From mboxrd@z Thu Jan  1 00:00:00 1970
From: Mumit Khan <khan@xraylith.wisc.edu>
To: egcs@cygnus.com
Subject: Re: Test result for 970814 on native sparc-sun-solaris2.5.1
Date: Tue, 19 Aug 1997 02:36:07 -0000
Message-id: <9708190234.AA05546@modi.xraylith.wisc.edu>
In-reply-to: m0x0cn4-0004ecC@ocean.lucon.org
X-SW-Source: 1997-08/0123.html

hjl@lucon.org (H.J. Lu) writes:
> > 
> >   In message <9708171708.AA03535@modi.xraylith.wisc.edu>you write:
> >   > btw, static template data members are not being instantiated breaking S
> GI
> >   > STL in libg++-2.8.0b6 (the allocator in particular).
> > I assume you're referring to the libstdc++ tests right?  Do you get an
> > error like:
> > 
> > /usr/ccs/bin/ld: Unsatisfied symbols:
> >    __default_alloc_template<false, 0>::end_free(data)
> >    __default_alloc_template<false, 0>::heap_size(data)
> >    __default_alloc_template<false, 0>::free_list(data)
> >    __default_alloc_template<false, 0>::start_free(data)
> > 

Yup. I instantiated these by hand just to test things, but then tmap
core dumped on me with '-g -O2'. The only time tmap works is when I
don't specify '-O*'. 

btw, the instantiations *are* there for i386-linux-gnulibc1, but not
for hppa1.1-hp-hpux10.20. "tmap" dies on either platform with -O*.

Mumit