[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [patch]: Stop connector socket-to-channel EOF flooding

On 09.04.19 10:03, Tilo Eckert wrote:

> Am 08.04.2019 um 20:54 schrieb g4-lisz@xxxxxxxxxxxx:
>>> I'm using connectors for a direct-tcp client. So this creates two
>>> connectors FD in --> channel out and vice versa.
>>> Now when the socket forwarding peer (not the ssh server) closes the
>>> connection, i.e. reading on the socket returns 0 = EOF, we end up in
>>> some loop with sending EOFs on the channel, until finally the client
>>> receives a channel close message.
>> I attached some test code to reproduce this.
>> Edit line 14 to your needs. It should connect to a service which does
>> not dicsonnecet the connenction from its side. I.e. SSH is fine. A
>> webserver will not do because it disconnects after answering the request.
>> Start a local sshd with 
>> shell1>  /usr/sbin/sshd -d -d -p 2222
>> shell2> ./fwconnector username localhost 2222
>> shell3> ssh admin@localhost -p 2022
>> Welcome to Debian...
>> ~: exit
>> On shell1 you will see the > 20 lines of "channel 0: rcvd eof"
> Your callback sends an EOF for every received EOF from the server. If
> your server does the same, you end up with EOF ping pong. You might want
> to check whether you already sent it before: channel->local_eof != 0
> Btw, if you close the channel on your end, an EOF is sent automatically
> if one has not been sent, yet.

Hi Tilo,

this is what my patch is doing: Checking for channel->local_eof != 0.

But it's also an issue to poll on a socket which has EOF from the peer's

Archive administrator: postmaster@lists.cynapses.org