NUC5i5RYH fTPM BIOS bug

[nop]

The BIOS of NUC5 has a bug which prevents the use of the fTPM (Intel PTT) on Linux.

The underlying problem is that the BIOS does not reserve the memory area 0xa2fff000-0xa2ffffff that it uses for the TPM control area.

I've detailed the tests, proof of the origin of the bug, and a temporary workaround, in this github bug report.

Please transmit this information to the BIOS developers so this 5 years old bug can finally be fixed.

[AndrewG_Intel]

Hello nop

Thank you for posting on the Intel® communities.

We would like to inform you that due to the Intel® NUC Kit NUC5i5RYH has been discontinued, Intel Customer Service no longer supports inquiries for it, but perhaps fellow community members have the knowledge to jump in and help. Unfortunately, since the Intel® NUC Kit NUC5i5RYH has been discontinued, there is no further support for reports of bugs. You may also find the Discontinued Products website helpful to address your request. Thank you for understanding.

https://www.intel.com/content/www/us/en/support/discontinued-products.html 

You can verify this product's discontinuance status at http://ark.intel.com/ > Product Status > "Discontinued". 

If you need any additional information, please submit a new question as this thread will no longer being monitored.

Best Regards,

Andrew G.

Intel Customer Support Technician

[nop]

@AndrewG_Intel 

Thanks for your response.

Other community members cannot "jump in and help" on a bug which must be fixed in the BIOS source code to which only Intel developers have access.

The NUC5 PTT fTPM never worked on Linux because of this bug, and your support completely ignored all bug reports regarding this when this device was not "discontinued". This means an advertised feature of this device was never correctly working.

I've spent days of my time debugging and precisely identifying a bug in your products. The least thing you can do is transmit this info so the BIOS developers can fix the 1 line of code needed. With the description I provided, it should not take more than 1 minute for a developer to fix this bug.

Besides the NUC5 BIOS still receives bug fixes regularly. Why not finally fix this 5 years old bug at the same time?

[AndrewG_Intel]

Hello nop

Thank you for your response and for taking the time to report this and for sharing your feedback and details about your debugging. We do appreciate your interest in making Intel's products better.

It is worth mentioning that for the Intel® NUC Kit NUC5i5RYH only Windows® 10/8.1/7 have been validated by Intel®. This can be confirmed in the below links:

Supported Operating Systems for Intel® NUC Products: https://www.intel.com/content/www/us/en/support/articles/000005628/intel-nuc.html

Linux* Support for Intel® NUC: https://www.intel.com/content/www/us/en/support/articles/000005499/intel-nuc.html

Nevertheless, we have shared your feedback with the proper team; however, we cannot promise any action will be taken (e.g. BIOS update releases) since this NUC has been discontinued and Linux operating systems have not been validated by Intel®.

We will proceed to close the thread now. Also, we as a community will not have further visibility on the status of the request. If you need any additional information, please submit a new question as this thread will no longer being monitored. Thank you for your understanding.

Best regards,

Andrew G.

Intel Customer Support Technician

[GerKel]

Whilst we're at it: Also declared 'unsupported' these NUC5I5RY# still receive faulty BIOS 'updates' so perhaps you could share this report with the proper team too?!

I used several NUC5I3RYH with archlinux and all was OK till I unfortunately updated the BIOS of one of them.
Seems like I did not have learned from the last desaster.
Remember the half-baked Version RY0368 which partially bricked one of my NUCs?

While with BIOS 366 there where no issues now with BIOS Ver. 385 I get the well for bad BIOS implementations known infamous:

kernel: nct6775: Found NCT6776D/F or compatible chip at 0x4e:0xa00
kernel: ACPI Warning: SystemIO range 0x0000000000000A05-0x0000000000000A06 conflicts with OpRegion 0x0000000000000A05-0x0000000000000A06 (\_TZ.TZ00.DEB3) (20200528/utaddress-204)
kernel: ACPI: If an ACPI driver is available for this device, you should use it instead of the native driver

in journal / logs according to the sensors chip NCT6776D/F.
Of course the chip stopped working and you cannot access its readings anymore. 

As most of my NUCs run passively cooled this is very bad as I need the temperature output of these sensors.

Both NUCs are 100% identical (beside the BIOS version), the OS is cloned from the one to the other so there are no differences in OS configuration settings. The only difference is the BIOS version.

No thanks to you as you now hide every old BIOS version from your customers *grrr* I was able to downgrade the BIOS of my affected NUC to Ver. 371 which I fortunately still had saved somewhere - and guess what:

With this BIOS RY0371 the ACPI error is gone and the sensors chip works again!

So the blame is definitely up to one of your latest BIOS version causing those annoying ACPI errors according to the sensors module 'nct6775' when using linux.

So dear Intel - could you please fix this - again - because I hate such error messages in my logs and I would be very grateful if I could read out the sensors on my board again using a recent BIOS version.

Thank you in advance.