Re: toolchain requirements submission

[Dan Kegel <dank at kegel dot com>]

Frank Meerkoetter wrote:
> I had to apply these two patches to get something usable:
>  
>  - The first one configures the toolchain in a way that it generates big
>    endian by default.
>    http://www.davehylands.com/linux/ARM-bigendian/gcc-uclibc-3.3-300-bigend.patch

I do link to one of Dave's bigendian patches from the crosstool page.
I haven't included it because I haven't gone to the trouble of
making sure the bigendianness default is configurable.

>  - The second one fixes a glibc-bug. The ARM optimized version of strlen
>    is brocken on big endian systems.
>    http://lists.arm.linux.org.uk/pipermail/linux-arm-kernel/2003-June/015964.html

That's now in crosstool-0.28-rc25.

> The resulting toolchain is able to compile a working helloworld
> programm. I will do more testing the next days. The main reason why
> we're using the prebuild Montavista toolchain right now is that it was 
> the only toolchain which really(TM) worked for us. We've generated arm-be
> toolchains before but code compiled with these sometimes failed in
> mysterious ways. So i'm quite eager to setup the regression tests.

Excellent!
- Dan

--
My technical stuff: http://kegel.com
My politics: see http://www.misleader.org for examples of why I'm for regime change

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