Hi,
I am now Running -- 2.4.22-1.2088.nptl (Fedora Test 3)
After executing test-starve---
>test-starve
>expecting to receive 10000000 signals
I am indeed using TOP:
---------------------------------------------------------------------
09:45:36 up 10:40, 4 users, load average: 2.08, 0.62, 0.21
101 processes: 97 sleeping, 4 running, 0 zombie, 0 stopped
CPU states: cpu user nice system irq softirq iowait idle
total 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0%
Mem: 514380k av, 500620k used, 13760k free, 0k shrd, 295000k
buff
275508k active, 192160k inactive
Swap: 2096472k av, 41204k used, 2055268k free 72136k
cached
PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME CPU COMMAND
28167 williams 17 0 300 300 280 S 96.5 0.0 0:55 0 test-starve
4001 root 16 0 271M 15M 5080 R 3.1 3.0 7:55 0 X
28166 williams 16 0 1224 1224 892 R 0.3 0.2 0:00 0 top
4161 williams 16 0 556 520 432 S 0.1 0.1 3:36 0 autorun
28129 williams 15 0 11532 11M 10000 S 0.1 2.1 0:00 0 kdeinit
1 root 16 0 376 376 324 S 0.0 0.0 0:04 0 init
2 root 15 0 0 0 0 SW 0.0 0.0 0:00 0 keventd
---------------------------------------------------------------------------
Now, I will try the M (display memory on) option for TOP:
---------------------------------------------------------------------------
09:52:18 up 10:47, 4 users, load average: 2.42, 1.44, 0.68
101 processes: 98 sleeping, 3 running, 0 zombie, 0 stopped
CPU states: cpu user nice system irq softirq iowait idle
total 0.0% 0.0% 0.0% 33.3% 33.4% 33.1% 0.0%
Mem: 514380k av, 505756k used, 8624k free, 0k shrd, 295424k
buff
275900k active, 196880k inactive
Swap: 2096472k av, 41200k used, 2055272k free 76384k
cached
PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME CPU COMMAND
28183 williams 17 0 304 304 284 S 98.1 0.0 0:59 0 test-starve
4161 williams 16 0 556 520 432 S 0.9 0.1 3:37 0 autorun
28185 williams 17 0 1224 1224 896 R 0.3 0.2 0:00 0 top
4001 root 15 0 271M 15M 5080 S 0.1 3.0 7:57 0 X
------------------------------------------------------------------------
After the test program finishes, it does return the CPU.
Thanks,
Ernesto
On Sat, 2003-10-18 at 05:10, Ingo Molnar wrote:
> On Tue, 14 Oct 2003, Ernest L. Williams Jr. wrote:
>
> > I see that is taking up to 97% of the CPU and not letting go.
> >
> > Hewlett Packard has the test program on their web site.
>
> I can see no starvation.
>
> are you looking at it via 'top'? If yes then press 'M' to show all threads
> - test-starve.c is a threaded application.
>
> Ingo
--
Ernest L. Williams Jr. <ernesto@xxxxxxxx>
