Hi Collin, On Thu, Sep 05, 2024 at 12:48:03PM +0200, Simon Josefsson wrote: > Collin Funk <collin.fu...@gmail.com> writes: > > > Can someone give it a look over before pushing? > > Could you see if it is feasible to write a self-test that check if this > code is doing the right thing? There is a syslogd self test already, so > maybe just make sure it is actually testing what this function is > responsible for doing.
There seems to be a test binary tests/readutmp built from tests/readutmp.c used in test script tests/utmp.sh to check if code similar to that in syslogd and talkd can read utmp information, I think. Perhaps you can also adjust that code to work the same way as in syslogd after your patch. >From a quick glance at the syslogd code, I'd expect that logging to a specific user would trigger the utmp-reading code. The syslogd configuration file used in tests/syslogd.sh does not use this, I think. Br, Erik
