[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <6A47BDC7-FB73-4799-BC6A-9C0C020E424D@oracle.com>
Date: Wed, 4 Oct 2023 13:27:15 +0000
From: Chuck Lever III <chuck.lever@...cle.com>
To: Lorenzo Bianconi <lorenzo.bianconi@...hat.com>
CC: Jakub Kicinski <kuba@...nel.org>, Lorenzo Bianconi <lorenzo@...nel.org>,
Linux NFS Mailing List <linux-nfs@...r.kernel.org>,
Jeff Layton
<jlayton@...nel.org>, Neil Brown <neilb@...e.de>,
"netdev@...r.kernel.org"
<netdev@...r.kernel.org>
Subject: Re: [PATCH v8 3/3] NFSD: add rpc_status netlink support
> On Oct 4, 2023, at 6:14 AM, Lorenzo Bianconi <lorenzo.bianconi@...hat.com> wrote:
>
>> On Mon, 11 Sep 2023 14:49:46 +0200 Lorenzo Bianconi wrote:
>>> + hdr = genlmsg_put(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq,
>>> + &nfsd_server_nl_family, NLM_F_MULTI,
>>> + NFSD_CMD_RPC_STATUS_GET);
>>> + if (!hdr)
>>> + return -ENOBUFS;
>>
>> Why NLM_F_MULTI? AFAIU that means "I'm splitting one object over
>> multiple messages". 99% of the time the right thing to do is change
>> what we consider to be "an object" rather than do F_MULTI. In theory
>> user space should re-constitute all the NLM_F_MULTI messages into as
>> single object, which none of YNL does today :(
>>
> ack, fine. I think we can get rid of it.
> @chuck: do you want me to send a patch or are you taking care of it?
Send a (tested) patch and I can squash it into this one.
--
Chuck Lever
Powered by blists - more mailing lists