Re: PR/38359 CVS commit: src/gnu/dist/gcc4/gcc/config/m68k

[Izumi Tsutsui <tsutsui%ceres.dti.ne.jp@localhost>]

mrg%eterna.com.au@localhost wrote:

>     Modified Files:
>       src/gnu/dist/gcc4/gcc/config/m68k: m68k.md
>     
>     Log Message:
>     Pull the follwoing fix from upstream:
>     http://gcc.gnu.org/viewcvs/trunk/gcc/config/m68k/m68k.md#rev117181
>     >       * config/m68k/m68k.md (negsf2, negdf2, negxf2): Use
>     >       -2147483647 - 1 instead of 0x80000000.
>     
>     Fixes "internal compiler error: in do_SUBST" on compiling
>     floating point ops with -msoft-float or -m68010 on LP64 hosts,
>     and may also close PR toolchain/38359.  Tested on alpha.
>
> oh really.  nice find.
> 
> i wonder if this means most of sun2 can be built with
> -O again?  i had disabled it but my build platform is
> amd64 so i was seeing errors _all over the place_ so i
> disabled it more globally.

I had enabled -O1 with some hacks and confirmed it working on i386,
http://mail-index.netbsd.org/source-changes/2008/04/27/msg005366.html
http://mail-index.netbsd.org/source-changes/2008/04/27/msg005367.html
but later I noticed autobuild amd64 and my alpha failed with
the ICE in do_SUBST, then I've tracked the bug on my alpha and
found the fix ;-)

Now sun2 build on alpha fails in libm by SIGFPE of lint, though..
---
Izumi Tsutsui