Hi Douglas, Thanks for your kindly support!
The root cause is parsing _PSD error. ======== WARNING: cpu_acpi: error parsing _PSD for CPU instance 0 WARNING: cpudrv_mach_pm_init: instance 0: unable to initialize P-state support ======== Although dmesg didn't tell us why, from ist0.dat, I'm sure there is a mistake in BIOS. That's why your CPU doesn't support speedstep and power management, but I'm not sure if it's the reason your laptop became power hungry. What I can tell is, cpu power management will prolong the battery life if it's enabled. Unexpectedly, there is a lot of debug/development codes in Sony's BIOS. Now, if you want to go ahead, you have to upgrade the BIOS on your own risk. I'm not suggesting you to do that, ;-). -Aubrey ________________________________ From: Douglas Atique [mailto:d3a...@gmail.com] Sent: Thursday, February 12, 2009 7:56 PM To: Li, Aubrey Cc: Napanda.Pemmaiah at Sun.COM; pm-discuss at opensolaris.org Subject: Re: [pm-discuss] Help with battery consumption Resending after a gmail outage... -- Douglas On Thu, Feb 12, 2009 at 11:54 AM, Douglas Atique <d3atiq at gmail.com<mailto:d3atiq at gmail.com>> wrote: Hi Aubrey, Please find the 4 files attached. CAVEAT: In the mean time, I have live upgraded to snv_107. If this is a problem, please let me know and I'll boot on snv_105 and resubmit the files. Cheers, Douglas On Thu, Feb 12, 2009 at 6:12 AM, Li, Aubrey <aubrey.li<http://aubrey.li>@intel.com<http://intel.com>> wrote: The error is caused by the incompatible Compiler and Disassembler. That means the AML in the BIOS doesn't match the tool you are using. Which version of iasl are you using? I didn't encounter the error. See the attachement. So, now, on the same box, I need Douglas run the following command: # ./acpidump --addr 0x7F6912E8 --length 0x00000240 > ist0.dat # ./acpidump --addr 0x7F691528 --length 0x000000DD > ist1.dat And send the two files to me. I also need some log messages: # dmesg > log1.txt # kstat -m cpu_info | grep supported_frequencies_Hz > log2.txt. 4 files, please, :) Thanks, -Aubrey ________________________________ From: Napanda.Pemmaiah at Sun.COM [mailto:Napanda.Pemmaiah at Sun.COM<mailto:napanda.pemma...@sun.com>] Sent: Thursday, February 12, 2009 8:18 AM To: Douglas Atique Cc: Li, Aubrey; pm-discuss at opensolaris.org<mailto:pm-discuss at opensolaris.org> Subject: Re: [pm-discuss] Help with battery consumption Tried looking into the ACPI tables. There were some method argument count mismatch errors for methods DBGC and BCEN. I don't know if those matters. I assume it should not, Aubrey might know better. Also looks like the ACPI P-State related information like PSS (which exports different supported freq's) is dynamically loaded and may have to do another acpidump to dump P-State related information. I would wait for Aubrey to confirm if the issues related to some method errors is a problem or not before dumping P-State related information. Thanks Anup On 02/11/09 07:38, Douglas Atique wrote: Hi Aubrey, So you can learn all about my ACPI with this tool? Cool :-) The output is attached. Regards, Douglas On Wed, Feb 11, 2009 at 3:12 PM, Li, Aubrey <aubrey.li<http://aubrey.li>@intel.com<http://intel.com>> wrote: Hi Douglas, Yeah, from cpuid info, the CPU on your box supports speedstep indeed. But if BIOS doesn't export any related information out to the OS, OS still can't do anything. So Let's go ahead to check your box: Please download and install acpidump. http://www.ibiblio.org/pub/solaris/blastwave/unstable/i386/5.9/acpidump-20071116,REV=2008.05.10-SunOS5.8-i386-CSW.pkg.gz And run the following command as root. # ./acpidump > dump.out Please send me the file "dump.out". Thanks, -Aubrey ________________________________ From: Douglas Atique [mailto:d3atiq at gmail.com<mailto:d3a...@gmail.com>] Sent: Wednesday, February 11, 2009 7:35 PM To: Li, Aubrey Cc: Mark Haywood; pm-discuss at opensolaris.org<mailto:pm-discuss at opensolaris.org> Subject: Re: [pm-discuss] Help with battery consumption Hi Aubrey, Apparently it supports "Enhanced SpeedStep" (as claimed by the manufacturer: http://www.vaio.sony.co.uk/view/ShowProduct.action?product=VGN-N38Z%2FW&productFilters=retired&site=voe_en_GB_cons&pageType=Overview&category=VN+N+Series). Shouldn't that work? I don't recall exactly from which build on the battery would go away so quicky. I remember that was after ZFS boot (snv_90) and I would say it was in the last few builds, but I cannot recall exactly. Does that help at all? Cheers, Douglas On Tue, Feb 10, 2009 at 1:27 PM, Li, Aubrey <aubrey.li<http://aubrey.li>@intel.com<http://intel.com>> wrote: Hi Douglas, As for cpu power mangement, if your laptop doesn't support Speedstep. I'm afraid I can't do any further help. Can you confirm that on snv_69 so that we can make sure speedstep doesn't have a regression. There are still other components on the motherboard, the related experts here might be able to help. Thanks, -Aubrey ________________________________ From: Douglas Atique [mailto:d3atiq at gmail.com<mailto:d3a...@gmail.com>] Sent: Tuesday, February 10, 2009 8:36 PM To: Li, Aubrey Cc: Mark Haywood; pm-discuss at opensolaris.org<mailto:pm-discuss at opensolaris.org> Subject: Re: [pm-discuss] Help with battery consumption Looks like this last e-mail didn't make it to the list. Did you get it? -- Douglas On Wed, Feb 4, 2009 at 11:59 AM, Douglas Atique <d3atiq at gmail.com<mailto:d3atiq at gmail.com>> wrote: Unfortunately, I must have a very cheap (in the worst sense) laptop... My BIOS (Phoenix R0100J4) doesn't allow me to control anything except boot order and power-on passwords. It doesn't even mention SpeedStep. Here's psrinfo output: # psrinfo -pv The physical processor has 2 virtual processors (0 1) x86 (GenuineIntel 6F2 family 6 model 15 step 2 clock 1667 MHz) Intel(r) Core(tm)2 CPU T5500 @ 1.66GHz # -- Douglas On Wed, Feb 4, 2009 at 1:18 AM, Li, Aubrey <aubrey.li<http://aubrey.li>@intel.com<http://intel.com>> wrote: It looks like EIST(speedstep) is not enabled, or it's not supported. Please check BIOS and give us "psrinfo -vp" report. -Aubrey ________________________________ From: Douglas Atique [mailto:d3atiq at gmail.com<mailto:d3a...@gmail.com>] Sent: Wednesday, February 04, 2009 1:02 AM To: Mark Haywood Cc: Li, Aubrey; pm-discuss at opensolaris.org<mailto:pm-discuss at opensolaris.org> Subject: Re: [pm-discuss] Help with battery consumption Looks like it was already enabled... -- Douglas # # Copyright 1996-2002 Sun Microsystems, Inc. All rights reserved. # Use is subject to license terms. # #pragma ident "@(#)power.conf 2.1 02/03/04 SMI" # # Power Management Configuration File # device-dependency-property removable-media /dev/fb autopm default autoS3 default cpu-threshold 1s cpupm enable # Auto-Shutdown Idle(min) Start/Finish(hh:mm) Behavior autoshutdown 30 9:00 9:00 noshutdown ~ ~ ~ ~ ~ ~ ~ "power.conf" 16 lines, 411 characters On Tue, Feb 3, 2009 at 4:58 PM, Mark Haywood <Mark.Haywood at sun.com<mailto:Mark.Haywood at sun.com>> wrote: Douglas Atique wrote: I have used this laptop with SXCE since snv_69. I have noticed this high battery consumption with more recent builds (say snv_100), but I am not exactly sure which one. I am not sure if cpupm is enabled. I didn't do anything to enable it. Here is an output of powertop running on my machine. Look at your /etc/power.conf file and see if you see a line like: cpupm enable Also, a line like: cpu-threshold 1s would be good. And make sure SpeedStep is enable in your BIOS setup. I assume that it is. Mark -- Douglas OpenSolaris PowerTOP version 1.1 Cn Avg residency P-states (frequencies) C0 (cpu running) (15.8%) 1662 Mhz 100.0% C1 1.3ms (84.2%) Wakeups-from-idle per second: 660.8 interval: 5.1s no ACPI power usage estimate available Top causes for wakeups: 40.0% (264.3) <kernel> : genunix`cv_wakeup 15.1% (100.0) <kernel> : genunix`clock 13.4% ( 88.4) <interrupt> : ath#0 7.1% ( 47.2) <interrupt> : i8042#0 5.7% ( 37.5) sched : <cross calls> 3.9% ( 26.0) <kernel> : genunix`realitexpire 1.8% ( 11.7) <kernel> : uhci`uhci_handle_root_hub_status_change 1.5% ( 10.1) <kernel> : ata`ghd_timeout 0.7% ( 4.9) <kernel> : myk`gem_mii_link_watcher 0.7% ( 4.7) <kernel> : genunix`lwp_timer_timeout On Tue, Feb 3, 2009 at 1:09 PM, Li, Aubrey <aubrey.li<http://aubrey.li> <http://aubrey.li>@intel.com<http://intel.com> <http://intel.com>> wrote: Douglas Atique wrote: > Hi, > > I have a Sony Vaio VGN-N38Z laptop which used to work well for > about 2 hours. Which build did you use to make your laptop working well for about 2 hours? >However, in the latest SXCE builds (I use > snv_105 now) I have noticed that once disconnected from AC > power, the battery indicator in the Gnome bar starts to > decrease almost like a clock and in about 20 minutes it is > already asking me to recharge. I started suspecting that some > power management-related change is the cause of this behavior > because the laptop has been working warm all the time. > > Can I do something to diagnose a possible problem? If anyone > wants to take a look, I offer my time for testing. Just let me know > what I have to do. > > Cheers, > Douglas Did you enable cpupm(speedstep)? What does powertop say? Thanks, -Aubrey ------------------------------------------------------------------------ _______________________________________________ pm-discuss mailing list pm-discuss at opensolaris.org<mailto:pm-discuss at opensolaris.org> http://mail.opensolaris.org/mailman/listinfo/pm-discuss ________________________________ _______________________________________________ pm-discuss mailing list pm-discuss at opensolaris.org<mailto:pm-discuss at opensolaris.org> http://mail.opensolaris.org/mailman/listinfo/pm-discuss -- Anup Pemmaiah Sun Microsystems -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://mail.opensolaris.org/pipermail/pm-discuss/attachments/20090213/8bb769cd/attachment.html>
