On Tue, Nov 16, 2021 at 12:40 AM Jean-Philippe Brucker jean-philippe@linaro.org wrote:
On Mon, Nov 15, 2021 at 10:10:03PM -0800, Andrii Nakryiko wrote:
On Fri, Nov 12, 2021 at 8:28 AM Jean-Philippe Brucker jean-philippe@linaro.org wrote:
On Fri, Nov 12, 2021 at 04:17:21PM +0000, Quentin Monnet wrote:
2021-11-12 15:51 UTC+0000 ~ Jean-Philippe Brucker jean-philippe@linaro.org
Commit be79505caf3f ("tools/runqslower: Install libbpf headers when building") uses the target libbpf to build the host bpftool, which doesn't work when cross-building:
make ARCH=arm64 CROSS_COMPILE=aarch64-linux-gnu- -C tools/bpf/runqslower O=/tmp/runqslower ... LINK /tmp/runqslower/bpftool/bpftool /usr/bin/ld: /tmp/runqslower/libbpf/libbpf.a(libbpf-in.o): Relocations in generic ELF (EM: 183) /usr/bin/ld: /tmp/runqslower/libbpf/libbpf.a: error adding symbols: file in wrong format collect2: error: ld returned 1 exit status
When cross-building, the target architecture differs from the host. The bpftool used for building runqslower is executed on the host, and thus must use a different libbpf than that used for runqslower itself. Remove the LIBBPF_OUTPUT and LIBBPF_DESTDIR parameters, so the bpftool build makes its own library if necessary.
In the selftests, pass the host bpftool, already a prerequisite for the runqslower recipe, as BPFTOOL_OUTPUT. The runqslower Makefile will use the bpftool that's already built for selftests instead of making a new one.
Fixes: be79505caf3f ("tools/runqslower: Install libbpf headers when building") Signed-off-by: Jean-Philippe Brucker jean-philippe@linaro.org
Reviewed-by: Quentin Monnet quentin@isovalent.com
I realised too late I should have cc-ed you on those patches, apologies for not doing so. Thank you for the fix!
No worries, I usually try to catch build issues in bpf-next but missed it this time. I'm still slowly working towards getting automated testing on Arm targets, which will catch regressions quicker.
Are you planning to contribute it to BPF CI? Or you meant you have your own separate system?
At the moment I'm looking at adding it to LKFT, which does regression testing on various Arm boards (https://lkft.linaro.org/about/). The only bit missing for that is support for cross-building tools with clang, which I'll send shortly.
Detecting regressions is good, but preventing regressions is even better. That's what we are doing with BPF CI. We create a Github PR and run BPF selftests before patches are applied (see [0] for the current list of patch sets being reviewed and thus tested). It would be good to have more support for various architectures there. We currently have x86-64 and s390x was added just recently. Regressions still happen, obviously (we don't also control all the other parts of the kernel so after tree merges things quite often are a bit unstable and broken, but the system has caught many issues already), but it's been a tremendous help and time saver. So consider contributing, if you care about those architectures.
[0] https://github.com/kernel-patches/bpf/pulls
Thanks, Jean