Firmware and acpi messages

While I do not suffer from this posts issues, I have noted on my dmesg occasional whinning about ACPI and bugs in ACPI.

My issue is that my Intel firmware is 2018 latest vintage, after which they’ve stopped supporting my 2013 mobo.

This thread reminded me that I have a time clock, i.e. it’s a matter of time before I suffer these problems. Therefore I would like to know if your are aware of means of updating firmware from 3rd party sources or otherwise by me manipulating it with latest h/w manufactuer updates. Eg my f/w has no nvme f/w. But I can use the efi commands to install it separately and then boot (I normally do it via ssd /boot). Loce to put this nvme driver into my f/w somehow.

The reason I persist with my mobo that is in the IT world ancient, is it’s dual Xeon e5-26XX with huge amount of grunt and still runs everything very nicely compared to even the latest mobos. In H/W terms nothing has really changed other then more modern GPU which I’ve replaced and internal latest generation CPU instruction upgrade relate to AI, crypto and so on I hardly need.

The thing that limits me continuing with my dinosaur is firmware updates. To me it seems like a waste of precious hardware that has plenty of life under linux left in it.

Would love to see a opensource fw system supported by linux.

'Tis a common tale but true. Vendors make money selling new hardware, supporting elderly hardware is a cost. Intel has had a long period of really marginal year over year performance gains. I used a SATA SSD in a 2007 system until the graphics hardware died last winter.

I suspect your elderly system can still be used as a space heater. At work we were offered use of a dual Xeon Intel workstation configured to do processing in cluster when not otherwise busy, At startup it went full-on cluster workload and then shut down due to UPS overloading before anyone could log in. We had to move it to a machine room with a much heavier UPS.

That is very common to see kernel apci related messages in dmesg, but it does not seem to matter.

There is a devel@ thread about an ACPI related bug in linux 6.3 series that might explain this. I haven’t read the full thread. But it seems like a low risk test to see if you’re affected by booting with:

module_blacklist=ucsi_acpi

The devel thread mentions there is a fix pending upstream, and that Fedora will carry a downstream patch to get the fix to users a little faster.

1 Like