H
hyc
I'm seeing a bunch of these messages on my system (2.6.18.3) which I
presume is telling me about ECC errors. Since the kernel hasn't aborted
I'm going to assume they're non-fatal, corrected errors. But can anyone
glean any more information out of them than that? I find it suspicious
that the address is almost always the same in the low order 20 bits.
But the CPU x Bank y doesn't seem to actually be telling me which DIMM
is the problem.
The machine is an Asus A8V Deluxe with X2 3800+ and 4 1GB ECC DIMMs.
Dec 3 06:03:31 mandolin kernel: Machine check events logged
Dec 3 06:08:31 mandolin kernel: Machine check events logged
Dec 3 06:13:31 mandolin kernel: Machine check events logged
Dec 3 06:18:31 mandolin kernel: Machine check events logged
Dec 3 06:18:46 mandolin mcelog: CPU 1 BANK 2 TSC 329b9009c0a92
Dec 3 06:18:46 mandolin mcelog: STATUS d000400000000863 MCGSTATUS 0
Dec 3 06:18:46 mandolin mcelog: CPU 0 BANK 2 TSC 32a37f396d7e5
Dec 3 06:18:46 mandolin mcelog: STATUS d000400000000863 MCGSTATUS 0
Dec 3 06:18:46 mandolin mcelog: CPU 0 BANK 4 TSC 32a37f396de57
Dec 3 06:18:46 mandolin mcelog: ADDR 970666e0
Dec 3 06:18:46 mandolin mcelog:
Dec 3 06:18:46 mandolin mcelog: STATUS 9433c00100000813 MCGSTATUS 0
Dec 3 06:18:46 mandolin mcelog: CPU 1 BANK 2 TSC 32a9a97b0038d
Dec 3 06:18:46 mandolin mcelog: STATUS d000400000000863 MCGSTATUS 0
Dec 3 06:18:46 mandolin mcelog: CPU 0 BANK 4 TSC 32a9a7318deb4
Dec 3 06:18:46 mandolin mcelog: ADDR b67666e0
Dec 3 06:18:46 mandolin mcelog:
Dec 3 06:18:46 mandolin mcelog: STATUS d433c00200000813 MCGSTATUS 0
Dec 3 06:18:46 mandolin mcelog: CPU 1 BANK 2 TSC 32b1d818d82f8
Dec 3 06:18:46 mandolin mcelog: STATUS 9000400000000863 MCGSTATUS 0
Dec 3 06:18:46 mandolin mcelog: CPU 0 BANK 2 TSC 32b1d5ce7def4
Dec 3 06:18:46 mandolin mcelog: STATUS d000400000000863 MCGSTATUS 0
Dec 3 06:18:46 mandolin mcelog: CPU 0 BANK 4 TSC 32b1d5ce7e606
Dec 3 06:18:46 mandolin mcelog: ADDR abf666e0
Dec 3 06:18:46 mandolin mcelog:
Dec 3 06:18:46 mandolin mcelog: STATUS 9433c00200000813 MCGSTATUS 0
presume is telling me about ECC errors. Since the kernel hasn't aborted
I'm going to assume they're non-fatal, corrected errors. But can anyone
glean any more information out of them than that? I find it suspicious
that the address is almost always the same in the low order 20 bits.
But the CPU x Bank y doesn't seem to actually be telling me which DIMM
is the problem.
The machine is an Asus A8V Deluxe with X2 3800+ and 4 1GB ECC DIMMs.
Dec 3 06:03:31 mandolin kernel: Machine check events logged
Dec 3 06:08:31 mandolin kernel: Machine check events logged
Dec 3 06:13:31 mandolin kernel: Machine check events logged
Dec 3 06:18:31 mandolin kernel: Machine check events logged
Dec 3 06:18:46 mandolin mcelog: CPU 1 BANK 2 TSC 329b9009c0a92
Dec 3 06:18:46 mandolin mcelog: STATUS d000400000000863 MCGSTATUS 0
Dec 3 06:18:46 mandolin mcelog: CPU 0 BANK 2 TSC 32a37f396d7e5
Dec 3 06:18:46 mandolin mcelog: STATUS d000400000000863 MCGSTATUS 0
Dec 3 06:18:46 mandolin mcelog: CPU 0 BANK 4 TSC 32a37f396de57
Dec 3 06:18:46 mandolin mcelog: ADDR 970666e0
Dec 3 06:18:46 mandolin mcelog:
Dec 3 06:18:46 mandolin mcelog: STATUS 9433c00100000813 MCGSTATUS 0
Dec 3 06:18:46 mandolin mcelog: CPU 1 BANK 2 TSC 32a9a97b0038d
Dec 3 06:18:46 mandolin mcelog: STATUS d000400000000863 MCGSTATUS 0
Dec 3 06:18:46 mandolin mcelog: CPU 0 BANK 4 TSC 32a9a7318deb4
Dec 3 06:18:46 mandolin mcelog: ADDR b67666e0
Dec 3 06:18:46 mandolin mcelog:
Dec 3 06:18:46 mandolin mcelog: STATUS d433c00200000813 MCGSTATUS 0
Dec 3 06:18:46 mandolin mcelog: CPU 1 BANK 2 TSC 32b1d818d82f8
Dec 3 06:18:46 mandolin mcelog: STATUS 9000400000000863 MCGSTATUS 0
Dec 3 06:18:46 mandolin mcelog: CPU 0 BANK 2 TSC 32b1d5ce7def4
Dec 3 06:18:46 mandolin mcelog: STATUS d000400000000863 MCGSTATUS 0
Dec 3 06:18:46 mandolin mcelog: CPU 0 BANK 4 TSC 32b1d5ce7e606
Dec 3 06:18:46 mandolin mcelog: ADDR abf666e0
Dec 3 06:18:46 mandolin mcelog:
Dec 3 06:18:46 mandolin mcelog: STATUS 9433c00200000813 MCGSTATUS 0