Hi folks,
Are the -d and -t options for powertop mutually exclusive? If so,
powertop should spew a usage message and exit with non-zero status.
If not, then something is wrong. I used two different time
intervals (3 and 10 seconds) and powertop produced no causes for
wakeups if -d was also supplied:
The output produced was:
# ./powertop -time=10 -d
Solaris PowerTOP 1.0 (C) 2007 Intel Corporation
Collecting data for 0 seconds
Cn Avg residency
C0 (cpu running) (-698443.3%)
C1 3.1ms (698543.3%)
P-states (frequencies)
2393 Mhz 100.0%
Wakeups-from-idle per second: 2274327.2 interval: 0.0s
Top causes for wakeups:
# echo $?
0
Clearly something isn't right since it collects data for "0" seconds
and the residency fields are unreasonable. I couldn't find an
outright man page for powertop on lesswatts.org or opensolaris.org
and didn't see anything other powertop-related documentation on the
web that could provide further insight. This was on a clovertown
system running snv_81 (the source for the powertop built was
downloaded this morning) with the following socket info:
# psrinfo -vp
The physical processor has 4 virtual processors (0 1 4 5)
x86 (GenuineIntel 6F4 family 6 model 15 step 4 clock 2400 MHz)
Intel(r) CPU @ 2.40GHz
The physical processor has 4 virtual processors (2 3 6 7)
x86 (GenuineIntel 6F4 family 6 model 15 step 4 clock 2400 MHz)
Intel(r) CPU @ 2.40GHz
Kindly let me know if this is a user-error kind of thing or if
something can be done to make powertop better.
Thanks much,
Pat
--
Pat Bredenberg
Solaris Quality Engineer
Sun Microsystems, Inc
patrick.bredenberg at sun.com
