Path: utzoo!telly!ddsw1!mcdchg!rutgers!labrea!mcnc!xanth!uunet!mstan!frank
From: frank@Morgan.COM (Frank Wortner)
Newsgroups: gnu.gcc.bug
Subject: Re: problem with making gcc on Sun4/SunOs 4.0
Message-ID: <111@hudson.Morgan.COM>
Date: 10 Nov 88 16:32:59 GMT
References: <8811091516.AA14509@sunbar.mc.duke.edu> <468@dmk3b1.UUCP>
Reply-To: frank@Morgan.COM (Frank Wortner)
Distribution: gnu
Organization: Morgan Stanley and Co., NY, NY
Lines: 30

In article <468@dmk3b1.UUCP> dmk@dmk3b1.UUCP (David Keaton) writes:
>In article <8811091516.AA14509@sunbar.mc.duke.edu> ndd@SUNBAR.MC.DUKE.EDU (Ned Danieley) writes:
>)in trying to make gcc on a 4/110 running SunOs 4.0, I received a
>)signal 6 from cc1 when compiling toplev.c:
>)
>)make CC=stage1/gcc CFLAGS="-g -O -Bstage1/"
>)(other subroutines)
>)stage1/gcc -g -O -Bstage1/ -sun4 -c  toplev.c
>)stage1/gcc: Program cc1 got fatal signal 6.
>)*** Error code 1
>
>Yes, it's a bug, but the bug is Sun's.  They have changed the meaning
>of the -B flag to cc.  Their equivalent of "-Bstage1/" would be
>"-Qpath stage1".

I'm afraid you are incorrect.  The bug is *not* in Sun's C compiler.
Notice that the error comes from *gcc*, not *cc*.   While you are
correct in stating that the meaning of the '-B' flag of the Sun
compiler is different, you are not correct in attributing the failure
of gcc to execute correctly to this defect.  (Whether Sun was correct
in changing -B to -Q is another story ;-))

The documentation for gcc 1.30 states that the code generator for the
sparc architecture has bugs and will be fixed in the next release.  I
think that is the cause of Ned's problems.

-- 
						Frank

"Computers are mistake amplifiers."