/hardware/intel/common/omx-components/videocodec/libvpx_internal/libvpx/vp8/common/arm/armv6/ |
D | vp8_variance_halfpixvar16x16_v_armv6.asm | 41 ldr r6, [r9, #0] ; load 4 src pixels from next row 45 mvn r6, r6 46 uhsub8 r4, r4, r6 49 usub8 r6, r4, r5 ; calculate difference 51 sel r7, r6, lr ; select bytes with positive difference 52 usub8 r6, r5, r4 ; calculate difference with reversed operands 54 sel r6, r6, lr ; select bytes with negative difference 58 usad8 r5, r6, lr ; calculate sum of negative differences 59 orr r6, r6, r7 ; differences of all 4 pixels 65 uxtb16 r5, r6 ; byte (two pixels) to halfwords [all …]
|
D | vp8_variance_halfpixvar16x16_h_armv6.asm | 40 ldr r6, [r0, #1] ; load 4 src pixels with 1 byte offset 44 mvn r6, r6 45 uhsub8 r4, r4, r6 48 usub8 r6, r4, r5 ; calculate difference 50 sel r7, r6, lr ; select bytes with positive difference 51 usub8 r6, r5, r4 ; calculate difference with reversed operands 53 sel r6, r6, lr ; select bytes with negative difference 57 usad8 r5, r6, lr ; calculate sum of negative differences 58 orr r6, r6, r7 ; differences of all 4 pixels 64 uxtb16 r5, r6 ; byte (two pixels) to halfwords [all …]
|
D | vp8_variance_halfpixvar16x16_hv_armv6.asm | 41 ldr r6, [r0, #1] ; load source pixels b, row N 46 mvn r6, r6 47 uhsub8 r4, r4, r6 59 usub8 r6, r4, r5 ; calculate difference 61 sel r7, r6, lr ; select bytes with positive difference 62 usub8 r6, r5, r4 ; calculate difference with reversed operands 64 sel r6, r6, lr ; select bytes with negative difference 68 usad8 r5, r6, lr ; calculate sum of negative differences 69 orr r6, r6, r7 ; differences of all 4 pixels 75 uxtb16 r5, r6 ; byte (two pixels) to halfwords [all …]
|
D | dequant_idct_v6.asm | 32 smulbb r6, r4, r5 38 strh r6, [r0], #2 41 smulbb r6, r4, r5 49 strh r6, [r0], #2 60 ldr r6, [r0, #8] 65 smulwt r9, r3, r6 66 smulwb r7, r3, r6 67 smulwt r10, r4, r6 68 smulwb r8, r4, r6 72 uadd16 r6, r6, r7 [all …]
|
D | idct_v6.asm | 38 ldr r6, [r0, #(4*2)] ; i5 | i4 42 smulbt r9, r5, r6 ; (ip[5] * cospi8sqrt2minus1) >> 16 43 smulbb r7, r5, r6 ; (ip[4] * cospi8sqrt2minus1) >> 16 44 smulwt r10, r4, r6 ; (ip[5] * sinpi8sqrt2) >> 16 45 smulwb r8, r4, r6 ; (ip[4] * sinpi8sqrt2) >> 16 50 uadd16 r6, r6, r7 ; 5c+5 | 4c+4 64 uadd16 r6, r6, r10 ; d 68 uadd16 r9, r10, r6 ; a+d 69 usub16 r10, r10, r6 ; a-d 70 uadd16 r6, r8, r7 ; b+c [all …]
|
D | loopfilter_v6.asm | 68 ldr r6, [sp, #36] ; load thresh address 77 ldrb r3, [r6] ; thresh 90 uqsub8 r6, r9, r10 ; p3 - p2 95 orr r6, r6, r7 ; abs (p3-p2) 97 uqsub8 lr, r6, r2 ; compare to limit. lr: vp8_filter_mask 99 uqsub8 r6, r11, r12 ; p1 - p0 104 orr r6, r6, r7 ; abs (p1-p0) 105 uqsub8 r7, r6, r2 ; compare to limit 106 uqsub8 r8, r6, r3 ; compare to thresh -- save r8 for later 109 uqsub8 r6, r11, r10 ; p1 - q1 [all …]
|
D | vp8_variance16x16_armv6.asm | 43 usub8 r6, r4, r5 ; calculate difference 45 sel r7, r6, lr ; select bytes with positive difference 48 sel r6, r9, lr ; select bytes with negative difference 52 usad8 r5, r6, lr ; calculate sum of negative differences 53 orr r6, r6, r7 ; differences of all 4 pixels 59 uxtb16 r5, r6 ; byte (two pixels) to halfwords 60 uxtb16 r10, r6, ror #8 ; another two pixels to halfwords 68 usub8 r6, r4, r5 ; calculate difference 69 sel r7, r6, lr ; select bytes with positive difference 71 sel r6, r9, lr ; select bytes with negative difference [all …]
|
D | dc_only_idct_add_v6.asm | 30 ldr r6, [r1], r2 37 uxtab16 r7, r0, r6 38 uxtab16 r6, r0, r6, ror #8 42 usat16 r6, #8, r6 44 orr r7, r7, r6, lsl #8 47 ldr r6, [r1] 52 uxtab16 r7, r0, r6 53 uxtab16 r6, r0, r6, ror #8 57 usat16 r6, #8, r6 59 orr r7, r7, r6, lsl #8
|
D | intra4x4_predict_v6.asm | 56 ldrb r6, [r1], r2 ; Left[2] 62 add r4, r4, r6 86 ldrb r6, [r1], r2 ; Left[2] 99 add r6, r6, r6, lsl #16 ; l[2|2] 116 sadd16 r1, r6, r10 ; l[2|2] + a[2|0] - [tl|tl] 117 sadd16 r2, r6, r11 ; l[2|2] + a[3|1] - [tl|tl] 180 ldrb r6, [r1], r2 ; Left[2] 185 add r10, r5, r6 ; l[1] + l[2] 186 add r11, r6, r7 ; l[2] + l[3] 193 add r6, r11, r7, lsl #1 ; l[2] + 2*l[3] + l[3] [all …]
|
D | bilinearfilter_v6.asm | 50 ldrb r6, [r0] ; load source data 59 pkhbt r6, r6, r7, lsl #16 ; src[1] | src[0] 62 smuad r6, r6, r5 ; apply the filter 73 add r6, r6, #0x40 ; round_shift_and_clamp 75 usat r6, #16, r6, asr #7 78 strh r6, [r1], r3 ; result is transposed and stored 88 ldrneb r6, [r0] ; load source data 114 ldrb r6, [r0] ; load data 119 strh r6, [r1], r3 ; store it to immediate buffer 162 ldr r6, [r0] ; load the data [all …]
|
/hardware/intel/common/omx-components/videocodec/libvpx_internal/libvpx/vpx_scale/arm/neon/ |
D | vp8_vpxyv12_extendframeborders_neon.asm | 35 add r6, r1, r3 ; dest_ptr2 = src_ptr2 + 1 (src_ptr1 + plane_width) 36 sub r2, r6, #1 ; src_ptr2 = src_ptr1 + plane_width - 1 63 vst1.8 {q2, q3}, [r6], lr 65 vst1.8 {q6, q7}, [r6], lr 67 vst1.8 {q10, q11}, [r6], lr 69 vst1.8 {q14, q15}, [r6], lr 81 …add r6, r1, r8 ; dest_ptr2 = src_ptr2 - plane_stride (src_ptr1 + (plane_h… 82 …sub r2, r6, lr ; src_ptr2 = src_ptr1 + (plane_height * plane_stride) - pl… 102 vst1.8 {q8, q9}, [r6]! 104 vst1.8 {q10, q11}, [r6]! [all …]
|
/hardware/intel/common/omx-components/videocodec/libvpx_internal/libvpx/vp9/common/arm/neon/ |
D | vp9_avg_neon.asm | 19 push {r4-r6, lr} 21 mov r6, r2 39 vld1.8 {q8-q9}, [r6@128]! 40 vld1.8 {q10-q11}, [r6@128], r4 49 pop {r4-r6, pc} 54 vld1.8 {q8-q9}, [r6@128], r3 55 vld1.8 {q10-q11}, [r6@128], r3 60 pld [r6] 62 pld [r6, r3] 68 pop {r4-r6, pc} [all …]
|
/hardware/intel/common/omx-components/videocodec/libvpx_internal/libvpx/vp8/encoder/arm/neon/ |
D | subtract_neon.asm | 32 ldr r6, [r0, #vp8_block_src_stride] 36 vld1.8 {d0}, [r3], r6 ;load src 38 vld1.8 {d2}, [r3], r6 40 vld1.8 {d4}, [r3], r6 42 vld1.8 {d6}, [r3], r6 70 mov r6, #32 ; "diff" stride x2 92 vst1.16 {q8}, [r0], r6 ;store diff 93 vst1.16 {q9}, [r5], r6 94 vst1.16 {q10}, [r0], r6 95 vst1.16 {q11}, [r5], r6 [all …]
|
/hardware/intel/common/omx-components/videocodec/libvpx_internal/libvpx/vp8/encoder/arm/armv6/ |
D | vp8_mse16x16_armv6.asm | 40 ldr r6, [r2, #0x0] ; load 4 ref pixels 44 usub8 r8, r5, r6 ; calculate difference 47 usub8 r9, r6, r5 ; calculate difference with reversed operands 53 usad8 r6, r8, lr ; calculate sum of negative differences 59 uxtb16 r6, r8 ; byte (two pixels) to halfwords 61 smlad r4, r6, r6, r4 ; dual signed multiply, add and accumulate (1) 64 ldr r6, [r2, #0x4] ; load 4 ref pixels 67 usub8 r8, r5, r6 ; calculate difference 69 usub8 r9, r6, r5 ; calculate difference with reversed operands 74 usad8 r6, r8, lr ; calculate sum of negative differences [all …]
|
D | vp8_short_fdct4x4_armv6.asm | 34 qadd16 r6, r4, r5 ; [i1+i2 | i0+i3] = [b1 | a1] without shift 43 smuad r4, r6, lr ; o0 = (i1+i2)*8 + (i0+i3)*8 44 smusd r5, r6, lr ; o2 = (i1+i2)*8 - (i0+i3)*8 46 smlad r6, r7, r12, r11 ; o1 = (c1 * 2217 + d1 * 5352 + 14500) 51 pkhbt r3, r4, r6, lsl #4 ; [o1 | o0], keep in register for PART 2 52 pkhbt r6, r5, r7, lsl #4 ; [o3 | o2] 54 str r6, [r1, #4] 59 qadd16 r6, r8, r9 ; [i5+i6 | i4+i7] = [b1 | a1] without shift 68 smuad r9, r6, lr ; o4 = (i5+i6)*8 + (i4+i7)*8 69 smusd r8, r6, lr ; o6 = (i5+i6)*8 - (i4+i7)*8 [all …]
|
D | vp8_subtract_armv6.asm | 33 ldr r6, [r0, #vp8_block_src_diff] 60 str r0, [r6, #0] ; diff 61 str r1, [r6, #4] ; diff 63 add r6, r6, r2, lsl #1 ; update diff pointer 90 ldr r6, [r1] ; usrc (A) 93 uxtb16 r8, r6 ; [s2 | s0] (A) 95 uxtb16 r10, r6, ror #8 ; [s3 | s1] (A) 98 usub16 r6, r8, r9 ; [d2 | d0] (A) 104 pkhbt r8, r6, r7, lsl #16 ; [d1 | d0] (A) 105 pkhtb r9, r7, r6, asr #16 ; [d3 | d2] (A) [all …]
|
/hardware/intel/common/omx-components/videocodec/libvpx_internal/libvpx/vp8/encoder/arm/armv5te/ |
D | vp8_packtokens_mbrow_armv5.asm | 50 ldr r6, _VP8_COMMON_MBrows_ 53 ldr r5, [r4, r6] ; load up mb_rows 81 ldrb r6, [r1, #tokenextra_token] ; t 84 add r4, r4, r6, lsl #3 ; a = vp8_coef_encodings + t 89 ldr r6, [r4, #vp8_token_value] ; v 101 lsl r12, r6, r4 ; r12 = v << 32 - n 112 mul r6, r4, r7 ; ((range-1) * pp[i>>1])) 120 add r4, r7, r6, lsr #8 ; 1 + (((range-1) * pp[i>>1]) >> 8) 126 clz r6, r4 127 sub r6, r6, #24 ; shift [all …]
|
D | vp8_packtokens_partitions_armv5.asm | 51 ldr r6, _VP8_COMMON_MBrows_ 54 ldr r5, [r4, r6] ; load up mb_rows 110 ldrb r6, [r1, #tokenextra_token] ; t 113 add r4, r4, r6, lsl #3 ; a = vp8_coef_encodings + t 118 ldr r6, [r4, #vp8_token_value] ; v 130 lsl r12, r6, r4 ; r12 = v << 32 - n 141 mul r6, r4, r7 ; ((range-1) * pp[i>>1])) 149 add r4, r7, r6, lsr #8 ; 1 + (((range-1) * pp[i>>1]) >> 8) 155 clz r6, r4 156 sub r6, r6, #24 ; shift [all …]
|
D | vp8_packtokens_armv5.asm | 60 ldrb r6, [r1, #tokenextra_token] ; t 63 add r4, r4, r6, lsl #3 ; a = vp8_coef_encodings + t 68 ldr r6, [r4, #vp8_token_value] ; v 80 lsl r12, r6, r4 ; r12 = v << 32 - n 91 mul r6, r4, r7 ; ((range-1) * pp[i>>1])) 99 add r4, r7, r6, lsr #8 ; 1 + (((range-1) * pp[i>>1]) >> 8) 105 clz r6, r4 106 sub r6, r6, #24 ; shift 110 adds r3, r3, r6 ; count += shift 111 lsl r5, r4, r6 ; range <<= shift [all …]
|
D | boolhuff_armv5te.asm | 71 mul r6, r4, r7 ; ((range-1) * probability) 74 add r4, r7, r6, lsr #8 ; 1 + (((range-1) * probability) >> 8) 80 clz r6, r4 81 sub r6, r6, #24 ; shift 85 adds r3, r3, r6 ; count += shift 86 lsl r5, r4, r6 ; range <<= shift 89 sub r6, r6, r3 ; offset = shift - count 90 sub r4, r6, #1 ; offset-1 113 rsb r4, r6, #24 ; 24-offset 117 lsl r2, r2, r6 ; lowvalue <<= offset [all …]
|
/hardware/samsung_slsi/exynos5/libswconverter/ |
D | csc_ARGB8888_to_YUV420SP_NEON.s | 15 @r6 temp7, nWidth16m 200 ldr r6, =0x000000FF 201 and r10, r10, r6 @R: (rgbIn[k] & 0xF800) >> 10; 202 ldr r6, =0x00FF0000 203 and r14, r14, r6 @R: (rgbIn[k] & 0xF800) >> 10; 212 ldr r6, =0x000000FF 213 and r10, r10, r6 @G: 214 ldr r6, =0x00FF0000 215 and r14, r14, r6 @G: 224 ldr r6, =0x000000FF [all …]
|
D | csc_tiled_to_linear_uv_neon.s | 60 @r6 j 81 mov r6, #0 88 add r12, r12, r6 102 add r6, r6, #16 103 cmp r6, r10 111 add r8, r8, r6, lsl #3 115 add r12, r12, r6 117 sub r14, r2, r6 @ r14 = width - j 149 mov r6, #0 152 add r8, r8, r6, lsl #3 [all …]
|
D | csc_tiled_to_linear_y_neon.s | 61 @r6 j 82 mov r6, #0 94 add r12, r12, r6 115 add r6, r6, #16 117 cmp r6, r10 126 add r8, r8, r6, lsl #4 130 add r12, r12, r6 132 sub r14, r2, r6 @ r14 = width - j 164 mov r6, #0 167 add r8, r8, r6, lsl #4 [all …]
|
D | csc_tiled_to_linear_uv_deinterleave_neon.s | 63 @r6 j 84 mov r6, #0 92 add r12, r12, r6, asr #1 117 add r6, r6, #16 119 cmp r6, r10 128 add r8, r8, r6, lsl #3 132 add r12, r12, r6, asr #1 135 sub r14, r3, r6 @ r14 = (width - j) / 2 171 mov r6, #0 174 add r8, r8, r6, lsl #3 [all …]
|
/hardware/intel/common/omx-components/videocodec/libvpx_internal/libvpx/vp8/common/arm/neon/ |
D | buildintrapredictorsmby_neon.asm | 58 sub r6, r0, r2 59 vld1.8 {q1}, [r6] 65 vmov.32 r6, d9[0] 67 add r12, r4, r6 82 ldrb r6, [r0], r2 87 add r12, r12, r6 92 ldrb r6, [r0], r2 97 add r12, r12, r6 102 ldrb r6, [r0], r2 107 add r12, r12, r6 [all …]
|