| /external/libvpx/vp8/encoder/arm/neon/ |
| D | variance_neon.asm | 29 vmov.i8 q8, #0 ;q8 - sum
49 vpadal.s16 q8, q11 ;calculate sum
55 vpadal.s16 q8, q12
58 vpadal.s16 q8, q13
61 vpadal.s16 q8, q14
68 vpaddl.s32 q0, q8 ;accumulate sum
102 vmov.i8 q8, #0 ;q8 - sum
119 vpadal.s16 q8, q11 ;calculate sum
125 vpadal.s16 q8, q12
128 vpadal.s16 q8, q13
[all …]
|
| D | vp8_subpixelvariance16x16s_neon.asm | 37 vmov.i8 q8, #0 ;q8 - sum
75 vpadal.s16 q8, q4 ;sum
81 vpadal.s16 q8, q5
84 vpadal.s16 q8, q6
87 vpadal.s16 q8, q7
91 vpadal.s16 q8, q0 ;sum
94 vpadal.s16 q8, q1
97 vpadal.s16 q8, q2
100 vpadal.s16 q8, q3
107 vpaddl.s32 q0, q8 ;accumulate sum
[all …]
|
| D | vp8_mse16x16_neon.asm | 31 vmov.i8 q7, #0 ;q7, q8, q9, q10 - sse
32 vmov.i8 q8, #0
50 vmlal.s16 q8, d23, d23
57 vmlal.s16 q8, d27, d27
63 vadd.u32 q7, q7, q8
85 vmov.i8 q8, #0 ;q8 - sum
102 vpadal.s16 q8, q11
108 vpadal.s16 q8, q12
111 vpadal.s16 q8, q13
114 vpadal.s16 q8, q14
[all …]
|
| D | vp8_subpixelvariance8x8_neon.asm | 51 vmull.u8 q8, d6, d0
61 vmlal.u8 q8, d7, d1
69 vqrshrn.u16 d24, q8, #7
78 vmull.u8 q8, d6, d0
90 vmlal.u8 q8, d7, d1
96 vqrshrn.u16 d28, q8, #7
120 vmull.u8 q8, d29, d0
129 vmlal.u8 q8, d30, d1
138 vqrshrn.u16 d29, q8, #7
159 vmov.i8 q8, #0 ;q8 - sum
[all …]
|
| D | fastfdct4x4_neon.asm | 56 vqdmulh.s16 q8, q4, d0[0]
60 vshr.s16 q8, q8, #1
62 …vadd.s16 q8, q4, q8 ;d16:temp2 = ((c1 * x_c1)>>16) + c1; d17:temp2 = ((d1 * x…
86 vqdmulh.s16 q8, q4, d0[0]
90 vshr.s16 q8, q8, #1
92 …vadd.s16 q8, q4, q8 ;d16:temp2 = ((c1 * x_c1)>>16) + c1; d17:temp2 = ((d1 * x…
|
| D | vp8_subpixelvariance16x16_neon.asm | 66 vmull.u8 q8, d3, d0
89 vmlal.u8 q8, d3, d1 ;(src_ptr[0] * Filter[1])
97 vqrshrn.u16 d15, q8, #7
202 vmull.u8 q8, d29, d0
211 vmlal.u8 q8, d31, d1
222 vqrshrn.u16 d9, q8, #7
254 vmull.u8 q8, d3, d0
277 vmlal.u8 q8, d3, d1 ;(src_ptr[0] * Filter[1])
285 vqrshrn.u16 d15, q8, #7
329 vmull.u8 q8, d29, d0
[all …]
|
| D | fastquantizeb_neon.asm | 56 vld1.s16 {q8, q9}, [r12] ;load quant_ptr [0-15]
63 vqdmulh.s16 q4, q8 ;y = ((Round + abs(z)) * Quant) >> 16
67 vceq.s16 q8, q8 ;set q8 to all 1
|
| D | subtract_neon.asm | 65 vsubl.u8 q8, d0, d2
74 vst1.16 {q8}, [r0]! ;store diff
115 vsubl.u8 q8, d0, d1
124 vst1.16 {q8}, [r0]! ;store diff
151 vsubl.u8 q8, d0, d1
160 vst1.16 {q8}, [r0]! ;store diff
|
| D | fastfdct8x4_neon.asm | 69 vqdmulh.s16 q8, q7, d0[1]
76 vshr.s16 q8, q8, #1
120 vqdmulh.s16 q8, q7, d0[1]
127 vshr.s16 q8, q8, #1
148 vclt.s16 q8, q4, #0
153 vsub.s16 q4, q4, q8
|
| /external/libvpx/vp8/common/arm/neon/ |
| D | sixtappredict8x4_neon.asm | 69 vmull.u8 q8, d8, d0
79 vmlsl.u8 q8, d29, d1
89 vmlsl.u8 q8, d29, d4
99 vmlal.u8 q8, d29, d2
109 vmlal.u8 q8, d29, d5
124 vqadd.s16 q8, q4
131 vqrshrun.s16 d23, q8, #7
145 vmull.u8 q8, d6, d0 ;(src_ptr[-2] * vp8_filter[0])
157 vmlsl.u8 q8, d27, d1 ;-(src_ptr[-1] * vp8_filter[1])
169 vmlsl.u8 q8, d27, d4 ;-(src_ptr[2] * vp8_filter[4])
[all …]
|
| D | sixtappredict8x8_neon.asm | 73 vmull.u8 q8, d8, d0
83 vmlsl.u8 q8, d29, d1
93 vmlsl.u8 q8, d29, d4
103 vmlal.u8 q8, d29, d2
113 vmlal.u8 q8, d29, d5
130 vqadd.s16 q8, q4
137 vqrshrun.s16 d23, q8, #7
158 vmull.u8 q8, d6, d0 ;(src_ptr[-2] * vp8_filter[0])
170 vmlsl.u8 q8, d27, d1 ;-(src_ptr[-1] * vp8_filter[1])
182 vmlsl.u8 q8, d27, d4 ;-(src_ptr[2] * vp8_filter[4])
[all …]
|
| D | sixtappredict4x4_neon.asm | 75 vmull.u8 q8, d20, d5
85 vmlal.u8 q8, d10, d0
92 vmlsl.u8 q8, d20, d1
99 vmlsl.u8 q8, d10, d4
106 vmlal.u8 q8, d20, d2
117 vqadd.s16 q8, q10
123 vqrshrun.s16 d28, q8, #7
140 vmull.u8 q8, d20, d5
152 vmlal.u8 q8, d10, d0
162 vmlsl.u8 q8, d20, d1
[all …]
|
| D | mbloopfilter_neon.asm | 48 vld1.u8 {q8}, [r0], r1 ; q1
64 vst1.u8 {q8}, [r12], r1 ; store oq1
169 vtrn.32 q4, q8
176 vtrn.16 q8, q10
180 vtrn.8 q7, q8
202 vtrn.32 q4, q8
210 vtrn.16 q8, q10
214 vtrn.8 q7, q8
283 vtrn.32 q4, q8
290 vtrn.16 q8, q10
[all …]
|
| D | sixtappredict16x16_neon.asm | 78 vmull.u8 q8, d6, d0 ;(src_ptr[-2] * vp8_filter[0])
89 vmlsl.u8 q8, d28, d1 ;-(src_ptr[-1] * vp8_filter[1])
105 vmlsl.u8 q8, d28, d4 ;-(src_ptr[2] * vp8_filter[4])
121 vmlal.u8 q8, d28, d5 ;(src_ptr[3] * vp8_filter[5])
137 vmlal.u8 q8, d28, d2 ;(src_ptr[0] * vp8_filter[2])
161 vqadd.s16 q8, q4 ;sum of all (src_data*filter_parameters)
175 vqrshrun.s16 d6, q8, #7 ;shift/round/saturate to u8
196 vabs.s32 q8, q6
247 vmull.u8 q8, d22, d3
254 vqadd.s16 q8, q4
[all …]
|
| D | loopfilter_neon.asm | 48 vld1.u8 {q8}, [r2], r1 ; q1
59 vst1.u8 {q8}, [r0], r1 ; store oq1
163 vtrn.32 q4, q8
170 vtrn.16 q8, q10
174 vtrn.8 q7, q8
247 vtrn.32 q4, q8
254 vtrn.16 q8, q10
258 vtrn.8 q7, q8
307 ; q8 q1
317 vabd.u8 q14, q8, q7 ; abs(q1 - q0)
[all …]
|
| D | bilinearpredict8x8_neon.asm | 48 vmull.u8 q8, d6, d0
58 vmlal.u8 q8, d7, d1
66 vqrshrn.u16 d24, q8, #7
75 vmull.u8 q8, d6, d0
87 vmlal.u8 q8, d7, d1
93 vqrshrn.u16 d28, q8, #7
118 vmull.u8 q8, d29, d0
127 vmlal.u8 q8, d30, d1
136 vqrshrn.u16 d9, q8, #7
|
| D | bilinearpredict16x16_neon.asm | 63 vmull.u8 q8, d3, d0
86 vmlal.u8 q8, d3, d1 ;(src_ptr[0] * vp8_filter[1])
94 vqrshrn.u16 d15, q8, #7
196 vmull.u8 q8, d29, d0
205 vmlal.u8 q8, d31, d1
216 vqrshrn.u16 d9, q8, #7
248 vmull.u8 q8, d3, d0
271 vmlal.u8 q8, d3, d1 ;(src_ptr[0] * vp8_filter[1])
279 vqrshrn.u16 d15, q8, #7
320 vmull.u8 q8, d29, d0
[all …]
|
| D | loopfiltersimplehorizontaledge_neon.asm | 39 vld1.u8 {q8}, [r0] ; q1
43 vabd.u8 q14, q5, q8 ; abs(p1 - q1)
52 veor q8, q8, q0 ; qs1: q1 offset to convert to a signed value
63 vqsub.s8 q4, q5, q8 ; q4: vp8_filter = vp8_signed_char_clamp(ps1-qs1)
|
| D | recon16x16mb_neon.asm | 31 vld1.16 {q8, q9}, [r1]! ;load data from diff_ptr
50 vadd.s16 q0, q0, q8 ;add Diff data and Pred data together
82 vld1.16 {q8, q9}, [r1]! ;load data from diff_ptr
97 vadd.s16 q0, q0, q8 ;add Diff data and Pred data together
|
| /external/libvpx/vp8/decoder/arm/neon/ |
| D | idct_dequant_dc_full_2x_neon.asm | 69 ; q8: 4 * cospi
73 vqdmulh.s16 q8, q4, d0[0] ; cospi8sqrt2minus1
87 vshr.s16 q8, q8, #1
92 vqadd.s16 q4, q4, q8
120 ; q8: 1 * sinpi : c1/temp1
124 vqdmulh.s16 q8, q5, d0[2] ; sinpi8sqrt2
141 ; q8: c1 = temp1 - temp2
143 vqsub.s16 q8, q8, q11
151 vqadd.s16 q5, q3, q8
152 vqsub.s16 q6, q3, q8
|
| D | idct_dequant_full_2x_neon.asm | 61 ; q8: 4 * cospi
65 vqdmulh.s16 q8, q4, d0[0] ; cospi8sqrt2minus1
79 vshr.s16 q8, q8, #1
84 vqadd.s16 q4, q4, q8
112 ; q8: 1 * sinpi : c1/temp1
116 vqdmulh.s16 q8, q5, d0[2] ; sinpi8sqrt2
133 ; q8: c1 = temp1 - temp2
135 vqsub.s16 q8, q8, q11
143 vqadd.s16 q5, q3, q8
144 vqsub.s16 q6, q3, q8
|
| /external/libvpx/vpx_scale/arm/neon/ |
| D | vp8_vpxyv12_copyframeyonly_neon.asm | 49 vld1.8 {q8, q9}, [r10]!
60 vst1.8 {q8, q9}, [r11]!
123 vmov q9, q8
134 vst1.8 {q8, q9}, [r5], lr
154 vld1.8 {q8, q9}, [r2]!
168 vst1.8 {q8, q9}, [r6]!
264 vst1.8 {q8}, [r5], lr
284 vld1.8 {q8, q9}, [r2]!
298 vst1.8 {q8, q9}, [r6]!
426 vld1.8 {q8, q9}, [r10]!
[all …]
|
| D | vp8_vpxyv12_extendframeborders_neon.asm | 65 vmov q9, q8
76 vst1.8 {q8, q9}, [r5], lr
96 vld1.8 {q8, q9}, [r2]!
110 vst1.8 {q8, q9}, [r6]!
185 vst1.8 {q8}, [r5], lr
206 vld1.8 {q8, q9}, [r2]!
216 vst1.8 {q8, q9}, [r6]!
347 vst1.8 {q8}, [r5], lr
367 vld1.8 {q8, q9}, [r2]!
381 vst1.8 {q8, q9}, [r6]!
[all …]
|
| D | vp8_vpxyv12_copyframe_func_neon.asm | 59 vld1.8 {q8, q9}, [r10]!
70 vst1.8 {q8, q9}, [r11]!
117 vld1.8 {q8, q9}, [r10]!
124 vst1.8 {q8, q9}, [r11]!
|
| D | vp8_vpxyv12_copysrcframe_func_neon.asm | 56 vld1.8 {q8, q9}, [r2]!
66 vst1.8 {q8, q9}, [r3]!
111 vld1.8 {q8, q9}, [r2]!
117 vst1.8 {q8, q9}, [r3]!
|