xref: /freebsd-10.2-release/sys/kern/subr_syscall.c

/*-
 * Copyright (c) 1990 The Regents of the University of California.
 * All rights reserved.
 *
 * This code is derived from software contributed to Berkeley by
 * the University of Utah, and William Jolitz.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 * 3. All advertising materials mentioning features or use of this software
 *    must display the following acknowledgement:
 *	This product includes software developed by the University of
 *	California, Berkeley and its contributors.
 * 4. Neither the name of the University nor the names of its contributors
 *    may be used to endorse or promote products derived from this software
 *    without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 *
 *	@(#)trap.c	7.4 (Berkeley) 5/13/91
 *
 * PATCHES MAGIC                LEVEL   PATCH THAT GOT US HERE
 * --------------------         -----   ----------------------
 * CURRENT PATCH LEVEL:         1       00137
 * --------------------         -----   ----------------------
 *
 * 08 Apr 93	Bruce Evans		Several VM system fixes
 * 		Paul Kranenburg		Add counter for vmstat
 */
static char rcsid[] = "$Header: /a/cvs/386BSD/src/sys/i386/i386/trap.c,v 1.1.1.1 1993/06/12 14:58:05 rgrimes Exp $";

/*
 * 386 Trap and System call handleing
 */

#include "machine/cpu.h"
#include "machine/psl.h"
#include "machine/reg.h"

#include "param.h"
#include "systm.h"
#include "proc.h"
#include "user.h"
#include "acct.h"
#include "kernel.h"
#ifdef KTRACE
#include "ktrace.h"
#endif

#include "vm/vm_param.h"
#include "vm/pmap.h"
#include "vm/vm_map.h"
#include "sys/vmmeter.h"

#include "machine/trap.h"

#ifdef	__GNUC__

/*
 * The "r" contraint could be "rm" except for fatal bugs in gas.  As usual,
 * we omit the size from the mov instruction to avoid nonfatal bugs in gas.
 */
#define	read_gs()	({ u_short gs; __asm("mov %%gs,%0" : "=r" (gs)); gs; })
#define	write_gs(gs)	__asm("mov %0,%%gs" : : "r" ((u_short) gs))

#else	/* not __GNUC__ */

u_short	read_gs		__P((void));
void	write_gs	__P((/* promoted u_short */ int gs));

#endif	/* __GNUC__ */

struct	sysent sysent[];
int	nsysent;
int dostacklimits;
unsigned rcr2();
extern short cpl;


/*
 * trap(frame):
 *	Exception, fault, and trap interface to BSD kernel. This
 * common code is called from assembly language IDT gate entry
 * routines that prepare a suitable stack frame, and restore this
 * frame after the exception has been processed. Note that the
 * effect is as if the arguments were passed call by reference.
 */

/*ARGSUSED*/
trap(frame)
	struct trapframe frame;
{
	register int i;
	register struct proc *p = curproc;
	struct timeval syst;
	int ucode, type, code, eva;

	frame.tf_eflags &= ~PSL_NT;	/* clear nested trap XXX */
	type = frame.tf_trapno;
#include "ddb.h"
#if NDDB > 0
	if (curpcb && curpcb->pcb_onfault) {
		if (frame.tf_trapno == T_BPTFLT
		    || frame.tf_trapno == T_TRCTRAP)
			if (kdb_trap (type, 0, &frame))
				return;
	}
#endif

/*pg("trap type %d code = %x eip = %x cs = %x eva = %x esp %x",
			frame.tf_trapno, frame.tf_err, frame.tf_eip,
			frame.tf_cs, rcr2(), frame.tf_esp);*/
if(curpcb == 0 || curproc == 0) goto we_re_toast;
	if (curpcb->pcb_onfault && frame.tf_trapno != T_PAGEFLT) {
		extern int _udatasel;

		if (read_gs() != (u_short) _udatasel)
			/*
			 * Some user has corrupted %gs but we depend on it in
			 * copyout() etc.  Fix it up and retry.
			 *
			 * (We don't preserve %fs or %gs, so users can change
			 * them to either _ucodesel, _udatasel or a not-present
			 * selector, possibly ORed with 0 to 3, making them
			 * volatile for other users.  Not preserving them saves
			 * time and doesn't lose functionality or open security
			 * holes.)
			 */
			write_gs(_udatasel);
		else
copyfault:
			frame.tf_eip = (int)curpcb->pcb_onfault;
		return;
	}

	syst = p->p_stime;
	if (ISPL(frame.tf_cs) == SEL_UPL) {
		type |= T_USER;
		p->p_regs = (int *)&frame;
		curpcb->pcb_flags |= FM_TRAP;	/* used by sendsig */
	}

	ucode=0;
	eva = rcr2();
	code = frame.tf_err;
	switch (type) {

	default:
	we_re_toast:
#ifdef KDB
		if (kdb_trap(&psl))
			return;
#endif
#if NDDB > 0
		if (kdb_trap (type, 0, &frame))
			return;
#endif

		printf("trap type %d code = %x eip = %x cs = %x eflags = %x ",
			frame.tf_trapno, frame.tf_err, frame.tf_eip,
			frame.tf_cs, frame.tf_eflags);
	eva = rcr2();
		printf("cr2 %x cpl %x\n", eva, cpl);
		/* type &= ~T_USER; */ /* XXX what the hell is this */
		panic("trap");
		/*NOTREACHED*/

	case T_SEGNPFLT|T_USER:
	case T_STKFLT|T_USER:
	case T_PROTFLT|T_USER:		/* protection fault */
		ucode = code + BUS_SEGM_FAULT ;
		i = SIGBUS;
		break;

	case T_PRIVINFLT|T_USER:	/* privileged instruction fault */
	case T_RESADFLT|T_USER:		/* reserved addressing fault */
	case T_RESOPFLT|T_USER:		/* reserved operand fault */
	case T_FPOPFLT|T_USER:		/* coprocessor operand fault */
		ucode = type &~ T_USER;
		i = SIGILL;
		break;

	case T_ASTFLT|T_USER:		/* Allow process switch */
		astoff();
		cnt.v_soft++;
		if ((p->p_flag & SOWEUPC) && p->p_stats->p_prof.pr_scale) {
			addupc(frame.tf_eip, &p->p_stats->p_prof, 1);
			p->p_flag &= ~SOWEUPC;
		}
		goto out;

	case T_DNA|T_USER:
#ifdef	NPX
		/* if a transparent fault (due to context switch "late") */
		if (npxdna()) return;
#endif
		i = math_emulate(&frame);
		if (i == 0) return;
		ucode = FPE_FPU_NP_TRAP;
		break;

	case T_BOUND|T_USER:
		ucode = FPE_SUBRNG_TRAP;
		i = SIGFPE;
		break;

	case T_OFLOW|T_USER:
		ucode = FPE_INTOVF_TRAP;
		i = SIGFPE;
		break;

	case T_DIVIDE|T_USER:
		ucode = FPE_INTDIV_TRAP;
		i = SIGFPE;
		break;

	case T_ARITHTRAP|T_USER:
		ucode = code;
		i = SIGFPE;
		break;

	case T_PAGEFLT:			/* allow page faults in kernel mode */
#if 0
		/* XXX - check only applies to 386's and 486's with WP off */
		if (code & PGEX_P) goto we_re_toast;
#endif

		/* fall into */
	case T_PAGEFLT|T_USER:		/* page fault */
	    {
		register vm_offset_t va;
		register struct vmspace *vm = p->p_vmspace;
		register vm_map_t map;
		int rv;
		vm_prot_t ftype;
		extern vm_map_t kernel_map;
		unsigned nss,v;

		va = trunc_page((vm_offset_t)eva);
		/*
		 * Avoid even looking at pde_v(va) for high va's.   va's
		 * above VM_MAX_KERNEL_ADDRESS don't correspond to normal
		 * PDE's (half of them correspond to APDEpde and half to
		 * an unmapped kernel PDE).  va's betweeen 0xFEC00000 and
		 * VM_MAX_KERNEL_ADDRESS correspond to unmapped kernel PDE's
		 * (XXX - why are only 3 initialized when 6 are required to
		 * reach VM_MAX_KERNEL_ADDRESS?).  Faulting in an unmapped
		 * kernel page table would give inconsistent PTD's.
		 *
		 * XXX - faulting in unmapped page tables wastes a page if
		 * va turns out to be invalid.
		 *
		 * XXX - should "kernel address space" cover the kernel page
		 * tables?  Might have same problem with PDEpde as with
		 * APDEpde (or there may be no problem with APDEpde).
		 */
		if (va > 0xFEBFF000) {
			rv = KERN_FAILURE;	/* becomes SIGBUS */
			goto nogo;
		}
		/*
		 * It is only a kernel address space fault iff:
		 * 	1. (type & T_USER) == 0  and
		 * 	2. pcb_onfault not set or
		 *	3. pcb_onfault set but supervisor space fault
		 * The last can occur during an exec() copyin where the
		 * argument space is lazy-allocated.
		 */
		if (type == T_PAGEFLT && va >= KERNBASE)
			map = kernel_map;
		else
			map = &vm->vm_map;
		if (code & PGEX_W)
			ftype = VM_PROT_READ | VM_PROT_WRITE;
		else
			ftype = VM_PROT_READ;

#ifdef DEBUG
		if (map == kernel_map && va == 0) {
			printf("trap: bad kernel access at %x\n", va);
			goto we_re_toast;
		}
#endif

		/*
		 * XXX: rude hack to make stack limits "work"
		 */
		nss = 0;
		if ((caddr_t)va >= vm->vm_maxsaddr && map != kernel_map
			&& dostacklimits) {
			nss = clrnd(btoc((unsigned)vm->vm_maxsaddr
				+ MAXSSIZ - (unsigned)va));
			if (nss > btoc(p->p_rlimit[RLIMIT_STACK].rlim_cur)) {
/*pg("trap rlimit %d, maxsaddr %x va %x ", nss, vm->vm_maxsaddr, va);*/
				rv = KERN_FAILURE;
				goto nogo;
			}
		}

		/* check if page table is mapped, if not, fault it first */
#define pde_v(v) (PTD[((v)>>PD_SHIFT)&1023].pd_v)
		if (!pde_v(va)) {
			v = trunc_page(vtopte(va));
			rv = vm_fault(map, v, ftype, FALSE);
			if (rv != KERN_SUCCESS) goto nogo;
			/* check if page table fault, increment wiring */
			vm_map_pageable(map, v, round_page(v+1), FALSE);
		} else v=0;
		rv = vm_fault(map, va, ftype, FALSE);
		if (rv == KERN_SUCCESS) {
			/*
			 * XXX: continuation of rude stack hack
			 */
			if (nss > vm->vm_ssize)
				vm->vm_ssize = nss;
			va = trunc_page(vtopte(va));
			/* for page table, increment wiring
			   as long as not a page table fault as well */
			if (!v && type != T_PAGEFLT)
			  vm_map_pageable(map, va, round_page(va+1), FALSE);
			if (type == T_PAGEFLT)
				return;
			goto out;
		}
nogo:
		if (type == T_PAGEFLT) {
			if (curpcb->pcb_onfault)
				goto copyfault;
			printf("vm_fault(%x, %x, %x, 0) -> %x\n",
			       map, va, ftype, rv);
			printf("  type %x, code %x\n",
			       type, code);
			goto we_re_toast;
		}
		i = (rv == KERN_PROTECTION_FAILURE) ? SIGBUS : SIGSEGV;
		break;
	    }

#if NDDB == 0
	case T_TRCTRAP:	 /* trace trap -- someone single stepping lcall's */
		frame.tf_eflags &= ~PSL_T;

			/* Q: how do we turn it on again? */
		return;
#endif

	case T_BPTFLT|T_USER:		/* bpt instruction fault */
	case T_TRCTRAP|T_USER:		/* trace trap */
		frame.tf_eflags &= ~PSL_T;
		i = SIGTRAP;
		break;

#include "isa.h"
#if	NISA > 0
	case T_NMI:
	case T_NMI|T_USER:
#if NDDB > 0
		/* NMI can be hooked up to a pushbutton for debugging */
		printf ("NMI ... going to debugger\n");
		if (kdb_trap (type, 0, &frame))
			return;
#endif
		/* machine/parity/power fail/"kitchen sink" faults */
		if(isa_nmi(code) == 0) return;
		else goto we_re_toast;
#endif
	}

	trapsignal(p, i, ucode);
	if ((type & T_USER) == 0)
		return;
out:
	while (i = CURSIG(p))
		psig(i);
	p->p_pri = p->p_usrpri;
	if (want_resched) {
		/*
		 * Since we are curproc, clock will normally just change
		 * our priority without moving us from one queue to another
		 * (since the running process is not on a queue.)
		 * If that happened after we setrq ourselves but before we
		 * swtch()'ed, we might not be on the queue indicated by
		 * our priority.
		 */
		(void) splclock();
		setrq(p);
		p->p_stats->p_ru.ru_nivcsw++;
		swtch();
		(void) splnone();
		while (i = CURSIG(p))
			psig(i);
	}
	if (p->p_stats->p_prof.pr_scale) {
		int ticks;
		struct timeval *tv = &p->p_stime;

		ticks = ((tv->tv_sec - syst.tv_sec) * 1000 +
			(tv->tv_usec - syst.tv_usec) / 1000) / (tick / 1000);
		if (ticks) {
#ifdef PROFTIMER
			extern int profscale;
			addupc(frame.tf_eip, &p->p_stats->p_prof,
			    ticks * profscale);
#else
			addupc(frame.tf_eip, &p->p_stats->p_prof, ticks);
#endif
		}
	}
	curpri = p->p_pri;
	curpcb->pcb_flags &= ~FM_TRAP;	/* used by sendsig */
}

/*
 * Compensate for 386 brain damage (missing URKR).
 * This is a little simpler than the pagefault handler in trap() because
 * it the page tables have already been faulted in and high addresses
 * are thrown out early for other reasons.
 */
int trapwrite(addr)
	unsigned addr;
{
	unsigned nss;
	struct proc *p;
	vm_offset_t va;
	struct vmspace *vm;

	va = trunc_page((vm_offset_t)addr);
	/*
	 * XXX - MAX is END.  Changed > to >= for temp. fix.
	 */
	if (va >= VM_MAXUSER_ADDRESS)
		return (1);
	/*
	 * XXX: rude stack hack adapted from trap().
	 */
	nss = 0;
	p = curproc;
	vm = p->p_vmspace;
	if ((caddr_t)va >= vm->vm_maxsaddr && dostacklimits) {
		nss = clrnd(btoc((unsigned)vm->vm_maxsaddr + MAXSSIZ
				 - (unsigned)va));
		if (nss > btoc(p->p_rlimit[RLIMIT_STACK].rlim_cur))
			return (1);
	}

	if (vm_fault(&vm->vm_map, va, VM_PROT_READ | VM_PROT_WRITE, FALSE)
	    != KERN_SUCCESS)
		return (1);

	/*
	 * XXX: continuation of rude stack hack
	 */
	if (nss > vm->vm_ssize)
		vm->vm_ssize = nss;

	return (0);
}

/*
 * syscall(frame):
 *	System call request from POSIX system call gate interface to kernel.
 * Like trap(), argument is call by reference.
 */
/*ARGSUSED*/
syscall(frame)
	volatile struct syscframe frame;
{
	register int *locr0 = ((int *)&frame);
	register caddr_t params;
	register int i;
	register struct sysent *callp;
	register struct proc *p = curproc;
	struct timeval syst;
	int error, opc;
	int args[8], rval[2];
	int code;

#ifdef lint
	r0 = 0; r0 = r0; r1 = 0; r1 = r1;
#endif
	syst = p->p_stime;
	if (ISPL(frame.sf_cs) != SEL_UPL)
		panic("syscall");

	code = frame.sf_eax;
	curpcb->pcb_flags &= ~FM_TRAP;	/* used by sendsig */
	p->p_regs = (int *)&frame;
	params = (caddr_t)frame.sf_esp + sizeof (int) ;

	/*
	 * Reconstruct pc, assuming lcall $X,y is 7 bytes, as it is always.
	 */
	opc = frame.sf_eip - 7;
	callp = (code >= nsysent) ? &sysent[63] : &sysent[code];
	if (callp == sysent) {
		i = fuword(params);
		params += sizeof (int);
		callp = (code >= nsysent) ? &sysent[63] : &sysent[code];
	}

	if ((i = callp->sy_narg * sizeof (int)) &&
	    (error = copyin(params, (caddr_t)args, (u_int)i))) {
		frame.sf_eax = error;
		frame.sf_eflags |= PSL_C;	/* carry bit */
#ifdef KTRACE
		if (KTRPOINT(p, KTR_SYSCALL))
			ktrsyscall(p->p_tracep, code, callp->sy_narg, &args);
#endif
		goto done;
	}
#ifdef KTRACE
	if (KTRPOINT(p, KTR_SYSCALL))
		ktrsyscall(p->p_tracep, code, callp->sy_narg, &args);
#endif
	rval[0] = 0;
	rval[1] = frame.sf_edx;
/*pg("%d. s %d\n", p->p_pid, code);*/
	error = (*callp->sy_call)(p, args, rval);
	if (error == ERESTART)
		frame.sf_eip = opc;
	else if (error != EJUSTRETURN) {
		if (error) {
/*pg("error %d", error);*/
			frame.sf_eax = error;
			frame.sf_eflags |= PSL_C;	/* carry bit */
		} else {
			frame.sf_eax = rval[0];
			frame.sf_edx = rval[1];
			frame.sf_eflags &= ~PSL_C;	/* carry bit */
		}
	}
	/* else if (error == EJUSTRETURN) */
		/* nothing to do */
done:
	/*
	 * Reinitialize proc pointer `p' as it may be different
	 * if this is a child returning from fork syscall.
	 */
	p = curproc;
	while (i = CURSIG(p))
		psig(i);
	p->p_pri = p->p_usrpri;
	if (want_resched) {
		/*
		 * Since we are curproc, clock will normally just change
		 * our priority without moving us from one queue to another
		 * (since the running process is not on a queue.)
		 * If that happened after we setrq ourselves but before we
		 * swtch()'ed, we might not be on the queue indicated by
		 * our priority.
		 */
		(void) splclock();
		setrq(p);
		p->p_stats->p_ru.ru_nivcsw++;
		swtch();
		(void) splnone();
		while (i = CURSIG(p))
			psig(i);
	}
	if (p->p_stats->p_prof.pr_scale) {
		int ticks;
		struct timeval *tv = &p->p_stime;

		ticks = ((tv->tv_sec - syst.tv_sec) * 1000 +
			(tv->tv_usec - syst.tv_usec) / 1000) / (tick / 1000);
		if (ticks) {
#ifdef PROFTIMER
			extern int profscale;
			addupc(frame.sf_eip, &p->p_stats->p_prof,
			    ticks * profscale);
#else
			addupc(frame.sf_eip, &p->p_stats->p_prof, ticks);
#endif
		}
	}
	curpri = p->p_pri;
#ifdef KTRACE
	if (KTRPOINT(p, KTR_SYSRET))
		ktrsysret(p->p_tracep, code, error, rval[0]);
#endif
#ifdef	DIAGNOSTICx
{ extern int _udatasel, _ucodesel;
	if (frame.sf_ss != _udatasel)
		printf("ss %x call %d\n", frame.sf_ss, code);
	if ((frame.sf_cs&0xffff) != _ucodesel)
		printf("cs %x call %d\n", frame.sf_cs, code);
	if (frame.sf_eip > VM_MAXUSER_ADDRESS) {
		printf("eip %x call %d\n", frame.sf_eip, code);
		frame.sf_eip = 0;
	}
}
#endif
}