[prev in list] [next in list] [prev in thread] [next in thread] 

List:       racket-users
Subject:    [plt-scheme] Unable to run the 103p1 version on SuSE 10
From:       mflatt () cs ! utah ! edu (Matthew Flatt)
Date:       2006-04-28 12:00:25
Message-ID: 20060428120012.CD22865007A () mail-svr1 ! cs ! utah ! edu
[Download RAW message or body]

At Fri, 28 Apr 2006 09:24:41 +0200, "Jean-Baka Domelevo-Entfellner" wrote:
> I had compiled it on Fedora Core with an old gcc (3.2), to avoid too
> many error messages you get using 4.x flavours. On my SuSE platform, I
> do not have such an old compiler, but I have access to a 3.3.6, which
> I used to compile some parts of the software (meaning "some
> directories"). Problems arise particularly in the wxxt subdirectory,
> because of some obsolete syntax. So the compiling process is somewhat
> untidy, but at the end I got it all compiled.
> And even if I could do more uniform a compilation (same version of gcc
> for all, same options), in my mind as soon as you compile it all
> without errors, the sofware should work, uh?

Sorry - I'm a bit unclear on the conclusions.

I think you're using versions 4.x and 3.3.6 of the compiler, and
specifically version 4.x for the MzScheme part. If that's right, I
wonder whether using 3.3.6 for the MzScheme part would help.

Also, it's not clear to me whether you tried removing -O2 from the
makefile in <builddir>/mzscheme/src/Makefile, and whether that helped.

> P.S. How do you know the -O2 option is unsupported ? Can't this be
> removed by the configuree process ?

Removing -O2 might work around old code with bad aliasing (where "bad"
means "aliasing that the C99 specification says a compiler can
ignore"). Really, it's the -fstrict-aliasing part of -O2 that I think
it may be helpful to remove. Having -O2 imply -fstrict-aliasing is a
relatively new feature of gcc.

At the time that v103p1 was created and packaged, removing -O2 was not
necessary or helpful, so that's why `configure' didn't remove it. Of
course, I'm still not sure whether it helps, but let me know.

Thanks,
Matthew



[prev in list] [next in list] [prev in thread] [next in thread]