Speex dev - Feb 2008 - VisualDSP++ with enabled 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

On Tue 19 Feb 2008 11:14, Voss, Stefan, AEAV22 pondered:
> 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++ ?   
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

-Robin

Robin Getz a ?crit :
> On Tue 19 Feb 2008 11:14, 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 can see there was a guy Michael Shatz who tried it,
but no news...

I'm currently 'translating' GNU assembly constraints to VDSP.
I'm far from finished, but first attempts show no improvement over the C 
compiler.
This is rather strange and would suggest memory bandwith limitation 
rather than code efficiency.
I will investigate further by placing all in L1 SRAM.

I'm using speex because of wide-band, and I'm pretty disapointed by 
performance considering it's slower than ADI's MP3 codec (10/20 MIPS for
decoding/encoding of 44100 Hz @ 64 kbit/s).

But maybe Jean-Marc would say the comparison is point-less ?

Best regards.

-- 
Stephane Lesage
ATEIS International