RGB Fusion version B22.0329.1 and above causing system hang with HWi

[Ultraboss]

Hi there,
recently updated rgb fusion on 14/04/2022, then launched HWi as usual to check sensors, then after few minutes, got a black screen, system has hung.
I needed to power off completely (unplug sector) to be able to boot again, then relaunched all. Got a "bsod whea uncorrectable error" this time...
What I have updated was : rgb fusion, app center, Kaspersky, and some windows update.
Trying to update my old HWi version (7.03) to 7.22, same issue.
My first guess was to try running HWi without rgb fusion because I don't launch it all the time.
So running HWi 7.22 only was OK
Running rgb fusion only was OK
Both running causes system hang after few hours of pc idle...
I validated that rgb fusion above B22.0329.1 is in cause, rollbacked to B21.1116.1 was OK !
Maybe it is due to new ene ddr driver or kind from gigabyte (there are new led animation for ddr).
I can't attach debug log file because it is too big.

So it is not an issue with HWi directly, but just to inform users that are using a gigabyte/aorus mb with rgb fusion and HWi
I will also report this on gigabyte forum.

 

[Martin]

Try to compress the Debug File before attaching it here, I'd still like to have a look at it.
Yes, please report this GIGABYTE as well.

 

[Ultraboss]

It has been already compressed, but it is around 42Mio
You can download it there : link removed

 

[Martin]

Looks like the problem might be a conflict when both tools access the ISL69138 PWM VRM.
In such case disabling monitoring of this entire sensor in HWiNFO or excluding the SMBus address 41 might mitigate the problem. But this should still be looked into by GIGABYTE.

 

[Ultraboss]

Thanks for the analysis !
Can you tell me why I have already the address 0x69 disabled in my settings ?
How did you found that it is the ISL69138 sensor on 0x41 ?

 

[Martin]

I don't know why you have that address disabled
I saw 0x41 in the Debug File you provided.

 

[Ultraboss]

I remember that HWi automatically ask to disable some addresses for compatibility or known issues. But maybe no more on newer versions.
For 0x41 I meant how did you "guess" it could be an access conflict on the ISL69138 sensor ? Just for my personal knowledge

 

[Martin]

Access of that sensor was the last operation I saw in the DBG file.

 

[Ultraboss]

I checked another debug file, but this one ends on :

SM2
Sensor Disp End

I would try another catch...

 

[Martin]

Try to exclude that address and you'll see if it still crashes.

 

[Ultraboss]

I have been running 1 day without system hang, by disabling address 0x41

 

[Martin]

Thanks for the feedback. So it must be some collision when both tools access the same device, but I doubt RGB Fusion supports any synchronization mechanisms with others.

 

[Ultraboss]

Support ticket opened on Gigabyte side, I will keep you in touch.
Thanks for the help

 

[Ultraboss]

Hi there,
it looks like new version B22.0713.1 fixes the issue with SMBus as I am able to reenable VRM monitoring without any hang

 

[Martin]

Nice Thanks for letting us know!