8.8. 配置引导加载器（Bootloader）

It is probably already functional, but it is always good to know how to configure and install the bootloader in case it “disappears“. This can occur after an installation or an upgrade of another operating system, such as Windows. The following information can also help you to modify the bootloader configuration if needed.

8.8.1. 辨识磁盘

文化 udev 与 /dev/

/dev/ 文件夹通常保存被称为 “特殊的” 文件，用于代表系统的周边设备 (见专栏 基本知识 设备访问权限)。曾经纳入所有可能用到的特殊文件。这种处理方式有若干缺点，限制设备数量是其中之一，每种设备只能有一个 (因为清单名称被限制)，且不可能知道实际上该用哪个特殊文件。

如今，以动态方式管理特殊文件，并且更适合热切换电脑设备。内核与 udev (第 9.11.3 节 “udev 如何工作”) 合作在需要的时候添加与删除文件以映射出现或消失的设备。所以，/dev/ 不需要维持不变，反而采用 RAM 保存的文件系统清除或纳入相关的条目。

The kernel communicates lots of information about any newly added device and hands out a pair of major/minor numbers to identify it. With this udevd can create the special file under the name and with the permissions that it wants. It can also create aliases and perform additional actions (such as initialization or registration tasks). udevd's behavior is driven by a large set of (customizable) rules usually found in /lib/udev/rules.d/ and /etc/udev/rules.d/.

动态给予名称，就能够让相同的名称永远指向特定设备，不必在意链接的设备或顺序，尤其在使用 USB 周边时更为便利。第一个磁盘的第一个分区称为 /dev/sda1 可供向后兼容，喜欢的话也可称为 /dev/root-partition，或两者共用，因为 udevd 可以配置成自动添加符号链接。

In ancient times, some kernel modules did automatically load when you tried to access the corresponding device file. This is no longer the case, and the peripheral's special file no longer exists prior to loading the module; this is no big deal, since most modules are loaded on boot thanks to automatic hardware detection. But for undetectable peripherals (such as very old disk drives or PS/2 mice), this doesn't work. Consider adding the modules, floppy, psmouse and mousedev to /etc/modules or /etc/modules-load.d/ in order to force loading them on boot.

配置引导加载器必须辨认硬盘及其分区。Linux 使用 “block” 特殊文件保存在 /dev/ 文件夹内。由 Debian Squeeze 版本开始使用源自 Linux 内核的硬盘命名架构，所有的硬盘 (IDE/PATA、SATA、SCSI、USB、IEEE 1394) 都以 /dev/sd* 表示。

每个分区由其磁盘代号表示：如，/dev/sda1 是第一个磁盘的第一个分区，而 /dev/sdb3 是第二个磁盘的第三个分区。

PC 架构 (或 “i386”，包括新出现的 “amd64”) 不再受限于使用 “MS-DOS” 分区表格式，每个磁盘可以拥有超过 4 个 “主要” 分区。在此架构下要超过此限制，必须添加 “扩展” 分区，然后就能使用添加的 “次要” 分区。这些次要分区由 5 开始编号。所以，第一个磁盘的次要分区可以是 /dev/sda5，然后是 /dev/sda6，余此类推。

Another restriction of the MS-DOS partition table format is that it only allows disks up to 2 TiB in size, which was becoming a real problem with recent disks.

A new partition table format called GPT (GUID Partition Table) loosens these constraints on the number of partitions (it allows up to 128 partitions when using standard settings) and on the size of the disks (up to 8 ZiB, which is more than 8 billion terabytes). If you intend to create many physical partitions on the same disk, you should therefore ensure that you are creating the partition table in the GPT format when partitioning your disk.

QUICK LOOK The Globally Unique Identifier (GUID)

Using GPT the partition type is represented as a 16 bytes long universally unique identifier, also known as Globally Unique Identifier (GUID), which is part of the UEFI standard (see 笔记 UEFI，一个 BIOS 的现代替代品). Compared to the hex code used to determine the partition type in the MS-DOS table format, you will have a hard time trying to remember these identifiers. Fortunately there are lists with all the IDs which you can use to create partition recipes to use with sfdisk or for preseeding the Debian installer.

→ https://en.wikipedia.org/wiki/GUID_Partition_Table#Partition_type_GUIDs

It is not always easy to remember what disk is connected to which SATA controller, or in third position in the SCSI chain, especially since the naming of hotplugged hard drives (which includes among others most SATA disks and external disks) can change from one boot to another. Fortunately, udev creates, in addition to /dev/sd*, symbolic links with a fixed name, which you could then use if you wished to identify a hard drive in a non-ambiguous manner. These symbolic links are stored in /dev/disk/by-id/. On a machine with two physical disks, for example, one could find the following:

mirexpress:/dev/disk/by-id# ls -l
total 0
lrwxrwxrwx 1 root root  9 23 jul. 08:58 ata-STM3500418AS_9VM3L3KP -> ../../sda
lrwxrwxrwx 1 root root 10 23 jul. 08:58 ata-STM3500418AS_9VM3L3KP-part1 -> ../../sda1
lrwxrwxrwx 1 root root 10 23 jul. 08:58 ata-STM3500418AS_9VM3L3KP-part2 -> ../../sda2
[...]
lrwxrwxrwx 1 root root  9 23 jul. 08:58 ata-WDC_WD5001AALS-00L3B2_WD-WCAT00241697 -> ../../sdb
lrwxrwxrwx 1 root root 10 23 jul. 08:58 ata-WDC_WD5001AALS-00L3B2_WD-WCAT00241697-part1 -> ../../sdb1
lrwxrwxrwx 1 root root 10 23 jul. 08:58 ata-WDC_WD5001AALS-00L3B2_WD-WCAT00241697-part2 -> ../../sdb2
[...]
lrwxrwxrwx 1 root root  9 23 jul. 08:58 scsi-SATA_STM3500418AS_9VM3L3KP -> ../../sda
lrwxrwxrwx 1 root root 10 23 jul. 08:58 scsi-SATA_STM3500418AS_9VM3L3KP-part1 -> ../../sda1
lrwxrwxrwx 1 root root 10 23 jul. 08:58 scsi-SATA_STM3500418AS_9VM3L3KP-part2 -> ../../sda2
[...]
lrwxrwxrwx 1 root root  9 23 jul. 08:58 scsi-SATA_WDC_WD5001AALS-_WD-WCAT00241697 -> ../../sdb
lrwxrwxrwx 1 root root 10 23 jul. 08:58 scsi-SATA_WDC_WD5001AALS-_WD-WCAT00241697-part1 -> ../../sdb1
lrwxrwxrwx 1 root root 10 23 jul. 08:58 scsi-SATA_WDC_WD5001AALS-_WD-WCAT00241697-part2 -> ../../sdb2
[...]
lrwxrwxrwx 1 root root  9 23 jul. 16:48 usb-LaCie_iamaKey_3ed00e26ccc11a-0:0 -> ../../sdc
lrwxrwxrwx 1 root root 10 23 jul. 16:48 usb-LaCie_iamaKey_3ed00e26ccc11a-0:0-part1 -> ../../sdc1
lrwxrwxrwx 1 root root 10 23 jul. 16:48 usb-LaCie_iamaKey_3ed00e26ccc11a-0:0-part2 -> ../../sdc2
[...]
lrwxrwxrwx 1 root root  9 23 jul. 08:58 wwn-0x5000c50015c4842f -> ../../sda
lrwxrwxrwx 1 root root 10 23 jul. 08:58 wwn-0x5000c50015c4842f-part1 -> ../../sda1
[...]
mirexpress:/dev/disk/by-id#

Note that some disks are listed several times (because they behave simultaneously as ATA disks and SCSI disks), but the relevant information is mainly in the model and serial numbers of the disks, from which you can find the peripheral file. While the links in /dev/disk/by-id/ are created using the device' serial number and physical path, there are more convenience links in e.g. /dev/disk/by-label/ (based on given labels), /dev/disk/by-uuid/ (based on unique identifiers, which can change when reformatting a device using mkfs.* or mkswap), /dev/disk/by-path/ (based on shortest physical path), and /dev/disk/by-partlabel/ and /dev/disk/by-partuuid/ (only partitions with GPT labels and their unique identifiers). If you use these links, e.g. in /etc/fstab, always prefer unique identifiers over labels. You can also obtain and change this information for each partition or device using the lsblk and blkid commands.

下面的配置根据同样的设置而来：一个 SATA 磁盘，第一个区分安装了旧的Windows系统、第二个分区安装了 Debian GNU/Linux。

ALTERNATIVE The LILO bootloader

LILO (Linux 加载器，LInux LOader) 是最旧的引导加载器 — 稳定但粗糙。把启动内核的物理地址写入到MBR中，所以每次更新 LILO (或其配置文件) 都必须再运行命令 lilo。忘记的话，移除或更新内核后，新的内核就不在磁盘的原来位置，以致系统无法启动。

LILO 的配置档是 /etc/lilo.conf；一个简单的标准配置文件示例如下。

例 8.4. LILO 配置文件

# The disk on which LILO should be installed.
# By indicating the disk and not a partition.
# you order LILO to be installed on the MBR.
boot=/dev/sda
# the partition that contains Debian
root=/dev/sda2
# the item to be loaded by default
default=Linux

# the most recent kernel image
image=/vmlinuz
  label=Linux
  initrd=/initrd.img
  read-only

# Old kernel (if the newly installed kernel doesn't boot)
image=/vmlinuz.old
  label=LinuxOLD
  initrd=/initrd.img.old
  read-only
  optional

# only for Linux/Windows dual boot
other=/dev/sda1
  label=Windows

As per request of the maintainer and upstream author of lilo, the bootloader has been removed in Debian 11 Bullseye and it is unlikely to make a comeback.

8.8.2. GRUB 2 配置

GRUB (GRand Unified Bootloader) is more recent. It is not necessary to invoke it after each update of the kernel; GRUB knows how to read the filesystems and find the position of the kernel on the disk by itself. To install it on the MBR of the first disk, simply type grub-install /dev/sda. This will overwrite the MBR, so be careful not to overwrite the wrong location. While it is also possible to install GRUB into a partition boot record, beware that it is usually a mistake and doing grub-install /dev/sda1 has not the same meaning as grub-install /dev/sda.

说明 GRUB 的磁盘名称

GRUB can only identify hard drives based on information provided by the BIOS. (hd0) corresponds to the first disk thus detected, (hd1) the second, etc. In most cases, this order corresponds exactly to the usual order of disks under Linux, but problems can occur when you associate SCSI and IDE disks. GRUB used to store the correspondences that it detected in the file /boot/grub/device.map. Nowadays, it avoids this problem using universally unique identifier (UUIDs) or file system labels when generating grub.cfg. However, the device map file is not obsolete yet, since it can be used to override when the current environment is different from the one on boot. If you find errors there (because you know that your BIOS detects drives in a different order), correct them manually and run grub-install again. grub-mkdevicemap can help creating a device.map file from which to start.

分区表在 GRUB 内也有特定的名称。通常使用 MS-DOS 格式的 “经典” 分区，第一个磁盘的第一个分区标记为 (hd0,msdos1)、第二个分区标记为 (hd0,msdos2)，余此类推。

GRUB 2 configuration is stored in /boot/grub/grub.cfg, but this file (in Debian) is generated from others. Be careful not to modify it by hand, since such local modifications will be lost the next time update-grub is run (which may occur upon update of various packages). The most common modifications of the /boot/grub/grub.cfg file (to add command line parameters to the kernel or change the duration that the menu is displayed, for example) are made through the variables in /etc/default/grub. To add entries to the menu, you can either create a /boot/grub/custom.cfg file or modify the /etc/grub.d/40_custom file. For more complex configurations, you can modify other files in /etc/grub.d, or add to them; these scripts should return configuration snippets, possibly by making use of external programs. These scripts are the ones that will update the list of kernels to boot: 10_linux takes into consideration the installed Linux kernels; 20_linux_xen takes into account Xen virtual systems, and 30_os-prober adds other existing operating systems (Windows, OS X, Hurd), kernel images, and BIOS/EFI access options to the menu.

CAUTION error: symbol `grub_*' not found

One of the most often reported issues when using GRUB is that users get an error like error symbol `grub_calloc' not found and they are unable to boot the system anymore. Most of the time this error is caused by installing an updated version of GRUB that causes new modules in /boot/grub to be incompatible with old core images in the boot sector that your firmware jumps to when booting your machine. This happens on systems that are configured to run grub-install to a target device that is not actually the one that the firmware uses to boot your system (e.g. after replacing disks or moving them around - one can see the issue when running dpkg-reconfigure grub-pc showing the wrong target device). With the release of Debian 11 Bullseye (the change was also populated to Buster) the update process now checks for this issue and exits with an error in case it finds such a constellation.

→ https://bugs.debian.org/bug=966575#95

Setting up grub-pc (2.02+dfsg1-20+deb10u4) ...
/dev/disk/by-id/[..] does not exist, so cannot grub-install to it!
You must correct your GRUB install devices before proceeding:

  DEBIAN_FRONTEND=dialog dpkg --configure grub-pc
  dpkg --configure -a
dpkg: error processing package grub-pc (--configure):
 installed grub-pc package post-installation script subprocess returned error exit status 1

To fix the issue and continue the upgrade procedure, just follow the advice and run the commands given by the error message as root.

8.8.3. Using GRUB with EFI and Secure Boot

Using GRUB to boot either a traditional BIOS system (legacy or UEFI-CSM) or a UEFI system is quite different. Fortunately the user doesn't need to know the differences because Debian provides different packages for each purpose and the installer automatically cares about which one(s) to choose. The grub-pc package is chosen for legacy systems, where GRUB is installed into the MBR, while UEFI systems require grub-efi-arch, where GRUB is installed into the EFI System Partition (ESP). The latter requires a GTP partition table as well as an EFI partition.

→ https://wiki.debian.org/Grub2#UEFI_vs_BIOS_boot

To switch an existing system (supporting UEFI) from legacy to UEFI boot mode not only requires to switch the GRUB packages on the system, but also to adjust the partition table and the to create an EFI partition (probably including resizing existing partitions to create the necessary free space). It is therefore quite an elaborate process and we cannot cover it here. Fortunately, there are some manuals by bloggers describing the necessary procedures.

If you are using a system with “Secure Boot“ enabled and have installed shim-signed (see sidebar 文化安全引导和 shim 引导程序), you must also install grub-efi-arch-signed. This package is not pulled in automatically, only if the installation of recommended package has been enabled.