Re: [PATCH bpf-next v11 11/11] selftests/bpf: lsm_cgroup functional test

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Wed, Jun 29, 2022 at 2:29 PM Stanislav Fomichev <sdf@xxxxxxxxxx> wrote:
>
> On Wed, Jun 29, 2022 at 1:31 PM Stanislav Fomichev <sdf@xxxxxxxxxx> wrote:
> >
> > On Wed, Jun 29, 2022 at 1:26 PM Alexei Starovoitov
> > <alexei.starovoitov@xxxxxxxxx> wrote:
> > >
> > > On Tue, Jun 28, 2022 at 10:43 AM Stanislav Fomichev <sdf@xxxxxxxxxx> wrote:
> > > > +
> > > > +static void test_lsm_cgroup_functional(void)
> > >
> > > It fails BPF CI on s390:
> > >
> > > test_lsm_cgroup_functional:FAIL:attach alloc_prog_fd unexpected error:
> > > -524 (errno 524)
> > > test_lsm_cgroup_functional:FAIL:detach_create unexpected
> > > detach_create: actual -2 < expected 0
> > > test_lsm_cgroup_functional:FAIL:detach_alloc unexpected detach_alloc:
> > > actual -2 < expected 0
> > > test_lsm_cgroup_functional:FAIL:detach_clone unexpected detach_clone:
> > > actual -2 < expected 0
> > >
> > > https://github.com/kernel-patches/bpf/runs/7100626120?check_suite_focus=true
> > >
> > > but I pushed it to bpf-next anyway.
> > > Thanks a lot for this work and please follow up with a fix.
> >
> > Thanks, I'll take a look!
>
> Looks like this needs a blacklist entry in
> https://github.com/kernel-patches/vmtest/blob/master/travis-ci/vmtest/configs/blacklist/BLACKLIST-latest.s390x
>
> Or, I can make tests more flexible by doing the following
> (copy-pasting into gmail, so tabs are broken):
>
> diff --git a/tools/testing/selftests/bpf/prog_tests/lsm_cgroup.c
> b/tools/testing/selftests/bpf/prog_tests/lsm_cgroup.c
> index d40810a742fa..904b02a17598 100644
> --- a/tools/testing/selftests/bpf/prog_tests/lsm_cgroup.c
> +++ b/tools/testing/selftests/bpf/prog_tests/lsm_cgroup.c
> @@ -100,6 +100,10 @@ static void test_lsm_cgroup_functional(void)
>   ASSERT_EQ(query_prog_cnt(cgroup_fd, "bpf_lsm_sk_alloc_security"), 0,
> "prog count");
>   ASSERT_EQ(query_prog_cnt(cgroup_fd, NULL), 0, "total prog count");
>   err = bpf_prog_attach(alloc_prog_fd, cgroup_fd, BPF_LSM_CGROUP, 0);
> + if (err < 0 && errno == ENOTSUPP) {
> + test__skip();
> + return;
> + }
>   if (!ASSERT_OK(err, "attach alloc_prog_fd"))
>   goto detach_cgroup;
>   ASSERT_EQ(query_prog_cnt(cgroup_fd, "bpf_lsm_sk_alloc_security"), 1,
> "prog count");
>
> Any preference?

Ahh. Right. s390 lacks bpf trampoline support.
We've been blacklisting the tests manually,
but if it's fixable this way it's better to do it in the test,
so when s390 (and other archs) gain trampoline support
the test will be executed automatically without waiting for CI
maintainers to unlist the tests.



[Index of Archives]     [Linux Samsung SoC]     [Linux Rockchip SoC]     [Linux Actions SoC]     [Linux for Synopsys ARC Processors]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]


  Powered by Linux