|
Bugcheck 9F A Win 8.1 Pro x64 PC of mine has a nasty habit of blanking the display, and not waking up afterwards without a held down PWR button or a bugcheck restart.
This morning I changed the "Turn off the display:" to "Never" and saved it, but this evening, it's back to 20 mins. Anyone knows of some Random Access Gremlins that randomly change settings?
In advance settings, pretty much every power saving option is either disabled or off.
Can someone take a look at the attached dump files? I suspect it's the video driver (AMD HD7750 based video) but I'm mot certain.
I also have a regular MEMORY.DMP, but even compressed, it's too big to be attached. (75 MB)
TIA
|
|
dib22 join:2002-01-27 Kansas City, MO 1 edit |
dib22
Member
2014-Dec-3 12:27 am
said by aurgathor:I suspect it's the video driver (AMD HD7750 based video) but I'm mot certain. I see it randomly on one of my desktops with AMD graphics (R9 270) as well... still haven't tracked it down but also have placed blame on the amd drivers since it doesn't ever happen on a desktop with nvidia. Keep waiting for amd/ati to release the next version of catalyst, however the black locked screen of doom has not happened often enough to try the beta video drivers yet. edit: the machines are also running Win 8.1 Pro x64 both intel chipsets one with amd graphics card one with nvidia |
|
|
to aurgathor
yes, I'd agree that it does look like it's caused by this driver. Looks like the version of the driver you have is from July. I see that they just released an updated one in December. Worth giving that a shot. » support.amd.com/en-us/do ··· 8.1+-+64 |
|
|
|
Well, I did try several versions, including the then latest from AMDs site to no avail. At the end, I let windowsupdate install a driver over an older AMD version
and that seemed to have fixed 9f bugchecks. (I haven't had a single bugcheck in about 3 weeks)
I believe MS does some of their own testing on 3rd party SW that is released through windowsupdate, so while those drivers may not have all the features, or the latest features, they can be missing some highly irritating bugs, too.
|
|
|
Weird. Any driver you would've downloaded from AMD's website should've been the same as that one or newer. Anyway, glad it's fixed said by aurgathor:I believe MS does some of their own testing on 3rd party SW that is released through windowsupdate, so while those drivers may not have all the features, or the latest features, they can be missing some highly irritating bugs, too That is not correct. Microsoft does no testing of these drivers. That's all up to the company submitting them. They do have to pass a certain set of tests before they are posted on Windows Update, which should include power management tests. Full disclosure: I work for Microsoft. |
|
dib22 join:2002-01-27 Kansas City, MO |
to aurgathor
Just for information, I've not had the black screen of doom since updating drivers to 14.501.1003-141120a-178000C (the Dec 2014 amd/ati release). |
|
|
to RightHere
Well, one thing I can say with 100% certainty that the issue had disappeared after installing the driver from WU among a handful of other patches. So it's entirely possible that the root cause was fixed somewhere else in that batch of patches, although I find that somewhat unlikely.
In any case, I still have all the drivers downloaded and tried from AMD, and I can try them again to see if 9F will come back with those drivers.
I do not work for Microsoft, so I don't know what's needed for drivers to be signed by MS and available from WU, but I'm guessing passing HCK is one of them. |
|
|
RightHere
Premium Member
2014-Dec-29 10:59 pm
said by aurgathor:it's entirely possible that the root cause was fixed somewhere else in that batch of patches, although I find that somewhat unlikely. I'd agree, certainly possible. In any case, I still have all the drivers downloaded and tried from AMD, and I can try them again to see if 9F will come back with those drivers. No, please don't! You've got it working now, don't jinx it! I don't know what's needed for drivers to be signed by MS and available from WU, but I'm guessing passing HCK is one of them. Correct. The point I was trying to make is that the testing is not performed by Microsoft. More info for anyone interested - » msdn.microsoft.com/en-us ··· hh833788 |
|
|
Well, 9F is back. However, I may have found something interesting in the event log wherein the power policy was reset -- that would explain why I'm seeing that "Turn off display" is being changed to 20 mins.... How can I figure out who calls WinSAT.exe? The only things I did today with the system: a) installed HDTUne Pro Trial b) disconnected the optical drive and connected an SSD instead c) rebooted said PC with GPARTED and formatted above SSD with it d) rebooted, then tested said SSD with HDTune e) removed SSD, reconnected optical f) let it idle, and when I tried to use it, it wouldn't wake up
The analysis is pretty much the same, pointing to the atikmpag driver (attached)
|
|
|
It's setting the same policy, which is "power saver" policy. As long as the 20m timeout is getting saved in that policy, this shouldn't be a problem. So there may be another issue with something changing the policy, or that setting not sticking for some reason. As far as your bugcheck, ok, NOW you should go get the latest display driver from AMD |
|
|
The 20 mins timeout is actually a major issue. If the display is turned off for any reason, it is not possible to turn it back without a bugcheck or powercycle. I just tested that by setting it to 5 mins, and lo and behold, I ended up with another 9F. I'm thinking there may be something to do with the video card's firmware because it's a weird iron: a Visiontek low power card that is certified to work in SFF Dells. I think that PC needs a video card that draws 25W or less. A plain vanilla 6670 based card would not work in that PC, but a Dell certified and recommended 7750 would. Go figure... In any case, I'll try the latest driver, but I'm at least 90% certain about the outcome.... |
|