Re: [PATCHv2 bpf-next 8/8] selftests/bpf: Test user memory kfunc errors

[Jiri Olsa <olsajiri@gmail.com>]

On Tue, Nov 03, 2026 at 08:39:48AM +0000, bot+bpf-ci@kernel.org wrote:
> > selftests/bpf: Test user memory kfunc errors
> 
> TL;DR: Test nit only: the user_mem_kprobe() selftest looks redundant with
> user_mem_nonsleepable(), since both have the same body and expected failure
> and load as the same program type. One may be droppable.
> 
> > diff --git a/tools/testing/selftests/bpf/progs/user_mem.c b/tools/testing/selftests/bpf/progs/user_mem.c
> > --- a/tools/testing/selftests/bpf/progs/user_mem.c
> > +++ b/tools/testing/selftests/bpf/progs/user_mem.c
> 
> [ ... ]
> 
> > @@ -80,3 +80,32 @@ int user_mem_test(struct pt_regs *ctx)
> > +SEC("?uprobe//proc/self/exe:user_mem_trigger")
> > +__failure
> > +__msg("program must be sleepable to call sleepable kfunc bpf_user_anon_mmap")
> > +__msg("cannot be used in non-sleepable program")
> > +int user_mem_nonsleepable(struct pt_regs *ctx)
> > +{
> > +	bpf_user_anon_mmap(4096, 1);
> > +	return 0;
> > +}
> > +
> > +SEC("?kprobe")
> > +__failure
> > +__msg("program must be sleepable to call sleepable kfunc bpf_user_anon_mmap")
> > +__msg("cannot be used in non-sleepable program")
> > +int user_mem_kprobe(struct pt_regs *ctx)
> > +{
> > +	bpf_user_anon_mmap(4096, 1);
> > +	return 0;
> > +}
> 
> This isn't a bug, but user_mem_kprobe() has the same body and the same
> __failure and __msg() annotations as user_mem_nonsleepable() above it.
> 
> Since SEC("kprobe") and SEC("uprobe") both load as BPF_PROG_TYPE_KPROBE
> and the attach target is not used at load time, does user_mem_kprobe()
> test anything that user_mem_nonsleepable() does not, or could one of the
> two be dropped?

yep, will drop one

jirka