<?xml version="1.0" encoding="latin1" ?>
<!DOCTYPE erlref SYSTEM "erlref.dtd">
<erlref>
<header>
<copyright>
<year>1996</year><year>2010</year>
<holder>Ericsson AB. All Rights Reserved.</holder>
</copyright>
<legalnotice>
The contents of this file are subject to the Erlang Public License,
Version 1.1, (the "License"); you may not use this file except in
compliance with the License. You should have received a copy of the
Erlang Public License along with this software. If not, it can be
retrieved online at http://www.erlang.org/.
Software distributed under the License is distributed on an "AS IS"
basis, WITHOUT WARRANTY OF ANY KIND, either express or implied. See
the License for the specific language governing rights and limitations
under the License.
</legalnotice>
<title>erlang</title>
<prepared></prepared>
<docno></docno>
<date></date>
<rev></rev>
<file>erlang.xml</file>
</header>
<module>erlang</module>
<modulesummary>The Erlang BIFs</modulesummary>
<description>
<p>By convention, most built-in functions (BIFs) are seen as being
in the module <c>erlang</c>. A number of the BIFs are viewed more
or less as part of the Erlang programming language and are
<em>auto-imported</em>. Thus, it is not necessary to specify
the module name and both the calls <c>atom_to_list(Erlang)</c> and
<c>erlang:atom_to_list(Erlang)</c> are identical.</p>
<p>In the text, auto-imported BIFs are listed without module prefix.
BIFs listed with module prefix are not auto-imported.</p>
<p>BIFs may fail for a variety of reasons. All BIFs fail with
reason <c>badarg</c> if they are called with arguments of an
incorrect type. The other reasons that may make BIFs fail are
described in connection with the description of each individual
BIF.</p>
<p>Some BIFs may be used in guard tests, these are marked with
"Allowed in guard tests".</p>
</description>
<section>
<title>DATA TYPES</title>
<marker id="iolist_definition"></marker>
<code type="none">
ext_binary()
a binary data object,
structured according to the Erlang external term format
iodata() = iolist() | binary()
iolist() = [char() | binary() | iolist()]
a binary is allowed as the tail of the list</code>
</section>
<funcs>
<func>
<name>abs(Number) -> int() | float()</name>
<fsummary>Arithmetical absolute value</fsummary>
<type>
<v>Number = number()</v>
</type>
<desc>
<p>Returns an integer or float which is the arithmetical
absolute value of <c>Number</c>.</p>
<pre>
> <input>abs(-3.33).</input>
3.33
> <input>abs(-3).</input>
3</pre>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>erlang:adler32(Data) -> int()</name>
<fsummary>Compute adler32 checksum</fsummary>
<type>
<v>Data = iodata()</v>
</type>
<desc>
<p>Computes and returns the adler32 checksum for <c>Data</c>.</p>
</desc>
</func>
<func>
<name>erlang:adler32(OldAdler, Data) -> int()</name>
<fsummary>Compute adler32 checksum</fsummary>
<type>
<v>OldAdler = int()</v>
<v>Data = iodata()</v>
</type>
<desc>
<p>Continue computing the adler32 checksum by combining
the previous checksum, <c>OldAdler</c>, with the checksum of
<c>Data</c>.</p>
<p>The following code:</p>
<code>
X = erlang:adler32(Data1),
Y = erlang:adler32(X,Data2).
</code>
<p>- would assign the same value to <c>Y</c> as this would:</p>
<code>
Y = erlang:adler32([Data1,Data2]).
</code>
</desc>
</func>
<func>
<name>erlang:adler32_combine(FirstAdler, SecondAdler, SecondSize) -> int()</name>
<fsummary>Combine two adler32 checksums</fsummary>
<type>
<v>FirstAdler = SecondAdler = int()</v>
<v>SecondSize = int()</v>
</type>
<desc>
<p>Combines two previously computed adler32 checksums.
This computation requires the size of the data object for
the second checksum to be known.</p>
<p>The following code:</p>
<code>
Y = erlang:adler32(Data1),
Z = erlang:adler32(Y,Data2).
</code>
<p>- would assign the same value to <c>Z</c> as this would:</p>
<code>
X = erlang:adler32(Data1),
Y = erlang:adler32(Data2),
Z = erlang:adler32_combine(X,Y,iolist_size(Data2)).
</code>
</desc>
</func>
<func>
<name>erlang:append_element(Tuple1, Term) -> Tuple2</name>
<fsummary>Append an extra element to a tuple</fsummary>
<type>
<v>Tuple1 = Tuple2 = tuple()</v>
<v>Term = term()</v>
</type>
<desc>
<p>Returns a new tuple which has one element more than
<c>Tuple1</c>, and contains the elements in <c>Tuple1</c>
followed by <c>Term</c> as the last element. Semantically
equivalent to
<c>list_to_tuple(tuple_to_list(Tuple) ++ [Term])</c>, but much
faster.</p>
<pre>
> <input>erlang:append_element({one, two}, three).</input>
{one,two,three}</pre>
</desc>
</func>
<func>
<name>apply(Fun, Args) -> term() | empty()</name>
<fsummary>Apply a function to an argument list</fsummary>
<type>
<v>Fun = fun()</v>
<v>Args = [term()]</v>
</type>
<desc>
<p>Call a fun, passing the elements in <c>Args</c> as
arguments.</p>
<p>Note: If the number of elements in the arguments are known at
compile-time, the call is better written as
<c>Fun(Arg1, Arg2, ... ArgN)</c>.</p>
<warning>
<p>Earlier, <c>Fun</c> could also be given as
<c>{Module, Function}</c>, equivalent to
<c>apply(Module, Function, Args)</c>. This usage is
deprecated and will stop working in a future release of
Erlang/OTP.</p>
</warning>
</desc>
</func>
<func>
<name>apply(Module, Function, Args) -> term() | empty()</name>
<fsummary>Apply a function to an argument list</fsummary>
<type>
<v>Module = Function = atom()</v>
<v>Args = [term()]</v>
</type>
<desc>
<p>Returns the result of applying <c>Function</c> in
<c>Module</c> to <c>Args</c>. The applied function must
be exported from <c>Module</c>. The arity of the function is
the length of <c>Args</c>.</p>
<pre>
> <input>apply(lists, reverse, [[a, b, c]]).</input>
[c,b,a]</pre>
<p><c>apply</c> can be used to evaluate BIFs by using
the module name <c>erlang</c>.</p>
<pre>
> <input>apply(erlang, atom_to_list, ['Erlang']).</input>
"Erlang"</pre>
<p>Note: If the number of arguments are known at compile-time,
the call is better written as
<c>Module:Function(Arg1, Arg2, ..., ArgN)</c>.</p>
<p>Failure: <c>error_handler:undefined_function/3</c> is called
if the applied function is not exported. The error handler
can be redefined (see
<seealso marker="#process_flag/2">process_flag/2</seealso>).
If the <c>error_handler</c> is undefined, or if the user has
redefined the default <c>error_handler</c> so the replacement
module is undefined, an error with the reason <c>undef</c>
is generated.</p>
</desc>
</func>
<func>
<name>atom_to_binary(Atom, Encoding) -> binary()</name>
<fsummary>Return the binary representation of an atom</fsummary>
<type>
<v>Atom = atom()</v>
<v>Encoding = latin1 | utf8 | unicode</v>
</type>
<desc>
<p>Returns a binary which corresponds to the text
representation of <c>Atom</c>. If <c>Encoding</c>
is <c>latin1</c>, there will be one byte for each character
in the text representation. If <c>Encoding</c> is <c>utf8</c> or
<c>unicode</c>, the characters will encoded using UTF-8
(meaning that characters from 16#80 up to 0xFF will be
encode in two bytes).</p>
<note><p>Currently, <c>atom_to_binary(Atom, latin1)</c> can
never fail because the text representation of an atom can only contain
characters from 0 to 16#FF. In a future release, the text representation
of atoms might be allowed to contain any Unicode character
and <c>atom_to_binary(Atom, latin1)</c> will fail if the
text representation for the <c>Atom</c> contains a Unicode
character greater than 16#FF.</p></note>
<pre>
> <input>atom_to_binary('Erlang', latin1).</input>
<<"Erlang">></pre>
</desc>
</func>
<func>
<name>atom_to_list(Atom) -> string()</name>
<fsummary>Text representation of an atom</fsummary>
<type>
<v>Atom = atom()</v>
</type>
<desc>
<p>Returns a string which corresponds to the text
representation of <c>Atom</c>.</p>
<pre>
> <input>atom_to_list('Erlang').</input>
"Erlang"</pre>
</desc>
</func>
<func>
<name>binary_part(Subject, PosLen) -> binary()</name>
<fsummary>Extracts a part of a binary</fsummary>
<type>
<v>Subject = binary()</v>
<v>PosLen = {Start,Length}</v>
<v>Start = int()</v>
<v>Length = int()</v>
</type>
<desc>
<p>Extracts the part of the binary described by <c>PosLen</c>.</p>
<p>Negative length can be used to extract bytes at the end of a binary:</p>
<code>
1> Bin = <<1,2,3,4,5,6,7,8,9,10>>.
2> binary_part(Bin,{byte_size(Bin), -5)).
<<6,7,8,9,10>>
</code>
<p>If <c>PosLen</c> in any way references outside the binary, a <c>badarg</c> exception is raised.</p>
<p><c>Start</c> is zero-based, i.e:</p>
<code>
1> Bin = <<1,2,3>>
2> binary_part(Bin,{0,2}).
<<1,2>>
</code>
<p>See the STDLIB module <c>binary</c> for details about the <c>PosLen</c> semantics.</p>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>binary_part(Subject, Start, Length) -> binary()</name>
<fsummary>Extracts a part of a binary</fsummary>
<type>
<v>Subject = binary()</v>
<v>Start = int()</v>
<v>Length = int()</v>
</type>
<desc>
<p>The same as <c>binary_part(Subject, {Pos, Len})</c>.</p>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>binary_to_atom(Binary, Encoding) -> atom()</name>
<fsummary>Convert from text representation to an atom</fsummary>
<type>
<v>Binary = binary()</v>
<v>Encoding = latin1 | utf8 | unicode</v>
</type>
<desc>
<p>Returns the atom whose text representation is
<c>Binary</c>. If <c>Encoding</c> is <c>latin1</c>, no
translation of bytes in the binary is done. If <c>Encoding</c>
is <c>utf8</c> or <c>unicode</c>, the binary must contain
valid UTF-8 sequences; furthermore, only Unicode characters up
to 0xFF are allowed.</p>
<note><p><c>binary_to_atom(Binary, utf8)</c> will fail if
the binary contains Unicode characters greater than 16#FF.
In a future release, such Unicode characters might be allowed
and <c>binary_to_atom(Binary, utf8)</c>
will not fail in that case.</p></note>
<pre>
> <input>binary_to_atom(<<"Erlang">>, latin1).</input>
'Erlang'
> <input>binary_to_atom(<<1024/utf8>>, utf8).</input>
** exception error: bad argument
in function binary_to_atom/2
called as binary_to_atom(<<208,128>>,utf8)</pre>
</desc>
</func>
<func>
<name>binary_to_existing_atom(Binary, Encoding) -> atom()</name>
<fsummary>Convert from text representation to an atom</fsummary>
<type>
<v>Binary = binary()</v>
<v>Encoding = latin1 | utf8 | unicode</v>
</type>
<desc>
<p>Works like <seealso marker="#binary_to_atom/2">binary_to_atom/2</seealso>,
but the atom must already exist.</p>
<p>Failure: <c>badarg</c> if the atom does not already exist.</p>
</desc>
</func>
<func>
<name>binary_to_list(Binary) -> [char()]</name>
<fsummary>Convert a binary to a list</fsummary>
<type>
<v>Binary = binary()</v>
</type>
<desc>
<p>Returns a list of integers which correspond to the bytes of
<c>Binary</c>.</p>
</desc>
</func>
<func>
<name>binary_to_list(Binary, Start, Stop) -> [char()]</name>
<fsummary>Convert part of a binary to a list</fsummary>
<type>
<v>Binary = binary()</v>
<v>Start = Stop = 1..byte_size(Binary)</v>
</type>
<desc>
<p>As <c>binary_to_list/1</c>, but returns a list of integers
corresponding to the bytes from position <c>Start</c> to
position <c>Stop</c> in <c>Binary</c>. Positions in the
binary are numbered starting from 1.</p>
<note><p>This function's indexing style of using one-based indices for
binaries is deprecated. New code should use the functions in
the STDLIB module <c>binary</c> instead. They consequently
use the same (zero-based) style of indexing.</p></note>
</desc>
</func>
<func>
<name>bitstring_to_list(Bitstring) -> [char()|bitstring()]</name>
<fsummary>Convert a bitstring to a list</fsummary>
<type>
<v>Bitstring = bitstring()</v>
</type>
<desc>
<p>Returns a list of integers which correspond to the bytes of
<c>Bitstring</c>. If the number of bits in the binary is not
divisible by 8, the last element of the list will be a bitstring
containing the remaining bits (1 up to 7 bits).</p>
</desc>
</func>
<func>
<name>binary_to_term(Binary) -> term()</name>
<fsummary>Decode an Erlang external term format binary</fsummary>
<type>
<v>Binary = ext_binary()</v>
</type>
<desc>
<p>Returns an Erlang term which is the result of decoding
the binary object <c>Binary</c>, which must be encoded
according to the Erlang external term format.</p>
<warning>
<p>When decoding binaries from untrusted sources, consider using
<c>binary_to_term/2</c> to prevent denial of service attacks.</p>
</warning>
<p>See also
<seealso marker="#term_to_binary/1">term_to_binary/1</seealso>
and
<seealso marker="#binary_to_term/2">binary_to_term/2</seealso>.</p>
</desc>
</func>
<func>
<name>binary_to_term(Binary, Opts) -> term()</name>
<fsummary>Decode an Erlang external term format binary</fsummary>
<type>
<v>Opts = [safe]</v>
<v>Binary = ext_binary()</v>
</type>
<desc>
<p>As <c>binary_to_term/1</c>, but takes options that affect decoding
of the binary.</p>
<taglist>
<tag><c>safe</c></tag>
<item>
<p>Use this option when receiving binaries from an untrusted
source.</p>
<p>When enabled, it prevents decoding data that may be used to
attack the Erlang system. In the event of receiving unsafe
data, decoding fails with a badarg error.</p>
<p>Currently, this prevents creation of new atoms directly,
creation of new atoms indirectly (as they are embedded in
certain structures like pids, refs, funs, etc.), and creation of
new external function references. None of those resources are
currently garbage collected, so unchecked creation of them can
exhaust available memory.</p>
</item>
</taglist>
<p>Failure: <c>badarg</c> if <c>safe</c> is specified and unsafe data
is decoded.</p>
<p>See also
<seealso marker="#term_to_binary/1">term_to_binary/1</seealso>,
<seealso marker="#binary_to_term/1">binary_to_term/1</seealso>,
and <seealso marker="#list_to_existing_atom/1">
list_to_existing_atom/1</seealso>.</p>
</desc>
</func>
<func>
<name>bit_size(Bitstring) -> int()</name>
<fsummary>Return the size of a bitstring</fsummary>
<type>
<v>Bitstring = bitstring()</v>
</type>
<desc>
<p>Returns an integer which is the size in bits of <c>Bitstring</c>.</p>
<pre>
> <input>bit_size(<<433:16,3:3>>).</input>
19
> <input>bit_size(<<1,2,3>>).</input>
24</pre>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>erlang:bump_reductions(Reductions) -> void()</name>
<fsummary>Increment the reduction counter</fsummary>
<type>
<v>Reductions = int()</v>
</type>
<desc>
<p>This implementation-dependent function increments
the reduction counter for the calling process. In the Beam
emulator, the reduction counter is normally incremented by
one for each function and BIF call, and a context switch is
forced when the counter reaches the maximum number of reductions
for a process (2000 reductions in R12B).</p>
<warning>
<p>This BIF might be removed in a future version of the Beam
machine without prior warning. It is unlikely to be
implemented in other Erlang implementations.</p>
</warning>
</desc>
</func>
<func>
<name>byte_size(Bitstring) -> int()</name>
<fsummary>Return the size of a bitstring (or binary)</fsummary>
<type>
<v>Bitstring = bitstring()</v>
</type>
<desc>
<p>Returns an integer which is the number of bytes needed to contain
<c>Bitstring</c>. (That is, if the number of bits in <c>Bitstring</c> is not
divisible by 8, the resulting number of bytes will be rounded <em>up</em>.)</p>
<pre>
> <input>byte_size(<<433:16,3:3>>).</input>
3
> <input>byte_size(<<1,2,3>>).</input>
3</pre>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>erlang:cancel_timer(TimerRef) -> Time | false</name>
<fsummary>Cancel a timer</fsummary>
<type>
<v>TimerRef = reference()</v>
<v>Time = int()</v>
</type>
<desc>
<p>Cancels a timer, where <c>TimerRef</c> was returned by
either
<seealso marker="#send_after/3">erlang:send_after/3</seealso>
or
<seealso marker="#start_timer/3">erlang:start_timer/3</seealso>.
If the timer is there to be removed, the function returns
the time in milliseconds left until the timer would have expired,
otherwise <c>false</c> (which means that <c>TimerRef</c> was
never a timer, that it has already been cancelled, or that it
has already delivered its message).</p>
<p>See also
<seealso marker="#send_after/3">erlang:send_after/3</seealso>,
<seealso marker="#start_timer/3">erlang:start_timer/3</seealso>,
and
<seealso marker="#read_timer/1">erlang:read_timer/1</seealso>.</p>
<p>Note: Cancelling a timer does not guarantee that the message
has not already been delivered to the message queue.</p>
</desc>
</func>
<func>
<name>check_process_code(Pid, Module) -> bool()</name>
<fsummary>Check if a process is executing old code for a module</fsummary>
<type>
<v>Pid = pid()</v>
<v>Module = atom()</v>
</type>
<desc>
<p>Returns <c>true</c> if the process <c>Pid</c> is executing
old code for <c>Module</c>. That is, if the current call of
the process executes old code for this module, or if the
process has references to old code for this module, or if the
process contains funs that references old code for this
module. Otherwise, it returns <c>false</c>.</p>
<pre>
> <input>check_process_code(Pid, lists).</input>
false</pre>
<p>See also <seealso marker="kernel:code">code(3)</seealso>.</p>
</desc>
</func>
<func>
<name>concat_binary(ListOfBinaries)</name>
<fsummary>Concatenate a list of binaries (deprecated)</fsummary>
<desc>
<p>Do not use; use
<seealso marker="#list_to_binary/1">list_to_binary/1</seealso>
instead.</p>
</desc>
</func>
<func>
<name>erlang:crc32(Data) -> int()</name>
<fsummary>Compute crc32 (IEEE 802.3) checksum</fsummary>
<type>
<v>Data = iodata()</v>
</type>
<desc>
<p>Computes and returns the crc32 (IEEE 802.3 style) checksum for <c>Data</c>.</p>
</desc>
</func>
<func>
<name>erlang:crc32(OldCrc, Data) -> int()</name>
<fsummary>Compute crc32 (IEEE 802.3) checksum</fsummary>
<type>
<v>OldCrc = int()</v>
<v>Data = iodata()</v>
</type>
<desc>
<p>Continue computing the crc32 checksum by combining
the previous checksum, <c>OldCrc</c>, with the checksum of
<c>Data</c>.</p>
<p>The following code:</p>
<code>
X = erlang:crc32(Data1),
Y = erlang:crc32(X,Data2).
</code>
<p>- would assign the same value to <c>Y</c> as this would:</p>
<code>
Y = erlang:crc32([Data1,Data2]).
</code>
</desc>
</func>
<func>
<name>erlang:crc32_combine(FirstCrc, SecondCrc, SecondSize) -> int()</name>
<fsummary>Combine two crc32 (IEEE 802.3) checksums</fsummary>
<type>
<v>FirstCrc = SecondCrc = int()</v>
<v>SecondSize = int()</v>
</type>
<desc>
<p>Combines two previously computed crc32 checksums.
This computation requires the size of the data object for
the second checksum to be known.</p>
<p>The following code:</p>
<code>
Y = erlang:crc32(Data1),
Z = erlang:crc32(Y,Data2).
</code>
<p>- would assign the same value to <c>Z</c> as this would:</p>
<code>
X = erlang:crc32(Data1),
Y = erlang:crc32(Data2),
Z = erlang:crc32_combine(X,Y,iolist_size(Data2)).
</code>
</desc>
</func>
<func>
<name>date() -> {Year, Month, Day}</name>
<fsummary>Current date</fsummary>
<type>
<v>Year = Month = Day = int()</v>
</type>
<desc>
<p>Returns the current date as <c>{Year, Month, Day}</c>.</p>
<p>The time zone and daylight saving time correction depend on
the underlying OS.</p>
<pre>
> <input>date().</input>
{1995,2,19}</pre>
</desc>
</func>
<func>
<name>erlang:decode_packet(Type,Bin,Options) -> {ok,Packet,Rest} | {more,Length} | {error,Reason}</name>
<fsummary>Extracts a protocol packet from a binary</fsummary>
<type>
<v>Bin = binary()</v>
<v>Options = [Opt]</v>
<v>Packet = binary() | HttpPacket</v>
<v>Rest = binary()</v>
<v>Length = int() | undefined</v>
<v>Reason = term()</v>
<v> Type, Opt -- see below</v>
<v></v>
<v>HttpPacket = HttpRequest | HttpResponse | HttpHeader | http_eoh | HttpError</v>
<v>HttpRequest = {http_request, HttpMethod, HttpUri, HttpVersion}</v>
<v>HttpResponse = {http_response, HttpVersion, integer(), HttpString}</v>
<v>HttpHeader = {http_header, int(), HttpField, Reserved=term(), Value=HttpString}</v>
<v>HttpError = {http_error, HttpString}</v>
<v>HttpMethod = HttpMethodAtom | HttpString</v>
<v>HttpMethodAtom = 'OPTIONS' | 'GET' | 'HEAD' | 'POST' | 'PUT' | 'DELETE' | 'TRACE'</v>
<v>HttpUri = '*' | {absoluteURI, http|https, Host=HttpString, Port=int()|undefined, Path=HttpString} |
{scheme, Scheme=HttpString, HttpString} | {abs_path, HttpString} | HttpString</v>
<v>HttpVersion = {Major=int(), Minor=int()}</v>
<v>HttpString = string() | binary()</v>
<v>HttpField = HttpFieldAtom | HttpString</v>
<v>HttpFieldAtom = 'Cache-Control' | 'Connection' | 'Date' | 'Pragma' | 'Transfer-Encoding' | 'Upgrade' | 'Via' | 'Accept' | 'Accept-Charset' | 'Accept-Encoding' | 'Accept-Language' | 'Authorization' | 'From' | 'Host' | 'If-Modified-Since' | 'If-Match' | 'If-None-Match' | 'If-Range' | 'If-Unmodified-Since' | 'Max-Forwards' | 'Proxy-Authorization' | 'Range' | 'Referer' | 'User-Agent' | 'Age' | 'Location' | 'Proxy-Authenticate' | 'Public' | 'Retry-After' | 'Server' | 'Vary' | 'Warning' | 'Www-Authenticate' | 'Allow' | 'Content-Base' | 'Content-Encoding' | 'Content-Language' | 'Content-Length' | 'Content-Location' | 'Content-Md5' | 'Content-Range' | 'Content-Type' | 'Etag' | 'Expires' | 'Last-Modified' | 'Accept-Ranges' | 'Set-Cookie' | 'Set-Cookie2' | 'X-Forwarded-For' | 'Cookie' | 'Keep-Alive' | 'Proxy-Connection'</v>
<v></v>
</type>
<desc>
<p>Decodes the binary <c>Bin</c> according to the packet
protocol specified by <c>Type</c>. Very similar to the packet
handling done by sockets with the option {packet,Type}.</p>
<p>If an entire packet is contained in <c>Bin</c> it is
returned together with the remainder of the binary as
<c>{ok,Packet,Rest}</c>.</p>
<p>If <c>Bin</c> does not contain the entire packet,
<c>{more,Length}</c> is returned. <c>Length</c> is either the
expected <em>total size</em> of the packet or <c>undefined</c>
if the expected packet size is not known. <c>decode_packet</c>
can then be called again with more data added.</p>
<p>If the packet does not conform to the protocol format
<c>{error,Reason}</c> is returned.</p>
<p>The following values of <c>Type</c> are valid:</p>
<taglist>
<tag><c>raw | 0</c></tag>
<item>
<p>No packet handling is done. Entire binary is
returned unless it is empty.</p>
</item>
<tag><c>1 | 2 | 4</c></tag>
<item>
<p>Packets consist of a header specifying the number of
bytes in the packet, followed by that number of bytes.
The length of header can be one, two, or four bytes;
the order of the bytes is big-endian. The header
will be stripped off when the packet is returned.</p>
</item>
<tag><c>line</c></tag>
<item>
<p>A packet is a line terminated with newline. The
newline character is included in the returned packet
unless the line was truncated according to the option
<c>line_length</c>.</p>
</item>
<tag><c>asn1 | cdr | sunrm | fcgi | tpkt</c></tag>
<item>
<p>The header is <em>not</em> stripped off.</p>
<p>The meanings of the packet types are as follows:</p>
<taglist>
<tag><c>asn1</c> - ASN.1 BER</tag><item></item>
<tag><c>sunrm</c> - Sun's RPC encoding</tag><item></item>
<tag><c>cdr</c> - CORBA (GIOP 1.1)</tag><item></item>
<tag><c>fcgi</c> - Fast CGI</tag><item></item>
<tag><c>tpkt</c> - TPKT format [RFC1006]</tag><item></item>
</taglist>
</item>
<tag><c>http | httph | http_bin | httph_bin</c></tag>
<item>
<p>The Hypertext Transfer Protocol. The packets
are returned with the format according to
<c>HttpPacket</c> described above. A packet is either a
request, a response, a header or an end of header
mark. Invalid lines are returned as <c>HttpError</c>.</p>
<p>Recognized request methods and header fields are returned as atoms.
Others are returned as strings.</p>
<p>The protocol type <c>http</c> should only be used for
the first line when a <c>HttpRequest</c> or a
<c>HttpResponse</c> is expected. The following calls
should use <c>httph</c> to get <c>HttpHeader</c>'s until
<c>http_eoh</c> is returned that marks the end of the
headers and the beginning of any following message body.</p>
<p>The variants <c>http_bin</c> and <c>httph_bin</c> will return
strings (<c>HttpString</c>) as binaries instead of lists.</p>
</item>
</taglist>
<p>The following options are available:</p>
<taglist>
<tag><c>{packet_size, int()}</c></tag>
<item><p>Sets the max allowed size of the packet body. If
the packet header indicates that the length of the
packet is longer than the max allowed length, the packet
is considered invalid. Default is 0 which means no
size limit.</p>
</item>
<tag><c>{line_length, int()}</c></tag>
<item><p>Applies only to line oriented protocols
(<c>line</c>, <c>http</c>). Lines longer than this
will be truncated.</p>
</item>
</taglist>
<pre>
> <input>erlang:decode_packet(1,<<3,"abcd">>,[]).</input>
{ok,<<"abc">>,<<"d">>}
> <input>erlang:decode_packet(1,<<5,"abcd">>,[]).</input>
{more,6}</pre>
</desc>
</func>
<func>
<name>delete_module(Module) -> true | undefined</name>
<fsummary>Make the current code for a module old</fsummary>
<type>
<v>Module = atom()</v>
</type>
<desc>
<p>Makes the current code for <c>Module</c> become old code, and
deletes all references for this module from the export table.
Returns <c>undefined</c> if the module does not exist,
otherwise <c>true</c>.</p>
<warning>
<p>This BIF is intended for the code server (see
<seealso marker="kernel:code">code(3)</seealso>) and should not be
used elsewhere.</p>
</warning>
<p>Failure: <c>badarg</c> if there is already an old version of
<c>Module</c>.</p>
</desc>
</func>
<func>
<name>demonitor(MonitorRef) -> true</name>
<fsummary>Stop monitoring</fsummary>
<type>
<v>MonitorRef = reference()</v>
</type>
<desc>
<p>If <c>MonitorRef</c> is a reference which the calling process
obtained by calling
<seealso marker="#monitor/2">monitor/2</seealso>,
this monitoring is turned off. If the monitoring is already
turned off, nothing happens.</p>
<p>Once <c>demonitor(MonitorRef)</c> has returned it is
guaranteed that no <c>{'DOWN', MonitorRef, _, _, _}</c> message
due to the monitor will be placed in the callers message queue
in the future. A <c>{'DOWN', MonitorRef, _, _, _}</c> message
might have been placed in the callers message queue prior to
the call, though. Therefore, in most cases, it is advisable
to remove such a <c>'DOWN'</c> message from the message queue
after monitoring has been stopped.
<seealso marker="#demonitor/2">demonitor(MonitorRef, [flush])</seealso> can be used instead of
<c>demonitor(MonitorRef)</c> if this cleanup is wanted.</p>
<note>
<p>Prior to OTP release R11B (erts version 5.5) <c>demonitor/1</c>
behaved completely asynchronous, i.e., the monitor was active
until the "demonitor signal" reached the monitored entity. This
had one undesirable effect, though. You could never know when
you were guaranteed <em>not</em> to receive a <c>DOWN</c> message
due to the monitor.</p>
<p>Current behavior can be viewed as two combined operations:
asynchronously send a "demonitor signal" to the monitored entity
and ignore any future results of the monitor. </p>
</note>
<p>Failure: It is an error if <c>MonitorRef</c> refers to a
monitoring started by another process. Not all such cases are
cheap to check; if checking is cheap, the call fails with
<c>badarg</c> (for example if <c>MonitorRef</c> is a remote
reference).</p>
</desc>
</func>
<func>
<name>demonitor(MonitorRef, OptionList) -> true|false</name>
<fsummary>Stop monitoring</fsummary>
<type>
<v>MonitorRef = reference()</v>
<v>OptionList = [Option]</v>
<v>Option = flush</v>
<v>Option = info</v>
</type>
<desc>
<p>The returned value is <c>true</c> unless <c>info</c> is part
of <c>OptionList</c>.
</p>
<p><c>demonitor(MonitorRef, [])</c> is equivalent to
<seealso marker="#demonitor/1">demonitor(MonitorRef)</seealso>.</p>
<p>Currently the following <c>Option</c>s are valid:</p>
<taglist>
<tag><c>flush</c></tag>
<item>
<p>Remove (one) <c>{_, MonitorRef, _, _, _}</c> message,
if there is one, from the callers message queue after
monitoring has been stopped.</p>
<p>Calling <c>demonitor(MonitorRef, [flush])</c>
is equivalent to the following, but more efficient:</p>
<code type="none">
demonitor(MonitorRef),
receive
{_, MonitorRef, _, _, _} ->
true
after 0 ->
true
end</code>
</item>
<tag><c>info</c></tag>
<item>
<p>The returned value is one of the following:</p>
<taglist>
<tag><c>true</c></tag>
<item><p>The monitor was found and removed. In this case
no <c>'DOWN'</c> message due to this monitor have
been nor will be placed in the message queue
of the caller.
</p>
</item>
<tag><c>false</c></tag>
<item><p>The monitor was not found and could not be removed.
This probably because someone already has placed a
<c>'DOWN'</c> message corresponding to this monitor
in the callers message queue.
</p>
</item>
</taglist>
<p>If the <c>info</c> option is combined with the <c>flush</c>
option, <c>false</c> will be returned if a flush was needed;
otherwise, <c>true</c>.
</p>
</item>
</taglist>
<note>
<p>More options may be added in the future.</p>
</note>
<p>Failure: <c>badarg</c> if <c>OptionList</c> is not a list, or
if <c>Option</c> is not a valid option, or the same failure as for
<seealso marker="#demonitor/1">demonitor/1</seealso></p>
</desc>
</func>
<func>
<name>disconnect_node(Node) -> bool() | ignored</name>
<fsummary>Force the disconnection of a node</fsummary>
<type>
<v>Node = atom()</v>
</type>
<desc>
<p>Forces the disconnection of a node. This will appear to
the node <c>Node</c> as if the local node has crashed. This
BIF is mainly used in the Erlang network authentication
protocols. Returns <c>true</c> if disconnection succeeds,
otherwise <c>false</c>. If the local node is not alive,
the function returns <c>ignored</c>.</p>
</desc>
</func>
<func>
<name>erlang:display(Term) -> true</name>
<fsummary>Print a term on standard output</fsummary>
<type>
<v>Term = term()</v>
</type>
<desc>
<p>Prints a text representation of <c>Term</c> on the standard
output.</p>
<warning>
<p>This BIF is intended for debugging only.</p>
</warning>
</desc>
</func>
<func>
<name>element(N, Tuple) -> term()</name>
<fsummary>Get Nth element of a tuple</fsummary>
<type>
<v>N = 1..tuple_size(Tuple)</v>
<v>Tuple = tuple()</v>
</type>
<desc>
<p>Returns the <c>N</c>th element (numbering from 1) of
<c>Tuple</c>.</p>
<pre>
> <input>element(2, {a, b, c}).</input>
b</pre>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>erase() -> [{Key, Val}]</name>
<fsummary>Return and delete the process dictionary</fsummary>
<type>
<v>Key = Val = term()</v>
</type>
<desc>
<p>Returns the process dictionary and deletes it.</p>
<pre>
> <input>put(key1, {1, 2, 3}),</input>
<input>put(key2, [a, b, c]),</input>
<input>erase().</input>
[{key1,{1,2,3}},{key2,[a,b,c]}]</pre>
</desc>
</func>
<func>
<name>erase(Key) -> Val | undefined</name>
<fsummary>Return and delete a value from the process dictionary</fsummary>
<type>
<v>Key = Val = term()</v>
</type>
<desc>
<p>Returns the value <c>Val</c> associated with <c>Key</c> and
deletes it from the process dictionary. Returns
<c>undefined</c> if no value is associated with <c>Key</c>.</p>
<pre>
> <input>put(key1, {merry, lambs, are, playing}),</input>
<input>X = erase(key1),</input>
<input>{X, erase(key1)}.</input>
{{merry,lambs,are,playing},undefined}</pre>
</desc>
</func>
<func>
<name>error(Reason)</name>
<fsummary>Stop execution with a given reason</fsummary>
<type>
<v>Reason = term()</v>
</type>
<desc>
<p>Stops the execution of the calling process with the reason
<c>Reason</c>, where <c>Reason</c> is any term. The actual
exit reason will be <c>{Reason, Where}</c>, where <c>Where</c>
is a list of the functions most recently called (the current
function first). Since evaluating this function causes
the process to terminate, it has no return value.</p>
<pre>
> <input>catch error(foobar).</input>
{'EXIT',{foobar,[{erl_eval,do_apply,5},
{erl_eval,expr,5},
{shell,exprs,6},
{shell,eval_exprs,6},
{shell,eval_loop,3}]}}</pre>
</desc>
</func>
<func>
<name>error(Reason, Args)</name>
<fsummary>Stop execution with a given reason</fsummary>
<type>
<v>Reason = term()</v>
<v>Args = [term()]</v>
</type>
<desc>
<p>Stops the execution of the calling process with the reason
<c>Reason</c>, where <c>Reason</c> is any term. The actual
exit reason will be <c>{Reason, Where}</c>, where <c>Where</c>
is a list of the functions most recently called (the current
function first). <c>Args</c> is expected to be the list of
arguments for the current function; in Beam it will be used
to provide the actual arguments for the current function in
the <c>Where</c> term. Since evaluating this function causes
the process to terminate, it has no return value.</p>
</desc>
</func>
<func>
<name>exit(Reason)</name>
<fsummary>Stop execution with a given reason</fsummary>
<type>
<v>Reason = term()</v>
</type>
<desc>
<p>Stops the execution of the calling process with the exit
reason <c>Reason</c>, where <c>Reason</c> is any term. Since
evaluating this function causes the process to terminate, it
has no return value.</p>
<pre>
> <input>exit(foobar).</input>
** exception exit: foobar
> <input>catch exit(foobar).</input>
{'EXIT',foobar}</pre>
</desc>
</func>
<func>
<name>exit(Pid, Reason) -> true</name>
<fsummary>Send an exit signal to a process</fsummary>
<type>
<v>Pid = pid()</v>
<v>Reason = term()</v>
</type>
<desc>
<p>Sends an exit signal with exit reason <c>Reason</c> to
the process <c>Pid</c>.</p>
<p>The following behavior apply if <c>Reason</c> is any term
except <c>normal</c> or <c>kill</c>:</p>
<p>If <c>Pid</c> is not trapping exits, <c>Pid</c> itself will
exit with exit reason <c>Reason</c>. If <c>Pid</c> is trapping
exits, the exit signal is transformed into a message
<c>{'EXIT', From, Reason}</c> and delivered to the message
queue of <c>Pid</c>. <c>From</c> is the pid of the process
which sent the exit signal. See also
<seealso marker="#process_flag/2">process_flag/2</seealso>.</p>
<p>If <c>Reason</c> is the atom <c>normal</c>, <c>Pid</c> will
not exit. If it is trapping exits, the exit signal is
transformed into a message <c>{'EXIT', From, normal}</c>
and delivered to its message queue.</p>
<p>If <c>Reason</c> is the atom <c>kill</c>, that is if
<c>exit(Pid, kill)</c> is called, an untrappable exit signal
is sent to <c>Pid</c> which will unconditionally exit with
exit reason <c>killed</c>.</p>
</desc>
</func>
<func>
<name>float(Number) -> float()</name>
<fsummary>Convert a number to a float</fsummary>
<type>
<v>Number = number()</v>
</type>
<desc>
<p>Returns a float by converting <c>Number</c> to a float.</p>
<pre>
> <input>float(55).</input>
55.0</pre>
<p>Allowed in guard tests.</p>
<note>
<p>Note that if used on the top-level in a guard, it will
test whether the argument is a floating point number; for
clarity, use
<seealso marker="#is_float/1">is_float/1</seealso> instead.</p>
<p>When <c>float/1</c> is used in an expression in a guard,
such as '<c>float(A) == 4.0</c>', it converts a number as
described above.</p>
</note>
</desc>
</func>
<func>
<name>float_to_list(Float) -> string()</name>
<fsummary>Text representation of a float</fsummary>
<type>
<v>Float = float()</v>
</type>
<desc>
<p>Returns a string which corresponds to the text
representation of <c>Float</c>.</p>
<pre>
> <input>float_to_list(7.0).</input>
"7.00000000000000000000e+00"</pre>
</desc>
</func>
<func>
<name>erlang:fun_info(Fun) -> [{Item, Info}]</name>
<fsummary>Information about a fun</fsummary>
<type>
<v>Fun = fun()</v>
<v>Item, Info -- see below</v>
</type>
<desc>
<p>Returns a list containing information about the fun
<c>Fun</c>. Each element of the list is a tuple. The order of
the tuples is not defined, and more tuples may be added in a
future release.</p>
<warning>
<p>This BIF is mainly intended for debugging, but it can
occasionally be useful in library functions that might need
to verify, for instance, the arity of a fun.</p>
</warning>
<p>There are two types of funs with slightly different
semantics:</p>
<p>A fun created by <c>fun M:F/A</c> is called an
<em>external</em> fun. Calling it will always call the
function <c>F</c> with arity <c>A</c> in the latest code for
module <c>M</c>. Note that module <c>M</c> does not even need
to be loaded when the fun <c>fun M:F/A</c> is created.</p>
<p>All other funs are called <em>local</em>. When a local fun
is called, the same version of the code that created the fun
will be called (even if newer version of the module has been
loaded).</p>
<p>The following elements will always be present in the list
for both local and external funs:</p>
<taglist>
<tag><c>{type, Type}</c></tag>
<item>
<p><c>Type</c> is either <c>local</c> or <c>external</c>.</p>
</item>
<tag><c>{module, Module}</c></tag>
<item>
<p><c>Module</c> (an atom) is the module name.</p>
<p>If <c>Fun</c> is a local fun, <c>Module</c> is the module
in which the fun is defined.</p>
<p>If <c>Fun</c> is an external fun, <c>Module</c> is the
module that the fun refers to.</p>
</item>
<tag><c>{name, Name}</c></tag>
<item>
<p><c>Name</c> (an atom) is a function name.</p>
<p>If <c>Fun</c> is a local fun, <c>Name</c> is the name
of the local function that implements the fun.
(This name was generated by the compiler, and is generally
only of informational use. As it is a local function, it
is not possible to call it directly.)
If no code is currently loaded for the fun, <c>[]</c>
will be returned instead of an atom.</p>
<p>If <c>Fun</c> is an external fun, <c>Name</c> is the name
of the exported function that the fun refers to.</p>
</item>
<tag><c>{arity, Arity}</c></tag>
<item>
<p><c>Arity</c> is the number of arguments that the fun
should be called with.</p>
</item>
<tag><c>{env, Env}</c></tag>
<item>
<p><c>Env</c> (a list) is the environment or free variables
for the fun. (For external funs, the returned list is
always empty.)</p>
</item>
</taglist>
<p>The following elements will only be present in the list if
<c>Fun</c> is local:</p>
<taglist>
<tag><c>{pid, Pid}</c></tag>
<item>
<p><c>Pid</c> is the pid of the process that originally
created the fun.</p>
</item>
<tag><c>{index, Index}</c></tag>
<item>
<p><c>Index</c> (an integer) is an index into the module's
fun table.</p>
</item>
<tag><c>{new_index, Index}</c></tag>
<item>
<p><c>Index</c> (an integer) is an index into the module's
fun table.</p>
</item>
<tag><c>{new_uniq, Uniq}</c></tag>
<item>
<p><c>Uniq</c> (a binary) is a unique value for this fun.</p>
</item>
<tag><c>{uniq, Uniq}</c></tag>
<item>
<p><c>Uniq</c> (an integer) is a unique value for this fun.</p>
</item>
</taglist>
</desc>
</func>
<func>
<name>erlang:fun_info(Fun, Item) -> {Item, Info}</name>
<fsummary>Information about a fun</fsummary>
<type>
<v>Fun = fun()</v>
<v>Item, Info -- see below</v>
</type>
<desc>
<p>Returns information about <c>Fun</c> as specified by
<c>Item</c>, in the form <c>{Item,Info}</c>.</p>
<p>For any fun, <c>Item</c> can be any of the atoms
<c>module</c>, <c>name</c>, <c>arity</c>, or <c>env</c>.</p>
<p>For a local fun, <c>Item</c> can also be any of the atoms
<c>index</c>, <c>new_index</c>, <c>new_uniq</c>,
<c>uniq</c>, and <c>pid</c>. For an external fun, the value
of any of these items is always the atom <c>undefined</c>.</p>
<p>See
<seealso marker="#fun_info/1">erlang:fun_info/1</seealso>.</p>
</desc>
</func>
<func>
<name>erlang:fun_to_list(Fun) -> string()</name>
<fsummary>Text representation of a fun</fsummary>
<type>
<v>Fun = fun()</v>
</type>
<desc>
<p>Returns a string which corresponds to the text
representation of <c>Fun</c>.</p>
</desc>
</func>
<func>
<name>erlang:function_exported(Module, Function, Arity) -> bool()</name>
<fsummary>Check if a function is exported and loaded</fsummary>
<type>
<v>Module = Function = atom()</v>
<v>Arity = int()</v>
</type>
<desc>
<p>Returns <c>true</c> if the module <c>Module</c> is loaded
and contains an exported function <c>Function/Arity</c>;
otherwise <c>false</c>.</p>
<p>Returns <c>false</c> for any BIF (functions implemented in C
rather than in Erlang).</p>
</desc>
</func>
<func>
<name>garbage_collect() -> true</name>
<fsummary>Force an immediate garbage collection of the calling process</fsummary>
<desc>
<p>Forces an immediate garbage collection of the currently
executing process. The function should not be used, unless
it has been noticed -- or there are good reasons to suspect --
that the spontaneous garbage collection will occur too late
or not at all. Improper use may seriously degrade system
performance.</p>
<p>Compatibility note: In versions of OTP prior to R7,
the garbage collection took place at the next context switch,
not immediately. To force a context switch after a call to
<c>erlang:garbage_collect()</c>, it was sufficient to make
any function call.</p>
</desc>
</func>
<func>
<name>garbage_collect(Pid) -> bool()</name>
<fsummary>Force an immediate garbage collection of a process</fsummary>
<type>
<v>Pid = pid()</v>
</type>
<desc>
<p>Works like <c>erlang:garbage_collect()</c> but on any
process. The same caveats apply. Returns <c>false</c> if
<c>Pid</c> refers to a dead process; <c>true</c> otherwise.</p>
</desc>
</func>
<func>
<name>get() -> [{Key, Val}]</name>
<fsummary>Return the process dictionary</fsummary>
<type>
<v>Key = Val = term()</v>
</type>
<desc>
<p>Returns the process dictionary as a list of
<c>{Key, Val}</c> tuples.</p>
<pre>
> <input>put(key1, merry),</input>
<input>put(key2, lambs),</input>
<input>put(key3, {are, playing}),</input>
<input>get().</input>
[{key1,merry},{key2,lambs},{key3,{are,playing}}]</pre>
</desc>
</func>
<func>
<name>get(Key) -> Val | undefined</name>
<fsummary>Return a value from the process dictionary</fsummary>
<type>
<v>Key = Val = term()</v>
</type>
<desc>
<p>Returns the value <c>Val</c>associated with <c>Key</c> in
the process dictionary, or <c>undefined</c> if <c>Key</c>
does not exist.</p>
<pre>
> <input>put(key1, merry),</input>
<input>put(key2, lambs),</input>
<input>put({any, [valid, term]}, {are, playing}),</input>
<input>get({any, [valid, term]}).</input>
{are,playing}</pre>
</desc>
</func>
<func>
<name>erlang:get_cookie() -> Cookie | nocookie</name>
<fsummary>Get the magic cookie of the local node</fsummary>
<type>
<v>Cookie = atom()</v>
</type>
<desc>
<p>Returns the magic cookie of the local node, if the node is
alive; otherwise the atom <c>nocookie</c>.</p>
</desc>
</func>
<func>
<name>get_keys(Val) -> [Key]</name>
<fsummary>Return a list of keys from the process dictionary</fsummary>
<type>
<v>Val = Key = term()</v>
</type>
<desc>
<p>Returns a list of keys which are associated with the value
<c>Val</c> in the process dictionary.</p>
<pre>
> <input>put(mary, {1, 2}),</input>
<input>put(had, {1, 2}),</input>
<input>put(a, {1, 2}),</input>
<input>put(little, {1, 2}),</input>
<input>put(dog, {1, 3}),</input>
<input>put(lamb, {1, 2}),</input>
<input>get_keys({1, 2}).</input>
[mary,had,a,little,lamb]</pre>
</desc>
</func>
<func>
<name>erlang:get_stacktrace() -> [{Module, Function, Arity | Args}]</name>
<fsummary>Get the call stack back-trace of the last exception</fsummary>
<type>
<v>Module = Function = atom()</v>
<v>Arity = int()</v>
<v>Args = [term()]</v>
</type>
<desc>
<p>Get the call stack back-trace (<em>stacktrace</em>) of the last
exception in the calling process as a list of
<c>{Module,Function,Arity}</c> tuples.
The <c>Arity</c> field in the first tuple may be the argument
list of that function call instead of an arity integer,
depending on the exception.</p>
<p>If there has not been any exceptions in a process, the
stacktrace is []. After a code change for the process,
the stacktrace may also be reset to [].</p>
<p>The stacktrace is the same data as the <c>catch</c> operator
returns, for example:</p>
<p><c>{'EXIT',{badarg,Stacktrace}} = catch abs(x)</c></p>
<p>See also
<seealso marker="#error/1">erlang:error/1</seealso> and
<seealso marker="#error/2">erlang:error/2</seealso>.</p>
</desc>
</func>
<func>
<name>group_leader() -> GroupLeader</name>
<fsummary>Get the group leader for the calling process</fsummary>
<type>
<v>GroupLeader = pid()</v>
</type>
<desc>
<p>Returns the pid of the group leader for the process which
evaluates the function.</p>
<p>Every process is a member of some process group and all
groups have a <em>group leader</em>. All IO from the group
is channeled to the group leader. When a new process is
spawned, it gets the same group leader as the spawning
process. Initially, at system start-up, <c>init</c> is both
its own group leader and the group leader of all processes.</p>
</desc>
</func>
<func>
<name>group_leader(GroupLeader, Pid) -> true</name>
<fsummary>Set the group leader for a process</fsummary>
<type>
<v>GroupLeader = Pid = pid()</v>
</type>
<desc>
<p>Sets the group leader of <c>Pid</c> to <c>GroupLeader</c>.
Typically, this is used when a processes started from a
certain shell should have another group leader than
<c>init</c>.</p>
<p>See also
<seealso marker="#group_leader/0">group_leader/0</seealso>.</p>
</desc>
</func>
<func>
<name>halt()</name>
<fsummary>Halt the Erlang runtime system and indicate normal exit to the calling environment</fsummary>
<desc>
<p>Halts the Erlang runtime system and indicates normal exit to
the calling environment. Has no return value.</p>
<pre>
> <input>halt().</input>
os_prompt%</pre>
</desc>
</func>
<func>
<name>halt(Status)</name>
<fsummary>Halt the Erlang runtime system</fsummary>
<type>
<v>Status = int()>=0 | string()</v>
</type>
<desc>
<p><c>Status</c> must be a non-negative integer, or a string.
Halts the Erlang runtime system. Has no return value.
If <c>Status</c> is an integer, it is returned as an exit
status of Erlang to the calling environment.
If <c>Status</c> is a string, produces an Erlang crash dump
with <c>String</c> as slogan, and then exits with a non-zero
status code.</p>
<p>Note that on many platforms, only the status codes 0-255 are
supported by the operating system.</p>
</desc>
</func>
<func>
<name>erlang:hash(Term, Range) -> Hash</name>
<fsummary>Hash function (deprecated)</fsummary>
<desc>
<p>Returns a hash value for <c>Term</c> within the range
<c>1..Range</c>. The allowed range is 1..2^27-1.</p>
<warning>
<p>This BIF is deprecated as the hash value may differ on
different architectures. Also the hash values for integer
terms larger than 2^27 as well as large binaries are very
poor. The BIF is retained for backward compatibility
reasons (it may have been used to hash records into a file),
but all new code should use one of the BIFs
<c>erlang:phash/2</c> or <c>erlang:phash2/1,2</c> instead.</p>
</warning>
</desc>
</func>
<func>
<name>hd(List) -> term()</name>
<fsummary>Head of a list</fsummary>
<type>
<v>List = [term()]</v>
</type>
<desc>
<p>Returns the head of <c>List</c>, that is, the first element.</p>
<pre>
> <input>hd([1,2,3,4,5]).</input>
1</pre>
<p>Allowed in guard tests.</p>
<p>Failure: <c>badarg</c> if <c>List</c> is the empty list [].</p>
</desc>
</func>
<func>
<name>erlang:hibernate(Module, Function, Args)</name>
<fsummary>Hibernate a process until a message is sent to it</fsummary>
<type>
<v>Module = Function = atom()</v>
<v>Args = [term()]</v>
</type>
<desc>
<p>Puts the calling process into a wait state where its memory
allocation has been reduced as much as possible, which is
useful if the process does not expect to receive any messages
in the near future.</p>
<p>The process will be awaken when a message is sent to it, and
control will resume in <c>Module:Function</c> with
the arguments given by <c>Args</c> with the call stack
emptied, meaning that the process will terminate when that
function returns. Thus <c>erlang:hibernate/3</c> will never
return to its caller.</p>
<p>If the process has any message in its message queue,
the process will be awaken immediately in the same way as
described above.</p>
<p>In more technical terms, what <c>erlang:hibernate/3</c> does
is the following. It discards the call stack for the process.
Then it garbage collects the process. After the garbage
collection, all live data is in one continuous heap. The heap
is then shrunken to the exact same size as the live data
which it holds (even if that size is less than the minimum
heap size for the process).</p>
<p>If the size of the live data in the process is less than
the minimum heap size, the first garbage collection occurring
after the process has been awaken will ensure that the heap
size is changed to a size not smaller than the minimum heap
size.</p>
<p>Note that emptying the call stack means that any surrounding
<c>catch</c> is removed and has to be re-inserted after
hibernation. One effect of this is that processes started
using <c>proc_lib</c> (also indirectly, such as
<c>gen_server</c> processes), should use
<seealso marker="stdlib:proc_lib#hibernate/3">proc_lib:hibernate/3</seealso>
instead to ensure that the exception handler continues to work
when the process wakes up.</p>
</desc>
</func>
<func>
<name>integer_to_list(Integer) -> string()</name>
<fsummary>Text representation of an integer</fsummary>
<type>
<v>Integer = int()</v>
</type>
<desc>
<p>Returns a string which corresponds to the text
representation of <c>Integer</c>.</p>
<pre>
> <input>integer_to_list(77).</input>
"77"</pre>
</desc>
</func>
<func>
<name>integer_to_list(Integer, Base) -> string()</name>
<fsummary>Text representation of an integer</fsummary>
<type>
<v>Integer = int()</v>
<v>Base = 2..36</v>
</type>
<desc>
<p>Returns a string which corresponds to the text
representation of <c>Integer</c> in base <c>Base</c>.</p>
<pre>
> <input>integer_to_list(1023, 16).</input>
"3FF"</pre>
</desc>
</func>
<func>
<name>iolist_to_binary(IoListOrBinary) -> binary()</name>
<fsummary>Convert an iolist to a binary</fsummary>
<type>
<v>IoListOrBinary = iolist() | binary()</v>
</type>
<desc>
<p>Returns a binary which is made from the integers and
binaries in <c>IoListOrBinary</c>.</p>
<pre>
> <input>Bin1 = <<1,2,3>>.</input>
<<1,2,3>>
> <input>Bin2 = <<4,5>>.</input>
<<4,5>>
> <input>Bin3 = <<6>>.</input>
<<6>>
> <input>iolist_to_binary([Bin1,1,[2,3,Bin2],4|Bin3]).</input>
<<1,2,3,1,2,3,4,5,4,6>></pre>
</desc>
</func>
<func>
<name>iolist_size(Item) -> int()</name>
<fsummary>Size of an iolist</fsummary>
<type>
<v>Item = iolist() | binary()</v>
</type>
<desc>
<p>Returns an integer which is the size in bytes
of the binary that would be the result of
<c>iolist_to_binary(Item)</c>.</p>
<pre>
> <input>iolist_size([1,2|<<3,4>>]).</input>
4</pre>
</desc>
</func>
<func>
<name>is_alive() -> bool()</name>
<fsummary>Check whether the local node is alive</fsummary>
<desc>
<p>Returns <c>true</c> if the local node is alive; that is, if
the node can be part of a distributed system. Otherwise, it
returns <c>false</c>.</p>
</desc>
</func>
<func>
<name>is_atom(Term) -> bool()</name>
<fsummary>Check whether a term is an atom</fsummary>
<type>
<v>Term = term()</v>
</type>
<desc>
<p>Returns <c>true</c> if <c>Term</c> is an atom;
otherwise returns <c>false</c>.</p>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>is_binary(Term) -> bool()</name>
<fsummary>Check whether a term is a binary</fsummary>
<type>
<v>Term = term()</v>
</type>
<desc>
<p>Returns <c>true</c> if <c>Term</c> is a binary;
otherwise returns <c>false</c>.</p>
<p>A binary always contains a complete number of bytes.</p>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>is_bitstring(Term) -> bool()</name>
<fsummary>Check whether a term is a bitstring</fsummary>
<type>
<v>Term = term()</v>
</type>
<desc>
<p>Returns <c>true</c> if <c>Term</c> is a bitstring (including a binary);
otherwise returns <c>false</c>.</p>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>is_boolean(Term) -> bool()</name>
<fsummary>Check whether a term is a boolean</fsummary>
<type>
<v>Term = term()</v>
</type>
<desc>
<p>Returns <c>true</c> if <c>Term</c> is
either the atom <c>true</c> or the atom <c>false</c>
(i.e. a boolean); otherwise returns <c>false</c>.</p>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>erlang:is_builtin(Module, Function, Arity) -> bool()</name>
<fsummary>Check if a function is a BIF implemented in C</fsummary>
<type>
<v>Module = Function = atom()</v>
<v>Arity = int()</v>
</type>
<desc>
<p>Returns <c>true</c> if <c>Module:Function/Arity</c> is
a BIF implemented in C; otherwise returns <c>false</c>.
This BIF is useful for builders of cross reference tools.</p>
</desc>
</func>
<func>
<name>is_float(Term) -> bool()</name>
<fsummary>Check whether a term is a float</fsummary>
<type>
<v>Term = term()</v>
</type>
<desc>
<p>Returns <c>true</c> if <c>Term</c> is a floating point
number; otherwise returns <c>false</c>.</p>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>is_function(Term) -> bool()</name>
<fsummary>Check whether a term is a fun</fsummary>
<type>
<v>Term = term()</v>
</type>
<desc>
<p>Returns <c>true</c> if <c>Term</c> is a fun; otherwise
returns <c>false</c>.</p>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>is_function(Term, Arity) -> bool()</name>
<fsummary>Check whether a term is a fun with a given arity</fsummary>
<type>
<v>Term = term()</v>
<v>Arity = int()</v>
</type>
<desc>
<p>Returns <c>true</c> if <c>Term</c> is a fun that can be
applied with <c>Arity</c> number of arguments; otherwise
returns <c>false</c>.</p>
<p>Allowed in guard tests.</p>
<warning>
<p>Currently, <c>is_function/2</c> will also return
<c>true</c> if the first argument is a tuple fun (a tuple
containing two atoms). In a future release, tuple funs will
no longer be supported and <c>is_function/2</c> will return
<c>false</c> if given a tuple fun.</p>
</warning>
</desc>
</func>
<func>
<name>is_integer(Term) -> bool()</name>
<fsummary>Check whether a term is an integer</fsummary>
<type>
<v>Term = term()</v>
</type>
<desc>
<p>Returns <c>true</c> if <c>Term</c> is an integer;
otherwise returns <c>false</c>.</p>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>is_list(Term) -> bool()</name>
<fsummary>Check whether a term is a list</fsummary>
<type>
<v>Term = term()</v>
</type>
<desc>
<p>Returns <c>true</c> if <c>Term</c> is a list with
zero or more elements; otherwise returns <c>false</c>.</p>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>is_number(Term) -> bool()</name>
<fsummary>Check whether a term is a number</fsummary>
<type>
<v>Term = term()</v>
</type>
<desc>
<p>Returns <c>true</c> if <c>Term</c> is either an integer or a
floating point number; otherwise returns <c>false</c>.</p>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>is_pid(Term) -> bool()</name>
<fsummary>Check whether a term is a pid</fsummary>
<type>
<v>Term = term()</v>
</type>
<desc>
<p>Returns <c>true</c> if <c>Term</c> is a pid (process
identifier); otherwise returns <c>false</c>.</p>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>is_port(Term) -> bool()</name>
<fsummary>Check whether a term is a port</fsummary>
<type>
<v>Term = term()</v>
</type>
<desc>
<p>Returns <c>true</c> if <c>Term</c> is a port identifier;
otherwise returns <c>false</c>.</p>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>is_process_alive(Pid) -> bool()</name>
<fsummary>Check whether a process is alive</fsummary>
<type>
<v>Pid = pid()</v>
</type>
<desc>
<p>
<c>Pid</c> must refer to a process at the local node.
Returns <c>true</c> if the process exists and is alive, that
is, is not exiting and has not exited. Otherwise, returns
<c>false</c>.
</p>
</desc>
</func>
<func>
<name>is_record(Term, RecordTag) -> bool()</name>
<fsummary>Check whether a term appears to be a record</fsummary>
<type>
<v>Term = term()</v>
<v>RecordTag = atom()</v>
</type>
<desc>
<p>Returns <c>true</c> if <c>Term</c> is a tuple and its first
element is <c>RecordTag</c>. Otherwise, returns <c>false</c>.</p>
<note>
<p>Normally the compiler treats calls to <c>is_record/2</c>
specially. It emits code to verify that <c>Term</c> is a
tuple, that its first element is <c>RecordTag</c>, and that
the size is correct. However, if the <c>RecordTag</c> is
not a literal atom, the <c>is_record/2</c> BIF will be
called instead and the size of the tuple will not be
verified.</p>
</note>
<p>Allowed in guard tests, if <c>RecordTag</c> is a literal
atom.</p>
</desc>
</func>
<func>
<name>is_record(Term, RecordTag, Size) -> bool()</name>
<fsummary>Check whether a term appears to be a record</fsummary>
<type>
<v>Term = term()</v>
<v>RecordTag = atom()</v>
<v>Size = int()</v>
</type>
<desc>
<p><c>RecordTag</c> must be an atom. Returns <c>true</c> if
<c>Term</c> is a tuple, its first element is <c>RecordTag</c>,
and its size is <c>Size</c>. Otherwise, returns <c>false</c>.</p>
<p>Allowed in guard tests, provided that <c>RecordTag</c> is
a literal atom and <c>Size</c> is a literal integer.</p>
<note>
<p>This BIF is documented for completeness. In most cases
<c>is_record/2</c> should be used.</p>
</note>
</desc>
</func>
<func>
<name>is_reference(Term) -> bool()</name>
<fsummary>Check whether a term is a reference</fsummary>
<type>
<v>Term = term()</v>
</type>
<desc>
<p>Returns <c>true</c> if <c>Term</c> is a reference;
otherwise returns <c>false</c>.</p>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>is_tuple(Term) -> bool()</name>
<fsummary>Check whether a term is a tuple</fsummary>
<type>
<v>Term = term()</v>
</type>
<desc>
<p>Returns <c>true</c> if <c>Term</c> is a tuple;
otherwise returns <c>false</c>.</p>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>length(List) -> int()</name>
<fsummary>Length of a list</fsummary>
<type>
<v>List = [term()]</v>
</type>
<desc>
<p>Returns the length of <c>List</c>.</p>
<pre>
> <input>length([1,2,3,4,5,6,7,8,9]).</input>
9</pre>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>link(Pid) -> true</name>
<fsummary>Create a link to another process (or port)</fsummary>
<type>
<v>Pid = pid() | port()</v>
</type>
<desc>
<p>Creates a link between the calling process and another
process (or port) <c>Pid</c>, if there is not such a link
already. If a process attempts to create a link to itself,
nothing is done. Returns <c>true</c>.</p>
<p>If <c>Pid</c> does not exist, the behavior of the BIF depends
on if the calling process is trapping exits or not (see
<seealso marker="#process_flag/2">process_flag/2</seealso>):</p>
<list type="bulleted">
<item>If the calling process is not trapping exits, and
checking <c>Pid</c> is cheap -- that is, if <c>Pid</c> is
local -- <c>link/1</c> fails with reason <c>noproc</c>.</item>
<item>Otherwise, if the calling process is trapping exits,
and/or <c>Pid</c> is remote, <c>link/1</c> returns
<c>true</c>, but an exit signal with reason <c>noproc</c>
is sent to the calling process.</item>
</list>
</desc>
</func>
<func>
<name>list_to_atom(String) -> atom()</name>
<fsummary>Convert from text representation to an atom</fsummary>
<type>
<v>String = string()</v>
</type>
<desc>
<p>Returns the atom whose text representation is <c>String</c>.</p>
<pre>
> <input>list_to_atom("Erlang").</input>
'Erlang'</pre>
</desc>
</func>
<func>
<name>list_to_binary(IoList) -> binary()</name>
<fsummary>Convert a list to a binary</fsummary>
<type>
<v>IoList = iolist()</v>
</type>
<desc>
<p>Returns a binary which is made from the integers and
binaries in <c>IoList</c>.</p>
<pre>
> <input>Bin1 = <<1,2,3>>.</input>
<<1,2,3>>
> <input>Bin2 = <<4,5>>.</input>
<<4,5>>
> <input>Bin3 = <<6>>.</input>
<<6>>
> <input>list_to_binary([Bin1,1,[2,3,Bin2],4|Bin3]).</input>
<<1,2,3,1,2,3,4,5,4,6>></pre>
</desc>
</func>
<func>
<name>list_to_bitstring(BitstringList) -> bitstring()</name>
<fsummary>Convert a list to a bitstring</fsummary>
<type>
<v>BitstringList = [BitstringList | bitstring() | char()]</v>
</type>
<desc>
<p>Returns a bitstring which is made from the integers and
bitstrings in <c>BitstringList</c>. (The last tail in <c>BitstringList</c>
is allowed to be a bitstring.)</p>
<pre>
> <input>Bin1 = <<1,2,3>>.</input>
<<1,2,3>>
> <input>Bin2 = <<4,5>>.</input>
<<4,5>>
> <input>Bin3 = <<6,7:4,>>.</input>
<<6>>
> <input>list_to_binary([Bin1,1,[2,3,Bin2],4|Bin3]).</input>
<<1,2,3,1,2,3,4,5,4,6,7:46>></pre>
</desc>
</func>
<func>
<name>list_to_existing_atom(String) -> atom()</name>
<fsummary>Convert from text representation to an atom</fsummary>
<type>
<v>String = string()</v>
</type>
<desc>
<p>Returns the atom whose text representation is <c>String</c>,
but only if there already exists such atom.</p>
<p>Failure: <c>badarg</c> if there does not already exist an atom
whose text representation is <c>String</c>.</p>
</desc>
</func>
<func>
<name>list_to_float(String) -> float()</name>
<fsummary>Convert from text representation to a float</fsummary>
<type>
<v>String = string()</v>
</type>
<desc>
<p>Returns the float whose text representation is <c>String</c>.</p>
<pre>
> <input>list_to_float("2.2017764e+0").</input>
2.2017764</pre>
<p>Failure: <c>badarg</c> if <c>String</c> contains a bad
representation of a float.</p>
</desc>
</func>
<func>
<name>list_to_integer(String) -> int()</name>
<fsummary>Convert from text representation to an integer</fsummary>
<type>
<v>String = string()</v>
</type>
<desc>
<p>Returns an integer whose text representation is
<c>String</c>.</p>
<pre>
> <input>list_to_integer("123").</input>
123</pre>
<p>Failure: <c>badarg</c> if <c>String</c> contains a bad
representation of an integer.</p>
</desc>
</func>
<func>
<name>list_to_integer(String, Base) -> int()</name>
<fsummary>Convert from text representation to an integer</fsummary>
<type>
<v>String = string()</v>
<v>Base = 2..36</v>
</type>
<desc>
<p>Returns an integer whose text representation in base
<c>Base</c> is <c>String</c>.</p>
<pre>
> <input>list_to_integer("3FF", 16).</input>
1023</pre>
<p>Failure: <c>badarg</c> if <c>String</c> contains a bad
representation of an integer.</p>
</desc>
</func>
<func>
<name>list_to_pid(String) -> pid()</name>
<fsummary>Convert from text representation to a pid</fsummary>
<type>
<v>String = string()</v>
</type>
<desc>
<p>Returns a pid whose text representation is <c>String</c>.</p>
<warning>
<p>This BIF is intended for debugging and for use in
the Erlang operating system. It should not be used in
application programs.</p>
</warning>
<pre>
> <input>list_to_pid("<0.4.1>").</input>
<0.4.1></pre>
<p>Failure: <c>badarg</c> if <c>String</c> contains a bad
representation of a pid.</p>
</desc>
</func>
<func>
<name>list_to_tuple(List) -> tuple()</name>
<fsummary>Convert a list to a tuple</fsummary>
<type>
<v>List = [term()]</v>
</type>
<desc>
<p>Returns a tuple which corresponds to <c>List</c>. <c>List</c>
can contain any Erlang terms.</p>
<pre>
> <input>list_to_tuple([share, ['Ericsson_B', 163]]).</input>
{share, ['Ericsson_B', 163]}</pre>
</desc>
</func>
<func>
<name>load_module(Module, Binary) -> {module, Module} | {error, Reason}</name>
<fsummary>Load object code for a module</fsummary>
<type>
<v>Module = atom()</v>
<v>Binary = binary()</v>
<v>Reason = badfile | not_purged | badfile</v>
</type>
<desc>
<p>If <c>Binary</c> contains the object code for the module
<c>Module</c>, this BIF loads that object code. Also, if
the code for the module <c>Module</c> already exists, all
export references are replaced so they point to the newly
loaded code. The previously loaded code is kept in the system
as old code, as there may still be processes which are
executing that code. It returns either
<c>{module, Module}</c>, or <c>{error, Reason}</c> if loading
fails. <c>Reason</c> is one of the following:</p>
<taglist>
<tag><c>badfile</c></tag>
<item>
<p>The object code in <c>Binary</c> has an incorrect format.</p>
</item>
<tag><c>not_purged</c></tag>
<item>
<p><c>Binary</c> contains a module which cannot be loaded
because old code for this module already exists.</p>
</item>
<tag><c>badfile</c></tag>
<item>
<p>The object code contains code for another module than
<c>Module</c></p>
</item>
</taglist>
<warning>
<p>This BIF is intended for the code server (see
<seealso marker="kernel:code">code(3)</seealso>) and should not be
used elsewhere.</p>
</warning>
</desc>
</func>
<func>
<name>erlang:load_nif(Path, LoadInfo) -> ok | {error, {Reason, Text}}</name>
<fsummary>Load NIF library</fsummary>
<type>
<v>Path = string()</v>
<v>LoadInfo = term()</v>
<v>Reason = load_failed | bad_lib | load | reload |
upgrade | old_code</v>
<v>Text = string()</v>
</type>
<desc>
<note>
<p>In releases older than OTP R14B, NIFs were an
experimental feature. Versions of OTP older than R14B might
have different and possibly incompatible NIF semantics and
interfaces. For example, in R13B03 the return value on
failure was
<c>{error,Reason,Text}</c>.</p>
</note>
<p>Loads and links a dynamic library containing native
implemented functions (NIFs) for a module. <c>Path</c> is a
file path to the sharable object/dynamic library file minus
the OS-dependent file extension (.so for Unix and .dll for
Windows). See <seealso marker="erl_nif">erl_nif</seealso>
on how to implement a NIF library.</p>
<p><c>LoadInfo</c> can be any term. It will be passed on to
the library as part of the initialization. A good practice is
to include a module version number to support future code
upgrade scenarios.</p>
<p>The call to <c>load_nif/2</c> must be made
<em>directly</em> from the Erlang code of the module that the
NIF library belongs to.</p>
<p>It returns either <c>ok</c>, or <c>{error,{Reason,Text}}</c>
if loading fails. <c>Reason</c> is one of the atoms below,
while <c>Text</c> is a human readable string that may give
some more information about the failure.</p>
<taglist>
<tag><c>load_failed</c></tag>
<item>
<p>The OS failed to load the NIF library.</p>
</item>
<tag><c>bad_lib</c></tag>
<item>
<p>The library did not fulfil the requirements as a NIF
library of the calling module.</p>
</item>
<tag><c>load | reload | upgrade</c></tag>
<item>
<p>The corresponding library callback was not successful.</p>
</item>
<tag><c>old_code</c></tag>
<item>
<p>The call to <c>load_nif/2</c> was made from the old
code of a module that has been upgraded. This is not
allowed.</p>
</item>
</taglist>
</desc>
</func>
<func>
<name>erlang:loaded() -> [Module]</name>
<fsummary>List of all loaded modules</fsummary>
<type>
<v>Module = atom()</v>
</type>
<desc>
<p>Returns a list of all loaded Erlang modules (current and/or
old code), including preloaded modules.</p>
<p>See also <seealso marker="kernel:code">code(3)</seealso>.</p>
</desc>
</func>
<func>
<name>erlang:localtime() -> {Date, Time}</name>
<fsummary>Current local date and time</fsummary>
<type>
<v>Date = {Year, Month, Day}</v>
<v>Time = {Hour, Minute, Second}</v>
<v> Year = Month = Day = Hour = Minute = Second = int()</v>
</type>
<desc>
<p>Returns the current local date and time
<c>{{Year, Month, Day}, {Hour, Minute, Second}}</c>.</p>
<p>The time zone and daylight saving time correction depend
on the underlying OS.</p>
<pre>
> <input>erlang:localtime().</input>
{{1996,11,6},{14,45,17}}</pre>
</desc>
</func>
<func>
<name>erlang:localtime_to_universaltime({Date1, Time1}) -> {Date2, Time2}</name>
<fsummary>Convert from local to Universal Time Coordinated (UTC) date and time</fsummary>
<type>
<v>Date1 = Date2 = {Year, Month, Day}</v>
<v>Time1 = Time2 = {Hour, Minute, Second}</v>
<v> Year = Month = Day = Hour = Minute = Second = int()</v>
</type>
<desc>
<p>Converts local date and time to Universal Time Coordinated
(UTC), if this is supported by the underlying OS. Otherwise,
no conversion is done and <c>{Date1, Time1}</c> is returned.</p>
<pre>
> <input>erlang:localtime_to_universaltime({{1996,11,6},{14,45,17}}).</input>
{{1996,11,6},{13,45,17}}</pre>
<p>Failure: <c>badarg</c> if <c>Date1</c> or <c>Time1</c> do
not denote a valid date or time.</p>
</desc>
</func>
<func>
<name>erlang:localtime_to_universaltime({Date1, Time1}, IsDst) -> {Date2, Time2}</name>
<fsummary>Convert from local to Universal Time Coordinated (UTC) date and time</fsummary>
<type>
<v>Date1 = Date2 = {Year, Month, Day}</v>
<v>Time1 = Time2 = {Hour, Minute, Second}</v>
<v> Year = Month = Day = Hour = Minute = Second = int()</v>
<v>IsDst = true | false | undefined</v>
</type>
<desc>
<p>Converts local date and time to Universal Time Coordinated
(UTC) just like <c>erlang:localtime_to_universaltime/1</c>,
but the caller decides if daylight saving time is active or
not.</p>
<p>If <c>IsDst == true</c> the <c>{Date1, Time1}</c> is during
daylight saving time, if <c>IsDst == false</c> it is not,
and if <c>IsDst == undefined</c> the underlying OS may
guess, which is the same as calling
<c>erlang:localtime_to_universaltime({Date1, Time1})</c>.</p>
<pre>
> <input>erlang:localtime_to_universaltime({{1996,11,6},{14,45,17}}, true).</input>
{{1996,11,6},{12,45,17}}
> <input>erlang:localtime_to_universaltime({{1996,11,6},{14,45,17}}, false).</input>
{{1996,11,6},{13,45,17}}
> <input>erlang:localtime_to_universaltime({{1996,11,6},{14,45,17}}, undefined).</input>
{{1996,11,6},{13,45,17}}</pre>
<p>Failure: <c>badarg</c> if <c>Date1</c> or <c>Time1</c> do
not denote a valid date or time.</p>
</desc>
</func>
<func>
<name>make_ref() -> reference()</name>
<fsummary>Return an almost unique reference</fsummary>
<desc>
<p>Returns an almost unique reference.</p>
<p>The returned reference will re-occur after approximately 2^82
calls; therefore it is unique enough for practical purposes.</p>
<pre>
> <input>make_ref().</input>
#Ref<0.0.0.135></pre>
</desc>
</func>
<func>
<name>erlang:make_tuple(Arity, InitialValue) -> tuple()</name>
<fsummary>Create a new tuple of a given arity</fsummary>
<type>
<v>Arity = int()</v>
<v>InitialValue = term()</v>
</type>
<desc>
<p>Returns a new tuple of the given <c>Arity</c>, where all
elements are <c>InitialValue</c>.</p>
<pre>
> <input>erlang:make_tuple(4, []).</input>
{[],[],[],[]}</pre>
</desc>
</func>
<func>
<name>erlang:make_tuple(Arity, Default, InitList) -> tuple()</name>
<fsummary>Create a new tuple with given arity and contents</fsummary>
<type>
<v>Arity = int()</v>
<v>Default = term()</v>
<v>InitList = [{Position,term()}]</v>
<v>Position = integer()</v>
</type>
<desc>
<p><c>erlang:make_tuple</c> first creates a tuple of size <c>Arity</c>
where each element has the value <c>Default</c>. It then fills
in values from <c>InitList</c>. Each list element in <c>InitList</c>
must be a two-tuple where the first element is a position in the
newly created tuple and the second element is any term. If a position
occurs more than once in the list, the term corresponding to
last occurrence will be used.</p>
<pre>
> <input>erlang:make_tuple(5, [], [{2,ignored},{5,zz},{2,aa}]).</input>
{{[],aa,[],[],zz}</pre>
</desc>
</func>
<func>
<name>max(Term1, Term2) -> Maximum</name>
<fsummary>Return the largest of two term</fsummary>
<type>
<v>Term1 = Term2 = Maximum = term()</v>
</type>
<desc>
<p>Return the largest of <c>Term1</c> and <c>Term2</c>;
if the terms compares equal, <c>Term1</c> will be returned.</p>
</desc>
</func>
<func>
<name>erlang:md5(Data) -> Digest</name>
<fsummary>Compute an MD5 message digest</fsummary>
<type>
<v>Data = iodata()</v>
<v>Digest = binary()</v>
</type>
<desc>
<p>Computes an <c>MD5</c> message digest from <c>Data</c>, where
the length of the digest is 128 bits (16 bytes). <c>Data</c>
is a binary or a list of small integers and binaries.</p>
<p>See The MD5 Message Digest Algorithm (RFC 1321) for more
information about MD5.</p>
<warning><p>The MD5 Message Digest Algorithm is <em>not</em> considered
safe for code-signing or software integrity purposes.</p></warning>
</desc>
</func>
<func>
<name>erlang:md5_final(Context) -> Digest</name>
<fsummary>Finish the update of an MD5 context and return the computed MD5 message digest</fsummary>
<type>
<v>Context = Digest = binary()</v>
</type>
<desc>
<p>Finishes the update of an MD5 <c>Context</c> and returns
the computed <c>MD5</c> message digest.</p>
</desc>
</func>
<func>
<name>erlang:md5_init() -> Context</name>
<fsummary>Create an MD5 context</fsummary>
<type>
<v>Context = binary()</v>
</type>
<desc>
<p>Creates an MD5 context, to be used in subsequent calls to
<c>md5_update/2</c>.</p>
</desc>
</func>
<func>
<name>erlang:md5_update(Context, Data) -> NewContext</name>
<fsummary>Update an MD5 context with data, and return a new context</fsummary>
<type>
<v>Data = iodata()</v>
<v>Context = NewContext = binary()</v>
</type>
<desc>
<p>Updates an MD5 <c>Context</c> with <c>Data</c>, and returns
a <c>NewContext</c>.</p>
</desc>
</func>
<func>
<name>erlang:memory() -> [{Type, Size}]</name>
<fsummary>Information about dynamically allocated memory</fsummary>
<type>
<v>Type, Size -- see below</v>
</type>
<desc>
<p>Returns a list containing information about memory
dynamically allocated by the Erlang emulator. Each element of
the list is a tuple <c>{Type, Size}</c>. The first element
<c>Type</c>is an atom describing memory type. The second
element <c>Size</c>is memory size in bytes. A description of
each memory type follows:</p>
<taglist>
<tag><c>total</c></tag>
<item>
<p>The total amount of memory currently allocated, which is
the same as the sum of memory size for <c>processes</c>
and <c>system</c>.</p>
</item>
<tag><c>processes</c></tag>
<item>
<p>The total amount of memory currently allocated by
the Erlang processes.</p>
</item>
<tag><c>processes_used</c></tag>
<item>
<p>The total amount of memory currently used by the Erlang
processes.</p>
<p>This memory is part of the memory presented as
<c>processes</c> memory.</p>
</item>
<tag><c>system</c></tag>
<item>
<p>The total amount of memory currently allocated by
the emulator that is not directly related to any Erlang
process.</p>
<p>Memory presented as <c>processes</c> is not included in
this memory.</p>
</item>
<tag><c>atom</c></tag>
<item>
<p>The total amount of memory currently allocated for atoms.</p>
<p>This memory is part of the memory presented as
<c>system</c> memory.</p>
</item>
<tag><c>atom_used</c></tag>
<item>
<p>The total amount of memory currently used for atoms.</p>
<p>This memory is part of the memory presented as
<c>atom</c> memory.</p>
</item>
<tag><c>binary</c></tag>
<item>
<p>The total amount of memory currently allocated for
binaries.</p>
<p>This memory is part of the memory presented as
<c>system</c> memory.</p>
</item>
<tag><c>code</c></tag>
<item>
<p>The total amount of memory currently allocated for
Erlang code.</p>
<p>This memory is part of the memory presented as
<c>system</c> memory.</p>
</item>
<tag><c>ets</c></tag>
<item>
<p>The total amount of memory currently allocated for ets
tables.</p>
<p>This memory is part of the memory presented as
<c>system</c> memory.</p>
</item>
<tag><c>maximum</c></tag>
<item>
<p>The maximum total amount of memory allocated since
the emulator was started.</p>
<p>This tuple is only present when the emulator is run with
instrumentation.</p>
<p>For information on how to run the emulator with
instrumentation see
<seealso marker="tools:instrument">instrument(3)</seealso>
and/or <seealso marker="erts:erl">erl(1)</seealso>.</p>
</item>
</taglist>
<note>
<p>The <c>system</c> value is not complete. Some allocated
memory that should be part of the <c>system</c> value are
not.</p>
<p>When the emulator is run with instrumentation,
the <c>system</c> value is more accurate, but memory
directly allocated by <c>malloc</c> (and friends) are still
not part of the <c>system</c> value. Direct calls to
<c>malloc</c> are only done from OS specific runtime
libraries and perhaps from user implemented Erlang drivers
that do not use the memory allocation functions in
the driver interface.</p>
<p>Since the <c>total</c> value is the sum of <c>processes</c>
and <c>system</c> the error in <c>system</c> will propagate
to the <c>total</c> value.</p>
<p>The different amounts of memory that are summed are
<em>not</em> gathered atomically which also introduce
an error in the result.</p>
</note>
<p>The different values has the following relation to each
other. Values beginning with an uppercase letter is not part
of the result.</p>
<code type="none">
total = processes + system
processes = processes_used + ProcessesNotUsed
system = atom + binary + code + ets + OtherSystem
atom = atom_used + AtomNotUsed
RealTotal = processes + RealSystem
RealSystem = system + MissedSystem</code>
<p>More tuples in the returned list may be added in the future.</p>
<note>
<p>The <c>total</c> value is supposed to be the total amount
of memory dynamically allocated by the emulator. Shared
libraries, the code of the emulator itself, and
the emulator stack(s) are not supposed to be included. That
is, the <c>total</c> value is <em>not</em> supposed to be
equal to the total size of all pages mapped to the emulator.
Furthermore, due to fragmentation and pre-reservation of
memory areas, the size of the memory segments which contain
the dynamically allocated memory blocks can be substantially
larger than the total size of the dynamically allocated
memory blocks.</p>
</note>
<note>
<p>
Since erts version 5.6.4 <c>erlang:memory/0</c> requires that
all <seealso marker="erts:erts_alloc">erts_alloc(3)</seealso>
allocators are enabled (default behaviour).
</p>
</note>
<p>Failure:</p>
<taglist>
<tag><c>notsup</c></tag>
<item>
If an <seealso marker="erts:erts_alloc">erts_alloc(3)</seealso>
allocator has been disabled.
</item>
</taglist>
</desc>
</func>
<func>
<name>erlang:memory(Type | [Type]) -> Size | [{Type, Size}]</name>
<fsummary>Information about dynamically allocated memory</fsummary>
<type>
<v>Type, Size -- see below</v>
</type>
<desc>
<p>Returns the memory size in bytes allocated for memory of
type <c>Type</c>. The argument can also be given as a list
of <c>Type</c> atoms, in which case a corresponding list of
<c>{Type, Size}</c> tuples is returned.</p>
<note>
<p>
Since erts version 5.6.4 <c>erlang:memory/1</c> requires that
all <seealso marker="erts:erts_alloc">erts_alloc(3)</seealso>
allocators are enabled (default behaviour).
</p>
</note>
<p>Failures:</p>
<taglist>
<tag><c>badarg</c></tag>
<item>
If <c>Type</c> is not one of the memory types listed in the
documentation of
<seealso marker="#memory/0">erlang:memory/0</seealso>.
</item>
<tag><c>badarg</c></tag>
<item>
If <c>maximum</c> is passed as <c>Type</c> and the emulator
is not run in instrumented mode.
</item>
<tag><c>notsup</c></tag>
<item>
If an <seealso marker="erts:erts_alloc">erts_alloc(3)</seealso>
allocator has been disabled.
</item>
</taglist>
<p>See also
<seealso marker="#memory/0">erlang:memory/0</seealso>.</p>
</desc>
</func>
<func>
<name>min(Term1, Term2) -> Minimum</name>
<fsummary>Return the smallest of two term</fsummary>
<type>
<v>Term1 = Term2 = Minimum = term()</v>
</type>
<desc>
<p>Return the smallest of <c>Term1</c> and <c>Term2</c>;
if the terms compare equal, <c>Term1</c> will be returned.</p>
</desc>
</func>
<func>
<name>module_loaded(Module) -> bool()</name>
<fsummary>Check if a module is loaded</fsummary>
<type>
<v>Module = atom()</v>
</type>
<desc>
<p>Returns <c>true</c> if the module <c>Module</c> is loaded,
otherwise returns <c>false</c>. It does not attempt to load
the module.</p>
<warning>
<p>This BIF is intended for the code server (see
<seealso marker="kernel:code">code(3)</seealso>) and should not be
used elsewhere.</p>
</warning>
</desc>
</func>
<func>
<name>monitor(Type, Item) -> MonitorRef</name>
<fsummary>Start monitoring</fsummary>
<type>
<v>Type = process</v>
<v>Item = pid() | {RegName, Node} | RegName</v>
<v> RegName = atom()</v>
<v> Node = node()</v>
<v>MonitorRef = reference()</v>
</type>
<desc>
<p>The calling process starts monitoring <c>Item</c> which is
an object of type <c>Type</c>.</p>
<p>Currently only processes can be monitored, i.e. the only
allowed <c>Type</c> is <c>process</c>, but other types may be
allowed in the future.</p>
<p><c>Item</c> can be:</p>
<taglist>
<tag><c>pid()</c></tag>
<item>
<p>The pid of the process to monitor.</p>
</item>
<tag><c>{RegName, Node}</c></tag>
<item>
<p>A tuple consisting of a registered name of a process and
a node name. The process residing on the node <c>Node</c>
with the registered name <c>RegName</c> will be monitored.</p>
</item>
<tag><c>RegName</c></tag>
<item>
<p>The process locally registered as <c>RegName</c> will be
monitored.</p>
</item>
</taglist>
<note>
<p>When a process is monitored by registered name, the process
that has the registered name at the time when
<c>monitor/2</c> is called will be monitored.
The monitor will not be effected, if the registered name is
unregistered.</p>
</note>
<p>A <c>'DOWN'</c> message will be sent to the monitoring
process if <c>Item</c> dies, if <c>Item</c> does not exist,
or if the connection is lost to the node which <c>Item</c>
resides on. A <c>'DOWN'</c> message has the following pattern:</p>
<code type="none">
{'DOWN', MonitorRef, Type, Object, Info}</code>
<p>where <c>MonitorRef</c> and <c>Type</c> are the same as
described above, and:</p>
<taglist>
<tag><c>Object</c></tag>
<item>
<p>A reference to the monitored object:</p>
<list type="bulleted">
<item>the pid of the monitored process, if <c>Item</c> was
specified as a pid.</item>
<item><c>{RegName, Node}</c>, if <c>Item</c> was specified as
<c>{RegName, Node}</c>.</item>
<item><c>{RegName, Node}</c>, if <c>Item</c> was specified as
<c>RegName</c>. <c>Node</c> will in this case be the
name of the local node (<c>node()</c>).</item>
</list>
</item>
<tag><c>Info</c></tag>
<item>
<p>Either the exit reason of the process, <c>noproc</c>
(non-existing process), or <c>noconnection</c> (no
connection to <c>Node</c>).</p>
</item>
</taglist>
<note>
<p>If/when <c>monitor/2</c> is extended (e.g. to
handle other item types than <c>process</c>), other
possible values for <c>Object</c>, and <c>Info</c> in the
<c>'DOWN'</c> message will be introduced.</p>
</note>
<p>The monitoring is turned off either when the <c>'DOWN'</c>
message is sent, or when
<seealso marker="#demonitor/1">demonitor/1</seealso>
is called.</p>
<p>If an attempt is made to monitor a process on an older node
(where remote process monitoring is not implemented or one
where remote process monitoring by registered name is not
implemented), the call fails with <c>badarg</c>.</p>
<p>Making several calls to <c>monitor/2</c> for the same
<c>Item</c> is not an error; it results in as many, completely
independent, monitorings.</p>
<note>
<p>The format of the <c>'DOWN'</c> message changed in the 5.2
version of the emulator (OTP release R9B) for monitor <em>by registered name</em>. The <c>Object</c> element of
the <c>'DOWN'</c> message could in earlier versions
sometimes be the pid of the monitored process and sometimes
be the registered name. Now the <c>Object</c> element is
always a tuple consisting of the registered name and
the node name. Processes on new nodes (emulator version 5.2
or greater) will always get <c>'DOWN'</c> messages on
the new format even if they are monitoring processes on old
nodes. Processes on old nodes will always get <c>'DOWN'</c>
messages on the old format.</p>
</note>
</desc>
</func>
<func>
<name>monitor_node(Node, Flag) -> true</name>
<fsummary>Monitor the status of a node</fsummary>
<type>
<v>Node = node()</v>
<v>Flag = bool()</v>
</type>
<desc>
<p>Monitors the status of the node <c>Node</c>. If <c>Flag</c>
is <c>true</c>, monitoring is turned on; if <c>Flag</c> is
<c>false</c>, monitoring is turned off.</p>
<p>Making several calls to <c>monitor_node(Node, true)</c> for
the same <c>Node</c> is not an error; it results in as many,
completely independent, monitorings.</p>
<p>If <c>Node</c> fails or does not exist, the message
<c>{nodedown, Node}</c> is delivered to the process. If a
process has made two calls to <c>monitor_node(Node, true)</c>
and <c>Node</c> terminates, two <c>nodedown</c> messages are
delivered to the process. If there is no connection to
<c>Node</c>, there will be an attempt to create one. If this
fails, a <c>nodedown</c> message is delivered.</p>
<p>Nodes connected through hidden connections can be monitored
as any other node.</p>
<p>Failure: <c>badarg</c>if the local node is not alive.</p>
</desc>
</func>
<func>
<name>erlang:monitor_node(Node, Flag, Options) -> true</name>
<fsummary>Monitor the status of a node</fsummary>
<type>
<v>Node = node()</v>
<v>Flag = bool()</v>
<v>Options = [Option]</v>
<v>Option = allow_passive_connect</v>
</type>
<desc>
<p>Behaves as <c>monitor_node/2</c> except that it allows an
extra option to be given, namely <c>allow_passive_connect</c>.
The option allows the BIF to wait the normal net connection
timeout for the <em>monitored node</em> to connect itself,
even if it cannot be actively connected from this node
(i.e. it is blocked). The state where this might be useful can
only be achieved by using the kernel option
<c>dist_auto_connect once</c>. If that kernel option is not
used, the <c>allow_passive_connect</c> option has no
effect.</p>
<note>
<p>The <c>allow_passive_connect</c> option is used
internally and is seldom needed in applications where the
network topology and the kernel options in effect is known in
advance.</p>
</note>
<p>Failure: <c>badarg</c> if the local node is not alive or the
option list is malformed.</p>
</desc>
</func>
<func>
<name>erlang:nif_error(Reason)</name>
<fsummary>Stop execution with a given reason</fsummary>
<type>
<v>Reason = term()</v>
</type>
<desc>
<p>Works exactly like
<seealso marker="#error/1">erlang:error/1</seealso>,
but Dialyzer thinks that this BIF will return an arbitrary term.
When used in a stub function for a NIF to generate an
exception when the NIF library is not loaded, Dialyzer
will not generate false warnings.</p>
</desc>
</func>
<func>
<name>erlang:nif_error(Reason, Args)</name>
<fsummary>Stop execution with a given reason</fsummary>
<type>
<v>Reason = term()</v>
<v>Args = [term()]</v>
</type>
<desc>
<p>Works exactly like
<seealso marker="#error/2">erlang:error/2</seealso>,
but Dialyzer thinks that this BIF will return an arbitrary term.
When used in a stub function for a NIF to generate an
exception when the NIF library is not loaded, Dialyzer
will not generate false warnings.</p>
</desc>
</func>
<func>
<name>node() -> Node</name>
<fsummary>Name of the local node</fsummary>
<type>
<v>Node = node()</v>
</type>
<desc>
<p>Returns the name of the local node. If the node is not alive,
<c>nonode@nohost</c> is returned instead.</p>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>node(Arg) -> Node</name>
<fsummary>At which node is a pid, port or reference located</fsummary>
<type>
<v>Arg = pid() | port() | reference()</v>
<v>Node = node()</v>
</type>
<desc>
<p>Returns the node where <c>Arg</c> is located. <c>Arg</c> can
be a pid, a reference, or a port. If the local node is not
alive, <c>nonode@nohost</c> is returned.</p>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>nodes() -> Nodes</name>
<fsummary>All visible nodes in the system</fsummary>
<type>
<v>Nodes = [node()]</v>
</type>
<desc>
<p>Returns a list of all visible nodes in the system, excluding
the local node. Same as <c>nodes(visible)</c>.</p>
</desc>
</func>
<func>
<name>nodes(Arg | [Arg]) -> Nodes</name>
<fsummary>All nodes of a certain type in the system</fsummary>
<type>
<v>Arg = visible | hidden | connected | this | known</v>
<v>Nodes = [node()]</v>
</type>
<desc>
<p>Returns a list of nodes according to argument given.
The result returned when the argument is a list, is the list
of nodes satisfying the disjunction(s) of the list elements.</p>
<p><c>Arg</c> can be any of the following:</p>
<taglist>
<tag><c>visible</c></tag>
<item>
<p>Nodes connected to this node through normal connections.</p>
</item>
<tag><c>hidden</c></tag>
<item>
<p>Nodes connected to this node through hidden connections.</p>
</item>
<tag><c>connected</c></tag>
<item>
<p>All nodes connected to this node.</p>
</item>
<tag><c>this</c></tag>
<item>
<p>This node.</p>
</item>
<tag><c>known</c></tag>
<item>
<p>Nodes which are known to this node, i.e., connected,
previously connected, etc.</p>
</item>
</taglist>
<p>Some equalities: <c>[node()] = nodes(this)</c>,
<c>nodes(connected) = nodes([visible, hidden])</c>, and
<c>nodes() = nodes(visible)</c>.</p>
<p>If the local node is not alive,
<c>nodes(this) == nodes(known) == [nonode@nohost]</c>, for
any other <c>Arg</c> the empty list [] is returned.</p>
</desc>
</func>
<func>
<name>now() -> {MegaSecs, Secs, MicroSecs}</name>
<fsummary>Elapsed time since 00:00 GMT</fsummary>
<type>
<v>MegaSecs = Secs = MicroSecs = int()</v>
</type>
<desc>
<p>Returns the tuple <c>{MegaSecs, Secs, MicroSecs}</c> which is
the elapsed time since 00:00 GMT, January 1, 1970 (zero hour)
on the assumption that the underlying OS supports this.
Otherwise, some other point in time is chosen. It is also
guaranteed that subsequent calls to this BIF returns
continuously increasing values. Hence, the return value from
<c>now()</c> can be used to generate unique time-stamps,
and if it is called in a tight loop on a fast machine
the time of the node can become skewed.</p>
<p>It can only be used to check the local time of day if
the time-zone info of the underlying operating system is
properly configured.</p>
</desc>
</func>
<func>
<name>open_port(PortName, PortSettings) -> port()</name>
<fsummary>Open a port</fsummary>
<type>
<v>PortName = {spawn, Command} | {spawn_driver, Command} | {spawn_executable, Command} | {fd, In, Out}</v>
<v> Command = string()</v>
<v> In = Out = int()</v>
<v>PortSettings = [Opt]</v>
<v> Opt = {packet, N} | stream | {line, L} | {cd, Dir} | {env, Env} | {args, [ string() ]} | {arg0, string()} | exit_status | use_stdio | nouse_stdio | stderr_to_stdout | in | out | binary | eof</v>
<v> N = 1 | 2 | 4</v>
<v> L = int()</v>
<v> Dir = string()</v>
<v> Env = [{Name, Val}]</v>
<v> Name = string()</v>
<v> Val = string() | false</v>
</type>
<desc>
<p>Returns a port identifier as the result of opening a
new Erlang port. A port can be seen as an external Erlang
process. <c>PortName</c> is one of the following:</p>
<taglist>
<tag><c>{spawn, Command}</c></tag>
<item>
<p>Starts an external program. <c>Command</c> is the name
of the external program which will be run. <c>Command</c>
runs outside the Erlang work space unless an Erlang
driver with the name <c>Command</c> is found. If found,
that driver will be started. A driver runs in the Erlang
workspace, which means that it is linked with the Erlang
runtime system.</p>
<p>When starting external programs on Solaris, the system
call <c>vfork</c> is used in preference to <c>fork</c>
for performance reasons, although it has a history of
being less robust. If there are problems with using
<c>vfork</c>, setting the environment variable
<c>ERL_NO_VFORK</c> to any value will cause <c>fork</c>
to be used instead.</p>
<p>For external programs, the <c>PATH</c> is searched
(or an equivalent method is used to find programs,
depending on operating system). This is done by invoking
the shell och certain platforms. The first space
separated token of the command will be considered as the
name of the executable (or driver). This (among other
things) makes this option unsuitable for running
programs having spaces in file or directory names. Use
{spawn_executable, Command} instead if spaces in executable
file names is desired.</p>
</item>
<tag><c>{spawn_driver, Command}</c></tag>
<item>
<p>Works like <c>{spawn, Command}</c>, but demands the
first (space separated) token of the command to be the name of a
loaded driver. If no driver with that name is loaded, a
<c>badarg</c> error is raised.</p>
</item>
<tag><c>{spawn_executable, Command}</c></tag>
<item>
<p>Works like <c>{spawn, Command}</c>, but only runs
external executables. The <c>Command</c> in its whole
is used as the name of the executable, including any
spaces. If arguments are to be passed, the
<c>args</c> and <c>arg0</c> <c>PortSettings</c> can be used.</p>
<p>The shell is not usually invoked to start the
program, it's executed directly. Neither is the
<c>PATH</c> (or equivalent) searched. To find a program
in the PATH to execute, use <seealso
marker="kernel:os#find_executable/1">os:find_executable/1</seealso>.</p>
<p>Only if a shell script or <c>.bat</c> file is
executed, the appropriate command interpreter will
implicitly be invoked, but there will still be no
command argument expansion or implicit PATH search.</p>
<p>If the <c>Command</c> cannot be run, an error
exception, with the posix error code as the reason, is
raised. The error reason may differ between operating
systems. Typically the error <c>enoent</c> is raised
when one tries to run a program that is not found and
<c>eaccess</c> is raised when the given file is not
executable.</p>
</item>
<tag><c>{fd, In, Out}</c></tag>
<item>
<p>Allows an Erlang process to access any currently opened
file descriptors used by Erlang. The file descriptor
<c>In</c> can be used for standard input, and the file
descriptor <c>Out</c> for standard output. It is only
used for various servers in the Erlang operating system
(<c>shell</c> and <c>user</c>). Hence, its use is very
limited.</p>
</item>
</taglist>
<p><c>PortSettings</c> is a list of settings for the port.
Valid settings are:</p>
<taglist>
<tag><c>{packet, N}</c></tag>
<item>
<p>Messages are preceded by their length, sent in <c>N</c>
bytes, with the most significant byte first. Valid values
for <c>N</c> are 1, 2, or 4.</p>
</item>
<tag><c>stream</c></tag>
<item>
<p>Output messages are sent without packet lengths. A
user-defined protocol must be used between the Erlang
process and the external object.</p>
</item>
<tag><c>{line, L}</c></tag>
<item>
<p>Messages are delivered on a per line basis. Each line
(delimited by the OS-dependent newline sequence) is
delivered in one single message. The message data format
is <c>{Flag, Line}</c>, where <c>Flag</c> is either
<c>eol</c> or <c>noeol</c> and <c>Line</c> is the actual
data delivered (without the newline sequence).</p>
<p><c>L</c> specifies the maximum line length in bytes.
Lines longer than this will be delivered in more than one
message, with the <c>Flag</c> set to <c>noeol</c> for all
but the last message. If end of file is encountered
anywhere else than immediately following a newline
sequence, the last line will also be delivered with
the <c>Flag</c> set to <c>noeol</c>. In all other cases,
lines are delivered with <c>Flag</c> set to <c>eol</c>.</p>
<p>The <c>{packet, N}</c> and <c>{line, L}</c> settings are
mutually exclusive.</p>
</item>
<tag><c>{cd, Dir}</c></tag>
<item>
<p>This is only valid for <c>{spawn, Command}</c> and
<c>{spawn_executable, Command}</c>.
The external program starts using <c>Dir</c> as its
working directory. <c>Dir</c> must be a string. Not
available on VxWorks.</p>
</item>
<tag><c>{env, Env}</c></tag>
<item>
<p>This is only valid for <c>{spawn, Command}</c> and
<c>{spawn_executable, Command}</c>.
The environment of the started process is extended using
the environment specifications in <c>Env</c>.</p>
<p><c>Env</c> should be a list of tuples <c>{Name, Val}</c>,
where <c>Name</c> is the name of an environment variable,
and <c>Val</c> is the value it is to have in the spawned
port process. Both <c>Name</c> and <c>Val</c> must be
strings. The one exception is <c>Val</c> being the atom
<c>false</c> (in analogy with <c>os:getenv/1</c>), which
removes the environment variable. Not available on
VxWorks.</p>
</item>
<tag><c>{args, [ string() ]}</c></tag>
<item>
<p>This option is only valid for <c>{spawn_executable, Command}</c>
and specifies arguments to the executable. Each argument
is given as a separate string and (on Unix) eventually
ends up as one element each in the argument vector. On
other platforms, similar behavior is mimicked.</p>
<p>The arguments are not expanded by the shell prior to
being supplied to the executable, most notably this
means that file wildcard expansion will not happen. Use
<seealso
marker="stdlib:filelib#wildcard/1">filelib:wildcard/1</seealso>
to expand wildcards for the arguments. Note that even if
the program is a Unix shell script, meaning that the
shell will ultimately be invoked, wildcard expansion
will not happen and the script will be provided with the
untouched arguments. On Windows®, wildcard expansion
is always up to the program itself, why this isn't an
issue.</p>
<p>Note also that the actual executable name (a.k.a. <c>argv[0]</c>)
should not be given in this list. The proper executable name will
automatically be used as argv[0] where applicable.</p>
<p>If one, for any reason, wants to explicitly set the
program name in the argument vector, the <c>arg0</c>
option can be used.</p>
</item>
<tag><c>{arg0, string()}</c></tag>
<item>
<p>This option is only valid for <c>{spawn_executable, Command}</c>
and explicitly specifies the program name argument when
running an executable. This might in some circumstances,
on some operating systems, be desirable. How the program
responds to this is highly system dependent and no specific
effect is guaranteed.</p>
</item>
<tag><c>exit_status</c></tag>
<item>
<p>This is only valid for <c>{spawn, Command}</c> where
<c>Command</c> refers to an external program, and for
<c>{spawn_executable, Command}</c>.</p>
<p>When the external process connected to the port exits, a
message of the form <c>{Port,{exit_status,Status}}</c> is
sent to the connected process, where <c>Status</c> is the
exit status of the external process. If the program
aborts, on Unix the same convention is used as the shells
do (i.e., 128+signal).</p>
<p>If the <c>eof</c> option has been given as well,
the <c>eof</c> message and the <c>exit_status</c> message
appear in an unspecified order.</p>
<p>If the port program closes its stdout without exiting,
the <c>exit_status</c> option will not work.</p>
</item>
<tag><c>use_stdio</c></tag>
<item>
<p>This is only valid for <c>{spawn, Command}</c> and
<c>{spawn_executable, Command}</c>. It
allows the standard input and output (file descriptors 0
and 1) of the spawned (UNIX) process for communication
with Erlang.</p>
</item>
<tag><c>nouse_stdio</c></tag>
<item>
<p>The opposite of <c>use_stdio</c>. Uses file descriptors
3 and 4 for communication with Erlang.</p>
</item>
<tag><c>stderr_to_stdout</c></tag>
<item>
<p>Affects ports to external programs. The executed program
gets its standard error file redirected to its standard
output file. <c>stderr_to_stdout</c> and
<c>nouse_stdio</c> are mutually exclusive.</p>
</item>
<tag><c>overlapped_io</c></tag>
<item>
<p>Affects ports to external programs on Windows® only.
The standard input and standard output handles of the port program
will, if this option is supplied, be opened with the flag
FILE_FLAG_OVERLAPPED, so that the port program can (and has to) do
overlapped I/O on its standard handles. This is not normally
the case for simple port programs, but an option of value for the
experienced Windows programmer. <em>On all other platforms, this
option is silently discarded</em>.</p>
</item>
<tag><c>in</c></tag>
<item>
<p>The port can only be used for input.</p>
</item>
<tag><c>out</c></tag>
<item>
<p>The port can only be used for output.</p>
</item>
<tag><c>binary</c></tag>
<item>
<p>All IO from the port are binary data objects as opposed
to lists of bytes.</p>
</item>
<tag><c>eof</c></tag>
<item>
<p>The port will not be closed at the end of the file and
produce an exit signal. Instead, it will remain open and
a <c>{Port, eof}</c> message will be sent to the process
holding the port.</p>
</item>
<tag><c>hide</c></tag>
<item>
<p>When running on Windows, suppress creation of a new
console window when spawning the port program.
(This option has no effect on other platforms.)</p>
</item>
</taglist>
<p>The default is <c>stream</c> for all types of port and
<c>use_stdio</c> for spawned ports.</p>
<p>Failure: If the port cannot be opened, the exit reason is
<c>badarg</c>, <c>system_limit</c>, or the Posix error code which
most closely describes the error, or <c>einval</c> if no Posix code
is appropriate:</p>
<taglist>
<tag><c>badarg</c></tag>
<item>
<p>Bad input arguments to <c>open_port</c>.</p>
</item>
<tag><c>system_limit</c></tag>
<item>
<p>All available ports in the Erlang emulator are in use.</p>
</item>
<tag><c>enomem</c></tag>
<item>
<p>There was not enough memory to create the port.</p>
</item>
<tag><c>eagain</c></tag>
<item>
<p>There are no more available operating system processes.</p>
</item>
<tag><c>enametoolong</c></tag>
<item>
<p>The external command given was too long.</p>
</item>
<tag><c>emfile</c></tag>
<item>
<p>There are no more available file descriptors (for the operating system process
that the Erlang emulator runs in).</p>
</item>
<tag><c>enfile</c></tag>
<item>
<p>The file table is full (for the entire operating system).</p>
</item>
<tag><c>eacces</c></tag>
<item>
<p>The <c>Command</c> given in <c>{spawn_executable, Command}</c> does not point out an executable file.</p>
</item>
<tag><c>enoent</c></tag>
<item>
<p>The <c>Command</c> given in <c>{spawn_executable, Command}</c> does not point out an existing file.</p>
</item>
</taglist>
<p>During use of a port opened using <c>{spawn, Name}</c>,
<c>{spawn_driver, Name}</c> or <c>{spawn_executable, Name}</c>,
errors arising when sending messages to it are reported to
the owning process using signals of the form
<c>{'EXIT', Port, PosixCode}</c>. See <c>file(3)</c> for
possible values of <c>PosixCode</c>.</p>
<p><marker id="ERL_MAX_PORTS"></marker>
The maximum number of ports that can be open at the same
time is 1024 by default, but can be configured by
the environment variable <c>ERL_MAX_PORTS</c>.</p>
</desc>
</func>
<func>
<name>erlang:phash(Term, Range) -> Hash</name>
<fsummary>Portable hash function</fsummary>
<type>
<v>Term = term()</v>
<v>Range = 1..2^32</v>
<v>Hash = 1..Range</v>
</type>
<desc>
<p>Portable hash function that will give the same hash for
the same Erlang term regardless of machine architecture and
ERTS version (the BIF was introduced in ERTS 4.9.1.1). Range
can be between 1 and 2^32, the function returns a hash value
for <c>Term</c> within the range <c>1..Range</c>.</p>
<p>This BIF could be used instead of the old deprecated
<c>erlang:hash/2</c> BIF, as it calculates better hashes for
all data-types, but consider using <c>phash2/1,2</c> instead.</p>
</desc>
</func>
<func>
<name>erlang:phash2(Term [, Range]) -> Hash</name>
<fsummary>Portable hash function</fsummary>
<type>
<v>Term = term()</v>
<v>Range = 1..2^32</v>
<v>Hash = 0..Range-1</v>
</type>
<desc>
<p>Portable hash function that will give the same hash for
the same Erlang term regardless of machine architecture and
ERTS version (the BIF was introduced in ERTS 5.2). Range can
be between 1 and 2^32, the function returns a hash value for
<c>Term</c> within the range <c>0..Range-1</c>. When called
without the <c>Range</c> argument, a value in the range
<c>0..2^27-1</c> is returned.</p>
<p>This BIF should always be used for hashing terms. It
distributes small integers better than <c>phash/2</c>, and
it is faster for bignums and binaries.</p>
<p>Note that the range <c>0..Range-1</c> is different from
the range of <c>phash/2</c> (<c>1..Range</c>).</p>
</desc>
</func>
<func>
<name>pid_to_list(Pid) -> string()</name>
<fsummary>Text representation of a pid</fsummary>
<type>
<v>Pid = pid()</v>
</type>
<desc>
<p>Returns a string which corresponds to the text
representation of <c>Pid</c>.</p>
<warning>
<p>This BIF is intended for debugging and for use in
the Erlang operating system. It should not be used in
application programs.</p>
</warning>
</desc>
</func>
<func>
<name>port_close(Port) -> true</name>
<fsummary>Close an open port</fsummary>
<type>
<v>Port = port() | atom()</v>
</type>
<desc>
<p>Closes an open port. Roughly the same as
<c>Port ! {self(), close}</c> except for the error behaviour
(see below), and that the port does <em>not</em> reply with
<c>{Port, closed}</c>. Any process may close a port with
<c>port_close/1</c>, not only the port owner (the connected
process).</p>
<p>For comparison: <c>Port ! {self(), close}</c> fails with
<c>badarg</c> if <c>Port</c> cannot be sent to (i.e.,
<c>Port</c> refers neither to a port nor to a process). If
<c>Port</c> is a closed port nothing happens. If <c>Port</c>
is an open port and the calling process is the port owner,
the port replies with <c>{Port, closed}</c> when all buffers
have been flushed and the port really closes, but if
the calling process is not the port owner the <em>port owner</em> fails with <c>badsig</c>.</p>
<p>Note that any process can close a port using
<c>Port ! {PortOwner, close}</c> just as if it itself was
the port owner, but the reply always goes to the port owner.</p>
<p>In short: <c>port_close(Port)</c> has a cleaner and more
logical behaviour than <c>Port ! {self(), close}</c>.</p>
<p>Failure: <c>badarg</c> if <c>Port</c> is not an open port or
the registered name of an open port.</p>
</desc>
</func>
<func>
<name>port_command(Port, Data) -> true</name>
<fsummary>Send data to a port</fsummary>
<type>
<v>Port = port() | atom()</v>
<v>Data = iodata()</v>
</type>
<desc>
<p>Sends data to a port. Same as
<c>Port ! {self(), {command, Data}}</c> except for the error
behaviour (see below). Any process may send data to a port
with <c>port_command/2</c>, not only the port owner
(the connected process).</p>
<p>For comparison: <c>Port ! {self(), {command, Data}}</c>
fails with <c>badarg</c> if <c>Port</c> cannot be sent to
(i.e., <c>Port</c> refers neither to a port nor to a process).
If <c>Port</c> is a closed port the data message disappears
without a sound. If <c>Port</c> is open and the calling
process is not the port owner, the <em>port owner</em> fails
with <c>badsig</c>. The port owner fails with <c>badsig</c>
also if <c>Data</c> is not a valid IO list.</p>
<p>Note that any process can send to a port using
<c>Port ! {PortOwner, {command, Data}}</c> just as if it
itself was the port owner.</p>
<p>In short: <c>port_command(Port, Data)</c> has a cleaner and
more logical behaviour than
<c>Port ! {self(), {command, Data}}</c>.</p>
<p>If the port is busy, the calling process will be suspended
until the port is not busy anymore.</p>
<p>Failures:</p>
<taglist>
<tag><c>badarg</c></tag>
<item>
If <c>Port</c> is not an open port or the registered name
of an open port.
</item>
<tag><c>badarg</c></tag>
<item>
If <c>Data</c> is not a valid io list.
</item>
</taglist>
</desc>
</func>
<func>
<name>port_command(Port, Data, OptionList) -> true|false</name>
<fsummary>Send data to a port</fsummary>
<type>
<v>Port = port() | atom()</v>
<v>Data = iodata()</v>
<v>OptionList = [Option]</v>
<v>Option = force</v>
<v>Option = nosuspend</v>
</type>
<desc>
<p>Sends data to a port. <c>port_command(Port, Data, [])</c>
equals <c>port_command(Port, Data)</c>.</p>
<p>If the port command is aborted <c>false</c> is returned;
otherwise, <c>true</c> is returned.</p>
<p>If the port is busy, the calling process will be suspended
until the port is not busy anymore.</p>
<p>Currently the following <c>Option</c>s are valid:</p>
<taglist>
<tag><c>force</c></tag>
<item>The calling process will not be suspended if the port is
busy; instead, the port command is forced through. The
call will fail with a <c>notsup</c> exception if the
driver of the port does not support this. For more
information see the
<seealso marker="driver_entry#driver_flags"><![CDATA[ERL_DRV_FLAG_SOFT_BUSY]]></seealso>
driver flag.
</item>
<tag><c>nosuspend</c></tag>
<item>The calling process will not be suspended if the port is
busy; instead, the port command is aborted and
<c>false</c> is returned.
</item>
</taglist>
<note>
<p>More options may be added in the future.</p>
</note>
<p>Failures:</p>
<taglist>
<tag><c>badarg</c></tag>
<item>
If <c>Port</c> is not an open port or the registered name
of an open port.
</item>
<tag><c>badarg</c></tag>
<item>
If <c>Data</c> is not a valid io list.
</item>
<tag><c>badarg</c></tag>
<item>
If <c>OptionList</c> is not a valid option list.
</item>
<tag><c>notsup</c></tag>
<item>
If the <c>force</c> option has been passed, but the
driver of the port does not allow forcing through
a busy port.
</item>
</taglist>
</desc>
</func>
<func>
<name>port_connect(Port, Pid) -> true</name>
<fsummary>Set the owner of a port</fsummary>
<type>
<v>Port = port() | atom()</v>
<v>Pid = pid()</v>
</type>
<desc>
<p>Sets the port owner (the connected port) to <c>Pid</c>.
Roughly the same as <c>Port ! {self(), {connect, Pid}}</c>
except for the following:</p>
<list type="bulleted">
<item>
<p>The error behavior differs, see below.</p>
</item>
<item>
<p>The port does <em>not</em> reply with
<c>{Port,connected}</c>.</p>
</item>
<item>
<p>The new port owner gets linked to the port.</p>
</item>
</list>
<p>The old port owner stays linked to the port and have to call
<c>unlink(Port)</c> if this is not desired. Any process may
set the port owner to be any process with
<c>port_connect/2</c>.</p>
<p>For comparison: <c>Port ! {self(), {connect, Pid}}</c> fails
with <c>badarg</c> if <c>Port</c> cannot be sent to (i.e.,
<c>Port</c> refers neither to a port nor to a process). If
<c>Port</c> is a closed port nothing happens. If <c>Port</c>
is an open port and the calling process is the port owner,
the port replies with <c>{Port, connected}</c> to the old
port owner. Note that the old port owner is still linked to
the port, and that the new is not. If <c>Port</c> is an open
port and the calling process is not the port owner,
the <em>port owner</em> fails with <c>badsig</c>. The port
owner fails with <c>badsig</c> also if <c>Pid</c> is not an
existing local pid.</p>
<p>Note that any process can set the port owner using
<c>Port ! {PortOwner, {connect, Pid}}</c> just as if it
itself was the port owner, but the reply always goes to
the port owner.</p>
<p>In short: <c>port_connect(Port, Pid)</c> has a cleaner and
more logical behaviour than
<c>Port ! {self(),{connect,Pid}}</c>.</p>
<p>Failure: <c>badarg</c> if <c>Port</c> is not an open port
or the registered name of an open port, or if <c>Pid</c> is
not an existing local pid.</p>
</desc>
</func>
<func>
<name>port_control(Port, Operation, Data) -> Res</name>
<fsummary>Perform a synchronous control operation on a port</fsummary>
<type>
<v>Port = port() | atom()</v>
<v>Operation = int()</v>
<v>Data = Res = iodata()</v>
</type>
<desc>
<p>Performs a synchronous control operation on a port.
The meaning of <c>Operation</c> and <c>Data</c> depends on
the port, i.e., on the port driver. Not all port drivers
support this control feature.</p>
<p>Returns: a list of integers in the range 0 through 255, or a
binary, depending on the port driver. The meaning of
the returned data also depends on the port driver.</p>
<p>Failure: <c>badarg</c> if <c>Port</c> is not an open port or
the registered name of an open port, if <c>Operation</c>
cannot fit in a 32-bit integer, if the port driver does not
support synchronous control operations, or if the port driver
so decides for any reason (probably something wrong with
<c>Operation</c> or <c>Data</c>).</p>
</desc>
</func>
<func>
<name>erlang:port_call(Port, Operation, Data) -> term()</name>
<fsummary>Synchronous call to a port with term data</fsummary>
<type>
<v>Port = port() | atom()</v>
<v>Operation = int()</v>
<v>Data = term()</v>
</type>
<desc>
<p>Performs a synchronous call to a port. The meaning of
<c>Operation</c> and <c>Data</c> depends on the port, i.e.,
on the port driver. Not all port drivers support this feature.</p>
<p><c>Port</c> is a port identifier, referring to a driver.</p>
<p><c>Operation</c> is an integer, which is passed on to
the driver.</p>
<p><c>Data</c> is any Erlang term. This data is converted to
binary term format and sent to the port.</p>
<p>Returns: a term from the driver. The meaning of the returned
data also depends on the port driver.</p>
<p>Failure: <c>badarg</c> if <c>Port</c> is not an open port or
the registered name of an open port, if <c>Operation</c>
cannot fit in a 32-bit integer, if the port driver does not
support synchronous control operations, or if the port driver
so decides for any reason (probably something wrong with
<c>Operation</c> or <c>Data</c>).</p>
</desc>
</func>
<func>
<name>erlang:port_info(Port) -> [{Item, Info}] | undefined</name>
<fsummary>Information about a port</fsummary>
<type>
<v>Port = port() | atom()</v>
<v>Item, Info -- see below</v>
</type>
<desc>
<p>Returns a list containing tuples with information about
the <c>Port</c>, or <c>undefined</c> if the port is not open.
The order of the tuples is not defined, nor are all the
tuples mandatory.</p>
<taglist>
<tag><c>{registered_name, RegName}</c></tag>
<item>
<p><c>RegName</c> (an atom) is the registered name of
the port. If the port has no registered name, this tuple
is not present in the list.</p>
</item>
<tag><c>{id, Index}</c></tag>
<item>
<p><c>Index</c> (an integer) is the internal index of the
port. This index may be used to separate ports.</p>
</item>
<tag><c>{connected, Pid}</c></tag>
<item>
<p><c>Pid</c> is the process connected to the port.</p>
</item>
<tag><c>{links, Pids}</c></tag>
<item>
<p><c>Pids</c> is a list of pids to which processes the
port is linked.</p>
</item>
<tag><c>{name, String}</c></tag>
<item>
<p><c>String</c> is the command name set by
<c>open_port</c>.</p>
</item>
<tag><c>{input, Bytes}</c></tag>
<item>
<p><c>Bytes</c> is the total number of bytes read from
the port.</p>
</item>
<tag><c>{output, Bytes}</c></tag>
<item>
<p><c>Bytes</c> is the total number of bytes written to
the port.</p>
</item>
</taglist>
<p>Failure: <c>badarg</c> if <c>Port</c> is not a local port.</p>
</desc>
</func>
<func>
<name>erlang:port_info(Port, Item) -> {Item, Info} | undefined | []</name>
<fsummary>Information about a port</fsummary>
<type>
<v>Port = port() | atom()</v>
<v>Item, Info -- see below</v>
</type>
<desc>
<p>Returns information about <c>Port</c> as specified
by <c>Item</c>, or <c>undefined</c> if the port is not open.
Also, if <c>Item == registered_name</c> and the port has no
registered name, [] is returned.</p>
<p>For valid values of <c>Item</c>, and corresponding
values of <c>Info</c>, see
<seealso marker="#port_info/1">erlang:port_info/1</seealso>.</p>
<p>Failure: <c>badarg</c> if <c>Port</c> is not a local port.</p>
</desc>
</func>
<func>
<name>erlang:port_to_list(Port) -> string()</name>
<fsummary>Text representation of a port identifier</fsummary>
<type>
<v>Port = port()</v>
</type>
<desc>
<p>Returns a string which corresponds to the text
representation of the port identifier <c>Port</c>.</p>
<warning>
<p>This BIF is intended for debugging and for use in
the Erlang operating system. It should not be used in
application programs.</p>
</warning>
</desc>
</func>
<func>
<name>erlang:ports() -> [port()]</name>
<fsummary>All open ports</fsummary>
<desc>
<p>Returns a list of all ports on the local node.</p>
</desc>
</func>
<func>
<name>pre_loaded() -> [Module]</name>
<fsummary>List of all pre-loaded modules</fsummary>
<type>
<v>Module = atom()</v>
</type>
<desc>
<p>Returns a list of Erlang modules which are pre-loaded in
the system. As all loading of code is done through the file
system, the file system must have been loaded previously.
Hence, at least the module <c>init</c> must be pre-loaded.</p>
</desc>
</func>
<func>
<name>erlang:process_display(Pid, Type) -> void()</name>
<fsummary>Write information about a local process on standard error</fsummary>
<type>
<v>Pid = pid()</v>
<v>Type = backtrace</v>
</type>
<desc>
<p>Writes information about the local process <c>Pid</c> on
standard error. The currently allowed value for the atom
<c>Type</c> is <c>backtrace</c>, which shows the contents of
the call stack, including information about the call chain, with
the current function printed first. The format of the output
is not further defined.</p>
</desc>
</func>
<func>
<name>process_flag(Flag, Value) -> OldValue</name>
<fsummary>Set process flags for the calling process</fsummary>
<type>
<v>Flag, Value, OldValue -- see below</v>
</type>
<desc>
<p>Sets certain flags for the process which calls this
function. Returns the old value of the flag.</p>
<taglist>
<tag><c>process_flag(trap_exit, Boolean)</c></tag>
<item>
<p>When <c>trap_exit</c> is set to <c>true</c>, exit signals
arriving to a process are converted to <c>{'EXIT', From, Reason}</c> messages, which can be received as ordinary
messages. If <c>trap_exit</c> is set to <c>false</c>, the
process exits if it receives an exit signal other than
<c>normal</c> and the exit signal is propagated to its
linked processes. Application processes should normally
not trap exits.</p>
<p>See also <seealso marker="#exit/2">exit/2</seealso>.</p>
</item>
<tag><c>process_flag(error_handler, Module)</c></tag>
<item>
<p>This is used by a process to redefine the error handler
for undefined function calls and undefined registered
processes. Inexperienced users should not use this flag
since code auto-loading is dependent on the correct
operation of the error handling module.</p>
</item>
<tag><c>process_flag(min_heap_size, MinHeapSize)</c></tag>
<item>
<p>This changes the minimum heap size for the calling
process.</p>
</item>
<tag><c>process_flag(min_bin_vheap_size, MinBinVHeapSize)</c></tag>
<item>
<p>This changes the minimum binary virtual heap size for the calling
process.</p>
</item>
<tag><marker id="process_flag_priority"><c>process_flag(priority, Level)</c></marker></tag>
<item>
<p>This sets the process priority. <c>Level</c> is an atom.
There are currently four priority levels: <c>low</c>,
<c>normal</c>, <c>high</c>, and <c>max</c>. The default
priority level is <c>normal</c>. <em>NOTE</em>: The
<c>max</c> priority level is reserved for internal use in
the Erlang runtime system, and should <em>not</em> be used
by others.
</p>
<p>Internally in each priority level processes are scheduled
in a round robin fashion.
</p>
<p>Execution of processes on priority <c>normal</c> and
priority <c>low</c> will be interleaved. Processes on
priority <c>low</c> will be selected for execution less
frequently than processes on priority <c>normal</c>.
</p>
<p>When there are runnable processes on priority <c>high</c>
no processes on priority <c>low</c>, or <c>normal</c> will
be selected for execution. Note, however, that this does
<em>not</em> mean that no processes on priority <c>low</c>,
or <c>normal</c> will be able to run when there are
processes on priority <c>high</c> running. On the runtime
system with SMP support there might be more processes running
in parallel than processes on priority <c>high</c>, i.e.,
a <c>low</c>, and a <c>high</c> priority process might
execute at the same time.
</p>
<p>When there are runnable processes on priority <c>max</c>
no processes on priority <c>low</c>, <c>normal</c>, or
<c>high</c> will be selected for execution. As with the
<c>high</c> priority, processes on lower priorities might
execute in parallel with processes on priority <c>max</c>.
</p>
<p>Scheduling is preemptive. Regardless of priority, a process
is preempted when it has consumed more than a certain amount
of reductions since the last time it was selected for
execution.
</p>
<p><em>NOTE</em>: You should not depend on the scheduling
to remain exactly as it is today. Scheduling, at least on
the runtime system with SMP support, is very likely to be
modified in the future in order to better utilize available
processor cores.
</p>
<p>There is currently <em>no</em> automatic mechanism for
avoiding priority inversion, such as priority inheritance,
or priority ceilings. When using priorities you have
to take this into account and handle such scenarios by
yourself.
</p>
<p>Making calls from a <c>high</c> priority process into code
that you don't have control over may cause the <c>high</c>
priority process to wait for a processes with lower
priority, i.e., effectively decreasing the priority of the
<c>high</c> priority process during the call. Even if this
isn't the case with one version of the code that you don't
have under your control, it might be the case in a future
version of it. This might, for example, happen if a
<c>high</c> priority process triggers code loading, since
the code server runs on priority <c>normal</c>.
</p>
<p>Other priorities than <c>normal</c> are normally not needed.
When other priorities are used, they need to be used
with care, especially the <c>high</c> priority <em>must</em>
be used with care. A process on <c>high</c> priority should
only perform work for short periods of time. Busy looping for
long periods of time in a <c>high</c> priority process will
most likely cause problems, since there are important servers
in OTP running on priority <c>normal</c>.
</p>
</item>
<tag><c>process_flag(save_calls, N)</c></tag>
<item>
<p>When there are runnable processes on priority <c>max</c>
no processes on priority <c>low</c>, <c>normal</c>, or
<c>high</c> will be selected for execution. As with the
<c>high</c> priority, processes on lower priorities might
execute in parallel with processes on priority <c>max</c>.
</p>
<p><c>N</c> must be an integer in the interval 0..10000.
If <c>N</c> > 0, call saving is made active for the
process, which means that information about the <c>N</c>
most recent global function calls, BIF calls, sends and
receives made by the process are saved in a list, which
can be retrieved with
<c>process_info(Pid, last_calls)</c>. A global function
call is one in which the module of the function is
explicitly mentioned. Only a fixed amount of information
is saved: a tuple <c>{Module, Function, Arity}</c> for
function calls, and the mere atoms <c>send</c>,
<c>'receive'</c> and <c>timeout</c> for sends and receives
(<c>'receive'</c> when a message is received and
<c>timeout</c> when a receive times out). If <c>N</c> = 0,
call saving is disabled for the process, which is the
default. Whenever the size of the call saving list is set,
its contents are reset.</p>
</item>
<tag><c>process_flag(sensitive, Boolean)</c></tag>
<item>
<p>Set or clear the <c>sensitive</c> flag for the current process.
When a process has been marked as sensitive by calling
<c>process_flag(sensitive, true)</c>, features in the run-time
system that can be used for examining the data and/or inner working
of the process are silently disabled.</p>
<p>Features that are disabled include (but are not limited to)
the following:</p>
<p>Tracing: Trace flags can still be set for the process, but no
trace messages of any kind will be generated.
(If the <c>sensitive</c> flag is turned off, trace messages will
again be generated if there are any trace flags set.)</p>
<p>Sequential tracing: The sequential trace token will be propagated
as usual, but no sequential trace messages will be generated.</p>
<p><c>process_info/1,2</c> cannot be used to read out the message
queue or the process dictionary (both will be returned as empty lists).</p>
<p>Stack back-traces cannot be displayed for the process.</p>
<p>In crash dumps, the stack, messages, and the process dictionary
will be omitted.</p>
<p>If <c>{save_calls,N}</c> has been set for the process, no
function calls will be saved to the call saving list.
(The call saving list will not be cleared; furthermore, send, receive,
and timeout events will still be added to the list.)</p>
</item>
</taglist>
</desc>
</func>
<func>
<name>process_flag(Pid, Flag, Value) -> OldValue</name>
<fsummary>Set process flags for a process</fsummary>
<type>
<v>Pid = pid()</v>
<v>Flag, Value, OldValue -- see below</v>
</type>
<desc>
<p>Sets certain flags for the process <c>Pid</c>, in the same
manner as
<seealso marker="#process_flag/2">process_flag/2</seealso>.
Returns the old value of the flag. The allowed values for
<c>Flag</c> are only a subset of those allowed in
<c>process_flag/2</c>, namely: <c>save_calls</c>.</p>
<p>Failure: <c>badarg</c> if <c>Pid</c> is not a local process.</p>
</desc>
</func>
<func>
<name>process_info(Pid) -> InfoResult</name>
<fsummary>Information about a process</fsummary>
<type>
<v>Pid = pid()</v>
<v>Item = atom()</v>
<v>Info = term()</v>
<v>InfoTuple = {Item, Info}</v>
<v>InfoTupleList = [InfoTuple]</v>
<v>InfoResult = InfoTupleList | undefined</v>
</type>
<desc>
<p>Returns a list containing <c>InfoTuple</c>s with
miscellaneous information about the process identified by
<c>Pid</c>, or <c>undefined</c> if the process is not alive.
</p>
<p>
The order of the <c>InfoTuple</c>s is not defined, nor
are all the <c>InfoTuple</c>s mandatory. The <c>InfoTuple</c>s
part of the result may be changed without prior notice.
Currently <c>InfoTuple</c>s with the following <c>Item</c>s
are part of the result:
<c>current_function</c>, <c>initial_call</c>, <c>status</c>,
<c>message_queue_len</c>, <c>messages</c>, <c>links</c>,
<c>dictionary</c>, <c>trap_exit</c>, <c>error_handler</c>,
<c>priority</c>, <c>group_leader</c>, <c>total_heap_size</c>,
<c>heap_size</c>, <c>stack_size</c>, <c>reductions</c>, and
<c>garbage_collection</c>.
If the process identified by <c>Pid</c> has a registered name
also an <c>InfoTuple</c> with <c>Item == registered_name</c>
will appear.
</p>
<p>See <seealso marker="#process_info/2">process_info/2</seealso>
for information about specific <c>InfoTuple</c>s.</p>
<warning>
<p>This BIF is intended for <em>debugging only</em>, use
<seealso marker="#process_info/2">process_info/2</seealso>
for all other purposes.
</p>
</warning>
<p>Failure: <c>badarg</c> if <c>Pid</c> is not a local process.</p>
</desc>
</func>
<func>
<name>process_info(Pid, ItemSpec) -> InfoResult</name>
<fsummary>Information about a process</fsummary>
<type>
<v>Pid = pid()</v>
<v>Item = atom()</v>
<v>Info = term()</v>
<v>ItemList = [Item]</v>
<v>ItemSpec = Item | ItemList</v>
<v>InfoTuple = {Item, Info}</v>
<v>InfoTupleList = [InfoTuple]</v>
<v>InfoResult = InfoTuple | InfoTupleList | undefined | []</v>
</type>
<desc>
<p>Returns information about the process identified by <c>Pid</c>
as specified by the <c>ItemSpec</c>, or <c>undefined</c> if the
process is not alive.
</p>
<p>If the process is alive and <c>ItemSpec</c> is a single
<c>Item</c>, the returned value is the corresponding
<c>InfoTuple</c> unless <c>ItemSpec == registered_name</c>
and the process has no registered name. In this case
<c>[]</c> is returned. This strange behavior is due to
historical reasons, and is kept for backward compatibility.
</p>
<p>If <c>ItemSpec</c> is an <c>ItemList</c>, the result is an
<c>InfoTupleList</c>. The <c>InfoTuple</c>s in the
<c>InfoTupleList</c> will appear with the corresponding
<c>Item</c>s in the same order as the <c>Item</c>s appeared
in the <c>ItemList</c>. Valid <c>Item</c>s may appear multiple
times in the <c>ItemList</c>.
</p>
<note><p>If <c>registered_name</c> is part of an <c>ItemList</c>
and the process has no name registered a
<c>{registered_name, []}</c> <c>InfoTuple</c> <em>will</em>
appear in the resulting <c>InfoTupleList</c>. This
behavior is different than when
<c>ItemSpec == registered_name</c>, and than when
<c>process_info/1</c> is used.
</p></note>
<p>Currently the following <c>InfoTuple</c>s with corresponding
<c>Item</c>s are valid:</p>
<taglist>
<tag><c>{backtrace, Bin}</c></tag>
<item>
<p>The binary <c>Bin</c> contains the same information as
the output from
<c>erlang:process_display(Pid, backtrace)</c>. Use
<c>binary_to_list/1</c> to obtain the string of characters
from the binary.</p>
</item>
<tag><c>{binary, BinInfo}</c></tag>
<item>
<p><c>BinInfo</c> is a list containing miscellaneous information
about binaries currently being referred to by this process.
This <c>InfoTuple</c> may be changed or removed without prior
notice.</p>
</item>
<tag><c>{catchlevel, CatchLevel}</c></tag>
<item>
<p><c>CatchLevel</c> is the number of currently active
catches in this process. This <c>InfoTuple</c> may be
changed or removed without prior notice.</p>
</item>
<tag><c>{current_function, {Module, Function, Args}}</c></tag>
<item>
<p><c>Module</c>, <c>Function</c>, <c>Args</c> is
the current function call of the process.</p>
</item>
<tag><c>{dictionary, Dictionary}</c></tag>
<item>
<p><c>Dictionary</c> is the dictionary of the process.</p>
</item>
<tag><c>{error_handler, Module}</c></tag>
<item>
<p><c>Module</c> is the error handler module used by
the process (for undefined function calls, for example).</p>
</item>
<tag><c>{garbage_collection, GCInfo}</c></tag>
<item>
<p><c>GCInfo</c> is a list which contains miscellaneous
information about garbage collection for this process.
The content of <c>GCInfo</c> may be changed without
prior notice.</p>
</item>
<tag><c>{group_leader, GroupLeader}</c></tag>
<item>
<p><c>GroupLeader</c> is group leader for the IO of
the process.</p>
</item>
<tag><c>{heap_size, Size}</c></tag>
<item>
<p><c>Size</c> is the size in words of youngest heap generation
of the process. This generation currently include the stack
of the process. This information is highly implementation
dependent, and may change if the implementation change.
</p>
</item>
<tag><c>{initial_call, {Module, Function, Arity}}</c></tag>
<item>
<p><c>Module</c>, <c>Function</c>, <c>Arity</c> is
the initial function call with which the process was
spawned.</p>
</item>
<tag><c>{links, Pids}</c></tag>
<item>
<p><c>Pids</c> is a list of pids, with processes to
which the process has a link.</p>
</item>
<tag><c>{last_calls, false|Calls}</c></tag>
<item>
<p>The value is <c>false</c> if call saving is not active
for the process (see
<seealso marker="#process_flag/3">process_flag/3</seealso>).
If call saving is active, a list is returned, in which
the last element is the most recent called.</p>
</item>
<tag><c>{memory, Size}</c></tag>
<item>
<p><c>Size</c> is the size in bytes of the process. This
includes call stack, heap and internal structures.</p>
</item>
<tag><c>{message_binary, BinInfo}</c></tag>
<item>
<p><c>BinInfo</c> is a list containing miscellaneous information
about binaries currently being referred to by the message
area. This <c>InfoTuple</c> is only valid on an emulator
using the hybrid heap type. This <c>InfoTuple</c> may be
changed or removed without prior notice.</p>
</item>
<tag><c>{message_queue_len, MessageQueueLen}</c></tag>
<item>
<p><c>MessageQueueLen</c> is the number of messages
currently in the message queue of the process. This is
the length of the list <c>MessageQueue</c> returned as
the info item <c>messages</c> (see below).</p>
</item>
<tag><c>{messages, MessageQueue}</c></tag>
<item>
<p><c>MessageQueue</c> is a list of the messages to
the process, which have not yet been processed.</p>
</item>
<tag><c>{min_heap_size, MinHeapSize}</c></tag>
<item>
<p><c>MinHeapSize</c> is the minimum heap size for the process.</p>
</item>
<tag><c>{min_bin_vheap_size, MinBinVHeapSize}</c></tag>
<item>
<p><c>MinBinVHeapSize</c> is the minimum binary virtual heap size for the process.</p>
</item>
<tag><c>{monitored_by, Pids}</c></tag>
<item>
<p>A list of pids that are monitoring the process (with
<c>monitor/2</c>).</p>
</item>
<tag><c>{monitors, Monitors}</c></tag>
<item>
<p>A list of monitors (started by <c>monitor/2</c>)
that are active for the process. For a local process
monitor or a remote process monitor by pid, the list item
is <c>{process, Pid}</c>, and for a remote process
monitor by name, the list item is
<c>{process, {RegName, Node}}</c>.</p>
</item>
<tag><c>{priority, Level}</c></tag>
<item>
<p><c>Level</c> is the current priority level for
the process. For more information on priorities see
<seealso marker="#process_flag_priority">process_flag(priority, Level)</seealso>.</p>
</item>
<tag><c>{reductions, Number}</c></tag>
<item>
<p><c>Number</c> is the number of reductions executed by
the process.</p>
</item>
<tag><c>{registered_name, Atom}</c></tag>
<item>
<p><c>Atom</c> is the registered name of the process. If
the process has no registered name, this tuple is not
present in the list.</p>
</item>
<tag><c>{sequential_trace_token, [] | SequentialTraceToken}</c></tag>
<item>
<p><c>SequentialTraceToken</c> the sequential trace token for
the process. This <c>InfoTuple</c> may be changed or removed
without prior notice.</p>
</item>
<tag><c>{stack_size, Size}</c></tag>
<item>
<p><c>Size</c> is the stack size of the process in words.</p>
</item>
<tag><c>{status, Status}</c></tag>
<item>
<p><c>Status</c> is the status of the process. <c>Status</c>
is <c>waiting</c> (waiting for a message), <c>running</c>,
<c>runnable</c> (ready to run, but another process is
running), or <c>suspended</c> (suspended on a "busy" port
or by the <c>erlang:suspend_process/[1,2]</c> BIF).</p>
</item>
<tag><c>{suspending, SuspendeeList}</c></tag>
<item>
<p><c>SuspendeeList</c> is a list of <c>{Suspendee,
ActiveSuspendCount, OutstandingSuspendCount}</c> tuples.
<c>Suspendee</c> is the pid of a process that have been or is to
be suspended by the process identified by <c>Pid</c> via the
<seealso marker="#suspend_process/2">erlang:suspend_process/2</seealso>
BIF, or the
<seealso marker="#suspend_process/1">erlang:suspend_process/1</seealso>
BIF. <c>ActiveSuspendCount</c> is the number of times the
<c>Suspendee</c> has been suspended by <c>Pid</c>.
<c>OutstandingSuspendCount</c> is the number of not yet
completed suspend requests sent by <c>Pid</c>. That is,
if <c>ActiveSuspendCount /= 0</c>, <c>Suspendee</c> is
currently in the suspended state, and if
<c>OutstandingSuspendCount /= 0</c> the <c>asynchronous</c>
option of <c>erlang:suspend_process/2</c> has been used and
the suspendee has not yet been suspended by <c>Pid</c>.
Note that the <c>ActiveSuspendCount</c> and
<c>OutstandingSuspendCount</c> are not the total suspend count
on <c>Suspendee</c>, only the parts contributed by <c>Pid</c>.
</p>
</item>
<tag><c>{total_heap_size, Size}</c></tag>
<item>
<p><c>Size</c> is the total size in words of all heap
fragments of the process. This currently include the stack
of the process.
</p>
</item>
<tag><c>{trace, InternalTraceFlags}</c></tag>
<item>
<p><c>InternalTraceFlags</c> is an integer representing
internal trace flag for this process. This <c>InfoTuple</c>
may be changed or removed without prior notice.</p>
</item>
<tag><c>{trap_exit, Boolean}</c></tag>
<item>
<p><c>Boolean</c> is <c>true</c> if the process is trapping
exits, otherwise it is <c>false</c>.</p>
</item>
</taglist>
<p>Note however, that not all implementations support every one
of the above <c>Items</c>.</p>
<p>Failure: <c>badarg</c> if <c>Pid</c> is not a local process,
or if <c>Item</c> is not a valid <c>Item</c>.</p>
</desc>
</func>
<func>
<name>processes() -> [pid()]</name>
<fsummary>All processes</fsummary>
<desc>
<p>Returns a list of process identifiers corresponding to
all the processes currently existing on the local node.
</p>
<p>Note that a process that is exiting, exists but is not alive, i.e.,
<c>is_process_alive/1</c> will return <c>false</c> for a process
that is exiting, but its process identifier will be part
of the result returned from <c>processes/0</c>.
</p>
<pre>
> <input>processes().</input>
[<0.0.0>,<0.2.0>,<0.4.0>,<0.5.0>,<0.7.0>,<0.8.0>]</pre>
</desc>
</func>
<func>
<name>purge_module(Module) -> void()</name>
<fsummary>Remove old code for a module</fsummary>
<type>
<v>Module = atom()</v>
</type>
<desc>
<p>Removes old code for <c>Module</c>. Before this BIF is used,
<c>erlang:check_process_code/2</c> should be called to check
that no processes are executing old code in the module.</p>
<warning>
<p>This BIF is intended for the code server (see
<seealso marker="kernel:code">code(3)</seealso>) and should not be
used elsewhere.</p>
</warning>
<p>Failure: <c>badarg</c> if there is no old code for
<c>Module</c>.</p>
</desc>
</func>
<func>
<name>put(Key, Val) -> OldVal | undefined</name>
<fsummary>Add a new value to the process dictionary</fsummary>
<type>
<v>Key = Val = OldVal = term()</v>
</type>
<desc>
<p>Adds a new <c>Key</c> to the process dictionary, associated
with the value <c>Val</c>, and returns <c>undefined</c>. If
<c>Key</c> already exists, the old value is deleted and
replaced by <c>Val</c> and the function returns the old value.</p>
<note>
<p>The values stored when <c>put</c> is evaluated within
the scope of a <c>catch</c> will not be retracted if a
<c>throw</c> is evaluated, or if an error occurs.</p>
</note>
<pre>
> <input>X = put(name, walrus), Y = put(name, carpenter),</input>
<input>Z = get(name),</input>
<input>{X, Y, Z}.</input>
{undefined,walrus,carpenter}</pre>
</desc>
</func>
<func>
<name>erlang:raise(Class, Reason, Stacktrace)</name>
<fsummary>Stop execution with an exception of given class, reason and call stack backtrace</fsummary>
<type>
<v>Class = error | exit | throw</v>
<v>Reason = term()</v>
<v>Stacktrace = [{Module, Function, Arity | Args} | {Fun, Args}]</v>
<v> Module = Function = atom()</v>
<v> Arity = int()</v>
<v> Args = [term()]</v>
<v> Fun = [fun()]</v>
</type>
<desc>
<p>Stops the execution of the calling process with an
exception of given class, reason and call stack backtrace
(<em>stacktrace</em>).</p>
<warning>
<p>This BIF is intended for debugging and for use in
the Erlang operating system. In general, it should
be avoided in applications, unless you know
very well what you are doing.</p>
</warning>
<p><c>Class</c> is one of <c>error</c>, <c>exit</c> or
<c>throw</c>, so if it were not for the stacktrace
<c>erlang:raise(Class, Reason, Stacktrace)</c> is
equivalent to <c>erlang:Class(Reason)</c>.
<c>Reason</c> is any term and <c>Stacktrace</c> is a list as
returned from <c>get_stacktrace()</c>, that is a list of
3-tuples <c>{Module, Function, Arity | Args}</c> where
<c>Module</c> and <c>Function</c> are atoms and the third
element is an integer arity or an argument list. The
stacktrace may also contain <c>{Fun, Args}</c> tuples where
<c>Fun</c> is a local fun and <c>Args</c> is an argument list.</p>
<p>The stacktrace is used as the exception stacktrace for the
calling process; it will be truncated to the current
maximum stacktrace depth.</p>
<p>Because evaluating this function causes the process to
terminate, it has no return value - unless the arguments are
invalid, in which case the function <em>returns the error reason</em>, that is <c>badarg</c>. If you want to be
really sure not to return you can call
<c>error(erlang:raise(Class, Reason, Stacktrace))</c>
and hope to distinguish exceptions later.</p>
</desc>
</func>
<func>
<name>erlang:read_timer(TimerRef) -> int() | false</name>
<fsummary>Number of milliseconds remaining for a timer</fsummary>
<type>
<v>TimerRef = reference()</v>
</type>
<desc>
<p><c>TimerRef</c> is a timer reference returned by
<seealso marker="#send_after/3">erlang:send_after/3</seealso>
or
<seealso marker="#start_timer/3">erlang:start_timer/3</seealso>.
If the timer is active, the function returns the time in
milliseconds left until the timer will expire, otherwise
<c>false</c> (which means that <c>TimerRef</c> was never a
timer, that it has been cancelled, or that it has already
delivered its message).</p>
<p>See also
<seealso marker="#send_after/3">erlang:send_after/3</seealso>,
<seealso marker="#start_timer/3">erlang:start_timer/3</seealso>,
and
<seealso marker="#cancel_timer/1">erlang:cancel_timer/1</seealso>.</p>
</desc>
</func>
<func>
<name>erlang:ref_to_list(Ref) -> string()</name>
<fsummary>Text representation of a reference</fsummary>
<type>
<v>Ref = reference()</v>
</type>
<desc>
<p>Returns a string which corresponds to the text
representation of <c>Ref</c>.</p>
<warning>
<p>This BIF is intended for debugging and for use in
the Erlang operating system. It should not be used in
application programs.</p>
</warning>
</desc>
</func>
<func>
<name>register(RegName, Pid | Port) -> true</name>
<fsummary>Register a name for a pid (or port)</fsummary>
<type>
<v>RegName = atom()</v>
<v>Pid = pid()</v>
<v>Port = port()</v>
</type>
<desc>
<p>Associates the name <c>RegName</c> with a pid or a port
identifier. <c>RegName</c>, which must be an atom, can be used
instead of the pid / port identifier in the send operator
(<c>RegName ! Message</c>).</p>
<pre>
> <input>register(db, Pid).</input>
true</pre>
<p>Failure: <c>badarg</c> if <c>Pid</c> is not an existing,
local process or port, if <c>RegName</c> is already in use,
if the process or port is already registered (already has a
name), or if <c>RegName</c> is the atom <c>undefined</c>.</p>
</desc>
</func>
<func>
<name>registered() -> [RegName]</name>
<fsummary>All registered names</fsummary>
<type>
<v>RegName = atom()</v>
</type>
<desc>
<p>Returns a list of names which have been registered using
<seealso marker="#register/2">register/2</seealso>.</p>
<pre>
> <input>registered().</input>
[code_server, file_server, init, user, my_db]</pre>
</desc>
</func>
<func>
<name>erlang:resume_process(Suspendee) -> true</name>
<fsummary>Resume a suspended process</fsummary>
<type>
<v>Suspendee = pid()</v>
</type>
<desc>
<p>Decreases the suspend count on the process identified by
<c>Suspendee</c>. <c>Suspendee</c> should previously have been
suspended via
<seealso marker="#suspend_process/2">erlang:suspend_process/2</seealso>,
or
<seealso marker="#suspend_process/1">erlang:suspend_process/1</seealso>
by the process calling <c>erlang:resume_process(Suspendee)</c>. When
the suspend count on <c>Suspendee</c> reach zero, <c>Suspendee</c>
will be resumed, i.e., the state of the <c>Suspendee</c> is changed
from suspended into the state <c>Suspendee</c> was in before it was
suspended.
</p>
<warning>
<p>This BIF is intended for debugging only.</p>
</warning>
<p>Failures:</p>
<taglist>
<tag><c>badarg</c></tag>
<item>
If <c>Suspendee</c> isn't a process identifier.
</item>
<tag><c>badarg</c></tag>
<item>
If the process calling <c>erlang:resume_process/1</c> had
not previously increased the suspend count on the process
identified by <c>Suspendee</c>.
</item>
<tag><c>badarg</c></tag>
<item>
If the process identified by <c>Suspendee</c> is not alive.
</item>
</taglist>
</desc>
</func>
<func>
<name>round(Number) -> int()</name>
<fsummary>Return an integer by rounding a number</fsummary>
<type>
<v>Number = number()</v>
</type>
<desc>
<p>Returns an integer by rounding <c>Number</c>.</p>
<pre>
> <input>round(5.5).</input>
6</pre>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>self() -> pid()</name>
<fsummary>Pid of the calling process</fsummary>
<desc>
<p>Returns the pid (process identifier) of the calling process.</p>
<pre>
> <input>self().</input>
<0.26.0></pre>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>erlang:send(Dest, Msg) -> Msg</name>
<fsummary>Send a message</fsummary>
<type>
<v>Dest = pid() | port() | RegName | {RegName, Node}</v>
<v>Msg = term()</v>
<v> RegName = atom()</v>
<v> Node = node()</v>
</type>
<desc>
<p>Sends a message and returns <c>Msg</c>. This is the same as
<c>Dest ! Msg</c>.</p>
<p><c>Dest</c> may be a remote or local pid, a (local) port, a
locally registered name, or a tuple <c>{RegName, Node}</c>
for a registered name at another node.</p>
</desc>
</func>
<func>
<name>erlang:send(Dest, Msg, [Option]) -> Res</name>
<fsummary>Send a message conditionally</fsummary>
<type>
<v>Dest = pid() | port() | RegName | {RegName, Node}</v>
<v> RegName = atom()</v>
<v> Node = node()</v>
<v>Msg = term()</v>
<v>Option = nosuspend | noconnect</v>
<v>Res = ok | nosuspend | noconnect</v>
</type>
<desc>
<p>Sends a message and returns <c>ok</c>, or does not send
the message but returns something else (see below). Otherwise
the same as
<seealso marker="#send/2">erlang:send/2</seealso>. See
also
<seealso marker="#send_nosuspend/2">erlang:send_nosuspend/2,3</seealso>.
for more detailed explanation and warnings.</p>
<p>The possible options are:</p>
<taglist>
<tag><c>nosuspend</c></tag>
<item>
<p>If the sender would have to be suspended to do the send,
<c>nosuspend</c> is returned instead.</p>
</item>
<tag><c>noconnect</c></tag>
<item>
<p>If the destination node would have to be auto-connected
before doing the send, <c>noconnect</c> is returned
instead.</p>
</item>
</taglist>
<warning>
<p>As with <c>erlang:send_nosuspend/2,3</c>: Use with extreme
care!</p>
</warning>
</desc>
</func>
<func>
<name>erlang:send_after(Time, Dest, Msg) -> TimerRef</name>
<fsummary>Start a timer</fsummary>
<type>
<v>Time = int()</v>
<v> 0 <= Time <= 4294967295</v>
<v>Dest = pid() | RegName </v>
<v> LocalPid = pid() (of a process, alive or dead, on the local node)</v>
<v>Msg = term()</v>
<v>TimerRef = reference()</v>
</type>
<desc>
<p>Starts a timer which will send the message <c>Msg</c>
to <c>Dest</c> after <c>Time</c> milliseconds.</p>
<p>If <c>Dest</c> is an atom, it is supposed to be the name of
a registered process. The process referred to by the name is
looked up at the time of delivery. No error is given if
the name does not refer to a process.</p>
<p>If <c>Dest</c> is a pid, the timer will be automatically
canceled if the process referred to by the pid is not alive,
or when the process exits. This feature was introduced in
erts version 5.4.11. Note that timers will not be
automatically canceled when <c>Dest</c> is an atom.</p>
<p>See also
<seealso marker="#start_timer/3">erlang:start_timer/3</seealso>,
<seealso marker="#cancel_timer/1">erlang:cancel_timer/1</seealso>,
and
<seealso marker="#read_timer/1">erlang:read_timer/1</seealso>.</p>
<p>Failure: <c>badarg</c> if the arguments does not satisfy
the requirements specified above.</p>
</desc>
</func>
<func>
<name>erlang:send_nosuspend(Dest, Msg) -> bool()</name>
<fsummary>Try to send a message without ever blocking</fsummary>
<type>
<v>Dest = pid() | port() | RegName | {RegName, Node}</v>
<v> RegName = atom()</v>
<v> Node = node()</v>
<v>Msg = term()</v>
</type>
<desc>
<p>The same as
<seealso marker="#send/3">erlang:send(Dest, Msg, [nosuspend])</seealso>, but returns <c>true</c> if
the message was sent and <c>false</c> if the message was not
sent because the sender would have had to be suspended.</p>
<p>This function is intended for send operations towards an
unreliable remote node without ever blocking the sending
(Erlang) process. If the connection to the remote node
(usually not a real Erlang node, but a node written in C or
Java) is overloaded, this function <em>will not send the message</em> but return <c>false</c> instead.</p>
<p>The same happens, if <c>Dest</c> refers to a local port that
is busy. For all other destinations (allowed for the ordinary
send operator <c>'!'</c>) this function sends the message and
returns <c>true</c>.</p>
<p>This function is only to be used in very rare circumstances
where a process communicates with Erlang nodes that can
disappear without any trace causing the TCP buffers and
the drivers queue to be over-full before the node will actually
be shut down (due to tick timeouts) by <c>net_kernel</c>. The
normal reaction to take when this happens is some kind of
premature shutdown of the other node.</p>
<p>Note that ignoring the return value from this function would
result in <em>unreliable</em> message passing, which is
contradictory to the Erlang programming model. The message is
<em>not</em> sent if this function returns <c>false</c>.</p>
<p>Note also that in many systems, transient states of
overloaded queues are normal. The fact that this function
returns <c>false</c> does not in any way mean that the other
node is guaranteed to be non-responsive, it could be a
temporary overload. Also a return value of <c>true</c> does
only mean that the message could be sent on the (TCP) channel
without blocking, the message is not guaranteed to have
arrived at the remote node. Also in the case of a disconnected
non-responsive node, the return value is <c>true</c> (mimics
the behaviour of the <c>!</c> operator). The expected
behaviour as well as the actions to take when the function
returns <c>false</c> are application and hardware specific.</p>
<warning>
<p>Use with extreme care!</p>
</warning>
</desc>
</func>
<func>
<name>erlang:send_nosuspend(Dest, Msg, Options) -> bool()</name>
<fsummary>Try to send a message without ever blocking</fsummary>
<type>
<v>Dest = pid() | port() | RegName | {RegName, Node}</v>
<v> RegName = atom()</v>
<v> Node = node()</v>
<v>Msg = term()</v>
<v>Option = noconnect</v>
</type>
<desc>
<p>The same as
<seealso marker="#send/3">erlang:send(Dest, Msg, [nosuspend | Options])</seealso>,
but with boolean return value.</p>
<p>This function behaves like
<seealso marker="#send_nosuspend/2">erlang:send_nosuspend/2)</seealso>,
but takes a third parameter, a list of options. The only
currently implemented option is <c>noconnect</c>. The option
<c>noconnect</c> makes the function return <c>false</c> if
the remote node is not currently reachable by the local
node. The normal behaviour is to try to connect to the node,
which may stall the process for a shorter period. The use of
the <c>noconnect</c> option makes it possible to be
absolutely sure not to get even the slightest delay when
sending to a remote process. This is especially useful when
communicating with nodes who expect to always be
the connecting part (i.e. nodes written in C or Java).</p>
<p>Whenever the function returns <c>false</c> (either when a
suspend would occur or when <c>noconnect</c> was specified and
the node was not already connected), the message is guaranteed
<em>not</em> to have been sent.</p>
<warning>
<p>Use with extreme care!</p>
</warning>
</desc>
</func>
<func>
<name>erlang:set_cookie(Node, Cookie) -> true</name>
<fsummary>Set the magic cookie of a node</fsummary>
<type>
<v>Node = node()</v>
<v>Cookie = atom()</v>
</type>
<desc>
<p>Sets the magic cookie of <c>Node</c> to the atom
<c>Cookie</c>. If <c>Node</c> is the local node, the function
also sets the cookie of all other unknown nodes to
<c>Cookie</c> (see
<seealso marker="doc/reference_manual:distributed">Distributed Erlang</seealso> in the Erlang Reference Manual).</p>
<p>Failure: <c>function_clause</c> if the local node is not
alive.</p>
</desc>
</func>
<func>
<name>setelement(Index, Tuple1, Value) -> Tuple2</name>
<fsummary>Set Nth element of a tuple</fsummary>
<type>
<v>Index = 1..tuple_size(Tuple1)</v>
<v>Tuple1 = Tuple2 = tuple()</v>
<v>Value = term()</v>
</type>
<desc>
<p>Returns a tuple which is a copy of the argument <c>Tuple1</c>
with the element given by the integer argument <c>Index</c>
(the first element is the element with index 1) replaced by
the argument <c>Value</c>.</p>
<pre>
> <input>setelement(2, {10, green, bottles}, red).</input>
{10,red,bottles}</pre>
</desc>
</func>
<func>
<name>size(Item) -> int()</name>
<fsummary>Size of a tuple or binary</fsummary>
<type>
<v>Item = tuple() | binary()</v>
</type>
<desc>
<p>Returns an integer which is the size of the argument
<c>Item</c>, which must be either a tuple or a binary.</p>
<pre>
> <input>size({morni, mulle, bwange}).</input>
3</pre>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>spawn(Fun) -> pid()</name>
<fsummary>Create a new process with a fun as entry point</fsummary>
<type>
<v>Fun = fun()</v>
</type>
<desc>
<p>Returns the pid of a new process started by the application
of <c>Fun</c> to the empty list <c>[]</c>. Otherwise works
like <seealso marker="#spawn/3">spawn/3</seealso>.</p>
</desc>
</func>
<func>
<name>spawn(Node, Fun) -> pid()</name>
<fsummary>Create a new process with a fun as entry point on a given node</fsummary>
<type>
<v>Node = node()</v>
<v>Fun = fun()</v>
</type>
<desc>
<p>Returns the pid of a new process started by the application
of <c>Fun</c> to the empty list <c>[]</c> on <c>Node</c>. If
<c>Node</c> does not exist, a useless pid is returned.
Otherwise works like
<seealso marker="#spawn/3">spawn/3</seealso>.</p>
</desc>
</func>
<func>
<name>spawn(Module, Function, Args) -> pid()</name>
<fsummary>Create a new process with a function as entry point</fsummary>
<type>
<v>Module = Function = atom()</v>
<v>Args = [term()]</v>
</type>
<desc>
<p>Returns the pid of a new process started by the application
of <c>Module:Function</c> to <c>Args</c>. The new process
created will be placed in the system scheduler queue and be
run some time later.</p>
<p><c>error_handler:undefined_function(Module, Function, Args)</c> is evaluated by the new process if
<c>Module:Function/Arity</c> does not exist (where
<c>Arity</c> is the length of <c>Args</c>). The error handler
can be redefined (see
<seealso marker="#process_flag/2">process_flag/2</seealso>).
If <c>error_handler</c> is undefined, or the user has
redefined the default <c>error_handler</c> its replacement is
undefined, a failure with the reason <c>undef</c> will occur.</p>
<pre>
> <input>spawn(speed, regulator, [high_speed, thin_cut]).</input>
<0.13.1></pre>
</desc>
</func>
<func>
<name>spawn(Node, Module, Function, Args) -> pid()</name>
<fsummary>Create a new process with a function as entry point on a given node</fsummary>
<type>
<v>Node = node()</v>
<v>Module = Function = atom()</v>
<v>Args = [term()]</v>
</type>
<desc>
<p>Returns the pid of a new process started by the application
of <c>Module:Function</c> to <c>Args</c> on <c>Node</c>. If
<c>Node</c> does not exists, a useless pid is returned.
Otherwise works like
<seealso marker="#spawn/3">spawn/3</seealso>.</p>
</desc>
</func>
<func>
<name>spawn_link(Fun) -> pid()</name>
<fsummary>Create and link to a new process with a fun as entry point</fsummary>
<type>
<v>Fun = fun()</v>
</type>
<desc>
<p>Returns the pid of a new process started by the application
of <c>Fun</c> to the empty list []. A link is created between
the calling process and the new process, atomically.
Otherwise works like
<seealso marker="#spawn/3">spawn/3</seealso>.</p>
</desc>
</func>
<func>
<name>spawn_link(Node, Fun) -> pid()</name>
<fsummary>Create and link to a new process with a fun as entry point on a specified node</fsummary>
<type>
<v>Node = node()</v>
<v>Fun = fun()</v>
</type>
<desc>
<p>Returns the pid of a new process started by the application
of <c>Fun</c> to the empty list [] on <c>Node</c>. A link is
created between the calling process and the new process,
atomically. If <c>Node</c> does not exist, a useless pid is
returned (and due to the link, an exit signal with exit
reason <c>noconnection</c> will be received). Otherwise works
like <seealso marker="#spawn/3">spawn/3</seealso>.</p>
</desc>
</func>
<func>
<name>spawn_link(Module, Function, Args) -> pid()</name>
<fsummary>Create and link to a new process with a function as entry point</fsummary>
<type>
<v>Module = Function = atom()</v>
<v>Args = [term()]</v>
</type>
<desc>
<p>Returns the pid of a new process started by the application
of <c>Module:Function</c> to <c>Args</c>. A link is created
between the calling process and the new process, atomically.
Otherwise works like
<seealso marker="#spawn/3">spawn/3</seealso>.</p>
</desc>
</func>
<func>
<name>spawn_link(Node, Module, Function, Args) -> pid()</name>
<fsummary>Create and link to a new process with a function as entry point on a given node</fsummary>
<type>
<v>Node = node()</v>
<v>Module = Function = atom()</v>
<v>Args = [term()]</v>
</type>
<desc>
<p>Returns the pid of a new process started by the application
of <c>Module:Function</c> to <c>Args</c> on <c>Node</c>. A
link is created between the calling process and the new
process, atomically. If <c>Node</c> does not exist, a useless
pid is returned (and due to the link, an exit signal with exit
reason <c>noconnection</c> will be received). Otherwise works
like <seealso marker="#spawn/3">spawn/3</seealso>.</p>
</desc>
</func>
<func>
<name>spawn_monitor(Fun) -> {pid(),reference()}</name>
<fsummary>Create and monitor a new process with a fun as entry point</fsummary>
<type>
<v>Fun = fun()</v>
</type>
<desc>
<p>Returns the pid of a new process started by the application
of <c>Fun</c> to the empty list [] and reference for a monitor
created to the new process.
Otherwise works like
<seealso marker="#spawn/3">spawn/3</seealso>.</p>
</desc>
</func>
<func>
<name>spawn_monitor(Module, Function, Args) -> {pid(),reference()}</name>
<fsummary>Create and monitor a new process with a function as entry point</fsummary>
<type>
<v>Module = Function = atom()</v>
<v>Args = [term()]</v>
</type>
<desc>
<p>A new process is started by the application
of <c>Module:Function</c> to <c>Args</c>, and the process is
monitored at the same time. Returns the pid and a reference
for the monitor.
Otherwise works like
<seealso marker="#spawn/3">spawn/3</seealso>.</p>
</desc>
</func>
<func>
<name>spawn_opt(Fun, [Option]) -> pid() | {pid(),reference()}</name>
<fsummary>Create a new process with a fun as entry point</fsummary>
<type>
<v>Fun = fun()</v>
<v>Option = link | monitor | {priority, Level} | {fullsweep_after, Number} | {min_heap_size, Size} | {min_bin_vheap_size, VSize}</v>
<v> Level = low | normal | high</v>
<v> Number = int()</v>
<v> Size = int()</v>
<v> VSize = int()</v>
</type>
<desc>
<p>Returns the pid of a new process started by the application
of <c>Fun</c> to the empty list <c>[]</c>. Otherwise
works like
<seealso marker="#spawn_opt/4">spawn_opt/4</seealso>.</p>
<p>If the option <c>monitor</c> is given, the newly created
process will be monitored and both the pid and reference for
the monitor will be returned.</p>
</desc>
</func>
<func>
<name>spawn_opt(Node, Fun, [Option]) -> pid()</name>
<fsummary>Create a new process with a fun as entry point on a given node</fsummary>
<type>
<v>Node = node()</v>
<v>Fun = fun()</v>
<v>Option = link | {priority, Level} | {fullsweep_after, Number} | {min_heap_size, Size} | {min_bin_vheap_size, VSize}</v>
<v> Level = low | normal | high</v>
<v> Number = int()</v>
<v> Size = int()</v>
<v> VSize = int()</v>
</type>
<desc>
<p>Returns the pid of a new process started by the application
of <c>Fun</c> to the empty list <c>[]</c> on <c>Node</c>. If
<c>Node</c> does not exist, a useless pid is returned.
Otherwise works like
<seealso marker="#spawn_opt/4">spawn_opt/4</seealso>.</p>
</desc>
</func>
<func>
<name>spawn_opt(Module, Function, Args, [Option]) -> pid() | {pid(),reference()}</name>
<fsummary>Create a new process with a function as entry point</fsummary>
<type>
<v>Module = Function = atom()</v>
<v>Args = [term()]</v>
<v>Option = link | monitor | {priority, Level} | {fullsweep_after, Number} | {min_heap_size, Size} | {min_bin_vheap_size, VSize}</v>
<v> Level = low | normal | high</v>
<v> Number = int()</v>
<v> Size = int()</v>
<v> VSize = int()</v>
</type>
<desc>
<p>Works exactly like
<seealso marker="#spawn/3">spawn/3</seealso>, except that an
extra option list is given when creating the process.</p>
<p>If the option <c>monitor</c> is given, the newly created
process will be monitored and both the pid and reference for
the monitor will be returned.</p>
<taglist>
<tag><c>link</c></tag>
<item>
<p>Sets a link to the parent process (like
<c>spawn_link/3</c> does).</p>
</item>
<tag><c>monitor</c></tag>
<item>
<p>Monitor the new process (just like
<seealso marker="#monitor/2">monitor/2</seealso> does).</p>
</item>
<tag><c>{priority, Level}</c></tag>
<item>
<p>Sets the priority of the new process. Equivalent to
executing
<seealso marker="#process_flag_priority">process_flag(priority, Level)</seealso> in the start function of the new process,
except that the priority will be set before the process is
selected for execution for the first time. For more information
on priorities see
<seealso marker="#process_flag_priority">process_flag(priority, Level)</seealso>.</p>
</item>
<tag><c>{fullsweep_after, Number}</c></tag>
<item>
<p>This option is only useful for performance tuning.
In general, you should not use this option unless you
know that there is problem with execution times and/or
memory consumption, and you should measure to make sure
that the option improved matters.
</p>
<p>The Erlang runtime system uses a generational garbage
collection scheme, using an "old heap" for data that has
survived at least one garbage collection. When there is
no more room on the old heap, a fullsweep garbage
collection will be done.</p>
<p>The <c>fullsweep_after</c> option makes it possible to
specify the maximum number of generational collections
before forcing a fullsweep even if there is still room on
the old heap. Setting the number to zero effectively
disables the general collection algorithm, meaning that
all live data is copied at every garbage collection.</p>
<p>Here are a few cases when it could be useful to change
<c>fullsweep_after</c>. Firstly, if binaries that are no
longer used should be thrown away as soon as possible.
(Set <c>Number</c> to zero.) Secondly, a process that
mostly have short-lived data will be fullsweeped seldom
or never, meaning that the old heap will contain mostly
garbage. To ensure a fullsweep once in a while, set
<c>Number</c> to a suitable value such as 10 or 20.
Thirdly, in embedded systems with limited amount of RAM
and no virtual memory, one might want to preserve memory
by setting <c>Number</c> to zero. (The value may be set
globally, see
<seealso marker="#system_flag/2">erlang:system_flag/2</seealso>.)</p>
</item>
<tag><c>{min_heap_size, Size}</c></tag>
<item>
<p>This option is only useful for performance tuning.
In general, you should not use this option unless you
know that there is problem with execution times and/or
memory consumption, and you should measure to make sure
that the option improved matters.
</p>
<p>Gives a minimum heap size in words. Setting this value
higher than the system default might speed up some
processes because less garbage collection is done.
Setting too high value, however, might waste memory and
slow down the system due to worse data locality.
Therefore, it is recommended to use this option only for
fine-tuning an application and to measure the execution
time with various <c>Size</c> values.</p>
</item>
<tag><c>{min_bin_vheap_size, VSize}</c></tag>
<item>
<p>This option is only useful for performance tuning.
In general, you should not use this option unless you
know that there is problem with execution times and/or
memory consumption, and you should measure to make sure
that the option improved matters.
</p>
<p>Gives a minimum binary virtual heap size in words. Setting this value
higher than the system default might speed up some
processes because less garbage collection is done.
Setting too high value, however, might waste memory.
Therefore, it is recommended to use this option only for
fine-tuning an application and to measure the execution
time with various <c>VSize</c> values.</p>
</item>
</taglist>
</desc>
</func>
<func>
<name>spawn_opt(Node, Module, Function, Args, [Option]) -> pid()</name>
<fsummary>Create a new process with a function as entry point on a given node</fsummary>
<type>
<v>Node = node()</v>
<v>Module = Function = atom()</v>
<v>Args = [term()]</v>
<v>Option = link | {priority, Level} | {fullsweep_after, Number} | {min_heap_size, Size} | {min_bin_vheap_size, VSize}</v>
<v> Level = low | normal | high</v>
<v> Number = int()</v>
<v> Size = int()</v>
<v> VSize = int()</v>
</type>
<desc>
<p>Returns the pid of a new process started by the application
of <c>Module:Function</c> to <c>Args</c> on <c>Node</c>. If
<c>Node</c> does not exist, a useless pid is returned.
Otherwise works like
<seealso marker="#spawn_opt/4">spawn_opt/4</seealso>.</p>
</desc>
</func>
<func>
<name>split_binary(Bin, Pos) -> {Bin1, Bin2}</name>
<fsummary>Split a binary into two</fsummary>
<type>
<v>Bin = Bin1 = Bin2 = binary()</v>
<v>Pos = 0..byte_size(Bin)</v>
</type>
<desc>
<p>Returns a tuple containing the binaries which are the result
of splitting <c>Bin</c> into two parts at position <c>Pos</c>.
This is not a destructive operation. After the operation,
there will be three binaries altogether.</p>
<pre>
> <input>B = list_to_binary("0123456789").</input>
<<"0123456789">>
> <input>byte_size(B).</input>
10
> <input>{B1, B2} = split_binary(B,3).</input>
{<<"012">>,<<"3456789">>}
> <input>byte_size(B1).</input>
3
> <input>byte_size(B2).</input>
7</pre>
</desc>
</func>
<func>
<name>erlang:start_timer(Time, Dest, Msg) -> TimerRef</name>
<fsummary>Start a timer</fsummary>
<type>
<v>Time = int()</v>
<v> 0 <= Time <= 4294967295</v>
<v>Dest = LocalPid | RegName </v>
<v> LocalPid = pid() (of a process, alive or dead, on the local node)</v>
<v> RegName = atom()</v>
<v>Msg = term()</v>
<v>TimerRef = reference()</v>
</type>
<desc>
<p>Starts a timer which will send the message
<c>{timeout, TimerRef, Msg}</c> to <c>Dest</c>
after <c>Time</c> milliseconds.</p>
<p>If <c>Dest</c> is an atom, it is supposed to be the name of
a registered process. The process referred to by the name is
looked up at the time of delivery. No error is given if
the name does not refer to a process.</p>
<p>If <c>Dest</c> is a pid, the timer will be automatically
canceled if the process referred to by the pid is not alive,
or when the process exits. This feature was introduced in
erts version 5.4.11. Note that timers will not be
automatically canceled when <c>Dest</c> is an atom.</p>
<p>See also
<seealso marker="#send_after/3">erlang:send_after/3</seealso>,
<seealso marker="#cancel_timer/1">erlang:cancel_timer/1</seealso>,
and
<seealso marker="#read_timer/1">erlang:read_timer/1</seealso>.</p>
<p>Failure: <c>badarg</c> if the arguments does not satisfy
the requirements specified above.</p>
</desc>
</func>
<func>
<name>statistics(Type) -> Res</name>
<fsummary>Information about the system</fsummary>
<type>
<v>Type, Res -- see below</v>
</type>
<desc>
<p>Returns information about the system as specified by
<c>Type</c>:</p>
<taglist>
<tag><c>context_switches</c></tag>
<item>
<p>Returns <c>{ContextSwitches, 0}</c>, where
<c>ContextSwitches</c> is the total number of context
switches since the system started.</p>
</item>
<tag><marker id="statistics_exact_reductions"><c>exact_reductions</c></marker></tag>
<item>
<p>Returns
<c>{Total_Exact_Reductions, Exact_Reductions_Since_Last_Call}</c>.</p>
<p><em>NOTE:</em><c>statistics(exact_reductions)</c> is
a more expensive operation than
<seealso marker="#statistics_reductions">statistics(reductions)</seealso>
especially on an Erlang machine with SMP support.</p>
</item>
<tag><c>garbage_collection</c></tag>
<item>
<p>Returns <c>{Number_of_GCs, Words_Reclaimed, 0}</c>. This
information may not be valid for all implementations.</p>
</item>
<tag><c>io</c></tag>
<item>
<p>Returns <c>{{input, Input}, {output, Output}}</c>,
where <c>Input</c> is the total number of bytes received
through ports, and <c>Output</c> is the total number of
bytes output to ports.</p>
</item>
<tag><marker id="statistics_reductions"><c>reductions</c></marker></tag>
<item>
<p>Returns
<c>{Total_Reductions, Reductions_Since_Last_Call}</c>.</p>
<p><em>NOTE:</em> From erts version 5.5 (OTP release R11B)
this value does not include reductions performed in current
time slices of currently scheduled processes. If an
exact value is wanted, use
<seealso marker="#statistics_exact_reductions">statistics(exact_reductions)</seealso>.</p>
</item>
<tag><c>run_queue</c></tag>
<item>
<p>Returns the length of the run queue, that is, the number
of processes that are ready to run.</p>
</item>
<tag><c>runtime</c></tag>
<item>
<p>Returns <c>{Total_Run_Time, Time_Since_Last_Call}</c>.
Note that the run-time is the sum of the run-time for all
threads in the Erlang run-time system and may therefore be greater
than the wall-clock time.</p>
</item>
<tag><c>wall_clock</c></tag>
<item>
<p>Returns
<c>{Total_Wallclock_Time, Wallclock_Time_Since_Last_Call}</c>.
<c>wall_clock</c> can be used in the same manner as
<c>runtime</c>, except that real time is measured as
opposed to runtime or CPU time.</p>
</item>
</taglist>
<p>All times are in milliseconds.</p>
<pre>
> <input>statistics(runtime).</input>
{1690,1620}
> <input>statistics(reductions).</input>
{2046,11}
> <input>statistics(garbage_collection).</input>
{85,23961,0}</pre>
</desc>
</func>
<func>
<name>erlang:suspend_process(Suspendee, OptList) -> true | false</name>
<fsummary>Suspend a process</fsummary>
<type>
<v>Suspendee = pid()</v>
<v>OptList = [Opt]</v>
<v>Opt = atom()</v>
</type>
<desc>
<p>Increases the suspend count on the process identified by
<c>Suspendee</c> and puts it in the suspended state if it isn't
already in the suspended state. A suspended process will not be
scheduled for execution until the process has been resumed.
</p>
<p>A process can be suspended by multiple processes and can
be suspended multiple times by a single process. A suspended
process will not leave the suspended state until its suspend
count reach zero. The suspend count of <c>Suspendee</c> is
decreased when
<seealso marker="#resume_process/1">erlang:resume_process(Suspendee)</seealso>
is called by the same process that called
<c>erlang:suspend_process(Suspendee)</c>. All increased suspend
counts on other processes acquired by a process will automatically be
decreased when the process terminates.</p>
<p>Currently the following options (<c>Opt</c>s) are available:</p>
<taglist>
<tag><c>asynchronous</c></tag>
<item>
A suspend request is sent to the process identified by
<c>Suspendee</c>. <c>Suspendee</c> will eventually suspend
unless it is resumed before it was able to suspend. The caller
of <c>erlang:suspend_process/2</c> will return immediately,
regardless of whether the <c>Suspendee</c> has suspended yet
or not. Note that the point in time when the <c>Suspendee</c>
will actually suspend cannot be deduced from other events
in the system. The only guarantee given is that the
<c>Suspendee</c> will <em>eventually</em> suspend (unless it
is resumed). If the <c>asynchronous</c> option has <em>not</em>
been passed, the caller of <c>erlang:suspend_process/2</c> will
be blocked until the <c>Suspendee</c> has actually suspended.
</item>
<tag><c>unless_suspending</c></tag>
<item>
The process identified by <c>Suspendee</c> will be suspended
unless the calling process already is suspending the
<c>Suspendee</c>. If <c>unless_suspending</c> is combined
with the <c>asynchronous</c> option, a suspend request will be
sent unless the calling process already is suspending the
<c>Suspendee</c> or if a suspend request already has been sent
and is in transit. If the calling process already is suspending
the <c>Suspendee</c>, or if combined with the <c>asynchronous</c>
option and a send request already is in transit,
<c>false</c> is returned and the suspend count on <c>Suspendee</c>
will remain unchanged.
</item>
</taglist>
<p>If the suspend count on the process identified by
<c>Suspendee</c> was increased, <c>true</c> is returned; otherwise,
<c>false</c> is returned.</p>
<warning>
<p>This BIF is intended for debugging only.</p>
</warning>
<p>Failures:</p>
<taglist>
<tag><c>badarg</c></tag>
<item>
If <c>Suspendee</c> isn't a process identifier.
</item>
<tag><c>badarg</c></tag>
<item>
If the process identified by <c>Suspendee</c> is same the process as
the process calling <c>erlang:suspend_process/2</c>.
</item>
<tag><c>badarg</c></tag>
<item>
If the process identified by <c>Suspendee</c> is not alive.
</item>
<tag><c>badarg</c></tag>
<item>
If the process identified by <c>Suspendee</c> resides on another node.
</item>
<tag><c>badarg</c></tag>
<item>
If <c>OptList</c> isn't a proper list of valid <c>Opt</c>s.
</item>
<tag><c>system_limit</c></tag>
<item>
If the process identified by <c>Suspendee</c> has been suspended more
times by the calling process than can be represented by the
currently used internal data structures. The current system limit
is larger than 2 000 000 000 suspends, and it will never be less
than that.
</item>
</taglist>
</desc>
</func>
<func>
<name>erlang:suspend_process(Suspendee) -> true</name>
<fsummary>Suspend a process</fsummary>
<type>
<v>Suspendee = pid()</v>
</type>
<desc>
<p>Suspends the process identified by <c>Suspendee</c>. The
same as calling
<seealso marker="#suspend_process/2">erlang:suspend_process(Suspendee, [])</seealso>. For more information see the documentation of <seealso marker="#suspend_process/2">erlang:suspend_process/2</seealso>.
</p>
<warning>
<p>This BIF is intended for debugging only.</p>
</warning>
</desc>
</func>
<func>
<name>erlang:system_flag(Flag, Value) -> OldValue</name>
<fsummary>Set system flags</fsummary>
<type>
<v>Flag, Value, OldValue -- see below</v>
</type>
<desc>
<p>Sets various system properties of the Erlang node. Returns
the old value of the flag.</p>
<taglist>
<tag><c>erlang:system_flag(backtrace_depth, Depth)</c></tag>
<item>
<p>Sets the maximum depth of call stack back-traces in the
exit reason element of <c>'EXIT'</c> tuples.</p>
</item>
<tag><marker id="system_flag_cpu_topology"><c>erlang:system_flag(cpu_topology, CpuTopology)</c></marker></tag>
<item>
<p>Sets the user defined <c>CpuTopology</c>. The user defined
CPU topology will override any automatically detected
CPU topology. By passing <c>undefined</c> as <c>CpuTopology</c>
the system will revert back to the CPU topology automatically
detected. The returned value equals the value returned
from <c>erlang:system_info(cpu_topology)</c> before the
change was made.
</p>
<p>The CPU topology is used when binding schedulers to logical
processors. If schedulers are already bound when the CPU
topology is changed, the schedulers will be sent a request
to rebind according to the new CPU topology.
</p>
<p>The user defined CPU topology can also be set by passing
the <seealso marker="erl#+sct">+sct</seealso> command
line argument to <c>erl</c>.
</p>
<p>For information on the <c>CpuTopology</c> type
and more, see the documentation of
<seealso marker="#system_info_cpu_topology">erlang:system_info(cpu_topology)</seealso>,
the <c>erl</c> <seealso marker="erl#+sct">+sct</seealso>
emulator flag, and
<seealso marker="#system_flag_scheduler_bind_type">erlang:system_flag(scheduler_bind_type, How)</seealso>.
</p>
</item>
<tag><c>erlang:system_flag(fullsweep_after, Number)</c></tag>
<item>
<p><c>Number</c> is a non-negative integer which indicates
how many times generational garbage collections can be
done without forcing a fullsweep collection. The value
applies to new processes; processes already running are
not affected.</p>
<p>In low-memory systems (especially without virtual
memory), setting the value to 0 can help to conserve
memory.</p>
<p>An alternative way to set this value is through the
(operating system) environment variable
<c>ERL_FULLSWEEP_AFTER</c>.</p>
</item>
<tag><c>erlang:system_flag(min_heap_size, MinHeapSize)</c></tag>
<item>
<p>Sets the default minimum heap size for processes. The
size is given in words. The new <c>min_heap_size</c> only
effects processes spawned after the change of
<c>min_heap_size</c> has been made.
The <c>min_heap_size</c> can be set for individual
processes by use of
<seealso marker="#spawn_opt/4">spawn_opt/N</seealso> or
<seealso marker="#process_flag/2">process_flag/2</seealso>. </p>
</item>
<tag><c>erlang:system_flag(min_bin_vheap_size, MinBinVHeapSize)</c></tag>
<item>
<p>Sets the default minimum binary virtual heap size for processes. The
size is given in words. The new <c>min_bin_vhheap_size</c> only
effects processes spawned after the change of
<c>min_bin_vhheap_size</c> has been made.
The <c>min_bin_vheap_size</c> can be set for individual
processes by use of
<seealso marker="#spawn_opt/4">spawn_opt/N</seealso> or
<seealso marker="#process_flag/2">process_flag/2</seealso>. </p>
</item>
<tag><marker id="system_flag_multi_scheduling"><c>erlang:system_flag(multi_scheduling, BlockState)</c></marker></tag>
<item>
<p><c>BlockState = block | unblock</c></p>
<p>If multi-scheduling is enabled, more than one scheduler
thread is used by the emulator. Multi-scheduling can be
blocked. When multi-scheduling has been blocked, only
one scheduler thread will schedule Erlang processes.</p>
<p>If <c>BlockState =:= block</c>, multi-scheduling will
be blocked. If <c>BlockState =:= unblock</c> and no-one
else is blocking multi-scheduling and this process has
only blocked one time, multi-scheduling will be unblocked.
One process can block multi-scheduling multiple times.
If a process has blocked multiple times, it has to
unblock exactly as many times as it has blocked before it
has released its multi-scheduling block. If a process that
has blocked multi-scheduling exits, it will release its
blocking of multi-scheduling.</p>
<p>The return values are <c>disabled</c>, <c>blocked</c>,
or <c>enabled</c>. The returned value describes the
state just after the call to
<c>erlang:system_flag(multi_scheduling, BlockState)</c>
has been made. The return values are described in the
documentation of <seealso marker="#system_info_multi_scheduling">erlang:system_info(multi_scheduling)</seealso>.</p>
<p><em>NOTE</em>: Blocking of multi-scheduling should normally
not be needed. If you feel that you need to
block multi-scheduling, think through the
problem at least a couple of times again.
Blocking multi-scheduling should only be used
as a last resort since it will most likely be
a <em>very inefficient</em> way to solve the
problem.</p>
<p>See also <seealso marker="#system_info_multi_scheduling">erlang:system_info(multi_scheduling)</seealso>,
<seealso marker="#system_info_multi_scheduling_blockers">erlang:system_info(multi_scheduling_blockers)</seealso>, and
<seealso marker="#system_info_schedulers">erlang:system_info(schedulers)</seealso>.</p>
</item>
<tag><marker id="system_flag_scheduler_bind_type"><c>erlang:system_flag(scheduler_bind_type, How)</c></marker></tag>
<item>
<p>Controls if and how schedulers are bound to logical
processors.</p>
<p>When <c>erlang:system_flag(scheduler_bind_type, How)</c> is
called, an asynchronous signal is sent to all schedulers
online which causes them to try to bind or unbind as requested.
<em>NOTE:</em> If a scheduler fails to bind, this
will often be silently ignored. This since it isn't always
possible to verify valid logical processor identifiers. If
an error is reported, it will be reported to the
<c>error_logger</c>. If you want to verify that the
schedulers actually have bound as requested, call
<seealso marker="#system_info_scheduler_bindings">erlang:system_info(scheduler_bindings)</seealso>.
</p>
<p>Schedulers can currently only be bound on newer Linux,
Solaris, and Windows systems, but more systems will be
supported in the future.
</p>
<p>In order for the runtime system to be able to bind schedulers,
the CPU topology needs to be known. If the runtime system fails
to automatically detect the CPU topology, it can be defined.
For more information on how to define the CPU topology, see
<seealso marker="#system_flag_cpu_topology">erlang:system_flag(cpu_topology, CpuTopology)</seealso>.
</p>
<p>The runtime system will by default bind schedulers to logical
processors using the <c>default_bind</c> bind type if the amount
of schedulers are at least equal to the amount of logical
processors configured, binding of schedulers is supported,
and a CPU topology is available at startup.
</p>
<p><em>NOTE:</em> If the Erlang runtime system is the only
operating system process that binds threads to logical processors,
this improves the performance of the runtime system. However,
if other operating system processes (as for example another Erlang
runtime system) also bind threads to logical processors, there
might be a performance penalty instead. If this is the case you,
are are advised to unbind the schedulers using the
<seealso marker="erl#+sbt">+sbtu</seealso> command line argument,
or <c>erlang:system_flag(scheduler_bind_type, unbound)</c>.</p>
<p>Schedulers can be bound in different ways. The <c>How</c>
argument determines how schedulers are bound. <c>How</c> can
currently be one of:</p>
<taglist>
<tag><c>unbound</c></tag>
<item>
<p>Schedulers will not be bound to logical processors, i.e.,
the operating system decides where the scheduler threads
execute, and when to migrate them. This is the default.</p>
</item>
<tag><c>no_spread</c></tag>
<item>
<p>Schedulers with close scheduler identifiers will be bound
as close as possible in hardware.</p>
</item>
<tag><c>thread_spread</c></tag>
<item>
<p>Thread refers to hardware threads (e.g. Intels
hyper-threads). Schedulers with low scheduler identifiers,
will be bound to the first hardware thread of each core,
then schedulers with higher scheduler identifiers will be
bound to the second hardware thread of each core, etc.</p>
</item>
<tag><c>processor_spread</c></tag>
<item>
<p>Schedulers will be spread like <c>thread_spread</c>, but
also over physical processor chips.</p>
</item>
<tag><c>spread</c></tag>
<item>
<p>Schedulers will be spread as much as possible.</p>
</item>
<tag><c>no_node_thread_spread</c></tag>
<item>
<p>Like <c>thread_spread</c>, but if multiple NUMA
(Non-Uniform Memory Access) nodes exists,
schedulers will be spread over one NUMA node at a time,
i.e., all logical processors of one NUMA node will
be bound to schedulers in sequence.</p>
</item>
<tag><c>no_node_processor_spread</c></tag>
<item>
<p>Like <c>processor_spread</c>, but if multiple NUMA
nodes exists, schedulers will be spread over one
NUMA node at a time, i.e., all logical processors of
one NUMA node will be bound to schedulers in sequence.</p>
</item>
<tag><c>thread_no_node_processor_spread</c></tag>
<item>
<p>A combination of <c>thread_spread</c>, and
<c>no_node_processor_spread</c>. Schedulers will be
spread over hardware threads across NUMA nodes, but
schedulers will only be spread over processors internally
in one NUMA node at a time.</p>
</item>
<tag><c>default_bind</c></tag>
<item>
<p>Binds schedulers the default way. Currently the default
is <c>thread_no_node_processor_spread</c> (which might change
in the future).</p>
</item>
</taglist>
<p>How schedulers are bound matters. For example, in
situations when there are fewer running processes than
schedulers online, the runtime system tries to migrate
processes to schedulers with low scheduler identifiers.
The more the schedulers are spread over the hardware,
the more resources will be available to the runtime
system in such situations.
</p>
<p>The value returned equals <c>How</c> before the
<c>scheduler_bind_type</c> flag was changed.</p>
<p>Failure:</p>
<taglist>
<tag><c>notsup</c></tag>
<item>
<p>If binding of schedulers is not supported.</p>
</item>
<tag><c>badarg</c></tag>
<item>
<p>If <c>How</c> isn't one of the documented alternatives.</p>
</item>
<tag><c>badarg</c></tag>
<item>
<p>If no CPU topology information is available.</p>
</item>
</taglist>
<p>The scheduler bind type can also be set by passing
the <seealso marker="erl#+sbt">+sbt</seealso> command
line argument to <c>erl</c>.
</p>
<p>For more information, see
<seealso marker="#system_info_scheduler_bind_type">erlang:system_info(scheduler_bind_type)</seealso>,
<seealso marker="#system_info_scheduler_bindings">erlang:system_info(scheduler_bindings)</seealso>,
the <c>erl</c> <seealso marker="erl#+sbt">+sbt</seealso>
emulator flag, and
<seealso marker="#system_flag_cpu_topology">erlang:system_flag(cpu_topology, CpuTopology)</seealso>.
</p>
</item>
<tag><marker id="system_flag_schedulers_online"><c>erlang:system_flag(schedulers_online, SchedulersOnline)</c></marker></tag>
<item>
<p>Sets the amount of schedulers online. Valid range is
<![CDATA[1 <= SchedulerId <= erlang:system_info(schedulers)]]>.
</p>
<p>For more information see,
<seealso marker="#system_info_schedulers">erlang:system_info(schedulers)</seealso>,
and
<seealso marker="#system_info_schedulers_online">erlang:system_info(schedulers_online)</seealso>.
</p>
</item>
<tag><c>erlang:system_flag(trace_control_word, TCW)</c></tag>
<item>
<p>Sets the value of the node's trace control word to
<c>TCW</c>. <c>TCW</c> should be an unsigned integer. For
more information see documentation of the
<seealso marker="erts:match_spec#set_tcw">set_tcw</seealso>
function in the match specification documentation in the
ERTS User's Guide.</p>
</item>
</taglist>
<note>
<p>The <c>schedulers</c> option has been removed as
of erts version 5.5.3. The number of scheduler
threads is determined at emulator boot time, and
cannot be changed after that.</p>
</note>
</desc>
</func>
<func>
<name>erlang:system_info(Type) -> Res</name>
<fsummary>Information about the system</fsummary>
<type>
<v>Type, Res -- see below</v>
</type>
<desc>
<p>Returns various information about the current system
(emulator) as specified by <c>Type</c>:</p>
<taglist>
<tag><marker id="system_info_allocated_areas"><c>allocated_areas</c></marker></tag>
<item>
<p>Returns a list of tuples with information about
miscellaneous allocated memory areas.</p>
<p>Each tuple contains an atom describing type of memory as
first element and amount of allocated memory in bytes as
second element. In those cases when there is information
present about allocated and used memory, a third element
is present. This third element contains the amount of
used memory in bytes.</p>
<p><c>erlang:system_info(allocated_areas)</c> is intended
for debugging, and the content is highly implementation
dependent. The content of the results will therefore
change when needed without prior notice.</p>
<p><em>Note:</em> The sum of these values is <em>not</em>
the total amount of memory allocated by the emulator.
Some values are part of other values, and some memory
areas are not part of the result. If you are interested
in the total amount of memory allocated by the emulator
see <seealso marker="#memory/0">erlang:memory/0,1</seealso>.</p>
</item>
<tag><marker id="system_info_allocator"><c>allocator</c></marker></tag>
<item>
<p>Returns <c>{Allocator, Version, Features, Settings}.</c></p>
<p>Types:</p>
<list type="bulleted">
<item><c>Allocator = undefined | glibc</c></item>
<item><c>Version = [int()]</c></item>
<item><c>Features = [atom()]</c></item>
<item><c>Settings = [{Subsystem, [{Parameter, Value}]}]</c></item>
<item><c>Subsystem = atom()</c></item>
<item><c>Parameter = atom()</c></item>
<item><c>Value = term()</c></item>
</list>
<p>Explanation:</p>
<list type="bulleted">
<item>
<p><c>Allocator</c> corresponds to the <c>malloc()</c>
implementation used. If <c>Allocator</c> equals
<c>undefined</c>, the <c>malloc()</c> implementation
used could not be identified. Currently
<c>glibc</c> can be identified.</p>
</item>
<item>
<p><c>Version</c> is a list of integers (but not a
string) representing the version of
the <c>malloc()</c> implementation used.</p>
</item>
<item>
<p><c>Features</c> is a list of atoms representing
allocation features used.</p>
</item>
<item>
<p><c>Settings</c> is a list of subsystems, their
configurable parameters, and used values. Settings
may differ between different combinations of
platforms, allocators, and allocation features.
Memory sizes are given in bytes.</p>
</item>
</list>
<p>See also "System Flags Effecting erts_alloc" in
<seealso marker="erts:erts_alloc#flags">erts_alloc(3)</seealso>.</p>
</item>
<tag><marker id="system_info_alloc_util_allocators"><c>alloc_util_allocators</c></marker></tag>
<item>
<p>Returns a list of the names of all allocators
using the ERTS internal <c>alloc_util</c> framework
as atoms. For more information see the
<seealso marker="erts:erts_alloc#alloc_util">"the
alloc_util framework" section in the
erts_alloc(3)</seealso> documentation.
</p>
</item>
<tag><marker id="system_info_allocator_tuple"><c>{allocator, Alloc}</c></marker></tag>
<item>
<p>Returns information about the specified allocator.
As of erts version 5.6.1 the return value is a list
of <c>{instance, InstanceNo, InstanceInfo}</c> tuples
where <c>InstanceInfo</c> contains information about
a specific instance of the allocator.
If <c>Alloc</c> is not a recognized allocator,
<c>undefined</c> is returned. If <c>Alloc</c> is disabled,
<c>false</c> is returned.</p>
<p><em>Note:</em> The information returned is highly
implementation dependent and may be changed, or removed
at any time without prior notice. It was initially
intended as a tool when developing new allocators, but
since it might be of interest for others it has been
briefly documented.</p>
<p>The recognized allocators are listed in
<seealso marker="erts:erts_alloc">erts_alloc(3)</seealso>.
After reading the <c>erts_alloc(3)</c> documentation,
the returned information
should more or less speak for itself. But it can be worth
explaining some things. Call counts are presented by two
values. The first value is giga calls, and the second
value is calls. <c>mbcs</c>, and <c>sbcs</c> are
abbreviations for, respectively, multi-block carriers, and
single-block carriers. Sizes are presented in bytes. When
it is not a size that is presented, it is the amount of
something. Sizes and amounts are often presented by three
values, the first is current value, the second is maximum
value since the last call to
<c>erlang:system_info({allocator, Alloc})</c>, and
the third is maximum value since the emulator was started.
If only one value is present, it is the current value.
<c>fix_alloc</c> memory block types are presented by two
values. The first value is memory pool size and
the second value used memory size.</p>
</item>
<tag><marker id="system_info_allocator_sizes"><c>{allocator_sizes, Alloc}</c></marker></tag>
<item>
<p>Returns various size information for the specified
allocator. The information returned is a subset of the
information returned by
<seealso marker="#system_info_allocator_tuple">erlang:system_info({allocator, Alloc})</seealso>.
</p>
</item>
<tag><c>build_type</c></tag>
<item>
<p>Returns an atom describing the build type of the runtime
system. This is normally the atom <c>opt</c> for optimized.
Other possible return values are <c>debug</c>, <c>purify</c>,
<c>quantify</c>, <c>purecov</c>, <c>gcov</c>, <c>valgrind</c>,
<c>gprof</c>, and <c>lcnt</c>. Possible return values
may be added and/or removed at any time without prior notice.
</p>
</item>
<tag><c>c_compiler_used</c></tag>
<item>
<p>Returns a two-tuple describing the C compiler used when
compiling the runtime system. The first element is an
atom describing the name of the compiler, or <c>undefined</c>
if unknown. The second element is a term describing the
version of the compiler, or <c>undefined</c> if unknown.
</p>
</item>
<tag><c>check_io</c></tag>
<item>
<p>Returns a list containing miscellaneous information
regarding the emulators internal I/O checking. Note,
the content of the returned list may vary between
platforms and over time. The only thing guaranteed is
that a list is returned.</p>
</item>
<tag><c>compat_rel</c></tag>
<item>
<p>Returns the compatibility mode of the local node as
an integer. The integer returned represents the
Erlang/OTP release which the current emulator has been
set to be backward compatible with. The compatibility
mode can be configured at startup by using the command
line flag <c>+R</c>, see
<seealso marker="erts:erl#compat_rel">erl(1)</seealso>.</p>
</item>
<tag><marker id="system_info_cpu_topology"><c>cpu_topology</c></marker></tag>
<item>
<p>Returns the <c>CpuTopology</c> which currently is used by the
emulator. The CPU topology is used when binding schedulers
to logical processors. The CPU topology used is the user defined
CPU topology if such exist; otherwise, the automatically
detected CPU topology if such exist. If no CPU topology
exist <c>undefined</c> is returned.</p>
<p>Types:</p>
<list type="bulleted">
<item><c>CpuTopology = LevelEntryList | undefined</c></item>
<item><c>LevelEntryList = [LevelEntry]</c> (all
<c>LevelEntry</c>s of a <c>LevelEntryList</c>
must contain the same <c>LevelTag</c>, except
on the top level where both <c>node</c> and
<c>processor</c> <c>LevelTag</c>s may co-exist)</item>
<item><c>LevelEntry = {LevelTag, SubLevel}
| {LevelTag, InfoList, SubLevel}</c>
(<c>{LevelTag, SubLevel}
== {LevelTag, [], SubLevel}</c>)</item>
<item><c>LevelTag = node|processor|core|thread</c>
(more <c>LevelTag</c>s may be introduced in
the future)</item>
<item><c>SubLevel = [LevelEntry] | LogicalCpuId</c></item>
<item><c>LogicalCpuId = {logical, integer()}</c></item>
<item><c>InfoList = []</c> (the <c>InfoList</c>
may be extended in the future)</item>
</list>
<p><c>node</c> refers to NUMA (non-uniform memory access)
nodes, and <c>thread</c> refers to hardware threads
(e.g. Intels hyper-threads).</p>
<p>A level in the <c>CpuTopology</c> term can be omitted if
only one entry exists and the <c>InfoList</c> is empty.
</p>
<p><c>thread</c> can only be a sub level to <c>core</c>.
<c>core</c> can be a sub level to either <c>processor</c>
or <c>node</c>. <c>processor</c> can either be on the
top level or a sub level to <c>node</c>. <c>node</c>
can either be on the top level or a sub level to
<c>processor</c>. That is, NUMA nodes can be processor
internal or processor external. A CPU topology can
consist of a mix of processor internal and external
NUMA nodes, as long as each logical CPU belongs to one
and only one NUMA node. Cache hierarchy is not part of
the <c>CpuTopology</c> type yet, but will be in the
future. Other things may also make it into the CPU
topology in the future. In other words, expect the
<c>CpuTopology</c> type to change.
</p>
</item>
<tag><marker id="system_info_cpu_topology_defined"><c>{cpu_topology, defined}</c></marker></tag>
<item>
<p>Returns the user defined <c>CpuTopology</c>. For more
information see the documentation of
<seealso marker="#system_flag_cpu_topology">erlang:system_flag(cpu_topology, CpuTopology)</seealso>
and the documentation of the
<seealso marker="#system_info_cpu_topology">cpu_topology</seealso>
argument.
</p>
</item>
<tag><marker id="system_info_cpu_topology_detected"><c>{cpu_topology, detected}</c></marker></tag>
<item>
<p>Returns the automatically detected <c>CpuTopology</c>. The
emulator currently only detects the CPU topology on some newer
Linux, Solaris, and Windows systems. On Windows system with
more than 32 logical processors the CPU topology is not detected.
</p>
<p>For more information see the documentation of the
<seealso marker="#system_info_cpu_topology">cpu_topology</seealso>
argument.
</p>
</item>
<tag><c>{cpu_topology, used}</c></tag>
<item>
<p>Returns the <c>CpuTopology</c> which is used by the
emulator. For more information see the
documentation of the
<seealso marker="#system_info_cpu_topology">cpu_topology</seealso>
argument.
</p>
</item>
<tag><c>creation</c></tag>
<item>
<p>Returns the creation of the local node as an integer.
The creation is changed when a node is restarted. The
creation of a node is stored in process identifiers, port
identifiers, and references. This makes it (to some
extent) possible to distinguish between identifiers from
different incarnations of a node. Currently valid
creations are integers in the range 1..3, but this may
(probably will) change in the future. If the node is not
alive, 0 is returned.</p>
</item>
<tag><c>debug_compiled</c></tag>
<item>
<p>Returns <c>true</c> if the emulator has been debug
compiled; otherwise, <c>false</c>.
</p>
</item>
<tag><c>dist</c></tag>
<item>
<p>Returns a binary containing a string of distribution
information formatted as in Erlang crash dumps. For more
information see the <seealso marker="erts:crash_dump">"How to interpret the Erlang crash dumps"</seealso>
chapter in the ERTS User's Guide.</p>
</item>
<tag><c>dist_ctrl</c></tag>
<item>
<p>Returns a list of tuples
<c>{Node, ControllingEntity}</c>, one entry for each
connected remote node. The <c>Node</c> is the name of the
node and the <c>ControllingEntity</c> is the port or pid
responsible for the communication to that node. More
specifically, the <c>ControllingEntity</c> for nodes
connected via TCP/IP (the normal case) is the socket
actually used in communication with the specific node.</p>
</item>
<tag><c>driver_version</c></tag>
<item>
<p>Returns a string containing the erlang driver version
used by the runtime system. It will be on the form
<seealso marker="erts:erl_driver#version_management">"<major ver>.<minor ver>"</seealso>.</p>
</item>
<tag><c>elib_malloc</c></tag>
<item>
<p>This option will be removed in a future release.
The return value will always be <c>false</c> since
the elib_malloc allocator has been removed.</p>
</item>
<tag><c>fullsweep_after</c></tag>
<item>
<p>Returns <c>{fullsweep_after, int()}</c> which is the
<c>fullsweep_after</c> garbage collection setting used
by default. For more information see
<c>garbage_collection</c> described below.</p>
</item>
<tag><c>garbage_collection</c></tag>
<item>
<p>Returns a list describing the default garbage collection
settings. A process spawned on the local node by a
<c>spawn</c> or <c>spawn_link</c> will use these
garbage collection settings. The default settings can be
changed by use of
<seealso marker="#system_flag/2">system_flag/2</seealso>.
<seealso marker="#spawn_opt/4">spawn_opt/4</seealso>
can spawn a process that does not use the default
settings.</p>
</item>
<tag><c>global_heaps_size</c></tag>
<item>
<p>Returns the current size of the shared (global) heap.</p>
</item>
<tag><c>heap_sizes</c></tag>
<item>
<p>Returns a list of integers representing valid heap sizes
in words. All Erlang heaps are sized from sizes in this
list.</p>
</item>
<tag><c>heap_type</c></tag>
<item>
<p>Returns the heap type used by the current emulator.
Currently the following heap types exist:</p>
<taglist>
<tag><c>private</c></tag>
<item>
<p>Each process has a heap reserved for its use and no
references between heaps of different processes are
allowed. Messages passed between processes are copied
between heaps.</p>
</item>
<tag><c>shared</c></tag>
<item>
<p>One heap for use by all processes. Messages passed
between processes are passed by reference.</p>
</item>
<tag><c>hybrid</c></tag>
<item>
<p>A hybrid of the <c>private</c> and <c>shared</c> heap
types. A shared heap as well as private heaps are
used.</p>
</item>
</taglist>
</item>
<tag><c>info</c></tag>
<item>
<p>Returns a binary containing a string of miscellaneous
system information formatted as in Erlang crash dumps.
For more information see the
<seealso marker="erts:crash_dump">"How to interpret the Erlang crash dumps"</seealso> chapter in the ERTS
User's Guide.</p>
</item>
<tag><c>kernel_poll</c></tag>
<item>
<p>Returns <c>true</c> if the emulator uses some kind of
kernel-poll implementation; otherwise, <c>false</c>.</p>
</item>
<tag><c>loaded</c></tag>
<item>
<p>Returns a binary containing a string of loaded module
information formatted as in Erlang crash dumps. For more
information see the <seealso marker="erts:crash_dump">"How to interpret the Erlang crash dumps"</seealso> chapter
in the ERTS User's Guide.</p>
</item>
<tag><marker id="logical_processors"><c>logical_processors</c></marker></tag>
<item>
<p>Returns the detected number of logical processors configured
on the system. The return value is either an integer, or
the atom <c>unknown</c> if the emulator wasn't able to
detect logical processors configured.
</p>
</item>
<tag><marker id="logical_processors_available"><c>logical_processors_available</c></marker></tag>
<item>
<p>Returns the detected number of logical processors available to
the Erlang runtime system. The return value is either an
integer, or the atom <c>unknown</c> if the emulator wasn't
able to detect logical processors available. The number
of logical processors available is less than or equal to
the number of <seealso marker="#logical_processors_online">logical
processors online</seealso>.
</p>
</item>
<tag><marker id="logical_processors_online"><c>logical_processors_online</c></marker></tag>
<item>
<p>Returns the detected number of logical processors online on
the system. The return value is either an integer,
or the atom <c>unknown</c> if the emulator wasn't able to
detect logical processors online. The number of logical
processors online is less than or equal to the number of
<seealso marker="#logical_processors">logical processors
configured</seealso>.
</p>
</item>
<tag><c>machine</c></tag>
<item>
<p>Returns a string containing the Erlang machine name.</p>
</item>
<tag><c>min_heap_size</c></tag>
<item>
<p>Returns <c>{min_heap_size, MinHeapSize}</c> where <c>MinHeapSize</c> is the current system wide
minimum heap size for spawned processes.</p>
</item>
<tag><c>min_bin_vheap_size</c></tag>
<item>
<p>Returns <c>{min_bin_vheap_size, MinBinVHeapSize}</c> where <c>MinBinVHeapSize</c> is the current system wide
minimum binary virtual heap size for spawned processes.</p>
</item>
<tag><c>modified_timing_level</c></tag>
<item>
<p>Returns the modified timing level (an integer) if
modified timing has been enabled; otherwise,
<c>undefined</c>. See the <c>+T</c> command line flag
in the documentation of the
<seealso marker="erts:erl#+T">erl(1)</seealso>
command for more information on modified timing.</p>
</item>
<tag><marker id="system_info_multi_scheduling"><c>multi_scheduling</c></marker></tag>
<item>
<p>Returns <c>disabled</c>, <c>blocked</c>, or <c>enabled</c>.
A description of the return values:</p>
<taglist>
<tag><c>disabled</c></tag>
<item>
<p>The emulator has only one scheduler thread. The
emulator does not have SMP support, or have been
started with only one scheduler thread.</p>
</item>
<tag><c>blocked</c></tag>
<item>
<p>The emulator has more than one scheduler thread,
but all scheduler threads but one have been blocked,
i.e., only one scheduler thread will schedule
Erlang processes and execute Erlang code.</p>
</item>
<tag><c>enabled</c></tag>
<item>
<p>The emulator has more than one scheduler thread,
and no scheduler threads have been blocked, i.e.,
all available scheduler threads will schedule
Erlang processes and execute Erlang code.</p>
</item>
</taglist>
<p>See also <seealso marker="#system_flag_multi_scheduling">erlang:system_flag(multi_scheduling, BlockState)</seealso>,
<seealso marker="#system_info_multi_scheduling_blockers">erlang:system_info(multi_scheduling_blockers)</seealso>, and
<seealso marker="#system_info_schedulers">erlang:system_info(schedulers)</seealso>.</p>
</item>
<tag><marker id="system_info_multi_scheduling_blockers"><c>multi_scheduling_blockers</c></marker></tag>
<item>
<p>Returns a list of <c>PID</c>s when multi-scheduling
is blocked; otherwise, the empty list. The <c>PID</c>s
in the list is <c>PID</c>s of the processes currently
blocking multi-scheduling. A <c>PID</c> will only be
present once in the list, even if the corresponding
process has blocked multiple times.</p>
<p>See also <seealso marker="#system_flag_multi_scheduling">erlang:system_flag(multi_scheduling, BlockState)</seealso>,
<seealso marker="#system_info_multi_scheduling">erlang:system_info(multi_scheduling)</seealso>, and
<seealso marker="#system_info_schedulers">erlang:system_info(schedulers)</seealso>.</p>
</item>
<tag><marker id="system_info_otp_release"><c>otp_release</c></marker></tag>
<item>
<p>Returns a string containing the OTP release number.</p>
</item>
<tag><c>process_count</c></tag>
<item>
<p>Returns the number of processes currently existing at
the local node as an integer. The same value as
<c>length(processes())</c> returns.</p>
</item>
<tag><c>process_limit</c></tag>
<item>
<p>Returns the maximum number of concurrently existing
processes at the local node as an integer. This limit
can be configured at startup by using the command line
flag <c>+P</c>, see
<seealso marker="erts:erl#max_processes">erl(1)</seealso>.</p>
</item>
<tag><c>procs</c></tag>
<item>
<p>Returns a binary containing a string of process and port
information formatted as in Erlang crash dumps. For more
information see the <seealso marker="erts:crash_dump">"How to interpret the Erlang crash dumps"</seealso> chapter
in the ERTS User's Guide.</p>
</item>
<tag><marker id="system_info_scheduler_bind_type"><c>scheduler_bind_type</c></marker></tag>
<item>
<p>Returns information on how user has requested
schedulers to be bound or not bound.</p>
<p><em>NOTE:</em> Even though user has requested
schedulers to be bound via
<seealso marker="#system_flag_scheduler_bind_type">erlang:system_flag(scheduler_bind_type, How)</seealso>,
they might have silently failed to bind. In order to
inspect actual scheduler bindings call
<seealso marker="#system_info_scheduler_bindings">erlang:system_info(scheduler_bindings)</seealso>.
</p>
<p>For more information, see
<seealso marker="#system_flag_scheduler_bind_type">erlang:system_flag(scheduler_bind_type, How)</seealso>, and
<seealso marker="#system_info_scheduler_bindings">erlang:system_info(scheduler_bindings)</seealso>.
</p>
</item>
<tag><marker id="system_info_scheduler_bindings"><c>scheduler_bindings</c></marker></tag>
<item>
<p>Returns information on currently used scheduler
bindings.</p>
<p>A tuple of a size equal to
<seealso marker="#system_info_schedulers">erlang:system_info(schedulers)</seealso> is returned. The elements of the tuple are integers
or the atom <c>unbound</c>. Logical processor identifiers
are represented as integers. The <c>N</c>th
element of the tuple equals the current binding for
the scheduler with the scheduler identifier equal to
<c>N</c>. E.g., if the schedulers have been bound,
<c>element(erlang:system_info(scheduler_id),
erlang:system_info(scheduler_bindings))</c> will return
the identifier of the logical processor that the calling
process is executing on.
</p>
<p>Note that only schedulers online can be bound to logical
processors.</p>
<p>For more information, see
<seealso marker="#system_flag_scheduler_bind_type">erlang:system_flag(scheduler_bind_type, How)</seealso>,
<seealso marker="#system_info_schedulers_online">erlang:system_info(schedulers_online)</seealso>.
</p>
</item>
<tag><marker id="system_info_scheduler_id"><c>scheduler_id</c></marker></tag>
<item>
<p>Returns the scheduler id (<c>SchedulerId</c>) of the
scheduler thread that the calling process is executing
on. <c>SchedulerId</c> is a positive integer; where
<c><![CDATA[1 <= SchedulerId <= erlang:system_info(schedulers)]]></c>. See also
<seealso marker="#system_info_schedulers">erlang:system_info(schedulers)</seealso>.</p>
</item>
<tag><marker id="system_info_schedulers"><c>schedulers</c></marker></tag>
<item>
<p>Returns the number of scheduler threads used by
the emulator. Scheduler threads online schedules Erlang
processes and Erlang ports, and execute Erlang code
and Erlang linked in driver code.</p>
<p>The number of scheduler threads is determined at
emulator boot time and cannot be changed after
that. The amount of schedulers online can
however be changed at any time.</p>
<p>See also <seealso marker="#system_flag_schedulers_online">erlang:system_flag(schedulers_online, SchedulersOnline)</seealso>,
<seealso marker="#system_info_schedulers_online">erlang:system_info(schedulers_online)</seealso>,
<seealso marker="#system_info_scheduler_id">erlang:system_info(scheduler_id)</seealso>,
<seealso marker="#system_flag_multi_scheduling">erlang:system_flag(multi_scheduling, BlockState)</seealso>,
<seealso marker="#system_info_multi_scheduling">erlang:system_info(multi_scheduling)</seealso>, and
and <seealso marker="#system_info_multi_scheduling_blockers">erlang:system_info(multi_scheduling_blockers)</seealso>.</p>
</item>
<tag><marker id="system_info_schedulers_online"><c>schedulers_online</c></marker></tag>
<item>
<p>Returns the amount of schedulers online. The scheduler
identifiers of schedulers online satisfy the following
relationship:
<c><![CDATA[1 <= SchedulerId <= erlang:system_info(schedulers_online)]]></c>.
</p>
<p>For more information, see
<seealso marker="#system_info_schedulers">erlang:system_info(schedulers)</seealso>,
and
<seealso marker="#system_flag_schedulers_online">erlang:system_flag(schedulers_online, SchedulersOnline)</seealso>.
</p>
</item>
<tag><c>smp_support</c></tag>
<item>
<p>Returns <c>true</c> if the emulator has been compiled
with smp support; otherwise, <c>false</c>.</p>
</item>
<tag><c>system_version</c></tag>
<item>
<p>Returns a string containing version number and
some important properties such as the number of schedulers.</p>
</item>
<tag><c>system_architecture</c></tag>
<item>
<p>Returns a string containing the processor and OS
architecture the emulator is built for.</p>
</item>
<tag><c>threads</c></tag>
<item>
<p>Returns <c>true</c> if the emulator has been compiled
with thread support; otherwise, <c>false</c> is
returned.</p>
</item>
<tag><marker id="system_info_thread_pool_size"><c>thread_pool_size</c></marker></tag>
<item>
<p>Returns the number of async threads in the async thread
pool used for asynchronous driver calls
(<seealso marker="erts:erl_driver#driver_async">driver_async()</seealso>)
as an integer.</p>
</item>
<tag><c>trace_control_word</c></tag>
<item>
<p>Returns the value of the node's trace control word.
For more information see documentation of the function
<c>get_tcw</c> in "Match Specifications in Erlang",
<seealso marker="erts:match_spec#get_tcw">ERTS User's Guide</seealso>.</p>
</item>
<tag><marker id="update_cpu_info"><c>update_cpu_info</c></marker></tag>
<item>
<p>The runtime system rereads the CPU information available and
updates its internally stored information about the
<seealso marker="#system_info_cpu_topology_detected">detected CPU
topology</seealso> and the amount of logical processors
<seealso marker="#logical_processors">configured</seealso>,
<seealso marker="#logical_processors_online">online</seealso>, and
<seealso marker="#logical_processors_available">available</seealso>.
If the CPU information has changed since the last time it was read,
the atom <c>changed</c> is returned; otherwise, the atom
<c>unchanged</c> is returned. If the CPU information has changed
you probably want to
<seealso marker="#system_flag_schedulers_online">adjust the amount
of schedulers online</seealso>. You typically want to have as
many schedulers online as
<seealso marker="#logical_processors_available">logical processors
available</seealso>.
</p>
</item>
<tag><marker id="system_info_version"><c>version</c></marker></tag>
<item>
<p>Returns a string containing the version number of the
emulator.</p>
</item>
<tag><c>wordsize</c></tag>
<item>
<p>Same as <c>{wordsize, internal}</c></p>
</item>
<tag><c>{wordsize, internal}</c></tag>
<item>
<p>Returns the size of Erlang term words in bytes as an
integer, i.e. on a 32-bit architecture 4 is returned,
and on a pure 64-bit architecture 8 is returned. On a
halfword 64-bit emulator, 4 is returned, as the Erlang
terms are stored using a virtual wordsize of half the
systems wordsize.</p>
</item>
<tag><c>{wordsize, external}</c></tag>
<item>
<p>Returns the true wordsize of the emulator, i.e. the size
of a pointer, in bytes as an integer. On a pure 32-bit
architecture 4 is returned, on both a halfword and pure
64-bit architecture, 8 is returned.</p>
</item>
</taglist>
<note>
<p>The <c>scheduler</c> argument has changed name to
<c>scheduler_id</c>. This in order to avoid mixup with
the <c>schedulers</c> argument. The <c>scheduler</c>
argument was introduced in ERTS version 5.5 and renamed
in ERTS version 5.5.1.</p>
</note>
</desc>
</func>
<func>
<name>erlang:system_monitor() -> MonSettings</name>
<fsummary>Current system performance monitoring settings</fsummary>
<type>
<v>MonSettings -> {MonitorPid, Options} | undefined</v>
<v> MonitorPid = pid()</v>
<v> Options = [Option]</v>
<v> Option = {long_gc, Time} | {large_heap, Size} | busy_port | busy_dist_port</v>
<v> Time = Size = int()</v>
</type>
<desc>
<p>Returns the current system monitoring settings set by
<seealso marker="#system_monitor/2">erlang:system_monitor/2</seealso>
as <c>{MonitorPid, Options}</c>, or <c>undefined</c> if there
are no settings. The order of the options may be different
from the one that was set.</p>
</desc>
</func>
<func>
<name>erlang:system_monitor(undefined | {MonitorPid, Options}) -> MonSettings</name>
<fsummary>Set or clear system performance monitoring options</fsummary>
<type>
<v>MonitorPid, Options, MonSettings -- see below</v>
</type>
<desc>
<p>When called with the argument <c>undefined</c>, all
system performance monitoring settings are cleared.</p>
<p>Calling the function with <c>{MonitorPid, Options}</c> as
argument, is the same as calling
<seealso marker="#system_monitor/2">erlang:system_monitor(MonitorPid, Options)</seealso>.</p>
<p>Returns the previous system monitor settings just like
<seealso marker="#system_monitor/0">erlang:system_monitor/0</seealso>.</p>
</desc>
</func>
<func>
<name>erlang:system_monitor(MonitorPid, [Option]) -> MonSettings</name>
<fsummary>Set system performance monitoring options</fsummary>
<type>
<v>MonitorPid = pid()</v>
<v>Option = {long_gc, Time} | {large_heap, Size} | busy_port | busy_dist_port</v>
<v> Time = Size = int()</v>
<v>MonSettings = {OldMonitorPid, [Option]}</v>
<v> OldMonitorPid = pid()</v>
</type>
<desc>
<p>Sets system performance monitoring options. <c>MonitorPid</c>
is a local pid that will receive system monitor messages, and
the second argument is a list of monitoring options:</p>
<taglist>
<tag><c>{long_gc, Time}</c></tag>
<item>
<p>If a garbage collection in the system takes at least
<c>Time</c> wallclock milliseconds, a message
<c>{monitor, GcPid, long_gc, Info}</c> is sent to
<c>MonitorPid</c>. <c>GcPid</c> is the pid that was
garbage collected and <c>Info</c> is a list of two-element
tuples describing the result of the garbage collection.
One of the tuples is <c>{timeout, GcTime}</c> where
<c>GcTime</c> is the actual time for the garbage
collection in milliseconds. The other tuples are
tagged with <c>heap_size</c>, <c>heap_block_size</c>,
<c>stack_size</c>, <c>mbuf_size</c>, <c>old_heap_size</c>,
and <c>old_heap_block_size</c>. These tuples are
explained in the documentation of the
<seealso marker="#gc_start">gc_start</seealso>
trace message (see
<seealso marker="#trace/3">erlang:trace/3</seealso>).
New tuples may be added, and the order of the tuples in
the <c>Info</c> list may be changed at any time without prior
notice.
</p>
</item>
<tag><c>{large_heap, Size}</c></tag>
<item>
<p>If a garbage collection in the system results in
the allocated size of a heap being at least <c>Size</c>
words, a message <c>{monitor, GcPid, large_heap, Info}</c>
is sent to <c>MonitorPid</c>. <c>GcPid</c> and <c>Info</c>
are the same as for <c>long_gc</c> above, except that
the tuple tagged with <c>timeout</c> is not present.
<em>Note</em>: As of erts version 5.6 the monitor message
is sent if the sum of the sizes of all memory blocks allocated
for all heap generations is equal to or larger than <c>Size</c>.
Previously the monitor message was sent if the memory block
allocated for the youngest generation was equal to or larger
than <c>Size</c>.
</p>
</item>
<tag><c>busy_port</c></tag>
<item>
<p>If a process in the system gets suspended because it
sends to a busy port, a message
<c>{monitor, SusPid, busy_port, Port}</c> is sent to
<c>MonitorPid</c>. <c>SusPid</c> is the pid that got
suspended when sending to <c>Port</c>.</p>
</item>
<tag><c>busy_dist_port</c></tag>
<item>
<p>If a process in the system gets suspended because it
sends to a process on a remote node whose inter-node
communication was handled by a busy port, a message
<c>{monitor, SusPid, busy_dist_port, Port}</c> is sent to
<c>MonitorPid</c>. <c>SusPid</c> is the pid that got
suspended when sending through the inter-node
communication port <c>Port</c>.</p>
</item>
</taglist>
<p>Returns the previous system monitor settings just like
<seealso marker="#system_monitor/0">erlang:system_monitor/0</seealso>.</p>
<note>
<p>If a monitoring process gets so large that it itself
starts to cause system monitor messages when garbage
collecting, the messages will enlarge the process's
message queue and probably make the problem worse.</p>
<p>Keep the monitoring process neat and do not set the system
monitor limits too tight.</p>
</note>
<p>Failure: <c>badarg</c> if <c>MonitorPid</c> does not exist.</p>
</desc>
</func>
<func>
<name>erlang:system_profile() -> ProfilerSettings</name>
<fsummary>Current system profiling settings</fsummary>
<type>
<v>ProfilerSettings -> {ProfilerPid, Options} | undefined</v>
<v> ProfilerPid = pid() | port()</v>
<v> Options = [Option]</v>
<v> Option = runnable_procs | runnable_ports | scheduler | exclusive</v>
</type>
<desc>
<p>Returns the current system profiling settings set by
<seealso marker="#system_profile/2">erlang:system_profile/2</seealso>
as <c>{ProfilerPid, Options}</c>, or <c>undefined</c> if there
are no settings. The order of the options may be different
from the one that was set.</p>
</desc>
</func>
<func>
<name>erlang:system_profile(ProfilerPid, Options) -> ProfilerSettings</name>
<fsummary>Current system profiling settings</fsummary>
<type>
<v>ProfilerSettings -> {ProfilerPid, Options} | undefined</v>
<v> ProfilerPid = pid() | port()</v>
<v> Options = [Option]</v>
<v> Option = runnable_procs | runnable_ports | scheduler | exclusive</v>
</type>
<desc>
<p>Sets system profiler options. <c>ProfilerPid</c>
is a local pid or port that will receive profiling messages. The
receiver is excluded from all profiling.
The second argument is a list of profiling options:</p>
<taglist>
<tag><c>runnable_procs</c></tag>
<item>
<p>If a process is put into or removed from the run queue a message,
<c>{profile, Pid, State, Mfa, Ts}</c>, is sent to
<c>ProfilerPid</c>. Running processes that is reinserted into the
run queue after having been preemptively scheduled out will not trigger this
message.
</p>
</item>
<tag><c>runnable_ports</c></tag>
<item>
<p>If a port is put into or removed from the run queue a message,
<c>{profile, Port, State, 0, Ts}</c>, is sent to
<c>ProfilerPid</c>.
</p>
</item>
<tag><c>scheduler</c></tag>
<item>
<p>If a scheduler is put to sleep or awoken a message,
<c>{profile, scheduler, Id, State, NoScheds, Ts}</c>, is sent
to <c>ProfilerPid</c>.
</p>
</item>
<tag><c>exclusive</c></tag>
<item>
<p>
If a synchronous call to a port from a process is done, the
calling process is considered not runnable during the call
runtime to the port. The calling process is notified as
<c>inactive</c> and subsequently <c>active</c> when the port
callback returns.
</p>
</item>
</taglist>
<note><p><c>erlang:system_profile</c> is considered experimental and
its behaviour may change in the future.</p>
</note>
</desc>
</func>
<func>
<name>term_to_binary(Term) -> ext_binary()</name>
<fsummary>Encode a term to an Erlang external term format binary</fsummary>
<type>
<v>Term = term()</v>
</type>
<desc>
<p>Returns a binary data object which is the result of encoding
<c>Term</c> according to the Erlang external term format.</p>
<p>This can be used for a variety of purposes, for example
writing a term to a file in an efficient way, or sending an
Erlang term to some type of communications channel not
supported by distributed Erlang.</p>
<p>See also
<seealso marker="#binary_to_term/1">binary_to_term/1</seealso>.</p>
</desc>
</func>
<func>
<name>term_to_binary(Term, [Option]) -> ext_binary()</name>
<fsummary>Encode a term to en Erlang external term format binary</fsummary>
<type>
<v>Term = term()</v>
<v>Option = compressed | {compressed,Level} | {minor_version,Version}</v>
</type>
<desc>
<p>Returns a binary data object which is the result of encoding
<c>Term</c> according to the Erlang external term format.</p>
<p>If the option <c>compressed</c> is provided, the external
term format will be compressed. The compressed format is
automatically recognized by <c>binary_to_term/1</c> in R7B and later.</p>
<p>It is also possible to specify a compression level by giving
the option <c>{compressed,Level}</c>, where <c>Level</c> is an
integer from 0 through 9. <c>0</c> means that no compression
will be done (it is the same as not giving any <c>compressed</c> option);
<c>1</c> will take the least time but may not compress as well as
the higher levels; <c>9</c> will take the most time and may produce
a smaller result. Note the "mays" in the preceding sentence; depending
on the input term, level 9 compression may or may not produce a smaller
result than level 1 compression.</p>
<p>Currently, <c>compressed</c> gives the same result as
<c>{compressed,6}</c>.</p>
<p>The option <c>{minor_version,Version}</c> can be use to control
some details of the encoding. This option was
introduced in R11B-4. Currently, the allowed values for <c>Version</c>
are <c>0</c> and <c>1</c>.</p>
<p><c>{minor_version,1}</c> forces any floats in the term to be encoded
in a more space-efficient and exact way (namely in the 64-bit IEEE format,
rather than converted to a textual representation). <c>binary_to_term/1</c>
in R11B-4 and later is able decode the new representation.</p>
<p><c>{minor_version,0}</c> is currently the default, meaning that floats
will be encoded using a textual representation; this option is useful if
you want to ensure that releases prior to R11B-4 can decode resulting
binary.</p>
<p>See also
<seealso marker="#binary_to_term/1">binary_to_term/1</seealso>.</p>
</desc>
</func>
<func>
<name>throw(Any)</name>
<fsummary>Throw an exception</fsummary>
<type>
<v>Any = term()</v>
</type>
<desc>
<p>A non-local return from a function. If evaluated within a
<c>catch</c>, <c>catch</c> will return the value <c>Any</c>.</p>
<pre>
> <input>catch throw({hello, there}).</input>
{hello,there}</pre>
<p>Failure: <c>nocatch</c> if not evaluated within a catch.</p>
</desc>
</func>
<func>
<name>time() -> {Hour, Minute, Second}</name>
<fsummary>Current time</fsummary>
<type>
<v>Hour = Minute = Second = int()</v>
</type>
<desc>
<p>Returns the current time as <c>{Hour, Minute, Second}</c>.</p>
<p>The time zone and daylight saving time correction depend on
the underlying OS.</p>
<pre>
> <input>time().</input>
{9,42,44}</pre>
</desc>
</func>
<func>
<name>tl(List1) -> List2</name>
<fsummary>Tail of a list</fsummary>
<type>
<v>List1 = List2 = [term()]</v>
</type>
<desc>
<p>Returns the tail of <c>List1</c>, that is, the list minus
the first element.</p>
<pre>
> <input>tl([geesties, guilies, beasties]).</input>
[guilies, beasties]</pre>
<p>Allowed in guard tests.</p>
<p>Failure: <c>badarg</c> if <c>List</c> is the empty list [].</p>
</desc>
</func>
<func>
<name>erlang:trace(PidSpec, How, FlagList) -> int()</name>
<fsummary>Set trace flags for a process or processes</fsummary>
<type>
<v>PidSpec = pid() | existing | new | all</v>
<v>How = bool()</v>
<v>FlagList = [Flag]</v>
<v> Flag -- see below</v>
</type>
<desc>
<p>Turns on (if <c>How == true</c>) or off (if
<c>How == false</c>) the trace flags in <c>FlagList</c> for
the process or processes represented by <c>PidSpec</c>.</p>
<p><c>PidSpec</c> is either a pid for a local process, or one of
the following atoms:</p>
<taglist>
<tag><c>existing</c></tag>
<item>
<p>All processes currently existing.</p>
</item>
<tag><c>new</c></tag>
<item>
<p>All processes that will be created in the future.</p>
</item>
<tag><c>all</c></tag>
<item>
<p>All currently existing processes and all processes that
will be created in the future.</p>
</item>
</taglist>
<p><c>FlagList</c> can contain any number of the following
flags (the "message tags" refers to the list of messages
following below):</p>
<taglist>
<tag><c>all</c></tag>
<item>
<p>Set all trace flags except <c>{tracer, Tracer}</c> and
<c>cpu_timestamp</c> that are in their nature different
than the others.</p>
</item>
<tag><c>send</c></tag>
<item>
<p>Trace sending of messages.</p>
<p>Message tags: <c>send</c>,
<c>send_to_non_existing_process</c>.</p>
</item>
<tag><c>'receive'</c></tag>
<item>
<p>Trace receiving of messages.</p>
<p>Message tags: <c>'receive'</c>.</p>
</item>
<tag><c>procs</c></tag>
<item>
<p>Trace process related events.</p>
<p>Message tags: <c>spawn</c>, <c>exit</c>,
<c>register</c>, <c>unregister</c>, <c>link</c>,
<c>unlink</c>, <c>getting_linked</c>,
<c>getting_unlinked</c>.</p>
</item>
<tag><c>call</c></tag>
<item>
<p>Trace certain function calls. Specify which function
calls to trace by calling
<seealso marker="#trace_pattern/3">erlang:trace_pattern/3</seealso>.</p>
<p>Message tags: <c>call</c>, <c>return_from</c>.</p>
</item>
<tag><c>silent</c></tag>
<item>
<p>Used in conjunction with the <c>call</c> trace flag.
The <c>call</c>, <c>return_from</c> and <c>return_to</c>
trace messages are inhibited if this flag is set,
but if there are match specs they are executed as normal.</p>
<p>Silent mode is inhibited by executing
<c>erlang:trace(_, false, [silent|_])</c>,
or by a match spec executing the <c>{silent, false}</c>
function.</p>
<p>The <c>silent</c> trace flag facilitates setting up
a trace on many or even all processes in the system.
Then the interesting trace can be activated and
deactivated using the <c>{silent,Bool}</c>
match spec function, giving a high degree
of control of which functions with which
arguments that triggers the trace.</p>
<p>Message tags: <c>call</c>, <c>return_from</c>,
<c>return_to</c>. Or rather, the absence of.</p>
</item>
<tag><c>return_to</c></tag>
<item>
<p>Used in conjunction with the <c>call</c> trace flag.
Trace the actual return from a traced function back to
its caller. Only works for functions traced with
the <c>local</c> option to
<seealso marker="#trace_pattern/3">erlang:trace_pattern/3</seealso>.</p>
<p>The semantics is that a trace message is sent when a
call traced function actually returns, that is, when a
chain of tail recursive calls is ended. There will be
only one trace message sent per chain of tail recursive
calls, why the properties of tail recursiveness for
function calls are kept while tracing with this flag.
Using <c>call</c> and <c>return_to</c> trace together
makes it possible to know exactly in which function a
process executes at any time.</p>
<p>To get trace messages containing return values from
functions, use the <c>{return_trace}</c> match_spec
action instead.</p>
<p>Message tags: <c>return_to</c>.</p>
</item>
<tag><c>running</c></tag>
<item>
<p>Trace scheduling of processes.</p>
<p>Message tags: <c>in</c>, and <c>out</c>.</p>
</item>
<tag><c>exiting</c></tag>
<item>
<p>Trace scheduling of an exiting processes.</p>
<p>Message tags: <c>in_exiting</c>, <c>out_exiting</c>, and
<c>out_exited</c>.</p>
</item>
<tag><c>garbage_collection</c></tag>
<item>
<p>Trace garbage collections of processes.</p>
<p>Message tags: <c>gc_start</c>, <c>gc_end</c>.</p>
</item>
<tag><c>timestamp</c></tag>
<item>
<p>Include a time stamp in all trace messages. The time
stamp (Ts) is of the same form as returned by
<c>erlang:now()</c>.</p>
</item>
<tag><c>cpu_timestamp</c></tag>
<item>
<p>A global trace flag for the Erlang node that makes all
trace timestamps be in CPU time, not wallclock. It is
only allowed with <c>PidSpec==all</c>. If the host
machine operating system does not support high resolution
CPU time measurements, <c>trace/3</c> exits with
<c>badarg</c>.</p>
</item>
<tag><c>arity</c></tag>
<item>
<p>Used in conjunction with the <c>call</c> trace flag.
<c>{M, F, Arity}</c> will be specified instead of
<c>{M, F, Args}</c> in call trace messages.</p>
</item>
<tag><c>set_on_spawn</c></tag>
<item>
<p>Makes any process created by a traced process inherit
its trace flags, including the <c>set_on_spawn</c> flag.</p>
</item>
<tag><c>set_on_first_spawn</c></tag>
<item>
<p>Makes the first process created by a traced process
inherit its trace flags, excluding
the <c>set_on_first_spawn</c> flag.</p>
</item>
<tag><c>set_on_link</c></tag>
<item>
<p>Makes any process linked by a traced process inherit its
trace flags, including the <c>set_on_link</c> flag.</p>
</item>
<tag><c>set_on_first_link</c></tag>
<item>
<p>Makes the first process linked to by a traced process
inherit its trace flags, excluding
the <c>set_on_first_link</c> flag.</p>
</item>
<tag><c>{tracer, Tracer}</c></tag>
<item>
<p>Specify where to send the trace messages. <c>Tracer</c>
must be the pid of a local process or the port identifier
of a local port. If this flag is not given, trace
messages will be sent to the process that called
<c>erlang:trace/3</c>.</p>
</item>
</taglist>
<p>The effect of combining <c>set_on_first_link</c> with
<c>set_on_link</c> is the same as having
<c>set_on_first_link</c> alone. Likewise for
<c>set_on_spawn</c> and <c>set_on_first_spawn</c>.</p>
<p>If the <c>timestamp</c> flag is not given, the tracing
process will receive the trace messages described below.
<c>Pid</c> is the pid of the traced process in which
the traced event has occurred. The third element of the tuple
is the message tag.</p>
<p>If the <c>timestamp</c> flag is given, the first element of
the tuple will be <c>trace_ts</c> instead and the timestamp
is added last in the tuple.</p>
<taglist>
<tag><c>{trace, Pid, 'receive', Msg}</c></tag>
<item>
<p>When <c>Pid</c> receives the message <c>Msg</c>.</p>
</item>
<tag><c>{trace, Pid, send, Msg, To}</c></tag>
<item>
<p>When <c>Pid</c> sends the message <c>Msg</c> to
the process <c>To</c>.</p>
</item>
<tag><c>{trace, Pid, send_to_non_existing_process, Msg, To}</c></tag>
<item>
<p>When <c>Pid</c> sends the message <c>Msg</c> to
the non-existing process <c>To</c>.</p>
</item>
<tag><c>{trace, Pid, call, {M, F, Args}}</c></tag>
<item>
<p>When <c>Pid</c> calls a traced function. The return
values of calls are never supplied, only the call and its
arguments.</p>
<p>Note that the trace flag <c>arity</c> can be used to
change the contents of this message, so that <c>Arity</c>
is specified instead of <c>Args</c>.</p>
</item>
<tag><c>{trace, Pid, return_to, {M, F, Arity}}</c></tag>
<item>
<p>When <c>Pid</c> returns <em>to</em> the specified
function. This trace message is sent if both
the <c>call</c> and the <c>return_to</c> flags are set,
and the function is set to be traced on <em>local</em>
function calls. The message is only sent when returning
from a chain of tail recursive function calls where at
least one call generated a <c>call</c> trace message
(that is, the functions match specification matched and
<c>{message, false}</c> was not an action).</p>
</item>
<tag><c>{trace, Pid, return_from, {M, F, Arity}, ReturnValue}</c></tag>
<item>
<p>When <c>Pid</c> returns <em>from</em> the specified
function. This trace message is sent if the <c>call</c>
flag is set, and the function has a match specification
with a <c>return_trace</c> or <c>exception_trace</c> action.</p>
</item>
<tag><c>{trace, Pid, exception_from, {M, F, Arity}, {Class, Value}}</c></tag>
<item>
<p>When <c>Pid</c> exits <em>from</em> the specified
function due to an exception. This trace message is sent
if the <c>call</c> flag is set, and the function has
a match specification with an <c>exception_trace</c> action.</p>
</item>
<tag><c>{trace, Pid, spawn, Pid2, {M, F, Args}}</c></tag>
<item>
<p>When <c>Pid</c> spawns a new process <c>Pid2</c> with
the specified function call as entry point.</p>
<p>Note that <c>Args</c> is supposed to be the argument
list, but may be any term in the case of an erroneous
spawn.</p>
</item>
<tag><c>{trace, Pid, exit, Reason}</c></tag>
<item>
<p>When <c>Pid</c> exits with reason <c>Reason</c>.</p>
</item>
<tag><c>{trace, Pid, link, Pid2}</c></tag>
<item>
<p>When <c>Pid</c> links to a process <c>Pid2</c>.</p>
</item>
<tag><c>{trace, Pid, unlink, Pid2}</c></tag>
<item>
<p>When <c>Pid</c> removes the link from a process
<c>Pid2</c>.</p>
</item>
<tag><c>{trace, Pid, getting_linked, Pid2}</c></tag>
<item>
<p>When <c>Pid</c> gets linked to a process <c>Pid2</c>.</p>
</item>
<tag><c>{trace, Pid, getting_unlinked, Pid2}</c></tag>
<item>
<p>When <c>Pid</c> gets unlinked from a process <c>Pid2</c>.</p>
</item>
<tag><c>{trace, Pid, register, RegName}</c></tag>
<item>
<p>When <c>Pid</c> gets the name <c>RegName</c> registered.</p>
</item>
<tag><c>{trace, Pid, unregister, RegName}</c></tag>
<item>
<p>When <c>Pid</c> gets the name <c>RegName</c> unregistered.
Note that this is done automatically when a registered
process exits.</p>
</item>
<tag><c>{trace, Pid, in, {M, F, Arity} | 0}</c></tag>
<item>
<p>When <c>Pid</c> is scheduled to run. The process will
run in function <c>{M, F, Arity}</c>. On some rare
occasions the current function cannot be determined, then
the last element <c>Arity</c> is 0.</p>
</item>
<tag><c>{trace, Pid, out, {M, F, Arity} | 0}</c></tag>
<item>
<p>When <c>Pid</c> is scheduled out. The process was
running in function {M, F, Arity}. On some rare occasions
the current function cannot be determined, then the last
element <c>Arity</c> is 0.</p>
</item>
<tag><marker id="gc_start"><c>{trace, Pid, gc_start, Info}</c></marker></tag>
<item>
<p>Sent when garbage collection is about to be started.
<c>Info</c> is a list of two-element tuples, where
the first element is a key, and the second is the value.
You should not depend on the tuples have any defined
order. Currently, the following keys are defined:</p>
<taglist>
<tag><c>heap_size</c></tag>
<item>The size of the used part of the heap.</item>
<tag><c>heap_block_size</c></tag>
<item>The size of the memory block used for storing
the heap and the stack.</item>
<tag><c>old_heap_size</c></tag>
<item>The size of the used part of the old heap.</item>
<tag><c>old_heap_block_size</c></tag>
<item>The size of the memory block used for storing
the old heap.</item>
<tag><c>stack_size</c></tag>
<item>The actual size of the stack.</item>
<tag><c>recent_size</c></tag>
<item>The size of the data that survived the previous garbage
collection.</item>
<tag><c>mbuf_size</c></tag>
<item>The combined size of message buffers associated with
the process.</item>
<tag><c>bin_vheap_size</c></tag>
<item>The total size of unique off-heap binaries referenced from the process heap.</item>
<tag><c>bin_vheap_block_size</c></tag>
<item>The total size of binaries, in words, allowed in the virtual
heap in the process before doing a garbage collection. </item>
<tag><c>bin_old_vheap_size</c></tag>
<item>The total size of unique off-heap binaries referenced from the process old heap.</item>
<tag><c>bin_vheap_block_size</c></tag>
<item>The total size of binaries, in words, allowed in the virtual
old heap in the process before doing a garbage collection. </item>
</taglist>
<p>All sizes are in words.</p>
</item>
<tag><c>{trace, Pid, gc_end, Info}</c></tag>
<item>
<p>Sent when garbage collection is finished. <c>Info</c>
contains the same kind of list as in the <c>gc_start</c>
message, but the sizes reflect the new sizes after
garbage collection.</p>
</item>
</taglist>
<p>If the tracing process dies, the flags will be silently
removed.</p>
<p>Only one process can trace a particular process. For this
reason, attempts to trace an already traced process will fail.</p>
<p>Returns: A number indicating the number of processes that
matched <c>PidSpec</c>. If <c>PidSpec</c> is a pid,
the return value will be <c>1</c>. If <c>PidSpec</c> is
<c>all</c> or <c>existing</c> the return value will be
the number of processes running, excluding tracer processes.
If <c>PidSpec</c> is <c>new</c>, the return value will be
<c>0</c>.</p>
<p>Failure: If specified arguments are not supported. For
example <c>cpu_timestamp</c> is not supported on all
platforms.</p>
</desc>
</func>
<func>
<name>erlang:trace_delivered(Tracee) -> Ref</name>
<fsummary>Notification when trace has been delivered</fsummary>
<type>
<v>Tracee = pid() | all</v>
<v>Ref = reference()</v>
</type>
<desc>
<p>The delivery of trace messages is dislocated on the time-line
compared to other events in the system. If you know that the
<c>Tracee</c> has passed some specific point in its execution,
and you want to know when at least all trace messages
corresponding to events up to this point have reached the tracer
you can use <c>erlang:trace_delivered(Tracee)</c>. A
<c>{trace_delivered, Tracee, Ref}</c> message is sent to
the caller of <c>erlang:trace_delivered(Tracee)</c> when it
is guaranteed that all trace messages have been delivered to
the tracer up to the point that the <c>Tracee</c> had reached
at the time of the call to
<c>erlang:trace_delivered(Tracee)</c>.</p>
<p>Note that the <c>trace_delivered</c> message does <em>not</em>
imply that trace messages have been delivered; instead, it implies
that all trace messages that <em>should</em> be delivered have
been delivered. It is not an error if <c>Tracee</c> isn't, and
hasn't been traced by someone, but if this is the case,
<em>no</em> trace messages will have been delivered when the
<c>trace_delivered</c> message arrives.</p>
<p>Note that <c>Tracee</c> has to refer to a process currently,
or previously existing on the same node as the caller of
<c>erlang:trace_delivered(Tracee)</c> resides on.
The special <c>Tracee</c> atom <c>all</c> denotes all processes
that currently are traced in the node.</p>
<p>An example: Process <c>A</c> is tracee, port <c>B</c> is
tracer, and process <c>C</c> is the port owner of <c>B</c>.
<c>C</c> wants to close <c>B</c> when <c>A</c> exits. <c>C</c>
can ensure that the trace isn't truncated by calling
<c>erlang:trace_delivered(A)</c> when <c>A</c> exits and wait
for the <c>{trace_delivered, A, Ref}</c> message before closing
<c>B</c>.</p>
<p>Failure: <c>badarg</c> if <c>Tracee</c> does not refer to a
process (dead or alive) on the same node as the caller of
<c>erlang:trace_delivered(Tracee)</c> resides on.</p>
</desc>
</func>
<func>
<name>erlang:trace_info(PidOrFunc, Item) -> Res</name>
<fsummary>Trace information about a process or function</fsummary>
<type>
<v>PidOrFunc = pid() | new | {Module, Function, Arity} | on_load</v>
<v> Module = Function = atom()</v>
<v> Arity = int()</v>
<v>Item, Res -- see below</v>
</type>
<desc>
<p>Returns trace information about a process or function.</p>
<p>To get information about a process, <c>PidOrFunc</c> should
be a pid or the atom <c>new</c>. The atom <c>new</c> means
that the default trace state for processes to be created will
be returned. <c>Item</c> must have one of the following
values:</p>
<taglist>
<tag><c>flags</c></tag>
<item>
<p>Return a list of atoms indicating what kind of traces is
enabled for the process. The list will be empty if no
traces are enabled, and one or more of the followings
atoms if traces are enabled: <c>send</c>,
<c>'receive'</c>, <c>set_on_spawn</c>, <c>call</c>,
<c>return_to</c>, <c>procs</c>, <c>set_on_first_spawn</c>,
<c>set_on_link</c>, <c>running</c>,
<c>garbage_collection</c>, <c>timestamp</c>, and
<c>arity</c>. The order is arbitrary.</p>
</item>
<tag><c>tracer</c></tag>
<item>
<p>Return the identifier for process or port tracing this
process. If this process is not being traced, the return
value will be <c>[]</c>.</p>
</item>
</taglist>
<p>To get information about a function, <c>PidOrFunc</c> should
be a three-element tuple: <c>{Module, Function, Arity}</c> or
the atom <c>on_load</c>. No wildcards are allowed. Returns
<c>undefined</c> if the function does not exist or
<c>false</c> if the function is not traced at all. <c>Item</c>
must have one of the following values:</p>
<taglist>
<tag><c>traced</c></tag>
<item>
<p>Return <c>global</c> if this function is traced on
global function calls, <c>local</c> if this function is
traced on local function calls (i.e local and global
function calls), and <c>false</c> if neither local nor
global function calls are traced.</p>
</item>
<tag><c>match_spec</c></tag>
<item>
<p>Return the match specification for this function, if it
has one. If the function is locally or globally traced but
has no match specification defined, the returned value
is <c>[]</c>.</p>
</item>
<tag><c>meta</c></tag>
<item>
<p>Return the meta trace tracer process or port for this
function, if it has one. If the function is not meta
traced the returned value is <c>false</c>, and if
the function is meta traced but has once detected that
the tracer proc is invalid, the returned value is [].</p>
</item>
<tag><c>meta_match_spec</c></tag>
<item>
<p>Return the meta trace match specification for this
function, if it has one. If the function is meta traced
but has no match specification defined, the returned
value is <c>[]</c>.</p>
</item>
<tag><c>call_count</c></tag>
<item>
<p>Return the call count value for this function or
<c>true</c> for the pseudo function <c>on_load</c> if call
count tracing is active. Return <c>false</c> otherwise.
See also
<seealso marker="#trace_pattern/3">erlang:trace_pattern/3</seealso>.</p>
</item>
<tag><c>call_time</c></tag>
<item>
<p>Return the call time values for this function or
<c>true</c> for the pseudo function <c>on_load</c> if call
time tracing is active. Returns <c>false</c> otherwise.
The call time values returned, <c>[{Pid, Count, S, Us}]</c>,
is a list of each process that has executed the function and its specific counters.
See also
<seealso marker="#trace_pattern/3">erlang:trace_pattern/3</seealso>.</p>
</item>
<tag><c>all</c></tag>
<item>
<p>Return a list containing the <c>{Item, Value}</c> tuples
for all other items, or return <c>false</c> if no tracing
is active for this function.</p>
</item>
</taglist>
<p>The actual return value will be <c>{Item, Value}</c>, where
<c>Value</c> is the requested information as described above.
If a pid for a dead process was given, or the name of a
non-existing function, <c>Value</c> will be <c>undefined</c>.</p>
<p>If <c>PidOrFunc</c> is the <c>on_load</c>, the information
returned refers to the default value for code that will be
loaded.</p>
</desc>
</func>
<func>
<name>erlang:trace_pattern(MFA, MatchSpec) -> int()</name>
<fsummary>Set trace patterns for global call tracing</fsummary>
<desc>
<p>The same as
<seealso marker="#trace_pattern/3">erlang:trace_pattern(MFA, MatchSpec, [])</seealso>,
retained for backward compatibility.</p>
</desc>
</func>
<func>
<name>erlang:trace_pattern(MFA, MatchSpec, FlagList) -> int()</name>
<fsummary>Set trace patterns for tracing of function calls</fsummary>
<type>
<v>MFA, MatchSpec, FlagList -- see below</v>
</type>
<desc>
<p>This BIF is used to enable or disable call tracing for
exported functions. It must be combined with
<seealso marker="#trace/3">erlang:trace/3</seealso>
to set the <c>call</c> trace flag for one or more processes.</p>
<p>Conceptually, call tracing works like this: Inside
the Erlang virtual machine there is a set of processes to be
traced and a set of functions to be traced. Tracing will be
enabled on the intersection of the set. That is, if a process
included in the traced process set calls a function included
in the traced function set, the trace action will be taken.
Otherwise, nothing will happen.</p>
<p>Use
<seealso marker="#trace/3">erlang:trace/3</seealso> to
add or remove one or more processes to the set of traced
processes. Use <c>erlang:trace_pattern/2</c> to add or remove
exported functions to the set of traced functions.</p>
<p>The <c>erlang:trace_pattern/3</c> BIF can also add match
specifications to an exported function. A match specification
comprises a pattern that the arguments to the function must
match, a guard expression which must evaluate to <c>true</c>
and an action to be performed. The default action is to send a
trace message. If the pattern does not match or the guard
fails, the action will not be executed.</p>
<p>The <c>MFA</c> argument should be a tuple like
<c>{Module, Function, Arity}</c> or the atom <c>on_load</c>
(described below). It can be the module, function, and arity
for an exported function (or a BIF in any module).
The <c>'_'</c> atom can be used to mean any of that kind.
Wildcards can be used in any of the following ways:</p>
<taglist>
<tag><c>{Module,Function,'_'}</c></tag>
<item>
<p>All exported functions of any arity named <c>Function</c>
in module <c>Module</c>.</p>
</item>
<tag><c>{Module,'_','_'}</c></tag>
<item>
<p>All exported functions in module <c>Module</c>.</p>
</item>
<tag><c>{'_','_','_'}</c></tag>
<item>
<p>All exported functions in all loaded modules.</p>
</item>
</taglist>
<p>Other combinations, such as <c>{Module,'_',Arity}</c>, are
not allowed. Local functions will match wildcards only if
the <c>local</c> option is in the <c>FlagList</c>.</p>
<p>If the <c>MFA</c> argument is the atom <c>on_load</c>,
the match specification and flag list will be used on all
modules that are newly loaded.</p>
<p>The <c>MatchSpec</c> argument can take any of the following
forms:</p>
<taglist>
<tag><c>false</c></tag>
<item>
<p>Disable tracing for the matching function(s). Any match
specification will be removed.</p>
</item>
<tag><c>true</c></tag>
<item>
<p>Enable tracing for the matching function(s).</p>
</item>
<tag><c>MatchSpecList</c></tag>
<item>
<p>A list of match specifications. An empty list is
equivalent to <c>true</c>. See the ERTS User's Guide
for a description of match specifications.</p>
</item>
<tag><c>restart</c></tag>
<item>
<p>For the <c>FlagList</c> option <c>call_count</c> and <c>call_time</c>:
restart the existing counters. The behaviour is undefined
for other <c>FlagList</c> options.</p>
</item>
<tag><c>pause</c></tag>
<item>
<p>For the <c>FlagList</c> option <c>call_count</c> and <c>call_time</c>: pause
the existing counters. The behaviour is undefined for
other <c>FlagList</c> options.</p>
</item>
</taglist>
<p>The <c>FlagList</c> parameter is a list of options.
The following options are allowed:</p>
<taglist>
<tag><c>global</c></tag>
<item>
<p>Turn on or off call tracing for global function calls
(that is, calls specifying the module explicitly). Only
exported functions will match and only global calls will
generate trace messages. This is the default.</p>
</item>
<tag><c>local</c></tag>
<item>
<p>Turn on or off call tracing for all types of function
calls. Trace messages will be sent whenever any of
the specified functions are called, regardless of how they
are called. If the <c>return_to</c> flag is set for
the process, a <c>return_to</c> message will also be sent
when this function returns to its caller.</p>
</item>
<tag><c>meta | {meta, Pid}</c></tag>
<item>
<p>Turn on or off meta tracing for all types of function
calls. Trace messages will be sent to the tracer process
or port <c>Pid</c> whenever any of the specified
functions are called, regardless of how they are called.
If no <c>Pid</c> is specified, <c>self()</c> is used as a
default tracer process.</p>
<p>Meta tracing traces all processes and does not care
about the process trace flags set by <c>trace/3</c>,
the trace flags are instead fixed to
<c>[call, timestamp]</c>.</p>
<p>The match spec function <c>{return_trace}</c> works with
meta trace and send its trace message to the same tracer
process.</p>
</item>
<tag><c>call_count</c></tag>
<item>
<p>Starts (<c>MatchSpec == true</c>) or stops
(<c>MatchSpec == false</c>) call count tracing for all
types of function calls. For every function a counter is
incremented when the function is called, in any process.
No process trace flags need to be activated.</p>
<p>If call count tracing is started while already running,
the count is restarted from zero. Running counters can be
paused with <c>MatchSpec == pause</c>. Paused and running
counters can be restarted from zero with
<c>MatchSpec == restart</c>.</p>
<p>The counter value can be read with
<seealso marker="#trace_info/2">erlang:trace_info/2</seealso>.</p>
</item>
<tag><c>call_time</c></tag>
<item>
<p>Starts (<c>MatchSpec == true</c>) or stops
(<c>MatchSpec == false</c>) call time tracing for all
types of function calls. For every function a counter is
incremented when the function is called. Time spent in the function
is accumulated in two other counters, seconds and micro-seconds.
The counters are stored for each call traced process.</p>
<p>If call time tracing is started while already running,
the count and time is restarted from zero. Running counters can be
paused with <c>MatchSpec == pause</c>. Paused and running
counters can be restarted from zero with
<c>MatchSpec == restart</c>.</p>
<p>The counter value can be read with
<seealso marker="#trace_info/2">erlang:trace_info/2</seealso>.</p>
</item>
</taglist>
<p>The <c>global</c> and <c>local</c> options are mutually
exclusive and <c>global</c> is the default (if no options are
specified). The <c>call_count</c> and <c>meta</c> options
perform a kind of local tracing, and can also not be combined
with <c>global</c>. A function can be either globally or
locally traced. If global tracing is specified for a
specified set of functions; local, meta, call time and call count
tracing for the matching set of local functions will be
disabled, and vice versa.</p>
<p>When disabling trace, the option must match the type of trace
that is set on the function, so that local tracing must be
disabled with the <c>local</c> option and global tracing with
the <c>global</c> option (or no option at all), and so forth.</p>
<p>There is no way to directly change part of a match
specification list. If a function has a match specification,
you can replace it with a completely new one. If you need to
change an existing match specification, use the
<seealso marker="#trace_info/2">erlang:trace_info/2</seealso>
BIF to retrieve the existing match specification.</p>
<p>Returns the number of exported functions that matched
the <c>MFA</c> argument. This will be zero if none matched at
all.</p>
</desc>
</func>
<func>
<name>trunc(Number) -> int()</name>
<fsummary>Return an integer by the truncating a number</fsummary>
<type>
<v>Number = number()</v>
</type>
<desc>
<p>Returns an integer by the truncating <c>Number</c>.</p>
<pre>
> <input>trunc(5.5).</input>
5</pre>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>tuple_size(Tuple) -> int()</name>
<fsummary>Return the size of a tuple</fsummary>
<type>
<v>Tuple = tuple()</v>
</type>
<desc>
<p>Returns an integer which is the number of elements in <c>Tuple</c>.</p>
<pre>
> <input>tuple_size({morni, mulle, bwange}).</input>
3</pre>
<p>Allowed in guard tests.</p>
</desc>
</func>
<func>
<name>tuple_to_list(Tuple) -> [term()]</name>
<fsummary>Convert a tuple to a list</fsummary>
<type>
<v>Tuple = tuple()</v>
</type>
<desc>
<p>Returns a list which corresponds to <c>Tuple</c>.
<c>Tuple</c> may contain any Erlang terms.</p>
<pre>
> <input>tuple_to_list({share, {'Ericsson_B', 163}}).</input>
[share,{'Ericsson_B',163}]</pre>
</desc>
</func>
<func>
<name>erlang:universaltime() -> {Date, Time}</name>
<fsummary>Current date and time according to Universal Time Coordinated (UTC)</fsummary>
<type>
<v>Date = {Year, Month, Day}</v>
<v>Time = {Hour, Minute, Second}</v>
<v> Year = Month = Day = Hour = Minute = Second = int()</v>
</type>
<desc>
<p>Returns the current date and time according to Universal
Time Coordinated (UTC), also called GMT, in the form
<c>{{Year, Month, Day}, {Hour, Minute, Second}}</c> if
supported by the underlying operating system. If not,
<c>erlang:universaltime()</c> is equivalent to
<c>erlang:localtime()</c>.</p>
<pre>
> <input>erlang:universaltime().</input>
{{1996,11,6},{14,18,43}}</pre>
</desc>
</func>
<func>
<name>erlang:universaltime_to_localtime({Date1, Time1}) -> {Date2, Time2}</name>
<fsummary>Convert from Universal Time Coordinated (UTC) to local date and time</fsummary>
<type>
<v>Date1 = Date2 = {Year, Month, Day}</v>
<v>Time1 = Time2 = {Hour, Minute, Second}</v>
<v> Year = Month = Day = Hour = Minute = Second = int()</v>
</type>
<desc>
<p>Converts Universal Time Coordinated (UTC) date and time to
local date and time, if this is supported by the underlying
OS. Otherwise, no conversion is done, and
<c>{Date1, Time1}</c> is returned.</p>
<pre>
> <input>erlang:universaltime_to_localtime({{1996,11,6},{14,18,43}}).</input>
{{1996,11,7},{15,18,43}}</pre>
<p>Failure: <c>badarg</c> if <c>Date1</c> or <c>Time1</c> do
not denote a valid date or time.</p>
</desc>
</func>
<func>
<name>unlink(Id) -> true</name>
<fsummary>Remove a link, if there is one, to another process or port</fsummary>
<type>
<v>Id = pid() | port()</v>
</type>
<desc>
<p>Removes the link, if there is one, between the calling
process and the process or port referred to by <c>Id</c>.</p>
<p>Returns <c>true</c> and does not fail, even if there is no
link to <c>Id</c>, or if <c>Id</c> does not exist.</p>
<p>Once <c>unlink(Id)</c> has returned it is guaranteed that
the link between the caller and the entity referred to by
<c>Id</c> has no effect on the caller in the future (unless
the link is setup again). If caller is trapping exits, an
<c>{'EXIT', Id, _}</c> message due to the link might have
been placed in the callers message queue prior to the call,
though. Note, the <c>{'EXIT', Id, _}</c> message can be the
result of the link, but can also be the result of <c>Id</c>
calling <c>exit/2</c>. Therefore, it <em>may</em> be
appropriate to cleanup the message queue when trapping exits
after the call to <c>unlink(Id)</c>, as follow:</p>
<code type="none">
unlink(Id),
receive
{'EXIT', Id, _} ->
true
after 0 ->
true
end</code>
<note>
<p>Prior to OTP release R11B (erts version 5.5) <c>unlink/1</c>
behaved completely asynchronous, i.e., the link was active
until the "unlink signal" reached the linked entity. This
had one undesirable effect, though. You could never know when
you were guaranteed <em>not</em> to be effected by the link.</p>
<p>Current behavior can be viewed as two combined operations:
asynchronously send an "unlink signal" to the linked entity
and ignore any future results of the link.</p>
</note>
</desc>
</func>
<func>
<name>unregister(RegName) -> true</name>
<fsummary>Remove the registered name for a process (or port)</fsummary>
<type>
<v>RegName = atom()</v>
</type>
<desc>
<p>Removes the registered name <c>RegName</c>, associated with a
pid or a port identifier.</p>
<pre>
> <input>unregister(db).</input>
true</pre>
<p>Users are advised not to unregister system processes.</p>
<p>Failure: <c>badarg</c> if <c>RegName</c> is not a registered
name.</p>
</desc>
</func>
<func>
<name>whereis(RegName) -> pid() | port() | undefined</name>
<fsummary>Get the pid (or port) with a given registered name</fsummary>
<desc>
<p>Returns the pid or port identifier with the registered name
<c>RegName</c>. Returns <c>undefined</c> if the name is not
registered.</p>
<pre>
> <input>whereis(db).</input>
<0.43.0></pre>
</desc>
</func>
<func>
<name>erlang:yield() -> true</name>
<fsummary>Let other processes get a chance to execute</fsummary>
<desc>
<p>Voluntarily let other processes (if any) get a chance to
execute. Using <c>erlang:yield()</c> is similar to
<c>receive after 1 -> ok end</c>, except that <c>yield()</c>
is faster.</p>
<warning><p>There is seldom or never any need to use this BIF,
especially in the SMP-emulator as other processes will have a
chance to run in another scheduler thread anyway.
Using this BIF without a thorough grasp of how the scheduler
works may cause performance degradation.</p></warning>
</desc>
</func>
</funcs>
</erlref>
