Culprit is a Lenovo Thinkpad E540 (20C600) machine. To solve firmware related vulnerabilities, I installed when available the 2.26 BIOS version from Lenovo and corresponding kernel patches from Ubuntu. Lenovo recalled the BIOS to return to the 2.25 version (which I did) due to firmware flaws found by Intel. Ubuntu issued new kernels which I installed too.
Since then, I'm getting tons of MCE errors:
georges@espineux:~$ ll /var/log/mce*
-rw-r--r-- 1 root root 1474070 févr. 14 10:19 /var/log/mcelog
-rw-r--r-- 1 root root 150947 janv. 26 18:22 /var/log/mcelog.OLD
georges@espineux:~$ dmesg
...
[ 47.508238] hid-generic 0005:0000:0000.0001: input,hidraw0: BLUETOOTH HID v0.00 Mouse [Bluetooth Mouse] on 7c:5c:f8:29:e2:c5
[ 350.737496] mce: [Hardware Error]: Machine check events logged
[ 352.193960] mce: [Hardware Error]: Machine check events logged
[ 572.826735] mce_notify_irq: 1 callbacks suppressed
[ 572.826739] mce: [Hardware Error]: Machine check events logged
[ 748.950959] mce: [Hardware Error]: Machine check events logged
[ 1134.519367] mce: [Hardware Error]: Machine check events logged
[ 1134.570062] mce: [Hardware Error]: Machine check events logged
[ 1277.080985] mce: [Hardware Error]: Machine check events logged
[ 1384.073759] mce: [Hardware Error]: Machine check events logged
[ 1392.788087] mce: [Hardware Error]: Machine check events logged
[ 1458.624036] mce: [Hardware Error]: Machine check events logged
[ 1519.699866] mce: [Hardware Error]: Machine check events logged
[ 1887.815406] mce: [Hardware Error]: Machine check events logged
[ 1930.099318] mce: [Hardware Error]: Machine check events logged
[ 1963.753506] mce: [Hardware Error]: Machine check events logged
[ 2545.724632] mce: [Hardware Error]: Machine check events logged
[ 4390.100756] mce: [Hardware Error]: Machine check events logged
[ 5227.708637] mce: [Hardware Error]: Machine check events logged
[ 6228.314537] mce: [Hardware Error]: Machine check events logged
[ 6590.448387] mce: [Hardware Error]: Machine check events logged
[ 6646.747585] mce: [Hardware Error]: Machine check events logged
[ 6651.198427] mce: [Hardware Error]: Machine check events logged
[ 6657.065077] mce: [Hardware Error]: Machine check events logged
[ 6732.614503] mce_notify_irq: 1 callbacks suppressed
[ 6732.614504] mce: [Hardware Error]: Machine check events logged
[ 6762.581883] mce: [Hardware Error]: Machine check events logged
[ 7052.204827] mce: [Hardware Error]: Machine check events logged
[ 7297.349344] mce: [Hardware Error]: Machine check events logged
[ 7299.973406] mce: [Hardware Error]: Machine check events logged
[ 7872.192674] mce_notify_irq: 1 callbacks suppressed
[ 7872.192678] mce: [Hardware Error]: Machine check events logged
[ 8113.366287] mce: [Hardware Error]: Machine check events logged
[ 8274.092477] mce: [Hardware Error]: Machine check events logged
[ 8871.384635] mce: [Hardware Error]: Machine check events logged
[ 9107.149254] mce: [Hardware Error]: Machine check events logged
[ 9252.802930] mce: [Hardware Error]: Machine check events logged
[ 9292.786247] mce: [Hardware Error]: Machine check events logged
[ 9727.664285] mce: [Hardware Error]: Machine check events logged
[ 9995.683721] mce: [Hardware Error]: Machine check events logged
[10296.785257] mce: [Hardware Error]: Machine check events logged
[10396.229639] mce: [Hardware Error]: Machine check events logged
[10416.685511] mce: [Hardware Error]: Machine check events logged
[10477.233152] mce: [Hardware Error]: Machine check events logged
[10578.752495] mce: [Hardware Error]: Machine check events logged
I've checked every piece of hardware i could : 48 hours memory test without any error, Tried to decode some of the TSC to find more using this: https://unix.stackexchange.com/questions/165222/mce-error-mca-internal-parity-error but to no avail.
I've used the Lenovo provided hardware tests without anything wrong. I've tried to run the 17.10.1 Ubuntu on an usb stick and got no errors. So I bet these (mainly parity errors) are harmless but they could bury deep a real problem. So what are my options? Is there a solution from Ubuntu in the pipeline? Did I miss something?
P.S. : I was told this question could be a duplicate. Alas, it is not as the temperature never goes above 54°C, both thermald and intel-microcode are installed and running.
The error is quite always an internal parity error like the following one. they affect all 4 CPU in various percentages.
$ tail /var/log/mcelog
CPU 0 BANK 0 TSC 1a086c5a3800
TIME 1518850510 Sat Feb 17 07:55:10 2018
MCG status:
MCi status:
Corrected error
Error enabled
MCA: Internal parity error
STATUS 90000040000f0005 MCGSTATUS 0
MCGCAP c07 APICID 0 SOCKETID 0
CPUID Vendor Intel Family 6 Model 60
Thank you for your help.
