V3 PATCH: Use -rpath, not --rpath, when testing

Phil Edwards pedwards@disaster.jaj.com
Mon Oct 30 07:06:00 GMT 2000

On Mon, Oct 30, 2000 at 02:42:12AM -0200, Alexandre Oliva wrote:
> On Oct 30, 2000, Mark Mitchell <mark@codesourcery.com> wrote:
> > Since you're a libtool expert, would you mind putting together a patch
> > for libstdc++-v3/mkcheck.in?
> I'd only mind if you're in a hurry to get that in.  I'm in quite a
> tight schedule right now, and I don't expect to have time to look into
> this stuff this week :-(

Since I have been making huge rewrites to mkcheck.in over the last few days
(haven't checked them in since one part is still broken), I can take a
shot at this... after finishing up three days of email and the wwwdocs stuff.

One of the changes, btw, was switching --rpath to -R, but of course that
doesn't work on IRIX.

> > (I don't think we really need *everything* static; just libstdc++ if
> > that's easy to arrange.)
> It isn't, unless you break libtool's encapsulation and look for
> libstdc++.a.  I don't see any value in testing both static and shared
> versions; I'd rather just defer to whatever choices the builder makes
> at configure time (i.e., --enable or --disable-shared).

Another of my changes was for configure.in to SUBST the enable_shared and
enable_static results, and for mkcheck.in to be aware of them and only
test the libraries which are enabled (both, of course, by default).


pedwards at disaster dot jaj dot com  |  pme at sources dot redhat dot com
devphil at several other less interesting addresses in various dot domains
The gods do not protect fools.  Fools are protected by more capable fools.

More information about the Gcc-patches mailing list