Philippe Mathieu-Daudé <f4...@amsat.org> writes:
> On 07/13/2018 09:17 AM, Alex Bennée wrote: >> Hi, >> >> Unfortunately this series grew a little while I was re-basing as I >> noticed a couple of bugs. An upstream change in debootstrap which I'd >> been happily using for Ubuntu images caused a breakage which I've >> hopefully now fixed: >> >> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=903657 >> >> However since I upgraded by devbox I hadn't clocked to the fact I >> should have been using the stable debootstrap anyway but our ugly >> shell version check failed to pick it up. That's now fixed and I can >> now bootstrap away. >> >> I grabbed a couple of Philippe's patches while I was at it as they >> were relevant to the other changes. It does mean the un-reviewed count >> has gone up slightly: >> >> patch docker/base debian tricore on qemu debian9.patch >> patch docker/split configure_qemu from build_qemu.patch >> patch docker/move make check into check_qemu helper.patch >> patch docker/gracefully skip check_qemu.patch >> patch docker/Makefile.include don t include partial images.patch >> patch docker/disable debian powerpc user cross.patch >> patch docker/add test unit runner.patch >> patch docker/add expansion for docker test FOO to Makefile.patch >> patch docker/drop QEMU_TARGET check fallback in EXECUTABLE.patch >> patch docker/add hint to docker.py check.patch >> patch docker/add commentary to debian bootstrap.docker.patch >> patch docker/ignore distro versioning of debootstrap.patch > > [Note] > > I hit: > > $ make docker-binfmt-image-debian-powerpc-user > BUILD binfmt debian-powerpc-user (debootstrapped) > Cloning into './debootstrap.git'... > remote: Enumerating objects: 3262, done. > remote: Counting objects: 100% (3262/3262), done. > remote: Compressing objects: 100% (902/902), done. > remote: Total 3262 (delta 2353), reused 3245 (delta 2339) > Receiving objects: 100% (3262/3262), 531.08 KiB | 335.00 KiB/s, done. > Resolving deltas: 100% (2353/2353), done. > The command '/bin/sh -c /debootstrap/debootstrap --second-stage' > returned a non-zero code: 139 > Traceback (most recent call last): > File "tests/docker/docker.py", line 536, in <module> > sys.exit(main()) > File "tests/docker/docker.py", line 533, in main > return args.cmdobj.run(args, argv) > File "tests/docker/docker.py", line 348, in run > extra_files_cksum=cksum) > File "tests/docker/docker.py", line 231, in build_image > quiet=quiet) > File "tests/docker/docker.py", line 158, in _do_check > return subprocess.check_call(self._command + cmd, **kwargs) > File "/usr/lib/python2.7/subprocess.py", line 190, in check_call > raise CalledProcessError(retcode, cmd) > subprocess.CalledProcessError: Command '['docker', 'build', '-t', > 'qemu:debian-powerpc-user', '-f', > '/tmp/docker_buildPNDAtz/tmpIVpPn7.docker', '/tmp/docker_buildPNDAtz']' > returned non-zero exit status 139 > make: *** [tests/docker/Makefile.include:62: > docker-binfmt-image-debian-powerpc-user] Error 1 > > "returned non-zero exit status 139" seems to be qemu-user crashing for > not being statistically linked. What system are you building on? Basically --static is always safer but as long as there isn't a clash between host/guest file system layouts it can work which is why I don't enforce it (although earlier patches tried to). > Once rebuilt after ./configure --static, it worked: > > $ make docker-binfmt-image-debian-powerpc-user V=1 > [...] > I: Extracting zlib1g... > ppc-linux-user/qemu-ppc had no associated libraries (static build?) > Sending build context to Docker daemon 226.5MB > Step 1/7 : FROM scratch > Step 2/7 : ADD . / > Step 3/7 : RUN sed -i 's/in_target mount/echo not for docker in_target > mount/g' /debootstrap/functions > Step 4/7 : RUN /debootstrap/debootstrap --second-stage > I: Installing core packages... > [...] -- Alex Bennée