Need to ditch nForce4 chipset.. any suggestions?

[MDesigner]

I've been having random lockups on my computer. I've swapped out every part except the power supply, which I'm pretty damn sure would not cause system lockups. The voltages are stable. I've swapped out the CPU with a known-working CPU. I've tried known-working RAM. Nothing fixes it. I've been through THREE different motherboards among two different manufacturers. The only thing in common here is the nForce4 chipset. So I'm ditching it.

I have two options.. go for socket AM2 and an nForce 500 series chipset (anyone know if nForce 500 series is improved over the horrible nForce4?), OR, wait a few weeks(?) for Intel's new Conroe (I forget the name they changed it to) and get one of those. Or I suppose I could get a dual core Pentium D "Presler" CPUs..but with Conroe around the corner, that may be a waste.

Suggestions, anyone?

 

[furballi]

Don't believe the lockup is caused by NF4. Have worked on many NF4 chipsets with 754 and 939. All are stable if I don't install the Nvidia FW and Nvidia IDE driver. Go to the Nvidia website and download the latest NF4 chipset/audio drivers.

The NF4 chip runs HOT. If you have poor case cooling, then you may encounter random lockup. Reset BIOS to default and retest.

The NF4 is a proven chipset. I would wait at least 6 months before jumping into AM2. Same logic with NF 500 and Conroe. Let other people work out the bugs.

 

[MDesigner]

Interesting. Can you recommend a good temp monitoring software? MBM5 used to be good..till they stopped development on it. Now it doesn't support newer boards.

I should say, it apparently is not 100% random. Twice, I was able to replicate the lockup by loading up a graphics heavy web page, and scrolling it up and down.. that locked me up. Another time, I minimized a window and it locked up on me. Maybe the video card is behind all this??

 

[furballi]

Monitoring chipset temp can be tricky because it may not be supported by your MB. If you can hang on to the chipset heat sink for more than 30 seconds, then it's probably receiving adequate cooling.

It appears that your problem is video related (driver and/or hardware).

 

[MDesigner]

Could be. It's either the PSU or the video card. Every other part has been swapped out for another for testing purposes. The PSU is an Enermax Liberty 500W, and the vid card is an eVGA GeForce 7900GT. Since the lockups happen more often on my dual core Athlon, I might put that back in tonight and swap the 7900GT for a 6600GT (known working), and see how it fares.

 

[furballi]

Use the latest chipset AND video drivers from the Nvidia website. DO NOT install the FW and IDE driver. I'd wipe the HDD and clean install Windows with the latest drivers.

I would install the drivers in the following order: 1.sound. 2.chipset. 3.video.

 

[MDesigner]

Ugh, the latest nVidia video drivers (91.31) suck. I'll be using 84.21.. you couldn't pay me to use the latest. They're buggier than 84.xx.

OK so I'll wipe the system clean again.. I won't install the IDE drivers this time. I never install the nVidia firewall, it's crap. Any reason you recommend installing the sound before the chipset? I usually do chipset before anything else.

 

[furballi]

Based on personal experience with the NF4 boards. No hard core scientific test to support my approach.

 

[MDesigner]

Well, since you've had more luck with NF4 than I have, I'll install sound first. Then chipset (minus IDE.. and minus audio too since I have my own sound card), then the 84.21 video drivers.

 

[MDesigner]

Oh wait.. I can't test my other vid card This motherboard has no AGP slots, and my 6600GT is AGP.
ARGH!!

EDIT: I'm going to buy a $40 PCI Express vid card.. it'll be good to have around for future testing purposes.

 

[MDesigner]

OK, here's something interesting.

This image: http://lisastoner.com/Main/SkullDrinker-small.gif

causes lockups on my machine CONSISTENTLY.. but only when viewed in Firefox. I finally got a reset (not a lockup) which left a crash dump behind, which I'm going to debug now to find out which driver caused it.

 

[furballi]

The NF4 LAN driver work fine with my rig. You could try another PCI network card to isolate the problem. Scrolling a web page should not affect the LAN.

 

[MDesigner]

Here's the cause of the recent reboot caused by the above image in Firefox. I have no idea what win32k.sys is.. but it looks like a function call to GreSetLayout() caused the crash. Sounds like a graphics related API call?

*******************************************************************************
* *
* Bugcheck Analysis *
* *
*******************************************************************************

KERNEL_MODE_EXCEPTION_NOT_HANDLED_M (1000008e)
This is a very common bugcheck. Usually the exception address pinpoints
the driver/function that caused the problem. Always note this address
as well as the link date of the driver/image that contains this address.
Some common problems are exception code 0x80000003. This means a hard
coded breakpoint or assertion was hit, but this system was booted
/NODEBUG. This is not supposed to happen as developers should never have
hardcoded breakpoints in retail code, but ...
If this happens, make sure a debugger gets connected, and the
system is booted /DEBUG. This will let us see why this breakpoint is
happening.
Arguments:
Arg1: c0000005, The exception code that was not handled
Arg2: bf8243e0, The address that the exception occurred at
Arg3: b6ee3608, Trap Frame
Arg4: 00000000

Debugging Details:
------------------


EXCEPTION_CODE: (NTSTATUS) 0xc0000005 - The instruction at "0x%08lx" referenced memory at "0x%08lx". The memory could not be "%s".

FAULTING_IP:
win32k!GreSetLayout+2f
bf8243e0 56 push esi

TRAP_FRAME: b6ee3608 -- (.trap ffffffffb6ee3608)
ErrCode = 00000002
eax=00ffffff ebx=b6ee36d0 ecx=b6ee36d0 edx=00000000 esi=e353de00 edi=e2dafcdc
eip=bf8243e0 esp=b6ee367c ebp=b6ee368c iopl=0 nv up ei pl zr na po nc
cs=0008 ss=0010 ds=0023 es=0023 fs=0030 gs=0000 efl=00010246
win32k!GreSetLayout+0x2f:
bf8243e0 56 push esi
Resetting default scope

CUSTOMER_CRASH_COUNT: 1

DEFAULT_BUCKET_ID: DRIVER_FAULT

BUGCHECK_STR: 0x8E

LAST_CONTROL_TRANSFER: from bf830d74 to bf8243e0

STACK_TEXT:
b6ee368c bf830d74 e353ddf8 e2dafcd4 00000002 win32k!GreSetLayout+0x2f
b6ee36f0 bf85fe5f 52010f74 00000004 0518dff8 win32k!NtUserBuildHwndList+0x131
b6ee3874 bf86039b 52010f74 00000000 00000028 win32k!xxxTooltipWndProc+0x198
b6ee38ec 8053c808 52010f74 00000000 00000028 win32k!AnimateFade+0xb3
b6ee3934 7c90eb94 badb0d00 0012ef0c 88c941a0 nt!ObpPushStackInfo+0x75
WARNING: Frame IP not in any known module. Following frames may be wrong.
b6ee3940 88c941a0 00000018 b6ee3c7c 00000000 0x7c90eb94
b6ee3944 00000000 b6ee3c7c 00000000 0012fcbc 0x88c941a0


STACK_COMMAND: .bugcheck ; kb

FOLLOWUP_IP:
win32k!GreSetLayout+2f
bf8243e0 56 push esi

FAULTING_SOURCE_CODE:


SYMBOL_STACK_INDEX: 0

FOLLOWUP_NAME: MachineOwner

SYMBOL_NAME: win32k!GreSetLayout+2f

MODULE_NAME: win32k

IMAGE_NAME: win32k.sys

DEBUG_FLR_IMAGE_TIMESTAMP: 43446a58

FAILURE_BUCKET_ID: 0x8E_win32k!GreSetLayout+2f

BUCKET_ID: 0x8E_win32k!GreSetLayout+2f

Followup: MachineOwner
---------

 

[MDesigner]

Just a note, I was able to make the desktop lock up while opening that image in Firefox via Remote Desktop. So basically I was viewing the desktop machine remotely using my laptop. I didn't really expect it to lock up. I'll run another test.. on the desktop, bring up the image but cover it with another window. I need to determine if it's the actual display of the image that's causing the crash, or just drawing it in the browser window, regardless of whether it's actually visible

EDIT: OK, brought the image up (locally, not via Remote Desktop) and covered it immediately with a notepad window. It ran for several minutes like this. Then I moved the notepad window out of the way, the machine locked up within a few seconds. So that answers that...

 

[pkrush]

Sounds like you may have been bitten by the bad 7900gt bug like me. Bought the 7900gt, put it in, everything worked fine for a while and then I started getting huge amounts of lockups in 3d and eventually 2d too. Put my 6600gt back in and everything's completely stable. eVGA's just released a fixed version of the card, but I can't RMA it now since I'm about a thousand miles away at my dad's at the moment.

 

[MDesigner]

Oh.. you mean there was a bad batch of 7900GTs?? Interesting. Well, I ordered a cheapo PCI express video card just to drop into my system to test it out.. JUST to be 100% sure this is the prob. Then I'll RMA the video card immediately.

 

[furballi]

Try to re-create the problem with IE and/or Opera. If okay, then I'd isolate the video/video driver.

 

[MDesigner]

Already done It works perfectly in IE. Firefox must use some different API to draw graphics that's causing a lockup. Also, I've determined it's not the nVidia drivers either.. I booted into safe mode, and brought up the graphic.. same result, lockup. So it's the video card hardware, I'd say. I don't like to assume things, especially when it comes to this stuff, so I'll wait till I get my test video card in the mail before I make a final conclusion.

 

[Operandi]

There was compatibility problem with the nForce4 chipset and a certain run of Maxtor (I think?) hard drives that would cause data corruption and just general instability. I would look into that, and make sure that?s not the problem.

 

[MDesigner]

Nope, I have Seagate.

The problem you're referring to is actually just an nForce4 chipset issue.. some people get a bad board, others don't. My first Abit AN8 was fine. My second one had the nForce4 SATA/IDE corruption problem..basically any data xfer'd via the nForce SATA/IDE controller was corrupted if it was large (over 300MB or so).. the CRC checks failed. But if I plugged my drives into the SiliconImage (BIOS) SATA ports, it was just fine. It's just one of those things you have to deal with, pack up the motherboard, and RMA it. There's no solution for it, it's a hardware defect.

 

[Operandi]

Originally posted by: MDesigner
Nope, I have Seagate.

The problem you're referring to is actually just an nForce4 chipset issue.. some people get a bad board, others don't. My first Abit AN8 was fine. My second one had the nForce4 SATA/IDE corruption problem..basically any data xfer'd via the nForce SATA/IDE controller was corrupted if it was large (over 300MB or so).. the CRC checks failed. But if I plugged my drives into the SiliconImage (BIOS) SATA ports, it was just fine. It's just one of those things you have to deal with, pack up the motherboard, and RMA it. There's no solution for it, it's a hardware defect.

I'm not sure about that... I had to deal with the problem myself and was able to track it down to certain line of hard drives within a certain range of serial numbers.

There was a firmware update for the drive that supposedly addressed the issue but I just decided to buy back the drive and swap it with a different brand. After that the machine was rock solid.

Either way it seems that's not whats causing your issue.

 

[furballi]

Never had any problem with FF and NF4. Make sure you've deleted all profiles and reload FF. Do not test with any extension. Do not import old data into FF. The latest version can be buggy. I'm using 1.0.0.0...perfectly stable.

If Opera is also okay, then the problem is coming from FF.

 

[MDesigner]

Originally posted by: Operandi

Originally posted by: MDesigner
Nope, I have Seagate.

The problem you're referring to is actually just an nForce4 chipset issue.. some people get a bad board, others don't. My first Abit AN8 was fine. My second one had the nForce4 SATA/IDE corruption problem..basically any data xfer'd via the nForce SATA/IDE controller was corrupted if it was large (over 300MB or so).. the CRC checks failed. But if I plugged my drives into the SiliconImage (BIOS) SATA ports, it was just fine. It's just one of those things you have to deal with, pack up the motherboard, and RMA it. There's no solution for it, it's a hardware defect.

I'm not sure about that... I had to deal with the problem myself and was able to track it down to certain line of hard drives within a certain range of serial numbers.

There was a firmware update for the drive that supposedly addressed the issue but I just decided to buy back the drive and swap it with a different brand. After that the machine was rock solid.

Either way it seems that's not whats causing your issue.

No kidding? Interesting. I did read about some people having issues with Maxtor and NF4.. but I also saw some people with other brand drives having the same data corruption issue. I've been using Seagate for a few years now.. I dumped Maxtor some time ago. Now Seagate bought out Maxtor, so I'm wondering what's going to happen there.

 

[MDesigner]

Originally posted by: furballi
Never had any problem with FF and NF4. Make sure you've deleted all profiles and reload FF. Do not test with any extension. Do not import old data into FF. The latest version can be buggy. I'm using 1.0.0.0...perfectly stable.

If Opera is also okay, then the problem is coming from FF.

Although the image locks up only in Firefox, I'd hesitate to say that Firefox itself is actually causing the lockup. I have the exact same version of Firefox on my laptop, and the image does not cause any issues at all. Besides, if it were some problem with Firefox, then I would just expect FF to crash and not bring the whole system to a halt. The whole system locking up is caused by hardware and/or driver issues. A program like Firefox should never cause such a thing.

I think I've also ruled out the power supply. If it were the power supply, I think the lockups/resets would be more random. The fact that I can view a specific image in Firefox and make the system lockup every time pretty much rules out any power supply and/or voltage problems. My money's on the video card.

 

[furballi]

If you don't have issue with games, then it must be FF. It's very easy to corrupt FF with an old profile and/or extension. A clean install of FF is in order.

There is nothing taxing about a browser. IE works, and I suspect Opera is also okay.