Re: [PATCH bpf-next v2 2/3] bpf, netns: Keep attached programs in bpf_prog_array

On Tue, Jun 23, 2020 at 11:24 PM CEST, Martin KaFai Lau wrote:

On Tue, Jun 23, 2020 at 10:59:37PM +0200, Jakub Sitnicki wrote:

quoted

On Tue, Jun 23, 2020 at 09:33 PM CEST, Martin KaFai Lau wrote:

quoted

On Tue, Jun 23, 2020 at 12:34:58PM +0200, Jakub Sitnicki wrote:

[ ... ]

quoted

@@ -93,8 +108,16 @@ static int bpf_netns_link_update_prog(struct bpf_link *link,
 		goto out_unlock;
 	}

+	run_array = rcu_dereference_protected(net->bpf.run_array[type],
+					      lockdep_is_held(&netns_bpf_mutex));
+	if (run_array)
+		ret = bpf_prog_array_replace_item(run_array, link->prog, new_prog);
+	else

When will this happen?

This will never happen, unless there is a bug. As long as there is a
link attached, run_array should never be detached (null). Because it can
be handled gracefully, we fail the bpf(LINK_UPDATE) syscall.

Your question makes me think that perhaps it should trigger a warning,
with WARN_ON_ONCE, to signal clearly to the reader that this is an
unexpected state.

WDYT?

Thanks for confirming and the explanation.

If it will never happen, I would skip the "if (run_array)".  That
will help the code reading in the future.

I would not WARN also.

Best code is no code :-)

I realized that bpf_prog_array_replace_item() cannot fail either, unless
there is a bug how we compile the prog_array. So I plan to remove that
error check as well.

Thanks for feedback.