The dtrace xcalls are likely caused by powertop(1) itself, it's a side
effect of watching probes that fire very often on a per CPU basis on
large systems (the idle_state_transition probe on this case).
The usual cause of higher consumption, when not shown by powertop(1), is
high CPU usage. John, can you check if you have anything unusual in
prstat(1)?
Rafael
On 01/ 3/11 07:48 AM, Milan Jurik - Oracle Czech wrote:
Hi John,
interesting amount of dtrace_xcall_func(), do you have some dtrace
scripts running at that time? If you compare b154 and b155 with
powertop, do you see some significant difference?
Best regards,
Milan
John Martin píše v pá 31. 12. 2010 v 12:33 -0500:
Ever since upgrading from b154 to b155 my power meter shows
the average power use for an idle desktop to be about 20W higher and
the power level jumps around quite a bit with swings as high
as an additional 50W. Any known causes? Typical PowerTOP output:
OpenSolaris PowerTOP version 1.2
C-states (idle power) Avg Residency P-states (frequencies)
C0 (cpu running) (17.5%) 1596 Mhz 75.3%
C1 1.3ms (38.2%) 1729 Mhz 0.0%
C2 0.0ms (0.0%) 1862 Mhz 0.0%
C3 0.0ms (44.3%) 1995 Mhz 0.0%
2128 Mhz 0.0%
2261 Mhz 0.0%
2394 Mhz 0.0%
2527 Mhz 0.0%
2660 Mhz 0.0%
2793 Mhz 0.0%
2926 Mhz 0.0%
3059 Mhz 0.0%
3192 Mhz 0.0%
3325 Mhz 0.0%
3425
Mhz(turbo) 24.7%
Wakeups-from-idle per second: 139250.4 interval: 5.0s
no ACPI power usage estimate available
Top causes for wakeups:
32.5% (45217.1) sched : <xcalls> unix`dtrace_xcall_func
2.2% (3007.1) sched : <xcalls>
unix`speedstep_pstate_transition
0.4% (610.8)<kernel> : genunix`clock
0.4% (495.6)<kernel> : genunix`cv_wakeup
0.2% (305.4)<kernel> : SDC`sysdc_update
0.1% (141.0)<kernel> :
uhci`uhci_handle_root_hub_status_change
0.1% (113.6)<interrupt> : ahci#0
0.1% ( 74.4)<kernel> : cpudrv`cpudrv_monitor_disp
0.0% ( 69.0) intrd : <xcalls> unix`hati_demap_func
0.0% ( 47.2)<kernel> :
ehci`ehci_handle_root_hub_status_change
0.0% ( 40.6) mixer_applet2 : <xcalls>
unix`speedstep_pstate_transition
0.0% ( 36.8)<kernel> : uhci`uhci_cmd_timeout_hdlr
0.0% ( 36.0) zpool-rpool :<xcalls>
unix`speedstep_pstate_transition
0.0% ( 30.8) Xorg :<xcalls>
unix`speedstep_pstate_transition
0.0% ( 24.4)<kernel> : genunix`schedpaging
0.0% ( 22.6) updatemanagernot :<xcalls>
unix`speedstep_pstate_transition
0.0% ( 17.6) fsflush :<xcalls>
unix`speedstep_pstate_transition
0.0% ( 11.2)<kernel> : sd`sd_pm_idletimeout_handler
0.0% ( 6.6) nwam-manager :<xcalls>
unix`speedstep_pstate_transition
0.0% ( 6.2)<interrupt> : nvidia#0
_______________________________________________
pm-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pm-discuss
_______________________________________________
pm-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pm-discuss
