Vista SP2 INTERNAL_POWER_ERROR bluescreen on hibernate
I applied SP2 with no apparent problems, then tried suspend/resume and hibernate (since hibernation had been problematic on this machine after SP1 was applied (HP dv2214 with Vista Home Premium-- prior to SP1, hibernation and suspend/resume worked perfectly. With SP1, about half the time hibernate would fail during shutdown and would result in a restart with the 'Windows did not shut down...' screen).
With SP2 applied, on the first couple of trials hibernation (and suspend/resume) worked fine. On third tray, the laptop produced bluescreen with INTERNAL_POWER_ERROR displayed during the shutdown during hibernation, then restarted with 'Windows did not shut down successfully' screen. This seems like the same behavior as SP1, though with SP1 I don't recall ever seeing the INTERNAL_POWER_ERROR bluescreen.
This is Vista SP2 applied to Vista Home Premium SP1. Unfortunately I wasn't able to copy the stop error before it timed out and went to the 'Windows did not shut down sucessfully' screen, although I read that a crash dump was prepared. Any suggestions?
Tutte le risposte
- Do you have a memory.dmp file in C:\Windows directory?
There doesn't appear to be.... I did however install the Feedback Data Collector and was able to reproduce a similar problem. This time the machine hung with fans running when it began shutdown after I initiated hibernate.. I noticed that hibernate appears to work correctly (I tried at least a dozen times) UNITL the machine has been suspended/resumed. Then (after resuming from suspend) shutdown during hibernate is likely to fail. Could be a coincidence, but maybe significant.
I've seen many reports of hibernation issues (many of which describe failures to initiate hibernation or failures on resume from hibernate) but the key here in my scenario is that hibernate fails during the shutdown process. Also it often works.. but as I described above it seems much more likely to fail if the hibernation has been initiated once the machine has previously been suspended, then resumed. I had similar problems since Vista SP1 (the machine as delivered, with Vista pre-SP1, hibernated perfectly) but never noticed the suspend/resume as a significant factor... at that time it just seemed like randomly hibernate would fail, so I eventually uninstalled SP1 hoping the problem would be addressed in SP2.
Searching on INTERNAL_POWER_ERROR produced http://support.microsoft.com/kb/942524 which describes:
Stop error message when you put a Windows Vista-based computer in hibernation: "STOP: 0x000000A0 (0x00000009, 0xc0000001, 0x00000001, 0x00000000) INTERNAL_POWER_ERROR (a0)"
Hotfix Download Available
I wonder if this hotfix got left out of SP2, or is a similar one required for the Nvidia SATA controllers (hotfix mentions Intel SATA 2.5" disk controllers-- my machine uses Nvidia storage controller).Burdy,
My suggestion would be a couple of things:
First, you can go on HP's website and see if there's any downloads for your exact machine that is a required update for any nVidia controller, Video card adaptor, and see if reinstalling those two drivers help alleviate these issues under SP2.
Second, do you have the ability to restore your machine back to factory-installed defaults? If so, you can check under "Programs and Features" to see if HP had, in fact, installed that particular hotfix on that machine. Or you could ask in the HP forums this question.
Last, please make sure you update at least the following components to your machine: BIOS, nForce controller, Video, and nForce LAN drivers (I looked at your system on HP's site and they have two separate nForce downloads, make sure you update both. I also recommend getting the newest updates from nVidia, as HP doesn't update their software on a regular basis).
Let us know what you find out.
Best,
Kamika007z
Thanks for the suggestions Kamika007z,
Back when I first experienced this behavior (though I hadn't observed the sleep / hibernate interaction) I made sure I had all the latest HP updates (including bios and nForce chipset as well as networking drivers, video drivers; in fact I applied all hp updates). I double checked to make sure there were no recent updates and I'm current with all the updates.
Per Microsoft, the hotfix came out May '08 (after Vista SP1) for Intel SATA 2.5" disk controller machines, so it would not have been on my AMD-based HP laptop (which came out with Vista as delivered back in Feb. '07). But it sure looks like there is a similar scenario affecting the NVidia chipset...
Also, I can confirm that *if I use hibernate exclusively, and do not use SLEEP* I do not see the problem-- for the last 3 days I used hibernate at least a dozen times successfully. This morning, as a test, I suspended the machine by selecting SLEEP. The machine went into sleep mode (power light flashing) and then I was able to resume without a problem. THEN I selected hibernate, and saw the following scenario: screen went black, hard disk activity (as machine started hibernation), wireless light went from blue (enabled) to orange (disabled), then NO more hard disk activity, machine appeared to hang for several minutes followed by INTERNAL_POWER_ERROR bluescreen (STOP: 0x000000A0), and restart with "Windows did not shut down successfully". I note that I can also successfully use sleep/resume and as long as I do not try to hibernate, I can use sleep. HOWEVER I cannot successfully hibernate if I have used SLEEP since the last cold boot.Burdy-
This is an interesting scenario and I appreciate the thoroughness you have shown so far. Are there any other error codes you see along with the STOP 0xA0? That would get us in the right direction.
You may also want to try to disable devices (Bluetooth, WiFi, etc) and see if it repros. This could help narrow down the issue.
Thanks,
John
I also wanted to clarify something. Which of the following work? I put in parenteses what I understood so far.
1. Repeated sleep/resume without any hibernation. (Works)
2. Repeated hibernation without any sleep/resume. (Works)
3. Repeated hibernation followed by sleep. (Fails)
4. Repeated sleep/resume followed by hibernate. (Works)
5. Repeated sleep/resume followed by hibernate followed by sleep (?)
Thanks,
John
Two more things:
1. Uninstall Sp2 and confirm the issue(s) no longer repro
2. On Sp1, install http://support.microsoft.com/kb/954434 and see if the issue(s) begin to repro again.
Thanks again
John
Hi John,
I may not have made myself clear, but the case that fails is *hibernation* after sleep/resume (it fails to shut down during hibernate, but bluescreens with INTERNAL_POWER_ERROR and restarts).
On my setup, your test cases would read (assuming each has begun by starting the machine from a complete shutdown state):
1. Yes - Repeated sleep/resume without any hibernation. (Works)
2. Yes - Repeated hibernation without any sleep/resume. (Works)
3. No-- I verified that repeated sleep/resume works on my machine after repeated hibernation
4. No-- This is the case that makes hibernation fail--> sleep/resume followed by hibernate
5. No-- Because of 4, it's game over after the hibernation attempt.
Thanks for the clarification, Burdy. It was more likely me misreading you than you being unclear
I look forward to hearing about any more details (read: numbers from the BSOD) from the 0xA0 as well as the repro results of that KB. In the meantime, we are working on our own on-site repro.
Thanks,
John
Hi John,
As I mentioned in post 6, the issues with hibernation first began when I applied SP1 (via windows update) to my 'stock' HP dv2214us laptop which came pre-installed with Vista shortly after Vista became GA. As delivered, I never saw any suspend/resume issues or hibernate issues with Vista-- it worked great (much more reliable than my XP desktop).
Immediately after SP1 was applied (via Windows Update), I had problems with getting crashes on hibernate, though I didn't notice, or look for, the 'hibernate following suspend/resume will crash' dependency. I couldn't live without hibernate, so I uninstalled SP1..... and I *still* had problems with hibernate BSOD. I assumed at the time that something that SP1 installed did not get 'undone' when SP1 was uninstalled. I really liked the way that my machine was able to quickly suspend/resume and hibernate with Vista (pre-SP1), and since I read of others having hibernation issues with SP1 (includig one user who resorted to factor restore to get normal operation), I used the 'factory restore' partition on my dv2214us and voila-- with the machine *restored to the factory state*, Vista again functioned perfectly with suspend/resume and hibernate. I gave up on SP1 but was happy since I got suspend/resume/hibernate working again.
Flash foward to December, in order to see if the issue had finally been resolved with SP2, I then (this is on 12/5/08) re-applied SP1 (via windows update), verified that the suspend/resume/hibernate crashes still occurred (they did) and went on to apply SP2 (again via Windows Update). I then installed the FDC tool (and reported this as a bug on Connect) and was able to notice the dependency on the sequence of suspend/resume followed by hibernate to produce the INTERNAL_POWER_ERROR bluescreen.
So... unless you think this would *really* be helpful, I don't think the issues would go away by uninstalling SP2 (since they first surfaced with SP1). In order to verify the 'no repro' I'd have to do a factory restore (very painful and long process to get up and running again, since all apps need to be re-installed), and I'm not sure what kb954434 would do, since it doesn't seem to apply to my scenario (my computer never successfully hibernates-- kb954434 seems to address issues of resuming from hibernation). Anything besides this that I can do to help?
Since the behavior seems to be completely repeatable, I can create the BSOD at will. I notice that it takes at least 4-5 minutes for the BSOD to appear after initiating hibernate:
The numbers following the STOP '0x000000A0' are 0x00000009 0xC0000001 0x00000001 0x00000000
These are repeatable also.
I notice the screen says a crash dump is being prepared for analysis, but I can't find it... there is a minidump directory in windows but it appears to be empty. I attached the output of the FDC tool to my bug report on Connect a few days ago. I can re-run it if necessary.
Do you have the latest BIOS for this system? 0xC0000001 usually is a BIOS issue.
Thanks,
John
Yes, I've kept it current and it matches the latest on HP's site (F.39, 11-2007).
I assume you saw my note about this hotfix for Vista SP1 for INTEL SATA 2.5" disk controller -based machines. The description of the problem matches my scenario as good as any I've seen, since it describes a failure to turn off during hibernate (not a resume from hibernation):
http://support.microsoft.com/kb/942524:
Stop error message when you put a Windows Vista-based computer in hibernation: "STOP: 0x000000A0 (0x00000009, 0xc0000001, 0x00000001, 0x00000000) INTERNAL_POWER_ERROR (a0)"
Hotfix Download AvailableI just now note the data match those produced by my machines BSOD exactly... my AMD based dv2214us uses nVidia (nForce) controller... could a similar fix be required here for AMD machines using nForce hw??
- I have this same type of problem on Dell's Opti 755's. I have updated using Dell Drivers and tried to use Intels latest Drivers. I haven't found a workaround except for changing the Power Mode in the the BIOS. I had to change S3 to S1, again this to me is a workaround and not a fix. I have had to do this on about 25 machines so far.
- PaulMoose said:
I have this same type of problem on Dell's Opti 755's. I have updated using Dell Drivers and tried to use Intels latest Drivers. I haven't found a workaround except for changing the Power Mode in the the BIOS. I had to change S3 to S1, again this to me is a workaround and not a fix. I have had to do this on about 25 machines so far.
Just wondering, if you are getting the INTERNAL_POWER_ERROR with the hex crash data shown above, have you looked into the SP1 hotfix (kb942524)? The failure I get produces the same data and error codes, but the hotfix is for Intel, not Nvidia chipsets. - I'm having the same exact problem (INTERNAL_POWER_ERROR, same hex codes), with the same circumstances you described, on my HP dv2210us with an nvidia chipset. The issue has been consistant, even before I installed SP1. I haven't tried SP2 yet. I've been trying to solve this one for a little over a year now. I can't believe more people aren't having this problem and complaining.
- For what its worth folks, had a similar problem and found a solution. For me, it turned out to be the system trying to turn off my networking controller to "save power". See my full blog post. I include some links to pages I've poured over trying to fix this as well as some of the solutions I tried which did not help.
As I said - for what its worth.
- I have an Acer TravelMate 7720g and also have the same problem with SP2? Looks like its a microsoft issue...
- John E [MSFT]2 said:
Thanks for the clarification, Burdy. It was more likely me misreading you than you being unclear
I look forward to hearing about any more details (read: numbers from the BSOD) from the 0xA0 as well as the repro results of that KB. In the meantime, we are working on our own on-site repro.
Thanks,
John
John,Has any progress been made toward the resolution of this issue? I am also experiencing it, each time that I hibernate.Thanks very much.
Aaron Alton | thehobt.blogspot.com I have the same problem that Burdy, after I installed SP1 the problem occurred on my HP laptop DV2201, I have AMD motherboard and video controller, all my drivers are updated. I even installed SP2 but it did not help... If someone find out how we can fix this problem please post it !
Graham.- I also had this problem when I tried to put my computer in hibernate mode. It has only occurred once so far. I don't have the hibernate option on my shutdown menu, instead I have a "deep sleep" that the computer will go into when I choose sleep. However I sometimes will choose to hibernate from the command line like this:
shutdown /h
Anyway I did that earlier this month and I saw the INTERNAL_POWER_ERROR blue screen shortly after that.
The exact message is the very same as described in KB 942524 , however it's noted that:
"This issue occurs if the computer has Intel 82801G ICH7-M Serial ATA storage controllers that are connected to the 2.5-inch hard disks."
I have a 3.5" SATA disk, SAMSUNG HD753LJ ATA Device connected to Intel(R) ICH10 Family 4 port Serial ATA Storage Controller 1 - 3A20. I believe I have native command query (NCQ) enabled. My motherboard is a Gigabyte GA-EP45-UD3P.
I'm running Vista SP2 x86. For the most part I can hibernate fine. And I also routinely run Memtest86+, never a fail.
Mini080309-01.dmp 8/3/2009 8:07:32 PM INTERNAL_POWER_ERROR 0x000000a0 0x00000009 0xc0000001 0x00000001 0x00000000 ntkrnlpa.exe ntkrnlpa.exe+cdb0d NT Kernel & System Microsoft® Windows® Operating System Microsoft Corporation 6.0.6002.18005 (lh_sp2rtm.090410-1830)
Thanks
- Burdy, i am experiencing this EXACT issue on my hp 3000z laptop post installation of SP1 and thus SP2. Did you ever come to a resolution? i tried posted suggestions of disabling "allow the computer to put this device to sleep" for the nVidia networking controller. This solved my "can not hibernate at all" issue but alas it still BSOD with "INTERNAL_POWER_ERROR" if i sleep THEN hibernate- which is the default functionailty in Vista- so i had to set the "Hibernate After" setting to like 10 days to ensure the battery would die before ever reaching that time. lame- but the only workaround i find. Would LOVE the functionality of Vista pre-SP1 which worked a treat!
Cheers,
Q - Q, sorry no. In fact, the same exact behavior persists in the Windows 7 release candidate.
- Some more interesting findings:
1. bad news here: http://support.microsoft.com/kb/969853/
these are the exact stop codes i have gotten since the install of SP1. No solution from microsoft yet.
2. i began to suspect the video card drivers after further online reading. So i changed my graphics driver to "Standard VGA Controller."
Interesting results...
a. i am now unable to Sleep because the option was removed from the Shut Down menu as well as my Power Control Panel.
b. repeated hibernation succeeds as before except now i get a flashing underscore cursor at the upper left of an otherwise black screen while it is entering hibernation.
Those two things tell me that the video driver drastically effects Power management. I was previously using the most current Vista32 drivers from HP (softpak sp34747) for my specific laptop (CTO V3000z) (released 2-15-2007). I then tried the most current nForce 430 drivers from nVidia version 15.01G (released July 2007) which did not come with "Geforce Go 6150" drivers but a Geforce 6150 and Geforce Go 6100. Unfortunately, All tested drivers resulted the same as the HP driver with the INTERNAL_POWER_ERROR and stop codes.
drivers tested:
1. HP Softpak 34747 NVIDIA GeForce Go 6150 (7.15.10.9759 Rev. B) (1/13/07)
2. nVidia 15.01G GeForce 6150 (7.15.11.5824) (4-26-2007)
3. nVidia 15.01G GeForce Go 6100 (7.15.11.5824) (4-26-2007)
4. Geforce 6150 (MS - WDDM) (7.14.10.9687)
5. Geforce 6100 nForce 430 (MS - WDDM) (7.14.10.9687) (8-22-2006)
6. Geforce 6100 nForce 400 (MS - WDDM) (7.14.10.9687) (8-22-2006)
My newest solution is to remove "Sleep" options altogether from power profiles by manually typing "Never" in the "Sleep AFter" fields and setting Hibernate to activate when sleep normally would. This way i retain battery life longer and on the same schedule i was used to. The only drawback is slightly longer "sleep" and "wake" times. But no bluescreens and prolonged battery sell that solution for me until i find a better one.
Until next time...
Q - :(
This happened again a few days ago:
Mini100909-01.dmp 10/9/2009 8:11:27 PM INTERNAL_POWER_ERROR 0x000000a0 0x00000009 0xc0000001 0x00000001 0x00000000 ntkrnlpa.exe ntkrnlpa.exe+cdb0d NT Kernel & System Microsoft® Windows® Operating System Microsoft Corporation 6.0.6002.18005 (lh_sp2rtm.090410-1830)
Same address in stack:
ntkrnlpa.exe ntkrnlpa.exe+2d7c81 0x8224f000 0x82608000 0x003b9000 0x49e0199e 4/11/2009 12:16:30 AM Microsoft® Windows® Operating System NT Kernel & System 6.0.6002.18005 (lh_sp2rtm.090410-1830) Microsoft Corporation C:\Windows\system32\ntkrnlpa.exe
Since that crash I have updated my Gigabyte BIOS to a beta BIOS, F10f, and then I switched to AHCI mode. The beta BIOS includes AHCI 1.20e. Then I went to intel's website and downloaded their AHCI driver. Now I've got this:
Intel(R) ICH10R SATA AHCI Controller, iaStor.sys, 8.9.0.1023, 6/4/2009
So I'll let everyone know how it goes. I also updated my graphics card driver, nvidia 7600gs; although I doubt that has anything to do with it, it can't hurt.
notes:
Anyone switching from IDE mode to AHCI mode, your computer will bluescreen (unrelated to the issue in this thread) unless you enable the AHCI driver _beforehand_. If you're going to try AHCI do this as admin:
sc config msahci start= boot
you should see "[SC] ChangeServiceConfig SUCCESS"
Here's hoping I don't lose any more work...
