re: sparc64 vs. gcc 4.8

> One other way to go about it: compare the individual object files
> produced by each compiler before they're linked into a monolithic
> kernel. Wouldn't one expect most of them to be identical? Or is GCC code
> generation between 4.5 and 4.8 so different that they'll all change?

i would be *very* surprised if any file was the same.  except maybe
the .s and .S files :-)

> The object files that differ from the same source but different
> compiler versions would, one hopes, constrain the bug search space.

this is true, and i haven't checked, but i didn't consider it a
real possibility.


.mrg.