The main idea is to boot a kernel, passing it all needed parameters. This is what the boot loader part of PALO does (see Section 4, “How to use PALO at early boot stage?”). Once it has been called by the firmware, it will load the Linux Kernel in memory, passing to it the given arguments, and tell the processor to branch to its entry point. This will begin the execution of the kernel on the PA-RISC computer.
The PALO management tool can transform the usual
vmlinux into a PA-RISC bootable
lifimage, including or not
RAMDISK or NFSROOT support.
However, it can also make a hard disk drive bootable, specifying
the console output and the root device.
We are going to see all these points precisely.
What must be kept in mind is that vmlinux
is the kernel alone, which is not bootable as is.
It needs PALO to be turned into a bootable
lifimage for CD or network boot,
or to be loaded at boot time from a prepared hard
disk drive. Have a look at Glossary about these words.
Quoting a well known PA/Linux hacker:
People often try to put a lifimage in /boot, or a vmlinux on the network boot server. | ||
| --Richard Hirst | ||
Which is obviously wrong.