1 @c Copyright (C) 1988, 1989, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
2 @c 2000, 2001, 2002 Free Software Foundation, Inc.
3 @c This is part of the GCC manual.
4 @c For copying conditions, see the file gcc.texi.
8 Copyright @copyright{} 1988, 1989, 1992, 1993, 1994, 1995, 1996, 1997,
9 1998, 1999, 2000, 2001, 2002 Free Software Foundation, Inc.
11 Permission is granted to copy, distribute and/or modify this document
12 under the terms of the GNU Free Documentation License, Version 1.1 or
13 any later version published by the Free Software Foundation; with the
14 Invariant Sections being ``GNU General Public License'' and ``Funding
15 Free Software'', the Front-Cover texts being (a) (see below), and with
16 the Back-Cover Texts being (b) (see below). A copy of the license is
17 included in the gfdl(7) man page.
19 (a) The FSF's Front-Cover Text is:
23 (b) The FSF's Back-Cover Text is:
25 You have freedom to copy and modify this GNU Manual, like GNU
26 software. Copies published by the Free Software Foundation raise
27 funds for GNU development.
29 @c Set file name and title for the man page.
31 @settitle GNU project C and C++ compiler
33 gcc [@option{-c}|@option{-S}|@option{-E}] [@option{-std=}@var{standard}]
34 [@option{-g}] [@option{-pg}] [@option{-O}@var{level}]
35 [@option{-W}@var{warn}@dots{}] [@option{-pedantic}]
36 [@option{-I}@var{dir}@dots{}] [@option{-L}@var{dir}@dots{}]
37 [@option{-D}@var{macro}[=@var{defn}]@dots{}] [@option{-U}@var{macro}]
38 [@option{-f}@var{option}@dots{}] [@option{-m}@var{machine-option}@dots{}]
39 [@option{-o} @var{outfile}] @var{infile}@dots{}
41 Only the most useful options are listed here; see below for the
42 remainder. @samp{g++} accepts mostly the same options as @samp{gcc}.
45 gpl(7), gfdl(7), fsf-funding(7),
46 cpp(1), gcov(1), g77(1), as(1), ld(1), gdb(1), adb(1), dbx(1), sdb(1)
47 and the Info entries for @file{gcc}, @file{cpp}, @file{g77}, @file{as},
48 @file{ld}, @file{binutils} and @file{gdb}.
51 For instructions on reporting bugs, see
52 @w{@uref{http://gcc.gnu.org/bugs.html}}. Use of the @command{gccbug}
53 script to report bugs is recommended.
56 See the Info entry for @command{gcc}, or
57 @w{@uref{http://gcc.gnu.org/onlinedocs/gcc/Contributors.html}},
58 for contributors to GCC@.
63 @chapter GCC Command Options
64 @cindex GCC command options
65 @cindex command options
66 @cindex options, GCC command
68 @c man begin DESCRIPTION
70 When you invoke GCC, it normally does preprocessing, compilation,
71 assembly and linking. The ``overall options'' allow you to stop this
72 process at an intermediate stage. For example, the @option{-c} option
73 says not to run the linker. Then the output consists of object files
74 output by the assembler.
76 Other options are passed on to one stage of processing. Some options
77 control the preprocessor and others the compiler itself. Yet other
78 options control the assembler and linker; most of these are not
79 documented here, since you rarely need to use any of them.
81 @cindex C compilation options
82 Most of the command line options that you can use with GCC are useful
83 for C programs; when an option is only useful with another language
84 (usually C++), the explanation says so explicitly. If the description
85 for a particular option does not mention a source language, you can use
86 that option with all supported languages.
88 @cindex C++ compilation options
89 @xref{Invoking G++,,Compiling C++ Programs}, for a summary of special
90 options for compiling C++ programs.
92 @cindex grouping options
93 @cindex options, grouping
94 The @command{gcc} program accepts options and file names as operands. Many
95 options have multi-letter names; therefore multiple single-letter options
96 may @emph{not} be grouped: @option{-dr} is very different from @w{@samp{-d
99 @cindex order of options
100 @cindex options, order
101 You can mix options and other arguments. For the most part, the order
102 you use doesn't matter. Order does matter when you use several options
103 of the same kind; for example, if you specify @option{-L} more than once,
104 the directories are searched in the order specified.
106 Many options have long names starting with @samp{-f} or with
107 @samp{-W}---for example, @option{-fforce-mem},
108 @option{-fstrength-reduce}, @option{-Wformat} and so on. Most of
109 these have both positive and negative forms; the negative form of
110 @option{-ffoo} would be @option{-fno-foo}. This manual documents
111 only one of these two forms, whichever one is not the default.
115 @xref{Option Index}, for an index to GCC's options.
118 * Option Summary:: Brief list of all options, without explanations.
119 * Overall Options:: Controlling the kind of output:
120 an executable, object files, assembler files,
121 or preprocessed source.
122 * Invoking G++:: Compiling C++ programs.
123 * C Dialect Options:: Controlling the variant of C language compiled.
124 * C++ Dialect Options:: Variations on C++.
125 * Objective-C Dialect Options:: Variations on Objective-C.
126 * Language Independent Options:: Controlling how diagnostics should be
128 * Warning Options:: How picky should the compiler be?
129 * Debugging Options:: Symbol tables, measurements, and debugging dumps.
130 * Optimize Options:: How much optimization?
131 * Preprocessor Options:: Controlling header files and macro definitions.
132 Also, getting dependency information for Make.
133 * Assembler Options:: Passing options to the assembler.
134 * Link Options:: Specifying libraries and so on.
135 * Directory Options:: Where to find header files and libraries.
136 Where to find the compiler executable files.
137 * Spec Files:: How to pass switches to sub-processes.
138 * Target Options:: Running a cross-compiler, or an old version of GCC.
139 * Submodel Options:: Specifying minor hardware or convention variations,
140 such as 68010 vs 68020.
141 * Code Gen Options:: Specifying conventions for function calls, data layout
143 * Environment Variables:: Env vars that affect GCC.
144 * Running Protoize:: Automatically adding or removing function prototypes.
150 @section Option Summary
152 Here is a summary of all the options, grouped by type. Explanations are
153 in the following sections.
156 @item Overall Options
157 @xref{Overall Options,,Options Controlling the Kind of Output}.
159 -c -S -E -o @var{file} -pipe -pass-exit-codes -x @var{language} @gol
160 -v --target-help --help}
162 @item C Language Options
163 @xref{C Dialect Options,,Options Controlling C Dialect}.
165 -ansi -std=@var{standard} -aux-info @var{filename} @gol
166 -fno-asm -fno-builtin -fno-builtin-@var{function} @gol
167 -fhosted -ffreestanding @gol
168 -trigraphs -traditional -traditional-cpp @gol
169 -fallow-single-precision -fcond-mismatch @gol
170 -fsigned-bitfields -fsigned-char @gol
171 -funsigned-bitfields -funsigned-char @gol
172 -fwritable-strings -fshort-wchar}
174 @item C++ Language Options
175 @xref{C++ Dialect Options,,Options Controlling C++ Dialect}.
177 -fno-access-control -fcheck-new -fconserve-space @gol
178 -fno-const-strings -fdollars-in-identifiers @gol
179 -fno-elide-constructors @gol
180 -fno-enforce-eh-specs -fexternal-templates @gol
181 -falt-external-templates @gol
182 -ffor-scope -fno-for-scope -fno-gnu-keywords @gol
183 -fno-implicit-templates @gol
184 -fno-implicit-inline-templates @gol
185 -fno-implement-inlines -fms-extensions @gol
186 -fno-nonansi-builtins -fno-operator-names @gol
187 -fno-optional-diags -fpermissive @gol
188 -frepo -fno-rtti -fstats -ftemplate-depth-@var{n} @gol
189 -fuse-cxa-atexit -fvtable-gc -fno-weak -nostdinc++ @gol
190 -fno-default-inline -Wctor-dtor-privacy @gol
191 -Wnon-virtual-dtor -Wreorder @gol
192 -Weffc++ -Wno-deprecated @gol
193 -Wno-non-template-friend -Wold-style-cast @gol
194 -Woverloaded-virtual -Wno-pmf-conversions @gol
195 -Wsign-promo -Wsynth}
197 @item Objective-C Language Options
198 @xref{Objective-C Dialect Options,,Options Controlling Objective-C Dialect}.
200 -fconstant-string-class=@var{class-name} @gol
201 -fgnu-runtime -fnext-runtime -gen-decls @gol
202 -Wno-protocol -Wselector}
204 @item Language Independent Options
205 @xref{Language Independent Options,,Options to Control Diagnostic Messages Formatting}.
207 -fmessage-length=@var{n} @gol
208 -fdiagnostics-show-location=@r{[}once@r{|}every-line@r{]}}
210 @item Warning Options
211 @xref{Warning Options,,Options to Request or Suppress Warnings}.
213 -fsyntax-only -pedantic -pedantic-errors @gol
214 -w -W -Wall -Waggregate-return @gol
215 -Wcast-align -Wcast-qual -Wchar-subscripts -Wcomment @gol
216 -Wconversion -Wno-deprecated-declarations @gol
217 -Wdisabled-optimization -Wdiv-by-zero -Werror @gol
218 -Wfloat-equal -Wformat -Wformat=2 @gol
219 -Wformat-nonliteral -Wformat-security @gol
220 -Wimplicit -Wimplicit-int @gol
221 -Wimplicit-function-declaration @gol
222 -Werror-implicit-function-declaration @gol
223 -Wimport -Winline @gol
224 -Wlarger-than-@var{len} -Wlong-long @gol
225 -Wmain -Wmissing-braces -Wmissing-declarations @gol
226 -Wmissing-format-attribute -Wmissing-noreturn @gol
227 -Wmultichar -Wno-format-extra-args -Wno-format-y2k @gol
228 -Wno-import -Wpacked -Wpadded @gol
229 -Wparentheses -Wpointer-arith -Wredundant-decls @gol
230 -Wreturn-type -Wsequence-point -Wshadow @gol
231 -Wsign-compare -Wswitch -Wsystem-headers @gol
232 -Wtrigraphs -Wundef -Wuninitialized @gol
233 -Wunknown-pragmas -Wunreachable-code @gol
234 -Wunused -Wunused-function -Wunused-label -Wunused-parameter @gol
235 -Wunused-value -Wunused-variable -Wwrite-strings}
237 @item C-only Warning Options
239 -Wbad-function-cast -Wmissing-prototypes -Wnested-externs @gol
240 -Wstrict-prototypes -Wtraditional}
242 @item Debugging Options
243 @xref{Debugging Options,,Options for Debugging Your Program or GCC}.
245 -d@var{letters} -dumpspecs -dumpmachine -dumpversion @gol
246 -fdump-unnumbered -fdump-translation-unit@r{[}-@var{n}@r{]} @gol
247 -fdump-class-hierarchy@r{[}-@var{n}@r{]} @gol
248 -fdump-tree-original@r{[}-@var{n}@r{]} -fdump-tree-optimized@r{[}-@var{n}@r{]} @gol
249 -fdump-tree-inlined@r{[}-@var{n}@r{]} @gol
250 -fmem-report -fpretend-float @gol
251 -fprofile-arcs -ftest-coverage -ftime-report @gol
252 -g -g@var{level} -gcoff -gdwarf -gdwarf-1 -gdwarf-1+ -gdwarf-2 @gol
253 -ggdb -gstabs -gstabs+ -gvms -gxcoff -gxcoff+ @gol
254 -p -pg -print-file-name=@var{library} -print-libgcc-file-name @gol
255 -print-multi-directory -print-multi-lib @gol
256 -print-prog-name=@var{program} -print-search-dirs -Q @gol
259 @item Optimization Options
260 @xref{Optimize Options,,Options that Control Optimization}.
262 -falign-functions=@var{n} -falign-jumps=@var{n} @gol
263 -falign-labels=@var{n} -falign-loops=@var{n} @gol
264 -fbranch-probabilities -fcaller-saves -fcprop-registers @gol
265 -fcse-follow-jumps -fcse-skip-blocks -fdata-sections @gol
266 -fdelayed-branch -fdelete-null-pointer-checks @gol
267 -fexpensive-optimizations -ffast-math -ffloat-store @gol
268 -fforce-addr -fforce-mem -ffunction-sections @gol
269 -fgcse -fgcse-lm -fgcse-sm @gol
270 -finline-functions -finline-limit=@var{n} -fkeep-inline-functions @gol
271 -fkeep-static-consts -fmerge-constants -fmerge-all-constants @gol
272 -fmove-all-movables -fno-default-inline -fno-defer-pop @gol
273 -fno-function-cse -fno-guess-branch-probability @gol
274 -fno-inline -fno-math-errno -fno-peephole -fno-peephole2 @gol
275 -funsafe-math-optimizations -fno-trapping-math @gol
276 -fomit-frame-pointer -foptimize-register-move @gol
277 -foptimize-sibling-calls -fprefetch-loop-arrays @gol
278 -freduce-all-givs -fregmove -frename-registers @gol
279 -frerun-cse-after-loop -frerun-loop-opt @gol
280 -fschedule-insns -fschedule-insns2 @gol
281 -fsingle-precision-constant -fssa -fssa-ccp -fssa-dce @gol
282 -fstrength-reduce -fstrict-aliasing -fthread-jumps -ftrapv @gol
283 -funroll-all-loops -funroll-loops @gol
284 --param @var{name}=@var{value}
285 -O -O0 -O1 -O2 -O3 -Os}
287 @item Preprocessor Options
288 @xref{Preprocessor Options,,Options Controlling the Preprocessor}.
290 -$ -A@var{question}=@var{answer} -A-@var{question}@r{[}=@var{answer}@r{]} @gol
291 -C -dD -dI -dM -dN @gol
292 -D@var{macro}@r{[}=@var{defn}@r{]} -E -H @gol
293 -idirafter @var{dir} @gol
294 -include @var{file} -imacros @var{file} @gol
295 -iprefix @var{file} -iwithprefix @var{dir} @gol
296 -iwithprefixbefore @var{dir} -isystem @var{dir} @gol
297 -M -MM -MF -MG -MP -MQ -MT -nostdinc -P -remap @gol
298 -trigraphs -undef -U@var{macro} -Wp,@var{option}}
300 @item Assembler Option
301 @xref{Assembler Options,,Passing Options to the Assembler}.
306 @xref{Link Options,,Options for Linking}.
308 @var{object-file-name} -l@var{library} @gol
309 -nostartfiles -nodefaultlibs -nostdlib @gol
310 -s -static -static-libgcc -shared -shared-libgcc -symbolic @gol
311 -Wl,@var{option} -Xlinker @var{option} @gol
314 @item Directory Options
315 @xref{Directory Options,,Options for Directory Search}.
317 -B@var{prefix} -I@var{dir} -I- -L@var{dir} -specs=@var{file}}
320 @c I wrote this xref this way to avoid overfull hbox. -- rms
321 @xref{Target Options}.
323 -b @var{machine} -V @var{version}}
325 @item Machine Dependent Options
326 @xref{Submodel Options,,Hardware Models and Configurations}.
328 @emph{M680x0 Options}
330 -m68000 -m68020 -m68020-40 -m68020-60 -m68030 -m68040 @gol
331 -m68060 -mcpu32 -m5200 -m68881 -mbitfield -mc68000 -mc68020 @gol
332 -mfpa -mnobitfield -mrtd -mshort -msoft-float -mpcrel @gol
333 -malign-int -mstrict-align}
335 @emph{M68hc1x Options}
337 -m6811 -m6812 -m68hc11 -m68hc12 @gol
338 -mauto-incdec -mshort -msoft-reg-count=@var{count}}
346 -mcpu=@var{cpu-type} @gol
347 -mtune=@var{cpu-type} @gol
348 -mcmodel=@var{code-model} @gol
350 -mapp-regs -mbroken-saverestore -mcypress @gol
351 -mepilogue -mfaster-structs -mflat @gol
352 -mfpu -mhard-float -mhard-quad-float @gol
353 -mimpure-text -mlive-g0 -mno-app-regs @gol
354 -mno-epilogue -mno-faster-structs -mno-flat -mno-fpu @gol
355 -mno-impure-text -mno-stack-bias -mno-unaligned-doubles @gol
356 -msoft-float -msoft-quad-float -msparclite -mstack-bias @gol
357 -msupersparc -munaligned-doubles -mv8}
359 @emph{Convex Options}
361 -mc1 -mc2 -mc32 -mc34 -mc38 @gol
362 -margcount -mnoargcount @gol
363 -mlong32 -mlong64 @gol
364 -mvolatile-cache -mvolatile-nocache}
366 @emph{AMD29K Options}
368 -m29000 -m29050 -mbw -mnbw -mdw -mndw @gol
369 -mlarge -mnormal -msmall @gol
370 -mkernel-registers -mno-reuse-arg-regs @gol
371 -mno-stack-check -mno-storem-bug @gol
372 -mreuse-arg-regs -msoft-float -mstack-check @gol
373 -mstorem-bug -muser-registers}
377 -mapcs-frame -mno-apcs-frame @gol
378 -mapcs-26 -mapcs-32 @gol
379 -mapcs-stack-check -mno-apcs-stack-check @gol
380 -mapcs-float -mno-apcs-float @gol
381 -mapcs-reentrant -mno-apcs-reentrant @gol
382 -msched-prolog -mno-sched-prolog @gol
383 -mlittle-endian -mbig-endian -mwords-little-endian @gol
384 -malignment-traps -mno-alignment-traps @gol
385 -msoft-float -mhard-float -mfpe @gol
386 -mthumb-interwork -mno-thumb-interwork @gol
387 -mcpu=@var{name} -march=@var{name} -mfpe=@var{name} @gol
388 -mstructure-size-boundary=@var{n} @gol
389 -mbsd -mxopen -mno-symrename @gol
390 -mabort-on-noreturn @gol
391 -mlong-calls -mno-long-calls @gol
392 -msingle-pic-base -mno-single-pic-base @gol
393 -mpic-register=@var{reg} @gol
394 -mnop-fun-dllimport @gol
395 -mpoke-function-name @gol
397 -mtpcs-frame -mtpcs-leaf-frame @gol
398 -mcaller-super-interworking -mcallee-super-interworking }
400 @emph{MN10200 Options}
404 @emph{MN10300 Options}
406 -mmult-bug -mno-mult-bug @gol
407 -mam33 -mno-am33 @gol
410 @emph{M32R/D Options}
412 -m32rx -m32r -mcode-model=@var{model-type} -msdata=@var{sdata-type} @gol
417 -m88000 -m88100 -m88110 -mbig-pic @gol
418 -mcheck-zero-division -mhandle-large-shift @gol
419 -midentify-revision -mno-check-zero-division @gol
420 -mno-ocs-debug-info -mno-ocs-frame-position @gol
421 -mno-optimize-arg-area -mno-serialize-volatile @gol
422 -mno-underscores -mocs-debug-info @gol
423 -mocs-frame-position -moptimize-arg-area @gol
424 -mserialize-volatile -mshort-data-@var{num} -msvr3 @gol
425 -msvr4 -mtrap-large-shift -muse-div-instruction @gol
426 -mversion-03.00 -mwarn-passed-structs}
428 @emph{RS/6000 and PowerPC Options}
430 -mcpu=@var{cpu-type} @gol
431 -mtune=@var{cpu-type} @gol
432 -mpower -mno-power -mpower2 -mno-power2 @gol
433 -mpowerpc -mpowerpc64 -mno-powerpc @gol
434 -maltivec -mno-altivec @gol
435 -mpowerpc-gpopt -mno-powerpc-gpopt @gol
436 -mpowerpc-gfxopt -mno-powerpc-gfxopt @gol
437 -mnew-mnemonics -mold-mnemonics @gol
438 -mfull-toc -mminimal-toc -mno-fp-in-toc -mno-sum-in-toc @gol
439 -m64 -m32 -mxl-call -mno-xl-call -mpe @gol
440 -msoft-float -mhard-float -mmultiple -mno-multiple @gol
441 -mstring -mno-string -mupdate -mno-update @gol
442 -mfused-madd -mno-fused-madd -mbit-align -mno-bit-align @gol
443 -mstrict-align -mno-strict-align -mrelocatable @gol
444 -mno-relocatable -mrelocatable-lib -mno-relocatable-lib @gol
445 -mtoc -mno-toc -mlittle -mlittle-endian -mbig -mbig-endian @gol
446 -mcall-aix -mcall-sysv -mcall-netbsd @gol
447 -maix-struct-return -msvr4-struct-return
449 -mprototype -mno-prototype @gol
450 -msim -mmvme -mads -myellowknife -memb -msdata @gol
451 -msdata=@var{opt} -mvxworks -G @var{num} -pthread}
455 -mcall-lib-mul -mfp-arg-in-fpregs -mfp-arg-in-gregs @gol
456 -mfull-fp-blocks -mhc-struct-return -min-line-mul @gol
457 -mminimum-fp-blocks -mnohc-struct-return}
461 -mabicalls -march=@var{cpu-type} -mtune=@var{cpu=type} @gol
462 -mcpu=@var{cpu-type} -membedded-data -muninit-const-in-rodata @gol
463 -membedded-pic -mfp32 -mfp64 -mfused-madd -mno-fused-madd @gol
464 -mgas -mgp32 -mgp64 @gol
465 -mgpopt -mhalf-pic -mhard-float -mint64 -mips1 @gol
466 -mips2 -mips3 -mips4 -mlong64 -mlong32 -mlong-calls -mmemcpy @gol
467 -mmips-as -mmips-tfile -mno-abicalls @gol
468 -mno-embedded-data -mno-uninit-const-in-rodata @gol
469 -mno-embedded-pic -mno-gpopt -mno-long-calls @gol
470 -mno-memcpy -mno-mips-tfile -mno-rnames -mno-stats @gol
471 -mrnames -msoft-float @gol
472 -m4650 -msingle-float -mmad @gol
473 -mstats -EL -EB -G @var{num} -nocpp @gol
474 -mabi=32 -mabi=n32 -mabi=64 -mabi=eabi @gol
475 -mfix7000 -mno-crt0 -mflush-func=@var{func} -mno-flush-func}
477 @emph{i386 and x86-64 Options}
479 -mcpu=@var{cpu-type} -march=@var{cpu-type} -mfpmath=@var{unit} @gol
480 -masm=@var{dialect} -mno-fancy-math-387 @gol
481 -mno-fp-ret-in-387 -msoft-float -msvr3-shlib @gol
482 -mno-wide-multiply -mrtd -malign-double @gol
483 -mpreferred-stack-boundary=@var{num} @gol
484 -mmmx -msse -msse2 -msse-math -m3dnow @gol
485 -mthreads -mno-align-stringops -minline-all-stringops @gol
486 -mpush-args -maccumulate-outgoing-args -m128bit-long-double @gol
487 -m96bit-long-double -mregparm=@var{num} -momit-leaf-frame-pointer @gol
493 -march=@var{architecture-type} @gol
494 -mbig-switch -mdisable-fpregs -mdisable-indexing @gol
495 -mfast-indirect-calls -mgas -mjump-in-delay @gol
496 -mlong-load-store -mno-big-switch -mno-disable-fpregs @gol
497 -mno-disable-indexing -mno-fast-indirect-calls -mno-gas @gol
498 -mno-jump-in-delay -mno-long-load-store @gol
499 -mno-portable-runtime -mno-soft-float @gol
500 -mno-space-regs -msoft-float -mpa-risc-1-0 @gol
501 -mpa-risc-1-1 -mpa-risc-2-0 -mportable-runtime @gol
502 -mschedule=@var{cpu-type} -mspace-regs}
504 @emph{Intel 960 Options}
506 -m@var{cpu-type} -masm-compat -mclean-linkage @gol
507 -mcode-align -mcomplex-addr -mleaf-procedures @gol
508 -mic-compat -mic2.0-compat -mic3.0-compat @gol
509 -mintel-asm -mno-clean-linkage -mno-code-align @gol
510 -mno-complex-addr -mno-leaf-procedures @gol
511 -mno-old-align -mno-strict-align -mno-tail-call @gol
512 -mnumerics -mold-align -msoft-float -mstrict-align @gol
515 @emph{DEC Alpha Options}
517 -mno-fp-regs -msoft-float -malpha-as -mgas @gol
518 -mieee -mieee-with-inexact -mieee-conformant @gol
519 -mfp-trap-mode=@var{mode} -mfp-rounding-mode=@var{mode} @gol
520 -mtrap-precision=@var{mode} -mbuild-constants @gol
521 -mcpu=@var{cpu-type} -mtune=@var{cpu-type} @gol
522 -mbwx -mmax -mfix -mcix @gol
523 -mfloat-vax -mfloat-ieee @gol
524 -mexplicit-relocs -msmall-data -mlarge-data @gol
525 -mmemory-latency=@var{time}}
527 @emph{DEC Alpha/VMS Options}
531 @emph{Clipper Options}
535 @emph{H8/300 Options}
537 -mrelax -mh -ms -mint32 -malign-300}
541 -m1 -m2 -m3 -m3e @gol
542 -m4-nofpu -m4-single-only -m4-single -m4 @gol
543 -m5-64media -m5-64media-nofpu @gol
544 -m5-32media -m5-32media-nofpu @gol
545 -m5-compact -m5-compact-nofpu @gol
546 -mb -ml -mdalign -mrelax @gol
547 -mbigtable -mfmovd -mhitachi -mnomacsave @gol
548 -mieee -misize -mpadstruct -mspace @gol
549 -mprefergot -musermode}
551 @emph{System V Options}
553 -Qy -Qn -YP,@var{paths} -Ym,@var{dir}}
558 -mmangle-cpu -mcpu=@var{cpu} -mtext=@var{text-section} @gol
559 -mdata=@var{data-section} -mrodata=@var{readonly-data-section}}
561 @emph{TMS320C3x/C4x Options}
563 -mcpu=@var{cpu} -mbig -msmall -mregparm -mmemparm @gol
564 -mfast-fix -mmpyi -mbk -mti -mdp-isr-reload @gol
565 -mrpts=@var{count} -mrptb -mdb -mloop-unsigned @gol
566 -mparallel-insns -mparallel-mpy -mpreserve-float}
570 -mlong-calls -mno-long-calls -mep -mno-ep @gol
571 -mprolog-function -mno-prolog-function -mspace @gol
572 -mtda=@var{n} -msda=@var{n} -mzda=@var{n} @gol
577 -m32032 -m32332 -m32532 -m32081 -m32381 @gol
578 -mmult-add -mnomult-add -msoft-float -mrtd -mnortd @gol
579 -mregparam -mnoregparam -msb -mnosb @gol
580 -mbitfield -mnobitfield -mhimem -mnohimem}
584 -mmcu=@var{mcu} -msize -minit-stack=@var{n} -mno-interrupts @gol
585 -mcall-prologues -mno-tablejump -mtiny-stack}
589 -mhardlit -mno-hardlit -mdiv -mno-div -mrelax-immediates @gol
590 -mno-relax-immediates -mwide-bitfields -mno-wide-bitfields @gol
591 -m4byte-functions -mno-4byte-functions -mcallgraph-data @gol
592 -mno-callgraph-data -mslow-bytes -mno-slow-bytes -mno-lsim @gol
593 -mlittle-endian -mbig-endian -m210 -m340 -mstack-increment}
597 -mlibfuncs -mno-libfuncs -mepsilon -mno-epsilon -mabi=gnu @gol
598 -mabi=mmixware -mzero-extend -mknuthdiv -mtoplevel-symbols @gol
599 -melf -mbranch-predict -mno-branch-predict -mbase-addresses @gol
604 -mbig-endian -mlittle-endian -mgnu-as -mgnu-ld -mno-pic @gol
605 -mvolatile-asm-stop -mb-step -mregister-names -mno-sdata @gol
606 -mconstant-gp -mauto-pic -minline-divide-min-latency @gol
607 -minline-divide-max-throughput -mno-dwarf2-asm @gol
608 -mfixed-range=@var{register-range}}
612 -mextmem -mextmemory -monchip -mno-asm-optimize -masm-optimize @gol
613 -mbranch-cost=@var{n} -mcond-exec=@var{n}}
615 @emph{S/390 and zSeries Options}
617 -mhard-float -msoft-float -mbackchain -mno-backchain @gol
618 -msmall-exec -mno-small-exec -mmvcle -mno-mvcle @gol
619 -m64 -m31 -mdebug -mno-debug}
623 -mcpu=@var{cpu} -march=@var{cpu} -mtune=@var{cpu} @gol
624 -mmax-stack-frame=@var{n} -melinux-stacksize=@var{n} @gol
625 -metrax4 -metrax100 -mpdebug -mcc-init -mno-side-effects @gol
626 -mstack-align -mdata-align -mconst-align @gol
627 -m32-bit -m16-bit -m8-bit -mno-prologue-epilogue -mno-gotplt @gol
628 -melf -maout -melinux -mlinux -sim -sim2}
630 @emph{PDP-11 Options}
632 -mfpu -msoft-float -mac0 -mno-ac0 -m40 -m45 -m10 @gol
633 -mbcopy -mbcopy-builtin -mint32 -mno-int16 @gol
634 -mint16 -mno-int32 -mfloat32 -mno-float64 @gol
635 -mfloat64 -mno-float32 -mabshi -mno-abshi @gol
636 -mbranch-expensive -mbranch-cheap @gol
637 -msplit -mno-split -munix-asm -mdec-asm}
639 @emph{Xstormy16 Options}
643 @emph{Xtensa Options}
645 -mbig-endian -mlittle-endian @gol
646 -mdensity -mno-density @gol
647 -mmac16 -mno-mac16 @gol
648 -mmul16 -mno-mul16 @gol
649 -mmul32 -mno-mul32 @gol
651 -mminmax -mno-minmax @gol
652 -msext -mno-sext @gol
653 -mbooleans -mno-booleans @gol
654 -mhard-float -msoft-float @gol
655 -mfused-madd -mno-fused-madd @gol
656 -mserialize-volatile -mno-serialize-volatile @gol
657 -mtext-section-literals -mno-text-section-literals @gol
658 -mtarget-align -mno-target-align @gol
659 -mlongcalls -mno-longcalls}
661 @item Code Generation Options
662 @xref{Code Gen Options,,Options for Code Generation Conventions}.
664 -fcall-saved-@var{reg} -fcall-used-@var{reg} @gol
665 -ffixed-@var{reg} -fexceptions @gol
666 -fnon-call-exceptions -funwind-tables @gol
667 -fasynchronous-unwind-tables @gol
668 -finhibit-size-directive -finstrument-functions @gol
669 -fno-common -fno-ident -fno-gnu-linker @gol
670 -fpcc-struct-return -fpic -fPIC @gol
671 -freg-struct-return -fshared-data -fshort-enums @gol
672 -fshort-double -fvolatile @gol
673 -fvolatile-global -fvolatile-static @gol
674 -fverbose-asm -fpack-struct -fstack-check @gol
675 -fstack-limit-register=@var{reg} -fstack-limit-symbol=@var{sym} @gol
676 -fargument-alias -fargument-noalias @gol
677 -fargument-noalias-global -fleading-underscore}
681 * Overall Options:: Controlling the kind of output:
682 an executable, object files, assembler files,
683 or preprocessed source.
684 * C Dialect Options:: Controlling the variant of C language compiled.
685 * C++ Dialect Options:: Variations on C++.
686 * Objective-C Dialect Options:: Variations on Objective-C.
687 * Language Independent Options:: Controlling how diagnostics should be
689 * Warning Options:: How picky should the compiler be?
690 * Debugging Options:: Symbol tables, measurements, and debugging dumps.
691 * Optimize Options:: How much optimization?
692 * Preprocessor Options:: Controlling header files and macro definitions.
693 Also, getting dependency information for Make.
694 * Assembler Options:: Passing options to the assembler.
695 * Link Options:: Specifying libraries and so on.
696 * Directory Options:: Where to find header files and libraries.
697 Where to find the compiler executable files.
698 * Spec Files:: How to pass switches to sub-processes.
699 * Target Options:: Running a cross-compiler, or an old version of GCC.
702 @node Overall Options
703 @section Options Controlling the Kind of Output
705 Compilation can involve up to four stages: preprocessing, compilation
706 proper, assembly and linking, always in that order. The first three
707 stages apply to an individual source file, and end by producing an
708 object file; linking combines all the object files (those newly
709 compiled, and those specified as input) into an executable file.
711 @cindex file name suffix
712 For any given input file, the file name suffix determines what kind of
717 C source code which must be preprocessed.
720 C source code which should not be preprocessed.
723 C++ source code which should not be preprocessed.
726 Objective-C source code. Note that you must link with the library
727 @file{libobjc.a} to make an Objective-C program work.
730 Objective-C source code which should not be preprocessed.
733 C header file (not to be compiled or linked).
737 @itemx @var{file}.cxx
738 @itemx @var{file}.cpp
739 @itemx @var{file}.c++
741 C++ source code which must be preprocessed. Note that in @samp{.cxx},
742 the last two letters must both be literally @samp{x}. Likewise,
743 @samp{.C} refers to a literal capital C@.
746 @itemx @var{file}.for
747 @itemx @var{file}.FOR
748 Fortran source code which should not be preprocessed.
751 @itemx @var{file}.fpp
752 @itemx @var{file}.FPP
753 Fortran source code which must be preprocessed (with the traditional
757 Fortran source code which must be preprocessed with a RATFOR
758 preprocessor (not included with GCC)@.
760 @xref{Overall Options,,Options Controlling the Kind of Output, g77,
761 Using and Porting GNU Fortran}, for more details of the handling of
764 @c FIXME: Descriptions of Java file types.
771 Ada source code file which contains a library unit declaration (a
772 declaration of a package, subprogram, or generic, or a generic
773 instantiation), or a library unit renaming declaration (a package,
774 generic, or subprogram renaming declaration). Such files are also
777 @itemx @var{file}.adb
778 Ada source code file containing a library unit body (a subprogram or
779 package body). Such files are also called @dfn{bodies}.
781 @c GCC also knows about some suffixes for languages not yet included:
787 @itemx @var{file}.chi
788 CHILL source code (preprocessed with the traditional preprocessor).
794 Assembler code which must be preprocessed.
797 An object file to be fed straight into linking.
798 Any file name with no recognized suffix is treated this way.
802 You can specify the input language explicitly with the @option{-x} option:
805 @item -x @var{language}
806 Specify explicitly the @var{language} for the following input files
807 (rather than letting the compiler choose a default based on the file
808 name suffix). This option applies to all following input files until
809 the next @option{-x} option. Possible values for @var{language} are:
811 c c-header cpp-output
813 objective-c objc-cpp-output
814 assembler assembler-with-cpp
817 f77 f77-cpp-input ratfor
822 Turn off any specification of a language, so that subsequent files are
823 handled according to their file name suffixes (as they are if @option{-x}
824 has not been used at all).
826 @item -pass-exit-codes
827 @opindex pass-exit-codes
828 Normally the @command{gcc} program will exit with the code of 1 if any
829 phase of the compiler returns a non-success return code. If you specify
830 @option{-pass-exit-codes}, the @command{gcc} program will instead return with
831 numerically highest error produced by any phase that returned an error
835 If you only want some of the stages of compilation, you can use
836 @option{-x} (or filename suffixes) to tell @command{gcc} where to start, and
837 one of the options @option{-c}, @option{-S}, or @option{-E} to say where
838 @command{gcc} is to stop. Note that some combinations (for example,
839 @samp{-x cpp-output -E}) instruct @command{gcc} to do nothing at all.
844 Compile or assemble the source files, but do not link. The linking
845 stage simply is not done. The ultimate output is in the form of an
846 object file for each source file.
848 By default, the object file name for a source file is made by replacing
849 the suffix @samp{.c}, @samp{.i}, @samp{.s}, etc., with @samp{.o}.
851 Unrecognized input files, not requiring compilation or assembly, are
856 Stop after the stage of compilation proper; do not assemble. The output
857 is in the form of an assembler code file for each non-assembler input
860 By default, the assembler file name for a source file is made by
861 replacing the suffix @samp{.c}, @samp{.i}, etc., with @samp{.s}.
863 Input files that don't require compilation are ignored.
867 Stop after the preprocessing stage; do not run the compiler proper. The
868 output is in the form of preprocessed source code, which is sent to the
871 Input files which don't require preprocessing are ignored.
873 @cindex output file option
876 Place output in file @var{file}. This applies regardless to whatever
877 sort of output is being produced, whether it be an executable file,
878 an object file, an assembler file or preprocessed C code.
880 Since only one output file can be specified, it does not make sense to
881 use @option{-o} when compiling more than one input file, unless you are
882 producing an executable file as output.
884 If @option{-o} is not specified, the default is to put an executable file
885 in @file{a.out}, the object file for @file{@var{source}.@var{suffix}} in
886 @file{@var{source}.o}, its assembler file in @file{@var{source}.s}, and
887 all preprocessed C source on standard output.
891 Print (on standard error output) the commands executed to run the stages
892 of compilation. Also print the version number of the compiler driver
893 program and of the preprocessor and the compiler proper.
897 Use pipes rather than temporary files for communication between the
898 various stages of compilation. This fails to work on some systems where
899 the assembler is unable to read from a pipe; but the GNU assembler has
904 Print (on the standard output) a description of the command line options
905 understood by @command{gcc}. If the @option{-v} option is also specified
906 then @option{--help} will also be passed on to the various processes
907 invoked by @command{gcc}, so that they can display the command line options
908 they accept. If the @option{-W} option is also specified then command
909 line options which have no documentation associated with them will also
914 Print (on the standard output) a description of target specific command
915 line options for each tool.
919 @section Compiling C++ Programs
921 @cindex suffixes for C++ source
922 @cindex C++ source file suffixes
923 C++ source files conventionally use one of the suffixes @samp{.C},
924 @samp{.cc}, @samp{.cpp}, @samp{.c++}, @samp{.cp}, or @samp{.cxx};
925 preprocessed C++ files use the suffix @samp{.ii}. GCC recognizes
926 files with these names and compiles them as C++ programs even if you
927 call the compiler the same way as for compiling C programs (usually with
928 the name @command{gcc}).
932 However, C++ programs often require class libraries as well as a
933 compiler that understands the C++ language---and under some
934 circumstances, you might want to compile programs from standard input,
935 or otherwise without a suffix that flags them as C++ programs.
936 @command{g++} is a program that calls GCC with the default language
937 set to C++, and automatically specifies linking against the C++
938 library. On many systems, @command{g++} is also
939 installed with the name @command{c++}.
941 @cindex invoking @command{g++}
942 When you compile C++ programs, you may specify many of the same
943 command-line options that you use for compiling programs in any
944 language; or command-line options meaningful for C and related
945 languages; or options that are meaningful only for C++ programs.
946 @xref{C Dialect Options,,Options Controlling C Dialect}, for
947 explanations of options for languages related to C@.
948 @xref{C++ Dialect Options,,Options Controlling C++ Dialect}, for
949 explanations of options that are meaningful only for C++ programs.
951 @node C Dialect Options
952 @section Options Controlling C Dialect
953 @cindex dialect options
954 @cindex language dialect options
955 @cindex options, dialect
957 The following options control the dialect of C (or languages derived
958 from C, such as C++ and Objective-C) that the compiler accepts:
965 In C mode, support all ISO C89 programs. In C++ mode,
966 remove GNU extensions that conflict with ISO C++.
968 This turns off certain features of GCC that are incompatible with ISO
969 C89 (when compiling C code), or of standard C++ (when compiling C++ code),
970 such as the @code{asm} and @code{typeof} keywords, and
971 predefined macros such as @code{unix} and @code{vax} that identify the
972 type of system you are using. It also enables the undesirable and
973 rarely used ISO trigraph feature. For the C compiler,
974 it disables recognition of C++ style @samp{//} comments as well as
975 the @code{inline} keyword.
977 The alternate keywords @code{__asm__}, @code{__extension__},
978 @code{__inline__} and @code{__typeof__} continue to work despite
979 @option{-ansi}. You would not want to use them in an ISO C program, of
980 course, but it is useful to put them in header files that might be included
981 in compilations done with @option{-ansi}. Alternate predefined macros
982 such as @code{__unix__} and @code{__vax__} are also available, with or
983 without @option{-ansi}.
985 The @option{-ansi} option does not cause non-ISO programs to be
986 rejected gratuitously. For that, @option{-pedantic} is required in
987 addition to @option{-ansi}. @xref{Warning Options}.
989 The macro @code{__STRICT_ANSI__} is predefined when the @option{-ansi}
990 option is used. Some header files may notice this macro and refrain
991 from declaring certain functions or defining certain macros that the
992 ISO standard doesn't call for; this is to avoid interfering with any
993 programs that might use these names for other things.
995 Functions which would normally be built in but do not have semantics
996 defined by ISO C (such as @code{alloca} and @code{ffs}) are not built-in
997 functions with @option{-ansi} is used. @xref{Other Builtins,,Other
998 built-in functions provided by GCC}, for details of the functions
1003 Determine the language standard. This option is currently only
1004 supported when compiling C@. A value for this option must be provided;
1010 ISO C89 (same as @option{-ansi}).
1012 @item iso9899:199409
1013 ISO C89 as modified in amendment 1.
1019 ISO C99. Note that this standard is not yet fully supported; see
1020 @w{@uref{http://gcc.gnu.org/c99status.html}} for more information. The
1021 names @samp{c9x} and @samp{iso9899:199x} are deprecated.
1024 Default, ISO C89 plus GNU extensions (including some C99 features).
1028 ISO C99 plus GNU extensions. When ISO C99 is fully implemented in GCC,
1029 this will become the default. The name @samp{gnu9x} is deprecated.
1033 Even when this option is not specified, you can still use some of the
1034 features of newer standards in so far as they do not conflict with
1035 previous C standards. For example, you may use @code{__restrict__} even
1036 when @option{-std=c99} is not specified.
1038 The @option{-std} options specifying some version of ISO C have the same
1039 effects as @option{-ansi}, except that features that were not in ISO C89
1040 but are in the specified version (for example, @samp{//} comments and
1041 the @code{inline} keyword in ISO C99) are not disabled.
1043 @xref{Standards,,Language Standards Supported by GCC}, for details of
1044 these standard versions.
1046 @item -aux-info @var{filename}
1048 Output to the given filename prototyped declarations for all functions
1049 declared and/or defined in a translation unit, including those in header
1050 files. This option is silently ignored in any language other than C@.
1052 Besides declarations, the file indicates, in comments, the origin of
1053 each declaration (source file and line), whether the declaration was
1054 implicit, prototyped or unprototyped (@samp{I}, @samp{N} for new or
1055 @samp{O} for old, respectively, in the first character after the line
1056 number and the colon), and whether it came from a declaration or a
1057 definition (@samp{C} or @samp{F}, respectively, in the following
1058 character). In the case of function definitions, a K&R-style list of
1059 arguments followed by their declarations is also provided, inside
1060 comments, after the declaration.
1064 Do not recognize @code{asm}, @code{inline} or @code{typeof} as a
1065 keyword, so that code can use these words as identifiers. You can use
1066 the keywords @code{__asm__}, @code{__inline__} and @code{__typeof__}
1067 instead. @option{-ansi} implies @option{-fno-asm}.
1069 In C++, this switch only affects the @code{typeof} keyword, since
1070 @code{asm} and @code{inline} are standard keywords. You may want to
1071 use the @option{-fno-gnu-keywords} flag instead, which has the same
1072 effect. In C99 mode (@option{-std=c99} or @option{-std=gnu99}), this
1073 switch only affects the @code{asm} and @code{typeof} keywords, since
1074 @code{inline} is a standard keyword in ISO C99.
1077 @itemx -fno-builtin-@var{function} @r{(C and Objective-C only)}
1078 @opindex fno-builtin
1079 @cindex built-in functions
1080 Don't recognize built-in functions that do not begin with
1081 @samp{__builtin_} as prefix. @xref{Other Builtins,,Other built-in
1082 functions provided by GCC}, for details of the functions affected,
1083 including those which are not built-in functions when @option{-ansi} or
1084 @option{-std} options for strict ISO C conformance are used because they
1085 do not have an ISO standard meaning.
1087 GCC normally generates special code to handle certain built-in functions
1088 more efficiently; for instance, calls to @code{alloca} may become single
1089 instructions that adjust the stack directly, and calls to @code{memcpy}
1090 may become inline copy loops. The resulting code is often both smaller
1091 and faster, but since the function calls no longer appear as such, you
1092 cannot set a breakpoint on those calls, nor can you change the behavior
1093 of the functions by linking with a different library.
1095 In C++, @option{-fno-builtin} is always in effect. The @option{-fbuiltin}
1096 option has no effect. Therefore, in C++, the only way to get the
1097 optimization benefits of built-in functions is to call the function
1098 using the @samp{__builtin_} prefix. The GNU C++ Standard Library uses
1099 built-in functions to implement many functions (like
1100 @code{std::strchr}), so that you automatically get efficient code.
1102 With the @option{-fno-builtin-@var{function}} option, not available
1103 when compiling C++, only the built-in function @var{function} is
1104 disabled. @var{function} must not begin with @samp{__builtin_}. If a
1105 function is named this is not built-in in this version of GCC, this
1106 option is ignored. There is no corresponding
1107 @option{-fbuiltin-@var{function}} option; if you wish to enable
1108 built-in functions selectively when using @option{-fno-builtin} or
1109 @option{-ffreestanding}, you may define macros such as:
1112 #define abs(n) __builtin_abs ((n))
1113 #define strcpy(d, s) __builtin_strcpy ((d), (s))
1118 @cindex hosted environment
1120 Assert that compilation takes place in a hosted environment. This implies
1121 @option{-fbuiltin}. A hosted environment is one in which the
1122 entire standard library is available, and in which @code{main} has a return
1123 type of @code{int}. Examples are nearly everything except a kernel.
1124 This is equivalent to @option{-fno-freestanding}.
1126 @item -ffreestanding
1127 @opindex ffreestanding
1128 @cindex hosted environment
1130 Assert that compilation takes place in a freestanding environment. This
1131 implies @option{-fno-builtin}. A freestanding environment
1132 is one in which the standard library may not exist, and program startup may
1133 not necessarily be at @code{main}. The most obvious example is an OS kernel.
1134 This is equivalent to @option{-fno-hosted}.
1136 @xref{Standards,,Language Standards Supported by GCC}, for details of
1137 freestanding and hosted environments.
1141 Support ISO C trigraphs. The @option{-ansi} option (and @option{-std}
1142 options for strict ISO C conformance) implies @option{-trigraphs}.
1144 @cindex traditional C language
1145 @cindex C language, traditional
1147 @opindex traditional
1148 Attempt to support some aspects of traditional C compilers.
1153 All @code{extern} declarations take effect globally even if they
1154 are written inside of a function definition. This includes implicit
1155 declarations of functions.
1158 The newer keywords @code{typeof}, @code{inline}, @code{signed}, @code{const}
1159 and @code{volatile} are not recognized. (You can still use the
1160 alternative keywords such as @code{__typeof__}, @code{__inline__}, and
1164 Comparisons between pointers and integers are always allowed.
1167 Integer types @code{unsigned short} and @code{unsigned char} promote
1168 to @code{unsigned int}.
1171 Out-of-range floating point literals are not an error.
1174 Certain constructs which ISO regards as a single invalid preprocessing
1175 number, such as @samp{0xe-0xd}, are treated as expressions instead.
1178 String ``constants'' are not necessarily constant; they are stored in
1179 writable space, and identical looking constants are allocated
1180 separately. (This is the same as the effect of
1181 @option{-fwritable-strings}.)
1183 @cindex @code{longjmp} and automatic variables
1185 All automatic variables not declared @code{register} are preserved by
1186 @code{longjmp}. Ordinarily, GNU C follows ISO C: automatic variables
1187 not declared @code{volatile} may be clobbered.
1192 @cindex escape sequences, traditional
1193 The character escape sequences @samp{\x} and @samp{\a} evaluate as the
1194 literal characters @samp{x} and @samp{a} respectively. Without
1195 @w{@option{-traditional}}, @samp{\x} is a prefix for the hexadecimal
1196 representation of a character, and @samp{\a} produces a bell.
1199 This option is deprecated and may be removed.
1201 You may wish to use @option{-fno-builtin} as well as @option{-traditional}
1202 if your program uses names that are normally GNU C built-in functions for
1203 other purposes of its own.
1205 You cannot use @option{-traditional} if you include any header files that
1206 rely on ISO C features. Some vendors are starting to ship systems with
1207 ISO C header files and you cannot use @option{-traditional} on such
1208 systems to compile files that include any system headers.
1210 The @option{-traditional} option also enables @option{-traditional-cpp}.
1212 @item -traditional-cpp
1213 @opindex traditional-cpp
1214 Attempt to support some aspects of traditional C preprocessors.
1215 See the GNU CPP manual for details.
1217 @item -fcond-mismatch
1218 @opindex fcond-mismatch
1219 Allow conditional expressions with mismatched types in the second and
1220 third arguments. The value of such an expression is void. This option
1221 is not supported for C++.
1223 @item -funsigned-char
1224 @opindex funsigned-char
1225 Let the type @code{char} be unsigned, like @code{unsigned char}.
1227 Each kind of machine has a default for what @code{char} should
1228 be. It is either like @code{unsigned char} by default or like
1229 @code{signed char} by default.
1231 Ideally, a portable program should always use @code{signed char} or
1232 @code{unsigned char} when it depends on the signedness of an object.
1233 But many programs have been written to use plain @code{char} and
1234 expect it to be signed, or expect it to be unsigned, depending on the
1235 machines they were written for. This option, and its inverse, let you
1236 make such a program work with the opposite default.
1238 The type @code{char} is always a distinct type from each of
1239 @code{signed char} or @code{unsigned char}, even though its behavior
1240 is always just like one of those two.
1243 @opindex fsigned-char
1244 Let the type @code{char} be signed, like @code{signed char}.
1246 Note that this is equivalent to @option{-fno-unsigned-char}, which is
1247 the negative form of @option{-funsigned-char}. Likewise, the option
1248 @option{-fno-signed-char} is equivalent to @option{-funsigned-char}.
1250 @item -fsigned-bitfields
1251 @itemx -funsigned-bitfields
1252 @itemx -fno-signed-bitfields
1253 @itemx -fno-unsigned-bitfields
1254 @opindex fsigned-bitfields
1255 @opindex funsigned-bitfields
1256 @opindex fno-signed-bitfields
1257 @opindex fno-unsigned-bitfields
1258 These options control whether a bit-field is signed or unsigned, when the
1259 declaration does not use either @code{signed} or @code{unsigned}. By
1260 default, such a bit-field is signed, because this is consistent: the
1261 basic integer types such as @code{int} are signed types.
1263 However, when @option{-traditional} is used, bit-fields are all unsigned
1266 @item -fwritable-strings
1267 @opindex fwritable-strings
1268 Store string constants in the writable data segment and don't uniquize
1269 them. This is for compatibility with old programs which assume they can
1270 write into string constants. The option @option{-traditional} also has
1273 Writing into string constants is a very bad idea; ``constants'' should
1276 @item -fallow-single-precision
1277 @opindex fallow-single-precision
1278 Do not promote single precision math operations to double precision,
1279 even when compiling with @option{-traditional}.
1281 Traditional K&R C promotes all floating point operations to double
1282 precision, regardless of the sizes of the operands. On the
1283 architecture for which you are compiling, single precision may be faster
1284 than double precision. If you must use @option{-traditional}, but want
1285 to use single precision operations when the operands are single
1286 precision, use this option. This option has no effect when compiling
1287 with ISO or GNU C conventions (the default).
1290 @opindex fshort-wchar
1291 Override the underlying type for @samp{wchar_t} to be @samp{short
1292 unsigned int} instead of the default for the target. This option is
1293 useful for building programs to run under WINE@.
1296 @node C++ Dialect Options
1297 @section Options Controlling C++ Dialect
1299 @cindex compiler options, C++
1300 @cindex C++ options, command line
1301 @cindex options, C++
1302 This section describes the command-line options that are only meaningful
1303 for C++ programs; but you can also use most of the GNU compiler options
1304 regardless of what language your program is in. For example, you
1305 might compile a file @code{firstClass.C} like this:
1308 g++ -g -frepo -O -c firstClass.C
1312 In this example, only @option{-frepo} is an option meant
1313 only for C++ programs; you can use the other options with any
1314 language supported by GCC@.
1316 Here is a list of options that are @emph{only} for compiling C++ programs:
1319 @item -fno-access-control
1320 @opindex fno-access-control
1321 Turn off all access checking. This switch is mainly useful for working
1322 around bugs in the access control code.
1326 Check that the pointer returned by @code{operator new} is non-null
1327 before attempting to modify the storage allocated. The current Working
1328 Paper requires that @code{operator new} never return a null pointer, so
1329 this check is normally unnecessary.
1331 An alternative to using this option is to specify that your
1332 @code{operator new} does not throw any exceptions; if you declare it
1333 @samp{throw()}, G++ will check the return value. See also @samp{new
1336 @item -fconserve-space
1337 @opindex fconserve-space
1338 Put uninitialized or runtime-initialized global variables into the
1339 common segment, as C does. This saves space in the executable at the
1340 cost of not diagnosing duplicate definitions. If you compile with this
1341 flag and your program mysteriously crashes after @code{main()} has
1342 completed, you may have an object that is being destroyed twice because
1343 two definitions were merged.
1345 This option is no longer useful on most targets, now that support has
1346 been added for putting variables into BSS without making them common.
1348 @item -fno-const-strings
1349 @opindex fno-const-strings
1350 Give string constants type @code{char *} instead of type @code{const
1351 char *}. By default, G++ uses type @code{const char *} as required by
1352 the standard. Even if you use @option{-fno-const-strings}, you cannot
1353 actually modify the value of a string constant, unless you also use
1354 @option{-fwritable-strings}.
1356 This option might be removed in a future release of G++. For maximum
1357 portability, you should structure your code so that it works with
1358 string constants that have type @code{const char *}.
1360 @item -fdollars-in-identifiers
1361 @opindex fdollars-in-identifiers
1362 Accept @samp{$} in identifiers. You can also explicitly prohibit use of
1363 @samp{$} with the option @option{-fno-dollars-in-identifiers}. (GNU C allows
1364 @samp{$} by default on most target systems, but there are a few exceptions.)
1365 Traditional C allowed the character @samp{$} to form part of
1366 identifiers. However, ISO C and C++ forbid @samp{$} in identifiers.
1368 @item -fno-elide-constructors
1369 @opindex fno-elide-constructors
1370 The C++ standard allows an implementation to omit creating a temporary
1371 which is only used to initialize another object of the same type.
1372 Specifying this option disables that optimization, and forces G++ to
1373 call the copy constructor in all cases.
1375 @item -fno-enforce-eh-specs
1376 @opindex fno-enforce-eh-specs
1377 Don't check for violation of exception specifications at runtime. This
1378 option violates the C++ standard, but may be useful for reducing code
1379 size in production builds, much like defining @samp{NDEBUG}. The compiler
1380 will still optimize based on the exception specifications.
1382 @item -fexternal-templates
1383 @opindex fexternal-templates
1385 Cause @samp{#pragma interface} and @samp{implementation} to apply to
1386 template instantiation; template instances are emitted or not according
1387 to the location of the template definition. @xref{Template
1388 Instantiation}, for more information.
1390 This option is deprecated.
1392 @item -falt-external-templates
1393 @opindex falt-external-templates
1394 Similar to @option{-fexternal-templates}, but template instances are
1395 emitted or not according to the place where they are first instantiated.
1396 @xref{Template Instantiation}, for more information.
1398 This option is deprecated.
1401 @itemx -fno-for-scope
1403 @opindex fno-for-scope
1404 If @option{-ffor-scope} is specified, the scope of variables declared in
1405 a @i{for-init-statement} is limited to the @samp{for} loop itself,
1406 as specified by the C++ standard.
1407 If @option{-fno-for-scope} is specified, the scope of variables declared in
1408 a @i{for-init-statement} extends to the end of the enclosing scope,
1409 as was the case in old versions of G++, and other (traditional)
1410 implementations of C++.
1412 The default if neither flag is given to follow the standard,
1413 but to allow and give a warning for old-style code that would
1414 otherwise be invalid, or have different behavior.
1416 @item -fno-gnu-keywords
1417 @opindex fno-gnu-keywords
1418 Do not recognize @code{typeof} as a keyword, so that code can use this
1419 word as an identifier. You can use the keyword @code{__typeof__} instead.
1420 @option{-ansi} implies @option{-fno-gnu-keywords}.
1422 @item -fno-implicit-templates
1423 @opindex fno-implicit-templates
1424 Never emit code for non-inline templates which are instantiated
1425 implicitly (i.e.@: by use); only emit code for explicit instantiations.
1426 @xref{Template Instantiation}, for more information.
1428 @item -fno-implicit-inline-templates
1429 @opindex fno-implicit-inline-templates
1430 Don't emit code for implicit instantiations of inline templates, either.
1431 The default is to handle inlines differently so that compiles with and
1432 without optimization will need the same set of explicit instantiations.
1434 @item -fno-implement-inlines
1435 @opindex fno-implement-inlines
1436 To save space, do not emit out-of-line copies of inline functions
1437 controlled by @samp{#pragma implementation}. This will cause linker
1438 errors if these functions are not inlined everywhere they are called.
1440 @item -fms-extensions
1441 @opindex fms-extensions
1442 Disable pedantic warnings about constructs used in MFC, such as implicit
1443 int and getting a pointer to member function via non-standard syntax.
1445 @item -fno-nonansi-builtins
1446 @opindex fno-nonansi-builtins
1447 Disable built-in declarations of functions that are not mandated by
1448 ANSI/ISO C@. These include @code{ffs}, @code{alloca}, @code{_exit},
1449 @code{index}, @code{bzero}, @code{conjf}, and other related functions.
1451 @item -fno-operator-names
1452 @opindex fno-operator-names
1453 Do not treat the operator name keywords @code{and}, @code{bitand},
1454 @code{bitor}, @code{compl}, @code{not}, @code{or} and @code{xor} as
1455 synonyms as keywords.
1457 @item -fno-optional-diags
1458 @opindex fno-optional-diags
1459 Disable diagnostics that the standard says a compiler does not need to
1460 issue. Currently, the only such diagnostic issued by G++ is the one for
1461 a name having multiple meanings within a class.
1464 @opindex fpermissive
1465 Downgrade messages about nonconformant code from errors to warnings. By
1466 default, G++ effectively sets @option{-pedantic-errors} without
1467 @option{-pedantic}; this option reverses that. This behavior and this
1468 option are superseded by @option{-pedantic}, which works as it does for GNU C@.
1472 Enable automatic template instantiation at link time. This option also
1473 implies @option{-fno-implicit-templates}. @xref{Template
1474 Instantiation}, for more information.
1478 Disable generation of information about every class with virtual
1479 functions for use by the C++ runtime type identification features
1480 (@samp{dynamic_cast} and @samp{typeid}). If you don't use those parts
1481 of the language, you can save some space by using this flag. Note that
1482 exception handling uses the same information, but it will generate it as
1487 Emit statistics about front-end processing at the end of the compilation.
1488 This information is generally only useful to the G++ development team.
1490 @item -ftemplate-depth-@var{n}
1491 @opindex ftemplate-depth
1492 Set the maximum instantiation depth for template classes to @var{n}.
1493 A limit on the template instantiation depth is needed to detect
1494 endless recursions during template class instantiation. ANSI/ISO C++
1495 conforming programs must not rely on a maximum depth greater than 17.
1497 @item -fuse-cxa-atexit
1498 @opindex fuse-cxa-atexit
1499 Register destructors for objects with static storage duration with the
1500 @code{__cxa_atexit} function rather than the @code{atexit} function.
1501 This option is required for fully standards-compliant handling of static
1502 destructors, but will only work if your C library supports
1503 @code{__cxa_atexit}.
1507 Emit special relocations for vtables and virtual function references
1508 so that the linker can identify unused virtual functions and zero out
1509 vtable slots that refer to them. This is most useful with
1510 @option{-ffunction-sections} and @option{-Wl,--gc-sections}, in order to
1511 also discard the functions themselves.
1513 This optimization requires GNU as and GNU ld. Not all systems support
1514 this option. @option{-Wl,--gc-sections} is ignored without @option{-static}.
1518 Do not use weak symbol support, even if it is provided by the linker.
1519 By default, G++ will use weak symbols if they are available. This
1520 option exists only for testing, and should not be used by end-users;
1521 it will result in inferior code and has no benefits. This option may
1522 be removed in a future release of G++.
1526 Do not search for header files in the standard directories specific to
1527 C++, but do still search the other standard directories. (This option
1528 is used when building the C++ library.)
1531 In addition, these optimization, warning, and code generation options
1532 have meanings only for C++ programs:
1535 @item -fno-default-inline
1536 @opindex fno-default-inline
1537 Do not assume @samp{inline} for functions defined inside a class scope.
1538 @xref{Optimize Options,,Options That Control Optimization}. Note that these
1539 functions will have linkage like inline functions; they just won't be
1542 @item -Wctor-dtor-privacy @r{(C++ only)}
1543 @opindex Wctor-dtor-privacy
1544 Warn when a class seems unusable, because all the constructors or
1545 destructors in a class are private and the class has no friends or
1546 public static member functions.
1548 @item -Wnon-virtual-dtor @r{(C++ only)}
1549 @opindex Wnon-virtual-dtor
1550 Warn when a class declares a non-virtual destructor that should probably
1551 be virtual, because it looks like the class will be used polymorphically.
1553 @item -Wreorder @r{(C++ only)}
1555 @cindex reordering, warning
1556 @cindex warning for reordering of member initializers
1557 Warn when the order of member initializers given in the code does not
1558 match the order in which they must be executed. For instance:
1564 A(): j (0), i (1) @{ @}
1568 Here the compiler will warn that the member initializers for @samp{i}
1569 and @samp{j} will be rearranged to match the declaration order of the
1573 The following @option{-W@dots{}} options are not affected by @option{-Wall}.
1576 @item -Weffc++ @r{(C++ only)}
1578 Warn about violations of the following style guidelines from Scott Meyers'
1579 @cite{Effective C++} book:
1583 Item 11: Define a copy constructor and an assignment operator for classes
1584 with dynamically allocated memory.
1587 Item 12: Prefer initialization to assignment in constructors.
1590 Item 14: Make destructors virtual in base classes.
1593 Item 15: Have @code{operator=} return a reference to @code{*this}.
1596 Item 23: Don't try to return a reference when you must return an object.
1600 and about violations of the following style guidelines from Scott Meyers'
1601 @cite{More Effective C++} book:
1605 Item 6: Distinguish between prefix and postfix forms of increment and
1606 decrement operators.
1609 Item 7: Never overload @code{&&}, @code{||}, or @code{,}.
1613 If you use this option, you should be aware that the standard library
1614 headers do not obey all of these guidelines; you can use @samp{grep -v}
1615 to filter out those warnings.
1617 @item -Wno-deprecated @r{(C++ only)}
1618 @opindex Wno-deprecated
1619 Do not warn about usage of deprecated features. @xref{Deprecated Features}.
1621 @item -Wno-non-template-friend @r{(C++ only)}
1622 @opindex Wno-non-template-friend
1623 Disable warnings when non-templatized friend functions are declared
1624 within a template. With the advent of explicit template specification
1625 support in G++, if the name of the friend is an unqualified-id (i.e.,
1626 @samp{friend foo(int)}), the C++ language specification demands that the
1627 friend declare or define an ordinary, nontemplate function. (Section
1628 14.5.3). Before G++ implemented explicit specification, unqualified-ids
1629 could be interpreted as a particular specialization of a templatized
1630 function. Because this non-conforming behavior is no longer the default
1631 behavior for G++, @option{-Wnon-template-friend} allows the compiler to
1632 check existing code for potential trouble spots, and is on by default.
1633 This new compiler behavior can be turned off with
1634 @option{-Wno-non-template-friend} which keeps the conformant compiler code
1635 but disables the helpful warning.
1637 @item -Wold-style-cast @r{(C++ only)}
1638 @opindex Wold-style-cast
1639 Warn if an old-style (C-style) cast to a non-void type is used within
1640 a C++ program. The new-style casts (@samp{static_cast},
1641 @samp{reinterpret_cast}, and @samp{const_cast}) are less vulnerable to
1642 unintended effects, and much easier to grep for.
1644 @item -Woverloaded-virtual @r{(C++ only)}
1645 @opindex Woverloaded-virtual
1646 @cindex overloaded virtual fn, warning
1647 @cindex warning for overloaded virtual fn
1648 Warn when a function declaration hides virtual functions from a
1649 base class. For example, in:
1656 struct B: public A @{
1661 the @code{A} class version of @code{f} is hidden in @code{B}, and code
1669 will fail to compile.
1671 @item -Wno-pmf-conversions @r{(C++ only)}
1672 @opindex Wno-pmf-conversions
1673 Disable the diagnostic for converting a bound pointer to member function
1676 @item -Wsign-promo @r{(C++ only)}
1677 @opindex Wsign-promo
1678 Warn when overload resolution chooses a promotion from unsigned or
1679 enumeral type to a signed type over a conversion to an unsigned type of
1680 the same size. Previous versions of G++ would try to preserve
1681 unsignedness, but the standard mandates the current behavior.
1683 @item -Wsynth @r{(C++ only)}
1685 @cindex warning for synthesized methods
1686 @cindex synthesized methods, warning
1687 Warn when G++'s synthesis behavior does not match that of cfront. For
1693 A& operator = (int);
1703 In this example, G++ will synthesize a default @samp{A& operator =
1704 (const A&);}, while cfront will use the user-defined @samp{operator =}.
1707 @node Objective-C Dialect Options
1708 @section Options Controlling Objective-C Dialect
1710 @cindex compiler options, Objective-C
1711 @cindex Objective-C options, command line
1712 @cindex options, Objective-C
1713 This section describes the command-line options that are only meaningful
1714 for Objective-C programs; but you can also use most of the GNU compiler
1715 options regardless of what language your program is in. For example,
1716 you might compile a file @code{some_class.m} like this:
1719 gcc -g -fgnu-runtime -O -c some_class.m
1723 In this example, only @option{-fgnu-runtime} is an option meant only for
1724 Objective-C programs; you can use the other options with any language
1727 Here is a list of options that are @emph{only} for compiling Objective-C
1731 @item -fconstant-string-class=@var{class-name}
1732 @opindex fconstant-string-class
1733 Use @var{class-name} as the name of the class to instantiate for each
1734 literal string specified with the syntax @code{@@"@dots{}"}. The default
1735 class name is @code{NXConstantString}.
1738 @opindex fgnu-runtime
1739 Generate object code compatible with the standard GNU Objective-C
1740 runtime. This is the default for most types of systems.
1742 @item -fnext-runtime
1743 @opindex fnext-runtime
1744 Generate output compatible with the NeXT runtime. This is the default
1745 for NeXT-based systems, including Darwin and Mac OS X@.
1749 Dump interface declarations for all classes seen in the source file to a
1750 file named @file{@var{sourcename}.decl}.
1753 @opindex Wno-protocol
1754 Do not warn if methods required by a protocol are not implemented
1755 in the class adopting it.
1759 Warn if a selector has multiple methods of different types defined.
1761 @c not documented because only avail via -Wp
1762 @c @item -print-objc-runtime-info
1766 @node Language Independent Options
1767 @section Options to Control Diagnostic Messages Formatting
1768 @cindex options to control diagnostics formatting
1769 @cindex diagnostic messages
1770 @cindex message formatting
1772 Traditionally, diagnostic messages have been formatted irrespective of
1773 the output device's aspect (e.g.@: its width, @dots{}). The options described
1774 below can be used to control the diagnostic messages formatting
1775 algorithm, e.g.@: how many characters per line, how often source location
1776 information should be reported. Right now, only the C++ front end can
1777 honor these options. However it is expected, in the near future, that
1778 the remaining front ends would be able to digest them correctly.
1781 @item -fmessage-length=@var{n}
1782 @opindex fmessage-length
1783 Try to format error messages so that they fit on lines of about @var{n}
1784 characters. The default is 72 characters for @command{g++} and 0 for the rest of
1785 the front ends supported by GCC@. If @var{n} is zero, then no
1786 line-wrapping will be done; each error message will appear on a single
1789 @opindex fdiagnostics-show-location
1790 @item -fdiagnostics-show-location=once
1791 Only meaningful in line-wrapping mode. Instructs the diagnostic messages
1792 reporter to emit @emph{once} source location information; that is, in
1793 case the message is too long to fit on a single physical line and has to
1794 be wrapped, the source location won't be emitted (as prefix) again,
1795 over and over, in subsequent continuation lines. This is the default
1798 @item -fdiagnostics-show-location=every-line
1799 Only meaningful in line-wrapping mode. Instructs the diagnostic
1800 messages reporter to emit the same source location information (as
1801 prefix) for physical lines that result from the process of breaking
1802 a message which is too long to fit on a single line.
1806 @node Warning Options
1807 @section Options to Request or Suppress Warnings
1808 @cindex options to control warnings
1809 @cindex warning messages
1810 @cindex messages, warning
1811 @cindex suppressing warnings
1813 Warnings are diagnostic messages that report constructions which
1814 are not inherently erroneous but which are risky or suggest there
1815 may have been an error.
1817 You can request many specific warnings with options beginning @samp{-W},
1818 for example @option{-Wimplicit} to request warnings on implicit
1819 declarations. Each of these specific warning options also has a
1820 negative form beginning @samp{-Wno-} to turn off warnings;
1821 for example, @option{-Wno-implicit}. This manual lists only one of the
1822 two forms, whichever is not the default.
1824 These options control the amount and kinds of warnings produced by GCC:
1827 @cindex syntax checking
1829 @opindex fsyntax-only
1830 Check the code for syntax errors, but don't do anything beyond that.
1834 Issue all the warnings demanded by strict ISO C and ISO C++;
1835 reject all programs that use forbidden extensions, and some other
1836 programs that do not follow ISO C and ISO C++. For ISO C, follows the
1837 version of the ISO C standard specified by any @option{-std} option used.
1839 Valid ISO C and ISO C++ programs should compile properly with or without
1840 this option (though a rare few will require @option{-ansi} or a
1841 @option{-std} option specifying the required version of ISO C)@. However,
1842 without this option, certain GNU extensions and traditional C and C++
1843 features are supported as well. With this option, they are rejected.
1845 @option{-pedantic} does not cause warning messages for use of the
1846 alternate keywords whose names begin and end with @samp{__}. Pedantic
1847 warnings are also disabled in the expression that follows
1848 @code{__extension__}. However, only system header files should use
1849 these escape routes; application programs should avoid them.
1850 @xref{Alternate Keywords}.
1852 Some users try to use @option{-pedantic} to check programs for strict ISO
1853 C conformance. They soon find that it does not do quite what they want:
1854 it finds some non-ISO practices, but not all---only those for which
1855 ISO C @emph{requires} a diagnostic, and some others for which
1856 diagnostics have been added.
1858 A feature to report any failure to conform to ISO C might be useful in
1859 some instances, but would require considerable additional work and would
1860 be quite different from @option{-pedantic}. We don't have plans to
1861 support such a feature in the near future.
1863 Where the standard specified with @option{-std} represents a GNU
1864 extended dialect of C, such as @samp{gnu89} or @samp{gnu99}, there is a
1865 corresponding @dfn{base standard}, the version of ISO C on which the GNU
1866 extended dialect is based. Warnings from @option{-pedantic} are given
1867 where they are required by the base standard. (It would not make sense
1868 for such warnings to be given only for features not in the specified GNU
1869 C dialect, since by definition the GNU dialects of C include all
1870 features the compiler supports with the given option, and there would be
1871 nothing to warn about.)
1873 @item -pedantic-errors
1874 @opindex pedantic-errors
1875 Like @option{-pedantic}, except that errors are produced rather than
1880 Inhibit all warning messages.
1884 Inhibit warning messages about the use of @samp{#import}.
1886 @item -Wchar-subscripts
1887 @opindex Wchar-subscripts
1888 Warn if an array subscript has type @code{char}. This is a common cause
1889 of error, as programmers often forget that this type is signed on some
1894 Warn whenever a comment-start sequence @samp{/*} appears in a @samp{/*}
1895 comment, or whenever a Backslash-Newline appears in a @samp{//} comment.
1899 Check calls to @code{printf} and @code{scanf}, etc., to make sure that
1900 the arguments supplied have types appropriate to the format string
1901 specified, and that the conversions specified in the format string make
1902 sense. This includes standard functions, and others specified by format
1903 attributes (@pxref{Function Attributes}), in the @code{printf},
1904 @code{scanf}, @code{strftime} and @code{strfmon} (an X/Open extension,
1905 not in the C standard) families.
1907 The formats are checked against the format features supported by GNU
1908 libc version 2.2. These include all ISO C89 and C99 features, as well
1909 as features from the Single Unix Specification and some BSD and GNU
1910 extensions. Other library implementations may not support all these
1911 features; GCC does not support warning about features that go beyond a
1912 particular library's limitations. However, if @option{-pedantic} is used
1913 with @option{-Wformat}, warnings will be given about format features not
1914 in the selected standard version (but not for @code{strfmon} formats,
1915 since those are not in any version of the C standard). @xref{C Dialect
1916 Options,,Options Controlling C Dialect}.
1918 @option{-Wformat} is included in @option{-Wall}. For more control over some
1919 aspects of format checking, the options @option{-Wno-format-y2k},
1920 @option{-Wno-format-extra-args}, @option{-Wformat-nonliteral},
1921 @option{-Wformat-security} and @option{-Wformat=2} are available, but are
1922 not included in @option{-Wall}.
1924 @item -Wno-format-y2k
1925 @opindex Wno-format-y2k
1926 If @option{-Wformat} is specified, do not warn about @code{strftime}
1927 formats which may yield only a two-digit year.
1929 @item -Wno-format-extra-args
1930 @opindex Wno-format-extra-args
1931 If @option{-Wformat} is specified, do not warn about excess arguments to a
1932 @code{printf} or @code{scanf} format function. The C standard specifies
1933 that such arguments are ignored.
1935 Where the unused arguments lie between used arguments that are
1936 specified with @samp{$} operand number specifications, normally
1937 warnings are still given, since the implementation could not know what
1938 type to pass to @code{va_arg} to skip the unused arguments. However,
1939 in the case of @code{scanf} formats, this option will suppress the
1940 warning if the unused arguments are all pointers, since the Single
1941 Unix Specification says that such unused arguments are allowed.
1943 @item -Wformat-nonliteral
1944 @opindex Wformat-nonliteral
1945 If @option{-Wformat} is specified, also warn if the format string is not a
1946 string literal and so cannot be checked, unless the format function
1947 takes its format arguments as a @code{va_list}.
1949 @item -Wformat-security
1950 @opindex Wformat-security
1951 If @option{-Wformat} is specified, also warn about uses of format
1952 functions that represent possible security problems. At present, this
1953 warns about calls to @code{printf} and @code{scanf} functions where the
1954 format string is not a string literal and there are no format arguments,
1955 as in @code{printf (foo);}. This may be a security hole if the format
1956 string came from untrusted input and contains @samp{%n}. (This is
1957 currently a subset of what @option{-Wformat-nonliteral} warns about, but
1958 in future warnings may be added to @option{-Wformat-security} that are not
1959 included in @option{-Wformat-nonliteral}.)
1963 Enable @option{-Wformat} plus format checks not included in
1964 @option{-Wformat}. Currently equivalent to @samp{-Wformat
1965 -Wformat-nonliteral -Wformat-security}.
1967 @item -Wimplicit-int
1968 @opindex Wimplicit-int
1969 Warn when a declaration does not specify a type.
1971 @item -Wimplicit-function-declaration
1972 @itemx -Werror-implicit-function-declaration
1973 @opindex Wimplicit-function-declaration
1974 @opindex Werror-implicit-function-declaration
1975 Give a warning (or error) whenever a function is used before being
1980 Same as @option{-Wimplicit-int} and @option{-Wimplicit-function-declaration}.
1984 Warn if the type of @samp{main} is suspicious. @samp{main} should be a
1985 function with external linkage, returning int, taking either zero
1986 arguments, two, or three arguments of appropriate types.
1988 @item -Wmissing-braces
1989 @opindex Wmissing-braces
1990 Warn if an aggregate or union initializer is not fully bracketed. In
1991 the following example, the initializer for @samp{a} is not fully
1992 bracketed, but that for @samp{b} is fully bracketed.
1995 int a[2][2] = @{ 0, 1, 2, 3 @};
1996 int b[2][2] = @{ @{ 0, 1 @}, @{ 2, 3 @} @};
2000 @opindex Wparentheses
2001 Warn if parentheses are omitted in certain contexts, such
2002 as when there is an assignment in a context where a truth value
2003 is expected, or when operators are nested whose precedence people
2004 often get confused about.
2006 Also warn about constructions where there may be confusion to which
2007 @code{if} statement an @code{else} branch belongs. Here is an example of
2022 In C, every @code{else} branch belongs to the innermost possible @code{if}
2023 statement, which in this example is @code{if (b)}. This is often not
2024 what the programmer expected, as illustrated in the above example by
2025 indentation the programmer chose. When there is the potential for this
2026 confusion, GCC will issue a warning when this flag is specified.
2027 To eliminate the warning, add explicit braces around the innermost
2028 @code{if} statement so there is no way the @code{else} could belong to
2029 the enclosing @code{if}. The resulting code would look like this:
2045 @item -Wsequence-point
2046 @opindex Wsequence-point
2047 Warn about code that may have undefined semantics because of violations
2048 of sequence point rules in the C standard.
2050 The C standard defines the order in which expressions in a C program are
2051 evaluated in terms of @dfn{sequence points}, which represent a partial
2052 ordering between the execution of parts of the program: those executed
2053 before the sequence point, and those executed after it. These occur
2054 after the evaluation of a full expression (one which is not part of a
2055 larger expression), after the evaluation of the first operand of a
2056 @code{&&}, @code{||}, @code{? :} or @code{,} (comma) operator, before a
2057 function is called (but after the evaluation of its arguments and the
2058 expression denoting the called function), and in certain other places.
2059 Other than as expressed by the sequence point rules, the order of
2060 evaluation of subexpressions of an expression is not specified. All
2061 these rules describe only a partial order rather than a total order,
2062 since, for example, if two functions are called within one expression
2063 with no sequence point between them, the order in which the functions
2064 are called is not specified. However, the standards committee have
2065 ruled that function calls do not overlap.
2067 It is not specified when between sequence points modifications to the
2068 values of objects take effect. Programs whose behavior depends on this
2069 have undefined behavior; the C standard specifies that ``Between the
2070 previous and next sequence point an object shall have its stored value
2071 modified at most once by the evaluation of an expression. Furthermore,
2072 the prior value shall be read only to determine the value to be
2073 stored.''. If a program breaks these rules, the results on any
2074 particular implementation are entirely unpredictable.
2076 Examples of code with undefined behavior are @code{a = a++;}, @code{a[n]
2077 = b[n++]} and @code{a[i++] = i;}. Some more complicated cases are not
2078 diagnosed by this option, and it may give an occasional false positive
2079 result, but in general it has been found fairly effective at detecting
2080 this sort of problem in programs.
2082 The present implementation of this option only works for C programs. A
2083 future implementation may also work for C++ programs.
2085 The C standard is worded confusingly, therefore there is some debate
2086 over the precise meaning of the sequence point rules in subtle cases.
2087 Links to discussions of the problem, including proposed formal
2088 definitions, may be found on our readings page, at
2089 @w{@uref{http://gcc.gnu.org/readings.html}}.
2092 @opindex Wreturn-type
2093 Warn whenever a function is defined with a return-type that defaults to
2094 @code{int}. Also warn about any @code{return} statement with no
2095 return-value in a function whose return-type is not @code{void}.
2097 For C++, a function without return type always produces a diagnostic
2098 message, even when @option{-Wno-return-type} is specified. The only
2099 exceptions are @samp{main} and functions defined in system headers.
2103 Warn whenever a @code{switch} statement has an index of enumeral type
2104 and lacks a @code{case} for one or more of the named codes of that
2105 enumeration. (The presence of a @code{default} label prevents this
2106 warning.) @code{case} labels outside the enumeration range also
2107 provoke warnings when this option is used.
2111 Warn if any trigraphs are encountered that might change the meaning of
2112 the program (trigraphs within comments are not warned about).
2114 @item -Wunused-function
2115 @opindex Wunused-function
2116 Warn whenever a static function is declared but not defined or a
2117 non\-inline static function is unused.
2119 @item -Wunused-label
2120 @opindex Wunused-label
2121 Warn whenever a label is declared but not used.
2123 To suppress this warning use the @samp{unused} attribute
2124 (@pxref{Variable Attributes}).
2126 @item -Wunused-parameter
2127 @opindex Wunused-parameter
2128 Warn whenever a function parameter is unused aside from its declaration.
2130 To suppress this warning use the @samp{unused} attribute
2131 (@pxref{Variable Attributes}).
2133 @item -Wunused-variable
2134 @opindex Wunused-variable
2135 Warn whenever a local variable or non-constant static variable is unused
2136 aside from its declaration
2138 To suppress this warning use the @samp{unused} attribute
2139 (@pxref{Variable Attributes}).
2141 @item -Wunused-value
2142 @opindex Wunused-value
2143 Warn whenever a statement computes a result that is explicitly not used.
2145 To suppress this warning cast the expression to @samp{void}.
2149 All all the above @option{-Wunused} options combined.
2151 In order to get a warning about an unused function parameter, you must
2152 either specify @samp{-W -Wunused} or separately specify
2153 @option{-Wunused-parameter}.
2155 @item -Wuninitialized
2156 @opindex Wuninitialized
2157 Warn if an automatic variable is used without first being initialized or
2158 if a variable may be clobbered by a @code{setjmp} call.
2160 These warnings are possible only in optimizing compilation,
2161 because they require data flow information that is computed only
2162 when optimizing. If you don't specify @option{-O}, you simply won't
2165 These warnings occur only for variables that are candidates for
2166 register allocation. Therefore, they do not occur for a variable that
2167 is declared @code{volatile}, or whose address is taken, or whose size
2168 is other than 1, 2, 4 or 8 bytes. Also, they do not occur for
2169 structures, unions or arrays, even when they are in registers.
2171 Note that there may be no warning about a variable that is used only
2172 to compute a value that itself is never used, because such
2173 computations may be deleted by data flow analysis before the warnings
2176 These warnings are made optional because GCC is not smart
2177 enough to see all the reasons why the code might be correct
2178 despite appearing to have an error. Here is one example of how
2199 If the value of @code{y} is always 1, 2 or 3, then @code{x} is
2200 always initialized, but GCC doesn't know this. Here is
2201 another common case:
2206 if (change_y) save_y = y, y = new_y;
2208 if (change_y) y = save_y;
2213 This has no bug because @code{save_y} is used only if it is set.
2215 @cindex @code{longjmp} warnings
2216 This option also warns when a non-volatile automatic variable might be
2217 changed by a call to @code{longjmp}. These warnings as well are possible
2218 only in optimizing compilation.
2220 The compiler sees only the calls to @code{setjmp}. It cannot know
2221 where @code{longjmp} will be called; in fact, a signal handler could
2222 call it at any point in the code. As a result, you may get a warning
2223 even when there is in fact no problem because @code{longjmp} cannot
2224 in fact be called at the place which would cause a problem.
2226 Some spurious warnings can be avoided if you declare all the functions
2227 you use that never return as @code{noreturn}. @xref{Function
2230 @item -Wreorder @r{(C++ only)}
2232 @cindex reordering, warning
2233 @cindex warning for reordering of member initializers
2234 Warn when the order of member initializers given in the code does not
2235 match the order in which they must be executed. For instance:
2237 @item -Wunknown-pragmas
2238 @opindex Wunknown-pragmas
2239 @cindex warning for unknown pragmas
2240 @cindex unknown pragmas, warning
2241 @cindex pragmas, warning of unknown
2242 Warn when a #pragma directive is encountered which is not understood by
2243 GCC@. If this command line option is used, warnings will even be issued
2244 for unknown pragmas in system header files. This is not the case if
2245 the warnings were only enabled by the @option{-Wall} command line option.
2249 All of the above @samp{-W} options combined. This enables all the
2250 warnings about constructions that some users consider questionable, and
2251 that are easy to avoid (or modify to prevent the warning), even in
2252 conjunction with macros.
2255 @opindex Wno-div-by-zero
2256 @opindex Wdiv-by-zero
2257 Warn about compile-time integer division by zero. This is default. To
2258 inhibit the warning messages, use @option{-Wno-div-by-zero}. Floating
2259 point division by zero is not warned about, as it can be a legitimate
2260 way of obtaining infinities and NaNs.
2263 @opindex Wno-multichar
2265 Warn if a multicharacter constant (@samp{'FOOF'}) is used. This is
2266 default. To inhibit the warning messages, use @option{-Wno-multichar}.
2267 Usually they indicate a typo in the user's code, as they have
2268 implementation-defined values, and should not be used in portable code.
2270 @item -Wsystem-headers
2271 @opindex Wsystem-headers
2272 @cindex warnings from system headers
2273 @cindex system headers, warnings from
2274 Print warning messages for constructs found in system header files.
2275 Warnings from system headers are normally suppressed, on the assumption
2276 that they usually do not indicate real problems and would only make the
2277 compiler output harder to read. Using this command line option tells
2278 GCC to emit warnings from system headers as if they occurred in user
2279 code. However, note that using @option{-Wall} in conjunction with this
2280 option will @emph{not} warn about unknown pragmas in system
2281 headers---for that, @option{-Wunknown-pragmas} must also be used.
2284 The following @option{-W@dots{}} options are not implied by @option{-Wall}.
2285 Some of them warn about constructions that users generally do not
2286 consider questionable, but which occasionally you might wish to check
2287 for; others warn about constructions that are necessary or hard to avoid
2288 in some cases, and there is no simple way to modify the code to suppress
2294 Print extra warning messages for these events:
2298 A function can return either with or without a value. (Falling
2299 off the end of the function body is considered returning without
2300 a value.) For example, this function would evoke such a
2314 An expression-statement or the left-hand side of a comma expression
2315 contains no side effects.
2316 To suppress the warning, cast the unused expression to void.
2317 For example, an expression such as @samp{x[i,j]} will cause a warning,
2318 but @samp{x[(void)i,j]} will not.
2321 An unsigned value is compared against zero with @samp{<} or @samp{<=}.
2324 A comparison like @samp{x<=y<=z} appears; this is equivalent to
2325 @samp{(x<=y ? 1 : 0) <= z}, which is a different interpretation from
2326 that of ordinary mathematical notation.
2329 Storage-class specifiers like @code{static} are not the first things in
2330 a declaration. According to the C Standard, this usage is obsolescent.
2333 The return type of a function has a type qualifier such as @code{const}.
2334 Such a type qualifier has no effect, since the value returned by a
2335 function is not an lvalue. (But don't warn about the GNU extension of
2336 @code{volatile void} return types. That extension will be warned about
2337 if @option{-pedantic} is specified.)
2340 If @option{-Wall} or @option{-Wunused} is also specified, warn about unused
2344 A comparison between signed and unsigned values could produce an
2345 incorrect result when the signed value is converted to unsigned.
2346 (But don't warn if @option{-Wno-sign-compare} is also specified.)
2349 An aggregate has a partly bracketed initializer.
2350 For example, the following code would evoke such a warning,
2351 because braces are missing around the initializer for @code{x.h}:
2354 struct s @{ int f, g; @};
2355 struct t @{ struct s h; int i; @};
2356 struct t x = @{ 1, 2, 3 @};
2360 An aggregate has an initializer which does not initialize all members.
2361 For example, the following code would cause such a warning, because
2362 @code{x.h} would be implicitly initialized to zero:
2365 struct s @{ int f, g, h; @};
2366 struct s x = @{ 3, 4 @};
2371 @opindex Wfloat-equal
2372 Warn if floating point values are used in equality comparisons.
2374 The idea behind this is that sometimes it is convenient (for the
2375 programmer) to consider floating-point values as approximations to
2376 infinitely precise real numbers. If you are doing this, then you need
2377 to compute (by analysing the code, or in some other way) the maximum or
2378 likely maximum error that the computation introduces, and allow for it
2379 when performing comparisons (and when producing output, but that's a
2380 different problem). In particular, instead of testing for equality, you
2381 would check to see whether the two values have ranges that overlap; and
2382 this is done with the relational operators, so equality comparisons are
2385 @item -Wtraditional @r{(C only)}
2386 @opindex Wtraditional
2387 Warn about certain constructs that behave differently in traditional and
2388 ISO C@. Also warn about ISO C constructs that have no traditional C
2389 equivalent, and/or problematic constructs which should be avoided.
2393 Macro parameters that appear within string literals in the macro body.
2394 In traditional C macro replacement takes place within string literals,
2395 but does not in ISO C@.
2398 In traditional C, some preprocessor directives did not exist.
2399 Traditional preprocessors would only consider a line to be a directive
2400 if the @samp{#} appeared in column 1 on the line. Therefore
2401 @option{-Wtraditional} warns about directives that traditional C
2402 understands but would ignore because the @samp{#} does not appear as the
2403 first character on the line. It also suggests you hide directives like
2404 @samp{#pragma} not understood by traditional C by indenting them. Some
2405 traditional implementations would not recognize @samp{#elif}, so it
2406 suggests avoiding it altogether.
2409 A function-like macro that appears without arguments.
2412 The unary plus operator.
2415 The @samp{U} integer constant suffix, or the @samp{F} or @samp{L} floating point
2416 constant suffixes. (Traditional C does support the @samp{L} suffix on integer
2417 constants.) Note, these suffixes appear in macros defined in the system
2418 headers of most modern systems, e.g.@: the @samp{_MIN}/@samp{_MAX} macros in @code{<limits.h>}.
2419 Use of these macros in user code might normally lead to spurious
2420 warnings, however gcc's integrated preprocessor has enough context to
2421 avoid warning in these cases.
2424 A function declared external in one block and then used after the end of
2428 A @code{switch} statement has an operand of type @code{long}.
2431 A non-@code{static} function declaration follows a @code{static} one.
2432 This construct is not accepted by some traditional C compilers.
2435 The ISO type of an integer constant has a different width or
2436 signedness from its traditional type. This warning is only issued if
2437 the base of the constant is ten. I.e.@: hexadecimal or octal values, which
2438 typically represent bit patterns, are not warned about.
2441 Usage of ISO string concatenation is detected.
2444 Initialization of automatic aggregates.
2447 Identifier conflicts with labels. Traditional C lacks a separate
2448 namespace for labels.
2451 Initialization of unions. If the initializer is zero, the warning is
2452 omitted. This is done under the assumption that the zero initializer in
2453 user code appears conditioned on e.g.@: @code{__STDC__} to avoid missing
2454 initializer warnings and relies on default initialization to zero in the
2458 Conversions by prototypes between fixed/floating point values and vice
2459 versa. The absence of these prototypes when compiling with traditional
2460 C would cause serious problems. This is a subset of the possible
2461 conversion warnings, for the full set use @option{-Wconversion}.
2466 Warn if an undefined identifier is evaluated in an @samp{#if} directive.
2470 Warn whenever a local variable shadows another local variable, parameter or
2471 global variable or whenever a built-in function is shadowed.
2473 @item -Wlarger-than-@var{len}
2474 @opindex Wlarger-than
2475 Warn whenever an object of larger than @var{len} bytes is defined.
2477 @item -Wpointer-arith
2478 @opindex Wpointer-arith
2479 Warn about anything that depends on the ``size of'' a function type or
2480 of @code{void}. GNU C assigns these types a size of 1, for
2481 convenience in calculations with @code{void *} pointers and pointers
2484 @item -Wbad-function-cast @r{(C only)}
2485 @opindex Wbad-function-cast
2486 Warn whenever a function call is cast to a non-matching type.
2487 For example, warn if @code{int malloc()} is cast to @code{anything *}.
2491 Warn whenever a pointer is cast so as to remove a type qualifier from
2492 the target type. For example, warn if a @code{const char *} is cast
2493 to an ordinary @code{char *}.
2496 @opindex Wcast-align
2497 Warn whenever a pointer is cast such that the required alignment of the
2498 target is increased. For example, warn if a @code{char *} is cast to
2499 an @code{int *} on machines where integers can only be accessed at
2500 two- or four-byte boundaries.
2502 @item -Wwrite-strings
2503 @opindex Wwrite-strings
2504 When compiling C, give string constants the type @code{const
2505 char[@var{length}]} so that
2506 copying the address of one into a non-@code{const} @code{char *}
2507 pointer will get a warning; when compiling C++, warn about the
2508 deprecated conversion from string constants to @code{char *}.
2509 These warnings will help you find at
2510 compile time code that can try to write into a string constant, but
2511 only if you have been very careful about using @code{const} in
2512 declarations and prototypes. Otherwise, it will just be a nuisance;
2513 this is why we did not make @option{-Wall} request these warnings.
2516 @opindex Wconversion
2517 Warn if a prototype causes a type conversion that is different from what
2518 would happen to the same argument in the absence of a prototype. This
2519 includes conversions of fixed point to floating and vice versa, and
2520 conversions changing the width or signedness of a fixed point argument
2521 except when the same as the default promotion.
2523 Also, warn if a negative integer constant expression is implicitly
2524 converted to an unsigned type. For example, warn about the assignment
2525 @code{x = -1} if @code{x} is unsigned. But do not warn about explicit
2526 casts like @code{(unsigned) -1}.
2528 @item -Wsign-compare
2529 @opindex Wsign-compare
2530 @cindex warning for comparison of signed and unsigned values
2531 @cindex comparison of signed and unsigned values, warning
2532 @cindex signed and unsigned values, comparison warning
2533 Warn when a comparison between signed and unsigned values could produce
2534 an incorrect result when the signed value is converted to unsigned.
2535 This warning is also enabled by @option{-W}; to get the other warnings
2536 of @option{-W} without this warning, use @samp{-W -Wno-sign-compare}.
2538 @item -Waggregate-return
2539 @opindex Waggregate-return
2540 Warn if any functions that return structures or unions are defined or
2541 called. (In languages where you can return an array, this also elicits
2544 @item -Wstrict-prototypes @r{(C only)}
2545 @opindex Wstrict-prototypes
2546 Warn if a function is declared or defined without specifying the
2547 argument types. (An old-style function definition is permitted without
2548 a warning if preceded by a declaration which specifies the argument
2551 @item -Wmissing-prototypes @r{(C only)}
2552 @opindex Wmissing-prototypes
2553 Warn if a global function is defined without a previous prototype
2554 declaration. This warning is issued even if the definition itself
2555 provides a prototype. The aim is to detect global functions that fail
2556 to be declared in header files.
2558 @item -Wmissing-declarations
2559 @opindex Wmissing-declarations
2560 Warn if a global function is defined without a previous declaration.
2561 Do so even if the definition itself provides a prototype.
2562 Use this option to detect global functions that are not declared in
2565 @item -Wmissing-noreturn
2566 @opindex Wmissing-noreturn
2567 Warn about functions which might be candidates for attribute @code{noreturn}.
2568 Note these are only possible candidates, not absolute ones. Care should
2569 be taken to manually verify functions actually do not ever return before
2570 adding the @code{noreturn} attribute, otherwise subtle code generation
2571 bugs could be introduced. You will not get a warning for @code{main} in
2572 hosted C environments.
2574 @item -Wmissing-format-attribute
2575 @opindex Wmissing-format-attribute
2577 If @option{-Wformat} is enabled, also warn about functions which might be
2578 candidates for @code{format} attributes. Note these are only possible
2579 candidates, not absolute ones. GCC will guess that @code{format}
2580 attributes might be appropriate for any function that calls a function
2581 like @code{vprintf} or @code{vscanf}, but this might not always be the
2582 case, and some functions for which @code{format} attributes are
2583 appropriate may not be detected. This option has no effect unless
2584 @option{-Wformat} is enabled (possibly by @option{-Wall}).
2586 @item -Wno-deprecated-declarations
2587 @opindex Wno-deprecated-declarations
2588 Do not warn about uses of functions, variables, and types marked as
2589 deprecated by using the @code{deprecated} attribute.
2590 (@pxref{Function Attributes}, @pxref{Variable Attributes},
2591 @pxref{Type Attributes}.)
2595 Warn if a structure is given the packed attribute, but the packed
2596 attribute has no effect on the layout or size of the structure.
2597 Such structures may be mis-aligned for little benefit. For
2598 instance, in this code, the variable @code{f.x} in @code{struct bar}
2599 will be misaligned even though @code{struct bar} does not itself
2600 have the packed attribute:
2607 @} __attribute__((packed));
2617 Warn if padding is included in a structure, either to align an element
2618 of the structure or to align the whole structure. Sometimes when this
2619 happens it is possible to rearrange the fields of the structure to
2620 reduce the padding and so make the structure smaller.
2622 @item -Wredundant-decls
2623 @opindex Wredundant-decls
2624 Warn if anything is declared more than once in the same scope, even in
2625 cases where multiple declaration is valid and changes nothing.
2627 @item -Wnested-externs @r{(C only)}
2628 @opindex Wnested-externs
2629 Warn if an @code{extern} declaration is encountered within a function.
2631 @item -Wunreachable-code
2632 @opindex Wunreachable-code
2633 Warn if the compiler detects that code will never be executed.
2635 This option is intended to warn when the compiler detects that at
2636 least a whole line of source code will never be executed, because
2637 some condition is never satisfied or because it is after a
2638 procedure that never returns.
2640 It is possible for this option to produce a warning even though there
2641 are circumstances under which part of the affected line can be executed,
2642 so care should be taken when removing apparently-unreachable code.
2644 For instance, when a function is inlined, a warning may mean that the
2645 line is unreachable in only one inlined copy of the function.
2647 This option is not made part of @option{-Wall} because in a debugging
2648 version of a program there is often substantial code which checks
2649 correct functioning of the program and is, hopefully, unreachable
2650 because the program does work. Another common use of unreachable
2651 code is to provide behavior which is selectable at compile-time.
2655 Warn if a function can not be inlined and it was declared as inline.
2659 @opindex Wno-long-long
2660 Warn if @samp{long long} type is used. This is default. To inhibit
2661 the warning messages, use @option{-Wno-long-long}. Flags
2662 @option{-Wlong-long} and @option{-Wno-long-long} are taken into account
2663 only when @option{-pedantic} flag is used.
2665 @item -Wdisabled-optimization
2666 @opindex Wdisabled-optimization
2667 Warn if a requested optimization pass is disabled. This warning does
2668 not generally indicate that there is anything wrong with your code; it
2669 merely indicates that GCC's optimizers were unable to handle the code
2670 effectively. Often, the problem is that your code is too big or too
2671 complex; GCC will refuse to optimize programs when the optimization
2672 itself is likely to take inordinate amounts of time.
2676 Make all warnings into errors.
2679 @node Debugging Options
2680 @section Options for Debugging Your Program or GCC
2681 @cindex options, debugging
2682 @cindex debugging information options
2684 GCC has various special options that are used for debugging
2685 either your program or GCC:
2690 Produce debugging information in the operating system's native format
2691 (stabs, COFF, XCOFF, or DWARF)@. GDB can work with this debugging
2694 On most systems that use stabs format, @option{-g} enables use of extra
2695 debugging information that only GDB can use; this extra information
2696 makes debugging work better in GDB but will probably make other debuggers
2698 refuse to read the program. If you want to control for certain whether
2699 to generate the extra information, use @option{-gstabs+}, @option{-gstabs},
2700 @option{-gxcoff+}, @option{-gxcoff}, @option{-gdwarf-1+}, @option{-gdwarf-1},
2701 or @option{-gvms} (see below).
2703 Unlike most other C compilers, GCC allows you to use @option{-g} with
2704 @option{-O}. The shortcuts taken by optimized code may occasionally
2705 produce surprising results: some variables you declared may not exist
2706 at all; flow of control may briefly move where you did not expect it;
2707 some statements may not be executed because they compute constant
2708 results or their values were already at hand; some statements may
2709 execute in different places because they were moved out of loops.
2711 Nevertheless it proves possible to debug optimized output. This makes
2712 it reasonable to use the optimizer for programs that might have bugs.
2714 The following options are useful when GCC is generated with the
2715 capability for more than one debugging format.
2719 Produce debugging information for use by GDB@. This means to use the
2720 most expressive format available (DWARF 2, stabs, or the native format
2721 if neither of those are supported), including GDB extensions if at all
2726 Produce debugging information in stabs format (if that is supported),
2727 without GDB extensions. This is the format used by DBX on most BSD
2728 systems. On MIPS, Alpha and System V Release 4 systems this option
2729 produces stabs debugging output which is not understood by DBX or SDB@.
2730 On System V Release 4 systems this option requires the GNU assembler.
2734 Produce debugging information in stabs format (if that is supported),
2735 using GNU extensions understood only by the GNU debugger (GDB)@. The
2736 use of these extensions is likely to make other debuggers crash or
2737 refuse to read the program.
2741 Produce debugging information in COFF format (if that is supported).
2742 This is the format used by SDB on most System V systems prior to
2747 Produce debugging information in XCOFF format (if that is supported).
2748 This is the format used by the DBX debugger on IBM RS/6000 systems.
2752 Produce debugging information in XCOFF format (if that is supported),
2753 using GNU extensions understood only by the GNU debugger (GDB)@. The
2754 use of these extensions is likely to make other debuggers crash or
2755 refuse to read the program, and may cause assemblers other than the GNU
2756 assembler (GAS) to fail with an error.
2760 Produce debugging information in DWARF version 1 format (if that is
2761 supported). This is the format used by SDB on most System V Release 4
2766 Produce debugging information in DWARF version 1 format (if that is
2767 supported), using GNU extensions understood only by the GNU debugger
2768 (GDB)@. The use of these extensions is likely to make other debuggers
2769 crash or refuse to read the program.
2773 Produce debugging information in DWARF version 2 format (if that is
2774 supported). This is the format used by DBX on IRIX 6.
2778 Produce debugging information in VMS debug format (if that is
2779 supported). This is the format used by DEBUG on VMS systems.
2782 @itemx -ggdb@var{level}
2783 @itemx -gstabs@var{level}
2784 @itemx -gcoff@var{level}
2785 @itemx -gxcoff@var{level}
2786 @itemx -gdwarf@var{level}
2787 @itemx -gdwarf-2@var{level}
2788 @itemx -gvms@var{level}
2789 Request debugging information and also use @var{level} to specify how
2790 much information. The default level is 2.
2792 Level 1 produces minimal information, enough for making backtraces in
2793 parts of the program that you don't plan to debug. This includes
2794 descriptions of functions and external variables, but no information
2795 about local variables and no line numbers.
2797 Level 3 includes extra information, such as all the macro definitions
2798 present in the program. Some debuggers support macro expansion when
2799 you use @option{-g3}.
2804 Generate extra code to write profile information suitable for the
2805 analysis program @code{prof}. You must use this option when compiling
2806 the source files you want data about, and you must also use it when
2809 @cindex @code{gprof}
2812 Generate extra code to write profile information suitable for the
2813 analysis program @code{gprof}. You must use this option when compiling
2814 the source files you want data about, and you must also use it when
2820 Generate extra code to write profile information for basic blocks, which will
2821 record the number of times each basic block is executed, the basic block start
2822 address, and the function name containing the basic block. If @option{-g} is
2823 used, the line number and filename of the start of the basic block will also be
2824 recorded. If not overridden by the machine description, the default action is
2825 to append to the text file @file{bb.out}.
2827 This data could be analyzed by a program like @code{tcov}. Note,
2828 however, that the format of the data is not what @code{tcov} expects.
2829 Eventually GNU @code{gprof} should be extended to process this data.
2833 Makes the compiler print out each function name as it is compiled, and
2834 print some statistics about each pass when it finishes.
2837 @opindex ftime-report
2838 Makes the compiler print some statistics about the time consumed by each
2839 pass when it finishes.
2842 @opindex fmem-report
2843 Makes the compiler print some statistics about permanent memory
2844 allocation when it finishes.
2846 @item -fprofile-arcs
2847 @opindex fprofile-arcs
2848 Instrument @dfn{arcs} during compilation to generate coverage data
2849 or for profile-directed block ordering. During execution the program
2850 records how many times each branch is executed and how many times it is
2851 taken. When the compiled program exits it saves this data to a file
2852 called @file{@var{sourcename}.da} for each source file.
2854 For profile-directed block ordering, compile the program with
2855 @option{-fprofile-arcs} plus optimization and code generation options,
2856 generate the arc profile information by running the program on a
2857 selected workload, and then compile the program again with the same
2858 optimization and code generation options plus
2859 @option{-fbranch-probabilities} (@pxref{Optimize Options,,Options that
2860 Control Optimization}).
2862 The other use of @option{-fprofile-arcs} is for use with @code{gcov},
2863 when it is used with the @option{-ftest-coverage} option. GCC
2864 supports two methods of determining code coverage: the options that
2865 support @code{gcov}, and options @option{-a} and @option{-ax}, which
2866 write information to text files. The options that support @code{gcov}
2867 do not need to instrument every arc in the program, so a program compiled
2868 with them runs faster than a program compiled with @option{-a}, which
2869 adds instrumentation code to every basic block in the program. The
2870 tradeoff: since @code{gcov} does not have execution counts for all
2871 branches, it must start with the execution counts for the instrumented
2872 branches, and then iterate over the program flow graph until the entire
2873 graph has been solved. Hence, @code{gcov} runs a little more slowly than
2874 a program which uses information from @option{-a} and @option{-ax}.
2876 With @option{-fprofile-arcs}, for each function of your program GCC
2877 creates a program flow graph, then finds a spanning tree for the graph.
2878 Only arcs that are not on the spanning tree have to be instrumented: the
2879 compiler adds code to count the number of times that these arcs are
2880 executed. When an arc is the only exit or only entrance to a block, the
2881 instrumentation code can be added to the block; otherwise, a new basic
2882 block must be created to hold the instrumentation code.
2884 This option makes it possible to estimate branch probabilities and to
2885 calculate basic block execution counts. In general, basic block
2886 execution counts as provided by @option{-a} do not give enough
2887 information to estimate all branch probabilities.
2890 @item -ftest-coverage
2891 @opindex ftest-coverage
2892 Create data files for the @code{gcov} code-coverage utility
2893 (@pxref{Gcov,, @code{gcov}: a GCC Test Coverage Program}).
2894 The data file names begin with the name of your source file:
2897 @item @var{sourcename}.bb
2898 A mapping from basic blocks to line numbers, which @code{gcov} uses to
2899 associate basic block execution counts with line numbers.
2901 @item @var{sourcename}.bbg
2902 A list of all arcs in the program flow graph. This allows @code{gcov}
2903 to reconstruct the program flow graph, so that it can compute all basic
2904 block and arc execution counts from the information in the
2905 @code{@var{sourcename}.da} file.
2908 Use @option{-ftest-coverage} with @option{-fprofile-arcs}; the latter
2909 option adds instrumentation to the program, which then writes
2910 execution counts to another data file:
2913 @item @var{sourcename}.da
2914 Runtime arc execution counts, used in conjunction with the arc
2915 information in the file @code{@var{sourcename}.bbg}.
2918 Coverage data will map better to the source files if
2919 @option{-ftest-coverage} is used without optimization.
2921 @item -d@var{letters}
2923 Says to make debugging dumps during compilation at times specified by
2924 @var{letters}. This is used for debugging the compiler. The file names
2925 for most of the dumps are made by appending a pass number and a word to
2926 the source file name (e.g. @file{foo.c.00.rtl} or @file{foo.c.01.sibling}).
2927 Here are the possible letters for use in @var{letters}, and their meanings:
2932 Annotate the assembler output with miscellaneous debugging information.
2935 Dump after computing branch probabilities, to @file{@var{file}.14.bp}.
2938 Dump after block reordering, to @file{@var{file}.29.bbro}.
2941 Dump after instruction combination, to the file @file{@var{file}.16.combine}.
2944 Dump after the first if conversion, to the file @file{@var{file}.17.ce}.
2947 Dump after delayed branch scheduling, to @file{@var{file}.31.dbr}.
2950 Dump all macro definitions, at the end of preprocessing, in addition to
2954 Dump after SSA optimizations, to @file{@var{file}.04.ssa} and
2955 @file{@var{file}.07.ussa}.
2958 Dump after the second if conversion, to @file{@var{file}.26.ce2}.
2961 Dump after life analysis, to @file{@var{file}.15.life}.
2964 Dump after purging @code{ADDRESSOF} codes, to @file{@var{file}.09.addressof}.
2967 Dump after global register allocation, to @file{@var{file}.21.greg}.
2970 Dump after finalization of EH handling code, to @file{@var{file}.02.eh}.
2973 Dump after reg-to-stack conversion, to @file{@var{file}.28.stack}.
2976 Dump after post-reload optimizations, to @file{@var{file}.22.postreload}.
2979 Dump after GCSE, to @file{@var{file}.10.gcse}.
2982 Dump after sibling call optimizations, to @file{@var{file}.01.sibling}.
2985 Dump after the first jump optimization, to @file{@var{file}.03.jump}.
2988 Dump after conversion from registers to stack, to @file{@var{file}.32.stack}.
2991 Dump after local register allocation, to @file{@var{file}.20.lreg}.
2994 Dump after loop optimization, to @file{@var{file}.11.loop}.
2997 Dump after performing the machine dependent reorganisation pass, to
2998 @file{@var{file}.30.mach}.
3001 Dump after register renumbering, to @file{@var{file}.25.rnreg}.
3004 Dump after the register move pass, to @file{@var{file}.18.regmove}.
3007 Dump after RTL generation, to @file{@var{file}.00.rtl}.
3010 Dump after the second scheduling pass, to @file{@var{file}.27.sched2}.
3013 Dump after CSE (including the jump optimization that sometimes follows
3014 CSE), to @file{@var{file}.08.cse}.
3017 Dump after the first scheduling pass, to @file{@var{file}.19.sched}.
3020 Dump after the second CSE pass (including the jump optimization that
3021 sometimes follows CSE), to @file{@var{file}.12.cse2}.
3024 Dump after the second flow pass, to @file{@var{file}.23.flow2}.
3027 Dump after SSA dead code elimination, to @file{@var{file}.06.ssadce}.
3030 Dump after the peephole pass, to @file{@var{file}.24.peephole2}.
3033 Produce all the dumps listed above.
3036 Print statistics on memory usage, at the end of the run, to
3040 Annotate the assembler output with a comment indicating which
3041 pattern and alternative was used. The length of each instruction is
3045 Dump the RTL in the assembler output as a comment before each instruction.
3046 Also turns on @option{-dp} annotation.
3049 For each of the other indicated dump files (except for
3050 @file{@var{file}.00.rtl}), dump a representation of the control flow graph
3051 suitable for viewing with VCG to @file{@var{file}.@var{pass}.vcg}.
3054 Just generate RTL for a function instead of compiling it. Usually used
3058 Dump debugging information during parsing, to standard error.
3061 @item -fdump-unnumbered
3062 @opindex fdump-unnumbered
3063 When doing debugging dumps (see @option{-d} option above), suppress instruction
3064 numbers and line number note output. This makes it more feasible to
3065 use diff on debugging dumps for compiler invocations with different
3066 options, in particular with and without @option{-g}.
3068 @item -fdump-translation-unit @r{(C and C++ only)}
3069 @itemx -fdump-translation-unit-@var{options} @r{(C and C++ only)}
3070 @opindex fdump-translation-unit
3071 Dump a representation of the tree structure for the entire translation
3072 unit to a file. The file name is made by appending @file{.tu} to the
3073 source file name. If the @samp{-@var{options}} form is used, @var{options}
3074 controls the details of the dump as described for the
3075 @option{-fdump-tree} options.
3077 @item -fdump-class-hierarchy @r{(C++ only)}
3078 @itemx -fdump-class-hierarchy-@var{options} @r{(C++ only)}
3079 @opindex fdump-class-hierarchy
3080 Dump a representation of each class's hierarchy and virtual function
3081 table layout to a file. The file name is made by appending @file{.class}
3082 to the source file name. If the @samp{-@var{options}} form is used,
3083 @var{options} controls the details of the dump as described for the
3084 @option{-fdump-tree} options.
3086 @item -fdump-tree-@var{switch} @r{(C++ only)}
3087 @itemx -fdump-tree-@var{switch}-@var{options} @r{(C++ only)}
3089 Control the dumping at various stages of processing the intermediate
3090 language tree to a file. The file name is generated by appending a switch
3091 specific suffix to the source file name. If the @samp{-@var{options}}
3092 form is used, @var{options} is a list of @samp{-} separated options that
3093 control the details of the dump. Not all options are applicable to all
3094 dumps, those which are not meaningful will be ignored. The following
3095 options are available
3099 Print the address of each node. Usually this is not meaningful as it
3100 changes according to the environment and source file. Its primary use
3101 is for tying up a dump file with a debug environment.
3103 Inhibit dumping of members of a scope or body of a function merely
3104 because that scope has been reached. Only dump such items when they
3105 are directly reachable by some other path.
3107 Turn on all options.
3110 The following tree dumps are possible:
3113 Dump before any tree based optimization, to @file{@var{file}.original}.
3115 Dump after all tree based optimization, to @file{@var{file}.optimized}.
3117 Dump after function inlining, to @file{@var{file}.inlined}.
3120 @item -fpretend-float
3121 @opindex fpretend-float
3122 When running a cross-compiler, pretend that the target machine uses the
3123 same floating point format as the host machine. This causes incorrect
3124 output of the actual floating constants, but the actual instruction
3125 sequence will probably be the same as GCC would make when running on
3130 Store the usual ``temporary'' intermediate files permanently; place them
3131 in the current directory and name them based on the source file. Thus,
3132 compiling @file{foo.c} with @samp{-c -save-temps} would produce files
3133 @file{foo.i} and @file{foo.s}, as well as @file{foo.o}. This creates a
3134 preprocessed @file{foo.i} output file even though the compiler now
3135 normally uses an integrated preprocessor.
3139 Report the CPU time taken by each subprocess in the compilation
3140 sequence. For C source files, this is the compiler proper and assembler
3141 (plus the linker if linking is done). The output looks like this:
3148 The first number on each line is the ``user time,'' that is time spent
3149 executing the program itself. The second number is ``system time,''
3150 time spent executing operating system routines on behalf of the program.
3151 Both numbers are in seconds.
3153 @item -print-file-name=@var{library}
3154 @opindex print-file-name
3155 Print the full absolute name of the library file @var{library} that
3156 would be used when linking---and don't do anything else. With this
3157 option, GCC does not compile or link anything; it just prints the
3160 @item -print-multi-directory
3161 @opindex print-multi-directory
3162 Print the directory name corresponding to the multilib selected by any
3163 other switches present in the command line. This directory is supposed
3164 to exist in @env{GCC_EXEC_PREFIX}.
3166 @item -print-multi-lib
3167 @opindex print-multi-lib
3168 Print the mapping from multilib directory names to compiler switches
3169 that enable them. The directory name is separated from the switches by
3170 @samp{;}, and each switch starts with an @samp{@@} instead of the
3171 @samp{-}, without spaces between multiple switches. This is supposed to
3172 ease shell-processing.
3174 @item -print-prog-name=@var{program}
3175 @opindex print-prog-name
3176 Like @option{-print-file-name}, but searches for a program such as @samp{cpp}.
3178 @item -print-libgcc-file-name
3179 @opindex print-libgcc-file-name
3180 Same as @option{-print-file-name=libgcc.a}.
3182 This is useful when you use @option{-nostdlib} or @option{-nodefaultlibs}
3183 but you do want to link with @file{libgcc.a}. You can do
3186 gcc -nostdlib @var{files}@dots{} `gcc -print-libgcc-file-name`
3189 @item -print-search-dirs
3190 @opindex print-search-dirs
3191 Print the name of the configured installation directory and a list of
3192 program and library directories gcc will search---and don't do anything else.
3194 This is useful when gcc prints the error message
3195 @samp{installation problem, cannot exec cpp0: No such file or directory}.
3196 To resolve this you either need to put @file{cpp0} and the other compiler
3197 components where gcc expects to find them, or you can set the environment
3198 variable @env{GCC_EXEC_PREFIX} to the directory where you installed them.
3199 Don't forget the trailing '/'.
3200 @xref{Environment Variables}.
3203 @opindex dumpmachine
3204 Print the compiler's target machine (for example,
3205 @samp{i686-pc-linux-gnu})---and don't do anything else.
3208 @opindex dumpversion
3209 Print the compiler version (for example, @samp{3.0})---and don't do
3214 Print the compiler's built-in specs---and don't do anything else. (This
3215 is used when GCC itself is being built.) @xref{Spec Files}.
3218 @node Optimize Options
3219 @section Options That Control Optimization
3220 @cindex optimize options
3221 @cindex options, optimization
3223 These options control various sorts of optimizations:
3230 Optimize. Optimizing compilation takes somewhat more time, and a lot
3231 more memory for a large function.
3233 Without @option{-O}, the compiler's goal is to reduce the cost of
3234 compilation and to make debugging produce the expected results.
3235 Statements are independent: if you stop the program with a breakpoint
3236 between statements, you can then assign a new value to any variable or
3237 change the program counter to any other statement in the function and
3238 get exactly the results you would expect from the source code.
3240 With @option{-O}, the compiler tries to reduce code size and execution
3241 time, without performing any optimizations that take a great deal of
3246 Optimize even more. GCC performs nearly all supported optimizations
3247 that do not involve a space-speed tradeoff. The compiler does not
3248 perform loop unrolling or function inlining when you specify @option{-O2}.
3249 As compared to @option{-O}, this option increases both compilation time
3250 and the performance of the generated code.
3252 @option{-O2} turns on all optional optimizations except for loop unrolling,
3253 function inlining, and register renaming. It also turns on the
3254 @option{-fforce-mem} option on all machines and frame pointer elimination
3255 on machines where doing so does not interfere with debugging.
3257 Please note the warning under @option{-fgcse} about
3258 invoking @option{-O2} on programs that use computed gotos.
3262 Optimize yet more. @option{-O3} turns on all optimizations specified by
3263 @option{-O2} and also turns on the @option{-finline-functions} and
3264 @option{-frename-registers} options.
3272 Optimize for size. @option{-Os} enables all @option{-O2} optimizations that
3273 do not typically increase code size. It also performs further
3274 optimizations designed to reduce code size.
3276 If you use multiple @option{-O} options, with or without level numbers,
3277 the last such option is the one that is effective.
3280 Options of the form @option{-f@var{flag}} specify machine-independent
3281 flags. Most flags have both positive and negative forms; the negative
3282 form of @option{-ffoo} would be @option{-fno-foo}. In the table below,
3283 only one of the forms is listed---the one which is not the default.
3284 You can figure out the other form by either removing @samp{no-} or
3289 @opindex ffloat-store
3290 Do not store floating point variables in registers, and inhibit other
3291 options that might change whether a floating point value is taken from a
3294 @cindex floating point precision
3295 This option prevents undesirable excess precision on machines such as
3296 the 68000 where the floating registers (of the 68881) keep more
3297 precision than a @code{double} is supposed to have. Similarly for the
3298 x86 architecture. For most programs, the excess precision does only
3299 good, but a few programs rely on the precise definition of IEEE floating
3300 point. Use @option{-ffloat-store} for such programs, after modifying
3301 them to store all pertinent intermediate computations into variables.
3303 @item -fno-default-inline
3304 @opindex fno-default-inline
3305 Do not make member functions inline by default merely because they are
3306 defined inside the class scope (C++ only). Otherwise, when you specify
3307 @w{@option{-O}}, member functions defined inside class scope are compiled
3308 inline by default; i.e., you don't need to add @samp{inline} in front of
3309 the member function name.
3311 @item -fno-defer-pop
3312 @opindex fno-defer-pop
3313 Always pop the arguments to each function call as soon as that function
3314 returns. For machines which must pop arguments after a function call,
3315 the compiler normally lets arguments accumulate on the stack for several
3316 function calls and pops them all at once.
3320 Force memory operands to be copied into registers before doing
3321 arithmetic on them. This produces better code by making all memory
3322 references potential common subexpressions. When they are not common
3323 subexpressions, instruction combination should eliminate the separate
3324 register-load. The @option{-O2} option turns on this option.
3327 @opindex fforce-addr
3328 Force memory address constants to be copied into registers before
3329 doing arithmetic on them. This may produce better code just as
3330 @option{-fforce-mem} may.
3332 @item -fomit-frame-pointer
3333 @opindex fomit-frame-pointer
3334 Don't keep the frame pointer in a register for functions that
3335 don't need one. This avoids the instructions to save, set up and
3336 restore frame pointers; it also makes an extra register available
3337 in many functions. @strong{It also makes debugging impossible on
3340 On some machines, such as the VAX, this flag has no effect, because
3341 the standard calling sequence automatically handles the frame pointer
3342 and nothing is saved by pretending it doesn't exist. The
3343 machine-description macro @code{FRAME_POINTER_REQUIRED} controls
3344 whether a target machine supports this flag. @xref{Registers,,Register
3345 Usage, gccint, GNU Compiler Collection (GCC) Internals}.
3347 @item -foptimize-sibling-calls
3348 @opindex foptimize-sibling-calls
3349 Optimize sibling and tail recursive calls.
3353 This option generates traps for signed overflow on addition, subtraction,
3354 multiplication operations.
3358 Don't pay attention to the @code{inline} keyword. Normally this option
3359 is used to keep the compiler from expanding any functions inline.
3360 Note that if you are not optimizing, no functions can be expanded inline.
3362 @item -finline-functions
3363 @opindex finline-functions
3364 Integrate all simple functions into their callers. The compiler
3365 heuristically decides which functions are simple enough to be worth
3366 integrating in this way.
3368 If all calls to a given function are integrated, and the function is
3369 declared @code{static}, then the function is normally not output as
3370 assembler code in its own right.
3372 @item -finline-limit=@var{n}
3373 @opindex finline-limit
3374 By default, gcc limits the size of functions that can be inlined. This flag
3375 allows the control of this limit for functions that are explicitly marked as
3376 inline (ie marked with the inline keyword or defined within the class
3377 definition in c++). @var{n} is the size of functions that can be inlined in
3378 number of pseudo instructions (not counting parameter handling). The default
3379 value of @var{n} is 600.
3380 Increasing this value can result in more inlined code at
3381 the cost of compilation time and memory consumption. Decreasing usually makes
3382 the compilation faster and less code will be inlined (which presumably
3383 means slower programs). This option is particularly useful for programs that
3384 use inlining heavily such as those based on recursive templates with C++.
3386 @emph{Note:} pseudo instruction represents, in this particular context, an
3387 abstract measurement of function's size. In no way, it represents a count
3388 of assembly instructions and as such its exact meaning might change from one
3389 release to an another.
3391 @item -fkeep-inline-functions
3392 @opindex fkeep-inline-functions
3393 Even if all calls to a given function are integrated, and the function
3394 is declared @code{static}, nevertheless output a separate run-time
3395 callable version of the function. This switch does not affect
3396 @code{extern inline} functions.
3398 @item -fkeep-static-consts
3399 @opindex fkeep-static-consts
3400 Emit variables declared @code{static const} when optimization isn't turned
3401 on, even if the variables aren't referenced.
3403 GCC enables this option by default. If you want to force the compiler to
3404 check if the variable was referenced, regardless of whether or not
3405 optimization is turned on, use the @option{-fno-keep-static-consts} option.
3407 @item -fmerge-constants
3408 Attempt to merge identical constants (string constants and floating point
3409 constants) accross compilation units.
3411 This option is default for optimized compilation if assembler and linker
3412 support it. Use @option{-fno-merge-constants} to inhibit this behavior.
3414 @item -fmerge-all-constants
3415 Attempt to merge identical constants and identical variables.
3417 This option implies @option{-fmerge-constants}. In addition to
3418 @option{-fmerge-constants} this considers e.g. even constant initialized
3419 arrays or initialized constant variables with integral or floating point
3420 types. Languages like C or C++ require each non-automatic variable to
3421 have distinct location, so using this option will result in non-conforming
3424 @item -fno-function-cse
3425 @opindex fno-function-cse
3426 Do not put function addresses in registers; make each instruction that
3427 calls a constant function contain the function's address explicitly.
3429 This option results in less efficient code, but some strange hacks
3430 that alter the assembler output may be confused by the optimizations
3431 performed when this option is not used.
3435 Sets @option{-fno-math-errno}, @option{-funsafe-math-optimizations}, and @*
3436 @option{-fno-trapping-math}.
3438 This option causes the preprocessor macro @code{__FAST_MATH__} to be defined.
3440 This option should never be turned on by any @option{-O} option since
3441 it can result in incorrect output for programs which depend on
3442 an exact implementation of IEEE or ISO rules/specifications for
3445 @item -fno-math-errno
3446 @opindex fno-math-errno
3447 Do not set ERRNO after calling math functions that are executed
3448 with a single instruction, e.g., sqrt. A program that relies on
3449 IEEE exceptions for math error handling may want to use this flag
3450 for speed while maintaining IEEE arithmetic compatibility.
3452 This option should never be turned on by any @option{-O} option since
3453 it can result in incorrect output for programs which depend on
3454 an exact implementation of IEEE or ISO rules/specifications for
3457 The default is @option{-fmath-errno}.
3459 @item -funsafe-math-optimizations
3460 @opindex funsafe-math-optimizations
3461 Allow optimizations for floating-point arithmetic that (a) assume
3462 that arguments and results are valid and (b) may violate IEEE or
3463 ANSI standards. When used at link-time, it may include libraries
3464 or startup files that change the default FPU control word or other
3465 similar optimizations.
3467 This option should never be turned on by any @option{-O} option since
3468 it can result in incorrect output for programs which depend on
3469 an exact implementation of IEEE or ISO rules/specifications for
3472 The default is @option{-fno-unsafe-math-optimizations}.
3474 @item -fno-trapping-math
3475 @opindex fno-trapping-math
3476 Compile code assuming that floating-point operations cannot generate
3477 user-visible traps. Setting this option may allow faster code
3478 if one relies on ``non-stop'' IEEE arithmetic, for example.
3480 This option should never be turned on by any @option{-O} option since
3481 it can result in incorrect output for programs which depend on
3482 an exact implementation of IEEE or ISO rules/specifications for
3485 The default is @option{-ftrapping-math}.
3488 The following options control specific optimizations. The @option{-O2}
3489 option turns on all of these optimizations except @option{-funroll-loops}
3490 and @option{-funroll-all-loops}. On most machines, the @option{-O} option
3491 turns on the @option{-fthread-jumps} and @option{-fdelayed-branch} options,
3492 but specific machines may handle it differently.
3494 You can use the following flags in the rare cases when ``fine-tuning''
3495 of optimizations to be performed is desired.
3497 Not all of the optimizations performed by GCC have @option{-f} options
3501 @item -fstrength-reduce
3502 @opindex fstrength-reduce
3503 Perform the optimizations of loop strength reduction and
3504 elimination of iteration variables.
3506 @item -fthread-jumps
3507 @opindex fthread-jumps
3508 Perform optimizations where we check to see if a jump branches to a
3509 location where another comparison subsumed by the first is found. If
3510 so, the first branch is redirected to either the destination of the
3511 second branch or a point immediately following it, depending on whether
3512 the condition is known to be true or false.
3514 @item -fcse-follow-jumps
3515 @opindex fcse-follow-jumps
3516 In common subexpression elimination, scan through jump instructions
3517 when the target of the jump is not reached by any other path. For
3518 example, when CSE encounters an @code{if} statement with an
3519 @code{else} clause, CSE will follow the jump when the condition
3522 @item -fcse-skip-blocks
3523 @opindex fcse-skip-blocks
3524 This is similar to @option{-fcse-follow-jumps}, but causes CSE to
3525 follow jumps which conditionally skip over blocks. When CSE
3526 encounters a simple @code{if} statement with no else clause,
3527 @option{-fcse-skip-blocks} causes CSE to follow the jump around the
3528 body of the @code{if}.
3530 @item -frerun-cse-after-loop
3531 @opindex frerun-cse-after-loop
3532 Re-run common subexpression elimination after loop optimizations has been
3535 @item -frerun-loop-opt
3536 @opindex frerun-loop-opt
3537 Run the loop optimizer twice.
3541 Perform a global common subexpression elimination pass.
3542 This pass also performs global constant and copy propagation.
3544 @emph{Note:} When compiling a program using computed gotos, a GCC
3545 extension, you may get better runtime performance if you disable
3546 the global common subexpression elmination pass by adding
3547 @option{-fno-gcse} to the command line.
3551 When @option{-fgcse-lm} is enabled, global common subexpression elimination will
3552 attempt to move loads which are only killed by stores into themselves. This
3553 allows a loop containing a load/store sequence to be changed to a load outside
3554 the loop, and a copy/store within the loop.
3558 When @option{-fgcse-sm} is enabled, A store motion pass is run after global common
3559 subexpression elimination. This pass will attempt to move stores out of loops.
3560 When used in conjunction with @option{-fgcse-lm}, loops containing a load/store sequence
3561 can be changed to a load before the loop and a store after the loop.
3563 @item -fdelete-null-pointer-checks
3564 @opindex fdelete-null-pointer-checks
3565 Use global dataflow analysis to identify and eliminate useless checks
3566 for null pointers. The compiler assumes that dereferencing a null
3567 pointer would have halted the program. If a pointer is checked after
3568 it has already been dereferenced, it cannot be null.
3570 In some environments, this assumption is not true, and programs can
3571 safely dereference null pointers. Use
3572 @option{-fno-delete-null-pointer-checks} to disable this optimization
3573 for programs which depend on that behavior.
3575 @item -fexpensive-optimizations
3576 @opindex fexpensive-optimizations
3577 Perform a number of minor optimizations that are relatively expensive.
3579 @item -foptimize-register-move
3581 @opindex foptimize-register-move
3583 Attempt to reassign register numbers in move instructions and as
3584 operands of other simple instructions in order to maximize the amount of
3585 register tying. This is especially helpful on machines with two-operand
3586 instructions. GCC enables this optimization by default with @option{-O2}
3589 Note @option{-fregmove} and @option{-foptimize-register-move} are the same
3592 @item -fdelayed-branch
3593 @opindex fdelayed-branch
3594 If supported for the target machine, attempt to reorder instructions
3595 to exploit instruction slots available after delayed branch
3598 @item -fschedule-insns
3599 @opindex fschedule-insns
3600 If supported for the target machine, attempt to reorder instructions to
3601 eliminate execution stalls due to required data being unavailable. This
3602 helps machines that have slow floating point or memory load instructions
3603 by allowing other instructions to be issued until the result of the load
3604 or floating point instruction is required.
3606 @item -fschedule-insns2
3607 @opindex fschedule-insns2
3608 Similar to @option{-fschedule-insns}, but requests an additional pass of
3609 instruction scheduling after register allocation has been done. This is
3610 especially useful on machines with a relatively small number of
3611 registers and where memory load instructions take more than one cycle.
3613 @item -ffunction-sections
3614 @itemx -fdata-sections
3615 @opindex ffunction-sections
3616 @opindex fdata-sections
3617 Place each function or data item into its own section in the output
3618 file if the target supports arbitrary sections. The name of the
3619 function or the name of the data item determines the section's name
3622 Use these options on systems where the linker can perform optimizations
3623 to improve locality of reference in the instruction space. HPPA
3624 processors running HP-UX and Sparc processors running Solaris 2 have
3625 linkers with such optimizations. Other systems using the ELF object format
3626 as well as AIX may have these optimizations in the future.
3628 Only use these options when there are significant benefits from doing
3629 so. When you specify these options, the assembler and linker will
3630 create larger object and executable files and will also be slower.
3631 You will not be able to use @code{gprof} on all systems if you
3632 specify this option and you may have problems with debugging if
3633 you specify both this option and @option{-g}.
3635 @item -fcaller-saves
3636 @opindex fcaller-saves
3637 Enable values to be allocated in registers that will be clobbered by
3638 function calls, by emitting extra instructions to save and restore the
3639 registers around such calls. Such allocation is done only when it
3640 seems to result in better code than would otherwise be produced.
3642 This option is always enabled by default on certain machines, usually
3643 those which have no call-preserved registers to use instead.
3645 For all machines, optimization level 2 and higher enables this flag by
3648 @item -funroll-loops
3649 @opindex funroll-loops
3650 Unroll loops whose number of iterations can be determined at compile
3651 time or upon entry to the loop. @option{-funroll-loops} implies both
3652 @option{-fstrength-reduce} and @option{-frerun-cse-after-loop}. This
3653 option makes code larger, and may or may not make it run faster.
3655 @item -funroll-all-loops
3656 @opindex funroll-all-loops
3657 Unroll all loops, even if their number of iterations is uncertain when
3658 the loop is entered. This usually makes programs run more slowly.
3659 @option{-funroll-all-loops} implies the same options as
3660 @option{-funroll-loops},
3662 @item -fprefetch-loop-arrays
3663 @opindex fprefetch-loop-arrays
3664 If supported by the target machine, generate instructions to prefetch
3665 memory to improve the performance of loops that access large arrays.
3667 @item -fmove-all-movables
3668 @opindex fmove-all-movables
3669 Forces all invariant computations in loops to be moved
3672 @item -freduce-all-givs
3673 @opindex freduce-all-givs
3674 Forces all general-induction variables in loops to be
3677 @emph{Note:} When compiling programs written in Fortran,
3678 @option{-fmove-all-movables} and @option{-freduce-all-givs} are enabled
3679 by default when you use the optimizer.
3681 These options may generate better or worse code; results are highly
3682 dependent on the structure of loops within the source code.
3684 These two options are intended to be removed someday, once
3685 they have helped determine the efficacy of various
3686 approaches to improving loop optimizations.
3688 Please let us (@w{@email{gcc@@gcc.gnu.org}} and @w{@email{fortran@@gnu.org}})
3689 know how use of these options affects
3690 the performance of your production code.
3691 We're very interested in code that runs @emph{slower}
3692 when these options are @emph{enabled}.
3695 @itemx -fno-peephole2
3696 @opindex fno-peephole
3697 @opindex fno-peephole2
3698 Disable any machine-specific peephole optimizations. The difference
3699 between @option{-fno-peephole} and @option{-fno-peephole2} is in how they
3700 are implemented in the compiler; some targets use one, some use the
3701 other, a few use both.
3703 @item -fbranch-probabilities
3704 @opindex fbranch-probabilities
3705 After running a program compiled with @option{-fprofile-arcs}
3706 (@pxref{Debugging Options,, Options for Debugging Your Program or
3707 @command{gcc}}), you can compile it a second time using
3708 @option{-fbranch-probabilities}, to improve optimizations based on
3709 the number of times each branch was taken. When the program
3710 compiled with @option{-fprofile-arcs} exits it saves arc execution
3711 counts to a file called @file{@var{sourcename}.da} for each source
3712 file The information in this data file is very dependent on the
3713 structure of the generated code, so you must use the same source code
3714 and the same optimization options for both compilations.
3716 With @option{-fbranch-probabilities}, GCC puts a @samp{REG_EXEC_COUNT}
3717 note on the first instruction of each basic block, and a
3718 @samp{REG_BR_PROB} note on each @samp{JUMP_INSN} and @samp{CALL_INSN}.
3719 These can be used to improve optimization. Currently, they are only
3720 used in one place: in @file{reorg.c}, instead of guessing which path a
3721 branch is mostly to take, the @samp{REG_BR_PROB} values are used to
3722 exactly determine which path is taken more often.
3724 @item -fno-guess-branch-probability
3725 @opindex fno-guess-branch-probability
3726 Do not guess branch probabilities using a randomized model.
3728 Sometimes gcc will opt to use a randomized model to guess branch
3729 probabilities, when none are available from either profiling feedback
3730 (@option{-fprofile-arcs}) or @samp{__builtin_expect}. This means that
3731 different runs of the compiler on the same program may produce different
3734 In a hard real-time system, people don't want different runs of the
3735 compiler to produce code that has different behavior; minimizing
3736 non-determinism is of paramount import. This switch allows users to
3737 reduce non-determinism, possibly at the expense of inferior
3740 @item -fstrict-aliasing
3741 @opindex fstrict-aliasing
3742 Allows the compiler to assume the strictest aliasing rules applicable to
3743 the language being compiled. For C (and C++), this activates
3744 optimizations based on the type of expressions. In particular, an
3745 object of one type is assumed never to reside at the same address as an
3746 object of a different type, unless the types are almost the same. For
3747 example, an @code{unsigned int} can alias an @code{int}, but not a
3748 @code{void*} or a @code{double}. A character type may alias any other
3751 Pay special attention to code like this:
3764 The practice of reading from a different union member than the one most
3765 recently written to (called ``type-punning'') is common. Even with
3766 @option{-fstrict-aliasing}, type-punning is allowed, provided the memory
3767 is accessed through the union type. So, the code above will work as
3768 expected. However, this code might not:
3779 Every language that wishes to perform language-specific alias analysis
3780 should define a function that computes, given an @code{tree}
3781 node, an alias set for the node. Nodes in different alias sets are not
3782 allowed to alias. For an example, see the C front-end function
3783 @code{c_get_alias_set}.
3785 @item -falign-functions
3786 @itemx -falign-functions=@var{n}
3787 @opindex falign-functions
3788 Align the start of functions to the next power-of-two greater than
3789 @var{n}, skipping up to @var{n} bytes. For instance,
3790 @option{-falign-functions=32} aligns functions to the next 32-byte
3791 boundary, but @option{-falign-functions=24} would align to the next
3792 32-byte boundary only if this can be done by skipping 23 bytes or less.
3794 @option{-fno-align-functions} and @option{-falign-functions=1} are
3795 equivalent and mean that functions will not be aligned.
3797 Some assemblers only support this flag when @var{n} is a power of two;
3798 in that case, it is rounded up.
3800 If @var{n} is not specified, use a machine-dependent default.
3802 @item -falign-labels
3803 @itemx -falign-labels=@var{n}
3804 @opindex falign-labels
3805 Align all branch targets to a power-of-two boundary, skipping up to
3806 @var{n} bytes like @option{-falign-functions}. This option can easily
3807 make code slower, because it must insert dummy operations for when the
3808 branch target is reached in the usual flow of the code.
3810 If @option{-falign-loops} or @option{-falign-jumps} are applicable and
3811 are greater than this value, then their values are used instead.
3813 If @var{n} is not specified, use a machine-dependent default which is
3814 very likely to be @samp{1}, meaning no alignment.
3817 @itemx -falign-loops=@var{n}
3818 @opindex falign-loops
3819 Align loops to a power-of-two boundary, skipping up to @var{n} bytes
3820 like @option{-falign-functions}. The hope is that the loop will be
3821 executed many times, which will make up for any execution of the dummy
3824 If @var{n} is not specified, use a machine-dependent default.
3827 @itemx -falign-jumps=@var{n}
3828 @opindex falign-jumps
3829 Align branch targets to a power-of-two boundary, for branch targets
3830 where the targets can only be reached by jumping, skipping up to @var{n}
3831 bytes like @option{-falign-functions}. In this case, no dummy operations
3834 If @var{n} is not specified, use a machine-dependent default.
3838 Perform optimizations in static single assignment form. Each function's
3839 flow graph is translated into SSA form, optimizations are performed, and
3840 the flow graph is translated back from SSA form. Users should not
3841 specify this option, since it is not yet ready for production use.
3845 Perform Sparse Conditional Constant Propagation in SSA form. Requires
3846 @option{-fssa}. Like @option{-fssa}, this is an experimental feature.
3850 Perform aggressive dead-code elimination in SSA form. Requires @option{-fssa}.
3851 Like @option{-fssa}, this is an experimental feature.
3853 @item -fsingle-precision-constant
3854 @opindex fsingle-precision-constant
3855 Treat floating point constant as single precision constant instead of
3856 implicitly converting it to double precision constant.
3858 @item -frename-registers
3859 @opindex frename-registers
3860 Attempt to avoid false dependencies in scheduled code by making use
3861 of registers left over after register allocation. This optimization
3862 will most benefit processors with lots of registers. It can, however,
3863 make debugging impossible, since variables will no longer stay in
3864 a ``home register''.
3866 @item -fno-cprop-registers
3867 @opindex fno-cprop-registers
3868 After register allocation and post-register allocation instruction splitting,
3869 we perform a copy-propagation pass to try to reduce scheduling dependencies
3870 and occasionally eliminate the copy.
3872 @item --param @var{name}=@var{value}
3874 In some places, GCC uses various constants to control the amount of
3875 optimization that is done. For example, GCC will not inline functions
3876 that contain more that a certain number of instructions. You can
3877 control some of these constants on the command-line using the
3878 @option{--param} option.
3880 In each case, the @var{value} is an integer. The allowable choices for
3881 @var{name} are given in the following table:
3884 @item max-delay-slot-insn-search
3885 The maximum number of instructions to consider when looking for an
3886 instruction to fill a delay slot. If more than this arbitrary number of
3887 instructions is searched, the time savings from filling the delay slot
3888 will be minimal so stop searching. Increasing values mean more
3889 aggressive optimization, making the compile time increase with probably
3890 small improvement in executable run time.
3892 @item max-delay-slot-live-search
3893 When trying to fill delay slots, the maximum number of instructions to
3894 consider when searching for a block with valid live register
3895 information. Increasing this arbitrarily chosen value means more
3896 aggressive optimization, increasing the compile time. This parameter
3897 should be removed when the delay slot code is rewritten to maintain the
3900 @item max-gcse-memory
3901 The approximate maximum amount of memory that will be allocated in
3902 order to perform the global common subexpression elimination
3903 optimization. If more memory than specified is required, the
3904 optimization will not be done.
3906 @item max-gcse-passes
3907 The maximum number of passes of GCSE to run.
3909 @item max-pending-list-length
3910 The maximum number of pending dependencies scheduling will allow
3911 before flushing the current state and starting over. Large functions
3912 with few branches or calls can create excessively large lists which
3913 needlessly consume memory and resources.
3915 @item max-inline-insns
3916 If an function contains more than this many instructions, it
3917 will not be inlined. This option is precisely equivalent to
3918 @option{-finline-limit}.
3923 @node Preprocessor Options
3924 @section Options Controlling the Preprocessor
3925 @cindex preprocessor options
3926 @cindex options, preprocessor
3928 These options control the C preprocessor, which is run on each C source
3929 file before actual compilation.
3931 If you use the @option{-E} option, nothing is done except preprocessing.
3932 Some of these options make sense only together with @option{-E} because
3933 they cause the preprocessor output to be unsuitable for actual
3937 @item -include @var{file}
3939 Process @var{file} as input before processing the regular input file.
3940 In effect, the contents of @var{file} are compiled first. Any @option{-D}
3941 and @option{-U} options on the command line are always processed before
3942 @option{-include @var{file}}, regardless of the order in which they are
3943 written. All the @option{-include} and @option{-imacros} options are
3944 processed in the order in which they are written.
3946 @item -imacros @var{file}
3948 Process @var{file} as input, discarding the resulting output, before
3949 processing the regular input file. Because the output generated from
3950 @var{file} is discarded, the only effect of @option{-imacros @var{file}}
3951 is to make the macros defined in @var{file} available for use in the
3952 main input. All the @option{-include} and @option{-imacros} options are
3953 processed in the order in which they are written.
3955 @item -idirafter @var{dir}
3957 @cindex second include path
3958 Add the directory @var{dir} to the second include path. The directories
3959 on the second include path are searched when a header file is not found
3960 in any of the directories in the main include path (the one that
3961 @option{-I} adds to).
3963 @item -iprefix @var{prefix}
3965 Specify @var{prefix} as the prefix for subsequent @option{-iwithprefix}
3968 @item -iwithprefix @var{dir}
3969 @opindex iwithprefix
3970 Add a directory to the second include path. The directory's name is
3971 made by concatenating @var{prefix} and @var{dir}, where @var{prefix} was
3972 specified previously with @option{-iprefix}. If you have not specified a
3973 prefix yet, the directory containing the installed passes of the
3974 compiler is used as the default.
3976 @item -iwithprefixbefore @var{dir}
3977 @opindex iwithprefixbefore
3978 Add a directory to the main include path. The directory's name is made
3979 by concatenating @var{prefix} and @var{dir}, as in the case of
3980 @option{-iwithprefix}.
3982 @item -isystem @var{dir}
3984 Add a directory to the beginning of the second include path, marking it
3985 as a system directory, so that it gets the same special treatment as
3986 is applied to the standard system directories.
3990 Do not search the standard system directories for header files. Only
3991 the directories you have specified with @option{-I} options (and the
3992 current directory, if appropriate) are searched. @xref{Directory
3993 Options}, for information on @option{-I}.
3995 By using both @option{-nostdinc} and @option{-I-}, you can limit the include-file
3996 search path to only those directories you specify explicitly.
4000 When searching for a header file in a directory, remap file names if a
4001 file named @file{header.gcc} exists in that directory. This can be used
4002 to work around limitations of file systems with file name restrictions.
4003 The @file{header.gcc} file should contain a series of lines with two
4004 tokens on each line: the first token is the name to map, and the second
4005 token is the actual name to use.
4009 Do not predefine any nonstandard macros. (Including architecture flags).
4013 Run only the C preprocessor. Preprocess all the C source files
4014 specified and output the results to standard output or to the
4015 specified output file.
4019 Tell the preprocessor not to discard comments. Used with the
4024 Tell the preprocessor not to generate @samp{#line} directives.
4025 Used with the @option{-E} option.
4028 @cindex dependencies, make
4031 Instead of outputting the result of preprocessing, output a rule
4032 suitable for @command{make} describing the dependencies of the main
4033 source file. The preprocessor outputs one @command{make} rule containing
4034 the object file name for that source file, a colon, and the names of all
4035 the included files, including those coming from @option{-include} or
4036 @option{-imacros} command line options.
4038 Unless specified explicitly (with @option{-MT} or @option{-MQ}), the
4039 object file name consists of the basename of the source file with any
4040 suffix replaced with object file suffix. If there are many included
4041 files then the rule is split into several lines using @samp{\}-newline.
4042 The rule has no commands.
4044 Passing @option{-M} to the driver implies @option{-E}.
4048 Like @option{-M} but do not mention header files that are found in
4049 system header directories, nor header files that are included,
4050 directly or indirectly, from such a header.
4052 This implies that the choice of angle brackets or double quotes in an
4053 @samp{#include} directive does not in itself determine whether that
4054 header will appear in @option{-MM} dependency output. This is a
4055 slight change in semantics from GCC versions 3.0 and earlier.
4059 @option{-MD} is equivalent to @option{-M -MF @var{file}}, except that
4060 @option{-E} is not implied. The driver determines @var{file} based on
4061 whether an @option{-o} option is given. If it is, the driver uses its
4062 argument but with a suffix of @file{.d}, otherwise it take the
4063 basename of the input file and applies a @file{.d} suffix.
4065 If @option{-MD} is used in conjunction with @option{-E}, any
4066 @option{-o} switch is understood to specify the dependency output file
4067 (but @pxref{-MF}), but if used without @option{-E}, each @option{-o}
4068 is understood to specify a target object file.
4070 Since @option{-E} is not implied, @option{-MD} can be used to generate
4071 a dependency output file as a side-effect of the compilation process.
4073 With Mach, you can use the utility @code{md} to merge multiple
4074 dependency files into a single dependency file suitable for using with
4075 the @samp{make} command.
4079 Like @option{-MD} except mention only user header files, not system
4082 @item -MF @var{file}
4085 When used with @option{-M} or @option{-MM}, specifies a
4086 file to write the dependencies to. If no @option{-MF} switch is given
4087 the preprocessor sends the rules to the same place it would have sent
4088 preprocessed output.
4090 When used with the driver options @option{-MD} or @option{-MMD},
4091 @option{-MF} overrides the default dependency output file.
4093 Another way to specify output of a @code{make} rule is by setting
4094 the environment variable @env{DEPENDENCIES_OUTPUT} (@pxref{Environment
4099 When used with @option{-M} or @option{-MM}, @option{-MG} says to treat missing
4100 header files as generated files and assume they live in the same
4101 directory as the source file. It suppresses preprocessed output, as a
4102 missing header file is ordinarily an error.
4104 This feature is used in automatic updating of makefiles.
4108 This option instructs CPP to add a phony target for each dependency
4109 other than the main file, causing each to depend on nothing. These
4110 dummy rules work around errors @code{make} gives if you remove header
4111 files without updating the @code{Makefile} to match.
4113 This is typical output:-
4116 /tmp/test.o: /tmp/test.c /tmp/test.h
4121 @item -MQ @var{target}
4122 @item -MT @var{target}
4125 By default CPP uses the main file name, including any path, and appends
4126 the object suffix, normally ``.o'', to it to obtain the name of the
4127 target for dependency generation. With @option{-MT} you can specify a
4128 target yourself, overriding the default one.
4130 If you want multiple targets, you can specify them as a single argument
4131 to @option{-MT}, or use multiple @option{-MT} options.
4133 The targets you specify are output in the order they appear on the
4134 command line. @option{-MQ} is identical to @option{-MT}, except that the
4135 target name is quoted for Make, but with @option{-MT} it isn't. For
4136 example, @option{-MT '$(objpfx)foo.o'} gives
4139 $(objpfx)foo.o: /tmp/foo.c
4142 but @option{-MQ '$(objpfx)foo.o'} gives
4145 $$(objpfx)foo.o: /tmp/foo.c
4148 The default target is automatically quoted, as if it were given with
4153 Print the name of each header file used, in addition to other normal
4156 @item -A@var{question}(@var{answer})
4158 Assert the answer @var{answer} for @var{question}, in case it is tested
4159 with a preprocessing conditional such as @samp{#if
4160 #@var{question}(@var{answer})}. @option{-A-} disables the standard
4161 assertions that normally describe the target machine.
4165 Define macro @var{macro} with the string @samp{1} as its definition.
4167 @item -D@var{macro}=@var{defn}
4168 Define macro @var{macro} as @var{defn}. All instances of @option{-D} on
4169 the command line are processed before any @option{-U} options.
4171 Any @option{-D} and @option{-U} options on the command line are processed in
4172 order, and always before @option{-imacros @var{file}}, regardless of the
4173 order in which they are written.
4177 Undefine macro @var{macro}. @option{-U} options are evaluated after all
4178 @option{-D} options, but before any @option{-include} and @option{-imacros}
4181 Any @option{-D} and @option{-U} options on the command line are processed in
4182 order, and always before @option{-imacros @var{file}}, regardless of the
4183 order in which they are written.
4187 Tell the preprocessor to output only a list of the macro definitions
4188 that are in effect at the end of preprocessing. Used with the @option{-E}
4193 Tell the preprocessing to pass all macro definitions into the output, in
4194 their proper sequence in the rest of the output.
4198 Like @option{-dD} except that the macro arguments and contents are omitted.
4199 Only @samp{#define @var{name}} is included in the output.
4203 Output @samp{#include} directives in addition to the result of
4206 @item -fpreprocessed
4207 @opindex fpreprocessed
4208 Indicate to the preprocessor that the input file has already been
4209 preprocessed. This suppresses things like macro expansion, trigraph
4210 conversion, escaped newline splicing, and processing of most directives.
4211 The preprocessor still recognizes and removes comments, so that you can
4212 pass a file preprocessed with @option{-C} to the compiler without
4213 problems. In this mode the integrated preprocessor is little more than
4214 a tokenizer for the front ends.
4216 @option{-fpreprocessed} is implicit if the input file has one of the
4217 extensions @samp{i}, @samp{ii} or @samp{mi}. These are the extensions
4218 that GCC uses for preprocessed files created by @option{-save-temps}.
4222 Process ISO standard trigraph sequences. These are three-character
4223 sequences, all starting with @samp{??}, that are defined by ISO C to
4224 stand for single characters. For example, @samp{??/} stands for
4225 @samp{\}, so @samp{'??/n'} is a character constant for a newline. By
4226 default, GCC ignores trigraphs, but in standard-conforming modes it
4227 converts them. See the @option{-std} and @option{-ansi} options.
4229 The nine trigraph sequences are
4232 @expansion{} @samp{[}
4235 @expansion{} @samp{]}
4238 @expansion{} @samp{@{}
4241 @expansion{} @samp{@}}
4244 @expansion{} @samp{#}
4247 @expansion{} @samp{\}
4250 @expansion{} @samp{^}
4253 @expansion{} @samp{|}
4256 @expansion{} @samp{~}
4260 Trigraph support is not popular, so many compilers do not implement it
4261 properly. Portable code should not rely on trigraphs being either
4262 converted or ignored.
4264 @item -Wp,@var{option}
4266 Pass @var{option} as an option to the preprocessor. If @var{option}
4267 contains commas, it is split into multiple options at the commas.
4270 @node Assembler Options
4271 @section Passing Options to the Assembler
4273 @c prevent bad page break with this line
4274 You can pass options to the assembler.
4277 @item -Wa,@var{option}
4279 Pass @var{option} as an option to the assembler. If @var{option}
4280 contains commas, it is split into multiple options at the commas.
4284 @section Options for Linking
4285 @cindex link options
4286 @cindex options, linking
4288 These options come into play when the compiler links object files into
4289 an executable output file. They are meaningless if the compiler is
4290 not doing a link step.
4294 @item @var{object-file-name}
4295 A file name that does not end in a special recognized suffix is
4296 considered to name an object file or library. (Object files are
4297 distinguished from libraries by the linker according to the file
4298 contents.) If linking is done, these object files are used as input
4307 If any of these options is used, then the linker is not run, and
4308 object file names should not be used as arguments. @xref{Overall
4312 @item -l@var{library}
4313 @itemx -l @var{library}
4315 Search the library named @var{library} when linking. (The second
4316 alternative with the library as a separate argument is only for
4317 POSIX compliance and is not recommended.)
4319 It makes a difference where in the command you write this option; the
4320 linker searches and processes libraries and object files in the order they
4321 are specified. Thus, @samp{foo.o -lz bar.o} searches library @samp{z}
4322 after file @file{foo.o} but before @file{bar.o}. If @file{bar.o} refers
4323 to functions in @samp{z}, those functions may not be loaded.
4325 The linker searches a standard list of directories for the library,
4326 which is actually a file named @file{lib@var{library}.a}. The linker
4327 then uses this file as if it had been specified precisely by name.
4329 The directories searched include several standard system directories
4330 plus any that you specify with @option{-L}.
4332 Normally the files found this way are library files---archive files
4333 whose members are object files. The linker handles an archive file by
4334 scanning through it for members which define symbols that have so far
4335 been referenced but not defined. But if the file that is found is an
4336 ordinary object file, it is linked in the usual fashion. The only
4337 difference between using an @option{-l} option and specifying a file name
4338 is that @option{-l} surrounds @var{library} with @samp{lib} and @samp{.a}
4339 and searches several directories.
4343 You need this special case of the @option{-l} option in order to
4344 link an Objective-C program.
4347 @opindex nostartfiles
4348 Do not use the standard system startup files when linking.
4349 The standard system libraries are used normally, unless @option{-nostdlib}
4350 or @option{-nodefaultlibs} is used.
4352 @item -nodefaultlibs
4353 @opindex nodefaultlibs
4354 Do not use the standard system libraries when linking.
4355 Only the libraries you specify will be passed to the linker.
4356 The standard startup files are used normally, unless @option{-nostartfiles}
4357 is used. The compiler may generate calls to memcmp, memset, and memcpy
4358 for System V (and ISO C) environments or to bcopy and bzero for
4359 BSD environments. These entries are usually resolved by entries in
4360 libc. These entry points should be supplied through some other
4361 mechanism when this option is specified.
4365 Do not use the standard system startup files or libraries when linking.
4366 No startup files and only the libraries you specify will be passed to
4367 the linker. The compiler may generate calls to memcmp, memset, and memcpy
4368 for System V (and ISO C) environments or to bcopy and bzero for
4369 BSD environments. These entries are usually resolved by entries in
4370 libc. These entry points should be supplied through some other
4371 mechanism when this option is specified.
4373 @cindex @option{-lgcc}, use with @option{-nostdlib}
4374 @cindex @option{-nostdlib} and unresolved references
4375 @cindex unresolved references and @option{-nostdlib}
4376 @cindex @option{-lgcc}, use with @option{-nodefaultlibs}
4377 @cindex @option{-nodefaultlibs} and unresolved references
4378 @cindex unresolved references and @option{-nodefaultlibs}
4379 One of the standard libraries bypassed by @option{-nostdlib} and
4380 @option{-nodefaultlibs} is @file{libgcc.a}, a library of internal subroutines
4381 that GCC uses to overcome shortcomings of particular machines, or special
4382 needs for some languages.
4383 (@xref{Interface,,Interfacing to GCC Output,gccint,GNU Compiler
4384 Collection (GCC) Internals},
4385 for more discussion of @file{libgcc.a}.)
4386 In most cases, you need @file{libgcc.a} even when you want to avoid
4387 other standard libraries. In other words, when you specify @option{-nostdlib}
4388 or @option{-nodefaultlibs} you should usually specify @option{-lgcc} as well.
4389 This ensures that you have no unresolved references to internal GCC
4390 library subroutines. (For example, @samp{__main}, used to ensure C++
4391 constructors will be called; @pxref{Collect2,,@code{collect2}, gccint,
4392 GNU Compiler Collection (GCC) Internals}.)
4396 Remove all symbol table and relocation information from the executable.
4400 On systems that support dynamic linking, this prevents linking with the shared
4401 libraries. On other systems, this option has no effect.
4405 Produce a shared object which can then be linked with other objects to
4406 form an executable. Not all systems support this option. For predictable
4407 results, you must also specify the same set of options that were used to
4408 generate code (@option{-fpic}, @option{-fPIC}, or model suboptions)
4409 when you specify this option.@footnote{On some systems, @samp{gcc -shared}
4410 needs to build supplementary stub code for constructors to work. On
4411 multi-libbed systems, @samp{gcc -shared} must select the correct support
4412 libraries to link against. Failing to supply the correct flags may lead
4413 to subtle defects. Supplying them in cases where they are not necessary
4416 @item -shared-libgcc
4417 @itemx -static-libgcc
4418 @opindex shared-libgcc
4419 @opindex static-libgcc
4420 On systems that provide @file{libgcc} as a shared library, these options
4421 force the use of either the shared or static version respectively.
4422 If no shared version of @file{libgcc} was built when the compiler was
4423 configured, these options have no effect.
4425 There are several situations in which an application should use the
4426 shared @file{libgcc} instead of the static version. The most common
4427 of these is when the application wishes to throw and catch exceptions
4428 across different shared libraries. In that case, each of the libraries
4429 as well as the application itself should use the shared @file{libgcc}.
4431 Therefore, the G++ and GCJ drivers automatically add
4432 @option{-shared-libgcc} whenever you build a shared library or a main
4433 executable, because C++ and Java programs typically use exceptions, so
4434 this is the right thing to do.
4436 If, instead, you use the GCC driver to create shared libraries, you may
4437 find that they will not always be linked with the shared @file{libgcc}.
4438 If GCC finds, at its configuration time, that you have a GNU linker that
4439 does not support option @option{--eh-frame-hdr}, it will link the shared
4440 version of @file{libgcc} into shared libraries by default. Otherwise,
4441 it will take advantage of the linker and optimize away the linking with
4442 the shared version of @file{libgcc}, linking with the static version of
4443 libgcc by default. This allows exceptions to propagate through such
4444 shared libraries, without incurring relocation costs at library load
4447 However, if a library or main executable is supposed to throw or catch
4448 exceptions, you must link it using the G++ or GCJ driver, as appropriate
4449 for the languages used in the program, or using the option
4450 @option{-shared-libgcc}, such that it is linked with the shared
4455 Bind references to global symbols when building a shared object. Warn
4456 about any unresolved references (unless overridden by the link editor
4457 option @samp{-Xlinker -z -Xlinker defs}). Only a few systems support
4460 @item -Xlinker @var{option}
4462 Pass @var{option} as an option to the linker. You can use this to
4463 supply system-specific linker options which GCC does not know how to
4466 If you want to pass an option that takes an argument, you must use
4467 @option{-Xlinker} twice, once for the option and once for the argument.
4468 For example, to pass @option{-assert definitions}, you must write
4469 @samp{-Xlinker -assert -Xlinker definitions}. It does not work to write
4470 @option{-Xlinker "-assert definitions"}, because this passes the entire
4471 string as a single argument, which is not what the linker expects.
4473 @item -Wl,@var{option}
4475 Pass @var{option} as an option to the linker. If @var{option} contains
4476 commas, it is split into multiple options at the commas.
4478 @item -u @var{symbol}
4480 Pretend the symbol @var{symbol} is undefined, to force linking of
4481 library modules to define it. You can use @option{-u} multiple times with
4482 different symbols to force loading of additional library modules.
4485 @node Directory Options
4486 @section Options for Directory Search
4487 @cindex directory options
4488 @cindex options, directory search
4491 These options specify directories to search for header files, for
4492 libraries and for parts of the compiler:
4497 Add the directory @var{dir} to the head of the list of directories to be
4498 searched for header files. This can be used to override a system header
4499 file, substituting your own version, since these directories are
4500 searched before the system header file directories. However, you should
4501 not use this option to add directories that contain vendor-supplied
4502 system header files (use @option{-isystem} for that). If you use more than
4503 one @option{-I} option, the directories are scanned in left-to-right
4504 order; the standard system directories come after.
4506 If a standard system include directory, or a directory specified with
4507 @option{-isystem}, is also specified with @option{-I}, it will be
4508 searched only in the position requested by @option{-I}. Also, it will
4509 not be considered a system include directory. If that directory really
4510 does contain system headers, there is a good chance that they will
4511 break. For instance, if GCC's installation procedure edited the headers
4512 in @file{/usr/include} to fix bugs, @samp{-I/usr/include} will cause the
4513 original, buggy headers to be found instead of the corrected ones. GCC
4514 will issue a warning when a system include directory is hidden in this
4519 Any directories you specify with @option{-I} options before the @option{-I-}
4520 option are searched only for the case of @samp{#include "@var{file}"};
4521 they are not searched for @samp{#include <@var{file}>}.
4523 If additional directories are specified with @option{-I} options after
4524 the @option{-I-}, these directories are searched for all @samp{#include}
4525 directives. (Ordinarily @emph{all} @option{-I} directories are used
4528 In addition, the @option{-I-} option inhibits the use of the current
4529 directory (where the current input file came from) as the first search
4530 directory for @samp{#include "@var{file}"}. There is no way to
4531 override this effect of @option{-I-}. With @option{-I.} you can specify
4532 searching the directory which was current when the compiler was
4533 invoked. That is not exactly the same as what the preprocessor does
4534 by default, but it is often satisfactory.
4536 @option{-I-} does not inhibit the use of the standard system directories
4537 for header files. Thus, @option{-I-} and @option{-nostdinc} are
4542 Add directory @var{dir} to the list of directories to be searched
4545 @item -B@var{prefix}
4547 This option specifies where to find the executables, libraries,
4548 include files, and data files of the compiler itself.
4550 The compiler driver program runs one or more of the subprograms
4551 @file{cpp}, @file{cc1}, @file{as} and @file{ld}. It tries
4552 @var{prefix} as a prefix for each program it tries to run, both with and
4553 without @samp{@var{machine}/@var{version}/} (@pxref{Target Options}).
4555 For each subprogram to be run, the compiler driver first tries the
4556 @option{-B} prefix, if any. If that name is not found, or if @option{-B}
4557 was not specified, the driver tries two standard prefixes, which are
4558 @file{/usr/lib/gcc/} and @file{/usr/local/lib/gcc-lib/}. If neither of
4559 those results in a file name that is found, the unmodified program
4560 name is searched for using the directories specified in your
4561 @env{PATH} environment variable.
4563 The compiler will check to see if the path provided by the @option{-B}
4564 refers to a directory, and if necessary it will add a directory
4565 separator character at the end of the path.
4567 @option{-B} prefixes that effectively specify directory names also apply
4568 to libraries in the linker, because the compiler translates these
4569 options into @option{-L} options for the linker. They also apply to
4570 includes files in the preprocessor, because the compiler translates these
4571 options into @option{-isystem} options for the preprocessor. In this case,
4572 the compiler appends @samp{include} to the prefix.
4574 The run-time support file @file{libgcc.a} can also be searched for using
4575 the @option{-B} prefix, if needed. If it is not found there, the two
4576 standard prefixes above are tried, and that is all. The file is left
4577 out of the link if it is not found by those means.
4579 Another way to specify a prefix much like the @option{-B} prefix is to use
4580 the environment variable @env{GCC_EXEC_PREFIX}. @xref{Environment
4583 As a special kludge, if the path provided by @option{-B} is
4584 @file{[dir/]stage@var{N}/}, where @var{N} is a number in the range 0 to
4585 9, then it will be replaced by @file{[dir/]include}. This is to help
4586 with boot-strapping the compiler.
4588 @item -specs=@var{file}
4590 Process @var{file} after the compiler reads in the standard @file{specs}
4591 file, in order to override the defaults that the @file{gcc} driver
4592 program uses when determining what switches to pass to @file{cc1},
4593 @file{cc1plus}, @file{as}, @file{ld}, etc. More than one
4594 @option{-specs=@var{file}} can be specified on the command line, and they
4595 are processed in order, from left to right.
4601 @section Specifying subprocesses and the switches to pass to them
4603 @command{gcc} is a driver program. It performs its job by invoking a
4604 sequence of other programs to do the work of compiling, assembling and
4605 linking. GCC interprets its command-line parameters and uses these to
4606 deduce which programs it should invoke, and which command-line options
4607 it ought to place on their command lines. This behavior is controlled
4608 by @dfn{spec strings}. In most cases there is one spec string for each
4609 program that GCC can invoke, but a few programs have multiple spec
4610 strings to control their behavior. The spec strings built into GCC can
4611 be overridden by using the @option{-specs=} command-line switch to specify
4614 @dfn{Spec files} are plaintext files that are used to construct spec
4615 strings. They consist of a sequence of directives separated by blank
4616 lines. The type of directive is determined by the first non-whitespace
4617 character on the line and it can be one of the following:
4620 @item %@var{command}
4621 Issues a @var{command} to the spec file processor. The commands that can
4625 @item %include <@var{file}>
4627 Search for @var{file} and insert its text at the current point in the
4630 @item %include_noerr <@var{file}>
4631 @cindex %include_noerr
4632 Just like @samp{%include}, but do not generate an error message if the include
4633 file cannot be found.
4635 @item %rename @var{old_name} @var{new_name}
4637 Rename the spec string @var{old_name} to @var{new_name}.
4641 @item *[@var{spec_name}]:
4642 This tells the compiler to create, override or delete the named spec
4643 string. All lines after this directive up to the next directive or
4644 blank line are considered to be the text for the spec string. If this
4645 results in an empty string then the spec will be deleted. (Or, if the
4646 spec did not exist, then nothing will happened.) Otherwise, if the spec
4647 does not currently exist a new spec will be created. If the spec does
4648 exist then its contents will be overridden by the text of this
4649 directive, unless the first character of that text is the @samp{+}
4650 character, in which case the text will be appended to the spec.
4652 @item [@var{suffix}]:
4653 Creates a new @samp{[@var{suffix}] spec} pair. All lines after this directive
4654 and up to the next directive or blank line are considered to make up the
4655 spec string for the indicated suffix. When the compiler encounters an
4656 input file with the named suffix, it will processes the spec string in
4657 order to work out how to compile that file. For example:
4664 This says that any input file whose name ends in @samp{.ZZ} should be
4665 passed to the program @samp{z-compile}, which should be invoked with the
4666 command-line switch @option{-input} and with the result of performing the
4667 @samp{%i} substitution. (See below.)
4669 As an alternative to providing a spec string, the text that follows a
4670 suffix directive can be one of the following:
4673 @item @@@var{language}
4674 This says that the suffix is an alias for a known @var{language}. This is
4675 similar to using the @option{-x} command-line switch to GCC to specify a
4676 language explicitly. For example:
4683 Says that .ZZ files are, in fact, C++ source files.
4686 This causes an error messages saying:
4689 @var{name} compiler not installed on this system.
4693 GCC already has an extensive list of suffixes built into it.
4694 This directive will add an entry to the end of the list of suffixes, but
4695 since the list is searched from the end backwards, it is effectively
4696 possible to override earlier entries using this technique.
4700 GCC has the following spec strings built into it. Spec files can
4701 override these strings or create their own. Note that individual
4702 targets can also add their own spec strings to this list.
4705 asm Options to pass to the assembler
4706 asm_final Options to pass to the assembler post-processor
4707 cpp Options to pass to the C preprocessor
4708 cc1 Options to pass to the C compiler
4709 cc1plus Options to pass to the C++ compiler
4710 endfile Object files to include at the end of the link
4711 link Options to pass to the linker
4712 lib Libraries to include on the command line to the linker
4713 libgcc Decides which GCC support library to pass to the linker
4714 linker Sets the name of the linker
4715 predefines Defines to be passed to the C preprocessor
4716 signed_char Defines to pass to CPP to say whether @code{char} is signed
4718 startfile Object files to include at the start of the link
4721 Here is a small example of a spec file:
4727 --start-group -lgcc -lc -leval1 --end-group %(old_lib)
4730 This example renames the spec called @samp{lib} to @samp{old_lib} and
4731 then overrides the previous definition of @samp{lib} with a new one.
4732 The new definition adds in some extra command-line options before
4733 including the text of the old definition.
4735 @dfn{Spec strings} are a list of command-line options to be passed to their
4736 corresponding program. In addition, the spec strings can contain
4737 @samp{%}-prefixed sequences to substitute variable text or to
4738 conditionally insert text into the command line. Using these constructs
4739 it is possible to generate quite complex command lines.
4741 Here is a table of all defined @samp{%}-sequences for spec
4742 strings. Note that spaces are not generated automatically around the
4743 results of expanding these sequences. Therefore you can concatenate them
4744 together or combine them with constant text in a single argument.
4748 Substitute one @samp{%} into the program name or argument.
4751 Substitute the name of the input file being processed.
4754 Substitute the basename of the input file being processed.
4755 This is the substring up to (and not including) the last period
4756 and not including the directory.
4759 This is the same as @samp{%b}, but include the file suffix (text after
4763 Marks the argument containing or following the @samp{%d} as a
4764 temporary file name, so that that file will be deleted if GCC exits
4765 successfully. Unlike @samp{%g}, this contributes no text to the
4768 @item %g@var{suffix}
4769 Substitute a file name that has suffix @var{suffix} and is chosen
4770 once per compilation, and mark the argument in the same way as
4771 @samp{%d}. To reduce exposure to denial-of-service attacks, the file
4772 name is now chosen in a way that is hard to predict even when previously
4773 chosen file names are known. For example, @samp{%g.s @dots{} %g.o @dots{} %g.s}
4774 might turn into @samp{ccUVUUAU.s ccXYAXZ12.o ccUVUUAU.s}. @var{suffix} matches
4775 the regexp @samp{[.A-Za-z]*} or the special string @samp{%O}, which is
4776 treated exactly as if @samp{%O} had been preprocessed. Previously, @samp{%g}
4777 was simply substituted with a file name chosen once per compilation,
4778 without regard to any appended suffix (which was therefore treated
4779 just like ordinary text), making such attacks more likely to succeed.
4781 @item %u@var{suffix}
4782 Like @samp{%g}, but generates a new temporary file name even if
4783 @samp{%u@var{suffix}} was already seen.
4785 @item %U@var{suffix}
4786 Substitutes the last file name generated with @samp{%u@var{suffix}}, generating a
4787 new one if there is no such last file name. In the absence of any
4788 @samp{%u@var{suffix}}, this is just like @samp{%g@var{suffix}}, except they don't share
4789 the same suffix @emph{space}, so @samp{%g.s @dots{} %U.s @dots{} %g.s @dots{} %U.s}
4790 would involve the generation of two distinct file names, one
4791 for each @samp{%g.s} and another for each @samp{%U.s}. Previously, @samp{%U} was
4792 simply substituted with a file name chosen for the previous @samp{%u},
4793 without regard to any appended suffix.
4795 @item %j@var{SUFFIX}
4796 Substitutes the name of the @code{HOST_BIT_BUCKET}, if any, and if it is
4797 writable, and if save-temps is off; otherwise, substitute the name
4798 of a temporary file, just like @samp{%u}. This temporary file is not
4799 meant for communication between processes, but rather as a junk
4802 @item %.@var{SUFFIX}
4803 Substitutes @var{.SUFFIX} for the suffixes of a matched switch's args
4804 when it is subsequently output with @samp{%*}. @var{SUFFIX} is
4805 terminated by the next space or %.
4808 Marks the argument containing or following the @samp{%w} as the
4809 designated output file of this compilation. This puts the argument
4810 into the sequence of arguments that @samp{%o} will substitute later.
4813 Substitutes the names of all the output files, with spaces
4814 automatically placed around them. You should write spaces
4815 around the @samp{%o} as well or the results are undefined.
4816 @samp{%o} is for use in the specs for running the linker.
4817 Input files whose names have no recognized suffix are not compiled
4818 at all, but they are included among the output files, so they will
4822 Substitutes the suffix for object files. Note that this is
4823 handled specially when it immediately follows @samp{%g, %u, or %U},
4824 because of the need for those to form complete file names. The
4825 handling is such that @samp{%O} is treated exactly as if it had already
4826 been substituted, except that @samp{%g, %u, and %U} do not currently
4827 support additional @var{suffix} characters following @samp{%O} as they would
4828 following, for example, @samp{.o}.
4831 Substitutes the standard macro predefinitions for the
4832 current target machine. Use this when running @code{cpp}.
4835 Like @samp{%p}, but puts @samp{__} before and after the name of each
4836 predefined macro, except for macros that start with @samp{__} or with
4837 @samp{_@var{L}}, where @var{L} is an uppercase letter. This is for ISO
4841 Substitute a @option{-iprefix} option made from @env{GCC_EXEC_PREFIX}.
4844 Current argument is the name of a library or startup file of some sort.
4845 Search for that file in a standard list of directories and substitute
4846 the full name found.
4849 Print @var{str} as an error message. @var{str} is terminated by a newline.
4850 Use this when inconsistent options are detected.
4853 Output @samp{-} if the input for the current command is coming from a pipe.
4856 Substitute the contents of spec string @var{name} at this point.
4859 Like @samp{%(@dots{})} but put @samp{__} around @option{-D} arguments.
4861 @item %x@{@var{option}@}
4862 Accumulate an option for @samp{%X}.
4865 Output the accumulated linker options specified by @option{-Wl} or a @samp{%x}
4869 Output the accumulated assembler options specified by @option{-Wa}.
4872 Output the accumulated preprocessor options specified by @option{-Wp}.
4875 Substitute the major version number of GCC@.
4876 (For version 2.9.5, this is 2.)
4879 Substitute the minor version number of GCC@.
4880 (For version 2.9.5, this is 9.)
4883 Substitute the patch level number of GCC@.
4884 (For version 2.9.5, this is 5.)
4887 Process the @code{asm} spec. This is used to compute the
4888 switches to be passed to the assembler.
4891 Process the @code{asm_final} spec. This is a spec string for
4892 passing switches to an assembler post-processor, if such a program is
4896 Process the @code{link} spec. This is the spec for computing the
4897 command line passed to the linker. Typically it will make use of the
4898 @samp{%L %G %S %D and %E} sequences.
4901 Dump out a @option{-L} option for each directory that GCC believes might
4902 contain startup files. If the target supports multilibs then the
4903 current multilib directory will be prepended to each of these paths.
4906 Output the multilib directory with directory separators replaced with
4907 @samp{_}. If multilib directories are not set, or the multilib directory is
4908 @file{.} then this option emits nothing.
4911 Process the @code{lib} spec. This is a spec string for deciding which
4912 libraries should be included on the command line to the linker.
4915 Process the @code{libgcc} spec. This is a spec string for deciding
4916 which GCC support library should be included on the command line to the linker.
4919 Process the @code{startfile} spec. This is a spec for deciding which
4920 object files should be the first ones passed to the linker. Typically
4921 this might be a file named @file{crt0.o}.
4924 Process the @code{endfile} spec. This is a spec string that specifies
4925 the last object files that will be passed to the linker.
4928 Process the @code{cpp} spec. This is used to construct the arguments
4929 to be passed to the C preprocessor.
4932 Process the @code{signed_char} spec. This is intended to be used
4933 to tell cpp whether a char is signed. It typically has the definition:
4935 %@{funsigned-char:-D__CHAR_UNSIGNED__@}
4939 Process the @code{cc1} spec. This is used to construct the options to be
4940 passed to the actual C compiler (@samp{cc1}).
4943 Process the @code{cc1plus} spec. This is used to construct the options to be
4944 passed to the actual C++ compiler (@samp{cc1plus}).
4947 Substitute the variable part of a matched option. See below.
4948 Note that each comma in the substituted string is replaced by
4952 Substitutes the @code{-S} switch, if that switch was given to GCC@.
4953 If that switch was not specified, this substitutes nothing. Note that
4954 the leading dash is omitted when specifying this option, and it is
4955 automatically inserted if the substitution is performed. Thus the spec
4956 string @samp{%@{foo@}} would match the command-line option @option{-foo}
4957 and would output the command line option @option{-foo}.
4959 @item %W@{@code{S}@}
4960 Like %@{@code{S}@} but mark last argument supplied within as a file to be
4963 @item %@{@code{S}*@}
4964 Substitutes all the switches specified to GCC whose names start
4965 with @code{-S}, but which also take an argument. This is used for
4966 switches like @option{-o}, @option{-D}, @option{-I}, etc.
4967 GCC considers @option{-o foo} as being
4968 one switch whose names starts with @samp{o}. %@{o*@} would substitute this
4969 text, including the space. Thus two arguments would be generated.
4971 @item %@{^@code{S}*@}
4972 Like %@{@code{S}*@}, but don't put a blank between a switch and its
4973 argument. Thus %@{^o*@} would only generate one argument, not two.
4975 @item %@{@code{S}*&@code{T}*@}
4976 Like %@{@code{S}*@}, but preserve order of @code{S} and @code{T} options
4977 (the order of @code{S} and @code{T} in the spec is not significant).
4978 There can be any number of ampersand-separated variables; for each the
4979 wild card is optional. Useful for CPP as @samp{%@{D*&U*&A*@}}.
4981 @item %@{<@code{S}@}
4982 Remove all occurrences of @code{-S} from the command line. Note---this
4983 command is position dependent. @samp{%} commands in the spec string
4984 before this option will see @code{-S}, @samp{%} commands in the spec
4985 string after this option will not.
4987 @item %@{@code{S}*:@code{X}@}
4988 Substitutes @code{X} if one or more switches whose names start with
4989 @code{-S} are specified to GCC@. Note that the tail part of the
4990 @code{-S} option (i.e.@: the part matched by the @samp{*}) will be substituted
4991 for each occurrence of @samp{%*} within @code{X}.
4993 @item %@{@code{S}:@code{X}@}
4994 Substitutes @code{X}, but only if the @samp{-S} switch was given to GCC@.
4996 @item %@{!@code{S}:@code{X}@}
4997 Substitutes @code{X}, but only if the @samp{-S} switch was @emph{not} given to GCC@.
4999 @item %@{|@code{S}:@code{X}@}
5000 Like %@{@code{S}:@code{X}@}, but if no @code{S} switch, substitute @samp{-}.
5002 @item %@{|!@code{S}:@code{X}@}
5003 Like %@{!@code{S}:@code{X}@}, but if there is an @code{S} switch, substitute @samp{-}.
5005 @item %@{.@code{S}:@code{X}@}
5006 Substitutes @code{X}, but only if processing a file with suffix @code{S}.
5008 @item %@{!.@code{S}:@code{X}@}
5009 Substitutes @code{X}, but only if @emph{not} processing a file with suffix @code{S}.
5011 @item %@{@code{S}|@code{P}:@code{X}@}
5012 Substitutes @code{X} if either @code{-S} or @code{-P} was given to GCC@. This may be
5013 combined with @samp{!} and @samp{.} sequences as well, although they
5014 have a stronger binding than the @samp{|}. For example a spec string
5018 %@{.c:-foo@} %@{!.c:-bar@} %@{.c|d:-baz@} %@{!.c|d:-boggle@}
5021 will output the following command-line options from the following input
5022 command-line options:
5027 -d fred.c -foo -baz -boggle
5028 -d jim.d -bar -baz -boggle
5033 The conditional text @code{X} in a %@{@code{S}:@code{X}@} or
5034 %@{!@code{S}:@code{X}@} construct may contain other nested @samp{%} constructs
5035 or spaces, or even newlines. They are processed as usual, as described
5038 The @option{-O}, @option{-f}, @option{-m}, and @option{-W}
5039 switches are handled specifically in these
5040 constructs. If another value of @option{-O} or the negated form of a @option{-f}, @option{-m}, or
5041 @option{-W} switch is found later in the command line, the earlier switch
5042 value is ignored, except with @{@code{S}*@} where @code{S} is just one
5043 letter, which passes all matching options.
5045 The character @samp{|} at the beginning of the predicate text is used to indicate
5046 that a command should be piped to the following command, but only if @option{-pipe}
5049 It is built into GCC which switches take arguments and which do not.
5050 (You might think it would be useful to generalize this to allow each
5051 compiler's spec to say which switches take arguments. But this cannot
5052 be done in a consistent fashion. GCC cannot even decide which input
5053 files have been specified without knowing which switches take arguments,
5054 and it must know which input files to compile in order to tell which
5057 GCC also knows implicitly that arguments starting in @option{-l} are to be
5058 treated as compiler output files, and passed to the linker in their
5059 proper position among the other output files.
5061 @c man begin OPTIONS
5063 @node Target Options
5064 @section Specifying Target Machine and Compiler Version
5065 @cindex target options
5066 @cindex cross compiling
5067 @cindex specifying machine version
5068 @cindex specifying compiler version and target machine
5069 @cindex compiler version, specifying
5070 @cindex target machine, specifying
5072 By default, GCC compiles code for the same type of machine that you
5073 are using. However, it can also be installed as a cross-compiler, to
5074 compile for some other type of machine. In fact, several different
5075 configurations of GCC, for different target machines, can be
5076 installed side by side. Then you specify which one to use with the
5079 In addition, older and newer versions of GCC can be installed side
5080 by side. One of them (probably the newest) will be the default, but
5081 you may sometimes wish to use another.
5084 @item -b @var{machine}
5086 The argument @var{machine} specifies the target machine for compilation.
5087 This is useful when you have installed GCC as a cross-compiler.
5089 The value to use for @var{machine} is the same as was specified as the
5090 machine type when configuring GCC as a cross-compiler. For
5091 example, if a cross-compiler was configured with @samp{configure
5092 i386v}, meaning to compile for an 80386 running System V, then you
5093 would specify @option{-b i386v} to run that cross compiler.
5095 When you do not specify @option{-b}, it normally means to compile for
5096 the same type of machine that you are using.
5098 @item -V @var{version}
5100 The argument @var{version} specifies which version of GCC to run.
5101 This is useful when multiple versions are installed. For example,
5102 @var{version} might be @samp{2.0}, meaning to run GCC version 2.0.
5104 The default version, when you do not specify @option{-V}, is the last
5105 version of GCC that you installed.
5108 The @option{-b} and @option{-V} options actually work by controlling part of
5109 the file name used for the executable files and libraries used for
5110 compilation. A given version of GCC, for a given target machine, is
5111 normally kept in the directory @file{/usr/local/lib/gcc-lib/@var{machine}/@var{version}}.
5113 Thus, sites can customize the effect of @option{-b} or @option{-V} either by
5114 changing the names of these directories or adding alternate names (or
5115 symbolic links). If in directory @file{/usr/local/lib/gcc-lib/} the
5116 file @file{80386} is a link to the file @file{i386v}, then @option{-b
5117 80386} becomes an alias for @option{-b i386v}.
5119 In one respect, the @option{-b} or @option{-V} do not completely change
5120 to a different compiler: the top-level driver program @command{gcc}
5121 that you originally invoked continues to run and invoke the other
5122 executables (preprocessor, compiler per se, assembler and linker)
5123 that do the real work. However, since no real work is done in the
5124 driver program, it usually does not matter that the driver program
5125 in use is not the one for the specified target. It is common for the
5126 interface to the other executables to change incompatibly between
5127 compiler versions, so unless the version specified is very close to that
5128 of the driver (for example, @option{-V 3.0} with a driver program from GCC
5129 version 3.0.1), use of @option{-V} may not work; for example, using
5130 @option{-V 2.95.2} will not work with a driver program from GCC 3.0.
5132 The only way that the driver program depends on the target machine is
5133 in the parsing and handling of special machine-specific options.
5134 However, this is controlled by a file which is found, along with the
5135 other executables, in the directory for the specified version and
5136 target machine. As a result, a single installed driver program adapts
5137 to any specified target machine, and sufficiently similar compiler
5140 The driver program executable does control one significant thing,
5141 however: the default version and target machine. Therefore, you can
5142 install different instances of the driver program, compiled for
5143 different targets or versions, under different names.
5145 For example, if the driver for version 2.0 is installed as @command{ogcc}
5146 and that for version 2.1 is installed as @command{gcc}, then the command
5147 @command{gcc} will use version 2.1 by default, while @command{ogcc} will use
5148 2.0 by default. However, you can choose either version with either
5149 command with the @option{-V} option.
5151 @node Submodel Options
5152 @section Hardware Models and Configurations
5153 @cindex submodel options
5154 @cindex specifying hardware config
5155 @cindex hardware models and configurations, specifying
5156 @cindex machine dependent options
5158 Earlier we discussed the standard option @option{-b} which chooses among
5159 different installed compilers for completely different target
5160 machines, such as VAX vs.@: 68000 vs.@: 80386.
5162 In addition, each of these target machine types can have its own
5163 special options, starting with @samp{-m}, to choose among various
5164 hardware models or configurations---for example, 68010 vs 68020,
5165 floating coprocessor or none. A single installed version of the
5166 compiler can compile for any model or configuration, according to the
5169 Some configurations of the compiler also support additional special
5170 options, usually for compatibility with other compilers on the same
5173 These options are defined by the macro @code{TARGET_SWITCHES} in the
5174 machine description. The default for the options is also defined by
5175 that macro, which enables you to change the defaults.
5189 * RS/6000 and PowerPC Options::
5192 * i386 and x86-64 Options::
5194 * Intel 960 Options::
5195 * DEC Alpha Options::
5196 * DEC Alpha/VMS Options::
5200 * System V Options::
5201 * TMS320C3x/C4x Options::
5209 * S/390 and zSeries Options::
5213 * Xstormy16 Options::
5217 @node M680x0 Options
5218 @subsection M680x0 Options
5219 @cindex M680x0 options
5221 These are the @samp{-m} options defined for the 68000 series. The default
5222 values for these options depends on which style of 68000 was selected when
5223 the compiler was configured; the defaults for the most common choices are
5231 Generate output for a 68000. This is the default
5232 when the compiler is configured for 68000-based systems.
5234 Use this option for microcontrollers with a 68000 or EC000 core,
5235 including the 68008, 68302, 68306, 68307, 68322, 68328 and 68356.
5241 Generate output for a 68020. This is the default
5242 when the compiler is configured for 68020-based systems.
5246 Generate output containing 68881 instructions for floating point.
5247 This is the default for most 68020 systems unless @option{--nfp} was
5248 specified when the compiler was configured.
5252 Generate output for a 68030. This is the default when the compiler is
5253 configured for 68030-based systems.
5257 Generate output for a 68040. This is the default when the compiler is
5258 configured for 68040-based systems.
5260 This option inhibits the use of 68881/68882 instructions that have to be
5261 emulated by software on the 68040. Use this option if your 68040 does not
5262 have code to emulate those instructions.
5266 Generate output for a 68060. This is the default when the compiler is
5267 configured for 68060-based systems.
5269 This option inhibits the use of 68020 and 68881/68882 instructions that
5270 have to be emulated by software on the 68060. Use this option if your 68060
5271 does not have code to emulate those instructions.
5275 Generate output for a CPU32. This is the default
5276 when the compiler is configured for CPU32-based systems.
5278 Use this option for microcontrollers with a
5279 CPU32 or CPU32+ core, including the 68330, 68331, 68332, 68333, 68334,
5280 68336, 68340, 68341, 68349 and 68360.
5284 Generate output for a 520X ``coldfire'' family cpu. This is the default
5285 when the compiler is configured for 520X-based systems.
5287 Use this option for microcontroller with a 5200 core, including
5288 the MCF5202, MCF5203, MCF5204 and MCF5202.
5293 Generate output for a 68040, without using any of the new instructions.
5294 This results in code which can run relatively efficiently on either a
5295 68020/68881 or a 68030 or a 68040. The generated code does use the
5296 68881 instructions that are emulated on the 68040.
5300 Generate output for a 68060, without using any of the new instructions.
5301 This results in code which can run relatively efficiently on either a
5302 68020/68881 or a 68030 or a 68040. The generated code does use the
5303 68881 instructions that are emulated on the 68060.
5307 Generate output containing Sun FPA instructions for floating point.
5310 @opindex msoft-float
5311 Generate output containing library calls for floating point.
5312 @strong{Warning:} the requisite libraries are not available for all m68k
5313 targets. Normally the facilities of the machine's usual C compiler are
5314 used, but this can't be done directly in cross-compilation. You must
5315 make your own arrangements to provide suitable library functions for
5316 cross-compilation. The embedded targets @samp{m68k-*-aout} and
5317 @samp{m68k-*-coff} do provide software floating point support.
5321 Consider type @code{int} to be 16 bits wide, like @code{short int}.
5324 @opindex mnobitfield
5325 Do not use the bit-field instructions. The @option{-m68000}, @option{-mcpu32}
5326 and @option{-m5200} options imply @w{@option{-mnobitfield}}.
5330 Do use the bit-field instructions. The @option{-m68020} option implies
5331 @option{-mbitfield}. This is the default if you use a configuration
5332 designed for a 68020.
5336 Use a different function-calling convention, in which functions
5337 that take a fixed number of arguments return with the @code{rtd}
5338 instruction, which pops their arguments while returning. This
5339 saves one instruction in the caller since there is no need to pop
5340 the arguments there.
5342 This calling convention is incompatible with the one normally
5343 used on Unix, so you cannot use it if you need to call libraries
5344 compiled with the Unix compiler.
5346 Also, you must provide function prototypes for all functions that
5347 take variable numbers of arguments (including @code{printf});
5348 otherwise incorrect code will be generated for calls to those
5351 In addition, seriously incorrect code will result if you call a
5352 function with too many arguments. (Normally, extra arguments are
5353 harmlessly ignored.)
5355 The @code{rtd} instruction is supported by the 68010, 68020, 68030,
5356 68040, 68060 and CPU32 processors, but not by the 68000 or 5200.
5359 @itemx -mno-align-int
5361 @opindex mno-align-int
5362 Control whether GCC aligns @code{int}, @code{long}, @code{long long},
5363 @code{float}, @code{double}, and @code{long double} variables on a 32-bit
5364 boundary (@option{-malign-int}) or a 16-bit boundary (@option{-mno-align-int}).
5365 Aligning variables on 32-bit boundaries produces code that runs somewhat
5366 faster on processors with 32-bit busses at the expense of more memory.
5368 @strong{Warning:} if you use the @option{-malign-int} switch, GCC will
5369 align structures containing the above types differently than
5370 most published application binary interface specifications for the m68k.
5374 Use the pc-relative addressing mode of the 68000 directly, instead of
5375 using a global offset table. At present, this option implies @option{-fpic},
5376 allowing at most a 16-bit offset for pc-relative addressing. @option{-fPIC} is
5377 not presently supported with @option{-mpcrel}, though this could be supported for
5378 68020 and higher processors.
5380 @item -mno-strict-align
5381 @itemx -mstrict-align
5382 @opindex mno-strict-align
5383 @opindex mstrict-align
5384 Do not (do) assume that unaligned memory references will be handled by
5389 @node M68hc1x Options
5390 @subsection M68hc1x Options
5391 @cindex M68hc1x options
5393 These are the @samp{-m} options defined for the 68hc11 and 68hc12
5394 microcontrollers. The default values for these options depends on
5395 which style of microcontroller was selected when the compiler was configured;
5396 the defaults for the most common choices are given below.
5403 Generate output for a 68HC11. This is the default
5404 when the compiler is configured for 68HC11-based systems.
5410 Generate output for a 68HC12. This is the default
5411 when the compiler is configured for 68HC12-based systems.
5414 @opindex mauto-incdec
5415 Enable the use of 68HC12 pre and post auto-increment and auto-decrement
5420 Consider type @code{int} to be 16 bits wide, like @code{short int}.
5422 @item -msoft-reg-count=@var{count}
5423 @opindex msoft-reg-count
5424 Specify the number of pseudo-soft registers which are used for the
5425 code generation. The maximum number is 32. Using more pseudo-soft
5426 register may or may not result in better code depending on the program.
5427 The default is 4 for 68HC11 and 2 for 68HC12.
5432 @subsection VAX Options
5435 These @samp{-m} options are defined for the VAX:
5440 Do not output certain jump instructions (@code{aobleq} and so on)
5441 that the Unix assembler for the VAX cannot handle across long
5446 Do output those jump instructions, on the assumption that you
5447 will assemble with the GNU assembler.
5451 Output code for g-format floating point numbers instead of d-format.
5455 @subsection SPARC Options
5456 @cindex SPARC options
5458 These @samp{-m} switches are supported on the SPARC:
5463 @opindex mno-app-regs
5465 Specify @option{-mapp-regs} to generate output using the global registers
5466 2 through 4, which the SPARC SVR4 ABI reserves for applications. This
5469 To be fully SVR4 ABI compliant at the cost of some performance loss,
5470 specify @option{-mno-app-regs}. You should compile libraries and system
5471 software with this option.
5476 @opindex mhard-float
5477 Generate output containing floating point instructions. This is the
5483 @opindex msoft-float
5484 Generate output containing library calls for floating point.
5485 @strong{Warning:} the requisite libraries are not available for all SPARC
5486 targets. Normally the facilities of the machine's usual C compiler are
5487 used, but this cannot be done directly in cross-compilation. You must make
5488 your own arrangements to provide suitable library functions for
5489 cross-compilation. The embedded targets @samp{sparc-*-aout} and
5490 @samp{sparclite-*-*} do provide software floating point support.
5492 @option{-msoft-float} changes the calling convention in the output file;
5493 therefore, it is only useful if you compile @emph{all} of a program with
5494 this option. In particular, you need to compile @file{libgcc.a}, the
5495 library that comes with GCC, with @option{-msoft-float} in order for
5498 @item -mhard-quad-float
5499 @opindex mhard-quad-float
5500 Generate output containing quad-word (long double) floating point
5503 @item -msoft-quad-float
5504 @opindex msoft-quad-float
5505 Generate output containing library calls for quad-word (long double)
5506 floating point instructions. The functions called are those specified
5507 in the SPARC ABI@. This is the default.
5509 As of this writing, there are no sparc implementations that have hardware
5510 support for the quad-word floating point instructions. They all invoke
5511 a trap handler for one of these instructions, and then the trap handler
5512 emulates the effect of the instruction. Because of the trap handler overhead,
5513 this is much slower than calling the ABI library routines. Thus the
5514 @option{-msoft-quad-float} option is the default.
5518 @opindex mno-epilogue
5520 With @option{-mepilogue} (the default), the compiler always emits code for
5521 function exit at the end of each function. Any function exit in
5522 the middle of the function (such as a return statement in C) will
5523 generate a jump to the exit code at the end of the function.
5525 With @option{-mno-epilogue}, the compiler tries to emit exit code inline
5526 at every function exit.
5532 With @option{-mflat}, the compiler does not generate save/restore instructions
5533 and will use a ``flat'' or single register window calling convention.
5534 This model uses %i7 as the frame pointer and is compatible with the normal
5535 register window model. Code from either may be intermixed.
5536 The local registers and the input registers (0--5) are still treated as
5537 ``call saved'' registers and will be saved on the stack as necessary.
5539 With @option{-mno-flat} (the default), the compiler emits save/restore
5540 instructions (except for leaf functions) and is the normal mode of operation.
5542 @item -mno-unaligned-doubles
5543 @itemx -munaligned-doubles
5544 @opindex mno-unaligned-doubles
5545 @opindex munaligned-doubles
5546 Assume that doubles have 8 byte alignment. This is the default.
5548 With @option{-munaligned-doubles}, GCC assumes that doubles have 8 byte
5549 alignment only if they are contained in another type, or if they have an
5550 absolute address. Otherwise, it assumes they have 4 byte alignment.
5551 Specifying this option avoids some rare compatibility problems with code
5552 generated by other compilers. It is not the default because it results
5553 in a performance loss, especially for floating point code.
5555 @item -mno-faster-structs
5556 @itemx -mfaster-structs
5557 @opindex mno-faster-structs
5558 @opindex mfaster-structs
5559 With @option{-mfaster-structs}, the compiler assumes that structures
5560 should have 8 byte alignment. This enables the use of pairs of
5561 @code{ldd} and @code{std} instructions for copies in structure
5562 assignment, in place of twice as many @code{ld} and @code{st} pairs.
5563 However, the use of this changed alignment directly violates the Sparc
5564 ABI@. Thus, it's intended only for use on targets where the developer
5565 acknowledges that their resulting code will not be directly in line with
5566 the rules of the ABI@.
5572 These two options select variations on the SPARC architecture.
5574 By default (unless specifically configured for the Fujitsu SPARClite),
5575 GCC generates code for the v7 variant of the SPARC architecture.
5577 @option{-mv8} will give you SPARC v8 code. The only difference from v7
5578 code is that the compiler emits the integer multiply and integer
5579 divide instructions which exist in SPARC v8 but not in SPARC v7.
5581 @option{-msparclite} will give you SPARClite code. This adds the integer
5582 multiply, integer divide step and scan (@code{ffs}) instructions which
5583 exist in SPARClite but not in SPARC v7.
5585 These options are deprecated and will be deleted in a future GCC release.
5586 They have been replaced with @option{-mcpu=xxx}.
5591 @opindex msupersparc
5592 These two options select the processor for which the code is optimized.
5594 With @option{-mcypress} (the default), the compiler optimizes code for the
5595 Cypress CY7C602 chip, as used in the SparcStation/SparcServer 3xx series.
5596 This is also appropriate for the older SparcStation 1, 2, IPX etc.
5598 With @option{-msupersparc} the compiler optimizes code for the SuperSparc cpu, as
5599 used in the SparcStation 10, 1000 and 2000 series. This flag also enables use
5600 of the full SPARC v8 instruction set.
5602 These options are deprecated and will be deleted in a future GCC release.
5603 They have been replaced with @option{-mcpu=xxx}.
5605 @item -mcpu=@var{cpu_type}
5607 Set the instruction set, register set, and instruction scheduling parameters
5608 for machine type @var{cpu_type}. Supported values for @var{cpu_type} are
5609 @samp{v7}, @samp{cypress}, @samp{v8}, @samp{supersparc}, @samp{sparclite},
5610 @samp{hypersparc}, @samp{sparclite86x}, @samp{f930}, @samp{f934},
5611 @samp{sparclet}, @samp{tsc701}, @samp{v9}, and @samp{ultrasparc}.
5613 Default instruction scheduling parameters are used for values that select
5614 an architecture and not an implementation. These are @samp{v7}, @samp{v8},
5615 @samp{sparclite}, @samp{sparclet}, @samp{v9}.
5617 Here is a list of each supported architecture and their supported
5622 v8: supersparc, hypersparc
5623 sparclite: f930, f934, sparclite86x
5628 @item -mtune=@var{cpu_type}
5630 Set the instruction scheduling parameters for machine type
5631 @var{cpu_type}, but do not set the instruction set or register set that the
5632 option @option{-mcpu=@var{cpu_type}} would.
5634 The same values for @option{-mcpu=@var{cpu_type}} can be used for
5635 @option{-mtune=@var{cpu_type}}, but the only useful values are those
5636 that select a particular cpu implementation. Those are @samp{cypress},
5637 @samp{supersparc}, @samp{hypersparc}, @samp{f930}, @samp{f934},
5638 @samp{sparclite86x}, @samp{tsc701}, and @samp{ultrasparc}.
5642 These @samp{-m} switches are supported in addition to the above
5643 on the SPARCLET processor.
5646 @item -mlittle-endian
5647 @opindex mlittle-endian
5648 Generate code for a processor running in little-endian mode.
5652 Treat register @code{%g0} as a normal register.
5653 GCC will continue to clobber it as necessary but will not assume
5654 it always reads as 0.
5656 @item -mbroken-saverestore
5657 @opindex mbroken-saverestore
5658 Generate code that does not use non-trivial forms of the @code{save} and
5659 @code{restore} instructions. Early versions of the SPARCLET processor do
5660 not correctly handle @code{save} and @code{restore} instructions used with
5661 arguments. They correctly handle them used without arguments. A @code{save}
5662 instruction used without arguments increments the current window pointer
5663 but does not allocate a new stack frame. It is assumed that the window
5664 overflow trap handler will properly handle this case as will interrupt
5668 These @samp{-m} switches are supported in addition to the above
5669 on SPARC V9 processors in 64-bit environments.
5672 @item -mlittle-endian
5673 @opindex mlittle-endian
5674 Generate code for a processor running in little-endian mode.
5680 Generate code for a 32-bit or 64-bit environment.
5681 The 32-bit environment sets int, long and pointer to 32 bits.
5682 The 64-bit environment sets int to 32 bits and long and pointer
5685 @item -mcmodel=medlow
5686 @opindex mcmodel=medlow
5687 Generate code for the Medium/Low code model: the program must be linked
5688 in the low 32 bits of the address space. Pointers are 64 bits.
5689 Programs can be statically or dynamically linked.
5691 @item -mcmodel=medmid
5692 @opindex mcmodel=medmid
5693 Generate code for the Medium/Middle code model: the program must be linked
5694 in the low 44 bits of the address space, the text segment must be less than
5695 2G bytes, and data segment must be within 2G of the text segment.
5696 Pointers are 64 bits.
5698 @item -mcmodel=medany
5699 @opindex mcmodel=medany
5700 Generate code for the Medium/Anywhere code model: the program may be linked
5701 anywhere in the address space, the text segment must be less than
5702 2G bytes, and data segment must be within 2G of the text segment.
5703 Pointers are 64 bits.
5705 @item -mcmodel=embmedany
5706 @opindex mcmodel=embmedany
5707 Generate code for the Medium/Anywhere code model for embedded systems:
5708 assume a 32-bit text and a 32-bit data segment, both starting anywhere
5709 (determined at link time). Register %g4 points to the base of the
5710 data segment. Pointers are still 64 bits.
5711 Programs are statically linked, PIC is not supported.
5714 @itemx -mno-stack-bias
5715 @opindex mstack-bias
5716 @opindex mno-stack-bias
5717 With @option{-mstack-bias}, GCC assumes that the stack pointer, and
5718 frame pointer if present, are offset by @minus{}2047 which must be added back
5719 when making stack frame references.
5720 Otherwise, assume no such offset is present.
5723 @node Convex Options
5724 @subsection Convex Options
5725 @cindex Convex options
5727 These @samp{-m} options are defined for Convex:
5732 Generate output for C1. The code will run on any Convex machine.
5733 The preprocessor symbol @code{__convex__c1__} is defined.
5737 Generate output for C2. Uses instructions not available on C1.
5738 Scheduling and other optimizations are chosen for max performance on C2.
5739 The preprocessor symbol @code{__convex_c2__} is defined.
5743 Generate output for C32xx. Uses instructions not available on C1.
5744 Scheduling and other optimizations are chosen for max performance on C32.
5745 The preprocessor symbol @code{__convex_c32__} is defined.
5749 Generate output for C34xx. Uses instructions not available on C1.
5750 Scheduling and other optimizations are chosen for max performance on C34.
5751 The preprocessor symbol @code{__convex_c34__} is defined.
5755 Generate output for C38xx. Uses instructions not available on C1.
5756 Scheduling and other optimizations are chosen for max performance on C38.
5757 The preprocessor symbol @code{__convex_c38__} is defined.
5761 Generate code which puts an argument count in the word preceding each
5762 argument list. This is compatible with regular CC, and a few programs
5763 may need the argument count word. GDB and other source-level debuggers
5764 do not need it; this info is in the symbol table.
5767 @opindex mnoargcount
5768 Omit the argument count word. This is the default.
5770 @item -mvolatile-cache
5771 @opindex mvolatile-cache
5772 Allow volatile references to be cached. This is the default.
5774 @item -mvolatile-nocache
5775 @opindex mvolatile-nocache
5776 Volatile references bypass the data cache, going all the way to memory.
5777 This is only needed for multi-processor code that does not use standard
5778 synchronization instructions. Making non-volatile references to volatile
5779 locations will not necessarily work.
5783 Type long is 32 bits, the same as type int. This is the default.
5787 Type long is 64 bits, the same as type long long. This option is useless,
5788 because no library support exists for it.
5791 @node AMD29K Options
5792 @subsection AMD29K Options
5793 @cindex AMD29K options
5795 These @samp{-m} options are defined for the AMD Am29000:
5800 @cindex DW bit (29k)
5801 Generate code that assumes the @code{DW} bit is set, i.e., that byte and
5802 halfword operations are directly supported by the hardware. This is the
5807 Generate code that assumes the @code{DW} bit is not set.
5811 @cindex byte writes (29k)
5812 Generate code that assumes the system supports byte and halfword write
5813 operations. This is the default.
5817 Generate code that assumes the systems does not support byte and
5818 halfword write operations. @option{-mnbw} implies @option{-mndw}.
5822 @cindex memory model (29k)
5823 Use a small memory model that assumes that all function addresses are
5824 either within a single 256 KB segment or at an absolute address of less
5825 than 256k. This allows the @code{call} instruction to be used instead
5826 of a @code{const}, @code{consth}, @code{calli} sequence.
5830 Use the normal memory model: Generate @code{call} instructions only when
5831 calling functions in the same file and @code{calli} instructions
5832 otherwise. This works if each file occupies less than 256 KB but allows
5833 the entire executable to be larger than 256 KB@. This is the default.
5837 Always use @code{calli} instructions. Specify this option if you expect
5838 a single file to compile into more than 256 KB of code.
5842 @cindex processor selection (29k)
5843 Generate code for the Am29050.
5847 Generate code for the Am29000. This is the default.
5849 @item -mkernel-registers
5850 @opindex mkernel-registers
5851 @cindex kernel and user registers (29k)
5852 Generate references to registers @code{gr64-gr95} instead of to
5853 registers @code{gr96-gr127}. This option can be used when compiling
5854 kernel code that wants a set of global registers disjoint from that used
5857 Note that when this option is used, register names in @samp{-f} flags
5858 must use the normal, user-mode, names.
5860 @item -muser-registers
5861 @opindex muser-registers
5862 Use the normal set of global registers, @code{gr96-gr127}. This is the
5866 @itemx -mno-stack-check
5867 @opindex mstack-check
5868 @opindex mno-stack-check
5869 @cindex stack checks (29k)
5870 Insert (or do not insert) a call to @code{__msp_check} after each stack
5871 adjustment. This is often used for kernel code.
5874 @itemx -mno-storem-bug
5875 @opindex mstorem-bug
5876 @opindex mno-storem-bug
5877 @cindex storem bug (29k)
5878 @option{-mstorem-bug} handles 29k processors which cannot handle the
5879 separation of a mtsrim insn and a storem instruction (most 29000 chips
5880 to date, but not the 29050).
5882 @item -mno-reuse-arg-regs
5883 @itemx -mreuse-arg-regs
5884 @opindex mno-reuse-arg-regs
5885 @opindex mreuse-arg-regs
5886 @option{-mno-reuse-arg-regs} tells the compiler to only use incoming argument
5887 registers for copying out arguments. This helps detect calling a function
5888 with fewer arguments than it was declared with.
5890 @item -mno-impure-text
5891 @itemx -mimpure-text
5892 @opindex mno-impure-text
5893 @opindex mimpure-text
5894 @option{-mimpure-text}, used in addition to @option{-shared}, tells the compiler to
5895 not pass @option{-assert pure-text} to the linker when linking a shared object.
5898 @opindex msoft-float
5899 Generate output containing library calls for floating point.
5900 @strong{Warning:} the requisite libraries are not part of GCC@.
5901 Normally the facilities of the machine's usual C compiler are used, but
5902 this can't be done directly in cross-compilation. You must make your
5903 own arrangements to provide suitable library functions for
5908 Do not generate multm or multmu instructions. This is useful for some embedded
5909 systems which do not have trap handlers for these instructions.
5913 @subsection ARM Options
5916 These @samp{-m} options are defined for Advanced RISC Machines (ARM)
5921 @opindex mapcs-frame
5922 Generate a stack frame that is compliant with the ARM Procedure Call
5923 Standard for all functions, even if this is not strictly necessary for
5924 correct execution of the code. Specifying @option{-fomit-frame-pointer}
5925 with this option will cause the stack frames not to be generated for
5926 leaf functions. The default is @option{-mno-apcs-frame}.
5930 This is a synonym for @option{-mapcs-frame}.
5934 Generate code for a processor running with a 26-bit program counter,
5935 and conforming to the function calling standards for the APCS 26-bit
5936 option. This option replaces the @option{-m2} and @option{-m3} options
5937 of previous releases of the compiler.
5941 Generate code for a processor running with a 32-bit program counter,
5942 and conforming to the function calling standards for the APCS 32-bit
5943 option. This option replaces the @option{-m6} option of previous releases
5947 @c not currently implemented
5948 @item -mapcs-stack-check
5949 @opindex mapcs-stack-check
5950 Generate code to check the amount of stack space available upon entry to
5951 every function (that actually uses some stack space). If there is
5952 insufficient space available then either the function
5953 @samp{__rt_stkovf_split_small} or @samp{__rt_stkovf_split_big} will be
5954 called, depending upon the amount of stack space required. The run time
5955 system is required to provide these functions. The default is
5956 @option{-mno-apcs-stack-check}, since this produces smaller code.
5958 @c not currently implemented
5960 @opindex mapcs-float
5961 Pass floating point arguments using the float point registers. This is
5962 one of the variants of the APCS@. This option is recommended if the
5963 target hardware has a floating point unit or if a lot of floating point
5964 arithmetic is going to be performed by the code. The default is
5965 @option{-mno-apcs-float}, since integer only code is slightly increased in
5966 size if @option{-mapcs-float} is used.
5968 @c not currently implemented
5969 @item -mapcs-reentrant
5970 @opindex mapcs-reentrant
5971 Generate reentrant, position independent code. The default is
5972 @option{-mno-apcs-reentrant}.
5975 @item -mthumb-interwork
5976 @opindex mthumb-interwork
5977 Generate code which supports calling between the ARM and Thumb
5978 instruction sets. Without this option the two instruction sets cannot
5979 be reliably used inside one program. The default is
5980 @option{-mno-thumb-interwork}, since slightly larger code is generated
5981 when @option{-mthumb-interwork} is specified.
5983 @item -mno-sched-prolog
5984 @opindex mno-sched-prolog
5985 Prevent the reordering of instructions in the function prolog, or the
5986 merging of those instruction with the instructions in the function's
5987 body. This means that all functions will start with a recognizable set
5988 of instructions (or in fact one of a choice from a small set of
5989 different function prologues), and this information can be used to
5990 locate the start if functions inside an executable piece of code. The
5991 default is @option{-msched-prolog}.
5994 @opindex mhard-float
5995 Generate output containing floating point instructions. This is the
5999 @opindex msoft-float
6000 Generate output containing library calls for floating point.
6001 @strong{Warning:} the requisite libraries are not available for all ARM
6002 targets. Normally the facilities of the machine's usual C compiler are
6003 used, but this cannot be done directly in cross-compilation. You must make
6004 your own arrangements to provide suitable library functions for
6007 @option{-msoft-float} changes the calling convention in the output file;
6008 therefore, it is only useful if you compile @emph{all} of a program with
6009 this option. In particular, you need to compile @file{libgcc.a}, the
6010 library that comes with GCC, with @option{-msoft-float} in order for
6013 @item -mlittle-endian
6014 @opindex mlittle-endian
6015 Generate code for a processor running in little-endian mode. This is
6016 the default for all standard configurations.
6019 @opindex mbig-endian
6020 Generate code for a processor running in big-endian mode; the default is
6021 to compile code for a little-endian processor.
6023 @item -mwords-little-endian
6024 @opindex mwords-little-endian
6025 This option only applies when generating code for big-endian processors.
6026 Generate code for a little-endian word order but a big-endian byte
6027 order. That is, a byte order of the form @samp{32107654}. Note: this
6028 option should only be used if you require compatibility with code for
6029 big-endian ARM processors generated by versions of the compiler prior to
6032 @item -malignment-traps
6033 @opindex malignment-traps
6034 Generate code that will not trap if the MMU has alignment traps enabled.
6035 On ARM architectures prior to ARMv4, there were no instructions to
6036 access half-word objects stored in memory. However, when reading from
6037 memory a feature of the ARM architecture allows a word load to be used,
6038 even if the address is unaligned, and the processor core will rotate the
6039 data as it is being loaded. This option tells the compiler that such
6040 misaligned accesses will cause a MMU trap and that it should instead
6041 synthesise the access as a series of byte accesses. The compiler can
6042 still use word accesses to load half-word data if it knows that the
6043 address is aligned to a word boundary.
6045 This option is ignored when compiling for ARM architecture 4 or later,
6046 since these processors have instructions to directly access half-word
6049 @item -mno-alignment-traps
6050 @opindex mno-alignment-traps
6051 Generate code that assumes that the MMU will not trap unaligned
6052 accesses. This produces better code when the target instruction set
6053 does not have half-word memory operations (i.e.@: implementations prior to
6056 Note that you cannot use this option to access unaligned word objects,
6057 since the processor will only fetch one 32-bit aligned object from
6060 The default setting for most targets is @option{-mno-alignment-traps}, since
6061 this produces better code when there are no half-word memory
6062 instructions available.
6064 @item -mshort-load-bytes
6065 @itemx -mno-short-load-words
6066 @opindex mshort-load-bytes
6067 @opindex mno-short-load-words
6068 These are deprecated aliases for @option{-malignment-traps}.
6070 @item -mno-short-load-bytes
6071 @itemx -mshort-load-words
6072 @opindex mno-short-load-bytes
6073 @opindex mshort-load-words
6074 This are deprecated aliases for @option{-mno-alignment-traps}.
6078 This option only applies to RISC iX@. Emulate the native BSD-mode
6079 compiler. This is the default if @option{-ansi} is not specified.
6083 This option only applies to RISC iX@. Emulate the native X/Open-mode
6086 @item -mno-symrename
6087 @opindex mno-symrename
6088 This option only applies to RISC iX@. Do not run the assembler
6089 post-processor, @samp{symrename}, after code has been assembled.
6090 Normally it is necessary to modify some of the standard symbols in
6091 preparation for linking with the RISC iX C library; this option
6092 suppresses this pass. The post-processor is never run when the
6093 compiler is built for cross-compilation.
6095 @item -mcpu=@var{name}
6097 This specifies the name of the target ARM processor. GCC uses this name
6098 to determine what kind of instructions it can emit when generating
6099 assembly code. Permissible names are: @samp{arm2}, @samp{arm250},
6100 @samp{arm3}, @samp{arm6}, @samp{arm60}, @samp{arm600}, @samp{arm610},
6101 @samp{arm620}, @samp{arm7}, @samp{arm7m}, @samp{arm7d}, @samp{arm7dm},
6102 @samp{arm7di}, @samp{arm7dmi}, @samp{arm70}, @samp{arm700},
6103 @samp{arm700i}, @samp{arm710}, @samp{arm710c}, @samp{arm7100},
6104 @samp{arm7500}, @samp{arm7500fe}, @samp{arm7tdmi}, @samp{arm8},
6105 @samp{strongarm}, @samp{strongarm110}, @samp{strongarm1100},
6106 @samp{arm8}, @samp{arm810}, @samp{arm9}, @samp{arm9e}, @samp{arm920},
6107 @samp{arm920t}, @samp{arm940t}, @samp{arm9tdmi}, @samp{arm10tdmi},
6108 @samp{arm1020t}, @samp{xscale}.
6110 @itemx -mtune=@var{name}
6112 This option is very similar to the @option{-mcpu=} option, except that
6113 instead of specifying the actual target processor type, and hence
6114 restricting which instructions can be used, it specifies that GCC should
6115 tune the performance of the code as if the target were of the type
6116 specified in this option, but still choosing the instructions that it
6117 will generate based on the cpu specified by a @option{-mcpu=} option.
6118 For some ARM implementations better performance can be obtained by using
6121 @item -march=@var{name}
6123 This specifies the name of the target ARM architecture. GCC uses this
6124 name to determine what kind of instructions it can emit when generating
6125 assembly code. This option can be used in conjunction with or instead
6126 of the @option{-mcpu=} option. Permissible names are: @samp{armv2},
6127 @samp{armv2a}, @samp{armv3}, @samp{armv3m}, @samp{armv4}, @samp{armv4t},
6128 @samp{armv5}, @samp{armv5t}, @samp{armv5te}.
6130 @item -mfpe=@var{number}
6131 @itemx -mfp=@var{number}
6134 This specifies the version of the floating point emulation available on
6135 the target. Permissible values are 2 and 3. @option{-mfp=} is a synonym
6136 for @option{-mfpe=}, for compatibility with older versions of GCC@.
6138 @item -mstructure-size-boundary=@var{n}
6139 @opindex mstructure-size-boundary
6140 The size of all structures and unions will be rounded up to a multiple
6141 of the number of bits set by this option. Permissible values are 8 and
6142 32. The default value varies for different toolchains. For the COFF
6143 targeted toolchain the default value is 8. Specifying the larger number
6144 can produce faster, more efficient code, but can also increase the size
6145 of the program. The two values are potentially incompatible. Code
6146 compiled with one value cannot necessarily expect to work with code or
6147 libraries compiled with the other value, if they exchange information
6148 using structures or unions.
6150 @item -mabort-on-noreturn
6151 @opindex mabort-on-noreturn
6152 Generate a call to the function @code{abort} at the end of a
6153 @code{noreturn} function. It will be executed if the function tries to
6157 @itemx -mno-long-calls
6158 @opindex mlong-calls
6159 @opindex mno-long-calls
6160 Tells the compiler to perform function calls by first loading the
6161 address of the function into a register and then performing a subroutine
6162 call on this register. This switch is needed if the target function
6163 will lie outside of the 64 megabyte addressing range of the offset based
6164 version of subroutine call instruction.
6166 Even if this switch is enabled, not all function calls will be turned
6167 into long calls. The heuristic is that static functions, functions
6168 which have the @samp{short-call} attribute, functions that are inside
6169 the scope of a @samp{#pragma no_long_calls} directive and functions whose
6170 definitions have already been compiled within the current compilation
6171 unit, will not be turned into long calls. The exception to this rule is
6172 that weak function definitions, functions with the @samp{long-call}
6173 attribute or the @samp{section} attribute, and functions that are within
6174 the scope of a @samp{#pragma long_calls} directive, will always be
6175 turned into long calls.
6177 This feature is not enabled by default. Specifying
6178 @option{-mno-long-calls} will restore the default behavior, as will
6179 placing the function calls within the scope of a @samp{#pragma
6180 long_calls_off} directive. Note these switches have no effect on how
6181 the compiler generates code to handle function calls via function
6184 @item -mnop-fun-dllimport
6185 @opindex mnop-fun-dllimport
6186 Disable support for the @code{dllimport} attribute.
6188 @item -msingle-pic-base
6189 @opindex msingle-pic-base
6190 Treat the register used for PIC addressing as read-only, rather than
6191 loading it in the prologue for each function. The run-time system is
6192 responsible for initializing this register with an appropriate value
6193 before execution begins.
6195 @item -mpic-register=@var{reg}
6196 @opindex mpic-register
6197 Specify the register to be used for PIC addressing. The default is R10
6198 unless stack-checking is enabled, when R9 is used.
6200 @item -mpoke-function-name
6201 @opindex mpoke-function-name
6202 Write the name of each function into the text section, directly
6203 preceding the function prologue. The generated code is similar to this:
6207 .ascii "arm_poke_function_name", 0
6210 .word 0xff000000 + (t1 - t0)
6211 arm_poke_function_name
6213 stmfd sp!, @{fp, ip, lr, pc@}
6217 When performing a stack backtrace, code can inspect the value of
6218 @code{pc} stored at @code{fp + 0}. If the trace function then looks at
6219 location @code{pc - 12} and the top 8 bits are set, then we know that
6220 there is a function name embedded immediately preceding this location
6221 and has length @code{((pc[-3]) & 0xff000000)}.
6225 Generate code for the 16-bit Thumb instruction set. The default is to
6226 use the 32-bit ARM instruction set.
6229 @opindex mtpcs-frame
6230 Generate a stack frame that is compliant with the Thumb Procedure Call
6231 Standard for all non-leaf functions. (A leaf function is one that does
6232 not call any other functions.) The default is @option{-mno-tpcs-frame}.
6234 @item -mtpcs-leaf-frame
6235 @opindex mtpcs-leaf-frame
6236 Generate a stack frame that is compliant with the Thumb Procedure Call
6237 Standard for all leaf functions. (A leaf function is one that does
6238 not call any other functions.) The default is @option{-mno-apcs-leaf-frame}.
6240 @item -mcallee-super-interworking
6241 @opindex mcallee-super-interworking
6242 Gives all externally visible functions in the file being compiled an ARM
6243 instruction set header which switches to Thumb mode before executing the
6244 rest of the function. This allows these functions to be called from
6245 non-interworking code.
6247 @item -mcaller-super-interworking
6248 @opindex mcaller-super-interworking
6249 Allows calls via function pointers (including virtual functions) to
6250 execute correctly regardless of whether the target code has been
6251 compiled for interworking or not. There is a small overhead in the cost
6252 of executing a function pointer if this option is enabled.
6256 @node MN10200 Options
6257 @subsection MN10200 Options
6258 @cindex MN10200 options
6259 These @option{-m} options are defined for Matsushita MN10200 architectures:
6264 Indicate to the linker that it should perform a relaxation optimization pass
6265 to shorten branches, calls and absolute memory addresses. This option only
6266 has an effect when used on the command line for the final link step.
6268 This option makes symbolic debugging impossible.
6271 @node MN10300 Options
6272 @subsection MN10300 Options
6273 @cindex MN10300 options
6274 These @option{-m} options are defined for Matsushita MN10300 architectures:
6279 Generate code to avoid bugs in the multiply instructions for the MN10300
6280 processors. This is the default.
6283 @opindex mno-mult-bug
6284 Do not generate code to avoid bugs in the multiply instructions for the
6289 Generate code which uses features specific to the AM33 processor.
6293 Do not generate code which uses features specific to the AM33 processor. This
6298 Do not link in the C run-time initialization object file.
6302 Indicate to the linker that it should perform a relaxation optimization pass
6303 to shorten branches, calls and absolute memory addresses. This option only
6304 has an effect when used on the command line for the final link step.
6306 This option makes symbolic debugging impossible.
6310 @node M32R/D Options
6311 @subsection M32R/D Options
6312 @cindex M32R/D options
6314 These @option{-m} options are defined for Mitsubishi M32R/D architectures:
6319 Generate code for the M32R/X@.
6323 Generate code for the M32R@. This is the default.
6325 @item -mcode-model=small
6326 @opindex mcode-model=small
6327 Assume all objects live in the lower 16MB of memory (so that their addresses
6328 can be loaded with the @code{ld24} instruction), and assume all subroutines
6329 are reachable with the @code{bl} instruction.
6330 This is the default.
6332 The addressability of a particular object can be set with the
6333 @code{model} attribute.
6335 @item -mcode-model=medium
6336 @opindex mcode-model=medium
6337 Assume objects may be anywhere in the 32-bit address space (the compiler
6338 will generate @code{seth/add3} instructions to load their addresses), and
6339 assume all subroutines are reachable with the @code{bl} instruction.
6341 @item -mcode-model=large
6342 @opindex mcode-model=large
6343 Assume objects may be anywhere in the 32-bit address space (the compiler
6344 will generate @code{seth/add3} instructions to load their addresses), and
6345 assume subroutines may not be reachable with the @code{bl} instruction
6346 (the compiler will generate the much slower @code{seth/add3/jl}
6347 instruction sequence).
6350 @opindex msdata=none
6351 Disable use of the small data area. Variables will be put into
6352 one of @samp{.data}, @samp{bss}, or @samp{.rodata} (unless the
6353 @code{section} attribute has been specified).
6354 This is the default.
6356 The small data area consists of sections @samp{.sdata} and @samp{.sbss}.
6357 Objects may be explicitly put in the small data area with the
6358 @code{section} attribute using one of these sections.
6361 @opindex msdata=sdata
6362 Put small global and static data in the small data area, but do not
6363 generate special code to reference them.
6367 Put small global and static data in the small data area, and generate
6368 special instructions to reference them.
6372 @cindex smaller data references
6373 Put global and static objects less than or equal to @var{num} bytes
6374 into the small data or bss sections instead of the normal data or bss
6375 sections. The default value of @var{num} is 8.
6376 The @option{-msdata} option must be set to one of @samp{sdata} or @samp{use}
6377 for this option to have any effect.
6379 All modules should be compiled with the same @option{-G @var{num}} value.
6380 Compiling with different values of @var{num} may or may not work; if it
6381 doesn't the linker will give an error message---incorrect code will not be
6387 @subsection M88K Options
6388 @cindex M88k options
6390 These @samp{-m} options are defined for Motorola 88k architectures:
6395 Generate code that works well on both the m88100 and the
6400 Generate code that works best for the m88100, but that also
6405 Generate code that works best for the m88110, and may not run
6410 Obsolete option to be removed from the next revision.
6413 @item -midentify-revision
6414 @opindex midentify-revision
6415 @cindex identifying source, compiler (88k)
6416 Include an @code{ident} directive in the assembler output recording the
6417 source file name, compiler name and version, timestamp, and compilation
6420 @item -mno-underscores
6421 @opindex mno-underscores
6422 @cindex underscores, avoiding (88k)
6423 In assembler output, emit symbol names without adding an underscore
6424 character at the beginning of each name. The default is to use an
6425 underscore as prefix on each name.
6427 @item -mocs-debug-info
6428 @itemx -mno-ocs-debug-info
6429 @opindex mocs-debug-info
6430 @opindex mno-ocs-debug-info
6432 @cindex debugging, 88k OCS
6433 Include (or omit) additional debugging information (about registers used
6434 in each stack frame) as specified in the 88open Object Compatibility
6435 Standard, ``OCS''@. This extra information allows debugging of code that
6436 has had the frame pointer eliminated. The default for DG/UX, SVr4, and
6437 Delta 88 SVr3.2 is to include this information; other 88k configurations
6438 omit this information by default.
6440 @item -mocs-frame-position
6441 @opindex mocs-frame-position
6442 @cindex register positions in frame (88k)
6443 When emitting COFF debugging information for automatic variables and
6444 parameters stored on the stack, use the offset from the canonical frame
6445 address, which is the stack pointer (register 31) on entry to the
6446 function. The DG/UX, SVr4, Delta88 SVr3.2, and BCS configurations use
6447 @option{-mocs-frame-position}; other 88k configurations have the default
6448 @option{-mno-ocs-frame-position}.
6450 @item -mno-ocs-frame-position
6451 @opindex mno-ocs-frame-position
6452 @cindex register positions in frame (88k)
6453 When emitting COFF debugging information for automatic variables and
6454 parameters stored on the stack, use the offset from the frame pointer
6455 register (register 30). When this option is in effect, the frame
6456 pointer is not eliminated when debugging information is selected by the
6459 @item -moptimize-arg-area
6460 @opindex moptimize-arg-area
6461 @cindex arguments in frame (88k)
6462 Save space by reorganizing the stack frame. This option generates code
6463 that does not agree with the 88open specifications, but uses less
6466 @itemx -mno-optimize-arg-area
6467 @opindex mno-optimize-arg-area
6468 Do not reorganize the stack frame to save space. This is the default.
6469 The generated conforms to the specification, but uses more memory.
6471 @item -mshort-data-@var{num}
6472 @opindex mshort-data
6473 @cindex smaller data references (88k)
6474 @cindex r0-relative references (88k)
6475 Generate smaller data references by making them relative to @code{r0},
6476 which allows loading a value using a single instruction (rather than the
6477 usual two). You control which data references are affected by
6478 specifying @var{num} with this option. For example, if you specify
6479 @option{-mshort-data-512}, then the data references affected are those
6480 involving displacements of less than 512 bytes.
6481 @option{-mshort-data-@var{num}} is not effective for @var{num} greater
6484 @item -mserialize-volatile
6485 @opindex mserialize-volatile
6486 @itemx -mno-serialize-volatile
6487 @opindex mno-serialize-volatile
6488 @cindex sequential consistency on 88k
6489 Do, or don't, generate code to guarantee sequential consistency
6490 of volatile memory references. By default, consistency is
6493 The order of memory references made by the MC88110 processor does
6494 not always match the order of the instructions requesting those
6495 references. In particular, a load instruction may execute before
6496 a preceding store instruction. Such reordering violates
6497 sequential consistency of volatile memory references, when there
6498 are multiple processors. When consistency must be guaranteed,
6499 GCC generates special instructions, as needed, to force
6500 execution in the proper order.
6502 The MC88100 processor does not reorder memory references and so
6503 always provides sequential consistency. However, by default, GCC
6504 generates the special instructions to guarantee consistency
6505 even when you use @option{-m88100}, so that the code may be run on an
6506 MC88110 processor. If you intend to run your code only on the
6507 MC88100 processor, you may use @option{-mno-serialize-volatile}.
6509 The extra code generated to guarantee consistency may affect the
6510 performance of your application. If you know that you can safely
6511 forgo this guarantee, you may use @option{-mno-serialize-volatile}.
6517 @cindex assembler syntax, 88k
6519 Turn on (@option{-msvr4}) or off (@option{-msvr3}) compiler extensions
6520 related to System V release 4 (SVr4). This controls the following:
6524 Which variant of the assembler syntax to emit.
6526 @option{-msvr4} makes the C preprocessor recognize @samp{#pragma weak}
6527 that is used on System V release 4.
6529 @option{-msvr4} makes GCC issue additional declaration directives used in
6533 @option{-msvr4} is the default for the m88k-motorola-sysv4 and
6534 m88k-dg-dgux m88k configurations. @option{-msvr3} is the default for all
6535 other m88k configurations.
6537 @item -mversion-03.00
6538 @opindex mversion-03.00
6539 This option is obsolete, and is ignored.
6540 @c ??? which asm syntax better for GAS? option there too?
6542 @item -mno-check-zero-division
6543 @itemx -mcheck-zero-division
6544 @opindex mno-check-zero-division
6545 @opindex mcheck-zero-division
6546 @cindex zero division on 88k
6547 Do, or don't, generate code to guarantee that integer division by
6548 zero will be detected. By default, detection is guaranteed.
6550 Some models of the MC88100 processor fail to trap upon integer
6551 division by zero under certain conditions. By default, when
6552 compiling code that might be run on such a processor, GCC
6553 generates code that explicitly checks for zero-valued divisors
6554 and traps with exception number 503 when one is detected. Use of
6555 @option{-mno-check-zero-division} suppresses such checking for code
6556 generated to run on an MC88100 processor.
6558 GCC assumes that the MC88110 processor correctly detects all instances
6559 of integer division by zero. When @option{-m88110} is specified, no
6560 explicit checks for zero-valued divisors are generated, and both
6561 @option{-mcheck-zero-division} and @option{-mno-check-zero-division} are
6564 @item -muse-div-instruction
6565 @opindex muse-div-instruction
6566 @cindex divide instruction, 88k
6567 Use the div instruction for signed integer division on the
6568 MC88100 processor. By default, the div instruction is not used.
6570 On the MC88100 processor the signed integer division instruction
6571 div) traps to the operating system on a negative operand. The
6572 operating system transparently completes the operation, but at a
6573 large cost in execution time. By default, when compiling code
6574 that might be run on an MC88100 processor, GCC emulates signed
6575 integer division using the unsigned integer division instruction
6576 divu), thereby avoiding the large penalty of a trap to the
6577 operating system. Such emulation has its own, smaller, execution
6578 cost in both time and space. To the extent that your code's
6579 important signed integer division operations are performed on two
6580 nonnegative operands, it may be desirable to use the div
6581 instruction directly.
6583 On the MC88110 processor the div instruction (also known as the
6584 divs instruction) processes negative operands without trapping to
6585 the operating system. When @option{-m88110} is specified,
6586 @option{-muse-div-instruction} is ignored, and the div instruction is used
6587 for signed integer division.
6589 Note that the result of dividing @code{INT_MIN} by @minus{}1 is undefined. In
6590 particular, the behavior of such a division with and without
6591 @option{-muse-div-instruction} may differ.
6593 @item -mtrap-large-shift
6594 @itemx -mhandle-large-shift
6595 @opindex mtrap-large-shift
6596 @opindex mhandle-large-shift
6597 @cindex bit shift overflow (88k)
6598 @cindex large bit shifts (88k)
6599 Include code to detect bit-shifts of more than 31 bits; respectively,
6600 trap such shifts or emit code to handle them properly. By default GCC
6601 makes no special provision for large bit shifts.
6603 @item -mwarn-passed-structs
6604 @opindex mwarn-passed-structs
6605 @cindex structure passing (88k)
6606 Warn when a function passes a struct as an argument or result.
6607 Structure-passing conventions have changed during the evolution of the C
6608 language, and are often the source of portability problems. By default,
6609 GCC issues no such warning.
6612 @c break page here to avoid unsightly interparagraph stretch.
6616 @node RS/6000 and PowerPC Options
6617 @subsection IBM RS/6000 and PowerPC Options
6618 @cindex RS/6000 and PowerPC Options
6619 @cindex IBM RS/6000 and PowerPC Options
6621 These @samp{-m} options are defined for the IBM RS/6000 and PowerPC:
6629 @itemx -mpowerpc-gpopt
6630 @itemx -mno-powerpc-gpopt
6631 @itemx -mpowerpc-gfxopt
6632 @itemx -mno-powerpc-gfxopt
6634 @itemx -mno-powerpc64
6640 @opindex mno-powerpc
6641 @opindex mpowerpc-gpopt
6642 @opindex mno-powerpc-gpopt
6643 @opindex mpowerpc-gfxopt
6644 @opindex mno-powerpc-gfxopt
6646 @opindex mno-powerpc64
6647 GCC supports two related instruction set architectures for the
6648 RS/6000 and PowerPC@. The @dfn{POWER} instruction set are those
6649 instructions supported by the @samp{rios} chip set used in the original
6650 RS/6000 systems and the @dfn{PowerPC} instruction set is the
6651 architecture of the Motorola MPC5xx, MPC6xx, MPC8xx microprocessors, and
6652 the IBM 4xx microprocessors.
6654 Neither architecture is a subset of the other. However there is a
6655 large common subset of instructions supported by both. An MQ
6656 register is included in processors supporting the POWER architecture.
6658 You use these options to specify which instructions are available on the
6659 processor you are using. The default value of these options is
6660 determined when configuring GCC@. Specifying the
6661 @option{-mcpu=@var{cpu_type}} overrides the specification of these
6662 options. We recommend you use the @option{-mcpu=@var{cpu_type}} option
6663 rather than the options listed above.
6665 The @option{-mpower} option allows GCC to generate instructions that
6666 are found only in the POWER architecture and to use the MQ register.
6667 Specifying @option{-mpower2} implies @option{-power} and also allows GCC
6668 to generate instructions that are present in the POWER2 architecture but
6669 not the original POWER architecture.
6671 The @option{-mpowerpc} option allows GCC to generate instructions that
6672 are found only in the 32-bit subset of the PowerPC architecture.
6673 Specifying @option{-mpowerpc-gpopt} implies @option{-mpowerpc} and also allows
6674 GCC to use the optional PowerPC architecture instructions in the
6675 General Purpose group, including floating-point square root. Specifying
6676 @option{-mpowerpc-gfxopt} implies @option{-mpowerpc} and also allows GCC to
6677 use the optional PowerPC architecture instructions in the Graphics
6678 group, including floating-point select.
6680 The @option{-mpowerpc64} option allows GCC to generate the additional
6681 64-bit instructions that are found in the full PowerPC64 architecture
6682 and to treat GPRs as 64-bit, doubleword quantities. GCC defaults to
6683 @option{-mno-powerpc64}.
6685 If you specify both @option{-mno-power} and @option{-mno-powerpc}, GCC
6686 will use only the instructions in the common subset of both
6687 architectures plus some special AIX common-mode calls, and will not use
6688 the MQ register. Specifying both @option{-mpower} and @option{-mpowerpc}
6689 permits GCC to use any instruction from either architecture and to
6690 allow use of the MQ register; specify this for the Motorola MPC601.
6692 @item -mnew-mnemonics
6693 @itemx -mold-mnemonics
6694 @opindex mnew-mnemonics
6695 @opindex mold-mnemonics
6696 Select which mnemonics to use in the generated assembler code. With
6697 @option{-mnew-mnemonics}, GCC uses the assembler mnemonics defined for
6698 the PowerPC architecture. With @option{-mold-mnemonics} it uses the
6699 assembler mnemonics defined for the POWER architecture. Instructions
6700 defined in only one architecture have only one mnemonic; GCC uses that
6701 mnemonic irrespective of which of these options is specified.
6703 GCC defaults to the mnemonics appropriate for the architecture in
6704 use. Specifying @option{-mcpu=@var{cpu_type}} sometimes overrides the
6705 value of these option. Unless you are building a cross-compiler, you
6706 should normally not specify either @option{-mnew-mnemonics} or
6707 @option{-mold-mnemonics}, but should instead accept the default.
6709 @item -mcpu=@var{cpu_type}
6711 Set architecture type, register usage, choice of mnemonics, and
6712 instruction scheduling parameters for machine type @var{cpu_type}.
6713 Supported values for @var{cpu_type} are @samp{rios}, @samp{rios1},
6714 @samp{rsc}, @samp{rios2}, @samp{rs64a}, @samp{601}, @samp{602},
6715 @samp{603}, @samp{603e}, @samp{604}, @samp{604e}, @samp{620},
6716 @samp{630}, @samp{740}, @samp{7400}, @samp{7450}, @samp{750},
6717 @samp{power}, @samp{power2}, @samp{powerpc}, @samp{403}, @samp{505},
6718 @samp{801}, @samp{821}, @samp{823}, and @samp{860} and @samp{common}.
6720 @option{-mcpu=common} selects a completely generic processor. Code
6721 generated under this option will run on any POWER or PowerPC processor.
6722 GCC will use only the instructions in the common subset of both
6723 architectures, and will not use the MQ register. GCC assumes a generic
6724 processor model for scheduling purposes.
6726 @option{-mcpu=power}, @option{-mcpu=power2}, @option{-mcpu=powerpc}, and
6727 @option{-mcpu=powerpc64} specify generic POWER, POWER2, pure 32-bit
6728 PowerPC (i.e., not MPC601), and 64-bit PowerPC architecture machine
6729 types, with an appropriate, generic processor model assumed for
6730 scheduling purposes.
6732 The other options specify a specific processor. Code generated under
6733 those options will run best on that processor, and may not run at all on
6736 The @option{-mcpu} options automatically enable or disable other
6737 @option{-m} options as follows:
6741 @option{-mno-power}, @option{-mno-powerc}
6748 @option{-mpower}, @option{-mno-powerpc}, @option{-mno-new-mnemonics}
6763 @option{-mno-power}, @option{-mpowerpc}, @option{-mnew-mnemonics}
6766 @option{-mpower}, @option{-mpowerpc}, @option{-mnew-mnemonics}
6771 @option{-mno-power}, @option{-mpowerpc}, @option{-mnew-mnemonics}, @option{-msoft-float}
6774 @item -mtune=@var{cpu_type}
6776 Set the instruction scheduling parameters for machine type
6777 @var{cpu_type}, but do not set the architecture type, register usage, or
6778 choice of mnemonics, as @option{-mcpu=@var{cpu_type}} would. The same
6779 values for @var{cpu_type} are used for @option{-mtune} as for
6780 @option{-mcpu}. If both are specified, the code generated will use the
6781 architecture, registers, and mnemonics set by @option{-mcpu}, but the
6782 scheduling parameters set by @option{-mtune}.
6787 @opindex mno-altivec
6788 These switches enable or disable the use of built-in functions that
6789 allow access to the AltiVec instruction set. You may also need to set
6790 @option{-mabi=altivec} to adjust the current ABI with AltiVec ABI
6794 @itemx -mno-fp-in-toc
6795 @itemx -mno-sum-in-toc
6796 @itemx -mminimal-toc
6798 @opindex mno-fp-in-toc
6799 @opindex mno-sum-in-toc
6800 @opindex mminimal-toc
6801 Modify generation of the TOC (Table Of Contents), which is created for
6802 every executable file. The @option{-mfull-toc} option is selected by
6803 default. In that case, GCC will allocate at least one TOC entry for
6804 each unique non-automatic variable reference in your program. GCC
6805 will also place floating-point constants in the TOC@. However, only
6806 16,384 entries are available in the TOC@.
6808 If you receive a linker error message that saying you have overflowed
6809 the available TOC space, you can reduce the amount of TOC space used
6810 with the @option{-mno-fp-in-toc} and @option{-mno-sum-in-toc} options.
6811 @option{-mno-fp-in-toc} prevents GCC from putting floating-point
6812 constants in the TOC and @option{-mno-sum-in-toc} forces GCC to
6813 generate code to calculate the sum of an address and a constant at
6814 run-time instead of putting that sum into the TOC@. You may specify one
6815 or both of these options. Each causes GCC to produce very slightly
6816 slower and larger code at the expense of conserving TOC space.
6818 If you still run out of space in the TOC even when you specify both of
6819 these options, specify @option{-mminimal-toc} instead. This option causes
6820 GCC to make only one TOC entry for every file. When you specify this
6821 option, GCC will produce code that is slower and larger but which
6822 uses extremely little TOC space. You may wish to use this option
6823 only on files that contain less frequently executed code.
6829 Enable 64-bit AIX ABI and calling convention: 64-bit pointers, 64-bit
6830 @code{long} type, and the infrastructure needed to support them.
6831 Specifying @option{-maix64} implies @option{-mpowerpc64} and
6832 @option{-mpowerpc}, while @option{-maix32} disables the 64-bit ABI and
6833 implies @option{-mno-powerpc64}. GCC defaults to @option{-maix32}.
6838 @opindex mno-xl-call
6839 On AIX, pass floating-point arguments to prototyped functions beyond the
6840 register save area (RSA) on the stack in addition to argument FPRs. The
6841 AIX calling convention was extended but not initially documented to
6842 handle an obscure K&R C case of calling a function that takes the
6843 address of its arguments with fewer arguments than declared. AIX XL
6844 compilers access floating point arguments which do not fit in the
6845 RSA from the stack when a subroutine is compiled without
6846 optimization. Because always storing floating-point arguments on the
6847 stack is inefficient and rarely needed, this option is not enabled by
6848 default and only is necessary when calling subroutines compiled by AIX
6849 XL compilers without optimization.
6853 Support @dfn{IBM RS/6000 SP} @dfn{Parallel Environment} (PE)@. Link an
6854 application written to use message passing with special startup code to
6855 enable the application to run. The system must have PE installed in the
6856 standard location (@file{/usr/lpp/ppe.poe/}), or the @file{specs} file
6857 must be overridden with the @option{-specs=} option to specify the
6858 appropriate directory location. The Parallel Environment does not
6859 support threads, so the @option{-mpe} option and the @option{-pthread}
6860 option are incompatible.
6864 @opindex msoft-float
6865 @opindex mhard-float
6866 Generate code that does not use (uses) the floating-point register set.
6867 Software floating point emulation is provided if you use the
6868 @option{-msoft-float} option, and pass the option to GCC when linking.
6871 @itemx -mno-multiple
6873 @opindex mno-multiple
6874 Generate code that uses (does not use) the load multiple word
6875 instructions and the store multiple word instructions. These
6876 instructions are generated by default on POWER systems, and not
6877 generated on PowerPC systems. Do not use @option{-mmultiple} on little
6878 endian PowerPC systems, since those instructions do not work when the
6879 processor is in little endian mode. The exceptions are PPC740 and
6880 PPC750 which permit the instructions usage in little endian mode.
6886 Generate code that uses (does not use) the load string instructions
6887 and the store string word instructions to save multiple registers and
6888 do small block moves. These instructions are generated by default on
6889 POWER systems, and not generated on PowerPC systems. Do not use
6890 @option{-mstring} on little endian PowerPC systems, since those
6891 instructions do not work when the processor is in little endian mode.
6892 The exceptions are PPC740 and PPC750 which permit the instructions
6893 usage in little endian mode.
6899 Generate code that uses (does not use) the load or store instructions
6900 that update the base register to the address of the calculated memory
6901 location. These instructions are generated by default. If you use
6902 @option{-mno-update}, there is a small window between the time that the
6903 stack pointer is updated and the address of the previous frame is
6904 stored, which means code that walks the stack frame across interrupts or
6905 signals may get corrupted data.
6908 @itemx -mno-fused-madd
6909 @opindex mfused-madd
6910 @opindex mno-fused-madd
6911 Generate code that uses (does not use) the floating point multiply and
6912 accumulate instructions. These instructions are generated by default if
6913 hardware floating is used.
6915 @item -mno-bit-align
6917 @opindex mno-bit-align
6919 On System V.4 and embedded PowerPC systems do not (do) force structures
6920 and unions that contain bit-fields to be aligned to the base type of the
6923 For example, by default a structure containing nothing but 8
6924 @code{unsigned} bit-fields of length 1 would be aligned to a 4 byte
6925 boundary and have a size of 4 bytes. By using @option{-mno-bit-align},
6926 the structure would be aligned to a 1 byte boundary and be one byte in
6929 @item -mno-strict-align
6930 @itemx -mstrict-align
6931 @opindex mno-strict-align
6932 @opindex mstrict-align
6933 On System V.4 and embedded PowerPC systems do not (do) assume that
6934 unaligned memory references will be handled by the system.
6937 @itemx -mno-relocatable
6938 @opindex mrelocatable
6939 @opindex mno-relocatable
6940 On embedded PowerPC systems generate code that allows (does not allow)
6941 the program to be relocated to a different address at runtime. If you
6942 use @option{-mrelocatable} on any module, all objects linked together must
6943 be compiled with @option{-mrelocatable} or @option{-mrelocatable-lib}.
6945 @item -mrelocatable-lib
6946 @itemx -mno-relocatable-lib
6947 @opindex mrelocatable-lib
6948 @opindex mno-relocatable-lib
6949 On embedded PowerPC systems generate code that allows (does not allow)
6950 the program to be relocated to a different address at runtime. Modules
6951 compiled with @option{-mrelocatable-lib} can be linked with either modules
6952 compiled without @option{-mrelocatable} and @option{-mrelocatable-lib} or
6953 with modules compiled with the @option{-mrelocatable} options.
6959 On System V.4 and embedded PowerPC systems do not (do) assume that
6960 register 2 contains a pointer to a global area pointing to the addresses
6961 used in the program.
6964 @itemx -mlittle-endian
6966 @opindex mlittle-endian
6967 On System V.4 and embedded PowerPC systems compile code for the
6968 processor in little endian mode. The @option{-mlittle-endian} option is
6969 the same as @option{-mlittle}.
6974 @opindex mbig-endian
6975 On System V.4 and embedded PowerPC systems compile code for the
6976 processor in big endian mode. The @option{-mbig-endian} option is
6977 the same as @option{-mbig}.
6981 On System V.4 and embedded PowerPC systems compile code using calling
6982 conventions that adheres to the March 1995 draft of the System V
6983 Application Binary Interface, PowerPC processor supplement. This is the
6984 default unless you configured GCC using @samp{powerpc-*-eabiaix}.
6986 @item -mcall-sysv-eabi
6987 @opindex mcall-sysv-eabi
6988 Specify both @option{-mcall-sysv} and @option{-meabi} options.
6990 @item -mcall-sysv-noeabi
6991 @opindex mcall-sysv-noeabi
6992 Specify both @option{-mcall-sysv} and @option{-mno-eabi} options.
6996 On System V.4 and embedded PowerPC systems compile code using calling
6997 conventions that are similar to those used on AIX@. This is the
6998 default if you configured GCC using @samp{powerpc-*-eabiaix}.
7000 @item -mcall-solaris
7001 @opindex mcall-solaris
7002 On System V.4 and embedded PowerPC systems compile code for the Solaris
7006 @opindex mcall-linux
7007 On System V.4 and embedded PowerPC systems compile code for the
7008 Linux-based GNU system.
7012 On System V.4 and embedded PowerPC systems compile code for the
7013 Hurd-based GNU system.
7016 @opindex mcall-netbsd
7017 On System V.4 and embedded PowerPC systems compile code for the
7018 NetBSD operating system.
7020 @item -maix-struct-return
7021 @opindex maix-struct-return
7022 Return all structures in memory (as specified by the AIX ABI)@.
7024 @item -msvr4-struct-return
7025 @opindex msvr4-struct-return
7026 Return structures smaller than 8 bytes in registers (as specified by the
7030 @opindex mabi=altivec
7031 Extend the current ABI with AltiVec ABI extensions. This does not
7032 change the default ABI, instead it adds the AltiVec ABI extensions to
7036 @itemx -mno-prototype
7038 @opindex mno-prototype
7039 On System V.4 and embedded PowerPC systems assume that all calls to
7040 variable argument functions are properly prototyped. Otherwise, the
7041 compiler must insert an instruction before every non prototyped call to
7042 set or clear bit 6 of the condition code register (@var{CR}) to
7043 indicate whether floating point values were passed in the floating point
7044 registers in case the function takes a variable arguments. With
7045 @option{-mprototype}, only calls to prototyped variable argument functions
7046 will set or clear the bit.
7050 On embedded PowerPC systems, assume that the startup module is called
7051 @file{sim-crt0.o} and that the standard C libraries are @file{libsim.a} and
7052 @file{libc.a}. This is the default for @samp{powerpc-*-eabisim}.
7057 On embedded PowerPC systems, assume that the startup module is called
7058 @file{crt0.o} and the standard C libraries are @file{libmvme.a} and
7063 On embedded PowerPC systems, assume that the startup module is called
7064 @file{crt0.o} and the standard C libraries are @file{libads.a} and
7068 @opindex myellowknife
7069 On embedded PowerPC systems, assume that the startup module is called
7070 @file{crt0.o} and the standard C libraries are @file{libyk.a} and
7075 On System V.4 and embedded PowerPC systems, specify that you are
7076 compiling for a VxWorks system.
7080 On embedded PowerPC systems, set the @var{PPC_EMB} bit in the ELF flags
7081 header to indicate that @samp{eabi} extended relocations are used.
7087 On System V.4 and embedded PowerPC systems do (do not) adhere to the
7088 Embedded Applications Binary Interface (eabi) which is a set of
7089 modifications to the System V.4 specifications. Selecting @option{-meabi}
7090 means that the stack is aligned to an 8 byte boundary, a function
7091 @code{__eabi} is called to from @code{main} to set up the eabi
7092 environment, and the @option{-msdata} option can use both @code{r2} and
7093 @code{r13} to point to two separate small data areas. Selecting
7094 @option{-mno-eabi} means that the stack is aligned to a 16 byte boundary,
7095 do not call an initialization function from @code{main}, and the
7096 @option{-msdata} option will only use @code{r13} to point to a single
7097 small data area. The @option{-meabi} option is on by default if you
7098 configured GCC using one of the @samp{powerpc*-*-eabi*} options.
7101 @opindex msdata=eabi
7102 On System V.4 and embedded PowerPC systems, put small initialized
7103 @code{const} global and static data in the @samp{.sdata2} section, which
7104 is pointed to by register @code{r2}. Put small initialized
7105 non-@code{const} global and static data in the @samp{.sdata} section,
7106 which is pointed to by register @code{r13}. Put small uninitialized
7107 global and static data in the @samp{.sbss} section, which is adjacent to
7108 the @samp{.sdata} section. The @option{-msdata=eabi} option is
7109 incompatible with the @option{-mrelocatable} option. The
7110 @option{-msdata=eabi} option also sets the @option{-memb} option.
7113 @opindex msdata=sysv
7114 On System V.4 and embedded PowerPC systems, put small global and static
7115 data in the @samp{.sdata} section, which is pointed to by register
7116 @code{r13}. Put small uninitialized global and static data in the
7117 @samp{.sbss} section, which is adjacent to the @samp{.sdata} section.
7118 The @option{-msdata=sysv} option is incompatible with the
7119 @option{-mrelocatable} option.
7121 @item -msdata=default
7123 @opindex msdata=default
7125 On System V.4 and embedded PowerPC systems, if @option{-meabi} is used,
7126 compile code the same as @option{-msdata=eabi}, otherwise compile code the
7127 same as @option{-msdata=sysv}.
7130 @opindex msdata-data
7131 On System V.4 and embedded PowerPC systems, put small global and static
7132 data in the @samp{.sdata} section. Put small uninitialized global and
7133 static data in the @samp{.sbss} section. Do not use register @code{r13}
7134 to address small data however. This is the default behavior unless
7135 other @option{-msdata} options are used.
7139 @opindex msdata=none
7141 On embedded PowerPC systems, put all initialized global and static data
7142 in the @samp{.data} section, and all uninitialized data in the
7143 @samp{.bss} section.
7147 @cindex smaller data references (PowerPC)
7148 @cindex .sdata/.sdata2 references (PowerPC)
7149 On embedded PowerPC systems, put global and static items less than or
7150 equal to @var{num} bytes into the small data or bss sections instead of
7151 the normal data or bss section. By default, @var{num} is 8. The
7152 @option{-G @var{num}} switch is also passed to the linker.
7153 All modules should be compiled with the same @option{-G @var{num}} value.
7156 @itemx -mno-regnames
7158 @opindex mno-regnames
7159 On System V.4 and embedded PowerPC systems do (do not) emit register
7160 names in the assembly language output using symbolic forms.
7164 Adds support for multithreading with the @dfn{pthreads} library.
7165 This option sets flags for both the preprocessor and linker.
7170 @subsection IBM RT Options
7172 @cindex IBM RT options
7174 These @samp{-m} options are defined for the IBM RT PC:
7178 @opindex min-line-mul
7179 Use an in-line code sequence for integer multiplies. This is the
7182 @item -mcall-lib-mul
7183 @opindex mcall-lib-mul
7184 Call @code{lmul$$} for integer multiples.
7186 @item -mfull-fp-blocks
7187 @opindex mfull-fp-blocks
7188 Generate full-size floating point data blocks, including the minimum
7189 amount of scratch space recommended by IBM@. This is the default.
7191 @item -mminimum-fp-blocks
7192 @opindex mminimum-fp-blocks
7193 Do not include extra scratch space in floating point data blocks. This
7194 results in smaller code, but slower execution, since scratch space must
7195 be allocated dynamically.
7197 @cindex @file{varargs.h} and RT PC
7198 @cindex @file{stdarg.h} and RT PC
7199 @item -mfp-arg-in-fpregs
7200 @opindex mfp-arg-in-fpregs
7201 Use a calling sequence incompatible with the IBM calling convention in
7202 which floating point arguments are passed in floating point registers.
7203 Note that @code{varargs.h} and @code{stdarg.h} will not work with
7204 floating point operands if this option is specified.
7206 @item -mfp-arg-in-gregs
7207 @opindex mfp-arg-in-gregs
7208 Use the normal calling convention for floating point arguments. This is
7211 @item -mhc-struct-return
7212 @opindex mhc-struct-return
7213 Return structures of more than one word in memory, rather than in a
7214 register. This provides compatibility with the MetaWare HighC (hc)
7215 compiler. Use the option @option{-fpcc-struct-return} for compatibility
7216 with the Portable C Compiler (pcc).
7218 @item -mnohc-struct-return
7219 @opindex mnohc-struct-return
7220 Return some structures of more than one word in registers, when
7221 convenient. This is the default. For compatibility with the
7222 IBM-supplied compilers, use the option @option{-fpcc-struct-return} or the
7223 option @option{-mhc-struct-return}.
7227 @subsection MIPS Options
7228 @cindex MIPS options
7230 These @samp{-m} options are defined for the MIPS family of computers:
7234 @item -march=@var{cpu-type}
7236 Assume the defaults for the machine type @var{cpu-type} when generating
7237 instructions. The choices for @var{cpu-type} are @samp{r2000}, @samp{r3000},
7238 @samp{r3900}, @samp{r4000}, @samp{r4100}, @samp{r4300}, @samp{r4400},
7239 @samp{r4600}, @samp{r4650}, @samp{r5000}, @samp{r6000}, @samp{r8000},
7240 and @samp{orion}. Additionally, the @samp{r2000}, @samp{r3000},
7241 @samp{r4000}, @samp{r5000}, and @samp{r6000} can be abbreviated as
7242 @samp{r2k} (or @samp{r2K}), @samp{r3k}, etc.
7244 @item -mtune=@var{cpu-type}
7246 Assume the defaults for the machine type @var{cpu-type} when scheduling
7247 instructions. The choices for @var{cpu-type} are @samp{r2000}, @samp{r3000},
7248 @samp{r3900}, @samp{r4000}, @samp{r4100}, @samp{r4300}, @samp{r4400},
7249 @samp{r4600}, @samp{r4650}, @samp{r5000}, @samp{r6000}, @samp{r8000},
7250 and @samp{orion}. Additionally, the @samp{r2000}, @samp{r3000},
7251 @samp{r4000}, @samp{r5000}, and @samp{r6000} can be abbreviated as
7252 @samp{r2k} (or @samp{r2K}), @samp{r3k}, etc. While picking a specific
7253 @var{cpu-type} will schedule things appropriately for that particular
7254 chip, the compiler will not generate any code that does not meet level 1
7255 of the MIPS ISA (instruction set architecture) without a @option{-mipsX}
7256 or @option{-mabi} switch being used.
7258 @item -mcpu=@var{cpu-type}
7260 This is identical to specifying both @option{-march} and @option{-mtune}.
7264 Issue instructions from level 1 of the MIPS ISA@. This is the default.
7265 @samp{r3000} is the default @var{cpu-type} at this ISA level.
7269 Issue instructions from level 2 of the MIPS ISA (branch likely, square
7270 root instructions). @samp{r6000} is the default @var{cpu-type} at this
7275 Issue instructions from level 3 of the MIPS ISA (64-bit instructions).
7276 @samp{r4000} is the default @var{cpu-type} at this ISA level.
7280 Issue instructions from level 4 of the MIPS ISA (conditional move,
7281 prefetch, enhanced FPU instructions). @samp{r8000} is the default
7282 @var{cpu-type} at this ISA level.
7286 Assume that 32 32-bit floating point registers are available. This is
7291 Assume that 32 64-bit floating point registers are available. This is
7292 the default when the @option{-mips3} option is used.
7295 @itemx -mno-fused-madd
7296 @opindex mfused-madd
7297 @opindex mno-fused-madd
7298 Generate code that uses (does not use) the floating point multiply and
7299 accumulate instructions, when they are available. These instructions
7300 are generated by default if they are available, but this may be
7301 undesirable if the extra precision causes problems or on certain chips
7302 in the mode where denormals are rounded to zero where denormals
7303 generated by multiply and accumulate instructions cause exceptions
7308 Assume that 32 32-bit general purpose registers are available. This is
7313 Assume that 32 64-bit general purpose registers are available. This is
7314 the default when the @option{-mips3} option is used.
7318 Force int and long types to be 64 bits wide. See @option{-mlong32} for an
7319 explanation of the default, and the width of pointers.
7323 Force long types to be 64 bits wide. See @option{-mlong32} for an
7324 explanation of the default, and the width of pointers.
7328 Force long, int, and pointer types to be 32 bits wide.
7330 If none of @option{-mlong32}, @option{-mlong64}, or @option{-mint64} are set,
7331 the size of ints, longs, and pointers depends on the ABI and ISA chosen.
7332 For @option{-mabi=32}, and @option{-mabi=n32}, ints and longs are 32 bits
7333 wide. For @option{-mabi=64}, ints are 32 bits, and longs are 64 bits wide.
7334 For @option{-mabi=eabi} and either @option{-mips1} or @option{-mips2}, ints
7335 and longs are 32 bits wide. For @option{-mabi=eabi} and higher ISAs, ints
7336 are 32 bits, and longs are 64 bits wide. The width of pointer types is
7337 the smaller of the width of longs or the width of general purpose
7338 registers (which in turn depends on the ISA)@.
7350 Generate code for the indicated ABI@. The default instruction level is
7351 @option{-mips1} for @samp{32}, @option{-mips3} for @samp{n32}, and
7352 @option{-mips4} otherwise. Conversely, with @option{-mips1} or
7353 @option{-mips2}, the default ABI is @samp{32}; otherwise, the default ABI
7358 Generate code for the MIPS assembler, and invoke @file{mips-tfile} to
7359 add normal debug information. This is the default for all
7360 platforms except for the OSF/1 reference platform, using the OSF/rose
7361 object format. If the either of the @option{-gstabs} or @option{-gstabs+}
7362 switches are used, the @file{mips-tfile} program will encapsulate the
7363 stabs within MIPS ECOFF@.
7367 Generate code for the GNU assembler. This is the default on the OSF/1
7368 reference platform, using the OSF/rose object format. Also, this is
7369 the default if the configure option @option{--with-gnu-as} is used.
7371 @item -msplit-addresses
7372 @itemx -mno-split-addresses
7373 @opindex msplit-addresses
7374 @opindex mno-split-addresses
7375 Generate code to load the high and low parts of address constants separately.
7376 This allows GCC to optimize away redundant loads of the high order
7377 bits of addresses. This optimization requires GNU as and GNU ld.
7378 This optimization is enabled by default for some embedded targets where
7379 GNU as and GNU ld are standard.
7385 The @option{-mrnames} switch says to output code using the MIPS software
7386 names for the registers, instead of the hardware names (ie, @var{a0}
7387 instead of @var{$4}). The only known assembler that supports this option
7388 is the Algorithmics assembler.
7394 The @option{-mgpopt} switch says to write all of the data declarations
7395 before the instructions in the text section, this allows the MIPS
7396 assembler to generate one word memory references instead of using two
7397 words for short global or static data items. This is on by default if
7398 optimization is selected.
7404 For each non-inline function processed, the @option{-mstats} switch
7405 causes the compiler to emit one line to the standard error file to
7406 print statistics about the program (number of registers saved, stack
7413 The @option{-mmemcpy} switch makes all block moves call the appropriate
7414 string function (@samp{memcpy} or @samp{bcopy}) instead of possibly
7415 generating inline code.
7418 @itemx -mno-mips-tfile
7419 @opindex mmips-tfile
7420 @opindex mno-mips-tfile
7421 The @option{-mno-mips-tfile} switch causes the compiler not
7422 postprocess the object file with the @file{mips-tfile} program,
7423 after the MIPS assembler has generated it to add debug support. If
7424 @file{mips-tfile} is not run, then no local variables will be
7425 available to the debugger. In addition, @file{stage2} and
7426 @file{stage3} objects will have the temporary file names passed to the
7427 assembler embedded in the object file, which means the objects will
7428 not compare the same. The @option{-mno-mips-tfile} switch should only
7429 be used when there are bugs in the @file{mips-tfile} program that
7430 prevents compilation.
7433 @opindex msoft-float
7434 Generate output containing library calls for floating point.
7435 @strong{Warning:} the requisite libraries are not part of GCC@.
7436 Normally the facilities of the machine's usual C compiler are used, but
7437 this can't be done directly in cross-compilation. You must make your
7438 own arrangements to provide suitable library functions for
7442 @opindex mhard-float
7443 Generate output containing floating point instructions. This is the
7444 default if you use the unmodified sources.
7447 @itemx -mno-abicalls
7449 @opindex mno-abicalls
7450 Emit (or do not emit) the pseudo operations @samp{.abicalls},
7451 @samp{.cpload}, and @samp{.cprestore} that some System V.4 ports use for
7452 position independent code.
7455 @itemx -mno-long-calls
7456 @opindex mlong-calls
7457 @opindex mno-long-calls
7458 Do all calls with the @samp{JALR} instruction, which requires
7459 loading up a function's address into a register before the call.
7460 You need to use this switch, if you call outside of the current
7461 512 megabyte segment to functions that are not through pointers.
7464 @itemx -mno-half-pic
7466 @opindex mno-half-pic
7467 Put pointers to extern references into the data section and load them
7468 up, rather than put the references in the text section.
7470 @item -membedded-pic
7471 @itemx -mno-embedded-pic
7472 @opindex membedded-pic
7473 @opindex mno-embedded-pic
7474 Generate PIC code suitable for some embedded systems. All calls are
7475 made using PC relative address, and all data is addressed using the $gp
7476 register. No more than 65536 bytes of global data may be used. This
7477 requires GNU as and GNU ld which do most of the work. This currently
7478 only works on targets which use ECOFF; it does not work with ELF@.
7480 @item -membedded-data
7481 @itemx -mno-embedded-data
7482 @opindex membedded-data
7483 @opindex mno-embedded-data
7484 Allocate variables to the read-only data section first if possible, then
7485 next in the small data section if possible, otherwise in data. This gives
7486 slightly slower code than the default, but reduces the amount of RAM required
7487 when executing, and thus may be preferred for some embedded systems.
7489 @item -muninit-const-in-rodata
7490 @itemx -mno-uninit-const-in-rodata
7491 @opindex muninit-const-in-rodata
7492 @opindex mno-uninit-const-in-rodata
7493 When used together with @option{-membedded-data}, it will always store uninitialized
7494 const variables in the read-only data section.
7496 @item -msingle-float
7497 @itemx -mdouble-float
7498 @opindex msingle-float
7499 @opindex mdouble-float
7500 The @option{-msingle-float} switch tells gcc to assume that the floating
7501 point coprocessor only supports single precision operations, as on the
7502 @samp{r4650} chip. The @option{-mdouble-float} switch permits gcc to use
7503 double precision operations. This is the default.
7509 Permit use of the @samp{mad}, @samp{madu} and @samp{mul} instructions,
7510 as on the @samp{r4650} chip.
7514 Turns on @option{-msingle-float}, @option{-mmad}, and, at least for now,
7515 @option{-mcpu=r4650}.
7521 Enable 16-bit instructions.
7525 Use the entry and exit pseudo ops. This option can only be used with
7530 Compile code for the processor in little endian mode.
7531 The requisite libraries are assumed to exist.
7535 Compile code for the processor in big endian mode.
7536 The requisite libraries are assumed to exist.
7540 @cindex smaller data references (MIPS)
7541 @cindex gp-relative references (MIPS)
7542 Put global and static items less than or equal to @var{num} bytes into
7543 the small data or bss sections instead of the normal data or bss
7544 section. This allows the assembler to emit one word memory reference
7545 instructions based on the global pointer (@var{gp} or @var{$28}),
7546 instead of the normal two words used. By default, @var{num} is 8 when
7547 the MIPS assembler is used, and 0 when the GNU assembler is used. The
7548 @option{-G @var{num}} switch is also passed to the assembler and linker.
7549 All modules should be compiled with the same @option{-G @var{num}}
7554 Tell the MIPS assembler to not run its preprocessor over user
7555 assembler files (with a @samp{.s} suffix) when assembling them.
7559 Pass an option to gas which will cause nops to be inserted if
7560 the read of the destination register of an mfhi or mflo instruction
7561 occurs in the following two instructions.
7565 Do not include the default crt0.
7567 @item -mflush-func=@var{func}
7568 @itemx -mno-flush-func
7569 @opindex mflush-func
7570 Specifies the function to call to flush the I and D caches, or to not
7571 call any such function. If called, the function must take the same
7572 arguments as the common @code{_flush_func()}, that is, the address of the
7573 memory range for which the cache is being flushed, the size of the
7574 memory range, and the number 3 (to flush both caches). The default
7575 depends on the target gcc was configured for, but commonly is either
7576 @samp{_flush_func} or @samp{__cpu_flush}.
7579 These options are defined by the macro
7580 @code{TARGET_SWITCHES} in the machine description. The default for the
7581 options is also defined by that macro, which enables you to change the
7584 @node i386 and x86-64 Options
7585 @subsection Intel 386 and AMD x86-64 Options
7586 @cindex i386 Options
7587 @cindex x86-64 Options
7588 @cindex Intel 386 Options
7589 @cindex AMD x86-64 Options
7591 These @samp{-m} options are defined for the i386 and x86-64 family of
7595 @item -mcpu=@var{cpu-type}
7597 Tune to @var{cpu-type} everything applicable about the generated code, except
7598 for the ABI and the set of available instructions. The choices for
7599 @var{cpu-type} are @samp{i386}, @samp{i486}, @samp{i586}, @samp{i686},
7600 @samp{pentium}, @samp{pentium-mmx}, @samp{pentiumpro}, @samp{pentium2},
7601 @samp{pentium3}, @samp{pentium4}, @samp{k6}, @samp{k6-2}, @samp{k6-3},
7602 @samp{athlon}, @samp{athlon-tbird}, @samp{athlon-4}, @samp{athlon-xp}
7603 and @samp{athlon-mp}.
7605 While picking a specific @var{cpu-type} will schedule things appropriately
7606 for that particular chip, the compiler will not generate any code that
7607 does not run on the i386 without the @option{-march=@var{cpu-type}} option
7608 being used. @samp{i586} is equivalent to @samp{pentium} and @samp{i686}
7609 is equivalent to @samp{pentiumpro}. @samp{k6} and @samp{athlon} are the
7610 AMD chips as opposed to the Intel ones.
7612 @item -march=@var{cpu-type}
7614 Generate instructions for the machine type @var{cpu-type}. The choices
7615 for @var{cpu-type} are the same as for @option{-mcpu}. Moreover,
7616 specifying @option{-march=@var{cpu-type}} implies @option{-mcpu=@var{cpu-type}}.
7625 @opindex mpentiumpro
7626 These options are synonyms for @option{-mcpu=i386}, @option{-mcpu=i486},
7627 @option{-mcpu=pentium}, and @option{-mcpu=pentiumpro} respectively.
7628 These synonyms are deprecated.
7630 @item -mfpmath=@var{unit}
7632 generate floating point arithmetics for selected unit @var{unit}. the choices
7637 Use the standard 387 floating point coprocessor present majority of chips and
7638 emulated otherwise. Code compiled with this option will run almost everywhere.
7639 The temporary results are computed in 80bit precesion instead of precision
7640 specified by the type resulting in slightly different results compared to most
7641 of other chips. See @option{-ffloat-store} for more detailed description.
7643 This is the default choice for i386 compiler.
7646 Use scalar floating point instructions present in the SSE instruction set.
7647 This instruction set is supported by Pentium3 and newer chips, in the AMD line
7648 by Athlon-4, Athlon-xp and Athlon-mp chips. The earlier version of SSE
7649 instruction set supports only single precision arithmetics, thus the double and
7650 extended precision arithmetics is still done using 387. Later version, present
7651 only in Pentium4 and the future AMD x86-64 chips supports double precision
7654 For i387 you need to use @option{-march=@var{cpu-type}}, @option{-msse} or
7655 @option{-msse2} switches to enable SSE extensions and make this option
7656 effective. For x86-64 compiler, these extensions are enabled by default.
7658 The resulting code should be considerably faster in majority of cases and avoid
7659 the numerical instability problems of 387 code, but may break some existing
7660 code that expects temporaries to be 80bit.
7662 This is the default choice for x86-64 compiler.
7665 Attempt to utilize both instruction sets at once. This effectivly double the
7666 amount of available registers and on chips with separate execution units for
7667 387 and SSE the execution resources too. Use this option with care, as it is
7668 still experimental, because gcc register allocator does not model separate
7669 functional units well resulting in instable performance.
7672 @item -masm=@var{dialect}
7673 @opindex masm=@var{dialect}
7674 Output asm instructions using selected @var{dialect}. Supported choices are
7675 @samp{intel} or @samp{att} (the default one).
7680 @opindex mno-ieee-fp
7681 Control whether or not the compiler uses IEEE floating point
7682 comparisons. These handle correctly the case where the result of a
7683 comparison is unordered.
7686 @opindex msoft-float
7687 Generate output containing library calls for floating point.
7688 @strong{Warning:} the requisite libraries are not part of GCC@.
7689 Normally the facilities of the machine's usual C compiler are used, but
7690 this can't be done directly in cross-compilation. You must make your
7691 own arrangements to provide suitable library functions for
7694 On machines where a function returns floating point results in the 80387
7695 register stack, some floating point opcodes may be emitted even if
7696 @option{-msoft-float} is used.
7698 @item -mno-fp-ret-in-387
7699 @opindex mno-fp-ret-in-387
7700 Do not use the FPU registers for return values of functions.
7702 The usual calling convention has functions return values of types
7703 @code{float} and @code{double} in an FPU register, even if there
7704 is no FPU@. The idea is that the operating system should emulate
7707 The option @option{-mno-fp-ret-in-387} causes such values to be returned
7708 in ordinary CPU registers instead.
7710 @item -mno-fancy-math-387
7711 @opindex mno-fancy-math-387
7712 Some 387 emulators do not support the @code{sin}, @code{cos} and
7713 @code{sqrt} instructions for the 387. Specify this option to avoid
7714 generating those instructions. This option is the default on FreeBSD@.
7715 As of revision 2.6.1, these instructions are not generated unless you
7716 also use the @option{-funsafe-math-optimizations} switch.
7718 @item -malign-double
7719 @itemx -mno-align-double
7720 @opindex malign-double
7721 @opindex mno-align-double
7722 Control whether GCC aligns @code{double}, @code{long double}, and
7723 @code{long long} variables on a two word boundary or a one word
7724 boundary. Aligning @code{double} variables on a two word boundary will
7725 produce code that runs somewhat faster on a @samp{Pentium} at the
7726 expense of more memory.
7728 @item -m128bit-long-double
7729 @opindex m128bit-long-double
7730 Control the size of @code{long double} type. i386 application binary interface
7731 specify the size to be 12 bytes, while modern architectures (Pentium and newer)
7732 prefer @code{long double} aligned to 8 or 16 byte boundary. This is
7733 impossible to reach with 12 byte long doubles in the array accesses.
7735 @strong{Warning:} if you use the @option{-m128bit-long-double} switch, the
7736 structures and arrays containing @code{long double} will change their size as
7737 well as function calling convention for function taking @code{long double}
7740 @item -m96bit-long-double
7741 @opindex m96bit-long-double
7742 Set the size of @code{long double} to 96 bits as required by the i386
7743 application binary interface. This is the default.
7746 @itemx -mno-svr3-shlib
7747 @opindex msvr3-shlib
7748 @opindex mno-svr3-shlib
7749 Control whether GCC places uninitialized local variables into the
7750 @code{bss} or @code{data} segments. @option{-msvr3-shlib} places them
7751 into @code{bss}. These options are meaningful only on System V Release 3.
7755 Use a different function-calling convention, in which functions that
7756 take a fixed number of arguments return with the @code{ret} @var{num}
7757 instruction, which pops their arguments while returning. This saves one
7758 instruction in the caller since there is no need to pop the arguments
7761 You can specify that an individual function is called with this calling
7762 sequence with the function attribute @samp{stdcall}. You can also
7763 override the @option{-mrtd} option by using the function attribute
7764 @samp{cdecl}. @xref{Function Attributes}.
7766 @strong{Warning:} this calling convention is incompatible with the one
7767 normally used on Unix, so you cannot use it if you need to call
7768 libraries compiled with the Unix compiler.
7770 Also, you must provide function prototypes for all functions that
7771 take variable numbers of arguments (including @code{printf});
7772 otherwise incorrect code will be generated for calls to those
7775 In addition, seriously incorrect code will result if you call a
7776 function with too many arguments. (Normally, extra arguments are
7777 harmlessly ignored.)
7779 @item -mregparm=@var{num}
7781 Control how many registers are used to pass integer arguments. By
7782 default, no registers are used to pass arguments, and at most 3
7783 registers can be used. You can control this behavior for a specific
7784 function by using the function attribute @samp{regparm}.
7785 @xref{Function Attributes}.
7787 @strong{Warning:} if you use this switch, and
7788 @var{num} is nonzero, then you must build all modules with the same
7789 value, including any libraries. This includes the system libraries and
7792 @item -mpreferred-stack-boundary=@var{num}
7793 @opindex mpreferred-stack-boundary
7794 Attempt to keep the stack boundary aligned to a 2 raised to @var{num}
7795 byte boundary. If @option{-mpreferred-stack-boundary} is not specified,
7796 the default is 4 (16 bytes or 128 bits), except when optimizing for code
7797 size (@option{-Os}), in which case the default is the minimum correct
7798 alignment (4 bytes for x86, and 8 bytes for x86-64).
7800 On Pentium and PentiumPro, @code{double} and @code{long double} values
7801 should be aligned to an 8 byte boundary (see @option{-malign-double}) or
7802 suffer significant run time performance penalties. On Pentium III, the
7803 Streaming SIMD Extension (SSE) data type @code{__m128} suffers similar
7804 penalties if it is not 16 byte aligned.
7806 To ensure proper alignment of this values on the stack, the stack boundary
7807 must be as aligned as that required by any value stored on the stack.
7808 Further, every function must be generated such that it keeps the stack
7809 aligned. Thus calling a function compiled with a higher preferred
7810 stack boundary from a function compiled with a lower preferred stack
7811 boundary will most likely misalign the stack. It is recommended that
7812 libraries that use callbacks always use the default setting.
7814 This extra alignment does consume extra stack space, and generally
7815 increases code size. Code that is sensitive to stack space usage, such
7816 as embedded systems and operating system kernels, may want to reduce the
7817 preferred alignment to @option{-mpreferred-stack-boundary=2}.
7833 These switches enable or disable the use of built-in functions that allow
7834 direct access to the MMX, SSE and 3Dnow extensions of the instruction set.
7836 @xref{X86 Built-in Functions}, for details of the functions enabled
7837 and disabled by these switches.
7840 @itemx -mno-push-args
7842 @opindex mno-push-args
7843 Use PUSH operations to store outgoing parameters. This method is shorter
7844 and usually equally fast as method using SUB/MOV operations and is enabled
7845 by default. In some cases disabling it may improve performance because of
7846 improved scheduling and reduced dependencies.
7848 @item -maccumulate-outgoing-args
7849 @opindex maccumulate-outgoing-args
7850 If enabled, the maximum amount of space required for outgoing arguments will be
7851 computed in the function prologue. This is faster on most modern CPUs
7852 because of reduced dependencies, improved scheduling and reduced stack usage
7853 when preferred stack boundary is not equal to 2. The drawback is a notable
7854 increase in code size. This switch implies @option{-mno-push-args}.
7858 Support thread-safe exception handling on @samp{Mingw32}. Code that relies
7859 on thread-safe exception handling must compile and link all code with the
7860 @option{-mthreads} option. When compiling, @option{-mthreads} defines
7861 @option{-D_MT}; when linking, it links in a special thread helper library
7862 @option{-lmingwthrd} which cleans up per thread exception handling data.
7864 @item -mno-align-stringops
7865 @opindex mno-align-stringops
7866 Do not align destination of inlined string operations. This switch reduces
7867 code size and improves performance in case the destination is already aligned,
7868 but gcc don't know about it.
7870 @item -minline-all-stringops
7871 @opindex minline-all-stringops
7872 By default GCC inlines string operations only when destination is known to be
7873 aligned at least to 4 byte boundary. This enables more inlining, increase code
7874 size, but may improve performance of code that depends on fast memcpy, strlen
7875 and memset for short lengths.
7877 @item -momit-leaf-frame-pointer
7878 @opindex momit-leaf-frame-pointer
7879 Don't keep the frame pointer in a register for leaf functions. This
7880 avoids the instructions to save, set up and restore frame pointers and
7881 makes an extra register available in leaf functions. The option
7882 @option{-fomit-frame-pointer} removes the frame pointer for all functions
7883 which might make debugging harder.
7886 These @samp{-m} switches are supported in addition to the above
7887 on AMD x86-64 processors in 64-bit environments.
7894 Generate code for a 32-bit or 64-bit environment.
7895 The 32-bit environment sets int, long and pointer to 32 bits and
7896 generates code that runs on any i386 system.
7897 The 64-bit environment sets int to 32 bits and long and pointer
7898 to 64 bits and generates code for AMD's x86-64 architecture.
7901 @opindex no-red-zone
7902 Do not use a so called red zone for x86-64 code. The red zone is mandated
7903 by the x86-64 ABI, it is a 128-byte area beyond the location of the
7904 stack pointer that will not be modified by signal or interrupt handlers
7905 and therefore can be used for temporary data without adjusting the stack
7906 pointer. The flag @option{-mno-red-zone} disables this red zone.
7910 @subsection HPPA Options
7911 @cindex HPPA Options
7913 These @samp{-m} options are defined for the HPPA family of computers:
7916 @item -march=@var{architecture-type}
7918 Generate code for the specified architecture. The choices for
7919 @var{architecture-type} are @samp{1.0} for PA 1.0, @samp{1.1} for PA
7920 1.1, and @samp{2.0} for PA 2.0 processors. Refer to
7921 @file{/usr/lib/sched.models} on an HP-UX system to determine the proper
7922 architecture option for your machine. Code compiled for lower numbered
7923 architectures will run on higher numbered architectures, but not the
7926 PA 2.0 support currently requires gas snapshot 19990413 or later. The
7927 next release of binutils (current is 2.9.1) will probably contain PA 2.0
7931 @itemx -mpa-risc-1-1
7932 @itemx -mpa-risc-2-0
7933 @opindex mpa-risc-1-0
7934 @opindex mpa-risc-1-1
7935 @opindex mpa-risc-2-0
7936 Synonyms for @option{-march=1.0}, @option{-march=1.1}, and @option{-march=2.0} respectively.
7939 @opindex mbig-switch
7940 Generate code suitable for big switch tables. Use this option only if
7941 the assembler/linker complain about out of range branches within a switch
7944 @item -mjump-in-delay
7945 @opindex mjump-in-delay
7946 Fill delay slots of function calls with unconditional jump instructions
7947 by modifying the return pointer for the function call to be the target
7948 of the conditional jump.
7950 @item -mdisable-fpregs
7951 @opindex mdisable-fpregs
7952 Prevent floating point registers from being used in any manner. This is
7953 necessary for compiling kernels which perform lazy context switching of
7954 floating point registers. If you use this option and attempt to perform
7955 floating point operations, the compiler will abort.
7957 @item -mdisable-indexing
7958 @opindex mdisable-indexing
7959 Prevent the compiler from using indexing address modes. This avoids some
7960 rather obscure problems when compiling MIG generated code under MACH@.
7962 @item -mno-space-regs
7963 @opindex mno-space-regs
7964 Generate code that assumes the target has no space registers. This allows
7965 GCC to generate faster indirect calls and use unscaled index address modes.
7967 Such code is suitable for level 0 PA systems and kernels.
7969 @item -mfast-indirect-calls
7970 @opindex mfast-indirect-calls
7971 Generate code that assumes calls never cross space boundaries. This
7972 allows GCC to emit code which performs faster indirect calls.
7974 This option will not work in the presence of shared libraries or nested
7977 @item -mlong-load-store
7978 @opindex mlong-load-store
7979 Generate 3-instruction load and store sequences as sometimes required by
7980 the HP-UX 10 linker. This is equivalent to the @samp{+k} option to
7983 @item -mportable-runtime
7984 @opindex mportable-runtime
7985 Use the portable calling conventions proposed by HP for ELF systems.
7989 Enable the use of assembler directives only GAS understands.
7991 @item -mschedule=@var{cpu-type}
7993 Schedule code according to the constraints for the machine type
7994 @var{cpu-type}. The choices for @var{cpu-type} are @samp{700}
7995 @samp{7100}, @samp{7100LC}, @samp{7200}, and @samp{8000}. Refer to
7996 @file{/usr/lib/sched.models} on an HP-UX system to determine the
7997 proper scheduling option for your machine.
8000 @opindex mlinker-opt
8001 Enable the optimization pass in the HPUX linker. Note this makes symbolic
8002 debugging impossible. It also triggers a bug in the HPUX 8 and HPUX 9 linkers
8003 in which they give bogus error messages when linking some programs.
8006 @opindex msoft-float
8007 Generate output containing library calls for floating point.
8008 @strong{Warning:} the requisite libraries are not available for all HPPA
8009 targets. Normally the facilities of the machine's usual C compiler are
8010 used, but this cannot be done directly in cross-compilation. You must make
8011 your own arrangements to provide suitable library functions for
8012 cross-compilation. The embedded target @samp{hppa1.1-*-pro}
8013 does provide software floating point support.
8015 @option{-msoft-float} changes the calling convention in the output file;
8016 therefore, it is only useful if you compile @emph{all} of a program with
8017 this option. In particular, you need to compile @file{libgcc.a}, the
8018 library that comes with GCC, with @option{-msoft-float} in order for
8022 @node Intel 960 Options
8023 @subsection Intel 960 Options
8025 These @samp{-m} options are defined for the Intel 960 implementations:
8028 @item -m@var{cpu-type}
8036 Assume the defaults for the machine type @var{cpu-type} for some of
8037 the other options, including instruction scheduling, floating point
8038 support, and addressing modes. The choices for @var{cpu-type} are
8039 @samp{ka}, @samp{kb}, @samp{mc}, @samp{ca}, @samp{cf},
8040 @samp{sa}, and @samp{sb}.
8047 @opindex msoft-float
8048 The @option{-mnumerics} option indicates that the processor does support
8049 floating-point instructions. The @option{-msoft-float} option indicates
8050 that floating-point support should not be assumed.
8052 @item -mleaf-procedures
8053 @itemx -mno-leaf-procedures
8054 @opindex mleaf-procedures
8055 @opindex mno-leaf-procedures
8056 Do (or do not) attempt to alter leaf procedures to be callable with the
8057 @code{bal} instruction as well as @code{call}. This will result in more
8058 efficient code for explicit calls when the @code{bal} instruction can be
8059 substituted by the assembler or linker, but less efficient code in other
8060 cases, such as calls via function pointers, or using a linker that doesn't
8061 support this optimization.
8064 @itemx -mno-tail-call
8066 @opindex mno-tail-call
8067 Do (or do not) make additional attempts (beyond those of the
8068 machine-independent portions of the compiler) to optimize tail-recursive
8069 calls into branches. You may not want to do this because the detection of
8070 cases where this is not valid is not totally complete. The default is
8071 @option{-mno-tail-call}.
8073 @item -mcomplex-addr
8074 @itemx -mno-complex-addr
8075 @opindex mcomplex-addr
8076 @opindex mno-complex-addr
8077 Assume (or do not assume) that the use of a complex addressing mode is a
8078 win on this implementation of the i960. Complex addressing modes may not
8079 be worthwhile on the K-series, but they definitely are on the C-series.
8080 The default is currently @option{-mcomplex-addr} for all processors except
8084 @itemx -mno-code-align
8085 @opindex mcode-align
8086 @opindex mno-code-align
8087 Align code to 8-byte boundaries for faster fetching (or don't bother).
8088 Currently turned on by default for C-series implementations only.
8091 @item -mclean-linkage
8092 @itemx -mno-clean-linkage
8093 @opindex mclean-linkage
8094 @opindex mno-clean-linkage
8095 These options are not fully implemented.
8099 @itemx -mic2.0-compat
8100 @itemx -mic3.0-compat
8102 @opindex mic2.0-compat
8103 @opindex mic3.0-compat
8104 Enable compatibility with iC960 v2.0 or v3.0.
8108 @opindex masm-compat
8110 Enable compatibility with the iC960 assembler.
8112 @item -mstrict-align
8113 @itemx -mno-strict-align
8114 @opindex mstrict-align
8115 @opindex mno-strict-align
8116 Do not permit (do permit) unaligned accesses.
8120 Enable structure-alignment compatibility with Intel's gcc release version
8121 1.3 (based on gcc 1.37). This option implies @option{-mstrict-align}.
8123 @item -mlong-double-64
8124 @opindex mlong-double-64
8125 Implement type @samp{long double} as 64-bit floating point numbers.
8126 Without the option @samp{long double} is implemented by 80-bit
8127 floating point numbers. The only reason we have it because there is
8128 no 128-bit @samp{long double} support in @samp{fp-bit.c} yet. So it
8129 is only useful for people using soft-float targets. Otherwise, we
8130 should recommend against use of it.
8134 @node DEC Alpha Options
8135 @subsection DEC Alpha Options
8137 These @samp{-m} options are defined for the DEC Alpha implementations:
8140 @item -mno-soft-float
8142 @opindex mno-soft-float
8143 @opindex msoft-float
8144 Use (do not use) the hardware floating-point instructions for
8145 floating-point operations. When @option{-msoft-float} is specified,
8146 functions in @file{libgcc.a} will be used to perform floating-point
8147 operations. Unless they are replaced by routines that emulate the
8148 floating-point operations, or compiled in such a way as to call such
8149 emulations routines, these routines will issue floating-point
8150 operations. If you are compiling for an Alpha without floating-point
8151 operations, you must ensure that the library is built so as not to call
8154 Note that Alpha implementations without floating-point operations are
8155 required to have floating-point registers.
8160 @opindex mno-fp-regs
8161 Generate code that uses (does not use) the floating-point register set.
8162 @option{-mno-fp-regs} implies @option{-msoft-float}. If the floating-point
8163 register set is not used, floating point operands are passed in integer
8164 registers as if they were integers and floating-point results are passed
8165 in @code{$0} instead of @code{$f0}. This is a non-standard calling sequence,
8166 so any function with a floating-point argument or return value called by code
8167 compiled with @option{-mno-fp-regs} must also be compiled with that
8170 A typical use of this option is building a kernel that does not use,
8171 and hence need not save and restore, any floating-point registers.
8175 The Alpha architecture implements floating-point hardware optimized for
8176 maximum performance. It is mostly compliant with the IEEE floating
8177 point standard. However, for full compliance, software assistance is
8178 required. This option generates code fully IEEE compliant code
8179 @emph{except} that the @var{inexact-flag} is not maintained (see below).
8180 If this option is turned on, the preprocessor macro @code{_IEEE_FP} is
8181 defined during compilation. The resulting code is less efficient but is
8182 able to correctly support denormalized numbers and exceptional IEEE
8183 values such as not-a-number and plus/minus infinity. Other Alpha
8184 compilers call this option @option{-ieee_with_no_inexact}.
8186 @item -mieee-with-inexact
8187 @opindex mieee-with-inexact
8188 This is like @option{-mieee} except the generated code also maintains
8189 the IEEE @var{inexact-flag}. Turning on this option causes the
8190 generated code to implement fully-compliant IEEE math. In addition to
8191 @code{_IEEE_FP}, @code{_IEEE_FP_EXACT} is defined as a preprocessor
8192 macro. On some Alpha implementations the resulting code may execute
8193 significantly slower than the code generated by default. Since there is
8194 very little code that depends on the @var{inexact-flag}, you should
8195 normally not specify this option. Other Alpha compilers call this
8196 option @option{-ieee_with_inexact}.
8198 @item -mfp-trap-mode=@var{trap-mode}
8199 @opindex mfp-trap-mode
8200 This option controls what floating-point related traps are enabled.
8201 Other Alpha compilers call this option @option{-fptm @var{trap-mode}}.
8202 The trap mode can be set to one of four values:
8206 This is the default (normal) setting. The only traps that are enabled
8207 are the ones that cannot be disabled in software (e.g., division by zero
8211 In addition to the traps enabled by @samp{n}, underflow traps are enabled
8215 Like @samp{su}, but the instructions are marked to be safe for software
8216 completion (see Alpha architecture manual for details).
8219 Like @samp{su}, but inexact traps are enabled as well.
8222 @item -mfp-rounding-mode=@var{rounding-mode}
8223 @opindex mfp-rounding-mode
8224 Selects the IEEE rounding mode. Other Alpha compilers call this option
8225 @option{-fprm @var{rounding-mode}}. The @var{rounding-mode} can be one
8230 Normal IEEE rounding mode. Floating point numbers are rounded towards
8231 the nearest machine number or towards the even machine number in case
8235 Round towards minus infinity.
8238 Chopped rounding mode. Floating point numbers are rounded towards zero.
8241 Dynamic rounding mode. A field in the floating point control register
8242 (@var{fpcr}, see Alpha architecture reference manual) controls the
8243 rounding mode in effect. The C library initializes this register for
8244 rounding towards plus infinity. Thus, unless your program modifies the
8245 @var{fpcr}, @samp{d} corresponds to round towards plus infinity.
8248 @item -mtrap-precision=@var{trap-precision}
8249 @opindex mtrap-precision
8250 In the Alpha architecture, floating point traps are imprecise. This
8251 means without software assistance it is impossible to recover from a
8252 floating trap and program execution normally needs to be terminated.
8253 GCC can generate code that can assist operating system trap handlers
8254 in determining the exact location that caused a floating point trap.
8255 Depending on the requirements of an application, different levels of
8256 precisions can be selected:
8260 Program precision. This option is the default and means a trap handler
8261 can only identify which program caused a floating point exception.
8264 Function precision. The trap handler can determine the function that
8265 caused a floating point exception.
8268 Instruction precision. The trap handler can determine the exact
8269 instruction that caused a floating point exception.
8272 Other Alpha compilers provide the equivalent options called
8273 @option{-scope_safe} and @option{-resumption_safe}.
8275 @item -mieee-conformant
8276 @opindex mieee-conformant
8277 This option marks the generated code as IEEE conformant. You must not
8278 use this option unless you also specify @option{-mtrap-precision=i} and either
8279 @option{-mfp-trap-mode=su} or @option{-mfp-trap-mode=sui}. Its only effect
8280 is to emit the line @samp{.eflag 48} in the function prologue of the
8281 generated assembly file. Under DEC Unix, this has the effect that
8282 IEEE-conformant math library routines will be linked in.
8284 @item -mbuild-constants
8285 @opindex mbuild-constants
8286 Normally GCC examines a 32- or 64-bit integer constant to
8287 see if it can construct it from smaller constants in two or three
8288 instructions. If it cannot, it will output the constant as a literal and
8289 generate code to load it from the data segment at runtime.
8291 Use this option to require GCC to construct @emph{all} integer constants
8292 using code, even if it takes more instructions (the maximum is six).
8294 You would typically use this option to build a shared library dynamic
8295 loader. Itself a shared library, it must relocate itself in memory
8296 before it can find the variables and constants in its own data segment.
8302 Select whether to generate code to be assembled by the vendor-supplied
8303 assembler (@option{-malpha-as}) or by the GNU assembler @option{-mgas}.
8321 Indicate whether GCC should generate code to use the optional BWX,
8322 CIX, FIX and MAX instruction sets. The default is to use the instruction
8323 sets supported by the CPU type specified via @option{-mcpu=} option or that
8324 of the CPU on which GCC was built if none was specified.
8329 @opindex mfloat-ieee
8330 Generate code that uses (does not use) VAX F and G floating point
8331 arithmetic instead of IEEE single and double precision.
8333 @item -mexplicit-relocs
8334 @itemx -mno-explicit-relocs
8335 @opindex mexplicit-relocs
8336 @opindex mno-explicit-relocs
8337 Older Alpha assemblers provided no way to generate symbol relocations
8338 except via assembler macros. Use of these macros does not allow
8339 optimial instruction scheduling. GNU binutils as of version 2.12
8340 supports a new syntax that allows the compiler to explicitly mark
8341 which relocations should apply to which instructions. This option
8342 is mostly useful for debugging, as GCC detects the capabilities of
8343 the assembler when it is built and sets the default accordingly.
8347 @opindex msmall-data
8348 @opindex mlarge-data
8349 When @option{-mexplicit-relocs} is in effect, static data is
8350 accessed via @dfn{gp-relative} relocations. When @option{-msmall-data}
8351 is used, objects 8 bytes long or smaller are placed in a @dfn{small data area}
8352 (the @code{.sdata} and @code{.sbss} sections) and are accessed via
8353 16-bit relocations off of the @code{$gp} register. This limits the
8354 size of the small data area to 64KB, but allows the variables to be
8355 directly accessed via a single instruction.
8357 The default is @option{-mlarge-data}. With this option the data area
8358 is limited to just below 2GB. Programs that require more than 2GB of
8359 data must use @code{malloc} or @code{mmap} to allocate the data in the
8360 heap instead of in the program's data segment.
8362 When generating code for shared libraries, @option{-fpic} implies
8363 @option{-msmall-data} and @option{-fPIC} implies @option{-mlarge-data}.
8365 @item -mcpu=@var{cpu_type}
8367 Set the instruction set and instruction scheduling parameters for
8368 machine type @var{cpu_type}. You can specify either the @samp{EV}
8369 style name or the corresponding chip number. GCC supports scheduling
8370 parameters for the EV4, EV5 and EV6 family of processors and will
8371 choose the default values for the instruction set from the processor
8372 you specify. If you do not specify a processor type, GCC will default
8373 to the processor on which the compiler was built.
8375 Supported values for @var{cpu_type} are
8381 Schedules as an EV4 and has no instruction set extensions.
8385 Schedules as an EV5 and has no instruction set extensions.
8389 Schedules as an EV5 and supports the BWX extension.
8394 Schedules as an EV5 and supports the BWX and MAX extensions.
8398 Schedules as an EV6 and supports the BWX, FIX, and MAX extensions.
8402 Schedules as an EV6 and supports the BWX, CIX, FIX, and MAX extensions.
8405 @item -mtune=@var{cpu_type}
8407 Set only the instruction scheduling parameters for machine type
8408 @var{cpu_type}. The instruction set is not changed.
8410 @item -mmemory-latency=@var{time}
8411 @opindex mmemory-latency
8412 Sets the latency the scheduler should assume for typical memory
8413 references as seen by the application. This number is highly
8414 dependent on the memory access patterns used by the application
8415 and the size of the external cache on the machine.
8417 Valid options for @var{time} are
8421 A decimal number representing clock cycles.
8427 The compiler contains estimates of the number of clock cycles for
8428 ``typical'' EV4 & EV5 hardware for the Level 1, 2 & 3 caches
8429 (also called Dcache, Scache, and Bcache), as well as to main memory.
8430 Note that L3 is only valid for EV5.
8435 @node DEC Alpha/VMS Options
8436 @subsection DEC Alpha/VMS Options
8438 These @samp{-m} options are defined for the DEC Alpha/VMS implementations:
8441 @item -mvms-return-codes
8442 @opindex mvms-return-codes
8443 Return VMS condition codes from main. The default is to return POSIX
8444 style condition (e.g.@ error) codes.
8447 @node Clipper Options
8448 @subsection Clipper Options
8450 These @samp{-m} options are defined for the Clipper implementations:
8455 Produce code for a C300 Clipper processor. This is the default.
8459 Produce code for a C400 Clipper processor, i.e.@: use floating point
8463 @node H8/300 Options
8464 @subsection H8/300 Options
8466 These @samp{-m} options are defined for the H8/300 implementations:
8471 Shorten some address references at link time, when possible; uses the
8472 linker option @option{-relax}. @xref{H8/300,, @code{ld} and the H8/300,
8473 ld.info, Using ld}, for a fuller description.
8477 Generate code for the H8/300H@.
8481 Generate code for the H8/S@.
8485 Generate code for the H8/S2600. This switch must be used with @option{-ms}.
8489 Make @code{int} data 32 bits by default.
8493 On the H8/300H and H8/S, use the same alignment rules as for the H8/300.
8494 The default for the H8/300H and H8/S is to align longs and floats on 4
8496 @option{-malign-300} causes them to be aligned on 2 byte boundaries.
8497 This option has no effect on the H8/300.
8501 @subsection SH Options
8503 These @samp{-m} options are defined for the SH implementations:
8508 Generate code for the SH1.
8512 Generate code for the SH2.
8516 Generate code for the SH3.
8520 Generate code for the SH3e.
8524 Generate code for the SH4 without a floating-point unit.
8526 @item -m4-single-only
8527 @opindex m4-single-only
8528 Generate code for the SH4 with a floating-point unit that only
8529 supports single-precision arithmetic.
8533 Generate code for the SH4 assuming the floating-point unit is in
8534 single-precision mode by default.
8538 Generate code for the SH4.
8542 Compile code for the processor in big endian mode.
8546 Compile code for the processor in little endian mode.
8550 Align doubles at 64-bit boundaries. Note that this changes the calling
8551 conventions, and thus some functions from the standard C library will
8552 not work unless you recompile it first with @option{-mdalign}.
8556 Shorten some address references at link time, when possible; uses the
8557 linker option @option{-relax}.
8561 Use 32-bit offsets in @code{switch} tables. The default is to use
8566 Enable the use of the instruction @code{fmovd}.
8570 Comply with the calling conventions defined by Hitachi.
8574 Mark the @code{MAC} register as call-clobbered, even if
8575 @option{-mhitachi} is given.
8579 Increase IEEE-compliance of floating-point code.
8583 Dump instruction size and location in the assembly code.
8587 This option is deprecated. It pads structures to multiple of 4 bytes,
8588 which is incompatible with the SH ABI@.
8592 Optimize for space instead of speed. Implied by @option{-Os}.
8596 When generating position-independent code, emit function calls using
8597 the Global Offset Table instead of the Procedure Linkage Table.
8601 Generate a library function call to invalidate instruction cache
8602 entries, after fixing up a trampoline. This library function call
8603 doesn't assume it can write to the whole memory address space. This
8604 is the default when the target is @code{sh-*-linux*}.
8607 @node System V Options
8608 @subsection Options for System V
8610 These additional options are available on System V Release 4 for
8611 compatibility with other compilers on those systems:
8616 Create a shared object.
8617 It is recommended that @option{-symbolic} or @option{-shared} be used instead.
8621 Identify the versions of each tool used by the compiler, in a
8622 @code{.ident} assembler directive in the output.
8626 Refrain from adding @code{.ident} directives to the output file (this is
8629 @item -YP,@var{dirs}
8631 Search the directories @var{dirs}, and no others, for libraries
8632 specified with @option{-l}.
8636 Look in the directory @var{dir} to find the M4 preprocessor.
8637 The assembler uses this option.
8638 @c This is supposed to go with a -Yd for predefined M4 macro files, but
8639 @c the generic assembler that comes with Solaris takes just -Ym.
8642 @node TMS320C3x/C4x Options
8643 @subsection TMS320C3x/C4x Options
8644 @cindex TMS320C3x/C4x Options
8646 These @samp{-m} options are defined for TMS320C3x/C4x implementations:
8650 @item -mcpu=@var{cpu_type}
8652 Set the instruction set, register set, and instruction scheduling
8653 parameters for machine type @var{cpu_type}. Supported values for
8654 @var{cpu_type} are @samp{c30}, @samp{c31}, @samp{c32}, @samp{c40}, and
8655 @samp{c44}. The default is @samp{c40} to generate code for the
8660 @itemx -msmall-memory
8662 @opindex mbig-memory
8664 @opindex msmall-memory
8666 Generates code for the big or small memory model. The small memory
8667 model assumed that all data fits into one 64K word page. At run-time
8668 the data page (DP) register must be set to point to the 64K page
8669 containing the .bss and .data program sections. The big memory model is
8670 the default and requires reloading of the DP register for every direct
8677 Allow (disallow) allocation of general integer operands into the block
8684 Enable (disable) generation of code using decrement and branch,
8685 DBcond(D), instructions. This is enabled by default for the C4x. To be
8686 on the safe side, this is disabled for the C3x, since the maximum
8687 iteration count on the C3x is @math{2^23 + 1} (but who iterates loops more than
8688 @math{2^23} times on the C3x?). Note that GCC will try to reverse a loop so
8689 that it can utilise the decrement and branch instruction, but will give
8690 up if there is more than one memory reference in the loop. Thus a loop
8691 where the loop counter is decremented can generate slightly more
8692 efficient code, in cases where the RPTB instruction cannot be utilised.
8694 @item -mdp-isr-reload
8696 @opindex mdp-isr-reload
8698 Force the DP register to be saved on entry to an interrupt service
8699 routine (ISR), reloaded to point to the data section, and restored on
8700 exit from the ISR@. This should not be required unless someone has
8701 violated the small memory model by modifying the DP register, say within
8708 For the C3x use the 24-bit MPYI instruction for integer multiplies
8709 instead of a library call to guarantee 32-bit results. Note that if one
8710 of the operands is a constant, then the multiplication will be performed
8711 using shifts and adds. If the @option{-mmpyi} option is not specified for the C3x,
8712 then squaring operations are performed inline instead of a library call.
8715 @itemx -mno-fast-fix
8717 @opindex mno-fast-fix
8718 The C3x/C4x FIX instruction to convert a floating point value to an
8719 integer value chooses the nearest integer less than or equal to the
8720 floating point value rather than to the nearest integer. Thus if the
8721 floating point number is negative, the result will be incorrectly
8722 truncated an additional code is necessary to detect and correct this
8723 case. This option can be used to disable generation of the additional
8724 code required to correct the result.
8730 Enable (disable) generation of repeat block sequences using the RPTB
8731 instruction for zero overhead looping. The RPTB construct is only used
8732 for innermost loops that do not call functions or jump across the loop
8733 boundaries. There is no advantage having nested RPTB loops due to the
8734 overhead required to save and restore the RC, RS, and RE registers.
8735 This is enabled by default with @option{-O2}.
8737 @item -mrpts=@var{count}
8741 Enable (disable) the use of the single instruction repeat instruction
8742 RPTS@. If a repeat block contains a single instruction, and the loop
8743 count can be guaranteed to be less than the value @var{count}, GCC will
8744 emit a RPTS instruction instead of a RPTB@. If no value is specified,
8745 then a RPTS will be emitted even if the loop count cannot be determined
8746 at compile time. Note that the repeated instruction following RPTS does
8747 not have to be reloaded from memory each iteration, thus freeing up the
8748 CPU buses for operands. However, since interrupts are blocked by this
8749 instruction, it is disabled by default.
8751 @item -mloop-unsigned
8752 @itemx -mno-loop-unsigned
8753 @opindex mloop-unsigned
8754 @opindex mno-loop-unsigned
8755 The maximum iteration count when using RPTS and RPTB (and DB on the C40)
8756 is @math{2^31 + 1} since these instructions test if the iteration count is
8757 negative to terminate the loop. If the iteration count is unsigned
8758 there is a possibility than the @math{2^31 + 1} maximum iteration count may be
8759 exceeded. This switch allows an unsigned iteration count.
8763 Try to emit an assembler syntax that the TI assembler (asm30) is happy
8764 with. This also enforces compatibility with the API employed by the TI
8765 C3x C compiler. For example, long doubles are passed as structures
8766 rather than in floating point registers.
8772 Generate code that uses registers (stack) for passing arguments to functions.
8773 By default, arguments are passed in registers where possible rather
8774 than by pushing arguments on to the stack.
8776 @item -mparallel-insns
8777 @itemx -mno-parallel-insns
8778 @opindex mparallel-insns
8779 @opindex mno-parallel-insns
8780 Allow the generation of parallel instructions. This is enabled by
8781 default with @option{-O2}.
8783 @item -mparallel-mpy
8784 @itemx -mno-parallel-mpy
8785 @opindex mparallel-mpy
8786 @opindex mno-parallel-mpy
8787 Allow the generation of MPY||ADD and MPY||SUB parallel instructions,
8788 provided @option{-mparallel-insns} is also specified. These instructions have
8789 tight register constraints which can pessimize the code generation
8795 @subsection V850 Options
8796 @cindex V850 Options
8798 These @samp{-m} options are defined for V850 implementations:
8802 @itemx -mno-long-calls
8803 @opindex mlong-calls
8804 @opindex mno-long-calls
8805 Treat all calls as being far away (near). If calls are assumed to be
8806 far away, the compiler will always load the functions address up into a
8807 register, and call indirect through the pointer.
8813 Do not optimize (do optimize) basic blocks that use the same index
8814 pointer 4 or more times to copy pointer into the @code{ep} register, and
8815 use the shorter @code{sld} and @code{sst} instructions. The @option{-mep}
8816 option is on by default if you optimize.
8818 @item -mno-prolog-function
8819 @itemx -mprolog-function
8820 @opindex mno-prolog-function
8821 @opindex mprolog-function
8822 Do not use (do use) external functions to save and restore registers at
8823 the prolog and epilog of a function. The external functions are slower,
8824 but use less code space if more than one function saves the same number
8825 of registers. The @option{-mprolog-function} option is on by default if
8830 Try to make the code as small as possible. At present, this just turns
8831 on the @option{-mep} and @option{-mprolog-function} options.
8835 Put static or global variables whose size is @var{n} bytes or less into
8836 the tiny data area that register @code{ep} points to. The tiny data
8837 area can hold up to 256 bytes in total (128 bytes for byte references).
8841 Put static or global variables whose size is @var{n} bytes or less into
8842 the small data area that register @code{gp} points to. The small data
8843 area can hold up to 64 kilobytes.
8847 Put static or global variables whose size is @var{n} bytes or less into
8848 the first 32 kilobytes of memory.
8852 Specify that the target processor is the V850.
8855 @opindex mbig-switch
8856 Generate code suitable for big switch tables. Use this option only if
8857 the assembler/linker complain about out of range branches within a switch
8862 @subsection ARC Options
8865 These options are defined for ARC implementations:
8870 Compile code for little endian mode. This is the default.
8874 Compile code for big endian mode.
8877 @opindex mmangle-cpu
8878 Prepend the name of the cpu to all public symbol names.
8879 In multiple-processor systems, there are many ARC variants with different
8880 instruction and register set characteristics. This flag prevents code
8881 compiled for one cpu to be linked with code compiled for another.
8882 No facility exists for handling variants that are ``almost identical''.
8883 This is an all or nothing option.
8885 @item -mcpu=@var{cpu}
8887 Compile code for ARC variant @var{cpu}.
8888 Which variants are supported depend on the configuration.
8889 All variants support @option{-mcpu=base}, this is the default.
8891 @item -mtext=@var{text-section}
8892 @itemx -mdata=@var{data-section}
8893 @itemx -mrodata=@var{readonly-data-section}
8897 Put functions, data, and readonly data in @var{text-section},
8898 @var{data-section}, and @var{readonly-data-section} respectively
8899 by default. This can be overridden with the @code{section} attribute.
8900 @xref{Variable Attributes}.
8905 @subsection NS32K Options
8906 @cindex NS32K options
8908 These are the @samp{-m} options defined for the 32000 series. The default
8909 values for these options depends on which style of 32000 was selected when
8910 the compiler was configured; the defaults for the most common choices are
8918 Generate output for a 32032. This is the default
8919 when the compiler is configured for 32032 and 32016 based systems.
8925 Generate output for a 32332. This is the default
8926 when the compiler is configured for 32332-based systems.
8932 Generate output for a 32532. This is the default
8933 when the compiler is configured for 32532-based systems.
8937 Generate output containing 32081 instructions for floating point.
8938 This is the default for all systems.
8942 Generate output containing 32381 instructions for floating point. This
8943 also implies @option{-m32081}. The 32381 is only compatible with the 32332
8944 and 32532 cpus. This is the default for the pc532-netbsd configuration.
8948 Try and generate multiply-add floating point instructions @code{polyF}
8949 and @code{dotF}. This option is only available if the @option{-m32381}
8950 option is in effect. Using these instructions requires changes to
8951 register allocation which generally has a negative impact on
8952 performance. This option should only be enabled when compiling code
8953 particularly likely to make heavy use of multiply-add instructions.
8956 @opindex mnomulti-add
8957 Do not try and generate multiply-add floating point instructions
8958 @code{polyF} and @code{dotF}. This is the default on all platforms.
8961 @opindex msoft-float
8962 Generate output containing library calls for floating point.
8963 @strong{Warning:} the requisite libraries may not be available.
8966 @opindex mnobitfield
8967 Do not use the bit-field instructions. On some machines it is faster to
8968 use shifting and masking operations. This is the default for the pc532.
8972 Do use the bit-field instructions. This is the default for all platforms
8977 Use a different function-calling convention, in which functions
8978 that take a fixed number of arguments return pop their
8979 arguments on return with the @code{ret} instruction.
8981 This calling convention is incompatible with the one normally
8982 used on Unix, so you cannot use it if you need to call libraries
8983 compiled with the Unix compiler.
8985 Also, you must provide function prototypes for all functions that
8986 take variable numbers of arguments (including @code{printf});
8987 otherwise incorrect code will be generated for calls to those
8990 In addition, seriously incorrect code will result if you call a
8991 function with too many arguments. (Normally, extra arguments are
8992 harmlessly ignored.)
8994 This option takes its name from the 680x0 @code{rtd} instruction.
8999 Use a different function-calling convention where the first two arguments
9000 are passed in registers.
9002 This calling convention is incompatible with the one normally
9003 used on Unix, so you cannot use it if you need to call libraries
9004 compiled with the Unix compiler.
9007 @opindex mnoregparam
9008 Do not pass any arguments in registers. This is the default for all
9013 It is OK to use the sb as an index register which is always loaded with
9014 zero. This is the default for the pc532-netbsd target.
9018 The sb register is not available for use or has not been initialized to
9019 zero by the run time system. This is the default for all targets except
9020 the pc532-netbsd. It is also implied whenever @option{-mhimem} or
9021 @option{-fpic} is set.
9025 Many ns32000 series addressing modes use displacements of up to 512MB@.
9026 If an address is above 512MB then displacements from zero can not be used.
9027 This option causes code to be generated which can be loaded above 512MB@.
9028 This may be useful for operating systems or ROM code.
9032 Assume code will be loaded in the first 512MB of virtual address space.
9033 This is the default for all platforms.
9039 @subsection AVR Options
9042 These options are defined for AVR implementations:
9045 @item -mmcu=@var{mcu}
9047 Specify ATMEL AVR instruction set or MCU type.
9049 Instruction set avr1 is for the minimal AVR core, not supported by the C
9050 compiler, only for assembler programs (MCU types: at90s1200, attiny10,
9051 attiny11, attiny12, attiny15, attiny28).
9053 Instruction set avr2 (default) is for the classic AVR core with up to
9054 8K program memory space (MCU types: at90s2313, at90s2323, attiny22,
9055 at90s2333, at90s2343, at90s4414, at90s4433, at90s4434, at90s8515,
9056 at90c8534, at90s8535).
9058 Instruction set avr3 is for the classic AVR core with up to 128K program
9059 memory space (MCU types: atmega103, atmega603, at43usb320, at76c711).
9061 Instruction set avr4 is for the enhanced AVR core with up to 8K program
9062 memory space (MCU types: atmega8, atmega83, atmega85).
9064 Instruction set avr5 is for the enhanced AVR core with up to 128K program
9065 memory space (MCU types: atmega16, atmega161, atmega163, atmega32, atmega323,
9066 atmega64, atmega128, at43usb355, at94k).
9070 Output instruction sizes to the asm file.
9072 @item -minit-stack=@var{N}
9073 @opindex minit-stack
9074 Specify the initial stack address, which may be a symbol or numeric value,
9075 @samp{__stack} is the default.
9077 @item -mno-interrupts
9078 @opindex mno-interrupts
9079 Generated code is not compatible with hardware interrupts.
9080 Code size will be smaller.
9082 @item -mcall-prologues
9083 @opindex mcall-prologues
9084 Functions prologues/epilogues expanded as call to appropriate
9085 subroutines. Code size will be smaller.
9087 @item -mno-tablejump
9088 @opindex mno-tablejump
9089 Do not generate tablejump insns which sometimes increase code size.
9092 @opindex mtiny-stack
9093 Change only the low 8 bits of the stack pointer.
9097 @subsection MCore Options
9098 @cindex MCore options
9100 These are the @samp{-m} options defined for the Motorola M*Core
9110 @opindex mno-hardlit
9111 Inline constants into the code stream if it can be done in two
9112 instructions or less.
9120 Use the divide instruction. (Enabled by default).
9122 @item -mrelax-immediate
9123 @itemx -mrelax-immediate
9124 @itemx -mno-relax-immediate
9125 @opindex mrelax-immediate
9126 @opindex mrelax-immediate
9127 @opindex mno-relax-immediate
9128 Allow arbitrary sized immediates in bit operations.
9130 @item -mwide-bitfields
9131 @itemx -mwide-bitfields
9132 @itemx -mno-wide-bitfields
9133 @opindex mwide-bitfields
9134 @opindex mwide-bitfields
9135 @opindex mno-wide-bitfields
9136 Always treat bit-fields as int-sized.
9138 @item -m4byte-functions
9139 @itemx -m4byte-functions
9140 @itemx -mno-4byte-functions
9141 @opindex m4byte-functions
9142 @opindex m4byte-functions
9143 @opindex mno-4byte-functions
9144 Force all functions to be aligned to a four byte boundary.
9146 @item -mcallgraph-data
9147 @itemx -mcallgraph-data
9148 @itemx -mno-callgraph-data
9149 @opindex mcallgraph-data
9150 @opindex mcallgraph-data
9151 @opindex mno-callgraph-data
9152 Emit callgraph information.
9156 @itemx -mno-slow-bytes
9157 @opindex mslow-bytes
9158 @opindex mslow-bytes
9159 @opindex mno-slow-bytes
9160 Prefer word access when reading byte quantities.
9162 @item -mlittle-endian
9163 @itemx -mlittle-endian
9165 @opindex mlittle-endian
9166 @opindex mlittle-endian
9167 @opindex mbig-endian
9168 Generate code for a little endian target.
9176 Generate code for the 210 processor.
9180 @subsection IA-64 Options
9181 @cindex IA-64 Options
9183 These are the @samp{-m} options defined for the Intel IA-64 architecture.
9187 @opindex mbig-endian
9188 Generate code for a big endian target. This is the default for HPUX@.
9190 @item -mlittle-endian
9191 @opindex mlittle-endian
9192 Generate code for a little endian target. This is the default for AIX5
9199 Generate (or don't) code for the GNU assembler. This is the default.
9200 @c Also, this is the default if the configure option @option{--with-gnu-as}
9207 Generate (or don't) code for the GNU linker. This is the default.
9208 @c Also, this is the default if the configure option @option{--with-gnu-ld}
9213 Generate code that does not use a global pointer register. The result
9214 is not position independent code, and violates the IA-64 ABI@.
9216 @item -mvolatile-asm-stop
9217 @itemx -mno-volatile-asm-stop
9218 @opindex mvolatile-asm-stop
9219 @opindex mno-volatile-asm-stop
9220 Generate (or don't) a stop bit immediately before and after volatile asm
9225 Generate code that works around Itanium B step errata.
9227 @item -mregister-names
9228 @itemx -mno-register-names
9229 @opindex mregister-names
9230 @opindex mno-register-names
9231 Generate (or don't) @samp{in}, @samp{loc}, and @samp{out} register names for
9232 the stacked registers. This may make assembler output more readable.
9238 Disable (or enable) optimizations that use the small data section. This may
9239 be useful for working around optimizer bugs.
9242 @opindex mconstant-gp
9243 Generate code that uses a single constant global pointer value. This is
9244 useful when compiling kernel code.
9248 Generate code that is self-relocatable. This implies @option{-mconstant-gp}.
9249 This is useful when compiling firmware code.
9251 @item -minline-divide-min-latency
9252 @opindex minline-divide-min-latency
9253 Generate code for inline divides using the minimum latency algorithm.
9255 @item -minline-divide-max-throughput
9256 @opindex minline-divide-max-throughput
9257 Generate code for inline divides using the maximum throughput algorithm.
9259 @item -mno-dwarf2-asm
9261 @opindex mno-dwarf2-asm
9262 @opindex mdwarf2-asm
9263 Don't (or do) generate assembler code for the DWARF2 line number debugging
9264 info. This may be useful when not using the GNU assembler.
9266 @item -mfixed-range=@var{register-range}
9267 @opindex mfixed-range
9268 Generate code treating the given register range as fixed registers.
9269 A fixed register is one that the register allocator can not use. This is
9270 useful when compiling kernel code. A register range is specified as
9271 two registers separated by a dash. Multiple register ranges can be
9272 specified separated by a comma.
9276 @subsection D30V Options
9277 @cindex D30V Options
9279 These @samp{-m} options are defined for D30V implementations:
9284 Link the @samp{.text}, @samp{.data}, @samp{.bss}, @samp{.strings},
9285 @samp{.rodata}, @samp{.rodata1}, @samp{.data1} sections into external
9286 memory, which starts at location @code{0x80000000}.
9290 Same as the @option{-mextmem} switch.
9294 Link the @samp{.text} section into onchip text memory, which starts at
9295 location @code{0x0}. Also link @samp{.data}, @samp{.bss},
9296 @samp{.strings}, @samp{.rodata}, @samp{.rodata1}, @samp{.data1} sections
9297 into onchip data memory, which starts at location @code{0x20000000}.
9299 @item -mno-asm-optimize
9300 @itemx -masm-optimize
9301 @opindex mno-asm-optimize
9302 @opindex masm-optimize
9303 Disable (enable) passing @option{-O} to the assembler when optimizing.
9304 The assembler uses the @option{-O} option to automatically parallelize
9305 adjacent short instructions where possible.
9307 @item -mbranch-cost=@var{n}
9308 @opindex mbranch-cost
9309 Increase the internal costs of branches to @var{n}. Higher costs means
9310 that the compiler will issue more instructions to avoid doing a branch.
9313 @item -mcond-exec=@var{n}
9315 Specify the maximum number of conditionally executed instructions that
9316 replace a branch. The default is 4.
9319 @node S/390 and zSeries Options
9320 @subsection S/390 and zSeries Options
9321 @cindex S/390 and zSeries Options
9323 These are the @samp{-m} options defined for the S/390 and zSeries architecture.
9328 @opindex mhard-float
9329 @opindex msoft-float
9330 Use (do not use) the hardware floating-point instructions and registers
9331 for floating-point operations. When @option{-msoft-float} is specified,
9332 functions in @file{libgcc.a} will be used to perform floating-point
9333 operations. When @option{-mhard-float} is specified, the compiler
9334 generates IEEE floating-point instructions. This is the default.
9337 @itemx -mno-backchain
9339 @opindex mno-backchain
9340 Generate (or do not generate) code which maintains an explicit
9341 backchain within the stack frame that points to the caller's frame.
9342 This is currently needed to allow debugging. The default is to
9343 generate the backchain.
9346 @itemx -mno-small-exec
9347 @opindex msmall-exec
9348 @opindex mno-small-exec
9349 Generate (or do not generate) code using the @code{bras} instruction
9350 to do subroutine calls.
9351 This only works reliably if the total executable size does not
9352 exceed 64k. The default is to use the @code{basr} instruction instead,
9353 which does not have this limitation.
9359 When @option{-m31} is specified, generate code compliant to the
9360 Linux for S/390 ABI@. When @option{-m64} is specified, generate
9361 code compliant to the Linux for zSeries ABI@. This allows GCC in
9362 particular to generate 64-bit instructions. For the @samp{s390}
9363 targets, the default is @option{-m31}, while the @samp{s390x}
9364 targets default to @option{-m64}.
9370 Generate (or do not generate) code using the @code{mvcle} instruction
9371 to perform block moves. When @option{-mno-mvcle} is specifed,
9372 use a @code{mvc} loop instead. This is the default.
9378 Print (or do not print) additional debug information when compiling.
9379 The default is to not print debug information.
9384 @subsection CRIS Options
9385 @cindex CRIS Options
9387 These options are defined specifically for the CRIS ports.
9390 @item -march=@var{architecture-type}
9391 @itemx -mcpu=@var{architecture-type}
9394 Generate code for the specified architecture. The choices for
9395 @var{architecture-type} are @samp{v3}, @samp{v8} and @samp{v10} for
9396 respectively ETRAX@w{ }4, ETRAX@w{ }100, and ETRAX@w{ }100@w{ }LX.
9397 Default is @samp{v0} except for cris-axis-linux-gnu, where the default is
9400 @item -mtune=@var{architecture-type}
9402 Tune to @var{architecture-type} everything applicable about the generated
9403 code, except for the ABI and the set of available instructions. The
9404 choices for @var{architecture-type} are the same as for
9405 @option{-march=@var{architecture-type}}.
9407 @item -mmax-stack-frame=@var{n}
9408 @opindex mmax-stack-frame
9409 Warn when the stack frame of a function exceeds @var{n} bytes.
9411 @item -melinux-stacksize=@var{n}
9412 @opindex melinux-stacksize
9413 Only available with the @samp{cris-axis-aout} target. Arranges for
9414 indications in the program to the kernel loader that the stack of the
9415 program should be set to @var{n} bytes.
9421 The options @option{-metrax4} and @option{-metrax100} are synonyms for
9422 @option{-march=v3} and @option{-march=v8} respectively.
9426 Enable CRIS-specific verbose debug-related information in the assembly
9427 code. This option also has the effect to turn off the @samp{#NO_APP}
9428 formatted-code indicator to the assembler at the beginning of the
9433 Do not use condition-code results from previous instruction; always emit
9434 compare and test instructions before use of condition codes.
9436 @item -mno-side-effects
9437 @opindex mno-side-effects
9438 Do not emit instructions with side-effects in addressing modes other than
9442 @itemx -mno-stack-align
9444 @itemx -mno-data-align
9445 @itemx -mconst-align
9446 @itemx -mno-const-align
9447 @opindex mstack-align
9448 @opindex mno-stack-align
9449 @opindex mdata-align
9450 @opindex mno-data-align
9451 @opindex mconst-align
9452 @opindex mno-const-align
9453 These options (no-options) arranges (eliminate arrangements) for the
9454 stack-frame, individual data and constants to be aligned for the maximum
9455 single data access size for the chosen CPU model. The default is to
9456 arrange for 32-bit alignment. ABI details such as structure layout are
9457 not affected by these options.
9465 Similar to the stack- data- and const-align options above, these options
9466 arrange for stack-frame, writable data and constants to all be 32-bit,
9467 16-bit or 8-bit aligned. The default is 32-bit alignment.
9469 @item -mno-prologue-epilogue
9470 @itemx -mprologue-epilogue
9471 @opindex mno-prologue-epilogue
9472 @opindex mprologue-epilogue
9473 With @option{-mno-prologue-epilogue}, the normal function prologue and
9474 epilogue that sets up the stack-frame are omitted and no return
9475 instructions or return sequences are generated in the code. Use this
9476 option only together with visual inspection of the compiled code: no
9477 warnings or errors are generated when call-saved registers must be saved,
9478 or storage for local variable needs to be allocated.
9484 With @option{-fpic} and @option{-fPIC}, don't generate (do generate)
9485 instruction sequences that load addresses for functions from the PLT part
9486 of the GOT rather than (traditional on other architectures) calls to the
9487 PLT. The default is @option{-mgotplt}.
9491 Legacy no-op option only recognized with the cris-axis-aout target.
9495 Legacy no-op option only recognized with the cris-axis-elf and
9496 cris-axis-linux-gnu targets.
9500 Only recognized with the cris-axis-aout target, where it selects a
9501 GNU/linux-like multilib, include files and instruction set for
9506 Legacy no-op option only recognized with the cris-axis-linux-gnu target.
9510 This option, recognized for the cris-axis-aout and cris-axis-elf arranges
9511 to link with input-output functions from a simulator library. Code,
9512 initialized data and zero-initialized data are allocated consecutively.
9516 Like @option{-sim}, but pass linker options to locate initialized data at
9517 0x40000000 and zero-initialized data at 0x80000000.
9521 @subsection MMIX Options
9522 @cindex MMIX Options
9524 These options are defined for the MMIX:
9528 @itemx -mno-libfuncs
9530 @opindex mno-libfuncs
9531 Specify that intrinsic library functions are being compiled, passing all
9532 values in registers, no matter the size.
9537 @opindex mno-epsilon
9538 Generate floating-point comparison instructions that compare with respect
9539 to the @code{rE} epsilon register.
9541 @item -mabi=mmixware
9543 @opindex mabi-mmixware
9545 Generate code that passes function parameters and return values that (in
9546 the called function) are seen as registers @code{$0} and up, as opposed to
9547 the GNU ABI which uses global registers @code{$231} and up.
9550 @itemx -mno-zero-extend
9551 @opindex mzero-extend
9552 @opindex mno-zero-extend
9553 When reading data from memory in sizes shorter than 64 bits, use (do not
9554 use) zero-extending load instructions by default, rather than
9555 sign-extending ones.
9558 @itemx -mno-knuthdiv
9560 @opindex mno-knuthdiv
9561 Make the result of a division yielding a remainder have the same sign as
9562 the divisor. With the default, @option{-mno-knuthdiv}, the sign of the
9563 remainder follows the sign of the dividend. Both methods are
9564 arithmetically valid, the latter being almost exclusively used.
9566 @item -mtoplevel-symbols
9567 @itemx -mno-toplevel-symbols
9568 @opindex mtoplevel-symbols
9569 @opindex mno-toplevel-symbols
9570 Prepend (do not prepend) a @samp{:} to all global symbols, so the assembly
9571 code can be used with the @code{PREFIX} assembly directive.
9575 Generate an executable in the ELF format, rather than the default
9576 @samp{mmo} format used by the @command{mmix} simulator.
9578 @item -mbranch-predict
9579 @itemx -mno-branch-predict
9580 @opindex mbranch-predict
9581 @opindex mno-branch-predict
9582 Use (do not use) the probable-branch instructions, when static branch
9583 prediction indicates a probable branch.
9585 @item -mbase-addresses
9586 @itemx -mno-base-addresses
9587 @opindex mbase-addresses
9588 @opindex mno-base-addresses
9589 Generate (do not generate) code that uses @emph{base addresses}. Using a
9590 base address automatically generates a request (handled by the assembler
9591 and the linker) for a constant to be set up in a global register. The
9592 register is used for one or more base address requests within the range 0
9593 to 255 from the value held in the register. The generally leads to short
9594 and fast code, but the number of different data items that can be
9595 addressed is limited. This means that a program that uses lots of static
9596 data may require @option{-mno-base-addresses}.
9599 @node PDP-11 Options
9600 @subsection PDP-11 Options
9601 @cindex PDP-11 Options
9603 These options are defined for the PDP-11:
9608 Use hardware FPP floating point. This is the default. (FIS floating
9609 point on the PDP-11/40 is not supported.)
9612 @opindex msoft-float
9613 Do not use hardware floating point.
9617 Return floating-point results in ac0 (fr0 in Unix assembler syntax).
9621 Return floating-point results in memory. This is the default.
9625 Generate code for a PDP-11/40.
9629 Generate code for a PDP-11/45. This is the default.
9633 Generate code for a PDP-11/10.
9635 @item -mbcopy-builtin
9636 @opindex bcopy-builtin
9637 Use inline @code{movstrhi} patterns for copying memory. This is the
9642 Do not use inline @code{movstrhi} patterns for copying memory.
9648 Use 16-bit @code{int}. This is the default.
9654 Use 32-bit @code{int}.
9659 @opindex mno-float32
9660 Use 64-bit @code{float}. This is the default.
9665 @opindex mno-float64
9666 Use 32-bit @code{float}.
9670 Use @code{abshi2} pattern. This is the default.
9674 Do not use @code{abshi2} pattern.
9676 @item -mbranch-expensive
9677 @opindex mbranch-expensive
9678 Pretend that branches are expensive. This is for experimenting with
9679 code generation only.
9681 @item -mbranch-cheap
9682 @opindex mbranch-cheap
9683 Do not pretend that branches are expensive. This is the default.
9687 Generate code for a system with split I&D.
9691 Generate code for a system without split I&D. This is the default.
9695 Use Unix assembler syntax. This is the default when configured for
9700 Use DEC assembler syntax. This is the default when configured for any
9701 PDP-11 target other than @samp{pdp11-*-bsd}.
9704 @node Xstormy16 Options
9705 @subsection Xstormy16 Options
9706 @cindex Xstormy16 Options
9708 These options are defined for Xstormy16:
9713 Choose startup files and linker script suitable for the simulator.
9716 @node Xtensa Options
9717 @subsection Xtensa Options
9718 @cindex Xtensa Options
9720 The Xtensa architecture is designed to support many different
9721 configurations. The compiler's default options can be set to match a
9722 particular Xtensa configuration by copying a configuration file into the
9723 GCC sources when building GCC@. The options below may be used to
9724 override the default options.
9728 @itemx -mlittle-endian
9729 @opindex mbig-endian
9730 @opindex mlittle-endian
9731 Specify big-endian or little-endian byte ordering for the target Xtensa
9737 @opindex mno-density
9738 Enable or disable use of the optional Xtensa code density instructions.
9744 Enable or disable use of the Xtensa MAC16 option. When enabled, GCC
9745 will generate MAC16 instructions from standard C code, with the
9746 limitation that it will use neither the MR register file nor any
9747 instruction that operates on the MR registers. When this option is
9748 disabled, GCC will translate 16-bit multiply/accumulate operations to a
9749 combination of core instructions and library calls, depending on whether
9750 any other multiplier options are enabled.
9756 Enable or disable use of the 16-bit integer multiplier option. When
9757 enabled, the compiler will generate 16-bit multiply instructions for
9758 multiplications of 16 bits or smaller in standard C code. When this
9759 option is disabled, the compiler will either use 32-bit multiply or
9760 MAC16 instructions if they are available or generate library calls to
9761 perform the multiply operations using shifts and adds.
9767 Enable or disable use of the 32-bit integer multiplier option. When
9768 enabled, the compiler will generate 32-bit multiply instructions for
9769 multiplications of 32 bits or smaller in standard C code. When this
9770 option is disabled, the compiler will generate library calls to perform
9771 the multiply operations using either shifts and adds or 16-bit multiply
9772 instructions if they are available.
9778 Enable or disable use of the optional normalization shift amount
9779 (@code{NSA}) instructions to implement the built-in @code{ffs} function.
9785 Enable or disable use of the optional minimum and maximum value
9792 Enable or disable use of the optional sign extend (@code{SEXT})
9796 @itemx -mno-booleans
9798 @opindex mno-booleans
9799 Enable or disable support for the boolean register file used by Xtensa
9800 coprocessors. This is not typically useful by itself but may be
9801 required for other options that make use of the boolean registers (e.g.,
9802 the floating-point option).
9806 @opindex mhard-float
9807 @opindex msoft-float
9808 Enable or disable use of the floating-point option. When enabled, GCC
9809 generates floating-point instructions for 32-bit @code{float}
9810 operations. When this option is disabled, GCC generates library calls
9811 to emulate 32-bit floating-point operations using integer instructions.
9812 Regardless of this option, 64-bit @code{double} operations are always
9813 emulated with calls to library functions.
9816 @itemx -mno-fused-madd
9817 @opindex mfused-madd
9818 @opindex mno-fused-madd
9819 Enable or disable use of fused multiply/add and multiply/subtract
9820 instructions in the floating-point option. This has no effect if the
9821 floating-point option is not also enabled. Disabling fused multiply/add
9822 and multiply/subtract instructions forces the compiler to use separate
9823 instructions for the multiply and add/subtract operations. This may be
9824 desirable in some cases where strict IEEE 754-compliant results are
9825 required: the fused multiply add/subtract instructions do not round the
9826 intermediate result, thereby producing results with @emph{more} bits of
9827 precision than specified by the IEEE standard. Disabling fused multiply
9828 add/subtract instructions also ensures that the program output is not
9829 sensitive to the compiler's ability to combine multiply and add/subtract
9832 @item -mserialize-volatile
9833 @itemx -mno-serialize-volatile
9834 @opindex mserialize-volatile
9835 @opindex mno-serialize-volatile
9836 When this option is enabled, GCC inserts @code{MEMW} instructions before
9837 @code{volatile} memory references to guarantee sequential consistency.
9838 The default is @option{-mserialize-volatile}. Use
9839 @option{-mno-serialize-volatile} to omit the @code{MEMW} instructions.
9841 @item -mtext-section-literals
9842 @itemx -mno-text-section-literals
9843 @opindex mtext-section-literals
9844 @opindex mno-text-section-literals
9845 Control the treatment of literal pools. The default is
9846 @option{-mno-text-section-literals}, which places literals in a separate
9847 section in the output file. This allows the literal pool to be placed
9848 in a data RAM/ROM, and it also allows the linker to combine literal
9849 pools from separate object files to remove redundant literals and
9850 improve code size. With @option{-mtext-section-literals}, the literals
9851 are interspersed in the text section in order to keep them as close as
9852 possible to their references. This may be necessary for large assembly
9855 @item -mtarget-align
9856 @itemx -mno-target-align
9857 @opindex mtarget-align
9858 @opindex mno-target-align
9859 When this option is enabled, GCC instructs the assembler to
9860 automatically align instructions to reduce branch penalties at the
9861 expense of some code density. The assembler attempts to widen density
9862 instructions to align branch targets and the instructions following call
9863 instructions. If there are not enough preceding safe density
9864 instructions to align a target, no widening will be performed. The
9865 default is @option{-mtarget-align}. These options do not affect the
9866 treatment of auto-aligned instructions like @code{LOOP}, which the
9867 assembler will always align, either by widening density instructions or
9868 by inserting no-op instructions.
9871 @itemx -mno-longcalls
9873 @opindex mno-longcalls
9874 When this option is enabled, GCC instructs the assembler to translate
9875 direct calls to indirect calls unless it can determine that the target
9876 of a direct call is in the range allowed by the call instruction. This
9877 translation typically occurs for calls to functions in other source
9878 files. Specifically, the assembler translates a direct @code{CALL}
9879 instruction into an @code{L32R} followed by a @code{CALLX} instruction.
9880 The default is @option{-mno-longcalls}. This option should be used in
9881 programs where the call target can potentially be out of range. This
9882 option is implemented in the assembler, not the compiler, so the
9883 assembly code generated by GCC will still show direct call
9884 instructions---look at the disassembled object code to see the actual
9885 instructions. Note that the assembler will use an indirect call for
9886 every cross-file call, not just those that really will be out of range.
9889 @node Code Gen Options
9890 @section Options for Code Generation Conventions
9891 @cindex code generation conventions
9892 @cindex options, code generation
9893 @cindex run-time options
9895 These machine-independent options control the interface conventions
9896 used in code generation.
9898 Most of them have both positive and negative forms; the negative form
9899 of @option{-ffoo} would be @option{-fno-foo}. In the table below, only
9900 one of the forms is listed---the one which is not the default. You
9901 can figure out the other form by either removing @samp{no-} or adding
9906 @opindex fexceptions
9907 Enable exception handling. Generates extra code needed to propagate
9908 exceptions. For some targets, this implies GCC will generate frame
9909 unwind information for all functions, which can produce significant data
9910 size overhead, although it does not affect execution. If you do not
9911 specify this option, GCC will enable it by default for languages like
9912 C++ which normally require exception handling, and disable it for
9913 languages like C that do not normally require it. However, you may need
9914 to enable this option when compiling C code that needs to interoperate
9915 properly with exception handlers written in C++. You may also wish to
9916 disable this option if you are compiling older C++ programs that don't
9917 use exception handling.
9919 @item -fnon-call-exceptions
9920 @opindex fnon-call-exceptions
9921 Generate code that allows trapping instructions to throw exceptions.
9922 Note that this requires platform-specific runtime support that does
9923 not exist everywhere. Moreover, it only allows @emph{trapping}
9924 instructions to throw exceptions, i.e.@: memory references or floating
9925 point instructions. It does not allow exceptions to be thrown from
9926 arbitrary signal handlers such as @code{SIGALRM}.
9928 @item -funwind-tables
9929 @opindex funwind-tables
9930 Similar to @option{-fexceptions}, except that it will just generate any needed
9931 static data, but will not affect the generated code in any other way.
9932 You will normally not enable this option; instead, a language processor
9933 that needs this handling would enable it on your behalf.
9935 @item -fasynchronous-unwind-tables
9936 @opindex funwind-tables
9937 Generate unwind table in dwarf2 format, if supported by target machine. The
9938 table is exact at each instruction boundary, so it can be used for stack
9939 unwinding from asynchronous events (such as debugger or garbage collector).
9941 @item -fpcc-struct-return
9942 @opindex fpcc-struct-return
9943 Return ``short'' @code{struct} and @code{union} values in memory like
9944 longer ones, rather than in registers. This convention is less
9945 efficient, but it has the advantage of allowing intercallability between
9946 GCC-compiled files and files compiled with other compilers.
9948 The precise convention for returning structures in memory depends
9949 on the target configuration macros.
9951 Short structures and unions are those whose size and alignment match
9952 that of some integer type.
9954 @item -freg-struct-return
9955 @opindex freg-struct-return
9956 Return @code{struct} and @code{union} values in registers when possible.
9957 This is more efficient for small structures than
9958 @option{-fpcc-struct-return}.
9960 If you specify neither @option{-fpcc-struct-return} nor
9961 @option{-freg-struct-return}, GCC defaults to whichever convention is
9962 standard for the target. If there is no standard convention, GCC
9963 defaults to @option{-fpcc-struct-return}, except on targets where GCC is
9964 the principal compiler. In those cases, we can choose the standard, and
9965 we chose the more efficient register return alternative.
9968 @opindex fshort-enums
9969 Allocate to an @code{enum} type only as many bytes as it needs for the
9970 declared range of possible values. Specifically, the @code{enum} type
9971 will be equivalent to the smallest integer type which has enough room.
9973 @item -fshort-double
9974 @opindex fshort-double
9975 Use the same size for @code{double} as for @code{float}.
9978 @opindex fshared-data
9979 Requests that the data and non-@code{const} variables of this
9980 compilation be shared data rather than private data. The distinction
9981 makes sense only on certain operating systems, where shared data is
9982 shared between processes running the same program, while private data
9983 exists in one copy per process.
9987 In C, allocate even uninitialized global variables in the data section of the
9988 object file, rather than generating them as common blocks. This has the
9989 effect that if the same variable is declared (without @code{extern}) in
9990 two different compilations, you will get an error when you link them.
9991 The only reason this might be useful is if you wish to verify that the
9992 program will work on other systems which always work this way.
9996 Ignore the @samp{#ident} directive.
9998 @item -fno-gnu-linker
9999 @opindex fno-gnu-linker
10000 Do not output global initializations (such as C++ constructors and
10001 destructors) in the form used by the GNU linker (on systems where the GNU
10002 linker is the standard method of handling them). Use this option when
10003 you want to use a non-GNU linker, which also requires using the
10004 @command{collect2} program to make sure the system linker includes
10005 constructors and destructors. (@command{collect2} is included in the GCC
10006 distribution.) For systems which @emph{must} use @command{collect2}, the
10007 compiler driver @command{gcc} is configured to do this automatically.
10009 @item -finhibit-size-directive
10010 @opindex finhibit-size-directive
10011 Don't output a @code{.size} assembler directive, or anything else that
10012 would cause trouble if the function is split in the middle, and the
10013 two halves are placed at locations far apart in memory. This option is
10014 used when compiling @file{crtstuff.c}; you should not need to use it
10017 @item -fverbose-asm
10018 @opindex fverbose-asm
10019 Put extra commentary information in the generated assembly code to
10020 make it more readable. This option is generally only of use to those
10021 who actually need to read the generated assembly code (perhaps while
10022 debugging the compiler itself).
10024 @option{-fno-verbose-asm}, the default, causes the
10025 extra information to be omitted and is useful when comparing two assembler
10030 Consider all memory references through pointers to be volatile.
10032 @item -fvolatile-global
10033 @opindex fvolatile-global
10034 Consider all memory references to extern and global data items to
10035 be volatile. GCC does not consider static data items to be volatile
10036 because of this switch.
10038 @item -fvolatile-static
10039 @opindex fvolatile-static
10040 Consider all memory references to static data to be volatile.
10044 @cindex global offset table
10046 Generate position-independent code (PIC) suitable for use in a shared
10047 library, if supported for the target machine. Such code accesses all
10048 constant addresses through a global offset table (GOT)@. The dynamic
10049 loader resolves the GOT entries when the program starts (the dynamic
10050 loader is not part of GCC; it is part of the operating system). If
10051 the GOT size for the linked executable exceeds a machine-specific
10052 maximum size, you get an error message from the linker indicating that
10053 @option{-fpic} does not work; in that case, recompile with @option{-fPIC}
10054 instead. (These maximums are 16k on the m88k, 8k on the Sparc, and 32k
10055 on the m68k and RS/6000. The 386 has no such limit.)
10057 Position-independent code requires special support, and therefore works
10058 only on certain machines. For the 386, GCC supports PIC for System V
10059 but not for the Sun 386i. Code generated for the IBM RS/6000 is always
10060 position-independent.
10064 If supported for the target machine, emit position-independent code,
10065 suitable for dynamic linking and avoiding any limit on the size of the
10066 global offset table. This option makes a difference on the m68k, m88k,
10069 Position-independent code requires special support, and therefore works
10070 only on certain machines.
10072 @item -ffixed-@var{reg}
10074 Treat the register named @var{reg} as a fixed register; generated code
10075 should never refer to it (except perhaps as a stack pointer, frame
10076 pointer or in some other fixed role).
10078 @var{reg} must be the name of a register. The register names accepted
10079 are machine-specific and are defined in the @code{REGISTER_NAMES}
10080 macro in the machine description macro file.
10082 This flag does not have a negative form, because it specifies a
10085 @item -fcall-used-@var{reg}
10086 @opindex fcall-used
10087 Treat the register named @var{reg} as an allocable register that is
10088 clobbered by function calls. It may be allocated for temporaries or
10089 variables that do not live across a call. Functions compiled this way
10090 will not save and restore the register @var{reg}.
10092 It is an error to used this flag with the frame pointer or stack pointer.
10093 Use of this flag for other registers that have fixed pervasive roles in
10094 the machine's execution model will produce disastrous results.
10096 This flag does not have a negative form, because it specifies a
10099 @item -fcall-saved-@var{reg}
10100 @opindex fcall-saved
10101 Treat the register named @var{reg} as an allocable register saved by
10102 functions. It may be allocated even for temporaries or variables that
10103 live across a call. Functions compiled this way will save and restore
10104 the register @var{reg} if they use it.
10106 It is an error to used this flag with the frame pointer or stack pointer.
10107 Use of this flag for other registers that have fixed pervasive roles in
10108 the machine's execution model will produce disastrous results.
10110 A different sort of disaster will result from the use of this flag for
10111 a register in which function values may be returned.
10113 This flag does not have a negative form, because it specifies a
10116 @item -fpack-struct
10117 @opindex fpack-struct
10118 Pack all structure members together without holes. Usually you would
10119 not want to use this option, since it makes the code suboptimal, and
10120 the offsets of structure members won't agree with system libraries.
10122 @item -finstrument-functions
10123 @opindex finstrument-functions
10124 Generate instrumentation calls for entry and exit to functions. Just
10125 after function entry and just before function exit, the following
10126 profiling functions will be called with the address of the current
10127 function and its call site. (On some platforms,
10128 @code{__builtin_return_address} does not work beyond the current
10129 function, so the call site information may not be available to the
10130 profiling functions otherwise.)
10133 void __cyg_profile_func_enter (void *this_fn,
10135 void __cyg_profile_func_exit (void *this_fn,
10139 The first argument is the address of the start of the current function,
10140 which may be looked up exactly in the symbol table.
10142 This instrumentation is also done for functions expanded inline in other
10143 functions. The profiling calls will indicate where, conceptually, the
10144 inline function is entered and exited. This means that addressable
10145 versions of such functions must be available. If all your uses of a
10146 function are expanded inline, this may mean an additional expansion of
10147 code size. If you use @samp{extern inline} in your C code, an
10148 addressable version of such functions must be provided. (This is
10149 normally the case anyways, but if you get lucky and the optimizer always
10150 expands the functions inline, you might have gotten away without
10151 providing static copies.)
10153 A function may be given the attribute @code{no_instrument_function}, in
10154 which case this instrumentation will not be done. This can be used, for
10155 example, for the profiling functions listed above, high-priority
10156 interrupt routines, and any functions from which the profiling functions
10157 cannot safely be called (perhaps signal handlers, if the profiling
10158 routines generate output or allocate memory).
10160 @item -fstack-check
10161 @opindex fstack-check
10162 Generate code to verify that you do not go beyond the boundary of the
10163 stack. You should specify this flag if you are running in an
10164 environment with multiple threads, but only rarely need to specify it in
10165 a single-threaded environment since stack overflow is automatically
10166 detected on nearly all systems if there is only one stack.
10168 Note that this switch does not actually cause checking to be done; the
10169 operating system must do that. The switch causes generation of code
10170 to ensure that the operating system sees the stack being extended.
10172 @item -fstack-limit-register=@var{reg}
10173 @itemx -fstack-limit-symbol=@var{sym}
10174 @itemx -fno-stack-limit
10175 @opindex fstack-limit-register
10176 @opindex fstack-limit-symbol
10177 @opindex fno-stack-limit
10178 Generate code to ensure that the stack does not grow beyond a certain value,
10179 either the value of a register or the address of a symbol. If the stack
10180 would grow beyond the value, a signal is raised. For most targets,
10181 the signal is raised before the stack overruns the boundary, so
10182 it is possible to catch the signal without taking special precautions.
10184 For instance, if the stack starts at absolute address @samp{0x80000000}
10185 and grows downwards, you can use the flags
10186 @option{-fstack-limit-symbol=__stack_limit} and
10187 @option{-Wl,--defsym,__stack_limit=0x7ffe0000} to enforce a stack limit
10188 of 128KB@. Note that this may only work with the GNU linker.
10190 @cindex aliasing of parameters
10191 @cindex parameters, aliased
10192 @item -fargument-alias
10193 @itemx -fargument-noalias
10194 @itemx -fargument-noalias-global
10195 @opindex fargument-alias
10196 @opindex fargument-noalias
10197 @opindex fargument-noalias-global
10198 Specify the possible relationships among parameters and between
10199 parameters and global data.
10201 @option{-fargument-alias} specifies that arguments (parameters) may
10202 alias each other and may alias global storage.@*
10203 @option{-fargument-noalias} specifies that arguments do not alias
10204 each other, but may alias global storage.@*
10205 @option{-fargument-noalias-global} specifies that arguments do not
10206 alias each other and do not alias global storage.
10208 Each language will automatically use whatever option is required by
10209 the language standard. You should not need to use these options yourself.
10211 @item -fleading-underscore
10212 @opindex fleading-underscore
10213 This option and its counterpart, @option{-fno-leading-underscore}, forcibly
10214 change the way C symbols are represented in the object file. One use
10215 is to help link with legacy assembly code.
10217 Be warned that you should know what you are doing when invoking this
10218 option, and that not all targets provide complete support for it.
10223 @node Environment Variables
10224 @section Environment Variables Affecting GCC
10225 @cindex environment variables
10227 @c man begin ENVIRONMENT
10229 This section describes several environment variables that affect how GCC
10230 operates. Some of them work by specifying directories or prefixes to use
10231 when searching for various kinds of files. Some are used to specify other
10232 aspects of the compilation environment.
10234 Note that you can also specify places to search using options such as
10235 @option{-B}, @option{-I} and @option{-L} (@pxref{Directory Options}). These
10236 take precedence over places specified using environment variables, which
10237 in turn take precedence over those specified by the configuration of GCC@.
10238 @xref{Driver,, Controlling the Compilation Driver @file{gcc}, gccint,
10239 GNU Compiler Collection (GCC) Internals}.
10244 @c @itemx LC_COLLATE
10246 @c @itemx LC_MONETARY
10247 @c @itemx LC_NUMERIC
10252 @c @findex LC_COLLATE
10253 @findex LC_MESSAGES
10254 @c @findex LC_MONETARY
10255 @c @findex LC_NUMERIC
10259 These environment variables control the way that GCC uses
10260 localization information that allow GCC to work with different
10261 national conventions. GCC inspects the locale categories
10262 @env{LC_CTYPE} and @env{LC_MESSAGES} if it has been configured to do
10263 so. These locale categories can be set to any value supported by your
10264 installation. A typical value is @samp{en_UK} for English in the United
10267 The @env{LC_CTYPE} environment variable specifies character
10268 classification. GCC uses it to determine the character boundaries in
10269 a string; this is needed for some multibyte encodings that contain quote
10270 and escape characters that would otherwise be interpreted as a string
10273 The @env{LC_MESSAGES} environment variable specifies the language to
10274 use in diagnostic messages.
10276 If the @env{LC_ALL} environment variable is set, it overrides the value
10277 of @env{LC_CTYPE} and @env{LC_MESSAGES}; otherwise, @env{LC_CTYPE}
10278 and @env{LC_MESSAGES} default to the value of the @env{LANG}
10279 environment variable. If none of these variables are set, GCC
10280 defaults to traditional C English behavior.
10284 If @env{TMPDIR} is set, it specifies the directory to use for temporary
10285 files. GCC uses temporary files to hold the output of one stage of
10286 compilation which is to be used as input to the next stage: for example,
10287 the output of the preprocessor, which is the input to the compiler
10290 @item GCC_EXEC_PREFIX
10291 @findex GCC_EXEC_PREFIX
10292 If @env{GCC_EXEC_PREFIX} is set, it specifies a prefix to use in the
10293 names of the subprograms executed by the compiler. No slash is added
10294 when this prefix is combined with the name of a subprogram, but you can
10295 specify a prefix that ends with a slash if you wish.
10297 If @env{GCC_EXEC_PREFIX} is not set, GCC will attempt to figure out
10298 an appropriate prefix to use based on the pathname it was invoked with.
10300 If GCC cannot find the subprogram using the specified prefix, it
10301 tries looking in the usual places for the subprogram.
10303 The default value of @env{GCC_EXEC_PREFIX} is
10304 @file{@var{prefix}/lib/gcc-lib/} where @var{prefix} is the value
10305 of @code{prefix} when you ran the @file{configure} script.
10307 Other prefixes specified with @option{-B} take precedence over this prefix.
10309 This prefix is also used for finding files such as @file{crt0.o} that are
10312 In addition, the prefix is used in an unusual way in finding the
10313 directories to search for header files. For each of the standard
10314 directories whose name normally begins with @samp{/usr/local/lib/gcc-lib}
10315 (more precisely, with the value of @env{GCC_INCLUDE_DIR}), GCC tries
10316 replacing that beginning with the specified prefix to produce an
10317 alternate directory name. Thus, with @option{-Bfoo/}, GCC will search
10318 @file{foo/bar} where it would normally search @file{/usr/local/lib/bar}.
10319 These alternate directories are searched first; the standard directories
10322 @item COMPILER_PATH
10323 @findex COMPILER_PATH
10324 The value of @env{COMPILER_PATH} is a colon-separated list of
10325 directories, much like @env{PATH}. GCC tries the directories thus
10326 specified when searching for subprograms, if it can't find the
10327 subprograms using @env{GCC_EXEC_PREFIX}.
10330 @findex LIBRARY_PATH
10331 The value of @env{LIBRARY_PATH} is a colon-separated list of
10332 directories, much like @env{PATH}. When configured as a native compiler,
10333 GCC tries the directories thus specified when searching for special
10334 linker files, if it can't find them using @env{GCC_EXEC_PREFIX}. Linking
10335 using GCC also uses these directories when searching for ordinary
10336 libraries for the @option{-l} option (but directories specified with
10337 @option{-L} come first).
10339 @item C_INCLUDE_PATH
10340 @itemx CPLUS_INCLUDE_PATH
10341 @itemx OBJC_INCLUDE_PATH
10342 @findex C_INCLUDE_PATH
10343 @findex CPLUS_INCLUDE_PATH
10344 @findex OBJC_INCLUDE_PATH
10345 @c @itemx OBJCPLUS_INCLUDE_PATH
10346 These environment variables pertain to particular languages. Each
10347 variable's value is a colon-separated list of directories, much like
10348 @env{PATH}. When GCC searches for header files, it tries the
10349 directories listed in the variable for the language you are using, after
10350 the directories specified with @option{-I} but before the standard header
10353 @item DEPENDENCIES_OUTPUT
10354 @findex DEPENDENCIES_OUTPUT
10355 @cindex dependencies for make as output
10356 If this variable is set, its value specifies how to output dependencies
10357 for Make based on the header files processed by the compiler. This
10358 output looks much like the output from the @option{-M} option
10359 (@pxref{Preprocessor Options}), but it goes to a separate file, and is
10360 in addition to the usual results of compilation.
10362 The value of @env{DEPENDENCIES_OUTPUT} can be just a file name, in
10363 which case the Make rules are written to that file, guessing the target
10364 name from the source file name. Or the value can have the form
10365 @samp{@var{file} @var{target}}, in which case the rules are written to
10366 file @var{file} using @var{target} as the target name.
10370 @cindex locale definition
10371 This variable is used to pass locale information to the compiler. One way in
10372 which this information is used is to determine the character set to be used
10373 when character literals, string literals and comments are parsed in C and C++.
10374 When the compiler is configured to allow multibyte characters,
10375 the following values for @env{LANG} are recognized:
10379 Recognize JIS characters.
10381 Recognize SJIS characters.
10383 Recognize EUCJP characters.
10386 If @env{LANG} is not defined, or if it has some other value, then the
10387 compiler will use mblen and mbtowc as defined by the default locale to
10388 recognize and translate multibyte characters.
10393 @node Running Protoize
10394 @section Running Protoize
10396 The program @code{protoize} is an optional part of GCC@. You can use
10397 it to add prototypes to a program, thus converting the program to ISO
10398 C in one respect. The companion program @code{unprotoize} does the
10399 reverse: it removes argument types from any prototypes that are found.
10401 When you run these programs, you must specify a set of source files as
10402 command line arguments. The conversion programs start out by compiling
10403 these files to see what functions they define. The information gathered
10404 about a file @var{foo} is saved in a file named @file{@var{foo}.X}.
10406 After scanning comes actual conversion. The specified files are all
10407 eligible to be converted; any files they include (whether sources or
10408 just headers) are eligible as well.
10410 But not all the eligible files are converted. By default,
10411 @code{protoize} and @code{unprotoize} convert only source and header
10412 files in the current directory. You can specify additional directories
10413 whose files should be converted with the @option{-d @var{directory}}
10414 option. You can also specify particular files to exclude with the
10415 @option{-x @var{file}} option. A file is converted if it is eligible, its
10416 directory name matches one of the specified directory names, and its
10417 name within the directory has not been excluded.
10419 Basic conversion with @code{protoize} consists of rewriting most
10420 function definitions and function declarations to specify the types of
10421 the arguments. The only ones not rewritten are those for varargs
10424 @code{protoize} optionally inserts prototype declarations at the
10425 beginning of the source file, to make them available for any calls that
10426 precede the function's definition. Or it can insert prototype
10427 declarations with block scope in the blocks where undeclared functions
10430 Basic conversion with @code{unprotoize} consists of rewriting most
10431 function declarations to remove any argument types, and rewriting
10432 function definitions to the old-style pre-ISO form.
10434 Both conversion programs print a warning for any function declaration or
10435 definition that they can't convert. You can suppress these warnings
10438 The output from @code{protoize} or @code{unprotoize} replaces the
10439 original source file. The original file is renamed to a name ending
10440 with @samp{.save} (for DOS, the saved filename ends in @samp{.sav}
10441 without the original @samp{.c} suffix). If the @samp{.save} (@samp{.sav}
10442 for DOS) file already exists, then the source file is simply discarded.
10444 @code{protoize} and @code{unprotoize} both depend on GCC itself to
10445 scan the program and collect information about the functions it uses.
10446 So neither of these programs will work until GCC is installed.
10448 Here is a table of the options you can use with @code{protoize} and
10449 @code{unprotoize}. Each option works with both programs unless
10453 @item -B @var{directory}
10454 Look for the file @file{SYSCALLS.c.X} in @var{directory}, instead of the
10455 usual directory (normally @file{/usr/local/lib}). This file contains
10456 prototype information about standard system functions. This option
10457 applies only to @code{protoize}.
10459 @item -c @var{compilation-options}
10460 Use @var{compilation-options} as the options when running @code{gcc} to
10461 produce the @samp{.X} files. The special option @option{-aux-info} is
10462 always passed in addition, to tell @code{gcc} to write a @samp{.X} file.
10464 Note that the compilation options must be given as a single argument to
10465 @code{protoize} or @code{unprotoize}. If you want to specify several
10466 @code{gcc} options, you must quote the entire set of compilation options
10467 to make them a single word in the shell.
10469 There are certain @code{gcc} arguments that you cannot use, because they
10470 would produce the wrong kind of output. These include @option{-g},
10471 @option{-O}, @option{-c}, @option{-S}, and @option{-o} If you include these in
10472 the @var{compilation-options}, they are ignored.
10475 Rename files to end in @samp{.C} (@samp{.cc} for DOS-based file
10476 systems) instead of @samp{.c}. This is convenient if you are converting
10477 a C program to C++. This option applies only to @code{protoize}.
10480 Add explicit global declarations. This means inserting explicit
10481 declarations at the beginning of each source file for each function
10482 that is called in the file and was not declared. These declarations
10483 precede the first function definition that contains a call to an
10484 undeclared function. This option applies only to @code{protoize}.
10486 @item -i @var{string}
10487 Indent old-style parameter declarations with the string @var{string}.
10488 This option applies only to @code{protoize}.
10490 @code{unprotoize} converts prototyped function definitions to old-style
10491 function definitions, where the arguments are declared between the
10492 argument list and the initial @samp{@{}. By default, @code{unprotoize}
10493 uses five spaces as the indentation. If you want to indent with just
10494 one space instead, use @option{-i " "}.
10497 Keep the @samp{.X} files. Normally, they are deleted after conversion
10501 Add explicit local declarations. @code{protoize} with @option{-l} inserts
10502 a prototype declaration for each function in each block which calls the
10503 function without any declaration. This option applies only to
10507 Make no real changes. This mode just prints information about the conversions
10508 that would have been done without @option{-n}.
10511 Make no @samp{.save} files. The original files are simply deleted.
10512 Use this option with caution.
10514 @item -p @var{program}
10515 Use the program @var{program} as the compiler. Normally, the name
10516 @file{gcc} is used.
10519 Work quietly. Most warnings are suppressed.
10522 Print the version number, just like @option{-v} for @code{gcc}.
10525 If you need special compiler options to compile one of your program's
10526 source files, then you should generate that file's @samp{.X} file
10527 specially, by running @code{gcc} on that source file with the
10528 appropriate options and the option @option{-aux-info}. Then run
10529 @code{protoize} on the entire set of files. @code{protoize} will use
10530 the existing @samp{.X} file because it is newer than the source file.
10534 gcc -Dfoo=bar file1.c -aux-info file1.X
10539 You need to include the special files along with the rest in the
10540 @code{protoize} command, even though their @samp{.X} files already
10541 exist, because otherwise they won't get converted.
10543 @xref{Protoize Caveats}, for more information on how to use
10544 @code{protoize} successfully.