Re: Cross Compiler question (the essence of a Canadian CrossCompiler)

["J. Robert Wyatt" <jamwyatt at cisco dot com>]

The liberals, conservatives, and NDP were the party names when this term was
likely coined ... Since I left, they renamed a few things ;-)

So, that being said, if I wanted to use crosstool for this second compiler,
what is the "preferred" way to have crosstool do the work of generating the
host=target compiler? Would that be setting an environment variable like
"HOST" to be equal to the same as the target that was generated in the first
build? As well, I guess that the "RESULT_TOP should be named different too
to avoid overwritting the first build?

Thanks,
Robert


On 8/3/05 7:01 PM, "Dan Kegel" <dank@kegel.com> wrote:

> J. Robert Wyatt wrote:
>> I'd like to build a gcc compiler on a i686 system for ARMv5b that runs on
>> the ARMv5b. I've got the corsstool-0.37 to build the armv5b compiler that
>> runs on i686, but the next step is to use that compiler to build a compiler
>> that can run natively on the ARM platform (Linux 2.4.21 if that is of
>> interest). This last step is confusing since I don't see the three step
>> requirement.
>> 
>> Basically, the --build is the i686 system and the --host and --target are
>> one in the same ... want a compiler that runs on ARM to generate for ARM. Is
>> this essentially a Canadian Cross compiler? If so, do I really have to build
>> three compilers, when the first two are really the same?
> 
> You're not really building a canadian cross,
> since your host and target are the same.
> (Canada had three political parties, I think, is where the name came from.)
> So you can get away with building just two compilers.
> 

------
Want more information?  See the CrossGCC FAQ, http://www.objsw.com/CrossGCC/
Want to unsubscribe? Send a note to crossgcc-unsubscribe@sources.redhat.com