Re: [PATCH] af_unix: Allow Unix sockets to raise SIGURG

[Shoaib Rao <rao.shoaib@oracle.com>]

On 1/29/21 11:06 AM, Jakub Kicinski wrote:
> On Fri, 29 Jan 2021 09:56:48 -0800 Shoaib Rao wrote:
>> On 1/25/21 3:36 PM, Jakub Kicinski wrote:
>>> On Fri, 22 Jan 2021 15:06:37 +0000 Matthew Wilcox (Oracle) wrote:
>>>> From: Rao Shoaib <rao.shoaib@oracle.com>
>>>>
>>>> TCP sockets allow SIGURG to be sent to the process holding the other
>>>> end of the socket.  Extend Unix sockets to have the same ability.
>>>>
>>>> The API is the same in that the sender uses sendmsg() with MSG_OOB to
>>>> raise SIGURG.  Unix sockets behave in the same way as TCP sockets with
>>>> SO_OOBINLINE set.
>>> Noob question, if we only want to support the inline mode, why don't we
>>> require SO_OOBINLINE to have been called on @other? Wouldn't that
>>> provide more consistent behavior across address families?
>>>
>>> With the current implementation the receiver will also not see MSG_OOB
>>> set in msg->msg_flags, right?
>> SO_OOBINLINE does not control the delivery of signal, It controls how
>> OOB Byte is delivered. It may not be obvious but this change does not
>> deliver any Byte, just a signal. So, as long as sendmsg flag contains
>> MSG_OOB, signal will be delivered just like it happens for TCP.
> Not as far as I can read this code. If MSG_OOB is set the data from the
> message used to be discarded, and EOPNOTSUPP returned. Now the data gets
> queued to the socket, and will be read inline.

Data was discarded because the flag was not supported, this patch 
changes that but does not support any urgent data.

OOB data has some semantics that would have to be followed and if we 
support SO_OOBINLINE we would have to support NOT SO_OOBINLINE.

One can argue that we add a socket option to allow this OR just do what 
TCP does.

Shoaib


>
> Sure, you also add firing of the signal, which is fine. The removal of
> the error check is the code I'm pointing at, so to speak.
That is the change in behavior that this change is making.
>
>>>> SIGURG is ignored by default, so applications which do not know about this
>>>> feature will be unaffected.  In addition to installing a SIGURG handler,
>>>> the receiving application must call F_SETOWN or F_SETOWN_EX to indicate
>>>> which process or thread should receive the signal.
>>>>
>>>> Signed-off-by: Rao Shoaib <rao.shoaib@oracle.com>
>>>> Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
>>>> ---
>>>>    net/unix/af_unix.c | 5 +++--
>>>>    1 file changed, 3 insertions(+), 2 deletions(-)
>>>>
>>>> diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
>>>> index 41c3303c3357..849dff688c2c 100644
>>>> --- a/net/unix/af_unix.c
>>>> +++ b/net/unix/af_unix.c
>>>> @@ -1837,8 +1837,6 @@ static int unix_stream_sendmsg(struct socket *sock, struct msghdr *msg,
>>>>    		return err;
>>>>    
>>>>    	err = -EOPNOTSUPP;
>>>> -	if (msg->msg_flags&MSG_OOB)
>>>> -		goto out_err;
>>>>    
>>>>    	if (msg->msg_namelen) {
>>>>    		err = sk->sk_state == TCP_ESTABLISHED ? -EISCONN : -EOPNOTSUPP;
>>>> @@ -1903,6 +1901,9 @@ static int unix_stream_sendmsg(struct socket *sock, struct msghdr *msg,
>>>>    		sent += size;
>>>>    	}
>>>>    
>>>> +	if (msg->msg_flags & MSG_OOB)
>>>> +		sk_send_sigurg(other);
>>>> +
>>>>    	scm_destroy(&scm);
>>>>    
>>>>    	return sent;