Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Solved HWinfo Vega 64 intermittent crash
#1
Since installing HWinfo64 Beta 561 3293 I'm getting my display driver crashed when starting HWinfo intermittently. Black screen. Reset doesn't help, need to kill the PC and restart.

This is with Radeon Drivers 17.11.4 on a Vega 64 Reference (with Morpheus II cooler swap, but standard Vega 64 BIOS.

Edit: trying to capture a crash with debug logging enabled.
Reply
#2
Did it perhaps start to crash when you upgraded to Crimson 17.11.4 ?
It seems that this is because AMD has finally unblocked I2C access to VRMs on Vega with latest drivers.
I made some tests and can now see all GPU VRM details for both Core and Memory VRMs without problems.
But on the other hand, some users started to experience crashes during HWiNFO startup, which happen right when HWiNFO tries to access those VRMs.
Disabling the "GPU I2C Support" option should resolve this problem.
Reply
#3
This problem should be resolved in HWiNFO v5.70, which by default disables GPU I2C scan on Vega to avoid a potential crash.
For those who would like to take the risk of monitoring VRM parameters, use the "GPU I2C Support Force" option to override this safety feature and force HWiNFO to access the VRMs.
Reply
#4
okay. I ran into the same issue with a vega56, on hwi 5.6.x.
Reply
#5
And I almost disassembled my pc already, looking for the crash, I thought the 2nd mainboard is defective
Reply
#6
I realize this is a zombie thread, but I'm still having this issue with 5.82. My system config is very odd, and I'm wondering if that's the problem: I have a Vega FE and a Titan Xp in the same machine. I've attempted to disable gpu i2c support entirely through settings, and I still get this hard crash. Does anyone have any ideas?
Reply
#7
AFAIK there are no issues running HWiNFO on Vega GPUs, all problems were solved in the past. So most probably your problem is your hybrid GPU configuration.
If you attach the HWiNFO Debug File of the crash, I can have a look at that.
Reply
#8
Apologies for the delayed reply.  Here's my latest debug and log file fresh off a crash.


Attached Files
.csv   logfile.CSV (Size: 974.62 KB / Downloads: 0)
.dbg   HWiNFO64.DBG (Size: 1.1 MB / Downloads: 1)
Reply
#9
Is the attached DBG file capturing the crash? Because I cannot see any problem in that dump.
Note that if you have HWiNFO configured for autostart with Windows then a subsequent start of HWiNFO after a reboot will overwrite the previously produced DBG.
Reply
#10
(05-08-2018, 09:00 AM)Martin Wrote: Is the attached DBG file capturing the crash? Because I cannot see any problem in that dump.
Note that if you have HWiNFO configured for autostart with Windows then a subsequent start of HWiNFO after a reboot will overwrite the previously produced DBG.

Yes, it's set to autostart.  I will recrash the system, and get back to you.
Reply
#11
sorry for the delay again, but here's a crash with autostart disabled.


Attached Files
.zip   HWiNFO64 (2).zip (Size: 873.6 KB / Downloads: 1)
Reply
#12
Was the "Debug Write Direct" option enabled when you created the dump?
Reply
#13
yes it was enabled.
Reply
#14
Well, it looks like the dump doesn't provide enough data to determine the problem.
Try to disable monitoring of some sensors (hit Del over their headings) and see which one of them disabled doesn't cause the crash.
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)