Linux kernel file directory

Kernel/Traditional compilation

This article is an introduction to building custom kernels from kernel.org sources. This method of compiling kernels is the traditional method common to all distributions. It can be, depending on your background, more complicated than using the Kernels/Arch Build System. Consider the Arch Build System tools are developed and maintained to make repeatable compilation tasks efficient and safe.

Contents

Preparation

It is not necessary (or recommended) to use the root account or root privileges (i.e. via Sudo) for kernel preparation.

Install the core packages

Install the base-devel package group, which contains necessary packages such as make and gcc . It is also recommended to install the following packages, as listed in the default Arch kernel PKGBUILD: xmlto , kmod , inetutils , bc , libelf , git , cpio , perl , tar , xz .

Create a kernel compilation directory

It is recommended to create a separate build directory for your kernel(s). In this example, the directory kernelbuild will be created in the home directory:

Download the kernel source

Download the kernel source from https://www.kernel.org. This should be the tarball ( tar.xz ) file for your chosen kernel.

It can be downloaded by simply right-clicking the tar.xz link in your browser and selecting Save Link As. , or any other number of ways via alternative graphical or command-line tools that utilise HTTP, TFTP, Rsync, or Git.

In the following command-line example, wget has been installed and is used inside the

/kernelbuild directory to obtain kernel 4.8.6:

You should also verify the correctness of the download before trusting it. First grab the signature, then use that to grab the fingerprint of the signing key, then use the fingerprint to obtain the actual signing key:

Note the signature was generated for the tar archive (i.e. extension .tar ), not the compressed .tar.xz file that you have downloaded. You need to decompress the latter without untarring it. Verify that you have xz installed, then you can proceed like so:

Do not proceed if this does not result in output that includes the string «Good signature».

If wget was not used inside the build directory, it will be necessary to move the tarball into it, e.g.

Unpack the kernel source

Within the build directory, unpack the kernel tarball:

To finalise the preparation, ensure that the kernel tree is absolutely clean; do not rely on the source tree being clean after unpacking. To do so, first change into the new kernel source directory created, and then run the make mrproper command:

Kernel configuration

This is the most crucial step in customizing the default kernel to reflect your computer’s precise specifications. Kernel configuration is set in its .config file, which includes the use of Kernel modules. By setting the options in .config properly, your kernel and computer will function most efficiently.

You can do a mixture of two things:

• Use the default Arch settings from an official kernel (recommended)
• Manually configure the kernel options (optional, advanced and not recommended)

Default Arch configuration

This method will create a .config file for the custom kernel using the default Arch kernel settings. If a stock Arch kernel is running, you can use the following command inside the custom kernel source directory:

Otherwise, the default configuration can be found online in the official Arch Linux kernel package.

Advanced configuration

There are several tools available to fine-tune the kernel configuration, which provide an alternative to otherwise spending hours manually configuring each and every one of the options available during compilation.

Those tools are:

• make menuconfig : Command-line ncurses interface superseded by nconfig
• make nconfig : Newer ncurses interface for the command-line
• make xconfig : User-friendly graphical interface that requires packagekit-qt5 to be installed as a dependency. This is the recommended method — especially for less experienced users — as it is easier to navigate, and information about each option is also displayed.
• make gconfig : Graphical configuration similar to xconfig but using gtk. This requires gtk2 , glib2 and libgladeAUR .

The chosen method should be run inside the kernel source directory, and all will either create a new .config file, or overwrite an existing one where present. All optional configurations will be automatically enabled, although any newer configuration options (i.e. with an older kernel .config ) may not be automatically selected.

Once the changes have been made save the .config file. It is a good idea to make a backup copy outside the source directory. You may need to do this multiple times before you get all the options right.

If unsure, only change a few options between compilations. If you cannot boot your newly built kernel, see the list of necessary config items here.

Running lspci -k # from liveCD lists names of kernel modules in use. Most importantly, you must maintain cgroups support. This is necessary for systemd. For more detailed information, see Gentoo:Kernel/Gentoo Kernel Configuration Guide and Gentoo:Intel#Kernel or Gentoo:Ryzen#Kernel for Intel or AMD Ryzen processors.

Compilation

Compilation time will vary from as little as fifteen minutes to over an hour, depending on your kernel configuration and processor capability. Once the .config file has been set for the custom kernel, within the source directory run the following command to compile:

Installation

Install the modules

Once the kernel has been compiled, the modules for it must follow. First build the modules:

Then install the modules. As root or with root privileges, run the following command to do so:

This will copy the compiled modules into /lib/modules/ — . For example, for kernel version 4.8 installed above, they would be copied to /lib/modules/4.8.6-ARCH . This keeps the modules for individual kernels used separated.

Copy the kernel to /boot directory

The kernel compilation process will generate a compressed bzImage (big zImage) of that kernel, which must be copied to the /boot directory and renamed in the process. Provided the name is prefixed with vmlinuz- , you may name the kernel as you wish. In the examples below, the installed and compiled 4.8 kernel has been copied over and renamed to vmlinuz-linux48 :

Make initial RAM disk

If you do not know what making an initial RAM disk is, see Initramfs on Wikipedia and mkinitcpio.

Automated preset method

An existing mkinitcpio preset can be copied and modified so that the custom kernel initramfs images can be generated in the same way as for an official kernel. This is useful where intending to recompile the kernel (e.g. where updated). In the example below, the preset file for the stock Arch kernel will be copied and modified for kernel 4.8, installed above.

First, copy the existing preset file, renaming it to match the name of the custom kernel specified as a suffix to /boot/vmlinuz- when copying the bzImage (in this case, linux48 ):

Second, edit the file and amend for the custom kernel. Note (again) that the ALL_kver= parameter also matches the name of the custom kernel specified when copying the bzImage :

Finally, generate the initramfs images for the custom kernel in the same way as for an official kernel:

Manual method

Rather than use a preset file, mkinitcpio can also be used to generate an initramfs file manually. The syntax of the command is:

• -k ( —kernel ): Specifies the modules to use when generating the initramfs image. The name will be the same as the name of the custom kernel source directory (and the modules directory for it, located in /usr/lib/modules/ ).
• -g ( —generate ): Specifies the name of the initramfs file to generate in the /boot directory. Again, using the naming convention mentioned above is recommended.

For example, the command for the 4.8 custom kernel installed above would be:

Copy System.map

The System.map file is not required for booting Linux. It is a type of «phone directory» list of functions in a particular build of a kernel. The System.map contains a list of kernel symbols (i.e function names, variable names etc) and their corresponding addresses. This «symbol-name to address mapping» is used by:

• Some processes like klogd, ksymoops, etc.
• By OOPS handler when information has to be dumped to the screen during a kernel crash (i.e info like in which function it has crashed).

If your /boot is on a filesystem which supports symlinks (i.e., not FAT32), copy System.map to /boot , appending your kernel’s name to the destination file. Then create a symlink from /boot/System.map to point to /boot/System.map- :

After completing all steps above, you should have the following 3 files and 1 soft symlink in your /boot directory along with any other previously existing files:

• Kernel: vmlinuz-
• Initramfs: Initramfs- .img
• System Map: System.map-
• System Map kernel symlink

Bootloader configuration

Add an entry for your new kernel in your bootloader’s configuration file. See Arch boot process#Feature comparison for possible boot loaders, their wiki articles and other information.

Источник

Directory Entries¶

In an ext4 filesystem, a directory is more or less a flat file that maps an arbitrary byte string (usually ASCII) to an inode number on the filesystem. There can be many directory entries across the filesystem that reference the same inode number–these are known as hard links, and that is why hard links cannot reference files on other filesystems. As such, directory entries are found by reading the data block(s) associated with a directory file for the particular directory entry that is desired.

Linear (Classic) Directories¶

By default, each directory lists its entries in an “almost-linear” array. I write “almost” because it’s not a linear array in the memory sense because directory entries are not split across filesystem blocks. Therefore, it is more accurate to say that a directory is a series of data blocks and that each block contains a linear array of directory entries. The end of each per-block array is signified by reaching the end of the block; the last entry in the block has a record length that takes it all the way to the end of the block. The end of the entire directory is of course signified by reaching the end of the file. Unused directory entries are signified by inode = 0. By default the filesystem uses struct ext4_dir_entry_2 for directory entries unless the “filetype” feature flag is not set, in which case it uses struct ext4_dir_entry .

The original directory entry format is struct ext4_dir_entry , which is at most 263 bytes long, though on disk you’ll need to reference dirent.rec_len to know for sure.

Number of the inode that this directory entry points to.

Length of this directory entry. Must be a multiple of 4.

Length of the file name.

Since file names cannot be longer than 255 bytes, the new directory entry format shortens the name_len field and uses the space for a file type flag, probably to avoid having to load every inode during directory tree traversal. This format is ext4_dir_entry_2 , which is at most 263 bytes long, though on disk you’ll need to reference dirent.rec_len to know for sure.

Number of the inode that this directory entry points to.

Length of this directory entry.

Length of the file name.

File type code, see ftype table below.

The directory file type is one of the following values:

Character device file.

Block device file.

To support directories that are both encrypted and casefolded directories, we must also include hash information in the directory entry. We append ext4_extended_dir_entry_2 to ext4_dir_entry_2 except for the entries for dot and dotdot, which are kept the same. The structure follows immediately after name and is included in the size listed by rec_len If a directory entry uses this extension, it may be up to 271 bytes.

The hash of the directory name

The minor hash of the directory name

In order to add checksums to these classic directory blocks, a phony struct ext4_dir_entry is placed at the end of each leaf block to hold the checksum. The directory entry is 12 bytes long. The inode number and name_len fields are set to zero to fool old software into ignoring an apparently empty directory entry, and the checksum is stored in the place where the name normally goes. The structure is struct ext4_dir_entry_tail :

Inode number, which must be zero.

Length of this directory entry, which must be 12.

Length of the file name, which must be zero.

File type, which must be 0xDE.

Directory leaf block checksum.

The leaf directory block checksum is calculated against the FS UUID, the directory’s inode number, the directory’s inode generation number, and the entire directory entry block up to (but not including) the fake directory entry.

Hash Tree Directories¶

A linear array of directory entries isn’t great for performance, so a new feature was added to ext3 to provide a faster (but peculiar) balanced tree keyed off a hash of the directory entry name. If the EXT4_INDEX_FL (0x1000) flag is set in the inode, this directory uses a hashed btree (htree) to organize and find directory entries. For backwards read-only compatibility with ext2, this tree is actually hidden inside the directory file, masquerading as “empty” directory data blocks! It was stated previously that the end of the linear directory entry table was signified with an entry pointing to inode 0; this is (ab)used to fool the old linear-scan algorithm into thinking that the rest of the directory block is empty so that it moves on.

The root of the tree always lives in the first data block of the directory. By ext2 custom, the ‘.’ and ‘..’ entries must appear at the beginning of this first block, so they are put here as two struct ext4_dir_entry_2 s and not stored in the tree. The rest of the root node contains metadata about the tree and finally a hash->block map to find nodes that are lower in the htree. If dx_root.info.indirect_levels is non-zero then the htree has two levels; the data block pointed to by the root node’s map is an interior node, which is indexed by a minor hash. Interior nodes in this tree contains a zeroed out struct ext4_dir_entry_2 followed by a minor_hash->block map to find leafe nodes. Leaf nodes contain a linear array of all struct ext4_dir_entry_2 ; all of these entries (presumably) hash to the same value. If there is an overflow, the entries simply overflow into the next leaf node, and the least-significant bit of the hash (in the interior node map) that gets us to this next leaf node is set.

To traverse the directory as a htree, the code calculates the hash of the desired file name and uses it to find the corresponding block number. If the tree is flat, the block is a linear array of directory entries that can be searched; otherwise, the minor hash of the file name is computed and used against this second block to find the corresponding third block number. That third block number will be a linear array of directory entries.

To traverse the directory as a linear array (such as the old code does), the code simply reads every data block in the directory. The blocks used for the htree will appear to have no entries (aside from ‘.’ and ‘..’) and so only the leaf nodes will appear to have any interesting content.

The root of the htree is in struct dx_root , which is the full length of a data block:

Источник