A glossary of various terms and acronyms related to the Linux kernel. If you know something, please create yourself an account (UserPreferences) and add a term in alphabetical order. Don't be afraid to improve other people's explanations, the goal is to have a high quality document for the readers of this site. If everybody who reads this page adds a term each week, this glossary should be complete within a few months...

Index

Anchor(0-9)

0-9

---

2Q algorithm

MM algorithm based on two areas, one managed as a FIFO queue, and one as an LRU list.

8259 PIC

Outdated interrupt controller present on Intel hardware.

Anchor(A)

A

---

AIO

Asynchronous IO, IO that is performed without the issuing process blocking on IO completion.

Anticipatory Scheduler

A disk IO scheduler that leaves the disk idle after a read, in anticipation of the next read.

Anchor(C)

C

---

CBQ

Class Based Queueing, a hierarchical packet fair queueing qdisc. [http://www.icir.org/floyd/cbq.html CBQ Homepage]

Classifier

(also: filter or tcf) classifies a network packet by inspecting it, used by QDiscs.

Context switch

switching the CPU from running one thread to running another thread.

Copy-on-Write

(also: COW) reuse and share existing objects and copy them not until a modification is required.

Current

a kernel variable which points to the task_struct structure of the process currently running on this CPU.

Anchor(D)

D

---

Device Mapper

A technology for presenting arbitrary groupings of underlying sectors on physical devices in a consistent logical fashion usable by higher level algorithms. Heavily used by kernel technologies such as LVM.

Dwarf

Debugging Information Format

Anchor(E)

E

---

ELF

Executable Linkable Format, a popular binary format, the default for Linux on most architectures.

ematch

Extended Match, small classification helper attached to classifiers.

G

• GRUB: GRand Unified Bootloader, a popular bootloader for Linux, BSD, and other OSes.

H

• HTB: Hierachical Token Bucket, a qdisc based on TBF and CBQ [http://luxik.cdi.cz/~devik/qos/htb/manual/theory.htm HTB Theory]

I

• ISR: interrupt service routine, the function in each device driver that gets called when an interrupt happens.

J

• Jiffies: An incrementing counter representing system "uptime" in ticks - or the number of timer interrupts since boot. Ultimately the entire original concept of a jiffy will likely vanish as systems use timer events only when necessary and become "jiffyless".

K

• kswapd: a kernel thread that frees up memory by evicting data from caches and paging out userspace memory, part of the virtual memory subsystem.

L

• Linux Device Drivers, 2nd Edition: [http://www.xml.com/ldd/chapter/book/].

• LKM: Linux Kernel Module. A (often dynamically loadable at system runtime) kernel extension ("driver") to support, for example, some kind of new hardware device or generic software abstraction.

• LKML: Linux Kernel Mailing List. The primary virtual watering hole (meeting ground) for kernel developers to share ideas and bounce opinions off oneanother during the course of the kernel development process. FAQ at [http://www.tux.org/lkml/].

• LSM: Linux Security Module. a security framework for providing different security levels.
• LVM: Logical volume management. A technology for providing an arbitrary logical view of underlying data storage in a fashion supporting resizing and restructuring of storage on the fly. Currently in version 2, originally written by Sistina (now Redhat).

• LXR: a cross-reference tool that can be used to navigate the Linux kernel source code, available at [http://lxr.linux.no/ lxr.linux.no].

M

• mem_map: A contiguous virtual array of struct pages representing the entirity of physical memory pages available within a system.
• MMU: memory management unit, part of the CPU that is needed for virtual memory.

N

• netlink: Communication protocol between kernel and userspace

P

• Page cache: a cache of file data and filesystem metadata, dynamically grown an shrunk depending on other memory use.
• Page table: data structure used by the MMU to translate virtual memory addresses to physical memory addresses.
• PFN: page frame number, index into the mem_map[] array which describes physical memory pages.

• PGD: Page Global Directory, the top level of the page table tree. The page table hierarchy is pgd -> pud -> pmd -> pte.

• PID: process identifier (POSIX thread identifier)
• PMD: Page Mid-level Directory, note that pmds are folded into pgds on systems with 2 level page tables.
• Process descriptor: kernel data structure that describes/accounts process data related to a single process.
• PTE: Page Table Entry
• PUD: Page Upper Directory, note that puds are folded into pmds, except on systems with 4-levels page tables.

Q

• QDisc: [http://lartc.org/howto/lartc.qdisc.html Queueing Discipline], queues packets before they are sent out to the network device, enforces QoS requirements, provides traffic shaping and prioritizing capabilities.

• QoS: Quality of Service, method to define the importance/priority of network services

R

• RCU: Read Copy Update, a mechanism for ["SMPSynchronisation"]
• Rlimit: resource limit, eg. "maximum amount of virtual memory" or "maximum number of processes". Can be per process or per user.

S

• Semaphore: a lock mechanism that works per process context, see ["SMPSynchronisation"]

• Scheduler: the part of the kernel that chooses a suitable process to run on the cpu, see the [http://lxr.linux.no/ident?i=schedule schedule()] function.

• Shared/Paged Socket Buffer: (also: pskb) Socket Buffer with uncontinuous data buffer, used for zero copy, TSO and Scatter/Gather capable network cards.
• Slab cache: a fast, SMP scalable kernel memory allocator.
• Socket Buffer: (also: skb) data structure used to hold the data and attributes of a network packet.
• SoftIRQ: kind of bottom half rarely used.
• Spin lock: a simple SMP lock, see ["SMPSynchronisation"]

• Swap token: a token to temporarily protect a process from pageout, an alternative approach to memory scheduling, thrashing control. See the [http://www.cs.wm.edu/~sjiang/token.pdf Token Based Thrashing Control] paper by Song Jiang and the [http://linux-mm.org/wiki/SwapTokenTuning Linux-MM wiki].

• System call: (also: syscall) the way a program transitions from userspace into kernel space, to call a kernel space function.
• System.map: symbol table used by ksymoops to resolve numbers to function names in Oops. Also used by ps and top for WCHAN field.

T

• TASK_RUNNING: State of the task that can run but is not necessarily running.
• TBF: Token Bucket Filter, a qdisc used for rate limiting
• TGID: Task group identifier (POSIX process identifier)

U

• Use-once: the page replacement algorithm used by the Linux 2.6 kernel, based on the ideas behind the 2Q page replacement algorithm, also see the [http://linux-mm.org/wiki/AdvancedPageReplacement AdvancedPageReplacement] page.

V

• VDSO: Virtual Dynamically-linked Shared Object, a kernel-provided shared library that helps userspace perform a few kernel actions without the overhead of a system call, as well as automatically chosing the most efficient syscall mechanism. Also called the "vsyscall page".
• VFS: Virtual File System, an interface through which multiple filesystems can be hooked into the kernel.
• Virtual memory: every process in the system gets its own memory address space, independant of the other processes.
• Vsyscall page: see VDSO.

X

• [http://xen.sf.net Xen]: a paravirtualisation engine for Linux, an efficient way to run multiple Linux OSes on one computer. Also runs BSD, Plan9 and other OSes.

• XIP, Execute In Place: the ability to run an executable directly from the filesystem (usually ROM or flash), instead of loading it into memory.

Z

• Zero-Copy: A special networking code path where data is sent to the network directly from userspace memory; this avoids unnecessary copying of data and improves performance.