d/freebsd-skq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
aa9bc3b171
freebsd-skq/sys/vm/vm_object.h
Konstantin Belousov 2a339d9e3d Add implementation of robust mutexes, hopefully close enough to the 
intention of the POSIX IEEE Std 1003.1TM-2008/Cor 1-2013.

A robust mutex is guaranteed to be cleared by the system upon either
thread or process owner termination while the mutex is held.  The next
mutex locker is then notified about inconsistent mutex state and can
execute (or abandon) corrective actions.

The patch mostly consists of small changes here and there, adding
neccessary checks for the inconsistent and abandoned conditions into
existing paths.  Additionally, the thread exit handler was extended to
iterate over the userspace-maintained list of owned robust mutexes,
unlocking and marking as terminated each of them.

The list of owned robust mutexes cannot be maintained atomically
synchronous with the mutex lock state (it is possible in kernel, but
is too expensive).  Instead, for the duration of lock or unlock
operation, the current mutex is remembered in a special slot that is
also checked by the kernel at thread termination.

Kernel must be aware about the per-thread location of the heads of
robust mutex lists and the current active mutex slot.  When a thread
touches a robust mutex for the first time, a new umtx op syscall is
issued which informs about location of lists heads.

The umtx sleep queues for PP and PI mutexes are split between
non-robust and robust.

Somewhat unrelated changes in the patch:
1. Style.
2. The fix for proper tdfind() call use in umtxq_sleep_pi() for shared
   pi mutexes.
3. Removal of the userspace struct pthread_mutex m_owner field.
4. The sysctl kern.ipc.umtx_vnode_persistent is added, which controls
   the lifetime of the shared mutex associated with a vnode' page.

Reviewed by:	jilles (previous version, supposedly the objection was fixed)
Discussed with:	brooks, Martin Simmons <martin@lispworks.com> (some aspects)
Tested by:	pho
Sponsored by:	The FreeBSD Foundation
2016-05-17 09:56:22 +00:00

336 lines
12 KiB
C
Raw Blame History

/*-
* Copyright (c) 1991, 1993
* The Regents of the University of California. All rights reserved.
*
* This code is derived from software contributed to Berkeley by
* The Mach Operating System project at Carnegie-Mellon University.
*
* 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.
*
* from: @(#)vm_object.h 8.3 (Berkeley) 1/12/94
*
*
* Copyright (c) 1987, 1990 Carnegie-Mellon University.
* All rights reserved.
*
* Authors: Avadis Tevanian, Jr., Michael Wayne Young
*
* Permission to use, copy, modify and distribute this software and
* its documentation is hereby granted, provided that both the copyright
* notice and this permission notice appear in all copies of the
* software, derivative works or modified versions, and any portions
* thereof, and that both notices appear in supporting documentation.
*
* CARNEGIE MELLON ALLOWS FREE USE OF THIS SOFTWARE IN ITS "AS IS"
* CONDITION. CARNEGIE MELLON DISCLAIMS ANY LIABILITY OF ANY KIND
* FOR ANY DAMAGES WHATSOEVER RESULTING FROM THE USE OF THIS SOFTWARE.
*
* Carnegie Mellon requests users of this software to return to
*
* Software Distribution Coordinator or Software.Distribution@CS.CMU.EDU
* School of Computer Science
* Carnegie Mellon University
* Pittsburgh PA 15213-3890
*
* any improvements or extensions that they make and grant Carnegie the
* rights to redistribute these changes.
*
* $FreeBSD$
*/
/*
* Virtual memory object module definitions.
*/
#ifndef _VM_OBJECT_
#define _VM_OBJECT_
#include <sys/queue.h>
#include <sys/_lock.h>
#include <sys/_mutex.h>
#include <sys/_rwlock.h>
#include <vm/_vm_radix.h>
/*
* Types defined:
*
* vm_object_t Virtual memory object.
*
* The root of cached pages pool is protected by both the per-object lock
* and the free pages queue mutex.
* On insert in the cache radix trie, the per-object lock is expected
* to be already held and the free pages queue mutex will be
* acquired during the operation too.
* On remove and lookup from the cache radix trie, only the free
* pages queue mutex is expected to be locked.
* These rules allow for reliably checking for the presence of cached
* pages with only the per-object lock held, thereby reducing contention
* for the free pages queue mutex.
*
* List of locks
* (c) const until freed
* (o) per-object lock
* (f) free pages queue mutex
*
*/
struct vm_object {
struct rwlock lock;
TAILQ_ENTRY(vm_object) object_list; /* list of all objects */
LIST_HEAD(, vm_object) shadow_head; /* objects that this is a shadow for */
LIST_ENTRY(vm_object) shadow_list; /* chain of shadow objects */
TAILQ_HEAD(respgs, vm_page) memq; /* list of resident pages */
struct vm_radix rtree; /* root of the resident page radix trie*/
vm_pindex_t size; /* Object size */
int generation; /* generation ID */
int ref_count; /* How many refs?? */
int shadow_count; /* how many objects that this is a shadow for */
vm_memattr_t memattr; /* default memory attribute for pages */
objtype_t type; /* type of pager */
u_short flags; /* see below */
u_short pg_color; /* (c) color of first page in obj */
u_int paging_in_progress; /* Paging (in or out) so don't collapse or destroy */
int resident_page_count; /* number of resident pages */
struct vm_object *backing_object; /* object that I'm a shadow of */
vm_ooffset_t backing_object_offset;/* Offset in backing object */
TAILQ_ENTRY(vm_object) pager_object_list; /* list of all objects of this pager type */
LIST_HEAD(, vm_reserv) rvq; /* list of reservations */
struct vm_radix cache; /* (o + f) root of the cache page radix trie */
void *handle;
union {
/*
* VNode pager
*
* vnp_size - current size of file
*/
struct {
off_t vnp_size;
vm_ooffset_t writemappings;
} vnp;
/*
* Device pager
*
* devp_pglist - list of allocated pages
*/
struct {
TAILQ_HEAD(, vm_page) devp_pglist;
struct cdev_pager_ops *ops;
struct cdev *dev;
} devp;
/*
* SG pager
*
* sgp_pglist - list of allocated pages
*/
struct {
TAILQ_HEAD(, vm_page) sgp_pglist;
} sgp;
/*
* Swap pager
*
* swp_tmpfs - back-pointer to the tmpfs vnode,
* if any, which uses the vm object
* as backing store. The handle
* cannot be reused for linking,
* because the vnode can be
* reclaimed and recreated, making
* the handle changed and hash-chain
* invalid.
*
* swp_bcount - number of swap 'swblock' metablocks, each
* contains up to 16 swapblk assignments.
* see vm/swap_pager.h
*/
struct {
void *swp_tmpfs;
int swp_bcount;
} swp;
} un_pager;
struct ucred *cred;
vm_ooffset_t charge;
void *umtx_data;
};
/*
* Flags
*/
#define OBJ_FICTITIOUS 0x0001 /* (c) contains fictitious pages */
#define OBJ_UNMANAGED 0x0002 /* (c) contains unmanaged pages */
#define OBJ_ACTIVE 0x0004 /* active objects */
#define OBJ_DEAD 0x0008 /* dead objects (during rundown) */
#define OBJ_NOSPLIT 0x0010 /* dont split this object */
#define OBJ_UMTXDEAD 0x0020 /* umtx pshared was terminated */
#define OBJ_PIPWNT 0x0040 /* paging in progress wanted */
#define OBJ_MIGHTBEDIRTY 0x0100 /* object might be dirty, only for vnode */
#define OBJ_TMPFS_NODE 0x0200 /* object belongs to tmpfs VREG node */
#define OBJ_TMPFS_DIRTY 0x0400 /* dirty tmpfs obj */
#define OBJ_COLORED 0x1000 /* pg_color is defined */
#define OBJ_ONEMAPPING 0x2000 /* One USE (a single, non-forked) mapping flag */
#define OBJ_DISCONNECTWNT 0x4000 /* disconnect from vnode wanted */
#define OBJ_TMPFS 0x8000 /* has tmpfs vnode allocated */
#define IDX_TO_OFF(idx) (((vm_ooffset_t)(idx)) << PAGE_SHIFT)
#define OFF_TO_IDX(off) ((vm_pindex_t)(((vm_ooffset_t)(off)) >> PAGE_SHIFT))
#ifdef _KERNEL
#define OBJPC_SYNC 0x1 /* sync I/O */
#define OBJPC_INVAL 0x2 /* invalidate */
#define OBJPC_NOSYNC 0x4 /* skip if VPO_NOSYNC */
/*
* The following options are supported by vm_object_page_remove().
*/
#define OBJPR_CLEANONLY 0x1 /* Don't remove dirty pages. */
#define OBJPR_NOTMAPPED 0x2 /* Don't unmap pages. */
TAILQ_HEAD(object_q, vm_object);
extern struct object_q vm_object_list; /* list of allocated objects */
extern struct mtx vm_object_list_mtx; /* lock for object list and count */
extern struct vm_object kernel_object_store;
extern struct vm_object kmem_object_store;
#define kernel_object (&kernel_object_store)
#define kmem_object (&kmem_object_store)
#define VM_OBJECT_ASSERT_LOCKED(object) \
rw_assert(&(object)->lock, RA_LOCKED)
#define VM_OBJECT_ASSERT_RLOCKED(object) \
rw_assert(&(object)->lock, RA_RLOCKED)
#define VM_OBJECT_ASSERT_WLOCKED(object) \
rw_assert(&(object)->lock, RA_WLOCKED)
#define VM_OBJECT_ASSERT_UNLOCKED(object) \
rw_assert(&(object)->lock, RA_UNLOCKED)
#define VM_OBJECT_LOCK_DOWNGRADE(object) \
rw_downgrade(&(object)->lock)
#define VM_OBJECT_RLOCK(object) \
rw_rlock(&(object)->lock)
#define VM_OBJECT_RUNLOCK(object) \
rw_runlock(&(object)->lock)
#define VM_OBJECT_SLEEP(object, wchan, pri, wmesg, timo) \
rw_sleep((wchan), &(object)->lock, (pri), (wmesg), (timo))
#define VM_OBJECT_TRYRLOCK(object) \
rw_try_rlock(&(object)->lock)
#define VM_OBJECT_TRYWLOCK(object) \
rw_try_wlock(&(object)->lock)
#define VM_OBJECT_TRYUPGRADE(object) \
rw_try_upgrade(&(object)->lock)
#define VM_OBJECT_WLOCK(object) \
rw_wlock(&(object)->lock)
#define VM_OBJECT_WOWNED(object) \
rw_wowned(&(object)->lock)
#define VM_OBJECT_WUNLOCK(object) \
rw_wunlock(&(object)->lock)
/*
* The object must be locked or thread private.
*/
static __inline void
vm_object_set_flag(vm_object_t object, u_short bits)
{
object->flags |= bits;
}
/*
* Conditionally set the object's color, which (1) enables the allocation
* of physical memory reservations for anonymous objects and larger-than-
* superpage-sized named objects and (2) determines the first page offset
* within the object at which a reservation may be allocated. In other
* words, the color determines the alignment of the object with respect
* to the largest superpage boundary. When mapping named objects, like
* files or POSIX shared memory objects, the color should be set to zero
* before a virtual address is selected for the mapping. In contrast,
* for anonymous objects, the color may be set after the virtual address
* is selected.
*
* The object must be locked.
*/
static __inline void
vm_object_color(vm_object_t object, u_short color)
{
if ((object->flags & OBJ_COLORED) == 0) {
object->pg_color = color;
object->flags |= OBJ_COLORED;
}
}
void vm_object_clear_flag(vm_object_t object, u_short bits);
void vm_object_pip_add(vm_object_t object, short i);
void vm_object_pip_subtract(vm_object_t object, short i);
void vm_object_pip_wakeup(vm_object_t object);
void vm_object_pip_wakeupn(vm_object_t object, short i);
void vm_object_pip_wait(vm_object_t object, char *waitid);
static __inline boolean_t
vm_object_cache_is_empty(vm_object_t object)
{
return (vm_radix_is_empty(&object->cache));
}
void umtx_shm_object_init(vm_object_t object);
void umtx_shm_object_terminated(vm_object_t object);
extern int umtx_shm_vnobj_persistent;
vm_object_t vm_object_allocate (objtype_t, vm_pindex_t);
boolean_t vm_object_coalesce(vm_object_t, vm_ooffset_t, vm_size_t, vm_size_t,
boolean_t);
void vm_object_collapse (vm_object_t);
void vm_object_deallocate (vm_object_t);
void vm_object_destroy (vm_object_t);
void vm_object_terminate (vm_object_t);
void vm_object_set_writeable_dirty (vm_object_t);
void vm_object_init (void);
void vm_object_madvise(vm_object_t, vm_pindex_t, vm_pindex_t, int);
boolean_t vm_object_page_clean(vm_object_t object, vm_ooffset_t start,
vm_ooffset_t end, int flags);
void vm_object_page_noreuse(vm_object_t object, vm_pindex_t start,
vm_pindex_t end);
void vm_object_page_remove(vm_object_t object, vm_pindex_t start,
vm_pindex_t end, int options);
boolean_t vm_object_populate(vm_object_t, vm_pindex_t, vm_pindex_t);
void vm_object_print(long addr, boolean_t have_addr, long count, char *modif);
void vm_object_reference (vm_object_t);
void vm_object_reference_locked(vm_object_t);
int vm_object_set_memattr(vm_object_t object, vm_memattr_t memattr);
void vm_object_shadow (vm_object_t *, vm_ooffset_t *, vm_size_t);
void vm_object_split(vm_map_entry_t);
boolean_t vm_object_sync(vm_object_t, vm_ooffset_t, vm_size_t, boolean_t,
boolean_t);
void vm_object_unwire(vm_object_t object, vm_ooffset_t offset,
vm_size_t length, uint8_t queue);
struct vnode *vm_object_vnode(vm_object_t object);
#endif /* _KERNEL */
#endif /* _VM_OBJECT_ */
Reference in New Issue View Git Blame Copy Permalink