Date: Wed, 14 Oct 2015 11:29:27 +0300 From: Jukka Ukkonen <jau789@gmail.com> To: John-Mark Gurney <jmg@funkthat.com> Cc: freebsd-arm@freebsd.org Subject: Re: machine/acle-compat.h missing Message-ID: <561E1267.4000404@gmail.com> In-Reply-To: <20151013162244.GS67524@funkthat.com> References: <561B4A25.2030703@gmail.com> <20151012232952.GR67524@funkthat.com> <qdnhjn.nw50ar.1hge18i-qmf@smtp.gmail.com> <20151013162244.GS67524@funkthat.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 10/13/15 19:22, John-Mark Gurney wrote: > You should not need to set CC and CXX to build, and these are probably > causing your build failures... I don't have a RPi2, but the instructions > at: https://wiki.freebsd.org/FreeBSD/arm/Raspberry%20Pi%202%20image > should be enough, notice the following: > # make TARGET_ARCH=armv6 UBLDR_LOADADDR=0x2000000 buildworld > # make TARGET_ARCH=armv6 KERNCONF=RPI2 buildkernel Well, there is nothing different to what I have been doing unless we consider my setting the target architecture to armv6hf and use of an explicit source root path as somehow error prone. I tested again building with the default compiler on 10-stable. It identifies itself like this... > cc -v FreeBSD clang version 3.4.1 (tags/RELEASE_34/dot1-final 208032) 20140512 Target: x86_64-unknown-freebsd10.2 Thread model: posix Selected GCC installation: Now the error does not happen while building the new version of clang, but still this does not exactly count as a huge success either... --- sbin.all__D --- cc -O3 -pipe -std=gnu99 -Wsystem-headers -Werror -Wall -Wno-format-y2k -Wno-uninitialized -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-unused-local-typedef -Wno-switch -Wno-switch-enum -Wno-knr-promoted-parameter -Qunused-arguments -c /tank/fish/jau/Xcomp/src.head/sbin/clri/clri.c -o clri.o --- tests.all__D --- --- _sub.all --- ===> tests/sys/pjdfstest/tests/granular (buildfiles) --- _sub.buildfiles --- mv: Kyuafile.auto.tmp: No such file or directory --- share.all__D --- colldef -I /tank/fish/jau/Xcomp/src.head/share/colldef -o uk_UA.ISO8859-5.out /tank/fish/jau/Xcomp/src.head/share/colldef/uk_UA.ISO8859-5.src --- tests.all__D --- *** [Kyuafile.auto] Error code 1 make[7]: stopped in /tank/fish/jau/Xcomp/src.head/tests/sys/pjdfstest/tests/ftruncate 1 error make[7]: stopped in /tank/fish/jau/Xcomp/src.head/tests/sys/pjdfstest/tests/ftruncate --- libexec.all__D --- gzip -cn /tank/fish/jau/Xcomp/src.head/contrib/atf/atf-sh/atf-check.1 > atf-check.1.gz --- tests.all__D --- *** [_sub.buildfiles] Error code 2 Just to clarify things and to make sure we are not comparing apples to oranges... Are you building for RPI2 on 10-stable? Do you use jail containers for the builds or plain separate source (and object) root path? Jailed builds have apparently sometimes behaved differently from builds using the actual host environment. > Once you have those built, then comes how to install it... One method > is to do: > make TARGET_ARCH=argv6 installworld DESTDIR=<someemptydir> -DNO_ROOT > tar -czf rpi2.tgz @<someemptydir>/METALOG My install phase has always been just fine. At the moment there does not seem to be a big risk that I would even have to test it again in any predictable future, because the build phase falls on its face. Maybe there is no need to worry about the install phase just yet. Thanks anyway. --jau
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?561E1267.4000404>