On 05/03/2017 08:16 AM, Jesper Dangaard Brouer wrote:
On Tue, 02 May 2017 23:10:04 +0200
Daniel Borkmann <dan...@iogearbox.net> wrote:
On 05/02/2017 02:31 PM, Jesper Dangaard Brouer wrote:
This series improves and fixes bpf ELF loader and programs under
samples/bpf. The bpf_load.c created some hard to debug issues when
the struct (bpf_map_def) used in the ELF maps section format changed
in commit fb30d4b71214 ("bpf: Add tests for map-in-map").
This was hotfixed in commit 409526bea3c3 ("samples/bpf: bpf_load.c
detect and abort if ELF maps section size is wrong") by detecting the
issue and aborting the program.
In most situations the bpf-loader should be able to handle these kind
of changes to the struct size. This patch series aim to do proper
backward and forward compabilility handling when loading ELF files.
This series also adjust the callback that was introduced in commit
9fd63d05f3e8 ("bpf: Allow bpf sample programs (*_user.c) to change
bpf_map_def") to use the new bpf_map_data structure, before more users
start to use this callback.
Hoping these changes can make the merge window, as above mentioned
commits have not been merged yet, and it would be good to avoid users
hitting these issues.
Overall, set looks good to me. The last patch doesn't have a
user yet, so probably better to drop it until there is an actual
user in the tree.
The reason for simply exporting map_data[] was that in patch 3, the
data-struct (bpf_map_data) is already exposed, thus users can already
grab and store those into a separate data structure. Thus, it seemed
natural to simply export/expose the map_data[] array directly. Guess,
I could have combined patch 4 and 3. As patch-3 uses the data struct,
but in an indirect way.
To Daniel, if you still feel we should drop patch 4, then let me know.
It is only the other patches that are time critical, as patch 4 is
trivial to introduce once the first sample program uses this directly
(instead of indirectly through the callback).
Ah, if there's still code coming from your side that will make use
of it, I'm fine with this, though it's usually best to add it with
an actual user. Given it's sample code, I don't mind too much if that
happens at a later point in time.
Thanks,
Daniel
