xref: /onnv-gate/usr/src/uts/common/sys/zone.h (revision 12725:334fd88ae67c)
10Sstevel@tonic-gate /*
20Sstevel@tonic-gate  * CDDL HEADER START
30Sstevel@tonic-gate  *
40Sstevel@tonic-gate  * The contents of this file are subject to the terms of the
51676Sjpk  * Common Development and Distribution License (the "License").
61676Sjpk  * You may not use this file except in compliance with the License.
70Sstevel@tonic-gate  *
80Sstevel@tonic-gate  * You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
90Sstevel@tonic-gate  * or http://www.opensolaris.org/os/licensing.
100Sstevel@tonic-gate  * See the License for the specific language governing permissions
110Sstevel@tonic-gate  * and limitations under the License.
120Sstevel@tonic-gate  *
130Sstevel@tonic-gate  * When distributing Covered Code, include this CDDL HEADER in each
140Sstevel@tonic-gate  * file and include the License file at usr/src/OPENSOLARIS.LICENSE.
150Sstevel@tonic-gate  * If applicable, add the following below this CDDL HEADER, with the
160Sstevel@tonic-gate  * fields enclosed by brackets "[]" replaced with your own identifying
170Sstevel@tonic-gate  * information: Portions Copyright [yyyy] [name of copyright owner]
180Sstevel@tonic-gate  *
190Sstevel@tonic-gate  * CDDL HEADER END
200Sstevel@tonic-gate  */
210Sstevel@tonic-gate /*
2212273SCasper.Dik@Sun.COM  * Copyright (c) 2003, 2010, Oracle and/or its affiliates. All rights reserved.
230Sstevel@tonic-gate  */
240Sstevel@tonic-gate 
250Sstevel@tonic-gate #ifndef _SYS_ZONE_H
260Sstevel@tonic-gate #define	_SYS_ZONE_H
270Sstevel@tonic-gate 
280Sstevel@tonic-gate #include <sys/types.h>
290Sstevel@tonic-gate #include <sys/mutex.h>
300Sstevel@tonic-gate #include <sys/param.h>
310Sstevel@tonic-gate #include <sys/rctl.h>
322677Sml93401 #include <sys/ipc_rctl.h>
330Sstevel@tonic-gate #include <sys/pset.h>
341676Sjpk #include <sys/tsol/label.h>
353448Sdh155122 #include <sys/cred.h>
363448Sdh155122 #include <sys/netstack.h>
372267Sdp #include <sys/uadmin.h>
3810910SRobert.Harris@Sun.COM #include <sys/ksynch.h>
390Sstevel@tonic-gate 
400Sstevel@tonic-gate #ifdef	__cplusplus
410Sstevel@tonic-gate extern "C" {
420Sstevel@tonic-gate #endif
430Sstevel@tonic-gate 
440Sstevel@tonic-gate /*
450Sstevel@tonic-gate  * NOTE
460Sstevel@tonic-gate  *
470Sstevel@tonic-gate  * The contents of this file are private to the implementation of
480Sstevel@tonic-gate  * Solaris and are subject to change at any time without notice.
490Sstevel@tonic-gate  * Applications and drivers using these interfaces may fail to
500Sstevel@tonic-gate  * run on future releases.
510Sstevel@tonic-gate  */
520Sstevel@tonic-gate 
530Sstevel@tonic-gate /* Available both in kernel and for user space */
540Sstevel@tonic-gate 
550Sstevel@tonic-gate /* zone id restrictions and special ids */
560Sstevel@tonic-gate #define	MAX_ZONEID	9999
570Sstevel@tonic-gate #define	MIN_USERZONEID	1	/* lowest user-creatable zone ID */
580Sstevel@tonic-gate #define	MIN_ZONEID	0	/* minimum zone ID on system */
590Sstevel@tonic-gate #define	GLOBAL_ZONEID	0
600Sstevel@tonic-gate #define	ZONEID_WIDTH	4	/* for printf */
610Sstevel@tonic-gate 
621676Sjpk /*
631676Sjpk  * Special zoneid_t token to refer to all zones.
641676Sjpk  */
651676Sjpk #define	ALL_ZONES	(-1)
661676Sjpk 
670Sstevel@tonic-gate /* system call subcodes */
683448Sdh155122 #define	ZONE_CREATE		0
693448Sdh155122 #define	ZONE_DESTROY		1
703448Sdh155122 #define	ZONE_GETATTR		2
713448Sdh155122 #define	ZONE_ENTER		3
723448Sdh155122 #define	ZONE_LIST		4
733448Sdh155122 #define	ZONE_SHUTDOWN		5
743448Sdh155122 #define	ZONE_LOOKUP		6
753448Sdh155122 #define	ZONE_BOOT		7
763448Sdh155122 #define	ZONE_VERSION		8
773448Sdh155122 #define	ZONE_SETATTR		9
783448Sdh155122 #define	ZONE_ADD_DATALINK	10
793448Sdh155122 #define	ZONE_DEL_DATALINK	11
803448Sdh155122 #define	ZONE_CHECK_DATALINK	12
813448Sdh155122 #define	ZONE_LIST_DATALINK	13
820Sstevel@tonic-gate 
830Sstevel@tonic-gate /* zone attributes */
840Sstevel@tonic-gate #define	ZONE_ATTR_ROOT		1
850Sstevel@tonic-gate #define	ZONE_ATTR_NAME		2
860Sstevel@tonic-gate #define	ZONE_ATTR_STATUS	3
870Sstevel@tonic-gate #define	ZONE_ATTR_PRIVSET	4
880Sstevel@tonic-gate #define	ZONE_ATTR_UNIQID	5
890Sstevel@tonic-gate #define	ZONE_ATTR_POOLID	6
900Sstevel@tonic-gate #define	ZONE_ATTR_INITPID	7
911676Sjpk #define	ZONE_ATTR_SLBL		8
922267Sdp #define	ZONE_ATTR_INITNAME	9
932267Sdp #define	ZONE_ATTR_BOOTARGS	10
942712Snn35248 #define	ZONE_ATTR_BRAND		11
953247Sgjelinek #define	ZONE_ATTR_PHYS_MCAP	12
963247Sgjelinek #define	ZONE_ATTR_SCHED_CLASS	13
973448Sdh155122 #define	ZONE_ATTR_FLAGS		14
988662SJordan.Vaughan@Sun.com #define	ZONE_ATTR_HOSTID	15
9912633Sjohn.levon@sun.com #define	ZONE_ATTR_FS_ALLOWED	16
1002712Snn35248 
1012712Snn35248 /* Start of the brand-specific attribute namespace */
1022712Snn35248 #define	ZONE_ATTR_BRAND_ATTRS	32768
1030Sstevel@tonic-gate 
10412633Sjohn.levon@sun.com #define	ZONE_FS_ALLOWED_MAX	1024
10512633Sjohn.levon@sun.com 
1061166Sdstaff #define	ZONE_EVENT_CHANNEL	"com.sun:zones:status"
1071166Sdstaff #define	ZONE_EVENT_STATUS_CLASS	"status"
1081166Sdstaff #define	ZONE_EVENT_STATUS_SUBCLASS	"change"
1091166Sdstaff 
1101166Sdstaff #define	ZONE_EVENT_UNINITIALIZED	"uninitialized"
1115880Snordmark #define	ZONE_EVENT_INITIALIZED		"initialized"
1121166Sdstaff #define	ZONE_EVENT_READY		"ready"
1131166Sdstaff #define	ZONE_EVENT_RUNNING		"running"
1141166Sdstaff #define	ZONE_EVENT_SHUTTING_DOWN	"shutting_down"
1151166Sdstaff 
1161166Sdstaff #define	ZONE_CB_NAME		"zonename"
1171166Sdstaff #define	ZONE_CB_NEWSTATE	"newstate"
1181166Sdstaff #define	ZONE_CB_OLDSTATE	"oldstate"
1191166Sdstaff #define	ZONE_CB_TIMESTAMP	"when"
1201166Sdstaff #define	ZONE_CB_ZONEID		"zoneid"
1211166Sdstaff 
1222712Snn35248 /*
1232712Snn35248  * Exit values that may be returned by scripts or programs invoked by various
1242712Snn35248  * zone commands.
1252712Snn35248  *
1262712Snn35248  * These are defined as:
1272712Snn35248  *
1282712Snn35248  *	ZONE_SUBPROC_OK
1292712Snn35248  *	===============
1302712Snn35248  *	The subprocess completed successfully.
1312712Snn35248  *
1322712Snn35248  *	ZONE_SUBPROC_USAGE
1332712Snn35248  *	==================
1342712Snn35248  *	The subprocess failed with a usage message, or a usage message should
1352712Snn35248  *	be output in its behalf.
1362712Snn35248  *
1372712Snn35248  *	ZONE_SUBPROC_NOTCOMPLETE
1382712Snn35248  *	========================
1392712Snn35248  *	The subprocess did not complete, but the actions performed by the
1402712Snn35248  *	subprocess require no recovery actions by the user.
1412712Snn35248  *
1422712Snn35248  *	For example, if the subprocess were called by "zoneadm install," the
1432712Snn35248  *	installation of the zone did not succeed but the user need not perform
1442712Snn35248  *	a "zoneadm uninstall" before attempting another install.
1452712Snn35248  *
1462712Snn35248  *	ZONE_SUBPROC_FATAL
1472712Snn35248  *	==================
1482712Snn35248  *	The subprocess failed in a fatal manner, usually one that will require
1492712Snn35248  *	some type of recovery action by the user.
1502712Snn35248  *
1512712Snn35248  *	For example, if the subprocess were called by "zoneadm install," the
1522712Snn35248  *	installation of the zone did not succeed and the user will need to
1532712Snn35248  *	perform a "zoneadm uninstall" before another install attempt is
1542712Snn35248  *	possible.
1552712Snn35248  *
1562712Snn35248  *	The non-success exit values are large to avoid accidental collision
1572712Snn35248  *	with values used internally by some commands (e.g. "Z_ERR" and
1582712Snn35248  *	"Z_USAGE" as used by zoneadm.)
1592712Snn35248  */
1602712Snn35248 #define	ZONE_SUBPROC_OK			0
1612712Snn35248 #define	ZONE_SUBPROC_USAGE		253
1622712Snn35248 #define	ZONE_SUBPROC_NOTCOMPLETE	254
1632712Snn35248 #define	ZONE_SUBPROC_FATAL		255
1642712Snn35248 
1650Sstevel@tonic-gate #ifdef _SYSCALL32
1660Sstevel@tonic-gate typedef struct {
1670Sstevel@tonic-gate 	caddr32_t zone_name;
1680Sstevel@tonic-gate 	caddr32_t zone_root;
1690Sstevel@tonic-gate 	caddr32_t zone_privs;
170813Sdp 	size32_t zone_privssz;
1710Sstevel@tonic-gate 	caddr32_t rctlbuf;
1720Sstevel@tonic-gate 	size32_t rctlbufsz;
1730Sstevel@tonic-gate 	caddr32_t extended_error;
174789Sahrens 	caddr32_t zfsbuf;
175789Sahrens 	size32_t  zfsbufsz;
1762110Srica 	int match;			/* match level */
1772110Srica 	uint32_t doi;			/* DOI for label */
1782110Srica 	caddr32_t label;		/* label associated with zone */
1793448Sdh155122 	int flags;
1800Sstevel@tonic-gate } zone_def32;
1810Sstevel@tonic-gate #endif
1820Sstevel@tonic-gate typedef struct {
1830Sstevel@tonic-gate 	const char *zone_name;
1840Sstevel@tonic-gate 	const char *zone_root;
1850Sstevel@tonic-gate 	const struct priv_set *zone_privs;
186813Sdp 	size_t zone_privssz;
1870Sstevel@tonic-gate 	const char *rctlbuf;
1880Sstevel@tonic-gate 	size_t rctlbufsz;
1890Sstevel@tonic-gate 	int *extended_error;
190789Sahrens 	const char *zfsbuf;
191789Sahrens 	size_t zfsbufsz;
1921676Sjpk 	int match;			/* match level */
1932110Srica 	uint32_t doi;			/* DOI for label */
1941676Sjpk 	const bslabel_t *label;		/* label associated with zone */
1953448Sdh155122 	int flags;
1960Sstevel@tonic-gate } zone_def;
1970Sstevel@tonic-gate 
1980Sstevel@tonic-gate /* extended error information */
1990Sstevel@tonic-gate #define	ZE_UNKNOWN	0	/* No extended error info */
2000Sstevel@tonic-gate #define	ZE_CHROOTED	1	/* tried to zone_create from chroot */
2010Sstevel@tonic-gate #define	ZE_AREMOUNTS	2	/* there are mounts within the zone */
2024791Ston #define	ZE_LABELINUSE	3	/* label is already in use by some other zone */
2030Sstevel@tonic-gate 
2040Sstevel@tonic-gate /* zone_status */
2050Sstevel@tonic-gate typedef enum {
2060Sstevel@tonic-gate 	ZONE_IS_UNINITIALIZED = 0,
2075880Snordmark 	ZONE_IS_INITIALIZED,
2080Sstevel@tonic-gate 	ZONE_IS_READY,
2090Sstevel@tonic-gate 	ZONE_IS_BOOTING,
2100Sstevel@tonic-gate 	ZONE_IS_RUNNING,
2110Sstevel@tonic-gate 	ZONE_IS_SHUTTING_DOWN,
2120Sstevel@tonic-gate 	ZONE_IS_EMPTY,
2130Sstevel@tonic-gate 	ZONE_IS_DOWN,
2140Sstevel@tonic-gate 	ZONE_IS_DYING,
2150Sstevel@tonic-gate 	ZONE_IS_DEAD
2160Sstevel@tonic-gate } zone_status_t;
2170Sstevel@tonic-gate #define	ZONE_MIN_STATE		ZONE_IS_UNINITIALIZED
2180Sstevel@tonic-gate #define	ZONE_MAX_STATE		ZONE_IS_DEAD
2190Sstevel@tonic-gate 
2200Sstevel@tonic-gate /*
2210Sstevel@tonic-gate  * Valid commands which may be issued by zoneadm to zoneadmd.  The kernel also
2220Sstevel@tonic-gate  * communicates with zoneadmd, but only uses Z_REBOOT and Z_HALT.
2230Sstevel@tonic-gate  */
2240Sstevel@tonic-gate typedef enum zone_cmd {
2252712Snn35248 	Z_READY, Z_BOOT, Z_FORCEBOOT, Z_REBOOT, Z_HALT, Z_NOTE_UNINSTALLING,
2262712Snn35248 	Z_MOUNT, Z_FORCEMOUNT, Z_UNMOUNT
2270Sstevel@tonic-gate } zone_cmd_t;
2280Sstevel@tonic-gate 
2290Sstevel@tonic-gate /*
2300Sstevel@tonic-gate  * The structure of a request to zoneadmd.
2310Sstevel@tonic-gate  */
2320Sstevel@tonic-gate typedef struct zone_cmd_arg {
2330Sstevel@tonic-gate 	uint64_t	uniqid;		/* unique "generation number" */
2340Sstevel@tonic-gate 	zone_cmd_t	cmd;		/* requested action */
2352267Sdp 	uint32_t	_pad;		/* need consistent 32/64 bit alignmt */
2360Sstevel@tonic-gate 	char locale[MAXPATHLEN];	/* locale in which to render messages */
2372267Sdp 	char bootbuf[BOOTARGS_MAX];	/* arguments passed to zone_boot() */
2380Sstevel@tonic-gate } zone_cmd_arg_t;
2390Sstevel@tonic-gate 
2400Sstevel@tonic-gate /*
2410Sstevel@tonic-gate  * Structure of zoneadmd's response to a request.  A NULL return value means
2420Sstevel@tonic-gate  * the caller should attempt to restart zoneadmd and retry.
2430Sstevel@tonic-gate  */
2440Sstevel@tonic-gate typedef struct zone_cmd_rval {
2450Sstevel@tonic-gate 	int rval;			/* return value of request */
2460Sstevel@tonic-gate 	char errbuf[1];	/* variable-sized buffer containing error messages */
2470Sstevel@tonic-gate } zone_cmd_rval_t;
2480Sstevel@tonic-gate 
2490Sstevel@tonic-gate /*
2500Sstevel@tonic-gate  * The zone support infrastructure uses the zone name as a component
2510Sstevel@tonic-gate  * of unix domain (AF_UNIX) sockets, which are limited to 108 characters
2520Sstevel@tonic-gate  * in length, so ZONENAME_MAX is limited by that.
2530Sstevel@tonic-gate  */
2540Sstevel@tonic-gate #define	ZONENAME_MAX		64
2550Sstevel@tonic-gate 
2560Sstevel@tonic-gate #define	GLOBAL_ZONENAME		"global"
2570Sstevel@tonic-gate 
2580Sstevel@tonic-gate /*
2590Sstevel@tonic-gate  * Extended Regular expression (see regex(5)) which matches all valid zone
2600Sstevel@tonic-gate  * names.
2610Sstevel@tonic-gate  */
2620Sstevel@tonic-gate #define	ZONENAME_REGEXP		"[a-zA-Z0-9][-_.a-zA-Z0-9]{0,62}"
2630Sstevel@tonic-gate 
2640Sstevel@tonic-gate /*
2650Sstevel@tonic-gate  * Where the zones support infrastructure places temporary files.
2660Sstevel@tonic-gate  */
2670Sstevel@tonic-gate #define	ZONES_TMPDIR		"/var/run/zones"
2680Sstevel@tonic-gate 
2690Sstevel@tonic-gate /*
2700Sstevel@tonic-gate  * The path to the door used by clients to communicate with zoneadmd.
2710Sstevel@tonic-gate  */
2720Sstevel@tonic-gate #define	ZONE_DOOR_PATH		ZONES_TMPDIR "/%s.zoneadmd_door"
2730Sstevel@tonic-gate 
2743448Sdh155122 /* zone_flags */
2753448Sdh155122 #define	ZF_HASHED_LABEL		0x2	/* zone has a unique label */
2763448Sdh155122 #define	ZF_IS_SCRATCH		0x4	/* scratch zone */
2773448Sdh155122 #define	ZF_NET_EXCL		0x8	/* Zone has an exclusive IP stack */
2783448Sdh155122 
2793448Sdh155122 /* zone_create flags */
2803448Sdh155122 #define	ZCF_NET_EXCL		0x1	/* Create a zone with exclusive IP */
2813448Sdh155122 
2820Sstevel@tonic-gate #ifdef _KERNEL
2830Sstevel@tonic-gate /*
2840Sstevel@tonic-gate  * We need to protect the definition of 'list_t' from userland applications and
2850Sstevel@tonic-gate  * libraries which may be defining ther own versions.
2860Sstevel@tonic-gate  */
2870Sstevel@tonic-gate #include <sys/list.h>
2880Sstevel@tonic-gate 
2890Sstevel@tonic-gate #define	GLOBAL_ZONEUNIQID	0	/* uniqid of the global zone */
2900Sstevel@tonic-gate 
2910Sstevel@tonic-gate struct pool;
2922712Snn35248 struct brand;
2930Sstevel@tonic-gate 
294789Sahrens /*
295789Sahrens  * Structure to record list of ZFS datasets exported to a zone.
296789Sahrens  */
297789Sahrens typedef struct zone_dataset {
298789Sahrens 	char		*zd_dataset;
299789Sahrens 	list_node_t	zd_linkage;
300789Sahrens } zone_dataset_t;
301789Sahrens 
3023247Sgjelinek /*
3033247Sgjelinek  * structure for zone kstats
3043247Sgjelinek  */
3053247Sgjelinek typedef struct zone_kstat {
3063247Sgjelinek 	kstat_named_t zk_zonename;
3073247Sgjelinek 	kstat_named_t zk_usage;
3083247Sgjelinek 	kstat_named_t zk_value;
3093247Sgjelinek } zone_kstat_t;
3103247Sgjelinek 
3113792Sakolb struct cpucap;
3123792Sakolb 
3130Sstevel@tonic-gate typedef struct zone {
3140Sstevel@tonic-gate 	/*
3150Sstevel@tonic-gate 	 * zone_name is never modified once set.
3160Sstevel@tonic-gate 	 */
3170Sstevel@tonic-gate 	char		*zone_name;	/* zone's configuration name */
3180Sstevel@tonic-gate 	/*
3190Sstevel@tonic-gate 	 * zone_nodename and zone_domain are never freed once allocated.
3200Sstevel@tonic-gate 	 */
3210Sstevel@tonic-gate 	char		*zone_nodename;	/* utsname.nodename equivalent */
3220Sstevel@tonic-gate 	char		*zone_domain;	/* srpc_domain equivalent */
3230Sstevel@tonic-gate 	/*
3248662SJordan.Vaughan@Sun.com 	 * zone_hostid is used for per-zone hostid emulation.
3258662SJordan.Vaughan@Sun.com 	 * Currently it isn't modified after it's set (so no locks protect
3268662SJordan.Vaughan@Sun.com 	 * accesses), but that might have to change when we allow
3278662SJordan.Vaughan@Sun.com 	 * administrators to change running zones' properties.
3288662SJordan.Vaughan@Sun.com 	 *
3298662SJordan.Vaughan@Sun.com 	 * The global zone's zone_hostid must always be HW_INVALID_HOSTID so
3308662SJordan.Vaughan@Sun.com 	 * that zone_get_hostid() will function correctly.
3318662SJordan.Vaughan@Sun.com 	 */
3328662SJordan.Vaughan@Sun.com 	uint32_t	zone_hostid;	/* zone's hostid, HW_INVALID_HOSTID */
3338662SJordan.Vaughan@Sun.com 					/* if not emulated */
3348662SJordan.Vaughan@Sun.com 	/*
3350Sstevel@tonic-gate 	 * zone_lock protects the following fields of a zone_t:
3360Sstevel@tonic-gate 	 * 	zone_ref
3370Sstevel@tonic-gate 	 * 	zone_cred_ref
3380Sstevel@tonic-gate 	 * 	zone_ntasks
3390Sstevel@tonic-gate 	 * 	zone_flags
3400Sstevel@tonic-gate 	 * 	zone_zsd
34112273SCasper.Dik@Sun.COM 	 *	zone_pfexecd
3420Sstevel@tonic-gate 	 */
3430Sstevel@tonic-gate 	kmutex_t	zone_lock;
3440Sstevel@tonic-gate 	/*
3450Sstevel@tonic-gate 	 * zone_linkage is the zone's linkage into the active or
3460Sstevel@tonic-gate 	 * death-row list.  The field is protected by zonehash_lock.
3470Sstevel@tonic-gate 	 */
3480Sstevel@tonic-gate 	list_node_t	zone_linkage;
3490Sstevel@tonic-gate 	zoneid_t	zone_id;	/* ID of zone */
3500Sstevel@tonic-gate 	uint_t		zone_ref;	/* count of zone_hold()s on zone */
3510Sstevel@tonic-gate 	uint_t		zone_cred_ref;	/* count of zone_hold_cred()s on zone */
3520Sstevel@tonic-gate 	/*
3530Sstevel@tonic-gate 	 * zone_rootvp and zone_rootpath can never be modified once set.
3540Sstevel@tonic-gate 	 */
3550Sstevel@tonic-gate 	struct vnode	*zone_rootvp;	/* zone's root vnode */
3560Sstevel@tonic-gate 	char		*zone_rootpath;	/* Path to zone's root + '/' */
3570Sstevel@tonic-gate 	ushort_t	zone_flags;	/* misc flags */
3580Sstevel@tonic-gate 	zone_status_t	zone_status;	/* protected by zone_status_lock */
3590Sstevel@tonic-gate 	uint_t		zone_ntasks;	/* number of tasks executing in zone */
3600Sstevel@tonic-gate 	kmutex_t	zone_nlwps_lock; /* protects zone_nlwps, and *_nlwps */
3610Sstevel@tonic-gate 					/* counters in projects and tasks */
3620Sstevel@tonic-gate 					/* that are within the zone */
3630Sstevel@tonic-gate 	rctl_qty_t	zone_nlwps;	/* number of lwps in zone */
3640Sstevel@tonic-gate 	rctl_qty_t	zone_nlwps_ctl; /* protected by zone_rctls->rcs_lock */
3652677Sml93401 	rctl_qty_t	zone_shmmax;	/* System V shared memory usage */
3662677Sml93401 	ipc_rqty_t	zone_ipc;	/* System V IPC id resource usage */
3670Sstevel@tonic-gate 
3680Sstevel@tonic-gate 	uint_t		zone_rootpathlen; /* strlen(zone_rootpath) + 1 */
3690Sstevel@tonic-gate 	uint32_t	zone_shares;	/* FSS shares allocated to zone */
3700Sstevel@tonic-gate 	rctl_set_t	*zone_rctls;	/* zone-wide (zone.*) rctls */
3713247Sgjelinek 	kmutex_t	zone_mem_lock;	/* protects zone_locked_mem and */
3722768Ssl108498 					/* kpd_locked_mem for all */
3733247Sgjelinek 					/* projects in zone. */
3743247Sgjelinek 					/* Also protects zone_max_swap */
3752768Ssl108498 					/* grab after p_lock, before rcs_lock */
3763247Sgjelinek 	rctl_qty_t	zone_locked_mem;	/* bytes of locked memory in */
3773247Sgjelinek 						/* zone */
3783247Sgjelinek 	rctl_qty_t	zone_locked_mem_ctl;	/* Current locked memory */
3792768Ssl108498 						/* limit.  Protected by */
3802768Ssl108498 						/* zone_rctls->rcs_lock */
3813247Sgjelinek 	rctl_qty_t	zone_max_swap; /* bytes of swap reserved by zone */
3823247Sgjelinek 	rctl_qty_t	zone_max_swap_ctl;	/* current swap limit. */
3833247Sgjelinek 						/* Protected by */
3843247Sgjelinek 						/* zone_rctls->rcs_lock */
38512633Sjohn.levon@sun.com 	kmutex_t	zone_rctl_lock;	/* protects zone_max_lofi */
38612633Sjohn.levon@sun.com 	rctl_qty_t	zone_max_lofi; /* lofi devs for zone */
38712633Sjohn.levon@sun.com 	rctl_qty_t	zone_max_lofi_ctl;	/* current lofi limit. */
38812633Sjohn.levon@sun.com 						/* Protected by */
38912633Sjohn.levon@sun.com 						/* zone_rctls->rcs_lock */
3900Sstevel@tonic-gate 	list_t		zone_zsd;	/* list of Zone-Specific Data values */
3910Sstevel@tonic-gate 	kcondvar_t	zone_cv;	/* used to signal state changes */
3920Sstevel@tonic-gate 	struct proc	*zone_zsched;	/* Dummy kernel "zsched" process */
3930Sstevel@tonic-gate 	pid_t		zone_proc_initpid; /* pid of "init" for this zone */
3942267Sdp 	char		*zone_initname;	/* fs path to 'init' */
3950Sstevel@tonic-gate 	int		zone_boot_err;  /* for zone_boot() if boot fails */
3960Sstevel@tonic-gate 	char		*zone_bootargs;	/* arguments passed via zone_boot() */
3973247Sgjelinek 	uint64_t	zone_phys_mcap;	/* physical memory cap */
3980Sstevel@tonic-gate 	/*
3990Sstevel@tonic-gate 	 * zone_kthreads is protected by zone_status_lock.
4000Sstevel@tonic-gate 	 */
4010Sstevel@tonic-gate 	kthread_t	*zone_kthreads;	/* kernel threads in zone */
4020Sstevel@tonic-gate 	struct priv_set	*zone_privset;	/* limit set for zone */
4030Sstevel@tonic-gate 	/*
4040Sstevel@tonic-gate 	 * zone_vfslist is protected by vfs_list_lock().
4050Sstevel@tonic-gate 	 */
4060Sstevel@tonic-gate 	struct vfs	*zone_vfslist;	/* list of FS's mounted in zone */
4070Sstevel@tonic-gate 	uint64_t	zone_uniqid;	/* unique zone generation number */
4080Sstevel@tonic-gate 	struct cred	*zone_kcred;	/* kcred-like, zone-limited cred */
4090Sstevel@tonic-gate 	/*
4100Sstevel@tonic-gate 	 * zone_pool is protected by pool_lock().
4110Sstevel@tonic-gate 	 */
4120Sstevel@tonic-gate 	struct pool	*zone_pool;	/* pool the zone is bound to */
4130Sstevel@tonic-gate 	hrtime_t	zone_pool_mod;	/* last pool bind modification time */
4140Sstevel@tonic-gate 	/* zone_psetid is protected by cpu_lock */
4150Sstevel@tonic-gate 	psetid_t	zone_psetid;	/* pset the zone is bound to */
4160Sstevel@tonic-gate 	/*
4170Sstevel@tonic-gate 	 * The following two can be read without holding any locks.  They are
4180Sstevel@tonic-gate 	 * updated under cpu_lock.
4190Sstevel@tonic-gate 	 */
4200Sstevel@tonic-gate 	int		zone_ncpus;  /* zone's idea of ncpus */
4210Sstevel@tonic-gate 	int		zone_ncpus_online; /* zone's idea of ncpus_online */
422789Sahrens 	/*
423789Sahrens 	 * List of ZFS datasets exported to this zone.
424789Sahrens 	 */
425789Sahrens 	list_t		zone_datasets;	/* list of datasets */
4261676Sjpk 
4272110Srica 	ts_label_t	*zone_slabel;	/* zone sensitivity label */
4282110Srica 	int		zone_match;	/* require label match for packets */
4291676Sjpk 	tsol_mlp_list_t zone_mlps;	/* MLPs on zone-private addresses */
4302267Sdp 
4312712Snn35248 	boolean_t	zone_restart_init;	/* Restart init if it dies? */
4322712Snn35248 	struct brand	*zone_brand;		/* zone's brand */
4334888Seh208807 	void 		*zone_brand_data;	/* store brand specific data */
4343247Sgjelinek 	id_t		zone_defaultcid;	/* dflt scheduling class id */
4353247Sgjelinek 	kstat_t		*zone_swapresv_kstat;
4363247Sgjelinek 	kstat_t		*zone_lockedmem_kstat;
4373448Sdh155122 	/*
4383448Sdh155122 	 * zone_dl_list is protected by zone_lock
4393448Sdh155122 	 */
44010616SSebastien.Roy@Sun.COM 	list_t		zone_dl_list;
4413448Sdh155122 	netstack_t	*zone_netstack;
4423792Sakolb 	struct cpucap	*zone_cpucap;	/* CPU caps data */
4434197Spaulson 	/*
4444197Spaulson 	 * Solaris Auditing per-zone audit context
4454197Spaulson 	 */
4464197Spaulson 	struct au_kcontext	*zone_audit_kctxt;
44710910SRobert.Harris@Sun.COM 	/*
44810910SRobert.Harris@Sun.COM 	 * For private use by mntfs.
44910910SRobert.Harris@Sun.COM 	 */
45010910SRobert.Harris@Sun.COM 	struct mntelem	*zone_mntfs_db;
45110910SRobert.Harris@Sun.COM 	krwlock_t	zone_mntfs_db_lock;
45212273SCasper.Dik@Sun.COM 
45312273SCasper.Dik@Sun.COM 	struct klpd_reg		*zone_pfexecd;
45412633Sjohn.levon@sun.com 
45512633Sjohn.levon@sun.com 	char		*zone_fs_allowed;
456*12725SMenno.Lageman@Sun.COM 	rctl_qty_t	zone_nprocs;	/* number of processes in the zone */
457*12725SMenno.Lageman@Sun.COM 	rctl_qty_t	zone_nprocs_ctl;	/* current limit protected by */
458*12725SMenno.Lageman@Sun.COM 						/* zone_rctls->rcs_lock */
459*12725SMenno.Lageman@Sun.COM 	kstat_t		*zone_nprocs_kstat;
4600Sstevel@tonic-gate } zone_t;
4610Sstevel@tonic-gate 
4620Sstevel@tonic-gate /*
4630Sstevel@tonic-gate  * Special value of zone_psetid to indicate that pools are disabled.
4640Sstevel@tonic-gate  */
4650Sstevel@tonic-gate #define	ZONE_PS_INVAL	PS_MYID
4660Sstevel@tonic-gate 
4673448Sdh155122 
4680Sstevel@tonic-gate extern zone_t zone0;
4690Sstevel@tonic-gate extern zone_t *global_zone;
4700Sstevel@tonic-gate extern uint_t maxzones;
4710Sstevel@tonic-gate extern rctl_hndl_t rc_zone_nlwps;
472*12725SMenno.Lageman@Sun.COM extern rctl_hndl_t rc_zone_nprocs;
4730Sstevel@tonic-gate 
474789Sahrens extern long zone(int, void *, void *, void *, void *);
4750Sstevel@tonic-gate extern void zone_zsd_init(void);
4760Sstevel@tonic-gate extern void zone_init(void);
4770Sstevel@tonic-gate extern void zone_hold(zone_t *);
4780Sstevel@tonic-gate extern void zone_rele(zone_t *);
4790Sstevel@tonic-gate extern void zone_cred_hold(zone_t *);
4800Sstevel@tonic-gate extern void zone_cred_rele(zone_t *);
4810Sstevel@tonic-gate extern void zone_task_hold(zone_t *);
4820Sstevel@tonic-gate extern void zone_task_rele(zone_t *);
4830Sstevel@tonic-gate extern zone_t *zone_find_by_id(zoneid_t);
4841676Sjpk extern zone_t *zone_find_by_label(const ts_label_t *);
4850Sstevel@tonic-gate extern zone_t *zone_find_by_name(char *);
4861769Scarlsonj extern zone_t *zone_find_by_any_path(const char *, boolean_t);
4870Sstevel@tonic-gate extern zone_t *zone_find_by_path(const char *);
4880Sstevel@tonic-gate extern zoneid_t getzoneid(void);
4893448Sdh155122 extern zone_t *zone_find_by_id_nolock(zoneid_t);
49010616SSebastien.Roy@Sun.COM extern int zone_datalink_walk(zoneid_t, int (*)(datalink_id_t, void *), void *);
49110616SSebastien.Roy@Sun.COM extern int zone_check_datalink(zoneid_t *, datalink_id_t);
4920Sstevel@tonic-gate 
4930Sstevel@tonic-gate /*
4940Sstevel@tonic-gate  * Zone-specific data (ZSD) APIs
4950Sstevel@tonic-gate  */
4960Sstevel@tonic-gate /*
4970Sstevel@tonic-gate  * The following is what code should be initializing its zone_key_t to if it
4980Sstevel@tonic-gate  * calls zone_getspecific() without necessarily knowing that zone_key_create()
4990Sstevel@tonic-gate  * has been called on the key.
5000Sstevel@tonic-gate  */
5010Sstevel@tonic-gate #define	ZONE_KEY_UNINITIALIZED	0
5020Sstevel@tonic-gate 
5030Sstevel@tonic-gate typedef uint_t zone_key_t;
5040Sstevel@tonic-gate 
5050Sstevel@tonic-gate extern void	zone_key_create(zone_key_t *, void *(*)(zoneid_t),
5060Sstevel@tonic-gate     void (*)(zoneid_t, void *), void (*)(zoneid_t, void *));
5070Sstevel@tonic-gate extern int 	zone_key_delete(zone_key_t);
5080Sstevel@tonic-gate extern void	*zone_getspecific(zone_key_t, zone_t *);
5090Sstevel@tonic-gate extern int	zone_setspecific(zone_key_t, zone_t *, const void *);
5100Sstevel@tonic-gate 
5110Sstevel@tonic-gate /*
5120Sstevel@tonic-gate  * The definition of a zsd_entry is truly private to zone.c and is only
5130Sstevel@tonic-gate  * placed here so it can be shared with mdb.
5145880Snordmark  *
5155880Snordmark  * State maintained for each zone times each registered key, which tracks
5165880Snordmark  * the state of the create, shutdown and destroy callbacks.
5175880Snordmark  *
5185880Snordmark  * zsd_flags is used to keep track of pending actions to avoid holding locks
5195880Snordmark  * when calling the create/shutdown/destroy callbacks, since doing so
5205880Snordmark  * could lead to deadlocks.
5210Sstevel@tonic-gate  */
5220Sstevel@tonic-gate struct zsd_entry {
5230Sstevel@tonic-gate 	zone_key_t		zsd_key;	/* Key used to lookup value */
5240Sstevel@tonic-gate 	void			*zsd_data;	/* Caller-managed value */
5250Sstevel@tonic-gate 	/*
5260Sstevel@tonic-gate 	 * Callbacks to be executed when a zone is created, shutdown, and
5270Sstevel@tonic-gate 	 * destroyed, respectively.
5280Sstevel@tonic-gate 	 */
5290Sstevel@tonic-gate 	void			*(*zsd_create)(zoneid_t);
5300Sstevel@tonic-gate 	void			(*zsd_shutdown)(zoneid_t, void *);
5310Sstevel@tonic-gate 	void			(*zsd_destroy)(zoneid_t, void *);
5320Sstevel@tonic-gate 	list_node_t		zsd_linkage;
5335880Snordmark 	uint16_t 		zsd_flags;	/* See below */
5345880Snordmark 	kcondvar_t		zsd_cv;
5350Sstevel@tonic-gate };
5360Sstevel@tonic-gate 
5370Sstevel@tonic-gate /*
5385880Snordmark  * zsd_flags
5395880Snordmark  */
5405880Snordmark #define	ZSD_CREATE_NEEDED	0x0001
5415880Snordmark #define	ZSD_CREATE_INPROGRESS	0x0002
5425880Snordmark #define	ZSD_CREATE_COMPLETED	0x0004
5435880Snordmark #define	ZSD_SHUTDOWN_NEEDED	0x0010
5445880Snordmark #define	ZSD_SHUTDOWN_INPROGRESS	0x0020
5455880Snordmark #define	ZSD_SHUTDOWN_COMPLETED	0x0040
5465880Snordmark #define	ZSD_DESTROY_NEEDED	0x0100
5475880Snordmark #define	ZSD_DESTROY_INPROGRESS	0x0200
5485880Snordmark #define	ZSD_DESTROY_COMPLETED	0x0400
5495880Snordmark 
5505880Snordmark #define	ZSD_CREATE_ALL	\
5515880Snordmark 	(ZSD_CREATE_NEEDED|ZSD_CREATE_INPROGRESS|ZSD_CREATE_COMPLETED)
5525880Snordmark #define	ZSD_SHUTDOWN_ALL	\
5535880Snordmark 	(ZSD_SHUTDOWN_NEEDED|ZSD_SHUTDOWN_INPROGRESS|ZSD_SHUTDOWN_COMPLETED)
5545880Snordmark #define	ZSD_DESTROY_ALL	\
5555880Snordmark 	(ZSD_DESTROY_NEEDED|ZSD_DESTROY_INPROGRESS|ZSD_DESTROY_COMPLETED)
5565880Snordmark 
5575880Snordmark #define	ZSD_ALL_INPROGRESS \
5585880Snordmark 	(ZSD_CREATE_INPROGRESS|ZSD_SHUTDOWN_INPROGRESS|ZSD_DESTROY_INPROGRESS)
5595880Snordmark 
5605880Snordmark /*
5610Sstevel@tonic-gate  * Macros to help with zone visibility restrictions.
5620Sstevel@tonic-gate  */
5630Sstevel@tonic-gate 
5640Sstevel@tonic-gate /*
5650Sstevel@tonic-gate  * Is process in the global zone?
5660Sstevel@tonic-gate  */
5670Sstevel@tonic-gate #define	INGLOBALZONE(p) \
5680Sstevel@tonic-gate 	((p)->p_zone == global_zone)
5690Sstevel@tonic-gate 
5700Sstevel@tonic-gate /*
5710Sstevel@tonic-gate  * Can process view objects in given zone?
5720Sstevel@tonic-gate  */
5730Sstevel@tonic-gate #define	HASZONEACCESS(p, zoneid) \
5740Sstevel@tonic-gate 	((p)->p_zone->zone_id == (zoneid) || INGLOBALZONE(p))
5750Sstevel@tonic-gate 
5760Sstevel@tonic-gate /*
5770Sstevel@tonic-gate  * Convenience macro to see if a resolved path is visible from within a
5780Sstevel@tonic-gate  * given zone.
5790Sstevel@tonic-gate  *
5800Sstevel@tonic-gate  * The basic idea is that the first (zone_rootpathlen - 1) bytes of the
5810Sstevel@tonic-gate  * two strings must be equal.  Since the rootpathlen has a trailing '/',
5820Sstevel@tonic-gate  * we want to skip everything in the path up to (but not including) the
5830Sstevel@tonic-gate  * trailing '/'.
5840Sstevel@tonic-gate  */
5850Sstevel@tonic-gate #define	ZONE_PATH_VISIBLE(path, zone) \
5860Sstevel@tonic-gate 	(strncmp((path), (zone)->zone_rootpath,		\
5870Sstevel@tonic-gate 	    (zone)->zone_rootpathlen - 1) == 0)
5880Sstevel@tonic-gate 
5890Sstevel@tonic-gate /*
5900Sstevel@tonic-gate  * Convenience macro to go from the global view of a path to that seen
5910Sstevel@tonic-gate  * from within said zone.  It is the responsibility of the caller to
5920Sstevel@tonic-gate  * ensure that the path is a resolved one (ie, no '..'s or '.'s), and is
5930Sstevel@tonic-gate  * in fact visible from within the zone.
5940Sstevel@tonic-gate  */
5950Sstevel@tonic-gate #define	ZONE_PATH_TRANSLATE(path, zone)	\
5960Sstevel@tonic-gate 	(ASSERT(ZONE_PATH_VISIBLE(path, zone)),	\
5970Sstevel@tonic-gate 	(path) + (zone)->zone_rootpathlen - 2)
5980Sstevel@tonic-gate 
5990Sstevel@tonic-gate /*
6000Sstevel@tonic-gate  * Special processes visible in all zones.
6010Sstevel@tonic-gate  */
6020Sstevel@tonic-gate #define	ZONE_SPECIALPID(x)	 ((x) == 0 || (x) == 1)
6030Sstevel@tonic-gate 
6040Sstevel@tonic-gate /*
6050Sstevel@tonic-gate  * Zone-safe version of thread_create() to be used when the caller wants to
6060Sstevel@tonic-gate  * create a kernel thread to run within the current zone's context.
6070Sstevel@tonic-gate  */
6080Sstevel@tonic-gate extern kthread_t *zthread_create(caddr_t, size_t, void (*)(), void *, size_t,
6090Sstevel@tonic-gate     pri_t);
6100Sstevel@tonic-gate extern void zthread_exit(void);
6110Sstevel@tonic-gate 
6120Sstevel@tonic-gate /*
6130Sstevel@tonic-gate  * Functions for an external observer to register interest in a zone's status
6140Sstevel@tonic-gate  * change.  Observers will be woken up when the zone status equals the status
6150Sstevel@tonic-gate  * argument passed in (in the case of zone_status_timedwait, the function may
6160Sstevel@tonic-gate  * also return because of a timeout; zone_status_wait_sig may return early due
6170Sstevel@tonic-gate  * to a signal being delivered; zone_status_timedwait_sig may return for any of
6180Sstevel@tonic-gate  * the above reasons).
6190Sstevel@tonic-gate  *
6200Sstevel@tonic-gate  * Otherwise these behave identically to cv_timedwait(), cv_wait(), and
6210Sstevel@tonic-gate  * cv_wait_sig() respectively.
6220Sstevel@tonic-gate  */
6230Sstevel@tonic-gate extern clock_t zone_status_timedwait(zone_t *, clock_t, zone_status_t);
6240Sstevel@tonic-gate extern clock_t zone_status_timedwait_sig(zone_t *, clock_t, zone_status_t);
6250Sstevel@tonic-gate extern void zone_status_wait(zone_t *, zone_status_t);
6260Sstevel@tonic-gate extern int zone_status_wait_sig(zone_t *, zone_status_t);
6270Sstevel@tonic-gate 
6280Sstevel@tonic-gate /*
6290Sstevel@tonic-gate  * Get the status  of the zone (at the time it was called).  The state may
6300Sstevel@tonic-gate  * have progressed by the time it is returned.
6310Sstevel@tonic-gate  */
6320Sstevel@tonic-gate extern zone_status_t zone_status_get(zone_t *);
6330Sstevel@tonic-gate 
6340Sstevel@tonic-gate /*
6358662SJordan.Vaughan@Sun.com  * Safely get the hostid of the specified zone (defaults to machine's hostid
6368662SJordan.Vaughan@Sun.com  * if the specified zone doesn't emulate a hostid).  Passing NULL retrieves
6378662SJordan.Vaughan@Sun.com  * the global zone's (i.e., physical system's) hostid.
6388662SJordan.Vaughan@Sun.com  */
6398662SJordan.Vaughan@Sun.com extern uint32_t zone_get_hostid(zone_t *);
6408662SJordan.Vaughan@Sun.com 
6418662SJordan.Vaughan@Sun.com /*
6420Sstevel@tonic-gate  * Get the "kcred" credentials corresponding to the given zone.
6430Sstevel@tonic-gate  */
6440Sstevel@tonic-gate extern struct cred *zone_get_kcred(zoneid_t);
6450Sstevel@tonic-gate 
6460Sstevel@tonic-gate /*
6470Sstevel@tonic-gate  * Get/set the pool the zone is currently bound to.
6480Sstevel@tonic-gate  */
6490Sstevel@tonic-gate extern struct pool *zone_pool_get(zone_t *);
6500Sstevel@tonic-gate extern void zone_pool_set(zone_t *, struct pool *);
6510Sstevel@tonic-gate 
6520Sstevel@tonic-gate /*
6530Sstevel@tonic-gate  * Get/set the pset the zone is currently using.
6540Sstevel@tonic-gate  */
6550Sstevel@tonic-gate extern psetid_t zone_pset_get(zone_t *);
6560Sstevel@tonic-gate extern void zone_pset_set(zone_t *, psetid_t);
6570Sstevel@tonic-gate 
6580Sstevel@tonic-gate /*
6590Sstevel@tonic-gate  * Get the number of cpus/online-cpus visible from the given zone.
6600Sstevel@tonic-gate  */
6610Sstevel@tonic-gate extern int zone_ncpus_get(zone_t *);
6620Sstevel@tonic-gate extern int zone_ncpus_online_get(zone_t *);
6630Sstevel@tonic-gate 
6640Sstevel@tonic-gate /*
665789Sahrens  * Returns true if the named pool/dataset is visible in the current zone.
666789Sahrens  */
667789Sahrens extern int zone_dataset_visible(const char *, int *);
668789Sahrens 
669789Sahrens /*
6702267Sdp  * zone version of kadmin()
6710Sstevel@tonic-gate  */
6722267Sdp extern int zone_kadmin(int, int, const char *, cred_t *);
6730Sstevel@tonic-gate extern void zone_shutdown_global(void);
6740Sstevel@tonic-gate 
6750Sstevel@tonic-gate extern void mount_in_progress(void);
6760Sstevel@tonic-gate extern void mount_completed(void);
6770Sstevel@tonic-gate 
6780Sstevel@tonic-gate extern int zone_walk(int (*)(zone_t *, void *), void *);
6790Sstevel@tonic-gate 
6802768Ssl108498 extern rctl_hndl_t rc_zone_locked_mem;
6813247Sgjelinek extern rctl_hndl_t rc_zone_max_swap;
68212633Sjohn.levon@sun.com extern rctl_hndl_t rc_zone_max_lofi;
6832768Ssl108498 
6840Sstevel@tonic-gate #endif	/* _KERNEL */
6850Sstevel@tonic-gate 
6860Sstevel@tonic-gate #ifdef	__cplusplus
6870Sstevel@tonic-gate }
6880Sstevel@tonic-gate #endif
6890Sstevel@tonic-gate 
6900Sstevel@tonic-gate #endif	/* _SYS_ZONE_H */
691