NAME
boot —
system bootstrapping
procedures
SYNOPSIS
b |
[dev
[(cntrl,
unit, part)]]
[file]
[-adqsv] |
DESCRIPTION
Power fail and crash
recovery
Normally, the system will reboot itself at power-up or after crashes. An
automatic consistency check of the file systems will be performed as described
in
fsck(8), and unless this
fails, the system will resume multi-user operations.
Selecting the device
and kernel to boot
Normally, the
b command alone is sufficient to boot the
system, as the PROM chooses a default boot device
dev if
none is specified. The PROM chooses the first device present on the system
from the following ordered list:
sd SCSI disk
ie Intel Ethernet
ec 3Com Ethernet
Unless specified, the controller number
cntrl, unit number
unit, and partition number
part
default to zero, which is almost always correct.
The controller number can be specified if there is more than one of the given
device in the system. For example, use “ie(1,,)” to boot off of
the second Intel Ethernet in the system.
The unit number specifies one of the many devices attached to a controller. The
exact meaning and values vary depending on the device name. For example,
“sd(,18,)” boots the disk at target 6 on the first SCSI
controller, 18 being the target number 6, multiplied by 4, and given in
hexadecimal.
The partition number specifies one of the many partitions on a device. The exact
meaning and values vary depending on the device name. For example,
“sd(,18,1)” boots the second partition on the disk at target 6 on
the first SCSI controller.
The PROM only loads a first-stage boot program, currently either
/usr/mdec/bootxx (for a disk boot), or
/usr/mdec/bootyy (for a network boot). This first-stage boot
program then loads the second-stage boot program from the same device,
currently either
/usr/mdec/ufsboot (for a disk boot), or
/usr/mdec/netboot (for a network boot).
The second-stage boot program will then attempt to load the kernel named
file (or
vmunix if none is specified).
The second-stage disk boot program
/usr/mdec/ufsboot loads
the kernel from the same device that it was loaded from, while the
second-stage network boot program
/usr/mdec/netboot will
load the kernel from the NFS root as determined by the procedure described in
diskless(8).
Boot program options
-
-
- -a
- Prompt for the root file system device, the system crash
dump device, and the path to
init(8).
-
-
- -d
- Bring the system up in debug mode. Here it waits for a
kernel debugger connect; see
ddb(4).
-
-
- -q
- Boot the system in quiet mode.
-
-
- -s
- Bring the system up in single-user mode.
-
-
- -v
- Boot the system in verbose mode.
Other flags are currently ignored.
At any time you can break back to the ROM by pressing the ‘L1’ and
‘a’ keys at the same time (if the console is a serial port the
same is achieved by sending a ‘break’). If you do this
accidentally you can continue whatever was in progress by typing
‘c’ followed by the return key.
FILES
- /netbsd
- system code
- /usr/mdec/bootxx
- first-level boot block for disks
- /usr/mdec/bootyy
- first-level boot block for NFS (diskless) boot
- /usr/mdec/netboot
- boot program for NFS (diskless) boot
- /usr/mdec/ufsboot
- second-level boot program for UFS disks
- /usr/sbin/installboot
- program to install bootxx on a disk
SEE ALSO
crash(8),
disklabel(8),
fsck(8),
halt(8),
init(8),
rc(8),
shutdown(8),
syslogd(8)