NAME
kextd — loads kexts on demand from kernel or client processes
SYNOPSIS
kextd [-c] [-d] [-f] [-h] [-j] [-r directory] … [-v [1-6]] [-x]
DESCRIPTION
kextd runs as a daemon process to handle requests from the kernel to load
kernel extensions (kexts). For proper operation kextd must be invoked as
the superuser.
kextd should not be considered a formal interface in the Darwin OS or in
Mac OS X. Neither its usage nor its presence should be relied on by any
software or software installer.
The options available are these:
-c Ignore any repository cache files and scan all kext bundles to
gather information. If this option is not given, kextd will
attempt to use cache files and to create them if they are out of
date or don’t exist.
-d Debug mode; run in the foreground without forking as a daemon
process.
-f Don’t fork a child task when loading kexts. This option is
present for debugging kextd and should never be used otherwise.
-h Print a help message describing each option flag.
-j Don’t jettison kernel linker; load NDRVs in the kernel and exit,
allowing the kernel to continue handling all load requests.
This option is used in startup scripts for install CDs, along
with a properly built mkext cache, in order to speed startup
from the CD.
-r directory
Use directory as a repository of known kexts for handling load
requests. This is not recursive; only the directory’s immediate
contents are scanned. By default kextd uses only the kexts in
/System/Library/Extensions.
-v [1-6]
Verbose mode; print information about the kext scanning and
loading process (note that this is generally not useful when
unloading). See the man page for kextload(8) to see what each
verbose level prints.
-x Run kextd in safe boot mode (indicating startup with the Shift
key held down). Kexts that don’t specify a proper value for the
OSBundleRequired info dictionary property will not be loaded.
This option implies the use of the -c option.
DIAGNOSTICS
kextd normally never exits. If an error occurs it exits with a nonzero
status. If invoked with the -j option it exits with a zero status when
no error occurs, or a nonzero status if an error does occur.
kextd logs all error and verbose messages to the system log, or to the
console if the system log facility isn’t available. When running in
debug mode all output is printed to the standard output and error
streams.
Next One 
NAME
rpc.lockd — NFS file locking daemon
SYNOPSIS
rpc.lockd [-d debug_level] [-g grace period] [-x statd cache period]
DESCRIPTION
The rpc.lockd utility provides monitored and unmonitored file and record
locking services in an NFS environment. To monitor the status of hosts
requesting locks, the locking daemon typically operates in conjunction
with rpc.statd(8).
Options and operands available for rpc.lockd:
-d The -d option causes debugging information to be written to sys-
log, recording all RPC transactions to the daemon. These mes-
sages are logged with level LOG_DEBUG and facility LOG_DAEMON.
Specifying a debug_level of 1 results in the generation of one
log line per protocol operation. Higher debug levels can be
specified, causing display of operation arguments and internal
operations of the daemon.
-g The -g option allow to specify the grace period, in seconds.
During the grace period rpc.lockd only accepts requests from
hosts which are reinitialising locks which existed before the
server restart. Default is 30 seconds.
-w The -w option tells rpc.lockd to wait until the first client
locking request is made before starting the locking daemon(s).
This may be used on NFS clients to defer starting the NFS locking
daemons until it is known that they will be needed. (Note:
rpc.statd(8) will also be started if it isn’t already running)
-x The -x option tells rpc.lockd how long to cache state records for
contacting client rpc.statd(8) implementations. Setting it to
zero will disable the cache which will make lock and unlock
requests from a single client more expensive because of addi-
tional interaction with the client’s statd. The default cache
time is 60 seconds.
Error conditions are logged to syslog, irrespective of the debug level,
using log level LOG_ERR and facility LOG_DAEMON.
The rpc.lockd utility must NOT be invoked by inetd(8) because the proto-
col assumes that the daemon will run from system start time. Instead, it
should be configured in rc.conf(5) to run at system startup.
FILES
/usr/include/rpcsvc/nlm_prot.x RPC protocol specification for the network lock manager protocol.
Next One