Commit | Line | Data |
---|---|---|
2571583a | 1 | @c Copyright (C) 1996-2017 Free Software Foundation, Inc. |
252b5132 RH |
2 | @c This is part of the GAS manual. |
3 | @c For copying conditions, see the file as.texinfo. | |
4 | ||
5 | @ifset GENERIC | |
6 | @page | |
7 | @node ARM-Dependent | |
8 | @chapter ARM Dependent Features | |
9 | @end ifset | |
10 | ||
11 | @ifclear GENERIC | |
12 | @node Machine Dependencies | |
13 | @chapter ARM Dependent Features | |
14 | @end ifclear | |
15 | ||
16 | @cindex ARM support | |
17 | @cindex Thumb support | |
18 | @menu | |
19 | * ARM Options:: Options | |
20 | * ARM Syntax:: Syntax | |
21 | * ARM Floating Point:: Floating Point | |
22 | * ARM Directives:: ARM Machine Directives | |
23 | * ARM Opcodes:: Opcodes | |
6057a28f | 24 | * ARM Mapping Symbols:: Mapping Symbols |
7da4f750 | 25 | * ARM Unwinding Tutorial:: Unwinding |
252b5132 RH |
26 | @end menu |
27 | ||
28 | @node ARM Options | |
29 | @section Options | |
30 | @cindex ARM options (none) | |
31 | @cindex options for ARM (none) | |
adcf07e6 | 32 | |
252b5132 | 33 | @table @code |
adcf07e6 | 34 | |
03b1477f | 35 | @cindex @code{-mcpu=} command line option, ARM |
92081f48 | 36 | @item -mcpu=@var{processor}[+@var{extension}@dots{}] |
252b5132 RH |
37 | This option specifies the target processor. The assembler will issue an |
38 | error message if an attempt is made to assemble an instruction which | |
03b1477f | 39 | will not execute on the target processor. The following processor names are |
34bca508 | 40 | recognized: |
03b1477f RE |
41 | @code{arm1}, |
42 | @code{arm2}, | |
43 | @code{arm250}, | |
44 | @code{arm3}, | |
45 | @code{arm6}, | |
46 | @code{arm60}, | |
47 | @code{arm600}, | |
48 | @code{arm610}, | |
49 | @code{arm620}, | |
50 | @code{arm7}, | |
51 | @code{arm7m}, | |
52 | @code{arm7d}, | |
53 | @code{arm7dm}, | |
54 | @code{arm7di}, | |
55 | @code{arm7dmi}, | |
56 | @code{arm70}, | |
57 | @code{arm700}, | |
58 | @code{arm700i}, | |
59 | @code{arm710}, | |
60 | @code{arm710t}, | |
61 | @code{arm720}, | |
62 | @code{arm720t}, | |
63 | @code{arm740t}, | |
64 | @code{arm710c}, | |
65 | @code{arm7100}, | |
66 | @code{arm7500}, | |
67 | @code{arm7500fe}, | |
68 | @code{arm7t}, | |
69 | @code{arm7tdmi}, | |
1ff4677c | 70 | @code{arm7tdmi-s}, |
03b1477f RE |
71 | @code{arm8}, |
72 | @code{arm810}, | |
73 | @code{strongarm}, | |
74 | @code{strongarm1}, | |
75 | @code{strongarm110}, | |
76 | @code{strongarm1100}, | |
77 | @code{strongarm1110}, | |
78 | @code{arm9}, | |
79 | @code{arm920}, | |
80 | @code{arm920t}, | |
81 | @code{arm922t}, | |
82 | @code{arm940t}, | |
83 | @code{arm9tdmi}, | |
7fac0536 NC |
84 | @code{fa526} (Faraday FA526 processor), |
85 | @code{fa626} (Faraday FA626 processor), | |
03b1477f | 86 | @code{arm9e}, |
7de9afa2 | 87 | @code{arm926e}, |
1ff4677c | 88 | @code{arm926ej-s}, |
03b1477f RE |
89 | @code{arm946e-r0}, |
90 | @code{arm946e}, | |
db8ac8f9 | 91 | @code{arm946e-s}, |
03b1477f RE |
92 | @code{arm966e-r0}, |
93 | @code{arm966e}, | |
db8ac8f9 PB |
94 | @code{arm966e-s}, |
95 | @code{arm968e-s}, | |
03b1477f | 96 | @code{arm10t}, |
db8ac8f9 | 97 | @code{arm10tdmi}, |
03b1477f RE |
98 | @code{arm10e}, |
99 | @code{arm1020}, | |
100 | @code{arm1020t}, | |
7de9afa2 | 101 | @code{arm1020e}, |
db8ac8f9 | 102 | @code{arm1022e}, |
1ff4677c | 103 | @code{arm1026ej-s}, |
4a58c4bd NC |
104 | @code{fa606te} (Faraday FA606TE processor), |
105 | @code{fa616te} (Faraday FA616TE processor), | |
7fac0536 | 106 | @code{fa626te} (Faraday FA626TE processor), |
4a58c4bd | 107 | @code{fmp626} (Faraday FMP626 processor), |
7fac0536 | 108 | @code{fa726te} (Faraday FA726TE processor), |
1ff4677c RE |
109 | @code{arm1136j-s}, |
110 | @code{arm1136jf-s}, | |
db8ac8f9 PB |
111 | @code{arm1156t2-s}, |
112 | @code{arm1156t2f-s}, | |
0dd132b6 NC |
113 | @code{arm1176jz-s}, |
114 | @code{arm1176jzf-s}, | |
115 | @code{mpcore}, | |
116 | @code{mpcorenovfp}, | |
b38f9f31 | 117 | @code{cortex-a5}, |
c90460e4 | 118 | @code{cortex-a7}, |
62b3e311 | 119 | @code{cortex-a8}, |
15290f0a | 120 | @code{cortex-a9}, |
dbb1f804 | 121 | @code{cortex-a15}, |
ed5491b9 | 122 | @code{cortex-a17}, |
6735952f | 123 | @code{cortex-a32}, |
43cdc0a8 | 124 | @code{cortex-a35}, |
4469186b KT |
125 | @code{cortex-a53}, |
126 | @code{cortex-a57}, | |
127 | @code{cortex-a72}, | |
362a3eba | 128 | @code{cortex-a73}, |
62b3e311 | 129 | @code{cortex-r4}, |
307c948d | 130 | @code{cortex-r4f}, |
70a8bc5b | 131 | @code{cortex-r5}, |
132 | @code{cortex-r7}, | |
5f474010 | 133 | @code{cortex-r8}, |
b19ea8d2 | 134 | @code{cortex-m33}, |
ce1b0a45 | 135 | @code{cortex-m23}, |
a715796b | 136 | @code{cortex-m7}, |
7ef07ba0 | 137 | @code{cortex-m4}, |
62b3e311 | 138 | @code{cortex-m3}, |
5b19eaba NC |
139 | @code{cortex-m1}, |
140 | @code{cortex-m0}, | |
ce32bd10 | 141 | @code{cortex-m0plus}, |
246496bb | 142 | @code{exynos-m1}, |
ea0d6bb9 PT |
143 | @code{marvell-pj4}, |
144 | @code{marvell-whitney}, | |
145 | @code{xgene1}, | |
146 | @code{xgene2}, | |
03b1477f RE |
147 | @code{ep9312} (ARM920 with Cirrus Maverick coprocessor), |
148 | @code{i80200} (Intel XScale processor) | |
e16bb312 | 149 | @code{iwmmxt} (Intel(r) XScale processor with Wireless MMX(tm) technology coprocessor) |
03b1477f | 150 | and |
34bca508 | 151 | @code{xscale}. |
03b1477f RE |
152 | The special name @code{all} may be used to allow the |
153 | assembler to accept instructions valid for any ARM processor. | |
154 | ||
34bca508 L |
155 | In addition to the basic instruction set, the assembler can be told to |
156 | accept various extension mnemonics that extend the processor using the | |
03b1477f | 157 | co-processor instruction space. For example, @code{-mcpu=arm920+maverick} |
34bca508 | 158 | is equivalent to specifying @code{-mcpu=ep9312}. |
69133863 | 159 | |
34bca508 | 160 | Multiple extensions may be specified, separated by a @code{+}. The |
69133863 MGD |
161 | extensions should be specified in ascending alphabetical order. |
162 | ||
34bca508 | 163 | Some extensions may be restricted to particular architectures; this is |
60e5ef9f MGD |
164 | documented in the list of extensions below. |
165 | ||
34bca508 L |
166 | Extension mnemonics may also be removed from those the assembler accepts. |
167 | This is done be prepending @code{no} to the option that adds the extension. | |
168 | Extensions that are removed should be listed after all extensions which have | |
169 | been added, again in ascending alphabetical order. For example, | |
69133863 MGD |
170 | @code{-mcpu=ep9312+nomaverick} is equivalent to specifying @code{-mcpu=arm920}. |
171 | ||
172 | ||
eea54501 | 173 | The following extensions are currently supported: |
ea0d6bb9 | 174 | @code{crc} |
bca38921 MGD |
175 | @code{crypto} (Cryptography Extensions for v8-A architecture, implies @code{fp+simd}), |
176 | @code{fp} (Floating Point Extensions for v8-A architecture), | |
177 | @code{idiv} (Integer Divide Extensions for v7-A and v7-R architectures), | |
69133863 MGD |
178 | @code{iwmmxt}, |
179 | @code{iwmmxt2}, | |
ea0d6bb9 | 180 | @code{xscale}, |
69133863 | 181 | @code{maverick}, |
ea0d6bb9 PT |
182 | @code{mp} (Multiprocessing Extensions for v7-A and v7-R |
183 | architectures), | |
b2a5fbdc | 184 | @code{os} (Operating System for v6M architecture), |
f4c65163 | 185 | @code{sec} (Security Extensions for v6K and v7-A architectures), |
bca38921 | 186 | @code{simd} (Advanced SIMD Extensions for v8-A architecture, implies @code{fp}), |
34bca508 | 187 | @code{virt} (Virtualization Extensions for v7-A architecture, implies |
90ec0d68 | 188 | @code{idiv}), |
33eaf5de | 189 | @code{pan} (Privileged Access Never Extensions for v8-A architecture), |
4d1464f2 MW |
190 | @code{ras} (Reliability, Availability and Serviceability extensions |
191 | for v8-A architecture), | |
d6b4b13e MW |
192 | @code{rdma} (ARMv8.1 Advanced SIMD extensions for v8-A architecture, implies |
193 | @code{simd}) | |
03b1477f | 194 | and |
69133863 | 195 | @code{xscale}. |
03b1477f RE |
196 | |
197 | @cindex @code{-march=} command line option, ARM | |
92081f48 | 198 | @item -march=@var{architecture}[+@var{extension}@dots{}] |
252b5132 RH |
199 | This option specifies the target architecture. The assembler will issue |
200 | an error message if an attempt is made to assemble an instruction which | |
34bca508 L |
201 | will not execute on the target architecture. The following architecture |
202 | names are recognized: | |
03b1477f RE |
203 | @code{armv1}, |
204 | @code{armv2}, | |
205 | @code{armv2a}, | |
206 | @code{armv2s}, | |
207 | @code{armv3}, | |
208 | @code{armv3m}, | |
209 | @code{armv4}, | |
210 | @code{armv4xm}, | |
211 | @code{armv4t}, | |
212 | @code{armv4txm}, | |
213 | @code{armv5}, | |
214 | @code{armv5t}, | |
215 | @code{armv5txm}, | |
216 | @code{armv5te}, | |
09d92015 | 217 | @code{armv5texp}, |
c5f98204 | 218 | @code{armv6}, |
1ddd7f43 | 219 | @code{armv6j}, |
0dd132b6 NC |
220 | @code{armv6k}, |
221 | @code{armv6z}, | |
f33026a9 | 222 | @code{armv6kz}, |
b2a5fbdc MGD |
223 | @code{armv6-m}, |
224 | @code{armv6s-m}, | |
62b3e311 | 225 | @code{armv7}, |
c450d570 | 226 | @code{armv7-a}, |
c9fb6e58 | 227 | @code{armv7ve}, |
c450d570 PB |
228 | @code{armv7-r}, |
229 | @code{armv7-m}, | |
9e3c6df6 | 230 | @code{armv7e-m}, |
bca38921 | 231 | @code{armv8-a}, |
a5932920 | 232 | @code{armv8.1-a}, |
56a1b672 | 233 | @code{armv8.2-a}, |
a12fd8e1 | 234 | @code{armv8.3-a}, |
e16bb312 | 235 | @code{iwmmxt} |
ea0d6bb9 | 236 | @code{iwmmxt2} |
03b1477f RE |
237 | and |
238 | @code{xscale}. | |
239 | If both @code{-mcpu} and | |
240 | @code{-march} are specified, the assembler will use | |
241 | the setting for @code{-mcpu}. | |
242 | ||
243 | The architecture option can be extended with the same instruction set | |
244 | extension options as the @code{-mcpu} option. | |
245 | ||
246 | @cindex @code{-mfpu=} command line option, ARM | |
247 | @item -mfpu=@var{floating-point-format} | |
248 | ||
249 | This option specifies the floating point format to assemble for. The | |
250 | assembler will issue an error message if an attempt is made to assemble | |
34bca508 | 251 | an instruction which will not execute on the target floating point unit. |
03b1477f RE |
252 | The following format options are recognized: |
253 | @code{softfpa}, | |
254 | @code{fpe}, | |
bc89618b RE |
255 | @code{fpe2}, |
256 | @code{fpe3}, | |
03b1477f RE |
257 | @code{fpa}, |
258 | @code{fpa10}, | |
259 | @code{fpa11}, | |
260 | @code{arm7500fe}, | |
261 | @code{softvfp}, | |
262 | @code{softvfp+vfp}, | |
263 | @code{vfp}, | |
264 | @code{vfp10}, | |
265 | @code{vfp10-r0}, | |
266 | @code{vfp9}, | |
267 | @code{vfpxd}, | |
62f3b8c8 PB |
268 | @code{vfpv2}, |
269 | @code{vfpv3}, | |
270 | @code{vfpv3-fp16}, | |
271 | @code{vfpv3-d16}, | |
272 | @code{vfpv3-d16-fp16}, | |
273 | @code{vfpv3xd}, | |
274 | @code{vfpv3xd-d16}, | |
275 | @code{vfpv4}, | |
276 | @code{vfpv4-d16}, | |
f0cd0667 | 277 | @code{fpv4-sp-d16}, |
a715796b TG |
278 | @code{fpv5-sp-d16}, |
279 | @code{fpv5-d16}, | |
bca38921 | 280 | @code{fp-armv8}, |
09d92015 MM |
281 | @code{arm1020t}, |
282 | @code{arm1020e}, | |
b1cc4aeb | 283 | @code{arm1136jf-s}, |
62f3b8c8 PB |
284 | @code{maverick}, |
285 | @code{neon}, | |
d5e0ba9c RE |
286 | @code{neon-vfpv3}, |
287 | @code{neon-fp16}, | |
bca38921 MGD |
288 | @code{neon-vfpv4}, |
289 | @code{neon-fp-armv8}, | |
081e4c7d MW |
290 | @code{crypto-neon-fp-armv8}, |
291 | @code{neon-fp-armv8.1} | |
d6b4b13e | 292 | and |
081e4c7d | 293 | @code{crypto-neon-fp-armv8.1}. |
03b1477f RE |
294 | |
295 | In addition to determining which instructions are assembled, this option | |
296 | also affects the way in which the @code{.double} assembler directive behaves | |
297 | when assembling little-endian code. | |
298 | ||
34bca508 | 299 | The default is dependent on the processor selected. For Architecture 5 or |
d5e0ba9c | 300 | later, the default is to assemble for VFP instructions; for earlier |
03b1477f | 301 | architectures the default is to assemble for FPA instructions. |
adcf07e6 | 302 | |
252b5132 RH |
303 | @cindex @code{-mthumb} command line option, ARM |
304 | @item -mthumb | |
03b1477f | 305 | This option specifies that the assembler should start assembling Thumb |
34bca508 | 306 | instructions; that is, it should behave as though the file starts with a |
03b1477f | 307 | @code{.code 16} directive. |
adcf07e6 | 308 | |
252b5132 RH |
309 | @cindex @code{-mthumb-interwork} command line option, ARM |
310 | @item -mthumb-interwork | |
311 | This option specifies that the output generated by the assembler should | |
312 | be marked as supporting interworking. | |
adcf07e6 | 313 | |
52970753 NC |
314 | @cindex @code{-mimplicit-it} command line option, ARM |
315 | @item -mimplicit-it=never | |
316 | @itemx -mimplicit-it=always | |
317 | @itemx -mimplicit-it=arm | |
318 | @itemx -mimplicit-it=thumb | |
319 | The @code{-mimplicit-it} option controls the behavior of the assembler when | |
320 | conditional instructions are not enclosed in IT blocks. | |
321 | There are four possible behaviors. | |
322 | If @code{never} is specified, such constructs cause a warning in ARM | |
323 | code and an error in Thumb-2 code. | |
324 | If @code{always} is specified, such constructs are accepted in both | |
325 | ARM and Thumb-2 code, where the IT instruction is added implicitly. | |
326 | If @code{arm} is specified, such constructs are accepted in ARM code | |
327 | and cause an error in Thumb-2 code. | |
328 | If @code{thumb} is specified, such constructs cause a warning in ARM | |
329 | code and are accepted in Thumb-2 code. If you omit this option, the | |
330 | behavior is equivalent to @code{-mimplicit-it=arm}. | |
e07e6e58 | 331 | |
5a5829dd NS |
332 | @cindex @code{-mapcs-26} command line option, ARM |
333 | @cindex @code{-mapcs-32} command line option, ARM | |
334 | @item -mapcs-26 | |
335 | @itemx -mapcs-32 | |
336 | These options specify that the output generated by the assembler should | |
252b5132 RH |
337 | be marked as supporting the indicated version of the Arm Procedure. |
338 | Calling Standard. | |
adcf07e6 | 339 | |
077b8428 NC |
340 | @cindex @code{-matpcs} command line option, ARM |
341 | @item -matpcs | |
34bca508 | 342 | This option specifies that the output generated by the assembler should |
077b8428 NC |
343 | be marked as supporting the Arm/Thumb Procedure Calling Standard. If |
344 | enabled this option will cause the assembler to create an empty | |
345 | debugging section in the object file called .arm.atpcs. Debuggers can | |
346 | use this to determine the ABI being used by. | |
347 | ||
adcf07e6 | 348 | @cindex @code{-mapcs-float} command line option, ARM |
252b5132 | 349 | @item -mapcs-float |
1be59579 | 350 | This indicates the floating point variant of the APCS should be |
252b5132 | 351 | used. In this variant floating point arguments are passed in FP |
550262c4 | 352 | registers rather than integer registers. |
adcf07e6 NC |
353 | |
354 | @cindex @code{-mapcs-reentrant} command line option, ARM | |
252b5132 RH |
355 | @item -mapcs-reentrant |
356 | This indicates that the reentrant variant of the APCS should be used. | |
357 | This variant supports position independent code. | |
adcf07e6 | 358 | |
33a392fb PB |
359 | @cindex @code{-mfloat-abi=} command line option, ARM |
360 | @item -mfloat-abi=@var{abi} | |
361 | This option specifies that the output generated by the assembler should be | |
362 | marked as using specified floating point ABI. | |
363 | The following values are recognized: | |
364 | @code{soft}, | |
365 | @code{softfp} | |
366 | and | |
367 | @code{hard}. | |
368 | ||
d507cf36 PB |
369 | @cindex @code{-eabi=} command line option, ARM |
370 | @item -meabi=@var{ver} | |
371 | This option specifies which EABI version the produced object files should | |
372 | conform to. | |
b45619c0 | 373 | The following values are recognized: |
3a4a14e9 PB |
374 | @code{gnu}, |
375 | @code{4} | |
d507cf36 | 376 | and |
3a4a14e9 | 377 | @code{5}. |
d507cf36 | 378 | |
252b5132 RH |
379 | @cindex @code{-EB} command line option, ARM |
380 | @item -EB | |
381 | This option specifies that the output generated by the assembler should | |
382 | be marked as being encoded for a big-endian processor. | |
adcf07e6 | 383 | |
080bb7bb NC |
384 | Note: If a program is being built for a system with big-endian data |
385 | and little-endian instructions then it should be assembled with the | |
386 | @option{-EB} option, (all of it, code and data) and then linked with | |
387 | the @option{--be8} option. This will reverse the endianness of the | |
388 | instructions back to little-endian, but leave the data as big-endian. | |
389 | ||
252b5132 RH |
390 | @cindex @code{-EL} command line option, ARM |
391 | @item -EL | |
392 | This option specifies that the output generated by the assembler should | |
393 | be marked as being encoded for a little-endian processor. | |
adcf07e6 | 394 | |
252b5132 RH |
395 | @cindex @code{-k} command line option, ARM |
396 | @cindex PIC code generation for ARM | |
397 | @item -k | |
a349d9dd PB |
398 | This option specifies that the output of the assembler should be marked |
399 | as position-independent code (PIC). | |
adcf07e6 | 400 | |
845b51d6 PB |
401 | @cindex @code{--fix-v4bx} command line option, ARM |
402 | @item --fix-v4bx | |
403 | Allow @code{BX} instructions in ARMv4 code. This is intended for use with | |
404 | the linker option of the same name. | |
405 | ||
278df34e NS |
406 | @cindex @code{-mwarn-deprecated} command line option, ARM |
407 | @item -mwarn-deprecated | |
408 | @itemx -mno-warn-deprecated | |
409 | Enable or disable warnings about using deprecated options or | |
410 | features. The default is to warn. | |
411 | ||
2e6976a8 DG |
412 | @cindex @code{-mccs} command line option, ARM |
413 | @item -mccs | |
414 | Turns on CodeComposer Studio assembly syntax compatibility mode. | |
415 | ||
8b2d793c NC |
416 | @cindex @code{-mwarn-syms} command line option, ARM |
417 | @item -mwarn-syms | |
418 | @itemx -mno-warn-syms | |
419 | Enable or disable warnings about symbols that match the names of ARM | |
420 | instructions. The default is to warn. | |
421 | ||
252b5132 RH |
422 | @end table |
423 | ||
424 | ||
425 | @node ARM Syntax | |
426 | @section Syntax | |
427 | @menu | |
cab7e4d9 | 428 | * ARM-Instruction-Set:: Instruction Set |
252b5132 RH |
429 | * ARM-Chars:: Special Characters |
430 | * ARM-Regs:: Register Names | |
b6895b4f | 431 | * ARM-Relocations:: Relocations |
99f1a7a7 | 432 | * ARM-Neon-Alignment:: NEON Alignment Specifiers |
252b5132 RH |
433 | @end menu |
434 | ||
cab7e4d9 NC |
435 | @node ARM-Instruction-Set |
436 | @subsection Instruction Set Syntax | |
437 | Two slightly different syntaxes are support for ARM and THUMB | |
438 | instructions. The default, @code{divided}, uses the old style where | |
439 | ARM and THUMB instructions had their own, separate syntaxes. The new, | |
440 | @code{unified} syntax, which can be selected via the @code{.syntax} | |
441 | directive, and has the following main features: | |
442 | ||
9e6f3811 AS |
443 | @itemize @bullet |
444 | @item | |
cab7e4d9 NC |
445 | Immediate operands do not require a @code{#} prefix. |
446 | ||
9e6f3811 | 447 | @item |
cab7e4d9 NC |
448 | The @code{IT} instruction may appear, and if it does it is validated |
449 | against subsequent conditional affixes. In ARM mode it does not | |
450 | generate machine code, in THUMB mode it does. | |
451 | ||
9e6f3811 | 452 | @item |
cab7e4d9 NC |
453 | For ARM instructions the conditional affixes always appear at the end |
454 | of the instruction. For THUMB instructions conditional affixes can be | |
455 | used, but only inside the scope of an @code{IT} instruction. | |
456 | ||
9e6f3811 | 457 | @item |
cab7e4d9 NC |
458 | All of the instructions new to the V6T2 architecture (and later) are |
459 | available. (Only a few such instructions can be written in the | |
460 | @code{divided} syntax). | |
461 | ||
9e6f3811 | 462 | @item |
cab7e4d9 NC |
463 | The @code{.N} and @code{.W} suffixes are recognized and honored. |
464 | ||
9e6f3811 | 465 | @item |
cab7e4d9 NC |
466 | All instructions set the flags if and only if they have an @code{s} |
467 | affix. | |
9e6f3811 | 468 | @end itemize |
cab7e4d9 | 469 | |
252b5132 RH |
470 | @node ARM-Chars |
471 | @subsection Special Characters | |
472 | ||
473 | @cindex line comment character, ARM | |
474 | @cindex ARM line comment character | |
7c31ae13 NC |
475 | The presence of a @samp{@@} anywhere on a line indicates the start of |
476 | a comment that extends to the end of that line. | |
477 | ||
478 | If a @samp{#} appears as the first character of a line then the whole | |
479 | line is treated as a comment, but in this case the line could also be | |
480 | a logical line number directive (@pxref{Comments}) or a preprocessor | |
481 | control command (@pxref{Preprocessing}). | |
550262c4 NC |
482 | |
483 | @cindex line separator, ARM | |
484 | @cindex statement separator, ARM | |
485 | @cindex ARM line separator | |
a349d9dd PB |
486 | The @samp{;} character can be used instead of a newline to separate |
487 | statements. | |
550262c4 NC |
488 | |
489 | @cindex immediate character, ARM | |
490 | @cindex ARM immediate character | |
491 | Either @samp{#} or @samp{$} can be used to indicate immediate operands. | |
252b5132 RH |
492 | |
493 | @cindex identifiers, ARM | |
494 | @cindex ARM identifiers | |
495 | *TODO* Explain about /data modifier on symbols. | |
496 | ||
497 | @node ARM-Regs | |
498 | @subsection Register Names | |
499 | ||
500 | @cindex ARM register names | |
501 | @cindex register names, ARM | |
502 | *TODO* Explain about ARM register naming, and the predefined names. | |
503 | ||
b6895b4f PB |
504 | @node ARM-Relocations |
505 | @subsection ARM relocation generation | |
506 | ||
507 | @cindex data relocations, ARM | |
508 | @cindex ARM data relocations | |
509 | Specific data relocations can be generated by putting the relocation name | |
510 | in parentheses after the symbol name. For example: | |
511 | ||
512 | @smallexample | |
513 | .word foo(TARGET1) | |
514 | @end smallexample | |
515 | ||
516 | This will generate an @samp{R_ARM_TARGET1} relocation against the symbol | |
517 | @var{foo}. | |
518 | The following relocations are supported: | |
519 | @code{GOT}, | |
520 | @code{GOTOFF}, | |
521 | @code{TARGET1}, | |
522 | @code{TARGET2}, | |
523 | @code{SBREL}, | |
524 | @code{TLSGD}, | |
525 | @code{TLSLDM}, | |
526 | @code{TLSLDO}, | |
0855e32b NS |
527 | @code{TLSDESC}, |
528 | @code{TLSCALL}, | |
b43420e6 NC |
529 | @code{GOTTPOFF}, |
530 | @code{GOT_PREL} | |
b6895b4f PB |
531 | and |
532 | @code{TPOFF}. | |
533 | ||
534 | For compatibility with older toolchains the assembler also accepts | |
3da1d841 NC |
535 | @code{(PLT)} after branch targets. On legacy targets this will |
536 | generate the deprecated @samp{R_ARM_PLT32} relocation. On EABI | |
537 | targets it will encode either the @samp{R_ARM_CALL} or | |
538 | @samp{R_ARM_JUMP24} relocation, as appropriate. | |
b6895b4f PB |
539 | |
540 | @cindex MOVW and MOVT relocations, ARM | |
541 | Relocations for @samp{MOVW} and @samp{MOVT} instructions can be generated | |
542 | by prefixing the value with @samp{#:lower16:} and @samp{#:upper16} | |
b45619c0 | 543 | respectively. For example to load the 32-bit address of foo into r0: |
252b5132 | 544 | |
b6895b4f PB |
545 | @smallexample |
546 | MOVW r0, #:lower16:foo | |
547 | MOVT r0, #:upper16:foo | |
548 | @end smallexample | |
252b5132 | 549 | |
72d98d16 MG |
550 | Relocations @samp{R_ARM_THM_ALU_ABS_G0_NC}, @samp{R_ARM_THM_ALU_ABS_G1_NC}, |
551 | @samp{R_ARM_THM_ALU_ABS_G2_NC} and @samp{R_ARM_THM_ALU_ABS_G3_NC} can be | |
552 | generated by prefixing the value with @samp{#:lower0_7:#}, | |
553 | @samp{#:lower8_15:#}, @samp{#:upper0_7:#} and @samp{#:upper8_15:#} | |
554 | respectively. For example to load the 32-bit address of foo into r0: | |
555 | ||
556 | @smallexample | |
557 | MOVS r0, #:upper8_15:#foo | |
558 | LSLS r0, r0, #8 | |
559 | ADDS r0, #:upper0_7:#foo | |
560 | LSLS r0, r0, #8 | |
561 | ADDS r0, #:lower8_15:#foo | |
562 | LSLS r0, r0, #8 | |
563 | ADDS r0, #:lower0_7:#foo | |
564 | @end smallexample | |
565 | ||
ba724cfc NC |
566 | @node ARM-Neon-Alignment |
567 | @subsection NEON Alignment Specifiers | |
568 | ||
569 | @cindex alignment for NEON instructions | |
570 | Some NEON load/store instructions allow an optional address | |
571 | alignment qualifier. | |
572 | The ARM documentation specifies that this is indicated by | |
573 | @samp{@@ @var{align}}. However GAS already interprets | |
574 | the @samp{@@} character as a "line comment" start, | |
575 | so @samp{: @var{align}} is used instead. For example: | |
576 | ||
577 | @smallexample | |
578 | vld1.8 @{q0@}, [r0, :128] | |
579 | @end smallexample | |
580 | ||
581 | @node ARM Floating Point | |
582 | @section Floating Point | |
583 | ||
584 | @cindex floating point, ARM (@sc{ieee}) | |
585 | @cindex ARM floating point (@sc{ieee}) | |
586 | The ARM family uses @sc{ieee} floating-point numbers. | |
587 | ||
252b5132 RH |
588 | @node ARM Directives |
589 | @section ARM Machine Directives | |
590 | ||
591 | @cindex machine directives, ARM | |
592 | @cindex ARM machine directives | |
593 | @table @code | |
594 | ||
4a6bc624 NS |
595 | @c AAAAAAAAAAAAAAAAAAAAAAAAA |
596 | ||
2b841ec2 | 597 | @ifclear ELF |
4a6bc624 NS |
598 | @cindex @code{.2byte} directive, ARM |
599 | @cindex @code{.4byte} directive, ARM | |
600 | @cindex @code{.8byte} directive, ARM | |
601 | @item .2byte @var{expression} [, @var{expression}]* | |
602 | @itemx .4byte @var{expression} [, @var{expression}]* | |
603 | @itemx .8byte @var{expression} [, @var{expression}]* | |
604 | These directives write 2, 4 or 8 byte values to the output section. | |
2b841ec2 | 605 | @end ifclear |
4a6bc624 NS |
606 | |
607 | @cindex @code{.align} directive, ARM | |
adcf07e6 NC |
608 | @item .align @var{expression} [, @var{expression}] |
609 | This is the generic @var{.align} directive. For the ARM however if the | |
610 | first argument is zero (ie no alignment is needed) the assembler will | |
611 | behave as if the argument had been 2 (ie pad to the next four byte | |
062b7c0c | 612 | boundary). This is for compatibility with ARM's own assembler. |
adcf07e6 | 613 | |
4a6bc624 NS |
614 | @cindex @code{.arch} directive, ARM |
615 | @item .arch @var{name} | |
616 | Select the target architecture. Valid values for @var{name} are the same as | |
617 | for the @option{-march} commandline option. | |
252b5132 | 618 | |
34bca508 | 619 | Specifying @code{.arch} clears any previously selected architecture |
69133863 MGD |
620 | extensions. |
621 | ||
622 | @cindex @code{.arch_extension} directive, ARM | |
623 | @item .arch_extension @var{name} | |
34bca508 L |
624 | Add or remove an architecture extension to the target architecture. Valid |
625 | values for @var{name} are the same as those accepted as architectural | |
69133863 MGD |
626 | extensions by the @option{-mcpu} commandline option. |
627 | ||
628 | @code{.arch_extension} may be used multiple times to add or remove extensions | |
629 | incrementally to the architecture being compiled for. | |
630 | ||
4a6bc624 NS |
631 | @cindex @code{.arm} directive, ARM |
632 | @item .arm | |
633 | This performs the same action as @var{.code 32}. | |
252b5132 | 634 | |
4a6bc624 | 635 | @c BBBBBBBBBBBBBBBBBBBBBBBBBB |
0bbf2aa4 | 636 | |
4a6bc624 NS |
637 | @cindex @code{.bss} directive, ARM |
638 | @item .bss | |
639 | This directive switches to the @code{.bss} section. | |
0bbf2aa4 | 640 | |
4a6bc624 NS |
641 | @c CCCCCCCCCCCCCCCCCCCCCCCCCC |
642 | ||
643 | @cindex @code{.cantunwind} directive, ARM | |
644 | @item .cantunwind | |
645 | Prevents unwinding through the current function. No personality routine | |
646 | or exception table data is required or permitted. | |
647 | ||
648 | @cindex @code{.code} directive, ARM | |
649 | @item .code @code{[16|32]} | |
650 | This directive selects the instruction set being generated. The value 16 | |
651 | selects Thumb, with the value 32 selecting ARM. | |
652 | ||
653 | @cindex @code{.cpu} directive, ARM | |
654 | @item .cpu @var{name} | |
655 | Select the target processor. Valid values for @var{name} are the same as | |
656 | for the @option{-mcpu} commandline option. | |
657 | ||
34bca508 | 658 | Specifying @code{.cpu} clears any previously selected architecture |
69133863 MGD |
659 | extensions. |
660 | ||
4a6bc624 NS |
661 | @c DDDDDDDDDDDDDDDDDDDDDDDDDD |
662 | ||
663 | @cindex @code{.dn} and @code{.qn} directives, ARM | |
f467aa98 | 664 | @item @var{name} .dn @var{register name} [@var{.type}] [[@var{index}]] |
1f9bb1ca | 665 | @itemx @var{name} .qn @var{register name} [@var{.type}] [[@var{index}]] |
23753660 JB |
666 | |
667 | The @code{dn} and @code{qn} directives are used to create typed | |
668 | and/or indexed register aliases for use in Advanced SIMD Extension | |
669 | (Neon) instructions. The former should be used to create aliases | |
670 | of double-precision registers, and the latter to create aliases of | |
671 | quad-precision registers. | |
672 | ||
673 | If these directives are used to create typed aliases, those aliases can | |
674 | be used in Neon instructions instead of writing types after the mnemonic | |
675 | or after each operand. For example: | |
676 | ||
677 | @smallexample | |
678 | x .dn d2.f32 | |
679 | y .dn d3.f32 | |
680 | z .dn d4.f32[1] | |
681 | vmul x,y,z | |
682 | @end smallexample | |
683 | ||
684 | This is equivalent to writing the following: | |
685 | ||
686 | @smallexample | |
687 | vmul.f32 d2,d3,d4[1] | |
688 | @end smallexample | |
689 | ||
690 | Aliases created using @code{dn} or @code{qn} can be destroyed using | |
691 | @code{unreq}. | |
692 | ||
4a6bc624 | 693 | @c EEEEEEEEEEEEEEEEEEEEEEEEEE |
252b5132 | 694 | |
4a6bc624 NS |
695 | @cindex @code{.eabi_attribute} directive, ARM |
696 | @item .eabi_attribute @var{tag}, @var{value} | |
697 | Set the EABI object attribute @var{tag} to @var{value}. | |
252b5132 | 698 | |
4a6bc624 NS |
699 | The @var{tag} is either an attribute number, or one of the following: |
700 | @code{Tag_CPU_raw_name}, @code{Tag_CPU_name}, @code{Tag_CPU_arch}, | |
701 | @code{Tag_CPU_arch_profile}, @code{Tag_ARM_ISA_use}, | |
75375b3e | 702 | @code{Tag_THUMB_ISA_use}, @code{Tag_FP_arch}, @code{Tag_WMMX_arch}, |
4a6bc624 NS |
703 | @code{Tag_Advanced_SIMD_arch}, @code{Tag_PCS_config}, |
704 | @code{Tag_ABI_PCS_R9_use}, @code{Tag_ABI_PCS_RW_data}, | |
705 | @code{Tag_ABI_PCS_RO_data}, @code{Tag_ABI_PCS_GOT_use}, | |
706 | @code{Tag_ABI_PCS_wchar_t}, @code{Tag_ABI_FP_rounding}, | |
707 | @code{Tag_ABI_FP_denormal}, @code{Tag_ABI_FP_exceptions}, | |
708 | @code{Tag_ABI_FP_user_exceptions}, @code{Tag_ABI_FP_number_model}, | |
75375b3e | 709 | @code{Tag_ABI_align_needed}, @code{Tag_ABI_align_preserved}, |
4a6bc624 NS |
710 | @code{Tag_ABI_enum_size}, @code{Tag_ABI_HardFP_use}, |
711 | @code{Tag_ABI_VFP_args}, @code{Tag_ABI_WMMX_args}, | |
712 | @code{Tag_ABI_optimization_goals}, @code{Tag_ABI_FP_optimization_goals}, | |
713 | @code{Tag_compatibility}, @code{Tag_CPU_unaligned_access}, | |
75375b3e | 714 | @code{Tag_FP_HP_extension}, @code{Tag_ABI_FP_16bit_format}, |
cd21e546 | 715 | @code{Tag_MPextension_use}, @code{Tag_DIV_use}, |
4a6bc624 NS |
716 | @code{Tag_nodefaults}, @code{Tag_also_compatible_with}, |
717 | @code{Tag_conformance}, @code{Tag_T2EE_use}, | |
cd21e546 | 718 | @code{Tag_Virtualization_use} |
4a6bc624 NS |
719 | |
720 | The @var{value} is either a @code{number}, @code{"string"}, or | |
721 | @code{number, "string"} depending on the tag. | |
722 | ||
75375b3e | 723 | Note - the following legacy values are also accepted by @var{tag}: |
34bca508 | 724 | @code{Tag_VFP_arch}, @code{Tag_ABI_align8_needed}, |
75375b3e MGD |
725 | @code{Tag_ABI_align8_preserved}, @code{Tag_VFP_HP_extension}, |
726 | ||
4a6bc624 NS |
727 | @cindex @code{.even} directive, ARM |
728 | @item .even | |
729 | This directive aligns to an even-numbered address. | |
730 | ||
731 | @cindex @code{.extend} directive, ARM | |
732 | @cindex @code{.ldouble} directive, ARM | |
733 | @item .extend @var{expression} [, @var{expression}]* | |
734 | @itemx .ldouble @var{expression} [, @var{expression}]* | |
735 | These directives write 12byte long double floating-point values to the | |
736 | output section. These are not compatible with current ARM processors | |
737 | or ABIs. | |
738 | ||
739 | @c FFFFFFFFFFFFFFFFFFFFFFFFFF | |
740 | ||
741 | @anchor{arm_fnend} | |
742 | @cindex @code{.fnend} directive, ARM | |
743 | @item .fnend | |
744 | Marks the end of a function with an unwind table entry. The unwind index | |
745 | table entry is created when this directive is processed. | |
252b5132 | 746 | |
4a6bc624 NS |
747 | If no personality routine has been specified then standard personality |
748 | routine 0 or 1 will be used, depending on the number of unwind opcodes | |
749 | required. | |
750 | ||
751 | @anchor{arm_fnstart} | |
752 | @cindex @code{.fnstart} directive, ARM | |
753 | @item .fnstart | |
754 | Marks the start of a function with an unwind table entry. | |
755 | ||
756 | @cindex @code{.force_thumb} directive, ARM | |
252b5132 RH |
757 | @item .force_thumb |
758 | This directive forces the selection of Thumb instructions, even if the | |
759 | target processor does not support those instructions | |
760 | ||
4a6bc624 NS |
761 | @cindex @code{.fpu} directive, ARM |
762 | @item .fpu @var{name} | |
763 | Select the floating-point unit to assemble for. Valid values for @var{name} | |
764 | are the same as for the @option{-mfpu} commandline option. | |
252b5132 | 765 | |
4a6bc624 NS |
766 | @c GGGGGGGGGGGGGGGGGGGGGGGGGG |
767 | @c HHHHHHHHHHHHHHHHHHHHHHHHHH | |
e1da3f5b | 768 | |
4a6bc624 NS |
769 | @cindex @code{.handlerdata} directive, ARM |
770 | @item .handlerdata | |
771 | Marks the end of the current function, and the start of the exception table | |
772 | entry for that function. Anything between this directive and the | |
773 | @code{.fnend} directive will be added to the exception table entry. | |
774 | ||
775 | Must be preceded by a @code{.personality} or @code{.personalityindex} | |
776 | directive. | |
777 | ||
778 | @c IIIIIIIIIIIIIIIIIIIIIIIIII | |
c921be7d NC |
779 | |
780 | @cindex @code{.inst} directive, ARM | |
781 | @item .inst @var{opcode} [ , @dots{} ] | |
1f9bb1ca AS |
782 | @itemx .inst.n @var{opcode} [ , @dots{} ] |
783 | @itemx .inst.w @var{opcode} [ , @dots{} ] | |
c921be7d NC |
784 | Generates the instruction corresponding to the numerical value @var{opcode}. |
785 | @code{.inst.n} and @code{.inst.w} allow the Thumb instruction size to be | |
786 | specified explicitly, overriding the normal encoding rules. | |
787 | ||
4a6bc624 NS |
788 | @c JJJJJJJJJJJJJJJJJJJJJJJJJJ |
789 | @c KKKKKKKKKKKKKKKKKKKKKKKKKK | |
790 | @c LLLLLLLLLLLLLLLLLLLLLLLLLL | |
791 | ||
792 | @item .ldouble @var{expression} [, @var{expression}]* | |
793 | See @code{.extend}. | |
5395a469 | 794 | |
252b5132 RH |
795 | @cindex @code{.ltorg} directive, ARM |
796 | @item .ltorg | |
797 | This directive causes the current contents of the literal pool to be | |
798 | dumped into the current section (which is assumed to be the .text | |
799 | section) at the current location (aligned to a word boundary). | |
3d0c9500 NC |
800 | @code{GAS} maintains a separate literal pool for each section and each |
801 | sub-section. The @code{.ltorg} directive will only affect the literal | |
802 | pool of the current section and sub-section. At the end of assembly | |
803 | all remaining, un-empty literal pools will automatically be dumped. | |
804 | ||
805 | Note - older versions of @code{GAS} would dump the current literal | |
806 | pool any time a section change occurred. This is no longer done, since | |
807 | it prevents accurate control of the placement of literal pools. | |
252b5132 | 808 | |
4a6bc624 | 809 | @c MMMMMMMMMMMMMMMMMMMMMMMMMM |
252b5132 | 810 | |
4a6bc624 NS |
811 | @cindex @code{.movsp} directive, ARM |
812 | @item .movsp @var{reg} [, #@var{offset}] | |
813 | Tell the unwinder that @var{reg} contains an offset from the current | |
814 | stack pointer. If @var{offset} is not specified then it is assumed to be | |
815 | zero. | |
7ed4c4c5 | 816 | |
4a6bc624 NS |
817 | @c NNNNNNNNNNNNNNNNNNNNNNNNNN |
818 | @c OOOOOOOOOOOOOOOOOOOOOOOOOO | |
7ed4c4c5 | 819 | |
4a6bc624 NS |
820 | @cindex @code{.object_arch} directive, ARM |
821 | @item .object_arch @var{name} | |
822 | Override the architecture recorded in the EABI object attribute section. | |
823 | Valid values for @var{name} are the same as for the @code{.arch} directive. | |
824 | Typically this is useful when code uses runtime detection of CPU features. | |
7ed4c4c5 | 825 | |
4a6bc624 NS |
826 | @c PPPPPPPPPPPPPPPPPPPPPPPPPP |
827 | ||
828 | @cindex @code{.packed} directive, ARM | |
829 | @item .packed @var{expression} [, @var{expression}]* | |
830 | This directive writes 12-byte packed floating-point values to the | |
831 | output section. These are not compatible with current ARM processors | |
832 | or ABIs. | |
833 | ||
ea4cff4f | 834 | @anchor{arm_pad} |
4a6bc624 NS |
835 | @cindex @code{.pad} directive, ARM |
836 | @item .pad #@var{count} | |
837 | Generate unwinder annotations for a stack adjustment of @var{count} bytes. | |
838 | A positive value indicates the function prologue allocated stack space by | |
839 | decrementing the stack pointer. | |
7ed4c4c5 NC |
840 | |
841 | @cindex @code{.personality} directive, ARM | |
842 | @item .personality @var{name} | |
843 | Sets the personality routine for the current function to @var{name}. | |
844 | ||
845 | @cindex @code{.personalityindex} directive, ARM | |
846 | @item .personalityindex @var{index} | |
847 | Sets the personality routine for the current function to the EABI standard | |
848 | routine number @var{index} | |
849 | ||
4a6bc624 NS |
850 | @cindex @code{.pool} directive, ARM |
851 | @item .pool | |
852 | This is a synonym for .ltorg. | |
7ed4c4c5 | 853 | |
4a6bc624 NS |
854 | @c QQQQQQQQQQQQQQQQQQQQQQQQQQ |
855 | @c RRRRRRRRRRRRRRRRRRRRRRRRRR | |
856 | ||
857 | @cindex @code{.req} directive, ARM | |
858 | @item @var{name} .req @var{register name} | |
859 | This creates an alias for @var{register name} called @var{name}. For | |
860 | example: | |
861 | ||
862 | @smallexample | |
863 | foo .req r0 | |
864 | @end smallexample | |
865 | ||
866 | @c SSSSSSSSSSSSSSSSSSSSSSSSSS | |
7ed4c4c5 | 867 | |
7da4f750 | 868 | @anchor{arm_save} |
7ed4c4c5 NC |
869 | @cindex @code{.save} directive, ARM |
870 | @item .save @var{reglist} | |
871 | Generate unwinder annotations to restore the registers in @var{reglist}. | |
872 | The format of @var{reglist} is the same as the corresponding store-multiple | |
873 | instruction. | |
874 | ||
875 | @smallexample | |
876 | @exdent @emph{core registers} | |
877 | .save @{r4, r5, r6, lr@} | |
878 | stmfd sp!, @{r4, r5, r6, lr@} | |
879 | @exdent @emph{FPA registers} | |
880 | .save f4, 2 | |
881 | sfmfd f4, 2, [sp]! | |
882 | @exdent @emph{VFP registers} | |
883 | .save @{d8, d9, d10@} | |
fa073d69 | 884 | fstmdx sp!, @{d8, d9, d10@} |
7ed4c4c5 NC |
885 | @exdent @emph{iWMMXt registers} |
886 | .save @{wr10, wr11@} | |
887 | wstrd wr11, [sp, #-8]! | |
888 | wstrd wr10, [sp, #-8]! | |
889 | or | |
890 | .save wr11 | |
891 | wstrd wr11, [sp, #-8]! | |
892 | .save wr10 | |
893 | wstrd wr10, [sp, #-8]! | |
894 | @end smallexample | |
895 | ||
7da4f750 | 896 | @anchor{arm_setfp} |
7ed4c4c5 NC |
897 | @cindex @code{.setfp} directive, ARM |
898 | @item .setfp @var{fpreg}, @var{spreg} [, #@var{offset}] | |
4a6bc624 | 899 | Make all unwinder annotations relative to a frame pointer. Without this |
7ed4c4c5 NC |
900 | the unwinder will use offsets from the stack pointer. |
901 | ||
a5b82cbe | 902 | The syntax of this directive is the same as the @code{add} or @code{mov} |
7ed4c4c5 NC |
903 | instruction used to set the frame pointer. @var{spreg} must be either |
904 | @code{sp} or mentioned in a previous @code{.movsp} directive. | |
905 | ||
906 | @smallexample | |
907 | .movsp ip | |
908 | mov ip, sp | |
909 | @dots{} | |
910 | .setfp fp, ip, #4 | |
a5b82cbe | 911 | add fp, ip, #4 |
7ed4c4c5 NC |
912 | @end smallexample |
913 | ||
4a6bc624 NS |
914 | @cindex @code{.secrel32} directive, ARM |
915 | @item .secrel32 @var{expression} [, @var{expression}]* | |
916 | This directive emits relocations that evaluate to the section-relative | |
917 | offset of each expression's symbol. This directive is only supported | |
918 | for PE targets. | |
919 | ||
cab7e4d9 NC |
920 | @cindex @code{.syntax} directive, ARM |
921 | @item .syntax [@code{unified} | @code{divided}] | |
922 | This directive sets the Instruction Set Syntax as described in the | |
923 | @ref{ARM-Instruction-Set} section. | |
924 | ||
4a6bc624 NS |
925 | @c TTTTTTTTTTTTTTTTTTTTTTTTTT |
926 | ||
927 | @cindex @code{.thumb} directive, ARM | |
928 | @item .thumb | |
929 | This performs the same action as @var{.code 16}. | |
930 | ||
931 | @cindex @code{.thumb_func} directive, ARM | |
932 | @item .thumb_func | |
933 | This directive specifies that the following symbol is the name of a | |
934 | Thumb encoded function. This information is necessary in order to allow | |
935 | the assembler and linker to generate correct code for interworking | |
936 | between Arm and Thumb instructions and should be used even if | |
937 | interworking is not going to be performed. The presence of this | |
938 | directive also implies @code{.thumb} | |
939 | ||
33eaf5de | 940 | This directive is not necessary when generating EABI objects. On these |
4a6bc624 NS |
941 | targets the encoding is implicit when generating Thumb code. |
942 | ||
943 | @cindex @code{.thumb_set} directive, ARM | |
944 | @item .thumb_set | |
945 | This performs the equivalent of a @code{.set} directive in that it | |
946 | creates a symbol which is an alias for another symbol (possibly not yet | |
947 | defined). This directive also has the added property in that it marks | |
948 | the aliased symbol as being a thumb function entry point, in the same | |
949 | way that the @code{.thumb_func} directive does. | |
950 | ||
0855e32b NS |
951 | @cindex @code{.tlsdescseq} directive, ARM |
952 | @item .tlsdescseq @var{tls-variable} | |
953 | This directive is used to annotate parts of an inlined TLS descriptor | |
954 | trampoline. Normally the trampoline is provided by the linker, and | |
955 | this directive is not needed. | |
956 | ||
4a6bc624 NS |
957 | @c UUUUUUUUUUUUUUUUUUUUUUUUUU |
958 | ||
959 | @cindex @code{.unreq} directive, ARM | |
960 | @item .unreq @var{alias-name} | |
961 | This undefines a register alias which was previously defined using the | |
962 | @code{req}, @code{dn} or @code{qn} directives. For example: | |
963 | ||
964 | @smallexample | |
965 | foo .req r0 | |
966 | .unreq foo | |
967 | @end smallexample | |
968 | ||
969 | An error occurs if the name is undefined. Note - this pseudo op can | |
970 | be used to delete builtin in register name aliases (eg 'r0'). This | |
971 | should only be done if it is really necessary. | |
972 | ||
7ed4c4c5 | 973 | @cindex @code{.unwind_raw} directive, ARM |
4a6bc624 | 974 | @item .unwind_raw @var{offset}, @var{byte1}, @dots{} |
33eaf5de | 975 | Insert one of more arbitrary unwind opcode bytes, which are known to adjust |
7ed4c4c5 NC |
976 | the stack pointer by @var{offset} bytes. |
977 | ||
978 | For example @code{.unwind_raw 4, 0xb1, 0x01} is equivalent to | |
979 | @code{.save @{r0@}} | |
980 | ||
4a6bc624 | 981 | @c VVVVVVVVVVVVVVVVVVVVVVVVVV |
ee065d83 | 982 | |
4a6bc624 NS |
983 | @cindex @code{.vsave} directive, ARM |
984 | @item .vsave @var{vfp-reglist} | |
985 | Generate unwinder annotations to restore the VFP registers in @var{vfp-reglist} | |
986 | using FLDMD. Also works for VFPv3 registers | |
987 | that are to be restored using VLDM. | |
988 | The format of @var{vfp-reglist} is the same as the corresponding store-multiple | |
989 | instruction. | |
ee065d83 | 990 | |
4a6bc624 NS |
991 | @smallexample |
992 | @exdent @emph{VFP registers} | |
993 | .vsave @{d8, d9, d10@} | |
994 | fstmdd sp!, @{d8, d9, d10@} | |
995 | @exdent @emph{VFPv3 registers} | |
996 | .vsave @{d15, d16, d17@} | |
997 | vstm sp!, @{d15, d16, d17@} | |
998 | @end smallexample | |
e04befd0 | 999 | |
4a6bc624 NS |
1000 | Since FLDMX and FSTMX are now deprecated, this directive should be |
1001 | used in favour of @code{.save} for saving VFP registers for ARMv6 and above. | |
e04befd0 | 1002 | |
4a6bc624 NS |
1003 | @c WWWWWWWWWWWWWWWWWWWWWWWWWW |
1004 | @c XXXXXXXXXXXXXXXXXXXXXXXXXX | |
1005 | @c YYYYYYYYYYYYYYYYYYYYYYYYYY | |
1006 | @c ZZZZZZZZZZZZZZZZZZZZZZZZZZ | |
ee065d83 | 1007 | |
252b5132 RH |
1008 | @end table |
1009 | ||
1010 | @node ARM Opcodes | |
1011 | @section Opcodes | |
1012 | ||
1013 | @cindex ARM opcodes | |
1014 | @cindex opcodes for ARM | |
49a5575c NC |
1015 | @code{@value{AS}} implements all the standard ARM opcodes. It also |
1016 | implements several pseudo opcodes, including several synthetic load | |
34bca508 | 1017 | instructions. |
252b5132 | 1018 | |
49a5575c NC |
1019 | @table @code |
1020 | ||
1021 | @cindex @code{NOP} pseudo op, ARM | |
1022 | @item NOP | |
1023 | @smallexample | |
1024 | nop | |
1025 | @end smallexample | |
252b5132 | 1026 | |
49a5575c NC |
1027 | This pseudo op will always evaluate to a legal ARM instruction that does |
1028 | nothing. Currently it will evaluate to MOV r0, r0. | |
252b5132 | 1029 | |
49a5575c | 1030 | @cindex @code{LDR reg,=<label>} pseudo op, ARM |
34bca508 | 1031 | @item LDR |
252b5132 RH |
1032 | @smallexample |
1033 | ldr <register> , = <expression> | |
1034 | @end smallexample | |
1035 | ||
1036 | If expression evaluates to a numeric constant then a MOV or MVN | |
1037 | instruction will be used in place of the LDR instruction, if the | |
1038 | constant can be generated by either of these instructions. Otherwise | |
1039 | the constant will be placed into the nearest literal pool (if it not | |
1040 | already there) and a PC relative LDR instruction will be generated. | |
1041 | ||
49a5575c NC |
1042 | @cindex @code{ADR reg,<label>} pseudo op, ARM |
1043 | @item ADR | |
1044 | @smallexample | |
1045 | adr <register> <label> | |
1046 | @end smallexample | |
1047 | ||
1048 | This instruction will load the address of @var{label} into the indicated | |
1049 | register. The instruction will evaluate to a PC relative ADD or SUB | |
1050 | instruction depending upon where the label is located. If the label is | |
1051 | out of range, or if it is not defined in the same file (and section) as | |
1052 | the ADR instruction, then an error will be generated. This instruction | |
1053 | will not make use of the literal pool. | |
1054 | ||
1055 | @cindex @code{ADRL reg,<label>} pseudo op, ARM | |
34bca508 | 1056 | @item ADRL |
49a5575c NC |
1057 | @smallexample |
1058 | adrl <register> <label> | |
1059 | @end smallexample | |
1060 | ||
1061 | This instruction will load the address of @var{label} into the indicated | |
a349d9dd | 1062 | register. The instruction will evaluate to one or two PC relative ADD |
49a5575c NC |
1063 | or SUB instructions depending upon where the label is located. If a |
1064 | second instruction is not needed a NOP instruction will be generated in | |
1065 | its place, so that this instruction is always 8 bytes long. | |
1066 | ||
1067 | If the label is out of range, or if it is not defined in the same file | |
1068 | (and section) as the ADRL instruction, then an error will be generated. | |
1069 | This instruction will not make use of the literal pool. | |
1070 | ||
1071 | @end table | |
1072 | ||
252b5132 RH |
1073 | For information on the ARM or Thumb instruction sets, see @cite{ARM |
1074 | Software Development Toolkit Reference Manual}, Advanced RISC Machines | |
1075 | Ltd. | |
1076 | ||
6057a28f NC |
1077 | @node ARM Mapping Symbols |
1078 | @section Mapping Symbols | |
1079 | ||
1080 | The ARM ELF specification requires that special symbols be inserted | |
1081 | into object files to mark certain features: | |
1082 | ||
1083 | @table @code | |
1084 | ||
1085 | @cindex @code{$a} | |
1086 | @item $a | |
1087 | At the start of a region of code containing ARM instructions. | |
1088 | ||
1089 | @cindex @code{$t} | |
1090 | @item $t | |
1091 | At the start of a region of code containing THUMB instructions. | |
1092 | ||
1093 | @cindex @code{$d} | |
1094 | @item $d | |
1095 | At the start of a region of data. | |
1096 | ||
1097 | @end table | |
1098 | ||
1099 | The assembler will automatically insert these symbols for you - there | |
1100 | is no need to code them yourself. Support for tagging symbols ($b, | |
1101 | $f, $p and $m) which is also mentioned in the current ARM ELF | |
1102 | specification is not implemented. This is because they have been | |
1103 | dropped from the new EABI and so tools cannot rely upon their | |
1104 | presence. | |
1105 | ||
7da4f750 MM |
1106 | @node ARM Unwinding Tutorial |
1107 | @section Unwinding | |
1108 | ||
1109 | The ABI for the ARM Architecture specifies a standard format for | |
1110 | exception unwind information. This information is used when an | |
1111 | exception is thrown to determine where control should be transferred. | |
1112 | In particular, the unwind information is used to determine which | |
1113 | function called the function that threw the exception, and which | |
1114 | function called that one, and so forth. This information is also used | |
1115 | to restore the values of callee-saved registers in the function | |
1116 | catching the exception. | |
1117 | ||
1118 | If you are writing functions in assembly code, and those functions | |
1119 | call other functions that throw exceptions, you must use assembly | |
1120 | pseudo ops to ensure that appropriate exception unwind information is | |
1121 | generated. Otherwise, if one of the functions called by your assembly | |
1122 | code throws an exception, the run-time library will be unable to | |
1123 | unwind the stack through your assembly code and your program will not | |
1124 | behave correctly. | |
1125 | ||
1126 | To illustrate the use of these pseudo ops, we will examine the code | |
1127 | that G++ generates for the following C++ input: | |
1128 | ||
1129 | @verbatim | |
1130 | void callee (int *); | |
1131 | ||
34bca508 L |
1132 | int |
1133 | caller () | |
7da4f750 MM |
1134 | { |
1135 | int i; | |
1136 | callee (&i); | |
34bca508 | 1137 | return i; |
7da4f750 MM |
1138 | } |
1139 | @end verbatim | |
1140 | ||
1141 | This example does not show how to throw or catch an exception from | |
1142 | assembly code. That is a much more complex operation and should | |
1143 | always be done in a high-level language, such as C++, that directly | |
1144 | supports exceptions. | |
1145 | ||
1146 | The code generated by one particular version of G++ when compiling the | |
1147 | example above is: | |
1148 | ||
1149 | @verbatim | |
1150 | _Z6callerv: | |
1151 | .fnstart | |
1152 | .LFB2: | |
1153 | @ Function supports interworking. | |
1154 | @ args = 0, pretend = 0, frame = 8 | |
1155 | @ frame_needed = 1, uses_anonymous_args = 0 | |
1156 | stmfd sp!, {fp, lr} | |
1157 | .save {fp, lr} | |
1158 | .LCFI0: | |
1159 | .setfp fp, sp, #4 | |
1160 | add fp, sp, #4 | |
1161 | .LCFI1: | |
1162 | .pad #8 | |
1163 | sub sp, sp, #8 | |
1164 | .LCFI2: | |
1165 | sub r3, fp, #8 | |
1166 | mov r0, r3 | |
1167 | bl _Z6calleePi | |
1168 | ldr r3, [fp, #-8] | |
1169 | mov r0, r3 | |
1170 | sub sp, fp, #4 | |
1171 | ldmfd sp!, {fp, lr} | |
1172 | bx lr | |
1173 | .LFE2: | |
1174 | .fnend | |
1175 | @end verbatim | |
1176 | ||
1177 | Of course, the sequence of instructions varies based on the options | |
1178 | you pass to GCC and on the version of GCC in use. The exact | |
1179 | instructions are not important since we are focusing on the pseudo ops | |
1180 | that are used to generate unwind information. | |
1181 | ||
1182 | An important assumption made by the unwinder is that the stack frame | |
1183 | does not change during the body of the function. In particular, since | |
1184 | we assume that the assembly code does not itself throw an exception, | |
1185 | the only point where an exception can be thrown is from a call, such | |
1186 | as the @code{bl} instruction above. At each call site, the same saved | |
1187 | registers (including @code{lr}, which indicates the return address) | |
1188 | must be located in the same locations relative to the frame pointer. | |
1189 | ||
1190 | The @code{.fnstart} (@pxref{arm_fnstart,,.fnstart pseudo op}) pseudo | |
1191 | op appears immediately before the first instruction of the function | |
1192 | while the @code{.fnend} (@pxref{arm_fnend,,.fnend pseudo op}) pseudo | |
1193 | op appears immediately after the last instruction of the function. | |
34bca508 | 1194 | These pseudo ops specify the range of the function. |
7da4f750 MM |
1195 | |
1196 | Only the order of the other pseudos ops (e.g., @code{.setfp} or | |
1197 | @code{.pad}) matters; their exact locations are irrelevant. In the | |
1198 | example above, the compiler emits the pseudo ops with particular | |
1199 | instructions. That makes it easier to understand the code, but it is | |
1200 | not required for correctness. It would work just as well to emit all | |
1201 | of the pseudo ops other than @code{.fnend} in the same order, but | |
1202 | immediately after @code{.fnstart}. | |
1203 | ||
1204 | The @code{.save} (@pxref{arm_save,,.save pseudo op}) pseudo op | |
1205 | indicates registers that have been saved to the stack so that they can | |
1206 | be restored before the function returns. The argument to the | |
1207 | @code{.save} pseudo op is a list of registers to save. If a register | |
1208 | is ``callee-saved'' (as specified by the ABI) and is modified by the | |
1209 | function you are writing, then your code must save the value before it | |
1210 | is modified and restore the original value before the function | |
1211 | returns. If an exception is thrown, the run-time library restores the | |
1212 | values of these registers from their locations on the stack before | |
1213 | returning control to the exception handler. (Of course, if an | |
1214 | exception is not thrown, the function that contains the @code{.save} | |
1215 | pseudo op restores these registers in the function epilogue, as is | |
1216 | done with the @code{ldmfd} instruction above.) | |
1217 | ||
1218 | You do not have to save callee-saved registers at the very beginning | |
1219 | of the function and you do not need to use the @code{.save} pseudo op | |
1220 | immediately following the point at which the registers are saved. | |
1221 | However, if you modify a callee-saved register, you must save it on | |
1222 | the stack before modifying it and before calling any functions which | |
1223 | might throw an exception. And, you must use the @code{.save} pseudo | |
1224 | op to indicate that you have done so. | |
1225 | ||
1226 | The @code{.pad} (@pxref{arm_pad,,.pad}) pseudo op indicates a | |
1227 | modification of the stack pointer that does not save any registers. | |
1228 | The argument is the number of bytes (in decimal) that are subtracted | |
1229 | from the stack pointer. (On ARM CPUs, the stack grows downwards, so | |
1230 | subtracting from the stack pointer increases the size of the stack.) | |
1231 | ||
1232 | The @code{.setfp} (@pxref{arm_setfp,,.setfp pseudo op}) pseudo op | |
1233 | indicates the register that contains the frame pointer. The first | |
1234 | argument is the register that is set, which is typically @code{fp}. | |
1235 | The second argument indicates the register from which the frame | |
1236 | pointer takes its value. The third argument, if present, is the value | |
1237 | (in decimal) added to the register specified by the second argument to | |
1238 | compute the value of the frame pointer. You should not modify the | |
1239 | frame pointer in the body of the function. | |
1240 | ||
1241 | If you do not use a frame pointer, then you should not use the | |
1242 | @code{.setfp} pseudo op. If you do not use a frame pointer, then you | |
1243 | should avoid modifying the stack pointer outside of the function | |
1244 | prologue. Otherwise, the run-time library will be unable to find | |
1245 | saved registers when it is unwinding the stack. | |
1246 | ||
1247 | The pseudo ops described above are sufficient for writing assembly | |
1248 | code that calls functions which may throw exceptions. If you need to | |
1249 | know more about the object-file format used to represent unwind | |
1250 | information, you may consult the @cite{Exception Handling ABI for the | |
1251 | ARM Architecture} available from @uref{http://infocenter.arm.com}. | |
91f68a68 | 1252 |