#include "libavutil/aarch64/asm.S"
function ff_ps_add_squares_neon, export=1
-1: ld1 {v0.4S,v1.4S}, [x1], #32
- fmul v0.4S, v0.4S, v0.4S
- fmul v1.4S, v1.4S, v1.4S
- faddp v2.4S, v0.4S, v1.4S
- ld1 {v3.4S}, [x0]
- fadd v3.4S, v3.4S, v2.4S
- st1 {v3.4S}, [x0], #16
- subs w2, w2, #4
- b.gt 1b
+1: ld1 {v0.4s,v1.4s}, [x1], #32
+ fmul v0.4s, v0.4s, v0.4s
+ fmul v1.4s, v1.4s, v1.4s
+ faddp v2.4s, v0.4s, v1.4s
+ ld1 {v3.4s}, [x0]
+ fadd v3.4s, v3.4s, v2.4s
+ st1 {v3.4s}, [x0], #16
+ subs w2, w2, #4
+ b.gt 1b
ret
endfunc
function ff_ps_mul_pair_single_neon, export=1
-1: ld1 {v0.4S,v1.4S}, [x1], #32
- ld1 {v2.4S}, [x2], #16
- zip1 v3.4S, v2.4S, v2.4S
- zip2 v4.4S, v2.4S, v2.4S
- fmul v0.4S, v0.4S, v3.4S
- fmul v1.4S, v1.4S, v4.4S
- st1 {v0.4S,v1.4S}, [x0], #32
- subs w3, w3, #4
- b.gt 1b
+1: ld1 {v0.4s,v1.4s}, [x1], #32
+ ld1 {v2.4s}, [x2], #16
+ zip1 v3.4s, v2.4s, v2.4s
+ zip2 v4.4s, v2.4s, v2.4s
+ fmul v0.4s, v0.4s, v3.4s
+ fmul v1.4s, v1.4s, v4.4s
+ st1 {v0.4s,v1.4s}, [x0], #32
+ subs w3, w3, #4
+ b.gt 1b
ret
endfunc
function ff_ps_stereo_interpolate_neon, export=1
- ld1 {v0.4S}, [x2]
- ld1 {v1.4S}, [x3]
- zip1 v4.4S, v0.4S, v0.4S
- zip2 v5.4S, v0.4S, v0.4S
- zip1 v6.4S, v1.4S, v1.4S
- zip2 v7.4S, v1.4S, v1.4S
-1: ld1 {v2.2S}, [x0]
- ld1 {v3.2S}, [x1]
- fadd v4.4S, v4.4S, v6.4S
- fadd v5.4S, v5.4S, v7.4S
- mov v2.D[1], v2.D[0]
- mov v3.D[1], v3.D[0]
- fmul v2.4S, v2.4S, v4.4S
- fmla v2.4S, v3.4S, v5.4S
- st1 {v2.D}[0], [x0], #8
- st1 {v2.D}[1], [x1], #8
- subs w4, w4, #1
- b.gt 1b
+ ld1 {v0.4s}, [x2]
+ ld1 {v1.4s}, [x3]
+ zip1 v4.4s, v0.4s, v0.4s
+ zip2 v5.4s, v0.4s, v0.4s
+ zip1 v6.4s, v1.4s, v1.4s
+ zip2 v7.4s, v1.4s, v1.4s
+1: ld1 {v2.2s}, [x0]
+ ld1 {v3.2s}, [x1]
+ fadd v4.4s, v4.4s, v6.4s
+ fadd v5.4s, v5.4s, v7.4s
+ mov v2.d[1], v2.d[0]
+ mov v3.d[1], v3.d[0]
+ fmul v2.4s, v2.4s, v4.4s
+ fmla v2.4s, v3.4s, v5.4s
+ st1 {v2.d}[0], [x0], #8
+ st1 {v2.d}[1], [x1], #8
+ subs w4, w4, #1
+ b.gt 1b
ret
endfunc
function ff_ps_stereo_interpolate_ipdopd_neon, export=1
- ld1 {v0.4S,v1.4S}, [x2]
- ld1 {v6.4S,v7.4S}, [x3]
- fneg v2.4S, v1.4S
- fneg v3.4S, v7.4S
- zip1 v16.4S, v0.4S, v0.4S
- zip2 v17.4S, v0.4S, v0.4S
- zip1 v18.4S, v2.4S, v1.4S
- zip2 v19.4S, v2.4S, v1.4S
- zip1 v20.4S, v6.4S, v6.4S
- zip2 v21.4S, v6.4S, v6.4S
- zip1 v22.4S, v3.4S, v7.4S
- zip2 v23.4S, v3.4S, v7.4S
-1: ld1 {v2.2S}, [x0]
- ld1 {v3.2S}, [x1]
- fadd v16.4S, v16.4S, v20.4S
- fadd v17.4S, v17.4S, v21.4S
- mov v2.D[1], v2.D[0]
- mov v3.D[1], v3.D[0]
- fmul v4.4S, v2.4S, v16.4S
- fmla v4.4S, v3.4S, v17.4S
- fadd v18.4S, v18.4S, v22.4S
- fadd v19.4S, v19.4S, v23.4S
- ext v2.16B, v2.16B, v2.16B, #4
- ext v3.16B, v3.16B, v3.16B, #4
- fmla v4.4S, v2.4S, v18.4S
- fmla v4.4S, v3.4S, v19.4S
- st1 {v4.D}[0], [x0], #8
- st1 {v4.D}[1], [x1], #8
- subs w4, w4, #1
- b.gt 1b
+ ld1 {v0.4s,v1.4s}, [x2]
+ ld1 {v6.4s,v7.4s}, [x3]
+ fneg v2.4s, v1.4s
+ fneg v3.4s, v7.4s
+ zip1 v16.4s, v0.4s, v0.4s
+ zip2 v17.4s, v0.4s, v0.4s
+ zip1 v18.4s, v2.4s, v1.4s
+ zip2 v19.4s, v2.4s, v1.4s
+ zip1 v20.4s, v6.4s, v6.4s
+ zip2 v21.4s, v6.4s, v6.4s
+ zip1 v22.4s, v3.4s, v7.4s
+ zip2 v23.4s, v3.4s, v7.4s
+1: ld1 {v2.2s}, [x0]
+ ld1 {v3.2s}, [x1]
+ fadd v16.4s, v16.4s, v20.4s
+ fadd v17.4s, v17.4s, v21.4s
+ mov v2.d[1], v2.d[0]
+ mov v3.d[1], v3.d[0]
+ fmul v4.4s, v2.4s, v16.4s
+ fmla v4.4s, v3.4s, v17.4s
+ fadd v18.4s, v18.4s, v22.4s
+ fadd v19.4s, v19.4s, v23.4s
+ ext v2.16b, v2.16b, v2.16b, #4
+ ext v3.16b, v3.16b, v3.16b, #4
+ fmla v4.4s, v2.4s, v18.4s
+ fmla v4.4s, v3.4s, v19.4s
+ st1 {v4.d}[0], [x0], #8
+ st1 {v4.d}[1], [x1], #8
+ subs w4, w4, #1
+ b.gt 1b
ret
endfunc
function ff_ps_hybrid_analysis_neon, export=1
- lsl x3, x3, #3
- ld2 {v0.4S,v1.4S}, [x1], #32
- ld2 {v2.2S,v3.2S}, [x1], #16
- ld1 {v24.2S}, [x1], #8
- ld2 {v4.2S,v5.2S}, [x1], #16
- ld2 {v6.4S,v7.4S}, [x1]
- rev64 v6.4S, v6.4S
- rev64 v7.4S, v7.4S
- ext v6.16B, v6.16B, v6.16B, #8
- ext v7.16B, v7.16B, v7.16B, #8
- rev64 v4.2S, v4.2S
- rev64 v5.2S, v5.2S
- mov v2.D[1], v3.D[0]
- mov v4.D[1], v5.D[0]
- mov v5.D[1], v2.D[0]
- mov v3.D[1], v4.D[0]
- fadd v16.4S, v0.4S, v6.4S
- fadd v17.4S, v1.4S, v7.4S
- fsub v18.4S, v1.4S, v7.4S
- fsub v19.4S, v0.4S, v6.4S
- fadd v22.4S, v2.4S, v4.4S
- fsub v23.4S, v5.4S, v3.4S
- trn1 v20.2D, v22.2D, v23.2D // {re4+re8, re5+re7, im8-im4, im7-im5}
- trn2 v21.2D, v22.2D, v23.2D // {im4+im8, im5+im7, re4-re8, re5-re7}
-1: ld2 {v2.4S,v3.4S}, [x2], #32
- ld2 {v4.2S,v5.2S}, [x2], #16
- ld1 {v6.2S}, [x2], #8
- add x2, x2, #8
- mov v4.D[1], v5.D[0]
- mov v6.S[1], v6.S[0]
- fmul v6.2S, v6.2S, v24.2S
- fmul v0.4S, v2.4S, v16.4S
- fmul v1.4S, v2.4S, v17.4S
- fmls v0.4S, v3.4S, v18.4S
- fmla v1.4S, v3.4S, v19.4S
- fmla v0.4S, v4.4S, v20.4S
- fmla v1.4S, v4.4S, v21.4S
- faddp v0.4S, v0.4S, v1.4S
- faddp v0.4S, v0.4S, v0.4S
- fadd v0.2S, v0.2S, v6.2S
- st1 {v0.2S}, [x0], x3
- subs w4, w4, #1
- b.gt 1b
+ lsl x3, x3, #3
+ ld2 {v0.4s,v1.4s}, [x1], #32
+ ld2 {v2.2s,v3.2s}, [x1], #16
+ ld1 {v24.2s}, [x1], #8
+ ld2 {v4.2s,v5.2s}, [x1], #16
+ ld2 {v6.4s,v7.4s}, [x1]
+ rev64 v6.4s, v6.4s
+ rev64 v7.4s, v7.4s
+ ext v6.16b, v6.16b, v6.16b, #8
+ ext v7.16b, v7.16b, v7.16b, #8
+ rev64 v4.2s, v4.2s
+ rev64 v5.2s, v5.2s
+ mov v2.d[1], v3.d[0]
+ mov v4.d[1], v5.d[0]
+ mov v5.d[1], v2.d[0]
+ mov v3.d[1], v4.d[0]
+ fadd v16.4s, v0.4s, v6.4s
+ fadd v17.4s, v1.4s, v7.4s
+ fsub v18.4s, v1.4s, v7.4s
+ fsub v19.4s, v0.4s, v6.4s
+ fadd v22.4s, v2.4s, v4.4s
+ fsub v23.4s, v5.4s, v3.4s
+ trn1 v20.2d, v22.2d, v23.2d // {re4+re8, re5+re7, im8-im4, im7-im5}
+ trn2 v21.2d, v22.2d, v23.2d // {im4+im8, im5+im7, re4-re8, re5-re7}
+1: ld2 {v2.4s,v3.4s}, [x2], #32
+ ld2 {v4.2s,v5.2s}, [x2], #16
+ ld1 {v6.2s}, [x2], #8
+ add x2, x2, #8
+ mov v4.d[1], v5.d[0]
+ mov v6.s[1], v6.s[0]
+ fmul v6.2s, v6.2s, v24.2s
+ fmul v0.4s, v2.4s, v16.4s
+ fmul v1.4s, v2.4s, v17.4s
+ fmls v0.4s, v3.4s, v18.4s
+ fmla v1.4s, v3.4s, v19.4s
+ fmla v0.4s, v4.4s, v20.4s
+ fmla v1.4s, v4.4s, v21.4s
+ faddp v0.4s, v0.4s, v1.4s
+ faddp v0.4s, v0.4s, v0.4s
+ fadd v0.2s, v0.2s, v6.2s
+ st1 {v0.2s}, [x0], x3
+ subs w4, w4, #1
+ b.gt 1b
ret
endfunc
projects / platform / upstream / ffmpeg.git / blobdiff