|
Message
From: Bill Cox <bill@v...>
Date: Wed, 12 Nov 2003 07:05:18 -0500
Subject: Re: [openrisc] eCos on OpenRISC
Scott Furman wrote:
> Matjaz Breskvar wrote:
>
>> i've seen quite some activity and work beeing done on ecos platform. can
>> sombody recomend the most stable version of eCos for openrisc
>> architecture?
>>
>> it seams that most of the original work from Scott has been merged into
>> eCos cvs (with some enhancments) but there is no entry for OpenRISC in
>> 'ecos.db'.
>
> AFAIK, the latest eCos code consists of the contents of the official
> eCos repository overlaid with a patch that I submitted to the
> ecos-patches mailing list in June. Despite repeated requests to the
> eCos maintainers, these patches have never been entered into the
> official eCos repository. My interpretation of the situation is that
> OpenRISC is too marginal a platform to attract their attention. At
> one time I had advocated that we use only the official eCos repository
> rather than creating an OpenCores fork of eCos, but that was assuming
> responsive eCos maintainers. So, now I tend to agree with your
> suggestion that OpenCores retain a stable snapshot of the OpenRISC
> eCos port.
>
> A zip of the official eCos repository contents is here (including a
> patch to the ecos.db file)::
> http://sources.redhat.com/ml/ecos-patches/2003-04/msg00115.shtml
>
> The afforementioned patch (not yet entered into the official eCos
> repository) is here:
> http://sources.redhat.com/ml/ecos-patches/2003-09/msg00051.shtml
>
> Incidentally, if anyone is wondering why there have been no additional
> patches to the OpenRISC eCos port in months, it is because I haven't
> encountered any architecture-specific bugs after months of use.
>
> -Scott
This is a real problem. I've been following this group for a while, and
I've also been talking to various guys looking for embedded uPs. I
think the OpenRISC core is going to proliferate very fast. It basically
needs a bit more maturing, and a few successes in silicon, not just
FPGAs. The other thing it has to have is a RTOS.
The eCos guys would be doing themselves a real service by taking
OpenRISC seriously. The majority of ASIC users out there that I talk to
have a real problem with the $6M for ARM, and the other solutions don't
seem to work very well. I know of at least three companies considering
using the OpenRISC core, if it can mature in time. A good RTOS is a key
factor for each.
Bill
|
 |
Browse
