boot
—
system bootstrapping procedures
The NetBSD kernel is started by placing it near the
beginning of physical memory and transferring to the entry point. Since the
system is not reenterable, it is necessary to read it in from disk or the
network each time it is to be bootstrapped.
FPGA devices lose their content upon loss of power and must be re-programmed.
This is usually accomplished via a JTAG cable programmer, or using a FLASH
device. Once the eMIPS processor has been programmed into the FPGA, the system
will normally boot itself. An automatic consistency check of the file systems
will be performed, and unless this fails, the system will resume multi-user
operations.
On these systems the kernel is normally read in from disk. The path used for
automatic booting is /netbsd
, but it can be overridden
in the optional file /boot.config
stored in the root
of the disk filesystem. The user can also type a different filename and a
combination of boot options.
On this system there is no local disk and booting over the network is the only
option. Network boot can also be explicitly selected by hitting
f
at the boot loader prompt.
At the boot loader prompt, the user may boot NetBSD with
a number of options that are passed to the kernel unmodified. Parsing of these
options is defined in the file /sys/boot_flag.h
Some
additional flags are as follows.
e
- Empty -- this argument requests NetBSD to ignore
all symbolic information and empty its initial symbol table.
n
- Name -- Prompt for the root file system device, the system crash dump
device, and the path to
init(8).
N
- NoName -- Do not prompt for the root file system device, the system crash
dump device, and the path to
init(8). If the
configured-in devices are present, use them.
s
- Single -- Boot only to single-user mode.
The boot
command is currently under development.