per-connection sshd doesn't always pass on SIGQUIT

[Philippe Cerfon]

On Wed, Jan 11, 2023 at 11:08 PM Damien Miller <djm at mindrot.org> wrote:
> Does it solve your problem?

Well there was no real problem in the strict sense. I've just noticed
that and wondered why in that single case, the child continues to run,
while in all other cases it's terminated, too.

So I thought it would make sense to have consistent behavior.


I don't fully agree with Bob:

If a killed parent process makes it necessary to also kill the child
in the other cases (which AFAIU sshd does manually), then it should be
also done when the parent is killed because of QUIT - or it shouldn't
be necessary in any case.

Whether the child needs to be terminated with QUIT as well, is of
course another topic.


Cheers,
Philippe