1 @c Copyright 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004
2 @c Free Software Foundation, Inc.
3 @c This is part of the GAS manual.
4 @c For copying conditions, see the file as.texinfo.
9 @chapter ARM Dependent Features
13 @node Machine Dependencies
14 @chapter ARM Dependent Features
20 * ARM Options:: Options
22 * ARM Floating Point:: Floating Point
23 * ARM Directives:: ARM Machine Directives
24 * ARM Opcodes:: Opcodes
25 * ARM Mapping Symbols:: Mapping Symbols
30 @cindex ARM options (none)
31 @cindex options for ARM (none)
35 @cindex @code{-mcpu=} command line option, ARM
36 @item -mcpu=@var{processor}[+@var{extension}@dots{}]
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
39 will not execute on the target processor. The following processor names are
110 @code{ep9312} (ARM920 with Cirrus Maverick coprocessor),
111 @code{i80200} (Intel XScale processor)
112 @code{iwmmxt} (Intel(r) XScale processor with Wireless MMX(tm) technology coprocessor)
115 The special name @code{all} may be used to allow the
116 assembler to accept instructions valid for any ARM processor.
118 In addition to the basic instruction set, the assembler can be told to
119 accept various extension mnemonics that extend the processor using the
120 co-processor instruction space. For example, @code{-mcpu=arm920+maverick}
121 is equivalent to specifying @code{-mcpu=ep9312}. The following extensions
122 are currently supported:
128 @cindex @code{-march=} command line option, ARM
129 @item -march=@var{architecture}[+@var{extension}@dots{}]
130 This option specifies the target architecture. The assembler will issue
131 an error message if an attempt is made to assemble an instruction which
132 will not execute on the target architecture. The following architecture
133 names are recognized:
157 If both @code{-mcpu} and
158 @code{-march} are specified, the assembler will use
159 the setting for @code{-mcpu}.
161 The architecture option can be extended with the same instruction set
162 extension options as the @code{-mcpu} option.
164 @cindex @code{-mfpu=} command line option, ARM
165 @item -mfpu=@var{floating-point-format}
167 This option specifies the floating point format to assemble for. The
168 assembler will issue an error message if an attempt is made to assemble
169 an instruction which will not execute on the target floating point unit.
170 The following format options are recognized:
192 In addition to determining which instructions are assembled, this option
193 also affects the way in which the @code{.double} assembler directive behaves
194 when assembling little-endian code.
196 The default is dependent on the processor selected. For Architecture 5 or
197 later, the default is to assembler for VFP instructions; for earlier
198 architectures the default is to assemble for FPA instructions.
200 @cindex @code{-mthumb} command line option, ARM
202 This option specifies that the assembler should start assembling Thumb
203 instructions; that is, it should behave as though the file starts with a
204 @code{.code 16} directive.
206 @cindex @code{-mthumb-interwork} command line option, ARM
207 @item -mthumb-interwork
208 This option specifies that the output generated by the assembler should
209 be marked as supporting interworking.
211 @cindex @code{-mapcs} command line option, ARM
212 @item -mapcs @code{[26|32]}
213 This option specifies that the output generated by the assembler should
214 be marked as supporting the indicated version of the Arm Procedure.
217 @cindex @code{-matpcs} command line option, ARM
219 This option specifies that the output generated by the assembler should
220 be marked as supporting the Arm/Thumb Procedure Calling Standard. If
221 enabled this option will cause the assembler to create an empty
222 debugging section in the object file called .arm.atpcs. Debuggers can
223 use this to determine the ABI being used by.
225 @cindex @code{-mapcs-float} command line option, ARM
227 This indicates the floating point variant of the APCS should be
228 used. In this variant floating point arguments are passed in FP
229 registers rather than integer registers.
231 @cindex @code{-mapcs-reentrant} command line option, ARM
232 @item -mapcs-reentrant
233 This indicates that the reentrant variant of the APCS should be used.
234 This variant supports position independent code.
236 @cindex @code{-mfloat-abi=} command line option, ARM
237 @item -mfloat-abi=@var{abi}
238 This option specifies that the output generated by the assembler should be
239 marked as using specified floating point ABI.
240 The following values are recognized:
246 @cindex @code{-eabi=} command line option, ARM
247 @item -meabi=@var{ver}
248 This option specifies which EABI version the produced object files should
250 The following values are recognised:
255 @cindex @code{-EB} command line option, ARM
257 This option specifies that the output generated by the assembler should
258 be marked as being encoded for a big-endian processor.
260 @cindex @code{-EL} command line option, ARM
262 This option specifies that the output generated by the assembler should
263 be marked as being encoded for a little-endian processor.
265 @cindex @code{-k} command line option, ARM
266 @cindex PIC code generation for ARM
268 This option specifies that the output of the assembler should be marked
269 as position-independent code (PIC).
277 * ARM-Chars:: Special Characters
278 * ARM-Regs:: Register Names
282 @subsection Special Characters
284 @cindex line comment character, ARM
285 @cindex ARM line comment character
286 The presence of a @samp{@@} on a line indicates the start of a comment
287 that extends to the end of the current line. If a @samp{#} appears as
288 the first character of a line, the whole line is treated as a comment.
290 @cindex line separator, ARM
291 @cindex statement separator, ARM
292 @cindex ARM line separator
293 The @samp{;} character can be used instead of a newline to separate
296 @cindex immediate character, ARM
297 @cindex ARM immediate character
298 Either @samp{#} or @samp{$} can be used to indicate immediate operands.
300 @cindex identifiers, ARM
301 @cindex ARM identifiers
302 *TODO* Explain about /data modifier on symbols.
305 @subsection Register Names
307 @cindex ARM register names
308 @cindex register names, ARM
309 *TODO* Explain about ARM register naming, and the predefined names.
311 @node ARM Floating Point
312 @section Floating Point
314 @cindex floating point, ARM (@sc{ieee})
315 @cindex ARM floating point (@sc{ieee})
316 The ARM family uses @sc{ieee} floating-point numbers.
321 @section ARM Machine Directives
323 @cindex machine directives, ARM
324 @cindex ARM machine directives
327 @cindex @code{align} directive, ARM
328 @item .align @var{expression} [, @var{expression}]
329 This is the generic @var{.align} directive. For the ARM however if the
330 first argument is zero (ie no alignment is needed) the assembler will
331 behave as if the argument had been 2 (ie pad to the next four byte
332 boundary). This is for compatibility with ARM's own assembler.
334 @cindex @code{req} directive, ARM
335 @item @var{name} .req @var{register name}
336 This creates an alias for @var{register name} called @var{name}. For
343 @cindex @code{unreq} directive, ARM
344 @item .unreq @var{alias-name}
345 This undefines a register alias which was previously defined using the
346 @code{req} directive. For example:
353 An error occurs if the name is undefined. Note - this pseudo op can
354 be used to delete builtin in register name aliases (eg 'r0'). This
355 should only be done if it is really necessary.
357 @cindex @code{code} directive, ARM
358 @item .code @code{[16|32]}
359 This directive selects the instruction set being generated. The value 16
360 selects Thumb, with the value 32 selecting ARM.
362 @cindex @code{thumb} directive, ARM
364 This performs the same action as @var{.code 16}.
366 @cindex @code{arm} directive, ARM
368 This performs the same action as @var{.code 32}.
370 @cindex @code{force_thumb} directive, ARM
372 This directive forces the selection of Thumb instructions, even if the
373 target processor does not support those instructions
375 @cindex @code{thumb_func} directive, ARM
377 This directive specifies that the following symbol is the name of a
378 Thumb encoded function. This information is necessary in order to allow
379 the assembler and linker to generate correct code for interworking
380 between Arm and Thumb instructions and should be used even if
381 interworking is not going to be performed. The presence of this
382 directive also implies @code{.thumb}
384 @cindex @code{thumb_set} directive, ARM
386 This performs the equivalent of a @code{.set} directive in that it
387 creates a symbol which is an alias for another symbol (possibly not yet
388 defined). This directive also has the added property in that it marks
389 the aliased symbol as being a thumb function entry point, in the same
390 way that the @code{.thumb_func} directive does.
392 @cindex @code{.ltorg} directive, ARM
394 This directive causes the current contents of the literal pool to be
395 dumped into the current section (which is assumed to be the .text
396 section) at the current location (aligned to a word boundary).
397 @code{GAS} maintains a separate literal pool for each section and each
398 sub-section. The @code{.ltorg} directive will only affect the literal
399 pool of the current section and sub-section. At the end of assembly
400 all remaining, un-empty literal pools will automatically be dumped.
402 Note - older versions of @code{GAS} would dump the current literal
403 pool any time a section change occurred. This is no longer done, since
404 it prevents accurate control of the placement of literal pools.
406 @cindex @code{.pool} directive, ARM
408 This is a synonym for .ltorg.
410 @cindex @code{.fnstart} directive, ARM
411 @item .unwind_fnstart
412 Marks the start of a function with an unwind table entry.
414 @cindex @code{.fnend} directive, ARM
416 Marks the end of a function with an unwind table entry. The unwind index
417 table entry is created when this directive is processed.
419 If no personality routine has been specified then standard personality
420 routine 0 or 1 will be used, depending on the number of unwind opcodes
423 @cindex @code{.cantunwind} directive, ARM
425 Prevents unwinding through the current function. No personality routine
426 or exception table data is required or permitted.
428 @cindex @code{.personality} directive, ARM
429 @item .personality @var{name}
430 Sets the personality routine for the current function to @var{name}.
432 @cindex @code{.personalityindex} directive, ARM
433 @item .personalityindex @var{index}
434 Sets the personality routine for the current function to the EABI standard
435 routine number @var{index}
437 @cindex @code{.handlerdata} directive, ARM
439 Marks the end of the current function, and the start of the exception table
440 entry for that function. Anything between this directive and the
441 @code{.fnend} directive will be added to the exception table entry.
443 Must be preceded by a @code{.personality} or @code{.personalityindex}
446 @cindex @code{.save} directive, ARM
447 @item .save @var{reglist}
448 Generate unwinder annotations to restore the registers in @var{reglist}.
449 The format of @var{reglist} is the same as the corresponding store-multiple
453 @exdent @emph{core registers}
454 .save @{r4, r5, r6, lr@}
455 stmfd sp!, @{r4, r5, r6, lr@}
456 @exdent @emph{FPA registers}
459 @exdent @emph{VFP registers}
460 .save @{d8, d9, d10@}
461 fstmdf sp!, @{d8, d9, d10@}
462 @exdent @emph{iWMMXt registers}
464 wstrd wr11, [sp, #-8]!
465 wstrd wr10, [sp, #-8]!
468 wstrd wr11, [sp, #-8]!
470 wstrd wr10, [sp, #-8]!
473 @cindex @code{.pad} directive, ARM
474 @item .pad #@var{count}
475 Generate unwinder annotations for a stack adjustment of @var{count} bytes.
476 A positive value indicates the function prologue allocated stack space by
477 decrementing the stack pointer.
479 @cindex @code{.movsp} directive, ARM
480 @item .movsp @var{reg}
481 Tell the unwinder that @var{reg} contains the current stack pointer.
483 @cindex @code{.setfp} directive, ARM
484 @item .setfp @var{fpreg}, @var{spreg} [, #@var{offset}]
485 Make all unwinder annotations relaive to a frame pointer. Without this
486 the unwinder will use offsets from the stack pointer.
488 The syntax of this directive is the same as the @code{sub} or @code{mov}
489 instruction used to set the frame pointer. @var{spreg} must be either
490 @code{sp} or mentioned in a previous @code{.movsp} directive.
500 @cindex @code{.unwind_raw} directive, ARM
501 @item .raw @var{offset}, @var{byte1}, @dots{}
502 Insert one of more arbitary unwind opcode bytes, which are known to adjust
503 the stack pointer by @var{offset} bytes.
505 For example @code{.unwind_raw 4, 0xb1, 0x01} is equivalent to
508 @cindex @code{.cpu} directive, ARM
509 @item .cpu @var{name}
510 Select the target processor. Valid values for @var{name} are the same as
511 for the @option{-mcpu} commandline option.
513 @cindex @code{.arch} directive, ARM
514 @item .arch @var{name}
515 Select the target architecture. Valid values for @var{name} are the same as
516 for the @option{-march} commandline option.
518 @cindex @code{.fpu} directive, ARM
519 @item .fpu @var{name}
520 Select the floating point unit to assemble for. Valid values for @var{name}
521 are the same as for the @option{-mfpu} commandline option.
523 @cindex @code{.eabi_attribute} directive, ARM
524 @item .eabi_attribute @var{tag}, @var{value}
525 Set the EABI object attribute number @var{tag} to @var{value}. The value
526 is either a @code{number}, @code{"string"}, or @code{number, "string"}
527 depending on the tag.
535 @cindex opcodes for ARM
536 @code{@value{AS}} implements all the standard ARM opcodes. It also
537 implements several pseudo opcodes, including several synthetic load
542 @cindex @code{NOP} pseudo op, ARM
548 This pseudo op will always evaluate to a legal ARM instruction that does
549 nothing. Currently it will evaluate to MOV r0, r0.
551 @cindex @code{LDR reg,=<label>} pseudo op, ARM
554 ldr <register> , = <expression>
557 If expression evaluates to a numeric constant then a MOV or MVN
558 instruction will be used in place of the LDR instruction, if the
559 constant can be generated by either of these instructions. Otherwise
560 the constant will be placed into the nearest literal pool (if it not
561 already there) and a PC relative LDR instruction will be generated.
563 @cindex @code{ADR reg,<label>} pseudo op, ARM
566 adr <register> <label>
569 This instruction will load the address of @var{label} into the indicated
570 register. The instruction will evaluate to a PC relative ADD or SUB
571 instruction depending upon where the label is located. If the label is
572 out of range, or if it is not defined in the same file (and section) as
573 the ADR instruction, then an error will be generated. This instruction
574 will not make use of the literal pool.
576 @cindex @code{ADRL reg,<label>} pseudo op, ARM
579 adrl <register> <label>
582 This instruction will load the address of @var{label} into the indicated
583 register. The instruction will evaluate to one or two PC relative ADD
584 or SUB instructions depending upon where the label is located. If a
585 second instruction is not needed a NOP instruction will be generated in
586 its place, so that this instruction is always 8 bytes long.
588 If the label is out of range, or if it is not defined in the same file
589 (and section) as the ADRL instruction, then an error will be generated.
590 This instruction will not make use of the literal pool.
594 For information on the ARM or Thumb instruction sets, see @cite{ARM
595 Software Development Toolkit Reference Manual}, Advanced RISC Machines
598 @node ARM Mapping Symbols
599 @section Mapping Symbols
601 The ARM ELF specification requires that special symbols be inserted
602 into object files to mark certain features:
608 At the start of a region of code containing ARM instructions.
612 At the start of a region of code containing THUMB instructions.
616 At the start of a region of data.
620 The assembler will automatically insert these symbols for you - there
621 is no need to code them yourself. Support for tagging symbols ($b,
622 $f, $p and $m) which is also mentioned in the current ARM ELF
623 specification is not implemented. This is because they have been
624 dropped from the new EABI and so tools cannot rely upon their