gmp 4.3.0 failures on Intel Darwin9

[Jack Howarth]

On Thu, Apr 23, 2009 at 07:18:29PM +0200, Torbjorn Granlund wrote:
> Jack Howarth <howarth at bromo.med.uc.edu> writes:
> 
>     I see the same problem without the patch. The stock Apple gcc 4.0.1
>   compiler doesn't understand the -mtune=core2 option so it ends up using
>   -mtune=k8. The Apple gcc-4.2.1 compiler does and configure ends up using
>   -mtune=core2. However both builds yield the same set of testsuite failures
>   (none of which exist in gmp 4.2.1). Do you have access to a Intel based
>   Mac to test gmp 4.3.0 directly?
> 
> A few Macs are used for the automated GMP builds.  One of the Macs
> is a Core2.
> 
> sill$ uname -a
> Darwin sill.swox.se 8.11.1 Darwin Kernel Version 8.11.1: Wed Oct 10
> 18:23:28 PDT 2007; root:xnu-792.25.20~1/RELEASE_I386 i386 i386
> 
> sill$ gcc -v
> [snip]
> gcc version 4.0.1 (Apple Computer, Inc. build 5367)
> 
> Apple's compiler tools fail to impress.  See http://gmplib.org/macos.html
> for a sample of the problems they have caused.  It is possible that
> you've run into another bug.
> 
> Did you remember to clean up your build directory between incompatible
> builds?
> 
> -- 
> Torbjörn

Torbjörn,
    As a general rule, the fink developers had adopted the approach that
only the latest Xcode release for a given Mac OS release has to be
considered. We have depreciated support for Mac OS X 10.3 so that leaves
Xcode 2.5 for Mac OS X 10.4 and Xcode 3.1.2 for Mac OS X 10.5. I would
only worry about bugs against those two. Keep in mind that with Xcode 3.1.2,
you do have the option of requiring gmp to build against the gcc-4.2 compiler
if you want a newer gcc that understands -mtune=core2.
                 Jack
ps The fink build system always cleans its buildroot before a new build.
I'll try the current svn of gmp tonight. If it it fixed, can you point
me to the svn checkin that did that so we can use that as a patch?