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 -### --help --target-help --version}
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 -fms-extensions @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
174 @item C++ Language Options
175 @xref{C++ Dialect Options,,Options Controlling C++ Dialect}.
177 -fabi-version=@var{n} -fno-access-control -fcheck-new @gol
178 -fconserve-space -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 -Wabi -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 -Wundeclared-selector}
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 -Wno-div-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 -Wno-endif-labels @gol
224 -Wlarger-than-@var{len} -Wlong-long @gol
225 -Wmain -Wmissing-braces @gol
226 -Wmissing-format-attribute -Wmissing-noreturn @gol
227 -Wno-multichar -Wno-format-extra-args -Wno-format-y2k @gol
228 -Wno-import -Wnonnull -Wpacked -Wpadded @gol
229 -Wparentheses -Wpointer-arith -Wredundant-decls @gol
230 -Wreturn-type -Wsequence-point -Wshadow @gol
231 -Wsign-compare -Wstrict-aliasing @gol
232 -Wswitch -Wswitch-default -Wswitch-enum @gol
233 -Wsystem-headers -Wtrigraphs -Wundef -Wuninitialized @gol
234 -Wunknown-pragmas -Wunreachable-code @gol
235 -Wunused -Wunused-function -Wunused-label -Wunused-parameter @gol
236 -Wunused-value -Wunused-variable -Wwrite-strings}
238 @item C-only Warning Options
240 -Wbad-function-cast -Wmissing-declarations @gol
241 -Wmissing-prototypes -Wnested-externs @gol
242 -Wstrict-prototypes -Wtraditional}
244 @item Debugging Options
245 @xref{Debugging Options,,Options for Debugging Your Program or GCC}.
247 -d@var{letters} -dumpspecs -dumpmachine -dumpversion @gol
248 -fdump-unnumbered -fdump-translation-unit@r{[}-@var{n}@r{]} @gol
249 -fdump-class-hierarchy@r{[}-@var{n}@r{]} @gol
250 -fdump-tree-original@r{[}-@var{n}@r{]} -fdump-tree-optimized@r{[}-@var{n}@r{]} @gol
251 -fdump-tree-inlined@r{[}-@var{n}@r{]} @gol
252 -feliminate-dwarf2-dups -fmem-report @gol
253 -fprofile-arcs -fsched-verbose=@var{n} @gol
254 -ftest-coverage -ftime-report @gol
255 -g -g@var{level} -gcoff -gdwarf -gdwarf-1 -gdwarf-1+ -gdwarf-2 @gol
256 -ggdb -gstabs -gstabs+ -gvms -gxcoff -gxcoff+ @gol
257 -p -pg -print-file-name=@var{library} -print-libgcc-file-name @gol
258 -print-multi-directory -print-multi-lib @gol
259 -print-prog-name=@var{program} -print-search-dirs -Q @gol
262 @item Optimization Options
263 @xref{Optimize Options,,Options that Control Optimization}.
265 -falign-functions=@var{n} -falign-jumps=@var{n} @gol
266 -falign-labels=@var{n} -falign-loops=@var{n} @gol
268 -fbranch-probabilities -fcaller-saves -fcprop-registers @gol
269 -fcse-follow-jumps -fcse-skip-blocks -fdata-sections @gol
270 -fdelayed-branch -fdelete-null-pointer-checks @gol
271 -fexpensive-optimizations -ffast-math -ffloat-store @gol
272 -fforce-addr -fforce-mem -ffunction-sections @gol
273 -fgcse -fgcse-lm -fgcse-sm -floop-optimize -fcrossjumping @gol
274 -fif-conversion -fif-conversion2 @gol
275 -finline-functions -finline-limit=@var{n} -fkeep-inline-functions @gol
276 -fkeep-static-consts -fmerge-constants -fmerge-all-constants @gol
277 -fmove-all-movables -fnew-ra -fno-branch-count-reg @gol
278 -fno-default-inline -fno-defer-pop @gol
279 -fno-function-cse -fno-guess-branch-probability @gol
280 -fno-inline -fno-math-errno -fno-peephole -fno-peephole2 @gol
281 -funsafe-math-optimizations -ffinite-math-only @gol
282 -fno-trapping-math -fno-zero-initialized-in-bss @gol
283 -fomit-frame-pointer -foptimize-register-move @gol
284 -foptimize-sibling-calls -fprefetch-loop-arrays @gol
285 -freduce-all-givs -fregmove -frename-registers @gol
286 -freorder-blocks -freorder-functions @gol
287 -frerun-cse-after-loop -frerun-loop-opt @gol
288 -fschedule-insns -fschedule-insns2 @gol
289 -fno-sched-interblock -fno-sched-spec -fsched-spec-load @gol
290 -fsched-spec-load-dangerous -fsignaling-nans @gol
291 -fsingle-precision-constant -fssa -fssa-ccp -fssa-dce @gol
292 -fstrength-reduce -fstrict-aliasing -ftracer -fthread-jumps @gol
293 -ftrapv -funroll-all-loops -funroll-loops @gol
294 --param @var{name}=@var{value}
295 -O -O0 -O1 -O2 -O3 -Os}
297 @item Preprocessor Options
298 @xref{Preprocessor Options,,Options Controlling the Preprocessor}.
300 -$ -A@var{question}=@var{answer} -A-@var{question}@r{[}=@var{answer}@r{]} @gol
301 -C -dD -dI -dM -dN @gol
302 -D@var{macro}@r{[}=@var{defn}@r{]} -E -H @gol
303 -idirafter @var{dir} @gol
304 -include @var{file} -imacros @var{file} @gol
305 -iprefix @var{file} -iwithprefix @var{dir} @gol
306 -iwithprefixbefore @var{dir} -isystem @var{dir} @gol
307 -M -MM -MF -MG -MP -MQ -MT -nostdinc -P -remap @gol
308 -trigraphs -undef -U@var{macro} -Wp,@var{option}}
310 @item Assembler Option
311 @xref{Assembler Options,,Passing Options to the Assembler}.
316 @xref{Link Options,,Options for Linking}.
318 @var{object-file-name} -l@var{library} @gol
319 -nostartfiles -nodefaultlibs -nostdlib @gol
320 -s -static -static-libgcc -shared -shared-libgcc -symbolic @gol
321 -Wl,@var{option} -Xlinker @var{option} @gol
324 @item Directory Options
325 @xref{Directory Options,,Options for Directory Search}.
327 -B@var{prefix} -I@var{dir} -I- -L@var{dir} -specs=@var{file}}
330 @c I wrote this xref this way to avoid overfull hbox. -- rms
331 @xref{Target Options}.
333 -V @var{version} -b @var{machine}}
335 @item Machine Dependent Options
336 @xref{Submodel Options,,Hardware Models and Configurations}.
338 @emph{M680x0 Options}
340 -m68000 -m68020 -m68020-40 -m68020-60 -m68030 -m68040 @gol
341 -m68060 -mcpu32 -m5200 -m68881 -mbitfield -mc68000 -mc68020 @gol
342 -mfpa -mnobitfield -mrtd -mshort -msoft-float -mpcrel @gol
343 -malign-int -mstrict-align}
345 @emph{M68hc1x Options}
347 -m6811 -m6812 -m68hc11 -m68hc12 @gol
348 -mauto-incdec -minmax -mlong-calls -mshort @gol
349 -msoft-reg-count=@var{count}}
357 -mcpu=@var{cpu-type} @gol
358 -mtune=@var{cpu-type} @gol
359 -mcmodel=@var{code-model} @gol
361 -mapp-regs -mbroken-saverestore -mcypress @gol
362 -mfaster-structs -mflat @gol
363 -mfpu -mhard-float -mhard-quad-float @gol
364 -mimpure-text -mlive-g0 -mno-app-regs @gol
365 -mno-faster-structs -mno-flat -mno-fpu @gol
366 -mno-impure-text -mno-stack-bias -mno-unaligned-doubles @gol
367 -msoft-float -msoft-quad-float -msparclite -mstack-bias @gol
368 -msupersparc -munaligned-doubles -mv8}
370 @emph{Convex Options}
372 -mc1 -mc2 -mc32 -mc34 -mc38 @gol
373 -margcount -mnoargcount @gol
374 -mlong32 -mlong64 @gol
375 -mvolatile-cache -mvolatile-nocache}
377 @emph{AMD29K Options}
379 -m29000 -m29050 -mbw -mnbw -mdw -mndw @gol
380 -mlarge -mnormal -msmall @gol
381 -mkernel-registers -mno-reuse-arg-regs @gol
382 -mno-stack-check -mno-storem-bug @gol
383 -mreuse-arg-regs -msoft-float -mstack-check @gol
384 -mstorem-bug -muser-registers}
388 -mapcs-frame -mno-apcs-frame @gol
389 -mapcs-26 -mapcs-32 @gol
390 -mapcs-stack-check -mno-apcs-stack-check @gol
391 -mapcs-float -mno-apcs-float @gol
392 -mapcs-reentrant -mno-apcs-reentrant @gol
393 -msched-prolog -mno-sched-prolog @gol
394 -mlittle-endian -mbig-endian -mwords-little-endian @gol
395 -malignment-traps -mno-alignment-traps @gol
396 -msoft-float -mhard-float -mfpe @gol
397 -mthumb-interwork -mno-thumb-interwork @gol
398 -mcpu=@var{name} -march=@var{name} -mfpe=@var{name} @gol
399 -mstructure-size-boundary=@var{n} @gol
400 -mbsd -mxopen -mno-symrename @gol
401 -mabort-on-noreturn @gol
402 -mlong-calls -mno-long-calls @gol
403 -msingle-pic-base -mno-single-pic-base @gol
404 -mpic-register=@var{reg} @gol
405 -mnop-fun-dllimport @gol
406 -mpoke-function-name @gol
408 -mtpcs-frame -mtpcs-leaf-frame @gol
409 -mcaller-super-interworking -mcallee-super-interworking }
411 @emph{MN10200 Options}
415 @emph{MN10300 Options}
417 -mmult-bug -mno-mult-bug @gol
418 -mam33 -mno-am33 @gol
421 @emph{M32R/D Options}
423 -m32rx -m32r -mcode-model=@var{model-type} -msdata=@var{sdata-type} @gol
428 -m88000 -m88100 -m88110 -mbig-pic @gol
429 -mcheck-zero-division -mhandle-large-shift @gol
430 -midentify-revision -mno-check-zero-division @gol
431 -mno-ocs-debug-info -mno-ocs-frame-position @gol
432 -mno-optimize-arg-area -mno-serialize-volatile @gol
433 -mno-underscores -mocs-debug-info @gol
434 -mocs-frame-position -moptimize-arg-area @gol
435 -mserialize-volatile -mshort-data-@var{num} -msvr3 @gol
436 -msvr4 -mtrap-large-shift -muse-div-instruction @gol
437 -mversion-03.00 -mwarn-passed-structs}
439 @emph{RS/6000 and PowerPC Options}
441 -mcpu=@var{cpu-type} @gol
442 -mtune=@var{cpu-type} @gol
443 -mpower -mno-power -mpower2 -mno-power2 @gol
444 -mpowerpc -mpowerpc64 -mno-powerpc @gol
445 -maltivec -mno-altivec @gol
446 -mpowerpc-gpopt -mno-powerpc-gpopt @gol
447 -mpowerpc-gfxopt -mno-powerpc-gfxopt @gol
448 -mnew-mnemonics -mold-mnemonics @gol
449 -mfull-toc -mminimal-toc -mno-fp-in-toc -mno-sum-in-toc @gol
450 -m64 -m32 -mxl-call -mno-xl-call -mpe @gol
451 -msoft-float -mhard-float -mmultiple -mno-multiple @gol
452 -mstring -mno-string -mupdate -mno-update @gol
453 -mfused-madd -mno-fused-madd -mbit-align -mno-bit-align @gol
454 -mstrict-align -mno-strict-align -mrelocatable @gol
455 -mno-relocatable -mrelocatable-lib -mno-relocatable-lib @gol
456 -mtoc -mno-toc -mlittle -mlittle-endian -mbig -mbig-endian @gol
457 -mcall-aix -mcall-sysv -mcall-netbsd @gol
458 -maix-struct-return -msvr4-struct-return @gol
459 -mabi=altivec -mabi=no-altivec @gol
460 -mabi=spe -mabi=no-spe @gol
461 -misel=yes -misel=no @gol
462 -mprototype -mno-prototype @gol
463 -msim -mmvme -mads -myellowknife -memb -msdata @gol
464 -msdata=@var{opt} -mvxworks -mwindiss -G @var{num} -pthread}
468 -mcall-lib-mul -mfp-arg-in-fpregs -mfp-arg-in-gregs @gol
469 -mfull-fp-blocks -mhc-struct-return -min-line-mul @gol
470 -mminimum-fp-blocks -mnohc-struct-return}
474 -mabicalls -march=@var{cpu-type} -mtune=@var{cpu=type} @gol
475 -mcpu=@var{cpu-type} -membedded-data -muninit-const-in-rodata @gol
476 -membedded-pic -mfp32 -mfp64 -mfused-madd -mno-fused-madd @gol
477 -mgas -mgp32 -mgp64 @gol
478 -mgpopt -mhalf-pic -mhard-float -mint64 -mips1 @gol
479 -mips2 -mips3 -mips4 -mlong64 -mlong32 -mlong-calls -mmemcpy @gol
480 -mmips-as -mmips-tfile -mno-abicalls @gol
481 -mno-embedded-data -mno-uninit-const-in-rodata @gol
482 -mno-embedded-pic -mno-gpopt -mno-long-calls @gol
483 -mno-memcpy -mno-mips-tfile -mno-rnames -mno-stats @gol
484 -mrnames -msoft-float @gol
485 -m4650 -msingle-float -mmad @gol
486 -mstats -EL -EB -G @var{num} -nocpp @gol
487 -mabi=32 -mabi=n32 -mabi=64 -mabi=eabi @gol
488 -mfix7000 -mno-crt0 -mflush-func=@var{func} -mno-flush-func @gol
489 -mbranch-likely -mno-branch-likely}
491 @emph{i386 and x86-64 Options}
493 -mcpu=@var{cpu-type} -march=@var{cpu-type} -mfpmath=@var{unit} @gol
494 -masm=@var{dialect} -mno-fancy-math-387 @gol
495 -mno-fp-ret-in-387 -msoft-float -msvr3-shlib @gol
496 -mno-wide-multiply -mrtd -malign-double @gol
497 -mpreferred-stack-boundary=@var{num} @gol
498 -mmmx -msse -msse2 -m3dnow @gol
499 -mthreads -mno-align-stringops -minline-all-stringops @gol
500 -mpush-args -maccumulate-outgoing-args -m128bit-long-double @gol
501 -m96bit-long-double -mregparm=@var{num} -momit-leaf-frame-pointer @gol
503 -mcmodel=@var{code-model} @gol
508 -march=@var{architecture-type} @gol
509 -mbig-switch -mdisable-fpregs -mdisable-indexing @gol
510 -mfast-indirect-calls -mgas -mgnu-ld -mhp-ld @gol
511 -mjump-in-delay -mlinker-opt -mlong-calls @gol
512 -mlong-load-store -mno-big-switch -mno-disable-fpregs @gol
513 -mno-disable-indexing -mno-fast-indirect-calls -mno-gas @gol
514 -mno-jump-in-delay -mno-long-load-store @gol
515 -mno-portable-runtime -mno-soft-float @gol
516 -mno-space-regs -msoft-float -mpa-risc-1-0 @gol
517 -mpa-risc-1-1 -mpa-risc-2-0 -mportable-runtime @gol
518 -mschedule=@var{cpu-type} -mspace-regs -msio -mwsio}
520 @emph{Intel 960 Options}
522 -m@var{cpu-type} -masm-compat -mclean-linkage @gol
523 -mcode-align -mcomplex-addr -mleaf-procedures @gol
524 -mic-compat -mic2.0-compat -mic3.0-compat @gol
525 -mintel-asm -mno-clean-linkage -mno-code-align @gol
526 -mno-complex-addr -mno-leaf-procedures @gol
527 -mno-old-align -mno-strict-align -mno-tail-call @gol
528 -mnumerics -mold-align -msoft-float -mstrict-align @gol
531 @emph{DEC Alpha Options}
533 -mno-fp-regs -msoft-float -malpha-as -mgas @gol
534 -mieee -mieee-with-inexact -mieee-conformant @gol
535 -mfp-trap-mode=@var{mode} -mfp-rounding-mode=@var{mode} @gol
536 -mtrap-precision=@var{mode} -mbuild-constants @gol
537 -mcpu=@var{cpu-type} -mtune=@var{cpu-type} @gol
538 -mbwx -mmax -mfix -mcix @gol
539 -mfloat-vax -mfloat-ieee @gol
540 -mexplicit-relocs -msmall-data -mlarge-data @gol
541 -mmemory-latency=@var{time}}
543 @emph{DEC Alpha/VMS Options}
547 @emph{Clipper Options}
551 @emph{H8/300 Options}
553 -mrelax -mh -ms -mint32 -malign-300}
557 -m1 -m2 -m3 -m3e @gol
558 -m4-nofpu -m4-single-only -m4-single -m4 @gol
559 -m5-64media -m5-64media-nofpu @gol
560 -m5-32media -m5-32media-nofpu @gol
561 -m5-compact -m5-compact-nofpu @gol
562 -mb -ml -mdalign -mrelax @gol
563 -mbigtable -mfmovd -mhitachi -mnomacsave @gol
564 -mieee -misize -mpadstruct -mspace @gol
565 -mprefergot -musermode}
567 @emph{System V Options}
569 -Qy -Qn -YP,@var{paths} -Ym,@var{dir}}
574 -mmangle-cpu -mcpu=@var{cpu} -mtext=@var{text-section} @gol
575 -mdata=@var{data-section} -mrodata=@var{readonly-data-section}}
577 @emph{TMS320C3x/C4x Options}
579 -mcpu=@var{cpu} -mbig -msmall -mregparm -mmemparm @gol
580 -mfast-fix -mmpyi -mbk -mti -mdp-isr-reload @gol
581 -mrpts=@var{count} -mrptb -mdb -mloop-unsigned @gol
582 -mparallel-insns -mparallel-mpy -mpreserve-float}
586 -mlong-calls -mno-long-calls -mep -mno-ep @gol
587 -mprolog-function -mno-prolog-function -mspace @gol
588 -mtda=@var{n} -msda=@var{n} -mzda=@var{n} @gol
589 -mapp-regs -mno-app-regs @gol
590 -mdisable-callt -mno-disable-callt @gol
596 -m32032 -m32332 -m32532 -m32081 -m32381 @gol
597 -mmult-add -mnomult-add -msoft-float -mrtd -mnortd @gol
598 -mregparam -mnoregparam -msb -mnosb @gol
599 -mbitfield -mnobitfield -mhimem -mnohimem}
603 -mmcu=@var{mcu} -msize -minit-stack=@var{n} -mno-interrupts @gol
604 -mcall-prologues -mno-tablejump -mtiny-stack}
608 -mhardlit -mno-hardlit -mdiv -mno-div -mrelax-immediates @gol
609 -mno-relax-immediates -mwide-bitfields -mno-wide-bitfields @gol
610 -m4byte-functions -mno-4byte-functions -mcallgraph-data @gol
611 -mno-callgraph-data -mslow-bytes -mno-slow-bytes -mno-lsim @gol
612 -mlittle-endian -mbig-endian -m210 -m340 -mstack-increment}
616 -mlibfuncs -mno-libfuncs -mepsilon -mno-epsilon -mabi=gnu @gol
617 -mabi=mmixware -mzero-extend -mknuthdiv -mtoplevel-symbols @gol
618 -melf -mbranch-predict -mno-branch-predict -mbase-addresses @gol
619 -mno-base-addresses -msingle-exit -mno-single-exit}
623 -mbig-endian -mlittle-endian -mgnu-as -mgnu-ld -mno-pic @gol
624 -mvolatile-asm-stop -mb-step -mregister-names -mno-sdata @gol
625 -mconstant-gp -mauto-pic -minline-float-divide-min-latency @gol
626 -minline-float-divide-max-throughput -minline-int-divide-min-latency @gol
627 -minline-int-divide-max-throughput -mno-dwarf2-asm @gol
628 -mfixed-range=@var{register-range}}
632 -mextmem -mextmemory -monchip -mno-asm-optimize @gol
633 -masm-optimize -mbranch-cost=@var{n} -mcond-exec=@var{n}}
635 @emph{S/390 and zSeries Options}
637 -mhard-float -msoft-float -mbackchain -mno-backchain @gol
638 -msmall-exec -mno-small-exec -mmvcle -mno-mvcle @gol
639 -m64 -m31 -mdebug -mno-debug}
643 -mcpu=@var{cpu} -march=@var{cpu} -mtune=@var{cpu} @gol
644 -mmax-stack-frame=@var{n} -melinux-stacksize=@var{n} @gol
645 -metrax4 -metrax100 -mpdebug -mcc-init -mno-side-effects @gol
646 -mstack-align -mdata-align -mconst-align @gol
647 -m32-bit -m16-bit -m8-bit -mno-prologue-epilogue -mno-gotplt @gol
648 -melf -maout -melinux -mlinux -sim -sim2}
650 @emph{PDP-11 Options}
652 -mfpu -msoft-float -mac0 -mno-ac0 -m40 -m45 -m10 @gol
653 -mbcopy -mbcopy-builtin -mint32 -mno-int16 @gol
654 -mint16 -mno-int32 -mfloat32 -mno-float64 @gol
655 -mfloat64 -mno-float32 -mabshi -mno-abshi @gol
656 -mbranch-expensive -mbranch-cheap @gol
657 -msplit -mno-split -munix-asm -mdec-asm}
659 @emph{Xstormy16 Options}
663 @emph{Xtensa Options}
665 -mbig-endian -mlittle-endian @gol
666 -mdensity -mno-density @gol
667 -mmac16 -mno-mac16 @gol
668 -mmul16 -mno-mul16 @gol
669 -mmul32 -mno-mul32 @gol
671 -mminmax -mno-minmax @gol
672 -msext -mno-sext @gol
673 -mbooleans -mno-booleans @gol
674 -mhard-float -msoft-float @gol
675 -mfused-madd -mno-fused-madd @gol
676 -mserialize-volatile -mno-serialize-volatile @gol
677 -mtext-section-literals -mno-text-section-literals @gol
678 -mtarget-align -mno-target-align @gol
679 -mlongcalls -mno-longcalls}
681 @item Code Generation Options
682 @xref{Code Gen Options,,Options for Code Generation Conventions}.
684 -fcall-saved-@var{reg} -fcall-used-@var{reg} @gol
685 -ffixed-@var{reg} -fexceptions @gol
686 -fnon-call-exceptions -funwind-tables @gol
687 -fasynchronous-unwind-tables @gol
688 -finhibit-size-directive -finstrument-functions @gol
689 -fno-common -fno-ident -fno-gnu-linker @gol
690 -fpcc-struct-return -fpic -fPIC @gol
691 -freg-struct-return -fshared-data -fshort-enums @gol
692 -fshort-double -fshort-wchar -fvolatile @gol
693 -fvolatile-global -fvolatile-static @gol
694 -fverbose-asm -fpack-struct -fstack-check @gol
695 -fstack-limit-register=@var{reg} -fstack-limit-symbol=@var{sym} @gol
696 -fargument-alias -fargument-noalias @gol
697 -fargument-noalias-global -fleading-underscore @gol
698 -ftls-model=@var{model}}
702 * Overall Options:: Controlling the kind of output:
703 an executable, object files, assembler files,
704 or preprocessed source.
705 * C Dialect Options:: Controlling the variant of C language compiled.
706 * C++ Dialect Options:: Variations on C++.
707 * Objective-C Dialect Options:: Variations on Objective-C.
708 * Language Independent Options:: Controlling how diagnostics should be
710 * Warning Options:: How picky should the compiler be?
711 * Debugging Options:: Symbol tables, measurements, and debugging dumps.
712 * Optimize Options:: How much optimization?
713 * Preprocessor Options:: Controlling header files and macro definitions.
714 Also, getting dependency information for Make.
715 * Assembler Options:: Passing options to the assembler.
716 * Link Options:: Specifying libraries and so on.
717 * Directory Options:: Where to find header files and libraries.
718 Where to find the compiler executable files.
719 * Spec Files:: How to pass switches to sub-processes.
720 * Target Options:: Running a cross-compiler, or an old version of GCC.
723 @node Overall Options
724 @section Options Controlling the Kind of Output
726 Compilation can involve up to four stages: preprocessing, compilation
727 proper, assembly and linking, always in that order. The first three
728 stages apply to an individual source file, and end by producing an
729 object file; linking combines all the object files (those newly
730 compiled, and those specified as input) into an executable file.
732 @cindex file name suffix
733 For any given input file, the file name suffix determines what kind of
738 C source code which must be preprocessed.
741 C source code which should not be preprocessed.
744 C++ source code which should not be preprocessed.
747 Objective-C source code. Note that you must link with the library
748 @file{libobjc.a} to make an Objective-C program work.
751 Objective-C source code which should not be preprocessed.
754 C header file (not to be compiled or linked).
758 @itemx @var{file}.cxx
759 @itemx @var{file}.cpp
760 @itemx @var{file}.c++
762 C++ source code which must be preprocessed. Note that in @samp{.cxx},
763 the last two letters must both be literally @samp{x}. Likewise,
764 @samp{.C} refers to a literal capital C@.
767 @itemx @var{file}.for
768 @itemx @var{file}.FOR
769 Fortran source code which should not be preprocessed.
772 @itemx @var{file}.fpp
773 @itemx @var{file}.FPP
774 Fortran source code which must be preprocessed (with the traditional
778 Fortran source code which must be preprocessed with a RATFOR
779 preprocessor (not included with GCC)@.
781 @xref{Overall Options,,Options Controlling the Kind of Output, g77,
782 Using and Porting GNU Fortran}, for more details of the handling of
785 @c FIXME: Descriptions of Java file types.
792 Ada source code file which contains a library unit declaration (a
793 declaration of a package, subprogram, or generic, or a generic
794 instantiation), or a library unit renaming declaration (a package,
795 generic, or subprogram renaming declaration). Such files are also
798 @itemx @var{file}.adb
799 Ada source code file containing a library unit body (a subprogram or
800 package body). Such files are also called @dfn{bodies}.
802 @c GCC also knows about some suffixes for languages not yet included:
811 Assembler code which must be preprocessed.
814 An object file to be fed straight into linking.
815 Any file name with no recognized suffix is treated this way.
819 You can specify the input language explicitly with the @option{-x} option:
822 @item -x @var{language}
823 Specify explicitly the @var{language} for the following input files
824 (rather than letting the compiler choose a default based on the file
825 name suffix). This option applies to all following input files until
826 the next @option{-x} option. Possible values for @var{language} are:
828 c c-header cpp-output
830 objective-c objc-cpp-output
831 assembler assembler-with-cpp
833 f77 f77-cpp-input ratfor
839 Turn off any specification of a language, so that subsequent files are
840 handled according to their file name suffixes (as they are if @option{-x}
841 has not been used at all).
843 @item -pass-exit-codes
844 @opindex pass-exit-codes
845 Normally the @command{gcc} program will exit with the code of 1 if any
846 phase of the compiler returns a non-success return code. If you specify
847 @option{-pass-exit-codes}, the @command{gcc} program will instead return with
848 numerically highest error produced by any phase that returned an error
852 If you only want some of the stages of compilation, you can use
853 @option{-x} (or filename suffixes) to tell @command{gcc} where to start, and
854 one of the options @option{-c}, @option{-S}, or @option{-E} to say where
855 @command{gcc} is to stop. Note that some combinations (for example,
856 @samp{-x cpp-output -E}) instruct @command{gcc} to do nothing at all.
861 Compile or assemble the source files, but do not link. The linking
862 stage simply is not done. The ultimate output is in the form of an
863 object file for each source file.
865 By default, the object file name for a source file is made by replacing
866 the suffix @samp{.c}, @samp{.i}, @samp{.s}, etc., with @samp{.o}.
868 Unrecognized input files, not requiring compilation or assembly, are
873 Stop after the stage of compilation proper; do not assemble. The output
874 is in the form of an assembler code file for each non-assembler input
877 By default, the assembler file name for a source file is made by
878 replacing the suffix @samp{.c}, @samp{.i}, etc., with @samp{.s}.
880 Input files that don't require compilation are ignored.
884 Stop after the preprocessing stage; do not run the compiler proper. The
885 output is in the form of preprocessed source code, which is sent to the
888 Input files which don't require preprocessing are ignored.
890 @cindex output file option
893 Place output in file @var{file}. This applies regardless to whatever
894 sort of output is being produced, whether it be an executable file,
895 an object file, an assembler file or preprocessed C code.
897 Since only one output file can be specified, it does not make sense to
898 use @option{-o} when compiling more than one input file, unless you are
899 producing an executable file as output.
901 If @option{-o} is not specified, the default is to put an executable file
902 in @file{a.out}, the object file for @file{@var{source}.@var{suffix}} in
903 @file{@var{source}.o}, its assembler file in @file{@var{source}.s}, and
904 all preprocessed C source on standard output.
908 Print (on standard error output) the commands executed to run the stages
909 of compilation. Also print the version number of the compiler driver
910 program and of the preprocessor and the compiler proper.
914 Like @option{-v} except the commands are not executed and all command
915 arguments are quoted. This is useful for shell scripts to capture the
916 driver-generated command lines.
920 Use pipes rather than temporary files for communication between the
921 various stages of compilation. This fails to work on some systems where
922 the assembler is unable to read from a pipe; but the GNU assembler has
927 Print (on the standard output) a description of the command line options
928 understood by @command{gcc}. If the @option{-v} option is also specified
929 then @option{--help} will also be passed on to the various processes
930 invoked by @command{gcc}, so that they can display the command line options
931 they accept. If the @option{-W} option is also specified then command
932 line options which have no documentation associated with them will also
937 Print (on the standard output) a description of target specific command
938 line options for each tool.
942 Display the version number and copyrights of the invoked GCC.
946 @section Compiling C++ Programs
948 @cindex suffixes for C++ source
949 @cindex C++ source file suffixes
950 C++ source files conventionally use one of the suffixes @samp{.C},
951 @samp{.cc}, @samp{.cpp}, @samp{.c++}, @samp{.cp}, or @samp{.cxx};
952 preprocessed C++ files use the suffix @samp{.ii}. GCC recognizes
953 files with these names and compiles them as C++ programs even if you
954 call the compiler the same way as for compiling C programs (usually with
955 the name @command{gcc}).
959 However, C++ programs often require class libraries as well as a
960 compiler that understands the C++ language---and under some
961 circumstances, you might want to compile programs from standard input,
962 or otherwise without a suffix that flags them as C++ programs.
963 @command{g++} is a program that calls GCC with the default language
964 set to C++, and automatically specifies linking against the C++
965 library. On many systems, @command{g++} is also
966 installed with the name @command{c++}.
968 @cindex invoking @command{g++}
969 When you compile C++ programs, you may specify many of the same
970 command-line options that you use for compiling programs in any
971 language; or command-line options meaningful for C and related
972 languages; or options that are meaningful only for C++ programs.
973 @xref{C Dialect Options,,Options Controlling C Dialect}, for
974 explanations of options for languages related to C@.
975 @xref{C++ Dialect Options,,Options Controlling C++ Dialect}, for
976 explanations of options that are meaningful only for C++ programs.
978 @node C Dialect Options
979 @section Options Controlling C Dialect
980 @cindex dialect options
981 @cindex language dialect options
982 @cindex options, dialect
984 The following options control the dialect of C (or languages derived
985 from C, such as C++ and Objective-C) that the compiler accepts:
992 In C mode, support all ISO C90 programs. In C++ mode,
993 remove GNU extensions that conflict with ISO C++.
995 This turns off certain features of GCC that are incompatible with ISO
996 C90 (when compiling C code), or of standard C++ (when compiling C++ code),
997 such as the @code{asm} and @code{typeof} keywords, and
998 predefined macros such as @code{unix} and @code{vax} that identify the
999 type of system you are using. It also enables the undesirable and
1000 rarely used ISO trigraph feature. For the C compiler,
1001 it disables recognition of C++ style @samp{//} comments as well as
1002 the @code{inline} keyword.
1004 The alternate keywords @code{__asm__}, @code{__extension__},
1005 @code{__inline__} and @code{__typeof__} continue to work despite
1006 @option{-ansi}. You would not want to use them in an ISO C program, of
1007 course, but it is useful to put them in header files that might be included
1008 in compilations done with @option{-ansi}. Alternate predefined macros
1009 such as @code{__unix__} and @code{__vax__} are also available, with or
1010 without @option{-ansi}.
1012 The @option{-ansi} option does not cause non-ISO programs to be
1013 rejected gratuitously. For that, @option{-pedantic} is required in
1014 addition to @option{-ansi}. @xref{Warning Options}.
1016 The macro @code{__STRICT_ANSI__} is predefined when the @option{-ansi}
1017 option is used. Some header files may notice this macro and refrain
1018 from declaring certain functions or defining certain macros that the
1019 ISO standard doesn't call for; this is to avoid interfering with any
1020 programs that might use these names for other things.
1022 Functions which would normally be built in but do not have semantics
1023 defined by ISO C (such as @code{alloca} and @code{ffs}) are not built-in
1024 functions with @option{-ansi} is used. @xref{Other Builtins,,Other
1025 built-in functions provided by GCC}, for details of the functions
1030 Determine the language standard. This option is currently only
1031 supported when compiling C or C++. A value for this option must be
1032 provided; possible values are
1037 ISO C90 (same as @option{-ansi}).
1039 @item iso9899:199409
1040 ISO C90 as modified in amendment 1.
1046 ISO C99. Note that this standard is not yet fully supported; see
1047 @w{@uref{http://gcc.gnu.org/c99status.html}} for more information. The
1048 names @samp{c9x} and @samp{iso9899:199x} are deprecated.
1051 Default, ISO C90 plus GNU extensions (including some C99 features).
1055 ISO C99 plus GNU extensions. When ISO C99 is fully implemented in GCC,
1056 this will become the default. The name @samp{gnu9x} is deprecated.
1059 The 1998 ISO C++ standard plus amendments.
1062 The same as @option{-std=c++98} plus GNU extensions. This is the
1063 default for C++ code.
1066 Even when this option is not specified, you can still use some of the
1067 features of newer standards in so far as they do not conflict with
1068 previous C standards. For example, you may use @code{__restrict__} even
1069 when @option{-std=c99} is not specified.
1071 The @option{-std} options specifying some version of ISO C have the same
1072 effects as @option{-ansi}, except that features that were not in ISO C90
1073 but are in the specified version (for example, @samp{//} comments and
1074 the @code{inline} keyword in ISO C99) are not disabled.
1076 @xref{Standards,,Language Standards Supported by GCC}, for details of
1077 these standard versions.
1079 @item -aux-info @var{filename}
1081 Output to the given filename prototyped declarations for all functions
1082 declared and/or defined in a translation unit, including those in header
1083 files. This option is silently ignored in any language other than C@.
1085 Besides declarations, the file indicates, in comments, the origin of
1086 each declaration (source file and line), whether the declaration was
1087 implicit, prototyped or unprototyped (@samp{I}, @samp{N} for new or
1088 @samp{O} for old, respectively, in the first character after the line
1089 number and the colon), and whether it came from a declaration or a
1090 definition (@samp{C} or @samp{F}, respectively, in the following
1091 character). In the case of function definitions, a K&R-style list of
1092 arguments followed by their declarations is also provided, inside
1093 comments, after the declaration.
1097 Do not recognize @code{asm}, @code{inline} or @code{typeof} as a
1098 keyword, so that code can use these words as identifiers. You can use
1099 the keywords @code{__asm__}, @code{__inline__} and @code{__typeof__}
1100 instead. @option{-ansi} implies @option{-fno-asm}.
1102 In C++, this switch only affects the @code{typeof} keyword, since
1103 @code{asm} and @code{inline} are standard keywords. You may want to
1104 use the @option{-fno-gnu-keywords} flag instead, which has the same
1105 effect. In C99 mode (@option{-std=c99} or @option{-std=gnu99}), this
1106 switch only affects the @code{asm} and @code{typeof} keywords, since
1107 @code{inline} is a standard keyword in ISO C99.
1110 @itemx -fno-builtin-@var{function}
1111 @opindex fno-builtin
1112 @cindex built-in functions
1113 Don't recognize built-in functions that do not begin with
1114 @samp{__builtin_} as prefix. @xref{Other Builtins,,Other built-in
1115 functions provided by GCC}, for details of the functions affected,
1116 including those which are not built-in functions when @option{-ansi} or
1117 @option{-std} options for strict ISO C conformance are used because they
1118 do not have an ISO standard meaning.
1120 GCC normally generates special code to handle certain built-in functions
1121 more efficiently; for instance, calls to @code{alloca} may become single
1122 instructions that adjust the stack directly, and calls to @code{memcpy}
1123 may become inline copy loops. The resulting code is often both smaller
1124 and faster, but since the function calls no longer appear as such, you
1125 cannot set a breakpoint on those calls, nor can you change the behavior
1126 of the functions by linking with a different library.
1128 With the @option{-fno-builtin-@var{function}} option
1129 only the built-in function @var{function} is
1130 disabled. @var{function} must not begin with @samp{__builtin_}. If a
1131 function is named this is not built-in in this version of GCC, this
1132 option is ignored. There is no corresponding
1133 @option{-fbuiltin-@var{function}} option; if you wish to enable
1134 built-in functions selectively when using @option{-fno-builtin} or
1135 @option{-ffreestanding}, you may define macros such as:
1138 #define abs(n) __builtin_abs ((n))
1139 #define strcpy(d, s) __builtin_strcpy ((d), (s))
1144 @cindex hosted environment
1146 Assert that compilation takes place in a hosted environment. This implies
1147 @option{-fbuiltin}. A hosted environment is one in which the
1148 entire standard library is available, and in which @code{main} has a return
1149 type of @code{int}. Examples are nearly everything except a kernel.
1150 This is equivalent to @option{-fno-freestanding}.
1152 @item -ffreestanding
1153 @opindex ffreestanding
1154 @cindex hosted environment
1156 Assert that compilation takes place in a freestanding environment. This
1157 implies @option{-fno-builtin}. A freestanding environment
1158 is one in which the standard library may not exist, and program startup may
1159 not necessarily be at @code{main}. The most obvious example is an OS kernel.
1160 This is equivalent to @option{-fno-hosted}.
1162 @xref{Standards,,Language Standards Supported by GCC}, for details of
1163 freestanding and hosted environments.
1165 @item -fms-extensions
1166 @opindex fms-extensions
1167 Accept some non-standard constructs used in Microsoft header files.
1171 Support ISO C trigraphs. The @option{-ansi} option (and @option{-std}
1172 options for strict ISO C conformance) implies @option{-trigraphs}.
1174 @cindex traditional C language
1175 @cindex C language, traditional
1177 @itemx -traditional-cpp
1178 @opindex traditional-cpp
1179 @opindex traditional
1180 Formerly, these options caused GCC to attempt to emulate a pre-standard
1181 C compiler. They are now only supported with the @option{-E} switch.
1182 The preprocessor continues to support a pre-standard mode. See the GNU
1183 CPP manual for details.
1185 @item -fcond-mismatch
1186 @opindex fcond-mismatch
1187 Allow conditional expressions with mismatched types in the second and
1188 third arguments. The value of such an expression is void. This option
1189 is not supported for C++.
1191 @item -funsigned-char
1192 @opindex funsigned-char
1193 Let the type @code{char} be unsigned, like @code{unsigned char}.
1195 Each kind of machine has a default for what @code{char} should
1196 be. It is either like @code{unsigned char} by default or like
1197 @code{signed char} by default.
1199 Ideally, a portable program should always use @code{signed char} or
1200 @code{unsigned char} when it depends on the signedness of an object.
1201 But many programs have been written to use plain @code{char} and
1202 expect it to be signed, or expect it to be unsigned, depending on the
1203 machines they were written for. This option, and its inverse, let you
1204 make such a program work with the opposite default.
1206 The type @code{char} is always a distinct type from each of
1207 @code{signed char} or @code{unsigned char}, even though its behavior
1208 is always just like one of those two.
1211 @opindex fsigned-char
1212 Let the type @code{char} be signed, like @code{signed char}.
1214 Note that this is equivalent to @option{-fno-unsigned-char}, which is
1215 the negative form of @option{-funsigned-char}. Likewise, the option
1216 @option{-fno-signed-char} is equivalent to @option{-funsigned-char}.
1218 @item -fsigned-bitfields
1219 @itemx -funsigned-bitfields
1220 @itemx -fno-signed-bitfields
1221 @itemx -fno-unsigned-bitfields
1222 @opindex fsigned-bitfields
1223 @opindex funsigned-bitfields
1224 @opindex fno-signed-bitfields
1225 @opindex fno-unsigned-bitfields
1226 These options control whether a bit-field is signed or unsigned, when the
1227 declaration does not use either @code{signed} or @code{unsigned}. By
1228 default, such a bit-field is signed, because this is consistent: the
1229 basic integer types such as @code{int} are signed types.
1231 @item -fwritable-strings
1232 @opindex fwritable-strings
1233 Store string constants in the writable data segment and don't uniquize
1234 them. This is for compatibility with old programs which assume they can
1235 write into string constants.
1237 Writing into string constants is a very bad idea; ``constants'' should
1241 @node C++ Dialect Options
1242 @section Options Controlling C++ Dialect
1244 @cindex compiler options, C++
1245 @cindex C++ options, command line
1246 @cindex options, C++
1247 This section describes the command-line options that are only meaningful
1248 for C++ programs; but you can also use most of the GNU compiler options
1249 regardless of what language your program is in. For example, you
1250 might compile a file @code{firstClass.C} like this:
1253 g++ -g -frepo -O -c firstClass.C
1257 In this example, only @option{-frepo} is an option meant
1258 only for C++ programs; you can use the other options with any
1259 language supported by GCC@.
1261 Here is a list of options that are @emph{only} for compiling C++ programs:
1265 @item -fabi-version=@var{n}
1266 @opindex fabi-version
1267 Use version @var{n} of the C++ ABI. Version 1 is the version of the C++
1268 ABI that first appeared in G++ 3.2. Version 0 will always be the
1269 version that conforms most closely to the C++ ABI specification.
1270 Therefore, the ABI obtained using version 0 will change as ABI bugs are
1273 The default is version 1.
1275 @item -fno-access-control
1276 @opindex fno-access-control
1277 Turn off all access checking. This switch is mainly useful for working
1278 around bugs in the access control code.
1282 Check that the pointer returned by @code{operator new} is non-null
1283 before attempting to modify the storage allocated. The current Working
1284 Paper requires that @code{operator new} never return a null pointer, so
1285 this check is normally unnecessary.
1287 An alternative to using this option is to specify that your
1288 @code{operator new} does not throw any exceptions; if you declare it
1289 @samp{throw()}, G++ will check the return value. See also @samp{new
1292 @item -fconserve-space
1293 @opindex fconserve-space
1294 Put uninitialized or runtime-initialized global variables into the
1295 common segment, as C does. This saves space in the executable at the
1296 cost of not diagnosing duplicate definitions. If you compile with this
1297 flag and your program mysteriously crashes after @code{main()} has
1298 completed, you may have an object that is being destroyed twice because
1299 two definitions were merged.
1301 This option is no longer useful on most targets, now that support has
1302 been added for putting variables into BSS without making them common.
1304 @item -fno-const-strings
1305 @opindex fno-const-strings
1306 Give string constants type @code{char *} instead of type @code{const
1307 char *}. By default, G++ uses type @code{const char *} as required by
1308 the standard. Even if you use @option{-fno-const-strings}, you cannot
1309 actually modify the value of a string constant, unless you also use
1310 @option{-fwritable-strings}.
1312 This option might be removed in a future release of G++. For maximum
1313 portability, you should structure your code so that it works with
1314 string constants that have type @code{const char *}.
1316 @item -fdollars-in-identifiers
1317 @opindex fdollars-in-identifiers
1318 Accept @samp{$} in identifiers. You can also explicitly prohibit use of
1319 @samp{$} with the option @option{-fno-dollars-in-identifiers}. (GNU C allows
1320 @samp{$} by default on most target systems, but there are a few exceptions.)
1321 Traditional C allowed the character @samp{$} to form part of
1322 identifiers. However, ISO C and C++ forbid @samp{$} in identifiers.
1324 @item -fno-elide-constructors
1325 @opindex fno-elide-constructors
1326 The C++ standard allows an implementation to omit creating a temporary
1327 which is only used to initialize another object of the same type.
1328 Specifying this option disables that optimization, and forces G++ to
1329 call the copy constructor in all cases.
1331 @item -fno-enforce-eh-specs
1332 @opindex fno-enforce-eh-specs
1333 Don't check for violation of exception specifications at runtime. This
1334 option violates the C++ standard, but may be useful for reducing code
1335 size in production builds, much like defining @samp{NDEBUG}. The compiler
1336 will still optimize based on the exception specifications.
1338 @item -fexternal-templates
1339 @opindex fexternal-templates
1341 Cause @samp{#pragma interface} and @samp{implementation} to apply to
1342 template instantiation; template instances are emitted or not according
1343 to the location of the template definition. @xref{Template
1344 Instantiation}, for more information.
1346 This option is deprecated.
1348 @item -falt-external-templates
1349 @opindex falt-external-templates
1350 Similar to @option{-fexternal-templates}, but template instances are
1351 emitted or not according to the place where they are first instantiated.
1352 @xref{Template Instantiation}, for more information.
1354 This option is deprecated.
1357 @itemx -fno-for-scope
1359 @opindex fno-for-scope
1360 If @option{-ffor-scope} is specified, the scope of variables declared in
1361 a @i{for-init-statement} is limited to the @samp{for} loop itself,
1362 as specified by the C++ standard.
1363 If @option{-fno-for-scope} is specified, the scope of variables declared in
1364 a @i{for-init-statement} extends to the end of the enclosing scope,
1365 as was the case in old versions of G++, and other (traditional)
1366 implementations of C++.
1368 The default if neither flag is given to follow the standard,
1369 but to allow and give a warning for old-style code that would
1370 otherwise be invalid, or have different behavior.
1372 @item -fno-gnu-keywords
1373 @opindex fno-gnu-keywords
1374 Do not recognize @code{typeof} as a keyword, so that code can use this
1375 word as an identifier. You can use the keyword @code{__typeof__} instead.
1376 @option{-ansi} implies @option{-fno-gnu-keywords}.
1378 @item -fno-implicit-templates
1379 @opindex fno-implicit-templates
1380 Never emit code for non-inline templates which are instantiated
1381 implicitly (i.e.@: by use); only emit code for explicit instantiations.
1382 @xref{Template Instantiation}, for more information.
1384 @item -fno-implicit-inline-templates
1385 @opindex fno-implicit-inline-templates
1386 Don't emit code for implicit instantiations of inline templates, either.
1387 The default is to handle inlines differently so that compiles with and
1388 without optimization will need the same set of explicit instantiations.
1390 @item -fno-implement-inlines
1391 @opindex fno-implement-inlines
1392 To save space, do not emit out-of-line copies of inline functions
1393 controlled by @samp{#pragma implementation}. This will cause linker
1394 errors if these functions are not inlined everywhere they are called.
1396 @item -fms-extensions
1397 @opindex fms-extensions
1398 Disable pedantic warnings about constructs used in MFC, such as implicit
1399 int and getting a pointer to member function via non-standard syntax.
1401 @item -fno-nonansi-builtins
1402 @opindex fno-nonansi-builtins
1403 Disable built-in declarations of functions that are not mandated by
1404 ANSI/ISO C@. These include @code{ffs}, @code{alloca}, @code{_exit},
1405 @code{index}, @code{bzero}, @code{conjf}, and other related functions.
1407 @item -fno-operator-names
1408 @opindex fno-operator-names
1409 Do not treat the operator name keywords @code{and}, @code{bitand},
1410 @code{bitor}, @code{compl}, @code{not}, @code{or} and @code{xor} as
1411 synonyms as keywords.
1413 @item -fno-optional-diags
1414 @opindex fno-optional-diags
1415 Disable diagnostics that the standard says a compiler does not need to
1416 issue. Currently, the only such diagnostic issued by G++ is the one for
1417 a name having multiple meanings within a class.
1420 @opindex fpermissive
1421 Downgrade messages about nonconformant code from errors to warnings. By
1422 default, G++ effectively sets @option{-pedantic-errors} without
1423 @option{-pedantic}; this option reverses that. This behavior and this
1424 option are superseded by @option{-pedantic}, which works as it does for GNU C@.
1428 Enable automatic template instantiation at link time. This option also
1429 implies @option{-fno-implicit-templates}. @xref{Template
1430 Instantiation}, for more information.
1434 Disable generation of information about every class with virtual
1435 functions for use by the C++ runtime type identification features
1436 (@samp{dynamic_cast} and @samp{typeid}). If you don't use those parts
1437 of the language, you can save some space by using this flag. Note that
1438 exception handling uses the same information, but it will generate it as
1443 Emit statistics about front-end processing at the end of the compilation.
1444 This information is generally only useful to the G++ development team.
1446 @item -ftemplate-depth-@var{n}
1447 @opindex ftemplate-depth
1448 Set the maximum instantiation depth for template classes to @var{n}.
1449 A limit on the template instantiation depth is needed to detect
1450 endless recursions during template class instantiation. ANSI/ISO C++
1451 conforming programs must not rely on a maximum depth greater than 17.
1453 @item -fuse-cxa-atexit
1454 @opindex fuse-cxa-atexit
1455 Register destructors for objects with static storage duration with the
1456 @code{__cxa_atexit} function rather than the @code{atexit} function.
1457 This option is required for fully standards-compliant handling of static
1458 destructors, but will only work if your C library supports
1459 @code{__cxa_atexit}.
1463 Emit special relocations for vtables and virtual function references
1464 so that the linker can identify unused virtual functions and zero out
1465 vtable slots that refer to them. This is most useful with
1466 @option{-ffunction-sections} and @option{-Wl,--gc-sections}, in order to
1467 also discard the functions themselves.
1469 This optimization requires GNU as and GNU ld. Not all systems support
1470 this option. @option{-Wl,--gc-sections} is ignored without @option{-static}.
1474 Do not use weak symbol support, even if it is provided by the linker.
1475 By default, G++ will use weak symbols if they are available. This
1476 option exists only for testing, and should not be used by end-users;
1477 it will result in inferior code and has no benefits. This option may
1478 be removed in a future release of G++.
1482 Do not search for header files in the standard directories specific to
1483 C++, but do still search the other standard directories. (This option
1484 is used when building the C++ library.)
1487 In addition, these optimization, warning, and code generation options
1488 have meanings only for C++ programs:
1491 @item -fno-default-inline
1492 @opindex fno-default-inline
1493 Do not assume @samp{inline} for functions defined inside a class scope.
1494 @xref{Optimize Options,,Options That Control Optimization}. Note that these
1495 functions will have linkage like inline functions; they just won't be
1498 @item -Wabi @r{(C++ only)}
1500 Warn when G++ generates code that is probably not compatible with the
1501 vendor-neutral C++ ABI. Although an effort has been made to warn about
1502 all such cases, there are probably some cases that are not warned about,
1503 even though G++ is generating incompatible code. There may also be
1504 cases where warnings are emitted even though the code that is generated
1507 You should rewrite your code to avoid these warnings if you are
1508 concerned about the fact that code generated by G++ may not be binary
1509 compatible with code generated by other compilers.
1511 The known incompatibilites at this point include:
1516 Incorrect handling of tail-padding for bit-fields. G++ may attempt to
1517 pack data into the same byte as a base class. For example:
1520 struct A @{ virtual void f(); int f1 : 1; @};
1521 struct B : public A @{ int f2 : 1; @};
1525 In this case, G++ will place @code{B::f2} into the same byte
1526 as@code{A::f1}; other compilers will not. You can avoid this problem
1527 by explicitly padding @code{A} so that its size is a multiple of the
1528 byte size on your platform; that will cause G++ and other compilers to
1529 layout @code{B} identically.
1532 Incorrect handling of tail-padding for virtual bases. G++ does not use
1533 tail padding when laying out virtual bases. For example:
1536 struct A @{ virtual void f(); char c1; @};
1537 struct B @{ B(); char c2; @};
1538 struct C : public A, public virtual B @{@};
1542 In this case, G++ will not place @code{B} into the tail-padding for
1543 @code{A}; other compilers will. You can avoid this problem by
1544 explicitly padding @code{A} so that its size is a multiple of its
1545 alignment (ignoring virtual base classes); that will cause G++ and other
1546 compilers to layout @code{C} identically.
1549 Incorrect handling of bit-fields with declared widths greater than that
1550 of their underlying types, when the bit-fields appear in a union. For
1554 union U @{ int i : 4096; @};
1558 Assuming that an @code{int} does not have 4096 bits, G++ will make the
1559 union too small by the number of bits in an @code{int}.
1562 Empty classes can be placed at incorrect offsets. For example:
1572 struct C : public B, public A @{@};
1576 G++ will place the @code{A} base class of @code{C} at a non-zero offset;
1577 it should be placed at offset zero. G++ mistakenly believes that the
1578 @code{A} data member of @code{B} is already at offset zero.
1581 Names of template functions whose types involve @code{typename} or
1582 template template parameters can be mangled incorrectly.
1585 template <typename Q>
1586 void f(typename Q::X) @{@}
1588 template <template <typename> class Q>
1589 void f(typename Q<int>::X) @{@}
1593 Instantions of these templates may be mangled incorrectly.
1597 @item -Wctor-dtor-privacy @r{(C++ only)}
1598 @opindex Wctor-dtor-privacy
1599 Warn when a class seems unusable, because all the constructors or
1600 destructors in a class are private and the class has no friends or
1601 public static member functions. This warning is enabled by default.
1603 @item -Wnon-virtual-dtor @r{(C++ only)}
1604 @opindex Wnon-virtual-dtor
1605 Warn when a class declares a non-virtual destructor that should probably
1606 be virtual, because it looks like the class will be used polymorphically.
1607 This warning is enabled by @option{-Wall}.
1609 @item -Wreorder @r{(C++ only)}
1611 @cindex reordering, warning
1612 @cindex warning for reordering of member initializers
1613 Warn when the order of member initializers given in the code does not
1614 match the order in which they must be executed. For instance:
1620 A(): j (0), i (1) @{ @}
1624 Here the compiler will warn that the member initializers for @samp{i}
1625 and @samp{j} will be rearranged to match the declaration order of the
1626 members. This warning is enabled by @option{-Wall}.
1629 The following @option{-W@dots{}} options are not affected by @option{-Wall}.
1632 @item -Weffc++ @r{(C++ only)}
1634 Warn about violations of the following style guidelines from Scott Meyers'
1635 @cite{Effective C++} book:
1639 Item 11: Define a copy constructor and an assignment operator for classes
1640 with dynamically allocated memory.
1643 Item 12: Prefer initialization to assignment in constructors.
1646 Item 14: Make destructors virtual in base classes.
1649 Item 15: Have @code{operator=} return a reference to @code{*this}.
1652 Item 23: Don't try to return a reference when you must return an object.
1656 and about violations of the following style guidelines from Scott Meyers'
1657 @cite{More Effective C++} book:
1661 Item 6: Distinguish between prefix and postfix forms of increment and
1662 decrement operators.
1665 Item 7: Never overload @code{&&}, @code{||}, or @code{,}.
1669 If you use this option, you should be aware that the standard library
1670 headers do not obey all of these guidelines; you can use @samp{grep -v}
1671 to filter out those warnings.
1673 @item -Wno-deprecated @r{(C++ only)}
1674 @opindex Wno-deprecated
1675 Do not warn about usage of deprecated features. @xref{Deprecated Features}.
1677 @item -Wno-non-template-friend @r{(C++ only)}
1678 @opindex Wno-non-template-friend
1679 Disable warnings when non-templatized friend functions are declared
1680 within a template. With the advent of explicit template specification
1681 support in G++, if the name of the friend is an unqualified-id (i.e.,
1682 @samp{friend foo(int)}), the C++ language specification demands that the
1683 friend declare or define an ordinary, nontemplate function. (Section
1684 14.5.3). Before G++ implemented explicit specification, unqualified-ids
1685 could be interpreted as a particular specialization of a templatized
1686 function. Because this non-conforming behavior is no longer the default
1687 behavior for G++, @option{-Wnon-template-friend} allows the compiler to
1688 check existing code for potential trouble spots, and is on by default.
1689 This new compiler behavior can be turned off with
1690 @option{-Wno-non-template-friend} which keeps the conformant compiler code
1691 but disables the helpful warning.
1693 @item -Wold-style-cast @r{(C++ only)}
1694 @opindex Wold-style-cast
1695 Warn if an old-style (C-style) cast to a non-void type is used within
1696 a C++ program. The new-style casts (@samp{static_cast},
1697 @samp{reinterpret_cast}, and @samp{const_cast}) are less vulnerable to
1698 unintended effects, and much easier to grep for.
1700 @item -Woverloaded-virtual @r{(C++ only)}
1701 @opindex Woverloaded-virtual
1702 @cindex overloaded virtual fn, warning
1703 @cindex warning for overloaded virtual fn
1704 Warn when a function declaration hides virtual functions from a
1705 base class. For example, in:
1712 struct B: public A @{
1717 the @code{A} class version of @code{f} is hidden in @code{B}, and code
1725 will fail to compile.
1727 @item -Wno-pmf-conversions @r{(C++ only)}
1728 @opindex Wno-pmf-conversions
1729 Disable the diagnostic for converting a bound pointer to member function
1732 @item -Wsign-promo @r{(C++ only)}
1733 @opindex Wsign-promo
1734 Warn when overload resolution chooses a promotion from unsigned or
1735 enumeral type to a signed type over a conversion to an unsigned type of
1736 the same size. Previous versions of G++ would try to preserve
1737 unsignedness, but the standard mandates the current behavior.
1739 @item -Wsynth @r{(C++ only)}
1741 @cindex warning for synthesized methods
1742 @cindex synthesized methods, warning
1743 Warn when G++'s synthesis behavior does not match that of cfront. For
1749 A& operator = (int);
1759 In this example, G++ will synthesize a default @samp{A& operator =
1760 (const A&);}, while cfront will use the user-defined @samp{operator =}.
1763 @node Objective-C Dialect Options
1764 @section Options Controlling Objective-C Dialect
1766 @cindex compiler options, Objective-C
1767 @cindex Objective-C options, command line
1768 @cindex options, Objective-C
1769 This section describes the command-line options that are only meaningful
1770 for Objective-C programs; but you can also use most of the GNU compiler
1771 options regardless of what language your program is in. For example,
1772 you might compile a file @code{some_class.m} like this:
1775 gcc -g -fgnu-runtime -O -c some_class.m
1779 In this example, only @option{-fgnu-runtime} is an option meant only for
1780 Objective-C programs; you can use the other options with any language
1783 Here is a list of options that are @emph{only} for compiling Objective-C
1787 @item -fconstant-string-class=@var{class-name}
1788 @opindex fconstant-string-class
1789 Use @var{class-name} as the name of the class to instantiate for each
1790 literal string specified with the syntax @code{@@"@dots{}"}. The default
1791 class name is @code{NXConstantString}.
1794 @opindex fgnu-runtime
1795 Generate object code compatible with the standard GNU Objective-C
1796 runtime. This is the default for most types of systems.
1798 @item -fnext-runtime
1799 @opindex fnext-runtime
1800 Generate output compatible with the NeXT runtime. This is the default
1801 for NeXT-based systems, including Darwin and Mac OS X@. The macro
1802 @code{__NEXT_RUNTIME__} is predefined if (and only if) this option is
1807 Dump interface declarations for all classes seen in the source file to a
1808 file named @file{@var{sourcename}.decl}.
1811 @opindex Wno-protocol
1812 If a class is declared to implement a protocol, a warning is issued for
1813 every method in the protocol that is not implemented by the class. The
1814 default behavior is to issue a warning for every method not explicitly
1815 implemented in the class, even if a method implementation is inherited
1816 from the superclass. If you use the @code{-Wno-protocol} option, then
1817 methods inherited from the superclass are considered to be implemented,
1818 and no warning is issued for them.
1822 Warn if multiple methods of different types for the same selector are
1823 found during compilation. The check is performed on the list of methods
1824 in the final stage of compilation. Additionally, a check is performed
1825 that for each selector appearing in a @code{@@selector(@dots{})}
1826 expression, a corresponding method with that selector has been found
1827 during compilation. Because these checks scan the method table only at
1828 the end of compilation, these warnings are not produced if the final
1829 stage of compilation is not reached, for example because an error is
1830 found during compilation, or because the @code{-fsyntax-only} option is
1833 @item -Wundeclared-selector
1834 @opindex Wundeclared-selector
1835 Warn if a @code{@@selector(@dots{})} expression referring to an
1836 undeclared selector is found. A selector is considered undeclared if no
1837 method with that name has been declared (explicitly, in an
1838 @code{@@interface} or @code{@@protocol} declaration, or implicitly, in
1839 an @code{@@implementation} section) before the
1840 @code{@@selector(@dots{})} expression. This option always performs its
1841 checks as soon as a @code{@@selector(@dots{})} expression is found
1842 (while @code{-Wselector} only performs its checks in the final stage of
1843 compilation), and so additionally enforces the coding style convention
1844 that methods and selectors must be declared before being used.
1846 @c not documented because only avail via -Wp
1847 @c @item -print-objc-runtime-info
1851 @node Language Independent Options
1852 @section Options to Control Diagnostic Messages Formatting
1853 @cindex options to control diagnostics formatting
1854 @cindex diagnostic messages
1855 @cindex message formatting
1857 Traditionally, diagnostic messages have been formatted irrespective of
1858 the output device's aspect (e.g.@: its width, @dots{}). The options described
1859 below can be used to control the diagnostic messages formatting
1860 algorithm, e.g.@: how many characters per line, how often source location
1861 information should be reported. Right now, only the C++ front end can
1862 honor these options. However it is expected, in the near future, that
1863 the remaining front ends would be able to digest them correctly.
1866 @item -fmessage-length=@var{n}
1867 @opindex fmessage-length
1868 Try to format error messages so that they fit on lines of about @var{n}
1869 characters. The default is 72 characters for @command{g++} and 0 for the rest of
1870 the front ends supported by GCC@. If @var{n} is zero, then no
1871 line-wrapping will be done; each error message will appear on a single
1874 @opindex fdiagnostics-show-location
1875 @item -fdiagnostics-show-location=once
1876 Only meaningful in line-wrapping mode. Instructs the diagnostic messages
1877 reporter to emit @emph{once} source location information; that is, in
1878 case the message is too long to fit on a single physical line and has to
1879 be wrapped, the source location won't be emitted (as prefix) again,
1880 over and over, in subsequent continuation lines. This is the default
1883 @item -fdiagnostics-show-location=every-line
1884 Only meaningful in line-wrapping mode. Instructs the diagnostic
1885 messages reporter to emit the same source location information (as
1886 prefix) for physical lines that result from the process of breaking
1887 a message which is too long to fit on a single line.
1891 @node Warning Options
1892 @section Options to Request or Suppress Warnings
1893 @cindex options to control warnings
1894 @cindex warning messages
1895 @cindex messages, warning
1896 @cindex suppressing warnings
1898 Warnings are diagnostic messages that report constructions which
1899 are not inherently erroneous but which are risky or suggest there
1900 may have been an error.
1902 You can request many specific warnings with options beginning @samp{-W},
1903 for example @option{-Wimplicit} to request warnings on implicit
1904 declarations. Each of these specific warning options also has a
1905 negative form beginning @samp{-Wno-} to turn off warnings;
1906 for example, @option{-Wno-implicit}. This manual lists only one of the
1907 two forms, whichever is not the default.
1909 The following options control the amount and kinds of warnings produced
1910 by GCC; for further, language-specific options also refer to
1911 @ref{C++ Dialect Options} and @ref{Objective-C Dialect Options}.
1914 @cindex syntax checking
1916 @opindex fsyntax-only
1917 Check the code for syntax errors, but don't do anything beyond that.
1921 Issue all the warnings demanded by strict ISO C and ISO C++;
1922 reject all programs that use forbidden extensions, and some other
1923 programs that do not follow ISO C and ISO C++. For ISO C, follows the
1924 version of the ISO C standard specified by any @option{-std} option used.
1926 Valid ISO C and ISO C++ programs should compile properly with or without
1927 this option (though a rare few will require @option{-ansi} or a
1928 @option{-std} option specifying the required version of ISO C)@. However,
1929 without this option, certain GNU extensions and traditional C and C++
1930 features are supported as well. With this option, they are rejected.
1932 @option{-pedantic} does not cause warning messages for use of the
1933 alternate keywords whose names begin and end with @samp{__}. Pedantic
1934 warnings are also disabled in the expression that follows
1935 @code{__extension__}. However, only system header files should use
1936 these escape routes; application programs should avoid them.
1937 @xref{Alternate Keywords}.
1939 Some users try to use @option{-pedantic} to check programs for strict ISO
1940 C conformance. They soon find that it does not do quite what they want:
1941 it finds some non-ISO practices, but not all---only those for which
1942 ISO C @emph{requires} a diagnostic, and some others for which
1943 diagnostics have been added.
1945 A feature to report any failure to conform to ISO C might be useful in
1946 some instances, but would require considerable additional work and would
1947 be quite different from @option{-pedantic}. We don't have plans to
1948 support such a feature in the near future.
1950 Where the standard specified with @option{-std} represents a GNU
1951 extended dialect of C, such as @samp{gnu89} or @samp{gnu99}, there is a
1952 corresponding @dfn{base standard}, the version of ISO C on which the GNU
1953 extended dialect is based. Warnings from @option{-pedantic} are given
1954 where they are required by the base standard. (It would not make sense
1955 for such warnings to be given only for features not in the specified GNU
1956 C dialect, since by definition the GNU dialects of C include all
1957 features the compiler supports with the given option, and there would be
1958 nothing to warn about.)
1960 @item -pedantic-errors
1961 @opindex pedantic-errors
1962 Like @option{-pedantic}, except that errors are produced rather than
1967 Inhibit all warning messages.
1971 Inhibit warning messages about the use of @samp{#import}.
1973 @item -Wchar-subscripts
1974 @opindex Wchar-subscripts
1975 Warn if an array subscript has type @code{char}. This is a common cause
1976 of error, as programmers often forget that this type is signed on some
1981 Warn whenever a comment-start sequence @samp{/*} appears in a @samp{/*}
1982 comment, or whenever a Backslash-Newline appears in a @samp{//} comment.
1986 Check calls to @code{printf} and @code{scanf}, etc., to make sure that
1987 the arguments supplied have types appropriate to the format string
1988 specified, and that the conversions specified in the format string make
1989 sense. This includes standard functions, and others specified by format
1990 attributes (@pxref{Function Attributes}), in the @code{printf},
1991 @code{scanf}, @code{strftime} and @code{strfmon} (an X/Open extension,
1992 not in the C standard) families.
1994 The formats are checked against the format features supported by GNU
1995 libc version 2.2. These include all ISO C90 and C99 features, as well
1996 as features from the Single Unix Specification and some BSD and GNU
1997 extensions. Other library implementations may not support all these
1998 features; GCC does not support warning about features that go beyond a
1999 particular library's limitations. However, if @option{-pedantic} is used
2000 with @option{-Wformat}, warnings will be given about format features not
2001 in the selected standard version (but not for @code{strfmon} formats,
2002 since those are not in any version of the C standard). @xref{C Dialect
2003 Options,,Options Controlling C Dialect}.
2005 Since @option{-Wformat} also checks for null format arguments for
2006 several functions, @option{-Wformat} also implies @option{-Wnonnull}.
2008 @option{-Wformat} is included in @option{-Wall}. For more control over some
2009 aspects of format checking, the options @option{-Wno-format-y2k},
2010 @option{-Wno-format-extra-args}, @option{-Wno-format-zero-length},
2011 @option{-Wformat-nonliteral}, @option{-Wformat-security}, and
2012 @option{-Wformat=2} are available, but are not included in @option{-Wall}.
2014 @item -Wno-format-y2k
2015 @opindex Wno-format-y2k
2016 If @option{-Wformat} is specified, do not warn about @code{strftime}
2017 formats which may yield only a two-digit year.
2019 @item -Wno-format-extra-args
2020 @opindex Wno-format-extra-args
2021 If @option{-Wformat} is specified, do not warn about excess arguments to a
2022 @code{printf} or @code{scanf} format function. The C standard specifies
2023 that such arguments are ignored.
2025 Where the unused arguments lie between used arguments that are
2026 specified with @samp{$} operand number specifications, normally
2027 warnings are still given, since the implementation could not know what
2028 type to pass to @code{va_arg} to skip the unused arguments. However,
2029 in the case of @code{scanf} formats, this option will suppress the
2030 warning if the unused arguments are all pointers, since the Single
2031 Unix Specification says that such unused arguments are allowed.
2033 @item -Wno-format-zero-length
2034 @opindex Wno-format-zero-length
2035 If @option{-Wformat} is specified, do not warn about zero-length formats.
2036 The C standard specifies that zero-length formats are allowed.
2038 @item -Wformat-nonliteral
2039 @opindex Wformat-nonliteral
2040 If @option{-Wformat} is specified, also warn if the format string is not a
2041 string literal and so cannot be checked, unless the format function
2042 takes its format arguments as a @code{va_list}.
2044 @item -Wformat-security
2045 @opindex Wformat-security
2046 If @option{-Wformat} is specified, also warn about uses of format
2047 functions that represent possible security problems. At present, this
2048 warns about calls to @code{printf} and @code{scanf} functions where the
2049 format string is not a string literal and there are no format arguments,
2050 as in @code{printf (foo);}. This may be a security hole if the format
2051 string came from untrusted input and contains @samp{%n}. (This is
2052 currently a subset of what @option{-Wformat-nonliteral} warns about, but
2053 in future warnings may be added to @option{-Wformat-security} that are not
2054 included in @option{-Wformat-nonliteral}.)
2058 Enable @option{-Wformat} plus format checks not included in
2059 @option{-Wformat}. Currently equivalent to @samp{-Wformat
2060 -Wformat-nonliteral -Wformat-security}.
2064 Enable warning about passing a null pointer for arguments marked as
2065 requiring a non-null value by the @code{nonnull} function attribute.
2067 @option{-Wnonnull} is included in @option{-Wall} and @option{-Wformat}. It
2068 can be disabled with the @option{-Wno-nonnull} option.
2070 @item -Wimplicit-int
2071 @opindex Wimplicit-int
2072 Warn when a declaration does not specify a type.
2074 @item -Wimplicit-function-declaration
2075 @itemx -Werror-implicit-function-declaration
2076 @opindex Wimplicit-function-declaration
2077 @opindex Werror-implicit-function-declaration
2078 Give a warning (or error) whenever a function is used before being
2083 Same as @option{-Wimplicit-int} and @option{-Wimplicit-function-declaration}.
2087 Warn if the type of @samp{main} is suspicious. @samp{main} should be a
2088 function with external linkage, returning int, taking either zero
2089 arguments, two, or three arguments of appropriate types.
2091 @item -Wmissing-braces
2092 @opindex Wmissing-braces
2093 Warn if an aggregate or union initializer is not fully bracketed. In
2094 the following example, the initializer for @samp{a} is not fully
2095 bracketed, but that for @samp{b} is fully bracketed.
2098 int a[2][2] = @{ 0, 1, 2, 3 @};
2099 int b[2][2] = @{ @{ 0, 1 @}, @{ 2, 3 @} @};
2103 @opindex Wparentheses
2104 Warn if parentheses are omitted in certain contexts, such
2105 as when there is an assignment in a context where a truth value
2106 is expected, or when operators are nested whose precedence people
2107 often get confused about.
2109 Also warn about constructions where there may be confusion to which
2110 @code{if} statement an @code{else} branch belongs. Here is an example of
2125 In C, every @code{else} branch belongs to the innermost possible @code{if}
2126 statement, which in this example is @code{if (b)}. This is often not
2127 what the programmer expected, as illustrated in the above example by
2128 indentation the programmer chose. When there is the potential for this
2129 confusion, GCC will issue a warning when this flag is specified.
2130 To eliminate the warning, add explicit braces around the innermost
2131 @code{if} statement so there is no way the @code{else} could belong to
2132 the enclosing @code{if}. The resulting code would look like this:
2148 @item -Wsequence-point
2149 @opindex Wsequence-point
2150 Warn about code that may have undefined semantics because of violations
2151 of sequence point rules in the C standard.
2153 The C standard defines the order in which expressions in a C program are
2154 evaluated in terms of @dfn{sequence points}, which represent a partial
2155 ordering between the execution of parts of the program: those executed
2156 before the sequence point, and those executed after it. These occur
2157 after the evaluation of a full expression (one which is not part of a
2158 larger expression), after the evaluation of the first operand of a
2159 @code{&&}, @code{||}, @code{? :} or @code{,} (comma) operator, before a
2160 function is called (but after the evaluation of its arguments and the
2161 expression denoting the called function), and in certain other places.
2162 Other than as expressed by the sequence point rules, the order of
2163 evaluation of subexpressions of an expression is not specified. All
2164 these rules describe only a partial order rather than a total order,
2165 since, for example, if two functions are called within one expression
2166 with no sequence point between them, the order in which the functions
2167 are called is not specified. However, the standards committee have
2168 ruled that function calls do not overlap.
2170 It is not specified when between sequence points modifications to the
2171 values of objects take effect. Programs whose behavior depends on this
2172 have undefined behavior; the C standard specifies that ``Between the
2173 previous and next sequence point an object shall have its stored value
2174 modified at most once by the evaluation of an expression. Furthermore,
2175 the prior value shall be read only to determine the value to be
2176 stored.''. If a program breaks these rules, the results on any
2177 particular implementation are entirely unpredictable.
2179 Examples of code with undefined behavior are @code{a = a++;}, @code{a[n]
2180 = b[n++]} and @code{a[i++] = i;}. Some more complicated cases are not
2181 diagnosed by this option, and it may give an occasional false positive
2182 result, but in general it has been found fairly effective at detecting
2183 this sort of problem in programs.
2185 The present implementation of this option only works for C programs. A
2186 future implementation may also work for C++ programs.
2188 The C standard is worded confusingly, therefore there is some debate
2189 over the precise meaning of the sequence point rules in subtle cases.
2190 Links to discussions of the problem, including proposed formal
2191 definitions, may be found on our readings page, at
2192 @w{@uref{http://gcc.gnu.org/readings.html}}.
2195 @opindex Wreturn-type
2196 Warn whenever a function is defined with a return-type that defaults to
2197 @code{int}. Also warn about any @code{return} statement with no
2198 return-value in a function whose return-type is not @code{void}.
2200 For C++, a function without return type always produces a diagnostic
2201 message, even when @option{-Wno-return-type} is specified. The only
2202 exceptions are @samp{main} and functions defined in system headers.
2206 Warn whenever a @code{switch} statement has an index of enumeral type
2207 and lacks a @code{case} for one or more of the named codes of that
2208 enumeration. (The presence of a @code{default} label prevents this
2209 warning.) @code{case} labels outside the enumeration range also
2210 provoke warnings when this option is used.
2212 @item -Wswitch-default
2213 @opindex Wswitch-switch
2214 Warn whenever a @code{switch} statement does not have a @code{default}
2218 @opindex Wswitch-enum
2219 Warn whenever a @code{switch} statement has an index of enumeral type
2220 and lacks a @code{case} for one or more of the named codes of that
2221 enumeration. @code{case} labels outside the enumeration range also
2222 provoke warnings when this option is used.
2226 Warn if any trigraphs are encountered that might change the meaning of
2227 the program (trigraphs within comments are not warned about).
2229 @item -Wunused-function
2230 @opindex Wunused-function
2231 Warn whenever a static function is declared but not defined or a
2232 non\-inline static function is unused.
2234 @item -Wunused-label
2235 @opindex Wunused-label
2236 Warn whenever a label is declared but not used.
2238 To suppress this warning use the @samp{unused} attribute
2239 (@pxref{Variable Attributes}).
2241 @item -Wunused-parameter
2242 @opindex Wunused-parameter
2243 Warn whenever a function parameter is unused aside from its declaration.
2245 To suppress this warning use the @samp{unused} attribute
2246 (@pxref{Variable Attributes}).
2248 @item -Wunused-variable
2249 @opindex Wunused-variable
2250 Warn whenever a local variable or non-constant static variable is unused
2251 aside from its declaration
2253 To suppress this warning use the @samp{unused} attribute
2254 (@pxref{Variable Attributes}).
2256 @item -Wunused-value
2257 @opindex Wunused-value
2258 Warn whenever a statement computes a result that is explicitly not used.
2260 To suppress this warning cast the expression to @samp{void}.
2264 All the above @option{-Wunused} options combined.
2266 In order to get a warning about an unused function parameter, you must
2267 either specify @samp{-W -Wunused} or separately specify
2268 @option{-Wunused-parameter}.
2270 @item -Wuninitialized
2271 @opindex Wuninitialized
2272 Warn if an automatic variable is used without first being initialized or
2273 if a variable may be clobbered by a @code{setjmp} call.
2275 These warnings are possible only in optimizing compilation,
2276 because they require data flow information that is computed only
2277 when optimizing. If you don't specify @option{-O}, you simply won't
2280 These warnings occur only for variables that are candidates for
2281 register allocation. Therefore, they do not occur for a variable that
2282 is declared @code{volatile}, or whose address is taken, or whose size
2283 is other than 1, 2, 4 or 8 bytes. Also, they do not occur for
2284 structures, unions or arrays, even when they are in registers.
2286 Note that there may be no warning about a variable that is used only
2287 to compute a value that itself is never used, because such
2288 computations may be deleted by data flow analysis before the warnings
2291 These warnings are made optional because GCC is not smart
2292 enough to see all the reasons why the code might be correct
2293 despite appearing to have an error. Here is one example of how
2314 If the value of @code{y} is always 1, 2 or 3, then @code{x} is
2315 always initialized, but GCC doesn't know this. Here is
2316 another common case:
2321 if (change_y) save_y = y, y = new_y;
2323 if (change_y) y = save_y;
2328 This has no bug because @code{save_y} is used only if it is set.
2330 @cindex @code{longjmp} warnings
2331 This option also warns when a non-volatile automatic variable might be
2332 changed by a call to @code{longjmp}. These warnings as well are possible
2333 only in optimizing compilation.
2335 The compiler sees only the calls to @code{setjmp}. It cannot know
2336 where @code{longjmp} will be called; in fact, a signal handler could
2337 call it at any point in the code. As a result, you may get a warning
2338 even when there is in fact no problem because @code{longjmp} cannot
2339 in fact be called at the place which would cause a problem.
2341 Some spurious warnings can be avoided if you declare all the functions
2342 you use that never return as @code{noreturn}. @xref{Function
2345 @item -Wunknown-pragmas
2346 @opindex Wunknown-pragmas
2347 @cindex warning for unknown pragmas
2348 @cindex unknown pragmas, warning
2349 @cindex pragmas, warning of unknown
2350 Warn when a #pragma directive is encountered which is not understood by
2351 GCC@. If this command line option is used, warnings will even be issued
2352 for unknown pragmas in system header files. This is not the case if
2353 the warnings were only enabled by the @option{-Wall} command line option.
2355 @item -Wstrict-aliasing
2356 @opindex Wstrict-aliasing
2357 This option is only active when @option{-fstrict-aliasing} is active.
2358 It warns about code which might break the strict aliasing rules that the
2359 compiler is using for optimization. The warning does not catch all
2360 cases, but does attempt to catch the more common pitfalls. It is
2361 included in @option{-Wall}.
2365 All of the above @samp{-W} options combined. This enables all the
2366 warnings about constructions that some users consider questionable, and
2367 that are easy to avoid (or modify to prevent the warning), even in
2368 conjunction with macros. This also enables some language-specific
2369 warnings described in @ref{C++ Dialect Options} and
2370 @ref{Objective-C Dialect Options}.
2373 The following @option{-W@dots{}} options are not implied by @option{-Wall}.
2374 Some of them warn about constructions that users generally do not
2375 consider questionable, but which occasionally you might wish to check
2376 for; others warn about constructions that are necessary or hard to avoid
2377 in some cases, and there is no simple way to modify the code to suppress
2383 Print extra warning messages for these events:
2387 A function can return either with or without a value. (Falling
2388 off the end of the function body is considered returning without
2389 a value.) For example, this function would evoke such a
2403 An expression-statement or the left-hand side of a comma expression
2404 contains no side effects.
2405 To suppress the warning, cast the unused expression to void.
2406 For example, an expression such as @samp{x[i,j]} will cause a warning,
2407 but @samp{x[(void)i,j]} will not.
2410 An unsigned value is compared against zero with @samp{<} or @samp{<=}.
2413 A comparison like @samp{x<=y<=z} appears; this is equivalent to
2414 @samp{(x<=y ? 1 : 0) <= z}, which is a different interpretation from
2415 that of ordinary mathematical notation.
2418 Storage-class specifiers like @code{static} are not the first things in
2419 a declaration. According to the C Standard, this usage is obsolescent.
2422 The return type of a function has a type qualifier such as @code{const}.
2423 Such a type qualifier has no effect, since the value returned by a
2424 function is not an lvalue. (But don't warn about the GNU extension of
2425 @code{volatile void} return types. That extension will be warned about
2426 if @option{-pedantic} is specified.)
2429 If @option{-Wall} or @option{-Wunused} is also specified, warn about unused
2433 A comparison between signed and unsigned values could produce an
2434 incorrect result when the signed value is converted to unsigned.
2435 (But don't warn if @option{-Wno-sign-compare} is also specified.)
2438 An aggregate has a partly bracketed initializer.
2439 For example, the following code would evoke such a warning,
2440 because braces are missing around the initializer for @code{x.h}:
2443 struct s @{ int f, g; @};
2444 struct t @{ struct s h; int i; @};
2445 struct t x = @{ 1, 2, 3 @};
2449 An aggregate has an initializer which does not initialize all members.
2450 For example, the following code would cause such a warning, because
2451 @code{x.h} would be implicitly initialized to zero:
2454 struct s @{ int f, g, h; @};
2455 struct s x = @{ 3, 4 @};
2459 @item -Wno-div-by-zero
2460 @opindex Wno-div-by-zero
2461 @opindex Wdiv-by-zero
2462 Do not warn about compile-time integer division by zero. Floating point
2463 division by zero is not warned about, as it can be a legitimate way of
2464 obtaining infinities and NaNs.
2466 @item -Wsystem-headers
2467 @opindex Wsystem-headers
2468 @cindex warnings from system headers
2469 @cindex system headers, warnings from
2470 Print warning messages for constructs found in system header files.
2471 Warnings from system headers are normally suppressed, on the assumption
2472 that they usually do not indicate real problems and would only make the
2473 compiler output harder to read. Using this command line option tells
2474 GCC to emit warnings from system headers as if they occurred in user
2475 code. However, note that using @option{-Wall} in conjunction with this
2476 option will @emph{not} warn about unknown pragmas in system
2477 headers---for that, @option{-Wunknown-pragmas} must also be used.
2480 @opindex Wfloat-equal
2481 Warn if floating point values are used in equality comparisons.
2483 The idea behind this is that sometimes it is convenient (for the
2484 programmer) to consider floating-point values as approximations to
2485 infinitely precise real numbers. If you are doing this, then you need
2486 to compute (by analysing the code, or in some other way) the maximum or
2487 likely maximum error that the computation introduces, and allow for it
2488 when performing comparisons (and when producing output, but that's a
2489 different problem). In particular, instead of testing for equality, you
2490 would check to see whether the two values have ranges that overlap; and
2491 this is done with the relational operators, so equality comparisons are
2494 @item -Wtraditional @r{(C only)}
2495 @opindex Wtraditional
2496 Warn about certain constructs that behave differently in traditional and
2497 ISO C@. Also warn about ISO C constructs that have no traditional C
2498 equivalent, and/or problematic constructs which should be avoided.
2502 Macro parameters that appear within string literals in the macro body.
2503 In traditional C macro replacement takes place within string literals,
2504 but does not in ISO C@.
2507 In traditional C, some preprocessor directives did not exist.
2508 Traditional preprocessors would only consider a line to be a directive
2509 if the @samp{#} appeared in column 1 on the line. Therefore
2510 @option{-Wtraditional} warns about directives that traditional C
2511 understands but would ignore because the @samp{#} does not appear as the
2512 first character on the line. It also suggests you hide directives like
2513 @samp{#pragma} not understood by traditional C by indenting them. Some
2514 traditional implementations would not recognize @samp{#elif}, so it
2515 suggests avoiding it altogether.
2518 A function-like macro that appears without arguments.
2521 The unary plus operator.
2524 The @samp{U} integer constant suffix, or the @samp{F} or @samp{L} floating point
2525 constant suffixes. (Traditional C does support the @samp{L} suffix on integer
2526 constants.) Note, these suffixes appear in macros defined in the system
2527 headers of most modern systems, e.g.@: the @samp{_MIN}/@samp{_MAX} macros in @code{<limits.h>}.
2528 Use of these macros in user code might normally lead to spurious
2529 warnings, however gcc's integrated preprocessor has enough context to
2530 avoid warning in these cases.
2533 A function declared external in one block and then used after the end of
2537 A @code{switch} statement has an operand of type @code{long}.
2540 A non-@code{static} function declaration follows a @code{static} one.
2541 This construct is not accepted by some traditional C compilers.
2544 The ISO type of an integer constant has a different width or
2545 signedness from its traditional type. This warning is only issued if
2546 the base of the constant is ten. I.e.@: hexadecimal or octal values, which
2547 typically represent bit patterns, are not warned about.
2550 Usage of ISO string concatenation is detected.
2553 Initialization of automatic aggregates.
2556 Identifier conflicts with labels. Traditional C lacks a separate
2557 namespace for labels.
2560 Initialization of unions. If the initializer is zero, the warning is
2561 omitted. This is done under the assumption that the zero initializer in
2562 user code appears conditioned on e.g.@: @code{__STDC__} to avoid missing
2563 initializer warnings and relies on default initialization to zero in the
2567 Conversions by prototypes between fixed/floating point values and vice
2568 versa. The absence of these prototypes when compiling with traditional
2569 C would cause serious problems. This is a subset of the possible
2570 conversion warnings, for the full set use @option{-Wconversion}.
2573 Use of ISO C style function definitions. This warning intentionally is
2574 @emph{not} issued for prototype declarations or variadic functions
2575 because these ISO C features will appear in your code when using
2576 libiberty's traditional C compatibility macros, @code{PARAMS} and
2577 @code{VPARAMS}. This warning is also bypassed for nested functions
2578 because that feature is already a gcc extension and thus not relevant to
2579 traditional C compatibility.
2584 Warn if an undefined identifier is evaluated in an @samp{#if} directive.
2586 @item -Wendif-labels
2587 @opindex Wendif-labels
2588 Warn whenever an @samp{#else} or an @samp{#endif} are followed by text.
2592 Warn whenever a local variable shadows another local variable, parameter or
2593 global variable or whenever a built-in function is shadowed.
2595 @item -Wlarger-than-@var{len}
2596 @opindex Wlarger-than
2597 Warn whenever an object of larger than @var{len} bytes is defined.
2599 @item -Wpointer-arith
2600 @opindex Wpointer-arith
2601 Warn about anything that depends on the ``size of'' a function type or
2602 of @code{void}. GNU C assigns these types a size of 1, for
2603 convenience in calculations with @code{void *} pointers and pointers
2606 @item -Wbad-function-cast @r{(C only)}
2607 @opindex Wbad-function-cast
2608 Warn whenever a function call is cast to a non-matching type.
2609 For example, warn if @code{int malloc()} is cast to @code{anything *}.
2613 Warn whenever a pointer is cast so as to remove a type qualifier from
2614 the target type. For example, warn if a @code{const char *} is cast
2615 to an ordinary @code{char *}.
2618 @opindex Wcast-align
2619 Warn whenever a pointer is cast such that the required alignment of the
2620 target is increased. For example, warn if a @code{char *} is cast to
2621 an @code{int *} on machines where integers can only be accessed at
2622 two- or four-byte boundaries.
2624 @item -Wwrite-strings
2625 @opindex Wwrite-strings
2626 When compiling C, give string constants the type @code{const
2627 char[@var{length}]} so that
2628 copying the address of one into a non-@code{const} @code{char *}
2629 pointer will get a warning; when compiling C++, warn about the
2630 deprecated conversion from string constants to @code{char *}.
2631 These warnings will help you find at
2632 compile time code that can try to write into a string constant, but
2633 only if you have been very careful about using @code{const} in
2634 declarations and prototypes. Otherwise, it will just be a nuisance;
2635 this is why we did not make @option{-Wall} request these warnings.
2638 @opindex Wconversion
2639 Warn if a prototype causes a type conversion that is different from what
2640 would happen to the same argument in the absence of a prototype. This
2641 includes conversions of fixed point to floating and vice versa, and
2642 conversions changing the width or signedness of a fixed point argument
2643 except when the same as the default promotion.
2645 Also, warn if a negative integer constant expression is implicitly
2646 converted to an unsigned type. For example, warn about the assignment
2647 @code{x = -1} if @code{x} is unsigned. But do not warn about explicit
2648 casts like @code{(unsigned) -1}.
2650 @item -Wsign-compare
2651 @opindex Wsign-compare
2652 @cindex warning for comparison of signed and unsigned values
2653 @cindex comparison of signed and unsigned values, warning
2654 @cindex signed and unsigned values, comparison warning
2655 Warn when a comparison between signed and unsigned values could produce
2656 an incorrect result when the signed value is converted to unsigned.
2657 This warning is also enabled by @option{-W}; to get the other warnings
2658 of @option{-W} without this warning, use @samp{-W -Wno-sign-compare}.
2660 @item -Waggregate-return
2661 @opindex Waggregate-return
2662 Warn if any functions that return structures or unions are defined or
2663 called. (In languages where you can return an array, this also elicits
2666 @item -Wstrict-prototypes @r{(C only)}
2667 @opindex Wstrict-prototypes
2668 Warn if a function is declared or defined without specifying the
2669 argument types. (An old-style function definition is permitted without
2670 a warning if preceded by a declaration which specifies the argument
2673 @item -Wmissing-prototypes @r{(C only)}
2674 @opindex Wmissing-prototypes
2675 Warn if a global function is defined without a previous prototype
2676 declaration. This warning is issued even if the definition itself
2677 provides a prototype. The aim is to detect global functions that fail
2678 to be declared in header files.
2680 @item -Wmissing-declarations
2681 @opindex Wmissing-declarations
2682 Warn if a global function is defined without a previous declaration.
2683 Do so even if the definition itself provides a prototype.
2684 Use this option to detect global functions that are not declared in
2687 @item -Wmissing-noreturn
2688 @opindex Wmissing-noreturn
2689 Warn about functions which might be candidates for attribute @code{noreturn}.
2690 Note these are only possible candidates, not absolute ones. Care should
2691 be taken to manually verify functions actually do not ever return before
2692 adding the @code{noreturn} attribute, otherwise subtle code generation
2693 bugs could be introduced. You will not get a warning for @code{main} in
2694 hosted C environments.
2696 @item -Wmissing-format-attribute
2697 @opindex Wmissing-format-attribute
2699 If @option{-Wformat} is enabled, also warn about functions which might be
2700 candidates for @code{format} attributes. Note these are only possible
2701 candidates, not absolute ones. GCC will guess that @code{format}
2702 attributes might be appropriate for any function that calls a function
2703 like @code{vprintf} or @code{vscanf}, but this might not always be the
2704 case, and some functions for which @code{format} attributes are
2705 appropriate may not be detected. This option has no effect unless
2706 @option{-Wformat} is enabled (possibly by @option{-Wall}).
2708 @item -Wno-multichar
2709 @opindex Wno-multichar
2711 Do not warn if a multicharacter constant (@samp{'FOOF'}) is used.
2712 Usually they indicate a typo in the user's code, as they have
2713 implementation-defined values, and should not be used in portable code.
2715 @item -Wno-deprecated-declarations
2716 @opindex Wno-deprecated-declarations
2717 Do not warn about uses of functions, variables, and types marked as
2718 deprecated by using the @code{deprecated} attribute.
2719 (@pxref{Function Attributes}, @pxref{Variable Attributes},
2720 @pxref{Type Attributes}.)
2724 Warn if a structure is given the packed attribute, but the packed
2725 attribute has no effect on the layout or size of the structure.
2726 Such structures may be mis-aligned for little benefit. For
2727 instance, in this code, the variable @code{f.x} in @code{struct bar}
2728 will be misaligned even though @code{struct bar} does not itself
2729 have the packed attribute:
2736 @} __attribute__((packed));
2746 Warn if padding is included in a structure, either to align an element
2747 of the structure or to align the whole structure. Sometimes when this
2748 happens it is possible to rearrange the fields of the structure to
2749 reduce the padding and so make the structure smaller.
2751 @item -Wredundant-decls
2752 @opindex Wredundant-decls
2753 Warn if anything is declared more than once in the same scope, even in
2754 cases where multiple declaration is valid and changes nothing.
2756 @item -Wnested-externs @r{(C only)}
2757 @opindex Wnested-externs
2758 Warn if an @code{extern} declaration is encountered within a function.
2760 @item -Wunreachable-code
2761 @opindex Wunreachable-code
2762 Warn if the compiler detects that code will never be executed.
2764 This option is intended to warn when the compiler detects that at
2765 least a whole line of source code will never be executed, because
2766 some condition is never satisfied or because it is after a
2767 procedure that never returns.
2769 It is possible for this option to produce a warning even though there
2770 are circumstances under which part of the affected line can be executed,
2771 so care should be taken when removing apparently-unreachable code.
2773 For instance, when a function is inlined, a warning may mean that the
2774 line is unreachable in only one inlined copy of the function.
2776 This option is not made part of @option{-Wall} because in a debugging
2777 version of a program there is often substantial code which checks
2778 correct functioning of the program and is, hopefully, unreachable
2779 because the program does work. Another common use of unreachable
2780 code is to provide behavior which is selectable at compile-time.
2784 Warn if a function can not be inlined and it was declared as inline.
2788 @opindex Wno-long-long
2789 Warn if @samp{long long} type is used. This is default. To inhibit
2790 the warning messages, use @option{-Wno-long-long}. Flags
2791 @option{-Wlong-long} and @option{-Wno-long-long} are taken into account
2792 only when @option{-pedantic} flag is used.
2794 @item -Wdisabled-optimization
2795 @opindex Wdisabled-optimization
2796 Warn if a requested optimization pass is disabled. This warning does
2797 not generally indicate that there is anything wrong with your code; it
2798 merely indicates that GCC's optimizers were unable to handle the code
2799 effectively. Often, the problem is that your code is too big or too
2800 complex; GCC will refuse to optimize programs when the optimization
2801 itself is likely to take inordinate amounts of time.
2805 Make all warnings into errors.
2808 @node Debugging Options
2809 @section Options for Debugging Your Program or GCC
2810 @cindex options, debugging
2811 @cindex debugging information options
2813 GCC has various special options that are used for debugging
2814 either your program or GCC:
2819 Produce debugging information in the operating system's native format
2820 (stabs, COFF, XCOFF, or DWARF)@. GDB can work with this debugging
2823 On most systems that use stabs format, @option{-g} enables use of extra
2824 debugging information that only GDB can use; this extra information
2825 makes debugging work better in GDB but will probably make other debuggers
2827 refuse to read the program. If you want to control for certain whether
2828 to generate the extra information, use @option{-gstabs+}, @option{-gstabs},
2829 @option{-gxcoff+}, @option{-gxcoff}, @option{-gdwarf-1+}, @option{-gdwarf-1},
2830 or @option{-gvms} (see below).
2832 Unlike most other C compilers, GCC allows you to use @option{-g} with
2833 @option{-O}. The shortcuts taken by optimized code may occasionally
2834 produce surprising results: some variables you declared may not exist
2835 at all; flow of control may briefly move where you did not expect it;
2836 some statements may not be executed because they compute constant
2837 results or their values were already at hand; some statements may
2838 execute in different places because they were moved out of loops.
2840 Nevertheless it proves possible to debug optimized output. This makes
2841 it reasonable to use the optimizer for programs that might have bugs.
2843 The following options are useful when GCC is generated with the
2844 capability for more than one debugging format.
2848 Produce debugging information for use by GDB@. This means to use the
2849 most expressive format available (DWARF 2, stabs, or the native format
2850 if neither of those are supported), including GDB extensions if at all
2855 Produce debugging information in stabs format (if that is supported),
2856 without GDB extensions. This is the format used by DBX on most BSD
2857 systems. On MIPS, Alpha and System V Release 4 systems this option
2858 produces stabs debugging output which is not understood by DBX or SDB@.
2859 On System V Release 4 systems this option requires the GNU assembler.
2863 Produce debugging information in stabs format (if that is supported),
2864 using GNU extensions understood only by the GNU debugger (GDB)@. The
2865 use of these extensions is likely to make other debuggers crash or
2866 refuse to read the program.
2870 Produce debugging information in COFF format (if that is supported).
2871 This is the format used by SDB on most System V systems prior to
2876 Produce debugging information in XCOFF format (if that is supported).
2877 This is the format used by the DBX debugger on IBM RS/6000 systems.
2881 Produce debugging information in XCOFF format (if that is supported),
2882 using GNU extensions understood only by the GNU debugger (GDB)@. The
2883 use of these extensions is likely to make other debuggers crash or
2884 refuse to read the program, and may cause assemblers other than the GNU
2885 assembler (GAS) to fail with an error.
2889 Produce debugging information in DWARF version 1 format (if that is
2890 supported). This is the format used by SDB on most System V Release 4
2893 This option is deprecated.
2897 Produce debugging information in DWARF version 1 format (if that is
2898 supported), using GNU extensions understood only by the GNU debugger
2899 (GDB)@. The use of these extensions is likely to make other debuggers
2900 crash or refuse to read the program.
2902 This option is deprecated.
2906 Produce debugging information in DWARF version 2 format (if that is
2907 supported). This is the format used by DBX on IRIX 6.
2911 Produce debugging information in VMS debug format (if that is
2912 supported). This is the format used by DEBUG on VMS systems.
2915 @itemx -ggdb@var{level}
2916 @itemx -gstabs@var{level}
2917 @itemx -gcoff@var{level}
2918 @itemx -gxcoff@var{level}
2919 @itemx -gvms@var{level}
2920 Request debugging information and also use @var{level} to specify how
2921 much information. The default level is 2.
2923 Level 1 produces minimal information, enough for making backtraces in
2924 parts of the program that you don't plan to debug. This includes
2925 descriptions of functions and external variables, but no information
2926 about local variables and no line numbers.
2928 Level 3 includes extra information, such as all the macro definitions
2929 present in the program. Some debuggers support macro expansion when
2930 you use @option{-g3}.
2932 Note that in order to avoid confusion between DWARF1 debug level 2,
2933 and DWARF2, neither @option{-gdwarf} nor @option{-gdwarf-2} accept
2934 a concatenated debug level. Instead use an additional @option{-g@var{level}}
2935 option to change the debug level for DWARF1 or DWARF2.
2937 @item -feliminate-dwarf2-dups
2938 @opindex feliminate-dwarf2-dups
2939 Compress DWARF2 debugging information by eliminating duplicated
2940 information about each symbol. This option only makes sense when
2941 generating DWARF2 debugging information with @option{-gdwarf-2}.
2946 Generate extra code to write profile information suitable for the
2947 analysis program @code{prof}. You must use this option when compiling
2948 the source files you want data about, and you must also use it when
2951 @cindex @code{gprof}
2954 Generate extra code to write profile information suitable for the
2955 analysis program @code{gprof}. You must use this option when compiling
2956 the source files you want data about, and you must also use it when
2961 Makes the compiler print out each function name as it is compiled, and
2962 print some statistics about each pass when it finishes.
2965 @opindex ftime-report
2966 Makes the compiler print some statistics about the time consumed by each
2967 pass when it finishes.
2970 @opindex fmem-report
2971 Makes the compiler print some statistics about permanent memory
2972 allocation when it finishes.
2974 @item -fprofile-arcs
2975 @opindex fprofile-arcs
2976 Instrument @dfn{arcs} during compilation to generate coverage data or
2977 for profile-directed block ordering. During execution the program
2978 records how many times each branch is executed and how many times it is
2979 taken. When the compiled program exits it saves this data to a file
2980 called @file{@var{auxname}.da} for each source file. @var{auxname} is
2981 generated from the name of the output file, if explicitly specified and
2982 it is not the final executable, otherwise it is the basename of the
2983 source file. In both cases any suffix is removed (e.g. @file{foo.da}
2984 for input file @file{dir/foo.c}, or @file{dir/foo.da} for output file
2985 specified as @option{-o dir/foo.o}).
2987 For profile-directed block ordering, compile the program with
2988 @option{-fprofile-arcs} plus optimization and code generation options,
2989 generate the arc profile information by running the program on a
2990 selected workload, and then compile the program again with the same
2991 optimization and code generation options plus
2992 @option{-fbranch-probabilities} (@pxref{Optimize Options,,Options that
2993 Control Optimization}).
2995 The other use of @option{-fprofile-arcs} is for use with @code{gcov},
2996 when it is used with the @option{-ftest-coverage} option.
2998 With @option{-fprofile-arcs}, for each function of your program GCC
2999 creates a program flow graph, then finds a spanning tree for the graph.
3000 Only arcs that are not on the spanning tree have to be instrumented: the
3001 compiler adds code to count the number of times that these arcs are
3002 executed. When an arc is the only exit or only entrance to a block, the
3003 instrumentation code can be added to the block; otherwise, a new basic
3004 block must be created to hold the instrumentation code.
3007 @item -ftest-coverage
3008 @opindex ftest-coverage
3009 Create data files for the @code{gcov} code-coverage utility
3010 (@pxref{Gcov,, @code{gcov}: a GCC Test Coverage Program}). See
3011 @option{-fprofile-arcs} option above for a description of @var{auxname}.
3014 @item @var{auxname}.bb
3015 A mapping from basic blocks to line numbers, which @code{gcov} uses to
3016 associate basic block execution counts with line numbers.
3018 @item @var{auxname}.bbg
3019 A list of all arcs in the program flow graph. This allows @code{gcov}
3020 to reconstruct the program flow graph, so that it can compute all basic
3021 block and arc execution counts from the information in the
3022 @file{@var{auxname}.da} file.
3025 Use @option{-ftest-coverage} with @option{-fprofile-arcs}; the latter
3026 option adds instrumentation to the program, which then writes
3027 execution counts to another data file:
3030 @item @var{auxname}.da
3031 Runtime arc execution counts, used in conjunction with the arc
3032 information in the file @file{@var{auxname}.bbg}.
3035 Coverage data will map better to the source files if
3036 @option{-ftest-coverage} is used without optimization.
3038 @item -d@var{letters}
3040 Says to make debugging dumps during compilation at times specified by
3041 @var{letters}. This is used for debugging the compiler. The file names
3042 for most of the dumps are made by appending a pass number and a word to
3043 the @var{dumpname}. @var{dumpname} is generated from the name of the
3044 output file, if explicitly specified and it is not an executable,
3045 otherwise it is the basename of the source file. In both cases any
3046 suffix is removed (e.g. @file{foo.00.rtl} or @file{foo.01.sibling}).
3047 Here are the possible letters for use in @var{letters}, and their
3053 Annotate the assembler output with miscellaneous debugging information.
3056 Dump after computing branch probabilities, to @file{@var{file}.15.bp}.
3059 Dump after block reordering, to @file{@var{file}.30.bbro}.
3062 Dump after instruction combination, to the file @file{@var{file}.17.combine}.
3065 Dump after the first if conversion, to the file @file{@var{file}.18.ce}.
3068 Dump after delayed branch scheduling, to @file{@var{file}.32.dbr}.
3071 Dump all macro definitions, at the end of preprocessing, in addition to
3075 Dump after SSA optimizations, to @file{@var{file}.04.ssa} and
3076 @file{@var{file}.07.ussa}.
3079 Dump after the second if conversion, to @file{@var{file}.27.ce2}.
3082 Dump after life analysis, to @file{@var{file}.16.life}.
3085 Dump after purging @code{ADDRESSOF} codes, to @file{@var{file}.10.addressof}.
3088 Dump after global register allocation, to @file{@var{file}.22.greg}.
3091 Dump after finalization of EH handling code, to @file{@var{file}.02.eh}.
3094 Dump after reg-to-stack conversion, to @file{@var{file}.29.stack}.
3097 Dump after post-reload optimizations, to @file{@var{file}.23.postreload}.
3100 Dump after GCSE, to @file{@var{file}.11.gcse}.
3103 Dump after sibling call optimizations, to @file{@var{file}.01.sibling}.
3106 Dump after the first jump optimization, to @file{@var{file}.03.jump}.
3109 Dump after conversion from registers to stack, to @file{@var{file}.33.stack}.
3112 Dump after local register allocation, to @file{@var{file}.21.lreg}.
3115 Dump after loop optimization, to @file{@var{file}.12.loop}.
3118 Dump after performing the machine dependent reorganisation pass, to
3119 @file{@var{file}.31.mach}.
3122 Dump after register renumbering, to @file{@var{file}.26.rnreg}.
3125 Dump after the register move pass, to @file{@var{file}.19.regmove}.
3128 Dump after RTL generation, to @file{@var{file}.00.rtl}.
3131 Dump after the second scheduling pass, to @file{@var{file}.28.sched2}.
3134 Dump after CSE (including the jump optimization that sometimes follows
3135 CSE), to @file{@var{file}.09.cse}.
3138 Dump after the first scheduling pass, to @file{@var{file}.20.sched}.
3141 Dump after the second CSE pass (including the jump optimization that
3142 sometimes follows CSE), to @file{@var{file}.13.cse2}.
3145 Dump after null pointer ellimination pass ti @file{@var{file}.08.null}.
3148 Dump after the second flow pass, to @file{@var{file}.24.flow2}.
3151 Dump after SSA dead code elimination, to @file{@var{file}.06.ssadce}.
3154 Dump after the peephole pass, to @file{@var{file}.25.peephole2}.
3157 Produce all the dumps listed above.
3160 Print statistics on memory usage, at the end of the run, to
3164 Annotate the assembler output with a comment indicating which
3165 pattern and alternative was used. The length of each instruction is
3169 Dump the RTL in the assembler output as a comment before each instruction.
3170 Also turns on @option{-dp} annotation.
3173 For each of the other indicated dump files (except for
3174 @file{@var{file}.00.rtl}), dump a representation of the control flow graph
3175 suitable for viewing with VCG to @file{@var{file}.@var{pass}.vcg}.
3178 Just generate RTL for a function instead of compiling it. Usually used
3182 Dump debugging information during parsing, to standard error.
3185 @item -fdump-unnumbered
3186 @opindex fdump-unnumbered
3187 When doing debugging dumps (see @option{-d} option above), suppress instruction
3188 numbers and line number note output. This makes it more feasible to
3189 use diff on debugging dumps for compiler invocations with different
3190 options, in particular with and without @option{-g}.
3192 @item -fdump-translation-unit @r{(C and C++ only)}
3193 @itemx -fdump-translation-unit-@var{options} @r{(C and C++ only)}
3194 @opindex fdump-translation-unit
3195 Dump a representation of the tree structure for the entire translation
3196 unit to a file. The file name is made by appending @file{.tu} to the
3197 source file name. If the @samp{-@var{options}} form is used, @var{options}
3198 controls the details of the dump as described for the
3199 @option{-fdump-tree} options.
3201 @item -fdump-class-hierarchy @r{(C++ only)}
3202 @itemx -fdump-class-hierarchy-@var{options} @r{(C++ only)}
3203 @opindex fdump-class-hierarchy
3204 Dump a representation of each class's hierarchy and virtual function
3205 table layout to a file. The file name is made by appending @file{.class}
3206 to the source file name. If the @samp{-@var{options}} form is used,
3207 @var{options} controls the details of the dump as described for the
3208 @option{-fdump-tree} options.
3210 @item -fdump-tree-@var{switch} @r{(C++ only)}
3211 @itemx -fdump-tree-@var{switch}-@var{options} @r{(C++ only)}
3213 Control the dumping at various stages of processing the intermediate
3214 language tree to a file. The file name is generated by appending a switch
3215 specific suffix to the source file name. If the @samp{-@var{options}}
3216 form is used, @var{options} is a list of @samp{-} separated options that
3217 control the details of the dump. Not all options are applicable to all
3218 dumps, those which are not meaningful will be ignored. The following
3219 options are available
3223 Print the address of each node. Usually this is not meaningful as it
3224 changes according to the environment and source file. Its primary use
3225 is for tying up a dump file with a debug environment.
3227 Inhibit dumping of members of a scope or body of a function merely
3228 because that scope has been reached. Only dump such items when they
3229 are directly reachable by some other path.
3231 Turn on all options.
3234 The following tree dumps are possible:
3237 Dump before any tree based optimization, to @file{@var{file}.original}.
3239 Dump after all tree based optimization, to @file{@var{file}.optimized}.
3241 Dump after function inlining, to @file{@var{file}.inlined}.
3244 @item -fsched-verbose=@var{n}
3245 @opindex fsched-verbose
3246 On targets that use instruction scheduling, this option controls the
3247 amount of debugging output the scheduler prints. This information is
3248 written to standard error, unless @option{-dS} or @option{-dR} is
3249 specified, in which case it is output to the usual dump
3250 listing file, @file{.sched} or @file{.sched2} respectively. However
3251 for @var{n} greater than nine, the output is always printed to standard
3254 For @var{n} greater than zero, @option{-fsched-verbose} outputs the
3255 same information as @option{-dRS}. For @var{n} greater than one, it
3256 also output basic block probabilities, detailed ready list information
3257 and unit/insn info. For @var{n} greater than two, it includes RTL
3258 at abort point, control-flow and regions info. And for @var{n} over
3259 four, @option{-fsched-verbose} also includes dependence info.
3263 Store the usual ``temporary'' intermediate files permanently; place them
3264 in the current directory and name them based on the source file. Thus,
3265 compiling @file{foo.c} with @samp{-c -save-temps} would produce files
3266 @file{foo.i} and @file{foo.s}, as well as @file{foo.o}. This creates a
3267 preprocessed @file{foo.i} output file even though the compiler now
3268 normally uses an integrated preprocessor.
3272 Report the CPU time taken by each subprocess in the compilation
3273 sequence. For C source files, this is the compiler proper and assembler
3274 (plus the linker if linking is done). The output looks like this:
3281 The first number on each line is the ``user time,'' that is time spent
3282 executing the program itself. The second number is ``system time,''
3283 time spent executing operating system routines on behalf of the program.
3284 Both numbers are in seconds.
3286 @item -print-file-name=@var{library}
3287 @opindex print-file-name
3288 Print the full absolute name of the library file @var{library} that
3289 would be used when linking---and don't do anything else. With this
3290 option, GCC does not compile or link anything; it just prints the
3293 @item -print-multi-directory
3294 @opindex print-multi-directory
3295 Print the directory name corresponding to the multilib selected by any
3296 other switches present in the command line. This directory is supposed
3297 to exist in @env{GCC_EXEC_PREFIX}.
3299 @item -print-multi-lib
3300 @opindex print-multi-lib
3301 Print the mapping from multilib directory names to compiler switches
3302 that enable them. The directory name is separated from the switches by
3303 @samp{;}, and each switch starts with an @samp{@@} instead of the
3304 @samp{-}, without spaces between multiple switches. This is supposed to
3305 ease shell-processing.
3307 @item -print-prog-name=@var{program}
3308 @opindex print-prog-name
3309 Like @option{-print-file-name}, but searches for a program such as @samp{cpp}.
3311 @item -print-libgcc-file-name
3312 @opindex print-libgcc-file-name
3313 Same as @option{-print-file-name=libgcc.a}.
3315 This is useful when you use @option{-nostdlib} or @option{-nodefaultlibs}
3316 but you do want to link with @file{libgcc.a}. You can do
3319 gcc -nostdlib @var{files}@dots{} `gcc -print-libgcc-file-name`
3322 @item -print-search-dirs
3323 @opindex print-search-dirs
3324 Print the name of the configured installation directory and a list of
3325 program and library directories gcc will search---and don't do anything else.
3327 This is useful when gcc prints the error message
3328 @samp{installation problem, cannot exec cpp0: No such file or directory}.
3329 To resolve this you either need to put @file{cpp0} and the other compiler
3330 components where gcc expects to find them, or you can set the environment
3331 variable @env{GCC_EXEC_PREFIX} to the directory where you installed them.
3332 Don't forget the trailing '/'.
3333 @xref{Environment Variables}.
3336 @opindex dumpmachine
3337 Print the compiler's target machine (for example,
3338 @samp{i686-pc-linux-gnu})---and don't do anything else.
3341 @opindex dumpversion
3342 Print the compiler version (for example, @samp{3.0})---and don't do
3347 Print the compiler's built-in specs---and don't do anything else. (This
3348 is used when GCC itself is being built.) @xref{Spec Files}.
3351 @node Optimize Options
3352 @section Options That Control Optimization
3353 @cindex optimize options
3354 @cindex options, optimization
3356 These options control various sorts of optimizations:
3363 Optimize. Optimizing compilation takes somewhat more time, and a lot
3364 more memory for a large function.
3366 Without @option{-O}, the compiler's goal is to reduce the cost of
3367 compilation and to make debugging produce the expected results.
3368 Statements are independent: if you stop the program with a breakpoint
3369 between statements, you can then assign a new value to any variable or
3370 change the program counter to any other statement in the function and
3371 get exactly the results you would expect from the source code.
3373 With @option{-O}, the compiler tries to reduce code size and execution
3374 time, without performing any optimizations that take a great deal of
3379 Optimize even more. GCC performs nearly all supported optimizations
3380 that do not involve a space-speed tradeoff. The compiler does not
3381 perform loop unrolling or function inlining when you specify @option{-O2}.
3382 As compared to @option{-O}, this option increases both compilation time
3383 and the performance of the generated code.
3385 @option{-O2} turns on all optional optimizations except for loop
3386 unrolling, function inlining, and register renaming. It also turns on
3387 the @option{-fforce-mem} and @option{-fstrict-aliasing} option on all
3388 machines and frame pointer elimination on machines where doing so does
3389 not interfere with debugging.
3391 Please note the warning under @option{-fgcse} about
3392 invoking @option{-O2} on programs that use computed gotos.
3396 Optimize yet more. @option{-O3} turns on all optimizations specified by
3397 @option{-O2} and also turns on the @option{-finline-functions} and
3398 @option{-frename-registers} options.
3406 Optimize for size. @option{-Os} enables all @option{-O2} optimizations that
3407 do not typically increase code size. It also performs further
3408 optimizations designed to reduce code size.
3410 If you use multiple @option{-O} options, with or without level numbers,
3411 the last such option is the one that is effective.
3414 Options of the form @option{-f@var{flag}} specify machine-independent
3415 flags. Most flags have both positive and negative forms; the negative
3416 form of @option{-ffoo} would be @option{-fno-foo}. In the table below,
3417 only one of the forms is listed---the one which is not the default.
3418 You can figure out the other form by either removing @samp{no-} or
3423 @opindex ffloat-store
3424 Do not store floating point variables in registers, and inhibit other
3425 options that might change whether a floating point value is taken from a
3428 @cindex floating point precision
3429 This option prevents undesirable excess precision on machines such as
3430 the 68000 where the floating registers (of the 68881) keep more
3431 precision than a @code{double} is supposed to have. Similarly for the
3432 x86 architecture. For most programs, the excess precision does only
3433 good, but a few programs rely on the precise definition of IEEE floating
3434 point. Use @option{-ffloat-store} for such programs, after modifying
3435 them to store all pertinent intermediate computations into variables.
3437 @item -fno-default-inline
3438 @opindex fno-default-inline
3439 Do not make member functions inline by default merely because they are
3440 defined inside the class scope (C++ only). Otherwise, when you specify
3441 @w{@option{-O}}, member functions defined inside class scope are compiled
3442 inline by default; i.e., you don't need to add @samp{inline} in front of
3443 the member function name.
3445 @item -fno-defer-pop
3446 @opindex fno-defer-pop
3447 Always pop the arguments to each function call as soon as that function
3448 returns. For machines which must pop arguments after a function call,
3449 the compiler normally lets arguments accumulate on the stack for several
3450 function calls and pops them all at once.
3454 Force memory operands to be copied into registers before doing
3455 arithmetic on them. This produces better code by making all memory
3456 references potential common subexpressions. When they are not common
3457 subexpressions, instruction combination should eliminate the separate
3458 register-load. The @option{-O2} option turns on this option.
3461 @opindex fforce-addr
3462 Force memory address constants to be copied into registers before
3463 doing arithmetic on them. This may produce better code just as
3464 @option{-fforce-mem} may.
3466 @item -fomit-frame-pointer
3467 @opindex fomit-frame-pointer
3468 Don't keep the frame pointer in a register for functions that
3469 don't need one. This avoids the instructions to save, set up and
3470 restore frame pointers; it also makes an extra register available
3471 in many functions. @strong{It also makes debugging impossible on
3474 On some machines, such as the VAX, this flag has no effect, because
3475 the standard calling sequence automatically handles the frame pointer
3476 and nothing is saved by pretending it doesn't exist. The
3477 machine-description macro @code{FRAME_POINTER_REQUIRED} controls
3478 whether a target machine supports this flag. @xref{Registers,,Register
3479 Usage, gccint, GNU Compiler Collection (GCC) Internals}.
3481 @item -foptimize-sibling-calls
3482 @opindex foptimize-sibling-calls
3483 Optimize sibling and tail recursive calls.
3487 This option generates traps for signed overflow on addition, subtraction,
3488 multiplication operations.
3492 Don't pay attention to the @code{inline} keyword. Normally this option
3493 is used to keep the compiler from expanding any functions inline.
3494 Note that if you are not optimizing, no functions can be expanded inline.
3496 @item -finline-functions
3497 @opindex finline-functions
3498 Integrate all simple functions into their callers. The compiler
3499 heuristically decides which functions are simple enough to be worth
3500 integrating in this way.
3502 If all calls to a given function are integrated, and the function is
3503 declared @code{static}, then the function is normally not output as
3504 assembler code in its own right.
3506 @item -finline-limit=@var{n}
3507 @opindex finline-limit
3508 By default, gcc limits the size of functions that can be inlined. This flag
3509 allows the control of this limit for functions that are explicitly marked as
3510 inline (ie marked with the inline keyword or defined within the class
3511 definition in c++). @var{n} is the size of functions that can be inlined in
3512 number of pseudo instructions (not counting parameter handling). The default
3513 value of @var{n} is 600.
3514 Increasing this value can result in more inlined code at
3515 the cost of compilation time and memory consumption. Decreasing usually makes
3516 the compilation faster and less code will be inlined (which presumably
3517 means slower programs). This option is particularly useful for programs that
3518 use inlining heavily such as those based on recursive templates with C++.
3520 @emph{Note:} pseudo instruction represents, in this particular context, an
3521 abstract measurement of function's size. In no way, it represents a count
3522 of assembly instructions and as such its exact meaning might change from one
3523 release to an another.
3525 @item -fkeep-inline-functions
3526 @opindex fkeep-inline-functions
3527 Even if all calls to a given function are integrated, and the function
3528 is declared @code{static}, nevertheless output a separate run-time
3529 callable version of the function. This switch does not affect
3530 @code{extern inline} functions.
3532 @item -fkeep-static-consts
3533 @opindex fkeep-static-consts
3534 Emit variables declared @code{static const} when optimization isn't turned
3535 on, even if the variables aren't referenced.
3537 GCC enables this option by default. If you want to force the compiler to
3538 check if the variable was referenced, regardless of whether or not
3539 optimization is turned on, use the @option{-fno-keep-static-consts} option.
3541 @item -fmerge-constants
3542 Attempt to merge identical constants (string constants and floating point
3543 constants) accross compilation units.
3545 This option is default for optimized compilation if assembler and linker
3546 support it. Use @option{-fno-merge-constants} to inhibit this behavior.
3548 @item -fmerge-all-constants
3549 Attempt to merge identical constants and identical variables.
3551 This option implies @option{-fmerge-constants}. In addition to
3552 @option{-fmerge-constants} this considers e.g. even constant initialized
3553 arrays or initialized constant variables with integral or floating point
3554 types. Languages like C or C++ require each non-automatic variable to
3555 have distinct location, so using this option will result in non-conforming
3560 Use a graph coloring register allocator. Currently this option is meant
3561 for testing, so we are interested to hear about miscompilations with
3564 @item -fno-branch-count-reg
3565 @opindex fno-branch-count-reg
3566 Do not use ``decrement and branch'' instructions on a count register,
3567 but instead generate a sequence of instructions that decrement a
3568 register, compare it against zero, then branch based upon the result.
3569 This option is only meaningful on architectures that support such
3570 instructions, which include x86, PowerPC, IA-64 and S/390.
3572 @item -fno-function-cse
3573 @opindex fno-function-cse
3574 Do not put function addresses in registers; make each instruction that
3575 calls a constant function contain the function's address explicitly.
3577 This option results in less efficient code, but some strange hacks
3578 that alter the assembler output may be confused by the optimizations
3579 performed when this option is not used.
3583 Sets @option{-fno-math-errno}, @option{-funsafe-math-optimizations}, @*
3584 @option{-fno-trapping-math}, @option{-ffinite-math-only} and @*
3585 @option{-fno-signaling-nans}.
3587 This option causes the preprocessor macro @code{__FAST_MATH__} to be defined.
3589 This option should never be turned on by any @option{-O} option since
3590 it can result in incorrect output for programs which depend on
3591 an exact implementation of IEEE or ISO rules/specifications for
3594 @item -fno-math-errno
3595 @opindex fno-math-errno
3596 Do not set ERRNO after calling math functions that are executed
3597 with a single instruction, e.g., sqrt. A program that relies on
3598 IEEE exceptions for math error handling may want to use this flag
3599 for speed while maintaining IEEE arithmetic compatibility.
3601 This option should never be turned on by any @option{-O} option since
3602 it can result in incorrect output for programs which depend on
3603 an exact implementation of IEEE or ISO rules/specifications for
3606 The default is @option{-fmath-errno}.
3608 @item -funsafe-math-optimizations
3609 @opindex funsafe-math-optimizations
3610 Allow optimizations for floating-point arithmetic that (a) assume
3611 that arguments and results are valid and (b) may violate IEEE or
3612 ANSI standards. When used at link-time, it may include libraries
3613 or startup files that change the default FPU control word or other
3614 similar optimizations.
3616 This option should never be turned on by any @option{-O} option since
3617 it can result in incorrect output for programs which depend on
3618 an exact implementation of IEEE or ISO rules/specifications for
3621 The default is @option{-fno-unsafe-math-optimizations}.
3623 @item -ffinite-math-only
3624 @opindex ffinite-math-only
3625 Allow optimizations for floating-point arithmetic that assume
3626 that arguments and results are not NaNs or +-Infs.
3628 This option should never be turned on by any @option{-O} option since
3629 it can result in incorrect output for programs which depend on
3630 an exact implementation of IEEE or ISO rules/specifications.
3632 The default is @option{-fno-finite-math-only}.
3634 @item -fno-trapping-math
3635 @opindex fno-trapping-math
3636 Compile code assuming that floating-point operations cannot generate
3637 user-visible traps. These traps include division by zero, overflow,
3638 underflow, inexact result and invalid operation. This option implies
3639 @option{-fno-signaling-nans}. Setting this option may allow faster
3640 code if one relies on ``non-stop'' IEEE arithmetic, for example.
3642 This option should never be turned on by any @option{-O} option since
3643 it can result in incorrect output for programs which depend on
3644 an exact implementation of IEEE or ISO rules/specifications for
3647 The default is @option{-ftrapping-math}.
3649 @item -fsignaling-nans
3650 @opindex fsignaling-nans
3651 Compile code assuming that IEEE signaling NaNs may generate user-visible
3652 traps during floating-point operations. Setting this option disables
3653 optimizations that may change the number of exceptions visible with
3654 signaling NaNs. This option implies @option{-ftrapping-math}.
3656 This option causes the preprocessor macro @code{__SUPPORT_SNAN__} to
3659 The default is @option{-fno-signaling-nans}.
3661 This option is experimental and does not currently guarantee to
3662 disable all GCC optimizations that affect signaling NaN behavior.
3664 @item -fno-zero-initialized-in-bss
3665 @opindex fno-zero-initialized-in-bss
3666 If the target supports a BSS section, GCC by default puts variables that
3667 are initialized to zero into BSS@. This can save space in the resulting
3670 This option turns off this behavior because some programs explicitly
3671 rely on variables going to the data section. E.g., so that the
3672 resulting executable can find the beginning of that section and/or make
3673 assumptions based on that.
3675 The default is @option{-fzero-initialized-in-bss}.
3677 @item -fbounds-check
3678 @opindex fbounds-check
3679 For front-ends that support it, generate additional code to check that
3680 indices used to access arrays are within the declared range. This is
3681 currenly only supported by the Java and Fortran 77 front-ends, where
3682 this option defaults to true and false respectively.
3686 The following options control specific optimizations. The @option{-O2}
3687 option turns on all of these optimizations except @option{-funroll-loops}
3688 and @option{-funroll-all-loops}. On most machines, the @option{-O} option
3689 turns on the @option{-fthread-jumps} and @option{-fdelayed-branch} options,
3690 but specific machines may handle it differently.
3692 You can use the following flags in the rare cases when ``fine-tuning''
3693 of optimizations to be performed is desired.
3695 Not all of the optimizations performed by GCC have @option{-f} options
3699 @item -fstrength-reduce
3700 @opindex fstrength-reduce
3701 Perform the optimizations of loop strength reduction and
3702 elimination of iteration variables.
3704 @item -fthread-jumps
3705 @opindex fthread-jumps
3706 Perform optimizations where we check to see if a jump branches to a
3707 location where another comparison subsumed by the first is found. If
3708 so, the first branch is redirected to either the destination of the
3709 second branch or a point immediately following it, depending on whether
3710 the condition is known to be true or false.
3712 @item -fcse-follow-jumps
3713 @opindex fcse-follow-jumps
3714 In common subexpression elimination, scan through jump instructions
3715 when the target of the jump is not reached by any other path. For
3716 example, when CSE encounters an @code{if} statement with an
3717 @code{else} clause, CSE will follow the jump when the condition
3720 @item -fcse-skip-blocks
3721 @opindex fcse-skip-blocks
3722 This is similar to @option{-fcse-follow-jumps}, but causes CSE to
3723 follow jumps which conditionally skip over blocks. When CSE
3724 encounters a simple @code{if} statement with no else clause,
3725 @option{-fcse-skip-blocks} causes CSE to follow the jump around the
3726 body of the @code{if}.
3728 @item -frerun-cse-after-loop
3729 @opindex frerun-cse-after-loop
3730 Re-run common subexpression elimination after loop optimizations has been
3733 @item -frerun-loop-opt
3734 @opindex frerun-loop-opt
3735 Run the loop optimizer twice.
3739 Perform a global common subexpression elimination pass.
3740 This pass also performs global constant and copy propagation.
3742 @emph{Note:} When compiling a program using computed gotos, a GCC
3743 extension, you may get better runtime performance if you disable
3744 the global common subexpression elmination pass by adding
3745 @option{-fno-gcse} to the command line.
3749 When @option{-fgcse-lm} is enabled, global common subexpression elimination will
3750 attempt to move loads which are only killed by stores into themselves. This
3751 allows a loop containing a load/store sequence to be changed to a load outside
3752 the loop, and a copy/store within the loop.
3756 When @option{-fgcse-sm} is enabled, A store motion pass is run after global common
3757 subexpression elimination. This pass will attempt to move stores out of loops.
3758 When used in conjunction with @option{-fgcse-lm}, loops containing a load/store sequence
3759 can be changed to a load before the loop and a store after the loop.
3761 @item -floop-optimize
3762 @opindex floop-optimize
3763 Perform loop optimizations: move constant expressions out of loops, simplify
3764 exit test conditions and optionally do strength-reduction and loop unrolling as
3767 @item -fcrossjumping
3768 @opindex crossjumping
3769 Perform cross-jumping transformation. This transformation unifies equivalent code and save code size. The
3770 resulting code may or may not perform better than without cross-jumping.
3772 @item -fif-conversion
3773 @opindex if-conversion
3774 Attempt to transform conditional jumps into branch-less equivalents. This
3775 include use of conditional moves, min, max, set flags and abs instructions, and
3776 some tricks doable by standard arithmetics. The use of conditional execution
3777 on chips where it is available is controlled by @code{if-conversion2}.
3779 @item -fif-conversion2
3780 @opindex if-conversion2
3781 Use conditional execution (where available) to transform conditional jumps into
3782 branch-less equivalents.
3784 @item -fdelete-null-pointer-checks
3785 @opindex fdelete-null-pointer-checks
3786 Use global dataflow analysis to identify and eliminate useless checks
3787 for null pointers. The compiler assumes that dereferencing a null
3788 pointer would have halted the program. If a pointer is checked after
3789 it has already been dereferenced, it cannot be null.
3791 In some environments, this assumption is not true, and programs can
3792 safely dereference null pointers. Use
3793 @option{-fno-delete-null-pointer-checks} to disable this optimization
3794 for programs which depend on that behavior.
3796 @item -fexpensive-optimizations
3797 @opindex fexpensive-optimizations
3798 Perform a number of minor optimizations that are relatively expensive.
3800 @item -foptimize-register-move
3802 @opindex foptimize-register-move
3804 Attempt to reassign register numbers in move instructions and as
3805 operands of other simple instructions in order to maximize the amount of
3806 register tying. This is especially helpful on machines with two-operand
3807 instructions. GCC enables this optimization by default with @option{-O2}
3810 Note @option{-fregmove} and @option{-foptimize-register-move} are the same
3813 @item -fdelayed-branch
3814 @opindex fdelayed-branch
3815 If supported for the target machine, attempt to reorder instructions
3816 to exploit instruction slots available after delayed branch
3819 @item -fschedule-insns
3820 @opindex fschedule-insns
3821 If supported for the target machine, attempt to reorder instructions to
3822 eliminate execution stalls due to required data being unavailable. This
3823 helps machines that have slow floating point or memory load instructions
3824 by allowing other instructions to be issued until the result of the load
3825 or floating point instruction is required.
3827 @item -fschedule-insns2
3828 @opindex fschedule-insns2
3829 Similar to @option{-fschedule-insns}, but requests an additional pass of
3830 instruction scheduling after register allocation has been done. This is
3831 especially useful on machines with a relatively small number of
3832 registers and where memory load instructions take more than one cycle.
3834 @item -fno-sched-interblock
3835 @opindex fno-sched-interblock
3836 Don't schedule instructions across basic blocks. This is normally
3837 enabled by default when scheduling before register allocation, i.e.@:
3838 with @option{-fschedule-insns} or at @option{-O2} or higher.
3840 @item -fno-sched-spec
3841 @opindex fno-sched-spec
3842 Don't allow speculative motion of non-load instructions. This is normally
3843 enabled by default when scheduling before register allocation, i.e.@:
3844 with @option{-fschedule-insns} or at @option{-O2} or higher.
3846 @item -fsched-spec-load
3847 @opindex fsched-spec-load
3848 Allow speculative motion of some load instructions. This only makes
3849 sense when scheduling before register allocation, i.e.@: with
3850 @option{-fschedule-insns} or at @option{-O2} or higher.
3852 @item -fsched-spec-load-dangerous
3853 @opindex fsched-spec-load-dangerous
3854 Allow speculative motion of more load instructions. This only makes
3855 sense when scheduling before register allocation, i.e.@: with
3856 @option{-fschedule-insns} or at @option{-O2} or higher.
3858 @item -ffunction-sections
3859 @itemx -fdata-sections
3860 @opindex ffunction-sections
3861 @opindex fdata-sections
3862 Place each function or data item into its own section in the output
3863 file if the target supports arbitrary sections. The name of the
3864 function or the name of the data item determines the section's name
3867 Use these options on systems where the linker can perform optimizations
3868 to improve locality of reference in the instruction space. HPPA
3869 processors running HP-UX and SPARC processors running Solaris 2 have
3870 linkers with such optimizations. Other systems using the ELF object format
3871 as well as AIX may have these optimizations in the future.
3873 Only use these options when there are significant benefits from doing
3874 so. When you specify these options, the assembler and linker will
3875 create larger object and executable files and will also be slower.
3876 You will not be able to use @code{gprof} on all systems if you
3877 specify this option and you may have problems with debugging if
3878 you specify both this option and @option{-g}.
3880 @item -fcaller-saves
3881 @opindex fcaller-saves
3882 Enable values to be allocated in registers that will be clobbered by
3883 function calls, by emitting extra instructions to save and restore the
3884 registers around such calls. Such allocation is done only when it
3885 seems to result in better code than would otherwise be produced.
3887 This option is always enabled by default on certain machines, usually
3888 those which have no call-preserved registers to use instead.
3890 For all machines, optimization level 2 and higher enables this flag by
3895 Perform tail duplication to enlarge superblock size. This transformation
3896 simplifies the control flow of the function allowing other optimizations to do
3899 @item -funroll-loops
3900 @opindex funroll-loops
3901 Unroll loops whose number of iterations can be determined at compile
3902 time or upon entry to the loop. @option{-funroll-loops} implies both
3903 @option{-fstrength-reduce} and @option{-frerun-cse-after-loop}. This
3904 option makes code larger, and may or may not make it run faster.
3906 @item -funroll-all-loops
3907 @opindex funroll-all-loops
3908 Unroll all loops, even if their number of iterations is uncertain when
3909 the loop is entered. This usually makes programs run more slowly.
3910 @option{-funroll-all-loops} implies the same options as
3911 @option{-funroll-loops},
3913 @item -fprefetch-loop-arrays
3914 @opindex fprefetch-loop-arrays
3915 If supported by the target machine, generate instructions to prefetch
3916 memory to improve the performance of loops that access large arrays.
3918 @item -fmove-all-movables
3919 @opindex fmove-all-movables
3920 Forces all invariant computations in loops to be moved
3923 @item -freduce-all-givs
3924 @opindex freduce-all-givs
3925 Forces all general-induction variables in loops to be
3928 @emph{Note:} When compiling programs written in Fortran,
3929 @option{-fmove-all-movables} and @option{-freduce-all-givs} are enabled
3930 by default when you use the optimizer.
3932 These options may generate better or worse code; results are highly
3933 dependent on the structure of loops within the source code.
3935 These two options are intended to be removed someday, once
3936 they have helped determine the efficacy of various
3937 approaches to improving loop optimizations.
3939 Please let us (@w{@email{gcc@@gcc.gnu.org}} and @w{@email{fortran@@gnu.org}})
3940 know how use of these options affects
3941 the performance of your production code.
3942 We're very interested in code that runs @emph{slower}
3943 when these options are @emph{enabled}.
3946 @itemx -fno-peephole2
3947 @opindex fno-peephole
3948 @opindex fno-peephole2
3949 Disable any machine-specific peephole optimizations. The difference
3950 between @option{-fno-peephole} and @option{-fno-peephole2} is in how they
3951 are implemented in the compiler; some targets use one, some use the
3952 other, a few use both.
3954 @item -fbranch-probabilities
3955 @opindex fbranch-probabilities
3956 After running a program compiled with @option{-fprofile-arcs}
3957 (@pxref{Debugging Options,, Options for Debugging Your Program or
3958 @command{gcc}}), you can compile it a second time using
3959 @option{-fbranch-probabilities}, to improve optimizations based on
3960 the number of times each branch was taken. When the program
3961 compiled with @option{-fprofile-arcs} exits it saves arc execution
3962 counts to a file called @file{@var{sourcename}.da} for each source
3963 file The information in this data file is very dependent on the
3964 structure of the generated code, so you must use the same source code
3965 and the same optimization options for both compilations.
3967 With @option{-fbranch-probabilities}, GCC puts a @samp{REG_EXEC_COUNT}
3968 note on the first instruction of each basic block, and a
3969 @samp{REG_BR_PROB} note on each @samp{JUMP_INSN} and @samp{CALL_INSN}.
3970 These can be used to improve optimization. Currently, they are only
3971 used in one place: in @file{reorg.c}, instead of guessing which path a
3972 branch is mostly to take, the @samp{REG_BR_PROB} values are used to
3973 exactly determine which path is taken more often.
3975 @item -fno-guess-branch-probability
3976 @opindex fno-guess-branch-probability
3977 Do not guess branch probabilities using a randomized model.
3979 Sometimes gcc will opt to use a randomized model to guess branch
3980 probabilities, when none are available from either profiling feedback
3981 (@option{-fprofile-arcs}) or @samp{__builtin_expect}. This means that
3982 different runs of the compiler on the same program may produce different
3985 In a hard real-time system, people don't want different runs of the
3986 compiler to produce code that has different behavior; minimizing
3987 non-determinism is of paramount import. This switch allows users to
3988 reduce non-determinism, possibly at the expense of inferior
3991 @item -freorder-blocks
3992 @opindex freorder-blocks
3993 Reorder basic blocks in the compiled function in order to reduce number of
3994 taken branches and improve code locality.
3996 @item -freorder-functions
3997 @opindex freorder-functions
3998 Reorder basic blocks in the compiled function in order to reduce number of
3999 taken branches and improve code locality. This is implemented by using special
4000 subsections @code{text.hot} for most frequently executed functions and
4001 @code{text.unlikely} for unlikely executed functions. Reordering is done by
4002 the linker so object file format must support named sections and linker must
4003 place them in resonable way.
4005 Also profile feedback must be available in to make this option effective. See
4006 @option{-fprofile-arcs} for details.
4008 @item -fstrict-aliasing
4009 @opindex fstrict-aliasing
4010 Allows the compiler to assume the strictest aliasing rules applicable to
4011 the language being compiled. For C (and C++), this activates
4012 optimizations based on the type of expressions. In particular, an
4013 object of one type is assumed never to reside at the same address as an
4014 object of a different type, unless the types are almost the same. For
4015 example, an @code{unsigned int} can alias an @code{int}, but not a
4016 @code{void*} or a @code{double}. A character type may alias any other
4019 Pay special attention to code like this:
4032 The practice of reading from a different union member than the one most
4033 recently written to (called ``type-punning'') is common. Even with
4034 @option{-fstrict-aliasing}, type-punning is allowed, provided the memory
4035 is accessed through the union type. So, the code above will work as
4036 expected. However, this code might not:
4047 Every language that wishes to perform language-specific alias analysis
4048 should define a function that computes, given an @code{tree}
4049 node, an alias set for the node. Nodes in different alias sets are not
4050 allowed to alias. For an example, see the C front-end function
4051 @code{c_get_alias_set}.
4053 For all machines, optimization level 2 and higher enables this flag by
4056 @item -falign-functions
4057 @itemx -falign-functions=@var{n}
4058 @opindex falign-functions
4059 Align the start of functions to the next power-of-two greater than
4060 @var{n}, skipping up to @var{n} bytes. For instance,
4061 @option{-falign-functions=32} aligns functions to the next 32-byte
4062 boundary, but @option{-falign-functions=24} would align to the next
4063 32-byte boundary only if this can be done by skipping 23 bytes or less.
4065 @option{-fno-align-functions} and @option{-falign-functions=1} are
4066 equivalent and mean that functions will not be aligned.
4068 Some assemblers only support this flag when @var{n} is a power of two;
4069 in that case, it is rounded up.
4071 If @var{n} is not specified, use a machine-dependent default.
4073 @item -falign-labels
4074 @itemx -falign-labels=@var{n}
4075 @opindex falign-labels
4076 Align all branch targets to a power-of-two boundary, skipping up to
4077 @var{n} bytes like @option{-falign-functions}. This option can easily
4078 make code slower, because it must insert dummy operations for when the
4079 branch target is reached in the usual flow of the code.
4081 If @option{-falign-loops} or @option{-falign-jumps} are applicable and
4082 are greater than this value, then their values are used instead.
4084 If @var{n} is not specified, use a machine-dependent default which is
4085 very likely to be @samp{1}, meaning no alignment.
4088 @itemx -falign-loops=@var{n}
4089 @opindex falign-loops
4090 Align loops to a power-of-two boundary, skipping up to @var{n} bytes
4091 like @option{-falign-functions}. The hope is that the loop will be
4092 executed many times, which will make up for any execution of the dummy
4095 If @var{n} is not specified, use a machine-dependent default.
4098 @itemx -falign-jumps=@var{n}
4099 @opindex falign-jumps
4100 Align branch targets to a power-of-two boundary, for branch targets
4101 where the targets can only be reached by jumping, skipping up to @var{n}
4102 bytes like @option{-falign-functions}. In this case, no dummy operations
4105 If @var{n} is not specified, use a machine-dependent default.
4109 Perform optimizations in static single assignment form. Each function's
4110 flow graph is translated into SSA form, optimizations are performed, and
4111 the flow graph is translated back from SSA form. Users should not
4112 specify this option, since it is not yet ready for production use.
4116 Perform Sparse Conditional Constant Propagation in SSA form. Requires
4117 @option{-fssa}. Like @option{-fssa}, this is an experimental feature.
4121 Perform aggressive dead-code elimination in SSA form. Requires @option{-fssa}.
4122 Like @option{-fssa}, this is an experimental feature.
4124 @item -fsingle-precision-constant
4125 @opindex fsingle-precision-constant
4126 Treat floating point constant as single precision constant instead of
4127 implicitly converting it to double precision constant.
4129 @item -frename-registers
4130 @opindex frename-registers
4131 Attempt to avoid false dependencies in scheduled code by making use
4132 of registers left over after register allocation. This optimization
4133 will most benefit processors with lots of registers. It can, however,
4134 make debugging impossible, since variables will no longer stay in
4135 a ``home register''.
4137 @item -fno-cprop-registers
4138 @opindex fno-cprop-registers
4139 After register allocation and post-register allocation instruction splitting,
4140 we perform a copy-propagation pass to try to reduce scheduling dependencies
4141 and occasionally eliminate the copy.
4143 @item --param @var{name}=@var{value}
4145 In some places, GCC uses various constants to control the amount of
4146 optimization that is done. For example, GCC will not inline functions
4147 that contain more that a certain number of instructions. You can
4148 control some of these constants on the command-line using the
4149 @option{--param} option.
4151 In each case, the @var{value} is an integer. The allowable choices for
4152 @var{name} are given in the following table:
4155 @item max-delay-slot-insn-search
4156 The maximum number of instructions to consider when looking for an
4157 instruction to fill a delay slot. If more than this arbitrary number of
4158 instructions is searched, the time savings from filling the delay slot
4159 will be minimal so stop searching. Increasing values mean more
4160 aggressive optimization, making the compile time increase with probably
4161 small improvement in executable run time.
4163 @item max-delay-slot-live-search
4164 When trying to fill delay slots, the maximum number of instructions to
4165 consider when searching for a block with valid live register
4166 information. Increasing this arbitrarily chosen value means more
4167 aggressive optimization, increasing the compile time. This parameter
4168 should be removed when the delay slot code is rewritten to maintain the
4171 @item max-gcse-memory
4172 The approximate maximum amount of memory that will be allocated in
4173 order to perform the global common subexpression elimination
4174 optimization. If more memory than specified is required, the
4175 optimization will not be done.
4177 @item max-gcse-passes
4178 The maximum number of passes of GCSE to run.
4180 @item max-pending-list-length
4181 The maximum number of pending dependencies scheduling will allow
4182 before flushing the current state and starting over. Large functions
4183 with few branches or calls can create excessively large lists which
4184 needlessly consume memory and resources.
4186 @item max-inline-insns
4187 If an function contains more than this many instructions, it
4188 will not be inlined. This option is precisely equivalent to
4189 @option{-finline-limit}.
4191 @item max-unrolled-insns
4192 The maximum number of instructions that a loop should have if that loop
4193 is unrolled, and if the loop is unrolled, it determines how many times
4194 the loop code is unrolled.
4196 @item hot-bb-count-fraction
4197 Select fraction of the maximal count of repetitions of basic block in program
4198 given basic block needs to have to be considered hot.
4200 @item hot-bb-frequency-fraction
4201 Select fraction of the maximal frequency of executions of basic block in
4202 function given basic block needs to have to be considered hot
4204 @item tracer-dynamic-coverage
4205 @itemx tracer-dynamic-coverage-feedback
4207 This value is used to limit superblock formation once given percentage of
4208 executed instructions is covered. This limits unnecesary code size expansion.
4210 The @option{tracer-dynamic-coverage-feedback} is used only when profile
4211 feedback is available. The real profiles (as opposed to statically estimated
4212 ones) are much less balanced allowing the threshold to be larger value.
4214 @item tracer-max-code-growth
4215 Stop tail duplication once code growth has reached given percentage. This is
4216 rather hokey argument, as most of the duplicates will be elliminated later in
4217 cross jumping, so it may be set to much higher values than is the desired code
4220 @item tracer-min-branch-ratio
4222 Stop reverse growth when the reverse probability of best edge is less than this
4223 threshold (in percent).
4225 @item tracer-min-branch-ratio
4226 @itemx tracer-min-branch-ratio-feedback
4228 Stop forward growth if the best edge do have probability lower than this
4231 Similary to @option{tracer-dynamic-coverage} two values are present, one for
4232 compilation for profile feedback and one for compilation without. The value
4233 for compilation with profile feedback needs to be more conservative (higher) in
4234 order to make tracer effective.
4236 @item ggc-min-expand
4238 GCC uses a garbage collector to manage its own memory allocation. This
4239 parameter specifies the minimum percentage by which the garbage
4240 collector's heap should be allowed to expand between collections.
4241 Tuning this may improve compilation speed; it has no effect on code
4244 The default is 30%. Setting this parameter and
4245 @option{ggc-min-heapsize} to zero causes a full collection to occur at
4246 every opportunity. This is extremely slow, but can be useful for
4249 @item ggc-min-heapsize
4251 Minimum size of the garbage collector's heap before it begins bothering
4252 to collect garbage. The first collection occurs after the heap expands
4253 by @option{ggc-min-expand}% beyond @option{ggc-min-heapsize}. Again,
4254 tuning this may improve compilation speed, and has no effect on code
4257 The default is 4096 (four megabytes). Setting this parameter very large
4258 effectively disables garbage collection. Setting this parameter and
4259 @option{ggc-min-expand} to zero causes a full collection to occur at
4265 @node Preprocessor Options
4266 @section Options Controlling the Preprocessor
4267 @cindex preprocessor options
4268 @cindex options, preprocessor
4270 These options control the C preprocessor, which is run on each C source
4271 file before actual compilation.
4273 If you use the @option{-E} option, nothing is done except preprocessing.
4274 Some of these options make sense only together with @option{-E} because
4275 they cause the preprocessor output to be unsuitable for actual
4279 You can use @option{-Wp,@var{option}} to bypass the compiler driver
4280 and pass @var{option} directly through to the preprocessor. If
4281 @var{option} contains commas, it is split into multiple options at the
4282 commas. However, many options are modified, translated or interpreted
4283 by the compiler driver before being passed to the preprocessor, and
4284 @option{-Wp} forcibly bypasses this phase. The preprocessor's direct
4285 interface is undocumented and subject to change, so whenever possible
4286 you should avoid using @option{-Wp} and let the driver handle the
4289 @include cppopts.texi
4291 @node Assembler Options
4292 @section Passing Options to the Assembler
4294 @c prevent bad page break with this line
4295 You can pass options to the assembler.
4298 @item -Wa,@var{option}
4300 Pass @var{option} as an option to the assembler. If @var{option}
4301 contains commas, it is split into multiple options at the commas.
4305 @section Options for Linking
4306 @cindex link options
4307 @cindex options, linking
4309 These options come into play when the compiler links object files into
4310 an executable output file. They are meaningless if the compiler is
4311 not doing a link step.
4315 @item @var{object-file-name}
4316 A file name that does not end in a special recognized suffix is
4317 considered to name an object file or library. (Object files are
4318 distinguished from libraries by the linker according to the file
4319 contents.) If linking is done, these object files are used as input
4328 If any of these options is used, then the linker is not run, and
4329 object file names should not be used as arguments. @xref{Overall
4333 @item -l@var{library}
4334 @itemx -l @var{library}
4336 Search the library named @var{library} when linking. (The second
4337 alternative with the library as a separate argument is only for
4338 POSIX compliance and is not recommended.)
4340 It makes a difference where in the command you write this option; the
4341 linker searches and processes libraries and object files in the order they
4342 are specified. Thus, @samp{foo.o -lz bar.o} searches library @samp{z}
4343 after file @file{foo.o} but before @file{bar.o}. If @file{bar.o} refers
4344 to functions in @samp{z}, those functions may not be loaded.
4346 The linker searches a standard list of directories for the library,
4347 which is actually a file named @file{lib@var{library}.a}. The linker
4348 then uses this file as if it had been specified precisely by name.
4350 The directories searched include several standard system directories
4351 plus any that you specify with @option{-L}.
4353 Normally the files found this way are library files---archive files
4354 whose members are object files. The linker handles an archive file by
4355 scanning through it for members which define symbols that have so far
4356 been referenced but not defined. But if the file that is found is an
4357 ordinary object file, it is linked in the usual fashion. The only
4358 difference between using an @option{-l} option and specifying a file name
4359 is that @option{-l} surrounds @var{library} with @samp{lib} and @samp{.a}
4360 and searches several directories.
4364 You need this special case of the @option{-l} option in order to
4365 link an Objective-C program.
4368 @opindex nostartfiles
4369 Do not use the standard system startup files when linking.
4370 The standard system libraries are used normally, unless @option{-nostdlib}
4371 or @option{-nodefaultlibs} is used.
4373 @item -nodefaultlibs
4374 @opindex nodefaultlibs
4375 Do not use the standard system libraries when linking.
4376 Only the libraries you specify will be passed to the linker.
4377 The standard startup files are used normally, unless @option{-nostartfiles}
4378 is used. The compiler may generate calls to memcmp, memset, and memcpy
4379 for System V (and ISO C) environments or to bcopy and bzero for
4380 BSD environments. These entries are usually resolved by entries in
4381 libc. These entry points should be supplied through some other
4382 mechanism when this option is specified.
4386 Do not use the standard system startup files or libraries when linking.
4387 No startup files and only the libraries you specify will be passed to
4388 the linker. The compiler may generate calls to memcmp, memset, and memcpy
4389 for System V (and ISO C) environments or to bcopy and bzero for
4390 BSD environments. These entries are usually resolved by entries in
4391 libc. These entry points should be supplied through some other
4392 mechanism when this option is specified.
4394 @cindex @option{-lgcc}, use with @option{-nostdlib}
4395 @cindex @option{-nostdlib} and unresolved references
4396 @cindex unresolved references and @option{-nostdlib}
4397 @cindex @option{-lgcc}, use with @option{-nodefaultlibs}
4398 @cindex @option{-nodefaultlibs} and unresolved references
4399 @cindex unresolved references and @option{-nodefaultlibs}
4400 One of the standard libraries bypassed by @option{-nostdlib} and
4401 @option{-nodefaultlibs} is @file{libgcc.a}, a library of internal subroutines
4402 that GCC uses to overcome shortcomings of particular machines, or special
4403 needs for some languages.
4404 (@xref{Interface,,Interfacing to GCC Output,gccint,GNU Compiler
4405 Collection (GCC) Internals},
4406 for more discussion of @file{libgcc.a}.)
4407 In most cases, you need @file{libgcc.a} even when you want to avoid
4408 other standard libraries. In other words, when you specify @option{-nostdlib}
4409 or @option{-nodefaultlibs} you should usually specify @option{-lgcc} as well.
4410 This ensures that you have no unresolved references to internal GCC
4411 library subroutines. (For example, @samp{__main}, used to ensure C++
4412 constructors will be called; @pxref{Collect2,,@code{collect2}, gccint,
4413 GNU Compiler Collection (GCC) Internals}.)
4417 Remove all symbol table and relocation information from the executable.
4421 On systems that support dynamic linking, this prevents linking with the shared
4422 libraries. On other systems, this option has no effect.
4426 Produce a shared object which can then be linked with other objects to
4427 form an executable. Not all systems support this option. For predictable
4428 results, you must also specify the same set of options that were used to
4429 generate code (@option{-fpic}, @option{-fPIC}, or model suboptions)
4430 when you specify this option.@footnote{On some systems, @samp{gcc -shared}
4431 needs to build supplementary stub code for constructors to work. On
4432 multi-libbed systems, @samp{gcc -shared} must select the correct support
4433 libraries to link against. Failing to supply the correct flags may lead
4434 to subtle defects. Supplying them in cases where they are not necessary
4437 @item -shared-libgcc
4438 @itemx -static-libgcc
4439 @opindex shared-libgcc
4440 @opindex static-libgcc
4441 On systems that provide @file{libgcc} as a shared library, these options
4442 force the use of either the shared or static version respectively.
4443 If no shared version of @file{libgcc} was built when the compiler was
4444 configured, these options have no effect.
4446 There are several situations in which an application should use the
4447 shared @file{libgcc} instead of the static version. The most common
4448 of these is when the application wishes to throw and catch exceptions
4449 across different shared libraries. In that case, each of the libraries
4450 as well as the application itself should use the shared @file{libgcc}.
4452 Therefore, the G++ and GCJ drivers automatically add
4453 @option{-shared-libgcc} whenever you build a shared library or a main
4454 executable, because C++ and Java programs typically use exceptions, so
4455 this is the right thing to do.
4457 If, instead, you use the GCC driver to create shared libraries, you may
4458 find that they will not always be linked with the shared @file{libgcc}.
4459 If GCC finds, at its configuration time, that you have a GNU linker that
4460 does not support option @option{--eh-frame-hdr}, it will link the shared
4461 version of @file{libgcc} into shared libraries by default. Otherwise,
4462 it will take advantage of the linker and optimize away the linking with
4463 the shared version of @file{libgcc}, linking with the static version of
4464 libgcc by default. This allows exceptions to propagate through such
4465 shared libraries, without incurring relocation costs at library load
4468 However, if a library or main executable is supposed to throw or catch
4469 exceptions, you must link it using the G++ or GCJ driver, as appropriate
4470 for the languages used in the program, or using the option
4471 @option{-shared-libgcc}, such that it is linked with the shared
4476 Bind references to global symbols when building a shared object. Warn
4477 about any unresolved references (unless overridden by the link editor
4478 option @samp{-Xlinker -z -Xlinker defs}). Only a few systems support
4481 @item -Xlinker @var{option}
4483 Pass @var{option} as an option to the linker. You can use this to
4484 supply system-specific linker options which GCC does not know how to
4487 If you want to pass an option that takes an argument, you must use
4488 @option{-Xlinker} twice, once for the option and once for the argument.
4489 For example, to pass @option{-assert definitions}, you must write
4490 @samp{-Xlinker -assert -Xlinker definitions}. It does not work to write
4491 @option{-Xlinker "-assert definitions"}, because this passes the entire
4492 string as a single argument, which is not what the linker expects.
4494 @item -Wl,@var{option}
4496 Pass @var{option} as an option to the linker. If @var{option} contains
4497 commas, it is split into multiple options at the commas.
4499 @item -u @var{symbol}
4501 Pretend the symbol @var{symbol} is undefined, to force linking of
4502 library modules to define it. You can use @option{-u} multiple times with
4503 different symbols to force loading of additional library modules.
4506 @node Directory Options
4507 @section Options for Directory Search
4508 @cindex directory options
4509 @cindex options, directory search
4512 These options specify directories to search for header files, for
4513 libraries and for parts of the compiler:
4518 Add the directory @var{dir} to the head of the list of directories to be
4519 searched for header files. This can be used to override a system header
4520 file, substituting your own version, since these directories are
4521 searched before the system header file directories. However, you should
4522 not use this option to add directories that contain vendor-supplied
4523 system header files (use @option{-isystem} for that). If you use more than
4524 one @option{-I} option, the directories are scanned in left-to-right
4525 order; the standard system directories come after.
4527 If a standard system include directory, or a directory specified with
4528 @option{-isystem}, is also specified with @option{-I}, the @option{-I}
4529 option will be ignored. The directory will still be searched but as a
4530 system directory at its normal position in the system include chain.
4531 This is to ensure that GCC's procedure to fix buggy system headers and
4532 the ordering for the include_next directive are not inadvertantly changed.
4533 If you really need to change the search order for system directories,
4534 use the @option{-nostdinc} and/or @option{-isystem} options.
4538 Any directories you specify with @option{-I} options before the @option{-I-}
4539 option are searched only for the case of @samp{#include "@var{file}"};
4540 they are not searched for @samp{#include <@var{file}>}.
4542 If additional directories are specified with @option{-I} options after
4543 the @option{-I-}, these directories are searched for all @samp{#include}
4544 directives. (Ordinarily @emph{all} @option{-I} directories are used
4547 In addition, the @option{-I-} option inhibits the use of the current
4548 directory (where the current input file came from) as the first search
4549 directory for @samp{#include "@var{file}"}. There is no way to
4550 override this effect of @option{-I-}. With @option{-I.} you can specify
4551 searching the directory which was current when the compiler was
4552 invoked. That is not exactly the same as what the preprocessor does
4553 by default, but it is often satisfactory.
4555 @option{-I-} does not inhibit the use of the standard system directories
4556 for header files. Thus, @option{-I-} and @option{-nostdinc} are
4561 Add directory @var{dir} to the list of directories to be searched
4564 @item -B@var{prefix}
4566 This option specifies where to find the executables, libraries,
4567 include files, and data files of the compiler itself.
4569 The compiler driver program runs one or more of the subprograms
4570 @file{cpp}, @file{cc1}, @file{as} and @file{ld}. It tries
4571 @var{prefix} as a prefix for each program it tries to run, both with and
4572 without @samp{@var{machine}/@var{version}/} (@pxref{Target Options}).
4574 For each subprogram to be run, the compiler driver first tries the
4575 @option{-B} prefix, if any. If that name is not found, or if @option{-B}
4576 was not specified, the driver tries two standard prefixes, which are
4577 @file{/usr/lib/gcc/} and @file{/usr/local/lib/gcc-lib/}. If neither of
4578 those results in a file name that is found, the unmodified program
4579 name is searched for using the directories specified in your
4580 @env{PATH} environment variable.
4582 The compiler will check to see if the path provided by the @option{-B}
4583 refers to a directory, and if necessary it will add a directory
4584 separator character at the end of the path.
4586 @option{-B} prefixes that effectively specify directory names also apply
4587 to libraries in the linker, because the compiler translates these
4588 options into @option{-L} options for the linker. They also apply to
4589 includes files in the preprocessor, because the compiler translates these
4590 options into @option{-isystem} options for the preprocessor. In this case,
4591 the compiler appends @samp{include} to the prefix.
4593 The run-time support file @file{libgcc.a} can also be searched for using
4594 the @option{-B} prefix, if needed. If it is not found there, the two
4595 standard prefixes above are tried, and that is all. The file is left
4596 out of the link if it is not found by those means.
4598 Another way to specify a prefix much like the @option{-B} prefix is to use
4599 the environment variable @env{GCC_EXEC_PREFIX}. @xref{Environment
4602 As a special kludge, if the path provided by @option{-B} is
4603 @file{[dir/]stage@var{N}/}, where @var{N} is a number in the range 0 to
4604 9, then it will be replaced by @file{[dir/]include}. This is to help
4605 with boot-strapping the compiler.
4607 @item -specs=@var{file}
4609 Process @var{file} after the compiler reads in the standard @file{specs}
4610 file, in order to override the defaults that the @file{gcc} driver
4611 program uses when determining what switches to pass to @file{cc1},
4612 @file{cc1plus}, @file{as}, @file{ld}, etc. More than one
4613 @option{-specs=@var{file}} can be specified on the command line, and they
4614 are processed in order, from left to right.
4620 @section Specifying subprocesses and the switches to pass to them
4622 @command{gcc} is a driver program. It performs its job by invoking a
4623 sequence of other programs to do the work of compiling, assembling and
4624 linking. GCC interprets its command-line parameters and uses these to
4625 deduce which programs it should invoke, and which command-line options
4626 it ought to place on their command lines. This behavior is controlled
4627 by @dfn{spec strings}. In most cases there is one spec string for each
4628 program that GCC can invoke, but a few programs have multiple spec
4629 strings to control their behavior. The spec strings built into GCC can
4630 be overridden by using the @option{-specs=} command-line switch to specify
4633 @dfn{Spec files} are plaintext files that are used to construct spec
4634 strings. They consist of a sequence of directives separated by blank
4635 lines. The type of directive is determined by the first non-whitespace
4636 character on the line and it can be one of the following:
4639 @item %@var{command}
4640 Issues a @var{command} to the spec file processor. The commands that can
4644 @item %include <@var{file}>
4646 Search for @var{file} and insert its text at the current point in the
4649 @item %include_noerr <@var{file}>
4650 @cindex %include_noerr
4651 Just like @samp{%include}, but do not generate an error message if the include
4652 file cannot be found.
4654 @item %rename @var{old_name} @var{new_name}
4656 Rename the spec string @var{old_name} to @var{new_name}.
4660 @item *[@var{spec_name}]:
4661 This tells the compiler to create, override or delete the named spec
4662 string. All lines after this directive up to the next directive or
4663 blank line are considered to be the text for the spec string. If this
4664 results in an empty string then the spec will be deleted. (Or, if the
4665 spec did not exist, then nothing will happened.) Otherwise, if the spec
4666 does not currently exist a new spec will be created. If the spec does
4667 exist then its contents will be overridden by the text of this
4668 directive, unless the first character of that text is the @samp{+}
4669 character, in which case the text will be appended to the spec.
4671 @item [@var{suffix}]:
4672 Creates a new @samp{[@var{suffix}] spec} pair. All lines after this directive
4673 and up to the next directive or blank line are considered to make up the
4674 spec string for the indicated suffix. When the compiler encounters an
4675 input file with the named suffix, it will processes the spec string in
4676 order to work out how to compile that file. For example:
4683 This says that any input file whose name ends in @samp{.ZZ} should be
4684 passed to the program @samp{z-compile}, which should be invoked with the
4685 command-line switch @option{-input} and with the result of performing the
4686 @samp{%i} substitution. (See below.)
4688 As an alternative to providing a spec string, the text that follows a
4689 suffix directive can be one of the following:
4692 @item @@@var{language}
4693 This says that the suffix is an alias for a known @var{language}. This is
4694 similar to using the @option{-x} command-line switch to GCC to specify a
4695 language explicitly. For example:
4702 Says that .ZZ files are, in fact, C++ source files.
4705 This causes an error messages saying:
4708 @var{name} compiler not installed on this system.
4712 GCC already has an extensive list of suffixes built into it.
4713 This directive will add an entry to the end of the list of suffixes, but
4714 since the list is searched from the end backwards, it is effectively
4715 possible to override earlier entries using this technique.
4719 GCC has the following spec strings built into it. Spec files can
4720 override these strings or create their own. Note that individual
4721 targets can also add their own spec strings to this list.
4724 asm Options to pass to the assembler
4725 asm_final Options to pass to the assembler post-processor
4726 cpp Options to pass to the C preprocessor
4727 cc1 Options to pass to the C compiler
4728 cc1plus Options to pass to the C++ compiler
4729 endfile Object files to include at the end of the link
4730 link Options to pass to the linker
4731 lib Libraries to include on the command line to the linker
4732 libgcc Decides which GCC support library to pass to the linker
4733 linker Sets the name of the linker
4734 predefines Defines to be passed to the C preprocessor
4735 signed_char Defines to pass to CPP to say whether @code{char} is signed
4737 startfile Object files to include at the start of the link
4740 Here is a small example of a spec file:
4746 --start-group -lgcc -lc -leval1 --end-group %(old_lib)
4749 This example renames the spec called @samp{lib} to @samp{old_lib} and
4750 then overrides the previous definition of @samp{lib} with a new one.
4751 The new definition adds in some extra command-line options before
4752 including the text of the old definition.
4754 @dfn{Spec strings} are a list of command-line options to be passed to their
4755 corresponding program. In addition, the spec strings can contain
4756 @samp{%}-prefixed sequences to substitute variable text or to
4757 conditionally insert text into the command line. Using these constructs
4758 it is possible to generate quite complex command lines.
4760 Here is a table of all defined @samp{%}-sequences for spec
4761 strings. Note that spaces are not generated automatically around the
4762 results of expanding these sequences. Therefore you can concatenate them
4763 together or combine them with constant text in a single argument.
4767 Substitute one @samp{%} into the program name or argument.
4770 Substitute the name of the input file being processed.
4773 Substitute the basename of the input file being processed.
4774 This is the substring up to (and not including) the last period
4775 and not including the directory.
4778 This is the same as @samp{%b}, but include the file suffix (text after
4782 Marks the argument containing or following the @samp{%d} as a
4783 temporary file name, so that that file will be deleted if GCC exits
4784 successfully. Unlike @samp{%g}, this contributes no text to the
4787 @item %g@var{suffix}
4788 Substitute a file name that has suffix @var{suffix} and is chosen
4789 once per compilation, and mark the argument in the same way as
4790 @samp{%d}. To reduce exposure to denial-of-service attacks, the file
4791 name is now chosen in a way that is hard to predict even when previously
4792 chosen file names are known. For example, @samp{%g.s @dots{} %g.o @dots{} %g.s}
4793 might turn into @samp{ccUVUUAU.s ccXYAXZ12.o ccUVUUAU.s}. @var{suffix} matches
4794 the regexp @samp{[.A-Za-z]*} or the special string @samp{%O}, which is
4795 treated exactly as if @samp{%O} had been preprocessed. Previously, @samp{%g}
4796 was simply substituted with a file name chosen once per compilation,
4797 without regard to any appended suffix (which was therefore treated
4798 just like ordinary text), making such attacks more likely to succeed.
4800 @item %u@var{suffix}
4801 Like @samp{%g}, but generates a new temporary file name even if
4802 @samp{%u@var{suffix}} was already seen.
4804 @item %U@var{suffix}
4805 Substitutes the last file name generated with @samp{%u@var{suffix}}, generating a
4806 new one if there is no such last file name. In the absence of any
4807 @samp{%u@var{suffix}}, this is just like @samp{%g@var{suffix}}, except they don't share
4808 the same suffix @emph{space}, so @samp{%g.s @dots{} %U.s @dots{} %g.s @dots{} %U.s}
4809 would involve the generation of two distinct file names, one
4810 for each @samp{%g.s} and another for each @samp{%U.s}. Previously, @samp{%U} was
4811 simply substituted with a file name chosen for the previous @samp{%u},
4812 without regard to any appended suffix.
4814 @item %j@var{SUFFIX}
4815 Substitutes the name of the @code{HOST_BIT_BUCKET}, if any, and if it is
4816 writable, and if save-temps is off; otherwise, substitute the name
4817 of a temporary file, just like @samp{%u}. This temporary file is not
4818 meant for communication between processes, but rather as a junk
4821 @item %.@var{SUFFIX}
4822 Substitutes @var{.SUFFIX} for the suffixes of a matched switch's args
4823 when it is subsequently output with @samp{%*}. @var{SUFFIX} is
4824 terminated by the next space or %.
4827 Marks the argument containing or following the @samp{%w} as the
4828 designated output file of this compilation. This puts the argument
4829 into the sequence of arguments that @samp{%o} will substitute later.
4832 Substitutes the names of all the output files, with spaces
4833 automatically placed around them. You should write spaces
4834 around the @samp{%o} as well or the results are undefined.
4835 @samp{%o} is for use in the specs for running the linker.
4836 Input files whose names have no recognized suffix are not compiled
4837 at all, but they are included among the output files, so they will
4841 Substitutes the suffix for object files. Note that this is
4842 handled specially when it immediately follows @samp{%g, %u, or %U},
4843 because of the need for those to form complete file names. The
4844 handling is such that @samp{%O} is treated exactly as if it had already
4845 been substituted, except that @samp{%g, %u, and %U} do not currently
4846 support additional @var{suffix} characters following @samp{%O} as they would
4847 following, for example, @samp{.o}.
4850 Substitutes the standard macro predefinitions for the
4851 current target machine. Use this when running @code{cpp}.
4854 Like @samp{%p}, but puts @samp{__} before and after the name of each
4855 predefined macro, except for macros that start with @samp{__} or with
4856 @samp{_@var{L}}, where @var{L} is an uppercase letter. This is for ISO
4860 Substitute a @option{-iprefix} option made from @env{GCC_EXEC_PREFIX}.
4863 Current argument is the name of a library or startup file of some sort.
4864 Search for that file in a standard list of directories and substitute
4865 the full name found.
4868 Print @var{str} as an error message. @var{str} is terminated by a newline.
4869 Use this when inconsistent options are detected.
4872 Output @samp{-} if the input for the current command is coming from a pipe.
4875 Substitute the contents of spec string @var{name} at this point.
4878 Like @samp{%(@dots{})} but put @samp{__} around @option{-D} arguments.
4880 @item %x@{@var{option}@}
4881 Accumulate an option for @samp{%X}.
4884 Output the accumulated linker options specified by @option{-Wl} or a @samp{%x}
4888 Output the accumulated assembler options specified by @option{-Wa}.
4891 Output the accumulated preprocessor options specified by @option{-Wp}.
4894 Substitute the major version number of GCC@.
4895 (For version 2.9.5, this is 2.)
4898 Substitute the minor version number of GCC@.
4899 (For version 2.9.5, this is 9.)
4902 Substitute the patch level number of GCC@.
4903 (For version 2.9.5, this is 5.)
4906 Process the @code{asm} spec. This is used to compute the
4907 switches to be passed to the assembler.
4910 Process the @code{asm_final} spec. This is a spec string for
4911 passing switches to an assembler post-processor, if such a program is
4915 Process the @code{link} spec. This is the spec for computing the
4916 command line passed to the linker. Typically it will make use of the
4917 @samp{%L %G %S %D and %E} sequences.
4920 Dump out a @option{-L} option for each directory that GCC believes might
4921 contain startup files. If the target supports multilibs then the
4922 current multilib directory will be prepended to each of these paths.
4925 Output the multilib directory with directory separators replaced with
4926 @samp{_}. If multilib directories are not set, or the multilib directory is
4927 @file{.} then this option emits nothing.
4930 Process the @code{lib} spec. This is a spec string for deciding which
4931 libraries should be included on the command line to the linker.
4934 Process the @code{libgcc} spec. This is a spec string for deciding
4935 which GCC support library should be included on the command line to the linker.
4938 Process the @code{startfile} spec. This is a spec for deciding which
4939 object files should be the first ones passed to the linker. Typically
4940 this might be a file named @file{crt0.o}.
4943 Process the @code{endfile} spec. This is a spec string that specifies
4944 the last object files that will be passed to the linker.
4947 Process the @code{cpp} spec. This is used to construct the arguments
4948 to be passed to the C preprocessor.
4951 Process the @code{signed_char} spec. This is intended to be used
4952 to tell cpp whether a char is signed. It typically has the definition:
4954 %@{funsigned-char:-D__CHAR_UNSIGNED__@}
4958 Process the @code{cc1} spec. This is used to construct the options to be
4959 passed to the actual C compiler (@samp{cc1}).
4962 Process the @code{cc1plus} spec. This is used to construct the options to be
4963 passed to the actual C++ compiler (@samp{cc1plus}).
4966 Substitute the variable part of a matched option. See below.
4967 Note that each comma in the substituted string is replaced by
4971 Substitutes the @code{-S} switch, if that switch was given to GCC@.
4972 If that switch was not specified, this substitutes nothing. Note that
4973 the leading dash is omitted when specifying this option, and it is
4974 automatically inserted if the substitution is performed. Thus the spec
4975 string @samp{%@{foo@}} would match the command-line option @option{-foo}
4976 and would output the command line option @option{-foo}.
4978 @item %W@{@code{S}@}
4979 Like %@{@code{S}@} but mark last argument supplied within as a file to be
4982 @item %@{@code{S}*@}
4983 Substitutes all the switches specified to GCC whose names start
4984 with @code{-S}, but which also take an argument. This is used for
4985 switches like @option{-o}, @option{-D}, @option{-I}, etc.
4986 GCC considers @option{-o foo} as being
4987 one switch whose names starts with @samp{o}. %@{o*@} would substitute this
4988 text, including the space. Thus two arguments would be generated.
4990 @item %@{^@code{S}*@}
4991 Like %@{@code{S}*@}, but don't put a blank between a switch and its
4992 argument. Thus %@{^o*@} would only generate one argument, not two.
4994 @item %@{@code{S}*&@code{T}*@}
4995 Like %@{@code{S}*@}, but preserve order of @code{S} and @code{T} options
4996 (the order of @code{S} and @code{T} in the spec is not significant).
4997 There can be any number of ampersand-separated variables; for each the
4998 wild card is optional. Useful for CPP as @samp{%@{D*&U*&A*@}}.
5000 @item %@{<@code{S}@}
5001 Remove all occurrences of @code{-S} from the command line. Note---this
5002 command is position dependent. @samp{%} commands in the spec string
5003 before this option will see @code{-S}, @samp{%} commands in the spec
5004 string after this option will not.
5006 @item %@{@code{S}*:@code{X}@}
5007 Substitutes @code{X} if one or more switches whose names start with
5008 @code{-S} are specified to GCC@. Note that the tail part of the
5009 @code{-S} option (i.e.@: the part matched by the @samp{*}) will be substituted
5010 for each occurrence of @samp{%*} within @code{X}.
5012 @item %@{@code{S}:@code{X}@}
5013 Substitutes @code{X}, but only if the @samp{-S} switch was given to GCC@.
5015 @item %@{!@code{S}:@code{X}@}
5016 Substitutes @code{X}, but only if the @samp{-S} switch was @emph{not} given to GCC@.
5018 @item %@{|@code{S}:@code{X}@}
5019 Like %@{@code{S}:@code{X}@}, but if no @code{S} switch, substitute @samp{-}.
5021 @item %@{|!@code{S}:@code{X}@}
5022 Like %@{!@code{S}:@code{X}@}, but if there is an @code{S} switch, substitute @samp{-}.
5024 @item %@{.@code{S}:@code{X}@}
5025 Substitutes @code{X}, but only if processing a file with suffix @code{S}.
5027 @item %@{!.@code{S}:@code{X}@}
5028 Substitutes @code{X}, but only if @emph{not} processing a file with suffix @code{S}.
5030 @item %@{@code{S}|@code{P}:@code{X}@}
5031 Substitutes @code{X} if either @code{-S} or @code{-P} was given to GCC@. This may be
5032 combined with @samp{!} and @samp{.} sequences as well, although they
5033 have a stronger binding than the @samp{|}. For example a spec string
5037 %@{.c:-foo@} %@{!.c:-bar@} %@{.c|d:-baz@} %@{!.c|d:-boggle@}
5040 will output the following command-line options from the following input
5041 command-line options:
5046 -d fred.c -foo -baz -boggle
5047 -d jim.d -bar -baz -boggle
5052 The conditional text @code{X} in a %@{@code{S}:@code{X}@} or
5053 %@{!@code{S}:@code{X}@} construct may contain other nested @samp{%} constructs
5054 or spaces, or even newlines. They are processed as usual, as described
5057 The @option{-O}, @option{-f}, @option{-m}, and @option{-W}
5058 switches are handled specifically in these
5059 constructs. If another value of @option{-O} or the negated form of a @option{-f}, @option{-m}, or
5060 @option{-W} switch is found later in the command line, the earlier switch
5061 value is ignored, except with @{@code{S}*@} where @code{S} is just one
5062 letter, which passes all matching options.
5064 The character @samp{|} at the beginning of the predicate text is used to indicate
5065 that a command should be piped to the following command, but only if @option{-pipe}
5068 It is built into GCC which switches take arguments and which do not.
5069 (You might think it would be useful to generalize this to allow each
5070 compiler's spec to say which switches take arguments. But this cannot
5071 be done in a consistent fashion. GCC cannot even decide which input
5072 files have been specified without knowing which switches take arguments,
5073 and it must know which input files to compile in order to tell which
5076 GCC also knows implicitly that arguments starting in @option{-l} are to be
5077 treated as compiler output files, and passed to the linker in their
5078 proper position among the other output files.
5080 @c man begin OPTIONS
5082 @node Target Options
5083 @section Specifying Target Machine and Compiler Version
5084 @cindex target options
5085 @cindex cross compiling
5086 @cindex specifying machine version
5087 @cindex specifying compiler version and target machine
5088 @cindex compiler version, specifying
5089 @cindex target machine, specifying
5091 The usual way to run GCC is to run the executable called @file{gcc}, or
5092 @file{<machine>-gcc} when cross-compiling, or
5093 @file{<machine>-gcc-<version>} to run a version other than the one that
5094 was installed last. Sometimes this is inconvenient, so GCC provides
5095 options that will switch to another cross-compiler or version.
5098 @item -b @var{machine}
5100 The argument @var{machine} specifies the target machine for compilation.
5102 The value to use for @var{machine} is the same as was specified as the
5103 machine type when configuring GCC as a cross-compiler. For
5104 example, if a cross-compiler was configured with @samp{configure
5105 i386v}, meaning to compile for an 80386 running System V, then you
5106 would specify @option{-b i386v} to run that cross compiler.
5108 @item -V @var{version}
5110 The argument @var{version} specifies which version of GCC to run.
5111 This is useful when multiple versions are installed. For example,
5112 @var{version} might be @samp{2.0}, meaning to run GCC version 2.0.
5115 The @option{-V} and @option{-b} options work by running the
5116 @file{<machine>-gcc-<version>} executable, so there's no real reason to
5117 use them if you can just run that directly.
5119 @node Submodel Options
5120 @section Hardware Models and Configurations
5121 @cindex submodel options
5122 @cindex specifying hardware config
5123 @cindex hardware models and configurations, specifying
5124 @cindex machine dependent options
5126 Earlier we discussed the standard option @option{-b} which chooses among
5127 different installed compilers for completely different target
5128 machines, such as VAX vs.@: 68000 vs.@: 80386.
5130 In addition, each of these target machine types can have its own
5131 special options, starting with @samp{-m}, to choose among various
5132 hardware models or configurations---for example, 68010 vs 68020,
5133 floating coprocessor or none. A single installed version of the
5134 compiler can compile for any model or configuration, according to the
5137 Some configurations of the compiler also support additional special
5138 options, usually for compatibility with other compilers on the same
5141 These options are defined by the macro @code{TARGET_SWITCHES} in the
5142 machine description. The default for the options is also defined by
5143 that macro, which enables you to change the defaults.
5157 * RS/6000 and PowerPC Options::
5160 * i386 and x86-64 Options::
5162 * Intel 960 Options::
5163 * DEC Alpha Options::
5164 * DEC Alpha/VMS Options::
5168 * System V Options::
5169 * TMS320C3x/C4x Options::
5177 * S/390 and zSeries Options::
5181 * Xstormy16 Options::
5185 @node M680x0 Options
5186 @subsection M680x0 Options
5187 @cindex M680x0 options
5189 These are the @samp{-m} options defined for the 68000 series. The default
5190 values for these options depends on which style of 68000 was selected when
5191 the compiler was configured; the defaults for the most common choices are
5199 Generate output for a 68000. This is the default
5200 when the compiler is configured for 68000-based systems.
5202 Use this option for microcontrollers with a 68000 or EC000 core,
5203 including the 68008, 68302, 68306, 68307, 68322, 68328 and 68356.
5209 Generate output for a 68020. This is the default
5210 when the compiler is configured for 68020-based systems.
5214 Generate output containing 68881 instructions for floating point.
5215 This is the default for most 68020 systems unless @option{--nfp} was
5216 specified when the compiler was configured.
5220 Generate output for a 68030. This is the default when the compiler is
5221 configured for 68030-based systems.
5225 Generate output for a 68040. This is the default when the compiler is
5226 configured for 68040-based systems.
5228 This option inhibits the use of 68881/68882 instructions that have to be
5229 emulated by software on the 68040. Use this option if your 68040 does not
5230 have code to emulate those instructions.
5234 Generate output for a 68060. This is the default when the compiler is
5235 configured for 68060-based systems.
5237 This option inhibits the use of 68020 and 68881/68882 instructions that
5238 have to be emulated by software on the 68060. Use this option if your 68060
5239 does not have code to emulate those instructions.
5243 Generate output for a CPU32. This is the default
5244 when the compiler is configured for CPU32-based systems.
5246 Use this option for microcontrollers with a
5247 CPU32 or CPU32+ core, including the 68330, 68331, 68332, 68333, 68334,
5248 68336, 68340, 68341, 68349 and 68360.
5252 Generate output for a 520X ``coldfire'' family cpu. This is the default
5253 when the compiler is configured for 520X-based systems.
5255 Use this option for microcontroller with a 5200 core, including
5256 the MCF5202, MCF5203, MCF5204 and MCF5202.
5261 Generate output for a 68040, without using any of the new instructions.
5262 This results in code which can run relatively efficiently on either a
5263 68020/68881 or a 68030 or a 68040. The generated code does use the
5264 68881 instructions that are emulated on the 68040.
5268 Generate output for a 68060, without using any of the new instructions.
5269 This results in code which can run relatively efficiently on either a
5270 68020/68881 or a 68030 or a 68040. The generated code does use the
5271 68881 instructions that are emulated on the 68060.
5275 Generate output containing Sun FPA instructions for floating point.
5278 @opindex msoft-float
5279 Generate output containing library calls for floating point.
5280 @strong{Warning:} the requisite libraries are not available for all m68k
5281 targets. Normally the facilities of the machine's usual C compiler are
5282 used, but this can't be done directly in cross-compilation. You must
5283 make your own arrangements to provide suitable library functions for
5284 cross-compilation. The embedded targets @samp{m68k-*-aout} and
5285 @samp{m68k-*-coff} do provide software floating point support.
5289 Consider type @code{int} to be 16 bits wide, like @code{short int}.
5292 @opindex mnobitfield
5293 Do not use the bit-field instructions. The @option{-m68000}, @option{-mcpu32}
5294 and @option{-m5200} options imply @w{@option{-mnobitfield}}.
5298 Do use the bit-field instructions. The @option{-m68020} option implies
5299 @option{-mbitfield}. This is the default if you use a configuration
5300 designed for a 68020.
5304 Use a different function-calling convention, in which functions
5305 that take a fixed number of arguments return with the @code{rtd}
5306 instruction, which pops their arguments while returning. This
5307 saves one instruction in the caller since there is no need to pop
5308 the arguments there.
5310 This calling convention is incompatible with the one normally
5311 used on Unix, so you cannot use it if you need to call libraries
5312 compiled with the Unix compiler.
5314 Also, you must provide function prototypes for all functions that
5315 take variable numbers of arguments (including @code{printf});
5316 otherwise incorrect code will be generated for calls to those
5319 In addition, seriously incorrect code will result if you call a
5320 function with too many arguments. (Normally, extra arguments are
5321 harmlessly ignored.)
5323 The @code{rtd} instruction is supported by the 68010, 68020, 68030,
5324 68040, 68060 and CPU32 processors, but not by the 68000 or 5200.
5327 @itemx -mno-align-int
5329 @opindex mno-align-int
5330 Control whether GCC aligns @code{int}, @code{long}, @code{long long},
5331 @code{float}, @code{double}, and @code{long double} variables on a 32-bit
5332 boundary (@option{-malign-int}) or a 16-bit boundary (@option{-mno-align-int}).
5333 Aligning variables on 32-bit boundaries produces code that runs somewhat
5334 faster on processors with 32-bit busses at the expense of more memory.
5336 @strong{Warning:} if you use the @option{-malign-int} switch, GCC will
5337 align structures containing the above types differently than
5338 most published application binary interface specifications for the m68k.
5342 Use the pc-relative addressing mode of the 68000 directly, instead of
5343 using a global offset table. At present, this option implies @option{-fpic},
5344 allowing at most a 16-bit offset for pc-relative addressing. @option{-fPIC} is
5345 not presently supported with @option{-mpcrel}, though this could be supported for
5346 68020 and higher processors.
5348 @item -mno-strict-align
5349 @itemx -mstrict-align
5350 @opindex mno-strict-align
5351 @opindex mstrict-align
5352 Do not (do) assume that unaligned memory references will be handled by
5357 @node M68hc1x Options
5358 @subsection M68hc1x Options
5359 @cindex M68hc1x options
5361 These are the @samp{-m} options defined for the 68hc11 and 68hc12
5362 microcontrollers. The default values for these options depends on
5363 which style of microcontroller was selected when the compiler was configured;
5364 the defaults for the most common choices are given below.
5371 Generate output for a 68HC11. This is the default
5372 when the compiler is configured for 68HC11-based systems.
5378 Generate output for a 68HC12. This is the default
5379 when the compiler is configured for 68HC12-based systems.
5382 @opindex mauto-incdec
5383 Enable the use of 68HC12 pre and post auto-increment and auto-decrement
5390 Enable the use of 68HC12 min and max instructions.
5393 @itemx -mno-long-calls
5394 @opindex mlong-calls
5395 @opindex mno-long-calls
5396 Treat all calls as being far away (near). If calls are assumed to be
5397 far away, the compiler will use the @code{call} instruction to
5398 call a function and the @code{rtc} instruction for returning.
5402 Consider type @code{int} to be 16 bits wide, like @code{short int}.
5404 @item -msoft-reg-count=@var{count}
5405 @opindex msoft-reg-count
5406 Specify the number of pseudo-soft registers which are used for the
5407 code generation. The maximum number is 32. Using more pseudo-soft
5408 register may or may not result in better code depending on the program.
5409 The default is 4 for 68HC11 and 2 for 68HC12.
5414 @subsection VAX Options
5417 These @samp{-m} options are defined for the VAX:
5422 Do not output certain jump instructions (@code{aobleq} and so on)
5423 that the Unix assembler for the VAX cannot handle across long
5428 Do output those jump instructions, on the assumption that you
5429 will assemble with the GNU assembler.
5433 Output code for g-format floating point numbers instead of d-format.
5437 @subsection SPARC Options
5438 @cindex SPARC options
5440 These @samp{-m} switches are supported on the SPARC:
5445 @opindex mno-app-regs
5447 Specify @option{-mapp-regs} to generate output using the global registers
5448 2 through 4, which the SPARC SVR4 ABI reserves for applications. This
5451 To be fully SVR4 ABI compliant at the cost of some performance loss,
5452 specify @option{-mno-app-regs}. You should compile libraries and system
5453 software with this option.
5458 @opindex mhard-float
5459 Generate output containing floating point instructions. This is the
5465 @opindex msoft-float
5466 Generate output containing library calls for floating point.
5467 @strong{Warning:} the requisite libraries are not available for all SPARC
5468 targets. Normally the facilities of the machine's usual C compiler are
5469 used, but this cannot be done directly in cross-compilation. You must make
5470 your own arrangements to provide suitable library functions for
5471 cross-compilation. The embedded targets @samp{sparc-*-aout} and
5472 @samp{sparclite-*-*} do provide software floating point support.
5474 @option{-msoft-float} changes the calling convention in the output file;
5475 therefore, it is only useful if you compile @emph{all} of a program with
5476 this option. In particular, you need to compile @file{libgcc.a}, the
5477 library that comes with GCC, with @option{-msoft-float} in order for
5480 @item -mhard-quad-float
5481 @opindex mhard-quad-float
5482 Generate output containing quad-word (long double) floating point
5485 @item -msoft-quad-float
5486 @opindex msoft-quad-float
5487 Generate output containing library calls for quad-word (long double)
5488 floating point instructions. The functions called are those specified
5489 in the SPARC ABI@. This is the default.
5491 As of this writing, there are no sparc implementations that have hardware
5492 support for the quad-word floating point instructions. They all invoke
5493 a trap handler for one of these instructions, and then the trap handler
5494 emulates the effect of the instruction. Because of the trap handler overhead,
5495 this is much slower than calling the ABI library routines. Thus the
5496 @option{-msoft-quad-float} option is the default.
5502 With @option{-mflat}, the compiler does not generate save/restore instructions
5503 and will use a ``flat'' or single register window calling convention.
5504 This model uses %i7 as the frame pointer and is compatible with the normal
5505 register window model. Code from either may be intermixed.
5506 The local registers and the input registers (0--5) are still treated as
5507 ``call saved'' registers and will be saved on the stack as necessary.
5509 With @option{-mno-flat} (the default), the compiler emits save/restore
5510 instructions (except for leaf functions) and is the normal mode of operation.
5512 @item -mno-unaligned-doubles
5513 @itemx -munaligned-doubles
5514 @opindex mno-unaligned-doubles
5515 @opindex munaligned-doubles
5516 Assume that doubles have 8 byte alignment. This is the default.
5518 With @option{-munaligned-doubles}, GCC assumes that doubles have 8 byte
5519 alignment only if they are contained in another type, or if they have an
5520 absolute address. Otherwise, it assumes they have 4 byte alignment.
5521 Specifying this option avoids some rare compatibility problems with code
5522 generated by other compilers. It is not the default because it results
5523 in a performance loss, especially for floating point code.
5525 @item -mno-faster-structs
5526 @itemx -mfaster-structs
5527 @opindex mno-faster-structs
5528 @opindex mfaster-structs
5529 With @option{-mfaster-structs}, the compiler assumes that structures
5530 should have 8 byte alignment. This enables the use of pairs of
5531 @code{ldd} and @code{std} instructions for copies in structure
5532 assignment, in place of twice as many @code{ld} and @code{st} pairs.
5533 However, the use of this changed alignment directly violates the SPARC
5534 ABI@. Thus, it's intended only for use on targets where the developer
5535 acknowledges that their resulting code will not be directly in line with
5536 the rules of the ABI@.
5542 These two options select variations on the SPARC architecture.
5544 By default (unless specifically configured for the Fujitsu SPARClite),
5545 GCC generates code for the v7 variant of the SPARC architecture.
5547 @option{-mv8} will give you SPARC v8 code. The only difference from v7
5548 code is that the compiler emits the integer multiply and integer
5549 divide instructions which exist in SPARC v8 but not in SPARC v7.
5551 @option{-msparclite} will give you SPARClite code. This adds the integer
5552 multiply, integer divide step and scan (@code{ffs}) instructions which
5553 exist in SPARClite but not in SPARC v7.
5555 These options are deprecated and will be deleted in a future GCC release.
5556 They have been replaced with @option{-mcpu=xxx}.
5561 @opindex msupersparc
5562 These two options select the processor for which the code is optimized.
5564 With @option{-mcypress} (the default), the compiler optimizes code for the
5565 Cypress CY7C602 chip, as used in the SPARCStation/SPARCServer 3xx series.
5566 This is also appropriate for the older SPARCStation 1, 2, IPX etc.
5568 With @option{-msupersparc} the compiler optimizes code for the SuperSPARC cpu, as
5569 used in the SPARCStation 10, 1000 and 2000 series. This flag also enables use
5570 of the full SPARC v8 instruction set.
5572 These options are deprecated and will be deleted in a future GCC release.
5573 They have been replaced with @option{-mcpu=xxx}.
5575 @item -mcpu=@var{cpu_type}
5577 Set the instruction set, register set, and instruction scheduling parameters
5578 for machine type @var{cpu_type}. Supported values for @var{cpu_type} are
5579 @samp{v7}, @samp{cypress}, @samp{v8}, @samp{supersparc}, @samp{sparclite},
5580 @samp{hypersparc}, @samp{sparclite86x}, @samp{f930}, @samp{f934},
5581 @samp{sparclet}, @samp{tsc701}, @samp{v9}, and @samp{ultrasparc}.
5583 Default instruction scheduling parameters are used for values that select
5584 an architecture and not an implementation. These are @samp{v7}, @samp{v8},
5585 @samp{sparclite}, @samp{sparclet}, @samp{v9}.
5587 Here is a list of each supported architecture and their supported
5592 v8: supersparc, hypersparc
5593 sparclite: f930, f934, sparclite86x
5598 @item -mtune=@var{cpu_type}
5600 Set the instruction scheduling parameters for machine type
5601 @var{cpu_type}, but do not set the instruction set or register set that the
5602 option @option{-mcpu=@var{cpu_type}} would.
5604 The same values for @option{-mcpu=@var{cpu_type}} can be used for
5605 @option{-mtune=@var{cpu_type}}, but the only useful values are those
5606 that select a particular cpu implementation. Those are @samp{cypress},
5607 @samp{supersparc}, @samp{hypersparc}, @samp{f930}, @samp{f934},
5608 @samp{sparclite86x}, @samp{tsc701}, and @samp{ultrasparc}.
5612 These @samp{-m} switches are supported in addition to the above
5613 on the SPARCLET processor.
5616 @item -mlittle-endian
5617 @opindex mlittle-endian
5618 Generate code for a processor running in little-endian mode.
5622 Treat register @code{%g0} as a normal register.
5623 GCC will continue to clobber it as necessary but will not assume
5624 it always reads as 0.
5626 @item -mbroken-saverestore
5627 @opindex mbroken-saverestore
5628 Generate code that does not use non-trivial forms of the @code{save} and
5629 @code{restore} instructions. Early versions of the SPARCLET processor do
5630 not correctly handle @code{save} and @code{restore} instructions used with
5631 arguments. They correctly handle them used without arguments. A @code{save}
5632 instruction used without arguments increments the current window pointer
5633 but does not allocate a new stack frame. It is assumed that the window
5634 overflow trap handler will properly handle this case as will interrupt
5638 These @samp{-m} switches are supported in addition to the above
5639 on SPARC V9 processors in 64-bit environments.
5642 @item -mlittle-endian
5643 @opindex mlittle-endian
5644 Generate code for a processor running in little-endian mode.
5650 Generate code for a 32-bit or 64-bit environment.
5651 The 32-bit environment sets int, long and pointer to 32 bits.
5652 The 64-bit environment sets int to 32 bits and long and pointer
5655 @item -mcmodel=medlow
5656 @opindex mcmodel=medlow
5657 Generate code for the Medium/Low code model: the program must be linked
5658 in the low 32 bits of the address space. Pointers are 64 bits.
5659 Programs can be statically or dynamically linked.
5661 @item -mcmodel=medmid
5662 @opindex mcmodel=medmid
5663 Generate code for the Medium/Middle code model: the program must be linked
5664 in the low 44 bits of the address space, the text segment must be less than
5665 2G bytes, and data segment must be within 2G of the text segment.
5666 Pointers are 64 bits.
5668 @item -mcmodel=medany
5669 @opindex mcmodel=medany
5670 Generate code for the Medium/Anywhere code model: the program may be linked
5671 anywhere in the address space, the text segment must be less than
5672 2G bytes, and data segment must be within 2G of the text segment.
5673 Pointers are 64 bits.
5675 @item -mcmodel=embmedany
5676 @opindex mcmodel=embmedany
5677 Generate code for the Medium/Anywhere code model for embedded systems:
5678 assume a 32-bit text and a 32-bit data segment, both starting anywhere
5679 (determined at link time). Register %g4 points to the base of the
5680 data segment. Pointers are still 64 bits.
5681 Programs are statically linked, PIC is not supported.
5684 @itemx -mno-stack-bias
5685 @opindex mstack-bias
5686 @opindex mno-stack-bias
5687 With @option{-mstack-bias}, GCC assumes that the stack pointer, and
5688 frame pointer if present, are offset by @minus{}2047 which must be added back
5689 when making stack frame references.
5690 Otherwise, assume no such offset is present.
5693 @node Convex Options
5694 @subsection Convex Options
5695 @cindex Convex options
5697 These @samp{-m} options are defined for Convex:
5702 Generate output for C1. The code will run on any Convex machine.
5703 The preprocessor symbol @code{__convex__c1__} is defined.
5707 Generate output for C2. Uses instructions not available on C1.
5708 Scheduling and other optimizations are chosen for max performance on C2.
5709 The preprocessor symbol @code{__convex_c2__} is defined.
5713 Generate output for C32xx. Uses instructions not available on C1.
5714 Scheduling and other optimizations are chosen for max performance on C32.
5715 The preprocessor symbol @code{__convex_c32__} is defined.
5719 Generate output for C34xx. Uses instructions not available on C1.
5720 Scheduling and other optimizations are chosen for max performance on C34.
5721 The preprocessor symbol @code{__convex_c34__} is defined.
5725 Generate output for C38xx. Uses instructions not available on C1.
5726 Scheduling and other optimizations are chosen for max performance on C38.
5727 The preprocessor symbol @code{__convex_c38__} is defined.
5731 Generate code which puts an argument count in the word preceding each
5732 argument list. This is compatible with regular CC, and a few programs
5733 may need the argument count word. GDB and other source-level debuggers
5734 do not need it; this info is in the symbol table.
5737 @opindex mnoargcount
5738 Omit the argument count word. This is the default.
5740 @item -mvolatile-cache
5741 @opindex mvolatile-cache
5742 Allow volatile references to be cached. This is the default.
5744 @item -mvolatile-nocache
5745 @opindex mvolatile-nocache
5746 Volatile references bypass the data cache, going all the way to memory.
5747 This is only needed for multi-processor code that does not use standard
5748 synchronization instructions. Making non-volatile references to volatile
5749 locations will not necessarily work.
5753 Type long is 32 bits, the same as type int. This is the default.
5757 Type long is 64 bits, the same as type long long. This option is useless,
5758 because no library support exists for it.
5761 @node AMD29K Options
5762 @subsection AMD29K Options
5763 @cindex AMD29K options
5765 These @samp{-m} options are defined for the AMD Am29000:
5770 @cindex DW bit (29k)
5771 Generate code that assumes the @code{DW} bit is set, i.e., that byte and
5772 halfword operations are directly supported by the hardware. This is the
5777 Generate code that assumes the @code{DW} bit is not set.
5781 @cindex byte writes (29k)
5782 Generate code that assumes the system supports byte and halfword write
5783 operations. This is the default.
5787 Generate code that assumes the systems does not support byte and
5788 halfword write operations. @option{-mnbw} implies @option{-mndw}.
5792 @cindex memory model (29k)
5793 Use a small memory model that assumes that all function addresses are
5794 either within a single 256 KB segment or at an absolute address of less
5795 than 256k. This allows the @code{call} instruction to be used instead
5796 of a @code{const}, @code{consth}, @code{calli} sequence.
5800 Use the normal memory model: Generate @code{call} instructions only when
5801 calling functions in the same file and @code{calli} instructions
5802 otherwise. This works if each file occupies less than 256 KB but allows
5803 the entire executable to be larger than 256 KB@. This is the default.
5807 Always use @code{calli} instructions. Specify this option if you expect
5808 a single file to compile into more than 256 KB of code.
5812 @cindex processor selection (29k)
5813 Generate code for the Am29050.
5817 Generate code for the Am29000. This is the default.
5819 @item -mkernel-registers
5820 @opindex mkernel-registers
5821 @cindex kernel and user registers (29k)
5822 Generate references to registers @code{gr64-gr95} instead of to
5823 registers @code{gr96-gr127}. This option can be used when compiling
5824 kernel code that wants a set of global registers disjoint from that used
5827 Note that when this option is used, register names in @samp{-f} flags
5828 must use the normal, user-mode, names.
5830 @item -muser-registers
5831 @opindex muser-registers
5832 Use the normal set of global registers, @code{gr96-gr127}. This is the
5836 @itemx -mno-stack-check
5837 @opindex mstack-check
5838 @opindex mno-stack-check
5839 @cindex stack checks (29k)
5840 Insert (or do not insert) a call to @code{__msp_check} after each stack
5841 adjustment. This is often used for kernel code.
5844 @itemx -mno-storem-bug
5845 @opindex mstorem-bug
5846 @opindex mno-storem-bug
5847 @cindex storem bug (29k)
5848 @option{-mstorem-bug} handles 29k processors which cannot handle the
5849 separation of a mtsrim insn and a storem instruction (most 29000 chips
5850 to date, but not the 29050).
5852 @item -mno-reuse-arg-regs
5853 @itemx -mreuse-arg-regs
5854 @opindex mno-reuse-arg-regs
5855 @opindex mreuse-arg-regs
5856 @option{-mno-reuse-arg-regs} tells the compiler to only use incoming argument
5857 registers for copying out arguments. This helps detect calling a function
5858 with fewer arguments than it was declared with.
5860 @item -mno-impure-text
5861 @itemx -mimpure-text
5862 @opindex mno-impure-text
5863 @opindex mimpure-text
5864 @option{-mimpure-text}, used in addition to @option{-shared}, tells the compiler to
5865 not pass @option{-assert pure-text} to the linker when linking a shared object.
5868 @opindex msoft-float
5869 Generate output containing library calls for floating point.
5870 @strong{Warning:} the requisite libraries are not part of GCC@.
5871 Normally the facilities of the machine's usual C compiler are used, but
5872 this can't be done directly in cross-compilation. You must make your
5873 own arrangements to provide suitable library functions for
5878 Do not generate multm or multmu instructions. This is useful for some embedded
5879 systems which do not have trap handlers for these instructions.
5883 @subsection ARM Options
5886 These @samp{-m} options are defined for Advanced RISC Machines (ARM)
5891 @opindex mapcs-frame
5892 Generate a stack frame that is compliant with the ARM Procedure Call
5893 Standard for all functions, even if this is not strictly necessary for
5894 correct execution of the code. Specifying @option{-fomit-frame-pointer}
5895 with this option will cause the stack frames not to be generated for
5896 leaf functions. The default is @option{-mno-apcs-frame}.
5900 This is a synonym for @option{-mapcs-frame}.
5904 Generate code for a processor running with a 26-bit program counter,
5905 and conforming to the function calling standards for the APCS 26-bit
5906 option. This option replaces the @option{-m2} and @option{-m3} options
5907 of previous releases of the compiler.
5911 Generate code for a processor running with a 32-bit program counter,
5912 and conforming to the function calling standards for the APCS 32-bit
5913 option. This option replaces the @option{-m6} option of previous releases
5917 @c not currently implemented
5918 @item -mapcs-stack-check
5919 @opindex mapcs-stack-check
5920 Generate code to check the amount of stack space available upon entry to
5921 every function (that actually uses some stack space). If there is
5922 insufficient space available then either the function
5923 @samp{__rt_stkovf_split_small} or @samp{__rt_stkovf_split_big} will be
5924 called, depending upon the amount of stack space required. The run time
5925 system is required to provide these functions. The default is
5926 @option{-mno-apcs-stack-check}, since this produces smaller code.
5928 @c not currently implemented
5930 @opindex mapcs-float
5931 Pass floating point arguments using the float point registers. This is
5932 one of the variants of the APCS@. This option is recommended if the
5933 target hardware has a floating point unit or if a lot of floating point
5934 arithmetic is going to be performed by the code. The default is
5935 @option{-mno-apcs-float}, since integer only code is slightly increased in
5936 size if @option{-mapcs-float} is used.
5938 @c not currently implemented
5939 @item -mapcs-reentrant
5940 @opindex mapcs-reentrant
5941 Generate reentrant, position independent code. The default is
5942 @option{-mno-apcs-reentrant}.
5945 @item -mthumb-interwork
5946 @opindex mthumb-interwork
5947 Generate code which supports calling between the ARM and Thumb
5948 instruction sets. Without this option the two instruction sets cannot
5949 be reliably used inside one program. The default is
5950 @option{-mno-thumb-interwork}, since slightly larger code is generated
5951 when @option{-mthumb-interwork} is specified.
5953 @item -mno-sched-prolog
5954 @opindex mno-sched-prolog
5955 Prevent the reordering of instructions in the function prolog, or the
5956 merging of those instruction with the instructions in the function's
5957 body. This means that all functions will start with a recognizable set
5958 of instructions (or in fact one of a choice from a small set of
5959 different function prologues), and this information can be used to
5960 locate the start if functions inside an executable piece of code. The
5961 default is @option{-msched-prolog}.
5964 @opindex mhard-float
5965 Generate output containing floating point instructions. This is the
5969 @opindex msoft-float
5970 Generate output containing library calls for floating point.
5971 @strong{Warning:} the requisite libraries are not available for all ARM
5972 targets. Normally the facilities of the machine's usual C compiler are
5973 used, but this cannot be done directly in cross-compilation. You must make
5974 your own arrangements to provide suitable library functions for
5977 @option{-msoft-float} changes the calling convention in the output file;
5978 therefore, it is only useful if you compile @emph{all} of a program with
5979 this option. In particular, you need to compile @file{libgcc.a}, the
5980 library that comes with GCC, with @option{-msoft-float} in order for
5983 @item -mlittle-endian
5984 @opindex mlittle-endian
5985 Generate code for a processor running in little-endian mode. This is
5986 the default for all standard configurations.
5989 @opindex mbig-endian
5990 Generate code for a processor running in big-endian mode; the default is
5991 to compile code for a little-endian processor.
5993 @item -mwords-little-endian
5994 @opindex mwords-little-endian
5995 This option only applies when generating code for big-endian processors.
5996 Generate code for a little-endian word order but a big-endian byte
5997 order. That is, a byte order of the form @samp{32107654}. Note: this
5998 option should only be used if you require compatibility with code for
5999 big-endian ARM processors generated by versions of the compiler prior to
6002 @item -malignment-traps
6003 @opindex malignment-traps
6004 Generate code that will not trap if the MMU has alignment traps enabled.
6005 On ARM architectures prior to ARMv4, there were no instructions to
6006 access half-word objects stored in memory. However, when reading from
6007 memory a feature of the ARM architecture allows a word load to be used,
6008 even if the address is unaligned, and the processor core will rotate the
6009 data as it is being loaded. This option tells the compiler that such
6010 misaligned accesses will cause a MMU trap and that it should instead
6011 synthesize the access as a series of byte accesses. The compiler can
6012 still use word accesses to load half-word data if it knows that the
6013 address is aligned to a word boundary.
6015 This option is ignored when compiling for ARM architecture 4 or later,
6016 since these processors have instructions to directly access half-word
6019 @item -mno-alignment-traps
6020 @opindex mno-alignment-traps
6021 Generate code that assumes that the MMU will not trap unaligned
6022 accesses. This produces better code when the target instruction set
6023 does not have half-word memory operations (i.e.@: implementations prior to
6026 Note that you cannot use this option to access unaligned word objects,
6027 since the processor will only fetch one 32-bit aligned object from
6030 The default setting for most targets is @option{-mno-alignment-traps}, since
6031 this produces better code when there are no half-word memory
6032 instructions available.
6034 @item -mshort-load-bytes
6035 @itemx -mno-short-load-words
6036 @opindex mshort-load-bytes
6037 @opindex mno-short-load-words
6038 These are deprecated aliases for @option{-malignment-traps}.
6040 @item -mno-short-load-bytes
6041 @itemx -mshort-load-words
6042 @opindex mno-short-load-bytes
6043 @opindex mshort-load-words
6044 This are deprecated aliases for @option{-mno-alignment-traps}.
6048 This option only applies to RISC iX@. Emulate the native BSD-mode
6049 compiler. This is the default if @option{-ansi} is not specified.
6053 This option only applies to RISC iX@. Emulate the native X/Open-mode
6056 @item -mno-symrename
6057 @opindex mno-symrename
6058 This option only applies to RISC iX@. Do not run the assembler
6059 post-processor, @samp{symrename}, after code has been assembled.
6060 Normally it is necessary to modify some of the standard symbols in
6061 preparation for linking with the RISC iX C library; this option
6062 suppresses this pass. The post-processor is never run when the
6063 compiler is built for cross-compilation.
6065 @item -mcpu=@var{name}
6067 This specifies the name of the target ARM processor. GCC uses this name
6068 to determine what kind of instructions it can emit when generating
6069 assembly code. Permissible names are: @samp{arm2}, @samp{arm250},
6070 @samp{arm3}, @samp{arm6}, @samp{arm60}, @samp{arm600}, @samp{arm610},
6071 @samp{arm620}, @samp{arm7}, @samp{arm7m}, @samp{arm7d}, @samp{arm7dm},
6072 @samp{arm7di}, @samp{arm7dmi}, @samp{arm70}, @samp{arm700},
6073 @samp{arm700i}, @samp{arm710}, @samp{arm710c}, @samp{arm7100},
6074 @samp{arm7500}, @samp{arm7500fe}, @samp{arm7tdmi}, @samp{arm8},
6075 @samp{strongarm}, @samp{strongarm110}, @samp{strongarm1100},
6076 @samp{arm8}, @samp{arm810}, @samp{arm9}, @samp{arm9e}, @samp{arm920},
6077 @samp{arm920t}, @samp{arm940t}, @samp{arm9tdmi}, @samp{arm10tdmi},
6078 @samp{arm1020t}, @samp{xscale}.
6080 @itemx -mtune=@var{name}
6082 This option is very similar to the @option{-mcpu=} option, except that
6083 instead of specifying the actual target processor type, and hence
6084 restricting which instructions can be used, it specifies that GCC should
6085 tune the performance of the code as if the target were of the type
6086 specified in this option, but still choosing the instructions that it
6087 will generate based on the cpu specified by a @option{-mcpu=} option.
6088 For some ARM implementations better performance can be obtained by using
6091 @item -march=@var{name}
6093 This specifies the name of the target ARM architecture. GCC uses this
6094 name to determine what kind of instructions it can emit when generating
6095 assembly code. This option can be used in conjunction with or instead
6096 of the @option{-mcpu=} option. Permissible names are: @samp{armv2},
6097 @samp{armv2a}, @samp{armv3}, @samp{armv3m}, @samp{armv4}, @samp{armv4t},
6098 @samp{armv5}, @samp{armv5t}, @samp{armv5te}.
6100 @item -mfpe=@var{number}
6101 @itemx -mfp=@var{number}
6104 This specifies the version of the floating point emulation available on
6105 the target. Permissible values are 2 and 3. @option{-mfp=} is a synonym
6106 for @option{-mfpe=}, for compatibility with older versions of GCC@.
6108 @item -mstructure-size-boundary=@var{n}
6109 @opindex mstructure-size-boundary
6110 The size of all structures and unions will be rounded up to a multiple
6111 of the number of bits set by this option. Permissible values are 8 and
6112 32. The default value varies for different toolchains. For the COFF
6113 targeted toolchain the default value is 8. Specifying the larger number
6114 can produce faster, more efficient code, but can also increase the size
6115 of the program. The two values are potentially incompatible. Code
6116 compiled with one value cannot necessarily expect to work with code or
6117 libraries compiled with the other value, if they exchange information
6118 using structures or unions.
6120 @item -mabort-on-noreturn
6121 @opindex mabort-on-noreturn
6122 Generate a call to the function @code{abort} at the end of a
6123 @code{noreturn} function. It will be executed if the function tries to
6127 @itemx -mno-long-calls
6128 @opindex mlong-calls
6129 @opindex mno-long-calls
6130 Tells the compiler to perform function calls by first loading the
6131 address of the function into a register and then performing a subroutine
6132 call on this register. This switch is needed if the target function
6133 will lie outside of the 64 megabyte addressing range of the offset based
6134 version of subroutine call instruction.
6136 Even if this switch is enabled, not all function calls will be turned
6137 into long calls. The heuristic is that static functions, functions
6138 which have the @samp{short-call} attribute, functions that are inside
6139 the scope of a @samp{#pragma no_long_calls} directive and functions whose
6140 definitions have already been compiled within the current compilation
6141 unit, will not be turned into long calls. The exception to this rule is
6142 that weak function definitions, functions with the @samp{long-call}
6143 attribute or the @samp{section} attribute, and functions that are within
6144 the scope of a @samp{#pragma long_calls} directive, will always be
6145 turned into long calls.
6147 This feature is not enabled by default. Specifying
6148 @option{-mno-long-calls} will restore the default behavior, as will
6149 placing the function calls within the scope of a @samp{#pragma
6150 long_calls_off} directive. Note these switches have no effect on how
6151 the compiler generates code to handle function calls via function
6154 @item -mnop-fun-dllimport
6155 @opindex mnop-fun-dllimport
6156 Disable support for the @code{dllimport} attribute.
6158 @item -msingle-pic-base
6159 @opindex msingle-pic-base
6160 Treat the register used for PIC addressing as read-only, rather than
6161 loading it in the prologue for each function. The run-time system is
6162 responsible for initializing this register with an appropriate value
6163 before execution begins.
6165 @item -mpic-register=@var{reg}
6166 @opindex mpic-register
6167 Specify the register to be used for PIC addressing. The default is R10
6168 unless stack-checking is enabled, when R9 is used.
6170 @item -mpoke-function-name
6171 @opindex mpoke-function-name
6172 Write the name of each function into the text section, directly
6173 preceding the function prologue. The generated code is similar to this:
6177 .ascii "arm_poke_function_name", 0
6180 .word 0xff000000 + (t1 - t0)
6181 arm_poke_function_name
6183 stmfd sp!, @{fp, ip, lr, pc@}
6187 When performing a stack backtrace, code can inspect the value of
6188 @code{pc} stored at @code{fp + 0}. If the trace function then looks at
6189 location @code{pc - 12} and the top 8 bits are set, then we know that
6190 there is a function name embedded immediately preceding this location
6191 and has length @code{((pc[-3]) & 0xff000000)}.
6195 Generate code for the 16-bit Thumb instruction set. The default is to
6196 use the 32-bit ARM instruction set.
6199 @opindex mtpcs-frame
6200 Generate a stack frame that is compliant with the Thumb Procedure Call
6201 Standard for all non-leaf functions. (A leaf function is one that does
6202 not call any other functions.) The default is @option{-mno-tpcs-frame}.
6204 @item -mtpcs-leaf-frame
6205 @opindex mtpcs-leaf-frame
6206 Generate a stack frame that is compliant with the Thumb Procedure Call
6207 Standard for all leaf functions. (A leaf function is one that does
6208 not call any other functions.) The default is @option{-mno-apcs-leaf-frame}.
6210 @item -mcallee-super-interworking
6211 @opindex mcallee-super-interworking
6212 Gives all externally visible functions in the file being compiled an ARM
6213 instruction set header which switches to Thumb mode before executing the
6214 rest of the function. This allows these functions to be called from
6215 non-interworking code.
6217 @item -mcaller-super-interworking
6218 @opindex mcaller-super-interworking
6219 Allows calls via function pointers (including virtual functions) to
6220 execute correctly regardless of whether the target code has been
6221 compiled for interworking or not. There is a small overhead in the cost
6222 of executing a function pointer if this option is enabled.
6226 @node MN10200 Options
6227 @subsection MN10200 Options
6228 @cindex MN10200 options
6229 These @option{-m} options are defined for Matsushita MN10200 architectures:
6234 Indicate to the linker that it should perform a relaxation optimization pass
6235 to shorten branches, calls and absolute memory addresses. This option only
6236 has an effect when used on the command line for the final link step.
6238 This option makes symbolic debugging impossible.
6241 @node MN10300 Options
6242 @subsection MN10300 Options
6243 @cindex MN10300 options
6244 These @option{-m} options are defined for Matsushita MN10300 architectures:
6249 Generate code to avoid bugs in the multiply instructions for the MN10300
6250 processors. This is the default.
6253 @opindex mno-mult-bug
6254 Do not generate code to avoid bugs in the multiply instructions for the
6259 Generate code which uses features specific to the AM33 processor.
6263 Do not generate code which uses features specific to the AM33 processor. This
6268 Do not link in the C run-time initialization object file.
6272 Indicate to the linker that it should perform a relaxation optimization pass
6273 to shorten branches, calls and absolute memory addresses. This option only
6274 has an effect when used on the command line for the final link step.
6276 This option makes symbolic debugging impossible.
6280 @node M32R/D Options
6281 @subsection M32R/D Options
6282 @cindex M32R/D options
6284 These @option{-m} options are defined for Mitsubishi M32R/D architectures:
6289 Generate code for the M32R/X@.
6293 Generate code for the M32R@. This is the default.
6295 @item -mcode-model=small
6296 @opindex mcode-model=small
6297 Assume all objects live in the lower 16MB of memory (so that their addresses
6298 can be loaded with the @code{ld24} instruction), and assume all subroutines
6299 are reachable with the @code{bl} instruction.
6300 This is the default.
6302 The addressability of a particular object can be set with the
6303 @code{model} attribute.
6305 @item -mcode-model=medium
6306 @opindex mcode-model=medium
6307 Assume objects may be anywhere in the 32-bit address space (the compiler
6308 will generate @code{seth/add3} instructions to load their addresses), and
6309 assume all subroutines are reachable with the @code{bl} instruction.
6311 @item -mcode-model=large
6312 @opindex mcode-model=large
6313 Assume objects may be anywhere in the 32-bit address space (the compiler
6314 will generate @code{seth/add3} instructions to load their addresses), and
6315 assume subroutines may not be reachable with the @code{bl} instruction
6316 (the compiler will generate the much slower @code{seth/add3/jl}
6317 instruction sequence).
6320 @opindex msdata=none
6321 Disable use of the small data area. Variables will be put into
6322 one of @samp{.data}, @samp{bss}, or @samp{.rodata} (unless the
6323 @code{section} attribute has been specified).
6324 This is the default.
6326 The small data area consists of sections @samp{.sdata} and @samp{.sbss}.
6327 Objects may be explicitly put in the small data area with the
6328 @code{section} attribute using one of these sections.
6331 @opindex msdata=sdata
6332 Put small global and static data in the small data area, but do not
6333 generate special code to reference them.
6337 Put small global and static data in the small data area, and generate
6338 special instructions to reference them.
6342 @cindex smaller data references
6343 Put global and static objects less than or equal to @var{num} bytes
6344 into the small data or bss sections instead of the normal data or bss
6345 sections. The default value of @var{num} is 8.
6346 The @option{-msdata} option must be set to one of @samp{sdata} or @samp{use}
6347 for this option to have any effect.
6349 All modules should be compiled with the same @option{-G @var{num}} value.
6350 Compiling with different values of @var{num} may or may not work; if it
6351 doesn't the linker will give an error message---incorrect code will not be
6357 @subsection M88K Options
6358 @cindex M88k options
6360 These @samp{-m} options are defined for Motorola 88k architectures:
6365 Generate code that works well on both the m88100 and the
6370 Generate code that works best for the m88100, but that also
6375 Generate code that works best for the m88110, and may not run
6380 Obsolete option to be removed from the next revision.
6383 @item -midentify-revision
6384 @opindex midentify-revision
6385 @cindex identifying source, compiler (88k)
6386 Include an @code{ident} directive in the assembler output recording the
6387 source file name, compiler name and version, timestamp, and compilation
6390 @item -mno-underscores
6391 @opindex mno-underscores
6392 @cindex underscores, avoiding (88k)
6393 In assembler output, emit symbol names without adding an underscore
6394 character at the beginning of each name. The default is to use an
6395 underscore as prefix on each name.
6397 @item -mocs-debug-info
6398 @itemx -mno-ocs-debug-info
6399 @opindex mocs-debug-info
6400 @opindex mno-ocs-debug-info
6402 @cindex debugging, 88k OCS
6403 Include (or omit) additional debugging information (about registers used
6404 in each stack frame) as specified in the 88open Object Compatibility
6405 Standard, ``OCS''@. This extra information allows debugging of code that
6406 has had the frame pointer eliminated. The default for DG/UX, SVr4, and
6407 Delta 88 SVr3.2 is to include this information; other 88k configurations
6408 omit this information by default.
6410 @item -mocs-frame-position
6411 @opindex mocs-frame-position
6412 @cindex register positions in frame (88k)
6413 When emitting COFF debugging information for automatic variables and
6414 parameters stored on the stack, use the offset from the canonical frame
6415 address, which is the stack pointer (register 31) on entry to the
6416 function. The DG/UX, SVr4, Delta88 SVr3.2, and BCS configurations use
6417 @option{-mocs-frame-position}; other 88k configurations have the default
6418 @option{-mno-ocs-frame-position}.
6420 @item -mno-ocs-frame-position
6421 @opindex mno-ocs-frame-position
6422 @cindex register positions in frame (88k)
6423 When emitting COFF debugging information for automatic variables and
6424 parameters stored on the stack, use the offset from the frame pointer
6425 register (register 30). When this option is in effect, the frame
6426 pointer is not eliminated when debugging information is selected by the
6429 @item -moptimize-arg-area
6430 @opindex moptimize-arg-area
6431 @cindex arguments in frame (88k)
6432 Save space by reorganizing the stack frame. This option generates code
6433 that does not agree with the 88open specifications, but uses less
6436 @itemx -mno-optimize-arg-area
6437 @opindex mno-optimize-arg-area
6438 Do not reorganize the stack frame to save space. This is the default.
6439 The generated conforms to the specification, but uses more memory.
6441 @item -mshort-data-@var{num}
6442 @opindex mshort-data
6443 @cindex smaller data references (88k)
6444 @cindex r0-relative references (88k)
6445 Generate smaller data references by making them relative to @code{r0},
6446 which allows loading a value using a single instruction (rather than the
6447 usual two). You control which data references are affected by
6448 specifying @var{num} with this option. For example, if you specify
6449 @option{-mshort-data-512}, then the data references affected are those
6450 involving displacements of less than 512 bytes.
6451 @option{-mshort-data-@var{num}} is not effective for @var{num} greater
6454 @item -mserialize-volatile
6455 @opindex mserialize-volatile
6456 @itemx -mno-serialize-volatile
6457 @opindex mno-serialize-volatile
6458 @cindex sequential consistency on 88k
6459 Do, or don't, generate code to guarantee sequential consistency
6460 of volatile memory references. By default, consistency is
6463 The order of memory references made by the MC88110 processor does
6464 not always match the order of the instructions requesting those
6465 references. In particular, a load instruction may execute before
6466 a preceding store instruction. Such reordering violates
6467 sequential consistency of volatile memory references, when there
6468 are multiple processors. When consistency must be guaranteed,
6469 GCC generates special instructions, as needed, to force
6470 execution in the proper order.
6472 The MC88100 processor does not reorder memory references and so
6473 always provides sequential consistency. However, by default, GCC
6474 generates the special instructions to guarantee consistency
6475 even when you use @option{-m88100}, so that the code may be run on an
6476 MC88110 processor. If you intend to run your code only on the
6477 MC88100 processor, you may use @option{-mno-serialize-volatile}.
6479 The extra code generated to guarantee consistency may affect the
6480 performance of your application. If you know that you can safely
6481 forgo this guarantee, you may use @option{-mno-serialize-volatile}.
6487 @cindex assembler syntax, 88k
6489 Turn on (@option{-msvr4}) or off (@option{-msvr3}) compiler extensions
6490 related to System V release 4 (SVr4). This controls the following:
6494 Which variant of the assembler syntax to emit.
6496 @option{-msvr4} makes the C preprocessor recognize @samp{#pragma weak}
6497 that is used on System V release 4.
6499 @option{-msvr4} makes GCC issue additional declaration directives used in
6503 @option{-msvr4} is the default for the m88k-motorola-sysv4 and
6504 m88k-dg-dgux m88k configurations. @option{-msvr3} is the default for all
6505 other m88k configurations.
6507 @item -mversion-03.00
6508 @opindex mversion-03.00
6509 This option is obsolete, and is ignored.
6510 @c ??? which asm syntax better for GAS? option there too?
6512 @item -mno-check-zero-division
6513 @itemx -mcheck-zero-division
6514 @opindex mno-check-zero-division
6515 @opindex mcheck-zero-division
6516 @cindex zero division on 88k
6517 Do, or don't, generate code to guarantee that integer division by
6518 zero will be detected. By default, detection is guaranteed.
6520 Some models of the MC88100 processor fail to trap upon integer
6521 division by zero under certain conditions. By default, when
6522 compiling code that might be run on such a processor, GCC
6523 generates code that explicitly checks for zero-valued divisors
6524 and traps with exception number 503 when one is detected. Use of
6525 @option{-mno-check-zero-division} suppresses such checking for code
6526 generated to run on an MC88100 processor.
6528 GCC assumes that the MC88110 processor correctly detects all instances
6529 of integer division by zero. When @option{-m88110} is specified, no
6530 explicit checks for zero-valued divisors are generated, and both
6531 @option{-mcheck-zero-division} and @option{-mno-check-zero-division} are
6534 @item -muse-div-instruction
6535 @opindex muse-div-instruction
6536 @cindex divide instruction, 88k
6537 Use the div instruction for signed integer division on the
6538 MC88100 processor. By default, the div instruction is not used.
6540 On the MC88100 processor the signed integer division instruction
6541 div) traps to the operating system on a negative operand. The
6542 operating system transparently completes the operation, but at a
6543 large cost in execution time. By default, when compiling code
6544 that might be run on an MC88100 processor, GCC emulates signed
6545 integer division using the unsigned integer division instruction
6546 divu), thereby avoiding the large penalty of a trap to the
6547 operating system. Such emulation has its own, smaller, execution
6548 cost in both time and space. To the extent that your code's
6549 important signed integer division operations are performed on two
6550 nonnegative operands, it may be desirable to use the div
6551 instruction directly.
6553 On the MC88110 processor the div instruction (also known as the
6554 divs instruction) processes negative operands without trapping to
6555 the operating system. When @option{-m88110} is specified,
6556 @option{-muse-div-instruction} is ignored, and the div instruction is used
6557 for signed integer division.
6559 Note that the result of dividing @code{INT_MIN} by @minus{}1 is undefined. In
6560 particular, the behavior of such a division with and without
6561 @option{-muse-div-instruction} may differ.
6563 @item -mtrap-large-shift
6564 @itemx -mhandle-large-shift
6565 @opindex mtrap-large-shift
6566 @opindex mhandle-large-shift
6567 @cindex bit shift overflow (88k)
6568 @cindex large bit shifts (88k)
6569 Include code to detect bit-shifts of more than 31 bits; respectively,
6570 trap such shifts or emit code to handle them properly. By default GCC
6571 makes no special provision for large bit shifts.
6573 @item -mwarn-passed-structs
6574 @opindex mwarn-passed-structs
6575 @cindex structure passing (88k)
6576 Warn when a function passes a struct as an argument or result.
6577 Structure-passing conventions have changed during the evolution of the C
6578 language, and are often the source of portability problems. By default,
6579 GCC issues no such warning.
6582 @c break page here to avoid unsightly interparagraph stretch.
6586 @node RS/6000 and PowerPC Options
6587 @subsection IBM RS/6000 and PowerPC Options
6588 @cindex RS/6000 and PowerPC Options
6589 @cindex IBM RS/6000 and PowerPC Options
6591 These @samp{-m} options are defined for the IBM RS/6000 and PowerPC:
6599 @itemx -mpowerpc-gpopt
6600 @itemx -mno-powerpc-gpopt
6601 @itemx -mpowerpc-gfxopt
6602 @itemx -mno-powerpc-gfxopt
6604 @itemx -mno-powerpc64
6610 @opindex mno-powerpc
6611 @opindex mpowerpc-gpopt
6612 @opindex mno-powerpc-gpopt
6613 @opindex mpowerpc-gfxopt
6614 @opindex mno-powerpc-gfxopt
6616 @opindex mno-powerpc64
6617 GCC supports two related instruction set architectures for the
6618 RS/6000 and PowerPC@. The @dfn{POWER} instruction set are those
6619 instructions supported by the @samp{rios} chip set used in the original
6620 RS/6000 systems and the @dfn{PowerPC} instruction set is the
6621 architecture of the Motorola MPC5xx, MPC6xx, MPC8xx microprocessors, and
6622 the IBM 4xx microprocessors.
6624 Neither architecture is a subset of the other. However there is a
6625 large common subset of instructions supported by both. An MQ
6626 register is included in processors supporting the POWER architecture.
6628 You use these options to specify which instructions are available on the
6629 processor you are using. The default value of these options is
6630 determined when configuring GCC@. Specifying the
6631 @option{-mcpu=@var{cpu_type}} overrides the specification of these
6632 options. We recommend you use the @option{-mcpu=@var{cpu_type}} option
6633 rather than the options listed above.
6635 The @option{-mpower} option allows GCC to generate instructions that
6636 are found only in the POWER architecture and to use the MQ register.
6637 Specifying @option{-mpower2} implies @option{-power} and also allows GCC
6638 to generate instructions that are present in the POWER2 architecture but
6639 not the original POWER architecture.
6641 The @option{-mpowerpc} option allows GCC to generate instructions that
6642 are found only in the 32-bit subset of the PowerPC architecture.
6643 Specifying @option{-mpowerpc-gpopt} implies @option{-mpowerpc} and also allows
6644 GCC to use the optional PowerPC architecture instructions in the
6645 General Purpose group, including floating-point square root. Specifying
6646 @option{-mpowerpc-gfxopt} implies @option{-mpowerpc} and also allows GCC to
6647 use the optional PowerPC architecture instructions in the Graphics
6648 group, including floating-point select.
6650 The @option{-mpowerpc64} option allows GCC to generate the additional
6651 64-bit instructions that are found in the full PowerPC64 architecture
6652 and to treat GPRs as 64-bit, doubleword quantities. GCC defaults to
6653 @option{-mno-powerpc64}.
6655 If you specify both @option{-mno-power} and @option{-mno-powerpc}, GCC
6656 will use only the instructions in the common subset of both
6657 architectures plus some special AIX common-mode calls, and will not use
6658 the MQ register. Specifying both @option{-mpower} and @option{-mpowerpc}
6659 permits GCC to use any instruction from either architecture and to
6660 allow use of the MQ register; specify this for the Motorola MPC601.
6662 @item -mnew-mnemonics
6663 @itemx -mold-mnemonics
6664 @opindex mnew-mnemonics
6665 @opindex mold-mnemonics
6666 Select which mnemonics to use in the generated assembler code. With
6667 @option{-mnew-mnemonics}, GCC uses the assembler mnemonics defined for
6668 the PowerPC architecture. With @option{-mold-mnemonics} it uses the
6669 assembler mnemonics defined for the POWER architecture. Instructions
6670 defined in only one architecture have only one mnemonic; GCC uses that
6671 mnemonic irrespective of which of these options is specified.
6673 GCC defaults to the mnemonics appropriate for the architecture in
6674 use. Specifying @option{-mcpu=@var{cpu_type}} sometimes overrides the
6675 value of these option. Unless you are building a cross-compiler, you
6676 should normally not specify either @option{-mnew-mnemonics} or
6677 @option{-mold-mnemonics}, but should instead accept the default.
6679 @item -mcpu=@var{cpu_type}
6681 Set architecture type, register usage, choice of mnemonics, and
6682 instruction scheduling parameters for machine type @var{cpu_type}.
6683 Supported values for @var{cpu_type} are @samp{rios}, @samp{rios1},
6684 @samp{rsc}, @samp{rios2}, @samp{rs64a}, @samp{601}, @samp{602},
6685 @samp{603}, @samp{603e}, @samp{604}, @samp{604e}, @samp{620},
6686 @samp{630}, @samp{740}, @samp{7400}, @samp{7450}, @samp{750},
6687 @samp{power}, @samp{power2}, @samp{powerpc}, @samp{403}, @samp{505},
6688 @samp{801}, @samp{821}, @samp{823}, and @samp{860} and @samp{common}.
6690 @option{-mcpu=common} selects a completely generic processor. Code
6691 generated under this option will run on any POWER or PowerPC processor.
6692 GCC will use only the instructions in the common subset of both
6693 architectures, and will not use the MQ register. GCC assumes a generic
6694 processor model for scheduling purposes.
6696 @option{-mcpu=power}, @option{-mcpu=power2}, @option{-mcpu=powerpc}, and
6697 @option{-mcpu=powerpc64} specify generic POWER, POWER2, pure 32-bit
6698 PowerPC (i.e., not MPC601), and 64-bit PowerPC architecture machine
6699 types, with an appropriate, generic processor model assumed for
6700 scheduling purposes.
6702 The other options specify a specific processor. Code generated under
6703 those options will run best on that processor, and may not run at all on
6706 The @option{-mcpu} options automatically enable or disable other
6707 @option{-m} options as follows:
6711 @option{-mno-power}, @option{-mno-powerc}
6718 @option{-mpower}, @option{-mno-powerpc}, @option{-mno-new-mnemonics}
6733 @option{-mno-power}, @option{-mpowerpc}, @option{-mnew-mnemonics}
6736 @option{-mpower}, @option{-mpowerpc}, @option{-mnew-mnemonics}
6741 @option{-mno-power}, @option{-mpowerpc}, @option{-mnew-mnemonics}, @option{-msoft-float}
6744 @item -mtune=@var{cpu_type}
6746 Set the instruction scheduling parameters for machine type
6747 @var{cpu_type}, but do not set the architecture type, register usage, or
6748 choice of mnemonics, as @option{-mcpu=@var{cpu_type}} would. The same
6749 values for @var{cpu_type} are used for @option{-mtune} as for
6750 @option{-mcpu}. If both are specified, the code generated will use the
6751 architecture, registers, and mnemonics set by @option{-mcpu}, but the
6752 scheduling parameters set by @option{-mtune}.
6757 @opindex mno-altivec
6758 These switches enable or disable the use of built-in functions that
6759 allow access to the AltiVec instruction set. You may also need to set
6760 @option{-mabi=altivec} to adjust the current ABI with AltiVec ABI
6765 Extend the current ABI with SPE ABI extensions. This does not change
6766 the default ABI, instead it adds the SPE ABI extensions to the current
6770 @opindex mabi=no-spe
6771 Disable Booke SPE ABI extensions for the current ABI.
6773 @item -misel=@var{yes/no}
6776 This switch enables or disables the generation of ISEL instructions.
6779 @itemx -mno-fp-in-toc
6780 @itemx -mno-sum-in-toc
6781 @itemx -mminimal-toc
6783 @opindex mno-fp-in-toc
6784 @opindex mno-sum-in-toc
6785 @opindex mminimal-toc
6786 Modify generation of the TOC (Table Of Contents), which is created for
6787 every executable file. The @option{-mfull-toc} option is selected by
6788 default. In that case, GCC will allocate at least one TOC entry for
6789 each unique non-automatic variable reference in your program. GCC
6790 will also place floating-point constants in the TOC@. However, only
6791 16,384 entries are available in the TOC@.
6793 If you receive a linker error message that saying you have overflowed
6794 the available TOC space, you can reduce the amount of TOC space used
6795 with the @option{-mno-fp-in-toc} and @option{-mno-sum-in-toc} options.
6796 @option{-mno-fp-in-toc} prevents GCC from putting floating-point
6797 constants in the TOC and @option{-mno-sum-in-toc} forces GCC to
6798 generate code to calculate the sum of an address and a constant at
6799 run-time instead of putting that sum into the TOC@. You may specify one
6800 or both of these options. Each causes GCC to produce very slightly
6801 slower and larger code at the expense of conserving TOC space.
6803 If you still run out of space in the TOC even when you specify both of
6804 these options, specify @option{-mminimal-toc} instead. This option causes
6805 GCC to make only one TOC entry for every file. When you specify this
6806 option, GCC will produce code that is slower and larger but which
6807 uses extremely little TOC space. You may wish to use this option
6808 only on files that contain less frequently executed code.
6814 Enable 64-bit AIX ABI and calling convention: 64-bit pointers, 64-bit
6815 @code{long} type, and the infrastructure needed to support them.
6816 Specifying @option{-maix64} implies @option{-mpowerpc64} and
6817 @option{-mpowerpc}, while @option{-maix32} disables the 64-bit ABI and
6818 implies @option{-mno-powerpc64}. GCC defaults to @option{-maix32}.
6823 @opindex mno-xl-call
6824 On AIX, pass floating-point arguments to prototyped functions beyond the
6825 register save area (RSA) on the stack in addition to argument FPRs. The
6826 AIX calling convention was extended but not initially documented to
6827 handle an obscure K&R C case of calling a function that takes the
6828 address of its arguments with fewer arguments than declared. AIX XL
6829 compilers access floating point arguments which do not fit in the
6830 RSA from the stack when a subroutine is compiled without
6831 optimization. Because always storing floating-point arguments on the
6832 stack is inefficient and rarely needed, this option is not enabled by
6833 default and only is necessary when calling subroutines compiled by AIX
6834 XL compilers without optimization.
6838 Support @dfn{IBM RS/6000 SP} @dfn{Parallel Environment} (PE)@. Link an
6839 application written to use message passing with special startup code to
6840 enable the application to run. The system must have PE installed in the
6841 standard location (@file{/usr/lpp/ppe.poe/}), or the @file{specs} file
6842 must be overridden with the @option{-specs=} option to specify the
6843 appropriate directory location. The Parallel Environment does not
6844 support threads, so the @option{-mpe} option and the @option{-pthread}
6845 option are incompatible.
6849 @opindex msoft-float
6850 @opindex mhard-float
6851 Generate code that does not use (uses) the floating-point register set.
6852 Software floating point emulation is provided if you use the
6853 @option{-msoft-float} option, and pass the option to GCC when linking.
6856 @itemx -mno-multiple
6858 @opindex mno-multiple
6859 Generate code that uses (does not use) the load multiple word
6860 instructions and the store multiple word instructions. These
6861 instructions are generated by default on POWER systems, and not
6862 generated on PowerPC systems. Do not use @option{-mmultiple} on little
6863 endian PowerPC systems, since those instructions do not work when the
6864 processor is in little endian mode. The exceptions are PPC740 and
6865 PPC750 which permit the instructions usage in little endian mode.
6871 Generate code that uses (does not use) the load string instructions
6872 and the store string word instructions to save multiple registers and
6873 do small block moves. These instructions are generated by default on
6874 POWER systems, and not generated on PowerPC systems. Do not use
6875 @option{-mstring} on little endian PowerPC systems, since those
6876 instructions do not work when the processor is in little endian mode.
6877 The exceptions are PPC740 and PPC750 which permit the instructions
6878 usage in little endian mode.
6884 Generate code that uses (does not use) the load or store instructions
6885 that update the base register to the address of the calculated memory
6886 location. These instructions are generated by default. If you use
6887 @option{-mno-update}, there is a small window between the time that the
6888 stack pointer is updated and the address of the previous frame is
6889 stored, which means code that walks the stack frame across interrupts or
6890 signals may get corrupted data.
6893 @itemx -mno-fused-madd
6894 @opindex mfused-madd
6895 @opindex mno-fused-madd
6896 Generate code that uses (does not use) the floating point multiply and
6897 accumulate instructions. These instructions are generated by default if
6898 hardware floating is used.
6900 @item -mno-bit-align
6902 @opindex mno-bit-align
6904 On System V.4 and embedded PowerPC systems do not (do) force structures
6905 and unions that contain bit-fields to be aligned to the base type of the
6908 For example, by default a structure containing nothing but 8
6909 @code{unsigned} bit-fields of length 1 would be aligned to a 4 byte
6910 boundary and have a size of 4 bytes. By using @option{-mno-bit-align},
6911 the structure would be aligned to a 1 byte boundary and be one byte in
6914 @item -mno-strict-align
6915 @itemx -mstrict-align
6916 @opindex mno-strict-align
6917 @opindex mstrict-align
6918 On System V.4 and embedded PowerPC systems do not (do) assume that
6919 unaligned memory references will be handled by the system.
6922 @itemx -mno-relocatable
6923 @opindex mrelocatable
6924 @opindex mno-relocatable
6925 On embedded PowerPC systems generate code that allows (does not allow)
6926 the program to be relocated to a different address at runtime. If you
6927 use @option{-mrelocatable} on any module, all objects linked together must
6928 be compiled with @option{-mrelocatable} or @option{-mrelocatable-lib}.
6930 @item -mrelocatable-lib
6931 @itemx -mno-relocatable-lib
6932 @opindex mrelocatable-lib
6933 @opindex mno-relocatable-lib
6934 On embedded PowerPC systems generate code that allows (does not allow)
6935 the program to be relocated to a different address at runtime. Modules
6936 compiled with @option{-mrelocatable-lib} can be linked with either modules
6937 compiled without @option{-mrelocatable} and @option{-mrelocatable-lib} or
6938 with modules compiled with the @option{-mrelocatable} options.
6944 On System V.4 and embedded PowerPC systems do not (do) assume that
6945 register 2 contains a pointer to a global area pointing to the addresses
6946 used in the program.
6949 @itemx -mlittle-endian
6951 @opindex mlittle-endian
6952 On System V.4 and embedded PowerPC systems compile code for the
6953 processor in little endian mode. The @option{-mlittle-endian} option is
6954 the same as @option{-mlittle}.
6959 @opindex mbig-endian
6960 On System V.4 and embedded PowerPC systems compile code for the
6961 processor in big endian mode. The @option{-mbig-endian} option is
6962 the same as @option{-mbig}.
6966 On System V.4 and embedded PowerPC systems compile code using calling
6967 conventions that adheres to the March 1995 draft of the System V
6968 Application Binary Interface, PowerPC processor supplement. This is the
6969 default unless you configured GCC using @samp{powerpc-*-eabiaix}.
6971 @item -mcall-sysv-eabi
6972 @opindex mcall-sysv-eabi
6973 Specify both @option{-mcall-sysv} and @option{-meabi} options.
6975 @item -mcall-sysv-noeabi
6976 @opindex mcall-sysv-noeabi
6977 Specify both @option{-mcall-sysv} and @option{-mno-eabi} options.
6981 On System V.4 and embedded PowerPC systems compile code using calling
6982 conventions that are similar to those used on AIX@. This is the
6983 default if you configured GCC using @samp{powerpc-*-eabiaix}.
6985 @item -mcall-solaris
6986 @opindex mcall-solaris
6987 On System V.4 and embedded PowerPC systems compile code for the Solaris
6991 @opindex mcall-linux
6992 On System V.4 and embedded PowerPC systems compile code for the
6993 Linux-based GNU system.
6997 On System V.4 and embedded PowerPC systems compile code for the
6998 Hurd-based GNU system.
7001 @opindex mcall-netbsd
7002 On System V.4 and embedded PowerPC systems compile code for the
7003 NetBSD operating system.
7005 @item -maix-struct-return
7006 @opindex maix-struct-return
7007 Return all structures in memory (as specified by the AIX ABI)@.
7009 @item -msvr4-struct-return
7010 @opindex msvr4-struct-return
7011 Return structures smaller than 8 bytes in registers (as specified by the
7015 @opindex mabi=altivec
7016 Extend the current ABI with AltiVec ABI extensions. This does not
7017 change the default ABI, instead it adds the AltiVec ABI extensions to
7020 @item -mabi=no-altivec
7021 @opindex mabi=no-altivec
7022 Disable AltiVec ABI extensions for the current ABI.
7025 @itemx -mno-prototype
7027 @opindex mno-prototype
7028 On System V.4 and embedded PowerPC systems assume that all calls to
7029 variable argument functions are properly prototyped. Otherwise, the
7030 compiler must insert an instruction before every non prototyped call to
7031 set or clear bit 6 of the condition code register (@var{CR}) to
7032 indicate whether floating point values were passed in the floating point
7033 registers in case the function takes a variable arguments. With
7034 @option{-mprototype}, only calls to prototyped variable argument functions
7035 will set or clear the bit.
7039 On embedded PowerPC systems, assume that the startup module is called
7040 @file{sim-crt0.o} and that the standard C libraries are @file{libsim.a} and
7041 @file{libc.a}. This is the default for @samp{powerpc-*-eabisim}.
7046 On embedded PowerPC systems, assume that the startup module is called
7047 @file{crt0.o} and the standard C libraries are @file{libmvme.a} and
7052 On embedded PowerPC systems, assume that the startup module is called
7053 @file{crt0.o} and the standard C libraries are @file{libads.a} and
7057 @opindex myellowknife
7058 On embedded PowerPC systems, assume that the startup module is called
7059 @file{crt0.o} and the standard C libraries are @file{libyk.a} and
7064 On System V.4 and embedded PowerPC systems, specify that you are
7065 compiling for a VxWorks system.
7069 Specify that you are compiling for the WindISS simulation environment.
7073 On embedded PowerPC systems, set the @var{PPC_EMB} bit in the ELF flags
7074 header to indicate that @samp{eabi} extended relocations are used.
7080 On System V.4 and embedded PowerPC systems do (do not) adhere to the
7081 Embedded Applications Binary Interface (eabi) which is a set of
7082 modifications to the System V.4 specifications. Selecting @option{-meabi}
7083 means that the stack is aligned to an 8 byte boundary, a function
7084 @code{__eabi} is called to from @code{main} to set up the eabi
7085 environment, and the @option{-msdata} option can use both @code{r2} and
7086 @code{r13} to point to two separate small data areas. Selecting
7087 @option{-mno-eabi} means that the stack is aligned to a 16 byte boundary,
7088 do not call an initialization function from @code{main}, and the
7089 @option{-msdata} option will only use @code{r13} to point to a single
7090 small data area. The @option{-meabi} option is on by default if you
7091 configured GCC using one of the @samp{powerpc*-*-eabi*} options.
7094 @opindex msdata=eabi
7095 On System V.4 and embedded PowerPC systems, put small initialized
7096 @code{const} global and static data in the @samp{.sdata2} section, which
7097 is pointed to by register @code{r2}. Put small initialized
7098 non-@code{const} global and static data in the @samp{.sdata} section,
7099 which is pointed to by register @code{r13}. Put small uninitialized
7100 global and static data in the @samp{.sbss} section, which is adjacent to
7101 the @samp{.sdata} section. The @option{-msdata=eabi} option is
7102 incompatible with the @option{-mrelocatable} option. The
7103 @option{-msdata=eabi} option also sets the @option{-memb} option.
7106 @opindex msdata=sysv
7107 On System V.4 and embedded PowerPC systems, put small global and static
7108 data in the @samp{.sdata} section, which is pointed to by register
7109 @code{r13}. Put small uninitialized global and static data in the
7110 @samp{.sbss} section, which is adjacent to the @samp{.sdata} section.
7111 The @option{-msdata=sysv} option is incompatible with the
7112 @option{-mrelocatable} option.
7114 @item -msdata=default
7116 @opindex msdata=default
7118 On System V.4 and embedded PowerPC systems, if @option{-meabi} is used,
7119 compile code the same as @option{-msdata=eabi}, otherwise compile code the
7120 same as @option{-msdata=sysv}.
7123 @opindex msdata-data
7124 On System V.4 and embedded PowerPC systems, put small global and static
7125 data in the @samp{.sdata} section. Put small uninitialized global and
7126 static data in the @samp{.sbss} section. Do not use register @code{r13}
7127 to address small data however. This is the default behavior unless
7128 other @option{-msdata} options are used.
7132 @opindex msdata=none
7134 On embedded PowerPC systems, put all initialized global and static data
7135 in the @samp{.data} section, and all uninitialized data in the
7136 @samp{.bss} section.
7140 @cindex smaller data references (PowerPC)
7141 @cindex .sdata/.sdata2 references (PowerPC)
7142 On embedded PowerPC systems, put global and static items less than or
7143 equal to @var{num} bytes into the small data or bss sections instead of
7144 the normal data or bss section. By default, @var{num} is 8. The
7145 @option{-G @var{num}} switch is also passed to the linker.
7146 All modules should be compiled with the same @option{-G @var{num}} value.
7149 @itemx -mno-regnames
7151 @opindex mno-regnames
7152 On System V.4 and embedded PowerPC systems do (do not) emit register
7153 names in the assembly language output using symbolic forms.
7156 @itemx -mno-longcall
7158 @opindex mno-longcall
7159 Default to making all function calls via pointers, so that functions
7160 which reside further than 64 megabytes (67,108,864 bytes) from the
7161 current location can be called. This setting can be overridden by the
7162 @code{shortcall} function attribute, or by @code{#pragma longcall(0)}.
7164 Some linkers are capable of detecting out-of-range calls and generating
7165 glue code on the fly. On these systems, long calls are unnecessary and
7166 generate slower code. As of this writing, the AIX linker can do this,
7167 as can the GNU linker for PowerPC/64. It is planned to add this feature
7168 to the GNU linker for 32-bit PowerPC systems as well.
7170 In the future, we may cause GCC to ignore all longcall specifications
7171 when the linker is known to generate glue.
7175 Adds support for multithreading with the @dfn{pthreads} library.
7176 This option sets flags for both the preprocessor and linker.
7181 @subsection IBM RT Options
7183 @cindex IBM RT options
7185 These @samp{-m} options are defined for the IBM RT PC:
7189 @opindex min-line-mul
7190 Use an in-line code sequence for integer multiplies. This is the
7193 @item -mcall-lib-mul
7194 @opindex mcall-lib-mul
7195 Call @code{lmul$$} for integer multiples.
7197 @item -mfull-fp-blocks
7198 @opindex mfull-fp-blocks
7199 Generate full-size floating point data blocks, including the minimum
7200 amount of scratch space recommended by IBM@. This is the default.
7202 @item -mminimum-fp-blocks
7203 @opindex mminimum-fp-blocks
7204 Do not include extra scratch space in floating point data blocks. This
7205 results in smaller code, but slower execution, since scratch space must
7206 be allocated dynamically.
7208 @cindex @file{stdarg.h} and RT PC
7209 @item -mfp-arg-in-fpregs
7210 @opindex mfp-arg-in-fpregs
7211 Use a calling sequence incompatible with the IBM calling convention in
7212 which floating point arguments are passed in floating point registers.
7213 Note that @code{stdarg.h} will not work with floating point operands
7214 if this option is specified.
7216 @item -mfp-arg-in-gregs
7217 @opindex mfp-arg-in-gregs
7218 Use the normal calling convention for floating point arguments. This is
7221 @item -mhc-struct-return
7222 @opindex mhc-struct-return
7223 Return structures of more than one word in memory, rather than in a
7224 register. This provides compatibility with the MetaWare HighC (hc)
7225 compiler. Use the option @option{-fpcc-struct-return} for compatibility
7226 with the Portable C Compiler (pcc).
7228 @item -mnohc-struct-return
7229 @opindex mnohc-struct-return
7230 Return some structures of more than one word in registers, when
7231 convenient. This is the default. For compatibility with the
7232 IBM-supplied compilers, use the option @option{-fpcc-struct-return} or the
7233 option @option{-mhc-struct-return}.
7237 @subsection MIPS Options
7238 @cindex MIPS options
7240 These @samp{-m} options are defined for the MIPS family of computers:
7244 @item -march=@var{arch}
7246 Generate code that will run on @var{arch}, which can be the name of a
7247 generic MIPS ISA, or the name of a particular processor. The ISA names
7248 are: @samp{mips1}, @samp{mips2}, @samp{mips3}, @samp{mips4}, @samp{mips32}
7249 and @samp{mips64}. The processor names are: @samp{r2000},
7250 @samp{r3000}, @samp{r3900}, @samp{r4000}, @samp{vr4100}, @samp{vr4300},
7251 @samp{r4400}, @samp{r4600}, @samp{r4650}, @samp{vr5000}, @samp{r6000},
7252 @samp{r8000}, @samp{4kc}, @samp{4kp}, @samp{5kc}, @samp{20kc},
7253 @samp{orion}, and @samp{sb1}. The special value @samp{from-abi} selects the
7254 most compatible architecture for the selected ABI (that is,
7255 @samp{mips1} for 32-bit ABIs and @samp{mips3} for 64-bit ABIs)@.
7257 In processor names, a final @samp{000} can be abbreviated as @samp{k}
7258 (for example, @samp{-march=r2k}). Prefixes are optional, and
7259 @samp{vr} may be written @samp{r}.
7261 GCC defines two macros based on the value of this option. The first
7262 is @samp{_MIPS_ARCH}, which gives the name of target architecture, as
7263 a string. The second has the form @samp{_MIPS_ARCH_@var{foo}},
7264 where @var{foo} is the capitialized value of @samp{_MIPS_ARCH}@.
7265 For example, @samp{-march=r2000} will set @samp{_MIPS_ARCH}
7266 to @samp{"r2000"} and define the macro @samp{_MIPS_ARCH_R2000}.
7268 Note that the @samp{_MIPS_ARCH} macro uses the processor names given
7269 above. In other words, it will have the full prefix and will not
7270 abbreviate @samp{000} as @samp{k}. In the case of @samp{from-abi},
7271 the macro names the resolved architecture (either @samp{"mips1"} or
7272 @samp{"mips3"}). It names the default architecture when no
7273 @option{-march} option is given.
7275 @item -mtune=@var{arch}
7277 Optimize for @var{arch}. Among other things, this option controls
7278 the way instructions are scheduled, and the perceived cost of arithmetic
7279 operations. The list of @var{arch} values is the same as for
7282 When this option is not used, GCC will optimize for the processor
7283 specified by @option{-march}. By using @option{-march} and
7284 @option{-mtune} together, it is possible to generate code that will
7285 run on a family of processors, but optimize the code for one
7286 particular member of that family.
7288 @samp{-mtune} defines the macros @samp{_MIPS_TUNE} and
7289 @samp{_MIPS_TUNE_@var{foo}}, which work in the same way as the
7290 @samp{-march} ones described above.
7294 Equivalent to @samp{-march=mips1}.
7298 Equivalent to @samp{-march=mips2}.
7302 Equivalent to @samp{-march=mips3}.
7306 Equivalent to @samp{-march=mips4}.
7310 Equivalent to @samp{-march=mips32}.
7314 Equivalent to @samp{-march=mips64}.
7317 @itemx -mno-fused-madd
7318 @opindex mfused-madd
7319 @opindex mno-fused-madd
7320 Generate code that uses (does not use) the floating point multiply and
7321 accumulate instructions, when they are available. These instructions
7322 are generated by default if they are available, but this may be
7323 undesirable if the extra precision causes problems or on certain chips
7324 in the mode where denormals are rounded to zero where denormals
7325 generated by multiply and accumulate instructions cause exceptions
7330 Assume that floating point registers are 32 bits wide.
7334 Assume that floating point registers are 64 bits wide.
7338 Assume that general purpose registers are 32 bits wide.
7342 Assume that general purpose registers are 64 bits wide.
7346 Force int and long types to be 64 bits wide. See @option{-mlong32} for an
7347 explanation of the default, and the width of pointers.
7351 Force long types to be 64 bits wide. See @option{-mlong32} for an
7352 explanation of the default, and the width of pointers.
7356 Force long, int, and pointer types to be 32 bits wide.
7358 The default size of ints, longs and pointers depends on the ABI@. All
7359 the supported ABIs use 32-bit ints. The n64 ABI uses 64-bit longs, as
7360 does the 64-bit Cygnus EABI; the others use 32-bit longs. Pointers
7361 are the same size as longs, or the same size as integer registers,
7362 whichever is smaller.
7376 Generate code for the given ABI@.
7378 Note that there are two embedded ABIs: @option{-mabi=eabi}
7379 selects the one defined by Cygnus while @option{-meabi=meabi}
7380 selects the one defined by MIPS@. Both these ABIs have
7381 32-bit and 64-bit variants. Normally, GCC will generate
7382 64-bit code when you select a 64-bit architecture, but you
7383 can use @option{-mgp32} to get 32-bit code instead.
7387 Generate code for the MIPS assembler, and invoke @file{mips-tfile} to
7388 add normal debug information. This is the default for all
7389 platforms except for the OSF/1 reference platform, using the OSF/rose
7390 object format. If the either of the @option{-gstabs} or @option{-gstabs+}
7391 switches are used, the @file{mips-tfile} program will encapsulate the
7392 stabs within MIPS ECOFF@.
7396 Generate code for the GNU assembler. This is the default on the OSF/1
7397 reference platform, using the OSF/rose object format. Also, this is
7398 the default if the configure option @option{--with-gnu-as} is used.
7400 @item -msplit-addresses
7401 @itemx -mno-split-addresses
7402 @opindex msplit-addresses
7403 @opindex mno-split-addresses
7404 Generate code to load the high and low parts of address constants separately.
7405 This allows GCC to optimize away redundant loads of the high order
7406 bits of addresses. This optimization requires GNU as and GNU ld.
7407 This optimization is enabled by default for some embedded targets where
7408 GNU as and GNU ld are standard.
7414 The @option{-mrnames} switch says to output code using the MIPS software
7415 names for the registers, instead of the hardware names (ie, @var{a0}
7416 instead of @var{$4}). The only known assembler that supports this option
7417 is the Algorithmics assembler.
7423 The @option{-mgpopt} switch says to write all of the data declarations
7424 before the instructions in the text section, this allows the MIPS
7425 assembler to generate one word memory references instead of using two
7426 words for short global or static data items. This is on by default if
7427 optimization is selected.
7433 For each non-inline function processed, the @option{-mstats} switch
7434 causes the compiler to emit one line to the standard error file to
7435 print statistics about the program (number of registers saved, stack
7442 The @option{-mmemcpy} switch makes all block moves call the appropriate
7443 string function (@samp{memcpy} or @samp{bcopy}) instead of possibly
7444 generating inline code.
7447 @itemx -mno-mips-tfile
7448 @opindex mmips-tfile
7449 @opindex mno-mips-tfile
7450 The @option{-mno-mips-tfile} switch causes the compiler not
7451 postprocess the object file with the @file{mips-tfile} program,
7452 after the MIPS assembler has generated it to add debug support. If
7453 @file{mips-tfile} is not run, then no local variables will be
7454 available to the debugger. In addition, @file{stage2} and
7455 @file{stage3} objects will have the temporary file names passed to the
7456 assembler embedded in the object file, which means the objects will
7457 not compare the same. The @option{-mno-mips-tfile} switch should only
7458 be used when there are bugs in the @file{mips-tfile} program that
7459 prevents compilation.
7462 @opindex msoft-float
7463 Generate output containing library calls for floating point.
7464 @strong{Warning:} the requisite libraries are not part of GCC@.
7465 Normally the facilities of the machine's usual C compiler are used, but
7466 this can't be done directly in cross-compilation. You must make your
7467 own arrangements to provide suitable library functions for
7471 @opindex mhard-float
7472 Generate output containing floating point instructions. This is the
7473 default if you use the unmodified sources.
7476 @itemx -mno-abicalls
7478 @opindex mno-abicalls
7479 Emit (or do not emit) the pseudo operations @samp{.abicalls},
7480 @samp{.cpload}, and @samp{.cprestore} that some System V.4 ports use for
7481 position independent code.
7484 @itemx -mno-long-calls
7485 @opindex mlong-calls
7486 @opindex mno-long-calls
7487 Do all calls with the @samp{JALR} instruction, which requires
7488 loading up a function's address into a register before the call.
7489 You need to use this switch, if you call outside of the current
7490 512 megabyte segment to functions that are not through pointers.
7493 @itemx -mno-half-pic
7495 @opindex mno-half-pic
7496 Put pointers to extern references into the data section and load them
7497 up, rather than put the references in the text section.
7499 @item -membedded-pic
7500 @itemx -mno-embedded-pic
7501 @opindex membedded-pic
7502 @opindex mno-embedded-pic
7503 Generate PIC code suitable for some embedded systems. All calls are
7504 made using PC relative address, and all data is addressed using the $gp
7505 register. No more than 65536 bytes of global data may be used. This
7506 requires GNU as and GNU ld which do most of the work. This currently
7507 only works on targets which use ECOFF; it does not work with ELF@.
7509 @item -membedded-data
7510 @itemx -mno-embedded-data
7511 @opindex membedded-data
7512 @opindex mno-embedded-data
7513 Allocate variables to the read-only data section first if possible, then
7514 next in the small data section if possible, otherwise in data. This gives
7515 slightly slower code than the default, but reduces the amount of RAM required
7516 when executing, and thus may be preferred for some embedded systems.
7518 @item -muninit-const-in-rodata
7519 @itemx -mno-uninit-const-in-rodata
7520 @opindex muninit-const-in-rodata
7521 @opindex mno-uninit-const-in-rodata
7522 When used together with @option{-membedded-data}, it will always store uninitialized
7523 const variables in the read-only data section.
7525 @item -msingle-float
7526 @itemx -mdouble-float
7527 @opindex msingle-float
7528 @opindex mdouble-float
7529 The @option{-msingle-float} switch tells gcc to assume that the floating
7530 point coprocessor only supports single precision operations, as on the
7531 @samp{r4650} chip. The @option{-mdouble-float} switch permits gcc to use
7532 double precision operations. This is the default.
7538 Permit use of the @samp{mad}, @samp{madu} and @samp{mul} instructions,
7539 as on the @samp{r4650} chip.
7543 Turns on @option{-msingle-float}, @option{-mmad}, and, at least for now,
7544 @option{-mcpu=r4650}.
7550 Enable 16-bit instructions.
7554 Use the entry and exit pseudo ops. This option can only be used with
7559 Compile code for the processor in little endian mode.
7560 The requisite libraries are assumed to exist.
7564 Compile code for the processor in big endian mode.
7565 The requisite libraries are assumed to exist.
7569 @cindex smaller data references (MIPS)
7570 @cindex gp-relative references (MIPS)
7571 Put global and static items less than or equal to @var{num} bytes into
7572 the small data or bss sections instead of the normal data or bss
7573 section. This allows the assembler to emit one word memory reference
7574 instructions based on the global pointer (@var{gp} or @var{$28}),
7575 instead of the normal two words used. By default, @var{num} is 8 when
7576 the MIPS assembler is used, and 0 when the GNU assembler is used. The
7577 @option{-G @var{num}} switch is also passed to the assembler and linker.
7578 All modules should be compiled with the same @option{-G @var{num}}
7583 Tell the MIPS assembler to not run its preprocessor over user
7584 assembler files (with a @samp{.s} suffix) when assembling them.
7588 Pass an option to gas which will cause nops to be inserted if
7589 the read of the destination register of an mfhi or mflo instruction
7590 occurs in the following two instructions.
7594 Do not include the default crt0.
7596 @item -mflush-func=@var{func}
7597 @itemx -mno-flush-func
7598 @opindex mflush-func
7599 Specifies the function to call to flush the I and D caches, or to not
7600 call any such function. If called, the function must take the same
7601 arguments as the common @code{_flush_func()}, that is, the address of the
7602 memory range for which the cache is being flushed, the size of the
7603 memory range, and the number 3 (to flush both caches). The default
7604 depends on the target gcc was configured for, but commonly is either
7605 @samp{_flush_func} or @samp{__cpu_flush}.
7607 @item -mbranch-likely
7608 @itemx -mno-branch-likely
7609 @opindex mbranch-likely
7610 @opindex mno-branch-likely
7611 Enable or disable use of Branch Likely instructions, regardless of the
7612 default for the selected architecture. By default, Branch Likely
7613 instructions may be generated if they are supported by the selected
7614 architecture. An exception is for the MIPS32 and MIPS64 architectures
7615 and processors which implement those architectures; for those, Branch
7616 Likely instructions will not be generated by default because the MIPS32
7617 and MIPS64 architectures specifically deprecate their use.
7620 @node i386 and x86-64 Options
7621 @subsection Intel 386 and AMD x86-64 Options
7622 @cindex i386 Options
7623 @cindex x86-64 Options
7624 @cindex Intel 386 Options
7625 @cindex AMD x86-64 Options
7627 These @samp{-m} options are defined for the i386 and x86-64 family of
7631 @item -mcpu=@var{cpu-type}
7633 Tune to @var{cpu-type} everything applicable about the generated code, except
7634 for the ABI and the set of available instructions. The choices for
7635 @var{cpu-type} are @samp{i386}, @samp{i486}, @samp{i586}, @samp{i686},
7636 @samp{pentium}, @samp{pentium-mmx}, @samp{pentiumpro}, @samp{pentium2},
7637 @samp{pentium3}, @samp{pentium4}, @samp{k6}, @samp{k6-2}, @samp{k6-3},
7638 @samp{athlon}, @samp{athlon-tbird}, @samp{athlon-4}, @samp{athlon-xp},
7639 @samp{athlon-mp}, @samp{winchip-c6}, @samp{winchip2} and @samp{c3}.
7641 While picking a specific @var{cpu-type} will schedule things appropriately
7642 for that particular chip, the compiler will not generate any code that
7643 does not run on the i386 without the @option{-march=@var{cpu-type}} option
7644 being used. @samp{i586} is equivalent to @samp{pentium} and @samp{i686}
7645 is equivalent to @samp{pentiumpro}. @samp{k6} and @samp{athlon} are the
7646 AMD chips as opposed to the Intel ones.
7648 @item -march=@var{cpu-type}
7650 Generate instructions for the machine type @var{cpu-type}. The choices
7651 for @var{cpu-type} are the same as for @option{-mcpu}. Moreover,
7652 specifying @option{-march=@var{cpu-type}} implies @option{-mcpu=@var{cpu-type}}.
7661 @opindex mpentiumpro
7662 These options are synonyms for @option{-mcpu=i386}, @option{-mcpu=i486},
7663 @option{-mcpu=pentium}, and @option{-mcpu=pentiumpro} respectively.
7664 These synonyms are deprecated.
7666 @item -mfpmath=@var{unit}
7668 generate floating point arithmetics for selected unit @var{unit}. the choices
7673 Use the standard 387 floating point coprocessor present majority of chips and
7674 emulated otherwise. Code compiled with this option will run almost everywhere.
7675 The temporary results are computed in 80bit precesion instead of precision
7676 specified by the type resulting in slightly different results compared to most
7677 of other chips. See @option{-ffloat-store} for more detailed description.
7679 This is the default choice for i386 compiler.
7682 Use scalar floating point instructions present in the SSE instruction set.
7683 This instruction set is supported by Pentium3 and newer chips, in the AMD line
7684 by Athlon-4, Athlon-xp and Athlon-mp chips. The earlier version of SSE
7685 instruction set supports only single precision arithmetics, thus the double and
7686 extended precision arithmetics is still done using 387. Later version, present
7687 only in Pentium4 and the future AMD x86-64 chips supports double precision
7690 For i387 you need to use @option{-march=@var{cpu-type}}, @option{-msse} or
7691 @option{-msse2} switches to enable SSE extensions and make this option
7692 effective. For x86-64 compiler, these extensions are enabled by default.
7694 The resulting code should be considerably faster in majority of cases and avoid
7695 the numerical instability problems of 387 code, but may break some existing
7696 code that expects temporaries to be 80bit.
7698 This is the default choice for x86-64 compiler.
7701 Attempt to utilize both instruction sets at once. This effectivly double the
7702 amount of available registers and on chips with separate execution units for
7703 387 and SSE the execution resources too. Use this option with care, as it is
7704 still experimental, because gcc register allocator does not model separate
7705 functional units well resulting in instable performance.
7708 @item -masm=@var{dialect}
7709 @opindex masm=@var{dialect}
7710 Output asm instructions using selected @var{dialect}. Supported choices are
7711 @samp{intel} or @samp{att} (the default one).
7716 @opindex mno-ieee-fp
7717 Control whether or not the compiler uses IEEE floating point
7718 comparisons. These handle correctly the case where the result of a
7719 comparison is unordered.
7722 @opindex msoft-float
7723 Generate output containing library calls for floating point.
7724 @strong{Warning:} the requisite libraries are not part of GCC@.
7725 Normally the facilities of the machine's usual C compiler are used, but
7726 this can't be done directly in cross-compilation. You must make your
7727 own arrangements to provide suitable library functions for
7730 On machines where a function returns floating point results in the 80387
7731 register stack, some floating point opcodes may be emitted even if
7732 @option{-msoft-float} is used.
7734 @item -mno-fp-ret-in-387
7735 @opindex mno-fp-ret-in-387
7736 Do not use the FPU registers for return values of functions.
7738 The usual calling convention has functions return values of types
7739 @code{float} and @code{double} in an FPU register, even if there
7740 is no FPU@. The idea is that the operating system should emulate
7743 The option @option{-mno-fp-ret-in-387} causes such values to be returned
7744 in ordinary CPU registers instead.
7746 @item -mno-fancy-math-387
7747 @opindex mno-fancy-math-387
7748 Some 387 emulators do not support the @code{sin}, @code{cos} and
7749 @code{sqrt} instructions for the 387. Specify this option to avoid
7750 generating those instructions. This option is the default on FreeBSD,
7751 OpenBSD and NetBSD@. This option is overridden when @option{-march}
7752 indicates that the target cpu will always have an FPU and so the
7753 instruction will not need emulation. As of revision 2.6.1, these
7754 instructions are not generated unless you also use the
7755 @option{-funsafe-math-optimizations} switch.
7757 @item -malign-double
7758 @itemx -mno-align-double
7759 @opindex malign-double
7760 @opindex mno-align-double
7761 Control whether GCC aligns @code{double}, @code{long double}, and
7762 @code{long long} variables on a two word boundary or a one word
7763 boundary. Aligning @code{double} variables on a two word boundary will
7764 produce code that runs somewhat faster on a @samp{Pentium} at the
7765 expense of more memory.
7767 @strong{Warning:} if you use the @samp{-malign-double} switch,
7768 structures containing the above types will be aligned differently than
7769 the published application binary interface specifications for the 386.
7771 @item -m128bit-long-double
7772 @opindex m128bit-long-double
7773 Control the size of @code{long double} type. i386 application binary interface
7774 specify the size to be 12 bytes, while modern architectures (Pentium and newer)
7775 prefer @code{long double} aligned to 8 or 16 byte boundary. This is
7776 impossible to reach with 12 byte long doubles in the array accesses.
7778 @strong{Warning:} if you use the @option{-m128bit-long-double} switch, the
7779 structures and arrays containing @code{long double} will change their size as
7780 well as function calling convention for function taking @code{long double}
7783 @item -m96bit-long-double
7784 @opindex m96bit-long-double
7785 Set the size of @code{long double} to 96 bits as required by the i386
7786 application binary interface. This is the default.
7789 @itemx -mno-svr3-shlib
7790 @opindex msvr3-shlib
7791 @opindex mno-svr3-shlib
7792 Control whether GCC places uninitialized local variables into the
7793 @code{bss} or @code{data} segments. @option{-msvr3-shlib} places them
7794 into @code{bss}. These options are meaningful only on System V Release 3.
7798 Use a different function-calling convention, in which functions that
7799 take a fixed number of arguments return with the @code{ret} @var{num}
7800 instruction, which pops their arguments while returning. This saves one
7801 instruction in the caller since there is no need to pop the arguments
7804 You can specify that an individual function is called with this calling
7805 sequence with the function attribute @samp{stdcall}. You can also
7806 override the @option{-mrtd} option by using the function attribute
7807 @samp{cdecl}. @xref{Function Attributes}.
7809 @strong{Warning:} this calling convention is incompatible with the one
7810 normally used on Unix, so you cannot use it if you need to call
7811 libraries compiled with the Unix compiler.
7813 Also, you must provide function prototypes for all functions that
7814 take variable numbers of arguments (including @code{printf});
7815 otherwise incorrect code will be generated for calls to those
7818 In addition, seriously incorrect code will result if you call a
7819 function with too many arguments. (Normally, extra arguments are
7820 harmlessly ignored.)
7822 @item -mregparm=@var{num}
7824 Control how many registers are used to pass integer arguments. By
7825 default, no registers are used to pass arguments, and at most 3
7826 registers can be used. You can control this behavior for a specific
7827 function by using the function attribute @samp{regparm}.
7828 @xref{Function Attributes}.
7830 @strong{Warning:} if you use this switch, and
7831 @var{num} is nonzero, then you must build all modules with the same
7832 value, including any libraries. This includes the system libraries and
7835 @item -mpreferred-stack-boundary=@var{num}
7836 @opindex mpreferred-stack-boundary
7837 Attempt to keep the stack boundary aligned to a 2 raised to @var{num}
7838 byte boundary. If @option{-mpreferred-stack-boundary} is not specified,
7839 the default is 4 (16 bytes or 128 bits), except when optimizing for code
7840 size (@option{-Os}), in which case the default is the minimum correct
7841 alignment (4 bytes for x86, and 8 bytes for x86-64).
7843 On Pentium and PentiumPro, @code{double} and @code{long double} values
7844 should be aligned to an 8 byte boundary (see @option{-malign-double}) or
7845 suffer significant run time performance penalties. On Pentium III, the
7846 Streaming SIMD Extension (SSE) data type @code{__m128} suffers similar
7847 penalties if it is not 16 byte aligned.
7849 To ensure proper alignment of this values on the stack, the stack boundary
7850 must be as aligned as that required by any value stored on the stack.
7851 Further, every function must be generated such that it keeps the stack
7852 aligned. Thus calling a function compiled with a higher preferred
7853 stack boundary from a function compiled with a lower preferred stack
7854 boundary will most likely misalign the stack. It is recommended that
7855 libraries that use callbacks always use the default setting.
7857 This extra alignment does consume extra stack space, and generally
7858 increases code size. Code that is sensitive to stack space usage, such
7859 as embedded systems and operating system kernels, may want to reduce the
7860 preferred alignment to @option{-mpreferred-stack-boundary=2}.
7876 These switches enable or disable the use of built-in functions that allow
7877 direct access to the MMX, SSE and 3Dnow extensions of the instruction set.
7879 @xref{X86 Built-in Functions}, for details of the functions enabled
7880 and disabled by these switches.
7882 To have SSE/SSE2 instructions generated automatically from floating-point
7883 code, see @option{-mfpmath=sse}.
7886 @itemx -mno-push-args
7888 @opindex mno-push-args
7889 Use PUSH operations to store outgoing parameters. This method is shorter
7890 and usually equally fast as method using SUB/MOV operations and is enabled
7891 by default. In some cases disabling it may improve performance because of
7892 improved scheduling and reduced dependencies.
7894 @item -maccumulate-outgoing-args
7895 @opindex maccumulate-outgoing-args
7896 If enabled, the maximum amount of space required for outgoing arguments will be
7897 computed in the function prologue. This is faster on most modern CPUs
7898 because of reduced dependencies, improved scheduling and reduced stack usage
7899 when preferred stack boundary is not equal to 2. The drawback is a notable
7900 increase in code size. This switch implies @option{-mno-push-args}.
7904 Support thread-safe exception handling on @samp{Mingw32}. Code that relies
7905 on thread-safe exception handling must compile and link all code with the
7906 @option{-mthreads} option. When compiling, @option{-mthreads} defines
7907 @option{-D_MT}; when linking, it links in a special thread helper library
7908 @option{-lmingwthrd} which cleans up per thread exception handling data.
7910 @item -mno-align-stringops
7911 @opindex mno-align-stringops
7912 Do not align destination of inlined string operations. This switch reduces
7913 code size and improves performance in case the destination is already aligned,
7914 but gcc don't know about it.
7916 @item -minline-all-stringops
7917 @opindex minline-all-stringops
7918 By default GCC inlines string operations only when destination is known to be
7919 aligned at least to 4 byte boundary. This enables more inlining, increase code
7920 size, but may improve performance of code that depends on fast memcpy, strlen
7921 and memset for short lengths.
7923 @item -momit-leaf-frame-pointer
7924 @opindex momit-leaf-frame-pointer
7925 Don't keep the frame pointer in a register for leaf functions. This
7926 avoids the instructions to save, set up and restore frame pointers and
7927 makes an extra register available in leaf functions. The option
7928 @option{-fomit-frame-pointer} removes the frame pointer for all functions
7929 which might make debugging harder.
7932 These @samp{-m} switches are supported in addition to the above
7933 on AMD x86-64 processors in 64-bit environments.
7940 Generate code for a 32-bit or 64-bit environment.
7941 The 32-bit environment sets int, long and pointer to 32 bits and
7942 generates code that runs on any i386 system.
7943 The 64-bit environment sets int to 32 bits and long and pointer
7944 to 64 bits and generates code for AMD's x86-64 architecture.
7947 @opindex no-red-zone
7948 Do not use a so called red zone for x86-64 code. The red zone is mandated
7949 by the x86-64 ABI, it is a 128-byte area beyond the location of the
7950 stack pointer that will not be modified by signal or interrupt handlers
7951 and therefore can be used for temporary data without adjusting the stack
7952 pointer. The flag @option{-mno-red-zone} disables this red zone.
7954 @item -mcmodel=small
7955 @opindex mcmodel=small
7956 Generate code for the small code model: the program and its symbols must
7957 be linked in the lower 2 GB of the address space. Pointers are 64 bits.
7958 Programs can be statically or dynamically linked. This is the default
7961 @item -mcmodel=kernel
7962 @opindex mcmodel=kernel
7963 Generate code for the kernel code model. The kernel runs in the
7964 negative 2 GB of the address space.
7965 This model has to be used for Linux kernel code.
7967 @item -mcmodel=medium
7968 @opindex mcmodel=medium
7969 Generate code for the medium model: The program is linked in the lower 2
7970 GB of the address space but symbols can be located anywhere in the
7971 address space. Programs can be statically or dynamically linked, but
7972 building of shared libraries are not supported with the medium model.
7974 @item -mcmodel=large
7975 @opindex mcmodel=large
7976 Generate code for the large model: This model makes no assumptions
7977 about addresses and sizes of sections. Currently GCC does not implement
7982 @subsection HPPA Options
7983 @cindex HPPA Options
7985 These @samp{-m} options are defined for the HPPA family of computers:
7988 @item -march=@var{architecture-type}
7990 Generate code for the specified architecture. The choices for
7991 @var{architecture-type} are @samp{1.0} for PA 1.0, @samp{1.1} for PA
7992 1.1, and @samp{2.0} for PA 2.0 processors. Refer to
7993 @file{/usr/lib/sched.models} on an HP-UX system to determine the proper
7994 architecture option for your machine. Code compiled for lower numbered
7995 architectures will run on higher numbered architectures, but not the
7998 PA 2.0 support currently requires gas snapshot 19990413 or later. The
7999 next release of binutils (current is 2.9.1) will probably contain PA 2.0
8003 @itemx -mpa-risc-1-1
8004 @itemx -mpa-risc-2-0
8005 @opindex mpa-risc-1-0
8006 @opindex mpa-risc-1-1
8007 @opindex mpa-risc-2-0
8008 Synonyms for @option{-march=1.0}, @option{-march=1.1}, and @option{-march=2.0} respectively.
8011 @opindex mbig-switch
8012 Generate code suitable for big switch tables. Use this option only if
8013 the assembler/linker complain about out of range branches within a switch
8016 @item -mjump-in-delay
8017 @opindex mjump-in-delay
8018 Fill delay slots of function calls with unconditional jump instructions
8019 by modifying the return pointer for the function call to be the target
8020 of the conditional jump.
8022 @item -mdisable-fpregs
8023 @opindex mdisable-fpregs
8024 Prevent floating point registers from being used in any manner. This is
8025 necessary for compiling kernels which perform lazy context switching of
8026 floating point registers. If you use this option and attempt to perform
8027 floating point operations, the compiler will abort.
8029 @item -mdisable-indexing
8030 @opindex mdisable-indexing
8031 Prevent the compiler from using indexing address modes. This avoids some
8032 rather obscure problems when compiling MIG generated code under MACH@.
8034 @item -mno-space-regs
8035 @opindex mno-space-regs
8036 Generate code that assumes the target has no space registers. This allows
8037 GCC to generate faster indirect calls and use unscaled index address modes.
8039 Such code is suitable for level 0 PA systems and kernels.
8041 @item -mfast-indirect-calls
8042 @opindex mfast-indirect-calls
8043 Generate code that assumes calls never cross space boundaries. This
8044 allows GCC to emit code which performs faster indirect calls.
8046 This option will not work in the presence of shared libraries or nested
8049 @item -mlong-load-store
8050 @opindex mlong-load-store
8051 Generate 3-instruction load and store sequences as sometimes required by
8052 the HP-UX 10 linker. This is equivalent to the @samp{+k} option to
8055 @item -mportable-runtime
8056 @opindex mportable-runtime
8057 Use the portable calling conventions proposed by HP for ELF systems.
8061 Enable the use of assembler directives only GAS understands.
8063 @item -mschedule=@var{cpu-type}
8065 Schedule code according to the constraints for the machine type
8066 @var{cpu-type}. The choices for @var{cpu-type} are @samp{700}
8067 @samp{7100}, @samp{7100LC}, @samp{7200}, @samp{7300} and @samp{8000}. Refer
8068 to @file{/usr/lib/sched.models} on an HP-UX system to determine the
8069 proper scheduling option for your machine. The default scheduling is
8073 @opindex mlinker-opt
8074 Enable the optimization pass in the HP-UX linker. Note this makes symbolic
8075 debugging impossible. It also triggers a bug in the HP-UX 8 and HP-UX 9
8076 linkers in which they give bogus error messages when linking some programs.
8079 @opindex msoft-float
8080 Generate output containing library calls for floating point.
8081 @strong{Warning:} the requisite libraries are not available for all HPPA
8082 targets. Normally the facilities of the machine's usual C compiler are
8083 used, but this cannot be done directly in cross-compilation. You must make
8084 your own arrangements to provide suitable library functions for
8085 cross-compilation. The embedded target @samp{hppa1.1-*-pro}
8086 does provide software floating point support.
8088 @option{-msoft-float} changes the calling convention in the output file;
8089 therefore, it is only useful if you compile @emph{all} of a program with
8090 this option. In particular, you need to compile @file{libgcc.a}, the
8091 library that comes with GCC, with @option{-msoft-float} in order for
8096 Generate the predefine, @code{_SIO}, for server IO. The default is
8097 @option{-mwsio}. This generates the predefines, @code{__hp9000s700},
8098 @code{__hp9000s700__} and @code{_WSIO}, for workstation IO. These
8099 options are available under HP-UX and HI-UX.
8103 Use GNU ld specific options. This passes @option{-shared} to ld when
8104 building a shared library. It is the default when GCC is configured,
8105 explicitly or implicitly, with the GNU linker. This option does not
8106 have any affect on which ld is called, it only changes what parameters
8107 are passed to that ld. The ld that is called is determined by the
8108 @option{--with-ld} configure option, gcc's program search path, and
8109 finally by the user's @env{PATH}. The linker used by GCC can be printed
8110 using @samp{which `gcc -print-prog-name=ld`}.
8114 Use HP ld specific options. This passes @option{-b} to ld when building
8115 a shared library and passes @option{+Accept TypeMismatch} to ld on all
8116 links. It is the default when GCC is configured, explicitly or
8117 implicitly, with the HP linker. This option does not have any affect on
8118 which ld is called, it only changes what parameters are passed to that
8119 ld. The ld that is called is determined by the @option{--with-ld}
8120 configure option, gcc's program search path, and finally by the user's
8121 @env{PATH}. The linker used by GCC can be printed using @samp{which
8122 `gcc -print-prog-name=ld`}.
8125 @opindex mno-long-calls
8126 Generate code that uses long call sequences. This ensures that a call
8127 is always able to reach linker generated stubs. The default is to generate
8128 long calls only when the distance from the call site to the beginning
8129 of the function or translation unit, as the case may be, exceeds a
8130 predefined limit set by the branch type being used. The limits for
8131 normal calls are 7,600,000 and 240,000 bytes, respectively for the
8132 PA 2.0 and PA 1.X architectures. Sibcalls are always limited at
8135 Distances are measured from the beginning of functions when using the
8136 @option{-ffunction-sections} option, or when using the @option{-mgas}
8137 and @option{-mno-portable-runtime} options together under HP-UX with
8140 It is normally not desirable to use this option as it will degrade
8141 performance. However, it may be useful in large applications,
8142 particularly when partial linking is used to build the application.
8144 The types of long calls used depends on the capabilities of the
8145 assembler and linker, and the type of code being generated. The
8146 impact on systems that support long absolute calls, and long pic
8147 symbol-difference or pc-relative calls should be relatively small.
8148 However, an indirect call is used on 32-bit ELF systems in pic code
8149 and it is quite long.
8153 @node Intel 960 Options
8154 @subsection Intel 960 Options
8156 These @samp{-m} options are defined for the Intel 960 implementations:
8159 @item -m@var{cpu-type}
8167 Assume the defaults for the machine type @var{cpu-type} for some of
8168 the other options, including instruction scheduling, floating point
8169 support, and addressing modes. The choices for @var{cpu-type} are
8170 @samp{ka}, @samp{kb}, @samp{mc}, @samp{ca}, @samp{cf},
8171 @samp{sa}, and @samp{sb}.
8178 @opindex msoft-float
8179 The @option{-mnumerics} option indicates that the processor does support
8180 floating-point instructions. The @option{-msoft-float} option indicates
8181 that floating-point support should not be assumed.
8183 @item -mleaf-procedures
8184 @itemx -mno-leaf-procedures
8185 @opindex mleaf-procedures
8186 @opindex mno-leaf-procedures
8187 Do (or do not) attempt to alter leaf procedures to be callable with the
8188 @code{bal} instruction as well as @code{call}. This will result in more
8189 efficient code for explicit calls when the @code{bal} instruction can be
8190 substituted by the assembler or linker, but less efficient code in other
8191 cases, such as calls via function pointers, or using a linker that doesn't
8192 support this optimization.
8195 @itemx -mno-tail-call
8197 @opindex mno-tail-call
8198 Do (or do not) make additional attempts (beyond those of the
8199 machine-independent portions of the compiler) to optimize tail-recursive
8200 calls into branches. You may not want to do this because the detection of
8201 cases where this is not valid is not totally complete. The default is
8202 @option{-mno-tail-call}.
8204 @item -mcomplex-addr
8205 @itemx -mno-complex-addr
8206 @opindex mcomplex-addr
8207 @opindex mno-complex-addr
8208 Assume (or do not assume) that the use of a complex addressing mode is a
8209 win on this implementation of the i960. Complex addressing modes may not
8210 be worthwhile on the K-series, but they definitely are on the C-series.
8211 The default is currently @option{-mcomplex-addr} for all processors except
8215 @itemx -mno-code-align
8216 @opindex mcode-align
8217 @opindex mno-code-align
8218 Align code to 8-byte boundaries for faster fetching (or don't bother).
8219 Currently turned on by default for C-series implementations only.
8222 @item -mclean-linkage
8223 @itemx -mno-clean-linkage
8224 @opindex mclean-linkage
8225 @opindex mno-clean-linkage
8226 These options are not fully implemented.
8230 @itemx -mic2.0-compat
8231 @itemx -mic3.0-compat
8233 @opindex mic2.0-compat
8234 @opindex mic3.0-compat
8235 Enable compatibility with iC960 v2.0 or v3.0.
8239 @opindex masm-compat
8241 Enable compatibility with the iC960 assembler.
8243 @item -mstrict-align
8244 @itemx -mno-strict-align
8245 @opindex mstrict-align
8246 @opindex mno-strict-align
8247 Do not permit (do permit) unaligned accesses.
8251 Enable structure-alignment compatibility with Intel's gcc release version
8252 1.3 (based on gcc 1.37). This option implies @option{-mstrict-align}.
8254 @item -mlong-double-64
8255 @opindex mlong-double-64
8256 Implement type @samp{long double} as 64-bit floating point numbers.
8257 Without the option @samp{long double} is implemented by 80-bit
8258 floating point numbers. The only reason we have it because there is
8259 no 128-bit @samp{long double} support in @samp{fp-bit.c} yet. So it
8260 is only useful for people using soft-float targets. Otherwise, we
8261 should recommend against use of it.
8265 @node DEC Alpha Options
8266 @subsection DEC Alpha Options
8268 These @samp{-m} options are defined for the DEC Alpha implementations:
8271 @item -mno-soft-float
8273 @opindex mno-soft-float
8274 @opindex msoft-float
8275 Use (do not use) the hardware floating-point instructions for
8276 floating-point operations. When @option{-msoft-float} is specified,
8277 functions in @file{libgcc.a} will be used to perform floating-point
8278 operations. Unless they are replaced by routines that emulate the
8279 floating-point operations, or compiled in such a way as to call such
8280 emulations routines, these routines will issue floating-point
8281 operations. If you are compiling for an Alpha without floating-point
8282 operations, you must ensure that the library is built so as not to call
8285 Note that Alpha implementations without floating-point operations are
8286 required to have floating-point registers.
8291 @opindex mno-fp-regs
8292 Generate code that uses (does not use) the floating-point register set.
8293 @option{-mno-fp-regs} implies @option{-msoft-float}. If the floating-point
8294 register set is not used, floating point operands are passed in integer
8295 registers as if they were integers and floating-point results are passed
8296 in @code{$0} instead of @code{$f0}. This is a non-standard calling sequence,
8297 so any function with a floating-point argument or return value called by code
8298 compiled with @option{-mno-fp-regs} must also be compiled with that
8301 A typical use of this option is building a kernel that does not use,
8302 and hence need not save and restore, any floating-point registers.
8306 The Alpha architecture implements floating-point hardware optimized for
8307 maximum performance. It is mostly compliant with the IEEE floating
8308 point standard. However, for full compliance, software assistance is
8309 required. This option generates code fully IEEE compliant code
8310 @emph{except} that the @var{inexact-flag} is not maintained (see below).
8311 If this option is turned on, the preprocessor macro @code{_IEEE_FP} is
8312 defined during compilation. The resulting code is less efficient but is
8313 able to correctly support denormalized numbers and exceptional IEEE
8314 values such as not-a-number and plus/minus infinity. Other Alpha
8315 compilers call this option @option{-ieee_with_no_inexact}.
8317 @item -mieee-with-inexact
8318 @opindex mieee-with-inexact
8319 This is like @option{-mieee} except the generated code also maintains
8320 the IEEE @var{inexact-flag}. Turning on this option causes the
8321 generated code to implement fully-compliant IEEE math. In addition to
8322 @code{_IEEE_FP}, @code{_IEEE_FP_EXACT} is defined as a preprocessor
8323 macro. On some Alpha implementations the resulting code may execute
8324 significantly slower than the code generated by default. Since there is
8325 very little code that depends on the @var{inexact-flag}, you should
8326 normally not specify this option. Other Alpha compilers call this
8327 option @option{-ieee_with_inexact}.
8329 @item -mfp-trap-mode=@var{trap-mode}
8330 @opindex mfp-trap-mode
8331 This option controls what floating-point related traps are enabled.
8332 Other Alpha compilers call this option @option{-fptm @var{trap-mode}}.
8333 The trap mode can be set to one of four values:
8337 This is the default (normal) setting. The only traps that are enabled
8338 are the ones that cannot be disabled in software (e.g., division by zero
8342 In addition to the traps enabled by @samp{n}, underflow traps are enabled
8346 Like @samp{su}, but the instructions are marked to be safe for software
8347 completion (see Alpha architecture manual for details).
8350 Like @samp{su}, but inexact traps are enabled as well.
8353 @item -mfp-rounding-mode=@var{rounding-mode}
8354 @opindex mfp-rounding-mode
8355 Selects the IEEE rounding mode. Other Alpha compilers call this option
8356 @option{-fprm @var{rounding-mode}}. The @var{rounding-mode} can be one
8361 Normal IEEE rounding mode. Floating point numbers are rounded towards
8362 the nearest machine number or towards the even machine number in case
8366 Round towards minus infinity.
8369 Chopped rounding mode. Floating point numbers are rounded towards zero.
8372 Dynamic rounding mode. A field in the floating point control register
8373 (@var{fpcr}, see Alpha architecture reference manual) controls the
8374 rounding mode in effect. The C library initializes this register for
8375 rounding towards plus infinity. Thus, unless your program modifies the
8376 @var{fpcr}, @samp{d} corresponds to round towards plus infinity.
8379 @item -mtrap-precision=@var{trap-precision}
8380 @opindex mtrap-precision
8381 In the Alpha architecture, floating point traps are imprecise. This
8382 means without software assistance it is impossible to recover from a
8383 floating trap and program execution normally needs to be terminated.
8384 GCC can generate code that can assist operating system trap handlers
8385 in determining the exact location that caused a floating point trap.
8386 Depending on the requirements of an application, different levels of
8387 precisions can be selected:
8391 Program precision. This option is the default and means a trap handler
8392 can only identify which program caused a floating point exception.
8395 Function precision. The trap handler can determine the function that
8396 caused a floating point exception.
8399 Instruction precision. The trap handler can determine the exact
8400 instruction that caused a floating point exception.
8403 Other Alpha compilers provide the equivalent options called
8404 @option{-scope_safe} and @option{-resumption_safe}.
8406 @item -mieee-conformant
8407 @opindex mieee-conformant
8408 This option marks the generated code as IEEE conformant. You must not
8409 use this option unless you also specify @option{-mtrap-precision=i} and either
8410 @option{-mfp-trap-mode=su} or @option{-mfp-trap-mode=sui}. Its only effect
8411 is to emit the line @samp{.eflag 48} in the function prologue of the
8412 generated assembly file. Under DEC Unix, this has the effect that
8413 IEEE-conformant math library routines will be linked in.
8415 @item -mbuild-constants
8416 @opindex mbuild-constants
8417 Normally GCC examines a 32- or 64-bit integer constant to
8418 see if it can construct it from smaller constants in two or three
8419 instructions. If it cannot, it will output the constant as a literal and
8420 generate code to load it from the data segment at runtime.
8422 Use this option to require GCC to construct @emph{all} integer constants
8423 using code, even if it takes more instructions (the maximum is six).
8425 You would typically use this option to build a shared library dynamic
8426 loader. Itself a shared library, it must relocate itself in memory
8427 before it can find the variables and constants in its own data segment.
8433 Select whether to generate code to be assembled by the vendor-supplied
8434 assembler (@option{-malpha-as}) or by the GNU assembler @option{-mgas}.
8452 Indicate whether GCC should generate code to use the optional BWX,
8453 CIX, FIX and MAX instruction sets. The default is to use the instruction
8454 sets supported by the CPU type specified via @option{-mcpu=} option or that
8455 of the CPU on which GCC was built if none was specified.
8460 @opindex mfloat-ieee
8461 Generate code that uses (does not use) VAX F and G floating point
8462 arithmetic instead of IEEE single and double precision.
8464 @item -mexplicit-relocs
8465 @itemx -mno-explicit-relocs
8466 @opindex mexplicit-relocs
8467 @opindex mno-explicit-relocs
8468 Older Alpha assemblers provided no way to generate symbol relocations
8469 except via assembler macros. Use of these macros does not allow
8470 optimial instruction scheduling. GNU binutils as of version 2.12
8471 supports a new syntax that allows the compiler to explicitly mark
8472 which relocations should apply to which instructions. This option
8473 is mostly useful for debugging, as GCC detects the capabilities of
8474 the assembler when it is built and sets the default accordingly.
8478 @opindex msmall-data
8479 @opindex mlarge-data
8480 When @option{-mexplicit-relocs} is in effect, static data is
8481 accessed via @dfn{gp-relative} relocations. When @option{-msmall-data}
8482 is used, objects 8 bytes long or smaller are placed in a @dfn{small data area}
8483 (the @code{.sdata} and @code{.sbss} sections) and are accessed via
8484 16-bit relocations off of the @code{$gp} register. This limits the
8485 size of the small data area to 64KB, but allows the variables to be
8486 directly accessed via a single instruction.
8488 The default is @option{-mlarge-data}. With this option the data area
8489 is limited to just below 2GB. Programs that require more than 2GB of
8490 data must use @code{malloc} or @code{mmap} to allocate the data in the
8491 heap instead of in the program's data segment.
8493 When generating code for shared libraries, @option{-fpic} implies
8494 @option{-msmall-data} and @option{-fPIC} implies @option{-mlarge-data}.
8496 @item -mcpu=@var{cpu_type}
8498 Set the instruction set and instruction scheduling parameters for
8499 machine type @var{cpu_type}. You can specify either the @samp{EV}
8500 style name or the corresponding chip number. GCC supports scheduling
8501 parameters for the EV4, EV5 and EV6 family of processors and will
8502 choose the default values for the instruction set from the processor
8503 you specify. If you do not specify a processor type, GCC will default
8504 to the processor on which the compiler was built.
8506 Supported values for @var{cpu_type} are
8512 Schedules as an EV4 and has no instruction set extensions.
8516 Schedules as an EV5 and has no instruction set extensions.
8520 Schedules as an EV5 and supports the BWX extension.
8525 Schedules as an EV5 and supports the BWX and MAX extensions.
8529 Schedules as an EV6 and supports the BWX, FIX, and MAX extensions.
8533 Schedules as an EV6 and supports the BWX, CIX, FIX, and MAX extensions.
8536 @item -mtune=@var{cpu_type}
8538 Set only the instruction scheduling parameters for machine type
8539 @var{cpu_type}. The instruction set is not changed.
8541 @item -mmemory-latency=@var{time}
8542 @opindex mmemory-latency
8543 Sets the latency the scheduler should assume for typical memory
8544 references as seen by the application. This number is highly
8545 dependent on the memory access patterns used by the application
8546 and the size of the external cache on the machine.
8548 Valid options for @var{time} are
8552 A decimal number representing clock cycles.
8558 The compiler contains estimates of the number of clock cycles for
8559 ``typical'' EV4 & EV5 hardware for the Level 1, 2 & 3 caches
8560 (also called Dcache, Scache, and Bcache), as well as to main memory.
8561 Note that L3 is only valid for EV5.
8566 @node DEC Alpha/VMS Options
8567 @subsection DEC Alpha/VMS Options
8569 These @samp{-m} options are defined for the DEC Alpha/VMS implementations:
8572 @item -mvms-return-codes
8573 @opindex mvms-return-codes
8574 Return VMS condition codes from main. The default is to return POSIX
8575 style condition (e.g.@ error) codes.
8578 @node Clipper Options
8579 @subsection Clipper Options
8581 These @samp{-m} options are defined for the Clipper implementations:
8586 Produce code for a C300 Clipper processor. This is the default.
8590 Produce code for a C400 Clipper processor, i.e.@: use floating point
8594 @node H8/300 Options
8595 @subsection H8/300 Options
8597 These @samp{-m} options are defined for the H8/300 implementations:
8602 Shorten some address references at link time, when possible; uses the
8603 linker option @option{-relax}. @xref{H8/300,, @code{ld} and the H8/300,
8604 ld.info, Using ld}, for a fuller description.
8608 Generate code for the H8/300H@.
8612 Generate code for the H8S@.
8616 Generate code for the H8S/2600. This switch must be used with @option{-ms}.
8620 Make @code{int} data 32 bits by default.
8624 On the H8/300H and H8S, use the same alignment rules as for the H8/300.
8625 The default for the H8/300H and H8S is to align longs and floats on 4
8627 @option{-malign-300} causes them to be aligned on 2 byte boundaries.
8628 This option has no effect on the H8/300.
8632 @subsection SH Options
8634 These @samp{-m} options are defined for the SH implementations:
8639 Generate code for the SH1.
8643 Generate code for the SH2.
8647 Generate code for the SH3.
8651 Generate code for the SH3e.
8655 Generate code for the SH4 without a floating-point unit.
8657 @item -m4-single-only
8658 @opindex m4-single-only
8659 Generate code for the SH4 with a floating-point unit that only
8660 supports single-precision arithmetic.
8664 Generate code for the SH4 assuming the floating-point unit is in
8665 single-precision mode by default.
8669 Generate code for the SH4.
8673 Compile code for the processor in big endian mode.
8677 Compile code for the processor in little endian mode.
8681 Align doubles at 64-bit boundaries. Note that this changes the calling
8682 conventions, and thus some functions from the standard C library will
8683 not work unless you recompile it first with @option{-mdalign}.
8687 Shorten some address references at link time, when possible; uses the
8688 linker option @option{-relax}.
8692 Use 32-bit offsets in @code{switch} tables. The default is to use
8697 Enable the use of the instruction @code{fmovd}.
8701 Comply with the calling conventions defined by Hitachi.
8705 Mark the @code{MAC} register as call-clobbered, even if
8706 @option{-mhitachi} is given.
8710 Increase IEEE-compliance of floating-point code.
8714 Dump instruction size and location in the assembly code.
8718 This option is deprecated. It pads structures to multiple of 4 bytes,
8719 which is incompatible with the SH ABI@.
8723 Optimize for space instead of speed. Implied by @option{-Os}.
8727 When generating position-independent code, emit function calls using
8728 the Global Offset Table instead of the Procedure Linkage Table.
8732 Generate a library function call to invalidate instruction cache
8733 entries, after fixing up a trampoline. This library function call
8734 doesn't assume it can write to the whole memory address space. This
8735 is the default when the target is @code{sh-*-linux*}.
8738 @node System V Options
8739 @subsection Options for System V
8741 These additional options are available on System V Release 4 for
8742 compatibility with other compilers on those systems:
8747 Create a shared object.
8748 It is recommended that @option{-symbolic} or @option{-shared} be used instead.
8752 Identify the versions of each tool used by the compiler, in a
8753 @code{.ident} assembler directive in the output.
8757 Refrain from adding @code{.ident} directives to the output file (this is
8760 @item -YP,@var{dirs}
8762 Search the directories @var{dirs}, and no others, for libraries
8763 specified with @option{-l}.
8767 Look in the directory @var{dir} to find the M4 preprocessor.
8768 The assembler uses this option.
8769 @c This is supposed to go with a -Yd for predefined M4 macro files, but
8770 @c the generic assembler that comes with Solaris takes just -Ym.
8773 @node TMS320C3x/C4x Options
8774 @subsection TMS320C3x/C4x Options
8775 @cindex TMS320C3x/C4x Options
8777 These @samp{-m} options are defined for TMS320C3x/C4x implementations:
8781 @item -mcpu=@var{cpu_type}
8783 Set the instruction set, register set, and instruction scheduling
8784 parameters for machine type @var{cpu_type}. Supported values for
8785 @var{cpu_type} are @samp{c30}, @samp{c31}, @samp{c32}, @samp{c40}, and
8786 @samp{c44}. The default is @samp{c40} to generate code for the
8791 @itemx -msmall-memory
8793 @opindex mbig-memory
8795 @opindex msmall-memory
8797 Generates code for the big or small memory model. The small memory
8798 model assumed that all data fits into one 64K word page. At run-time
8799 the data page (DP) register must be set to point to the 64K page
8800 containing the .bss and .data program sections. The big memory model is
8801 the default and requires reloading of the DP register for every direct
8808 Allow (disallow) allocation of general integer operands into the block
8815 Enable (disable) generation of code using decrement and branch,
8816 DBcond(D), instructions. This is enabled by default for the C4x. To be
8817 on the safe side, this is disabled for the C3x, since the maximum
8818 iteration count on the C3x is @math{2^{23} + 1} (but who iterates loops more than
8819 @math{2^{23}} times on the C3x?). Note that GCC will try to reverse a loop so
8820 that it can utilize the decrement and branch instruction, but will give
8821 up if there is more than one memory reference in the loop. Thus a loop
8822 where the loop counter is decremented can generate slightly more
8823 efficient code, in cases where the RPTB instruction cannot be utilized.
8825 @item -mdp-isr-reload
8827 @opindex mdp-isr-reload
8829 Force the DP register to be saved on entry to an interrupt service
8830 routine (ISR), reloaded to point to the data section, and restored on
8831 exit from the ISR@. This should not be required unless someone has
8832 violated the small memory model by modifying the DP register, say within
8839 For the C3x use the 24-bit MPYI instruction for integer multiplies
8840 instead of a library call to guarantee 32-bit results. Note that if one
8841 of the operands is a constant, then the multiplication will be performed
8842 using shifts and adds. If the @option{-mmpyi} option is not specified for the C3x,
8843 then squaring operations are performed inline instead of a library call.
8846 @itemx -mno-fast-fix
8848 @opindex mno-fast-fix
8849 The C3x/C4x FIX instruction to convert a floating point value to an
8850 integer value chooses the nearest integer less than or equal to the
8851 floating point value rather than to the nearest integer. Thus if the
8852 floating point number is negative, the result will be incorrectly
8853 truncated an additional code is necessary to detect and correct this
8854 case. This option can be used to disable generation of the additional
8855 code required to correct the result.
8861 Enable (disable) generation of repeat block sequences using the RPTB
8862 instruction for zero overhead looping. The RPTB construct is only used
8863 for innermost loops that do not call functions or jump across the loop
8864 boundaries. There is no advantage having nested RPTB loops due to the
8865 overhead required to save and restore the RC, RS, and RE registers.
8866 This is enabled by default with @option{-O2}.
8868 @item -mrpts=@var{count}
8872 Enable (disable) the use of the single instruction repeat instruction
8873 RPTS@. If a repeat block contains a single instruction, and the loop
8874 count can be guaranteed to be less than the value @var{count}, GCC will
8875 emit a RPTS instruction instead of a RPTB@. If no value is specified,
8876 then a RPTS will be emitted even if the loop count cannot be determined
8877 at compile time. Note that the repeated instruction following RPTS does
8878 not have to be reloaded from memory each iteration, thus freeing up the
8879 CPU buses for operands. However, since interrupts are blocked by this
8880 instruction, it is disabled by default.
8882 @item -mloop-unsigned
8883 @itemx -mno-loop-unsigned
8884 @opindex mloop-unsigned
8885 @opindex mno-loop-unsigned
8886 The maximum iteration count when using RPTS and RPTB (and DB on the C40)
8887 is @math{2^{31} + 1} since these instructions test if the iteration count is
8888 negative to terminate the loop. If the iteration count is unsigned
8889 there is a possibility than the @math{2^{31} + 1} maximum iteration count may be
8890 exceeded. This switch allows an unsigned iteration count.
8894 Try to emit an assembler syntax that the TI assembler (asm30) is happy
8895 with. This also enforces compatibility with the API employed by the TI
8896 C3x C compiler. For example, long doubles are passed as structures
8897 rather than in floating point registers.
8903 Generate code that uses registers (stack) for passing arguments to functions.
8904 By default, arguments are passed in registers where possible rather
8905 than by pushing arguments on to the stack.
8907 @item -mparallel-insns
8908 @itemx -mno-parallel-insns
8909 @opindex mparallel-insns
8910 @opindex mno-parallel-insns
8911 Allow the generation of parallel instructions. This is enabled by
8912 default with @option{-O2}.
8914 @item -mparallel-mpy
8915 @itemx -mno-parallel-mpy
8916 @opindex mparallel-mpy
8917 @opindex mno-parallel-mpy
8918 Allow the generation of MPY||ADD and MPY||SUB parallel instructions,
8919 provided @option{-mparallel-insns} is also specified. These instructions have
8920 tight register constraints which can pessimize the code generation
8926 @subsection V850 Options
8927 @cindex V850 Options
8929 These @samp{-m} options are defined for V850 implementations:
8933 @itemx -mno-long-calls
8934 @opindex mlong-calls
8935 @opindex mno-long-calls
8936 Treat all calls as being far away (near). If calls are assumed to be
8937 far away, the compiler will always load the functions address up into a
8938 register, and call indirect through the pointer.
8944 Do not optimize (do optimize) basic blocks that use the same index
8945 pointer 4 or more times to copy pointer into the @code{ep} register, and
8946 use the shorter @code{sld} and @code{sst} instructions. The @option{-mep}
8947 option is on by default if you optimize.
8949 @item -mno-prolog-function
8950 @itemx -mprolog-function
8951 @opindex mno-prolog-function
8952 @opindex mprolog-function
8953 Do not use (do use) external functions to save and restore registers at
8954 the prolog and epilog of a function. The external functions are slower,
8955 but use less code space if more than one function saves the same number
8956 of registers. The @option{-mprolog-function} option is on by default if
8961 Try to make the code as small as possible. At present, this just turns
8962 on the @option{-mep} and @option{-mprolog-function} options.
8966 Put static or global variables whose size is @var{n} bytes or less into
8967 the tiny data area that register @code{ep} points to. The tiny data
8968 area can hold up to 256 bytes in total (128 bytes for byte references).
8972 Put static or global variables whose size is @var{n} bytes or less into
8973 the small data area that register @code{gp} points to. The small data
8974 area can hold up to 64 kilobytes.
8978 Put static or global variables whose size is @var{n} bytes or less into
8979 the first 32 kilobytes of memory.
8983 Specify that the target processor is the V850.
8986 @opindex mbig-switch
8987 Generate code suitable for big switch tables. Use this option only if
8988 the assembler/linker complain about out of range branches within a switch
8993 This option will cause r2 and r5 to be used in the code generated by
8994 the compiler. This setting is the default.
8997 @opindex -mno-app-regs
8998 This option will cause r2 and r5 to be treated as fixed registers.
9002 Specify that the target processor is the V850E. The preprocessor
9003 constant @samp{__v850e__} will be defined if this option is used.
9005 If neither @option{-mv850} nor @option{-mv850e} are defined
9006 then a default target processor will be chosen and the relevant
9007 @samp{__v850*__} preprocessor constant will be defined.
9009 The preprocessor constants @samp{__v850} and @samp{__v851__} are always
9010 defined, regardless of which processor variant is the target.
9012 @item -mdisable-callt
9013 @opindex -mdisable-callt
9014 This option will suppress generation of the CALLT instruction for the
9015 v850e flavors of the v850 architecture. The default is
9016 @option{-mno-disable-callt} which allows the CALLT instruction to be used.
9021 @subsection ARC Options
9024 These options are defined for ARC implementations:
9029 Compile code for little endian mode. This is the default.
9033 Compile code for big endian mode.
9036 @opindex mmangle-cpu
9037 Prepend the name of the cpu to all public symbol names.
9038 In multiple-processor systems, there are many ARC variants with different
9039 instruction and register set characteristics. This flag prevents code
9040 compiled for one cpu to be linked with code compiled for another.
9041 No facility exists for handling variants that are ``almost identical''.
9042 This is an all or nothing option.
9044 @item -mcpu=@var{cpu}
9046 Compile code for ARC variant @var{cpu}.
9047 Which variants are supported depend on the configuration.
9048 All variants support @option{-mcpu=base}, this is the default.
9050 @item -mtext=@var{text-section}
9051 @itemx -mdata=@var{data-section}
9052 @itemx -mrodata=@var{readonly-data-section}
9056 Put functions, data, and readonly data in @var{text-section},
9057 @var{data-section}, and @var{readonly-data-section} respectively
9058 by default. This can be overridden with the @code{section} attribute.
9059 @xref{Variable Attributes}.
9064 @subsection NS32K Options
9065 @cindex NS32K options
9067 These are the @samp{-m} options defined for the 32000 series. The default
9068 values for these options depends on which style of 32000 was selected when
9069 the compiler was configured; the defaults for the most common choices are
9077 Generate output for a 32032. This is the default
9078 when the compiler is configured for 32032 and 32016 based systems.
9084 Generate output for a 32332. This is the default
9085 when the compiler is configured for 32332-based systems.
9091 Generate output for a 32532. This is the default
9092 when the compiler is configured for 32532-based systems.
9096 Generate output containing 32081 instructions for floating point.
9097 This is the default for all systems.
9101 Generate output containing 32381 instructions for floating point. This
9102 also implies @option{-m32081}. The 32381 is only compatible with the 32332
9103 and 32532 cpus. This is the default for the pc532-netbsd configuration.
9107 Try and generate multiply-add floating point instructions @code{polyF}
9108 and @code{dotF}. This option is only available if the @option{-m32381}
9109 option is in effect. Using these instructions requires changes to
9110 register allocation which generally has a negative impact on
9111 performance. This option should only be enabled when compiling code
9112 particularly likely to make heavy use of multiply-add instructions.
9115 @opindex mnomulti-add
9116 Do not try and generate multiply-add floating point instructions
9117 @code{polyF} and @code{dotF}. This is the default on all platforms.
9120 @opindex msoft-float
9121 Generate output containing library calls for floating point.
9122 @strong{Warning:} the requisite libraries may not be available.
9124 @item -mieee-compare
9125 @itemx -mno-ieee-compare
9126 @opindex mieee-compare
9127 @opindex mno-ieee-compare
9128 Control whether or not the compiler uses IEEE floating point
9129 comparisons. These handle correctly the case where the result of a
9130 comparison is unordered.
9131 @strong{Warning:} the requisite kernel support may not be available.
9134 @opindex mnobitfield
9135 Do not use the bit-field instructions. On some machines it is faster to
9136 use shifting and masking operations. This is the default for the pc532.
9140 Do use the bit-field instructions. This is the default for all platforms
9145 Use a different function-calling convention, in which functions
9146 that take a fixed number of arguments return pop their
9147 arguments on return with the @code{ret} instruction.
9149 This calling convention is incompatible with the one normally
9150 used on Unix, so you cannot use it if you need to call libraries
9151 compiled with the Unix compiler.
9153 Also, you must provide function prototypes for all functions that
9154 take variable numbers of arguments (including @code{printf});
9155 otherwise incorrect code will be generated for calls to those
9158 In addition, seriously incorrect code will result if you call a
9159 function with too many arguments. (Normally, extra arguments are
9160 harmlessly ignored.)
9162 This option takes its name from the 680x0 @code{rtd} instruction.
9167 Use a different function-calling convention where the first two arguments
9168 are passed in registers.
9170 This calling convention is incompatible with the one normally
9171 used on Unix, so you cannot use it if you need to call libraries
9172 compiled with the Unix compiler.
9175 @opindex mnoregparam
9176 Do not pass any arguments in registers. This is the default for all
9181 It is OK to use the sb as an index register which is always loaded with
9182 zero. This is the default for the pc532-netbsd target.
9186 The sb register is not available for use or has not been initialized to
9187 zero by the run time system. This is the default for all targets except
9188 the pc532-netbsd. It is also implied whenever @option{-mhimem} or
9189 @option{-fpic} is set.
9193 Many ns32000 series addressing modes use displacements of up to 512MB@.
9194 If an address is above 512MB then displacements from zero can not be used.
9195 This option causes code to be generated which can be loaded above 512MB@.
9196 This may be useful for operating systems or ROM code.
9200 Assume code will be loaded in the first 512MB of virtual address space.
9201 This is the default for all platforms.
9207 @subsection AVR Options
9210 These options are defined for AVR implementations:
9213 @item -mmcu=@var{mcu}
9215 Specify ATMEL AVR instruction set or MCU type.
9217 Instruction set avr1 is for the minimal AVR core, not supported by the C
9218 compiler, only for assembler programs (MCU types: at90s1200, attiny10,
9219 attiny11, attiny12, attiny15, attiny28).
9221 Instruction set avr2 (default) is for the classic AVR core with up to
9222 8K program memory space (MCU types: at90s2313, at90s2323, attiny22,
9223 at90s2333, at90s2343, at90s4414, at90s4433, at90s4434, at90s8515,
9224 at90c8534, at90s8535).
9226 Instruction set avr3 is for the classic AVR core with up to 128K program
9227 memory space (MCU types: atmega103, atmega603, at43usb320, at76c711).
9229 Instruction set avr4 is for the enhanced AVR core with up to 8K program
9230 memory space (MCU types: atmega8, atmega83, atmega85).
9232 Instruction set avr5 is for the enhanced AVR core with up to 128K program
9233 memory space (MCU types: atmega16, atmega161, atmega163, atmega32, atmega323,
9234 atmega64, atmega128, at43usb355, at94k).
9238 Output instruction sizes to the asm file.
9240 @item -minit-stack=@var{N}
9241 @opindex minit-stack
9242 Specify the initial stack address, which may be a symbol or numeric value,
9243 @samp{__stack} is the default.
9245 @item -mno-interrupts
9246 @opindex mno-interrupts
9247 Generated code is not compatible with hardware interrupts.
9248 Code size will be smaller.
9250 @item -mcall-prologues
9251 @opindex mcall-prologues
9252 Functions prologues/epilogues expanded as call to appropriate
9253 subroutines. Code size will be smaller.
9255 @item -mno-tablejump
9256 @opindex mno-tablejump
9257 Do not generate tablejump insns which sometimes increase code size.
9260 @opindex mtiny-stack
9261 Change only the low 8 bits of the stack pointer.
9265 @subsection MCore Options
9266 @cindex MCore options
9268 These are the @samp{-m} options defined for the Motorola M*Core
9278 @opindex mno-hardlit
9279 Inline constants into the code stream if it can be done in two
9280 instructions or less.
9288 Use the divide instruction. (Enabled by default).
9290 @item -mrelax-immediate
9291 @itemx -mrelax-immediate
9292 @itemx -mno-relax-immediate
9293 @opindex mrelax-immediate
9294 @opindex mrelax-immediate
9295 @opindex mno-relax-immediate
9296 Allow arbitrary sized immediates in bit operations.
9298 @item -mwide-bitfields
9299 @itemx -mwide-bitfields
9300 @itemx -mno-wide-bitfields
9301 @opindex mwide-bitfields
9302 @opindex mwide-bitfields
9303 @opindex mno-wide-bitfields
9304 Always treat bit-fields as int-sized.
9306 @item -m4byte-functions
9307 @itemx -m4byte-functions
9308 @itemx -mno-4byte-functions
9309 @opindex m4byte-functions
9310 @opindex m4byte-functions
9311 @opindex mno-4byte-functions
9312 Force all functions to be aligned to a four byte boundary.
9314 @item -mcallgraph-data
9315 @itemx -mcallgraph-data
9316 @itemx -mno-callgraph-data
9317 @opindex mcallgraph-data
9318 @opindex mcallgraph-data
9319 @opindex mno-callgraph-data
9320 Emit callgraph information.
9324 @itemx -mno-slow-bytes
9325 @opindex mslow-bytes
9326 @opindex mslow-bytes
9327 @opindex mno-slow-bytes
9328 Prefer word access when reading byte quantities.
9330 @item -mlittle-endian
9331 @itemx -mlittle-endian
9333 @opindex mlittle-endian
9334 @opindex mlittle-endian
9335 @opindex mbig-endian
9336 Generate code for a little endian target.
9344 Generate code for the 210 processor.
9348 @subsection IA-64 Options
9349 @cindex IA-64 Options
9351 These are the @samp{-m} options defined for the Intel IA-64 architecture.
9355 @opindex mbig-endian
9356 Generate code for a big endian target. This is the default for HP-UX@.
9358 @item -mlittle-endian
9359 @opindex mlittle-endian
9360 Generate code for a little endian target. This is the default for AIX5
9367 Generate (or don't) code for the GNU assembler. This is the default.
9368 @c Also, this is the default if the configure option @option{--with-gnu-as}
9375 Generate (or don't) code for the GNU linker. This is the default.
9376 @c Also, this is the default if the configure option @option{--with-gnu-ld}
9381 Generate code that does not use a global pointer register. The result
9382 is not position independent code, and violates the IA-64 ABI@.
9384 @item -mvolatile-asm-stop
9385 @itemx -mno-volatile-asm-stop
9386 @opindex mvolatile-asm-stop
9387 @opindex mno-volatile-asm-stop
9388 Generate (or don't) a stop bit immediately before and after volatile asm
9393 Generate code that works around Itanium B step errata.
9395 @item -mregister-names
9396 @itemx -mno-register-names
9397 @opindex mregister-names
9398 @opindex mno-register-names
9399 Generate (or don't) @samp{in}, @samp{loc}, and @samp{out} register names for
9400 the stacked registers. This may make assembler output more readable.
9406 Disable (or enable) optimizations that use the small data section. This may
9407 be useful for working around optimizer bugs.
9410 @opindex mconstant-gp
9411 Generate code that uses a single constant global pointer value. This is
9412 useful when compiling kernel code.
9416 Generate code that is self-relocatable. This implies @option{-mconstant-gp}.
9417 This is useful when compiling firmware code.
9419 @item -minline-float-divide-min-latency
9420 @opindex minline-float-divide-min-latency
9421 Generate code for inline divides of floating point values
9422 using the minimum latency algorithm.
9424 @item -minline-float-divide-max-throughput
9425 @opindex minline-float-divide-max-throughput
9426 Generate code for inline divides of floating point values
9427 using the maximum throughput algorithm.
9429 @item -minline-int-divide-min-latency
9430 @opindex minline-int-divide-min-latency
9431 Generate code for inline divides of integer values
9432 using the minimum latency algorithm.
9434 @item -minline-int-divide-max-throughput
9435 @opindex minline-int-divide-max-throughput
9436 Generate code for inline divides of integer values
9437 using the maximum throughput algorithm.
9439 @item -mno-dwarf2-asm
9441 @opindex mno-dwarf2-asm
9442 @opindex mdwarf2-asm
9443 Don't (or do) generate assembler code for the DWARF2 line number debugging
9444 info. This may be useful when not using the GNU assembler.
9446 @item -mfixed-range=@var{register-range}
9447 @opindex mfixed-range
9448 Generate code treating the given register range as fixed registers.
9449 A fixed register is one that the register allocator can not use. This is
9450 useful when compiling kernel code. A register range is specified as
9451 two registers separated by a dash. Multiple register ranges can be
9452 specified separated by a comma.
9456 @subsection D30V Options
9457 @cindex D30V Options
9459 These @samp{-m} options are defined for D30V implementations:
9464 Link the @samp{.text}, @samp{.data}, @samp{.bss}, @samp{.strings},
9465 @samp{.rodata}, @samp{.rodata1}, @samp{.data1} sections into external
9466 memory, which starts at location @code{0x80000000}.
9470 Same as the @option{-mextmem} switch.
9474 Link the @samp{.text} section into onchip text memory, which starts at
9475 location @code{0x0}. Also link @samp{.data}, @samp{.bss},
9476 @samp{.strings}, @samp{.rodata}, @samp{.rodata1}, @samp{.data1} sections
9477 into onchip data memory, which starts at location @code{0x20000000}.
9479 @item -mno-asm-optimize
9480 @itemx -masm-optimize
9481 @opindex mno-asm-optimize
9482 @opindex masm-optimize
9483 Disable (enable) passing @option{-O} to the assembler when optimizing.
9484 The assembler uses the @option{-O} option to automatically parallelize
9485 adjacent short instructions where possible.
9487 @item -mbranch-cost=@var{n}
9488 @opindex mbranch-cost
9489 Increase the internal costs of branches to @var{n}. Higher costs means
9490 that the compiler will issue more instructions to avoid doing a branch.
9493 @item -mcond-exec=@var{n}
9495 Specify the maximum number of conditionally executed instructions that
9496 replace a branch. The default is 4.
9499 @node S/390 and zSeries Options
9500 @subsection S/390 and zSeries Options
9501 @cindex S/390 and zSeries Options
9503 These are the @samp{-m} options defined for the S/390 and zSeries architecture.
9508 @opindex mhard-float
9509 @opindex msoft-float
9510 Use (do not use) the hardware floating-point instructions and registers
9511 for floating-point operations. When @option{-msoft-float} is specified,
9512 functions in @file{libgcc.a} will be used to perform floating-point
9513 operations. When @option{-mhard-float} is specified, the compiler
9514 generates IEEE floating-point instructions. This is the default.
9517 @itemx -mno-backchain
9519 @opindex mno-backchain
9520 Generate (or do not generate) code which maintains an explicit
9521 backchain within the stack frame that points to the caller's frame.
9522 This is currently needed to allow debugging. The default is to
9523 generate the backchain.
9526 @itemx -mno-small-exec
9527 @opindex msmall-exec
9528 @opindex mno-small-exec
9529 Generate (or do not generate) code using the @code{bras} instruction
9530 to do subroutine calls.
9531 This only works reliably if the total executable size does not
9532 exceed 64k. The default is to use the @code{basr} instruction instead,
9533 which does not have this limitation.
9539 When @option{-m31} is specified, generate code compliant to the
9540 Linux for S/390 ABI@. When @option{-m64} is specified, generate
9541 code compliant to the Linux for zSeries ABI@. This allows GCC in
9542 particular to generate 64-bit instructions. For the @samp{s390}
9543 targets, the default is @option{-m31}, while the @samp{s390x}
9544 targets default to @option{-m64}.
9550 Generate (or do not generate) code using the @code{mvcle} instruction
9551 to perform block moves. When @option{-mno-mvcle} is specifed,
9552 use a @code{mvc} loop instead. This is the default.
9558 Print (or do not print) additional debug information when compiling.
9559 The default is to not print debug information.
9564 @subsection CRIS Options
9565 @cindex CRIS Options
9567 These options are defined specifically for the CRIS ports.
9570 @item -march=@var{architecture-type}
9571 @itemx -mcpu=@var{architecture-type}
9574 Generate code for the specified architecture. The choices for
9575 @var{architecture-type} are @samp{v3}, @samp{v8} and @samp{v10} for
9576 respectively ETRAX@w{ }4, ETRAX@w{ }100, and ETRAX@w{ }100@w{ }LX.
9577 Default is @samp{v0} except for cris-axis-linux-gnu, where the default is
9580 @item -mtune=@var{architecture-type}
9582 Tune to @var{architecture-type} everything applicable about the generated
9583 code, except for the ABI and the set of available instructions. The
9584 choices for @var{architecture-type} are the same as for
9585 @option{-march=@var{architecture-type}}.
9587 @item -mmax-stack-frame=@var{n}
9588 @opindex mmax-stack-frame
9589 Warn when the stack frame of a function exceeds @var{n} bytes.
9591 @item -melinux-stacksize=@var{n}
9592 @opindex melinux-stacksize
9593 Only available with the @samp{cris-axis-aout} target. Arranges for
9594 indications in the program to the kernel loader that the stack of the
9595 program should be set to @var{n} bytes.
9601 The options @option{-metrax4} and @option{-metrax100} are synonyms for
9602 @option{-march=v3} and @option{-march=v8} respectively.
9606 Enable CRIS-specific verbose debug-related information in the assembly
9607 code. This option also has the effect to turn off the @samp{#NO_APP}
9608 formatted-code indicator to the assembler at the beginning of the
9613 Do not use condition-code results from previous instruction; always emit
9614 compare and test instructions before use of condition codes.
9616 @item -mno-side-effects
9617 @opindex mno-side-effects
9618 Do not emit instructions with side-effects in addressing modes other than
9622 @itemx -mno-stack-align
9624 @itemx -mno-data-align
9625 @itemx -mconst-align
9626 @itemx -mno-const-align
9627 @opindex mstack-align
9628 @opindex mno-stack-align
9629 @opindex mdata-align
9630 @opindex mno-data-align
9631 @opindex mconst-align
9632 @opindex mno-const-align
9633 These options (no-options) arranges (eliminate arrangements) for the
9634 stack-frame, individual data and constants to be aligned for the maximum
9635 single data access size for the chosen CPU model. The default is to
9636 arrange for 32-bit alignment. ABI details such as structure layout are
9637 not affected by these options.
9645 Similar to the stack- data- and const-align options above, these options
9646 arrange for stack-frame, writable data and constants to all be 32-bit,
9647 16-bit or 8-bit aligned. The default is 32-bit alignment.
9649 @item -mno-prologue-epilogue
9650 @itemx -mprologue-epilogue
9651 @opindex mno-prologue-epilogue
9652 @opindex mprologue-epilogue
9653 With @option{-mno-prologue-epilogue}, the normal function prologue and
9654 epilogue that sets up the stack-frame are omitted and no return
9655 instructions or return sequences are generated in the code. Use this
9656 option only together with visual inspection of the compiled code: no
9657 warnings or errors are generated when call-saved registers must be saved,
9658 or storage for local variable needs to be allocated.
9664 With @option{-fpic} and @option{-fPIC}, don't generate (do generate)
9665 instruction sequences that load addresses for functions from the PLT part
9666 of the GOT rather than (traditional on other architectures) calls to the
9667 PLT. The default is @option{-mgotplt}.
9671 Legacy no-op option only recognized with the cris-axis-aout target.
9675 Legacy no-op option only recognized with the cris-axis-elf and
9676 cris-axis-linux-gnu targets.
9680 Only recognized with the cris-axis-aout target, where it selects a
9681 GNU/linux-like multilib, include files and instruction set for
9686 Legacy no-op option only recognized with the cris-axis-linux-gnu target.
9690 This option, recognized for the cris-axis-aout and cris-axis-elf arranges
9691 to link with input-output functions from a simulator library. Code,
9692 initialized data and zero-initialized data are allocated consecutively.
9696 Like @option{-sim}, but pass linker options to locate initialized data at
9697 0x40000000 and zero-initialized data at 0x80000000.
9701 @subsection MMIX Options
9702 @cindex MMIX Options
9704 These options are defined for the MMIX:
9708 @itemx -mno-libfuncs
9710 @opindex mno-libfuncs
9711 Specify that intrinsic library functions are being compiled, passing all
9712 values in registers, no matter the size.
9717 @opindex mno-epsilon
9718 Generate floating-point comparison instructions that compare with respect
9719 to the @code{rE} epsilon register.
9721 @item -mabi=mmixware
9723 @opindex mabi-mmixware
9725 Generate code that passes function parameters and return values that (in
9726 the called function) are seen as registers @code{$0} and up, as opposed to
9727 the GNU ABI which uses global registers @code{$231} and up.
9730 @itemx -mno-zero-extend
9731 @opindex mzero-extend
9732 @opindex mno-zero-extend
9733 When reading data from memory in sizes shorter than 64 bits, use (do not
9734 use) zero-extending load instructions by default, rather than
9735 sign-extending ones.
9738 @itemx -mno-knuthdiv
9740 @opindex mno-knuthdiv
9741 Make the result of a division yielding a remainder have the same sign as
9742 the divisor. With the default, @option{-mno-knuthdiv}, the sign of the
9743 remainder follows the sign of the dividend. Both methods are
9744 arithmetically valid, the latter being almost exclusively used.
9746 @item -mtoplevel-symbols
9747 @itemx -mno-toplevel-symbols
9748 @opindex mtoplevel-symbols
9749 @opindex mno-toplevel-symbols
9750 Prepend (do not prepend) a @samp{:} to all global symbols, so the assembly
9751 code can be used with the @code{PREFIX} assembly directive.
9755 Generate an executable in the ELF format, rather than the default
9756 @samp{mmo} format used by the @command{mmix} simulator.
9758 @item -mbranch-predict
9759 @itemx -mno-branch-predict
9760 @opindex mbranch-predict
9761 @opindex mno-branch-predict
9762 Use (do not use) the probable-branch instructions, when static branch
9763 prediction indicates a probable branch.
9765 @item -mbase-addresses
9766 @itemx -mno-base-addresses
9767 @opindex mbase-addresses
9768 @opindex mno-base-addresses
9769 Generate (do not generate) code that uses @emph{base addresses}. Using a
9770 base address automatically generates a request (handled by the assembler
9771 and the linker) for a constant to be set up in a global register. The
9772 register is used for one or more base address requests within the range 0
9773 to 255 from the value held in the register. The generally leads to short
9774 and fast code, but the number of different data items that can be
9775 addressed is limited. This means that a program that uses lots of static
9776 data may require @option{-mno-base-addresses}.
9779 @itemx -mno-single-exit
9780 @opindex msingle-exit
9781 @opindex mno-single-exit
9782 Force (do not force) generated code to have a single exit point in each
9786 @node PDP-11 Options
9787 @subsection PDP-11 Options
9788 @cindex PDP-11 Options
9790 These options are defined for the PDP-11:
9795 Use hardware FPP floating point. This is the default. (FIS floating
9796 point on the PDP-11/40 is not supported.)
9799 @opindex msoft-float
9800 Do not use hardware floating point.
9804 Return floating-point results in ac0 (fr0 in Unix assembler syntax).
9808 Return floating-point results in memory. This is the default.
9812 Generate code for a PDP-11/40.
9816 Generate code for a PDP-11/45. This is the default.
9820 Generate code for a PDP-11/10.
9822 @item -mbcopy-builtin
9823 @opindex bcopy-builtin
9824 Use inline @code{movstrhi} patterns for copying memory. This is the
9829 Do not use inline @code{movstrhi} patterns for copying memory.
9835 Use 16-bit @code{int}. This is the default.
9841 Use 32-bit @code{int}.
9846 @opindex mno-float32
9847 Use 64-bit @code{float}. This is the default.
9852 @opindex mno-float64
9853 Use 32-bit @code{float}.
9857 Use @code{abshi2} pattern. This is the default.
9861 Do not use @code{abshi2} pattern.
9863 @item -mbranch-expensive
9864 @opindex mbranch-expensive
9865 Pretend that branches are expensive. This is for experimenting with
9866 code generation only.
9868 @item -mbranch-cheap
9869 @opindex mbranch-cheap
9870 Do not pretend that branches are expensive. This is the default.
9874 Generate code for a system with split I&D.
9878 Generate code for a system without split I&D. This is the default.
9882 Use Unix assembler syntax. This is the default when configured for
9887 Use DEC assembler syntax. This is the default when configured for any
9888 PDP-11 target other than @samp{pdp11-*-bsd}.
9891 @node Xstormy16 Options
9892 @subsection Xstormy16 Options
9893 @cindex Xstormy16 Options
9895 These options are defined for Xstormy16:
9900 Choose startup files and linker script suitable for the simulator.
9903 @node Xtensa Options
9904 @subsection Xtensa Options
9905 @cindex Xtensa Options
9907 The Xtensa architecture is designed to support many different
9908 configurations. The compiler's default options can be set to match a
9909 particular Xtensa configuration by copying a configuration file into the
9910 GCC sources when building GCC@. The options below may be used to
9911 override the default options.
9915 @itemx -mlittle-endian
9916 @opindex mbig-endian
9917 @opindex mlittle-endian
9918 Specify big-endian or little-endian byte ordering for the target Xtensa
9924 @opindex mno-density
9925 Enable or disable use of the optional Xtensa code density instructions.
9931 Enable or disable use of the Xtensa MAC16 option. When enabled, GCC
9932 will generate MAC16 instructions from standard C code, with the
9933 limitation that it will use neither the MR register file nor any
9934 instruction that operates on the MR registers. When this option is
9935 disabled, GCC will translate 16-bit multiply/accumulate operations to a
9936 combination of core instructions and library calls, depending on whether
9937 any other multiplier options are enabled.
9943 Enable or disable use of the 16-bit integer multiplier option. When
9944 enabled, the compiler will generate 16-bit multiply instructions for
9945 multiplications of 16 bits or smaller in standard C code. When this
9946 option is disabled, the compiler will either use 32-bit multiply or
9947 MAC16 instructions if they are available or generate library calls to
9948 perform the multiply operations using shifts and adds.
9954 Enable or disable use of the 32-bit integer multiplier option. When
9955 enabled, the compiler will generate 32-bit multiply instructions for
9956 multiplications of 32 bits or smaller in standard C code. When this
9957 option is disabled, the compiler will generate library calls to perform
9958 the multiply operations using either shifts and adds or 16-bit multiply
9959 instructions if they are available.
9965 Enable or disable use of the optional normalization shift amount
9966 (@code{NSA}) instructions to implement the built-in @code{ffs} function.
9972 Enable or disable use of the optional minimum and maximum value
9979 Enable or disable use of the optional sign extend (@code{SEXT})
9983 @itemx -mno-booleans
9985 @opindex mno-booleans
9986 Enable or disable support for the boolean register file used by Xtensa
9987 coprocessors. This is not typically useful by itself but may be
9988 required for other options that make use of the boolean registers (e.g.,
9989 the floating-point option).
9993 @opindex mhard-float
9994 @opindex msoft-float
9995 Enable or disable use of the floating-point option. When enabled, GCC
9996 generates floating-point instructions for 32-bit @code{float}
9997 operations. When this option is disabled, GCC generates library calls
9998 to emulate 32-bit floating-point operations using integer instructions.
9999 Regardless of this option, 64-bit @code{double} operations are always
10000 emulated with calls to library functions.
10003 @itemx -mno-fused-madd
10004 @opindex mfused-madd
10005 @opindex mno-fused-madd
10006 Enable or disable use of fused multiply/add and multiply/subtract
10007 instructions in the floating-point option. This has no effect if the
10008 floating-point option is not also enabled. Disabling fused multiply/add
10009 and multiply/subtract instructions forces the compiler to use separate
10010 instructions for the multiply and add/subtract operations. This may be
10011 desirable in some cases where strict IEEE 754-compliant results are
10012 required: the fused multiply add/subtract instructions do not round the
10013 intermediate result, thereby producing results with @emph{more} bits of
10014 precision than specified by the IEEE standard. Disabling fused multiply
10015 add/subtract instructions also ensures that the program output is not
10016 sensitive to the compiler's ability to combine multiply and add/subtract
10019 @item -mserialize-volatile
10020 @itemx -mno-serialize-volatile
10021 @opindex mserialize-volatile
10022 @opindex mno-serialize-volatile
10023 When this option is enabled, GCC inserts @code{MEMW} instructions before
10024 @code{volatile} memory references to guarantee sequential consistency.
10025 The default is @option{-mserialize-volatile}. Use
10026 @option{-mno-serialize-volatile} to omit the @code{MEMW} instructions.
10028 @item -mtext-section-literals
10029 @itemx -mno-text-section-literals
10030 @opindex mtext-section-literals
10031 @opindex mno-text-section-literals
10032 Control the treatment of literal pools. The default is
10033 @option{-mno-text-section-literals}, which places literals in a separate
10034 section in the output file. This allows the literal pool to be placed
10035 in a data RAM/ROM, and it also allows the linker to combine literal
10036 pools from separate object files to remove redundant literals and
10037 improve code size. With @option{-mtext-section-literals}, the literals
10038 are interspersed in the text section in order to keep them as close as
10039 possible to their references. This may be necessary for large assembly
10042 @item -mtarget-align
10043 @itemx -mno-target-align
10044 @opindex mtarget-align
10045 @opindex mno-target-align
10046 When this option is enabled, GCC instructs the assembler to
10047 automatically align instructions to reduce branch penalties at the
10048 expense of some code density. The assembler attempts to widen density
10049 instructions to align branch targets and the instructions following call
10050 instructions. If there are not enough preceding safe density
10051 instructions to align a target, no widening will be performed. The
10052 default is @option{-mtarget-align}. These options do not affect the
10053 treatment of auto-aligned instructions like @code{LOOP}, which the
10054 assembler will always align, either by widening density instructions or
10055 by inserting no-op instructions.
10058 @itemx -mno-longcalls
10059 @opindex mlongcalls
10060 @opindex mno-longcalls
10061 When this option is enabled, GCC instructs the assembler to translate
10062 direct calls to indirect calls unless it can determine that the target
10063 of a direct call is in the range allowed by the call instruction. This
10064 translation typically occurs for calls to functions in other source
10065 files. Specifically, the assembler translates a direct @code{CALL}
10066 instruction into an @code{L32R} followed by a @code{CALLX} instruction.
10067 The default is @option{-mno-longcalls}. This option should be used in
10068 programs where the call target can potentially be out of range. This
10069 option is implemented in the assembler, not the compiler, so the
10070 assembly code generated by GCC will still show direct call
10071 instructions---look at the disassembled object code to see the actual
10072 instructions. Note that the assembler will use an indirect call for
10073 every cross-file call, not just those that really will be out of range.
10076 @node Code Gen Options
10077 @section Options for Code Generation Conventions
10078 @cindex code generation conventions
10079 @cindex options, code generation
10080 @cindex run-time options
10082 These machine-independent options control the interface conventions
10083 used in code generation.
10085 Most of them have both positive and negative forms; the negative form
10086 of @option{-ffoo} would be @option{-fno-foo}. In the table below, only
10087 one of the forms is listed---the one which is not the default. You
10088 can figure out the other form by either removing @samp{no-} or adding
10093 @opindex fexceptions
10094 Enable exception handling. Generates extra code needed to propagate
10095 exceptions. For some targets, this implies GCC will generate frame
10096 unwind information for all functions, which can produce significant data
10097 size overhead, although it does not affect execution. If you do not
10098 specify this option, GCC will enable it by default for languages like
10099 C++ which normally require exception handling, and disable it for
10100 languages like C that do not normally require it. However, you may need
10101 to enable this option when compiling C code that needs to interoperate
10102 properly with exception handlers written in C++. You may also wish to
10103 disable this option if you are compiling older C++ programs that don't
10104 use exception handling.
10106 @item -fnon-call-exceptions
10107 @opindex fnon-call-exceptions
10108 Generate code that allows trapping instructions to throw exceptions.
10109 Note that this requires platform-specific runtime support that does
10110 not exist everywhere. Moreover, it only allows @emph{trapping}
10111 instructions to throw exceptions, i.e.@: memory references or floating
10112 point instructions. It does not allow exceptions to be thrown from
10113 arbitrary signal handlers such as @code{SIGALRM}.
10115 @item -funwind-tables
10116 @opindex funwind-tables
10117 Similar to @option{-fexceptions}, except that it will just generate any needed
10118 static data, but will not affect the generated code in any other way.
10119 You will normally not enable this option; instead, a language processor
10120 that needs this handling would enable it on your behalf.
10122 @item -fasynchronous-unwind-tables
10123 @opindex funwind-tables
10124 Generate unwind table in dwarf2 format, if supported by target machine. The
10125 table is exact at each instruction boundary, so it can be used for stack
10126 unwinding from asynchronous events (such as debugger or garbage collector).
10128 @item -fpcc-struct-return
10129 @opindex fpcc-struct-return
10130 Return ``short'' @code{struct} and @code{union} values in memory like
10131 longer ones, rather than in registers. This convention is less
10132 efficient, but it has the advantage of allowing intercallability between
10133 GCC-compiled files and files compiled with other compilers, particularly
10134 the Portable C Compiler (pcc).
10136 The precise convention for returning structures in memory depends
10137 on the target configuration macros.
10139 Short structures and unions are those whose size and alignment match
10140 that of some integer type.
10142 @strong{Warning:} code compiled with the @option{-fpcc-struct-return}
10143 switch is not binary compatible with code compiled with the
10144 @option{-freg-struct-return} switch.
10145 Use it to conform to a non-default application binary interface.
10147 @item -freg-struct-return
10148 @opindex freg-struct-return
10149 Return @code{struct} and @code{union} values in registers when possible.
10150 This is more efficient for small structures than
10151 @option{-fpcc-struct-return}.
10153 If you specify neither @option{-fpcc-struct-return} nor
10154 @option{-freg-struct-return}, GCC defaults to whichever convention is
10155 standard for the target. If there is no standard convention, GCC
10156 defaults to @option{-fpcc-struct-return}, except on targets where GCC is
10157 the principal compiler. In those cases, we can choose the standard, and
10158 we chose the more efficient register return alternative.
10160 @strong{Warning:} code compiled with the @option{-freg-struct-return}
10161 switch is not binary compatible with code compiled with the
10162 @option{-fpcc-struct-return} switch.
10163 Use it to conform to a non-default application binary interface.
10165 @item -fshort-enums
10166 @opindex fshort-enums
10167 Allocate to an @code{enum} type only as many bytes as it needs for the
10168 declared range of possible values. Specifically, the @code{enum} type
10169 will be equivalent to the smallest integer type which has enough room.
10171 @strong{Warning:} the @option{-fshort-enums} switch causes GCC to generate
10172 code that is not binary compatible with code generated without that switch.
10173 Use it to conform to a non-default application binary interface.
10175 @item -fshort-double
10176 @opindex fshort-double
10177 Use the same size for @code{double} as for @code{float}.
10179 @strong{Warning:} the @option{-fshort-double} switch causes GCC to generate
10180 code that is not binary compatible with code generated without that switch.
10181 Use it to conform to a non-default application binary interface.
10183 @item -fshort-wchar
10184 @opindex fshort-wchar
10185 Override the underlying type for @samp{wchar_t} to be @samp{short
10186 unsigned int} instead of the default for the target. This option is
10187 useful for building programs to run under WINE@.
10189 @strong{Warning:} the @option{-fshort-wchar} switch causes GCC to generate
10190 code that is not binary compatible with code generated without that switch.
10191 Use it to conform to a non-default application binary interface.
10193 @item -fshared-data
10194 @opindex fshared-data
10195 Requests that the data and non-@code{const} variables of this
10196 compilation be shared data rather than private data. The distinction
10197 makes sense only on certain operating systems, where shared data is
10198 shared between processes running the same program, while private data
10199 exists in one copy per process.
10202 @opindex fno-common
10203 In C, allocate even uninitialized global variables in the data section of the
10204 object file, rather than generating them as common blocks. This has the
10205 effect that if the same variable is declared (without @code{extern}) in
10206 two different compilations, you will get an error when you link them.
10207 The only reason this might be useful is if you wish to verify that the
10208 program will work on other systems which always work this way.
10212 Ignore the @samp{#ident} directive.
10214 @item -fno-gnu-linker
10215 @opindex fno-gnu-linker
10216 Do not output global initializations (such as C++ constructors and
10217 destructors) in the form used by the GNU linker (on systems where the GNU
10218 linker is the standard method of handling them). Use this option when
10219 you want to use a non-GNU linker, which also requires using the
10220 @command{collect2} program to make sure the system linker includes
10221 constructors and destructors. (@command{collect2} is included in the GCC
10222 distribution.) For systems which @emph{must} use @command{collect2}, the
10223 compiler driver @command{gcc} is configured to do this automatically.
10225 @item -finhibit-size-directive
10226 @opindex finhibit-size-directive
10227 Don't output a @code{.size} assembler directive, or anything else that
10228 would cause trouble if the function is split in the middle, and the
10229 two halves are placed at locations far apart in memory. This option is
10230 used when compiling @file{crtstuff.c}; you should not need to use it
10233 @item -fverbose-asm
10234 @opindex fverbose-asm
10235 Put extra commentary information in the generated assembly code to
10236 make it more readable. This option is generally only of use to those
10237 who actually need to read the generated assembly code (perhaps while
10238 debugging the compiler itself).
10240 @option{-fno-verbose-asm}, the default, causes the
10241 extra information to be omitted and is useful when comparing two assembler
10246 Consider all memory references through pointers to be volatile.
10248 @item -fvolatile-global
10249 @opindex fvolatile-global
10250 Consider all memory references to extern and global data items to
10251 be volatile. GCC does not consider static data items to be volatile
10252 because of this switch.
10254 @item -fvolatile-static
10255 @opindex fvolatile-static
10256 Consider all memory references to static data to be volatile.
10260 @cindex global offset table
10262 Generate position-independent code (PIC) suitable for use in a shared
10263 library, if supported for the target machine. Such code accesses all
10264 constant addresses through a global offset table (GOT)@. The dynamic
10265 loader resolves the GOT entries when the program starts (the dynamic
10266 loader is not part of GCC; it is part of the operating system). If
10267 the GOT size for the linked executable exceeds a machine-specific
10268 maximum size, you get an error message from the linker indicating that
10269 @option{-fpic} does not work; in that case, recompile with @option{-fPIC}
10270 instead. (These maximums are 16k on the m88k, 8k on the SPARC, and 32k
10271 on the m68k and RS/6000. The 386 has no such limit.)
10273 Position-independent code requires special support, and therefore works
10274 only on certain machines. For the 386, GCC supports PIC for System V
10275 but not for the Sun 386i. Code generated for the IBM RS/6000 is always
10276 position-independent.
10280 If supported for the target machine, emit position-independent code,
10281 suitable for dynamic linking and avoiding any limit on the size of the
10282 global offset table. This option makes a difference on the m68k, m88k,
10285 Position-independent code requires special support, and therefore works
10286 only on certain machines.
10288 @item -ffixed-@var{reg}
10290 Treat the register named @var{reg} as a fixed register; generated code
10291 should never refer to it (except perhaps as a stack pointer, frame
10292 pointer or in some other fixed role).
10294 @var{reg} must be the name of a register. The register names accepted
10295 are machine-specific and are defined in the @code{REGISTER_NAMES}
10296 macro in the machine description macro file.
10298 This flag does not have a negative form, because it specifies a
10301 @item -fcall-used-@var{reg}
10302 @opindex fcall-used
10303 Treat the register named @var{reg} as an allocable register that is
10304 clobbered by function calls. It may be allocated for temporaries or
10305 variables that do not live across a call. Functions compiled this way
10306 will not save and restore the register @var{reg}.
10308 It is an error to used this flag with the frame pointer or stack pointer.
10309 Use of this flag for other registers that have fixed pervasive roles in
10310 the machine's execution model will produce disastrous results.
10312 This flag does not have a negative form, because it specifies a
10315 @item -fcall-saved-@var{reg}
10316 @opindex fcall-saved
10317 Treat the register named @var{reg} as an allocable register saved by
10318 functions. It may be allocated even for temporaries or variables that
10319 live across a call. Functions compiled this way will save and restore
10320 the register @var{reg} if they use it.
10322 It is an error to used this flag with the frame pointer or stack pointer.
10323 Use of this flag for other registers that have fixed pervasive roles in
10324 the machine's execution model will produce disastrous results.
10326 A different sort of disaster will result from the use of this flag for
10327 a register in which function values may be returned.
10329 This flag does not have a negative form, because it specifies a
10332 @item -fpack-struct
10333 @opindex fpack-struct
10334 Pack all structure members together without holes.
10336 @strong{Warning:} the @option{-fpack-struct} switch causes GCC to generate
10337 code that is not binary compatible with code generated without that switch.
10338 Additionally, it makes the code suboptimial.
10339 Use it to conform to a non-default application binary interface.
10341 @item -finstrument-functions
10342 @opindex finstrument-functions
10343 Generate instrumentation calls for entry and exit to functions. Just
10344 after function entry and just before function exit, the following
10345 profiling functions will be called with the address of the current
10346 function and its call site. (On some platforms,
10347 @code{__builtin_return_address} does not work beyond the current
10348 function, so the call site information may not be available to the
10349 profiling functions otherwise.)
10352 void __cyg_profile_func_enter (void *this_fn,
10354 void __cyg_profile_func_exit (void *this_fn,
10358 The first argument is the address of the start of the current function,
10359 which may be looked up exactly in the symbol table.
10361 This instrumentation is also done for functions expanded inline in other
10362 functions. The profiling calls will indicate where, conceptually, the
10363 inline function is entered and exited. This means that addressable
10364 versions of such functions must be available. If all your uses of a
10365 function are expanded inline, this may mean an additional expansion of
10366 code size. If you use @samp{extern inline} in your C code, an
10367 addressable version of such functions must be provided. (This is
10368 normally the case anyways, but if you get lucky and the optimizer always
10369 expands the functions inline, you might have gotten away without
10370 providing static copies.)
10372 A function may be given the attribute @code{no_instrument_function}, in
10373 which case this instrumentation will not be done. This can be used, for
10374 example, for the profiling functions listed above, high-priority
10375 interrupt routines, and any functions from which the profiling functions
10376 cannot safely be called (perhaps signal handlers, if the profiling
10377 routines generate output or allocate memory).
10379 @item -fstack-check
10380 @opindex fstack-check
10381 Generate code to verify that you do not go beyond the boundary of the
10382 stack. You should specify this flag if you are running in an
10383 environment with multiple threads, but only rarely need to specify it in
10384 a single-threaded environment since stack overflow is automatically
10385 detected on nearly all systems if there is only one stack.
10387 Note that this switch does not actually cause checking to be done; the
10388 operating system must do that. The switch causes generation of code
10389 to ensure that the operating system sees the stack being extended.
10391 @item -fstack-limit-register=@var{reg}
10392 @itemx -fstack-limit-symbol=@var{sym}
10393 @itemx -fno-stack-limit
10394 @opindex fstack-limit-register
10395 @opindex fstack-limit-symbol
10396 @opindex fno-stack-limit
10397 Generate code to ensure that the stack does not grow beyond a certain value,
10398 either the value of a register or the address of a symbol. If the stack
10399 would grow beyond the value, a signal is raised. For most targets,
10400 the signal is raised before the stack overruns the boundary, so
10401 it is possible to catch the signal without taking special precautions.
10403 For instance, if the stack starts at absolute address @samp{0x80000000}
10404 and grows downwards, you can use the flags
10405 @option{-fstack-limit-symbol=__stack_limit} and
10406 @option{-Wl,--defsym,__stack_limit=0x7ffe0000} to enforce a stack limit
10407 of 128KB@. Note that this may only work with the GNU linker.
10409 @cindex aliasing of parameters
10410 @cindex parameters, aliased
10411 @item -fargument-alias
10412 @itemx -fargument-noalias
10413 @itemx -fargument-noalias-global
10414 @opindex fargument-alias
10415 @opindex fargument-noalias
10416 @opindex fargument-noalias-global
10417 Specify the possible relationships among parameters and between
10418 parameters and global data.
10420 @option{-fargument-alias} specifies that arguments (parameters) may
10421 alias each other and may alias global storage.@*
10422 @option{-fargument-noalias} specifies that arguments do not alias
10423 each other, but may alias global storage.@*
10424 @option{-fargument-noalias-global} specifies that arguments do not
10425 alias each other and do not alias global storage.
10427 Each language will automatically use whatever option is required by
10428 the language standard. You should not need to use these options yourself.
10430 @item -fleading-underscore
10431 @opindex fleading-underscore
10432 This option and its counterpart, @option{-fno-leading-underscore}, forcibly
10433 change the way C symbols are represented in the object file. One use
10434 is to help link with legacy assembly code.
10436 @strong{Warning:} the @option{-fleading-underscore} switch causes GCC to
10437 generate code that is not binary compatible with code generated without that
10438 switch. Use it to conform to a non-default application binary interface.
10439 Not all targets provide complete support for this switch.
10441 @item -ftls-model=@var{model}
10442 Alter the thread-local storage model to be used (@pxref{Thread-Local}).
10443 The @var{model} argument should be one of @code{global-dynamic},
10444 @code{local-dynamic}, @code{initial-exec} or @code{local-exec}.
10446 The default without @option{-fpic} is @code{initial-exec}; with
10447 @option{-fpic} the default is @code{global-dynamic}.
10452 @node Environment Variables
10453 @section Environment Variables Affecting GCC
10454 @cindex environment variables
10456 @c man begin ENVIRONMENT
10458 This section describes several environment variables that affect how GCC
10459 operates. Some of them work by specifying directories or prefixes to use
10460 when searching for various kinds of files. Some are used to specify other
10461 aspects of the compilation environment.
10463 Note that you can also specify places to search using options such as
10464 @option{-B}, @option{-I} and @option{-L} (@pxref{Directory Options}). These
10465 take precedence over places specified using environment variables, which
10466 in turn take precedence over those specified by the configuration of GCC@.
10467 @xref{Driver,, Controlling the Compilation Driver @file{gcc}, gccint,
10468 GNU Compiler Collection (GCC) Internals}.
10473 @c @itemx LC_COLLATE
10475 @c @itemx LC_MONETARY
10476 @c @itemx LC_NUMERIC
10481 @c @findex LC_COLLATE
10482 @findex LC_MESSAGES
10483 @c @findex LC_MONETARY
10484 @c @findex LC_NUMERIC
10488 These environment variables control the way that GCC uses
10489 localization information that allow GCC to work with different
10490 national conventions. GCC inspects the locale categories
10491 @env{LC_CTYPE} and @env{LC_MESSAGES} if it has been configured to do
10492 so. These locale categories can be set to any value supported by your
10493 installation. A typical value is @samp{en_UK} for English in the United
10496 The @env{LC_CTYPE} environment variable specifies character
10497 classification. GCC uses it to determine the character boundaries in
10498 a string; this is needed for some multibyte encodings that contain quote
10499 and escape characters that would otherwise be interpreted as a string
10502 The @env{LC_MESSAGES} environment variable specifies the language to
10503 use in diagnostic messages.
10505 If the @env{LC_ALL} environment variable is set, it overrides the value
10506 of @env{LC_CTYPE} and @env{LC_MESSAGES}; otherwise, @env{LC_CTYPE}
10507 and @env{LC_MESSAGES} default to the value of the @env{LANG}
10508 environment variable. If none of these variables are set, GCC
10509 defaults to traditional C English behavior.
10513 If @env{TMPDIR} is set, it specifies the directory to use for temporary
10514 files. GCC uses temporary files to hold the output of one stage of
10515 compilation which is to be used as input to the next stage: for example,
10516 the output of the preprocessor, which is the input to the compiler
10519 @item GCC_EXEC_PREFIX
10520 @findex GCC_EXEC_PREFIX
10521 If @env{GCC_EXEC_PREFIX} is set, it specifies a prefix to use in the
10522 names of the subprograms executed by the compiler. No slash is added
10523 when this prefix is combined with the name of a subprogram, but you can
10524 specify a prefix that ends with a slash if you wish.
10526 If @env{GCC_EXEC_PREFIX} is not set, GCC will attempt to figure out
10527 an appropriate prefix to use based on the pathname it was invoked with.
10529 If GCC cannot find the subprogram using the specified prefix, it
10530 tries looking in the usual places for the subprogram.
10532 The default value of @env{GCC_EXEC_PREFIX} is
10533 @file{@var{prefix}/lib/gcc-lib/} where @var{prefix} is the value
10534 of @code{prefix} when you ran the @file{configure} script.
10536 Other prefixes specified with @option{-B} take precedence over this prefix.
10538 This prefix is also used for finding files such as @file{crt0.o} that are
10541 In addition, the prefix is used in an unusual way in finding the
10542 directories to search for header files. For each of the standard
10543 directories whose name normally begins with @samp{/usr/local/lib/gcc-lib}
10544 (more precisely, with the value of @env{GCC_INCLUDE_DIR}), GCC tries
10545 replacing that beginning with the specified prefix to produce an
10546 alternate directory name. Thus, with @option{-Bfoo/}, GCC will search
10547 @file{foo/bar} where it would normally search @file{/usr/local/lib/bar}.
10548 These alternate directories are searched first; the standard directories
10551 @item COMPILER_PATH
10552 @findex COMPILER_PATH
10553 The value of @env{COMPILER_PATH} is a colon-separated list of
10554 directories, much like @env{PATH}. GCC tries the directories thus
10555 specified when searching for subprograms, if it can't find the
10556 subprograms using @env{GCC_EXEC_PREFIX}.
10559 @findex LIBRARY_PATH
10560 The value of @env{LIBRARY_PATH} is a colon-separated list of
10561 directories, much like @env{PATH}. When configured as a native compiler,
10562 GCC tries the directories thus specified when searching for special
10563 linker files, if it can't find them using @env{GCC_EXEC_PREFIX}. Linking
10564 using GCC also uses these directories when searching for ordinary
10565 libraries for the @option{-l} option (but directories specified with
10566 @option{-L} come first).
10570 @cindex locale definition
10571 This variable is used to pass locale information to the compiler. One way in
10572 which this information is used is to determine the character set to be used
10573 when character literals, string literals and comments are parsed in C and C++.
10574 When the compiler is configured to allow multibyte characters,
10575 the following values for @env{LANG} are recognized:
10579 Recognize JIS characters.
10581 Recognize SJIS characters.
10583 Recognize EUCJP characters.
10586 If @env{LANG} is not defined, or if it has some other value, then the
10587 compiler will use mblen and mbtowc as defined by the default locale to
10588 recognize and translate multibyte characters.
10592 Some additional environments variables affect the behavior of the
10595 @include cppenv.texi
10599 @node Running Protoize
10600 @section Running Protoize
10602 The program @code{protoize} is an optional part of GCC@. You can use
10603 it to add prototypes to a program, thus converting the program to ISO
10604 C in one respect. The companion program @code{unprotoize} does the
10605 reverse: it removes argument types from any prototypes that are found.
10607 When you run these programs, you must specify a set of source files as
10608 command line arguments. The conversion programs start out by compiling
10609 these files to see what functions they define. The information gathered
10610 about a file @var{foo} is saved in a file named @file{@var{foo}.X}.
10612 After scanning comes actual conversion. The specified files are all
10613 eligible to be converted; any files they include (whether sources or
10614 just headers) are eligible as well.
10616 But not all the eligible files are converted. By default,
10617 @code{protoize} and @code{unprotoize} convert only source and header
10618 files in the current directory. You can specify additional directories
10619 whose files should be converted with the @option{-d @var{directory}}
10620 option. You can also specify particular files to exclude with the
10621 @option{-x @var{file}} option. A file is converted if it is eligible, its
10622 directory name matches one of the specified directory names, and its
10623 name within the directory has not been excluded.
10625 Basic conversion with @code{protoize} consists of rewriting most
10626 function definitions and function declarations to specify the types of
10627 the arguments. The only ones not rewritten are those for varargs
10630 @code{protoize} optionally inserts prototype declarations at the
10631 beginning of the source file, to make them available for any calls that
10632 precede the function's definition. Or it can insert prototype
10633 declarations with block scope in the blocks where undeclared functions
10636 Basic conversion with @code{unprotoize} consists of rewriting most
10637 function declarations to remove any argument types, and rewriting
10638 function definitions to the old-style pre-ISO form.
10640 Both conversion programs print a warning for any function declaration or
10641 definition that they can't convert. You can suppress these warnings
10644 The output from @code{protoize} or @code{unprotoize} replaces the
10645 original source file. The original file is renamed to a name ending
10646 with @samp{.save} (for DOS, the saved filename ends in @samp{.sav}
10647 without the original @samp{.c} suffix). If the @samp{.save} (@samp{.sav}
10648 for DOS) file already exists, then the source file is simply discarded.
10650 @code{protoize} and @code{unprotoize} both depend on GCC itself to
10651 scan the program and collect information about the functions it uses.
10652 So neither of these programs will work until GCC is installed.
10654 Here is a table of the options you can use with @code{protoize} and
10655 @code{unprotoize}. Each option works with both programs unless
10659 @item -B @var{directory}
10660 Look for the file @file{SYSCALLS.c.X} in @var{directory}, instead of the
10661 usual directory (normally @file{/usr/local/lib}). This file contains
10662 prototype information about standard system functions. This option
10663 applies only to @code{protoize}.
10665 @item -c @var{compilation-options}
10666 Use @var{compilation-options} as the options when running @code{gcc} to
10667 produce the @samp{.X} files. The special option @option{-aux-info} is
10668 always passed in addition, to tell @code{gcc} to write a @samp{.X} file.
10670 Note that the compilation options must be given as a single argument to
10671 @code{protoize} or @code{unprotoize}. If you want to specify several
10672 @code{gcc} options, you must quote the entire set of compilation options
10673 to make them a single word in the shell.
10675 There are certain @code{gcc} arguments that you cannot use, because they
10676 would produce the wrong kind of output. These include @option{-g},
10677 @option{-O}, @option{-c}, @option{-S}, and @option{-o} If you include these in
10678 the @var{compilation-options}, they are ignored.
10681 Rename files to end in @samp{.C} (@samp{.cc} for DOS-based file
10682 systems) instead of @samp{.c}. This is convenient if you are converting
10683 a C program to C++. This option applies only to @code{protoize}.
10686 Add explicit global declarations. This means inserting explicit
10687 declarations at the beginning of each source file for each function
10688 that is called in the file and was not declared. These declarations
10689 precede the first function definition that contains a call to an
10690 undeclared function. This option applies only to @code{protoize}.
10692 @item -i @var{string}
10693 Indent old-style parameter declarations with the string @var{string}.
10694 This option applies only to @code{protoize}.
10696 @code{unprotoize} converts prototyped function definitions to old-style
10697 function definitions, where the arguments are declared between the
10698 argument list and the initial @samp{@{}. By default, @code{unprotoize}
10699 uses five spaces as the indentation. If you want to indent with just
10700 one space instead, use @option{-i " "}.
10703 Keep the @samp{.X} files. Normally, they are deleted after conversion
10707 Add explicit local declarations. @code{protoize} with @option{-l} inserts
10708 a prototype declaration for each function in each block which calls the
10709 function without any declaration. This option applies only to
10713 Make no real changes. This mode just prints information about the conversions
10714 that would have been done without @option{-n}.
10717 Make no @samp{.save} files. The original files are simply deleted.
10718 Use this option with caution.
10720 @item -p @var{program}
10721 Use the program @var{program} as the compiler. Normally, the name
10722 @file{gcc} is used.
10725 Work quietly. Most warnings are suppressed.
10728 Print the version number, just like @option{-v} for @code{gcc}.
10731 If you need special compiler options to compile one of your program's
10732 source files, then you should generate that file's @samp{.X} file
10733 specially, by running @code{gcc} on that source file with the
10734 appropriate options and the option @option{-aux-info}. Then run
10735 @code{protoize} on the entire set of files. @code{protoize} will use
10736 the existing @samp{.X} file because it is newer than the source file.
10740 gcc -Dfoo=bar file1.c -aux-info file1.X
10745 You need to include the special files along with the rest in the
10746 @code{protoize} command, even though their @samp{.X} files already
10747 exist, because otherwise they won't get converted.
10749 @xref{Protoize Caveats}, for more information on how to use
10750 @code{protoize} successfully.