MCA: CPU 1 COR (1) internal parity error

[Rutgzven]

Errata for 4th gen Intel CPU:s (Haswell):

HSD131. Spurious Corrected Errors May be Reported
Problem: Due this erratum, spurious corrected errors may be logged in the MC0_STATUS register
with the valid (bit 63) set, the uncorrected error (bit 61) not set, a Model Specific Error
Code (bits [31:16]) of 0x000F, and an MCA Error Code (bits [15:0]) of 0x0005. If CMCI
is enabled, these spurious corrected errors also signal interrupts
Implication: When this erratum occurs, software may see corrected errors that are benign. These
corrected errors may be safely ignored.
Workaround: None identified.
Status: For the steppings affected, see the Summary Table of Changes.

http://www.intel.com/content/dam/ww...-core-family-desktop-specification-update.pdf

Might be related?

edit: i have done the bit calculations and concluded that HEX error codes that
Starts with one 8, 9, C or D, and ends with F0005 falls into the description of this errata.


My conclusion (so far):

If you have one of the CPU models described in the errata, and get an error of:

MCA: Bank x, Status 0x8xxxxxxxxxxF0005
MCA: Bank x, Status 0x9xxxxxxxxxxF0005
MCA: Bank x, Status 0xCxxxxxxxxxxF0005
MCA: Bank x, Status 0xDxxxxxxxxxxF0005

...it could be due to this error in the CPU.

 

[Rutgzven]

http://svnweb.freebsd.org/base?view=revision&revision=269052

Code:

253            /*
254                   * Skip spurious corrected parity errors generated by desktop Haswell
255                   * (see HSD131 erratum) unless reporting is enabled.
256                   * Note that these errors also have been observed with DO-stepping,
257                   * while the revision 014 desktop Haswell specification update only
258                   * talks about CO-stepping.
259                   */
260                  if (rec->mr_cpu_vendor_id == CPU_VENDOR_INTEL &&
261                      rec->mr_cpu_id == 0x306c3 && rec->mr_bank == 0 &&
262                      rec->mr_status == 0x90000040000f0005 && !intel6h_HSD131)
263                          return (1);
264         
265                  return (0);

 

[no_connection]

Thanks.
Good thing it is benign.
So now I can start using VMs again.

Bad thing is filling up the log.