# What's behind the following ACPI error messages?

I am seeing the following ACPI error messages during boot for quite some time now. Can someone please explain what they mean and help me evaluate if they are of importance? Or even worth a bug report?

[0.079616] Error parsing PCC subspaces from PCCT
[0.104800] ACPI Error: Needed type [Reference], found [Integer] ffff8d2fcfbc90d8 (20170728/exresop-103)
[0.104819] ACPI Exception: AE_AML_OPERAND_TYPE, While resolving operands for [OpcodeName unavailable] (20170728/dswexec-461)
[0.104834] ACPI Error: Method parse/execution failed \_PR.CPU0._PDC, AE_AML_OPERAND_TYPE (20170728/psparse-550)


Any hints? Thanks!

edit retag close merge delete

Sort by » oldest newest most voted

I'll second the notion that firmware writers do not always create correct ACPI tables -- and the PCCT has had some changes to it very recently. What the messages say is that the ACPI tables have some incorrect content; without looking at the tables in detail, I'm guessing that at least the PCCT, and possibly the DSDT (or some SSDT) tables have a few entries that should really be cleaned up. I collect this kind of goofiness, by the way, so I'd love to have a copy of everything in /sys/firmware/acpi/tables, if you have a chance; it helps with my work on the ACPI standards committee.

The entries you have shown are probably harmless. In general, if there are ACPI errors, that means the kernel does not know how to find some bit of hardware and make it start up. And, in general, the kernel sets up reasonable defaults when there are errors. The PCC subspace error, for example, means that the kernel cannot communicate with firmware about something (it's used in several places) properly. If it is the PCC subspace used for adjusting CPU frequency (one of the possibilities -- we don't know from the messages alone), it might mean that the CPU always runs at the fastest possible speed. That's not necessarily a problem, though, even on a laptop where it might use more battery, because the firmware may take care of adjusting CPU frequency independently. You might never notice.

The first line shown is the PCC subspace problem; at least one of the subspaces could not be set up. The last three lines are probably part of a single error in the DSDT where CPU power adjustment capabilities are described, and ACPI does not understand the description. Again, the kernel default will be to just run at full power and let firmware adjust it if it can.

I would definitely complain to the hardware vendor about this. Any time there is an ACPI error, the hardware is not working at its best. It will work, and work reasonably well, but may not be able to take advantage of all the cool features the vendor has built in. If the vendor is responsive, they'll already have a firmware update available. If they are not, then filing a bug with them could help.

more

Thank yo uso much for your valuable input on this topic. I am sure I'll find an email address on your website once it's up and running again. I will then send you all the information needed.

BTW: CPU stepping is working properly.

I may need to send a bug report to Lenovo. Do you think they are interested in such stuff, or even listening? Do they have a public tracker for user feedback?

Again, many thanks!

( 2017-12-14 11:05:57 -0500 )edit

I don't know how to post a problem to Lenovo, personally. They should be interested in stuff like this, and I would hope they are; whether they listen or not, I don't know. OTOH, they won't know if they aren't told.

Just send the info to my ahs3 at fedoraproject.org address; that should work fine. Thanks!

( 2017-12-14 12:54:29 -0500 )edit

My BIOS firmware is from 03/2017 and the manufacturer's most recent one is dated 09/2017. Maybe I should flash that one...

( 2017-12-15 09:56:06 -0500 )edit

There are often ACPI error messages. It seems that the BIOS writers tend to be somewhat sloppy with following specs and only test on Windows. Most of the time the messages can just be ignored unless you know something isn't working properly.

more