Bug 79341 - [macbookpro] Firmware bug PCI: Using host bridge windows from ACPI
Summary: [macbookpro] Firmware bug PCI: Using host bridge windows from ACPI
Status: CLOSED DOCUMENTED
Alias: None
Product: ACPI
Classification: Unclassified
Component: BIOS (show other bugs)
Hardware: Intel Linux
: P1 normal
Assignee: Xing Zhengjun
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2014-07-01 03:24 UTC by Chris Murphy
Modified: 2014-12-02 14:42 UTC (History)
1 user (show)

See Also:
Kernel Version: 3.16.0-0.rc2.git1.1.fc21.x86_64
Subsystem:
Regression: No
Bisected commit-id:


Attachments
journal (1.95 MB, text/plain)
2014-07-01 03:26 UTC, Chris Murphy
Details
acpidump (185.57 KB, text/plain)
2014-07-01 03:26 UTC, Chris Murphy
Details
dmidecode (28.11 KB, text/plain)
2014-07-01 03:27 UTC, Chris Murphy
Details
dmesg (131.69 KB, text/plain)
2014-07-01 03:27 UTC, Chris Murphy
Details
lspci (16.66 KB, text/plain)
2014-07-01 03:28 UTC, Chris Murphy
Details

Description Chris Murphy 2014-07-01 03:24:58 UTC
MacbookPro 8,2 booted in EFI mode reports the following ACPI related bug and warning messages:

[    0.273602] [Firmware Bug]: ACPI: BIOS _OSI(Linux) query ignored
[    2.320358] PCI: Using host bridge windows from ACPI; if necessary, use "pci=nocrs" and report a bug

[   13.214077] ACPI Warning: SystemIO range 0x000000000000efa0-0x000000000000efbf conflicts with OpRegion 0x000000000000efa0-0x000000000000efa
f (\_SB_.PCI0.SBUS.SMBI) (20140424/utaddress-258)
[   13.215423] ACPI: If an ACPI driver is available for this device, you should use it instead of the native driver
[   13.299527] microcode: CPU0 sig=0x206a7, pf=0x10, revision=0x1a

Hardware has latest firmware update from Apple.

The main thing that concerns me is there are mcelog events due to high temps with minimal workloads. So at the moment I have minimal confidence I should even be running Linux on this hardware considering how hot it gets, how high the fans run, and how much the kernel seems to complain.

[  600.718678] rawhide.localdomain mcelog[939]: MCE 2
[  600.719811] rawhide.localdomain mcelog[939]: CPU 6 THERMAL EVENT TSC b7ea03e138
[  600.720928] rawhide.localdomain mcelog[939]: TIME 1404182780 Mon Jun 30 22:46:20 2014
[  600.722102] rawhide.localdomain mcelog[939]: Processor 6 below trip temperature. Throttling disabled
[  600.723250] rawhide.localdomain mcelog[939]: STATUS 88010282 MCGSTATUS 0
[  600.724393] rawhide.localdomain mcelog[939]: MCGCAP c09 APICID 5 SOCKETID 0
[  600.725410] rawhide.localdomain mcelog[939]: CPUID Vendor Intel Family 6 Model 42
[  600.726418] rawhide.localdomain mcelog[939]: Hardware event. This is not a software error.
Comment 1 Chris Murphy 2014-07-01 03:26:36 UTC
Created attachment 141611 [details]
journal

journalctl -l -b -o short-monotonic
Comment 2 Chris Murphy 2014-07-01 03:26:50 UTC
Created attachment 141621 [details]
acpidump
Comment 3 Chris Murphy 2014-07-01 03:27:15 UTC
Created attachment 141641 [details]
dmidecode
Comment 4 Chris Murphy 2014-07-01 03:27:53 UTC
Created attachment 141651 [details]
dmesg
Comment 5 Chris Murphy 2014-07-01 03:28:21 UTC
Created attachment 141661 [details]
lspci
Comment 6 Chris Murphy 2014-07-01 03:32:09 UTC
The overheating messages and mcelog messages are regressions since kernel 3.9.0, they didn't occur with 3.8 and older kernels.
https://bugzilla.redhat.com/show_bug.cgi?id=924570
Comment 7 Zhang Rui 2014-12-02 14:42:23 UTC
For the "[Firmware Bug]: ACPI: BIOS _OSI(Linux) query ignored" message, this just indicates a BIOS issue and we can ignore it.
For the "PCI: Using host bridge windows from ACPI; if necessary, use "pci=nocrs" and report a bug", this is just a kernel information message, which suggests user to file a bug if the current setting doesn't work and boot option "pci=nocrs" is needed.
And for the ACPI warning message, you can also ignore it only if some thing does not work.

For the overheating messages, I don't know why this is an regression because there is no code change between 3.4 and 3.11. But as the subject of this bug report does not reflect this, I suggest you to file a new bug report about the overheating message against Power/Thermal category and attach the output of "grep . /sys/class/thermal/*/*" and dmesg both in good (3.8) and bad kernel (3.9).


Bug closed.

Note You need to log in before you can comment on or make changes to this bug.