1 @c Copyright (C) 1988, 1989, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
2 @c 2000, 2001, 2002, 2003, 2004, 2005 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, 2003, 2004, 2005 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 1.2 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), as(1), ld(1), gdb(1), adb(1), dbx(1), sdb(1)
47 and the Info entries for @file{gcc}, @file{cpp}, @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}}.
55 See the Info entry for @command{gcc}, or
56 @w{@uref{http://gcc.gnu.org/onlinedocs/gcc/Contributors.html}},
57 for contributors to GCC@.
62 @chapter GCC Command Options
63 @cindex GCC command options
64 @cindex command options
65 @cindex options, GCC command
67 @c man begin DESCRIPTION
68 When you invoke GCC, it normally does preprocessing, compilation,
69 assembly and linking. The ``overall options'' allow you to stop this
70 process at an intermediate stage. For example, the @option{-c} option
71 says not to run the linker. Then the output consists of object files
72 output by the assembler.
74 Other options are passed on to one stage of processing. Some options
75 control the preprocessor and others the compiler itself. Yet other
76 options control the assembler and linker; most of these are not
77 documented here, since you rarely need to use any of them.
79 @cindex C compilation options
80 Most of the command line options that you can use with GCC are useful
81 for C programs; when an option is only useful with another language
82 (usually C++), the explanation says so explicitly. If the description
83 for a particular option does not mention a source language, you can use
84 that option with all supported languages.
86 @cindex C++ compilation options
87 @xref{Invoking G++,,Compiling C++ Programs}, for a summary of special
88 options for compiling C++ programs.
90 @cindex grouping options
91 @cindex options, grouping
92 The @command{gcc} program accepts options and file names as operands. Many
93 options have multi-letter names; therefore multiple single-letter options
94 may @emph{not} be grouped: @option{-dr} is very different from @w{@samp{-d
97 @cindex order of options
98 @cindex options, order
99 You can mix options and other arguments. For the most part, the order
100 you use doesn't matter. Order does matter when you use several options
101 of the same kind; for example, if you specify @option{-L} more than once,
102 the directories are searched in the order specified.
104 Many options have long names starting with @samp{-f} or with
105 @samp{-W}---for example, @option{-fforce-mem},
106 @option{-fstrength-reduce}, @option{-Wformat} and so on. Most of
107 these have both positive and negative forms; the negative form of
108 @option{-ffoo} would be @option{-fno-foo}. This manual documents
109 only one of these two forms, whichever one is not the default.
113 @xref{Option Index}, for an index to GCC's options.
116 * Option Summary:: Brief list of all options, without explanations.
117 * Overall Options:: Controlling the kind of output:
118 an executable, object files, assembler files,
119 or preprocessed source.
120 * Invoking G++:: Compiling C++ programs.
121 * C Dialect Options:: Controlling the variant of C language compiled.
122 * C++ Dialect Options:: Variations on C++.
123 * Objective-C and Objective-C++ Dialect Options:: Variations on Objective-C
125 * Language Independent Options:: Controlling how diagnostics should be
127 * Warning Options:: How picky should the compiler be?
128 * Debugging Options:: Symbol tables, measurements, and debugging dumps.
129 * Optimize Options:: How much optimization?
130 * Preprocessor Options:: Controlling header files and macro definitions.
131 Also, getting dependency information for Make.
132 * Assembler Options:: Passing options to the assembler.
133 * Link Options:: Specifying libraries and so on.
134 * Directory Options:: Where to find header files and libraries.
135 Where to find the compiler executable files.
136 * Spec Files:: How to pass switches to sub-processes.
137 * Target Options:: Running a cross-compiler, or an old version of GCC.
138 * Submodel Options:: Specifying minor hardware or convention variations,
139 such as 68010 vs 68020.
140 * Code Gen Options:: Specifying conventions for function calls, data layout
142 * Environment Variables:: Env vars that affect GCC.
143 * Precompiled Headers:: Compiling a header once, and using it many times.
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}.
158 @gccoptlist{-c -S -E -o @var{file} -combine -pipe -pass-exit-codes @gol
159 -x @var{language} -v -### --help --target-help --version}
161 @item C Language Options
162 @xref{C Dialect Options,,Options Controlling C Dialect}.
163 @gccoptlist{-ansi -std=@var{standard} -aux-info @var{filename} @gol
164 -fno-asm -fno-builtin -fno-builtin-@var{function} @gol
165 -fhosted -ffreestanding -fms-extensions @gol
166 -trigraphs -no-integrated-cpp -traditional -traditional-cpp @gol
167 -fallow-single-precision -fcond-mismatch @gol
168 -fsigned-bitfields -fsigned-char @gol
169 -funsigned-bitfields -funsigned-char}
171 @item C++ Language Options
172 @xref{C++ Dialect Options,,Options Controlling C++ Dialect}.
173 @gccoptlist{-fabi-version=@var{n} -fno-access-control -fcheck-new @gol
174 -fconserve-space -fno-const-strings @gol
175 -fno-elide-constructors @gol
176 -fno-enforce-eh-specs @gol
177 -ffor-scope -fno-for-scope -fno-gnu-keywords @gol
178 -fno-implicit-templates @gol
179 -fno-implicit-inline-templates @gol
180 -fno-implement-inlines -fms-extensions @gol
181 -fno-nonansi-builtins -fno-operator-names @gol
182 -fno-optional-diags -fpermissive @gol
183 -frepo -fno-rtti -fstats -ftemplate-depth-@var{n} @gol
184 -fno-threadsafe-statics -fuse-cxa-atexit -fno-weak -nostdinc++ @gol
185 -fno-default-inline -fvisibility-inlines-hidden @gol
186 -Wabi -Wctor-dtor-privacy @gol
187 -Wnon-virtual-dtor -Wreorder @gol
188 -Weffc++ -Wno-deprecated @gol
189 -Wno-non-template-friend -Wold-style-cast @gol
190 -Woverloaded-virtual -Wno-pmf-conversions @gol
191 -Wsign-promo -Wsynth}
193 @item Objective-C and Objective-C++ Language Options
194 @xref{Objective-C and Objective-C++ Dialect Options,,Options Controlling
195 Objective-C and Objective-C++ Dialects}.
197 -fconstant-string-class=@var{class-name} @gol
198 -fgnu-runtime -fnext-runtime @gol
199 -fno-nil-receivers @gol
200 -fobjc-exceptions @gol
201 -freplace-objc-classes @gol
204 -Wno-protocol -Wselector -Wundeclared-selector}
206 @item Language Independent Options
207 @xref{Language Independent Options,,Options to Control Diagnostic Messages Formatting}.
208 @gccoptlist{-fmessage-length=@var{n} @gol
209 -fdiagnostics-show-location=@r{[}once@r{|}every-line@r{]}}
211 @item Warning Options
212 @xref{Warning Options,,Options to Request or Suppress Warnings}.
213 @gccoptlist{-fsyntax-only -pedantic -pedantic-errors @gol
214 -w -Wextra -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 -Wendif-labels @gol
218 -Werror -Werror-implicit-function-declaration @gol
219 -Wfatal-errors -Wfloat-equal -Wformat -Wformat=2 @gol
220 -Wno-format-extra-args -Wformat-nonliteral @gol
221 -Wformat-security -Wformat-y2k @gol
222 -Wimplicit -Wimplicit-function-declaration -Wimplicit-int @gol
223 -Wimport -Wno-import -Winit-self -Winline @gol
224 -Wno-invalid-offsetof -Winvalid-pch @gol
225 -Wlarger-than-@var{len} -Wlong-long @gol
226 -Wmain -Wmissing-braces -Wmissing-field-initializers @gol
227 -Wmissing-format-attribute -Wmissing-include-dirs @gol
228 -Wmissing-noreturn @gol
229 -Wno-multichar -Wnonnull -Wpacked -Wpadded @gol
230 -Wparentheses -Wpointer-arith -Wredundant-decls @gol
231 -Wreturn-type -Wsequence-point -Wshadow @gol
232 -Wsign-compare -Wstrict-aliasing -Wstrict-aliasing=2 @gol
233 -Wswitch -Wswitch-default -Wswitch-enum @gol
234 -Wsystem-headers -Wtrigraphs -Wundef -Wuninitialized @gol
235 -Wunknown-pragmas -Wunreachable-code @gol
236 -Wunused -Wunused-function -Wunused-label -Wunused-parameter @gol
237 -Wunused-value -Wunused-variable -Wwrite-strings @gol
240 @item C-only Warning Options
241 @gccoptlist{-Wbad-function-cast -Wmissing-declarations @gol
242 -Wmissing-prototypes -Wnested-externs -Wold-style-definition @gol
243 -Wstrict-prototypes -Wtraditional @gol
244 -Wdeclaration-after-statement -Wno-pointer-sign}
246 @item Debugging Options
247 @xref{Debugging Options,,Options for Debugging Your Program or GCC}.
248 @gccoptlist{-d@var{letters} -dumpspecs -dumpmachine -dumpversion @gol
249 -fdump-unnumbered -fdump-translation-unit@r{[}-@var{n}@r{]} @gol
250 -fdump-class-hierarchy@r{[}-@var{n}@r{]} @gol
251 -fdump-ipa-all -fdump-ipa-cgraph @gol
253 -fdump-tree-original@r{[}-@var{n}@r{]} @gol
254 -fdump-tree-optimized@r{[}-@var{n}@r{]} @gol
255 -fdump-tree-inlined@r{[}-@var{n}@r{]} @gol
256 -fdump-tree-cfg -fdump-tree-vcg -fdump-tree-alias @gol
258 -fdump-tree-ssa@r{[}-@var{n}@r{]} -fdump-tree-pre@r{[}-@var{n}@r{]} @gol
259 -fdump-tree-ccp@r{[}-@var{n}@r{]} -fdump-tree-dce@r{[}-@var{n}@r{]} @gol
260 -fdump-tree-gimple@r{[}-raw@r{]} -fdump-tree-mudflap@r{[}-@var{n}@r{]} @gol
261 -fdump-tree-dom@r{[}-@var{n}@r{]} @gol
262 -fdump-tree-dse@r{[}-@var{n}@r{]} @gol
263 -fdump-tree-phiopt@r{[}-@var{n}@r{]} @gol
264 -fdump-tree-forwprop@r{[}-@var{n}@r{]} @gol
265 -fdump-tree-copyrename@r{[}-@var{n}@r{]} @gol
266 -fdump-tree-nrv -fdump-tree-vect @gol
267 -fdump-tree-sra@r{[}-@var{n}@r{]} @gol
268 -fdump-tree-fre@r{[}-@var{n}@r{]} @gol
269 -feliminate-dwarf2-dups -feliminate-unused-debug-types @gol
270 -feliminate-unused-debug-symbols -fmem-report -fprofile-arcs -ftree-based-profiling @gol
271 -frandom-seed=@var{string} -fsched-verbose=@var{n} @gol
272 -ftest-coverage -ftime-report -fvar-tracking @gol
273 -g -g@var{level} -gcoff -gdwarf-2 @gol
274 -ggdb -gstabs -gstabs+ -gvms -gxcoff -gxcoff+ @gol
275 -p -pg -print-file-name=@var{library} -print-libgcc-file-name @gol
276 -print-multi-directory -print-multi-lib @gol
277 -print-prog-name=@var{program} -print-search-dirs -Q @gol
280 @item Optimization Options
281 @xref{Optimize Options,,Options that Control Optimization}.
282 @gccoptlist{-falign-functions=@var{n} -falign-jumps=@var{n} @gol
283 -falign-labels=@var{n} -falign-loops=@var{n} @gol
284 -fbounds-check -fmudflap -fmudflapth -fmudflapir @gol
285 -fbranch-probabilities -fprofile-values -fvpt -fbranch-target-load-optimize @gol
286 -fbranch-target-load-optimize2 -fbtr-bb-exclusive @gol
287 -fcaller-saves -fcprop-registers @gol
288 -fcse-follow-jumps -fcse-skip-blocks -fdata-sections @gol
289 -fdelayed-branch -fdelete-null-pointer-checks @gol
290 -fexpensive-optimizations -ffast-math -ffloat-store @gol
291 -fforce-addr -fforce-mem -ffunction-sections @gol
292 -fgcse -fgcse-lm -fgcse-sm -fgcse-las -fgcse-after-reload @gol
293 -floop-optimize -fcrossjumping -fif-conversion -fif-conversion2 @gol
294 -finline-functions -finline-limit=@var{n} -fkeep-inline-functions @gol
295 -fkeep-static-consts -fmerge-constants -fmerge-all-constants @gol
296 -fmodulo-sched -fno-branch-count-reg @gol
297 -fno-default-inline -fno-defer-pop -floop-optimize2 -fmove-loop-invariants @gol
298 -fno-function-cse -fno-guess-branch-probability @gol
299 -fno-inline -fno-math-errno -fno-peephole -fno-peephole2 @gol
300 -funsafe-math-optimizations -ffinite-math-only @gol
301 -fno-trapping-math -fno-zero-initialized-in-bss @gol
302 -fomit-frame-pointer -foptimize-register-move @gol
303 -foptimize-sibling-calls -fprefetch-loop-arrays @gol
304 -fprofile-generate -fprofile-use @gol
305 -fregmove -frename-registers @gol
306 -freorder-blocks -freorder-blocks-and-partition -freorder-functions @gol
307 -frerun-cse-after-loop -frerun-loop-opt @gol
308 -frounding-math -fschedule-insns -fschedule-insns2 @gol
309 -fno-sched-interblock -fno-sched-spec -fsched-spec-load @gol
310 -fsched-spec-load-dangerous @gol
311 -fsched-stalled-insns=@var{n} -sched-stalled-insns-dep=@var{n} @gol
312 -fsched2-use-superblocks @gol
313 -fsched2-use-traces -freschedule-modulo-scheduled-loops @gol
314 -fsignaling-nans -fsingle-precision-constant -fspeculative-prefetching @gol
315 -fstrength-reduce -fstrict-aliasing -ftracer -fthread-jumps @gol
316 -funroll-all-loops -funroll-loops -fpeel-loops @gol
317 -fsplit-ivs-in-unroller -funswitch-loops @gol
318 -fvariable-expansion-in-unroller @gol
319 -ftree-pre -ftree-ccp -ftree-dce -ftree-loop-optimize @gol
320 -ftree-loop-linear -ftree-loop-im -ftree-loop-ivcanon -fivopts @gol
321 -ftree-dominator-opts -ftree-dse -ftree-copyrename @gol
322 -ftree-ch -ftree-sra -ftree-ter -ftree-lrs -ftree-fre -ftree-vectorize @gol
323 --param @var{name}=@var{value}
324 -O -O0 -O1 -O2 -O3 -Os}
326 @item Preprocessor Options
327 @xref{Preprocessor Options,,Options Controlling the Preprocessor}.
328 @gccoptlist{-A@var{question}=@var{answer} @gol
329 -A-@var{question}@r{[}=@var{answer}@r{]} @gol
330 -C -dD -dI -dM -dN @gol
331 -D@var{macro}@r{[}=@var{defn}@r{]} -E -H @gol
332 -idirafter @var{dir} @gol
333 -include @var{file} -imacros @var{file} @gol
334 -iprefix @var{file} -iwithprefix @var{dir} @gol
335 -iwithprefixbefore @var{dir} -isystem @var{dir} @gol
336 -M -MM -MF -MG -MP -MQ -MT -nostdinc @gol
337 -P -fworking-directory -remap @gol
338 -trigraphs -undef -U@var{macro} -Wp,@var{option} @gol
339 -Xpreprocessor @var{option}}
341 @item Assembler Option
342 @xref{Assembler Options,,Passing Options to the Assembler}.
343 @gccoptlist{-Wa,@var{option} -Xassembler @var{option}}
346 @xref{Link Options,,Options for Linking}.
347 @gccoptlist{@var{object-file-name} -l@var{library} @gol
348 -nostartfiles -nodefaultlibs -nostdlib -pie @gol
349 -s -static -static-libgcc -shared -shared-libgcc -symbolic @gol
350 -Wl,@var{option} -Xlinker @var{option} @gol
353 @item Directory Options
354 @xref{Directory Options,,Options for Directory Search}.
355 @gccoptlist{-B@var{prefix} -I@var{dir} -iquote@var{dir} -L@var{dir} -specs=@var{file} -I-}
358 @c I wrote this xref this way to avoid overfull hbox. -- rms
359 @xref{Target Options}.
360 @gccoptlist{-V @var{version} -b @var{machine}}
362 @item Machine Dependent Options
363 @xref{Submodel Options,,Hardware Models and Configurations}.
364 @c This list is ordered alphanumerically by subsection name.
365 @c Try and put the significant identifier (CPU or system) first,
366 @c so users have a clue at guessing where the ones they want will be.
369 @gccoptlist{-EB -EL @gol
370 -mmangle-cpu -mcpu=@var{cpu} -mtext=@var{text-section} @gol
371 -mdata=@var{data-section} -mrodata=@var{readonly-data-section}}
374 @gccoptlist{-mapcs-frame -mno-apcs-frame @gol
375 -mabi=@var{name} @gol
376 -mapcs-stack-check -mno-apcs-stack-check @gol
377 -mapcs-float -mno-apcs-float @gol
378 -mapcs-reentrant -mno-apcs-reentrant @gol
379 -msched-prolog -mno-sched-prolog @gol
380 -mlittle-endian -mbig-endian -mwords-little-endian @gol
381 -mfloat-abi=@var{name} -msoft-float -mhard-float -mfpe @gol
382 -mthumb-interwork -mno-thumb-interwork @gol
383 -mcpu=@var{name} -march=@var{name} -mfpu=@var{name} @gol
384 -mstructure-size-boundary=@var{n} @gol
385 -mabort-on-noreturn @gol
386 -mlong-calls -mno-long-calls @gol
387 -msingle-pic-base -mno-single-pic-base @gol
388 -mpic-register=@var{reg} @gol
389 -mnop-fun-dllimport @gol
390 -mcirrus-fix-invalid-insns -mno-cirrus-fix-invalid-insns @gol
391 -mpoke-function-name @gol
393 -mtpcs-frame -mtpcs-leaf-frame @gol
394 -mcaller-super-interworking -mcallee-super-interworking}
397 @gccoptlist{-mmcu=@var{mcu} -msize -minit-stack=@var{n} -mno-interrupts @gol
398 -mcall-prologues -mno-tablejump -mtiny-stack -mint8}
401 @gccoptlist{-mcpu=@var{cpu} -march=@var{cpu} -mtune=@var{cpu} @gol
402 -mmax-stack-frame=@var{n} -melinux-stacksize=@var{n} @gol
403 -metrax4 -metrax100 -mpdebug -mcc-init -mno-side-effects @gol
404 -mstack-align -mdata-align -mconst-align @gol
405 -m32-bit -m16-bit -m8-bit -mno-prologue-epilogue -mno-gotplt @gol
406 -melf -maout -melinux -mlinux -sim -sim2 @gol
407 -mmul-bug-workaround -mno-mul-bug-workaround}
409 @emph{Darwin Options}
410 @gccoptlist{-all_load -allowable_client -arch -arch_errors_fatal @gol
411 -arch_only -bind_at_load -bundle -bundle_loader @gol
412 -client_name -compatibility_version -current_version @gol
414 -dependency-file -dylib_file -dylinker_install_name @gol
415 -dynamic -dynamiclib -exported_symbols_list @gol
416 -filelist -flat_namespace -force_cpusubtype_ALL @gol
417 -force_flat_namespace -headerpad_max_install_names @gol
418 -image_base -init -install_name -keep_private_externs @gol
419 -multi_module -multiply_defined -multiply_defined_unused @gol
420 -noall_load -no_dead_strip_inits_and_terms @gol
421 -nofixprebinding -nomultidefs -noprebind -noseglinkedit @gol
422 -pagezero_size -prebind -prebind_all_twolevel_modules @gol
423 -private_bundle -read_only_relocs -sectalign @gol
424 -sectobjectsymbols -whyload -seg1addr @gol
425 -sectcreate -sectobjectsymbols -sectorder @gol
426 -segaddr -segs_read_only_addr -segs_read_write_addr @gol
427 -seg_addr_table -seg_addr_table_filename -seglinkedit @gol
428 -segprot -segs_read_only_addr -segs_read_write_addr @gol
429 -single_module -static -sub_library -sub_umbrella @gol
430 -twolevel_namespace -umbrella -undefined @gol
431 -unexported_symbols_list -weak_reference_mismatches @gol
432 -whatsloaded -F -gused -gfull -mone-byte-bool}
434 @emph{DEC Alpha Options}
435 @gccoptlist{-mno-fp-regs -msoft-float -malpha-as -mgas @gol
436 -mieee -mieee-with-inexact -mieee-conformant @gol
437 -mfp-trap-mode=@var{mode} -mfp-rounding-mode=@var{mode} @gol
438 -mtrap-precision=@var{mode} -mbuild-constants @gol
439 -mcpu=@var{cpu-type} -mtune=@var{cpu-type} @gol
440 -mbwx -mmax -mfix -mcix @gol
441 -mfloat-vax -mfloat-ieee @gol
442 -mexplicit-relocs -msmall-data -mlarge-data @gol
443 -msmall-text -mlarge-text @gol
444 -mmemory-latency=@var{time}}
446 @emph{DEC Alpha/VMS Options}
447 @gccoptlist{-mvms-return-codes}
450 @gccoptlist{-mgpr-32 -mgpr-64 -mfpr-32 -mfpr-64 @gol
451 -mhard-float -msoft-float @gol
452 -malloc-cc -mfixed-cc -mdword -mno-dword @gol
453 -mdouble -mno-double @gol
454 -mmedia -mno-media -mmuladd -mno-muladd @gol
455 -mfdpic -minline-plt -mgprel-ro -multilib-library-pic @gol
456 -mlinked-fp -mlong-calls -malign-labels @gol
457 -mlibrary-pic -macc-4 -macc-8 @gol
458 -mpack -mno-pack -mno-eflags -mcond-move -mno-cond-move @gol
459 -mscc -mno-scc -mcond-exec -mno-cond-exec @gol
460 -mvliw-branch -mno-vliw-branch @gol
461 -mmulti-cond-exec -mno-multi-cond-exec -mnested-cond-exec @gol
462 -mno-nested-cond-exec -mtomcat-stats @gol
466 @emph{H8/300 Options}
467 @gccoptlist{-mrelax -mh -ms -mn -mint32 -malign-300}
470 @gccoptlist{-march=@var{architecture-type} @gol
471 -mbig-switch -mdisable-fpregs -mdisable-indexing @gol
472 -mfast-indirect-calls -mgas -mgnu-ld -mhp-ld @gol
473 -mfixed-range=@var{register-range} @gol
474 -mjump-in-delay -mlinker-opt -mlong-calls @gol
475 -mlong-load-store -mno-big-switch -mno-disable-fpregs @gol
476 -mno-disable-indexing -mno-fast-indirect-calls -mno-gas @gol
477 -mno-jump-in-delay -mno-long-load-store @gol
478 -mno-portable-runtime -mno-soft-float @gol
479 -mno-space-regs -msoft-float -mpa-risc-1-0 @gol
480 -mpa-risc-1-1 -mpa-risc-2-0 -mportable-runtime @gol
481 -mschedule=@var{cpu-type} -mspace-regs -msio -mwsio @gol
482 -munix=@var{unix-std} -nolibdld -static -threads}
484 @emph{i386 and x86-64 Options}
485 @gccoptlist{-mtune=@var{cpu-type} -march=@var{cpu-type} @gol
486 -mfpmath=@var{unit} @gol
487 -masm=@var{dialect} -mno-fancy-math-387 @gol
488 -mno-fp-ret-in-387 -msoft-float -msvr3-shlib @gol
489 -mno-wide-multiply -mrtd -malign-double @gol
490 -mpreferred-stack-boundary=@var{num} @gol
491 -mmmx -msse -msse2 -msse3 -m3dnow @gol
492 -mthreads -mno-align-stringops -minline-all-stringops @gol
493 -mpush-args -maccumulate-outgoing-args -m128bit-long-double @gol
494 -m96bit-long-double -mregparm=@var{num} -momit-leaf-frame-pointer @gol
495 -mno-red-zone -mno-tls-direct-seg-refs @gol
496 -mcmodel=@var{code-model} @gol
500 @gccoptlist{-mbig-endian -mlittle-endian -mgnu-as -mgnu-ld -mno-pic @gol
501 -mvolatile-asm-stop -mb-step -mregister-names -mno-sdata @gol
502 -mconstant-gp -mauto-pic -minline-float-divide-min-latency @gol
503 -minline-float-divide-max-throughput @gol
504 -minline-int-divide-min-latency @gol
505 -minline-int-divide-max-throughput @gol
506 -minline-sqrt-min-latency -minline-sqrt-max-throughput @gol
507 -mno-dwarf2-asm -mearly-stop-bits @gol
508 -mfixed-range=@var{register-range} -mtls-size=@var{tls-size} @gol
509 -mtune=@var{cpu-type} -mt -pthread -milp32 -mlp64}
511 @emph{M32R/D Options}
512 @gccoptlist{-m32r2 -m32rx -m32r @gol
514 -malign-loops -mno-align-loops @gol
515 -missue-rate=@var{number} @gol
516 -mbranch-cost=@var{number} @gol
517 -mmodel=@var{code-size-model-type} @gol
518 -msdata=@var{sdata-type} @gol
519 -mno-flush-func -mflush-func=@var{name} @gol
520 -mno-flush-trap -mflush-trap=@var{number} @gol
523 @emph{M680x0 Options}
524 @gccoptlist{-m68000 -m68020 -m68020-40 -m68020-60 -m68030 -m68040 @gol
525 -m68060 -mcpu32 -m5200 -m68881 -mbitfield -mc68000 -mc68020 @gol
526 -mnobitfield -mrtd -mshort -msoft-float -mpcrel @gol
527 -malign-int -mstrict-align -msep-data -mno-sep-data @gol
528 -mshared-library-id=n -mid-shared-library -mno-id-shared-library}
530 @emph{M68hc1x Options}
531 @gccoptlist{-m6811 -m6812 -m68hc11 -m68hc12 -m68hcs12 @gol
532 -mauto-incdec -minmax -mlong-calls -mshort @gol
533 -msoft-reg-count=@var{count}}
536 @gccoptlist{-mhardlit -mno-hardlit -mdiv -mno-div -mrelax-immediates @gol
537 -mno-relax-immediates -mwide-bitfields -mno-wide-bitfields @gol
538 -m4byte-functions -mno-4byte-functions -mcallgraph-data @gol
539 -mno-callgraph-data -mslow-bytes -mno-slow-bytes -mno-lsim @gol
540 -mlittle-endian -mbig-endian -m210 -m340 -mstack-increment}
543 @gccoptlist{-EL -EB -march=@var{arch} -mtune=@var{arch} @gol
544 -mips1 -mips2 -mips3 -mips4 -mips32 -mips32r2 -mips64 @gol
545 -mips16 -mno-mips16 -mabi=@var{abi} -mabicalls -mno-abicalls @gol
546 -mxgot -mno-xgot -mgp32 -mgp64 -mfp32 -mfp64 @gol
547 -mhard-float -msoft-float -msingle-float -mdouble-float @gol
548 -mpaired-single -mips3d @gol
549 -mint64 -mlong64 -mlong32 @gol
550 -G@var{num} -membedded-data -mno-embedded-data @gol
551 -muninit-const-in-rodata -mno-uninit-const-in-rodata @gol
552 -msplit-addresses -mno-split-addresses @gol
553 -mexplicit-relocs -mno-explicit-relocs @gol
554 -mcheck-zero-division -mno-check-zero-division @gol
555 -mdivide-traps -mdivide-breaks @gol
556 -mmemcpy -mno-memcpy -mlong-calls -mno-long-calls @gol
557 -mmad -mno-mad -mfused-madd -mno-fused-madd -nocpp @gol
558 -mfix-r4000 -mno-fix-r4000 -mfix-r4400 -mno-fix-r4400 @gol
559 -mfix-vr4120 -mno-fix-vr4120 -mfix-sb1 -mno-fix-sb1 @gol
560 -mflush-func=@var{func} -mno-flush-func @gol
561 -mbranch-likely -mno-branch-likely @gol
562 -mfp-exceptions -mno-fp-exceptions @gol
563 -mvr4130-align -mno-vr4130-align}
566 @gccoptlist{-mlibfuncs -mno-libfuncs -mepsilon -mno-epsilon -mabi=gnu @gol
567 -mabi=mmixware -mzero-extend -mknuthdiv -mtoplevel-symbols @gol
568 -melf -mbranch-predict -mno-branch-predict -mbase-addresses @gol
569 -mno-base-addresses -msingle-exit -mno-single-exit}
571 @emph{MN10300 Options}
572 @gccoptlist{-mmult-bug -mno-mult-bug @gol
573 -mam33 -mno-am33 @gol
574 -mam33-2 -mno-am33-2 @gol
578 @gccoptlist{-m32032 -m32332 -m32532 -m32081 -m32381 @gol
579 -mmult-add -mnomult-add -msoft-float -mrtd -mnortd @gol
580 -mregparam -mnoregparam -msb -mnosb @gol
581 -mbitfield -mnobitfield -mhimem -mnohimem}
583 @emph{PDP-11 Options}
584 @gccoptlist{-mfpu -msoft-float -mac0 -mno-ac0 -m40 -m45 -m10 @gol
585 -mbcopy -mbcopy-builtin -mint32 -mno-int16 @gol
586 -mint16 -mno-int32 -mfloat32 -mno-float64 @gol
587 -mfloat64 -mno-float32 -mabshi -mno-abshi @gol
588 -mbranch-expensive -mbranch-cheap @gol
589 -msplit -mno-split -munix-asm -mdec-asm}
591 @emph{PowerPC Options}
592 See RS/6000 and PowerPC Options.
594 @emph{RS/6000 and PowerPC Options}
595 @gccoptlist{-mcpu=@var{cpu-type} @gol
596 -mtune=@var{cpu-type} @gol
597 -mpower -mno-power -mpower2 -mno-power2 @gol
598 -mpowerpc -mpowerpc64 -mno-powerpc @gol
599 -maltivec -mno-altivec @gol
600 -mpowerpc-gpopt -mno-powerpc-gpopt @gol
601 -mpowerpc-gfxopt -mno-powerpc-gfxopt @gol
602 -mnew-mnemonics -mold-mnemonics @gol
603 -mfull-toc -mminimal-toc -mno-fp-in-toc -mno-sum-in-toc @gol
604 -m64 -m32 -mxl-call -mno-xl-call -mpe @gol
605 -malign-power -malign-natural @gol
606 -msoft-float -mhard-float -mmultiple -mno-multiple @gol
607 -mstring -mno-string -mupdate -mno-update @gol
608 -mfused-madd -mno-fused-madd -mbit-align -mno-bit-align @gol
609 -mstrict-align -mno-strict-align -mrelocatable @gol
610 -mno-relocatable -mrelocatable-lib -mno-relocatable-lib @gol
611 -mtoc -mno-toc -mlittle -mlittle-endian -mbig -mbig-endian @gol
612 -mdynamic-no-pic @gol
613 -mprioritize-restricted-insns=@var{priority} @gol
614 -msched-costly-dep=@var{dependence_type} @gol
615 -minsert-sched-nops=@var{scheme} @gol
616 -mcall-sysv -mcall-netbsd @gol
617 -maix-struct-return -msvr4-struct-return @gol
618 -mabi=altivec -mabi=no-altivec @gol
619 -mabi=spe -mabi=no-spe @gol
620 -misel=yes -misel=no @gol
621 -mspe=yes -mspe=no @gol
622 -mfloat-gprs=yes -mfloat-gprs=no -mfloat-gprs=single -mfloat-gprs=double @gol
623 -mprototype -mno-prototype @gol
624 -msim -mmvme -mads -myellowknife -memb -msdata @gol
625 -msdata=@var{opt} -mvxworks -mwindiss -G @var{num} -pthread}
627 @emph{S/390 and zSeries Options}
628 @gccoptlist{-mtune=@var{cpu-type} -march=@var{cpu-type} @gol
629 -mhard-float -msoft-float -mbackchain -mno-backchain @gol
630 -mpacked-stack -mno-packed-stack @gol
631 -msmall-exec -mno-small-exec -mmvcle -mno-mvcle @gol
632 -m64 -m31 -mdebug -mno-debug -mesa -mzarch @gol
633 -mtpf-trace -mno-tpf-trace -mfused-madd -mno-fused-madd @gol
634 -mwarn-framesize -mwarn-dynamicstack -mstack-size -mstack-guard}
637 @gccoptlist{-m1 -m2 -m2e -m3 -m3e @gol
638 -m4-nofpu -m4-single-only -m4-single -m4 @gol
639 -m4a-nofpu -m4a-single-only -m4a-single -m4a -m4al @gol
640 -m5-64media -m5-64media-nofpu @gol
641 -m5-32media -m5-32media-nofpu @gol
642 -m5-compact -m5-compact-nofpu @gol
643 -mb -ml -mdalign -mrelax @gol
644 -mbigtable -mfmovd -mhitachi -mrenesas -mno-renesas -mnomacsave @gol
645 -mieee -misize -mpadstruct -mspace @gol
646 -mprefergot -musermode}
649 @gccoptlist{-mcpu=@var{cpu-type} @gol
650 -mtune=@var{cpu-type} @gol
651 -mcmodel=@var{code-model} @gol
652 -m32 -m64 -mapp-regs -mno-app-regs @gol
653 -mfaster-structs -mno-faster-structs @gol
654 -mfpu -mno-fpu -mhard-float -msoft-float @gol
655 -mhard-quad-float -msoft-quad-float @gol
656 -mimpure-text -mno-impure-text -mlittle-endian @gol
657 -mstack-bias -mno-stack-bias @gol
658 -munaligned-doubles -mno-unaligned-doubles @gol
659 -mv8plus -mno-v8plus -mvis -mno-vis
662 @emph{System V Options}
663 @gccoptlist{-Qy -Qn -YP,@var{paths} -Ym,@var{dir}}
665 @emph{TMS320C3x/C4x Options}
666 @gccoptlist{-mcpu=@var{cpu} -mbig -msmall -mregparm -mmemparm @gol
667 -mfast-fix -mmpyi -mbk -mti -mdp-isr-reload @gol
668 -mrpts=@var{count} -mrptb -mdb -mloop-unsigned @gol
669 -mparallel-insns -mparallel-mpy -mpreserve-float}
672 @gccoptlist{-mlong-calls -mno-long-calls -mep -mno-ep @gol
673 -mprolog-function -mno-prolog-function -mspace @gol
674 -mtda=@var{n} -msda=@var{n} -mzda=@var{n} @gol
675 -mapp-regs -mno-app-regs @gol
676 -mdisable-callt -mno-disable-callt @gol
682 @gccoptlist{-mg -mgnu -munix}
684 @emph{x86-64 Options}
685 See i386 and x86-64 Options.
687 @emph{Xstormy16 Options}
690 @emph{Xtensa Options}
691 @gccoptlist{-mconst16 -mno-const16 @gol
692 -mfused-madd -mno-fused-madd @gol
693 -mtext-section-literals -mno-text-section-literals @gol
694 -mtarget-align -mno-target-align @gol
695 -mlongcalls -mno-longcalls}
697 @emph{zSeries Options}
698 See S/390 and zSeries Options.
700 @item Code Generation Options
701 @xref{Code Gen Options,,Options for Code Generation Conventions}.
702 @gccoptlist{-fcall-saved-@var{reg} -fcall-used-@var{reg} @gol
703 -ffixed-@var{reg} -fexceptions @gol
704 -fnon-call-exceptions -funwind-tables @gol
705 -fasynchronous-unwind-tables @gol
706 -finhibit-size-directive -finstrument-functions @gol
707 -fno-common -fno-ident @gol
708 -fpcc-struct-return -fpic -fPIC -fpie -fPIE @gol
709 -freg-struct-return -fshared-data -fshort-enums @gol
710 -fshort-double -fshort-wchar @gol
711 -fverbose-asm -fpack-struct[=@var{n}] -fstack-check @gol
712 -fstack-limit-register=@var{reg} -fstack-limit-symbol=@var{sym} @gol
713 -fargument-alias -fargument-noalias @gol
714 -fargument-noalias-global -fleading-underscore @gol
715 -ftls-model=@var{model} @gol
716 -ftrapv -fwrapv -fbounds-check @gol
721 * Overall Options:: Controlling the kind of output:
722 an executable, object files, assembler files,
723 or preprocessed source.
724 * C Dialect Options:: Controlling the variant of C language compiled.
725 * C++ Dialect Options:: Variations on C++.
726 * Objective-C and Objective-C++ Dialect Options:: Variations on Objective-C
728 * Language Independent Options:: Controlling how diagnostics should be
730 * Warning Options:: How picky should the compiler be?
731 * Debugging Options:: Symbol tables, measurements, and debugging dumps.
732 * Optimize Options:: How much optimization?
733 * Preprocessor Options:: Controlling header files and macro definitions.
734 Also, getting dependency information for Make.
735 * Assembler Options:: Passing options to the assembler.
736 * Link Options:: Specifying libraries and so on.
737 * Directory Options:: Where to find header files and libraries.
738 Where to find the compiler executable files.
739 * Spec Files:: How to pass switches to sub-processes.
740 * Target Options:: Running a cross-compiler, or an old version of GCC.
743 @node Overall Options
744 @section Options Controlling the Kind of Output
746 Compilation can involve up to four stages: preprocessing, compilation
747 proper, assembly and linking, always in that order. GCC is capable of
748 preprocessing and compiling several files either into several
749 assembler input files, or into one assembler input file; then each
750 assembler input file produces an object file, and linking combines all
751 the object files (those newly compiled, and those specified as input)
752 into an executable file.
754 @cindex file name suffix
755 For any given input file, the file name suffix determines what kind of
760 C source code which must be preprocessed.
763 C source code which should not be preprocessed.
766 C++ source code which should not be preprocessed.
769 Objective-C source code. Note that you must link with the @file{libobjc}
770 library to make an Objective-C program work.
773 Objective-C source code which should not be preprocessed.
777 Objective-C++ source code. Note that you must link with the @file{libobjc}
778 library to make an Objective-C++ program work. Note that @samp{.M} refers
779 to a literal capital M@.
782 Objective-C++ source code which should not be preprocessed.
785 C, C++, Objective-C or Objective-C++ header file to be turned into a
790 @itemx @var{file}.cxx
791 @itemx @var{file}.cpp
792 @itemx @var{file}.CPP
793 @itemx @var{file}.c++
795 C++ source code which must be preprocessed. Note that in @samp{.cxx},
796 the last two letters must both be literally @samp{x}. Likewise,
797 @samp{.C} refers to a literal capital C@.
801 C++ header file to be turned into a precompiled header.
804 @itemx @var{file}.for
805 @itemx @var{file}.FOR
806 Fortran source code which should not be preprocessed.
809 @itemx @var{file}.fpp
810 @itemx @var{file}.FPP
811 Fortran source code which must be preprocessed (with the traditional
815 Fortran source code which must be preprocessed with a RATFOR
816 preprocessor (not included with GCC)@.
819 @itemx @var{file}.f95
820 Fortran 90/95 source code which should not be preprocessed.
822 @c FIXME: Descriptions of Java file types.
829 Ada source code file which contains a library unit declaration (a
830 declaration of a package, subprogram, or generic, or a generic
831 instantiation), or a library unit renaming declaration (a package,
832 generic, or subprogram renaming declaration). Such files are also
835 @itemx @var{file}.adb
836 Ada source code file containing a library unit body (a subprogram or
837 package body). Such files are also called @dfn{bodies}.
839 @c GCC also knows about some suffixes for languages not yet included:
848 Assembler code which must be preprocessed.
851 An object file to be fed straight into linking.
852 Any file name with no recognized suffix is treated this way.
856 You can specify the input language explicitly with the @option{-x} option:
859 @item -x @var{language}
860 Specify explicitly the @var{language} for the following input files
861 (rather than letting the compiler choose a default based on the file
862 name suffix). This option applies to all following input files until
863 the next @option{-x} option. Possible values for @var{language} are:
865 c c-header c-cpp-output
866 c++ c++-header c++-cpp-output
867 objective-c objective-c-header objective-c-cpp-output
868 objective-c++ objective-c++-header objective-c++-cpp-output
869 assembler assembler-with-cpp
871 f77 f77-cpp-input ratfor
878 Turn off any specification of a language, so that subsequent files are
879 handled according to their file name suffixes (as they are if @option{-x}
880 has not been used at all).
882 @item -pass-exit-codes
883 @opindex pass-exit-codes
884 Normally the @command{gcc} program will exit with the code of 1 if any
885 phase of the compiler returns a non-success return code. If you specify
886 @option{-pass-exit-codes}, the @command{gcc} program will instead return with
887 numerically highest error produced by any phase that returned an error
891 If you only want some of the stages of compilation, you can use
892 @option{-x} (or filename suffixes) to tell @command{gcc} where to start, and
893 one of the options @option{-c}, @option{-S}, or @option{-E} to say where
894 @command{gcc} is to stop. Note that some combinations (for example,
895 @samp{-x cpp-output -E}) instruct @command{gcc} to do nothing at all.
900 Compile or assemble the source files, but do not link. The linking
901 stage simply is not done. The ultimate output is in the form of an
902 object file for each source file.
904 By default, the object file name for a source file is made by replacing
905 the suffix @samp{.c}, @samp{.i}, @samp{.s}, etc., with @samp{.o}.
907 Unrecognized input files, not requiring compilation or assembly, are
912 Stop after the stage of compilation proper; do not assemble. The output
913 is in the form of an assembler code file for each non-assembler input
916 By default, the assembler file name for a source file is made by
917 replacing the suffix @samp{.c}, @samp{.i}, etc., with @samp{.s}.
919 Input files that don't require compilation are ignored.
923 Stop after the preprocessing stage; do not run the compiler proper. The
924 output is in the form of preprocessed source code, which is sent to the
927 Input files which don't require preprocessing are ignored.
929 @cindex output file option
932 Place output in file @var{file}. This applies regardless to whatever
933 sort of output is being produced, whether it be an executable file,
934 an object file, an assembler file or preprocessed C code.
936 If @option{-o} is not specified, the default is to put an executable
937 file in @file{a.out}, the object file for
938 @file{@var{source}.@var{suffix}} in @file{@var{source}.o}, its
939 assembler file in @file{@var{source}.s}, a precompiled header file in
940 @file{@var{source}.@var{suffix}.gch}, and all preprocessed C source on
945 Print (on standard error output) the commands executed to run the stages
946 of compilation. Also print the version number of the compiler driver
947 program and of the preprocessor and the compiler proper.
951 Like @option{-v} except the commands are not executed and all command
952 arguments are quoted. This is useful for shell scripts to capture the
953 driver-generated command lines.
957 Use pipes rather than temporary files for communication between the
958 various stages of compilation. This fails to work on some systems where
959 the assembler is unable to read from a pipe; but the GNU assembler has
964 If you are compiling multiple source files, this option tells the driver
965 to pass all the source files to the compiler at once (for those
966 languages for which the compiler can handle this). This will allow
967 intermodule analysis (IMA) to be performed by the compiler. Currently the only
968 language for which this is supported is C@. If you pass source files for
969 multiple languages to the driver, using this option, the driver will invoke
970 the compiler(s) that support IMA once each, passing each compiler all the
971 source files appropriate for it. For those languages that do not support
972 IMA this option will be ignored, and the compiler will be invoked once for
973 each source file in that language. If you use this option in conjunction
974 with @option{-save-temps}, the compiler will generate multiple
976 (one for each source file), but only one (combined) @file{.o} or
981 Print (on the standard output) a description of the command line options
982 understood by @command{gcc}. If the @option{-v} option is also specified
983 then @option{--help} will also be passed on to the various processes
984 invoked by @command{gcc}, so that they can display the command line options
985 they accept. If the @option{-Wextra} option is also specified then command
986 line options which have no documentation associated with them will also
991 Print (on the standard output) a description of target specific command
992 line options for each tool.
996 Display the version number and copyrights of the invoked GCC@.
1000 @section Compiling C++ Programs
1002 @cindex suffixes for C++ source
1003 @cindex C++ source file suffixes
1004 C++ source files conventionally use one of the suffixes @samp{.C},
1005 @samp{.cc}, @samp{.cpp}, @samp{.CPP}, @samp{.c++}, @samp{.cp}, or
1006 @samp{.cxx}; C++ header files often use @samp{.hh} or @samp{.H}; and
1007 preprocessed C++ files use the suffix @samp{.ii}. GCC recognizes
1008 files with these names and compiles them as C++ programs even if you
1009 call the compiler the same way as for compiling C programs (usually
1010 with the name @command{gcc}).
1014 However, C++ programs often require class libraries as well as a
1015 compiler that understands the C++ language---and under some
1016 circumstances, you might want to compile programs or header files from
1017 standard input, or otherwise without a suffix that flags them as C++
1018 programs. You might also like to precompile a C header file with a
1019 @samp{.h} extension to be used in C++ compilations. @command{g++} is a
1020 program that calls GCC with the default language set to C++, and
1021 automatically specifies linking against the C++ library. On many
1022 systems, @command{g++} is also installed with the name @command{c++}.
1024 @cindex invoking @command{g++}
1025 When you compile C++ programs, you may specify many of the same
1026 command-line options that you use for compiling programs in any
1027 language; or command-line options meaningful for C and related
1028 languages; or options that are meaningful only for C++ programs.
1029 @xref{C Dialect Options,,Options Controlling C Dialect}, for
1030 explanations of options for languages related to C@.
1031 @xref{C++ Dialect Options,,Options Controlling C++ Dialect}, for
1032 explanations of options that are meaningful only for C++ programs.
1034 @node C Dialect Options
1035 @section Options Controlling C Dialect
1036 @cindex dialect options
1037 @cindex language dialect options
1038 @cindex options, dialect
1040 The following options control the dialect of C (or languages derived
1041 from C, such as C++, Objective-C and Objective-C++) that the compiler
1045 @cindex ANSI support
1049 In C mode, support all ISO C90 programs. In C++ mode,
1050 remove GNU extensions that conflict with ISO C++.
1052 This turns off certain features of GCC that are incompatible with ISO
1053 C90 (when compiling C code), or of standard C++ (when compiling C++ code),
1054 such as the @code{asm} and @code{typeof} keywords, and
1055 predefined macros such as @code{unix} and @code{vax} that identify the
1056 type of system you are using. It also enables the undesirable and
1057 rarely used ISO trigraph feature. For the C compiler,
1058 it disables recognition of C++ style @samp{//} comments as well as
1059 the @code{inline} keyword.
1061 The alternate keywords @code{__asm__}, @code{__extension__},
1062 @code{__inline__} and @code{__typeof__} continue to work despite
1063 @option{-ansi}. You would not want to use them in an ISO C program, of
1064 course, but it is useful to put them in header files that might be included
1065 in compilations done with @option{-ansi}. Alternate predefined macros
1066 such as @code{__unix__} and @code{__vax__} are also available, with or
1067 without @option{-ansi}.
1069 The @option{-ansi} option does not cause non-ISO programs to be
1070 rejected gratuitously. For that, @option{-pedantic} is required in
1071 addition to @option{-ansi}. @xref{Warning Options}.
1073 The macro @code{__STRICT_ANSI__} is predefined when the @option{-ansi}
1074 option is used. Some header files may notice this macro and refrain
1075 from declaring certain functions or defining certain macros that the
1076 ISO standard doesn't call for; this is to avoid interfering with any
1077 programs that might use these names for other things.
1079 Functions which would normally be built in but do not have semantics
1080 defined by ISO C (such as @code{alloca} and @code{ffs}) are not built-in
1081 functions with @option{-ansi} is used. @xref{Other Builtins,,Other
1082 built-in functions provided by GCC}, for details of the functions
1087 Determine the language standard. This option is currently only
1088 supported when compiling C or C++. A value for this option must be
1089 provided; possible values are
1094 ISO C90 (same as @option{-ansi}).
1096 @item iso9899:199409
1097 ISO C90 as modified in amendment 1.
1103 ISO C99. Note that this standard is not yet fully supported; see
1104 @w{@uref{http://gcc.gnu.org/c99status.html}} for more information. The
1105 names @samp{c9x} and @samp{iso9899:199x} are deprecated.
1108 Default, ISO C90 plus GNU extensions (including some C99 features).
1112 ISO C99 plus GNU extensions. When ISO C99 is fully implemented in GCC,
1113 this will become the default. The name @samp{gnu9x} is deprecated.
1116 The 1998 ISO C++ standard plus amendments.
1119 The same as @option{-std=c++98} plus GNU extensions. This is the
1120 default for C++ code.
1123 Even when this option is not specified, you can still use some of the
1124 features of newer standards in so far as they do not conflict with
1125 previous C standards. For example, you may use @code{__restrict__} even
1126 when @option{-std=c99} is not specified.
1128 The @option{-std} options specifying some version of ISO C have the same
1129 effects as @option{-ansi}, except that features that were not in ISO C90
1130 but are in the specified version (for example, @samp{//} comments and
1131 the @code{inline} keyword in ISO C99) are not disabled.
1133 @xref{Standards,,Language Standards Supported by GCC}, for details of
1134 these standard versions.
1136 @item -aux-info @var{filename}
1138 Output to the given filename prototyped declarations for all functions
1139 declared and/or defined in a translation unit, including those in header
1140 files. This option is silently ignored in any language other than C@.
1142 Besides declarations, the file indicates, in comments, the origin of
1143 each declaration (source file and line), whether the declaration was
1144 implicit, prototyped or unprototyped (@samp{I}, @samp{N} for new or
1145 @samp{O} for old, respectively, in the first character after the line
1146 number and the colon), and whether it came from a declaration or a
1147 definition (@samp{C} or @samp{F}, respectively, in the following
1148 character). In the case of function definitions, a K&R-style list of
1149 arguments followed by their declarations is also provided, inside
1150 comments, after the declaration.
1154 Do not recognize @code{asm}, @code{inline} or @code{typeof} as a
1155 keyword, so that code can use these words as identifiers. You can use
1156 the keywords @code{__asm__}, @code{__inline__} and @code{__typeof__}
1157 instead. @option{-ansi} implies @option{-fno-asm}.
1159 In C++, this switch only affects the @code{typeof} keyword, since
1160 @code{asm} and @code{inline} are standard keywords. You may want to
1161 use the @option{-fno-gnu-keywords} flag instead, which has the same
1162 effect. In C99 mode (@option{-std=c99} or @option{-std=gnu99}), this
1163 switch only affects the @code{asm} and @code{typeof} keywords, since
1164 @code{inline} is a standard keyword in ISO C99.
1167 @itemx -fno-builtin-@var{function}
1168 @opindex fno-builtin
1169 @cindex built-in functions
1170 Don't recognize built-in functions that do not begin with
1171 @samp{__builtin_} as prefix. @xref{Other Builtins,,Other built-in
1172 functions provided by GCC}, for details of the functions affected,
1173 including those which are not built-in functions when @option{-ansi} or
1174 @option{-std} options for strict ISO C conformance are used because they
1175 do not have an ISO standard meaning.
1177 GCC normally generates special code to handle certain built-in functions
1178 more efficiently; for instance, calls to @code{alloca} may become single
1179 instructions that adjust the stack directly, and calls to @code{memcpy}
1180 may become inline copy loops. The resulting code is often both smaller
1181 and faster, but since the function calls no longer appear as such, you
1182 cannot set a breakpoint on those calls, nor can you change the behavior
1183 of the functions by linking with a different library. In addition,
1184 when a function is recognized as a built-in function, GCC may use
1185 information about that function to warn about problems with calls to
1186 that function, or to generate more efficient code, even if the
1187 resulting code still contains calls to that function. For example,
1188 warnings are given with @option{-Wformat} for bad calls to
1189 @code{printf}, when @code{printf} is built in, and @code{strlen} is
1190 known not to modify global memory.
1192 With the @option{-fno-builtin-@var{function}} option
1193 only the built-in function @var{function} is
1194 disabled. @var{function} must not begin with @samp{__builtin_}. If a
1195 function is named this is not built-in in this version of GCC, this
1196 option is ignored. There is no corresponding
1197 @option{-fbuiltin-@var{function}} option; if you wish to enable
1198 built-in functions selectively when using @option{-fno-builtin} or
1199 @option{-ffreestanding}, you may define macros such as:
1202 #define abs(n) __builtin_abs ((n))
1203 #define strcpy(d, s) __builtin_strcpy ((d), (s))
1208 @cindex hosted environment
1210 Assert that compilation takes place in a hosted environment. This implies
1211 @option{-fbuiltin}. A hosted environment is one in which the
1212 entire standard library is available, and in which @code{main} has a return
1213 type of @code{int}. Examples are nearly everything except a kernel.
1214 This is equivalent to @option{-fno-freestanding}.
1216 @item -ffreestanding
1217 @opindex ffreestanding
1218 @cindex hosted environment
1220 Assert that compilation takes place in a freestanding environment. This
1221 implies @option{-fno-builtin}. A freestanding environment
1222 is one in which the standard library may not exist, and program startup may
1223 not necessarily be at @code{main}. The most obvious example is an OS kernel.
1224 This is equivalent to @option{-fno-hosted}.
1226 @xref{Standards,,Language Standards Supported by GCC}, for details of
1227 freestanding and hosted environments.
1229 @item -fms-extensions
1230 @opindex fms-extensions
1231 Accept some non-standard constructs used in Microsoft header files.
1233 Some cases of unnamed fields in structures and unions are only
1234 accepted with this option. @xref{Unnamed Fields,,Unnamed struct/union
1235 fields within structs/unions}, for details.
1239 Support ISO C trigraphs. The @option{-ansi} option (and @option{-std}
1240 options for strict ISO C conformance) implies @option{-trigraphs}.
1242 @item -no-integrated-cpp
1243 @opindex no-integrated-cpp
1244 Performs a compilation in two passes: preprocessing and compiling. This
1245 option allows a user supplied "cc1", "cc1plus", or "cc1obj" via the
1246 @option{-B} option. The user supplied compilation step can then add in
1247 an additional preprocessing step after normal preprocessing but before
1248 compiling. The default is to use the integrated cpp (internal cpp)
1250 The semantics of this option will change if "cc1", "cc1plus", and
1251 "cc1obj" are merged.
1253 @cindex traditional C language
1254 @cindex C language, traditional
1256 @itemx -traditional-cpp
1257 @opindex traditional-cpp
1258 @opindex traditional
1259 Formerly, these options caused GCC to attempt to emulate a pre-standard
1260 C compiler. They are now only supported with the @option{-E} switch.
1261 The preprocessor continues to support a pre-standard mode. See the GNU
1262 CPP manual for details.
1264 @item -fcond-mismatch
1265 @opindex fcond-mismatch
1266 Allow conditional expressions with mismatched types in the second and
1267 third arguments. The value of such an expression is void. This option
1268 is not supported for C++.
1270 @item -funsigned-char
1271 @opindex funsigned-char
1272 Let the type @code{char} be unsigned, like @code{unsigned char}.
1274 Each kind of machine has a default for what @code{char} should
1275 be. It is either like @code{unsigned char} by default or like
1276 @code{signed char} by default.
1278 Ideally, a portable program should always use @code{signed char} or
1279 @code{unsigned char} when it depends on the signedness of an object.
1280 But many programs have been written to use plain @code{char} and
1281 expect it to be signed, or expect it to be unsigned, depending on the
1282 machines they were written for. This option, and its inverse, let you
1283 make such a program work with the opposite default.
1285 The type @code{char} is always a distinct type from each of
1286 @code{signed char} or @code{unsigned char}, even though its behavior
1287 is always just like one of those two.
1290 @opindex fsigned-char
1291 Let the type @code{char} be signed, like @code{signed char}.
1293 Note that this is equivalent to @option{-fno-unsigned-char}, which is
1294 the negative form of @option{-funsigned-char}. Likewise, the option
1295 @option{-fno-signed-char} is equivalent to @option{-funsigned-char}.
1297 @item -fsigned-bitfields
1298 @itemx -funsigned-bitfields
1299 @itemx -fno-signed-bitfields
1300 @itemx -fno-unsigned-bitfields
1301 @opindex fsigned-bitfields
1302 @opindex funsigned-bitfields
1303 @opindex fno-signed-bitfields
1304 @opindex fno-unsigned-bitfields
1305 These options control whether a bit-field is signed or unsigned, when the
1306 declaration does not use either @code{signed} or @code{unsigned}. By
1307 default, such a bit-field is signed, because this is consistent: the
1308 basic integer types such as @code{int} are signed types.
1311 @node C++ Dialect Options
1312 @section Options Controlling C++ Dialect
1314 @cindex compiler options, C++
1315 @cindex C++ options, command line
1316 @cindex options, C++
1317 This section describes the command-line options that are only meaningful
1318 for C++ programs; but you can also use most of the GNU compiler options
1319 regardless of what language your program is in. For example, you
1320 might compile a file @code{firstClass.C} like this:
1323 g++ -g -frepo -O -c firstClass.C
1327 In this example, only @option{-frepo} is an option meant
1328 only for C++ programs; you can use the other options with any
1329 language supported by GCC@.
1331 Here is a list of options that are @emph{only} for compiling C++ programs:
1335 @item -fabi-version=@var{n}
1336 @opindex fabi-version
1337 Use version @var{n} of the C++ ABI@. Version 2 is the version of the
1338 C++ ABI that first appeared in G++ 3.4. Version 1 is the version of
1339 the C++ ABI that first appeared in G++ 3.2. Version 0 will always be
1340 the version that conforms most closely to the C++ ABI specification.
1341 Therefore, the ABI obtained using version 0 will change as ABI bugs
1344 The default is version 2.
1346 @item -fno-access-control
1347 @opindex fno-access-control
1348 Turn off all access checking. This switch is mainly useful for working
1349 around bugs in the access control code.
1353 Check that the pointer returned by @code{operator new} is non-null
1354 before attempting to modify the storage allocated. This check is
1355 normally unnecessary because the C++ standard specifies that
1356 @code{operator new} will only return @code{0} if it is declared
1357 @samp{throw()}, in which case the compiler will always check the
1358 return value even without this option. In all other cases, when
1359 @code{operator new} has a non-empty exception specification, memory
1360 exhaustion is signalled by throwing @code{std::bad_alloc}. See also
1361 @samp{new (nothrow)}.
1363 @item -fconserve-space
1364 @opindex fconserve-space
1365 Put uninitialized or runtime-initialized global variables into the
1366 common segment, as C does. This saves space in the executable at the
1367 cost of not diagnosing duplicate definitions. If you compile with this
1368 flag and your program mysteriously crashes after @code{main()} has
1369 completed, you may have an object that is being destroyed twice because
1370 two definitions were merged.
1372 This option is no longer useful on most targets, now that support has
1373 been added for putting variables into BSS without making them common.
1375 @item -fno-const-strings
1376 @opindex fno-const-strings
1377 Give string constants type @code{char *} instead of type @code{const
1378 char *}. By default, G++ uses type @code{const char *} as required by
1379 the standard. Even if you use @option{-fno-const-strings}, you cannot
1380 actually modify the value of a string constant.
1382 This option might be removed in a future release of G++. For maximum
1383 portability, you should structure your code so that it works with
1384 string constants that have type @code{const char *}.
1386 @item -fno-elide-constructors
1387 @opindex fno-elide-constructors
1388 The C++ standard allows an implementation to omit creating a temporary
1389 which is only used to initialize another object of the same type.
1390 Specifying this option disables that optimization, and forces G++ to
1391 call the copy constructor in all cases.
1393 @item -fno-enforce-eh-specs
1394 @opindex fno-enforce-eh-specs
1395 Don't check for violation of exception specifications at runtime. This
1396 option violates the C++ standard, but may be useful for reducing code
1397 size in production builds, much like defining @samp{NDEBUG}. The compiler
1398 will still optimize based on the exception specifications.
1401 @itemx -fno-for-scope
1403 @opindex fno-for-scope
1404 If @option{-ffor-scope} is specified, the scope of variables declared in
1405 a @i{for-init-statement} is limited to the @samp{for} loop itself,
1406 as specified by the C++ standard.
1407 If @option{-fno-for-scope} is specified, the scope of variables declared in
1408 a @i{for-init-statement} extends to the end of the enclosing scope,
1409 as was the case in old versions of G++, and other (traditional)
1410 implementations of C++.
1412 The default if neither flag is given to follow the standard,
1413 but to allow and give a warning for old-style code that would
1414 otherwise be invalid, or have different behavior.
1416 @item -fno-gnu-keywords
1417 @opindex fno-gnu-keywords
1418 Do not recognize @code{typeof} as a keyword, so that code can use this
1419 word as an identifier. You can use the keyword @code{__typeof__} instead.
1420 @option{-ansi} implies @option{-fno-gnu-keywords}.
1422 @item -fno-implicit-templates
1423 @opindex fno-implicit-templates
1424 Never emit code for non-inline templates which are instantiated
1425 implicitly (i.e.@: by use); only emit code for explicit instantiations.
1426 @xref{Template Instantiation}, for more information.
1428 @item -fno-implicit-inline-templates
1429 @opindex fno-implicit-inline-templates
1430 Don't emit code for implicit instantiations of inline templates, either.
1431 The default is to handle inlines differently so that compiles with and
1432 without optimization will need the same set of explicit instantiations.
1434 @item -fno-implement-inlines
1435 @opindex fno-implement-inlines
1436 To save space, do not emit out-of-line copies of inline functions
1437 controlled by @samp{#pragma implementation}. This will cause linker
1438 errors if these functions are not inlined everywhere they are called.
1440 @item -fms-extensions
1441 @opindex fms-extensions
1442 Disable pedantic warnings about constructs used in MFC, such as implicit
1443 int and getting a pointer to member function via non-standard syntax.
1445 @item -fno-nonansi-builtins
1446 @opindex fno-nonansi-builtins
1447 Disable built-in declarations of functions that are not mandated by
1448 ANSI/ISO C@. These include @code{ffs}, @code{alloca}, @code{_exit},
1449 @code{index}, @code{bzero}, @code{conjf}, and other related functions.
1451 @item -fno-operator-names
1452 @opindex fno-operator-names
1453 Do not treat the operator name keywords @code{and}, @code{bitand},
1454 @code{bitor}, @code{compl}, @code{not}, @code{or} and @code{xor} as
1455 synonyms as keywords.
1457 @item -fno-optional-diags
1458 @opindex fno-optional-diags
1459 Disable diagnostics that the standard says a compiler does not need to
1460 issue. Currently, the only such diagnostic issued by G++ is the one for
1461 a name having multiple meanings within a class.
1464 @opindex fpermissive
1465 Downgrade some diagnostics about nonconformant code from errors to
1466 warnings. Thus, using @option{-fpermissive} will allow some
1467 nonconforming code to compile.
1471 Enable automatic template instantiation at link time. This option also
1472 implies @option{-fno-implicit-templates}. @xref{Template
1473 Instantiation}, for more information.
1477 Disable generation of information about every class with virtual
1478 functions for use by the C++ runtime type identification features
1479 (@samp{dynamic_cast} and @samp{typeid}). If you don't use those parts
1480 of the language, you can save some space by using this flag. Note that
1481 exception handling uses the same information, but it will generate it as
1486 Emit statistics about front-end processing at the end of the compilation.
1487 This information is generally only useful to the G++ development team.
1489 @item -ftemplate-depth-@var{n}
1490 @opindex ftemplate-depth
1491 Set the maximum instantiation depth for template classes to @var{n}.
1492 A limit on the template instantiation depth is needed to detect
1493 endless recursions during template class instantiation. ANSI/ISO C++
1494 conforming programs must not rely on a maximum depth greater than 17.
1496 @item -fno-threadsafe-statics
1497 @opindex fno-threadsafe-statics
1498 Do not emit the extra code to use the routines specified in the C++
1499 ABI for thread-safe initialization of local statics. You can use this
1500 option to reduce code size slightly in code that doesn't need to be
1503 @item -fuse-cxa-atexit
1504 @opindex fuse-cxa-atexit
1505 Register destructors for objects with static storage duration with the
1506 @code{__cxa_atexit} function rather than the @code{atexit} function.
1507 This option is required for fully standards-compliant handling of static
1508 destructors, but will only work if your C library supports
1509 @code{__cxa_atexit}.
1511 @item -fvisibility-inlines-hidden
1512 @opindex fvisibility-inlines-hidden
1513 Causes all inlined methods to be marked with
1514 @code{__attribute__ ((visibility ("hidden")))} so that they do not
1515 appear in the export table of a DSO and do not require a PLT indirection
1516 when used within the DSO@. Enabling this option can have a dramatic effect
1517 on load and link times of a DSO as it massively reduces the size of the
1518 dynamic export table when the library makes heavy use of templates. While
1519 it can cause bloating through duplication of code within each DSO where
1520 it is used, often the wastage is less than the considerable space occupied
1521 by a long symbol name in the export table which is typical when using
1522 templates and namespaces. For even more savings, combine with the
1523 @option{-fvisibility=hidden} switch.
1527 Do not use weak symbol support, even if it is provided by the linker.
1528 By default, G++ will use weak symbols if they are available. This
1529 option exists only for testing, and should not be used by end-users;
1530 it will result in inferior code and has no benefits. This option may
1531 be removed in a future release of G++.
1535 Do not search for header files in the standard directories specific to
1536 C++, but do still search the other standard directories. (This option
1537 is used when building the C++ library.)
1540 In addition, these optimization, warning, and code generation options
1541 have meanings only for C++ programs:
1544 @item -fno-default-inline
1545 @opindex fno-default-inline
1546 Do not assume @samp{inline} for functions defined inside a class scope.
1547 @xref{Optimize Options,,Options That Control Optimization}. Note that these
1548 functions will have linkage like inline functions; they just won't be
1551 @item -Wabi @r{(C++ only)}
1553 Warn when G++ generates code that is probably not compatible with the
1554 vendor-neutral C++ ABI@. Although an effort has been made to warn about
1555 all such cases, there are probably some cases that are not warned about,
1556 even though G++ is generating incompatible code. There may also be
1557 cases where warnings are emitted even though the code that is generated
1560 You should rewrite your code to avoid these warnings if you are
1561 concerned about the fact that code generated by G++ may not be binary
1562 compatible with code generated by other compilers.
1564 The known incompatibilities at this point include:
1569 Incorrect handling of tail-padding for bit-fields. G++ may attempt to
1570 pack data into the same byte as a base class. For example:
1573 struct A @{ virtual void f(); int f1 : 1; @};
1574 struct B : public A @{ int f2 : 1; @};
1578 In this case, G++ will place @code{B::f2} into the same byte
1579 as@code{A::f1}; other compilers will not. You can avoid this problem
1580 by explicitly padding @code{A} so that its size is a multiple of the
1581 byte size on your platform; that will cause G++ and other compilers to
1582 layout @code{B} identically.
1585 Incorrect handling of tail-padding for virtual bases. G++ does not use
1586 tail padding when laying out virtual bases. For example:
1589 struct A @{ virtual void f(); char c1; @};
1590 struct B @{ B(); char c2; @};
1591 struct C : public A, public virtual B @{@};
1595 In this case, G++ will not place @code{B} into the tail-padding for
1596 @code{A}; other compilers will. You can avoid this problem by
1597 explicitly padding @code{A} so that its size is a multiple of its
1598 alignment (ignoring virtual base classes); that will cause G++ and other
1599 compilers to layout @code{C} identically.
1602 Incorrect handling of bit-fields with declared widths greater than that
1603 of their underlying types, when the bit-fields appear in a union. For
1607 union U @{ int i : 4096; @};
1611 Assuming that an @code{int} does not have 4096 bits, G++ will make the
1612 union too small by the number of bits in an @code{int}.
1615 Empty classes can be placed at incorrect offsets. For example:
1625 struct C : public B, public A @{@};
1629 G++ will place the @code{A} base class of @code{C} at a nonzero offset;
1630 it should be placed at offset zero. G++ mistakenly believes that the
1631 @code{A} data member of @code{B} is already at offset zero.
1634 Names of template functions whose types involve @code{typename} or
1635 template template parameters can be mangled incorrectly.
1638 template <typename Q>
1639 void f(typename Q::X) @{@}
1641 template <template <typename> class Q>
1642 void f(typename Q<int>::X) @{@}
1646 Instantiations of these templates may be mangled incorrectly.
1650 @item -Wctor-dtor-privacy @r{(C++ only)}
1651 @opindex Wctor-dtor-privacy
1652 Warn when a class seems unusable because all the constructors or
1653 destructors in that class are private, and it has neither friends nor
1654 public static member functions.
1656 @item -Wnon-virtual-dtor @r{(C++ only)}
1657 @opindex Wnon-virtual-dtor
1658 Warn when a class appears to be polymorphic, thereby requiring a virtual
1659 destructor, yet it declares a non-virtual one.
1660 This warning is enabled by @option{-Wall}.
1662 @item -Wreorder @r{(C++ only)}
1664 @cindex reordering, warning
1665 @cindex warning for reordering of member initializers
1666 Warn when the order of member initializers given in the code does not
1667 match the order in which they must be executed. For instance:
1673 A(): j (0), i (1) @{ @}
1677 The compiler will rearrange the member initializers for @samp{i}
1678 and @samp{j} to match the declaration order of the members, emitting
1679 a warning to that effect. This warning is enabled by @option{-Wall}.
1682 The following @option{-W@dots{}} options are not affected by @option{-Wall}.
1685 @item -Weffc++ @r{(C++ only)}
1687 Warn about violations of the following style guidelines from Scott Meyers'
1688 @cite{Effective C++} book:
1692 Item 11: Define a copy constructor and an assignment operator for classes
1693 with dynamically allocated memory.
1696 Item 12: Prefer initialization to assignment in constructors.
1699 Item 14: Make destructors virtual in base classes.
1702 Item 15: Have @code{operator=} return a reference to @code{*this}.
1705 Item 23: Don't try to return a reference when you must return an object.
1709 Also warn about violations of the following style guidelines from
1710 Scott Meyers' @cite{More Effective C++} book:
1714 Item 6: Distinguish between prefix and postfix forms of increment and
1715 decrement operators.
1718 Item 7: Never overload @code{&&}, @code{||}, or @code{,}.
1722 When selecting this option, be aware that the standard library
1723 headers do not obey all of these guidelines; use @samp{grep -v}
1724 to filter out those warnings.
1726 @item -Wno-deprecated @r{(C++ only)}
1727 @opindex Wno-deprecated
1728 Do not warn about usage of deprecated features. @xref{Deprecated Features}.
1730 @item -Wno-non-template-friend @r{(C++ only)}
1731 @opindex Wno-non-template-friend
1732 Disable warnings when non-templatized friend functions are declared
1733 within a template. Since the advent of explicit template specification
1734 support in G++, if the name of the friend is an unqualified-id (i.e.,
1735 @samp{friend foo(int)}), the C++ language specification demands that the
1736 friend declare or define an ordinary, nontemplate function. (Section
1737 14.5.3). Before G++ implemented explicit specification, unqualified-ids
1738 could be interpreted as a particular specialization of a templatized
1739 function. Because this non-conforming behavior is no longer the default
1740 behavior for G++, @option{-Wnon-template-friend} allows the compiler to
1741 check existing code for potential trouble spots and is on by default.
1742 This new compiler behavior can be turned off with
1743 @option{-Wno-non-template-friend} which keeps the conformant compiler code
1744 but disables the helpful warning.
1746 @item -Wold-style-cast @r{(C++ only)}
1747 @opindex Wold-style-cast
1748 Warn if an old-style (C-style) cast to a non-void type is used within
1749 a C++ program. The new-style casts (@samp{static_cast},
1750 @samp{reinterpret_cast}, and @samp{const_cast}) are less vulnerable to
1751 unintended effects and much easier to search for.
1753 @item -Woverloaded-virtual @r{(C++ only)}
1754 @opindex Woverloaded-virtual
1755 @cindex overloaded virtual fn, warning
1756 @cindex warning for overloaded virtual fn
1757 Warn when a function declaration hides virtual functions from a
1758 base class. For example, in:
1765 struct B: public A @{
1770 the @code{A} class version of @code{f} is hidden in @code{B}, and code
1778 will fail to compile.
1780 @item -Wno-pmf-conversions @r{(C++ only)}
1781 @opindex Wno-pmf-conversions
1782 Disable the diagnostic for converting a bound pointer to member function
1785 @item -Wsign-promo @r{(C++ only)}
1786 @opindex Wsign-promo
1787 Warn when overload resolution chooses a promotion from unsigned or
1788 enumerated type to a signed type, over a conversion to an unsigned type of
1789 the same size. Previous versions of G++ would try to preserve
1790 unsignedness, but the standard mandates the current behavior.
1792 @item -Wsynth @r{(C++ only)}
1794 @cindex warning for synthesized methods
1795 @cindex synthesized methods, warning
1796 Warn when G++'s synthesis behavior does not match that of cfront. For
1802 A& operator = (int);
1812 In this example, G++ will synthesize a default @samp{A& operator =
1813 (const A&);}, while cfront will use the user-defined @samp{operator =}.
1816 @node Objective-C and Objective-C++ Dialect Options
1817 @section Options Controlling Objective-C and Objective-C++ Dialects
1819 @cindex compiler options, Objective-C and Objective-C++
1820 @cindex Objective-C and Objective-C++ options, command line
1821 @cindex options, Objective-C and Objective-C++
1822 (NOTE: This manual does not describe the Objective-C and Objective-C++
1823 languages themselves. See @xref{Standards,,Language Standards
1824 Supported by GCC}, for references.)
1826 This section describes the command-line options that are only meaningful
1827 for Objective-C and Objective-C++ programs, but you can also use most of
1828 the language-independent GNU compiler options.
1829 For example, you might compile a file @code{some_class.m} like this:
1832 gcc -g -fgnu-runtime -O -c some_class.m
1836 In this example, @option{-fgnu-runtime} is an option meant only for
1837 Objective-C and Objective-C++ programs; you can use the other options with
1838 any language supported by GCC@.
1840 Note that since Objective-C is an extension of the C language, Objective-C
1841 compilations may also use options specific to the C front-end (e.g.,
1842 @option{-Wtraditional}). Similarly, Objective-C++ compilations may use
1843 C++-specific options (e.g., @option{-Wabi}).
1845 Here is a list of options that are @emph{only} for compiling Objective-C
1846 and Objective-C++ programs:
1849 @item -fconstant-string-class=@var{class-name}
1850 @opindex fconstant-string-class
1851 Use @var{class-name} as the name of the class to instantiate for each
1852 literal string specified with the syntax @code{@@"@dots{}"}. The default
1853 class name is @code{NXConstantString} if the GNU runtime is being used, and
1854 @code{NSConstantString} if the NeXT runtime is being used (see below). The
1855 @option{-fconstant-cfstrings} option, if also present, will override the
1856 @option{-fconstant-string-class} setting and cause @code{@@"@dots{}"} literals
1857 to be laid out as constant CoreFoundation strings.
1860 @opindex fgnu-runtime
1861 Generate object code compatible with the standard GNU Objective-C
1862 runtime. This is the default for most types of systems.
1864 @item -fnext-runtime
1865 @opindex fnext-runtime
1866 Generate output compatible with the NeXT runtime. This is the default
1867 for NeXT-based systems, including Darwin and Mac OS X@. The macro
1868 @code{__NEXT_RUNTIME__} is predefined if (and only if) this option is
1871 @item -fno-nil-receivers
1872 @opindex fno-nil-receivers
1873 Assume that all Objective-C message dispatches (e.g.,
1874 @code{[receiver message:arg]}) in this translation unit ensure that the receiver
1875 is not @code{nil}. This allows for more efficient entry points in the runtime
1876 to be used. Currently, this option is only available in conjunction with
1877 the NeXT runtime on Mac OS X 10.3 and later.
1879 @item -fobjc-exceptions
1880 @opindex fobjc-exceptions
1881 Enable syntactic support for structured exception handling in Objective-C,
1882 similar to what is offered by C++ and Java. Currently, this option is only
1883 available in conjunction with the NeXT runtime on Mac OS X 10.3 and later.
1891 @@catch (AnObjCClass *exc) @{
1898 @@catch (AnotherClass *exc) @{
1901 @@catch (id allOthers) @{
1911 The @code{@@throw} statement may appear anywhere in an Objective-C or
1912 Objective-C++ program; when used inside of a @code{@@catch} block, the
1913 @code{@@throw} may appear without an argument (as shown above), in which case
1914 the object caught by the @code{@@catch} will be rethrown.
1916 Note that only (pointers to) Objective-C objects may be thrown and
1917 caught using this scheme. When an object is thrown, it will be caught
1918 by the nearest @code{@@catch} clause capable of handling objects of that type,
1919 analogously to how @code{catch} blocks work in C++ and Java. A
1920 @code{@@catch(id @dots{})} clause (as shown above) may also be provided to catch
1921 any and all Objective-C exceptions not caught by previous @code{@@catch}
1924 The @code{@@finally} clause, if present, will be executed upon exit from the
1925 immediately preceding @code{@@try @dots{} @@catch} section. This will happen
1926 regardless of whether any exceptions are thrown, caught or rethrown
1927 inside the @code{@@try @dots{} @@catch} section, analogously to the behavior
1928 of the @code{finally} clause in Java.
1930 There are several caveats to using the new exception mechanism:
1934 Although currently designed to be binary compatible with @code{NS_HANDLER}-style
1935 idioms provided by the @code{NSException} class, the new
1936 exceptions can only be used on Mac OS X 10.3 (Panther) and later
1937 systems, due to additional functionality needed in the (NeXT) Objective-C
1941 As mentioned above, the new exceptions do not support handling
1942 types other than Objective-C objects. Furthermore, when used from
1943 Objective-C++, the Objective-C exception model does not interoperate with C++
1944 exceptions at this time. This means you cannot @code{@@throw} an exception
1945 from Objective-C and @code{catch} it in C++, or vice versa
1946 (i.e., @code{throw @dots{} @@catch}).
1949 The @option{-fobjc-exceptions} switch also enables the use of synchronization
1950 blocks for thread-safe execution:
1953 @@synchronized (ObjCClass *guard) @{
1958 Upon entering the @code{@@synchronized} block, a thread of execution shall
1959 first check whether a lock has been placed on the corresponding @code{guard}
1960 object by another thread. If it has, the current thread shall wait until
1961 the other thread relinquishes its lock. Once @code{guard} becomes available,
1962 the current thread will place its own lock on it, execute the code contained in
1963 the @code{@@synchronized} block, and finally relinquish the lock (thereby
1964 making @code{guard} available to other threads).
1966 Unlike Java, Objective-C does not allow for entire methods to be marked
1967 @code{@@synchronized}. Note that throwing exceptions out of
1968 @code{@@synchronized} blocks is allowed, and will cause the guarding object
1969 to be unlocked properly.
1971 @item -freplace-objc-classes
1972 @opindex freplace-objc-classes
1973 Emit a special marker instructing @command{ld(1)} not to statically link in
1974 the resulting object file, and allow @command{dyld(1)} to load it in at
1975 run time instead. This is used in conjunction with the Fix-and-Continue
1976 debugging mode, where the object file in question may be recompiled and
1977 dynamically reloaded in the course of program execution, without the need
1978 to restart the program itself. Currently, Fix-and-Continue functionality
1979 is only available in conjunction with the NeXT runtime on Mac OS X 10.3
1984 When compiling for the NeXT runtime, the compiler ordinarily replaces calls
1985 to @code{objc_getClass("@dots{}")} (when the name of the class is known at
1986 compile time) with static class references that get initialized at load time,
1987 which improves run-time performance. Specifying the @option{-fzero-link} flag
1988 suppresses this behavior and causes calls to @code{objc_getClass("@dots{}")}
1989 to be retained. This is useful in Zero-Link debugging mode, since it allows
1990 for individual class implementations to be modified during program execution.
1994 Dump interface declarations for all classes seen in the source file to a
1995 file named @file{@var{sourcename}.decl}.
1998 @opindex Wno-protocol
1999 If a class is declared to implement a protocol, a warning is issued for
2000 every method in the protocol that is not implemented by the class. The
2001 default behavior is to issue a warning for every method not explicitly
2002 implemented in the class, even if a method implementation is inherited
2003 from the superclass. If you use the @option{-Wno-protocol} option, then
2004 methods inherited from the superclass are considered to be implemented,
2005 and no warning is issued for them.
2009 Warn if multiple methods of different types for the same selector are
2010 found during compilation. The check is performed on the list of methods
2011 in the final stage of compilation. Additionally, a check is performed
2012 for each selector appearing in a @code{@@selector(@dots{})}
2013 expression, and a corresponding method for that selector has been found
2014 during compilation. Because these checks scan the method table only at
2015 the end of compilation, these warnings are not produced if the final
2016 stage of compilation is not reached, for example because an error is
2017 found during compilation, or because the @option{-fsyntax-only} option is
2020 @item -Wundeclared-selector
2021 @opindex Wundeclared-selector
2022 Warn if a @code{@@selector(@dots{})} expression referring to an
2023 undeclared selector is found. A selector is considered undeclared if no
2024 method with that name has been declared before the
2025 @code{@@selector(@dots{})} expression, either explicitly in an
2026 @code{@@interface} or @code{@@protocol} declaration, or implicitly in
2027 an @code{@@implementation} section. This option always performs its
2028 checks as soon as a @code{@@selector(@dots{})} expression is found,
2029 while @option{-Wselector} only performs its checks in the final stage of
2030 compilation. This also enforces the coding style convention
2031 that methods and selectors must be declared before being used.
2033 @item -print-objc-runtime-info
2034 @opindex print-objc-runtime-info
2035 Generate C header describing the largest structure that is passed by
2040 @node Language Independent Options
2041 @section Options to Control Diagnostic Messages Formatting
2042 @cindex options to control diagnostics formatting
2043 @cindex diagnostic messages
2044 @cindex message formatting
2046 Traditionally, diagnostic messages have been formatted irrespective of
2047 the output device's aspect (e.g.@: its width, @dots{}). The options described
2048 below can be used to control the diagnostic messages formatting
2049 algorithm, e.g.@: how many characters per line, how often source location
2050 information should be reported. Right now, only the C++ front end can
2051 honor these options. However it is expected, in the near future, that
2052 the remaining front ends would be able to digest them correctly.
2055 @item -fmessage-length=@var{n}
2056 @opindex fmessage-length
2057 Try to format error messages so that they fit on lines of about @var{n}
2058 characters. The default is 72 characters for @command{g++} and 0 for the rest of
2059 the front ends supported by GCC@. If @var{n} is zero, then no
2060 line-wrapping will be done; each error message will appear on a single
2063 @opindex fdiagnostics-show-location
2064 @item -fdiagnostics-show-location=once
2065 Only meaningful in line-wrapping mode. Instructs the diagnostic messages
2066 reporter to emit @emph{once} source location information; that is, in
2067 case the message is too long to fit on a single physical line and has to
2068 be wrapped, the source location won't be emitted (as prefix) again,
2069 over and over, in subsequent continuation lines. This is the default
2072 @item -fdiagnostics-show-location=every-line
2073 Only meaningful in line-wrapping mode. Instructs the diagnostic
2074 messages reporter to emit the same source location information (as
2075 prefix) for physical lines that result from the process of breaking
2076 a message which is too long to fit on a single line.
2080 @node Warning Options
2081 @section Options to Request or Suppress Warnings
2082 @cindex options to control warnings
2083 @cindex warning messages
2084 @cindex messages, warning
2085 @cindex suppressing warnings
2087 Warnings are diagnostic messages that report constructions which
2088 are not inherently erroneous but which are risky or suggest there
2089 may have been an error.
2091 You can request many specific warnings with options beginning @samp{-W},
2092 for example @option{-Wimplicit} to request warnings on implicit
2093 declarations. Each of these specific warning options also has a
2094 negative form beginning @samp{-Wno-} to turn off warnings;
2095 for example, @option{-Wno-implicit}. This manual lists only one of the
2096 two forms, whichever is not the default.
2098 The following options control the amount and kinds of warnings produced
2099 by GCC; for further, language-specific options also refer to
2100 @ref{C++ Dialect Options} and @ref{Objective-C and Objective-C++ Dialect
2104 @cindex syntax checking
2106 @opindex fsyntax-only
2107 Check the code for syntax errors, but don't do anything beyond that.
2111 Issue all the warnings demanded by strict ISO C and ISO C++;
2112 reject all programs that use forbidden extensions, and some other
2113 programs that do not follow ISO C and ISO C++. For ISO C, follows the
2114 version of the ISO C standard specified by any @option{-std} option used.
2116 Valid ISO C and ISO C++ programs should compile properly with or without
2117 this option (though a rare few will require @option{-ansi} or a
2118 @option{-std} option specifying the required version of ISO C)@. However,
2119 without this option, certain GNU extensions and traditional C and C++
2120 features are supported as well. With this option, they are rejected.
2122 @option{-pedantic} does not cause warning messages for use of the
2123 alternate keywords whose names begin and end with @samp{__}. Pedantic
2124 warnings are also disabled in the expression that follows
2125 @code{__extension__}. However, only system header files should use
2126 these escape routes; application programs should avoid them.
2127 @xref{Alternate Keywords}.
2129 Some users try to use @option{-pedantic} to check programs for strict ISO
2130 C conformance. They soon find that it does not do quite what they want:
2131 it finds some non-ISO practices, but not all---only those for which
2132 ISO C @emph{requires} a diagnostic, and some others for which
2133 diagnostics have been added.
2135 A feature to report any failure to conform to ISO C might be useful in
2136 some instances, but would require considerable additional work and would
2137 be quite different from @option{-pedantic}. We don't have plans to
2138 support such a feature in the near future.
2140 Where the standard specified with @option{-std} represents a GNU
2141 extended dialect of C, such as @samp{gnu89} or @samp{gnu99}, there is a
2142 corresponding @dfn{base standard}, the version of ISO C on which the GNU
2143 extended dialect is based. Warnings from @option{-pedantic} are given
2144 where they are required by the base standard. (It would not make sense
2145 for such warnings to be given only for features not in the specified GNU
2146 C dialect, since by definition the GNU dialects of C include all
2147 features the compiler supports with the given option, and there would be
2148 nothing to warn about.)
2150 @item -pedantic-errors
2151 @opindex pedantic-errors
2152 Like @option{-pedantic}, except that errors are produced rather than
2157 Inhibit all warning messages.
2161 Inhibit warning messages about the use of @samp{#import}.
2163 @item -Wchar-subscripts
2164 @opindex Wchar-subscripts
2165 Warn if an array subscript has type @code{char}. This is a common cause
2166 of error, as programmers often forget that this type is signed on some
2168 This warning is enabled by @option{-Wall}.
2172 Warn whenever a comment-start sequence @samp{/*} appears in a @samp{/*}
2173 comment, or whenever a Backslash-Newline appears in a @samp{//} comment.
2174 This warning is enabled by @option{-Wall}.
2176 @item -Wfatal-errors
2177 @opindex Wfatal-errors
2178 This option causes the compiler to abort compilation on the first error
2179 occurred rather than trying to keep going and printing further error
2184 @opindex ffreestanding
2185 @opindex fno-builtin
2186 Check calls to @code{printf} and @code{scanf}, etc., to make sure that
2187 the arguments supplied have types appropriate to the format string
2188 specified, and that the conversions specified in the format string make
2189 sense. This includes standard functions, and others specified by format
2190 attributes (@pxref{Function Attributes}), in the @code{printf},
2191 @code{scanf}, @code{strftime} and @code{strfmon} (an X/Open extension,
2192 not in the C standard) families (or other target-specific families).
2193 Which functions are checked without format attributes having been
2194 specified depends on the standard version selected, and such checks of
2195 functions without the attribute specified are disabled by
2196 @option{-ffreestanding} or @option{-fno-builtin}.
2198 The formats are checked against the format features supported by GNU
2199 libc version 2.2. These include all ISO C90 and C99 features, as well
2200 as features from the Single Unix Specification and some BSD and GNU
2201 extensions. Other library implementations may not support all these
2202 features; GCC does not support warning about features that go beyond a
2203 particular library's limitations. However, if @option{-pedantic} is used
2204 with @option{-Wformat}, warnings will be given about format features not
2205 in the selected standard version (but not for @code{strfmon} formats,
2206 since those are not in any version of the C standard). @xref{C Dialect
2207 Options,,Options Controlling C Dialect}.
2209 Since @option{-Wformat} also checks for null format arguments for
2210 several functions, @option{-Wformat} also implies @option{-Wnonnull}.
2212 @option{-Wformat} is included in @option{-Wall}. For more control over some
2213 aspects of format checking, the options @option{-Wformat-y2k},
2214 @option{-Wno-format-extra-args}, @option{-Wno-format-zero-length},
2215 @option{-Wformat-nonliteral}, @option{-Wformat-security}, and
2216 @option{-Wformat=2} are available, but are not included in @option{-Wall}.
2219 @opindex Wformat-y2k
2220 If @option{-Wformat} is specified, also warn about @code{strftime}
2221 formats which may yield only a two-digit year.
2223 @item -Wno-format-extra-args
2224 @opindex Wno-format-extra-args
2225 If @option{-Wformat} is specified, do not warn about excess arguments to a
2226 @code{printf} or @code{scanf} format function. The C standard specifies
2227 that such arguments are ignored.
2229 Where the unused arguments lie between used arguments that are
2230 specified with @samp{$} operand number specifications, normally
2231 warnings are still given, since the implementation could not know what
2232 type to pass to @code{va_arg} to skip the unused arguments. However,
2233 in the case of @code{scanf} formats, this option will suppress the
2234 warning if the unused arguments are all pointers, since the Single
2235 Unix Specification says that such unused arguments are allowed.
2237 @item -Wno-format-zero-length
2238 @opindex Wno-format-zero-length
2239 If @option{-Wformat} is specified, do not warn about zero-length formats.
2240 The C standard specifies that zero-length formats are allowed.
2242 @item -Wformat-nonliteral
2243 @opindex Wformat-nonliteral
2244 If @option{-Wformat} is specified, also warn if the format string is not a
2245 string literal and so cannot be checked, unless the format function
2246 takes its format arguments as a @code{va_list}.
2248 @item -Wformat-security
2249 @opindex Wformat-security
2250 If @option{-Wformat} is specified, also warn about uses of format
2251 functions that represent possible security problems. At present, this
2252 warns about calls to @code{printf} and @code{scanf} functions where the
2253 format string is not a string literal and there are no format arguments,
2254 as in @code{printf (foo);}. This may be a security hole if the format
2255 string came from untrusted input and contains @samp{%n}. (This is
2256 currently a subset of what @option{-Wformat-nonliteral} warns about, but
2257 in future warnings may be added to @option{-Wformat-security} that are not
2258 included in @option{-Wformat-nonliteral}.)
2262 Enable @option{-Wformat} plus format checks not included in
2263 @option{-Wformat}. Currently equivalent to @samp{-Wformat
2264 -Wformat-nonliteral -Wformat-security -Wformat-y2k}.
2268 Warn about passing a null pointer for arguments marked as
2269 requiring a non-null value by the @code{nonnull} function attribute.
2271 @option{-Wnonnull} is included in @option{-Wall} and @option{-Wformat}. It
2272 can be disabled with the @option{-Wno-nonnull} option.
2274 @item -Winit-self @r{(C, C++, Objective-C and Objective-C++ only)}
2276 Warn about uninitialized variables which are initialized with themselves.
2277 Note this option can only be used with the @option{-Wuninitialized} option,
2278 which in turn only works with @option{-O1} and above.
2280 For example, GCC will warn about @code{i} being uninitialized in the
2281 following snippet only when @option{-Winit-self} has been specified:
2292 @item -Wimplicit-int
2293 @opindex Wimplicit-int
2294 Warn when a declaration does not specify a type.
2295 This warning is enabled by @option{-Wall}.
2297 @item -Wimplicit-function-declaration
2298 @itemx -Werror-implicit-function-declaration
2299 @opindex Wimplicit-function-declaration
2300 @opindex Werror-implicit-function-declaration
2301 Give a warning (or error) whenever a function is used before being
2302 declared. The form @option{-Wno-error-implicit-function-declaration}
2304 This warning is enabled by @option{-Wall} (as a warning, not an error).
2308 Same as @option{-Wimplicit-int} and @option{-Wimplicit-function-declaration}.
2309 This warning is enabled by @option{-Wall}.
2313 Warn if the type of @samp{main} is suspicious. @samp{main} should be a
2314 function with external linkage, returning int, taking either zero
2315 arguments, two, or three arguments of appropriate types.
2316 This warning is enabled by @option{-Wall}.
2318 @item -Wmissing-braces
2319 @opindex Wmissing-braces
2320 Warn if an aggregate or union initializer is not fully bracketed. In
2321 the following example, the initializer for @samp{a} is not fully
2322 bracketed, but that for @samp{b} is fully bracketed.
2325 int a[2][2] = @{ 0, 1, 2, 3 @};
2326 int b[2][2] = @{ @{ 0, 1 @}, @{ 2, 3 @} @};
2329 This warning is enabled by @option{-Wall}.
2331 @item -Wmissing-include-dirs @r{(C, C++, Objective-C and Objective-C++ only)}
2332 @opindex Wmissing-include-dirs
2333 Warn if a user-supplied include directory does not exist.
2336 @opindex Wparentheses
2337 Warn if parentheses are omitted in certain contexts, such
2338 as when there is an assignment in a context where a truth value
2339 is expected, or when operators are nested whose precedence people
2340 often get confused about. Only the warning for an assignment used as
2341 a truth value is supported when compiling C++; the other warnings are
2342 only supported when compiling C@.
2344 Also warn if a comparison like @samp{x<=y<=z} appears; this is
2345 equivalent to @samp{(x<=y ? 1 : 0) <= z}, which is a different
2346 interpretation from that of ordinary mathematical notation.
2348 Also warn about constructions where there may be confusion to which
2349 @code{if} statement an @code{else} branch belongs. Here is an example of
2364 In C, every @code{else} branch belongs to the innermost possible @code{if}
2365 statement, which in this example is @code{if (b)}. This is often not
2366 what the programmer expected, as illustrated in the above example by
2367 indentation the programmer chose. When there is the potential for this
2368 confusion, GCC will issue a warning when this flag is specified.
2369 To eliminate the warning, add explicit braces around the innermost
2370 @code{if} statement so there is no way the @code{else} could belong to
2371 the enclosing @code{if}. The resulting code would look like this:
2387 This warning is enabled by @option{-Wall}.
2389 @item -Wsequence-point
2390 @opindex Wsequence-point
2391 Warn about code that may have undefined semantics because of violations
2392 of sequence point rules in the C standard.
2394 The C standard defines the order in which expressions in a C program are
2395 evaluated in terms of @dfn{sequence points}, which represent a partial
2396 ordering between the execution of parts of the program: those executed
2397 before the sequence point, and those executed after it. These occur
2398 after the evaluation of a full expression (one which is not part of a
2399 larger expression), after the evaluation of the first operand of a
2400 @code{&&}, @code{||}, @code{? :} or @code{,} (comma) operator, before a
2401 function is called (but after the evaluation of its arguments and the
2402 expression denoting the called function), and in certain other places.
2403 Other than as expressed by the sequence point rules, the order of
2404 evaluation of subexpressions of an expression is not specified. All
2405 these rules describe only a partial order rather than a total order,
2406 since, for example, if two functions are called within one expression
2407 with no sequence point between them, the order in which the functions
2408 are called is not specified. However, the standards committee have
2409 ruled that function calls do not overlap.
2411 It is not specified when between sequence points modifications to the
2412 values of objects take effect. Programs whose behavior depends on this
2413 have undefined behavior; the C standard specifies that ``Between the
2414 previous and next sequence point an object shall have its stored value
2415 modified at most once by the evaluation of an expression. Furthermore,
2416 the prior value shall be read only to determine the value to be
2417 stored.''. If a program breaks these rules, the results on any
2418 particular implementation are entirely unpredictable.
2420 Examples of code with undefined behavior are @code{a = a++;}, @code{a[n]
2421 = b[n++]} and @code{a[i++] = i;}. Some more complicated cases are not
2422 diagnosed by this option, and it may give an occasional false positive
2423 result, but in general it has been found fairly effective at detecting
2424 this sort of problem in programs.
2426 The present implementation of this option only works for C programs. A
2427 future implementation may also work for C++ programs.
2429 The C standard is worded confusingly, therefore there is some debate
2430 over the precise meaning of the sequence point rules in subtle cases.
2431 Links to discussions of the problem, including proposed formal
2432 definitions, may be found on the GCC readings page, at
2433 @w{@uref{http://gcc.gnu.org/readings.html}}.
2435 This warning is enabled by @option{-Wall}.
2438 @opindex Wreturn-type
2439 Warn whenever a function is defined with a return-type that defaults to
2440 @code{int}. Also warn about any @code{return} statement with no
2441 return-value in a function whose return-type is not @code{void}.
2443 For C, also warn if the return type of a function has a type qualifier
2444 such as @code{const}. Such a type qualifier has no effect, since the
2445 value returned by a function is not an lvalue. ISO C prohibits
2446 qualified @code{void} return types on function definitions, so such
2447 return types always receive a warning even without this option.
2449 For C++, a function without return type always produces a diagnostic
2450 message, even when @option{-Wno-return-type} is specified. The only
2451 exceptions are @samp{main} and functions defined in system headers.
2453 This warning is enabled by @option{-Wall}.
2457 Warn whenever a @code{switch} statement has an index of enumerated type
2458 and lacks a @code{case} for one or more of the named codes of that
2459 enumeration. (The presence of a @code{default} label prevents this
2460 warning.) @code{case} labels outside the enumeration range also
2461 provoke warnings when this option is used.
2462 This warning is enabled by @option{-Wall}.
2464 @item -Wswitch-default
2465 @opindex Wswitch-switch
2466 Warn whenever a @code{switch} statement does not have a @code{default}
2470 @opindex Wswitch-enum
2471 Warn whenever a @code{switch} statement has an index of enumerated type
2472 and lacks a @code{case} for one or more of the named codes of that
2473 enumeration. @code{case} labels outside the enumeration range also
2474 provoke warnings when this option is used.
2478 Warn if any trigraphs are encountered that might change the meaning of
2479 the program (trigraphs within comments are not warned about).
2480 This warning is enabled by @option{-Wall}.
2482 @item -Wunused-function
2483 @opindex Wunused-function
2484 Warn whenever a static function is declared but not defined or a
2485 non\-inline static function is unused.
2486 This warning is enabled by @option{-Wall}.
2488 @item -Wunused-label
2489 @opindex Wunused-label
2490 Warn whenever a label is declared but not used.
2491 This warning is enabled by @option{-Wall}.
2493 To suppress this warning use the @samp{unused} attribute
2494 (@pxref{Variable Attributes}).
2496 @item -Wunused-parameter
2497 @opindex Wunused-parameter
2498 Warn whenever a function parameter is unused aside from its declaration.
2500 To suppress this warning use the @samp{unused} attribute
2501 (@pxref{Variable Attributes}).
2503 @item -Wunused-variable
2504 @opindex Wunused-variable
2505 Warn whenever a local variable or non-constant static variable is unused
2506 aside from its declaration
2507 This warning is enabled by @option{-Wall}.
2509 To suppress this warning use the @samp{unused} attribute
2510 (@pxref{Variable Attributes}).
2512 @item -Wunused-value
2513 @opindex Wunused-value
2514 Warn whenever a statement computes a result that is explicitly not used.
2515 This warning is enabled by @option{-Wall}.
2517 To suppress this warning cast the expression to @samp{void}.
2521 All the above @option{-Wunused} options combined.
2523 In order to get a warning about an unused function parameter, you must
2524 either specify @samp{-Wextra -Wunused} (note that @samp{-Wall} implies
2525 @samp{-Wunused}), or separately specify @option{-Wunused-parameter}.
2527 @item -Wuninitialized
2528 @opindex Wuninitialized
2529 Warn if an automatic variable is used without first being initialized or
2530 if a variable may be clobbered by a @code{setjmp} call.
2532 These warnings are possible only in optimizing compilation,
2533 because they require data flow information that is computed only
2534 when optimizing. If you don't specify @option{-O}, you simply won't
2537 If you want to warn about code which uses the uninitialized value of the
2538 variable in its own initializer, use the @option{-Winit-self} option.
2540 These warnings occur only for variables that are candidates for
2541 register allocation. Therefore, they do not occur for a variable that
2542 is declared @code{volatile}, or whose address is taken, or whose size
2543 is other than 1, 2, 4 or 8 bytes. Also, they do not occur for
2544 structures, unions or arrays, even when they are in registers.
2546 Note that there may be no warning about a variable that is used only
2547 to compute a value that itself is never used, because such
2548 computations may be deleted by data flow analysis before the warnings
2551 These warnings are made optional because GCC is not smart
2552 enough to see all the reasons why the code might be correct
2553 despite appearing to have an error. Here is one example of how
2574 If the value of @code{y} is always 1, 2 or 3, then @code{x} is
2575 always initialized, but GCC doesn't know this. Here is
2576 another common case:
2581 if (change_y) save_y = y, y = new_y;
2583 if (change_y) y = save_y;
2588 This has no bug because @code{save_y} is used only if it is set.
2590 @cindex @code{longjmp} warnings
2591 This option also warns when a non-volatile automatic variable might be
2592 changed by a call to @code{longjmp}. These warnings as well are possible
2593 only in optimizing compilation.
2595 The compiler sees only the calls to @code{setjmp}. It cannot know
2596 where @code{longjmp} will be called; in fact, a signal handler could
2597 call it at any point in the code. As a result, you may get a warning
2598 even when there is in fact no problem because @code{longjmp} cannot
2599 in fact be called at the place which would cause a problem.
2601 Some spurious warnings can be avoided if you declare all the functions
2602 you use that never return as @code{noreturn}. @xref{Function
2605 This warning is enabled by @option{-Wall}.
2607 @item -Wunknown-pragmas
2608 @opindex Wunknown-pragmas
2609 @cindex warning for unknown pragmas
2610 @cindex unknown pragmas, warning
2611 @cindex pragmas, warning of unknown
2612 Warn when a #pragma directive is encountered which is not understood by
2613 GCC@. If this command line option is used, warnings will even be issued
2614 for unknown pragmas in system header files. This is not the case if
2615 the warnings were only enabled by the @option{-Wall} command line option.
2617 @item -Wstrict-aliasing
2618 @opindex Wstrict-aliasing
2619 This option is only active when @option{-fstrict-aliasing} is active.
2620 It warns about code which might break the strict aliasing rules that the
2621 compiler is using for optimization. The warning does not catch all
2622 cases, but does attempt to catch the more common pitfalls. It is
2623 included in @option{-Wall}.
2625 @item -Wstrict-aliasing=2
2626 @opindex Wstrict-aliasing=2
2627 This option is only active when @option{-fstrict-aliasing} is active.
2628 It warns about all code which might break the strict aliasing rules that the
2629 compiler is using for optimization. This warning catches all cases, but
2630 it will also give a warning for some ambiguous cases that are safe.
2634 All of the above @samp{-W} options combined. This enables all the
2635 warnings about constructions that some users consider questionable, and
2636 that are easy to avoid (or modify to prevent the warning), even in
2637 conjunction with macros. This also enables some language-specific
2638 warnings described in @ref{C++ Dialect Options} and
2639 @ref{Objective-C and Objective-C++ Dialect Options}.
2642 The following @option{-W@dots{}} options are not implied by @option{-Wall}.
2643 Some of them warn about constructions that users generally do not
2644 consider questionable, but which occasionally you might wish to check
2645 for; others warn about constructions that are necessary or hard to avoid
2646 in some cases, and there is no simple way to modify the code to suppress
2653 (This option used to be called @option{-W}. The older name is still
2654 supported, but the newer name is more descriptive.) Print extra warning
2655 messages for these events:
2659 A function can return either with or without a value. (Falling
2660 off the end of the function body is considered returning without
2661 a value.) For example, this function would evoke such a
2675 An expression-statement or the left-hand side of a comma expression
2676 contains no side effects.
2677 To suppress the warning, cast the unused expression to void.
2678 For example, an expression such as @samp{x[i,j]} will cause a warning,
2679 but @samp{x[(void)i,j]} will not.
2682 An unsigned value is compared against zero with @samp{<} or @samp{>=}.
2685 Storage-class specifiers like @code{static} are not the first things in
2686 a declaration. According to the C Standard, this usage is obsolescent.
2689 If @option{-Wall} or @option{-Wunused} is also specified, warn about unused
2693 A comparison between signed and unsigned values could produce an
2694 incorrect result when the signed value is converted to unsigned.
2695 (But don't warn if @option{-Wno-sign-compare} is also specified.)
2698 An aggregate has an initializer which does not initialize all members.
2699 This warning can be independently controlled by
2700 @option{-Wmissing-field-initializers}.
2703 A function parameter is declared without a type specifier in K&R-style
2711 An empty body occurs in an @samp{if} or @samp{else} statement.
2714 A pointer is compared against integer zero with @samp{<}, @samp{<=},
2715 @samp{>}, or @samp{>=}.
2718 A variable might be changed by @samp{longjmp} or @samp{vfork}.
2721 Any of several floating-point events that often indicate errors, such as
2722 overflow, underflow, loss of precision, etc.
2724 @item @r{(C++ only)}
2725 An enumerator and a non-enumerator both appear in a conditional expression.
2727 @item @r{(C++ only)}
2728 A non-static reference or non-static @samp{const} member appears in a
2729 class without constructors.
2731 @item @r{(C++ only)}
2732 Ambiguous virtual bases.
2734 @item @r{(C++ only)}
2735 Subscripting an array which has been declared @samp{register}.
2737 @item @r{(C++ only)}
2738 Taking the address of a variable which has been declared @samp{register}.
2740 @item @r{(C++ only)}
2741 A base class is not initialized in a derived class' copy constructor.
2744 @item -Wno-div-by-zero
2745 @opindex Wno-div-by-zero
2746 @opindex Wdiv-by-zero
2747 Do not warn about compile-time integer division by zero. Floating point
2748 division by zero is not warned about, as it can be a legitimate way of
2749 obtaining infinities and NaNs.
2751 @item -Wsystem-headers
2752 @opindex Wsystem-headers
2753 @cindex warnings from system headers
2754 @cindex system headers, warnings from
2755 Print warning messages for constructs found in system header files.
2756 Warnings from system headers are normally suppressed, on the assumption
2757 that they usually do not indicate real problems and would only make the
2758 compiler output harder to read. Using this command line option tells
2759 GCC to emit warnings from system headers as if they occurred in user
2760 code. However, note that using @option{-Wall} in conjunction with this
2761 option will @emph{not} warn about unknown pragmas in system
2762 headers---for that, @option{-Wunknown-pragmas} must also be used.
2765 @opindex Wfloat-equal
2766 Warn if floating point values are used in equality comparisons.
2768 The idea behind this is that sometimes it is convenient (for the
2769 programmer) to consider floating-point values as approximations to
2770 infinitely precise real numbers. If you are doing this, then you need
2771 to compute (by analyzing the code, or in some other way) the maximum or
2772 likely maximum error that the computation introduces, and allow for it
2773 when performing comparisons (and when producing output, but that's a
2774 different problem). In particular, instead of testing for equality, you
2775 would check to see whether the two values have ranges that overlap; and
2776 this is done with the relational operators, so equality comparisons are
2779 @item -Wtraditional @r{(C only)}
2780 @opindex Wtraditional
2781 Warn about certain constructs that behave differently in traditional and
2782 ISO C@. Also warn about ISO C constructs that have no traditional C
2783 equivalent, and/or problematic constructs which should be avoided.
2787 Macro parameters that appear within string literals in the macro body.
2788 In traditional C macro replacement takes place within string literals,
2789 but does not in ISO C@.
2792 In traditional C, some preprocessor directives did not exist.
2793 Traditional preprocessors would only consider a line to be a directive
2794 if the @samp{#} appeared in column 1 on the line. Therefore
2795 @option{-Wtraditional} warns about directives that traditional C
2796 understands but would ignore because the @samp{#} does not appear as the
2797 first character on the line. It also suggests you hide directives like
2798 @samp{#pragma} not understood by traditional C by indenting them. Some
2799 traditional implementations would not recognize @samp{#elif}, so it
2800 suggests avoiding it altogether.
2803 A function-like macro that appears without arguments.
2806 The unary plus operator.
2809 The @samp{U} integer constant suffix, or the @samp{F} or @samp{L} floating point
2810 constant suffixes. (Traditional C does support the @samp{L} suffix on integer
2811 constants.) Note, these suffixes appear in macros defined in the system
2812 headers of most modern systems, e.g.@: the @samp{_MIN}/@samp{_MAX} macros in @code{<limits.h>}.
2813 Use of these macros in user code might normally lead to spurious
2814 warnings, however GCC's integrated preprocessor has enough context to
2815 avoid warning in these cases.
2818 A function declared external in one block and then used after the end of
2822 A @code{switch} statement has an operand of type @code{long}.
2825 A non-@code{static} function declaration follows a @code{static} one.
2826 This construct is not accepted by some traditional C compilers.
2829 The ISO type of an integer constant has a different width or
2830 signedness from its traditional type. This warning is only issued if
2831 the base of the constant is ten. I.e.@: hexadecimal or octal values, which
2832 typically represent bit patterns, are not warned about.
2835 Usage of ISO string concatenation is detected.
2838 Initialization of automatic aggregates.
2841 Identifier conflicts with labels. Traditional C lacks a separate
2842 namespace for labels.
2845 Initialization of unions. If the initializer is zero, the warning is
2846 omitted. This is done under the assumption that the zero initializer in
2847 user code appears conditioned on e.g.@: @code{__STDC__} to avoid missing
2848 initializer warnings and relies on default initialization to zero in the
2852 Conversions by prototypes between fixed/floating point values and vice
2853 versa. The absence of these prototypes when compiling with traditional
2854 C would cause serious problems. This is a subset of the possible
2855 conversion warnings, for the full set use @option{-Wconversion}.
2858 Use of ISO C style function definitions. This warning intentionally is
2859 @emph{not} issued for prototype declarations or variadic functions
2860 because these ISO C features will appear in your code when using
2861 libiberty's traditional C compatibility macros, @code{PARAMS} and
2862 @code{VPARAMS}. This warning is also bypassed for nested functions
2863 because that feature is already a GCC extension and thus not relevant to
2864 traditional C compatibility.
2867 @item -Wdeclaration-after-statement @r{(C only)}
2868 @opindex Wdeclaration-after-statement
2869 Warn when a declaration is found after a statement in a block. This
2870 construct, known from C++, was introduced with ISO C99 and is by default
2871 allowed in GCC@. It is not supported by ISO C90 and was not supported by
2872 GCC versions before GCC 3.0. @xref{Mixed Declarations}.
2876 Warn if an undefined identifier is evaluated in an @samp{#if} directive.
2878 @item -Wendif-labels
2879 @opindex Wendif-labels
2880 Warn whenever an @samp{#else} or an @samp{#endif} are followed by text.
2884 Warn whenever a local variable shadows another local variable, parameter or
2885 global variable or whenever a built-in function is shadowed.
2887 @item -Wlarger-than-@var{len}
2888 @opindex Wlarger-than
2889 Warn whenever an object of larger than @var{len} bytes is defined.
2891 @item -Wpointer-arith
2892 @opindex Wpointer-arith
2893 Warn about anything that depends on the ``size of'' a function type or
2894 of @code{void}. GNU C assigns these types a size of 1, for
2895 convenience in calculations with @code{void *} pointers and pointers
2898 @item -Wbad-function-cast @r{(C only)}
2899 @opindex Wbad-function-cast
2900 Warn whenever a function call is cast to a non-matching type.
2901 For example, warn if @code{int malloc()} is cast to @code{anything *}.
2905 Warn whenever a pointer is cast so as to remove a type qualifier from
2906 the target type. For example, warn if a @code{const char *} is cast
2907 to an ordinary @code{char *}.
2910 @opindex Wcast-align
2911 Warn whenever a pointer is cast such that the required alignment of the
2912 target is increased. For example, warn if a @code{char *} is cast to
2913 an @code{int *} on machines where integers can only be accessed at
2914 two- or four-byte boundaries.
2916 @item -Wwrite-strings
2917 @opindex Wwrite-strings
2918 When compiling C, give string constants the type @code{const
2919 char[@var{length}]} so that
2920 copying the address of one into a non-@code{const} @code{char *}
2921 pointer will get a warning; when compiling C++, warn about the
2922 deprecated conversion from string constants to @code{char *}.
2923 These warnings will help you find at
2924 compile time code that can try to write into a string constant, but
2925 only if you have been very careful about using @code{const} in
2926 declarations and prototypes. Otherwise, it will just be a nuisance;
2927 this is why we did not make @option{-Wall} request these warnings.
2930 @opindex Wconversion
2931 Warn if a prototype causes a type conversion that is different from what
2932 would happen to the same argument in the absence of a prototype. This
2933 includes conversions of fixed point to floating and vice versa, and
2934 conversions changing the width or signedness of a fixed point argument
2935 except when the same as the default promotion.
2937 Also, warn if a negative integer constant expression is implicitly
2938 converted to an unsigned type. For example, warn about the assignment
2939 @code{x = -1} if @code{x} is unsigned. But do not warn about explicit
2940 casts like @code{(unsigned) -1}.
2942 @item -Wsign-compare
2943 @opindex Wsign-compare
2944 @cindex warning for comparison of signed and unsigned values
2945 @cindex comparison of signed and unsigned values, warning
2946 @cindex signed and unsigned values, comparison warning
2947 Warn when a comparison between signed and unsigned values could produce
2948 an incorrect result when the signed value is converted to unsigned.
2949 This warning is also enabled by @option{-Wextra}; to get the other warnings
2950 of @option{-Wextra} without this warning, use @samp{-Wextra -Wno-sign-compare}.
2952 @item -Waggregate-return
2953 @opindex Waggregate-return
2954 Warn if any functions that return structures or unions are defined or
2955 called. (In languages where you can return an array, this also elicits
2958 @item -Wstrict-prototypes @r{(C only)}
2959 @opindex Wstrict-prototypes
2960 Warn if a function is declared or defined without specifying the
2961 argument types. (An old-style function definition is permitted without
2962 a warning if preceded by a declaration which specifies the argument
2965 @item -Wold-style-definition @r{(C only)}
2966 @opindex Wold-style-definition
2967 Warn if an old-style function definition is used. A warning is given
2968 even if there is a previous prototype.
2970 @item -Wmissing-prototypes @r{(C only)}
2971 @opindex Wmissing-prototypes
2972 Warn if a global function is defined without a previous prototype
2973 declaration. This warning is issued even if the definition itself
2974 provides a prototype. The aim is to detect global functions that fail
2975 to be declared in header files.
2977 @item -Wmissing-declarations @r{(C only)}
2978 @opindex Wmissing-declarations
2979 Warn if a global function is defined without a previous declaration.
2980 Do so even if the definition itself provides a prototype.
2981 Use this option to detect global functions that are not declared in
2984 @item -Wmissing-field-initializers
2985 @opindex Wmissing-field-initializers
2988 Warn if a structure's initializer has some fields missing. For
2989 example, the following code would cause such a warning, because
2990 @code{x.h} is implicitly zero:
2993 struct s @{ int f, g, h; @};
2994 struct s x = @{ 3, 4 @};
2997 This option does not warn about designated initializers, so the following
2998 modification would not trigger a warning:
3001 struct s @{ int f, g, h; @};
3002 struct s x = @{ .f = 3, .g = 4 @};
3005 This warning is included in @option{-Wextra}. To get other @option{-Wextra}
3006 warnings without this one, use @samp{-Wextra -Wno-missing-field-initializers}.
3008 @item -Wmissing-noreturn
3009 @opindex Wmissing-noreturn
3010 Warn about functions which might be candidates for attribute @code{noreturn}.
3011 Note these are only possible candidates, not absolute ones. Care should
3012 be taken to manually verify functions actually do not ever return before
3013 adding the @code{noreturn} attribute, otherwise subtle code generation
3014 bugs could be introduced. You will not get a warning for @code{main} in
3015 hosted C environments.
3017 @item -Wmissing-format-attribute
3018 @opindex Wmissing-format-attribute
3020 If @option{-Wformat} is enabled, also warn about functions which might be
3021 candidates for @code{format} attributes. Note these are only possible
3022 candidates, not absolute ones. GCC will guess that @code{format}
3023 attributes might be appropriate for any function that calls a function
3024 like @code{vprintf} or @code{vscanf}, but this might not always be the
3025 case, and some functions for which @code{format} attributes are
3026 appropriate may not be detected. This option has no effect unless
3027 @option{-Wformat} is enabled (possibly by @option{-Wall}).
3029 @item -Wno-multichar
3030 @opindex Wno-multichar
3032 Do not warn if a multicharacter constant (@samp{'FOOF'}) is used.
3033 Usually they indicate a typo in the user's code, as they have
3034 implementation-defined values, and should not be used in portable code.
3036 @item -Wno-deprecated-declarations
3037 @opindex Wno-deprecated-declarations
3038 Do not warn about uses of functions, variables, and types marked as
3039 deprecated by using the @code{deprecated} attribute.
3040 (@pxref{Function Attributes}, @pxref{Variable Attributes},
3041 @pxref{Type Attributes}.)
3045 Warn if a structure is given the packed attribute, but the packed
3046 attribute has no effect on the layout or size of the structure.
3047 Such structures may be mis-aligned for little benefit. For
3048 instance, in this code, the variable @code{f.x} in @code{struct bar}
3049 will be misaligned even though @code{struct bar} does not itself
3050 have the packed attribute:
3057 @} __attribute__((packed));
3067 Warn if padding is included in a structure, either to align an element
3068 of the structure or to align the whole structure. Sometimes when this
3069 happens it is possible to rearrange the fields of the structure to
3070 reduce the padding and so make the structure smaller.
3072 @item -Wredundant-decls
3073 @opindex Wredundant-decls
3074 Warn if anything is declared more than once in the same scope, even in
3075 cases where multiple declaration is valid and changes nothing.
3077 @item -Wnested-externs @r{(C only)}
3078 @opindex Wnested-externs
3079 Warn if an @code{extern} declaration is encountered within a function.
3081 @item -Wunreachable-code
3082 @opindex Wunreachable-code
3083 Warn if the compiler detects that code will never be executed.
3085 This option is intended to warn when the compiler detects that at
3086 least a whole line of source code will never be executed, because
3087 some condition is never satisfied or because it is after a
3088 procedure that never returns.
3090 It is possible for this option to produce a warning even though there
3091 are circumstances under which part of the affected line can be executed,
3092 so care should be taken when removing apparently-unreachable code.
3094 For instance, when a function is inlined, a warning may mean that the
3095 line is unreachable in only one inlined copy of the function.
3097 This option is not made part of @option{-Wall} because in a debugging
3098 version of a program there is often substantial code which checks
3099 correct functioning of the program and is, hopefully, unreachable
3100 because the program does work. Another common use of unreachable
3101 code is to provide behavior which is selectable at compile-time.
3105 Warn if a function can not be inlined and it was declared as inline.
3106 Even with this option, the compiler will not warn about failures to
3107 inline functions declared in system headers.
3109 The compiler uses a variety of heuristics to determine whether or not
3110 to inline a function. For example, the compiler takes into account
3111 the size of the function being inlined and the the amount of inlining
3112 that has already been done in the current function. Therefore,
3113 seemingly insignificant changes in the source program can cause the
3114 warnings produced by @option{-Winline} to appear or disappear.
3116 @item -Wno-invalid-offsetof @r{(C++ only)}
3117 @opindex Wno-invalid-offsetof
3118 Suppress warnings from applying the @samp{offsetof} macro to a non-POD
3119 type. According to the 1998 ISO C++ standard, applying @samp{offsetof}
3120 to a non-POD type is undefined. In existing C++ implementations,
3121 however, @samp{offsetof} typically gives meaningful results even when
3122 applied to certain kinds of non-POD types. (Such as a simple
3123 @samp{struct} that fails to be a POD type only by virtue of having a
3124 constructor.) This flag is for users who are aware that they are
3125 writing nonportable code and who have deliberately chosen to ignore the
3128 The restrictions on @samp{offsetof} may be relaxed in a future version
3129 of the C++ standard.
3132 @opindex Winvalid-pch
3133 Warn if a precompiled header (@pxref{Precompiled Headers}) is found in
3134 the search path but can't be used.
3138 @opindex Wno-long-long
3139 Warn if @samp{long long} type is used. This is default. To inhibit
3140 the warning messages, use @option{-Wno-long-long}. Flags
3141 @option{-Wlong-long} and @option{-Wno-long-long} are taken into account
3142 only when @option{-pedantic} flag is used.
3144 @item -Wvariadic-macros
3145 @opindex Wvariadic-macros
3146 @opindex Wno-variadic-macros
3147 Warn if variadic macros are used in pedantic ISO C90 mode, or the GNU
3148 alternate syntax when in pedantic ISO C99 mode. This is default.
3149 To inhibit the warning messages, use @option{-Wno-variadic-macros}.
3151 @item -Wdisabled-optimization
3152 @opindex Wdisabled-optimization
3153 Warn if a requested optimization pass is disabled. This warning does
3154 not generally indicate that there is anything wrong with your code; it
3155 merely indicates that GCC's optimizers were unable to handle the code
3156 effectively. Often, the problem is that your code is too big or too
3157 complex; GCC will refuse to optimize programs when the optimization
3158 itself is likely to take inordinate amounts of time.
3160 @item -Wno-pointer-sign
3161 @opindex Wno-pointer-sign
3162 Don't warn for pointer argument passing or assignment with different signedness.
3163 Only useful in the negative form since this warning is enabled by default.
3164 This option is only supported for C and Objective-C@.
3168 Make all warnings into errors.
3171 @node Debugging Options
3172 @section Options for Debugging Your Program or GCC
3173 @cindex options, debugging
3174 @cindex debugging information options
3176 GCC has various special options that are used for debugging
3177 either your program or GCC:
3182 Produce debugging information in the operating system's native format
3183 (stabs, COFF, XCOFF, or DWARF 2)@. GDB can work with this debugging
3186 On most systems that use stabs format, @option{-g} enables use of extra
3187 debugging information that only GDB can use; this extra information
3188 makes debugging work better in GDB but will probably make other debuggers
3190 refuse to read the program. If you want to control for certain whether
3191 to generate the extra information, use @option{-gstabs+}, @option{-gstabs},
3192 @option{-gxcoff+}, @option{-gxcoff}, or @option{-gvms} (see below).
3194 GCC allows you to use @option{-g} with
3195 @option{-O}. The shortcuts taken by optimized code may occasionally
3196 produce surprising results: some variables you declared may not exist
3197 at all; flow of control may briefly move where you did not expect it;
3198 some statements may not be executed because they compute constant
3199 results or their values were already at hand; some statements may
3200 execute in different places because they were moved out of loops.
3202 Nevertheless it proves possible to debug optimized output. This makes
3203 it reasonable to use the optimizer for programs that might have bugs.
3205 The following options are useful when GCC is generated with the
3206 capability for more than one debugging format.
3210 Produce debugging information for use by GDB@. This means to use the
3211 most expressive format available (DWARF 2, stabs, or the native format
3212 if neither of those are supported), including GDB extensions if at all
3217 Produce debugging information in stabs format (if that is supported),
3218 without GDB extensions. This is the format used by DBX on most BSD
3219 systems. On MIPS, Alpha and System V Release 4 systems this option
3220 produces stabs debugging output which is not understood by DBX or SDB@.
3221 On System V Release 4 systems this option requires the GNU assembler.
3223 @item -feliminate-unused-debug-symbols
3224 @opindex feliminate-unused-debug-symbols
3225 Produce debugging information in stabs format (if that is supported),
3226 for only symbols that are actually used.
3230 Produce debugging information in stabs format (if that is supported),
3231 using GNU extensions understood only by the GNU debugger (GDB)@. The
3232 use of these extensions is likely to make other debuggers crash or
3233 refuse to read the program.
3237 Produce debugging information in COFF format (if that is supported).
3238 This is the format used by SDB on most System V systems prior to
3243 Produce debugging information in XCOFF format (if that is supported).
3244 This is the format used by the DBX debugger on IBM RS/6000 systems.
3248 Produce debugging information in XCOFF format (if that is supported),
3249 using GNU extensions understood only by the GNU debugger (GDB)@. The
3250 use of these extensions is likely to make other debuggers crash or
3251 refuse to read the program, and may cause assemblers other than the GNU
3252 assembler (GAS) to fail with an error.
3256 Produce debugging information in DWARF version 2 format (if that is
3257 supported). This is the format used by DBX on IRIX 6. With this
3258 option, GCC uses features of DWARF version 3 when they are useful;
3259 version 3 is upward compatible with version 2, but may still cause
3260 problems for older debuggers.
3264 Produce debugging information in VMS debug format (if that is
3265 supported). This is the format used by DEBUG on VMS systems.
3268 @itemx -ggdb@var{level}
3269 @itemx -gstabs@var{level}
3270 @itemx -gcoff@var{level}
3271 @itemx -gxcoff@var{level}
3272 @itemx -gvms@var{level}
3273 Request debugging information and also use @var{level} to specify how
3274 much information. The default level is 2.
3276 Level 1 produces minimal information, enough for making backtraces in
3277 parts of the program that you don't plan to debug. This includes
3278 descriptions of functions and external variables, but no information
3279 about local variables and no line numbers.
3281 Level 3 includes extra information, such as all the macro definitions
3282 present in the program. Some debuggers support macro expansion when
3283 you use @option{-g3}.
3285 @option{-gdwarf-2} does not accept a concatenated debug level, because
3286 GCC used to support an option @option{-gdwarf} that meant to generate
3287 debug information in version 1 of the DWARF format (which is very
3288 different from version 2), and it would have been too confusing. That
3289 debug format is long obsolete, but the option cannot be changed now.
3290 Instead use an additional @option{-g@var{level}} option to change the
3291 debug level for DWARF2.
3293 @item -feliminate-dwarf2-dups
3294 @opindex feliminate-dwarf2-dups
3295 Compress DWARF2 debugging information by eliminating duplicated
3296 information about each symbol. This option only makes sense when
3297 generating DWARF2 debugging information with @option{-gdwarf-2}.
3299 @cindex @command{prof}
3302 Generate extra code to write profile information suitable for the
3303 analysis program @command{prof}. You must use this option when compiling
3304 the source files you want data about, and you must also use it when
3307 @cindex @command{gprof}
3310 Generate extra code to write profile information suitable for the
3311 analysis program @command{gprof}. You must use this option when compiling
3312 the source files you want data about, and you must also use it when
3317 Makes the compiler print out each function name as it is compiled, and
3318 print some statistics about each pass when it finishes.
3321 @opindex ftime-report
3322 Makes the compiler print some statistics about the time consumed by each
3323 pass when it finishes.
3326 @opindex fmem-report
3327 Makes the compiler print some statistics about permanent memory
3328 allocation when it finishes.
3330 @item -fprofile-arcs
3331 @opindex fprofile-arcs
3332 Add code so that program flow @dfn{arcs} are instrumented. During
3333 execution the program records how many times each branch and call is
3334 executed and how many times it is taken or returns. When the compiled
3335 program exits it saves this data to a file called
3336 @file{@var{auxname}.gcda} for each source file. The data may be used for
3337 profile-directed optimizations (@option{-fbranch-probabilities}), or for
3338 test coverage analysis (@option{-ftest-coverage}). Each object file's
3339 @var{auxname} is generated from the name of the output file, if
3340 explicitly specified and it is not the final executable, otherwise it is
3341 the basename of the source file. In both cases any suffix is removed
3342 (e.g.@: @file{foo.gcda} for input file @file{dir/foo.c}, or
3343 @file{dir/foo.gcda} for output file specified as @option{-o dir/foo.o}).
3348 Compile the source files with @option{-fprofile-arcs} plus optimization
3349 and code generation options. For test coverage analysis, use the
3350 additional @option{-ftest-coverage} option. You do not need to profile
3351 every source file in a program.
3354 Link your object files with @option{-lgcov} or @option{-fprofile-arcs}
3355 (the latter implies the former).
3358 Run the program on a representative workload to generate the arc profile
3359 information. This may be repeated any number of times. You can run
3360 concurrent instances of your program, and provided that the file system
3361 supports locking, the data files will be correctly updated. Also
3362 @code{fork} calls are detected and correctly handled (double counting
3366 For profile-directed optimizations, compile the source files again with
3367 the same optimization and code generation options plus
3368 @option{-fbranch-probabilities} (@pxref{Optimize Options,,Options that
3369 Control Optimization}).
3372 For test coverage analysis, use @command{gcov} to produce human readable
3373 information from the @file{.gcno} and @file{.gcda} files. Refer to the
3374 @command{gcov} documentation for further information.
3378 With @option{-fprofile-arcs}, for each function of your program GCC
3379 creates a program flow graph, then finds a spanning tree for the graph.
3380 Only arcs that are not on the spanning tree have to be instrumented: the
3381 compiler adds code to count the number of times that these arcs are
3382 executed. When an arc is the only exit or only entrance to a block, the
3383 instrumentation code can be added to the block; otherwise, a new basic
3384 block must be created to hold the instrumentation code.
3386 @item -ftree-based-profiling
3387 @opindex ftree-based-profiling
3388 This option is used in addition to @option{-fprofile-arcs} or
3389 @option{-fbranch-probabilities} to control whether those optimizations
3390 are performed on a tree-based or rtl-based internal representation.
3391 If you use this option when compiling with @option{-fprofile-arcs},
3392 you must also use it when compiling later with @option{-fbranch-probabilities}.
3393 Currently the tree-based optimization is in an early stage of
3394 development, and this option is recommended only for those people
3395 working on improving it.
3398 @item -ftest-coverage
3399 @opindex ftest-coverage
3400 Produce a notes file that the @command{gcov} code-coverage utility
3401 (@pxref{Gcov,, @command{gcov}---a Test Coverage Program}) can use to
3402 show program coverage. Each source file's note file is called
3403 @file{@var{auxname}.gcno}. Refer to the @option{-fprofile-arcs} option
3404 above for a description of @var{auxname} and instructions on how to
3405 generate test coverage data. Coverage data will match the source files
3406 more closely, if you do not optimize.
3408 @item -d@var{letters}
3409 @item -fdump-rtl-@var{pass}
3411 Says to make debugging dumps during compilation at times specified by
3412 @var{letters}. This is used for debugging the RTL-based passes of the
3413 compiler. The file names for most of the dumps are made by appending a
3414 pass number and a word to the @var{dumpname}. @var{dumpname} is generated
3415 from the name of the output file, if explicitly specified and it is not
3416 an executable, otherwise it is the basename of the source file.
3418 Most debug dumps can be enabled either passing a letter to the @option{-d}
3419 option, or with a long @option{-fdump-rtl} switch; here are the possible
3420 letters for use in @var{letters} and @var{pass}, and their meanings:
3425 Annotate the assembler output with miscellaneous debugging information.
3428 @itemx -fdump-rtl-bp
3430 @opindex fdump-rtl-bp
3431 Dump after computing branch probabilities, to @file{@var{file}.09.bp}.
3434 @itemx -fdump-rtl-bbro
3436 @opindex fdump-rtl-bbro
3437 Dump after block reordering, to @file{@var{file}.30.bbro}.
3440 @itemx -fdump-rtl-combine
3442 @opindex fdump-rtl-combine
3443 Dump after instruction combination, to the file @file{@var{file}.17.combine}.
3446 @itemx -fdump-rtl-ce1
3447 @itemx -fdump-rtl-ce2
3449 @opindex fdump-rtl-ce1
3450 @opindex fdump-rtl-ce2
3451 @option{-dC} and @option{-fdump-rtl-ce1} enable dumping after the
3452 first if conversion, to the file @file{@var{file}.11.ce1}. @option{-dC}
3453 and @option{-fdump-rtl-ce2} enable dumping after the second if
3454 conversion, to the file @file{@var{file}.18.ce2}.
3457 @itemx -fdump-rtl-btl
3458 @itemx -fdump-rtl-dbr
3460 @opindex fdump-rtl-btl
3461 @opindex fdump-rtl-dbr
3462 @option{-dd} and @option{-fdump-rtl-btl} enable dumping after branch
3463 target load optimization, to to @file{@var{file}.31.btl}. @option{-dd}
3464 and @option{-fdump-rtl-dbr} enable dumping after delayed branch
3465 scheduling, to @file{@var{file}.36.dbr}.
3469 Dump all macro definitions, at the end of preprocessing, in addition to
3473 @itemx -fdump-rtl-ce3
3475 @opindex fdump-rtl-ce3
3476 Dump after the third if conversion, to @file{@var{file}.28.ce3}.
3479 @itemx -fdump-rtl-cfg
3480 @itemx -fdump-rtl-life
3482 @opindex fdump-rtl-cfg
3483 @opindex fdump-rtl-life
3484 @option{-df} and @option{-fdump-rtl-cfg} enable dumping after control
3485 and data flow analysis, to @file{@var{file}.08.cfg}. @option{-df}
3486 and @option{-fdump-rtl-cfg} enable dumping dump after life analysis,
3487 to @file{@var{file}.16.life}.
3490 @itemx -fdump-rtl-greg
3492 @opindex fdump-rtl-greg
3493 Dump after global register allocation, to @file{@var{file}.23.greg}.
3496 @itemx -fdump-rtl-gcse
3497 @itemx -fdump-rtl-bypass
3499 @opindex fdump-rtl-gcse
3500 @opindex fdump-rtl-bypass
3501 @option{-dG} and @option{-fdump-rtl-gcse} enable dumping after GCSE, to
3502 @file{@var{file}.05.gcse}. @option{-dG} and @option{-fdump-rtl-bypass}
3503 enable dumping after jump bypassing and control flow optimizations, to
3504 @file{@var{file}.07.bypass}.
3507 @itemx -fdump-rtl-eh
3509 @opindex fdump-rtl-eh
3510 Dump after finalization of EH handling code, to @file{@var{file}.02.eh}.
3513 @itemx -fdump-rtl-sibling
3515 @opindex fdump-rtl-sibling
3516 Dump after sibling call optimizations, to @file{@var{file}.01.sibling}.
3519 @itemx -fdump-rtl-jump
3521 @opindex fdump-rtl-jump
3522 Dump after the first jump optimization, to @file{@var{file}.03.jump}.
3525 @itemx -fdump-rtl-stack
3527 @opindex fdump-rtl-stack
3528 Dump after conversion from registers to stack, to @file{@var{file}.33.stack}.
3531 @itemx -fdump-rtl-lreg
3533 @opindex fdump-rtl-lreg
3534 Dump after local register allocation, to @file{@var{file}.22.lreg}.
3537 @itemx -fdump-rtl-loop
3538 @itemx -fdump-rtl-loop2
3540 @opindex fdump-rtl-loop
3541 @opindex fdump-rtl-loop2
3542 @option{-dL} and @option{-fdump-rtl-loop} enable dumping after the first
3543 loop optimization pass, to @file{@var{file}.06.loop}. @option{-dL} and
3544 @option{-fdump-rtl-loop2} enable dumping after the second pass, to
3545 @file{@var{file}.13.loop2}.
3548 @itemx -fdump-rtl-sms
3550 @opindex fdump-rtl-sms
3551 Dump after modulo scheduling, to @file{@var{file}.20.sms}.
3554 @itemx -fdump-rtl-mach
3556 @opindex fdump-rtl-mach
3557 Dump after performing the machine dependent reorganization pass, to
3558 @file{@var{file}.35.mach}.
3561 @itemx -fdump-rtl-rnreg
3563 @opindex fdump-rtl-rnreg
3564 Dump after register renumbering, to @file{@var{file}.29.rnreg}.
3567 @itemx -fdump-rtl-regmove
3569 @opindex fdump-rtl-regmove
3570 Dump after the register move pass, to @file{@var{file}.19.regmove}.
3573 @itemx -fdump-rtl-postreload
3575 @opindex fdump-rtl-postreload
3576 Dump after post-reload optimizations, to @file{@var{file}.24.postreload}.
3579 @itemx -fdump-rtl-expand
3581 @opindex fdump-rtl-expand
3582 Dump after RTL generation, to @file{@var{file}.00.expand}.
3585 @itemx -fdump-rtl-sched2
3587 @opindex fdump-rtl-sched2
3588 Dump after the second scheduling pass, to @file{@var{file}.32.sched2}.
3591 @itemx -fdump-rtl-cse
3593 @opindex fdump-rtl-cse
3594 Dump after CSE (including the jump optimization that sometimes follows
3595 CSE), to @file{@var{file}.04.cse}.
3598 @itemx -fdump-rtl-sched
3600 @opindex fdump-rtl-sched
3601 Dump after the first scheduling pass, to @file{@var{file}.21.sched}.
3604 @itemx -fdump-rtl-cse2
3606 @opindex fdump-rtl-cse2
3607 Dump after the second CSE pass (including the jump optimization that
3608 sometimes follows CSE), to @file{@var{file}.15.cse2}.
3611 @itemx -fdump-rtl-tracer
3613 @opindex fdump-rtl-tracer
3614 Dump after running tracer, to @file{@var{file}.12.tracer}.
3617 @itemx -fdump-rtl-vpt
3618 @itemx -fdump-rtl-vartrack
3620 @opindex fdump-rtl-vpt
3621 @opindex fdump-rtl-vartrack
3622 @option{-dV} and @option{-fdump-rtl-vpt} enable dumping after the value
3623 profile transformations, to @file{@var{file}.10.vpt}. @option{-dV}
3624 and @option{-fdump-rtl-vartrack} enable dumping after variable tracking,
3625 to @file{@var{file}.34.vartrack}.
3628 @itemx -fdump-rtl-flow2
3630 @opindex fdump-rtl-flow2
3631 Dump after the second flow pass, to @file{@var{file}.26.flow2}.
3634 @itemx -fdump-rtl-peephole2
3636 @opindex fdump-rtl-peephole2
3637 Dump after the peephole pass, to @file{@var{file}.27.peephole2}.
3640 @itemx -fdump-rtl-web
3642 @opindex fdump-rtl-web
3643 Dump after live range splitting, to @file{@var{file}.14.web}.
3646 @itemx -fdump-rtl-all
3648 @opindex fdump-rtl-all
3649 Produce all the dumps listed above.
3653 Produce a core dump whenever an error occurs.
3657 Print statistics on memory usage, at the end of the run, to
3662 Annotate the assembler output with a comment indicating which
3663 pattern and alternative was used. The length of each instruction is
3668 Dump the RTL in the assembler output as a comment before each instruction.
3669 Also turns on @option{-dp} annotation.
3673 For each of the other indicated dump files (either with @option{-d} or
3674 @option{-fdump-rtl-@var{pass}}), dump a representation of the control flow
3675 graph suitable for viewing with VCG to @file{@var{file}.@var{pass}.vcg}.
3679 Just generate RTL for a function instead of compiling it. Usually used
3680 with @samp{r} (@option{-fdump-rtl-expand}).
3684 Dump debugging information during parsing, to standard error.
3687 @item -fdump-unnumbered
3688 @opindex fdump-unnumbered
3689 When doing debugging dumps (see @option{-d} option above), suppress instruction
3690 numbers and line number note output. This makes it more feasible to
3691 use diff on debugging dumps for compiler invocations with different
3692 options, in particular with and without @option{-g}.
3694 @item -fdump-translation-unit @r{(C and C++ only)}
3695 @itemx -fdump-translation-unit-@var{options} @r{(C and C++ only)}
3696 @opindex fdump-translation-unit
3697 Dump a representation of the tree structure for the entire translation
3698 unit to a file. The file name is made by appending @file{.tu} to the
3699 source file name. If the @samp{-@var{options}} form is used, @var{options}
3700 controls the details of the dump as described for the
3701 @option{-fdump-tree} options.
3703 @item -fdump-class-hierarchy @r{(C++ only)}
3704 @itemx -fdump-class-hierarchy-@var{options} @r{(C++ only)}
3705 @opindex fdump-class-hierarchy
3706 Dump a representation of each class's hierarchy and virtual function
3707 table layout to a file. The file name is made by appending @file{.class}
3708 to the source file name. If the @samp{-@var{options}} form is used,
3709 @var{options} controls the details of the dump as described for the
3710 @option{-fdump-tree} options.
3712 @item -fdump-ipa-@var{switch}
3714 Control the dumping at various stages of inter-procedural analysis
3715 language tree to a file. The file name is generated by appending a switch
3716 specific suffix to the source file name. The following dumps are possible:
3720 Enables all inter-procedural analysis dumps; currently the only produced
3721 dump is the @samp{cgraph} dump.
3724 Dumps information about call-graph optimization, unused function removal,
3725 and inlining decisions.
3728 @item -fdump-tree-@var{switch} @r{(C and C++ only)}
3729 @itemx -fdump-tree-@var{switch}-@var{options} @r{(C and C++ only)}
3731 Control the dumping at various stages of processing the intermediate
3732 language tree to a file. The file name is generated by appending a switch
3733 specific suffix to the source file name. If the @samp{-@var{options}}
3734 form is used, @var{options} is a list of @samp{-} separated options that
3735 control the details of the dump. Not all options are applicable to all
3736 dumps, those which are not meaningful will be ignored. The following
3737 options are available
3741 Print the address of each node. Usually this is not meaningful as it
3742 changes according to the environment and source file. Its primary use
3743 is for tying up a dump file with a debug environment.
3745 Inhibit dumping of members of a scope or body of a function merely
3746 because that scope has been reached. Only dump such items when they
3747 are directly reachable by some other path. When dumping pretty-printed
3748 trees, this option inhibits dumping the bodies of control structures.
3750 Print a raw representation of the tree. By default, trees are
3751 pretty-printed into a C-like representation.
3753 Enable more detailed dumps (not honored by every dump option).
3755 Enable dumping various statistics about the pass (not honored by every dump
3758 Enable showing basic block boundaries (disabled in raw dumps).
3760 Enable showing virtual operands for every statement.
3762 Enable showing line numbers for statements.
3764 Enable showing the unique ID (@code{DECL_UID}) for each variable.
3766 Turn on all options, except @option{raw}, @option{slim} and @option{lineno}.
3769 The following tree dumps are possible:
3773 Dump before any tree based optimization, to @file{@var{file}.original}.
3776 Dump after all tree based optimization, to @file{@var{file}.optimized}.
3779 Dump after function inlining, to @file{@var{file}.inlined}.
3782 @opindex fdump-tree-gimple
3783 Dump each function before and after the gimplification pass to a file. The
3784 file name is made by appending @file{.gimple} to the source file name.
3787 @opindex fdump-tree-cfg
3788 Dump the control flow graph of each function to a file. The file name is
3789 made by appending @file{.cfg} to the source file name.
3792 @opindex fdump-tree-vcg
3793 Dump the control flow graph of each function to a file in VCG format. The
3794 file name is made by appending @file{.vcg} to the source file name. Note
3795 that if the file contains more than one function, the generated file cannot
3796 be used directly by VCG@. You will need to cut and paste each function's
3797 graph into its own separate file first.
3800 @opindex fdump-tree-ch
3801 Dump each function after copying loop headers. The file name is made by
3802 appending @file{.ch} to the source file name.
3805 @opindex fdump-tree-ssa
3806 Dump SSA related information to a file. The file name is made by appending
3807 @file{.ssa} to the source file name.
3810 @opindex fdump-tree-alias
3811 Dump aliasing information for each function. The file name is made by
3812 appending @file{.alias} to the source file name.
3815 @opindex fdump-tree-ccp
3816 Dump each function after CCP@. The file name is made by appending
3817 @file{.ccp} to the source file name.
3820 @opindex fdump-tree-pre
3821 Dump trees after partial redundancy elimination. The file name is made
3822 by appending @file{.pre} to the source file name.
3825 @opindex fdump-tree-fre
3826 Dump trees after full redundancy elimination. The file name is made
3827 by appending @file{.fre} to the source file name.
3830 @opindex fdump-tree-dce
3831 Dump each function after dead code elimination. The file name is made by
3832 appending @file{.dce} to the source file name.
3835 @opindex fdump-tree-mudflap
3836 Dump each function after adding mudflap instrumentation. The file name is
3837 made by appending @file{.mudflap} to the source file name.
3840 @opindex fdump-tree-sra
3841 Dump each function after performing scalar replacement of aggregates. The
3842 file name is made by appending @file{.sra} to the source file name.
3845 @opindex fdump-tree-dom
3846 Dump each function after applying dominator tree optimizations. The file
3847 name is made by appending @file{.dom} to the source file name.
3850 @opindex fdump-tree-dse
3851 Dump each function after applying dead store elimination. The file
3852 name is made by appending @file{.dse} to the source file name.
3855 @opindex fdump-tree-phiopt
3856 Dump each function after optimizing PHI nodes into straightline code. The file
3857 name is made by appending @file{.phiopt} to the source file name.
3860 @opindex fdump-tree-forwprop
3861 Dump each function after forward propagating single use variables. The file
3862 name is made by appending @file{.forwprop} to the source file name.
3865 @opindex fdump-tree-copyrename
3866 Dump each function after applying the copy rename optimization. The file
3867 name is made by appending @file{.copyrename} to the source file name.
3870 @opindex fdump-tree-nrv
3871 Dump each function after applying the named return value optimization on
3872 generic trees. The file name is made by appending @file{.nrv} to the source
3876 @opindex fdump-tree-vect
3877 Dump each function after applying vectorization of loops. The file name is
3878 made by appending @file{.vect} to the source file name.
3881 @opindex fdump-tree-all
3882 Enable all the available tree dumps with the flags provided in this option.
3885 @item -frandom-seed=@var{string}
3886 @opindex frandom-string
3887 This option provides a seed that GCC uses when it would otherwise use
3888 random numbers. It is used to generate certain symbol names
3889 that have to be different in every compiled file. It is also used to
3890 place unique stamps in coverage data files and the object files that
3891 produce them. You can use the @option{-frandom-seed} option to produce
3892 reproducibly identical object files.
3894 The @var{string} should be different for every file you compile.
3896 @item -fsched-verbose=@var{n}
3897 @opindex fsched-verbose
3898 On targets that use instruction scheduling, this option controls the
3899 amount of debugging output the scheduler prints. This information is
3900 written to standard error, unless @option{-dS} or @option{-dR} is
3901 specified, in which case it is output to the usual dump
3902 listing file, @file{.sched} or @file{.sched2} respectively. However
3903 for @var{n} greater than nine, the output is always printed to standard
3906 For @var{n} greater than zero, @option{-fsched-verbose} outputs the
3907 same information as @option{-dRS}. For @var{n} greater than one, it
3908 also output basic block probabilities, detailed ready list information
3909 and unit/insn info. For @var{n} greater than two, it includes RTL
3910 at abort point, control-flow and regions info. And for @var{n} over
3911 four, @option{-fsched-verbose} also includes dependence info.
3915 Store the usual ``temporary'' intermediate files permanently; place them
3916 in the current directory and name them based on the source file. Thus,
3917 compiling @file{foo.c} with @samp{-c -save-temps} would produce files
3918 @file{foo.i} and @file{foo.s}, as well as @file{foo.o}. This creates a
3919 preprocessed @file{foo.i} output file even though the compiler now
3920 normally uses an integrated preprocessor.
3922 When used in combination with the @option{-x} command line option,
3923 @option{-save-temps} is sensible enough to avoid over writing an
3924 input source file with the same extension as an intermediate file.
3925 The corresponding intermediate file may be obtained by renaming the
3926 source file before using @option{-save-temps}.
3930 Report the CPU time taken by each subprocess in the compilation
3931 sequence. For C source files, this is the compiler proper and assembler
3932 (plus the linker if linking is done). The output looks like this:
3939 The first number on each line is the ``user time'', that is time spent
3940 executing the program itself. The second number is ``system time'',
3941 time spent executing operating system routines on behalf of the program.
3942 Both numbers are in seconds.
3944 @item -fvar-tracking
3945 @opindex fvar-tracking
3946 Run variable tracking pass. It computes where variables are stored at each
3947 position in code. Better debugging information is then generated
3948 (if the debugging information format supports this information).
3950 It is enabled by default when compiling with optimization (@option{-Os},
3951 @option{-O}, @option{-O2}, ...), debugging information (@option{-g}) and
3952 the debug info format supports it.
3954 @item -print-file-name=@var{library}
3955 @opindex print-file-name
3956 Print the full absolute name of the library file @var{library} that
3957 would be used when linking---and don't do anything else. With this
3958 option, GCC does not compile or link anything; it just prints the
3961 @item -print-multi-directory
3962 @opindex print-multi-directory
3963 Print the directory name corresponding to the multilib selected by any
3964 other switches present in the command line. This directory is supposed
3965 to exist in @env{GCC_EXEC_PREFIX}.
3967 @item -print-multi-lib
3968 @opindex print-multi-lib
3969 Print the mapping from multilib directory names to compiler switches
3970 that enable them. The directory name is separated from the switches by
3971 @samp{;}, and each switch starts with an @samp{@@} instead of the
3972 @samp{-}, without spaces between multiple switches. This is supposed to
3973 ease shell-processing.
3975 @item -print-prog-name=@var{program}
3976 @opindex print-prog-name
3977 Like @option{-print-file-name}, but searches for a program such as @samp{cpp}.
3979 @item -print-libgcc-file-name
3980 @opindex print-libgcc-file-name
3981 Same as @option{-print-file-name=libgcc.a}.
3983 This is useful when you use @option{-nostdlib} or @option{-nodefaultlibs}
3984 but you do want to link with @file{libgcc.a}. You can do
3987 gcc -nostdlib @var{files}@dots{} `gcc -print-libgcc-file-name`
3990 @item -print-search-dirs
3991 @opindex print-search-dirs
3992 Print the name of the configured installation directory and a list of
3993 program and library directories @command{gcc} will search---and don't do anything else.
3995 This is useful when @command{gcc} prints the error message
3996 @samp{installation problem, cannot exec cpp0: No such file or directory}.
3997 To resolve this you either need to put @file{cpp0} and the other compiler
3998 components where @command{gcc} expects to find them, or you can set the environment
3999 variable @env{GCC_EXEC_PREFIX} to the directory where you installed them.
4000 Don't forget the trailing @samp{/}.
4001 @xref{Environment Variables}.
4004 @opindex dumpmachine
4005 Print the compiler's target machine (for example,
4006 @samp{i686-pc-linux-gnu})---and don't do anything else.
4009 @opindex dumpversion
4010 Print the compiler version (for example, @samp{3.0})---and don't do
4015 Print the compiler's built-in specs---and don't do anything else. (This
4016 is used when GCC itself is being built.) @xref{Spec Files}.
4018 @item -feliminate-unused-debug-types
4019 @opindex feliminate-unused-debug-types
4020 Normally, when producing DWARF2 output, GCC will emit debugging
4021 information for all types declared in a compilation
4022 unit, regardless of whether or not they are actually used
4023 in that compilation unit. Sometimes this is useful, such as
4024 if, in the debugger, you want to cast a value to a type that is
4025 not actually used in your program (but is declared). More often,
4026 however, this results in a significant amount of wasted space.
4027 With this option, GCC will avoid producing debug symbol output
4028 for types that are nowhere used in the source file being compiled.
4031 @node Optimize Options
4032 @section Options That Control Optimization
4033 @cindex optimize options
4034 @cindex options, optimization
4036 These options control various sorts of optimizations.
4038 Without any optimization option, the compiler's goal is to reduce the
4039 cost of compilation and to make debugging produce the expected
4040 results. Statements are independent: if you stop the program with a
4041 breakpoint between statements, you can then assign a new value to any
4042 variable or change the program counter to any other statement in the
4043 function and get exactly the results you would expect from the source
4046 Turning on optimization flags makes the compiler attempt to improve
4047 the performance and/or code size at the expense of compilation time
4048 and possibly the ability to debug the program.
4050 The compiler performs optimization based on the knowledge it has of
4051 the program. Optimization levels @option{-O2} and above, in
4052 particular, enable @emph{unit-at-a-time} mode, which allows the
4053 compiler to consider information gained from later functions in
4054 the file when compiling a function. Compiling multiple files at
4055 once to a single output file in @emph{unit-at-a-time} mode allows
4056 the compiler to use information gained from all of the files when
4057 compiling each of them.
4059 Not all optimizations are controlled directly by a flag. Only
4060 optimizations that have a flag are listed.
4067 Optimize. Optimizing compilation takes somewhat more time, and a lot
4068 more memory for a large function.
4070 With @option{-O}, the compiler tries to reduce code size and execution
4071 time, without performing any optimizations that take a great deal of
4074 @option{-O} turns on the following optimization flags:
4075 @gccoptlist{-fdefer-pop @gol
4076 -fdelayed-branch @gol
4077 -fguess-branch-probability @gol
4078 -fcprop-registers @gol
4079 -floop-optimize @gol
4080 -fif-conversion @gol
4081 -fif-conversion2 @gol
4087 -ftree-live_range_split @gol
4089 -ftree-copyrename @gol
4094 @option{-O} also turns on @option{-fomit-frame-pointer} on machines
4095 where doing so does not interfere with debugging.
4099 Optimize even more. GCC performs nearly all supported optimizations
4100 that do not involve a space-speed tradeoff. The compiler does not
4101 perform loop unrolling or function inlining when you specify @option{-O2}.
4102 As compared to @option{-O}, this option increases both compilation time
4103 and the performance of the generated code.
4105 @option{-O2} turns on all optimization flags specified by @option{-O}. It
4106 also turns on the following optimization flags:
4107 @gccoptlist{-fthread-jumps @gol
4109 -foptimize-sibling-calls @gol
4110 -fcse-follow-jumps -fcse-skip-blocks @gol
4111 -fgcse -fgcse-lm @gol
4112 -fexpensive-optimizations @gol
4113 -fstrength-reduce @gol
4114 -frerun-cse-after-loop -frerun-loop-opt @gol
4118 -fschedule-insns -fschedule-insns2 @gol
4119 -fsched-interblock -fsched-spec @gol
4121 -fstrict-aliasing @gol
4122 -fdelete-null-pointer-checks @gol
4123 -freorder-blocks -freorder-functions @gol
4124 -funit-at-a-time @gol
4125 -falign-functions -falign-jumps @gol
4126 -falign-loops -falign-labels @gol
4129 Please note the warning under @option{-fgcse} about
4130 invoking @option{-O2} on programs that use computed gotos.
4134 Optimize yet more. @option{-O3} turns on all optimizations specified by
4135 @option{-O2} and also turns on the @option{-finline-functions},
4136 @option{-funswitch-loops} and @option{-fgcse-after-reload} options.
4140 Do not optimize. This is the default.
4144 Optimize for size. @option{-Os} enables all @option{-O2} optimizations that
4145 do not typically increase code size. It also performs further
4146 optimizations designed to reduce code size.
4148 @option{-Os} disables the following optimization flags:
4149 @gccoptlist{-falign-functions -falign-jumps -falign-loops @gol
4150 -falign-labels -freorder-blocks -freorder-blocks-and-partition -fprefetch-loop-arrays}
4152 If you use multiple @option{-O} options, with or without level numbers,
4153 the last such option is the one that is effective.
4156 Options of the form @option{-f@var{flag}} specify machine-independent
4157 flags. Most flags have both positive and negative forms; the negative
4158 form of @option{-ffoo} would be @option{-fno-foo}. In the table
4159 below, only one of the forms is listed---the one you typically will
4160 use. You can figure out the other form by either removing @samp{no-}
4163 The following options control specific optimizations. They are either
4164 activated by @option{-O} options or are related to ones that are. You
4165 can use the following flags in the rare cases when ``fine-tuning'' of
4166 optimizations to be performed is desired.
4169 @item -fno-default-inline
4170 @opindex fno-default-inline
4171 Do not make member functions inline by default merely because they are
4172 defined inside the class scope (C++ only). Otherwise, when you specify
4173 @w{@option{-O}}, member functions defined inside class scope are compiled
4174 inline by default; i.e., you don't need to add @samp{inline} in front of
4175 the member function name.
4177 @item -fno-defer-pop
4178 @opindex fno-defer-pop
4179 Always pop the arguments to each function call as soon as that function
4180 returns. For machines which must pop arguments after a function call,
4181 the compiler normally lets arguments accumulate on the stack for several
4182 function calls and pops them all at once.
4184 Disabled at levels @option{-O}, @option{-O2}, @option{-O3}, @option{-Os}.
4188 Force memory operands to be copied into registers before doing
4189 arithmetic on them. This produces better code by making all memory
4190 references potential common subexpressions. When they are not common
4191 subexpressions, instruction combination should eliminate the separate
4194 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
4197 @opindex fforce-addr
4198 Force memory address constants to be copied into registers before
4199 doing arithmetic on them. This may produce better code just as
4200 @option{-fforce-mem} may.
4202 @item -fomit-frame-pointer
4203 @opindex fomit-frame-pointer
4204 Don't keep the frame pointer in a register for functions that
4205 don't need one. This avoids the instructions to save, set up and
4206 restore frame pointers; it also makes an extra register available
4207 in many functions. @strong{It also makes debugging impossible on
4210 On some machines, such as the VAX, this flag has no effect, because
4211 the standard calling sequence automatically handles the frame pointer
4212 and nothing is saved by pretending it doesn't exist. The
4213 machine-description macro @code{FRAME_POINTER_REQUIRED} controls
4214 whether a target machine supports this flag. @xref{Registers,,Register
4215 Usage, gccint, GNU Compiler Collection (GCC) Internals}.
4217 Enabled at levels @option{-O}, @option{-O2}, @option{-O3}, @option{-Os}.
4219 @item -foptimize-sibling-calls
4220 @opindex foptimize-sibling-calls
4221 Optimize sibling and tail recursive calls.
4223 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
4227 Don't pay attention to the @code{inline} keyword. Normally this option
4228 is used to keep the compiler from expanding any functions inline.
4229 Note that if you are not optimizing, no functions can be expanded inline.
4231 @item -finline-functions
4232 @opindex finline-functions
4233 Integrate all simple functions into their callers. The compiler
4234 heuristically decides which functions are simple enough to be worth
4235 integrating in this way.
4237 If all calls to a given function are integrated, and the function is
4238 declared @code{static}, then the function is normally not output as
4239 assembler code in its own right.
4241 Enabled at level @option{-O3}.
4243 @item -finline-limit=@var{n}
4244 @opindex finline-limit
4245 By default, GCC limits the size of functions that can be inlined. This flag
4246 allows the control of this limit for functions that are explicitly marked as
4247 inline (i.e., marked with the inline keyword or defined within the class
4248 definition in c++). @var{n} is the size of functions that can be inlined in
4249 number of pseudo instructions (not counting parameter handling). The default
4250 value of @var{n} is 600.
4251 Increasing this value can result in more inlined code at
4252 the cost of compilation time and memory consumption. Decreasing usually makes
4253 the compilation faster and less code will be inlined (which presumably
4254 means slower programs). This option is particularly useful for programs that
4255 use inlining heavily such as those based on recursive templates with C++.
4257 Inlining is actually controlled by a number of parameters, which may be
4258 specified individually by using @option{--param @var{name}=@var{value}}.
4259 The @option{-finline-limit=@var{n}} option sets some of these parameters
4263 @item max-inline-insns-single
4264 is set to @var{n}/2.
4265 @item max-inline-insns-auto
4266 is set to @var{n}/2.
4267 @item min-inline-insns
4268 is set to 130 or @var{n}/4, whichever is smaller.
4269 @item max-inline-insns-rtl
4273 See below for a documentation of the individual
4274 parameters controlling inlining.
4276 @emph{Note:} pseudo instruction represents, in this particular context, an
4277 abstract measurement of function's size. In no way, it represents a count
4278 of assembly instructions and as such its exact meaning might change from one
4279 release to an another.
4281 @item -fkeep-inline-functions
4282 @opindex fkeep-inline-functions
4283 In C, emit @code{static} functions that are declared @code{inline}
4284 into the object file, even if the function has been inlined into all
4285 of its callers. This switch does not affect functions using the
4286 @code{extern inline} extension in GNU C@. In C++, emit any and all
4287 inline functions into the object file.
4289 @item -fkeep-static-consts
4290 @opindex fkeep-static-consts
4291 Emit variables declared @code{static const} when optimization isn't turned
4292 on, even if the variables aren't referenced.
4294 GCC enables this option by default. If you want to force the compiler to
4295 check if the variable was referenced, regardless of whether or not
4296 optimization is turned on, use the @option{-fno-keep-static-consts} option.
4298 @item -fmerge-constants
4299 Attempt to merge identical constants (string constants and floating point
4300 constants) across compilation units.
4302 This option is the default for optimized compilation if the assembler and
4303 linker support it. Use @option{-fno-merge-constants} to inhibit this
4306 Enabled at levels @option{-O}, @option{-O2}, @option{-O3}, @option{-Os}.
4308 @item -fmerge-all-constants
4309 Attempt to merge identical constants and identical variables.
4311 This option implies @option{-fmerge-constants}. In addition to
4312 @option{-fmerge-constants} this considers e.g.@: even constant initialized
4313 arrays or initialized constant variables with integral or floating point
4314 types. Languages like C or C++ require each non-automatic variable to
4315 have distinct location, so using this option will result in non-conforming
4318 @item -fmodulo-sched
4319 @opindex fmodulo-sched
4320 Perform swing modulo scheduling immediately before the first scheduling
4321 pass. This pass looks at innermost loops and reorders their
4322 instructions by overlapping different iterations.
4324 @item -fno-branch-count-reg
4325 @opindex fno-branch-count-reg
4326 Do not use ``decrement and branch'' instructions on a count register,
4327 but instead generate a sequence of instructions that decrement a
4328 register, compare it against zero, then branch based upon the result.
4329 This option is only meaningful on architectures that support such
4330 instructions, which include x86, PowerPC, IA-64 and S/390.
4332 The default is @option{-fbranch-count-reg}, enabled when
4333 @option{-fstrength-reduce} is enabled.
4335 @item -fno-function-cse
4336 @opindex fno-function-cse
4337 Do not put function addresses in registers; make each instruction that
4338 calls a constant function contain the function's address explicitly.
4340 This option results in less efficient code, but some strange hacks
4341 that alter the assembler output may be confused by the optimizations
4342 performed when this option is not used.
4344 The default is @option{-ffunction-cse}
4346 @item -fno-zero-initialized-in-bss
4347 @opindex fno-zero-initialized-in-bss
4348 If the target supports a BSS section, GCC by default puts variables that
4349 are initialized to zero into BSS@. This can save space in the resulting
4352 This option turns off this behavior because some programs explicitly
4353 rely on variables going to the data section. E.g., so that the
4354 resulting executable can find the beginning of that section and/or make
4355 assumptions based on that.
4357 The default is @option{-fzero-initialized-in-bss}.
4359 @item -fbounds-check
4360 @opindex fbounds-check
4361 For front-ends that support it, generate additional code to check that
4362 indices used to access arrays are within the declared range. This is
4363 currently only supported by the Java and Fortran front-ends, where
4364 this option defaults to true and false respectively.
4366 @item -fmudflap -fmudflapth -fmudflapir
4370 @cindex bounds checking
4372 For front-ends that support it (C and C++), instrument all risky
4373 pointer/array dereferencing operations, some standard library
4374 string/heap functions, and some other associated constructs with
4375 range/validity tests. Modules so instrumented should be immune to
4376 buffer overflows, invalid heap use, and some other classes of C/C++
4377 programming errors. The instrumentation relies on a separate runtime
4378 library (@file{libmudflap}), which will be linked into a program if
4379 @option{-fmudflap} is given at link time. Run-time behavior of the
4380 instrumented program is controlled by the @env{MUDFLAP_OPTIONS}
4381 environment variable. See @code{env MUDFLAP_OPTIONS=-help a.out}
4384 Use @option{-fmudflapth} instead of @option{-fmudflap} to compile and to
4385 link if your program is multi-threaded. Use @option{-fmudflapir}, in
4386 addition to @option{-fmudflap} or @option{-fmudflapth}, if
4387 instrumentation should ignore pointer reads. This produces less
4388 instrumentation (and therefore faster execution) and still provides
4389 some protection against outright memory corrupting writes, but allows
4390 erroneously read data to propagate within a program.
4392 @item -fstrength-reduce
4393 @opindex fstrength-reduce
4394 Perform the optimizations of loop strength reduction and
4395 elimination of iteration variables.
4397 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
4399 @item -fthread-jumps
4400 @opindex fthread-jumps
4401 Perform optimizations where we check to see if a jump branches to a
4402 location where another comparison subsumed by the first is found. If
4403 so, the first branch is redirected to either the destination of the
4404 second branch or a point immediately following it, depending on whether
4405 the condition is known to be true or false.
4407 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
4409 @item -fcse-follow-jumps
4410 @opindex fcse-follow-jumps
4411 In common subexpression elimination, scan through jump instructions
4412 when the target of the jump is not reached by any other path. For
4413 example, when CSE encounters an @code{if} statement with an
4414 @code{else} clause, CSE will follow the jump when the condition
4417 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
4419 @item -fcse-skip-blocks
4420 @opindex fcse-skip-blocks
4421 This is similar to @option{-fcse-follow-jumps}, but causes CSE to
4422 follow jumps which conditionally skip over blocks. When CSE
4423 encounters a simple @code{if} statement with no else clause,
4424 @option{-fcse-skip-blocks} causes CSE to follow the jump around the
4425 body of the @code{if}.
4427 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
4429 @item -frerun-cse-after-loop
4430 @opindex frerun-cse-after-loop
4431 Re-run common subexpression elimination after loop optimizations has been
4434 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
4436 @item -frerun-loop-opt
4437 @opindex frerun-loop-opt
4438 Run the loop optimizer twice.
4440 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
4444 Perform a global common subexpression elimination pass.
4445 This pass also performs global constant and copy propagation.
4447 @emph{Note:} When compiling a program using computed gotos, a GCC
4448 extension, you may get better runtime performance if you disable
4449 the global common subexpression elimination pass by adding
4450 @option{-fno-gcse} to the command line.
4452 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
4456 When @option{-fgcse-lm} is enabled, global common subexpression elimination will
4457 attempt to move loads which are only killed by stores into themselves. This
4458 allows a loop containing a load/store sequence to be changed to a load outside
4459 the loop, and a copy/store within the loop.
4461 Enabled by default when gcse is enabled.
4465 When @option{-fgcse-sm} is enabled, a store motion pass is run after
4466 global common subexpression elimination. This pass will attempt to move
4467 stores out of loops. When used in conjunction with @option{-fgcse-lm},
4468 loops containing a load/store sequence can be changed to a load before
4469 the loop and a store after the loop.
4471 Not enabled at any optimization level.
4475 When @option{-fgcse-las} is enabled, the global common subexpression
4476 elimination pass eliminates redundant loads that come after stores to the
4477 same memory location (both partial and full redundancies).
4479 Not enabled at any optimization level.
4481 @item -fgcse-after-reload
4482 @opindex fgcse-after-reload
4483 When @option{-fgcse-after-reload} is enabled, a redundant load elimination
4484 pass is performed after reload. The purpose of this pass is to cleanup
4487 @item -floop-optimize
4488 @opindex floop-optimize
4489 Perform loop optimizations: move constant expressions out of loops, simplify
4490 exit test conditions and optionally do strength-reduction as well.
4492 Enabled at levels @option{-O}, @option{-O2}, @option{-O3}, @option{-Os}.
4494 @item -floop-optimize2
4495 @opindex floop-optimize2
4496 Perform loop optimizations using the new loop optimizer. The optimizations
4497 (loop unrolling, peeling and unswitching, loop invariant motion) are enabled
4500 @item -fcrossjumping
4501 @opindex crossjumping
4502 Perform cross-jumping transformation. This transformation unifies equivalent code and save code size. The
4503 resulting code may or may not perform better than without cross-jumping.
4505 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
4507 @item -fif-conversion
4508 @opindex if-conversion
4509 Attempt to transform conditional jumps into branch-less equivalents. This
4510 include use of conditional moves, min, max, set flags and abs instructions, and
4511 some tricks doable by standard arithmetics. The use of conditional execution
4512 on chips where it is available is controlled by @code{if-conversion2}.
4514 Enabled at levels @option{-O}, @option{-O2}, @option{-O3}, @option{-Os}.
4516 @item -fif-conversion2
4517 @opindex if-conversion2
4518 Use conditional execution (where available) to transform conditional jumps into
4519 branch-less equivalents.
4521 Enabled at levels @option{-O}, @option{-O2}, @option{-O3}, @option{-Os}.
4523 @item -fdelete-null-pointer-checks
4524 @opindex fdelete-null-pointer-checks
4525 Use global dataflow analysis to identify and eliminate useless checks
4526 for null pointers. The compiler assumes that dereferencing a null
4527 pointer would have halted the program. If a pointer is checked after
4528 it has already been dereferenced, it cannot be null.
4530 In some environments, this assumption is not true, and programs can
4531 safely dereference null pointers. Use
4532 @option{-fno-delete-null-pointer-checks} to disable this optimization
4533 for programs which depend on that behavior.
4535 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
4537 @item -fexpensive-optimizations
4538 @opindex fexpensive-optimizations
4539 Perform a number of minor optimizations that are relatively expensive.
4541 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
4543 @item -foptimize-register-move
4545 @opindex foptimize-register-move
4547 Attempt to reassign register numbers in move instructions and as
4548 operands of other simple instructions in order to maximize the amount of
4549 register tying. This is especially helpful on machines with two-operand
4552 Note @option{-fregmove} and @option{-foptimize-register-move} are the same
4555 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
4557 @item -fdelayed-branch
4558 @opindex fdelayed-branch
4559 If supported for the target machine, attempt to reorder instructions
4560 to exploit instruction slots available after delayed branch
4563 Enabled at levels @option{-O}, @option{-O2}, @option{-O3}, @option{-Os}.
4565 @item -fschedule-insns
4566 @opindex fschedule-insns
4567 If supported for the target machine, attempt to reorder instructions to
4568 eliminate execution stalls due to required data being unavailable. This
4569 helps machines that have slow floating point or memory load instructions
4570 by allowing other instructions to be issued until the result of the load
4571 or floating point instruction is required.
4573 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
4575 @item -fschedule-insns2
4576 @opindex fschedule-insns2
4577 Similar to @option{-fschedule-insns}, but requests an additional pass of
4578 instruction scheduling after register allocation has been done. This is
4579 especially useful on machines with a relatively small number of
4580 registers and where memory load instructions take more than one cycle.
4582 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
4584 @item -fno-sched-interblock
4585 @opindex fno-sched-interblock
4586 Don't schedule instructions across basic blocks. This is normally
4587 enabled by default when scheduling before register allocation, i.e.@:
4588 with @option{-fschedule-insns} or at @option{-O2} or higher.
4590 @item -fno-sched-spec
4591 @opindex fno-sched-spec
4592 Don't allow speculative motion of non-load instructions. This is normally
4593 enabled by default when scheduling before register allocation, i.e.@:
4594 with @option{-fschedule-insns} or at @option{-O2} or higher.
4596 @item -fsched-spec-load
4597 @opindex fsched-spec-load
4598 Allow speculative motion of some load instructions. This only makes
4599 sense when scheduling before register allocation, i.e.@: with
4600 @option{-fschedule-insns} or at @option{-O2} or higher.
4602 @item -fsched-spec-load-dangerous
4603 @opindex fsched-spec-load-dangerous
4604 Allow speculative motion of more load instructions. This only makes
4605 sense when scheduling before register allocation, i.e.@: with
4606 @option{-fschedule-insns} or at @option{-O2} or higher.
4608 @item -fsched-stalled-insns=@var{n}
4609 @opindex fsched-stalled-insns
4610 Define how many insns (if any) can be moved prematurely from the queue
4611 of stalled insns into the ready list, during the second scheduling pass.
4613 @item -fsched-stalled-insns-dep=@var{n}
4614 @opindex fsched-stalled-insns-dep
4615 Define how many insn groups (cycles) will be examined for a dependency
4616 on a stalled insn that is candidate for premature removal from the queue
4617 of stalled insns. Has an effect only during the second scheduling pass,
4618 and only if @option{-fsched-stalled-insns} is used and its value is not zero.
4620 @item -fsched2-use-superblocks
4621 @opindex fsched2-use-superblocks
4622 When scheduling after register allocation, do use superblock scheduling
4623 algorithm. Superblock scheduling allows motion across basic block boundaries
4624 resulting on faster schedules. This option is experimental, as not all machine
4625 descriptions used by GCC model the CPU closely enough to avoid unreliable
4626 results from the algorithm.
4628 This only makes sense when scheduling after register allocation, i.e.@: with
4629 @option{-fschedule-insns2} or at @option{-O2} or higher.
4631 @item -fsched2-use-traces
4632 @opindex fsched2-use-traces
4633 Use @option{-fsched2-use-superblocks} algorithm when scheduling after register
4634 allocation and additionally perform code duplication in order to increase the
4635 size of superblocks using tracer pass. See @option{-ftracer} for details on
4638 This mode should produce faster but significantly longer programs. Also
4639 without @option{-fbranch-probabilities} the traces constructed may not
4640 match the reality and hurt the performance. This only makes
4641 sense when scheduling after register allocation, i.e.@: with
4642 @option{-fschedule-insns2} or at @option{-O2} or higher.
4644 @item -freschedule-modulo-scheduled-loops
4645 @opindex fscheduling-in-modulo-scheduled-loops
4646 The modulo scheduling comes before the traditional scheduling, if a loop was modulo scheduled
4647 we may want to prevent the later scheduling passes from changing its schedule, we use this
4648 option to control that.
4650 @item -fcaller-saves
4651 @opindex fcaller-saves
4652 Enable values to be allocated in registers that will be clobbered by
4653 function calls, by emitting extra instructions to save and restore the
4654 registers around such calls. Such allocation is done only when it
4655 seems to result in better code than would otherwise be produced.
4657 This option is always enabled by default on certain machines, usually
4658 those which have no call-preserved registers to use instead.
4660 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
4663 Perform Partial Redundancy Elimination (PRE) on trees. This flag is
4664 enabled by default at @option{-O2} and @option{-O3}.
4667 Perform Full Redundancy Elimination (FRE) on trees. The difference
4668 between FRE and PRE is that FRE only considers expressions
4669 that are computed on all paths leading to the redundant computation.
4670 This analysis faster than PRE, though it exposes fewer redundancies.
4671 This flag is enabled by default at @option{-O} and higher.
4674 Perform sparse conditional constant propagation (CCP) on trees. This flag
4675 is enabled by default at @option{-O} and higher.
4678 Perform dead code elimination (DCE) on trees. This flag is enabled by
4679 default at @option{-O} and higher.
4681 @item -ftree-dominator-opts
4682 Perform dead code elimination (DCE) on trees. This flag is enabled by
4683 default at @option{-O} and higher.
4686 Perform loop header copying on trees. This is beneficial since it increases
4687 effectiveness of code motion optimizations. It also saves one jump. This flag
4688 is enabled by default at @option{-O} and higher. It is not enabled
4689 for @option{-Os}, since it usually increases code size.
4691 @item -ftree-loop-optimize
4692 Perform loop optimizations on trees. This flag is enabled by default
4693 at @option{-O} and higher.
4695 @item -ftree-loop-linear
4696 Perform linear loop transformations on tree. This flag can improve cache
4697 performance and allow further loop optimizations to take place.
4700 Perform loop invariant motion on trees. This pass moves only invartiants that
4701 would be hard to handle on rtl level (function calls, operations that expand to
4702 nontrivial sequences of insns). With @option{-funswitch-loops} it also moves
4703 operands of conditions that are invariant out of the loop, so that we can use
4704 just trivial invariantness analysis in loop unswitching. The pass also includes
4708 Create a canonical counter for number of iterations in the loop for that
4709 determining number of iterations requires complicated analysis. Later
4710 optimizations then may determine the number easily. Useful especially
4711 in connection with unrolling.
4714 Perform induction variable optimizations (strength reduction, induction
4715 variable merging and induction variable elimination) on trees.
4718 Perform scalar replacement of aggregates. This pass replaces structure
4719 references with scalars to prevent committing structures to memory too
4720 early. This flag is enabled by default at @option{-O} and higher.
4722 @item -ftree-copyrename
4723 Perform copy renaming on trees. This pass attempts to rename compiler
4724 temporaries to other variables at copy locations, usually resulting in
4725 variable names which more closely resemble the original variables. This flag
4726 is enabled by default at @option{-O} and higher.
4729 Perform temporary expression replacement during the SSA->normal phase. Single
4730 use/single def temporaries are replaced at their use location with their
4731 defining expression. This results in non-GIMPLE code, but gives the expanders
4732 much more complex trees to work on resulting in better RTL generation. This is
4733 enabled by default at @option{-O} and higher.
4736 Perform live range splitting during the SSA->normal phase. Distinct live
4737 ranges of a variable are split into unique variables, allowing for better
4738 optimization later. This is enabled by default at @option{-O} and higher.
4740 @item -ftree-vectorize
4741 Perform loop vectorization on trees.
4745 Perform tail duplication to enlarge superblock size. This transformation
4746 simplifies the control flow of the function allowing other optimizations to do
4749 @item -funroll-loops
4750 @opindex funroll-loops
4751 Unroll loops whose number of iterations can be determined at compile
4752 time or upon entry to the loop. @option{-funroll-loops} implies both
4753 @option{-fstrength-reduce} and @option{-frerun-cse-after-loop}. This
4754 option makes code larger, and may or may not make it run faster.
4756 @item -funroll-all-loops
4757 @opindex funroll-all-loops
4758 Unroll all loops, even if their number of iterations is uncertain when
4759 the loop is entered. This usually makes programs run more slowly.
4760 @option{-funroll-all-loops} implies the same options as
4761 @option{-funroll-loops},
4763 @item -fsplit-ivs-in-unroller
4764 @opindex -fsplit-ivs-in-unroller
4765 Enables expressing of values of induction variables in later iterations
4766 of the unrolled loop using the value in the first iteration. This breaks
4767 long dependency chains, thus improving efficiency of the scheduling passes
4768 (for best results, @option{-fweb} should be used as well).
4770 Combination of @option{-fweb} and CSE is often sufficient to obtain the
4771 same effect. However in cases the loop body is more complicated than
4772 a single basic block, this is not reliable. It also does not work at all
4773 on some of the architectures due to restrictions in the CSE pass.
4775 This optimization is enabled by default.
4777 @item -fvariable-expansion-in-unroller
4778 @opindex -fvariable-expansion-in-unroller
4779 With this option, the compiler will create multiple copies of some
4780 local variables when unrolling a loop which can result in superior code.
4782 @item -fprefetch-loop-arrays
4783 @opindex fprefetch-loop-arrays
4784 If supported by the target machine, generate instructions to prefetch
4785 memory to improve the performance of loops that access large arrays.
4787 These options may generate better or worse code; results are highly
4788 dependent on the structure of loops within the source code.
4791 @itemx -fno-peephole2
4792 @opindex fno-peephole
4793 @opindex fno-peephole2
4794 Disable any machine-specific peephole optimizations. The difference
4795 between @option{-fno-peephole} and @option{-fno-peephole2} is in how they
4796 are implemented in the compiler; some targets use one, some use the
4797 other, a few use both.
4799 @option{-fpeephole} is enabled by default.
4800 @option{-fpeephole2} enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
4802 @item -fno-guess-branch-probability
4803 @opindex fno-guess-branch-probability
4804 Do not guess branch probabilities using heuristics.
4806 GCC will use heuristics to guess branch probabilities if they are
4807 not provided by profiling feedback (@option{-fprofile-arcs}). These
4808 heuristics are based on the control flow graph. If some branch probabilities
4809 are specified by @samp{__builtin_expect}, then the heuristics will be
4810 used to guess branch probabilities for the rest of the control flow graph,
4811 taking the @samp{__builtin_expect} info into account. The interactions
4812 between the heuristics and @samp{__builtin_expect} can be complex, and in
4813 some cases, it may be useful to disable the heuristics so that the effects
4814 of @samp{__builtin_expect} are easier to understand.
4816 The default is @option{-fguess-branch-probability} at levels
4817 @option{-O}, @option{-O2}, @option{-O3}, @option{-Os}.
4819 @item -freorder-blocks
4820 @opindex freorder-blocks
4821 Reorder basic blocks in the compiled function in order to reduce number of
4822 taken branches and improve code locality.
4824 Enabled at levels @option{-O2}, @option{-O3}.
4826 @item -freorder-blocks-and-partition
4827 @opindex freorder-blocks-and-partition
4828 In addition to reordering basic blocks in the compiled function, in order
4829 to reduce number of taken branches, partitions hot and cold basic blocks
4830 into separate sections of the assembly and .o files, to improve
4831 paging and cache locality performance.
4833 This optimization is automatically turned off in the presence of
4834 exception handling, for linkonce sections, for functions with a user-defined
4835 section attribute and on any architecture that does not support named
4838 @item -freorder-functions
4839 @opindex freorder-functions
4840 Reorder functions in the object file in order to
4841 improve code locality. This is implemented by using special
4842 subsections @code{.text.hot} for most frequently executed functions and
4843 @code{.text.unlikely} for unlikely executed functions. Reordering is done by
4844 the linker so object file format must support named sections and linker must
4845 place them in a reasonable way.
4847 Also profile feedback must be available in to make this option effective. See
4848 @option{-fprofile-arcs} for details.
4850 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
4852 @item -fstrict-aliasing
4853 @opindex fstrict-aliasing
4854 Allows the compiler to assume the strictest aliasing rules applicable to
4855 the language being compiled. For C (and C++), this activates
4856 optimizations based on the type of expressions. In particular, an
4857 object of one type is assumed never to reside at the same address as an
4858 object of a different type, unless the types are almost the same. For
4859 example, an @code{unsigned int} can alias an @code{int}, but not a
4860 @code{void*} or a @code{double}. A character type may alias any other
4863 Pay special attention to code like this:
4876 The practice of reading from a different union member than the one most
4877 recently written to (called ``type-punning'') is common. Even with
4878 @option{-fstrict-aliasing}, type-punning is allowed, provided the memory
4879 is accessed through the union type. So, the code above will work as
4880 expected. However, this code might not:
4891 Every language that wishes to perform language-specific alias analysis
4892 should define a function that computes, given an @code{tree}
4893 node, an alias set for the node. Nodes in different alias sets are not
4894 allowed to alias. For an example, see the C front-end function
4895 @code{c_get_alias_set}.
4897 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
4899 @item -falign-functions
4900 @itemx -falign-functions=@var{n}
4901 @opindex falign-functions
4902 Align the start of functions to the next power-of-two greater than
4903 @var{n}, skipping up to @var{n} bytes. For instance,
4904 @option{-falign-functions=32} aligns functions to the next 32-byte
4905 boundary, but @option{-falign-functions=24} would align to the next
4906 32-byte boundary only if this can be done by skipping 23 bytes or less.
4908 @option{-fno-align-functions} and @option{-falign-functions=1} are
4909 equivalent and mean that functions will not be aligned.
4911 Some assemblers only support this flag when @var{n} is a power of two;
4912 in that case, it is rounded up.
4914 If @var{n} is not specified or is zero, use a machine-dependent default.
4916 Enabled at levels @option{-O2}, @option{-O3}.
4918 @item -falign-labels
4919 @itemx -falign-labels=@var{n}
4920 @opindex falign-labels
4921 Align all branch targets to a power-of-two boundary, skipping up to
4922 @var{n} bytes like @option{-falign-functions}. This option can easily
4923 make code slower, because it must insert dummy operations for when the
4924 branch target is reached in the usual flow of the code.
4926 @option{-fno-align-labels} and @option{-falign-labels=1} are
4927 equivalent and mean that labels will not be aligned.
4929 If @option{-falign-loops} or @option{-falign-jumps} are applicable and
4930 are greater than this value, then their values are used instead.
4932 If @var{n} is not specified or is zero, use a machine-dependent default
4933 which is very likely to be @samp{1}, meaning no alignment.
4935 Enabled at levels @option{-O2}, @option{-O3}.
4938 @itemx -falign-loops=@var{n}
4939 @opindex falign-loops
4940 Align loops to a power-of-two boundary, skipping up to @var{n} bytes
4941 like @option{-falign-functions}. The hope is that the loop will be
4942 executed many times, which will make up for any execution of the dummy
4945 @option{-fno-align-loops} and @option{-falign-loops=1} are
4946 equivalent and mean that loops will not be aligned.
4948 If @var{n} is not specified or is zero, use a machine-dependent default.
4950 Enabled at levels @option{-O2}, @option{-O3}.
4953 @itemx -falign-jumps=@var{n}
4954 @opindex falign-jumps
4955 Align branch targets to a power-of-two boundary, for branch targets
4956 where the targets can only be reached by jumping, skipping up to @var{n}
4957 bytes like @option{-falign-functions}. In this case, no dummy operations
4960 @option{-fno-align-jumps} and @option{-falign-jumps=1} are
4961 equivalent and mean that loops will not be aligned.
4963 If @var{n} is not specified or is zero, use a machine-dependent default.
4965 Enabled at levels @option{-O2}, @option{-O3}.
4967 @item -funit-at-a-time
4968 @opindex funit-at-a-time
4969 Parse the whole compilation unit before starting to produce code.
4970 This allows some extra optimizations to take place but consumes
4971 more memory (in general). There are some compatibility issues
4972 with @emph{unit-at-at-time} mode:
4975 enabling @emph{unit-at-a-time} mode may change the order
4976 in which functions, variables, and top-level @code{asm} statements
4977 are emitted, and will likely break code relying on some particular
4978 ordering. The majority of such top-level @code{asm} statements,
4979 though, can be replaced by @code{section} attributes.
4982 @emph{unit-at-a-time} mode removes unreferenced static variables
4983 and functions are removed. This may result in undefined references
4984 when an @code{asm} statement refers directly to variables or functions
4985 that are otherwise unused. In that case either the variable/function
4986 shall be listed as an operand of the @code{asm} statement operand or,
4987 in the case of top-level @code{asm} statements the attribute @code{used}
4988 shall be used on the declaration.
4991 Static functions now can use non-standard passing conventions that
4992 may break @code{asm} statements calling functions directly. Again,
4993 attribute @code{used} will prevent this behavior.
4996 As a temporary workaround, @option{-fno-unit-at-a-time} can be used,
4997 but this scheme may not be supported by future releases of GCC@.
4999 Enabled at levels @option{-O2}, @option{-O3}.
5003 Constructs webs as commonly used for register allocation purposes and assign
5004 each web individual pseudo register. This allows the register allocation pass
5005 to operate on pseudos directly, but also strengthens several other optimization
5006 passes, such as CSE, loop optimizer and trivial dead code remover. It can,
5007 however, make debugging impossible, since variables will no longer stay in a
5010 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os},
5011 on targets where the default format for debugging information supports
5014 @item -fno-cprop-registers
5015 @opindex fno-cprop-registers
5016 After register allocation and post-register allocation instruction splitting,
5017 we perform a copy-propagation pass to try to reduce scheduling dependencies
5018 and occasionally eliminate the copy.
5020 Disabled at levels @option{-O}, @option{-O2}, @option{-O3}, @option{-Os}.
5022 @item -fprofile-generate
5023 @opindex fprofile-generate
5025 Enable options usually used for instrumenting application to produce
5026 profile useful for later recompilation with profile feedback based
5027 optimization. You must use @option{-fprofile-generate} both when
5028 compiling and when linking your program.
5030 The following options are enabled: @code{-fprofile-arcs}, @code{-fprofile-values}, @code{-fvpt}.
5033 @opindex fprofile-use
5034 Enable profile feedback directed optimizations, and optimizations
5035 generally profitable only with profile feedback available.
5037 The following options are enabled: @code{-fbranch-probabilities},
5038 @code{-fvpt}, @code{-funroll-loops}, @code{-fpeel-loops}, @code{-ftracer}.
5042 The following options control compiler behavior regarding floating
5043 point arithmetic. These options trade off between speed and
5044 correctness. All must be specifically enabled.
5048 @opindex ffloat-store
5049 Do not store floating point variables in registers, and inhibit other
5050 options that might change whether a floating point value is taken from a
5053 @cindex floating point precision
5054 This option prevents undesirable excess precision on machines such as
5055 the 68000 where the floating registers (of the 68881) keep more
5056 precision than a @code{double} is supposed to have. Similarly for the
5057 x86 architecture. For most programs, the excess precision does only
5058 good, but a few programs rely on the precise definition of IEEE floating
5059 point. Use @option{-ffloat-store} for such programs, after modifying
5060 them to store all pertinent intermediate computations into variables.
5064 Sets @option{-fno-math-errno}, @option{-funsafe-math-optimizations}, @*
5065 @option{-fno-trapping-math}, @option{-ffinite-math-only},
5066 @option{-fno-rounding-math} and @option{-fno-signaling-nans}.
5068 This option causes the preprocessor macro @code{__FAST_MATH__} to be defined.
5070 This option should never be turned on by any @option{-O} option since
5071 it can result in incorrect output for programs which depend on
5072 an exact implementation of IEEE or ISO rules/specifications for
5075 @item -fno-math-errno
5076 @opindex fno-math-errno
5077 Do not set ERRNO after calling math functions that are executed
5078 with a single instruction, e.g., sqrt. A program that relies on
5079 IEEE exceptions for math error handling may want to use this flag
5080 for speed while maintaining IEEE arithmetic compatibility.
5082 This option should never be turned on by any @option{-O} option since
5083 it can result in incorrect output for programs which depend on
5084 an exact implementation of IEEE or ISO rules/specifications for
5087 The default is @option{-fmath-errno}.
5089 @item -funsafe-math-optimizations
5090 @opindex funsafe-math-optimizations
5091 Allow optimizations for floating-point arithmetic that (a) assume
5092 that arguments and results are valid and (b) may violate IEEE or
5093 ANSI standards. When used at link-time, it may include libraries
5094 or startup files that change the default FPU control word or other
5095 similar optimizations.
5097 This option should never be turned on by any @option{-O} option since
5098 it can result in incorrect output for programs which depend on
5099 an exact implementation of IEEE or ISO rules/specifications for
5102 The default is @option{-fno-unsafe-math-optimizations}.
5104 @item -ffinite-math-only
5105 @opindex ffinite-math-only
5106 Allow optimizations for floating-point arithmetic that assume
5107 that arguments and results are not NaNs or +-Infs.
5109 This option should never be turned on by any @option{-O} option since
5110 it can result in incorrect output for programs which depend on
5111 an exact implementation of IEEE or ISO rules/specifications.
5113 The default is @option{-fno-finite-math-only}.
5115 @item -fno-trapping-math
5116 @opindex fno-trapping-math
5117 Compile code assuming that floating-point operations cannot generate
5118 user-visible traps. These traps include division by zero, overflow,
5119 underflow, inexact result and invalid operation. This option implies
5120 @option{-fno-signaling-nans}. Setting this option may allow faster
5121 code if one relies on ``non-stop'' IEEE arithmetic, for example.
5123 This option should never be turned on by any @option{-O} option since
5124 it can result in incorrect output for programs which depend on
5125 an exact implementation of IEEE or ISO rules/specifications for
5128 The default is @option{-ftrapping-math}.
5130 @item -frounding-math
5131 @opindex frounding-math
5132 Disable transformations and optimizations that assume default floating
5133 point rounding behavior. This is round-to-zero for all floating point
5134 to integer conversions, and round-to-nearest for all other arithmetic
5135 truncations. This option should be specified for programs that change
5136 the FP rounding mode dynamically, or that may be executed with a
5137 non-default rounding mode. This option disables constant folding of
5138 floating point expressions at compile-time (which may be affected by
5139 rounding mode) and arithmetic transformations that are unsafe in the
5140 presence of sign-dependent rounding modes.
5142 The default is @option{-fno-rounding-math}.
5144 This option is experimental and does not currently guarantee to
5145 disable all GCC optimizations that are affected by rounding mode.
5146 Future versions of GCC may provide finer control of this setting
5147 using C99's @code{FENV_ACCESS} pragma. This command line option
5148 will be used to specify the default state for @code{FENV_ACCESS}.
5150 @item -fsignaling-nans
5151 @opindex fsignaling-nans
5152 Compile code assuming that IEEE signaling NaNs may generate user-visible
5153 traps during floating-point operations. Setting this option disables
5154 optimizations that may change the number of exceptions visible with
5155 signaling NaNs. This option implies @option{-ftrapping-math}.
5157 This option causes the preprocessor macro @code{__SUPPORT_SNAN__} to
5160 The default is @option{-fno-signaling-nans}.
5162 This option is experimental and does not currently guarantee to
5163 disable all GCC optimizations that affect signaling NaN behavior.
5165 @item -fsingle-precision-constant
5166 @opindex fsingle-precision-constant
5167 Treat floating point constant as single precision constant instead of
5168 implicitly converting it to double precision constant.
5173 The following options control optimizations that may improve
5174 performance, but are not enabled by any @option{-O} options. This
5175 section includes experimental options that may produce broken code.
5178 @item -fbranch-probabilities
5179 @opindex fbranch-probabilities
5180 After running a program compiled with @option{-fprofile-arcs}
5181 (@pxref{Debugging Options,, Options for Debugging Your Program or
5182 @command{gcc}}), you can compile it a second time using
5183 @option{-fbranch-probabilities}, to improve optimizations based on
5184 the number of times each branch was taken. When the program
5185 compiled with @option{-fprofile-arcs} exits it saves arc execution
5186 counts to a file called @file{@var{sourcename}.gcda} for each source
5187 file The information in this data file is very dependent on the
5188 structure of the generated code, so you must use the same source code
5189 and the same optimization options for both compilations.
5191 With @option{-fbranch-probabilities}, GCC puts a
5192 @samp{REG_BR_PROB} note on each @samp{JUMP_INSN} and @samp{CALL_INSN}.
5193 These can be used to improve optimization. Currently, they are only
5194 used in one place: in @file{reorg.c}, instead of guessing which path a
5195 branch is mostly to take, the @samp{REG_BR_PROB} values are used to
5196 exactly determine which path is taken more often.
5198 @item -fprofile-values
5199 @opindex fprofile-values
5200 If combined with @option{-fprofile-arcs}, it adds code so that some
5201 data about values of expressions in the program is gathered.
5203 With @option{-fbranch-probabilities}, it reads back the data gathered
5204 from profiling values of expressions and adds @samp{REG_VALUE_PROFILE}
5205 notes to instructions for their later usage in optimizations.
5207 Enabled with @option{-fprofile-generate} and @option{-fprofile-use}.
5211 If combined with @option{-fprofile-arcs}, it instructs the compiler to add
5212 a code to gather information about values of expressions.
5214 With @option{-fbranch-probabilities}, it reads back the data gathered
5215 and actually performs the optimizations based on them.
5216 Currently the optimizations include specialization of division operation
5217 using the knowledge about the value of the denominator.
5219 @item -fspeculative-prefetching
5220 @opindex fspeculative-prefetching
5221 If combined with @option{-fprofile-arcs}, it instructs the compiler to add
5222 a code to gather information about addresses of memory references in the
5225 With @option{-fbranch-probabilities}, it reads back the data gathered
5226 and issues prefetch instructions according to them. In addition to the opportunities
5227 noticed by @option{-fprefetch-loop-arrays}, it also notices more complicated
5228 memory access patterns---for example accesses to the data stored in linked
5229 list whose elements are usually allocated sequentially.
5231 In order to prevent issuing double prefetches, usage of
5232 @option{-fspeculative-prefetching} implies @option{-fno-prefetch-loop-arrays}.
5234 Enabled with @option{-fprofile-generate} and @option{-fprofile-use}.
5236 @item -frename-registers
5237 @opindex frename-registers
5238 Attempt to avoid false dependencies in scheduled code by making use
5239 of registers left over after register allocation. This optimization
5240 will most benefit processors with lots of registers. Depending on the
5241 debug information format adopted by the target, however, it can
5242 make debugging impossible, since variables will no longer stay in
5243 a ``home register''.
5245 Not enabled by default at any level because it has known bugs.
5249 Perform tail duplication to enlarge superblock size. This transformation
5250 simplifies the control flow of the function allowing other optimizations to do
5253 Enabled with @option{-fprofile-use}.
5255 @item -funroll-loops
5256 @opindex funroll-loops
5257 Unroll loops whose number of iterations can be determined at compile time or
5258 upon entry to the loop. @option{-funroll-loops} implies
5259 @option{-frerun-cse-after-loop}. It also turns on complete loop peeling
5260 (i.e.@: complete removal of loops with small constant number of iterations).
5261 This option makes code larger, and may or may not make it run faster.
5263 Enabled with @option{-fprofile-use}.
5265 @item -funroll-all-loops
5266 @opindex funroll-all-loops
5267 Unroll all loops, even if their number of iterations is uncertain when
5268 the loop is entered. This usually makes programs run more slowly.
5269 @option{-funroll-all-loops} implies the same options as
5270 @option{-funroll-loops}.
5273 @opindex fpeel-loops
5274 Peels the loops for that there is enough information that they do not
5275 roll much (from profile feedback). It also turns on complete loop peeling
5276 (i.e.@: complete removal of loops with small constant number of iterations).
5278 Enabled with @option{-fprofile-use}.
5280 @item -fmove-loop-invariants
5281 @opindex fmove-loop-invariants
5282 Enables the loop invariant motion pass in the new loop optimizer. Enabled
5283 at level @option{-O1}
5285 @item -funswitch-loops
5286 @opindex funswitch-loops
5287 Move branches with loop invariant conditions out of the loop, with duplicates
5288 of the loop on both branches (modified according to result of the condition).
5290 @item -fprefetch-loop-arrays
5291 @opindex fprefetch-loop-arrays
5292 If supported by the target machine, generate instructions to prefetch
5293 memory to improve the performance of loops that access large arrays.
5295 Disabled at level @option{-Os}.
5297 @item -ffunction-sections
5298 @itemx -fdata-sections
5299 @opindex ffunction-sections
5300 @opindex fdata-sections
5301 Place each function or data item into its own section in the output
5302 file if the target supports arbitrary sections. The name of the
5303 function or the name of the data item determines the section's name
5306 Use these options on systems where the linker can perform optimizations
5307 to improve locality of reference in the instruction space. Most systems
5308 using the ELF object format and SPARC processors running Solaris 2 have
5309 linkers with such optimizations. AIX may have these optimizations in
5312 Only use these options when there are significant benefits from doing
5313 so. When you specify these options, the assembler and linker will
5314 create larger object and executable files and will also be slower.
5315 You will not be able to use @code{gprof} on all systems if you
5316 specify this option and you may have problems with debugging if
5317 you specify both this option and @option{-g}.
5319 @item -fbranch-target-load-optimize
5320 @opindex fbranch-target-load-optimize
5321 Perform branch target register load optimization before prologue / epilogue
5323 The use of target registers can typically be exposed only during reload,
5324 thus hoisting loads out of loops and doing inter-block scheduling needs
5325 a separate optimization pass.
5327 @item -fbranch-target-load-optimize2
5328 @opindex fbranch-target-load-optimize2
5329 Perform branch target register load optimization after prologue / epilogue
5332 @item -fbtr-bb-exclusive
5333 @opindex fbtr-bb-exclusive
5334 When performing branch target register load optimization, don't reuse
5335 branch target registers in within any basic block.
5337 @item --param @var{name}=@var{value}
5339 In some places, GCC uses various constants to control the amount of
5340 optimization that is done. For example, GCC will not inline functions
5341 that contain more that a certain number of instructions. You can
5342 control some of these constants on the command-line using the
5343 @option{--param} option.
5345 The names of specific parameters, and the meaning of the values, are
5346 tied to the internals of the compiler, and are subject to change
5347 without notice in future releases.
5349 In each case, the @var{value} is an integer. The allowable choices for
5350 @var{name} are given in the following table:
5353 @item sra-max-structure-size
5354 The maximum structure size, in bytes, at which the scalar replacement
5355 of aggregates (SRA) optimization will perform block copies. The
5356 default value, 0, implies that GCC will select the most appropriate
5359 @item sra-field-structure-ratio
5360 The threshold ratio (as a percentage) between instantiated fields and
5361 the complete structure size. We say that if the ratio of the number
5362 of bytes in instantiated fields to the number of bytes in the complete
5363 structure exceeds this parameter, then block copies are not used. The
5366 @item max-crossjump-edges
5367 The maximum number of incoming edges to consider for crossjumping.
5368 The algorithm used by @option{-fcrossjumping} is @math{O(N^2)} in
5369 the number of edges incoming to each block. Increasing values mean
5370 more aggressive optimization, making the compile time increase with
5371 probably small improvement in executable size.
5373 @item min-crossjump-insns
5374 The minimum number of instructions which must be matched at the end
5375 of two blocks before crossjumping will be performed on them. This
5376 value is ignored in the case where all instructions in the block being
5377 crossjumped from are matched. The default value is 5.
5379 @item max-goto-duplication-insns
5380 The maximum number of instructions to duplicate to a block that jumps
5381 to a computed goto. To avoid @math{O(N^2)} behavior in a number of
5382 passes, GCC factors computed gotos early in the compilation process,
5383 and unfactors them as late as possible. Only computed jumps at the
5384 end of a basic blocks with no more than max-goto-duplication-insns are
5385 unfactored. The default value is 8.
5387 @item max-delay-slot-insn-search
5388 The maximum number of instructions to consider when looking for an
5389 instruction to fill a delay slot. If more than this arbitrary number of
5390 instructions is searched, the time savings from filling the delay slot
5391 will be minimal so stop searching. Increasing values mean more
5392 aggressive optimization, making the compile time increase with probably
5393 small improvement in executable run time.
5395 @item max-delay-slot-live-search
5396 When trying to fill delay slots, the maximum number of instructions to
5397 consider when searching for a block with valid live register
5398 information. Increasing this arbitrarily chosen value means more
5399 aggressive optimization, increasing the compile time. This parameter
5400 should be removed when the delay slot code is rewritten to maintain the
5403 @item max-gcse-memory
5404 The approximate maximum amount of memory that will be allocated in
5405 order to perform the global common subexpression elimination
5406 optimization. If more memory than specified is required, the
5407 optimization will not be done.
5409 @item max-gcse-passes
5410 The maximum number of passes of GCSE to run. The default is 1.
5412 @item max-pending-list-length
5413 The maximum number of pending dependencies scheduling will allow
5414 before flushing the current state and starting over. Large functions
5415 with few branches or calls can create excessively large lists which
5416 needlessly consume memory and resources.
5418 @item max-inline-insns-single
5419 Several parameters control the tree inliner used in gcc.
5420 This number sets the maximum number of instructions (counted in GCC's
5421 internal representation) in a single function that the tree inliner
5422 will consider for inlining. This only affects functions declared
5423 inline and methods implemented in a class declaration (C++).
5424 The default value is 500.
5426 @item max-inline-insns-auto
5427 When you use @option{-finline-functions} (included in @option{-O3}),
5428 a lot of functions that would otherwise not be considered for inlining
5429 by the compiler will be investigated. To those functions, a different
5430 (more restrictive) limit compared to functions declared inline can
5432 The default value is 120.
5434 @item large-function-insns
5435 The limit specifying really large functions. For functions larger than this
5436 limit after inlining inlining is constrained by
5437 @option{--param large-function-growth}. This parameter is useful primarily
5438 to avoid extreme compilation time caused by non-linear algorithms used by the
5440 This parameter is ignored when @option{-funit-at-a-time} is not used.
5441 The default value is 3000.
5443 @item large-function-growth
5444 Specifies maximal growth of large function caused by inlining in percents.
5445 This parameter is ignored when @option{-funit-at-a-time} is not used.
5446 The default value is 100 which limits large function growth to 2.0 times
5449 @item inline-unit-growth
5450 Specifies maximal overall growth of the compilation unit caused by inlining.
5451 This parameter is ignored when @option{-funit-at-a-time} is not used.
5452 The default value is 50 which limits unit growth to 1.5 times the original
5455 @item max-inline-insns-recursive
5456 @itemx max-inline-insns-recursive-auto
5457 Specifies maximum number of instructions out-of-line copy of self recursive inline
5458 function can grow into by performing recursive inlining.
5460 For functions declared inline @option{--param max-inline-insns-recursive} is
5461 taken into acount. For function not declared inline, recursive inlining
5462 happens only when @option{-finline-functions} (included in @option{-O3}) is
5463 enabled and @option{--param max-inline-insns-recursive-auto} is used. The
5464 default value is 500.
5466 @item max-inline-recursive-depth
5467 @itemx max-inline-recursive-depth-auto
5468 Specifies maximum recursion depth used by the recursive inlining.
5470 For functions declared inline @option{--param max-inline-recursive-depth} is
5471 taken into acount. For function not declared inline, recursive inlining
5472 happens only when @option{-finline-functions} (included in @option{-O3}) is
5473 enabled and @option{--param max-inline-recursive-depth-auto} is used. The
5474 default value is 500.
5476 @item max-unrolled-insns
5477 The maximum number of instructions that a loop should have if that loop
5478 is unrolled, and if the loop is unrolled, it determines how many times
5479 the loop code is unrolled.
5481 @item max-average-unrolled-insns
5482 The maximum number of instructions biased by probabilities of their execution
5483 that a loop should have if that loop is unrolled, and if the loop is unrolled,
5484 it determines how many times the loop code is unrolled.
5486 @item max-unroll-times
5487 The maximum number of unrollings of a single loop.
5489 @item max-peeled-insns
5490 The maximum number of instructions that a loop should have if that loop
5491 is peeled, and if the loop is peeled, it determines how many times
5492 the loop code is peeled.
5494 @item max-peel-times
5495 The maximum number of peelings of a single loop.
5497 @item max-completely-peeled-insns
5498 The maximum number of insns of a completely peeled loop.
5500 @item max-completely-peel-times
5501 The maximum number of iterations of a loop to be suitable for complete peeling.
5503 @item max-unswitch-insns
5504 The maximum number of insns of an unswitched loop.
5506 @item max-unswitch-level
5507 The maximum number of branches unswitched in a single loop.
5510 The minimum cost of an expensive expression in the loop invariant motion.
5512 @item iv-consider-all-candidates-bound
5513 Bound on number of candidates for induction variables below that
5514 all candidates are considered for each use in induction variable
5515 optimizations. Only the most relevant candidates are considered
5516 if there are more candidates, to avoid quadratic time complexity.
5518 @item iv-max-considered-uses
5519 The induction variable optimizations give up on loops that contain more
5520 induction variable uses.
5522 @item iv-always-prune-cand-set-bound
5523 If number of candidates in the set is smaller than this value,
5524 we always try to remove unnecessary ivs from the set during its
5525 optimization when a new iv is added to the set.
5527 @item max-iterations-to-track
5529 The maximum number of iterations of a loop the brute force algorithm
5530 for analysis of # of iterations of the loop tries to evaluate.
5532 @item hot-bb-count-fraction
5533 Select fraction of the maximal count of repetitions of basic block in program
5534 given basic block needs to have to be considered hot.
5536 @item hot-bb-frequency-fraction
5537 Select fraction of the maximal frequency of executions of basic block in
5538 function given basic block needs to have to be considered hot
5540 @item tracer-dynamic-coverage
5541 @itemx tracer-dynamic-coverage-feedback
5543 This value is used to limit superblock formation once the given percentage of
5544 executed instructions is covered. This limits unnecessary code size
5547 The @option{tracer-dynamic-coverage-feedback} is used only when profile
5548 feedback is available. The real profiles (as opposed to statically estimated
5549 ones) are much less balanced allowing the threshold to be larger value.
5551 @item tracer-max-code-growth
5552 Stop tail duplication once code growth has reached given percentage. This is
5553 rather hokey argument, as most of the duplicates will be eliminated later in
5554 cross jumping, so it may be set to much higher values than is the desired code
5557 @item tracer-min-branch-ratio
5559 Stop reverse growth when the reverse probability of best edge is less than this
5560 threshold (in percent).
5562 @item tracer-min-branch-ratio
5563 @itemx tracer-min-branch-ratio-feedback
5565 Stop forward growth if the best edge do have probability lower than this
5568 Similarly to @option{tracer-dynamic-coverage} two values are present, one for
5569 compilation for profile feedback and one for compilation without. The value
5570 for compilation with profile feedback needs to be more conservative (higher) in
5571 order to make tracer effective.
5573 @item max-cse-path-length
5575 Maximum number of basic blocks on path that cse considers. The default is 10.
5577 @item global-var-threshold
5579 Counts the number of function calls (@var{n}) and the number of
5580 call-clobbered variables (@var{v}). If @var{n}x@var{v} is larger than this limit, a
5581 single artificial variable will be created to represent all the
5582 call-clobbered variables at function call sites. This artificial
5583 variable will then be made to alias every call-clobbered variable.
5584 (done as @code{int * size_t} on the host machine; beware overflow).
5586 @item max-aliased-vops
5588 Maximum number of virtual operands allowed to represent aliases
5589 before triggering the alias grouping heuristic. Alias grouping
5590 reduces compile times and memory consumption needed for aliasing at
5591 the expense of precision loss in alias information.
5593 @item ggc-min-expand
5595 GCC uses a garbage collector to manage its own memory allocation. This
5596 parameter specifies the minimum percentage by which the garbage
5597 collector's heap should be allowed to expand between collections.
5598 Tuning this may improve compilation speed; it has no effect on code
5601 The default is 30% + 70% * (RAM/1GB) with an upper bound of 100% when
5602 RAM >= 1GB@. If @code{getrlimit} is available, the notion of "RAM" is
5603 the smallest of actual RAM and @code{RLIMIT_DATA} or @code{RLIMIT_AS}. If
5604 GCC is not able to calculate RAM on a particular platform, the lower
5605 bound of 30% is used. Setting this parameter and
5606 @option{ggc-min-heapsize} to zero causes a full collection to occur at
5607 every opportunity. This is extremely slow, but can be useful for
5610 @item ggc-min-heapsize
5612 Minimum size of the garbage collector's heap before it begins bothering
5613 to collect garbage. The first collection occurs after the heap expands
5614 by @option{ggc-min-expand}% beyond @option{ggc-min-heapsize}. Again,
5615 tuning this may improve compilation speed, and has no effect on code
5618 The default is the smaller of RAM/8, RLIMIT_RSS, or a limit which
5619 tries to ensure that RLIMIT_DATA or RLIMIT_AS are not exceeded, but
5620 with a lower bound of 4096 (four megabytes) and an upper bound of
5621 131072 (128 megabytes). If GCC is not able to calculate RAM on a
5622 particular platform, the lower bound is used. Setting this parameter
5623 very large effectively disables garbage collection. Setting this
5624 parameter and @option{ggc-min-expand} to zero causes a full collection
5625 to occur at every opportunity.
5627 @item max-reload-search-insns
5628 The maximum number of instruction reload should look backward for equivalent
5629 register. Increasing values mean more aggressive optimization, making the
5630 compile time increase with probably slightly better performance. The default
5633 @item max-cselib-memory-location
5634 The maximum number of memory locations cselib should take into acount.
5635 Increasing values mean more aggressive optimization, making the compile time
5636 increase with probably slightly better performance. The default value is 500.
5638 @item reorder-blocks-duplicate
5639 @itemx reorder-blocks-duplicate-feedback
5641 Used by basic block reordering pass to decide whether to use unconditional
5642 branch or duplicate the code on its destination. Code is duplicated when its
5643 estimated size is smaller than this value multiplied by the estimated size of
5644 unconditional jump in the hot spots of the program.
5646 The @option{reorder-block-duplicate-feedback} is used only when profile
5647 feedback is available and may be set to higher values than
5648 @option{reorder-block-duplicate} since information about the hot spots is more
5651 @item max-sched-region-blocks
5652 The maximum number of blocks in a region to be considered for
5653 interblock scheduling. The default value is 10.
5655 @item max-sched-region-insns
5656 The maximum number of insns in a region to be considered for
5657 interblock scheduling. The default value is 100.
5659 @item max-last-value-rtl
5661 The maximum size measured as number of RTLs that can be recorded in an expression
5662 in combiner for a pseudo register as last known value of that register. The default
5665 @item integer-share-limit
5666 Small integer constants can use a shared data structure, reducing the
5667 compiler's memory usage and increasing its speed. This sets the maximum
5668 value of a shared integer constant's. The default value is 256.
5673 @node Preprocessor Options
5674 @section Options Controlling the Preprocessor
5675 @cindex preprocessor options
5676 @cindex options, preprocessor
5678 These options control the C preprocessor, which is run on each C source
5679 file before actual compilation.
5681 If you use the @option{-E} option, nothing is done except preprocessing.
5682 Some of these options make sense only together with @option{-E} because
5683 they cause the preprocessor output to be unsuitable for actual
5688 You can use @option{-Wp,@var{option}} to bypass the compiler driver
5689 and pass @var{option} directly through to the preprocessor. If
5690 @var{option} contains commas, it is split into multiple options at the
5691 commas. However, many options are modified, translated or interpreted
5692 by the compiler driver before being passed to the preprocessor, and
5693 @option{-Wp} forcibly bypasses this phase. The preprocessor's direct
5694 interface is undocumented and subject to change, so whenever possible
5695 you should avoid using @option{-Wp} and let the driver handle the
5698 @item -Xpreprocessor @var{option}
5699 @opindex preprocessor
5700 Pass @var{option} as an option to the preprocessor. You can use this to
5701 supply system-specific preprocessor options which GCC does not know how to
5704 If you want to pass an option that takes an argument, you must use
5705 @option{-Xpreprocessor} twice, once for the option and once for the argument.
5708 @include cppopts.texi
5710 @node Assembler Options
5711 @section Passing Options to the Assembler
5713 @c prevent bad page break with this line
5714 You can pass options to the assembler.
5717 @item -Wa,@var{option}
5719 Pass @var{option} as an option to the assembler. If @var{option}
5720 contains commas, it is split into multiple options at the commas.
5722 @item -Xassembler @var{option}
5724 Pass @var{option} as an option to the assembler. You can use this to
5725 supply system-specific assembler options which GCC does not know how to
5728 If you want to pass an option that takes an argument, you must use
5729 @option{-Xassembler} twice, once for the option and once for the argument.
5734 @section Options for Linking
5735 @cindex link options
5736 @cindex options, linking
5738 These options come into play when the compiler links object files into
5739 an executable output file. They are meaningless if the compiler is
5740 not doing a link step.
5744 @item @var{object-file-name}
5745 A file name that does not end in a special recognized suffix is
5746 considered to name an object file or library. (Object files are
5747 distinguished from libraries by the linker according to the file
5748 contents.) If linking is done, these object files are used as input
5757 If any of these options is used, then the linker is not run, and
5758 object file names should not be used as arguments. @xref{Overall
5762 @item -l@var{library}
5763 @itemx -l @var{library}
5765 Search the library named @var{library} when linking. (The second
5766 alternative with the library as a separate argument is only for
5767 POSIX compliance and is not recommended.)
5769 It makes a difference where in the command you write this option; the
5770 linker searches and processes libraries and object files in the order they
5771 are specified. Thus, @samp{foo.o -lz bar.o} searches library @samp{z}
5772 after file @file{foo.o} but before @file{bar.o}. If @file{bar.o} refers
5773 to functions in @samp{z}, those functions may not be loaded.
5775 The linker searches a standard list of directories for the library,
5776 which is actually a file named @file{lib@var{library}.a}. The linker
5777 then uses this file as if it had been specified precisely by name.
5779 The directories searched include several standard system directories
5780 plus any that you specify with @option{-L}.
5782 Normally the files found this way are library files---archive files
5783 whose members are object files. The linker handles an archive file by
5784 scanning through it for members which define symbols that have so far
5785 been referenced but not defined. But if the file that is found is an
5786 ordinary object file, it is linked in the usual fashion. The only
5787 difference between using an @option{-l} option and specifying a file name
5788 is that @option{-l} surrounds @var{library} with @samp{lib} and @samp{.a}
5789 and searches several directories.
5793 You need this special case of the @option{-l} option in order to
5794 link an Objective-C or Objective-C++ program.
5797 @opindex nostartfiles
5798 Do not use the standard system startup files when linking.
5799 The standard system libraries are used normally, unless @option{-nostdlib}
5800 or @option{-nodefaultlibs} is used.
5802 @item -nodefaultlibs
5803 @opindex nodefaultlibs
5804 Do not use the standard system libraries when linking.
5805 Only the libraries you specify will be passed to the linker.
5806 The standard startup files are used normally, unless @option{-nostartfiles}
5807 is used. The compiler may generate calls to @code{memcmp},
5808 @code{memset}, @code{memcpy} and @code{memmove}.
5809 These entries are usually resolved by entries in
5810 libc. These entry points should be supplied through some other
5811 mechanism when this option is specified.
5815 Do not use the standard system startup files or libraries when linking.
5816 No startup files and only the libraries you specify will be passed to
5817 the linker. The compiler may generate calls to @code{memcmp}, @code{memset},
5818 @code{memcpy} and @code{memmove}.
5819 These entries are usually resolved by entries in
5820 libc. These entry points should be supplied through some other
5821 mechanism when this option is specified.
5823 @cindex @option{-lgcc}, use with @option{-nostdlib}
5824 @cindex @option{-nostdlib} and unresolved references
5825 @cindex unresolved references and @option{-nostdlib}
5826 @cindex @option{-lgcc}, use with @option{-nodefaultlibs}
5827 @cindex @option{-nodefaultlibs} and unresolved references
5828 @cindex unresolved references and @option{-nodefaultlibs}
5829 One of the standard libraries bypassed by @option{-nostdlib} and
5830 @option{-nodefaultlibs} is @file{libgcc.a}, a library of internal subroutines
5831 that GCC uses to overcome shortcomings of particular machines, or special
5832 needs for some languages.
5833 (@xref{Interface,,Interfacing to GCC Output,gccint,GNU Compiler
5834 Collection (GCC) Internals},
5835 for more discussion of @file{libgcc.a}.)
5836 In most cases, you need @file{libgcc.a} even when you want to avoid
5837 other standard libraries. In other words, when you specify @option{-nostdlib}
5838 or @option{-nodefaultlibs} you should usually specify @option{-lgcc} as well.
5839 This ensures that you have no unresolved references to internal GCC
5840 library subroutines. (For example, @samp{__main}, used to ensure C++
5841 constructors will be called; @pxref{Collect2,,@code{collect2}, gccint,
5842 GNU Compiler Collection (GCC) Internals}.)
5846 Produce a position independent executable on targets which support it.
5847 For predictable results, you must also specify the same set of options
5848 that were used to generate code (@option{-fpie}, @option{-fPIE},
5849 or model suboptions) when you specify this option.
5853 Remove all symbol table and relocation information from the executable.
5857 On systems that support dynamic linking, this prevents linking with the shared
5858 libraries. On other systems, this option has no effect.
5862 Produce a shared object which can then be linked with other objects to
5863 form an executable. Not all systems support this option. For predictable
5864 results, you must also specify the same set of options that were used to
5865 generate code (@option{-fpic}, @option{-fPIC}, or model suboptions)
5866 when you specify this option.@footnote{On some systems, @samp{gcc -shared}
5867 needs to build supplementary stub code for constructors to work. On
5868 multi-libbed systems, @samp{gcc -shared} must select the correct support
5869 libraries to link against. Failing to supply the correct flags may lead
5870 to subtle defects. Supplying them in cases where they are not necessary
5873 @item -shared-libgcc
5874 @itemx -static-libgcc
5875 @opindex shared-libgcc
5876 @opindex static-libgcc
5877 On systems that provide @file{libgcc} as a shared library, these options
5878 force the use of either the shared or static version respectively.
5879 If no shared version of @file{libgcc} was built when the compiler was
5880 configured, these options have no effect.
5882 There are several situations in which an application should use the
5883 shared @file{libgcc} instead of the static version. The most common
5884 of these is when the application wishes to throw and catch exceptions
5885 across different shared libraries. In that case, each of the libraries
5886 as well as the application itself should use the shared @file{libgcc}.
5888 Therefore, the G++ and GCJ drivers automatically add
5889 @option{-shared-libgcc} whenever you build a shared library or a main
5890 executable, because C++ and Java programs typically use exceptions, so
5891 this is the right thing to do.
5893 If, instead, you use the GCC driver to create shared libraries, you may
5894 find that they will not always be linked with the shared @file{libgcc}.
5895 If GCC finds, at its configuration time, that you have a non-GNU linker
5896 or a GNU linker that does not support option @option{--eh-frame-hdr},
5897 it will link the shared version of @file{libgcc} into shared libraries
5898 by default. Otherwise, it will take advantage of the linker and optimize
5899 away the linking with the shared version of @file{libgcc}, linking with
5900 the static version of libgcc by default. This allows exceptions to
5901 propagate through such shared libraries, without incurring relocation
5902 costs at library load time.
5904 However, if a library or main executable is supposed to throw or catch
5905 exceptions, you must link it using the G++ or GCJ driver, as appropriate
5906 for the languages used in the program, or using the option
5907 @option{-shared-libgcc}, such that it is linked with the shared
5912 Bind references to global symbols when building a shared object. Warn
5913 about any unresolved references (unless overridden by the link editor
5914 option @samp{-Xlinker -z -Xlinker defs}). Only a few systems support
5917 @item -Xlinker @var{option}
5919 Pass @var{option} as an option to the linker. You can use this to
5920 supply system-specific linker options which GCC does not know how to
5923 If you want to pass an option that takes an argument, you must use
5924 @option{-Xlinker} twice, once for the option and once for the argument.
5925 For example, to pass @option{-assert definitions}, you must write
5926 @samp{-Xlinker -assert -Xlinker definitions}. It does not work to write
5927 @option{-Xlinker "-assert definitions"}, because this passes the entire
5928 string as a single argument, which is not what the linker expects.
5930 @item -Wl,@var{option}
5932 Pass @var{option} as an option to the linker. If @var{option} contains
5933 commas, it is split into multiple options at the commas.
5935 @item -u @var{symbol}
5937 Pretend the symbol @var{symbol} is undefined, to force linking of
5938 library modules to define it. You can use @option{-u} multiple times with
5939 different symbols to force loading of additional library modules.
5942 @node Directory Options
5943 @section Options for Directory Search
5944 @cindex directory options
5945 @cindex options, directory search
5948 These options specify directories to search for header files, for
5949 libraries and for parts of the compiler:
5954 Add the directory @var{dir} to the head of the list of directories to be
5955 searched for header files. This can be used to override a system header
5956 file, substituting your own version, since these directories are
5957 searched before the system header file directories. However, you should
5958 not use this option to add directories that contain vendor-supplied
5959 system header files (use @option{-isystem} for that). If you use more than
5960 one @option{-I} option, the directories are scanned in left-to-right
5961 order; the standard system directories come after.
5963 If a standard system include directory, or a directory specified with
5964 @option{-isystem}, is also specified with @option{-I}, the @option{-I}
5965 option will be ignored. The directory will still be searched but as a
5966 system directory at its normal position in the system include chain.
5967 This is to ensure that GCC's procedure to fix buggy system headers and
5968 the ordering for the include_next directive are not inadvertently changed.
5969 If you really need to change the search order for system directories,
5970 use the @option{-nostdinc} and/or @option{-isystem} options.
5972 @item -iquote@var{dir}
5974 Add the directory @var{dir} to the head of the list of directories to
5975 be searched for header files only for the case of @samp{#include
5976 "@var{file}"}; they are not searched for @samp{#include <@var{file}>},
5977 otherwise just like @option{-I}.
5981 Add directory @var{dir} to the list of directories to be searched
5984 @item -B@var{prefix}
5986 This option specifies where to find the executables, libraries,
5987 include files, and data files of the compiler itself.
5989 The compiler driver program runs one or more of the subprograms
5990 @file{cpp}, @file{cc1}, @file{as} and @file{ld}. It tries
5991 @var{prefix} as a prefix for each program it tries to run, both with and
5992 without @samp{@var{machine}/@var{version}/} (@pxref{Target Options}).
5994 For each subprogram to be run, the compiler driver first tries the
5995 @option{-B} prefix, if any. If that name is not found, or if @option{-B}
5996 was not specified, the driver tries two standard prefixes, which are
5997 @file{/usr/lib/gcc/} and @file{/usr/local/lib/gcc/}. If neither of
5998 those results in a file name that is found, the unmodified program
5999 name is searched for using the directories specified in your
6000 @env{PATH} environment variable.
6002 The compiler will check to see if the path provided by the @option{-B}
6003 refers to a directory, and if necessary it will add a directory
6004 separator character at the end of the path.
6006 @option{-B} prefixes that effectively specify directory names also apply
6007 to libraries in the linker, because the compiler translates these
6008 options into @option{-L} options for the linker. They also apply to
6009 includes files in the preprocessor, because the compiler translates these
6010 options into @option{-isystem} options for the preprocessor. In this case,
6011 the compiler appends @samp{include} to the prefix.
6013 The run-time support file @file{libgcc.a} can also be searched for using
6014 the @option{-B} prefix, if needed. If it is not found there, the two
6015 standard prefixes above are tried, and that is all. The file is left
6016 out of the link if it is not found by those means.
6018 Another way to specify a prefix much like the @option{-B} prefix is to use
6019 the environment variable @env{GCC_EXEC_PREFIX}. @xref{Environment
6022 As a special kludge, if the path provided by @option{-B} is
6023 @file{[dir/]stage@var{N}/}, where @var{N} is a number in the range 0 to
6024 9, then it will be replaced by @file{[dir/]include}. This is to help
6025 with boot-strapping the compiler.
6027 @item -specs=@var{file}
6029 Process @var{file} after the compiler reads in the standard @file{specs}
6030 file, in order to override the defaults that the @file{gcc} driver
6031 program uses when determining what switches to pass to @file{cc1},
6032 @file{cc1plus}, @file{as}, @file{ld}, etc. More than one
6033 @option{-specs=@var{file}} can be specified on the command line, and they
6034 are processed in order, from left to right.
6038 This option has been deprecated. Please use @option{-iquote} instead for
6039 @option{-I} directories before the @option{-I-} and remove the @option{-I-}.
6040 Any directories you specify with @option{-I} options before the @option{-I-}
6041 option are searched only for the case of @samp{#include "@var{file}"};
6042 they are not searched for @samp{#include <@var{file}>}.
6044 If additional directories are specified with @option{-I} options after
6045 the @option{-I-}, these directories are searched for all @samp{#include}
6046 directives. (Ordinarily @emph{all} @option{-I} directories are used
6049 In addition, the @option{-I-} option inhibits the use of the current
6050 directory (where the current input file came from) as the first search
6051 directory for @samp{#include "@var{file}"}. There is no way to
6052 override this effect of @option{-I-}. With @option{-I.} you can specify
6053 searching the directory which was current when the compiler was
6054 invoked. That is not exactly the same as what the preprocessor does
6055 by default, but it is often satisfactory.
6057 @option{-I-} does not inhibit the use of the standard system directories
6058 for header files. Thus, @option{-I-} and @option{-nostdinc} are
6065 @section Specifying subprocesses and the switches to pass to them
6068 @command{gcc} is a driver program. It performs its job by invoking a
6069 sequence of other programs to do the work of compiling, assembling and
6070 linking. GCC interprets its command-line parameters and uses these to
6071 deduce which programs it should invoke, and which command-line options
6072 it ought to place on their command lines. This behavior is controlled
6073 by @dfn{spec strings}. In most cases there is one spec string for each
6074 program that GCC can invoke, but a few programs have multiple spec
6075 strings to control their behavior. The spec strings built into GCC can
6076 be overridden by using the @option{-specs=} command-line switch to specify
6079 @dfn{Spec files} are plaintext files that are used to construct spec
6080 strings. They consist of a sequence of directives separated by blank
6081 lines. The type of directive is determined by the first non-whitespace
6082 character on the line and it can be one of the following:
6085 @item %@var{command}
6086 Issues a @var{command} to the spec file processor. The commands that can
6090 @item %include <@var{file}>
6092 Search for @var{file} and insert its text at the current point in the
6095 @item %include_noerr <@var{file}>
6096 @cindex %include_noerr
6097 Just like @samp{%include}, but do not generate an error message if the include
6098 file cannot be found.
6100 @item %rename @var{old_name} @var{new_name}
6102 Rename the spec string @var{old_name} to @var{new_name}.
6106 @item *[@var{spec_name}]:
6107 This tells the compiler to create, override or delete the named spec
6108 string. All lines after this directive up to the next directive or
6109 blank line are considered to be the text for the spec string. If this
6110 results in an empty string then the spec will be deleted. (Or, if the
6111 spec did not exist, then nothing will happened.) Otherwise, if the spec
6112 does not currently exist a new spec will be created. If the spec does
6113 exist then its contents will be overridden by the text of this
6114 directive, unless the first character of that text is the @samp{+}
6115 character, in which case the text will be appended to the spec.
6117 @item [@var{suffix}]:
6118 Creates a new @samp{[@var{suffix}] spec} pair. All lines after this directive
6119 and up to the next directive or blank line are considered to make up the
6120 spec string for the indicated suffix. When the compiler encounters an
6121 input file with the named suffix, it will processes the spec string in
6122 order to work out how to compile that file. For example:
6129 This says that any input file whose name ends in @samp{.ZZ} should be
6130 passed to the program @samp{z-compile}, which should be invoked with the
6131 command-line switch @option{-input} and with the result of performing the
6132 @samp{%i} substitution. (See below.)
6134 As an alternative to providing a spec string, the text that follows a
6135 suffix directive can be one of the following:
6138 @item @@@var{language}
6139 This says that the suffix is an alias for a known @var{language}. This is
6140 similar to using the @option{-x} command-line switch to GCC to specify a
6141 language explicitly. For example:
6148 Says that .ZZ files are, in fact, C++ source files.
6151 This causes an error messages saying:
6154 @var{name} compiler not installed on this system.
6158 GCC already has an extensive list of suffixes built into it.
6159 This directive will add an entry to the end of the list of suffixes, but
6160 since the list is searched from the end backwards, it is effectively
6161 possible to override earlier entries using this technique.
6165 GCC has the following spec strings built into it. Spec files can
6166 override these strings or create their own. Note that individual
6167 targets can also add their own spec strings to this list.
6170 asm Options to pass to the assembler
6171 asm_final Options to pass to the assembler post-processor
6172 cpp Options to pass to the C preprocessor
6173 cc1 Options to pass to the C compiler
6174 cc1plus Options to pass to the C++ compiler
6175 endfile Object files to include at the end of the link
6176 link Options to pass to the linker
6177 lib Libraries to include on the command line to the linker
6178 libgcc Decides which GCC support library to pass to the linker
6179 linker Sets the name of the linker
6180 predefines Defines to be passed to the C preprocessor
6181 signed_char Defines to pass to CPP to say whether @code{char} is signed
6183 startfile Object files to include at the start of the link
6186 Here is a small example of a spec file:
6192 --start-group -lgcc -lc -leval1 --end-group %(old_lib)
6195 This example renames the spec called @samp{lib} to @samp{old_lib} and
6196 then overrides the previous definition of @samp{lib} with a new one.
6197 The new definition adds in some extra command-line options before
6198 including the text of the old definition.
6200 @dfn{Spec strings} are a list of command-line options to be passed to their
6201 corresponding program. In addition, the spec strings can contain
6202 @samp{%}-prefixed sequences to substitute variable text or to
6203 conditionally insert text into the command line. Using these constructs
6204 it is possible to generate quite complex command lines.
6206 Here is a table of all defined @samp{%}-sequences for spec
6207 strings. Note that spaces are not generated automatically around the
6208 results of expanding these sequences. Therefore you can concatenate them
6209 together or combine them with constant text in a single argument.
6213 Substitute one @samp{%} into the program name or argument.
6216 Substitute the name of the input file being processed.
6219 Substitute the basename of the input file being processed.
6220 This is the substring up to (and not including) the last period
6221 and not including the directory.
6224 This is the same as @samp{%b}, but include the file suffix (text after
6228 Marks the argument containing or following the @samp{%d} as a
6229 temporary file name, so that that file will be deleted if GCC exits
6230 successfully. Unlike @samp{%g}, this contributes no text to the
6233 @item %g@var{suffix}
6234 Substitute a file name that has suffix @var{suffix} and is chosen
6235 once per compilation, and mark the argument in the same way as
6236 @samp{%d}. To reduce exposure to denial-of-service attacks, the file
6237 name is now chosen in a way that is hard to predict even when previously
6238 chosen file names are known. For example, @samp{%g.s @dots{} %g.o @dots{} %g.s}
6239 might turn into @samp{ccUVUUAU.s ccXYAXZ12.o ccUVUUAU.s}. @var{suffix} matches
6240 the regexp @samp{[.A-Za-z]*} or the special string @samp{%O}, which is
6241 treated exactly as if @samp{%O} had been preprocessed. Previously, @samp{%g}
6242 was simply substituted with a file name chosen once per compilation,
6243 without regard to any appended suffix (which was therefore treated
6244 just like ordinary text), making such attacks more likely to succeed.
6246 @item %u@var{suffix}
6247 Like @samp{%g}, but generates a new temporary file name even if
6248 @samp{%u@var{suffix}} was already seen.
6250 @item %U@var{suffix}
6251 Substitutes the last file name generated with @samp{%u@var{suffix}}, generating a
6252 new one if there is no such last file name. In the absence of any
6253 @samp{%u@var{suffix}}, this is just like @samp{%g@var{suffix}}, except they don't share
6254 the same suffix @emph{space}, so @samp{%g.s @dots{} %U.s @dots{} %g.s @dots{} %U.s}
6255 would involve the generation of two distinct file names, one
6256 for each @samp{%g.s} and another for each @samp{%U.s}. Previously, @samp{%U} was
6257 simply substituted with a file name chosen for the previous @samp{%u},
6258 without regard to any appended suffix.
6260 @item %j@var{suffix}
6261 Substitutes the name of the @code{HOST_BIT_BUCKET}, if any, and if it is
6262 writable, and if save-temps is off; otherwise, substitute the name
6263 of a temporary file, just like @samp{%u}. This temporary file is not
6264 meant for communication between processes, but rather as a junk
6267 @item %|@var{suffix}
6268 @itemx %m@var{suffix}
6269 Like @samp{%g}, except if @option{-pipe} is in effect. In that case
6270 @samp{%|} substitutes a single dash and @samp{%m} substitutes nothing at
6271 all. These are the two most common ways to instruct a program that it
6272 should read from standard input or write to standard output. If you
6273 need something more elaborate you can use an @samp{%@{pipe:@code{X}@}}
6274 construct: see for example @file{f/lang-specs.h}.
6276 @item %.@var{SUFFIX}
6277 Substitutes @var{.SUFFIX} for the suffixes of a matched switch's args
6278 when it is subsequently output with @samp{%*}. @var{SUFFIX} is
6279 terminated by the next space or %.
6282 Marks the argument containing or following the @samp{%w} as the
6283 designated output file of this compilation. This puts the argument
6284 into the sequence of arguments that @samp{%o} will substitute later.
6287 Substitutes the names of all the output files, with spaces
6288 automatically placed around them. You should write spaces
6289 around the @samp{%o} as well or the results are undefined.
6290 @samp{%o} is for use in the specs for running the linker.
6291 Input files whose names have no recognized suffix are not compiled
6292 at all, but they are included among the output files, so they will
6296 Substitutes the suffix for object files. Note that this is
6297 handled specially when it immediately follows @samp{%g, %u, or %U},
6298 because of the need for those to form complete file names. The
6299 handling is such that @samp{%O} is treated exactly as if it had already
6300 been substituted, except that @samp{%g, %u, and %U} do not currently
6301 support additional @var{suffix} characters following @samp{%O} as they would
6302 following, for example, @samp{.o}.
6305 Substitutes the standard macro predefinitions for the
6306 current target machine. Use this when running @code{cpp}.
6309 Like @samp{%p}, but puts @samp{__} before and after the name of each
6310 predefined macro, except for macros that start with @samp{__} or with
6311 @samp{_@var{L}}, where @var{L} is an uppercase letter. This is for ISO
6315 Substitute any of @option{-iprefix} (made from @env{GCC_EXEC_PREFIX}),
6316 @option{-isysroot} (made from @env{TARGET_SYSTEM_ROOT}), and
6317 @option{-isystem} (made from @env{COMPILER_PATH} and @option{-B} options)
6321 Current argument is the name of a library or startup file of some sort.
6322 Search for that file in a standard list of directories and substitute
6323 the full name found.
6326 Print @var{str} as an error message. @var{str} is terminated by a newline.
6327 Use this when inconsistent options are detected.
6330 Substitute the contents of spec string @var{name} at this point.
6333 Like @samp{%(@dots{})} but put @samp{__} around @option{-D} arguments.
6335 @item %x@{@var{option}@}
6336 Accumulate an option for @samp{%X}.
6339 Output the accumulated linker options specified by @option{-Wl} or a @samp{%x}
6343 Output the accumulated assembler options specified by @option{-Wa}.
6346 Output the accumulated preprocessor options specified by @option{-Wp}.
6349 Process the @code{asm} spec. This is used to compute the
6350 switches to be passed to the assembler.
6353 Process the @code{asm_final} spec. This is a spec string for
6354 passing switches to an assembler post-processor, if such a program is
6358 Process the @code{link} spec. This is the spec for computing the
6359 command line passed to the linker. Typically it will make use of the
6360 @samp{%L %G %S %D and %E} sequences.
6363 Dump out a @option{-L} option for each directory that GCC believes might
6364 contain startup files. If the target supports multilibs then the
6365 current multilib directory will be prepended to each of these paths.
6368 Output the multilib directory with directory separators replaced with
6369 @samp{_}. If multilib directories are not set, or the multilib directory is
6370 @file{.} then this option emits nothing.
6373 Process the @code{lib} spec. This is a spec string for deciding which
6374 libraries should be included on the command line to the linker.
6377 Process the @code{libgcc} spec. This is a spec string for deciding
6378 which GCC support library should be included on the command line to the linker.
6381 Process the @code{startfile} spec. This is a spec for deciding which
6382 object files should be the first ones passed to the linker. Typically
6383 this might be a file named @file{crt0.o}.
6386 Process the @code{endfile} spec. This is a spec string that specifies
6387 the last object files that will be passed to the linker.
6390 Process the @code{cpp} spec. This is used to construct the arguments
6391 to be passed to the C preprocessor.
6394 Process the @code{cc1} spec. This is used to construct the options to be
6395 passed to the actual C compiler (@samp{cc1}).
6398 Process the @code{cc1plus} spec. This is used to construct the options to be
6399 passed to the actual C++ compiler (@samp{cc1plus}).
6402 Substitute the variable part of a matched option. See below.
6403 Note that each comma in the substituted string is replaced by
6407 Remove all occurrences of @code{-S} from the command line. Note---this
6408 command is position dependent. @samp{%} commands in the spec string
6409 before this one will see @code{-S}, @samp{%} commands in the spec string
6410 after this one will not.
6412 @item %:@var{function}(@var{args})
6413 Call the named function @var{function}, passing it @var{args}.
6414 @var{args} is first processed as a nested spec string, then split
6415 into an argument vector in the usual fashion. The function returns
6416 a string which is processed as if it had appeared literally as part
6417 of the current spec.
6419 The following built-in spec functions are provided:
6422 @item @code{if-exists}
6423 The @code{if-exists} spec function takes one argument, an absolute
6424 pathname to a file. If the file exists, @code{if-exists} returns the
6425 pathname. Here is a small example of its usage:
6429 crt0%O%s %:if-exists(crti%O%s) crtbegin%O%s
6432 @item @code{if-exists-else}
6433 The @code{if-exists-else} spec function is similar to the @code{if-exists}
6434 spec function, except that it takes two arguments. The first argument is
6435 an absolute pathname to a file. If the file exists, @code{if-exists-else}
6436 returns the pathname. If it does not exist, it returns the second argument.
6437 This way, @code{if-exists-else} can be used to select one file or another,
6438 based on the existence of the first. Here is a small example of its usage:
6442 crt0%O%s %:if-exists(crti%O%s) \
6443 %:if-exists-else(crtbeginT%O%s crtbegin%O%s)
6446 @item @code{replace-outfile}
6447 The @code{replace-outfile} spec function takes two arguments. It looks for the
6448 first argument in the outfiles array and replaces it with the second argument. Here
6449 is a small example of its usage:
6452 %@{fgnu-runtime:%:replace-outfile(-lobjc -lobjc-gnu)@}
6458 Substitutes the @code{-S} switch, if that switch was given to GCC@.
6459 If that switch was not specified, this substitutes nothing. Note that
6460 the leading dash is omitted when specifying this option, and it is
6461 automatically inserted if the substitution is performed. Thus the spec
6462 string @samp{%@{foo@}} would match the command-line option @option{-foo}
6463 and would output the command line option @option{-foo}.
6465 @item %W@{@code{S}@}
6466 Like %@{@code{S}@} but mark last argument supplied within as a file to be
6469 @item %@{@code{S}*@}
6470 Substitutes all the switches specified to GCC whose names start
6471 with @code{-S}, but which also take an argument. This is used for
6472 switches like @option{-o}, @option{-D}, @option{-I}, etc.
6473 GCC considers @option{-o foo} as being
6474 one switch whose names starts with @samp{o}. %@{o*@} would substitute this
6475 text, including the space. Thus two arguments would be generated.
6477 @item %@{@code{S}*&@code{T}*@}
6478 Like %@{@code{S}*@}, but preserve order of @code{S} and @code{T} options
6479 (the order of @code{S} and @code{T} in the spec is not significant).
6480 There can be any number of ampersand-separated variables; for each the
6481 wild card is optional. Useful for CPP as @samp{%@{D*&U*&A*@}}.
6483 @item %@{@code{S}:@code{X}@}
6484 Substitutes @code{X}, if the @samp{-S} switch was given to GCC@.
6486 @item %@{!@code{S}:@code{X}@}
6487 Substitutes @code{X}, if the @samp{-S} switch was @emph{not} given to GCC@.
6489 @item %@{@code{S}*:@code{X}@}
6490 Substitutes @code{X} if one or more switches whose names start with
6491 @code{-S} are specified to GCC@. Normally @code{X} is substituted only
6492 once, no matter how many such switches appeared. However, if @code{%*}
6493 appears somewhere in @code{X}, then @code{X} will be substituted once
6494 for each matching switch, with the @code{%*} replaced by the part of
6495 that switch that matched the @code{*}.
6497 @item %@{.@code{S}:@code{X}@}
6498 Substitutes @code{X}, if processing a file with suffix @code{S}.
6500 @item %@{!.@code{S}:@code{X}@}
6501 Substitutes @code{X}, if @emph{not} processing a file with suffix @code{S}.
6503 @item %@{@code{S}|@code{P}:@code{X}@}
6504 Substitutes @code{X} if either @code{-S} or @code{-P} was given to GCC@.
6505 This may be combined with @samp{!}, @samp{.}, and @code{*} sequences as well,
6506 although they have a stronger binding than the @samp{|}. If @code{%*}
6507 appears in @code{X}, all of the alternatives must be starred, and only
6508 the first matching alternative is substituted.
6510 For example, a spec string like this:
6513 %@{.c:-foo@} %@{!.c:-bar@} %@{.c|d:-baz@} %@{!.c|d:-boggle@}
6516 will output the following command-line options from the following input
6517 command-line options:
6522 -d fred.c -foo -baz -boggle
6523 -d jim.d -bar -baz -boggle
6526 @item %@{S:X; T:Y; :D@}
6528 If @code{S} was given to GCC, substitutes @code{X}; else if @code{T} was
6529 given to GCC, substitutes @code{Y}; else substitutes @code{D}. There can
6530 be as many clauses as you need. This may be combined with @code{.},
6531 @code{!}, @code{|}, and @code{*} as needed.
6536 The conditional text @code{X} in a %@{@code{S}:@code{X}@} or similar
6537 construct may contain other nested @samp{%} constructs or spaces, or
6538 even newlines. They are processed as usual, as described above.
6539 Trailing white space in @code{X} is ignored. White space may also
6540 appear anywhere on the left side of the colon in these constructs,
6541 except between @code{.} or @code{*} and the corresponding word.
6543 The @option{-O}, @option{-f}, @option{-m}, and @option{-W} switches are
6544 handled specifically in these constructs. If another value of
6545 @option{-O} or the negated form of a @option{-f}, @option{-m}, or
6546 @option{-W} switch is found later in the command line, the earlier
6547 switch value is ignored, except with @{@code{S}*@} where @code{S} is
6548 just one letter, which passes all matching options.
6550 The character @samp{|} at the beginning of the predicate text is used to
6551 indicate that a command should be piped to the following command, but
6552 only if @option{-pipe} is specified.
6554 It is built into GCC which switches take arguments and which do not.
6555 (You might think it would be useful to generalize this to allow each
6556 compiler's spec to say which switches take arguments. But this cannot
6557 be done in a consistent fashion. GCC cannot even decide which input
6558 files have been specified without knowing which switches take arguments,
6559 and it must know which input files to compile in order to tell which
6562 GCC also knows implicitly that arguments starting in @option{-l} are to be
6563 treated as compiler output files, and passed to the linker in their
6564 proper position among the other output files.
6566 @c man begin OPTIONS
6568 @node Target Options
6569 @section Specifying Target Machine and Compiler Version
6570 @cindex target options
6571 @cindex cross compiling
6572 @cindex specifying machine version
6573 @cindex specifying compiler version and target machine
6574 @cindex compiler version, specifying
6575 @cindex target machine, specifying
6577 The usual way to run GCC is to run the executable called @file{gcc}, or
6578 @file{<machine>-gcc} when cross-compiling, or
6579 @file{<machine>-gcc-<version>} to run a version other than the one that
6580 was installed last. Sometimes this is inconvenient, so GCC provides
6581 options that will switch to another cross-compiler or version.
6584 @item -b @var{machine}
6586 The argument @var{machine} specifies the target machine for compilation.
6588 The value to use for @var{machine} is the same as was specified as the
6589 machine type when configuring GCC as a cross-compiler. For
6590 example, if a cross-compiler was configured with @samp{configure
6591 i386v}, meaning to compile for an 80386 running System V, then you
6592 would specify @option{-b i386v} to run that cross compiler.
6594 @item -V @var{version}
6596 The argument @var{version} specifies which version of GCC to run.
6597 This is useful when multiple versions are installed. For example,
6598 @var{version} might be @samp{2.0}, meaning to run GCC version 2.0.
6601 The @option{-V} and @option{-b} options work by running the
6602 @file{<machine>-gcc-<version>} executable, so there's no real reason to
6603 use them if you can just run that directly.
6605 @node Submodel Options
6606 @section Hardware Models and Configurations
6607 @cindex submodel options
6608 @cindex specifying hardware config
6609 @cindex hardware models and configurations, specifying
6610 @cindex machine dependent options
6612 Earlier we discussed the standard option @option{-b} which chooses among
6613 different installed compilers for completely different target
6614 machines, such as VAX vs.@: 68000 vs.@: 80386.
6616 In addition, each of these target machine types can have its own
6617 special options, starting with @samp{-m}, to choose among various
6618 hardware models or configurations---for example, 68010 vs 68020,
6619 floating coprocessor or none. A single installed version of the
6620 compiler can compile for any model or configuration, according to the
6623 Some configurations of the compiler also support additional special
6624 options, usually for compatibility with other compilers on the same
6627 These options are defined by the macro @code{TARGET_SWITCHES} in the
6628 machine description. The default for the options is also defined by
6629 that macro, which enables you to change the defaults.
6631 @c This list is ordered alphanumerically by subsection name.
6632 @c It should be the same order and spelling as these options are listed
6633 @c in Machine Dependent Options
6641 * DEC Alpha Options::
6642 * DEC Alpha/VMS Options::
6646 * i386 and x86-64 Options::
6658 * RS/6000 and PowerPC Options::
6659 * S/390 and zSeries Options::
6662 * System V Options::
6663 * TMS320C3x/C4x Options::
6667 * Xstormy16 Options::
6673 @subsection ARC Options
6676 These options are defined for ARC implementations:
6681 Compile code for little endian mode. This is the default.
6685 Compile code for big endian mode.
6688 @opindex mmangle-cpu
6689 Prepend the name of the cpu to all public symbol names.
6690 In multiple-processor systems, there are many ARC variants with different
6691 instruction and register set characteristics. This flag prevents code
6692 compiled for one cpu to be linked with code compiled for another.
6693 No facility exists for handling variants that are ``almost identical''.
6694 This is an all or nothing option.
6696 @item -mcpu=@var{cpu}
6698 Compile code for ARC variant @var{cpu}.
6699 Which variants are supported depend on the configuration.
6700 All variants support @option{-mcpu=base}, this is the default.
6702 @item -mtext=@var{text-section}
6703 @itemx -mdata=@var{data-section}
6704 @itemx -mrodata=@var{readonly-data-section}
6708 Put functions, data, and readonly data in @var{text-section},
6709 @var{data-section}, and @var{readonly-data-section} respectively
6710 by default. This can be overridden with the @code{section} attribute.
6711 @xref{Variable Attributes}.
6716 @subsection ARM Options
6719 These @samp{-m} options are defined for Advanced RISC Machines (ARM)
6723 @item -mabi=@var{name}
6725 Generate code for the specified ABI@. Permissible values are: @samp{apcs-gnu},
6726 @samp{atpcs}, @samp{aapcs} and @samp{iwmmxt}.
6729 @opindex mapcs-frame
6730 Generate a stack frame that is compliant with the ARM Procedure Call
6731 Standard for all functions, even if this is not strictly necessary for
6732 correct execution of the code. Specifying @option{-fomit-frame-pointer}
6733 with this option will cause the stack frames not to be generated for
6734 leaf functions. The default is @option{-mno-apcs-frame}.
6738 This is a synonym for @option{-mapcs-frame}.
6741 @c not currently implemented
6742 @item -mapcs-stack-check
6743 @opindex mapcs-stack-check
6744 Generate code to check the amount of stack space available upon entry to
6745 every function (that actually uses some stack space). If there is
6746 insufficient space available then either the function
6747 @samp{__rt_stkovf_split_small} or @samp{__rt_stkovf_split_big} will be
6748 called, depending upon the amount of stack space required. The run time
6749 system is required to provide these functions. The default is
6750 @option{-mno-apcs-stack-check}, since this produces smaller code.
6752 @c not currently implemented
6754 @opindex mapcs-float
6755 Pass floating point arguments using the float point registers. This is
6756 one of the variants of the APCS@. This option is recommended if the
6757 target hardware has a floating point unit or if a lot of floating point
6758 arithmetic is going to be performed by the code. The default is
6759 @option{-mno-apcs-float}, since integer only code is slightly increased in
6760 size if @option{-mapcs-float} is used.
6762 @c not currently implemented
6763 @item -mapcs-reentrant
6764 @opindex mapcs-reentrant
6765 Generate reentrant, position independent code. The default is
6766 @option{-mno-apcs-reentrant}.
6769 @item -mthumb-interwork
6770 @opindex mthumb-interwork
6771 Generate code which supports calling between the ARM and Thumb
6772 instruction sets. Without this option the two instruction sets cannot
6773 be reliably used inside one program. The default is
6774 @option{-mno-thumb-interwork}, since slightly larger code is generated
6775 when @option{-mthumb-interwork} is specified.
6777 @item -mno-sched-prolog
6778 @opindex mno-sched-prolog
6779 Prevent the reordering of instructions in the function prolog, or the
6780 merging of those instruction with the instructions in the function's
6781 body. This means that all functions will start with a recognizable set
6782 of instructions (or in fact one of a choice from a small set of
6783 different function prologues), and this information can be used to
6784 locate the start if functions inside an executable piece of code. The
6785 default is @option{-msched-prolog}.
6788 @opindex mhard-float
6789 Generate output containing floating point instructions. This is the
6793 @opindex msoft-float
6794 Generate output containing library calls for floating point.
6795 @strong{Warning:} the requisite libraries are not available for all ARM
6796 targets. Normally the facilities of the machine's usual C compiler are
6797 used, but this cannot be done directly in cross-compilation. You must make
6798 your own arrangements to provide suitable library functions for
6801 @option{-msoft-float} changes the calling convention in the output file;
6802 therefore, it is only useful if you compile @emph{all} of a program with
6803 this option. In particular, you need to compile @file{libgcc.a}, the
6804 library that comes with GCC, with @option{-msoft-float} in order for
6807 @item -mfloat-abi=@var{name}
6809 Specifies which ABI to use for floating point values. Permissible values
6810 are: @samp{soft}, @samp{softfp} and @samp{hard}.
6812 @samp{soft} and @samp{hard} are equivalent to @option{-msoft-float}
6813 and @option{-mhard-float} respectively. @samp{softfp} allows the generation
6814 of floating point instructions, but still uses the soft-float calling
6817 @item -mlittle-endian
6818 @opindex mlittle-endian
6819 Generate code for a processor running in little-endian mode. This is
6820 the default for all standard configurations.
6823 @opindex mbig-endian
6824 Generate code for a processor running in big-endian mode; the default is
6825 to compile code for a little-endian processor.
6827 @item -mwords-little-endian
6828 @opindex mwords-little-endian
6829 This option only applies when generating code for big-endian processors.
6830 Generate code for a little-endian word order but a big-endian byte
6831 order. That is, a byte order of the form @samp{32107654}. Note: this
6832 option should only be used if you require compatibility with code for
6833 big-endian ARM processors generated by versions of the compiler prior to
6836 @item -mcpu=@var{name}
6838 This specifies the name of the target ARM processor. GCC uses this name
6839 to determine what kind of instructions it can emit when generating
6840 assembly code. Permissible names are: @samp{arm2}, @samp{arm250},
6841 @samp{arm3}, @samp{arm6}, @samp{arm60}, @samp{arm600}, @samp{arm610},
6842 @samp{arm620}, @samp{arm7}, @samp{arm7m}, @samp{arm7d}, @samp{arm7dm},
6843 @samp{arm7di}, @samp{arm7dmi}, @samp{arm70}, @samp{arm700},
6844 @samp{arm700i}, @samp{arm710}, @samp{arm710c}, @samp{arm7100},
6845 @samp{arm7500}, @samp{arm7500fe}, @samp{arm7tdmi}, @samp{arm7tdmi-s},
6846 @samp{arm8}, @samp{strongarm}, @samp{strongarm110}, @samp{strongarm1100},
6847 @samp{arm8}, @samp{arm810}, @samp{arm9}, @samp{arm9e}, @samp{arm920},
6848 @samp{arm920t}, @samp{arm922t}, @samp{arm946e-s}, @samp{arm966e-s},
6849 @samp{arm968e-s}, @samp{arm926ej-s}, @samp{arm940t}, @samp{arm9tdmi},
6850 @samp{arm10tdmi}, @samp{arm1020t}, @samp{arm1026ej-s},
6851 @samp{arm10e}, @samp{arm1020e}, @samp{arm1022e},
6852 @samp{arm1136j-s}, @samp{arm1136jf-s}, @samp{mpcore}, @samp{mpcorenovfp},
6853 @samp{arm1176jz-s}, @samp{arm1176jzf-s}, @samp{xscale}, @samp{iwmmxt},
6856 @itemx -mtune=@var{name}
6858 This option is very similar to the @option{-mcpu=} option, except that
6859 instead of specifying the actual target processor type, and hence
6860 restricting which instructions can be used, it specifies that GCC should
6861 tune the performance of the code as if the target were of the type
6862 specified in this option, but still choosing the instructions that it
6863 will generate based on the cpu specified by a @option{-mcpu=} option.
6864 For some ARM implementations better performance can be obtained by using
6867 @item -march=@var{name}
6869 This specifies the name of the target ARM architecture. GCC uses this
6870 name to determine what kind of instructions it can emit when generating
6871 assembly code. This option can be used in conjunction with or instead
6872 of the @option{-mcpu=} option. Permissible names are: @samp{armv2},
6873 @samp{armv2a}, @samp{armv3}, @samp{armv3m}, @samp{armv4}, @samp{armv4t},
6874 @samp{armv5}, @samp{armv5t}, @samp{armv5te}, @samp{armv6}, @samp{armv6j},
6875 @samp{iwmmxt}, @samp{ep9312}.
6877 @item -mfpu=@var{name}
6878 @itemx -mfpe=@var{number}
6879 @itemx -mfp=@var{number}
6883 This specifies what floating point hardware (or hardware emulation) is
6884 available on the target. Permissible names are: @samp{fpa}, @samp{fpe2},
6885 @samp{fpe3}, @samp{maverick}, @samp{vfp}. @option{-mfp} and @option{-mfpe}
6886 are synonyms for @option{-mfpu}=@samp{fpe}@var{number}, for compatibility
6887 with older versions of GCC@.
6889 If @option{-msoft-float} is specified this specifies the format of
6890 floating point values.
6892 @item -mstructure-size-boundary=@var{n}
6893 @opindex mstructure-size-boundary
6894 The size of all structures and unions will be rounded up to a multiple
6895 of the number of bits set by this option. Permissible values are 8, 32
6896 and 64. The default value varies for different toolchains. For the COFF
6897 targeted toolchain the default value is 8. A value of 64 is only allowed
6898 if the underlying ABI supports it.
6900 Specifying the larger number can produce faster, more efficient code, but
6901 can also increase the size of the program. Different values are potentially
6902 incompatible. Code compiled with one value cannot necessarily expect to
6903 work with code or libraries compiled with another value, if they exchange
6904 information using structures or unions.
6906 @item -mabort-on-noreturn
6907 @opindex mabort-on-noreturn
6908 Generate a call to the function @code{abort} at the end of a
6909 @code{noreturn} function. It will be executed if the function tries to
6913 @itemx -mno-long-calls
6914 @opindex mlong-calls
6915 @opindex mno-long-calls
6916 Tells the compiler to perform function calls by first loading the
6917 address of the function into a register and then performing a subroutine
6918 call on this register. This switch is needed if the target function
6919 will lie outside of the 64 megabyte addressing range of the offset based
6920 version of subroutine call instruction.
6922 Even if this switch is enabled, not all function calls will be turned
6923 into long calls. The heuristic is that static functions, functions
6924 which have the @samp{short-call} attribute, functions that are inside
6925 the scope of a @samp{#pragma no_long_calls} directive and functions whose
6926 definitions have already been compiled within the current compilation
6927 unit, will not be turned into long calls. The exception to this rule is
6928 that weak function definitions, functions with the @samp{long-call}
6929 attribute or the @samp{section} attribute, and functions that are within
6930 the scope of a @samp{#pragma long_calls} directive, will always be
6931 turned into long calls.
6933 This feature is not enabled by default. Specifying
6934 @option{-mno-long-calls} will restore the default behavior, as will
6935 placing the function calls within the scope of a @samp{#pragma
6936 long_calls_off} directive. Note these switches have no effect on how
6937 the compiler generates code to handle function calls via function
6940 @item -mnop-fun-dllimport
6941 @opindex mnop-fun-dllimport
6942 Disable support for the @code{dllimport} attribute.
6944 @item -msingle-pic-base
6945 @opindex msingle-pic-base
6946 Treat the register used for PIC addressing as read-only, rather than
6947 loading it in the prologue for each function. The run-time system is
6948 responsible for initializing this register with an appropriate value
6949 before execution begins.
6951 @item -mpic-register=@var{reg}
6952 @opindex mpic-register
6953 Specify the register to be used for PIC addressing. The default is R10
6954 unless stack-checking is enabled, when R9 is used.
6956 @item -mcirrus-fix-invalid-insns
6957 @opindex mcirrus-fix-invalid-insns
6958 @opindex mno-cirrus-fix-invalid-insns
6959 Insert NOPs into the instruction stream to in order to work around
6960 problems with invalid Maverick instruction combinations. This option
6961 is only valid if the @option{-mcpu=ep9312} option has been used to
6962 enable generation of instructions for the Cirrus Maverick floating
6963 point co-processor. This option is not enabled by default, since the
6964 problem is only present in older Maverick implementations. The default
6965 can be re-enabled by use of the @option{-mno-cirrus-fix-invalid-insns}
6968 @item -mpoke-function-name
6969 @opindex mpoke-function-name
6970 Write the name of each function into the text section, directly
6971 preceding the function prologue. The generated code is similar to this:
6975 .ascii "arm_poke_function_name", 0
6978 .word 0xff000000 + (t1 - t0)
6979 arm_poke_function_name
6981 stmfd sp!, @{fp, ip, lr, pc@}
6985 When performing a stack backtrace, code can inspect the value of
6986 @code{pc} stored at @code{fp + 0}. If the trace function then looks at
6987 location @code{pc - 12} and the top 8 bits are set, then we know that
6988 there is a function name embedded immediately preceding this location
6989 and has length @code{((pc[-3]) & 0xff000000)}.
6993 Generate code for the 16-bit Thumb instruction set. The default is to
6994 use the 32-bit ARM instruction set.
6997 @opindex mtpcs-frame
6998 Generate a stack frame that is compliant with the Thumb Procedure Call
6999 Standard for all non-leaf functions. (A leaf function is one that does
7000 not call any other functions.) The default is @option{-mno-tpcs-frame}.
7002 @item -mtpcs-leaf-frame
7003 @opindex mtpcs-leaf-frame
7004 Generate a stack frame that is compliant with the Thumb Procedure Call
7005 Standard for all leaf functions. (A leaf function is one that does
7006 not call any other functions.) The default is @option{-mno-apcs-leaf-frame}.
7008 @item -mcallee-super-interworking
7009 @opindex mcallee-super-interworking
7010 Gives all externally visible functions in the file being compiled an ARM
7011 instruction set header which switches to Thumb mode before executing the
7012 rest of the function. This allows these functions to be called from
7013 non-interworking code.
7015 @item -mcaller-super-interworking
7016 @opindex mcaller-super-interworking
7017 Allows calls via function pointers (including virtual functions) to
7018 execute correctly regardless of whether the target code has been
7019 compiled for interworking or not. There is a small overhead in the cost
7020 of executing a function pointer if this option is enabled.
7025 @subsection AVR Options
7028 These options are defined for AVR implementations:
7031 @item -mmcu=@var{mcu}
7033 Specify ATMEL AVR instruction set or MCU type.
7035 Instruction set avr1 is for the minimal AVR core, not supported by the C
7036 compiler, only for assembler programs (MCU types: at90s1200, attiny10,
7037 attiny11, attiny12, attiny15, attiny28).
7039 Instruction set avr2 (default) is for the classic AVR core with up to
7040 8K program memory space (MCU types: at90s2313, at90s2323, attiny22,
7041 at90s2333, at90s2343, at90s4414, at90s4433, at90s4434, at90s8515,
7042 at90c8534, at90s8535).
7044 Instruction set avr3 is for the classic AVR core with up to 128K program
7045 memory space (MCU types: atmega103, atmega603, at43usb320, at76c711).
7047 Instruction set avr4 is for the enhanced AVR core with up to 8K program
7048 memory space (MCU types: atmega8, atmega83, atmega85).
7050 Instruction set avr5 is for the enhanced AVR core with up to 128K program
7051 memory space (MCU types: atmega16, atmega161, atmega163, atmega32, atmega323,
7052 atmega64, atmega128, at43usb355, at94k).
7056 Output instruction sizes to the asm file.
7058 @item -minit-stack=@var{N}
7059 @opindex minit-stack
7060 Specify the initial stack address, which may be a symbol or numeric value,
7061 @samp{__stack} is the default.
7063 @item -mno-interrupts
7064 @opindex mno-interrupts
7065 Generated code is not compatible with hardware interrupts.
7066 Code size will be smaller.
7068 @item -mcall-prologues
7069 @opindex mcall-prologues
7070 Functions prologues/epilogues expanded as call to appropriate
7071 subroutines. Code size will be smaller.
7073 @item -mno-tablejump
7074 @opindex mno-tablejump
7075 Do not generate tablejump insns which sometimes increase code size.
7078 @opindex mtiny-stack
7079 Change only the low 8 bits of the stack pointer.
7083 Assume int to be 8 bit integer. This affects the sizes of all types: A
7084 char will be 1 byte, an int will be 1 byte, an long will be 2 bytes
7085 and long long will be 4 bytes. Please note that this option does not
7086 comply to the C standards, but it will provide you with smaller code
7091 @subsection CRIS Options
7092 @cindex CRIS Options
7094 These options are defined specifically for the CRIS ports.
7097 @item -march=@var{architecture-type}
7098 @itemx -mcpu=@var{architecture-type}
7101 Generate code for the specified architecture. The choices for
7102 @var{architecture-type} are @samp{v3}, @samp{v8} and @samp{v10} for
7103 respectively ETRAX@w{ }4, ETRAX@w{ }100, and ETRAX@w{ }100@w{ }LX@.
7104 Default is @samp{v0} except for cris-axis-linux-gnu, where the default is
7107 @item -mtune=@var{architecture-type}
7109 Tune to @var{architecture-type} everything applicable about the generated
7110 code, except for the ABI and the set of available instructions. The
7111 choices for @var{architecture-type} are the same as for
7112 @option{-march=@var{architecture-type}}.
7114 @item -mmax-stack-frame=@var{n}
7115 @opindex mmax-stack-frame
7116 Warn when the stack frame of a function exceeds @var{n} bytes.
7118 @item -melinux-stacksize=@var{n}
7119 @opindex melinux-stacksize
7120 Only available with the @samp{cris-axis-aout} target. Arranges for
7121 indications in the program to the kernel loader that the stack of the
7122 program should be set to @var{n} bytes.
7128 The options @option{-metrax4} and @option{-metrax100} are synonyms for
7129 @option{-march=v3} and @option{-march=v8} respectively.
7131 @item -mmul-bug-workaround
7132 @itemx -mno-mul-bug-workaround
7133 @opindex mmul-bug-workaround
7134 @opindex mno-mul-bug-workaround
7135 Work around a bug in the @code{muls} and @code{mulu} instructions for CPU
7136 models where it applies. This option is active by default.
7140 Enable CRIS-specific verbose debug-related information in the assembly
7141 code. This option also has the effect to turn off the @samp{#NO_APP}
7142 formatted-code indicator to the assembler at the beginning of the
7147 Do not use condition-code results from previous instruction; always emit
7148 compare and test instructions before use of condition codes.
7150 @item -mno-side-effects
7151 @opindex mno-side-effects
7152 Do not emit instructions with side-effects in addressing modes other than
7156 @itemx -mno-stack-align
7158 @itemx -mno-data-align
7159 @itemx -mconst-align
7160 @itemx -mno-const-align
7161 @opindex mstack-align
7162 @opindex mno-stack-align
7163 @opindex mdata-align
7164 @opindex mno-data-align
7165 @opindex mconst-align
7166 @opindex mno-const-align
7167 These options (no-options) arranges (eliminate arrangements) for the
7168 stack-frame, individual data and constants to be aligned for the maximum
7169 single data access size for the chosen CPU model. The default is to
7170 arrange for 32-bit alignment. ABI details such as structure layout are
7171 not affected by these options.
7179 Similar to the stack- data- and const-align options above, these options
7180 arrange for stack-frame, writable data and constants to all be 32-bit,
7181 16-bit or 8-bit aligned. The default is 32-bit alignment.
7183 @item -mno-prologue-epilogue
7184 @itemx -mprologue-epilogue
7185 @opindex mno-prologue-epilogue
7186 @opindex mprologue-epilogue
7187 With @option{-mno-prologue-epilogue}, the normal function prologue and
7188 epilogue that sets up the stack-frame are omitted and no return
7189 instructions or return sequences are generated in the code. Use this
7190 option only together with visual inspection of the compiled code: no
7191 warnings or errors are generated when call-saved registers must be saved,
7192 or storage for local variable needs to be allocated.
7198 With @option{-fpic} and @option{-fPIC}, don't generate (do generate)
7199 instruction sequences that load addresses for functions from the PLT part
7200 of the GOT rather than (traditional on other architectures) calls to the
7201 PLT@. The default is @option{-mgotplt}.
7205 Legacy no-op option only recognized with the cris-axis-aout target.
7209 Legacy no-op option only recognized with the cris-axis-elf and
7210 cris-axis-linux-gnu targets.
7214 Only recognized with the cris-axis-aout target, where it selects a
7215 GNU/linux-like multilib, include files and instruction set for
7220 Legacy no-op option only recognized with the cris-axis-linux-gnu target.
7224 This option, recognized for the cris-axis-aout and cris-axis-elf arranges
7225 to link with input-output functions from a simulator library. Code,
7226 initialized data and zero-initialized data are allocated consecutively.
7230 Like @option{-sim}, but pass linker options to locate initialized data at
7231 0x40000000 and zero-initialized data at 0x80000000.
7234 @node Darwin Options
7235 @subsection Darwin Options
7236 @cindex Darwin options
7238 These options are defined for all architectures running the Darwin operating
7241 FSF GCC on Darwin does not create ``fat'' object files; it will create
7242 an object file for the single architecture that it was built to
7243 target. Apple's GCC on Darwin does create ``fat'' files if multiple
7244 @option{-arch} options are used; it does so by running the compiler or
7245 linker multiple times and joining the results together with
7248 The subtype of the file created (like @samp{ppc7400} or @samp{ppc970} or
7249 @samp{i686}) is determined by the flags that specify the ISA
7250 that GCC is targetting, like @option{-mcpu} or @option{-march}. The
7251 @option{-force_cpusubtype_ALL} option can be used to override this.
7253 The Darwin tools vary in their behaviour when presented with an ISA
7254 mismatch. The assembler, @file{as}, will only permit instructions to
7255 be used that are valid for the subtype of the file it is generating,
7256 so you cannot put 64-bit instructions in an @samp{ppc750} object file.
7257 The linker for shared libraries, @file{/usr/bin/libtool}, will fail
7258 and print an error if asked to create a shared library with a less
7259 restrictive subtype than its input files (for instance, trying to put
7260 a @samp{ppc970} object file in a @samp{ppc7400} library). The linker
7261 for executables, @file{ld}, will quietly give the executable the most
7262 restrictive subtype of any of its input files.
7267 Add the framework directory @var{dir} to the head of the list of
7268 directories to be searched for header files. These directories are
7269 interleaved with those specified by @option{-I} options and are
7270 scanned in a left-to-right order.
7272 A framework directory is a directory with frameworks in it. A
7273 framework is a directory with a @samp{"Headers"} and/or
7274 @samp{"PrivateHeaders"} directory contained directly in it that ends
7275 in @samp{".framework"}. The name of a framework is the name of this
7276 directory excluding the @samp{".framework"}. Headers associated with
7277 the framework are found in one of those two directories, with
7278 @samp{"Headers"} being searched first. A subframework is a framework
7279 directory that is in a framework's @samp{"Frameworks"} directory.
7280 Includes of subframework headers can only appear in a header of a
7281 framework that contains the subframework, or in a sibling subframework
7282 header. Two subframeworks are siblings if they occur in the same
7283 framework. A subframework should not have the same name as a
7284 framework, a warning will be issued if this is violated. Currently a
7285 subframework cannot have subframeworks, in the future, the mechanism
7286 may be extended to support this. The standard frameworks can be found
7287 in @samp{"/System/Library/Frameworks"} and
7288 @samp{"/Library/Frameworks"}. An example include looks like
7289 @code{#include <Framework/header.h>}, where @samp{Framework} denotes
7290 the name of the framework and header.h is found in the
7291 @samp{"PrivateHeaders"} or @samp{"Headers"} directory.
7295 Emit debugging information for symbols that are used. For STABS
7296 debugging format, this enables @option{-feliminate-unused-debug-symbols}.
7297 This is by default ON@.
7301 Emit debugging information for all symbols and types.
7303 @item -mone-byte-bool
7304 @opindex -mone-byte-bool
7305 Override the defaults for @samp{bool} so that @samp{sizeof(bool)==1}.
7306 By default @samp{sizeof(bool)} is @samp{4} when compiling for
7307 Darwin/PowerPC and @samp{1} when compiling for Darwin/x86, so this
7308 option has no effect on x86.
7310 @strong{Warning:} The @option{-mone-byte-bool} switch causes GCC
7311 to generate code that is not binary compatible with code generated
7312 without that switch. Using this switch may require recompiling all
7313 other modules in a program, including system libraries. Use this
7314 switch to conform to a non-default data model.
7316 @item -mfix-and-continue
7317 @itemx -ffix-and-continue
7318 @itemx -findirect-data
7319 @opindex mfix-and-continue
7320 @opindex ffix-and-continue
7321 @opindex findirect-data
7322 Generate code suitable for fast turn around development. Needed to
7323 enable gdb to dynamically load @code{.o} files into already running
7324 programs. @option{-findirect-data} and @option{-ffix-and-continue}
7325 are provided for backwards compatibility.
7329 Loads all members of static archive libraries.
7330 See man ld(1) for more information.
7332 @item -arch_errors_fatal
7333 @opindex arch_errors_fatal
7334 Cause the errors having to do with files that have the wrong architecture
7338 @opindex bind_at_load
7339 Causes the output file to be marked such that the dynamic linker will
7340 bind all undefined references when the file is loaded or launched.
7344 Produce a Mach-o bundle format file.
7345 See man ld(1) for more information.
7347 @item -bundle_loader @var{executable}
7348 @opindex bundle_loader
7349 This option specifies the @var{executable} that will be loading the build
7350 output file being linked. See man ld(1) for more information.
7353 @opindex -dynamiclib
7354 When passed this option, GCC will produce a dynamic library instead of
7355 an executable when linking, using the Darwin @file{libtool} command.
7357 @item -force_cpusubtype_ALL
7358 @opindex -force_cpusubtype_ALL
7359 This causes GCC's output file to have the @var{ALL} subtype, instead of
7360 one controlled by the @option{-mcpu} or @option{-march} option.
7362 @item -allowable_client @var{client_name}
7364 @itemx -compatibility_version
7365 @itemx -current_version
7367 @itemx -dependency-file
7369 @itemx -dylinker_install_name
7371 @itemx -exported_symbols_list
7373 @itemx -flat_namespace
7374 @itemx -force_flat_namespace
7375 @itemx -headerpad_max_install_names
7378 @itemx -install_name
7379 @itemx -keep_private_externs
7380 @itemx -multi_module
7381 @itemx -multiply_defined
7382 @itemx -multiply_defined_unused
7384 @itemx -no_dead_strip_inits_and_terms
7385 @itemx -nofixprebinding
7388 @itemx -noseglinkedit
7389 @itemx -pagezero_size
7391 @itemx -prebind_all_twolevel_modules
7392 @itemx -private_bundle
7393 @itemx -read_only_relocs
7395 @itemx -sectobjectsymbols
7399 @itemx -sectobjectsymbols
7402 @itemx -segs_read_only_addr
7403 @itemx -segs_read_write_addr
7404 @itemx -seg_addr_table
7405 @itemx -seg_addr_table_filename
7408 @itemx -segs_read_only_addr
7409 @itemx -segs_read_write_addr
7410 @itemx -single_module
7413 @itemx -sub_umbrella
7414 @itemx -twolevel_namespace
7417 @itemx -unexported_symbols_list
7418 @itemx -weak_reference_mismatches
7421 @opindex allowable_client
7422 @opindex client_name
7423 @opindex compatibility_version
7424 @opindex current_version
7426 @opindex dependency-file
7428 @opindex dylinker_install_name
7430 @opindex exported_symbols_list
7432 @opindex flat_namespace
7433 @opindex force_flat_namespace
7434 @opindex headerpad_max_install_names
7437 @opindex install_name
7438 @opindex keep_private_externs
7439 @opindex multi_module
7440 @opindex multiply_defined
7441 @opindex multiply_defined_unused
7443 @opindex no_dead_strip_inits_and_terms
7444 @opindex nofixprebinding
7445 @opindex nomultidefs
7447 @opindex noseglinkedit
7448 @opindex pagezero_size
7450 @opindex prebind_all_twolevel_modules
7451 @opindex private_bundle
7452 @opindex read_only_relocs
7454 @opindex sectobjectsymbols
7458 @opindex sectobjectsymbols
7461 @opindex segs_read_only_addr
7462 @opindex segs_read_write_addr
7463 @opindex seg_addr_table
7464 @opindex seg_addr_table_filename
7465 @opindex seglinkedit
7467 @opindex segs_read_only_addr
7468 @opindex segs_read_write_addr
7469 @opindex single_module
7471 @opindex sub_library
7472 @opindex sub_umbrella
7473 @opindex twolevel_namespace
7476 @opindex unexported_symbols_list
7477 @opindex weak_reference_mismatches
7478 @opindex whatsloaded
7480 These options are passed to the Darwin linker. The Darwin linker man page
7481 describes them in detail.
7484 @node DEC Alpha Options
7485 @subsection DEC Alpha Options
7487 These @samp{-m} options are defined for the DEC Alpha implementations:
7490 @item -mno-soft-float
7492 @opindex mno-soft-float
7493 @opindex msoft-float
7494 Use (do not use) the hardware floating-point instructions for
7495 floating-point operations. When @option{-msoft-float} is specified,
7496 functions in @file{libgcc.a} will be used to perform floating-point
7497 operations. Unless they are replaced by routines that emulate the
7498 floating-point operations, or compiled in such a way as to call such
7499 emulations routines, these routines will issue floating-point
7500 operations. If you are compiling for an Alpha without floating-point
7501 operations, you must ensure that the library is built so as not to call
7504 Note that Alpha implementations without floating-point operations are
7505 required to have floating-point registers.
7510 @opindex mno-fp-regs
7511 Generate code that uses (does not use) the floating-point register set.
7512 @option{-mno-fp-regs} implies @option{-msoft-float}. If the floating-point
7513 register set is not used, floating point operands are passed in integer
7514 registers as if they were integers and floating-point results are passed
7515 in @code{$0} instead of @code{$f0}. This is a non-standard calling sequence,
7516 so any function with a floating-point argument or return value called by code
7517 compiled with @option{-mno-fp-regs} must also be compiled with that
7520 A typical use of this option is building a kernel that does not use,
7521 and hence need not save and restore, any floating-point registers.
7525 The Alpha architecture implements floating-point hardware optimized for
7526 maximum performance. It is mostly compliant with the IEEE floating
7527 point standard. However, for full compliance, software assistance is
7528 required. This option generates code fully IEEE compliant code
7529 @emph{except} that the @var{inexact-flag} is not maintained (see below).
7530 If this option is turned on, the preprocessor macro @code{_IEEE_FP} is
7531 defined during compilation. The resulting code is less efficient but is
7532 able to correctly support denormalized numbers and exceptional IEEE
7533 values such as not-a-number and plus/minus infinity. Other Alpha
7534 compilers call this option @option{-ieee_with_no_inexact}.
7536 @item -mieee-with-inexact
7537 @opindex mieee-with-inexact
7538 This is like @option{-mieee} except the generated code also maintains
7539 the IEEE @var{inexact-flag}. Turning on this option causes the
7540 generated code to implement fully-compliant IEEE math. In addition to
7541 @code{_IEEE_FP}, @code{_IEEE_FP_EXACT} is defined as a preprocessor
7542 macro. On some Alpha implementations the resulting code may execute
7543 significantly slower than the code generated by default. Since there is
7544 very little code that depends on the @var{inexact-flag}, you should
7545 normally not specify this option. Other Alpha compilers call this
7546 option @option{-ieee_with_inexact}.
7548 @item -mfp-trap-mode=@var{trap-mode}
7549 @opindex mfp-trap-mode
7550 This option controls what floating-point related traps are enabled.
7551 Other Alpha compilers call this option @option{-fptm @var{trap-mode}}.
7552 The trap mode can be set to one of four values:
7556 This is the default (normal) setting. The only traps that are enabled
7557 are the ones that cannot be disabled in software (e.g., division by zero
7561 In addition to the traps enabled by @samp{n}, underflow traps are enabled
7565 Like @samp{su}, but the instructions are marked to be safe for software
7566 completion (see Alpha architecture manual for details).
7569 Like @samp{su}, but inexact traps are enabled as well.
7572 @item -mfp-rounding-mode=@var{rounding-mode}
7573 @opindex mfp-rounding-mode
7574 Selects the IEEE rounding mode. Other Alpha compilers call this option
7575 @option{-fprm @var{rounding-mode}}. The @var{rounding-mode} can be one
7580 Normal IEEE rounding mode. Floating point numbers are rounded towards
7581 the nearest machine number or towards the even machine number in case
7585 Round towards minus infinity.
7588 Chopped rounding mode. Floating point numbers are rounded towards zero.
7591 Dynamic rounding mode. A field in the floating point control register
7592 (@var{fpcr}, see Alpha architecture reference manual) controls the
7593 rounding mode in effect. The C library initializes this register for
7594 rounding towards plus infinity. Thus, unless your program modifies the
7595 @var{fpcr}, @samp{d} corresponds to round towards plus infinity.
7598 @item -mtrap-precision=@var{trap-precision}
7599 @opindex mtrap-precision
7600 In the Alpha architecture, floating point traps are imprecise. This
7601 means without software assistance it is impossible to recover from a
7602 floating trap and program execution normally needs to be terminated.
7603 GCC can generate code that can assist operating system trap handlers
7604 in determining the exact location that caused a floating point trap.
7605 Depending on the requirements of an application, different levels of
7606 precisions can be selected:
7610 Program precision. This option is the default and means a trap handler
7611 can only identify which program caused a floating point exception.
7614 Function precision. The trap handler can determine the function that
7615 caused a floating point exception.
7618 Instruction precision. The trap handler can determine the exact
7619 instruction that caused a floating point exception.
7622 Other Alpha compilers provide the equivalent options called
7623 @option{-scope_safe} and @option{-resumption_safe}.
7625 @item -mieee-conformant
7626 @opindex mieee-conformant
7627 This option marks the generated code as IEEE conformant. You must not
7628 use this option unless you also specify @option{-mtrap-precision=i} and either
7629 @option{-mfp-trap-mode=su} or @option{-mfp-trap-mode=sui}. Its only effect
7630 is to emit the line @samp{.eflag 48} in the function prologue of the
7631 generated assembly file. Under DEC Unix, this has the effect that
7632 IEEE-conformant math library routines will be linked in.
7634 @item -mbuild-constants
7635 @opindex mbuild-constants
7636 Normally GCC examines a 32- or 64-bit integer constant to
7637 see if it can construct it from smaller constants in two or three
7638 instructions. If it cannot, it will output the constant as a literal and
7639 generate code to load it from the data segment at runtime.
7641 Use this option to require GCC to construct @emph{all} integer constants
7642 using code, even if it takes more instructions (the maximum is six).
7644 You would typically use this option to build a shared library dynamic
7645 loader. Itself a shared library, it must relocate itself in memory
7646 before it can find the variables and constants in its own data segment.
7652 Select whether to generate code to be assembled by the vendor-supplied
7653 assembler (@option{-malpha-as}) or by the GNU assembler @option{-mgas}.
7671 Indicate whether GCC should generate code to use the optional BWX,
7672 CIX, FIX and MAX instruction sets. The default is to use the instruction
7673 sets supported by the CPU type specified via @option{-mcpu=} option or that
7674 of the CPU on which GCC was built if none was specified.
7679 @opindex mfloat-ieee
7680 Generate code that uses (does not use) VAX F and G floating point
7681 arithmetic instead of IEEE single and double precision.
7683 @item -mexplicit-relocs
7684 @itemx -mno-explicit-relocs
7685 @opindex mexplicit-relocs
7686 @opindex mno-explicit-relocs
7687 Older Alpha assemblers provided no way to generate symbol relocations
7688 except via assembler macros. Use of these macros does not allow
7689 optimal instruction scheduling. GNU binutils as of version 2.12
7690 supports a new syntax that allows the compiler to explicitly mark
7691 which relocations should apply to which instructions. This option
7692 is mostly useful for debugging, as GCC detects the capabilities of
7693 the assembler when it is built and sets the default accordingly.
7697 @opindex msmall-data
7698 @opindex mlarge-data
7699 When @option{-mexplicit-relocs} is in effect, static data is
7700 accessed via @dfn{gp-relative} relocations. When @option{-msmall-data}
7701 is used, objects 8 bytes long or smaller are placed in a @dfn{small data area}
7702 (the @code{.sdata} and @code{.sbss} sections) and are accessed via
7703 16-bit relocations off of the @code{$gp} register. This limits the
7704 size of the small data area to 64KB, but allows the variables to be
7705 directly accessed via a single instruction.
7707 The default is @option{-mlarge-data}. With this option the data area
7708 is limited to just below 2GB@. Programs that require more than 2GB of
7709 data must use @code{malloc} or @code{mmap} to allocate the data in the
7710 heap instead of in the program's data segment.
7712 When generating code for shared libraries, @option{-fpic} implies
7713 @option{-msmall-data} and @option{-fPIC} implies @option{-mlarge-data}.
7717 @opindex msmall-text
7718 @opindex mlarge-text
7719 When @option{-msmall-text} is used, the compiler assumes that the
7720 code of the entire program (or shared library) fits in 4MB, and is
7721 thus reachable with a branch instruction. When @option{-msmall-data}
7722 is used, the compiler can assume that all local symbols share the
7723 same @code{$gp} value, and thus reduce the number of instructions
7724 required for a function call from 4 to 1.
7726 The default is @option{-mlarge-text}.
7728 @item -mcpu=@var{cpu_type}
7730 Set the instruction set and instruction scheduling parameters for
7731 machine type @var{cpu_type}. You can specify either the @samp{EV}
7732 style name or the corresponding chip number. GCC supports scheduling
7733 parameters for the EV4, EV5 and EV6 family of processors and will
7734 choose the default values for the instruction set from the processor
7735 you specify. If you do not specify a processor type, GCC will default
7736 to the processor on which the compiler was built.
7738 Supported values for @var{cpu_type} are
7744 Schedules as an EV4 and has no instruction set extensions.
7748 Schedules as an EV5 and has no instruction set extensions.
7752 Schedules as an EV5 and supports the BWX extension.
7757 Schedules as an EV5 and supports the BWX and MAX extensions.
7761 Schedules as an EV6 and supports the BWX, FIX, and MAX extensions.
7765 Schedules as an EV6 and supports the BWX, CIX, FIX, and MAX extensions.
7768 @item -mtune=@var{cpu_type}
7770 Set only the instruction scheduling parameters for machine type
7771 @var{cpu_type}. The instruction set is not changed.
7773 @item -mmemory-latency=@var{time}
7774 @opindex mmemory-latency
7775 Sets the latency the scheduler should assume for typical memory
7776 references as seen by the application. This number is highly
7777 dependent on the memory access patterns used by the application
7778 and the size of the external cache on the machine.
7780 Valid options for @var{time} are
7784 A decimal number representing clock cycles.
7790 The compiler contains estimates of the number of clock cycles for
7791 ``typical'' EV4 & EV5 hardware for the Level 1, 2 & 3 caches
7792 (also called Dcache, Scache, and Bcache), as well as to main memory.
7793 Note that L3 is only valid for EV5.
7798 @node DEC Alpha/VMS Options
7799 @subsection DEC Alpha/VMS Options
7801 These @samp{-m} options are defined for the DEC Alpha/VMS implementations:
7804 @item -mvms-return-codes
7805 @opindex mvms-return-codes
7806 Return VMS condition codes from main. The default is to return POSIX
7807 style condition (e.g.@ error) codes.
7811 @subsection FRV Options
7818 Only use the first 32 general purpose registers.
7823 Use all 64 general purpose registers.
7828 Use only the first 32 floating point registers.
7833 Use all 64 floating point registers
7836 @opindex mhard-float
7838 Use hardware instructions for floating point operations.
7841 @opindex msoft-float
7843 Use library routines for floating point operations.
7848 Dynamically allocate condition code registers.
7853 Do not try to dynamically allocate condition code registers, only
7854 use @code{icc0} and @code{fcc0}.
7859 Change ABI to use double word insns.
7864 Do not use double word instructions.
7869 Use floating point double instructions.
7874 Do not use floating point double instructions.
7879 Use media instructions.
7884 Do not use media instructions.
7889 Use multiply and add/subtract instructions.
7894 Do not use multiply and add/subtract instructions.
7899 Select the FDPIC ABI, that uses function descriptors to represent
7900 pointers to functions. Without any PIC/PIE-related options, it
7901 implies @option{-fPIE}. With @option{-fpic} or @option{-fpie}, it
7902 assumes GOT entries and small data are within a 12-bit range from the
7903 GOT base address; with @option{-fPIC} or @option{-fPIE}, GOT offsets
7904 are computed with 32 bits.
7907 @opindex minline-plt
7909 Enable inlining of PLT entries in function calls to functions that are
7910 not known to bind locally. It has no effect without @option{-mfdpic}.
7911 It's enabled by default if optimizing for speed and compiling for
7912 shared libraries (i.e., @option{-fPIC} or @option{-fpic}), or when an
7913 optimization option such as @option{-O3} or above is present in the
7919 Assume a large TLS segment when generating thread-local code.
7924 Do not assume a large TLS segment when generating thread-local code.
7929 Enable the use of @code{GPREL} relocations in the FDPIC ABI for data
7930 that is known to be in read-only sections. It's enabled by default,
7931 except for @option{-fpic} or @option{-fpie}: even though it may help
7932 make the global offset table smaller, it trades 1 instruction for 4.
7933 With @option{-fPIC} or @option{-fPIE}, it trades 3 instructions for 4,
7934 one of which may be shared by multiple symbols, and it avoids the need
7935 for a GOT entry for the referenced symbol, so it's more likely to be a
7936 win. If it is not, @option{-mno-gprel-ro} can be used to disable it.
7938 @item -multilib-library-pic
7939 @opindex multilib-library-pic
7941 Link with the (library, not FD) pic libraries. It's implied by
7942 @option{-mlibrary-pic}, as well as by @option{-fPIC} and
7943 @option{-fpic} without @option{-mfdpic}. You should never have to use
7949 Follow the EABI requirement of always creating a frame pointer whenever
7950 a stack frame is allocated. This option is enabled by default and can
7951 be disabled with @option{-mno-linked-fp}.
7954 @opindex mlong-calls
7956 Use indirect addressing to call functions outside the current
7957 compilation unit. This allows the functions to be placed anywhere
7958 within the 32-bit address space.
7960 @item -malign-labels
7961 @opindex malign-labels
7963 Try to align labels to an 8-byte boundary by inserting nops into the
7964 previous packet. This option only has an effect when VLIW packing
7965 is enabled. It doesn't create new packets; it merely adds nops to
7969 @opindex mlibrary-pic
7971 Generate position-independent EABI code.
7976 Use only the first four media accumulator registers.
7981 Use all eight media accumulator registers.
7986 Pack VLIW instructions.
7991 Do not pack VLIW instructions.
7996 Do not mark ABI switches in e_flags.
8001 Enable the use of conditional-move instructions (default).
8003 This switch is mainly for debugging the compiler and will likely be removed
8004 in a future version.
8006 @item -mno-cond-move
8007 @opindex mno-cond-move
8009 Disable the use of conditional-move instructions.
8011 This switch is mainly for debugging the compiler and will likely be removed
8012 in a future version.
8017 Enable the use of conditional set instructions (default).
8019 This switch is mainly for debugging the compiler and will likely be removed
8020 in a future version.
8025 Disable the use of conditional set instructions.
8027 This switch is mainly for debugging the compiler and will likely be removed
8028 in a future version.
8033 Enable the use of conditional execution (default).
8035 This switch is mainly for debugging the compiler and will likely be removed
8036 in a future version.
8038 @item -mno-cond-exec
8039 @opindex mno-cond-exec
8041 Disable the use of conditional execution.
8043 This switch is mainly for debugging the compiler and will likely be removed
8044 in a future version.
8047 @opindex mvliw-branch
8049 Run a pass to pack branches into VLIW instructions (default).
8051 This switch is mainly for debugging the compiler and will likely be removed
8052 in a future version.
8054 @item -mno-vliw-branch
8055 @opindex mno-vliw-branch
8057 Do not run a pass to pack branches into VLIW instructions.
8059 This switch is mainly for debugging the compiler and will likely be removed
8060 in a future version.
8062 @item -mmulti-cond-exec
8063 @opindex mmulti-cond-exec
8065 Enable optimization of @code{&&} and @code{||} in conditional execution
8068 This switch is mainly for debugging the compiler and will likely be removed
8069 in a future version.
8071 @item -mno-multi-cond-exec
8072 @opindex mno-multi-cond-exec
8074 Disable optimization of @code{&&} and @code{||} in conditional execution.
8076 This switch is mainly for debugging the compiler and will likely be removed
8077 in a future version.
8079 @item -mnested-cond-exec
8080 @opindex mnested-cond-exec
8082 Enable nested conditional execution optimizations (default).
8084 This switch is mainly for debugging the compiler and will likely be removed
8085 in a future version.
8087 @item -mno-nested-cond-exec
8088 @opindex mno-nested-cond-exec
8090 Disable nested conditional execution optimizations.
8092 This switch is mainly for debugging the compiler and will likely be removed
8093 in a future version.
8095 @item -mtomcat-stats
8096 @opindex mtomcat-stats
8098 Cause gas to print out tomcat statistics.
8100 @item -mcpu=@var{cpu}
8103 Select the processor type for which to generate code. Possible values are
8104 @samp{frv}, @samp{fr550}, @samp{tomcat}, @samp{fr500}, @samp{fr450},
8105 @samp{fr405}, @samp{fr400}, @samp{fr300} and @samp{simple}.
8109 @node H8/300 Options
8110 @subsection H8/300 Options
8112 These @samp{-m} options are defined for the H8/300 implementations:
8117 Shorten some address references at link time, when possible; uses the
8118 linker option @option{-relax}. @xref{H8/300,, @code{ld} and the H8/300,
8119 ld, Using ld}, for a fuller description.
8123 Generate code for the H8/300H@.
8127 Generate code for the H8S@.
8131 Generate code for the H8S and H8/300H in the normal mode. This switch
8132 must be used either with @option{-mh} or @option{-ms}.
8136 Generate code for the H8S/2600. This switch must be used with @option{-ms}.
8140 Make @code{int} data 32 bits by default.
8144 On the H8/300H and H8S, use the same alignment rules as for the H8/300.
8145 The default for the H8/300H and H8S is to align longs and floats on 4
8147 @option{-malign-300} causes them to be aligned on 2 byte boundaries.
8148 This option has no effect on the H8/300.
8152 @subsection HPPA Options
8153 @cindex HPPA Options
8155 These @samp{-m} options are defined for the HPPA family of computers:
8158 @item -march=@var{architecture-type}
8160 Generate code for the specified architecture. The choices for
8161 @var{architecture-type} are @samp{1.0} for PA 1.0, @samp{1.1} for PA
8162 1.1, and @samp{2.0} for PA 2.0 processors. Refer to
8163 @file{/usr/lib/sched.models} on an HP-UX system to determine the proper
8164 architecture option for your machine. Code compiled for lower numbered
8165 architectures will run on higher numbered architectures, but not the
8169 @itemx -mpa-risc-1-1
8170 @itemx -mpa-risc-2-0
8171 @opindex mpa-risc-1-0
8172 @opindex mpa-risc-1-1
8173 @opindex mpa-risc-2-0
8174 Synonyms for @option{-march=1.0}, @option{-march=1.1}, and @option{-march=2.0} respectively.
8177 @opindex mbig-switch
8178 Generate code suitable for big switch tables. Use this option only if
8179 the assembler/linker complain about out of range branches within a switch
8182 @item -mjump-in-delay
8183 @opindex mjump-in-delay
8184 Fill delay slots of function calls with unconditional jump instructions
8185 by modifying the return pointer for the function call to be the target
8186 of the conditional jump.
8188 @item -mdisable-fpregs
8189 @opindex mdisable-fpregs
8190 Prevent floating point registers from being used in any manner. This is
8191 necessary for compiling kernels which perform lazy context switching of
8192 floating point registers. If you use this option and attempt to perform
8193 floating point operations, the compiler will abort.
8195 @item -mdisable-indexing
8196 @opindex mdisable-indexing
8197 Prevent the compiler from using indexing address modes. This avoids some
8198 rather obscure problems when compiling MIG generated code under MACH@.
8200 @item -mno-space-regs
8201 @opindex mno-space-regs
8202 Generate code that assumes the target has no space registers. This allows
8203 GCC to generate faster indirect calls and use unscaled index address modes.
8205 Such code is suitable for level 0 PA systems and kernels.
8207 @item -mfast-indirect-calls
8208 @opindex mfast-indirect-calls
8209 Generate code that assumes calls never cross space boundaries. This
8210 allows GCC to emit code which performs faster indirect calls.
8212 This option will not work in the presence of shared libraries or nested
8215 @item -mfixed-range=@var{register-range}
8216 @opindex mfixed-range
8217 Generate code treating the given register range as fixed registers.
8218 A fixed register is one that the register allocator can not use. This is
8219 useful when compiling kernel code. A register range is specified as
8220 two registers separated by a dash. Multiple register ranges can be
8221 specified separated by a comma.
8223 @item -mlong-load-store
8224 @opindex mlong-load-store
8225 Generate 3-instruction load and store sequences as sometimes required by
8226 the HP-UX 10 linker. This is equivalent to the @samp{+k} option to
8229 @item -mportable-runtime
8230 @opindex mportable-runtime
8231 Use the portable calling conventions proposed by HP for ELF systems.
8235 Enable the use of assembler directives only GAS understands.
8237 @item -mschedule=@var{cpu-type}
8239 Schedule code according to the constraints for the machine type
8240 @var{cpu-type}. The choices for @var{cpu-type} are @samp{700}
8241 @samp{7100}, @samp{7100LC}, @samp{7200}, @samp{7300} and @samp{8000}. Refer
8242 to @file{/usr/lib/sched.models} on an HP-UX system to determine the
8243 proper scheduling option for your machine. The default scheduling is
8247 @opindex mlinker-opt
8248 Enable the optimization pass in the HP-UX linker. Note this makes symbolic
8249 debugging impossible. It also triggers a bug in the HP-UX 8 and HP-UX 9
8250 linkers in which they give bogus error messages when linking some programs.
8253 @opindex msoft-float
8254 Generate output containing library calls for floating point.
8255 @strong{Warning:} the requisite libraries are not available for all HPPA
8256 targets. Normally the facilities of the machine's usual C compiler are
8257 used, but this cannot be done directly in cross-compilation. You must make
8258 your own arrangements to provide suitable library functions for
8259 cross-compilation. The embedded target @samp{hppa1.1-*-pro}
8260 does provide software floating point support.
8262 @option{-msoft-float} changes the calling convention in the output file;
8263 therefore, it is only useful if you compile @emph{all} of a program with
8264 this option. In particular, you need to compile @file{libgcc.a}, the
8265 library that comes with GCC, with @option{-msoft-float} in order for
8270 Generate the predefine, @code{_SIO}, for server IO@. The default is
8271 @option{-mwsio}. This generates the predefines, @code{__hp9000s700},
8272 @code{__hp9000s700__} and @code{_WSIO}, for workstation IO@. These
8273 options are available under HP-UX and HI-UX@.
8277 Use GNU ld specific options. This passes @option{-shared} to ld when
8278 building a shared library. It is the default when GCC is configured,
8279 explicitly or implicitly, with the GNU linker. This option does not
8280 have any affect on which ld is called, it only changes what parameters
8281 are passed to that ld. The ld that is called is determined by the
8282 @option{--with-ld} configure option, GCC's program search path, and
8283 finally by the user's @env{PATH}. The linker used by GCC can be printed
8284 using @samp{which `gcc -print-prog-name=ld`}.
8288 Use HP ld specific options. This passes @option{-b} to ld when building
8289 a shared library and passes @option{+Accept TypeMismatch} to ld on all
8290 links. It is the default when GCC is configured, explicitly or
8291 implicitly, with the HP linker. This option does not have any affect on
8292 which ld is called, it only changes what parameters are passed to that
8293 ld. The ld that is called is determined by the @option{--with-ld}
8294 configure option, GCC's program search path, and finally by the user's
8295 @env{PATH}. The linker used by GCC can be printed using @samp{which
8296 `gcc -print-prog-name=ld`}.
8299 @opindex mno-long-calls
8300 Generate code that uses long call sequences. This ensures that a call
8301 is always able to reach linker generated stubs. The default is to generate
8302 long calls only when the distance from the call site to the beginning
8303 of the function or translation unit, as the case may be, exceeds a
8304 predefined limit set by the branch type being used. The limits for
8305 normal calls are 7,600,000 and 240,000 bytes, respectively for the
8306 PA 2.0 and PA 1.X architectures. Sibcalls are always limited at
8309 Distances are measured from the beginning of functions when using the
8310 @option{-ffunction-sections} option, or when using the @option{-mgas}
8311 and @option{-mno-portable-runtime} options together under HP-UX with
8314 It is normally not desirable to use this option as it will degrade
8315 performance. However, it may be useful in large applications,
8316 particularly when partial linking is used to build the application.
8318 The types of long calls used depends on the capabilities of the
8319 assembler and linker, and the type of code being generated. The
8320 impact on systems that support long absolute calls, and long pic
8321 symbol-difference or pc-relative calls should be relatively small.
8322 However, an indirect call is used on 32-bit ELF systems in pic code
8323 and it is quite long.
8325 @item -munix=@var{unix-std}
8327 Generate compiler predefines and select a startfile for the specified
8328 UNIX standard. The choices for @var{unix-std} are @samp{93}, @samp{95}
8329 and @samp{98}. @samp{93} is supported on all HP-UX versions. @samp{95}
8330 is available on HP-UX 10.10 and later. @samp{98} is available on HP-UX
8331 11.11 and later. The default values are @samp{93} for HP-UX 10.00,
8332 @samp{95} for HP-UX 10.10 though to 11.00, and @samp{98} for HP-UX 11.11
8335 @option{-munix=93} provides the same predefines as GCC 3.3 and 3.4.
8336 @option{-munix=95} provides additional predefines for @code{XOPEN_UNIX}
8337 and @code{_XOPEN_SOURCE_EXTENDED}, and the startfile @file{unix95.o}.
8338 @option{-munix=98} provides additional predefines for @code{_XOPEN_UNIX},
8339 @code{_XOPEN_SOURCE_EXTENDED}, @code{_INCLUDE__STDC_A1_SOURCE} and
8340 @code{_INCLUDE_XOPEN_SOURCE_500}, and the startfile @file{unix98.o}.
8342 It is @emph{important} to note that this option changes the interfaces
8343 for various library routines. It also affects the operational behavior
8344 of the C library. Thus, @emph{extreme} care is needed in using this
8347 Library code that is intended to operate with more than one UNIX
8348 standard must test, set and restore the variable @var{__xpg4_extended_mask}
8349 as appropriate. Most GNU software doesn't provide this capability.
8353 Suppress the generation of link options to search libdld.sl when the
8354 @option{-static} option is specified on HP-UX 10 and later.
8358 The HP-UX implementation of setlocale in libc has a dependency on
8359 libdld.sl. There isn't an archive version of libdld.sl. Thus,
8360 when the @option{-static} option is specified, special link options
8361 are needed to resolve this dependency.
8363 On HP-UX 10 and later, the GCC driver adds the necessary options to
8364 link with libdld.sl when the @option{-static} option is specified.
8365 This causes the resulting binary to be dynamic. On the 64-bit port,
8366 the linkers generate dynamic binaries by default in any case. The
8367 @option{-nolibdld} option can be used to prevent the GCC driver from
8368 adding these link options.
8372 Add support for multithreading with the @dfn{dce thread} library
8373 under HP-UX@. This option sets flags for both the preprocessor and
8377 @node i386 and x86-64 Options
8378 @subsection Intel 386 and AMD x86-64 Options
8379 @cindex i386 Options
8380 @cindex x86-64 Options
8381 @cindex Intel 386 Options
8382 @cindex AMD x86-64 Options
8384 These @samp{-m} options are defined for the i386 and x86-64 family of
8388 @item -mtune=@var{cpu-type}
8390 Tune to @var{cpu-type} everything applicable about the generated code, except
8391 for the ABI and the set of available instructions. The choices for
8395 Original Intel's i386 CPU@.
8397 Intel's i486 CPU@. (No scheduling is implemented for this chip.)
8399 Intel Pentium CPU with no MMX support.
8401 Intel PentiumMMX CPU based on Pentium core with MMX instruction set support.
8402 @item i686, pentiumpro
8403 Intel PentiumPro CPU@.
8405 Intel Pentium2 CPU based on PentiumPro core with MMX instruction set support.
8406 @item pentium3, pentium3m
8407 Intel Pentium3 CPU based on PentiumPro core with MMX and SSE instruction set
8410 Low power version of Intel Pentium3 CPU with MMX, SSE and SSE2 instruction set
8411 support. Used by Centrino notebooks.
8412 @item pentium4, pentium4m
8413 Intel Pentium4 CPU with MMX, SSE and SSE2 instruction set support.
8415 Improved version of Intel Pentium4 CPU with MMX, SSE, SSE2 and SSE3 instruction
8418 Improved version of Intel Pentium4 CPU with 64-bit extensions, MMX, SSE,
8419 SSE2 and SSE3 instruction set support.
8421 AMD K6 CPU with MMX instruction set support.
8423 Improved versions of AMD K6 CPU with MMX and 3dNOW! instruction set support.
8424 @item athlon, athlon-tbird
8425 AMD Athlon CPU with MMX, 3dNOW!, enhanced 3dNOW! and SSE prefetch instructions
8427 @item athlon-4, athlon-xp, athlon-mp
8428 Improved AMD Athlon CPU with MMX, 3dNOW!, enhanced 3dNOW! and full SSE
8429 instruction set support.
8430 @item k8, opteron, athlon64, athlon-fx
8431 AMD K8 core based CPUs with x86-64 instruction set support. (This supersets
8432 MMX, SSE, SSE2, 3dNOW!, enhanced 3dNOW! and 64-bit instruction set extensions.)
8434 IDT Winchip C6 CPU, dealt in same way as i486 with additional MMX instruction
8437 IDT Winchip2 CPU, dealt in same way as i486 with additional MMX and 3dNOW!
8438 instruction set support.
8440 Via C3 CPU with MMX and 3dNOW! instruction set support. (No scheduling is
8441 implemented for this chip.)
8443 Via C3-2 CPU with MMX and SSE instruction set support. (No scheduling is
8444 implemented for this chip.)
8447 While picking a specific @var{cpu-type} will schedule things appropriately
8448 for that particular chip, the compiler will not generate any code that
8449 does not run on the i386 without the @option{-march=@var{cpu-type}} option
8452 @item -march=@var{cpu-type}
8454 Generate instructions for the machine type @var{cpu-type}. The choices
8455 for @var{cpu-type} are the same as for @option{-mtune}. Moreover,
8456 specifying @option{-march=@var{cpu-type}} implies @option{-mtune=@var{cpu-type}}.
8458 @item -mcpu=@var{cpu-type}
8460 A deprecated synonym for @option{-mtune}.
8469 @opindex mpentiumpro
8470 These options are synonyms for @option{-mtune=i386}, @option{-mtune=i486},
8471 @option{-mtune=pentium}, and @option{-mtune=pentiumpro} respectively.
8472 These synonyms are deprecated.
8474 @item -mfpmath=@var{unit}
8476 Generate floating point arithmetics for selected unit @var{unit}. The choices
8481 Use the standard 387 floating point coprocessor present majority of chips and
8482 emulated otherwise. Code compiled with this option will run almost everywhere.
8483 The temporary results are computed in 80bit precision instead of precision
8484 specified by the type resulting in slightly different results compared to most
8485 of other chips. See @option{-ffloat-store} for more detailed description.
8487 This is the default choice for i386 compiler.
8490 Use scalar floating point instructions present in the SSE instruction set.
8491 This instruction set is supported by Pentium3 and newer chips, in the AMD line
8492 by Athlon-4, Athlon-xp and Athlon-mp chips. The earlier version of SSE
8493 instruction set supports only single precision arithmetics, thus the double and
8494 extended precision arithmetics is still done using 387. Later version, present
8495 only in Pentium4 and the future AMD x86-64 chips supports double precision
8498 For the i386 compiler, you need to use @option{-march=@var{cpu-type}}, @option{-msse}
8499 or @option{-msse2} switches to enable SSE extensions and make this option
8500 effective. For the x86-64 compiler, these extensions are enabled by default.
8502 The resulting code should be considerably faster in the majority of cases and avoid
8503 the numerical instability problems of 387 code, but may break some existing
8504 code that expects temporaries to be 80bit.
8506 This is the default choice for the x86-64 compiler.
8509 Attempt to utilize both instruction sets at once. This effectively double the
8510 amount of available registers and on chips with separate execution units for
8511 387 and SSE the execution resources too. Use this option with care, as it is
8512 still experimental, because the GCC register allocator does not model separate
8513 functional units well resulting in instable performance.
8516 @item -masm=@var{dialect}
8517 @opindex masm=@var{dialect}
8518 Output asm instructions using selected @var{dialect}. Supported choices are
8519 @samp{intel} or @samp{att} (the default one).
8524 @opindex mno-ieee-fp
8525 Control whether or not the compiler uses IEEE floating point
8526 comparisons. These handle correctly the case where the result of a
8527 comparison is unordered.
8530 @opindex msoft-float
8531 Generate output containing library calls for floating point.
8532 @strong{Warning:} the requisite libraries are not part of GCC@.
8533 Normally the facilities of the machine's usual C compiler are used, but
8534 this can't be done directly in cross-compilation. You must make your
8535 own arrangements to provide suitable library functions for
8538 On machines where a function returns floating point results in the 80387
8539 register stack, some floating point opcodes may be emitted even if
8540 @option{-msoft-float} is used.
8542 @item -mno-fp-ret-in-387
8543 @opindex mno-fp-ret-in-387
8544 Do not use the FPU registers for return values of functions.
8546 The usual calling convention has functions return values of types
8547 @code{float} and @code{double} in an FPU register, even if there
8548 is no FPU@. The idea is that the operating system should emulate
8551 The option @option{-mno-fp-ret-in-387} causes such values to be returned
8552 in ordinary CPU registers instead.
8554 @item -mno-fancy-math-387
8555 @opindex mno-fancy-math-387
8556 Some 387 emulators do not support the @code{sin}, @code{cos} and
8557 @code{sqrt} instructions for the 387. Specify this option to avoid
8558 generating those instructions. This option is the default on FreeBSD,
8559 OpenBSD and NetBSD@. This option is overridden when @option{-march}
8560 indicates that the target cpu will always have an FPU and so the
8561 instruction will not need emulation. As of revision 2.6.1, these
8562 instructions are not generated unless you also use the
8563 @option{-funsafe-math-optimizations} switch.
8565 @item -malign-double
8566 @itemx -mno-align-double
8567 @opindex malign-double
8568 @opindex mno-align-double
8569 Control whether GCC aligns @code{double}, @code{long double}, and
8570 @code{long long} variables on a two word boundary or a one word
8571 boundary. Aligning @code{double} variables on a two word boundary will
8572 produce code that runs somewhat faster on a @samp{Pentium} at the
8573 expense of more memory.
8575 @strong{Warning:} if you use the @option{-malign-double} switch,
8576 structures containing the above types will be aligned differently than
8577 the published application binary interface specifications for the 386
8578 and will not be binary compatible with structures in code compiled
8579 without that switch.
8581 @item -m96bit-long-double
8582 @itemx -m128bit-long-double
8583 @opindex m96bit-long-double
8584 @opindex m128bit-long-double
8585 These switches control the size of @code{long double} type. The i386
8586 application binary interface specifies the size to be 96 bits,
8587 so @option{-m96bit-long-double} is the default in 32 bit mode.
8589 Modern architectures (Pentium and newer) would prefer @code{long double}
8590 to be aligned to an 8 or 16 byte boundary. In arrays or structures
8591 conforming to the ABI, this would not be possible. So specifying a
8592 @option{-m128bit-long-double} will align @code{long double}
8593 to a 16 byte boundary by padding the @code{long double} with an additional
8596 In the x86-64 compiler, @option{-m128bit-long-double} is the default choice as
8597 its ABI specifies that @code{long double} is to be aligned on 16 byte boundary.
8599 Notice that neither of these options enable any extra precision over the x87
8600 standard of 80 bits for a @code{long double}.
8602 @strong{Warning:} if you override the default value for your target ABI, the
8603 structures and arrays containing @code{long double} variables will change
8604 their size as well as function calling convention for function taking
8605 @code{long double} will be modified. Hence they will not be binary
8606 compatible with arrays or structures in code compiled without that switch.
8610 @itemx -mno-svr3-shlib
8611 @opindex msvr3-shlib
8612 @opindex mno-svr3-shlib
8613 Control whether GCC places uninitialized local variables into the
8614 @code{bss} or @code{data} segments. @option{-msvr3-shlib} places them
8615 into @code{bss}. These options are meaningful only on System V Release 3.
8619 Use a different function-calling convention, in which functions that
8620 take a fixed number of arguments return with the @code{ret} @var{num}
8621 instruction, which pops their arguments while returning. This saves one
8622 instruction in the caller since there is no need to pop the arguments
8625 You can specify that an individual function is called with this calling
8626 sequence with the function attribute @samp{stdcall}. You can also
8627 override the @option{-mrtd} option by using the function attribute
8628 @samp{cdecl}. @xref{Function Attributes}.
8630 @strong{Warning:} this calling convention is incompatible with the one
8631 normally used on Unix, so you cannot use it if you need to call
8632 libraries compiled with the Unix compiler.
8634 Also, you must provide function prototypes for all functions that
8635 take variable numbers of arguments (including @code{printf});
8636 otherwise incorrect code will be generated for calls to those
8639 In addition, seriously incorrect code will result if you call a
8640 function with too many arguments. (Normally, extra arguments are
8641 harmlessly ignored.)
8643 @item -mregparm=@var{num}
8645 Control how many registers are used to pass integer arguments. By
8646 default, no registers are used to pass arguments, and at most 3
8647 registers can be used. You can control this behavior for a specific
8648 function by using the function attribute @samp{regparm}.
8649 @xref{Function Attributes}.
8651 @strong{Warning:} if you use this switch, and
8652 @var{num} is nonzero, then you must build all modules with the same
8653 value, including any libraries. This includes the system libraries and
8656 @item -mpreferred-stack-boundary=@var{num}
8657 @opindex mpreferred-stack-boundary
8658 Attempt to keep the stack boundary aligned to a 2 raised to @var{num}
8659 byte boundary. If @option{-mpreferred-stack-boundary} is not specified,
8660 the default is 4 (16 bytes or 128 bits), except when optimizing for code
8661 size (@option{-Os}), in which case the default is the minimum correct
8662 alignment (4 bytes for x86, and 8 bytes for x86-64).
8664 On Pentium and PentiumPro, @code{double} and @code{long double} values
8665 should be aligned to an 8 byte boundary (see @option{-malign-double}) or
8666 suffer significant run time performance penalties. On Pentium III, the
8667 Streaming SIMD Extension (SSE) data type @code{__m128} suffers similar
8668 penalties if it is not 16 byte aligned.
8670 To ensure proper alignment of this values on the stack, the stack boundary
8671 must be as aligned as that required by any value stored on the stack.
8672 Further, every function must be generated such that it keeps the stack
8673 aligned. Thus calling a function compiled with a higher preferred
8674 stack boundary from a function compiled with a lower preferred stack
8675 boundary will most likely misalign the stack. It is recommended that
8676 libraries that use callbacks always use the default setting.
8678 This extra alignment does consume extra stack space, and generally
8679 increases code size. Code that is sensitive to stack space usage, such
8680 as embedded systems and operating system kernels, may want to reduce the
8681 preferred alignment to @option{-mpreferred-stack-boundary=2}.
8699 These switches enable or disable the use of built-in functions that allow
8700 direct access to the MMX, SSE, SSE2, SSE3 and 3Dnow extensions of the
8703 @xref{X86 Built-in Functions}, for details of the functions enabled
8704 and disabled by these switches.
8706 To have SSE/SSE2 instructions generated automatically from floating-point
8707 code, see @option{-mfpmath=sse}.
8710 @itemx -mno-push-args
8712 @opindex mno-push-args
8713 Use PUSH operations to store outgoing parameters. This method is shorter
8714 and usually equally fast as method using SUB/MOV operations and is enabled
8715 by default. In some cases disabling it may improve performance because of
8716 improved scheduling and reduced dependencies.
8718 @item -maccumulate-outgoing-args
8719 @opindex maccumulate-outgoing-args
8720 If enabled, the maximum amount of space required for outgoing arguments will be
8721 computed in the function prologue. This is faster on most modern CPUs
8722 because of reduced dependencies, improved scheduling and reduced stack usage
8723 when preferred stack boundary is not equal to 2. The drawback is a notable
8724 increase in code size. This switch implies @option{-mno-push-args}.
8728 Support thread-safe exception handling on @samp{Mingw32}. Code that relies
8729 on thread-safe exception handling must compile and link all code with the
8730 @option{-mthreads} option. When compiling, @option{-mthreads} defines
8731 @option{-D_MT}; when linking, it links in a special thread helper library
8732 @option{-lmingwthrd} which cleans up per thread exception handling data.
8734 @item -mno-align-stringops
8735 @opindex mno-align-stringops
8736 Do not align destination of inlined string operations. This switch reduces
8737 code size and improves performance in case the destination is already aligned,
8738 but GCC doesn't know about it.
8740 @item -minline-all-stringops
8741 @opindex minline-all-stringops
8742 By default GCC inlines string operations only when destination is known to be
8743 aligned at least to 4 byte boundary. This enables more inlining, increase code
8744 size, but may improve performance of code that depends on fast memcpy, strlen
8745 and memset for short lengths.
8747 @item -momit-leaf-frame-pointer
8748 @opindex momit-leaf-frame-pointer
8749 Don't keep the frame pointer in a register for leaf functions. This
8750 avoids the instructions to save, set up and restore frame pointers and
8751 makes an extra register available in leaf functions. The option
8752 @option{-fomit-frame-pointer} removes the frame pointer for all functions
8753 which might make debugging harder.
8755 @item -mtls-direct-seg-refs
8756 @itemx -mno-tls-direct-seg-refs
8757 @opindex mtls-direct-seg-refs
8758 Controls whether TLS variables may be accessed with offsets from the
8759 TLS segment register (@code{%gs} for 32-bit, @code{%fs} for 64-bit),
8760 or whether the thread base pointer must be added. Whether or not this
8761 is legal depends on the operating system, and whether it maps the
8762 segment to cover the entire TLS area.
8764 For systems that use GNU libc, the default is on.
8767 These @samp{-m} switches are supported in addition to the above
8768 on AMD x86-64 processors in 64-bit environments.
8775 Generate code for a 32-bit or 64-bit environment.
8776 The 32-bit environment sets int, long and pointer to 32 bits and
8777 generates code that runs on any i386 system.
8778 The 64-bit environment sets int to 32 bits and long and pointer
8779 to 64 bits and generates code for AMD's x86-64 architecture.
8782 @opindex no-red-zone
8783 Do not use a so called red zone for x86-64 code. The red zone is mandated
8784 by the x86-64 ABI, it is a 128-byte area beyond the location of the
8785 stack pointer that will not be modified by signal or interrupt handlers
8786 and therefore can be used for temporary data without adjusting the stack
8787 pointer. The flag @option{-mno-red-zone} disables this red zone.
8789 @item -mcmodel=small
8790 @opindex mcmodel=small
8791 Generate code for the small code model: the program and its symbols must
8792 be linked in the lower 2 GB of the address space. Pointers are 64 bits.
8793 Programs can be statically or dynamically linked. This is the default
8796 @item -mcmodel=kernel
8797 @opindex mcmodel=kernel
8798 Generate code for the kernel code model. The kernel runs in the
8799 negative 2 GB of the address space.
8800 This model has to be used for Linux kernel code.
8802 @item -mcmodel=medium
8803 @opindex mcmodel=medium
8804 Generate code for the medium model: The program is linked in the lower 2
8805 GB of the address space but symbols can be located anywhere in the
8806 address space. Programs can be statically or dynamically linked, but
8807 building of shared libraries are not supported with the medium model.
8809 @item -mcmodel=large
8810 @opindex mcmodel=large
8811 Generate code for the large model: This model makes no assumptions
8812 about addresses and sizes of sections. Currently GCC does not implement
8817 @subsection IA-64 Options
8818 @cindex IA-64 Options
8820 These are the @samp{-m} options defined for the Intel IA-64 architecture.
8824 @opindex mbig-endian
8825 Generate code for a big endian target. This is the default for HP-UX@.
8827 @item -mlittle-endian
8828 @opindex mlittle-endian
8829 Generate code for a little endian target. This is the default for AIX5
8836 Generate (or don't) code for the GNU assembler. This is the default.
8837 @c Also, this is the default if the configure option @option{--with-gnu-as}
8844 Generate (or don't) code for the GNU linker. This is the default.
8845 @c Also, this is the default if the configure option @option{--with-gnu-ld}
8850 Generate code that does not use a global pointer register. The result
8851 is not position independent code, and violates the IA-64 ABI@.
8853 @item -mvolatile-asm-stop
8854 @itemx -mno-volatile-asm-stop
8855 @opindex mvolatile-asm-stop
8856 @opindex mno-volatile-asm-stop
8857 Generate (or don't) a stop bit immediately before and after volatile asm
8862 Generate code that works around Itanium B step errata.
8864 @item -mregister-names
8865 @itemx -mno-register-names
8866 @opindex mregister-names
8867 @opindex mno-register-names
8868 Generate (or don't) @samp{in}, @samp{loc}, and @samp{out} register names for
8869 the stacked registers. This may make assembler output more readable.
8875 Disable (or enable) optimizations that use the small data section. This may
8876 be useful for working around optimizer bugs.
8879 @opindex mconstant-gp
8880 Generate code that uses a single constant global pointer value. This is
8881 useful when compiling kernel code.
8885 Generate code that is self-relocatable. This implies @option{-mconstant-gp}.
8886 This is useful when compiling firmware code.
8888 @item -minline-float-divide-min-latency
8889 @opindex minline-float-divide-min-latency
8890 Generate code for inline divides of floating point values
8891 using the minimum latency algorithm.
8893 @item -minline-float-divide-max-throughput
8894 @opindex minline-float-divide-max-throughput
8895 Generate code for inline divides of floating point values
8896 using the maximum throughput algorithm.
8898 @item -minline-int-divide-min-latency
8899 @opindex minline-int-divide-min-latency
8900 Generate code for inline divides of integer values
8901 using the minimum latency algorithm.
8903 @item -minline-int-divide-max-throughput
8904 @opindex minline-int-divide-max-throughput
8905 Generate code for inline divides of integer values
8906 using the maximum throughput algorithm.
8908 @item -minline-sqrt-min-latency
8909 @opindex minline-sqrt-min-latency
8910 Generate code for inline square roots
8911 using the minimum latency algorithm.
8913 @item -minline-sqrt-max-throughput
8914 @opindex minline-sqrt-max-throughput
8915 Generate code for inline square roots
8916 using the maximum throughput algorithm.
8918 @item -mno-dwarf2-asm
8920 @opindex mno-dwarf2-asm
8921 @opindex mdwarf2-asm
8922 Don't (or do) generate assembler code for the DWARF2 line number debugging
8923 info. This may be useful when not using the GNU assembler.
8925 @item -mearly-stop-bits
8926 @itemx -mno-early-stop-bits
8927 @opindex mearly-stop-bits
8928 @opindex mno-early-stop-bits
8929 Allow stop bits to be placed earlier than immediately preceding the
8930 instruction that triggered the stop bit. This can improve instruction
8931 scheduling, but does not always do so.
8933 @item -mfixed-range=@var{register-range}
8934 @opindex mfixed-range
8935 Generate code treating the given register range as fixed registers.
8936 A fixed register is one that the register allocator can not use. This is
8937 useful when compiling kernel code. A register range is specified as
8938 two registers separated by a dash. Multiple register ranges can be
8939 specified separated by a comma.
8941 @item -mtls-size=@var{tls-size}
8943 Specify bit size of immediate TLS offsets. Valid values are 14, 22, and
8946 @item -mtune-arch=@var{cpu-type}
8948 Tune the instruction scheduling for a particular CPU, Valid values are
8949 itanium, itanium1, merced, itanium2, and mckinley.
8955 Add support for multithreading using the POSIX threads library. This
8956 option sets flags for both the preprocessor and linker. It does
8957 not affect the thread safety of object code produced by the compiler or
8958 that of libraries supplied with it. These are HP-UX specific flags.
8964 Generate code for a 32-bit or 64-bit environment.
8965 The 32-bit environment sets int, long and pointer to 32 bits.
8966 The 64-bit environment sets int to 32 bits and long and pointer
8967 to 64 bits. These are HP-UX specific flags.
8971 @node M32R/D Options
8972 @subsection M32R/D Options
8973 @cindex M32R/D options
8975 These @option{-m} options are defined for Renesas M32R/D architectures:
8980 Generate code for the M32R/2@.
8984 Generate code for the M32R/X@.
8988 Generate code for the M32R@. This is the default.
8991 @opindex mmodel=small
8992 Assume all objects live in the lower 16MB of memory (so that their addresses
8993 can be loaded with the @code{ld24} instruction), and assume all subroutines
8994 are reachable with the @code{bl} instruction.
8995 This is the default.
8997 The addressability of a particular object can be set with the
8998 @code{model} attribute.
9000 @item -mmodel=medium
9001 @opindex mmodel=medium
9002 Assume objects may be anywhere in the 32-bit address space (the compiler
9003 will generate @code{seth/add3} instructions to load their addresses), and
9004 assume all subroutines are reachable with the @code{bl} instruction.
9007 @opindex mmodel=large
9008 Assume objects may be anywhere in the 32-bit address space (the compiler
9009 will generate @code{seth/add3} instructions to load their addresses), and
9010 assume subroutines may not be reachable with the @code{bl} instruction
9011 (the compiler will generate the much slower @code{seth/add3/jl}
9012 instruction sequence).
9015 @opindex msdata=none
9016 Disable use of the small data area. Variables will be put into
9017 one of @samp{.data}, @samp{bss}, or @samp{.rodata} (unless the
9018 @code{section} attribute has been specified).
9019 This is the default.
9021 The small data area consists of sections @samp{.sdata} and @samp{.sbss}.
9022 Objects may be explicitly put in the small data area with the
9023 @code{section} attribute using one of these sections.
9026 @opindex msdata=sdata
9027 Put small global and static data in the small data area, but do not
9028 generate special code to reference them.
9032 Put small global and static data in the small data area, and generate
9033 special instructions to reference them.
9037 @cindex smaller data references
9038 Put global and static objects less than or equal to @var{num} bytes
9039 into the small data or bss sections instead of the normal data or bss
9040 sections. The default value of @var{num} is 8.
9041 The @option{-msdata} option must be set to one of @samp{sdata} or @samp{use}
9042 for this option to have any effect.
9044 All modules should be compiled with the same @option{-G @var{num}} value.
9045 Compiling with different values of @var{num} may or may not work; if it
9046 doesn't the linker will give an error message---incorrect code will not be
9051 Makes the M32R specific code in the compiler display some statistics
9052 that might help in debugging programs.
9055 @opindex malign-loops
9056 Align all loops to a 32-byte boundary.
9058 @item -mno-align-loops
9059 @opindex mno-align-loops
9060 Do not enforce a 32-byte alignment for loops. This is the default.
9062 @item -missue-rate=@var{number}
9063 @opindex missue-rate=@var{number}
9064 Issue @var{number} instructions per cycle. @var{number} can only be 1
9067 @item -mbranch-cost=@var{number}
9068 @opindex mbranch-cost=@var{number}
9069 @var{number} can only be 1 or 2. If it is 1 then branches will be
9070 preferred over conditional code, if it is 2, then the opposite will
9073 @item -mflush-trap=@var{number}
9074 @opindex mflush-trap=@var{number}
9075 Specifies the trap number to use to flush the cache. The default is
9076 12. Valid numbers are between 0 and 15 inclusive.
9078 @item -mno-flush-trap
9079 @opindex mno-flush-trap
9080 Specifies that the cache cannot be flushed by using a trap.
9082 @item -mflush-func=@var{name}
9083 @opindex mflush-func=@var{name}
9084 Specifies the name of the operating system function to call to flush
9085 the cache. The default is @emph{_flush_cache}, but a function call
9086 will only be used if a trap is not available.
9088 @item -mno-flush-func
9089 @opindex mno-flush-func
9090 Indicates that there is no OS function for flushing the cache.
9094 @node M680x0 Options
9095 @subsection M680x0 Options
9096 @cindex M680x0 options
9098 These are the @samp{-m} options defined for the 68000 series. The default
9099 values for these options depends on which style of 68000 was selected when
9100 the compiler was configured; the defaults for the most common choices are
9108 Generate output for a 68000. This is the default
9109 when the compiler is configured for 68000-based systems.
9111 Use this option for microcontrollers with a 68000 or EC000 core,
9112 including the 68008, 68302, 68306, 68307, 68322, 68328 and 68356.
9118 Generate output for a 68020. This is the default
9119 when the compiler is configured for 68020-based systems.
9123 Generate output containing 68881 instructions for floating point.
9124 This is the default for most 68020 systems unless @option{--nfp} was
9125 specified when the compiler was configured.
9129 Generate output for a 68030. This is the default when the compiler is
9130 configured for 68030-based systems.
9134 Generate output for a 68040. This is the default when the compiler is
9135 configured for 68040-based systems.
9137 This option inhibits the use of 68881/68882 instructions that have to be
9138 emulated by software on the 68040. Use this option if your 68040 does not
9139 have code to emulate those instructions.
9143 Generate output for a 68060. This is the default when the compiler is
9144 configured for 68060-based systems.
9146 This option inhibits the use of 68020 and 68881/68882 instructions that
9147 have to be emulated by software on the 68060. Use this option if your 68060
9148 does not have code to emulate those instructions.
9152 Generate output for a CPU32. This is the default
9153 when the compiler is configured for CPU32-based systems.
9155 Use this option for microcontrollers with a
9156 CPU32 or CPU32+ core, including the 68330, 68331, 68332, 68333, 68334,
9157 68336, 68340, 68341, 68349 and 68360.
9161 Generate output for a 520X ``coldfire'' family cpu. This is the default
9162 when the compiler is configured for 520X-based systems.
9164 Use this option for microcontroller with a 5200 core, including
9165 the MCF5202, MCF5203, MCF5204 and MCF5202.
9170 Generate output for a 68040, without using any of the new instructions.
9171 This results in code which can run relatively efficiently on either a
9172 68020/68881 or a 68030 or a 68040. The generated code does use the
9173 68881 instructions that are emulated on the 68040.
9177 Generate output for a 68060, without using any of the new instructions.
9178 This results in code which can run relatively efficiently on either a
9179 68020/68881 or a 68030 or a 68040. The generated code does use the
9180 68881 instructions that are emulated on the 68060.
9183 @opindex msoft-float
9184 Generate output containing library calls for floating point.
9185 @strong{Warning:} the requisite libraries are not available for all m68k
9186 targets. Normally the facilities of the machine's usual C compiler are
9187 used, but this can't be done directly in cross-compilation. You must
9188 make your own arrangements to provide suitable library functions for
9189 cross-compilation. The embedded targets @samp{m68k-*-aout} and
9190 @samp{m68k-*-coff} do provide software floating point support.
9194 Consider type @code{int} to be 16 bits wide, like @code{short int}.
9195 Additionally, parameters passed on the stack are also aligned to a
9196 16-bit boundary even on targets whose API mandates promotion to 32-bit.
9199 @opindex mnobitfield
9200 Do not use the bit-field instructions. The @option{-m68000}, @option{-mcpu32}
9201 and @option{-m5200} options imply @w{@option{-mnobitfield}}.
9205 Do use the bit-field instructions. The @option{-m68020} option implies
9206 @option{-mbitfield}. This is the default if you use a configuration
9207 designed for a 68020.
9211 Use a different function-calling convention, in which functions
9212 that take a fixed number of arguments return with the @code{rtd}
9213 instruction, which pops their arguments while returning. This
9214 saves one instruction in the caller since there is no need to pop
9215 the arguments there.
9217 This calling convention is incompatible with the one normally
9218 used on Unix, so you cannot use it if you need to call libraries
9219 compiled with the Unix compiler.
9221 Also, you must provide function prototypes for all functions that
9222 take variable numbers of arguments (including @code{printf});
9223 otherwise incorrect code will be generated for calls to those
9226 In addition, seriously incorrect code will result if you call a
9227 function with too many arguments. (Normally, extra arguments are
9228 harmlessly ignored.)
9230 The @code{rtd} instruction is supported by the 68010, 68020, 68030,
9231 68040, 68060 and CPU32 processors, but not by the 68000 or 5200.
9234 @itemx -mno-align-int
9236 @opindex mno-align-int
9237 Control whether GCC aligns @code{int}, @code{long}, @code{long long},
9238 @code{float}, @code{double}, and @code{long double} variables on a 32-bit
9239 boundary (@option{-malign-int}) or a 16-bit boundary (@option{-mno-align-int}).
9240 Aligning variables on 32-bit boundaries produces code that runs somewhat
9241 faster on processors with 32-bit busses at the expense of more memory.
9243 @strong{Warning:} if you use the @option{-malign-int} switch, GCC will
9244 align structures containing the above types differently than
9245 most published application binary interface specifications for the m68k.
9249 Use the pc-relative addressing mode of the 68000 directly, instead of
9250 using a global offset table. At present, this option implies @option{-fpic},
9251 allowing at most a 16-bit offset for pc-relative addressing. @option{-fPIC} is
9252 not presently supported with @option{-mpcrel}, though this could be supported for
9253 68020 and higher processors.
9255 @item -mno-strict-align
9256 @itemx -mstrict-align
9257 @opindex mno-strict-align
9258 @opindex mstrict-align
9259 Do not (do) assume that unaligned memory references will be handled by
9263 Generate code that allows the data segment to be located in a different
9264 area of memory from the text segment. This allows for execute in place in
9265 an environment without virtual memory management. This option implies
9269 Generate code that assumes that the data segment follows the text segment.
9270 This is the default.
9272 @item -mid-shared-library
9273 Generate code that supports shared libraries via the library ID method.
9274 This allows for execute in place and shared libraries in an environment
9275 without virtual memory management. This option implies @option{-fPIC}.
9277 @item -mno-id-shared-library
9278 Generate code that doesn't assume ID based shared libraries are being used.
9279 This is the default.
9281 @item -mshared-library-id=n
9282 Specified the identification number of the ID based shared library being
9283 compiled. Specifying a value of 0 will generate more compact code, specifying
9284 other values will force the allocation of that number to the current
9285 library but is no more space or time efficient than omitting this option.
9289 @node M68hc1x Options
9290 @subsection M68hc1x Options
9291 @cindex M68hc1x options
9293 These are the @samp{-m} options defined for the 68hc11 and 68hc12
9294 microcontrollers. The default values for these options depends on
9295 which style of microcontroller was selected when the compiler was configured;
9296 the defaults for the most common choices are given below.
9303 Generate output for a 68HC11. This is the default
9304 when the compiler is configured for 68HC11-based systems.
9310 Generate output for a 68HC12. This is the default
9311 when the compiler is configured for 68HC12-based systems.
9317 Generate output for a 68HCS12.
9320 @opindex mauto-incdec
9321 Enable the use of 68HC12 pre and post auto-increment and auto-decrement
9328 Enable the use of 68HC12 min and max instructions.
9331 @itemx -mno-long-calls
9332 @opindex mlong-calls
9333 @opindex mno-long-calls
9334 Treat all calls as being far away (near). If calls are assumed to be
9335 far away, the compiler will use the @code{call} instruction to
9336 call a function and the @code{rtc} instruction for returning.
9340 Consider type @code{int} to be 16 bits wide, like @code{short int}.
9342 @item -msoft-reg-count=@var{count}
9343 @opindex msoft-reg-count
9344 Specify the number of pseudo-soft registers which are used for the
9345 code generation. The maximum number is 32. Using more pseudo-soft
9346 register may or may not result in better code depending on the program.
9347 The default is 4 for 68HC11 and 2 for 68HC12.
9352 @subsection MCore Options
9353 @cindex MCore options
9355 These are the @samp{-m} options defined for the Motorola M*Core
9363 @opindex mno-hardlit
9364 Inline constants into the code stream if it can be done in two
9365 instructions or less.
9371 Use the divide instruction. (Enabled by default).
9373 @item -mrelax-immediate
9374 @itemx -mno-relax-immediate
9375 @opindex mrelax-immediate
9376 @opindex mno-relax-immediate
9377 Allow arbitrary sized immediates in bit operations.
9379 @item -mwide-bitfields
9380 @itemx -mno-wide-bitfields
9381 @opindex mwide-bitfields
9382 @opindex mno-wide-bitfields
9383 Always treat bit-fields as int-sized.
9385 @item -m4byte-functions
9386 @itemx -mno-4byte-functions
9387 @opindex m4byte-functions
9388 @opindex mno-4byte-functions
9389 Force all functions to be aligned to a four byte boundary.
9391 @item -mcallgraph-data
9392 @itemx -mno-callgraph-data
9393 @opindex mcallgraph-data
9394 @opindex mno-callgraph-data
9395 Emit callgraph information.
9398 @itemx -mno-slow-bytes
9399 @opindex mslow-bytes
9400 @opindex mno-slow-bytes
9401 Prefer word access when reading byte quantities.
9403 @item -mlittle-endian
9405 @opindex mlittle-endian
9406 @opindex mbig-endian
9407 Generate code for a little endian target.
9413 Generate code for the 210 processor.
9417 @subsection MIPS Options
9418 @cindex MIPS options
9424 Generate big-endian code.
9428 Generate little-endian code. This is the default for @samp{mips*el-*-*}
9431 @item -march=@var{arch}
9433 Generate code that will run on @var{arch}, which can be the name of a
9434 generic MIPS ISA, or the name of a particular processor.
9436 @samp{mips1}, @samp{mips2}, @samp{mips3}, @samp{mips4},
9437 @samp{mips32}, @samp{mips32r2}, and @samp{mips64}.
9438 The processor names are:
9439 @samp{4kc}, @samp{4kp}, @samp{5kc}, @samp{20kc},
9441 @samp{r2000}, @samp{r3000}, @samp{r3900}, @samp{r4000}, @samp{r4400},
9442 @samp{r4600}, @samp{r4650}, @samp{r6000}, @samp{r8000}, @samp{rm7000},
9446 @samp{vr4100}, @samp{vr4111}, @samp{vr4120}, @samp{vr4130}, @samp{vr4300},
9447 @samp{vr5000}, @samp{vr5400} and @samp{vr5500}.
9448 The special value @samp{from-abi} selects the
9449 most compatible architecture for the selected ABI (that is,
9450 @samp{mips1} for 32-bit ABIs and @samp{mips3} for 64-bit ABIs)@.
9452 In processor names, a final @samp{000} can be abbreviated as @samp{k}
9453 (for example, @samp{-march=r2k}). Prefixes are optional, and
9454 @samp{vr} may be written @samp{r}.
9456 GCC defines two macros based on the value of this option. The first
9457 is @samp{_MIPS_ARCH}, which gives the name of target architecture, as
9458 a string. The second has the form @samp{_MIPS_ARCH_@var{foo}},
9459 where @var{foo} is the capitalized value of @samp{_MIPS_ARCH}@.
9460 For example, @samp{-march=r2000} will set @samp{_MIPS_ARCH}
9461 to @samp{"r2000"} and define the macro @samp{_MIPS_ARCH_R2000}.
9463 Note that the @samp{_MIPS_ARCH} macro uses the processor names given
9464 above. In other words, it will have the full prefix and will not
9465 abbreviate @samp{000} as @samp{k}. In the case of @samp{from-abi},
9466 the macro names the resolved architecture (either @samp{"mips1"} or
9467 @samp{"mips3"}). It names the default architecture when no
9468 @option{-march} option is given.
9470 @item -mtune=@var{arch}
9472 Optimize for @var{arch}. Among other things, this option controls
9473 the way instructions are scheduled, and the perceived cost of arithmetic
9474 operations. The list of @var{arch} values is the same as for
9477 When this option is not used, GCC will optimize for the processor
9478 specified by @option{-march}. By using @option{-march} and
9479 @option{-mtune} together, it is possible to generate code that will
9480 run on a family of processors, but optimize the code for one
9481 particular member of that family.
9483 @samp{-mtune} defines the macros @samp{_MIPS_TUNE} and
9484 @samp{_MIPS_TUNE_@var{foo}}, which work in the same way as the
9485 @samp{-march} ones described above.
9489 Equivalent to @samp{-march=mips1}.
9493 Equivalent to @samp{-march=mips2}.
9497 Equivalent to @samp{-march=mips3}.
9501 Equivalent to @samp{-march=mips4}.
9505 Equivalent to @samp{-march=mips32}.
9509 Equivalent to @samp{-march=mips32r2}.
9513 Equivalent to @samp{-march=mips64}.
9519 Use (do not use) the MIPS16 ISA@.
9531 Generate code for the given ABI@.
9533 Note that the EABI has a 32-bit and a 64-bit variant. GCC normally
9534 generates 64-bit code when you select a 64-bit architecture, but you
9535 can use @option{-mgp32} to get 32-bit code instead.
9537 For information about the O64 ABI, see
9538 @w{@uref{http://gcc.gnu.org/projects/mipso64-abi.html}}.
9541 @itemx -mno-abicalls
9543 @opindex mno-abicalls
9544 Generate (do not generate) SVR4-style position-independent code.
9545 @option{-mabicalls} is the default for SVR4-based systems.
9551 Lift (do not lift) the usual restrictions on the size of the global
9554 GCC normally uses a single instruction to load values from the GOT@.
9555 While this is relatively efficient, it will only work if the GOT
9556 is smaller than about 64k. Anything larger will cause the linker
9557 to report an error such as:
9559 @cindex relocation truncated to fit (MIPS)
9561 relocation truncated to fit: R_MIPS_GOT16 foobar
9564 If this happens, you should recompile your code with @option{-mxgot}.
9565 It should then work with very large GOTs, although it will also be
9566 less efficient, since it will take three instructions to fetch the
9567 value of a global symbol.
9569 Note that some linkers can create multiple GOTs. If you have such a
9570 linker, you should only need to use @option{-mxgot} when a single object
9571 file accesses more than 64k's worth of GOT entries. Very few do.
9573 These options have no effect unless GCC is generating position
9578 Assume that general-purpose registers are 32 bits wide.
9582 Assume that general-purpose registers are 64 bits wide.
9586 Assume that floating-point registers are 32 bits wide.
9590 Assume that floating-point registers are 64 bits wide.
9593 @opindex mhard-float
9594 Use floating-point coprocessor instructions.
9597 @opindex msoft-float
9598 Do not use floating-point coprocessor instructions. Implement
9599 floating-point calculations using library calls instead.
9601 @item -msingle-float
9602 @opindex msingle-float
9603 Assume that the floating-point coprocessor only supports single-precision
9606 @itemx -mdouble-float
9607 @opindex mdouble-float
9608 Assume that the floating-point coprocessor supports double-precision
9609 operations. This is the default.
9611 @itemx -mpaired-single
9612 @itemx -mno-paired-single
9613 @opindex mpaired-single
9614 @opindex mno-paired-single
9615 Use (do not use) paired-single floating-point instructions.
9616 @xref{MIPS Paired-Single Support}. This option can only be used
9617 when generating 64-bit code and requires hardware floating-point
9618 support to be enabled.
9624 Use (do not use) the MIPS-3D ASE@. @xref{MIPS-3D Built-in Functions}.
9625 The option @option{-mips3d} implies @option{-mpaired-single}.
9629 Force @code{int} and @code{long} types to be 64 bits wide. See
9630 @option{-mlong32} for an explanation of the default and the way
9631 that the pointer size is determined.
9633 This option has been deprecated and will be removed in a future release.
9637 Force @code{long} types to be 64 bits wide. See @option{-mlong32} for
9638 an explanation of the default and the way that the pointer size is
9643 Force @code{long}, @code{int}, and pointer types to be 32 bits wide.
9645 The default size of @code{int}s, @code{long}s and pointers depends on
9646 the ABI@. All the supported ABIs use 32-bit @code{int}s. The n64 ABI
9647 uses 64-bit @code{long}s, as does the 64-bit EABI; the others use
9648 32-bit @code{long}s. Pointers are the same size as @code{long}s,
9649 or the same size as integer registers, whichever is smaller.
9653 @cindex smaller data references (MIPS)
9654 @cindex gp-relative references (MIPS)
9655 Put global and static items less than or equal to @var{num} bytes into
9656 the small data or bss section instead of the normal data or bss section.
9657 This allows the data to be accessed using a single instruction.
9659 All modules should be compiled with the same @option{-G @var{num}}
9662 @item -membedded-data
9663 @itemx -mno-embedded-data
9664 @opindex membedded-data
9665 @opindex mno-embedded-data
9666 Allocate variables to the read-only data section first if possible, then
9667 next in the small data section if possible, otherwise in data. This gives
9668 slightly slower code than the default, but reduces the amount of RAM required
9669 when executing, and thus may be preferred for some embedded systems.
9671 @item -muninit-const-in-rodata
9672 @itemx -mno-uninit-const-in-rodata
9673 @opindex muninit-const-in-rodata
9674 @opindex mno-uninit-const-in-rodata
9675 Put uninitialized @code{const} variables in the read-only data section.
9676 This option is only meaningful in conjunction with @option{-membedded-data}.
9678 @item -msplit-addresses
9679 @itemx -mno-split-addresses
9680 @opindex msplit-addresses
9681 @opindex mno-split-addresses
9682 Enable (disable) use of the @code{%hi()} and @code{%lo()} assembler
9683 relocation operators. This option has been superceded by
9684 @option{-mexplicit-relocs} but is retained for backwards compatibility.
9686 @item -mexplicit-relocs
9687 @itemx -mno-explicit-relocs
9688 @opindex mexplicit-relocs
9689 @opindex mno-explicit-relocs
9690 Use (do not use) assembler relocation operators when dealing with symbolic
9691 addresses. The alternative, selected by @option{-mno-explicit-relocs},
9692 is to use assembler macros instead.
9694 @option{-mexplicit-relocs} is the default if GCC was configured
9695 to use an assembler that supports relocation operators.
9697 @item -mcheck-zero-division
9698 @itemx -mno-check-zero-division
9699 @opindex mcheck-zero-division
9700 @opindex mno-check-zero-division
9701 Trap (do not trap) on integer division by zero. The default is
9702 @option{-mcheck-zero-division}.
9704 @item -mdivide-traps
9705 @itemx -mdivide-breaks
9706 @opindex mdivide-traps
9707 @opindex mdivide-breaks
9708 MIPS systems check for division by zero by generating either a
9709 conditional trap or a break instruction. Using traps results in
9710 smaller code, but is only supported on MIPS II and later. Also, some
9711 versions of the Linux kernel have a bug that prevents trap from
9712 generating the proper signal (@code{SIGFPE}). Use @option{-mdivide-traps} to
9713 allow conditional traps on architectures that support them and
9714 @option{-mdivide-breaks} to force the use of breaks.
9716 The default is usually @option{-mdivide-traps}, but this can be
9717 overridden at configure time using @option{--with-divide=breaks}.
9718 Divide-by-zero checks can be completely disabled using
9719 @option{-mno-check-zero-division}.
9725 Force (do not force) the use of @code{memcpy()} for non-trivial block
9726 moves. The default is @option{-mno-memcpy}, which allows GCC to inline
9727 most constant-sized copies.
9730 @itemx -mno-long-calls
9731 @opindex mlong-calls
9732 @opindex mno-long-calls
9733 Disable (do not disable) use of the @code{jal} instruction. Calling
9734 functions using @code{jal} is more efficient but requires the caller
9735 and callee to be in the same 256 megabyte segment.
9737 This option has no effect on abicalls code. The default is
9738 @option{-mno-long-calls}.
9744 Enable (disable) use of the @code{mad}, @code{madu} and @code{mul}
9745 instructions, as provided by the R4650 ISA@.
9748 @itemx -mno-fused-madd
9749 @opindex mfused-madd
9750 @opindex mno-fused-madd
9751 Enable (disable) use of the floating point multiply-accumulate
9752 instructions, when they are available. The default is
9753 @option{-mfused-madd}.
9755 When multiply-accumulate instructions are used, the intermediate
9756 product is calculated to infinite precision and is not subject to
9757 the FCSR Flush to Zero bit. This may be undesirable in some
9762 Tell the MIPS assembler to not run its preprocessor over user
9763 assembler files (with a @samp{.s} suffix) when assembling them.
9766 @itemx -mno-fix-r4000
9768 @opindex mno-fix-r4000
9769 Work around certain R4000 CPU errata:
9772 A double-word or a variable shift may give an incorrect result if executed
9773 immediately after starting an integer division.
9775 A double-word or a variable shift may give an incorrect result if executed
9776 while an integer multiplication is in progress.
9778 An integer division may give an incorrect result if started in a delay slot
9779 of a taken branch or a jump.
9783 @itemx -mno-fix-r4400
9785 @opindex mno-fix-r4400
9786 Work around certain R4400 CPU errata:
9789 A double-word or a variable shift may give an incorrect result if executed
9790 immediately after starting an integer division.
9794 @itemx -mno-fix-vr4120
9795 @opindex mfix-vr4120
9796 Work around certain VR4120 errata:
9799 @code{dmultu} does not always produce the correct result.
9801 @code{div} and @code{ddiv} do not always produce the correct result if one
9802 of the operands is negative.
9804 The workarounds for the division errata rely on special functions in
9805 @file{libgcc.a}. At present, these functions are only provided by
9806 the @code{mips64vr*-elf} configurations.
9808 Other VR4120 errata require a nop to be inserted between certain pairs of
9809 instructions. These errata are handled by the assembler, not by GCC itself.
9814 Work around certain SB-1 CPU core errata.
9815 (This flag currently works around the SB-1 revision 2
9816 ``F1'' and ``F2'' floating point errata.)
9818 @item -mflush-func=@var{func}
9819 @itemx -mno-flush-func
9820 @opindex mflush-func
9821 Specifies the function to call to flush the I and D caches, or to not
9822 call any such function. If called, the function must take the same
9823 arguments as the common @code{_flush_func()}, that is, the address of the
9824 memory range for which the cache is being flushed, the size of the
9825 memory range, and the number 3 (to flush both caches). The default
9826 depends on the target GCC was configured for, but commonly is either
9827 @samp{_flush_func} or @samp{__cpu_flush}.
9829 @item -mbranch-likely
9830 @itemx -mno-branch-likely
9831 @opindex mbranch-likely
9832 @opindex mno-branch-likely
9833 Enable or disable use of Branch Likely instructions, regardless of the
9834 default for the selected architecture. By default, Branch Likely
9835 instructions may be generated if they are supported by the selected
9836 architecture. An exception is for the MIPS32 and MIPS64 architectures
9837 and processors which implement those architectures; for those, Branch
9838 Likely instructions will not be generated by default because the MIPS32
9839 and MIPS64 architectures specifically deprecate their use.
9841 @item -mfp-exceptions
9842 @itemx -mno-fp-exceptions
9843 @opindex mfp-exceptions
9844 Specifies whether FP exceptions are enabled. This affects how we schedule
9845 FP instructions for some processors. The default is that FP exceptions are
9848 For instance, on the SB-1, if FP exceptions are disabled, and we are emitting
9849 64-bit code, then we can use both FP pipes. Otherwise, we can only use one
9852 @item -mvr4130-align
9853 @itemx -mno-vr4130-align
9854 @opindex mvr4130-align
9855 The VR4130 pipeline is two-way superscalar, but can only issue two
9856 instructions together if the first one is 8-byte aligned. When this
9857 option is enabled, GCC will align pairs of instructions that it
9858 thinks should execute in parallel.
9860 This option only has an effect when optimizing for the VR4130.
9861 It normally makes code faster, but at the expense of making it bigger.
9862 It is enabled by default at optimization level @option{-O3}.
9866 @subsection MMIX Options
9867 @cindex MMIX Options
9869 These options are defined for the MMIX:
9873 @itemx -mno-libfuncs
9875 @opindex mno-libfuncs
9876 Specify that intrinsic library functions are being compiled, passing all
9877 values in registers, no matter the size.
9882 @opindex mno-epsilon
9883 Generate floating-point comparison instructions that compare with respect
9884 to the @code{rE} epsilon register.
9886 @item -mabi=mmixware
9888 @opindex mabi-mmixware
9890 Generate code that passes function parameters and return values that (in
9891 the called function) are seen as registers @code{$0} and up, as opposed to
9892 the GNU ABI which uses global registers @code{$231} and up.
9895 @itemx -mno-zero-extend
9896 @opindex mzero-extend
9897 @opindex mno-zero-extend
9898 When reading data from memory in sizes shorter than 64 bits, use (do not
9899 use) zero-extending load instructions by default, rather than
9900 sign-extending ones.
9903 @itemx -mno-knuthdiv
9905 @opindex mno-knuthdiv
9906 Make the result of a division yielding a remainder have the same sign as
9907 the divisor. With the default, @option{-mno-knuthdiv}, the sign of the
9908 remainder follows the sign of the dividend. Both methods are
9909 arithmetically valid, the latter being almost exclusively used.
9911 @item -mtoplevel-symbols
9912 @itemx -mno-toplevel-symbols
9913 @opindex mtoplevel-symbols
9914 @opindex mno-toplevel-symbols
9915 Prepend (do not prepend) a @samp{:} to all global symbols, so the assembly
9916 code can be used with the @code{PREFIX} assembly directive.
9920 Generate an executable in the ELF format, rather than the default
9921 @samp{mmo} format used by the @command{mmix} simulator.
9923 @item -mbranch-predict
9924 @itemx -mno-branch-predict
9925 @opindex mbranch-predict
9926 @opindex mno-branch-predict
9927 Use (do not use) the probable-branch instructions, when static branch
9928 prediction indicates a probable branch.
9930 @item -mbase-addresses
9931 @itemx -mno-base-addresses
9932 @opindex mbase-addresses
9933 @opindex mno-base-addresses
9934 Generate (do not generate) code that uses @emph{base addresses}. Using a
9935 base address automatically generates a request (handled by the assembler
9936 and the linker) for a constant to be set up in a global register. The
9937 register is used for one or more base address requests within the range 0
9938 to 255 from the value held in the register. The generally leads to short
9939 and fast code, but the number of different data items that can be
9940 addressed is limited. This means that a program that uses lots of static
9941 data may require @option{-mno-base-addresses}.
9944 @itemx -mno-single-exit
9945 @opindex msingle-exit
9946 @opindex mno-single-exit
9947 Force (do not force) generated code to have a single exit point in each
9951 @node MN10300 Options
9952 @subsection MN10300 Options
9953 @cindex MN10300 options
9955 These @option{-m} options are defined for Matsushita MN10300 architectures:
9960 Generate code to avoid bugs in the multiply instructions for the MN10300
9961 processors. This is the default.
9964 @opindex mno-mult-bug
9965 Do not generate code to avoid bugs in the multiply instructions for the
9970 Generate code which uses features specific to the AM33 processor.
9974 Do not generate code which uses features specific to the AM33 processor. This
9979 Do not link in the C run-time initialization object file.
9983 Indicate to the linker that it should perform a relaxation optimization pass
9984 to shorten branches, calls and absolute memory addresses. This option only
9985 has an effect when used on the command line for the final link step.
9987 This option makes symbolic debugging impossible.
9991 @subsection NS32K Options
9992 @cindex NS32K options
9994 These are the @samp{-m} options defined for the 32000 series. The default
9995 values for these options depends on which style of 32000 was selected when
9996 the compiler was configured; the defaults for the most common choices are
10004 Generate output for a 32032. This is the default
10005 when the compiler is configured for 32032 and 32016 based systems.
10011 Generate output for a 32332. This is the default
10012 when the compiler is configured for 32332-based systems.
10018 Generate output for a 32532. This is the default
10019 when the compiler is configured for 32532-based systems.
10023 Generate output containing 32081 instructions for floating point.
10024 This is the default for all systems.
10028 Generate output containing 32381 instructions for floating point. This
10029 also implies @option{-m32081}. The 32381 is only compatible with the 32332
10030 and 32532 cpus. This is the default for the pc532-netbsd configuration.
10033 @opindex mmulti-add
10034 Try and generate multiply-add floating point instructions @code{polyF}
10035 and @code{dotF}. This option is only available if the @option{-m32381}
10036 option is in effect. Using these instructions requires changes to
10037 register allocation which generally has a negative impact on
10038 performance. This option should only be enabled when compiling code
10039 particularly likely to make heavy use of multiply-add instructions.
10041 @item -mnomulti-add
10042 @opindex mnomulti-add
10043 Do not try and generate multiply-add floating point instructions
10044 @code{polyF} and @code{dotF}. This is the default on all platforms.
10047 @opindex msoft-float
10048 Generate output containing library calls for floating point.
10049 @strong{Warning:} the requisite libraries may not be available.
10051 @item -mieee-compare
10052 @itemx -mno-ieee-compare
10053 @opindex mieee-compare
10054 @opindex mno-ieee-compare
10055 Control whether or not the compiler uses IEEE floating point
10056 comparisons. These handle correctly the case where the result of a
10057 comparison is unordered.
10058 @strong{Warning:} the requisite kernel support may not be available.
10061 @opindex mnobitfield
10062 Do not use the bit-field instructions. On some machines it is faster to
10063 use shifting and masking operations. This is the default for the pc532.
10067 Do use the bit-field instructions. This is the default for all platforms
10072 Use a different function-calling convention, in which functions
10073 that take a fixed number of arguments return pop their
10074 arguments on return with the @code{ret} instruction.
10076 This calling convention is incompatible with the one normally
10077 used on Unix, so you cannot use it if you need to call libraries
10078 compiled with the Unix compiler.
10080 Also, you must provide function prototypes for all functions that
10081 take variable numbers of arguments (including @code{printf});
10082 otherwise incorrect code will be generated for calls to those
10085 In addition, seriously incorrect code will result if you call a
10086 function with too many arguments. (Normally, extra arguments are
10087 harmlessly ignored.)
10089 This option takes its name from the 680x0 @code{rtd} instruction.
10094 Use a different function-calling convention where the first two arguments
10095 are passed in registers.
10097 This calling convention is incompatible with the one normally
10098 used on Unix, so you cannot use it if you need to call libraries
10099 compiled with the Unix compiler.
10102 @opindex mnoregparam
10103 Do not pass any arguments in registers. This is the default for all
10108 It is OK to use the sb as an index register which is always loaded with
10109 zero. This is the default for the pc532-netbsd target.
10113 The sb register is not available for use or has not been initialized to
10114 zero by the run time system. This is the default for all targets except
10115 the pc532-netbsd. It is also implied whenever @option{-mhimem} or
10116 @option{-fpic} is set.
10120 Many ns32000 series addressing modes use displacements of up to 512MB@.
10121 If an address is above 512MB then displacements from zero can not be used.
10122 This option causes code to be generated which can be loaded above 512MB@.
10123 This may be useful for operating systems or ROM code.
10127 Assume code will be loaded in the first 512MB of virtual address space.
10128 This is the default for all platforms.
10132 @node PDP-11 Options
10133 @subsection PDP-11 Options
10134 @cindex PDP-11 Options
10136 These options are defined for the PDP-11:
10141 Use hardware FPP floating point. This is the default. (FIS floating
10142 point on the PDP-11/40 is not supported.)
10145 @opindex msoft-float
10146 Do not use hardware floating point.
10150 Return floating-point results in ac0 (fr0 in Unix assembler syntax).
10154 Return floating-point results in memory. This is the default.
10158 Generate code for a PDP-11/40.
10162 Generate code for a PDP-11/45. This is the default.
10166 Generate code for a PDP-11/10.
10168 @item -mbcopy-builtin
10169 @opindex bcopy-builtin
10170 Use inline @code{movmemhi} patterns for copying memory. This is the
10175 Do not use inline @code{movmemhi} patterns for copying memory.
10181 Use 16-bit @code{int}. This is the default.
10187 Use 32-bit @code{int}.
10190 @itemx -mno-float32
10192 @opindex mno-float32
10193 Use 64-bit @code{float}. This is the default.
10196 @itemx -mno-float64
10198 @opindex mno-float64
10199 Use 32-bit @code{float}.
10203 Use @code{abshi2} pattern. This is the default.
10207 Do not use @code{abshi2} pattern.
10209 @item -mbranch-expensive
10210 @opindex mbranch-expensive
10211 Pretend that branches are expensive. This is for experimenting with
10212 code generation only.
10214 @item -mbranch-cheap
10215 @opindex mbranch-cheap
10216 Do not pretend that branches are expensive. This is the default.
10220 Generate code for a system with split I&D@.
10224 Generate code for a system without split I&D@. This is the default.
10228 Use Unix assembler syntax. This is the default when configured for
10229 @samp{pdp11-*-bsd}.
10233 Use DEC assembler syntax. This is the default when configured for any
10234 PDP-11 target other than @samp{pdp11-*-bsd}.
10237 @node PowerPC Options
10238 @subsection PowerPC Options
10239 @cindex PowerPC options
10241 These are listed under @xref{RS/6000 and PowerPC Options}.
10243 @node RS/6000 and PowerPC Options
10244 @subsection IBM RS/6000 and PowerPC Options
10245 @cindex RS/6000 and PowerPC Options
10246 @cindex IBM RS/6000 and PowerPC Options
10248 These @samp{-m} options are defined for the IBM RS/6000 and PowerPC:
10255 @itemx -mno-powerpc
10256 @itemx -mpowerpc-gpopt
10257 @itemx -mno-powerpc-gpopt
10258 @itemx -mpowerpc-gfxopt
10259 @itemx -mno-powerpc-gfxopt
10261 @itemx -mno-powerpc64
10265 @opindex mno-power2
10267 @opindex mno-powerpc
10268 @opindex mpowerpc-gpopt
10269 @opindex mno-powerpc-gpopt
10270 @opindex mpowerpc-gfxopt
10271 @opindex mno-powerpc-gfxopt
10272 @opindex mpowerpc64
10273 @opindex mno-powerpc64
10274 GCC supports two related instruction set architectures for the
10275 RS/6000 and PowerPC@. The @dfn{POWER} instruction set are those
10276 instructions supported by the @samp{rios} chip set used in the original
10277 RS/6000 systems and the @dfn{PowerPC} instruction set is the
10278 architecture of the Motorola MPC5xx, MPC6xx, MPC8xx microprocessors, and
10279 the IBM 4xx microprocessors.
10281 Neither architecture is a subset of the other. However there is a
10282 large common subset of instructions supported by both. An MQ
10283 register is included in processors supporting the POWER architecture.
10285 You use these options to specify which instructions are available on the
10286 processor you are using. The default value of these options is
10287 determined when configuring GCC@. Specifying the
10288 @option{-mcpu=@var{cpu_type}} overrides the specification of these
10289 options. We recommend you use the @option{-mcpu=@var{cpu_type}} option
10290 rather than the options listed above.
10292 The @option{-mpower} option allows GCC to generate instructions that
10293 are found only in the POWER architecture and to use the MQ register.
10294 Specifying @option{-mpower2} implies @option{-power} and also allows GCC
10295 to generate instructions that are present in the POWER2 architecture but
10296 not the original POWER architecture.
10298 The @option{-mpowerpc} option allows GCC to generate instructions that
10299 are found only in the 32-bit subset of the PowerPC architecture.
10300 Specifying @option{-mpowerpc-gpopt} implies @option{-mpowerpc} and also allows
10301 GCC to use the optional PowerPC architecture instructions in the
10302 General Purpose group, including floating-point square root. Specifying
10303 @option{-mpowerpc-gfxopt} implies @option{-mpowerpc} and also allows GCC to
10304 use the optional PowerPC architecture instructions in the Graphics
10305 group, including floating-point select.
10307 The @option{-mpowerpc64} option allows GCC to generate the additional
10308 64-bit instructions that are found in the full PowerPC64 architecture
10309 and to treat GPRs as 64-bit, doubleword quantities. GCC defaults to
10310 @option{-mno-powerpc64}.
10312 If you specify both @option{-mno-power} and @option{-mno-powerpc}, GCC
10313 will use only the instructions in the common subset of both
10314 architectures plus some special AIX common-mode calls, and will not use
10315 the MQ register. Specifying both @option{-mpower} and @option{-mpowerpc}
10316 permits GCC to use any instruction from either architecture and to
10317 allow use of the MQ register; specify this for the Motorola MPC601.
10319 @item -mnew-mnemonics
10320 @itemx -mold-mnemonics
10321 @opindex mnew-mnemonics
10322 @opindex mold-mnemonics
10323 Select which mnemonics to use in the generated assembler code. With
10324 @option{-mnew-mnemonics}, GCC uses the assembler mnemonics defined for
10325 the PowerPC architecture. With @option{-mold-mnemonics} it uses the
10326 assembler mnemonics defined for the POWER architecture. Instructions
10327 defined in only one architecture have only one mnemonic; GCC uses that
10328 mnemonic irrespective of which of these options is specified.
10330 GCC defaults to the mnemonics appropriate for the architecture in
10331 use. Specifying @option{-mcpu=@var{cpu_type}} sometimes overrides the
10332 value of these option. Unless you are building a cross-compiler, you
10333 should normally not specify either @option{-mnew-mnemonics} or
10334 @option{-mold-mnemonics}, but should instead accept the default.
10336 @item -mcpu=@var{cpu_type}
10338 Set architecture type, register usage, choice of mnemonics, and
10339 instruction scheduling parameters for machine type @var{cpu_type}.
10340 Supported values for @var{cpu_type} are @samp{401}, @samp{403},
10341 @samp{405}, @samp{405fp}, @samp{440}, @samp{440fp}, @samp{505},
10342 @samp{601}, @samp{602}, @samp{603}, @samp{603e}, @samp{604},
10343 @samp{604e}, @samp{620}, @samp{630}, @samp{740}, @samp{7400},
10344 @samp{7450}, @samp{750}, @samp{801}, @samp{821}, @samp{823},
10345 @samp{860}, @samp{970}, @samp{common}, @samp{ec603e}, @samp{G3},
10346 @samp{G4}, @samp{G5}, @samp{power}, @samp{power2}, @samp{power3},
10347 @samp{power4}, @samp{power5}, @samp{powerpc}, @samp{powerpc64},
10348 @samp{rios}, @samp{rios1}, @samp{rios2}, @samp{rsc}, and @samp{rs64a}.
10350 @option{-mcpu=common} selects a completely generic processor. Code
10351 generated under this option will run on any POWER or PowerPC processor.
10352 GCC will use only the instructions in the common subset of both
10353 architectures, and will not use the MQ register. GCC assumes a generic
10354 processor model for scheduling purposes.
10356 @option{-mcpu=power}, @option{-mcpu=power2}, @option{-mcpu=powerpc}, and
10357 @option{-mcpu=powerpc64} specify generic POWER, POWER2, pure 32-bit
10358 PowerPC (i.e., not MPC601), and 64-bit PowerPC architecture machine
10359 types, with an appropriate, generic processor model assumed for
10360 scheduling purposes.
10362 The other options specify a specific processor. Code generated under
10363 those options will run best on that processor, and may not run at all on
10366 The @option{-mcpu} options automatically enable or disable the
10367 following options: @option{-maltivec}, @option{-mhard-float},
10368 @option{-mmfcrf}, @option{-mmultiple}, @option{-mnew-mnemonics},
10369 @option{-mpower}, @option{-mpower2}, @option{-mpowerpc64},
10370 @option{-mpowerpc-gpopt}, @option{-mpowerpc-gfxopt},
10371 @option{-mstring}. The particular options set for any particular CPU
10372 will vary between compiler versions, depending on what setting seems
10373 to produce optimal code for that CPU; it doesn't necessarily reflect
10374 the actual hardware's capabilities. If you wish to set an individual
10375 option to a particular value, you may specify it after the
10376 @option{-mcpu} option, like @samp{-mcpu=970 -mno-altivec}.
10378 On AIX, the @option{-maltivec} and @option{-mpowerpc64} options are
10379 not enabled or disabled by the @option{-mcpu} option at present, since
10380 AIX does not have full support for these options. You may still
10381 enable or disable them individually if you're sure it'll work in your
10384 @item -mtune=@var{cpu_type}
10386 Set the instruction scheduling parameters for machine type
10387 @var{cpu_type}, but do not set the architecture type, register usage, or
10388 choice of mnemonics, as @option{-mcpu=@var{cpu_type}} would. The same
10389 values for @var{cpu_type} are used for @option{-mtune} as for
10390 @option{-mcpu}. If both are specified, the code generated will use the
10391 architecture, registers, and mnemonics set by @option{-mcpu}, but the
10392 scheduling parameters set by @option{-mtune}.
10395 @itemx -mno-altivec
10397 @opindex mno-altivec
10398 Generate code that uses (does not use) AltiVec instructions, and also
10399 enable the use of built-in functions that allow more direct access to
10400 the AltiVec instruction set. You may also need to set
10401 @option{-mabi=altivec} to adjust the current ABI with AltiVec ABI
10406 Extend the current ABI with SPE ABI extensions. This does not change
10407 the default ABI, instead it adds the SPE ABI extensions to the current
10411 @opindex mabi=no-spe
10412 Disable Booke SPE ABI extensions for the current ABI@.
10414 @item -misel=@var{yes/no}
10417 This switch enables or disables the generation of ISEL instructions.
10419 @item -mspe=@var{yes/no}
10422 This switch enables or disables the generation of SPE simd
10425 @item -mfloat-gprs=@var{yes/single/double/no}
10426 @itemx -mfloat-gprs
10427 @opindex mfloat-gprs
10428 This switch enables or disables the generation of floating point
10429 operations on the general purpose registers for architectures that
10432 The argument @var{yes} or @var{single} enables the use of
10433 single-precision floating point operations.
10435 The argument @var{double} enables the use of single and
10436 double-precision floating point operations.
10438 The argument @var{no} disables floating point operations on the
10439 general purpose registers.
10441 This option is currently only available on the MPC854x.
10447 Generate code for 32-bit or 64-bit environments of Darwin and SVR4
10448 targets (including GNU/Linux). The 32-bit environment sets int, long
10449 and pointer to 32 bits and generates code that runs on any PowerPC
10450 variant. The 64-bit environment sets int to 32 bits and long and
10451 pointer to 64 bits, and generates code for PowerPC64, as for
10452 @option{-mpowerpc64}.
10455 @itemx -mno-fp-in-toc
10456 @itemx -mno-sum-in-toc
10457 @itemx -mminimal-toc
10459 @opindex mno-fp-in-toc
10460 @opindex mno-sum-in-toc
10461 @opindex mminimal-toc
10462 Modify generation of the TOC (Table Of Contents), which is created for
10463 every executable file. The @option{-mfull-toc} option is selected by
10464 default. In that case, GCC will allocate at least one TOC entry for
10465 each unique non-automatic variable reference in your program. GCC
10466 will also place floating-point constants in the TOC@. However, only
10467 16,384 entries are available in the TOC@.
10469 If you receive a linker error message that saying you have overflowed
10470 the available TOC space, you can reduce the amount of TOC space used
10471 with the @option{-mno-fp-in-toc} and @option{-mno-sum-in-toc} options.
10472 @option{-mno-fp-in-toc} prevents GCC from putting floating-point
10473 constants in the TOC and @option{-mno-sum-in-toc} forces GCC to
10474 generate code to calculate the sum of an address and a constant at
10475 run-time instead of putting that sum into the TOC@. You may specify one
10476 or both of these options. Each causes GCC to produce very slightly
10477 slower and larger code at the expense of conserving TOC space.
10479 If you still run out of space in the TOC even when you specify both of
10480 these options, specify @option{-mminimal-toc} instead. This option causes
10481 GCC to make only one TOC entry for every file. When you specify this
10482 option, GCC will produce code that is slower and larger but which
10483 uses extremely little TOC space. You may wish to use this option
10484 only on files that contain less frequently executed code.
10490 Enable 64-bit AIX ABI and calling convention: 64-bit pointers, 64-bit
10491 @code{long} type, and the infrastructure needed to support them.
10492 Specifying @option{-maix64} implies @option{-mpowerpc64} and
10493 @option{-mpowerpc}, while @option{-maix32} disables the 64-bit ABI and
10494 implies @option{-mno-powerpc64}. GCC defaults to @option{-maix32}.
10497 @itemx -mno-xl-call
10499 @opindex mno-xl-call
10500 On AIX, pass floating-point arguments to prototyped functions beyond the
10501 register save area (RSA) on the stack in addition to argument FPRs. The
10502 AIX calling convention was extended but not initially documented to
10503 handle an obscure K&R C case of calling a function that takes the
10504 address of its arguments with fewer arguments than declared. AIX XL
10505 compilers access floating point arguments which do not fit in the
10506 RSA from the stack when a subroutine is compiled without
10507 optimization. Because always storing floating-point arguments on the
10508 stack is inefficient and rarely needed, this option is not enabled by
10509 default and only is necessary when calling subroutines compiled by AIX
10510 XL compilers without optimization.
10514 Support @dfn{IBM RS/6000 SP} @dfn{Parallel Environment} (PE)@. Link an
10515 application written to use message passing with special startup code to
10516 enable the application to run. The system must have PE installed in the
10517 standard location (@file{/usr/lpp/ppe.poe/}), or the @file{specs} file
10518 must be overridden with the @option{-specs=} option to specify the
10519 appropriate directory location. The Parallel Environment does not
10520 support threads, so the @option{-mpe} option and the @option{-pthread}
10521 option are incompatible.
10523 @item -malign-natural
10524 @itemx -malign-power
10525 @opindex malign-natural
10526 @opindex malign-power
10527 On AIX, Darwin, and 64-bit PowerPC GNU/Linux, the option
10528 @option{-malign-natural} overrides the ABI-defined alignment of larger
10529 types, such as floating-point doubles, on their natural size-based boundary.
10530 The option @option{-malign-power} instructs GCC to follow the ABI-specified
10531 alignment rules. GCC defaults to the standard alignment defined in the ABI@.
10534 @itemx -mhard-float
10535 @opindex msoft-float
10536 @opindex mhard-float
10537 Generate code that does not use (uses) the floating-point register set.
10538 Software floating point emulation is provided if you use the
10539 @option{-msoft-float} option, and pass the option to GCC when linking.
10542 @itemx -mno-multiple
10544 @opindex mno-multiple
10545 Generate code that uses (does not use) the load multiple word
10546 instructions and the store multiple word instructions. These
10547 instructions are generated by default on POWER systems, and not
10548 generated on PowerPC systems. Do not use @option{-mmultiple} on little
10549 endian PowerPC systems, since those instructions do not work when the
10550 processor is in little endian mode. The exceptions are PPC740 and
10551 PPC750 which permit the instructions usage in little endian mode.
10556 @opindex mno-string
10557 Generate code that uses (does not use) the load string instructions
10558 and the store string word instructions to save multiple registers and
10559 do small block moves. These instructions are generated by default on
10560 POWER systems, and not generated on PowerPC systems. Do not use
10561 @option{-mstring} on little endian PowerPC systems, since those
10562 instructions do not work when the processor is in little endian mode.
10563 The exceptions are PPC740 and PPC750 which permit the instructions
10564 usage in little endian mode.
10569 @opindex mno-update
10570 Generate code that uses (does not use) the load or store instructions
10571 that update the base register to the address of the calculated memory
10572 location. These instructions are generated by default. If you use
10573 @option{-mno-update}, there is a small window between the time that the
10574 stack pointer is updated and the address of the previous frame is
10575 stored, which means code that walks the stack frame across interrupts or
10576 signals may get corrupted data.
10579 @itemx -mno-fused-madd
10580 @opindex mfused-madd
10581 @opindex mno-fused-madd
10582 Generate code that uses (does not use) the floating point multiply and
10583 accumulate instructions. These instructions are generated by default if
10584 hardware floating is used.
10586 @item -mno-bit-align
10588 @opindex mno-bit-align
10589 @opindex mbit-align
10590 On System V.4 and embedded PowerPC systems do not (do) force structures
10591 and unions that contain bit-fields to be aligned to the base type of the
10594 For example, by default a structure containing nothing but 8
10595 @code{unsigned} bit-fields of length 1 would be aligned to a 4 byte
10596 boundary and have a size of 4 bytes. By using @option{-mno-bit-align},
10597 the structure would be aligned to a 1 byte boundary and be one byte in
10600 @item -mno-strict-align
10601 @itemx -mstrict-align
10602 @opindex mno-strict-align
10603 @opindex mstrict-align
10604 On System V.4 and embedded PowerPC systems do not (do) assume that
10605 unaligned memory references will be handled by the system.
10607 @item -mrelocatable
10608 @itemx -mno-relocatable
10609 @opindex mrelocatable
10610 @opindex mno-relocatable
10611 On embedded PowerPC systems generate code that allows (does not allow)
10612 the program to be relocated to a different address at runtime. If you
10613 use @option{-mrelocatable} on any module, all objects linked together must
10614 be compiled with @option{-mrelocatable} or @option{-mrelocatable-lib}.
10616 @item -mrelocatable-lib
10617 @itemx -mno-relocatable-lib
10618 @opindex mrelocatable-lib
10619 @opindex mno-relocatable-lib
10620 On embedded PowerPC systems generate code that allows (does not allow)
10621 the program to be relocated to a different address at runtime. Modules
10622 compiled with @option{-mrelocatable-lib} can be linked with either modules
10623 compiled without @option{-mrelocatable} and @option{-mrelocatable-lib} or
10624 with modules compiled with the @option{-mrelocatable} options.
10630 On System V.4 and embedded PowerPC systems do not (do) assume that
10631 register 2 contains a pointer to a global area pointing to the addresses
10632 used in the program.
10635 @itemx -mlittle-endian
10637 @opindex mlittle-endian
10638 On System V.4 and embedded PowerPC systems compile code for the
10639 processor in little endian mode. The @option{-mlittle-endian} option is
10640 the same as @option{-mlittle}.
10643 @itemx -mbig-endian
10645 @opindex mbig-endian
10646 On System V.4 and embedded PowerPC systems compile code for the
10647 processor in big endian mode. The @option{-mbig-endian} option is
10648 the same as @option{-mbig}.
10650 @item -mdynamic-no-pic
10651 @opindex mdynamic-no-pic
10652 On Darwin and Mac OS X systems, compile code so that it is not
10653 relocatable, but that its external references are relocatable. The
10654 resulting code is suitable for applications, but not shared
10657 @item -mprioritize-restricted-insns=@var{priority}
10658 @opindex mprioritize-restricted-insns
10659 This option controls the priority that is assigned to
10660 dispatch-slot restricted instructions during the second scheduling
10661 pass. The argument @var{priority} takes the value @var{0/1/2} to assign
10662 @var{no/highest/second-highest} priority to dispatch slot restricted
10665 @item -msched-costly-dep=@var{dependence_type}
10666 @opindex msched-costly-dep
10667 This option controls which dependences are considered costly
10668 by the target during instruction scheduling. The argument
10669 @var{dependence_type} takes one of the following values:
10670 @var{no}: no dependence is costly,
10671 @var{all}: all dependences are costly,
10672 @var{true_store_to_load}: a true dependence from store to load is costly,
10673 @var{store_to_load}: any dependence from store to load is costly,
10674 @var{number}: any dependence which latency >= @var{number} is costly.
10676 @item -minsert-sched-nops=@var{scheme}
10677 @opindex minsert-sched-nops
10678 This option controls which nop insertion scheme will be used during
10679 the second scheduling pass. The argument @var{scheme} takes one of the
10681 @var{no}: Don't insert nops.
10682 @var{pad}: Pad with nops any dispatch group which has vacant issue slots,
10683 according to the scheduler's grouping.
10684 @var{regroup_exact}: Insert nops to force costly dependent insns into
10685 separate groups. Insert exactly as many nops as needed to force an insn
10686 to a new group, according to the estimated processor grouping.
10687 @var{number}: Insert nops to force costly dependent insns into
10688 separate groups. Insert @var{number} nops to force an insn to a new group.
10691 @opindex mcall-sysv
10692 On System V.4 and embedded PowerPC systems compile code using calling
10693 conventions that adheres to the March 1995 draft of the System V
10694 Application Binary Interface, PowerPC processor supplement. This is the
10695 default unless you configured GCC using @samp{powerpc-*-eabiaix}.
10697 @item -mcall-sysv-eabi
10698 @opindex mcall-sysv-eabi
10699 Specify both @option{-mcall-sysv} and @option{-meabi} options.
10701 @item -mcall-sysv-noeabi
10702 @opindex mcall-sysv-noeabi
10703 Specify both @option{-mcall-sysv} and @option{-mno-eabi} options.
10705 @item -mcall-solaris
10706 @opindex mcall-solaris
10707 On System V.4 and embedded PowerPC systems compile code for the Solaris
10711 @opindex mcall-linux
10712 On System V.4 and embedded PowerPC systems compile code for the
10713 Linux-based GNU system.
10717 On System V.4 and embedded PowerPC systems compile code for the
10718 Hurd-based GNU system.
10720 @item -mcall-netbsd
10721 @opindex mcall-netbsd
10722 On System V.4 and embedded PowerPC systems compile code for the
10723 NetBSD operating system.
10725 @item -maix-struct-return
10726 @opindex maix-struct-return
10727 Return all structures in memory (as specified by the AIX ABI)@.
10729 @item -msvr4-struct-return
10730 @opindex msvr4-struct-return
10731 Return structures smaller than 8 bytes in registers (as specified by the
10734 @item -mabi=altivec
10735 @opindex mabi=altivec
10736 Extend the current ABI with AltiVec ABI extensions. This does not
10737 change the default ABI, instead it adds the AltiVec ABI extensions to
10740 @item -mabi=no-altivec
10741 @opindex mabi=no-altivec
10742 Disable AltiVec ABI extensions for the current ABI@.
10745 @itemx -mno-prototype
10746 @opindex mprototype
10747 @opindex mno-prototype
10748 On System V.4 and embedded PowerPC systems assume that all calls to
10749 variable argument functions are properly prototyped. Otherwise, the
10750 compiler must insert an instruction before every non prototyped call to
10751 set or clear bit 6 of the condition code register (@var{CR}) to
10752 indicate whether floating point values were passed in the floating point
10753 registers in case the function takes a variable arguments. With
10754 @option{-mprototype}, only calls to prototyped variable argument functions
10755 will set or clear the bit.
10759 On embedded PowerPC systems, assume that the startup module is called
10760 @file{sim-crt0.o} and that the standard C libraries are @file{libsim.a} and
10761 @file{libc.a}. This is the default for @samp{powerpc-*-eabisim}.
10766 On embedded PowerPC systems, assume that the startup module is called
10767 @file{crt0.o} and the standard C libraries are @file{libmvme.a} and
10772 On embedded PowerPC systems, assume that the startup module is called
10773 @file{crt0.o} and the standard C libraries are @file{libads.a} and
10776 @item -myellowknife
10777 @opindex myellowknife
10778 On embedded PowerPC systems, assume that the startup module is called
10779 @file{crt0.o} and the standard C libraries are @file{libyk.a} and
10784 On System V.4 and embedded PowerPC systems, specify that you are
10785 compiling for a VxWorks system.
10789 Specify that you are compiling for the WindISS simulation environment.
10793 On embedded PowerPC systems, set the @var{PPC_EMB} bit in the ELF flags
10794 header to indicate that @samp{eabi} extended relocations are used.
10800 On System V.4 and embedded PowerPC systems do (do not) adhere to the
10801 Embedded Applications Binary Interface (eabi) which is a set of
10802 modifications to the System V.4 specifications. Selecting @option{-meabi}
10803 means that the stack is aligned to an 8 byte boundary, a function
10804 @code{__eabi} is called to from @code{main} to set up the eabi
10805 environment, and the @option{-msdata} option can use both @code{r2} and
10806 @code{r13} to point to two separate small data areas. Selecting
10807 @option{-mno-eabi} means that the stack is aligned to a 16 byte boundary,
10808 do not call an initialization function from @code{main}, and the
10809 @option{-msdata} option will only use @code{r13} to point to a single
10810 small data area. The @option{-meabi} option is on by default if you
10811 configured GCC using one of the @samp{powerpc*-*-eabi*} options.
10814 @opindex msdata=eabi
10815 On System V.4 and embedded PowerPC systems, put small initialized
10816 @code{const} global and static data in the @samp{.sdata2} section, which
10817 is pointed to by register @code{r2}. Put small initialized
10818 non-@code{const} global and static data in the @samp{.sdata} section,
10819 which is pointed to by register @code{r13}. Put small uninitialized
10820 global and static data in the @samp{.sbss} section, which is adjacent to
10821 the @samp{.sdata} section. The @option{-msdata=eabi} option is
10822 incompatible with the @option{-mrelocatable} option. The
10823 @option{-msdata=eabi} option also sets the @option{-memb} option.
10826 @opindex msdata=sysv
10827 On System V.4 and embedded PowerPC systems, put small global and static
10828 data in the @samp{.sdata} section, which is pointed to by register
10829 @code{r13}. Put small uninitialized global and static data in the
10830 @samp{.sbss} section, which is adjacent to the @samp{.sdata} section.
10831 The @option{-msdata=sysv} option is incompatible with the
10832 @option{-mrelocatable} option.
10834 @item -msdata=default
10836 @opindex msdata=default
10838 On System V.4 and embedded PowerPC systems, if @option{-meabi} is used,
10839 compile code the same as @option{-msdata=eabi}, otherwise compile code the
10840 same as @option{-msdata=sysv}.
10843 @opindex msdata-data
10844 On System V.4 and embedded PowerPC systems, put small global and static
10845 data in the @samp{.sdata} section. Put small uninitialized global and
10846 static data in the @samp{.sbss} section. Do not use register @code{r13}
10847 to address small data however. This is the default behavior unless
10848 other @option{-msdata} options are used.
10852 @opindex msdata=none
10854 On embedded PowerPC systems, put all initialized global and static data
10855 in the @samp{.data} section, and all uninitialized data in the
10856 @samp{.bss} section.
10860 @cindex smaller data references (PowerPC)
10861 @cindex .sdata/.sdata2 references (PowerPC)
10862 On embedded PowerPC systems, put global and static items less than or
10863 equal to @var{num} bytes into the small data or bss sections instead of
10864 the normal data or bss section. By default, @var{num} is 8. The
10865 @option{-G @var{num}} switch is also passed to the linker.
10866 All modules should be compiled with the same @option{-G @var{num}} value.
10869 @itemx -mno-regnames
10871 @opindex mno-regnames
10872 On System V.4 and embedded PowerPC systems do (do not) emit register
10873 names in the assembly language output using symbolic forms.
10876 @itemx -mno-longcall
10878 @opindex mno-longcall
10879 Default to making all function calls indirectly, using a register, so
10880 that functions which reside further than 32 megabytes (33,554,432
10881 bytes) from the current location can be called. This setting can be
10882 overridden by the @code{shortcall} function attribute, or by
10883 @code{#pragma longcall(0)}.
10885 Some linkers are capable of detecting out-of-range calls and generating
10886 glue code on the fly. On these systems, long calls are unnecessary and
10887 generate slower code. As of this writing, the AIX linker can do this,
10888 as can the GNU linker for PowerPC/64. It is planned to add this feature
10889 to the GNU linker for 32-bit PowerPC systems as well.
10891 On Darwin/PPC systems, @code{#pragma longcall} will generate ``jbsr
10892 callee, L42'', plus a ``branch island'' (glue code). The two target
10893 addresses represent the callee and the ``branch island''. The
10894 Darwin/PPC linker will prefer the first address and generate a ``bl
10895 callee'' if the PPC ``bl'' instruction will reach the callee directly;
10896 otherwise, the linker will generate ``bl L42'' to call the ``branch
10897 island''. The ``branch island'' is appended to the body of the
10898 calling function; it computes the full 32-bit address of the callee
10901 On Mach-O (Darwin) systems, this option directs the compiler emit to
10902 the glue for every direct call, and the Darwin linker decides whether
10903 to use or discard it.
10905 In the future, we may cause GCC to ignore all longcall specifications
10906 when the linker is known to generate glue.
10910 Adds support for multithreading with the @dfn{pthreads} library.
10911 This option sets flags for both the preprocessor and linker.
10915 @node S/390 and zSeries Options
10916 @subsection S/390 and zSeries Options
10917 @cindex S/390 and zSeries Options
10919 These are the @samp{-m} options defined for the S/390 and zSeries architecture.
10923 @itemx -msoft-float
10924 @opindex mhard-float
10925 @opindex msoft-float
10926 Use (do not use) the hardware floating-point instructions and registers
10927 for floating-point operations. When @option{-msoft-float} is specified,
10928 functions in @file{libgcc.a} will be used to perform floating-point
10929 operations. When @option{-mhard-float} is specified, the compiler
10930 generates IEEE floating-point instructions. This is the default.
10933 @itemx -mno-backchain
10934 @opindex mbackchain
10935 @opindex mno-backchain
10936 Store (do not store) the address of the caller's frame as backchain pointer
10937 into the callee's stack frame.
10938 A backchain may be needed to allow debugging using tools that do not understand
10939 DWARF-2 call frame information.
10940 When @option{-mno-packed-stack} is in effect, the backchain pointer is stored
10941 at the bottom of the stack frame; when @option{-mpacked-stack} is in effect,
10942 the backchain is placed into the topmost word of the 96/160 byte register
10945 In general, code compiled with @option{-mbackchain} is call-compatible with
10946 code compiled with @option{-mmo-backchain}; however, use of the backchain
10947 for debugging purposes usually requires that the whole binary is built with
10948 @option{-mbackchain}. Note that the combination of @option{-mbackchain},
10949 @option{-mpacked-stack} and @option{-mhard-float} is not supported. In order
10950 to build a linux kernel use @option{-msoft-float}.
10952 The default is to not maintain the backchain.
10954 @item -mpacked-stack
10955 @item -mno-packed-stack
10956 @opindex mpacked-stack
10957 @opindex mno-packed-stack
10958 Use (do not use) the packed stack layout. When @option{-mno-packed-stack} is
10959 specified, the compiler uses the all fields of the 96/160 byte register save
10960 area only for their default purpose; unused fields still take up stack space.
10961 When @option{-mpacked-stack} is specified, register save slots are densely
10962 packed at the top of the register save area; unused space is reused for other
10963 purposes, allowing for more efficient use of the available stack space.
10964 However, when @option{-mbackchain} is also in effect, the topmost word of
10965 the save area is always used to store the backchain, and the return address
10966 register is always saved two words below the backchain.
10968 As long as the stack frame backchain is not used, code generated with
10969 @option{-mpacked-stack} is call-compatible with code generated with
10970 @option{-mno-packed-stack}. Note that some non-FSF releases of GCC 2.95 for
10971 S/390 or zSeries generated code that uses the stack frame backchain at run
10972 time, not just for debugging purposes. Such code is not call-compatible
10973 with code compiled with @option{-mpacked-stack}. Also, note that the
10974 combination of @option{-mbackchain},
10975 @option{-mpacked-stack} and @option{-mhard-float} is not supported. In order
10976 to build a linux kernel use @option{-msoft-float}.
10978 The default is to not use the packed stack layout.
10981 @itemx -mno-small-exec
10982 @opindex msmall-exec
10983 @opindex mno-small-exec
10984 Generate (or do not generate) code using the @code{bras} instruction
10985 to do subroutine calls.
10986 This only works reliably if the total executable size does not
10987 exceed 64k. The default is to use the @code{basr} instruction instead,
10988 which does not have this limitation.
10994 When @option{-m31} is specified, generate code compliant to the
10995 GNU/Linux for S/390 ABI@. When @option{-m64} is specified, generate
10996 code compliant to the GNU/Linux for zSeries ABI@. This allows GCC in
10997 particular to generate 64-bit instructions. For the @samp{s390}
10998 targets, the default is @option{-m31}, while the @samp{s390x}
10999 targets default to @option{-m64}.
11005 When @option{-mzarch} is specified, generate code using the
11006 instructions available on z/Architecture.
11007 When @option{-mesa} is specified, generate code using the
11008 instructions available on ESA/390. Note that @option{-mesa} is
11009 not possible with @option{-m64}.
11010 When generating code compliant to the GNU/Linux for S/390 ABI,
11011 the default is @option{-mesa}. When generating code compliant
11012 to the GNU/Linux for zSeries ABI, the default is @option{-mzarch}.
11018 Generate (or do not generate) code using the @code{mvcle} instruction
11019 to perform block moves. When @option{-mno-mvcle} is specified,
11020 use a @code{mvc} loop instead. This is the default.
11026 Print (or do not print) additional debug information when compiling.
11027 The default is to not print debug information.
11029 @item -march=@var{cpu-type}
11031 Generate code that will run on @var{cpu-type}, which is the name of a system
11032 representing a certain processor type. Possible values for
11033 @var{cpu-type} are @samp{g5}, @samp{g6}, @samp{z900}, and @samp{z990}.
11034 When generating code using the instructions available on z/Architecture,
11035 the default is @option{-march=z900}. Otherwise, the default is
11036 @option{-march=g5}.
11038 @item -mtune=@var{cpu-type}
11040 Tune to @var{cpu-type} everything applicable about the generated code,
11041 except for the ABI and the set of available instructions.
11042 The list of @var{cpu-type} values is the same as for @option{-march}.
11043 The default is the value used for @option{-march}.
11046 @itemx -mno-tpf-trace
11047 @opindex mtpf-trace
11048 @opindex mno-tpf-trace
11049 Generate code that adds (does not add) in TPF OS specific branches to trace
11050 routines in the operating system. This option is off by default, even
11051 when compiling for the TPF OS@.
11054 @itemx -mno-fused-madd
11055 @opindex mfused-madd
11056 @opindex mno-fused-madd
11057 Generate code that uses (does not use) the floating point multiply and
11058 accumulate instructions. These instructions are generated by default if
11059 hardware floating point is used.
11061 @item -mwarn-framesize=@var{framesize}
11062 @opindex mwarn-framesize
11063 Emit a warning if the current function exceeds the given frame size. Because
11064 this is a compile time check it doesn't need to be a real problem when the program
11065 runs. It is intended to identify functions which most probably cause
11066 a stack overflow. It is useful to be used in an environment with limited stack
11067 size e.g.@: the linux kernel.
11069 @item -mwarn-dynamicstack
11070 @opindex mwarn-dynamicstack
11071 Emit a warning if the function calls alloca or uses dynamically
11072 sized arrays. This is generally a bad idea with a limited stack size.
11074 @item -mstack-guard=@var{stack-guard}
11075 @item -mstack-size=@var{stack-size}
11076 @opindex mstack-guard
11077 @opindex mstack-size
11078 These arguments always have to be used in conjunction. If they are present the s390
11079 back end emits additional instructions in the function prologue which trigger a trap
11080 if the stack size is @var{stack-guard} bytes above the @var{stack-size}
11081 (remember that the stack on s390 grows downward). These options are intended to
11082 be used to help debugging stack overflow problems. The additionally emitted code
11083 cause only little overhead and hence can also be used in production like systems
11084 without greater performance degradation. The given values have to be exact
11085 powers of 2 and @var{stack-size} has to be greater than @var{stack-guard}.
11086 In order to be efficient the extra code makes the assumption that the stack starts
11087 at an address aligned to the value given by @var{stack-size}.
11091 @subsection SH Options
11093 These @samp{-m} options are defined for the SH implementations:
11098 Generate code for the SH1.
11102 Generate code for the SH2.
11105 Generate code for the SH2e.
11109 Generate code for the SH3.
11113 Generate code for the SH3e.
11117 Generate code for the SH4 without a floating-point unit.
11119 @item -m4-single-only
11120 @opindex m4-single-only
11121 Generate code for the SH4 with a floating-point unit that only
11122 supports single-precision arithmetic.
11126 Generate code for the SH4 assuming the floating-point unit is in
11127 single-precision mode by default.
11131 Generate code for the SH4.
11135 Generate code for the SH4al-dsp, or for a SH4a in such a way that the
11136 floating-point unit is not used.
11138 @item -m4a-single-only
11139 @opindex m4a-single-only
11140 Generate code for the SH4a, in such a way that no double-precision
11141 floating point operations are used.
11144 @opindex m4a-single
11145 Generate code for the SH4a assuming the floating-point unit is in
11146 single-precision mode by default.
11150 Generate code for the SH4a.
11154 Same as @option{-m4a-nofpu}, except that it implicitly passes
11155 @option{-dsp} to the assembler. GCC doesn't generate any DSP
11156 instructions at the moment.
11160 Compile code for the processor in big endian mode.
11164 Compile code for the processor in little endian mode.
11168 Align doubles at 64-bit boundaries. Note that this changes the calling
11169 conventions, and thus some functions from the standard C library will
11170 not work unless you recompile it first with @option{-mdalign}.
11174 Shorten some address references at link time, when possible; uses the
11175 linker option @option{-relax}.
11179 Use 32-bit offsets in @code{switch} tables. The default is to use
11184 Enable the use of the instruction @code{fmovd}.
11188 Comply with the calling conventions defined by Renesas.
11192 Comply with the calling conventions defined by Renesas.
11196 Comply with the calling conventions defined for GCC before the Renesas
11197 conventions were available. This option is the default for all
11198 targets of the SH toolchain except for @samp{sh-symbianelf}.
11201 @opindex mnomacsave
11202 Mark the @code{MAC} register as call-clobbered, even if
11203 @option{-mhitachi} is given.
11207 Increase IEEE-compliance of floating-point code.
11211 Dump instruction size and location in the assembly code.
11214 @opindex mpadstruct
11215 This option is deprecated. It pads structures to multiple of 4 bytes,
11216 which is incompatible with the SH ABI@.
11220 Optimize for space instead of speed. Implied by @option{-Os}.
11223 @opindex mprefergot
11224 When generating position-independent code, emit function calls using
11225 the Global Offset Table instead of the Procedure Linkage Table.
11229 Generate a library function call to invalidate instruction cache
11230 entries, after fixing up a trampoline. This library function call
11231 doesn't assume it can write to the whole memory address space. This
11232 is the default when the target is @code{sh-*-linux*}.
11235 @node SPARC Options
11236 @subsection SPARC Options
11237 @cindex SPARC options
11239 These @samp{-m} options are supported on the SPARC:
11242 @item -mno-app-regs
11244 @opindex mno-app-regs
11246 Specify @option{-mapp-regs} to generate output using the global registers
11247 2 through 4, which the SPARC SVR4 ABI reserves for applications. This
11250 To be fully SVR4 ABI compliant at the cost of some performance loss,
11251 specify @option{-mno-app-regs}. You should compile libraries and system
11252 software with this option.
11255 @itemx -mhard-float
11257 @opindex mhard-float
11258 Generate output containing floating point instructions. This is the
11262 @itemx -msoft-float
11264 @opindex msoft-float
11265 Generate output containing library calls for floating point.
11266 @strong{Warning:} the requisite libraries are not available for all SPARC
11267 targets. Normally the facilities of the machine's usual C compiler are
11268 used, but this cannot be done directly in cross-compilation. You must make
11269 your own arrangements to provide suitable library functions for
11270 cross-compilation. The embedded targets @samp{sparc-*-aout} and
11271 @samp{sparclite-*-*} do provide software floating point support.
11273 @option{-msoft-float} changes the calling convention in the output file;
11274 therefore, it is only useful if you compile @emph{all} of a program with
11275 this option. In particular, you need to compile @file{libgcc.a}, the
11276 library that comes with GCC, with @option{-msoft-float} in order for
11279 @item -mhard-quad-float
11280 @opindex mhard-quad-float
11281 Generate output containing quad-word (long double) floating point
11284 @item -msoft-quad-float
11285 @opindex msoft-quad-float
11286 Generate output containing library calls for quad-word (long double)
11287 floating point instructions. The functions called are those specified
11288 in the SPARC ABI@. This is the default.
11290 As of this writing, there are no SPARC implementations that have hardware
11291 support for the quad-word floating point instructions. They all invoke
11292 a trap handler for one of these instructions, and then the trap handler
11293 emulates the effect of the instruction. Because of the trap handler overhead,
11294 this is much slower than calling the ABI library routines. Thus the
11295 @option{-msoft-quad-float} option is the default.
11297 @item -mno-unaligned-doubles
11298 @itemx -munaligned-doubles
11299 @opindex mno-unaligned-doubles
11300 @opindex munaligned-doubles
11301 Assume that doubles have 8 byte alignment. This is the default.
11303 With @option{-munaligned-doubles}, GCC assumes that doubles have 8 byte
11304 alignment only if they are contained in another type, or if they have an
11305 absolute address. Otherwise, it assumes they have 4 byte alignment.
11306 Specifying this option avoids some rare compatibility problems with code
11307 generated by other compilers. It is not the default because it results
11308 in a performance loss, especially for floating point code.
11310 @item -mno-faster-structs
11311 @itemx -mfaster-structs
11312 @opindex mno-faster-structs
11313 @opindex mfaster-structs
11314 With @option{-mfaster-structs}, the compiler assumes that structures
11315 should have 8 byte alignment. This enables the use of pairs of
11316 @code{ldd} and @code{std} instructions for copies in structure
11317 assignment, in place of twice as many @code{ld} and @code{st} pairs.
11318 However, the use of this changed alignment directly violates the SPARC
11319 ABI@. Thus, it's intended only for use on targets where the developer
11320 acknowledges that their resulting code will not be directly in line with
11321 the rules of the ABI@.
11323 @item -mimpure-text
11324 @opindex mimpure-text
11325 @option{-mimpure-text}, used in addition to @option{-shared}, tells
11326 the compiler to not pass @option{-z text} to the linker when linking a
11327 shared object. Using this option, you can link position-dependent
11328 code into a shared object.
11330 @option{-mimpure-text} suppresses the ``relocations remain against
11331 allocatable but non-writable sections'' linker error message.
11332 However, the necessary relocations will trigger copy-on-write, and the
11333 shared object is not actually shared across processes. Instead of
11334 using @option{-mimpure-text}, you should compile all source code with
11335 @option{-fpic} or @option{-fPIC}.
11337 This option is only available on SunOS and Solaris.
11339 @item -mcpu=@var{cpu_type}
11341 Set the instruction set, register set, and instruction scheduling parameters
11342 for machine type @var{cpu_type}. Supported values for @var{cpu_type} are
11343 @samp{v7}, @samp{cypress}, @samp{v8}, @samp{supersparc}, @samp{sparclite},
11344 @samp{f930}, @samp{f934}, @samp{hypersparc}, @samp{sparclite86x},
11345 @samp{sparclet}, @samp{tsc701}, @samp{v9}, @samp{ultrasparc}, and
11346 @samp{ultrasparc3}.
11348 Default instruction scheduling parameters are used for values that select
11349 an architecture and not an implementation. These are @samp{v7}, @samp{v8},
11350 @samp{sparclite}, @samp{sparclet}, @samp{v9}.
11352 Here is a list of each supported architecture and their supported
11357 v8: supersparc, hypersparc
11358 sparclite: f930, f934, sparclite86x
11360 v9: ultrasparc, ultrasparc3
11363 By default (unless configured otherwise), GCC generates code for the V7
11364 variant of the SPARC architecture. With @option{-mcpu=cypress}, the compiler
11365 additionally optimizes it for the Cypress CY7C602 chip, as used in the
11366 SPARCStation/SPARCServer 3xx series. This is also appropriate for the older
11367 SPARCStation 1, 2, IPX etc.
11369 With @option{-mcpu=v8}, GCC generates code for the V8 variant of the SPARC
11370 architecture. The only difference from V7 code is that the compiler emits
11371 the integer multiply and integer divide instructions which exist in SPARC-V8
11372 but not in SPARC-V7. With @option{-mcpu=supersparc}, the compiler additionally
11373 optimizes it for the SuperSPARC chip, as used in the SPARCStation 10, 1000 and
11376 With @option{-mcpu=sparclite}, GCC generates code for the SPARClite variant of
11377 the SPARC architecture. This adds the integer multiply, integer divide step
11378 and scan (@code{ffs}) instructions which exist in SPARClite but not in SPARC-V7.
11379 With @option{-mcpu=f930}, the compiler additionally optimizes it for the
11380 Fujitsu MB86930 chip, which is the original SPARClite, with no FPU@. With
11381 @option{-mcpu=f934}, the compiler additionally optimizes it for the Fujitsu
11382 MB86934 chip, which is the more recent SPARClite with FPU@.
11384 With @option{-mcpu=sparclet}, GCC generates code for the SPARClet variant of
11385 the SPARC architecture. This adds the integer multiply, multiply/accumulate,
11386 integer divide step and scan (@code{ffs}) instructions which exist in SPARClet
11387 but not in SPARC-V7. With @option{-mcpu=tsc701}, the compiler additionally
11388 optimizes it for the TEMIC SPARClet chip.
11390 With @option{-mcpu=v9}, GCC generates code for the V9 variant of the SPARC
11391 architecture. This adds 64-bit integer and floating-point move instructions,
11392 3 additional floating-point condition code registers and conditional move
11393 instructions. With @option{-mcpu=ultrasparc}, the compiler additionally
11394 optimizes it for the Sun UltraSPARC I/II chips. With
11395 @option{-mcpu=ultrasparc3}, the compiler additionally optimizes it for the
11396 Sun UltraSPARC III chip.
11398 @item -mtune=@var{cpu_type}
11400 Set the instruction scheduling parameters for machine type
11401 @var{cpu_type}, but do not set the instruction set or register set that the
11402 option @option{-mcpu=@var{cpu_type}} would.
11404 The same values for @option{-mcpu=@var{cpu_type}} can be used for
11405 @option{-mtune=@var{cpu_type}}, but the only useful values are those
11406 that select a particular cpu implementation. Those are @samp{cypress},
11407 @samp{supersparc}, @samp{hypersparc}, @samp{f930}, @samp{f934},
11408 @samp{sparclite86x}, @samp{tsc701}, @samp{ultrasparc}, and
11409 @samp{ultrasparc3}.
11414 @opindex mno-v8plus
11415 With @option{-mv8plus}, GCC generates code for the SPARC-V8+ ABI@. The
11416 difference from the V8 ABI is that the global and out registers are
11417 considered 64-bit wide. This is enabled by default on Solaris in 32-bit
11418 mode for all SPARC-V9 processors.
11424 With @option{-mvis}, GCC generates code that takes advantage of the UltraSPARC
11425 Visual Instruction Set extensions. The default is @option{-mno-vis}.
11428 These @samp{-m} options are supported in addition to the above
11429 on SPARC-V9 processors in 64-bit environments:
11432 @item -mlittle-endian
11433 @opindex mlittle-endian
11434 Generate code for a processor running in little-endian mode. It is only
11435 available for a few configurations and most notably not on Solaris.
11441 Generate code for a 32-bit or 64-bit environment.
11442 The 32-bit environment sets int, long and pointer to 32 bits.
11443 The 64-bit environment sets int to 32 bits and long and pointer
11446 @item -mcmodel=medlow
11447 @opindex mcmodel=medlow
11448 Generate code for the Medium/Low code model: 64-bit addresses, programs
11449 must be linked in the low 32 bits of memory. Programs can be statically
11450 or dynamically linked.
11452 @item -mcmodel=medmid
11453 @opindex mcmodel=medmid
11454 Generate code for the Medium/Middle code model: 64-bit addresses, programs
11455 must be linked in the low 44 bits of memory, the text and data segments must
11456 be less than 2GB in size and the data segment must be located within 2GB of
11459 @item -mcmodel=medany
11460 @opindex mcmodel=medany
11461 Generate code for the Medium/Anywhere code model: 64-bit addresses, programs
11462 may be linked anywhere in memory, the text and data segments must be less
11463 than 2GB in size and the data segment must be located within 2GB of the
11466 @item -mcmodel=embmedany
11467 @opindex mcmodel=embmedany
11468 Generate code for the Medium/Anywhere code model for embedded systems:
11469 64-bit addresses, the text and data segments must be less than 2GB in
11470 size, both starting anywhere in memory (determined at link time). The
11471 global register %g4 points to the base of the data segment. Programs
11472 are statically linked and PIC is not supported.
11475 @itemx -mno-stack-bias
11476 @opindex mstack-bias
11477 @opindex mno-stack-bias
11478 With @option{-mstack-bias}, GCC assumes that the stack pointer, and
11479 frame pointer if present, are offset by @minus{}2047 which must be added back
11480 when making stack frame references. This is the default in 64-bit mode.
11481 Otherwise, assume no such offset is present.
11484 These switches are supported in addition to the above on Solaris:
11489 Add support for multithreading using the Solaris threads library. This
11490 option sets flags for both the preprocessor and linker. This option does
11491 not affect the thread safety of object code produced by the compiler or
11492 that of libraries supplied with it.
11496 Add support for multithreading using the POSIX threads library. This
11497 option sets flags for both the preprocessor and linker. This option does
11498 not affect the thread safety of object code produced by the compiler or
11499 that of libraries supplied with it.
11502 @node System V Options
11503 @subsection Options for System V
11505 These additional options are available on System V Release 4 for
11506 compatibility with other compilers on those systems:
11511 Create a shared object.
11512 It is recommended that @option{-symbolic} or @option{-shared} be used instead.
11516 Identify the versions of each tool used by the compiler, in a
11517 @code{.ident} assembler directive in the output.
11521 Refrain from adding @code{.ident} directives to the output file (this is
11524 @item -YP,@var{dirs}
11526 Search the directories @var{dirs}, and no others, for libraries
11527 specified with @option{-l}.
11529 @item -Ym,@var{dir}
11531 Look in the directory @var{dir} to find the M4 preprocessor.
11532 The assembler uses this option.
11533 @c This is supposed to go with a -Yd for predefined M4 macro files, but
11534 @c the generic assembler that comes with Solaris takes just -Ym.
11537 @node TMS320C3x/C4x Options
11538 @subsection TMS320C3x/C4x Options
11539 @cindex TMS320C3x/C4x Options
11541 These @samp{-m} options are defined for TMS320C3x/C4x implementations:
11545 @item -mcpu=@var{cpu_type}
11547 Set the instruction set, register set, and instruction scheduling
11548 parameters for machine type @var{cpu_type}. Supported values for
11549 @var{cpu_type} are @samp{c30}, @samp{c31}, @samp{c32}, @samp{c40}, and
11550 @samp{c44}. The default is @samp{c40} to generate code for the
11555 @itemx -msmall-memory
11557 @opindex mbig-memory
11559 @opindex msmall-memory
11561 Generates code for the big or small memory model. The small memory
11562 model assumed that all data fits into one 64K word page. At run-time
11563 the data page (DP) register must be set to point to the 64K page
11564 containing the .bss and .data program sections. The big memory model is
11565 the default and requires reloading of the DP register for every direct
11572 Allow (disallow) allocation of general integer operands into the block
11573 count register BK@.
11579 Enable (disable) generation of code using decrement and branch,
11580 DBcond(D), instructions. This is enabled by default for the C4x. To be
11581 on the safe side, this is disabled for the C3x, since the maximum
11582 iteration count on the C3x is @math{2^{23} + 1} (but who iterates loops more than
11583 @math{2^{23}} times on the C3x?). Note that GCC will try to reverse a loop so
11584 that it can utilize the decrement and branch instruction, but will give
11585 up if there is more than one memory reference in the loop. Thus a loop
11586 where the loop counter is decremented can generate slightly more
11587 efficient code, in cases where the RPTB instruction cannot be utilized.
11589 @item -mdp-isr-reload
11591 @opindex mdp-isr-reload
11593 Force the DP register to be saved on entry to an interrupt service
11594 routine (ISR), reloaded to point to the data section, and restored on
11595 exit from the ISR@. This should not be required unless someone has
11596 violated the small memory model by modifying the DP register, say within
11603 For the C3x use the 24-bit MPYI instruction for integer multiplies
11604 instead of a library call to guarantee 32-bit results. Note that if one
11605 of the operands is a constant, then the multiplication will be performed
11606 using shifts and adds. If the @option{-mmpyi} option is not specified for the C3x,
11607 then squaring operations are performed inline instead of a library call.
11610 @itemx -mno-fast-fix
11612 @opindex mno-fast-fix
11613 The C3x/C4x FIX instruction to convert a floating point value to an
11614 integer value chooses the nearest integer less than or equal to the
11615 floating point value rather than to the nearest integer. Thus if the
11616 floating point number is negative, the result will be incorrectly
11617 truncated an additional code is necessary to detect and correct this
11618 case. This option can be used to disable generation of the additional
11619 code required to correct the result.
11625 Enable (disable) generation of repeat block sequences using the RPTB
11626 instruction for zero overhead looping. The RPTB construct is only used
11627 for innermost loops that do not call functions or jump across the loop
11628 boundaries. There is no advantage having nested RPTB loops due to the
11629 overhead required to save and restore the RC, RS, and RE registers.
11630 This is enabled by default with @option{-O2}.
11632 @item -mrpts=@var{count}
11636 Enable (disable) the use of the single instruction repeat instruction
11637 RPTS@. If a repeat block contains a single instruction, and the loop
11638 count can be guaranteed to be less than the value @var{count}, GCC will
11639 emit a RPTS instruction instead of a RPTB@. If no value is specified,
11640 then a RPTS will be emitted even if the loop count cannot be determined
11641 at compile time. Note that the repeated instruction following RPTS does
11642 not have to be reloaded from memory each iteration, thus freeing up the
11643 CPU buses for operands. However, since interrupts are blocked by this
11644 instruction, it is disabled by default.
11646 @item -mloop-unsigned
11647 @itemx -mno-loop-unsigned
11648 @opindex mloop-unsigned
11649 @opindex mno-loop-unsigned
11650 The maximum iteration count when using RPTS and RPTB (and DB on the C40)
11651 is @math{2^{31} + 1} since these instructions test if the iteration count is
11652 negative to terminate the loop. If the iteration count is unsigned
11653 there is a possibility than the @math{2^{31} + 1} maximum iteration count may be
11654 exceeded. This switch allows an unsigned iteration count.
11658 Try to emit an assembler syntax that the TI assembler (asm30) is happy
11659 with. This also enforces compatibility with the API employed by the TI
11660 C3x C compiler. For example, long doubles are passed as structures
11661 rather than in floating point registers.
11667 Generate code that uses registers (stack) for passing arguments to functions.
11668 By default, arguments are passed in registers where possible rather
11669 than by pushing arguments on to the stack.
11671 @item -mparallel-insns
11672 @itemx -mno-parallel-insns
11673 @opindex mparallel-insns
11674 @opindex mno-parallel-insns
11675 Allow the generation of parallel instructions. This is enabled by
11676 default with @option{-O2}.
11678 @item -mparallel-mpy
11679 @itemx -mno-parallel-mpy
11680 @opindex mparallel-mpy
11681 @opindex mno-parallel-mpy
11682 Allow the generation of MPY||ADD and MPY||SUB parallel instructions,
11683 provided @option{-mparallel-insns} is also specified. These instructions have
11684 tight register constraints which can pessimize the code generation
11685 of large functions.
11690 @subsection V850 Options
11691 @cindex V850 Options
11693 These @samp{-m} options are defined for V850 implementations:
11697 @itemx -mno-long-calls
11698 @opindex mlong-calls
11699 @opindex mno-long-calls
11700 Treat all calls as being far away (near). If calls are assumed to be
11701 far away, the compiler will always load the functions address up into a
11702 register, and call indirect through the pointer.
11708 Do not optimize (do optimize) basic blocks that use the same index
11709 pointer 4 or more times to copy pointer into the @code{ep} register, and
11710 use the shorter @code{sld} and @code{sst} instructions. The @option{-mep}
11711 option is on by default if you optimize.
11713 @item -mno-prolog-function
11714 @itemx -mprolog-function
11715 @opindex mno-prolog-function
11716 @opindex mprolog-function
11717 Do not use (do use) external functions to save and restore registers
11718 at the prologue and epilogue of a function. The external functions
11719 are slower, but use less code space if more than one function saves
11720 the same number of registers. The @option{-mprolog-function} option
11721 is on by default if you optimize.
11725 Try to make the code as small as possible. At present, this just turns
11726 on the @option{-mep} and @option{-mprolog-function} options.
11728 @item -mtda=@var{n}
11730 Put static or global variables whose size is @var{n} bytes or less into
11731 the tiny data area that register @code{ep} points to. The tiny data
11732 area can hold up to 256 bytes in total (128 bytes for byte references).
11734 @item -msda=@var{n}
11736 Put static or global variables whose size is @var{n} bytes or less into
11737 the small data area that register @code{gp} points to. The small data
11738 area can hold up to 64 kilobytes.
11740 @item -mzda=@var{n}
11742 Put static or global variables whose size is @var{n} bytes or less into
11743 the first 32 kilobytes of memory.
11747 Specify that the target processor is the V850.
11750 @opindex mbig-switch
11751 Generate code suitable for big switch tables. Use this option only if
11752 the assembler/linker complain about out of range branches within a switch
11757 This option will cause r2 and r5 to be used in the code generated by
11758 the compiler. This setting is the default.
11760 @item -mno-app-regs
11761 @opindex mno-app-regs
11762 This option will cause r2 and r5 to be treated as fixed registers.
11766 Specify that the target processor is the V850E1. The preprocessor
11767 constants @samp{__v850e1__} and @samp{__v850e__} will be defined if
11768 this option is used.
11772 Specify that the target processor is the V850E@. The preprocessor
11773 constant @samp{__v850e__} will be defined if this option is used.
11775 If neither @option{-mv850} nor @option{-mv850e} nor @option{-mv850e1}
11776 are defined then a default target processor will be chosen and the
11777 relevant @samp{__v850*__} preprocessor constant will be defined.
11779 The preprocessor constants @samp{__v850} and @samp{__v851__} are always
11780 defined, regardless of which processor variant is the target.
11782 @item -mdisable-callt
11783 @opindex mdisable-callt
11784 This option will suppress generation of the CALLT instruction for the
11785 v850e and v850e1 flavors of the v850 architecture. The default is
11786 @option{-mno-disable-callt} which allows the CALLT instruction to be used.
11791 @subsection VAX Options
11792 @cindex VAX options
11794 These @samp{-m} options are defined for the VAX:
11799 Do not output certain jump instructions (@code{aobleq} and so on)
11800 that the Unix assembler for the VAX cannot handle across long
11805 Do output those jump instructions, on the assumption that you
11806 will assemble with the GNU assembler.
11810 Output code for g-format floating point numbers instead of d-format.
11813 @node x86-64 Options
11814 @subsection x86-64 Options
11815 @cindex x86-64 options
11817 These are listed under @xref{i386 and x86-64 Options}.
11819 @node Xstormy16 Options
11820 @subsection Xstormy16 Options
11821 @cindex Xstormy16 Options
11823 These options are defined for Xstormy16:
11828 Choose startup files and linker script suitable for the simulator.
11831 @node Xtensa Options
11832 @subsection Xtensa Options
11833 @cindex Xtensa Options
11835 These options are supported for Xtensa targets:
11839 @itemx -mno-const16
11841 @opindex mno-const16
11842 Enable or disable use of @code{CONST16} instructions for loading
11843 constant values. The @code{CONST16} instruction is currently not a
11844 standard option from Tensilica. When enabled, @code{CONST16}
11845 instructions are always used in place of the standard @code{L32R}
11846 instructions. The use of @code{CONST16} is enabled by default only if
11847 the @code{L32R} instruction is not available.
11850 @itemx -mno-fused-madd
11851 @opindex mfused-madd
11852 @opindex mno-fused-madd
11853 Enable or disable use of fused multiply/add and multiply/subtract
11854 instructions in the floating-point option. This has no effect if the
11855 floating-point option is not also enabled. Disabling fused multiply/add
11856 and multiply/subtract instructions forces the compiler to use separate
11857 instructions for the multiply and add/subtract operations. This may be
11858 desirable in some cases where strict IEEE 754-compliant results are
11859 required: the fused multiply add/subtract instructions do not round the
11860 intermediate result, thereby producing results with @emph{more} bits of
11861 precision than specified by the IEEE standard. Disabling fused multiply
11862 add/subtract instructions also ensures that the program output is not
11863 sensitive to the compiler's ability to combine multiply and add/subtract
11866 @item -mtext-section-literals
11867 @itemx -mno-text-section-literals
11868 @opindex mtext-section-literals
11869 @opindex mno-text-section-literals
11870 Control the treatment of literal pools. The default is
11871 @option{-mno-text-section-literals}, which places literals in a separate
11872 section in the output file. This allows the literal pool to be placed
11873 in a data RAM/ROM, and it also allows the linker to combine literal
11874 pools from separate object files to remove redundant literals and
11875 improve code size. With @option{-mtext-section-literals}, the literals
11876 are interspersed in the text section in order to keep them as close as
11877 possible to their references. This may be necessary for large assembly
11880 @item -mtarget-align
11881 @itemx -mno-target-align
11882 @opindex mtarget-align
11883 @opindex mno-target-align
11884 When this option is enabled, GCC instructs the assembler to
11885 automatically align instructions to reduce branch penalties at the
11886 expense of some code density. The assembler attempts to widen density
11887 instructions to align branch targets and the instructions following call
11888 instructions. If there are not enough preceding safe density
11889 instructions to align a target, no widening will be performed. The
11890 default is @option{-mtarget-align}. These options do not affect the
11891 treatment of auto-aligned instructions like @code{LOOP}, which the
11892 assembler will always align, either by widening density instructions or
11893 by inserting no-op instructions.
11896 @itemx -mno-longcalls
11897 @opindex mlongcalls
11898 @opindex mno-longcalls
11899 When this option is enabled, GCC instructs the assembler to translate
11900 direct calls to indirect calls unless it can determine that the target
11901 of a direct call is in the range allowed by the call instruction. This
11902 translation typically occurs for calls to functions in other source
11903 files. Specifically, the assembler translates a direct @code{CALL}
11904 instruction into an @code{L32R} followed by a @code{CALLX} instruction.
11905 The default is @option{-mno-longcalls}. This option should be used in
11906 programs where the call target can potentially be out of range. This
11907 option is implemented in the assembler, not the compiler, so the
11908 assembly code generated by GCC will still show direct call
11909 instructions---look at the disassembled object code to see the actual
11910 instructions. Note that the assembler will use an indirect call for
11911 every cross-file call, not just those that really will be out of range.
11914 @node zSeries Options
11915 @subsection zSeries Options
11916 @cindex zSeries options
11918 These are listed under @xref{S/390 and zSeries Options}.
11920 @node Code Gen Options
11921 @section Options for Code Generation Conventions
11922 @cindex code generation conventions
11923 @cindex options, code generation
11924 @cindex run-time options
11926 These machine-independent options control the interface conventions
11927 used in code generation.
11929 Most of them have both positive and negative forms; the negative form
11930 of @option{-ffoo} would be @option{-fno-foo}. In the table below, only
11931 one of the forms is listed---the one which is not the default. You
11932 can figure out the other form by either removing @samp{no-} or adding
11936 @item -fbounds-check
11937 @opindex fbounds-check
11938 For front-ends that support it, generate additional code to check that
11939 indices used to access arrays are within the declared range. This is
11940 currently only supported by the Java and Fortran 77 front-ends, where
11941 this option defaults to true and false respectively.
11945 This option generates traps for signed overflow on addition, subtraction,
11946 multiplication operations.
11950 This option instructs the compiler to assume that signed arithmetic
11951 overflow of addition, subtraction and multiplication wraps around
11952 using twos-complement representation. This flag enables some optimizations
11953 and disables other. This option is enabled by default for the Java
11954 front-end, as required by the Java language specification.
11957 @opindex fexceptions
11958 Enable exception handling. Generates extra code needed to propagate
11959 exceptions. For some targets, this implies GCC will generate frame
11960 unwind information for all functions, which can produce significant data
11961 size overhead, although it does not affect execution. If you do not
11962 specify this option, GCC will enable it by default for languages like
11963 C++ which normally require exception handling, and disable it for
11964 languages like C that do not normally require it. However, you may need
11965 to enable this option when compiling C code that needs to interoperate
11966 properly with exception handlers written in C++. You may also wish to
11967 disable this option if you are compiling older C++ programs that don't
11968 use exception handling.
11970 @item -fnon-call-exceptions
11971 @opindex fnon-call-exceptions
11972 Generate code that allows trapping instructions to throw exceptions.
11973 Note that this requires platform-specific runtime support that does
11974 not exist everywhere. Moreover, it only allows @emph{trapping}
11975 instructions to throw exceptions, i.e.@: memory references or floating
11976 point instructions. It does not allow exceptions to be thrown from
11977 arbitrary signal handlers such as @code{SIGALRM}.
11979 @item -funwind-tables
11980 @opindex funwind-tables
11981 Similar to @option{-fexceptions}, except that it will just generate any needed
11982 static data, but will not affect the generated code in any other way.
11983 You will normally not enable this option; instead, a language processor
11984 that needs this handling would enable it on your behalf.
11986 @item -fasynchronous-unwind-tables
11987 @opindex fasynchronous-unwind-tables
11988 Generate unwind table in dwarf2 format, if supported by target machine. The
11989 table is exact at each instruction boundary, so it can be used for stack
11990 unwinding from asynchronous events (such as debugger or garbage collector).
11992 @item -fpcc-struct-return
11993 @opindex fpcc-struct-return
11994 Return ``short'' @code{struct} and @code{union} values in memory like
11995 longer ones, rather than in registers. This convention is less
11996 efficient, but it has the advantage of allowing intercallability between
11997 GCC-compiled files and files compiled with other compilers, particularly
11998 the Portable C Compiler (pcc).
12000 The precise convention for returning structures in memory depends
12001 on the target configuration macros.
12003 Short structures and unions are those whose size and alignment match
12004 that of some integer type.
12006 @strong{Warning:} code compiled with the @option{-fpcc-struct-return}
12007 switch is not binary compatible with code compiled with the
12008 @option{-freg-struct-return} switch.
12009 Use it to conform to a non-default application binary interface.
12011 @item -freg-struct-return
12012 @opindex freg-struct-return
12013 Return @code{struct} and @code{union} values in registers when possible.
12014 This is more efficient for small structures than
12015 @option{-fpcc-struct-return}.
12017 If you specify neither @option{-fpcc-struct-return} nor
12018 @option{-freg-struct-return}, GCC defaults to whichever convention is
12019 standard for the target. If there is no standard convention, GCC
12020 defaults to @option{-fpcc-struct-return}, except on targets where GCC is
12021 the principal compiler. In those cases, we can choose the standard, and
12022 we chose the more efficient register return alternative.
12024 @strong{Warning:} code compiled with the @option{-freg-struct-return}
12025 switch is not binary compatible with code compiled with the
12026 @option{-fpcc-struct-return} switch.
12027 Use it to conform to a non-default application binary interface.
12029 @item -fshort-enums
12030 @opindex fshort-enums
12031 Allocate to an @code{enum} type only as many bytes as it needs for the
12032 declared range of possible values. Specifically, the @code{enum} type
12033 will be equivalent to the smallest integer type which has enough room.
12035 @strong{Warning:} the @option{-fshort-enums} switch causes GCC to generate
12036 code that is not binary compatible with code generated without that switch.
12037 Use it to conform to a non-default application binary interface.
12039 @item -fshort-double
12040 @opindex fshort-double
12041 Use the same size for @code{double} as for @code{float}.
12043 @strong{Warning:} the @option{-fshort-double} switch causes GCC to generate
12044 code that is not binary compatible with code generated without that switch.
12045 Use it to conform to a non-default application binary interface.
12047 @item -fshort-wchar
12048 @opindex fshort-wchar
12049 Override the underlying type for @samp{wchar_t} to be @samp{short
12050 unsigned int} instead of the default for the target. This option is
12051 useful for building programs to run under WINE@.
12053 @strong{Warning:} the @option{-fshort-wchar} switch causes GCC to generate
12054 code that is not binary compatible with code generated without that switch.
12055 Use it to conform to a non-default application binary interface.
12057 @item -fshared-data
12058 @opindex fshared-data
12059 Requests that the data and non-@code{const} variables of this
12060 compilation be shared data rather than private data. The distinction
12061 makes sense only on certain operating systems, where shared data is
12062 shared between processes running the same program, while private data
12063 exists in one copy per process.
12066 @opindex fno-common
12067 In C, allocate even uninitialized global variables in the data section of the
12068 object file, rather than generating them as common blocks. This has the
12069 effect that if the same variable is declared (without @code{extern}) in
12070 two different compilations, you will get an error when you link them.
12071 The only reason this might be useful is if you wish to verify that the
12072 program will work on other systems which always work this way.
12076 Ignore the @samp{#ident} directive.
12078 @item -finhibit-size-directive
12079 @opindex finhibit-size-directive
12080 Don't output a @code{.size} assembler directive, or anything else that
12081 would cause trouble if the function is split in the middle, and the
12082 two halves are placed at locations far apart in memory. This option is
12083 used when compiling @file{crtstuff.c}; you should not need to use it
12086 @item -fverbose-asm
12087 @opindex fverbose-asm
12088 Put extra commentary information in the generated assembly code to
12089 make it more readable. This option is generally only of use to those
12090 who actually need to read the generated assembly code (perhaps while
12091 debugging the compiler itself).
12093 @option{-fno-verbose-asm}, the default, causes the
12094 extra information to be omitted and is useful when comparing two assembler
12099 @cindex global offset table
12101 Generate position-independent code (PIC) suitable for use in a shared
12102 library, if supported for the target machine. Such code accesses all
12103 constant addresses through a global offset table (GOT)@. The dynamic
12104 loader resolves the GOT entries when the program starts (the dynamic
12105 loader is not part of GCC; it is part of the operating system). If
12106 the GOT size for the linked executable exceeds a machine-specific
12107 maximum size, you get an error message from the linker indicating that
12108 @option{-fpic} does not work; in that case, recompile with @option{-fPIC}
12109 instead. (These maximums are 8k on the SPARC and 32k
12110 on the m68k and RS/6000. The 386 has no such limit.)
12112 Position-independent code requires special support, and therefore works
12113 only on certain machines. For the 386, GCC supports PIC for System V
12114 but not for the Sun 386i. Code generated for the IBM RS/6000 is always
12115 position-independent.
12119 If supported for the target machine, emit position-independent code,
12120 suitable for dynamic linking and avoiding any limit on the size of the
12121 global offset table. This option makes a difference on the m68k,
12122 PowerPC and SPARC@.
12124 Position-independent code requires special support, and therefore works
12125 only on certain machines.
12131 These options are similar to @option{-fpic} and @option{-fPIC}, but
12132 generated position independent code can be only linked into executables.
12133 Usually these options are used when @option{-pie} GCC option will be
12134 used during linking.
12136 @item -ffixed-@var{reg}
12138 Treat the register named @var{reg} as a fixed register; generated code
12139 should never refer to it (except perhaps as a stack pointer, frame
12140 pointer or in some other fixed role).
12142 @var{reg} must be the name of a register. The register names accepted
12143 are machine-specific and are defined in the @code{REGISTER_NAMES}
12144 macro in the machine description macro file.
12146 This flag does not have a negative form, because it specifies a
12149 @item -fcall-used-@var{reg}
12150 @opindex fcall-used
12151 Treat the register named @var{reg} as an allocable register that is
12152 clobbered by function calls. It may be allocated for temporaries or
12153 variables that do not live across a call. Functions compiled this way
12154 will not save and restore the register @var{reg}.
12156 It is an error to used this flag with the frame pointer or stack pointer.
12157 Use of this flag for other registers that have fixed pervasive roles in
12158 the machine's execution model will produce disastrous results.
12160 This flag does not have a negative form, because it specifies a
12163 @item -fcall-saved-@var{reg}
12164 @opindex fcall-saved
12165 Treat the register named @var{reg} as an allocable register saved by
12166 functions. It may be allocated even for temporaries or variables that
12167 live across a call. Functions compiled this way will save and restore
12168 the register @var{reg} if they use it.
12170 It is an error to used this flag with the frame pointer or stack pointer.
12171 Use of this flag for other registers that have fixed pervasive roles in
12172 the machine's execution model will produce disastrous results.
12174 A different sort of disaster will result from the use of this flag for
12175 a register in which function values may be returned.
12177 This flag does not have a negative form, because it specifies a
12180 @item -fpack-struct[=@var{n}]
12181 @opindex fpack-struct
12182 Without a value specified, pack all structure members together without
12183 holes. When a value is specified (which must be a small power of two), pack
12184 structure members according to this value, representing the maximum
12185 alignment (that is, objects with default alignment requirements larger than
12186 this will be output potentially unaligned at the next fitting location.
12188 @strong{Warning:} the @option{-fpack-struct} switch causes GCC to generate
12189 code that is not binary compatible with code generated without that switch.
12190 Additionally, it makes the code suboptimal.
12191 Use it to conform to a non-default application binary interface.
12193 @item -finstrument-functions
12194 @opindex finstrument-functions
12195 Generate instrumentation calls for entry and exit to functions. Just
12196 after function entry and just before function exit, the following
12197 profiling functions will be called with the address of the current
12198 function and its call site. (On some platforms,
12199 @code{__builtin_return_address} does not work beyond the current
12200 function, so the call site information may not be available to the
12201 profiling functions otherwise.)
12204 void __cyg_profile_func_enter (void *this_fn,
12206 void __cyg_profile_func_exit (void *this_fn,
12210 The first argument is the address of the start of the current function,
12211 which may be looked up exactly in the symbol table.
12213 This instrumentation is also done for functions expanded inline in other
12214 functions. The profiling calls will indicate where, conceptually, the
12215 inline function is entered and exited. This means that addressable
12216 versions of such functions must be available. If all your uses of a
12217 function are expanded inline, this may mean an additional expansion of
12218 code size. If you use @samp{extern inline} in your C code, an
12219 addressable version of such functions must be provided. (This is
12220 normally the case anyways, but if you get lucky and the optimizer always
12221 expands the functions inline, you might have gotten away without
12222 providing static copies.)
12224 A function may be given the attribute @code{no_instrument_function}, in
12225 which case this instrumentation will not be done. This can be used, for
12226 example, for the profiling functions listed above, high-priority
12227 interrupt routines, and any functions from which the profiling functions
12228 cannot safely be called (perhaps signal handlers, if the profiling
12229 routines generate output or allocate memory).
12231 @item -fstack-check
12232 @opindex fstack-check
12233 Generate code to verify that you do not go beyond the boundary of the
12234 stack. You should specify this flag if you are running in an
12235 environment with multiple threads, but only rarely need to specify it in
12236 a single-threaded environment since stack overflow is automatically
12237 detected on nearly all systems if there is only one stack.
12239 Note that this switch does not actually cause checking to be done; the
12240 operating system must do that. The switch causes generation of code
12241 to ensure that the operating system sees the stack being extended.
12243 @item -fstack-limit-register=@var{reg}
12244 @itemx -fstack-limit-symbol=@var{sym}
12245 @itemx -fno-stack-limit
12246 @opindex fstack-limit-register
12247 @opindex fstack-limit-symbol
12248 @opindex fno-stack-limit
12249 Generate code to ensure that the stack does not grow beyond a certain value,
12250 either the value of a register or the address of a symbol. If the stack
12251 would grow beyond the value, a signal is raised. For most targets,
12252 the signal is raised before the stack overruns the boundary, so
12253 it is possible to catch the signal without taking special precautions.
12255 For instance, if the stack starts at absolute address @samp{0x80000000}
12256 and grows downwards, you can use the flags
12257 @option{-fstack-limit-symbol=__stack_limit} and
12258 @option{-Wl,--defsym,__stack_limit=0x7ffe0000} to enforce a stack limit
12259 of 128KB@. Note that this may only work with the GNU linker.
12261 @cindex aliasing of parameters
12262 @cindex parameters, aliased
12263 @item -fargument-alias
12264 @itemx -fargument-noalias
12265 @itemx -fargument-noalias-global
12266 @opindex fargument-alias
12267 @opindex fargument-noalias
12268 @opindex fargument-noalias-global
12269 Specify the possible relationships among parameters and between
12270 parameters and global data.
12272 @option{-fargument-alias} specifies that arguments (parameters) may
12273 alias each other and may alias global storage.@*
12274 @option{-fargument-noalias} specifies that arguments do not alias
12275 each other, but may alias global storage.@*
12276 @option{-fargument-noalias-global} specifies that arguments do not
12277 alias each other and do not alias global storage.
12279 Each language will automatically use whatever option is required by
12280 the language standard. You should not need to use these options yourself.
12282 @item -fleading-underscore
12283 @opindex fleading-underscore
12284 This option and its counterpart, @option{-fno-leading-underscore}, forcibly
12285 change the way C symbols are represented in the object file. One use
12286 is to help link with legacy assembly code.
12288 @strong{Warning:} the @option{-fleading-underscore} switch causes GCC to
12289 generate code that is not binary compatible with code generated without that
12290 switch. Use it to conform to a non-default application binary interface.
12291 Not all targets provide complete support for this switch.
12293 @item -ftls-model=@var{model}
12294 Alter the thread-local storage model to be used (@pxref{Thread-Local}).
12295 The @var{model} argument should be one of @code{global-dynamic},
12296 @code{local-dynamic}, @code{initial-exec} or @code{local-exec}.
12298 The default without @option{-fpic} is @code{initial-exec}; with
12299 @option{-fpic} the default is @code{global-dynamic}.
12301 @item -fvisibility=@var{default|internal|hidden|protected}
12302 @opindex fvisibility
12303 Set the default ELF image symbol visibility to the specified option---all
12304 symbols will be marked with this unless overridden within the code.
12305 Using this feature can very substantially improve linking and
12306 load times of shared object libraries, produce more optimized
12307 code, provide near-perfect API export and prevent symbol clashes.
12308 It is @strong{strongly} recommended that you use this in any shared objects
12311 Despite the nomenclature, @code{default} always means public ie;
12312 available to be linked against from outside the shared object.
12313 @code{protected} and @code{internal} are pretty useless in real-world
12314 usage so the only other commonly used option will be @code{hidden}.
12315 The default if @option{-fvisibility} isn't specified is
12316 @code{default}, i.e., make every
12317 symbol public---this causes the same behavior as previous versions of
12320 A good explanation of the benefits offered by ensuring ELF
12321 symbols have the correct visibility is given by ``How To Write
12322 Shared Libraries'' by Ulrich Drepper (which can be found at
12323 @w{@uref{http://people.redhat.com/~drepper/}})---however a superior
12324 solution made possible by this option to marking things hidden when
12325 the default is public is to make the default hidden and mark things
12326 public. This is the norm with DLL's on Windows and with @option{-fvisibility=hidden}
12327 and @code{__attribute__ ((visibility("default")))} instead of
12328 @code{__declspec(dllexport)} you get almost identical semantics with
12329 identical syntax. This is a great boon to those working with
12330 cross-platform projects.
12332 For those adding visibility support to existing code, you may find
12333 @samp{#pragma GCC visibility} of use. This works by you enclosing
12334 the declarations you wish to set visibility for with (for example)
12335 @samp{#pragma GCC visibility push(hidden)} and
12336 @samp{#pragma GCC visibility pop}. These can be nested up to sixteen
12337 times. Bear in mind that symbol visibility should be viewed @strong{as
12338 part of the API interface contract} and thus all new code should
12339 always specify visibility when it is not the default ie; declarations
12340 only for use within the local DSO should @strong{always} be marked explicitly
12341 as hidden as so to avoid PLT indirection overheads---making this
12342 abundantly clear also aids readability and self-documentation of the code.
12343 Note that due to ISO C++ specification requirements, operator new and
12344 operator delete must always be of default visibility.
12346 An overview of these techniques, their benefits and how to use them
12347 is at @w{@uref{http://www.nedprod.com/programs/gccvisibility.html}}.
12353 @node Environment Variables
12354 @section Environment Variables Affecting GCC
12355 @cindex environment variables
12357 @c man begin ENVIRONMENT
12358 This section describes several environment variables that affect how GCC
12359 operates. Some of them work by specifying directories or prefixes to use
12360 when searching for various kinds of files. Some are used to specify other
12361 aspects of the compilation environment.
12363 Note that you can also specify places to search using options such as
12364 @option{-B}, @option{-I} and @option{-L} (@pxref{Directory Options}). These
12365 take precedence over places specified using environment variables, which
12366 in turn take precedence over those specified by the configuration of GCC@.
12367 @xref{Driver,, Controlling the Compilation Driver @file{gcc}, gccint,
12368 GNU Compiler Collection (GCC) Internals}.
12373 @c @itemx LC_COLLATE
12375 @c @itemx LC_MONETARY
12376 @c @itemx LC_NUMERIC
12381 @c @findex LC_COLLATE
12382 @findex LC_MESSAGES
12383 @c @findex LC_MONETARY
12384 @c @findex LC_NUMERIC
12388 These environment variables control the way that GCC uses
12389 localization information that allow GCC to work with different
12390 national conventions. GCC inspects the locale categories
12391 @env{LC_CTYPE} and @env{LC_MESSAGES} if it has been configured to do
12392 so. These locale categories can be set to any value supported by your
12393 installation. A typical value is @samp{en_GB.UTF-8} for English in the United
12394 Kingdom encoded in UTF-8.
12396 The @env{LC_CTYPE} environment variable specifies character
12397 classification. GCC uses it to determine the character boundaries in
12398 a string; this is needed for some multibyte encodings that contain quote
12399 and escape characters that would otherwise be interpreted as a string
12402 The @env{LC_MESSAGES} environment variable specifies the language to
12403 use in diagnostic messages.
12405 If the @env{LC_ALL} environment variable is set, it overrides the value
12406 of @env{LC_CTYPE} and @env{LC_MESSAGES}; otherwise, @env{LC_CTYPE}
12407 and @env{LC_MESSAGES} default to the value of the @env{LANG}
12408 environment variable. If none of these variables are set, GCC
12409 defaults to traditional C English behavior.
12413 If @env{TMPDIR} is set, it specifies the directory to use for temporary
12414 files. GCC uses temporary files to hold the output of one stage of
12415 compilation which is to be used as input to the next stage: for example,
12416 the output of the preprocessor, which is the input to the compiler
12419 @item GCC_EXEC_PREFIX
12420 @findex GCC_EXEC_PREFIX
12421 If @env{GCC_EXEC_PREFIX} is set, it specifies a prefix to use in the
12422 names of the subprograms executed by the compiler. No slash is added
12423 when this prefix is combined with the name of a subprogram, but you can
12424 specify a prefix that ends with a slash if you wish.
12426 If @env{GCC_EXEC_PREFIX} is not set, GCC will attempt to figure out
12427 an appropriate prefix to use based on the pathname it was invoked with.
12429 If GCC cannot find the subprogram using the specified prefix, it
12430 tries looking in the usual places for the subprogram.
12432 The default value of @env{GCC_EXEC_PREFIX} is
12433 @file{@var{prefix}/lib/gcc/} where @var{prefix} is the value
12434 of @code{prefix} when you ran the @file{configure} script.
12436 Other prefixes specified with @option{-B} take precedence over this prefix.
12438 This prefix is also used for finding files such as @file{crt0.o} that are
12441 In addition, the prefix is used in an unusual way in finding the
12442 directories to search for header files. For each of the standard
12443 directories whose name normally begins with @samp{/usr/local/lib/gcc}
12444 (more precisely, with the value of @env{GCC_INCLUDE_DIR}), GCC tries
12445 replacing that beginning with the specified prefix to produce an
12446 alternate directory name. Thus, with @option{-Bfoo/}, GCC will search
12447 @file{foo/bar} where it would normally search @file{/usr/local/lib/bar}.
12448 These alternate directories are searched first; the standard directories
12451 @item COMPILER_PATH
12452 @findex COMPILER_PATH
12453 The value of @env{COMPILER_PATH} is a colon-separated list of
12454 directories, much like @env{PATH}. GCC tries the directories thus
12455 specified when searching for subprograms, if it can't find the
12456 subprograms using @env{GCC_EXEC_PREFIX}.
12459 @findex LIBRARY_PATH
12460 The value of @env{LIBRARY_PATH} is a colon-separated list of
12461 directories, much like @env{PATH}. When configured as a native compiler,
12462 GCC tries the directories thus specified when searching for special
12463 linker files, if it can't find them using @env{GCC_EXEC_PREFIX}. Linking
12464 using GCC also uses these directories when searching for ordinary
12465 libraries for the @option{-l} option (but directories specified with
12466 @option{-L} come first).
12470 @cindex locale definition
12471 This variable is used to pass locale information to the compiler. One way in
12472 which this information is used is to determine the character set to be used
12473 when character literals, string literals and comments are parsed in C and C++.
12474 When the compiler is configured to allow multibyte characters,
12475 the following values for @env{LANG} are recognized:
12479 Recognize JIS characters.
12481 Recognize SJIS characters.
12483 Recognize EUCJP characters.
12486 If @env{LANG} is not defined, or if it has some other value, then the
12487 compiler will use mblen and mbtowc as defined by the default locale to
12488 recognize and translate multibyte characters.
12492 Some additional environments variables affect the behavior of the
12495 @include cppenv.texi
12499 @node Precompiled Headers
12500 @section Using Precompiled Headers
12501 @cindex precompiled headers
12502 @cindex speed of compilation
12504 Often large projects have many header files that are included in every
12505 source file. The time the compiler takes to process these header files
12506 over and over again can account for nearly all of the time required to
12507 build the project. To make builds faster, GCC allows users to
12508 `precompile' a header file; then, if builds can use the precompiled
12509 header file they will be much faster.
12511 @strong{Caution:} There are a few known situations where GCC will
12512 crash when trying to use a precompiled header. If you have trouble
12513 with a precompiled header, you should remove the precompiled header
12514 and compile without it. In addition, please use GCC's on-line
12515 defect-tracking system to report any problems you encounter with
12516 precompiled headers. @xref{Bugs}.
12518 To create a precompiled header file, simply compile it as you would any
12519 other file, if necessary using the @option{-x} option to make the driver
12520 treat it as a C or C++ header file. You will probably want to use a
12521 tool like @command{make} to keep the precompiled header up-to-date when
12522 the headers it contains change.
12524 A precompiled header file will be searched for when @code{#include} is
12525 seen in the compilation. As it searches for the included file
12526 (@pxref{Search Path,,Search Path,cpp,The C Preprocessor}) the
12527 compiler looks for a precompiled header in each directory just before it
12528 looks for the include file in that directory. The name searched for is
12529 the name specified in the @code{#include} with @samp{.gch} appended. If
12530 the precompiled header file can't be used, it is ignored.
12532 For instance, if you have @code{#include "all.h"}, and you have
12533 @file{all.h.gch} in the same directory as @file{all.h}, then the
12534 precompiled header file will be used if possible, and the original
12535 header will be used otherwise.
12537 Alternatively, you might decide to put the precompiled header file in a
12538 directory and use @option{-I} to ensure that directory is searched
12539 before (or instead of) the directory containing the original header.
12540 Then, if you want to check that the precompiled header file is always
12541 used, you can put a file of the same name as the original header in this
12542 directory containing an @code{#error} command.
12544 This also works with @option{-include}. So yet another way to use
12545 precompiled headers, good for projects not designed with precompiled
12546 header files in mind, is to simply take most of the header files used by
12547 a project, include them from another header file, precompile that header
12548 file, and @option{-include} the precompiled header. If the header files
12549 have guards against multiple inclusion, they will be skipped because
12550 they've already been included (in the precompiled header).
12552 If you need to precompile the same header file for different
12553 languages, targets, or compiler options, you can instead make a
12554 @emph{directory} named like @file{all.h.gch}, and put each precompiled
12555 header in the directory, perhaps using @option{-o}. It doesn't matter
12556 what you call the files in the directory, every precompiled header in
12557 the directory will be considered. The first precompiled header
12558 encountered in the directory that is valid for this compilation will
12559 be used; they're searched in no particular order.
12561 There are many other possibilities, limited only by your imagination,
12562 good sense, and the constraints of your build system.
12564 A precompiled header file can be used only when these conditions apply:
12568 Only one precompiled header can be used in a particular compilation.
12571 A precompiled header can't be used once the first C token is seen. You
12572 can have preprocessor directives before a precompiled header; you can
12573 even include a precompiled header from inside another header, so long as
12574 there are no C tokens before the @code{#include}.
12577 The precompiled header file must be produced for the same language as
12578 the current compilation. You can't use a C precompiled header for a C++
12582 The precompiled header file must be produced by the same compiler
12583 version and configuration as the current compilation is using.
12584 The easiest way to guarantee this is to use the same compiler binary
12585 for creating and using precompiled headers.
12588 Any macros defined before the precompiled header is included must
12589 either be defined in the same way as when the precompiled header was
12590 generated, or must not affect the precompiled header, which usually
12591 means that the they don't appear in the precompiled header at all.
12593 The @option{-D} option is one way to define a macro before a
12594 precompiled header is included; using a @code{#define} can also do it.
12595 There are also some options that define macros implicitly, like
12596 @option{-O} and @option{-Wdeprecated}; the same rule applies to macros
12599 @item If debugging information is output when using the precompiled
12600 header, using @option{-g} or similar, the same kind of debugging information
12601 must have been output when building the precompiled header. However,
12602 a precompiled header built using @option{-g} can be used in a compilation
12603 when no debugging information is being output.
12605 @item The same @option{-m} options must generally be used when building
12606 and using the precompiled header. @xref{Submodel Options},
12607 for any cases where this rule is relaxed.
12609 @item Each of the following options must be the same when building and using
12610 the precompiled header:
12612 @gccoptlist{-fexceptions -funit-at-a-time}
12615 Some other command-line options starting with @option{-f},
12616 @option{-p}, or @option{-O} must be defined in the same way as when
12617 the precompiled header was generated. At present, it's not clear
12618 which options are safe to change and which are not; the safest choice
12619 is to use exactly the same options when generating and using the
12620 precompiled header. The following are known to be safe:
12622 @gccoptlist{-fpreprocessed -pedantic-errors}
12626 For all of these except the last, the compiler will automatically
12627 ignore the precompiled header if the conditions aren't met. If you
12628 find an option combination that doesn't work and doesn't cause the
12629 precompiled header to be ignored, please consider filing a bug report,
12632 If you do use differing options when generating and using the
12633 precompiled header, the actual behavior will be a mixture of the
12634 behavior for the options. For instance, if you use @option{-g} to
12635 generate the precompiled header but not when using it, you may or may
12636 not get debugging information for routines in the precompiled header.
12638 @node Running Protoize
12639 @section Running Protoize
12641 The program @code{protoize} is an optional part of GCC@. You can use
12642 it to add prototypes to a program, thus converting the program to ISO
12643 C in one respect. The companion program @code{unprotoize} does the
12644 reverse: it removes argument types from any prototypes that are found.
12646 When you run these programs, you must specify a set of source files as
12647 command line arguments. The conversion programs start out by compiling
12648 these files to see what functions they define. The information gathered
12649 about a file @var{foo} is saved in a file named @file{@var{foo}.X}.
12651 After scanning comes actual conversion. The specified files are all
12652 eligible to be converted; any files they include (whether sources or
12653 just headers) are eligible as well.
12655 But not all the eligible files are converted. By default,
12656 @code{protoize} and @code{unprotoize} convert only source and header
12657 files in the current directory. You can specify additional directories
12658 whose files should be converted with the @option{-d @var{directory}}
12659 option. You can also specify particular files to exclude with the
12660 @option{-x @var{file}} option. A file is converted if it is eligible, its
12661 directory name matches one of the specified directory names, and its
12662 name within the directory has not been excluded.
12664 Basic conversion with @code{protoize} consists of rewriting most
12665 function definitions and function declarations to specify the types of
12666 the arguments. The only ones not rewritten are those for varargs
12669 @code{protoize} optionally inserts prototype declarations at the
12670 beginning of the source file, to make them available for any calls that
12671 precede the function's definition. Or it can insert prototype
12672 declarations with block scope in the blocks where undeclared functions
12675 Basic conversion with @code{unprotoize} consists of rewriting most
12676 function declarations to remove any argument types, and rewriting
12677 function definitions to the old-style pre-ISO form.
12679 Both conversion programs print a warning for any function declaration or
12680 definition that they can't convert. You can suppress these warnings
12683 The output from @code{protoize} or @code{unprotoize} replaces the
12684 original source file. The original file is renamed to a name ending
12685 with @samp{.save} (for DOS, the saved filename ends in @samp{.sav}
12686 without the original @samp{.c} suffix). If the @samp{.save} (@samp{.sav}
12687 for DOS) file already exists, then the source file is simply discarded.
12689 @code{protoize} and @code{unprotoize} both depend on GCC itself to
12690 scan the program and collect information about the functions it uses.
12691 So neither of these programs will work until GCC is installed.
12693 Here is a table of the options you can use with @code{protoize} and
12694 @code{unprotoize}. Each option works with both programs unless
12698 @item -B @var{directory}
12699 Look for the file @file{SYSCALLS.c.X} in @var{directory}, instead of the
12700 usual directory (normally @file{/usr/local/lib}). This file contains
12701 prototype information about standard system functions. This option
12702 applies only to @code{protoize}.
12704 @item -c @var{compilation-options}
12705 Use @var{compilation-options} as the options when running @command{gcc} to
12706 produce the @samp{.X} files. The special option @option{-aux-info} is
12707 always passed in addition, to tell @command{gcc} to write a @samp{.X} file.
12709 Note that the compilation options must be given as a single argument to
12710 @code{protoize} or @code{unprotoize}. If you want to specify several
12711 @command{gcc} options, you must quote the entire set of compilation options
12712 to make them a single word in the shell.
12714 There are certain @command{gcc} arguments that you cannot use, because they
12715 would produce the wrong kind of output. These include @option{-g},
12716 @option{-O}, @option{-c}, @option{-S}, and @option{-o} If you include these in
12717 the @var{compilation-options}, they are ignored.
12720 Rename files to end in @samp{.C} (@samp{.cc} for DOS-based file
12721 systems) instead of @samp{.c}. This is convenient if you are converting
12722 a C program to C++. This option applies only to @code{protoize}.
12725 Add explicit global declarations. This means inserting explicit
12726 declarations at the beginning of each source file for each function
12727 that is called in the file and was not declared. These declarations
12728 precede the first function definition that contains a call to an
12729 undeclared function. This option applies only to @code{protoize}.
12731 @item -i @var{string}
12732 Indent old-style parameter declarations with the string @var{string}.
12733 This option applies only to @code{protoize}.
12735 @code{unprotoize} converts prototyped function definitions to old-style
12736 function definitions, where the arguments are declared between the
12737 argument list and the initial @samp{@{}. By default, @code{unprotoize}
12738 uses five spaces as the indentation. If you want to indent with just
12739 one space instead, use @option{-i " "}.
12742 Keep the @samp{.X} files. Normally, they are deleted after conversion
12746 Add explicit local declarations. @code{protoize} with @option{-l} inserts
12747 a prototype declaration for each function in each block which calls the
12748 function without any declaration. This option applies only to
12752 Make no real changes. This mode just prints information about the conversions
12753 that would have been done without @option{-n}.
12756 Make no @samp{.save} files. The original files are simply deleted.
12757 Use this option with caution.
12759 @item -p @var{program}
12760 Use the program @var{program} as the compiler. Normally, the name
12761 @file{gcc} is used.
12764 Work quietly. Most warnings are suppressed.
12767 Print the version number, just like @option{-v} for @command{gcc}.
12770 If you need special compiler options to compile one of your program's
12771 source files, then you should generate that file's @samp{.X} file
12772 specially, by running @command{gcc} on that source file with the
12773 appropriate options and the option @option{-aux-info}. Then run
12774 @code{protoize} on the entire set of files. @code{protoize} will use
12775 the existing @samp{.X} file because it is newer than the source file.
12779 gcc -Dfoo=bar file1.c -aux-info file1.X
12784 You need to include the special files along with the rest in the
12785 @code{protoize} command, even though their @samp{.X} files already
12786 exist, because otherwise they won't get converted.
12788 @xref{Protoize Caveats}, for more information on how to use
12789 @code{protoize} successfully.