subr_syscall.c revision 111033
1/*-
2 * Copyright (C) 1994, David Greenman
3 * Copyright (c) 1990, 1993
4 *	The Regents of the University of California.  All rights reserved.
5 *
6 * This code is derived from software contributed to Berkeley by
7 * the University of Utah, and William Jolitz.
8 *
9 * Redistribution and use in source and binary forms, with or without
10 * modification, are permitted provided that the following conditions
11 * are met:
12 * 1. Redistributions of source code must retain the above copyright
13 *    notice, this list of conditions and the following disclaimer.
14 * 2. Redistributions in binary form must reproduce the above copyright
15 *    notice, this list of conditions and the following disclaimer in the
16 *    documentation and/or other materials provided with the distribution.
17 * 3. All advertising materials mentioning features or use of this software
18 *    must display the following acknowledgement:
19 *	This product includes software developed by the University of
20 *	California, Berkeley and its contributors.
21 * 4. Neither the name of the University nor the names of its contributors
22 *    may be used to endorse or promote products derived from this software
23 *    without specific prior written permission.
24 *
25 * THIS SOFTWARE IS PROVIDED BY THE REGENTS 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 REGENTS 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 *	from: @(#)trap.c	7.4 (Berkeley) 5/13/91
38 * $FreeBSD: head/sys/kern/subr_trap.c 111033 2003-02-17 09:58:11Z jeff $
39 */
40
41#include "opt_mac.h"
42#ifdef __i386__
43#include "opt_npx.h"
44#endif
45
46#include <sys/param.h>
47#include <sys/bus.h>
48#include <sys/kernel.h>
49#include <sys/lock.h>
50#include <sys/mac.h>
51#include <sys/mutex.h>
52#include <sys/proc.h>
53#include <sys/kse.h>
54#include <sys/ktr.h>
55#include <sys/resourcevar.h>
56#include <sys/sched.h>
57#include <sys/signalvar.h>
58#include <sys/systm.h>
59#include <sys/vmmeter.h>
60#include <machine/cpu.h>
61#include <machine/pcb.h>
62
63/*
64 * Define the code needed before returning to user mode, for
65 * trap and syscall.
66 *
67 * MPSAFE
68 */
69void
70userret(td, frame, oticks)
71	struct thread *td;
72	struct trapframe *frame;
73	u_int oticks;
74{
75	struct proc *p = td->td_proc;
76
77	CTR3(KTR_SYSC, "userret: thread %p (pid %d, %s)", td, p->p_pid,
78            p->p_comm);
79#ifdef INVARIANTS
80	/* Check that we called signotify() enough. */
81	mtx_lock(&Giant);
82	PROC_LOCK(p);
83	mtx_lock_spin(&sched_lock);
84	if (SIGPENDING(p) && ((p->p_sflag & PS_NEEDSIGCHK) == 0 ||
85	    (td->td_flags & TDF_ASTPENDING) == 0))
86		printf("failed to set signal flags properly for ast()\n");
87	mtx_unlock_spin(&sched_lock);
88	PROC_UNLOCK(p);
89	mtx_unlock(&Giant);
90#endif
91
92	/*
93	 * Let the scheduler adjust our priority etc.
94	 */
95	sched_userret(td);
96
97	/*
98	 * We need to check to see if we have to exit or wait due to a
99	 * single threading requirement or some other STOP condition.
100	 * Don't bother doing all the work if the stop bits are not set
101	 * at this time.. If we miss it, we miss it.. no big deal.
102	 */
103	if (P_SHOULDSTOP(p)) {
104		PROC_LOCK(p);
105		thread_suspend_check(0);	/* Can suspend or kill */
106		PROC_UNLOCK(p);
107	}
108
109	/*
110	 * Do special thread processing, e.g. upcall tweaking and such.
111	 */
112	if (p->p_flag & P_KSES) {
113		thread_userret(td, frame);
114	}
115
116	/*
117	 * Charge system time if profiling.
118	 *
119	 * XXX should move PS_PROFIL to a place that can obviously be
120	 * accessed safely without sched_lock.
121	 */
122	if (p->p_sflag & PS_PROFIL) {
123		quad_t ticks;
124
125		mtx_lock_spin(&sched_lock);
126		ticks = td->td_sticks - oticks;
127		mtx_unlock_spin(&sched_lock);
128		addupc_task(td, TRAPF_PC(frame), (u_int)ticks * psratio);
129	}
130}
131
132/*
133 * Process an asynchronous software trap.
134 * This is relatively easy.
135 * This function will return with preemption disabled.
136 */
137void
138ast(struct trapframe *framep)
139{
140	struct thread *td;
141	struct proc *p;
142	struct kse *ke;
143	struct ksegrp *kg;
144	struct rlimit *rlim;
145	u_int prticks, sticks;
146	int sflag;
147	int flags;
148	int sig;
149#if defined(DEV_NPX) && !defined(SMP)
150	int ucode;
151#endif
152
153	td = curthread;
154	p = td->td_proc;
155	kg = td->td_ksegrp;
156
157	CTR3(KTR_SYSC, "ast: thread %p (pid %d, %s)", td, p->p_pid,
158            p->p_comm);
159	KASSERT(TRAPF_USERMODE(framep), ("ast in kernel mode"));
160#ifdef WITNESS
161	if (witness_list(td))
162		panic("Returning to user mode with mutex(s) held");
163#endif
164	mtx_assert(&Giant, MA_NOTOWNED);
165	mtx_assert(&sched_lock, MA_NOTOWNED);
166	td->td_frame = framep;
167
168	/*
169	 * This updates the p_sflag's for the checks below in one
170	 * "atomic" operation with turning off the astpending flag.
171	 * If another AST is triggered while we are handling the
172	 * AST's saved in sflag, the astpending flag will be set and
173	 * ast() will be called again.
174	 */
175	mtx_lock_spin(&sched_lock);
176	ke = td->td_kse;
177	sticks = td->td_sticks;
178	flags = td->td_flags;
179	sflag = p->p_sflag;
180	p->p_sflag &= ~(PS_ALRMPEND | PS_NEEDSIGCHK | PS_PROFPEND | PS_XCPU);
181#ifdef MAC
182	p->p_sflag &= ~PS_MACPEND;
183#endif
184	td->td_flags &= ~(TDF_ASTPENDING | TDF_NEEDRESCHED | TDF_OWEUPC);
185	cnt.v_soft++;
186	prticks = 0;
187	if (flags & TDF_OWEUPC && sflag & PS_PROFIL) {
188		prticks = p->p_stats->p_prof.pr_ticks;
189		p->p_stats->p_prof.pr_ticks = 0;
190	}
191	mtx_unlock_spin(&sched_lock);
192	/*
193	 * XXXKSE While the fact that we owe a user profiling
194	 * tick is stored per KSE in this code, the statistics
195	 * themselves are still stored per process.
196	 * This should probably change, by which I mean that
197	 * possibly the location of both might change.
198	 */
199
200	if (td->td_ucred != p->p_ucred)
201		cred_update_thread(td);
202	if (flags & TDF_OWEUPC && sflag & PS_PROFIL)
203		addupc_task(td, p->p_stats->p_prof.pr_addr, prticks);
204	if (sflag & PS_ALRMPEND) {
205		PROC_LOCK(p);
206		psignal(p, SIGVTALRM);
207		PROC_UNLOCK(p);
208	}
209#if defined(DEV_NPX) && !defined(SMP)
210	if (PCPU_GET(curpcb)->pcb_flags & PCB_NPXTRAP) {
211		atomic_clear_int(&PCPU_GET(curpcb)->pcb_flags,
212		    PCB_NPXTRAP);
213		ucode = npxtrap();
214		if (ucode != -1) {
215			trapsignal(p, SIGFPE, ucode);
216		}
217	}
218#endif
219	if (sflag & PS_PROFPEND) {
220		PROC_LOCK(p);
221		psignal(p, SIGPROF);
222		PROC_UNLOCK(p);
223	}
224	if (sflag & PS_XCPU) {
225		PROC_LOCK(p);
226		rlim = &p->p_rlimit[RLIMIT_CPU];
227		if (p->p_runtime.sec >= rlim->rlim_max)
228			killproc(p, "exceeded maximum CPU limit");
229		else {
230			psignal(p, SIGXCPU);
231			mtx_lock_spin(&sched_lock);
232			if (p->p_cpulimit < rlim->rlim_max)
233				p->p_cpulimit += 5;
234			mtx_unlock_spin(&sched_lock);
235		}
236		PROC_UNLOCK(p);
237	}
238#ifdef MAC
239	if (sflag & PS_MACPEND)
240		mac_thread_userret(td);
241#endif
242	if (flags & TDF_NEEDRESCHED) {
243		mtx_lock_spin(&sched_lock);
244		sched_prio(td, kg->kg_user_pri);
245		p->p_stats->p_ru.ru_nivcsw++;
246		mi_switch();
247		mtx_unlock_spin(&sched_lock);
248	}
249	if (sflag & PS_NEEDSIGCHK) {
250		int sigs;
251
252		sigs = 0;
253		PROC_LOCK(p);
254		while ((sig = cursig(td)) != 0) {
255			postsig(sig);
256			sigs++;
257		}
258		PROC_UNLOCK(p);
259		if (p->p_flag & P_KSES && sigs)
260			thread_signal_upcall(td);
261	}
262
263	userret(td, framep, sticks);
264#ifdef DIAGNOSTIC
265	cred_free_thread(td);
266#endif
267	mtx_assert(&Giant, MA_NOTOWNED);
268}
269