gpg-agent timer tick

Alan Jenkins alan-jenkins at tuffmail.co.uk
Fri Jan 11 18:09:11 CET 2008


gpg-agent has a 2 second timer tick.

    alan at singularity:~$ gpg-agent --daemon
    GPG_AGENT_INFO=/tmp/gpg-21C85K/S.gpg-agent:28978:1; export
    GPG_AGENT_INFO;

    alan at singularity:~$ sudo powertop -d -t 5 | grep gpg-agent
       0.2% (  0.6)         gpg-agent : schedule_timeout (process_timeout)


    gpg-agent:1557:handle_connections(int listen_fd, int listen_fd_ssh):
    time_ev = pth_event (PTH_EVENT_TIME, pth_timeout (2, 0));

This is potentially a Bad Thing for CPU power consumption, because it 
reduces the amount of time the CPU can spend in deep sleep states.  I 
saw something similar in ssh-agent, but I couldn't interest anyone in my 
patch.  Let me know if you're interested here and I'll submit a patch.

Specifically, gpg-agent uses the timer tick to handle the case where the 
"command-line" argument is present:

gpg-agent --daemon [command-line]

In that case it runs the command given.  The tick handler is used to 
check, every two seconds, whether the command is still running; if not, 
gpg-agent quits.

A better way to do this is to run the command as a child of gpg-agent, 
and rely on the SIGCHLD signal instead of a timer tick.

The timer tick is also used to check the Scdaemon (whatever that is :-) 
is still running.  However, since it use waitpid() to check this, 
Scdaemon must already be being run a child, so that shouldn't be a problem.

I noticed this problem because I got interested the powertop project, 
and gpg-agent gets started by some package management software I use.

Thanks,

Alan



More information about the Gnupg-devel mailing list