Re: [PATCHv2 bpf-next 6/6] selftests/bpf: Fix get_func_ip offset test for CONFIG_X86_KERNEL_IBT

[Jiri Olsa <olsajiri@gmail.com>]

On Tue, Aug 16, 2022 at 09:00:20AM +0200, Jiri Olsa wrote:
> On Mon, Aug 15, 2022 at 08:25:51PM -0700, Andrii Nakryiko wrote:
> > On Thu, Aug 11, 2022 at 2:16 AM Jiri Olsa <jolsa@kernel.org> wrote:
> > >
> > > With CONFIG_X86_KERNEL_IBT enabled the test for kprobe with offset
> > > won't work because of the extra endbr instruction.
> > >
> > > As suggested by Andrii adding CONFIG_X86_KERNEL_IBT detection
> > > and using appropriate offset value based on that.
> > >
> > > Also removing test7 program, because it does the same as test6.
> > >
> > > Suggested-by: Andrii Nakryiko <andrii@kernel.org>
> > > Signed-off-by: Jiri Olsa <jolsa@kernel.org>
> > > ---
> > >  .../bpf/prog_tests/get_func_ip_test.c         | 62 +++++++++++++++----
> > >  .../selftests/bpf/progs/get_func_ip_test.c    | 20 +++---
> > >  2 files changed, 60 insertions(+), 22 deletions(-)
> > >
> > > diff --git a/tools/testing/selftests/bpf/prog_tests/get_func_ip_test.c b/tools/testing/selftests/bpf/prog_tests/get_func_ip_test.c
> > > index 938dbd4d7c2f..a4dab2fa2258 100644
> > > --- a/tools/testing/selftests/bpf/prog_tests/get_func_ip_test.c
> > > +++ b/tools/testing/selftests/bpf/prog_tests/get_func_ip_test.c
> > > @@ -2,7 +2,9 @@
> > >  #include <test_progs.h>
> > >  #include "get_func_ip_test.skel.h"
> > >
> > > -void test_get_func_ip_test(void)
> > > +static int config_ibt;
> > > +
> > > +static void test_function_entry(void)
> > >  {
> > >         struct get_func_ip_test *skel = NULL;
> > >         int err, prog_fd;
> > > @@ -12,14 +14,6 @@ void test_get_func_ip_test(void)
> > >         if (!ASSERT_OK_PTR(skel, "get_func_ip_test__open"))
> > >                 return;
> > >
> > > -       /* test6 is x86_64 specifc because of the instruction
> > > -        * offset, disabling it for all other archs
> > > -        */
> > > -#ifndef __x86_64__
> > > -       bpf_program__set_autoload(skel->progs.test6, false);
> > > -       bpf_program__set_autoload(skel->progs.test7, false);
> > > -#endif
> > > -
> > >         err = get_func_ip_test__load(skel);
> > >         if (!ASSERT_OK(err, "get_func_ip_test__load"))
> > >                 goto cleanup;
> > > @@ -38,16 +32,62 @@ void test_get_func_ip_test(void)
> > >
> > >         ASSERT_OK(err, "test_run");
> > >
> > > +       config_ibt = skel->bss->config_ibt;
> > 
> > skel->bss->config_ibt isn't actually necessary, you can just check
> > skel->kconfig->CONFIG_X86_KERNEL_IBT directly. And you won't need to
> > trigger BPF program unnecessary, libbpf will fill out kconfig section
> > during object/skeleton load phase.
> 
> nice, did not know that ;-) will remove it

actually.. to get kconfig datasec generated in BTF the CONFIG_X86_KERNEL_IBT
symbol needs to be used.. so I came up with adding unused function:

  int unused(void)
  {
        return CONFIG_X86_KERNEL_IBT ? 0 : 1;
  }

but I wonder having the config_ibt variable is better and more clear

thoughts?

thanks,
jirka