When emerging libperl on an amd64 system without +multilib, libperl fails the compile phase with the following error: `sh cflags "optimize='-O0 -march=nocona'" doio.o` -fPIC doio.c CCCMD = x86_64-pc-linux-gnu-gcc -DPERL_CORE -c -fno-strict-aliasing -pipe -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -O0 -march=nocona -Wall doio.c:32: error: conflicting types for 'shmat' /usr/include/sys/../gentoo-multilib/default/sys/shm.h:58: error: previous declaration of 'shmat' was here doio.c:32: error: conflicting types for 'shmat' /usr/include/sys/../gentoo-multilib/default/sys/shm.h:58: error: previous declaration of 'shmat' was here make: *** [doio.o] Error 1 !!! ERROR: sys-devel/libperl-5.8.5 failed. !!! Function src_compile, Line 218, Exitcode 2 !!! Unable to make libperl.so !!! If you need support, post the topmost build error, NOT this status message. Adding -DHAS_SHMAT_PROTOTYPE to the CFLAGS solves the issue. # cat /etc/portage/bashrc ([[ "$EBUILD_PHASE" = "compile" ]] && [[ $PN = "libperl" ]]) && export CFLAGS="${CFLAGS} -DHAS_SHMAT_PROTOTYPE" Reproducible: Always Steps to Reproduce: 1. 2. 3.
Perl, additionaly, has an identical problem. `sh cflags "optimize='-O0 -march=nocona'" doio.o` -fPIC doio.c CCCMD = gcc -DPERL_CORE -c -fno-strict-aliasing -pipe -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -O0 -march=nocona -Wall doio.c:32: error: conflicting types for 'shmat' /usr/include/sys/../gentoo-multilib/default/sys/shm.h:58: error: previous declaration of 'shmat' was here doio.c:32: error: conflicting types for 'shmat' /usr/include/sys/../gentoo-multilib/default/sys/shm.h:58: error: previous declaration of 'shmat' was here make: *** [doio.o] Error 1 !!! ERROR: dev-lang/perl-5.8.5-r5 failed. !!! Function src_compile, Line 244, Exitcode 2 !!! Unable to make !!! If you need support, post the topmost build error, NOT this status message.
I had the same failure on an amd64 non-multilib system. (this error is new)
I am having the same issue on a multilib system. Also tried upgrading and downgrading libperl and get same errors.
Is this still a problem? This is probably related to bug #87560. You may need to re-emerge linux-headers and/or glibc to sort your headers out.
Not a problems for me anymore.
As far as I know amd64 just p.masked this version.
I'm unable to reproduce this and nobody has responded to say they still have the problem so I'm marking this fixed. Re-open if the problem re-emerges.
Confirmed fixed on a systems where it was showing up before.