Nils Wallménius bc6c08c4b9 opus: only put the mdct function in iram for cf, ... doesn't make any speed diff on pp Change-Id: Ia6484e2a2bbf72d33a583470918c0fa267fd9b97		2012-10-17 20:55:49 +02:00
..
_kiss_fft_guts.h	opus: arm asm for C_MULC	2012-10-10 22:54:02 +02:00
arch.h
bands.c
bands.h
celt.c	opus: speed up mdct overlap add and copying	2012-10-07 00:31:08 +02:00
celt.h
celt_lpc.c
celt_lpc.h
cwrs.c	opus: put arrays frequently used by pulse decoding on the stack	2012-10-04 19:20:55 +02:00
cwrs.h
ecintrin.h
entcode.c
entcode.h
entdec.c
entdec.h
entenc.c
entenc.h
fixed_generic.h	opus: cf inline asm for MULT32_32_Q31	2012-10-06 23:43:05 +02:00
float_cast.h
kiss_fft.c
kiss_fft.h	opus: put hot code in iram	2012-10-16 17:52:53 +02:00
laplace.c
laplace.h
mathops.c
mathops.h
mdct.c	opus: don't rely on stride being 1<<shift	2012-10-16 21:03:22 +02:00
mdct.h	opus: only put the mdct function in iram for cf,	2012-10-17 20:55:49 +02:00
mfrngcod.h
modes.c
modes.h
os_support.h
pitch.c
pitch.h
quant_bands.c
quant_bands.h
rate.c
rate.h
stack_alloc.h
static_modes_fixed.h	opus: put some const tables and structs in iram	2012-09-25 17:19:05 +02:00
vq.c	opus: put arrays frequently used by pulse decoding on the stack	2012-10-04 19:20:55 +02:00
vq.h