Re: 2.2.8 testsuite failures on mac os x 10.6.3 with gcc-4.5

[Peter O'Gorman]

On 06/06/2010 04:05 AM, Ralf Wildenhues wrote:

> > Do we really want to test this with no object files?

> I think the workaround we used at the time for compilers like this was
> to empty whole_archive_flag_spec, although cheaper workaround could in
> principle be used (add dummy object, extract only the first convenience
> archive).  It bothers me to see that GCC reintroduces this bug, however.

Ugh, well, I am not going to remove the -force_load support from 
libtool, it's the best way to get debugging working with convenience 
archives. I'll look at adding a dummy object if there is no other object 
on the link line (but I don't think it's urgent for an OS that does not 
ship with an affected compiler).

Looks like gfortran has always done this, but on linux --whole-archive 
foo.a --no-whole-archive does not need to be -Wl, quoted, so gfortran 
sees the foo.a and counts it as an input file. I'll ask on the gcc list 
about this later.

I have a patch for gcj that appears to work, waiting to ensure it 
doesn't cause regressions before posting it to gcc-patches.

On another topic - do we care about test failures that only happen with 
SElinux in enforcing mode (at least tests/demo-exec.test fails on Fedora 
13)?

Thanks,
Peter