boot — system bootstrapping procedures
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, and unless this fails, the system will resume multi-user operations.
Cold starts. Most i386 PCs attempt to boot first from floppy disk drive 0 (sometimes known as drive A:) and, failing that, from hard disk drive 0 (sometimes known as drive C:, or as drive 0x80 to the BIOS). Some BIOSes allow you to change this default sequence, and may also include a CD-ROM drive as a boot device.
By default, a three-stage bootstrap is employed, and control is automatically passed from the boot blocks (bootstrap stages one and two) to a separate third-stage bootstrap program, loader(8). This third stage provides more sophisticated control over the booting process than it is possible to achieve in the boot blocks, which are constrained by occupying limited fixed space on a given disk or slice.
However, it is possible to dispense with the third stage altogether, either by specifying a kernel name in the boot block parameter file, /boot.config, or, unless option −n is set, by hitting a key during a brief pause (while one of the characters -, \, |, or / is displayed) before loader(8) is invoked. Booting will also be attempted at stage two, if the third stage cannot be loaded.
Make note of the fact that /boot.config is read only from the ’a’ partition. As a result, slices which are missing an ’a’ parition require user intervention during the boot process.
The remainder of this subsection deals only with the boot blocks. The loader(8) program is documented separately.
After the boot blocks have been loaded, you should see a prompt similar to the following:
The automatic boot will attempt to load /boot/loader from partition ’a’ of either the floppy or the hard disk. This boot may be aborted by typing any character on the keyboard at the ’boot:’ prompt. At this time, the following input will be accepted:
Give a short listing of the files in the root directory of the default boot device, as a hint about available boot files. (A ? may also be specified as the last segment of a path, in which case the listing will be of the relevant subdirectory.)
slice,]part) filename [−aCcDdghmnPprsv]
Specify boot file and flags.
The drive number as recognized by the BIOS. 0 for the first drive, 1 for the second drive, etc.
The type of controller to boot from. Note that the controller is required to have BIOS support since the BIOS services are used to load the boot file image.
The supported interfaces are:
ST506, IDE, ESDI, RLL disks on a WD100 or lookalike controller
5 1/4" or 3 1/2" High density floppies
SCSI disk on any supported SCSI controller
The unit number of the drive on the interface being used. 0 for the first drive, 1 for the second drive, etc.
The partition letter inside the BSD portion of the disk. See bsdlabel(8). By convention, only partition ’a’ contains a bootable image. If sliced disks are used (’’fdisk partitions’’), any slice (1 for the first slice, 2 for the second slice, etc.) can be booted from, with the default (if not specified) being the active slice or, otherwise, the first FreeBSD slice. If slice is specified as 0, the first FreeBSD slice (also known as ’’compatibility’’ slice) is booted from.
The pathname of the file to boot (relative to the root directory on the specified partition). Defaults to /kernel. Symbolic links are not supported (hard links are).
during kernel initialization, ask for the device to mount as the root file system.
boot from CDROM.
run UserConfig to modify hardware parameters for the loaded kernel. If the kernel was built with one of USERCONFIG, INTRO_USERCONFIG, VISUAL_USERCONFIG options, remain in UserConfig regardless of any quit commands present in the script.
toggle single and dual console configurations. In the single configuration the console will be either the internal display or the serial port, depending on the state of the −h option below. In the dual console configuration, both the internal display and the serial port will become the console at the same time, regardless of the state of the −h option. However, the dual console configuration takes effect only during the boot prompt. Once the kernel is loaded, the console specified by the −h option becomes the only console.
enter the DDB kernel debugger (see ddb(4)) as early as possible in kernel initialization.
use the GDB remote debugging protocol.
toggle internal and serial consoles. You can use this to switch console devices. For instance, if you boot from the internal console, you can use the −h option to force the kernel to use the serial port as its console device. Alternatively, if you boot from the serial port, you can use this option to force the kernel to use the internal display as the console instead. The serial port driver sio(4) has a flag to override this option. If that flag is set, the serial port will always be used as the console, regardless of the −h option described here. See the man page for sio(4) for more details.
mute the console.
ignore key press to interrupt boot before loader(8) is invoked.
probe the keyboard. If no keyboard is found, the −D and −h options are automatically set.
pause after each attached device during the device probing phase.
use the statically configured default for the device containing the root file system (see config(8)). Normally, the root file system is on the device that the kernel was loaded from.
boot into single-user mode; if the console is marked as ’’insecure’’ (see ttys(5)), the root password must be entered.
be verbose during device probing (and later).
You may put a BIOS drive number, a controller type, a unit number, a partition, a kernel file name, and any valid option in /boot.config to set defaults. Enter them in one line just as you type at the ’boot:’ prompt.
parameters for the boot blocks (optional)
first stage bootstrap file
second stage bootstrap file
third stage bootstrap
typical non-default kernel (optional)
When disk-related errors occur, these are reported by the second-stage bootstrap using the same error codes returned by the BIOS, for example ’’Disk error 0x1 (lba=0x12345678)’’. Here is a partial list of these error codes:
Address mark not found
Sector not found
DMA attempt across 64K boundary
Uncorrectable CRC/ECC error
NOTE: On older machines, or otherwise where EDD support (disk packet interface support) is not available, all boot-related files and structures (including the kernel) that need to be accessed during the boot phase must reside on the disk at or below cylinder 1023 (as the BIOS understands the geometry). When a ’’Disk error 0x1’’ is reported by the second-stage bootstrap, it generally means that this requirement has not been adhered to.
The bsdlabel(5) format used by this version of BSD is quite different from that of other architectures.
Due to space constraints, the keyboard probe initiated by the −P option is simply a test that the BIOS has detected an ’’extended’’ keyboard. If an ’’XT/AT’’ keyboard (with no F11 and F12 keys, etc.) is attached, the probe will fail.
BSD September 23, 2004 BSD