Mercurial > libavcodec.hg
comparison arm/mdct_neon.S @ 12047:c80c7a717156 libavcodec
Remove vestiges of radix-2 FFT
Patch (mostly) by Loren Merritt
| author | mru |
|---|---|
| date | Thu, 01 Jul 2010 23:21:42 +0000 |
| parents | 361a5fcb4393 |
| children |
comparison
equal
deleted
inserted
replaced
| 12046:ae57be2ef58c | 12047:c80c7a717156 |
|---|---|
| 29 | 29 |
| 30 function ff_imdct_half_neon, export=1 | 30 function ff_imdct_half_neon, export=1 |
| 31 push {r4-r8,lr} | 31 push {r4-r8,lr} |
| 32 | 32 |
| 33 mov r12, #1 | 33 mov r12, #1 |
| 34 ldr lr, [r0, #28] @ mdct_bits | 34 ldr lr, [r0, #20] @ mdct_bits |
| 35 ldr r4, [r0, #32] @ tcos | 35 ldr r4, [r0, #24] @ tcos |
| 36 ldr r3, [r0, #8] @ revtab | 36 ldr r3, [r0, #8] @ revtab |
| 37 lsl r12, r12, lr @ n = 1 << nbits | 37 lsl r12, r12, lr @ n = 1 << nbits |
| 38 lsr lr, r12, #2 @ n4 = n >> 2 | 38 lsr lr, r12, #2 @ n4 = n >> 2 |
| 39 add r7, r2, r12, lsl #1 | 39 add r7, r2, r12, lsl #1 |
| 40 mov r12, #-16 | 40 mov r12, #-16 |
| 74 mov r4, r0 | 74 mov r4, r0 |
| 75 mov r6, r1 | 75 mov r6, r1 |
| 76 bl ff_fft_calc_neon | 76 bl ff_fft_calc_neon |
| 77 | 77 |
| 78 mov r12, #1 | 78 mov r12, #1 |
| 79 ldr lr, [r4, #28] @ mdct_bits | 79 ldr lr, [r4, #20] @ mdct_bits |
| 80 ldr r4, [r4, #32] @ tcos | 80 ldr r4, [r4, #24] @ tcos |
| 81 lsl r12, r12, lr @ n = 1 << nbits | 81 lsl r12, r12, lr @ n = 1 << nbits |
| 82 lsr lr, r12, #3 @ n8 = n >> 3 | 82 lsr lr, r12, #3 @ n8 = n >> 3 |
| 83 | 83 |
| 84 add r4, r4, lr, lsl #3 | 84 add r4, r4, lr, lsl #3 |
| 85 add r6, r6, lr, lsl #3 | 85 add r6, r6, lr, lsl #3 |
| 125 endfunc | 125 endfunc |
| 126 | 126 |
| 127 function ff_imdct_calc_neon, export=1 | 127 function ff_imdct_calc_neon, export=1 |
| 128 push {r4-r6,lr} | 128 push {r4-r6,lr} |
| 129 | 129 |
| 130 ldr r3, [r0, #28] | 130 ldr r3, [r0, #20] |
| 131 mov r4, #1 | 131 mov r4, #1 |
| 132 mov r5, r1 | 132 mov r5, r1 |
| 133 lsl r4, r4, r3 | 133 lsl r4, r4, r3 |
| 134 add r1, r1, r4 | 134 add r1, r1, r4 |
| 135 | 135 |
| 162 | 162 |
| 163 function ff_mdct_calc_neon, export=1 | 163 function ff_mdct_calc_neon, export=1 |
| 164 push {r4-r10,lr} | 164 push {r4-r10,lr} |
| 165 | 165 |
| 166 mov r12, #1 | 166 mov r12, #1 |
| 167 ldr lr, [r0, #28] @ mdct_bits | 167 ldr lr, [r0, #20] @ mdct_bits |
| 168 ldr r4, [r0, #32] @ tcos | 168 ldr r4, [r0, #24] @ tcos |
| 169 ldr r3, [r0, #8] @ revtab | 169 ldr r3, [r0, #8] @ revtab |
| 170 lsl lr, r12, lr @ n = 1 << nbits | 170 lsl lr, r12, lr @ n = 1 << nbits |
| 171 add r7, r2, lr @ in4u | 171 add r7, r2, lr @ in4u |
| 172 sub r9, r7, #16 @ in4d | 172 sub r9, r7, #16 @ in4d |
| 173 add r2, r7, lr, lsl #1 @ in3u | 173 add r2, r7, lr, lsl #1 @ in3u |
| 251 mov r4, r0 | 251 mov r4, r0 |
| 252 mov r6, r1 | 252 mov r6, r1 |
| 253 bl ff_fft_calc_neon | 253 bl ff_fft_calc_neon |
| 254 | 254 |
| 255 mov r12, #1 | 255 mov r12, #1 |
| 256 ldr lr, [r4, #28] @ mdct_bits | 256 ldr lr, [r4, #20] @ mdct_bits |
| 257 ldr r4, [r4, #32] @ tcos | 257 ldr r4, [r4, #24] @ tcos |
| 258 lsl r12, r12, lr @ n = 1 << nbits | 258 lsl r12, r12, lr @ n = 1 << nbits |
| 259 lsr lr, r12, #3 @ n8 = n >> 3 | 259 lsr lr, r12, #3 @ n8 = n >> 3 |
| 260 | 260 |
| 261 add r4, r4, lr, lsl #3 | 261 add r4, r4, lr, lsl #3 |
| 262 add r6, r6, lr, lsl #3 | 262 add r6, r6, lr, lsl #3 |