Re: [3.0] Bootstrap failure - ASM_SPEC conflict with gas

[Matt Schalit <mschalit at pacbell dot net>]

Philip Blundell wrote:
> 
> >I solved this problem by installing binutils 2.11,
> >which simply ignores any -Q option sent to it, unlike
> >2.10.1, which errored out.
> >
> >Once again, gcc-bugs was no help at all.
> 
> Since there was no bug in gcc in this instance, is that so surprising?
> 
> p.


Hi Phil,

Thanks for the reply.

Let me see if I can think this out for Uw7.

  1)  Gcc sends -Qy to gas.
  2)  Gas didn't have a -Qy option, ever, until this month.
  3)  Gcc sends an option to gas that was not valid for years.

  4)  You claim that's not a gcc bug.


I'm not sure I see your reasoning.



It seems clear that gcc was tuned so that it would only 
compile with SCO's native as, which has a -Qy option.

It seems clear that gas's setting -Qy to "ignore" is there
to workaround bad source code.

I'm just lucky that gcc happens to be the bad source code that
gets the benefit of the excellent work being done by the binutils
crew.

Questions to the binutils list get crushed by avid programmers.
Questions to the gcc-bugs list get ignored more than 50% of the time.
Those who resort to GNATS get even less help.  Some groups work hard,
though, on their GNATS reports, like libstdc++-v3.  That's appreciated.
Those are just some observations for a year of lurking.

Regards,
Matthew