path: root/erts/doc/src/erl_nif.xml

<?xml version="1.0" encoding="utf-8" ?>
<!DOCTYPE cref SYSTEM "cref.dtd">

<cref>
  <header>
    <copyright>
      <year>2001</year><year>2018</year>
      <holder>Ericsson AB. All Rights Reserved.</holder>
    </copyright>
    <legalnotice>
      Licensed under the Apache License, Version 2.0 (the "License");
      you may not use this file except in compliance with the License.
      You may obtain a copy of the License at
 
          http://www.apache.org/licenses/LICENSE-2.0

      Unless required by applicable law or agreed to in writing, software
      distributed under the License is distributed on an "AS IS" BASIS,
      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
      See the License for the specific language governing permissions and
      limitations under the License.
    
    </legalnotice>

    <title>erl_nif</title>
    <prepared>Sverker Eriksson</prepared>
    <responsible>Sverker Eriksson</responsible>
    <docno>1</docno>
    <approved></approved>
    <checked></checked>
    <date>2009-11-17</date>
    <rev>PA1</rev>
    <file>erl_nif.xml</file>
  </header>
  <lib>erl_nif</lib>
  <libsummary>API functions for an Erlang NIF library.</libsummary>
  <description>
    <p>A NIF library contains native implementation of some functions
      of an Erlang module. The native implemented functions (NIFs) are
      called like any other functions without any difference to the
      caller. Each NIF must have an implementation in Erlang that
      is invoked if the function is called before the NIF library
      is successfully loaded. A typical such stub implementation
      is to throw an exception. But it can also be used as a fallback
      implementation if the NIF library is not implemented for some
      architecture.</p>

    <warning>
      <marker id="WARNING"/>
      <p><em>Use this functionality with extreme care.</em></p>
      <p>A native function is executed as a direct extension of the
        native code of the VM. Execution is not made in a safe environment.
        The VM <em>cannot</em> provide the same services as provided when
        executing Erlang code, such as pre-emptive scheduling or memory
        protection. If the native function does not behave well, the whole
        VM will misbehave.</p>
      <list type="bulleted">
        <item>
          <p>A native function that crash will crash the whole VM.</p>
        </item>
        <item>
          <p>An erroneously implemented native function can cause a VM
            internal state inconsistency, which can cause a crash of the VM,
            or miscellaneous misbehaviors of the VM at any point after the
            call to the native function.</p>
        </item>
        <item>
          <p>A native function doing <seealso marker="#lengthy_work">lengthy
            work</seealso> before returning degrades responsiveness of the VM,
            and can cause miscellaneous strange behaviors. Such strange
            behaviors include, but are not limited to, extreme memory usage,
            and bad load balancing between schedulers. Strange behaviors that
            can occur because of lengthy work can also vary between Erlang/OTP
            releases.</p>
        </item>
      </list>
    </warning>

    <p>A minimal example of a NIF library can look as follows:</p>

    <code type="none">
/* niftest.c */
#include &lt;erl_nif.h&gt;

static ERL_NIF_TERM hello(ErlNifEnv* env, int argc, const ERL_NIF_TERM argv[])
{
    return enif_make_string(env, "Hello world!", ERL_NIF_LATIN1);
}

static ErlNifFunc nif_funcs[] =
{
    {"hello", 0, hello}
};

ERL_NIF_INIT(niftest,nif_funcs,NULL,NULL,NULL,NULL)</code>

    <p>The Erlang module can look as follows:</p>

    <code type="none">
-module(niftest).

-export([init/0, hello/0]).

init() ->
      erlang:load_nif("./niftest", 0).

hello() ->
      "NIF library not loaded".</code>

    <p>Compile and test can look as follows (on Linux):</p>

    <code type="none">
$> gcc -fPIC -shared -o niftest.so niftest.c -I $ERL_ROOT/usr/include/
$> erl

1> c(niftest).
{ok,niftest}
2> niftest:hello().
"NIF library not loaded"
3> niftest:init().
ok
4> niftest:hello().
"Hello world!"</code>

    <p>A better solution for a real module is to take advantage of the new
      directive <c>on_load</c> (see section
      <seealso marker="doc/reference_manual:code_loading#on_load">Running a
      Function When a Module is Loaded</seealso> in the Erlang Reference
      Manual) to load the NIF library automatically when the module is
      loaded.</p>

    <note>
      <p>A NIF does not have to be exported, it can be local to the module.
        However, unused local stub functions will be optimized
        away by the compiler, causing loading of the NIF library to fail.</p>
    </note>

    <p>Once loaded, a NIF library is persistent. It will not be unloaded
      until the module code version that it belongs to is purged.</p>
  </description>

  <section>
  <title>Functionality</title>
    <p>All interaction between NIF code and the Erlang runtime system is
      performed by calling NIF API functions. Functions exist
      for the following functionality:</p>

    <taglist>
      <tag>Read and write Erlang terms</tag>
      <item>
        <p>Any Erlang terms can be passed to a NIF as function arguments and
          be returned as function return values. The terms are of C-type
          <seealso marker="#ERL_NIF_TERM"><c>ERL_NIF_TERM</c></seealso> and can
          only be read or written using API functions. Most functions to read
          the content of a term are prefixed <c>enif_get_</c> and usually return
          <c>true</c> (or <c>false</c>) if the term is of the expected type (or
          not). The functions to write terms are all prefixed <c>enif_make_</c>
          and usually
          return the created <c>ERL_NIF_TERM</c>. There are also some functions
          to query terms, like <c>enif_is_atom</c>, <c>enif_is_identical</c>,
          and <c>enif_compare</c>.</p>
        <p>All terms of type <c>ERL_NIF_TERM</c> belong to an environment of
          type <seealso marker="#ErlNifEnv"><c>ErlNifEnv</c></seealso>. The
          lifetime of a term is controlled by the lifetime of its environment
          object. All API functions that read or write terms has the
          environment that the term belongs to as the first function
          argument.</p>
      </item>
      <tag>Binaries</tag>
      <item>
        <p>Terms of type binary are accessed with the help of struct type
          <seealso marker="#ErlNifBinary"><c>ErlNifBinary</c></seealso>,
          which contains a pointer (<c>data</c>) to the raw binary data and the
          length (<c>size</c>) of the data in bytes. Both <c>data</c> and
          <c>size</c> are read-only and are only to be written using calls to
          API functions. Instances of <c>ErlNifBinary</c> are, however, always
          allocated by the user (usually as local variables).</p>
        <p>The raw data pointed to by <c>data</c> is only mutable after a call
          to <seealso marker="#enif_alloc_binary">
          <c>enif_alloc_binary</c></seealso> or
          <seealso marker="#enif_realloc_binary">
          <c>enif_realloc_binary</c></seealso>. All other functions that
          operate on a binary leave the data as read-only.
          A mutable binary must in the end either be freed with
          <seealso marker="#enif_release_binary">
          <c>enif_release_binary</c></seealso>
          or made read-only by transferring it to an Erlang term with
          <seealso marker="#enif_make_binary"><c>enif_make_binary</c></seealso>.
          However, it does not have to occur in the same NIF call. Read-only
          binaries do not have to be released.</p>
        <p><seealso marker="#enif_make_new_binary">
          <c>enif_make_new_binary</c></seealso> can be used as a shortcut to
          allocate and return a binary in the same NIF call.</p>
        <p>Binaries are sequences of whole bytes. Bitstrings with an arbitrary
          bit length have no support yet.</p>
      </item>
      <tag><marker id="resource_objects"/>Resource objects</tag>
      <item>
        <p>The use of resource objects is a safe way to return pointers to
          native data structures from a NIF. A resource object is
          only a block of memory allocated with
          <seealso marker="#enif_alloc_resource">
          <c>enif_alloc_resource</c></seealso>.
          A handle ("safe pointer") to this memory block can then be returned
          to Erlang by the use of
          <seealso marker="#enif_make_resource">
          <c>enif_make_resource</c></seealso>.
          The term returned by <c>enif_make_resource</c> is opaque in nature.
          It can be stored and passed between processes, but
          the only real end usage is to pass it back as an argument to a NIF.
          The NIF can then call <seealso marker="#enif_get_resource">
          <c>enif_get_resource</c></seealso> and get back a pointer to the
          memory block, which is guaranteed to still be valid. A resource
          object is not deallocated until the last handle term
          is garbage collected by the VM and the resource is released with
          <seealso marker="#enif_release_resource">
          <c>enif_release_resource</c></seealso>
          (not necessarily in that order).</p>
        <p>All resource objects are created as instances of some <em>resource
          type</em>. This makes resources from different modules to be
          distinguishable. A resource type is created by calling
          <seealso marker="#enif_open_resource_type">
          <c>enif_open_resource_type</c></seealso> when a library is loaded.
          Objects of that resource type can then later be allocated and
          <c>enif_get_resource</c> verifies that the resource is of the
          expected type. A resource type can have a user-supplied destructor
          function, which is automatically called when resources of that type
          are released (by either the garbage collector or
          <c>enif_release_resource</c>). Resource types are uniquely identified
          by a supplied name string and the name of the implementing module.</p>
        <marker id="enif_resource_example"/>
        <p>The following is a template example of how to create and return a
          resource object.</p>
        <code type="none">
ERL_NIF_TERM term;
MyStruct* obj = enif_alloc_resource(my_resource_type, sizeof(MyStruct));

/* initialize struct ... */

term = enif_make_resource(env, obj);

if (keep_a_reference_of_our_own) {
    /* store 'obj' in static variable, private data or other resource object */
}
else {
    enif_release_resource(obj);
    /* resource now only owned by "Erlang" */
}
return term;</code>
        <p>Notice that once <c>enif_make_resource</c> creates the term to
          return to Erlang, the code can choose to either keep its own
          native pointer to the allocated struct and release it later, or
          release it immediately and rely only on the garbage collector
          to deallocate the resource object eventually when it collects
          the term.</p> 
        <p>Another use of resource objects is to create binary terms with
          user-defined memory management.
          <seealso marker="#enif_make_resource_binary">
          <c>enif_make_resource_binary</c></seealso>
          creates a binary term that is connected to a resource object. The
          destructor of the resource is called when the binary is garbage
          collected, at which time the binary data can be released. An example
          of this can be a binary term consisting of data from a <c>mmap</c>'ed
          file. The destructor can then do <c>munmap</c> to release the memory
          region.</p>
        <p>Resource types support upgrade in runtime by allowing a loaded NIF
          library to take over an already existing resource type and by that
          "inherit" all existing objects of that type. The destructor of the
          new library is thereafter called for the inherited objects and the
          library with the old destructor function can be safely unloaded.
          Existing resource objects, of a module that is upgraded, must either
          be deleted or taken over by the new NIF library. The unloading of a
          library is postponed as long as there exist resource objects with a
          destructor function in the library.</p>
      </item>
      <tag>Module upgrade and static data</tag>
      <item>
	<p>A loaded NIF library is tied to the Erlang module instance
	that loaded it. If the module is upgraded, the new module instance
	needs to load its own NIF library (or maybe choose not to). The new
	module instance can, however, choose to load the exact same NIF library
	as the old code if it wants to. Sharing the dynamic library means that
	static data defined by the library is shared as well. To avoid
	unintentionally shared static data between module instances, each Erlang
	module version can keep its own private data. This private data can be
	set when the NIF library is loaded and later retrieved by calling
	<seealso marker="#enif_priv_data"><c>enif_priv_data</c></seealso>.</p>
      </item>
      <tag>Threads and concurrency</tag>
      <item>
        <p>A NIF is thread-safe without any explicit synchronization as
          long as it acts as a pure function and only reads the supplied
          arguments. When you write to a shared state either through
          static variables or <seealso marker="#enif_priv_data">
          <c>enif_priv_data</c></seealso>, you need to supply your own explicit
          synchronization. This includes terms in process-independent
          environments that are shared between threads. Resource objects also
          require synchronization if you treat them as mutable.</p>
        <p>The library initialization callbacks <c>load</c> and
          <c>upgrade</c> are thread-safe even for shared state data.</p>
      </item>
      <tag><marker id="version_management"/>Version Management</tag>
      <item>
        <p>When a NIF library is built, information about the NIF API version
          is compiled into the library. When a NIF library is loaded, the
          runtime system verifies that the library is of a compatible version.
          <c>erl_nif.h</c> defines the following:</p>
        <taglist>
          <tag><c>ERL_NIF_MAJOR_VERSION</c></tag>
          <item>
            <p>Incremented when NIF library incompatible changes are made to the
              Erlang runtime system. Normally it suffices to recompile the NIF
              library when the <c>ERL_NIF_MAJOR_VERSION</c> has changed, but it
              can, under rare circumstances, mean that NIF libraries must be
              slightly modified. If so, this will of course be documented.</p>
          </item>
          <tag><c>ERL_NIF_MINOR_VERSION</c></tag>
          <item>
            <p>Incremented when new features are added. The runtime system uses
              the minor version to determine what features to use.</p>
          </item>
        </taglist>
        <p>The runtime system normally refuses to load a NIF library if
          the major versions differ, or if the major versions are equal and
          the minor version used by the NIF library is greater than the one
          used by the runtime system. Old NIF libraries with lower major
          versions are, however, allowed after a bump of the major version
          during a transition period of two major releases. Such old NIF
          libraries can however fail if deprecated features are used.</p>
      </item>
      <tag><marker id="time_measurement"/>Time Measurement</tag>
      <item>
        <p>Support for time measurement in NIF libraries:</p>
        <list type="bulleted">
          <item><seealso marker="#ErlNifTime">
            <c>ErlNifTime</c></seealso></item>
          <item><seealso marker="#ErlNifTimeUnit">
            <c>ErlNifTimeUnit</c></seealso></item>
          <item><seealso marker="#enif_monotonic_time">
            <c>enif_monotonic_time()</c></seealso></item>
          <item><seealso marker="#enif_time_offset">
            <c>enif_time_offset()</c></seealso></item>
          <item><seealso marker="#enif_convert_time_unit">
            <c>enif_convert_time_unit()</c></seealso></item>
      </list>
      </item>
      <tag><marker id="enif_ioq"/>I/O Queues</tag>
      <item>
        <p>The Erlang nif library contains function for easily working
          with I/O vectors as used by the unix system call <c>writev</c>.
          The I/O Queue is not thread safe, so some other synchronization
          mechanism has to be used.</p>
          <list type="bulleted">
            <item><seealso marker="#SysIOVec">
            <c>SysIOVec</c></seealso></item>
            <item><seealso marker="#ErlNifIOVec">
            <c>ErlNifIOVec</c></seealso></item>
            <item><seealso marker="#enif_ioq_create">
            <c>enif_ioq_create()</c></seealso></item>
            <item><seealso marker="#enif_ioq_destroy">
            <c>enif_ioq_destroy()</c></seealso></item>
            <item><seealso marker="#enif_ioq_enq_binary">
            <c>enif_ioq_enq_binary()</c></seealso></item>
            <item><seealso marker="#enif_ioq_enqv">
            <c>enif_ioq_enqv()</c></seealso></item>
            <item><seealso marker="#enif_ioq_deq">
            <c>enif_ioq_deq()</c></seealso></item>
            <item><seealso marker="#enif_ioq_peek">
            <c>enif_ioq_peek()</c></seealso></item>
            <item><seealso marker="#enif_ioq_peek_head">
            <c>enif_ioq_peek_head()</c></seealso></item>
            <item><seealso marker="#enif_inspect_iovec">
            <c>enif_inspect_iovec()</c></seealso></item>
            <item><seealso marker="#enif_free_iovec">
            <c>enif_free_iovec()</c></seealso></item>
          </list>
        <p>Typical usage when writing to a file descriptor looks like this:</p>
          <code type="none"><![CDATA[
int writeiovec(ErlNifEnv *env, ERL_NIF_TERM term, ERL_NIF_TERM *tail,
               ErlNifIOQueue *q, int fd) {

    ErlNifIOVec vec, *iovec = &vec;
    SysIOVec *sysiovec;
    int saved_errno;
    int iovcnt, n;

    if (!enif_inspect_iovec(env, 64, term, tail, &iovec))
        return -2;

    if (enif_ioq_size(q) > 0) {
        /* If the I/O queue contains data we enqueue the iovec and
           then peek the data to write out of the queue. */
        if (!enif_ioq_enqv(q, iovec, 0))
            return -3;

        sysiovec = enif_ioq_peek(q, &iovcnt);
    } else {
        /* If the I/O queue is empty we skip the trip through it. */
        iovcnt = iovec->iovcnt;
        sysiovec = iovec->iov;
    }

    /* Attempt to write the data */
    n = writev(fd, sysiovec, iovcnt);
    saved_errno = errno;

    if (enif_ioq_size(q) == 0) {
        /* If the I/O queue was initially empty we enqueue any
           remaining data into the queue for writing later. */
        if (n >= 0 && !enif_ioq_enqv(q, iovec, n))
            return -3;
    } else {
        /* Dequeue any data that was written from the queue. */
        if (n > 0 && !enif_ioq_deq(q, n, NULL))
            return -4;
    }

    /* return n, which is either number of bytes written or -1 if
       some error happened */
    errno = saved_errno;
    return n;
}]]></code>
      </item>
      <tag><marker id="lengthy_work"/>Long-running NIFs</tag>
      <item>
        <p>As mentioned in the <seealso marker="#WARNING">warning</seealso> text
          at the beginning of this manual page, it is of <em>vital
          importance</em> that a native function returns relatively fast. It is
          difficult to give an exact maximum amount of time that a native
          function is allowed to work, but usually a well-behaving native
          function is to return to its caller within 1 millisecond. This can be
          achieved using different approaches. If you have full control over the
          code to execute in the native function, the best approach is to
          divide the work into multiple chunks of work and call the native
          function multiple times. This is, however, not always possible, for
          example when calling third-party libraries.</p>
        <p>The <seealso marker="#enif_consume_timeslice">
          <c>enif_consume_timeslice()</c></seealso> function can be used to
          inform the runtime system about the length of the NIF call.
          It is typically always to be used unless the NIF executes very
          fast.</p>
        <p>If the NIF call is too lengthy, this must be handled in one of
          the following ways to avoid degraded responsiveness, scheduler load
          balancing problems, and other strange behaviors:</p>
        <taglist>
          <tag>Yielding NIF</tag>
          <item>
            <p>If the functionality of a long-running NIF can be split so that
              its work can be achieved through a series of shorter NIF calls,
              the application has two options:</p>
            <list type="bulleted">
              <item>
                <p>Make that series of NIF calls from the Erlang level.</p>
              </item>
              <item>
                <p>Call a NIF that first performs a chunk of the work, then
                  invokes the <seealso marker="#enif_schedule_nif">
                  <c>enif_schedule_nif</c></seealso> function to schedule
                  another NIF call to perform the next chunk. The final call
                  scheduled in this manner can then return the overall
                  result.</p>
              </item>
            </list>
            <p>Breaking up a long-running function in this manner enables the
              VM to regain control between calls to the NIFs.</p>
            <p>This approach is always preferred over the other alternatives
              described below. This both from a performance perspective and
              a system characteristics perspective.</p>
          </item>
          <tag>Threaded NIF</tag>
          <item>
            <p>This is accomplished by dispatching the work to another thread
              managed by the NIF library, return from the NIF, and wait for
              the result. The thread can send the result back to the Erlang
              process using <seealso marker="#enif_send">
              <c>enif_send</c></seealso>.
              Information about thread primitives is provided below.</p>
          </item>
          <tag><marker id="dirty_nifs"/>Dirty NIF</tag>
          <item>
            <note>
              <p>Dirty NIF support is available only when the emulator is
	        configured with dirty scheduler support. As of ERTS version
		9.0, dirty scheduler support is enabled by default on the
		runtime system with SMP support. The Erlang runtime without
		SMP support does <em>not</em> support dirty schedulers even
		when the dirty scheduler support is explicitly enabled. To
		check at runtime for the presence of dirty scheduler threads,
		code can use the <seealso marker="#enif_system_info">
                <c>enif_system_info()</c></seealso> API function.</p>
            </note>
            <p>A NIF that cannot be split and cannot execute in a millisecond
              or less is called a "dirty NIF", as it performs work that the
              ordinary schedulers of the Erlang runtime system cannot handle cleanly.
              Applications that make use of such functions must indicate to the
              runtime that the functions are dirty so they can be handled 
              specially. This is handled by executing dirty jobs on a separate
	      set of schedulers called dirty schedulers. A dirty NIF executing
	      on a dirty scheduler does not have the same duration restriction
	      as a normal NIF.
            </p>

            <p>
              It is important to classify the dirty job correct. An I/O bound
	      job should be classified as such, and a CPU bound job should be
	      classified as such. If you should classify CPU bound jobs
	      as I/O bound jobs, dirty I/O schedulers might starve ordinary
	      schedulers. I/O bound jobs are expected to either block waiting
	      for I/O, and/or spend a limited amount of time moving data.
            </p>

            <p>
              To schedule a dirty NIF for execution, the application has two options:</p>
            <list type="bulleted">
              <item>
                <p>Set the appropriate flags value for the dirty NIF in its
                  <seealso marker="#ErlNifFunc"> <c>ErlNifFunc</c></seealso>
                  entry.</p>
              </item>
              <item>
                <p>Call <seealso marker="#enif_schedule_nif">
                  <c>enif_schedule_nif</c></seealso>, pass to it a pointer
                  to the dirty NIF to be executed, and indicate with argument
                  <c>flags</c> whether it expects the operation to be CPU-bound
                  or I/O-bound.</p>
              </item>
            </list>
            <p>A job that alternates between I/O bound and CPU bound can be
              reclassified and rescheduled using <c>enif_schedule_nif</c> so
              that it executes on the correct type of dirty scheduler at all
              times. For more information see the documentation of the
              <c>erl(1)</c> command line arguments
              <seealso marker="erl#+SDcpu"><c>+SDcpu</c></seealso>,
              and <seealso marker="erl#+SDio"><c>+SDio</c></seealso>.</p>
            <p>While a process executes a dirty NIF, some operations that
              communicate with it can take a very long time to complete.
              Suspend or garbage collection of a process executing a dirty
              NIF cannot be done until the dirty NIF has returned. Thus, other
              processes waiting for such operations to complete might
              have to wait for a very long time. Blocking multi-scheduling, that
              is, calling <seealso marker="erlang#system_flag_multi_scheduling">
              <c>erlang:system_flag(multi_scheduling, block)</c></seealso>, can
              also take a very long time to complete. This is because all ongoing
              dirty operations on all dirty schedulers must complete before
              the block operation can complete.</p>
            <p>Many operations communicating with a process executing a
              dirty NIF can, however, complete while it executes the
              dirty NIF. For example, retrieving information about it through
              <seealso marker="erlang:process_info/1">
              <c>erlang:process_info</c></seealso>, setting its group leader,
              register/unregister its name, and so on.</p>
            <p>Termination of a process executing a dirty NIF can only be
              completed up to a certain point while it executes the dirty NIF.
              All Erlang resources, such as its registered name and its ETS
              tables, are released. All links and monitors are triggered. The
              execution of the NIF is, however, <em>not</em> stopped. The NIF
              can safely continue execution, allocate heap memory, and so on,
              but it is of course better to stop executing as soon as possible.
              The NIF can check whether a current process is alive using
              <seealso marker="#enif_is_current_process_alive">
              <c>enif_is_current_process_alive</c></seealso>. Communication
              using <seealso marker="#enif_send"><c>enif_send</c></seealso> and
              <seealso marker="#enif_port_command">
              <c>enif_port_command</c></seealso> is also dropped when the
              sending process is not alive. Deallocation of certain internal
              resources, such as process heap and process control block, is
              delayed until the dirty NIF has completed.</p>
          </item>
        </taglist>
      </item>
    </taglist>
  </section>

  <section>
    <title>Initialization</title>
    <taglist>
      <tag><marker id="ERL_NIF_INIT"/><c>ERL_NIF_INIT(MODULE,
        ErlNifFunc funcs[], load, NULL, upgrade, unload)</c></tag>
      <item>
        <p>This is the magic macro to initialize a NIF library. It
          is to be evaluated in global file scope.</p>
        <p><c>MODULE</c> is the name of the Erlang module as an
          identifier without string quotations. It is stringified by
          the macro.</p>
        <p><c>funcs</c> is a static array of function descriptors for
          all the implemented NIFs in this library.</p>
        <p><c>load</c>, <c>upgrade</c> and <c>unload</c>
          are pointers to functions. One of <c>load</c> or
          <c>upgrade</c> is called to initialize the library.
          <c>unload</c> is called to release the library. All are
          described individually below.</p>
       <p>The fourth argument <c>NULL</c> is ignored. It
          was earlier used for the deprecated <c>reload</c> callback
          which is no longer supported since OTP 20.</p>
        <p>If compiling a NIF for static inclusion through
          <c>--enable-static-nifs</c>, you must define <c>STATIC_ERLANG_NIF</c>
           before the <c>ERL_NIF_INIT</c> declaration.</p>
      </item>
      <tag><marker id="load"/><c>int (*load)(ErlNifEnv* env, void** priv_data,
        ERL_NIF_TERM load_info)</c></tag>
      <item>
        <p><c>load</c> is called when the NIF library is loaded
          and no previously loaded library exists for this module.</p>
        <p><c>*priv_data</c> can be set to point to some private data
          if the library needs to keep a state between NIF
          calls. <c>enif_priv_data</c> returns this pointer.
          <c>*priv_data</c> is initialized to <c>NULL</c> when <c>load</c> is
          called.</p> 
        <p><c>load_info</c> is the second argument to <seealso
          marker="erlang#load_nif-2"><c>erlang:load_nif/2</c></seealso>.</p>
        <p>The library fails to load if <c>load</c> returns
          anything other than <c>0</c>. <c>load</c> can be <c>NULL</c> if
          initialization is not needed.</p> 
      </item>
      <tag><marker id="upgrade"/><c>int (*upgrade)(ErlNifEnv* env, void**
        priv_data, void** old_priv_data, ERL_NIF_TERM load_info)</c></tag>
      <item>
        <p><c>upgrade</c> is called when the NIF library is loaded
          and there is old code of this module with a loaded NIF library.</p>
        <p>Works as <c>load</c>, except that <c>*old_priv_data</c> already
          contains the value set by the last call to <c>load</c> or
          <c>upgrade</c> for the old module code. <c>*priv_data</c> is
          initialized to <c>NULL</c> when <c>upgrade</c> is called. It is
          allowed to write to both <c>*priv_data</c> and
          <c>*old_priv_data.</c></p> 
        <p>The library fails to load if <c>upgrade</c> returns
           anything other than <c>0</c> or if <c>upgrade</c> is <c>NULL</c>.</p>
      </item>
      <tag><marker id="unload"/><c>void (*unload)(ErlNifEnv* env, void*
        priv_data)</c></tag>
      <item>
        <p><c>unload</c> is called when the module code that
          the NIF library belongs to is purged as old. New code of the same
          module may or may not exist.</p>
      </item>
    </taglist>
  </section>

  <section>
    <title>Data Types</title>
    <taglist>
      <tag><marker id="ERL_NIF_TERM"/><c>ERL_NIF_TERM</c></tag>
      <item>
        <p>Variables of type <c>ERL_NIF_TERM</c> can refer to any Erlang term.
          This is an opaque type and values of it can only by used either as
          arguments to API functions or as return values from NIFs. All
          <c>ERL_NIF_TERM</c>s belong to an environment
          (<seealso marker="#ErlNifEnv"><c>ErlNifEnv</c></seealso>).
          A term cannot be destructed individually, it is valid until its
          environment is destructed.</p>
      </item>
      <tag><marker id="ErlNifEnv"/><c>ErlNifEnv</c></tag>
      <item>
        <p><c>ErlNifEnv</c> represents an environment that can host Erlang
          terms. All terms in an environment are valid as long as the
          environment is valid. <c>ErlNifEnv</c> is an opaque type; pointers to
          it can only be passed on to API functions. Two types of environments
          exist:</p>
        <taglist>
          <tag>Process-bound environment</tag>
          <item>
            <p>Passed as the first argument to all NIFs. All function arguments
              passed to a NIF belong to that environment. The return value from
              a NIF must also be a term belonging to the same environment.</p>
            <p>A process-bound environment contains transient information
              about the calling Erlang process. The environment is only valid
              in the thread where it was supplied as argument until the NIF
              returns. It is thus useless and dangerous to store pointers to
              process-bound environments between NIF calls.</p>
          </item>
          <tag>Process-independent environment</tag>
          <item>
            <p>Created by calling <seealso marker="#enif_alloc_env">
              <c>enif_alloc_env</c></seealso>. This environment can be
              used to store terms between NIF calls and to send terms with
              <seealso marker="#enif_send"><c>enif_send</c></seealso>. A
              process-independent environment with all its terms is valid until
              you explicitly invalidate it with
              <seealso marker="#enif_free_env"><c>enif_free_env</c></seealso>
              or <c>enif_send</c>.</p>
          </item>
        </taglist>
        <p>All contained terms of a list/tuple/map must belong to the same
          environment as the list/tuple/map itself. Terms can be copied between
          environments with
          <seealso marker="#enif_make_copy"><c>enif_make_copy</c></seealso>.</p>
      </item>
      <tag><marker id="ErlNifFunc"/><c>ErlNifFunc</c></tag>
      <item>
        <code type="none">
typedef struct {
    const char* name;
    unsigned arity;
    ERL_NIF_TERM (*fptr)(ErlNifEnv* env, int argc, const ERL_NIF_TERM argv[]);
    unsigned flags;
} ErlNifFunc;</code>
        <p>Describes a NIF by its name, arity, and implementation.</p>
        <taglist>
          <tag><c>fptr</c></tag>
          <item>
            <p>A pointer to the function that implements the NIF.</p>
          </item>
          <tag><c>argv</c></tag>
          <item>
            <p>Contains the function arguments passed to the NIF.</p>
          </item>
          <tag><c>argc</c></tag>
          <item>
            <p>The array length, that is, the function arity. <c>argv[N-1]</c>
              thus denotes the Nth argument to the NIF. Notice that the argument
              <c>argc</c> allows for the same C function to implement several
              Erlang functions with different arity (but probably with the same
              name).</p>
          </item>
          <tag><c>flags</c></tag>
          <item>
            <p>Is <c>0</c> for a regular NIF (and so its value can be omitted
              for statically initialized <c>ErlNifFunc</c> instances).</p>
            <p><c>flags</c> can be used to indicate that the NIF is a
              <seealso marker="#dirty_nifs">dirty NIF</seealso> that is to be
              executed on a dirty scheduler thread.</p>
            <p>If the dirty NIF is expected to be CPU-bound, its <c>flags</c>
              field is to be set to <c>ERL_NIF_DIRTY_JOB_CPU_BOUND</c> or
              <c>ERL_NIF_DIRTY_JOB_IO_BOUND</c>.</p>
            <note>
              <p>If one of the <c>ERL_NIF_DIRTY_JOB_*_BOUND</c> flags is set,
                and the runtime system has no support for dirty schedulers,
                the runtime system refuses to load the NIF library.</p>
            </note>
          </item>
        </taglist>
      </item>
      <tag><marker id="ErlNifBinary"/><c>ErlNifBinary</c></tag>
      <item>
        <code type="none">
typedef struct {
    unsigned size;
    unsigned char* data;
} ErlNifBinary;</code>
        <p><c>ErlNifBinary</c> contains transient information about an
          inspected binary term. <c>data</c> is a pointer to a buffer
          of <c>size</c> bytes with the raw content of the binary.</p>
          <p>Notice that <c>ErlNifBinary</c> is a semi-opaque type and you are
          only allowed to read fields <c>size</c> and <c>data</c>.</p>
      </item>
      <tag><marker id="ErlNifBinaryToTerm"/><c>ErlNifBinaryToTerm</c></tag>
      <item>
        <p>An enumeration of the options that can be specified to
          <seealso marker="#enif_binary_to_term">
          <c>enif_binary_to_term</c></seealso>.
          For default behavior, use value <c>0</c>.</p>
        <p>When receiving data from untrusted sources, use option
          <c>ERL_NIF_BIN2TERM_SAFE</c>.</p>
      </item>
      <tag><marker id="ErlNifMonitor"/><c>ErlNifMonitor</c></tag>
      <item>
        <p>This is an opaque data type that identifies a monitor.</p>
        <p>The nif writer is to provide the memory for storing the
          monitor when calling <seealso marker="#enif_monitor_process">
          <c>enif_monitor_process</c></seealso>. The
          address of the data is not stored by the runtime system, so
          <c>ErlNifMonitor</c> can be used as any other data, it
          can be copied, moved in memory, forgotten, and so on.
	  To compare two monitors, <seealso marker="#enif_compare_monitors">
          <c>enif_compare_monitors</c></seealso> must be used.</p>
      </item>
      <tag><marker id="ErlNifPid"/><c>ErlNifPid</c></tag>
      <item>
        <p>A process identifier (pid). In contrast to pid terms (instances of
          <c>ERL_NIF_TERM</c>), <c>ErlNifPid</c>s are self-contained and not
          bound to any <seealso marker="#ErlNifEnv">environment</seealso>.
          <c>ErlNifPid</c> is an opaque type.</p>
      </item>
      <tag><marker id="ErlNifPort"/><c>ErlNifPort</c></tag>
      <item>
        <p>A port identifier. In contrast to port ID terms (instances of
          <c>ERL_NIF_TERM</c>), <c>ErlNifPort</c>s are self-contained and not
          bound to any <seealso marker="#ErlNifEnv">environment</seealso>.
          <c>ErlNifPort</c> is an opaque type.</p>
      </item>
      <tag><marker id="ErlNifResourceType"/><c>ErlNifResourceType</c></tag>
      <item>
        <p>Each instance of <c>ErlNifResourceType</c> represents a class of
          memory-managed resource objects that can be garbage collected.
          Each resource type has a unique name and a destructor function that
          is called when objects of its type are released.</p>
      </item>
      <tag><marker id="ErlNifResourceTypeInit"/><c>ErlNifResourceTypeInit</c></tag>
      <item>
        <code type="none">
typedef struct {
    ErlNifResourceDtor* dtor;
    ErlNifResourceStop* stop;
    ErlNifResourceDown* down;
} ErlNifResourceTypeInit;</code>
        <p>Initialization structure read by <seealso marker="#enif_open_resource_type_x">
	enif_open_resource_type_x</seealso>.</p>
      </item>
      <tag><marker id="ErlNifResourceDtor"/><c>ErlNifResourceDtor</c></tag>
      <item>
        <code type="none">
typedef void ErlNifResourceDtor(ErlNifEnv* env, void* obj);</code>
        <p>The function prototype of a resource destructor function.</p>
	<p>The <c>obj</c> argument is a pointer to the resource. The only
	allowed use for the resource in the destructor is to access its
	user data one final time. The destructor is guaranteed to be the
	last callback before the resource is deallocated.</p>
      </item>
      <tag><marker id="ErlNifResourceDown"/><c>ErlNifResourceDown</c></tag>
      <item>
        <code type="none">
typedef void ErlNifResourceDown(ErlNifEnv* env, void* obj, ErlNifPid* pid, ErlNifMonitor* mon);</code>
        <p>The function prototype of a resource down function,
	  called on the behalf of <seealso marker="#enif_monitor_process">
	  enif_monitor_process</seealso>. <c>obj</c> is the resource, <c>pid</c>
	  is the identity of the monitored process that is exiting, and <c>mon</c>
	  is the identity of the monitor.
	</p>
      </item>
      <tag><marker id="ErlNifResourceStop"/><c>ErlNifResourceStop</c></tag>
      <item>
        <code type="none">
typedef void ErlNifResourceStop(ErlNifEnv* env, void* obj, ErlNifEvent event, int is_direct_call);</code>
        <p>The function prototype of a resource stop function,
	  called on the behalf of <seealso marker="#enif_select">
	  enif_select</seealso>. <c>obj</c> is the resource, <c>event</c> is OS event,
	  <c>is_direct_call</c> is true if the call is made directly from <c>enif_select</c>
	  or false if it is a scheduled call (potentially from another thread).</p>
      </item>
      <tag><marker id="ErlNifCharEncoding"/><c>ErlNifCharEncoding</c></tag>
      <item>
        <code type="none">
typedef enum {
    ERL_NIF_LATIN1
}ErlNifCharEncoding;</code>
        <p>The character encoding used in strings and atoms. The only
          supported encoding is <c>ERL_NIF_LATIN1</c> for
          ISO Latin-1 (8-bit ASCII).</p>
      </item>
      <tag><marker id="ErlNifSysInfo"/><c>ErlNifSysInfo</c></tag>
      <item>
        <p>Used by <seealso marker="#enif_system_info">
          <c>enif_system_info</c></seealso> to return information about the
          runtime system. Contains the same content as
          <seealso marker="erl_driver#ErlDrvSysInfo">
          <c>ErlDrvSysInfo</c></seealso>.</p>
      </item>
      <tag><marker id="ErlNifSInt64"/><c>ErlNifSInt64</c></tag>
      <item>
        <p>A native signed 64-bit integer type.</p>
      </item>
      <tag><marker id="ErlNifUInt64"/><c>ErlNifUInt64</c></tag>
      <item>
        <p>A native unsigned 64-bit integer type.</p>
      </item>
      <tag><marker id="ErlNifTime"/><c>ErlNifTime</c></tag>
      <item>
        <p>A signed 64-bit integer type for representation of time.</p>
      </item>
      <tag><marker id="ErlNifTimeUnit"/><c>ErlNifTimeUnit</c></tag>
      <item>
        <p>An enumeration of time units supported by the NIF API:</p>
        <taglist>
          <tag><c>ERL_NIF_SEC</c></tag>
          <item>Seconds</item>
          <tag><c>ERL_NIF_MSEC</c></tag>
          <item>Milliseconds</item>
          <tag><c>ERL_NIF_USEC</c></tag>
          <item>Microseconds</item>
          <tag><c>ERL_NIF_NSEC</c></tag>
          <item>Nanoseconds</item>
        </taglist>
      </item>
      <tag><marker id="ErlNifUniqueInteger"/><c>ErlNifUniqueInteger</c></tag>
      <item>
        <p>An enumeration of the properties that can be requested from
          <seealso marker="#enif_make_unique_integer">
          <c>enif_make_unique_integer</c></seealso>.
          For default properties, use value <c>0</c>.</p>
        <taglist>
          <tag><c>ERL_NIF_UNIQUE_POSITIVE</c></tag>
          <item>
            <p>Return only positive integers.</p>
          </item>
          <tag><c>ERL_NIF_UNIQUE_MONOTONIC</c></tag>
          <item>
            <p>Return only <seealso
              marker="time_correction#Strictly_Monotonically_Increasing">
              strictly monotonically increasing</seealso> integer corresponding
              to creation time.</p>
          </item>
        </taglist>
      </item>
      <tag><marker id="ErlNifHash"/><c>ErlNifHash</c></tag>
      <item>
        <p>An enumeration of the supported hash types that can be generated
          using <seealso marker="#enif_hash"><c>enif_hash</c></seealso>.
        </p>
        <taglist>
          <tag><c>ERL_NIF_INTERNAL_HASH</c></tag>
          <item>
            <p>Non-portable hash function that only guarantees the same hash
              for the same term within one Erlang VM instance.</p>
            <p>It takes 32-bit salt values and generates hashes within <c>0..2^32-1</c>.</p>
          </item>
          <tag><c>ERL_NIF_PHASH2</c></tag>
          <item>
            <p>Portable hash function that gives the same hash for the
              same Erlang term regardless of machine architecture and ERTS version.</p>
            <p><em>It ignores salt values</em> and generates hashes within <c>0..2^27-1</c>.</p>
            <p>Slower than <c>ERL_NIF_INTERNAL_HASH.</c>
              It corresponds to <seealso marker="erlang#phash2-1"><c>erlang:phash2/1</c></seealso>.
            </p>
          </item>
        </taglist>
      </item>
      <tag><marker id="SysIOVec"/><c>SysIOVec</c></tag>
      <item>
        <p>A system I/O vector, as used by <c>writev</c> on
          Unix and <c>WSASend</c> on Win32. It is used in
          <c>ErlNifIOVec</c> and by
          <seealso marker="#enif_ioq_peek"><c>enif_ioq_peek</c></seealso>.</p>
      </item>
      <tag><marker id="ErlNifIOVec"/><c>ErlNifIOVec</c></tag>
      <item>
        <code type="none">
typedef struct {
  int iovcnt;
  size_t size;
  SysIOVec* iov;
} ErlNifIOVec;</code>
        <p>An I/O vector containing <c>iovcnt</c> <c>SysIOVec</c>s
          pointing to the data. It is used by
          <seealso marker="#enif_inspect_iovec">
            <c>enif_inspect_iovec</c></seealso> and
          <seealso marker="#enif_ioq_enqv">
            <c>enif_ioq_enqv</c></seealso>.</p>
      </item>
      <tag><marker id="ErlNifIOQueueOpts"/><c>ErlNifIOQueueOpts</c></tag>
      <item>
        Options to configure a <c>ErlNifIOQueue</c>.
        <taglist>
          <tag>ERL_NIF_IOQ_NORMAL</tag>
          <item><p>Create a normal I/O Queue</p></item>
        </taglist>
      </item>
    </taglist>
  </section>

  <funcs>
    <func>
      <name><ret>void *</ret><nametext>enif_alloc(size_t size)</nametext></name>
      <fsummary>Allocate dynamic memory.</fsummary>
      <desc>
        <p>Allocates memory of <c>size</c> bytes.</p>
        <p>Returns <c>NULL</c> if the allocation fails.</p>
        <p>The returned pointer is suitably aligned for any built-in type that
          fit in the allocated memory.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_alloc_binary(size_t size, ErlNifBinary* bin)</nametext>
      </name>
      <fsummary>Create a new binary.</fsummary>
      <desc>
        <p>Allocates a new binary of size <c>size</c> bytes.
          Initializes the structure pointed to by <c>bin</c> to
          refer to the allocated binary. The binary must either be released by
          <seealso marker="#enif_release_binary">
          <c>enif_release_binary</c></seealso> 
          or ownership transferred to an Erlang term with
          <seealso marker="#enif_make_binary"><c>enif_make_binary</c></seealso>.
          An allocated (and owned) <c>ErlNifBinary</c> can be kept between NIF
          calls.</p>
        <p>If you do not need to reallocate or keep the data alive across NIF
          calls, consider using <seealso marker="#enif_make_new_binary">
          <c>enif_make_new_binary</c></seealso> instead as it will allocate
          small binaries on the process heap when possible.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if allocation
          fails.</p>
      </desc>
    </func>

    <func>
      <name><ret>ErlNifEnv *</ret><nametext>enif_alloc_env()</nametext></name>
      <fsummary>Create a new environment.</fsummary>
      <desc>
        <p>Allocates a new process-independent environment. The environment can
          be used to hold terms that are not bound to any process. Such terms
          can later be copied to a process environment with
          <seealso marker="#enif_make_copy"><c>enif_make_copy</c></seealso> or
          be sent to a process as a message with <seealso marker="#enif_send">
          <c>enif_send</c></seealso>.</p>
        <p>Returns pointer to the new environment.</p>
      </desc>
    </func>

    <func>
      <name><ret>void *</ret><nametext>enif_alloc_resource(ErlNifResourceType*
        type, unsigned size)</nametext></name>
      <fsummary>Allocate a memory-managed resource object.</fsummary>
      <desc>
        <p>Allocates a memory-managed resource object of type <c>type</c> and
          size <c>size</c> bytes.</p>
      </desc>
    </func>

    <func>
      <name><ret>size_t</ret><nametext>enif_binary_to_term(ErlNifEnv *env,
        const unsigned char* data, size_t size, ERL_NIF_TERM *term,
        ErlNifBinaryToTerm opts)</nametext></name>
      <fsummary>Create a term from the external format.</fsummary>
      <desc>
        <p>Creates a term that is the result of decoding the binary data at
          <c>data</c>, which must be encoded according to the Erlang external
          term format. No more than <c>size</c> bytes are read from <c>data</c>.
          Argument <c>opts</c> corresponds to the second argument to
          <seealso marker="erlang#binary_to_term-2">
          <c>erlang:binary_to_term/2</c></seealso> and must be either <c>0</c>
          or <c>ERL_NIF_BIN2TERM_SAFE</c>.</p>
        <p>On success, stores the resulting term at <c>*term</c> and returns
          the number of bytes read. Returns <c>0</c> if decoding fails or if
          <c>opts</c> is invalid.</p>
        <p>See also <seealso marker="#ErlNifBinaryToTerm">
          <c>ErlNifBinaryToTerm</c></seealso>,
          <seealso marker="erlang#binary_to_term-2">
          <c>erlang:binary_to_term/2</c></seealso>, and
          <seealso marker="#enif_term_to_binary">
          <c>enif_term_to_binary</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret><nametext>enif_clear_env(ErlNifEnv* env)</nametext>
      </name>
      <fsummary>Clear an environment for reuse.</fsummary>
      <desc>
        <p>Frees all terms in an environment and clears it for reuse.
          The environment must have been allocated with
          <seealso marker="#enif_alloc_env"><c>enif_alloc_env</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_compare(ERL_NIF_TERM lhs, ERL_NIF_TERM rhs)</nametext>
      </name>
      <fsummary>Compare two terms.</fsummary>
      <desc>
        <p>Returns an integer &lt; <c>0</c> if <c>lhs</c> &lt; <c>rhs</c>,
          <c>0</c> if <c>lhs</c> = <c>rhs</c>, and &gt; <c>0</c> if
          <c>lhs</c> &gt; <c>rhs</c>. Corresponds to the Erlang
          operators <c>==</c>, <c>/=</c>, <c>=&lt;</c>, <c>&lt;</c>,
          <c>&gt;=</c>, and <c>&gt;</c> (but <em>not</em> <c>=:=</c> or
          <c>=/=</c>).</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_compare_monitors(const ErlNifMonitor
        *monitor1, const ErlNifMonitor *monitor2)</nametext></name>
      <fsummary>Compare two monitors.</fsummary>
      <desc>
        <marker id="enif_compare_monitors"></marker>
        <p>Compares two <seealso marker="#ErlNifMonitor"><c>ErlNifMonitor</c></seealso>s.
          Can also be used to imply some artificial order on monitors,
          for whatever reason.</p>
        <p>Returns <c>0</c> if <c>monitor1</c> and <c>monitor2</c> are equal,
          &lt; <c>0</c> if <c>monitor1</c> &lt; <c>monitor2</c>, and
          &gt; <c>0</c> if <c>monitor1</c> &gt; <c>monitor2</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret>
        <nametext>enif_cond_broadcast(ErlNifCond *cnd)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_cond_broadcast">
          <c>erl_drv_cond_broadcast</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>ErlNifCond *</ret>
        <nametext>enif_cond_create(char *name)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_cond_create">
          <c>erl_drv_cond_create</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret>
        <nametext>enif_cond_destroy(ErlNifCond *cnd)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_cond_destroy">
          <c>erl_drv_cond_destroy</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>char*</ret>
        <nametext>enif_cond_name(ErlNifCond* cnd)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_cond_name">
          <c>erl_drv_cond_name</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret>
        <nametext>enif_cond_signal(ErlNifCond *cnd)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_cond_signal">
          <c>erl_drv_cond_signal</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret>
        <nametext>enif_cond_wait(ErlNifCond *cnd, ErlNifMutex *mtx)</nametext>
      </name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_cond_wait">
          <c>erl_drv_cond_wait</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_consume_timeslice(ErlNifEnv *env, int percent)</nametext>
      </name>
      <fsummary></fsummary>
      <desc>
        <p>Gives the runtime system a hint about how much CPU time the current
          NIF call has consumed since the last hint, or since the start of the
          NIF if no previous hint has been specified. The time is specified as a
          percent of the timeslice that a process is allowed to execute
          Erlang code until it can be suspended to give time for other runnable
          processes. The scheduling timeslice is not an exact entity, but can
          usually be approximated to about 1 millisecond.</p>
        <p>Notice that it is up to the runtime system to determine if and how
          to use this information. Implementations on some platforms can use
          other means to determine consumed CPU time. Lengthy NIFs should
          regardless of this frequently call <c>enif_consume_timeslice</c> to
          determine if it is allowed to continue execution.</p>
        <p>Argument <c>percent</c> must be an integer between 1 and 100. This
          function must only be called from a NIF-calling thread, and argument
          <c>env</c> must be the environment of the calling process.</p>
        <p>Returns <c>1</c> if the timeslice is exhausted, otherwise <c>0</c>.
          If <c>1</c> is returned, the NIF is to return as soon as possible in
          order for the process to yield.</p>
        <p>This function is provided to better support co-operative scheduling,
          improve system responsiveness, and make it easier to prevent
          misbehaviors of the VM because of a NIF monopolizing a scheduler
          thread. It can be used to divide <seealso marker="#lengthy_work">
          length work</seealso> into a number of repeated NIF calls without the
          need to create threads.</p>
        <p>See also the <seealso marker="#WARNING">warning</seealso> text at
          the beginning of this manual page.</p>
      </desc>
    </func>

    <func>
      <name><ret>ErlNifTime</ret><nametext>enif_convert_time_unit(ErlNifTime
        val, ErlNifTimeUnit from, ErlNifTimeUnit to)</nametext></name>
      <fsummary>Convert time unit of a time value.</fsummary>
      <desc>
        <marker id="enif_convert_time_unit"></marker>
        <p>Converts the <c>val</c> value of time unit <c>from</c> to
          the corresponding value of time unit <c>to</c>. The result is
          rounded using the floor function.</p>
        <taglist>
          <tag><c>val</c></tag>
          <item>Value to convert time unit for.</item>
          <tag><c>from</c></tag>
          <item>Time unit of <c>val</c>.</item>
          <tag><c>to</c></tag>
          <item>Time unit of returned value.</item>
        </taglist>
        <p>Returns <c>ERL_NIF_TIME_ERROR</c> if called with an invalid
          time unit argument.</p>
        <p>See also <seealso marker="#ErlNifTime"><c>ErlNifTime</c></seealso>
          and
          <seealso marker="#ErlNifTimeUnit"><c>ErlNifTimeUnit</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret>
        <nametext>enif_cpu_time(ErlNifEnv *)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Returns the CPU time in the same format as
          <seealso marker="erlang#timestamp-0">
          <c>erlang:timestamp()</c></seealso>.
          The CPU time is the time the current logical CPU has spent executing
          since some arbitrary point in the past. If the OS does not support
          fetching this value, <c>enif_cpu_time</c> invokes
          <seealso marker="#enif_make_badarg">
          <c>enif_make_badarg</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_demonitor_process(ErlNifEnv* env, void* obj,
      const ErlNifMonitor* mon)</nametext></name>
      <fsummary>Cancel a process monitor.</fsummary>
      <desc>
        <marker id="enif_demonitor_process"></marker>
        <p>Cancels a monitor created earlier with <seealso marker="#enif_monitor_process">
	<c>enif_monitor_process</c></seealso>. Argument <c>obj</c> is a pointer
	to the resource holding the monitor and	<c>*mon</c> identifies the monitor.</p>
        <p>Returns <c>0</c> if the monitor was successfully identified and removed.
	Returns	a non-zero value if the monitor could not be identified, which means
	it was either</p>
	<list type="bulleted">
          <item>never created for this resource</item>
          <item>already cancelled</item>
          <item>already triggered</item>
          <item>just about to be triggered by a concurrent thread</item>
        </list>
        <p>This function is only thread-safe when the emulator with SMP support
          is used. It can only be used in a non-SMP emulator from a NIF-calling
          thread.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_equal_tids(ErlNifTid tid1, ErlNifTid tid2)</nametext>
      </name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_equal_tids">
          <c>erl_drv_equal_tids</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_fprintf(FILE *stream, const char *format, ...)</nametext></name>
      <fsummary>Format strings and Erlang terms.</fsummary>
      <desc>
        <p>Similar to <c>fprintf</c> but this format string also accepts
        <c>"%T"</c>, which formats Erlang terms of type
	<seealso marker="#ERL_NIF_TERM"><c>ERL_NIF_TERM</c></seealso>.</p>
	<p>This function is primarily intended for debugging purpose. It is not
	  recommended to print very large terms with <c>%T</c>. The function may
	  change <c>errno</c>, even if successful.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret><nametext>enif_free(void* ptr)</nametext></name>
      <fsummary>Free dynamic memory.</fsummary>
      <desc>
        <p>Frees memory allocated by
          <seealso marker="#enif_alloc"><c>enif_alloc</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret>
        <nametext>enif_free_env(ErlNifEnv* env)</nametext></name>
      <fsummary>Free an environment allocated with enif_alloc_env.</fsummary>
      <desc>
        <p>Frees an environment allocated with
          <seealso marker="#enif_alloc_env"><c>enif_alloc_env</c></seealso>.
          All terms created in the environment are freed as well.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret>
        <nametext>enif_free_iovec(ErlNifIOvec* iov)</nametext></name>
      <fsummary>Free an ErlIOVec</fsummary>
      <desc>
        <p>Frees an io vector returned from
          <seealso marker="#enif_inspect_iovec">
            <c>enif_inspect_iovec</c></seealso>.
          This is needed only if a <c>NULL</c> environment is passed to
          <seealso marker="#enif_inspect_iovec">
          <c>enif_inspect_iovec</c></seealso>.</p>
          <code type="none"><![CDATA[
ErlNifIOVec *iovec = NULL;
size_t max_elements = 128;
ERL_NIF_TERM tail;
if (!enif_inspect_iovec(NULL, max_elements, term, &tail, &iovec))
  return 0;

// Do things with the iovec

/* Free the iovector, possibly in another thread or nif function call */
enif_free_iovec(iovec);]]></code>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_get_atom(ErlNifEnv* env, ERL_NIF_TERM
        term, char* buf, unsigned size, ErlNifCharEncoding encode)</nametext>
      </name>
      <fsummary>Get the text representation of an atom term.</fsummary>
      <desc>
        <p>Writes a <c>NULL</c>-terminated string in the buffer pointed to by
          <c>buf</c> of size <c>size</c>, consisting of the string
          representation of the atom <c>term</c> with encoding
          <seealso marker="#ErlNifCharEncoding">encode</seealso>.</p>
        <p>Returns the number of bytes written (including terminating
          <c>NULL</c> character) or <c>0</c> if <c>term</c> is not an atom with
           maximum length of <c>size-1</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_get_atom_length(ErlNifEnv* env,
        ERL_NIF_TERM term, unsigned* len, ErlNifCharEncoding encode)</nametext>
      </name>
      <fsummary>Get the length of atom <c>term</c>.</fsummary>
      <desc>
        <p>Sets <c>*len</c> to the length (number of bytes excluding
          terminating <c>NULL</c> character) of the atom <c>term</c> with
          encoding <c>encode</c>.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if <c>term</c> is not
          an atom.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_get_double(ErlNifEnv* env,
        ERL_NIF_TERM term, double* dp)</nametext></name>
      <fsummary>Read a floating-point number term.</fsummary>
      <desc>
        <p>Sets <c>*dp</c> to the floating-point value of <c>term</c>.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if <c>term</c> is not
          a float.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_get_int(ErlNifEnv* env, ERL_NIF_TERM
        term, int* ip)</nametext></name>
      <fsummary>Read an integer term.</fsummary>
      <desc>
        <p>Sets <c>*ip</c> to the integer value of <c>term</c>.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if <c>term</c> is not
          an integer or is outside the bounds of type <c>int</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_get_int64(ErlNifEnv* env, ERL_NIF_TERM
        term, ErlNifSInt64* ip)</nametext></name>
      <fsummary>Read a 64-bit integer term.</fsummary>
      <desc>
        <p>Sets <c>*ip</c> to the integer value of <c>term</c>.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if <c>term</c> is not
          an integer or is outside the bounds of a signed 64-bit integer.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_get_local_pid(ErlNifEnv* env,
        ERL_NIF_TERM term, ErlNifPid* pid)</nametext></name>
      <fsummary>Read a local pid term.</fsummary>
      <desc>
        <p>If <c>term</c> is the pid of a node local process, this function
          initializes the pid variable <c>*pid</c> from it and returns
          <c>true</c>. Otherwise returns <c>false</c>. No check is done to see
          if the process is alive.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_get_local_port(ErlNifEnv* env,
        ERL_NIF_TERM term, ErlNifPort* port_id)</nametext></name>
      <fsummary>Read a local port term.</fsummary>
      <desc>
        <p>If <c>term</c> identifies a node local port, this function
          initializes the port variable <c>*port_id</c> from it and returns
          <c>true</c>. Otherwise returns <c>false</c>. No check is done to see
          if the port is alive.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_get_list_cell(ErlNifEnv* env,
        ERL_NIF_TERM list, ERL_NIF_TERM* head, ERL_NIF_TERM* tail)</nametext>
      </name>
      <fsummary>Get head and tail from a list.</fsummary>
      <desc>
        <p>Sets <c>*head</c> and <c>*tail</c> from list <c>list</c>.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if it is
          not a list or the list is empty.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_get_list_length(ErlNifEnv* env,
        ERL_NIF_TERM term, unsigned* len)</nametext></name>
      <fsummary>Get the length of list <c>term</c>.</fsummary>
      <desc>
        <p>Sets <c>*len</c> to the length of list <c>term</c>.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if <c>term</c> is
          not a proper list.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_get_long(ErlNifEnv* env, ERL_NIF_TERM
        term, long int* ip)</nametext></name>
      <fsummary>Read a long integer term.</fsummary>
      <desc>
        <p>Sets <c>*ip</c> to the long integer value of <c>term</c>.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if <c>term</c> is
          not an integer or is outside the bounds of type <c>long int</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_get_map_size(ErlNifEnv* env,
        ERL_NIF_TERM term, size_t *size)</nametext></name>
      <fsummary>Read the size of a map term.</fsummary>
      <desc>
        <p>Sets <c>*size</c> to the number of key-value pairs in the map
          <c>term</c>.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if <c>term</c> is
          not a map.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_get_map_value(ErlNifEnv* env,
        ERL_NIF_TERM map, ERL_NIF_TERM key, ERL_NIF_TERM* value)</nametext>
      </name>
      <fsummary>Get the value of a key in a map.</fsummary>
      <desc>
        <p>Sets <c>*value</c> to the value associated with <c>key</c> in the
          map <c>map</c>.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if <c>map</c> is not
          a map or if <c>map</c> does not contain <c>key</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_get_resource(ErlNifEnv* env,
        ERL_NIF_TERM term, ErlNifResourceType* type, void** objp)</nametext>
      </name>
      <fsummary>Get the pointer to a resource object.</fsummary>
      <desc>
        <p>Sets <c>*objp</c> to point to the resource object referred to by
          <c>term</c>.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if <c>term</c> is
          not a handle to a resource object of type <c>type</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_get_string(ErlNifEnv* env,
        ERL_NIF_TERM list, char* buf, unsigned size,
        ErlNifCharEncoding encode)</nametext></name>
      <fsummary>Get a C-string from a list.</fsummary>
      <desc>
        <p>Writes a <c>NULL</c>-terminated string in the buffer pointed to by
          <c>buf</c> with size <c>size</c>, consisting of the characters
          in the string <c>list</c>. The characters are written using encoding
          <seealso marker="#ErlNifCharEncoding">encode</seealso>.</p>
        <p>Returns one of the following:</p>
        <list type="bulleted">
          <item>The number of bytes written (including terminating <c>NULL</c>
            character)</item>
          <item><c>-size</c> if the string was truncated because of buffer
            space</item>
          <item><c>0</c> if <c>list</c> is not a string that can be encoded
            with <c>encode</c> or if <c>size</c> was &lt; <c>1</c>.</item>
        </list>
        <p>The written string is always <c>NULL</c>-terminated, unless buffer
          <c>size</c> is &lt; <c>1</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_get_tuple(ErlNifEnv* env, ERL_NIF_TERM
        term, int* arity, const ERL_NIF_TERM** array)</nametext></name>
      <fsummary>Inspect the elements of a tuple.</fsummary>
      <desc>
        <p>If <c>term</c> is a tuple, this function sets <c>*array</c> to point
          to an array containing the elements of the tuple, and sets
          <c>*arity</c> to the number of elements. Notice that the array
          is read-only and <c>(*array)[N-1]</c> is the Nth element of
          the tuple. <c>*array</c> is undefined if the arity of the tuple
          is zero.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if <c>term</c> is
          not a tuple.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_get_uint(ErlNifEnv* env, ERL_NIF_TERM
        term, unsigned int* ip)</nametext></name>
      <fsummary>Read an unsigned integer term.</fsummary>
      <desc>
        <p>Sets <c>*ip</c> to the unsigned integer value of <c>term</c>.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if <c>term</c> is
          not an unsigned integer or is outside the bounds of type
          <c>unsigned int</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_get_uint64(ErlNifEnv* env,
        ERL_NIF_TERM term, ErlNifUInt64* ip)</nametext></name>
      <fsummary>Read an unsigned 64-bit integer term.</fsummary>
      <desc>
        <p>Sets <c>*ip</c> to the unsigned integer value of <c>term</c>.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if <c>term</c> is
          not an unsigned integer or is outside the bounds of an unsigned
          64-bit integer.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_get_ulong(ErlNifEnv* env, ERL_NIF_TERM
        term, unsigned long* ip)</nametext></name>
      <fsummary>Read an unsigned integer term.</fsummary>
      <desc>
        <p>Sets <c>*ip</c> to the unsigned long integer value of
          <c>term</c>.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if <c>term</c> is
          not an unsigned integer or is outside the bounds of type
          <c>unsigned long</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_getenv(const char* key, char* value,
        size_t *value_size)</nametext></name>
      <fsummary>Get the value of an environment variable.</fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_getenv">
          <c>erl_drv_getenv</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_has_pending_exception(ErlNifEnv* env,
        ERL_NIF_TERM* reason)</nametext></name>
      <fsummary>Check if an exception has been raised.</fsummary>
      <desc>
        <p>Returns <c>true</c> if a pending exception is associated with the
          environment <c>env</c>. If <c>reason</c> is a <c>NULL</c> pointer,
          ignore it. Otherwise, if a pending exception associated with
          <c>env</c> exists, set <c>*reason</c> to the value of the exception
	  term. For example, if <seealso marker="#enif_make_badarg">
          <c>enif_make_badarg</c></seealso> is called to set a pending
          <c>badarg</c> exception, a later call to
          <c>enif_has_pending_exception(env, &amp;reason)</c> sets
          <c>*reason</c> to the atom <c>badarg</c>, then return <c>true</c>.</p>
        <p>See also <seealso marker="#enif_make_badarg">
          <c>enif_make_badarg</c></seealso> and
          <seealso marker="#enif_raise_exception">
          <c>enif_raise_exception</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name>
        <ret>ErlNifUInt64</ret>
        <nametext>enif_hash(ErlNifHash type, ERL_NIF_TERM term, ErlNifUInt64 salt)</nametext>
      </name>
      <fsummary>Hash terms.</fsummary>
      <desc>
        <p>Hashes <c>term</c> according to the specified
          <seealso marker="#ErlNifHash"><c>ErlNifHash</c></seealso> <c>type</c>.</p>
        <p>Ranges of taken salt (if any) and returned value depend on the hash type.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_inspect_binary(ErlNifEnv* env,
        ERL_NIF_TERM bin_term, ErlNifBinary* bin)</nametext></name>
      <fsummary>Inspect the content of a binary.</fsummary>
      <desc>
        <p>Initializes the structure pointed to by <c>bin</c> with information
          about binary term <c>bin_term</c>.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if <c>bin_term</c>
          is not a binary.</p>
      </desc> 
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_inspect_iolist_as_binary(ErlNifEnv*
        env, ERL_NIF_TERM term, ErlNifBinary* bin)</nametext></name>
      <fsummary>Inspect the content of an iolist.</fsummary>
      <desc>
        <p>Initializes the structure pointed to by <c>bin</c> with a
          continuous buffer with the same byte content as <c>iolist</c>. As
          with <c>inspect_binary</c>, the data pointed to by <c>bin</c> is
          transient and does not need to be released.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if <c>iolist</c> is
          not an iolist.</p>
      </desc> 
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_inspect_iovec(ErlNifEnv*
        env, size_t max_elements, ERL_NIF_TERM iovec_term, ERL_NIF_TERM* tail,
        ErlNifIOVec** iovec)</nametext></name>
      <fsummary>Inspect a list of binaries as an ErlNifIOVec.</fsummary>
      <desc>
        <p>Fills <c>iovec</c> with the list of binaries provided in
           <c>iovec_term</c>. The number of elements handled in the call is
           limited to <c>max_elements</c>, and <c>tail</c> is set to the
           remainder of the list. Note that the output may be longer than
           <c>max_elements</c> on some platforms.
        </p>
        <p>To create a list of binaries from an arbitrary iolist, use
           <seealso marker="erts:erlang#iolist_to_iovec/1">
           <c>erlang:iolist_to_iovec/1</c></seealso>.</p>
        <p>When calling this function, <c>iovec</c> should contain a pointer to
           <c>NULL</c> or a ErlNifIOVec structure that should be used if
           possible. e.g.
        </p>
        <code type="none">
/* Don't use a pre-allocated structure */
ErlNifIOVec *iovec = NULL;
enif_inspect_iovec(env, max_elements, term, &amp;tail, &amp;iovec);

/* Use a stack-allocated vector as an optimization for vectors with few elements */
ErlNifIOVec vec, *iovec = &amp;vec;
enif_inspect_iovec(env, max_elements, term, &amp;tail, &amp;iovec);
</code>
        <p>The contents of the <c>iovec</c> is valid until the called nif
           function returns. If the <c>iovec</c> should be valid after the nif
           call returns, it is possible to call this function with a
           <c>NULL</c> environment. If no environment is given the <c>iovec</c>
           owns the data in the vector and it has to be explicitly freed using
           <seealso marker="#enif_free_iovec"><c>enif_free_iovec</c>
           </seealso>.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if <c>iovec_term</c>
           not an iovec.</p>
      </desc>
    </func>

    <func>
      <name><ret>ErlNifIOQueue *</ret>
      <nametext>enif_ioq_create(ErlNifIOQueueOpts opts)</nametext></name>
      <fsummary>Create a new IO Queue</fsummary>
      <desc>
        <p>Create a new I/O Queue that can be used to store data.
          <c>opts</c> has to be set to <c>ERL_NIF_IOQ_NORMAL</c>.
        </p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret>
      <nametext>enif_ioq_destroy(ErlNifIOQueue *q)</nametext></name>
      <fsummary>Destroy an IO Queue and free it's content</fsummary>
      <desc>
        <p>Destroy the I/O queue and free all of it's contents</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
      <nametext>enif_ioq_deq(ErlNifIOQueue *q, size_t count, size_t *size)</nametext></name>
      <fsummary>Dequeue count bytes from the IO Queue</fsummary>
      <desc>
        <p>Dequeue <c>count</c> bytes from the I/O queue.
          If <c>size</c> is not <c>NULL</c>, the new size of the queue
          is placed there.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if the I/O does
          not contain <c>count</c> bytes. On failure the queue is left un-altered.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
      <nametext>enif_ioq_enq_binary(ErlNifIOQueue *q, ErlNifBinary *bin, size_t skip)</nametext></name>
      <fsummary>Enqueue the binary into the IO Queue</fsummary>
      <desc>
        <p>Enqueue the <c>bin</c> into <c>q</c> skipping the first <c>skip</c> bytes.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if <c>skip</c> is greater
        than the size of <c>bin</c>. Any ownership of the binary data is transferred
        to the queue and <c>bin</c> is to be considered read-only for the rest of the NIF
        call and then as released.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
      <nametext>enif_ioq_enqv(ErlNifIOQueue *q, ErlNifIOVec *iovec, size_t skip)</nametext></name>
      <fsummary>Enqueue the iovec into the IO Queue</fsummary>
      <desc>
        <p>Enqueue the <c>iovec</c> into <c>q</c> skipping the first <c>skip</c> bytes.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if <c>skip</c> is greater
        than the size of <c>iovec</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>SysIOVec *</ret>
      <nametext>enif_ioq_peek(ErlNifIOQueue *q, int *iovlen)</nametext></name>
      <fsummary>Peek inside the IO Queue</fsummary>
      <desc>
        <p>Get the I/O queue as a pointer to an array of <c>SysIOVec</c>s.
           It also returns the number of elements in <c>iovlen</c>.</p>
        <p>Nothing is removed from the queue by this function, that must be done
          with <seealso marker="#enif_ioq_deq"><c>enif_ioq_deq</c></seealso>.</p>
         <p>The returned array is suitable to use with the Unix system
          call <c>writev</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
      <nametext>enif_ioq_peek_head(ErlNifEnv *env, ErlNifIOQueue *q, size_t *size, ERL_NIF_TERM *bin_term)</nametext></name>
      <fsummary>Peek the head of the IO Queue.</fsummary>
      <desc>
        <p>Get the head of the IO Queue as a binary term.</p>
        <p>If <c>size</c> is not <c>NULL</c>, the size of the head is placed
           there.</p>
        <p>Nothing is removed from the queue by this function, that must be done
           with <seealso marker="#enif_ioq_deq"><c>enif_ioq_deq</c></seealso>.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if the queue is
           empty.</p>
      </desc>
    </func>

    <func>
      <name><ret>size_t</ret>
      <nametext>enif_ioq_size(ErlNifIOQueue *q)</nametext></name>
      <fsummary>Get the current size of the IO Queue</fsummary>
      <desc>
        <p>Get the size of <c>q</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_is_atom(ErlNifEnv* env, ERL_NIF_TERM term)</nametext>
      </name>
      <fsummary>Determine if a term is an atom.</fsummary>
      <desc>
        <p>Returns <c>true</c> if <c>term</c> is an atom.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_is_binary(ErlNifEnv* env, ERL_NIF_TERM term)</nametext>
      </name>
      <fsummary>Determine if a term is a binary.</fsummary>
      <desc>
        <p>Returns <c>true</c> if <c>term</c> is a binary.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_is_current_process_alive(ErlNifEnv* env)</nametext>
      </name>
      <fsummary>Determine if currently executing process is alive.</fsummary>
      <desc>
        <p>Returns <c>true</c> if the currently executing process is currently
          alive, otherwise <c>false</c>.</p>
        <p>This function can only be used from a NIF-calling thread, and with
          an environment corresponding to currently executing processes.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_is_empty_list(ErlNifEnv* env,
        ERL_NIF_TERM term)</nametext></name>
      <fsummary>Determine if a term is an empty list.</fsummary>
      <desc>
        <p>Returns <c>true</c> if <c>term</c> is an empty list.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_is_exception(ErlNifEnv* env,
        ERL_NIF_TERM term)</nametext></name>
      <fsummary>Determine if a term is an exception.</fsummary>
      <desc><marker id="enif_is_exception"/>
      <p>Return true if <c>term</c> is an exception.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_is_fun(ErlNifEnv* env, ERL_NIF_TERM
        term)</nametext></name>
      <fsummary>Determine if a term is a fun.</fsummary>
      <desc>
        <p>Returns <c>true</c> if <c>term</c> is a fun.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_is_identical(ERL_NIF_TERM lhs,
        ERL_NIF_TERM rhs)</nametext></name>
      <fsummary>Erlang operator =:=.</fsummary>
      <desc>
        <p>Returns <c>true</c> if the two terms are identical. Corresponds to
          the Erlang operators <c>=:=</c> and <c>=/=</c>.</p>
      </desc> 
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_is_list(ErlNifEnv* env, ERL_NIF_TERM term)</nametext>
      </name>
      <fsummary>Determine if a term is a list.</fsummary>
      <desc>
        <p>Returns <c>true</c> if <c>term</c> is a list.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_is_map(ErlNifEnv* env, ERL_NIF_TERM
        term)</nametext></name>
      <fsummary>Determine if a term is a map.</fsummary>
      <desc>
        <p>Returns <c>true</c> if <c>term</c> is a map, otherwise
          <c>false</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_is_number(ErlNifEnv* env, ERL_NIF_TERM
        term)</nametext></name>
      <fsummary>Determine if a term is a number (integer or float).</fsummary>
      <desc>
        <p>Returns <c>true</c> if <c>term</c> is a number.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_is_pid(ErlNifEnv* env, ERL_NIF_TERM term)</nametext>
      </name>
      <fsummary>Determine if a term is a pid.</fsummary>
      <desc>
        <p>Returns <c>true</c> if <c>term</c> is a pid.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_is_port(ErlNifEnv* env, ERL_NIF_TERM term)</nametext>
      </name>
      <fsummary>Determine if a term is a port.</fsummary>
      <desc>
        <p>Returns <c>true</c> if <c>term</c> is a port.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_is_port_alive(ErlNifEnv* env,
        ErlNifPort *port_id)</nametext></name>
      <fsummary>Determine if a local port is alive.</fsummary>
      <desc>
        <p>Returns <c>true</c> if <c>port_id</c> is alive.</p>
        <p>This function is only thread-safe when the emulator with SMP support
          is used. It can only be used in a non-SMP emulator from a NIF-calling
          thread.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_is_process_alive(ErlNifEnv* env,
        ErlNifPid *pid)</nametext></name>
      <fsummary>Determine if a local process is alive.</fsummary>
      <desc>
        <p>Returns <c>true</c> if <c>pid</c> is alive.</p>
        <p>This function is only thread-safe when the emulator with SMP support
          is used. It can only be used in a non-SMP emulator from a NIF-calling
          thread.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_is_ref(ErlNifEnv* env, ERL_NIF_TERM term)</nametext>
      </name>
      <fsummary>Determine if a term is a reference.</fsummary>
      <desc>
        <p>Returns <c>true</c> if <c>term</c> is a reference.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_is_tuple(ErlNifEnv* env, ERL_NIF_TERM term)</nametext>
      </name>
      <fsummary>Determine if a term is a tuple.</fsummary>
      <desc>
        <p>Returns <c>true</c> if <c>term</c> is a tuple.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_keep_resource(void* obj)</nametext>
      </name>
      <fsummary>Add a reference to a resource object.</fsummary>
      <desc>
        <p>Adds a reference to resource object <c>obj</c> obtained from
          <seealso marker="#enif_alloc_resource">
          <c>enif_alloc_resource</c></seealso>. Each call to
          <c>enif_keep_resource</c> for an object must be balanced by a call to
          <seealso marker="#enif_release_resource">
          <c>enif_release_resource</c></seealso>
          before the object is destructed.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret>
        <nametext>enif_make_atom(ErlNifEnv* env, const char* name)</nametext>
      </name>
      <fsummary>Create an atom term.</fsummary>
      <desc>
        <p>Creates an atom term from the <c>NULL</c>-terminated C-string
          <c>name</c> with ISO Latin-1 encoding. If the length of <c>name</c>
          exceeds the maximum length allowed for an atom (255 characters),
          <c>enif_make_atom</c> invokes <seealso marker="#enif_make_badarg">
          <c>enif_make_badarg</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_atom_len(ErlNifEnv* env,
        const char* name, size_t len)</nametext></name>
      <fsummary>Create an atom term.</fsummary>
      <desc>
        <p>Create an atom term from the string <c>name</c> with length
          <c>len</c>. <c>NULL</c> characters are treated as any other
          characters. If <c>len</c> exceeds the maximum length
          allowed for an atom (255 characters), <c>enif_make_atom</c> invokes
          <seealso marker="#enif_make_badarg">
          <c>enif_make_badarg</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret>
        <nametext>enif_make_badarg(ErlNifEnv* env)</nametext></name>
      <fsummary>Make a badarg exception.</fsummary>
      <desc>
        <p>Makes a <c>badarg</c> exception to be returned from a NIF, and
          associates it with environment <c>env</c>. Once a NIF or any function
          it calls invokes <c>enif_make_badarg</c>, the runtime ensures that a
          <c>badarg</c> exception is raised when the NIF returns, even if the
          NIF attempts to return a non-exception term instead.</p>
        <p>The return value from <c>enif_make_badarg</c> can be used only as
          the return value from the NIF that invoked it (directly or indirectly)
          or be passed to <seealso marker="#enif_is_exception">
          <c>enif_is_exception</c></seealso>, but not to any other NIF API
          function.</p>
        <p>See also <seealso marker="#enif_has_pending_exception">
          <c>enif_has_pending_exception</c></seealso> and
          <seealso marker="#enif_raise_exception">
          <c>enif_raise_exception</c></seealso>.</p>
        <note>
          <p>Before ERTS 7.0 (Erlang/OTP 18), the return value
            from <c>enif_make_badarg</c> had to be returned from the NIF. This
            requirement is now lifted as the return value from the NIF is
            ignored if <c>enif_make_badarg</c> has been invoked.</p>
        </note>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret>
        <nametext>enif_make_binary(ErlNifEnv* env, ErlNifBinary* bin)</nametext>
      </name>
      <fsummary>Make a binary term.</fsummary>
      <desc>
        <p>Makes a binary term from <c>bin</c>. Any ownership of
          the binary data is transferred to the created term and
          <c>bin</c> is to be considered read-only for the rest of the NIF
          call and then as released.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_copy(ErlNifEnv* dst_env,
        ERL_NIF_TERM src_term)</nametext></name>
      <fsummary>Make a copy of a term.</fsummary>
      <desc>
        <p>Makes a copy of term <c>src_term</c>. The copy is created in
          environment <c>dst_env</c>. The source term can be located in any
          environment.</p>
      </desc>        
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret>
        <nametext>enif_make_double(ErlNifEnv* env, double d)</nametext></name>
      <fsummary>Create a floating-point term.</fsummary>
      <desc>
        <p>Creates a floating-point term from a <c>double</c>. If argument
          <c>double</c> is not finite or is NaN, <c>enif_make_double</c>
          invokes <seealso marker="#enif_make_badarg">
          <c>enif_make_badarg</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_make_existing_atom(ErlNifEnv* env,
        const char* name, ERL_NIF_TERM* atom, ErlNifCharEncoding
        encode)</nametext></name>
      <fsummary>Create an existing atom term.</fsummary>
      <desc>
        <p>Tries to create the term of an already existing atom from
          the <c>NULL</c>-terminated C-string <c>name</c> with encoding
          <seealso marker="#ErlNifCharEncoding">encode</seealso>.</p>
        <p>If the atom already exists, this function stores the term in
          <c>*atom</c> and returns <c>true</c>, otherwise <c>false</c>.
          Also returns <c>false</c> if the length of <c>name</c> exceeds the
          maximum length allowed for an atom (255 characters).</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_make_existing_atom_len(ErlNifEnv* env,
        const char* name, size_t len, ERL_NIF_TERM* atom, ErlNifCharEncoding
        encoding)</nametext></name>
      <fsummary>Create an existing atom term.</fsummary>
      <desc>
        <p>Tries to create the term of an already existing atom from the
          string <c>name</c> with length <c>len</c> and encoding
          <seealso marker="#ErlNifCharEncoding">encode</seealso>. <c>NULL</c>
          characters are treated as any other characters.</p>
        <p>If the atom already exists, this function stores the term in
          <c>*atom</c> and returns <c>true</c>, otherwise <c>false</c>.
          Also returns <c>false</c> if <c>len</c> exceeds the maximum length
          allowed for an atom (255 characters).</p>
      </desc>
    </func>

    <func><name><ret>ERL_NIF_TERM</ret>
      <nametext>enif_make_int(ErlNifEnv* env, int i)</nametext></name>
      <fsummary>Create an integer term.</fsummary>
      <desc>
        <p>Creates an integer term.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret>
        <nametext>enif_make_int64(ErlNifEnv* env, ErlNifSInt64 i)</nametext>
      </name>
      <fsummary>Create an integer term.</fsummary>
      <desc>
        <p>Creates an integer term from a signed 64-bit integer.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret>
        <nametext>enif_make_list(ErlNifEnv* env, unsigned cnt, ...)</nametext>
      </name>
      <fsummary>Create a list term.</fsummary>
      <desc>
        <p>Creates an ordinary list term of length <c>cnt</c>. Expects
          <c>cnt</c> number of arguments (after <c>cnt</c>) of type
          <c>ERL_NIF_TERM</c> as the elements of the list.</p>
        <p>Returns an empty list if <c>cnt</c> is 0.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret>
        <nametext>enif_make_list1(ErlNifEnv* env, ERL_NIF_TERM e1)</nametext>
      </name>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_list2(ErlNifEnv* env,
        ERL_NIF_TERM e1, ERL_NIF_TERM e2)</nametext></name>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_list3(ErlNifEnv* env,
        ERL_NIF_TERM e1, ERL_NIF_TERM e2, ERL_NIF_TERM e3)</nametext></name>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_list4(ErlNifEnv* env,
        ERL_NIF_TERM e1, ..., ERL_NIF_TERM e4)</nametext></name>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_list5(ErlNifEnv* env,
        ERL_NIF_TERM e1, ..., ERL_NIF_TERM e5)</nametext></name>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_list6(ErlNifEnv* env,
        ERL_NIF_TERM e1, ..., ERL_NIF_TERM e6)</nametext></name>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_list7(ErlNifEnv* env,
        ERL_NIF_TERM e1, ..., ERL_NIF_TERM e7)</nametext></name>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_list8(ErlNifEnv* env,
        ERL_NIF_TERM e1, ..., ERL_NIF_TERM e8)</nametext></name>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_list9(ErlNifEnv* env,
        ERL_NIF_TERM e1, ..., ERL_NIF_TERM e9)</nametext></name>
      <fsummary>Create a list term.</fsummary>
      <desc>
        <p>Creates an ordinary list term with length indicated by the
          function name. Prefer these functions (macros) over the variadic
          <c>enif_make_list</c> to get a compile-time error if the number of
          arguments does not match.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_list_cell(ErlNifEnv*
        env, ERL_NIF_TERM head, ERL_NIF_TERM tail)</nametext></name>
      <fsummary>Create a list cell.</fsummary>
      <desc>
        <p>Creates a list cell <c>[head | tail]</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret>
        <nametext>enif_make_list_from_array(ErlNifEnv* env, const ERL_NIF_TERM
          arr[], unsigned cnt)</nametext></name>
      <fsummary>Create a list term from an array.</fsummary>
      <desc>
        <p>Creates an ordinary list containing the elements of array <c>arr</c>
          of length <c>cnt</c>.</p>
        <p>Returns an empty list if <c>cnt</c> is 0.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret>
        <nametext>enif_make_long(ErlNifEnv* env, long int i)</nametext></name>
      <fsummary>Create an integer term from a long int.</fsummary>
      <desc>
        <p>Creates an integer term from a <c>long int</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_make_map_put(ErlNifEnv* env,
        ERL_NIF_TERM map_in, ERL_NIF_TERM key, ERL_NIF_TERM value,
        ERL_NIF_TERM* map_out)</nametext></name>
      <fsummary>Insert key-value pair in map.</fsummary>
      <desc>
        <p>Makes a copy of map <c>map_in</c> and inserts <c>key</c> with
          <c>value</c>. If <c>key</c> already exists in <c>map_in</c>, the old
          associated value is replaced by <c>value</c>.</p>
        <p>If successful, this function sets <c>*map_out</c> to the new map and
          returns <c>true</c>. Returns <c>false</c> if <c>map_in</c> is not a
          map.</p>
        <p>The <c>map_in</c> term must belong to environment <c>env</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_make_map_remove(ErlNifEnv* env,
        ERL_NIF_TERM map_in, ERL_NIF_TERM key, ERL_NIF_TERM* map_out)</nametext>
      </name>
      <fsummary>Remove key from map.</fsummary>
      <desc>
        <p>If map <c>map_in</c> contains <c>key</c>, this function makes a copy
          of <c>map_in</c> in <c>*map_out</c>, and removes <c>key</c> and the
          associated value. If map <c>map_in</c> does not contain <c>key</c>,
          <c>*map_out</c> is set to <c>map_in</c>.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if <c>map_in</c> is
          not a map.</p>
        <p>The <c>map_in</c> term must belong to environment <c>env</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_make_map_update(ErlNifEnv* env,
        ERL_NIF_TERM map_in, ERL_NIF_TERM key, ERL_NIF_TERM new_value,
        ERL_NIF_TERM* map_out)</nametext></name>
      <fsummary>Replace value for key in map.</fsummary>
      <desc>
        <p>Makes a copy of map <c>map_in</c> and replace the old associated
          value for <c>key</c> with <c>new_value</c>.</p>
        <p>If successful, this function sets <c>*map_out</c> to the new map and
          returns <c>true</c>. Returns <c>false</c> if <c>map_in</c> is not a
          map or if it does not contain <c>key</c>.</p>
        <p>The <c>map_in</c> term must belong to environment <c>env</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_make_map_from_arrays(ErlNifEnv* env, ERL_NIF_TERM keys[],
        ERL_NIF_TERM values[], size_t cnt, ERL_NIF_TERM *map_out)</nametext></name>
      <fsummary>Make map term from the given keys and values.</fsummary>
      <desc>
        <p>Makes a map term from the given keys and values.</p>
        <p>If successful, this function sets <c>*map_out</c> to the new map and
          returns <c>true</c>. Returns <c>false</c> there are any duplicate
          keys.</p>
        <p>All keys and values must belong to <c>env</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>unsigned char *</ret><nametext>enif_make_new_binary(ErlNifEnv*
        env, size_t size, ERL_NIF_TERM* termp)</nametext></name>
      <fsummary>Allocate and create a new binary term.</fsummary>
      <desc>
        <p>Allocates a binary of size <c>size</c> bytes and creates an owning
          term. The binary data is mutable until the calling NIF returns.
          This is a quick way to create a new binary without having to use
          <seealso marker="#ErlNifBinary"><c>ErlNifBinary</c></seealso>.
          The drawbacks are that the binary cannot be kept between NIF calls
          and it cannot be reallocated.</p>
        <p>Returns a pointer to the raw binary data and sets
          <c>*termp</c> to the binary term.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret>
        <nametext>enif_make_new_map(ErlNifEnv* env)</nametext></name>
      <fsummary>Make an empty map term.</fsummary>
      <desc>
        <p>Makes an empty map term.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret>
        <nametext>enif_make_pid(ErlNifEnv* env, const ErlNifPid* pid)</nametext>
      </name>
      <fsummary>Make a pid term.</fsummary>
      <desc>
        <p>Makes a pid term from <c>*pid</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret>
        <nametext>enif_make_ref(ErlNifEnv* env)</nametext></name>
      <fsummary>Create a reference.</fsummary>
      <desc>
        <p>Creates a reference like <seealso marker="erlang#make_ref-0">
          <c>erlang:make_ref/0</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret>
        <nametext>enif_make_resource(ErlNifEnv* env, void* obj)</nametext>
      </name>
      <fsummary>Create an opaque handle to a resource object.</fsummary>
      <desc>
        <p>Creates an opaque handle to a memory-managed resource object
          obtained by <seealso marker="#enif_alloc_resource">
          <c>enif_alloc_resource</c></seealso>. No ownership transfer is done,
          as the resource object still needs to be released by
          <seealso marker="#enif_release_resource">
          <c>enif_release_resource</c></seealso>. However, notice that the call
          to <c>enif_release_resource</c> can occur immediately after obtaining
          the term from <c>enif_make_resource</c>, in which case the resource
          object is deallocated when the term is garbage collected. For more
          details, see the <seealso marker="#enif_resource_example">example of
          creating and returning a resource object</seealso> in the User's
          Guide.</p>
	<note>
	  <p>Since ERTS 9.0 (OTP-20.0), resource terms have a defined behavior
	  when compared and serialized through <c>term_to_binary</c> or passed
	  between nodes.</p>
	  <list type="bulleted">
	    <item>
	      <p>Two resource terms will compare equal iff they
	      would yield the same resource object pointer when passed to
	      <seealso marker="#enif_get_resource"><c>enif_get_resource</c></seealso>.</p>
	    </item>
	    <item>
	      <p>A resource term can be serialized with <c>term_to_binary</c> and later
	      be fully recreated if the resource object is still alive when
	      <c>binary_to_term</c> is called. A <em>stale</em> resource term will be
	      returned from <c>binary_to_term</c> if the resource object has
	      been deallocated. <seealso marker="#enif_get_resource"><c>enif_get_resource</c></seealso>
	      will return false for stale resource terms.</p>
	      <p>The same principles of serialization apply when passing
	      resource terms in messages to remote nodes and back again. A
	      resource term will act stale on all nodes except the node where
	      its resource object is still alive in memory.</p>
	    </item>
	  </list>
	  <p>Before ERTS 9.0 (OTP-20.0), all resource terms did
	  compare equal to each other and to empty binaries (<c>&lt;&lt;&gt;&gt;</c>).
	  If serialized, they would be recreated as plain empty binaries.</p>
	</note>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret>
        <nametext>enif_make_resource_binary(ErlNifEnv* env, void* obj, const
        void* data, size_t size)</nametext></name>
      <fsummary>Create a custom binary term.</fsummary>
      <desc>
        <p>Creates a binary term that is memory-managed by a resource object
          <c>obj</c> obtained by <seealso marker="#enif_alloc_resource">
          <c>enif_alloc_resource</c></seealso>. The returned binary term
          consists of <c>size</c> bytes pointed to by <c>data</c>. This raw
          binary data must be kept readable and unchanged until the destructor
          of the resource is called. The binary data can be stored external to
          the resource object, in which case the destructor is responsible
          for releasing the data.</p>
        <p>Several binary terms can be managed by the same resource object. The
          destructor is not called until the last binary is garbage collected.
          This can be useful to return different parts of a larger binary
          buffer.</p>
        <p>As with <seealso marker="#enif_make_resource">
          <c>enif_make_resource</c></seealso>, no ownership transfer is done.
          The resource still needs to be released with
          <seealso marker="#enif_release_resource">
          <c>enif_release_resource</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_make_reverse_list(ErlNifEnv* env, ERL_NIF_TERM list_in,
        ERL_NIF_TERM *list_out)</nametext></name>
      <fsummary>Create the reverse of a list.</fsummary>
      <desc>
        <p>Sets <c>*list_out</c> to the reverse list of the list <c>list_in</c>
          and returns <c>true</c>, or returns <c>false</c> if <c>list_in</c> is
          not a list.</p>
        <p>This function is only to be used on short lists, as a copy is
          created of the list, which is not released until after the NIF
          returns.</p>
        <p>The <c>list_in</c> term must belong to environment <c>env</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_string(ErlNifEnv* env,
        const char* string, ErlNifCharEncoding encoding)</nametext></name>
      <fsummary>Create a string.</fsummary>
      <desc>
        <p>Creates a list containing the characters of the
          <c>NULL</c>-terminated string <c>string</c> with encoding
          <seealso marker="#ErlNifCharEncoding">encoding</seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_string_len(ErlNifEnv*
        env, const char* string, size_t len, ErlNifCharEncoding
        encoding)</nametext></name>
      <fsummary>Create a string.</fsummary>
      <desc>
        <p>Creates a list containing the characters of the string <c>string</c>
          with length <c>len</c> and encoding
          <seealso marker="#ErlNifCharEncoding">encoding</seealso>.
          <c>NULL</c> characters are treated as any other characters.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_sub_binary(ErlNifEnv*
        env, ERL_NIF_TERM bin_term, size_t pos, size_t size)</nametext></name>
      <fsummary>Make a subbinary term.</fsummary>
      <desc>
        <p>Makes a subbinary of binary <c>bin_term</c>, starting at
          zero-based position <c>pos</c> with a length of <c>size</c> bytes.
          <c>bin_term</c> must be a binary or bitstring. <c>pos+size</c> must
          be less or equal to the number of whole bytes in <c>bin_term</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_tuple(ErlNifEnv* env,
        unsigned cnt, ...)</nametext></name>
      <fsummary>Creates a tuple term.</fsummary>
      <desc>
        <p>Creates a tuple term of arity <c>cnt</c>. Expects <c>cnt</c> number
          of arguments (after <c>cnt</c>) of type <c>ERL_NIF_TERM</c> as the
          elements of the tuple.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_tuple1(ErlNifEnv* env,
        ERL_NIF_TERM e1)</nametext></name>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_tuple2(ErlNifEnv* env,
        ERL_NIF_TERM e1, ERL_NIF_TERM e2)</nametext></name>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_tuple3(ErlNifEnv* env,
        ERL_NIF_TERM e1, ERL_NIF_TERM e2, ERL_NIF_TERM e3)</nametext></name>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_tuple4(ErlNifEnv* env,
        ERL_NIF_TERM e1, ..., ERL_NIF_TERM e4)</nametext></name>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_tuple5(ErlNifEnv* env,
        ERL_NIF_TERM e1, ..., ERL_NIF_TERM e5)</nametext></name>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_tuple6(ErlNifEnv* env,
        ERL_NIF_TERM e1, ..., ERL_NIF_TERM e6)</nametext></name>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_tuple7(ErlNifEnv* env,
        ERL_NIF_TERM e1, ..., ERL_NIF_TERM e7)</nametext></name>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_tuple8(ErlNifEnv* env,
        ERL_NIF_TERM e1, ..., ERL_NIF_TERM e8)</nametext></name>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_tuple9(ErlNifEnv* env,
        ERL_NIF_TERM e1, ..., ERL_NIF_TERM e9)</nametext></name>
      <fsummary>Create a tuple term.</fsummary>
      <desc>
        <p>Creates a tuple term with length indicated by the
          function name. Prefer these functions (macros) over the variadic
          <c>enif_make_tuple</c> to get a compile-time error if the number of
          arguments does not match.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret>
        <nametext>enif_make_tuple_from_array(ErlNifEnv* env, const ERL_NIF_TERM
        arr[], unsigned cnt)</nametext></name>
      <fsummary>Create a tuple term from an array.</fsummary>
      <desc>
        <p>Creates a tuple containing the elements of array <c>arr</c>
          of length <c>cnt</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret>
        <nametext>enif_make_uint(ErlNifEnv* env, unsigned int i)</nametext>
      </name>
      <fsummary>Create an unsigned integer term.</fsummary>
      <desc>
        <p>Creates an integer term from an <c>unsigned int</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret>
        <nametext>enif_make_uint64(ErlNifEnv* env, ErlNifUInt64 i)</nametext>
      </name>
      <fsummary>Create an unsigned integer term.</fsummary>
      <desc>
        <p>Creates an integer term from an unsigned 64-bit integer.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret>
        <nametext>enif_make_ulong(ErlNifEnv* env, unsigned long i)</nametext>
      </name>
      <fsummary>Create an integer term from an unsigned long int.</fsummary>
      <desc>
        <p>Creates an integer term from an <c>unsigned long int</c>.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_make_unique_integer(ErlNifEnv
        *env, ErlNifUniqueInteger properties)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Returns a unique integer with the same properties as specified by
          <seealso marker="erlang#unique_integer-1">
          <c>erlang:unique_integer/1</c></seealso>.</p>
        <p><c>env</c> is the environment to create the integer in.</p>
        <p><c>ERL_NIF_UNIQUE_POSITIVE</c> and <c>ERL_NIF_UNIQUE_MONOTONIC</c>
          can be passed as the second argument to change the properties of the
          integer returned. They can be combined by OR:ing the two values
          together.</p>
        <p>See also <seealso marker="#ErlNifUniqueInteger">
          <c>ErlNifUniqueInteger</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_map_iterator_create(ErlNifEnv *env,
        ERL_NIF_TERM map, ErlNifMapIterator *iter, ErlNifMapIteratorEntry
        entry)</nametext></name>
      <fsummary>Create a map iterator.</fsummary>
      <desc>
        <p>Creates an iterator for the map <c>map</c> by initializing the
          structure pointed to by <c>iter</c>. Argument <c>entry</c> determines
          the start position of the iterator: <c>ERL_NIF_MAP_ITERATOR_FIRST</c>
          or <c>ERL_NIF_MAP_ITERATOR_LAST</c>.</p>
        <p>Returns <c>true</c> on success, or false if <c>map</c> is not a
          map.</p>
        <p>A map iterator is only useful during the lifetime of environment
          <c>env</c> that the <c>map</c> belongs to. The iterator must be
          destroyed by calling <seealso marker="#enif_map_iterator_destroy">
          <c>enif_map_iterator_destroy</c></seealso>:</p>
        <code type="none">
ERL_NIF_TERM key, value;
ErlNifMapIterator iter;
enif_map_iterator_create(env, my_map, &amp;iter, ERL_NIF_MAP_ITERATOR_FIRST);

while (enif_map_iterator_get_pair(env, &amp;iter, &amp;key, &amp;value)) {
    do_something(key,value);
    enif_map_iterator_next(env, &amp;iter);
}
enif_map_iterator_destroy(env, &amp;iter);</code>
        <note>
          <p>The key-value pairs of a map have no defined iteration order.
            The only guarantee is that the iteration order of a single map
            instance is preserved during the lifetime of the environment that
            the map belongs to.</p>
        </note>
      </desc>
    </func>

    <func>
      <name><ret>void</ret><nametext>enif_map_iterator_destroy(ErlNifEnv *env,
        ErlNifMapIterator *iter)</nametext></name>
      <fsummary>Destroy a map iterator.</fsummary>
      <desc>
        <p>Destroys a map iterator created by
          <seealso marker="#enif_map_iterator_create">
          <c>enif_map_iterator_create</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_map_iterator_get_pair(ErlNifEnv *env,
        ErlNifMapIterator *iter, ERL_NIF_TERM *key, ERL_NIF_TERM
        *value)</nametext></name>
      <fsummary>Get key and value at current map iterator position.</fsummary>
      <desc>
        <p>Gets key and value terms at the current map iterator position.</p>
        <p>On success, sets <c>*key</c> and <c>*value</c> and returns
          <c>true</c>. Returns <c>false</c> if the iterator is positioned at
          head (before first entry) or tail (beyond last entry).</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_map_iterator_is_head(ErlNifEnv *env,
        ErlNifMapIterator *iter)</nametext></name>
      <fsummary>Check if map iterator is positioned before first.</fsummary>
      <desc>
        <p>Returns <c>true</c> if map iterator <c>iter</c> is positioned
          before the first entry.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_map_iterator_is_tail(ErlNifEnv *env,
        ErlNifMapIterator *iter)</nametext></name>
      <fsummary>Check if map iterator is positioned after last.</fsummary>
      <desc>
        <p>Returns <c>true</c> if map iterator <c>iter</c> is positioned
          after the last entry.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_map_iterator_next(ErlNifEnv *env,
        ErlNifMapIterator *iter)</nametext></name>
      <fsummary>Increment map iterator to point to next entry.</fsummary>
      <desc>
        <p>Increments map iterator to point to the next key-value entry.</p>
        <p>Returns <c>true</c> if the iterator is now positioned at a valid
          key-value entry, or <c>false</c> if the iterator is positioned at
          the tail (beyond the last entry).</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_map_iterator_prev(ErlNifEnv *env,
        ErlNifMapIterator *iter)</nametext></name>
      <fsummary>Decrement map iterator to point to previous entry.</fsummary>
      <desc>
        <p>Decrements map iterator to point to the previous key-value entry.</p>
        <p>Returns <c>true</c> if the iterator is now positioned at a valid
          key-value entry, or <c>false</c> if the iterator is positioned at
          the head (before the first entry).</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_monitor_process(ErlNifEnv* env, void* obj,
      const ErlNifPid* target_pid, ErlNifMonitor* mon)</nametext></name>
      <fsummary>Monitor a process from a resource.</fsummary>
      <desc>
        <marker id="enif_monitor_process"></marker>
        <p>Starts monitoring a process from a resource. When a process is
          monitored, a process exit results in a call to the provided
          <seealso marker="#ErlNifResourceDown">
          <c>down</c></seealso> callback associated with the resource type.</p>
	<p>Argument <c>obj</c> is pointer to the resource to hold the monitor and
	<c>*target_pid</c> identifies the local process to be monitored.</p>
	<p>If <c>mon</c> is not <c>NULL</c>, a successful call stores the
	identity of the monitor	in the
	<seealso marker="#ErlNifMonitor"><c>ErlNifMonitor</c></seealso>
	struct pointed to by <c>mon</c>. This identifier is used to refer to the
	monitor for later removal with
	<seealso marker="#enif_demonitor_process"><c>enif_demonitor_process</c></seealso>
	or compare with
	<seealso marker="#enif_compare_monitors"><c>enif_compare_monitors</c></seealso>.
	A monitor is automatically removed when it triggers or when
	the resource is deallocated.</p>
        <p>Returns <c>0</c> on success, &lt; 0 if no <c>down</c> callback is
          provided, and &gt; 0 if the process is no longer alive.</p>
        <p>This function is only thread-safe when the emulator with SMP support
          is used. It can only be used in a non-SMP emulator from a NIF-calling
          thread.</p>
      </desc>
    </func>

    <func>
      <name><ret>ErlNifTime</ret>
        <nametext>enif_monotonic_time(ErlNifTimeUnit time_unit)</nametext>
      </name>
      <fsummary>Get Erlang monotonic time.</fsummary>
      <desc>
        <marker id="enif_monotonic_time"></marker>
        <p>Returns the current
          <seealso marker="time_correction#Erlang_Monotonic_Time">
          Erlang monotonic time</seealso>. Notice that it is not uncommon with
          negative values.</p>
        <p><c>time_unit</c> is the time unit of the returned value.</p>
        <p>Returns <c>ERL_NIF_TIME_ERROR</c> if called with an invalid time
          unit argument, or if called from a thread that is not a scheduler
          thread.</p>
        <p>See also <seealso marker="#ErlNifTime"><c>ErlNifTime</c></seealso>
          and <seealso marker="#ErlNifTimeUnit"><c>ErlNifTimeUnit</c></seealso>.
        </p>
      </desc>
    </func>

    <func>
      <name><ret>ErlNifMutex *</ret>
        <nametext>enif_mutex_create(char *name)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_mutex_create">
          <c>erl_drv_mutex_create</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret>
        <nametext>enif_mutex_destroy(ErlNifMutex *mtx)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_mutex_destroy">
          <c>erl_drv_mutex_destroy</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret>
        <nametext>enif_mutex_lock(ErlNifMutex *mtx)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_mutex_lock">
          <c>erl_drv_mutex_lock</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>char*</ret>
        <nametext>enif_mutex_name(ErlNifMutex* mtx)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_mutex_name">
          <c>erl_drv_mutex_name</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
      <nametext>enif_mutex_trylock(ErlNifMutex *mtx)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_mutex_trylock">
          <c>erl_drv_mutex_trylock</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret>
        <nametext>enif_mutex_unlock(ErlNifMutex *mtx)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_mutex_unlock">
          <c>erl_drv_mutex_unlock</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret>
        <nametext>enif_now_time(ErlNifEnv *env)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Returns an <seealso marker="erlang#now-0">
          <c>erlang:now()</c></seealso> time stamp.</p>
        <p><em>This function is deprecated.</em></p>
      </desc>
    </func>

    <func>
      <name><ret>ErlNifResourceType *</ret>
        <nametext>enif_open_resource_type(ErlNifEnv* env, const char*
        module_str, const char* name, ErlNifResourceDtor* dtor,
        ErlNifResourceFlags flags, ErlNifResourceFlags* tried)</nametext>
      </name>
      <fsummary>Create or takeover a resource type.</fsummary>
      <desc>
        <p>Creates or takes over a resource type identified by the string
          <c>name</c> and gives it the destructor function pointed to by
          <seealso marker="#ErlNifResourceDtor"><c>dtor</c></seealso>.
          Argument <c>flags</c> can have the following values:</p>
        <taglist>
          <tag><c>ERL_NIF_RT_CREATE</c></tag>
          <item>Creates a new resource type that does not already exist.</item>
          <tag><c>ERL_NIF_RT_TAKEOVER</c></tag>
          <item>Opens an existing resource type and takes over ownership of all
            its instances. The supplied destructor <c>dtor</c> is called both
            for existing instances and new instances not yet created by the
            calling NIF library.</item>
        </taglist>
        <p>The two flag values can be combined with bitwise OR. The resource
          type name is local to the calling module. Argument <c>module_str</c>
          is not (yet) used and must be <c>NULL</c>. <c>dtor</c> can be
          <c>NULL</c> if no destructor is needed.</p>
        <p>On success, the function returns a pointer to the resource type and
          <c>*tried</c> is set to either <c>ERL_NIF_RT_CREATE</c> or
          <c>ERL_NIF_RT_TAKEOVER</c> to indicate what was done. On failure,
          returns <c>NULL</c> and sets <c>*tried</c> to <c>flags</c>.
          It is allowed to set <c>tried</c> to <c>NULL</c>.</p>
        <p>Notice that <c>enif_open_resource_type</c> is only allowed to be
          called in the two callbacks
          <seealso marker="#load"><c>load</c></seealso> and
          <seealso marker="#upgrade"><c>upgrade</c></seealso>.</p>
	<p>See also <seealso marker="#enif_open_resource_type_x">
	  <c>enif_open_resource_type_x</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>ErlNifResourceType *</ret>
        <nametext>enif_open_resource_type_x(ErlNifEnv* env, const char* name,
	const ErlNifResourceTypeInit* init,
        ErlNifResourceFlags flags, ErlNifResourceFlags* tried)</nametext>
      </name>
      <fsummary>Create or takeover a resource type.</fsummary>
      <desc>
	<p>Same as <seealso marker="#enif_open_resource_type"><c>enif_open_resource_type</c></seealso>
	  except it accepts additional callback functions for resource types that are
	  used together with <seealso marker="#enif_select"><c>enif_select</c></seealso>
	  and <seealso marker="#enif_monitor_process"><c>enif_monitor_process</c></seealso>.</p>
	<p>Argument <c>init</c> is a pointer to an
	  <seealso marker="#ErlNifResourceTypeInit"><c>ErlNifResourceTypeInit</c></seealso>
	  structure that contains the function pointers for destructor, down and stop callbacks
	  for the resource type.</p>
      </desc>
    </func>

    <func>
     <name><ret>int</ret><nametext>enif_port_command(ErlNifEnv* env, const
       ErlNifPort* to_port, ErlNifEnv *msg_env, ERL_NIF_TERM msg)</nametext>
     </name>
     <fsummary>Send a port_command to to_port.</fsummary>
     <desc>
       <p>Works as <seealso marker="erlang#port_command-2">
         <c>erlang:port_command/2</c></seealso>,
         except that it is always completely asynchronous.</p>
       <taglist>
         <tag><c>env</c></tag>
         <item>The environment of the calling process. Must not be
           <c>NULL</c>.</item>
         <tag><c>*to_port</c></tag>
         <item>The port ID of the receiving port. The port ID is to refer to a
           port on the local node.</item>
         <tag><c>msg_env</c></tag>
         <item>The environment of the message term. Can be a process-independent
           environment allocated with <seealso marker="#enif_alloc_env">
           <c>enif_alloc_env</c></seealso> or <c>NULL</c>.</item>
         <tag><c>msg</c></tag>
         <item>The message term to send. The same limitations apply as on the
           payload to <seealso marker="erlang#port_command-2">
           <c>erlang:port_command/2</c></seealso>.</item>
       </taglist>
       <p>Using a <c>msg_env</c> of <c>NULL</c> is an optimization, which
         groups together calls to <c>enif_alloc_env</c>, <c>enif_make_copy</c>,
         <c>enif_port_command</c>, and <c>enif_free_env</c> into one call.
         This optimization is only useful when a majority of the terms are to
         be copied from <c>env</c> to <c>msg_env</c>.</p>
       <p>Returns <c>true</c> if the command is successfully sent. Returns
         <c>false</c> if the command fails, for example:</p>
       <list type="bulleted">
         <item><c>*to_port</c> does not refer to a local port.</item>
         <item>The currently executing process (that is, the sender) is not
           alive.</item>
         <item><c>msg</c> is invalid.</item>
       </list>
       <p>See also <seealso marker="#enif_get_local_port">
         <c>enif_get_local_port</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void *</ret>
        <nametext>enif_priv_data(ErlNifEnv* env)</nametext></name>
      <fsummary>Get the private data of a NIF library.</fsummary>
      <desc>
        <p>Returns the pointer to the private data that was set by
          <seealso marker="#load"><c>load</c></seealso> or
          <seealso marker="#upgrade"><c>upgrade</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_raise_exception(ErlNifEnv*
        env, ERL_NIF_TERM reason)</nametext></name>
      <fsummary>Raise a NIF error exception.</fsummary>
      <desc>
        <p>Creates an error exception with the term <c>reason</c> to be
          returned from a NIF, and associates it with environment <c>env</c>.
          Once a NIF or any function it calls invokes
          <c>enif_raise_exception</c>, the runtime ensures that the exception
          it creates is raised when the NIF returns, even if the NIF attempts
          to return a non-exception term instead.</p>
        <p>The return value from <c>enif_raise_exception</c> can only be used
          as the return value from the NIF that invoked it (directly or
          indirectly) or be passed to <seealso marker="#enif_is_exception">
          <c>enif_is_exception</c></seealso>, but not to any other NIF API
          function.</p>
        <p>See also <seealso marker="#enif_has_pending_exception">
          <c>enif_has_pending_exception</c></seealso> and
          <seealso marker="#enif_make_badarg">
          <c>enif_make_badarg</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void *</ret>
        <nametext>enif_realloc(void* ptr, size_t size)</nametext></name>
      <fsummary>Reallocate dynamic memory.</fsummary>
      <desc>
        <p>Reallocates memory allocated by
          <seealso marker="#enif_alloc"><c>enif_alloc</c></seealso> to
          <c>size</c> bytes.</p>
        <p>Returns <c>NULL</c> if the reallocation fails.</p>
        <p>The returned pointer is suitably aligned for any built-in type that
          fit in the allocated memory.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_realloc_binary(ErlNifBinary* bin, size_t size)</nametext>
      </name>
      <fsummary>Change the size of a binary.</fsummary>
      <desc>
        <p>Changes the size of a binary <c>bin</c>. The source binary
          can be read-only, in which case it is left untouched and
          a mutable copy is allocated and assigned to <c>*bin</c>.</p>
        <p>Returns <c>true</c> on success, or <c>false</c> if memory allocation
          failed.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret>
        <nametext>enif_release_binary(ErlNifBinary* bin)</nametext></name>
      <fsummary>Release a binary.</fsummary>
      <desc>
        <p>Releases a binary obtained from
          <seealso marker="#enif_alloc_binary">
          <c>enif_alloc_binary</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret>
        <nametext>enif_release_resource(void* obj)</nametext></name>
      <fsummary>Release a resource object.</fsummary>
      <desc>
        <p>Removes a reference to resource object <c>obj</c> obtained from
          <seealso marker="#enif_alloc_resource">
          <c>enif_alloc_resource</c></seealso>.
          The resource object is destructed when the last reference is removed.
          Each call to <c>enif_release_resource</c> must correspond to a
          previous call to <c>enif_alloc_resource</c> or
          <seealso marker="#enif_keep_resource">
          <c>enif_keep_resource</c></seealso>.
          References made by <seealso marker="#enif_make_resource">
          <c>enif_make_resource</c></seealso>
          can only be removed by the garbage collector.</p>
      </desc>
    </func>

    <func>
      <name><ret>ErlNifRWLock *</ret>
        <nametext>enif_rwlock_create(char *name)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_rwlock_create">
          <c>erl_drv_rwlock_create</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret>
        <nametext>enif_rwlock_destroy(ErlNifRWLock *rwlck)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_rwlock_destroy">
          <c>erl_drv_rwlock_destroy</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>char*</ret>
        <nametext>enif_rwlock_name(ErlNifRWLock* rwlck)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_rwlock_name">
          <c>erl_drv_rwlock_name</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret>
        <nametext>enif_rwlock_rlock(ErlNifRWLock *rwlck)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_rwlock_rlock">
          <c>erl_drv_rwlock_rlock</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret>
        <nametext>enif_rwlock_runlock(ErlNifRWLock *rwlck)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_rwlock_runlock">
          <c>erl_drv_rwlock_runlock</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret>
        <nametext>enif_rwlock_rwlock(ErlNifRWLock *rwlck)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_rwlock_rwlock">
          <c>erl_drv_rwlock_rwlock</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret>
        <nametext>enif_rwlock_rwunlock(ErlNifRWLock *rwlck)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_rwlock_rwunlock">
          <c>erl_drv_rwlock_rwunlock</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_rwlock_tryrlock(ErlNifRWLock *rwlck)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_rwlock_tryrlock">
          <c>erl_drv_rwlock_tryrlock</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_rwlock_tryrwlock(ErlNifRWLock *rwlck)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_rwlock_tryrwlock">
          <c>erl_drv_rwlock_tryrwlock</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>ERL_NIF_TERM</ret><nametext>enif_schedule_nif(ErlNifEnv* env,
        const char* fun_name, int flags, ERL_NIF_TERM (*fp)(ErlNifEnv* env, int
        argc, const ERL_NIF_TERM argv[]), int argc, const ERL_NIF_TERM
        argv[])</nametext></name>
      <fsummary>Schedule a NIF for execution.</fsummary>
      <desc>
        <p>Schedules NIF <c>fp</c> to execute. This function allows an
          application to break up long-running work into multiple regular NIF
          calls or to schedule a <seealso marker="#dirty_nifs">
          dirty NIF</seealso> to execute on a dirty scheduler thread.</p>
        <taglist>
          <tag><c>fun_name</c></tag>
          <item>
            <p>Provides a name for the NIF that is scheduled for execution.
              If it cannot be converted to an atom, <c>enif_schedule_nif</c>
              returns a <c>badarg</c> exception.</p>
          </item>
          <tag><c>flags</c></tag>
          <item>
            <p>Must be set to <c>0</c> for a regular NIF. If the emulator was
              built with dirty scheduler support enabled,
              <c>flags</c> can be set to either
              <c>ERL_NIF_DIRTY_JOB_CPU_BOUND</c> if the job is expected to be
              CPU-bound, or <c>ERL_NIF_DIRTY_JOB_IO_BOUND</c> for
              jobs that will be I/O-bound. If dirty scheduler threads are not
              available in the emulator, an attempt to schedule such a job
              results in a <c>notsup</c> exception.</p>
          </item>
          <tag><c>argc</c> and <c>argv</c></tag>
          <item>
            <p>Can either be the originals passed into the calling NIF,
              or can be values created by the calling NIF.</p>
          </item>
        </taglist>
        <p>The calling NIF must use the return value of
          <c>enif_schedule_nif</c> as its own return value.</p>
        <p>Be aware that <c>enif_schedule_nif</c>, as its name implies, only
          schedules the NIF for future execution. The calling NIF does not
          block waiting for the scheduled NIF to execute and return. This means
          that the calling NIF cannot expect to receive the scheduled NIF
          return value and use it for further operations.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_select(ErlNifEnv* env, ErlNifEvent event, enum ErlNifSelectFlags mode,
	void* obj, const ErlNifPid* pid, ERL_NIF_TERM ref)</nametext>
      </name>
      <fsummary>Manage subscription on IO event.</fsummary>
      <desc>
        <p>This function can be used to receive asynchronous notifications
	  when OS-specific event objects become ready for either read or write operations.</p>
	<p>Argument <c>event</c> identifies the event object. On Unix
	  systems, the functions <c>select</c>/<c>poll</c> are used. The event
	  object must be a socket, pipe or other file descriptor object that
	  <c>select</c>/<c>poll</c> can use.</p>
	<p>Argument <c>mode</c> describes the type of events to wait for. It can be
	  <c>ERL_NIF_SELECT_READ</c>, <c>ERL_NIF_SELECT_WRITE</c> or a bitwise
	  OR combination to wait for both. It can also be <c>ERL_NIF_SELECT_STOP</c>
	  or <c>ERL_NIF_SELECT_CANCEL</c> which are described further
	  below. When a read or write event is triggered,
	  a notification message like this is sent to the process identified by
	  <c>pid</c>:</p>
	<code type="none">{select, Obj, Ref, ready_input | ready_output}</code>
	<p><c>ready_input</c> or <c>ready_output</c> indicates if the event object
	  is ready for reading or writing.</p>
	<p>Argument <c>pid</c> may be <c>NULL</c> to indicate the calling process.</p>
	<p>Argument <c>obj</c> is a resource object obtained from
          <seealso marker="#enif_alloc_resource"><c>enif_alloc_resource</c></seealso>.
	  The purpose of the resource objects is as a container of the event object
	  to manage its state and lifetime. A handle to the resource is received
	  in the notification message as <c>Obj</c>.</p>
	<p>Argument <c>ref</c> must be either a reference obtained from
	  <seealso marker="erlang#make_ref-0"><c>erlang:make_ref/0</c></seealso>
	  or the atom <c>undefined</c>. It will be passed as <c>Ref</c> in the notifications.
	  If a selective <c>receive</c> statement is used to wait for the notification
	  then a reference created just before the <c>receive</c> will exploit a runtime
	  optimization that bypasses all earlier received messages in the queue.</p>
	<p>The notifications are one-shot only. To receive further notifications of the same
	  type (read or write), repeated calls to <c>enif_select</c> must be made
	  after receiving each notification.</p>
	<p><c>ERL_NIF_SELECT_CANCEL</c> can be used to cancel previously
	  selected events. It must be used in a bitwise OR combination with
	  <c>ERL_NIF_SELECT_READ</c> and/or <c>ERL_NIF_SELECT_WRITE</c> to
	  indicate which type of event to cancel. The return value will
	  tell if the event was actualy cancelled or if a notification may
	  already have been sent.</p>
	<p>Use <c>ERL_NIF_SELECT_STOP</c> as <c>mode</c> in order to safely
	  close an event object that has been passed to <c>enif_select</c>. The
	  <seealso marker="#ErlNifResourceStop"><c>stop</c></seealso> callback
	  of the resource <c>obj</c> will be called when it is safe to close
	  the event object. This safe way of closing event objects must be used
	  even if all notifications have been received (or cancelled) and no
	  further calls to <c>enif_select</c> have been made.
	  <c>ERL_NIF_SELECT_STOP</c> will first cancel any selected events
	  before it calls or schedules the <c>stop</c> callback.</p>
	<p>The first call to <c>enif_select</c> for a specific OS <c>event</c> will establish
	  a relation between the event object and the containing resource. All subsequent calls
	  for an <c>event</c> must pass its containing resource as argument
	  <c>obj</c>. The relation is dissolved when <c>enif_select</c> has
	  been called with <c>mode</c> as <c>ERL_NIF_SELECT_STOP</c> and the
	  corresponding <c>stop</c> callback has returned. A resource can contain
	  several event objects but one event object can only be contained within
	  one resource. A resource will not be destructed until all its contained relations
	  have been dissolved.</p>
        <note>
	  <p>Use <seealso marker="#enif_monitor_process"><c>enif_monitor_process</c></seealso>
	  together with <c>enif_select</c> to detect failing Erlang
	  processes and prevent them from causing permanent leakage of resources
	  and their contained OS event objects.</p>
	</note>
	<p>Returns a non-negative value on success where the following bits can be set:</p>
        <taglist>
          <tag><c>ERL_NIF_SELECT_STOP_CALLED</c></tag>
          <item>The stop callback was called directly by <c>enif_select</c>.</item>
          <tag><c>ERL_NIF_SELECT_STOP_SCHEDULED</c></tag>
          <item>The stop callback was scheduled to run on some other thread
	  or later by this thread.</item>
	  <tag><c>ERL_NIF_SELECT_READ_CANCELLED</c></tag>
          <item>A read event was cancelled by <c>ERL_NIF_SELECT_CANCEL</c> or
	  <c>ERL_NIF_SELECT_STOP</c> and is guaranteed not to generate a
	  <c>ready_input</c> notification message.</item>
	  <tag><c>ERL_NIF_SELECT_WRITE_CANCELLED</c></tag>
          <item>A write event was cancelled by <c>ERL_NIF_SELECT_CANCEL</c> or
	  <c>ERL_NIF_SELECT_STOP</c> and is guaranteed not to generate a
	  <c>ready_output</c> notification message.</item>
	</taglist>
	<p>Returns a negative value if the call failed where the following bits can be set:</p>
        <taglist>
          <tag><c>ERL_NIF_SELECT_INVALID_EVENT</c></tag>
          <item>Argument <c>event</c> is not a valid OS event object.</item>
          <tag><c>ERL_NIF_SELECT_FAILED</c></tag>
          <item>The system call failed to add the event object to the poll set.</item>
        </taglist>
        <note>
          <p>Use bitwise AND to test for specific bits in the return value.
	  New significant bits may be added in future releases to give more detailed
	  information for both failed and successful calls. Do NOT use equality tests
	  like <c>==</c>, as that may cause your application to stop working.</p>
	  <p>Example:</p>
	  <code type="none">
retval = enif_select(env, fd, ERL_NIF_SELECT_STOP, resource, ref);
if (retval &lt; 0) {
    /* handle error */
}
/* Success! */
if (retval &amp; ERL_NIF_SELECT_STOP_CALLED) {
    /* ... */
}
</code>
        </note>
	<note><p>The mode flag <c>ERL_NIF_SELECT_CANCEL</c> and the return flags
	  <c>ERL_NIF_SELECT_READ_CANCELLED</c> and
	  <c>ERL_NIF_SELECT_WRITE_CANCELLED</c> were introduced in erts-11.0
	  (OTP-22.0).</p>
	</note>
      </desc>
    </func>

    <func>
      <name><ret>ErlNifPid *</ret>
        <nametext>enif_self(ErlNifEnv* caller_env, ErlNifPid* pid)</nametext>
      </name>
      <fsummary>Get the pid of the calling process.</fsummary>
      <desc>
        <p>Initializes the <seealso marker="#ErlNifPid"><c>ErlNifPid</c></seealso>
	variable at <c>*pid</c> to represent the calling process.</p>
        <p>Returns <c>pid</c> if successful, or NULL if <c>caller_env</c> is not
	a <seealso marker="#ErlNifEnv">process-bound environment</seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_send(ErlNifEnv* env, ErlNifPid* to_pid,
        ErlNifEnv* msg_env, ERL_NIF_TERM msg)</nametext></name>
      <fsummary>Send a message to a process.</fsummary>
      <desc>
        <p>Sends a message to a process.</p>
        <taglist>
          <tag><c>env</c></tag>
          <item>The environment of the calling process. Must be <c>NULL</c>
            only if calling from a created thread.</item>
          <tag><c>*to_pid</c></tag>
          <item>The pid of the receiving process. The pid is to refer to a
            process on the local node.</item>
          <tag><c>msg_env</c></tag>
          <item>The environment of the message term. Must be a
            process-independent environment allocated with
            <seealso marker="#enif_alloc_env"><c>enif_alloc_env</c></seealso>
            or NULL.</item>
          <tag><c>msg</c></tag>
          <item>The message term to send.</item>
        </taglist>
        <p>Returns <c>true</c> if the message is successfully sent. Returns
          <c>false</c> if the send operation fails, that is:</p>
        <list type="bulleted">
          <item><c>*to_pid</c> does not refer to an alive local process.</item>
          <item>The currently executing process (that is, the sender) is not
           alive.</item>
        </list>
        <p>The message environment <c>msg_env</c> with all its terms (including
          <c>msg</c>) is invalidated by a successful call to <c>enif_send</c>.
          The environment is to either be freed with
          <seealso marker="#enif_free_env">
          <c>enif_free_env</c></seealso> of cleared for reuse with
          <seealso marker="#enif_clear_env"><c>enif_clear_env</c></seealso>.</p>
        <p>If <c>msg_env</c> is set to <c>NULL</c>, the <c>msg</c> term is
          copied and the original term and its environment is still valid after
          the call.</p>
        <p>This function is only thread-safe when the emulator with SMP support
          is used. It can only be used in a non-SMP emulator from a NIF-calling
          thread.</p>
        <note>
          <p>Passing <c>msg_env</c> as <c>NULL</c> is only supported as from
            ERTS 8.0 (Erlang/OTP 19).</p>
        </note>
      </desc>
    </func>

    <func>
      <name><ret>unsigned</ret>
        <nametext>enif_sizeof_resource(void* obj)</nametext></name>
      <fsummary>Get the byte size of a resource object.</fsummary>
      <desc>
        <p>Gets the byte size of resource object <c>obj</c> obtained by
          <seealso marker="#enif_alloc_resource">
          <c>enif_alloc_resource</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_snprintf(char *str, size_t size, const
        char *format, ...)</nametext></name>
      <fsummary>Format strings and Erlang terms.</fsummary>
      <desc>
        <p>Similar to <c>snprintf</c> but this format string also accepts
        <c>"%T"</c>, which formats Erlang terms of type
	<seealso marker="#ERL_NIF_TERM"><c>ERL_NIF_TERM</c></seealso>.</p>
	<p>This function is primarily intended for debugging purpose. It is not
	  recommended to print very large terms with <c>%T</c>. The function may
	  change <c>errno</c>, even if successful.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret><nametext>enif_system_info(ErlNifSysInfo
        *sys_info_ptr, size_t size)</nametext></name>
      <fsummary>Get information about the Erlang runtime system.</fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#driver_system_info">
          <c>driver_system_info</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret><nametext>enif_term_to_binary(ErlNifEnv *env,
        ERL_NIF_TERM term, ErlNifBinary *bin)</nametext></name>
      <fsummary>Convert a term to the external format.</fsummary>
      <desc>
        <p>Allocates a new binary with <seealso marker="#enif_alloc_binary">
          <c>enif_alloc_binary</c></seealso> and stores the result of encoding
          <c>term</c> according to the Erlang external term format.</p>
      <p>Returns <c>true</c> on success, or <c>false</c> if the allocation
        fails.</p>
      <p>See also <seealso marker="erlang#term_to_binary-1">
        <c>erlang:term_to_binary/1</c></seealso> and
        <seealso marker="#enif_binary_to_term">
        <c>enif_binary_to_term</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_thread_create(char *name,ErlNifTid
        *tid,void * (*func)(void *),void *args,ErlNifThreadOpts
        *opts)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_thread_create">
          <c>erl_drv_thread_create</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret>
        <nametext>enif_thread_exit(void *resp)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_thread_exit">
          <c>erl_drv_thread_exit</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_thread_join(ErlNifTid, void **respp)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_thread_join">
          <c>erl_drv_thread_join</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>char*</ret>
        <nametext>enif_thread_name(ErlNifTid tid)</nametext></name>
      <fsummary>Thread name</fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_thread_name">
          <c>erl_drv_thread_name</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>ErlNifThreadOpts *</ret>
        <nametext>enif_thread_opts_create(char *name)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_thread_opts_create">
          <c>erl_drv_thread_opts_create</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret>
        <nametext>enif_thread_opts_destroy(ErlNifThreadOpts *opts)</nametext>
      </name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_thread_opts_destroy">
          <c>erl_drv_thread_opts_destroy</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>ErlNifTid</ret>
        <nametext>enif_thread_self(void)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_thread_self">
          <c>erl_drv_thread_self</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
      <nametext>enif_thread_type(void)</nametext></name>
      <fsummary>Determine type of current thread</fsummary>
      <desc>
        <p>Determine the type of currently executing thread. A positive value
        indicates a scheduler thread while a negative value or zero indicates
        another type of thread. Currently the following specific types exist
        (which may be extended in the future):</p>
        <taglist>
          <tag><c>ERL_NIF_THR_UNDEFINED</c></tag>
          <item><p>Undefined thread that is not a scheduler thread.</p></item>
          <tag><c>ERL_NIF_THR_NORMAL_SCHEDULER</c></tag>
          <item><p>A normal scheduler thread.</p></item>
          <tag><c>ERL_NIF_THR_DIRTY_CPU_SCHEDULER</c></tag>
          <item><p>A dirty CPU scheduler thread.</p></item>
          <tag><c>ERL_NIF_THR_DIRTY_IO_SCHEDULER</c></tag>
          <item><p>A dirty I/O scheduler thread.</p></item>
        </taglist>
      </desc>
    </func>

    <func>
      <name><ret>ErlNifTime</ret>
        <nametext>enif_time_offset(ErlNifTimeUnit time_unit)</nametext></name>
      <fsummary>Get current time offset.</fsummary>
      <desc>
        <marker id="enif_time_offset"></marker>
        <p>Returns the current time offset between
          <seealso marker="time_correction#Erlang_Monotonic_Time">
          Erlang monotonic time</seealso> and
          <seealso marker="time_correction#Erlang_System_Time">
          Erlang system time</seealso>
          converted into the <c>time_unit</c> passed as argument.</p>
        <p><c>time_unit</c> is the time unit of the returned value.</p>
        <p>Returns <c>ERL_NIF_TIME_ERROR</c> if called with an invalid
          time unit argument or if called from a thread that is not a
          scheduler thread.</p>
        <p>See also <seealso marker="#ErlNifTime"><c>ErlNifTime</c></seealso>
          and
          <seealso marker="#ErlNifTimeUnit"><c>ErlNifTimeUnit</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void *</ret>
        <nametext>enif_tsd_get(ErlNifTSDKey key)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_tsd_get">
          <c>erl_drv_tsd_get</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_tsd_key_create(char *name, ErlNifTSDKey *key)</nametext>
      </name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_tsd_key_create">
          <c>erl_drv_tsd_key_create</c></seealso>.</p>
    </desc>
   </func>

    <func>
      <name><ret>void</ret>
        <nametext>enif_tsd_key_destroy(ErlNifTSDKey key)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_tsd_key_destroy">
          <c>erl_drv_tsd_key_destroy</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>void</ret>
        <nametext>enif_tsd_set(ErlNifTSDKey key, void *data)</nametext></name>
      <fsummary></fsummary>
      <desc>
        <p>Same as <seealso marker="erl_driver#erl_drv_tsd_set">
          <c>erl_drv_tsd_set</c></seealso>.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
      <nametext>enif_vfprintf(FILE *stream, const char *format, va_list ap)
      </nametext></name>
      <fsummary>Format strings and Erlang terms.</fsummary>
      <desc>
        <p>Equivalent to <seealso marker="#enif_fprintf"><c>enif_fprintf</c></seealso>
	except that its called with a <c>va_list</c> instead of a variable number of
	arguments.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
      <nametext>enif_vsnprintf(char *str, size_t size, const char *format, va_list ap)
      </nametext></name>
      <fsummary>Format strings and Erlang terms.</fsummary>
      <desc>
        <p>Equivalent to <seealso marker="#enif_snprintf"><c>enif_snprintf</c></seealso>
	except that its called with a <c>va_list</c> instead of a variable number of
	arguments.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_whereis_pid(ErlNifEnv *env,
          ERL_NIF_TERM name, ErlNifPid *pid)</nametext></name>
      <fsummary>Looks up a process by its registered name.</fsummary>
      <desc>
        <p>Looks up a process by its registered name.</p>
        <taglist>
          <tag><c>env</c></tag>
          <item>The environment of the calling process. Must be <c>NULL</c>
            only if calling from a created thread.</item>
          <tag><c>name</c></tag>
          <item>The name of a registered process, as an atom.</item>
          <tag><c>*pid</c></tag>
          <item>The <seealso marker="#ErlNifPid"><c>ErlNifPid</c></seealso>
            in which the resolved process id is stored.</item>
        </taglist>
        <p>On success, sets <c>*pid</c> to the local process registered with
          <c>name</c> and returns <c>true</c>. If <c>name</c> is not a
          registered process, or is not an atom, <c>false</c> is returned and
          <c>*pid</c> is unchanged.</p>
        <p>Works as <seealso marker="erlang#whereis-1">
          <c>erlang:whereis/1</c></seealso>, but restricted to processes. See
          <seealso marker="#enif_whereis_port"><c>enif_whereis_port</c></seealso>
          to resolve registered ports.</p>
      </desc>
    </func>

    <func>
      <name><ret>int</ret>
        <nametext>enif_whereis_port(ErlNifEnv *env,
          ERL_NIF_TERM name, ErlNifPort *port)</nametext></name>
      <fsummary>Looks up a port by its registered name.</fsummary>
      <desc>
        <p>Looks up a port by its registered name.</p>
        <taglist>
          <tag><c>env</c></tag>
          <item>The environment of the calling process. Must be <c>NULL</c>
            only if calling from a created thread.</item>
          <tag><c>name</c></tag>
          <item>The name of a registered port, as an atom.</item>
          <tag><c>*port</c></tag>
          <item>The <seealso marker="#ErlNifPort"><c>ErlNifPort</c></seealso>
            in which the resolved port id is stored.</item>
        </taglist>
        <p>On success, sets <c>*port</c> to the port registered with
          <c>name</c> and returns <c>true</c>. If <c>name</c> is not a
          registered port, or is not an atom, <c>false</c> is returned and
          <c>*port</c> is unchanged.</p>
        <p>Works as <seealso marker="erlang#whereis-1">
          <c>erlang:whereis/1</c></seealso>, but restricted to ports. See
          <seealso marker="#enif_whereis_pid"><c>enif_whereis_pid</c></seealso>
          to resolve registered processes.</p>
      </desc>
    </func>

  </funcs>

  <section>
    <title>See Also</title>
    <p><seealso marker="erlang#load_nif-2">
      <c>erlang:load_nif/2</c></seealso></p>
  </section>
</cref>