Messages in this thread | | | Date | Thu, 15 Jan 2015 10:34:48 +0100 (CET) | From | Jiri Kosina <> | Subject | Re: [PATCH 1/2 v2] ftrace: don't allow IPMODIFY without proper compiler support |
| |
On Thu, 15 Jan 2015, Masami Hiramatsu wrote:
> Hmm, if this binary doesn't support IPMODIFY, it should return -ENOTSUPP.
Good poing, will send v3.
> And also, IMHO, we'd better reject registering ftrace_ops with IPMODIFY > in this situation before updating hash table.
That would happen automatically if we return -ENOTSUPP, right? Because the check in ftrace_hash_move() would cause freeing of the new_hash and the error value would get propagated out to ftrace_hash_move() callers.
-- Jiri Kosina SUSE Labs
| |