On Fri, 28 Apr 2017 20:35:21 -0700
Alexei Starovoitov <alexei.starovoi...@gmail.com> wrote:
> On Fri, Apr 28, 2017 at 04:25:04PM +0200, Jesper Dangaard Brouer wrote:
> > The struct bpf_map_def was extended in commit fb30d4b71214 ("bpf: Add tests
> > for map-in-map") with member unsigned int inner_map_idx. This changed the
> > size
> > of the maps section in the generated ELF _kern.o files.
> >
> > Unfortunately the loader in bpf_load.c does not detect or handle this.
> > Thus,
> > older _kern.o files became incompatible, and caused hard-to-debug errors
> > where the syscall validation rejected BPF_MAP_CREATE request.
> >
> > This patch only detect the situation and aborts load_bpf_file(). It also
> > add code comments warning people that read this loader for inspiration
> > for these pitfalls.
> >
> > Fixes: fb30d4b71214 ("bpf: Add tests for map-in-map")
> > Signed-off-by: Jesper Dangaard Brouer <bro...@redhat.com>
>
> Thanks!
> Acked-by: Alexei Starovoitov <a...@kernel.org>
Thanks!
> > Is it worth to implement proper backward-compat loading of older ELF objects
> > with this bpf-loader?
>
> probably yes, since it looks like a bunch of code in samples/bpf/ still
> depend on it and some features are missing in tools/lib/bpf,
> so unless we actively work on improving libbpf.a
> we won't be able to get rid of this 'sample' loader for some time.
Okay, I'll work on that next week. Hoping I can make the merge window,
so we avoid having a non-backward compat bpf_load in a kernel release.
p.s. I'll be presenting about XDP in Sweden Thur+Friday next week:
https://lundlinuxcon.org/?page=current
--
Best regards,
Jesper Dangaard Brouer
MSc.CS, Principal Kernel Engineer at Red Hat
LinkedIn: http://www.linkedin.com/in/brouer
