process.c

linux内核源码

/*
 * Copyright (C) 2002- 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com)
 * Licensed under the GPL
 */

#include <stdlib.h>
#include <unistd.h>
#include <sched.h>
#include <errno.h>
#include <string.h>
#include <sys/mman.h>
#include <sys/ptrace.h>
#include <sys/wait.h>
#include <asm/unistd.h>
#include "as-layout.h"
#include "chan_user.h"
#include "kern_constants.h"
#include "mem.h"
#include "os.h"
#include "process.h"
#include "proc_mm.h"
#include "ptrace_user.h"
#include "registers.h"
#include "skas.h"
#include "skas_ptrace.h"
#include "user.h"
#include "sysdep/stub.h"

int is_skas_winch(int pid, int fd, void *data)
{
	if (pid != getpgrp())
		return 0;

	register_winch_irq(-1, fd, -1, data, 0);
	return 1;
}

static int ptrace_dump_regs(int pid)
{
        unsigned long regs[MAX_REG_NR];
        int i;

        if (ptrace(PTRACE_GETREGS, pid, 0, regs) < 0)
                return -errno;

	printk(UM_KERN_ERR "Stub registers -\n");
	for (i = 0; i < ARRAY_SIZE(regs); i++)
		printk(UM_KERN_ERR "\t%d - %lx\n", i, regs[i]);

        return 0;
}

/*
 * Signals that are OK to receive in the stub - we'll just continue it.
 * SIGWINCH will happen when UML is inside a detached screen.
 */
#define STUB_SIG_MASK ((1 << SIGVTALRM) | (1 << SIGWINCH))

/* Signals that the stub will finish with - anything else is an error */
#define STUB_DONE_MASK ((1 << SIGUSR1) | (1 << SIGTRAP))

void wait_stub_done(int pid)
{
	int n, status, err;

	while (1) {
		CATCH_EINTR(n = waitpid(pid, &status, WUNTRACED | __WALL));
		if ((n < 0) || !WIFSTOPPED(status))
			goto bad_wait;

		if (((1 << WSTOPSIG(status)) & STUB_SIG_MASK) == 0)
			break;

		err = ptrace(PTRACE_CONT, pid, 0, 0);
		if (err)
			panic("wait_stub_done : continue failed, errno = %d\n",
			      errno);
	}

	if (((1 << WSTOPSIG(status)) & STUB_DONE_MASK) != 0)
		return;

bad_wait:
	err = ptrace_dump_regs(pid);
	if (err)
		printk(UM_KERN_ERR "Failed to get registers from stub, "
		       "errno = %d\n", -err);
	panic("wait_stub_done : failed to wait for SIGUSR1/SIGTRAP, pid = %d, "
	      "n = %d, errno = %d, status = 0x%x\n", pid, n, errno, status);
}

extern unsigned long current_stub_stack(void);

void get_skas_faultinfo(int pid, struct faultinfo * fi)
{
	int err;

	if (ptrace_faultinfo) {
		err = ptrace(PTRACE_FAULTINFO, pid, 0, fi);
		if (err)
			panic("get_skas_faultinfo - PTRACE_FAULTINFO failed, "
			      "errno = %d\n", errno);

		/* Special handling for i386, which has different structs */
		if (sizeof(struct ptrace_faultinfo) < sizeof(struct faultinfo))
			memset((char *)fi + sizeof(struct ptrace_faultinfo), 0,
			       sizeof(struct faultinfo) -
			       sizeof(struct ptrace_faultinfo));
	}
	else {
		err = ptrace(PTRACE_CONT, pid, 0, SIGSEGV);
		if (err)
			panic("Failed to continue stub, pid = %d, errno = %d\n",
			      pid, errno);
		wait_stub_done(pid);

		/*
		 * faultinfo is prepared by the stub-segv-handler at start of
		 * the stub stack page. We just have to copy it.
		 */
		memcpy(fi, (void *)current_stub_stack(), sizeof(*fi));
	}
}

static void handle_segv(int pid, struct uml_pt_regs * regs)
{
	get_skas_faultinfo(pid, &regs->faultinfo);
	segv(regs->faultinfo, 0, 1, NULL);
}

/*
 * To use the same value of using_sysemu as the caller, ask it that value
 * (in local_using_sysemu
 */
static void handle_trap(int pid, struct uml_pt_regs *regs,
			int local_using_sysemu)
{
	int err, status;

	/* Mark this as a syscall */
	UPT_SYSCALL_NR(regs) = PT_SYSCALL_NR(regs->gp);

	if (!local_using_sysemu)
	{
		err = ptrace(PTRACE_POKEUSR, pid, PT_SYSCALL_NR_OFFSET,
			     __NR_getpid);
		if (err < 0)
			panic("handle_trap - nullifying syscall failed, "
			      "errno = %d\n", errno);

		err = ptrace(PTRACE_SYSCALL, pid, 0, 0);
		if (err < 0)
			panic("handle_trap - continuing to end of syscall "
			      "failed, errno = %d\n", errno);

		CATCH_EINTR(err = waitpid(pid, &status, WUNTRACED | __WALL));
		if ((err < 0) || !WIFSTOPPED(status) ||
		   (WSTOPSIG(status) != SIGTRAP + 0x80)) {
                        err = ptrace_dump_regs(pid);
                        if (err)
                                printk(UM_KERN_ERR "Failed to get registers "
				       "from process, errno = %d\n", -err);
			panic("handle_trap - failed to wait at end of syscall, "
			      "errno = %d, status = %d\n", errno, status);
                }
	}

	handle_syscall(regs);
}

extern int __syscall_stub_start;

static int userspace_tramp(void *stack)
{
	void *addr;
	int err;

	ptrace(PTRACE_TRACEME, 0, 0, 0);

	signal(SIGTERM, SIG_DFL);
	err = set_interval();
	if (err)
		panic("userspace_tramp - setting timer failed, errno = %d\n",
		      err);

	if (!proc_mm) {
		/*
		 * This has a pte, but it can't be mapped in with the usual
		 * tlb_flush mechanism because this is part of that mechanism
		 */
		int fd;
		unsigned long long offset;
		fd = phys_mapping(to_phys(&__syscall_stub_start), &offset);
		addr = mmap64((void *) STUB_CODE, UM_KERN_PAGE_SIZE,
			      PROT_EXEC, MAP_FIXED | MAP_PRIVATE, fd, offset);
		if (addr == MAP_FAILED) {
			printk(UM_KERN_ERR "mapping mmap stub at 0x%lx failed, "
			       "errno = %d\n", STUB_CODE, errno);
			exit(1);
		}

		if (stack != NULL) {
			fd = phys_mapping(to_phys(stack), &offset);
			addr = mmap((void *) STUB_DATA,
				    UM_KERN_PAGE_SIZE, PROT_READ | PROT_WRITE,
				    MAP_FIXED | MAP_SHARED, fd, offset);
			if (addr == MAP_FAILED) {
				printk(UM_KERN_ERR "mapping segfault stack "
				       "at 0x%lx failed, errno = %d\n",
				       STUB_DATA, errno);
				exit(1);
			}
		}
	}
	if (!ptrace_faultinfo && (stack != NULL)) {
		struct sigaction sa;

		unsigned long v = STUB_CODE +
				  (unsigned long) stub_segv_handler -
				  (unsigned long) &__syscall_stub_start;

		set_sigstack((void *) STUB_DATA, UM_KERN_PAGE_SIZE);
		sigemptyset(&sa.sa_mask);
		sigaddset(&sa.sa_mask, SIGIO);
		sigaddset(&sa.sa_mask, SIGWINCH);
		sigaddset(&sa.sa_mask, SIGVTALRM);
		sigaddset(&sa.sa_mask, SIGUSR1);
		sa.sa_flags = SA_ONSTACK;
		sa.sa_handler = (void *) v;
		sa.sa_restorer = NULL;
		if (sigaction(SIGSEGV, &sa, NULL) < 0)
			panic("userspace_tramp - setting SIGSEGV handler "
			      "failed - errno = %d\n", errno);
	}

	kill(os_getpid(), SIGSTOP);
	return 0;
}

/* Each element set once, and only accessed by a single processor anyway */
#undef NR_CPUS
#define NR_CPUS 1
int userspace_pid[NR_CPUS];

int start_userspace(unsigned long stub_stack)
{
	void *stack;
	unsigned long sp;
	int pid, status, n, flags;

	stack = mmap(NULL, UM_KERN_PAGE_SIZE,
		     PROT_READ | PROT_WRITE | PROT_EXEC,
		     MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
	if (stack == MAP_FAILED)
		panic("start_userspace : mmap failed, errno = %d", errno);
	sp = (unsigned long) stack + UM_KERN_PAGE_SIZE - sizeof(void *);

	flags = CLONE_FILES;
	if (proc_mm)
		flags |= CLONE_VM;
	else
		flags |= SIGCHLD;

	pid = clone(userspace_tramp, (void *) sp, flags, (void *) stub_stack);
	if (pid < 0)
		panic("start_userspace : clone failed, errno = %d", errno);

	do {
		CATCH_EINTR(n = waitpid(pid, &status, WUNTRACED | __WALL));
		if (n < 0)
			panic("start_userspace : wait failed, errno = %d",
			      errno);
	} while (WIFSTOPPED(status) && (WSTOPSIG(status) == SIGVTALRM));

	if (!WIFSTOPPED(status) || (WSTOPSIG(status) != SIGSTOP))
		panic("start_userspace : expected SIGSTOP, got status = %d",
		      status);

	if (ptrace(PTRACE_OLDSETOPTIONS, pid, NULL,
		   (void *) PTRACE_O_TRACESYSGOOD) < 0)
		panic("start_userspace : PTRACE_OLDSETOPTIONS failed, "
		      "errno = %d\n", errno);

	if (munmap(stack, UM_KERN_PAGE_SIZE) < 0)
		panic("start_userspace : munmap failed, errno = %d\n", errno);

	return pid;
}

void userspace(struct uml_pt_regs *regs)
{
	struct itimerval timer;
	unsigned long long nsecs, now;
	int err, status, op, pid = userspace_pid[0];
	/* To prevent races if using_sysemu changes under us.*/
	int local_using_sysemu;

	if (getitimer(ITIMER_VIRTUAL, &timer))
		printk("Failed to get itimer, errno = %d\n", errno);
	nsecs = timer.it_value.tv_sec * UM_NSEC_PER_SEC +
		timer.it_value.tv_usec * UM_NSEC_PER_USEC;
	nsecs += os_nsecs();

	while (1) {
		restore_registers(pid, regs);

		/* Now we set local_using_sysemu to be used for one loop */
		local_using_sysemu = get_using_sysemu();

		op = SELECT_PTRACE_OPERATION(local_using_sysemu,
					     singlestepping(NULL));