cp1"> boot8"> installboot8"> fdisk8"> x86/mbr8"> x86/boot8"> sun3/boot8"> macppc/boot8"> macppc/ofwboot8"> gpt8"> ] > This chapter collects various topics, in sparse order The boot manager is tied to the machine firmware, and not tied to an operating system. It loads and runs a boot loader, not necessarily the &os; boot loader. On systems with UEFI firmware, the boot manager comes with the firmware, and there is no &os; boot manager involved. On systems with OpenFirmware firmware, the boot manager is part of the firmware, and there is no &os; boot manager involved. See &man.macppcofwboot.8;. &os; provides boot managers for PC/AT and compatible systems with a "BIOS" instead of UEFI firmware. In the past, such systems had no real boot manager to speak of, although more recent systems often have menus for at least selecting which disk to boot from and in what order to attempt them. Sysinst, the &os; installation program usually installs the &os; boot manager on the hard disk. The &os; boot manager lives in unpartitioned space right in the first sector(s) of the disk. The installation programs of other operating systems, particularly if they re-partition the hard disk as part of their installation process, will usually overwrite this area with one of their own boot managers. When a disk has the GUID partition table, the &os; boot manager can be re-installed or reconfigured at a later time, if needed, with the &man.gpt.8; command. For example, assuming that your hard disk is wd0 and the &os; root partition is the partition with index 1: &rprompt; gpt biosboot -i 1 wd0 See &man.gpt.8; for other ways in which the &os; root partition can be specified. Substitute sd0 for wd0 if you have a SCSI disk, and change numbers as appropriate. This is unrelated to GUID partitions of type biosboot. Those have nothing to do with the &os; boot manager. When a disk has the "MBR"-style partition table, the &os; boot manager can be re-installed or reconfigured at a later time, if needed, with the &man.fdisk.8; command. For example: &rprompt; fdisk -B wd0 This installs the default boot manager. There are several others that can be chosen with fdisk's -c option. The choices of boot manager are listed in &man.mbr.8;. Again, substitute sd0 for wd0 if you have a SCSI disk, and change numbers as appropriate. Sometimes fdisk -B doesn't give the expected result (at least it happened to me), probably if you install/remove other operating systems. In this case, try running fdisk -i and then run again fdisk from &os;. If &os; doesn't boot from the hard disk, you can boot it from the installation disk/USB device and start the kernel on the hard disk. Insert the installation disk/USB device and, at the boot prompt, give the following command: > boot wd0a:netbsd This boots the kernel on the hard disk. (Use the correct device, for example sd0a for a SCSI disk.) Non-&os; boot managers (e.g. OS/2 Boot Manager, Partition Magic Boot Manager, Microsoft Windows Boot Manager, and the GNU GRand Unified Bootloader) are not in scope for this guide. See their own documentation for how to install and configure them. They are expected to load and run the &os; boot loader from a &os; root filesystem, in some fashion. The boot manager loads and runs the &os; boot loader. This program is generally described in &man.boot.8;, with extra architecture-specific information to be found on pages such as &man.x86boot.8;, &man.sun3boot.8;, or &man.macppcboot.8;. On systems with UEFI firmware, the UEFI Boot Manager finds and runs the &os; boot loader. Either you will have explicitly added an entry to the UEFI Boot Manager that points to a copy of the &os; boot loader, or UEFI will be looking in the well-known place that it looks for a fallback boot loader. In both cases, the &os; boot loader is a file stored on the EFI System Partition. A master copy of the &os; boot loader for UEFI can be found in /usr/mdec, and is named after the processor architecture, something like bootia32.efi, bootx64.efi, or bootaa64.efi. UEFI specifies the exact filename, and usually just the one(s) appropriate to your machine's processor architecture will exist in /usr/mdec. This file is usually copied to the /EFI/boot directory on the EFI System Partition, the place that UEFI looks for a fallback boot loader; and if it is missing from there, copy it back. For example, assuming that /dev/dk2 is your EFI System Partition: &rprompt; mount -t msdos /dev/dk2 /mnt &rprompt; cp -i /usr/mdec/boot*.efi /mnt/EFI/boot/ &rprompt; umount /mnt If, rather, you have explicitly added the &os; boot loader as an entry on the UEFI Boot Manager menu, copy the file to where that entry points. If you are multi-booting &os; and some other operating system, they cannot both use the /EFI/boot/bootarch.efi file. This is the case where you must have explicit entries in the EFI Boot Manager, and the &os; boot loader should be copied into /EFI/NetBSD on the EFI System Partition with an entry pointing at it. On non-UEFI systems, Sysinst, the &os; installation program again usually installs the &os; boot loader on the hard disk. It can also be re-installed or reconfigured at a later time, if needed. The &os; boot loader on these systems comes in two parts, a primary part and a secondary part. Each can be re-installed separately, according to need. The secondary part lives in a file (usually in the root directory) on the boot filesystem. A master copy of it is in the /usr/mdec directory. Re-install it by copying it into the root directory of whatever filesystem you are making bootable. For example, assuming that you are (re-)installing the loader on your dk3 disk wedge (and that is not your current root partition): &rprompt; mount /dev/dk3 /mnt &rprompt; cp -i /usr/mdec/boot /mnt/ &rprompt; umount /mnt The secondary part is specific to the machine type, and that is supposed to be part of its filename. Fortunately, the file can also be referred to by the generic name boot and the primary part knows to also look there. For best results, re-install the primary part after installing the secondary. It lives in a special area of the filesystem, not in any file. A master copy of it is in the /usr/mdec directory. It is installed with the &man.installboot.8; command. For example, assuming that /dev/dk3 is your root partition and that it is a FFSv2 partition: &rprompt; installboot /dev/dk3 /usr/mdec/boot_ffsv2 /boot The primary part is specific to the filesystem format, and must match it. The list of possible files to use in place of boot_ffsv2 can be found on the &man.installboot.8; page. Though this is not an operation that you need to perform frequently, it can be useful to know how to do it in case of need. Please be sure to know exactly what you are doing before performing this kind of operation. For example: &rprompt; dd if=/dev/zero of=/dev/rwd0c bs=8k count=1 The previous command deletes the disklabel (not the MBR partition table). To completely delete the disk, the whole device rwd0d must be used. For example: &rprompt; dd if=/dev/zero of=/dev/rwd0d bs=8k The commands above will only work as expected on the i386 and amd64 ports of &os;. On other ports, the whole device will end in c, not d (e.g. rwd0c). When the disk is partitioned with the GUID Partition Table (GPT), clearing the partition table is done with the &man.gpt.8; command: &rprompt; gpt destroy -r wd1 To output a sound from the speaker (for example at the end of a long script) the spkr driver can be used in the kernel config, which is mapped on /dev/speaker. For example: echo 'BPBPBPBPBP' > /dev/speaker The spkr device is not enabled in the generic kernel; a customized kernel is needed. If you forget root's password, not all is lost and you can still recover the system with the following steps: boot single user, mount / and change root's password. In detail: Boot single user: when the boot prompt appears and the five seconds countdown starts, give the following command: > boot -s At the following prompt Enter pathname of shell or RETURN for sh: press Enter. Write the following commands: &rprompt; fsck -y / &rprompt; mount -u / &rprompt; fsck -y /usr &rprompt; mount /usr Change root's password: &rprompt; passwd root Changing local password for root. New password: (not echoed) Retype new password: (not echoed) &rprompt; Exit the shell to go to multiuser mode. &rprompt; exit If you get the error Password file is busy, please see the section below. If you try to modify a password and you get the mysterious message Password file is busy, it probably means that the file /etc/ptmp has not been deleted from the system. This file is a temporary copy of the /etc/master.passwd file; check that you are not losing important information and then delete it: &rprompt; rm /etc/ptmp If the file /etc/ptmp exists you can also receive a warning message at system startup. For example: root: password file may be incorrect - /etc/ptmp exists This section describes how to add a new hard disk to an already working &os; system. In the following example a new SATA hard disk will be added. Other disk types can also be used, e.g. NVMe/M.2 by substituting wd# for ld#. When the SATA disk is physically plugged in to the SATA port, it is time to restart the computer and check that the device is correctly detected, using the dmesg command. This example kernel message is with a Seagate Barracuda drive: wd1 at atabus0 drive 1: <ST380013AS> wd1: drive supports 16-sector PIO transfers, LBA48 addressing wd1: 76293 MB, 155010 cyl, 16 head, 63 sec, 512 bytes/sect x 156250080 sectors wd1: 32-bit data port wd1: drive supports PIO mode 4, DMA mode 2, Ultra-DMA mode 6 (Ultra/133) Now the partitions can be created using the gpt command. First, check the current status of the disk: &rprompt; gpt show wd1 start size index contents 0 156250080 In this example the hard disk is blank, with no partition tables or disk labels. To create, for example, an EFI System Partition followed by a &os; partition, the commands are: &rprompt; gpt create wd1 &rprompt; gpt add -t efi -l EFI -s 1m wd1 Partition 1 added, use: dkctl wd1 addwedge <wedgename> xxx yyyyyyy <type> to create a wedge for it &rprompt; gpt add -t ffs -l netbsd-data wd1 Partition 2 added, use: dkctl wd1 addwedge <wedgename> xxx yyyyyyy <type> to create a wedge for it For simplicity, tell &os; to automatically re‐make the disk wedges for the newly defined partitions. This may not be necessary depending on the kernel configuration in use: &rprompt; dkctl wd1 makewedges Let us assume that the resultant disk wedges are dk2 and dk3. The device node names can be seen with dkctl wd1 listwedges. Most commands can also accept NAME=xxx where xxx is a GPT partition label. The final operation is the creation of the file systems for the newly defined partitions: &rprompt; newfs_msdos -F 32 -L EFISYS dk2 &rprompt; newfs -O2ea dk3 The disk is now ready for usage, and the partitions can be mounted. For example: &rprompt; mount /dev/dk3 /mnt &rprompt; mount /dev/dk2 /boot/efi In the following example a new SCSI controller and a new hard disk, connected to the controller, will be added. If you don't need to add a new controller, skip the relevant part and go to the hard disk configuration. The installation of an IDE/SATA hard disk using disklabel is identical; only the device name will be different (wd# instead of sd#). As always, before buying new hardware, consult the hardware compatibility list of &os; and make sure that the new device is supported by the system. When the SCSI controller has been physically installed in the system and the new hard disk has been connected, it's time to restart the computer and check that the device is correctly detected, using the dmesg command. This is the sample output for an NCR-875 controller: ncr0 at pci0 dev 15 function 0: ncr 53c875 fast20 wide scsi ncr0: interrupting at irq 10 ncr0: minsync=12, maxsync=137, maxoffs=16, 128 dwords burst, large dma fifo ncr0: single-ended, open drain IRQ driver, using on-chip SRAM ncr0: restart (scsi reset). scsibus0 at ncr0: 16 targets, 8 luns per target sd0(ncr0:2:0): 20.0 MB/s (50 ns, offset 15) sd0: 2063MB, 8188 cyl, 3 head, 172 sec, 512 bytes/sect x 4226725 sectors If the device doesn't appear in the output, check that it is supported by the kernel that you are using; if necessary, compile a customized kernel (see ). Now the partitions can be created using the fdisk command. First, check the current status of the disk: &rprompt; fdisk sd0 NetBSD disklabel disk geometry: cylinders: 8188 heads: 3 sectors/track: 172 (516 sectors/cylinder) BIOS disk geometry: cylinders: 524 heads: 128 sectors/track: 63 (8064 sectors/cylinder) Partition table: 0: sysid 6 (Primary 'big' DOS, 16-bit FAT (> 32MB)) start 63, size 4225473 (2063 MB), flag 0x0 beg: cylinder 0, head 1, sector 1 end: cylinder 523, head 127, sector 63 1: <UNUSED> 2: <UNUSED> 3: <UNUSED> In this example the hard disk already contains a DOS partition, which will be deleted and replaced with a native &os; partition. The command fdisk -u sd0 allows to modify interactively the partitions. The modified data will be written on the disk only before exiting and fdisk will request a confirmation before writing, so you can work relaxedly. The geometry of the disk reported by fdisk can appear confusing. Dmesg reports 4226725 sectors with 8188/3/172 for C/H/S, but 8188*3*172 gives 4225008 and not 4226725. What happens is that most modern disks don't have a fixed geometry and the number of sectors per track changes depending on the cylinder: the only interesting parameter is the number of sectors. The disk reports the C/H/S values but it's a fictitious geometry: the value 172 is the result of the total number of sectors (4226725) divided by 8188 and then by 3. To make things more confusing, the BIOS uses yet another fake geometry (C/H/S 524/128/63) which gives a total of 4225536, a value which is a better approximation to the real one than 425008. To partition the disk we will use the BIOS geometry, to maintain compatibility with other operating systems, although we will lose some sectors (4226725 - 4225536 = 1189 sectors = 594 KB). To create the BIOS partitions the command fdisk -u must be used; the result is the following: Partition table: 0: sysid 169 (NetBSD) start 63, size 4225473 (2063 MB), flag 0x0 beg: cylinder 0, head 1, sector 1 end: cylinder 523, head 127, sector 63 1: <UNUSED> 2: <UNUSED> 3: <UNUSED> Now it's time to create the disklabel for the &os; partition. The correct steps to do this are: &rprompt; disklabel sd0 > tempfile &rprompt; vi tempfile &rprompt; disklabel -R -r sd0 tempfile If you try to create the disklabel directly with &rprompt; disklabel -e sd0 you get the following message disklabel: ioctl DIOCWDINFO: No disk label on disk; use "disklabel -I" to install initial label because the disklabel does not yet exist on the disk. Now we create some disklabel partitions, editing the tempfile as already explained. The result is: # size offset fstype [fsize bsize cpg] a: 2048004 63 4.2BSD 1024 8192 16 # (Cyl. 0*- 3969*) c: 4226662 63 unused 0 0 # (Cyl. 0*- 8191*) d: 4226725 0 unused 0 0 # (Cyl. 0 - 8191*) e: 2178658 2048067 4.2BSD 1024 8192 16 # (Cyl. 3969*- 8191*) When the disklabel has been created it is possible to optimize it studying the output of the command newfs -N /dev/rsd0a, which warns about the existence of unallocated sectors at the end of a disklabel partition. The values reported by newfs can be used to adjust the sizes of the partitions with an iterative process. The final operation is the creation of the file systems for the newly defined partitions (a and e). &rprompt; newfs /dev/rsd0a &rprompt; newfs /dev/rsd0e The disk is now ready for usage, and the two partitions can be mounted. For example: &rprompt; mount /dev/sd0a /mnt If this succeeds, you may want to put an entry for the partition into /etc/fstab. First shutdown to single user, partitions still mounted rw (read-write); You can do that by just typing shutdown now while you are in multi user mode, or reboot with the -s option and make / and /dev read-writable by doing. &rprompt; mount -u / &rprompt; mount -u /dev Then: &rprompt; mkdir /newdev &rprompt; cd /newdev &rprompt; cp /dev/MAKEDEV* . &rprompt; sh ./MAKEDEV all &rprompt; cd / &rprompt; mv dev olddev &rprompt; mv newdev dev &rprompt; rm -r olddev Or if you fetched all the sources in /usr/src: &rprompt; mkdir /newdev &rprompt; cd /newdev &rprompt; cp /usr/src/etc/MAKEDEV.local . &rprompt; ( cd /usr/src/etc ; make MAKEDEV ) &rprompt; cp /usr/src/etc/obj*/MAKEDEV . &rprompt; sh ./MAKEDEV all &rprompt; cd / &rprompt; mv dev olddev; mv newdev dev &rprompt; rm -r olddev You can determine $arch by &rprompt; uname -m or &rprompt; sysctl hw.machine_arch