search for: bfin_asm

Hi I'm trying to integrate your speex codec on our custom Blackfin board and without uCLinux. I am using ADI-supplied VisualDSP++ IDE and corresponding toolchain. My question is: Is there anybody who ported speex with enabled BFIN_ASM to VisualDSP++ ? Best Regards, Stefan Voss -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.xiph.org/pipermail/speex-dev/attachments/20080219/cebf3d19/attachment.htm

...topic? http://lists.xiph.org/pipermail/speex-dev/2008-February/006561.html Michael tried to translate the GNU constraints to VDSP, but it shows no improvement. Is it only for performance or are there other things? I also had a look in the ADI-Speex in their SDK, but it is an older version and the BFIN_ASM does not work. There are no other options, or ASM translations. Is the VDSP-Compiler so good, that performance or other optimizations are not required? Best regards, Daniel

...:14, Voss, Stefan, AEAV22 pondered: >> I'm trying to integrate your speex codec on our custom Blackfin board and >> without uCLinux. I am using ADI-supplied VisualDSP++ IDE and corresponding >> toolchain. My question is: Is there anybody who ported speex with enabled >> BFIN_ASM to VisualDSP++ ? > > Nope - two choices: > - check out using gcc for your entire project > - use the old version of speex that is avaliable as a VDSP download. > http://www.analog.com/processors/platforms/sdk.html Hi Stefan, if you look at the mailing-list archives, you ca...

...-) diff --git a/configure.ac b/configure.ac index cc30d99..3179521 100644 --- a/configure.ac +++ b/configure.ac @@ -188,8 +188,10 @@ fi]) AC_ARG_ENABLE(blackfin-asm, [ --enable-blackfin-asm Make use of Blackfin assembly optimizations], [if test "$enableval" = yes; then AC_DEFINE([BFIN_ASM], , [Make use of Blackfin assembly optimizations]) - LDFLAGS="-Wl,-elf2flt=-s100000" fi]) +case $host_os in + uclinux) LDFLAGS="-Wl,-elf2flt=-s100000 $LDFLAGS";; +esac AC_ARG_ENABLE(fixed-point-debug, [ --enable-fixed-point-debug Debug fixed-point implementation], [if t...

...peex lib 1.1.11.1. I am trying to port speex lib to Blackfin processor. I am using VisualDSP++ 4.0. If I am compiling source codes with using floating point everything ok. When I am compiling with FIXED_POINT defined everything's ok and code works about two times faster. But when I am defining BFIN_ASM I am getting several compiling errors in Blackfin assembler functions. Here they are: 1) In the inline assembler functions (blackfin asm) in every loop there is a syntax error. May be I am wrong, but I can't understood what is written there(and compiler too). Example: __asm__ __volatile__...

...cific](https://cmake.org/cmake/help/latest/manual/cmake-packages.7.html#package-configuration-file) * Configured conditionals (`config.h`): * `WORDS_BIGENDIAN` * `FLOATING_POINT` * `FIXED_POINT` * `FIXED_POINT_DEBUG` * `DISABLE_FLOAT_API` * `_USE_SSE` * `ARM4_ASM` * `ARM5E_ASM` * `BFIN_ASM` * `TI_C55X` * `DISABLE_VBR` * `USE_SPEEXDSP` * `VORBIS_PSYCHO` * `HAVE_GETOPT_H` * `HAVE_GETOPT_LONG` * `VAR_ARRAYS` * `USE_ALLOCA` * `HAVE_ALLOCA_H` * `USE_SMALLFT` * `USE_KISS_FFT` * `USE_GPL_FFTW3` * `EXPORT` * Installation supported * Package config mode supported Ho...

...or more like the MS compilers. > If I am compiling source codes with using floating point everything > ok. > When I am compiling with FIXED_POINT defined everything's ok and code > works about two times faster. Strange that it's *only* 2x faster... > But when I am defining BFIN_ASM I am getting several compiling errors > in Blackfin assembler functions. > Here they are: > 1) In the inline assembler functions (blackfin asm) in every loop > there is a syntax error. May be I am wrong, but I can't understood > what is written there(and compiler too). > Examp...

...b/celt/arch.h @@ -122,6 +122,8 @@ static OPUS_INLINE opus_int16 SAT16(opus_int32 x) { #include "arm/fixed_armv5e.h" #elif defined (OPUS_ARM_INLINE_ASM) #include "arm/fixed_armv4.h" +#elif defined (OPUS_ARM64_INLINE_ASM) +#include "arm/fixed_arm64.h" #elif defined (BFIN_ASM) #include "fixed_bfin.h" #elif defined (TI_C5X_ASM) diff --git a/celt/arm/fixed_arm64.h b/celt/arm/fixed_arm64.h new file mode 100644 index 0000000..7fa413b --- /dev/null +++ b/celt/arm/fixed_arm64.h @@ -0,0 +1,75 @@ +/* Copyright (C) 2014 Vidyo */ +/* + Redistribution and use in sour...

...b/celt/arch.h @@ -122,6 +122,8 @@ static OPUS_INLINE opus_int16 SAT16(opus_int32 x) { #include "arm/fixed_armv5e.h" #elif defined (OPUS_ARM_INLINE_ASM) #include "arm/fixed_armv4.h" +#elif defined (OPUS_ARM64_INLINE_ASM) +#include "arm/fixed_arm64.h" #elif defined (BFIN_ASM) #include "fixed_bfin.h" #elif defined (TI_C5X_ASM) diff --git a/celt/arm/fixed_arm64.h b/celt/arm/fixed_arm64.h new file mode 100644 index 0000000..7fa413b --- /dev/null +++ b/celt/arm/fixed_arm64.h @@ -0,0 +1,75 @@ +/* Copyright (C) 2014 Vidyo */ +/* + Redistribution and use in sour...

As a note, with the latest versions, I get this "new" failure, when cross compiling for windows using mingw: CCLD speexdec.exe speexdec.o:speexdec.c:(.text+0x4af): undefined reference to `_imp__waveOutGetNumDevs at 0' speexdec.o:speexdec.c:(.text+0x54d): undefined reference to `_imp__waveOutOpen at 24' speexdec.o:speexdec.c:(.text+0x5aa): undefined reference to

...-#ifdef OPUS_ARM_INLINE_EDSP +#if defined(OPUS_ARM_INLINE_EDSP) && defined(FIXED_POINT) #include "arm/fixed_armv5e.h" -#elif defined (OPUS_ARM_INLINE_ASM) +#elif defined(OPUS_ARM_INLINE_ASM) && defined(FIXED_POINT) #include "arm/fixed_armv4.h" #elif defined (BFIN_ASM) #include "fixed_bfin.h" diff --git a/celt/arm/arm_celt_map.c b/celt/arm/arm_celt_map.c index 547a84d..a6313e7 100644 --- a/celt/arm/arm_celt_map.c +++ b/celt/arm/arm_celt_map.c @@ -42,8 +42,19 @@ opus_val32 (*const CELT_PITCH_XCORR_IMPL[OPUS_ARCHMASK+1])(const opus_val16 *, MAY_HAVE_...

Hello, I received feedback from engineers working on NE10 [1] that it would be better to use NE10 [1] for FFT optimizations for opus use cases. However, these FFT patches are currently in review and haven't been integrated into NE10 yet. While the FFT functions in NE10 are getting baked, I wanted to optimize the celt_pitch_xcorr (floating point only) and use it to introduce ARM NEON

This sequence of patches provides arm64 support for Opus. Tested on iOS, Android, and Ubuntu 14.04. The patch sequence was written on top of Viswanath Puttagunta's Ne10 patches, but all but the second ("Reorganize pitch_arm.h") should, I think, apply independently of it. It does depends on my previous intrinsics configury reorganization, however. Comments welcome. With this and

Here are my aarch64 patches rebased to the current tip of Opus master. They're largely the same as my previous patch set, with the addition of the final one (the Neon fixed-point implementation of xcorr_kernel). This replaces Viswanath's Neon fixed-point celt_pitch_xcorr, since xcorr_kernel is used in celt_fir and celt_iir as well. These have been tested for correctness under qemu