On 2010-05-27 16:59-0700 Alan W. Irwin wrote:
> Hazen, have you double-checked yet that isGUI is false for your latest
> tests? Of course, even if isGUI is false, QApplication may
> partially or totally ignore that so it still may be using up some X resource
> shared with firefox. Alternatively, the consumed resource shared with
> firefox may not be in X, but in some other system server.
> Anyhow, the next time you run the test, I suggest you look carefully at the
> results made by striking the "M" key while in top to sort the top results by
> memory usage. If I do that here, the top memory hog is the X server
> PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
> 3733 root 20 0 650m 225m 6424 S 0 11.2 191:27.66 Xorg
Actually, that was data for the wrong server, and since then I have rebooted
both my computers and started both X servers, and as usual the initial X
server memory requirements are small which makes it easier to see any change
in memory footprint.
So I ran "top" with the M command struck for both computers (just to be
sure) and ran the same LIMIT=2000 test again. Neither server showed the
slightest quiver in memory use (as expected if isGUI is false).
Hazen, do you know how to boot your computer without starting up your X
server? (I do that routinely and start X by hand when I need it rather than
automatically.) Anyhow, if you are able to avoid starting the X server on
boot, then it would be interesting to see the results of the LIMIT=2000 test
just from the Linux console with no X server (or any other desktop software)
running. If there is some system resource gradually being used up each time
x10c is run, then under the simplified Linux console conditions you might
have to use a much larger LIMIT value, but you should eventually still run
into a situation where you have used all the resource (whatever it is).
Also, if the X server is not running, and you still are somehow still using
isGUI true, I suspect x10c would bomb out immediately.
Regardless of whether you run the LIMIT test with the X server running or
not or with it on and also with it off, I would still be interested in your
"top" equivalent memory hogs during the test for each case.
Alan W. Irwin
Astronomical research affiliation with Department of Physics and Astronomy,
University of Victoria (astrowww.phys.uvic.ca).
Programming affiliations with the FreeEOS equation-of-state implementation
for stellar interiors (freeeos.sf.net); PLplot scientific plotting software
package (plplot.org); the libLASi project (unifont.org/lasi); the Loads of
Linux Links project (loll.sf.net); and the Linux Brochure Project