1/*-
2 * SPDX-License-Identifier: BSD-3-Clause
3 *
4 * Copyright (c) 2002 Poul-Henning Kamp
5 * Copyright (c) 2002 Networks Associates Technology, Inc.
6 * All rights reserved.
7 *
8 * This software was developed for the FreeBSD Project by Poul-Henning Kamp
9 * and NAI Labs, the Security Research Division of Network Associates, Inc.
10 * under DARPA/SPAWAR contract N66001-01-C-8035 ("CBOSS"), as part of the
11 * DARPA CHATS research program.
12 *
13 * Redistribution and use in source and binary forms, with or without
14 * modification, are permitted provided that the following conditions
15 * are met:
16 * 1. Redistributions of source code must retain the above copyright
17 *    notice, this list of conditions and the following disclaimer.
18 * 2. Redistributions in binary form must reproduce the above copyright
19 *    notice, this list of conditions and the following disclaimer in the
20 *    documentation and/or other materials provided with the distribution.
21 * 3. The names of the authors may not be used to endorse or promote
22 *    products derived from this software without specific prior written
23 *    permission.
24 *
25 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
26 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
27 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
28 * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
29 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
30 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
31 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
32 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
33 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
34 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
35 * SUCH DAMAGE.
36 */
37
38#include <sys/param.h>
39#include <sys/systm.h>
40#include <sys/kernel.h>
41#include <sys/eventhandler.h>
42#include <sys/malloc.h>
43#include <sys/bio.h>
44#include <sys/sysctl.h>
45#include <sys/proc.h>
46#include <sys/unistd.h>
47#include <sys/kthread.h>
48#include <sys/lock.h>
49#include <sys/mutex.h>
50#include <sys/sbuf.h>
51#include <sys/sched.h>
52#include <sys/sx.h>
53#include <geom/geom.h>
54#include <geom/geom_int.h>
55
56MALLOC_DEFINE(M_GEOM, "GEOM", "Geom data structures");
57
58struct sx topology_lock;
59
60static struct proc *g_proc;
61struct thread __read_mostly *g_up_td;
62struct thread __read_mostly *g_down_td;
63static struct thread __read_mostly *g_event_td;
64
65int __read_mostly g_debugflags;
66int __read_mostly g_collectstats = G_STATS_PROVIDERS;
67int g_shutdown;
68int g_notaste;
69
70/*
71 * G_UP and G_DOWN are the two threads which push I/O through the
72 * stack.
73 *
74 * Things are procesed in a FIFO order, but these threads could be
75 * part of I/O prioritization by deciding which bios/bioqs to service
76 * in what order.
77 *
78 * We have only one thread in each direction, it is believed that until
79 * a very non-trivial workload in the UP/DOWN path this will be enough,
80 * but more than one can actually be run without problems.
81 *
82 * Holding the "mymutex" is a debugging feature:  It prevents people
83 * from sleeping in the UP/DOWN I/O path by mistake or design (doing
84 * so almost invariably result in deadlocks since it stalls all I/O
85 * processing in the given direction.
86 */
87
88static void
89g_up_procbody(void *arg)
90{
91
92	thread_lock(g_up_td);
93	sched_prio(g_up_td, PRIBIO);
94	thread_unlock(g_up_td);
95	for(;;) {
96		g_io_schedule_up(g_up_td);
97	}
98}
99
100static void
101g_down_procbody(void *arg)
102{
103
104	thread_lock(g_down_td);
105	sched_prio(g_down_td, PRIBIO);
106	thread_unlock(g_down_td);
107	for(;;) {
108		g_io_schedule_down(g_down_td);
109	}
110}
111
112static void
113g_event_procbody(void *arg)
114{
115
116	thread_lock(g_event_td);
117	sched_prio(g_event_td, PRIBIO);
118	thread_unlock(g_event_td);
119	g_run_events();
120	/* NOTREACHED */
121}
122
123int
124g_is_geom_thread(struct thread *td)
125{
126
127	return (td == g_up_td || td == g_down_td || td == g_event_td);
128}
129
130static void
131geom_shutdown(void *foo __unused)
132{
133
134	g_shutdown = 1;
135}
136
137void
138g_init(void)
139{
140
141	g_trace(G_T_TOPOLOGY, "g_ignition");
142	sx_init(&topology_lock, "GEOM topology");
143	g_io_init();
144	g_event_init();
145	g_ctl_init();
146	kproc_kthread_add(g_event_procbody, NULL, &g_proc, &g_event_td,
147	    RFHIGHPID, 0, "geom", "g_event");
148	kproc_kthread_add(g_up_procbody, NULL, &g_proc, &g_up_td,
149	    RFHIGHPID, 0, "geom", "g_up");
150	kproc_kthread_add(g_down_procbody, NULL, &g_proc, &g_down_td,
151	    RFHIGHPID, 0, "geom", "g_down");
152	EVENTHANDLER_REGISTER(shutdown_pre_sync, geom_shutdown, NULL,
153		SHUTDOWN_PRI_FIRST);
154}
155
156static int
157sysctl_kern_geom_confany(struct sysctl_req *req, g_event_t *func, size_t *hint)
158{
159	size_t len = 0;
160	int error = 0;
161	struct sbuf *sb;
162
163	if (req->oldptr == NULL) {
164		sb = sbuf_new(NULL, NULL, PAGE_SIZE, SBUF_FIXEDLEN |
165		    SBUF_INCLUDENUL);
166		sbuf_set_drain(sb, sbuf_count_drain, &len);
167		g_waitfor_event(func, sb, M_WAITOK, NULL);
168		req->oldidx = *hint = len;
169	} else {
170		sb = sbuf_new(NULL, NULL, *hint, SBUF_AUTOEXTEND |
171		    SBUF_INCLUDENUL);
172		g_waitfor_event(func, sb, M_WAITOK, NULL);
173		*hint = sbuf_len(sb);
174		error = SYSCTL_OUT(req, sbuf_data(sb), sbuf_len(sb));
175	}
176	sbuf_delete(sb);
177	return error;
178}
179
180static int
181sysctl_kern_geom_conftxt(SYSCTL_HANDLER_ARGS)
182{
183	static size_t hint = PAGE_SIZE;
184
185	return (sysctl_kern_geom_confany(req, g_conftxt, &hint));
186}
187
188static int
189sysctl_kern_geom_confdot(SYSCTL_HANDLER_ARGS)
190{
191	static size_t hint = PAGE_SIZE;
192
193	return (sysctl_kern_geom_confany(req, g_confdot, &hint));
194}
195
196static int
197sysctl_kern_geom_confxml(SYSCTL_HANDLER_ARGS)
198{
199	static size_t hint = PAGE_SIZE;
200
201	return (sysctl_kern_geom_confany(req, g_confxml, &hint));
202}
203
204SYSCTL_NODE(_kern, OID_AUTO, geom, CTLFLAG_RW | CTLFLAG_MPSAFE, 0,
205    "GEOMetry management");
206
207SYSCTL_PROC(_kern_geom, OID_AUTO, confxml,
208    CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, 0, 0,
209    sysctl_kern_geom_confxml, "",
210    "Dump the GEOM config in XML");
211
212SYSCTL_PROC(_kern_geom, OID_AUTO, confdot,
213    CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, 0, 0,
214    sysctl_kern_geom_confdot, "",
215    "Dump the GEOM config in dot");
216
217SYSCTL_PROC(_kern_geom, OID_AUTO, conftxt,
218    CTLTYPE_STRING | CTLFLAG_RD | CTLFLAG_MPSAFE, 0, 0,
219    sysctl_kern_geom_conftxt, "",
220    "Dump the GEOM config in txt");
221
222SYSCTL_INT(_kern_geom, OID_AUTO, debugflags, CTLFLAG_RWTUN,
223	&g_debugflags, 0, "Set various trace levels for GEOM debugging");
224
225SYSCTL_INT(_kern_geom, OID_AUTO, notaste, CTLFLAG_RW,
226	&g_notaste, 0, "Prevent GEOM tasting");
227
228SYSCTL_INT(_kern_geom, OID_AUTO, collectstats, CTLFLAG_RW,
229	&g_collectstats, 0,
230	"Control statistics collection on GEOM providers and consumers");
231
232SYSCTL_INT(_debug_sizeof, OID_AUTO, g_class, CTLFLAG_RD,
233	SYSCTL_NULL_INT_PTR, sizeof(struct g_class), "sizeof(struct g_class)");
234SYSCTL_INT(_debug_sizeof, OID_AUTO, g_geom, CTLFLAG_RD,
235	SYSCTL_NULL_INT_PTR, sizeof(struct g_geom), "sizeof(struct g_geom)");
236SYSCTL_INT(_debug_sizeof, OID_AUTO, g_provider, CTLFLAG_RD,
237	SYSCTL_NULL_INT_PTR, sizeof(struct g_provider), "sizeof(struct g_provider)");
238SYSCTL_INT(_debug_sizeof, OID_AUTO, g_consumer, CTLFLAG_RD,
239	SYSCTL_NULL_INT_PTR, sizeof(struct g_consumer), "sizeof(struct g_consumer)");
240SYSCTL_INT(_debug_sizeof, OID_AUTO, g_bioq, CTLFLAG_RD,
241	SYSCTL_NULL_INT_PTR, sizeof(struct g_bioq), "sizeof(struct g_bioq)");
242