Re: [PATCH] fixes to consistency.vlad/vlad.exp

[Phil Edwards <phil at jaj dot com>]

On Mon, Aug 26, 2002 at 02:02:50PM -0400, Daniel Jacobowitz wrote:
> On Mon, Aug 26, 2002 at 01:53:06PM -0400, Phil Edwards wrote:
> > On Mon, Aug 26, 2002 at 07:15:56AM -0700, Janis Johnson wrote:
> > 
> > > We apparently need a set of tests that check that basic C++ runtime
> > > support is available outside of libstdc++.
> > 
> > Either that, and link against libsupc++, or move these tests into the
> > library's testsuite.  Hmmmm, but that makes consistency.vlad more complex.
> 
> I wonder - should "gcc -x c++" use "-lsupc++" the way that a plain
> "g++" would?

Dunno about "should".  It /could/ if the right -L option were passed.
Or just a path to the libsupc++.a file directly, since most (all?) linkers
will DTRT there.

The rule in general is to test language functionality in the compiler
testsuite, and library functionality in the library testsuite, and rarely
shall the twain meet, e.g., tweaks for handling of weak symbols are carefully
seperated.  Support for vtable could go into either category anyhow.

Speaking only for myself, I would be willing to make exceptions for this
rule in the case of consistency.vlad, given that it's not part of the usual
"make check" runs.


Phil

-- 
I would therefore like to posit that computing's central challenge, viz. "How
not to make a mess of it," has /not/ been met.
                                                 - Edsger Dijkstra, 1930-2002