master
  1/*-
  2 * SPDX-License-Identifier: BSD-4-Clause
  3 *
  4 * Copyright (c) 1995 Terrence R. Lambert
  5 * All rights reserved.
  6 *
  7 * Copyright (c) 1990, 1993
  8 *	The Regents of the University of California.  All rights reserved.
  9 * (c) UNIX System Laboratories, Inc.
 10 * All or some portions of this file are derived from material licensed
 11 * to the University of California by American Telephone and Telegraph
 12 * Co. or Unix System Laboratories, Inc. and are reproduced herein with
 13 * the permission of UNIX System Laboratories, Inc.
 14 *
 15 * Redistribution and use in source and binary forms, with or without
 16 * modification, are permitted provided that the following conditions
 17 * are met:
 18 * 1. Redistributions of source code must retain the above copyright
 19 *    notice, this list of conditions and the following disclaimer.
 20 * 2. Redistributions in binary form must reproduce the above copyright
 21 *    notice, this list of conditions and the following disclaimer in the
 22 *    documentation and/or other materials provided with the distribution.
 23 * 3. All advertising materials mentioning features or use of this software
 24 *    must display the following acknowledgement:
 25 *	This product includes software developed by the University of
 26 *	California, Berkeley and its contributors.
 27 * 4. Neither the name of the University nor the names of its contributors
 28 *    may be used to endorse or promote products derived from this software
 29 *    without specific prior written permission.
 30 *
 31 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
 32 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 33 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 34 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
 35 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 36 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 37 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 38 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 39 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 40 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 41 * SUCH DAMAGE.
 42 *
 43 *	@(#)kernel.h	8.3 (Berkeley) 1/21/94
 44 */
 45
 46#ifndef _SYS_KERNEL_H_
 47#define	_SYS_KERNEL_H_
 48
 49#include <sys/linker_set.h>
 50
 51#ifdef _KERNEL
 52
 53/* for intrhook and sysinit linked list below */
 54#include <sys/queue.h>
 55
 56/* for timestamping SYSINITs; other files may assume this is included here */
 57#include <sys/tslog.h>
 58
 59/* Global variables for the kernel. */
 60
 61/* 1.1 */
 62extern char kernelname[MAXPATHLEN];
 63
 64extern int tick;			/* usec per tick (1000000 / hz) */
 65extern int hz;				/* system clock's frequency */
 66extern int psratio;			/* ratio: prof / stat */
 67extern int stathz;			/* statistics clock's frequency */
 68extern int profhz;			/* profiling clock's frequency */
 69extern int profprocs;			/* number of process's profiling */
 70extern volatile int ticks;
 71
 72#endif /* _KERNEL */
 73
 74/*
 75 * Enumerated types for known system startup interfaces.
 76 *
 77 * Startup occurs in ascending numeric order; the list entries are
 78 * sorted prior to attempting startup to guarantee order.  Items
 79 * of the same level are arbitrated for order based on the 'order'
 80 * element.
 81 *
 82 * These numbers are arbitrary and are chosen ONLY for ordering; the
 83 * enumeration values are explicit rather than implicit to provide
 84 * for binary compatibility with inserted elements.
 85 *
 86 * The SI_SUB_LAST value must have the highest lexical value.
 87 */
 88enum sysinit_sub_id {
 89	SI_SUB_DUMMY		= 0x0000000,	/* not executed; for linker*/
 90	SI_SUB_DONE		= 0x0000001,	/* processed*/
 91	SI_SUB_TUNABLES		= 0x0700000,	/* establish tunable values */
 92	SI_SUB_COPYRIGHT	= 0x0800001,	/* first use of console*/
 93	SI_SUB_VM		= 0x1000000,	/* virtual memory system init */
 94	SI_SUB_COUNTER		= 0x1100000,	/* counter(9) is initialized */
 95	SI_SUB_KMEM		= 0x1800000,	/* kernel memory*/
 96	SI_SUB_HYPERVISOR	= 0x1A40000,	/*
 97						 * Hypervisor detection and
 98						 * virtualization support 
 99						 * setup.
100						 */
101	SI_SUB_WITNESS		= 0x1A80000,	/* witness initialization */
102	SI_SUB_MTX_POOL_DYNAMIC	= 0x1AC0000,	/* dynamic mutex pool */
103	SI_SUB_LOCK		= 0x1B00000,	/* various locks */
104	SI_SUB_EVENTHANDLER	= 0x1C00000,	/* eventhandler init */
105	SI_SUB_VNET_PRELINK	= 0x1E00000,	/* vnet init before modules */
106	SI_SUB_KLD		= 0x2000000,	/* KLD and module setup */
107	SI_SUB_KHELP		= 0x2080000,	/* khelp modules */
108	SI_SUB_CPU		= 0x2100000,	/* CPU resource(s)*/
109	SI_SUB_RACCT		= 0x2110000,	/* resource accounting */
110	SI_SUB_KDTRACE		= 0x2140000,	/* Kernel dtrace hooks */
111	SI_SUB_RANDOM		= 0x2160000,	/* random number generator */
112	SI_SUB_MAC		= 0x2180000,	/* TrustedBSD MAC subsystem */
113	SI_SUB_MAC_POLICY	= 0x21C0000,	/* TrustedBSD MAC policies */
114	SI_SUB_MAC_LATE		= 0x21D0000,	/* TrustedBSD MAC subsystem */
115	SI_SUB_VNET		= 0x21E0000,	/* vnet 0 */
116	SI_SUB_INTRINSIC	= 0x2200000,	/* proc 0*/
117	SI_SUB_VM_CONF		= 0x2300000,	/* config VM, set limits*/
118	SI_SUB_DDB_SERVICES	= 0x2380000,	/* capture, scripting, etc. */
119	SI_SUB_RUN_QUEUE	= 0x2400000,	/* set up run queue*/
120	SI_SUB_KTRACE		= 0x2480000,	/* ktrace */
121	SI_SUB_OPENSOLARIS	= 0x2490000,	/* OpenSolaris compatibility */
122	SI_SUB_AUDIT		= 0x24C0000,	/* audit */
123	SI_SUB_CREATE_INIT	= 0x2500000,	/* create init process*/
124	SI_SUB_SCHED_IDLE	= 0x2600000,	/* required idle procs */
125	SI_SUB_MBUF		= 0x2700000,	/* mbuf subsystem */
126	SI_SUB_INTR		= 0x2800000,	/* interrupt threads */
127	SI_SUB_TASKQ		= 0x2880000,	/* task queues */
128	SI_SUB_EPOCH		= 0x2888000,	/* epoch subsystem */
129#ifdef EARLY_AP_STARTUP
130	SI_SUB_SMP		= 0x2900000,	/* start the APs*/
131#endif
132	SI_SUB_SOFTINTR		= 0x2A00000,	/* start soft interrupt thread */
133	SI_SUB_DEVFS		= 0x2F00000,	/* devfs ready for devices */
134	SI_SUB_INIT_IF		= 0x3000000,	/* prep for net interfaces */
135	SI_SUB_NETGRAPH		= 0x3010000,	/* Let Netgraph initialize */
136	SI_SUB_DTRACE		= 0x3020000,	/* DTrace subsystem */
137	SI_SUB_DTRACE_PROVIDER	= 0x3048000,	/* DTrace providers */
138	SI_SUB_DTRACE_ANON	= 0x308C000,	/* DTrace anon enabling */
139	SI_SUB_DRIVERS		= 0x3100000,	/* Let Drivers initialize */
140	SI_SUB_CONFIGURE	= 0x3800000,	/* Configure devices */
141	SI_SUB_VFS		= 0x4000000,	/* virtual filesystem*/
142	SI_SUB_CLOCKS		= 0x4800000,	/* real time and stat clocks*/
143	SI_SUB_SYSV_SHM		= 0x6400000,	/* System V shared memory*/
144	SI_SUB_SYSV_SEM		= 0x6800000,	/* System V semaphores*/
145	SI_SUB_SYSV_MSG		= 0x6C00000,	/* System V message queues*/
146	SI_SUB_P1003_1B		= 0x6E00000,	/* P1003.1B realtime */
147	SI_SUB_PSEUDO		= 0x7000000,	/* pseudo devices*/
148	SI_SUB_EXEC		= 0x7400000,	/* execve() handlers */
149	SI_SUB_PROTO_BEGIN	= 0x8000000,	/* VNET initialization */
150	SI_SUB_PROTO_PFIL	= 0x8100000,	/* Initialize pfil before FWs */
151	SI_SUB_PROTO_IF		= 0x8400000,	/* interfaces*/
152	SI_SUB_PROTO_DOMAININIT	= 0x8600000,	/* domain registration system */
153	SI_SUB_PROTO_MC		= 0x8700000,	/* Multicast */
154	SI_SUB_PROTO_DOMAIN	= 0x8800000,	/* domains (address families?)*/
155	SI_SUB_PROTO_FIREWALL	= 0x8806000,	/* Firewalls */
156	SI_SUB_PROTO_IFATTACHDOMAIN = 0x8808000,/* domain dependent data init */
157	SI_SUB_PROTO_END	= 0x8ffffff,	/* VNET helper functions */
158	SI_SUB_KPROF		= 0x9000000,	/* kernel profiling*/
159	SI_SUB_KICK_SCHEDULER	= 0xa000000,	/* start the timeout events*/
160	SI_SUB_INT_CONFIG_HOOKS	= 0xa800000,	/* Interrupts enabled config */
161	SI_SUB_ROOT_CONF	= 0xb000000,	/* Find root devices */
162	SI_SUB_INTRINSIC_POST	= 0xd000000,	/* proc 0 cleanup*/
163	SI_SUB_SYSCALLS		= 0xd800000,	/* register system calls */
164	SI_SUB_VNET_DONE	= 0xdc00000,	/* vnet registration complete */
165	SI_SUB_KTHREAD_INIT	= 0xe000000,	/* init process*/
166	SI_SUB_KTHREAD_PAGE	= 0xe400000,	/* pageout daemon*/
167	SI_SUB_KTHREAD_VM	= 0xe800000,	/* vm daemon*/
168	SI_SUB_KTHREAD_BUF	= 0xea00000,	/* buffer daemon*/
169	SI_SUB_KTHREAD_UPDATE	= 0xec00000,	/* update daemon*/
170	SI_SUB_KTHREAD_IDLE	= 0xee00000,	/* idle procs*/
171#ifndef EARLY_AP_STARTUP
172	SI_SUB_SMP		= 0xf000000,	/* start the APs*/
173#endif	
174	SI_SUB_RACCTD		= 0xf100000,	/* start racctd*/
175	SI_SUB_LAST		= 0xfffffff	/* final initialization */
176};
177
178/*
179 * Some enumerated orders; "ANY" sorts last.
180 */
181enum sysinit_elem_order {
182	SI_ORDER_FIRST		= 0x0000000,	/* first*/
183	SI_ORDER_SECOND		= 0x0000001,	/* second*/
184	SI_ORDER_THIRD		= 0x0000002,	/* third*/
185	SI_ORDER_FOURTH		= 0x0000003,	/* fourth*/
186	SI_ORDER_FIFTH		= 0x0000004,	/* fifth*/
187	SI_ORDER_SIXTH		= 0x0000005,	/* sixth*/
188	SI_ORDER_SEVENTH	= 0x0000006,	/* seventh*/
189	SI_ORDER_EIGHTH		= 0x0000007,	/* eighth*/
190	SI_ORDER_MIDDLE		= 0x1000000,	/* somewhere in the middle */
191	SI_ORDER_ANY		= 0xfffffff	/* last*/
192};
193
194/*
195 * A system initialization call instance
196 *
197 * At the moment there is one instance of sysinit.  We probably do not
198 * want two which is why this code is if'd out, but we definitely want
199 * to discern SYSINIT's which take non-constant data pointers and
200 * SYSINIT's which take constant data pointers,
201 *
202 * The C_* macros take functions expecting const void * arguments
203 * while the non-C_* macros take functions expecting just void * arguments.
204 *
205 * With -Wcast-qual on, the compiler issues warnings:
206 *	- if we pass non-const data or functions taking non-const data
207 *	  to a C_* macro.
208 *
209 *	- if we pass const data to the normal macros
210 *
211 * However, no warning is issued if we pass a function taking const data
212 * through a normal non-const macro.  This is ok because the function is
213 * saying it won't modify the data so we don't care whether the data is
214 * modifiable or not.
215 */
216
217typedef void (*sysinit_nfunc_t)(void *);
218typedef void (*sysinit_cfunc_t)(const void *);
219
220struct sysinit {
221	enum sysinit_sub_id	subsystem;	/* subsystem identifier*/
222	enum sysinit_elem_order	order;		/* init order within subsystem*/
223	STAILQ_ENTRY(sysinit)	next;		/* singly-linked list */
224	sysinit_cfunc_t func;			/* function		*/
225	const void	*udata;			/* multiplexer/argument */
226};
227
228/*
229 * Default: no special processing
230 *
231 * The C_ version of SYSINIT is for data pointers to const
232 * data ( and functions taking data pointers to const data ).
233 * At the moment it is no different from SYSINIT and thus
234 * still results in warnings.
235 *
236 * The casts are necessary to have the compiler produce the
237 * correct warnings when -Wcast-qual is used.
238 *
239 */
240#ifdef TSLOG
241struct sysinit_tslog {
242	sysinit_cfunc_t func;
243	const void * data;
244	const char * name;
245};
246static inline void
247sysinit_tslog_shim(const void * data)
248{
249	const struct sysinit_tslog * x = data;
250
251	TSRAW(curthread, TS_ENTER, "SYSINIT", x->name);
252	(x->func)(x->data);
253	TSRAW(curthread, TS_EXIT, "SYSINIT", x->name);
254}
255#define	C_SYSINIT(uniquifier, subsystem, order, func, ident)	\
256	static struct sysinit_tslog uniquifier ## _sys_init_tslog = {	\
257		func,						\
258		(ident),					\
259		#uniquifier					\
260	};							\
261	static struct sysinit uniquifier ## _sys_init = {	\
262		subsystem,					\
263		order,						\
264		{ NULL },					\
265		sysinit_tslog_shim,				\
266		&uniquifier ## _sys_init_tslog			\
267	};							\
268	DATA_WSET(sysinit_set,uniquifier ## _sys_init)
269#else
270#define	C_SYSINIT(uniquifier, subsystem, order, func, ident)	\
271	static struct sysinit uniquifier ## _sys_init = {	\
272		subsystem,					\
273		order,						\
274		{ NULL },					\
275		func,						\
276		(ident)						\
277	};							\
278	DATA_WSET(sysinit_set,uniquifier ## _sys_init)
279#endif
280
281#define	SYSINIT(uniquifier, subsystem, order, func, ident)	\
282	C_SYSINIT(uniquifier, subsystem, order,			\
283	(sysinit_cfunc_t)(sysinit_nfunc_t)func, (void *)(ident))
284
285/*
286 * Called on module unload: no special processing
287 */
288#define	C_SYSUNINIT(uniquifier, subsystem, order, func, ident)	\
289	static struct sysinit uniquifier ## _sys_uninit = {	\
290		subsystem,					\
291		order,						\
292		{ NULL },					\
293		func,						\
294		(ident)						\
295	};							\
296	DATA_WSET(sysuninit_set,uniquifier ## _sys_uninit)
297
298#define	SYSUNINIT(uniquifier, subsystem, order, func, ident)	\
299	C_SYSUNINIT(uniquifier, subsystem, order,		\
300	(sysinit_cfunc_t)(sysinit_nfunc_t)func, (void *)(ident))
301
302void	sysinit_add(struct sysinit **set, struct sysinit **set_end);
303
304#ifdef _KERNEL
305
306/*
307 * Infrastructure for tunable 'constants'.  Value may be specified at compile
308 * time or kernel load time.  Rules relating tunables together can be placed
309 * in a SYSINIT function at SI_SUB_TUNABLES with SI_ORDER_ANY.
310 *
311 * WARNING: developers should never use the reserved suffixes specified in
312 * loader.conf(5) for any tunables or conflicts will result.
313 */
314
315/*
316 * int
317 * please avoid using for new tunables!
318 */
319extern void tunable_int_init(void *);
320struct tunable_int {
321	const char *path;
322	int *var;
323};
324#define	TUNABLE_INT(path, var)					\
325	static struct tunable_int __CONCAT(__tunable_int_, __LINE__) = { \
326		(path),						\
327		(var),						\
328	};							\
329	SYSINIT(__CONCAT(__Tunable_init_, __LINE__),		\
330	    SI_SUB_TUNABLES, SI_ORDER_MIDDLE, tunable_int_init,	\
331	    &__CONCAT(__tunable_int_, __LINE__))
332
333#define	TUNABLE_INT_FETCH(path, var)	getenv_int((path), (var))
334
335/*
336 * long
337 */
338extern void tunable_long_init(void *);
339struct tunable_long {
340	const char *path;
341	long *var;
342};
343#define	TUNABLE_LONG(path, var)					\
344	static struct tunable_long __CONCAT(__tunable_long_, __LINE__) = { \
345		(path),						\
346		(var),						\
347	};							\
348	SYSINIT(__CONCAT(__Tunable_init_, __LINE__),		\
349	    SI_SUB_TUNABLES, SI_ORDER_MIDDLE, tunable_long_init,\
350	    &__CONCAT(__tunable_long_, __LINE__))
351
352#define	TUNABLE_LONG_FETCH(path, var)	getenv_long((path), (var))
353
354/*
355 * unsigned long
356 */
357extern void tunable_ulong_init(void *);
358struct tunable_ulong {
359	const char *path;
360	unsigned long *var;
361};
362#define	TUNABLE_ULONG(path, var)				\
363	static struct tunable_ulong __CONCAT(__tunable_ulong_, __LINE__) = { \
364		(path),						\
365		(var),						\
366	};							\
367	SYSINIT(__CONCAT(__Tunable_init_, __LINE__),		\
368	    SI_SUB_TUNABLES, SI_ORDER_MIDDLE, tunable_ulong_init, \
369	    &__CONCAT(__tunable_ulong_, __LINE__))
370
371#define	TUNABLE_ULONG_FETCH(path, var)	getenv_ulong((path), (var))
372
373/*
374 * int64_t
375 */
376extern void tunable_int64_init(void *);
377struct tunable_int64 {
378	const char *path;
379	int64_t *var;
380};
381#define	TUNABLE_INT64(path, var)				\
382	static struct tunable_int64 __CONCAT(__tunable_int64_, __LINE__) = { \
383		(path),						\
384		(var),						\
385	};							\
386	SYSINIT(__CONCAT(__Tunable_init_, __LINE__),		\
387	    SI_SUB_TUNABLES, SI_ORDER_MIDDLE, tunable_int64_init, \
388	    &__CONCAT(__tunable_int64_, __LINE__))
389
390#define	TUNABLE_INT64_FETCH(path, var)	getenv_int64((path), (var))
391
392/*
393 * uint64_t
394 */
395extern void tunable_uint64_init(void *);
396struct tunable_uint64 {
397	const char *path;
398	uint64_t *var;
399};
400#define	TUNABLE_UINT64(path, var)				\
401	static struct tunable_uint64 __CONCAT(__tunable_uint64_, __LINE__) = { \
402		(path),						\
403		(var),						\
404	};							\
405	SYSINIT(__CONCAT(__Tunable_init_, __LINE__),		\
406	    SI_SUB_TUNABLES, SI_ORDER_MIDDLE, tunable_uint64_init, \
407	    &__CONCAT(__tunable_uint64_, __LINE__))
408
409#define	TUNABLE_UINT64_FETCH(path, var)	getenv_uint64((path), (var))
410
411/*
412 * quad
413 */
414extern void tunable_quad_init(void *);
415struct tunable_quad {
416	const char *path;
417	quad_t *var;
418};
419#define	TUNABLE_QUAD(path, var)					\
420	static struct tunable_quad __CONCAT(__tunable_quad_, __LINE__) = { \
421		(path),						\
422		(var),						\
423	};							\
424	SYSINIT(__CONCAT(__Tunable_init_, __LINE__),		\
425	    SI_SUB_TUNABLES, SI_ORDER_MIDDLE, tunable_quad_init, \
426	    &__CONCAT(__tunable_quad_, __LINE__))
427
428#define	TUNABLE_QUAD_FETCH(path, var)	getenv_quad((path), (var))
429
430/*
431 * bool
432 */
433extern void tunable_bool_init(void *);
434struct tunable_bool {
435	const char *path;
436	bool *var;
437};
438#define	TUNABLE_BOOL(path, var) \
439	static struct tunable_bool __CONCAT(__tunable_bool_, __LINE__) = { \
440		(path),						\
441		(var),						\
442	};							\
443	SYSINIT(__CONCAT(__Tunable_init_, __LINE__),		\
444	    SI_SUB_TUNABLES, SI_ORDER_MIDDLE, tunable_bool_init, \
445	    &__CONCAT(__tunable_bool_, __LINE__))
446
447#define	TUNABLE_BOOL_FETCH(path, var)	getenv_bool((path), (var))
448
449extern void tunable_str_init(void *);
450struct tunable_str {
451	const char *path;
452	char *var;
453	int size;
454};
455#define	TUNABLE_STR(path, var, size)				\
456	static struct tunable_str __CONCAT(__tunable_str_, __LINE__) = { \
457		(path),						\
458		(var),						\
459		(size),						\
460	};							\
461	SYSINIT(__CONCAT(__Tunable_init_, __LINE__),		\
462	    SI_SUB_TUNABLES, SI_ORDER_MIDDLE, tunable_str_init,	\
463	    &__CONCAT(__tunable_str_, __LINE__))
464
465#define	TUNABLE_STR_FETCH(path, var, size)			\
466	getenv_string((path), (var), (size))
467
468#endif /* _KERNEL */
469
470typedef void (*ich_func_t)(void *_arg);
471
472struct intr_config_hook {
473	STAILQ_ENTRY(intr_config_hook) ich_links;
474	uintptr_t	ich_state;
475#define ICHS_QUEUED	0x1
476#define ICHS_RUNNING	0x2
477#define	ICHS_DONE	0x3
478	ich_func_t	ich_func;
479	void		*ich_arg;
480};
481
482int	config_intrhook_establish(struct intr_config_hook *hook);
483void	config_intrhook_disestablish(struct intr_config_hook *hook);
484int	config_intrhook_drain(struct intr_config_hook *hook);
485void	config_intrhook_oneshot(ich_func_t _func, void *_arg);
486
487#endif /* !_SYS_KERNEL_H_*/