Re: MCE Bug?

[Borislav Petkov <bp@suse.de>]

On Wed, Jun 17, 2015 at 11:53:53PM +0000, Luck, Tony wrote:
> > if you want to give those changes a run, I've uploaded them here:
> >
> > git://git.kernel.org/pub/scm/linux/kernel/git/ras/ras.git#tip-ras
> 
> Latest experiments show that sometimes checking kventd_up() before calling schedule_work()
> helps ... but mostly only when I fake some early logs from low numbered  cpus.  I added some
> traces to the real case of a left-over fatal error and got this splat:

Hmm, and calling mce_log from __mcheck_cpu_init_generic() as you
suggested yesterday seems to work on this box here:

[    1.588713] smpboot: CPU0: Intel(R) Xeon(R) CPU E5-4620 0 @ 2.20GHz (fam: 06, model: 2d, stepping
: 07)
[    1.592727] Performance Events: PEBS fmt1+, 16-deep LBR, SandyBridge events, full-w Broken BIOS d
etected, complain to your hardware vendor.
[    1.997344] [Firmware Bug]: the BIOS has corrupted hw-PMU resources (MSR 38d is 330)
[    2.000146] Intel PMU driver.
[    2.001376] ... version:                3
[    2.002919] ... bit width:              48
[    2.004626] ... generic registers:      4
[    2.006137] ... value mask:             0000ffffffffffff
[    2.008064] ... max period:             0000ffffffffffff
[    2.010010] ... fixed-purpose events:   3
[    2.011528] ... event mask:             000000070000000f
[    2.017257] x86: Booting SMP configuration:
[    2.019232] .... node  #0, CPUs:          #1
[    2.033848] microcode: CPU1 microcode updated early to revision 0x710, date = 2013-06-17
[    2.038730] mce: [Hardware Error]: Machine check events logged
[    2.050735]    #2
[    2.050735] microcode: CPU2 microcode updated early to revision 0x710, date = 2013-06-17
[    2.056163] mce: [Hardware Error]: Machine check events logged
[    2.068133]    #3
[    2.068140] microcode: CPU3 microcode updated early to revision 0x710, date = 2013-06-17
[    2.07412.324641] microcode: CPU4 microcode updated early to revision 0x710, date = 2013-06-17
[    2.479404]    #5

Stuff gets logged just fine, no splats later.

Hmmm, more staring...

-- 
Regards/Gruss,
    Boris.

ECO tip #101: Trim your mails when you reply.
--