1 @c Copyright (C) 2009-2021 Free Software Foundation, Inc.
2 @c Contributed by ARM Ltd.
3 @c This is part of the GAS manual.
4 @c For copying conditions, see the file as.texinfo.
9 @node AArch64-Dependent
10 @chapter AArch64 Dependent Features
14 @node Machine Dependencies
15 @chapter AArch64 Dependent Features
18 @cindex AArch64 support
20 * AArch64 Options:: Options
21 * AArch64 Extensions:: Extensions
22 * AArch64 Syntax:: Syntax
23 * AArch64 Floating Point:: Floating Point
24 * AArch64 Directives:: AArch64 Machine Directives
25 * AArch64 Opcodes:: Opcodes
26 * AArch64 Mapping Symbols:: Mapping Symbols
31 @cindex AArch64 options (none)
32 @cindex options for AArch64 (none)
37 @cindex @option{-EB} command-line option, AArch64
39 This option specifies that the output generated by the assembler should
40 be marked as being encoded for a big-endian processor.
42 @cindex @option{-EL} command-line option, AArch64
44 This option specifies that the output generated by the assembler should
45 be marked as being encoded for a little-endian processor.
47 @cindex @option{-mabi=} command-line option, AArch64
49 Specify which ABI the source code uses. The recognized arguments
50 are: @code{ilp32} and @code{lp64}, which decides the generated object
51 file in ELF32 and ELF64 format respectively. The default is @code{lp64}.
53 @cindex @option{-mcpu=} command-line option, AArch64
54 @item -mcpu=@var{processor}[+@var{extension}@dots{}]
55 This option specifies the target processor. The assembler will issue an error
56 message if an attempt is made to assemble an instruction which will not execute
57 on the target processor. The following processor names are recognized:
90 The special name @code{all} may be used to allow the assembler to accept
91 instructions valid for any supported processor, including all optional
94 In addition to the basic instruction set, the assembler can be told to
95 accept, or restrict, various extension mnemonics that extend the
96 processor. @xref{AArch64 Extensions}.
98 If some implementations of a particular processor can have an
99 extension, then then those extensions are automatically enabled.
100 Consequently, you will not normally have to specify any additional
103 @cindex @option{-march=} command-line option, AArch64
104 @item -march=@var{architecture}[+@var{extension}@dots{}]
105 This option specifies the target architecture. The assembler will
106 issue an error message if an attempt is made to assemble an
107 instruction which will not execute on the target architecture. The
108 following architecture names are recognized: @code{armv8-a},
109 @code{armv8.1-a}, @code{armv8.2-a}, @code{armv8.3-a}, @code{armv8.4-a}
110 @code{armv8.5-a}, @code{armv8.6-a}, @code{armv8.7-a}, @code{armv8-r}, and
113 If both @option{-mcpu} and @option{-march} are specified, the
114 assembler will use the setting for @option{-mcpu}. If neither are
115 specified, the assembler will default to @option{-mcpu=all}.
117 The architecture option can be extended with the same instruction set
118 extension options as the @option{-mcpu} option. Unlike
119 @option{-mcpu}, extensions are not always enabled by default,
120 @xref{AArch64 Extensions}.
122 @cindex @code{-mverbose-error} command-line option, AArch64
123 @item -mverbose-error
124 This option enables verbose error messages for AArch64 gas. This option
125 is enabled by default.
127 @cindex @code{-mno-verbose-error} command-line option, AArch64
128 @item -mno-verbose-error
129 This option disables verbose error messages in AArch64 gas.
134 @node AArch64 Extensions
135 @section Architecture Extensions
137 The table below lists the permitted architecture extensions that are
138 supported by the assembler and the conditions under which they are
139 automatically enabled.
141 Multiple extensions may be specified, separated by a @code{+}.
142 Extension mnemonics may also be removed from those the assembler
143 accepts. This is done by prepending @code{no} to the option that adds
144 the extension. Extensions that are removed must be listed after all
145 extensions that have been added.
147 Enabling an extension that requires other extensions will
148 automatically cause those extensions to be enabled. Similarly,
149 disabling an extension that is required by other extensions will
150 automatically cause those extensions to be disabled.
152 @multitable @columnfractions .12 .17 .17 .54
153 @headitem Extension @tab Minimum Architecture @tab Enabled by default
155 @item @code{i8mm} @tab ARMv8.2-A @tab ARMv8.6-A or later
156 @tab Enable Int8 Matrix Multiply extension.
157 @item @code{f32mm} @tab ARMv8.2-A @tab No
158 @tab Enable F32 Matrix Multiply extension.
159 @item @code{f64mm} @tab ARMv8.2-A @tab No
160 @tab Enable F64 Matrix Multiply extension.
161 @item @code{bf16} @tab ARMv8.2-A @tab ARMv8.6-A or later
162 @tab Enable BFloat16 extension.
163 @item @code{compnum} @tab ARMv8.2-A @tab ARMv8.3-A or later
164 @tab Enable the complex number SIMD extensions. This implies
165 @code{fp16} and @code{simd}.
166 @item @code{crc} @tab ARMv8-A @tab ARMv8.1-A or later
167 @tab Enable CRC instructions.
168 @item @code{crypto} @tab ARMv8-A @tab No
169 @tab Enable cryptographic extensions. This implies @code{fp}, @code{simd}, @code{aes} and @code{sha2}.
170 @item @code{aes} @tab ARMv8-A @tab No
171 @tab Enable the AES cryptographic extensions. This implies @code{fp} and @code{simd}.
172 @item @code{sha2} @tab ARMv8-A @tab No
173 @tab Enable the SHA2 cryptographic extensions. This implies @code{fp} and @code{simd}.
174 @item @code{sha3} @tab ARMv8.2-A @tab No
175 @tab Enable the ARMv8.2-A SHA2 and SHA3 cryptographic extensions. This implies @code{fp}, @code{simd} and @code{sha2}.
176 @item @code{sm4} @tab ARMv8.2-A @tab No
177 @tab Enable the ARMv8.2-A SM3 and SM4 cryptographic extensions. This implies @code{fp} and @code{simd}.
178 @item @code{fp} @tab ARMv8-A @tab ARMv8-A or later
179 @tab Enable floating-point extensions.
180 @item @code{fp16} @tab ARMv8.2-A @tab ARMv8.2-A or later
181 @tab Enable ARMv8.2 16-bit floating-point support. This implies
183 @item @code{lor} @tab ARMv8-A @tab ARMv8.1-A or later
184 @tab Enable Limited Ordering Regions extensions.
185 @item @code{lse} @tab ARMv8-A @tab ARMv8.1-A or later
186 @tab Enable Large System extensions.
187 @item @code{pan} @tab ARMv8-A @tab ARMv8.1-A or later
188 @tab Enable Privileged Access Never support.
189 @item @code{profile} @tab ARMv8.2-A @tab No
190 @tab Enable statistical profiling extensions.
191 @item @code{ras} @tab ARMv8-A @tab ARMv8.2-A or later
192 @tab Enable the Reliability, Availability and Serviceability
194 @item @code{rcpc} @tab ARMv8.2-A @tab ARMv8.3-A or later
195 @tab Enable the weak release consistency extension.
196 @item @code{rdma} @tab ARMv8-A @tab ARMv8.1-A or later
197 @tab Enable ARMv8.1 Advanced SIMD extensions. This implies @code{simd}.
198 @item @code{simd} @tab ARMv8-A @tab ARMv8-A or later
199 @tab Enable Advanced SIMD extensions. This implies @code{fp}.
200 @item @code{sve} @tab ARMv8.2-A @tab Armv9-A or later
201 @tab Enable the Scalable Vector Extensions. This implies @code{fp16},
202 @code{simd} and @code{compnum}.
203 @item @code{dotprod} @tab ARMv8.2-A @tab ARMv8.4-A or later
204 @tab Enable the Dot Product extension. This implies @code{simd}.
205 @item @code{fp16fml} @tab ARMv8.2-A @tab ARMv8.4-A or later
206 @tab Enable ARMv8.2 16-bit floating-point multiplication variant support.
207 This implies @code{fp16}.
208 @item @code{sb} @tab ARMv8-A @tab ARMv8.5-A or later
209 @tab Enable the speculation barrier instruction sb.
210 @item @code{predres} @tab ARMv8-A @tab ARMv8.5-A or later
211 @tab Enable the Execution and Data and Prediction instructions.
212 @item @code{rng} @tab ARMv8.5-A @tab No
213 @tab Enable ARMv8.5-A random number instructions.
214 @item @code{ssbs} @tab ARMv8-A @tab ARMv8.5-A or later
215 @tab Enable Speculative Store Bypassing Safe state read and write.
216 @item @code{memtag} @tab ARMv8.5-A @tab No
217 @tab Enable ARMv8.5-A Memory Tagging Extensions.
218 @item @code{tme} @tab ARMv8-A @tab No
219 @tab Enable Transactional Memory Extensions.
220 @item @code{sve2} @tab ARMv8-A @tab Armv9-A or later
221 @tab Enable the SVE2 Extension.
222 @item @code{sve2-bitperm} @tab ARMv8-A @tab No
223 @tab Enable SVE2 BITPERM Extension.
224 @item @code{sve2-sm4} @tab ARMv8-A @tab No
225 @tab Enable SVE2 SM4 Extension.
226 @item @code{sve2-aes} @tab ARMv8-A @tab No
227 @tab Enable SVE2 AES Extension. This also enables the .Q->.B form of the
228 @code{pmullt} and @code{pmullb} instructions.
229 @item @code{sve2-sha3} @tab ARMv8-A @tab No
230 @tab Enable SVE2 SHA3 Extension.
231 @item @code{flagm} @tab ARMv8-A @tab ARMv8.4-A or later
232 @tab Enable Flag Manipulation instructions.
233 @item @code{ls64} @tab ARMv8.6-A @tab ARMv8.7-A or later
234 @tab Enable 64 Byte Loads/Stores.
235 @item @code{pauth} @tab ARMv8-A @tab No
236 @tab Enable Pointer Authentication.
242 * AArch64-Chars:: Special Characters
243 * AArch64-Regs:: Register Names
244 * AArch64-Relocations:: Relocations
248 @subsection Special Characters
250 @cindex line comment character, AArch64
251 @cindex AArch64 line comment character
252 The presence of a @samp{//} on a line indicates the start of a comment
253 that extends to the end of the current line. If a @samp{#} appears as
254 the first character of a line, the whole line is treated as a comment.
256 @cindex line separator, AArch64
257 @cindex statement separator, AArch64
258 @cindex AArch64 line separator
259 The @samp{;} character can be used instead of a newline to separate
262 @cindex immediate character, AArch64
263 @cindex AArch64 immediate character
264 The @samp{#} can be optionally used to indicate immediate operands.
267 @subsection Register Names
269 @cindex AArch64 register names
270 @cindex register names, AArch64
271 Please refer to the section @samp{4.4 Register Names} of
272 @samp{ARMv8 Instruction Set Overview}, which is available at
273 @uref{http://infocenter.arm.com}.
275 @node AArch64-Relocations
276 @subsection Relocations
278 @cindex relocations, AArch64
279 @cindex AArch64 relocations
280 @cindex MOVN, MOVZ and MOVK group relocations, AArch64
281 Relocations for @samp{MOVZ} and @samp{MOVK} instructions can be generated
282 by prefixing the label with @samp{#:abs_g2:} etc.
283 For example to load the 48-bit absolute address of @var{foo} into x0:
286 movz x0, #:abs_g2:foo // bits 32-47, overflow check
287 movk x0, #:abs_g1_nc:foo // bits 16-31, no overflow check
288 movk x0, #:abs_g0_nc:foo // bits 0-15, no overflow check
291 @cindex ADRP, ADD, LDR/STR group relocations, AArch64
292 Relocations for @samp{ADRP}, and @samp{ADD}, @samp{LDR} or @samp{STR}
293 instructions can be generated by prefixing the label with
294 @samp{:pg_hi21:} and @samp{#:lo12:} respectively.
296 For example to use 33-bit (+/-4GB) pc-relative addressing to
297 load the address of @var{foo} into x0:
300 adrp x0, :pg_hi21:foo
301 add x0, x0, #:lo12:foo
304 Or to load the value of @var{foo} into x0:
307 adrp x0, :pg_hi21:foo
308 ldr x0, [x0, #:lo12:foo]
311 Note that @samp{:pg_hi21:} is optional.
320 adrp x0, :pg_hi21:foo
323 @node AArch64 Floating Point
324 @section Floating Point
326 @cindex floating point, AArch64 (@sc{ieee})
327 @cindex AArch64 floating point (@sc{ieee})
328 The AArch64 architecture uses @sc{ieee} floating-point numbers.
330 @node AArch64 Directives
331 @section AArch64 Machine Directives
333 @cindex machine directives, AArch64
334 @cindex AArch64 machine directives
337 @c AAAAAAAAAAAAAAAAAAAAAAAAA
339 @cindex @code{.arch} directive, AArch64
340 @item .arch @var{name}
341 Select the target architecture. Valid values for @var{name} are the same as
342 for the @option{-march} command-line option.
344 Specifying @code{.arch} clears any previously selected architecture
347 @cindex @code{.arch_extension} directive, AArch64
348 @item .arch_extension @var{name}
349 Add or remove an architecture extension to the target architecture. Valid
350 values for @var{name} are the same as those accepted as architectural
351 extensions by the @option{-mcpu} command-line option.
353 @code{.arch_extension} may be used multiple times to add or remove extensions
354 incrementally to the architecture being compiled for.
356 @c BBBBBBBBBBBBBBBBBBBBBBBBBB
358 @cindex @code{.bss} directive, AArch64
360 This directive switches to the @code{.bss} section.
362 @c CCCCCCCCCCCCCCCCCCCCCCCCCC
364 @cindex @code{.cpu} directive, AArch64
365 @item .cpu @var{name}
366 Set the target processor. Valid values for @var{name} are the same as
367 those accepted by the @option{-mcpu=} command-line option.
369 @c DDDDDDDDDDDDDDDDDDDDDDDDDD
371 @cindex @code{.dword} directive, AArch64
372 @item .dword @var{expressions}
373 The @code{.dword} directive produces 64 bit values.
375 @c EEEEEEEEEEEEEEEEEEEEEEEEEE
377 @cindex @code{.even} directive, AArch64
379 The @code{.even} directive aligns the output on the next even byte
382 @c FFFFFFFFFFFFFFFFFFFFFFFFFF
384 @cindex @code{.float16} directive, AArch64
385 @item .float16 @var{value [,...,value_n]}
386 Place the half precision floating point representation of one or more
387 floating-point values into the current section.
388 The format used to encode the floating point values is always the
389 IEEE 754-2008 half precision floating point format.
391 @c GGGGGGGGGGGGGGGGGGGGGGGGGG
392 @c HHHHHHHHHHHHHHHHHHHHHHHHHH
393 @c IIIIIIIIIIIIIIIIIIIIIIIIII
395 @cindex @code{.inst} directive, AArch64
396 @item .inst @var{expressions}
397 Inserts the expressions into the output as if they were instructions,
400 @c JJJJJJJJJJJJJJJJJJJJJJJJJJ
401 @c KKKKKKKKKKKKKKKKKKKKKKKKKK
402 @c LLLLLLLLLLLLLLLLLLLLLLLLLL
404 @cindex @code{.ltorg} directive, AArch64
406 This directive causes the current contents of the literal pool to be
407 dumped into the current section (which is assumed to be the .text
408 section) at the current location (aligned to a word boundary).
409 GAS maintains a separate literal pool for each section and each
410 sub-section. The @code{.ltorg} directive will only affect the literal
411 pool of the current section and sub-section. At the end of assembly
412 all remaining, un-empty literal pools will automatically be dumped.
414 Note - older versions of GAS would dump the current literal
415 pool any time a section change occurred. This is no longer done, since
416 it prevents accurate control of the placement of literal pools.
418 @c MMMMMMMMMMMMMMMMMMMMMMMMMM
420 @c NNNNNNNNNNNNNNNNNNNNNNNNNN
421 @c OOOOOOOOOOOOOOOOOOOOOOOOOO
423 @c PPPPPPPPPPPPPPPPPPPPPPPPPP
425 @cindex @code{.pool} directive, AArch64
427 This is a synonym for .ltorg.
429 @c QQQQQQQQQQQQQQQQQQQQQQQQQQ
430 @c RRRRRRRRRRRRRRRRRRRRRRRRRR
432 @cindex @code{.req} directive, AArch64
433 @item @var{name} .req @var{register name}
434 This creates an alias for @var{register name} called @var{name}. For
441 ip0, ip1, lr and fp are automatically defined to
442 alias to X16, X17, X30 and X29 respectively.
444 @c SSSSSSSSSSSSSSSSSSSSSSSSSS
446 @c TTTTTTTTTTTTTTTTTTTTTTTTTT
448 @cindex @code{.tlsdescadd} directive, AArch64
449 @item @code{.tlsdescadd}
450 Emits a TLSDESC_ADD reloc on the next instruction.
452 @cindex @code{.tlsdesccall} directive, AArch64
453 @item @code{.tlsdesccall}
454 Emits a TLSDESC_CALL reloc on the next instruction.
456 @cindex @code{.tlsdescldr} directive, AArch64
457 @item @code{.tlsdescldr}
458 Emits a TLSDESC_LDR reloc on the next instruction.
460 @c UUUUUUUUUUUUUUUUUUUUUUUUUU
462 @cindex @code{.unreq} directive, AArch64
463 @item .unreq @var{alias-name}
464 This undefines a register alias which was previously defined using the
465 @code{req} directive. For example:
472 An error occurs if the name is undefined. Note - this pseudo op can
473 be used to delete builtin in register name aliases (eg 'w0'). This
474 should only be done if it is really necessary.
476 @c VVVVVVVVVVVVVVVVVVVVVVVVVV
478 @cindex @code{.variant_pcs} directive, AArch64
479 @item .variant_pcs @var{symbol}
480 This directive marks @var{symbol} referencing a function that may
481 follow a variant procedure call standard with different register
482 usage convention from the base procedure call standard.
484 @c WWWWWWWWWWWWWWWWWWWWWWWWWW
485 @c XXXXXXXXXXXXXXXXXXXXXXXXXX
487 @cindex @code{.xword} directive, AArch64
488 @item .xword @var{expressions}
489 The @code{.xword} directive produces 64 bit values. This is the same
490 as the @code{.dword} directive.
492 @c YYYYYYYYYYYYYYYYYYYYYYYYYY
493 @c ZZZZZZZZZZZZZZZZZZZZZZZZZZ
495 @cindex @code{.cfi_b_key_frame} directive, AArch64
496 @item @code{.cfi_b_key_frame}
497 The @code{.cfi_b_key_frame} directive inserts a 'B' character into the CIE
498 corresponding to the current frame's FDE, meaning that its return address has
499 been signed with the B-key. If two frames are signed with differing keys then
500 they will not share the same CIE. This information is intended to be used by
501 the stack unwinder in order to properly authenticate return addresses.
505 @node AArch64 Opcodes
508 @cindex AArch64 opcodes
509 @cindex opcodes for AArch64
510 GAS implements all the standard AArch64 opcodes. It also
511 implements several pseudo opcodes, including several synthetic load
516 @cindex @code{LDR reg,=<expr>} pseudo op, AArch64
519 ldr <register> , =<expression>
522 The constant expression will be placed into the nearest literal pool (if it not
523 already there) and a PC-relative LDR instruction will be generated.
527 For more information on the AArch64 instruction set and assembly language
528 notation, see @samp{ARMv8 Instruction Set Overview} available at
529 @uref{http://infocenter.arm.com}.
532 @node AArch64 Mapping Symbols
533 @section Mapping Symbols
535 The AArch64 ELF specification requires that special symbols be inserted
536 into object files to mark certain features:
542 At the start of a region of code containing AArch64 instructions.
546 At the start of a region of data.