This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug libstdc++/16651] [3.4/3.5 Regression] cross from powerpc-darwin to i686-darwin fails
- From: "mg_gentoo at yahoo dot com" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: 10 Aug 2004 19:12:00 -0000
- Subject: [Bug libstdc++/16651] [3.4/3.5 Regression] cross from powerpc-darwin to i686-darwin fails
- References: <20040720234951.16651.pinskia@gcc.gnu.org>
- Reply-to: gcc-bugzilla at gcc dot gnu dot org
------- Additional Comments From mg_gentoo at yahoo dot com 2004-08-10 19:11 -------
It is (though I can't mark it as such).
Just like in lv's case, the failure is that CHECK_MATH_SUPPORT (as opposed to my
case, where it's CHECK_COMPLEX_MATH_SUPPORT) is called after GCC_NO_EXECUTABLES.
Same bug though.
Should there be a GCC_NO_LINKER separated out from NO_EXECUTABLES to help
prevent this misuse situation? As long as we have the target binutils already,
we should be able to test the linker, afaik. We just cant run the target's
executables. The linker has (should have) the host == gcc's $build in a
non-canadian build. Certainly works here.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16651