gecko-dev/db/man/man.roff/db_internal.3

.ds TYPE C
.\"
.\" See the file LICENSE for redistribution information.
.\"
.\" Copyright (c) 1997, 1998
.\"	Sleepycat Software.  All rights reserved.
.\"
.\"	@(#)db_internal.so	10.19 (Sleepycat) 5/10/98
.\"
.\"
.\" See the file LICENSE for redistribution information.
.\"
.\" Copyright (c) 1997, 1998
.\"	Sleepycat Software.  All rights reserved.
.\"
.\"	@(#)macros.so	10.45 (Sleepycat) 5/4/98
.\"
.\" We don't want hyphenation for any HTML documents.
.ie '\*[HTML]'YES'\{\
.nh
\}
.el\{\
.ds Hy
.hy
..
.ds Nh
.nh
..
\}
.\" The alternative text macro
.\" This macro takes two arguments:
.\"	+ the text produced if this is a "C" manpage
.\"	+ the text produced if this is a "CXX" or "JAVA" manpage
.\"
.de Al
.ie '\*[TYPE]'C'\{\\$1
\}
.el\{\\$2
\}
..
.\" Scoped name macro.
.\" Produces a_b, a::b, a.b depending on language
.\" This macro takes two arguments:
.\"	+ the class or prefix (without underscore)
.\"	+ the name within the class or following the prefix
.de Sc
.ie '\*[TYPE]'C'\{\\$1_\\$2
\}
.el\{\
.ie '\*[TYPE]'CXX'\{\\$1::\\$2
\}
.el\{\\$1.\\$2
\}
\}
..
.\" Scoped name for Java.
.\" Produces Db.b, for Java, otherwise just b.  This macro is used for
.\" constants that must be scoped in Java, but are global otherwise.
.\" This macro takes two arguments:
.\"	+ the class
.\"	+ the name within the class or following the prefix
.de Sj
.ie '\*[TYPE]'JAVA'\{\
.TP 5
Db.\\$1\}
.el\{\
.TP 5
\\$1\}
..
.\" The general information text macro.
.de Gn
.ie '\*[TYPE]'C'\{The DB library is a family of groups of functions that provides a modular
programming interface to transactions and record-oriented file access.
The library includes support for transactions, locking, logging and file
page caching, as well as various indexed access methods.
Many of the functional groups (e.g., the file page caching functions)
are useful independent of the other DB functions,
although some functional groups are explicitly based on other functional
groups (e.g., transactions and logging).
\}
.el\{The DB library is a family of classes that provides a modular
programming interface to transactions and record-oriented file access.
The library includes support for transactions, locking, logging and file
page caching, as well as various indexed access methods.
Many of the classes (e.g., the file page caching class)
are useful independent of the other DB classes,
although some classes are explicitly based on other classes
(e.g., transactions and logging).
\}
For a general description of the DB package, see
.IR db_intro (3).
..
.\" The library error macro, the local error macro.
.\" These macros take one argument:
.\"	+ the function name.
.de Ee
The
.I \\$1
.ie '\*[TYPE]'C'\{function may fail and return
.I errno
\}
.el\{method may fail and throw a
.IR DbException (3)
.if '\*[TYPE]'CXX'\{
or return
.I errno
\}
\}
for any of the errors specified for the following DB and library functions:
..
.de Ec
In addition, the
.I \\$1
.ie '\*[TYPE]'C'\{function may fail and return
.I errno
\}
.el\{method may fail and throw a
.IR DbException (3)
.ie '\*[TYPE]'CXX'\{or return
.I errno
\}
.el\{encapsulating an
.I errno
\}
\}
for the following conditions:
..
.de Ea
[EAGAIN]
A lock was unavailable.
..
.de Eb
[EBUSY]
The shared memory region was in use and the force flag was not set.
..
.de Em
[EAGAIN]
The shared memory region was locked and (repeatedly) unavailable.
..
.de Ei
[EINVAL]
An invalid flag value or parameter was specified.
..
.de Es
[EACCES]
An attempt was made to modify a read-only database.
..
.de Et
The DB_THREAD flag was specified and spinlocks are not implemented for
this architecture.
..
.de Ep
[EPERM]
Database corruption was detected.
All subsequent database calls (other than
.ie '\*[TYPE]'C'\{\
.IR DB->close )
\}
.el\{\
.IR Db::close )
\}
will return EPERM.
..
.de Ek
.if '\*[TYPE]'CXX'\{\
Methods marked as returning
.I errno
will, by default, throw an exception that encapsulates the error information.
The default error behavior can be changed, see
.IR DbException (3).
\}
..
.\" The SEE ALSO text macro
.de Sa
.\" make the line long for nroff.
.if n .ll 72
.nh
.na
.IR db_archive (1),
.IR db_checkpoint (1),
.IR db_deadlock (1),
.IR db_dump (1),
.IR db_load (1),
.IR db_recover (1),
.IR db_stat (1),
.IR db_intro (3),
.ie '\*[TYPE]'C'\{\
.IR db_appinit (3),
.IR db_cursor (3),
.IR db_dbm (3),
.IR db_internal (3),
.IR db_lock (3),
.IR db_log (3),
.IR db_mpool (3),
.IR db_open (3),
.IR db_thread (3),
.IR db_txn (3)
\}
.el\{\
.IR db_internal (3),
.IR db_thread (3),
.IR Db (3),
.IR Dbc (3),
.IR DbEnv (3),
.IR DbException (3),
.IR DbInfo (3),
.IR DbLock (3),
.IR DbLockTab (3),
.IR DbLog (3),
.IR DbLsn (3),
.IR DbMpool (3),
.if !'\*[TYPE]'JAVA'\{\
.IR DbMpoolFile (3),
\}
.IR Dbt (3),
.IR DbTxn (3),
.IR DbTxnMgr (3)
\}
.ad
.Hy
..
.\" The function header macro.
.\" This macro takes one argument:
.\"	+ the function name.
.de Fn
.in 2
.I \\$1
.in
..
.\" The XXX_open function text macro, for merged create/open calls.
.\" This macro takes two arguments:
.\"	+ the interface, e.g., "transaction region"
.\"	+ the prefix, e.g., "txn" (or the class name for C++, e.g., "DbTxn")
.de Co
.ie '\*[TYPE]'C'\{\
.Fn \\$2_open
The
.I \\$2_open
function copies a pointer, to the \\$1 identified by the
.B directory
.IR dir ,
into the memory location referenced by
.IR regionp .
.PP
If the
.I dbenv
argument to
.I \\$2_open
was initialized using
.IR db_appinit ,
.I dir
is interpreted as described by
.IR db_appinit (3).
\}
.el\{\
.Fn \\$2::open
The
.I \\$2::open
.ie '\*[TYPE]'CXX'\{\
method copies a pointer, to the \\$1 identified by the
.B directory
.IR dir ,
into the memory location referenced by
.IR regionp .
\}
.el\{\
method returns a \\$1 identified by the
.B directory
.IR dir .
\}
.PP
If the
.I dbenv
argument to
.I \\$2::open
was initialized using
.IR DbEnv::appinit ,
.I dir
is interpreted as described by
.IR DbEnv (3).
\}
.PP
Otherwise,
if
.I dir
is not NULL,
it is interpreted relative to the current working directory of the process.
If
.I dir
is NULL,
the following environment variables are checked in order:
``TMPDIR'', ``TEMP'', and ``TMP''.
If one of them is set,
\\$1 files are created relative to the directory it specifies.
If none of them are set, the first possible one of the following
directories is used:
.IR /var/tmp ,
.IR /usr/tmp ,
.IR /temp ,
.IR /tmp ,
.I C:/temp
and
.IR C:/tmp .
.PP
All files associated with the \\$1 are created in this directory.
This directory must already exist when
.ie '\*[TYPE]'C'\{
\\$1_open
\}
.el\{\
\\$2::open
\}
is called.
If the \\$1 already exists,
the process must have permission to read and write the existing files.
If the \\$1 does not already exist,
it is optionally created and initialized.
..
.\" The common close language macro, for discarding created regions
.\" This macro takes one argument:
.\"	+ the function prefix, e.g., txn (the class name for C++, e.g., DbTxn)
.de Cc
In addition, if the
.I dir
argument to
.ie '\*[TYPE]'C'\{\
.ds Va db_appinit
.ds Vo \\$1_open
.ds Vu \\$1_unlink
\}
.el\{\
.ds Va DbEnv::appinit
.ds Vo \\$1::open
.ds Vu \\$1::unlink
\}
.I \\*(Vo
was NULL
and
.I dbenv
was not initialized using
.IR \\*(Va ,
.if '\\$1'memp'\{\
or the DB_MPOOL_PRIVATE flag was set,
\}
all files created for this shared region will be removed,
as if
.I \\*(Vu
were called.
.rm Va
.rm Vo
.rm Vu
..
.\" The DB_ENV information macro.
.\" This macro takes two arguments:
.\"	+ the function called to open, e.g., "txn_open"
.\"	+ the function called to close, e.g., "txn_close"
.de En
.ie '\*[TYPE]'C'\{\
based on the
.I dbenv
argument to
.IR \\$1 ,
which is a pointer to a structure of type DB_ENV (typedef'd in <db.h>).
Applications will normally use the same DB_ENV structure (initialized
by
.IR db_appinit (3)),
as an argument to all of the subsystems in the DB package.
.PP
References to the DB_ENV structure are maintained by DB,
so it may not be discarded until the last close function,
corresponding to an open function for which it was an argument,
has returned.
In order to ensure compatibility with future releases of DB, all fields of
the DB_ENV structure that are not explicitly set should be initialized to 0
before the first time the structure is used.
Do this by declaring the structure external or static, or by calling the C
library routine
.IR bzero (3)
or
.IR memset (3).
.PP
The fields of the DB_ENV structure used by
.I \\$1
are described below.
.if '\*[TYPE]'CXX'\{\
As references to the DB_ENV structure may be maintained by
.IR \\$1 ,
it is necessary that the DB_ENV structure and memory it references be valid
until the
.I \\$2
function is called.
\}
.ie '\\$1'db_appinit'\{The
.I dbenv
argument may not be NULL.
If any of the fields of the
.I dbenv
are set to 0,
defaults appropriate for the system are used where possible.
\}
.el\{If
.I dbenv
is NULL
or any of its fields are set to 0,
defaults appropriate for the system are used where possible.
\}
.PP
The following fields in the DB_ENV structure may be initialized before calling
.IR \\$1 :
\}
.el\{\
based on which set methods have been used.
It is expected that applications will use a single DbEnv object as the
argument to all of the subsystems in the DB package.
The fields of the DbEnv object used by
.I \\$1
are described below.
As references to the DbEnv object may be maintained by
.IR \\$1 ,
it is necessary that the DbEnv object and memory it references be valid
until the object is destroyed.
.ie '\\$1'appinit'\{\
The
.I dbenv
argument may not be NULL.
If any of the fields of the
.I dbenv
are set to 0,
defaults appropriate for the system are used where possible.
\}
.el\{\
Any of the DbEnv fields that are not explicitly set will default to
appropriate values.
\}
.PP
The following fields in the DbEnv object may be initialized, using the
appropriate set method, before calling
.IR \\$1 :
\}
..
.\" The DB_ENV common fields macros.
.de Se
.if '\*[TYPE]'JAVA'\{\
.TP 5
DbErrcall db_errcall;
.ns
.TP 5
String db_errpfx;
.ns
.TP 5
int db_verbose;
The error fields of the DbEnv behave as described for
.IR DbEnv (3).
\}
.ie '\*[TYPE]'CXX'\{\
.TP 5
void *(*db_errcall)(char *db_errpfx, char *buffer);
.ns
.TP 5
FILE *db_errfile;
.ns
.TP 5
const char *db_errpfx;
.ns
.TP 5
class ostream *db_error_stream;
.ns
.TP 5
int db_verbose;
The error fields of the DbEnv behave as described for
.IR DbEnv (3).
\}
.el\{\
void *(*db_errcall)(char *db_errpfx, char *buffer);
.ns
.TP 5
FILE *db_errfile;
.ns
.TP 5
const char *db_errpfx;
.ns
.TP 5
int db_verbose;
The error fields of the DB_ENV behave as described for
.IR db_appinit (3).
.sp
\}
..
.\" The open flags.
.de Fm
The
.I flags
and
.I mode
arguments specify how files will be opened and/or created when they
don't already exist.
The flags value is specified by
.BR or 'ing
together one or more of the following values:
.Sj DB_CREATE
Create any underlying files, as necessary.
If the files do not already exist and the DB_CREATE flag is not specified,
the call will fail.
..
.\" DB_THREAD open flag macro.
.\" This macro takes two arguments:
.\"	+ the open function name
.\"	+ the object it returns.
.de Ft
.TP 5
.Sj DB_THREAD
Cause the \\$2 handle returned by the
.I \\$1
.Al function method
to be useable by multiple threads within a single address space,
i.e., to be ``free-threaded''.
.if '\*[TYPE]'JAVA'\{\
Threading is assumed in the Java API,
so no special flags are required,
and DB functions will always behave as if the DB_THREAD flag was specified.
\}
..
.\" The mode macro.
.\" This macro takes one argument:
.\"	+ the subsystem name.
.de Mo
All files created by the \\$1 are created with mode
.I mode
(as described in
.IR chmod (2))
and modified by the process' umask value at the time of creation (see
.IR umask (2)).
The group ownership of created files is based on the system and directory
defaults, and is not further specified by DB.
..
.\" The application exits macro.
.\" This macro takes one argument:
.\"	+ the application name.
.de Ex
The
.I \\$1
utility exits 0 on success, and >0 if an error occurs.
..
.\" The application -h section.
.\" This macro takes one argument:
.\"	+ the application name
.de Dh
DB_HOME
If the
.B \-h
option is not specified and the environment variable
.I DB_HOME
is set, it is used as the path of the database home, as described in
.IR db_appinit (3).
..
.\" The function DB_HOME ENVIRONMENT VARIABLES section.
.\" This macro takes one argument:
.\"	+ the open function name
.de Eh
DB_HOME
If the
.I dbenv
argument to
.I \\$1
was initialized using
.IR db_appinit ,
the environment variable DB_HOME may be used as the path of the database
home for the interpretation of the
.I dir
argument to
.IR \\$1 ,
as described in
.IR db_appinit (3).
.if \\n(.$>1 \{Specifically,
.I \\$1
is affected by the configuration string value of \\$2.\}
..
.\" The function TMPDIR ENVIRONMENT VARIABLES section.
.\" This macro takes two arguments:
.\"	+ the interface, e.g., "transaction region"
.\"	+ the prefix, e.g., "txn" (or the class name for C++, e.g., "DbTxn")
.de Ev
TMPDIR
If the
.I dbenv
argument to
.ie '\*[TYPE]'C'\{\
.ds Vo \\$2_open
\}
.el\{\
.ds Vo \\$2::open
\}
.I \\*(Vo
was NULL or not initialized using
.IR db_appinit ,
the environment variable TMPDIR may be used as the directory in which to
create the \\$1,
as described in the
.I \\*(Vo
section above.
.rm Vo
..
.\" The unused flags macro.
.de Fl
The
.I flags
parameter is currently unused, and must be set to 0.
..
.\" The no-space TP macro.
.de Nt
.br
.ns
.TP 5
..
.\" The return values of the functions macros.
.\" Rc is the standard two-value return with a suffix for more values.
.\" Ro is the standard two-value return but there were previous values.
.\" Rt is the standard two-value return, returning errno, 0, or < 0.
.\" These macros take one argument:
.\"	+ the routine name
.de Rc
The
.I \\$1
.ie '\*[TYPE]'C'\{function returns the value of
.I errno
on failure,
0 on success,
\}
.el\{method throws a
.IR DbException (3)
.ie '\*[TYPE]'CXX'\{or returns the value of
.I errno
on failure,
0 on success,
\}
.el\{that encapsulates an
.I errno
on failure,
\}
\}
..
.de Ro
Otherwise, the
.I \\$1
.ie '\*[TYPE]'C'\{function returns the value of
.I errno
on failure and 0 on success.
\}
.el\{method throws a
.IR DbException (3)
.ie '\*[TYPE]'CXX'\{or returns the value of
.I errno
on failure and 0 on success.
\}
.el\{that encapsulates an
.I errno
on failure,
\}
\}
..
.de Rt
The
.I \\$1
.ie '\*[TYPE]'C'\{function returns the value of
.I errno
on failure and 0 on success.
\}
.el\{method throws a
.IR DbException (3)
.ie '\*[TYPE]'CXX'\{or returns the value of
.I errno
on failure and 0 on success.
\}
.el\{that encapsulates an
.I errno
on failure.
\}
\}
..
.\" The TXN id macro.
.de Tx
.IP
If the file is being accessed under transaction protection,
the
.I txnid
parameter is a transaction ID returned from
.IR txn_begin ,
otherwise, NULL.
..
.\" The XXX_unlink function text macro.
.\" This macro takes two arguments:
.\"	+ the interface, e.g., "transaction region"
.\"	+ the prefix (for C++, this is the class name)
.de Un
.ie '\*[TYPE]'C'\{\
.ds Va db_appinit
.ds Vc \\$2_close
.ds Vo \\$2_open
.ds Vu \\$2_unlink
\}
.el\{\
.ds Va DbEnv::appinit
.ds Vc \\$2::close
.ds Vo \\$2::open
.ds Vu \\$2::unlink
\}
.Fn \\*(Vu
The
.I \\*(Vu
.Al function method
destroys the \\$1 identified by the directory
.IR dir ,
removing all files used to implement the \\$1.
.ie '\\$2'log' \{(The log files themselves and the directory
.I dir
are not removed.)\}
.el \{(The directory
.I dir
is not removed.)\}
If there are processes that have called
.I \\*(Vo
without calling
.I \\*(Vc
(i.e., there are processes currently using the \\$1),
.I \\*(Vu
will fail without further action,
unless the force flag is set,
in which case
.I \\*(Vu
will attempt to remove the \\$1 files regardless of any processes
still using the \\$1.
.PP
The result of attempting to forcibly destroy the region when a process
has the region open is unspecified.
Processes using a shared memory region maintain an open file descriptor
for it.
On UNIX systems, the region removal should succeed
and processes that have already joined the region should continue to
run in the region without change,
however processes attempting to join the \\$1 will either fail or
attempt to create a new region.
On other systems, e.g., WNT, where the
.IR unlink (2)
system call will fail if any process has an open file descriptor
for the file,
the region removal will fail.
.PP
In the case of catastrophic or system failure,
database recovery must be performed (see
.IR db_recover (1)
or the DB_RECOVER and DB_RECOVER_FATAL flags to
.IR \\*(Va (3)).
Alternatively, if recovery is not required because no database state is
maintained across failures,
it is possible to clean up a \\$1 by removing all of the
files in the directory specified to the
.I \\*(Vo
.Al function, method,
as \\$1 files are never created in any directory other than the one
specified to
.IR \\*(Vo .
Note, however,
that this has the potential to remove files created by the other DB
subsystems in this database environment.
.PP
.Rt \\*(Vu
.rm Va
.rm Vo
.rm Vu
.rm Vc
..
.\" Signal paragraph for standard utilities.
.\" This macro takes one argument:
.\"	+ the utility name.
.de Si
The
.I \\$1
utility attaches to DB shared memory regions.
In order to avoid region corruption,
it should always be given the chance to detach and exit gracefully.
To cause
.I \\$1
to clean up after itself and exit,
send it an interrupt signal (SIGINT).
..
.\" Logging paragraph for standard utilities.
.\" This macro takes one argument:
.\"	+ the utility name.
.de Pi
.B \-L
Log the execution of the \\$1 utility to the specified file in the
following format, where ``###'' is the process ID, and the date is
the time the utility starting running.
.sp
\\$1: ### Wed Jun 15 01:23:45 EDT 1995
.sp
This file will be removed if the \\$1 utility exits gracefully.
..
.\" Malloc paragraph.
.\" This macro takes one argument:
.\"	+ the allocated object
.de Ma
.if !'\*[TYPE]'JAVA'\{\
\\$1 are created in allocated memory.
If
.I db_malloc
is non-NULL,
it is called to allocate the memory,
otherwise,
the library function
.IR malloc (3)
is used.
The function
.I db_malloc
must match the calling conventions of the
.IR malloc (3)
library routine.
Regardless,
the caller is responsible for deallocating the returned memory.
To deallocate the returned memory,
free each returned memory pointer;
pointers inside the memory do not need to be individually freed.
\}
..
.\" Underlying function paragraph.
.\" This macro takes two arguments:
.\"	+ the function name
.\"	+ the utility name
.de Uf
The
.I \\$1
.Al function method
is the underlying function used by the
.IR \\$2 (1)
utility.
See the source code for the
.I \\$2
utility for an example of using
.I \\$1
in a UNIX environment.
..
.\" Underlying function paragraph, for C++.
.\" This macro takes three arguments:
.\"	+ the C++ method name
.\"	+ the function name for C
.\"	+ the utility name
.de Ux
The
.I \\$1
method is based on the C
.I \\$2
function, which
is the underlying function used by the
.IR \\$3 (1)
utility.
See the source code for the
.I \\$3
utility for an example of using
.I \\$2
in a UNIX environment.
..
.\" ANSI C function replacement.
.de An
Replace all DB calls to the ANSI C X3.159-1989 (``ANSI C'') standard
.I \\$1
function with
.IR func ,
which must conform to the standard interface.
..
.ds aa "IEEE Std 1003.1b-1993 (``POSIX'')
.\" POSIX 1003.1 function replacement.
.de Po
Replace all DB calls to the \*(aa
.I \\$1
function with
.IR func ,
which must conform to the standard interface.
..
.TH DB_JUMP 3 "March 22, 1998"
.UC 7
.SH NAME
db_jump_set, db_value_set \- replace underlying DB functionality
.SH SYNOPSIS
.nf
.ft B
#include <db.h>

int
db_jump_set(void *func, int which);

int
db_value_set(int value, int which);
.ft R
.fi
.SH DESCRIPTION
.Gn
.PP
This manual page describes interfaces to tune or replace underlying system
functionality used by the DB library.
Few applications should have any need for these interfaces.
.PP
.Fn db_jump_set
The
.I db_jump_set
function enables applications to replace underlying DB library functionality
by replacing entries in a function call jump table.
The
.I which
argument specifies the entry to be replaced by the argument
.IR func .
.PP
The following values of
.I which
are supported:
.TP 5
DB_FUNC_CLOSE
.Po close
.TP 5
DB_FUNC_DIRFREE
The DB library requires the ability to return any memory allocated as part
of the routine which reads through a directory and creates a list of files
that that the directory contains (see DB_FUNC_DIRLIST).
The
.I func
argument must conform to the following interface:
.sp
.ti +5
int dirfree(char **namesp, int cnt);
.sp
The
.I namesp
and
.I cnt
arguments are the same values as were returned by the
DB_FUNC_DIRLIST function.
.sp
.Rt dirfree
.TP 5
DB_FUNC_DIRLIST
The DB library requires the ability to read through a directory and create
a list of files that that the directory contains.
The
.I func
argument must conform to the following interface:
.sp
.nf
.ti +5
int dirlist(const char *dir,
.ti +8
char ***namesp, int *cntp);
.fi
.sp
The
.I dir
argument is the name of the directory to be searched.
The function must return a pointer to an array of nul-terminated file
names in the memory location referenced by the argument
.IR namesp ,
and a count of the number of elements in the array in the memory location
referenced by
.IR cntp .
.sp
.Rt dirlist
.TP 5
DB_FUNC_EXISTS
The DB library requires the ability to determine if a file exists,
and optionally, if it is a file of type directory.
The
.I func
argument must conform to the following interface:
.sp
.ti +5
int exists(const char *path, int *isdirp);
.sp
The
.I path
argument is the pathname of the file to be checked.
.sp
If the
.I isdirp
argument is non-NULL,
it must be set to non-0 if
.I path
is a directory, and 0 if
.I path
is not a directory.
.sp
.Rt exists
.TP 5
DB_FUNC_FREE
.An free
.TP 5
DB_FUNC_FSYNC
.Po fsync
.TP 5
DB_FUNC_IOINFO
The DB library requires the ability to determine the size and I/O
characteristics of a file.
The
.I func
argument must conform to the following interface:
.sp
.nf
.ti +5
int ioinfo(const char *path, int fd,
.ti +8
u_int32_t *mbytesp, u_int32_t *bytesp,
.ti +8
u_int32_t *iosizep);
.fi
.sp
The
.I path
argument is the pathname of the file to be checked, and the
.I fd
argument is an open file descriptor on the file.
.sp
If the
.I mbytesp
and
.I bytesp
arguments are non-NULL, the
.I ioinfo
function must return in them the size of the file: the number of megabytes
in the file into the memory location referenced by the
.I mbytesp
argument,
and the number of bytes over and above that number of megabytes into the
memory location referenced by the
.I bytesp
argument.
.sp
In addition, if the
.I iosizep
argument is non-NULL, the
.I ioinfo
function must return the optimum granularity for I/O operations to the
file in the memory location referenced by it.
.sp
.Rt ioinfo
.TP 5
DB_FUNC_MALLOC
.An malloc
.TP 5
DB_FUNC_MAP
The DB library requires the ability to map a file into memory and to
create shared memory regions (which may or may not be backed by files).
The
.I func
argument must conform to the following interface:
.sp
.nf
.ti +5
int map(char *path, int fd, size_t len,
.ti +8
int is_region, int is_anonymous, int is_rdonly,
.ti +8
void **addr);
.fi
.sp
The
.I path
argument is the name of a file.
The
.I fd
argument is an open file descriptor on that file.
.sp
The
.I is_region
argument will be zero if the intention is to map a file into shared
memory.
In this case, the
.I map
function must map the first
.I len
bytes of the file into memory and return a pointer to the mapped
location in the memory location referenced by the argument
.IR addr .
In this case, the
.I is_anonymous
argument will always be zero.
The
.I is_rdonly
argument will be non-zero if the file is considered read-only by the caller.
.sp
The
.I is_region
argument will be non-zero if the memory is intended to be used as a
shared memory region for synchronization between DB threads/processes.
In this case, the returned memory may be of any kind (e.g., anonymous),
but must be able to support semaphores.
If the application has previously specified that regions are to be
instantiated in anonymous memory (see DB_REGION_ANON below),
or the region is being joined and is believed to have been allocated
in anonymous shared memory, the
.I is_anonymous
argument will be non-zero.
Additionally, the
.I path
and
.I fd
arguments may be ignored (although future
.I map
calls using the same
.I path
must return the same memory),
and the
.I is_rdonly
argument will always be zero.
.sp
By default, on UNIX systems, the DB library will use the \*(aa
.IR mmap (2)
interface to both map regular files into shared memory and create
shared memory regions.
If the application specifies that shared memory regions be instantiated
in anonymous memory (see DB_REGION_ANON below), the
.IR shmget (2)
shared memory segment interface will be used, where available, and
the MAP_ANON or MAP_ANONYMOUS options to
.IR mmap (2)
when
.IR shmget (2)
is not available.
.sp
When using
.IR shmget (2),
shared memory regions are named, and so multiple processes may share them.
When using the
.I mmap
MAP_ANON or MAP_ANONYMOUS options, shared memory regions are
.B not
named, and so may only be accessed by a single process and its threads.
.sp
.RS
.TP 5
HP/UX note:
The
.IR shmget (2)
interfaces are not used on HP/UX, even though they exist, as anonymous
memory allocated using
.IR shmget (2)
cannot be used to store semaphores.
.RE
.sp
.Rt map
.TP 5
DB_FUNC_OPEN
.Po open
.TP 5
DB_FUNC_READ
.Po read
.TP 5
DB_FUNC_REALLOC
.An realloc
.TP 5
DB_FUNC_RUNLINK
The DB library requires the ability to remove shared memory regions from
the system, whether or not they are backed by regular files.
The
.I func
argument must conform to the following interface:
.sp
.nf
.ti +5
int runlink(char *path);
.fi
.sp
The
.I path
argument is the path argument specified to the DB_FUNC_MAP function when
the region was mapped into memory.
.sp
.Rt runlink
.TP 5
DB_FUNC_SEEK
The DB library requires the ability to specify that a subsequent read from
or write to a file will occur at a specific location in that file.
The
.I func
argument must conform to the following interface:
.sp
.nf
.ti +5
int seek(int fd, size_t pgsize, db_pgno_t pageno,
.ti +8
u_int32_t relative, int rewind, int whence);
.fi
.sp
The
.I fd
argument is an open file descriptor on the file.
The
.I seek
function must cause a subsequent read from or write to the file to occur
at a byte offset specified by the calculation:
.sp
.ti +5
(pgsize * pageno) + relative
.sp
If
.I rewind
is non-zero, the byte offset is treated as a backwards seek, not a forwards one.
.sp
The
.I whence
argument specifies where in the file the byte offset is relative to,
as described by the \*(aa
.I lseek
system call.
.sp
.Rt seek
.TP 5
DB_FUNC_SLEEP
The DB library requires the ability to cause a process to suspend itself
for a period of time, relinquishing control of the processor to any other
waiting thread of control.
The
.I func
argument must conform to the following interface:
.sp
.ti +5
int sleep(u_long seconds, u_long microseconds);
.sp
The
.I seconds
and
.I microseconds
arguments specify the amount of time to wait until the suspending thread
of control should run again.
.sp
The
.I seconds
and
.I microseconds
arguments may not be normalized when the
.I sleep
function is called, i.e., the
.I microseconds
argument may be greater than 1000000.
.sp
.Rt sleep
.TP 5
DB_FUNC_UNLINK
.Po unlink
.TP 5
DB_FUNC_UNMAP
The DB library requires the ability to unmap a file or shared memory region
from memory.
The
.I func
argument must conform to the following interface:
.sp
.ti +5
int unmap(void *addr, size_t len);
.sp
The
.I addr
argument is the argument returned by the DB_FUNC_MAP function when the
file or region was mapped into memory, and the
.I len
argument is the same as the
.I len
argument specified to the DB_FUNC_MAP function when the file or region
was mapped into memory.
.sp
.Rt unmap
.TP 5
DB_FUNC_WRITE
.Po write
.TP 5
DB_FUNC_YIELD
The DB library requires the ability to yield the processor from the current
thread of control to any other waiting threads of control.
The
.I func
argument must conform to the following interface:
.sp
.ti +5
int yield(void);
.sp
.sp
The
.I yield
function must be able to cause the rescheduling all participants in the
current DB environment, whether threaded or not.
It may be incorrect to supply a thread
.I yield
function if more than a single process is operating in the DB environment.
This is because many thread-yield functions will not allow other processes
to run,
and the contested lock may be held by another process, not by another thread.
.sp
If no
.I yield
function is specified, or if the
.I yield
function returns an error,
the function specified by the DB_FUNC_SLEEP entry will be used instead or
subsequently, i.e., if no
.I yield
function is specified,
or it is possible for the
.I yield
function to fail, the
.I sleep
function
.B must
cause the processor to reschedule any waiting threads of control for execution.
.\".sp
.\".ft B
.\"Solaris architecture note:
.\".ft R
.\"Because of bugs in versions of Solaris before version 5.6,
.\"the DB library uses the
.\".IR sema_wait (3T)
.\"call instead of the
.\".IR sema_trywait (3T)
.\"call.
.\"For this reason, replacing the
.\".I yield
.\"function will have no effect on Solaris.
.sp
.Rt yield
.PP
.Rt db_jump_set
.PP
.Fn db_value_set
The
.I db_value_set
function enables applications to specify underlying DB library functionality.
The
.I which
argument specifies the information being set by the argument
.IR value .
.PP
The following values of
.I which
are supported:
.TP 5
DB_MUTEXLOCKS
Grant all requested mutual exclusion mutexes without testing for their
availability.
This flag should never be used for any other purpose than debugging.
.TP 5
DB_REGION_ANON
Setting
.I value
to a non-zero value specifies that shared memory regions are to be created
in anonymous memory, and not backed by a regular file.
DB_REGION_NAME differs from DB_REGION_ANON in that the former will fail if
the shared memory regions cannot be named, that is, if multiple processes
cannot use them.
See DB_FUNC_MAP for more information.
.TP 5
DB_REGION_INIT
In some applications, the expense of page-faulting the shared memory regions
can affect performance, e.g., when the page-fault occurs while holding a lock,
other lock requests can convoy and overall throughput may decrease.
Setting
.I value
to a non-zero value specifies that one byte be read from each 4K page of the
shared memory region when the region is initialized.
.TP 5
DB_REGION_NAME
Setting
.I value
to a non-zero value specifies that shared memory regions are to be created
in anonymous memory, and not backed by a regular file.
DB_REGION_NAME differs from DB_REGION_ANON in that the former will fail if
the shared memory regions cannot be named, that is, if multiple processes
cannot use them.
See DB_FUNC_MAP for more information.
.TP 5
DB_TSL_SPINS
Specify the number of times mutexes should spin without blocking.
.sp
This value defaults to 1 on uniprocessor systems and to 50 times the number
of processors on multiprocessor systems.
.PP
.Rt db_value_set
.SH ERRORS
The
.I db_jump_set
function may fail and return
.I errno
for the following conditions:
.TP 5
.Ei
.SH BUGS
No type checking is done of the
.I func
argument,
and specifying an invalid replacement routine will cause unpredictable
results.
.PP
Applications should be careful to replace related functions as a group
and at the same time.
Replacing DB_FUNC_MALLOC without replacing DB_FUNC_REALLOC is likely
to result in unpredictable results.
.PP
On Windows/95, files that are opened by multiple processes do not share
data correctly.  To tell Berkeley DB to use a named region of the paging
file to share memory instead, use:
.PP
.RS
db_value_set(1, DB_REGION_NAME);
.RE
.PP
You do not need to do this if your application can guarantee that
only one process will be accessing DB files.
.PP
On Windows/NT, sharing of data between processes through the paging file
does not work correctly, so you should not call
.IR db_value_set .
That will allow DB to use the file itself for sharing, which works
correctly.
.SH "SEE ALSO"
.Sa