tim at home work and in between.

things I didn't know but should have..

Can't get the terminal size right when using the console of a machine?

---

eval `/usr/openwin/bin/resize`
TERM=xterm ; export TERM

---

Your Solaris 10 machine crashes with a short panic string of "pci fatal error", afterwards you have a crash dump and no sign of any FMA ereports that look relevant from fmdump.. So how do you find out what the hardware reported that upset the system so much it panic'ed....

---

 ksh: mdb unix.0 vmcore.0
Loading modules: [ krtld genunix specfs dtrace ufs scsi_vhci sd mpt px ip sctp arp usba fcp fctl nca lofs audiosup md cpc random crypto zfs fcip logindmux ptm sppp nfs ipc ]
> ::walk ereportq_dump | ::ereport -v 
class='ereport.io.fire.pec.mfp'
ena=13e4662751600401
detector
    version=00
    scheme='dev'
    device-path='/pci@1e,600000'
primary=1
tlu-uele=00000000001fffff
tlu-uie=001fffff001fffff
tlu-uis=0000000000040000
tlu-uess=0000000000040000
tlu-rueh1l=0000000000000000
tlu-rueh2l=000000007e4151da
tlu-tueh1l=000000010000000f
tlu-tueh2l=0080004400000000

class='ereport.io.fire.pec.mfp'
ena=13e466e707800401
detector
    version=00
    scheme='dev'
    device-path='/pci@1e,600000'
primary=1
tlu-uele=00000000001fffff             
tlu-uie=001fffff001fffff
tlu-uis=0004000000040000
tlu-uess=0004000000040000
tlu-rueh1l=0000000000000000
tlu-rueh2l=000000002db088b7
tlu-tueh1l=0200000100000f0c
tlu-tueh2l=0000104400000000

---

I always used a much longer way of doing it...