This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
bootstrap/6598: configure identifies old/broken gcc as gcc and passes unknown options
- From: sdack at anitesystems dot de
- To: gcc-gnats at gcc dot gnu dot org
- Date: 7 May 2002 18:02:46 -0000
- Subject: bootstrap/6598: configure identifies old/broken gcc as gcc and passes unknown options
- Reply-to: sdack at anitesystems dot de
>Number: 6598
>Category: bootstrap
>Synopsis: configure identifies old/broken gcc as gcc and passes unknown options
>Confidential: no
>Severity: serious
>Priority: medium
>Responsible: unassigned
>State: open
>Class: sw-bug
>Submitter-Id: net
>Arrival-Date: Tue May 07 11:06:01 PDT 2002
>Closed-Date:
>Last-Modified:
>Originator: Sven Dack
>Release: gcc-3.0.4
>Organization:
>Environment:
m88k-motorola-sysv4
>Description:
Motorola ships their boxes with SVR4 and a master piece of compiler suite which is basically a GNU cc with their own assembler and linker but a broken 'cc' frontend. The compiler identifies itself as the GNU compiler through the builtin defines but doesn't accept all command line options. They even scratched --help and --version :-(. '-g' and '-O' works, -Wall doesn't which gets complained about during the bootstrap.
Now the problem is, that the usual configure script identifies this as a gcc but I think it rather shouldn't.
It actually is a gcc-2.2.2 or 2.4.5. Not very clear. It does compile things.
>How-To-Repeat:
Apply for a job at our company and make gcc run on m88k.
>Fix:
Have a configure option which disables checking of 'cc' beeing a GNU cc (and rather assume the worst).
>Release-Note:
>Audit-Trail:
>Unformatted: