xref: /openbmc/linux/kernel/Kconfig.preempt (revision 6d99a79c)
1
2choice
3	prompt "Preemption Model"
4	default PREEMPT_NONE
5
6config PREEMPT_NONE
7	bool "No Forced Preemption (Server)"
8	help
9	  This is the traditional Linux preemption model, geared towards
10	  throughput. It will still provide good latencies most of the
11	  time, but there are no guarantees and occasional longer delays
12	  are possible.
13
14	  Select this option if you are building a kernel for a server or
15	  scientific/computation system, or if you want to maximize the
16	  raw processing power of the kernel, irrespective of scheduling
17	  latencies.
18
19config PREEMPT_VOLUNTARY
20	bool "Voluntary Kernel Preemption (Desktop)"
21	depends on !ARCH_NO_PREEMPT
22	help
23	  This option reduces the latency of the kernel by adding more
24	  "explicit preemption points" to the kernel code. These new
25	  preemption points have been selected to reduce the maximum
26	  latency of rescheduling, providing faster application reactions,
27	  at the cost of slightly lower throughput.
28
29	  This allows reaction to interactive events by allowing a
30	  low priority process to voluntarily preempt itself even if it
31	  is in kernel mode executing a system call. This allows
32	  applications to run more 'smoothly' even when the system is
33	  under load.
34
35	  Select this if you are building a kernel for a desktop system.
36
37config PREEMPT
38	bool "Preemptible Kernel (Low-Latency Desktop)"
39	depends on !ARCH_NO_PREEMPT
40	select PREEMPT_COUNT
41	select UNINLINE_SPIN_UNLOCK if !ARCH_INLINE_SPIN_UNLOCK
42	help
43	  This option reduces the latency of the kernel by making
44	  all kernel code (that is not executing in a critical section)
45	  preemptible.  This allows reaction to interactive events by
46	  permitting a low priority process to be preempted involuntarily
47	  even if it is in kernel mode executing a system call and would
48	  otherwise not be about to reach a natural preemption point.
49	  This allows applications to run more 'smoothly' even when the
50	  system is under load, at the cost of slightly lower throughput
51	  and a slight runtime overhead to kernel code.
52
53	  Select this if you are building a kernel for a desktop or
54	  embedded system with latency requirements in the milliseconds
55	  range.
56
57endchoice
58
59config PREEMPT_COUNT
60       bool