Re: [avr-gcc-list] is tiny15 not c-programmable

[Joerg Wunsch]

"Bruce D. Lightner" <address@hidden> wrote:

Bruce,

others have already commented on your poor communication skills, so i
can just skip that here.

> I've been using "avr-gcc" to program the Atmel ATmega15L
> microcontroller for almost a year.

That's good to hear.

Your description sounds a bit `hacky', but is certainly doable.  I
wouldn't consider it to be ``gcc /ported/ to the RAM-less AVRs''
though.  (But you didn't claim that, OK.)

> (Unlike Joerg "IMHO" Wunsch, you will need to actually look at the
> assembly language listing file. [Ref: Joerg Wunsch email, Re:
> [avr-gcc-list] Help from a veteran?, Mon, 31 Mar 2003 10:00:05 +0200
> (MET DST)]

It's great how well you can misunderstand me, in particular since i'm
always telling people to look at the generated assembly code.  It's
only that i find these assembler listings bloated and unreadable, and
much prefer to read the code generated by the compiler instead (i. e.
the .s file, not the .lst file).  Also knowing the AVR opcodes is
usually not getting you really more information (and this is the only
value added by the .lst file).

> As for Joerg "IMHO" Wunsch's pronouncement:
> 
>    Nobody considered the effort worth a while to port gcc to an
>    architecture that doesn't even have RAM.  It might even be
>    impossible to do it.

> He once again has got it half-right!  The big volume (and in fact
> the big dollars) is in the cheap parts, like the ATtiny15.

I didn't speak about marketing volume, but about that nobody so far
volunteered to port gcc to these devices.  This includes yourself,
even though you gave some nice hints, it's far away from what i'd
consider a good port of gcc to them.  Of course, you're quite welcome
to contribute that port, why do you hesitate?  Are you waiting for a
job to be ``assigned'' to you?  That's not the way the opensource
community works.  If you want it, you need to start it yourself.

> As for it being "impossible", that's just another typical
> "Joerg-ism".

Can you even read English?  Unlike for me, i guess it's your native
language.  I wrote ``it might even be impossible''.  This is based on
the difficulties i've seen in porting gcc to the AVR itself already,
simply because gcc's design obviously never considered to apply to
Harvard architectures.  Judging from that, i have expressed my doubts
that it will be easily possible to adapt it to a RAM-less device.
Well, i didn't think of the option to divide the register block into a
block of usable `RAM', and one to be used as regular registers.  That
might solve the problem, but then, someone must be willing to jump
through the hoops of actually doing such a port.

Without someone volunteering, i don't see it happen.  I've been
volunteering to teach GNU binutils to emit yet another obsolescent
object file format (AVR COFF), but i won't be the person to touch gcc
except maybe for a minor bug report.  I don't see Denis and Marek
shouting ``Hurray'' either, so unless someone steps forward to do it,
it won't happen...
-- 
J"org Wunsch                                           Unix support engineer
address@hidden        http://www.interface-systems.de/~j/