Go Back   Cockos Incorporated Forums > REAPER Forums > REAPER Compatibility

Reply
 
Thread Tools Display Modes
Old 09-15-2015, 10:39 AM   #1
Bob Headroom
Human being with feelings
 
Join Date: Apr 2008
Location: UK
Posts: 262
Default PC keeps rebooting itself - can anyone help me diagnose the problem (Windows XP)?

Hi all

For a week or two now I have had a problem whereby my PC keeps rebooting itself. I've installed BlueScreenView so that I can check the MiniDumps, but it seems to give different fault codes and and blames different drivers....

I did not notice any kind of pattern to begin with. Today though, every time I opened Chrome it would reboot after about ten seconds. This happened quiet a few times. But then after running CCleaner and a few scans it sorted itself out and I could use Chrome again. For a while.... Then the reboot happens again, again seemingly randomly.

I have two spare sticks of RAM and have replaced the ones in the system with those. Still getting the problem. So that pretty much rules out RAM.

I've no idea whether the problem is hardware or software related, and because I don't really understand what BlueScreenView is telling me (especially because it tells me different and varied things) I'm not sure how to approach the problem.

I wondered if anyone here more knowledgeable about these things might be able to offer any advice and/or take a look at the MiniDumps to see what the heck's going on! Let me know what ever info you need form me and I will get it up here right away.

Cheers

Max

PS - I know WinXP is a fossil but for now I want to stick with it...

PPS - Unfortunately, after running CCleaner a lot of the Minidumps disappeared. I had saved four of them to a separate folder and so still have those, and I have one more as it has just recently rebooted again. So I currently only have 5 MiniDumps out of the many I should have.
Bob Headroom is offline   Reply With Quote
Old 09-15-2015, 11:08 AM   #2
tweezer
Human being with feelings
 
Join Date: Aug 2010
Location: S.E.England
Posts: 323
Default

Not saying this will be the answer to your problem but I had this happen to me a few years ago and it turned out to be the PSU on its last legs. I must say though that it wasn't when I opened any particular program. It would just happen.
tweezer is offline   Reply With Quote
Old 09-15-2015, 11:12 AM   #3
karbomusic
Human being with feelings
 
karbomusic's Avatar
 
Join Date: May 2009
Posts: 24,041
Default

Quote:
but it seems to give different fault codes and and blames different drivers....
Often a hardware issue/memory corruption/PSU etc. Can you point me to some of those minidumps?
__________________
If it requires a null test to find it, it is by definition minuscule.
karbomusic is online now   Reply With Quote
Old 09-15-2015, 12:08 PM   #4
JHughes
Human being with feelings
 
JHughes's Avatar
 
Join Date: Aug 2007
Location: Too close to Charlotte, NC
Posts: 3,350
Default

Bad power supply or overheating is my guess.
__________________
You can only Reverse, Switch, Invert or Flip POLARITY, not "PHASE".
JHughes is offline   Reply With Quote
Old 09-15-2015, 12:09 PM   #5
innuendo
Human being with feelings
 
Join Date: Nov 2013
Location: Jerusalem, Israel
Posts: 659
Default

Might be anything really. Easiest way to rule out hardware issues would be making a fresh installation on a separate drive and testing with that.
Prior to doing that, worth to check 2 things:
- overheating: check with Aida (built-in stress test) or with Prime95 (stress)+SIW (temperatures monitoring)
- HDD health: check with hddguardian and run a short self-test.

Also despite you ruled out the RAM, I would still go for a one-pass isolated memtest. You would need to create a bootable USB drive for that. Or use Hiren's boot CD, it has memtest built-in in the boot menu. You can use Rufus to create the bootable USB.
innuendo is offline   Reply With Quote
Old 09-15-2015, 12:49 PM   #6
sickamorz
Human being with feelings
 
Join Date: Feb 2015
Location: Philippines
Posts: 739
Default

I had this happen in several years ago on one of my computers.

Check you power supply that the fan is operational.
Have you removed the heat sink, if so are you sure you reinstalled the heat in the correct orientation. It is possible to install the heat sink in the wrong orientation on a xp.

Since you have an xp I assume you have a floppy disk drive, can you boot from a floppy disk drive? If the problem disappears, look at what software is loading at start up.
sickamorz is offline   Reply With Quote
Old 09-16-2015, 02:08 AM   #7
Bob Headroom
Human being with feelings
 
Join Date: Apr 2008
Location: UK
Posts: 262
Default

Hi all

I've stumbled across something today that might well prove helpful. I have noticed that it is one particular web page that is causing the PC to reboot instantly. I have been reading reviews for waterproof jackets and this morning I opened the page of a review I've been looking at previously. The PC rebooted instantly. I repeated it a couple of times with the exact same results. When the PC rebooted every time I opened Chrome yesterday, that particular webpage was one of my open tabs and is therefore likely to be the culprit.

I will post the webpage link here but - JUST IN CASE IT IS DODGY - please treat with caution. I don't think it is tbh, but I'm no expert. Whatsmore, I have opened it on the same PC using Opera 12.17 and it is fine. It opens without problem on my two tablets as well. But ultimately I don't want to cause problems for anyone else. I'm just posting the page link in case it might help diagnose the problem.

http://www.getoutwiththekids.co.uk/f...-reaction-lite -jacket/

All of that being said, the reboot problem has been going on for a week or two and I really cannot remember how if this page has been open on the browser at every reboot. I can't imagine it has been but you never know. Perhaps other webpages cause the same problem. For now, I will leave that page well alone and see if the problem reoccurs (and note what was happening at the time if it does). Hopefully a pattern will emerge....

@tweezer. Before today I too have wondered about the PSU. Not that I've ruled anything out at this stage. But with the direct link to that particular web page (and possibly others) I'm not so sure. It would seem more SW/OS related. Unless that webpage is somehow sending a piece of hardware into a frenzied meltdown???

@karbomusic. Sure, I can point you towards the Minidumps. What would be the best way of doing so? Should I upload the .dmp files to Google Drive and give you the link? Thanks for the kind offer

@JHughes. As above re the power supply. As for overheating, I've been keeping a keen eye on the temps and all seems well in that department. Plus, the reboot has happened quite quickly from a cold start.

@innuendo. I don't have a spare drive unfortunately... I'm hoping that this recent discovery might set me on the right path. But if not I'll have to see if I can get hold of one for testing purposes. As for overheating, see above. I did run a stress test on the GC as I wondered if that might be the culprit. It seemed to handle it fine,although I only ran it for 10-15 minutes. Perhaps I will try Prime95 if my other options run out. Saying that, I might run it anyway just to see how things are... For the HDs, I ran the error scan from HDTune and all seemed well. Does hddguardian do the same kind of thing or is it a different kind of test?

@Sickamorz. The fan on the PSU is fine, and no - no heat sinks etc have been removed. System temps are all looking fine. There's no floppy drive connector on my motherboard (DP35DP). I have checked my start up list of programs and it seems to be normal and as it should be.

THANKS TO EVERYONE FOR TAKING THE TIME TO RESPOND AND HELP ME WITH THIS. I REALLY APPRECIATE IT.
Bob Headroom is offline   Reply With Quote
Old 09-16-2015, 02:29 AM   #8
LightOfDay
Banned
 
Join Date: Jun 2015
Location: Lower Rhine Area, DE
Posts: 964
Default

that webpage isnt the reason for your rebooting. a webpage cant do that, not with modern browsers (= NOT Internet Explorer).

if your computer boots, you fire up Chrome, load that webpage and your computer boots instantenously and does so every time you do exactly the same its not the webpage, its your RAM.

because: every time the same adresses in RAM are used. that would explain why different other apps gives you the crashes seemingly randomly. its when they hit that spot of the RAM.

I put my money on faulty RAM. that would explain all of your experiences.
LightOfDay is offline   Reply With Quote
Old 09-16-2015, 02:51 AM   #9
innuendo
Human being with feelings
 
Join Date: Nov 2013
Location: Jerusalem, Israel
Posts: 659
Default

@Maxdread:
The "short test" for HDD is performed entirely in hardware, ie the HDD drive tests itself and reports back to the OS with the results. It doesn't matter what software you use to trigger the test. The test is only one part of checking HDD health. The other and more important part is reading SMART attributes of the drive. SMART is what HDD firmware knows about HDD health and error history. This information is provided as raw data and to make things worse, the format for this data differs across different manufacturers. So in order to make sense of it, you need a really clever software that would interpret this raw data for you. Every software that can read SMART does this, to a varying degree of correctness. I am recommending HDD guardian because its interpreting capabilities are superior to anything else I know of, including hd tune. And HDD guardian is free and open source.
innuendo is offline   Reply With Quote
Old 09-16-2015, 04:59 AM   #10
Bob Headroom
Human being with feelings
 
Join Date: Apr 2008
Location: UK
Posts: 262
Default

@Light of Day. I hear what you're saying and it does make sense, other than the fact that I have changed both sticks of RAM... So I can't see how that would be possible.

Could it be a problem with the page file or anything other than the physical RAM (virtual RAM etc)?

And just to confirm when you say: "your computer boots, you fire up Chrome, load that webpage and your computer boots instantenously and does so every time you do exactly the same"... This is indeed the case. Whether other web pages do the same is another matter and one I will now keep a keen eye on now. And whether it was that web page that caused the reboots when they first started happening (a week or two ago), I can't remember unfortunately.

@innuendo. OK, I will definitely give that a go then and will report back... As a quick aside, does it work for USB external drives as well? I've just bought a couple and would like to test them before using them...

Cheers guys
Bob Headroom is offline   Reply With Quote
Old 09-16-2015, 05:44 AM   #11
karbomusic
Human being with feelings
 
karbomusic's Avatar
 
Join Date: May 2009
Posts: 24,041
Default

Quote:
@karbomusic. Sure, I can point you towards the Minidumps. What would be the best way of doing so? Should I upload the .dmp files to Google Drive and give you the link?
That will work.

Also, it's always smart to look in the app and system event logs.
__________________
If it requires a null test to find it, it is by definition minuscule.
karbomusic is online now   Reply With Quote
Old 09-16-2015, 08:38 AM   #12
innuendo
Human being with feelings
 
Join Date: Nov 2013
Location: Jerusalem, Israel
Posts: 659
Default

Yes, it should work also with USB drives, as long as they are hard drives and not thumb-.
IMHO HDD or filesystem corruption are better explanation to your phenomenon than RAM because modern browsers use RAM randomization techniques to limit exploitation, so it's highly unlikely that the same page loads every time into same RAM addresses. What is more reasonable is that this page is cached on your HDD and this cache got corrupted. You can try cleaning your cache in chrome to figure wether this is true.
Another option is that there is something on that page that triggers a driver on your computer to misbehave. The most probable candidate would be the video driver.
But then it's all speculation, facts will follow as soon as you perform required tests. Which is RAM test, HDD test and SMART review, proper stress test, and also GPU stress test to rule out GPU hardware problems. You can get GPU testing software if you google "geeks3d gputest".
If neither gave any clue then fresh installation on a separate drive.

If you have no shortage of funds, I would recommend taking your machine to a repair shop rather than having to figure it out by yourself...
innuendo is offline   Reply With Quote
Old 09-16-2015, 09:04 AM   #13
karbomusic
Human being with feelings
 
karbomusic's Avatar
 
Join Date: May 2009
Posts: 24,041
Default

I'd still like to see the minidumps as well as looking at the app/sys logs. App/sys logs should be priority number one in every scenario.
__________________
If it requires a null test to find it, it is by definition minuscule.
karbomusic is online now   Reply With Quote
Old 09-16-2015, 09:10 AM   #14
Bob Headroom
Human being with feelings
 
Join Date: Apr 2008
Location: UK
Posts: 262
Default

Thanks for all this guys. As soon as I can get on my PC to move things forward and put some of this into place I will. And of course I will be back with updates. I really appreciate the help.

Quote:
Originally Posted by karbomusic View Post
I'd still like to see.....the app/sys logs. App/sys logs should be priority number one in every scenario.
What exactly do you mean by app/sys logs? And where will I find them etc?
Bob Headroom is offline   Reply With Quote
Old 09-16-2015, 09:28 AM   #15
karbomusic
Human being with feelings
 
karbomusic's Avatar
 
Join Date: May 2009
Posts: 24,041
Default

Quote:
Originally Posted by Max Dread View Post
Thanks for all this guys. As soon as I can get on my PC to move things forward and put some of this into place I will. And of course I will be back with updates. I really appreciate the help.



What exactly do you mean by app/sys logs? And where will I find them etc?
Start > Run then type eventvwr.msc then hit enter. An console window will appear and you can expand the folders to see the following logs:

Application
Security
System

You care about Application and System and any errors/warnings present that may be of interest to your issue. Back in the day when I owned a decent sized computer company I'd always give my techs a hard time because they would be trying a million things when the answer or clue was often right there in those logs. Not always but it's really troubleshooting 101 of checking the obvious first just to get it out of the way.
__________________
If it requires a null test to find it, it is by definition minuscule.
karbomusic is online now   Reply With Quote
Old 09-16-2015, 11:02 AM   #16
innuendo
Human being with feelings
 
Join Date: Nov 2013
Location: Jerusalem, Israel
Posts: 659
Default

Disclaimer: I'm not running a computer company, I'm just a humble tech on salary. Anyway my view is this: hardware usually is faster and easier to diagnose than software. That's why I generally start with the hardware when the source of the problem is unknown. Event viewer is not something I avoid, on the contrary. However in the case of a WinXP, possibly unmaintained, and which went through quite a few crashes, OS corruption is almost certainly the case to some degree. Fishing out of this mess potential cause of the problem in hand might require much more expertise on the part of the end user than running a few hardware diagnostics. That is why I recommended the latter to begin with. If we figure that the hardware is OK, we can look into software issues. In the matter of fact, I don't mind taking a look at it at this time, just I expect it to provide more questions than answers, while hardware tests give solid facts at their output.
innuendo is offline   Reply With Quote
Old 09-16-2015, 11:04 AM   #17
Jason Brian Merrill
Human being with feelings
 
Jason Brian Merrill's Avatar
 
Join Date: Jun 2006
Location: Northeastern PA, USA
Posts: 20,801
Default

what if the webpage is opening up with some kind of graphics-plugin thingy that requires something from a video card which either triggers a driver error or something?
__________________
Beliefs do not require respect. People do.
Jason Brian Merrill is online now   Reply With Quote
Old 09-16-2015, 11:43 AM   #18
karbomusic
Human being with feelings
 
karbomusic's Avatar
 
Join Date: May 2009
Posts: 24,041
Default

Quote:
Originally Posted by innuendo View Post
Disclaimer: I'm not running a computer company, I'm just a humble tech on salary. Anyway my view is this: hardware usually is faster and easier to diagnose than software. That's why I generally start with the hardware when the source of the problem is unknown. .
Sorry, wasn't implying anything about you just to be clear, it was just and old man flashback when troubleshooting. It the first and fastest place to look usually. It's something I even need remind myself of sometimes.
__________________
If it requires a null test to find it, it is by definition minuscule.
karbomusic is online now   Reply With Quote
Old 09-16-2015, 11:46 AM   #19
karbomusic
Human being with feelings
 
karbomusic's Avatar
 
Join Date: May 2009
Posts: 24,041
Default

Quote:
Originally Posted by Jason Brian Merrill View Post
what if the webpage is opening up with some kind of graphics-plugin thingy that requires something from a video card which either triggers a driver error or something?
Plausible.
__________________
If it requires a null test to find it, it is by definition minuscule.
karbomusic is online now   Reply With Quote
Old 09-17-2015, 05:21 AM   #20
Bob Headroom
Human being with feelings
 
Join Date: Apr 2008
Location: UK
Posts: 262
Default

Finally got some time to dedicate to this....

OK.... so first up I've zipped up some of the Minidumps and they can be found here:

https://drive.google.com/file/d/0B5T...ew?usp=sharing

I did recently change "Startup and Recovery > System Failure > Write Debugging Information" setting from "Small Memory Dump (64kb)" to "Kernel Memory Dump" and then back again. So there will be both types of dump in the zip. (Take that as you will!).


Next.... The app/sys logs. It's all Greek to me so I've taken some screen shots if anyone would mind taking a gander? I've also uploaded them to Google Drive.

Here's a link to several pages from Application:
https://drive.google.com/file/d/0B5T...ew?usp=sharing

What I would add though is that the crypt32 Error that can be seen can also be seen way back to the start of the event logs (August 2014).

And here's a link to several pages from System:
https://drive.google.com/file/d/0B5T...ew?usp=sharing

Lot's of errors dating back to when (*I THINK*) the problem began! I thought it had been going on for longer but time does play terrible tricks on my memory.


@Jason Brian Merrill. Interesting idea and one that does fit the pattern of behaviour at this stage (albeit that I cannot remember the details of the first few crashes, which is a real shame). I guess time will tell... Hopefully!

I'm now going to run some of the hardware tests as per Innuendo's posts and will report back with results in due course.

Thanks a billion

Max
Bob Headroom is offline   Reply With Quote
Old 09-17-2015, 05:34 AM   #21
ivansc
Human being with feelings
 
Join Date: Aug 2007
Location: Near Cambridge UK and Near Questembert, France
Posts: 18,654
Default

And of ourse when you swapped out the ram you need to make absolutely 100% sure the ram sticks are fully and correctly inserted.
I have been caught out like this so many times.
Run MEMTEST as well.... for the full series of tests.
Not boring if you run it whilst you are in bed asleep!
ivansc is offline   Reply With Quote
Old 09-17-2015, 05:37 AM   #22
Bob Headroom
Human being with feelings
 
Join Date: Apr 2008
Location: UK
Posts: 262
Default

OK - thanks. I'll run it while I sleep tonight. And if all is well, I might just stick the old RAM back in and run Memtest on that the next night. I might as well check everything while I'm here. And even if it does not relate to the problem it is still good "housekeeping".....

Cheers
Bob Headroom is offline   Reply With Quote
Old 09-17-2015, 07:47 AM   #23
Bob Headroom
Human being with feelings
 
Join Date: Apr 2008
Location: UK
Posts: 262
Default

A couple of updates. First, I have found another web page that reboots my PC when it opens.... This is the link (I'm certain now that there is nothing inherently wrong with the web page itself, but please treat with caution nonetheless):

http://www.techsupportalert.com/best...ng-utility.htm


Secondly, I forgot to mention this one... A few days ago I was listening to something from youtube and used the EMU Patchmix software to turn down the volume (let's say to -14). A little later the PC crashed and rebooted. When I next went to listen to music, I thought to myself how it was quieter than usual. I'd forgotten that I had turned down the fader in Patchmix. So I turned it back up to 0. The next time the PC crashed/rebooted, the volume on the fader was back to -14. This happened a few times.

Today I have had a similar experience. I re-arranged some icons on my desktop. After a crash/reboot, the icons returned to how they were before I re-arranged them. This also happened a few times.

Not sure if any of this is helpful or relevant but thought I would share in case it is.

Now going to try Prime95.

Cheers
Bob Headroom is offline   Reply With Quote
Old 09-17-2015, 07:55 AM   #24
mdmiller
Human being with feelings
 
Join Date: Mar 2013
Posts: 131
Default

Quote:
Originally Posted by Max Dread View Post

... So I turned it back up to 0. The next time the PC crashed/rebooted, the volume on the fader was back to -14. This happened a few times.

Today I have had a similar experience. I re-arranged some icons on my desktop. After a crash/reboot, the icons returned to how they were before I re-arranged them. This also happened a few times.

Cheers
This is to be expected after a crash since Windows saves settings when shutting down.

There are so many things that could be causing your problems it can be very difficult to troubleshoot. When I had similar problems in the past, I discovered that the caps on the MB were going bad. If they are bulging, it's time to replace them or get a new MB. I also experienced the same sort of issues when the power supply was on its last legs.
mdmiller is offline   Reply With Quote
Old 09-17-2015, 08:07 AM   #25
karbomusic
Human being with feelings
 
karbomusic's Avatar
 
Join Date: May 2009
Posts: 24,041
Default

Hmm... Was a pain, I had to go find XPSymbols which appears to now be a download....

Looks like it is drawing a font:

STACK_TEXT:
a4674a44 bf8f7cff e258f840 000002e5 e55c1000 win32k!cjFillPolygon+0x268
a4674b34 bf97c067 00000015 00000001 00000015 win32k!lQueryTTOutline+0xe4
a4674b68 bf97c203 e531c7c0 00000015 00000000 win32k!bValidRangeMORT+0x14
a4674b94 bf97a9bb e531c7c0 00000015 00000004 win32k!vSetClearTypeState__FONTCONTEXT+0x7f
a4674be0 bf9660ed 00000000 e531c7c0 00000015 win32k!vFillGLYPHDATA_ErrRecover+0x6e
a4674c10 bf951ad1 00000000 e531c7c0 00000015 win32k!RFONTOBJ::vInitEUDC+0x66
a4674cc4 bf948569 e531c7c0 00000000 00000002 win32k!GreMakeFontDir+0x109
a4674d3c 8054161c f901222c 00000015 00000182 win32k!pbmiConvertInfo+0x1e
a4674d64 7c90e4f4 badb0d00 00126fbc a43fcd98 nt!RtlIpv4StringToAddressExW+0x9d
WARNING: Frame IP not in any known module. Following frames may be wrong.
a4674d78 00000000 00000000 00000000 00000000 0x7c90e4f4


It's always the same stack/operation in the dumps. Couple things.... Which AV are you using and is it up to date. Could be related to video but I'm suspicious.

All of this could still be relating back to a PSU or something but I do find the above interesting if it crashes when Chrome is loading a web page so if it has any extensions/plugins, disable them to see what happens.
__________________
If it requires a null test to find it, it is by definition minuscule.

Last edited by karbomusic; 09-17-2015 at 09:26 AM.
karbomusic is online now   Reply With Quote
Old 09-17-2015, 03:44 PM   #26
innuendo
Human being with feelings
 
Join Date: Nov 2013
Location: Jerusalem, Israel
Posts: 659
Default

The event viewer logs confirm that you have some kind of OS corruption. Can not say to what extent because details are missing from these screenshots. Particularly interesting are DCOM errors and Service Control Manager errors. You can view the details when you double-click on a particular event. Details of all errors from 17/09/2015, between 12:13 and 12:17 should suffice. Plus maybe the DCOM errors from 12:42-13:00.

While this will provide a more precise characteristic of the OS corruption, chances are that we still won't know whether OS corruption is the source or the outcome of the crashes.

It would be reasonable to attempt to fix corruption, however I would not attempt this until review of HDD SMART. Because if your drive is collapsing then instead of attempting to fix OS, you should begin to backup data. BTW do you have free space on the OS partition?

Last edited by innuendo; 09-17-2015 at 04:03 PM.
innuendo is offline   Reply With Quote
Old 09-17-2015, 04:00 PM   #27
karbomusic
Human being with feelings
 
karbomusic's Avatar
 
Join Date: May 2009
Posts: 24,041
Default

Can you right-click and export the App and Sys logs as CSV? I can't tell anything from just a list other than the fact that I don't see any entries from FtDisk which is good but I need to see the details of the others. As per my previous post, the fact that every single crash is in the exact same spot is suspicious to me; mostly because the module that called into win32k.sys is unknown by the debugger, that means it is likely a third party hardware driver calling into win32k.sys. I'll see if I can figure out what it is but these minidumps don't appear carry that info.

Also, since you are providing so much info, feel free to PM the link to us.
__________________
If it requires a null test to find it, it is by definition minuscule.

Last edited by karbomusic; 09-17-2015 at 04:07 PM.
karbomusic is online now   Reply With Quote
Old 09-17-2015, 04:19 PM   #28
Bob Headroom
Human being with feelings
 
Join Date: Apr 2008
Location: UK
Posts: 262
Default

I didn't end up installing HDD guardian because it needed .NET Framework 4. I can't for the life of me remember why exactly, but I remember quite some time ago deciding not to install that when some other program needed it. HD Tune and WD Data LifeGuard Diagnostics both have SMART reports, and the latter also does an Extended test which I am now running. Will that suffice or is there still something HDD Guardian does which these two don't?

@karbomusic - your previous questions. I am running AVG Free for my AV. It is up to date. I also tried disabling all of the Chrome extensions I use but the crash still happened the same. I wonder if I should try installing Chrome completely and then re-installing it from scratch? One thing I did see is that Chrome has updated a couple of times this month and that has got me to wondering whether one of the recent updates is causing the problem. I've not been able to find anywhere yet though from where I can download previous versions of Chrome. I'll take another look tomorrow. That being said, I'm no expert (as I'm sure you've gathered) and me wondering whether a Chrome update could be the cause of all this could well be short sighted and/or naive. It's just that the crash is so tied in with Chrome, and the PC does not crash when opening those same pages in my old version of Opera. If I can't find an older version of Chrome to download, and if you guys think it is remotely possible that the Chrome update could be the problem, perhaps I should install another browser and use that for now and see if I get any more crashes...?

All that being said, I have saved the App and Sys logs as CSV as requested and perhaps that will shed some new light onto this:

https://drive.google.com/file/d/0B5T...ew?usp=sharing

https://drive.google.com/file/d/0B5T...ew?usp=sharing

I'm going to have to wimp out and get some sleep now but will be back in the morrow.

Thanks a trillion for all this. It really is ultra kind...
Bob Headroom is offline   Reply With Quote
Old 09-17-2015, 04:22 PM   #29
Bob Headroom
Human being with feelings
 
Join Date: Apr 2008
Location: UK
Posts: 262
Default

Quote:
Originally Posted by innuendo View Post
you should begin to backup data.
Yep, I'm on the case with that and have been for much of today. Good thinking!

Quote:
Originally Posted by innuendo View Post
BTW do you have free space on the OS partition?
It's a 30GB partition with 19.5GB free.
Bob Headroom is offline   Reply With Quote
Old 09-17-2015, 04:26 PM   #30
Bob Headroom
Human being with feelings
 
Join Date: Apr 2008
Location: UK
Posts: 262
Default

Another potentially stupid/naive/2,000 miles of the mark idea.... But is there any chance there could be something wrong with the page file? Should I try defragmenting it? I've never done that before...? Or should I go and get that sleep I just mentioned (and my coat while I'm at it)?
Bob Headroom is offline   Reply With Quote
Old 09-17-2015, 04:53 PM   #31
karbomusic
Human being with feelings
 
karbomusic's Avatar
 
Join Date: May 2009
Posts: 24,041
Default

Quote:
Originally Posted by Max Dread View Post
Another potentially stupid/naive/2,000 miles of the mark idea.... But is there any chance there could be something wrong with the page file? Should I try defragmenting it? I've never done that before...? Or should I go and get that sleep I just mentioned (and my coat while I'm at it)?
Scratch my comment about it being the same every time. I dug deeper. It is jumping around but RtlIpv4StringToAddressEx() does seem to be a common function that is being called much of the time but not the font thing....

https://msdn.microsoft.com/en-us/lib...(v=vs.85).aspx

All that ^function does is turn a string representation of an IP address and port number into a binary version. A few observations and guesses:

1. That function would never likely be called unless something needed access to the network. This implies we'd be needing the NIC because we wouldn't be translating an IP unless it was to access the network. Based on #2 could the NIC be tugging on current from the PSU who can't supply steady current?

2. As I look through the dumps it surely could be the PSU. Once a PSU becomes unstable enough due to age, it can't refresh quickly enough, this causes bits to flip which results in extremely strange memory corruption which you certainly have going on. Remember physical memory is really nothing more than incredibly large arrays of capacitors (in layman's terms) which constantly need to be refreshed to hold their values. Any dip in steady power and we have issues.

3. I found at least two references on the net of that function being called followed by a BSOD similar to yours and the issue being PSU. Anecdotal but interesting.

4. As innuendo stated, there could be some type of file/disk corruption. However, I think that should be provable via FtDisk entries in the event log, chkdisk throwing errors etc and running system file checker. So be sure to check that. There could potentially be a disk issue that gets by that but this will cover 98% of possibilities. You could also defragment drive C: because this will force sectors being moved around, if disk is corrupt either defrag will have issues or yield clues because there is chance the files I see involved get moved to other sectors in the process of the defrag.

5. You have drivers going back to 2007 such as Creative Labs. No biggie but 8 year old drivers have no idea about any newer OS code = potential bugginess. Albeit this is much lower on my list of suspicions.

You could potentially try this... Remove some peripherals that draw power and see if it continues to crash. Remove all USB devices or anything you think might help accomplish this task. If the PSU is borderline there is a small chance removing some of the load lightens the current load enough to return the machine to being stable. If so, replace the PSU but if it doesn't help at all this doesn't get the PSU off the hook entirely.

How old is this machine, my gut feeling is that it is very "tired" and if the PSU is 8 years old, dusty and been running that long, I'd surely suspect it though I'd check the above and couple things innuendo mentioned to.

Hope this helps.
__________________
If it requires a null test to find it, it is by definition minuscule.

Last edited by karbomusic; 09-17-2015 at 05:09 PM.
karbomusic is online now   Reply With Quote
Old 09-17-2015, 05:00 PM   #32
innuendo
Human being with feelings
 
Join Date: Nov 2013
Location: Jerusalem, Israel
Posts: 659
Default

You can use gsmartcontrol as an alternative to hddguardian. It is based on the same code but the GUI does not require .net. But since you are already running extended self test (which is essentially complete disk surface scan performed in hardware), wait until that completes, because it might present different data in SMART. That will take a few hours or days, depending on your drive's health and size. I would suggest to leave the HDD alone until that test completes. The software which triggered the test won't necessarily estimate correctly when it should end and won't necessarily be able to update test status in real time. Just quit and the re-open that software from time to time (or find a "refresh" control in it) and take a look at self-tests results. When you see the extended self-test result there, it's finished.

The WD software was extremely buggy last time I tried it, so I doubt it's usefulness. HD Tune was kinda OK I think but I would not rely on it. Just use gsmartcontrol.

AV software is often the cause of crashes as karbo mentioned. Chrome updates afaik are mandatory, so as soon as you download an older version, it will quietly update itself. Anyway it is better to complete the diagnostics prior to attempting repairs. And when you start with repairs, it should be done in a particular order. You should start with checkdisk, then depending on the diagnosis and post-checkdisk situation run OS repairs. Uninstalling AVG and cleaning up after it might be all you need to do. Userspace applications repairs are best to do in the end if required.
innuendo is offline   Reply With Quote
Old 09-17-2015, 05:33 PM   #33
innuendo
Human being with feelings
 
Join Date: Nov 2013
Location: Jerusalem, Israel
Posts: 659
Default

I'm not familiar with ftdisk events because when I started to work professionally in this field, XP had already become an antique, but if it's anything like the "disk" event in Win7 and later, then the lack of it does not prove that the disk is OK. Have seen so many machines without disk event but with collapsing disks.

The PSU hypothesis sounds plausible, however it should manifest itself in stress tests as well, particularly in CPU and GPU tests. That said, I mostly happen to service laptops which is like 95% of computers coming into the lab, so my experience with stationary machines is limited.
innuendo is offline   Reply With Quote
Old 09-17-2015, 06:13 PM   #34
innuendo
Human being with feelings
 
Join Date: Nov 2013
Location: Jerusalem, Israel
Posts: 659
Default

Took a look at your csv. Looks like some third-party services/drivers are corrupted, but no corruption affected core OS services.

There are 2 interesting things there:
1) nsynas32 error - which is a driver by synchrosoft
2) apple mobile device error

Both are probably not related to the crashes, but nevertheless worth to kick out of the equation. We will get to that after diagnostics if this will still be relevant.
innuendo is offline   Reply With Quote
Old 09-17-2015, 06:45 PM   #35
karbomusic
Human being with feelings
 
karbomusic's Avatar
 
Join Date: May 2009
Posts: 24,041
Default

Quote:
Originally Posted by innuendo View Post
I'm not familiar with ftdisk events
It's certainly good not to see them because I've rarely seen them without a disk completely failing a short time later. With or without those, if disk, it should be provable due to the frequency of the issue.


Quote:
The PSU hypothesis sounds plausible, however it should manifest itself in stress tests as well, particularly in CPU and GPU tests.
I just feel better (for whatever reason) if there were less components drawing current. That and there is the old school side-advantage of removing hardware and adding it back one by one. Academically and a short diversion....

Electrolytic caps are needed in the PSU to smooth out the DC and allow for spikes in demand. IOW, if the AC dipped for a short moment, those caps take over and drain their charge so that the computer doesn't actually see the dip provided the AC recovers before the cap drains. This is the same thing smoothing caps do in amplifiers so that large current demands due to the dynamics of the music don't cause the amp to sag too much due to power starvation. Think kick drum.

The thing about electrolytic capacitors is it is never if, but when because they always degrade over time by simply existing. That is so much the case that I would gander that many switches, routers etc. that mysteriously go bad, likely do because their caps die. It's that type of cap's Achilles' heel; they still get used because its the only size and cost effective way of achieving UF values high enough for this type of use. For example a 1000Uf electro cap might be half the size of an AA battery but a different type cap which would last much, much, much longer, of the exact same value would literally be the size of a can of beans!
__________________
If it requires a null test to find it, it is by definition minuscule.

Last edited by karbomusic; 09-17-2015 at 06:50 PM.
karbomusic is online now   Reply With Quote
Old 09-18-2015, 04:54 AM   #36
Softsynth
Human being with feelings
 
Softsynth's Avatar
 
Join Date: Jun 2015
Posts: 5,379
Default

Is the machine clean or clogged with dust?

Motherboards, PSUs and display cards made in the 2000s could have been made with chemically unstable capacitors, they have given electrolytic cap's a bad reputation.
These days motherboards and graphics cards are being made with more reliable "solid capacitors" and quality PSUs are being made with superior capacitors.

Electrolytic cap's in the PSU or motherboard could be the cause. Don't be tempted to replace capacitors, just get a replacement component. Tell tale signs are bulging/venting can tops or even leakage. but failing cap's do not have to be visibly stressed to be on the out.

Electrolytic Cap's can last for several decades with no discernible change, others can fail much sooner, depending on quality and heat. There was an issue with a batch of poor capacitors around the 2000s:
https://en.wikipedia.org/wiki/Capacitor_plague

Trouble shooting the PC I would switch it off, systematically disconnect components, inform the set up in BIOS if necessary. So extra hard drives if any in the machine. Additional memory modules, add in cards of any form. additional USB devices. Basically you want to strip back to bare bones and add components back in until stability issues rear their ugly head.
Softsynth is offline   Reply With Quote
Old 09-18-2015, 05:00 AM   #37
Bob Headroom
Human being with feelings
 
Join Date: Apr 2008
Location: UK
Posts: 262
Default

Rather annoyingly, I'm only getting snippets of time to dedicate to this. Hopefully more will come at the weekend....

With just a short window of time, I unplugged what I could from the PC. I did:

- the soundcard (I just disconnected the mini-molex thing that supplies it with power)
- the three CD/DVD writers I have (again I just unplugged the power connectors from the drives)
- one of my three hard drives (disconnected).

That's all I could think of disconnecting.... Aside from those, USB-wise I just have keyboard, mouse and wireless adapter. Do you think that was sufficient (let me know if you can think of other ways of reducing load)? Are there any other ways of testing the PSU? If not, I'll have to see if I can get hold of one - even if just to borrow - and trying plumbing it in to see what happens.


Now let me answer a few more questions from last nights posts, and give a few more details, before getting on to the testing side of things....

The age of the PC is around 7 years. I built it myself. The PSU is the original and so is also around 7 years old. It is a Seasonic SS-400ET (OEM version of the S-12 430W).

In terms of testing the HDs... I have three installed altogether. So far I have run the WD Data LifeGuard Diagnostics Extended Test on two of them (one being the OS drive). Both tests came back fine. What should I prioritise next? chkdisk? System File Checker? Or gsmartcontrol? When using gsmartcontrol should I run the Extended self-Test as well as the other two?

I have defrag'd the OS drive at least once since all these shenanigans kicked off.... I used JKdefrag/MyDefrag.

The Creative Labs driver is for my soundcard which EMU stopped supporting a long time ago. Hence that's the most up to date driver for it. As a quick aside, it is one of the reasons why I am sticking on to XP.

I will get on to testing with Prime95 as soon as I can. Which test would you recommend and for how long for the purposes of what we are doing?

I will run some GPU tests after that, as recommended previously above.

I have to dash now as my dogs need taking out! One last thing though - I really cannot afford much because I am almost out of work at the moment, but I would like to send you both a little thank you for the time and kindness you have shown me in all of this. If you can PM paypal details I can send you both enough for a pint over the weekend. It's not much, but at least extends a little way toward showing you haw grateful I am....

Cheers
Bob Headroom is offline   Reply With Quote
Old 09-18-2015, 06:12 AM   #38
ivansc
Human being with feelings
 
Join Date: Aug 2007
Location: Near Cambridge UK and Near Questembert, France
Posts: 18,654
Default

Hm. That does seem like a lot of devices hanging off our 430watt psu, even if it IS a Seasonic.

You could indeed be running the psu up to its limit, especially if it ha worked hard for 7 years.

Bearing in mind what you an buy a reasonable 500 watt supply for new these days, that might well be job one if the various tests you have already tried dont show anything else up.

And fwiw I used my Emu 1212m pci card up to and including win7 64bit using the beta drivers that are still out there -no problems at all.

I assume YOUR Emu card is that or an 0404 pci?
ivansc is offline   Reply With Quote
Old 09-18-2015, 07:20 AM   #39
karbomusic
Human being with feelings
 
karbomusic's Avatar
 
Join Date: May 2009
Posts: 24,041
Default

Just be aware I can't be 100% about the above, it's estimated guessing based on what info we have; meaning don't throw money at it unless you have to. That machine and OS have enough age on them that it will likely "nickel and dime" you as time passes.

Quote:
Originally Posted by Max Dread View Post
One last thing though - I really cannot afford much because I am almost out of work at the moment, but I would like to send you both a little thank you for the time and kindness you have shown me in all of this. If you can PM paypal details I can send you both enough for a pint over the weekend. It's not much, but at least extends a little way toward showing you haw grateful I am....

Cheers
Um.... no! I would not have answered for any other reason than helping and participating with other Reaper users. That's what we are here for; so just 'pay it forward' sometime when you have the chance.
__________________
If it requires a null test to find it, it is by definition minuscule.

Last edited by karbomusic; 09-18-2015 at 07:30 AM.
karbomusic is online now   Reply With Quote
Old 09-18-2015, 08:44 AM   #40
Bob Headroom
Human being with feelings
 
Join Date: Apr 2008
Location: UK
Posts: 262
Default

Quote:
Originally Posted by ivansc View Post
Hm. That does seem like a lot of devices hanging off our 430watt psu, even if it IS a Seasonic.
I spec'd it when I bought it and I'm pretty sure I'm ok with what I've got. I will double check though when I'm at my PC for longer and report back.

Quote:
Originally Posted by ivansc View Post
And fwiw I used my Emu 1212m pci card up to and including win7 64bit using the beta drivers that are still out there -no problems at all.

I assume YOUR Emu card is that or an 0404 pci?
I've got the 1820M. I think I might have read one or two forum posts about people who had success with Win7 64-bit. But I read other where people didn't. I will try it one of these days... But not at the moment as I just have too much on.

Quote:
Originally Posted by karbomusic View Post
That machine and OS have enough age on them that it will likely "nickel and dime" you as time passes.
I hear what you're saying, but up until this recent problem everything has been pretty much tip top with the system since I built it. And I really want to put off upgrading anything until I actual have a practical need to. If I'm totally honest, part of the reason for this is that I struggle with change and it really can quite stress me out. I go into a bit of a stress melt down. When the time comes I'll do it. And I'll be very happy to be able to have more than 4GB RAM; not that that's held me back so far...

Quote:
Originally Posted by karbomusic View Post
Um.... no! I would not have answered for any other reason than helping and participating with other Reaper users. That's what we are here for; so just 'pay it forward' sometime when you have the chance.
Thank you! I'm just not sure I have the knowledge in any area to help out much like you've helped me. Perhaps I could put the money toward a charity then, which you are more the welcome to choose if you have any causes that are particularly significant to you...


I'm home for a couple more hours then I'll be away for around 15-20 hours. Is there any test that requires that long and that I should prioritise and keep running while I'm away?

Prime95? (If so, which test)? And is it potentially dodgy or dangerous to run it whilst not at my PC (overheating, etc)?

Memtest? (If so, will it test BOTH sticks of RAM at once? Can it test FOUR sticks at once? I'm presuming it runs pre Windows load and therefore might see all four sticks if I put them in?

Any other test instead?

If I don't hear back in time, maybe I'll just switch off for now and run tests overnight once I'm back.

Thanks everyone.
Bob Headroom is offline   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -7. The time now is 07:57 PM.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2019, vBulletin Solutions Inc.