1 @c Copyright 1996, 1997, 1998, 1999, 2000, 2001, 2003
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
93 @code{ep9312} (ARM920 with Cirrus Maverick coprocessor),
94 @code{i80200} (Intel XScale processor)
95 @code{iwmmxt} (Intel(r) XScale processor with Wireless MMX(tm) technology coprocessor)
98 The special name @code{all} may be used to allow the
99 assembler to accept instructions valid for any ARM processor.
101 In addition to the basic instruction set, the assembler can be told to
102 accept various extension mnemonics that extend the processor using the
103 co-processor instruction space. For example, @code{-mcpu=arm920+maverick}
104 is equivalent to specifying @code{-mcpu=ep9312}. The following extensions
105 are currently supported:
111 @cindex @code{-march=} command line option, ARM
112 @item -march=@var{architecture}[+@var{extension}@dots{}]
113 This option specifies the target architecture. The assembler will issue
114 an error message if an attempt is made to assemble an instruction which
115 will not execute on the target architecture. The following architecture
116 names are recognized:
135 If both @code{-mcpu} and
136 @code{-march} are specified, the assembler will use
137 the setting for @code{-mcpu}.
139 The architecture option can be extended with the same instruction set
140 extension options as the @code{-mcpu} option.
142 @cindex @code{-mfpu=} command line option, ARM
143 @item -mfpu=@var{floating-point-format}
145 This option specifies the floating point format to assemble for. The
146 assembler will issue an error message if an attempt is made to assemble
147 an instruction which will not execute on the target floating point unit.
148 The following format options are recognized:
168 In addition to determining which instructions are assembled, this option
169 also affects the way in which the @code{.double} assembler directive behaves
170 when assembling little-endian code.
172 The default is dependent on the processor selected. For Architecture 5 or
173 later, the default is to assembler for VFP instructions; for earlier
174 architectures the default is to assemble for FPA instructions.
176 @cindex @code{-mthumb} command line option, ARM
178 This option specifies that the assembler should start assembling Thumb
179 instructions; that is, it should behave as though the file starts with a
180 @code{.code 16} directive.
182 @cindex @code{-mthumb-interwork} command line option, ARM
183 @item -mthumb-interwork
184 This option specifies that the output generated by the assembler should
185 be marked as supporting interworking.
187 @cindex @code{-mapcs} command line option, ARM
188 @item -mapcs @code{[26|32]}
189 This option specifies that the output generated by the assembler should
190 be marked as supporting the indicated version of the Arm Procedure.
193 @cindex @code{-matpcs} command line option, ARM
195 This option specifies that the output generated by the assembler should
196 be marked as supporting the Arm/Thumb Procedure Calling Standard. If
197 enabled this option will cause the assembler to create an empty
198 debugging section in the object file called .arm.atpcs. Debuggers can
199 use this to determine the ABI being used by.
201 @cindex @code{-mapcs-float} command line option, ARM
203 This indicates the floating point variant of the APCS should be
204 used. In this variant floating point arguments are passed in FP
205 registers rather than integer registers.
207 @cindex @code{-mapcs-reentrant} command line option, ARM
208 @item -mapcs-reentrant
209 This indicates that the reentrant variant of the APCS should be used.
210 This variant supports position independent code.
212 @cindex @code{-EB} command line option, ARM
214 This option specifies that the output generated by the assembler should
215 be marked as being encoded for a big-endian processor.
217 @cindex @code{-EL} command line option, ARM
219 This option specifies that the output generated by the assembler should
220 be marked as being encoded for a little-endian processor.
222 @cindex @code{-k} command line option, ARM
223 @cindex PIC code generation for ARM
225 This option specifies that the output of the assembler should be marked
226 as position-independent code (PIC).
228 @cindex @code{-moabi} command line option, ARM
230 This indicates that the code should be assembled using the old ARM ELF
231 conventions, based on a beta release release of the ARM-ELF
232 specifications, rather than the default conventions which are based on
233 the final release of the ARM-ELF specifications.
241 * ARM-Chars:: Special Characters
242 * ARM-Regs:: Register Names
246 @subsection Special Characters
248 @cindex line comment character, ARM
249 @cindex ARM line comment character
250 The presence of a @samp{@@} on a line indicates the start of a comment
251 that extends to the end of the current line. If a @samp{#} appears as
252 the first character of a line, the whole line is treated as a comment.
254 @cindex line separator, ARM
255 @cindex statement separator, ARM
256 @cindex ARM line separator
257 The @samp{;} character can be used instead of a newline to separate
260 @cindex immediate character, ARM
261 @cindex ARM immediate character
262 Either @samp{#} or @samp{$} can be used to indicate immediate operands.
264 @cindex identifiers, ARM
265 @cindex ARM identifiers
266 *TODO* Explain about /data modifier on symbols.
269 @subsection Register Names
271 @cindex ARM register names
272 @cindex register names, ARM
273 *TODO* Explain about ARM register naming, and the predefined names.
275 @node ARM Floating Point
276 @section Floating Point
278 @cindex floating point, ARM (@sc{ieee})
279 @cindex ARM floating point (@sc{ieee})
280 The ARM family uses @sc{ieee} floating-point numbers.
285 @section ARM Machine Directives
287 @cindex machine directives, ARM
288 @cindex ARM machine directives
291 @cindex @code{align} directive, ARM
292 @item .align @var{expression} [, @var{expression}]
293 This is the generic @var{.align} directive. For the ARM however if the
294 first argument is zero (ie no alignment is needed) the assembler will
295 behave as if the argument had been 2 (ie pad to the next four byte
296 boundary). This is for compatibility with ARM's own assembler.
298 @cindex @code{req} directive, ARM
299 @item @var{name} .req @var{register name}
300 This creates an alias for @var{register name} called @var{name}. For
307 @cindex @code{unreq} directive, ARM
308 @item .unreq @var{alias-name}
309 This undefines a register alias which was previously defined using the
310 @code{req} directive. For example:
317 An error occurs if the name is undefined. Note - this pseudo op can
318 be used to delete builtin in register name aliases (eg 'r0'). This
319 should only be done if it is really necessary.
321 @cindex @code{code} directive, ARM
322 @item .code @code{[16|32]}
323 This directive selects the instruction set being generated. The value 16
324 selects Thumb, with the value 32 selecting ARM.
326 @cindex @code{thumb} directive, ARM
328 This performs the same action as @var{.code 16}.
330 @cindex @code{arm} directive, ARM
332 This performs the same action as @var{.code 32}.
334 @cindex @code{force_thumb} directive, ARM
336 This directive forces the selection of Thumb instructions, even if the
337 target processor does not support those instructions
339 @cindex @code{thumb_func} directive, ARM
341 This directive specifies that the following symbol is the name of a
342 Thumb encoded function. This information is necessary in order to allow
343 the assembler and linker to generate correct code for interworking
344 between Arm and Thumb instructions and should be used even if
345 interworking is not going to be performed. The presence of this
346 directive also implies @code{.thumb}
348 @cindex @code{thumb_set} directive, ARM
350 This performs the equivalent of a @code{.set} directive in that it
351 creates a symbol which is an alias for another symbol (possibly not yet
352 defined). This directive also has the added property in that it marks
353 the aliased symbol as being a thumb function entry point, in the same
354 way that the @code{.thumb_func} directive does.
356 @cindex @code{.ltorg} directive, ARM
358 This directive causes the current contents of the literal pool to be
359 dumped into the current section (which is assumed to be the .text
360 section) at the current location (aligned to a word boundary).
361 @code{GAS} maintains a separate literal pool for each section and each
362 sub-section. The @code{.ltorg} directive will only affect the literal
363 pool of the current section and sub-section. At the end of assembly
364 all remaining, un-empty literal pools will automatically be dumped.
366 Note - older versions of @code{GAS} would dump the current literal
367 pool any time a section change occurred. This is no longer done, since
368 it prevents accurate control of the placement of literal pools.
370 @cindex @code{.pool} directive, ARM
372 This is a synonym for .ltorg.
380 @cindex opcodes for ARM
381 @code{@value{AS}} implements all the standard ARM opcodes. It also
382 implements several pseudo opcodes, including several synthetic load
387 @cindex @code{NOP} pseudo op, ARM
393 This pseudo op will always evaluate to a legal ARM instruction that does
394 nothing. Currently it will evaluate to MOV r0, r0.
396 @cindex @code{LDR reg,=<label>} pseudo op, ARM
399 ldr <register> , = <expression>
402 If expression evaluates to a numeric constant then a MOV or MVN
403 instruction will be used in place of the LDR instruction, if the
404 constant can be generated by either of these instructions. Otherwise
405 the constant will be placed into the nearest literal pool (if it not
406 already there) and a PC relative LDR instruction will be generated.
408 @cindex @code{ADR reg,<label>} pseudo op, ARM
411 adr <register> <label>
414 This instruction will load the address of @var{label} into the indicated
415 register. The instruction will evaluate to a PC relative ADD or SUB
416 instruction depending upon where the label is located. If the label is
417 out of range, or if it is not defined in the same file (and section) as
418 the ADR instruction, then an error will be generated. This instruction
419 will not make use of the literal pool.
421 @cindex @code{ADRL reg,<label>} pseudo op, ARM
424 adrl <register> <label>
427 This instruction will load the address of @var{label} into the indicated
428 register. The instruction will evaluate to one or two PC relative ADD
429 or SUB instructions depending upon where the label is located. If a
430 second instruction is not needed a NOP instruction will be generated in
431 its place, so that this instruction is always 8 bytes long.
433 If the label is out of range, or if it is not defined in the same file
434 (and section) as the ADRL instruction, then an error will be generated.
435 This instruction will not make use of the literal pool.
439 For information on the ARM or Thumb instruction sets, see @cite{ARM
440 Software Development Toolkit Reference Manual}, Advanced RISC Machines
443 @node ARM Mapping Symbols
444 @section Mapping Symbols
446 The ARM ELF specification requires that special symbols be inserted
447 into object files to mark certain features:
453 At the start of a region of code containing ARM instructions.
457 At the start of a region of code containing THUMB instructions.
461 At the start of a region of data.
465 The assembler will automatically insert these symbols for you - there
466 is no need to code them yourself. Support for tagging symbols ($b,
467 $f, $p and $m) which is also mentioned in the current ARM ELF
468 specification is not implemented. This is because they have been
469 dropped from the new EABI and so tools cannot rely upon their