Add preliminary support for x86-64 Linux binaries.

Differential Revision:	https://reviews.freebsd.org/D1076

sys/amd64/linux/Makefile

@@ -0,0 +1,17 @@
+# Makefile for syscall tables
+#
+# $FreeBSD$
+
+all:
+	@echo "make sysent only"
+
+sysent:  linux_sysent.c linux_syscall.h linux_proto.h linux_syscalls.c linux_systrace_args.c
+
+linux_sysent.c linux_syscall.h linux_proto.h linux_syscalls.c linux_systrace_args.c: \
+		../../kern/makesyscalls.sh syscalls.master syscalls.conf
+	-mv -f linux_sysent.c linux_sysent.c.bak
+	-mv -f linux_syscall.h linux_syscall.h.bak
+	-mv -f linux_proto.h linux_proto.h.bak
+	-mv -f linux_syscalls.c linux_syscalls.c.bak
+	-mv -f linux_systrace_args.c linux_systrace_args.c.bak
+	sh ../../kern/makesyscalls.sh syscalls.master syscalls.conf

sys/amd64/linux/linux.h

@@ -0,0 +1,666 @@
+/*-
+ * Copyright (c) 2013 Dmitry Chagin
+ * Copyright (c) 1994-1996 Søren Schmidt
+ * All rights reserved.
+ *
+ * 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
+ *    in this position and unchanged.
+ * 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. The name of the author may not be used to endorse or promote products
+ *    derived from this software without specific prior written permission
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 AUTHOR 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.
+ *
+ * $FreeBSD$
+ */
+
+#ifndef _AMD64_LINUX_H_
+#define	_AMD64_LINUX_H_
+
+#include <amd64/linux/linux_syscall.h>
+
+/*
+ * debugging support
+ */
+extern u_char linux_debug_map[];
+#define	ldebug(name)	isclr(linux_debug_map, LINUX_SYS_linux_ ## name)
+#define	ARGS(nm, fmt)	"linux(%ld/%ld): "#nm"("fmt")\n",			\
+			(long)td->td_proc->p_pid, (long)td->td_tid
+#define	LMSG(fmt)	"linux(%ld/%ld): "fmt"\n",				\
+			(long)td->td_proc->p_pid, (long)td->td_tid
+#define	LINUX_DTRACE	linuxulator
+
+#define	PTRIN(v)	(void *)(v)
+#define	PTROUT(v)	(uintptr_t)(v)
+
+#define	CP(src,dst,fld) do { (dst).fld = (src).fld; } while (0)
+#define	CP2(src,dst,sfld,dfld) do { (dst).dfld = (src).sfld; } while (0)
+#define	PTRIN_CP(src,dst,fld) \
+	do { (dst).fld = PTRIN((src).fld); } while (0)
+
+/*
+ * Provide a separate set of types for the Linux types.
+ */
+typedef int32_t		l_int;
+typedef int64_t		l_long;
+typedef int16_t		l_short;
+typedef uint32_t	l_uint;
+typedef uint64_t	l_ulong;
+typedef uint16_t	l_ushort;
+
+typedef l_ulong		l_uintptr_t;
+typedef l_long		l_clock_t;
+typedef l_int		l_daddr_t;
+typedef l_ulong		l_dev_t;
+typedef l_uint		l_gid_t;
+typedef l_uint		l_uid_t;
+typedef l_ulong		l_ino_t;
+typedef l_int		l_key_t;
+typedef l_long		l_loff_t;
+typedef l_uint		l_mode_t;
+typedef l_long		l_off_t;
+typedef l_int		l_pid_t;
+typedef l_ulong		l_size_t;
+typedef l_long		l_ssize_t;
+typedef l_long		l_suseconds_t;
+typedef l_long		l_time_t;
+typedef l_int		l_timer_t;
+typedef l_int		l_mqd_t;
+typedef l_size_t	l_socklen_t;
+
+typedef struct {
+	l_int		val[2];
+} l_fsid_t;
+
+typedef struct {
+	l_time_t	tv_sec;
+	l_suseconds_t	tv_usec;
+} l_timeval;
+
+#define	l_fd_set	fd_set
+
+/*
+ * Miscellaneous
+ */
+#define	LINUX_NAME_MAX		255
+#define	LINUX_CTL_MAXNAME	10
+
+#define LINUX_AT_COUNT		17	/* Count of used aux entry types. */
+
+struct l___sysctl_args
+{
+	l_uintptr_t	name;
+	l_int		nlen;
+	l_uintptr_t	oldval;
+	l_uintptr_t	oldlenp;
+	l_uintptr_t	newval;
+	l_size_t	newlen;
+	l_ulong		__spare[4];
+};
+
+/* Scheduling policies */
+#define	LINUX_SCHED_OTHER	0
+#define	LINUX_SCHED_FIFO	1
+#define	LINUX_SCHED_RR		2
+
+/* Resource limits */
+#define	LINUX_RLIMIT_CPU	0
+#define	LINUX_RLIMIT_FSIZE	1
+#define	LINUX_RLIMIT_DATA	2
+#define	LINUX_RLIMIT_STACK	3
+#define	LINUX_RLIMIT_CORE	4
+#define	LINUX_RLIMIT_RSS	5
+#define	LINUX_RLIMIT_NPROC	6
+#define	LINUX_RLIMIT_NOFILE	7
+#define	LINUX_RLIMIT_MEMLOCK	8
+#define	LINUX_RLIMIT_AS		9	/* Address space limit */
+
+#define	LINUX_RLIM_NLIMITS	10
+
+struct l_rlimit {
+	l_ulong		rlim_cur;
+	l_ulong		rlim_max;
+};
+
+/* mmap options */
+#define	LINUX_MAP_SHARED	0x0001
+#define	LINUX_MAP_PRIVATE	0x0002
+#define	LINUX_MAP_FIXED		0x0010
+#define	LINUX_MAP_ANON		0x0020
+#define	LINUX_MAP_GROWSDOWN	0x0100
+
+/*
+ * stat family of syscalls
+ */
+struct l_timespec {
+	l_time_t	tv_sec;
+	l_long		tv_nsec;
+};
+
+struct l_newstat {
+	l_dev_t		st_dev;
+	l_ino_t		st_ino;
+	l_ulong		st_nlink;
+	l_uint		st_mode;
+	l_uid_t		st_uid;
+	l_gid_t		st_gid;
+	l_uint		__st_pad1;
+	l_dev_t		st_rdev;
+	l_off_t		st_size;
+	l_long		st_blksize;
+	l_long		st_blocks;
+	struct l_timespec	st_atim;
+	struct l_timespec	st_mtim;
+	struct l_timespec	st_ctim;
+	l_long		__unused1;
+	l_long		__unused2;
+	l_long		__unused3;
+};
+
+/*
+ * Signalling
+ */
+#define	LINUX_SIGHUP		1
+#define	LINUX_SIGINT		2
+#define	LINUX_SIGQUIT		3
+#define	LINUX_SIGILL		4
+#define	LINUX_SIGTRAP		5
+#define	LINUX_SIGABRT		6
+#define	LINUX_SIGIOT		LINUX_SIGABRT
+#define	LINUX_SIGBUS		7
+#define	LINUX_SIGFPE		8
+#define	LINUX_SIGKILL		9
+#define	LINUX_SIGUSR1		10
+#define	LINUX_SIGSEGV		11
+#define	LINUX_SIGUSR2		12
+#define	LINUX_SIGPIPE		13
+#define	LINUX_SIGALRM		14
+#define	LINUX_SIGTERM		15
+#define	LINUX_SIGSTKFLT		16
+#define	LINUX_SIGCHLD		17
+#define	LINUX_SIGCONT		18
+#define	LINUX_SIGSTOP		19
+#define	LINUX_SIGTSTP		20
+#define	LINUX_SIGTTIN		21
+#define	LINUX_SIGTTOU		22
+#define	LINUX_SIGURG		23
+#define	LINUX_SIGXCPU		24
+#define	LINUX_SIGXFSZ		25
+#define	LINUX_SIGVTALRM		26
+#define	LINUX_SIGPROF		27
+#define	LINUX_SIGWINCH		28
+#define	LINUX_SIGIO		29
+#define	LINUX_SIGPOLL		LINUX_SIGIO
+#define	LINUX_SIGPWR		30
+#define	LINUX_SIGSYS		31
+#define	LINUX_SIGRTMIN		32
+
+#define	LINUX_SIGTBLSZ		31
+#define	LINUX_NSIG		64
+#define	LINUX_NBPW		64
+#define	LINUX_NSIG_WORDS	(LINUX_NSIG / LINUX_NBPW)
+
+/* sigaction flags */
+#define	LINUX_SA_NOCLDSTOP	0x00000001
+#define	LINUX_SA_NOCLDWAIT	0x00000002
+#define	LINUX_SA_SIGINFO	0x00000004
+#define	LINUX_SA_RESTORER	0x04000000
+#define	LINUX_SA_ONSTACK	0x08000000
+#define	LINUX_SA_RESTART	0x10000000
+#define	LINUX_SA_INTERRUPT	0x20000000
+#define	LINUX_SA_NOMASK		0x40000000
+#define	LINUX_SA_ONESHOT	0x80000000
+
+/* sigprocmask actions */
+#define	LINUX_SIG_BLOCK		0
+#define	LINUX_SIG_UNBLOCK	1
+#define	LINUX_SIG_SETMASK	2
+
+/* primitives to manipulate sigset_t */
+
+#define	LINUX_SIGEMPTYSET(set)					\
+	do {							\
+		(set).__bits[0] = 0;				\
+	} while(0)
+
+#define	LINUX_SIGISMEMBER(set, sig)				\
+	(1UL & ((set).__bits[0] >> _SIG_IDX(sig)))
+
+#define	LINUX_SIGADDSET(set, sig)				\
+	(set).__bits[0] |= 1UL << _SIG_IDX(sig)
+
+/* sigaltstack */
+#define	LINUX_MINSIGSTKSZ	2048
+#define	LINUX_SS_ONSTACK	1
+#define	LINUX_SS_DISABLE	2
+
+int linux_to_bsd_sigaltstack(int lsa);
+int bsd_to_linux_sigaltstack(int bsa);
+
+typedef void	(*l_handler_t)(l_int);
+
+typedef struct {
+	l_ulong __bits[LINUX_NSIG_WORDS];
+} l_sigset_t;
+
+typedef struct {
+	l_handler_t	lsa_handler;
+	l_ulong		lsa_flags;
+	l_uintptr_t	lsa_restorer;
+	l_sigset_t	lsa_mask;
+} l_sigaction_t;
+
+typedef struct {
+	l_uintptr_t	ss_sp;
+	l_int		ss_flags;
+	l_size_t	ss_size;
+} l_stack_t;
+
+struct l_fpstate {
+	u_int16_t cwd;
+	u_int16_t swd;
+	u_int16_t twd;
+	u_int16_t fop;
+	u_int64_t rip;
+	u_int64_t rdp;
+	u_int32_t mxcsr;
+	u_int32_t mxcsr_mask;
+	u_int32_t st_space[32];
+	u_int32_t xmm_space[64];
+	u_int32_t reserved2[24];
+};
+
+struct l_sigcontext {
+	l_ulong		sc_r8;
+	l_ulong		sc_r9;
+	l_ulong		sc_r10;
+	l_ulong		sc_r11;
+	l_ulong		sc_r12;
+	l_ulong		sc_r13;
+	l_ulong		sc_r14;
+	l_ulong		sc_r15;
+	l_ulong		sc_rdi;
+	l_ulong		sc_rsi;
+	l_ulong		sc_rbp;
+	l_ulong		sc_rbx;
+	l_ulong		sc_rdx;
+	l_ulong		sc_rax;
+	l_ulong		sc_rcx;
+	l_ulong		sc_rsp;
+	l_ulong		sc_rip;
+	l_ulong		sc_rflags;
+	l_ushort	sc_cs;
+	l_ushort	sc_gs;
+	l_ushort	sc_fs;
+	l_ushort	sc___pad0;
+	l_ulong		sc_err;
+	l_ulong		sc_trapno;
+	l_sigset_t	sc_mask;
+	l_ulong		sc_cr2;
+	struct l_fpstate *sc_fpstate;
+	l_ulong		sc_reserved1[8];
+};
+
+struct l_ucontext {
+	l_ulong		uc_flags;
+	l_uintptr_t	uc_link;
+	l_stack_t	uc_stack;
+	struct l_sigcontext	uc_mcontext;
+	l_sigset_t	uc_sigmask;
+};
+
+#define LINUX_SI_PREAMBLE_SIZE	(4 * sizeof(int))
+#define	LINUX_SI_MAX_SIZE	128
+#define	LINUX_SI_PAD_SIZE	((LINUX_SI_MAX_SIZE - \
+				    LINUX_SI_PREAMBLE_SIZE) / sizeof(l_int))
+typedef union l_sigval {
+	l_int		sival_int;
+	l_uintptr_t	sival_ptr;
+} l_sigval_t;
+
+typedef struct l_siginfo {
+	l_int		lsi_signo;
+	l_int		lsi_errno;
+	l_int		lsi_code;
+	union {
+		l_int	_pad[LINUX_SI_PAD_SIZE];
+
+		struct {
+			l_pid_t		_pid;
+			l_uid_t		_uid;
+		} _kill;
+
+		struct {
+			l_timer_t	_tid;
+			l_int		_overrun;
+			char		_pad[sizeof(l_uid_t) - sizeof(int)];
+			union l_sigval	_sigval;
+			l_uint		_sys_private;
+		} _timer;
+
+		struct {
+			l_pid_t		_pid;		/* sender's pid */
+			l_uid_t		_uid;		/* sender's uid */
+			union l_sigval	_sigval;
+		} _rt;
+
+		struct {
+			l_pid_t		_pid;		/* which child */
+			l_uid_t		_uid;		/* sender's uid */
+			l_int		_status;	/* exit code */
+			l_clock_t	_utime;
+			l_clock_t	_stime;
+		} _sigchld;
+
+		struct {
+			l_uintptr_t	_addr;	/* Faulting insn/memory ref. */
+		} _sigfault;
+
+		struct {
+			l_long		_band;	/* POLL_IN,POLL_OUT,POLL_MSG */
+			l_int		_fd;
+		} _sigpoll;
+	} _sifields;
+} l_siginfo_t;
+
+#define	lsi_pid		_sifields._kill._pid
+#define	lsi_uid		_sifields._kill._uid
+#define	lsi_tid		_sifields._timer._tid
+#define	lsi_overrun	_sifields._timer._overrun
+#define	lsi_sys_private	_sifields._timer._sys_private
+#define	lsi_status	_sifields._sigchld._status
+#define	lsi_utime	_sifields._sigchld._utime
+#define	lsi_stime	_sifields._sigchld._stime
+#define	lsi_value	_sifields._rt._sigval
+#define	lsi_int		_sifields._rt._sigval.sival_int
+#define	lsi_ptr		_sifields._rt._sigval.sival_ptr
+#define	lsi_addr	_sifields._sigfault._addr
+#define	lsi_band	_sifields._sigpoll._band
+#define	lsi_fd		_sifields._sigpoll._fd
+
+/*
+ * We make the stack look like Linux expects it when calling a signal
+ * handler, but use the BSD way of calling the handler and sigreturn().
+ * This means that we need to pass the pointer to the handler too.
+ * It is appended to the frame to not interfere with the rest of it.
+ */
+
+struct l_rt_sigframe {
+	struct l_ucontext	sf_sc;
+	struct l_siginfo	sf_si;
+	l_handler_t		sf_handler;
+};
+
+/*
+ * open/fcntl flags
+ */
+#define	LINUX_O_RDONLY		00000000
+#define	LINUX_O_WRONLY		00000001
+#define	LINUX_O_RDWR		00000002
+#define	LINUX_O_ACCMODE		00000003
+#define	LINUX_O_CREAT		00000100
+#define	LINUX_O_EXCL		00000200
+#define	LINUX_O_NOCTTY		00000400
+#define	LINUX_O_TRUNC		00001000
+#define	LINUX_O_APPEND		00002000
+#define	LINUX_O_NONBLOCK	00004000
+#define	LINUX_O_NDELAY		LINUX_O_NONBLOCK
+#define	LINUX_O_SYNC		00010000
+#define	LINUX_FASYNC		00020000
+#define	LINUX_O_DIRECT		00040000	/* Direct disk access hint */
+#define	LINUX_O_LARGEFILE	00100000
+#define	LINUX_O_DIRECTORY	00200000	/* Must be a directory */
+#define	LINUX_O_NOFOLLOW	00400000	/* Do not follow links */
+#define	LINUX_O_NOATIME		01000000
+#define	LINUX_O_CLOEXEC		02000000
+
+#define	LINUX_F_DUPFD		0
+#define	LINUX_F_GETFD		1
+#define	LINUX_F_SETFD		2
+#define	LINUX_F_GETFL		3
+#define	LINUX_F_SETFL		4
+#define	LINUX_F_GETLK		5
+#define	LINUX_F_SETLK		6
+#define	LINUX_F_SETLKW		7
+#define	LINUX_F_SETOWN		8
+#define	LINUX_F_GETOWN		9
+
+#define	LINUX_F_RDLCK		0
+#define	LINUX_F_WRLCK		1
+#define	LINUX_F_UNLCK		2
+
+/*
+ * mount flags
+ */
+#define	LINUX_MS_RDONLY		0x0001
+#define	LINUX_MS_NOSUID		0x0002
+#define	LINUX_MS_NODEV		0x0004
+#define	LINUX_MS_NOEXEC		0x0008
+#define	LINUX_MS_REMOUNT	0x0020
+
+/*
+ * SystemV IPC defines
+ */
+#define	LINUX_IPC_RMID		0
+#define	LINUX_IPC_SET		1
+#define	LINUX_IPC_STAT		2
+#define	LINUX_IPC_INFO		3
+
+#define	LINUX_SHM_LOCK		11
+#define	LINUX_SHM_UNLOCK	12
+#define	LINUX_SHM_STAT		13
+#define	LINUX_SHM_INFO		14
+
+#define	LINUX_SHM_RDONLY	0x1000
+#define	LINUX_SHM_RND		0x2000
+#define	LINUX_SHM_REMAP		0x4000
+
+/* semctl commands */
+#define	LINUX_GETPID		11
+#define	LINUX_GETVAL		12
+#define	LINUX_GETALL		13
+#define	LINUX_GETNCNT		14
+#define	LINUX_GETZCNT		15
+#define	LINUX_SETVAL		16
+#define	LINUX_SETALL		17
+#define	LINUX_SEM_STAT		18
+#define	LINUX_SEM_INFO		19
+
+union l_semun {
+	l_int		val;
+	l_uintptr_t	buf;
+	l_uintptr_t	array;
+	l_uintptr_t	__buf;
+	l_uintptr_t	__pad;
+};
+
+struct l_ipc_perm {
+	l_key_t		key;
+	l_uid_t		uid;
+	l_gid_t		gid;
+	l_uid_t		cuid;
+	l_gid_t		cgid;
+	l_ushort	mode;
+	l_ushort	seq;
+};
+
+/*
+ * Socket defines
+ */
+
+#define	LINUX_SOL_SOCKET	1
+#define	LINUX_SOL_IP		0
+#define	LINUX_SOL_IPX		256
+#define	LINUX_SOL_AX25		257
+#define	LINUX_SOL_TCP		6
+#define	LINUX_SOL_UDP		17
+
+#define	LINUX_SO_DEBUG		1
+#define	LINUX_SO_REUSEADDR	2
+#define	LINUX_SO_TYPE		3
+#define	LINUX_SO_ERROR		4
+#define	LINUX_SO_DONTROUTE	5
+#define	LINUX_SO_BROADCAST	6
+#define	LINUX_SO_SNDBUF		7
+#define	LINUX_SO_RCVBUF		8
+#define	LINUX_SO_KEEPALIVE	9
+#define	LINUX_SO_OOBINLINE	10
+#define	LINUX_SO_NO_CHECK	11
+#define	LINUX_SO_PRIORITY	12
+#define	LINUX_SO_LINGER		13
+#define	LINUX_SO_PASSCRED	16
+#define	LINUX_SO_PEERCRED	17
+#define	LINUX_SO_RCVLOWAT	18
+#define	LINUX_SO_SNDLOWAT	19
+#define	LINUX_SO_RCVTIMEO	20
+#define	LINUX_SO_SNDTIMEO	21
+#define	LINUX_SO_TIMESTAMP	29
+#define	LINUX_SO_ACCEPTCONN	30
+
+#define	LINUX_IP_TOS		1
+#define	LINUX_IP_TTL		2
+#define	LINUX_IP_HDRINCL	3
+#define	LINUX_IP_OPTIONS	4
+
+#define	LINUX_IP_MULTICAST_IF		32
+#define	LINUX_IP_MULTICAST_TTL		33
+#define	LINUX_IP_MULTICAST_LOOP		34
+#define	LINUX_IP_ADD_MEMBERSHIP		35
+#define	LINUX_IP_DROP_MEMBERSHIP	36
+
+struct l_sockaddr {
+	l_ushort	sa_family;
+	char		sa_data[14];
+};
+
+struct l_msghdr {
+	l_uintptr_t	msg_name;
+	l_int		msg_namelen;
+	l_uintptr_t	msg_iov;
+	l_size_t	msg_iovlen;
+	l_uintptr_t	msg_control;
+	l_size_t	msg_controllen;
+	l_uint		msg_flags;
+};
+
+struct l_cmsghdr {
+	l_size_t	cmsg_len;
+	l_int		cmsg_level;
+	l_int		cmsg_type;
+};
+
+struct l_ifmap {
+	l_ulong		mem_start;
+	l_ulong		mem_end;
+	l_ushort	base_addr;
+	u_char		irq;
+	u_char		dma;
+	u_char		port;
+} __packed;
+
+#define	LINUX_IFHWADDRLEN	6
+#define	LINUX_IFNAMSIZ		16
+
+struct l_ifreq {
+	union {
+		char	ifrn_name[LINUX_IFNAMSIZ];
+	} ifr_ifrn;
+
+	union {
+		struct l_sockaddr	ifru_addr;
+		struct l_sockaddr	ifru_dstaddr;
+		struct l_sockaddr	ifru_broadaddr;
+		struct l_sockaddr	ifru_netmask;
+		struct l_sockaddr	ifru_hwaddr;
+		l_short		ifru_flags[1];
+		l_int		ifru_metric;
+		l_int		ifru_mtu;
+		struct l_ifmap	ifru_map;
+		char		ifru_slave[LINUX_IFNAMSIZ];
+		l_uintptr_t	ifru_data;
+	} ifr_ifru;
+} __packed;
+
+#define	ifr_name	ifr_ifrn.ifrn_name	/* Interface name */
+#define	ifr_hwaddr	ifr_ifru.ifru_hwaddr	/* MAC address */
+
+struct l_ifconf {
+	int	ifc_len;
+	union {
+		l_uintptr_t	ifcu_buf;
+		l_uintptr_t	ifcu_req;
+	} ifc_ifcu;
+};
+
+#define	ifc_buf		ifc_ifcu.ifcu_buf
+#define	ifc_req		ifc_ifcu.ifcu_req
+
+/*
+ * poll()
+ */
+#define	LINUX_POLLIN		0x0001
+#define	LINUX_POLLPRI		0x0002
+#define	LINUX_POLLOUT		0x0004
+#define	LINUX_POLLERR		0x0008
+#define	LINUX_POLLHUP		0x0010
+#define	LINUX_POLLNVAL		0x0020
+#define	LINUX_POLLRDNORM	0x0040
+#define	LINUX_POLLRDBAND	0x0080
+#define	LINUX_POLLWRNORM	0x0100
+#define	LINUX_POLLWRBAND	0x0200
+#define	LINUX_POLLMSG		0x0400
+
+struct l_pollfd {
+	l_int		fd;
+	l_short		events;
+	l_short		revents;
+};
+
+
+#define	LINUX_CLONE_VM			0x00000100
+#define	LINUX_CLONE_FS			0x00000200
+#define	LINUX_CLONE_FILES		0x00000400
+#define	LINUX_CLONE_SIGHAND		0x00000800
+#define	LINUX_CLONE_PID			0x00001000	/* No longer exist in Linux */
+#define	LINUX_CLONE_VFORK		0x00004000
+#define	LINUX_CLONE_PARENT		0x00008000
+#define	LINUX_CLONE_THREAD		0x00010000
+#define	LINUX_CLONE_SETTLS		0x00080000
+#define	LINUX_CLONE_PARENT_SETTID	0x00100000
+#define	LINUX_CLONE_CHILD_CLEARTID	0x00200000
+#define	LINUX_CLONE_CHILD_SETTID	0x01000000
+
+#define LINUX_ARCH_SET_GS		0x1001
+#define LINUX_ARCH_SET_FS		0x1002
+#define LINUX_ARCH_GET_GS		0x1003
+#define LINUX_ARCH_GET_FS		0x1004
+
+#define	linux_copyout_rusage(r, u)	copyout(r, u, sizeof(*r))
+
+/* robust futexes */
+struct linux_robust_list {
+	l_uintptr_t			next;
+};
+ 
+struct linux_robust_list_head {
+	struct linux_robust_list	list;
+	l_long				futex_offset;
+	l_uintptr_t			pending_list;
+};
+
+#endif /* !_AMD64_LINUX_H_ */

sys/amd64/linux/linux_dummy.c

@@ -0,0 +1,154 @@
+/*-
+ * Copyright (c) 2013 Dmitry Chagin
+ * All rights reserved.
+ *
+ * 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
+ *    in this position and unchanged.
+ * 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.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 AUTHOR 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.
+ */
+
+#include <sys/cdefs.h>
+__FBSDID("$FreeBSD$");
+
+#include "opt_compat.h"
+#include "opt_kdtrace.h"
+
+#include <sys/param.h>
+#include <sys/kernel.h>
+#include <sys/sdt.h>
+#include <sys/systm.h>
+#include <sys/proc.h>
+
+#include <amd64/linux/linux.h>
+#include <amd64/linux/linux_proto.h>
+#include <compat/linux/linux_dtrace.h>
+#include <compat/linux/linux_util.h>
+
+/* DTrace init */
+LIN_SDT_PROVIDER_DECLARE(LINUX_DTRACE);
+
+DUMMY(mincore);
+DUMMY(sendfile);
+DUMMY(ptrace);
+DUMMY(syslog);
+DUMMY(setfsuid);
+DUMMY(setfsgid);
+DUMMY(sysfs);
+DUMMY(vhangup);
+DUMMY(pivot_root);
+DUMMY(adjtimex);
+DUMMY(swapoff);
+DUMMY(create_module);
+DUMMY(init_module);
+DUMMY(delete_module);
+DUMMY(get_kernel_syms);
+DUMMY(query_module);
+DUMMY(quotactl);
+DUMMY(nfsservctl);
+DUMMY(getpmsg);
+DUMMY(putpmsg);
+DUMMY(afs_syscall);
+DUMMY(tuxcall);
+DUMMY(security);
+DUMMY(set_thread_area);
+DUMMY(lookup_dcookie);
+DUMMY(epoll_create);
+DUMMY(epoll_ctl_old);
+DUMMY(epoll_wait_old);
+DUMMY(remap_file_pages);
+DUMMY(semtimedop);
+DUMMY(epoll_ctl);
+DUMMY(epoll_wait);
+DUMMY(mbind);
+DUMMY(get_mempolicy);
+DUMMY(set_mempolicy);
+DUMMY(mq_open);
+DUMMY(mq_unlink);
+DUMMY(mq_timedsend);
+DUMMY(mq_timedreceive);
+DUMMY(mq_notify);
+DUMMY(mq_getsetattr);
+DUMMY(kexec_load);
+DUMMY(add_key);
+DUMMY(request_key);
+DUMMY(keyctl);
+DUMMY(ioprio_set);
+DUMMY(ioprio_get);
+DUMMY(inotify_init);
+DUMMY(inotify_add_watch);
+DUMMY(inotify_rm_watch);
+DUMMY(migrate_pages);
+DUMMY(ppoll);
+DUMMY(unshare);
+DUMMY(splice);
+DUMMY(tee);
+DUMMY(sync_file_range);
+DUMMY(vmsplice);
+DUMMY(move_pages);
+DUMMY(utimensat);
+DUMMY(epoll_pwait);
+DUMMY(signalfd);
+DUMMY(timerfd);
+DUMMY(eventfd);
+DUMMY(fallocate);
+DUMMY(timerfd_settime);
+DUMMY(timerfd_gettime);
+DUMMY(signalfd4);
+DUMMY(eventfd2);
+DUMMY(epoll_create1);
+DUMMY(inotify_init1);
+DUMMY(preadv);
+DUMMY(pwritev);
+DUMMY(rt_tsigqueueinfo);
+DUMMY(perf_event_open);
+DUMMY(recvmmsg);
+DUMMY(fanotify_init);
+DUMMY(fanotify_mark);
+DUMMY(name_to_handle_at);
+DUMMY(open_by_handle_at);
+DUMMY(clock_adjtime);
+DUMMY(syncfs);
+DUMMY(sendmmsg);
+DUMMY(setns);
+DUMMY(process_vm_readv);
+DUMMY(process_vm_writev);
+DUMMY(kcmp);
+DUMMY(finit_module);
+
+#define DUMMY_XATTR(s)						\
+int								\
+linux_ ## s ## xattr(						\
+    struct thread *td, struct linux_ ## s ## xattr_args *arg)	\
+{								\
+								\
+	return (ENOATTR);					\
+}
+DUMMY_XATTR(set);
+DUMMY_XATTR(lset);
+DUMMY_XATTR(fset);
+DUMMY_XATTR(get);
+DUMMY_XATTR(lget);
+DUMMY_XATTR(fget);
+DUMMY_XATTR(list);
+DUMMY_XATTR(llist);
+DUMMY_XATTR(flist);
+DUMMY_XATTR(remove);
+DUMMY_XATTR(lremove);
+DUMMY_XATTR(fremove);

sys/amd64/linux/linux_genassym.c

@@ -0,0 +1,15 @@
+#include <sys/cdefs.h>
+__FBSDID("$FreeBSD$");
+
+#include <sys/param.h>
+#include <sys/assym.h>
+#include <sys/systm.h>
+
+#include <amd64/linux/linux.h>
+#include <compat/linux/linux_mib.h>
+
+ASSYM(LINUX_RT_SIGF_HANDLER, offsetof(struct l_rt_sigframe, sf_handler));
+ASSYM(LINUX_RT_SIGF_UC, offsetof(struct l_rt_sigframe, sf_sc));
+ASSYM(LINUX_RT_SIGF_SC, offsetof(struct l_ucontext, uc_mcontext));
+ASSYM(LINUX_VERSION_CODE, LINUX_VERSION_CODE);
+ASSYM(LINUX_SC_RSP, offsetof(struct l_sigcontext, sc_rsp));

sys/amd64/linux/linux_ipc64.h

@@ -0,0 +1,142 @@
+/*-
+ * Copyright (c) 2002 Maxim Sobolev <sobomax@FreeBSD.org>
+ * All rights reserved.
+ *
+ * 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
+ *    in this position and unchanged.
+ * 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. The name of the author may not be used to endorse or promote products
+ *    derived from this software without specific prior written permission
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 AUTHOR 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.
+ *
+ * $FreeBSD$
+ */
+
+#ifndef _AMD64_LINUX_LINUX_IPC64_H_
+#define	_AMD64_LINUX_LINUX_IPC64_H_
+
+/*
+ * The ipc64_perm structure for i386 architecture.
+ * Note extra padding because this structure is passed back and forth
+ * between kernel and user space.
+ *
+ * Pad space is left for:
+ * - 32-bit mode_t and seq
+ * - 2 miscellaneous 32-bit values
+ */
+
+struct l_ipc64_perm
+{
+	l_key_t		key;
+	l_uid_t		uid;
+	l_gid_t		gid;
+	l_uid_t		cuid;
+	l_gid_t		cgid;
+	l_mode_t	mode;
+	l_ushort	__pad1;
+	l_ushort	seq;
+	l_ushort	__pad2;
+	l_ulong		__unused1;
+	l_ulong		__unused2;
+};
+
+/*
+ * The msqid64_ds structure for i386 architecture.
+ * Note extra padding because this structure is passed back and forth
+ * between kernel and user space.
+ *
+ * Pad space is left for:
+ * - 64-bit time_t to solve y2038 problem
+ * - 2 miscellaneous 32-bit values
+ */
+
+struct l_msqid64_ds {
+	struct l_ipc64_perm msg_perm;
+	l_time_t	msg_stime;	/* last msgsnd time */
+	l_ulong		__unused1;
+	l_time_t	msg_rtime;	/* last msgrcv time */
+	l_ulong		__unused2;
+	l_time_t	msg_ctime;	/* last change time */
+	l_ulong		__unused3;
+	l_ulong		msg_cbytes;	/* current number of bytes on queue */
+	l_ulong		msg_qnum;	/* number of messages in queue */
+	l_ulong		msg_qbytes;	/* max number of bytes on queue */
+	l_pid_t		msg_lspid;	/* pid of last msgsnd */
+	l_pid_t		msg_lrpid;	/* last receive pid */
+	l_ulong		__unused4;
+	l_ulong		__unused5;
+};
+
+/*
+ * The semid64_ds structure for i386 architecture.
+ * Note extra padding because this structure is passed back and forth
+ * between kernel and user space.
+ *
+ * Pad space is left for:
+ * - 64-bit time_t to solve y2038 problem
+ * - 2 miscellaneous 32-bit values
+ */
+
+struct l_semid64_ds {
+	struct l_ipc64_perm sem_perm;	/* permissions */
+	l_time_t	sem_otime;	/* last semop time */
+	l_ulong		__unused1;
+	l_time_t	sem_ctime;	/* last change time */
+	l_ulong		__unused2;
+	l_ulong		sem_nsems;	/* no. of semaphores in array */
+	l_ulong		__unused3;
+	l_ulong		__unused4;
+};
+
+/*
+ * The shmid64_ds structure for i386 architecture.
+ * Note extra padding because this structure is passed back and forth
+ * between kernel and user space.
+ *
+ * Pad space is left for:
+ * - 64-bit time_t to solve y2038 problem
+ * - 2 miscellaneous 32-bit values
+ */
+
+struct l_shmid64_ds {
+	struct l_ipc64_perm shm_perm;	/* operation perms */
+	l_size_t	shm_segsz;	/* size of segment (bytes) */
+	l_time_t	shm_atime;	/* last attach time */
+	l_time_t	shm_dtime;	/* last detach time */
+	l_time_t	shm_ctime;	/* last change time */
+	l_pid_t		shm_cpid;	/* pid of creator */
+	l_pid_t		shm_lpid;	/* pid of last operator */
+	l_ulong		shm_nattch;	/* no. of current attaches */
+	l_ulong		__unused4;
+	l_ulong		__unused5;
+};
+
+struct l_shminfo64 {
+	l_ulong		shmmax;
+	l_ulong		shmmin;
+	l_ulong		shmmni;
+	l_ulong		shmseg;
+	l_ulong		shmall;
+	l_ulong   	__unused1;
+	l_ulong		__unused2;
+	l_ulong		__unused3;
+	l_ulong		__unused4;
+};
+
+#endif /* !_AMD64_LINUX_LINUX_IPC64_H_ */

sys/amd64/linux/linux_locore.s

@@ -0,0 +1,75 @@
+/* $FreeBSD$ */
+
+#include "linux_assym.h"			/* system definitions */
+#include <machine/asmacros.h>			/* miscellaneous asm macros */
+
+#include <amd64/linux/linux_syscall.h>		/* system call numbers */
+
+	.data
+
+	.globl linux_platform
+linux_platform:
+	.asciz "x86_64"
+
+
+	.text
+/*
+ * To avoid excess stack frame the signal trampoline code emulates
+ * the 'call' instruction.
+ */
+NON_GPROF_ENTRY(linux_rt_sigcode)
+	movq	%rsp, %rbx			/* preserve sigframe */
+	call	.getip
+.getip:
+	popq	%rax
+	add	$.startrtsigcode-.getip, %rax	/* ret address */
+	pushq	%rax
+	jmp	*LINUX_RT_SIGF_HANDLER(%rbx)
+.startrtsigcode:
+	movq	$LINUX_SYS_linux_rt_sigreturn,%rax   /* linux_rt_sigreturn() */
+	syscall					/* enter kernel with args */
+	hlt
+0:	jmp	0b
+
+NON_GPROF_ENTRY(__vdso_clock_gettime)
+	movq	$LINUX_SYS_linux_clock_gettime,%rax
+	syscall
+	ret
+.weak clock_gettime
+.set clock_gettime, __vdso_clock_gettime
+
+NON_GPROF_ENTRY(__vdso_time)
+	movq	$LINUX_SYS_linux_time,%rax
+	syscall
+	ret
+.weak time
+.set time, __vdso_time
+
+NON_GPROF_ENTRY(__vdso_gettimeofday)
+	movq	$LINUX_SYS_gettimeofday,%rax
+	syscall
+	ret
+.weak gettimeofday
+.set gettimeofday, __vdso_gettimeofday
+
+NON_GPROF_ENTRY(__vdso_getcpu)
+	movq	$-38,%rax	/* not implemented */
+	ret
+.weak getcpu
+.set getcpu, __vdso_getcpu
+
+
+	.section .note.Linux, "a",@note
+	.long 2f - 1f		/* namesz */
+	.balign 4
+	.long 4f - 3f		/* descsz */
+	.long 0
+1:
+	.asciz "Linux"
+2:
+	.balign 4
+3:
+	.long LINUX_VERSION_CODE
+4:
+	.balign 4
+	.previous

sys/amd64/linux/linux_machdep.c

@@ -0,0 +1,492 @@
+/*-
+ * Copyright (c) 2013 Dmitry Chagin
+ * Copyright (c) 2004 Tim J. Robbins
+ * Copyright (c) 2002 Doug Rabson
+ * Copyright (c) 2000 Marcel Moolenaar
+ * All rights reserved.
+ *
+ * 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
+ *    in this position and unchanged.
+ * 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. The name of the author may not be used to endorse or promote products
+ *    derived from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 AUTHOR 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.
+ */
+
+#include <sys/cdefs.h>
+__FBSDID("$FreeBSD$");
+
+#include <sys/param.h>
+#include <sys/kernel.h>
+#include <sys/systm.h>
+#include <sys/capability.h>
+#include <sys/dirent.h>
+#include <sys/file.h>
+#include <sys/fcntl.h>
+#include <sys/filedesc.h>
+#include <sys/clock.h>
+#include <sys/imgact.h>
+#include <sys/ktr.h>
+#include <sys/limits.h>
+#include <sys/lock.h>
+#include <sys/malloc.h>
+#include <sys/mman.h>
+#include <sys/mutex.h>
+#include <sys/priv.h>
+#include <sys/proc.h>
+#include <sys/resource.h>
+#include <sys/resourcevar.h>
+#include <sys/sched.h>
+#include <sys/syscallsubr.h>
+#include <sys/sysproto.h>
+#include <sys/vnode.h>
+#include <sys/unistd.h>
+#include <sys/wait.h>
+
+#include <security/mac/mac_framework.h>
+
+#include <ufs/ufs/extattr.h>
+#include <ufs/ufs/quota.h>
+#include <ufs/ufs/ufsmount.h>
+
+#include <machine/frame.h>
+#include <machine/md_var.h>
+#include <machine/pcb.h>
+#include <machine/psl.h>
+#include <machine/segments.h>
+#include <machine/specialreg.h>
+
+#include <vm/vm.h>
+#include <vm/pmap.h>
+#include <vm/vm_extern.h>
+#include <vm/vm_kern.h>
+#include <vm/vm_map.h>
+
+#include <amd64/linux/linux.h>
+#include <amd64/linux/linux_proto.h>
+#include <compat/linux/linux_ipc.h>
+#include <compat/linux/linux_file.h>
+#include <compat/linux/linux_misc.h>
+#include <compat/linux/linux_signal.h>
+#include <compat/linux/linux_util.h>
+#include <compat/linux/linux_emul.h>
+
+int
+linux_to_bsd_sigaltstack(int lsa)
+{
+	int bsa = 0;
+
+	if (lsa & LINUX_SS_DISABLE)
+		bsa |= SS_DISABLE;
+	if (lsa & LINUX_SS_ONSTACK)
+		bsa |= SS_ONSTACK;
+	return (bsa);
+}
+
+int
+bsd_to_linux_sigaltstack(int bsa)
+{
+	int lsa = 0;
+
+	if (bsa & SS_DISABLE)
+		lsa |= LINUX_SS_DISABLE;
+	if (bsa & SS_ONSTACK)
+		lsa |= LINUX_SS_ONSTACK;
+	return (lsa);
+}
+
+int
+linux_execve(struct thread *td, struct linux_execve_args *args)
+{
+	struct image_args eargs;
+	char *path;
+	int error;
+
+	LCONVPATHEXIST(td, args->path, &path);
+
+	LINUX_CTR(execve);
+
+	error = exec_copyin_args(&eargs, path, UIO_SYSSPACE, args->argp,
+	    args->envp);
+	free(path, M_TEMP);
+	if (error == 0)
+		error = linux_common_execve(td, &eargs);
+	return (error);
+}
+
+int
+linux_set_upcall_kse(struct thread *td, register_t stack)
+{
+
+	if (stack)
+		td->td_frame->tf_rsp = stack;
+
+	/*
+	 * The newly created Linux thread returns
+	 * to the user space by the same path that a parent do.
+	 */
+	td->td_frame->tf_rax = 0;
+	return (0);
+}
+
+#define STACK_SIZE  (2 * 1024 * 1024)
+#define GUARD_SIZE  (4 * PAGE_SIZE)
+
+int
+linux_mmap2(struct thread *td, struct linux_mmap2_args *args)
+{
+	struct proc *p = td->td_proc;
+	struct mmap_args /* {
+		caddr_t addr;
+		size_t len;
+		int prot;
+		int flags;
+		int fd;
+		long pad;
+		off_t pos;
+	} */ bsd_args;
+	int error;
+	struct file *fp;
+	cap_rights_t rights;
+
+	LINUX_CTR6(mmap2, "0x%lx, %ld, %ld, 0x%08lx, %ld, 0x%lx",
+	    args->addr, args->len, args->prot,
+	    args->flags, args->fd, args->pgoff);
+
+	error = 0;
+	bsd_args.flags = 0;
+	fp = NULL;
+
+	/*
+	 * Linux mmap(2):
+	 * You must specify exactly one of MAP_SHARED and MAP_PRIVATE
+	 */
+	if (! ((args->flags & LINUX_MAP_SHARED) ^
+	    (args->flags & LINUX_MAP_PRIVATE)))
+		return (EINVAL);
+
+	if (args->flags & LINUX_MAP_SHARED)
+		bsd_args.flags |= MAP_SHARED;
+	if (args->flags & LINUX_MAP_PRIVATE)
+		bsd_args.flags |= MAP_PRIVATE;
+	if (args->flags & LINUX_MAP_FIXED)
+		bsd_args.flags |= MAP_FIXED;
+	if (args->flags & LINUX_MAP_ANON)
+		bsd_args.flags |= MAP_ANON;
+	else
+		bsd_args.flags |= MAP_NOSYNC;
+	if (args->flags & LINUX_MAP_GROWSDOWN)
+		bsd_args.flags |= MAP_STACK;
+
+	/*
+	 * PROT_READ, PROT_WRITE, or PROT_EXEC implies PROT_READ and PROT_EXEC
+	 * on Linux/i386. We do this to ensure maximum compatibility.
+	 * Linux/ia64 does the same in i386 emulation mode.
+	 */
+	bsd_args.prot = args->prot;
+	if (bsd_args.prot & (PROT_READ | PROT_WRITE | PROT_EXEC))
+		bsd_args.prot |= PROT_READ | PROT_EXEC;
+
+	/* Linux does not check file descriptor when MAP_ANONYMOUS is set. */
+	bsd_args.fd = (bsd_args.flags & MAP_ANON) ? -1 : args->fd;
+	if (bsd_args.fd != -1) {
+		/*
+		 * Linux follows Solaris mmap(2) description:
+		 * The file descriptor fildes is opened with
+		 * read permission, regardless of the
+		 * protection options specified.
+		 */
+
+		error = fget(td, bsd_args.fd,
+		    cap_rights_init(&rights, CAP_MMAP), &fp);
+		if (error != 0 )
+			return (error);
+		if (fp->f_type != DTYPE_VNODE) {
+			fdrop(fp, td);
+			return (EINVAL);
+		}
+
+		/* Linux mmap() just fails for O_WRONLY files */
+		if (!(fp->f_flag & FREAD)) {
+			fdrop(fp, td);
+			return (EACCES);
+		}
+
+		fdrop(fp, td);
+	}
+
+	if (args->flags & LINUX_MAP_GROWSDOWN) {
+		/*
+		 * The Linux MAP_GROWSDOWN option does not limit auto
+		 * growth of the region.  Linux mmap with this option
+		 * takes as addr the inital BOS, and as len, the initial
+		 * region size.  It can then grow down from addr without
+		 * limit.  However, Linux threads has an implicit internal
+		 * limit to stack size of STACK_SIZE.  Its just not
+		 * enforced explicitly in Linux.  But, here we impose
+		 * a limit of (STACK_SIZE - GUARD_SIZE) on the stack
+		 * region, since we can do this with our mmap.
+		 *
+		 * Our mmap with MAP_STACK takes addr as the maximum
+		 * downsize limit on BOS, and as len the max size of
+		 * the region.  It then maps the top SGROWSIZ bytes,
+		 * and auto grows the region down, up to the limit
+		 * in addr.
+		 *
+		 * If we don't use the MAP_STACK option, the effect
+		 * of this code is to allocate a stack region of a
+		 * fixed size of (STACK_SIZE - GUARD_SIZE).
+		 */
+
+		if ((caddr_t)PTRIN(args->addr) + args->len >
+		    p->p_vmspace->vm_maxsaddr) {
+			/*
+			 * Some Linux apps will attempt to mmap
+			 * thread stacks near the top of their
+			 * address space.  If their TOS is greater
+			 * than vm_maxsaddr, vm_map_growstack()
+			 * will confuse the thread stack with the
+			 * process stack and deliver a SEGV if they
+			 * attempt to grow the thread stack past their
+			 * current stacksize rlimit.  To avoid this,
+			 * adjust vm_maxsaddr upwards to reflect
+			 * the current stacksize rlimit rather
+			 * than the maximum possible stacksize.
+			 * It would be better to adjust the
+			 * mmap'ed region, but some apps do not check
+			 * mmap's return value.
+			 */
+			PROC_LOCK(p);
+			p->p_vmspace->vm_maxsaddr = (char *)USRSTACK -
+			    lim_cur(p, RLIMIT_STACK);
+			PROC_UNLOCK(p);
+		}
+
+		/*
+		 * This gives us our maximum stack size and a new BOS.
+		 * If we're using VM_STACK, then mmap will just map
+		 * the top SGROWSIZ bytes, and let the stack grow down
+		 * to the limit at BOS.  If we're not using VM_STACK
+		 * we map the full stack, since we don't have a way
+		 * to autogrow it.
+		 */
+		if (args->len > STACK_SIZE - GUARD_SIZE) {
+			bsd_args.addr = (caddr_t)PTRIN(args->addr);
+			bsd_args.len = args->len;
+		} else {
+			bsd_args.addr = (caddr_t)PTRIN(args->addr) -
+			    (STACK_SIZE - GUARD_SIZE - args->len);
+			bsd_args.len = STACK_SIZE - GUARD_SIZE;
+		}
+	} else {
+		bsd_args.addr = (caddr_t)PTRIN(args->addr);
+		bsd_args.len  = args->len;
+	}
+	bsd_args.pos = (off_t)args->pgoff;
+
+	error = sys_mmap(td, &bsd_args);
+
+	LINUX_CTR2(mmap2, "return: %d (%p)",
+	    error, td->td_retval[0]);
+	return (error);
+}
+
+int
+linux_mprotect(struct thread *td, struct linux_mprotect_args *uap)
+{
+	struct mprotect_args bsd_args;
+
+	LINUX_CTR(mprotect);
+
+	bsd_args.addr = uap->addr;
+	bsd_args.len = uap->len;
+	bsd_args.prot = uap->prot;
+	if (bsd_args.prot & (PROT_READ | PROT_WRITE | PROT_EXEC))
+		bsd_args.prot |= PROT_READ | PROT_EXEC;
+	return (sys_mprotect(td, &bsd_args));
+}
+
+int
+linux_iopl(struct thread *td, struct linux_iopl_args *args)
+{
+	int error;
+
+	LINUX_CTR(iopl);
+
+	if (args->level > 3)
+		return (EINVAL);
+	if ((error = priv_check(td, PRIV_IO)) != 0)
+		return (error);
+	if ((error = securelevel_gt(td->td_ucred, 0)) != 0)
+		return (error);
+	td->td_frame->tf_rflags = (td->td_frame->tf_rflags & ~PSL_IOPL) |
+	    (args->level * (PSL_IOPL / 3));
+
+	return (0);
+}
+
+int
+linux_rt_sigsuspend(struct thread *td, struct linux_rt_sigsuspend_args *uap)
+{
+	l_sigset_t lmask;
+	sigset_t sigmask;
+	int error;
+
+	LINUX_CTR2(rt_sigsuspend, "%p, %ld",
+	    uap->newset, uap->sigsetsize);
+
+	if (uap->sigsetsize != sizeof(l_sigset_t))
+		return (EINVAL);
+
+	error = copyin(uap->newset, &lmask, sizeof(l_sigset_t));
+	if (error)
+		return (error);
+
+	linux_to_bsd_sigset(&lmask, &sigmask);
+	return (kern_sigsuspend(td, sigmask));
+}
+
+int
+linux_pause(struct thread *td, struct linux_pause_args *args)
+{
+	struct proc *p = td->td_proc;
+	sigset_t sigmask;
+
+	LINUX_CTR(pause);
+
+	PROC_LOCK(p);
+	sigmask = td->td_sigmask;
+	PROC_UNLOCK(p);
+	return (kern_sigsuspend(td, sigmask));
+}
+
+int
+linux_sigaltstack(struct thread *td, struct linux_sigaltstack_args *uap)
+{
+	stack_t ss, oss;
+	l_stack_t lss;
+	int error;
+
+	LINUX_CTR2(sigaltstack, "%p, %p", uap->uss, uap->uoss);
+
+	if (uap->uss != NULL) {
+		error = copyin(uap->uss, &lss, sizeof(l_stack_t));
+		if (error)
+			return (error);
+
+		ss.ss_sp = PTRIN(lss.ss_sp);
+		ss.ss_size = lss.ss_size;
+		ss.ss_flags = linux_to_bsd_sigaltstack(lss.ss_flags);
+	}
+	error = kern_sigaltstack(td, (uap->uss != NULL) ? &ss : NULL,
+	    (uap->uoss != NULL) ? &oss : NULL);
+	if (!error && uap->uoss != NULL) {
+		lss.ss_sp = PTROUT(oss.ss_sp);
+		lss.ss_size = oss.ss_size;
+		lss.ss_flags = bsd_to_linux_sigaltstack(oss.ss_flags);
+		error = copyout(&lss, uap->uoss, sizeof(l_stack_t));
+	}
+
+	return (error);
+}
+
+/* XXX do all */
+int
+linux_arch_prctl(struct thread *td, struct linux_arch_prctl_args *args)
+{
+	int error;
+	struct pcb *pcb;
+
+	LINUX_CTR2(arch_prctl, "0x%x, %p", args->code, args->addr);
+
+	error = ENOTSUP;
+	pcb = td->td_pcb;
+
+	switch (args->code) {
+	case LINUX_ARCH_GET_GS:
+		error = copyout(&pcb->pcb_gsbase, (unsigned long *)args->addr,
+		    sizeof(args->addr));
+		break;
+	case LINUX_ARCH_SET_GS:
+		if (args->addr >= VM_MAXUSER_ADDRESS)
+			return(EPERM);
+		break;
+	case LINUX_ARCH_GET_FS:
+		error = copyout(&pcb->pcb_fsbase, (unsigned long *)args->addr,
+		    sizeof(args->addr));
+		break;
+	case LINUX_ARCH_SET_FS:
+		error = linux_set_cloned_tls(td, (void *)args->addr);
+		break;
+	default:
+		error = EINVAL;
+	}
+	return (error);
+}
+
+int
+linux_set_cloned_tls(struct thread *td, void *desc)
+{
+	struct pcb *pcb;
+
+	if ((uint64_t)desc >= VM_MAXUSER_ADDRESS)
+		return (EPERM);
+
+	pcb = td->td_pcb;
+	pcb->pcb_fsbase = (register_t)desc;
+	td->td_frame->tf_fs = _ufssel;
+
+	return (0);
+}
+
+void
+linux_to_bsd_sigset(l_sigset_t *lss, sigset_t *bss)
+{
+	int b, l;
+
+	SIGEMPTYSET(*bss);
+	for (l = 1; l <= LINUX_NSIG; l++) {
+		if (LINUX_SIGISMEMBER(*lss, l)) {
+			if (l <= LINUX_SIGTBLSZ)
+				b = linux_to_bsd_signal[_SIG_IDX(l)];
+			else
+				b = l;
+			if (b)
+				SIGADDSET(*bss, b);
+		}
+	}
+}
+
+void
+bsd_to_linux_sigset(sigset_t *bss, l_sigset_t *lss)
+{
+	int b, l;
+
+	LINUX_SIGEMPTYSET(*lss);
+	for (b = 1; b <= LINUX_NSIG; b++) {
+		if (SIGISMEMBER(*bss, b)) {
+			if (b <= LINUX_SIGTBLSZ)
+				l = bsd_to_linux_signal[_SIG_IDX(b)];
+			else
+				l = b;
+			if (l)
+				LINUX_SIGADDSET(*lss, l);
+		}
+	}
+}

sys/amd64/linux/linux_support.s

@@ -0,0 +1,124 @@
+/*-
+ * Copyright (c) 2007 Konstantin Belousov
+ * All rights reserved.
+ *
+ * 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.
+ * 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.
+ *
+ * $FreeBSD$
+ */
+
+#include "linux_assym.h"		/* system definitions */
+#include <machine/asmacros.h>		/* miscellaneous asm macros */
+
+#include "assym.s"
+
+futex_fault:
+	movq	$0,PCB_ONFAULT(%r8)
+	movl	$-EFAULT,%eax
+	ret
+
+ENTRY(futex_xchgl)
+	movq	PCPU(CURPCB),%r8
+	movq	$futex_fault,PCB_ONFAULT(%r8)
+	movq	$VM_MAXUSER_ADDRESS-4,%rax
+	cmpq	%rax,%rsi
+	ja	futex_fault
+	xchgq	%rdi,(%rsi)
+	movq	%rdi,(%rdx)
+	xorq	%rax,%rax
+	movq	%rax,PCB_ONFAULT(%r8)
+	ret
+
+ENTRY(futex_addl)
+	movq	PCPU(CURPCB),%r8
+	movq	$futex_fault,PCB_ONFAULT(%r8)
+	movq	$VM_MAXUSER_ADDRESS-4,%rax
+	cmpq	%rax,%rsi
+	ja	futex_fault
+#ifdef SMP
+	lock
+#endif
+	xaddq	%rdi,(%rsi)
+	movq	%rdi,(%rdx)
+	xorq	%rax,%rax
+	movq	%rax,PCB_ONFAULT(%r8)
+	ret
+
+ENTRY(futex_orl)
+	movq	PCPU(CURPCB),%r8
+	movq	$futex_fault,PCB_ONFAULT(%r8)
+	movq	$VM_MAXUSER_ADDRESS-4,%rax
+	cmpq	%rax,%rsi
+	ja	futex_fault
+	movq	(%rsi),%rax
+1:	movq	%rax,%rcx
+	orq	%rdi,%rcx
+#ifdef SMP
+	lock
+#endif
+	cmpxchgq %rcx,(%rsi)
+	jnz	1b
+	movq	%rax,(%rdx)
+	xorq	%rax,%rax
+	movq	%rax,PCB_ONFAULT(%r8)
+	ret
+
+ENTRY(futex_andl)
+	movq	PCPU(CURPCB),%r8
+	movq	$futex_fault,PCB_ONFAULT(%r8)
+	movq	$VM_MAXUSER_ADDRESS-4,%rax
+	cmpq	%rax,%rsi
+	ja	futex_fault
+	movq	(%rsi),%rax
+1:	movq	%rax,%rcx
+	andq	%rdi,%rcx
+#ifdef SMP
+	lock
+#endif
+	cmpxchgq %rcx,(%rsi)
+	jnz	1b
+	movq	%rax,(%rdx)
+	xorq	%rax,%rax
+	movq	%rax,PCB_ONFAULT(%r8)
+	ret
+
+ENTRY(futex_xorl)
+	movq	PCPU(CURPCB),%r8
+	movq	$futex_fault,PCB_ONFAULT(%r8)
+	movq	$VM_MAXUSER_ADDRESS-4,%rax
+	cmpq	%rax,%rsi
+	ja	futex_fault
+	movq	(%rsi),%rax
+1:	movq	%rax,%rcx
+	xorq	%rdi,%rcx
+#ifdef SMP
+	lock
+#endif
+	cmpxchgq %rcx,(%rsi)
+	jnz	1b
+	movq	%rax,(%rdx)
+	xorq	%rax,%rax
+	movq	%rax,PCB_ONFAULT(%r8)
+	ret

sys/amd64/linux/linux_sysvec.c

@@ -0,0 +1,942 @@
+/*-
+ * Copyright (c) 2013 Dmitry Chagin
+ * Copyright (c) 2004 Tim J. Robbins
+ * Copyright (c) 2003 Peter Wemm
+ * Copyright (c) 2002 Doug Rabson
+ * Copyright (c) 1998-1999 Andrew Gallatin
+ * Copyright (c) 1994-1996 Søren Schmidt
+ * All rights reserved.
+ *
+ * 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
+ *    in this position and unchanged.
+ * 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. The name of the author may not be used to endorse or promote products
+ *    derived from this software without specific prior written permission
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 AUTHOR 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.
+ */
+
+#include <sys/cdefs.h>
+__FBSDID("$FreeBSD$");
+
+#include "opt_compat.h"
+
+#define	__ELF_WORD_SIZE	64
+
+#include <sys/param.h>
+#include <sys/systm.h>
+#include <sys/exec.h>
+#include <sys/fcntl.h>
+#include <sys/imgact.h>
+#include <sys/imgact_elf.h>
+#include <sys/kernel.h>
+#include <sys/ktr.h>
+#include <sys/lock.h>
+#include <sys/malloc.h>
+#include <sys/module.h>
+#include <sys/mutex.h>
+#include <sys/proc.h>
+#include <sys/resourcevar.h>
+#include <sys/signalvar.h>
+#include <sys/sysctl.h>
+#include <sys/syscallsubr.h>
+#include <sys/sysent.h>
+#include <sys/sysproto.h>
+#include <sys/vnode.h>
+#include <sys/eventhandler.h>
+
+#include <vm/vm.h>
+#include <vm/pmap.h>
+#include <vm/vm_extern.h>
+#include <vm/vm_map.h>
+#include <vm/vm_object.h>
+#include <vm/vm_page.h>
+#include <vm/vm_param.h>
+
+#include <machine/cpu.h>
+#include <machine/md_var.h>
+#include <machine/pcb.h>
+#include <machine/specialreg.h>
+
+#include <amd64/linux/linux.h>
+#include <amd64/linux/linux_proto.h>
+#include <compat/linux/linux_emul.h>
+#include <compat/linux/linux_futex.h>
+#include <compat/linux/linux_ioctl.h>
+#include <compat/linux/linux_mib.h>
+#include <compat/linux/linux_misc.h>
+#include <compat/linux/linux_signal.h>
+#include <compat/linux/linux_sysproto.h>
+#include <compat/linux/linux_util.h>
+#include <compat/linux/linux_vdso.h>
+
+MODULE_VERSION(linux64, 1);
+
+#if BYTE_ORDER == LITTLE_ENDIAN
+#define SHELLMAGIC      0x2123 /* #! */
+#else
+#define SHELLMAGIC      0x2321
+#endif
+
+#if defined(DEBUG)
+SYSCTL_PROC(_compat_linux, OID_AUTO, debug,
+	    CTLTYPE_STRING | CTLFLAG_RW,
+	    0, 0, linux_sysctl_debug, "A",
+	    "Linux 64 debugging control");
+#endif
+
+/*
+ * Allow the this functions to use the ldebug() facility
+ * even though they are not syscalls themselves. Map them
+ * to syscall 0. This is slightly less bogus than using
+ * ldebug(sigreturn).
+ */
+#define	LINUX_SYS_linux_rt_sendsig	0
+
+const char *linux_kplatform;
+static int linux_szsigcode;
+static vm_object_t linux_shared_page_obj;
+static char *linux_shared_page_mapping;
+extern char _binary_linux_locore_o_start;
+extern char _binary_linux_locore_o_end;
+
+extern struct sysent linux_sysent[LINUX_SYS_MAXSYSCALL];
+
+SET_DECLARE(linux_ioctl_handler_set, struct linux_ioctl_handler);
+
+static register_t * linux_copyout_strings(struct image_params *imgp);
+static int	elf_linux_fixup(register_t **stack_base,
+		    struct image_params *iparams);
+static boolean_t linux_trans_osrel(const Elf_Note *note, int32_t *osrel);
+static void	linux_vdso_install(void *param);
+static void	linux_vdso_deinstall(void *param);
+static void	linux_set_syscall_retval(struct thread *td, int error);
+static int	linux_fetch_syscall_args(struct thread *td, struct syscall_args *sa);
+static void	linux_exec_setregs(struct thread *td, struct image_params *imgp,
+		    u_long stack);
+
+/*
+ * Linux syscalls return negative errno's, we do positive and map them
+ * Reference:
+ *   FreeBSD: src/sys/sys/errno.h
+ *   Linux:   linux-2.6.17.8/include/asm-generic/errno-base.h
+ *            linux-2.6.17.8/include/asm-generic/errno.h
+ */
+static int bsd_to_linux_errno[ELAST + 1] = {
+	-0,  -1,  -2,  -3,  -4,  -5,  -6,  -7,  -8,  -9,
+	-10, -35, -12, -13, -14, -15, -16, -17, -18, -19,
+	-20, -21, -22, -23, -24, -25, -26, -27, -28, -29,
+	-30, -31, -32, -33, -34, -11,-115,-114, -88, -89,
+	-90, -91, -92, -93, -94, -95, -96, -97, -98, -99,
+	-100,-101,-102,-103,-104,-105,-106,-107,-108,-109,
+	-110,-111, -40, -36,-112,-113, -39, -11, -87,-122,
+	-116, -66,  -6,  -6,  -6,  -6,  -6, -37, -38,  -9,
+	  -6,  -6, -43, -42, -75,-125, -84, -95, -16, -74,
+	 -72, -67, -71
+};
+
+int bsd_to_linux_signal[LINUX_SIGTBLSZ] = {
+	LINUX_SIGHUP, LINUX_SIGINT, LINUX_SIGQUIT, LINUX_SIGILL,
+	LINUX_SIGTRAP, LINUX_SIGABRT, 0, LINUX_SIGFPE,
+	LINUX_SIGKILL, LINUX_SIGBUS, LINUX_SIGSEGV, LINUX_SIGSYS,
+	LINUX_SIGPIPE, LINUX_SIGALRM, LINUX_SIGTERM, LINUX_SIGURG,
+	LINUX_SIGSTOP, LINUX_SIGTSTP, LINUX_SIGCONT, LINUX_SIGCHLD,
+	LINUX_SIGTTIN, LINUX_SIGTTOU, LINUX_SIGIO, LINUX_SIGXCPU,
+	LINUX_SIGXFSZ, LINUX_SIGVTALRM, LINUX_SIGPROF, LINUX_SIGWINCH,
+	0, LINUX_SIGUSR1, LINUX_SIGUSR2
+};
+
+int linux_to_bsd_signal[LINUX_SIGTBLSZ] = {
+	SIGHUP, SIGINT, SIGQUIT, SIGILL,
+	SIGTRAP, SIGABRT, SIGBUS, SIGFPE,
+	SIGKILL, SIGUSR1, SIGSEGV, SIGUSR2,
+	SIGPIPE, SIGALRM, SIGTERM, SIGBUS,
+	SIGCHLD, SIGCONT, SIGSTOP, SIGTSTP,
+	SIGTTIN, SIGTTOU, SIGURG, SIGXCPU,
+	SIGXFSZ, SIGVTALRM, SIGPROF, SIGWINCH,
+	SIGIO, SIGURG, SIGSYS
+};
+
+#define LINUX_T_UNKNOWN  255
+static int _bsd_to_linux_trapcode[] = {
+	LINUX_T_UNKNOWN,	/* 0 */
+	6,			/* 1  T_PRIVINFLT */
+	LINUX_T_UNKNOWN,	/* 2 */
+	3,			/* 3  T_BPTFLT */
+	LINUX_T_UNKNOWN,	/* 4 */
+	LINUX_T_UNKNOWN,	/* 5 */
+	16,			/* 6  T_ARITHTRAP */
+	254,			/* 7  T_ASTFLT */
+	LINUX_T_UNKNOWN,	/* 8 */
+	13,			/* 9  T_PROTFLT */
+	1,			/* 10 T_TRCTRAP */
+	LINUX_T_UNKNOWN,	/* 11 */
+	14,			/* 12 T_PAGEFLT */
+	LINUX_T_UNKNOWN,	/* 13 */
+	17,			/* 14 T_ALIGNFLT */
+	LINUX_T_UNKNOWN,	/* 15 */
+	LINUX_T_UNKNOWN,	/* 16 */
+	LINUX_T_UNKNOWN,	/* 17 */
+	0,			/* 18 T_DIVIDE */
+	2,			/* 19 T_NMI */
+	4,			/* 20 T_OFLOW */
+	5,			/* 21 T_BOUND */
+	7,			/* 22 T_DNA */
+	8,			/* 23 T_DOUBLEFLT */
+	9,			/* 24 T_FPOPFLT */
+	10,			/* 25 T_TSSFLT */
+	11,			/* 26 T_SEGNPFLT */
+	12,			/* 27 T_STKFLT */
+	18,			/* 28 T_MCHK */
+	19,			/* 29 T_XMMFLT */
+	15			/* 30 T_RESERVED */
+};
+#define bsd_to_linux_trapcode(code) \
+    ((code)<sizeof(_bsd_to_linux_trapcode)/sizeof(*_bsd_to_linux_trapcode)? \
+     _bsd_to_linux_trapcode[(code)]: \
+     LINUX_T_UNKNOWN)
+
+LINUX_VDSO_SYM_INTPTR(linux_rt_sigcode);
+LINUX_VDSO_SYM_CHAR(linux_platform);
+
+/*
+ * If FreeBSD & Linux have a difference of opinion about what a trap
+ * means, deal with it here.
+ *
+ * MPSAFE
+ */
+static int
+translate_traps(int signal, int trap_code)
+{
+
+	if (signal != SIGBUS)
+		return signal;
+	switch (trap_code) {
+	case T_PROTFLT:
+	case T_TSSFLT:
+	case T_DOUBLEFLT:
+	case T_PAGEFLT:
+		return SIGSEGV;
+	default:
+		return signal;
+	}
+}
+
+static int
+linux_fetch_syscall_args(struct thread *td, struct syscall_args *sa)
+{
+	struct proc *p;
+	struct trapframe *frame;
+
+	p = td->td_proc;
+	frame = td->td_frame;
+
+	sa->args[0] = frame->tf_rdi;
+	sa->args[1] = frame->tf_rsi;
+	sa->args[2] = frame->tf_rdx;
+	sa->args[3] = frame->tf_rcx;
+	sa->args[4] = frame->tf_r8;
+	sa->args[5] = frame->tf_r9;
+	sa->code = frame->tf_rax;
+
+	if (sa->code >= p->p_sysent->sv_size) {
+		PROC_LOCK(p);
+		sigexit(td, SIGILL);
+	} else
+		sa->callp = &p->p_sysent->sv_table[sa->code];
+	sa->narg = sa->callp->sy_narg;
+
+	td->td_retval[0] = 0;
+	return (0);
+}
+
+static void
+linux_set_syscall_retval(struct thread *td, int error)
+{
+	struct trapframe *frame = td->td_frame;
+
+	/*
+	 * On Linux only %rcx and %r11 values are not preserved across
+	 * the syscall.
+	 * So, do not clobber %rdx and %r10
+	 */
+	td->td_retval[1] = frame->tf_rdx;
+	frame->tf_r10 = frame->tf_rcx;
+
+	cpu_set_syscall_retval(td, error);
+
+	 /* Restore all registers. */
+	set_pcb_flags(td->td_pcb, PCB_FULL_IRET);
+}
+
+static int
+elf_linux_fixup(register_t **stack_base, struct image_params *imgp)
+{
+	Elf_Auxargs *args;
+	Elf_Addr *base;
+	Elf_Addr *pos;
+	struct ps_strings *arginfo;
+	struct proc *p;
+
+	p = imgp->proc;
+	arginfo = (struct ps_strings *)p->p_sysent->sv_psstrings;
+
+	KASSERT(curthread->td_proc == imgp->proc,
+	    ("unsafe elf_linux_fixup(), should be curproc"));
+	base = (Elf64_Addr *)*stack_base;
+	args = (Elf64_Auxargs *)imgp->auxargs;
+	pos = base + (imgp->args->argc + imgp->args->envc + 2);
+
+	AUXARGS_ENTRY(pos, LINUX_AT_SYSINFO_EHDR,
+	    imgp->proc->p_sysent->sv_shared_page_base);
+	AUXARGS_ENTRY(pos, LINUX_AT_HWCAP, cpu_feature);
+	AUXARGS_ENTRY(pos, LINUX_AT_CLKTCK, stclohz);
+	AUXARGS_ENTRY(pos, AT_PHDR, args->phdr);
+	AUXARGS_ENTRY(pos, AT_PHENT, args->phent);
+	AUXARGS_ENTRY(pos, AT_PHNUM, args->phnum);
+	AUXARGS_ENTRY(pos, AT_PAGESZ, args->pagesz);
+	AUXARGS_ENTRY(pos, AT_BASE, args->base);
+	AUXARGS_ENTRY(pos, AT_FLAGS, args->flags);
+	AUXARGS_ENTRY(pos, AT_ENTRY, args->entry);
+	AUXARGS_ENTRY(pos, AT_UID, imgp->proc->p_ucred->cr_ruid);
+	AUXARGS_ENTRY(pos, AT_EUID, imgp->proc->p_ucred->cr_svuid);
+	AUXARGS_ENTRY(pos, AT_GID, imgp->proc->p_ucred->cr_rgid);
+	AUXARGS_ENTRY(pos, AT_EGID, imgp->proc->p_ucred->cr_svgid);
+	AUXARGS_ENTRY(pos, LINUX_AT_SECURE, 0);
+	AUXARGS_ENTRY(pos, LINUX_AT_PLATFORM, PTROUT(linux_platform));
+	if (args->execfd != -1)
+		AUXARGS_ENTRY(pos, AT_EXECFD, args->execfd);
+	AUXARGS_ENTRY(pos, AT_NULL, 0);
+	free(imgp->auxargs, M_TEMP);
+	imgp->auxargs = NULL;
+
+	base--;
+	suword(base, (uint64_t)imgp->args->argc);
+
+	*stack_base = (register_t *)base;
+	return (0);
+}
+
+/*
+ * Copy strings out to the new process address space, constructing new arg
+ * and env vector tables. Return a pointer to the base so that it can be used
+ * as the initial stack pointer.
+ */
+static register_t *
+linux_copyout_strings(struct image_params *imgp)
+{
+	int argc, envc;
+	char **vectp;
+	char *stringp, *destp;
+	register_t *stack_base;
+	struct ps_strings *arginfo;
+	struct proc *p;
+
+	/*
+	 * Calculate string base and vector table pointers.
+	 */
+	p = imgp->proc;
+	arginfo = (struct ps_strings *)p->p_sysent->sv_psstrings;
+	destp =	(caddr_t)arginfo - SPARE_USRSPACE -
+	    roundup((ARG_MAX - imgp->args->stringspace), sizeof(char *));
+
+	/*
+	 * If we have a valid auxargs ptr, prepare some room
+	 * on the stack.
+	 */
+	if (imgp->auxargs) {
+		/*
+		 * 'AT_COUNT*2' is size for the ELF Auxargs data. This is for
+		 * lower compatibility.
+		 */
+		imgp->auxarg_size = (imgp->auxarg_size) ? imgp->auxarg_size :
+		    (LINUX_AT_COUNT * 2);
+
+		/*
+		 * The '+ 2' is for the null pointers at the end of each of
+		 * the arg and env vector sets,and imgp->auxarg_size is room
+		 * for argument of Runtime loader.
+		 */
+		vectp = (char **)(destp - (imgp->args->argc +
+		    imgp->args->envc + 2 + imgp->auxarg_size) * sizeof(char *));
+
+	} else {
+		/*
+		 * The '+ 2' is for the null pointers at the end of each of
+		 * the arg and env vector sets
+		 */
+		vectp = (char **)(destp - (imgp->args->argc +
+		    imgp->args->envc + 2) * sizeof(char *));
+	}
+
+	/*
+	 * vectp also becomes our initial stack base
+	 */
+	stack_base = (register_t *)vectp;
+
+	stringp = imgp->args->begin_argv;
+	argc = imgp->args->argc;
+	envc = imgp->args->envc;
+
+	/*
+	 * Copy out strings - arguments and environment.
+	 */
+	copyout(stringp, destp, ARG_MAX - imgp->args->stringspace);
+
+	/*
+	 * Fill in "ps_strings" struct for ps, w, etc.
+	 */
+	suword(&arginfo->ps_argvstr, (long)(intptr_t)vectp);
+	suword(&arginfo->ps_nargvstr, argc);
+
+	/*
+	 * Fill in argument portion of vector table.
+	 */
+	for (; argc > 0; --argc) {
+		suword(vectp++, (long)(intptr_t)destp);
+		while (*stringp++ != 0)
+			destp++;
+		destp++;
+	}
+
+	/* a null vector table pointer separates the argp's from the envp's */
+	suword(vectp++, 0);
+
+	suword(&arginfo->ps_envstr, (long)(intptr_t)vectp);
+	suword(&arginfo->ps_nenvstr, envc);
+
+	/*
+	 * Fill in environment portion of vector table.
+	 */
+	for (; envc > 0; --envc) {
+		suword(vectp++, (long)(intptr_t)destp);
+		while (*stringp++ != 0)
+			destp++;
+		destp++;
+	}
+
+	/* end of vector table is a null pointer */
+	suword(vectp, 0);
+	return (stack_base);
+}
+
+/*
+ * Reset registers to default values on exec.
+ */
+static void
+linux_exec_setregs(struct thread *td, struct image_params *imgp, u_long stack)
+{
+	struct trapframe *regs = td->td_frame;
+	struct pcb *pcb = td->td_pcb;
+
+	mtx_lock(&dt_lock);
+	if (td->td_proc->p_md.md_ldt != NULL)
+		user_ldt_free(td);
+	else
+		mtx_unlock(&dt_lock);
+
+	pcb->pcb_fsbase = 0;
+	pcb->pcb_gsbase = 0;
+	clear_pcb_flags(pcb, PCB_32BIT);
+	pcb->pcb_initial_fpucw = __LINUX_NPXCW__;
+	set_pcb_flags(pcb, PCB_FULL_IRET);
+
+	bzero((char *)regs, sizeof(struct trapframe));
+	regs->tf_rip = imgp->entry_addr;
+	regs->tf_rsp = stack;
+	regs->tf_rflags = PSL_USER | (regs->tf_rflags & PSL_T);
+	regs->tf_ss = _udatasel;
+	regs->tf_cs = _ucodesel;
+	regs->tf_ds = _udatasel;
+	regs->tf_es = _udatasel;
+	regs->tf_fs = _ufssel;
+	regs->tf_gs = _ugssel;
+	regs->tf_flags = TF_HASSEGS;
+
+	/*
+	 * Reset the hardware debug registers if they were in use.
+	 * They won't have any meaning for the newly exec'd process.
+	 */
+	if (pcb->pcb_flags & PCB_DBREGS) {
+		pcb->pcb_dr0 = 0;
+		pcb->pcb_dr1 = 0;
+		pcb->pcb_dr2 = 0;
+		pcb->pcb_dr3 = 0;
+		pcb->pcb_dr6 = 0;
+		pcb->pcb_dr7 = 0;
+		if (pcb == curpcb) {
+			/*
+			 * Clear the debug registers on the running
+			 * CPU, otherwise they will end up affecting
+			 * the next process we switch to.
+			 */
+			reset_dbregs();
+		}
+		clear_pcb_flags(pcb, PCB_DBREGS);
+	}
+
+	/*
+	 * Drop the FP state if we hold it, so that the process gets a
+	 * clean FP state if it uses the FPU again.
+	 */
+	fpstate_drop(td);
+}
+
+/*
+ * Copied from amd64/amd64/machdep.c
+ *
+ * XXX fpu state need? don't think so
+ */
+int
+linux_rt_sigreturn(struct thread *td, struct linux_rt_sigreturn_args *args)
+{
+	struct proc *p;
+	struct l_ucontext uc;
+	struct l_sigcontext *context;
+	struct trapframe *regs;
+	unsigned long rflags;
+	int error;
+	ksiginfo_t ksi;
+
+	regs = td->td_frame;
+	error = copyin((void *)regs->tf_rbx, &uc, sizeof(uc));
+	if (error != 0)
+		return (error);
+
+	p = td->td_proc;
+	context = &uc.uc_mcontext;
+	rflags = context->sc_rflags;
+
+	/*
+	 * Don't allow users to change privileged or reserved flags.
+	 */
+	/*
+	 * XXX do allow users to change the privileged flag PSL_RF.
+	 * The cpu sets PSL_RF in tf_rflags for faults.  Debuggers
+	 * should sometimes set it there too.  tf_rflags is kept in
+	 * the signal context during signal handling and there is no
+	 * other place to remember it, so the PSL_RF bit may be
+	 * corrupted by the signal handler without us knowing.
+	 * Corruption of the PSL_RF bit at worst causes one more or
+	 * one less debugger trap, so allowing it is fairly harmless.
+	 */
+
+#define RFLAG_SECURE(ef, oef)     ((((ef) ^ (oef)) & ~PSL_USERCHANGE) == 0)
+	if (!RFLAG_SECURE(rflags & ~PSL_RF, regs->tf_rflags & ~PSL_RF)) {
+		printf("linux_rt_sigreturn: rflags = 0x%lx\n", rflags);
+		return (EINVAL);
+	}
+
+	/*
+	 * Don't allow users to load a valid privileged %cs.  Let the
+	 * hardware check for invalid selectors, excess privilege in
+	 * other selectors, invalid %eip's and invalid %esp's.
+	 */
+#define CS_SECURE(cs)           (ISPL(cs) == SEL_UPL)
+	if (!CS_SECURE(context->sc_cs)) {
+		printf("linux_rt_sigreturn: cs = 0x%x\n", context->sc_cs);
+		ksiginfo_init_trap(&ksi);
+		ksi.ksi_signo = SIGBUS;
+		ksi.ksi_code = BUS_OBJERR;
+		ksi.ksi_trapno = T_PROTFLT;
+		ksi.ksi_addr = (void *)regs->tf_rip;
+		trapsignal(td, &ksi);
+		return (EINVAL);
+	}
+
+	PROC_LOCK(p);
+	linux_to_bsd_sigset(&uc.uc_sigmask, &td->td_sigmask);
+	SIG_CANTMASK(td->td_sigmask);
+	signotify(td);
+	PROC_UNLOCK(p);
+
+	regs->tf_rdi    = context->sc_rdi;
+	regs->tf_rsi    = context->sc_rsi;
+	regs->tf_rdx    = context->sc_rdx;
+	regs->tf_rbp    = context->sc_rbp;
+	regs->tf_rbx    = context->sc_rbx;
+	regs->tf_rcx    = context->sc_rcx;
+	regs->tf_rax    = context->sc_rax;
+	regs->tf_rip    = context->sc_rip;
+	regs->tf_rsp    = context->sc_rsp;
+	regs->tf_r8     = context->sc_r8;
+	regs->tf_r9     = context->sc_r9;
+	regs->tf_r10    = context->sc_r10;
+	regs->tf_r11    = context->sc_r11;
+	regs->tf_r12    = context->sc_r12;
+	regs->tf_r13    = context->sc_r13;
+	regs->tf_r14    = context->sc_r14;
+	regs->tf_r15    = context->sc_r15;
+	regs->tf_cs     = context->sc_cs;
+	regs->tf_err    = context->sc_err;
+	regs->tf_rflags = rflags;
+
+	set_pcb_flags(td->td_pcb, PCB_FULL_IRET);
+	return (EJUSTRETURN);
+}
+
+/*
+ * copied from amd64/amd64/machdep.c
+ *
+ * Send an interrupt to process.
+ */
+static void
+linux_rt_sendsig(sig_t catcher, ksiginfo_t *ksi, sigset_t *mask)
+{
+	struct l_rt_sigframe sf, *sfp;
+	struct proc *p;
+	struct thread *td;
+	struct sigacts *psp;
+	caddr_t sp;
+	struct trapframe *regs;
+	int sig, code;
+	int oonstack;
+
+	td = curthread;
+	p = td->td_proc;
+	PROC_LOCK_ASSERT(p, MA_OWNED);
+	sig = ksi->ksi_signo;
+	psp = p->p_sigacts;
+	code = ksi->ksi_code;
+	mtx_assert(&psp->ps_mtx, MA_OWNED);
+	regs = td->td_frame;
+	oonstack = sigonstack(regs->tf_rsp);
+
+	LINUX_CTR4(rt_sendsig, "%p, %d, %p, %u",
+	    catcher, sig, mask, code);
+
+	/* Allocate space for the signal handler context. */
+	if ((td->td_pflags & TDP_ALTSTACK) != 0 && !oonstack &&
+	    SIGISMEMBER(psp->ps_sigonstack, sig)) {
+		sp = td->td_sigstk.ss_sp + td->td_sigstk.ss_size -
+		    sizeof(struct l_rt_sigframe);
+	} else
+		sp = (caddr_t)regs->tf_rsp - sizeof(struct l_rt_sigframe) - 128;
+	/* Align to 16 bytes. */
+	sfp = (struct l_rt_sigframe *)((unsigned long)sp & ~0xFul);
+	mtx_unlock(&psp->ps_mtx);
+
+	/* Translate the signal if appropriate. */
+	sig = BSD_TO_LINUX_SIGNAL(sig);
+
+	/* Save user context. */
+	bzero(&sf, sizeof(sf));
+	bsd_to_linux_sigset(mask, &sf.sf_sc.uc_sigmask);
+	bsd_to_linux_sigset(mask, &sf.sf_sc.uc_mcontext.sc_mask);
+
+	sf.sf_sc.uc_stack.ss_sp = PTROUT(td->td_sigstk.ss_sp);
+	sf.sf_sc.uc_stack.ss_size = td->td_sigstk.ss_size;
+	sf.sf_sc.uc_stack.ss_flags = (td->td_pflags & TDP_ALTSTACK)
+	    ? ((oonstack) ? LINUX_SS_ONSTACK : 0) : LINUX_SS_DISABLE;
+	PROC_UNLOCK(p);
+
+	sf.sf_sc.uc_mcontext.sc_rdi    = regs->tf_rdi;
+	sf.sf_sc.uc_mcontext.sc_rsi    = regs->tf_rsi;
+	sf.sf_sc.uc_mcontext.sc_rdx    = regs->tf_rdx;
+	sf.sf_sc.uc_mcontext.sc_rbp    = regs->tf_rbp;
+	sf.sf_sc.uc_mcontext.sc_rbx    = regs->tf_rbx;
+	sf.sf_sc.uc_mcontext.sc_rcx    = regs->tf_rcx;
+	sf.sf_sc.uc_mcontext.sc_rax    = regs->tf_rax;
+	sf.sf_sc.uc_mcontext.sc_rip    = regs->tf_rip;
+	sf.sf_sc.uc_mcontext.sc_rsp    = regs->tf_rsp;
+	sf.sf_sc.uc_mcontext.sc_r8     = regs->tf_r8;
+	sf.sf_sc.uc_mcontext.sc_r9     = regs->tf_r9;
+	sf.sf_sc.uc_mcontext.sc_r10    = regs->tf_r10;
+	sf.sf_sc.uc_mcontext.sc_r11    = regs->tf_r11;
+	sf.sf_sc.uc_mcontext.sc_r12    = regs->tf_r12;
+	sf.sf_sc.uc_mcontext.sc_r13    = regs->tf_r13;
+	sf.sf_sc.uc_mcontext.sc_r14    = regs->tf_r14;
+	sf.sf_sc.uc_mcontext.sc_r15    = regs->tf_r15;
+	sf.sf_sc.uc_mcontext.sc_cs     = regs->tf_cs;
+	sf.sf_sc.uc_mcontext.sc_rflags = regs->tf_rflags;
+	sf.sf_sc.uc_mcontext.sc_err    = regs->tf_err;
+	sf.sf_sc.uc_mcontext.sc_trapno = bsd_to_linux_trapcode(code);
+	sf.sf_sc.uc_mcontext.sc_cr2    = (register_t)ksi->ksi_addr;
+
+	/* Build the argument list for the signal handler. */
+	regs->tf_rdi = sig;			/* arg 1 in %rdi */
+	regs->tf_rax = 0;
+	regs->tf_rsi = (register_t)&sfp->sf_si;	/* arg 2 in %rsi */
+	regs->tf_rdx = (register_t)&sfp->sf_sc;	/* arg 3 in %rdx */
+
+	sf.sf_handler = catcher;
+	/* Fill in POSIX parts */
+	ksiginfo_to_lsiginfo(ksi, &sf.sf_si, sig);
+
+	/*
+	 * Copy the sigframe out to the user's stack.
+	 */
+	if (copyout(&sf, sfp, sizeof(*sfp)) != 0) {
+#ifdef DEBUG
+		printf("process %ld has trashed its stack\n", (long)p->p_pid);
+#endif
+		PROC_LOCK(p);
+		sigexit(td, SIGILL);
+	}
+
+	regs->tf_rsp = (long)sfp;
+	regs->tf_rip = linux_rt_sigcode;
+	regs->tf_rflags &= ~(PSL_T | PSL_D);
+	regs->tf_cs = _ucodesel;
+	set_pcb_flags(td->td_pcb, PCB_FULL_IRET);
+	PROC_LOCK(p);
+	mtx_lock(&psp->ps_mtx);
+}
+
+/*
+ * If a linux binary is exec'ing something, try this image activator
+ * first.  We override standard shell script execution in order to
+ * be able to modify the interpreter path.  We only do this if a linux
+ * binary is doing the exec, so we do not create an EXEC module for it.
+ */
+static int exec_linux_imgact_try(struct image_params *iparams);
+
+static int
+exec_linux_imgact_try(struct image_params *imgp)
+{
+	const char *head = (const char *)imgp->image_header;
+	char *rpath;
+	int error = -1, len;
+
+	/*
+	 * The interpreter for shell scripts run from a linux binary needs
+	 * to be located in /compat/linux if possible in order to recursively
+	 * maintain linux path emulation.
+	 */
+	if (((const short *)head)[0] == SHELLMAGIC) {
+		/*
+		 * Run our normal shell image activator.  If it succeeds
+		 * attempt to use the alternate path for the interpreter.
+		 * If an alternate path is found, use our stringspace
+		 * to store it.
+		 */
+		if ((error = exec_shell_imgact(imgp)) == 0) {
+			linux_emul_convpath(FIRST_THREAD_IN_PROC(imgp->proc),
+			    imgp->interpreter_name, UIO_SYSSPACE,
+			    &rpath, 0, AT_FDCWD);
+			if (rpath != NULL) {
+				len = strlen(rpath) + 1;
+
+				if (len <= MAXSHELLCMDLEN)
+					memcpy(imgp->interpreter_name,
+					    rpath, len);
+				free(rpath, M_TEMP);
+			}
+		}
+	}
+	return(error);
+}
+
+struct sysentvec elf_linux_sysvec = {
+	.sv_size	= LINUX_SYS_MAXSYSCALL,
+	.sv_table	= linux_sysent,
+	.sv_mask	= 0,
+	.sv_sigsize	= LINUX_SIGTBLSZ,
+	.sv_sigtbl	= bsd_to_linux_signal,
+	.sv_errsize	= ELAST + 1,
+	.sv_errtbl	= bsd_to_linux_errno,
+	.sv_transtrap	= translate_traps,
+	.sv_fixup	= elf_linux_fixup,
+	.sv_sendsig	= linux_rt_sendsig,
+	.sv_sigcode	= &_binary_linux_locore_o_start,
+	.sv_szsigcode	= &linux_szsigcode,
+	.sv_prepsyscall	= NULL,
+	.sv_name	= "Linux ELF64",
+	.sv_coredump	= elf64_coredump,
+	.sv_imgact_try	= exec_linux_imgact_try,
+	.sv_minsigstksz	= LINUX_MINSIGSTKSZ,
+	.sv_pagesize	= PAGE_SIZE,
+	.sv_minuser	= VM_MIN_ADDRESS,
+	.sv_maxuser	= VM_MAXUSER_ADDRESS,
+	.sv_usrstack	= USRSTACK,
+	.sv_psstrings	= PS_STRINGS,
+	.sv_stackprot	= VM_PROT_ALL,
+	.sv_copyout_strings = linux_copyout_strings,
+	.sv_setregs	= linux_exec_setregs,
+	.sv_fixlimit	= NULL,
+	.sv_maxssiz	= NULL,
+	.sv_flags	= SV_ABI_LINUX | SV_LP64 | SV_SHP,
+	.sv_set_syscall_retval = linux_set_syscall_retval,
+	.sv_fetch_syscall_args = linux_fetch_syscall_args,
+	.sv_syscallnames = NULL,
+	.sv_shared_page_base = SHAREDPAGE,
+	.sv_shared_page_len = PAGE_SIZE,
+	.sv_schedtail	= linux_schedtail,
+	.sv_thread_detach = linux_thread_detach
+};
+
+static void
+linux_vdso_install(void *param)
+{
+
+	linux_szsigcode = (&_binary_linux_locore_o_end - 
+	    &_binary_linux_locore_o_start);
+
+	if (linux_szsigcode > elf_linux_sysvec.sv_shared_page_len)
+		panic("Linux invalid vdso size\n");
+
+	__elfN(linux_vdso_fixup)(&elf_linux_sysvec);
+
+	linux_shared_page_obj = __elfN(linux_shared_page_init)
+	    (&linux_shared_page_mapping);
+
+	__elfN(linux_vdso_reloc)(&elf_linux_sysvec, SHAREDPAGE);
+
+	bcopy(elf_linux_sysvec.sv_sigcode, linux_shared_page_mapping,
+	    linux_szsigcode);
+	elf_linux_sysvec.sv_shared_page_obj = linux_shared_page_obj;
+
+	linux_kplatform = linux_shared_page_mapping +
+	    (linux_platform - (caddr_t)SHAREDPAGE);
+}
+SYSINIT(elf_linux_vdso_init, SI_SUB_EXEC, SI_ORDER_ANY,
+    (sysinit_cfunc_t)linux_vdso_install, NULL);
+
+static void
+linux_vdso_deinstall(void *param)
+{
+
+	__elfN(linux_shared_page_fini)(linux_shared_page_obj);
+};
+SYSUNINIT(elf_linux_vdso_uninit, SI_SUB_EXEC, SI_ORDER_FIRST,
+    (sysinit_cfunc_t)linux_vdso_deinstall, NULL);
+
+static char GNULINUX_ABI_VENDOR[] = "GNU";
+static int GNULINUX_ABI_DESC = 0;
+
+static boolean_t
+linux_trans_osrel(const Elf_Note *note, int32_t *osrel)
+{
+	const Elf32_Word *desc;
+	uintptr_t p;
+
+	p = (uintptr_t)(note + 1);
+	p += roundup2(note->n_namesz, sizeof(Elf32_Addr));
+
+	desc = (const Elf32_Word *)p;
+	if (desc[0] != GNULINUX_ABI_DESC)
+		return (FALSE);
+
+	/*
+	 * For linux we encode osrel as follows (see linux_mib.c):
+	 * VVVMMMIII (version, major, minor), see linux_mib.c.
+	 */
+	*osrel = desc[1] * 1000000 + desc[2] * 1000 + desc[3];
+
+	return (TRUE);
+}
+
+static Elf_Brandnote linux64_brandnote = {
+	.hdr.n_namesz	= sizeof(GNULINUX_ABI_VENDOR),
+	.hdr.n_descsz	= 16,
+	.hdr.n_type	= 1,
+	.vendor		= GNULINUX_ABI_VENDOR,
+	.flags		= BN_TRANSLATE_OSREL,
+	.trans_osrel	= linux_trans_osrel
+};
+
+static Elf64_Brandinfo linux_glibc2brand = {
+	.brand		= ELFOSABI_LINUX,
+	.machine	= EM_X86_64,
+	.compat_3_brand	= "Linux",
+	.emul_path	= "/compat/linux",
+	.interp_path	= "/lib64/ld-linux-x86-64.so.2",
+	.sysvec		= &elf_linux_sysvec,
+	.interp_newpath	= NULL,
+	.brand_note	= &linux64_brandnote,
+	.flags		= BI_CAN_EXEC_DYN | BI_BRAND_NOTE
+};
+
+static Elf64_Brandinfo linux_glibc2brandshort = {
+	.brand		= ELFOSABI_LINUX,
+	.machine	= EM_X86_64,
+	.compat_3_brand	= "Linux",
+	.emul_path	= "/compat/linux",
+	.interp_path	= "/lib64/ld-linux.so.2",
+	.sysvec		= &elf_linux_sysvec,
+	.interp_newpath	= NULL,
+	.brand_note	= &linux64_brandnote,
+	.flags		= BI_CAN_EXEC_DYN | BI_BRAND_NOTE
+};
+
+Elf64_Brandinfo *linux_brandlist[] = {
+	&linux_glibc2brand,
+	&linux_glibc2brandshort,
+	NULL
+};
+
+static int
+linux64_elf_modevent(module_t mod, int type, void *data)
+{
+	Elf64_Brandinfo **brandinfo;
+	int error;
+	struct linux_ioctl_handler **lihp;
+
+	error = 0;
+
+	switch(type) {
+	case MOD_LOAD:
+		for (brandinfo = &linux_brandlist[0]; *brandinfo != NULL;
+		     ++brandinfo)
+			if (elf64_insert_brand_entry(*brandinfo) < 0)
+				error = EINVAL;
+		if (error == 0) {
+			SET_FOREACH(lihp, linux_ioctl_handler_set)
+				linux_ioctl_register_handler(*lihp);
+			LIST_INIT(&futex_list);
+			mtx_init(&futex_mtx, "ftllk64", NULL, MTX_DEF);
+			stclohz = (stathz ? stathz : hz);
+			if (bootverbose)
+				printf("Linux x86-64 ELF exec handler installed\n");
+		} else
+			printf("cannot insert Linux x86-64 ELF brand handler\n");
+		break;
+	case MOD_UNLOAD:
+		for (brandinfo = &linux_brandlist[0]; *brandinfo != NULL;
+		     ++brandinfo)
+			if (elf64_brand_inuse(*brandinfo))
+				error = EBUSY;
+		if (error == 0) {
+			for (brandinfo = &linux_brandlist[0];
+			     *brandinfo != NULL; ++brandinfo)
+				if (elf64_remove_brand_entry(*brandinfo) < 0)
+					error = EINVAL;
+		}
+		if (error == 0) {
+			SET_FOREACH(lihp, linux_ioctl_handler_set)
+				linux_ioctl_unregister_handler(*lihp);
+			mtx_destroy(&futex_mtx);
+			if (bootverbose)
+				printf("Linux ELF exec handler removed\n");
+		} else
+			printf("Could not deinstall ELF interpreter entry\n");
+		break;
+	default:
+		return (EOPNOTSUPP);
+	}
+	return (error);
+}
+
+static moduledata_t linux64_elf_mod = {
+	"linux64elf",
+	linux64_elf_modevent,
+	0
+};
+
+DECLARE_MODULE_TIED(linux64elf, linux64_elf_mod, SI_SUB_EXEC, SI_ORDER_ANY);
+MODULE_DEPEND(linux64elf, linux_common, 1, 1, 1);

sys/amd64/linux/linux_vdso.lds.s

@@ -0,0 +1,69 @@
+/*
+ * Linker script for 64-bit vDSO.
+ * Copied from Linux kernel arch/x86/vdso/vdso-layout.lds.S
+ *
+ * $FreeBSD$
+ */
+
+SECTIONS
+{
+	. = . + SIZEOF_HEADERS;
+
+	.hash		: { *(.hash) }			:text
+	.gnu.hash	: { *(.gnu.hash) }
+	.dynsym		: { *(.dynsym) }
+	.dynstr		: { *(.dynstr) }
+	.gnu.version	: { *(.gnu.version) }
+	.gnu.version_d	: { *(.gnu.version_d) }
+	.gnu.version_r	: { *(.gnu.version_r) }
+
+	.note		: { *(.note.*) }		:text	:note
+
+	.eh_frame_hdr	: { *(.eh_frame_hdr) }		:text	:eh_frame_hdr
+	.eh_frame	: { KEEP (*(.eh_frame)) }	:text
+
+	.dynamic	: { *(.dynamic) }		:text	:dynamic
+
+	.rodata		: { *(.rodata*) }		:text
+	.data		: {
+	      *(.data*)
+	      *(.sdata*)
+	      *(.got.plt) *(.got)
+	      *(.gnu.linkonce.d.*)
+	      *(.bss*)
+	      *(.dynbss*)
+	      *(.gnu.linkonce.b.*)
+	}
+
+	.altinstructions	: { *(.altinstructions) }
+	.altinstr_replacement	: { *(.altinstr_replacement) }
+
+	. = ALIGN(0x100);
+	.text		: { *(.test .text*) }			:text	=0x90909090
+}
+
+PHDRS
+{
+	text		PT_LOAD		FLAGS(5) FILEHDR PHDRS; /* PF_R|PF_X */
+	dynamic		PT_DYNAMIC	FLAGS(4);		/* PF_R */
+	note		PT_NOTE		FLAGS(4);		/* PF_R */
+	eh_frame_hdr	PT_GNU_EH_FRAME;
+}
+
+VERSION
+{
+	LINUX_2.6 {
+	global:
+		time;
+		__vdso_time;
+		gettimeofday;
+		__vdso_gettimeofday;
+		getcpu;
+		__vdso_getcpu;
+		clock_gettime;
+		__vdso_clock_gettime;
+		linux_rt_sigcode;
+		linux_platform;
+	local: *;
+	};
+}

sys/amd64/linux/syscalls.conf

@@ -0,0 +1,11 @@
+# $FreeBSD$
+sysnames="linux_syscalls.c"
+sysproto="linux_proto.h"
+sysproto_h=_LINUX_SYSPROTO_H_
+syshdr="linux_syscall.h"
+syssw="linux_sysent.c"
+sysmk="/dev/null"
+syscallprefix="LINUX_SYS_"
+switchname="linux_sysent"
+namesname="linux_syscallnames"
+systrace="linux_systrace_args.c"

sys/amd64/linux/syscalls.master

@@ -0,0 +1,503 @@
+ $FreeBSD$
+
+;	@(#)syscalls.master	8.1 (Berkeley) 7/19/93
+; System call name/number master file (or rather, slave, from LINUX).
+; Processed to create linux_sysent.c, linux_proto.h and linux_syscall.h.
+
+; Columns: number audit type nargs name alt{name,tag,rtyp}/comments
+;	number	system call number, must be in order
+;	audit	the audit event associated with the system call
+;		A value of AUE_NULL means no auditing, but it also means that
+;		there is no audit event for the call at this time. For the
+;		case where the event exists, but we don't want auditing, the
+;		event should be #defined to AUE_NULL in audit_kevents.h.
+;	type	one of STD, OBSOL, UNIMPL
+;	name	psuedo-prototype of syscall routine
+;		If one of the following alts is different, then all appear:
+;	altname	name of system call if different
+;	alttag	name of args struct tag if different from [o]`name'"_args"
+;	altrtyp	return type if not int (bogus - syscalls always return int)
+;		for UNIMPL/OBSOL, name continues with comments
+
+; types:
+;	STD	always included
+;	OBSOL	obsolete, not included in system, only specifies name
+;	UNIMPL	not implemented, placeholder only
+
+#include <sys/param.h>
+#include <sys/sysent.h>
+#include <sys/sysproto.h>
+#include <compat/linux/linux_sysproto.h>
+#include <amd64/linux/linux.h>
+#include <amd64/linux/linux_proto.h>
+
+; Isn't pretty, but there seems to be no other way to trap nosys
+#define	nosys	linux_nosys
+
+; #ifdef's, etc. may be included, and are copied to the output files.
+
+0	AUE_NULL	NOPROTO	{ int read(int fd, char *buf, \
+				    u_int nbyte); }
+1	AUE_NULL	NOPROTO	{ int write(int fd, char *buf, \
+				    u_int nbyte); }
+2	AUE_OPEN_RWTC	STD	{ int linux_open(char *path, l_int flags, \
+				    l_int mode); }
+3	AUE_CLOSE	NOPROTO	{ int close(int fd); }
+4	AUE_STAT	STD	{ int linux_newstat(char *path, \
+				    struct l_newstat *buf); }
+5	AUE_FSTAT	STD	{ int linux_newfstat(l_uint fd, \
+				    struct l_newstat *buf); }
+6	AUE_LSTAT	STD	{ int linux_newlstat(char *path, \
+				    struct l_newstat *buf); }
+7	AUE_POLL	NOPROTO	{ int poll(struct pollfd*, \
+				    unsigned int nfds, int timeout); }
+8	AUE_LSEEK	STD	{ int linux_lseek(l_uint fdes, l_off_t off, \
+				    l_int whence); }
+9	AUE_MMAP	STD	{ int linux_mmap2(l_ulong addr, l_ulong len, \
+				    l_ulong prot, l_ulong flags, l_ulong fd, \
+				    l_ulong pgoff); }
+10	AUE_MPROTECT	STD	{ int linux_mprotect(caddr_t addr, int len, \
+				    int prot); }
+11	AUE_MUNMAP	NOPROTO	{ int munmap(caddr_t addr, int len); }
+12	AUE_NULL	STD	{ int linux_brk(l_ulong dsend); }
+13	AUE_NULL	STD	{ int linux_rt_sigaction(l_int sig, \
+				    l_sigaction_t *act, l_sigaction_t *oact, \
+				    l_size_t sigsetsize); }
+14	AUE_NULL	STD	{ int linux_rt_sigprocmask(l_int how, \
+				    l_sigset_t *mask, l_sigset_t *omask, \
+				    l_size_t sigsetsize); }
+15	AUE_NULL	STD	{ int linux_rt_sigreturn( \
+				    struct l_ucontext *ucp); }
+16	AUE_IOCTL	STD	{ int linux_ioctl(l_uint fd, l_uint cmd, \
+				    uintptr_t arg); }
+17	AUE_PREAD	STD	{ int linux_pread(l_uint fd, char *buf, \
+				    l_size_t nbyte, l_loff_t offset); }
+18	AUE_PWRITE	STD	{ int linux_pwrite(l_uint fd, char *buf, \
+				    l_size_t nbyte, l_loff_t offset); }
+19	AUE_READV	NOPROTO { int readv(int fd, struct iovec *iovp, \
+				    u_int iovcnt); }
+20	AUE_WRITEV	NOPROTO	{ int writev(int fd, struct iovec *iovp, \
+				    u_int iovcnt); }
+21	AUE_ACCESS	STD	{ int linux_access(char *path, l_int amode); }
+22	AUE_PIPE	STD	{ int linux_pipe(l_ulong *pipefds); }
+23	AUE_SELECT	STD	{ int linux_select(l_int nfds, \
+				    l_fd_set *readfds, l_fd_set *writefds, \
+				    l_fd_set *exceptfds, \
+				    struct l_timeval *timeout); }
+24	AUE_NULL	NOPROTO	{ int sched_yield(void); }
+25	AUE_NULL	STD	{ int linux_mremap(l_ulong addr, \
+				    l_ulong old_len, l_ulong new_len, \
+				    l_ulong flags, l_ulong new_addr); }
+26	AUE_MSYNC	STD	{ int linux_msync(l_ulong addr, \
+				    l_size_t len, l_int fl); }
+27	AUE_MINCORE	STD	{ int linux_mincore(l_ulong start, \
+				    l_size_t len, u_char *vec); }
+28	AUE_MADVISE	NOPROTO	{ int madvise(void *addr, size_t len, \
+				    int behav); }
+29	AUE_NULL	STD	{ int linux_shmget(l_key_t key, l_size_t size, \
+				    l_int shmflg); }
+30	AUE_NULL	STD	{ int linux_shmat(l_int shmid, char *shmaddr, \
+				    l_int shmflg); }
+31	AUE_NULL	STD	{ int linux_shmctl(l_int shmid, l_int cmd, \
+				    struct l_shmid_ds *buf); }
+32	AUE_DUP		NOPROTO	{ int dup(u_int fd); }
+33	AUE_DUP2	NOPROTO	{ int dup2(u_int from, u_int to); }
+34	AUE_NULL	STD	{ int linux_pause(void); }
+35	AUE_NULL	STD	{ int linux_nanosleep( \
+				    const struct l_timespec *rqtp, \
+				    struct l_timespec *rmtp); }
+36	AUE_GETITIMER	STD	{ int linux_getitimer(l_int which, \
+				    struct l_itimerval *itv); }
+37	AUE_NULL	STD	{ int linux_alarm(l_uint secs); }
+38	AUE_SETITIMER	STD	{ int linux_setitimer(l_int which, \
+				    struct l_itimerval *itv, \
+				    struct l_itimerval *oitv); }
+39	AUE_GETPID	STD	{ int linux_getpid(void); }
+40	AUE_SENDFILE	STD	{ int linux_sendfile(int out, int in, \
+				    l_long *offset, l_size_t count); }
+41	AUE_SOCKET	STD	{ int linux_socket(l_int domain, l_int type, \
+				    l_int protocol); }
+42	AUE_CONNECT	STD	{ int linux_connect(l_int s, l_uintptr_t name, \
+				    l_int namelen); }
+43	AUE_ACCEPT	STD	{ int linux_accept(l_int s, l_uintptr_t addr, \
+				    l_uintptr_t namelen); }
+44	AUE_SENDTO	STD	{ int linux_sendto(l_int s, l_uintptr_t msg, \
+				    l_int len, l_int flags, l_uintptr_t to, \
+				    l_int tolen); }
+45	AUE_RECVFROM	STD	{ int linux_recvfrom(l_int s, l_uintptr_t buf, \
+				    l_size_t len, l_int flags, l_uintptr_t from, \
+				    l_uintptr_t fromlen); }
+46	AUE_SENDMSG	STD	{ int linux_sendmsg(l_int s, l_uintptr_t msg, \
+				    l_int flags); }
+47	AUE_RECVMSG	STD	{ int linux_recvmsg(l_int s, l_uintptr_t msg, \
+				    l_int flags); }
+48	AUE_NULL	STD	{ int linux_shutdown(l_int s, l_int how); }
+49	AUE_BIND	STD	{ int linux_bind(l_int s, l_uintptr_t name, \
+				    l_int namelen); }
+50	AUE_LISTEN	STD	{ int linux_listen(l_int s, l_int backlog); }
+51	AUE_GETSOCKNAME	STD	{ int linux_getsockname(l_int s, \
+				    l_uintptr_t addr, l_uintptr_t namelen); }
+52	AUE_GETPEERNAME	STD	{ int linux_getpeername(l_int s, \
+				    l_uintptr_t addr, l_uintptr_t namelen); }
+53	AUE_SOCKETPAIR	STD	{ int linux_socketpair(l_int domain, \
+				    l_int type, l_int protocol, l_uintptr_t rsv); }
+54	AUE_SETSOCKOPT	STD	{ int linux_setsockopt(l_int s, l_int level, \
+				    l_int optname, l_uintptr_t optval, \
+				    l_int optlen); }
+55	AUE_GETSOCKOPT	STD	{ int linux_getsockopt(l_int s, l_int level, \
+				    l_int optname, l_uintptr_t optval, \
+				    l_uintptr_t optlen); }
+56	AUE_RFORK	STD	{ int linux_clone(l_int flags, void *stack, \
+				    void *parent_tidptr, void * child_tidptr, void *tls ); }
+57	AUE_FORK	STD	{ int linux_fork(void); }
+58	AUE_VFORK	STD	{ int linux_vfork(void); }
+59	AUE_EXECVE	STD	{ int linux_execve(char *path, char **argp, \
+				    char **envp); }
+60	AUE_EXIT	STD	{ void linux_exit(int rval); }
+61	AUE_WAIT4	STD	{ int linux_wait4(l_pid_t pid, \
+				    l_int *status, l_int options, \
+				    struct rusage *rusage); }
+62	AUE_KILL	STD	{ int linux_kill(l_int pid, l_int signum); }
+63	AUE_NULL	STD	{ int linux_newuname( \
+				    struct l_new_utsname *buf); }
+64	AUE_NULL	STD	{ int linux_semget(l_key_t key, \
+				    l_int nsems, l_int semflg); }
+65	AUE_NULL	STD	{ int linux_semop(l_int semid, \
+				    struct l_sembuf *tsops, l_uint nsops); }
+66	AUE_NULL	STD	{ int linux_semctl(l_int semid, \
+				    l_int semnum, l_int cmd, union l_semun arg); }
+67	AUE_NULL	STD	{ int linux_shmdt(char *shmaddr); }
+68	AUE_NULL	STD	{ int linux_msgget(l_key_t key, l_int msgflg); }
+69	AUE_NULL	STD	{ int linux_msgsnd(l_int msqid, \
+				    struct l_msgbuf *msgp, l_size_t msgsz, \
+				    l_int msgflg); }
+70	AUE_NULL	STD	{ int linux_msgrcv(l_int msqid, \
+				    struct l_msgbuf *msgp, l_size_t msgsz, \
+				    l_long msgtyp, l_int msgflg); }
+71	AUE_NULL	STD	{ int linux_msgctl(l_int msqid, l_int cmd, \
+				    struct l_msqid_ds *buf); }
+72	AUE_FCNTL	STD	{ int linux_fcntl(l_uint fd, l_uint cmd, \
+				    l_ulong arg); }
+73	AUE_FLOCK	NOPROTO	{ int flock(int fd, int how); }
+74	AUE_FSYNC	NOPROTO	{ int fsync(int fd); }
+75	AUE_NULL	STD	{ int linux_fdatasync(l_uint fd); }
+76	AUE_TRUNCATE	STD	{ int linux_truncate(char *path, \
+				    l_ulong length); }
+77	AUE_FTRUNCATE	STD	{ int linux_ftruncate(l_int fd, l_long length); }
+78	AUE_GETDIRENTRIES	STD { int linux_getdents(l_uint fd, void *dent, \
+				    l_uint count); }
+79	AUE_GETCWD	STD	{ int linux_getcwd(char *buf, \
+				    l_ulong bufsize); }
+80	AUE_CHDIR	STD	{ int linux_chdir(char *path); }
+81	AUE_FCHDIR	NOPROTO	{ int fchdir(int fd); }
+82	AUE_RENAME	STD	{ int linux_rename(char *from, char *to); }
+83	AUE_MKDIR	STD	{ int linux_mkdir(char *path, l_int mode); }
+84	AUE_RMDIR	STD	{ int linux_rmdir(char *path); }
+85	AUE_CREAT	STD	{ int linux_creat(char *path, \
+				    l_int mode); }
+86	AUE_LINK	STD	{ int linux_link(char *path, char *to); }
+87	AUE_UNLINK	STD	{ int linux_unlink(char *path); }
+88	AUE_SYMLINK	STD	{ int linux_symlink(char *path, char *to); }
+89	AUE_READLINK	STD	{ int linux_readlink(char *name, char *buf, \
+				    l_int count); }
+90	AUE_CHMOD	STD	{ int linux_chmod(char *path, \
+				    l_mode_t mode); }
+91	AUE_FCHMOD	NOPROTO	{ int fchmod(int fd, int mode); }
+92	AUE_LCHOWN	STD	{ int linux_chown(char *path, \
+				    l_uid_t uid, l_gid_t gid); }
+93	AUE_FCHOWN	NOPROTO	{ int fchown(int fd, int uid, int gid); }
+94	AUE_LCHOWN	STD	{ int linux_lchown(char *path, l_uid_t uid, \
+				    l_gid_t gid); }
+95	AUE_UMASK	NOPROTO	{ int umask(int newmask); }
+96	AUE_NULL	NOPROTO	{ int gettimeofday(struct l_timeval *tp, \
+				    struct timezone *tzp); }
+97	AUE_GETRLIMIT	STD	{ int linux_getrlimit(l_uint resource, \
+				    struct l_rlimit *rlim); }
+98	AUE_GETRUSAGE	NOPROTO	{ int getrusage(int who, struct rusage *rusage); }
+99	AUE_NULL	STD	{ int linux_sysinfo(struct l_sysinfo *info); }
+100	AUE_NULL	STD	{ int linux_times(struct l_times_argv *buf); }
+101	AUE_PTRACE	STD	{ int linux_ptrace(l_long req, l_long pid, \
+				    l_long addr, l_long data); }
+102	AUE_GETUID	STD	{ int linux_getuid(void); }
+103	AUE_NULL	STD	{ int linux_syslog(l_int type, char *buf, \
+				    l_int len); }
+104	AUE_GETGID	STD	{ int linux_getgid(void); }
+105	AUE_SETUID	NOPROTO	{ int setuid(uid_t uid); }
+106	AUE_SETGID	NOPROTO	{ int setgid(gid_t gid); }
+107	AUE_GETEUID	NOPROTO	{ int geteuid(void); }
+108	AUE_GETEGID	NOPROTO	{ int getegid(void); }
+109	AUE_SETPGRP	NOPROTO	{ int setpgid(int pid, int pgid); }
+110	AUE_GETPPID	STD	{ int linux_getppid(void); }
+111	AUE_GETPGRP	NOPROTO	{ int getpgrp(void); }
+112	AUE_SETSID	NOPROTO	{ int setsid(void); }
+113	AUE_SETREUID	NOPROTO	{ int setreuid(uid_t ruid, uid_t euid); }
+114	AUE_SETREGID	NOPROTO	{ int setregid(gid_t rgid, gid_t egid); }
+115	AUE_GETGROUPS	STD	{ int linux_getgroups(l_int gidsetsize, \
+				    l_gid_t *grouplist); }
+116	AUE_SETGROUPS	STD	{ int linux_setgroups(l_int gidsetsize, \
+				    l_gid_t *grouplist); }
+117	AUE_SETRESUID	NOPROTO	{ int setresuid(uid_t ruid, uid_t euid, \
+				    uid_t suid); }
+118	AUE_GETRESUID	NOPROTO	{ int getresuid(uid_t *ruid, uid_t *euid, \
+				    uid_t *suid); }
+119	AUE_SETRESGID	NOPROTO	{ int setresgid(gid_t rgid, gid_t egid, \
+				    gid_t sgid); }
+120	AUE_GETRESGID	NOPROTO	{ int getresgid(gid_t *rgid, gid_t *egid, \
+				    gid_t *sgid); }
+121	AUE_GETPGID	NOPROTO	{ int getpgid(int pid); }
+122	AUE_SETFSUID	STD	{ int linux_setfsuid(l_uid_t uid); }
+123	AUE_SETFSGID	STD	{ int linux_setfsgid(l_gid_t gid); }
+124	AUE_GETSID	STD	{ int linux_getsid(l_pid_t pid); }
+125	AUE_CAPGET	STD	{ int linux_capget(struct l_user_cap_header *hdrp, \
+				    struct l_user_cap_data *datap); }
+126	AUE_CAPSET	STD	{ int linux_capset(struct l_user_cap_header *hdrp, \
+				    struct l_user_cap_data *datap); }
+127	AUE_NULL	STD	{ int linux_rt_sigpending(l_sigset_t *set, \
+				    l_size_t sigsetsize); }
+128	AUE_NULL	STD	{ int linux_rt_sigtimedwait(l_sigset_t *mask, \
+				    l_siginfo_t *ptr, \
+				    struct l_timeval *timeout, \
+				    l_size_t sigsetsize); }
+129	AUE_NULL	STD	{ int linux_rt_sigqueueinfo(l_pid_t pid, l_int sig, \
+				    l_siginfo_t *info); }
+130	AUE_NULL	STD	{ int linux_rt_sigsuspend( \
+				    l_sigset_t *newset, \
+				    l_size_t sigsetsize); }
+131	AUE_NULL	STD	{ int linux_sigaltstack(l_stack_t *uss, \
+				    l_stack_t *uoss); }
+132	AUE_UTIME	STD	{ int linux_utime(char *fname, \
+				    struct l_utimbuf *times); }
+133	AUE_MKNOD	STD	{ int linux_mknod(char *path, l_int mode, \
+				    l_dev_t dev); }
+134	AUE_USELIB	UNIMPL	uselib
+135	AUE_PERSONALITY	STD	{ int linux_personality(l_ulong per); }
+136	AUE_NULL	STD	{ int linux_ustat(l_dev_t dev, \
+				    struct l_ustat *ubuf); }
+137	AUE_STATFS	STD	{ int linux_statfs(char *path, \
+				    struct l_statfs_buf *buf); }
+138	AUE_FSTATFS	STD	{ int linux_fstatfs(l_uint fd, \
+				    struct l_statfs_buf *buf); }
+139	AUE_NULL	STD	{ int linux_sysfs(l_int option, \
+				    l_ulong arg1, l_ulong arg2); }
+140	AUE_GETPRIORITY	STD	{ int linux_getpriority(int which, int who); }
+141	AUE_SETPRIORITY	NOPROTO	{ int setpriority(int which, int who, \
+				    int prio); }
+142	AUE_SCHED_SETPARAM	STD	{ int linux_sched_setparam(l_pid_t pid, \
+				    struct l_sched_param *param); }
+143	AUE_SCHED_GETPARAM	STD	{ int linux_sched_getparam(l_pid_t pid, \
+				    struct l_sched_param *param); }
+144	AUE_SCHED_SETSCHEDULER	STD { int linux_sched_setscheduler( \
+				    l_pid_t pid, l_int policy, \
+				    struct l_sched_param *param); }
+145	AUE_SCHED_GETSCHEDULER	STD { int linux_sched_getscheduler( \
+				    l_pid_t pid); }
+146	AUE_SCHED_GET_PRIORITY_MAX	STD { int linux_sched_get_priority_max( \
+				    l_int policy); }
+147	AUE_SCHED_GET_PRIORITY_MIN	STD { int linux_sched_get_priority_min( \
+				    l_int policy); }
+148	AUE_SCHED_RR_GET_INTERVAL	STD { int linux_sched_rr_get_interval(l_pid_t pid, \
+				    struct l_timespec *interval); }
+149	AUE_MLOCK	NOPROTO	{ int mlock(const void *addr, size_t len); }
+150	AUE_MUNLOCK	NOPROTO	{ int munlock(const void *addr, size_t len); }
+151	AUE_MLOCKALL	NOPROTO	{ int mlockall(int how); }
+152	AUE_MUNLOCKALL	NOPROTO	{ int munlockall(void); }
+153	AUE_NULL	STD	{ int linux_vhangup(void); }
+154	AUE_NULL	UNIMPL	modify_ldt
+155	AUE_PIVOT_ROOT	STD	{ int linux_pivot_root(void); }
+156	AUE_SYSCTL	STD	{ int linux_sysctl( \
+				    struct l___sysctl_args *args); }
+157	AUE_PRCTL	STD	{ int linux_prctl(l_int option, l_uintptr_t arg2, \
+				    l_uintptr_t arg3, l_uintptr_t arg4, \
+				    l_uintptr_t arg5); }
+158	AUE_PRCTL	STD	{ int linux_arch_prctl(l_int code, l_ulong addr); }
+159	AUE_ADJTIME	STD	{ int linux_adjtimex(void); }
+160	AUE_SETRLIMIT	STD	{ int linux_setrlimit(l_uint resource, \
+				    struct l_rlimit *rlim); }
+161	AUE_CHROOT	NOPROTO	{ int chroot(char *path); }
+162	AUE_SYNC	NOPROTO	{ int sync(void); }
+163	AUE_ACCT	NOPROTO	{ int acct(char *path); }
+164	AUE_SETTIMEOFDAY	NOPROTO	{ int settimeofday(struct l_timeval *tp, struct timezone *tzp); }
+165	AUE_MOUNT	STD	{ int linux_mount(char *specialfile, \
+				    char *dir, char *filesystemtype, \
+				    l_ulong rwflag, void *data); }
+166	AUE_UMOUNT	STD	{ int linux_umount(char *path, l_int flags); }
+167	AUE_SWAPON	NOPROTO	{ int swapon(char *name); }
+168	AUE_SWAPOFF	STD	{ int linux_swapoff(void); }
+169	AUE_REBOOT	STD	{ int linux_reboot(l_int magic1, \
+				    l_int magic2, l_uint cmd, void *arg); }
+170	AUE_SYSCTL	STD	{ int linux_sethostname(char *hostname, \
+				    l_uint len); }
+171	AUE_SYSCTL	STD	{ int linux_setdomainname(char *name, \
+				    l_int len); }
+172	AUE_NULL	STD	{ int linux_iopl(l_uint level); }
+173	AUE_NULL	UNIMPL	ioperm
+174	AUE_NULL	STD	{ int linux_create_module(void); }
+175	AUE_NULL	STD	{ int linux_init_module(void); }
+176	AUE_NULL	STD	{ int linux_delete_module(void); }
+177	AUE_NULL	STD	{ int linux_get_kernel_syms(void); }
+178	AUE_NULL	STD	{ int linux_query_module(void); }
+179	AUE_QUOTACTL	STD	{ int linux_quotactl(void); }
+180	AUE_NULL	STD	{ int linux_nfsservctl(void); }
+181	AUE_GETPMSG	STD	{ int linux_getpmsg(void); }
+182	AUE_PUTPMSG	STD	{ int linux_putpmsg(void); }
+183	AUE_NULL	STD	{ int linux_afs_syscall(void); }
+184	AUE_NULL	STD	{ int linux_tuxcall(void); }
+185	AUE_NULL	STD	{ int linux_security(void); }
+186	AUE_NULL	STD	{ int linux_gettid(void); }
+187	AUE_NULL	UNIMPL	linux_readahead
+188	AUE_NULL	STD	{ int linux_setxattr(void); }
+189	AUE_NULL	STD	{ int linux_lsetxattr(void); }
+190	AUE_NULL	STD	{ int linux_fsetxattr(void); }
+191	AUE_NULL	STD	{ int linux_getxattr(void); }
+192	AUE_NULL	STD	{ int linux_lgetxattr(void); }
+193	AUE_NULL	STD	{ int linux_fgetxattr(void); }
+194	AUE_NULL	STD	{ int linux_listxattr(void); }
+195	AUE_NULL	STD	{ int linux_llistxattr(void); }
+196	AUE_NULL	STD	{ int linux_flistxattr(void); }
+197	AUE_NULL	STD	{ int linux_removexattr(void); }
+198	AUE_NULL	STD	{ int linux_lremovexattr(void); }
+199	AUE_NULL	STD	{ int linux_fremovexattr(void); }
+200	AUE_NULL	STD	{ int linux_tkill(int tid, int sig); }
+201	AUE_NULL	STD	{ int linux_time(l_time_t *tm); }
+202	AUE_NULL	STD	{ int linux_sys_futex(void *uaddr, int op, int val, \
+				    struct l_timespec *timeout, void *uaddr2, int val3); }
+203	AUE_NULL	STD	{ int linux_sched_setaffinity(l_pid_t pid, l_uint len, \
+				    l_ulong *user_mask_ptr); }
+204	AUE_NULL	STD	{ int linux_sched_getaffinity(l_pid_t pid, l_uint len, \
+				    l_ulong *user_mask_ptr); }
+205	AUE_NULL	STD	{ int linux_set_thread_area(void); }
+206	AUE_NULL	UNIMPL	linux_io_setup
+207	AUE_NULL	UNIMPL	linux_io_destroy
+208	AUE_NULL	UNIMPL	linux_io_getevents
+209	AUE_NULL	UNIMPL	inux_io_submit
+210	AUE_NULL	UNIMPL	linux_io_cancel
+211	AUE_NULL	UNIMPL	linux_get_thread_area
+212	AUE_NULL	STD	{ int linux_lookup_dcookie(void); }
+213	AUE_NULL	STD	{ int linux_epoll_create(void); }
+214	AUE_NULL	STD	{ int linux_epoll_ctl_old(void); }
+215	AUE_NULL	STD	{ int linux_epoll_wait_old(void); }
+216	AUE_NULL	STD	{ int linux_remap_file_pages(void); }
+217	AUE_GETDIRENTRIES	STD { int linux_getdents64(l_uint fd, \
+				    void *dirent, l_uint count); }
+218	AUE_NULL	STD	{ int linux_set_tid_address(int *tidptr); }
+219	AUE_NULL	UNIMPL	restart_syscall
+220	AUE_NULL	STD	{ int linux_semtimedop(void); }
+221	AUE_NULL	STD	{ int linux_fadvise64(int fd, l_loff_t offset, \
+				    l_size_t len, int advice); }
+222	AUE_NULL	STD	{ int linux_timer_create(clockid_t clock_id, \
+				    struct sigevent *evp, l_timer_t *timerid); }
+223	AUE_NULL	STD	{ int linux_timer_settime(l_timer_t timerid, l_int flags, \
+				    const struct itimerspec *new, struct itimerspec *old); }
+224	AUE_NULL	STD	{ int linux_timer_gettime(l_timer_t timerid, struct itimerspec *setting); }
+225	AUE_NULL	STD	{ int linux_timer_getoverrun(l_timer_t timerid); }
+226	AUE_NULL	STD	{ int linux_timer_delete(l_timer_t timerid); }
+227	AUE_CLOCK_SETTIME	STD	{ int linux_clock_settime(clockid_t which, struct l_timespec *tp); }
+228	AUE_NULL	STD	{ int linux_clock_gettime(clockid_t which, struct l_timespec *tp); }
+229	AUE_NULL	STD	{ int linux_clock_getres(clockid_t which, struct l_timespec *tp); }
+230	AUE_NULL	STD	{ int linux_clock_nanosleep(clockid_t which, int flags, \
+				    struct l_timespec *rqtp, struct l_timespec *rmtp); }
+231	AUE_EXIT	STD	{ int linux_exit_group(int error_code); }
+232	AUE_NULL	STD	{ int linux_epoll_wait(void); }
+233	AUE_NULL	STD	{ int linux_epoll_ctl(void); }
+234	AUE_NULL	STD	{ int linux_tgkill(int tgid, int pid, int sig); }
+235	AUE_UTIMES	STD	{ int linux_utimes(char *fname, \
+				    struct l_timeval *tptr); }
+236	AUE_NULL	UNIMPL	vserver
+237	AUE_NULL	STD	{ int linux_mbind(void); }
+238	AUE_NULL	STD	{ int linux_set_mempolicy(void); }
+239	AUE_NULL	STD	{ int linux_get_mempolicy(void); }
+240	AUE_NULL	STD	{ int linux_mq_open(void); }
+241	AUE_NULL	STD	{ int linux_mq_unlink(void); }
+242	AUE_NULL	STD	{ int linux_mq_timedsend(void); }
+243	AUE_NULL	STD	{ int linux_mq_timedreceive(void); }
+244	AUE_NULL	STD	{ int linux_mq_notify(void); }
+245	AUE_NULL	STD	{ int linux_mq_getsetattr(void); }
+246	AUE_NULL	STD	{ int linux_kexec_load(void); }
+247	AUE_WAIT6	STD	{ int linux_waitid(int idtype, l_pid_t id, \
+				    l_siginfo_t *info, int options, \
+				    struct rusage *rusage); }
+248	AUE_NULL	STD	{ int linux_add_key(void); }
+249	AUE_NULL	STD	{ int linux_request_key(void); }
+250	AUE_NULL	STD	{ int linux_keyctl(void); }
+251	AUE_NULL	STD	{ int linux_ioprio_set(void); }
+252	AUE_NULL	STD	{ int linux_ioprio_get(void); }
+253	AUE_NULL	STD	{ int linux_inotify_init(void); }
+254	AUE_NULL	STD	{ int linux_inotify_add_watch(void); }
+255	AUE_NULL	STD	{ int linux_inotify_rm_watch(void); }
+256	AUE_NULL	STD	{ int linux_migrate_pages(void); }
+257	AUE_OPEN_RWTC	STD	{ int linux_openat(l_int dfd, const char *filename, \
+				    l_int flags, l_int mode); }
+258	AUE_MKDIRAT	STD	{ int linux_mkdirat(l_int dfd, const char *pathname, \
+				    l_int mode); }
+259	AUE_MKNODAT	STD	{ int linux_mknodat(l_int dfd, const char *filename, \
+				    l_int mode, l_uint dev); }
+260	AUE_FCHOWNAT	STD	{ int linux_fchownat(l_int dfd, const char *filename, \
+				    l_uid_t uid, l_gid_t gid, l_int flag); }
+261	AUE_FUTIMESAT	STD	{ int linux_futimesat(l_int dfd, char *filename, \
+				    struct l_timeval *utimes); }
+262	AUE_FSTATAT	STD	{ int linux_newfstatat(l_int dfd, char *pathname, \
+				    struct l_stat64 *statbuf, l_int flag); }
+263	AUE_UNLINKAT	STD	{ int linux_unlinkat(l_int dfd, const char *pathname, \
+				    l_int flag); }
+264	AUE_RENAMEAT	STD	{ int linux_renameat(l_int olddfd, const char *oldname, \
+				    l_int newdfd, const char *newname); }
+265	AUE_LINKAT	STD	{ int linux_linkat(l_int olddfd, const char *oldname, \
+				    l_int newdfd, const char *newname, l_int flag); }
+266	AUE_SYMLINKAT	STD	{ int linux_symlinkat(const char *oldname, l_int newdfd, \
+				    const char *newname); }
+267	AUE_READLINKAT	STD	{ int linux_readlinkat(l_int dfd, const char *path, \
+				    char *buf, l_int bufsiz); }
+268	AUE_FCHMODAT	STD	{ int linux_fchmodat(l_int dfd, const char *filename, \
+				    l_mode_t mode); }
+269	AUE_FACCESSAT	STD	{ int linux_faccessat(l_int dfd, const char *filename, \
+				    l_int amode, l_int flag); }
+270	AUE_SELECT	STD	{ int linux_pselect6(l_int nfds,			\
+				    l_fd_set *readfds, l_fd_set *writefds, l_fd_set *exceptfds,	\
+				    struct l_timespec *tsp, l_uintptr_t *sig); }
+271	AUE_NULL	STD	{ int linux_ppoll(void); }
+272	AUE_NULL	STD	{ int linux_unshare(void); }
+273	AUE_NULL	STD	{ int linux_set_robust_list(struct linux_robust_list_head *head, \
+				    l_size_t len); }
+274	AUE_NULL	STD	{ int linux_get_robust_list(l_int pid, struct linux_robust_list_head *head, \
+				    l_size_t *len); }
+275	AUE_NULL	STD	{ int linux_splice(void); }
+276	AUE_NULL	STD	{ int linux_tee(void); }
+277	AUE_NULL	STD	{ int linux_sync_file_range(void); }
+278	AUE_NULL	STD	{ int linux_vmsplice(void); }
+279	AUE_NULL	STD	{ int linux_move_pages(void); }
+280	AUE_NULL	STD	{ int linux_utimensat(void); }
+281	AUE_NULL	STD	{ int linux_epoll_pwait(void); }
+282	AUE_NULL	STD	{ int linux_signalfd(void); }
+283	AUE_NULL	STD	{ int linux_timerfd(void); }
+284	AUE_NULL	STD	{ int linux_eventfd(void); }
+285	AUE_NULL	STD	{ int linux_fallocate(void); }
+286	AUE_NULL	STD	{ int linux_timerfd_settime(void); }
+287	AUE_NULL	STD	{ int linux_timerfd_gettime(void); }
+288	AUE_ACCEPT	STD	{ int linux_accept4(l_int s, l_uintptr_t addr, \
+				    l_uintptr_t namelen, int flags); }
+289	AUE_NULL	STD	{ int linux_signalfd4(void); }
+290	AUE_NULL	STD	{ int linux_eventfd2(void); }
+291	AUE_NULL	STD	{ int linux_epoll_create1(void); }
+292	AUE_NULL	STD	{ int linux_dup3(l_int oldfd,		\
+				    l_int newfd, l_int flags); }
+293	AUE_NULL	STD	{ int linux_pipe2(l_int *pipefds, l_int flags); }
+294	AUE_NULL	STD	{ int linux_inotify_init1(void); }
+295	AUE_NULL	STD	{ int linux_preadv(void); }
+296	AUE_NULL	STD	{ int linux_pwritev(void); }
+297	AUE_NULL	STD	{ int linux_rt_tsigqueueinfo(void); }
+298	AUE_NULL	STD	{ int linux_perf_event_open(void); }
+299	AUE_NULL	STD	{ int linux_recvmmsg(void); }
+300	AUE_NULL	STD	{ int linux_fanotify_init(void); }
+301	AUE_NULL	STD	{ int linux_fanotify_mark(void); }
+302	AUE_NULL	STD	{ int linux_prlimit64(l_pid_t pid, l_uint resource, \
+				    struct rlimit *new, struct rlimit *old); }
+303	AUE_NULL	STD	{ int linux_name_to_handle_at(void); }
+304	AUE_NULL	STD	{ int linux_open_by_handle_at(void); }
+305	AUE_NULL	STD	{ int linux_clock_adjtime(void); }
+306	AUE_NULL	STD	{ int linux_syncfs(void); }
+307	AUE_NULL	STD	{ int linux_sendmmsg(void); }
+308	AUE_NULL	STD	{ int linux_setns(void); }
+309	AUE_NULL	STD	{ int linux_process_vm_readv(void); }
+310	AUE_NULL	STD	{ int linux_process_vm_writev(void); }
+311	AUE_NULL	STD	{ int linux_kcmp(void); }
+312	AUE_NULL	STD	{ int linux_finit_module(void); }

sys/modules/linux64/Makefile

@@ -0,0 +1,55 @@
+# $FreeBSD$
+
+.PATH: ${.CURDIR}/../../compat/linux ${.CURDIR}/../../${MACHINE_ARCH}/linux
+
+VDSO=	linux_vdso
+
+KMOD=	linux64
+SRCS=	linux_fork.c linux_dummy.c linux_file.c \
+	linux_futex.c linux_getcwd.c linux_ioctl.c linux_ipc.c \
+	linux_machdep.c linux_misc.c linux_signal.c \
+	linux_socket.c linux_stats.c linux_sysctl.c linux_sysent.c \
+	linux_sysvec.c linux_time.c linux_vdso.c linux_timer.c \
+	opt_inet6.h opt_compat.h opt_kdtrace.h opt_posix.h opt_usb.h \
+	vnode_if.h device_if.h bus_if.h assym.s \
+	linux_support.s
+DPSRCS=	linux_genassym.c
+
+# XXX: for assym.s
+SRCS+=  opt_kstack_pages.h opt_nfs.h opt_apic.h opt_hwpmc_hooks.h
+
+CLEANFILES=	linux_assym.h linux_genassym.o linux_locore.o
+
+OBJS=	${VDSO}.so
+
+linux_assym.h: linux_genassym.o
+	sh ${SYSDIR}/kern/genassym.sh linux_genassym.o > ${.TARGET}
+
+linux_locore.o: linux_locore.s linux_assym.h
+	${CC} -x assembler-with-cpp -DLOCORE -shared -mcmodel=small	\
+	-pipe -I. -I${SYSDIR} -Werror -Wall -fno-common -nostdinc 	\
+	-Wl,-T${.CURDIR}/../../${MACHINE_CPUARCH}/linux/${VDSO}.lds.s	\
+	-Wl,-soname=${VDSO}.so.1,-fPIC,-warn-common -nostdlib		\
+	    ${.IMPSRC} -o ${.TARGET}
+
+${VDSO}.so: linux_locore.o
+	${OBJCOPY} --input binary --output elf64-x86-64-freebsd		\
+	    -S -g --binary-architecture i386:x86-64 linux_locore.o ${.TARGET}
+
+linux_support.o: assym.s linux_assym.h
+	${CC} -c -x assembler-with-cpp -DLOCORE ${CFLAGS} \
+	    ${.IMPSRC} -o ${.TARGET}
+
+linux_genassym.o:
+	${CC} -c ${CFLAGS:N-fno-common} ${.IMPSRC}
+
+.if !defined(KERNBUILDDIR)
+.if defined(DEBUG)
+CFLAGS+=-DDEBUG
+.endif
+.if defined(KTR)
+CFLAGS+=-DKTR
+.endif
+.endif
+
+.include <bsd.kmod.mk>