Re: PATCH: Support x86 pseudo registers

[Pedro Alves <pedro at codesourcery dot com>]

On Friday 12 March 2010 16:04:45, Eli Zaretskii wrote:
> > So, should that sentence of the manual be relaxed?
> 
> Maybe, but frankly I don't really understand what it says, exactly.
> Does it mean that if the name does clash with the architecture, the
> architecture's meaning is used?

Yes.  That's what the code does too, hence this issue.

E.g., on archs with a real "$fp" register, in -fomit-frame-pointer
functions, $fp evals to the contents of the reg (which can
be anything), while on other archs it prints the unwinder's
notion of frame base.

> Anyway, are there any such conflicts in the current codebase?

There was one just now.  :-)

> > I guess this would be a good place to at least mention the x86 $sp
> > is always $esp or $rsp.
> 
> Yes, I think so.

-- 
Pedro Alves