On Thu, Jan 29, 2015 at 07:16:17PM +0100, Matthias Klose wrote:
> On 01/29/2015 07:12 PM, H.J. Lu wrote:
> > On Thu, Jan 29, 2015 at 10:05 AM, Matthias Klose <[email protected]> wrote:
> >> that fixes the build failure. ok to commit?
> >>
> >> 2015-01-29 Matthias Klose <[email protected]>
> >>
> >> * acinclude.m4 (GLIBCXX_ENABLE_VTABLE_VERIFY): Define VTV_CYGMIN
> >> unconditionally.
> >> * configure: Regenerate.
> >>
> >
> > This is wrong. You are checking vtv_cygmin before it is defined.
>
> my bad, somebody should obviously build both variants ... now testing this
> variant.
Even with this patch vtv_cygmin is uninitialized. It will be likely empty
string, unless somebody has vtv_cygmin in the environment, but it is better
to initialize it, see the patch I've posted.
> --- acinclude.m4 (revision 220257)
> +++ acinclude.m4 (working copy)
> @@ -2320,8 +2320,6 @@
> AC_MSG_CHECKING([for vtable verify support])
> AC_MSG_RESULT([$enable_vtable_verify])
>
> - AM_CONDITIONAL(VTV_CYGMIN, test $vtv_cygmin = yes)
> -
> if test $enable_vtable_verify = yes; then
> case ${target_os} in
> cygwin*|mingw32*)
> @@ -2341,6 +2339,8 @@
> VTV_CXXLINKFLAGS=
> fi
>
> + AM_CONDITIONAL(VTV_CYGMIN, test $vtv_cygmin = yes)
> +
> AC_SUBST(VTV_CXXFLAGS)
> AC_SUBST(VTV_PCH_CXXFLAGS)
> AC_SUBST(VTV_CXXLINKFLAGS)
>
Jakub
