IBM giving "vm_fault: pager read error"…?



  • Have an old IBM Intellistation M Pro here with a mix of old NICs in it.  It boots to the CD, gets past the initial screen of boot options, loads most of the device driver stuff, but when it tries to mount the root partition off the CD it dies with a rapidly scrolling series of: vm_fault: pager read error
      Haven't found anything specific to pfSense on this but the FreeBSD posts found through a Google search all seem to point to different I/O errors.
      So far have only:
          - set BIOS to factory defaults then tweaked slightly as appropriate, (e.g. disable ACPI, allow boot w/out kbd etc)
          - run Memtest (no errors)
          - tried booting FreeBSD safe mode.  Which works fine!  Or at least gets as far as the first NIC prompt.

    Anybody else run into anything like this and can save me the more extended troubleshooting?  :)



  • Hrm.  Letting it sit here at the initial NIC assignment following a safe mode boot and noticing some other errors showing up semi-randomly, (including more vm_fault errors):
    acd0: FAILURE - READ_BIG MEDIUM ERROR asc=0x11 ascq=0x00
    …and:
    _  ifconfig, uid 0: exited on signal 11 (core dumped)_

    So it's really looking like some sort of hardware problem, but to the best of my knowledge this machine has been running fine as an (old) workstation.  Maybe some funkyness with the NICs?  Besides the onboard Intel chipset, it has a Kingston 10Mb(!), a D-Link, a Netgear and a no-name with SMC chipset.
      I'll keep troubleshooting and posting as I go in case somebody finds it useful down the road, (or it rings a bell and somebody knows the answer.  :)



  • Found a BIOS "compatibility mode" option for the CD-ROM which let it boot normally once, but on reset it vm_fault'ed again, and it's doing that now even in safe mode.  So I think that was a red herring.
      In other news I discovered that somebody had apparently replaced the power supply awhile back with the wrong one (missing the "aux power" 5v connector) and the CPU speed DIP switch was set wrong.  But unfortunately fixing that stuff hasn't helped.  :(



  • M'kay.  It appears to have been a combination of a flakey CDROM drive, (the vm_fault errors when mounting root), and I needed to disable PCI bus mastering, (the other random errors while it was sitting there).


Log in to reply