This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Result of egcs 1.1.1 prereleases on Linux/x86
- To: robertl at dgii dot com (Robert Lipe)
- Subject: Re: Result of egcs 1.1.1 prereleases on Linux/x86
- From: hjl at lucon dot org (H.J. Lu)
- Date: Mon, 16 Nov 1998 08:17:46 -0800 (PST)
- Cc: egcs at cygnus dot com
>
> > Here is the result of egcs 1.1.1 prereleases on Linux/x86. With -fPIC,
> > I got 6 more failures:
> >
> > FAIL: gcc.c-torture/execute/980506-2.c execution, -O2
> > FAIL: gcc.c-torture/execute/980506-2.c execution, -O2 -g
> > FAIL: gcc.c-torture/execute/980506-2.c execution, -Os
> > FAIL: gcc.c-torture/execute/arith-rand.c compilation, -O2 -g
> > FAIL: g++.pt/friend7.C (test for excess errors)
> > FAIL: g++.pt/friend7.C Execution test
>
> Bizarre. On OpenServer (the ELF target is rather like Linux on x86) I
> don't get these failures on the 1.1.1 prerelease. I do see the 980506-2
> tests fail on the trunk on OSR5. In fact, I see 980506-2 fail in all
> modes (ELF, COFF, and PIC) and not just PIC. The other two aren't
My egcs 1.1.1 is sligtly modified. It may fix the non-PIC case.
> failing for me. Perhaps as or ld is doing something suprising.
>
Ooops.
FAIL: gcc.c-torture/execute/arith-rand.c compilation, -O2 -g
FAIL: g++.pt/friend7.C (test for excess errors)
FAIL: g++.pt/friend7.C Execution test
are false alarms. They just took too long to compile.
H.J.