This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Proposed targets to deprecate for 3.4
- From: Hans-Peter Nilsson <hp at bitrange dot com>
- To: Zack Weinberg <zack at codesourcery dot com>
- Cc: Nathanael Nerode <neroden at twcny dot rr dot com>, <gcc at gcc dot gnu dot org>
- Date: Wed, 1 Oct 2003 03:11:00 -0400 (EDT)
- Subject: Re: Proposed targets to deprecate for 3.4
On Tue, 30 Sep 2003, Zack Weinberg wrote:
> Hans-Peter Nilsson <hp@bitrange.com> writes:
> > g Port does not define TARGET_ASM_FUNCTION_(PRO|EPI)LOGUE.
> >
> > See mmix.c.
>
> There has got to be a better way to do that.
Perhaps a (new) TARGET_ASM_FUNCTION_BEGIN_PROLOGUE? If you're
opposed to TARGET_ASM_FUNCTION_END_PROLOGUE, look at alpha.
Hmm, maybe both uses could be replaced with unspec_volatile
insns. Can the scheduler still move insns past an
unspec_volatile?
> Incidentally, must we have eval: (c-set-style "gnu") at the bottom of
> files? It just makes emacs slow to visit them (since it has to pause
> and show the eval: to the user).
I think of that eval as a convenience. Maybe it isn't, and I
should fold chosing the mode into code in a mode-hook that looks
at the directory context. Do you visit those files often?
> > These would be nice to have in simtest-howto.html, but without
> > any interpretation of value.
>
> Patches are welcome.
Yes they are. :-)
> >> l Port can use ILP32 mode integer arithmetic.
> >> q Port can use LP64 mode integer arithmetic.
> >
> > Aren't those two exclusive in a "non-r" port?
>
> No. They might be switchable at compile time only.
Oh, right, you're picking on the CPU part of the ports, not the
port as a whole. Gotcha.
brgds, H-P