K
Krzysiek Setlak
TL;DR: I'd like to ask you guys for some guidance diagnosing "IRQL_NOT_LESS_OR_EQUAL" BSODs.
======================
Hi guys! It's my first post here (yeah, you guessed it, I registered specifically to post this here ;P ) and it will be lengthy... But it's all with best intentions to be as detailed as possible so I hope you bear with me.
The issue is as follows: when I stress my PC a bit by running a game that has some more substantial hardware requirements (recently it was Borderlands 3, but for example GTA V also managed to give my PC a hard time...) I get a BSOD (OK, GSOD to be exact as it's very, very green!
) and all the error description is "IRQL_NOT_LESS_OR_EQUAL". So no stopcode. This started happening after an extensive hardware upgrade, but not right after it - I managed to game on it for quite a bit and I played these exact titles that crash the PC now. I guess I got at least a solid 100 h of uninterrupted, correct operation with new hardware installed before it started crashing.
Of course I've done my best not being an ignorant slacker and googled around a bit before posting, but all I got was VERY general and I couldn't find any info on how to pinpoint which exact piece of hardware gives me my bad time. So here's the list of suspects in our little investigation ranked from most to least likely to be the culprit.
- Motherboard;
- PSU;
- RAM;
- GPU;
- CPU;
- filesystem / a hardware issue with storage.
Build
So the PC is based on a build I've composed waaaaay bac (2009 as far as I could establish...) and back then it was pretty ****... It server me well, but ultimately a graphics card not supporting DX11 i DX12 and a slow magnetic HDD had to go. So I presented myself with a challenge: I'll upgrade the rig, but squeezing an amount of bang for the buck that would make daddy proud. No mercy, it had all to be cheap and perform well. So here it is (for your convenience I marked what pieces of hardware are a part of the original build and which were installed recently):
[OLD] Motherboard: ASUS M4A79T Deluxe
[OLD] CPU: AMD Phenom II X4 955
[NEW] GPU: (AMD) Maxsun Radeon RX 580 2048SP
RAM:
[OLD] - 2 x Kingston 2048 MB DDR3 1333 MHz (RAM 9905403-011.A03LF 2048MB DIMM DDR3 1333MHz)
[NEW] - 2 x PLEXHD 4096 DDR3 1333 MHz (CAS latency: CL9; Memory timing sequence: 9-9-9-24)
[NEW] Storage:
- LVCARDS SSD 256 GB
[OLD] Case: CoolerMaster Elite 331 (+ szyba)
[OLD] PSU: CoolerMaster RS-520-ASAA-A1 520W
My take on the list of suspects
Motherboard: Main suspect. An old lady. Moreover, I've been clocking my CPU a tad up after the upgrade (core clock multiplier x16 (default) -> x17) and maybe the motherboard got a bit damaged because of this.
PSU: Another part that may well be dying of old age. The system only BSODs when CPU, GPU and RAM are all very busy at the same time, so this might indicate that the power supply doesn't manage.
RAM: Two modules are chinese. From my experience cheap Chinese RAM can be troublesome even if passing the Windows Memory Diagnostic test. What speaks in favor of innocence of RAM is that the problem started occurring after a while while I guess if RAM was borked, it wouldn't work correctly from the get-go.
GPU: Another part that is a suspect because of the nationality
Maybe some driver update introduced a bug or maybe it just burned in after a few hours and some manufacturing fault started to show its teeth?
CPU: On the list because of overclocking mentioned above. Maybe it's fried a little... But on the other hand if it was so, then a stress test of the CPU itself would crash the system.
filesystem or a hardware sotrage fault: It's here just because some articles on the web mention that IRQL_NOT_LESS_OR_EQUAL may be caused by this. But I doubt that, besides I've checked the FS on the drive multiple times and it was fine each time.
What do I really want from you?
I'd like to please get some solid advice on how to diagnose this and how to isolate the exact piece of hardware that's failing. Like, how to determine that it's the MoBo and not the PSU...
Please remember, that the challenge I've set for myself is to be cost effective and solve issues with more of brainpower and less of walletpower. Of course I could buy a brand new mid-tier rig and be done with it, but that's giving up and fighting the problem with my bare fists is part of the fun (perhaps more than using the rig after it's fixed ) So while I am prepared that maybe the only way is to give the PC to a professional PC service center, I think of it as a last resort and I prefer to tinker myself.
What I've tried so far
Windows memory diagnostic, MemTest x86 and trying my RAM dice in all possible configs (only old, only new, various combos of old and new in different slots)
The full diagnostic cycle revealed no problems with RAM. However, I know from experience this doesn't guarantee that everything is fine... At this point I'm quite confident it's not RAM.
Reducing / increasing voltages on the motherboard
I tried changing some voltages up and down and nothing changes. This doesn't mean this definitely cannot help as my changes were pretty uninformed - I'm a complete noob in all these north bridge, south bridge thingies.
Event Viewer
I've searched through the Windows Event Viewer (Event Viewer > Windows Logs > System) but to no avail.
Also, no memory dump was saved at C:\Windows\MEMDUMP.SYS.
Stress test with a benchmark
My rig has no problem completing the entire cycle of the "Superposition" benchmark on extreme settings (~20 FPS avg, but that's not the point, the point is to push the hardware). This is probably because it mostly pushes the GPU, but RAM - not so much. The problem seems to mostly occur when many subsystems are under pressure at one time, that is why the motherboard and the PSU are my main suspects.
Different GPU driver versions
I've tried installing two versions of drivers for my GPU with release dates spaced in time quite a bit - no difference.
Full system reinstall
I've reinstalled the system completely - no improvement.
Core dumps
I've recreated the error making two core dumps. They're here along with results of their analysis using WinDbg (X86) and output from MSInfo.
So that's all, folks. If you managed to push through all this ton of text - thanks and hope some of you guys will be able to help me!
Continue reading...
======================
Hi guys! It's my first post here (yeah, you guessed it, I registered specifically to post this here ;P ) and it will be lengthy... But it's all with best intentions to be as detailed as possible so I hope you bear with me.
The issue is as follows: when I stress my PC a bit by running a game that has some more substantial hardware requirements (recently it was Borderlands 3, but for example GTA V also managed to give my PC a hard time...) I get a BSOD (OK, GSOD to be exact as it's very, very green!
) and all the error description is "IRQL_NOT_LESS_OR_EQUAL". So no stopcode. This started happening after an extensive hardware upgrade, but not right after it - I managed to game on it for quite a bit and I played these exact titles that crash the PC now. I guess I got at least a solid 100 h of uninterrupted, correct operation with new hardware installed before it started crashing.Of course I've done my best not being an ignorant slacker and googled around a bit before posting, but all I got was VERY general and I couldn't find any info on how to pinpoint which exact piece of hardware gives me my bad time. So here's the list of suspects in our little investigation ranked from most to least likely to be the culprit.
- Motherboard;
- PSU;
- RAM;
- GPU;
- CPU;
- filesystem / a hardware issue with storage.
Build
So the PC is based on a build I've composed waaaaay bac (2009 as far as I could establish...) and back then it was pretty ****... It server me well, but ultimately a graphics card not supporting DX11 i DX12 and a slow magnetic HDD had to go. So I presented myself with a challenge: I'll upgrade the rig, but squeezing an amount of bang for the buck that would make daddy proud. No mercy, it had all to be cheap and perform well. So here it is (for your convenience I marked what pieces of hardware are a part of the original build and which were installed recently):
[OLD] Motherboard: ASUS M4A79T Deluxe
[OLD] CPU: AMD Phenom II X4 955
[NEW] GPU: (AMD) Maxsun Radeon RX 580 2048SP
RAM:
[OLD] - 2 x Kingston 2048 MB DDR3 1333 MHz (RAM 9905403-011.A03LF 2048MB DIMM DDR3 1333MHz)
[NEW] - 2 x PLEXHD 4096 DDR3 1333 MHz (CAS latency: CL9; Memory timing sequence: 9-9-9-24)
[NEW] Storage:
- LVCARDS SSD 256 GB
[OLD] Case: CoolerMaster Elite 331 (+ szyba)
[OLD] PSU: CoolerMaster RS-520-ASAA-A1 520W
My take on the list of suspects
Motherboard: Main suspect. An old lady. Moreover, I've been clocking my CPU a tad up after the upgrade (core clock multiplier x16 (default) -> x17) and maybe the motherboard got a bit damaged because of this.
PSU: Another part that may well be dying of old age. The system only BSODs when CPU, GPU and RAM are all very busy at the same time, so this might indicate that the power supply doesn't manage.
RAM: Two modules are chinese. From my experience cheap Chinese RAM can be troublesome even if passing the Windows Memory Diagnostic test. What speaks in favor of innocence of RAM is that the problem started occurring after a while while I guess if RAM was borked, it wouldn't work correctly from the get-go.
GPU: Another part that is a suspect because of the nationality
Maybe some driver update introduced a bug or maybe it just burned in after a few hours and some manufacturing fault started to show its teeth?CPU: On the list because of overclocking mentioned above. Maybe it's fried a little... But on the other hand if it was so, then a stress test of the CPU itself would crash the system.
filesystem or a hardware sotrage fault: It's here just because some articles on the web mention that IRQL_NOT_LESS_OR_EQUAL may be caused by this. But I doubt that, besides I've checked the FS on the drive multiple times and it was fine each time.
What do I really want from you?
I'd like to please get some solid advice on how to diagnose this and how to isolate the exact piece of hardware that's failing. Like, how to determine that it's the MoBo and not the PSU...
Please remember, that the challenge I've set for myself is to be cost effective and solve issues with more of brainpower and less of walletpower. Of course I could buy a brand new mid-tier rig and be done with it, but that's giving up and fighting the problem with my bare fists is part of the fun (perhaps more than using the rig after it's fixed
) So while I am prepared that maybe the only way is to give the PC to a professional PC service center, I think of it as a last resort and I prefer to tinker myself.What I've tried so far
Windows memory diagnostic, MemTest x86 and trying my RAM dice in all possible configs (only old, only new, various combos of old and new in different slots)
The full diagnostic cycle revealed no problems with RAM. However, I know from experience this doesn't guarantee that everything is fine... At this point I'm quite confident it's not RAM.
Reducing / increasing voltages on the motherboard
I tried changing some voltages up and down and nothing changes. This doesn't mean this definitely cannot help as my changes were pretty uninformed - I'm a complete noob in all these north bridge, south bridge thingies.
Event Viewer
I've searched through the Windows Event Viewer (Event Viewer > Windows Logs > System) but to no avail.
Also, no memory dump was saved at C:\Windows\MEMDUMP.SYS.
Stress test with a benchmark
My rig has no problem completing the entire cycle of the "Superposition" benchmark on extreme settings (~20 FPS avg, but that's not the point, the point is to push the hardware). This is probably because it mostly pushes the GPU, but RAM - not so much. The problem seems to mostly occur when many subsystems are under pressure at one time, that is why the motherboard and the PSU are my main suspects.
Different GPU driver versions
I've tried installing two versions of drivers for my GPU with release dates spaced in time quite a bit - no difference.
Full system reinstall
I've reinstalled the system completely - no improvement.
Core dumps
I've recreated the error making two core dumps. They're here along with results of their analysis using WinDbg (X86) and output from MSInfo.
So that's all, folks. If you managed to push through all this ton of text - thanks and hope some of you guys will be able to help me!
Continue reading...