This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: patch to improve current_function_sp_is_unchanging accuracy
- To: Richard Henderson <rth at cygnus dot com>
- Subject: Re: patch to improve current_function_sp_is_unchanging accuracy
- From: Jeffrey A Law <law at cygnus dot com>
- Date: Thu, 05 Nov 1998 23:04:08 -0700
- cc: John Wehle <john at feith dot com>, egcs-patches at cygnus dot com
- Reply-To: law at cygnus dot com
In message <19981105134430.A17842@dot.cygnus.com>you write:
> On Thu, Nov 05, 1998 at 01:59:02PM -0700, Jeffrey A Law wrote:
> > > * i386.md (call_value_pop): If we're not popping anything,
> > > defer to call_value.
> > > (call_pop): Likewise defer to call.
> > Shouldn't those code fragments be after the flag_pic handling?
>
> That happens in call_value -- there's a bit of code duplication
> between these four patterns.
Ah. That was my only concern. Seems like the right thing to do to me.
It's a lot easier to optimize when we generate reasonable code during the
tree->rtl translation :-) :-)
jeff