target/1613&Body=http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view&pr=1613&database=gcc

[Hallvard B Furuseth]

The following reply was made to PR target/1613; it has been noted by GNATS.

From: Hallvard B Furuseth <h.b.furuseth@usit.uio.no>
To: gcc-gnats@gcc.gnu.org, gcc-prs@gcc.gnu.org, rodrigc@gcc.gnu.org, gcc-bugs@gcc.gnu.org, nobody@gcc.gnu.org, soka@line.co.jp
Cc:  
Subject: Re: target/1613&Body= http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view&pr=1613&database=gcc
Date: Thu, 30 Aug 2001 21:52:30 +0200

 Hi,
 
 I just got a gcc-3.0.1 ICE in the same place (emit-rtl.c:542) on
 sparc-sun-solaris2.8 (ultrasparc) with this code:
 
 void foo(a, b, c, d, e, f, g, h, i, j, k, l, m, n, o)
   double a, b, c, d, e, f, g, h, i, j, k, l, m, n, o;
 {
   bar(a,b,c,d,e,f,g,h,i,j,k,l,m,n,o, a+b+c+d+e+f+g+h+i+j+k+l+m+n+o);
 }
 
 $ gcc3 -O2 -mcpu=ultrasparc -S -xcpp-output Bug.i 
 Bug.i: In function `foo':
 Bug.i:5: Internal compiler error in gen_reg_rtx, at emit-rtl.c:542
 
 
 However, foo.c doesn't crash gcc-3.0.1 on solaris.
 I can't test it linux now; maybe the original submitter can.
 
 PR 2099 is another case, this one for gcc version 3.1 20010226
 (experimental) on ultrasparc.  It crashes gcc-3.0.1 -O3 -mcpu=ultrasparc
 too.  It does not crash with -O2 instead of -O3.
 
 My gcc (which bootstrapped fine, with CC=gcc-2.95.3):
 
 host: sparc-sun-solaris2.8
 build: sparc-sun-solaris2.8
 target: sparc-sun-solaris2.8
 configured with: ../gcc-3.0.1/configure --prefix=/usit/bombur/hbf --program-suffix=3 --enable-version-specific-runtime-libs --enable-languages=c --quiet
 
 -- 
 Hallvard