Re: [gpsd-dev] 100% cpu usage?

[Chris Kuethe]

This smells like select in the main loop is returning right away.

strace/ktrace to see if that's the case, you can also compile with profiling to see if some code path is running really hot.

On Tue, Oct 29, 2013 at 7:34 PM, Eric S. Raymond <address@hidden> wrote:

Greg Troxel <address@hidden>:

> But, I noticed:
>
>    1456 nobody    39  -10  7828K 2484K CPU/1    215:08 99.02% 99.02% gpsd
>
> on one system, and
>
>   19758 nobody    37  -10    33M 2692K CPU/0     21:37 98.00% 98.00% gpsd
>
> on another.  This is just watching the ublox and passing time to ntpd.
>
> Is anyone else seeing this?

Dunno yet.  What tool and options should I use to get a comparable report?

And...the most likely thing in the code to cause a buzz is the PPS monitoring
loop.  Are you conditioning that out?
--
                <a href="" href="http://www.catb.org/~esr/" target="_blank">http://www.catb.org/~esr/">Eric S. Raymond</a>

--
GDB has a 'break' feature; why doesn't it have 'fix' too?