rw_try_upgrade (9)
Leading comments
Copyright (c) 2006 Gleb Smirnoff <glebius@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. 2. Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the doc...
NAME
rwlock rw_init rw_init_flags, rw_destroy rw_rlock rw_wlock rw_runlock rw_wunlock rw_unlock rw_try_rlock rw_try_upgrade rw_try_wlock rw_downgrade rw_sleep rw_initialized rw_wowned rw_assert RW_SYSINIT - kernel reader/writer lockSYNOPSIS
In sys/param.h In sys/lock.h In sys/rwlock.h Ft void Fn rw_init struct rwlock *rw const char *name Ft void Fn rw_init_flags struct rwlock *rw const char *name int opts Ft void Fn rw_destroy struct rwlock *rw Ft void Fn rw_rlock struct rwlock *rw Ft void Fn rw_wlock struct rwlock *rw Ft int Fn rw_try_rlock struct rwlock *rw Ft int Fn rw_try_wlock struct rwlock *rw Ft void Fn rw_runlock struct rwlock *rw Ft void Fn rw_wunlock struct rwlock *rw Ft void Fn rw_unlock struct rwlock *rw Ft int Fn rw_try_upgrade struct rwlock *rw Ft void Fn rw_downgrade struct rwlock *rw Ft int Fn rw_sleep void *chan struct rwlock *rw int priority const char *wmesg int timo Ft int Fn rw_initialized const struct rwlock *rw Ft int Fn rw_wowned const struct rwlock *rwoptions INVARIANTS options INVARIANT_SUPPORT Ft void Fn rw_assert const struct rwlock *rw int what In sys/kernel.h Fn RW_SYSINIT name struct rwlock *rw const char *desc
DESCRIPTION
Reader/writer locks allow shared access to protected data by multiple threads, or exclusive access by a single thread. The threads with shared access are known as readers since they only read the protected data. A thread with exclusive access is known as a writer since it can modify protected data.Although reader/writer locks look very similar to sx(9) locks, their usage pattern is different. Reader/writer locks can be treated as mutexes (see mutex(9)) with shared/exclusive semantics. Unlike sx(9), an ifconfig can be locked while holding a non-spin mutex, and an ifconfig cannot be held while sleeping. The ifconfig locks have priority propagation like mutexes, but priority can be propagated only to writers. This limitation comes from the fact that readers are anonymous. Another important property is that readers can always recurse, and exclusive locks can be made recursive selectively.
Macros and Functions
- Fn rw_init struct rwlock *rw const char *name
- Initialize structure located at Fa rw as reader/writer lock, described by name Fa name . The description is used solely for debugging purposes. This function must be called before any other operations on the lock.
- Fn rw_init_flags struct rwlock *rw const char *name int opts
-
Initialize the rw lock just like the
Fn rw_init
function, but specifying a set of optional flags to alter the
behaviour of
Fa rw ,
through the
Fa opts
argument.
It contains one or more of the following flags:
- RW_DUPOK
- Witness should not log messages about duplicate locks being acquired.
- RW_NOPROFILE
- Do not profile this lock.
- RW_NOWITNESS
- Instruct witness(4) to ignore this lock.
- RW_QUIET
- Do not log any operations for this lock via ktr(4).
- RW_RECURSE
- Allow threads to recursively acquire exclusive locks for Fa rw .
- Fn rw_rlock struct rwlock *rw
- Lock Fa rw as a reader. If any thread holds this lock exclusively, the current thread blocks, and its priority is propagated to the exclusive holder. The Fn rw_rlock function can be called when the thread has already acquired reader access on Fa rw . This is called ``recursing on a lock''
- Fn rw_wlock struct rwlock *rw
- Lock Fa rw as a writer. If there are any shared owners of the lock, the current thread blocks. The Fn rw_wlock function can be called recursively only if Fa rw has been initialized with the RW_RECURSE option enabled.
- Fn rw_try_rlock struct rwlock *rw
- Try to lock Fa rw as a reader. This function will return true if the operation succeeds, otherwise 0 will be returned.
- Fn rw_try_wlock struct rwlock *rw
- Try to lock Fa rw as a writer. This function will return true if the operation succeeds, otherwise 0 will be returned.
- Fn rw_runlock struct rwlock *rw
- This function releases a shared lock previously acquired by Fn rw_rlock .
- Fn rw_wunlock struct rwlock *rw
- This function releases an exclusive lock previously acquired by Fn rw_wlock .
- Fn rw_unlock struct rwlock *rw
- This function releases a shared lock previously acquired by Fn rw_rlock or an exclusive lock previously acquired by Fn rw_wlock .
- Fn rw_try_upgrade struct rwlock *rw
- Attempt to upgrade a single shared lock to an exclusive lock. The current thread must hold a shared lock of Fa rw . This will only succeed if the current thread holds the only shared lock on Fa rw , and it only holds a single shared lock. If the attempt succeeds Fn rw_try_upgrade will return a non-zero value, and the current thread will hold an exclusive lock. If the attempt fails Fn rw_try_upgrade will return zero, and the current thread will still hold a shared lock.
- Fn rw_downgrade struct rwlock *rw
- Convert an exclusive lock into a single shared lock. The current thread must hold an exclusive lock of Fa rw .
- Fn rw_sleep void *chan struct rwlock *rw int priority const char *wmesg int timo
- Atomically release Fa rw while waiting for an event. For more details on the parameters to this function, see sleep(9).
- Fn rw_initialized const struct rwlock *rw
- This function returns non-zero if Fa rw has been initialized, and zero otherwise.
- Fn rw_destroy struct rwlock *rw
- This functions destroys a lock previously initialized with Fn rw_init . The Fa rw lock must be unlocked.
- Fn rw_wowned const struct rwlock *rw
- This function returns a non-zero value if the current thread owns an exclusive lock on Fa rw .
- Fn rw_assert const struct rwlock *rw int what
-
This function allows assertions specified in
Fa what
to be made about
Fa rw .
If the assertions are not true and the kernel is compiled
with
options INVARIANTS
and
options INVARIANT_SUPPORT
the kernel will panic.
Currently the following base assertions are supported:
- RA_LOCKED
- Assert that current thread holds either a shared or exclusive lock of Fa rw .
- RA_RLOCKED
- Assert that current thread holds a shared lock of Fa rw .
- RA_WLOCKED
- Assert that current thread holds an exclusive lock of Fa rw .
- RA_UNLOCKED
- Assert that current thread holds neither a shared nor exclusive lock of Fa rw .
In addition, one of the following optional flags may be specified with RA_LOCKED RA_RLOCKED or RA_WLOCKED
- RA_RECURSED
- Assert that the current thread holds a recursive lock of Fa rw .
- RA_NOTRECURSED
- Assert that the current thread does not hold a recursive lock of Fa rw .
SEE ALSO
locking(9), mutex(9), panic(9), sema(9), sx(9)HISTORY
These functions appeared in Fx 7.0 .AUTHORS
An -nosplit The ifconfig facility was written by An John Baldwin . This manual page was written by An Gleb Smirnoff .BUGS
If WITNESS is not included in the kernel, then it is impossible to assert that the current thread does or does not hold a read lock. In the non- WITNESS case, the RA_LOCKED and RA_RLOCKED assertions merely check that some thread holds a read lock.Reader/writer is a bit of an awkward name. An ifconfig can also be called a ``Robert Watson'' lock if desired.