[m68k 11/13] fix symbolic address recognition during reload
Jeffrey Law
law@redhat.com
Mon Feb 5 17:49:00 GMT 2007
On Tue, 2007-01-30 at 14:56 +0100, Roman Zippel wrote:
> Hi,
>
> On Tue, 30 Jan 2007, Richard Sandiford wrote:
>
> > I'm not sure about this. We want to load the symbol or label
> > and add the offset separately, don't we? That's certainly
> > what the pre-reload code does.
> >
> > Can you give a few more details about how this fails? How do
> > we end up with (plus pic_offset_table_rtx (const (plus ...)))?
>
> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=385327
>
> The additional offset is generated later by converting the SI argument to
> a QI argument, but I have to check the exact details again myself.
> Note that all mentioned bugzilla references are different bugs.
I also wonder if we should be trying to prevent reload from changing
the mode of the argument from SI to QI mode -- it seems like we are
having numerous problems when arguments get squished down to QImode.
I thought we had a macro to prevent this kind of behavior.
Jeff
More information about the Gcc-patches
mailing list