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:
93 The special name @code{all} may be used to allow the assembler to accept
94 instructions valid for any supported processor, including all optional
97 In addition to the basic instruction set, the assembler can be told to
98 accept, or restrict, various extension mnemonics that extend the
99 processor. @xref{AArch64 Extensions}.
101 If some implementations of a particular processor can have an
102 extension, then then those extensions are automatically enabled.
103 Consequently, you will not normally have to specify any additional
106 @cindex @option{-march=} command-line option, AArch64
107 @item -march=@var{architecture}[+@var{extension}@dots{}]
108 This option specifies the target architecture. The assembler will
109 issue an error message if an attempt is made to assemble an
110 instruction which will not execute on the target architecture. The
111 following architecture names are recognized: @code{armv8-a},
112 @code{armv8.1-a}, @code{armv8.2-a}, @code{armv8.3-a}, @code{armv8.4-a}
113 @code{armv8.5-a}, @code{armv8.6-a}, @code{armv8.7-a}, @code{armv8-r}, and
116 If both @option{-mcpu} and @option{-march} are specified, the
117 assembler will use the setting for @option{-mcpu}. If neither are
118 specified, the assembler will default to @option{-mcpu=all}.
120 The architecture option can be extended with the same instruction set
121 extension options as the @option{-mcpu} option. Unlike
122 @option{-mcpu}, extensions are not always enabled by default,
123 @xref{AArch64 Extensions}.
125 @cindex @code{-mverbose-error} command-line option, AArch64
126 @item -mverbose-error
127 This option enables verbose error messages for AArch64 gas. This option
128 is enabled by default.
130 @cindex @code{-mno-verbose-error} command-line option, AArch64
131 @item -mno-verbose-error
132 This option disables verbose error messages in AArch64 gas.
137 @node AArch64 Extensions
138 @section Architecture Extensions
140 The table below lists the permitted architecture extensions that are
141 supported by the assembler and the conditions under which they are
142 automatically enabled.
144 Multiple extensions may be specified, separated by a @code{+}.
145 Extension mnemonics may also be removed from those the assembler
146 accepts. This is done by prepending @code{no} to the option that adds
147 the extension. Extensions that are removed must be listed after all
148 extensions that have been added.
150 Enabling an extension that requires other extensions will
151 automatically cause those extensions to be enabled. Similarly,
152 disabling an extension that is required by other extensions will
153 automatically cause those extensions to be disabled.
155 @multitable @columnfractions .12 .17 .17 .54
156 @headitem Extension @tab Minimum Architecture @tab Enabled by default
158 @item @code{aes} @tab ARMv8-A @tab No
159 @tab Enable the AES cryptographic extensions. This implies @code{fp} and
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 @code{fp16} and
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},
170 @code{aes} and @code{sha2}.
171 @item @code{dotprod} @tab ARMv8.2-A @tab ARMv8.4-A or later
172 @tab Enable the Dot Product extension. This implies @code{simd}.
173 @item @code{f32mm} @tab ARMv8.2-A @tab No
174 @tab Enable F32 Matrix Multiply extension. This implies @code{sve}.
175 @item @code{f64mm} @tab ARMv8.2-A @tab No
176 @tab Enable F64 Matrix Multiply extension. This implies @code{sve}.
177 @item @code{flagm} @tab ARMv8-A @tab ARMv8.4-A or later
178 @tab Enable Flag Manipulation instructions.
179 @item @code{fp16fml} @tab ARMv8.2-A @tab ARMv8.4-A or later
180 @tab Enable ARMv8.2 16-bit floating-point multiplication variant support. This
181 implies @code{fp} and @code{fp16}.
182 @item @code{fp16} @tab ARMv8.2-A @tab ARMv8.2-A or later
183 @tab Enable ARMv8.2 16-bit floating-point support. This implies @code{fp}.
184 @item @code{fp} @tab ARMv8-A @tab ARMv8-A or later
185 @tab Enable floating-point extensions.
186 @item @code{i8mm} @tab ARMv8.2-A @tab ARMv8.6-A or later
187 @tab Enable Int8 Matrix Multiply extension.
188 @item @code{lor} @tab ARMv8-A @tab ARMv8.1-A or later
189 @tab Enable Limited Ordering Regions extensions.
190 @item @code{ls64} @tab ARMv8.6-A @tab ARMv8.7-A or later
191 @tab Enable 64 Byte Loads/Stores.
192 @item @code{lse} @tab ARMv8-A @tab ARMv8.1-A or later
193 @tab Enable Large System extensions.
194 @item @code{memtag} @tab ARMv8.5-A @tab No
195 @tab Enable ARMv8.5-A Memory Tagging Extensions.
196 @item @code{pan} @tab ARMv8-A @tab ARMv8.1-A or later
197 @tab Enable Privileged Access Never support.
198 @item @code{pauth} @tab ARMv8-A @tab No
199 @tab Enable Pointer Authentication.
200 @item @code{predres} @tab ARMv8-A @tab ARMv8.5-A or later
201 @tab Enable the Execution and Data and Prediction instructions.
202 @item @code{profile} @tab ARMv8.2-A @tab No
203 @tab Enable statistical profiling extensions.
204 @item @code{ras} @tab ARMv8-A @tab ARMv8.2-A or later
205 @tab Enable the Reliability, Availability and Serviceability extension.
206 @item @code{rcpc} @tab ARMv8.2-A @tab ARMv8.3-A or later
207 @tab Enable the weak release consistency extension.
208 @item @code{rdma} @tab ARMv8-A @tab ARMv8.1-A or later
209 @tab Enable ARMv8.1 Advanced SIMD extensions. This implies @code{simd}.
210 @item @code{rng} @tab ARMv8.5-A @tab No
211 @tab Enable ARMv8.5-A random number instructions.
212 @item @code{sb} @tab ARMv8-A @tab ARMv8.5-A or later
213 @tab Enable the speculation barrier instruction sb.
214 @item @code{sha2} @tab ARMv8-A @tab No
215 @tab Enable the SHA2 cryptographic extensions. This implies @code{fp} and
217 @item @code{sha3} @tab ARMv8.2-A @tab No
218 @tab Enable the ARMv8.2-A SHA2 and SHA3 cryptographic extensions. This implies
219 @code{fp}, @code{simd} and @code{sha2}.
220 @item @code{simd} @tab ARMv8-A @tab ARMv8-A or later
221 @tab Enable Advanced SIMD extensions. This implies @code{fp}.
222 @item @code{sm4} @tab ARMv8.2-A @tab No
223 @tab Enable the ARMv8.2-A SM3 and SM4 cryptographic extensions. This implies
224 @code{fp} and @code{simd}.
225 @item @code{ssbs} @tab ARMv8-A @tab ARMv8.5-A or later
226 @tab Enable Speculative Store Bypassing Safe state read and write.
227 @item @code{sve} @tab ARMv8.2-A @tab Armv9-A or later
228 @tab Enable the Scalable Vector Extensions. This implies @code{fp16},
229 @code{simd} and @code{compnum}.
230 @item @code{sve2} @tab ARMv8-A @tab Armv9-A or later
231 @tab Enable the SVE2 Extension. This implies @code{sve}.
232 @item @code{sve2-aes} @tab ARMv8-A @tab No
233 @tab Enable SVE2 AES Extension. This also enables the .Q->.B form of the
234 @code{pmullt} and @code{pmullb} instructions. This implies @code{aes} and
236 @item @code{sve2-bitperm} @tab ARMv8-A @tab No
237 @tab Enable SVE2 BITPERM Extension.
238 @item @code{sve2-sha3} @tab ARMv8-A @tab No
239 @tab Enable SVE2 SHA3 Extension. This implies @code{sha3} and @code{sve2}.
240 @item @code{sve2-sm4} @tab ARMv8-A @tab No
241 @tab Enable SVE2 SM4 Extension. This implies @code{sm4} and @code{sve2}.
242 @item @code{tme} @tab ARMv8-A @tab No
243 @tab Enable Transactional Memory Extensions.
249 * AArch64-Chars:: Special Characters
250 * AArch64-Regs:: Register Names
251 * AArch64-Relocations:: Relocations
255 @subsection Special Characters
257 @cindex line comment character, AArch64
258 @cindex AArch64 line comment character
259 The presence of a @samp{//} on a line indicates the start of a comment
260 that extends to the end of the current line. If a @samp{#} appears as
261 the first character of a line, the whole line is treated as a comment.
263 @cindex line separator, AArch64
264 @cindex statement separator, AArch64
265 @cindex AArch64 line separator
266 The @samp{;} character can be used instead of a newline to separate
269 @cindex immediate character, AArch64
270 @cindex AArch64 immediate character
271 The @samp{#} can be optionally used to indicate immediate operands.
274 @subsection Register Names
276 @cindex AArch64 register names
277 @cindex register names, AArch64
278 Please refer to the section @samp{4.4 Register Names} of
279 @samp{ARMv8 Instruction Set Overview}, which is available at
280 @uref{http://infocenter.arm.com}.
282 @node AArch64-Relocations
283 @subsection Relocations
285 @cindex relocations, AArch64
286 @cindex AArch64 relocations
287 @cindex MOVN, MOVZ and MOVK group relocations, AArch64
288 Relocations for @samp{MOVZ} and @samp{MOVK} instructions can be generated
289 by prefixing the label with @samp{#:abs_g2:} etc.
290 For example to load the 48-bit absolute address of @var{foo} into x0:
293 movz x0, #:abs_g2:foo // bits 32-47, overflow check
294 movk x0, #:abs_g1_nc:foo // bits 16-31, no overflow check
295 movk x0, #:abs_g0_nc:foo // bits 0-15, no overflow check
298 @cindex ADRP, ADD, LDR/STR group relocations, AArch64
299 Relocations for @samp{ADRP}, and @samp{ADD}, @samp{LDR} or @samp{STR}
300 instructions can be generated by prefixing the label with
301 @samp{:pg_hi21:} and @samp{#:lo12:} respectively.
303 For example to use 33-bit (+/-4GB) pc-relative addressing to
304 load the address of @var{foo} into x0:
307 adrp x0, :pg_hi21:foo
308 add x0, x0, #:lo12:foo
311 Or to load the value of @var{foo} into x0:
314 adrp x0, :pg_hi21:foo
315 ldr x0, [x0, #:lo12:foo]
318 Note that @samp{:pg_hi21:} is optional.
327 adrp x0, :pg_hi21:foo
330 @node AArch64 Floating Point
331 @section Floating Point
333 @cindex floating point, AArch64 (@sc{ieee})
334 @cindex AArch64 floating point (@sc{ieee})
335 The AArch64 architecture uses @sc{ieee} floating-point numbers.
337 @node AArch64 Directives
338 @section AArch64 Machine Directives
340 @cindex machine directives, AArch64
341 @cindex AArch64 machine directives
344 @c AAAAAAAAAAAAAAAAAAAAAAAAA
346 @cindex @code{.arch} directive, AArch64
347 @item .arch @var{name}
348 Select the target architecture. Valid values for @var{name} are the same as
349 for the @option{-march} command-line option.
351 Specifying @code{.arch} clears any previously selected architecture
354 @cindex @code{.arch_extension} directive, AArch64
355 @item .arch_extension @var{name}
356 Add or remove an architecture extension to the target architecture. Valid
357 values for @var{name} are the same as those accepted as architectural
358 extensions by the @option{-mcpu} command-line option.
360 @code{.arch_extension} may be used multiple times to add or remove extensions
361 incrementally to the architecture being compiled for.
363 @c BBBBBBBBBBBBBBBBBBBBBBBBBB
365 @cindex @code{.bss} directive, AArch64
367 This directive switches to the @code{.bss} section.
369 @c CCCCCCCCCCCCCCCCCCCCCCCCCC
371 @cindex @code{.cpu} directive, AArch64
372 @item .cpu @var{name}
373 Set the target processor. Valid values for @var{name} are the same as
374 those accepted by the @option{-mcpu=} command-line option.
376 @c DDDDDDDDDDDDDDDDDDDDDDDDDD
378 @cindex @code{.dword} directive, AArch64
379 @item .dword @var{expressions}
380 The @code{.dword} directive produces 64 bit values.
382 @c EEEEEEEEEEEEEEEEEEEEEEEEEE
384 @cindex @code{.even} directive, AArch64
386 The @code{.even} directive aligns the output on the next even byte
389 @c FFFFFFFFFFFFFFFFFFFFFFFFFF
391 @cindex @code{.float16} directive, AArch64
392 @item .float16 @var{value [,...,value_n]}
393 Place the half precision floating point representation of one or more
394 floating-point values into the current section.
395 The format used to encode the floating point values is always the
396 IEEE 754-2008 half precision floating point format.
398 @c GGGGGGGGGGGGGGGGGGGGGGGGGG
399 @c HHHHHHHHHHHHHHHHHHHHHHHHHH
400 @c IIIIIIIIIIIIIIIIIIIIIIIIII
402 @cindex @code{.inst} directive, AArch64
403 @item .inst @var{expressions}
404 Inserts the expressions into the output as if they were instructions,
407 @c JJJJJJJJJJJJJJJJJJJJJJJJJJ
408 @c KKKKKKKKKKKKKKKKKKKKKKKKKK
409 @c LLLLLLLLLLLLLLLLLLLLLLLLLL
411 @cindex @code{.ltorg} directive, AArch64
413 This directive causes the current contents of the literal pool to be
414 dumped into the current section (which is assumed to be the .text
415 section) at the current location (aligned to a word boundary).
416 GAS maintains a separate literal pool for each section and each
417 sub-section. The @code{.ltorg} directive will only affect the literal
418 pool of the current section and sub-section. At the end of assembly
419 all remaining, un-empty literal pools will automatically be dumped.
421 Note - older versions of GAS would dump the current literal
422 pool any time a section change occurred. This is no longer done, since
423 it prevents accurate control of the placement of literal pools.
425 @c MMMMMMMMMMMMMMMMMMMMMMMMMM
427 @c NNNNNNNNNNNNNNNNNNNNNNNNNN
428 @c OOOOOOOOOOOOOOOOOOOOOOOOOO
430 @c PPPPPPPPPPPPPPPPPPPPPPPPPP
432 @cindex @code{.pool} directive, AArch64
434 This is a synonym for .ltorg.
436 @c QQQQQQQQQQQQQQQQQQQQQQQQQQ
437 @c RRRRRRRRRRRRRRRRRRRRRRRRRR
439 @cindex @code{.req} directive, AArch64
440 @item @var{name} .req @var{register name}
441 This creates an alias for @var{register name} called @var{name}. For
448 ip0, ip1, lr and fp are automatically defined to
449 alias to X16, X17, X30 and X29 respectively.
451 @c SSSSSSSSSSSSSSSSSSSSSSSSSS
453 @c TTTTTTTTTTTTTTTTTTTTTTTTTT
455 @cindex @code{.tlsdescadd} directive, AArch64
456 @item @code{.tlsdescadd}
457 Emits a TLSDESC_ADD reloc on the next instruction.
459 @cindex @code{.tlsdesccall} directive, AArch64
460 @item @code{.tlsdesccall}
461 Emits a TLSDESC_CALL reloc on the next instruction.
463 @cindex @code{.tlsdescldr} directive, AArch64
464 @item @code{.tlsdescldr}
465 Emits a TLSDESC_LDR reloc on the next instruction.
467 @c UUUUUUUUUUUUUUUUUUUUUUUUUU
469 @cindex @code{.unreq} directive, AArch64
470 @item .unreq @var{alias-name}
471 This undefines a register alias which was previously defined using the
472 @code{req} directive. For example:
479 An error occurs if the name is undefined. Note - this pseudo op can
480 be used to delete builtin in register name aliases (eg 'w0'). This
481 should only be done if it is really necessary.
483 @c VVVVVVVVVVVVVVVVVVVVVVVVVV
485 @cindex @code{.variant_pcs} directive, AArch64
486 @item .variant_pcs @var{symbol}
487 This directive marks @var{symbol} referencing a function that may
488 follow a variant procedure call standard with different register
489 usage convention from the base procedure call standard.
491 @c WWWWWWWWWWWWWWWWWWWWWWWWWW
492 @c XXXXXXXXXXXXXXXXXXXXXXXXXX
494 @cindex @code{.xword} directive, AArch64
495 @item .xword @var{expressions}
496 The @code{.xword} directive produces 64 bit values. This is the same
497 as the @code{.dword} directive.
499 @c YYYYYYYYYYYYYYYYYYYYYYYYYY
500 @c ZZZZZZZZZZZZZZZZZZZZZZZZZZ
502 @cindex @code{.cfi_b_key_frame} directive, AArch64
503 @item @code{.cfi_b_key_frame}
504 The @code{.cfi_b_key_frame} directive inserts a 'B' character into the CIE
505 corresponding to the current frame's FDE, meaning that its return address has
506 been signed with the B-key. If two frames are signed with differing keys then
507 they will not share the same CIE. This information is intended to be used by
508 the stack unwinder in order to properly authenticate return addresses.
512 @node AArch64 Opcodes
515 @cindex AArch64 opcodes
516 @cindex opcodes for AArch64
517 GAS implements all the standard AArch64 opcodes. It also
518 implements several pseudo opcodes, including several synthetic load
523 @cindex @code{LDR reg,=<expr>} pseudo op, AArch64
526 ldr <register> , =<expression>
529 The constant expression will be placed into the nearest literal pool (if it not
530 already there) and a PC-relative LDR instruction will be generated.
534 For more information on the AArch64 instruction set and assembly language
535 notation, see @samp{ARMv8 Instruction Set Overview} available at
536 @uref{http://infocenter.arm.com}.
539 @node AArch64 Mapping Symbols
540 @section Mapping Symbols
542 The AArch64 ELF specification requires that special symbols be inserted
543 into object files to mark certain features:
549 At the start of a region of code containing AArch64 instructions.
553 At the start of a region of data.