path: root/lib/compiler/doc/src/notes.xml

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

<chapter>
  <header>
    <copyright>
      <year>2004</year><year>2018</year>
      <holder>Ericsson AB. All Rights Reserved.</holder>
    </copyright>
    <legalnotice>
      Licensed under the Apache License, Version 2.0 (the "License");
      you may not use this file except in compliance with the License.
      You may obtain a copy of the License at
 
          http://www.apache.org/licenses/LICENSE-2.0

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

    </legalnotice>

    <title>Compiler Release Notes</title>
    <prepared>otp_appnotes</prepared>
    <docno>nil</docno>
    <date>nil</date>
    <rev>nil</rev>
    <file>notes.xml</file>
  </header>
  <p>This document describes the changes made to the Compiler
    application.</p>

<section><title>Compiler 7.4</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p><c>record_info/2</c> is a pseudo-function that
	    requires literal arguments known at compile time.
	    Therefore, the following usage is illegal: <c>fun
	    record/info/2</c>. The compiler would crash when during
	    compilation of that kind of code. Corrected to issue a
	    compilation error.</p>
          <p>
	    Own Id: OTP-15760 Aux Id: ERL-907 </p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    The compiler has been rewritten to internally use an
	    intermediate representation based on Static Single
	    Assignment (SSA). The new intermediate representation
	    makes more optimizations possible.</p>
          <p>
	    Most noticeable is that the binary matching optimizations
	    are now applicable in many more circumstances than
	    before.</p>
          <p>
	    Another noticeable change is that type optimizations are
	    now applied across local function calls, and will remove
	    a lot more redundant type tests than before.</p>
          <p>
	    Own Id: OTP-14894 Aux Id: ERL-714 </p>
        </item>
        <item>
	    <p>Funs are no longer created when they are only used
	    locally, greatly improving the performance of named funs
	    and "fun-wrapped" macros.</p>
          <p>
	    Own Id: OTP-15273 Aux Id: ERL-639 </p>
        </item>
        <item>
	    <p>All compiler options that can be given in the source
	    file can now also be given in the option list or from the
	    command line for <c>erlc</c>.</p>
	    <p>Specifically, the option
	    <c>{nowarn_deprecated_function,MFAs}</c> was only
	    recognized when given in the file with the attribute
	    <c>-compile()</c>. The option
	    <c>{nowarn_unused_function,FAs}</c> was incorrectly
	    documented to only work in a file, but it also worked
	    when given in the option list.</p>
          <p>
	    Own Id: OTP-15456</p>
        </item>
        <item>
	    <p> Do not allow function specifications for functions
	    residing in other modules. </p>
          <p>
	    *** POTENTIAL INCOMPATIBILITY ***</p>
          <p>
	    Own Id: OTP-15563 Aux Id: ERL-845, OTP-15562 </p>
        </item>
        <item>
	    <p>Internal documentation has now been added to the
	    <em>Erts</em> and <em>Compiler</em> applications.</p>
	    <p>The internal documents for <em>Erts</em> describe
	    miscellaneous interesting implementation details. Those
	    details can change at any time.</p>
	    <p>The internal documentation for <em>Compiler</em>
	    documents the API for the Core Erlang modules. While we
	    will not change those APIs without good reason, we don't
	    give the same guarantees about backward compatibility as
	    for the rest of the APIs in OTP.</p>
          <p>
	    Own Id: OTP-15715</p>
        </item>
        <item>
	    <p> There are new compiler options <c>nowarn_removed</c>
	    and <c>{nowarn_removed,Items}</c> to suppress warnings
	    for functions and modules that have been removed from
	    OTP.</p>
          <p>
	    Own Id: OTP-15749 Aux Id: ERL-904 </p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.3.2</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>An expression such as <c>(A / B) band 16#ff</c> would
	    crash the compiler.</p>
          <p>
	    Own Id: OTP-15518 Aux Id: ERL-829 </p>
        </item>
        <item>
	    <p>There could be an incorrect warning when the
	    <c>tuple_calls</c> option was given. The generated code
	    would be correct. Here is an example of code that would
	    trigger the warning:</p>
	    <p><c>(list_to_atom("prefix_" ++
	    atom_to_list(suffix))):doit(X)</c>.</p>
          <p>
	    Own Id: OTP-15552 Aux Id: ERL-838 </p>
        </item>
        <item>
	    <p>Optimize (again) Dialyzer's handling of
	    left-associative use of <c>andalso</c> and <c>orelse</c>
	    in guards.</p>
          <p>
	    Own Id: OTP-15577 Aux Id: ERL-851, PR-2141, PR-1944 </p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.3.1</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>An optimization that avoided allocation of a stack
	    frame for some <c>case</c> expressions was introduced in
	    OTP 21. (ERL-504/OTP-14808) It turns out that in rare
	    circumstances, this optimization is not safe. Therefore,
	    this optimization has been disabled.</p>
	    <p>A similar optimization will be included in OTP 22 in a
	    safe way.</p>
          <p>
	    Own Id: OTP-15501 Aux Id: ERL-807, ERL-514, OTP-14808 </p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.3</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>Fixed a rare internal consistency failure caused by a
	    bug in the <c>beam_jump</c> pass. (Thanks to Simon
	    Cornish for reporting this bug.)</p>
          <p>
	    Own Id: OTP-15400 Aux Id: ERL-759 </p>
        </item>
        <item>
	    <p>The compiler could fail with an internal consistency
	    check failure when compiling code that used the
	    <c>is_function/2</c> BIF.</p>
          <p>
	    Own Id: OTP-15435 Aux Id: ERL-778 </p>
        </item>
        <item>
	    <p>When an external fun was used, warnings for unused
	    variables could be suppressed.</p>
          <p>
	    Own Id: OTP-15437 Aux Id: ERL-762 </p>
        </item>
        <item>
	    <p>The compiler would crash when compiling an
	    <c>after</c> block that called <c>erlang:raise/3</c> like
	    this: <c>erlang:raise(Class, Stacktrace,
	    Stacktrace)</c></p>
          <p>
	    Own Id: OTP-15481</p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
	    <p>When specified, the <c>+{source,Name}</c> option will
	    now override the actual file name in stack traces,
	    instead of only affecting the return value of
	    <c>Mod:module_info()</c>.</p>
	    <p>The <c>+deterministic</c> flag will also affect stack
	    traces now, omitting all path information except the file
	    name, fixing a long-standing issue where deterministic
	    builds required deterministic paths.</p>
          <p>
	    Own Id: OTP-15245 Aux Id: ERL-706 </p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.2.7</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>Fixed a bug where incorrect code was generated
	    following a binary match guard.</p>
          <p>
	    Own Id: OTP-15353 Aux Id: ERL-753 </p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.2.6</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>In rare circumstances, the matched out tail of a
	    binary could be the entire original binary. (There was
	    partial correction to this problem in version 7.2.5 of
	    the compiler application.)</p>
          <p>
	    Own Id: OTP-15335 Aux Id: ERL-689, OTP-15219 </p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.2.5</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>Fixed a bug that prevented certain variable-sized
	    binary comprehensions from compiling.</p>
          <p>
	    Own Id: OTP-15186 Aux Id: ERL-665 </p>
        </item>
        <item>
	    <p>When compiling from Core Erlang, funs created in
	    certain expressions that were only used for their
	    side-effects were subtly broken.</p>
          <p>
	    Own Id: OTP-15188 Aux Id: ERL-658 </p>
        </item>
        <item>
	    <p>There could be an internal consistency failure when a
	    <c>receive</c> was nested in a
	    <c>try</c>/<c>catch</c>.</p>
          <p>
	    Own Id: OTP-15218 Aux Id: ERL-684 </p>
        </item>
        <item>
	    <p>In rare circumstances, the matched out tail of a
	    binary could be the entire original binary.</p>
          <p>
	    Own Id: OTP-15219 Aux Id: ERL-689 </p>
        </item>
        <item>
	    <p>When <c>is_map_key/2</c> was used in a guard together
	    with the <c>not/1</c> or <c>or/2</c> operators, the error
	    behavior could be wrong when <c>is_map_key/2</c> was
	    passed a non-map as the second argument. </p>
	    <p>In rare circumstances, compiling code that uses
	    <c>is_map_key/2</c> could cause an internal consistency
	    check failure.</p>
          <p>
	    Own Id: OTP-15227 Aux Id: ERL-699 </p>
        </item>
        <item>
	    <p>The compiler could crash when compiling a function
	    with multiple receives in multiple clauses.</p>
          <p>
	    Own Id: OTP-15235 Aux Id: ERL-703 </p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.2.4</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>Fix a regression in OTP-15204 that removed
	    <c>.beam</c> file metadata that some external build tools
	    relied on.</p>
          <p>
	    Own Id: OTP-15292</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.2.3</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>Fixed an issue where files compiled with the
	    <c>+deterministic</c> option differed if they were
	    compiled in a different directory but were otherwise
	    identical.</p>
          <p>
	    Own Id: OTP-15204 Aux Id: ERL-679 </p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.2.2</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>In rare cases involving matching of binary literal
	    strings, the compiler could optimize away code that
	    should be executed.</p>
          <p>
	    Own Id: OTP-15156 Aux Id: ERL-655 </p>
        </item>
        <item>
	    <p>There could be an internal consistency check failure
	    when compiling code that called <c>map_get(Key, Map)</c>
	    and then updated the same map.</p>
          <p>
	    Own Id: OTP-15157</p>
        </item>
        <item>
	    <p>In rare circumstances, the compiler could crash in
	    <c>beam_jump</c> when compiling a floating point
	    operation.</p>
          <p>
	    Own Id: OTP-15166 Aux Id: ERL-660 </p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.2.1</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>The could could crash when compiling a complicated
	    function that used the binary syntax.</p>
          <p>
	    Own Id: OTP-15150 Aux Id: ERL-650 </p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.2</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>Fixed an error in an optimization pass that caused
	    impossible tuple matching.</p>
          <p>
	    Own Id: OTP-14855 Aux Id: ERL-549 </p>
        </item>
        <item>
	    <p>The exception thrown when a list comprehension was
	    given a non-list term was not always correct.</p>
          <p>
	    Own Id: OTP-14992 Aux Id: ERL-572 </p>
        </item>
        <item>
	    <p>The compiler could produce incorrect code in rare
	    circumstances when the <c>[{inline,F/A}]</c> option was
	    used.</p>
          <p>
	    Own Id: OTP-15115 Aux Id: PR-1831 </p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
	    <p>Changed the default behaviour of <c>.erlang</c>
	    loading: <c>.erlang</c> is no longer loaded from the
	    current directory. <c>c:erlangrc(PathList)</c> can be
	    used to search and load an <c>.erlang</c> file from user
	    specified directories.</p> <p><c>escript</c>,
	    <c>erlc</c>, <c>dialyzer</c> and <c>typer</c> no longer
	    load an <c>.erlang</c> at all.</p>
          <p>
	    *** POTENTIAL INCOMPATIBILITY ***</p>
          <p>
	    Own Id: OTP-14439</p>
        </item>
        <item>
	    <p>Support for "tuple calls" have been removed from the
	    run-time system. Tuple calls was an undocumented and
	    unsupported feature which allowed the module argument for
	    an apply operation to be a tuple: <c>Var = dict:new(),
	    Var:size()</c>. This "feature" frequently caused
	    confusion, especially when such call failed. The
	    stacktrace would point out functions that don't exist in
	    the source code.</p>
	    <p>For legacy code that need to use parameterized modules
	    or tuple calls for some other reason, there is a new
	    compiler option called <c>tuple_calls</c>. When this
	    option is given, the compiler will generate extra code
	    that emulates the old behavior for calls where the module
	    is a variable.</p>
          <p>
	    *** POTENTIAL INCOMPATIBILITY ***</p>
          <p>
	    Own Id: OTP-14497</p>
        </item>
        <item>
	    <p>In code such as <c>example({ok, Val}) -&gt; {ok,
	    Val}.</c> a tuple would be built. The compiler will now
	    automatically rewrite the code to
	    <c>example({ok,Val}=Tuple) -&gt; Tuple.</c> which will
	    reduce code size, execution time, and remove GC
	    pressure.</p>
          <p>
	    Own Id: OTP-14505</p>
        </item>
        <item>
	    <p>The optimization of <c>case</c> expression where only
	    one of the case arms can execute successfully has been
	    improved.</p>
          <p>
	    Own Id: OTP-14525</p>
        </item>
        <item>
	    <p>Some uses of binary matching has been slightly
	    improved, eliminating unnecessary register shuffling.</p>
          <p>
	    Own Id: OTP-14594 Aux Id: ERL-444 </p>
        </item>
        <item>
	    <p>There is a new <c>{compile_info,Info}</c> option for
	    the compiler that allows BEAM-based languages such as
	    Elixir and LFE to add their own compiler versions.</p>
          <p>
	    Own Id: OTP-14615 Aux Id: PR-1558 </p>
        </item>
        <item>
	    <p>Loaded BEAM code in a 64-bit system requires less
	    memory because of better packing of operands for
	    instructions.</p>
	    <p>These memory savings were achieved by major
	    improvements to the <c>beam_makeops</c> scripts used when
	    building the run time system and BEAM compiler. There is
	    also new for documentation for <c>beam_makeops</c> that
	    describes how new BEAM instructions and loader
	    transformations can be implemented. The documentation is
	    found in here in a source directory or git repository:
	    erts/emulator/internal_doc/beam_makeops.md. An online
	    version can be found here:
	    https://github.com/erlang/otp/blob/master/erts/emulator/internal_doc/beam_makeops.md</p>
          <p>
	    Own Id: OTP-14626</p>
        </item>
        <item>
	    <p>Size calculations for binary constructions has been
	    somewhat optimized, producing smaller code.</p>
          <p>
	    Own Id: OTP-14654</p>
        </item>
        <item>
	    <p>When the value returned from a '<c>catch</c>'
	    expression is ignored, no stacktrace will be built if an
	    exception is caught. That will save time and produce less
	    garbage. There are also some minor optimizations of
	    '<c>try</c>/<c>catch</c>' both in the compiler and
	    run-time system.</p>
          <p>
	    Own Id: OTP-14683</p>
        </item>
        <item>
	    <p>There is a new syntax in '<c>try/catch</c>' for
	    retrieving the stacktrace without calling
	    '<c>erlang:get_stacktrace/0</c>'. See the reference
	    manual for a description of the new syntax. The
	    '<c>erlang:get_stacktrace/0</c>' BIF is now
	    deprecated.</p>
          <p>
	    Own Id: OTP-14692</p>
        </item>
        <item>
	    <p>The following is an internal change in the compiler,
	    that is not noticeable for normal use of the compiler:
	    The module <c>v3_life</c> has been removed. Its
	    functionality has been simplified and integrated into
	    <c>v3_codegen</c>.</p>
          <p>
	    Own Id: OTP-14712</p>
        </item>
        <item>
	    <p>The optimization of binary matching that delays
	    creation of sub binaries (see the Efficiency Guide) could
	    be thwarted by the argument order and could be necessary
	    to change the argument order. The compiler has now become
	    smarter and can handle any argument order.</p>
          <p>
	    Own Id: OTP-14774</p>
        </item>
        <item>
	    <p>When the compiler was faced with complex case
	    expressions it would unnecessarily allocate stack
	    elements and shuffle data between x and y registers.
	    Improved code generation to only allocate a stack frame
	    when strictly necessary.</p>
          <p>
	    Own Id: OTP-14808 Aux Id: ERL-514 </p>
        </item>
        <item>
	    <p>There is a new option '<c>makedep_side_effect</c>' for
	    the compiler and <c>-MMD</c> for '<c>erlc</c>' that
	    generates dependencies and continues to compile as
	    normal.</p>
          <p>
	    Own Id: OTP-14830</p>
        </item>
        <item>
	    <p>When compiling modules with huge functions, the
	    compiler would generate a lot of atoms for its internal,
	    sometimes so many that the atom table would overflow. The
	    compiler has been rewritten to generate far less internal
	    atoms to avoid filling the atom table.</p>
          <p>
	    Own Id: OTP-14968 Aux Id: ERL-563 </p>
        </item>
        <item>
	    <p>External funs with literal values for module, name,
	    and arity (e.g. <c>erlang:abs/1</c>) are now treated as
	    literals. That means more efficient code that produces
	    less garbage on the heap.</p>
          <p>
	    Own Id: OTP-15003</p>
        </item>
        <item>
	    <p>Two new guards BIFs operating on maps have been added:
	    <c>map_get/2</c> and <c>is_map_key/2</c>. They do the
	    same as <c>maps:get/2</c> and <c>maps:is_key/2</c>,
	    respectively, except that they are allowed to be used in
	    guards.</p>
          <p>
	    Own Id: OTP-15037 Aux Id: PR-1784, PR-1802 </p>
        </item>
        <item>
	    <p>A call or apply of a literal external fun will be
	    replaced with a direct call.</p>
          <p>
	    Own Id: OTP-15044 Aux Id: ERL-614 </p>
        </item>
        <item>
	    <p>Part of EEP-44 has been implemented.</p>
	    <p>There is a new predefined macro called
	    <c>OTP_RELEASE</c> which is an integer indicating the OTP
	    release number (its value is <c>21</c> in this
	    release).</p>
	    <p>There are new preprocessor directives
	    <c>-if(Condition).</c> and <c>-elif(Condition).</c>. The
	    <c>if/elif</c> supports the builtin function
	    <c>defined(Symbol)</c>.</p>
          <p>
	    Own Id: OTP-15087 Aux Id: PR-1810 </p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.1.5.2</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>Fix a regression in OTP-15204 that removed
	    <c>.beam</c> file metadata that some external build tools
	    relied on.</p>
          <p>
	    Own Id: OTP-15292</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.1.5.1</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>Fixed an issue where files compiled with the
	    <c>+deterministic</c> option differed if they were
	    compiled in a different directory but were otherwise
	    identical.</p>
          <p>
	    Own Id: OTP-15204 Aux Id: ERL-679 </p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.1.5</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>The internal compiler pass (<c>beam_validator</c>)
	    that validates the generated code has been
	    strengthened.</p>
	    <p>When compiling from BEAM assembly code, the
	    <c>beam_type</c> optimizer pass could make the code
	    unsafe. Corrected.</p>
          <p>
	    Own Id: OTP-14863</p>
        </item>
        <item>
	    <p>Corrected optimizations of integers matched out from
	    binaries and used in bit operations.</p>
          <p>
	    Own Id: OTP-14898</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.1.4</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>The '<c>deterministic</c>' option was not recognized
	    when given in a <c>-compile()</c> attribute in the source
	    code.</p>
          <p>
	    Own Id: OTP-14773 Aux Id: ERL-498 </p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.1.3</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>The compiler could issue an incorrect internal
	    consistency failure diagnostic for some complicated bit
	    syntax maches.</p>
          <p>
	    Own Id: OTP-14640 Aux Id: ERL-490 </p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.1.2</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>Fail labels on guard BIFs weren't taken into account
	    during an optimization pass, and a bug in the validation
	    pass sometimes prevented this from being noticed when a
	    fault occurred.</p>
          <p>
	    Own Id: OTP-14522 Aux Id: ERIERL-48 </p>
        </item>
        <item>
          <p>
	    When compiling from Core Erlang, an 'apply' with a nested
	    apply in the function position would be treated as an
	    invalid call. Corrected. (Thanks to Mikael Pettersson for
	    reporting this bug.)</p>
          <p>
	    Own Id: OTP-14526</p>
        </item>
        <item>
	    <p>Fixed checking of binary matching in the
	    <c>beam_validator</c> module to ensure that potential
	    compiler bugs are found at compile-time instead as
	    emulator crash at run-time.</p>
          <p>
	    Own Id: OTP-14591</p>
        </item>
        <item>
	    <p>There could be false warnings for
	    <c>erlang:get_stacktrace/0</c> being used outside of a
	    <c>try</c> block when using multiple <c>catch</c>
	    clauses.</p>
          <p>
	    Own Id: OTP-14600 Aux Id: ERL-478 </p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
	    <p> The Erlang code linter no longer checks that the
	    functions mentioned in <c>nowarn_deprecated_function</c>
	    options are declared in the module. </p>
          <p>
	    Own Id: OTP-14378</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.1.1</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>Fail labels on guard BIFs weren't taken into account
	    during an optimization pass, and a bug in the validation
	    pass sometimes prevented this from being noticed when a
	    fault occurred.</p>
          <p>
	    Own Id: OTP-14522 Aux Id: ERIERL-48 </p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.1</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>For many releases, it has been legal to override a BIF
	    with a local function having the same name. However,
	    calling a local function with the same name as guard BIF
	    as filter in a list comprehension was not allowed.</p>
          <p>
	    Own Id: OTP-13690</p>
        </item>
        <item>
	    <p>compile:forms/2 would not return the module name as
	    documented when one of the options '<c>from_core</c>',
	    '<c>from_asm</c>', or '<c>from_beam</c>' was given. Also,
	    the compiler would crash if one of those options was
	    combined with '<c>native</c>'.</p>
          <p>
	    Own Id: OTP-14408 Aux Id: ERL-417 </p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    Optimized test for tuples with an atom as first element.</p>
          <p>
	    Own Id: OTP-12148</p>
        </item>
        <item>
          <p>
	    Compilation of modules with huge literal binary strings
	    is now much faster.</p>
          <p>
	    Own Id: OTP-13794</p>
        </item>
        <item>
	    <p>Replaced usage of deprecated symbolic <seealso
	    marker="erts:erlang#type-time_unit"><c>time
	    unit</c></seealso> representations.</p>
          <p>
	    Own Id: OTP-13831 Aux Id: OTP-13735 </p>
        </item>
        <item>
	    <p>The undocumented and unsupported module
	    <c>sys_pre_expand</c> has been removed. As a partial
	    replacement for the functionality, there is a new
	    function <c>erl_internal:add_predefined_functions/1</c>
	    and <c>erl_expand_records</c> will now add a module
	    prefix to calls to BIFs and imported functions.</p>
          <p>
	    Own Id: OTP-13856</p>
        </item>
        <item>
	    <p>The internal compiler passes now start all generated
	    variables with "@" to avoid any conflicts with variables
	    in languages such as Elixir or LFE.</p>
          <p>
	    Own Id: OTP-13924</p>
        </item>
        <item>
	    <p>The function <c>fmod/2</c> has been added to the
	    <c>math</c> module.</p>
          <p>
	    Own Id: OTP-14000</p>
        </item>
        <item>
	    <p>Code generation for complicated guards have been
	    improved.</p>
          <p>
	    Own Id: OTP-14042</p>
        </item>
        <item>
          <p>
	    The compiler has new warnings for repeated identical map
	    keys.</p>
          <p>
	    A map expression such as,</p>
          <p>
	    <c> #{'a' => 1, 'b' => 2, 'a' => 3}.</c></p>
          <p>
	    will produce a warning for the repeated key 'a'.</p>
          <p>
	    Own Id: OTP-14058</p>
        </item>
        <item>
	    <p>By default, there will now be a warning when
	    <c>export_all</c> is used. The warning can be disabled
	    using <c>nowarn_export_all</c>.</p>
          <p>
	    Own Id: OTP-14071</p>
        </item>
        <item>
          <p>
	    Optimize maps pattern matching by only examining the
	    common keys in each clause first instead of all keys.
	    This will reduce the number of lookups of each key in
	    maps pattern matching.</p>
          <p>
	    Own Id: OTP-14072</p>
        </item>
        <item>
	    <p>There is a new '<c>deterministic</c>' option to omit
	    '<c>source</c>' and '<c>options</c>' tuples in the BEAM
	    file.</p>
          <p>
	    Own Id: OTP-14087</p>
        </item>
        <item>
          <p>
	    Analyzing modules with binary construction with huge
	    strings is now much faster. The compiler also compiles
	    such modules slightly faster.</p>
          <p>
	    Own Id: OTP-14125 Aux Id: ERL-308 </p>
        </item>
        <item>
	    <p>Atoms may now contain arbitrary Unicode
	    characters.</p>
          <p>
	    Own Id: OTP-14178</p>
        </item>
        <item>
	    <p><c>compile:file/2</c> now accepts the option
	    <c>extra_chunks</c> to include extra chunks in the BEAM
	    file.</p>
          <p>
	    Own Id: OTP-14221</p>
        </item>
        <item>
	    <p>The format of debug information that is stored in BEAM
	    files (when <c>debug_info</c> is used) has been changed.
	    The purpose of the change is to better support other
	    BEAM-based languages such as Elixir or LFE.</p>
	    <p>All tools included in OTP (dialyzer, debugger, cover,
	    and so on) will handle both the new format and the
	    previous format. Tools that retrieve the debug
	    information using <c>beam_lib:chunk(Beam,
	    [abstract_code])</c> will continue to work with both the
	    new and old format. Tools that call
	    <c>beam_lib:chunk(Beam, ["Abst"])</c> will not work with
	    the new format.</p>
	    <p>For more information, see the description of
	    <c>debug_info</c> in the documentation for
	    <c>beam_lib</c> and the description of the
	    <c>{debug_info,{Backend,Data}}</c> option in the
	    documentation for <c>compile</c>.</p>
          <p>
	    Own Id: OTP-14369 Aux Id: PR-1367 </p>
        </item>
        <item>
	    <p>In a future release, <c>erlang:get_stacktrace/0</c>
	    will probably only work when called from within a
	    '<c>try</c>' expression (otherwise it will return
	    <c>[]</c>.</p>
	    <p>To help prepare for that change, the compiler will now
	    by default warn if '<c>get_stacktrace/0</c>' is used in a
	    way that will not work in the future. Note that the
	    warning will not be issued if '<c>get_stacktrace/0</c>'
	    is used in a function that uses neither '<c>catch</c>'
	    nor '<c>try</c>' (because that could be a legal use if
	    the function is called from within a '<c>try</c>'.</p>
          <p>
	    Own Id: OTP-14401</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.0.4.1</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>Fail labels on guard BIFs weren't taken into account
	    during an optimization pass, and a bug in the validation
	    pass sometimes prevented this from being noticed when a
	    fault occurred.</p>
          <p>
	    Own Id: OTP-14522 Aux Id: ERIERL-48 </p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.0.4</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    Minor internal changes. A typo in the documentation was
	    also fixed.</p>
          <p>
	    Own Id: OTP-14240</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.0.3</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    Fixed a compiler crash when maps were matched.</p>
          <p>
	    Own Id: OTP-13931 Aux Id: ERL-266 </p>
        </item>
        <item>
          <p>
	    Fixed a compiler crash having to with the delayed
	    sub-creation optimization. (Thanks to Jose Valim for
	    reporting this bug.)</p>
          <p>
	    Own Id: OTP-13947 Aux Id: ERL-268 </p>
        </item>
        <item>
	    <p>The compiler option <c>inline_list_funcs</c>
	    accidentally turned off some other optimizations.</p>
          <p>
	    Own Id: OTP-13985</p>
        </item>
        <item>
	    <p>The compiler could sometimes generate spurious
	    warnings when inlining was enabled.</p>
          <p>
	    Own Id: OTP-14040 Aux Id: ERL-301 </p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.0.2</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    If the compiler fails to write the BEAM file, it will now
	    report the reason of the error for the write operation.</p>
          <p>
	    Own Id: OTP-13701</p>
        </item>
        <item>
          <p>
	    Fixed an internal compiler error. (Thanks to Svilen
	    Ivanov for reporting this bug.)</p>
          <p>
	    Own Id: OTP-13780 Aux Id: ERL-202 </p>
        </item>
        <item>
          <p>
	    The compiler could crash when trying to compile a
	    complicated expression with multiple catches all on one
	    line . (Thanks to Thomas Arts for reporting this bug.)</p>
          <p>
	    Own Id: OTP-13804 Aux Id: ERL-209 </p>
        </item>
        <item>
          <p>
	    Eliminated a few internal compiler failures.</p>
          <p>
	    Own Id: OTP-13863</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.0.1</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    A literal binary matching regression was introduced in
	    19.0 where a match could fail to resolve to the right
	    clause. This has now been fixed.</p>
          <p>
	    Own Id: OTP-13738</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 7.0</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p><c>compile:forms/1,2</c> would crash when used in a
	    working directory that had been deleted by another
	    process.</p>
          <p>
	    Own Id: OTP-13430 Aux Id: ERL-113 </p>
        </item>
        <item>
	    <p>Dialyzer no longer crashes when there is an invalid
	    function call such as <c>42(7)</c> in a module being
	    analyzed. The compiler will now warn for invalid function
	    calls such as <c>X = 42, x(7)</c>.</p>
          <p>
	    Own Id: OTP-13552 Aux Id: ERL-138 </p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    Optimization of tuple matching has been slightly
	    improved.</p>
          <p>
	    Own Id: OTP-12951</p>
        </item>
        <item>
	    <p>Five deprecated and undocumented functions in the
	    module <c>core_lib</c> have been removed. The functions
	    are: <c>get_anno/{1,2}</c>, <c>is_literal/1</c>,
	    <c>is_literal_list/1</c>, and <c>literal_value</c>. Use
	    the appropriate functions in the <c>cerl</c> module
	    instead.</p>
          <p>
	    Own Id: OTP-12979</p>
        </item>
        <item>
	    <p>The pre-processor can now expand the ?FUNCTION_NAME
	    and ?FUNCTION_ARITY macros.</p>
          <p>
	    Own Id: OTP-13059</p>
        </item>
        <item>
	    <p>The function mapfold/4 has been added to the
	    <c>cerl_trees</c> module.</p>
          <p>
	    Own Id: OTP-13280</p>
        </item>
        <item>
	    <p>Bitstring comprehensions have been generalized to
	    allow arbitrary expressions in the construction part.</p>
          <p>
	    Own Id: OTP-13289</p>
        </item>
        <item>
	    <p>The compiler will now produce warnings for binary
	    patterns that will never match (example:
	    <c>&lt;&lt;-1/unsigned&gt;&gt; = Bin</c>). </p>
          <p>
	    Own Id: OTP-13374 Aux Id: ERL-44 </p>
        </item>
        <item>
	    <p>The compiler will no longer put the compilation date
	    and time into BEAM files. That means that two BEAM files
	    compiled on the same computer from the same source code
	    and compilation options will be identical.</p>
	    <p>Note: If you want to find out whether a BEAM file on
	    disk is different from the loaded code, compared the MD5
	    value obtained from <c>Mod:module_info(md5)</c> with the
	    MD5 value obtained from
	    <c>beam_lib:md5(BeamFileForMod)</c></p>.
          <p>
	    *** POTENTIAL INCOMPATIBILITY ***</p>
          <p>
	    Own Id: OTP-13504</p>
        </item>
        <item>
	    <p>The function <c>compile:env_compiler_options/0</c> has
	    been added to allow tools to pick up the same default
	    compiler options as the compiler itself.</p>
          <p>
	    Own Id: OTP-13654</p>
        </item>
      </list>
    </section>

</section>


<section><title>Compiler 6.0.3.1</title>
    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>Fail labels on guard BIFs weren't taken into account
	    during an optimization pass, and a bug in the validation
	    pass sometimes prevented this from being noticed when a
	    fault occurred.</p>
          <p>
	    Own Id: OTP-14522 Aux Id: ERIERL-48 </p>
        </item>
      </list>
    </section>
</section>

<section><title>Compiler 6.0.3</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    An complicated guard expression in a function call could
	    crash the compiler. (Thanks to Thomas Arts for reporting
	    this bug.)</p>
          <p>
	    Own Id: OTP-13208</p>
        </item>
        <item>
	    <p>Constructing a map in a guard in a catch could crash
	    the compiler. (Thanks to Thomas Arts for reporting this
	    bug.)</p>
          <p>
	    Own Id: OTP-13223</p>
        </item>
        <item>
	    <p>Updating a fun as if it were a map would cause the
	    compiler to crash. (Thanks to Thomas Arts for reporting
	    this bug.)</p>
          <p>
	    Own Id: OTP-13231</p>
        </item>
        <item>
          <p>
	    Fix pretty printing of Core Maps</p>
          <p>
	    Literal maps could cause Dialyzer to crash when pretty
	    printing the results.</p>
          <p>
	    Own Id: OTP-13238</p>
        </item>
        <item>
          <p>
	    A complex combination of bit syntax matching operations
	    would cause an internal consistency check failure during
	    compilation. (Thanks to Jose Valim for reporting this
	    bug.)</p>
          <p>
	    Own Id: OTP-13309</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 6.0.2</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    Fix cerl_trees:label/2 bug with map K/V swap</p>
          <p>
	    Own Id: OTP-13091</p>
        </item>
        <item>
          <p>
	    Warnings produced when the '<c>bin_opt_info</c>' option
	    was given could sometimes lack filenames and line
	    numbers. (Thanks to José Valim for reporting this bug.)</p>
          <p>
	    Own Id: OTP-13113</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 6.0.1</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    Fix <c>get_map_elements</c> register corruption</p>
          <p>
	    Instruction <c>get_map_elements</c> might destroy target
	    registers when the fail-label is taken. Only seen for
	    patterns with two, and only two, target registers.
	    Specifically if we copy one register and then jump.</p>
          <p>
	    Own Id: OTP-12967</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 6.0</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    The compiler optimizes away building of terms that are
	    never actually used. As a result, the compiler in OTP 18
	    may produce more warnings for terms that are built but
	    not used than the compiler in OTP 17.</p>
          <p>
	    Own Id: OTP-12453</p>
        </item>
        <item>
          <p>
	    Using a map could incorrectly suppress warnings for
	    unused variables.</p>
          <p>
	    Own Id: OTP-12515</p>
        </item>
        <item>
          <p>
	    The compiler now properly reports unknown parse
	    transforms. That is, <c>undef</c> exceptions coming from
	    the parse transform itself is reported differently from
	    the absence of the parse transform.</p>
          <p>
	    Own Id: OTP-12723</p>
        </item>
        <item>
          <p>
	    Allow for 'creation of sub binary delayed' optimization
	    if maps instructions are in a clause.</p>
          <p>
	    Own Id: OTP-12758</p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    The <c>cerl</c> and <c>cerl_trees</c> modules in the
	    Compiler application are now documented.</p>
          <p>
	    Own Id: OTP-11978</p>
        </item>
        <item>
          <p>
	    The deprecated '<c>asm</c>' option has been removed.</p>
          <p>
	    *** POTENTIAL INCOMPATIBILITY ***</p>
          <p>
	    Own Id: OTP-12100</p>
        </item>
        <item>
          <p>
	    Support variables as Map keys in expressions and patterns</p>
	    <p>Erlang will accept any expression as keys in Map
	    expressions and it will accept literals or bound
	    variables as keys in Map patterns.</p>
          <p>
	    Own Id: OTP-12218</p>
        </item>
        <item>
          <p>
	    Infer Map type information in beam_type compiler
	    optimization pass.</p>
          <p>
	    Own Id: OTP-12253</p>
        </item>
        <item>
          <p>
	    Compiler optimizations have been improved.</p>
          <p>
	    Own Id: OTP-12393</p>
        </item>
        <item>
          <p>
	    Five undocumented functions in the module <c>core_lib</c>
	    have been deprecated and will be removed in the next
	    major release. The functions are: <c>get_anno/{1,2}</c>,
	    <c>is_literal/1</c>, <c>is_literal_list/1</c>, and
	    <c>literal_value</c>. Use the appropriate functions in
	    the <c>cerl</c> module instead.</p>
          <p>
	    Own Id: OTP-12497</p>
        </item>
        <item>
          <p>
	    Change some internal data structures to Maps in order to
	    speed up compilation time. Measured speed up is around
	    10%-15%.</p>
          <p>
	    Own Id: OTP-12774</p>
        </item>
        <item>
          <p>
	    Fix beam_bool pass for Maps instruction get_map_elements</p>
          <p>
	    Before beam_split the get_map_elements instruction is
	    still in blocks and the helper function in beam_jump did
	    not reflect this.</p>
          <p>
	    Own Id: OTP-12844 Aux Id: 17 </p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 5.0.4</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    Matching out a map from a record and then updating the
	    record could cause a 'badarg' exception at run-time.
	    (Thanks to Dmitry Aleksandrov for reporting this bug.)</p>
          <p>
	    Own Id: OTP-12402</p>
        </item>
        <item>
	    <p>The compiler would crash when compiling some complex,
	    nonsensical guards such as:</p>
	    <p> ... <c>when {{X}}, -X</c>...</p>
          <p>
	    Own Id: OTP-12410</p>
        </item>
        <item>
          <p>
	    In rare circumstances, using binary pattern in the value
	    part of a map pattern would cause the compiler to crash.</p>
          <p>
	    Own Id: OTP-12414</p>
        </item>
        <item>
	    <p>Case expressions where a map was wrapped in a tuple or
	    list such as:</p>
	    <p><c>case {a,Map} of</c><br/> <c>{a,#{k:=_}}=Tuple -&gt;
	    Tuple</c><br/> <c>end.</c></p>
	    <p>would be unsafely "optimized" to either cause an
	    exception at run-time or would return an empty map.</p>
          <p>
	    Own Id: OTP-12451</p>
        </item>
        <item>
	    <p>When a variable was compared to a literal map using
	    the '<c>==</c>' operator, the compiler would change the
	    operator to '<c>=:=</c>' since it is more efficient.
	    However, this optimization is not safe if the map literal
	    has numeric keys or values. The compiler will now only do
	    the optimization if all keys and values are
	    non-numeric.</p>
          <p>
	    Own Id: OTP-12456</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 5.0.3</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    Named funs with the same name and arity could get mixed
	    up with each other.</p>
          <p>
	    Own Id: OTP-12262</p>
        </item>
        <item>
          <p>
	    Coalesce map keys in dialyzer mode</p>
          <p>
	    This fixes a regression introduced in commit
	    805f9c89fc01220bc1bb0f27e1b68fd4eca688ba The problem
	    occurred with compounded map keys compiled with dialyzer
	    option turned on, '+dialyzer'.</p>
          <p>
	    Reported by: Ivan Uemlianin</p>
          <p>
	    Own Id: OTP-12347</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 5.0.2</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    Corrected a bug with incorrect code generation when
	    inlining was turned on.</p>
          <p>
	    Own Id: OTP-12132</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 5.0.1</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    A Dialyzer crash involving analysis of Map types has now
	    been fixed.</p>
          <p>
	    Own Id: OTP-11947</p>
        </item>
        <item>
	    <p>The compiler would fail to compile a file with a
	    latin-1 character in the false branch of an <c>-ifdef</c>
	    or <c>-indef</c>.</p>
          <p>
	    Own Id: OTP-11987</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 5.0</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    Line numbers would not be correct when a binary
	    construction such as
	    '<c>&lt;&lt;Bin/binary,...&gt;&gt;</c>' fails. (Thanks to
	    Stanislav Seletskiy for reporting this bug.)</p>
          <p>
	    Own Id: OTP-11572</p>
        </item>
        <item>
          <p>
	    The compiler now properly annotates the code in value in
	    the '<c>after</c>' clause for a '<c>try</c>' so that
	    Dialyzer no longer generates a false warning for an
	    unmatched return.</p>
          <p>
	    Own Id: OTP-11580</p>
        </item>
        <item>
          <p>
	    Some case statements where no clause would match could
	    cause an internal error in the compiler. (Thanks to Erik
	    Soe Sorensen for reporting this bug.)</p>
          <p>
	    Own Id: OTP-11610</p>
        </item>
        <item>
          <p>
	    With <c>--Wunmatched_returns</c>, dialyzer will no longer
	    warn when the value of a list comprehension is ignored,
	    provided that the each value in the list would be an
	    atomic value (such as integer or atoms, as opposed to
	    tuples and lists). Example: ignoring '<c>[io:format(...)
	    || ...]</c>' will not cause a warning, while ignoring
	    '<c>[file:close(Fd) || ...]</c>' will.</p>
          <p>
	    Own Id: OTP-11626</p>
        </item>
        <item>
          <p>
	    Matching out a binary and applying the binary as if it
	    were a fun would crash the run-time system. (Thanks to
	    Loïc Hoguin.)</p>
          <p>
	    Own Id: OTP-11672</p>
        </item>
        <item>
          <p>
	    Some local implementations of removing the last element
	    from a list are replaced by <c>lists:droplast/1</c>. Note
	    that this requires at least <c>stdlib-2.0</c>, which is
	    the stdlib version delivered in OTP 17.0. (Thanks to Hans
	    Svensson)</p>
          <p>
	    Own Id: OTP-11678</p>
        </item>
        <item>
          <p>
	    Allow all auto imports to be suppressed at once.
	    Introducing the no_auto_import attribute:
	    -compile(no_auto_import). Useful for code generation
	    tools that always use the qualified function names and
	    want to avoid the auto imported functions clashing with
	    local ones. (Thanks to José Valim.)</p>
          <p>
	    Own Id: OTP-11682</p>
        </item>
        <item>
          <p>
	    Application upgrade (appup) files are corrected for the
	    following applications: </p>
          <p>
	    <c>asn1, common_test, compiler, crypto, debugger,
	    dialyzer, edoc, eldap, erl_docgen, et, eunit, gs, hipe,
	    inets, observer, odbc, os_mon, otp_mibs, parsetools,
	    percept, public_key, reltool, runtime_tools, ssh,
	    syntax_tools, test_server, tools, typer, webtool, wx,
	    xmerl</c></p>
          <p>
	    A new test utility for testing appup files is added to
	    test_server. This is now used by most applications in
	    OTP.</p>
          <p>
	    (Thanks to Tobias Schlager)</p>
          <p>
	    Own Id: OTP-11744</p>
        </item>
        <item>
          <p>
	    Adapt 'asm' deprecation message to new version scheme.
	    (Thanks to Tuncer Ayaz)</p>
          <p>
	    Own Id: OTP-11751</p>
        </item>
        <item>
          <p>
	    A number of compiler errors where unusual or nonsensical
	    code would crash the compiler have been reported by Ulf
	    Norell and corrected by Anthony Ramine.</p>
          <p>
	    Own Id: OTP-11770</p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    Compilation times for modules with a huge number for
	    record accesses using the dot operator has been improved.</p>
          <p>
	    Own Id: OTP-10652</p>
        </item>
        <item>
          <p>
	    The compiler can generate somewhat better code by moving
	    let expressions into sequences. (Thanks to Anthony
	    Ramine.)</p>
          <p>
	    Own Id: OTP-11056</p>
        </item>
        <item>
          <p>
	    Forbid unsized fields in patterns of binary generators
	    and simplified v3_core's translation of bit string
	    generators. (Thanks to Anthony Ramine.)</p>
          <p>
	    Own Id: OTP-11186</p>
        </item>
        <item>
          <p>
	    Funs can now be a given a name. Thanks to to Richard
	    O'Keefe for the idea (EEP37) and to Anthony Ramine for
	    the implementation.</p>
          <p>
	    Own Id: OTP-11537</p>
        </item>
        <item>
          <p>
	    Using the <c>from_asm</c> option to produce a BEAM file
	    starting from BEAM assembly code would often fail because
	    early optimization passes would not understand
	    instructions that later optimization passes would
	    introduce. (Thanks to Anthony Ramine.)</p>
          <p>
	    Own Id: OTP-11544</p>
        </item>
        <item>
          <p>
	    The <c>.core</c> and <c>.S</c> extensions are now
	    documented in the <c>erlc</c> documentation, and the
	    '<c>from_core</c>' and '<c>from_asm</c>' options are now
	    documented in the compiler documentation. (Thanks to
	    Tuncer Ayaz.)</p>
          <p>
	    Own Id: OTP-11547</p>
        </item>
        <item>
	    <p>Optimization of case expressions that build tuples or
	    lists have been improved.</p>
          <p>
	    Own Id: OTP-11584</p>
        </item>
        <item>
          <p>
	    EEP43: New data type - Maps</p>
          <p>
	    With Maps you may for instance:</p>
          <taglist>
            <tag/> <item><c>M0 = #{ a =&gt; 1, b =&gt; 2}, % create
	      associations</c></item>
            <tag/><item><c>M1 = M0#{ a := 10 }, % update values</c></item>
            <tag/><item><c>M2 = M1#{ "hi" =&gt;
	      "hello"}, % add new associations</c></item>
            <tag/><item><c>#{ "hi" := V1, a := V2, b := V3} = M2.
               % match keys with values</c></item>
          </taglist>
		<p>
	    For information on how to use Maps please see Map Expressions in the
		<seealso marker="doc/reference_manual:expressions#map_expressions">
			Reference Manual</seealso>.</p>
          <p>
	    The current implementation is without the following
	    features:</p>
          <taglist>
            <tag/><item>No variable keys</item>
	    <tag/><item>No single value access</item>
            <tag/><item>No map comprehensions</item>
          </taglist>
          <p>
	    Note that Maps is <em>experimental</em> during OTP 17.0.</p>
          <p>
	    Own Id: OTP-11616</p>
        </item>
        <item>
          <p>
	    Some function specs are corrected or moved and some edoc
	    comments are corrected in order to allow use of edoc.
	    (Thanks to Pierre Fenoll)</p>
          <p>
	    Own Id: OTP-11702</p>
        </item>
        <item>
          <p>
	    Thanks to Anthony Ramine for several improvements to the
	    optimizations in the BEAM compiler and for cleaning up
	    the code the code that transforms list and binary
	    comprehensions to Core Erlang.</p>
          <p>
	    Own Id: OTP-11720</p>
        </item>
        <item>
          <p>
	    The default encoding for Erlang source files is now
	    UTF-8. As a temporary measure to ease the transition from
	    the old default of latin-1, if the compiler encounters
	    byte sequences that are not valid UTF-8 sequences, the
	    compiler will re-try the compilation in latin-1 mode.
	    This workaround will be removed in a future release.</p>
          <p>
	    Own Id: OTP-11791</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.9.4</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    Typo fix ambigous -&gt; ambiguous. Thanks to Leo Correa.</p>
          <p>
	    Own Id: OTP-11455</p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    Lift 'after' blocks to zeroary functions. Thanks to
	    Anthony Ramine.</p>
          <p>
	    Own Id: OTP-11267</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.9.3</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    Expressions such as <c>'B = is_integer(V), if B and B
	    -&gt; ok end'</c> would crash the compiler.</p>
          <p>
	    Own Id: OTP-11240</p>
        </item>
        <item>
          <p>
	    <c>compile:file2/2</c> with the option
	    <c>report_errors</c> could return ErrorInfo tuples with
	    only two elements, while the documentation says that the
	    ErrorInfo tuple always has three elements. Also updated
	    the documentation to add that the first element may be
	    '<c>none</c>' if no line number is applicable.</p>
          <p>
	    Own Id: OTP-11304 Aux Id: seq12412 </p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    Fix matching of floating point middle-endian machines.
	    Thanks to Johannes Weissl.</p>
          <p>
	    Own Id: OTP-11201</p>
        </item>
        <item>
          <p>
	    Restrict inlining of local fun references. Thanks to
	    Anthony Ramine.</p>
          <p>
	    Own Id: OTP-11211</p>
        </item>
        <item>
          <p>
	    Silence a misleading warning with some comprehensions.
	    Thanks to Anthony Ramine.</p>
          <p>
	    Own Id: OTP-11212</p>
        </item>
        <item>
          <p>
	    Forbid returning a match context in beam_validator.
	    Thanks to Anthony Ramine.</p>
          <p>
	    Own Id: OTP-11247</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.9.2</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    Compiling functions with complex boolean operations in
	    guards could be very slow. (Thanks to Magnus Muller for
	    reporting this issue.)</p>
          <p>
	    Own Id: OTP-10939</p>
        </item>
        <item>
          <p>
	    Certain guard expressions used in a receive statement
	    could cause the compiler to crash.</p>
          <p>
	    Own Id: OTP-11119 Aux Id: seq12342 </p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    Fix optimization of some binary comprehensions. Thanks to
	    Anthony Ramine.</p>
          <p>
	    Own Id: OTP-11005</p>
        </item>
        <item>
          <p>
	    Use a set to store ref registers in beam_receive. Thanks
	    to Anthony Ramine.</p>
          <p>
	    Own Id: OTP-11069</p>
        </item>
        <item>
          <p>
	    Fix renaming of bs_put_string instructions. Thanks to
	    Anthony Ramine.</p>
          <p>
	    Own Id: OTP-11129</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.9.1</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    The compiler would crash attempting to compile
	    expressions such as "element(2, not_tuple)".</p>
          <p>
	    Own Id: OTP-10794</p>
        </item>
        <item>
          <p>
	    Forbid multiple values in Core Erlang sequence arguments.
	    Thanks to José Valim and Anthony Ramine.</p>
          <p>
	    Own Id: OTP-10818</p>
        </item>
        <item>
          <p>
	    An unsafe optimization would cause the compiler to crash
	    with an internal error for certain complex code
	    sequences.</p>
          <p>
	    Own Id: OTP-10825 Aux Id: seq12247 </p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
	    <p>Integers in expression that will give a floating point
	    result (such as "<c>X / 2</c>" will now be converted to
	    floating point at compile-time. (Suggested by Richard
	    O'Keefe.)</p>
	    <p>Identical floating points constans in a module will
	    now be coalesced to one entry in the constant pool.</p>
          <p>
	    Own Id: OTP-10788</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.9</title>

    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    The compiler optimizations have been polished, so that
	    the code quality will be slightly better in some cases.</p>
          <p>
	    Own Id: OTP-10193</p>
        </item>
        <item>
	    <p> Support for Unicode has been implemented. </p>
          <p>
	    Own Id: OTP-10302</p>
        </item>
        <item>
	    <p>Where necessary a comment stating encoding has been
	    added to Erlang files. The comment is meant to be removed
	    in Erlang/OTP R17B when UTF-8 becomes the default
	    encoding. </p>
          <p>
	    Own Id: OTP-10630</p>
        </item>
        <item>
          <p>
	    Fix some wrong warnings triggered by the option
	    inline_list_funcs. Thanks to Anthony Ramine.</p>
          <p>
	    Own Id: OTP-10690</p>
        </item>
        <item>
          <p>
	    Forbid local fun variables in Core Erlang guards. Thanks
	    to Anthony Ramine.</p>
          <p>
	    Own Id: OTP-10706</p>
        </item>
        <item>
          <p>
	    Binary syntax matches could cause an internal consistency
	    error in in the compiler. (Thanks to Viktor Sovietov for
	    reporting this bug.)</p>
          <p>
	    Own Id: OTP-10724</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.8.2</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    Modules with very many functions would compile very
	    slowly.</p>
          <p>
	    Own Id: OTP-10123</p>
        </item>
        <item>
          <p>
	    <c>compile:forms/2</c> will now use a
	    {source,SourceFilePath} to set the source returned by
	    <c>module_info(compile)</c> (Thanks to José Valim)</p>
          <p>
	    Own Id: OTP-10150</p>
        </item>
        <item>
          <p>
	    A process which had enabled trap_exit would receive EXIT
	    messages after calling the compiler. (Thanks to Jeremy
	    Heater.)</p>
          <p>
	    Own Id: OTP-10171</p>
        </item>
        <item>
          <p>
	    Fix messages ordering with column numbers</p>
          <p>
	    Own Id: OTP-10183</p>
        </item>
        <item>
          <p>
	    sys_pre_expand: Fix BASE never being set</p>
          <p>
	    Commit a612e99fb5aaa934fe5a8591db0f083d7fa0b20a turned
	    module attributes from 2-tuples to 3-tuples but forgot to
	    update get_base/1, breaking BASE for parametric modules.</p>
          <p>
	    Own Id: OTP-10184</p>
        </item>
        <item>
          <p>
	    The compiler will now issue a warning if literal tuple
	    funs are used. For example, {erlang,is_tuple}(X) will now
	    generate a warning.</p>
          <p>
	    Own Id: OTP-10185</p>
        </item>
        <item>
          <p>
	    The compiler will now warn for illegal sizes for segments
	    in binary construction. For example,
	    &lt;&lt;X:(2.5)&gt;&gt; will now cause the compiler to
	    issue a warning.</p>
          <p>
	    Own Id: OTP-10197</p>
        </item>
        <item>
          <p>
	    Fix the erlc -MP flag</p>
          <p>
	    Because of a copy-and-paste error in erlc.c, the -MP flag
	    had the same effect as -MG. As a workaround, you had to
	    pass +makedep_phony to enable the MP option. This patch
	    makes -MP work as intended.</p>
          <p>
	    Own Id: OTP-10211</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.8.1</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    In rare circumstance, the compiler could crash when
	    compiling a case statement. (Thanks to Hakan Mattsson.)</p>
          <p>
	    Own Id: OTP-9842</p>
        </item>
        <item>
	    <p>Calling a guard test (such as is_list/1) from the
	    top-level in a guard, would cause a compiler crash if
	    there was a local definition with the same name.
	    Corrected to reject the program with an error
	    message.</p>
          <p>
	    Own Id: OTP-9866</p>
        </item>
        <item>
	    <p>Using <c>get/1</c> in a <c>try</c> block could in some
	    cases cause an internal compiler error. (Thanks to Eric
	    Merritt.)</p>
          <p>
	    Own Id: OTP-9867</p>
        </item>
        <item>
          <p>
	    An unexported on_load function would not get run if the
	    module was compiled with the <c>inline</c> option.
	    (Thanks to Yiannis Tsiouris.)</p>
          <p>
	    Own Id: OTP-9910</p>
        </item>
        <item>
          <p>
	    Fixed a discrepancy in compile_info</p>
          <p>
	    The BEAM disassembler used the atom 'none' to signify the
	    absence of a compile_info chunk in a .beam file. This
	    clashed with the type declaration of the compile_info
	    field of a #beam_file{} record as containing a list. Now
	    [] signifies the absence of this chunk. This simplifies
	    the code and avoids a dialyzer warning.</p>
          <p>
	    Own Id: OTP-9917</p>
        </item>
        <item>
          <p>
	    Fix typo in `compile' doc: unmatched parenthesis (Thanks
	    to Ricardo Catalinas Jiménez)</p>
          <p>
	    Own Id: OTP-9919</p>
        </item>
        <item>
	    <p>In a <c>try</c>...<c>catch</c> statement that always
	    returned <c>false</c>, the compiler would remove calls to
	    BIFs that could not cause an exception (such as
	    <c>put/2</c>). Example of such code: <c>try put(K, V),
	    false catch _:_ -&gt; false end.</c></p>
          <p>
	    Own Id: OTP-9982</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.8</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    Add '-callback' attributes in stdlib's behaviours</p>
          <p>
	    Replace the behaviour_info(callbacks) export in stdlib's
	    behaviours with -callback' attributes for all the
	    callbacks. Update the documentation with information on
	    the callback attribute Automatically generate
	    'behaviour_info' function from '-callback' attributes</p>
          <p>
	    'behaviour_info(callbacks)' is a special function that is
	    defined in a module which describes a behaviour and
	    returns a list of its callbacks.</p>
          <p>
	    This function is now automatically generated using the
	    '-callback' specs. An error is returned by lint if user
	    defines both '-callback' attributes and the
	    behaviour_info/1 function. If no type info is needed for
	    a callback use a generic spec for it. Add '-callback'
	    attribute to language syntax</p>
          <p>
	    Behaviours may define specs for their callbacks using the
	    familiar spec syntax, replacing the '-spec' keyword with
	    '-callback'. Simple lint checks are performed to ensure
	    that no callbacks are defined twice and all types
	    referred are declared.</p>
          <p>
	    These attributes can be then used by tools to provide
	    documentation to the behaviour or find discrepancies in
	    the callback definitions in the callback module.</p>
          <p>
	    Add callback specs into 'application' module in kernel
	    Add callback specs to tftp module following internet
	    documentation Add callback specs to inets_service module
	    following possibly deprecated comments</p>
          <p>
	    Own Id: OTP-9621</p>
        </item>
        <item>
          <p>
	    The calculation of the 'uniq' value for a fun (see
	    <c>erlang:fun_info/1</c>) was too weak and has been
	    strengthened. It used to be based on the only the code
	    for the fun body, but it is now based on the MD5 of the
	    BEAM code for the module.</p>
          <p>
	    Own Id: OTP-9667</p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
	    <p>Variables are now now allowed in '<c>fun M:F/A</c>' as
	    suggested by Richard O'Keefe in EEP-23.</p>
	    <p>The representation of '<c>fun M:F/A</c>' in the
	    abstract format has been changed in an incompatible way.
	    Tools that directly read or manipulate the abstract
	    format (such as parse transforms) may need to be updated.
	    The compiler can handle both the new and the old format
	    (i.e. extracting the abstract format from a pre-R15 BEAM
	    file and compiling it using compile:forms/1,2 will work).
	    The <c>syntax_tools</c> application can also handle both
	    formats.</p>
          <p>
	    *** POTENTIAL INCOMPATIBILITY ***</p>
          <p>
	    Own Id: OTP-9643</p>
        </item>
        <item>
          <p>
	    <c>filename:find_src/1,2</c> will now work on stripped
	    BEAM files (reported by Per Hedeland). The HiPE compiler
	    will also work on stripped BEAM files. The BEAM compiler
	    will no longer include compilation options given in the
	    source code itself in <c>M:module_info(compile)</c>
	    (because those options will be applied anyway if the
	    module is re-compiled).</p>
          <p>
	    Own Id: OTP-9752</p>
        </item>
        <item>
	    <p>Inlining binary matching could cause an internal
	    compiler error. (Thanks to Rene Kijewski for reporting
	    this bug.)</p>
          <p>
	    Own Id: OTP-9770</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.7.5</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    Compiler options given in the source code using a
	    <c>-compile()</c> attribute used to be included twice in
	    <c>Mod:module_info(compile)</c>. They are now only
	    included once at the beginning of the list of options.</p>
          <p>
	    Own Id: OTP-9534</p>
        </item>
        <item>
          <p>
	    beam_disasm: Handle stripped BEAM files</p>
          <p>
	    beam_disasm:file/1 would crash if asked to disassemble a
	    stripped BEAM file without an "Attr" chunk. (Thanks to
	    Haitao Li)</p>
          <p>
	    Own Id: OTP-9571</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.7.4</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    If a variable is matched out in binary matching and used
	    as the size for a binary element, it would seem to be
	    unbound if used in a subsequent match operation. (Thanks
	    to Bernard Duggan.)</p>
          <p>
	    Own Id: OTP-9134</p>
        </item>
        <item>
	    <p>Eliminate incorrect warning in
	    <c>sys_core_fold</c></p>
          <p>
	    Own Id: OTP-9152</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.7.3</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    The <c>-export_type()</c> directive is no longer included
	    among the attributes.</p>
          <p>
	    Own Id: OTP-8998</p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    The maximum number of allowed arguments for an Erlang
	    function has been lowered from 256 to 255, so that the
	    number of arguments can now fit in a byte.</p>
          <p>
	    Own Id: OTP-9049</p>
        </item>
        <item>
          <p>
	    Dependency generation for Makefiles has been added to the
	    compiler and erlc. See the manual pages for
	    <c>compile</c> and <c>erlc</c>. (Thanks to Jean-Sebastien
	    Pedron.)</p>
          <p>
	    Own Id: OTP-9065</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.7.2</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    Two compiler bugs (that would cause the compiler to
	    terminate) reported by Christopher Williams have been
	    fixed.</p>
          <p>
	    Own Id: OTP-8949</p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
	    <p>The compiler would translate binary comprehensions
	    containing tail segments in a way that would would
	    confuse Dialyzer. For instance:</p>
	    <p><c>[42 || &lt;&lt;_:8/integer, _/bits&gt;&gt; &lt;=
	    Bits]</c></p>
          <p>
	    would produce a Dialyzer warning.</p>
          <p>
	    Own Id: OTP-8864</p>
        </item>
        <item>
          <p>
	    Code such as <c>foo(A) -&gt; &lt;&lt;A:0&gt;&gt;</c>
	    would crash the compiler.</p>
          <p>
	    Own Id: OTP-8865</p>
        </item>
        <item>
          <p>
	    The compiler could fail with an internal error when
	    variables were exported from a receive block but the
	    return value of the receive block were not used. (Thanks
	    to Jim Engquist for reporting this error.)</p>
          <p>
	    Own Id: OTP-8888</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.7.1</title>

    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    Eliminated warnings for auto-imported BIF clashes.</p>
          <p>
	    Own Id: OTP-8840</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.7</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    Several problems in the inliner have been fixed.</p>
          <p>
	    Own Id: OTP-8552</p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    The module binary from EEP31 (and EEP9) is implemented.</p>
          <p>
	    Own Id: OTP-8217</p>
        </item>
        <item>
	    <p>Local and imported functions now override the
	    auto-imported BIFs when the names clash. The pre R14
	    behaviour was that auto-imported BIFs would override
	    local functions. To avoid that old programs change
	    behaviour, the following will generate an error:</p>
	    <list><item><p>Doing a call without explicit module name
	    to a local function having a name clashing with the name
	    of an auto-imported BIF that was present (and
	    auto-imported) before OTP R14A</p></item>
	    <item><p>Explicitly importing a function having a name
	    clashing with the name of an autoimported BIF that was
	    present (and autoimported) before OTP R14A</p></item>
	    <item><p>Using any form of the old compiler directive
	    <c>nowarn_bif_clash</c></p></item> </list> <p>If the BIF
	    was added or auto-imported in OTP R14A or later,
	    overriding it with an import or a local function will
	    only result in a warning,</p> <p>To resolve clashes, you
	    can either use the explicit module name <c>erlang</c> to
	    call the BIF, or you can remove the auto-import of that
	    specific BIF by using the new compiler directive
	    <c>-compile({no_auto_import,[F/A]}).</c>, which makes all
	    calls to the local or imported function without explicit
	    module name pass without warnings or errors.</p> <p>The
	    change makes it possible to add auto-imported BIFs
	    without breaking or silently changing old code in the
	    future. However some current code ingeniously utilizing
	    the old behaviour or the <c>nowarn_bif_clash</c> compiler
	    directive, might need changing to be accepted by the
	    compiler.</p>
          <p>
	    *** POTENTIAL INCOMPATIBILITY ***</p>
          <p>
	    Own Id: OTP-8579</p>
        </item>
        <item>
	    <p>The undocumented, unsupport, and deprecated function
	    <c>lists:flat_length/1</c> has been removed.</p>
          <p>
	    Own Id: OTP-8584</p>
        </item>
        <item>
	    <p>Nested records can now be accessed without
	    parenthesis. See the Reference Manual for examples.
	    (Thanks to YAMASHINA Hio and Tuncer Ayaz.)</p>
          <p>
	    Own Id: OTP-8597</p>
        </item>
        <item>
	    <p>It is now possible to suppress the warning in code
	    such as "<c>list_to_integer(S), ok</c>" by assigning the
	    ignored value "_" like this: "<c>_ = list_to_integer(S),
	    ok</c>".</p>
          <p>
	    Own Id: OTP-8602</p>
        </item>
        <item>
	    <p><c>receive</c> statements that can only read out a
	    newly created reference are now specially optimized so
	    that it will execute in constant time regardless of the
	    number of messages in the receive queue for the process.
	    That optimization will benefit calls to
	    <c>gen_server:call()</c>. (See <c>gen:do_call/4</c> for
	    an example of a receive statement that will be
	    optimized.)</p>
          <p>
	    Own Id: OTP-8623</p>
        </item>
        <item>
	    <p>The compiler optimizes record operations better.</p>
          <p>
	    Own Id: OTP-8668</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.6.5</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    Using complex boolean expressions in ifs could cause the
	    compiler to either crash or teminate with an internal
	    error. (Thanks to Simon Cornish.)</p>
          <p>
	    Own Id: OTP-8338</p>
        </item>
        <item>
	    <p>Bit string comprehensions can now be used in
	    parameterized modules. (Thanks to Jebu Ittiachen.)</p>
          <p>
	    Own Id: OTP-8447</p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    The expected return value for an on_load function has
	    been changed. (See the section about code loading in the
	    Reference manual.)</p>
          <p>
	    *** POTENTIAL INCOMPATIBILITY ***</p>
          <p>
	    Own Id: OTP-8339</p>
        </item>
        <item>
          <p>
	    In rare circumstances when using garbaging collecting
	    guard BIFs, the validation pass (beam_validator) would
	    signal that the code was unsafe, when it in fact was
	    correct. (Thanks to Kiran Khaladkar.)</p>
          <p>
	    Own Id: OTP-8378</p>
        </item>
        <item>
          <p>
	    The <c>-Werror</c> option for <c>erlc</c> and the
	    compiler option <c>warnings_as_errors</c> will cause
	    warnings to be treated as errors. (Thanks to Christopher
	    Faulet.)</p>
          <p>
	    Own Id: OTP-8382</p>
        </item>
        <item>
	    <p>Macros overloading has been implemented. (Thanks to
	    Christopher Faulet.)</p>
          <p>
	    Own Id: OTP-8388</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.6.4</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>The compiler's 'E' option now works with modules with
	    types and specifications.</p>
          <p>
	    Own Id: OTP-8238 Aux Id: OTP-8150 </p>
        </item>
        <item>
          <p>
	    Certain uses of binary matching in a
	    <c>begin</c>-<c>end</c> in a list comprehension could
	    cause the compiler to crash or generate incorrect code.</p>
          <p>
	    Own Id: OTP-8271</p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    The documentation is now built with open source tools
	    (xsltproc and fop) that exists on most platforms. One
	    visible change is that the frames are removed.</p>
          <p>
	    Own Id: OTP-8201</p>
        </item>
        <item>
          <p>
	    The compiler could crash if invalid calls to is_record/2
	    was used in (for example) a list comprehension. (Thanks
	    to Tobias Lindahl.)</p>
          <p>
	    Own Id: OTP-8269</p>
        </item>
        <item>
	    <p>The -on_load() directive can be used to run a function
	    when a module is loaded. It is documented in the section
	    about code loading in the Reference Manual.</p>
          <p>
	    Own Id: OTP-8295</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.6.3</title>

    <section><title>Improvements and New Features</title>
      <list>
        <item>
	    <p>Corrected liveness optimization to eliminate a
	    compiler crash that could occur when compiling bit syntax
	    construction code. (Thanks to Mikage Sawatari.)</p>
	    <p>Calling BIFs such as <c>length/1</c> in guard context
	    in a try/catch block could cause a compiler crash.
	    (Thanks to Paul Fisher.)</p>
	    <p>Using filter expressions containing <c>andalso</c> or
	    <c>orelse</c> in a list comprehension could cause a
	    compiler crash. (Thanks to Martin Engström.)</p>
          <p>
	    Own Id: OTP-8054</p>
        </item>
        <item>
          <p>
	    A guard with nested 'not' operators could cause the
	    compiler to crash. (Thanks to Tuncer Ayaz.)</p>
          <p>
	    Own Id: OTP-8131</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.6.2</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    The compiler would crash while compiling certain complex
	    function bodies containing <c>receive after</c> due to a
	    bug in the jump optimizer (a label that had only had
	    backward references could still be removed). (Thanks to
	    Vincent de Phily.)</p>
          <p>
	    Own Id: OTP-7980</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.6.1</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    Miscellaneous minor bugs fixed.</p>
          <p>
	    Own Id: OTP-7937</p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    There will be more efficient code if there is a clause
	    that matches the empty binary and no other clauses that
	    matches non-empty binaries.</p>
          <p>
	    Own Id: OTP-7924</p>
        </item>
        <item>
	    <p>There is new option to allow a module to have a module
	    name other than the filename. Do not use it unless you
	    know what you are doing.</p>
          <p>
	    Own Id: OTP-7927</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.6.0.1</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
            <p>Using <c>andalso</c>/<c>orelse</c> or record access in
            a <c>try</c>...<c>catch</c> could cause a compiler
            crash.</p>
            <p>Som large and complex functions could require
            extremely long compilation times (hours or days).</p>
          <p>
            Own Id: OTP-7905</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.6</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    For some complex guards which used
	    <c>andalso</c>/<c>orelse</c>, the compiler would crash.
	    (Thanks to Hunter Morris.)</p>
          <p>
	    Own Id: OTP-7679</p>
        </item>
        <item>
          <p>
	    Code that (incorrectly) used the the value of nested
	    applications of <c>setelement/3</c> in bit syntax
	    construction could crash the compiler.</p>
          <p>
	    Own Id: OTP-7690</p>
        </item>
        <item>
	    <p>Modules containing huge integers (consisting of
	    several hundreds of thousands of digits or more) could be
	    slow to compile. This problem has been corrected.</p>
          <p>
	    Own Id: OTP-7707 Aux Id: seq11129 </p>
        </item>
        <item>
	    <p>If the generator in a list comprehension is given a
	    non-list term, there will now be <c>function_clause</c>
	    exception instead of a <c>case_clause</c> exception (as
	    it was in all releases before R12B).</p>
          <p>
	    Own Id: OTP-7844</p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    The compiler could crash if the size for a binary segment
	    in matching was a complex literal such as binary or
	    tuple.</p>
          <p>
	    Own Id: OTP-7650</p>
        </item>
        <item>
          <p>
	    The compiler generates more compact and faster code for
	    matching of complex constants (such as constant lists and
	    tuples).</p>
          <p>
	    Own Id: OTP-7655</p>
        </item>
        <item>
          <p>
	    The undocumented, unsupported, and deprecated guard BIF
	    <c>is_constant/1</c> has been removed.</p>
          <p>
	    *** INCOMPATIBILITY with R12B ***</p>
          <p>
	    Own Id: OTP-7673</p>
        </item>
        <item>
	    <p>The compiler generates better code for many guard
	    expressions, and especially for guards that use
	    <c>andalso</c>/<c>orelse</c> or record fields.</p>
	    <p>(In technical terms, <c>andalso</c>/<c>orelse</c> in a
	    guard would case the creation of a stack frame and saving
	    of all x registers that could potentially be alive after
	    the guard and restoring all x registers before leaving
	    the guard. For certain guards, far too many x registers
	    were saved and subsequently restored. In this version of
	    the compiler, no stack frame is created and no x
	    registers are saved and restored.)</p>
          <p>
	    Own Id: OTP-7718</p>
        </item>
        <item>
	    <p>The default size for the resulting binary created by a
	    binary comprehension was 64Kb in R12B (it would grow if
	    needed). This was often far too much. In this release,
	    the default is changed to 256 bytes. Furthermore, for
	    most binary comprehensions without filters, the exact
	    size of the resulting binary can be calculated beforehand
	    and the compiler now generates code that does that
	    calculation.</p>
          <p>
	    Own Id: OTP-7737</p>
        </item>
        <item>
	    <p>The short-circuit operators <c>andalso</c> and
	    <c>orelse</c> no longer guarantees that their second
	    argument is either <c>true</c> or <c>false</c>. As a
	    consequence, <c>andalso</c>/<c>orelse</c> are now
	    tail-recursive.</p>
          <p>
	    *** POTENTIAL INCOMPATIBILITY ***</p>
          <p>
	    Own Id: OTP-7748</p>
        </item>
        <item>
	    <p>The compiler will refuse to a compile file where the
	    module name in the file differs from the output file
	    name.</p>
	    <p>When compiling using <c>erlc</c>, the current working
	    directory will no be included in the code path (unless
	    explicitly added using "-pa .").</p>
          <p>
	    *** POTENTIAL INCOMPATIBILITY ***</p>
          <p>
	    Own Id: OTP-7793</p>
        </item>
        <item>
	    <p>There will no longer be any warnings for list
	    comprehensions without generators, as such list
	    comprehension have turned out to be useful.</p>
          <p>
	    Own Id: OTP-7846</p>
        </item>
        <item>
	    <p>Warnings for obsolete guard tests are now turned on.
	    (That is, writing <c>list(L)</c> in a guard instead of
	    <c>is_list(L)</c> will generate a warning.)</p>
	    <p>The warnings can be turned off using the
	    <c>nowarn_obsolete_guard</c> option.</p>
          <p>
	    Own Id: OTP-7850</p>
        </item>
        <item>
	    <p>The copyright notices have been updated.</p>
          <p>
	    Own Id: OTP-7851</p>
        </item>
        <item>
	    <p>If a module contains an exported function with the
	    same name as an auto-imported BIF (such as
	    <c>length/1</c>), any calls to the BIF must have an
	    explicit <c>erlang:</c> prefix, or there will be a
	    compilation error (such calls would only generate a
	    warning in previous releases).</p>
	    <p>(The reason for the change is to avoid breaking code
	    in a future major release, R14 or R15, in which we plan
	    to make calls without a module prefix always call the
	    local function in the same module even if there is an
	    auto-imported BIF with the same name.)</p>
          <p>
	    *** POTENTIAL INCOMPATIBILITY ***</p>
          <p>
	    Own Id: OTP-7873</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.5.5</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>Matching on a zero-width segment in the bit syntax
	    would crash the compiler. (Thanks to Will.)</p>
          <p>
	    Own Id: OTP-7591</p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    In bit syntax expressions which started with a binary
	    segment, and was followed by at least two segments of
	    variable size, too little space could be allocated for
	    the binary, leading to memory corruption.</p>
          <p>
	    Own Id: OTP-7556</p>
        </item>
        <item>
	    <p>In user-defined attributes, <c>Name/Arity</c> is now
	    allowed and will be translated to <c>{Name,Arity}</c>.
	    (An implementation of EEP-24 by Richard O'Keefe.)</p>
	    <p>The <c>module_info/{0,1}</c> functions automatically
	    inserted into each compiled modules are now documented in
	    the Modules section in the Reference Manual.</p>
          <p>
	    Own Id: OTP-7586</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.5.4</title>

    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    Certain complex bit syntax matching operations matching
	    out binaries and having several clauses could give
	    incorrect results (the matched out binaries were too
	    short). (Thanks to Christian von Roques for bug report
	    and correction.)</p>
          <p>
	    Own Id: OTP-7498</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.5.3</title>

    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    New option <c>warn_export_all</c> to warn for a module
	    using <c>export_all</c>. (Thanks to Richard Carlsson.)</p>
          <p>
	    Own Id: OTP-7392</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.5.2.1</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
            In rare circumstances, the length/1 BIF (and a few other
            guard BIFs) would seem to return an incorrect value (of
            any type).</p>
          <p>
            Own Id: OTP-7345 Aux Id: seq10962 </p>
        </item>
      </list>
    </section>

</section>
<section><title>Compiler 4.5.2</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>A bug in the old inliner has been fixed. Some
	    undocumented functionality has been removed.</p>
          <p>
	    Own Id: OTP-7223</p>
        </item>
        <item>
	    <p>Matching several binary patterns in parallel using the
	    '=' operator is not allowed (an implementation
	    limitation), but the compiler did not reject all such
	    attempts (depending on the patterns, the generated code
	    might or might not work correctly). Now the compiler
	    rejects all binary patterns joined by '='.</p>
          <p>
	    Own Id: OTP-7227</p>
        </item>
        <item>
	    <p>Complex combinations of record operations and binary
	    matching could cause the compiler to crash. (Thanks to
	    Vladimir Klebansky.)</p>
          <p>
	    Own Id: OTP-7233</p>
        </item>
        <item>
          <p>
	    In rare circumstances, mixing binary matching clauses
	    with clauses matching other data types, the compiler
	    could crash.</p>
          <p>
	    Own Id: OTP-7240 Aux Id: seq10916 </p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.5.1.1</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    Corrected a compiler bug that could cause a complex
	    binary matching operation to fail when it shouldn't.
	    (Thanks to Tomas Stejskal.)</p>
          <p>
	    Own Id: OTP-7188</p>
        </item>
        <item>
          <p>
	    In unusual circumstances, the environment for a fun could
	    bind wrong values.</p>
          <p>
	    Own Id: OTP-7202 Aux Id: seq10887 </p>
        </item>
        <item>
	    <p>Long sequences of list comprehensions without
	    generators joined by the '++' operator would cause a code
	    expansion explosion, which could cause the compiler to
	    run out of memory. To resolve this problem, in
	    '<c>[...||...]++Expr</c>', <c>Expr</c> is now evaluated
	    before the list comprehension. This change <em>is</em>
	    backwards compatible (see the following note about
	    evaluation order if you have doubts).</p>
	    <p>Note about evaluation order: The Reference manual says
	    that subexpressions are evaluated <em>in any order</em>
	    before the expression itself. Therefore, in an expression
	    such as '<c>LeftExpr++RightExpr</c>', you should not
	    depend on <c>LeftExpr</c> being evaluated before
	    <c>RightExpr</c> or vice versa. The evaluation order is
	    only important if the expressions contains and/or depends
	    on operations with side-effects, such as message passing
	    or ETS operations.</p>
          <p>
	    Own Id: OTP-7206</p>
        </item>
      </list>
    </section>

</section>


<section><title>Compiler 4.5.1</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
          <p>
	    A match expression inside a function call could cause a
	    false "a term is constructed but never used" warning.</p>
          <p>
	    Own Id: OTP-7018 Aux Id: seq10824 </p>
        </item>
        <item>
	    <p>The compiler could crash if a binary tail was matched
	    out, and then used in a binary append operation. (Thanks
	    to Oleg Avdeev.)</p>
	    <p>Similarly, the compiler could crash if a binary tail
	    was matched out, and then used (incorrectly) in binary
	    construction in an integer field. (Thanks to Fredrik
	    Svahn.) Or was incorrectly used in a float field. Or was
	    used in a binary field with a given length. (Thanks to
	    Chih - Wei Yu.) </p>
          <p>
	    Own Id: OTP-7022</p>
        </item>
        <item>
          <p>
	    Matching an empty binary in a record and then using the
	    same record again could cause a compiler crash. (Thanks
	    to Fredrik Thulin.)</p>
          <p>
	    Own Id: OTP-7029</p>
        </item>
        <item>
	    <p>In rare circumstances, constants containing floating
	    points and integers could be confused. Example:</p>
	    <p><c>f(a) -> [1]; f(b) -> [1.0].</c></p>
	    <p>Both <c>f(a)</c> and <c>f(b)</c> would return
	    <c>[1]</c>.</p>
          <p>
	    Own Id: OTP-7073</p>
        </item>
        <item>
	    <p>Some bit syntax code such as</p>
	    <p><c>matching d(_,&lt;$lt;$gt;$gt;) -> one; d(0,&lt;$lt;D$gt;$gt;)
	    ->two.</c></p>
	    <p>could crash the compiler. (Thanks to Simon
	    Cornish.)</p>
          <p>
	    Own Id: OTP-7094</p>
        </item>
        <item>
          <p>
	    In unusual circumstances, a call to a fun could fail due
	    to an unsafe optimization. (Thanks to Simon Cornish.)</p>
          <p>
	    Own Id: OTP-7102</p>
        </item>
        <item>
          <p>
	    Bit syntax matching with a guard containing two or more
	    uses of andalso/orelse could cause the compiler to crash.
	    (Thanks to Mateusz Berezecki.)</p>
          <p>
	    Own Id: OTP-7113</p>
        </item>
        <item>
          <p>
	    This was only a problem if you generated or wrote your
	    own Core Erlang code: The Core Erlang optimizer code
	    could move nested calls such as
	    <c>erlang:'$lt;'(erlang:length(L), 2)</c> as case expression
	    into a guard, which would change the semantics. (Thanks
	    to Robert Virding.)</p>
          <p>
	    Own Id: OTP-7117</p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    The compiler could generate suboptimal code for record
	    updates if the record update code consisted of multiple
	    source code lines.</p>
          <p>
	    Own Id: OTP-7101</p>
        </item>
      </list>
    </section>

</section>

<section><title>Compiler 4.5</title>

    <section><title>Fixed Bugs and Malfunctions</title>
      <list>
        <item>
	    <p>The compiler used to allow that a binary field without
	    size could be used in other positions than at the end in
	    bit syntax pattern. For instance,
	    <c><![CDATA[<<B/binary,EmptyBinary/binary>> = Bin]]></c>
	    used to compile, but now the compilation will fail with
	    an an error message.</p>
	    <p>Also, it is now longer permitted to give a literal
	    string in a binary pattern a type or a size; for
	    instance, <c><![CDATA[<<"abc"/binary>> = Bin]]></c> will
	    no longer compile. (In previous releases, there would
	    always be a <c>badmatch</c> exception at run-time.)</p>
          <p>
	    Own Id: OTP-6885</p>
        </item>
      </list>
    </section>


    <section><title>Improvements and New Features</title>
      <list>
        <item>
          <p>
	    Bitstrings (bit-level) binaries and binary comprehensions
	    are now part of the language. See the Reference Manual.</p>
          <p>
	    Own Id: OTP-6558</p>
        </item>
        <item>
          <p>
	    The '<c>compressed</c>' option for the compiler has been
	    documented.</p>
          <p>
	    Own Id: OTP-6801</p>
        </item>
        <item>
	    <p>If the value of a list comprehension is not used, such
	    as in '<c>[do_something(X) || X &lt;- List], ok</c>', a
	    result list will no longer be built. For more details,
	    see the Efficiency Guide.</p>
	    <p>If the value of an expression is not used, and the
	    expression has no side effects except for possibly
	    throwing an exception, a warning will be generated.
	    Examples: '<c>self(),ok</c>' and
	    '<c>{error,Reason},ok</c>'.</p>
          <p>
	    Own Id: OTP-6824</p>
        </item>
        <item>
          <p>
	    Three new functions have been added to the <c>compile</c>
	    module: <c>noenv_file/2</c>, <c>noenv_forms/2</c>, and
	    <c>noenv_output_generated/1</c>.</p>
          <p>
	    Own Id: OTP-6829</p>
        </item>
        <item>
	    <p>Many bit syntax operations, both construction and
	    matching, are faster. For further information, see the
	    Efficiency Guide.</p>
          <p>
	    Own Id: OTP-6838</p>
        </item>
        <item>
	    <p>Literal lists, tuples, and binaries are no longer
	    constructed at run-time as they used to be, but are
	    stored in a per-module constant pool. Literals that are
	    used more than once are stored only once.</p>
	    <p>This is not a change to the language, only in the
	    details of its implementation. Therefore, the
	    implications of this change is described in the
	    Efficiency Guide.</p>
	    <p>Example 1: In the expression <c>element(BitNum-1,
	    {1,2,4,8,16,32,64,128})</c>, the tuple used to be
	    constructed every time the expression was executed, which
	    could be detrimental to performance in two ways if the
	    expression was executed in a loop: the time to build the
	    tuple itself and the time spent in garbage collections
	    because the heap filled up with garbage faster.</p>
	    <p>Example 2: Literal strings, such as <c>"abc"</c>, used
	    to be stored in the compiled code compactly as a byte
	    string and expanded to a list at run-time. Now all
	    strings will be stored expanded to lists (such as
	    <c>[$a,$b,$c]</c>) in the constant pool. That means that
	    the string will be faster to use at run-time, but that it
	    will require more space even when not used. If space is
	    an issue, you might want to use binary literals (that is,
	    <c>&lt;&lt;"abc"&lt;&lt;</c>) instead of string literals for
	    infrequently used long strings (such as error
	    messages).</p>
          <p>
	    Own Id: OTP-6850</p>
        </item>
        <item>
          <p>
	    Recursive calls now usually consume less stack than in
	    R11B. See the Efficiency Guide.</p>
          <p>
	    Own Id: OTP-6862 Aux Id: seq10746 </p>
        </item>
        <item>
	    <p>Two new guard BIFs have been introduced as a
	    recommended replacement for <c>size/1</c>. (The
	    <c>size/1</c> BIF will be removed no earlier than in
	    R14B.) The BIFs are <c>tuple_size/1</c> to calculate the
	    size of a tuple and <c>byte_size/1</c> to calculate the
	    number of bytes needed for the contents of the binary or
	    bitstring (rounded up to the nearest number of bytes if
	    necessary).</p>
	    <p>There is also a new <c>bit_size/1</c> BIF that returns
	    the exact number of bits that a binary or bitstring
	    contains.</p>
          <p>
	    Own Id: OTP-6902</p>
        </item>
        <item>
          <p>
	    The two internal functions <c>erl_bifs:is_bif/3</c> and
	    <c>erl_bifs:is_guard/3</c> have been removed. They were
	    unsupported, undocumented, and unmaintained.</p>
          <p>
	    Own Id: OTP-6966</p>
        </item>
      </list>
    </section>

</section>

  <section>
    <title>Compiler 4.4.5</title>

    <section>
      <title>Fixed Bugs and Malfunctions</title>
      <list type="bulleted">
        <item>
          <p>The compiler would crash if you tried to combine to
            non-list literals with '<c><![CDATA[++]]></c>' (for instance,
            <c><![CDATA[an_atom++"string"]]></c>).</p>
          <p>Own Id: OTP-6630 Aux Id: seq10635 </p>
        </item>
      </list>
    </section>

    <section>
      <title>Improvements and New Features</title>
      <list type="bulleted">
        <item>
          <p>Minor Makefile changes.</p>
          <p>Own Id: OTP-6689</p>
        </item>
      </list>
    </section>
  </section>

  <section>
    <title>Compiler 4.4.4</title>

    <section>
      <title>Fixed Bugs and Malfunctions</title>
      <list type="bulleted">
        <item>
          <p>Incorrect code could be generated for bit syntax matching
            if the old inliner was used with aggressive settings.</p>
          <p>Own Id: OTP-6461</p>
        </item>
      </list>
    </section>
  </section>

  <section>
    <title>Compiler 4.4.3</title>

    <section>
      <title>Fixed Bugs and Malfunctions</title>
      <list type="bulleted">
        <item>
          <p>The R10B compiler could generate unsafe
            <c><![CDATA[bs_save/bs_restore]]></c> instructions that could cause
            memory corruption. (The R11B compiler does not have that
            problem.) The erlang emulator will now refuse to load
            R10B-compiled modules that contain such unsafe
            <c><![CDATA[bs_save/bs_restore]]></c> instructions. In addition, the
            beam_validator module in the compiler will also reject
            such instructions (in case it is used to validate R10B
            code). (Thanks to Matthew Reilly.)</p>
          <p>Own Id: OTP-6386</p>
        </item>
      </list>
    </section>

    <section>
      <title>Improvements and New Features</title>
      <list type="bulleted">
        <item>
          <p>Directives for parse transforms that have been run are
            now removed from the abstract code stored when the
            debug_info option is given, to prevent the parse
            transforms to be run again.</p>
          <p>Own Id: OTP-5344</p>
        </item>
        <item>
          <p>Minor improvements in code generation for some guards
            expression involving boolean expressions.</p>
          <p>Own Id: OTP-6347</p>
        </item>
      </list>
    </section>
  </section>

  <section>
    <title>Compiler 4.4.2.1</title>

    <section>
      <title>Fixed Bugs and Malfunctions</title>
      <list type="bulleted">
        <item>
          <p>The compiler could generate incorrect code for bit syntax
            matching consisting of several clauses.</p>
          <p>Own Id: OTP-6392 Aux Id: seq10539 </p>
        </item>
      </list>
    </section>
  </section>

  <section>
    <title>Compiler 4.4.2</title>

    <section>
      <title>Fixed Bugs and Malfunctions</title>
      <list type="bulleted">
        <item>
          <p>Defining a fun itself containing a fun in an
            <c><![CDATA[after]]></c> block of a <c><![CDATA[try]]></c> would cause the
            compiler to crash or generate incorrect code. (Thanks to
            Tim Rath.)</p>
          <p>Shorter compilation times for modules containing with
            an extreme number of functions (10000 functions or more).</p>
          <p>(The compiled could generate deprecated instructions
            for certain bit syntax matching operations.)</p>
          <p>Own Id: OTP-6212 Aux Id: seq10446 </p>
        </item>
        <item>
          <p>Fixed several bugs that would cause warnings to be shown
            without file name and line number.</p>
          <p>Own Id: OTP-6260 Aux Id: seq10461 </p>
        </item>
      </list>
    </section>

    <section>
      <title>Improvements and New Features</title>
      <list type="bulleted">
        <item>
          <p>The <c><![CDATA[strict_record_tests]]></c> option is now default;
            that is, reading a field from a record using the
            <c><![CDATA[Record#record_tag.field]]></c> syntax will fail if
            <c><![CDATA[Record]]></c> is not a record of the correct type.</p>
          <p>If necessary, the record tests can be turned off by
            giving the <c><![CDATA[no_strict_record_tests]]></c> option. To avoid
            editing Makefiles, the environment variable
            <c><![CDATA[ERL_COMPILER_OPTIONS]]></c> can be set to
            "<c><![CDATA[no_strict_record_tests]]></c>".</p>
          <p>The <c><![CDATA[no_strict_record_tests]]></c> option will probably
            be removed in the R12B release.</p>
          <p>*** POTENTIAL INCOMPATIBILITY ***</p>
          <p>Own Id: OTP-6294</p>
        </item>
      </list>
    </section>
  </section>

  <section>
    <title>Compiler 4.4.1</title>

    <section>
      <title>Fixed Bugs and Malfunctions</title>
      <list type="bulleted">
        <item>
          <p>The compiler used to crash if a module contained code
            similar to '<c><![CDATA[fun(1=0) -> ok end]]></c>'. (Thanks to
            Richard Carlsson.)</p>
          <p>The compiler would spend really long time compiling
            bit syntax expressions such as
            '<c><![CDATA[<<1:(50*1024*1024)>>]]></c>' and produce a huge .beam
            file. Corrected.</p>
          <p>The compiler would compile list comprehensions with
            many generators really, really slow. (Thanks to Thomas
            Raes.)</p>
          <p>Module attributes would be stored in reverse order
            compared to the order in the source code. (Thus,
            <c><![CDATA[M:module_info(attributes)]]></c> would also return the
            attributes in reversed order.)</p>
          <p>Defining a fun in an <c><![CDATA[after]]></c> block of a
            <c><![CDATA[try]]></c> would cause the compiler to crash or generate
            incorrect code. (Thanks to Martin Bjorklund.)</p>
          <p>The combination of binary pattern and a guard with
            andalso/orelse could cause the compiler to crash.</p>
          <p>Own Id: OTP-6121 Aux Id: seq10400 </p>
        </item>
      </list>
    </section>
  </section>

  <section>
    <title>Compiler 4.4</title>

    <section>
      <title>Fixed Bugs and Malfunctions</title>
      <list type="bulleted">
        <item>
          <p>When a <c><![CDATA[.hrl]]></c> file is included using
            <c><![CDATA[-include_lib]]></c>, the include path is temporarily
            updated to include the directory the <c><![CDATA[.hrl]]></c> file was
            found in, which will allow that <c><![CDATA[.hrl]]></c> file to itself
            include files from the same directory using
            <c><![CDATA[-include]]></c>. (Thanks to Richard Carlsson.)</p>
          <p>Own Id: OTP-5944</p>
        </item>
      </list>
    </section>

    <section>
      <title>Improvements and New Features</title>
      <list type="bulleted">
        <item>
          <p>The <c><![CDATA[andalso]]></c> and <c><![CDATA[orelse]]></c> operators are
            now allowed to be used in guards. That also applies to
            match specifications.</p>
          <p>Own Id: OTP-5894 Aux Id: OTP-5149 </p>
        </item>
        <item>
          <p>When given the new option
            <c><![CDATA[strict_record_tests]]></c>, the compiler will generate
            code that verifies the record type for
            <c><![CDATA[R#record.field]]></c> operations in guards. Code that
            verifies record types in bodies has already been
            generated since R10B, but in this release there will be a
            <c><![CDATA[{badrecord,RecordTag}]]></c> instead of a
            <c><![CDATA[badmatch]]></c> if the record verification test fails.
            See <c><![CDATA[compile(3)]]></c> for more information.</p>
          <p>The Erlang shell always applies strict record tests.</p>
          <p>Own Id: OTP-5915 Aux Id: OTP-5714 </p>
        </item>
        <item>
          <p>The BIF <c><![CDATA[is_record/3]]></c> can now be used in guards.
            Also, <c><![CDATA[is_record/3]]></c> can now be called without an
            <c><![CDATA[erlang:]]></c> module prefix for consistency with the other
            <c><![CDATA[is_*]]></c> functions.</p>
          <p>Own Id: OTP-5916</p>
        </item>
        <item>
          <p>The compiler options <c><![CDATA[ignore_try]]></c> and
            <c><![CDATA[ignore_cond]]></c>, which allowed code that used
            unquoted <c><![CDATA[try]]></c> or <c><![CDATA[cond]]></c> as atoms or record
            tags, has been removed. Old code that depended on the
            options need to be revised to have occurrences of
            <c><![CDATA[try]]></c> or <c><![CDATA[cond]]></c> as atom or record tags
            single-quoted. (Note: Although <c><![CDATA[cond]]></c> is a reserved
            keyword, there is no <c><![CDATA[cond]]></c> statement. It might be
            introduced in a future release.)</p>
          <p>*** POTENTIAL INCOMPATIBILITY ***</p>
          <p>Own Id: OTP-6058</p>
        </item>
      </list>
    </section>
  </section>

  <section>
    <title>Compiler 4.3.12</title>

    <section>
      <title>Improvements and New Features</title>
      <list type="bulleted">
        <item>
          <p>The following code would crash the compiler: <c><![CDATA[case T of #r{s = ""} -> T #r{s = "x"} end]]></c>. (Thanks to
            Richard Carlsson.)</p>
          <p>The compiler could crash if binaries were constructed
            in certain guards involving boolean operators (including
            semicolon). (Thanks to Torbjorn Tornkvist.)</p>
          <p>Own Id: OTP-5872</p>
        </item>
        <item>
          <p>The compiler will now warn that the
            <c><![CDATA[megaco:format_versions/1]]></c> function is deprecated.</p>
          <p>Own Id: OTP-5976</p>
        </item>
      </list>
    </section>
  </section>

  <section>
    <title>Compiler 4.3.11</title>

    <section>
      <title>Improvements and New Features</title>
      <list type="bulleted">
        <item>
          <p>The compiler would assume that some patterns with
            aliases ('=') would not match if they were split into
            several lines. (Thanks to Peter Nagy/Mats Cronqvist.)</p>
          <p>Minor cleanups to eliminate Dialyzer warnings.</p>
          <p>Own Id: OTP-5791 Aux Id: seq10141 </p>
        </item>
      </list>
    </section>
  </section>

  <section>
    <title>Compiler 4.3.10</title>

    <section>
      <title>Fixed Bugs and Malfunctions</title>
      <list type="bulleted">
        <item>
          <p>When given the new option
            <c><![CDATA[strict_record_tests]]></c>, the compiler will generate
            code that verifies the record type for
            <c><![CDATA[R#record.field]]></c> operations (in body context only,
            not in guards). See the documentation for the
            <c><![CDATA[compile]]></c> module for more information.</p>
          <p>The beam validator pass of the compiler could crash
            given in rare circumstances when given certain
            combinations of catches and record operations. (Thanks to
            Mats Cronqvist.)</p>
          <p>Attributes containing binaries (such as -a(&lt;&lt;1,2,3&gt;&gt;))
            would crash the compiler. (Thanks to Roger Price.)</p>
          <p>Multiple behaviours in the same module will no longer
            generate a warning, unless one or more callbacks for the
            behaviours overlap. For instance, using both the
            <c><![CDATA[application]]></c> and <c><![CDATA[supervisor]]></c> behaviours
            in the same module will NOT generate any warning, but
            using <c><![CDATA[gen_server]]></c> and <c><![CDATA[gen_fsm]]></c> will.</p>
          <p>Own Id: OTP-5714 Aux Id: seq10073 </p>
        </item>
        <item>
          <p>The pre-processor used to complain that the macro
            definition <c><![CDATA[-define(S(S), ??S).]]></c> was circular,
            which it isn't. (Thanks to Richard Carlsson.)</p>
          <p>Own Id: OTP-5777</p>
        </item>
      </list>
    </section>
  </section>

  <section>
    <title>Compiler 4.3.9</title>

    <section>
      <title>Fixed Bugs and Malfunctions</title>
      <list type="bulleted">
        <item>
          <p>Updating at least two fields of a record with a literal
            string could cause the compiler to generate dangerous
            code that could cause a crash at run-time (e.g.
            <c><![CDATA[R#r{a="abc",b=1}]]></c>). (Thanks to Mikael Karlsson.)</p>
          <p>Unecessary tests (such as a 'case' with two case
            branches that were identical) could cause the compiler to
            crash. (Thanks to Fredrik Thulin.)</p>
          <p>The validation pass of the compiler could generate an
            error for correct code when floating point operations
            were used in try/catch statements.</p>
          <p>In bit syntax construction, any field following a
            binary field would always be marked as "aligned" (which
            may or may not be correct). That would cause the hipe
            native compiler to generate incorrect code if the field
            was in fact unaligned. (Thanks to Per Gustafsson.)</p>
          <p>Some complex guard expressions (such as <c><![CDATA[A#a.b==""; A#a.b==undefined]]></c>) would crash the compiler. (Thanks
            to Sean Hinde.)</p>
          <p>Compilation speed has been increased for modules with
            many functions and/or atoms (such as modules generated by
            the Asn1 application or other code generators).</p>
          <p>Own Id: OTP-5632 Aux Id: seq10057 </p>
        </item>
      </list>
    </section>
  </section>

  <section>
    <title>Compiler 4.3.8</title>

    <section>
      <title>Fixed Bugs and Malfunctions</title>
      <list type="bulleted">
        <item>
          <p>In some circumstances, having two try/catch constructs
            following each in a function body, would cause an
            internal error to be generated (when in fact the
            generated code was correct). (Thanks to Fredrik Thulin.)</p>
          <p>Incorrect calls such as <c><![CDATA[M:42()]]></c> would crash the
            compiler. The compiler now generates a warning. (Thanks
            to Ulf Wiger.)</p>
          <p>Own Id: OTP-5553</p>
        </item>
      </list>
    </section>

    <section>
      <title>Improvements and New Features</title>
      <list type="bulleted">
        <item>
          <p>The new <c><![CDATA[fun M:F/A]]></c> construct creates a fun that
            refers to the latest version of <c><![CDATA[M:F/A]]></c>. This syntax is
            meant to replace tuple funs <c><![CDATA[{M,F}]]></c> which have many
            problems.</p>
          <p>The new type test <c><![CDATA[is_function(Fun, A)]]></c> (which may be
            used in guards) test whether <c><![CDATA[Fun]]></c> is a fun that can be
            applied with <c><![CDATA[A]]></c> arguments. (Currently, <c><![CDATA[Fun]]></c> can
            also be a tuple fun.)</p>
          <p>Own Id: OTP-5584</p>
        </item>
      </list>
    </section>
  </section>

  <section>
    <title>Compiler 4.3.7</title>

    <section>
      <title>Improvements and New Features</title>
      <list type="bulleted">
        <item>
          <p>Further improvements of encrypted debug info: New option
            <c><![CDATA[encrypt_debug_info]]></c> for compiler.</p>
          <p>Own Id: OTP-5541 Aux Id: seq9837 </p>
        </item>
      </list>
    </section>
  </section>

  <section>
    <title>Compiler 4.3.6</title>

    <section>
      <title>Fixed Bugs and Malfunctions</title>
      <list type="bulleted">
        <item>
          <p>Fixed a bug in the validator of the generated code
            (beam_validator) which caused an internal compiler error
            even though the generated code was indeed correct.</p>
          <p>Own Id: OTP-5481 Aux Id: seq9798 </p>
        </item>
      </list>
    </section>

    <section>
      <title>Improvements and New Features</title>
      <list type="bulleted">
        <item>
          <p>It is now possible to encrypt the debug information in
            Beam files, to help keep the source code secret. See the
            documentation for <c><![CDATA[compile]]></c> on how to provide the key
            for encrypting, and the documentation for <c><![CDATA[beam_lib]]></c>
            on how to provide the key for decryption so that tools such
            as the Debugger, Xref, or Cover can be used.</p>
          <p>The <c><![CDATA[beam_lib:chunks/2]]></c> functions now accepts an
            additional chunk type <c><![CDATA[compile_info]]></c> to retrieve
            the compilation information directly as a term. (Thanks
            to Tobias Lindahl.)</p>
          <p>Own Id: OTP-5460 Aux Id: seq9787 </p>
        </item>
      </list>
    </section>
  </section>

  <section>
    <title>Compiler 4.3.5</title>

    <section>
      <title>Fixed Bugs and Malfunctions</title>
      <list type="bulleted">
        <item>
          <p>Complex functions could cause the internal validator in
            the compiler to generate an internal error even though
            the generated code was correct.</p>
          <p>Own Id: OTP-5436 Aux Id: seq9781 </p>
        </item>
      </list>
    </section>
  </section>

  <section>
    <title>Compiler 4.3.4</title>

    <section>
      <title>Fixed Bugs and Malfunctions</title>
      <list type="bulleted">
        <item>
          <p>In rare circumstances, incorrect code for record or
            tuple access could be generated. The incorrect code would
            either trigger an internal error in the compiler or cause
            an exception at run time. (Thanks to Martin Bjorklund.)</p>
          <p>Corrected a bug in in bit syntax matching where
            clauses could match in the wrong order. (Thanks to Ulf
            Wiger.)</p>
          <p>Own Id: OTP-5404 Aux Id: seq9767 </p>
        </item>
      </list>
    </section>
  </section>

  <section>
    <title>Compiler 4.3.3</title>

    <section>
      <title>Improvements and New Features</title>
      <list type="bulleted">
        <item>
          <p>Given bit syntax construction in certain complex
            contexts involving a catch, the compiler would either
            crash or terminate due to failure in an internal
            consistency check. (Thanks to Fredrik Thulin.)</p>
          <p>Matches such as
            <c><![CDATA[<<103133:64/float>> = <<103133:64/float>>]]></c>
            used to fail. Now they succeed.</p>
          <p>Shadowing of variables in bit syntax matches in fun heads
            such as in
            <c><![CDATA[L = 8, F = fun(<<L:L,B:L>>) -> B end]]></c> was
            handled incorrectly by the compiler. The fun used to be
            compiled as if it was written
            '<c><![CDATA[>fun(<<8:8,B:8>>)]]></c>, while it should be
            compiled in the same way as <c><![CDATA[fun(<<L:8,B:L>>)]]></c>.</p>
          <p>A bug in the validation pass has been corrected. It
            sometimes occurred when the compiler optimized by reusing
            code for causing an exception when the reused code was
            called from within catch or try-catch statements. Then the
            validator refused to approve the code and complained about
            <c><![CDATA[fun(<<L:L,B:L>>) -> B end]]></c> was handled
            incorrectly by the in the same way as
            <c><![CDATA[fun(<<L:8,B:L>>)]]></c>.</p>
          <p>A bug in the unknown_catch_try_state.</p>
          <p>Corrected a bug in the optimizer that would cause
            the compiler to crash. (Thanks to Peter-Henry Mander.)</p>
          <p>There are now warnings generated if a bit syntax
            construction will fail at run-time because of a type
            mismatch (e.g. <c><![CDATA[<<an_atom:8>>]]></c>).</p>
          <p>Own Id: OTP-5342 Aux Id: OTP-5118, OTP-5270, OTP-5323 </p>
        </item>
        <item>
          <p>Binary pattern matching such as
            <c><![CDATA[t(<<A:8>> = <<A:8>)]]></c> used to silently
            fail at runtime (i.e. never match). The compiler now
            generates an error for any such patterns.</p>
          <p>Own Id: OTP-5371</p>
        </item>
      </list>
    </section>
  </section>

  <section>
    <title>Compiler 4.3.2</title>

    <section>
      <title>Fixed Bugs and Malfunctions</title>
      <list type="bulleted">
        <item>
          <p>In rare cases, the code compiler code generate code
            for a tuple match that could crash the emulator if passed
            a term that was not a tuple.</p>
          <p>If a bit syntax construction failed within a catch,
            previously assigned variables could get the wrong value.</p>
          <p>The compiler now runs a validation pass on the
            generated code and aborts before writing a Beam file if
            any suspect code is found. In particular, the validation
            pass checks for incorrect code that may cause emulator
            crashes or other strange symptoms in the emulator.</p>
          <p>Some corrections to the unsupported feature
            parameterized modules by Richard Carlsson (HiPE).</p>
          <p>Own Id: OTP-5247 Aux Id: OTP-5235 </p>
        </item>
      </list>
    </section>
  </section>

  <section>
    <title>Compiler 4.3.1</title>

    <section>
      <title>Fixed Bugs and Malfunctions</title>
      <list type="bulleted">
        <item>
          <p>Corrected the release note regarding <c><![CDATA[try/catch]]></c> below.
            <c><![CDATA[try/catch]]></c> DOES work in the initial R10B release.</p>
          <p>A few minor issues code generation issues were corrected.
            Although the generated code was correct, it was slightly
            slower and larger than it needed to be.</p>
          <p>A debug printout (that could be seen in rare
            circumstances) has been removed.</p>
          <p><c><![CDATA[not record_test(not_a_tuple, RecordTag)]]></c> and
            similar expressions in a guard would fail.</p>
          <p>New options <c><![CDATA[basic_validation]]></c> and
            <c><![CDATA[strong_validation]]></c> to do a quick check of the code
            of a module.</p>
          <p>The <c><![CDATA[inline]]></c> option was not recognized if it
            appeared in a <c><![CDATA[-compile()]]></c> directive inside the
            module.</p>
          <p>Corrected some bugs in the undocumented feature
            "parameterized modules".</p>
          <p>Own Id: OTP-5198</p>
        </item>
        <item>
          <p>When the undocumented feature "parameterized modules" was
            used, the <c><![CDATA[?MODULE]]></c> macro did not work correctly.</p>
          <p>Own Id: OTP-5224</p>
        </item>
      </list>
    </section>
  </section>
</chapter>