[ITA] inetutils-1.5-1

[Charles Wilson]

Corinna Vinschen wrote:
> On Feb 24 22:07, Charles Wilson wrote:
>> I'm tossing my hat in for #3.  It's basically a new port, using the 
>> previous patches (1.3.2-37 vs. upstream 1.3.2) as a guide, because none of 
>> the patches seemed to apply cleanly, and there were other issues as well.
> 
> Yippee!  Charles, there is no number of gold stars in the known universe
> to express my gratitude :)))

You're welcome.

> For a start, maybe you should change the default motd.  I just couldn't
> think of something better way back when, but the message is rather
> boring, isn't it?

OK.  I hadn't really given those sorts of things much thought.

> And, maybe it's time to start to be more cautious by default and
> disable all service entries in /etc/defaults/etc/inetd.conf?

You're probably right. Existing users' inetd.confs will not get 
overwritten -- but they will have to manually edit them anyway, given 
the executable name changes.  New users...will have to read the README. 
Or perhaps I could make a REALLY smart iu-config


The 'chargen' service is blah blah blah...describe security issues...
Do you want to enable the 'chargen' service? [yes/no]


etc.


What do you think, further, of requiring tcp_wrappers, and having the 
default inetd.conf file explicitly use /usr/sbin/tcpd (even if the line 
is commented out and entirely disabled)?


> A few minutes ago I found a security problem with rshd.c which I just
> fixed by uploading 1.3.2-40, and which you copied verbatim into the 1.5
> rshd.c:
> 

Probably need something similar in rexecd, too. I'll take a look before 
posting 1.5-2.

> Since the check if seteuid fails is missing so far, you will run rsh
> commands under the SYSTEM account for every user on 2003 upwards!

Urk. That's bad.

> In -40 I changed the description in inetutils-1.3.2.README substantially
> to explain this problem.

I will track all of your changes as we go forward (you probably saw I 
copied in your README verbatim as inetutils.OLD-README).

>> (3) Added a new option to inetd: -T/--traditional-daemon, which does the 
>> regular fork/daemonize behavior.  This is used with the (also provided) 
>> sysvinit-style startup script, so that inetd can be run under the control 
>> of the sysvinit package's init daemon.  So now, there are THREE ways to run 
>> inetd as a service:
>>   a) install as a service using cygrunsrv (with the -D option)
>>   b) installed as a service under its own power
>>   c) as a slave to the init service, using /etc/rc.d/init.d/inetd (which
>>      uses the -T option when invoking inetd)
> 
> Given the problem with the SYSTEM account, maybe we should deprecate 
> usage b. 

Well, I kinda wanted to avoid a huge "flag day" thing where stuff just 
stopped working for people (well, except for the server executable names 
thing).

But I could definitely see "method (b) is supported NT, 2000, and XP 
only", encouraging people to use method (a) or (c).

Right now, neither your version nor my version of iu-config actually 
installs (or even offers to install) inetd as a service.  If the new 
iu-config became more like syslogd-config or sshd-host-config, then I 
could see it defaulting to method (a).

> Maybe an install script (iu-config?) could do something
> along the lines of the ssh-host-config script.

Hmmm...I've got an idea...cue ominous music...

> I would be willing
> to switch the ssh-host-config script from the "sshd_server" user name
> to something like "cygwin_svc" or so.

How about a new package, "cygwin-services-helper" or somesuch, that contains

(1) a script [*] derived from the appropriate portion of 
sshd-host-config, whose job is to create the appropriate priveleged user 
(I like 'cygwin_svc') -- unless it already exists under either name 
('cygwin_svc' or 'sshd_server').

(2) maybe another script [*] whose job is to ascertain whether such a 
user already exists, and return its name (or "" if not).

It would be up to the calling foo-config to use these two scripts 
appropriately.  And, of course, the user might have to enter the 
password for the priveleged user account twice: once when it is created, 
and then again (by foo-config) to install the service 'foo'.

Then, openssh (and inetutils, and syslog-ng, and sysvinit, ...) could 
all depend on the "cygwin-services-helper" package.

[*] or maybe a script function library somewhere like 
/usr/lib/cygwin-services/ that foo-config could 'source', and then call 
the functions directly.  This would help the "enter the password twice" 
problem...

> And maybe the iu-config script
> could re-use the sshd_server user if it already exists...

Right. See above.

BTW, with the new inetd.d/ support, sshd-host-config doesn't have to 
edit the inetd.conf file directly.  It can have a

     /etc/defaults/etc/inetd.d/sshd

file, that it either installs to /etc/inetd.d/ or not. (ditto /etc/xinetd.d)

--
Chuck