Re: config.sub trickeries (WAS: c4x - TMS320C30-40)

["Svein E. Seldal" <Svein dot Seldal at solidas dot com>]

Joel Sherrill wrote:
> bash-2.05$ sh -x /tmp/j1
> + ./config.sub tic4x-coff
> tic4x-unknown-coff
> + ./config.sub tic4x-elf
> tic4x-unknown-coff
> + ./config.sub tic4x-rtems
> tic4x-unknown-coff
> + ./config.sub c4x-coff
> tic4x-unknown-coff
> + ./config.sub c4x-elf
> tic4x-unknown-coff
> + ./config.sub c4x-rtems
> tic4x-unknown-coff

That's because you have not specified a company. First of all, IMHO 
these two must work:

> ./config.sub c4x
tic4x-unknown-coff
> ./config.sub tic4x
tic4x-unknown-coff

They need to do this because 1) the c4x alias to tic4x is recoginzed, 2) 
coff output is specified. Your propsed patch does not reply correctly 
for the c4x target.

However, given that you take away the star in line 925, this will work 
(as long as you specify a company (of unknown));

> ./config.sub c4x-unknown-elf
c4x-unknown-elf
> ./config.sub c4x-unknown-rtems
c4x-unknown-rtems

I have been looking at your patch, and it doesnt solve the problems for 
this target. I think I know what you are trying to do, but the patch 
breaks the compilation of the binutils, because its reported target 
triplet is "tic4x-unknown-none", which BFD does not understand. If we 
are going use the patch, bfd must be told that "none" equals "coff" - 
and this would be a hack.

I'm openminded to another solution as long as these conditions apply:

- "./config.sub tic4x" gives "tic4x-unknown-coff" *and* "./config.sub 
c4x"  gives "tic4x-unknown-coff"

I dont know how to accomplish yours and my wishes.. They seems to be 
mutally exclusive. :(


Svein