## ====================================================================
## Copyright (c) 1999-2003 Ralf S. Engelschall <rse@engelschall.com>
## Copyright (c) 1999-2003 The OSSP Project <http://www.ossp.org/>
##
## 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.
##
## 3. All advertising materials mentioning features or use of this
##    software must display the following acknowledgment:
##    "This product includes software developed by
##     Ralf S. Engelschall <rse@engelschall.com>."
##
## 4. Redistributions of any form whatsoever must retain the following
##    acknowledgment:
##    "This product includes software developed by
##     Ralf S. Engelschall <rse@engelschall.com>."
##
## THIS SOFTWARE IS PROVIDED BY RALF S. ENGELSCHALL ``AS IS'' AND ANY
## EXPRESSED 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 RALF S. ENGELSCHALL OR
## ITS 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.
## ====================================================================

##
## mm.pod -- Manpage
##

=pod

=head1 NAME

B<OSSP mm> - B<Shared Memory Allocation>

=head1 VERSION

OSSP mm MM_VERSION_STR

=head1 SYNOPSIS

 #include "mm.h"

B< Global Malloc-Replacement API>

 int     MM_create(size_t size, const char *file);
 int     MM_permission(mode_t mode, uid_t owner, gid_t group);
 void    MM_destroy(void);
 int     MM_lock(mm_lock_mode mode);
 int     MM_unlock(void);
 void   *MM_malloc(size_t size);
 void   *MM_realloc(void *ptr, size_t size);
 void    MM_free(void *ptr);
 void   *MM_calloc(size_t number, size_t size);
 char   *MM_strdup(const char *str);
 size_t  MM_sizeof(void *ptr);
 size_t  MM_maxsize(void);
 size_t  MM_available(void);
 char   *MM_error(void);

B< Standard Malloc-Style API>

 MM     *mm_create(size_t size, char *file);
 int     mm_permission(MM *mm, mode_t mode, uid_t owner, gid_t group);
 void    mm_destroy(MM *mm);
 int     mm_lock(MM *mm, mm_lock_mode mode);
 int     mm_unlock(MM *mm);
 void   *mm_malloc(MM *mm, size_t size);
 void   *mm_realloc(MM *mm, void *ptr, size_t size);
 void    mm_free(MM *mm, void *ptr);
 void   *mm_calloc(MM *mm, size_t number, size_t size);
 char   *mm_strdup(MM *mm, const char *str);
 size_t  mm_sizeof(void *ptr);
 size_t  mm_maxsize(void);
 size_t  mm_available(MM *mm);
 char   *mm_error(void);
 void    mm_display_info(MM *mm);

B< Low-level Shared Memory API>

 void   *mm_core_create(size_t size, char *file);
 int     mm_core_permission(void *core, mode_t mode, uid_t owner, gid_t group);
 void    mm_core_delete(void *core);
 int     mm_core_lock(void *core, mm_lock_mode mode);
 int     mm_core_unlock(void *core);
 size_t  mm_core_size(void *core);
 size_t  mm_core_maxsegsize(void);
 size_t  mm_core_align2page(size_t size);
 size_t  mm_core_align2click(size_t size);

B< Internal Library API>

 void    mm_lib_error_set(unsigned int, const char *str);
 char   *mm_lib_error_get(void);
 int     mm_lib_version(void);

=head1 DESCRIPTION

The B<OSSP mm> library is a 2-layer abstraction library which simplifies the usage
of shared memory between forked (and this way strongly related) processes
under Unix platforms. On the first (lower) layer it hides all platform
dependent implementation details (allocation and locking) when dealing with
shared memory segments and on the second (higher) layer it provides a
high-level malloc(3)-style API for a convenient and well known way to work
with data-structures inside those shared memory segments.

The abbreviation B<OSSP mm> is historically and originally comes from the phrase
``I<memory mapped>'' as used by the POSIX.1 mmap(2) function. Because this
facility is internally used by this library on most platforms to establish the
shared memory segments.

=head2 LIBRARY STRUCTURE

This library is structured into three main APIs which are internally based on
each other:

=over 4

=item B<Global Malloc-Replacement API>

This is the most high-level API which directly can be used as replacement API
for the POSIX.1 memory allocation API (malloc(2) and friends). This is
useful when converting I<heap> based data structures to I<shared memory>
based data structures without the need to change the code dramatically.  All
which is needed is to prefix the POSIX.1 memory allocation functions with
`C<MM_>', i.e. `C<malloc>' becomes `C<MM_malloc>', `C<strdup>' becomes
`C<MM_strdup>', etc. This API internally uses just a global `C<MM *>' pool for
calling the corresponding functions (those with prefix `C<mm_>') of the
I<Standard Malloc-Style API>.

=item B<Standard Malloc-Style API>

This is the standard high-level memory allocation API. Its interface is
similar to the I<Global Malloc-Replacement API> but it uses an explicit `C<MM *>'
pool to operate on. That is why every function of this API has an argument of
type `C<MM *>' as its first argument. This API provides a comfortable way to
work with small dynamically allocated shared memory chunks inside large
statically allocated shared memory segments. It is internally based on the
I<Low-Level Shared Memory API> for creating the underlaying shared memory
segment.

=item B<Low-Level Shared Memory API>

This is the basis of the whole B<OSSP mm> library. It provides low-level functions
for creating shared memory segments with mutual exclusion (in short I<mutex>)
capabilities in a portable way. Internally the shared memory and mutex
facility is implemented in various platform-dependent ways. A list of
implementation variants follows under the next topic.

=back

=head2 SHARED MEMORY IMPLEMENTATION

Internally the shared memory facility is implemented in various
platform-dependent ways. Each way has its own advantages and disadvantages
(in addition to the fact that some variants aren't available at all on some
platforms). The B<OSSP mm> library's configuration procedure tries hard to make a
good decision. The implemented variants are now given for overview and
background reasons with their advantages and disadvantages and in an ascending
order, i.e. the B<OSSP mm> configuration mechanism chooses the last available one
in the list as the preferred variant.

=over 4

=item Classical mmap(2) on temporary file (MMFILE)

I<Advantage:> maximum portable.
I<Disadvantage:> needs a temporary file on the filesystem.

=item mmap(2) via POSIX.1 shm_open(3) on temporary file (MMPOSX)

I<Advantage:> standardized by POSIX.1 and theoretically portable.
I<Disadvantage:> needs a temporary file on the filesystem and is
is usually not available on existing Unix platform.

=item SVR4-style mmap(2) on C</dev/zero> device (MMZERO)

I<Advantage:> widely available and mostly portable on SVR4 platforms.
I<Disadvantage:> needs the C</dev/zero> device and a mmap(2)
which supports memory mapping through this device.

=item SysV IPC shmget(2) (IPCSHM)

I<Advantage:> does not need a temporary file or external device.
I<Disadvantage:> although available on mostly all modern Unix platforms, it has
strong restrictions like the maximum size of a single shared memory segment (can
be as small as 100KB, but depends on the platform).

=item 4.4BSD-style mmap(2) via C<MAP_ANON> facility (MMANON)

I<Advantage:> does not need a temporary file or external device.
I<Disadvantage:> usually only available on BSD platforms and derivatives.

=back

=head2 LOCKING IMPLEMENTATION

As for the shared memory facility, internally the locking facility is
implemented in various platform-dependent ways. They are again listed
in ascending order, i.e. the B<OSSP mm> configuration mechanism chooses the
last available one in the list as the preferred variant. The list of
implemented variants is:

=over 4

=item 4.2BSD-style flock(2) on temporary file (FLOCK)

I<Advantage:> exists on a lot of platforms, especially on older Unix
derivates.  I<Disadvantage:> needs a temporary file on the filesystem and has
to re-open file-descriptors to it in each(!) fork(2)'ed child process.

=item SysV IPC semget(2) (IPCSEM)

I<Advantage:> exists on a lot of platforms and does not need a temporary file.
I<Disadvantage:> an unmeant termination of the application leads to a
semaphore leak because the facility does not allow a ``remove in advance''
trick (as the IPC shared memory facility does) for safe cleanups.

=item SVR4-style fcntl(2) on temporary file (FCNTL)

I<Advantage:> exists on a lot of platforms and is also the most powerful
variant (although not always the fastest one). I<Disadvantage:> needs a
temporary file.

=back

=head2 MEMORY ALLOCATION STRATEGY

The memory allocation strategy the I<Standard Malloc-Style API> functions use
internally is the following:

=over 4

=item B<Allocation>

If a chunk of memory has to be allocated, the internal list of free chunks
is searched for a minimal-size chunk which is larger or equal than the size of
the to be allocated chunk (a I<best fit> strategy).

If a chunk is found which matches this best-fit criteria, but is still a lot
larger than the requested size, it is split into two chunks: One with exactly
the requested size (which is the resulting chunk given back) and one with the
remaining size (which is immediately re-inserted into the list of free
chunks).

If no fitting chunk is found at all in the list of free chunks, a new one is
created from the spare area of the shared memory segment until the segment is
full (in which case an I<out of memory> error occurs).

=item B<Deallocation>

If a chunk of memory has to be deallocated, it is inserted in sorted manner
into the internal list of free chunks. The insertion operation automatically
merges the chunk with a previous and/or a next free chunk if possible, i.e.
if the free chunks stay physically seamless (one after another) in memory, to
automatically form larger free chunks out of smaller ones.

This way the shared memory segment is automatically defragmented when memory
is deallocated.

=back

This strategy reduces memory waste and fragmentation caused by small and
frequent allocations and deallocations to a minimum.

The internal implementation of the list of free chunks is not specially
optimized (for instance by using binary search trees or even I<splay> trees,
etc), because it is assumed that the total amount of entries in the list of
free chunks is always small (caused both by the fact that shared memory
segments are usually a lot smaller than heaps and the fact that we always
defragment by merging the free chunks if possible).

=head1 API FUNCTIONS

In the following, all API functions are described in detail The order
directly follows the one in the B<SYNOPSIS> section above.

=head2 Global Malloc-Replacement API

=over 4

=item int B<MM_create>(size_t I<size>, const char *I<file>);

This initializes the global shared memory pool with I<size> and I<file> and
has to be called I<before> any fork(2) operations are performed by the
application.

=item int B<MM_permission>(mode_t I<mode>, uid_t I<owner>, gid_t I<group>);

This sets the filesystem I<mode>, I<owner> and I<group> for the global shared
memory pool (has effects only if the underlaying shared memory segment
implementation is actually based on external auxiliary files).  The arguments
are directly passed through to chmod(2) and chown(2).

=item void B<MM_destroy>(void);

This destroys the global shared memory pool and should be called I<after> all
child processes were killed.

=item int B<MM_lock>(mm_lock_mode I<mode>);

This locks the global shared memory pool for the current process in order to
perform either shared/read-only (I<mode> is C<MM_LOCK_RD>) or
exclusive/read-write (I<mode> is C<MM_LOCK_RW>) critical operations inside the
global shared memory pool.

=item int B<MM_unlock>(void);

This unlocks the global shared memory pool for the current process after the
critical operations were performed inside the global shared memory pool.

=item void *B<MM_malloc>(size_t I<size>);

Identical to the POSIX.1 malloc(3) function but instead of allocating
memory from the I<heap> it allocates it from the global shared memory pool.

=item void B<MM_free>(void *I<ptr>);

Identical to the POSIX.1 free(3) function but instead of deallocating
memory in the I<heap> it deallocates it in the global shared memory pool.

=item void *B<MM_realloc>(void *I<ptr>, size_t I<size>);

Identical to the POSIX.1 realloc(3) function but instead of reallocating
memory in the I<heap> it reallocates it inside the global shared memory pool.

=item void *B<MM_calloc>(size_t I<number>, size_t I<size>);

Identical to the POSIX.1 calloc(3) function but instead of allocating and
initializing memory from the I<heap> it allocates and initializes it from the
global shared memory pool.

=item char *B<MM_strdup>(const char *I<str>);

Identical to the POSIX.1 strdup(3) function but instead of creating the
string copy in the I<heap> it creates it in the global shared memory pool.

=item size_t B<MM_sizeof>(const void *I<ptr>);

This function returns the size in bytes of the chunk starting at I<ptr> when
I<ptr> was previously allocated with MM_malloc(3). The result is undefined
if I<ptr> was not previously allocated with MM_malloc(3).

=item size_t B<MM_maxsize>(void);

This function returns the maximum size which is allowed
as the first argument to the MM_create(3) function.

=item size_t B<MM_available>(void);

Returns the amount in bytes of still available (free) memory in the global
shared memory pool.

=item char *B<MM_error>(void);

Returns the last error message which occurred inside the B<OSSP mm> library.

=back

=head2 Standard Malloc-Style API

=over 4

=item MM *B<mm_create>(size_t I<size>, const char *I<file>);

This creates a shared memory pool which has space for approximately a total of
I<size> bytes with the help of I<file>. Here I<file> is a filesystem path to a
file which need not to exist (and perhaps is never created because this
depends on the platform and chosen shared memory and mutex implementation).
The return value is a pointer to a C<MM> structure which should be treated as
opaque by the application. It describes the internals of the created shared
memory pool. In case of an error C<NULL> is returned.  A I<size> of 0 means to
allocate the maximum allowed size which is platform dependent and is between a
few KB and the soft limit of 64MB.

=item int B<mm_permission>(MM *I<mm>, mode_t I<mode>, uid_t I<owner>, gid_t I<group>);

This sets the filesystem I<mode>, I<owner> and I<group> for the shared memory
pool I<mm> (has effects only when the underlaying shared memory segment
implementation is actually based on external auxiliary files).  The arguments
are directly passed through to chmod(2) and chown(2).

=item void B<mm_destroy>(MM *I<mm>);

This destroys the complete shared memory pool I<mm> and with it all chunks
which were allocated in this pool. Additionally any created files on the
filesystem corresponding the to shared memory pool are unlinked.

=item int B<mm_lock>(MM *I<mm>, mm_lock_mode I<mode>);

This locks the shared memory pool I<mm> for the current process in order to
perform either shared/read-only (I<mode> is C<MM_LOCK_RD>) or
exclusive/read-write (I<mode> is C<MM_LOCK_RW>) critical operations inside the
global shared memory pool.

=item int B<mm_unlock>(MM *I<mm>);

This unlocks the shared memory pool I<mm> for the current process after
critical operations were performed inside the global shared memory pool.

=item void *B<mm_malloc>(MM *I<mm>, size_t I<size>);

This function allocates I<size> bytes from the shared memory pool I<mm> and
returns either a (virtual memory word aligned) pointer to it or C<NULL> in
case of an error (out of memory). It behaves like the POSIX.1 malloc(3)
function but instead of allocating memory from the I<heap> it allocates it
from the shared memory segment underlaying I<mm>.

=item void B<mm_free>(MM *I<mm>, void *I<ptr>);

This deallocates the chunk starting at I<ptr> in the shared memory pool I<mm>.
It behaves like the POSIX.1 free(3) function but instead of deallocating
memory from the I<heap> it deallocates it from the shared memory segment
underlaying I<mm>.

=item void *B<mm_realloc>(MM *I<mm>, void *I<ptr>, size_t I<size>);

This function reallocates the chunk starting at I<ptr> inside the shared
memory pool I<mm> with the new size of I<size> bytes.  It behaves like the
POSIX.1 realloc(3) function but instead of reallocating memory in the
I<heap> it reallocates it in the shared memory segment underlaying I<mm>.

=item void *B<mm_calloc>(MM *I<mm>, size_t I<number>, size_t I<size>);

This is similar to mm_malloc(3), but additionally clears the chunk. It behaves
like the POSIX.1 calloc(3) function.  It allocates space for I<number>
objects, each I<size> bytes in length from the shared memory pool I<mm>.  The
result is identical to calling mm_malloc(3) with an argument of ``I<number> *
I<size>'', with the exception that the allocated memory is initialized to nul
bytes.

=item char *B<mm_strdup>(MM *I<mm>, const char *I<str>);

This function behaves like the POSIX.1 strdup(3) function.  It allocates
sufficient memory inside the shared memory pool I<mm> for a copy of the string
I<str>, does the copy, and returns a pointer to it.  The pointer may
subsequently be used as an argument to the function mm_free(3). If
insufficient shared memory is available, C<NULL> is returned.

=item size_t B<mm_sizeof>(const void *I<ptr>);

This function returns the size in bytes of the chunk starting at I<ptr> when
I<ptr> was previously allocated with mm_malloc(3). The result is undefined
when I<ptr> was not previously allocated with mm_malloc(3).

=item size_t B<mm_maxsize>(void);

This function returns the maximum size which is allowed as the first argument
to the mm_create(3) function.

=item size_t B<mm_available>(MM *I<mm>);

Returns the amount in bytes of still available (free) memory in the
shared memory pool I<mm>.

=item char *B<mm_error>(void);

Returns the last error message which occurred inside the B<OSSP mm> library.

=item void B<mm_display_info>(MM *I<mm>);

This is debugging function which displays a summary page for the shared memory
pool I<mm> describing various internal sizes and counters.

=back

=head2 Low-Level Shared Memory API

=over 4

=item void *B<mm_core_create>(size_t I<size>, const char *I<file>);

This creates a shared memory area which is at least I<size> bytes in size with
the help of I<file>. The value I<size> has to be greater than 0 and less or
equal the value returned by mm_core_maxsegsize(3). Here I<file> is a
filesystem path to a file which need not to exist (and perhaps is never
created because this depends on the platform and chosen shared memory and
mutex implementation).  The return value is either a (virtual memory word
aligned) pointer to the shared memory segment or C<NULL> in case of an error.
The application is guaranteed to be able to access the shared memory segment
from byte 0 to byte I<size>-1 starting at the returned address.

=item int B<mm_core_permission>(void *I<core>, mode_t I<mode>, uid_t I<owner>, gid_t I<group>);

This sets the filesystem I<mode>, I<owner> and I<group> for the shared memory
segment I<code> (has effects only when the underlaying shared memory segment
implementation is actually based on external auxiliary files).  The arguments
are directly passed through to chmod(2) and chown(2).

=item void B<mm_core_delete>(void *I<core>);

This deletes a shared memory segment I<core> (as previously returned by a
mm_core_create(3) call). After this operation, accessing the segment starting
at I<core> is no longer allowed and will usually lead to a segmentation fault.

=item int B<mm_core_lock>(const void *I<core>, mm_lock_mode I<mode>);

This function acquires an advisory lock for the current process on the shared
memory segment I<core> for either shared/read-only (I<mode> is C<MM_LOCK_RD>)
or exclusive/read-write (I<mode> is C<MM_LOCK_RW>) critical operations between
fork(2)'ed child processes.

=item int B<mm_core_unlock>(const void *I<core>);

This function releases a previously acquired advisory lock for the current
process on the shared memory segment I<core>.

=item size_t B<mm_core_size>(const void *I<core>);

This returns the size in bytes of I<core>. This size is exactly the size which
was used for creating the shared memory area via mm_core_create(3). The
function is provided just for convenience reasons to not require the
application to remember the memory size behind I<core> itself.

=item size_t B<mm_core_maxsegsize>(void);

This returns the number of bytes of a maximum-size shared memory segment which
is allowed to allocate via the MM library. It is between a few KB and the soft
limit of 64MB.

=item size_t B<mm_core_align2page>(size_t I<size>);

This is just a utility function which can be used to align the number I<size>
to the next virtual memory I<page> boundary used by the underlaying platform.
The memory page boundary under Unix platforms is usually somewhere between
2048 and 16384 bytes. You do not have to align the I<size> arguments of other
B<OSSP mm> library functions yourself, because this is already done internally.
This function is exported by the B<OSSP mm> library just for convenience reasons in
case an application wants to perform similar calculations for other purposes.

=item size_t B<mm_core_align2word>(size_t I<size>);

This is another utility function which can be used to align the number I<size>
to the next virtual memory I<word> boundary used by the underlaying platform.
The memory word boundary under Unix platforms is usually somewhere between 4
and 16 bytes.  You do not have to align the I<size> arguments of other B<OSSP mm>
library functions yourself, because this is already done internally.  This
function is exported by the B<OSSP mm> library just for convenience reasons in case
an application wants to perform simular calculations for other purposes.

=back

=head2 Low-Level Shared Memory API

=over 4

=item void B<mm_lib_error_set>(unsigned int, const char *str);

This is a function which is used internally by the various MM function to set
an error string. It's usually not called directly from applications.

=item char *B<mm_lib_error_get>(void);

This is a function which is used internally by MM_error(3) and mm_error(3)
functions to get the current error string. It is usually not called directly
from applications.

=item int B<mm_lib_version>(void);

This function returns a hex-value ``0xI<V>I<RR>I<T>I<LL>'' which describes the
current B<OSSP mm> library version. I<V> is the version, I<RR> the revisions, I<LL>
the level and I<T> the type of the level (alphalevel=0, betalevel=1,
patchlevel=2, etc). For instance B<OSSP mm> version 1.0.4 is encoded as 0x100204.
The reason for this unusual mapping is that this way the version number is
steadily I<increasing>.

=back

=head1 RESTRICTIONS

The maximum size of a continuous shared memory segment one can allocate
depends on the underlaying platform. This cannot be changed, of course.  But
currently the high-level malloc(3)-style API just uses a single shared memory
segment as the underlaying data structure for an C<MM> object which means that
the maximum amount of memory an C<MM> object represents also depends on the
platform.

This could be changed in later versions by allowing at least the
high-level malloc(3)-style API to internally use multiple shared memory
segments to form the C<MM> object. This way C<MM> objects could have
arbitrary sizes, although the maximum size of an allocatable continous
chunk still is bounded by the maximum size of a shared memory segment.

=head1 SEE ALSO

mm-config(1).

malloc(3), calloc(3), realloc(3), strdup(3), free(3), mmap(2), shmget(2),
shmctl(2), flock(2), fcntl(2), semget(2), semctl(2), semop(2).

=head1 HOME

http://www.ossp.org/pkg/lib/mm/

=head1 HISTORY

This library was originally written in January 1999 by I<Ralf S.
Engelschall> <rse@engelschall.com> for use in the B<Extended API> (EAPI)
of the B<Apache> HTTP server project (see http://www.apache.org/), which
was originally invented for B<mod_ssl> (see http://www.modssl.org/).

Its base idea (a malloc-style API for handling shared memory) was originally
derived from the non-publically available I<mm_malloc> library written in
October 1997 by I<Charles Randall> <crandall@matchlogic.com> for MatchLogic,
Inc.

In 2000 this library joined the B<OSSP> project where all other software
development projects of I<Ralf S. Engelschall> are located.

=head1 AUTHOR

 Ralf S. Engelschall
 rse@engelschall.com
 www.engelschall.com

=cut