1	CPU Scheduler implementation hints for architecture specific code
2
3	Nick Piggin, 2005
4
5Context switch
6==============
71. Runqueue locking
8By default, the switch_to arch function is called with the runqueue
9locked. This is usually not a problem unless switch_to may need to
10take the runqueue lock. This is usually due to a wake up operation in
11the context switch. See arch/ia64/include/asm/system.h for an example.
12
13To request the scheduler call switch_to with the runqueue unlocked,
14you must `#define __ARCH_WANT_UNLOCKED_CTXSW` in a header file
15(typically the one where switch_to is defined).
16
17Unlocked context switches introduce only a very minor performance
18penalty to the core scheduler implementation in the CONFIG_SMP case.
19
202. Interrupt status
21By default, the switch_to arch function is called with interrupts
22disabled. Interrupts may be enabled over the call if it is likely to
23introduce a significant interrupt latency by adding the line
24`#define __ARCH_WANT_INTERRUPTS_ON_CTXSW` in the same place as for
25unlocked context switches. This define also implies
26`__ARCH_WANT_UNLOCKED_CTXSW`. See arch/arm/include/asm/system.h for an
27example.
28
29
30CPU idle
31========
32Your cpu_idle routines need to obey the following rules:
33
341. Preempt should now disabled over idle routines. Should only
35   be enabled to call schedule() then disabled again.
36
372. need_resched/TIF_NEED_RESCHED is only ever set, and will never
38   be cleared until the running task has called schedule(). Idle
39   threads need only ever query need_resched, and may never set or
40   clear it.
41
423. When cpu_idle finds (need_resched() == 'true'), it should call
43   schedule(). It should not call schedule() otherwise.
44
454. The only time interrupts need to be disabled when checking
46   need_resched is if we are about to sleep the processor until
47   the next interrupt (this doesn't provide any protection of
48   need_resched, it prevents losing an interrupt).
49
50	4a. Common problem with this type of sleep appears to be:
51	        local_irq_disable();
52	        if (!need_resched()) {
53	                local_irq_enable();
54	                *** resched interrupt arrives here ***
55	                __asm__("sleep until next interrupt");
56	        }
57
585. TIF_POLLING_NRFLAG can be set by idle routines that do not
59   need an interrupt to wake them up when need_resched goes high.
60   In other words, they must be periodically polling need_resched,
61   although it may be reasonable to do some background work or enter
62   a low CPU priority.
63
64   	5a. If TIF_POLLING_NRFLAG is set, and we do decide to enter
65	    an interrupt sleep, it needs to be cleared then a memory
66	    barrier issued (followed by a test of need_resched with
67	    interrupts disabled, as explained in 3).
68
69arch/x86/kernel/process.c has examples of both polling and
70sleeping idle functions.
71
72
73Possible arch/ problems
74=======================
75
76Possible arch problems I found (and either tried to fix or didn't):
77
78h8300 - Is such sleeping racy vs interrupts? (See #4a).
79        The H8/300 manual I found indicates yes, however disabling IRQs
80        over the sleep mean only NMIs can wake it up, so can't fix easily
81        without doing spin waiting.
82
83ia64 - is safe_halt call racy vs interrupts? (does it sleep?) (See #4a)
84
85sh64 - Is sleeping racy vs interrupts? (See #4a)
86
87sparc - IRQs on at this point(?), change local_irq_save to _disable.
88      - TODO: needs secondary CPUs to disable preempt (See #1)
89
90