Sun Fire[TM] V100 server: Diagnosing memory errors with CPU's 550MHz and above. (Doc ID 1010879.1)

Last updated on JULY 01, 2016

Applies to:

Sun Netra X1 Server - Version All Versions and later
Sun Fire V100 Server - Version All Versions and later
All Platforms

Symptoms

Error Checking and Correction(ECC) events on DIMMs installed in
memory slots 2 and 3, are not correctly diagnosed in Open Boot
PROM(OBP) 4.0.18, for the Sun Fire[TM] V100 server(flapjack_lite)
platform, and platforms sharing a similar architecture like the
Netra x1.

This has lead to the field mis-diagnosing the replacement part when
an ECC error occurs and in some instances has caused the entire
platform FRU to be swapped.

Typically, OBP can name a failing DIMM if the ECC error is within the
address range of 0x00000000-0x7fffffff.

However, when the UltraSPARC IIe processor was extended to scale
beyond 500MHz, the memory addressing capability was doubled to
the range 0x00000000-0xffffffff. (550-650MHZ CPU's) OBP does not
correctly identify an ECC event present above 0x7fffffff and
reports an invalid syndrome.

Typically, only DIMMs installed in DIMM slots 2 & 3 will be mapped
to this high address range unless DIMMs of mixed sizes are installed.

Cause

Sign In with your My Oracle Support account

Don't have a My Oracle Support account? Click to get started

My Oracle Support provides customers with access to over a
Million Knowledge Articles and hundreds of Community platforms