diff options
| author | Mans Rullgard <mans@mansr.com> | 2011-03-09 19:51:50 +0000 |
|---|---|---|
| committer | Mans Rullgard <mans@mansr.com> | 2011-03-24 16:30:49 +0000 |
| commit | f4855a904e4d3a12822739cda149262f98e8723a (patch) | |
| tree | f9a98438f7ed84b21096fec7a2d962360c4e5479 /libavcodec/arm/ac3dsp_neon.S | |
| parent | f71142496931d264ceeedfaac399eba07b7c5def (diff) | |
ac3enc: NEON optimised ac3_max_msb_abs_int16 and ac3_exponent_min
Diffstat (limited to 'libavcodec/arm/ac3dsp_neon.S')
| -rw-r--r-- | libavcodec/arm/ac3dsp_neon.S | 59 |
1 files changed, 59 insertions, 0 deletions
diff --git a/libavcodec/arm/ac3dsp_neon.S b/libavcodec/arm/ac3dsp_neon.S new file mode 100644 index 0000000000..e69a3c2013 --- /dev/null +++ b/libavcodec/arm/ac3dsp_neon.S @@ -0,0 +1,59 @@ +/* + * Copyright (c) 2011 Mans Rullgard <mans@mansr.com> + * + * This file is part of Libav. + * + * Libav is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * Libav is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with Libav; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#include "asm.S" + +function ff_ac3_max_msb_abs_int16_neon, export=1 + vmov.i16 q0, #0 + vmov.i16 q2, #0 +1: vld1.16 {q1}, [r0,:128]! + vabs.s16 q1, q1 + vld1.16 {q3}, [r0,:128]! + vabs.s16 q3, q3 + vorr q0, q0, q1 + vorr q2, q2, q3 + subs r1, r1, #16 + bgt 1b + vorr q0, q0, q2 + vorr d0, d0, d1 + vpmax.u16 d0, d0, d0 + vpmax.u16 d0, d0, d0 + vmov.u16 r0, d0[0] + bx lr +endfunc + +function ff_ac3_exponent_min_neon, export=1 + cmp r1, #0 + bxeq lr + push {lr} + mov r12, #256 +1: + vld1.8 {q0}, [r0,:128] + mov lr, r1 + add r3, r0, #256 +2: vld1.8 {q1}, [r3,:128], r12 + subs lr, lr, #1 + vmin.u8 q0, q0, q1 + bgt 2b + subs r2, r2, #16 + vst1.8 {q0}, [r0,:128]! + bgt 1b + pop {pc} +endfunc |