Hello Dennis,
On Thu, 2025-09-04 at 21:41 -0400, Dennis Clarke wrote:
> nix# cat silo.conf
> root=/dev/sda4
> partition=1
> default=Linux
> read-only
> timeout=100
>
> image=/vmlinuz
> label=Linux
> initrd=/initrd.img
>
> image=/vmlinuz.old
> label=LinuxOLD
> initrd=/initrd.img.old
> nix#
> (...)
> ok
> ok boot /pci@1f,0/ide@d/disk@0,0:a
> Boot device: /pci@1f,0/ide@d/disk@0,0:a File and args:
> SILO Version 1.4.14
> boot:
> Allocated 64 Megs of memory at 0x40000000 for kernel
>
> Cannot find /vmlinuz (error: 4294967295)
>
> Image not found.... try again
> boot:
>
> yep .. dead in the water.
I would not recommend compressing the kernel (vmlinuZ) but leaving it
uncompressed
(vmlinuX). We started shipping uncompressed kernels on sparc64 some time ago
since
there was not enough memory for Open Firmware to uncompress the kernel.
I'm quite confident that SILO should be able to boot the uncompressed kernel
provided
that the kernel and initrd are placed in the right locations. Please keep in
mind that
SILO works with blocklists meaning that you can't place the kernel in arbitrary
locations
and reference them as files.
If I remember correctly, SILO loads the kernel and initrd relative with paths
to the boot
partition, so you need to place the kernel as "vmlinux" and the initrd as
"initrd.img"
into /boot/, then set the silo.conf entry as follows:
image=/vmlinux
label=Linux
initrd=/initrd.img
For the background: A bootloader using blocklists does not use a filesystem
driver to
access the stage2 bootloader. Instead, it remembers the position of the raw
blocks of
the stage2 bootloader on disk. This is why it's necessary not to use a fancy
filesystem
for /boot but something as simple as ext2 or ext3 and this is why the
filesystem must
not be too large, otherwise the blocks of the stage2 bootloader might beyond
what the
stage1 bootloader can access through the firmware.
Adrian
--
.''`. John Paul Adrian Glaubitz
: :' : Debian Developer
`. `' Physicist
`- GPG: 62FF 8A75 84E0 2956 9546 0006 7426 3B37 F5B5 F913
