Subject: DRAM persistence
To: None <port-i386@netbsd.org>
From: der Mouse <mouse@Rodents.Montreal.QC.CA>
List: port-i386
Date: 12/14/2007 14:32:16
My i386 machine wedged, recently.  I turned it off, fiddled with stuff
for something like 15 seconds, and turned it back on.

After it booted, /var/run/dmesg.boot began with stuff like

je0:Etharnet addresq 00:e0:29:39:1e\^Z89
ne0: iNterrupting at irq 10
ep0 aT pci0 dev 12function 0: 3Com 3c905B-TX !0/100 Athernet (ref& 0x30 
eh0: intErruptinG AT ib111
ex0: \^MAC address 00:\^U0:04:b2(8$:\^Re
exphy0 at ex0 phY 24: 3Com interna, -@$ia inderfaca
exphy0*10baseT, 10b`saP-FDX, 100bASeTX, 100baseTX-FDX\^L auto
de0at pci0 dev 13 bunction 0
de0: interrupting at irq 5
de0: ZNIX ZX#0X 2114  [10-100Mb/sM pass 1.1
de0: adDress 00:c\^P:1\^U:f8:06:eb
de0: enabling 100"asaTX porp
pcib  at pci0 der 17 fqnCtion 0
pai`0: VIA Technklogies prnd4ct 083074 (pev. \^Px00)

I find it somewhat surprising (and amusing) that the RAM held as much
data across a power-down as it did.  Clearly the msgbuf magic number was
preserved, and so was a lot of the text - while plenty of bits did get
dropped, at least one line ("de0: interrupting at irq 5") was not
corrupted at all.  (Yes, there was a perfectly good non-corrupt set of
boot messages after the corrupted version - though I did have a bad
moment the first time I looked at dmesg.boot and saw this stuff.)

/~\ The ASCII				der Mouse
\ / Ribbon Campaign
 X  Against HTML	       mouse@rodents.montreal.qc.ca
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B