(0xC0000005) STATUS_ACCESS_VIOLATION any idea what caused it?? Or what it exactly is describing?? Internet searches find nothing.

Message boards : Number crunching : (0xC0000005) STATUS_ACCESS_VIOLATION any idea what caused it?? Or what it exactly is describing?? Internet searches find nothing.
Message board moderation

To post messages, you must log in.

1 · 2 · 3 · 4 . . . 5 · Next

AuthorMessage
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1923912 - Posted: 11 Mar 2018, 6:10:21 UTC

After I brought Numbskull back up this afternoon after being turned off for some room reconfiguration, I discovered that I started to trash about 47 tasks on all the gpus for no apparent reason. It was running fine before I turned the system off.

Does anyone know what causes the issue and have a solution for it?

I stopped BOINC after I saw the compute errors. I did a sfc /scannow and didn't find any issues as expected. I then decided to do a BOINC repair. That didn't accomplish anything since I was still making errors after a brief BOINC restart. Decided to reinstall the Nvidia drivers.

That might have been the trick to getting the errors to stop. Haven't seen any errors since I brought the system and BOINC back online. The system was running fine with no apparent failures. BOINC ran each gpu task to completion with a good exit signal. Just that each task is labelled as a compute error with a system dump appended.

Can't find any reference to the error online except to a game called Arma 3. Anyone have any comment or knowledge on the subject? TIA.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1923912 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14649
Credit: 200,643,578
RAC: 874
United Kingdom
Message 1923934 - Posted: 11 Mar 2018, 10:27:19 UTC - in response to Message 1923912.  

Anything with the 'shape' like 0xC0000005 is a low-level Microsoft operating system error. Googling it invariably leads to thousands and thousands of dumb questions where the OS threw a whoopsie during the running of this application, that application, or the other application.

The motherlode for starting your diagnostics is https://msdn.microsoft.com/en-us/library/cc704588.aspx. In this case, STATUS_ACCESS_VIOLATION is described as

The instruction at 0x%08lx referenced memory at 0x%08lx. The memory could not be %s.
You'll find that in full, with the blanks filled in, in the stderr for the tasks.

So, the application that's running is having problems with RAM. If it's an application which has run before, or is running on another machine, 47 consecutive failures points at RAM failure. Test it.
ID: 1923934 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1923966 - Posted: 11 Mar 2018, 15:49:03 UTC

Thanks Richard, I looked in Event Viewer for the error and clicked on the 'Help' function which then took me to a invalid Microsoft URL. Thanks for nothing Microsoft.

When you say the memory was an issue, are you saying the system memory or the gpu memory? The tasks that were running on the three gpus all errored. None of the cpu tasks errored. I find it hard to believe all three gpus simultaneously had memory issues. There had to be a common connection between all three.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1923966 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14649
Credit: 200,643,578
RAC: 874
United Kingdom
Message 1923977 - Posted: 11 Mar 2018, 16:59:41 UTC - in response to Message 1923966.  

I don't think the Microsoft OS has any knowledge of GPU memory: it will be referring to system memory. That page describes a coding file called 'ntstatus.h', which has existed since, well, the beginning of Windows NT.

Go up a level, and look at https://msdn.microsoft.com/en-us/library/cc231200.aspx. Your error has severity 'C' (for error), and code 5 (a very, very, early number in the grand scheme of things). It was probably defined in the very first draft of Windows NT.

Note the final part of the explanation: "The memory could not be %s." That %s could be replaced by 'written' - it could be a security protection error. I suggested testing RAM because it's relatively easy for the end-user to do, but from what you've said now, I agree it's less likely to be the problem. Is that machine up-to-date with security patches?
ID: 1923977 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1923978 - Posted: 11 Mar 2018, 17:05:06 UTC - in response to Message 1923977.  

Well I just made two more errors. This time on the cpu. So I am running a stress test on the memory now. It is up to date on security patches as far as I can tell. Only get the monthly rollups automatically.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1923978 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1923979 - Posted: 11 Mar 2018, 17:29:31 UTC

I just made a 300% coverage memory stress test on all 16GB on all cores and didn't have any errors. I have not blue-screened or system errored yet. I don't see any WHEA errors either.

I reseated the RAM in case picking the case up and moving it yesterday while building the new desk jostled anything. I have no idea what else to try other than going into the BIOS and detuning every memory parameter back to JEDEC 2133.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1923979 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1923983 - Posted: 11 Mar 2018, 18:03:13 UTC

Still getting errors and I finally did just get a memory blue screen. So the system is not happy with the memory now for some reason. Failing memory or the new cpu? I just threw the memory settings back to stock DDR4-2400 for Ryzen and see if that makes any difference.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1923983 · Report as offensive
Profile Gary Charpentier Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 25 Dec 00
Posts: 30608
Credit: 53,134,872
RAC: 32
United States
Message 1924013 - Posted: 11 Mar 2018, 19:34:48 UTC - in response to Message 1923983.  

Still getting errors and I finally did just get a memory blue screen. So the system is not happy with the memory now for some reason. Failing memory or the new cpu? I just threw the memory settings back to stock DDR4-2400 for Ryzen and see if that makes any difference.

Yes. Possible it didn't like the hot-cold-hot cycle and internally a chip has an intermittent fault from thermal expansion/contraction. If you have an extra known good stick of ram you could try swapping it and seeing if you can find the bad one.
ID: 1924013 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1924027 - Posted: 11 Mar 2018, 19:53:12 UTC - in response to Message 1924013.  

Been trying to find the post by AMD saying the IMC doesn't like temps lower than 20° C. I had the window open in the room as usual for cooling and the cold north breeze dropped the water temp down to 17° C. I am wondering if that could possibly be the reason for the errors. Never usually worry about too low temps, certainly the opposite with all computers running and daytime outside temps greater than 75° F. and the house and room temps start going past 82° F.

I've run for about an hour now with no further errors at 2400. Will bump it up a notch and see if it still stays stable and no errors.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1924027 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1924092 - Posted: 12 Mar 2018, 1:09:00 UTC

I think something in Windows got corrupted even though the sfc /scannow showed no issues. I rebooted the system after changing some BIOS values and Windows went to Automated Repair which it failed at repeatedly for several hours of me trying everything to get the system back. Finally decided I would just reinstall Windows and at least get a running OS again. Somehow that got botched too but it did present my original Windows installation to me again with no loss of data. Never had to deal with a unbootable Win 10 system before and never had to use any of the recovery options. They don't work all that well it seems. Wouldn't even use my System Image and Restore point on another separate physical drive. I actually had started installation of Linux and was going to pull the plug on Windows10, but changed my mind and decided to try the complete reinstall of Windows 10.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1924092 · Report as offensive
Profile Gary Charpentier Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 25 Dec 00
Posts: 30608
Credit: 53,134,872
RAC: 32
United States
Message 1924114 - Posted: 12 Mar 2018, 3:47:26 UTC - in response to Message 1924027.  

Been trying to find the post by AMD saying the IMC doesn't like temps lower than 20° C. I had the window open in the room as usual for cooling and the cold north breeze dropped the water temp down to 17° C. I am wondering if that could possibly be the reason for the errors. Never usually worry about too low temps, certainly the opposite with all computers running and daytime outside temps greater than 75° F. and the house and room temps start going past 82° F.

I've run for about an hour now with no further errors at 2400. Will bump it up a notch and see if it still stays stable and no errors.

Wasn't talking about running temps, but while off, the entire rig goes to room and then you turn it on and it goes to hot. Big change. Can cause the less than hair size wires inside the chip case to go sprong.
But I see you did a reinstall and things seem happy. Hope they stay happy.
ID: 1924114 · Report as offensive
Profile Jord
Volunteer tester
Avatar

Send message
Joined: 9 Jun 99
Posts: 15184
Credit: 4,362,181
RAC: 3
Netherlands
Message 1924152 - Posted: 12 Mar 2018, 11:22:32 UTC

I always found 0xC0000005 very difficult to trace. Have even got a BOINC FAQ for it, which may or may not give pointers, because this is such a difficult error to trace.
ID: 1924152 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1924198 - Posted: 12 Mar 2018, 15:42:27 UTC

I think that Gary is on to something. The system had been powered on ever since I replaced the cpu and hadn't been turned off until Saturday when I shut down both computers in the room for rebuilding a desk surface sufficient to handle another system. It never came back the same. Been unstable ever since. The other identical system (more or less) just powered up like nothing happened and is running the same as before. The system with the new cpu won't run at anything other than box stock. It corrupted the operating system. I checked the RAM and verified the CRC's and SPD settings to be sure the memory hadn't been corrupted. Everything is pointing to the IMC in the CPU going bad. The only way to know for sure is to replace the cpu again. I will have to purchase another chip as I don't want to go through the AMD RMA process again without advance shipment. If I can just ride this one out for another month and a half I would just buy the new 2700X cpu and replace the bad 1700 cpu through the usual AMD RMA process since I won't be time constrained.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1924198 · Report as offensive
Profile Karsten Vinding
Volunteer tester

Send message
Joined: 18 May 99
Posts: 239
Credit: 25,201,931
RAC: 11
Denmark
Message 1924214 - Posted: 12 Mar 2018, 16:32:19 UTC - in response to Message 1924198.  
Last modified: 12 Mar 2018, 16:33:25 UTC

I have to ask, mostly because I have had similar problems recently myself...:

Did you upgrade your BIOS about the time where you moved the PC?

I have had problems since I upgraded my BIOS from ver 3401 (AGESA 1071) to 3803 (AGESA 1000a). My board is the ASUS TUF B350M-PLUS GAMING.

The system became unstable and would crash. There was no regularity to the crashes. They happened at stock, overclocked, sometimes after few minutes sometimes the system would run for days.

But it allways crashed with a blue screen, never just froze.

As a last resort i set my ram to 2400MHz. My ram is Ryzen certified DDR 3200 (G.Skill FlareX), which I have been running at 3200 since day one without problems. But setting it to 2400 MHz removed the crashes.

I have been reading up on the problem, and its a known problem with the latest AGESA that it breaks some RAM (others get better). And if the ram is unstable peoble have seen corruption of their OS'es.

I am a little irritated that my, rather expensive, ram is now running below spech, but at least my system is now stable.

I'll wait for a fix, mostly because a downgrade is not really very straight forward on my board.
ID: 1924214 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1924221 - Posted: 12 Mar 2018, 16:54:44 UTC - in response to Message 1924214.  

I had been running the 3803 BIOS since it came out, so about 1 1/2 months. I only lost my ability to use 3333Mhz and had to fall back to 3200Mhz. Same memory timings mostly and still at CL14.

The memory was happier at much less voltage as well as with the SoC voltage with the 3803 BIOS.

No, I did not upgrade my BIOS at the time I shut off the computer for 4 hours.

I think I have been reading the same in the forums, Hynix memory got better at the detriment to Samsung memory.

I was up every couple of hours during the night checking on my latest adjustments to get the system stable again. That along with constantly adjusting all evening beforehand trying to find stability after I brought the system back up. The system is blue-screening in the ntoskernel.

I have had to drop back to stock latency timings at only 3000Mhz for the memory and also drop the cpu core clock down to 3.8Ghz from the normal 3.95Ghz.

I sure hope they release a better AGESA BIOS soon.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1924221 · Report as offensive
Profile Gary Charpentier Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 25 Dec 00
Posts: 30608
Credit: 53,134,872
RAC: 32
United States
Message 1924222 - Posted: 12 Mar 2018, 16:58:38 UTC - in response to Message 1924198.  

I think that Gary is on to something. The system had been powered on ever since I replaced the cpu and hadn't been turned off until Saturday when I shut down both computers in the room for rebuilding a desk surface sufficient to handle another system. It never came back the same. Been unstable ever since. The other identical system (more or less) just powered up like nothing happened and is running the same as before. The system with the new cpu won't run at anything other than box stock. It corrupted the operating system. I checked the RAM and verified the CRC's and SPD settings to be sure the memory hadn't been corrupted. Everything is pointing to the IMC in the CPU going bad. The only way to know for sure is to replace the cpu again. I will have to purchase another chip as I don't want to go through the AMD RMA process again without advance shipment. If I can just ride this one out for another month and a half I would just buy the new 2700X cpu and replace the bad 1700 cpu through the usual AMD RMA process since I won't be time constrained.

They have cans of freeze spray for testing just such things. https://www.techspray.com/electronic-testing

Been through that once or twice. Just guessing in your case, but easy to test. Also possible to have fractured a trace on the MB in replacing the CPU. Been through that also. A memtest came back saying it all had failed. Knew that the problem wasn't the RAM.
ID: 1924222 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1926790 - Posted: 26 Mar 2018, 23:58:16 UTC

Well I am still getting occasional (0xC0000005) STATUS_ACCESS_VIOLATION errors just about every day.

I have downclocked both the cpu and memory and still getting errors. I think the new cpu is the culprit. I think is a duff cpu. Considering it was a RMA replacement for a perfectly working chip with only the theoretical seg fault issue, I think it was a mistake to RMA my first cpu. I wonder if I just got someone else's RMA of a duff chip in return.

The only way to prove anything is to replace the chip with another and see if the problem persists. I would then have to look at motherboard or memory degradation as possibilities. So unless I want to take the system down and wait for another RMA'd cpu or I wait out a couple more months with the daily errors while waiting for the Ryzen+ cpus to arrive and spend some out of pocket money for a new retail cpu.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1926790 · Report as offensive
Profile Siran d'Vel'nahr
Volunteer tester
Avatar

Send message
Joined: 23 May 99
Posts: 7379
Credit: 44,181,323
RAC: 238
United States
Message 1926851 - Posted: 27 Mar 2018, 10:36:35 UTC - in response to Message 1926790.  

I wonder if I just got someone else's RMA of a duff chip in return.


Hi Keith,

I've been watching a lot of PC build YouTube videos from a guy that has a business, for over 25 years, building and maintaining PCs for companies. He has mentioned several times that getting a bad CPU is very, very rare. He says it is possible, but then anything is possible. I don't remember him ever saying that he got a bad CPU for any of his builds. He says that if everything done diagnostically falls down to just the motherboard, CPU and PSU that more than likely it's either mobo or PSU. If the PSU is eliminated, more than like it is the mobo.

If you are one of the unlucky ones to get a bad CPU and the RMA replacement turns out to be an RMA from someone else, I would change vendors, after settling up with the current vendor that is. :) Did the RMA CPU you got come in a factory sealed package? If it did, then I would suspect the mobo if all else has been diagnostically eliminated. If it didn't, I would get my money back from the vendor and get the CPU from somewhere else. CPUs cannot be fixed! I highly doubt that the vendor can factory seal a package. To me, a vendor sending out an RMA item from another of their customers is unscrupulous, especially something like a CPU. They would lose my business in a heart beat. :)

Just some food for thought. :)

Siran
CAPT Siran d'Vel'nahr - L L & P _\\//
Winders 11 OS? "What a piece of junk!" - L. Skywalker
"Logic is the cement of our civilization with which we ascend from chaos using reason as our guide." - T'Plana-hath
ID: 1926851 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1926886 - Posted: 28 Mar 2018, 14:26:59 UTC - in response to Message 1926851.  

Well I would argue that cpus don't degrade or die. Case in point. My brand new 6850K lasted all of 5 days. It died yesterday with no warning. Machine was running just fine crunching all cpu tasks because it had run through all the gpu tasks, went to take the trash cans to the curb and came back to a powered off computer. Q-code 00 on the display and cpu diagnostic led lit on the mobo. Took the AIO cold plate off the cpu, pulled the chip and put back in the Xeon E5-2620V4 and the system powered right back up from where it left of. Dead cpu - no question.

The RMA was direct from AMD through their Support Team. I changed the chip because pre-week 25 chips ( mine was week 7) had a manufacturing flaw that caused segmentation violation faults when heavily loaded like when crunching. The replacement chip was factory sealed and week 38.

But the system never behaved the same right from the start with the new cpu. Wouldn't run the clock and memory speeds that the first chip ran without issues and the newer chips were supposed to be able to reach higher speeds at less voltage than the early chips because of process improvements since the early production runs. That is what others that replaced their defective chips had reported.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1926886 · Report as offensive
Profile Karsten Vinding
Volunteer tester

Send message
Joined: 18 May 99
Posts: 239
Credit: 25,201,931
RAC: 11
Denmark
Message 1926940 - Posted: 28 Mar 2018, 20:10:25 UTC - in response to Message 1926886.  

As a followup, yesterday I installed the just released 3805 BIOS for my board.

I have set mem back to DOCP3200, and have been running the sytem since then without crashes.

Its still early days, as I did have uptimes at above 1 day on 3803 and DOCP3200, but most times it crashed within minutes.

So it _seems_ the 3805 BIOS is an improvement, at least in my setup.
ID: 1926940 · Report as offensive
1 · 2 · 3 · 4 . . . 5 · Next

Message boards : Number crunching : (0xC0000005) STATUS_ACCESS_VIOLATION any idea what caused it?? Or what it exactly is describing?? Internet searches find nothing.


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.