_u32_u16(m_16x8);
+
+ // place m into every 8 bit element of vector
+ m_32x4 = vmulq_u32(m_32x4, x1_32x4);
+ m_8x16 = vreinterpretq_u8_u32(m_32x4);
+ m0_8x8 = vget_low_u8(m_8x16);
+ m1_8x8 = vget_high_u8(m_8x16);
+
+ // multiply MUL_SYM(m, *s);
+ ms0_16x8 = vmull_u8(m0_8x8, s0_8x8);
+ ms1_16x8 = vmull_u8(m1_8x8, s1_8x8);
+ ms0_16x8 = vaddq_u16(ms0_16x8, x255_16x8);
+ ms1_16x8 = vaddq_u16(ms1_16x8, x255_16x8);
+ ms0_8x8 = vshrn_n_u16(ms0_16x8, 8);
+ ms1_8x8 = vshrn_n_u16(ms1_16x8, 8);
+ ms_8x16 = vcombine_u8(ms0_8x8, ms1_8x8);
+
+ // substract 256 - m*s
+ a_8x16 = vsubq_u8(x0_8x16, ms_8x16);
+ a_32x4 = vreinterpretq_u32_u8(a_8x16);
+
+ // shift alpha>>24 and place it into every 8bit element
+ a_32x4 = vshrq_n_u32(a_32x4, 24);
+ a_32x4 = vmulq_u32(a_32x4, x1_32x4);
+ a_8x16 = vreinterpretq_u8_u32(a_32x4);
+ a0_8x8 = vget_low_u8(a_8x16);
+ a1_8x8 = vget_high_u8(a_8x16);
+
+ // multiply MUL_256(a, *d)
+ temp0_16x8 = vmull_u8(a0_8x8, d0_8x8);
+ temp1_16x8 = vmull_u8(a1_8x8, d1_8x8);
+ temp0_8x8 = vshrn_n_u16(temp0_16x8,8);
+ temp1_8x8 = vshrn_n_u16(temp1_16x8,8);
+ temp_8x16 = vcombine_u8(temp0_8x8, temp1_8x8);
+ temp_32x4 = vreinterpretq_u32_u8(temp_8x16);
+
+ // if alpha is 0, replace a*d with d
+ cond_32x4 = vceqq_u32(a_32x4, x0_32x4);
+ ad_32x4 = vbslq_u32(cond_32x4, d_32x4, temp_32x4);
+
+ // add m*s
+ ms_32x4 = vreinterpretq_u32_u8(ms_8x16);
+ d_32x4 = vaddq_u32(ms_32x4, ad_32x4);
+
+ // save result
+ vst1q_u32(d, d_32x4);
+
+ d+=4;
+ s+=4;
+ m+=4;
+ }
+
int alpha;
- UNROLL8_PLD_WHILE(d, l, e,
- {
- alpha = *m;
- switch(alpha)
- {
- case 0:
- break;
- case 255:
- *d = *s;
- break;
- default:
- alpha++;
- *d = INTERP_256(alpha, *s, *d);
- break;
- }
- m++; s++; d++;
- });
+ DATA32 temp;
+
+ end += (l & 3);
+ while (d < end)
+ {
+ alpha = *m;
+ temp = MUL_SYM(alpha, *s);
+ alpha = 256 - (temp >> 24);
+ *d = temp + MUL_256(alpha, *d);
+ m++; s++; d++;
+ }
}
-#else
-// FIXME
-#define _op_blend_pas_mas_dp_neon NULL
-#define _op_blend_pan_mas_dp_neon NULL
-#endif
-#define _op_blend_p_mas_dp_neon _op_blend_pas_mas_dp_neon
+#define _op_blend_pan_mas_dp_neon _op_blend_pas_mas_dp_neon
#define _op_blend_p_mas_dpan_neon _op_blend_p_mas_dp_neon
#define _op_blend_pan_mas_dpan_neon _op_blend_pan_mas_dp_neon
--
