On 2018-07-03, Grant Edwards <grant.b.edwa...@gmail.com> wrote:
> On 2018-07-03, Grant Edwards <grant.b.edwa...@gmail.com> wrote:
>> On 2018-07-01, Marko Rauhamaa <ma...@pacujo.net> wrote:
>>> Gregory Ewing <greg.ew...@canterbury.ac.nz>:
>>>
>>>> I don't see how the address-reuse timeout can be a security measure,
>>>> because the process trying to take over the address can easily
>>>> circumvent it by setting SO_REUSEADDR.
>>>
>>> [...]
>>>
>>> Nevertheless, the later socket object cannot unilaterally take over a
>>> socket using SO_REUSEADDR. The earlier socket object must have set the
>>> same option previously.
>>
>> On what OS? In my experience, that's not true on Linux or BSD Unix.
>
> I was wrong. I just did a quick test on Linux, and it works the way
> Marko Hauhamaa describes.
Um that was supposed to be Rauhamaa... sorry about that.
> [Don't have handy access to a BSD system at the moment.]
Found a NetBSD 7.1 system to test on...
After killing a server with an active connection, a new server can
bind to the same socket regardless of the SO_REUSEADDR setting in
either server. I don't know if that's some sort of system
configuration setting or what...
--
Grant Edwards grant.b.edwards Yow! Where's the Coke
at machine? Tell me a joke!!
gmail.com
--
https://mail.python.org/mailman/listinfo/python-list
