Adjust primary/secondary platform lists

[Jack Howarth]

   In PR39888, there does exist initial attempts at implementing a libgcc_ext
for the darwin target which would allow us to have access to all of these
additional symbols that aren't exported by libgcc-10.4/10.5. It would solve
the most nagging problem on darwin of adding additional symbols to libgcc_s.
I would also note that as of 10.6, libgcc on darwin is actually subsumed into
libSystem and not actually used. This makes the creation of a libgcc_ext to
contain all of the symbols added after libgcc-10.5 even more critical.
                               Jack
ps I believe that the Darwin maintainers do agree with this approach of
adding libgcc_ext to solve the problem. It would nice if we could manage
to get it into gcc 4.5 and remove any issues on darwin with the patch
mentioned below.

On Tue, Sep 22, 2009 at 10:31:13AM -0700, Neil Vachharajani wrote:
> In light of this change, I was wondering if we could reconsider the
> patch discussed here:
> 
> http://gcc.gnu.org/ml/gcc-patches/2008-12/msg01259.html
> 
> The patch makes certain global variables used during value profiling
> thread local to avoid data races.  The patch was originally applied
> and then reverted due to a regression on Darwin.  Seeing as Darwin
> support is being moved to secondary, I was wondering if that somehow
> changes the status of the patch.  I was also curious if
> targetm.have_tls could be used to guard the modifications in the patch
> to prevent the regressions on darwin in the first place.
> 
> Neil
> 
> On Tue, Sep 22, 2009 at 8:43 AM, Mark Mitchell
> <mitchell@codesourcery.com> wrote:
> >
> > The SC has decided to remove powerpc-apple-darwin from the secondary
> > platform list (due to the fact that Apple is no longer supporting
> > Power in OS X and Apple's overall shift from Power to x86) and to move
> > i686-apple-darwin from the primary to the secondary platform list (due
> > to the lack of active maintenance of GCC, by Apple or otherwise, for
> > x86 OS X).
> >
> > Applied.
> >
> > --
> > Mark Mitchell
> > CodeSourcery
> > mark@codesourcery.com
> > (650) 331-3385 x713
> >
> > Index: gcc-4.5/criteria.html
> > ===================================================================
> > RCS file: /cvs/gcc/wwwdocs/htdocs/gcc-4.5/criteria.html,v
> > retrieving revision 1.1
> > diff -c -5 -p -r1.1 criteria.html
> > *** gcc-4.5/criteria.html       27 Mar 2009 21:08:49 -0000      1.1
> > --- gcc-4.5/criteria.html       22 Sep 2009 15:40:32 -0000
> > *************** application testing.</p>
> > *** 98,108 ****
> >  <p>The primary platforms are:</p>
> >  <ul>
> >  <li>arm-eabi</li>
> >  <li>i386-unknown-freebsd</li>
> >  <li>i686-pc-linux-gnu</li>
> > - <li>i686-apple-darwin</li>
> >  <li>mipsisa64-elf</li>
> >  <li>powerpc64-unknown-linux-gnu</li>
> >  <li>sparc-sun-solaris2.10</li>
> >  <li>x86_64-unknown-linux-gnu</li>
> >  </ul>
> > --- 98,107 ----
> > *************** application testing.</p>
> > *** 111,121 ****
> >
> >  <p>The secondary platforms are:</p>
> >  <ul>
> >  <li>hppa2.0w-hp-hpux11.11</li>
> >  <li>powerpc-ibm-aix5.2.0.0</li>
> > ! <li>powerpc-apple-darwin</li>
> >  <li>i686-pc-cygwin</li>
> >  <li>i686-mingw32</li>
> >  <li>ia64-unknown-linux-gnu</li>
> >  <li>s390-linux-gnu</li>
> >  </ul>
> > --- 110,120 ----
> >
> >  <p>The secondary platforms are:</p>
> >  <ul>
> >  <li>hppa2.0w-hp-hpux11.11</li>
> >  <li>powerpc-ibm-aix5.2.0.0</li>
> > ! <li>i686-apple-darwin</li>
> >  <li>i686-pc-cygwin</li>
> >  <li>i686-mingw32</li>
> >  <li>ia64-unknown-linux-gnu</li>
> >  <li>s390-linux-gnu</li>
> >  </ul>
> >
> 
> 
> 
> -- 
> Neil Vachharajani
> Google
> 650-214-1804