subr_syscall.c revision 110140
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 110140 2003-01-31 11:22:31Z tjr $
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#ifdef INVARIANTS
77	struct kse *ke;
78#endif
79	u_int64_t eticks;
80
81	CTR3(KTR_SYSC, "userret: thread %p (pid %d, %s)", td, p->p_pid,
82            p->p_comm);
83#ifdef INVARIANTS
84	/*
85	 * Check that we called signotify() enough.
86	 * XXXKSE this checking is bogus for threaded program,
87	 */
88	mtx_lock(&Giant);
89	PROC_LOCK(p);
90	mtx_lock_spin(&sched_lock);
91	ke = td->td_kse;
92	if (SIGPENDING(p) && ((p->p_sflag & PS_NEEDSIGCHK) == 0 ||
93	    (td->td_kse->ke_flags & KEF_ASTPENDING) == 0))
94		printf("failed to set signal flags properly for ast()\n");
95	mtx_unlock_spin(&sched_lock);
96	PROC_UNLOCK(p);
97	mtx_unlock(&Giant);
98#endif
99
100	/*
101	 * Let the scheduler adjust our priority etc.
102	 */
103	sched_userret(td);
104
105	/*
106	 * Charge system time if profiling.
107	 *
108	 * XXX should move PS_PROFIL to a place that can obviously be
109	 * accessed safely without sched_lock.
110	 */
111
112	if (p->p_sflag & PS_PROFIL) {
113		eticks = td->td_sticks - oticks;
114		addupc_task(td, TRAPF_PC(frame), (u_int)eticks * psratio);
115	}
116
117	/*
118	 * We need to check to see if we have to exit or wait due to a
119	 * single threading requirement or some other STOP condition.
120	 * Don't bother doing all the work if the stop bits are not set
121	 * at this time.. If we miss it, we miss it.. no big deal.
122	 */
123	if (P_SHOULDSTOP(p)) {
124		PROC_LOCK(p);
125		thread_suspend_check(0);	/* Can suspend or kill */
126		PROC_UNLOCK(p);
127	}
128
129	/*
130	 * Do special thread processing, e.g. upcall tweaking and such.
131	 */
132	if (p->p_flag & P_KSES) {
133		thread_userret(td, frame);
134	}
135}
136
137/*
138 * Process an asynchronous software trap.
139 * This is relatively easy.
140 * This function will return with preemption disabled.
141 */
142void
143ast(struct trapframe *framep)
144{
145	struct thread *td;
146	struct proc *p;
147	struct kse *ke;
148	struct ksegrp *kg;
149	struct rlimit *rlim;
150	u_int prticks, sticks;
151	int sflag;
152	int flags;
153	int tflags;
154	int sig;
155#if defined(DEV_NPX) && !defined(SMP)
156	int ucode;
157#endif
158
159	td = curthread;
160	p = td->td_proc;
161	kg = td->td_ksegrp;
162
163	CTR3(KTR_SYSC, "ast: thread %p (pid %d, %s)", td, p->p_pid,
164            p->p_comm);
165	KASSERT(TRAPF_USERMODE(framep), ("ast in kernel mode"));
166#ifdef WITNESS
167	if (witness_list(td))
168		panic("Returning to user mode with mutex(s) held");
169#endif
170	mtx_assert(&Giant, MA_NOTOWNED);
171	mtx_assert(&sched_lock, MA_NOTOWNED);
172	td->td_frame = framep;
173
174	/*
175	 * This updates the p_sflag's for the checks below in one
176	 * "atomic" operation with turning off the astpending flag.
177	 * If another AST is triggered while we are handling the
178	 * AST's saved in sflag, the astpending flag will be set and
179	 * ast() will be called again.
180	 */
181	mtx_lock_spin(&sched_lock);
182	ke = td->td_kse;
183	sticks = td->td_sticks;
184	tflags = td->td_flags;
185	flags = ke->ke_flags;
186	sflag = p->p_sflag;
187	p->p_sflag &= ~(PS_ALRMPEND | PS_NEEDSIGCHK | PS_PROFPEND | PS_XCPU);
188#ifdef MAC
189	p->p_sflag &= ~PS_MACPEND;
190#endif
191	ke->ke_flags &= ~(KEF_ASTPENDING | KEF_NEEDRESCHED);
192	td->td_flags &= ~(TDF_ASTPENDING | TDF_OWEUPC);
193	cnt.v_soft++;
194	prticks = 0;
195	if (tflags & TDF_OWEUPC && sflag & PS_PROFIL) {
196		prticks = td->td_prticks;
197		td->td_prticks = 0;
198	}
199	mtx_unlock_spin(&sched_lock);
200	/*
201	 * XXXKSE While the fact that we owe a user profiling
202	 * tick is stored per KSE in this code, the statistics
203	 * themselves are still stored per process.
204	 * This should probably change, by which I mean that
205	 * possibly the location of both might change.
206	 */
207
208	if (td->td_ucred != p->p_ucred)
209		cred_update_thread(td);
210	if (tflags & TDF_OWEUPC && sflag & PS_PROFIL) {
211		addupc_task(td, td->td_praddr, prticks);
212	}
213	if (sflag & PS_ALRMPEND) {
214		PROC_LOCK(p);
215		psignal(p, SIGVTALRM);
216		PROC_UNLOCK(p);
217	}
218#if defined(DEV_NPX) && !defined(SMP)
219	if (PCPU_GET(curpcb)->pcb_flags & PCB_NPXTRAP) {
220		atomic_clear_int(&PCPU_GET(curpcb)->pcb_flags,
221		    PCB_NPXTRAP);
222		ucode = npxtrap();
223		if (ucode != -1) {
224			trapsignal(p, SIGFPE, ucode);
225		}
226	}
227#endif
228	if (sflag & PS_PROFPEND) {
229		PROC_LOCK(p);
230		psignal(p, SIGPROF);
231		PROC_UNLOCK(p);
232	}
233	if (sflag & PS_XCPU) {
234		PROC_LOCK(p);
235		rlim = &p->p_rlimit[RLIMIT_CPU];
236		if (p->p_runtime.sec >= rlim->rlim_max)
237			killproc(p, "exceeded maximum CPU limit");
238		else {
239			psignal(p, SIGXCPU);
240			mtx_lock_spin(&sched_lock);
241			if (p->p_cpulimit < rlim->rlim_max)
242				p->p_cpulimit += 5;
243			mtx_unlock_spin(&sched_lock);
244		}
245		PROC_UNLOCK(p);
246	}
247#ifdef MAC
248	if (sflag & PS_MACPEND)
249		mac_thread_userret(td);
250#endif
251	if (flags & KEF_NEEDRESCHED) {
252		mtx_lock_spin(&sched_lock);
253		sched_prio(td, kg->kg_user_pri);
254		p->p_stats->p_ru.ru_nivcsw++;
255		mi_switch();
256		mtx_unlock_spin(&sched_lock);
257	}
258	if (sflag & PS_NEEDSIGCHK) {
259		PROC_LOCK(p);
260		while ((sig = cursig(td)) != 0)
261			postsig(sig);
262		PROC_UNLOCK(p);
263	}
264
265	userret(td, framep, sticks);
266#ifdef DIAGNOSTIC
267	cred_free_thread(td);
268#endif
269	mtx_assert(&Giant, MA_NOTOWNED);
270}
271