BSOD "Stop" and "page fault in non.paged area" w23k.sys and nv4_disp.sys

Status
Not open for further replies.

Theo343

Posts: 22   +0
BSOD "Stop" and "page fault in non-paged area" w32k.sys and nv4_disp.dll ++

I also get:
- irc_not_less_or_equal
- memory_management, system halt/stop

I've been having trouble since the start with this PC. Its newly bought SocketA and it started making trouble from the install of winxp where I first had trouble finding the right SATA(boot hdd) driver(nvidia mcps, not si or promise etc) to include on CD install since the floppy drive ive bought would work, just thje floppy led was lit all the time and not working. Tried many things to correct the floppy but ended up by making a custom install cd with nlite.

Many installation(winxp sp2 and without) didnt find hdd as I had trouble findnig the right sata driver, and many that did work got BSOD during "Hardware detection" process in the GUI setip, or around T-30 to 39.

Ive been working on the problem for a week now and im getting kinda exhausted and need some expert advice.

List of my HW:
-----------------------------------
Chenbro GamingBomb (watercooled system)
Tagan 480W PSU w/act. pfc
DFI NFII U400SG-AGF (DFI NFII Ultra Infinity is on warr rep for 7 months now)
XP-M2400@11.5*218/1.675(FJQC4)
Kingston HyperX PC3200 2*512MB 2-2-2-5
Leadtek/Winfast 6600GT-TDH 128MB, AGP.
250GB Seagate barracuda, SATA.
BenQ DW1640 16X, black
Razer Diamondback Chameleon
Sony Trinitron CRT 19"

What ive tried so far:
-----------------------------------
Ive tried a thousan things buit ill try to summarize here.
- Installed XP 12-15 times to get a running XP with nlite tool (started out to advance and had to make many cd sources to sucseed)
- Ended up with an nlite with winxp sp2 english and MS ide drivers(I think)
- There are 2 bios for this MB, ive tried both.
- Ive tried many safe settings in the Bios currently I have no shadowing, no agp fastwrite, CPU is at FSB 166, Ram is at spd 200, MP have been from 8-11, vcore is at 1.375(1.35v 35w cpu). but ive tried many many Bios tweaks.
- Ive done many things with the cooling to get optimal cooling allover and I never vet a temp outside 30 to 40+. (everything is watercooled, can list equip if wanted)
- Im running DX9.0c and Nvidia nforce UD 5.10, Forceware 78.01 also tried newest 81.84? 77.77 wouldnt install.
- Ive run memtest and many utils from Hirens Bootcd and Ultimate Bootcd
- Ive reseated ram and change sata controller on the hdd. Was running ram in dual channle seating and now trying single channel seating.
- ive been remodelling and optimizing the cooling system 5 times to make sure thats is not an issue. Some fans ar running at 5v and some at 12v, had some on 6v mod aswell(serial).
- Ive defragged the partitions(sys-10gb/data-230gb) with diskeeper lite 9.
- Ive ran chkdsk and other disk utils and it showed 48kB of bad cluster or something, never semmed to be an issue before adter running diskeeper..?
- tried to disable sound acc in dxdiag, many safe settings in forceware driver aswell
- turned off all powermanagement on HDD etc. SMART enabled and disabled.
- many of the memtest and dos diags never finished with a status cause the pc would hang in a black screen when I returned some hours later.

Symptoms:
- Bios behaves weirdly sometimes loosing HDD settings. And sometimes it shows HDD as 136GB(even windows sees it as 250GB) and sometimes 250GB
- Some games can run for a while like EVE-Online or Brothers In Arms demo, others like Empire III demo crash almost immediatly.
- Before(1 week old) it would run good if I didnt press it to much, but now it goes to BSOD after a while even if im not using it.
- As mentioned it seems it fails more on disk activity now than before aswell?
- Seems I cant find any source
- A weirt thing happened with Hirens Bootcd the other day. It got stack overflow on any program and only displayed jiberish chars on the screen until it stopped. Its also not able to run all programs and sometimes hangs after ramdrive load.
- 3dmark03 would finish the demo but when running the benchmark it got "corrupt datafile" its installed on c:(sys).
- Its also c: that chkdsk corrected a lot of issues that suddenly had appeared.
- 3dmark05 wont even run demo before it goed into BSOD with "System halted... Stop" and I get diff codes like 0x0000008e, ...4E, ...ED etc.
- I tried doing a disk check in Partition Magiq 8.5 and I got some sector there that I fixed and also error 701 that I couldnt find any info about.
- Sandra 2005 sr2 didnt finish memory benchmark first time but after reboot it did.


Provided:
- Ive uploaded some of my minidumps so if you would please take a look to give me some hints I would be very grateful.
- Ive uploaded dxdiag.txt

I dont have to much hardware to test with but ill see what I can do with som directed tips.

PS!
When I try to upload minidum.zip I get "document contains no data" :(
 

Attachments

  • Mini101305-07.doc
    88 KB · Views: 6
I had to rename the dumps to .doc, and got error on zip. So pls rename them to .dmp. these are far from all but some.
 
This is one of the things that happens if I try to install the Nvidia forceware 77.77 from leadtek/winfast. See attachment.
 

Attachments

  • fw77-77.JPG
    fw77-77.JPG
    35.7 KB · Views: 16
Hi,

I've processed your minidumps. Some crashes are at nVidia display card drivers and some are related to faulty ram. I have no idea whether Vidia Display card driver version 77.77 is stable or not as it is very new. I am sure that version 61.77 is stable.


Suggestion
1. Install nVidia Display Card Driver Version: 61.77 Release Date: July 27, 2004 http://www.nvidia.com/object/winxp_2k_61.77.html
2. Run memtest
 
Thanks, ill install the mentioned nvidia driver and i'll try to complete a ramtest.

I downgraded from 78.01 to 77.77 yesterday(had to give it 3 tries before successfull) and also installed other nforce chipsett drivers(from Unified 5.10 to guru3d remix 2.51) and changed from Nvidia Unified audio drivers to DFI Realtek audio drivers(MB specific) before i called it the night. I got some BSODS when running 3dmark05 after the changes, but it went on much further than ever.

And when i got up my PC was still alive but there was an error msg from AVG "Out of memory".

EDIT:
The 61.77 is very old. Over a year(2004-07-27), thats quite some time regarding 6600GT isnt too old. But i will ofc try it regardless :) Its actuallt the first FW that has DX9.0c support and it got 6800 support, hope it supports 6600GT aswell.
 
Memtest and forceware drivers.

Ive ran MS Windows memory diagnostic(boot floppy) and memtest86 v3.2, but i didnt have time to run them for more than an hour each so ill run them trough the night. Memtest86 v3.2 will run this night.

The FW 61.77 doesnt seemt to have support for 6600GT, only early 6800 models. Ive gathered som intel from major nvidia forums and these seems to be some of the ones i should try. Do you know which of these i should try? Ive gathered some comments from many threads etc.
66.93 - Currently I still get the best performance out of the old Omega tweaked 66.93s - apart from BF2 which doesn't want to know
71.90 - since this release i cannot get a single driver to work right for my card
77.30 - best performance drivers for the 6600 gt series
- best performance drivers for me, check out my rig.
77.74 - are the best for me the 8x.xx it randomly reboots when it goes into 3d mode
77.77 - 77.77 has been very good to me
- many agree
78.05 - You need to be using these (no reason made)
- work great. I had downclocked my card, and obtained the same or more 3dmark scores and css stress test with my old 77.76 or 77.77
81.84 - I have had no problems with my Leadtek 6600GT with any drivers. First time I have broken 50K in Aquamark. Got 53,080 with them.
- quite a few people have been having good results with the
Here are some Manufactor/cpu/driver stats from the same forum. Info on the system of those that adviced to use a certain version.
71.81 - XFX/XP-24
77.30 - asus/A64-32
77.74 - XFX/A64-32
77.77 - XFX/P4-30HT, BFG/A64-32, XFX/XP-28, PNY/P4-22
78.05 - XFX/p4-28, XFX/P4-26HT
78.11 - XFX/XP-24
im also attaching some new minidumps from today. It doesnt seem that every crash produce a minidump.

PS!
Could the reason be unstable power in the building? I sometimes notice flick in the lightbulbs some places in the house. but i guess many have that.

I dont have a UPS yet, but i will get a Online 3000VA after new year.
 
Further work.

As suspected the 61.77 did not support 6600GT, only 6800 series. I removed all nforce chipset drivers and 77.77 display driver.

I reinstalled with chipset remix2.51 thats an updated nforce1&package. And i also chose to install the IDE SW driver. I had to run the packager twice cause it failed the first time. And anfter everything is installed my OS is still using MS IDE drivers(just a curiosity).

Then i installed FW 77.30 which ive heard should be a stable and potent version for 6600GT.

Ive gotten som BSODS after this aswell and im including them here. For the night im gonna run memtest86 v3.2 and hope it survives without hang or blackscreen so i can see some results.

Is there any memtest that logs to floppy or something? I thought memtest could do that.

EDIT:
Hm, seems last BSODs havent been able to make minidumps, but one of the last BSOD messages was:

- A driver corrupted pool memory used for holding pages destined for disk. STOP: 0x000000DE....

Gonna run memtest while sleeping. Cya.
 
Hi Theo343

According to my record, nVidia display version 66.93 is unstable. Version 71.84 has one unconfirmed memory corruption and 77.72 have one unconfirmed bugcheck 8E. Probably the culprit is faulty ram.

Suggestion
1. Check the temperature of the CPU and make sure that it is not overheat (ie temperature < 60C)
2. Reseat the memory stick to another memory slot
3. Downclock the ram
4. Clean the dust inside the computer case
5. Make sure that the ram is compatible to the motherboard
If it still crashes, diagnostic which memory stick is faulty
1. Take out one memory stick. If windows does not crash, the removed memory stick is faulty.
2. If you have only one memory stick, replace the ram
 
Hi again cpc2400 and thanks for your tremendous service, makes me quite pleased to have you around here.

1. I have worked alot with the cooling system and ive never had any temps higher than 43+ celcius. Now im running at 30+ on cpu and all other(4 sensors) under 30. So temps have never been an issiue even on long lasting 100% load.

2. Ive reseated the memory twice and also tried single-channle and dual-channel seatings.

3. Ram is PC3200/200(400) and ive tried setting it at par with CPU FSB(166/333) and it didnt give any better outcome. So it runs at spd/200.

4. There is no dust in my system as it is new and i always think of everything from assembly to location not to get dust in my pc. atleast minumum. Use dustfilters aswell in inhales.

5. The RAM is comp. with the MB.

6. Now to the interesting part, ive ran som memtests during the night.
First i ran MS Windows Memory Diag. And it went ok on pass1. Then i changed to memtest86 v3.2(all floppies since ubcd and hirens gave me weird faults).

First at memtest it seemed fine it went for a couple of hours and no fail so far. Then i did som experiments.

1. i went into config and set "memory sizing" to "bios all" or "probe". On both those settings i started to get fails at once in the 1022.0MB area and up.

2. I set memory sizing back to normal and i changed adress range to all and choose all 9 tests and let ut run trough the night.

When i came back the screen was red full of errors and it was in around 600-622.0MB when i looked at the screen(individual fails setting) and 72%pass, 37% test random sequence and a total of 10.689.539 fails so far.

:eek:
this makes me wounder about a couple of things.

Questions:
1. Do this mean that the ram in slot 2 is defect since it failed around 622 and 1022? or do i have to test specifically each one?

2. it only showd where it failed when i was watching, no "failed at area this and that repport". How do i setup to get a better repport?

3. Ive had problems with floppydrive and bad sectors on disk, all new, could all be because of a bad memory stick? even bad sectors on a brand new 250GB s-ata?

4. Could it be that its the motherbnoard messing up all my hardware? or the PSU(which is qualuity brand 480w tagan)

Im gonna do som stresstests in windows with both of the memstick now, first without the one i took out now, then swap. And see id its any difference.

If one fails im gonne run only that in memtest86 again, if both give same errors im gone run both but one by one in memtest86 just to identify if all or just one is defect. Even though i gotta return both as a kit.

PS!
last two days ive started getting a lot of critical crashes on Firefox, i guess this is either ram or the "bad sector" on chksk. Im not very comfortable with the sectors on the chkdsk.. could it be that bad ram make "virtual" bad sectors?
 
Hi,

If memory occurs at kernel mode, windows crashes with blue screen. If memory error occurs at user mode, windows writes the dump to DrWatson and terminate the thread. Do you have a lot of crash at application within this week (ie check DrWatson log)?

C:\Documents and Settings\All Users\Application Data\Microsoft\Dr Watson\drwtsn32.log
C:\Documents and Settings\All Users\Application Data\Microsoft\Dr Watson\user.dmp
 
DrWatson

Ive gotten a lot of DrWatson(user mode) aswell. Im uploading the logfile zipped here.

It was a 2.9MB file and ive split it manually.
 
RAM Fail news.

Ive been swapping memory now and testing in windows. Running them alone.

The one in slot1 gave huge amounts of errors. Windows alone got error all along.
Then I swapped to the one that used to be in slot2 and so far everything is working great.

- Ive been able to watch the whole 3dmark05 demo without BSOD or hangups
- I could rund the whole 3dmark05 benchmark with a result of 3233 which isnt to bad for my sysconfig i guess.
- And so far no errors anywhere. gonna try some games aswell to see how they behave.

But the "bad sectors" on HDD worries me as i bought this last week.
 
Hi,

I have a quick check on your DrWatson log and several modules crashes with acccess violation. I believe that they are victims of faulty memory. You had better rename DrWatson log as it is acculmated. If your DrWatson log does not have any new entry within this week, your memory problem is fixed. Let me know if you still get BSOD
 
I was gonna tell you yesterday that finally everything seemd to work fine. But as i wrote it my computer freezed totally without any BSOD and errormsg.

But i was able to play a game the rest of the evening.

Ive had 2 or 3 BSODs after that with irq_not_less_or_equal or how it goes.. and tcpip.sys as the file.. never seen that before..

And today ive had 1 tcpip.sys BSOD and 2 total freezups when i used winamp together with firefox or thunderbirt. Never had that before on other working computers(use same software on my laptop).

The only thing that gave som kind of log was one fault on thunderbird. Im attaching that log.

EDIT:
Ive tried both latest winamp and mediaplayer10, and every time i play mp3 while im on the web i get a complete freezeup after about 10-15 mins.
 
I ran memtes86 for 9 hours on the functioning RAM stick and no errors.

I left my computer at the desktop and when i got back this afternoon i was frozen in the desktop picture. I restarted and startet firefox1.0.7 and thunderbird1.0.7.

Before they were able to start i got a BSOD with just STOP/HALT and no specific error msg. And i got no errorlog on this, minidump or drwatson.

I guess the one memory stick i dismounted was faulty, but i still have som serious problems even thoug its not as frequent as before.

Got any ideas?
 
The ram is PC3200 with native 2-2-2-5 timings and is running at native speed at 200Mhz. Why would it need to be downclocked?
 
Some cases donwclock the ram and resolve blue screen problem. You can try downclock the CPU as well.
 
Downclocked RAM and CPU

I kept getting system freezes especially when playing music, video, games almost without exception after 3-30 minutes use.
I also got freezes or BSODs on desktop program use but usually it could take hours or not happen at all.

First i downclocked the CPU to 166/333 i also tried different vcore and default(bios default is wrong, 1.525-def vs 1.350-cpu) since the MB doesnt recognize the CPU type which btw. shouldnt be a problem.

But i still got freezes and BSODs, then i downclocked the RAM aswell to go 1:1 with the CPU on 166/333 so the RAM and FSB also were running at 166/333. AGP has always been locked at 66.

Worked fine for 3 hours until i started a DX based game. 3 minutes and i got "page fault in non-paged area". Same happens when i play a movie og music.

This is starting to be a real problem. How can i diagnose this more precisly? I dont have any spare comp. hardware i can play with, i got a running server which i dont want to mess with. Im gonna get the RAM exchanged for a new kit, so it wont be the problem anymore but its only one stick that got errors in memtest.

And i think its strange that, since we resolved the one RAM stick fault i havent gotten any minidumps or any more drwatsn logs latly on these errors.
 
Hi,

It is hardware error. Open a new problem at hardware forum. Maybe some hardware experts can help you.
 
Status
Not open for further replies.
Back