Am Sat, 4 Oct 2014 15:58:48 -0700 Mark Johnston <markj_at_freebsd.org> schrieb: > On Sat, Oct 4, 2014 at 3:40 PM, Mark Johnston <markj_at_freebsd.org> wrote: > > On Sat, Oct 04, 2014 at 04:39:37PM -0400, Ryan Stone wrote: > >> On Sat, Oct 4, 2014 at 2:33 PM, Mark Johnston <markj_at_freebsd.org> wrote: > >> > On Sat, Oct 04, 2014 at 07:47:56PM +0200, O. Hartmann wrote: > >> >> Recent sources (Revision: 272529) fail to compile: > >> >> > >> >> [...] > >> >> cc -m32 -march=native -DCOMPAT_32BIT -isystem /usr/obj/usr/src/lib32/usr/include/ > >> >> -L/usr/obj/usr/src/lib32/usr/lib32 -B/usr/obj/usr/src/lib32/usr/lib32 -O2 -pipe > >> >> -O3 -O3 -pipe -DYP -I/usr/obj/usr/src/lib32/usr/include/rpcsvc -std=gnu99 > >> >> -fstack-protector -Wsystem-headers -Werror -Wno-pointer-sign -Wno-empty-body > >> >> -Wno-string-plus-int -Wno-unused-const-variable -Wno-tautological-compare > >> >> -Wno-unused-value -Wno-parentheses-equality -Wno-unused-function > >> >> -Wno-enum-conversion -Wno-switch -Wno-switch-enum -Wno-knr-promoted-parameter > >> >> -Wno-parentheses -Qunused-arguments -c /usr/src/lib/librpcsvc/yp_passwd.c -o > >> >> yp_passwd.o --- all_subdir_libproc --- --- libproc.so.3 > >> >> --- /usr/obj/usr/src/tmp/usr/bin/ld: skipping > >> >> incompatible /usr/obj/usr/src/tmp/usr/lib/libctf.so when searching for > >> > > >> > I'm confused by this message. Are you building with -DNO_CLEAN? Do you > >> > have anything in make.conf or src.conf, especially anything that's > >> > changed since libctf was rebuilt? > >> > > >> > You might try rebuilding libctf with > >> > > >> > $ cd /usr/src > >> > $ make -C cddl/lib/libctf clean all > >> > > >> > but I'm not sure why ld is ignoring the existing libctf.so. > >> > >> The failure is coming while building the lib32 compat libraries. Are > >> we not currently building a lib32 libctf.so? > > > > No, we do. One thing I've noticed is that cddl/lib is built after lib/ > > when compiling 32-bit libs, whereas cddl/lib is built first when building > > natively. > > Sorry, that's not even true. I misread a part of Makefile.inc1. > > I'm still not able to reproduce the problem, but it seems that the > patch here is appropriate: > http://people.freebsd.org/~markj/patches/libctf_prebuild.diff > > Oliver, could you give this a try? This patch doesn't apply with sources at rev. 272559, as it bails out. Investigating the situation shows, that my top-level Makefile.inc seems different from what the patch is supposed to expect: root_at_gate [src] patch -p1 < /tmp/libctf_prebuild.diff Hmm... Looks like a unified diff to me... The text leading up to this was: -------------------------- |diff --git a/Makefile.inc1 b/Makefile.inc1 |index 3b92aef..9350339 100644 |--- a/Makefile.inc1 |+++ b/Makefile.inc1 -------------------------- Patching file Makefile.inc1 using Plan A... Hunk #1 succeeded at 1536 with fuzz 1 (offset 2 lines). Hunk #2 failed at 1584. 1 out of 2 hunks failed--saving rejects to Makefile.inc1.rej done The patch at line 1530 ff: [...] lib/libgeom \ ${_cddl_lib_libumem} ${_cddl_lib_libnvpair} \ ${_cddl_lib_libuutil} \ ${_cddl_lib_libavl} \ ${_cddl_lib_libzfs_core} \ ${_cddl_lib_libctf} \ lib/libutil lib/libpjdlog ${_lib_libypclnt} lib/libz lib/msun \ ${_secure_lib_libcrypto} ${_lib_libldns} \ ${_secure_lib_libssh} ${_secure_lib_libssl} My source tree is clean and sober as "svn status" reports. A bit spooky, isn't it? Oliver
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:40:52 UTC