Segmentation fault in __uClibc_main on m68k

[Rob Landley]

On Sun, May 31, 2015 at 2:14 AM, Rich Felker <dalias at libc.org> wrote:
> On Sat, May 30, 2015 at 01:28:42PM +0300, Max Filippov wrote:
>> On Sat, May 30, 2015 at 1:12 PM, Rob Landley <rob at landley.net> wrote:
>> > I see a qemu-system-xtensa so might give that a try too.
>>
>> Please do. I'd be happy to answer any questions.
>> I also hope to port musl to xtensa this year.
>
> Would you like to be involved in binfmt/toolchain discussions going on
> now? My plan for nommu support with musl is to start with plain ELF,
> ET_DYN (PIE) form only, with both static and dynamic linking
> supported. This gives behavior identical to MMU-ful archs, but of
> course with no text sharing. Then we can add FDPIC as needed to get
> shared text; this requires both toolchain support and library/dynamic
> linker runtime changes in musl. With musl being something of a clean
> slate and aiming for minimal arch-specific code, I really don't want
> to carry forward legacy binfmts and associated support code, but I do
> want to understand if there are any particular properties of existing
> binfmts/ABIs/etc. that are important to developers/users, and I'd
> like to hear if you have any input on this from the Xtensa side.
>
> Rich

We have Jeff Dionne, founder of uclinux and developer of uc-libc
(which Erik forked to make uClibc, do "git log | cat" and look at the
next to last commit) contributing to the musl-sh2 discussion. I'm trying to
get the thread restarted and the existing messages posted to the musl
list so they're archived somewhere.

Jeff's CEO of a tech start in Japan these days so _really_ busy, but
still brilliant and the world's leading expert on a lot of nommu stuff
if you can get his time.

Not that I'm exactly unbiased here. He's my boss at $DAYJOB and I'm
appearing on a panel with him and the original superh processor
architect at linuxcon japan on thursday:
http://lccojapan2015.sched.org/event/860288ccda595208a5d7337d31c6075c#.VWv3sR-1XVN

Still, _that_ is the discussion you want to get in on...

Rob