Reply by Ulf Samuelsson●September 7, 20082008-09-07
"C. J. Clegg" <answer.to.newsgroup@nospam.com> skrev i meddelandet
news:5tt5c4puar410rh86d54v99jncgr021qpi@4ax.com...
>
> Good afternoon, Ulf.
>
> Back in November of 2007, in comp.arch.embedded, you wrote:
>
> "I see as many ARM9 users as ARM7 users today and neither the CM3 or
> the uC3k will meet the performance of the ARM9 and neither will run
> Linux/WinCE. - uCLinux? ...sigh!"
>
> I've only started reading about uCLinux over the last couple of days
> but I was wondering why you "sigh" at it.
>
> I'm told that the Atmel AVR32UC3A/B (and their associated eval boards,
> the EVK1100/1) cannot run Linux because the processor doesn't have an
> MMU... hence something like the AVR32-AP7, at a minimum, is needed.
>
> On the other hand, uCLinux is specifically designed to run on
> processors without MMUs. So, any reason why uCLinux cannot (or should
> not) run on the AVR32UCA/B?
uCLinux could run on any decent processor without an MMU.
Many applications which runs under Linux will not run under uCLinux,
so they have to be rewritten.
The memory handling sucks, so you probably need more memory
with a uCLinux than you need with a Linux.
The current 2.6 kernel can easily be > 1,6 MB (compressed)
so you do not fit into the flash anyway.
See: http://www.linuxjournal.com/article/7221
>
> Certainly uCLinux has its limitations but if it's the only flavor of
> Linux that will run on processors without an MMU, and the lack of MMU
> is the reason why the UC3x can't run Linux, then I would think that a
> uCLinux port for the UC3x would be welcomed by low-end processor
> developers... would it not?
>
Show me a project which is more cost efficient using uCLinux than Linux?
Even if you pay a tad bit more for such a CPU,
you also have to calculate the cost of maintenance to continously
port new versions of S/W from Linux to uCLinux.
Also the cost of delay of introduction due to the choice of uCLinux.
--
Best Regards,
Ulf Samuelsson
This is intended to be my personal opinion which may,
or may not be shared by my employer Atmel Nordic AB
Reply by Frank Buss●September 7, 20082008-09-07
C. J. Clegg wrote:
> Good afternoon, Ulf.
>
> Back in November of 2007, in comp.arch.embedded, you wrote:
>
> "I see as many ARM9 users as ARM7 users today and neither the CM3 or
> the uC3k will meet the performance of the ARM9 and neither will run
> Linux/WinCE. - uCLinux? ...sigh!"
>
> I've only started reading about uCLinux over the last couple of days
> but I was wondering why you "sigh" at it.
>
> I'm told that the Atmel AVR32UC3A/B (and their associated eval boards,
> the EVK1100/1) cannot run Linux because the processor doesn't have an
> MMU... hence something like the AVR32-AP7, at a minimum, is needed.
>
> On the other hand, uCLinux is specifically designed to run on
> processors without MMUs. So, any reason why uCLinux cannot (or should
> not) run on the AVR32UCA/B?
>
> Certainly uCLinux has its limitations but if it's the only flavor of
> Linux that will run on processors without an MMU, and the lack of MMU
> is the reason why the UC3x can't run Linux, then I would think that a
> uCLinux port for the UC3x would be welcomed by low-end processor
> developers... would it not?
On Sun, 07 Sep 2008 08:09:58 +0300, Paul Keinanen <keinanen@sci.fi>
wrote:
>Is there a specific reason, why you want to run something Unix
>look-alike services on low resource platforms ?
Good morning, Paul.
As a training exercise, mostly. I don't know very much that's worth
knowing about embedded Linux, and virtually nothing at all about
bringing Linux up on a new platform, and I've always been best at
learning by doing and not by reading some book or attending a class.
That said, I have a friend who is a customer support engineer at the
local Atmel office, and he says they do get inquiries and requests for
Linux on the AVR32UC3x.... just not enough to justify the porting
effort.
Reply by Paul Keinanen●September 7, 20082008-09-07
On Sat, 06 Sep 2008 17:42:25 -0400, C. J. Clegg
<answer.to.newsgroup@nospam.com> wrote:
>On the other hand, uCLinux is specifically designed to run on
>processors without MMUs. So, any reason why uCLinux cannot (or should
>not) run on the AVR32UCA/B?
>
>Certainly uCLinux has its limitations but if it's the only flavor of
>Linux that will run on processors without an MMU, and the lack of MMU
>is the reason why the UC3x can't run Linux, then I would think that a
>uCLinux port for the UC3x would be welcomed by low-end processor
>developers... would it not?
Is there a specific reason, why you want to run something Unix
look-alike services on low resource platforms ?
While this makes sense when trying to port some existing Unix/Linux
application to such platform, but for a completely new application,
what is the point ?
When designing completely new applications in a shared address space
environment, much simpler and lower overhead tasking can be used.
Paul
Reply by C. J. Clegg●September 6, 20082008-09-06
Good afternoon, Ulf.
Back in November of 2007, in comp.arch.embedded, you wrote:
"I see as many ARM9 users as ARM7 users today and neither the CM3 or
the uC3k will meet the performance of the ARM9 and neither will run
Linux/WinCE. - uCLinux? ...sigh!"
I've only started reading about uCLinux over the last couple of days
but I was wondering why you "sigh" at it.
I'm told that the Atmel AVR32UC3A/B (and their associated eval boards,
the EVK1100/1) cannot run Linux because the processor doesn't have an
MMU... hence something like the AVR32-AP7, at a minimum, is needed.
On the other hand, uCLinux is specifically designed to run on
processors without MMUs. So, any reason why uCLinux cannot (or should
not) run on the AVR32UCA/B?
Certainly uCLinux has its limitations but if it's the only flavor of
Linux that will run on processors without an MMU, and the lack of MMU
is the reason why the UC3x can't run Linux, then I would think that a
uCLinux port for the UC3x would be welcomed by low-end processor
developers... would it not?