bug#23348: [m4/libtool.m4] linker error for powerpc64le-unknown-linux-gnu
Sam James <[email protected]> Sun, 14 Jan 2024 03:42:48 +0000
| Newsgroups | gmane.comp.gnu.libtool.bugs |
|---|---|
| Organization | Gentoo |
| Message-ID | <[email protected]> |
Mike Frysinger <[email protected]> writes: > [[PGP Signed Part:Undecided]] > On 24 Apr 2016 00:46, Leno Hou wrote: >> 3) But when I set CHOST="powerpc64le-unkown-linux-gnu" in Gentoo and >> execute aclocal, automake, libtoolize --copy --force, autoconf. When I >> glance in the latest m4/libtool.m4 : > > i took another look at this patch. it isn't making sense to me, and your > commit message doesn't have an explanation as to why you need it. the only > thing you've written is "i looked at the file, and it looks incorrect". you liked it before! https://bugs.gentoo.org/581314 ;) > > do you have a build failure/log somewhere to share ? > >> if AC_TRY_EVAL(ac_compile); then >> case `/usr/bin/file conftest.o` in >> *32-bit*) >> case $host in >> ... >> powerpc64le-*linux*) >> LD="${LD-ld} -m elf32lppclinux" >> ;; >> powerpc64-*linux*) >> LD="${LD-ld} -m elf32ppclinux" >> ;; >> ... >> *64-bit*) >> case $host in >> ... >> powerpcle-*linux*) >> LD="${LD-ld} -m elf64lppc" >> ;; >> powerpc-*linux*) >> LD="${LD-ld} -m elf64ppc" >> ;; >> ... >> >> seems linked to $LD='$LD-ld -m elf64ppc' and failed to compile packages. > > this code is really for multilib situations. it's only needed when you want > to compile for a diff ABI than the toolchain is defaulting to. > > so if --host=powerpc64le-linux-gnu, it should produce 64-bit binaries, and > libtool doesn't have to do anything. that's why powerpc64 isn't listed in > the 64-bit section. > > adding it here would only matter if your powerpc64le-linux-gnu is producing > 32-bit binaries by default, and that doesn't make sense. and as you can > see from the arch list in this section, no other target does this. > -mike > > [[End of PGP Signed Part]]