On Mon, 24 Apr 2000, Jason R Thorpe wrote:

> When the port to the ARM was originally done, it looks like Mark and/or
> Neil cheated (just a little! :-) and relaxed the alignment constraints in
> the compiler (arm/netbsd.h I guess is where this was done).
> 
> Did you also relax this in the ELF toolchain?

That's how it's done at present, yes.  I more-or-less copied arm/netbsd.h,
and only changed those things that were obviously wrong.  The patch is at
<URL:http://cvs.tartarus.org/netbsd-arm26/sys/arch/arm26/patches/gcc-2.95.2>
or thereabouts, if you're interested.

> It's pretty important because, if you didn't:
> 
> 	- We have to make absolutely sure that all protocol headers
> 	  use the __packed__ attribute.

Do other ports do the same thing, then?  I would expect that structure
alignment on the Alpha, for instance, would be a lot more critical than on
the ARM.

> 	- We have to allocate a new magic number, since binaries are
> 	  now using a different ABI.

Different from what?  Aren't ELF and a.out magic numbers distinct anyway,
or is this something different?

-- 
Ben Harris
Unix Support, University of Cambridge Computing Service.