Re: [PATCH 3/8] [PATCH] Helper modules load-on-demand support for ctnetlink

[Patrick McHardy <kaber@trash.net>]

Pablo Neira Ayuso wrote:
> Patrick McHardy wrote:
>> Patrick McHardy wrote:
>>> This one doesn't apply:
>>>
>>> patching file include/net/netfilter/nf_conntrack_helper.h
>>> patching file net/netfilter/nf_conntrack_core.c
>>> Hunk #1 FAILED at 581.
>>> Hunk #2 succeeded at 753 (offset -5 lines).
>>> Hunk #3 succeeded at 764 (offset -5 lines).
>>> 1 out of 3 hunks FAILED -- saving rejects to file
>>> net/netfilter/nf_conntrack_core.c.rej
>>> patching file net/netfilter/nf_conntrack_helper.c
>>> patching file net/netfilter/nf_conntrack_netlink.c
>>> Hunk #2 succeeded at 1131 with fuzz 1.
>>> patching file net/netfilter/nfnetlink.c
>> Oops, I think I didn't refresh my tree, so it didn't include any
>> of the netfilter patches I sent to Dave :)
> 
> So, what do I have to do now? I'm cloning Dave's tree now to put the
> patches on top on it. Fine with it?

I think it was my mistake, let me try again ... yes, it applies
cleanly. Both Dave's tree or my tree should work, but use Dave's
if you want to be safe since I haven't updated mine yet.

>> The remaining issues, especially the nfnl_lock think, should be
>> fixed though.
> 
> OK. BTW, returning EOPNOTSUPP looks better to me. ENOENT is misleading
> since the user won't be able to differenciate between "I don't know
> anything about that helper" and "this conntrack entry does not exist".
> So, I prefer using EOPNOTSUPP.

Makes sense.