Bugzilla – Bug 117644
lilo ignores the timeout parameter
Last modified: 2006-01-31 17:26:31 UTC
I upgraded from SuSE Linux 9.3 to 10.0 RC1. The computer boots in the following order: 1) WinXP boot manager 2) Lilo During the installation I did not configured a boot loader, but started "lilo" manually after the update and copied 512 bytes from /dev/hda6 to the windows boot loader. The problem is, that lilo ignores my option "timeout = 30" which worked all the time. Instead of continuing after 30 sec it prints "3802 s" (what does this mean) and waits indefinitely. This is my /etc/lilo.conf menu-scheme = Wb:kw:Wb:Wb timeout = 30 lba32 change-rules reset read-only prompt default = linux26131 message = /boot/message boot = /dev/hda6 image = /boot/vmlinuz label = suse93 append = "selinux=0 splash=silent resume=/dev/hda5 desktop elevator=as showopts" vga = 791 initrd = /boot/initrd root = /dev/hdb6 image = /boot/vmlinuz-2.6.13.1 label = linux26131 root = /dev/hdb6 vga = 791 append = "selinux=0 splash=silent resume=/dev/hda5 desktop elevator=as showopts" optional ...
And when I installed 10.0 rc to 9.1 (works great in the end) ,it failed to properly configure my old working LILO. It deleted the line for the TIMEOUT completely. so I used Yast to set a new delay and , still no timeout. So I said to self , looks like the install bug in 9.1 is still there ( yup saw it last year). I decided to just flip it over to Grub using yast and it works perfect. So we have two Yast bugs in the Lilo department. BTW the symptoms look like no Lilo at all because it goes by so fast ! Preventing a dual boot selection by user good day. jerry
Torsten, can you please look into this one? I'm quite busy will all the kernel stuff...
I am experiencing something similar. I have timeout=80: system1: the picture of the house by the river is shown, with "2571s" indicated below the smiling chameleon. The HDD light is on permanently, and a timeout never occurs. The number does abruptly change to "2504s", and then "2?36s" (? = looks like the numbers were written on top of each other). system2: the picture of the house by the river is shown, with "1937s" indicated below the smiling chameleon. After a while the number changes abruptly, but a timeout never occurs.
What is the behaviour after a power cycle, going to lilo directly, maybe only via the "generic MBR code" from our install? I suspect different memory contents after the boot selector.
Lilo seems to be ok now. I could not detect the timeout-bug in Lilo anymore. So, this bug is fixed in 10.1 Beta 1 (lilo package lilo-22.7-7).