On Fri, May 20, 2005 at 12:19:24AM -0400, Bill Nottingham wrote:
Matt Domsch (Matt_Domsch(a)dell.com) said:
> + if [ "${DEV_IPMI}" = "1" ]; then
> + modprobe ipmi_devintf || RETVAL=2
> + if [ "${RETVAL}" != "2" ]; then
> + # Note, this really should be done by udev on 2.6
> + DEVMAJOR=`cat /proc/devices | awk '/ipmidev/{print $1}'`
> + mknod -m 0600 /dev/ipmi${INTF_NUM} c ${DEVMAJOR} 0 || RETVAL=2
> + ln -sf /dev/ipmi${INTF_NUM} /dev/ipmi || RETVAL=2
> + fi
> + fi
Why isn't the driver fixed to populate sysfs properly?
Just-in-time programming. Corey submitted class_simple support this
evening.
http://marc.theaimsgroup.com/?l=linux-kernel&m=111654581209165&
That'll handle the kernels where it exists, still need to handle the
cases where it doesn't. I'll add a test for existance of the dev
file, and fall back to creating the devnode here if it doesn't exist.
> +
> + start_watchdog
> + start_powercontrol
> + if [ "${IPMI_IMB}" = "1" ]; then
> + modprobe ipmi_imb || RETVAL=2
> + # FIXME create canonical /dev/foo entry
> + fi
> + if [ "${IPMI_RADISYS}" = "1" ]; then
> + modprobe ipmi_radisys || RETVAL=2
> + # FIXME create canonical /dev/foo entry
> + fi
> + return
How many, if any, of these modules can be detected for the
proper hardware?
You're thinking of using kudzu instead? Possible, though the data is
scattered throughout ACPI and SMBIOS tables, when it is known. That's
more than I was up for this week. Future incremental change?
Thanks,
Matt
--
Matt Domsch
Software Architect
Dell Linux Solutions
linux.dell.com &
www.dell.com/linux
Linux on Dell mailing lists @
http://lists.us.dell.com