--Signature_Fri__30_Sep_2005_11_52_15_-0400_vF4JZEDbt.swNLXE
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: quoted-printable

Hello,

> Hi, I noticed this logged in /var/adm/messages on Ultra 10 (dual boot
> Solaris/NetBSD) while running Solaris:
>=20
> SUNW,UltraSPARC-IIi: [ID 657638 kern.info] [AFT0] Corrected Memory
> Error detected by CPU0, errID 0x000058ad.4cfd3df5
> 	AFSR 0x00000000.00100000<CE> AFAR 0x00000000.0813a850
> 	AFSR.PSYND 0x0000(Score 05) AFSR.ETS 0x00 Fault_PC 0x1038ca4
> UDBH Syndrome 0x1 Memory Module DIMM1=20
> SUNW,UltraSPARC-IIi: [ID 601383 kern.info] [AFT0] errID
> 0x000058ad.4cfd3df5
> Corrected Memory Error on DIMM1 is Persistent=20
> SUNW,UltraSPARC-IIi: [ID 705825 kern.info] [AFT0] errID
> 0x000058ad.4cfd3df5 ECC
> Check Bit  0 was in error and corrected
>=20
> I am a bit confused, is this message related to ECC error correction
> or does this indicate faulty memory?

Yes and maybe. If it happens often you should probably look for
replacement RAM.

> Which brings me to another question: how would this affect NetBSD?
> What I mean could this result in system crashes/freezes that I've been
> seeing with threaded applications?

It should cause a trap, I have no idea how NetBSD handles it. If it
panics and X is running you'd effectively get a freeze.

have fun
Michael

--Signature_Fri__30_Sep_2005_11_52_15_-0400_vF4JZEDbt.swNLXE
Content-Type: application/pgp-signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (NetBSD)

iQEVAwUBQz1fL8pnzkX8Yg2nAQItFQf/TskRhfntqpbyfc+c8AbjQ2WzOkTc913G
pJlRmf3DDNVb1OSWsEJuavD931YxtN2nAw21vGqiJnvwFf19/ayZYG3HdhniZBLI
DxdDf760jr11DSdB3uwJ7hVogb1Cym7A5aSMFYFaSoCGot3Hjfw05KaXw5xHqbxu
U50BN2oMxZyrDnsD7g9QRKkW4BVC99YzBD9m1hFZ0mE5zGwJ5LzrvN6MouQToP5f
q6QlS67oGWnbNbdcLZIQYf25kTBoYvx/rG0ALxg4loGuMmoTF+PdIXP+FUtrUc3F
HX3gdzBgrylB35MZgYuT6Blg59sz5/Ha0pUQKlL8s0thFakbHgwa+Q==
=vJZ7
-----END PGP SIGNATURE-----

--Signature_Fri__30_Sep_2005_11_52_15_-0400_vF4JZEDbt.swNLXE--