|
Server IP : 161.35.85.9 / Your IP : 216.73.217.134 Web Server : nginx/1.30.1 System : Linux cheapdeb-ams3 7.0.9+deb14-amd64 #1 SMP PREEMPT_DYNAMIC Debian 7.0.9-1 (2026-05-22) x86_64 User : root ( 0) PHP Version : 8.2.18 Disable Function : NONE MySQL : OFF | cURL : ON | WGET : ON | Perl : ON | Python : ON Directory (0755) : /usr/share/doc/python3-ndg-httpsclient/../cpp-15/ |
| [ Home ] | [ C0mmand ] | [ Upload File ] |
|---|
<!DOCTYPE html>
<html lang="en">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
<title>
GCC 15 Release Series — Changes, New Features, and Fixes
- GNU Project</title>
<link rel="stylesheet" type="text/css" href="gcc.css">
</head>
<!-- GCC maintainers, please do not hesitate to contribute/update
entries concerning those part of GCC you maintain!
-->
<body>
<h1>GCC 15 Release Series<br>Changes, New Features, and Fixes</h1>
<p>
This page is a "brief" summary of some of the huge number of improvements
in GCC 15.
You may also want to check out our
<a href="porting_to.html">Porting to GCC 15</a> page and the
<a href="../onlinedocs/index.html#current">full GCC documentation</a>.
</p>
<!-- .................................................................. -->
<h2>Caveats</h2>
<ul>
<li>Support for Nios II targets, which was marked obsolete in GCC 14,
has now been removed entirely.
</li>
<li>In the AArch64 port, support for ILP32 (<code>-mabi=ilp32</code>) has
been deprecated and will be removed in a future release.
</li>
<li><code>{0}</code> initializer in C or C++ for unions no longer
guarantees clearing of the whole union (except for static storage
duration initialization), it just initializes the first
union member to zero. If initialization of the whole union including
padding bits is desirable, use <code>{}</code> (valid in C23 or C++)
or use <code>-fzero-init-padding-bits=unions</code> option to restore
old GCC behavior.</li>
</ul>
<!-- .................................................................. -->
<h2 id="general"><a href="#general">General Improvements</a></h2>
<ul>
<li>The default vectorizer cost model at <code>-O2</code> has been enhanced
to handle unknown tripcount. But it still disables vectorization of loops
when any runtime check for data dependence or alignment is required,
it also disables vectorization of epilogue loops but otherwise is equal
to the cheap cost model.
</li>
<li>The vectorizer now supports vectorizing loops with early exits where
the number of elements for the input pointers are unknown through peeling
for alignment. This is supported for only for loops with fixed vector
lengths.
</li>
<li><code>-ftime-report</code> now only reports monotonic run time instead of
system and user time. This reduces the overhead of the option significantly,
making it possible to use in standard build systems.
</li>
<li>Incremental Link-Time Optimizations significantly reduce average
recompilation time of LTO when doing small code edits
(e.g. editing a single function).
Enable with <a
href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/Optimize-Options.html#index-flto-incremental"
><code>-flto-incremental=</code></a>.
</li>
<li>
For offloading using OpenMP and OpenACC, issues preventing some
host–device architecture combinations have been resolved. In particular,
offloading from aarch64 hosts to nvptx devices is now supported. Additionally,
the support for using C++ in offload regions has been extended.
</li>
<li>
Improvements for compiling very large input files. The compile time for
large input files with <code>-Wmisleading-indentation</code> has been
significantly improved. The compiler can now track columnn numbers larger
than 4096. Very large source files have more accurate location reporting.
</li>
</ul>
<!-- .................................................................. -->
<h2 id="languages"><a href="#languages">New Languages and Language specific improvements</a></h2>
<h3 id="openmp"><a href="#openmp">OpenMP</a></h3>
<p>
See the
<a href="../projects/gomp/">GNU Offloading and Multi-Processing Project (GOMP)</a>
page for general information.
</p>
<ul>
<li>
Support for unified-shared memory has been added for some AMD and Nvidia
GPU devices, enabled when using the <code>unified_shared_memory</code>
clause to the <code>requires</code> directive.
The OpenMP 6.0 <code>self_maps</code> clause is also now supported.
For details,
see the offload-target specifics section in the
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/libgomp/Offload-Target-Specifics.html"
>GNU Offloading and Multi Processing Runtime Library Manual</a>.
</li>
<li>
GCC added <code>ompx_gnu_pinned_mem_alloc</code> as a <a
href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/libgomp/OMP_005fALLOCATOR.html">predefined
allocator</a> and, for C++, allocator class templates in the
<code>omp::allocator</code> namespace for the predefined allocators as
specified in the OpenMP specification 5.0, including
<code>omp::allocator::null_allocator</code> of OpenMP 6.0 and
<code>ompx::allocator::gnu_pinned_mem</code>; the allocator templates
can be used with C++ containers such as <code>std::vector</code>.
</li>
<li>
In C and Fortran, the <code>allocate</code> directive now supports
static variables; stack variables were previously supported in
those languages. C++ support is not available yet.
</li>
<li>
Offloading improvements:
On <a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/libgomp/nvptx.html">Nvidia
GPUs, writing to the terminal</a> from OpenMP target regions (but not from
OpenACC compute regions) is now also supported in Fortran; in C/C++ and
on AMD GPUs this was already supported before with both OpenMP and OpenACC.
Constructors and destructors on the device side for
<code>declare target</code> static aggregates are now handled.
</li>
<li>For Fortran, mapping derived-type variables with allocatable components
is now supported.</li>
<li>
The OpenMP 5.1 <code>unroll</code> and <code>tile</code>
loop-transforming constructs are now supported.
</li>
<li>OpenMP 5.0 metadirectives are now supported, as are OpenMP 5.1 dynamic
selectors in both <code>metadirective</code> and
<code>declare variant</code> (the latter with some restrictions).
</li>
<li>
The <code>interop</code> construct and the OpenMP interoperability API
routines for C, C++ and Fortran are now implemented, including the
OpenMP 6.0 additions. This includes foreign-runtime support for <a
href= "https://gcc.gnu.org/onlinedocs/gcc-15.1.0/libgomp/Foreign-runtime-support-for-Nvidia-GPUs.html"
>Cuda, Cuda Driver, and HIP on Nvida GPUs</a> and for <a
href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/libgomp/Foreign-runtime-support-for-AMD-GPUs.html"
>HIP and HSA on AMD GPUs</a>.
</li>
<li>
The OpenMP 5.1 <code>dispatch</code> construct has been implemented
with support for the <code>adjust_args</code> and <code>append_args</code>
clauses to the <code>declare variant</code> directive, including the
following OpenMP 6.0 additions: the <code>interop</code> clause to
<code>dispatch</code> and the syntax extensions to <code>append_args</code>
are supported.
</li>
<li>
OpenMP 6.0: The <a
href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/libgomp/omp_005fget_005fdevice_005ffrom_005fuid.html"
><code>get_device_from_uid</code></a> and <a
href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/libgomp/omp_005fget_005fuid_005ffrom_005fdevice.html">
<code>omp_get_uid_from_device</code></a> API routines have been added.
</li>
</ul>
<h3 id="cobol"><a href="#cobol">COBOL</a></h3>
<ul>
<li>GCC now includes an ISO COBOL compiler, gcobol. It has been
tested on x86-64 and AArch64 targets. It is not expected to work
on 32-bit systems. Efforts are underway to adapt it to other machine
architectures that support native 128-bit computation.</li>
<li>gcobol passes much of the NIST CCVS/85 test suite (except for parts
that are now obsolete). It uses ISO/IEC 1989:2023 as a
reference specification. Some parts of that document, notably
object-orientation features, are yet to be implemented. Beyond
ISO, gcobol recognizes some syntax common on other compilers,
with special attention given to IBM.
<p>More information about GCC COBOL can be found at
<a href="https://www.cobolworx.com">the COBOLworx website</a>.</li>
</ul>
<h3 id="ada"><a href="#ada">Ada</a></h3>
<ul>
<li>
GNAT now allows
the <a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gnat_rm/Attribute-Round.html#index-Round"><code>'Round</code></a>
attribute also for ordinary fixed-point types.
</li>
<li>
The new GNAT
attribute <a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gnat_rm/Attribute-Super.html#index-Super"><code>'Super</code></a>
can be applied to objects of tagged types in order to obtain a view
conversion to the most immediate specific parent type.
</li>
<li>
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gnat_rm/Mutably-Tagged-Types-with-Size_2019Class-Aspect.html">Mutably tagged types</a> with a defined size are now available
through the use of <code>Size'Class</code>. This allows defining a maximum
size for the tagged. Example:
<pre>
type Base is tagged null record with Size'Class => 16 * 8;
-- Size in bits (128 bits, or 16 bytes)
type Derived_Type is new Base with record Data_Field : Integer; end record;
-- ERROR if Derived_Type exceeds 16 bytes
</pre>
</li>
<li>
New <a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gnat_rm/Generalized-Finalization.html"><code>Finalizable</code>
</a> aspect. It is a GNAT language extension which serves as a lightweight
alternative to controlled types.
<pre>
type T is record
...
end record with Finalizable => (Initialize => Initialize,
Adjust => Adjust,
Finalize => Finalize,
Relaxed_Finalization => True);
procedure Adjust (Obj : in out T);
procedure Finalize (Obj : in out T);
procedure Initialize (Obj : in out T);
</pre>
</li>
<li>
The
aspect <a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gnat_rm/No_005fRaise-aspect.html"><code>No_Raise</code></a>
has been added, it declares that a subprogram cannot raise an exception.
</li>
<li>
The
aspect <a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gnat_rm/External_005fInitialization-Aspect.html"><code>External_Initialization</code></a>
has been added, it allows for data to be initialized using an external file
which is loaded during compilation time.
</li>
<li>
The
aspect <a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gnat_rm/Pragma-Exit_005fCases.html#index-Exit_005fCases"><code>Exit_Cases</code></a>
has been added to annotate functions and procedures with side effects in
SPARK
(see <a href="https://docs.adacore.com/spark2014-docs/html/lrm/subprograms.html#program-exit-aspects">SPARK
reference manual</a>) . It can be used to partition the input state into a
list of cases and specify, for each case, how the subprogram is allowed to
terminate.
</li>
<li>
Language extensions are enabled through the use of <code>pragma
Extensions_Allowed (On | Off | All_Extensions);</code> which has had its
syntax changed. An argument of <code>All_Extensions</code> has the same
effect as <code>On</code>, except
that <a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gnat_rm/Experimental-Language-Extensions.html">some
extra experimental extensions are enabled</a>.
</li>
<li>
Several new compilation flags have been added, some examples
include <a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gnat_ugn/Info-message-Control.html"><code>-gnatis</code></a>, <a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gnat_ugn/Warning-Message-Control.html#index--gnatw_002en-_0028gcc_0029"><code>-gnatw.n</code></a>, <a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gnat_ugn/Warning-Message-Control.html#index--gnatw_005fl-_0028gcc_0029"><code>-gnatw_l</code></a>
and <a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gnat_ugn/Warning-Message-Control.html#index--gnatw_002ev-_0028gcc_0029"><code>-gnatw.v</code></a>.
The internal debugging utilities for the compiler have also received a lot
of new options, please refer
to <a href="https://gcc.gnu.org/git/?p=gcc.git;a=blob;f=gcc/ada/debug.adb"><code>debug.adb</code></a>
for more information.
</li>
<li>
The diagnostics code has seen a major refactor, it now supports the sarif
format <code>-fdiagnostics-format=sarif-file</code> among other
improvements. More changes are expected in following releases.
</li>
<li>
<code>System.Image_A</code> has now printing routines to output address
information in HEX.
</li>
<li>
Several program units have had contracts added to them and SPARK analysis
has been enabled.
</li>
<li>
Support for FreeBSD, Android and aarch64 targets has been improved.
</li>
<li>
Task priorities on MinGW have been reworked.
</li>
<li>
The documentation has been rearanged for clarity, mainly the sections
related to tasking and platform dependent information.
</li>
<li>
REMOVAL: <code>Generic_Formal_Hash_Table</code> has been removed,
the <a href="https://docs.adacore.com/spark2014-docs/html/ug/en/source/spark_libraries.html#">SPARK
Library</a> is recommended as a substitute.
</li>
</ul>
<h3 id="c-family"><a href="#c-family">C family</a></h3>
<ul>
<li>A <a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/Statement-Attributes.html#index-musttail-statement-attribute">
<code>musttail</code> statement attribute</a> was added to enforce tail calls.</li>
<li><a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/Extended-Asm.html"
>Extended</a> inline assembler statements
can now be used with some limitations outside of functions as well.
New constraints have been added for defining symbols or using symbols
inside of inline assembler, and a new generic operand modifier has
been added to allow printing those regardless of PIC. For example:
<pre>
struct S { int a, b, c; };
extern foo (void);
extern char var;
int var2;
asm (".text; %cc0: mov %cc2, %%r0; .previous;"
".rodata: %cc1: .byte %3; .previous" : :
":" (foo), /* Tell compiler asm defines foo function. */
":" (&var), /* Tell compiler asm defines var variable. */
"-s" (var2), /* Tell compiler asm uses var2 variable. */
/* "s" would work too but might not work with -fpic. */
"i" (sizeof (struct S))); /* It is possible to pass constants to toplevel asm. */
</pre>
</li>
<li>The <code>"redzone"</code> clobber is now allowed in inline
assembler statements to describe that the assembler can overwrite
memory in the stack red zone (e.g. on x86-64 or PowerPC).</li>
<li>The <a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/Common-Function-Attributes.html#index-nonnull_005fif_005fnonzero-function-attribute">
nonnull_if_nonzero</a> function attribute has been added to describe
functions where some pointer parameter may be <code>NULL</code> only
if some other parameter is zero.</li>
<li>The <a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/Warning-Options.html#index-Wtrailing-whitespace_003d">
<code>-Wtrailing-whitespace=</code></a> and
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/Warning-Options.html#index-Wleading-whitespace_003d">
<code>-Wleading-whitespace=</code></a> options have been added to
diagnose certain whitespace characters at the end of source lines or
whitespace characters at the start of source lines violating certain
indentation styles.</li>
<li>The <a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/Warning-Options.html#index-Wheader-guard">
<code>-Wheader-guard</code></a> warning has been added and enabled
in <code>-Wall</code> to warn about some inconsistencies in header
file guarding macros.</li>
</ul>
<h3 id="c"><a href="#c">C</a></h3>
<ul>
<li>C23 by default: GCC 15 changes the default language version
for C compilation from
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/C-Dialect-Options.html#index-std-1">-std=gnu17</a>
to
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/C-Dialect-Options.html#index-std-1">-std=gnu23</a>.
If your code relies on older versions of the C standard, you will need to
either add
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/C-Dialect-Options.html#index-std-1">-std=</a>
to your build flags, or port your code; see <a href="porting_to.html#c23">the porting notes</a>.
</li>
<li>Some more C23 features have been implemented:
<ul>
<li><code>#embed</code> preprocessing directive support.</li>
<li>Support for <code>unsequenced</code> and <code>reproducible</code>
attributes.</li>
<li><code>__STDC_VERSION__</code> predefined macro value changed
for <code>-std=c23</code> or <code>-std=gnu23</code> to
<code>202311L</code>.</li>
</ul>
</li>
<li>Some new features from the upcoming C2Y revision of the ISO C
standard are supported with <code>-std=c2y</code>
and <code>-std=gnu2y</code>. Some of these features are also
supported as extensions when compiling for older language versions.
<ul>
<li>Generic selection expression with a type operand.</li>
<li>Support <code>++</code> and <code>--</code> on complex values.</li>
<li>Accessing byte arrays.</li>
<li><code>alignof</code> of an incomplete array type.</li>
<li>Obsolete implicitly octal literals and add delimited escape
sequences (just partially implemented, support for new syntax added
but nothing deprecated yet).</li>
<li>Named loops.</li>
<li>More Modern Bit Utilities (addition of
<code>__builtin_stdc_rotate_left</code> and
<code>__builtin_stdc_rotate_right</code> builtins for use in future
C library <code><stdbit.h></code> headers).</li>
<li>Case range expressions.</li>
<li><code>if</code> declarations.</li>
<li>Introduce complex literals.</li>
<li>Abs Without Undefined Behavior (addition of builtins for
use in future C library <code><stdlib.h></code> headers).</li>
<li>Allow zero length operations on null pointers (just the compiler
side, C library headers will need adjustments too if using
<code>nonnull</code> attribute).</li>
</ul>
</li>
</ul>
<h3 id="cxx"><a href="#cxx">C++</a></h3>
<ul>
<li>Several C++26 features have been implemented:
<ul>
<li><a href="https://wg21.link/P2558R2">P2558R2</a>, Add @, $, and ` to the basic
character set (<a href="https://gcc.gnu.org/PR110343">PR110343</a>)
</li>
<li><a href="https://wg21.link/P2552R3">P2552R3</a>, On the ignorability of
standard attributes (<a href="https://gcc.gnu.org/PR110345">PR110345</a>)
</li>
<li><a href="https://wg21.link/P2662R3">P2662R3</a>, Pack indexing
(<a href="https://gcc.gnu.org/PR113798">PR113798</a>)
</li>
<li><a href="https://wg21.link/P0609R3">P0609R3</a>, Attributes for structured
bindings (<a href="https://gcc.gnu.org/PR114456">PR114456</a>)
</li>
<li><a href="https://wg21.link/P2573R2">P2573R2</a>,
<code>= delete("reason");</code> (<a href="https://gcc.gnu.org/PR114458">PR114458</a>)
</li>
<li><a href="https://wg21.link/P2893R3">P2893R3</a>, Variadic friends
(<a href="https://gcc.gnu.org/PR114459">PR114459</a>)
</li>
<li><a href="https://wg21.link/P3034R1">P3034R1</a>, Disallow module declarations
to be macros (<a href="https://gcc.gnu.org/PR114461">PR114461</a>)
</li>
<li><a href="https://wg21.link/P2747R2">P2747R2</a>, <code>constexpr</code>
placement new (<a href="https://gcc.gnu.org/PR115744">PR115744</a>)
</li>
<li><a href="https://wg21.link/P0963R3">P0963R3</a>, Structured binding declaration
as a condition (<a href="https://gcc.gnu.org/PR115745">PR115745</a>)
</li>
<li><a href="https://wg21.link/P3144R2">P3144R2</a>, Deleting a pointer to an
incomplete type should be ill-formed
(<a href="https://gcc.gnu.org/PR115747">PR115747</a>)
</li>
<li><a href="https://wg21.link/P3176R0">P3176R0</a>, Oxford variadic comma
(<a href="https://gcc.gnu.org/PR117786">PR117786</a>)
</li>
<li><a href="https://wg21.link/P2865R5">P2865R5</a>, Removing deprecated array
comparisons (<a href="https://gcc.gnu.org/PR117788">PR117788</a>)
</li>
<li><a href="https://wg21.link/P1967R14">P1967R14</a>, <code>#embed</code>
(<a href="https://gcc.gnu.org/PR119065">PR119065</a>)
</li>
<li><a href="https://wg21.link/P3247R2">P3247R2</a>, Deprecating the notion of
trivial types (<a href="https://gcc.gnu.org/PR117787">PR117787</a>)
</li>
</ul>
</li>
<li>Several C++23 features have been implemented:
<ul>
<li><a href="https://wg21.link/P2615R1">P2615R1</a>, Meaningful exports
(<a href="https://gcc.gnu.org/PR107688">PR107688</a>)
</li>
<li><a href="https://wg21.link/P2718R0">P2718R0</a>, Wording for P2644R1
Fix for Range-based for Loop
(<a href="https://gcc.gnu.org/PR107637">PR107637</a>)
</li>
</ul>
</li>
<li>Several C++ Defect Reports have been resolved, e.g.:
<ul>
<li><a href="https://wg21.link/cwg36">DR 36</a>,
<em>using-declarations</em> in multiple-declaration contexts</li>
<li><a href="https://wg21.link/cwg882">DR 882</a>,
Defining main as deleted</li>
<li><a href="https://wg21.link/cwg1363">DR 1363</a>,
Triviality vs multiple default constructors</li>
<li><a href="https://wg21.link/cwg1496">DR 1496</a>,
Triviality with deleted and missing default constructors</li>
<li><a href="https://wg21.link/cwg2387">DR 2387</a>,
Linkage of const-qualified variable template</li>
<li><a href="https://wg21.link/cwg2521">DR 2521</a>,
User-defined literals and reserved identifiers</li>
<li><a href="https://wg21.link/cwg2627">DR 2627</a>,
Bit-fields and narrowing conversions</li>
<li><a href="https://wg21.link/cwg2819">DR 2819</a>,
Cast from null pointer value in a constant expression (C++26 only)</li>
<li><a href="https://wg21.link/cwg2867">DR 2867</a>,
Order of initialization for structured bindings</li>
<li><a href="https://wg21.link/cwg2918">DR 2918</a>,
Consideration of constraints for address of overloaded function</li>
</ul>
</li>
<li>Inline assembler statements now support
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/Asm-constexprs.html"><code>constexpr</code> generated strings</a>,
analoguous to <code>static_assert</code>.</li>
<li>
<!-- commit r15-2117-g313afcfdabeab3 -->
Qualified name lookup failure into the current instantiation, e.g.
<code>this->non_existent</code>, is now proactively diagnosed
when parsing a template.
</li>
<li>The <a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/C_002b_002b-Dialect-Options.html#index-fassume-sane-operators-new-delete">
<code>-fassume-sane-operators-new-delete</code></a> option has been
added and enabled by default. This option allows control over some
optimizations around calls to replaceable global operators new and
delete. If a program overrides those replaceable global operators and
the replaced definitions read or modify global state visible to the
rest of the program, programs might need to be compiled with
<code>-fno-assume-sane-operators-new-delete</code>.</li>
<li>
<!-- commit r15-871-gefaaae49b307fc -->
The <a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/Warning-Options.html#index-Wself-move">
<code>-Wself-move</code></a> warning now warns even in a
member-initializer-list.
</li>
<li>
<!-- commit r15-1953-gf0fb6b6acd805c -->
The support for Concepts TS was removed. <code>-fconcepts-ts</code> has no
effect anymore.
</li>
<li>
<!-- commit r15-2774-g596d1ed9d40b10 -->
A new option
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/C_002b_002b-Dialect-Options.html#index-Wtemplate-body">
<code>-Wtemplate-body</code></a> was added, which can be used to disable
diagnosing errors when parsing a template.
</li>
<li>
C++ Modules have been greatly improved.
</li>
<li>
<!-- commit r15-3433-g3775f71c8909b3 -->
C++11 attributes are now supported even in C++98.
</li>
<li>
New
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/Common-Type-Attributes.html#index-flag_005fenum-type-attribute">
<code>flag_enum</code></a> attribute to indicate that the enumerators
are used in bitwise operations; this suppresses a <code>-Wswitch</code>
warning.
</li>
<li>
The <a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/C_002b_002b-Dialect-Options.html#index-Wdangling-reference">
<code>-Wdangling-reference</code></a> warning has been improved: for
example, it doesn't warn for empty classes anymore.
</li>
<li>
The front end's handling of explicitly-defaulted functions has been
corrected to properly handle <em>[dcl.fct.def.default]</em>. The
new
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/C_002b_002b-Dialect-Options.html#index-Wdefaulted-function-deleted">
<code>-Wdefaulted-function-deleted</code></a> warning warns when an
explicitly defaulted function is deleted.
</li>
<li>
The implementation of <a href="https://wg21.link/cwg2789">DR 2789</a>
was refined.
</li>
<li>
<!-- commit r15-4050-g5dad738c1dd164 -->
Compilation time speed ups, e.g. by improving hashing of template
specializations.
</li>
<li>
<!-- commit r15-5109-g417b4cc9bf2180 -->
Support for <code>__builtin_operator_new</code> and
<code>__builtin_operator_delete</code> was added. See
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/New_002fDelete-Builtins.html">
the manual</a> for more info.
</li>
<li>
<!-- commit r15-6052-g12de1942a0a673 -->
More prvalues are evaluated at compile time
(<a href="https://gcc.gnu.org/git/?p=gcc.git;a=commitdiff;h=12de1942a0a673f9f2f1c2bfce4279a666061ffc">git</a>).
</li>
<li>
Various diagnostic improvements.
</li>
</ul>
<h4 id="libstdcxx"><a href="#libstdcxx">Runtime Library (libstdc++)</a></h4>
<ul>
<li>Debug assertions are now enabled by default for unoptimized builds.
Use <code>-D_GLIBCXX_NO_ASSERTIONS</code> to override this.
</li>
<li>Improved experimental support for C++26, including:
<ul>
<li><code>views::concat</code>.</li>
<li>Member <code>visit</code>.</li>
<li>Type-checking <code>std::format</code> args.</li>
</ul>
</li>
<li>Improved experimental support for C++23, including:
<ul>
<li>
<code>std</code> and <code>std.compat</code> modules
(also supported for C++20).
</li>
<li>
<code>std::flat_map</code> and <code>std::flat_set</code>.
</li>
<li>
Clarify handling of encodings in localized formatting of chrono types.
</li>
</ul>
</li>
</ul>
<h3 id="d"><a href="#d">D</a></h3>
<ul>
<li>Support for the D programming language has been updated to version
2.111.0 of the language and run-time library. Full changelog for this
release and previous releases can be found on the
<a href="https://dlang.org/changelog/2.111.0.html">dlang.org</a>
website.
</li>
<li>On supported targets, the version <code>GNU_CET</code> is now predefined
when the option <code>-fcf-protection</code> is used. The protection level
is also set in the traits key <code>__traits(getTargetInfo, "CET")</code>.
</li>
<li>A new option
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gdc/Runtime-Options.html#index-finclude-imports">
<code>-finclude-imports</code></a> was added, which tells the compiler to
include imported modules in the compilation, as if they were given on the
command-line.
</li>
</ul>
<h3 id="fortran"><a href="#fortran">Fortran</a></h3>
<ul>
<li>
Fortran 2018 and 2023 locality specifiers to <code>do concurrent</code> are
now supported.
</li>
<li>
Experimental support for <code>unsigned</code> modular integers,
enabled by <code>-funsigned</code>;
see <a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gfortran/Unsigned-integers.html">
gfortran documentation</a> for details. This follows
(<a href="https://j3-fortran.org/doc/year/24/24-116.txt">J3/24-116</a>).
With this option in force, the <code>selected_logical_kind</code>
intrinsic function and, in the <code>ISO_FORTRAN_ENV</code>
module, the named constants <code>logical{8,16,32,64}</code> and
<code>real16</code> were added. The <code>ISO_C_BINDING</code>
module has been extended accordingly.
</li>
<li>Missing commas separating descriptors in input/output format strings are no
longer permitted by default and are rejected at run-time unless -std=legacy
is used when compiling the main program unit. See Fortran 2023 constraint C1302.
</li>
<li>The Fortran module <code>*.mod</code> format generated by GCC 15 is
incompatible with the module format generated by GCC 8 - 14, but GCC
15 can for compatibility still read GCC 8 - 14 created module
files.</li>
<li>Coarray support has been reworked to allow access to components in derived
types that have not been compiled with coarray support enabled;
especially, when the derived type is in a binary only module. This has
changed the ABI and may lead to link-time errors with object files
generated with a previous GCC version and to be linked to the current
<code>caf_single</code> library. If this library is to be used, then
it is recommended to recompile all artifacts. The OpenCoarrays library
is not affected, because it provides backwards compatibility with the
older ABI.</li>
<li>
The <code>-Wexternal-interface-mismatch</code> option has been
added. This checks for mismatches between the argument lists in
dummy external arguments, and is implied by <code>-Wall</code>
and <code>-fc-prototypes-external</code> options.
</li>
<li>
The <code>-fc-prototypes</code> now also generates prototypes for
interoperable procedures with assumed shape and assumed rank
arguments that require the header file
<code><ISO_Fortran_binding.h></code>.
</li>
</ul>
<!-- <h3 id="go"><a href="#go">Go</a></h3> -->
<!-- .................................................................. -->
<!-- <h2 id="jit"><a href="#jit">libgccjit</a></h2> -->
<h3 id="modula2"><a href="#modula2">Modula-2</a></h3>
<ul>
<li>The keyword <code>FORWARD</code> has been implemented in the
compiler and is available by default in all dialects.
</li>
<li>The <code>SYSTEM</code> module now exports the
datatype <code>COFF_T</code> mapping onto the POSIX <code>off_t</code>
type. The size of this datatype can be controlled by the new
option <code>-fm2-file-offset-bits=</code>.
</li>
<li>Access to the GCC
builtins <code>clz</code>, <code>clzll</code>, <code>ctz</code>
and <code>ctzll</code> are now available from the
module <code>Builtins</code>.
</li>
</ul>
<!-- .................................................................. -->
<h2 id="targets"><a href="#targets">New Targets and Target Specific Improvements</a></h2>
<h3 id="aarch64"><a href="#aarch64">AArch64</a></h3>
<ul>
<li>Support has been added for the AArch64 MinGW target
(<code>aarch64-w64-mingw32</code>). At present, this target
supports C and C++ for base Armv8-A, but with some caveats:
<ul>
<li>Although most variadic functions work, the implementation
of them is not yet complete.
</li>
<li>C++ exception handling is not yet implemented.</li>
</ul>
Further work is planned for GCC 16.
</li>
<li>As noted above, support for ILP32 (<code>-mabi=ilp32</code>)
has been deprecated and will be removed in a future release.
<code>aarch64*-elf</code> targets no longer build the ILP32 multilibs.
</li>
<li>The following architecture level is now supported by
<code>-march</code> and related source-level constructs
(GCC identifiers in parentheses):
<ul>
<li>Armv9.5-A (<code>arm9.5-a</code>)</li>
</ul>
</li>
<li>The following CPUs are now supported by <code>-mcpu</code>,
<code>-mtune</code>, and related source-level constructs
(GCC identifiers in parentheses):
<ul>
<li>Apple A12 (<code>apple-a12</code>)</li>
<li>Apple M1 (<code>apple-m1</code>)</li>
<li>Apple M2 (<code>apple-m2</code>)</li>
<li>Apple M3 (<code>apple-m3</code>)</li>
<li>Arm Cortex-A520AE (<code>cortex-a520ae</code>)</li>
<li>Arm Cortex-A720AE (<code>cortex-a720ae</code>)</li>
<li>Arm Cortex-A725 (<code>cortex-a725</code>)</li>
<li>Arm Cortex-R82AE (<code>cortex-r82ae</code>)</li>
<li>Arm Cortex-X925 (<code>cortex-x925</code>)</li>
<li>Arm Neoverse N3 (<code>neoverse-n3</code>)</li>
<li>Arm Neoverse V3 (<code>neoverse-v3</code>)</li>
<li>Arm Neoverse V3AE (<code>neoverse-v3ae</code>)</li>
<li>FUJITSU-MONAKA (<code>fujitsu-monaka</code>)</li>
<li>NVIDIA Grace (<code>grace</code>)</li>
<li>NVIDIA Olympus (<code>olympus</code>)</li>
<li>Qualcomm Oryon-1 (<code>oryon-1</code>)</li>
</ul>
</li>
<li>The following features are now supported by <code>-march</code>,
<code>-mcpu</code>, and related source-level constructs
(GCC modifiers in parentheses):
<ul>
<li>FEAT_CPA (<code>+cpa</code>), enabled by default for
Arm9.5-A and above
</li>
<li>FEAT_FAMINMAX (<code>+faminmax</code>), enabled by default for
Arm9.5-A and above
</li>
<li>FEAT_FCMA (<code>+fcma</code>), enabled by default for Armv8.3-A
and above
</li>
<li>FEAT_FLAGM2 (<code>+flagm2</code>), enabled by default for
Armv8.5-A and above
</li>
<li>FEAT_FP8 (<code>+fp8</code>)</li>
<li>FEAT_FP8DOT2 (<code>+fp8dot2</code>)</li>
<li>FEAT_FP8DOT4 (<code>+fp8dot4</code>)</li>
<li>FEAT_FP8FMA (<code>+fp8fma</code>)</li>
<li>FEAT_FRINTTS (<code>+frintts</code>), enabled by default for
Armv8.5-A and above
</li>
<li>FEAT_JSCVT (<code>+jscvt</code>), enabled by default for
Armv8.3-A and above
</li>
<li>FEAT_LUT (<code>+lut</code>), enabled by default for
Arm9.5-A and above
</li>
<li>FEAT_LRCPC2 (<code>+rcpc2</code>), enabled by default for
Armv8.4-A and above
</li>
<li>FEAT_SME_B16B16 (<code>+sme-b16b16</code>)</li>
<li>FEAT_SME_F16F16 (<code>+sme-f16f16</code>)</li>
<li>FEAT_SME2p1 (<code>+sme2p1</code>)</li>
<li>FEAT_SSVE_FP8DOT2 (<code>+ssve-fp8dot2</code>)</li>
<li>FEAT_SSVE_FP8DOT4 (<code>+ssve-fp8dot4</code>)</li>
<li>FEAT_SSVE_FP8FMA (<code>+ssve-fp8fma</code>)</li>
<li>FEAT_SVE_B16B16 (<code>+sve-b16b16</code>)</li>
<li>FEAT_SVE2p1 (<code>+sve2p1</code>), enabled by default for
Armv9.4-A and above
</li>
<li>FEAT_WFXT (<code>+wfxt</code>), enabled by default for
Armv8.7-A and above
</li>
<li>FEAT_XS (<code>+xs</code>), enabled by default for
Armv8.7-A and above
</li>
</ul>
The features listed as being enabled by default for Armv8.7-A or earlier
were previously only selectable using the associated architecture level.
For example, FEAT_FCMA was previously selected by
<code>-march=armv8.3-a</code> and above (as it still is), but it wasn't
previously selectable independently.
</li>
<li><code>-mbranch-protection</code> has been extended to support
the Guarded Control Stack (GCS) extension. This support
is included in <code>-mbranch-protection=standard</code> and can
be enabled individually using <code>-mbranch-protection=gcs</code>.
</li>
<li>The following additional changes have been made to the
command-line options:
<ul>
<li>In order to align with other tools, the SME feature modifier
<code>+sme</code> no longer enables SVE. However, GCC does not
yet support using SME without SVE and instead rejects such
combinations with a “not implemented” error.
</li>
<li>The options <code>-mfix-cortex-a53-835769</code> and
<code>-mfix-cortex-a53-843419</code> are now silently ignored
if the selected architecture is incompatible with Cortex-A53.
This is particularly useful for toolchains that are configured
to apply the Cortex-A53 workarounds by default. For example,
all other things being equal, a toolchain configured with
<code>--enable-fix-cortex-a53-835769</code> now produces the
same code for <code>-mcpu=neoverse-n2</code> as a toolchain
configured without <code>--enable-fix-cortex-a53-835769</code>.
</li>
<li><code>-mcpu=native</code> now handles unrecognized heterogeneous
systems by detecting which individual architecture features are
supported by the CPUs. This matches the preexisting behavior for
unknown homogeneous systems.
</li>
<li>The first scheduling pass (<code>-fschedule-insns</code>) is no
longer enabled by default at <code>-O2</code> for AArch64 targets.
The pass is still enabled by default at <code>-O3</code> and
<code>-Ofast</code>.
</li>
</ul>
</li>
<li>Support has been added for the following features of the Arm C
Language Extensions
(<a href="https://github.com/ARM-software/acle">ACLE</a>):
<ul>
<li>guarded control stacks</li>
<li>lookup table instructions with 2-bit and 4-bit indices
(predefined macro
<code>__ARM_FEATURE_LUT</code>, enabled by <code>+lut</code>)
</li>
<li>floating-point absolute minimum and maximum instructions
(predefined macro <code>__ARM_FEATURE_FAMINMAX</code>,
enabled by <code>+faminmax</code>)
</li>
<li>FP8 conversions (predefined macro
<code>__ARM_FEATURE_FP8</code>, enabled by <code>+fp8</code>)
</li>
<li>FP8 2-way dot product to half precision instructions
(predefined macro <code>__ARM_FEATURE_FP8DOT2</code>,
enabled by <code>+fp8dot2</code>)
</li>
<li>FP8 4-way dot product to single precision instructions
(predefined macro <code>__ARM_FEATURE_FP8DOT4</code>,
enabled by <code>+fp8dot4</code>)
</li>
<li>FP8 multiply-accumulate to half precision and single precision
instructions (predefined macro <code>__ARM_FEATURE_FP8FMA</code>,
enabled by <code>+fp8fma</code>)
</li>
<li>SVE FP8 2-way dot product to half precision instructions
(predefined macro <code>__ARM_FEATURE_SSVE_FP8DOT2</code>,
enabled by <code>+ssve-fp8dot2</code>)
</li>
<li>SVE FP8 4-way dot product to single precision instructions
(predefined macro <code>__ARM_FEATURE_SSVE_FP8DOT4</code>,
enabled by <code>+ssve-fp8dot4</code>)
</li>
<li>SVE FP8 multiply-accumulate to half precision and single precision
instructions (predefined macro <code>__ARM_FEATURE_SSVE_FP8FMA</code>,
enabled by <code>+ssve-fp8fma</code>)
</li>
<li>SVE2.1 instructions (predefined macro
<code>__ARM_FEATURE_SVE2p1</code>, enabled by <code>+sve2p1</code>)
</li>
<li>SVE non-widening bfloat16 instructions
(predefined macro <code>__ARM_FEATURE_SVE_B16B16</code>,
enabled by <code>+sve-b16b16</code>)
</li>
<li>SME2.1 instructions (predefined macro
<code>__ARM_FEATURE_SME2p1</code>, enabled by <code>+sme2p1</code>)
</li>
<li>SME non-widening bfloat16 instructions
(predefined macro <code>__ARM_FEATURE_SME_B16B16</code>,
enabled by <code>+sme-b16b16</code>)
</li>
<li>SME half-precision instructions
(predefined macro <code>__ARM_FEATURE_SME_F16F16</code>,
enabled by <code>+sme-f16f16</code>)
</li>
<li>using C and C++ prefix operators, infix operators, and postfix
operators with scalable SVE ACLE types
(predefined macro <code>__ARM_FEATURE_SVE_VECTOR_OPERATORS==2</code>,
enabled by <code>+sve</code>)
</li>
<li><code>__fma</code> (in <code>arm_acle.h</code>)</li>
<li><code>__fmaf</code> (in <code>arm_acle.h</code>)</li>
<li><code>__chkfeat</code> (in <code>arm_acle.h</code>)</li>
</ul>
</li>
<li>In addition, the following changes have been made to preexisting
ACLE features:
<ul>
<li>The macros <code>__ARM_FEATURE_BF16</code> and
<code>__ARM_FEATURE_SVE_BF16</code> are now predefined when the
associated support is available. Previous versions of GCC provided
the associated intrinsics but did not predefine the macros.
</li>
<li>OpenMP tasks can now share scalable SVE vectors and predicates.
However, offloading of scalable vectors and predicates is not
supported.
</li>
<li>ACLE system register functions (such as <code>__arm_rsr</code>
and <code>__arm_wsr</code>) no longer try to enforce the minimum
architectural requirement.
</li>
<li>A warning is reported if code attempts to use the Function
Multi-Versioning feature. GCC's current implementation of this
feature is still experimental and it does not conform to the
ACLE specification.
</li>
</ul>
</li>
<li>Support has been added for the <code>indirect_return</code>
function-type attribute, which indicates that a function might return
via an indirect branch instead of via a normal return instruction.
</li>
<li>128-bit atomic operations have been extended to make use of
FEAT_LRCPC3 instructions, when support for the instructions is
detected at runtime.
</li>
<li>There have been many code-generation improvements to the AArch64 port.
Some examples are:
<ul>
<li>automatic use of AArch64 CRC instructions</li>
<li>automatic use of AArch64 saturating vector arithmetic
instructions
</li>
<li>better code generation of population counts</li>
<li>improved handling of floating-point and vector immediates</li>
<li>improved handling of vector permutations</li>
<li>more use of SVE instructions to optimize Advanced SIMD code</li>
<li>more folding and simplification of SVE ACLE intrinsics</li>
<li>improved CPU-specific tuning</li>
<li>improved register allocation, such as eliminating some
vector moves
</li>
</ul>
</li>
</ul>
<h3 id="amdgcn"><a href="#amdgcn">AMD GPU (GCN)</a></h3>
<ul>
<li>The standard C++ library (libstdc++) is now supported and enabled.</li>
<li>Experimental support for supporting generic devices has been added;
specifying <code>gfx9-generic</code>, <code>gfx10-3-generic</code>,
or <code>gfx11-generic</code> to
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/AMD-GCN-Options.html">
<code>-march=</code></a> will generate code that can run on all
devices of a series. Additionally, the following specific devices
are now have experimental support, all of which are compatible with a
listed generic: <code>gfx902</code>, <code>gfx904</code>,
<code>gfx909</code>, <code>gfx1031</code>, <code>gfx1032</code>,
<code>gfx1033</code>, <code>gfx1034</code>, <code>gfx1035</code>,
<code>gfx1101</code>, <code>gfx1102</code>, <code>gfx1150</code>,
and <code>gfx1151</code>. To use any of the listed new devices including
the generic ones, GCC has to be configured to build the runtime library
for the device. Note that generic support requires ROCm 6.4.0 (or newer).
For details, consult GCC's
<a href="https://gcc.gnu.org/install/specific.html#amdgcn-x-amdhsa">
installation notes</a>.</li>
<li>Support for Fiji (gfx803) devices has been removed (this was already
deprecated in GCC 14).</li>
</ul>
<!-- <h3 id="arc"><a href="#arc">ARC</a></h3> -->
<!-- <h3 id="arm"><a href="#arm">arm</a></h3> -->
<h3 id="avr"><a href="#avr">AVR</a></h3>
<ul>
<li>Support has been added for the <code>signal(<i>num</i>)</code>
and <code>interrupt(<i>num</i>)</code>
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/AVR-Function-Attributes.html#index-signal_0028num_0029-function-attribute_002c-AVR"
>function attributes</a>
that allow to specify the interrupt vector number <code><i>num</i></code>
as an argument.
It allows to use static functions as interrupt handlers, and also
functions defined in a C++ namespace.</li>
<li>Support has been added for the <code>noblock</code> function attribute.
It can be specified together with the <code>signal</code> attribute to
indicate that the interrupt service routine should start with a
<code>SEI</code> instruction to globally re-enable interrupts.
The difference to the <code>interrupt</code> attribute is that the
<code>noblock</code> attribute just acts like a flag and does not
impose a specific function name.</li>
<li>Support has been added for the <code>__builtin_avr_mask1</code>
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/AVR-Built-in-Functions.html#index-_005f_005fbuiltin_005favr_005fmask1"
>built-in function</a>. It can be used to compute some bit masks when
code like <code>1 << offset</code> is not fast enough.</li>
<li>Support has been added for a new 24-bit
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/Named-Address-Spaces.html#AVR-Named-Address-Spaces-1"
>named address space</a> <code>__flashx</code>.
It is similar to the <code>__memx</code> address space introduced in v4.7,
but reading is a bit more efficient since it only supports reading from
program memory. Objects in the <code>__flashx</code> address space are
located in the <code>.progmemx.data</code> section.</li>
<li>Apart from the built-in types <code>__int24</code> and
<code>__uint24</code> supported since v4.7, support has been added for the
<code>signed __int24</code> and <code>unsigned __int24</code> types.</li>
<li>Code generation for the 32-bit integer shifts with constant
offset has been improved. The code size may slightly increase even
when optimizing for code size with <code>-Os</code>.</li>
<li>Support has been added for a <em>compact vector table</em> as supported
by some AVR devices. It can be activated by the new command-line option
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/AVR-Options.html#index-mcvt"
><code>-mcvt</code></a>.
It links <code>crt<i>mcu</i>-cvt.o</code> as startup code which
is supported since AVR-LibC v2.3.</li>
<li>Support has been added for the new option
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/AVR-Options.html#index-mno-call-main"
><code>-mno-call-main</code></a>. Instead of calling <code>main</code>,
it will be located in section <code>.init9</code>.</li>
<li>New AVR specific optimizations have been added.
They can be controlled by the new command-line options
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/AVR-Options.html#index-mfuse-move"
><code>-mfuse-move</code></a>,
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/AVR-Options.html#index-msplit-ldst"
><code>-msplit-ldst</code></a>,
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/AVR-Options.html#index-msplit-bit-shift"
><code>-msplit-bit-shift</code></a> and
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/AVR-Options.html#index-muse-nonzero-bits"
><code>-muse-nonzero-bits</code></a>.</li>
</ul>
<h3 id="x86"><a href="#x86">IA-32/x86-64</a></h3>
<ul>
<li>New ISA extension support for Intel AMX-AVX512 was added.
AMX-AVX512 intrinsics are available via the <code>-mamx-avx512</code>
compiler switch.
</li>
<li>New ISA extension support for Intel AMX-FP8 was added.
AMX-FP8 intrinsics are available via the <code>-mamx-fp8</code>
compiler switch.
</li>
<li>New ISA extension support for Intel AMX-MOVRS was added.
AMX-MOVRS intrinsics are available via the <code>-mamx-movrs</code>
compiler switch.
</li>
<li>New ISA extension support for Intel AMX-TF32 was added.
AMX-TF32 intrinsics are available via the <code>-mamx-tf32</code>
compiler switch.
</li>
<li>New ISA extension support for Intel AMX-TRANSPOSE was added.
AMX-TRANSPOSE intrinsics are available via the <code>-mamx-transpose</code>
compiler switch.
</li>
<li>All of new feature support for Intel APX expect for CFCMOV was added,
including CCMP/CTEST, NF and ZU. APX support is available via the
<code>-mapxf</code> compiler switch.
</li>
<li>New ISA extension support for Intel AVX10.2 was added.
AVX10.2 intrinsics are available via the <code>-mavx10.2</code>
compiler switch.
</li>
<li>New ISA extension support for Intel MOVRS was added.
MOVRS intrinsics are available via the <code>-mmovrs</code>
compiler switch. MOVRS vector intrinsics are available via
the <code>-mmovrs -mavx10.2</code> compiler switches.
</li>
<li>EVEX version support for Intel SM4 was added.
New 512-bit SM4 intrinsics are available via the
<code>-msm4 -mavx10.2</code> compiler switches.
</li>
<li>GCC now supports the Intel CPU named Diamond Rapids through
<code>-march=diamondrapids</code>.
Based on Granite Rapids, the switch further enables the AMX-AVX512,
AMX-FP8, AMX-MOVRS, AMX-TF32, AMX-TRANSPOSE, APX_F, AVX10.2, AVX-IFMA,
AVX-NE-CONVERT, AVX-VNNI-INT16, AVX-VNNI-INT8, CMPccXADD, MOVRS, SHA512,
SM3, SM4, and USER_MSR ISA extensions.
</li>
<li>Support for Xeon Phi CPUs (a.k.a. Knight Landing and Knight Mill) were
removed in GCC 15. GCC will no longer accept <code>-march=knl</code>,
<code>-march=knm</code>, <code>-mavx5124fmaps</code>,
<code>-mavx5124vnniw</code>, <code>-mavx512er</code>,
<code>-mavx512pf</code>, <code>-mprefetchwt1</code>,
<code>-mtune=knl</code>, and <code>-mtune=knm</code> compiler switches.
</li>
<li><code>-mavx10.1-256</code>, <code>-mavx10.1-512</code>, and
<code>-mevex512</code> are deprecated. Meanwhile, <code>-mavx10.1</code>
enables AVX10.1 intrinsics with 512-bit vector support, while in GCC 14.1
and GCC 14.2, it only enables 256-bit vector support. GCC will emit a
warning when using these compiler switches. <code>-mavx10.1-256</code>,
<code>-mavx10.1-512</code>, and <code>-mevex512</code> will be removed in
GCC 16 together with the warning for the behavior change on
<code>-mavx10.1</code>.
</li>
<li>With the <code>-mveclibabi</code> compiler switch GCC is able to generate
vectorized calls to external libraries. GCC 15 newly supports generating
vectorized math calls to the math library from AMD Optimizing CPU Libraries
(AOCL LibM). This option is available through
<code>-mveclibabi=aocl</code>. GCC still supports generating calls to AMD
Core Math Library (ACML). However, that library is end-of-life and AOCL
offers many more vectorized functions.
</li>
</ul>
<h3 id="loongarch"><a href="#loongarch">LoongArch</a></h3>
<ul>
<li>Support has been added for target attributes and pragmas. For more details
on available attributes and pragmas, including their proper usage, please
refer to the provided
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/LoongArch-Function-Attributes.html#LoongArch-Function-Attributes-1">
documentation</a>.
</li>
<li>Support has been added for the new option
<a href="https://gcc.gnu.org/onlinedocs/gcc-15.1.0/gcc/LoongArch-Options.html#index-mannotate-tablejump">
<code>-mannotate-tablejump</code></a>. Which can create an annotation
section <code>.discard.tablejump_annotate</code> to correlate the
<code>jirl</code> instruction and the jump table.
</li>
<li>Add CRC expander to generate faster CRC.
</li>
<li>Add ABI names for FPR.
</li>
</ul>
<!-- <h3 id="mips"><a href="#mips">MIPS</a></h3> -->
<!-- <h3 id="mep"><a href="#mep">MeP</a></h3> -->
<!-- <h3 id="msp430"><a href="#msp430">MSP430</a></h3> -->
<!-- <h3 id="nds32"><a href="#nds32">NDS32</a></h3> -->
<!-- <h3 id="nios2"><a href="#nios2">Nios II</a></h3> -->
<h3 id="nvptx"><a href="#nvptx">NVPTX</a></h3>
<ul>
<li>The standard C++ library (libstdc++) is now supported and enabled.</li>
<li>GCC's nvptx target now supports constructors and destructors.
For this, a recent version of <a
href="https://gcc.gnu.org/install/specific.html#nvptx-x-none"
>nvptx-tools is required</a>.</li>
</ul>
<!-- <h3 id="hppa"><a href="#hppa">PA-RISC</a></h3> -->
<!-- <h3 id="powerpc"><a href="#powerpc">PowerPC / PowerPC64 / RS6000</a></h3> -->
<!-- <h3 id="s390"><a href="#s390">S/390, System z, IBM z Systems</a></h3> -->
<!-- <h3 id="riscv"><a href="#riscv">RISC-V</a></h3> -->
<!-- <h3 id="rx"><a href="#rx">RX</a></h3> -->
<h3 id="sh"><a href="#sh">SH</a></h3>
<ul>
<li>Bare metal <code>sh-elf</code> targets are now using the newer soft-fp
library for improved performance of floating-point emulation on CPUs
without hardware floating-point support.</li>
</ul>
<!-- <h3 id="sparc"><a href="#sparc">SPARC</a></h3> -->
<!-- <h3 id="Tile"><a href="#Tile">Tile</a></h3> -->
<!-- .................................................................. -->
<h2 id="os"><a href="#os">Operating Systems</a></h2>
<!-- <h3 id="aix"><a href="#aix">AIX</a></h3> -->
<h3>PowerPC Darwin</h3>
<ul>
<li>Fortran's IEEE modules are now supported on Darwin PowerPC.</li>
</ul>
<!-- <h3 id="fuchsia"><a href="#fuchsia">Fuchsia</a></h3> -->
<!-- <h3 id="dragonfly"><a href="#dragonfly">DragonFly BSD</a></h3> -->
<!-- <h3 id="freebsd"><a href="#freebsd">FreeBSD</a></h3> -->
<!-- <h3 id="gnulinux"><a href="#gnulinux">GNU/Linux</a></h3> -->
<!-- <h3 id="rtems"><a href="#rtems">RTEMS</a></h3> -->
<!-- <h3 id="solaris"><a href="#solaris">Solaris</a></h3> -->
<!-- <h3 id="vxmils"><a href="#vxmils">VxWorks MILS</a></h3> -->
<!-- <h3 id="windows"><a href="#windows">Windows</a></h3> -->
<!-- .................................................................. -->
<!-- <h2>Documentation improvements</h2> -->
<!-- .................................................................. -->
<!-- <h2 id="plugins"><a href="#plugins">Improvements for plugin authors</a></h2> -->
<!-- .................................................................. -->
<h2>Other significant improvements</h2>
<!-- <h3 id="uninitialized"><a href="#uninitialized">Eliminating uninitialized variables</a></h3> -->
<!-- .................................................................. -->
<h2 id="15.1"><a href="#15.1">GCC 15.1</a></h2>
<p>This is the <a href="https://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=15.0">list
of problem reports (PRs)</a> from GCC's bug tracking system that are
known to be fixed in the 15.1 release. This list might not be
complete (that is, it is possible that some PRs that have been fixed
are not listed here).</p>
<!-- .................................................................. -->
<!-- ==================================================================== -->
<div class="copyright">
<address>For questions related to the use of GCC,
please consult these web pages and the
<a href="https://gcc.gnu.org/onlinedocs/">GCC manuals</a>. If that fails,
the <a href="mailto:gcc-help@gcc.gnu.org">gcc-help@gcc.gnu.org</a>
mailing list might help.
Comments on these web pages and the development of GCC are welcome on our
developer list at <a href="mailto:gcc@gcc.gnu.org">gcc@gcc.gnu.org</a>.
All of <a href="https://gcc.gnu.org/lists.html">our lists</a>
have public archives.
</address>
<p>Copyright (C)
<a href="https://www.fsf.org">Free Software Foundation, Inc.</a>
Verbatim copying and distribution of this entire article is
permitted in any medium, provided this notice is preserved.</p>
<p>These pages are
<a href="https://gcc.gnu.org/about.html">maintained by the GCC team</a>.
Last modified 2025-04-25.</p><!-- IGNORE DIFF -->
</div>
<!-- ==================================================================== -->
</body>
</html>