Red Hat Bugzilla – Bug 43324
Please update to latest LILO, RH7.1 version doesn't handle many IDE controllers properly
Last modified: 2007-04-18 12:33:33 EDT
From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux 2.4.2-2 i686; en-US; rv:0.9+)
Description of problem:
I have a Soyo IV 440BX motherboard. It has onboard HighPoint 366 Ultra66
controler. I have also added a Promise Ultra100 PCI controller. The
version of LILO included with RH7.1 doesn't work with my hard drives
plugged into the Promise controller. I upgraded to the latest official
LILO 21.7.5 and everything works fine.
This is what the kernel sees as far as IDE controller hardware (notes on
the side are mine). Everything is correct:
/dev/hda - Standard 440BX Ultra33 controller (master)
/dev/hdb - ... (slave)
/dev/hdc - Standard 440BX Ultra33 controller (master)
/dev/hdd - ... (slave)
/dev/hde - HighPoint 366 Ultra66 controller (master)
/dev/hdf - ... (slave)
/dev/hdg - HighPoint 366 Ultra66 conroller (master)
/dev/hdh - ... (slave)
/dev/hdi - Promise Ultra100 controller (master)
/dev/hdj - ... (slave)
/dev/hdk - Promise Ultra100 conroller (master)
/dev/hdl - ... (slave)
Even though I have all that IDE controller hardware, I only have 3
hdc: TOSHIBA DVD-ROM SD-M1502, ATAPI CD/DVD-ROM drive
hdi: IBM-DTLA-307030, ATA DISK drive
hdk: IC35L040AVER07-0, ATA DISK drive
hdi: 60036480 sectors (30739 MB) w/1916KiB Cache, CHS=59560/16/63, UDMA(100)
hdk: 80418240 sectors (41174 MB) w/1916KiB Cache, CHS=79780/16/63, UDMA(100)
I have Windows 2000 installed on "hdi" and Red Hat 7.1 installed on "hdk".
/dev/hdk3 on / type ext2 (rw)
/dev/hdk1 on /boot type ext2 (rw)
menu-title=" Choose an OS "
Note: The "disk=" lines are required with both the RH7.1 version of lilo
and the latest lilo, 21.7.5. Their view of drives is swapped, and I get
"LI" on boot.
With the LILO that comes with Red Hat 7.1, when I run "lilo -v -v -v" I get
"Sorry, don't know how to handle device XXXX", where XXXX is the
major/minor number for /dev/hdi1.
I upgraded to the latest version of lilo, 21.7.5, the error message went
away (still using exact same lilo.conf), and everything works as expected.
Steps to Reproduce:
1. Create computer with same controller and drive layout as what I have.
Please upgrade the urgency of this bug.
In addition to the situation this user described, this bites us when we attempt
to create IDE RAID configurations where the /boot partition is mirrored across
multiple drives which include /dev/hdi. LILO correctly detects that /boot is
mirrored and attempts to update the boot blocks on all drives in the mirror.
When it gets to /dev/hdi you get the message:
Sorry, don't know how to handle device 0x3801.
If you have re-arranged the drives and you have the unfortunate situation of hdi
being the FIRST device in the mirror (slot 0 in the raid superblock), then LILO
won't update *any* of the superblocks.
*** This bug has been marked as a duplicate of 39365 ***