1 @c Copyright (C) 1988, 1989, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
2 @c 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008
3 @c Free Software Foundation, Inc.
4 @c This is part of the GCC manual.
5 @c For copying conditions, see the file gcc.texi.
12 @c man begin COPYRIGHT
13 Copyright @copyright{} 1988, 1989, 1992, 1993, 1994, 1995, 1996, 1997, 1998,
14 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008
15 Free Software Foundation, Inc.
17 Permission is granted to copy, distribute and/or modify this document
18 under the terms of the GNU Free Documentation License, Version 1.2 or
19 any later version published by the Free Software Foundation; with the
20 Invariant Sections being ``GNU General Public License'' and ``Funding
21 Free Software'', the Front-Cover texts being (a) (see below), and with
22 the Back-Cover Texts being (b) (see below). A copy of the license is
23 included in the gfdl(7) man page.
25 (a) The FSF's Front-Cover Text is:
29 (b) The FSF's Back-Cover Text is:
31 You have freedom to copy and modify this GNU Manual, like GNU
32 software. Copies published by the Free Software Foundation raise
33 funds for GNU development.
35 @c Set file name and title for the man page.
37 @settitle GNU project C and C++ compiler
39 gcc [@option{-c}|@option{-S}|@option{-E}] [@option{-std=}@var{standard}]
40 [@option{-g}] [@option{-pg}] [@option{-O}@var{level}]
41 [@option{-W}@var{warn}@dots{}] [@option{-pedantic}]
42 [@option{-I}@var{dir}@dots{}] [@option{-L}@var{dir}@dots{}]
43 [@option{-D}@var{macro}[=@var{defn}]@dots{}] [@option{-U}@var{macro}]
44 [@option{-f}@var{option}@dots{}] [@option{-m}@var{machine-option}@dots{}]
45 [@option{-o} @var{outfile}] [@@@var{file}] @var{infile}@dots{}
47 Only the most useful options are listed here; see below for the
48 remainder. @samp{g++} accepts mostly the same options as @samp{gcc}.
51 gpl(7), gfdl(7), fsf-funding(7),
52 cpp(1), gcov(1), as(1), ld(1), gdb(1), adb(1), dbx(1), sdb(1)
53 and the Info entries for @file{gcc}, @file{cpp}, @file{as},
54 @file{ld}, @file{binutils} and @file{gdb}.
57 For instructions on reporting bugs, see
61 See the Info entry for @command{gcc}, or
62 @w{@uref{http://gcc.gnu.org/onlinedocs/gcc/Contributors.html}},
63 for contributors to GCC@.
68 @chapter GCC Command Options
69 @cindex GCC command options
70 @cindex command options
71 @cindex options, GCC command
73 @c man begin DESCRIPTION
74 When you invoke GCC, it normally does preprocessing, compilation,
75 assembly and linking. The ``overall options'' allow you to stop this
76 process at an intermediate stage. For example, the @option{-c} option
77 says not to run the linker. Then the output consists of object files
78 output by the assembler.
80 Other options are passed on to one stage of processing. Some options
81 control the preprocessor and others the compiler itself. Yet other
82 options control the assembler and linker; most of these are not
83 documented here, since you rarely need to use any of them.
85 @cindex C compilation options
86 Most of the command line options that you can use with GCC are useful
87 for C programs; when an option is only useful with another language
88 (usually C++), the explanation says so explicitly. If the description
89 for a particular option does not mention a source language, you can use
90 that option with all supported languages.
92 @cindex C++ compilation options
93 @xref{Invoking G++,,Compiling C++ Programs}, for a summary of special
94 options for compiling C++ programs.
96 @cindex grouping options
97 @cindex options, grouping
98 The @command{gcc} program accepts options and file names as operands. Many
99 options have multi-letter names; therefore multiple single-letter options
100 may @emph{not} be grouped: @option{-dr} is very different from @w{@samp{-d
103 @cindex order of options
104 @cindex options, order
105 You can mix options and other arguments. For the most part, the order
106 you use doesn't matter. Order does matter when you use several
107 options of the same kind; for example, if you specify @option{-L} more
108 than once, the directories are searched in the order specified. Also,
109 the placement of the @option{-l} option is significant.
111 Many options have long names starting with @samp{-f} or with
112 @samp{-W}---for example,
113 @option{-fmove-loop-invariants}, @option{-Wformat} and so on. Most of
114 these have both positive and negative forms; the negative form of
115 @option{-ffoo} would be @option{-fno-foo}. This manual documents
116 only one of these two forms, whichever one is not the default.
120 @xref{Option Index}, for an index to GCC's options.
123 * Option Summary:: Brief list of all options, without explanations.
124 * Overall Options:: Controlling the kind of output:
125 an executable, object files, assembler files,
126 or preprocessed source.
127 * Invoking G++:: Compiling C++ programs.
128 * C Dialect Options:: Controlling the variant of C language compiled.
129 * C++ Dialect Options:: Variations on C++.
130 * Objective-C and Objective-C++ Dialect Options:: Variations on Objective-C
132 * Language Independent Options:: Controlling how diagnostics should be
134 * Warning Options:: How picky should the compiler be?
135 * Debugging Options:: Symbol tables, measurements, and debugging dumps.
136 * Optimize Options:: How much optimization?
137 * Preprocessor Options:: Controlling header files and macro definitions.
138 Also, getting dependency information for Make.
139 * Assembler Options:: Passing options to the assembler.
140 * Link Options:: Specifying libraries and so on.
141 * Directory Options:: Where to find header files and libraries.
142 Where to find the compiler executable files.
143 * Spec Files:: How to pass switches to sub-processes.
144 * Target Options:: Running a cross-compiler, or an old version of GCC.
145 * Submodel Options:: Specifying minor hardware or convention variations,
146 such as 68010 vs 68020.
147 * Code Gen Options:: Specifying conventions for function calls, data layout
149 * Environment Variables:: Env vars that affect GCC.
150 * Precompiled Headers:: Compiling a header once, and using it many times.
151 * Running Protoize:: Automatically adding or removing function prototypes.
157 @section Option Summary
159 Here is a summary of all the options, grouped by type. Explanations are
160 in the following sections.
163 @item Overall Options
164 @xref{Overall Options,,Options Controlling the Kind of Output}.
165 @gccoptlist{-c -S -E -o @var{file} -combine -pipe -pass-exit-codes @gol
166 -x @var{language} -v -### --help@r{[}=@var{class}@r{]} --target-help @gol
167 --version -wrapper@@@var{file}}
169 @item C Language Options
170 @xref{C Dialect Options,,Options Controlling C Dialect}.
171 @gccoptlist{-ansi -std=@var{standard} -fgnu89-inline @gol
172 -aux-info @var{filename} @gol
173 -fno-asm -fno-builtin -fno-builtin-@var{function} @gol
174 -fhosted -ffreestanding -fopenmp -fms-extensions @gol
175 -trigraphs -no-integrated-cpp -traditional -traditional-cpp @gol
176 -fallow-single-precision -fcond-mismatch -flax-vector-conversions @gol
177 -fsigned-bitfields -fsigned-char @gol
178 -funsigned-bitfields -funsigned-char}
180 @item C++ Language Options
181 @xref{C++ Dialect Options,,Options Controlling C++ Dialect}.
182 @gccoptlist{-fabi-version=@var{n} -fno-access-control -fcheck-new @gol
183 -fconserve-space -ffriend-injection @gol
184 -fno-elide-constructors @gol
185 -fno-enforce-eh-specs @gol
186 -ffor-scope -fno-for-scope -fno-gnu-keywords @gol
187 -fno-implicit-templates @gol
188 -fno-implicit-inline-templates @gol
189 -fno-implement-inlines -fms-extensions @gol
190 -fno-nonansi-builtins -fno-operator-names @gol
191 -fno-optional-diags -fpermissive @gol
192 -frepo -fno-rtti -fstats -ftemplate-depth-@var{n} @gol
193 -fno-threadsafe-statics -fuse-cxa-atexit -fno-weak -nostdinc++ @gol
194 -fno-default-inline -fvisibility-inlines-hidden @gol
195 -fvisibility-ms-compat @gol
196 -Wabi -Wctor-dtor-privacy @gol
197 -Wnon-virtual-dtor -Wreorder @gol
198 -Weffc++ -Wstrict-null-sentinel @gol
199 -Wno-non-template-friend -Wold-style-cast @gol
200 -Woverloaded-virtual -Wno-pmf-conversions @gol
203 @item Objective-C and Objective-C++ Language Options
204 @xref{Objective-C and Objective-C++ Dialect Options,,Options Controlling
205 Objective-C and Objective-C++ Dialects}.
206 @gccoptlist{-fconstant-string-class=@var{class-name} @gol
207 -fgnu-runtime -fnext-runtime @gol
208 -fno-nil-receivers @gol
209 -fobjc-call-cxx-cdtors @gol
210 -fobjc-direct-dispatch @gol
211 -fobjc-exceptions @gol
213 -freplace-objc-classes @gol
216 -Wassign-intercept @gol
217 -Wno-protocol -Wselector @gol
218 -Wstrict-selector-match @gol
219 -Wundeclared-selector}
221 @item Language Independent Options
222 @xref{Language Independent Options,,Options to Control Diagnostic Messages Formatting}.
223 @gccoptlist{-fmessage-length=@var{n} @gol
224 -fdiagnostics-show-location=@r{[}once@r{|}every-line@r{]} @gol
225 -fdiagnostics-show-option}
227 @item Warning Options
228 @xref{Warning Options,,Options to Request or Suppress Warnings}.
229 @gccoptlist{-fsyntax-only -pedantic -pedantic-errors @gol
230 -w -Wextra -Wall -Waddress -Waggregate-return -Warray-bounds @gol
231 -Wno-attributes -Wc++-compat -Wc++0x-compat -Wcast-align -Wcast-qual @gol
232 -Wchar-subscripts -Wclobbered -Wcomment @gol
233 -Wconversion -Wcoverage-mismatch -Wno-deprecated @gol
234 -Wno-deprecated-declarations -Wdisabled-optimization @gol
235 -Wdisallowed-function-list=@var{sym},@var{sym},@dots{} @gol
236 -Wno-div-by-zero -Wempty-body -Wenum-compare -Wno-endif-labels @gol
237 -Werror -Werror=* @gol
238 -Wfatal-errors -Wfloat-equal -Wformat -Wformat=2 @gol
239 -Wno-format-contains-nul -Wno-format-extra-args -Wformat-nonliteral @gol
240 -Wformat-security -Wformat-y2k @gol
241 -Wframe-larger-than=@var{len} -Wignored-qualifiers @gol
242 -Wimplicit -Wimplicit-function-declaration -Wimplicit-int @gol
243 -Winit-self -Winline @gol
244 -Wno-int-to-pointer-cast -Wno-invalid-offsetof @gol
245 -Winvalid-pch -Wlarger-than=@var{len} -Wunsafe-loop-optimizations @gol
246 -Wlogical-op -Wlong-long @gol
247 -Wmain -Wmissing-braces -Wmissing-field-initializers @gol
248 -Wmissing-format-attribute -Wmissing-include-dirs @gol
249 -Wmissing-noreturn -Wno-mudflap @gol
250 -Wno-multichar -Wnonnull -Wno-overflow @gol
251 -Woverlength-strings -Wpacked -Wpadded @gol
252 -Wparentheses -Wpointer-arith -Wno-pointer-to-int-cast @gol
253 -Wredundant-decls @gol
254 -Wreturn-type -Wsequence-point -Wshadow @gol
255 -Wsign-compare -Wsign-conversion -Wstack-protector @gol
256 -Wstrict-aliasing -Wstrict-aliasing=n @gol
257 -Wstrict-overflow -Wstrict-overflow=@var{n} @gol
258 -Wswitch -Wswitch-default -Wswitch-enum @gol
259 -Wsystem-headers -Wtrigraphs -Wtype-limits -Wundef -Wuninitialized @gol
260 -Wunknown-pragmas -Wno-pragmas -Wunreachable-code @gol
261 -Wunused -Wunused-function -Wunused-label -Wunused-parameter @gol
262 -Wunused-value -Wunused-variable @gol
263 -Wvariadic-macros -Wvla @gol
264 -Wvolatile-register-var -Wwrite-strings}
266 @item C and Objective-C-only Warning Options
267 @gccoptlist{-Wbad-function-cast -Wmissing-declarations @gol
268 -Wmissing-parameter-type -Wmissing-prototypes -Wnested-externs @gol
269 -Wold-style-declaration -Wold-style-definition @gol
270 -Wstrict-prototypes -Wtraditional -Wtraditional-conversion @gol
271 -Wdeclaration-after-statement -Wpointer-sign}
273 @item Debugging Options
274 @xref{Debugging Options,,Options for Debugging Your Program or GCC}.
275 @gccoptlist{-d@var{letters} -dumpspecs -dumpmachine -dumpversion @gol
276 -fdbg-cnt-list -fdbg-cnt=@var{counter-value-list} @gol
277 -fdump-noaddr -fdump-unnumbered -fdump-translation-unit@r{[}-@var{n}@r{]} @gol
278 -fdump-class-hierarchy@r{[}-@var{n}@r{]} @gol
279 -fdump-ipa-all -fdump-ipa-cgraph -fdump-ipa-inline @gol
280 -fdump-statistics @gol
282 -fdump-tree-original@r{[}-@var{n}@r{]} @gol
283 -fdump-tree-optimized@r{[}-@var{n}@r{]} @gol
284 -fdump-tree-cfg -fdump-tree-vcg -fdump-tree-alias @gol
286 -fdump-tree-ssa@r{[}-@var{n}@r{]} -fdump-tree-pre@r{[}-@var{n}@r{]} @gol
287 -fdump-tree-ccp@r{[}-@var{n}@r{]} -fdump-tree-dce@r{[}-@var{n}@r{]} @gol
288 -fdump-tree-gimple@r{[}-raw@r{]} -fdump-tree-mudflap@r{[}-@var{n}@r{]} @gol
289 -fdump-tree-dom@r{[}-@var{n}@r{]} @gol
290 -fdump-tree-dse@r{[}-@var{n}@r{]} @gol
291 -fdump-tree-phiopt@r{[}-@var{n}@r{]} @gol
292 -fdump-tree-forwprop@r{[}-@var{n}@r{]} @gol
293 -fdump-tree-copyrename@r{[}-@var{n}@r{]} @gol
294 -fdump-tree-nrv -fdump-tree-vect @gol
295 -fdump-tree-sink @gol
296 -fdump-tree-sra@r{[}-@var{n}@r{]} @gol
297 -fdump-tree-fre@r{[}-@var{n}@r{]} @gol
298 -fdump-tree-vrp@r{[}-@var{n}@r{]} @gol
299 -ftree-vectorizer-verbose=@var{n} @gol
300 -fdump-tree-storeccp@r{[}-@var{n}@r{]} @gol
301 -feliminate-dwarf2-dups -feliminate-unused-debug-types @gol
302 -feliminate-unused-debug-symbols -femit-class-debug-always @gol
303 -fmem-report -fpre-ipa-mem-report -fpost-ipa-mem-report -fprofile-arcs @gol
304 -frandom-seed=@var{string} -fsched-verbose=@var{n} @gol
305 -ftest-coverage -ftime-report -fvar-tracking @gol
306 -g -g@var{level} -gcoff -gdwarf-2 @gol
307 -ggdb -gstabs -gstabs+ -gvms -gxcoff -gxcoff+ @gol
308 -fno-merge-debug-strings -fdebug-prefix-map=@var{old}=@var{new} @gol
309 -femit-struct-debug-baseonly -femit-struct-debug-reduced @gol
310 -femit-struct-debug-detailed@r{[}=@var{spec-list}@r{]} @gol
311 -p -pg -print-file-name=@var{library} -print-libgcc-file-name @gol
312 -print-multi-directory -print-multi-lib @gol
313 -print-prog-name=@var{program} -print-search-dirs -Q @gol
314 -print-sysroot -print-sysroot-headers-suffix @gol
317 @item Optimization Options
318 @xref{Optimize Options,,Options that Control Optimization}.
320 -falign-functions[=@var{n}] -falign-jumps[=@var{n}] @gol
321 -falign-labels[=@var{n}] -falign-loops[=@var{n}] -fassociative-math @gol
322 -fauto-inc-dec -fbranch-probabilities -fbranch-target-load-optimize @gol
323 -fbranch-target-load-optimize2 -fbtr-bb-exclusive -fcaller-saves @gol
324 -fcheck-data-deps -fcprop-registers -fcrossjumping -fcse-follow-jumps @gol
325 -fcse-skip-blocks -fcx-fortran-rules -fcx-limited-range @gol
326 -fdata-sections -fdce -fdce @gol
327 -fdelayed-branch -fdelete-null-pointer-checks -fdse -fdse @gol
328 -fearly-inlining -fexpensive-optimizations -ffast-math @gol
329 -ffinite-math-only -ffloat-store -fforward-propagate @gol
330 -ffunction-sections -fgcse -fgcse-after-reload -fgcse-las -fgcse-lm @gol
331 -fgcse-sm -fif-conversion -fif-conversion2 -findirect-inlining @gol
332 -finline-functions -finline-functions-called-once -finline-limit=@var{n} @gol
333 -finline-small-functions -fipa-cp -fipa-marix-reorg -fipa-pta @gol
334 -fipa-pure-const -fipa-reference -fipa-struct-reorg @gol
335 -fipa-type-escape -fivopts -fkeep-inline-functions -fkeep-static-consts @gol
336 -fmerge-all-constants -fmerge-constants -fmodulo-sched @gol
337 -fmodulo-sched-allow-regmoves -fmove-loop-invariants -fmudflap @gol
338 -fmudflapir -fmudflapth -fno-branch-count-reg -fno-default-inline @gol
339 -fno-defer-pop -fno-function-cse -fno-guess-branch-probability @gol
340 -fno-inline -fno-math-errno -fno-peephole -fno-peephole2 @gol
341 -fno-sched-interblock -fno-sched-spec -fno-signed-zeros @gol
342 -fno-toplevel-reorder -fno-trapping-math -fno-zero-initialized-in-bss @gol
343 -fomit-frame-pointer -foptimize-register-move -foptimize-sibling-calls @gol
344 -fpeel-loops -fpredictive-commoning -fprefetch-loop-arrays @gol
345 -fprofile-dir=@var{path} -fprofile-generate -fprofile-generate=@var{path} @gol
346 -fprofile-use -fprofile-use=@var{path} -fprofile-values @gol
347 -freciprocal-math -fregmove -frename-registers -freorder-blocks @gol
348 -freorder-blocks-and-partition -freorder-functions @gol
349 -frerun-cse-after-loop -freschedule-modulo-scheduled-loops @gol
350 -frounding-math -frtl-abstract-sequences -fsched2-use-superblocks @gol
351 -fsched2-use-traces -fsched-spec-load -fsched-spec-load-dangerous @gol
352 -fsched-stalled-insns-dep[=@var{n}] -fsched-stalled-insns[=@var{n}] @gol
353 -fschedule-insns -fschedule-insns2 -fsection-anchors -fsee @gol
354 -fsignaling-nans -fsingle-precision-constant -fsplit-ivs-in-unroller @gol
355 -fsplit-wide-types -fstack-protector -fstack-protector-all @gol
356 -fstrict-aliasing -fstrict-overflow -fthread-jumps -ftracer @gol
357 -ftree-builtin-call-dce -ftree-ccp -ftree-ch -ftree-copy-prop @gol
358 -ftree-copyrename -ftree-dce @gol
359 -ftree-dominator-opts -ftree-dse -ftree-fre -ftree-loop-im @gol
360 -ftree-loop-distribution @gol
361 -ftree-loop-ivcanon -ftree-loop-linear -ftree-loop-optimize @gol
362 -ftree-parallelize-loops=@var{n} -ftree-pre -ftree-reassoc @gol
363 -ftree-sink -ftree-sra -ftree-store-ccp -ftree-switch-conversion @gol
364 -ftree-ter -ftree-vect-loop-version -ftree-vectorize -ftree-vrp @gol
365 -funit-at-a-time -funroll-all-loops -funroll-loops @gol
366 -funsafe-loop-optimizations -funsafe-math-optimizations -funswitch-loops @gol
367 -fvariable-expansion-in-unroller -fvect-cost-model -fvpt -fweb @gol
369 --param @var{name}=@var{value}
370 -O -O0 -O1 -O2 -O3 -Os}
372 @item Preprocessor Options
373 @xref{Preprocessor Options,,Options Controlling the Preprocessor}.
374 @gccoptlist{-A@var{question}=@var{answer} @gol
375 -A-@var{question}@r{[}=@var{answer}@r{]} @gol
376 -C -dD -dI -dM -dN @gol
377 -D@var{macro}@r{[}=@var{defn}@r{]} -E -H @gol
378 -idirafter @var{dir} @gol
379 -include @var{file} -imacros @var{file} @gol
380 -iprefix @var{file} -iwithprefix @var{dir} @gol
381 -iwithprefixbefore @var{dir} -isystem @var{dir} @gol
382 -imultilib @var{dir} -isysroot @var{dir} @gol
383 -M -MM -MF -MG -MP -MQ -MT -nostdinc @gol
384 -P -fworking-directory -remap @gol
385 -trigraphs -undef -U@var{macro} -Wp,@var{option} @gol
386 -Xpreprocessor @var{option}}
388 @item Assembler Option
389 @xref{Assembler Options,,Passing Options to the Assembler}.
390 @gccoptlist{-Wa,@var{option} -Xassembler @var{option}}
393 @xref{Link Options,,Options for Linking}.
394 @gccoptlist{@var{object-file-name} -l@var{library} @gol
395 -nostartfiles -nodefaultlibs -nostdlib -pie -rdynamic @gol
396 -s -static -static-libgcc -shared -shared-libgcc -symbolic @gol
397 -Wl,@var{option} -Xlinker @var{option} @gol
400 @item Directory Options
401 @xref{Directory Options,,Options for Directory Search}.
402 @gccoptlist{-B@var{prefix} -I@var{dir} -iquote@var{dir} -L@var{dir}
403 -specs=@var{file} -I- --sysroot=@var{dir}}
406 @c I wrote this xref this way to avoid overfull hbox. -- rms
407 @xref{Target Options}.
408 @gccoptlist{-V @var{version} -b @var{machine}}
410 @item Machine Dependent Options
411 @xref{Submodel Options,,Hardware Models and Configurations}.
412 @c This list is ordered alphanumerically by subsection name.
413 @c Try and put the significant identifier (CPU or system) first,
414 @c so users have a clue at guessing where the ones they want will be.
417 @gccoptlist{-EB -EL @gol
418 -mmangle-cpu -mcpu=@var{cpu} -mtext=@var{text-section} @gol
419 -mdata=@var{data-section} -mrodata=@var{readonly-data-section}}
422 @gccoptlist{-mapcs-frame -mno-apcs-frame @gol
423 -mabi=@var{name} @gol
424 -mapcs-stack-check -mno-apcs-stack-check @gol
425 -mapcs-float -mno-apcs-float @gol
426 -mapcs-reentrant -mno-apcs-reentrant @gol
427 -msched-prolog -mno-sched-prolog @gol
428 -mlittle-endian -mbig-endian -mwords-little-endian @gol
429 -mfloat-abi=@var{name} -msoft-float -mhard-float -mfpe @gol
430 -mthumb-interwork -mno-thumb-interwork @gol
431 -mcpu=@var{name} -march=@var{name} -mfpu=@var{name} @gol
432 -mstructure-size-boundary=@var{n} @gol
433 -mabort-on-noreturn @gol
434 -mlong-calls -mno-long-calls @gol
435 -msingle-pic-base -mno-single-pic-base @gol
436 -mpic-register=@var{reg} @gol
437 -mnop-fun-dllimport @gol
438 -mcirrus-fix-invalid-insns -mno-cirrus-fix-invalid-insns @gol
439 -mpoke-function-name @gol
441 -mtpcs-frame -mtpcs-leaf-frame @gol
442 -mcaller-super-interworking -mcallee-super-interworking @gol
446 @gccoptlist{-mmcu=@var{mcu} -msize -minit-stack=@var{n} -mno-interrupts @gol
447 -mcall-prologues -mno-tablejump -mtiny-stack -mint8}
449 @emph{Blackfin Options}
450 @gccoptlist{-mcpu=@var{cpu}@r{[}-@var{sirevision}@r{]} @gol
451 -msim -momit-leaf-frame-pointer -mno-omit-leaf-frame-pointer @gol
452 -mspecld-anomaly -mno-specld-anomaly -mcsync-anomaly -mno-csync-anomaly @gol
453 -mlow-64k -mno-low64k -mstack-check-l1 -mid-shared-library @gol
454 -mno-id-shared-library -mshared-library-id=@var{n} @gol
455 -mleaf-id-shared-library -mno-leaf-id-shared-library @gol
456 -msep-data -mno-sep-data -mlong-calls -mno-long-calls @gol
457 -mfast-fp -minline-plt -mmulticore -mcorea -mcoreb -msdram}
460 @gccoptlist{-mcpu=@var{cpu} -march=@var{cpu} -mtune=@var{cpu} @gol
461 -mmax-stack-frame=@var{n} -melinux-stacksize=@var{n} @gol
462 -metrax4 -metrax100 -mpdebug -mcc-init -mno-side-effects @gol
463 -mstack-align -mdata-align -mconst-align @gol
464 -m32-bit -m16-bit -m8-bit -mno-prologue-epilogue -mno-gotplt @gol
465 -melf -maout -melinux -mlinux -sim -sim2 @gol
466 -mmul-bug-workaround -mno-mul-bug-workaround}
469 @gccoptlist{-mmac -mpush-args}
471 @emph{Darwin Options}
472 @gccoptlist{-all_load -allowable_client -arch -arch_errors_fatal @gol
473 -arch_only -bind_at_load -bundle -bundle_loader @gol
474 -client_name -compatibility_version -current_version @gol
476 -dependency-file -dylib_file -dylinker_install_name @gol
477 -dynamic -dynamiclib -exported_symbols_list @gol
478 -filelist -flat_namespace -force_cpusubtype_ALL @gol
479 -force_flat_namespace -headerpad_max_install_names @gol
481 -image_base -init -install_name -keep_private_externs @gol
482 -multi_module -multiply_defined -multiply_defined_unused @gol
483 -noall_load -no_dead_strip_inits_and_terms @gol
484 -nofixprebinding -nomultidefs -noprebind -noseglinkedit @gol
485 -pagezero_size -prebind -prebind_all_twolevel_modules @gol
486 -private_bundle -read_only_relocs -sectalign @gol
487 -sectobjectsymbols -whyload -seg1addr @gol
488 -sectcreate -sectobjectsymbols -sectorder @gol
489 -segaddr -segs_read_only_addr -segs_read_write_addr @gol
490 -seg_addr_table -seg_addr_table_filename -seglinkedit @gol
491 -segprot -segs_read_only_addr -segs_read_write_addr @gol
492 -single_module -static -sub_library -sub_umbrella @gol
493 -twolevel_namespace -umbrella -undefined @gol
494 -unexported_symbols_list -weak_reference_mismatches @gol
495 -whatsloaded -F -gused -gfull -mmacosx-version-min=@var{version} @gol
496 -mkernel -mone-byte-bool}
498 @emph{DEC Alpha Options}
499 @gccoptlist{-mno-fp-regs -msoft-float -malpha-as -mgas @gol
500 -mieee -mieee-with-inexact -mieee-conformant @gol
501 -mfp-trap-mode=@var{mode} -mfp-rounding-mode=@var{mode} @gol
502 -mtrap-precision=@var{mode} -mbuild-constants @gol
503 -mcpu=@var{cpu-type} -mtune=@var{cpu-type} @gol
504 -mbwx -mmax -mfix -mcix @gol
505 -mfloat-vax -mfloat-ieee @gol
506 -mexplicit-relocs -msmall-data -mlarge-data @gol
507 -msmall-text -mlarge-text @gol
508 -mmemory-latency=@var{time}}
510 @emph{DEC Alpha/VMS Options}
511 @gccoptlist{-mvms-return-codes}
514 @gccoptlist{-mgpr-32 -mgpr-64 -mfpr-32 -mfpr-64 @gol
515 -mhard-float -msoft-float @gol
516 -malloc-cc -mfixed-cc -mdword -mno-dword @gol
517 -mdouble -mno-double @gol
518 -mmedia -mno-media -mmuladd -mno-muladd @gol
519 -mfdpic -minline-plt -mgprel-ro -multilib-library-pic @gol
520 -mlinked-fp -mlong-calls -malign-labels @gol
521 -mlibrary-pic -macc-4 -macc-8 @gol
522 -mpack -mno-pack -mno-eflags -mcond-move -mno-cond-move @gol
523 -moptimize-membar -mno-optimize-membar @gol
524 -mscc -mno-scc -mcond-exec -mno-cond-exec @gol
525 -mvliw-branch -mno-vliw-branch @gol
526 -mmulti-cond-exec -mno-multi-cond-exec -mnested-cond-exec @gol
527 -mno-nested-cond-exec -mtomcat-stats @gol
531 @emph{GNU/Linux Options}
532 @gccoptlist{-muclibc}
534 @emph{H8/300 Options}
535 @gccoptlist{-mrelax -mh -ms -mn -mint32 -malign-300}
538 @gccoptlist{-march=@var{architecture-type} @gol
539 -mbig-switch -mdisable-fpregs -mdisable-indexing @gol
540 -mfast-indirect-calls -mgas -mgnu-ld -mhp-ld @gol
541 -mfixed-range=@var{register-range} @gol
542 -mjump-in-delay -mlinker-opt -mlong-calls @gol
543 -mlong-load-store -mno-big-switch -mno-disable-fpregs @gol
544 -mno-disable-indexing -mno-fast-indirect-calls -mno-gas @gol
545 -mno-jump-in-delay -mno-long-load-store @gol
546 -mno-portable-runtime -mno-soft-float @gol
547 -mno-space-regs -msoft-float -mpa-risc-1-0 @gol
548 -mpa-risc-1-1 -mpa-risc-2-0 -mportable-runtime @gol
549 -mschedule=@var{cpu-type} -mspace-regs -msio -mwsio @gol
550 -munix=@var{unix-std} -nolibdld -static -threads}
552 @emph{i386 and x86-64 Options}
553 @gccoptlist{-mtune=@var{cpu-type} -march=@var{cpu-type} @gol
554 -mfpmath=@var{unit} @gol
555 -masm=@var{dialect} -mno-fancy-math-387 @gol
556 -mno-fp-ret-in-387 -msoft-float @gol
557 -mno-wide-multiply -mrtd -malign-double @gol
558 -mpreferred-stack-boundary=@var{num}
559 -mincoming-stack-boundary=@var{num}
560 -mcld -mcx16 -msahf -mrecip @gol
561 -mmmx -msse -msse2 -msse3 -mssse3 -msse4.1 -msse4.2 -msse4 @gol
563 -msse4a -m3dnow -mpopcnt -mabm -msse5 @gol
564 -mthreads -mno-align-stringops -minline-all-stringops @gol
565 -minline-stringops-dynamically -mstringop-strategy=@var{alg} @gol
566 -mpush-args -maccumulate-outgoing-args -m128bit-long-double @gol
567 -m96bit-long-double -mregparm=@var{num} -msseregparm @gol
568 -mveclibabi=@var{type} -mpc32 -mpc64 -mpc80 -mstackrealign @gol
569 -momit-leaf-frame-pointer -mno-red-zone -mno-tls-direct-seg-refs @gol
570 -mcmodel=@var{code-model} @gol
571 -m32 -m64 -mlarge-data-threshold=@var{num} @gol
572 -mfused-madd -mno-fused-madd}
575 @gccoptlist{-mbig-endian -mlittle-endian -mgnu-as -mgnu-ld -mno-pic @gol
576 -mvolatile-asm-stop -mregister-names -mno-sdata @gol
577 -mconstant-gp -mauto-pic -minline-float-divide-min-latency @gol
578 -minline-float-divide-max-throughput @gol
579 -minline-int-divide-min-latency @gol
580 -minline-int-divide-max-throughput @gol
581 -minline-sqrt-min-latency -minline-sqrt-max-throughput @gol
582 -mno-dwarf2-asm -mearly-stop-bits @gol
583 -mfixed-range=@var{register-range} -mtls-size=@var{tls-size} @gol
584 -mtune=@var{cpu-type} -mt -pthread -milp32 -mlp64 @gol
585 -mno-sched-br-data-spec -msched-ar-data-spec -mno-sched-control-spec @gol
586 -msched-br-in-data-spec -msched-ar-in-data-spec -msched-in-control-spec @gol
587 -msched-ldc -mno-sched-control-ldc -mno-sched-spec-verbose @gol
588 -mno-sched-prefer-non-data-spec-insns @gol
589 -mno-sched-prefer-non-control-spec-insns @gol
590 -mno-sched-count-spec-in-critical-path}
592 @emph{M32R/D Options}
593 @gccoptlist{-m32r2 -m32rx -m32r @gol
595 -malign-loops -mno-align-loops @gol
596 -missue-rate=@var{number} @gol
597 -mbranch-cost=@var{number} @gol
598 -mmodel=@var{code-size-model-type} @gol
599 -msdata=@var{sdata-type} @gol
600 -mno-flush-func -mflush-func=@var{name} @gol
601 -mno-flush-trap -mflush-trap=@var{number} @gol
605 @gccoptlist{-mcpu=@var{cpu} -msim -memregs=@var{number}}
607 @emph{M680x0 Options}
608 @gccoptlist{-march=@var{arch} -mcpu=@var{cpu} -mtune=@var{tune}
609 -m68000 -m68020 -m68020-40 -m68020-60 -m68030 -m68040 @gol
610 -m68060 -mcpu32 -m5200 -m5206e -m528x -m5307 -m5407 @gol
611 -mcfv4e -mbitfield -mno-bitfield -mc68000 -mc68020 @gol
612 -mnobitfield -mrtd -mno-rtd -mdiv -mno-div -mshort @gol
613 -mno-short -mhard-float -m68881 -msoft-float -mpcrel @gol
614 -malign-int -mstrict-align -msep-data -mno-sep-data @gol
615 -mshared-library-id=n -mid-shared-library -mno-id-shared-library @gol
618 @emph{M68hc1x Options}
619 @gccoptlist{-m6811 -m6812 -m68hc11 -m68hc12 -m68hcs12 @gol
620 -mauto-incdec -minmax -mlong-calls -mshort @gol
621 -msoft-reg-count=@var{count}}
624 @gccoptlist{-mhardlit -mno-hardlit -mdiv -mno-div -mrelax-immediates @gol
625 -mno-relax-immediates -mwide-bitfields -mno-wide-bitfields @gol
626 -m4byte-functions -mno-4byte-functions -mcallgraph-data @gol
627 -mno-callgraph-data -mslow-bytes -mno-slow-bytes -mno-lsim @gol
628 -mlittle-endian -mbig-endian -m210 -m340 -mstack-increment}
631 @gccoptlist{-EL -EB -march=@var{arch} -mtune=@var{arch} @gol
632 -mips1 -mips2 -mips3 -mips4 -mips32 -mips32r2 -mips64 @gol
633 -mips16 -mno-mips16 -mflip-mips16 @gol
634 -minterlink-mips16 -mno-interlink-mips16 @gol
635 -mabi=@var{abi} -mabicalls -mno-abicalls @gol
636 -mshared -mno-shared -mxgot -mno-xgot -mgp32 -mgp64 @gol
637 -mfp32 -mfp64 -mhard-float -msoft-float @gol
638 -msingle-float -mdouble-float -mdsp -mno-dsp -mdspr2 -mno-dspr2 @gol
639 -msmartmips -mno-smartmips @gol
640 -mpaired-single -mno-paired-single -mdmx -mno-mdmx @gol
641 -mips3d -mno-mips3d -mmt -mno-mt -mllsc -mno-llsc @gol
642 -mlong64 -mlong32 -msym32 -mno-sym32 @gol
643 -G@var{num} -mlocal-sdata -mno-local-sdata @gol
644 -mextern-sdata -mno-extern-sdata -mgpopt -mno-gopt @gol
645 -membedded-data -mno-embedded-data @gol
646 -muninit-const-in-rodata -mno-uninit-const-in-rodata @gol
647 -mcode-readable=@var{setting} @gol
648 -msplit-addresses -mno-split-addresses @gol
649 -mexplicit-relocs -mno-explicit-relocs @gol
650 -mcheck-zero-division -mno-check-zero-division @gol
651 -mdivide-traps -mdivide-breaks @gol
652 -mmemcpy -mno-memcpy -mlong-calls -mno-long-calls @gol
653 -mmad -mno-mad -mfused-madd -mno-fused-madd -nocpp @gol
654 -mfix-r4000 -mno-fix-r4000 -mfix-r4400 -mno-fix-r4400 @gol
655 -mfix-vr4120 -mno-fix-vr4120 -mfix-vr4130 -mno-fix-vr4130 @gol
656 -mfix-sb1 -mno-fix-sb1 @gol
657 -mflush-func=@var{func} -mno-flush-func @gol
658 -mbranch-cost=@var{num} -mbranch-likely -mno-branch-likely @gol
659 -mfp-exceptions -mno-fp-exceptions @gol
660 -mvr4130-align -mno-vr4130-align}
663 @gccoptlist{-mlibfuncs -mno-libfuncs -mepsilon -mno-epsilon -mabi=gnu @gol
664 -mabi=mmixware -mzero-extend -mknuthdiv -mtoplevel-symbols @gol
665 -melf -mbranch-predict -mno-branch-predict -mbase-addresses @gol
666 -mno-base-addresses -msingle-exit -mno-single-exit}
668 @emph{MN10300 Options}
669 @gccoptlist{-mmult-bug -mno-mult-bug @gol
670 -mam33 -mno-am33 @gol
671 -mam33-2 -mno-am33-2 @gol
672 -mreturn-pointer-on-d0 @gol
675 @emph{PDP-11 Options}
676 @gccoptlist{-mfpu -msoft-float -mac0 -mno-ac0 -m40 -m45 -m10 @gol
677 -mbcopy -mbcopy-builtin -mint32 -mno-int16 @gol
678 -mint16 -mno-int32 -mfloat32 -mno-float64 @gol
679 -mfloat64 -mno-float32 -mabshi -mno-abshi @gol
680 -mbranch-expensive -mbranch-cheap @gol
681 -msplit -mno-split -munix-asm -mdec-asm}
683 @emph{PowerPC Options}
684 See RS/6000 and PowerPC Options.
686 @emph{RS/6000 and PowerPC Options}
687 @gccoptlist{-mcpu=@var{cpu-type} @gol
688 -mtune=@var{cpu-type} @gol
689 -mpower -mno-power -mpower2 -mno-power2 @gol
690 -mpowerpc -mpowerpc64 -mno-powerpc @gol
691 -maltivec -mno-altivec @gol
692 -mpowerpc-gpopt -mno-powerpc-gpopt @gol
693 -mpowerpc-gfxopt -mno-powerpc-gfxopt @gol
694 -mmfcrf -mno-mfcrf -mpopcntb -mno-popcntb -mfprnd -mno-fprnd @gol
695 -mcmpb -mno-cmpb -mmfpgpr -mno-mfpgpr -mhard-dfp -mno-hard-dfp @gol
696 -mnew-mnemonics -mold-mnemonics @gol
697 -mfull-toc -mminimal-toc -mno-fp-in-toc -mno-sum-in-toc @gol
698 -m64 -m32 -mxl-compat -mno-xl-compat -mpe @gol
699 -malign-power -malign-natural @gol
700 -msoft-float -mhard-float -mmultiple -mno-multiple @gol
701 -mstring -mno-string -mupdate -mno-update @gol
702 -mfused-madd -mno-fused-madd -mbit-align -mno-bit-align @gol
703 -mstrict-align -mno-strict-align -mrelocatable @gol
704 -mno-relocatable -mrelocatable-lib -mno-relocatable-lib @gol
705 -mtoc -mno-toc -mlittle -mlittle-endian -mbig -mbig-endian @gol
706 -mdynamic-no-pic -maltivec -mswdiv @gol
707 -mprioritize-restricted-insns=@var{priority} @gol
708 -msched-costly-dep=@var{dependence_type} @gol
709 -minsert-sched-nops=@var{scheme} @gol
710 -mcall-sysv -mcall-netbsd @gol
711 -maix-struct-return -msvr4-struct-return @gol
712 -mabi=@var{abi-type} -msecure-plt -mbss-plt @gol
713 -misel -mno-isel @gol
714 -misel=yes -misel=no @gol
716 -mspe=yes -mspe=no @gol
718 -mvrsave -mno-vrsave @gol
719 -mmulhw -mno-mulhw @gol
720 -mdlmzb -mno-dlmzb @gol
721 -mfloat-gprs=yes -mfloat-gprs=no -mfloat-gprs=single -mfloat-gprs=double @gol
722 -mprototype -mno-prototype @gol
723 -msim -mmvme -mads -myellowknife -memb -msdata @gol
724 -msdata=@var{opt} -mvxworks -G @var{num} -pthread}
726 @emph{S/390 and zSeries Options}
727 @gccoptlist{-mtune=@var{cpu-type} -march=@var{cpu-type} @gol
728 -mhard-float -msoft-float -mlong-double-64 -mlong-double-128 @gol
729 -mbackchain -mno-backchain -mpacked-stack -mno-packed-stack @gol
730 -msmall-exec -mno-small-exec -mmvcle -mno-mvcle @gol
731 -m64 -m31 -mdebug -mno-debug -mesa -mzarch @gol
732 -mtpf-trace -mno-tpf-trace -mfused-madd -mno-fused-madd @gol
733 -mwarn-framesize -mwarn-dynamicstack -mstack-size -mstack-guard}
736 @gccoptlist{-meb -mel @gol
740 -mscore5 -mscore5u -mscore7 -mscore7d}
743 @gccoptlist{-m1 -m2 -m2e -m3 -m3e @gol
744 -m4-nofpu -m4-single-only -m4-single -m4 @gol
745 -m4a-nofpu -m4a-single-only -m4a-single -m4a -m4al @gol
746 -m5-64media -m5-64media-nofpu @gol
747 -m5-32media -m5-32media-nofpu @gol
748 -m5-compact -m5-compact-nofpu @gol
749 -mb -ml -mdalign -mrelax @gol
750 -mbigtable -mfmovd -mhitachi -mrenesas -mno-renesas -mnomacsave @gol
751 -mieee -mbitops -misize -minline-ic_invalidate -mpadstruct -mspace @gol
752 -mprefergot -musermode -multcost=@var{number} -mdiv=@var{strategy} @gol
753 -mdivsi3_libfunc=@var{name} -mfixed-range=@var{register-range} @gol
754 -madjust-unroll -mindexed-addressing -mgettrcost=@var{number} -mpt-fixed @gol
758 @gccoptlist{-mcpu=@var{cpu-type} @gol
759 -mtune=@var{cpu-type} @gol
760 -mcmodel=@var{code-model} @gol
761 -m32 -m64 -mapp-regs -mno-app-regs @gol
762 -mfaster-structs -mno-faster-structs @gol
763 -mfpu -mno-fpu -mhard-float -msoft-float @gol
764 -mhard-quad-float -msoft-quad-float @gol
765 -mimpure-text -mno-impure-text -mlittle-endian @gol
766 -mstack-bias -mno-stack-bias @gol
767 -munaligned-doubles -mno-unaligned-doubles @gol
768 -mv8plus -mno-v8plus -mvis -mno-vis
769 -threads -pthreads -pthread}
772 @gccoptlist{-mwarn-reloc -merror-reloc @gol
773 -msafe-dma -munsafe-dma @gol
775 -msmall-mem -mlarge-mem -mstdmain @gol
776 -mfixed-range=@var{register-range}}
778 @emph{System V Options}
779 @gccoptlist{-Qy -Qn -YP,@var{paths} -Ym,@var{dir}}
782 @gccoptlist{-mlong-calls -mno-long-calls -mep -mno-ep @gol
783 -mprolog-function -mno-prolog-function -mspace @gol
784 -mtda=@var{n} -msda=@var{n} -mzda=@var{n} @gol
785 -mapp-regs -mno-app-regs @gol
786 -mdisable-callt -mno-disable-callt @gol
792 @gccoptlist{-mg -mgnu -munix}
794 @emph{VxWorks Options}
795 @gccoptlist{-mrtp -non-static -Bstatic -Bdynamic @gol
796 -Xbind-lazy -Xbind-now}
798 @emph{x86-64 Options}
799 See i386 and x86-64 Options.
801 @emph{Xstormy16 Options}
804 @emph{Xtensa Options}
805 @gccoptlist{-mconst16 -mno-const16 @gol
806 -mfused-madd -mno-fused-madd @gol
807 -mserialize-volatile -mno-serialize-volatile @gol
808 -mtext-section-literals -mno-text-section-literals @gol
809 -mtarget-align -mno-target-align @gol
810 -mlongcalls -mno-longcalls}
812 @emph{zSeries Options}
813 See S/390 and zSeries Options.
815 @item Code Generation Options
816 @xref{Code Gen Options,,Options for Code Generation Conventions}.
817 @gccoptlist{-fcall-saved-@var{reg} -fcall-used-@var{reg} @gol
818 -ffixed-@var{reg} -fexceptions @gol
819 -fnon-call-exceptions -funwind-tables @gol
820 -fasynchronous-unwind-tables @gol
821 -finhibit-size-directive -finstrument-functions @gol
822 -finstrument-functions-exclude-function-list=@var{sym},@var{sym},@dots{} @gol
823 -finstrument-functions-exclude-file-list=@var{file},@var{file},@dots{} @gol
824 -fno-common -fno-ident @gol
825 -fpcc-struct-return -fpic -fPIC -fpie -fPIE @gol
826 -fno-jump-tables @gol
827 -frecord-gcc-switches @gol
828 -freg-struct-return -fshort-enums @gol
829 -fshort-double -fshort-wchar @gol
830 -fverbose-asm -fpack-struct[=@var{n}] -fstack-check @gol
831 -fstack-limit-register=@var{reg} -fstack-limit-symbol=@var{sym} @gol
832 -fno-stack-limit -fargument-alias -fargument-noalias @gol
833 -fargument-noalias-global -fargument-noalias-anything @gol
834 -fleading-underscore -ftls-model=@var{model} @gol
835 -ftrapv -fwrapv -fbounds-check @gol
840 * Overall Options:: Controlling the kind of output:
841 an executable, object files, assembler files,
842 or preprocessed source.
843 * C Dialect Options:: Controlling the variant of C language compiled.
844 * C++ Dialect Options:: Variations on C++.
845 * Objective-C and Objective-C++ Dialect Options:: Variations on Objective-C
847 * Language Independent Options:: Controlling how diagnostics should be
849 * Warning Options:: How picky should the compiler be?
850 * Debugging Options:: Symbol tables, measurements, and debugging dumps.
851 * Optimize Options:: How much optimization?
852 * Preprocessor Options:: Controlling header files and macro definitions.
853 Also, getting dependency information for Make.
854 * Assembler Options:: Passing options to the assembler.
855 * Link Options:: Specifying libraries and so on.
856 * Directory Options:: Where to find header files and libraries.
857 Where to find the compiler executable files.
858 * Spec Files:: How to pass switches to sub-processes.
859 * Target Options:: Running a cross-compiler, or an old version of GCC.
862 @node Overall Options
863 @section Options Controlling the Kind of Output
865 Compilation can involve up to four stages: preprocessing, compilation
866 proper, assembly and linking, always in that order. GCC is capable of
867 preprocessing and compiling several files either into several
868 assembler input files, or into one assembler input file; then each
869 assembler input file produces an object file, and linking combines all
870 the object files (those newly compiled, and those specified as input)
871 into an executable file.
873 @cindex file name suffix
874 For any given input file, the file name suffix determines what kind of
879 C source code which must be preprocessed.
882 C source code which should not be preprocessed.
885 C++ source code which should not be preprocessed.
888 Objective-C source code. Note that you must link with the @file{libobjc}
889 library to make an Objective-C program work.
892 Objective-C source code which should not be preprocessed.
896 Objective-C++ source code. Note that you must link with the @file{libobjc}
897 library to make an Objective-C++ program work. Note that @samp{.M} refers
898 to a literal capital M@.
901 Objective-C++ source code which should not be preprocessed.
904 C, C++, Objective-C or Objective-C++ header file to be turned into a
909 @itemx @var{file}.cxx
910 @itemx @var{file}.cpp
911 @itemx @var{file}.CPP
912 @itemx @var{file}.c++
914 C++ source code which must be preprocessed. Note that in @samp{.cxx},
915 the last two letters must both be literally @samp{x}. Likewise,
916 @samp{.C} refers to a literal capital C@.
920 Objective-C++ source code which must be preprocessed.
923 Objective-C++ source code which should not be preprocessed.
928 @itemx @var{file}.hxx
929 @itemx @var{file}.hpp
930 @itemx @var{file}.HPP
931 @itemx @var{file}.h++
932 @itemx @var{file}.tcc
933 C++ header file to be turned into a precompiled header.
936 @itemx @var{file}.for
937 @itemx @var{file}.ftn
938 Fixed form Fortran source code which should not be preprocessed.
941 @itemx @var{file}.FOR
942 @itemx @var{file}.fpp
943 @itemx @var{file}.FPP
944 @itemx @var{file}.FTN
945 Fixed form Fortran source code which must be preprocessed (with the traditional
949 @itemx @var{file}.f95
950 @itemx @var{file}.f03
951 @itemx @var{file}.f08
952 Free form Fortran source code which should not be preprocessed.
955 @itemx @var{file}.F95
956 @itemx @var{file}.F03
957 @itemx @var{file}.F08
958 Free form Fortran source code which must be preprocessed (with the
959 traditional preprocessor).
961 @c FIXME: Descriptions of Java file types.
968 Ada source code file which contains a library unit declaration (a
969 declaration of a package, subprogram, or generic, or a generic
970 instantiation), or a library unit renaming declaration (a package,
971 generic, or subprogram renaming declaration). Such files are also
975 Ada source code file containing a library unit body (a subprogram or
976 package body). Such files are also called @dfn{bodies}.
978 @c GCC also knows about some suffixes for languages not yet included:
990 Assembler code which must be preprocessed.
993 An object file to be fed straight into linking.
994 Any file name with no recognized suffix is treated this way.
998 You can specify the input language explicitly with the @option{-x} option:
1001 @item -x @var{language}
1002 Specify explicitly the @var{language} for the following input files
1003 (rather than letting the compiler choose a default based on the file
1004 name suffix). This option applies to all following input files until
1005 the next @option{-x} option. Possible values for @var{language} are:
1007 c c-header c-cpp-output
1008 c++ c++-header c++-cpp-output
1009 objective-c objective-c-header objective-c-cpp-output
1010 objective-c++ objective-c++-header objective-c++-cpp-output
1011 assembler assembler-with-cpp
1013 f77 f77-cpp-input f95 f95-cpp-input
1018 Turn off any specification of a language, so that subsequent files are
1019 handled according to their file name suffixes (as they are if @option{-x}
1020 has not been used at all).
1022 @item -pass-exit-codes
1023 @opindex pass-exit-codes
1024 Normally the @command{gcc} program will exit with the code of 1 if any
1025 phase of the compiler returns a non-success return code. If you specify
1026 @option{-pass-exit-codes}, the @command{gcc} program will instead return with
1027 numerically highest error produced by any phase that returned an error
1028 indication. The C, C++, and Fortran frontends return 4, if an internal
1029 compiler error is encountered.
1032 If you only want some of the stages of compilation, you can use
1033 @option{-x} (or filename suffixes) to tell @command{gcc} where to start, and
1034 one of the options @option{-c}, @option{-S}, or @option{-E} to say where
1035 @command{gcc} is to stop. Note that some combinations (for example,
1036 @samp{-x cpp-output -E}) instruct @command{gcc} to do nothing at all.
1041 Compile or assemble the source files, but do not link. The linking
1042 stage simply is not done. The ultimate output is in the form of an
1043 object file for each source file.
1045 By default, the object file name for a source file is made by replacing
1046 the suffix @samp{.c}, @samp{.i}, @samp{.s}, etc., with @samp{.o}.
1048 Unrecognized input files, not requiring compilation or assembly, are
1053 Stop after the stage of compilation proper; do not assemble. The output
1054 is in the form of an assembler code file for each non-assembler input
1057 By default, the assembler file name for a source file is made by
1058 replacing the suffix @samp{.c}, @samp{.i}, etc., with @samp{.s}.
1060 Input files that don't require compilation are ignored.
1064 Stop after the preprocessing stage; do not run the compiler proper. The
1065 output is in the form of preprocessed source code, which is sent to the
1068 Input files which don't require preprocessing are ignored.
1070 @cindex output file option
1073 Place output in file @var{file}. This applies regardless to whatever
1074 sort of output is being produced, whether it be an executable file,
1075 an object file, an assembler file or preprocessed C code.
1077 If @option{-o} is not specified, the default is to put an executable
1078 file in @file{a.out}, the object file for
1079 @file{@var{source}.@var{suffix}} in @file{@var{source}.o}, its
1080 assembler file in @file{@var{source}.s}, a precompiled header file in
1081 @file{@var{source}.@var{suffix}.gch}, and all preprocessed C source on
1086 Print (on standard error output) the commands executed to run the stages
1087 of compilation. Also print the version number of the compiler driver
1088 program and of the preprocessor and the compiler proper.
1092 Like @option{-v} except the commands are not executed and all command
1093 arguments are quoted. This is useful for shell scripts to capture the
1094 driver-generated command lines.
1098 Use pipes rather than temporary files for communication between the
1099 various stages of compilation. This fails to work on some systems where
1100 the assembler is unable to read from a pipe; but the GNU assembler has
1105 If you are compiling multiple source files, this option tells the driver
1106 to pass all the source files to the compiler at once (for those
1107 languages for which the compiler can handle this). This will allow
1108 intermodule analysis (IMA) to be performed by the compiler. Currently the only
1109 language for which this is supported is C@. If you pass source files for
1110 multiple languages to the driver, using this option, the driver will invoke
1111 the compiler(s) that support IMA once each, passing each compiler all the
1112 source files appropriate for it. For those languages that do not support
1113 IMA this option will be ignored, and the compiler will be invoked once for
1114 each source file in that language. If you use this option in conjunction
1115 with @option{-save-temps}, the compiler will generate multiple
1117 (one for each source file), but only one (combined) @file{.o} or
1122 Print (on the standard output) a description of the command line options
1123 understood by @command{gcc}. If the @option{-v} option is also specified
1124 then @option{--help} will also be passed on to the various processes
1125 invoked by @command{gcc}, so that they can display the command line options
1126 they accept. If the @option{-Wextra} option has also been specified
1127 (prior to the @option{--help} option), then command line options which
1128 have no documentation associated with them will also be displayed.
1131 @opindex target-help
1132 Print (on the standard output) a description of target-specific command
1133 line options for each tool. For some targets extra target-specific
1134 information may also be printed.
1136 @item --help=@var{class}@r{[},@var{qualifier}@r{]}
1137 Print (on the standard output) a description of the command line
1138 options understood by the compiler that fit into a specific class.
1139 The class can be one of @samp{optimizers}, @samp{warnings}, @samp{target},
1140 @samp{params}, or @var{language}:
1143 @item @samp{optimizers}
1144 This will display all of the optimization options supported by the
1147 @item @samp{warnings}
1148 This will display all of the options controlling warning messages
1149 produced by the compiler.
1152 This will display target-specific options. Unlike the
1153 @option{--target-help} option however, target-specific options of the
1154 linker and assembler will not be displayed. This is because those
1155 tools do not currently support the extended @option{--help=} syntax.
1158 This will display the values recognized by the @option{--param}
1161 @item @var{language}
1162 This will display the options supported for @var{language}, where
1163 @var{language} is the name of one of the languages supported in this
1167 This will display the options that are common to all languages.
1170 It is possible to further refine the output of the @option{--help=}
1171 option by adding a comma separated list of qualifiers after the
1172 class. These can be any from the following list:
1175 @item @samp{undocumented}
1176 Display only those options which are undocumented.
1179 Display options which take an argument that appears after an equal
1180 sign in the same continuous piece of text, such as:
1181 @samp{--help=target}.
1183 @item @samp{separate}
1184 Display options which take an argument that appears as a separate word
1185 following the original option, such as: @samp{-o output-file}.
1188 Thus for example to display all the undocumented target-specific
1189 switches supported by the compiler the following can be used:
1192 --help=target,undocumented
1195 The sense of a qualifier can be inverted by prefixing it with the
1196 @var{^} character, so for example to display all binary warning
1197 options (i.e., ones that are either on or off and that do not take an
1198 argument), which have a description the following can be used:
1201 --help=warnings,^joined,^undocumented
1204 A class can also be used as a qualifier, although this usually
1205 restricts the output by so much that there is nothing to display. One
1206 case where it does work however is when one of the classes is
1207 @var{target}. So for example to display all the target-specific
1208 optimization options the following can be used:
1211 --help=target,optimizers
1214 The @option{--help=} option can be repeated on the command line. Each
1215 successive use will display its requested class of options, skipping
1216 those that have already been displayed.
1218 If the @option{-Q} option appears on the command line before the
1219 @option{--help=} option, then the descriptive text displayed by
1220 @option{--help=} is changed. Instead of describing the displayed
1221 options, an indication is given as to whether the option is enabled,
1222 disabled or set to a specific value (assuming that the compiler
1223 knows this at the point where the @option{--help=} option is used).
1225 Here is a truncated example from the ARM port of @command{gcc}:
1228 % gcc -Q -mabi=2 --help=target -c
1229 The following options are target specific:
1231 -mabort-on-noreturn [disabled]
1235 The output is sensitive to the effects of previous command line
1236 options, so for example it is possible to find out which optimizations
1237 are enabled at @option{-O2} by using:
1240 -O2 --help=optimizers
1243 Alternatively you can discover which binary optimizations are enabled
1244 by @option{-O3} by using:
1247 gcc -c -Q -O3 --help=optimizers > /tmp/O3-opts
1248 gcc -c -Q -O2 --help=optimizers > /tmp/O2-opts
1249 diff /tmp/O2-opts /tmp/O3-opts | grep enabled
1254 Display the version number and copyrights of the invoked GCC@.
1258 Invoke all subcommands under a wrapper program. It takes a single
1259 comma separated list as an argument, which will be used to invoke
1263 gcc -c t.c -wrapper gdb,--args
1266 This will invoke all subprograms of gcc under "gdb --args",
1267 thus cc1 invocation will be "gdb --args cc1 ...".
1269 @include @value{srcdir}/../libiberty/at-file.texi
1273 @section Compiling C++ Programs
1275 @cindex suffixes for C++ source
1276 @cindex C++ source file suffixes
1277 C++ source files conventionally use one of the suffixes @samp{.C},
1278 @samp{.cc}, @samp{.cpp}, @samp{.CPP}, @samp{.c++}, @samp{.cp}, or
1279 @samp{.cxx}; C++ header files often use @samp{.hh}, @samp{.hpp},
1280 @samp{.H}, or (for shared template code) @samp{.tcc}; and
1281 preprocessed C++ files use the suffix @samp{.ii}. GCC recognizes
1282 files with these names and compiles them as C++ programs even if you
1283 call the compiler the same way as for compiling C programs (usually
1284 with the name @command{gcc}).
1288 However, the use of @command{gcc} does not add the C++ library.
1289 @command{g++} is a program that calls GCC and treats @samp{.c},
1290 @samp{.h} and @samp{.i} files as C++ source files instead of C source
1291 files unless @option{-x} is used, and automatically specifies linking
1292 against the C++ library. This program is also useful when
1293 precompiling a C header file with a @samp{.h} extension for use in C++
1294 compilations. On many systems, @command{g++} is also installed with
1295 the name @command{c++}.
1297 @cindex invoking @command{g++}
1298 When you compile C++ programs, you may specify many of the same
1299 command-line options that you use for compiling programs in any
1300 language; or command-line options meaningful for C and related
1301 languages; or options that are meaningful only for C++ programs.
1302 @xref{C Dialect Options,,Options Controlling C Dialect}, for
1303 explanations of options for languages related to C@.
1304 @xref{C++ Dialect Options,,Options Controlling C++ Dialect}, for
1305 explanations of options that are meaningful only for C++ programs.
1307 @node C Dialect Options
1308 @section Options Controlling C Dialect
1309 @cindex dialect options
1310 @cindex language dialect options
1311 @cindex options, dialect
1313 The following options control the dialect of C (or languages derived
1314 from C, such as C++, Objective-C and Objective-C++) that the compiler
1318 @cindex ANSI support
1322 In C mode, this is equivalent to @samp{-std=c89}. In C++ mode, it is
1323 equivalent to @samp{-std=c++98}.
1325 This turns off certain features of GCC that are incompatible with ISO
1326 C90 (when compiling C code), or of standard C++ (when compiling C++ code),
1327 such as the @code{asm} and @code{typeof} keywords, and
1328 predefined macros such as @code{unix} and @code{vax} that identify the
1329 type of system you are using. It also enables the undesirable and
1330 rarely used ISO trigraph feature. For the C compiler,
1331 it disables recognition of C++ style @samp{//} comments as well as
1332 the @code{inline} keyword.
1334 The alternate keywords @code{__asm__}, @code{__extension__},
1335 @code{__inline__} and @code{__typeof__} continue to work despite
1336 @option{-ansi}. You would not want to use them in an ISO C program, of
1337 course, but it is useful to put them in header files that might be included
1338 in compilations done with @option{-ansi}. Alternate predefined macros
1339 such as @code{__unix__} and @code{__vax__} are also available, with or
1340 without @option{-ansi}.
1342 The @option{-ansi} option does not cause non-ISO programs to be
1343 rejected gratuitously. For that, @option{-pedantic} is required in
1344 addition to @option{-ansi}. @xref{Warning Options}.
1346 The macro @code{__STRICT_ANSI__} is predefined when the @option{-ansi}
1347 option is used. Some header files may notice this macro and refrain
1348 from declaring certain functions or defining certain macros that the
1349 ISO standard doesn't call for; this is to avoid interfering with any
1350 programs that might use these names for other things.
1352 Functions that would normally be built in but do not have semantics
1353 defined by ISO C (such as @code{alloca} and @code{ffs}) are not built-in
1354 functions when @option{-ansi} is used. @xref{Other Builtins,,Other
1355 built-in functions provided by GCC}, for details of the functions
1360 Determine the language standard. @xref{Standards,,Language Standards
1361 Supported by GCC}, for details of these standard versions. This option
1362 is currently only supported when compiling C or C++.
1364 The compiler can accept several base standards, such as @samp{c89} or
1365 @samp{c++98}, and GNU dialects of those standards, such as
1366 @samp{gnu89} or @samp{gnu++98}. By specifing a base standard, the
1367 compiler will accept all programs following that standard and those
1368 using GNU extensions that do not contradict it. For example,
1369 @samp{-std=c89} turns off certain features of GCC that are
1370 incompatible with ISO C90, such as the @code{asm} and @code{typeof}
1371 keywords, but not other GNU extensions that do not have a meaning in
1372 ISO C90, such as omitting the middle term of a @code{?:}
1373 expression. On the other hand, by specifing a GNU dialect of a
1374 standard, all features the compiler support are enabled, even when
1375 those features change the meaning of the base standard and some
1376 strict-conforming programs may be rejected. The particular standard
1377 is used by @option{-pedantic} to identify which features are GNU
1378 extensions given that version of the standard. For example
1379 @samp{-std=gnu89 -pedantic} would warn about C++ style @samp{//}
1380 comments, while @samp{-std=gnu99 -pedantic} would not.
1382 A value for this option must be provided; possible values are
1387 Support all ISO C90 programs (certain GNU extensions that conflict
1388 with ISO C90 are disabled). Same as @option{-ansi} for C code.
1390 @item iso9899:199409
1391 ISO C90 as modified in amendment 1.
1397 ISO C99. Note that this standard is not yet fully supported; see
1398 @w{@uref{http://gcc.gnu.org/c99status.html}} for more information. The
1399 names @samp{c9x} and @samp{iso9899:199x} are deprecated.
1402 GNU dialect of ISO C90 (including some C99 features). This
1403 is the default for C code.
1407 GNU dialect of ISO C99. When ISO C99 is fully implemented in GCC,
1408 this will become the default. The name @samp{gnu9x} is deprecated.
1411 The 1998 ISO C++ standard plus amendments. Same as @option{-ansi} for
1415 GNU dialect of @option{-std=c++98}. This is the default for
1419 The working draft of the upcoming ISO C++0x standard. This option
1420 enables experimental features that are likely to be included in
1421 C++0x. The working draft is constantly changing, and any feature that is
1422 enabled by this flag may be removed from future versions of GCC if it is
1423 not part of the C++0x standard.
1426 GNU dialect of @option{-std=c++0x}. This option enables
1427 experimental features that may be removed in future versions of GCC.
1430 @item -fgnu89-inline
1431 @opindex fgnu89-inline
1432 The option @option{-fgnu89-inline} tells GCC to use the traditional
1433 GNU semantics for @code{inline} functions when in C99 mode.
1434 @xref{Inline,,An Inline Function is As Fast As a Macro}. This option
1435 is accepted and ignored by GCC versions 4.1.3 up to but not including
1436 4.3. In GCC versions 4.3 and later it changes the behavior of GCC in
1437 C99 mode. Using this option is roughly equivalent to adding the
1438 @code{gnu_inline} function attribute to all inline functions
1439 (@pxref{Function Attributes}).
1441 The option @option{-fno-gnu89-inline} explicitly tells GCC to use the
1442 C99 semantics for @code{inline} when in C99 or gnu99 mode (i.e., it
1443 specifies the default behavior). This option was first supported in
1444 GCC 4.3. This option is not supported in C89 or gnu89 mode.
1446 The preprocessor macros @code{__GNUC_GNU_INLINE__} and
1447 @code{__GNUC_STDC_INLINE__} may be used to check which semantics are
1448 in effect for @code{inline} functions. @xref{Common Predefined
1449 Macros,,,cpp,The C Preprocessor}.
1451 @item -aux-info @var{filename}
1453 Output to the given filename prototyped declarations for all functions
1454 declared and/or defined in a translation unit, including those in header
1455 files. This option is silently ignored in any language other than C@.
1457 Besides declarations, the file indicates, in comments, the origin of
1458 each declaration (source file and line), whether the declaration was
1459 implicit, prototyped or unprototyped (@samp{I}, @samp{N} for new or
1460 @samp{O} for old, respectively, in the first character after the line
1461 number and the colon), and whether it came from a declaration or a
1462 definition (@samp{C} or @samp{F}, respectively, in the following
1463 character). In the case of function definitions, a K&R-style list of
1464 arguments followed by their declarations is also provided, inside
1465 comments, after the declaration.
1469 Do not recognize @code{asm}, @code{inline} or @code{typeof} as a
1470 keyword, so that code can use these words as identifiers. You can use
1471 the keywords @code{__asm__}, @code{__inline__} and @code{__typeof__}
1472 instead. @option{-ansi} implies @option{-fno-asm}.
1474 In C++, this switch only affects the @code{typeof} keyword, since
1475 @code{asm} and @code{inline} are standard keywords. You may want to
1476 use the @option{-fno-gnu-keywords} flag instead, which has the same
1477 effect. In C99 mode (@option{-std=c99} or @option{-std=gnu99}), this
1478 switch only affects the @code{asm} and @code{typeof} keywords, since
1479 @code{inline} is a standard keyword in ISO C99.
1482 @itemx -fno-builtin-@var{function}
1483 @opindex fno-builtin
1484 @cindex built-in functions
1485 Don't recognize built-in functions that do not begin with
1486 @samp{__builtin_} as prefix. @xref{Other Builtins,,Other built-in
1487 functions provided by GCC}, for details of the functions affected,
1488 including those which are not built-in functions when @option{-ansi} or
1489 @option{-std} options for strict ISO C conformance are used because they
1490 do not have an ISO standard meaning.
1492 GCC normally generates special code to handle certain built-in functions
1493 more efficiently; for instance, calls to @code{alloca} may become single
1494 instructions that adjust the stack directly, and calls to @code{memcpy}
1495 may become inline copy loops. The resulting code is often both smaller
1496 and faster, but since the function calls no longer appear as such, you
1497 cannot set a breakpoint on those calls, nor can you change the behavior
1498 of the functions by linking with a different library. In addition,
1499 when a function is recognized as a built-in function, GCC may use
1500 information about that function to warn about problems with calls to
1501 that function, or to generate more efficient code, even if the
1502 resulting code still contains calls to that function. For example,
1503 warnings are given with @option{-Wformat} for bad calls to
1504 @code{printf}, when @code{printf} is built in, and @code{strlen} is
1505 known not to modify global memory.
1507 With the @option{-fno-builtin-@var{function}} option
1508 only the built-in function @var{function} is
1509 disabled. @var{function} must not begin with @samp{__builtin_}. If a
1510 function is named this is not built-in in this version of GCC, this
1511 option is ignored. There is no corresponding
1512 @option{-fbuiltin-@var{function}} option; if you wish to enable
1513 built-in functions selectively when using @option{-fno-builtin} or
1514 @option{-ffreestanding}, you may define macros such as:
1517 #define abs(n) __builtin_abs ((n))
1518 #define strcpy(d, s) __builtin_strcpy ((d), (s))
1523 @cindex hosted environment
1525 Assert that compilation takes place in a hosted environment. This implies
1526 @option{-fbuiltin}. A hosted environment is one in which the
1527 entire standard library is available, and in which @code{main} has a return
1528 type of @code{int}. Examples are nearly everything except a kernel.
1529 This is equivalent to @option{-fno-freestanding}.
1531 @item -ffreestanding
1532 @opindex ffreestanding
1533 @cindex hosted environment
1535 Assert that compilation takes place in a freestanding environment. This
1536 implies @option{-fno-builtin}. A freestanding environment
1537 is one in which the standard library may not exist, and program startup may
1538 not necessarily be at @code{main}. The most obvious example is an OS kernel.
1539 This is equivalent to @option{-fno-hosted}.
1541 @xref{Standards,,Language Standards Supported by GCC}, for details of
1542 freestanding and hosted environments.
1546 @cindex openmp parallel
1547 Enable handling of OpenMP directives @code{#pragma omp} in C/C++ and
1548 @code{!$omp} in Fortran. When @option{-fopenmp} is specified, the
1549 compiler generates parallel code according to the OpenMP Application
1550 Program Interface v2.5 @w{@uref{http://www.openmp.org/}}. This option
1551 implies @option{-pthread}, and thus is only supported on targets that
1552 have support for @option{-pthread}.
1554 @item -fms-extensions
1555 @opindex fms-extensions
1556 Accept some non-standard constructs used in Microsoft header files.
1558 Some cases of unnamed fields in structures and unions are only
1559 accepted with this option. @xref{Unnamed Fields,,Unnamed struct/union
1560 fields within structs/unions}, for details.
1564 Support ISO C trigraphs. The @option{-ansi} option (and @option{-std}
1565 options for strict ISO C conformance) implies @option{-trigraphs}.
1567 @item -no-integrated-cpp
1568 @opindex no-integrated-cpp
1569 Performs a compilation in two passes: preprocessing and compiling. This
1570 option allows a user supplied "cc1", "cc1plus", or "cc1obj" via the
1571 @option{-B} option. The user supplied compilation step can then add in
1572 an additional preprocessing step after normal preprocessing but before
1573 compiling. The default is to use the integrated cpp (internal cpp)
1575 The semantics of this option will change if "cc1", "cc1plus", and
1576 "cc1obj" are merged.
1578 @cindex traditional C language
1579 @cindex C language, traditional
1581 @itemx -traditional-cpp
1582 @opindex traditional-cpp
1583 @opindex traditional
1584 Formerly, these options caused GCC to attempt to emulate a pre-standard
1585 C compiler. They are now only supported with the @option{-E} switch.
1586 The preprocessor continues to support a pre-standard mode. See the GNU
1587 CPP manual for details.
1589 @item -fcond-mismatch
1590 @opindex fcond-mismatch
1591 Allow conditional expressions with mismatched types in the second and
1592 third arguments. The value of such an expression is void. This option
1593 is not supported for C++.
1595 @item -flax-vector-conversions
1596 @opindex flax-vector-conversions
1597 Allow implicit conversions between vectors with differing numbers of
1598 elements and/or incompatible element types. This option should not be
1601 @item -funsigned-char
1602 @opindex funsigned-char
1603 Let the type @code{char} be unsigned, like @code{unsigned char}.
1605 Each kind of machine has a default for what @code{char} should
1606 be. It is either like @code{unsigned char} by default or like
1607 @code{signed char} by default.
1609 Ideally, a portable program should always use @code{signed char} or
1610 @code{unsigned char} when it depends on the signedness of an object.
1611 But many programs have been written to use plain @code{char} and
1612 expect it to be signed, or expect it to be unsigned, depending on the
1613 machines they were written for. This option, and its inverse, let you
1614 make such a program work with the opposite default.
1616 The type @code{char} is always a distinct type from each of
1617 @code{signed char} or @code{unsigned char}, even though its behavior
1618 is always just like one of those two.
1621 @opindex fsigned-char
1622 Let the type @code{char} be signed, like @code{signed char}.
1624 Note that this is equivalent to @option{-fno-unsigned-char}, which is
1625 the negative form of @option{-funsigned-char}. Likewise, the option
1626 @option{-fno-signed-char} is equivalent to @option{-funsigned-char}.
1628 @item -fsigned-bitfields
1629 @itemx -funsigned-bitfields
1630 @itemx -fno-signed-bitfields
1631 @itemx -fno-unsigned-bitfields
1632 @opindex fsigned-bitfields
1633 @opindex funsigned-bitfields
1634 @opindex fno-signed-bitfields
1635 @opindex fno-unsigned-bitfields
1636 These options control whether a bit-field is signed or unsigned, when the
1637 declaration does not use either @code{signed} or @code{unsigned}. By
1638 default, such a bit-field is signed, because this is consistent: the
1639 basic integer types such as @code{int} are signed types.
1642 @node C++ Dialect Options
1643 @section Options Controlling C++ Dialect
1645 @cindex compiler options, C++
1646 @cindex C++ options, command line
1647 @cindex options, C++
1648 This section describes the command-line options that are only meaningful
1649 for C++ programs; but you can also use most of the GNU compiler options
1650 regardless of what language your program is in. For example, you
1651 might compile a file @code{firstClass.C} like this:
1654 g++ -g -frepo -O -c firstClass.C
1658 In this example, only @option{-frepo} is an option meant
1659 only for C++ programs; you can use the other options with any
1660 language supported by GCC@.
1662 Here is a list of options that are @emph{only} for compiling C++ programs:
1666 @item -fabi-version=@var{n}
1667 @opindex fabi-version
1668 Use version @var{n} of the C++ ABI@. Version 2 is the version of the
1669 C++ ABI that first appeared in G++ 3.4. Version 1 is the version of
1670 the C++ ABI that first appeared in G++ 3.2. Version 0 will always be
1671 the version that conforms most closely to the C++ ABI specification.
1672 Therefore, the ABI obtained using version 0 will change as ABI bugs
1675 The default is version 2.
1677 @item -fno-access-control
1678 @opindex fno-access-control
1679 Turn off all access checking. This switch is mainly useful for working
1680 around bugs in the access control code.
1684 Check that the pointer returned by @code{operator new} is non-null
1685 before attempting to modify the storage allocated. This check is
1686 normally unnecessary because the C++ standard specifies that
1687 @code{operator new} will only return @code{0} if it is declared
1688 @samp{throw()}, in which case the compiler will always check the
1689 return value even without this option. In all other cases, when
1690 @code{operator new} has a non-empty exception specification, memory
1691 exhaustion is signalled by throwing @code{std::bad_alloc}. See also
1692 @samp{new (nothrow)}.
1694 @item -fconserve-space
1695 @opindex fconserve-space
1696 Put uninitialized or runtime-initialized global variables into the
1697 common segment, as C does. This saves space in the executable at the
1698 cost of not diagnosing duplicate definitions. If you compile with this
1699 flag and your program mysteriously crashes after @code{main()} has
1700 completed, you may have an object that is being destroyed twice because
1701 two definitions were merged.
1703 This option is no longer useful on most targets, now that support has
1704 been added for putting variables into BSS without making them common.
1706 @item -ffriend-injection
1707 @opindex ffriend-injection
1708 Inject friend functions into the enclosing namespace, so that they are
1709 visible outside the scope of the class in which they are declared.
1710 Friend functions were documented to work this way in the old Annotated
1711 C++ Reference Manual, and versions of G++ before 4.1 always worked
1712 that way. However, in ISO C++ a friend function which is not declared
1713 in an enclosing scope can only be found using argument dependent
1714 lookup. This option causes friends to be injected as they were in
1717 This option is for compatibility, and may be removed in a future
1720 @item -fno-elide-constructors
1721 @opindex fno-elide-constructors
1722 The C++ standard allows an implementation to omit creating a temporary
1723 which is only used to initialize another object of the same type.
1724 Specifying this option disables that optimization, and forces G++ to
1725 call the copy constructor in all cases.
1727 @item -fno-enforce-eh-specs
1728 @opindex fno-enforce-eh-specs
1729 Don't generate code to check for violation of exception specifications
1730 at runtime. This option violates the C++ standard, but may be useful
1731 for reducing code size in production builds, much like defining
1732 @samp{NDEBUG}. This does not give user code permission to throw
1733 exceptions in violation of the exception specifications; the compiler
1734 will still optimize based on the specifications, so throwing an
1735 unexpected exception will result in undefined behavior.
1738 @itemx -fno-for-scope
1740 @opindex fno-for-scope
1741 If @option{-ffor-scope} is specified, the scope of variables declared in
1742 a @i{for-init-statement} is limited to the @samp{for} loop itself,
1743 as specified by the C++ standard.
1744 If @option{-fno-for-scope} is specified, the scope of variables declared in
1745 a @i{for-init-statement} extends to the end of the enclosing scope,
1746 as was the case in old versions of G++, and other (traditional)
1747 implementations of C++.
1749 The default if neither flag is given to follow the standard,
1750 but to allow and give a warning for old-style code that would
1751 otherwise be invalid, or have different behavior.
1753 @item -fno-gnu-keywords
1754 @opindex fno-gnu-keywords
1755 Do not recognize @code{typeof} as a keyword, so that code can use this
1756 word as an identifier. You can use the keyword @code{__typeof__} instead.
1757 @option{-ansi} implies @option{-fno-gnu-keywords}.
1759 @item -fno-implicit-templates
1760 @opindex fno-implicit-templates
1761 Never emit code for non-inline templates which are instantiated
1762 implicitly (i.e.@: by use); only emit code for explicit instantiations.
1763 @xref{Template Instantiation}, for more information.
1765 @item -fno-implicit-inline-templates
1766 @opindex fno-implicit-inline-templates
1767 Don't emit code for implicit instantiations of inline templates, either.
1768 The default is to handle inlines differently so that compiles with and
1769 without optimization will need the same set of explicit instantiations.
1771 @item -fno-implement-inlines
1772 @opindex fno-implement-inlines
1773 To save space, do not emit out-of-line copies of inline functions
1774 controlled by @samp{#pragma implementation}. This will cause linker
1775 errors if these functions are not inlined everywhere they are called.
1777 @item -fms-extensions
1778 @opindex fms-extensions
1779 Disable pedantic warnings about constructs used in MFC, such as implicit
1780 int and getting a pointer to member function via non-standard syntax.
1782 @item -fno-nonansi-builtins
1783 @opindex fno-nonansi-builtins
1784 Disable built-in declarations of functions that are not mandated by
1785 ANSI/ISO C@. These include @code{ffs}, @code{alloca}, @code{_exit},
1786 @code{index}, @code{bzero}, @code{conjf}, and other related functions.
1788 @item -fno-operator-names
1789 @opindex fno-operator-names
1790 Do not treat the operator name keywords @code{and}, @code{bitand},
1791 @code{bitor}, @code{compl}, @code{not}, @code{or} and @code{xor} as
1792 synonyms as keywords.
1794 @item -fno-optional-diags
1795 @opindex fno-optional-diags
1796 Disable diagnostics that the standard says a compiler does not need to
1797 issue. Currently, the only such diagnostic issued by G++ is the one for
1798 a name having multiple meanings within a class.
1801 @opindex fpermissive
1802 Downgrade some diagnostics about nonconformant code from errors to
1803 warnings. Thus, using @option{-fpermissive} will allow some
1804 nonconforming code to compile.
1808 Enable automatic template instantiation at link time. This option also
1809 implies @option{-fno-implicit-templates}. @xref{Template
1810 Instantiation}, for more information.
1814 Disable generation of information about every class with virtual
1815 functions for use by the C++ runtime type identification features
1816 (@samp{dynamic_cast} and @samp{typeid}). If you don't use those parts
1817 of the language, you can save some space by using this flag. Note that
1818 exception handling uses the same information, but it will generate it as
1819 needed. The @samp{dynamic_cast} operator can still be used for casts that
1820 do not require runtime type information, i.e.@: casts to @code{void *} or to
1821 unambiguous base classes.
1825 Emit statistics about front-end processing at the end of the compilation.
1826 This information is generally only useful to the G++ development team.
1828 @item -ftemplate-depth-@var{n}
1829 @opindex ftemplate-depth
1830 Set the maximum instantiation depth for template classes to @var{n}.
1831 A limit on the template instantiation depth is needed to detect
1832 endless recursions during template class instantiation. ANSI/ISO C++
1833 conforming programs must not rely on a maximum depth greater than 17.
1835 @item -fno-threadsafe-statics
1836 @opindex fno-threadsafe-statics
1837 Do not emit the extra code to use the routines specified in the C++
1838 ABI for thread-safe initialization of local statics. You can use this
1839 option to reduce code size slightly in code that doesn't need to be
1842 @item -fuse-cxa-atexit
1843 @opindex fuse-cxa-atexit
1844 Register destructors for objects with static storage duration with the
1845 @code{__cxa_atexit} function rather than the @code{atexit} function.
1846 This option is required for fully standards-compliant handling of static
1847 destructors, but will only work if your C library supports
1848 @code{__cxa_atexit}.
1850 @item -fno-use-cxa-get-exception-ptr
1851 @opindex fno-use-cxa-get-exception-ptr
1852 Don't use the @code{__cxa_get_exception_ptr} runtime routine. This
1853 will cause @code{std::uncaught_exception} to be incorrect, but is necessary
1854 if the runtime routine is not available.
1856 @item -fvisibility-inlines-hidden
1857 @opindex fvisibility-inlines-hidden
1858 This switch declares that the user does not attempt to compare
1859 pointers to inline methods where the addresses of the two functions
1860 were taken in different shared objects.
1862 The effect of this is that GCC may, effectively, mark inline methods with
1863 @code{__attribute__ ((visibility ("hidden")))} so that they do not
1864 appear in the export table of a DSO and do not require a PLT indirection
1865 when used within the DSO@. Enabling this option can have a dramatic effect
1866 on load and link times of a DSO as it massively reduces the size of the
1867 dynamic export table when the library makes heavy use of templates.
1869 The behavior of this switch is not quite the same as marking the
1870 methods as hidden directly, because it does not affect static variables
1871 local to the function or cause the compiler to deduce that
1872 the function is defined in only one shared object.
1874 You may mark a method as having a visibility explicitly to negate the
1875 effect of the switch for that method. For example, if you do want to
1876 compare pointers to a particular inline method, you might mark it as
1877 having default visibility. Marking the enclosing class with explicit
1878 visibility will have no effect.
1880 Explicitly instantiated inline methods are unaffected by this option
1881 as their linkage might otherwise cross a shared library boundary.
1882 @xref{Template Instantiation}.
1884 @item -fvisibility-ms-compat
1885 @opindex fvisibility-ms-compat
1886 This flag attempts to use visibility settings to make GCC's C++
1887 linkage model compatible with that of Microsoft Visual Studio.
1889 The flag makes these changes to GCC's linkage model:
1893 It sets the default visibility to @code{hidden}, like
1894 @option{-fvisibility=hidden}.
1897 Types, but not their members, are not hidden by default.
1900 The One Definition Rule is relaxed for types without explicit
1901 visibility specifications which are defined in more than one different
1902 shared object: those declarations are permitted if they would have
1903 been permitted when this option was not used.
1906 In new code it is better to use @option{-fvisibility=hidden} and
1907 export those classes which are intended to be externally visible.
1908 Unfortunately it is possible for code to rely, perhaps accidentally,
1909 on the Visual Studio behavior.
1911 Among the consequences of these changes are that static data members
1912 of the same type with the same name but defined in different shared
1913 objects will be different, so changing one will not change the other;
1914 and that pointers to function members defined in different shared
1915 objects may not compare equal. When this flag is given, it is a
1916 violation of the ODR to define types with the same name differently.
1920 Do not use weak symbol support, even if it is provided by the linker.
1921 By default, G++ will use weak symbols if they are available. This
1922 option exists only for testing, and should not be used by end-users;
1923 it will result in inferior code and has no benefits. This option may
1924 be removed in a future release of G++.
1928 Do not search for header files in the standard directories specific to
1929 C++, but do still search the other standard directories. (This option
1930 is used when building the C++ library.)
1933 In addition, these optimization, warning, and code generation options
1934 have meanings only for C++ programs:
1937 @item -fno-default-inline
1938 @opindex fno-default-inline
1939 Do not assume @samp{inline} for functions defined inside a class scope.
1940 @xref{Optimize Options,,Options That Control Optimization}. Note that these
1941 functions will have linkage like inline functions; they just won't be
1944 @item -Wabi @r{(C++ and Objective-C++ only)}
1947 Warn when G++ generates code that is probably not compatible with the
1948 vendor-neutral C++ ABI@. Although an effort has been made to warn about
1949 all such cases, there are probably some cases that are not warned about,
1950 even though G++ is generating incompatible code. There may also be
1951 cases where warnings are emitted even though the code that is generated
1954 You should rewrite your code to avoid these warnings if you are
1955 concerned about the fact that code generated by G++ may not be binary
1956 compatible with code generated by other compilers.
1958 The known incompatibilities at this point include:
1963 Incorrect handling of tail-padding for bit-fields. G++ may attempt to
1964 pack data into the same byte as a base class. For example:
1967 struct A @{ virtual void f(); int f1 : 1; @};
1968 struct B : public A @{ int f2 : 1; @};
1972 In this case, G++ will place @code{B::f2} into the same byte
1973 as@code{A::f1}; other compilers will not. You can avoid this problem
1974 by explicitly padding @code{A} so that its size is a multiple of the
1975 byte size on your platform; that will cause G++ and other compilers to
1976 layout @code{B} identically.
1979 Incorrect handling of tail-padding for virtual bases. G++ does not use
1980 tail padding when laying out virtual bases. For example:
1983 struct A @{ virtual void f(); char c1; @};
1984 struct B @{ B(); char c2; @};
1985 struct C : public A, public virtual B @{@};
1989 In this case, G++ will not place @code{B} into the tail-padding for
1990 @code{A}; other compilers will. You can avoid this problem by
1991 explicitly padding @code{A} so that its size is a multiple of its
1992 alignment (ignoring virtual base classes); that will cause G++ and other
1993 compilers to layout @code{C} identically.
1996 Incorrect handling of bit-fields with declared widths greater than that
1997 of their underlying types, when the bit-fields appear in a union. For
2001 union U @{ int i : 4096; @};
2005 Assuming that an @code{int} does not have 4096 bits, G++ will make the
2006 union too small by the number of bits in an @code{int}.
2009 Empty classes can be placed at incorrect offsets. For example:
2019 struct C : public B, public A @{@};
2023 G++ will place the @code{A} base class of @code{C} at a nonzero offset;
2024 it should be placed at offset zero. G++ mistakenly believes that the
2025 @code{A} data member of @code{B} is already at offset zero.
2028 Names of template functions whose types involve @code{typename} or
2029 template template parameters can be mangled incorrectly.
2032 template <typename Q>
2033 void f(typename Q::X) @{@}
2035 template <template <typename> class Q>
2036 void f(typename Q<int>::X) @{@}
2040 Instantiations of these templates may be mangled incorrectly.
2044 @item -Wctor-dtor-privacy @r{(C++ and Objective-C++ only)}
2045 @opindex Wctor-dtor-privacy
2046 @opindex Wno-ctor-dtor-privacy
2047 Warn when a class seems unusable because all the constructors or
2048 destructors in that class are private, and it has neither friends nor
2049 public static member functions.
2051 @item -Wnon-virtual-dtor @r{(C++ and Objective-C++ only)}
2052 @opindex Wnon-virtual-dtor
2053 @opindex Wno-non-virtual-dtor
2054 Warn when a class has virtual functions and accessible non-virtual
2055 destructor, in which case it would be possible but unsafe to delete
2056 an instance of a derived class through a pointer to the base class.
2057 This warning is also enabled if -Weffc++ is specified.
2059 @item -Wreorder @r{(C++ and Objective-C++ only)}
2061 @opindex Wno-reorder
2062 @cindex reordering, warning
2063 @cindex warning for reordering of member initializers
2064 Warn when the order of member initializers given in the code does not
2065 match the order in which they must be executed. For instance:
2071 A(): j (0), i (1) @{ @}
2075 The compiler will rearrange the member initializers for @samp{i}
2076 and @samp{j} to match the declaration order of the members, emitting
2077 a warning to that effect. This warning is enabled by @option{-Wall}.
2080 The following @option{-W@dots{}} options are not affected by @option{-Wall}.
2083 @item -Weffc++ @r{(C++ and Objective-C++ only)}
2086 Warn about violations of the following style guidelines from Scott Meyers'
2087 @cite{Effective C++} book:
2091 Item 11: Define a copy constructor and an assignment operator for classes
2092 with dynamically allocated memory.
2095 Item 12: Prefer initialization to assignment in constructors.
2098 Item 14: Make destructors virtual in base classes.
2101 Item 15: Have @code{operator=} return a reference to @code{*this}.
2104 Item 23: Don't try to return a reference when you must return an object.
2108 Also warn about violations of the following style guidelines from
2109 Scott Meyers' @cite{More Effective C++} book:
2113 Item 6: Distinguish between prefix and postfix forms of increment and
2114 decrement operators.
2117 Item 7: Never overload @code{&&}, @code{||}, or @code{,}.
2121 When selecting this option, be aware that the standard library
2122 headers do not obey all of these guidelines; use @samp{grep -v}
2123 to filter out those warnings.
2125 @item -Wstrict-null-sentinel @r{(C++ and Objective-C++ only)}
2126 @opindex Wstrict-null-sentinel
2127 @opindex Wno-strict-null-sentinel
2128 Warn also about the use of an uncasted @code{NULL} as sentinel. When
2129 compiling only with GCC this is a valid sentinel, as @code{NULL} is defined
2130 to @code{__null}. Although it is a null pointer constant not a null pointer,
2131 it is guaranteed to of the same size as a pointer. But this use is
2132 not portable across different compilers.
2134 @item -Wno-non-template-friend @r{(C++ and Objective-C++ only)}
2135 @opindex Wno-non-template-friend
2136 @opindex Wnon-template-friend
2137 Disable warnings when non-templatized friend functions are declared
2138 within a template. Since the advent of explicit template specification
2139 support in G++, if the name of the friend is an unqualified-id (i.e.,
2140 @samp{friend foo(int)}), the C++ language specification demands that the
2141 friend declare or define an ordinary, nontemplate function. (Section
2142 14.5.3). Before G++ implemented explicit specification, unqualified-ids
2143 could be interpreted as a particular specialization of a templatized
2144 function. Because this non-conforming behavior is no longer the default
2145 behavior for G++, @option{-Wnon-template-friend} allows the compiler to
2146 check existing code for potential trouble spots and is on by default.
2147 This new compiler behavior can be turned off with
2148 @option{-Wno-non-template-friend} which keeps the conformant compiler code
2149 but disables the helpful warning.
2151 @item -Wold-style-cast @r{(C++ and Objective-C++ only)}
2152 @opindex Wold-style-cast
2153 @opindex Wno-old-style-cast
2154 Warn if an old-style (C-style) cast to a non-void type is used within
2155 a C++ program. The new-style casts (@samp{dynamic_cast},
2156 @samp{static_cast}, @samp{reinterpret_cast}, and @samp{const_cast}) are
2157 less vulnerable to unintended effects and much easier to search for.
2159 @item -Woverloaded-virtual @r{(C++ and Objective-C++ only)}
2160 @opindex Woverloaded-virtual
2161 @opindex Wno-overloaded-virtual
2162 @cindex overloaded virtual fn, warning
2163 @cindex warning for overloaded virtual fn
2164 Warn when a function declaration hides virtual functions from a
2165 base class. For example, in:
2172 struct B: public A @{
2177 the @code{A} class version of @code{f} is hidden in @code{B}, and code
2185 will fail to compile.
2187 @item -Wno-pmf-conversions @r{(C++ and Objective-C++ only)}
2188 @opindex Wno-pmf-conversions
2189 @opindex Wpmf-conversions
2190 Disable the diagnostic for converting a bound pointer to member function
2193 @item -Wsign-promo @r{(C++ and Objective-C++ only)}
2194 @opindex Wsign-promo
2195 @opindex Wno-sign-promo
2196 Warn when overload resolution chooses a promotion from unsigned or
2197 enumerated type to a signed type, over a conversion to an unsigned type of
2198 the same size. Previous versions of G++ would try to preserve
2199 unsignedness, but the standard mandates the current behavior.
2204 A& operator = (int);
2214 In this example, G++ will synthesize a default @samp{A& operator =
2215 (const A&);}, while cfront will use the user-defined @samp{operator =}.
2218 @node Objective-C and Objective-C++ Dialect Options
2219 @section Options Controlling Objective-C and Objective-C++ Dialects
2221 @cindex compiler options, Objective-C and Objective-C++
2222 @cindex Objective-C and Objective-C++ options, command line
2223 @cindex options, Objective-C and Objective-C++
2224 (NOTE: This manual does not describe the Objective-C and Objective-C++
2225 languages themselves. See @xref{Standards,,Language Standards
2226 Supported by GCC}, for references.)
2228 This section describes the command-line options that are only meaningful
2229 for Objective-C and Objective-C++ programs, but you can also use most of
2230 the language-independent GNU compiler options.
2231 For example, you might compile a file @code{some_class.m} like this:
2234 gcc -g -fgnu-runtime -O -c some_class.m
2238 In this example, @option{-fgnu-runtime} is an option meant only for
2239 Objective-C and Objective-C++ programs; you can use the other options with
2240 any language supported by GCC@.
2242 Note that since Objective-C is an extension of the C language, Objective-C
2243 compilations may also use options specific to the C front-end (e.g.,
2244 @option{-Wtraditional}). Similarly, Objective-C++ compilations may use
2245 C++-specific options (e.g., @option{-Wabi}).
2247 Here is a list of options that are @emph{only} for compiling Objective-C
2248 and Objective-C++ programs:
2251 @item -fconstant-string-class=@var{class-name}
2252 @opindex fconstant-string-class
2253 Use @var{class-name} as the name of the class to instantiate for each
2254 literal string specified with the syntax @code{@@"@dots{}"}. The default
2255 class name is @code{NXConstantString} if the GNU runtime is being used, and
2256 @code{NSConstantString} if the NeXT runtime is being used (see below). The
2257 @option{-fconstant-cfstrings} option, if also present, will override the
2258 @option{-fconstant-string-class} setting and cause @code{@@"@dots{}"} literals
2259 to be laid out as constant CoreFoundation strings.
2262 @opindex fgnu-runtime
2263 Generate object code compatible with the standard GNU Objective-C
2264 runtime. This is the default for most types of systems.
2266 @item -fnext-runtime
2267 @opindex fnext-runtime
2268 Generate output compatible with the NeXT runtime. This is the default
2269 for NeXT-based systems, including Darwin and Mac OS X@. The macro
2270 @code{__NEXT_RUNTIME__} is predefined if (and only if) this option is
2273 @item -fno-nil-receivers
2274 @opindex fno-nil-receivers
2275 Assume that all Objective-C message dispatches (e.g.,
2276 @code{[receiver message:arg]}) in this translation unit ensure that the receiver
2277 is not @code{nil}. This allows for more efficient entry points in the runtime
2278 to be used. Currently, this option is only available in conjunction with
2279 the NeXT runtime on Mac OS X 10.3 and later.
2281 @item -fobjc-call-cxx-cdtors
2282 @opindex fobjc-call-cxx-cdtors
2283 For each Objective-C class, check if any of its instance variables is a
2284 C++ object with a non-trivial default constructor. If so, synthesize a
2285 special @code{- (id) .cxx_construct} instance method that will run
2286 non-trivial default constructors on any such instance variables, in order,
2287 and then return @code{self}. Similarly, check if any instance variable
2288 is a C++ object with a non-trivial destructor, and if so, synthesize a
2289 special @code{- (void) .cxx_destruct} method that will run
2290 all such default destructors, in reverse order.
2292 The @code{- (id) .cxx_construct} and/or @code{- (void) .cxx_destruct} methods
2293 thusly generated will only operate on instance variables declared in the
2294 current Objective-C class, and not those inherited from superclasses. It
2295 is the responsibility of the Objective-C runtime to invoke all such methods
2296 in an object's inheritance hierarchy. The @code{- (id) .cxx_construct} methods
2297 will be invoked by the runtime immediately after a new object
2298 instance is allocated; the @code{- (void) .cxx_destruct} methods will
2299 be invoked immediately before the runtime deallocates an object instance.
2301 As of this writing, only the NeXT runtime on Mac OS X 10.4 and later has
2302 support for invoking the @code{- (id) .cxx_construct} and
2303 @code{- (void) .cxx_destruct} methods.
2305 @item -fobjc-direct-dispatch
2306 @opindex fobjc-direct-dispatch
2307 Allow fast jumps to the message dispatcher. On Darwin this is
2308 accomplished via the comm page.
2310 @item -fobjc-exceptions
2311 @opindex fobjc-exceptions
2312 Enable syntactic support for structured exception handling in Objective-C,
2313 similar to what is offered by C++ and Java. This option is
2314 unavailable in conjunction with the NeXT runtime on Mac OS X 10.2 and
2323 @@catch (AnObjCClass *exc) @{
2330 @@catch (AnotherClass *exc) @{
2333 @@catch (id allOthers) @{
2343 The @code{@@throw} statement may appear anywhere in an Objective-C or
2344 Objective-C++ program; when used inside of a @code{@@catch} block, the
2345 @code{@@throw} may appear without an argument (as shown above), in which case
2346 the object caught by the @code{@@catch} will be rethrown.
2348 Note that only (pointers to) Objective-C objects may be thrown and
2349 caught using this scheme. When an object is thrown, it will be caught
2350 by the nearest @code{@@catch} clause capable of handling objects of that type,
2351 analogously to how @code{catch} blocks work in C++ and Java. A
2352 @code{@@catch(id @dots{})} clause (as shown above) may also be provided to catch
2353 any and all Objective-C exceptions not caught by previous @code{@@catch}
2356 The @code{@@finally} clause, if present, will be executed upon exit from the
2357 immediately preceding @code{@@try @dots{} @@catch} section. This will happen
2358 regardless of whether any exceptions are thrown, caught or rethrown
2359 inside the @code{@@try @dots{} @@catch} section, analogously to the behavior
2360 of the @code{finally} clause in Java.
2362 There are several caveats to using the new exception mechanism:
2366 Although currently designed to be binary compatible with @code{NS_HANDLER}-style
2367 idioms provided by the @code{NSException} class, the new
2368 exceptions can only be used on Mac OS X 10.3 (Panther) and later
2369 systems, due to additional functionality needed in the (NeXT) Objective-C
2373 As mentioned above, the new exceptions do not support handling
2374 types other than Objective-C objects. Furthermore, when used from
2375 Objective-C++, the Objective-C exception model does not interoperate with C++
2376 exceptions at this time. This means you cannot @code{@@throw} an exception
2377 from Objective-C and @code{catch} it in C++, or vice versa
2378 (i.e., @code{throw @dots{} @@catch}).
2381 The @option{-fobjc-exceptions} switch also enables the use of synchronization
2382 blocks for thread-safe execution:
2385 @@synchronized (ObjCClass *guard) @{
2390 Upon entering the @code{@@synchronized} block, a thread of execution shall
2391 first check whether a lock has been placed on the corresponding @code{guard}
2392 object by another thread. If it has, the current thread shall wait until
2393 the other thread relinquishes its lock. Once @code{guard} becomes available,
2394 the current thread will place its own lock on it, execute the code contained in
2395 the @code{@@synchronized} block, and finally relinquish the lock (thereby
2396 making @code{guard} available to other threads).
2398 Unlike Java, Objective-C does not allow for entire methods to be marked
2399 @code{@@synchronized}. Note that throwing exceptions out of
2400 @code{@@synchronized} blocks is allowed, and will cause the guarding object
2401 to be unlocked properly.
2405 Enable garbage collection (GC) in Objective-C and Objective-C++ programs.
2407 @item -freplace-objc-classes
2408 @opindex freplace-objc-classes
2409 Emit a special marker instructing @command{ld(1)} not to statically link in
2410 the resulting object file, and allow @command{dyld(1)} to load it in at
2411 run time instead. This is used in conjunction with the Fix-and-Continue
2412 debugging mode, where the object file in question may be recompiled and
2413 dynamically reloaded in the course of program execution, without the need
2414 to restart the program itself. Currently, Fix-and-Continue functionality
2415 is only available in conjunction with the NeXT runtime on Mac OS X 10.3
2420 When compiling for the NeXT runtime, the compiler ordinarily replaces calls
2421 to @code{objc_getClass("@dots{}")} (when the name of the class is known at
2422 compile time) with static class references that get initialized at load time,
2423 which improves run-time performance. Specifying the @option{-fzero-link} flag
2424 suppresses this behavior and causes calls to @code{objc_getClass("@dots{}")}
2425 to be retained. This is useful in Zero-Link debugging mode, since it allows
2426 for individual class implementations to be modified during program execution.
2430 Dump interface declarations for all classes seen in the source file to a
2431 file named @file{@var{sourcename}.decl}.
2433 @item -Wassign-intercept @r{(Objective-C and Objective-C++ only)}
2434 @opindex Wassign-intercept
2435 @opindex Wno-assign-intercept
2436 Warn whenever an Objective-C assignment is being intercepted by the
2439 @item -Wno-protocol @r{(Objective-C and Objective-C++ only)}
2440 @opindex Wno-protocol
2442 If a class is declared to implement a protocol, a warning is issued for
2443 every method in the protocol that is not implemented by the class. The
2444 default behavior is to issue a warning for every method not explicitly
2445 implemented in the class, even if a method implementation is inherited
2446 from the superclass. If you use the @option{-Wno-protocol} option, then
2447 methods inherited from the superclass are considered to be implemented,
2448 and no warning is issued for them.
2450 @item -Wselector @r{(Objective-C and Objective-C++ only)}
2452 @opindex Wno-selector
2453 Warn if multiple methods of different types for the same selector are
2454 found during compilation. The check is performed on the list of methods
2455 in the final stage of compilation. Additionally, a check is performed
2456 for each selector appearing in a @code{@@selector(@dots{})}
2457 expression, and a corresponding method for that selector has been found
2458 during compilation. Because these checks scan the method table only at
2459 the end of compilation, these warnings are not produced if the final
2460 stage of compilation is not reached, for example because an error is
2461 found during compilation, or because the @option{-fsyntax-only} option is
2464 @item -Wstrict-selector-match @r{(Objective-C and Objective-C++ only)}
2465 @opindex Wstrict-selector-match
2466 @opindex Wno-strict-selector-match
2467 Warn if multiple methods with differing argument and/or return types are
2468 found for a given selector when attempting to send a message using this
2469 selector to a receiver of type @code{id} or @code{Class}. When this flag
2470 is off (which is the default behavior), the compiler will omit such warnings
2471 if any differences found are confined to types which share the same size
2474 @item -Wundeclared-selector @r{(Objective-C and Objective-C++ only)}
2475 @opindex Wundeclared-selector
2476 @opindex Wno-undeclared-selector
2477 Warn if a @code{@@selector(@dots{})} expression referring to an
2478 undeclared selector is found. A selector is considered undeclared if no
2479 method with that name has been declared before the
2480 @code{@@selector(@dots{})} expression, either explicitly in an
2481 @code{@@interface} or @code{@@protocol} declaration, or implicitly in
2482 an @code{@@implementation} section. This option always performs its
2483 checks as soon as a @code{@@selector(@dots{})} expression is found,
2484 while @option{-Wselector} only performs its checks in the final stage of
2485 compilation. This also enforces the coding style convention
2486 that methods and selectors must be declared before being used.
2488 @item -print-objc-runtime-info
2489 @opindex print-objc-runtime-info
2490 Generate C header describing the largest structure that is passed by
2495 @node Language Independent Options
2496 @section Options to Control Diagnostic Messages Formatting
2497 @cindex options to control diagnostics formatting
2498 @cindex diagnostic messages
2499 @cindex message formatting
2501 Traditionally, diagnostic messages have been formatted irrespective of
2502 the output device's aspect (e.g.@: its width, @dots{}). The options described
2503 below can be used to control the diagnostic messages formatting
2504 algorithm, e.g.@: how many characters per line, how often source location
2505 information should be reported. Right now, only the C++ front end can
2506 honor these options. However it is expected, in the near future, that
2507 the remaining front ends would be able to digest them correctly.
2510 @item -fmessage-length=@var{n}
2511 @opindex fmessage-length
2512 Try to format error messages so that they fit on lines of about @var{n}
2513 characters. The default is 72 characters for @command{g++} and 0 for the rest of
2514 the front ends supported by GCC@. If @var{n} is zero, then no
2515 line-wrapping will be done; each error message will appear on a single
2518 @opindex fdiagnostics-show-location
2519 @item -fdiagnostics-show-location=once
2520 Only meaningful in line-wrapping mode. Instructs the diagnostic messages
2521 reporter to emit @emph{once} source location information; that is, in
2522 case the message is too long to fit on a single physical line and has to
2523 be wrapped, the source location won't be emitted (as prefix) again,
2524 over and over, in subsequent continuation lines. This is the default
2527 @item -fdiagnostics-show-location=every-line
2528 Only meaningful in line-wrapping mode. Instructs the diagnostic
2529 messages reporter to emit the same source location information (as
2530 prefix) for physical lines that result from the process of breaking
2531 a message which is too long to fit on a single line.
2533 @item -fdiagnostics-show-option
2534 @opindex fdiagnostics-show-option
2535 This option instructs the diagnostic machinery to add text to each
2536 diagnostic emitted, which indicates which command line option directly
2537 controls that diagnostic, when such an option is known to the
2538 diagnostic machinery.
2540 @item -Wcoverage-mismatch
2541 @opindex Wcoverage-mismatch
2542 Warn if feedback profiles do not match when using the
2543 @option{-fprofile-use} option.
2544 If a source file was changed between @option{-fprofile-gen} and
2545 @option{-fprofile-use}, the files with the profile feedback can fail
2546 to match the source file and GCC can not use the profile feedback
2547 information. By default, GCC emits an error message in this case.
2548 The option @option{-Wcoverage-mismatch} emits a warning instead of an
2549 error. GCC does not use appropriate feedback profiles, so using this
2550 option can result in poorly optimized code. This option is useful
2551 only in the case of very minor changes such as bug fixes to an
2556 @node Warning Options
2557 @section Options to Request or Suppress Warnings
2558 @cindex options to control warnings
2559 @cindex warning messages
2560 @cindex messages, warning
2561 @cindex suppressing warnings
2563 Warnings are diagnostic messages that report constructions which
2564 are not inherently erroneous but which are risky or suggest there
2565 may have been an error.
2567 The following language-independent options do not enable specific
2568 warnings but control the kinds of diagnostics produced by GCC.
2571 @cindex syntax checking
2573 @opindex fsyntax-only
2574 Check the code for syntax errors, but don't do anything beyond that.
2578 Inhibit all warning messages.
2583 Make all warnings into errors.
2588 Make the specified warning into an error. The specifier for a warning
2589 is appended, for example @option{-Werror=switch} turns the warnings
2590 controlled by @option{-Wswitch} into errors. This switch takes a
2591 negative form, to be used to negate @option{-Werror} for specific
2592 warnings, for example @option{-Wno-error=switch} makes
2593 @option{-Wswitch} warnings not be errors, even when @option{-Werror}
2594 is in effect. You can use the @option{-fdiagnostics-show-option}
2595 option to have each controllable warning amended with the option which
2596 controls it, to determine what to use with this option.
2598 Note that specifying @option{-Werror=}@var{foo} automatically implies
2599 @option{-W}@var{foo}. However, @option{-Wno-error=}@var{foo} does not
2602 @item -Wfatal-errors
2603 @opindex Wfatal-errors
2604 @opindex Wno-fatal-errors
2605 This option causes the compiler to abort compilation on the first error
2606 occurred rather than trying to keep going and printing further error
2611 You can request many specific warnings with options beginning
2612 @samp{-W}, for example @option{-Wimplicit} to request warnings on
2613 implicit declarations. Each of these specific warning options also
2614 has a negative form beginning @samp{-Wno-} to turn off warnings; for
2615 example, @option{-Wno-implicit}. This manual lists only one of the
2616 two forms, whichever is not the default. For further,
2617 language-specific options also refer to @ref{C++ Dialect Options} and
2618 @ref{Objective-C and Objective-C++ Dialect Options}.
2623 Issue all the warnings demanded by strict ISO C and ISO C++;
2624 reject all programs that use forbidden extensions, and some other
2625 programs that do not follow ISO C and ISO C++. For ISO C, follows the
2626 version of the ISO C standard specified by any @option{-std} option used.
2628 Valid ISO C and ISO C++ programs should compile properly with or without
2629 this option (though a rare few will require @option{-ansi} or a
2630 @option{-std} option specifying the required version of ISO C)@. However,
2631 without this option, certain GNU extensions and traditional C and C++
2632 features are supported as well. With this option, they are rejected.
2634 @option{-pedantic} does not cause warning messages for use of the
2635 alternate keywords whose names begin and end with @samp{__}. Pedantic
2636 warnings are also disabled in the expression that follows
2637 @code{__extension__}. However, only system header files should use
2638 these escape routes; application programs should avoid them.
2639 @xref{Alternate Keywords}.
2641 Some users try to use @option{-pedantic} to check programs for strict ISO
2642 C conformance. They soon find that it does not do quite what they want:
2643 it finds some non-ISO practices, but not all---only those for which
2644 ISO C @emph{requires} a diagnostic, and some others for which
2645 diagnostics have been added.
2647 A feature to report any failure to conform to ISO C might be useful in
2648 some instances, but would require considerable additional work and would
2649 be quite different from @option{-pedantic}. We don't have plans to
2650 support such a feature in the near future.
2652 Where the standard specified with @option{-std} represents a GNU
2653 extended dialect of C, such as @samp{gnu89} or @samp{gnu99}, there is a
2654 corresponding @dfn{base standard}, the version of ISO C on which the GNU
2655 extended dialect is based. Warnings from @option{-pedantic} are given
2656 where they are required by the base standard. (It would not make sense
2657 for such warnings to be given only for features not in the specified GNU
2658 C dialect, since by definition the GNU dialects of C include all
2659 features the compiler supports with the given option, and there would be
2660 nothing to warn about.)
2662 @item -pedantic-errors
2663 @opindex pedantic-errors
2664 Like @option{-pedantic}, except that errors are produced rather than
2670 This enables all the warnings about constructions that some users
2671 consider questionable, and that are easy to avoid (or modify to
2672 prevent the warning), even in conjunction with macros. This also
2673 enables some language-specific warnings described in @ref{C++ Dialect
2674 Options} and @ref{Objective-C and Objective-C++ Dialect Options}.
2676 @option{-Wall} turns on the following warning flags:
2678 @gccoptlist{-Waddress @gol
2679 -Warray-bounds @r{(only with} @option{-O2}@r{)} @gol
2681 -Wchar-subscripts @gol
2683 -Wimplicit-function-declaration @gol
2686 -Wmain @r{(only for C/ObjC and unless} @option{-ffreestanding}@r{)} @gol
2687 -Wmissing-braces @gol
2693 -Wsequence-point @gol
2694 -Wsign-compare @r{(only in C++)} @gol
2695 -Wstrict-aliasing @gol
2696 -Wstrict-overflow=1 @gol
2699 -Wuninitialized @r{(only with} @option{-O1} @r{and above)} @gol
2700 -Wunknown-pragmas @gol
2701 -Wunused-function @gol
2704 -Wunused-variable @gol
2705 -Wvolatile-register-var @gol
2708 Note that some warning flags are not implied by @option{-Wall}. Some of
2709 them warn about constructions that users generally do not consider
2710 questionable, but which occasionally you might wish to check for;
2711 others warn about constructions that are necessary or hard to avoid in
2712 some cases, and there is no simple way to modify the code to suppress
2713 the warning. Some of them are enabled by @option{-Wextra} but many of
2714 them must be enabled individually.
2720 This enables some extra warning flags that are not enabled by
2721 @option{-Wall}. (This option used to be called @option{-W}. The older
2722 name is still supported, but the newer name is more descriptive.)
2724 @gccoptlist{-Wclobbered @gol
2726 -Wignored-qualifiers @gol
2727 -Wmissing-field-initializers @gol
2728 -Wmissing-parameter-type @r{(C only)} @gol
2729 -Wold-style-declaration @r{(C only)} @gol
2730 -Woverride-init @gol
2733 -Wuninitialized @r{(only with} @option{-O1} @r{and above)} @gol
2734 -Wunused-parameter @r{(only with} @option{-Wunused} @r{or} @option{-Wall}@r{)} @gol
2737 The option @option{-Wextra} also prints warning messages for the
2743 A pointer is compared against integer zero with @samp{<}, @samp{<=},
2744 @samp{>}, or @samp{>=}.
2747 (C++ only) An enumerator and a non-enumerator both appear in a
2748 conditional expression.
2751 (C++ only) A non-static reference or non-static @samp{const} member
2752 appears in a class without constructors.
2755 (C++ only) Ambiguous virtual bases.
2758 (C++ only) Subscripting an array which has been declared @samp{register}.
2761 (C++ only) Taking the address of a variable which has been declared
2765 (C++ only) A base class is not initialized in a derived class' copy
2770 @item -Wchar-subscripts
2771 @opindex Wchar-subscripts
2772 @opindex Wno-char-subscripts
2773 Warn if an array subscript has type @code{char}. This is a common cause
2774 of error, as programmers often forget that this type is signed on some
2776 This warning is enabled by @option{-Wall}.
2780 @opindex Wno-comment
2781 Warn whenever a comment-start sequence @samp{/*} appears in a @samp{/*}
2782 comment, or whenever a Backslash-Newline appears in a @samp{//} comment.
2783 This warning is enabled by @option{-Wall}.
2788 @opindex ffreestanding
2789 @opindex fno-builtin
2790 Check calls to @code{printf} and @code{scanf}, etc., to make sure that
2791 the arguments supplied have types appropriate to the format string
2792 specified, and that the conversions specified in the format string make
2793 sense. This includes standard functions, and others specified by format
2794 attributes (@pxref{Function Attributes}), in the @code{printf},
2795 @code{scanf}, @code{strftime} and @code{strfmon} (an X/Open extension,
2796 not in the C standard) families (or other target-specific families).
2797 Which functions are checked without format attributes having been
2798 specified depends on the standard version selected, and such checks of
2799 functions without the attribute specified are disabled by
2800 @option{-ffreestanding} or @option{-fno-builtin}.
2802 The formats are checked against the format features supported by GNU
2803 libc version 2.2. These include all ISO C90 and C99 features, as well
2804 as features from the Single Unix Specification and some BSD and GNU
2805 extensions. Other library implementations may not support all these
2806 features; GCC does not support warning about features that go beyond a
2807 particular library's limitations. However, if @option{-pedantic} is used
2808 with @option{-Wformat}, warnings will be given about format features not
2809 in the selected standard version (but not for @code{strfmon} formats,
2810 since those are not in any version of the C standard). @xref{C Dialect
2811 Options,,Options Controlling C Dialect}.
2813 Since @option{-Wformat} also checks for null format arguments for
2814 several functions, @option{-Wformat} also implies @option{-Wnonnull}.
2816 @option{-Wformat} is included in @option{-Wall}. For more control over some
2817 aspects of format checking, the options @option{-Wformat-y2k},
2818 @option{-Wno-format-extra-args}, @option{-Wno-format-zero-length},
2819 @option{-Wformat-nonliteral}, @option{-Wformat-security}, and
2820 @option{-Wformat=2} are available, but are not included in @option{-Wall}.
2823 @opindex Wformat-y2k
2824 @opindex Wno-format-y2k
2825 If @option{-Wformat} is specified, also warn about @code{strftime}
2826 formats which may yield only a two-digit year.
2828 @item -Wno-format-contains-nul
2829 @opindex Wno-format-contains-nul
2830 @opindex Wformat-contains-nul
2831 If @option{-Wformat} is specified, do not warn about format strings that
2834 @item -Wno-format-extra-args
2835 @opindex Wno-format-extra-args
2836 @opindex Wformat-extra-args
2837 If @option{-Wformat} is specified, do not warn about excess arguments to a
2838 @code{printf} or @code{scanf} format function. The C standard specifies
2839 that such arguments are ignored.
2841 Where the unused arguments lie between used arguments that are
2842 specified with @samp{$} operand number specifications, normally
2843 warnings are still given, since the implementation could not know what
2844 type to pass to @code{va_arg} to skip the unused arguments. However,
2845 in the case of @code{scanf} formats, this option will suppress the
2846 warning if the unused arguments are all pointers, since the Single
2847 Unix Specification says that such unused arguments are allowed.
2849 @item -Wno-format-zero-length @r{(C and Objective-C only)}
2850 @opindex Wno-format-zero-length
2851 @opindex Wformat-zero-length
2852 If @option{-Wformat} is specified, do not warn about zero-length formats.
2853 The C standard specifies that zero-length formats are allowed.
2855 @item -Wformat-nonliteral
2856 @opindex Wformat-nonliteral
2857 @opindex Wno-format-nonliteral
2858 If @option{-Wformat} is specified, also warn if the format string is not a
2859 string literal and so cannot be checked, unless the format function
2860 takes its format arguments as a @code{va_list}.
2862 @item -Wformat-security
2863 @opindex Wformat-security
2864 @opindex Wno-format-security
2865 If @option{-Wformat} is specified, also warn about uses of format
2866 functions that represent possible security problems. At present, this
2867 warns about calls to @code{printf} and @code{scanf} functions where the
2868 format string is not a string literal and there are no format arguments,
2869 as in @code{printf (foo);}. This may be a security hole if the format
2870 string came from untrusted input and contains @samp{%n}. (This is
2871 currently a subset of what @option{-Wformat-nonliteral} warns about, but
2872 in future warnings may be added to @option{-Wformat-security} that are not
2873 included in @option{-Wformat-nonliteral}.)
2877 @opindex Wno-format=2
2878 Enable @option{-Wformat} plus format checks not included in
2879 @option{-Wformat}. Currently equivalent to @samp{-Wformat
2880 -Wformat-nonliteral -Wformat-security -Wformat-y2k}.
2882 @item -Wnonnull @r{(C and Objective-C only)}
2884 @opindex Wno-nonnull
2885 Warn about passing a null pointer for arguments marked as
2886 requiring a non-null value by the @code{nonnull} function attribute.
2888 @option{-Wnonnull} is included in @option{-Wall} and @option{-Wformat}. It
2889 can be disabled with the @option{-Wno-nonnull} option.
2891 @item -Winit-self @r{(C, C++, Objective-C and Objective-C++ only)}
2893 @opindex Wno-init-self
2894 Warn about uninitialized variables which are initialized with themselves.
2895 Note this option can only be used with the @option{-Wuninitialized} option,
2896 which in turn only works with @option{-O1} and above.
2898 For example, GCC will warn about @code{i} being uninitialized in the
2899 following snippet only when @option{-Winit-self} has been specified:
2910 @item -Wimplicit-int @r{(C and Objective-C only)}
2911 @opindex Wimplicit-int
2912 @opindex Wno-implicit-int
2913 Warn when a declaration does not specify a type.
2914 This warning is enabled by @option{-Wall}.
2916 @item -Wimplicit-function-declaration @r{(C and Objective-C only)}
2917 @opindex Wimplicit-function-declaration
2918 @opindex Wno-implicit-function-declaration
2919 Give a warning whenever a function is used before being declared. In
2920 C99 mode (@option{-std=c99} or @option{-std=gnu99}), this warning is
2921 enabled by default and it is made into an error by
2922 @option{-pedantic-errors}. This warning is also enabled by
2927 @opindex Wno-implicit
2928 Same as @option{-Wimplicit-int} and @option{-Wimplicit-function-declaration}.
2929 This warning is enabled by @option{-Wall}.
2931 @item -Wignored-qualifiers @r{(C and C++ only)}
2932 @opindex Wignored-qualifiers
2933 @opindex Wno-ignored-qualifiers
2934 Warn if the return type of a function has a type qualifier
2935 such as @code{const}. For ISO C such a type qualifier has no effect,
2936 since the value returned by a function is not an lvalue.
2937 For C++, the warning is only emitted for scalar types or @code{void}.
2938 ISO C prohibits qualified @code{void} return types on function
2939 definitions, so such return types always receive a warning
2940 even without this option.
2942 This warning is also enabled by @option{-Wextra}.
2947 Warn if the type of @samp{main} is suspicious. @samp{main} should be a
2948 function with external linkage, returning int, taking either zero
2949 arguments, two, or three arguments of appropriate types.
2950 This warning is enabled by @option{-Wall}.
2952 @item -Wmissing-braces
2953 @opindex Wmissing-braces
2954 @opindex Wno-missing-braces
2955 Warn if an aggregate or union initializer is not fully bracketed. In
2956 the following example, the initializer for @samp{a} is not fully
2957 bracketed, but that for @samp{b} is fully bracketed.
2960 int a[2][2] = @{ 0, 1, 2, 3 @};
2961 int b[2][2] = @{ @{ 0, 1 @}, @{ 2, 3 @} @};
2964 This warning is enabled by @option{-Wall}.
2966 @item -Wmissing-include-dirs @r{(C, C++, Objective-C and Objective-C++ only)}
2967 @opindex Wmissing-include-dirs
2968 @opindex Wno-missing-include-dirs
2969 Warn if a user-supplied include directory does not exist.
2972 @opindex Wparentheses
2973 @opindex Wno-parentheses
2974 Warn if parentheses are omitted in certain contexts, such
2975 as when there is an assignment in a context where a truth value
2976 is expected, or when operators are nested whose precedence people
2977 often get confused about.
2979 Also warn if a comparison like @samp{x<=y<=z} appears; this is
2980 equivalent to @samp{(x<=y ? 1 : 0) <= z}, which is a different
2981 interpretation from that of ordinary mathematical notation.
2983 Also warn about constructions where there may be confusion to which
2984 @code{if} statement an @code{else} branch belongs. Here is an example of
2999 In C/C++, every @code{else} branch belongs to the innermost possible
3000 @code{if} statement, which in this example is @code{if (b)}. This is
3001 often not what the programmer expected, as illustrated in the above
3002 example by indentation the programmer chose. When there is the
3003 potential for this confusion, GCC will issue a warning when this flag
3004 is specified. To eliminate the warning, add explicit braces around
3005 the innermost @code{if} statement so there is no way the @code{else}
3006 could belong to the enclosing @code{if}. The resulting code would
3023 This warning is enabled by @option{-Wall}.
3025 @item -Wsequence-point
3026 @opindex Wsequence-point
3027 @opindex Wno-sequence-point
3028 Warn about code that may have undefined semantics because of violations
3029 of sequence point rules in the C and C++ standards.
3031 The C and C++ standards defines the order in which expressions in a C/C++
3032 program are evaluated in terms of @dfn{sequence points}, which represent
3033 a partial ordering between the execution of parts of the program: those
3034 executed before the sequence point, and those executed after it. These
3035 occur after the evaluation of a full expression (one which is not part
3036 of a larger expression), after the evaluation of the first operand of a
3037 @code{&&}, @code{||}, @code{? :} or @code{,} (comma) operator, before a
3038 function is called (but after the evaluation of its arguments and the
3039 expression denoting the called function), and in certain other places.
3040 Other than as expressed by the sequence point rules, the order of
3041 evaluation of subexpressions of an expression is not specified. All
3042 these rules describe only a partial order rather than a total order,
3043 since, for example, if two functions are called within one expression
3044 with no sequence point between them, the order in which the functions
3045 are called is not specified. However, the standards committee have
3046 ruled that function calls do not overlap.
3048 It is not specified when between sequence points modifications to the
3049 values of objects take effect. Programs whose behavior depends on this
3050 have undefined behavior; the C and C++ standards specify that ``Between
3051 the previous and next sequence point an object shall have its stored
3052 value modified at most once by the evaluation of an expression.
3053 Furthermore, the prior value shall be read only to determine the value
3054 to be stored.''. If a program breaks these rules, the results on any
3055 particular implementation are entirely unpredictable.
3057 Examples of code with undefined behavior are @code{a = a++;}, @code{a[n]
3058 = b[n++]} and @code{a[i++] = i;}. Some more complicated cases are not
3059 diagnosed by this option, and it may give an occasional false positive
3060 result, but in general it has been found fairly effective at detecting
3061 this sort of problem in programs.
3063 The standard is worded confusingly, therefore there is some debate
3064 over the precise meaning of the sequence point rules in subtle cases.
3065 Links to discussions of the problem, including proposed formal
3066 definitions, may be found on the GCC readings page, at
3067 @w{@uref{http://gcc.gnu.org/readings.html}}.
3069 This warning is enabled by @option{-Wall} for C and C++.
3072 @opindex Wreturn-type
3073 @opindex Wno-return-type
3074 Warn whenever a function is defined with a return-type that defaults
3075 to @code{int}. Also warn about any @code{return} statement with no
3076 return-value in a function whose return-type is not @code{void}
3077 (falling off the end of the function body is considered returning
3078 without a value), and about a @code{return} statement with a
3079 expression in a function whose return-type is @code{void}.
3081 For C++, a function without return type always produces a diagnostic
3082 message, even when @option{-Wno-return-type} is specified. The only
3083 exceptions are @samp{main} and functions defined in system headers.
3085 This warning is enabled by @option{-Wall}.
3090 Warn whenever a @code{switch} statement has an index of enumerated type
3091 and lacks a @code{case} for one or more of the named codes of that
3092 enumeration. (The presence of a @code{default} label prevents this
3093 warning.) @code{case} labels outside the enumeration range also
3094 provoke warnings when this option is used.
3095 This warning is enabled by @option{-Wall}.
3097 @item -Wswitch-default
3098 @opindex Wswitch-default
3099 @opindex Wno-switch-default
3100 Warn whenever a @code{switch} statement does not have a @code{default}
3104 @opindex Wswitch-enum
3105 @opindex Wno-switch-enum
3106 Warn whenever a @code{switch} statement has an index of enumerated type
3107 and lacks a @code{case} for one or more of the named codes of that
3108 enumeration. @code{case} labels outside the enumeration range also
3109 provoke warnings when this option is used.
3113 @opindex Wno-trigraphs
3114 Warn if any trigraphs are encountered that might change the meaning of
3115 the program (trigraphs within comments are not warned about).
3116 This warning is enabled by @option{-Wall}.
3118 @item -Wunused-function
3119 @opindex Wunused-function
3120 @opindex Wno-unused-function
3121 Warn whenever a static function is declared but not defined or a
3122 non-inline static function is unused.
3123 This warning is enabled by @option{-Wall}.
3125 @item -Wunused-label
3126 @opindex Wunused-label
3127 @opindex Wno-unused-label
3128 Warn whenever a label is declared but not used.
3129 This warning is enabled by @option{-Wall}.
3131 To suppress this warning use the @samp{unused} attribute
3132 (@pxref{Variable Attributes}).
3134 @item -Wunused-parameter
3135 @opindex Wunused-parameter
3136 @opindex Wno-unused-parameter
3137 Warn whenever a function parameter is unused aside from its declaration.
3139 To suppress this warning use the @samp{unused} attribute
3140 (@pxref{Variable Attributes}).
3142 @item -Wunused-variable
3143 @opindex Wunused-variable
3144 @opindex Wno-unused-variable
3145 Warn whenever a local variable or non-constant static variable is unused
3146 aside from its declaration.
3147 This warning is enabled by @option{-Wall}.
3149 To suppress this warning use the @samp{unused} attribute
3150 (@pxref{Variable Attributes}).
3152 @item -Wunused-value
3153 @opindex Wunused-value
3154 @opindex Wno-unused-value
3155 Warn whenever a statement computes a result that is explicitly not
3156 used. To suppress this warning cast the unused expression to
3157 @samp{void}. This includes an expression-statement or the left-hand
3158 side of a comma expression that contains no side effects. For example,
3159 an expression such as @samp{x[i,j]} will cause a warning, while
3160 @samp{x[(void)i,j]} will not.
3162 This warning is enabled by @option{-Wall}.
3167 All the above @option{-Wunused} options combined.
3169 In order to get a warning about an unused function parameter, you must
3170 either specify @samp{-Wextra -Wunused} (note that @samp{-Wall} implies
3171 @samp{-Wunused}), or separately specify @option{-Wunused-parameter}.
3173 @item -Wuninitialized
3174 @opindex Wuninitialized
3175 @opindex Wno-uninitialized
3176 Warn if an automatic variable is used without first being initialized or
3177 if a variable may be clobbered by a @code{setjmp} call.
3179 These warnings are possible only in optimizing compilation,
3180 because they require data flow information that is computed only
3181 when optimizing. If you do not specify @option{-O}, you will not get
3182 these warnings. Instead, GCC will issue a warning about @option{-Wuninitialized}
3183 requiring @option{-O}.
3185 If you want to warn about code which uses the uninitialized value of the
3186 variable in its own initializer, use the @option{-Winit-self} option.
3188 These warnings occur for individual uninitialized or clobbered
3189 elements of structure, union or array variables as well as for
3190 variables which are uninitialized or clobbered as a whole. They do
3191 not occur for variables or elements declared @code{volatile}. Because
3192 these warnings depend on optimization, the exact variables or elements
3193 for which there are warnings will depend on the precise optimization
3194 options and version of GCC used.
3196 Note that there may be no warning about a variable that is used only
3197 to compute a value that itself is never used, because such
3198 computations may be deleted by data flow analysis before the warnings
3201 These warnings are made optional because GCC is not smart
3202 enough to see all the reasons why the code might be correct
3203 despite appearing to have an error. Here is one example of how
3224 If the value of @code{y} is always 1, 2 or 3, then @code{x} is
3225 always initialized, but GCC doesn't know this. Here is
3226 another common case:
3231 if (change_y) save_y = y, y = new_y;
3233 if (change_y) y = save_y;
3238 This has no bug because @code{save_y} is used only if it is set.
3240 @cindex @code{longjmp} warnings
3241 This option also warns when a non-volatile automatic variable might be
3242 changed by a call to @code{longjmp}. These warnings as well are possible
3243 only in optimizing compilation.
3245 The compiler sees only the calls to @code{setjmp}. It cannot know
3246 where @code{longjmp} will be called; in fact, a signal handler could
3247 call it at any point in the code. As a result, you may get a warning
3248 even when there is in fact no problem because @code{longjmp} cannot
3249 in fact be called at the place which would cause a problem.
3251 Some spurious warnings can be avoided if you declare all the functions
3252 you use that never return as @code{noreturn}. @xref{Function
3255 This warning is enabled by @option{-Wall} or @option{-Wextra} in
3256 optimizing compilations (@option{-O1} and above).
3258 @item -Wunknown-pragmas
3259 @opindex Wunknown-pragmas
3260 @opindex Wno-unknown-pragmas
3261 @cindex warning for unknown pragmas
3262 @cindex unknown pragmas, warning
3263 @cindex pragmas, warning of unknown
3264 Warn when a #pragma directive is encountered which is not understood by
3265 GCC@. If this command line option is used, warnings will even be issued
3266 for unknown pragmas in system header files. This is not the case if
3267 the warnings were only enabled by the @option{-Wall} command line option.
3270 @opindex Wno-pragmas
3272 Do not warn about misuses of pragmas, such as incorrect parameters,
3273 invalid syntax, or conflicts between pragmas. See also
3274 @samp{-Wunknown-pragmas}.
3276 @item -Wstrict-aliasing
3277 @opindex Wstrict-aliasing
3278 @opindex Wno-strict-aliasing
3279 This option is only active when @option{-fstrict-aliasing} is active.
3280 It warns about code which might break the strict aliasing rules that the
3281 compiler is using for optimization. The warning does not catch all
3282 cases, but does attempt to catch the more common pitfalls. It is
3283 included in @option{-Wall}.
3284 It is equivalent to @option{-Wstrict-aliasing=3}
3286 @item -Wstrict-aliasing=n
3287 @opindex Wstrict-aliasing=n
3288 @opindex Wno-strict-aliasing=n
3289 This option is only active when @option{-fstrict-aliasing} is active.
3290 It warns about code which might break the strict aliasing rules that the
3291 compiler is using for optimization.
3292 Higher levels correspond to higher accuracy (fewer false positives).
3293 Higher levels also correspond to more effort, similar to the way -O works.
3294 @option{-Wstrict-aliasing} is equivalent to @option{-Wstrict-aliasing=n},
3297 Level 1: Most aggressive, quick, least accurate.
3298 Possibly useful when higher levels
3299 do not warn but -fstrict-aliasing still breaks the code, as it has very few
3300 false negatives. However, it has many false positives.
3301 Warns for all pointer conversions between possibly incompatible types,
3302 even if never dereferenced. Runs in the frontend only.
3304 Level 2: Aggressive, quick, not too precise.
3305 May still have many false positives (not as many as level 1 though),
3306 and few false negatives (but possibly more than level 1).
3307 Unlike level 1, it only warns when an address is taken. Warns about
3308 incomplete types. Runs in the frontend only.
3310 Level 3 (default for @option{-Wstrict-aliasing}):
3311 Should have very few false positives and few false
3312 negatives. Slightly slower than levels 1 or 2 when optimization is enabled.
3313 Takes care of the common punn+dereference pattern in the frontend:
3314 @code{*(int*)&some_float}.
3315 If optimization is enabled, it also runs in the backend, where it deals
3316 with multiple statement cases using flow-sensitive points-to information.
3317 Only warns when the converted pointer is dereferenced.
3318 Does not warn about incomplete types.
3320 @item -Wstrict-overflow
3321 @itemx -Wstrict-overflow=@var{n}
3322 @opindex Wstrict-overflow
3323 @opindex Wno-strict-overflow
3324 This option is only active when @option{-fstrict-overflow} is active.
3325 It warns about cases where the compiler optimizes based on the
3326 assumption that signed overflow does not occur. Note that it does not
3327 warn about all cases where the code might overflow: it only warns
3328 about cases where the compiler implements some optimization. Thus
3329 this warning depends on the optimization level.
3331 An optimization which assumes that signed overflow does not occur is
3332 perfectly safe if the values of the variables involved are such that
3333 overflow never does, in fact, occur. Therefore this warning can
3334 easily give a false positive: a warning about code which is not
3335 actually a problem. To help focus on important issues, several
3336 warning levels are defined. No warnings are issued for the use of
3337 undefined signed overflow when estimating how many iterations a loop
3338 will require, in particular when determining whether a loop will be
3342 @item -Wstrict-overflow=1
3343 Warn about cases which are both questionable and easy to avoid. For
3344 example: @code{x + 1 > x}; with @option{-fstrict-overflow}, the
3345 compiler will simplify this to @code{1}. This level of
3346 @option{-Wstrict-overflow} is enabled by @option{-Wall}; higher levels
3347 are not, and must be explicitly requested.
3349 @item -Wstrict-overflow=2
3350 Also warn about other cases where a comparison is simplified to a
3351 constant. For example: @code{abs (x) >= 0}. This can only be
3352 simplified when @option{-fstrict-overflow} is in effect, because
3353 @code{abs (INT_MIN)} overflows to @code{INT_MIN}, which is less than
3354 zero. @option{-Wstrict-overflow} (with no level) is the same as
3355 @option{-Wstrict-overflow=2}.
3357 @item -Wstrict-overflow=3
3358 Also warn about other cases where a comparison is simplified. For
3359 example: @code{x + 1 > 1} will be simplified to @code{x > 0}.
3361 @item -Wstrict-overflow=4
3362 Also warn about other simplifications not covered by the above cases.
3363 For example: @code{(x * 10) / 5} will be simplified to @code{x * 2}.
3365 @item -Wstrict-overflow=5
3366 Also warn about cases where the compiler reduces the magnitude of a
3367 constant involved in a comparison. For example: @code{x + 2 > y} will
3368 be simplified to @code{x + 1 >= y}. This is reported only at the
3369 highest warning level because this simplification applies to many
3370 comparisons, so this warning level will give a very large number of
3374 @item -Warray-bounds
3375 @opindex Wno-array-bounds
3376 @opindex Warray-bounds
3377 This option is only active when @option{-ftree-vrp} is active
3378 (default for -O2 and above). It warns about subscripts to arrays
3379 that are always out of bounds. This warning is enabled by @option{-Wall}.
3381 @item -Wno-div-by-zero
3382 @opindex Wno-div-by-zero
3383 @opindex Wdiv-by-zero
3384 Do not warn about compile-time integer division by zero. Floating point
3385 division by zero is not warned about, as it can be a legitimate way of
3386 obtaining infinities and NaNs.
3388 @item -Wsystem-headers
3389 @opindex Wsystem-headers
3390 @opindex Wno-system-headers
3391 @cindex warnings from system headers
3392 @cindex system headers, warnings from
3393 Print warning messages for constructs found in system header files.
3394 Warnings from system headers are normally suppressed, on the assumption
3395 that they usually do not indicate real problems and would only make the
3396 compiler output harder to read. Using this command line option tells
3397 GCC to emit warnings from system headers as if they occurred in user
3398 code. However, note that using @option{-Wall} in conjunction with this
3399 option will @emph{not} warn about unknown pragmas in system
3400 headers---for that, @option{-Wunknown-pragmas} must also be used.
3403 @opindex Wfloat-equal
3404 @opindex Wno-float-equal
3405 Warn if floating point values are used in equality comparisons.
3407 The idea behind this is that sometimes it is convenient (for the
3408 programmer) to consider floating-point values as approximations to
3409 infinitely precise real numbers. If you are doing this, then you need
3410 to compute (by analyzing the code, or in some other way) the maximum or
3411 likely maximum error that the computation introduces, and allow for it
3412 when performing comparisons (and when producing output, but that's a
3413 different problem). In particular, instead of testing for equality, you
3414 would check to see whether the two values have ranges that overlap; and
3415 this is done with the relational operators, so equality comparisons are
3418 @item -Wtraditional @r{(C and Objective-C only)}
3419 @opindex Wtraditional
3420 @opindex Wno-traditional
3421 Warn about certain constructs that behave differently in traditional and
3422 ISO C@. Also warn about ISO C constructs that have no traditional C
3423 equivalent, and/or problematic constructs which should be avoided.
3427 Macro parameters that appear within string literals in the macro body.
3428 In traditional C macro replacement takes place within string literals,
3429 but does not in ISO C@.
3432 In traditional C, some preprocessor directives did not exist.
3433 Traditional preprocessors would only consider a line to be a directive
3434 if the @samp{#} appeared in column 1 on the line. Therefore
3435 @option{-Wtraditional} warns about directives that traditional C
3436 understands but would ignore because the @samp{#} does not appear as the
3437 first character on the line. It also suggests you hide directives like
3438 @samp{#pragma} not understood by traditional C by indenting them. Some
3439 traditional implementations would not recognize @samp{#elif}, so it
3440 suggests avoiding it altogether.
3443 A function-like macro that appears without arguments.
3446 The unary plus operator.
3449 The @samp{U} integer constant suffix, or the @samp{F} or @samp{L} floating point
3450 constant suffixes. (Traditional C does support the @samp{L} suffix on integer
3451 constants.) Note, these suffixes appear in macros defined in the system
3452 headers of most modern systems, e.g.@: the @samp{_MIN}/@samp{_MAX} macros in @code{<limits.h>}.
3453 Use of these macros in user code might normally lead to spurious
3454 warnings, however GCC's integrated preprocessor has enough context to
3455 avoid warning in these cases.
3458 A function declared external in one block and then used after the end of
3462 A @code{switch} statement has an operand of type @code{long}.
3465 A non-@code{static} function declaration follows a @code{static} one.
3466 This construct is not accepted by some traditional C compilers.
3469 The ISO type of an integer constant has a different width or
3470 signedness from its traditional type. This warning is only issued if
3471 the base of the constant is ten. I.e.@: hexadecimal or octal values, which
3472 typically represent bit patterns, are not warned about.
3475 Usage of ISO string concatenation is detected.
3478 Initialization of automatic aggregates.
3481 Identifier conflicts with labels. Traditional C lacks a separate
3482 namespace for labels.
3485 Initialization of unions. If the initializer is zero, the warning is
3486 omitted. This is done under the assumption that the zero initializer in
3487 user code appears conditioned on e.g.@: @code{__STDC__} to avoid missing
3488 initializer warnings and relies on default initialization to zero in the
3492 Conversions by prototypes between fixed/floating point values and vice
3493 versa. The absence of these prototypes when compiling with traditional
3494 C would cause serious problems. This is a subset of the possible
3495 conversion warnings, for the full set use @option{-Wtraditional-conversion}.
3498 Use of ISO C style function definitions. This warning intentionally is
3499 @emph{not} issued for prototype declarations or variadic functions
3500 because these ISO C features will appear in your code when using
3501 libiberty's traditional C compatibility macros, @code{PARAMS} and
3502 @code{VPARAMS}. This warning is also bypassed for nested functions
3503 because that feature is already a GCC extension and thus not relevant to
3504 traditional C compatibility.
3507 @item -Wtraditional-conversion @r{(C and Objective-C only)}
3508 @opindex Wtraditional-conversion
3509 @opindex Wno-traditional-conversion
3510 Warn if a prototype causes a type conversion that is different from what
3511 would happen to the same argument in the absence of a prototype. This
3512 includes conversions of fixed point to floating and vice versa, and
3513 conversions changing the width or signedness of a fixed point argument
3514 except when the same as the default promotion.
3516 @item -Wdeclaration-after-statement @r{(C and Objective-C only)}
3517 @opindex Wdeclaration-after-statement
3518 @opindex Wno-declaration-after-statement
3519 Warn when a declaration is found after a statement in a block. This
3520 construct, known from C++, was introduced with ISO C99 and is by default
3521 allowed in GCC@. It is not supported by ISO C90 and was not supported by
3522 GCC versions before GCC 3.0. @xref{Mixed Declarations}.
3527 Warn if an undefined identifier is evaluated in an @samp{#if} directive.
3529 @item -Wno-endif-labels
3530 @opindex Wno-endif-labels
3531 @opindex Wendif-labels
3532 Do not warn whenever an @samp{#else} or an @samp{#endif} are followed by text.
3537 Warn whenever a local variable shadows another local variable, parameter or
3538 global variable or whenever a built-in function is shadowed.
3540 @item -Wlarger-than=@var{len}
3541 @opindex Wlarger-than=@var{len}
3542 @opindex Wlarger-than-@var{len}
3543 Warn whenever an object of larger than @var{len} bytes is defined.
3545 @item -Wframe-larger-than=@var{len}
3546 @opindex Wframe-larger-than
3547 Warn if the size of a function frame is larger than @var{len} bytes.
3548 The computation done to determine the stack frame size is approximate
3549 and not conservative.
3550 The actual requirements may be somewhat greater than @var{len}
3551 even if you do not get a warning. In addition, any space allocated
3552 via @code{alloca}, variable-length arrays, or related constructs
3553 is not included by the compiler when determining
3554 whether or not to issue a warning.
3556 @item -Wunsafe-loop-optimizations
3557 @opindex Wunsafe-loop-optimizations
3558 @opindex Wno-unsafe-loop-optimizations
3559 Warn if the loop cannot be optimized because the compiler could not
3560 assume anything on the bounds of the loop indices. With
3561 @option{-funsafe-loop-optimizations} warn if the compiler made
3564 @item -Wpointer-arith
3565 @opindex Wpointer-arith
3566 @opindex Wno-pointer-arith
3567 Warn about anything that depends on the ``size of'' a function type or
3568 of @code{void}. GNU C assigns these types a size of 1, for
3569 convenience in calculations with @code{void *} pointers and pointers
3570 to functions. In C++, warn also when an arithmetic operation involves
3571 @code{NULL}. This warning is also enabled by @option{-pedantic}.
3574 @opindex Wtype-limits
3575 @opindex Wno-type-limits
3576 Warn if a comparison is always true or always false due to the limited
3577 range of the data type, but do not warn for constant expressions. For
3578 example, warn if an unsigned variable is compared against zero with
3579 @samp{<} or @samp{>=}. This warning is also enabled by
3582 @item -Wbad-function-cast @r{(C and Objective-C only)}
3583 @opindex Wbad-function-cast
3584 @opindex Wno-bad-function-cast
3585 Warn whenever a function call is cast to a non-matching type.
3586 For example, warn if @code{int malloc()} is cast to @code{anything *}.
3588 @item -Wc++-compat @r{(C and Objective-C only)}
3589 Warn about ISO C constructs that are outside of the common subset of
3590 ISO C and ISO C++, e.g.@: request for implicit conversion from
3591 @code{void *} to a pointer to non-@code{void} type.
3593 @item -Wc++0x-compat @r{(C++ and Objective-C++ only)}
3594 Warn about C++ constructs whose meaning differs between ISO C++ 1998 and
3595 ISO C++ 200x, e.g., identifiers in ISO C++ 1998 that will become keywords
3596 in ISO C++ 200x. This warning is enabled by @option{-Wall}.
3600 @opindex Wno-cast-qual
3601 Warn whenever a pointer is cast so as to remove a type qualifier from
3602 the target type. For example, warn if a @code{const char *} is cast
3603 to an ordinary @code{char *}.
3606 @opindex Wcast-align
3607 @opindex Wno-cast-align
3608 Warn whenever a pointer is cast such that the required alignment of the
3609 target is increased. For example, warn if a @code{char *} is cast to
3610 an @code{int *} on machines where integers can only be accessed at
3611 two- or four-byte boundaries.
3613 @item -Wwrite-strings
3614 @opindex Wwrite-strings
3615 @opindex Wno-write-strings
3616 When compiling C, give string constants the type @code{const
3617 char[@var{length}]} so that
3618 copying the address of one into a non-@code{const} @code{char *}
3619 pointer will get a warning; when compiling C++, warn about the
3620 deprecated conversion from string literals to @code{char *}. This
3621 warning, by default, is enabled for C++ programs.
3622 These warnings will help you find at
3623 compile time code that can try to write into a string constant, but
3624 only if you have been very careful about using @code{const} in
3625 declarations and prototypes. Otherwise, it will just be a nuisance;
3626 this is why we did not make @option{-Wall} request these warnings.
3630 @opindex Wno-clobbered
3631 Warn for variables that might be changed by @samp{longjmp} or
3632 @samp{vfork}. This warning is also enabled by @option{-Wextra}.
3635 @opindex Wconversion
3636 @opindex Wno-conversion
3637 Warn for implicit conversions that may alter a value. This includes
3638 conversions between real and integer, like @code{abs (x)} when
3639 @code{x} is @code{double}; conversions between signed and unsigned,
3640 like @code{unsigned ui = -1}; and conversions to smaller types, like
3641 @code{sqrtf (M_PI)}. Do not warn for explicit casts like @code{abs
3642 ((int) x)} and @code{ui = (unsigned) -1}, or if the value is not
3643 changed by the conversion like in @code{abs (2.0)}. Warnings about
3644 conversions between signed and unsigned integers can be disabled by
3645 using @option{-Wno-sign-conversion}.
3647 For C++, also warn for conversions between @code{NULL} and non-pointer
3648 types; confusing overload resolution for user-defined conversions; and
3649 conversions that will never use a type conversion operator:
3650 conversions to @code{void}, the same type, a base class or a reference
3651 to them. Warnings about conversions between signed and unsigned
3652 integers are disabled by default in C++ unless
3653 @option{-Wsign-conversion} is explicitly enabled.
3656 @opindex Wempty-body
3657 @opindex Wno-empty-body
3658 Warn if an empty body occurs in an @samp{if}, @samp{else} or @samp{do
3659 while} statement. Additionally, in C++, warn when an empty body occurs
3660 in a @samp{while} or @samp{for} statement with no whitespacing before
3661 the semicolon. This warning is also enabled by @option{-Wextra}.
3663 @item -Wenum-compare @r{(C++ and Objective-C++ only)}
3664 @opindex Wenum-compare
3665 @opindex Wno-enum-compare
3666 Warn about a comparison between values of different enum types. This
3667 warning is enabled by default.
3669 @item -Wsign-compare
3670 @opindex Wsign-compare
3671 @opindex Wno-sign-compare
3672 @cindex warning for comparison of signed and unsigned values
3673 @cindex comparison of signed and unsigned values, warning
3674 @cindex signed and unsigned values, comparison warning
3675 Warn when a comparison between signed and unsigned values could produce
3676 an incorrect result when the signed value is converted to unsigned.
3677 This warning is also enabled by @option{-Wextra}; to get the other warnings
3678 of @option{-Wextra} without this warning, use @samp{-Wextra -Wno-sign-compare}.
3680 @item -Wsign-conversion
3681 @opindex Wsign-conversion
3682 @opindex Wno-sign-conversion
3683 Warn for implicit conversions that may change the sign of an integer
3684 value, like assigning a signed integer expression to an unsigned
3685 integer variable. An explicit cast silences the warning. In C, this
3686 option is enabled also by @option{-Wconversion}.
3690 @opindex Wno-address
3691 Warn about suspicious uses of memory addresses. These include using
3692 the address of a function in a conditional expression, such as
3693 @code{void func(void); if (func)}, and comparisons against the memory
3694 address of a string literal, such as @code{if (x == "abc")}. Such
3695 uses typically indicate a programmer error: the address of a function
3696 always evaluates to true, so their use in a conditional usually
3697 indicate that the programmer forgot the parentheses in a function
3698 call; and comparisons against string literals result in unspecified
3699 behavior and are not portable in C, so they usually indicate that the
3700 programmer intended to use @code{strcmp}. This warning is enabled by
3704 @opindex Wlogical-op
3705 @opindex Wno-logical-op
3706 Warn about suspicious uses of logical operators in expressions.
3707 This includes using logical operators in contexts where a
3708 bit-wise operator is likely to be expected.
3710 @item -Waggregate-return
3711 @opindex Waggregate-return
3712 @opindex Wno-aggregate-return
3713 Warn if any functions that return structures or unions are defined or
3714 called. (In languages where you can return an array, this also elicits
3717 @item -Wno-attributes
3718 @opindex Wno-attributes
3719 @opindex Wattributes
3720 Do not warn if an unexpected @code{__attribute__} is used, such as
3721 unrecognized attributes, function attributes applied to variables,
3722 etc. This will not stop errors for incorrect use of supported
3725 @item -Wstrict-prototypes @r{(C and Objective-C only)}
3726 @opindex Wstrict-prototypes
3727 @opindex Wno-strict-prototypes
3728 Warn if a function is declared or defined without specifying the
3729 argument types. (An old-style function definition is permitted without
3730 a warning if preceded by a declaration which specifies the argument
3733 @item -Wold-style-declaration @r{(C and Objective-C only)}
3734 @opindex Wold-style-declaration
3735 @opindex Wno-old-style-declaration
3736 Warn for obsolescent usages, according to the C Standard, in a
3737 declaration. For example, warn if storage-class specifiers like
3738 @code{static} are not the first things in a declaration. This warning
3739 is also enabled by @option{-Wextra}.
3741 @item -Wold-style-definition @r{(C and Objective-C only)}
3742 @opindex Wold-style-definition
3743 @opindex Wno-old-style-definition
3744 Warn if an old-style function definition is used. A warning is given
3745 even if there is a previous prototype.
3747 @item -Wmissing-parameter-type @r{(C and Objective-C only)}
3748 @opindex Wmissing-parameter-type
3749 @opindex Wno-missing-parameter-type
3750 A function parameter is declared without a type specifier in K&R-style
3757 This warning is also enabled by @option{-Wextra}.
3759 @item -Wmissing-prototypes @r{(C and Objective-C only)}
3760 @opindex Wmissing-prototypes
3761 @opindex Wno-missing-prototypes
3762 Warn if a global function is defined without a previous prototype
3763 declaration. This warning is issued even if the definition itself
3764 provides a prototype. The aim is to detect global functions that fail
3765 to be declared in header files.
3767 @item -Wmissing-declarations
3768 @opindex Wmissing-declarations
3769 @opindex Wno-missing-declarations
3770 Warn if a global function is defined without a previous declaration.
3771 Do so even if the definition itself provides a prototype.
3772 Use this option to detect global functions that are not declared in
3773 header files. In C++, no warnings are issued for function templates,
3774 or for inline functions, or for functions in anonymous namespaces.
3776 @item -Wmissing-field-initializers
3777 @opindex Wmissing-field-initializers
3778 @opindex Wno-missing-field-initializers
3782 Warn if a structure's initializer has some fields missing. For
3783 example, the following code would cause such a warning, because
3784 @code{x.h} is implicitly zero:
3787 struct s @{ int f, g, h; @};
3788 struct s x = @{ 3, 4 @};
3791 This option does not warn about designated initializers, so the following
3792 modification would not trigger a warning:
3795 struct s @{ int f, g, h; @};
3796 struct s x = @{ .f = 3, .g = 4 @};
3799 This warning is included in @option{-Wextra}. To get other @option{-Wextra}
3800 warnings without this one, use @samp{-Wextra -Wno-missing-field-initializers}.
3802 @item -Wmissing-noreturn
3803 @opindex Wmissing-noreturn
3804 @opindex Wno-missing-noreturn
3805 Warn about functions which might be candidates for attribute @code{noreturn}.
3806 Note these are only possible candidates, not absolute ones. Care should
3807 be taken to manually verify functions actually do not ever return before
3808 adding the @code{noreturn} attribute, otherwise subtle code generation
3809 bugs could be introduced. You will not get a warning for @code{main} in
3810 hosted C environments.
3812 @item -Wmissing-format-attribute
3813 @opindex Wmissing-format-attribute
3814 @opindex Wno-missing-format-attribute
3817 Warn about function pointers which might be candidates for @code{format}
3818 attributes. Note these are only possible candidates, not absolute ones.
3819 GCC will guess that function pointers with @code{format} attributes that
3820 are used in assignment, initialization, parameter passing or return
3821 statements should have a corresponding @code{format} attribute in the
3822 resulting type. I.e.@: the left-hand side of the assignment or
3823 initialization, the type of the parameter variable, or the return type
3824 of the containing function respectively should also have a @code{format}
3825 attribute to avoid the warning.
3827 GCC will also warn about function definitions which might be
3828 candidates for @code{format} attributes. Again, these are only
3829 possible candidates. GCC will guess that @code{format} attributes
3830 might be appropriate for any function that calls a function like
3831 @code{vprintf} or @code{vscanf}, but this might not always be the
3832 case, and some functions for which @code{format} attributes are
3833 appropriate may not be detected.
3835 @item -Wno-multichar
3836 @opindex Wno-multichar
3838 Do not warn if a multicharacter constant (@samp{'FOOF'}) is used.
3839 Usually they indicate a typo in the user's code, as they have
3840 implementation-defined values, and should not be used in portable code.
3842 @item -Wnormalized=<none|id|nfc|nfkc>
3843 @opindex Wnormalized=
3846 @cindex character set, input normalization
3847 In ISO C and ISO C++, two identifiers are different if they are
3848 different sequences of characters. However, sometimes when characters
3849 outside the basic ASCII character set are used, you can have two
3850 different character sequences that look the same. To avoid confusion,
3851 the ISO 10646 standard sets out some @dfn{normalization rules} which
3852 when applied ensure that two sequences that look the same are turned into
3853 the same sequence. GCC can warn you if you are using identifiers which
3854 have not been normalized; this option controls that warning.
3856 There are four levels of warning that GCC supports. The default is
3857 @option{-Wnormalized=nfc}, which warns about any identifier which is
3858 not in the ISO 10646 ``C'' normalized form, @dfn{NFC}. NFC is the
3859 recommended form for most uses.
3861 Unfortunately, there are some characters which ISO C and ISO C++ allow
3862 in identifiers that when turned into NFC aren't allowable as
3863 identifiers. That is, there's no way to use these symbols in portable
3864 ISO C or C++ and have all your identifiers in NFC@.
3865 @option{-Wnormalized=id} suppresses the warning for these characters.
3866 It is hoped that future versions of the standards involved will correct
3867 this, which is why this option is not the default.
3869 You can switch the warning off for all characters by writing
3870 @option{-Wnormalized=none}. You would only want to do this if you
3871 were using some other normalization scheme (like ``D''), because
3872 otherwise you can easily create bugs that are literally impossible to see.
3874 Some characters in ISO 10646 have distinct meanings but look identical
3875 in some fonts or display methodologies, especially once formatting has
3876 been applied. For instance @code{\u207F}, ``SUPERSCRIPT LATIN SMALL
3877 LETTER N'', will display just like a regular @code{n} which has been
3878 placed in a superscript. ISO 10646 defines the @dfn{NFKC}
3879 normalization scheme to convert all these into a standard form as
3880 well, and GCC will warn if your code is not in NFKC if you use
3881 @option{-Wnormalized=nfkc}. This warning is comparable to warning
3882 about every identifier that contains the letter O because it might be
3883 confused with the digit 0, and so is not the default, but may be
3884 useful as a local coding convention if the programming environment is
3885 unable to be fixed to display these characters distinctly.
3887 @item -Wno-deprecated
3888 @opindex Wno-deprecated
3889 @opindex Wdeprecated
3890 Do not warn about usage of deprecated features. @xref{Deprecated Features}.
3892 @item -Wno-deprecated-declarations
3893 @opindex Wno-deprecated-declarations
3894 @opindex Wdeprecated-declarations
3895 Do not warn about uses of functions (@pxref{Function Attributes}),
3896 variables (@pxref{Variable Attributes}), and types (@pxref{Type
3897 Attributes}) marked as deprecated by using the @code{deprecated}
3901 @opindex Wno-overflow
3903 Do not warn about compile-time overflow in constant expressions.
3905 @item -Woverride-init @r{(C and Objective-C only)}
3906 @opindex Woverride-init
3907 @opindex Wno-override-init
3911 Warn if an initialized field without side effects is overridden when
3912 using designated initializers (@pxref{Designated Inits, , Designated
3915 This warning is included in @option{-Wextra}. To get other
3916 @option{-Wextra} warnings without this one, use @samp{-Wextra
3917 -Wno-override-init}.
3922 Warn if a structure is given the packed attribute, but the packed
3923 attribute has no effect on the layout or size of the structure.
3924 Such structures may be mis-aligned for little benefit. For
3925 instance, in this code, the variable @code{f.x} in @code{struct bar}
3926 will be misaligned even though @code{struct bar} does not itself
3927 have the packed attribute:
3934 @} __attribute__((packed));
3945 Warn if padding is included in a structure, either to align an element
3946 of the structure or to align the whole structure. Sometimes when this
3947 happens it is possible to rearrange the fields of the structure to
3948 reduce the padding and so make the structure smaller.
3950 @item -Wredundant-decls
3951 @opindex Wredundant-decls
3952 @opindex Wno-redundant-decls
3953 Warn if anything is declared more than once in the same scope, even in
3954 cases where multiple declaration is valid and changes nothing.
3956 @item -Wnested-externs @r{(C and Objective-C only)}
3957 @opindex Wnested-externs
3958 @opindex Wno-nested-externs
3959 Warn if an @code{extern} declaration is encountered within a function.
3961 @item -Wunreachable-code
3962 @opindex Wunreachable-code
3963 @opindex Wno-unreachable-code
3964 Warn if the compiler detects that code will never be executed.
3966 This option is intended to warn when the compiler detects that at
3967 least a whole line of source code will never be executed, because
3968 some condition is never satisfied or because it is after a
3969 procedure that never returns.
3971 It is possible for this option to produce a warning even though there
3972 are circumstances under which part of the affected line can be executed,
3973 so care should be taken when removing apparently-unreachable code.
3975 For instance, when a function is inlined, a warning may mean that the
3976 line is unreachable in only one inlined copy of the function.
3978 This option is not made part of @option{-Wall} because in a debugging
3979 version of a program there is often substantial code which checks
3980 correct functioning of the program and is, hopefully, unreachable
3981 because the program does work. Another common use of unreachable
3982 code is to provide behavior which is selectable at compile-time.
3987 Warn if a function can not be inlined and it was declared as inline.
3988 Even with this option, the compiler will not warn about failures to
3989 inline functions declared in system headers.
3991 The compiler uses a variety of heuristics to determine whether or not
3992 to inline a function. For example, the compiler takes into account
3993 the size of the function being inlined and the amount of inlining
3994 that has already been done in the current function. Therefore,
3995 seemingly insignificant changes in the source program can cause the
3996 warnings produced by @option{-Winline} to appear or disappear.
3998 @item -Wno-invalid-offsetof @r{(C++ and Objective-C++ only)}
3999 @opindex Wno-invalid-offsetof
4000 @opindex Winvalid-offsetof
4001 Suppress warnings from applying the @samp{offsetof} macro to a non-POD
4002 type. According to the 1998 ISO C++ standard, applying @samp{offsetof}
4003 to a non-POD type is undefined. In existing C++ implementations,
4004 however, @samp{offsetof} typically gives meaningful results even when
4005 applied to certain kinds of non-POD types. (Such as a simple
4006 @samp{struct} that fails to be a POD type only by virtue of having a
4007 constructor.) This flag is for users who are aware that they are
4008 writing nonportable code and who have deliberately chosen to ignore the
4011 The restrictions on @samp{offsetof} may be relaxed in a future version
4012 of the C++ standard.
4014 @item -Wno-int-to-pointer-cast @r{(C and Objective-C only)}
4015 @opindex Wno-int-to-pointer-cast
4016 @opindex Wint-to-pointer-cast
4017 Suppress warnings from casts to pointer type of an integer of a
4020 @item -Wno-pointer-to-int-cast @r{(C and Objective-C only)}
4021 @opindex Wno-pointer-to-int-cast
4022 @opindex Wpointer-to-int-cast
4023 Suppress warnings from casts from a pointer to an integer type of a
4027 @opindex Winvalid-pch
4028 @opindex Wno-invalid-pch
4029 Warn if a precompiled header (@pxref{Precompiled Headers}) is found in
4030 the search path but can't be used.
4034 @opindex Wno-long-long
4035 Warn if @samp{long long} type is used. This is default. To inhibit
4036 the warning messages, use @option{-Wno-long-long}. Flags
4037 @option{-Wlong-long} and @option{-Wno-long-long} are taken into account
4038 only when @option{-pedantic} flag is used.
4040 @item -Wvariadic-macros
4041 @opindex Wvariadic-macros
4042 @opindex Wno-variadic-macros
4043 Warn if variadic macros are used in pedantic ISO C90 mode, or the GNU
4044 alternate syntax when in pedantic ISO C99 mode. This is default.
4045 To inhibit the warning messages, use @option{-Wno-variadic-macros}.
4050 Warn if variable length array is used in the code.
4051 @option{-Wno-vla} will prevent the @option{-pedantic} warning of
4052 the variable length array.
4054 @item -Wvolatile-register-var
4055 @opindex Wvolatile-register-var
4056 @opindex Wno-volatile-register-var
4057 Warn if a register variable is declared volatile. The volatile
4058 modifier does not inhibit all optimizations that may eliminate reads
4059 and/or writes to register variables. This warning is enabled by
4062 @item -Wdisabled-optimization
4063 @opindex Wdisabled-optimization
4064 @opindex Wno-disabled-optimization
4065 Warn if a requested optimization pass is disabled. This warning does
4066 not generally indicate that there is anything wrong with your code; it
4067 merely indicates that GCC's optimizers were unable to handle the code
4068 effectively. Often, the problem is that your code is too big or too
4069 complex; GCC will refuse to optimize programs when the optimization
4070 itself is likely to take inordinate amounts of time.
4072 @item -Wpointer-sign @r{(C and Objective-C only)}
4073 @opindex Wpointer-sign
4074 @opindex Wno-pointer-sign
4075 Warn for pointer argument passing or assignment with different signedness.
4076 This option is only supported for C and Objective-C@. It is implied by
4077 @option{-Wall} and by @option{-pedantic}, which can be disabled with
4078 @option{-Wno-pointer-sign}.
4080 @item -Wstack-protector
4081 @opindex Wstack-protector
4082 @opindex Wno-stack-protector
4083 This option is only active when @option{-fstack-protector} is active. It
4084 warns about functions that will not be protected against stack smashing.
4087 @opindex Wno-mudflap
4088 Suppress warnings about constructs that cannot be instrumented by
4091 @item -Woverlength-strings
4092 @opindex Woverlength-strings
4093 @opindex Wno-overlength-strings
4094 Warn about string constants which are longer than the ``minimum
4095 maximum'' length specified in the C standard. Modern compilers
4096 generally allow string constants which are much longer than the
4097 standard's minimum limit, but very portable programs should avoid
4098 using longer strings.
4100 The limit applies @emph{after} string constant concatenation, and does
4101 not count the trailing NUL@. In C89, the limit was 509 characters; in
4102 C99, it was raised to 4095. C++98 does not specify a normative
4103 minimum maximum, so we do not diagnose overlength strings in C++@.
4105 This option is implied by @option{-pedantic}, and can be disabled with
4106 @option{-Wno-overlength-strings}.
4108 @item -Wdisallowed-function-list=@var{sym},@var{sym},@dots{}
4109 @opindex Wdisallowed-function-list
4111 If any of @var{sym} is called, GCC will issue a warning. This can be useful
4112 in enforcing coding conventions that ban calls to certain functions, for
4113 example, @code{alloca}, @code{malloc}, etc.
4116 @node Debugging Options
4117 @section Options for Debugging Your Program or GCC
4118 @cindex options, debugging
4119 @cindex debugging information options
4121 GCC has various special options that are used for debugging
4122 either your program or GCC:
4127 Produce debugging information in the operating system's native format
4128 (stabs, COFF, XCOFF, or DWARF 2)@. GDB can work with this debugging
4131 On most systems that use stabs format, @option{-g} enables use of extra
4132 debugging information that only GDB can use; this extra information
4133 makes debugging work better in GDB but will probably make other debuggers
4135 refuse to read the program. If you want to control for certain whether
4136 to generate the extra information, use @option{-gstabs+}, @option{-gstabs},
4137 @option{-gxcoff+}, @option{-gxcoff}, or @option{-gvms} (see below).
4139 GCC allows you to use @option{-g} with
4140 @option{-O}. The shortcuts taken by optimized code may occasionally
4141 produce surprising results: some variables you declared may not exist
4142 at all; flow of control may briefly move where you did not expect it;
4143 some statements may not be executed because they compute constant
4144 results or their values were already at hand; some statements may
4145 execute in different places because they were moved out of loops.
4147 Nevertheless it proves possible to debug optimized output. This makes
4148 it reasonable to use the optimizer for programs that might have bugs.
4150 The following options are useful when GCC is generated with the
4151 capability for more than one debugging format.
4155 Produce debugging information for use by GDB@. This means to use the
4156 most expressive format available (DWARF 2, stabs, or the native format
4157 if neither of those are supported), including GDB extensions if at all
4162 Produce debugging information in stabs format (if that is supported),
4163 without GDB extensions. This is the format used by DBX on most BSD
4164 systems. On MIPS, Alpha and System V Release 4 systems this option
4165 produces stabs debugging output which is not understood by DBX or SDB@.
4166 On System V Release 4 systems this option requires the GNU assembler.
4168 @item -feliminate-unused-debug-symbols
4169 @opindex feliminate-unused-debug-symbols
4170 Produce debugging information in stabs format (if that is supported),
4171 for only symbols that are actually used.
4173 @item -femit-class-debug-always
4174 Instead of emitting debugging information for a C++ class in only one
4175 object file, emit it in all object files using the class. This option
4176 should be used only with debuggers that are unable to handle the way GCC
4177 normally emits debugging information for classes because using this
4178 option will increase the size of debugging information by as much as a
4183 Produce debugging information in stabs format (if that is supported),
4184 using GNU extensions understood only by the GNU debugger (GDB)@. The
4185 use of these extensions is likely to make other debuggers crash or
4186 refuse to read the program.
4190 Produce debugging information in COFF format (if that is supported).
4191 This is the format used by SDB on most System V systems prior to
4196 Produce debugging information in XCOFF format (if that is supported).
4197 This is the format used by the DBX debugger on IBM RS/6000 systems.
4201 Produce debugging information in XCOFF format (if that is supported),
4202 using GNU extensions understood only by the GNU debugger (GDB)@. The
4203 use of these extensions is likely to make other debuggers crash or
4204 refuse to read the program, and may cause assemblers other than the GNU
4205 assembler (GAS) to fail with an error.
4209 Produce debugging information in DWARF version 2 format (if that is
4210 supported). This is the format used by DBX on IRIX 6. With this
4211 option, GCC uses features of DWARF version 3 when they are useful;
4212 version 3 is upward compatible with version 2, but may still cause
4213 problems for older debuggers.
4217 Produce debugging information in VMS debug format (if that is
4218 supported). This is the format used by DEBUG on VMS systems.
4221 @itemx -ggdb@var{level}
4222 @itemx -gstabs@var{level}
4223 @itemx -gcoff@var{level}
4224 @itemx -gxcoff@var{level}
4225 @itemx -gvms@var{level}
4226 Request debugging information and also use @var{level} to specify how
4227 much information. The default level is 2.
4229 Level 0 produces no debug information at all. Thus, @option{-g0} negates
4232 Level 1 produces minimal information, enough for making backtraces in
4233 parts of the program that you don't plan to debug. This includes
4234 descriptions of functions and external variables, but no information
4235 about local variables and no line numbers.
4237 Level 3 includes extra information, such as all the macro definitions
4238 present in the program. Some debuggers support macro expansion when
4239 you use @option{-g3}.
4241 @option{-gdwarf-2} does not accept a concatenated debug level, because
4242 GCC used to support an option @option{-gdwarf} that meant to generate
4243 debug information in version 1 of the DWARF format (which is very
4244 different from version 2), and it would have been too confusing. That
4245 debug format is long obsolete, but the option cannot be changed now.
4246 Instead use an additional @option{-g@var{level}} option to change the
4247 debug level for DWARF2.
4249 @item -feliminate-dwarf2-dups
4250 @opindex feliminate-dwarf2-dups
4251 Compress DWARF2 debugging information by eliminating duplicated
4252 information about each symbol. This option only makes sense when
4253 generating DWARF2 debugging information with @option{-gdwarf-2}.
4255 @item -femit-struct-debug-baseonly
4256 Emit debug information for struct-like types
4257 only when the base name of the compilation source file
4258 matches the base name of file in which the struct was defined.
4260 This option substantially reduces the size of debugging information,
4261 but at significant potential loss in type information to the debugger.
4262 See @option{-femit-struct-debug-reduced} for a less aggressive option.
4263 See @option{-femit-struct-debug-detailed} for more detailed control.
4265 This option works only with DWARF 2.
4267 @item -femit-struct-debug-reduced
4268 Emit debug information for struct-like types
4269 only when the base name of the compilation source file
4270 matches the base name of file in which the type was defined,
4271 unless the struct is a template or defined in a system header.
4273 This option significantly reduces the size of debugging information,
4274 with some potential loss in type information to the debugger.
4275 See @option{-femit-struct-debug-baseonly} for a more aggressive option.
4276 See @option{-femit-struct-debug-detailed} for more detailed control.
4278 This option works only with DWARF 2.
4280 @item -femit-struct-debug-detailed@r{[}=@var{spec-list}@r{]}
4281 Specify the struct-like types
4282 for which the compiler will generate debug information.
4283 The intent is to reduce duplicate struct debug information
4284 between different object files within the same program.
4286 This option is a detailed version of
4287 @option{-femit-struct-debug-reduced} and @option{-femit-struct-debug-baseonly},
4288 which will serve for most needs.
4290 A specification has the syntax
4291 [@samp{dir:}|@samp{ind:}][@samp{ord:}|@samp{gen:}](@samp{any}|@samp{sys}|@samp{base}|@samp{none})
4293 The optional first word limits the specification to
4294 structs that are used directly (@samp{dir:}) or used indirectly (@samp{ind:}).
4295 A struct type is used directly when it is the type of a variable, member.
4296 Indirect uses arise through pointers to structs.
4297 That is, when use of an incomplete struct would be legal, the use is indirect.
4299 @samp{struct one direct; struct two * indirect;}.
4301 The optional second word limits the specification to
4302 ordinary structs (@samp{ord:}) or generic structs (@samp{gen:}).
4303 Generic structs are a bit complicated to explain.
4304 For C++, these are non-explicit specializations of template classes,
4305 or non-template classes within the above.
4306 Other programming languages have generics,
4307 but @samp{-femit-struct-debug-detailed} does not yet implement them.
4309 The third word specifies the source files for those
4310 structs for which the compiler will emit debug information.
4311 The values @samp{none} and @samp{any} have the normal meaning.
4312 The value @samp{base} means that
4313 the base of name of the file in which the type declaration appears
4314 must match the base of the name of the main compilation file.
4315 In practice, this means that
4316 types declared in @file{foo.c} and @file{foo.h} will have debug information,
4317 but types declared in other header will not.
4318 The value @samp{sys} means those types satisfying @samp{base}
4319 or declared in system or compiler headers.
4321 You may need to experiment to determine the best settings for your application.
4323 The default is @samp{-femit-struct-debug-detailed=all}.
4325 This option works only with DWARF 2.
4327 @item -fno-merge-debug-strings
4328 @opindex fmerge-debug-strings
4329 @opindex fno-merge-debug-strings
4330 Direct the linker to merge together strings which are identical in
4331 different object files. This is not supported by all assemblers or
4332 linker. This decreases the size of the debug information in the
4333 output file at the cost of increasing link processing time. This is
4336 @item -fdebug-prefix-map=@var{old}=@var{new}
4337 @opindex fdebug-prefix-map
4338 When compiling files in directory @file{@var{old}}, record debugging
4339 information describing them as in @file{@var{new}} instead.
4341 @cindex @command{prof}
4344 Generate extra code to write profile information suitable for the
4345 analysis program @command{prof}. You must use this option when compiling
4346 the source files you want data about, and you must also use it when
4349 @cindex @command{gprof}
4352 Generate extra code to write profile information suitable for the
4353 analysis program @command{gprof}. You must use this option when compiling
4354 the source files you want data about, and you must also use it when
4359 Makes the compiler print out each function name as it is compiled, and
4360 print some statistics about each pass when it finishes.
4363 @opindex ftime-report
4364 Makes the compiler print some statistics about the time consumed by each
4365 pass when it finishes.
4368 @opindex fmem-report
4369 Makes the compiler print some statistics about permanent memory
4370 allocation when it finishes.
4372 @item -fpre-ipa-mem-report
4373 @opindex fpre-ipa-mem-report
4374 @item -fpost-ipa-mem-report
4375 @opindex fpost-ipa-mem-report
4376 Makes the compiler print some statistics about permanent memory
4377 allocation before or after interprocedural optimization.
4379 @item -fprofile-arcs
4380 @opindex fprofile-arcs
4381 Add code so that program flow @dfn{arcs} are instrumented. During
4382 execution the program records how many times each branch and call is
4383 executed and how many times it is taken or returns. When the compiled
4384 program exits it saves this data to a file called
4385 @file{@var{auxname}.gcda} for each source file. The data may be used for
4386 profile-directed optimizations (@option{-fbranch-probabilities}), or for
4387 test coverage analysis (@option{-ftest-coverage}). Each object file's
4388 @var{auxname} is generated from the name of the output file, if
4389 explicitly specified and it is not the final executable, otherwise it is
4390 the basename of the source file. In both cases any suffix is removed
4391 (e.g.@: @file{foo.gcda} for input file @file{dir/foo.c}, or
4392 @file{dir/foo.gcda} for output file specified as @option{-o dir/foo.o}).
4393 @xref{Cross-profiling}.
4395 @cindex @command{gcov}
4399 This option is used to compile and link code instrumented for coverage
4400 analysis. The option is a synonym for @option{-fprofile-arcs}
4401 @option{-ftest-coverage} (when compiling) and @option{-lgcov} (when
4402 linking). See the documentation for those options for more details.
4407 Compile the source files with @option{-fprofile-arcs} plus optimization
4408 and code generation options. For test coverage analysis, use the
4409 additional @option{-ftest-coverage} option. You do not need to profile
4410 every source file in a program.
4413 Link your object files with @option{-lgcov} or @option{-fprofile-arcs}
4414 (the latter implies the former).
4417 Run the program on a representative workload to generate the arc profile
4418 information. This may be repeated any number of times. You can run
4419 concurrent instances of your program, and provided that the file system
4420 supports locking, the data files will be correctly updated. Also
4421 @code{fork} calls are detected and correctly handled (double counting
4425 For profile-directed optimizations, compile the source files again with
4426 the same optimization and code generation options plus
4427 @option{-fbranch-probabilities} (@pxref{Optimize Options,,Options that
4428 Control Optimization}).
4431 For test coverage analysis, use @command{gcov} to produce human readable
4432 information from the @file{.gcno} and @file{.gcda} files. Refer to the
4433 @command{gcov} documentation for further information.
4437 With @option{-fprofile-arcs}, for each function of your program GCC
4438 creates a program flow graph, then finds a spanning tree for the graph.
4439 Only arcs that are not on the spanning tree have to be instrumented: the
4440 compiler adds code to count the number of times that these arcs are
4441 executed. When an arc is the only exit or only entrance to a block, the
4442 instrumentation code can be added to the block; otherwise, a new basic
4443 block must be created to hold the instrumentation code.
4446 @item -ftest-coverage
4447 @opindex ftest-coverage
4448 Produce a notes file that the @command{gcov} code-coverage utility
4449 (@pxref{Gcov,, @command{gcov}---a Test Coverage Program}) can use to
4450 show program coverage. Each source file's note file is called
4451 @file{@var{auxname}.gcno}. Refer to the @option{-fprofile-arcs} option
4452 above for a description of @var{auxname} and instructions on how to
4453 generate test coverage data. Coverage data will match the source files
4454 more closely, if you do not optimize.
4456 @item -fdbg-cnt-list
4457 @opindex fdbg-cnt-list
4458 Print the name and the counter upperbound for all debug counters.
4460 @item -fdbg-cnt=@var{counter-value-list}
4462 Set the internal debug counter upperbound. @var{counter-value-list}
4463 is a comma-separated list of @var{name}:@var{value} pairs
4464 which sets the upperbound of each debug counter @var{name} to @var{value}.
4465 All debug counters have the initial upperbound of @var{UINT_MAX},
4466 thus dbg_cnt() returns true always unless the upperbound is set by this option.
4467 e.g. With -fdbg-cnt=dce:10,tail_call:0
4468 dbg_cnt(dce) will return true only for first 10 invocations
4469 and dbg_cnt(tail_call) will return false always.
4471 @item -d@var{letters}
4472 @itemx -fdump-rtl-@var{pass}
4474 Says to make debugging dumps during compilation at times specified by
4475 @var{letters}. This is used for debugging the RTL-based passes of the
4476 compiler. The file names for most of the dumps are made by appending a
4477 pass number and a word to the @var{dumpname}. @var{dumpname} is generated
4478 from the name of the output file, if explicitly specified and it is not
4479 an executable, otherwise it is the basename of the source file. These
4480 switches may have different effects when @option{-E} is used for
4483 Most debug dumps can be enabled either passing a letter to the @option{-d}
4484 option, or with a long @option{-fdump-rtl} switch; here are the possible
4485 letters for use in @var{letters} and @var{pass}, and their meanings:
4490 Annotate the assembler output with miscellaneous debugging information.
4492 @item -fdump-rtl-bbro
4493 @opindex fdump-rtl-bbro
4494 Dump after block reordering, to @file{@var{file}.148r.bbro}.
4496 @item -fdump-rtl-combine
4497 @opindex fdump-rtl-combine
4498 Dump after the RTL instruction combination pass, to the file
4499 @file{@var{file}.129r.combine}.
4501 @item -fdump-rtl-ce1
4502 @itemx -fdump-rtl-ce2
4503 @opindex fdump-rtl-ce1
4504 @opindex fdump-rtl-ce2
4505 @option{-fdump-rtl-ce1} enable dumping after the
4506 first if conversion, to the file @file{@var{file}.117r.ce1}.
4507 @option{-fdump-rtl-ce2} enable dumping after the second if
4508 conversion, to the file @file{@var{file}.130r.ce2}.
4510 @item -fdump-rtl-btl
4511 @itemx -fdump-rtl-dbr
4512 @opindex fdump-rtl-btl
4513 @opindex fdump-rtl-dbr
4514 @option{-fdump-rtl-btl} enable dumping after branch
4515 target load optimization, to @file{@var{file}.31.btl}.
4516 @option{-fdump-rtl-dbr} enable dumping after delayed branch
4517 scheduling, to @file{@var{file}.36.dbr}.
4521 Dump all macro definitions, at the end of preprocessing, in addition to
4524 @item -fdump-rtl-ce3
4525 @opindex fdump-rtl-ce3
4526 Dump after the third if conversion, to @file{@var{file}.146r.ce3}.
4528 @item -fdump-rtl-cfg
4529 @itemx -fdump-rtl-life
4530 @opindex fdump-rtl-cfg
4531 @opindex fdump-rtl-life
4532 @option{-fdump-rtl-cfg} enable dumping after control
4533 and data flow analysis, to @file{@var{file}.116r.cfg}.
4534 @option{-fdump-rtl-cfg} enable dumping dump after life analysis,
4535 to @file{@var{file}.128r.life1} and @file{@var{file}.135r.life2}.
4537 @item -fdump-rtl-greg
4538 @opindex fdump-rtl-greg
4539 Dump after global register allocation, to @file{@var{file}.139r.greg}.
4541 @item -fdump-rtl-gcse
4542 @itemx -fdump-rtl-bypass
4543 @opindex fdump-rtl-gcse
4544 @opindex fdump-rtl-bypass
4545 @option{-fdump-rtl-gcse} enable dumping after GCSE, to
4546 @file{@var{file}.114r.gcse}. @option{-fdump-rtl-bypass}
4547 enable dumping after jump bypassing and control flow optimizations, to
4548 @file{@var{file}.115r.bypass}.
4551 @opindex fdump-rtl-eh
4552 Dump after finalization of EH handling code, to @file{@var{file}.02.eh}.
4554 @item -fdump-rtl-sibling
4555 @opindex fdump-rtl-sibling
4556 Dump after sibling call optimizations, to @file{@var{file}.106r.sibling}.
4558 @item -fdump-rtl-jump
4559 @opindex fdump-rtl-jump
4560 Dump after the first jump optimization, to @file{@var{file}.112r.jump}.
4562 @item -fdump-rtl-stack
4563 @opindex fdump-rtl-stack
4564 Dump after conversion from GCC's "flat register file" registers to the
4565 x87's stack-like registers, to @file{@var{file}.152r.stack}.
4567 @item -fdump-rtl-lreg
4568 @opindex fdump-rtl-lreg
4569 Dump after local register allocation, to @file{@var{file}.138r.lreg}.
4571 @item -fdump-rtl-loop2
4572 @opindex fdump-rtl-loop2
4573 @option{-dL} and @option{-fdump-rtl-loop2} enable dumping after the
4574 loop optimization pass, to @file{@var{file}.119r.loop2},
4575 @file{@var{file}.120r.loop2_init},
4576 @file{@var{file}.121r.loop2_invariant}, and
4577 @file{@var{file}.125r.loop2_done}.
4579 @item -fdump-rtl-sms
4580 @opindex fdump-rtl-sms
4581 Dump after modulo scheduling, to @file{@var{file}.136r.sms}.
4583 @item -fdump-rtl-mach
4584 @opindex fdump-rtl-mach
4585 Dump after performing the machine dependent reorganization pass, to
4586 @file{@var{file}.155r.mach} if that pass exists.
4588 @item -fdump-rtl-rnreg
4589 @opindex fdump-rtl-rnreg
4590 Dump after register renumbering, to @file{@var{file}.147r.rnreg}.
4592 @item -fdump-rtl-regmove
4593 @opindex fdump-rtl-regmove
4594 Dump after the register move pass, to @file{@var{file}.132r.regmove}.
4596 @item -fdump-rtl-postreload
4597 @opindex fdump-rtl-postreload
4598 Dump after post-reload optimizations, to @file{@var{file}.24.postreload}.
4600 @item -fdump-rtl-expand
4601 @opindex fdump-rtl-expand
4602 Dump after RTL generation, to @file{@var{file}.104r.expand}.
4604 @item -fdump-rtl-sched2
4605 @opindex fdump-rtl-sched2
4606 Dump after the second scheduling pass, to @file{@var{file}.149r.sched2}.
4608 @item -fdump-rtl-cse
4609 @opindex fdump-rtl-cse
4610 Dump after CSE (including the jump optimization that sometimes follows
4611 CSE), to @file{@var{file}.113r.cse}.
4613 @item -fdump-rtl-sched1
4614 @opindex fdump-rtl-sched1
4615 Dump after the first scheduling pass, to @file{@var{file}.136r.sched1}.
4617 @item -fdump-rtl-cse2
4618 @opindex fdump-rtl-cse2
4619 Dump after the second CSE pass (including the jump optimization that
4620 sometimes follows CSE), to @file{@var{file}.127r.cse2}.
4622 @item -fdump-rtl-tracer
4623 @opindex fdump-rtl-tracer
4624 Dump after running tracer, to @file{@var{file}.118r.tracer}.
4626 @item -fdump-rtl-vpt
4627 @itemx -fdump-rtl-vartrack
4628 @opindex fdump-rtl-vpt
4629 @opindex fdump-rtl-vartrack
4630 @option{-fdump-rtl-vpt} enable dumping after the value
4631 profile transformations, to @file{@var{file}.10.vpt}.
4632 @option{-fdump-rtl-vartrack} enable dumping after variable tracking,
4633 to @file{@var{file}.154r.vartrack}.
4635 @item -fdump-rtl-flow2
4636 @opindex fdump-rtl-flow2
4637 Dump after the second flow pass, to @file{@var{file}.142r.flow2}.
4639 @item -fdump-rtl-peephole2
4640 @opindex fdump-rtl-peephole2
4641 Dump after the peephole pass, to @file{@var{file}.145r.peephole2}.
4643 @item -fdump-rtl-web
4644 @opindex fdump-rtl-web
4645 Dump after live range splitting, to @file{@var{file}.126r.web}.
4647 @item -fdump-rtl-all
4648 @opindex fdump-rtl-all
4649 Produce all the dumps listed above.
4653 Produce a core dump whenever an error occurs.
4657 Print statistics on memory usage, at the end of the run, to
4662 Annotate the assembler output with a comment indicating which
4663 pattern and alternative was used. The length of each instruction is
4668 Dump the RTL in the assembler output as a comment before each instruction.
4669 Also turns on @option{-dp} annotation.
4673 For each of the other indicated dump files (either with @option{-d} or
4674 @option{-fdump-rtl-@var{pass}}), dump a representation of the control flow
4675 graph suitable for viewing with VCG to @file{@var{file}.@var{pass}.vcg}.
4679 Just generate RTL for a function instead of compiling it. Usually used
4680 with @samp{r} (@option{-fdump-rtl-expand}).
4684 Dump debugging information during parsing, to standard error.
4688 @opindex fdump-noaddr
4689 When doing debugging dumps (see @option{-d} option above), suppress
4690 address output. This makes it more feasible to use diff on debugging
4691 dumps for compiler invocations with different compiler binaries and/or
4692 different text / bss / data / heap / stack / dso start locations.
4694 @item -fdump-unnumbered
4695 @opindex fdump-unnumbered
4696 When doing debugging dumps (see @option{-d} option above), suppress instruction
4697 numbers and address output. This makes it more feasible to
4698 use diff on debugging dumps for compiler invocations with different
4699 options, in particular with and without @option{-g}.
4701 @item -fdump-translation-unit @r{(C++ only)}
4702 @itemx -fdump-translation-unit-@var{options} @r{(C++ only)}
4703 @opindex fdump-translation-unit
4704 Dump a representation of the tree structure for the entire translation
4705 unit to a file. The file name is made by appending @file{.tu} to the
4706 source file name. If the @samp{-@var{options}} form is used, @var{options}
4707 controls the details of the dump as described for the
4708 @option{-fdump-tree} options.
4710 @item -fdump-class-hierarchy @r{(C++ only)}
4711 @itemx -fdump-class-hierarchy-@var{options} @r{(C++ only)}
4712 @opindex fdump-class-hierarchy
4713 Dump a representation of each class's hierarchy and virtual function
4714 table layout to a file. The file name is made by appending @file{.class}
4715 to the source file name. If the @samp{-@var{options}} form is used,
4716 @var{options} controls the details of the dump as described for the
4717 @option{-fdump-tree} options.
4719 @item -fdump-ipa-@var{switch}
4721 Control the dumping at various stages of inter-procedural analysis
4722 language tree to a file. The file name is generated by appending a switch
4723 specific suffix to the source file name. The following dumps are possible:
4727 Enables all inter-procedural analysis dumps.
4730 Dumps information about call-graph optimization, unused function removal,
4731 and inlining decisions.
4734 Dump after function inlining.
4738 @item -fdump-statistics-@var{option}
4739 @opindex -fdump-statistics
4740 Enable and control dumping of pass statistics in a separate file. The
4741 file name is generated by appending a suffix ending in @samp{.statistics}
4742 to the source file name. If the @samp{-@var{option}} form is used,
4743 @samp{-stats} will cause counters to be summed over the whole compilation unit
4744 while @samp{-details} will dump every event as the passes generate them.
4745 The default with no option is to sum counters for each function compiled.
4747 @item -fdump-tree-@var{switch}
4748 @itemx -fdump-tree-@var{switch}-@var{options}
4750 Control the dumping at various stages of processing the intermediate
4751 language tree to a file. The file name is generated by appending a switch
4752 specific suffix to the source file name. If the @samp{-@var{options}}
4753 form is used, @var{options} is a list of @samp{-} separated options that
4754 control the details of the dump. Not all options are applicable to all
4755 dumps, those which are not meaningful will be ignored. The following
4756 options are available
4760 Print the address of each node. Usually this is not meaningful as it
4761 changes according to the environment and source file. Its primary use
4762 is for tying up a dump file with a debug environment.
4764 Inhibit dumping of members of a scope or body of a function merely
4765 because that scope has been reached. Only dump such items when they
4766 are directly reachable by some other path. When dumping pretty-printed
4767 trees, this option inhibits dumping the bodies of control structures.
4769 Print a raw representation of the tree. By default, trees are
4770 pretty-printed into a C-like representation.
4772 Enable more detailed dumps (not honored by every dump option).
4774 Enable dumping various statistics about the pass (not honored by every dump
4777 Enable showing basic block boundaries (disabled in raw dumps).
4779 Enable showing virtual operands for every statement.
4781 Enable showing line numbers for statements.
4783 Enable showing the unique ID (@code{DECL_UID}) for each variable.
4785 Enable showing the tree dump for each statement.
4787 Turn on all options, except @option{raw}, @option{slim}, @option{verbose}
4788 and @option{lineno}.
4791 The following tree dumps are possible:
4795 Dump before any tree based optimization, to @file{@var{file}.original}.
4798 Dump after all tree based optimization, to @file{@var{file}.optimized}.
4801 @opindex fdump-tree-gimple
4802 Dump each function before and after the gimplification pass to a file. The
4803 file name is made by appending @file{.gimple} to the source file name.
4806 @opindex fdump-tree-cfg
4807 Dump the control flow graph of each function to a file. The file name is
4808 made by appending @file{.cfg} to the source file name.
4811 @opindex fdump-tree-vcg
4812 Dump the control flow graph of each function to a file in VCG format. The
4813 file name is made by appending @file{.vcg} to the source file name. Note
4814 that if the file contains more than one function, the generated file cannot
4815 be used directly by VCG@. You will need to cut and paste each function's
4816 graph into its own separate file first.
4819 @opindex fdump-tree-ch
4820 Dump each function after copying loop headers. The file name is made by
4821 appending @file{.ch} to the source file name.
4824 @opindex fdump-tree-ssa
4825 Dump SSA related information to a file. The file name is made by appending
4826 @file{.ssa} to the source file name.
4829 @opindex fdump-tree-alias
4830 Dump aliasing information for each function. The file name is made by
4831 appending @file{.alias} to the source file name.
4834 @opindex fdump-tree-ccp
4835 Dump each function after CCP@. The file name is made by appending
4836 @file{.ccp} to the source file name.
4839 @opindex fdump-tree-storeccp
4840 Dump each function after STORE-CCP@. The file name is made by appending
4841 @file{.storeccp} to the source file name.
4844 @opindex fdump-tree-pre
4845 Dump trees after partial redundancy elimination. The file name is made
4846 by appending @file{.pre} to the source file name.
4849 @opindex fdump-tree-fre
4850 Dump trees after full redundancy elimination. The file name is made
4851 by appending @file{.fre} to the source file name.
4854 @opindex fdump-tree-copyprop
4855 Dump trees after copy propagation. The file name is made
4856 by appending @file{.copyprop} to the source file name.
4858 @item store_copyprop
4859 @opindex fdump-tree-store_copyprop
4860 Dump trees after store copy-propagation. The file name is made
4861 by appending @file{.store_copyprop} to the source file name.
4864 @opindex fdump-tree-dce
4865 Dump each function after dead code elimination. The file name is made by
4866 appending @file{.dce} to the source file name.
4869 @opindex fdump-tree-mudflap
4870 Dump each function after adding mudflap instrumentation. The file name is
4871 made by appending @file{.mudflap} to the source file name.
4874 @opindex fdump-tree-sra
4875 Dump each function after performing scalar replacement of aggregates. The
4876 file name is made by appending @file{.sra} to the source file name.
4879 @opindex fdump-tree-sink
4880 Dump each function after performing code sinking. The file name is made
4881 by appending @file{.sink} to the source file name.
4884 @opindex fdump-tree-dom
4885 Dump each function after applying dominator tree optimizations. The file
4886 name is made by appending @file{.dom} to the source file name.
4889 @opindex fdump-tree-dse
4890 Dump each function after applying dead store elimination. The file
4891 name is made by appending @file{.dse} to the source file name.
4894 @opindex fdump-tree-phiopt
4895 Dump each function after optimizing PHI nodes into straightline code. The file
4896 name is made by appending @file{.phiopt} to the source file name.
4899 @opindex fdump-tree-forwprop
4900 Dump each function after forward propagating single use variables. The file
4901 name is made by appending @file{.forwprop} to the source file name.
4904 @opindex fdump-tree-copyrename
4905 Dump each function after applying the copy rename optimization. The file
4906 name is made by appending @file{.copyrename} to the source file name.
4909 @opindex fdump-tree-nrv
4910 Dump each function after applying the named return value optimization on
4911 generic trees. The file name is made by appending @file{.nrv} to the source
4915 @opindex fdump-tree-vect
4916 Dump each function after applying vectorization of loops. The file name is
4917 made by appending @file{.vect} to the source file name.
4920 @opindex fdump-tree-vrp
4921 Dump each function after Value Range Propagation (VRP). The file name
4922 is made by appending @file{.vrp} to the source file name.
4925 @opindex fdump-tree-all
4926 Enable all the available tree dumps with the flags provided in this option.
4929 @item -ftree-vectorizer-verbose=@var{n}
4930 @opindex ftree-vectorizer-verbose
4931 This option controls the amount of debugging output the vectorizer prints.
4932 This information is written to standard error, unless
4933 @option{-fdump-tree-all} or @option{-fdump-tree-vect} is specified,
4934 in which case it is output to the usual dump listing file, @file{.vect}.
4935 For @var{n}=0 no diagnostic information is reported.
4936 If @var{n}=1 the vectorizer reports each loop that got vectorized,
4937 and the total number of loops that got vectorized.
4938 If @var{n}=2 the vectorizer also reports non-vectorized loops that passed
4939 the first analysis phase (vect_analyze_loop_form) - i.e.@: countable,
4940 inner-most, single-bb, single-entry/exit loops. This is the same verbosity
4941 level that @option{-fdump-tree-vect-stats} uses.
4942 Higher verbosity levels mean either more information dumped for each
4943 reported loop, or same amount of information reported for more loops:
4944 If @var{n}=3, alignment related information is added to the reports.
4945 If @var{n}=4, data-references related information (e.g.@: memory dependences,
4946 memory access-patterns) is added to the reports.
4947 If @var{n}=5, the vectorizer reports also non-vectorized inner-most loops
4948 that did not pass the first analysis phase (i.e., may not be countable, or
4949 may have complicated control-flow).
4950 If @var{n}=6, the vectorizer reports also non-vectorized nested loops.
4951 For @var{n}=7, all the information the vectorizer generates during its
4952 analysis and transformation is reported. This is the same verbosity level
4953 that @option{-fdump-tree-vect-details} uses.
4955 @item -frandom-seed=@var{string}
4956 @opindex frandom-string
4957 This option provides a seed that GCC uses when it would otherwise use
4958 random numbers. It is used to generate certain symbol names
4959 that have to be different in every compiled file. It is also used to
4960 place unique stamps in coverage data files and the object files that
4961 produce them. You can use the @option{-frandom-seed} option to produce
4962 reproducibly identical object files.
4964 The @var{string} should be different for every file you compile.
4966 @item -fsched-verbose=@var{n}
4967 @opindex fsched-verbose
4968 On targets that use instruction scheduling, this option controls the
4969 amount of debugging output the scheduler prints. This information is
4970 written to standard error, unless @option{-dS} or @option{-dR} is
4971 specified, in which case it is output to the usual dump
4972 listing file, @file{.sched} or @file{.sched2} respectively. However
4973 for @var{n} greater than nine, the output is always printed to standard
4976 For @var{n} greater than zero, @option{-fsched-verbose} outputs the
4977 same information as @option{-dRS}. For @var{n} greater than one, it
4978 also output basic block probabilities, detailed ready list information
4979 and unit/insn info. For @var{n} greater than two, it includes RTL
4980 at abort point, control-flow and regions info. And for @var{n} over
4981 four, @option{-fsched-verbose} also includes dependence info.
4985 Store the usual ``temporary'' intermediate files permanently; place them
4986 in the current directory and name them based on the source file. Thus,
4987 compiling @file{foo.c} with @samp{-c -save-temps} would produce files
4988 @file{foo.i} and @file{foo.s}, as well as @file{foo.o}. This creates a
4989 preprocessed @file{foo.i} output file even though the compiler now
4990 normally uses an integrated preprocessor.
4992 When used in combination with the @option{-x} command line option,
4993 @option{-save-temps} is sensible enough to avoid over writing an
4994 input source file with the same extension as an intermediate file.
4995 The corresponding intermediate file may be obtained by renaming the
4996 source file before using @option{-save-temps}.
5000 Report the CPU time taken by each subprocess in the compilation
5001 sequence. For C source files, this is the compiler proper and assembler
5002 (plus the linker if linking is done). The output looks like this:
5009 The first number on each line is the ``user time'', that is time spent
5010 executing the program itself. The second number is ``system time'',
5011 time spent executing operating system routines on behalf of the program.
5012 Both numbers are in seconds.
5014 @item -fvar-tracking
5015 @opindex fvar-tracking
5016 Run variable tracking pass. It computes where variables are stored at each
5017 position in code. Better debugging information is then generated
5018 (if the debugging information format supports this information).
5020 It is enabled by default when compiling with optimization (@option{-Os},
5021 @option{-O}, @option{-O2}, @dots{}), debugging information (@option{-g}) and
5022 the debug info format supports it.
5024 @item -print-file-name=@var{library}
5025 @opindex print-file-name
5026 Print the full absolute name of the library file @var{library} that
5027 would be used when linking---and don't do anything else. With this
5028 option, GCC does not compile or link anything; it just prints the
5031 @item -print-multi-directory
5032 @opindex print-multi-directory
5033 Print the directory name corresponding to the multilib selected by any
5034 other switches present in the command line. This directory is supposed
5035 to exist in @env{GCC_EXEC_PREFIX}.
5037 @item -print-multi-lib
5038 @opindex print-multi-lib
5039 Print the mapping from multilib directory names to compiler switches
5040 that enable them. The directory name is separated from the switches by
5041 @samp{;}, and each switch starts with an @samp{@@} instead of the
5042 @samp{-}, without spaces between multiple switches. This is supposed to
5043 ease shell-processing.
5045 @item -print-prog-name=@var{program}
5046 @opindex print-prog-name
5047 Like @option{-print-file-name}, but searches for a program such as @samp{cpp}.
5049 @item -print-libgcc-file-name
5050 @opindex print-libgcc-file-name
5051 Same as @option{-print-file-name=libgcc.a}.
5053 This is useful when you use @option{-nostdlib} or @option{-nodefaultlibs}
5054 but you do want to link with @file{libgcc.a}. You can do
5057 gcc -nostdlib @var{files}@dots{} `gcc -print-libgcc-file-name`
5060 @item -print-search-dirs
5061 @opindex print-search-dirs
5062 Print the name of the configured installation directory and a list of
5063 program and library directories @command{gcc} will search---and don't do anything else.
5065 This is useful when @command{gcc} prints the error message
5066 @samp{installation problem, cannot exec cpp0: No such file or directory}.
5067 To resolve this you either need to put @file{cpp0} and the other compiler
5068 components where @command{gcc} expects to find them, or you can set the environment
5069 variable @env{GCC_EXEC_PREFIX} to the directory where you installed them.
5070 Don't forget the trailing @samp{/}.
5071 @xref{Environment Variables}.
5073 @item -print-sysroot
5074 @opindex print-sysroot
5075 Print the target sysroot directory that will be used during
5076 compilation. This is the target sysroot specified either at configure
5077 time or or using the @option{--sysroot} option, possibly with an extra
5078 suffix that depends on compilation options. If no target sysroot is
5079 specified, the option prints nothing.
5081 @item -print-sysroot-headers-suffix
5082 @opindex print-sysroot-headers-suffix
5083 Print the suffix added to the target sysroot when searching for
5084 headers, or give an error if the compiler is not configured with such
5085 a suffix---and don't do anything else.
5088 @opindex dumpmachine
5089 Print the compiler's target machine (for example,
5090 @samp{i686-pc-linux-gnu})---and don't do anything else.
5093 @opindex dumpversion
5094 Print the compiler version (for example, @samp{3.0})---and don't do
5099 Print the compiler's built-in specs---and don't do anything else. (This
5100 is used when GCC itself is being built.) @xref{Spec Files}.
5102 @item -feliminate-unused-debug-types
5103 @opindex feliminate-unused-debug-types
5104 Normally, when producing DWARF2 output, GCC will emit debugging
5105 information for all types declared in a compilation
5106 unit, regardless of whether or not they are actually used
5107 in that compilation unit. Sometimes this is useful, such as
5108 if, in the debugger, you want to cast a value to a type that is
5109 not actually used in your program (but is declared). More often,
5110 however, this results in a significant amount of wasted space.
5111 With this option, GCC will avoid producing debug symbol output
5112 for types that are nowhere used in the source file being compiled.
5115 @node Optimize Options
5116 @section Options That Control Optimization
5117 @cindex optimize options
5118 @cindex options, optimization
5120 These options control various sorts of optimizations.
5122 Without any optimization option, the compiler's goal is to reduce the
5123 cost of compilation and to make debugging produce the expected
5124 results. Statements are independent: if you stop the program with a
5125 breakpoint between statements, you can then assign a new value to any
5126 variable or change the program counter to any other statement in the
5127 function and get exactly the results you would expect from the source
5130 Turning on optimization flags makes the compiler attempt to improve
5131 the performance and/or code size at the expense of compilation time
5132 and possibly the ability to debug the program.
5134 The compiler performs optimization based on the knowledge it has of the
5135 program. Compiling multiple files at once to a single output file mode allows
5136 the compiler to use information gained from all of the files when compiling
5139 Not all optimizations are controlled directly by a flag. Only
5140 optimizations that have a flag are listed.
5147 Optimize. Optimizing compilation takes somewhat more time, and a lot
5148 more memory for a large function.
5150 With @option{-O}, the compiler tries to reduce code size and execution
5151 time, without performing any optimizations that take a great deal of
5154 @option{-O} turns on the following optimization flags:
5157 -fcprop-registers @gol
5160 -fdelayed-branch @gol
5162 -fguess-branch-probability @gol
5163 -fif-conversion2 @gol
5164 -fif-conversion @gol
5165 -finline-small-functions @gol
5166 -fipa-pure-const @gol
5167 -fipa-reference @gol
5169 -fsplit-wide-types @gol
5170 -ftree-builtin-call-dce @gol
5173 -ftree-copyrename @gol
5175 -ftree-dominator-opts @gol
5182 @option{-O} also turns on @option{-fomit-frame-pointer} on machines
5183 where doing so does not interfere with debugging.
5187 Optimize even more. GCC performs nearly all supported optimizations
5188 that do not involve a space-speed tradeoff. The compiler does not
5189 perform loop unrolling or function inlining when you specify @option{-O2}.
5190 As compared to @option{-O}, this option increases both compilation time
5191 and the performance of the generated code.
5193 @option{-O2} turns on all optimization flags specified by @option{-O}. It
5194 also turns on the following optimization flags:
5195 @gccoptlist{-fthread-jumps @gol
5196 -falign-functions -falign-jumps @gol
5197 -falign-loops -falign-labels @gol
5200 -fcse-follow-jumps -fcse-skip-blocks @gol
5201 -fdelete-null-pointer-checks @gol
5202 -fexpensive-optimizations @gol
5203 -fgcse -fgcse-lm @gol
5204 -findirect-inlining @gol
5205 -foptimize-sibling-calls @gol
5208 -freorder-blocks -freorder-functions @gol
5209 -frerun-cse-after-loop @gol
5210 -fsched-interblock -fsched-spec @gol
5211 -fschedule-insns -fschedule-insns2 @gol
5212 -fstrict-aliasing -fstrict-overflow @gol
5213 -ftree-switch-conversion @gol
5217 Please note the warning under @option{-fgcse} about
5218 invoking @option{-O2} on programs that use computed gotos.
5222 Optimize yet more. @option{-O3} turns on all optimizations specified
5223 by @option{-O2} and also turns on the @option{-finline-functions},
5224 @option{-funswitch-loops}, @option{-fpredictive-commoning},
5225 @option{-fgcse-after-reload} and @option{-ftree-vectorize} options.
5229 Reduce compilation time and make debugging produce the expected
5230 results. This is the default.
5234 Optimize for size. @option{-Os} enables all @option{-O2} optimizations that
5235 do not typically increase code size. It also performs further
5236 optimizations designed to reduce code size.
5238 @option{-Os} disables the following optimization flags:
5239 @gccoptlist{-falign-functions -falign-jumps -falign-loops @gol
5240 -falign-labels -freorder-blocks -freorder-blocks-and-partition @gol
5241 -fprefetch-loop-arrays -ftree-vect-loop-version}
5243 If you use multiple @option{-O} options, with or without level numbers,
5244 the last such option is the one that is effective.
5247 Options of the form @option{-f@var{flag}} specify machine-independent
5248 flags. Most flags have both positive and negative forms; the negative
5249 form of @option{-ffoo} would be @option{-fno-foo}. In the table
5250 below, only one of the forms is listed---the one you typically will
5251 use. You can figure out the other form by either removing @samp{no-}
5254 The following options control specific optimizations. They are either
5255 activated by @option{-O} options or are related to ones that are. You
5256 can use the following flags in the rare cases when ``fine-tuning'' of
5257 optimizations to be performed is desired.
5260 @item -fno-default-inline
5261 @opindex fno-default-inline
5262 Do not make member functions inline by default merely because they are
5263 defined inside the class scope (C++ only). Otherwise, when you specify
5264 @w{@option{-O}}, member functions defined inside class scope are compiled
5265 inline by default; i.e., you don't need to add @samp{inline} in front of
5266 the member function name.
5268 @item -fno-defer-pop
5269 @opindex fno-defer-pop
5270 Always pop the arguments to each function call as soon as that function
5271 returns. For machines which must pop arguments after a function call,
5272 the compiler normally lets arguments accumulate on the stack for several
5273 function calls and pops them all at once.
5275 Disabled at levels @option{-O}, @option{-O2}, @option{-O3}, @option{-Os}.
5277 @item -fforward-propagate
5278 @opindex fforward-propagate
5279 Perform a forward propagation pass on RTL@. The pass tries to combine two
5280 instructions and checks if the result can be simplified. If loop unrolling
5281 is active, two passes are performed and the second is scheduled after
5284 This option is enabled by default at optimization levels @option{-O2},
5285 @option{-O3}, @option{-Os}.
5287 @item -fomit-frame-pointer
5288 @opindex fomit-frame-pointer
5289 Don't keep the frame pointer in a register for functions that
5290 don't need one. This avoids the instructions to save, set up and
5291 restore frame pointers; it also makes an extra register available
5292 in many functions. @strong{It also makes debugging impossible on
5295 On some machines, such as the VAX, this flag has no effect, because
5296 the standard calling sequence automatically handles the frame pointer
5297 and nothing is saved by pretending it doesn't exist. The
5298 machine-description macro @code{FRAME_POINTER_REQUIRED} controls
5299 whether a target machine supports this flag. @xref{Registers,,Register
5300 Usage, gccint, GNU Compiler Collection (GCC) Internals}.
5302 Enabled at levels @option{-O}, @option{-O2}, @option{-O3}, @option{-Os}.
5304 @item -foptimize-sibling-calls
5305 @opindex foptimize-sibling-calls
5306 Optimize sibling and tail recursive calls.
5308 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
5312 Don't pay attention to the @code{inline} keyword. Normally this option
5313 is used to keep the compiler from expanding any functions inline.
5314 Note that if you are not optimizing, no functions can be expanded inline.
5316 @item -finline-small-functions
5317 @opindex finline-small-functions
5318 Integrate functions into their callers when their body is smaller than expected
5319 function call code (so overall size of program gets smaller). The compiler
5320 heuristically decides which functions are simple enough to be worth integrating
5323 Enabled at level @option{-O2}.
5325 @item -findirect-inlining
5326 @opindex findirect-inlining
5327 Inline also indirect calls that are discovered to be known at compile
5328 time thanks to previous inlining. This option has any effect only
5329 when inlining itself is turned on by the @option{-finline-functions}
5330 or @option{-finline-small-functions} options.
5332 Enabled at level @option{-O2}.
5334 @item -finline-functions
5335 @opindex finline-functions
5336 Integrate all simple functions into their callers. The compiler
5337 heuristically decides which functions are simple enough to be worth
5338 integrating in this way.
5340 If all calls to a given function are integrated, and the function is
5341 declared @code{static}, then the function is normally not output as
5342 assembler code in its own right.
5344 Enabled at level @option{-O3}.
5346 @item -finline-functions-called-once
5347 @opindex finline-functions-called-once
5348 Consider all @code{static} functions called once for inlining into their
5349 caller even if they are not marked @code{inline}. If a call to a given
5350 function is integrated, then the function is not output as assembler code
5353 Enabled at levels @option{-O1}, @option{-O2}, @option{-O3} and @option{-Os}.
5355 @item -fearly-inlining
5356 @opindex fearly-inlining
5357 Inline functions marked by @code{always_inline} and functions whose body seems
5358 smaller than the function call overhead early before doing
5359 @option{-fprofile-generate} instrumentation and real inlining pass. Doing so
5360 makes profiling significantly cheaper and usually inlining faster on programs
5361 having large chains of nested wrapper functions.
5365 @item -finline-limit=@var{n}
5366 @opindex finline-limit
5367 By default, GCC limits the size of functions that can be inlined. This flag
5368 allows coarse control of this limit. @var{n} is the size of functions that
5369 can be inlined in number of pseudo instructions.
5371 Inlining is actually controlled by a number of parameters, which may be
5372 specified individually by using @option{--param @var{name}=@var{value}}.
5373 The @option{-finline-limit=@var{n}} option sets some of these parameters
5377 @item max-inline-insns-single
5378 is set to @var{n}/2.
5379 @item max-inline-insns-auto
5380 is set to @var{n}/2.
5383 See below for a documentation of the individual
5384 parameters controlling inlining and for the defaults of these parameters.
5386 @emph{Note:} there may be no value to @option{-finline-limit} that results
5387 in default behavior.
5389 @emph{Note:} pseudo instruction represents, in this particular context, an
5390 abstract measurement of function's size. In no way does it represent a count
5391 of assembly instructions and as such its exact meaning might change from one
5392 release to an another.
5394 @item -fkeep-inline-functions
5395 @opindex fkeep-inline-functions
5396 In C, emit @code{static} functions that are declared @code{inline}
5397 into the object file, even if the function has been inlined into all
5398 of its callers. This switch does not affect functions using the
5399 @code{extern inline} extension in GNU C89@. In C++, emit any and all
5400 inline functions into the object file.
5402 @item -fkeep-static-consts
5403 @opindex fkeep-static-consts
5404 Emit variables declared @code{static const} when optimization isn't turned
5405 on, even if the variables aren't referenced.
5407 GCC enables this option by default. If you want to force the compiler to
5408 check if the variable was referenced, regardless of whether or not
5409 optimization is turned on, use the @option{-fno-keep-static-consts} option.
5411 @item -fmerge-constants
5412 @opindex fmerge-constants
5413 Attempt to merge identical constants (string constants and floating point
5414 constants) across compilation units.
5416 This option is the default for optimized compilation if the assembler and
5417 linker support it. Use @option{-fno-merge-constants} to inhibit this
5420 Enabled at levels @option{-O}, @option{-O2}, @option{-O3}, @option{-Os}.
5422 @item -fmerge-all-constants
5423 @opindex fmerge-all-constants
5424 Attempt to merge identical constants and identical variables.
5426 This option implies @option{-fmerge-constants}. In addition to
5427 @option{-fmerge-constants} this considers e.g.@: even constant initialized
5428 arrays or initialized constant variables with integral or floating point
5429 types. Languages like C or C++ require each non-automatic variable to
5430 have distinct location, so using this option will result in non-conforming
5433 @item -fmodulo-sched
5434 @opindex fmodulo-sched
5435 Perform swing modulo scheduling immediately before the first scheduling
5436 pass. This pass looks at innermost loops and reorders their
5437 instructions by overlapping different iterations.
5439 @item -fmodulo-sched-allow-regmoves
5440 @opindex fmodulo-sched-allow-regmoves
5441 Perform more aggressive SMS based modulo scheduling with register moves
5442 allowed. By setting this flag certain anti-dependences edges will be
5443 deleted which will trigger the generation of reg-moves based on the
5444 life-range analysis. This option is effective only with
5445 @option{-fmodulo-sched} enabled.
5447 @item -fno-branch-count-reg
5448 @opindex fno-branch-count-reg
5449 Do not use ``decrement and branch'' instructions on a count register,
5450 but instead generate a sequence of instructions that decrement a
5451 register, compare it against zero, then branch based upon the result.
5452 This option is only meaningful on architectures that support such
5453 instructions, which include x86, PowerPC, IA-64 and S/390.
5455 The default is @option{-fbranch-count-reg}.
5457 @item -fno-function-cse
5458 @opindex fno-function-cse
5459 Do not put function addresses in registers; make each instruction that
5460 calls a constant function contain the function's address explicitly.
5462 This option results in less efficient code, but some strange hacks
5463 that alter the assembler output may be confused by the optimizations
5464 performed when this option is not used.
5466 The default is @option{-ffunction-cse}
5468 @item -fno-zero-initialized-in-bss
5469 @opindex fno-zero-initialized-in-bss
5470 If the target supports a BSS section, GCC by default puts variables that
5471 are initialized to zero into BSS@. This can save space in the resulting
5474 This option turns off this behavior because some programs explicitly
5475 rely on variables going to the data section. E.g., so that the
5476 resulting executable can find the beginning of that section and/or make
5477 assumptions based on that.
5479 The default is @option{-fzero-initialized-in-bss}.
5481 @item -fmudflap -fmudflapth -fmudflapir
5485 @cindex bounds checking
5487 For front-ends that support it (C and C++), instrument all risky
5488 pointer/array dereferencing operations, some standard library
5489 string/heap functions, and some other associated constructs with
5490 range/validity tests. Modules so instrumented should be immune to
5491 buffer overflows, invalid heap use, and some other classes of C/C++
5492 programming errors. The instrumentation relies on a separate runtime
5493 library (@file{libmudflap}), which will be linked into a program if
5494 @option{-fmudflap} is given at link time. Run-time behavior of the
5495 instrumented program is controlled by the @env{MUDFLAP_OPTIONS}
5496 environment variable. See @code{env MUDFLAP_OPTIONS=-help a.out}
5499 Use @option{-fmudflapth} instead of @option{-fmudflap} to compile and to
5500 link if your program is multi-threaded. Use @option{-fmudflapir}, in
5501 addition to @option{-fmudflap} or @option{-fmudflapth}, if
5502 instrumentation should ignore pointer reads. This produces less
5503 instrumentation (and therefore faster execution) and still provides
5504 some protection against outright memory corrupting writes, but allows
5505 erroneously read data to propagate within a program.
5507 @item -fthread-jumps
5508 @opindex fthread-jumps
5509 Perform optimizations where we check to see if a jump branches to a
5510 location where another comparison subsumed by the first is found. If
5511 so, the first branch is redirected to either the destination of the
5512 second branch or a point immediately following it, depending on whether
5513 the condition is known to be true or false.
5515 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
5517 @item -fsplit-wide-types
5518 @opindex fsplit-wide-types
5519 When using a type that occupies multiple registers, such as @code{long
5520 long} on a 32-bit system, split the registers apart and allocate them
5521 independently. This normally generates better code for those types,
5522 but may make debugging more difficult.
5524 Enabled at levels @option{-O}, @option{-O2}, @option{-O3},
5527 @item -fcse-follow-jumps
5528 @opindex fcse-follow-jumps
5529 In common subexpression elimination (CSE), scan through jump instructions
5530 when the target of the jump is not reached by any other path. For
5531 example, when CSE encounters an @code{if} statement with an
5532 @code{else} clause, CSE will follow the jump when the condition
5535 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
5537 @item -fcse-skip-blocks
5538 @opindex fcse-skip-blocks
5539 This is similar to @option{-fcse-follow-jumps}, but causes CSE to
5540 follow jumps which conditionally skip over blocks. When CSE
5541 encounters a simple @code{if} statement with no else clause,
5542 @option{-fcse-skip-blocks} causes CSE to follow the jump around the
5543 body of the @code{if}.
5545 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
5547 @item -frerun-cse-after-loop
5548 @opindex frerun-cse-after-loop
5549 Re-run common subexpression elimination after loop optimizations has been
5552 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
5556 Perform a global common subexpression elimination pass.
5557 This pass also performs global constant and copy propagation.
5559 @emph{Note:} When compiling a program using computed gotos, a GCC
5560 extension, you may get better runtime performance if you disable
5561 the global common subexpression elimination pass by adding
5562 @option{-fno-gcse} to the command line.
5564 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
5568 When @option{-fgcse-lm} is enabled, global common subexpression elimination will
5569 attempt to move loads which are only killed by stores into themselves. This
5570 allows a loop containing a load/store sequence to be changed to a load outside
5571 the loop, and a copy/store within the loop.
5573 Enabled by default when gcse is enabled.
5577 When @option{-fgcse-sm} is enabled, a store motion pass is run after
5578 global common subexpression elimination. This pass will attempt to move
5579 stores out of loops. When used in conjunction with @option{-fgcse-lm},
5580 loops containing a load/store sequence can be changed to a load before
5581 the loop and a store after the loop.
5583 Not enabled at any optimization level.
5587 When @option{-fgcse-las} is enabled, the global common subexpression
5588 elimination pass eliminates redundant loads that come after stores to the
5589 same memory location (both partial and full redundancies).
5591 Not enabled at any optimization level.
5593 @item -fgcse-after-reload
5594 @opindex fgcse-after-reload
5595 When @option{-fgcse-after-reload} is enabled, a redundant load elimination
5596 pass is performed after reload. The purpose of this pass is to cleanup
5599 @item -funsafe-loop-optimizations
5600 @opindex funsafe-loop-optimizations
5601 If given, the loop optimizer will assume that loop indices do not
5602 overflow, and that the loops with nontrivial exit condition are not
5603 infinite. This enables a wider range of loop optimizations even if
5604 the loop optimizer itself cannot prove that these assumptions are valid.
5605 Using @option{-Wunsafe-loop-optimizations}, the compiler will warn you
5606 if it finds this kind of loop.
5608 @item -fcrossjumping
5609 @opindex fcrossjumping
5610 Perform cross-jumping transformation. This transformation unifies equivalent code and save code size. The
5611 resulting code may or may not perform better than without cross-jumping.
5613 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
5615 @item -fauto-inc-dec
5616 @opindex fauto-inc-dec
5617 Combine increments or decrements of addresses with memory accesses.
5618 This pass is always skipped on architectures that do not have
5619 instructions to support this. Enabled by default at @option{-O} and
5620 higher on architectures that support this.
5624 Perform dead code elimination (DCE) on RTL@.
5625 Enabled by default at @option{-O} and higher.
5629 Perform dead store elimination (DSE) on RTL@.
5630 Enabled by default at @option{-O} and higher.
5632 @item -fif-conversion
5633 @opindex fif-conversion
5634 Attempt to transform conditional jumps into branch-less equivalents. This
5635 include use of conditional moves, min, max, set flags and abs instructions, and
5636 some tricks doable by standard arithmetics. The use of conditional execution
5637 on chips where it is available is controlled by @code{if-conversion2}.
5639 Enabled at levels @option{-O}, @option{-O2}, @option{-O3}, @option{-Os}.
5641 @item -fif-conversion2
5642 @opindex fif-conversion2
5643 Use conditional execution (where available) to transform conditional jumps into
5644 branch-less equivalents.
5646 Enabled at levels @option{-O}, @option{-O2}, @option{-O3}, @option{-Os}.
5648 @item -fdelete-null-pointer-checks
5649 @opindex fdelete-null-pointer-checks
5650 Use global dataflow analysis to identify and eliminate useless checks
5651 for null pointers. The compiler assumes that dereferencing a null
5652 pointer would have halted the program. If a pointer is checked after
5653 it has already been dereferenced, it cannot be null.
5655 In some environments, this assumption is not true, and programs can
5656 safely dereference null pointers. Use
5657 @option{-fno-delete-null-pointer-checks} to disable this optimization
5658 for programs which depend on that behavior.
5660 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
5662 @item -fexpensive-optimizations
5663 @opindex fexpensive-optimizations
5664 Perform a number of minor optimizations that are relatively expensive.
5666 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
5668 @item -foptimize-register-move
5670 @opindex foptimize-register-move
5672 Attempt to reassign register numbers in move instructions and as
5673 operands of other simple instructions in order to maximize the amount of
5674 register tying. This is especially helpful on machines with two-operand
5677 Note @option{-fregmove} and @option{-foptimize-register-move} are the same
5680 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
5682 @item -fdelayed-branch
5683 @opindex fdelayed-branch
5684 If supported for the target machine, attempt to reorder instructions
5685 to exploit instruction slots available after delayed branch
5688 Enabled at levels @option{-O}, @option{-O2}, @option{-O3}, @option{-Os}.
5690 @item -fschedule-insns
5691 @opindex fschedule-insns
5692 If supported for the target machine, attempt to reorder instructions to
5693 eliminate execution stalls due to required data being unavailable. This
5694 helps machines that have slow floating point or memory load instructions
5695 by allowing other instructions to be issued until the result of the load
5696 or floating point instruction is required.
5698 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
5700 @item -fschedule-insns2
5701 @opindex fschedule-insns2
5702 Similar to @option{-fschedule-insns}, but requests an additional pass of
5703 instruction scheduling after register allocation has been done. This is
5704 especially useful on machines with a relatively small number of
5705 registers and where memory load instructions take more than one cycle.
5707 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
5709 @item -fno-sched-interblock
5710 @opindex fno-sched-interblock
5711 Don't schedule instructions across basic blocks. This is normally
5712 enabled by default when scheduling before register allocation, i.e.@:
5713 with @option{-fschedule-insns} or at @option{-O2} or higher.
5715 @item -fno-sched-spec
5716 @opindex fno-sched-spec
5717 Don't allow speculative motion of non-load instructions. This is normally
5718 enabled by default when scheduling before register allocation, i.e.@:
5719 with @option{-fschedule-insns} or at @option{-O2} or higher.
5721 @item -fsched-spec-load
5722 @opindex fsched-spec-load
5723 Allow speculative motion of some load instructions. This only makes
5724 sense when scheduling before register allocation, i.e.@: with
5725 @option{-fschedule-insns} or at @option{-O2} or higher.
5727 @item -fsched-spec-load-dangerous
5728 @opindex fsched-spec-load-dangerous
5729 Allow speculative motion of more load instructions. This only makes
5730 sense when scheduling before register allocation, i.e.@: with
5731 @option{-fschedule-insns} or at @option{-O2} or higher.
5733 @item -fsched-stalled-insns
5734 @itemx -fsched-stalled-insns=@var{n}
5735 @opindex fsched-stalled-insns
5736 Define how many insns (if any) can be moved prematurely from the queue
5737 of stalled insns into the ready list, during the second scheduling pass.
5738 @option{-fno-sched-stalled-insns} means that no insns will be moved
5739 prematurely, @option{-fsched-stalled-insns=0} means there is no limit
5740 on how many queued insns can be moved prematurely.
5741 @option{-fsched-stalled-insns} without a value is equivalent to
5742 @option{-fsched-stalled-insns=1}.
5744 @item -fsched-stalled-insns-dep
5745 @itemx -fsched-stalled-insns-dep=@var{n}
5746 @opindex fsched-stalled-insns-dep
5747 Define how many insn groups (cycles) will be examined for a dependency
5748 on a stalled insn that is candidate for premature removal from the queue
5749 of stalled insns. This has an effect only during the second scheduling pass,
5750 and only if @option{-fsched-stalled-insns} is used.
5751 @option{-fno-sched-stalled-insns-dep} is equivalent to
5752 @option{-fsched-stalled-insns-dep=0}.
5753 @option{-fsched-stalled-insns-dep} without a value is equivalent to
5754 @option{-fsched-stalled-insns-dep=1}.
5756 @item -fsched2-use-superblocks
5757 @opindex fsched2-use-superblocks
5758 When scheduling after register allocation, do use superblock scheduling
5759 algorithm. Superblock scheduling allows motion across basic block boundaries
5760 resulting on faster schedules. This option is experimental, as not all machine
5761 descriptions used by GCC model the CPU closely enough to avoid unreliable
5762 results from the algorithm.
5764 This only makes sense when scheduling after register allocation, i.e.@: with
5765 @option{-fschedule-insns2} or at @option{-O2} or higher.
5767 @item -fsched2-use-traces
5768 @opindex fsched2-use-traces
5769 Use @option{-fsched2-use-superblocks} algorithm when scheduling after register
5770 allocation and additionally perform code duplication in order to increase the
5771 size of superblocks using tracer pass. See @option{-ftracer} for details on
5774 This mode should produce faster but significantly longer programs. Also
5775 without @option{-fbranch-probabilities} the traces constructed may not
5776 match the reality and hurt the performance. This only makes
5777 sense when scheduling after register allocation, i.e.@: with
5778 @option{-fschedule-insns2} or at @option{-O2} or higher.
5782 Eliminate redundant sign extension instructions and move the non-redundant
5783 ones to optimal placement using lazy code motion (LCM).
5785 @item -freschedule-modulo-scheduled-loops
5786 @opindex freschedule-modulo-scheduled-loops
5787 The modulo scheduling comes before the traditional scheduling, if a loop
5788 was modulo scheduled we may want to prevent the later scheduling passes
5789 from changing its schedule, we use this option to control that.
5791 @item -fcaller-saves
5792 @opindex fcaller-saves
5793 Enable values to be allocated in registers that will be clobbered by
5794 function calls, by emitting extra instructions to save and restore the
5795 registers around such calls. Such allocation is done only when it
5796 seems to result in better code than would otherwise be produced.
5798 This option is always enabled by default on certain machines, usually
5799 those which have no call-preserved registers to use instead.
5801 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
5803 @item -ftree-reassoc
5804 @opindex ftree-reassoc
5805 Perform reassociation on trees. This flag is enabled by default
5806 at @option{-O} and higher.
5810 Perform partial redundancy elimination (PRE) on trees. This flag is
5811 enabled by default at @option{-O2} and @option{-O3}.
5815 Perform full redundancy elimination (FRE) on trees. The difference
5816 between FRE and PRE is that FRE only considers expressions
5817 that are computed on all paths leading to the redundant computation.
5818 This analysis is faster than PRE, though it exposes fewer redundancies.
5819 This flag is enabled by default at @option{-O} and higher.
5821 @item -ftree-copy-prop
5822 @opindex ftree-copy-prop
5823 Perform copy propagation on trees. This pass eliminates unnecessary
5824 copy operations. This flag is enabled by default at @option{-O} and
5827 @item -fipa-pure-const
5828 @opindex fipa-pure-const
5829 Discover which functions are pure or constant.
5830 Enabled by default at @option{-O} and higher.
5832 @item -fipa-reference
5833 @opindex fipa-reference
5834 Discover which static variables do not escape cannot escape the
5836 Enabled by default at @option{-O} and higher.
5838 @item -fipa-struct-reorg
5839 @opindex fipa-struct-reorg
5840 Perform structure reorganization optimization, that change C-like structures
5841 layout in order to better utilize spatial locality. This transformation is
5842 affective for programs containing arrays of structures. Available in two
5843 compilation modes: profile-based (enabled with @option{-fprofile-generate})
5844 or static (which uses built-in heuristics). Require @option{-fipa-type-escape}
5845 to provide the safety of this transformation. It works only in whole program
5846 mode, so it requires @option{-fwhole-program} and @option{-combine} to be
5847 enabled. Structures considered @samp{cold} by this transformation are not
5848 affected (see @option{--param struct-reorg-cold-struct-ratio=@var{value}}).
5850 With this flag, the program debug info reflects a new structure layout.
5854 Perform interprocedural pointer analysis.
5858 Perform interprocedural constant propagation.
5859 This optimization analyzes the program to determine when values passed
5860 to functions are constants and then optimizes accordingly.
5861 This optimization can substantially increase performance
5862 if the application has constants passed to functions, but
5863 because this optimization can create multiple copies of functions,
5864 it may significantly increase code size.
5866 @item -fipa-matrix-reorg
5867 @opindex fipa-matrix-reorg
5868 Perform matrix flattening and transposing.
5869 Matrix flattening tries to replace a m-dimensional matrix
5870 with its equivalent n-dimensional matrix, where n < m.
5871 This reduces the level of indirection needed for accessing the elements
5872 of the matrix. The second optimization is matrix transposing that
5873 attemps to change the order of the matrix's dimensions in order to
5874 improve cache locality.
5875 Both optimizations need fwhole-program flag.
5876 Transposing is enabled only if profiling information is avaliable.
5881 Perform forward store motion on trees. This flag is
5882 enabled by default at @option{-O} and higher.
5886 Perform sparse conditional constant propagation (CCP) on trees. This
5887 pass only operates on local scalar variables and is enabled by default
5888 at @option{-O} and higher.
5890 @item -ftree-store-ccp
5891 @opindex ftree-store-ccp
5892 Perform sparse conditional constant propagation (CCP) on trees. This
5893 pass operates on both local scalar variables and memory stores and
5894 loads (global variables, structures, arrays, etc). This flag is
5895 enabled by default at @option{-O2} and higher.
5897 @item -ftree-switch-conversion
5898 Perform conversion of simple initializations in a switch to
5899 initializations from a scalar array. This flag is enabled by default
5900 at @option{-O2} and higher.
5904 Perform dead code elimination (DCE) on trees. This flag is enabled by
5905 default at @option{-O} and higher.
5907 @item -ftree-builtin-call-dce
5908 @opindex ftree-builtin-call-dce
5909 Perform conditional dead code elimination (DCE) for calls to builtin functions
5910 that may set @code{errno} but are otherwise side-effect free. This flag is
5911 enabled by default at @option{-O2} and higher if @option{-Os} is not also
5914 @item -ftree-dominator-opts
5915 @opindex ftree-dominator-opts
5916 Perform a variety of simple scalar cleanups (constant/copy
5917 propagation, redundancy elimination, range propagation and expression
5918 simplification) based on a dominator tree traversal. This also
5919 performs jump threading (to reduce jumps to jumps). This flag is
5920 enabled by default at @option{-O} and higher.
5924 Perform dead store elimination (DSE) on trees. A dead store is a store into
5925 a memory location which will later be overwritten by another store without
5926 any intervening loads. In this case the earlier store can be deleted. This
5927 flag is enabled by default at @option{-O} and higher.
5931 Perform loop header copying on trees. This is beneficial since it increases
5932 effectiveness of code motion optimizations. It also saves one jump. This flag
5933 is enabled by default at @option{-O} and higher. It is not enabled
5934 for @option{-Os}, since it usually increases code size.
5936 @item -ftree-loop-optimize
5937 @opindex ftree-loop-optimize
5938 Perform loop optimizations on trees. This flag is enabled by default
5939 at @option{-O} and higher.
5941 @item -ftree-loop-linear
5942 @opindex ftree-loop-linear
5943 Perform linear loop transformations on tree. This flag can improve cache
5944 performance and allow further loop optimizations to take place.
5946 @item -fcheck-data-deps
5947 @opindex fcheck-data-deps
5948 Compare the results of several data dependence analyzers. This option
5949 is used for debugging the data dependence analyzers.
5951 @item -ftree-loop-distribution
5952 Perform loop distribution. This flag can improve cache performance on
5953 big loop bodies and allow further loop optimizations, like
5954 parallelization or vectorization, to take place. For example, the loop
5971 @item -ftree-loop-im
5972 @opindex ftree-loop-im
5973 Perform loop invariant motion on trees. This pass moves only invariants that
5974 would be hard to handle at RTL level (function calls, operations that expand to
5975 nontrivial sequences of insns). With @option{-funswitch-loops} it also moves
5976 operands of conditions that are invariant out of the loop, so that we can use
5977 just trivial invariantness analysis in loop unswitching. The pass also includes
5980 @item -ftree-loop-ivcanon
5981 @opindex ftree-loop-ivcanon
5982 Create a canonical counter for number of iterations in the loop for that
5983 determining number of iterations requires complicated analysis. Later
5984 optimizations then may determine the number easily. Useful especially
5985 in connection with unrolling.
5989 Perform induction variable optimizations (strength reduction, induction
5990 variable merging and induction variable elimination) on trees.
5992 @item -ftree-parallelize-loops=n
5993 @opindex ftree-parallelize-loops
5994 Parallelize loops, i.e., split their iteration space to run in n threads.
5995 This is only possible for loops whose iterations are independent
5996 and can be arbitrarily reordered. The optimization is only
5997 profitable on multiprocessor machines, for loops that are CPU-intensive,
5998 rather than constrained e.g.@: by memory bandwidth. This option
5999 implies @option{-pthread}, and thus is only supported on targets
6000 that have support for @option{-pthread}.
6004 Perform scalar replacement of aggregates. This pass replaces structure
6005 references with scalars to prevent committing structures to memory too
6006 early. This flag is enabled by default at @option{-O} and higher.
6008 @item -ftree-copyrename
6009 @opindex ftree-copyrename
6010 Perform copy renaming on trees. This pass attempts to rename compiler
6011 temporaries to other variables at copy locations, usually resulting in
6012 variable names which more closely resemble the original variables. This flag
6013 is enabled by default at @option{-O} and higher.
6017 Perform temporary expression replacement during the SSA->normal phase. Single
6018 use/single def temporaries are replaced at their use location with their
6019 defining expression. This results in non-GIMPLE code, but gives the expanders
6020 much more complex trees to work on resulting in better RTL generation. This is
6021 enabled by default at @option{-O} and higher.
6023 @item -ftree-vectorize
6024 @opindex ftree-vectorize
6025 Perform loop vectorization on trees. This flag is enabled by default at
6028 @item -ftree-vect-loop-version
6029 @opindex ftree-vect-loop-version
6030 Perform loop versioning when doing loop vectorization on trees. When a loop
6031 appears to be vectorizable except that data alignment or data dependence cannot
6032 be determined at compile time then vectorized and non-vectorized versions of
6033 the loop are generated along with runtime checks for alignment or dependence
6034 to control which version is executed. This option is enabled by default
6035 except at level @option{-Os} where it is disabled.
6037 @item -fvect-cost-model
6038 @opindex fvect-cost-model
6039 Enable cost model for vectorization.
6043 Perform Value Range Propagation on trees. This is similar to the
6044 constant propagation pass, but instead of values, ranges of values are
6045 propagated. This allows the optimizers to remove unnecessary range
6046 checks like array bound checks and null pointer checks. This is
6047 enabled by default at @option{-O2} and higher. Null pointer check
6048 elimination is only done if @option{-fdelete-null-pointer-checks} is
6053 Perform tail duplication to enlarge superblock size. This transformation
6054 simplifies the control flow of the function allowing other optimizations to do
6057 @item -funroll-loops
6058 @opindex funroll-loops
6059 Unroll loops whose number of iterations can be determined at compile
6060 time or upon entry to the loop. @option{-funroll-loops} implies
6061 @option{-frerun-cse-after-loop}. This option makes code larger,
6062 and may or may not make it run faster.
6064 @item -funroll-all-loops
6065 @opindex funroll-all-loops
6066 Unroll all loops, even if their number of iterations is uncertain when
6067 the loop is entered. This usually makes programs run more slowly.
6068 @option{-funroll-all-loops} implies the same options as
6069 @option{-funroll-loops},
6071 @item -fsplit-ivs-in-unroller
6072 @opindex fsplit-ivs-in-unroller
6073 Enables expressing of values of induction variables in later iterations
6074 of the unrolled loop using the value in the first iteration. This breaks
6075 long dependency chains, thus improving efficiency of the scheduling passes.
6077 Combination of @option{-fweb} and CSE is often sufficient to obtain the
6078 same effect. However in cases the loop body is more complicated than
6079 a single basic block, this is not reliable. It also does not work at all
6080 on some of the architectures due to restrictions in the CSE pass.
6082 This optimization is enabled by default.
6084 @item -fvariable-expansion-in-unroller
6085 @opindex fvariable-expansion-in-unroller
6086 With this option, the compiler will create multiple copies of some
6087 local variables when unrolling a loop which can result in superior code.
6089 @item -fpredictive-commoning
6090 @opindex fpredictive-commoning
6091 Perform predictive commoning optimization, i.e., reusing computations
6092 (especially memory loads and stores) performed in previous
6093 iterations of loops.
6095 This option is enabled at level @option{-O3}.
6097 @item -fprefetch-loop-arrays
6098 @opindex fprefetch-loop-arrays
6099 If supported by the target machine, generate instructions to prefetch
6100 memory to improve the performance of loops that access large arrays.
6102 This option may generate better or worse code; results are highly
6103 dependent on the structure of loops within the source code.
6105 Disabled at level @option{-Os}.
6108 @itemx -fno-peephole2
6109 @opindex fno-peephole
6110 @opindex fno-peephole2
6111 Disable any machine-specific peephole optimizations. The difference
6112 between @option{-fno-peephole} and @option{-fno-peephole2} is in how they
6113 are implemented in the compiler; some targets use one, some use the
6114 other, a few use both.
6116 @option{-fpeephole} is enabled by default.
6117 @option{-fpeephole2} enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
6119 @item -fno-guess-branch-probability
6120 @opindex fno-guess-branch-probability
6121 Do not guess branch probabilities using heuristics.
6123 GCC will use heuristics to guess branch probabilities if they are
6124 not provided by profiling feedback (@option{-fprofile-arcs}). These
6125 heuristics are based on the control flow graph. If some branch probabilities
6126 are specified by @samp{__builtin_expect}, then the heuristics will be
6127 used to guess branch probabilities for the rest of the control flow graph,
6128 taking the @samp{__builtin_expect} info into account. The interactions
6129 between the heuristics and @samp{__builtin_expect} can be complex, and in
6130 some cases, it may be useful to disable the heuristics so that the effects
6131 of @samp{__builtin_expect} are easier to understand.
6133 The default is @option{-fguess-branch-probability} at levels
6134 @option{-O}, @option{-O2}, @option{-O3}, @option{-Os}.
6136 @item -freorder-blocks
6137 @opindex freorder-blocks
6138 Reorder basic blocks in the compiled function in order to reduce number of
6139 taken branches and improve code locality.
6141 Enabled at levels @option{-O2}, @option{-O3}.
6143 @item -freorder-blocks-and-partition
6144 @opindex freorder-blocks-and-partition
6145 In addition to reordering basic blocks in the compiled function, in order
6146 to reduce number of taken branches, partitions hot and cold basic blocks
6147 into separate sections of the assembly and .o files, to improve
6148 paging and cache locality performance.
6150 This optimization is automatically turned off in the presence of
6151 exception handling, for linkonce sections, for functions with a user-defined
6152 section attribute and on any architecture that does not support named
6155 @item -freorder-functions
6156 @opindex freorder-functions
6157 Reorder functions in the object file in order to
6158 improve code locality. This is implemented by using special
6159 subsections @code{.text.hot} for most frequently executed functions and
6160 @code{.text.unlikely} for unlikely executed functions. Reordering is done by
6161 the linker so object file format must support named sections and linker must
6162 place them in a reasonable way.
6164 Also profile feedback must be available in to make this option effective. See
6165 @option{-fprofile-arcs} for details.
6167 Enabled at levels @option{-O2}, @option{-O3}, @option{-Os}.
6169 @item -fstrict-aliasing
6170 @opindex fstrict-aliasing
6171 Allows the compiler to assume the strictest aliasing rules applicable to
6172 the language being compiled. For C (and C++), this activates
6173 optimizations based on the type of expressions. In particular, an
6174 object of one type is assumed never to reside at the same address as an
6175 object of a different type, unless the types are almost the same. For
6176 example, an @code{unsigned int} can alias an @code{int}, but not a
6177 @code{void*} or a @code{double}. A character type may alias any other
6180 @anchor{Type-punning}Pay special attention to code like this:
6193 The practice of reading from a different union member than the one most
6194 recently written to (called ``type-punning'') is common. Even with
6195 @option{-fstrict-aliasing}, type-punning is allowed, provided the memory
6196 is accessed through the union type. So, the code above will work as
6197 expected. @xref{Structures unions enumerations and bit-fields
6198 implementation}. However, this code might not:
6209 Similarly, access by taking the address, casting the resulting pointer
6210 and dereferencing the result has undefined behavior, even if the cast
6211 uses a union type, e.g.:
6215 return ((union a_union *) &d)->i;
6219 The @option{-fstrict-aliasing} option is enabled at levels
6220 @option{-O2}, @option{-O3}, @option{-Os}.
6222 @item -fstrict-overflow
6223 @opindex fstrict-overflow
6224 Allow the compiler to assume strict signed overflow rules, depending
6225 on the language being compiled. For C (and C++) this means that
6226 overflow when doing arithmetic with signed numbers is undefined, which
6227 means that the compiler may assume that it will not happen. This
6228 permits various optimizations. For example, the compiler will assume
6229 that an expression like @code{i + 10 > i} will always be true for
6230 signed @code{i}. This assumption is only valid if signed overflow is
6231 undefined, as the expression is false if @code{i + 10} overflows when
6232 using twos complement arithmetic. When this option is in effect any
6233 attempt to determine whether an operation on signed numbers will
6234 overflow must be written carefully to not actually involve overflow.
6236 This option also allows the compiler to assume strict pointer
6237 semantics: given a pointer to an object, if adding an offset to that
6238 pointer does not produce a pointer to the same object, the addition is
6239 undefined. This permits the compiler to conclude that @code{p + u >
6240 p} is always true for a pointer @code{p} and unsigned integer
6241 @code{u}. This assumption is only valid because pointer wraparound is
6242 undefined, as the expression is false if @code{p + u} overflows using
6243 twos complement arithmetic.
6245 See also the @option{-fwrapv} option. Using @option{-fwrapv} means
6246 that integer signed overflow is fully defined: it wraps. When
6247 @option{-fwrapv} is used, there is no difference between
6248 @option{-fstrict-overflow} and @option{-fno-strict-overflow} for
6249 integers. With @option{-fwrapv} certain types of overflow are
6250 permitted. For example, if the compiler gets an overflow when doing
6251 arithmetic on constants, the overflowed value can still be used with
6252 @option{-fwrapv}, but not otherwise.
6254 The @option{-fstrict-overflow} option is enabled at levels
6255 @option{-O2}, @option{-O3}, @option{-Os}.
6257 @item -falign-functions
6258 @itemx -falign-functions=@var{n}
6259 @opindex falign-functions
6260 Align the start of functions to the next power-of-two greater than
6261 @var{n}, skipping up to @var{n} bytes. For instance,
6262 @option{-falign-functions=32} aligns functions to the next 32-byte
6263 boundary, but @option{-falign-functions=24} would align to the next
6264 32-byte boundary only if this can be done by skipping 23 bytes or less.
6266 @option{-fno-align-functions} and @option{-falign-functions=1} are
6267 equivalent and mean that functions will not be aligned.
6269 Some assemblers only support this flag when @var{n} is a power of two;
6270 in that case, it is rounded up.
6272 If @var{n} is not specified or is zero, use a machine-dependent default.
6274 Enabled at levels @option{-O2}, @option{-O3}.
6276 @item -falign-labels
6277 @itemx -falign-labels=@var{n}
6278 @opindex falign-labels
6279 Align all branch targets to a power-of-two boundary, skipping up to
6280 @var{n} bytes like @option{-falign-functions}. This option can easily
6281 make code slower, because it must insert dummy operations for when the
6282 branch target is reached in the usual flow of the code.
6284 @option{-fno-align-labels} and @option{-falign-labels=1} are
6285 equivalent and mean that labels will not be aligned.
6287 If @option{-falign-loops} or @option{-falign-jumps} are applicable and
6288 are greater than this value, then their values are used instead.
6290 If @var{n} is not specified or is zero, use a machine-dependent default
6291 which is very likely to be @samp{1}, meaning no alignment.
6293 Enabled at levels @option{-O2}, @option{-O3}.
6296 @itemx -falign-loops=@var{n}
6297 @opindex falign-loops
6298 Align loops to a power-of-two boundary, skipping up to @var{n} bytes
6299 like @option{-falign-functions}. The hope is that the loop will be
6300 executed many times, which will make up for any execution of the dummy
6303 @option{-fno-align-loops} and @option{-falign-loops=1} are
6304 equivalent and mean that loops will not be aligned.
6306 If @var{n} is not specified or is zero, use a machine-dependent default.
6308 Enabled at levels @option{-O2}, @option{-O3}.
6311 @itemx -falign-jumps=@var{n}
6312 @opindex falign-jumps
6313 Align branch targets to a power-of-two boundary, for branch targets
6314 where the targets can only be reached by jumping, skipping up to @var{n}
6315 bytes like @option{-falign-functions}. In this case, no dummy operations
6318 @option{-fno-align-jumps} and @option{-falign-jumps=1} are
6319 equivalent and mean that loops will not be aligned.
6321 If @var{n} is not specified or is zero, use a machine-dependent default.
6323 Enabled at levels @option{-O2}, @option{-O3}.
6325 @item -funit-at-a-time
6326 @opindex funit-at-a-time
6327 This option is left for compatibility reasons. @option{-funit-at-a-time}
6328 has no effect, while @option{-fno-unit-at-a-time} implies
6329 @option{-fno-toplevel-reorder} and @option{-fno-section-anchors}.
6333 @item -fno-toplevel-reorder
6334 @opindex fno-toplevel-reorder
6335 Do not reorder top-level functions, variables, and @code{asm}
6336 statements. Output them in the same order that they appear in the
6337 input file. When this option is used, unreferenced static variables
6338 will not be removed. This option is intended to support existing code
6339 which relies on a particular ordering. For new code, it is better to
6342 Enabled at level @option{-O0}. When disabled explicitly, it also imply
6343 @option{-fno-section-anchors} that is otherwise enabled at @option{-O0} on some
6348 Constructs webs as commonly used for register allocation purposes and assign
6349 each web individual pseudo register. This allows the register allocation pass
6350 to operate on pseudos directly, but also strengthens several other optimization
6351 passes, such as CSE, loop optimizer and trivial dead code remover. It can,
6352 however, make debugging impossible, since variables will no longer stay in a
6355 Enabled by default with @option{-funroll-loops}.
6357 @item -fwhole-program
6358 @opindex fwhole-program
6359 Assume that the current compilation unit represents whole program being
6360 compiled. All public functions and variables with the exception of @code{main}
6361 and those merged by attribute @code{externally_visible} become static functions
6362 and in a affect gets more aggressively optimized by interprocedural optimizers.
6363 While this option is equivalent to proper use of @code{static} keyword for
6364 programs consisting of single file, in combination with option
6365 @option{--combine} this flag can be used to compile most of smaller scale C
6366 programs since the functions and variables become local for the whole combined
6367 compilation unit, not for the single source file itself.
6369 This option is not supported for Fortran programs.
6371 @item -fcprop-registers
6372 @opindex fcprop-registers
6373 After register allocation and post-register allocation instruction splitting,
6374 we perform a copy-propagation pass to try to reduce scheduling dependencies
6375 and occasionally eliminate the copy.
6377 Enabled at levels @option{-O}, @option{-O2}, @option{-O3}, @option{-Os}.
6379 @item -fprofile-dir=@var{path}
6380 @opindex fprofile-dir
6382 Set the directory to search the profile data files in to @var{path}.
6383 This option affects only the profile data generated by
6384 @option{-fprofile-generate}, @option{-ftest-coverage}, @option{-fprofile-arcs}
6385 and used by @option{-fprofile-use} and @option{-fbranch-probabilities}
6386 and its related options.
6387 By default, GCC will use the current directory as @var{path}
6388 thus the profile data file will appear in the same directory as the object file.
6390 @item -fprofile-generate
6391 @itemx -fprofile-generate=@var{path}
6392 @opindex fprofile-generate
6394 Enable options usually used for instrumenting application to produce
6395 profile useful for later recompilation with profile feedback based
6396 optimization. You must use @option{-fprofile-generate} both when
6397 compiling and when linking your program.
6399 The following options are enabled: @code{-fprofile-arcs}, @code{-fprofile-values}, @code{-fvpt}.
6401 If @var{path} is specified, GCC will look at the @var{path} to find
6402 the profile feeedback data files. See @option{-fprofile-dir}.
6405 @itemx -fprofile-use=@var{path}
6406 @opindex fprofile-use
6407 Enable profile feedback directed optimizations, and optimizations
6408 generally profitable only with profile feedback available.
6410 The following options are enabled: @code{-fbranch-probabilities}, @code{-fvpt},
6411 @code{-funroll-loops}, @code{-fpeel-loops}, @code{-ftracer}
6413 By default, GCC emits an error message if the feedback profiles do not
6414 match the source code. This error can be turned into a warning by using
6415 @option{-Wcoverage-mismatch}. Note this may result in poorly optimized
6418 If @var{path} is specified, GCC will look at the @var{path} to find
6419 the profile feedback data files. See @option{-fprofile-dir}.
6422 The following options control compiler behavior regarding floating
6423 point arithmetic. These options trade off between speed and
6424 correctness. All must be specifically enabled.
6428 @opindex ffloat-store
6429 Do not store floating point variables in registers, and inhibit other
6430 options that might change whether a floating point value is taken from a
6433 @cindex floating point precision
6434 This option prevents undesirable excess precision on machines such as
6435 the 68000 where the floating registers (of the 68881) keep more
6436 precision than a @code{double} is supposed to have. Similarly for the
6437 x86 architecture. For most programs, the excess precision does only
6438 good, but a few programs rely on the precise definition of IEEE floating
6439 point. Use @option{-ffloat-store} for such programs, after modifying
6440 them to store all pertinent intermediate computations into variables.
6444 Sets @option{-fno-math-errno}, @option{-funsafe-math-optimizations},
6445 @option{-ffinite-math-only}, @option{-fno-rounding-math},
6446 @option{-fno-signaling-nans} and @option{-fcx-limited-range}.
6448 This option causes the preprocessor macro @code{__FAST_MATH__} to be defined.
6450 This option is not turned on by any @option{-O} option since
6451 it can result in incorrect output for programs which depend on
6452 an exact implementation of IEEE or ISO rules/specifications for
6453 math functions. It may, however, yield faster code for programs
6454 that do not require the guarantees of these specifications.
6456 @item -fno-math-errno
6457 @opindex fno-math-errno
6458 Do not set ERRNO after calling math functions that are executed
6459 with a single instruction, e.g., sqrt. A program that relies on
6460 IEEE exceptions for math error handling may want to use this flag
6461 for speed while maintaining IEEE arithmetic compatibility.
6463 This option is not turned on by any @option{-O} option since
6464 it can result in incorrect output for programs which depend on
6465 an exact implementation of IEEE or ISO rules/specifications for
6466 math functions. It may, however, yield faster code for programs
6467 that do not require the guarantees of these specifications.
6469 The default is @option{-fmath-errno}.
6471 On Darwin systems, the math library never sets @code{errno}. There is
6472 therefore no reason for the compiler to consider the possibility that
6473 it might, and @option{-fno-math-errno} is the default.
6475 @item -funsafe-math-optimizations
6476 @opindex funsafe-math-optimizations
6478 Allow optimizations for floating-point arithmetic that (a) assume
6479 that arguments and results are valid and (b) may violate IEEE or
6480 ANSI standards. When used at link-time, it may include libraries
6481 or startup files that change the default FPU control word or other
6482 similar optimizations.
6484 This option is not turned on by any @option{-O} option since
6485 it can result in incorrect output for programs which depend on
6486 an exact implementation of IEEE or ISO rules/specifications for
6487 math functions. It may, however, yield faster code for programs
6488 that do not require the guarantees of these specifications.
6489 Enables @option{-fno-signed-zeros}, @option{-fno-trapping-math},
6490 @option{-fassociative-math} and @option{-freciprocal-math}.
6492 The default is @option{-fno-unsafe-math-optimizations}.
6494 @item -fassociative-math
6495 @opindex fassociative-math
6497 Allow re-association of operands in series of floating-point operations.
6498 This violates the ISO C and C++ language standard by possibly changing
6499 computation result. NOTE: re-ordering may change the sign of zero as
6500 well as ignore NaNs and inhibit or create underflow or overflow (and
6501 thus cannot be used on a code which relies on rounding behavior like
6502 @code{(x + 2**52) - 2**52)}. May also reorder floating-point comparisons
6503 and thus may not be used when ordered comparisons are required.
6504 This option requires that both @option{-fno-signed-zeros} and
6505 @option{-fno-trapping-math} be in effect. Moreover, it doesn't make
6506 much sense with @option{-frounding-math}.
6508 The default is @option{-fno-associative-math}.
6510 @item -freciprocal-math
6511 @opindex freciprocal-math
6513 Allow the reciprocal of a value to be used instead of dividing by
6514 the value if this enables optimizations. For example @code{x / y}
6515 can be replaced with @code{x * (1/y)} which is useful if @code{(1/y)}
6516 is subject to common subexpression elimination. Note that this loses
6517 precision and increases the number of flops operating on the value.
6519 The default is @option{-fno-reciprocal-math}.
6521 @item -ffinite-math-only
6522 @opindex ffinite-math-only
6523 Allow optimizations for floating-point arithmetic that assume
6524 that arguments and results are not NaNs or +-Infs.
6526 This option is not turned on by any @option{-O} option since
6527 it can result in incorrect output for programs which depend on
6528 an exact implementation of IEEE or ISO rules/specifications for
6529 math functions. It may, however, yield faster code for programs
6530 that do not require the guarantees of these specifications.
6532 The default is @option{-fno-finite-math-only}.
6534 @item -fno-signed-zeros
6535 @opindex fno-signed-zeros
6536 Allow optimizations for floating point arithmetic that ignore the
6537 signedness of zero. IEEE arithmetic specifies the behavior of
6538 distinct +0.0 and @minus{}0.0 values, which then prohibits simplification
6539 of expressions such as x+0.0 or 0.0*x (even with @option{-ffinite-math-only}).
6540 This option implies that the sign of a zero result isn't significant.
6542 The default is @option{-fsigned-zeros}.
6544 @item -fno-trapping-math
6545 @opindex fno-trapping-math
6546 Compile code assuming that floating-point operations cannot generate
6547 user-visible traps. These traps include division by zero, overflow,
6548 underflow, inexact result and invalid operation. This option requires
6549 that @option{-fno-signaling-nans} be in effect. Setting this option may
6550 allow faster code if one relies on ``non-stop'' IEEE arithmetic, for example.
6552 This option should never be turned on by any @option{-O} option since
6553 it can result in incorrect output for programs which depend on
6554 an exact implementation of IEEE or ISO rules/specifications for
6557 The default is @option{-ftrapping-math}.
6559 @item -frounding-math
6560 @opindex frounding-math
6561 Disable transformations and optimizations that assume default floating
6562 point rounding behavior. This is round-to-zero for all floating point
6563 to integer conversions, and round-to-nearest for all other arithmetic
6564 truncations. This option should be specified for programs that change
6565 the FP rounding mode dynamically, or that may be executed with a
6566 non-default rounding mode. This option disables constant folding of
6567 floating point expressions at compile-time (which may be affected by
6568 rounding mode) and arithmetic transformations that are unsafe in the
6569 presence of sign-dependent rounding modes.
6571 The default is @option{-fno-rounding-math}.
6573 This option is experimental and does not currently guarantee to
6574 disable all GCC optimizations that are affected by rounding mode.
6575 Future versions of GCC may provide finer control of this setting
6576 using C99's @code{FENV_ACCESS} pragma. This command line option
6577 will be used to specify the default state for @code{FENV_ACCESS}.
6579 @item -frtl-abstract-sequences
6580 @opindex frtl-abstract-sequences
6581 It is a size optimization method. This option is to find identical
6582 sequences of code, which can be turned into pseudo-procedures and
6583 then replace all occurrences with calls to the newly created
6584 subroutine. It is kind of an opposite of @option{-finline-functions}.
6585 This optimization runs at RTL level.
6587 @item -fsignaling-nans
6588 @opindex fsignaling-nans
6589 Compile code assuming that IEEE signaling NaNs may generate user-visible
6590 traps during floating-point operations. Setting this option disables
6591 optimizations that may change the number of exceptions visible with
6592 signaling NaNs. This option implies @option{-ftrapping-math}.
6594 This option causes the preprocessor macro @code{__SUPPORT_SNAN__} to
6597 The default is @option{-fno-signaling-nans}.
6599 This option is experimental and does not currently guarantee to
6600 disable all GCC optimizations that affect signaling NaN behavior.
6602 @item -fsingle-precision-constant
6603 @opindex fsingle-precision-constant
6604 Treat floating point constant as single precision constant instead of
6605 implicitly converting it to double precision constant.
6607 @item -fcx-limited-range
6608 @opindex fcx-limited-range
6609 When enabled, this option states that a range reduction step is not
6610 needed when performing complex division. Also, there is no checking
6611 whether the result of a complex multiplication or division is @code{NaN
6612 + I*NaN}, with an attempt to rescue the situation in that case. The
6613 default is @option{-fno-cx-limited-range}, but is enabled by
6614 @option{-ffast-math}.
6616 This option controls the default setting of the ISO C99
6617 @code{CX_LIMITED_RANGE} pragma. Nevertheless, the option applies to
6620 @item -fcx-fortran-rules
6621 @opindex fcx-fortran-rules
6622 Complex multiplication and division follow Fortran rules. Range
6623 reduction is done as part of complex division, but there is no checking
6624 whether the result of a complex multiplication or division is @code{NaN
6625 + I*NaN}, with an attempt to rescue the situation in that case.
6627 The default is @option{-fno-cx-fortran-rules}.
6631 The following options control optimizations that may improve
6632 performance, but are not enabled by any @option{-O} options. This
6633 section includes experimental options that may produce broken code.
6636 @item -fbranch-probabilities
6637 @opindex fbranch-probabilities
6638 After running a program compiled with @option{-fprofile-arcs}
6639 (@pxref{Debugging Options,, Options for Debugging Your Program or
6640 @command{gcc}}), you can compile it a second time using
6641 @option{-fbranch-probabilities}, to improve optimizations based on
6642 the number of times each branch was taken. When the program
6643 compiled with @option{-fprofile-arcs} exits it saves arc execution
6644 counts to a file called @file{@var{sourcename}.gcda} for each source
6645 file. The information in this data file is very dependent on the
6646 structure of the generated code, so you must use the same source code
6647 and the same optimization options for both compilations.
6649 With @option{-fbranch-probabilities}, GCC puts a
6650 @samp{REG_BR_PROB} note on each @samp{JUMP_INSN} and @samp{CALL_INSN}.
6651 These can be used to improve optimization. Currently, they are only
6652 used in one place: in @file{reorg.c}, instead of guessing which path a
6653 branch is mostly to take, the @samp{REG_BR_PROB} values are used to
6654 exactly determine which path is taken more often.
6656 @item -fprofile-values
6657 @opindex fprofile-values
6658 If combined with @option{-fprofile-arcs}, it adds code so that some
6659 data about values of expressions in the program is gathered.
6661 With @option{-fbranch-probabilities}, it reads back the data gathered
6662 from profiling values of expressions and adds @samp{REG_VALUE_PROFILE}
6663 notes to instructions for their later usage in optimizations.
6665 Enabled with @option{-fprofile-generate} and @option{-fprofile-use}.
6669 If combined with @option{-fprofile-arcs}, it instructs the compiler to add
6670 a code to gather information about values of expressions.
6672 With @option{-fbranch-probabilities}, it reads back the data gathered
6673 and actually performs the optimizations based on them.
6674 Currently the optimizations include specialization of division operation
6675 using the knowledge about the value of the denominator.
6677 @item -frename-registers
6678 @opindex frename-registers
6679 Attempt to avoid false dependencies in scheduled code by making use
6680 of registers left over after register allocation. This optimization
6681 will most benefit processors with lots of registers. Depending on the
6682 debug information format adopted by the target, however, it can
6683 make debugging impossible, since variables will no longer stay in
6684 a ``home register''.
6686 Enabled by default with @option{-funroll-loops}.
6690 Perform tail duplication to enlarge superblock size. This transformation
6691 simplifies the control flow of the function allowing other optimizations to do
6694 Enabled with @option{-fprofile-use}.
6696 @item -funroll-loops
6697 @opindex funroll-loops
6698 Unroll loops whose number of iterations can be determined at compile time or
6699 upon entry to the loop. @option{-funroll-loops} implies
6700 @option{-frerun-cse-after-loop}, @option{-fweb} and @option{-frename-registers}.
6701 It also turns on complete loop peeling (i.e.@: complete removal of loops with
6702 small constant number of iterations). This option makes code larger, and may
6703 or may not make it run faster.
6705 Enabled with @option{-fprofile-use}.
6707 @item -funroll-all-loops
6708 @opindex funroll-all-loops
6709 Unroll all loops, even if their number of iterations is uncertain when
6710 the loop is entered. This usually makes programs run more slowly.
6711 @option{-funroll-all-loops} implies the same options as
6712 @option{-funroll-loops}.
6715 @opindex fpeel-loops
6716 Peels the loops for that there is enough information that they do not
6717 roll much (from profile feedback). It also turns on complete loop peeling
6718 (i.e.@: complete removal of loops with small constant number of iterations).
6720 Enabled with @option{-fprofile-use}.
6722 @item -fmove-loop-invariants
6723 @opindex fmove-loop-invariants
6724 Enables the loop invariant motion pass in the RTL loop optimizer. Enabled
6725 at level @option{-O1}
6727 @item -funswitch-loops
6728 @opindex funswitch-loops
6729 Move branches with loop invariant conditions out of the loop, with duplicates
6730 of the loop on both branches (modified according to result of the condition).
6732 @item -ffunction-sections
6733 @itemx -fdata-sections
6734 @opindex ffunction-sections
6735 @opindex fdata-sections
6736 Place each function or data item into its own section in the output
6737 file if the target supports arbitrary sections. The name of the
6738 function or the name of the data item determines the section's name
6741 Use these options on systems where the linker can perform optimizations
6742 to improve locality of reference in the instruction space. Most systems
6743 using the ELF object format and SPARC processors running Solaris 2 have
6744 linkers with such optimizations. AIX may have these optimizations in
6747 Only use these options when there are significant benefits from doing
6748 so. When you specify these options, the assembler and linker will
6749 create larger object and executable files and will also be slower.
6750 You will not be able to use @code{gprof} on all systems if you
6751 specify this option and you may have problems with debugging if
6752 you specify both this option and @option{-g}.
6754 @item -fbranch-target-load-optimize
6755 @opindex fbranch-target-load-optimize
6756 Perform branch target register load optimization before prologue / epilogue
6758 The use of target registers can typically be exposed only during reload,
6759 thus hoisting loads out of loops and doing inter-block scheduling needs
6760 a separate optimization pass.
6762 @item -fbranch-target-load-optimize2
6763 @opindex fbranch-target-load-optimize2
6764 Perform branch target register load optimization after prologue / epilogue
6767 @item -fbtr-bb-exclusive
6768 @opindex fbtr-bb-exclusive
6769 When performing branch target register load optimization, don't reuse
6770 branch target registers in within any basic block.
6772 @item -fstack-protector
6773 @opindex fstack-protector
6774 Emit extra code to check for buffer overflows, such as stack smashing
6775 attacks. This is done by adding a guard variable to functions with
6776 vulnerable objects. This includes functions that call alloca, and
6777 functions with buffers larger than 8 bytes. The guards are initialized
6778 when a function is entered and then checked when the function exits.
6779 If a guard check fails, an error message is printed and the program exits.
6781 @item -fstack-protector-all
6782 @opindex fstack-protector-all
6783 Like @option{-fstack-protector} except that all functions are protected.
6785 @item -fsection-anchors
6786 @opindex fsection-anchors
6787 Try to reduce the number of symbolic address calculations by using
6788 shared ``anchor'' symbols to address nearby objects. This transformation
6789 can help to reduce the number of GOT entries and GOT accesses on some
6792 For example, the implementation of the following function @code{foo}:
6796 int foo (void) @{ return a + b + c; @}
6799 would usually calculate the addresses of all three variables, but if you
6800 compile it with @option{-fsection-anchors}, it will access the variables
6801 from a common anchor point instead. The effect is similar to the
6802 following pseudocode (which isn't valid C):
6807 register int *xr = &x;
6808 return xr[&a - &x] + xr[&b - &x] + xr[&c - &x];
6812 Not all targets support this option.
6814 @item --param @var{name}=@var{value}
6816 In some places, GCC uses various constants to control the amount of
6817 optimization that is done. For example, GCC will not inline functions
6818 that contain more that a certain number of instructions. You can
6819 control some of these constants on the command-line using the
6820 @option{--param} option.
6822 The names of specific parameters, and the meaning of the values, are
6823 tied to the internals of the compiler, and are subject to change
6824 without notice in future releases.
6826 In each case, the @var{value} is an integer. The allowable choices for
6827 @var{name} are given in the following table:
6830 @item sra-max-structure-size
6831 The maximum structure size, in bytes, at which the scalar replacement
6832 of aggregates (SRA) optimization will perform block copies. The
6833 default value, 0, implies that GCC will select the most appropriate
6836 @item sra-field-structure-ratio
6837 The threshold ratio (as a percentage) between instantiated fields and
6838 the complete structure size. We say that if the ratio of the number
6839 of bytes in instantiated fields to the number of bytes in the complete
6840 structure exceeds this parameter, then block copies are not used. The
6843 @item struct-reorg-cold-struct-ratio
6844 The threshold ratio (as a percentage) between a structure frequency
6845 and the frequency of the hottest structure in the program. This parameter
6846 is used by struct-reorg optimization enabled by @option{-fipa-struct-reorg}.
6847 We say that if the ratio of a structure frequency, calculated by profiling,
6848 to the hottest structure frequency in the program is less than this
6849 parameter, then structure reorganization is not applied to this structure.
6852 @item max-crossjump-edges
6853 The maximum number of incoming edges to consider for crossjumping.
6854 The algorithm used by @option{-fcrossjumping} is @math{O(N^2)} in
6855 the number of edges incoming to each block. Increasing values mean
6856 more aggressive optimization, making the compile time increase with
6857 probably small improvement in executable size.
6859 @item min-crossjump-insns
6860 The minimum number of instructions which must be matched at the end
6861 of two blocks before crossjumping will be performed on them. This
6862 value is ignored in the case where all instructions in the block being
6863 crossjumped from are matched. The default value is 5.
6865 @item max-grow-copy-bb-insns
6866 The maximum code size expansion factor when copying basic blocks
6867 instead of jumping. The expansion is relative to a jump instruction.
6868 The default value is 8.
6870 @item max-goto-duplication-insns
6871 The maximum number of instructions to duplicate to a block that jumps
6872 to a computed goto. To avoid @math{O(N^2)} behavior in a number of
6873 passes, GCC factors computed gotos early in the compilation process,
6874 and unfactors them as late as possible. Only computed jumps at the
6875 end of a basic blocks with no more than max-goto-duplication-insns are
6876 unfactored. The default value is 8.
6878 @item max-delay-slot-insn-search
6879 The maximum number of instructions to consider when looking for an
6880 instruction to fill a delay slot. If more than this arbitrary number of
6881 instructions is searched, the time savings from filling the delay slot
6882 will be minimal so stop searching. Increasing values mean more
6883 aggressive optimization, making the compile time increase with probably
6884 small improvement in executable run time.
6886 @item max-delay-slot-live-search
6887 When trying to fill delay slots, the maximum number of instructions to
6888 consider when searching for a block with valid live register
6889 information. Increasing this arbitrarily chosen value means more
6890 aggressive optimization, increasing the compile time. This parameter
6891 should be removed when the delay slot code is rewritten to maintain the
6894 @item max-gcse-memory
6895 The approximate maximum amount of memory that will be allocated in
6896 order to perform the global common subexpression elimination
6897 optimization. If more memory than specified is required, the
6898 optimization will not be done.
6900 @item max-gcse-passes
6901 The maximum number of passes of GCSE to run. The default is 1.
6903 @item max-pending-list-length
6904 The maximum number of pending dependencies scheduling will allow
6905 before flushing the current state and starting over. Large functions
6906 with few branches or calls can create excessively large lists which
6907 needlessly consume memory and resources.
6909 @item max-inline-insns-single
6910 Several parameters control the tree inliner used in gcc.
6911 This number sets the maximum number of instructions (counted in GCC's
6912 internal representation) in a single function that the tree inliner
6913 will consider for inlining. This only affects functions declared
6914 inline and methods implemented in a class declaration (C++).
6915 The default value is 450.
6917 @item max-inline-insns-auto
6918 When you use @option{-finline-functions} (included in @option{-O3}),
6919 a lot of functions that would otherwise not be considered for inlining
6920 by the compiler will be investigated. To those functions, a different
6921 (more restrictive) limit compared to functions declared inline can
6923 The default value is 90.
6925 @item large-function-insns
6926 The limit specifying really large functions. For functions larger than this
6927 limit after inlining inlining is constrained by
6928 @option{--param large-function-growth}. This parameter is useful primarily
6929 to avoid extreme compilation time caused by non-linear algorithms used by the
6931 The default value is 2700.
6933 @item large-function-growth
6934 Specifies maximal growth of large function caused by inlining in percents.
6935 The default value is 100 which limits large function growth to 2.0 times
6938 @item large-unit-insns
6939 The limit specifying large translation unit. Growth caused by inlining of
6940 units larger than this limit is limited by @option{--param inline-unit-growth}.
6941 For small units this might be too tight (consider unit consisting of function A
6942 that is inline and B that just calls A three time. If B is small relative to
6943 A, the growth of unit is 300\% and yet such inlining is very sane. For very
6944 large units consisting of small inlineable functions however the overall unit
6945 growth limit is needed to avoid exponential explosion of code size. Thus for
6946 smaller units, the size is increased to @option{--param large-unit-insns}
6947 before applying @option{--param inline-unit-growth}. The default is 10000
6949 @item inline-unit-growth
6950 Specifies maximal overall growth of the compilation unit caused by inlining.
6951 The default value is 30 which limits unit growth to 1.3 times the original
6954 @item large-stack-frame
6955 The limit specifying large stack frames. While inlining the algorithm is trying
6956 to not grow past this limit too much. Default value is 256 bytes.
6958 @item large-stack-frame-growth
6959 Specifies maximal growth of large stack frames caused by inlining in percents.
6960 The default value is 1000 which limits large stack frame growth to 11 times
6963 @item max-inline-insns-recursive
6964 @itemx max-inline-insns-recursive-auto
6965 Specifies maximum number of instructions out-of-line copy of self recursive inline
6966 function can grow into by performing recursive inlining.
6968 For functions declared inline @option{--param max-inline-insns-recursive} is
6969 taken into account. For function not declared inline, recursive inlining
6970 happens only when @option{-finline-functions} (included in @option{-O3}) is
6971 enabled and @option{--param max-inline-insns-recursive-auto} is used. The
6972 default value is 450.
6974 @item max-inline-recursive-depth
6975 @itemx max-inline-recursive-depth-auto
6976 Specifies maximum recursion depth used by the recursive inlining.
6978 For functions declared inline @option{--param max-inline-recursive-depth} is
6979 taken into account. For function not declared inline, recursive inlining
6980 happens only when @option{-finline-functions} (included in @option{-O3}) is
6981 enabled and @option{--param max-inline-recursive-depth-auto} is used. The
6984 @item min-inline-recursive-probability
6985 Recursive inlining is profitable only for function having deep recursion
6986 in average and can hurt for function having little recursion depth by
6987 increasing the prologue size or complexity of function body to other
6990 When profile feedback is available (see @option{-fprofile-generate}) the actual
6991 recursion depth can be guessed from probability that function will recurse via
6992 given call expression. This parameter limits inlining only to call expression
6993 whose probability exceeds given threshold (in percents). The default value is
6996 @item inline-call-cost
6997 Specify cost of call instruction relative to simple arithmetics operations
6998 (having cost of 1). Increasing this cost disqualifies inlining of non-leaf
6999 functions and at the same time increases size of leaf function that is believed to
7000 reduce function size by being inlined. In effect it increases amount of
7001 inlining for code having large abstraction penalty (many functions that just
7002 pass the arguments to other functions) and decrease inlining for code with low
7003 abstraction penalty. The default value is 12.
7005 @item min-vect-loop-bound
7006 The minimum number of iterations under which a loop will not get vectorized
7007 when @option{-ftree-vectorize} is used. The number of iterations after
7008 vectorization needs to be greater than the value specified by this option
7009 to allow vectorization. The default value is 0.
7011 @item max-unrolled-insns
7012 The maximum number of instructions that a loop should have if that loop
7013 is unrolled, and if the loop is unrolled, it determines how many times
7014 the loop code is unrolled.
7016 @item max-average-unrolled-insns
7017 The maximum number of instructions biased by probabilities of their execution
7018 that a loop should have if that loop is unrolled, and if the loop is unrolled,
7019 it determines how many times the loop code is unrolled.
7021 @item max-unroll-times
7022 The maximum number of unrollings of a single loop.
7024 @item max-peeled-insns
7025 The maximum number of instructions that a loop should have if that loop
7026 is peeled, and if the loop is peeled, it determines how many times
7027 the loop code is peeled.
7029 @item max-peel-times
7030 The maximum number of peelings of a single loop.
7032 @item max-completely-peeled-insns
7033 The maximum number of insns of a completely peeled loop.
7035 @item max-completely-peel-times
7036 The maximum number of iterations of a loop to be suitable for complete peeling.
7038 @item max-unswitch-insns
7039 The maximum number of insns of an unswitched loop.
7041 @item max-unswitch-level
7042 The maximum number of branches unswitched in a single loop.
7045 The minimum cost of an expensive expression in the loop invariant motion.
7047 @item iv-consider-all-candidates-bound
7048 Bound on number of candidates for induction variables below that
7049 all candidates are considered for each use in induction variable
7050 optimizations. Only the most relevant candidates are considered
7051 if there are more candidates, to avoid quadratic time complexity.
7053 @item iv-max-considered-uses
7054 The induction variable optimizations give up on loops that contain more
7055 induction variable uses.
7057 @item iv-always-prune-cand-set-bound
7058 If number of candidates in the set is smaller than this value,
7059 we always try to remove unnecessary ivs from the set during its
7060 optimization when a new iv is added to the set.
7062 @item scev-max-expr-size
7063 Bound on size of expressions used in the scalar evolutions analyzer.
7064 Large expressions slow the analyzer.
7066 @item omega-max-vars
7067 The maximum number of variables in an Omega constraint system.
7068 The default value is 128.
7070 @item omega-max-geqs
7071 The maximum number of inequalities in an Omega constraint system.
7072 The default value is 256.
7075 The maximum number of equalities in an Omega constraint system.
7076 The default value is 128.
7078 @item omega-max-wild-cards
7079 The maximum number of wildcard variables that the Omega solver will
7080 be able to insert. The default value is 18.
7082 @item omega-hash-table-size
7083 The size of the hash table in the Omega solver. The default value is
7086 @item omega-max-keys
7087 The maximal number of keys used by the Omega solver. The default
7090 @item omega-eliminate-redundant-constraints
7091 When set to 1, use expensive methods to eliminate all redundant
7092 constraints. The default value is 0.
7094 @item vect-max-version-for-alignment-checks
7095 The maximum number of runtime checks that can be performed when
7096 doing loop versioning for alignment in the vectorizer. See option
7097 ftree-vect-loop-version for more information.
7099 @item vect-max-version-for-alias-checks
7100 The maximum number of runtime checks that can be performed when
7101 doing loop versioning for alias in the vectorizer. See option
7102 ftree-vect-loop-version for more information.
7104 @item max-iterations-to-track
7106 The maximum number of iterations of a loop the brute force algorithm
7107 for analysis of # of iterations of the loop tries to evaluate.
7109 @item hot-bb-count-fraction
7110 Select fraction of the maximal count of repetitions of basic block in program
7111 given basic block needs to have to be considered hot.
7113 @item hot-bb-frequency-fraction
7114 Select fraction of the maximal frequency of executions of basic block in
7115 function given basic block needs to have to be considered hot
7117 @item max-predicted-iterations
7118 The maximum number of loop iterations we predict statically. This is useful
7119 in cases where function contain single loop with known bound and other loop
7120 with unknown. We predict the known number of iterations correctly, while
7121 the unknown number of iterations average to roughly 10. This means that the
7122 loop without bounds would appear artificially cold relative to the other one.
7124 @item align-threshold
7126 Select fraction of the maximal frequency of executions of basic block in
7127 function given basic block will get aligned.
7129 @item align-loop-iterations
7131 A loop expected to iterate at lest the selected number of iterations will get
7134 @item tracer-dynamic-coverage
7135 @itemx tracer-dynamic-coverage-feedback
7137 This value is used to limit superblock formation once the given percentage of
7138 executed instructions is covered. This limits unnecessary code size
7141 The @option{tracer-dynamic-coverage-feedback} is used only when profile
7142 feedback is available. The real profiles (as opposed to statically estimated
7143 ones) are much less balanced allowing the threshold to be larger value.
7145 @item tracer-max-code-growth
7146 Stop tail duplication once code growth has reached given percentage. This is
7147 rather hokey argument, as most of the duplicates will be eliminated later in
7148 cross jumping, so it may be set to much higher values than is the desired code
7151 @item tracer-min-branch-ratio
7153 Stop reverse growth when the reverse probability of best edge is less than this
7154 threshold (in percent).
7156 @item tracer-min-branch-ratio
7157 @itemx tracer-min-branch-ratio-feedback
7159 Stop forward growth if the best edge do have probability lower than this
7162 Similarly to @option{tracer-dynamic-coverage} two values are present, one for
7163 compilation for profile feedback and one for compilation without. The value
7164 for compilation with profile feedback needs to be more conservative (higher) in
7165 order to make tracer effective.
7167 @item max-cse-path-length
7169 Maximum number of basic blocks on path that cse considers. The default is 10.
7172 The maximum instructions CSE process before flushing. The default is 1000.
7174 @item max-aliased-vops
7176 Maximum number of virtual operands per function allowed to represent
7177 aliases before triggering the alias partitioning heuristic. Alias
7178 partitioning reduces compile times and memory consumption needed for
7179 aliasing at the expense of precision loss in alias information. The
7180 default value for this parameter is 100 for -O1, 500 for -O2 and 1000
7183 Notice that if a function contains more memory statements than the
7184 value of this parameter, it is not really possible to achieve this
7185 reduction. In this case, the compiler will use the number of memory
7186 statements as the value for @option{max-aliased-vops}.
7188 @item avg-aliased-vops
7190 Average number of virtual operands per statement allowed to represent
7191 aliases before triggering the alias partitioning heuristic. This
7192 works in conjunction with @option{max-aliased-vops}. If a function
7193 contains more than @option{max-aliased-vops} virtual operators, then
7194 memory symbols will be grouped into memory partitions until either the
7195 total number of virtual operators is below @option{max-aliased-vops}
7196 or the average number of virtual operators per memory statement is
7197 below @option{avg-aliased-vops}. The default value for this parameter
7198 is 1 for -O1 and -O2, and 3 for -O3.
7200 @item ggc-min-expand
7202 GCC uses a garbage collector to manage its own memory allocation. This
7203 parameter specifies the minimum percentage by which the garbage
7204 collector's heap should be allowed to expand between collections.
7205 Tuning this may improve compilation speed; it has no effect on code
7208 The default is 30% + 70% * (RAM/1GB) with an upper bound of 100% when
7209 RAM >= 1GB@. If @code{getrlimit} is available, the notion of "RAM" is
7210 the smallest of actual RAM and @code{RLIMIT_DATA} or @code{RLIMIT_AS}. If
7211 GCC is not able to calculate RAM on a particular platform, the lower
7212 bound of 30% is used. Setting this parameter and
7213 @option{ggc-min-heapsize} to zero causes a full collection to occur at
7214 every opportunity. This is extremely slow, but can be useful for
7217 @item ggc-min-heapsize
7219 Minimum size of the garbage collector's heap before it begins bothering
7220 to collect garbage. The first collection occurs after the heap expands
7221 by @option{ggc-min-expand}% beyond @option{ggc-min-heapsize}. Again,
7222 tuning this may improve compilation speed, and has no effect on code
7225 The default is the smaller of RAM/8, RLIMIT_RSS, or a limit which
7226 tries to ensure that RLIMIT_DATA or RLIMIT_AS are not exceeded, but
7227 with a lower bound of 4096 (four megabytes) and an upper bound of
7228 131072 (128 megabytes). If GCC is not able to calculate RAM on a
7229 particular platform, the lower bound is used. Setting this parameter
7230 very large effectively disables garbage collection. Setting this
7231 parameter and @option{ggc-min-expand} to zero causes a full collection
7232 to occur at every opportunity.
7234 @item max-reload-search-insns
7235 The maximum number of instruction reload should look backward for equivalent
7236 register. Increasing values mean more aggressive optimization, making the
7237 compile time increase with probably slightly better performance. The default
7240 @item max-cselib-memory-locations
7241 The maximum number of memory locations cselib should take into account.
7242 Increasing values mean more aggressive optimization, making the compile time
7243 increase with probably slightly better performance. The default value is 500.
7245 @item reorder-blocks-duplicate
7246 @itemx reorder-blocks-duplicate-feedback
7248 Used by basic block reordering pass to decide whether to use unconditional
7249 branch or duplicate the code on its destination. Code is duplicated when its
7250 estimated size is smaller than this value multiplied by the estimated size of
7251 unconditional jump in the hot spots of the program.
7253 The @option{reorder-block-duplicate-feedback} is used only when profile
7254 feedback is available and may be set to higher values than
7255 @option{reorder-block-duplicate} since information about the hot spots is more
7258 @item max-sched-ready-insns
7259 The maximum number of instructions ready to be issued the scheduler should
7260 consider at any given time during the first scheduling pass. Increasing
7261 values mean more thorough searches, making the compilation time increase
7262 with probably little benefit. The default value is 100.
7264 @item max-sched-region-blocks
7265 The maximum number of blocks in a region to be considered for
7266 interblock scheduling. The default value is 10.
7268 @item max-sched-region-insns
7269 The maximum number of insns in a region to be considered for
7270 interblock scheduling. The default value is 100.
7273 The minimum probability (in percents) of reaching a source block
7274 for interblock speculative scheduling. The default value is 40.
7276 @item max-sched-extend-regions-iters
7277 The maximum number of iterations through CFG to extend regions.
7278 0 - disable region extension,
7279 N - do at most N iterations.
7280 The default value is 0.
7282 @item max-sched-insn-conflict-delay
7283 The maximum conflict delay for an insn to be considered for speculative motion.
7284 The default value is 3.
7286 @item sched-spec-prob-cutoff
7287 The minimal probability of speculation success (in percents), so that
7288 speculative insn will be scheduled.
7289 The default value is 40.
7291 @item max-last-value-rtl
7293 The maximum size measured as number of RTLs that can be recorded in an expression
7294 in combiner for a pseudo register as last known value of that register. The default
7297 @item integer-share-limit
7298 Small integer constants can use a shared data structure, reducing the
7299 compiler's memory usage and increasing its speed. This sets the maximum
7300 value of a shared integer constant. The default value is 256.
7302 @item min-virtual-mappings
7303 Specifies the minimum number of virtual mappings in the incremental
7304 SSA updater that should be registered to trigger the virtual mappings
7305 heuristic defined by virtual-mappings-ratio. The default value is
7308 @item virtual-mappings-ratio
7309 If the number of virtual mappings is virtual-mappings-ratio bigger
7310 than the number of virtual symbols to be updated, then the incremental
7311 SSA updater switches to a full update for those symbols. The default
7314 @item ssp-buffer-size
7315 The minimum size of buffers (i.e.@: arrays) that will receive stack smashing
7316 protection when @option{-fstack-protection} is used.
7318 @item max-jump-thread-duplication-stmts
7319 Maximum number of statements allowed in a block that needs to be
7320 duplicated when threading jumps.
7322 @item max-fields-for-field-sensitive
7323 Maximum number of fields in a structure we will treat in
7324 a field sensitive manner during pointer analysis. The default is zero
7325 for -O0, and -O1 and 100 for -Os, -O2, and -O3.
7327 @item prefetch-latency
7328 Estimate on average number of instructions that are executed before
7329 prefetch finishes. The distance we prefetch ahead is proportional
7330 to this constant. Increasing this number may also lead to less
7331 streams being prefetched (see @option{simultaneous-prefetches}).
7333 @item simultaneous-prefetches
7334 Maximum number of prefetches that can run at the same time.
7336 @item l1-cache-line-size
7337 The size of cache line in L1 cache, in bytes.
7340 The size of L1 cache, in kilobytes.
7343 The size of L2 cache, in kilobytes.
7345 @item use-canonical-types
7346 Whether the compiler should use the ``canonical'' type system. By
7347 default, this should always be 1, which uses a more efficient internal
7348 mechanism for comparing types in C++ and Objective-C++. However, if
7349 bugs in the canonical type system are causing compilation failures,
7350 set this value to 0 to disable canonical types.
7352 @item switch-conversion-max-branch-ratio
7353 Switch initialization conversion will refuse to create arrays that are
7354 bigger than @option{switch-conversion-max-branch-ratio} times the number of
7355 branches in the switch.
7357 @item max-partial-antic-length
7358 Maximum length of the partial antic set computed during the tree
7359 partial redundancy elimination optimization (@option{-ftree-pre}) when
7360 optimizing at @option{-O3} and above. For some sorts of source code
7361 the enhanced partial redundancy elimination optimization can run away,
7362 consuming all of the memory available on the host machine. This
7363 parameter sets a limit on the length of the sets that are computed,
7364 which prevents the runaway behaviour. Setting a value of 0 for
7365 this paramter will allow an unlimited set length.
7367 @item sccvn-max-scc-size
7368 Maximum size of a strongly connected component (SCC) during SCCVN
7369 processing. If this limit is hit, SCCVN processing for the whole
7370 function will not be done and optimizations depending on it will
7371 be disabled. The default maximum SCC size is 10000.
7376 @node Preprocessor Options
7377 @section Options Controlling the Preprocessor
7378 @cindex preprocessor options
7379 @cindex options, preprocessor
7381 These options control the C preprocessor, which is run on each C source
7382 file before actual compilation.
7384 If you use the @option{-E} option, nothing is done except preprocessing.
7385 Some of these options make sense only together with @option{-E} because
7386 they cause the preprocessor output to be unsuitable for actual
7391 You can use @option{-Wp,@var{option}} to bypass the compiler driver
7392 and pass @var{option} directly through to the preprocessor. If
7393 @var{option} contains commas, it is split into multiple options at the
7394 commas. However, many options are modified, translated or interpreted
7395 by the compiler driver before being passed to the preprocessor, and
7396 @option{-Wp} forcibly bypasses this phase. The preprocessor's direct
7397 interface is undocumented and subject to change, so whenever possible
7398 you should avoid using @option{-Wp} and let the driver handle the
7401 @item -Xpreprocessor @var{option}
7402 @opindex preprocessor
7403 Pass @var{option} as an option to the preprocessor. You can use this to
7404 supply system-specific preprocessor options which GCC does not know how to
7407 If you want to pass an option that takes an argument, you must use
7408 @option{-Xpreprocessor} twice, once for the option and once for the argument.
7411 @include cppopts.texi
7413 @node Assembler Options
7414 @section Passing Options to the Assembler
7416 @c prevent bad page break with this line
7417 You can pass options to the assembler.
7420 @item -Wa,@var{option}
7422 Pass @var{option} as an option to the assembler. If @var{option}
7423 contains commas, it is split into multiple options at the commas.
7425 @item -Xassembler @var{option}
7427 Pass @var{option} as an option to the assembler. You can use this to
7428 supply system-specific assembler options which GCC does not know how to
7431 If you want to pass an option that takes an argument, you must use
7432 @option{-Xassembler} twice, once for the option and once for the argument.
7437 @section Options for Linking
7438 @cindex link options
7439 @cindex options, linking
7441 These options come into play when the compiler links object files into
7442 an executable output file. They are meaningless if the compiler is
7443 not doing a link step.
7447 @item @var{object-file-name}
7448 A file name that does not end in a special recognized suffix is
7449 considered to name an object file or library. (Object files are
7450 distinguished from libraries by the linker according to the file
7451 contents.) If linking is done, these object files are used as input
7460 If any of these options is used, then the linker is not run, and
7461 object file names should not be used as arguments. @xref{Overall
7465 @item -l@var{library}
7466 @itemx -l @var{library}
7468 Search the library named @var{library} when linking. (The second
7469 alternative with the library as a separate argument is only for
7470 POSIX compliance and is not recommended.)
7472 It makes a difference where in the command you write this option; the
7473 linker searches and processes libraries and object files in the order they
7474 are specified. Thus, @samp{foo.o -lz bar.o} searches library @samp{z}
7475 after file @file{foo.o} but before @file{bar.o}. If @file{bar.o} refers
7476 to functions in @samp{z}, those functions may not be loaded.
7478 The linker searches a standard list of directories for the library,
7479 which is actually a file named @file{lib@var{library}.a}. The linker
7480 then uses this file as if it had been specified precisely by name.
7482 The directories searched include several standard system directories
7483 plus any that you specify with @option{-L}.
7485 Normally the files found this way are library files---archive files
7486 whose members are object files. The linker handles an archive file by
7487 scanning through it for members which define symbols that have so far
7488 been referenced but not defined. But if the file that is found is an
7489 ordinary object file, it is linked in the usual fashion. The only
7490 difference between using an @option{-l} option and specifying a file name
7491 is that @option{-l} surrounds @var{library} with @samp{lib} and @samp{.a}
7492 and searches several directories.
7496 You need this special case of the @option{-l} option in order to
7497 link an Objective-C or Objective-C++ program.
7500 @opindex nostartfiles
7501 Do not use the standard system startup files when linking.
7502 The standard system libraries are used normally, unless @option{-nostdlib}
7503 or @option{-nodefaultlibs} is used.
7505 @item -nodefaultlibs
7506 @opindex nodefaultlibs
7507 Do not use the standard system libraries when linking.
7508 Only the libraries you specify will be passed to the linker.
7509 The standard startup files are used normally, unless @option{-nostartfiles}
7510 is used. The compiler may generate calls to @code{memcmp},
7511 @code{memset}, @code{memcpy} and @code{memmove}.
7512 These entries are usually resolved by entries in
7513 libc. These entry points should be supplied through some other
7514 mechanism when this option is specified.
7518 Do not use the standard system startup files or libraries when linking.
7519 No startup files and only the libraries you specify will be passed to
7520 the linker. The compiler may generate calls to @code{memcmp}, @code{memset},
7521 @code{memcpy} and @code{memmove}.
7522 These entries are usually resolved by entries in
7523 libc. These entry points should be supplied through some other
7524 mechanism when this option is specified.
7526 @cindex @option{-lgcc}, use with @option{-nostdlib}
7527 @cindex @option{-nostdlib} and unresolved references
7528 @cindex unresolved references and @option{-nostdlib}
7529 @cindex @option{-lgcc}, use with @option{-nodefaultlibs}
7530 @cindex @option{-nodefaultlibs} and unresolved references
7531 @cindex unresolved references and @option{-nodefaultlibs}
7532 One of the standard libraries bypassed by @option{-nostdlib} and
7533 @option{-nodefaultlibs} is @file{libgcc.a}, a library of internal subroutines
7534 that GCC uses to overcome shortcomings of particular machines, or special
7535 needs for some languages.
7536 (@xref{Interface,,Interfacing to GCC Output,gccint,GNU Compiler
7537 Collection (GCC) Internals},
7538 for more discussion of @file{libgcc.a}.)
7539 In most cases, you need @file{libgcc.a} even when you want to avoid
7540 other standard libraries. In other words, when you specify @option{-nostdlib}
7541 or @option{-nodefaultlibs} you should usually specify @option{-lgcc} as well.
7542 This ensures that you have no unresolved references to internal GCC
7543 library subroutines. (For example, @samp{__main}, used to ensure C++
7544 constructors will be called; @pxref{Collect2,,@code{collect2}, gccint,
7545 GNU Compiler Collection (GCC) Internals}.)
7549 Produce a position independent executable on targets which support it.
7550 For predictable results, you must also specify the same set of options
7551 that were used to generate code (@option{-fpie}, @option{-fPIE},
7552 or model suboptions) when you specify this option.
7556 Pass the flag @option{-export-dynamic} to the ELF linker, on targets
7557 that support it. This instructs the linker to add all symbols, not
7558 only used ones, to the dynamic symbol table. This option is needed
7559 for some uses of @code{dlopen} or to allow obtaining backtraces
7560 from within a program.
7564 Remove all symbol table and relocation information from the executable.
7568 On systems that support dynamic linking, this prevents linking with the shared
7569 libraries. On other systems, this option has no effect.
7573 Produce a shared object which can then be linked with other objects to
7574 form an executable. Not all systems support this option. For predictable
7575 results, you must also specify the same set of options that were used to
7576 generate code (@option{-fpic}, @option{-fPIC}, or model suboptions)
7577 when you specify this option.@footnote{On some systems, @samp{gcc -shared}
7578 needs to build supplementary stub code for constructors to work. On
7579 multi-libbed systems, @samp{gcc -shared} must select the correct support
7580 libraries to link against. Failing to supply the correct flags may lead
7581 to subtle defects. Supplying them in cases where they are not necessary
7584 @item -shared-libgcc
7585 @itemx -static-libgcc
7586 @opindex shared-libgcc
7587 @opindex static-libgcc
7588 On systems that provide @file{libgcc} as a shared library, these options
7589 force the use of either the shared or static version respectively.
7590 If no shared version of @file{libgcc} was built when the compiler was
7591 configured, these options have no effect.
7593 There are several situations in which an application should use the
7594 shared @file{libgcc} instead of the static version. The most common
7595 of these is when the application wishes to throw and catch exceptions
7596 across different shared libraries. In that case, each of the libraries
7597 as well as the application itself should use the shared @file{libgcc}.
7599 Therefore, the G++ and GCJ drivers automatically add
7600 @option{-shared-libgcc} whenever you build a shared library or a main
7601 executable, because C++ and Java programs typically use exceptions, so
7602 this is the right thing to do.
7604 If, instead, you use the GCC driver to create shared libraries, you may
7605 find that they will not always be linked with the shared @file{libgcc}.
7606 If GCC finds, at its configuration time, that you have a non-GNU linker
7607 or a GNU linker that does not support option @option{--eh-frame-hdr},
7608 it will link the shared version of @file{libgcc} into shared libraries
7609 by default. Otherwise, it will take advantage of the linker and optimize
7610 away the linking with the shared version of @file{libgcc}, linking with
7611 the static version of libgcc by default. This allows exceptions to
7612 propagate through such shared libraries, without incurring relocation
7613 costs at library load time.
7615 However, if a library or main executable is supposed to throw or catch
7616 exceptions, you must link it using the G++ or GCJ driver, as appropriate
7617 for the languages used in the program, or using the option
7618 @option{-shared-libgcc}, such that it is linked with the shared
7623 Bind references to global symbols when building a shared object. Warn
7624 about any unresolved references (unless overridden by the link editor
7625 option @samp{-Xlinker -z -Xlinker defs}). Only a few systems support
7628 @item -Xlinker @var{option}
7630 Pass @var{option} as an option to the linker. You can use this to
7631 supply system-specific linker options which GCC does not know how to
7634 If you want to pass an option that takes an argument, you must use
7635 @option{-Xlinker} twice, once for the option and once for the argument.
7636 For example, to pass @option{-assert definitions}, you must write
7637 @samp{-Xlinker -assert -Xlinker definitions}. It does not work to write
7638 @option{-Xlinker "-assert definitions"}, because this passes the entire
7639 string as a single argument, which is not what the linker expects.
7641 @item -Wl,@var{option}
7643 Pass @var{option} as an option to the linker. If @var{option} contains
7644 commas, it is split into multiple options at the commas.
7646 @item -u @var{symbol}
7648 Pretend the symbol @var{symbol} is undefined, to force linking of
7649 library modules to define it. You can use @option{-u} multiple times with
7650 different symbols to force loading of additional library modules.
7653 @node Directory Options
7654 @section Options for Directory Search
7655 @cindex directory options
7656 @cindex options, directory search
7659 These options specify directories to search for header files, for
7660 libraries and for parts of the compiler:
7665 Add the directory @var{dir} to the head of the list of directories to be
7666 searched for header files. This can be used to override a system header
7667 file, substituting your own version, since these directories are
7668 searched before the system header file directories. However, you should
7669 not use this option to add directories that contain vendor-supplied
7670 system header files (use @option{-isystem} for that). If you use more than
7671 one @option{-I} option, the directories are scanned in left-to-right
7672 order; the standard system directories come after.
7674 If a standard system include directory, or a directory specified with
7675 @option{-isystem}, is also specified with @option{-I}, the @option{-I}
7676 option will be ignored. The directory will still be searched but as a
7677 system directory at its normal position in the system include chain.
7678 This is to ensure that GCC's procedure to fix buggy system headers and
7679 the ordering for the include_next directive are not inadvertently changed.
7680 If you really need to change the search order for system directories,
7681 use the @option{-nostdinc} and/or @option{-isystem} options.
7683 @item -iquote@var{dir}
7685 Add the directory @var{dir} to the head of the list of directories to
7686 be searched for header files only for the case of @samp{#include
7687 "@var{file}"}; they are not searched for @samp{#include <@var{file}>},
7688 otherwise just like @option{-I}.
7692 Add directory @var{dir} to the list of directories to be searched
7695 @item -B@var{prefix}
7697 This option specifies where to find the executables, libraries,
7698 include files, and data files of the compiler itself.
7700 The compiler driver program runs one or more of the subprograms
7701 @file{cpp}, @file{cc1}, @file{as} and @file{ld}. It tries
7702 @var{prefix} as a prefix for each program it tries to run, both with and
7703 without @samp{@var{machine}/@var{version}/} (@pxref{Target Options}).
7705 For each subprogram to be run, the compiler driver first tries the
7706 @option{-B} prefix, if any. If that name is not found, or if @option{-B}
7707 was not specified, the driver tries two standard prefixes, which are
7708 @file{/usr/lib/gcc/} and @file{/usr/local/lib/gcc/}. If neither of
7709 those results in a file name that is found, the unmodified program
7710 name is searched for using the directories specified in your
7711 @env{PATH} environment variable.
7713 The compiler will check to see if the path provided by the @option{-B}
7714 refers to a directory, and if necessary it will add a directory
7715 separator character at the end of the path.
7717 @option{-B} prefixes that effectively specify directory names also apply
7718 to libraries in the linker, because the compiler translates these
7719 options into @option{-L} options for the linker. They also apply to
7720 includes files in the preprocessor, because the compiler translates these
7721 options into @option{-isystem} options for the preprocessor. In this case,
7722 the compiler appends @samp{include} to the prefix.
7724 The run-time support file @file{libgcc.a} can also be searched for using
7725 the @option{-B} prefix, if needed. If it is not found there, the two
7726 standard prefixes above are tried, and that is all. The file is left
7727 out of the link if it is not found by those means.
7729 Another way to specify a prefix much like the @option{-B} prefix is to use
7730 the environment variable @env{GCC_EXEC_PREFIX}. @xref{Environment
7733 As a special kludge, if the path provided by @option{-B} is
7734 @file{[dir/]stage@var{N}/}, where @var{N} is a number in the range 0 to
7735 9, then it will be replaced by @file{[dir/]include}. This is to help
7736 with boot-strapping the compiler.
7738 @item -specs=@var{file}
7740 Process @var{file} after the compiler reads in the standard @file{specs}
7741 file, in order to override the defaults that the @file{gcc} driver
7742 program uses when determining what switches to pass to @file{cc1},
7743 @file{cc1plus}, @file{as}, @file{ld}, etc. More than one
7744 @option{-specs=@var{file}} can be specified on the command line, and they
7745 are processed in order, from left to right.
7747 @item --sysroot=@var{dir}
7749 Use @var{dir} as the logical root directory for headers and libraries.
7750 For example, if the compiler would normally search for headers in
7751 @file{/usr/include} and libraries in @file{/usr/lib}, it will instead
7752 search @file{@var{dir}/usr/include} and @file{@var{dir}/usr/lib}.
7754 If you use both this option and the @option{-isysroot} option, then
7755 the @option{--sysroot} option will apply to libraries, but the
7756 @option{-isysroot} option will apply to header files.
7758 The GNU linker (beginning with version 2.16) has the necessary support
7759 for this option. If your linker does not support this option, the
7760 header file aspect of @option{--sysroot} will still work, but the
7761 library aspect will not.
7765 This option has been deprecated. Please use @option{-iquote} instead for
7766 @option{-I} directories before the @option{-I-} and remove the @option{-I-}.
7767 Any directories you specify with @option{-I} options before the @option{-I-}
7768 option are searched only for the case of @samp{#include "@var{file}"};
7769 they are not searched for @samp{#include <@var{file}>}.
7771 If additional directories are specified with @option{-I} options after
7772 the @option{-I-}, these directories are searched for all @samp{#include}
7773 directives. (Ordinarily @emph{all} @option{-I} directories are used
7776 In addition, the @option{-I-} option inhibits the use of the current
7777 directory (where the current input file came from) as the first search
7778 directory for @samp{#include "@var{file}"}. There is no way to
7779 override this effect of @option{-I-}. With @option{-I.} you can specify
7780 searching the directory which was current when the compiler was
7781 invoked. That is not exactly the same as what the preprocessor does
7782 by default, but it is often satisfactory.
7784 @option{-I-} does not inhibit the use of the standard system directories
7785 for header files. Thus, @option{-I-} and @option{-nostdinc} are
7792 @section Specifying subprocesses and the switches to pass to them
7795 @command{gcc} is a driver program. It performs its job by invoking a
7796 sequence of other programs to do the work of compiling, assembling and
7797 linking. GCC interprets its command-line parameters and uses these to
7798 deduce which programs it should invoke, and which command-line options
7799 it ought to place on their command lines. This behavior is controlled
7800 by @dfn{spec strings}. In most cases there is one spec string for each
7801 program that GCC can invoke, but a few programs have multiple spec
7802 strings to control their behavior. The spec strings built into GCC can
7803 be overridden by using the @option{-specs=} command-line switch to specify
7806 @dfn{Spec files} are plaintext files that are used to construct spec
7807 strings. They consist of a sequence of directives separated by blank
7808 lines. The type of directive is determined by the first non-whitespace
7809 character on the line and it can be one of the following:
7812 @item %@var{command}
7813 Issues a @var{command} to the spec file processor. The commands that can
7817 @item %include <@var{file}>
7819 Search for @var{file} and insert its text at the current point in the
7822 @item %include_noerr <@var{file}>
7823 @cindex %include_noerr
7824 Just like @samp{%include}, but do not generate an error message if the include
7825 file cannot be found.
7827 @item %rename @var{old_name} @var{new_name}
7829 Rename the spec string @var{old_name} to @var{new_name}.
7833 @item *[@var{spec_name}]:
7834 This tells the compiler to create, override or delete the named spec
7835 string. All lines after this directive up to the next directive or
7836 blank line are considered to be the text for the spec string. If this
7837 results in an empty string then the spec will be deleted. (Or, if the
7838 spec did not exist, then nothing will happened.) Otherwise, if the spec
7839 does not currently exist a new spec will be created. If the spec does
7840 exist then its contents will be overridden by the text of this
7841 directive, unless the first character of that text is the @samp{+}
7842 character, in which case the text will be appended to the spec.
7844 @item [@var{suffix}]:
7845 Creates a new @samp{[@var{suffix}] spec} pair. All lines after this directive
7846 and up to the next directive or blank line are considered to make up the
7847 spec string for the indicated suffix. When the compiler encounters an
7848 input file with the named suffix, it will processes the spec string in
7849 order to work out how to compile that file. For example:
7856 This says that any input file whose name ends in @samp{.ZZ} should be
7857 passed to the program @samp{z-compile}, which should be invoked with the
7858 command-line switch @option{-input} and with the result of performing the
7859 @samp{%i} substitution. (See below.)
7861 As an alternative to providing a spec string, the text that follows a
7862 suffix directive can be one of the following:
7865 @item @@@var{language}
7866 This says that the suffix is an alias for a known @var{language}. This is
7867 similar to using the @option{-x} command-line switch to GCC to specify a
7868 language explicitly. For example:
7875 Says that .ZZ files are, in fact, C++ source files.
7878 This causes an error messages saying:
7881 @var{name} compiler not installed on this system.
7885 GCC already has an extensive list of suffixes built into it.
7886 This directive will add an entry to the end of the list of suffixes, but
7887 since the list is searched from the end backwards, it is effectively
7888 possible to override earlier entries using this technique.
7892 GCC has the following spec strings built into it. Spec files can
7893 override these strings or create their own. Note that individual
7894 targets can also add their own spec strings to this list.
7897 asm Options to pass to the assembler
7898 asm_final Options to pass to the assembler post-processor
7899 cpp Options to pass to the C preprocessor
7900 cc1 Options to pass to the C compiler
7901 cc1plus Options to pass to the C++ compiler
7902 endfile Object files to include at the end of the link
7903 link Options to pass to the linker
7904 lib Libraries to include on the command line to the linker
7905 libgcc Decides which GCC support library to pass to the linker
7906 linker Sets the name of the linker
7907 predefines Defines to be passed to the C preprocessor
7908 signed_char Defines to pass to CPP to say whether @code{char} is signed
7910 startfile Object files to include at the start of the link
7913 Here is a small example of a spec file:
7919 --start-group -lgcc -lc -leval1 --end-group %(old_lib)
7922 This example renames the spec called @samp{lib} to @samp{old_lib} and
7923 then overrides the previous definition of @samp{lib} with a new one.
7924 The new definition adds in some extra command-line options before
7925 including the text of the old definition.
7927 @dfn{Spec strings} are a list of command-line options to be passed to their
7928 corresponding program. In addition, the spec strings can contain
7929 @samp{%}-prefixed sequences to substitute variable text or to
7930 conditionally insert text into the command line. Using these constructs
7931 it is possible to generate quite complex command lines.
7933 Here is a table of all defined @samp{%}-sequences for spec
7934 strings. Note that spaces are not generated automatically around the
7935 results of expanding these sequences. Therefore you can concatenate them
7936 together or combine them with constant text in a single argument.
7940 Substitute one @samp{%} into the program name or argument.
7943 Substitute the name of the input file being processed.
7946 Substitute the basename of the input file being processed.
7947 This is the substring up to (and not including) the last period
7948 and not including the directory.
7951 This is the same as @samp{%b}, but include the file suffix (text after
7955 Marks the argument containing or following the @samp{%d} as a
7956 temporary file name, so that that file will be deleted if GCC exits
7957 successfully. Unlike @samp{%g}, this contributes no text to the
7960 @item %g@var{suffix}
7961 Substitute a file name that has suffix @var{suffix} and is chosen
7962 once per compilation, and mark the argument in the same way as
7963 @samp{%d}. To reduce exposure to denial-of-service attacks, the file
7964 name is now chosen in a way that is hard to predict even when previously
7965 chosen file names are known. For example, @samp{%g.s @dots{} %g.o @dots{} %g.s}
7966 might turn into @samp{ccUVUUAU.s ccXYAXZ12.o ccUVUUAU.s}. @var{suffix} matches
7967 the regexp @samp{[.A-Za-z]*} or the special string @samp{%O}, which is
7968 treated exactly as if @samp{%O} had been preprocessed. Previously, @samp{%g}
7969 was simply substituted with a file name chosen once per compilation,
7970 without regard to any appended suffix (which was therefore treated
7971 just like ordinary text), making such attacks more likely to succeed.
7973 @item %u@var{suffix}
7974 Like @samp{%g}, but generates a new temporary file name even if
7975 @samp{%u@var{suffix}} was already seen.
7977 @item %U@var{suffix}
7978 Substitutes the last file name generated with @samp{%u@var{suffix}}, generating a
7979 new one if there is no such last file name. In the absence of any
7980 @samp{%u@var{suffix}}, this is just like @samp{%g@var{suffix}}, except they don't share
7981 the same suffix @emph{space}, so @samp{%g.s @dots{} %U.s @dots{} %g.s @dots{} %U.s}
7982 would involve the generation of two distinct file names, one
7983 for each @samp{%g.s} and another for each @samp{%U.s}. Previously, @samp{%U} was
7984 simply substituted with a file name chosen for the previous @samp{%u},
7985 without regard to any appended suffix.
7987 @item %j@var{suffix}
7988 Substitutes the name of the @code{HOST_BIT_BUCKET}, if any, and if it is
7989 writable, and if save-temps is off; otherwise, substitute the name
7990 of a temporary file, just like @samp{%u}. This temporary file is not
7991 meant for communication between processes, but rather as a junk
7994 @item %|@var{suffix}
7995 @itemx %m@var{suffix}
7996 Like @samp{%g}, except if @option{-pipe} is in effect. In that case
7997 @samp{%|} substitutes a single dash and @samp{%m} substitutes nothing at
7998 all. These are the two most common ways to instruct a program that it
7999 should read from standard input or write to standard output. If you
8000 need something more elaborate you can use an @samp{%@{pipe:@code{X}@}}
8001 construct: see for example @file{f/lang-specs.h}.
8003 @item %.@var{SUFFIX}
8004 Substitutes @var{.SUFFIX} for the suffixes of a matched switch's args
8005 when it is subsequently output with @samp{%*}. @var{SUFFIX} is
8006 terminated by the next space or %.
8009 Marks the argument containing or following the @samp{%w} as the
8010 designated output file of this compilation. This puts the argument
8011 into the sequence of arguments that @samp{%o} will substitute later.
8014 Substitutes the names of all the output files, with spaces
8015 automatically placed around them. You should write spaces
8016 around the @samp{%o} as well or the results are undefined.
8017 @samp{%o} is for use in the specs for running the linker.
8018 Input files whose names have no recognized suffix are not compiled
8019 at all, but they are included among the output files, so they will
8023 Substitutes the suffix for object files. Note that this is
8024 handled specially when it immediately follows @samp{%g, %u, or %U},
8025 because of the need for those to form complete file names. The
8026 handling is such that @samp{%O} is treated exactly as if it had already
8027 been substituted, except that @samp{%g, %u, and %U} do not currently
8028 support additional @var{suffix} characters following @samp{%O} as they would
8029 following, for example, @samp{.o}.
8032 Substitutes the standard macro predefinitions for the
8033 current target machine. Use this when running @code{cpp}.
8036 Like @samp{%p}, but puts @samp{__} before and after the name of each
8037 predefined macro, except for macros that start with @samp{__} or with
8038 @samp{_@var{L}}, where @var{L} is an uppercase letter. This is for ISO
8042 Substitute any of @option{-iprefix} (made from @env{GCC_EXEC_PREFIX}),
8043 @option{-isysroot} (made from @env{TARGET_SYSTEM_ROOT}),
8044 @option{-isystem} (made from @env{COMPILER_PATH} and @option{-B} options)
8045 and @option{-imultilib} as necessary.
8048 Current argument is the name of a library or startup file of some sort.
8049 Search for that file in a standard list of directories and substitute
8050 the full name found.
8053 Print @var{str} as an error message. @var{str} is terminated by a newline.
8054 Use this when inconsistent options are detected.
8057 Substitute the contents of spec string @var{name} at this point.
8060 Like @samp{%(@dots{})} but put @samp{__} around @option{-D} arguments.
8062 @item %x@{@var{option}@}
8063 Accumulate an option for @samp{%X}.
8066 Output the accumulated linker options specified by @option{-Wl} or a @samp{%x}
8070 Output the accumulated assembler options specified by @option{-Wa}.
8073 Output the accumulated preprocessor options specified by @option{-Wp}.
8076 Process the @code{asm} spec. This is used to compute the
8077 switches to be passed to the assembler.
8080 Process the @code{asm_final} spec. This is a spec string for
8081 passing switches to an assembler post-processor, if such a program is
8085 Process the @code{link} spec. This is the spec for computing the
8086 command line passed to the linker. Typically it will make use of the
8087 @samp{%L %G %S %D and %E} sequences.
8090 Dump out a @option{-L} option for each directory that GCC believes might
8091 contain startup files. If the target supports multilibs then the
8092 current multilib directory will be prepended to each of these paths.
8095 Process the @code{lib} spec. This is a spec string for deciding which
8096 libraries should be included on the command line to the linker.
8099 Process the @code{libgcc} spec. This is a spec string for deciding
8100 which GCC support library should be included on the command line to the linker.
8103 Process the @code{startfile} spec. This is a spec for deciding which
8104 object files should be the first ones passed to the linker. Typically
8105 this might be a file named @file{crt0.o}.
8108 Process the @code{endfile} spec. This is a spec string that specifies
8109 the last object files that will be passed to the linker.
8112 Process the @code{cpp} spec. This is used to construct the arguments
8113 to be passed to the C preprocessor.
8116 Process the @code{cc1} spec. This is used to construct the options to be
8117 passed to the actual C compiler (@samp{cc1}).
8120 Process the @code{cc1plus} spec. This is used to construct the options to be
8121 passed to the actual C++ compiler (@samp{cc1plus}).
8124 Substitute the variable part of a matched option. See below.
8125 Note that each comma in the substituted string is replaced by
8129 Remove all occurrences of @code{-S} from the command line. Note---this
8130 command is position dependent. @samp{%} commands in the spec string
8131 before this one will see @code{-S}, @samp{%} commands in the spec string
8132 after this one will not.
8134 @item %:@var{function}(@var{args})
8135 Call the named function @var{function}, passing it @var{args}.
8136 @var{args} is first processed as a nested spec string, then split
8137 into an argument vector in the usual fashion. The function returns
8138 a string which is processed as if it had appeared literally as part
8139 of the current spec.
8141 The following built-in spec functions are provided:
8145 The @code{getenv} spec function takes two arguments: an environment
8146 variable name and a string. If the environment variable is not
8147 defined, a fatal error is issued. Otherwise, the return value is the
8148 value of the environment variable concatenated with the string. For
8149 example, if @env{TOPDIR} is defined as @file{/path/to/top}, then:
8152 %:getenv(TOPDIR /include)
8155 expands to @file{/path/to/top/include}.
8157 @item @code{if-exists}
8158 The @code{if-exists} spec function takes one argument, an absolute
8159 pathname to a file. If the file exists, @code{if-exists} returns the
8160 pathname. Here is a small example of its usage:
8164 crt0%O%s %:if-exists(crti%O%s) crtbegin%O%s
8167 @item @code{if-exists-else}
8168 The @code{if-exists-else} spec function is similar to the @code{if-exists}
8169 spec function, except that it takes two arguments. The first argument is
8170 an absolute pathname to a file. If the file exists, @code{if-exists-else}
8171 returns the pathname. If it does not exist, it returns the second argument.
8172 This way, @code{if-exists-else} can be used to select one file or another,
8173 based on the existence of the first. Here is a small example of its usage:
8177 crt0%O%s %:if-exists(crti%O%s) \
8178 %:if-exists-else(crtbeginT%O%s crtbegin%O%s)
8181 @item @code{replace-outfile}
8182 The @code{replace-outfile} spec function takes two arguments. It looks for the
8183 first argument in the outfiles array and replaces it with the second argument. Here
8184 is a small example of its usage:
8187 %@{fgnu-runtime:%:replace-outfile(-lobjc -lobjc-gnu)@}
8190 @item @code{print-asm-header}
8191 The @code{print-asm-header} function takes no arguments and simply
8192 prints a banner like:
8198 Use "-Wa,OPTION" to pass "OPTION" to the assembler.
8201 It is used to separate compiler options from assembler options
8202 in the @option{--target-help} output.
8206 Substitutes the @code{-S} switch, if that switch was given to GCC@.
8207 If that switch was not specified, this substitutes nothing. Note that
8208 the leading dash is omitted when specifying this option, and it is
8209 automatically inserted if the substitution is performed. Thus the spec
8210 string @samp{%@{foo@}} would match the command-line option @option{-foo}
8211 and would output the command line option @option{-foo}.
8213 @item %W@{@code{S}@}
8214 Like %@{@code{S}@} but mark last argument supplied within as a file to be
8217 @item %@{@code{S}*@}
8218 Substitutes all the switches specified to GCC whose names start
8219 with @code{-S}, but which also take an argument. This is used for
8220 switches like @option{-o}, @option{-D}, @option{-I}, etc.
8221 GCC considers @option{-o foo} as being
8222 one switch whose names starts with @samp{o}. %@{o*@} would substitute this
8223 text, including the space. Thus two arguments would be generated.
8225 @item %@{@code{S}*&@code{T}*@}
8226 Like %@{@code{S}*@}, but preserve order of @code{S} and @code{T} options
8227 (the order of @code{S} and @code{T} in the spec is not significant).
8228 There can be any number of ampersand-separated variables; for each the
8229 wild card is optional. Useful for CPP as @samp{%@{D*&U*&A*@}}.
8231 @item %@{@code{S}:@code{X}@}
8232 Substitutes @code{X}, if the @samp{-S} switch was given to GCC@.
8234 @item %@{!@code{S}:@code{X}@}
8235 Substitutes @code{X}, if the @samp{-S} switch was @emph{not} given to GCC@.
8237 @item %@{@code{S}*:@code{X}@}
8238 Substitutes @code{X} if one or more switches whose names start with
8239 @code{-S} are specified to GCC@. Normally @code{X} is substituted only
8240 once, no matter how many such switches appeared. However, if @code{%*}
8241 appears somewhere in @code{X}, then @code{X} will be substituted once
8242 for each matching switch, with the @code{%*} replaced by the part of
8243 that switch that matched the @code{*}.
8245 @item %@{.@code{S}:@code{X}@}
8246 Substitutes @code{X}, if processing a file with suffix @code{S}.
8248 @item %@{!.@code{S}:@code{X}@}
8249 Substitutes @code{X}, if @emph{not} processing a file with suffix @code{S}.
8251 @item %@{,@code{S}:@code{X}@}
8252 Substitutes @code{X}, if processing a file for language @code{S}.
8254 @item %@{!,@code{S}:@code{X}@}
8255 Substitutes @code{X}, if not processing a file for language @code{S}.
8257 @item %@{@code{S}|@code{P}:@code{X}@}
8258 Substitutes @code{X} if either @code{-S} or @code{-P} was given to
8259 GCC@. This may be combined with @samp{!}, @samp{.}, @samp{,}, and
8260 @code{*} sequences as well, although they have a stronger binding than
8261 the @samp{|}. If @code{%*} appears in @code{X}, all of the
8262 alternatives must be starred, and only the first matching alternative
8265 For example, a spec string like this:
8268 %@{.c:-foo@} %@{!.c:-bar@} %@{.c|d:-baz@} %@{!.c|d:-boggle@}
8271 will output the following command-line options from the following input
8272 command-line options:
8277 -d fred.c -foo -baz -boggle
8278 -d jim.d -bar -baz -boggle
8281 @item %@{S:X; T:Y; :D@}
8283 If @code{S} was given to GCC, substitutes @code{X}; else if @code{T} was
8284 given to GCC, substitutes @code{Y}; else substitutes @code{D}. There can
8285 be as many clauses as you need. This may be combined with @code{.},
8286 @code{,}, @code{!}, @code{|}, and @code{*} as needed.
8291 The conditional text @code{X} in a %@{@code{S}:@code{X}@} or similar
8292 construct may contain other nested @samp{%} constructs or spaces, or
8293 even newlines. They are processed as usual, as described above.
8294 Trailing white space in @code{X} is ignored. White space may also
8295 appear anywhere on the left side of the colon in these constructs,
8296 except between @code{.} or @code{*} and the corresponding word.
8298 The @option{-O}, @option{-f}, @option{-m}, and @option{-W} switches are
8299 handled specifically in these constructs. If another value of
8300 @option{-O} or the negated form of a @option{-f}, @option{-m}, or
8301 @option{-W} switch is found later in the command line, the earlier
8302 switch value is ignored, except with @{@code{S}*@} where @code{S} is
8303 just one letter, which passes all matching options.
8305 The character @samp{|} at the beginning of the predicate text is used to
8306 indicate that a command should be piped to the following command, but
8307 only if @option{-pipe} is specified.
8309 It is built into GCC which switches take arguments and which do not.
8310 (You might think it would be useful to generalize this to allow each
8311 compiler's spec to say which switches take arguments. But this cannot
8312 be done in a consistent fashion. GCC cannot even decide which input
8313 files have been specified without knowing which switches take arguments,
8314 and it must know which input files to compile in order to tell which
8317 GCC also knows implicitly that arguments starting in @option{-l} are to be
8318 treated as compiler output files, and passed to the linker in their
8319 proper position among the other output files.
8321 @c man begin OPTIONS
8323 @node Target Options
8324 @section Specifying Target Machine and Compiler Version
8325 @cindex target options
8326 @cindex cross compiling
8327 @cindex specifying machine version
8328 @cindex specifying compiler version and target machine
8329 @cindex compiler version, specifying
8330 @cindex target machine, specifying
8332 The usual way to run GCC is to run the executable called @file{gcc}, or
8333 @file{<machine>-gcc} when cross-compiling, or
8334 @file{<machine>-gcc-<version>} to run a version other than the one that
8335 was installed last. Sometimes this is inconvenient, so GCC provides
8336 options that will switch to another cross-compiler or version.
8339 @item -b @var{machine}
8341 The argument @var{machine} specifies the target machine for compilation.
8343 The value to use for @var{machine} is the same as was specified as the
8344 machine type when configuring GCC as a cross-compiler. For
8345 example, if a cross-compiler was configured with @samp{configure
8346 arm-elf}, meaning to compile for an arm processor with elf binaries,
8347 then you would specify @option{-b arm-elf} to run that cross compiler.
8348 Because there are other options beginning with @option{-b}, the
8349 configuration must contain a hyphen.
8351 @item -V @var{version}
8353 The argument @var{version} specifies which version of GCC to run.
8354 This is useful when multiple versions are installed. For example,
8355 @var{version} might be @samp{4.0}, meaning to run GCC version 4.0.
8358 The @option{-V} and @option{-b} options work by running the
8359 @file{<machine>-gcc-<version>} executable, so there's no real reason to
8360 use them if you can just run that directly.
8362 @node Submodel Options
8363 @section Hardware Models and Configurations
8364 @cindex submodel options
8365 @cindex specifying hardware config
8366 @cindex hardware models and configurations, specifying
8367 @cindex machine dependent options
8369 Earlier we discussed the standard option @option{-b} which chooses among
8370 different installed compilers for completely different target
8371 machines, such as VAX vs.@: 68000 vs.@: 80386.
8373 In addition, each of these target machine types can have its own
8374 special options, starting with @samp{-m}, to choose among various
8375 hardware models or configurations---for example, 68010 vs 68020,
8376 floating coprocessor or none. A single installed version of the
8377 compiler can compile for any model or configuration, according to the
8380 Some configurations of the compiler also support additional special
8381 options, usually for compatibility with other compilers on the same
8384 @c This list is ordered alphanumerically by subsection name.
8385 @c It should be the same order and spelling as these options are listed
8386 @c in Machine Dependent Options
8392 * Blackfin Options::
8396 * DEC Alpha Options::
8397 * DEC Alpha/VMS Options::
8399 * GNU/Linux Options::
8402 * i386 and x86-64 Options::
8414 * RS/6000 and PowerPC Options::
8415 * S/390 and zSeries Options::
8420 * System V Options::
8425 * Xstormy16 Options::
8431 @subsection ARC Options
8434 These options are defined for ARC implementations:
8439 Compile code for little endian mode. This is the default.
8443 Compile code for big endian mode.
8446 @opindex mmangle-cpu
8447 Prepend the name of the cpu to all public symbol names.
8448 In multiple-processor systems, there are many ARC variants with different
8449 instruction and register set characteristics. This flag prevents code
8450 compiled for one cpu to be linked with code compiled for another.
8451 No facility exists for handling variants that are ``almost identical''.
8452 This is an all or nothing option.
8454 @item -mcpu=@var{cpu}
8456 Compile code for ARC variant @var{cpu}.
8457 Which variants are supported depend on the configuration.
8458 All variants support @option{-mcpu=base}, this is the default.
8460 @item -mtext=@var{text-section}
8461 @itemx -mdata=@var{data-section}
8462 @itemx -mrodata=@var{readonly-data-section}
8466 Put functions, data, and readonly data in @var{text-section},
8467 @var{data-section}, and @var{readonly-data-section} respectively
8468 by default. This can be overridden with the @code{section} attribute.
8469 @xref{Variable Attributes}.
8474 @subsection ARM Options
8477 These @samp{-m} options are defined for Advanced RISC Machines (ARM)
8481 @item -mabi=@var{name}
8483 Generate code for the specified ABI@. Permissible values are: @samp{apcs-gnu},
8484 @samp{atpcs}, @samp{aapcs}, @samp{aapcs-linux} and @samp{iwmmxt}.
8487 @opindex mapcs-frame
8488 Generate a stack frame that is compliant with the ARM Procedure Call
8489 Standard for all functions, even if this is not strictly necessary for
8490 correct execution of the code. Specifying @option{-fomit-frame-pointer}
8491 with this option will cause the stack frames not to be generated for
8492 leaf functions. The default is @option{-mno-apcs-frame}.
8496 This is a synonym for @option{-mapcs-frame}.
8499 @c not currently implemented
8500 @item -mapcs-stack-check
8501 @opindex mapcs-stack-check
8502 Generate code to check the amount of stack space available upon entry to
8503 every function (that actually uses some stack space). If there is
8504 insufficient space available then either the function
8505 @samp{__rt_stkovf_split_small} or @samp{__rt_stkovf_split_big} will be
8506 called, depending upon the amount of stack space required. The run time
8507 system is required to provide these functions. The default is
8508 @option{-mno-apcs-stack-check}, since this produces smaller code.
8510 @c not currently implemented
8512 @opindex mapcs-float
8513 Pass floating point arguments using the float point registers. This is
8514 one of the variants of the APCS@. This option is recommended if the
8515 target hardware has a floating point unit or if a lot of floating point
8516 arithmetic is going to be performed by the code. The default is
8517 @option{-mno-apcs-float}, since integer only code is slightly increased in
8518 size if @option{-mapcs-float} is used.
8520 @c not currently implemented
8521 @item -mapcs-reentrant
8522 @opindex mapcs-reentrant
8523 Generate reentrant, position independent code. The default is
8524 @option{-mno-apcs-reentrant}.
8527 @item -mthumb-interwork
8528 @opindex mthumb-interwork
8529 Generate code which supports calling between the ARM and Thumb
8530 instruction sets. Without this option the two instruction sets cannot
8531 be reliably used inside one program. The default is
8532 @option{-mno-thumb-interwork}, since slightly larger code is generated
8533 when @option{-mthumb-interwork} is specified.
8535 @item -mno-sched-prolog
8536 @opindex mno-sched-prolog
8537 Prevent the reordering of instructions in the function prolog, or the
8538 merging of those instruction with the instructions in the function's
8539 body. This means that all functions will start with a recognizable set
8540 of instructions (or in fact one of a choice from a small set of
8541 different function prologues), and this information can be used to
8542 locate the start if functions inside an executable piece of code. The
8543 default is @option{-msched-prolog}.
8546 @opindex mhard-float
8547 Generate output containing floating point instructions. This is the
8551 @opindex msoft-float
8552 Generate output containing library calls for floating point.
8553 @strong{Warning:} the requisite libraries are not available for all ARM
8554 targets. Normally the facilities of the machine's usual C compiler are
8555 used, but this cannot be done directly in cross-compilation. You must make
8556 your own arrangements to provide suitable library functions for
8559 @option{-msoft-float} changes the calling convention in the output file;
8560 therefore, it is only useful if you compile @emph{all} of a program with
8561 this option. In particular, you need to compile @file{libgcc.a}, the
8562 library that comes with GCC, with @option{-msoft-float} in order for
8565 @item -mfloat-abi=@var{name}
8567 Specifies which ABI to use for floating point values. Permissible values
8568 are: @samp{soft}, @samp{softfp} and @samp{hard}.
8570 @samp{soft} and @samp{hard} are equivalent to @option{-msoft-float}
8571 and @option{-mhard-float} respectively. @samp{softfp} allows the generation
8572 of floating point instructions, but still uses the soft-float calling
8575 @item -mlittle-endian
8576 @opindex mlittle-endian
8577 Generate code for a processor running in little-endian mode. This is
8578 the default for all standard configurations.
8581 @opindex mbig-endian
8582 Generate code for a processor running in big-endian mode; the default is
8583 to compile code for a little-endian processor.
8585 @item -mwords-little-endian
8586 @opindex mwords-little-endian
8587 This option only applies when generating code for big-endian processors.
8588 Generate code for a little-endian word order but a big-endian byte
8589 order. That is, a byte order of the form @samp{32107654}. Note: this
8590 option should only be used if you require compatibility with code for
8591 big-endian ARM processors generated by versions of the compiler prior to
8594 @item -mcpu=@var{name}
8596 This specifies the name of the target ARM processor. GCC uses this name
8597 to determine what kind of instructions it can emit when generating
8598 assembly code. Permissible names are: @samp{arm2}, @samp{arm250},
8599 @samp{arm3}, @samp{arm6}, @samp{arm60}, @samp{arm600}, @samp{arm610},
8600 @samp{arm620}, @samp{arm7}, @samp{arm7m}, @samp{arm7d}, @samp{arm7dm},
8601 @samp{arm7di}, @samp{arm7dmi}, @samp{arm70}, @samp{arm700},
8602 @samp{arm700i}, @samp{arm710}, @samp{arm710c}, @samp{arm7100},
8603 @samp{arm7500}, @samp{arm7500fe}, @samp{arm7tdmi}, @samp{arm7tdmi-s},
8604 @samp{arm8}, @samp{strongarm}, @samp{strongarm110}, @samp{strongarm1100},
8605 @samp{arm8}, @samp{arm810}, @samp{arm9}, @samp{arm9e}, @samp{arm920},
8606 @samp{arm920t}, @samp{arm922t}, @samp{arm946e-s}, @samp{arm966e-s},
8607 @samp{arm968e-s}, @samp{arm926ej-s}, @samp{arm940t}, @samp{arm9tdmi},
8608 @samp{arm10tdmi}, @samp{arm1020t}, @samp{arm1026ej-s},
8609 @samp{arm10e}, @samp{arm1020e}, @samp{arm1022e},
8610 @samp{arm1136j-s}, @samp{arm1136jf-s}, @samp{mpcore}, @samp{mpcorenovfp},
8611 @samp{arm1156t2-s}, @samp{arm1176jz-s}, @samp{arm1176jzf-s},
8612 @samp{cortex-a8}, @samp{cortex-r4}, @samp{cortex-m3}, @samp{cortex-m1},
8613 @samp{xscale}, @samp{iwmmxt}, @samp{ep9312}.
8615 @item -mtune=@var{name}
8617 This option is very similar to the @option{-mcpu=} option, except that
8618 instead of specifying the actual target processor type, and hence
8619 restricting which instructions can be used, it specifies that GCC should
8620 tune the performance of the code as if the target were of the type
8621 specified in this option, but still choosing the instructions that it
8622 will generate based on the cpu specified by a @option{-mcpu=} option.
8623 For some ARM implementations better performance can be obtained by using
8626 @item -march=@var{name}
8628 This specifies the name of the target ARM architecture. GCC uses this
8629 name to determine what kind of instructions it can emit when generating
8630 assembly code. This option can be used in conjunction with or instead
8631 of the @option{-mcpu=} option. Permissible names are: @samp{armv2},
8632 @samp{armv2a}, @samp{armv3}, @samp{armv3m}, @samp{armv4}, @samp{armv4t},
8633 @samp{armv5}, @samp{armv5t}, @samp{armv5te}, @samp{armv6}, @samp{armv6j},
8634 @samp{armv6t2}, @samp{armv6z}, @samp{armv6zk}, @samp{armv6-m},
8635 @samp{armv7}, @samp{armv7-a}, @samp{armv7-r}, @samp{armv7-m},
8636 @samp{iwmmxt}, @samp{ep9312}.
8638 @item -mfpu=@var{name}
8639 @itemx -mfpe=@var{number}
8640 @itemx -mfp=@var{number}
8644 This specifies what floating point hardware (or hardware emulation) is
8645 available on the target. Permissible names are: @samp{fpa}, @samp{fpe2},
8646 @samp{fpe3}, @samp{maverick}, @samp{vfp}. @option{-mfp} and @option{-mfpe}
8647 are synonyms for @option{-mfpu}=@samp{fpe}@var{number}, for compatibility
8648 with older versions of GCC@.
8650 If @option{-msoft-float} is specified this specifies the format of
8651 floating point values.
8653 @item -mstructure-size-boundary=@var{n}
8654 @opindex mstructure-size-boundary
8655 The size of all structures and unions will be rounded up to a multiple
8656 of the number of bits set by this option. Permissible values are 8, 32
8657 and 64. The default value varies for different toolchains. For the COFF
8658 targeted toolchain the default value is 8. A value of 64 is only allowed
8659 if the underlying ABI supports it.
8661 Specifying the larger number can produce faster, more efficient code, but
8662 can also increase the size of the program. Different values are potentially
8663 incompatible. Code compiled with one value cannot necessarily expect to
8664 work with code or libraries compiled with another value, if they exchange
8665 information using structures or unions.
8667 @item -mabort-on-noreturn
8668 @opindex mabort-on-noreturn
8669 Generate a call to the function @code{abort} at the end of a
8670 @code{noreturn} function. It will be executed if the function tries to
8674 @itemx -mno-long-calls
8675 @opindex mlong-calls
8676 @opindex mno-long-calls
8677 Tells the compiler to perform function calls by first loading the
8678 address of the function into a register and then performing a subroutine
8679 call on this register. This switch is needed if the target function
8680 will lie outside of the 64 megabyte addressing range of the offset based
8681 version of subroutine call instruction.
8683 Even if this switch is enabled, not all function calls will be turned
8684 into long calls. The heuristic is that static functions, functions
8685 which have the @samp{short-call} attribute, functions that are inside
8686 the scope of a @samp{#pragma no_long_calls} directive and functions whose
8687 definitions have already been compiled within the current compilation
8688 unit, will not be turned into long calls. The exception to this rule is
8689 that weak function definitions, functions with the @samp{long-call}
8690 attribute or the @samp{section} attribute, and functions that are within
8691 the scope of a @samp{#pragma long_calls} directive, will always be
8692 turned into long calls.
8694 This feature is not enabled by default. Specifying
8695 @option{-mno-long-calls} will restore the default behavior, as will
8696 placing the function calls within the scope of a @samp{#pragma
8697 long_calls_off} directive. Note these switches have no effect on how
8698 the compiler generates code to handle function calls via function
8701 @item -mnop-fun-dllimport
8702 @opindex mnop-fun-dllimport
8703 Disable support for the @code{dllimport} attribute.
8705 @item -msingle-pic-base
8706 @opindex msingle-pic-base
8707 Treat the register used for PIC addressing as read-only, rather than
8708 loading it in the prologue for each function. The run-time system is
8709 responsible for initializing this register with an appropriate value
8710 before execution begins.
8712 @item -mpic-register=@var{reg}
8713 @opindex mpic-register
8714 Specify the register to be used for PIC addressing. The default is R10
8715 unless stack-checking is enabled, when R9 is used.
8717 @item -mcirrus-fix-invalid-insns
8718 @opindex mcirrus-fix-invalid-insns
8719 @opindex mno-cirrus-fix-invalid-insns
8720 Insert NOPs into the instruction stream to in order to work around
8721 problems with invalid Maverick instruction combinations. This option
8722 is only valid if the @option{-mcpu=ep9312} option has been used to
8723 enable generation of instructions for the Cirrus Maverick floating
8724 point co-processor. This option is not enabled by default, since the
8725 problem is only present in older Maverick implementations. The default
8726 can be re-enabled by use of the @option{-mno-cirrus-fix-invalid-insns}
8729 @item -mpoke-function-name
8730 @opindex mpoke-function-name
8731 Write the name of each function into the text section, directly
8732 preceding the function prologue. The generated code is similar to this:
8736 .ascii "arm_poke_function_name", 0
8739 .word 0xff000000 + (t1 - t0)
8740 arm_poke_function_name
8742 stmfd sp!, @{fp, ip, lr, pc@}
8746 When performing a stack backtrace, code can inspect the value of
8747 @code{pc} stored at @code{fp + 0}. If the trace function then looks at
8748 location @code{pc - 12} and the top 8 bits are set, then we know that
8749 there is a function name embedded immediately preceding this location
8750 and has length @code{((pc[-3]) & 0xff000000)}.
8754 Generate code for the Thumb instruction set. The default is to
8755 use the 32-bit ARM instruction set.
8756 This option automatically enables either 16-bit Thumb-1 or
8757 mixed 16/32-bit Thumb-2 instructions based on the @option{-mcpu=@var{name}}
8758 and @option{-march=@var{name}} options.
8761 @opindex mtpcs-frame
8762 Generate a stack frame that is compliant with the Thumb Procedure Call
8763 Standard for all non-leaf functions. (A leaf function is one that does
8764 not call any other functions.) The default is @option{-mno-tpcs-frame}.
8766 @item -mtpcs-leaf-frame
8767 @opindex mtpcs-leaf-frame
8768 Generate a stack frame that is compliant with the Thumb Procedure Call
8769 Standard for all leaf functions. (A leaf function is one that does
8770 not call any other functions.) The default is @option{-mno-apcs-leaf-frame}.
8772 @item -mcallee-super-interworking
8773 @opindex mcallee-super-interworking
8774 Gives all externally visible functions in the file being compiled an ARM
8775 instruction set header which switches to Thumb mode before executing the
8776 rest of the function. This allows these functions to be called from
8777 non-interworking code.
8779 @item -mcaller-super-interworking
8780 @opindex mcaller-super-interworking
8781 Allows calls via function pointers (including virtual functions) to
8782 execute correctly regardless of whether the target code has been
8783 compiled for interworking or not. There is a small overhead in the cost
8784 of executing a function pointer if this option is enabled.
8786 @item -mtp=@var{name}
8788 Specify the access model for the thread local storage pointer. The valid
8789 models are @option{soft}, which generates calls to @code{__aeabi_read_tp},
8790 @option{cp15}, which fetches the thread pointer from @code{cp15} directly
8791 (supported in the arm6k architecture), and @option{auto}, which uses the
8792 best available method for the selected processor. The default setting is
8798 @subsection AVR Options
8801 These options are defined for AVR implementations:
8804 @item -mmcu=@var{mcu}
8806 Specify ATMEL AVR instruction set or MCU type.
8808 Instruction set avr1 is for the minimal AVR core, not supported by the C
8809 compiler, only for assembler programs (MCU types: at90s1200, attiny10,
8810 attiny11, attiny12, attiny15, attiny28).
8812 Instruction set avr2 (default) is for the classic AVR core with up to
8813 8K program memory space (MCU types: at90s2313, at90s2323, attiny22,
8814 at90s2333, at90s2343, at90s4414, at90s4433, at90s4434, at90s8515,
8815 at90c8534, at90s8535).
8817 Instruction set avr3 is for the classic AVR core with up to 128K program
8818 memory space (MCU types: atmega103, atmega603, at43usb320, at76c711).
8820 Instruction set avr4 is for the enhanced AVR core with up to 8K program
8821 memory space (MCU types: atmega8, atmega83, atmega85).
8823 Instruction set avr5 is for the enhanced AVR core with up to 128K program
8824 memory space (MCU types: atmega16, atmega161, atmega163, atmega32, atmega323,
8825 atmega64, atmega128, at43usb355, at94k).
8829 Output instruction sizes to the asm file.
8831 @item -minit-stack=@var{N}
8832 @opindex minit-stack
8833 Specify the initial stack address, which may be a symbol or numeric value,
8834 @samp{__stack} is the default.
8836 @item -mno-interrupts
8837 @opindex mno-interrupts
8838 Generated code is not compatible with hardware interrupts.
8839 Code size will be smaller.
8841 @item -mcall-prologues
8842 @opindex mcall-prologues
8843 Functions prologues/epilogues expanded as call to appropriate
8844 subroutines. Code size will be smaller.
8846 @item -mno-tablejump
8847 @opindex mno-tablejump
8848 Do not generate tablejump insns which sometimes increase code size.
8851 @opindex mtiny-stack
8852 Change only the low 8 bits of the stack pointer.
8856 Assume int to be 8 bit integer. This affects the sizes of all types: A
8857 char will be 1 byte, an int will be 1 byte, an long will be 2 bytes
8858 and long long will be 4 bytes. Please note that this option does not
8859 comply to the C standards, but it will provide you with smaller code
8863 @node Blackfin Options
8864 @subsection Blackfin Options
8865 @cindex Blackfin Options
8868 @item -mcpu=@var{cpu}@r{[}-@var{sirevision}@r{]}
8870 Specifies the name of the target Blackfin processor. Currently, @var{cpu}
8871 can be one of @samp{bf522}, @samp{bf523}, @samp{bf524},
8872 @samp{bf525}, @samp{bf526}, @samp{bf527},
8873 @samp{bf531}, @samp{bf532}, @samp{bf533}, @samp{bf534},
8874 @samp{bf536}, @samp{bf537}, @samp{bf538}, @samp{bf539},
8875 @samp{bf542}, @samp{bf544}, @samp{bf547}, @samp{bf548}, @samp{bf549},
8877 The optional @var{sirevision} specifies the silicon revision of the target
8878 Blackfin processor. Any workarounds available for the targeted silicon revision
8879 will be enabled. If @var{sirevision} is @samp{none}, no workarounds are enabled.
8880 If @var{sirevision} is @samp{any}, all workarounds for the targeted processor
8881 will be enabled. The @code{__SILICON_REVISION__} macro is defined to two
8882 hexadecimal digits representing the major and minor numbers in the silicon
8883 revision. If @var{sirevision} is @samp{none}, the @code{__SILICON_REVISION__}
8884 is not defined. If @var{sirevision} is @samp{any}, the
8885 @code{__SILICON_REVISION__} is defined to be @code{0xffff}.
8886 If this optional @var{sirevision} is not used, GCC assumes the latest known
8887 silicon revision of the targeted Blackfin processor.
8889 Support for @samp{bf561} is incomplete. For @samp{bf561},
8890 Only the processor macro is defined.
8891 Without this option, @samp{bf532} is used as the processor by default.
8892 The corresponding predefined processor macros for @var{cpu} is to
8893 be defined. And for @samp{bfin-elf} toolchain, this causes the hardware BSP
8894 provided by libgloss to be linked in if @option{-msim} is not given.
8898 Specifies that the program will be run on the simulator. This causes
8899 the simulator BSP provided by libgloss to be linked in. This option
8900 has effect only for @samp{bfin-elf} toolchain.
8901 Certain other options, such as @option{-mid-shared-library} and
8902 @option{-mfdpic}, imply @option{-msim}.
8904 @item -momit-leaf-frame-pointer
8905 @opindex momit-leaf-frame-pointer
8906 Don't keep the frame pointer in a register for leaf functions. This
8907 avoids the instructions to save, set up and restore frame pointers and
8908 makes an extra register available in leaf functions. The option
8909 @option{-fomit-frame-pointer} removes the frame pointer for all functions
8910 which might make debugging harder.
8912 @item -mspecld-anomaly
8913 @opindex mspecld-anomaly
8914 When enabled, the compiler will ensure that the generated code does not
8915 contain speculative loads after jump instructions. If this option is used,
8916 @code{__WORKAROUND_SPECULATIVE_LOADS} is defined.
8918 @item -mno-specld-anomaly
8919 @opindex mno-specld-anomaly
8920 Don't generate extra code to prevent speculative loads from occurring.
8922 @item -mcsync-anomaly
8923 @opindex mcsync-anomaly
8924 When enabled, the compiler will ensure that the generated code does not
8925 contain CSYNC or SSYNC instructions too soon after conditional branches.
8926 If this option is used, @code{__WORKAROUND_SPECULATIVE_SYNCS} is defined.
8928 @item -mno-csync-anomaly
8929 @opindex mno-csync-anomaly
8930 Don't generate extra code to prevent CSYNC or SSYNC instructions from
8931 occurring too soon after a conditional branch.
8935 When enabled, the compiler is free to take advantage of the knowledge that
8936 the entire program fits into the low 64k of memory.
8939 @opindex mno-low-64k
8940 Assume that the program is arbitrarily large. This is the default.
8942 @item -mstack-check-l1
8943 @opindex mstack-check-l1
8944 Do stack checking using information placed into L1 scratchpad memory by the
8947 @item -mid-shared-library
8948 @opindex mid-shared-library
8949 Generate code that supports shared libraries via the library ID method.
8950 This allows for execute in place and shared libraries in an environment
8951 without virtual memory management. This option implies @option{-fPIC}.
8952 With a @samp{bfin-elf} target, this option implies @option{-msim}.
8954 @item -mno-id-shared-library
8955 @opindex mno-id-shared-library
8956 Generate code that doesn't assume ID based shared libraries are being used.
8957 This is the default.
8959 @item -mleaf-id-shared-library
8960 @opindex mleaf-id-shared-library
8961 Generate code that supports shared libraries via the library ID method,
8962 but assumes that this library or executable won't link against any other
8963 ID shared libraries. That allows the compiler to use faster code for jumps
8966 @item -mno-leaf-id-shared-library
8967 @opindex mno-leaf-id-shared-library
8968 Do not assume that the code being compiled won't link against any ID shared
8969 libraries. Slower code will be generated for jump and call insns.
8971 @item -mshared-library-id=n
8972 @opindex mshared-library-id
8973 Specified the identification number of the ID based shared library being
8974 compiled. Specifying a value of 0 will generate more compact code, specifying
8975 other values will force the allocation of that number to the current
8976 library but is no more space or time efficient than omitting this option.
8980 Generate code that allows the data segment to be located in a different
8981 area of memory from the text segment. This allows for execute in place in
8982 an environment without virtual memory management by eliminating relocations
8983 against the text section.
8986 @opindex mno-sep-data
8987 Generate code that assumes that the data segment follows the text segment.
8988 This is the default.
8991 @itemx -mno-long-calls
8992 @opindex mlong-calls
8993 @opindex mno-long-calls
8994 Tells the compiler to perform function calls by first loading the
8995 address of the function into a register and then performing a subroutine
8996 call on this register. This switch is needed if the target function
8997 will lie outside of the 24 bit addressing range of the offset based
8998 version of subroutine call instruction.
9000 This feature is not enabled by default. Specifying
9001 @option{-mno-long-calls} will restore the default behavior. Note these
9002 switches have no effect on how the compiler generates code to handle
9003 function calls via function pointers.
9007 Link with the fast floating-point library. This library relaxes some of
9008 the IEEE floating-point standard's rules for checking inputs against
9009 Not-a-Number (NAN), in the interest of performance.
9012 @opindex minline-plt
9013 Enable inlining of PLT entries in function calls to functions that are
9014 not known to bind locally. It has no effect without @option{-mfdpic}.
9018 Build standalone application for multicore Blackfin processor. Proper
9019 start files and link scripts will be used to support multicore.
9020 This option defines @code{__BFIN_MULTICORE}. It can only be used with
9021 @option{-mcpu=bf561@r{[}-@var{sirevision}@r{]}}. It can be used with
9022 @option{-mcorea} or @option{-mcoreb}. If it's used without
9023 @option{-mcorea} or @option{-mcoreb}, single application/dual core
9024 programming model is used. In this model, the main function of Core B
9025 should be named as coreb_main. If it's used with @option{-mcorea} or
9026 @option{-mcoreb}, one application per core programming model is used.
9027 If this option is not used, single core application programming
9032 Build standalone application for Core A of BF561 when using
9033 one application per core programming model. Proper start files
9034 and link scripts will be used to support Core A. This option
9035 defines @code{__BFIN_COREA}. It must be used with @option{-mmulticore}.
9039 Build standalone application for Core B of BF561 when using
9040 one application per core programming model. Proper start files
9041 and link scripts will be used to support Core B. This option
9042 defines @code{__BFIN_COREB}. When this option is used, coreb_main
9043 should be used instead of main. It must be used with
9044 @option{-mmulticore}.
9048 Build standalone application for SDRAM. Proper start files and
9049 link scripts will be used to put the application into SDRAM.
9050 Loader should initialize SDRAM before loading the application
9051 into SDRAM. This option defines @code{__BFIN_SDRAM}.
9055 @subsection CRIS Options
9056 @cindex CRIS Options
9058 These options are defined specifically for the CRIS ports.
9061 @item -march=@var{architecture-type}
9062 @itemx -mcpu=@var{architecture-type}
9065 Generate code for the specified architecture. The choices for
9066 @var{architecture-type} are @samp{v3}, @samp{v8} and @samp{v10} for
9067 respectively ETRAX@w{ }4, ETRAX@w{ }100, and ETRAX@w{ }100@w{ }LX@.
9068 Default is @samp{v0} except for cris-axis-linux-gnu, where the default is
9071 @item -mtune=@var{architecture-type}
9073 Tune to @var{architecture-type} everything applicable about the generated
9074 code, except for the ABI and the set of available instructions. The
9075 choices for @var{architecture-type} are the same as for
9076 @option{-march=@var{architecture-type}}.
9078 @item -mmax-stack-frame=@var{n}
9079 @opindex mmax-stack-frame
9080 Warn when the stack frame of a function exceeds @var{n} bytes.
9086 The options @option{-metrax4} and @option{-metrax100} are synonyms for
9087 @option{-march=v3} and @option{-march=v8} respectively.
9089 @item -mmul-bug-workaround
9090 @itemx -mno-mul-bug-workaround
9091 @opindex mmul-bug-workaround
9092 @opindex mno-mul-bug-workaround
9093 Work around a bug in the @code{muls} and @code{mulu} instructions for CPU
9094 models where it applies. This option is active by default.
9098 Enable CRIS-specific verbose debug-related information in the assembly
9099 code. This option also has the effect to turn off the @samp{#NO_APP}
9100 formatted-code indicator to the assembler at the beginning of the
9105 Do not use condition-code results from previous instruction; always emit
9106 compare and test instructions before use of condition codes.
9108 @item -mno-side-effects
9109 @opindex mno-side-effects
9110 Do not emit instructions with side-effects in addressing modes other than
9114 @itemx -mno-stack-align
9116 @itemx -mno-data-align
9117 @itemx -mconst-align
9118 @itemx -mno-const-align
9119 @opindex mstack-align
9120 @opindex mno-stack-align
9121 @opindex mdata-align
9122 @opindex mno-data-align
9123 @opindex mconst-align
9124 @opindex mno-const-align
9125 These options (no-options) arranges (eliminate arrangements) for the
9126 stack-frame, individual data and constants to be aligned for the maximum
9127 single data access size for the chosen CPU model. The default is to
9128 arrange for 32-bit alignment. ABI details such as structure layout are
9129 not affected by these options.
9137 Similar to the stack- data- and const-align options above, these options
9138 arrange for stack-frame, writable data and constants to all be 32-bit,
9139 16-bit or 8-bit aligned. The default is 32-bit alignment.
9141 @item -mno-prologue-epilogue
9142 @itemx -mprologue-epilogue
9143 @opindex mno-prologue-epilogue
9144 @opindex mprologue-epilogue
9145 With @option{-mno-prologue-epilogue}, the normal function prologue and
9146 epilogue that sets up the stack-frame are omitted and no return
9147 instructions or return sequences are generated in the code. Use this
9148 option only together with visual inspection of the compiled code: no
9149 warnings or errors are generated when call-saved registers must be saved,
9150 or storage for local variable needs to be allocated.
9156 With @option{-fpic} and @option{-fPIC}, don't generate (do generate)
9157 instruction sequences that load addresses for functions from the PLT part
9158 of the GOT rather than (traditional on other architectures) calls to the
9159 PLT@. The default is @option{-mgotplt}.
9163 Legacy no-op option only recognized with the cris-axis-elf and
9164 cris-axis-linux-gnu targets.
9168 Legacy no-op option only recognized with the cris-axis-linux-gnu target.
9172 This option, recognized for the cris-axis-elf arranges
9173 to link with input-output functions from a simulator library. Code,
9174 initialized data and zero-initialized data are allocated consecutively.
9178 Like @option{-sim}, but pass linker options to locate initialized data at
9179 0x40000000 and zero-initialized data at 0x80000000.
9183 @subsection CRX Options
9186 These options are defined specifically for the CRX ports.
9192 Enable the use of multiply-accumulate instructions. Disabled by default.
9196 Push instructions will be used to pass outgoing arguments when functions
9197 are called. Enabled by default.
9200 @node Darwin Options
9201 @subsection Darwin Options
9202 @cindex Darwin options
9204 These options are defined for all architectures running the Darwin operating
9207 FSF GCC on Darwin does not create ``fat'' object files; it will create
9208 an object file for the single architecture that it was built to
9209 target. Apple's GCC on Darwin does create ``fat'' files if multiple
9210 @option{-arch} options are used; it does so by running the compiler or
9211 linker multiple times and joining the results together with
9214 The subtype of the file created (like @samp{ppc7400} or @samp{ppc970} or
9215 @samp{i686}) is determined by the flags that specify the ISA
9216 that GCC is targetting, like @option{-mcpu} or @option{-march}. The
9217 @option{-force_cpusubtype_ALL} option can be used to override this.
9219 The Darwin tools vary in their behavior when presented with an ISA
9220 mismatch. The assembler, @file{as}, will only permit instructions to
9221 be used that are valid for the subtype of the file it is generating,
9222 so you cannot put 64-bit instructions in an @samp{ppc750} object file.
9223 The linker for shared libraries, @file{/usr/bin/libtool}, will fail
9224 and print an error if asked to create a shared library with a less
9225 restrictive subtype than its input files (for instance, trying to put
9226 a @samp{ppc970} object file in a @samp{ppc7400} library). The linker
9227 for executables, @file{ld}, will quietly give the executable the most
9228 restrictive subtype of any of its input files.
9233 Add the framework directory @var{dir} to the head of the list of
9234 directories to be searched for header files. These directories are
9235 interleaved with those specified by @option{-I} options and are
9236 scanned in a left-to-right order.
9238 A framework directory is a directory with frameworks in it. A
9239 framework is a directory with a @samp{"Headers"} and/or
9240 @samp{"PrivateHeaders"} directory contained directly in it that ends
9241 in @samp{".framework"}. The name of a framework is the name of this
9242 directory excluding the @samp{".framework"}. Headers associated with
9243 the framework are found in one of those two directories, with
9244 @samp{"Headers"} being searched first. A subframework is a framework
9245 directory that is in a framework's @samp{"Frameworks"} directory.
9246 Includes of subframework headers can only appear in a header of a
9247 framework that contains the subframework, or in a sibling subframework
9248 header. Two subframeworks are siblings if they occur in the same
9249 framework. A subframework should not have the same name as a
9250 framework, a warning will be issued if this is violated. Currently a
9251 subframework cannot have subframeworks, in the future, the mechanism
9252 may be extended to support this. The standard frameworks can be found
9253 in @samp{"/System/Library/Frameworks"} and
9254 @samp{"/Library/Frameworks"}. An example include looks like
9255 @code{#include <Framework/header.h>}, where @samp{Framework} denotes
9256 the name of the framework and header.h is found in the
9257 @samp{"PrivateHeaders"} or @samp{"Headers"} directory.
9259 @item -iframework@var{dir}
9261 Like @option{-F} except the directory is a treated as a system
9262 directory. The main difference between this @option{-iframework} and
9263 @option{-F} is that with @option{-iframework} the compiler does not
9264 warn about constructs contained within header files found via
9265 @var{dir}. This option is valid only for the C family of languages.
9269 Emit debugging information for symbols that are used. For STABS
9270 debugging format, this enables @option{-feliminate-unused-debug-symbols}.
9271 This is by default ON@.
9275 Emit debugging information for all symbols and types.
9277 @item -mmacosx-version-min=@var{version}
9278 The earliest version of MacOS X that this executable will run on
9279 is @var{version}. Typical values of @var{version} include @code{10.1},
9280 @code{10.2}, and @code{10.3.9}.
9282 If the compiler was built to use the system's headers by default,
9283 then the default for this option is the system version on which the
9284 compiler is running, otherwise the default is to make choices which
9285 are compatible with as many systems and code bases as possible.
9289 Enable kernel development mode. The @option{-mkernel} option sets
9290 @option{-static}, @option{-fno-common}, @option{-fno-cxa-atexit},
9291 @option{-fno-exceptions}, @option{-fno-non-call-exceptions},
9292 @option{-fapple-kext}, @option{-fno-weak} and @option{-fno-rtti} where
9293 applicable. This mode also sets @option{-mno-altivec},
9294 @option{-msoft-float}, @option{-fno-builtin} and
9295 @option{-mlong-branch} for PowerPC targets.
9297 @item -mone-byte-bool
9298 @opindex mone-byte-bool
9299 Override the defaults for @samp{bool} so that @samp{sizeof(bool)==1}.
9300 By default @samp{sizeof(bool)} is @samp{4} when compiling for
9301 Darwin/PowerPC and @samp{1} when compiling for Darwin/x86, so this
9302 option has no effect on x86.
9304 @strong{Warning:} The @option{-mone-byte-bool} switch causes GCC
9305 to generate code that is not binary compatible with code generated
9306 without that switch. Using this switch may require recompiling all
9307 other modules in a program, including system libraries. Use this
9308 switch to conform to a non-default data model.
9310 @item -mfix-and-continue
9311 @itemx -ffix-and-continue
9312 @itemx -findirect-data
9313 @opindex mfix-and-continue
9314 @opindex ffix-and-continue
9315 @opindex findirect-data
9316 Generate code suitable for fast turn around development. Needed to
9317 enable gdb to dynamically load @code{.o} files into already running
9318 programs. @option{-findirect-data} and @option{-ffix-and-continue}
9319 are provided for backwards compatibility.
9323 Loads all members of static archive libraries.
9324 See man ld(1) for more information.
9326 @item -arch_errors_fatal
9327 @opindex arch_errors_fatal
9328 Cause the errors having to do with files that have the wrong architecture
9332 @opindex bind_at_load
9333 Causes the output file to be marked such that the dynamic linker will
9334 bind all undefined references when the file is loaded or launched.
9338 Produce a Mach-o bundle format file.
9339 See man ld(1) for more information.
9341 @item -bundle_loader @var{executable}
9342 @opindex bundle_loader
9343 This option specifies the @var{executable} that will be loading the build
9344 output file being linked. See man ld(1) for more information.
9348 When passed this option, GCC will produce a dynamic library instead of
9349 an executable when linking, using the Darwin @file{libtool} command.
9351 @item -force_cpusubtype_ALL
9352 @opindex force_cpusubtype_ALL
9353 This causes GCC's output file to have the @var{ALL} subtype, instead of
9354 one controlled by the @option{-mcpu} or @option{-march} option.
9356 @item -allowable_client @var{client_name}
9358 @itemx -compatibility_version
9359 @itemx -current_version
9361 @itemx -dependency-file
9363 @itemx -dylinker_install_name
9365 @itemx -exported_symbols_list
9367 @itemx -flat_namespace
9368 @itemx -force_flat_namespace
9369 @itemx -headerpad_max_install_names
9372 @itemx -install_name
9373 @itemx -keep_private_externs
9374 @itemx -multi_module
9375 @itemx -multiply_defined
9376 @itemx -multiply_defined_unused
9378 @itemx -no_dead_strip_inits_and_terms
9379 @itemx -nofixprebinding
9382 @itemx -noseglinkedit
9383 @itemx -pagezero_size
9385 @itemx -prebind_all_twolevel_modules
9386 @itemx -private_bundle
9387 @itemx -read_only_relocs
9389 @itemx -sectobjectsymbols
9393 @itemx -sectobjectsymbols
9396 @itemx -segs_read_only_addr
9397 @itemx -segs_read_write_addr
9398 @itemx -seg_addr_table
9399 @itemx -seg_addr_table_filename
9402 @itemx -segs_read_only_addr
9403 @itemx -segs_read_write_addr
9404 @itemx -single_module
9407 @itemx -sub_umbrella
9408 @itemx -twolevel_namespace
9411 @itemx -unexported_symbols_list
9412 @itemx -weak_reference_mismatches
9414 @opindex allowable_client
9415 @opindex client_name
9416 @opindex compatibility_version
9417 @opindex current_version
9419 @opindex dependency-file
9421 @opindex dylinker_install_name
9423 @opindex exported_symbols_list
9425 @opindex flat_namespace
9426 @opindex force_flat_namespace
9427 @opindex headerpad_max_install_names
9430 @opindex install_name
9431 @opindex keep_private_externs
9432 @opindex multi_module
9433 @opindex multiply_defined
9434 @opindex multiply_defined_unused
9436 @opindex no_dead_strip_inits_and_terms
9437 @opindex nofixprebinding
9438 @opindex nomultidefs
9440 @opindex noseglinkedit
9441 @opindex pagezero_size
9443 @opindex prebind_all_twolevel_modules
9444 @opindex private_bundle
9445 @opindex read_only_relocs
9447 @opindex sectobjectsymbols
9451 @opindex sectobjectsymbols
9454 @opindex segs_read_only_addr
9455 @opindex segs_read_write_addr
9456 @opindex seg_addr_table
9457 @opindex seg_addr_table_filename
9458 @opindex seglinkedit
9460 @opindex segs_read_only_addr
9461 @opindex segs_read_write_addr
9462 @opindex single_module
9464 @opindex sub_library
9465 @opindex sub_umbrella
9466 @opindex twolevel_namespace
9469 @opindex unexported_symbols_list
9470 @opindex weak_reference_mismatches
9471 @opindex whatsloaded
9472 These options are passed to the Darwin linker. The Darwin linker man page
9473 describes them in detail.
9476 @node DEC Alpha Options
9477 @subsection DEC Alpha Options
9479 These @samp{-m} options are defined for the DEC Alpha implementations:
9482 @item -mno-soft-float
9484 @opindex mno-soft-float
9485 @opindex msoft-float
9486 Use (do not use) the hardware floating-point instructions for
9487 floating-point operations. When @option{-msoft-float} is specified,
9488 functions in @file{libgcc.a} will be used to perform floating-point
9489 operations. Unless they are replaced by routines that emulate the
9490 floating-point operations, or compiled in such a way as to call such
9491 emulations routines, these routines will issue floating-point
9492 operations. If you are compiling for an Alpha without floating-point
9493 operations, you must ensure that the library is built so as not to call
9496 Note that Alpha implementations without floating-point operations are
9497 required to have floating-point registers.
9502 @opindex mno-fp-regs
9503 Generate code that uses (does not use) the floating-point register set.
9504 @option{-mno-fp-regs} implies @option{-msoft-float}. If the floating-point
9505 register set is not used, floating point operands are passed in integer
9506 registers as if they were integers and floating-point results are passed
9507 in @code{$0} instead of @code{$f0}. This is a non-standard calling sequence,
9508 so any function with a floating-point argument or return value called by code
9509 compiled with @option{-mno-fp-regs} must also be compiled with that
9512 A typical use of this option is building a kernel that does not use,
9513 and hence need not save and restore, any floating-point registers.
9517 The Alpha architecture implements floating-point hardware optimized for
9518 maximum performance. It is mostly compliant with the IEEE floating
9519 point standard. However, for full compliance, software assistance is
9520 required. This option generates code fully IEEE compliant code
9521 @emph{except} that the @var{inexact-flag} is not maintained (see below).
9522 If this option is turned on, the preprocessor macro @code{_IEEE_FP} is
9523 defined during compilation. The resulting code is less efficient but is
9524 able to correctly support denormalized numbers and exceptional IEEE
9525 values such as not-a-number and plus/minus infinity. Other Alpha
9526 compilers call this option @option{-ieee_with_no_inexact}.
9528 @item -mieee-with-inexact
9529 @opindex mieee-with-inexact
9530 This is like @option{-mieee} except the generated code also maintains
9531 the IEEE @var{inexact-flag}. Turning on this option causes the
9532 generated code to implement fully-compliant IEEE math. In addition to
9533 @code{_IEEE_FP}, @code{_IEEE_FP_EXACT} is defined as a preprocessor
9534 macro. On some Alpha implementations the resulting code may execute
9535 significantly slower than the code generated by default. Since there is
9536 very little code that depends on the @var{inexact-flag}, you should
9537 normally not specify this option. Other Alpha compilers call this
9538 option @option{-ieee_with_inexact}.
9540 @item -mfp-trap-mode=@var{trap-mode}
9541 @opindex mfp-trap-mode
9542 This option controls what floating-point related traps are enabled.
9543 Other Alpha compilers call this option @option{-fptm @var{trap-mode}}.
9544 The trap mode can be set to one of four values:
9548 This is the default (normal) setting. The only traps that are enabled
9549 are the ones that cannot be disabled in software (e.g., division by zero
9553 In addition to the traps enabled by @samp{n}, underflow traps are enabled
9557 Like @samp{u}, but the instructions are marked to be safe for software
9558 completion (see Alpha architecture manual for details).
9561 Like @samp{su}, but inexact traps are enabled as well.
9564 @item -mfp-rounding-mode=@var{rounding-mode}
9565 @opindex mfp-rounding-mode
9566 Selects the IEEE rounding mode. Other Alpha compilers call this option
9567 @option{-fprm @var{rounding-mode}}. The @var{rounding-mode} can be one
9572 Normal IEEE rounding mode. Floating point numbers are rounded towards
9573 the nearest machine number or towards the even machine number in case
9577 Round towards minus infinity.
9580 Chopped rounding mode. Floating point numbers are rounded towards zero.
9583 Dynamic rounding mode. A field in the floating point control register
9584 (@var{fpcr}, see Alpha architecture reference manual) controls the
9585 rounding mode in effect. The C library initializes this register for
9586 rounding towards plus infinity. Thus, unless your program modifies the
9587 @var{fpcr}, @samp{d} corresponds to round towards plus infinity.
9590 @item -mtrap-precision=@var{trap-precision}
9591 @opindex mtrap-precision
9592 In the Alpha architecture, floating point traps are imprecise. This
9593 means without software assistance it is impossible to recover from a
9594 floating trap and program execution normally needs to be terminated.
9595 GCC can generate code that can assist operating system trap handlers
9596 in determining the exact location that caused a floating point trap.
9597 Depending on the requirements of an application, different levels of
9598 precisions can be selected:
9602 Program precision. This option is the default and means a trap handler
9603 can only identify which program caused a floating point exception.
9606 Function precision. The trap handler can determine the function that
9607 caused a floating point exception.
9610 Instruction precision. The trap handler can determine the exact
9611 instruction that caused a floating point exception.
9614 Other Alpha compilers provide the equivalent options called
9615 @option{-scope_safe} and @option{-resumption_safe}.
9617 @item -mieee-conformant
9618 @opindex mieee-conformant
9619 This option marks the generated code as IEEE conformant. You must not
9620 use this option unless you also specify @option{-mtrap-precision=i} and either
9621 @option{-mfp-trap-mode=su} or @option{-mfp-trap-mode=sui}. Its only effect
9622 is to emit the line @samp{.eflag 48} in the function prologue of the
9623 generated assembly file. Under DEC Unix, this has the effect that
9624 IEEE-conformant math library routines will be linked in.
9626 @item -mbuild-constants
9627 @opindex mbuild-constants
9628 Normally GCC examines a 32- or 64-bit integer constant to
9629 see if it can construct it from smaller constants in two or three
9630 instructions. If it cannot, it will output the constant as a literal and
9631 generate code to load it from the data segment at runtime.
9633 Use this option to require GCC to construct @emph{all} integer constants
9634 using code, even if it takes more instructions (the maximum is six).
9636 You would typically use this option to build a shared library dynamic
9637 loader. Itself a shared library, it must relocate itself in memory
9638 before it can find the variables and constants in its own data segment.
9644 Select whether to generate code to be assembled by the vendor-supplied
9645 assembler (@option{-malpha-as}) or by the GNU assembler @option{-mgas}.
9663 Indicate whether GCC should generate code to use the optional BWX,
9664 CIX, FIX and MAX instruction sets. The default is to use the instruction
9665 sets supported by the CPU type specified via @option{-mcpu=} option or that
9666 of the CPU on which GCC was built if none was specified.
9671 @opindex mfloat-ieee
9672 Generate code that uses (does not use) VAX F and G floating point
9673 arithmetic instead of IEEE single and double precision.
9675 @item -mexplicit-relocs
9676 @itemx -mno-explicit-relocs
9677 @opindex mexplicit-relocs
9678 @opindex mno-explicit-relocs
9679 Older Alpha assemblers provided no way to generate symbol relocations
9680 except via assembler macros. Use of these macros does not allow
9681 optimal instruction scheduling. GNU binutils as of version 2.12
9682 supports a new syntax that allows the compiler to explicitly mark
9683 which relocations should apply to which instructions. This option
9684 is mostly useful for debugging, as GCC detects the capabilities of
9685 the assembler when it is built and sets the default accordingly.
9689 @opindex msmall-data
9690 @opindex mlarge-data
9691 When @option{-mexplicit-relocs} is in effect, static data is
9692 accessed via @dfn{gp-relative} relocations. When @option{-msmall-data}
9693 is used, objects 8 bytes long or smaller are placed in a @dfn{small data area}
9694 (the @code{.sdata} and @code{.sbss} sections) and are accessed via
9695 16-bit relocations off of the @code{$gp} register. This limits the
9696 size of the small data area to 64KB, but allows the variables to be
9697 directly accessed via a single instruction.
9699 The default is @option{-mlarge-data}. With this option the data area
9700 is limited to just below 2GB@. Programs that require more than 2GB of
9701 data must use @code{malloc} or @code{mmap} to allocate the data in the
9702 heap instead of in the program's data segment.
9704 When generating code for shared libraries, @option{-fpic} implies
9705 @option{-msmall-data} and @option{-fPIC} implies @option{-mlarge-data}.
9709 @opindex msmall-text
9710 @opindex mlarge-text
9711 When @option{-msmall-text} is used, the compiler assumes that the
9712 code of the entire program (or shared library) fits in 4MB, and is
9713 thus reachable with a branch instruction. When @option{-msmall-data}
9714 is used, the compiler can assume that all local symbols share the
9715 same @code{$gp} value, and thus reduce the number of instructions
9716 required for a function call from 4 to 1.
9718 The default is @option{-mlarge-text}.
9720 @item -mcpu=@var{cpu_type}
9722 Set the instruction set and instruction scheduling parameters for
9723 machine type @var{cpu_type}. You can specify either the @samp{EV}
9724 style name or the corresponding chip number. GCC supports scheduling
9725 parameters for the EV4, EV5 and EV6 family of processors and will
9726 choose the default values for the instruction set from the processor
9727 you specify. If you do not specify a processor type, GCC will default
9728 to the processor on which the compiler was built.
9730 Supported values for @var{cpu_type} are
9736 Schedules as an EV4 and has no instruction set extensions.
9740 Schedules as an EV5 and has no instruction set extensions.
9744 Schedules as an EV5 and supports the BWX extension.
9749 Schedules as an EV5 and supports the BWX and MAX extensions.
9753 Schedules as an EV6 and supports the BWX, FIX, and MAX extensions.
9757 Schedules as an EV6 and supports the BWX, CIX, FIX, and MAX extensions.
9760 @item -mtune=@var{cpu_type}
9762 Set only the instruction scheduling parameters for machine type
9763 @var{cpu_type}. The instruction set is not changed.
9765 @item -mmemory-latency=@var{time}
9766 @opindex mmemory-latency
9767 Sets the latency the scheduler should assume for typical memory
9768 references as seen by the application. This number is highly
9769 dependent on the memory access patterns used by the application
9770 and the size of the external cache on the machine.
9772 Valid options for @var{time} are
9776 A decimal number representing clock cycles.
9782 The compiler contains estimates of the number of clock cycles for
9783 ``typical'' EV4 & EV5 hardware for the Level 1, 2 & 3 caches
9784 (also called Dcache, Scache, and Bcache), as well as to main memory.
9785 Note that L3 is only valid for EV5.
9790 @node DEC Alpha/VMS Options
9791 @subsection DEC Alpha/VMS Options
9793 These @samp{-m} options are defined for the DEC Alpha/VMS implementations:
9796 @item -mvms-return-codes
9797 @opindex mvms-return-codes
9798 Return VMS condition codes from main. The default is to return POSIX
9799 style condition (e.g.@: error) codes.
9803 @subsection FRV Options
9810 Only use the first 32 general purpose registers.
9815 Use all 64 general purpose registers.
9820 Use only the first 32 floating point registers.
9825 Use all 64 floating point registers
9828 @opindex mhard-float
9830 Use hardware instructions for floating point operations.
9833 @opindex msoft-float
9835 Use library routines for floating point operations.
9840 Dynamically allocate condition code registers.
9845 Do not try to dynamically allocate condition code registers, only
9846 use @code{icc0} and @code{fcc0}.
9851 Change ABI to use double word insns.
9856 Do not use double word instructions.
9861 Use floating point double instructions.
9866 Do not use floating point double instructions.
9871 Use media instructions.
9876 Do not use media instructions.
9881 Use multiply and add/subtract instructions.
9886 Do not use multiply and add/subtract instructions.
9891 Select the FDPIC ABI, that uses function descriptors to represent
9892 pointers to functions. Without any PIC/PIE-related options, it
9893 implies @option{-fPIE}. With @option{-fpic} or @option{-fpie}, it
9894 assumes GOT entries and small data are within a 12-bit range from the
9895 GOT base address; with @option{-fPIC} or @option{-fPIE}, GOT offsets
9896 are computed with 32 bits.
9897 With a @samp{bfin-elf} target, this option implies @option{-msim}.
9900 @opindex minline-plt
9902 Enable inlining of PLT entries in function calls to functions that are
9903 not known to bind locally. It has no effect without @option{-mfdpic}.
9904 It's enabled by default if optimizing for speed and compiling for
9905 shared libraries (i.e., @option{-fPIC} or @option{-fpic}), or when an
9906 optimization option such as @option{-O3} or above is present in the
9912 Assume a large TLS segment when generating thread-local code.
9917 Do not assume a large TLS segment when generating thread-local code.
9922 Enable the use of @code{GPREL} relocations in the FDPIC ABI for data
9923 that is known to be in read-only sections. It's enabled by default,
9924 except for @option{-fpic} or @option{-fpie}: even though it may help
9925 make the global offset table smaller, it trades 1 instruction for 4.
9926 With @option{-fPIC} or @option{-fPIE}, it trades 3 instructions for 4,
9927 one of which may be shared by multiple symbols, and it avoids the need
9928 for a GOT entry for the referenced symbol, so it's more likely to be a
9929 win. If it is not, @option{-mno-gprel-ro} can be used to disable it.
9931 @item -multilib-library-pic
9932 @opindex multilib-library-pic
9934 Link with the (library, not FD) pic libraries. It's implied by
9935 @option{-mlibrary-pic}, as well as by @option{-fPIC} and
9936 @option{-fpic} without @option{-mfdpic}. You should never have to use
9942 Follow the EABI requirement of always creating a frame pointer whenever
9943 a stack frame is allocated. This option is enabled by default and can
9944 be disabled with @option{-mno-linked-fp}.
9947 @opindex mlong-calls
9949 Use indirect addressing to call functions outside the current
9950 compilation unit. This allows the functions to be placed anywhere
9951 within the 32-bit address space.
9953 @item -malign-labels
9954 @opindex malign-labels
9956 Try to align labels to an 8-byte boundary by inserting nops into the
9957 previous packet. This option only has an effect when VLIW packing
9958 is enabled. It doesn't create new packets; it merely adds nops to
9962 @opindex mlibrary-pic
9964 Generate position-independent EABI code.
9969 Use only the first four media accumulator registers.
9974 Use all eight media accumulator registers.
9979 Pack VLIW instructions.
9984 Do not pack VLIW instructions.
9989 Do not mark ABI switches in e_flags.
9994 Enable the use of conditional-move instructions (default).
9996 This switch is mainly for debugging the compiler and will likely be removed
9997 in a future version.
9999 @item -mno-cond-move
10000 @opindex mno-cond-move
10002 Disable the use of conditional-move instructions.
10004 This switch is mainly for debugging the compiler and will likely be removed
10005 in a future version.
10010 Enable the use of conditional set instructions (default).
10012 This switch is mainly for debugging the compiler and will likely be removed
10013 in a future version.
10018 Disable the use of conditional set instructions.
10020 This switch is mainly for debugging the compiler and will likely be removed
10021 in a future version.
10024 @opindex mcond-exec
10026 Enable the use of conditional execution (default).
10028 This switch is mainly for debugging the compiler and will likely be removed
10029 in a future version.
10031 @item -mno-cond-exec
10032 @opindex mno-cond-exec
10034 Disable the use of conditional execution.
10036 This switch is mainly for debugging the compiler and will likely be removed
10037 in a future version.
10039 @item -mvliw-branch
10040 @opindex mvliw-branch
10042 Run a pass to pack branches into VLIW instructions (default).
10044 This switch is mainly for debugging the compiler and will likely be removed
10045 in a future version.
10047 @item -mno-vliw-branch
10048 @opindex mno-vliw-branch
10050 Do not run a pass to pack branches into VLIW instructions.
10052 This switch is mainly for debugging the compiler and will likely be removed
10053 in a future version.
10055 @item -mmulti-cond-exec
10056 @opindex mmulti-cond-exec
10058 Enable optimization of @code{&&} and @code{||} in conditional execution
10061 This switch is mainly for debugging the compiler and will likely be removed
10062 in a future version.
10064 @item -mno-multi-cond-exec
10065 @opindex mno-multi-cond-exec
10067 Disable optimization of @code{&&} and @code{||} in conditional execution.
10069 This switch is mainly for debugging the compiler and will likely be removed
10070 in a future version.
10072 @item -mnested-cond-exec
10073 @opindex mnested-cond-exec
10075 Enable nested conditional execution optimizations (default).
10077 This switch is mainly for debugging the compiler and will likely be removed
10078 in a future version.
10080 @item -mno-nested-cond-exec
10081 @opindex mno-nested-cond-exec
10083 Disable nested conditional execution optimizations.
10085 This switch is mainly for debugging the compiler and will likely be removed
10086 in a future version.
10088 @item -moptimize-membar
10089 @opindex moptimize-membar
10091 This switch removes redundant @code{membar} instructions from the
10092 compiler generated code. It is enabled by default.
10094 @item -mno-optimize-membar
10095 @opindex mno-optimize-membar
10097 This switch disables the automatic removal of redundant @code{membar}
10098 instructions from the generated code.
10100 @item -mtomcat-stats
10101 @opindex mtomcat-stats
10103 Cause gas to print out tomcat statistics.
10105 @item -mcpu=@var{cpu}
10108 Select the processor type for which to generate code. Possible values are
10109 @samp{frv}, @samp{fr550}, @samp{tomcat}, @samp{fr500}, @samp{fr450},
10110 @samp{fr405}, @samp{fr400}, @samp{fr300} and @samp{simple}.
10114 @node GNU/Linux Options
10115 @subsection GNU/Linux Options
10117 These @samp{-m} options are defined for GNU/Linux targets:
10122 Use the GNU C library instead of uClibc. This is the default except
10123 on @samp{*-*-linux-*uclibc*} targets.
10127 Use uClibc instead of the GNU C library. This is the default on
10128 @samp{*-*-linux-*uclibc*} targets.
10131 @node H8/300 Options
10132 @subsection H8/300 Options
10134 These @samp{-m} options are defined for the H8/300 implementations:
10139 Shorten some address references at link time, when possible; uses the
10140 linker option @option{-relax}. @xref{H8/300,, @code{ld} and the H8/300,
10141 ld, Using ld}, for a fuller description.
10145 Generate code for the H8/300H@.
10149 Generate code for the H8S@.
10153 Generate code for the H8S and H8/300H in the normal mode. This switch
10154 must be used either with @option{-mh} or @option{-ms}.
10158 Generate code for the H8S/2600. This switch must be used with @option{-ms}.
10162 Make @code{int} data 32 bits by default.
10165 @opindex malign-300
10166 On the H8/300H and H8S, use the same alignment rules as for the H8/300.
10167 The default for the H8/300H and H8S is to align longs and floats on 4
10169 @option{-malign-300} causes them to be aligned on 2 byte boundaries.
10170 This option has no effect on the H8/300.
10174 @subsection HPPA Options
10175 @cindex HPPA Options
10177 These @samp{-m} options are defined for the HPPA family of computers:
10180 @item -march=@var{architecture-type}
10182 Generate code for the specified architecture. The choices for
10183 @var{architecture-type} are @samp{1.0} for PA 1.0, @samp{1.1} for PA
10184 1.1, and @samp{2.0} for PA 2.0 processors. Refer to
10185 @file{/usr/lib/sched.models} on an HP-UX system to determine the proper
10186 architecture option for your machine. Code compiled for lower numbered
10187 architectures will run on higher numbered architectures, but not the
10190 @item -mpa-risc-1-0
10191 @itemx -mpa-risc-1-1
10192 @itemx -mpa-risc-2-0
10193 @opindex mpa-risc-1-0
10194 @opindex mpa-risc-1-1
10195 @opindex mpa-risc-2-0
10196 Synonyms for @option{-march=1.0}, @option{-march=1.1}, and @option{-march=2.0} respectively.
10199 @opindex mbig-switch
10200 Generate code suitable for big switch tables. Use this option only if
10201 the assembler/linker complain about out of range branches within a switch
10204 @item -mjump-in-delay
10205 @opindex mjump-in-delay
10206 Fill delay slots of function calls with unconditional jump instructions
10207 by modifying the return pointer for the function call to be the target
10208 of the conditional jump.
10210 @item -mdisable-fpregs
10211 @opindex mdisable-fpregs
10212 Prevent floating point registers from being used in any manner. This is
10213 necessary for compiling kernels which perform lazy context switching of
10214 floating point registers. If you use this option and attempt to perform
10215 floating point operations, the compiler will abort.
10217 @item -mdisable-indexing
10218 @opindex mdisable-indexing
10219 Prevent the compiler from using indexing address modes. This avoids some
10220 rather obscure problems when compiling MIG generated code under MACH@.
10222 @item -mno-space-regs
10223 @opindex mno-space-regs
10224 Generate code that assumes the target has no space registers. This allows
10225 GCC to generate faster indirect calls and use unscaled index address modes.
10227 Such code is suitable for level 0 PA systems and kernels.
10229 @item -mfast-indirect-calls
10230 @opindex mfast-indirect-calls
10231 Generate code that assumes calls never cross space boundaries. This
10232 allows GCC to emit code which performs faster indirect calls.
10234 This option will not work in the presence of shared libraries or nested
10237 @item -mfixed-range=@var{register-range}
10238 @opindex mfixed-range
10239 Generate code treating the given register range as fixed registers.
10240 A fixed register is one that the register allocator can not use. This is
10241 useful when compiling kernel code. A register range is specified as
10242 two registers separated by a dash. Multiple register ranges can be
10243 specified separated by a comma.
10245 @item -mlong-load-store
10246 @opindex mlong-load-store
10247 Generate 3-instruction load and store sequences as sometimes required by
10248 the HP-UX 10 linker. This is equivalent to the @samp{+k} option to
10251 @item -mportable-runtime
10252 @opindex mportable-runtime
10253 Use the portable calling conventions proposed by HP for ELF systems.
10257 Enable the use of assembler directives only GAS understands.
10259 @item -mschedule=@var{cpu-type}
10261 Schedule code according to the constraints for the machine type
10262 @var{cpu-type}. The choices for @var{cpu-type} are @samp{700}
10263 @samp{7100}, @samp{7100LC}, @samp{7200}, @samp{7300} and @samp{8000}. Refer
10264 to @file{/usr/lib/sched.models} on an HP-UX system to determine the
10265 proper scheduling option for your machine. The default scheduling is
10269 @opindex mlinker-opt
10270 Enable the optimization pass in the HP-UX linker. Note this makes symbolic
10271 debugging impossible. It also triggers a bug in the HP-UX 8 and HP-UX 9
10272 linkers in which they give bogus error messages when linking some programs.
10275 @opindex msoft-float
10276 Generate output containing library calls for floating point.
10277 @strong{Warning:} the requisite libraries are not available for all HPPA
10278 targets. Normally the facilities of the machine's usual C compiler are
10279 used, but this cannot be done directly in cross-compilation. You must make
10280 your own arrangements to provide suitable library functions for
10283 @option{-msoft-float} changes the calling convention in the output file;
10284 therefore, it is only useful if you compile @emph{all} of a program with
10285 this option. In particular, you need to compile @file{libgcc.a}, the
10286 library that comes with GCC, with @option{-msoft-float} in order for
10291 Generate the predefine, @code{_SIO}, for server IO@. The default is
10292 @option{-mwsio}. This generates the predefines, @code{__hp9000s700},
10293 @code{__hp9000s700__} and @code{_WSIO}, for workstation IO@. These
10294 options are available under HP-UX and HI-UX@.
10298 Use GNU ld specific options. This passes @option{-shared} to ld when
10299 building a shared library. It is the default when GCC is configured,
10300 explicitly or implicitly, with the GNU linker. This option does not
10301 have any affect on which ld is called, it only changes what parameters
10302 are passed to that ld. The ld that is called is determined by the
10303 @option{--with-ld} configure option, GCC's program search path, and
10304 finally by the user's @env{PATH}. The linker used by GCC can be printed
10305 using @samp{which `gcc -print-prog-name=ld`}. This option is only available
10306 on the 64 bit HP-UX GCC, i.e.@: configured with @samp{hppa*64*-*-hpux*}.
10310 Use HP ld specific options. This passes @option{-b} to ld when building
10311 a shared library and passes @option{+Accept TypeMismatch} to ld on all
10312 links. It is the default when GCC is configured, explicitly or
10313 implicitly, with the HP linker. This option does not have any affect on
10314 which ld is called, it only changes what parameters are passed to that
10315 ld. The ld that is called is determined by the @option{--with-ld}
10316 configure option, GCC's program search path, and finally by the user's
10317 @env{PATH}. The linker used by GCC can be printed using @samp{which
10318 `gcc -print-prog-name=ld`}. This option is only available on the 64 bit
10319 HP-UX GCC, i.e.@: configured with @samp{hppa*64*-*-hpux*}.
10322 @opindex mno-long-calls
10323 Generate code that uses long call sequences. This ensures that a call
10324 is always able to reach linker generated stubs. The default is to generate
10325 long calls only when the distance from the call site to the beginning
10326 of the function or translation unit, as the case may be, exceeds a
10327 predefined limit set by the branch type being used. The limits for
10328 normal calls are 7,600,000 and 240,000 bytes, respectively for the
10329 PA 2.0 and PA 1.X architectures. Sibcalls are always limited at
10332 Distances are measured from the beginning of functions when using the
10333 @option{-ffunction-sections} option, or when using the @option{-mgas}
10334 and @option{-mno-portable-runtime} options together under HP-UX with
10337 It is normally not desirable to use this option as it will degrade
10338 performance. However, it may be useful in large applications,
10339 particularly when partial linking is used to build the application.
10341 The types of long calls used depends on the capabilities of the
10342 assembler and linker, and the type of code being generated. The
10343 impact on systems that support long absolute calls, and long pic
10344 symbol-difference or pc-relative calls should be relatively small.
10345 However, an indirect call is used on 32-bit ELF systems in pic code
10346 and it is quite long.
10348 @item -munix=@var{unix-std}
10350 Generate compiler predefines and select a startfile for the specified
10351 UNIX standard. The choices for @var{unix-std} are @samp{93}, @samp{95}
10352 and @samp{98}. @samp{93} is supported on all HP-UX versions. @samp{95}
10353 is available on HP-UX 10.10 and later. @samp{98} is available on HP-UX
10354 11.11 and later. The default values are @samp{93} for HP-UX 10.00,
10355 @samp{95} for HP-UX 10.10 though to 11.00, and @samp{98} for HP-UX 11.11
10358 @option{-munix=93} provides the same predefines as GCC 3.3 and 3.4.
10359 @option{-munix=95} provides additional predefines for @code{XOPEN_UNIX}
10360 and @code{_XOPEN_SOURCE_EXTENDED}, and the startfile @file{unix95.o}.
10361 @option{-munix=98} provides additional predefines for @code{_XOPEN_UNIX},
10362 @code{_XOPEN_SOURCE_EXTENDED}, @code{_INCLUDE__STDC_A1_SOURCE} and
10363 @code{_INCLUDE_XOPEN_SOURCE_500}, and the startfile @file{unix98.o}.
10365 It is @emph{important} to note that this option changes the interfaces
10366 for various library routines. It also affects the operational behavior
10367 of the C library. Thus, @emph{extreme} care is needed in using this
10370 Library code that is intended to operate with more than one UNIX
10371 standard must test, set and restore the variable @var{__xpg4_extended_mask}
10372 as appropriate. Most GNU software doesn't provide this capability.
10376 Suppress the generation of link options to search libdld.sl when the
10377 @option{-static} option is specified on HP-UX 10 and later.
10381 The HP-UX implementation of setlocale in libc has a dependency on
10382 libdld.sl. There isn't an archive version of libdld.sl. Thus,
10383 when the @option{-static} option is specified, special link options
10384 are needed to resolve this dependency.
10386 On HP-UX 10 and later, the GCC driver adds the necessary options to
10387 link with libdld.sl when the @option{-static} option is specified.
10388 This causes the resulting binary to be dynamic. On the 64-bit port,
10389 the linkers generate dynamic binaries by default in any case. The
10390 @option{-nolibdld} option can be used to prevent the GCC driver from
10391 adding these link options.
10395 Add support for multithreading with the @dfn{dce thread} library
10396 under HP-UX@. This option sets flags for both the preprocessor and
10400 @node i386 and x86-64 Options
10401 @subsection Intel 386 and AMD x86-64 Options
10402 @cindex i386 Options
10403 @cindex x86-64 Options
10404 @cindex Intel 386 Options
10405 @cindex AMD x86-64 Options
10407 These @samp{-m} options are defined for the i386 and x86-64 family of
10411 @item -mtune=@var{cpu-type}
10413 Tune to @var{cpu-type} everything applicable about the generated code, except
10414 for the ABI and the set of available instructions. The choices for
10415 @var{cpu-type} are:
10418 Produce code optimized for the most common IA32/AMD64/EM64T processors.
10419 If you know the CPU on which your code will run, then you should use
10420 the corresponding @option{-mtune} option instead of
10421 @option{-mtune=generic}. But, if you do not know exactly what CPU users
10422 of your application will have, then you should use this option.
10424 As new processors are deployed in the marketplace, the behavior of this
10425 option will change. Therefore, if you upgrade to a newer version of
10426 GCC, the code generated option will change to reflect the processors
10427 that were most common when that version of GCC was released.
10429 There is no @option{-march=generic} option because @option{-march}
10430 indicates the instruction set the compiler can use, and there is no
10431 generic instruction set applicable to all processors. In contrast,
10432 @option{-mtune} indicates the processor (or, in this case, collection of
10433 processors) for which the code is optimized.
10435 This selects the CPU to tune for at compilation time by determining
10436 the processor type of the compiling machine. Using @option{-mtune=native}
10437 will produce code optimized for the local machine under the constraints
10438 of the selected instruction set. Using @option{-march=native} will
10439 enable all instruction subsets supported by the local machine (hence
10440 the result might not run on different machines).
10442 Original Intel's i386 CPU@.
10444 Intel's i486 CPU@. (No scheduling is implemented for this chip.)
10445 @item i586, pentium
10446 Intel Pentium CPU with no MMX support.
10448 Intel PentiumMMX CPU based on Pentium core with MMX instruction set support.
10450 Intel PentiumPro CPU@.
10452 Same as @code{generic}, but when used as @code{march} option, PentiumPro
10453 instruction set will be used, so the code will run on all i686 family chips.
10455 Intel Pentium2 CPU based on PentiumPro core with MMX instruction set support.
10456 @item pentium3, pentium3m
10457 Intel Pentium3 CPU based on PentiumPro core with MMX and SSE instruction set
10460 Low power version of Intel Pentium3 CPU with MMX, SSE and SSE2 instruction set
10461 support. Used by Centrino notebooks.
10462 @item pentium4, pentium4m
10463 Intel Pentium4 CPU with MMX, SSE and SSE2 instruction set support.
10465 Improved version of Intel Pentium4 CPU with MMX, SSE, SSE2 and SSE3 instruction
10468 Improved version of Intel Pentium4 CPU with 64-bit extensions, MMX, SSE,
10469 SSE2 and SSE3 instruction set support.
10471 Intel Core2 CPU with 64-bit extensions, MMX, SSE, SSE2, SSE3 and SSSE3
10472 instruction set support.
10474 AMD K6 CPU with MMX instruction set support.
10476 Improved versions of AMD K6 CPU with MMX and 3dNOW!@: instruction set support.
10477 @item athlon, athlon-tbird
10478 AMD Athlon CPU with MMX, 3dNOW!, enhanced 3dNOW!@: and SSE prefetch instructions
10480 @item athlon-4, athlon-xp, athlon-mp
10481 Improved AMD Athlon CPU with MMX, 3dNOW!, enhanced 3dNOW!@: and full SSE
10482 instruction set support.
10483 @item k8, opteron, athlon64, athlon-fx
10484 AMD K8 core based CPUs with x86-64 instruction set support. (This supersets
10485 MMX, SSE, SSE2, 3dNOW!, enhanced 3dNOW!@: and 64-bit instruction set extensions.)
10486 @item k8-sse3, opteron-sse3, athlon64-sse3
10487 Improved versions of k8, opteron and athlon64 with SSE3 instruction set support.
10488 @item amdfam10, barcelona
10489 AMD Family 10h core based CPUs with x86-64 instruction set support. (This
10490 supersets MMX, SSE, SSE2, SSE3, SSE4A, 3dNOW!, enhanced 3dNOW!, ABM and 64-bit
10491 instruction set extensions.)
10493 IDT Winchip C6 CPU, dealt in same way as i486 with additional MMX instruction
10496 IDT Winchip2 CPU, dealt in same way as i486 with additional MMX and 3dNOW!@:
10497 instruction set support.
10499 Via C3 CPU with MMX and 3dNOW!@: instruction set support. (No scheduling is
10500 implemented for this chip.)
10502 Via C3-2 CPU with MMX and SSE instruction set support. (No scheduling is
10503 implemented for this chip.)
10505 Embedded AMD CPU with MMX and 3dNOW! instruction set support.
10508 While picking a specific @var{cpu-type} will schedule things appropriately
10509 for that particular chip, the compiler will not generate any code that
10510 does not run on the i386 without the @option{-march=@var{cpu-type}} option
10513 @item -march=@var{cpu-type}
10515 Generate instructions for the machine type @var{cpu-type}. The choices
10516 for @var{cpu-type} are the same as for @option{-mtune}. Moreover,
10517 specifying @option{-march=@var{cpu-type}} implies @option{-mtune=@var{cpu-type}}.
10519 @item -mcpu=@var{cpu-type}
10521 A deprecated synonym for @option{-mtune}.
10523 @item -mfpmath=@var{unit}
10525 Generate floating point arithmetics for selected unit @var{unit}. The choices
10526 for @var{unit} are:
10530 Use the standard 387 floating point coprocessor present majority of chips and
10531 emulated otherwise. Code compiled with this option will run almost everywhere.
10532 The temporary results are computed in 80bit precision instead of precision
10533 specified by the type resulting in slightly different results compared to most
10534 of other chips. See @option{-ffloat-store} for more detailed description.
10536 This is the default choice for i386 compiler.
10539 Use scalar floating point instructions present in the SSE instruction set.
10540 This instruction set is supported by Pentium3 and newer chips, in the AMD line
10541 by Athlon-4, Athlon-xp and Athlon-mp chips. The earlier version of SSE
10542 instruction set supports only single precision arithmetics, thus the double and
10543 extended precision arithmetics is still done using 387. Later version, present
10544 only in Pentium4 and the future AMD x86-64 chips supports double precision
10547 For the i386 compiler, you need to use @option{-march=@var{cpu-type}}, @option{-msse}
10548 or @option{-msse2} switches to enable SSE extensions and make this option
10549 effective. For the x86-64 compiler, these extensions are enabled by default.
10551 The resulting code should be considerably faster in the majority of cases and avoid
10552 the numerical instability problems of 387 code, but may break some existing
10553 code that expects temporaries to be 80bit.
10555 This is the default choice for the x86-64 compiler.
10560 Attempt to utilize both instruction sets at once. This effectively double the
10561 amount of available registers and on chips with separate execution units for
10562 387 and SSE the execution resources too. Use this option with care, as it is
10563 still experimental, because the GCC register allocator does not model separate
10564 functional units well resulting in instable performance.
10567 @item -masm=@var{dialect}
10568 @opindex masm=@var{dialect}
10569 Output asm instructions using selected @var{dialect}. Supported
10570 choices are @samp{intel} or @samp{att} (the default one). Darwin does
10571 not support @samp{intel}.
10574 @itemx -mno-ieee-fp
10576 @opindex mno-ieee-fp
10577 Control whether or not the compiler uses IEEE floating point
10578 comparisons. These handle correctly the case where the result of a
10579 comparison is unordered.
10582 @opindex msoft-float
10583 Generate output containing library calls for floating point.
10584 @strong{Warning:} the requisite libraries are not part of GCC@.
10585 Normally the facilities of the machine's usual C compiler are used, but
10586 this can't be done directly in cross-compilation. You must make your
10587 own arrangements to provide suitable library functions for
10590 On machines where a function returns floating point results in the 80387
10591 register stack, some floating point opcodes may be emitted even if
10592 @option{-msoft-float} is used.
10594 @item -mno-fp-ret-in-387
10595 @opindex mno-fp-ret-in-387
10596 Do not use the FPU registers for return values of functions.
10598 The usual calling convention has functions return values of types
10599 @code{float} and @code{double} in an FPU register, even if there
10600 is no FPU@. The idea is that the operating system should emulate
10603 The option @option{-mno-fp-ret-in-387} causes such values to be returned
10604 in ordinary CPU registers instead.
10606 @item -mno-fancy-math-387
10607 @opindex mno-fancy-math-387
10608 Some 387 emulators do not support the @code{sin}, @code{cos} and
10609 @code{sqrt} instructions for the 387. Specify this option to avoid
10610 generating those instructions. This option is the default on FreeBSD,
10611 OpenBSD and NetBSD@. This option is overridden when @option{-march}
10612 indicates that the target cpu will always have an FPU and so the
10613 instruction will not need emulation. As of revision 2.6.1, these
10614 instructions are not generated unless you also use the
10615 @option{-funsafe-math-optimizations} switch.
10617 @item -malign-double
10618 @itemx -mno-align-double
10619 @opindex malign-double
10620 @opindex mno-align-double
10621 Control whether GCC aligns @code{double}, @code{long double}, and
10622 @code{long long} variables on a two word boundary or a one word
10623 boundary. Aligning @code{double} variables on a two word boundary will
10624 produce code that runs somewhat faster on a @samp{Pentium} at the
10625 expense of more memory.
10627 On x86-64, @option{-malign-double} is enabled by default.
10629 @strong{Warning:} if you use the @option{-malign-double} switch,
10630 structures containing the above types will be aligned differently than
10631 the published application binary interface specifications for the 386
10632 and will not be binary compatible with structures in code compiled
10633 without that switch.
10635 @item -m96bit-long-double
10636 @itemx -m128bit-long-double
10637 @opindex m96bit-long-double
10638 @opindex m128bit-long-double
10639 These switches control the size of @code{long double} type. The i386
10640 application binary interface specifies the size to be 96 bits,
10641 so @option{-m96bit-long-double} is the default in 32 bit mode.
10643 Modern architectures (Pentium and newer) would prefer @code{long double}
10644 to be aligned to an 8 or 16 byte boundary. In arrays or structures
10645 conforming to the ABI, this would not be possible. So specifying a
10646 @option{-m128bit-long-double} will align @code{long double}
10647 to a 16 byte boundary by padding the @code{long double} with an additional
10650 In the x86-64 compiler, @option{-m128bit-long-double} is the default choice as
10651 its ABI specifies that @code{long double} is to be aligned on 16 byte boundary.
10653 Notice that neither of these options enable any extra precision over the x87
10654 standard of 80 bits for a @code{long double}.
10656 @strong{Warning:} if you override the default value for your target ABI, the
10657 structures and arrays containing @code{long double} variables will change
10658 their size as well as function calling convention for function taking
10659 @code{long double} will be modified. Hence they will not be binary
10660 compatible with arrays or structures in code compiled without that switch.
10662 @item -mmlarge-data-threshold=@var{number}
10663 @opindex mlarge-data-threshold=@var{number}
10664 When @option{-mcmodel=medium} is specified, the data greater than
10665 @var{threshold} are placed in large data section. This value must be the
10666 same across all object linked into the binary and defaults to 65535.
10670 Use a different function-calling convention, in which functions that
10671 take a fixed number of arguments return with the @code{ret} @var{num}
10672 instruction, which pops their arguments while returning. This saves one
10673 instruction in the caller since there is no need to pop the arguments
10676 You can specify that an individual function is called with this calling
10677 sequence with the function attribute @samp{stdcall}. You can also
10678 override the @option{-mrtd} option by using the function attribute
10679 @samp{cdecl}. @xref{Function Attributes}.
10681 @strong{Warning:} this calling convention is incompatible with the one
10682 normally used on Unix, so you cannot use it if you need to call
10683 libraries compiled with the Unix compiler.
10685 Also, you must provide function prototypes for all functions that
10686 take variable numbers of arguments (including @code{printf});
10687 otherwise incorrect code will be generated for calls to those
10690 In addition, seriously incorrect code will result if you call a
10691 function with too many arguments. (Normally, extra arguments are
10692 harmlessly ignored.)
10694 @item -mregparm=@var{num}
10696 Control how many registers are used to pass integer arguments. By
10697 default, no registers are used to pass arguments, and at most 3
10698 registers can be used. You can control this behavior for a specific
10699 function by using the function attribute @samp{regparm}.
10700 @xref{Function Attributes}.
10702 @strong{Warning:} if you use this switch, and
10703 @var{num} is nonzero, then you must build all modules with the same
10704 value, including any libraries. This includes the system libraries and
10708 @opindex msseregparm
10709 Use SSE register passing conventions for float and double arguments
10710 and return values. You can control this behavior for a specific
10711 function by using the function attribute @samp{sseregparm}.
10712 @xref{Function Attributes}.
10714 @strong{Warning:} if you use this switch then you must build all
10715 modules with the same value, including any libraries. This includes
10716 the system libraries and startup modules.
10725 Set 80387 floating-point precision to 32, 64 or 80 bits. When @option{-mpc32}
10726 is specified, the significands of results of floating-point operations are
10727 rounded to 24 bits (single precision); @option{-mpc64} rounds the the
10728 significands of results of floating-point operations to 53 bits (double
10729 precision) and @option{-mpc80} rounds the significands of results of
10730 floating-point operations to 64 bits (extended double precision), which is
10731 the default. When this option is used, floating-point operations in higher
10732 precisions are not available to the programmer without setting the FPU
10733 control word explicitly.
10735 Setting the rounding of floating-point operations to less than the default
10736 80 bits can speed some programs by 2% or more. Note that some mathematical
10737 libraries assume that extended precision (80 bit) floating-point operations
10738 are enabled by default; routines in such libraries could suffer significant
10739 loss of accuracy, typically through so-called "catastrophic cancellation",
10740 when this option is used to set the precision to less than extended precision.
10742 @item -mstackrealign
10743 @opindex mstackrealign
10744 Realign the stack at entry. On the Intel x86, the @option{-mstackrealign}
10745 option will generate an alternate prologue and epilogue that realigns the
10746 runtime stack if necessary. This supports mixing legacy codes that keep
10747 a 4-byte aligned stack with modern codes that keep a 16-byte stack for
10748 SSE compatibility. See also the attribute @code{force_align_arg_pointer},
10749 applicable to individual functions.
10751 @item -mpreferred-stack-boundary=@var{num}
10752 @opindex mpreferred-stack-boundary
10753 Attempt to keep the stack boundary aligned to a 2 raised to @var{num}
10754 byte boundary. If @option{-mpreferred-stack-boundary} is not specified,
10755 the default is 4 (16 bytes or 128 bits).
10757 @item -mincoming-stack-boundary=@var{num}
10758 @opindex mincoming-stack-boundary
10759 Assume the incoming stack is aligned to a 2 raised to @var{num} byte
10760 boundary. If @option{-mincoming-stack-boundary} is not specified,
10761 the one specified by @option{-mpreferred-stack-boundary} will be used.
10763 On Pentium and PentiumPro, @code{double} and @code{long double} values
10764 should be aligned to an 8 byte boundary (see @option{-malign-double}) or
10765 suffer significant run time performance penalties. On Pentium III, the
10766 Streaming SIMD Extension (SSE) data type @code{__m128} may not work
10767 properly if it is not 16 byte aligned.
10769 To ensure proper alignment of this values on the stack, the stack boundary
10770 must be as aligned as that required by any value stored on the stack.
10771 Further, every function must be generated such that it keeps the stack
10772 aligned. Thus calling a function compiled with a higher preferred
10773 stack boundary from a function compiled with a lower preferred stack
10774 boundary will most likely misalign the stack. It is recommended that
10775 libraries that use callbacks always use the default setting.
10777 This extra alignment does consume extra stack space, and generally
10778 increases code size. Code that is sensitive to stack space usage, such
10779 as embedded systems and operating system kernels, may want to reduce the
10780 preferred alignment to @option{-mpreferred-stack-boundary=2}.
10818 These switches enable or disable the use of instructions in the MMX,
10819 SSE, SSE2, SSE3, SSSE3, SSE4.1, AES, PCLMUL, SSE4A, SSE5, ABM or
10820 3DNow!@: extended instruction sets.
10821 These extensions are also available as built-in functions: see
10822 @ref{X86 Built-in Functions}, for details of the functions enabled and
10823 disabled by these switches.
10825 To have SSE/SSE2 instructions generated automatically from floating-point
10826 code (as opposed to 387 instructions), see @option{-mfpmath=sse}.
10828 These options will enable GCC to use these extended instructions in
10829 generated code, even without @option{-mfpmath=sse}. Applications which
10830 perform runtime CPU detection must compile separate files for each
10831 supported architecture, using the appropriate flags. In particular,
10832 the file containing the CPU detection code should be compiled without
10837 This option instructs GCC to emit a @code{cld} instruction in the prologue
10838 of functions that use string instructions. String instructions depend on
10839 the DF flag to select between autoincrement or autodecrement mode. While the
10840 ABI specifies the DF flag to be cleared on function entry, some operating
10841 systems violate this specification by not clearing the DF flag in their
10842 exception dispatchers. The exception handler can be invoked with the DF flag
10843 set which leads to wrong direction mode, when string instructions are used.
10844 This option can be enabled by default on 32-bit x86 targets by configuring
10845 GCC with the @option{--enable-cld} configure option. Generation of @code{cld}
10846 instructions can be suppressed with the @option{-mno-cld} compiler option
10851 This option will enable GCC to use CMPXCHG16B instruction in generated code.
10852 CMPXCHG16B allows for atomic operations on 128-bit double quadword (or oword)
10853 data types. This is useful for high resolution counters that could be updated
10854 by multiple processors (or cores). This instruction is generated as part of
10855 atomic built-in functions: see @ref{Atomic Builtins} for details.
10859 This option will enable GCC to use SAHF instruction in generated 64-bit code.
10860 Early Intel CPUs with Intel 64 lacked LAHF and SAHF instructions supported
10861 by AMD64 until introduction of Pentium 4 G1 step in December 2005. LAHF and
10862 SAHF are load and store instructions, respectively, for certain status flags.
10863 In 64-bit mode, SAHF instruction is used to optimize @code{fmod}, @code{drem}
10864 or @code{remainder} built-in functions: see @ref{Other Builtins} for details.
10868 This option will enable GCC to use RCPSS and RSQRTSS instructions (and their
10869 vectorized variants RCPPS and RSQRTPS) with an additional Newton-Rhapson step
10870 to increase precision instead of DIVSS and SQRTSS (and their vectorized
10871 variants) for single precision floating point arguments. These instructions
10872 are generated only when @option{-funsafe-math-optimizations} is enabled
10873 together with @option{-finite-math-only} and @option{-fno-trapping-math}.
10874 Note that while the throughput of the sequence is higher than the throughput
10875 of the non-reciprocal instruction, the precision of the sequence can be
10876 decreased by up to 2 ulp (i.e. the inverse of 1.0 equals 0.99999994).
10878 @item -mveclibabi=@var{type}
10879 @opindex mveclibabi
10880 Specifies the ABI type to use for vectorizing intrinsics using an
10881 external library. Supported types are @code{svml} for the Intel short
10882 vector math library and @code{acml} for the AMD math core library style
10883 of interfacing. GCC will currently emit calls to @code{vmldExp2},
10884 @code{vmldLn2}, @code{vmldLog102}, @code{vmldLog102}, @code{vmldPow2},
10885 @code{vmldTanh2}, @code{vmldTan2}, @code{vmldAtan2}, @code{vmldAtanh2},
10886 @code{vmldCbrt2}, @code{vmldSinh2}, @code{vmldSin2}, @code{vmldAsinh2},
10887 @code{vmldAsin2}, @code{vmldCosh2}, @code{vmldCos2}, @code{vmldAcosh2},
10888 @code{vmldAcos2}, @code{vmlsExp4}, @code{vmlsLn4}, @code{vmlsLog104},
10889 @code{vmlsLog104}, @code{vmlsPow4}, @code{vmlsTanh4}, @code{vmlsTan4},
10890 @code{vmlsAtan4}, @code{vmlsAtanh4}, @code{vmlsCbrt4}, @code{vmlsSinh4},
10891 @code{vmlsSin4}, @code{vmlsAsinh4}, @code{vmlsAsin4}, @code{vmlsCosh4},
10892 @code{vmlsCos4}, @code{vmlsAcosh4} and @code{vmlsAcos4} for corresponding
10893 function type when @option{-mveclibabi=svml} is used and @code{__vrd2_sin},
10894 @code{__vrd2_cos}, @code{__vrd2_exp}, @code{__vrd2_log}, @code{__vrd2_log2},
10895 @code{__vrd2_log10}, @code{__vrs4_sinf}, @code{__vrs4_cosf},
10896 @code{__vrs4_expf}, @code{__vrs4_logf}, @code{__vrs4_log2f},
10897 @code{__vrs4_log10f} and @code{__vrs4_powf} for corresponding function type
10898 when @option{-mveclibabi=acml} is used. Both @option{-ftree-vectorize} and
10899 @option{-funsafe-math-optimizations} have to be enabled. A SVML or ACML ABI
10900 compatible library will have to be specified at link time.
10903 @itemx -mno-push-args
10904 @opindex mpush-args
10905 @opindex mno-push-args
10906 Use PUSH operations to store outgoing parameters. This method is shorter
10907 and usually equally fast as method using SUB/MOV operations and is enabled
10908 by default. In some cases disabling it may improve performance because of
10909 improved scheduling and reduced dependencies.
10911 @item -maccumulate-outgoing-args
10912 @opindex maccumulate-outgoing-args
10913 If enabled, the maximum amount of space required for outgoing arguments will be
10914 computed in the function prologue. This is faster on most modern CPUs
10915 because of reduced dependencies, improved scheduling and reduced stack usage
10916 when preferred stack boundary is not equal to 2. The drawback is a notable
10917 increase in code size. This switch implies @option{-mno-push-args}.
10921 Support thread-safe exception handling on @samp{Mingw32}. Code that relies
10922 on thread-safe exception handling must compile and link all code with the
10923 @option{-mthreads} option. When compiling, @option{-mthreads} defines
10924 @option{-D_MT}; when linking, it links in a special thread helper library
10925 @option{-lmingwthrd} which cleans up per thread exception handling data.
10927 @item -mno-align-stringops
10928 @opindex mno-align-stringops
10929 Do not align destination of inlined string operations. This switch reduces
10930 code size and improves performance in case the destination is already aligned,
10931 but GCC doesn't know about it.
10933 @item -minline-all-stringops
10934 @opindex minline-all-stringops
10935 By default GCC inlines string operations only when destination is known to be
10936 aligned at least to 4 byte boundary. This enables more inlining, increase code
10937 size, but may improve performance of code that depends on fast memcpy, strlen
10938 and memset for short lengths.
10940 @item -minline-stringops-dynamically
10941 @opindex minline-stringops-dynamically
10942 For string operation of unknown size, inline runtime checks so for small
10943 blocks inline code is used, while for large blocks library call is used.
10945 @item -mstringop-strategy=@var{alg}
10946 @opindex mstringop-strategy=@var{alg}
10947 Overwrite internal decision heuristic about particular algorithm to inline
10948 string operation with. The allowed values are @code{rep_byte},
10949 @code{rep_4byte}, @code{rep_8byte} for expanding using i386 @code{rep} prefix
10950 of specified size, @code{byte_loop}, @code{loop}, @code{unrolled_loop} for
10951 expanding inline loop, @code{libcall} for always expanding library call.
10953 @item -momit-leaf-frame-pointer
10954 @opindex momit-leaf-frame-pointer
10955 Don't keep the frame pointer in a register for leaf functions. This
10956 avoids the instructions to save, set up and restore frame pointers and
10957 makes an extra register available in leaf functions. The option
10958 @option{-fomit-frame-pointer} removes the frame pointer for all functions
10959 which might make debugging harder.
10961 @item -mtls-direct-seg-refs
10962 @itemx -mno-tls-direct-seg-refs
10963 @opindex mtls-direct-seg-refs
10964 Controls whether TLS variables may be accessed with offsets from the
10965 TLS segment register (@code{%gs} for 32-bit, @code{%fs} for 64-bit),
10966 or whether the thread base pointer must be added. Whether or not this
10967 is legal depends on the operating system, and whether it maps the
10968 segment to cover the entire TLS area.
10970 For systems that use GNU libc, the default is on.
10973 @itemx -mno-fused-madd
10974 @opindex mfused-madd
10975 Enable automatic generation of fused floating point multiply-add instructions
10976 if the ISA supports such instructions. The -mfused-madd option is on by
10977 default. The fused multiply-add instructions have a different
10978 rounding behavior compared to executing a multiply followed by an add.
10981 These @samp{-m} switches are supported in addition to the above
10982 on AMD x86-64 processors in 64-bit environments.
10989 Generate code for a 32-bit or 64-bit environment.
10990 The 32-bit environment sets int, long and pointer to 32 bits and
10991 generates code that runs on any i386 system.
10992 The 64-bit environment sets int to 32 bits and long and pointer
10993 to 64 bits and generates code for AMD's x86-64 architecture. For
10994 darwin only the -m64 option turns off the @option{-fno-pic} and
10995 @option{-mdynamic-no-pic} options.
10997 @item -mno-red-zone
10998 @opindex no-red-zone
10999 Do not use a so called red zone for x86-64 code. The red zone is mandated
11000 by the x86-64 ABI, it is a 128-byte area beyond the location of the
11001 stack pointer that will not be modified by signal or interrupt handlers
11002 and therefore can be used for temporary data without adjusting the stack
11003 pointer. The flag @option{-mno-red-zone} disables this red zone.
11005 @item -mcmodel=small
11006 @opindex mcmodel=small
11007 Generate code for the small code model: the program and its symbols must
11008 be linked in the lower 2 GB of the address space. Pointers are 64 bits.
11009 Programs can be statically or dynamically linked. This is the default
11012 @item -mcmodel=kernel
11013 @opindex mcmodel=kernel
11014 Generate code for the kernel code model. The kernel runs in the
11015 negative 2 GB of the address space.
11016 This model has to be used for Linux kernel code.
11018 @item -mcmodel=medium
11019 @opindex mcmodel=medium
11020 Generate code for the medium model: The program is linked in the lower 2
11021 GB of the address space but symbols can be located anywhere in the
11022 address space. Programs can be statically or dynamically linked, but
11023 building of shared libraries are not supported with the medium model.
11025 @item -mcmodel=large
11026 @opindex mcmodel=large
11027 Generate code for the large model: This model makes no assumptions
11028 about addresses and sizes of sections.
11031 @node IA-64 Options
11032 @subsection IA-64 Options
11033 @cindex IA-64 Options
11035 These are the @samp{-m} options defined for the Intel IA-64 architecture.
11039 @opindex mbig-endian
11040 Generate code for a big endian target. This is the default for HP-UX@.
11042 @item -mlittle-endian
11043 @opindex mlittle-endian
11044 Generate code for a little endian target. This is the default for AIX5
11050 @opindex mno-gnu-as
11051 Generate (or don't) code for the GNU assembler. This is the default.
11052 @c Also, this is the default if the configure option @option{--with-gnu-as}
11058 @opindex mno-gnu-ld
11059 Generate (or don't) code for the GNU linker. This is the default.
11060 @c Also, this is the default if the configure option @option{--with-gnu-ld}
11065 Generate code that does not use a global pointer register. The result
11066 is not position independent code, and violates the IA-64 ABI@.
11068 @item -mvolatile-asm-stop
11069 @itemx -mno-volatile-asm-stop
11070 @opindex mvolatile-asm-stop
11071 @opindex mno-volatile-asm-stop
11072 Generate (or don't) a stop bit immediately before and after volatile asm
11075 @item -mregister-names
11076 @itemx -mno-register-names
11077 @opindex mregister-names
11078 @opindex mno-register-names
11079 Generate (or don't) @samp{in}, @samp{loc}, and @samp{out} register names for
11080 the stacked registers. This may make assembler output more readable.
11086 Disable (or enable) optimizations that use the small data section. This may
11087 be useful for working around optimizer bugs.
11089 @item -mconstant-gp
11090 @opindex mconstant-gp
11091 Generate code that uses a single constant global pointer value. This is
11092 useful when compiling kernel code.
11096 Generate code that is self-relocatable. This implies @option{-mconstant-gp}.
11097 This is useful when compiling firmware code.
11099 @item -minline-float-divide-min-latency
11100 @opindex minline-float-divide-min-latency
11101 Generate code for inline divides of floating point values
11102 using the minimum latency algorithm.
11104 @item -minline-float-divide-max-throughput
11105 @opindex minline-float-divide-max-throughput
11106 Generate code for inline divides of floating point values
11107 using the maximum throughput algorithm.
11109 @item -minline-int-divide-min-latency
11110 @opindex minline-int-divide-min-latency
11111 Generate code for inline divides of integer values
11112 using the minimum latency algorithm.
11114 @item -minline-int-divide-max-throughput
11115 @opindex minline-int-divide-max-throughput
11116 Generate code for inline divides of integer values
11117 using the maximum throughput algorithm.
11119 @item -minline-sqrt-min-latency
11120 @opindex minline-sqrt-min-latency
11121 Generate code for inline square roots
11122 using the minimum latency algorithm.
11124 @item -minline-sqrt-max-throughput
11125 @opindex minline-sqrt-max-throughput
11126 Generate code for inline square roots
11127 using the maximum throughput algorithm.
11129 @item -mno-dwarf2-asm
11130 @itemx -mdwarf2-asm
11131 @opindex mno-dwarf2-asm
11132 @opindex mdwarf2-asm
11133 Don't (or do) generate assembler code for the DWARF2 line number debugging
11134 info. This may be useful when not using the GNU assembler.
11136 @item -mearly-stop-bits
11137 @itemx -mno-early-stop-bits
11138 @opindex mearly-stop-bits
11139 @opindex mno-early-stop-bits
11140 Allow stop bits to be placed earlier than immediately preceding the
11141 instruction that triggered the stop bit. This can improve instruction
11142 scheduling, but does not always do so.
11144 @item -mfixed-range=@var{register-range}
11145 @opindex mfixed-range
11146 Generate code treating the given register range as fixed registers.
11147 A fixed register is one that the register allocator can not use. This is
11148 useful when compiling kernel code. A register range is specified as
11149 two registers separated by a dash. Multiple register ranges can be
11150 specified separated by a comma.
11152 @item -mtls-size=@var{tls-size}
11154 Specify bit size of immediate TLS offsets. Valid values are 14, 22, and
11157 @item -mtune=@var{cpu-type}
11159 Tune the instruction scheduling for a particular CPU, Valid values are
11160 itanium, itanium1, merced, itanium2, and mckinley.
11166 Add support for multithreading using the POSIX threads library. This
11167 option sets flags for both the preprocessor and linker. It does
11168 not affect the thread safety of object code produced by the compiler or
11169 that of libraries supplied with it. These are HP-UX specific flags.
11175 Generate code for a 32-bit or 64-bit environment.
11176 The 32-bit environment sets int, long and pointer to 32 bits.
11177 The 64-bit environment sets int to 32 bits and long and pointer
11178 to 64 bits. These are HP-UX specific flags.
11180 @item -mno-sched-br-data-spec
11181 @itemx -msched-br-data-spec
11182 @opindex mno-sched-br-data-spec
11183 @opindex msched-br-data-spec
11184 (Dis/En)able data speculative scheduling before reload.
11185 This will result in generation of the ld.a instructions and
11186 the corresponding check instructions (ld.c / chk.a).
11187 The default is 'disable'.
11189 @item -msched-ar-data-spec
11190 @itemx -mno-sched-ar-data-spec
11191 @opindex msched-ar-data-spec
11192 @opindex mno-sched-ar-data-spec
11193 (En/Dis)able data speculative scheduling after reload.
11194 This will result in generation of the ld.a instructions and
11195 the corresponding check instructions (ld.c / chk.a).
11196 The default is 'enable'.
11198 @item -mno-sched-control-spec
11199 @itemx -msched-control-spec
11200 @opindex mno-sched-control-spec
11201 @opindex msched-control-spec
11202 (Dis/En)able control speculative scheduling. This feature is
11203 available only during region scheduling (i.e.@: before reload).
11204 This will result in generation of the ld.s instructions and
11205 the corresponding check instructions chk.s .
11206 The default is 'disable'.
11208 @item -msched-br-in-data-spec
11209 @itemx -mno-sched-br-in-data-spec
11210 @opindex msched-br-in-data-spec
11211 @opindex mno-sched-br-in-data-spec
11212 (En/Dis)able speculative scheduling of the instructions that
11213 are dependent on the data speculative loads before reload.
11214 This is effective only with @option{-msched-br-data-spec} enabled.
11215 The default is 'enable'.
11217 @item -msched-ar-in-data-spec
11218 @itemx -mno-sched-ar-in-data-spec
11219 @opindex msched-ar-in-data-spec
11220 @opindex mno-sched-ar-in-data-spec
11221 (En/Dis)able speculative scheduling of the instructions that
11222 are dependent on the data speculative loads after reload.
11223 This is effective only with @option{-msched-ar-data-spec} enabled.
11224 The default is 'enable'.
11226 @item -msched-in-control-spec
11227 @itemx -mno-sched-in-control-spec
11228 @opindex msched-in-control-spec
11229 @opindex mno-sched-in-control-spec
11230 (En/Dis)able speculative scheduling of the instructions that
11231 are dependent on the control speculative loads.
11232 This is effective only with @option{-msched-control-spec} enabled.
11233 The default is 'enable'.
11236 @itemx -mno-sched-ldc
11237 @opindex msched-ldc
11238 @opindex mno-sched-ldc
11239 (En/Dis)able use of simple data speculation checks ld.c .
11240 If disabled, only chk.a instructions will be emitted to check
11241 data speculative loads.
11242 The default is 'enable'.
11244 @item -mno-sched-control-ldc
11245 @itemx -msched-control-ldc
11246 @opindex mno-sched-control-ldc
11247 @opindex msched-control-ldc
11248 (Dis/En)able use of ld.c instructions to check control speculative loads.
11249 If enabled, in case of control speculative load with no speculatively
11250 scheduled dependent instructions this load will be emitted as ld.sa and
11251 ld.c will be used to check it.
11252 The default is 'disable'.
11254 @item -mno-sched-spec-verbose
11255 @itemx -msched-spec-verbose
11256 @opindex mno-sched-spec-verbose
11257 @opindex msched-spec-verbose
11258 (Dis/En)able printing of the information about speculative motions.
11260 @item -mno-sched-prefer-non-data-spec-insns
11261 @itemx -msched-prefer-non-data-spec-insns
11262 @opindex mno-sched-prefer-non-data-spec-insns
11263 @opindex msched-prefer-non-data-spec-insns
11264 If enabled, data speculative instructions will be chosen for schedule
11265 only if there are no other choices at the moment. This will make
11266 the use of the data speculation much more conservative.
11267 The default is 'disable'.
11269 @item -mno-sched-prefer-non-control-spec-insns
11270 @itemx -msched-prefer-non-control-spec-insns
11271 @opindex mno-sched-prefer-non-control-spec-insns
11272 @opindex msched-prefer-non-control-spec-insns
11273 If enabled, control speculative instructions will be chosen for schedule
11274 only if there are no other choices at the moment. This will make
11275 the use of the control speculation much more conservative.
11276 The default is 'disable'.
11278 @item -mno-sched-count-spec-in-critical-path
11279 @itemx -msched-count-spec-in-critical-path
11280 @opindex mno-sched-count-spec-in-critical-path
11281 @opindex msched-count-spec-in-critical-path
11282 If enabled, speculative dependencies will be considered during
11283 computation of the instructions priorities. This will make the use of the
11284 speculation a bit more conservative.
11285 The default is 'disable'.
11290 @subsection M32C Options
11291 @cindex M32C options
11294 @item -mcpu=@var{name}
11296 Select the CPU for which code is generated. @var{name} may be one of
11297 @samp{r8c} for the R8C/Tiny series, @samp{m16c} for the M16C (up to
11298 /60) series, @samp{m32cm} for the M16C/80 series, or @samp{m32c} for
11299 the M32C/80 series.
11303 Specifies that the program will be run on the simulator. This causes
11304 an alternate runtime library to be linked in which supports, for
11305 example, file I/O@. You must not use this option when generating
11306 programs that will run on real hardware; you must provide your own
11307 runtime library for whatever I/O functions are needed.
11309 @item -memregs=@var{number}
11311 Specifies the number of memory-based pseudo-registers GCC will use
11312 during code generation. These pseudo-registers will be used like real
11313 registers, so there is a tradeoff between GCC's ability to fit the
11314 code into available registers, and the performance penalty of using
11315 memory instead of registers. Note that all modules in a program must
11316 be compiled with the same value for this option. Because of that, you
11317 must not use this option with the default runtime libraries gcc
11322 @node M32R/D Options
11323 @subsection M32R/D Options
11324 @cindex M32R/D options
11326 These @option{-m} options are defined for Renesas M32R/D architectures:
11331 Generate code for the M32R/2@.
11335 Generate code for the M32R/X@.
11339 Generate code for the M32R@. This is the default.
11341 @item -mmodel=small
11342 @opindex mmodel=small
11343 Assume all objects live in the lower 16MB of memory (so that their addresses
11344 can be loaded with the @code{ld24} instruction), and assume all subroutines
11345 are reachable with the @code{bl} instruction.
11346 This is the default.
11348 The addressability of a particular object can be set with the
11349 @code{model} attribute.
11351 @item -mmodel=medium
11352 @opindex mmodel=medium
11353 Assume objects may be anywhere in the 32-bit address space (the compiler
11354 will generate @code{seth/add3} instructions to load their addresses), and
11355 assume all subroutines are reachable with the @code{bl} instruction.
11357 @item -mmodel=large
11358 @opindex mmodel=large
11359 Assume objects may be anywhere in the 32-bit address space (the compiler
11360 will generate @code{seth/add3} instructions to load their addresses), and
11361 assume subroutines may not be reachable with the @code{bl} instruction
11362 (the compiler will generate the much slower @code{seth/add3/jl}
11363 instruction sequence).
11366 @opindex msdata=none
11367 Disable use of the small data area. Variables will be put into
11368 one of @samp{.data}, @samp{bss}, or @samp{.rodata} (unless the
11369 @code{section} attribute has been specified).
11370 This is the default.
11372 The small data area consists of sections @samp{.sdata} and @samp{.sbss}.
11373 Objects may be explicitly put in the small data area with the
11374 @code{section} attribute using one of these sections.
11376 @item -msdata=sdata
11377 @opindex msdata=sdata
11378 Put small global and static data in the small data area, but do not
11379 generate special code to reference them.
11382 @opindex msdata=use
11383 Put small global and static data in the small data area, and generate
11384 special instructions to reference them.
11388 @cindex smaller data references
11389 Put global and static objects less than or equal to @var{num} bytes
11390 into the small data or bss sections instead of the normal data or bss
11391 sections. The default value of @var{num} is 8.
11392 The @option{-msdata} option must be set to one of @samp{sdata} or @samp{use}
11393 for this option to have any effect.
11395 All modules should be compiled with the same @option{-G @var{num}} value.
11396 Compiling with different values of @var{num} may or may not work; if it
11397 doesn't the linker will give an error message---incorrect code will not be
11402 Makes the M32R specific code in the compiler display some statistics
11403 that might help in debugging programs.
11405 @item -malign-loops
11406 @opindex malign-loops
11407 Align all loops to a 32-byte boundary.
11409 @item -mno-align-loops
11410 @opindex mno-align-loops
11411 Do not enforce a 32-byte alignment for loops. This is the default.
11413 @item -missue-rate=@var{number}
11414 @opindex missue-rate=@var{number}
11415 Issue @var{number} instructions per cycle. @var{number} can only be 1
11418 @item -mbranch-cost=@var{number}
11419 @opindex mbranch-cost=@var{number}
11420 @var{number} can only be 1 or 2. If it is 1 then branches will be
11421 preferred over conditional code, if it is 2, then the opposite will
11424 @item -mflush-trap=@var{number}
11425 @opindex mflush-trap=@var{number}
11426 Specifies the trap number to use to flush the cache. The default is
11427 12. Valid numbers are between 0 and 15 inclusive.
11429 @item -mno-flush-trap
11430 @opindex mno-flush-trap
11431 Specifies that the cache cannot be flushed by using a trap.
11433 @item -mflush-func=@var{name}
11434 @opindex mflush-func=@var{name}
11435 Specifies the name of the operating system function to call to flush
11436 the cache. The default is @emph{_flush_cache}, but a function call
11437 will only be used if a trap is not available.
11439 @item -mno-flush-func
11440 @opindex mno-flush-func
11441 Indicates that there is no OS function for flushing the cache.
11445 @node M680x0 Options
11446 @subsection M680x0 Options
11447 @cindex M680x0 options
11449 These are the @samp{-m} options defined for M680x0 and ColdFire processors.
11450 The default settings depend on which architecture was selected when
11451 the compiler was configured; the defaults for the most common choices
11455 @item -march=@var{arch}
11457 Generate code for a specific M680x0 or ColdFire instruction set
11458 architecture. Permissible values of @var{arch} for M680x0
11459 architectures are: @samp{68000}, @samp{68010}, @samp{68020},
11460 @samp{68030}, @samp{68040}, @samp{68060} and @samp{cpu32}. ColdFire
11461 architectures are selected according to Freescale's ISA classification
11462 and the permissible values are: @samp{isaa}, @samp{isaaplus},
11463 @samp{isab} and @samp{isac}.
11465 gcc defines a macro @samp{__mcf@var{arch}__} whenever it is generating
11466 code for a ColdFire target. The @var{arch} in this macro is one of the
11467 @option{-march} arguments given above.
11469 When used together, @option{-march} and @option{-mtune} select code
11470 that runs on a family of similar processors but that is optimized
11471 for a particular microarchitecture.
11473 @item -mcpu=@var{cpu}
11475 Generate code for a specific M680x0 or ColdFire processor.
11476 The M680x0 @var{cpu}s are: @samp{68000}, @samp{68010}, @samp{68020},
11477 @samp{68030}, @samp{68040}, @samp{68060}, @samp{68302}, @samp{68332}
11478 and @samp{cpu32}. The ColdFire @var{cpu}s are given by the table
11479 below, which also classifies the CPUs into families:
11481 @multitable @columnfractions 0.20 0.80
11482 @item @strong{Family} @tab @strong{@samp{-mcpu} arguments}
11483 @item @samp{51qe} @tab @samp{51qe}
11484 @item @samp{5206} @tab @samp{5202} @samp{5204} @samp{5206}
11485 @item @samp{5206e} @tab @samp{5206e}
11486 @item @samp{5208} @tab @samp{5207} @samp{5208}
11487 @item @samp{5211a} @tab @samp{5210a} @samp{5211a}
11488 @item @samp{5213} @tab @samp{5211} @samp{5212} @samp{5213}
11489 @item @samp{5216} @tab @samp{5214} @samp{5216}
11490 @item @samp{52235} @tab @samp{52230} @samp{52231} @samp{52232} @samp{52233} @samp{52234} @samp{52235}
11491 @item @samp{5225} @tab @samp{5224} @samp{5225}
11492 @item @samp{5235} @tab @samp{5232} @samp{5233} @samp{5234} @samp{5235} @samp{523x}
11493 @item @samp{5249} @tab @samp{5249}
11494 @item @samp{5250} @tab @samp{5250}
11495 @item @samp{5271} @tab @samp{5270} @samp{5271}
11496 @item @samp{5272} @tab @samp{5272}
11497 @item @samp{5275} @tab @samp{5274} @samp{5275}
11498 @item @samp{5282} @tab @samp{5280} @samp{5281} @samp{5282} @samp{528x}
11499 @item @samp{5307} @tab @samp{5307}
11500 @item @samp{5329} @tab @samp{5327} @samp{5328} @samp{5329} @samp{532x}
11501 @item @samp{5373} @tab @samp{5372} @samp{5373} @samp{537x}
11502 @item @samp{5407} @tab @samp{5407}
11503 @item @samp{5475} @tab @samp{5470} @samp{5471} @samp{5472} @samp{5473} @samp{5474} @samp{5475} @samp{547x} @samp{5480} @samp{5481} @samp{5482} @samp{5483} @samp{5484} @samp{5485}
11506 @option{-mcpu=@var{cpu}} overrides @option{-march=@var{arch}} if
11507 @var{arch} is compatible with @var{cpu}. Other combinations of
11508 @option{-mcpu} and @option{-march} are rejected.
11510 gcc defines the macro @samp{__mcf_cpu_@var{cpu}} when ColdFire target
11511 @var{cpu} is selected. It also defines @samp{__mcf_family_@var{family}},
11512 where the value of @var{family} is given by the table above.
11514 @item -mtune=@var{tune}
11516 Tune the code for a particular microarchitecture, within the
11517 constraints set by @option{-march} and @option{-mcpu}.
11518 The M680x0 microarchitectures are: @samp{68000}, @samp{68010},
11519 @samp{68020}, @samp{68030}, @samp{68040}, @samp{68060}
11520 and @samp{cpu32}. The ColdFire microarchitectures
11521 are: @samp{cfv1}, @samp{cfv2}, @samp{cfv3}, @samp{cfv4} and @samp{cfv4e}.
11523 You can also use @option{-mtune=68020-40} for code that needs
11524 to run relatively well on 68020, 68030 and 68040 targets.
11525 @option{-mtune=68020-60} is similar but includes 68060 targets
11526 as well. These two options select the same tuning decisions as
11527 @option{-m68020-40} and @option{-m68020-60} respectively.
11529 gcc defines the macros @samp{__mc@var{arch}} and @samp{__mc@var{arch}__}
11530 when tuning for 680x0 architecture @var{arch}. It also defines
11531 @samp{mc@var{arch}} unless either @option{-ansi} or a non-GNU @option{-std}
11532 option is used. If gcc is tuning for a range of architectures,
11533 as selected by @option{-mtune=68020-40} or @option{-mtune=68020-60},
11534 it defines the macros for every architecture in the range.
11536 gcc also defines the macro @samp{__m@var{uarch}__} when tuning for
11537 ColdFire microarchitecture @var{uarch}, where @var{uarch} is one
11538 of the arguments given above.
11544 Generate output for a 68000. This is the default
11545 when the compiler is configured for 68000-based systems.
11546 It is equivalent to @option{-march=68000}.
11548 Use this option for microcontrollers with a 68000 or EC000 core,
11549 including the 68008, 68302, 68306, 68307, 68322, 68328 and 68356.
11553 Generate output for a 68010. This is the default
11554 when the compiler is configured for 68010-based systems.
11555 It is equivalent to @option{-march=68010}.
11561 Generate output for a 68020. This is the default
11562 when the compiler is configured for 68020-based systems.
11563 It is equivalent to @option{-march=68020}.
11567 Generate output for a 68030. This is the default when the compiler is
11568 configured for 68030-based systems. It is equivalent to
11569 @option{-march=68030}.
11573 Generate output for a 68040. This is the default when the compiler is
11574 configured for 68040-based systems. It is equivalent to
11575 @option{-march=68040}.
11577 This option inhibits the use of 68881/68882 instructions that have to be
11578 emulated by software on the 68040. Use this option if your 68040 does not
11579 have code to emulate those instructions.
11583 Generate output for a 68060. This is the default when the compiler is
11584 configured for 68060-based systems. It is equivalent to
11585 @option{-march=68060}.
11587 This option inhibits the use of 68020 and 68881/68882 instructions that
11588 have to be emulated by software on the 68060. Use this option if your 68060
11589 does not have code to emulate those instructions.
11593 Generate output for a CPU32. This is the default
11594 when the compiler is configured for CPU32-based systems.
11595 It is equivalent to @option{-march=cpu32}.
11597 Use this option for microcontrollers with a
11598 CPU32 or CPU32+ core, including the 68330, 68331, 68332, 68333, 68334,
11599 68336, 68340, 68341, 68349 and 68360.
11603 Generate output for a 520X ColdFire CPU@. This is the default
11604 when the compiler is configured for 520X-based systems.
11605 It is equivalent to @option{-mcpu=5206}, and is now deprecated
11606 in favor of that option.
11608 Use this option for microcontroller with a 5200 core, including
11609 the MCF5202, MCF5203, MCF5204 and MCF5206.
11613 Generate output for a 5206e ColdFire CPU@. The option is now
11614 deprecated in favor of the equivalent @option{-mcpu=5206e}.
11618 Generate output for a member of the ColdFire 528X family.
11619 The option is now deprecated in favor of the equivalent
11620 @option{-mcpu=528x}.
11624 Generate output for a ColdFire 5307 CPU@. The option is now deprecated
11625 in favor of the equivalent @option{-mcpu=5307}.
11629 Generate output for a ColdFire 5407 CPU@. The option is now deprecated
11630 in favor of the equivalent @option{-mcpu=5407}.
11634 Generate output for a ColdFire V4e family CPU (e.g.@: 547x/548x).
11635 This includes use of hardware floating point instructions.
11636 The option is equivalent to @option{-mcpu=547x}, and is now
11637 deprecated in favor of that option.
11641 Generate output for a 68040, without using any of the new instructions.
11642 This results in code which can run relatively efficiently on either a
11643 68020/68881 or a 68030 or a 68040. The generated code does use the
11644 68881 instructions that are emulated on the 68040.
11646 The option is equivalent to @option{-march=68020} @option{-mtune=68020-40}.
11650 Generate output for a 68060, without using any of the new instructions.
11651 This results in code which can run relatively efficiently on either a
11652 68020/68881 or a 68030 or a 68040. The generated code does use the
11653 68881 instructions that are emulated on the 68060.
11655 The option is equivalent to @option{-march=68020} @option{-mtune=68020-60}.
11659 @opindex mhard-float
11661 Generate floating-point instructions. This is the default for 68020
11662 and above, and for ColdFire devices that have an FPU@. It defines the
11663 macro @samp{__HAVE_68881__} on M680x0 targets and @samp{__mcffpu__}
11664 on ColdFire targets.
11667 @opindex msoft-float
11668 Do not generate floating-point instructions; use library calls instead.
11669 This is the default for 68000, 68010, and 68832 targets. It is also
11670 the default for ColdFire devices that have no FPU.
11676 Generate (do not generate) ColdFire hardware divide and remainder
11677 instructions. If @option{-march} is used without @option{-mcpu},
11678 the default is ``on'' for ColdFire architectures and ``off'' for M680x0
11679 architectures. Otherwise, the default is taken from the target CPU
11680 (either the default CPU, or the one specified by @option{-mcpu}). For
11681 example, the default is ``off'' for @option{-mcpu=5206} and ``on'' for
11682 @option{-mcpu=5206e}.
11684 gcc defines the macro @samp{__mcfhwdiv__} when this option is enabled.
11688 Consider type @code{int} to be 16 bits wide, like @code{short int}.
11689 Additionally, parameters passed on the stack are also aligned to a
11690 16-bit boundary even on targets whose API mandates promotion to 32-bit.
11694 Do not consider type @code{int} to be 16 bits wide. This is the default.
11697 @itemx -mno-bitfield
11698 @opindex mnobitfield
11699 @opindex mno-bitfield
11700 Do not use the bit-field instructions. The @option{-m68000}, @option{-mcpu32}
11701 and @option{-m5200} options imply @w{@option{-mnobitfield}}.
11705 Do use the bit-field instructions. The @option{-m68020} option implies
11706 @option{-mbitfield}. This is the default if you use a configuration
11707 designed for a 68020.
11711 Use a different function-calling convention, in which functions
11712 that take a fixed number of arguments return with the @code{rtd}
11713 instruction, which pops their arguments while returning. This
11714 saves one instruction in the caller since there is no need to pop
11715 the arguments there.
11717 This calling convention is incompatible with the one normally
11718 used on Unix, so you cannot use it if you need to call libraries
11719 compiled with the Unix compiler.
11721 Also, you must provide function prototypes for all functions that
11722 take variable numbers of arguments (including @code{printf});
11723 otherwise incorrect code will be generated for calls to those
11726 In addition, seriously incorrect code will result if you call a
11727 function with too many arguments. (Normally, extra arguments are
11728 harmlessly ignored.)
11730 The @code{rtd} instruction is supported by the 68010, 68020, 68030,
11731 68040, 68060 and CPU32 processors, but not by the 68000 or 5200.
11735 Do not use the calling conventions selected by @option{-mrtd}.
11736 This is the default.
11739 @itemx -mno-align-int
11740 @opindex malign-int
11741 @opindex mno-align-int
11742 Control whether GCC aligns @code{int}, @code{long}, @code{long long},
11743 @code{float}, @code{double}, and @code{long double} variables on a 32-bit
11744 boundary (@option{-malign-int}) or a 16-bit boundary (@option{-mno-align-int}).
11745 Aligning variables on 32-bit boundaries produces code that runs somewhat
11746 faster on processors with 32-bit busses at the expense of more memory.
11748 @strong{Warning:} if you use the @option{-malign-int} switch, GCC will
11749 align structures containing the above types differently than
11750 most published application binary interface specifications for the m68k.
11754 Use the pc-relative addressing mode of the 68000 directly, instead of
11755 using a global offset table. At present, this option implies @option{-fpic},
11756 allowing at most a 16-bit offset for pc-relative addressing. @option{-fPIC} is
11757 not presently supported with @option{-mpcrel}, though this could be supported for
11758 68020 and higher processors.
11760 @item -mno-strict-align
11761 @itemx -mstrict-align
11762 @opindex mno-strict-align
11763 @opindex mstrict-align
11764 Do not (do) assume that unaligned memory references will be handled by
11768 Generate code that allows the data segment to be located in a different
11769 area of memory from the text segment. This allows for execute in place in
11770 an environment without virtual memory management. This option implies
11773 @item -mno-sep-data
11774 Generate code that assumes that the data segment follows the text segment.
11775 This is the default.
11777 @item -mid-shared-library
11778 Generate code that supports shared libraries via the library ID method.
11779 This allows for execute in place and shared libraries in an environment
11780 without virtual memory management. This option implies @option{-fPIC}.
11782 @item -mno-id-shared-library
11783 Generate code that doesn't assume ID based shared libraries are being used.
11784 This is the default.
11786 @item -mshared-library-id=n
11787 Specified the identification number of the ID based shared library being
11788 compiled. Specifying a value of 0 will generate more compact code, specifying
11789 other values will force the allocation of that number to the current
11790 library but is no more space or time efficient than omitting this option.
11796 When generating position-independent code for ColdFire, generate code
11797 that works if the GOT has more than 8192 entries. This code is
11798 larger and slower than code generated without this option. On M680x0
11799 processors, this option is not needed; @option{-fPIC} suffices.
11801 GCC normally uses a single instruction to load values from the GOT@.
11802 While this is relatively efficient, it only works if the GOT
11803 is smaller than about 64k. Anything larger causes the linker
11804 to report an error such as:
11806 @cindex relocation truncated to fit (ColdFire)
11808 relocation truncated to fit: R_68K_GOT16O foobar
11811 If this happens, you should recompile your code with @option{-mxgot}.
11812 It should then work with very large GOTs. However, code generated with
11813 @option{-mxgot} is less efficient, since it takes 4 instructions to fetch
11814 the value of a global symbol.
11816 Note that some linkers, including newer versions of the GNU linker,
11817 can create multiple GOTs and sort GOT entries. If you have such a linker,
11818 you should only need to use @option{-mxgot} when compiling a single
11819 object file that accesses more than 8192 GOT entries. Very few do.
11821 These options have no effect unless GCC is generating
11822 position-independent code.
11826 @node M68hc1x Options
11827 @subsection M68hc1x Options
11828 @cindex M68hc1x options
11830 These are the @samp{-m} options defined for the 68hc11 and 68hc12
11831 microcontrollers. The default values for these options depends on
11832 which style of microcontroller was selected when the compiler was configured;
11833 the defaults for the most common choices are given below.
11840 Generate output for a 68HC11. This is the default
11841 when the compiler is configured for 68HC11-based systems.
11847 Generate output for a 68HC12. This is the default
11848 when the compiler is configured for 68HC12-based systems.
11854 Generate output for a 68HCS12.
11856 @item -mauto-incdec
11857 @opindex mauto-incdec
11858 Enable the use of 68HC12 pre and post auto-increment and auto-decrement
11865 Enable the use of 68HC12 min and max instructions.
11868 @itemx -mno-long-calls
11869 @opindex mlong-calls
11870 @opindex mno-long-calls
11871 Treat all calls as being far away (near). If calls are assumed to be
11872 far away, the compiler will use the @code{call} instruction to
11873 call a function and the @code{rtc} instruction for returning.
11877 Consider type @code{int} to be 16 bits wide, like @code{short int}.
11879 @item -msoft-reg-count=@var{count}
11880 @opindex msoft-reg-count
11881 Specify the number of pseudo-soft registers which are used for the
11882 code generation. The maximum number is 32. Using more pseudo-soft
11883 register may or may not result in better code depending on the program.
11884 The default is 4 for 68HC11 and 2 for 68HC12.
11888 @node MCore Options
11889 @subsection MCore Options
11890 @cindex MCore options
11892 These are the @samp{-m} options defined for the Motorola M*Core
11898 @itemx -mno-hardlit
11900 @opindex mno-hardlit
11901 Inline constants into the code stream if it can be done in two
11902 instructions or less.
11908 Use the divide instruction. (Enabled by default).
11910 @item -mrelax-immediate
11911 @itemx -mno-relax-immediate
11912 @opindex mrelax-immediate
11913 @opindex mno-relax-immediate
11914 Allow arbitrary sized immediates in bit operations.
11916 @item -mwide-bitfields
11917 @itemx -mno-wide-bitfields
11918 @opindex mwide-bitfields
11919 @opindex mno-wide-bitfields
11920 Always treat bit-fields as int-sized.
11922 @item -m4byte-functions
11923 @itemx -mno-4byte-functions
11924 @opindex m4byte-functions
11925 @opindex mno-4byte-functions
11926 Force all functions to be aligned to a four byte boundary.
11928 @item -mcallgraph-data
11929 @itemx -mno-callgraph-data
11930 @opindex mcallgraph-data
11931 @opindex mno-callgraph-data
11932 Emit callgraph information.
11935 @itemx -mno-slow-bytes
11936 @opindex mslow-bytes
11937 @opindex mno-slow-bytes
11938 Prefer word access when reading byte quantities.
11940 @item -mlittle-endian
11941 @itemx -mbig-endian
11942 @opindex mlittle-endian
11943 @opindex mbig-endian
11944 Generate code for a little endian target.
11950 Generate code for the 210 processor.
11954 @subsection MIPS Options
11955 @cindex MIPS options
11961 Generate big-endian code.
11965 Generate little-endian code. This is the default for @samp{mips*el-*-*}
11968 @item -march=@var{arch}
11970 Generate code that will run on @var{arch}, which can be the name of a
11971 generic MIPS ISA, or the name of a particular processor.
11973 @samp{mips1}, @samp{mips2}, @samp{mips3}, @samp{mips4},
11974 @samp{mips32}, @samp{mips32r2}, and @samp{mips64}.
11975 The processor names are:
11976 @samp{4kc}, @samp{4km}, @samp{4kp}, @samp{4ksc},
11977 @samp{4kec}, @samp{4kem}, @samp{4kep}, @samp{4ksd},
11978 @samp{5kc}, @samp{5kf},
11980 @samp{24kc}, @samp{24kf2_1}, @samp{24kf1_1},
11981 @samp{24kec}, @samp{24kef2_1}, @samp{24kef1_1},
11982 @samp{34kc}, @samp{34kf2_1}, @samp{34kf1_1},
11983 @samp{74kc}, @samp{74kf2_1}, @samp{74kf1_1}, @samp{74kf3_2},
11984 @samp{loongson2e}, @samp{loongson2f},
11987 @samp{r2000}, @samp{r3000}, @samp{r3900}, @samp{r4000}, @samp{r4400},
11988 @samp{r4600}, @samp{r4650}, @samp{r6000}, @samp{r8000},
11989 @samp{rm7000}, @samp{rm9000},
11992 @samp{vr4100}, @samp{vr4111}, @samp{vr4120}, @samp{vr4130}, @samp{vr4300},
11993 @samp{vr5000}, @samp{vr5400}, @samp{vr5500}
11995 The special value @samp{from-abi} selects the
11996 most compatible architecture for the selected ABI (that is,
11997 @samp{mips1} for 32-bit ABIs and @samp{mips3} for 64-bit ABIs)@.
11999 Native Linux/GNU toolchains also support the value @samp{native},
12000 which selects the best architecture option for the host processor.
12001 @option{-march=native} has no effect if GCC does not recognize
12004 In processor names, a final @samp{000} can be abbreviated as @samp{k}
12005 (for example, @samp{-march=r2k}). Prefixes are optional, and
12006 @samp{vr} may be written @samp{r}.
12008 Names of the form @samp{@var{n}f2_1} refer to processors with
12009 FPUs clocked at half the rate of the core, names of the form
12010 @samp{@var{n}f1_1} refer to processors with FPUs clocked at the same
12011 rate as the core, and names of the form @samp{@var{n}f3_2} refer to
12012 processors with FPUs clocked a ratio of 3:2 with respect to the core.
12013 For compatibility reasons, @samp{@var{n}f} is accepted as a synonym
12014 for @samp{@var{n}f2_1} while @samp{@var{n}x} and @samp{@var{b}fx} are
12015 accepted as synonyms for @samp{@var{n}f1_1}.
12017 GCC defines two macros based on the value of this option. The first
12018 is @samp{_MIPS_ARCH}, which gives the name of target architecture, as
12019 a string. The second has the form @samp{_MIPS_ARCH_@var{foo}},
12020 where @var{foo} is the capitalized value of @samp{_MIPS_ARCH}@.
12021 For example, @samp{-march=r2000} will set @samp{_MIPS_ARCH}
12022 to @samp{"r2000"} and define the macro @samp{_MIPS_ARCH_R2000}.
12024 Note that the @samp{_MIPS_ARCH} macro uses the processor names given
12025 above. In other words, it will have the full prefix and will not
12026 abbreviate @samp{000} as @samp{k}. In the case of @samp{from-abi},
12027 the macro names the resolved architecture (either @samp{"mips1"} or
12028 @samp{"mips3"}). It names the default architecture when no
12029 @option{-march} option is given.
12031 @item -mtune=@var{arch}
12033 Optimize for @var{arch}. Among other things, this option controls
12034 the way instructions are scheduled, and the perceived cost of arithmetic
12035 operations. The list of @var{arch} values is the same as for
12038 When this option is not used, GCC will optimize for the processor
12039 specified by @option{-march}. By using @option{-march} and
12040 @option{-mtune} together, it is possible to generate code that will
12041 run on a family of processors, but optimize the code for one
12042 particular member of that family.
12044 @samp{-mtune} defines the macros @samp{_MIPS_TUNE} and
12045 @samp{_MIPS_TUNE_@var{foo}}, which work in the same way as the
12046 @samp{-march} ones described above.
12050 Equivalent to @samp{-march=mips1}.
12054 Equivalent to @samp{-march=mips2}.
12058 Equivalent to @samp{-march=mips3}.
12062 Equivalent to @samp{-march=mips4}.
12066 Equivalent to @samp{-march=mips32}.
12070 Equivalent to @samp{-march=mips32r2}.
12074 Equivalent to @samp{-march=mips64}.
12079 @opindex mno-mips16
12080 Generate (do not generate) MIPS16 code. If GCC is targetting a
12081 MIPS32 or MIPS64 architecture, it will make use of the MIPS16e ASE@.
12083 MIPS16 code generation can also be controlled on a per-function basis
12084 by means of @code{mips16} and @code{nomips16} attributes.
12085 @xref{Function Attributes}, for more information.
12087 @item -mflip-mips16
12088 @opindex mflip-mips16
12089 Generate MIPS16 code on alternating functions. This option is provided
12090 for regression testing of mixed MIPS16/non-MIPS16 code generation, and is
12091 not intended for ordinary use in compiling user code.
12093 @item -minterlink-mips16
12094 @itemx -mno-interlink-mips16
12095 @opindex minterlink-mips16
12096 @opindex mno-interlink-mips16
12097 Require (do not require) that non-MIPS16 code be link-compatible with
12100 For example, non-MIPS16 code cannot jump directly to MIPS16 code;
12101 it must either use a call or an indirect jump. @option{-minterlink-mips16}
12102 therefore disables direct jumps unless GCC knows that the target of the
12103 jump is not MIPS16.
12115 Generate code for the given ABI@.
12117 Note that the EABI has a 32-bit and a 64-bit variant. GCC normally
12118 generates 64-bit code when you select a 64-bit architecture, but you
12119 can use @option{-mgp32} to get 32-bit code instead.
12121 For information about the O64 ABI, see
12122 @w{@uref{http://gcc.gnu.org/projects/mipso64-abi.html}}.
12124 GCC supports a variant of the o32 ABI in which floating-point registers
12125 are 64 rather than 32 bits wide. You can select this combination with
12126 @option{-mabi=32} @option{-mfp64}. This ABI relies on the @samp{mthc1}
12127 and @samp{mfhc1} instructions and is therefore only supported for
12128 MIPS32R2 processors.
12130 The register assignments for arguments and return values remain the
12131 same, but each scalar value is passed in a single 64-bit register
12132 rather than a pair of 32-bit registers. For example, scalar
12133 floating-point values are returned in @samp{$f0} only, not a
12134 @samp{$f0}/@samp{$f1} pair. The set of call-saved registers also
12135 remains the same, but all 64 bits are saved.
12138 @itemx -mno-abicalls
12140 @opindex mno-abicalls
12141 Generate (do not generate) code that is suitable for SVR4-style
12142 dynamic objects. @option{-mabicalls} is the default for SVR4-based
12147 Generate (do not generate) code that is fully position-independent,
12148 and that can therefore be linked into shared libraries. This option
12149 only affects @option{-mabicalls}.
12151 All @option{-mabicalls} code has traditionally been position-independent,
12152 regardless of options like @option{-fPIC} and @option{-fpic}. However,
12153 as an extension, the GNU toolchain allows executables to use absolute
12154 accesses for locally-binding symbols. It can also use shorter GP
12155 initialization sequences and generate direct calls to locally-defined
12156 functions. This mode is selected by @option{-mno-shared}.
12158 @option{-mno-shared} depends on binutils 2.16 or higher and generates
12159 objects that can only be linked by the GNU linker. However, the option
12160 does not affect the ABI of the final executable; it only affects the ABI
12161 of relocatable objects. Using @option{-mno-shared} will generally make
12162 executables both smaller and quicker.
12164 @option{-mshared} is the default.
12170 Lift (do not lift) the usual restrictions on the size of the global
12173 GCC normally uses a single instruction to load values from the GOT@.
12174 While this is relatively efficient, it will only work if the GOT
12175 is smaller than about 64k. Anything larger will cause the linker
12176 to report an error such as:
12178 @cindex relocation truncated to fit (MIPS)
12180 relocation truncated to fit: R_MIPS_GOT16 foobar
12183 If this happens, you should recompile your code with @option{-mxgot}.
12184 It should then work with very large GOTs, although it will also be
12185 less efficient, since it will take three instructions to fetch the
12186 value of a global symbol.
12188 Note that some linkers can create multiple GOTs. If you have such a
12189 linker, you should only need to use @option{-mxgot} when a single object
12190 file accesses more than 64k's worth of GOT entries. Very few do.
12192 These options have no effect unless GCC is generating position
12197 Assume that general-purpose registers are 32 bits wide.
12201 Assume that general-purpose registers are 64 bits wide.
12205 Assume that floating-point registers are 32 bits wide.
12209 Assume that floating-point registers are 64 bits wide.
12212 @opindex mhard-float
12213 Use floating-point coprocessor instructions.
12216 @opindex msoft-float
12217 Do not use floating-point coprocessor instructions. Implement
12218 floating-point calculations using library calls instead.
12220 @item -msingle-float
12221 @opindex msingle-float
12222 Assume that the floating-point coprocessor only supports single-precision
12225 @item -mdouble-float
12226 @opindex mdouble-float
12227 Assume that the floating-point coprocessor supports double-precision
12228 operations. This is the default.
12234 Use (do not use) @samp{ll}, @samp{sc}, and @samp{sync} instructions to
12235 implement atomic memory built-in functions. When neither option is
12236 specified, GCC will use the instructions if the target architecture
12239 @option{-mllsc} is useful if the runtime environment can emulate the
12240 instructions and @option{-mno-llsc} can be useful when compiling for
12241 nonstandard ISAs. You can make either option the default by
12242 configuring GCC with @option{--with-llsc} and @option{--without-llsc}
12243 respectively. @option{--with-llsc} is the default for some
12244 configurations; see the installation documentation for details.
12250 Use (do not use) revision 1 of the MIPS DSP ASE@.
12251 @xref{MIPS DSP Built-in Functions}. This option defines the
12252 preprocessor macro @samp{__mips_dsp}. It also defines
12253 @samp{__mips_dsp_rev} to 1.
12259 Use (do not use) revision 2 of the MIPS DSP ASE@.
12260 @xref{MIPS DSP Built-in Functions}. This option defines the
12261 preprocessor macros @samp{__mips_dsp} and @samp{__mips_dspr2}.
12262 It also defines @samp{__mips_dsp_rev} to 2.
12265 @itemx -mno-smartmips
12266 @opindex msmartmips
12267 @opindex mno-smartmips
12268 Use (do not use) the MIPS SmartMIPS ASE.
12270 @item -mpaired-single
12271 @itemx -mno-paired-single
12272 @opindex mpaired-single
12273 @opindex mno-paired-single
12274 Use (do not use) paired-single floating-point instructions.
12275 @xref{MIPS Paired-Single Support}. This option requires
12276 hardware floating-point support to be enabled.
12282 Use (do not use) MIPS Digital Media Extension instructions.
12283 This option can only be used when generating 64-bit code and requires
12284 hardware floating-point support to be enabled.
12289 @opindex mno-mips3d
12290 Use (do not use) the MIPS-3D ASE@. @xref{MIPS-3D Built-in Functions}.
12291 The option @option{-mips3d} implies @option{-mpaired-single}.
12297 Use (do not use) MT Multithreading instructions.
12301 Force @code{long} types to be 64 bits wide. See @option{-mlong32} for
12302 an explanation of the default and the way that the pointer size is
12307 Force @code{long}, @code{int}, and pointer types to be 32 bits wide.
12309 The default size of @code{int}s, @code{long}s and pointers depends on
12310 the ABI@. All the supported ABIs use 32-bit @code{int}s. The n64 ABI
12311 uses 64-bit @code{long}s, as does the 64-bit EABI; the others use
12312 32-bit @code{long}s. Pointers are the same size as @code{long}s,
12313 or the same size as integer registers, whichever is smaller.
12319 Assume (do not assume) that all symbols have 32-bit values, regardless
12320 of the selected ABI@. This option is useful in combination with
12321 @option{-mabi=64} and @option{-mno-abicalls} because it allows GCC
12322 to generate shorter and faster references to symbolic addresses.
12326 Put definitions of externally-visible data in a small data section
12327 if that data is no bigger than @var{num} bytes. GCC can then access
12328 the data more efficiently; see @option{-mgpopt} for details.
12330 The default @option{-G} option depends on the configuration.
12332 @item -mlocal-sdata
12333 @itemx -mno-local-sdata
12334 @opindex mlocal-sdata
12335 @opindex mno-local-sdata
12336 Extend (do not extend) the @option{-G} behavior to local data too,
12337 such as to static variables in C@. @option{-mlocal-sdata} is the
12338 default for all configurations.
12340 If the linker complains that an application is using too much small data,
12341 you might want to try rebuilding the less performance-critical parts with
12342 @option{-mno-local-sdata}. You might also want to build large
12343 libraries with @option{-mno-local-sdata}, so that the libraries leave
12344 more room for the main program.
12346 @item -mextern-sdata
12347 @itemx -mno-extern-sdata
12348 @opindex mextern-sdata
12349 @opindex mno-extern-sdata
12350 Assume (do not assume) that externally-defined data will be in
12351 a small data section if that data is within the @option{-G} limit.
12352 @option{-mextern-sdata} is the default for all configurations.
12354 If you compile a module @var{Mod} with @option{-mextern-sdata} @option{-G
12355 @var{num}} @option{-mgpopt}, and @var{Mod} references a variable @var{Var}
12356 that is no bigger than @var{num} bytes, you must make sure that @var{Var}
12357 is placed in a small data section. If @var{Var} is defined by another
12358 module, you must either compile that module with a high-enough
12359 @option{-G} setting or attach a @code{section} attribute to @var{Var}'s
12360 definition. If @var{Var} is common, you must link the application
12361 with a high-enough @option{-G} setting.
12363 The easiest way of satisfying these restrictions is to compile
12364 and link every module with the same @option{-G} option. However,
12365 you may wish to build a library that supports several different
12366 small data limits. You can do this by compiling the library with
12367 the highest supported @option{-G} setting and additionally using
12368 @option{-mno-extern-sdata} to stop the library from making assumptions
12369 about externally-defined data.
12375 Use (do not use) GP-relative accesses for symbols that are known to be
12376 in a small data section; see @option{-G}, @option{-mlocal-sdata} and
12377 @option{-mextern-sdata}. @option{-mgpopt} is the default for all
12380 @option{-mno-gpopt} is useful for cases where the @code{$gp} register
12381 might not hold the value of @code{_gp}. For example, if the code is
12382 part of a library that might be used in a boot monitor, programs that
12383 call boot monitor routines will pass an unknown value in @code{$gp}.
12384 (In such situations, the boot monitor itself would usually be compiled
12385 with @option{-G0}.)
12387 @option{-mno-gpopt} implies @option{-mno-local-sdata} and
12388 @option{-mno-extern-sdata}.
12390 @item -membedded-data
12391 @itemx -mno-embedded-data
12392 @opindex membedded-data
12393 @opindex mno-embedded-data
12394 Allocate variables to the read-only data section first if possible, then
12395 next in the small data section if possible, otherwise in data. This gives
12396 slightly slower code than the default, but reduces the amount of RAM required
12397 when executing, and thus may be preferred for some embedded systems.
12399 @item -muninit-const-in-rodata
12400 @itemx -mno-uninit-const-in-rodata
12401 @opindex muninit-const-in-rodata
12402 @opindex mno-uninit-const-in-rodata
12403 Put uninitialized @code{const} variables in the read-only data section.
12404 This option is only meaningful in conjunction with @option{-membedded-data}.
12406 @item -mcode-readable=@var{setting}
12407 @opindex mcode-readable
12408 Specify whether GCC may generate code that reads from executable sections.
12409 There are three possible settings:
12412 @item -mcode-readable=yes
12413 Instructions may freely access executable sections. This is the
12416 @item -mcode-readable=pcrel
12417 MIPS16 PC-relative load instructions can access executable sections,
12418 but other instructions must not do so. This option is useful on 4KSc
12419 and 4KSd processors when the code TLBs have the Read Inhibit bit set.
12420 It is also useful on processors that can be configured to have a dual
12421 instruction/data SRAM interface and that, like the M4K, automatically
12422 redirect PC-relative loads to the instruction RAM.
12424 @item -mcode-readable=no
12425 Instructions must not access executable sections. This option can be
12426 useful on targets that are configured to have a dual instruction/data
12427 SRAM interface but that (unlike the M4K) do not automatically redirect
12428 PC-relative loads to the instruction RAM.
12431 @item -msplit-addresses
12432 @itemx -mno-split-addresses
12433 @opindex msplit-addresses
12434 @opindex mno-split-addresses
12435 Enable (disable) use of the @code{%hi()} and @code{%lo()} assembler
12436 relocation operators. This option has been superseded by
12437 @option{-mexplicit-relocs} but is retained for backwards compatibility.
12439 @item -mexplicit-relocs
12440 @itemx -mno-explicit-relocs
12441 @opindex mexplicit-relocs
12442 @opindex mno-explicit-relocs
12443 Use (do not use) assembler relocation operators when dealing with symbolic
12444 addresses. The alternative, selected by @option{-mno-explicit-relocs},
12445 is to use assembler macros instead.
12447 @option{-mexplicit-relocs} is the default if GCC was configured
12448 to use an assembler that supports relocation operators.
12450 @item -mcheck-zero-division
12451 @itemx -mno-check-zero-division
12452 @opindex mcheck-zero-division
12453 @opindex mno-check-zero-division
12454 Trap (do not trap) on integer division by zero.
12456 The default is @option{-mcheck-zero-division}.
12458 @item -mdivide-traps
12459 @itemx -mdivide-breaks
12460 @opindex mdivide-traps
12461 @opindex mdivide-breaks
12462 MIPS systems check for division by zero by generating either a
12463 conditional trap or a break instruction. Using traps results in
12464 smaller code, but is only supported on MIPS II and later. Also, some
12465 versions of the Linux kernel have a bug that prevents trap from
12466 generating the proper signal (@code{SIGFPE}). Use @option{-mdivide-traps} to
12467 allow conditional traps on architectures that support them and
12468 @option{-mdivide-breaks} to force the use of breaks.
12470 The default is usually @option{-mdivide-traps}, but this can be
12471 overridden at configure time using @option{--with-divide=breaks}.
12472 Divide-by-zero checks can be completely disabled using
12473 @option{-mno-check-zero-division}.
12478 @opindex mno-memcpy
12479 Force (do not force) the use of @code{memcpy()} for non-trivial block
12480 moves. The default is @option{-mno-memcpy}, which allows GCC to inline
12481 most constant-sized copies.
12484 @itemx -mno-long-calls
12485 @opindex mlong-calls
12486 @opindex mno-long-calls
12487 Disable (do not disable) use of the @code{jal} instruction. Calling
12488 functions using @code{jal} is more efficient but requires the caller
12489 and callee to be in the same 256 megabyte segment.
12491 This option has no effect on abicalls code. The default is
12492 @option{-mno-long-calls}.
12498 Enable (disable) use of the @code{mad}, @code{madu} and @code{mul}
12499 instructions, as provided by the R4650 ISA@.
12502 @itemx -mno-fused-madd
12503 @opindex mfused-madd
12504 @opindex mno-fused-madd
12505 Enable (disable) use of the floating point multiply-accumulate
12506 instructions, when they are available. The default is
12507 @option{-mfused-madd}.
12509 When multiply-accumulate instructions are used, the intermediate
12510 product is calculated to infinite precision and is not subject to
12511 the FCSR Flush to Zero bit. This may be undesirable in some
12516 Tell the MIPS assembler to not run its preprocessor over user
12517 assembler files (with a @samp{.s} suffix) when assembling them.
12520 @itemx -mno-fix-r4000
12521 @opindex mfix-r4000
12522 @opindex mno-fix-r4000
12523 Work around certain R4000 CPU errata:
12526 A double-word or a variable shift may give an incorrect result if executed
12527 immediately after starting an integer division.
12529 A double-word or a variable shift may give an incorrect result if executed
12530 while an integer multiplication is in progress.
12532 An integer division may give an incorrect result if started in a delay slot
12533 of a taken branch or a jump.
12537 @itemx -mno-fix-r4400
12538 @opindex mfix-r4400
12539 @opindex mno-fix-r4400
12540 Work around certain R4400 CPU errata:
12543 A double-word or a variable shift may give an incorrect result if executed
12544 immediately after starting an integer division.
12548 @itemx -mno-fix-vr4120
12549 @opindex mfix-vr4120
12550 Work around certain VR4120 errata:
12553 @code{dmultu} does not always produce the correct result.
12555 @code{div} and @code{ddiv} do not always produce the correct result if one
12556 of the operands is negative.
12558 The workarounds for the division errata rely on special functions in
12559 @file{libgcc.a}. At present, these functions are only provided by
12560 the @code{mips64vr*-elf} configurations.
12562 Other VR4120 errata require a nop to be inserted between certain pairs of
12563 instructions. These errata are handled by the assembler, not by GCC itself.
12566 @opindex mfix-vr4130
12567 Work around the VR4130 @code{mflo}/@code{mfhi} errata. The
12568 workarounds are implemented by the assembler rather than by GCC,
12569 although GCC will avoid using @code{mflo} and @code{mfhi} if the
12570 VR4130 @code{macc}, @code{macchi}, @code{dmacc} and @code{dmacchi}
12571 instructions are available instead.
12574 @itemx -mno-fix-sb1
12576 Work around certain SB-1 CPU core errata.
12577 (This flag currently works around the SB-1 revision 2
12578 ``F1'' and ``F2'' floating point errata.)
12580 @item -mflush-func=@var{func}
12581 @itemx -mno-flush-func
12582 @opindex mflush-func
12583 Specifies the function to call to flush the I and D caches, or to not
12584 call any such function. If called, the function must take the same
12585 arguments as the common @code{_flush_func()}, that is, the address of the
12586 memory range for which the cache is being flushed, the size of the
12587 memory range, and the number 3 (to flush both caches). The default
12588 depends on the target GCC was configured for, but commonly is either
12589 @samp{_flush_func} or @samp{__cpu_flush}.
12591 @item mbranch-cost=@var{num}
12592 @opindex mbranch-cost
12593 Set the cost of branches to roughly @var{num} ``simple'' instructions.
12594 This cost is only a heuristic and is not guaranteed to produce
12595 consistent results across releases. A zero cost redundantly selects
12596 the default, which is based on the @option{-mtune} setting.
12598 @item -mbranch-likely
12599 @itemx -mno-branch-likely
12600 @opindex mbranch-likely
12601 @opindex mno-branch-likely
12602 Enable or disable use of Branch Likely instructions, regardless of the
12603 default for the selected architecture. By default, Branch Likely
12604 instructions may be generated if they are supported by the selected
12605 architecture. An exception is for the MIPS32 and MIPS64 architectures
12606 and processors which implement those architectures; for those, Branch
12607 Likely instructions will not be generated by default because the MIPS32
12608 and MIPS64 architectures specifically deprecate their use.
12610 @item -mfp-exceptions
12611 @itemx -mno-fp-exceptions
12612 @opindex mfp-exceptions
12613 Specifies whether FP exceptions are enabled. This affects how we schedule
12614 FP instructions for some processors. The default is that FP exceptions are
12617 For instance, on the SB-1, if FP exceptions are disabled, and we are emitting
12618 64-bit code, then we can use both FP pipes. Otherwise, we can only use one
12621 @item -mvr4130-align
12622 @itemx -mno-vr4130-align
12623 @opindex mvr4130-align
12624 The VR4130 pipeline is two-way superscalar, but can only issue two
12625 instructions together if the first one is 8-byte aligned. When this
12626 option is enabled, GCC will align pairs of instructions that it
12627 thinks should execute in parallel.
12629 This option only has an effect when optimizing for the VR4130.
12630 It normally makes code faster, but at the expense of making it bigger.
12631 It is enabled by default at optimization level @option{-O3}.
12635 @subsection MMIX Options
12636 @cindex MMIX Options
12638 These options are defined for the MMIX:
12642 @itemx -mno-libfuncs
12644 @opindex mno-libfuncs
12645 Specify that intrinsic library functions are being compiled, passing all
12646 values in registers, no matter the size.
12649 @itemx -mno-epsilon
12651 @opindex mno-epsilon
12652 Generate floating-point comparison instructions that compare with respect
12653 to the @code{rE} epsilon register.
12655 @item -mabi=mmixware
12657 @opindex mabi-mmixware
12659 Generate code that passes function parameters and return values that (in
12660 the called function) are seen as registers @code{$0} and up, as opposed to
12661 the GNU ABI which uses global registers @code{$231} and up.
12663 @item -mzero-extend
12664 @itemx -mno-zero-extend
12665 @opindex mzero-extend
12666 @opindex mno-zero-extend
12667 When reading data from memory in sizes shorter than 64 bits, use (do not
12668 use) zero-extending load instructions by default, rather than
12669 sign-extending ones.
12672 @itemx -mno-knuthdiv
12674 @opindex mno-knuthdiv
12675 Make the result of a division yielding a remainder have the same sign as
12676 the divisor. With the default, @option{-mno-knuthdiv}, the sign of the
12677 remainder follows the sign of the dividend. Both methods are
12678 arithmetically valid, the latter being almost exclusively used.
12680 @item -mtoplevel-symbols
12681 @itemx -mno-toplevel-symbols
12682 @opindex mtoplevel-symbols
12683 @opindex mno-toplevel-symbols
12684 Prepend (do not prepend) a @samp{:} to all global symbols, so the assembly
12685 code can be used with the @code{PREFIX} assembly directive.
12689 Generate an executable in the ELF format, rather than the default
12690 @samp{mmo} format used by the @command{mmix} simulator.
12692 @item -mbranch-predict
12693 @itemx -mno-branch-predict
12694 @opindex mbranch-predict
12695 @opindex mno-branch-predict
12696 Use (do not use) the probable-branch instructions, when static branch
12697 prediction indicates a probable branch.
12699 @item -mbase-addresses
12700 @itemx -mno-base-addresses
12701 @opindex mbase-addresses
12702 @opindex mno-base-addresses
12703 Generate (do not generate) code that uses @emph{base addresses}. Using a
12704 base address automatically generates a request (handled by the assembler
12705 and the linker) for a constant to be set up in a global register. The
12706 register is used for one or more base address requests within the range 0
12707 to 255 from the value held in the register. The generally leads to short
12708 and fast code, but the number of different data items that can be
12709 addressed is limited. This means that a program that uses lots of static
12710 data may require @option{-mno-base-addresses}.
12712 @item -msingle-exit
12713 @itemx -mno-single-exit
12714 @opindex msingle-exit
12715 @opindex mno-single-exit
12716 Force (do not force) generated code to have a single exit point in each
12720 @node MN10300 Options
12721 @subsection MN10300 Options
12722 @cindex MN10300 options
12724 These @option{-m} options are defined for Matsushita MN10300 architectures:
12729 Generate code to avoid bugs in the multiply instructions for the MN10300
12730 processors. This is the default.
12732 @item -mno-mult-bug
12733 @opindex mno-mult-bug
12734 Do not generate code to avoid bugs in the multiply instructions for the
12735 MN10300 processors.
12739 Generate code which uses features specific to the AM33 processor.
12743 Do not generate code which uses features specific to the AM33 processor. This
12746 @item -mreturn-pointer-on-d0
12747 @opindex mreturn-pointer-on-d0
12748 When generating a function which returns a pointer, return the pointer
12749 in both @code{a0} and @code{d0}. Otherwise, the pointer is returned
12750 only in a0, and attempts to call such functions without a prototype
12751 would result in errors. Note that this option is on by default; use
12752 @option{-mno-return-pointer-on-d0} to disable it.
12756 Do not link in the C run-time initialization object file.
12760 Indicate to the linker that it should perform a relaxation optimization pass
12761 to shorten branches, calls and absolute memory addresses. This option only
12762 has an effect when used on the command line for the final link step.
12764 This option makes symbolic debugging impossible.
12767 @node PDP-11 Options
12768 @subsection PDP-11 Options
12769 @cindex PDP-11 Options
12771 These options are defined for the PDP-11:
12776 Use hardware FPP floating point. This is the default. (FIS floating
12777 point on the PDP-11/40 is not supported.)
12780 @opindex msoft-float
12781 Do not use hardware floating point.
12785 Return floating-point results in ac0 (fr0 in Unix assembler syntax).
12789 Return floating-point results in memory. This is the default.
12793 Generate code for a PDP-11/40.
12797 Generate code for a PDP-11/45. This is the default.
12801 Generate code for a PDP-11/10.
12803 @item -mbcopy-builtin
12804 @opindex bcopy-builtin
12805 Use inline @code{movmemhi} patterns for copying memory. This is the
12810 Do not use inline @code{movmemhi} patterns for copying memory.
12816 Use 16-bit @code{int}. This is the default.
12822 Use 32-bit @code{int}.
12825 @itemx -mno-float32
12827 @opindex mno-float32
12828 Use 64-bit @code{float}. This is the default.
12831 @itemx -mno-float64
12833 @opindex mno-float64
12834 Use 32-bit @code{float}.
12838 Use @code{abshi2} pattern. This is the default.
12842 Do not use @code{abshi2} pattern.
12844 @item -mbranch-expensive
12845 @opindex mbranch-expensive
12846 Pretend that branches are expensive. This is for experimenting with
12847 code generation only.
12849 @item -mbranch-cheap
12850 @opindex mbranch-cheap
12851 Do not pretend that branches are expensive. This is the default.
12855 Generate code for a system with split I&D@.
12859 Generate code for a system without split I&D@. This is the default.
12863 Use Unix assembler syntax. This is the default when configured for
12864 @samp{pdp11-*-bsd}.
12868 Use DEC assembler syntax. This is the default when configured for any
12869 PDP-11 target other than @samp{pdp11-*-bsd}.
12872 @node PowerPC Options
12873 @subsection PowerPC Options
12874 @cindex PowerPC options
12876 These are listed under @xref{RS/6000 and PowerPC Options}.
12878 @node RS/6000 and PowerPC Options
12879 @subsection IBM RS/6000 and PowerPC Options
12880 @cindex RS/6000 and PowerPC Options
12881 @cindex IBM RS/6000 and PowerPC Options
12883 These @samp{-m} options are defined for the IBM RS/6000 and PowerPC:
12890 @itemx -mno-powerpc
12891 @itemx -mpowerpc-gpopt
12892 @itemx -mno-powerpc-gpopt
12893 @itemx -mpowerpc-gfxopt
12894 @itemx -mno-powerpc-gfxopt
12896 @itemx -mno-powerpc64
12900 @itemx -mno-popcntb
12908 @itemx -mno-hard-dfp
12912 @opindex mno-power2
12914 @opindex mno-powerpc
12915 @opindex mpowerpc-gpopt
12916 @opindex mno-powerpc-gpopt
12917 @opindex mpowerpc-gfxopt
12918 @opindex mno-powerpc-gfxopt
12919 @opindex mpowerpc64
12920 @opindex mno-powerpc64
12924 @opindex mno-popcntb
12930 @opindex mno-mfpgpr
12932 @opindex mno-hard-dfp
12933 GCC supports two related instruction set architectures for the
12934 RS/6000 and PowerPC@. The @dfn{POWER} instruction set are those
12935 instructions supported by the @samp{rios} chip set used in the original
12936 RS/6000 systems and the @dfn{PowerPC} instruction set is the
12937 architecture of the Freescale MPC5xx, MPC6xx, MPC8xx microprocessors, and
12938 the IBM 4xx, 6xx, and follow-on microprocessors.
12940 Neither architecture is a subset of the other. However there is a
12941 large common subset of instructions supported by both. An MQ
12942 register is included in processors supporting the POWER architecture.
12944 You use these options to specify which instructions are available on the
12945 processor you are using. The default value of these options is
12946 determined when configuring GCC@. Specifying the
12947 @option{-mcpu=@var{cpu_type}} overrides the specification of these
12948 options. We recommend you use the @option{-mcpu=@var{cpu_type}} option
12949 rather than the options listed above.
12951 The @option{-mpower} option allows GCC to generate instructions that
12952 are found only in the POWER architecture and to use the MQ register.
12953 Specifying @option{-mpower2} implies @option{-power} and also allows GCC
12954 to generate instructions that are present in the POWER2 architecture but
12955 not the original POWER architecture.
12957 The @option{-mpowerpc} option allows GCC to generate instructions that
12958 are found only in the 32-bit subset of the PowerPC architecture.
12959 Specifying @option{-mpowerpc-gpopt} implies @option{-mpowerpc} and also allows
12960 GCC to use the optional PowerPC architecture instructions in the
12961 General Purpose group, including floating-point square root. Specifying
12962 @option{-mpowerpc-gfxopt} implies @option{-mpowerpc} and also allows GCC to
12963 use the optional PowerPC architecture instructions in the Graphics
12964 group, including floating-point select.
12966 The @option{-mmfcrf} option allows GCC to generate the move from
12967 condition register field instruction implemented on the POWER4
12968 processor and other processors that support the PowerPC V2.01
12970 The @option{-mpopcntb} option allows GCC to generate the popcount and
12971 double precision FP reciprocal estimate instruction implemented on the
12972 POWER5 processor and other processors that support the PowerPC V2.02
12974 The @option{-mfprnd} option allows GCC to generate the FP round to
12975 integer instructions implemented on the POWER5+ processor and other
12976 processors that support the PowerPC V2.03 architecture.
12977 The @option{-mcmpb} option allows GCC to generate the compare bytes
12978 instruction implemented on the POWER6 processor and other processors
12979 that support the PowerPC V2.05 architecture.
12980 The @option{-mmfpgpr} option allows GCC to generate the FP move to/from
12981 general purpose register instructions implemented on the POWER6X
12982 processor and other processors that support the extended PowerPC V2.05
12984 The @option{-mhard-dfp} option allows GCC to generate the decimal floating
12985 point instructions implemented on some POWER processors.
12987 The @option{-mpowerpc64} option allows GCC to generate the additional
12988 64-bit instructions that are found in the full PowerPC64 architecture
12989 and to treat GPRs as 64-bit, doubleword quantities. GCC defaults to
12990 @option{-mno-powerpc64}.
12992 If you specify both @option{-mno-power} and @option{-mno-powerpc}, GCC
12993 will use only the instructions in the common subset of both
12994 architectures plus some special AIX common-mode calls, and will not use
12995 the MQ register. Specifying both @option{-mpower} and @option{-mpowerpc}
12996 permits GCC to use any instruction from either architecture and to
12997 allow use of the MQ register; specify this for the Motorola MPC601.
12999 @item -mnew-mnemonics
13000 @itemx -mold-mnemonics
13001 @opindex mnew-mnemonics
13002 @opindex mold-mnemonics
13003 Select which mnemonics to use in the generated assembler code. With
13004 @option{-mnew-mnemonics}, GCC uses the assembler mnemonics defined for
13005 the PowerPC architecture. With @option{-mold-mnemonics} it uses the
13006 assembler mnemonics defined for the POWER architecture. Instructions
13007 defined in only one architecture have only one mnemonic; GCC uses that
13008 mnemonic irrespective of which of these options is specified.
13010 GCC defaults to the mnemonics appropriate for the architecture in
13011 use. Specifying @option{-mcpu=@var{cpu_type}} sometimes overrides the
13012 value of these option. Unless you are building a cross-compiler, you
13013 should normally not specify either @option{-mnew-mnemonics} or
13014 @option{-mold-mnemonics}, but should instead accept the default.
13016 @item -mcpu=@var{cpu_type}
13018 Set architecture type, register usage, choice of mnemonics, and
13019 instruction scheduling parameters for machine type @var{cpu_type}.
13020 Supported values for @var{cpu_type} are @samp{401}, @samp{403},
13021 @samp{405}, @samp{405fp}, @samp{440}, @samp{440fp}, @samp{464}, @samp{464fp},
13022 @samp{505}, @samp{601}, @samp{602}, @samp{603}, @samp{603e}, @samp{604},
13023 @samp{604e}, @samp{620}, @samp{630}, @samp{740}, @samp{7400},
13024 @samp{7450}, @samp{750}, @samp{801}, @samp{821}, @samp{823},
13025 @samp{860}, @samp{970}, @samp{8540}, @samp{e300c2}, @samp{e300c3},
13026 @samp{e500mc}, @samp{ec603e}, @samp{G3}, @samp{G4}, @samp{G5},
13027 @samp{power}, @samp{power2}, @samp{power3}, @samp{power4},
13028 @samp{power5}, @samp{power5+}, @samp{power6}, @samp{power6x},
13029 @samp{common}, @samp{powerpc}, @samp{powerpc64}, @samp{rios},
13030 @samp{rios1}, @samp{rios2}, @samp{rsc}, and @samp{rs64}.
13032 @option{-mcpu=common} selects a completely generic processor. Code
13033 generated under this option will run on any POWER or PowerPC processor.
13034 GCC will use only the instructions in the common subset of both
13035 architectures, and will not use the MQ register. GCC assumes a generic
13036 processor model for scheduling purposes.
13038 @option{-mcpu=power}, @option{-mcpu=power2}, @option{-mcpu=powerpc}, and
13039 @option{-mcpu=powerpc64} specify generic POWER, POWER2, pure 32-bit
13040 PowerPC (i.e., not MPC601), and 64-bit PowerPC architecture machine
13041 types, with an appropriate, generic processor model assumed for
13042 scheduling purposes.
13044 The other options specify a specific processor. Code generated under
13045 those options will run best on that processor, and may not run at all on
13048 The @option{-mcpu} options automatically enable or disable the
13051 @gccoptlist{-maltivec -mfprnd -mhard-float -mmfcrf -mmultiple @gol
13052 -mnew-mnemonics -mpopcntb -mpower -mpower2 -mpowerpc64 @gol
13053 -mpowerpc-gpopt -mpowerpc-gfxopt -mstring -mmulhw -mdlmzb -mmfpgpr}
13055 The particular options set for any particular CPU will vary between
13056 compiler versions, depending on what setting seems to produce optimal
13057 code for that CPU; it doesn't necessarily reflect the actual hardware's
13058 capabilities. If you wish to set an individual option to a particular
13059 value, you may specify it after the @option{-mcpu} option, like
13060 @samp{-mcpu=970 -mno-altivec}.
13062 On AIX, the @option{-maltivec} and @option{-mpowerpc64} options are
13063 not enabled or disabled by the @option{-mcpu} option at present because
13064 AIX does not have full support for these options. You may still
13065 enable or disable them individually if you're sure it'll work in your
13068 @item -mtune=@var{cpu_type}
13070 Set the instruction scheduling parameters for machine type
13071 @var{cpu_type}, but do not set the architecture type, register usage, or
13072 choice of mnemonics, as @option{-mcpu=@var{cpu_type}} would. The same
13073 values for @var{cpu_type} are used for @option{-mtune} as for
13074 @option{-mcpu}. If both are specified, the code generated will use the
13075 architecture, registers, and mnemonics set by @option{-mcpu}, but the
13076 scheduling parameters set by @option{-mtune}.
13082 Generate code to compute division as reciprocal estimate and iterative
13083 refinement, creating opportunities for increased throughput. This
13084 feature requires: optional PowerPC Graphics instruction set for single
13085 precision and FRE instruction for double precision, assuming divides
13086 cannot generate user-visible traps, and the domain values not include
13087 Infinities, denormals or zero denominator.
13090 @itemx -mno-altivec
13092 @opindex mno-altivec
13093 Generate code that uses (does not use) AltiVec instructions, and also
13094 enable the use of built-in functions that allow more direct access to
13095 the AltiVec instruction set. You may also need to set
13096 @option{-mabi=altivec} to adjust the current ABI with AltiVec ABI
13102 @opindex mno-vrsave
13103 Generate VRSAVE instructions when generating AltiVec code.
13106 @opindex msecure-plt
13107 Generate code that allows ld and ld.so to build executables and shared
13108 libraries with non-exec .plt and .got sections. This is a PowerPC
13109 32-bit SYSV ABI option.
13113 Generate code that uses a BSS .plt section that ld.so fills in, and
13114 requires .plt and .got sections that are both writable and executable.
13115 This is a PowerPC 32-bit SYSV ABI option.
13121 This switch enables or disables the generation of ISEL instructions.
13123 @item -misel=@var{yes/no}
13124 This switch has been deprecated. Use @option{-misel} and
13125 @option{-mno-isel} instead.
13131 This switch enables or disables the generation of SPE simd
13137 @opindex mno-paired
13138 This switch enables or disables the generation of PAIRED simd
13141 @item -mspe=@var{yes/no}
13142 This option has been deprecated. Use @option{-mspe} and
13143 @option{-mno-spe} instead.
13145 @item -mfloat-gprs=@var{yes/single/double/no}
13146 @itemx -mfloat-gprs
13147 @opindex mfloat-gprs
13148 This switch enables or disables the generation of floating point
13149 operations on the general purpose registers for architectures that
13152 The argument @var{yes} or @var{single} enables the use of
13153 single-precision floating point operations.
13155 The argument @var{double} enables the use of single and
13156 double-precision floating point operations.
13158 The argument @var{no} disables floating point operations on the
13159 general purpose registers.
13161 This option is currently only available on the MPC854x.
13167 Generate code for 32-bit or 64-bit environments of Darwin and SVR4
13168 targets (including GNU/Linux). The 32-bit environment sets int, long
13169 and pointer to 32 bits and generates code that runs on any PowerPC
13170 variant. The 64-bit environment sets int to 32 bits and long and
13171 pointer to 64 bits, and generates code for PowerPC64, as for
13172 @option{-mpowerpc64}.
13175 @itemx -mno-fp-in-toc
13176 @itemx -mno-sum-in-toc
13177 @itemx -mminimal-toc
13179 @opindex mno-fp-in-toc
13180 @opindex mno-sum-in-toc
13181 @opindex mminimal-toc
13182 Modify generation of the TOC (Table Of Contents), which is created for
13183 every executable file. The @option{-mfull-toc} option is selected by
13184 default. In that case, GCC will allocate at least one TOC entry for
13185 each unique non-automatic variable reference in your program. GCC
13186 will also place floating-point constants in the TOC@. However, only
13187 16,384 entries are available in the TOC@.
13189 If you receive a linker error message that saying you have overflowed
13190 the available TOC space, you can reduce the amount of TOC space used
13191 with the @option{-mno-fp-in-toc} and @option{-mno-sum-in-toc} options.
13192 @option{-mno-fp-in-toc} prevents GCC from putting floating-point
13193 constants in the TOC and @option{-mno-sum-in-toc} forces GCC to
13194 generate code to calculate the sum of an address and a constant at
13195 run-time instead of putting that sum into the TOC@. You may specify one
13196 or both of these options. Each causes GCC to produce very slightly
13197 slower and larger code at the expense of conserving TOC space.
13199 If you still run out of space in the TOC even when you specify both of
13200 these options, specify @option{-mminimal-toc} instead. This option causes
13201 GCC to make only one TOC entry for every file. When you specify this
13202 option, GCC will produce code that is slower and larger but which
13203 uses extremely little TOC space. You may wish to use this option
13204 only on files that contain less frequently executed code.
13210 Enable 64-bit AIX ABI and calling convention: 64-bit pointers, 64-bit
13211 @code{long} type, and the infrastructure needed to support them.
13212 Specifying @option{-maix64} implies @option{-mpowerpc64} and
13213 @option{-mpowerpc}, while @option{-maix32} disables the 64-bit ABI and
13214 implies @option{-mno-powerpc64}. GCC defaults to @option{-maix32}.
13217 @itemx -mno-xl-compat
13218 @opindex mxl-compat
13219 @opindex mno-xl-compat
13220 Produce code that conforms more closely to IBM XL compiler semantics
13221 when using AIX-compatible ABI@. Pass floating-point arguments to
13222 prototyped functions beyond the register save area (RSA) on the stack
13223 in addition to argument FPRs. Do not assume that most significant
13224 double in 128-bit long double value is properly rounded when comparing
13225 values and converting to double. Use XL symbol names for long double
13228 The AIX calling convention was extended but not initially documented to
13229 handle an obscure K&R C case of calling a function that takes the
13230 address of its arguments with fewer arguments than declared. IBM XL
13231 compilers access floating point arguments which do not fit in the
13232 RSA from the stack when a subroutine is compiled without
13233 optimization. Because always storing floating-point arguments on the
13234 stack is inefficient and rarely needed, this option is not enabled by
13235 default and only is necessary when calling subroutines compiled by IBM
13236 XL compilers without optimization.
13240 Support @dfn{IBM RS/6000 SP} @dfn{Parallel Environment} (PE)@. Link an
13241 application written to use message passing with special startup code to
13242 enable the application to run. The system must have PE installed in the
13243 standard location (@file{/usr/lpp/ppe.poe/}), or the @file{specs} file
13244 must be overridden with the @option{-specs=} option to specify the
13245 appropriate directory location. The Parallel Environment does not
13246 support threads, so the @option{-mpe} option and the @option{-pthread}
13247 option are incompatible.
13249 @item -malign-natural
13250 @itemx -malign-power
13251 @opindex malign-natural
13252 @opindex malign-power
13253 On AIX, 32-bit Darwin, and 64-bit PowerPC GNU/Linux, the option
13254 @option{-malign-natural} overrides the ABI-defined alignment of larger
13255 types, such as floating-point doubles, on their natural size-based boundary.
13256 The option @option{-malign-power} instructs GCC to follow the ABI-specified
13257 alignment rules. GCC defaults to the standard alignment defined in the ABI@.
13259 On 64-bit Darwin, natural alignment is the default, and @option{-malign-power}
13263 @itemx -mhard-float
13264 @opindex msoft-float
13265 @opindex mhard-float
13266 Generate code that does not use (uses) the floating-point register set.
13267 Software floating point emulation is provided if you use the
13268 @option{-msoft-float} option, and pass the option to GCC when linking.
13271 @itemx -mno-multiple
13273 @opindex mno-multiple
13274 Generate code that uses (does not use) the load multiple word
13275 instructions and the store multiple word instructions. These
13276 instructions are generated by default on POWER systems, and not
13277 generated on PowerPC systems. Do not use @option{-mmultiple} on little
13278 endian PowerPC systems, since those instructions do not work when the
13279 processor is in little endian mode. The exceptions are PPC740 and
13280 PPC750 which permit the instructions usage in little endian mode.
13285 @opindex mno-string
13286 Generate code that uses (does not use) the load string instructions
13287 and the store string word instructions to save multiple registers and
13288 do small block moves. These instructions are generated by default on
13289 POWER systems, and not generated on PowerPC systems. Do not use
13290 @option{-mstring} on little endian PowerPC systems, since those
13291 instructions do not work when the processor is in little endian mode.
13292 The exceptions are PPC740 and PPC750 which permit the instructions
13293 usage in little endian mode.
13298 @opindex mno-update
13299 Generate code that uses (does not use) the load or store instructions
13300 that update the base register to the address of the calculated memory
13301 location. These instructions are generated by default. If you use
13302 @option{-mno-update}, there is a small window between the time that the
13303 stack pointer is updated and the address of the previous frame is
13304 stored, which means code that walks the stack frame across interrupts or
13305 signals may get corrupted data.
13308 @itemx -mno-fused-madd
13309 @opindex mfused-madd
13310 @opindex mno-fused-madd
13311 Generate code that uses (does not use) the floating point multiply and
13312 accumulate instructions. These instructions are generated by default if
13313 hardware floating is used.
13319 Generate code that uses (does not use) the half-word multiply and
13320 multiply-accumulate instructions on the IBM 405, 440 and 464 processors.
13321 These instructions are generated by default when targetting those
13328 Generate code that uses (does not use) the string-search @samp{dlmzb}
13329 instruction on the IBM 405, 440 and 464 processors. This instruction is
13330 generated by default when targetting those processors.
13332 @item -mno-bit-align
13334 @opindex mno-bit-align
13335 @opindex mbit-align
13336 On System V.4 and embedded PowerPC systems do not (do) force structures
13337 and unions that contain bit-fields to be aligned to the base type of the
13340 For example, by default a structure containing nothing but 8
13341 @code{unsigned} bit-fields of length 1 would be aligned to a 4 byte
13342 boundary and have a size of 4 bytes. By using @option{-mno-bit-align},
13343 the structure would be aligned to a 1 byte boundary and be one byte in
13346 @item -mno-strict-align
13347 @itemx -mstrict-align
13348 @opindex mno-strict-align
13349 @opindex mstrict-align
13350 On System V.4 and embedded PowerPC systems do not (do) assume that
13351 unaligned memory references will be handled by the system.
13353 @item -mrelocatable
13354 @itemx -mno-relocatable
13355 @opindex mrelocatable
13356 @opindex mno-relocatable
13357 On embedded PowerPC systems generate code that allows (does not allow)
13358 the program to be relocated to a different address at runtime. If you
13359 use @option{-mrelocatable} on any module, all objects linked together must
13360 be compiled with @option{-mrelocatable} or @option{-mrelocatable-lib}.
13362 @item -mrelocatable-lib
13363 @itemx -mno-relocatable-lib
13364 @opindex mrelocatable-lib
13365 @opindex mno-relocatable-lib
13366 On embedded PowerPC systems generate code that allows (does not allow)
13367 the program to be relocated to a different address at runtime. Modules
13368 compiled with @option{-mrelocatable-lib} can be linked with either modules
13369 compiled without @option{-mrelocatable} and @option{-mrelocatable-lib} or
13370 with modules compiled with the @option{-mrelocatable} options.
13376 On System V.4 and embedded PowerPC systems do not (do) assume that
13377 register 2 contains a pointer to a global area pointing to the addresses
13378 used in the program.
13381 @itemx -mlittle-endian
13383 @opindex mlittle-endian
13384 On System V.4 and embedded PowerPC systems compile code for the
13385 processor in little endian mode. The @option{-mlittle-endian} option is
13386 the same as @option{-mlittle}.
13389 @itemx -mbig-endian
13391 @opindex mbig-endian
13392 On System V.4 and embedded PowerPC systems compile code for the
13393 processor in big endian mode. The @option{-mbig-endian} option is
13394 the same as @option{-mbig}.
13396 @item -mdynamic-no-pic
13397 @opindex mdynamic-no-pic
13398 On Darwin and Mac OS X systems, compile code so that it is not
13399 relocatable, but that its external references are relocatable. The
13400 resulting code is suitable for applications, but not shared
13403 @item -mprioritize-restricted-insns=@var{priority}
13404 @opindex mprioritize-restricted-insns
13405 This option controls the priority that is assigned to
13406 dispatch-slot restricted instructions during the second scheduling
13407 pass. The argument @var{priority} takes the value @var{0/1/2} to assign
13408 @var{no/highest/second-highest} priority to dispatch slot restricted
13411 @item -msched-costly-dep=@var{dependence_type}
13412 @opindex msched-costly-dep
13413 This option controls which dependences are considered costly
13414 by the target during instruction scheduling. The argument
13415 @var{dependence_type} takes one of the following values:
13416 @var{no}: no dependence is costly,
13417 @var{all}: all dependences are costly,
13418 @var{true_store_to_load}: a true dependence from store to load is costly,
13419 @var{store_to_load}: any dependence from store to load is costly,
13420 @var{number}: any dependence which latency >= @var{number} is costly.
13422 @item -minsert-sched-nops=@var{scheme}
13423 @opindex minsert-sched-nops
13424 This option controls which nop insertion scheme will be used during
13425 the second scheduling pass. The argument @var{scheme} takes one of the
13427 @var{no}: Don't insert nops.
13428 @var{pad}: Pad with nops any dispatch group which has vacant issue slots,
13429 according to the scheduler's grouping.
13430 @var{regroup_exact}: Insert nops to force costly dependent insns into
13431 separate groups. Insert exactly as many nops as needed to force an insn
13432 to a new group, according to the estimated processor grouping.
13433 @var{number}: Insert nops to force costly dependent insns into
13434 separate groups. Insert @var{number} nops to force an insn to a new group.
13437 @opindex mcall-sysv
13438 On System V.4 and embedded PowerPC systems compile code using calling
13439 conventions that adheres to the March 1995 draft of the System V
13440 Application Binary Interface, PowerPC processor supplement. This is the
13441 default unless you configured GCC using @samp{powerpc-*-eabiaix}.
13443 @item -mcall-sysv-eabi
13444 @opindex mcall-sysv-eabi
13445 Specify both @option{-mcall-sysv} and @option{-meabi} options.
13447 @item -mcall-sysv-noeabi
13448 @opindex mcall-sysv-noeabi
13449 Specify both @option{-mcall-sysv} and @option{-mno-eabi} options.
13451 @item -mcall-solaris
13452 @opindex mcall-solaris
13453 On System V.4 and embedded PowerPC systems compile code for the Solaris
13457 @opindex mcall-linux
13458 On System V.4 and embedded PowerPC systems compile code for the
13459 Linux-based GNU system.
13463 On System V.4 and embedded PowerPC systems compile code for the
13464 Hurd-based GNU system.
13466 @item -mcall-netbsd
13467 @opindex mcall-netbsd
13468 On System V.4 and embedded PowerPC systems compile code for the
13469 NetBSD operating system.
13471 @item -maix-struct-return
13472 @opindex maix-struct-return
13473 Return all structures in memory (as specified by the AIX ABI)@.
13475 @item -msvr4-struct-return
13476 @opindex msvr4-struct-return
13477 Return structures smaller than 8 bytes in registers (as specified by the
13480 @item -mabi=@var{abi-type}
13482 Extend the current ABI with a particular extension, or remove such extension.
13483 Valid values are @var{altivec}, @var{no-altivec}, @var{spe},
13484 @var{no-spe}, @var{ibmlongdouble}, @var{ieeelongdouble}@.
13488 Extend the current ABI with SPE ABI extensions. This does not change
13489 the default ABI, instead it adds the SPE ABI extensions to the current
13493 @opindex mabi=no-spe
13494 Disable Booke SPE ABI extensions for the current ABI@.
13496 @item -mabi=ibmlongdouble
13497 @opindex mabi=ibmlongdouble
13498 Change the current ABI to use IBM extended precision long double.
13499 This is a PowerPC 32-bit SYSV ABI option.
13501 @item -mabi=ieeelongdouble
13502 @opindex mabi=ieeelongdouble
13503 Change the current ABI to use IEEE extended precision long double.
13504 This is a PowerPC 32-bit Linux ABI option.
13507 @itemx -mno-prototype
13508 @opindex mprototype
13509 @opindex mno-prototype
13510 On System V.4 and embedded PowerPC systems assume that all calls to
13511 variable argument functions are properly prototyped. Otherwise, the
13512 compiler must insert an instruction before every non prototyped call to
13513 set or clear bit 6 of the condition code register (@var{CR}) to
13514 indicate whether floating point values were passed in the floating point
13515 registers in case the function takes a variable arguments. With
13516 @option{-mprototype}, only calls to prototyped variable argument functions
13517 will set or clear the bit.
13521 On embedded PowerPC systems, assume that the startup module is called
13522 @file{sim-crt0.o} and that the standard C libraries are @file{libsim.a} and
13523 @file{libc.a}. This is the default for @samp{powerpc-*-eabisim}
13528 On embedded PowerPC systems, assume that the startup module is called
13529 @file{crt0.o} and the standard C libraries are @file{libmvme.a} and
13534 On embedded PowerPC systems, assume that the startup module is called
13535 @file{crt0.o} and the standard C libraries are @file{libads.a} and
13538 @item -myellowknife
13539 @opindex myellowknife
13540 On embedded PowerPC systems, assume that the startup module is called
13541 @file{crt0.o} and the standard C libraries are @file{libyk.a} and
13546 On System V.4 and embedded PowerPC systems, specify that you are
13547 compiling for a VxWorks system.
13551 On embedded PowerPC systems, set the @var{PPC_EMB} bit in the ELF flags
13552 header to indicate that @samp{eabi} extended relocations are used.
13558 On System V.4 and embedded PowerPC systems do (do not) adhere to the
13559 Embedded Applications Binary Interface (eabi) which is a set of
13560 modifications to the System V.4 specifications. Selecting @option{-meabi}
13561 means that the stack is aligned to an 8 byte boundary, a function
13562 @code{__eabi} is called to from @code{main} to set up the eabi
13563 environment, and the @option{-msdata} option can use both @code{r2} and
13564 @code{r13} to point to two separate small data areas. Selecting
13565 @option{-mno-eabi} means that the stack is aligned to a 16 byte boundary,
13566 do not call an initialization function from @code{main}, and the
13567 @option{-msdata} option will only use @code{r13} to point to a single
13568 small data area. The @option{-meabi} option is on by default if you
13569 configured GCC using one of the @samp{powerpc*-*-eabi*} options.
13572 @opindex msdata=eabi
13573 On System V.4 and embedded PowerPC systems, put small initialized
13574 @code{const} global and static data in the @samp{.sdata2} section, which
13575 is pointed to by register @code{r2}. Put small initialized
13576 non-@code{const} global and static data in the @samp{.sdata} section,
13577 which is pointed to by register @code{r13}. Put small uninitialized
13578 global and static data in the @samp{.sbss} section, which is adjacent to
13579 the @samp{.sdata} section. The @option{-msdata=eabi} option is
13580 incompatible with the @option{-mrelocatable} option. The
13581 @option{-msdata=eabi} option also sets the @option{-memb} option.
13584 @opindex msdata=sysv
13585 On System V.4 and embedded PowerPC systems, put small global and static
13586 data in the @samp{.sdata} section, which is pointed to by register
13587 @code{r13}. Put small uninitialized global and static data in the
13588 @samp{.sbss} section, which is adjacent to the @samp{.sdata} section.
13589 The @option{-msdata=sysv} option is incompatible with the
13590 @option{-mrelocatable} option.
13592 @item -msdata=default
13594 @opindex msdata=default
13596 On System V.4 and embedded PowerPC systems, if @option{-meabi} is used,
13597 compile code the same as @option{-msdata=eabi}, otherwise compile code the
13598 same as @option{-msdata=sysv}.
13601 @opindex msdata-data
13602 On System V.4 and embedded PowerPC systems, put small global
13603 data in the @samp{.sdata} section. Put small uninitialized global
13604 data in the @samp{.sbss} section. Do not use register @code{r13}
13605 to address small data however. This is the default behavior unless
13606 other @option{-msdata} options are used.
13610 @opindex msdata=none
13612 On embedded PowerPC systems, put all initialized global and static data
13613 in the @samp{.data} section, and all uninitialized data in the
13614 @samp{.bss} section.
13618 @cindex smaller data references (PowerPC)
13619 @cindex .sdata/.sdata2 references (PowerPC)
13620 On embedded PowerPC systems, put global and static items less than or
13621 equal to @var{num} bytes into the small data or bss sections instead of
13622 the normal data or bss section. By default, @var{num} is 8. The
13623 @option{-G @var{num}} switch is also passed to the linker.
13624 All modules should be compiled with the same @option{-G @var{num}} value.
13627 @itemx -mno-regnames
13629 @opindex mno-regnames
13630 On System V.4 and embedded PowerPC systems do (do not) emit register
13631 names in the assembly language output using symbolic forms.
13634 @itemx -mno-longcall
13636 @opindex mno-longcall
13637 By default assume that all calls are far away so that a longer more
13638 expensive calling sequence is required. This is required for calls
13639 further than 32 megabytes (33,554,432 bytes) from the current location.
13640 A short call will be generated if the compiler knows
13641 the call cannot be that far away. This setting can be overridden by
13642 the @code{shortcall} function attribute, or by @code{#pragma
13645 Some linkers are capable of detecting out-of-range calls and generating
13646 glue code on the fly. On these systems, long calls are unnecessary and
13647 generate slower code. As of this writing, the AIX linker can do this,
13648 as can the GNU linker for PowerPC/64. It is planned to add this feature
13649 to the GNU linker for 32-bit PowerPC systems as well.
13651 On Darwin/PPC systems, @code{#pragma longcall} will generate ``jbsr
13652 callee, L42'', plus a ``branch island'' (glue code). The two target
13653 addresses represent the callee and the ``branch island''. The
13654 Darwin/PPC linker will prefer the first address and generate a ``bl
13655 callee'' if the PPC ``bl'' instruction will reach the callee directly;
13656 otherwise, the linker will generate ``bl L42'' to call the ``branch
13657 island''. The ``branch island'' is appended to the body of the
13658 calling function; it computes the full 32-bit address of the callee
13661 On Mach-O (Darwin) systems, this option directs the compiler emit to
13662 the glue for every direct call, and the Darwin linker decides whether
13663 to use or discard it.
13665 In the future, we may cause GCC to ignore all longcall specifications
13666 when the linker is known to generate glue.
13670 Adds support for multithreading with the @dfn{pthreads} library.
13671 This option sets flags for both the preprocessor and linker.
13675 @node S/390 and zSeries Options
13676 @subsection S/390 and zSeries Options
13677 @cindex S/390 and zSeries Options
13679 These are the @samp{-m} options defined for the S/390 and zSeries architecture.
13683 @itemx -msoft-float
13684 @opindex mhard-float
13685 @opindex msoft-float
13686 Use (do not use) the hardware floating-point instructions and registers
13687 for floating-point operations. When @option{-msoft-float} is specified,
13688 functions in @file{libgcc.a} will be used to perform floating-point
13689 operations. When @option{-mhard-float} is specified, the compiler
13690 generates IEEE floating-point instructions. This is the default.
13692 @item -mlong-double-64
13693 @itemx -mlong-double-128
13694 @opindex mlong-double-64
13695 @opindex mlong-double-128
13696 These switches control the size of @code{long double} type. A size
13697 of 64bit makes the @code{long double} type equivalent to the @code{double}
13698 type. This is the default.
13701 @itemx -mno-backchain
13702 @opindex mbackchain
13703 @opindex mno-backchain
13704 Store (do not store) the address of the caller's frame as backchain pointer
13705 into the callee's stack frame.
13706 A backchain may be needed to allow debugging using tools that do not understand
13707 DWARF-2 call frame information.
13708 When @option{-mno-packed-stack} is in effect, the backchain pointer is stored
13709 at the bottom of the stack frame; when @option{-mpacked-stack} is in effect,
13710 the backchain is placed into the topmost word of the 96/160 byte register
13713 In general, code compiled with @option{-mbackchain} is call-compatible with
13714 code compiled with @option{-mmo-backchain}; however, use of the backchain
13715 for debugging purposes usually requires that the whole binary is built with
13716 @option{-mbackchain}. Note that the combination of @option{-mbackchain},
13717 @option{-mpacked-stack} and @option{-mhard-float} is not supported. In order
13718 to build a linux kernel use @option{-msoft-float}.
13720 The default is to not maintain the backchain.
13722 @item -mpacked-stack
13723 @itemx -mno-packed-stack
13724 @opindex mpacked-stack
13725 @opindex mno-packed-stack
13726 Use (do not use) the packed stack layout. When @option{-mno-packed-stack} is
13727 specified, the compiler uses the all fields of the 96/160 byte register save
13728 area only for their default purpose; unused fields still take up stack space.
13729 When @option{-mpacked-stack} is specified, register save slots are densely
13730 packed at the top of the register save area; unused space is reused for other
13731 purposes, allowing for more efficient use of the available stack space.
13732 However, when @option{-mbackchain} is also in effect, the topmost word of
13733 the save area is always used to store the backchain, and the return address
13734 register is always saved two words below the backchain.
13736 As long as the stack frame backchain is not used, code generated with
13737 @option{-mpacked-stack} is call-compatible with code generated with
13738 @option{-mno-packed-stack}. Note that some non-FSF releases of GCC 2.95 for
13739 S/390 or zSeries generated code that uses the stack frame backchain at run
13740 time, not just for debugging purposes. Such code is not call-compatible
13741 with code compiled with @option{-mpacked-stack}. Also, note that the
13742 combination of @option{-mbackchain},
13743 @option{-mpacked-stack} and @option{-mhard-float} is not supported. In order
13744 to build a linux kernel use @option{-msoft-float}.
13746 The default is to not use the packed stack layout.
13749 @itemx -mno-small-exec
13750 @opindex msmall-exec
13751 @opindex mno-small-exec
13752 Generate (or do not generate) code using the @code{bras} instruction
13753 to do subroutine calls.
13754 This only works reliably if the total executable size does not
13755 exceed 64k. The default is to use the @code{basr} instruction instead,
13756 which does not have this limitation.
13762 When @option{-m31} is specified, generate code compliant to the
13763 GNU/Linux for S/390 ABI@. When @option{-m64} is specified, generate
13764 code compliant to the GNU/Linux for zSeries ABI@. This allows GCC in
13765 particular to generate 64-bit instructions. For the @samp{s390}
13766 targets, the default is @option{-m31}, while the @samp{s390x}
13767 targets default to @option{-m64}.
13773 When @option{-mzarch} is specified, generate code using the
13774 instructions available on z/Architecture.
13775 When @option{-mesa} is specified, generate code using the
13776 instructions available on ESA/390. Note that @option{-mesa} is
13777 not possible with @option{-m64}.
13778 When generating code compliant to the GNU/Linux for S/390 ABI,
13779 the default is @option{-mesa}. When generating code compliant
13780 to the GNU/Linux for zSeries ABI, the default is @option{-mzarch}.
13786 Generate (or do not generate) code using the @code{mvcle} instruction
13787 to perform block moves. When @option{-mno-mvcle} is specified,
13788 use a @code{mvc} loop instead. This is the default unless optimizing for
13795 Print (or do not print) additional debug information when compiling.
13796 The default is to not print debug information.
13798 @item -march=@var{cpu-type}
13800 Generate code that will run on @var{cpu-type}, which is the name of a system
13801 representing a certain processor type. Possible values for
13802 @var{cpu-type} are @samp{g5}, @samp{g6}, @samp{z900}, and @samp{z990}.
13803 When generating code using the instructions available on z/Architecture,
13804 the default is @option{-march=z900}. Otherwise, the default is
13805 @option{-march=g5}.
13807 @item -mtune=@var{cpu-type}
13809 Tune to @var{cpu-type} everything applicable about the generated code,
13810 except for the ABI and the set of available instructions.
13811 The list of @var{cpu-type} values is the same as for @option{-march}.
13812 The default is the value used for @option{-march}.
13815 @itemx -mno-tpf-trace
13816 @opindex mtpf-trace
13817 @opindex mno-tpf-trace
13818 Generate code that adds (does not add) in TPF OS specific branches to trace
13819 routines in the operating system. This option is off by default, even
13820 when compiling for the TPF OS@.
13823 @itemx -mno-fused-madd
13824 @opindex mfused-madd
13825 @opindex mno-fused-madd
13826 Generate code that uses (does not use) the floating point multiply and
13827 accumulate instructions. These instructions are generated by default if
13828 hardware floating point is used.
13830 @item -mwarn-framesize=@var{framesize}
13831 @opindex mwarn-framesize
13832 Emit a warning if the current function exceeds the given frame size. Because
13833 this is a compile time check it doesn't need to be a real problem when the program
13834 runs. It is intended to identify functions which most probably cause
13835 a stack overflow. It is useful to be used in an environment with limited stack
13836 size e.g.@: the linux kernel.
13838 @item -mwarn-dynamicstack
13839 @opindex mwarn-dynamicstack
13840 Emit a warning if the function calls alloca or uses dynamically
13841 sized arrays. This is generally a bad idea with a limited stack size.
13843 @item -mstack-guard=@var{stack-guard}
13844 @itemx -mstack-size=@var{stack-size}
13845 @opindex mstack-guard
13846 @opindex mstack-size
13847 If these options are provided the s390 back end emits additional instructions in
13848 the function prologue which trigger a trap if the stack size is @var{stack-guard}
13849 bytes above the @var{stack-size} (remember that the stack on s390 grows downward).
13850 If the @var{stack-guard} option is omitted the smallest power of 2 larger than
13851 the frame size of the compiled function is chosen.
13852 These options are intended to be used to help debugging stack overflow problems.
13853 The additionally emitted code causes only little overhead and hence can also be
13854 used in production like systems without greater performance degradation. The given
13855 values have to be exact powers of 2 and @var{stack-size} has to be greater than
13856 @var{stack-guard} without exceeding 64k.
13857 In order to be efficient the extra code makes the assumption that the stack starts
13858 at an address aligned to the value given by @var{stack-size}.
13859 The @var{stack-guard} option can only be used in conjunction with @var{stack-size}.
13862 @node Score Options
13863 @subsection Score Options
13864 @cindex Score Options
13866 These options are defined for Score implementations:
13871 Compile code for big endian mode. This is the default.
13875 Compile code for little endian mode.
13879 Disable generate bcnz instruction.
13883 Enable generate unaligned load and store instruction.
13887 Enable the use of multiply-accumulate instructions. Disabled by default.
13891 Specify the SCORE5 as the target architecture.
13895 Specify the SCORE5U of the target architecture.
13899 Specify the SCORE7 as the target architecture. This is the default.
13903 Specify the SCORE7D as the target architecture.
13907 @subsection SH Options
13909 These @samp{-m} options are defined for the SH implementations:
13914 Generate code for the SH1.
13918 Generate code for the SH2.
13921 Generate code for the SH2e.
13925 Generate code for the SH3.
13929 Generate code for the SH3e.
13933 Generate code for the SH4 without a floating-point unit.
13935 @item -m4-single-only
13936 @opindex m4-single-only
13937 Generate code for the SH4 with a floating-point unit that only
13938 supports single-precision arithmetic.
13942 Generate code for the SH4 assuming the floating-point unit is in
13943 single-precision mode by default.
13947 Generate code for the SH4.
13951 Generate code for the SH4al-dsp, or for a SH4a in such a way that the
13952 floating-point unit is not used.
13954 @item -m4a-single-only
13955 @opindex m4a-single-only
13956 Generate code for the SH4a, in such a way that no double-precision
13957 floating point operations are used.
13960 @opindex m4a-single
13961 Generate code for the SH4a assuming the floating-point unit is in
13962 single-precision mode by default.
13966 Generate code for the SH4a.
13970 Same as @option{-m4a-nofpu}, except that it implicitly passes
13971 @option{-dsp} to the assembler. GCC doesn't generate any DSP
13972 instructions at the moment.
13976 Compile code for the processor in big endian mode.
13980 Compile code for the processor in little endian mode.
13984 Align doubles at 64-bit boundaries. Note that this changes the calling
13985 conventions, and thus some functions from the standard C library will
13986 not work unless you recompile it first with @option{-mdalign}.
13990 Shorten some address references at link time, when possible; uses the
13991 linker option @option{-relax}.
13995 Use 32-bit offsets in @code{switch} tables. The default is to use
14000 Enable the use of bit manipulation instructions on SH2A.
14004 Enable the use of the instruction @code{fmovd}.
14008 Comply with the calling conventions defined by Renesas.
14012 Comply with the calling conventions defined by Renesas.
14016 Comply with the calling conventions defined for GCC before the Renesas
14017 conventions were available. This option is the default for all
14018 targets of the SH toolchain except for @samp{sh-symbianelf}.
14021 @opindex mnomacsave
14022 Mark the @code{MAC} register as call-clobbered, even if
14023 @option{-mhitachi} is given.
14027 Increase IEEE-compliance of floating-point code.
14028 At the moment, this is equivalent to @option{-fno-finite-math-only}.
14029 When generating 16 bit SH opcodes, getting IEEE-conforming results for
14030 comparisons of NANs / infinities incurs extra overhead in every
14031 floating point comparison, therefore the default is set to
14032 @option{-ffinite-math-only}.
14034 @item -minline-ic_invalidate
14035 @opindex minline-ic_invalidate
14036 Inline code to invalidate instruction cache entries after setting up
14037 nested function trampolines.
14038 This option has no effect if -musermode is in effect and the selected
14039 code generation option (e.g. -m4) does not allow the use of the icbi
14041 If the selected code generation option does not allow the use of the icbi
14042 instruction, and -musermode is not in effect, the inlined code will
14043 manipulate the instruction cache address array directly with an associative
14044 write. This not only requires privileged mode, but it will also
14045 fail if the cache line had been mapped via the TLB and has become unmapped.
14049 Dump instruction size and location in the assembly code.
14052 @opindex mpadstruct
14053 This option is deprecated. It pads structures to multiple of 4 bytes,
14054 which is incompatible with the SH ABI@.
14058 Optimize for space instead of speed. Implied by @option{-Os}.
14061 @opindex mprefergot
14062 When generating position-independent code, emit function calls using
14063 the Global Offset Table instead of the Procedure Linkage Table.
14067 Don't generate privileged mode only code; implies -mno-inline-ic_invalidate
14068 if the inlined code would not work in user mode.
14069 This is the default when the target is @code{sh-*-linux*}.
14071 @item -multcost=@var{number}
14072 @opindex multcost=@var{number}
14073 Set the cost to assume for a multiply insn.
14075 @item -mdiv=@var{strategy}
14076 @opindex mdiv=@var{strategy}
14077 Set the division strategy to use for SHmedia code. @var{strategy} must be
14078 one of: call, call2, fp, inv, inv:minlat, inv20u, inv20l, inv:call,
14079 inv:call2, inv:fp .
14080 "fp" performs the operation in floating point. This has a very high latency,
14081 but needs only a few instructions, so it might be a good choice if
14082 your code has enough easily exploitable ILP to allow the compiler to
14083 schedule the floating point instructions together with other instructions.
14084 Division by zero causes a floating point exception.
14085 "inv" uses integer operations to calculate the inverse of the divisor,
14086 and then multiplies the dividend with the inverse. This strategy allows
14087 cse and hoisting of the inverse calculation. Division by zero calculates
14088 an unspecified result, but does not trap.
14089 "inv:minlat" is a variant of "inv" where if no cse / hoisting opportunities
14090 have been found, or if the entire operation has been hoisted to the same
14091 place, the last stages of the inverse calculation are intertwined with the
14092 final multiply to reduce the overall latency, at the expense of using a few
14093 more instructions, and thus offering fewer scheduling opportunities with
14095 "call" calls a library function that usually implements the inv:minlat
14097 This gives high code density for m5-*media-nofpu compilations.
14098 "call2" uses a different entry point of the same library function, where it
14099 assumes that a pointer to a lookup table has already been set up, which
14100 exposes the pointer load to cse / code hoisting optimizations.
14101 "inv:call", "inv:call2" and "inv:fp" all use the "inv" algorithm for initial
14102 code generation, but if the code stays unoptimized, revert to the "call",
14103 "call2", or "fp" strategies, respectively. Note that the
14104 potentially-trapping side effect of division by zero is carried by a
14105 separate instruction, so it is possible that all the integer instructions
14106 are hoisted out, but the marker for the side effect stays where it is.
14107 A recombination to fp operations or a call is not possible in that case.
14108 "inv20u" and "inv20l" are variants of the "inv:minlat" strategy. In the case
14109 that the inverse calculation was nor separated from the multiply, they speed
14110 up division where the dividend fits into 20 bits (plus sign where applicable),
14111 by inserting a test to skip a number of operations in this case; this test
14112 slows down the case of larger dividends. inv20u assumes the case of a such
14113 a small dividend to be unlikely, and inv20l assumes it to be likely.
14115 @item -mdivsi3_libfunc=@var{name}
14116 @opindex mdivsi3_libfunc=@var{name}
14117 Set the name of the library function used for 32 bit signed division to
14118 @var{name}. This only affect the name used in the call and inv:call
14119 division strategies, and the compiler will still expect the same
14120 sets of input/output/clobbered registers as if this option was not present.
14122 @item -mfixed-range=@var{register-range}
14123 @opindex mfixed-range
14124 Generate code treating the given register range as fixed registers.
14125 A fixed register is one that the register allocator can not use. This is
14126 useful when compiling kernel code. A register range is specified as
14127 two registers separated by a dash. Multiple register ranges can be
14128 specified separated by a comma.
14130 @item -madjust-unroll
14131 @opindex madjust-unroll
14132 Throttle unrolling to avoid thrashing target registers.
14133 This option only has an effect if the gcc code base supports the
14134 TARGET_ADJUST_UNROLL_MAX target hook.
14136 @item -mindexed-addressing
14137 @opindex mindexed-addressing
14138 Enable the use of the indexed addressing mode for SHmedia32/SHcompact.
14139 This is only safe if the hardware and/or OS implement 32 bit wrap-around
14140 semantics for the indexed addressing mode. The architecture allows the
14141 implementation of processors with 64 bit MMU, which the OS could use to
14142 get 32 bit addressing, but since no current hardware implementation supports
14143 this or any other way to make the indexed addressing mode safe to use in
14144 the 32 bit ABI, the default is -mno-indexed-addressing.
14146 @item -mgettrcost=@var{number}
14147 @opindex mgettrcost=@var{number}
14148 Set the cost assumed for the gettr instruction to @var{number}.
14149 The default is 2 if @option{-mpt-fixed} is in effect, 100 otherwise.
14153 Assume pt* instructions won't trap. This will generally generate better
14154 scheduled code, but is unsafe on current hardware. The current architecture
14155 definition says that ptabs and ptrel trap when the target anded with 3 is 3.
14156 This has the unintentional effect of making it unsafe to schedule ptabs /
14157 ptrel before a branch, or hoist it out of a loop. For example,
14158 __do_global_ctors, a part of libgcc that runs constructors at program
14159 startup, calls functions in a list which is delimited by @minus{}1. With the
14160 -mpt-fixed option, the ptabs will be done before testing against @minus{}1.
14161 That means that all the constructors will be run a bit quicker, but when
14162 the loop comes to the end of the list, the program crashes because ptabs
14163 loads @minus{}1 into a target register. Since this option is unsafe for any
14164 hardware implementing the current architecture specification, the default
14165 is -mno-pt-fixed. Unless the user specifies a specific cost with
14166 @option{-mgettrcost}, -mno-pt-fixed also implies @option{-mgettrcost=100};
14167 this deters register allocation using target registers for storing
14170 @item -minvalid-symbols
14171 @opindex minvalid-symbols
14172 Assume symbols might be invalid. Ordinary function symbols generated by
14173 the compiler will always be valid to load with movi/shori/ptabs or
14174 movi/shori/ptrel, but with assembler and/or linker tricks it is possible
14175 to generate symbols that will cause ptabs / ptrel to trap.
14176 This option is only meaningful when @option{-mno-pt-fixed} is in effect.
14177 It will then prevent cross-basic-block cse, hoisting and most scheduling
14178 of symbol loads. The default is @option{-mno-invalid-symbols}.
14181 @node SPARC Options
14182 @subsection SPARC Options
14183 @cindex SPARC options
14185 These @samp{-m} options are supported on the SPARC:
14188 @item -mno-app-regs
14190 @opindex mno-app-regs
14192 Specify @option{-mapp-regs} to generate output using the global registers
14193 2 through 4, which the SPARC SVR4 ABI reserves for applications. This
14196 To be fully SVR4 ABI compliant at the cost of some performance loss,
14197 specify @option{-mno-app-regs}. You should compile libraries and system
14198 software with this option.
14201 @itemx -mhard-float
14203 @opindex mhard-float
14204 Generate output containing floating point instructions. This is the
14208 @itemx -msoft-float
14210 @opindex msoft-float
14211 Generate output containing library calls for floating point.
14212 @strong{Warning:} the requisite libraries are not available for all SPARC
14213 targets. Normally the facilities of the machine's usual C compiler are
14214 used, but this cannot be done directly in cross-compilation. You must make
14215 your own arrangements to provide suitable library functions for
14216 cross-compilation. The embedded targets @samp{sparc-*-aout} and
14217 @samp{sparclite-*-*} do provide software floating point support.
14219 @option{-msoft-float} changes the calling convention in the output file;
14220 therefore, it is only useful if you compile @emph{all} of a program with
14221 this option. In particular, you need to compile @file{libgcc.a}, the
14222 library that comes with GCC, with @option{-msoft-float} in order for
14225 @item -mhard-quad-float
14226 @opindex mhard-quad-float
14227 Generate output containing quad-word (long double) floating point
14230 @item -msoft-quad-float
14231 @opindex msoft-quad-float
14232 Generate output containing library calls for quad-word (long double)
14233 floating point instructions. The functions called are those specified
14234 in the SPARC ABI@. This is the default.
14236 As of this writing, there are no SPARC implementations that have hardware
14237 support for the quad-word floating point instructions. They all invoke
14238 a trap handler for one of these instructions, and then the trap handler
14239 emulates the effect of the instruction. Because of the trap handler overhead,
14240 this is much slower than calling the ABI library routines. Thus the
14241 @option{-msoft-quad-float} option is the default.
14243 @item -mno-unaligned-doubles
14244 @itemx -munaligned-doubles
14245 @opindex mno-unaligned-doubles
14246 @opindex munaligned-doubles
14247 Assume that doubles have 8 byte alignment. This is the default.
14249 With @option{-munaligned-doubles}, GCC assumes that doubles have 8 byte
14250 alignment only if they are contained in another type, or if they have an
14251 absolute address. Otherwise, it assumes they have 4 byte alignment.
14252 Specifying this option avoids some rare compatibility problems with code
14253 generated by other compilers. It is not the default because it results
14254 in a performance loss, especially for floating point code.
14256 @item -mno-faster-structs
14257 @itemx -mfaster-structs
14258 @opindex mno-faster-structs
14259 @opindex mfaster-structs
14260 With @option{-mfaster-structs}, the compiler assumes that structures
14261 should have 8 byte alignment. This enables the use of pairs of
14262 @code{ldd} and @code{std} instructions for copies in structure
14263 assignment, in place of twice as many @code{ld} and @code{st} pairs.
14264 However, the use of this changed alignment directly violates the SPARC
14265 ABI@. Thus, it's intended only for use on targets where the developer
14266 acknowledges that their resulting code will not be directly in line with
14267 the rules of the ABI@.
14269 @item -mimpure-text
14270 @opindex mimpure-text
14271 @option{-mimpure-text}, used in addition to @option{-shared}, tells
14272 the compiler to not pass @option{-z text} to the linker when linking a
14273 shared object. Using this option, you can link position-dependent
14274 code into a shared object.
14276 @option{-mimpure-text} suppresses the ``relocations remain against
14277 allocatable but non-writable sections'' linker error message.
14278 However, the necessary relocations will trigger copy-on-write, and the
14279 shared object is not actually shared across processes. Instead of
14280 using @option{-mimpure-text}, you should compile all source code with
14281 @option{-fpic} or @option{-fPIC}.
14283 This option is only available on SunOS and Solaris.
14285 @item -mcpu=@var{cpu_type}
14287 Set the instruction set, register set, and instruction scheduling parameters
14288 for machine type @var{cpu_type}. Supported values for @var{cpu_type} are
14289 @samp{v7}, @samp{cypress}, @samp{v8}, @samp{supersparc}, @samp{sparclite},
14290 @samp{f930}, @samp{f934}, @samp{hypersparc}, @samp{sparclite86x},
14291 @samp{sparclet}, @samp{tsc701}, @samp{v9}, @samp{ultrasparc},
14292 @samp{ultrasparc3}, @samp{niagara} and @samp{niagara2}.
14294 Default instruction scheduling parameters are used for values that select
14295 an architecture and not an implementation. These are @samp{v7}, @samp{v8},
14296 @samp{sparclite}, @samp{sparclet}, @samp{v9}.
14298 Here is a list of each supported architecture and their supported
14303 v8: supersparc, hypersparc
14304 sparclite: f930, f934, sparclite86x
14306 v9: ultrasparc, ultrasparc3, niagara, niagara2
14309 By default (unless configured otherwise), GCC generates code for the V7
14310 variant of the SPARC architecture. With @option{-mcpu=cypress}, the compiler
14311 additionally optimizes it for the Cypress CY7C602 chip, as used in the
14312 SPARCStation/SPARCServer 3xx series. This is also appropriate for the older
14313 SPARCStation 1, 2, IPX etc.
14315 With @option{-mcpu=v8}, GCC generates code for the V8 variant of the SPARC
14316 architecture. The only difference from V7 code is that the compiler emits
14317 the integer multiply and integer divide instructions which exist in SPARC-V8
14318 but not in SPARC-V7. With @option{-mcpu=supersparc}, the compiler additionally
14319 optimizes it for the SuperSPARC chip, as used in the SPARCStation 10, 1000 and
14322 With @option{-mcpu=sparclite}, GCC generates code for the SPARClite variant of
14323 the SPARC architecture. This adds the integer multiply, integer divide step
14324 and scan (@code{ffs}) instructions which exist in SPARClite but not in SPARC-V7.
14325 With @option{-mcpu=f930}, the compiler additionally optimizes it for the
14326 Fujitsu MB86930 chip, which is the original SPARClite, with no FPU@. With
14327 @option{-mcpu=f934}, the compiler additionally optimizes it for the Fujitsu
14328 MB86934 chip, which is the more recent SPARClite with FPU@.
14330 With @option{-mcpu=sparclet}, GCC generates code for the SPARClet variant of
14331 the SPARC architecture. This adds the integer multiply, multiply/accumulate,
14332 integer divide step and scan (@code{ffs}) instructions which exist in SPARClet
14333 but not in SPARC-V7. With @option{-mcpu=tsc701}, the compiler additionally
14334 optimizes it for the TEMIC SPARClet chip.
14336 With @option{-mcpu=v9}, GCC generates code for the V9 variant of the SPARC
14337 architecture. This adds 64-bit integer and floating-point move instructions,
14338 3 additional floating-point condition code registers and conditional move
14339 instructions. With @option{-mcpu=ultrasparc}, the compiler additionally
14340 optimizes it for the Sun UltraSPARC I/II/IIi chips. With
14341 @option{-mcpu=ultrasparc3}, the compiler additionally optimizes it for the
14342 Sun UltraSPARC III/III+/IIIi/IIIi+/IV/IV+ chips. With
14343 @option{-mcpu=niagara}, the compiler additionally optimizes it for
14344 Sun UltraSPARC T1 chips. With @option{-mcpu=niagara2}, the compiler
14345 additionally optimizes it for Sun UltraSPARC T2 chips.
14347 @item -mtune=@var{cpu_type}
14349 Set the instruction scheduling parameters for machine type
14350 @var{cpu_type}, but do not set the instruction set or register set that the
14351 option @option{-mcpu=@var{cpu_type}} would.
14353 The same values for @option{-mcpu=@var{cpu_type}} can be used for
14354 @option{-mtune=@var{cpu_type}}, but the only useful values are those
14355 that select a particular cpu implementation. Those are @samp{cypress},
14356 @samp{supersparc}, @samp{hypersparc}, @samp{f930}, @samp{f934},
14357 @samp{sparclite86x}, @samp{tsc701}, @samp{ultrasparc},
14358 @samp{ultrasparc3}, @samp{niagara}, and @samp{niagara2}.
14363 @opindex mno-v8plus
14364 With @option{-mv8plus}, GCC generates code for the SPARC-V8+ ABI@. The
14365 difference from the V8 ABI is that the global and out registers are
14366 considered 64-bit wide. This is enabled by default on Solaris in 32-bit
14367 mode for all SPARC-V9 processors.
14373 With @option{-mvis}, GCC generates code that takes advantage of the UltraSPARC
14374 Visual Instruction Set extensions. The default is @option{-mno-vis}.
14377 These @samp{-m} options are supported in addition to the above
14378 on SPARC-V9 processors in 64-bit environments:
14381 @item -mlittle-endian
14382 @opindex mlittle-endian
14383 Generate code for a processor running in little-endian mode. It is only
14384 available for a few configurations and most notably not on Solaris and Linux.
14390 Generate code for a 32-bit or 64-bit environment.
14391 The 32-bit environment sets int, long and pointer to 32 bits.
14392 The 64-bit environment sets int to 32 bits and long and pointer
14395 @item -mcmodel=medlow
14396 @opindex mcmodel=medlow
14397 Generate code for the Medium/Low code model: 64-bit addresses, programs
14398 must be linked in the low 32 bits of memory. Programs can be statically
14399 or dynamically linked.
14401 @item -mcmodel=medmid
14402 @opindex mcmodel=medmid
14403 Generate code for the Medium/Middle code model: 64-bit addresses, programs
14404 must be linked in the low 44 bits of memory, the text and data segments must
14405 be less than 2GB in size and the data segment must be located within 2GB of
14408 @item -mcmodel=medany
14409 @opindex mcmodel=medany
14410 Generate code for the Medium/Anywhere code model: 64-bit addresses, programs
14411 may be linked anywhere in memory, the text and data segments must be less
14412 than 2GB in size and the data segment must be located within 2GB of the
14415 @item -mcmodel=embmedany
14416 @opindex mcmodel=embmedany
14417 Generate code for the Medium/Anywhere code model for embedded systems:
14418 64-bit addresses, the text and data segments must be less than 2GB in
14419 size, both starting anywhere in memory (determined at link time). The
14420 global register %g4 points to the base of the data segment. Programs
14421 are statically linked and PIC is not supported.
14424 @itemx -mno-stack-bias
14425 @opindex mstack-bias
14426 @opindex mno-stack-bias
14427 With @option{-mstack-bias}, GCC assumes that the stack pointer, and
14428 frame pointer if present, are offset by @minus{}2047 which must be added back
14429 when making stack frame references. This is the default in 64-bit mode.
14430 Otherwise, assume no such offset is present.
14433 These switches are supported in addition to the above on Solaris:
14438 Add support for multithreading using the Solaris threads library. This
14439 option sets flags for both the preprocessor and linker. This option does
14440 not affect the thread safety of object code produced by the compiler or
14441 that of libraries supplied with it.
14445 Add support for multithreading using the POSIX threads library. This
14446 option sets flags for both the preprocessor and linker. This option does
14447 not affect the thread safety of object code produced by the compiler or
14448 that of libraries supplied with it.
14452 This is a synonym for @option{-pthreads}.
14456 @subsection SPU Options
14457 @cindex SPU options
14459 These @samp{-m} options are supported on the SPU:
14463 @itemx -merror-reloc
14464 @opindex mwarn-reloc
14465 @opindex merror-reloc
14467 The loader for SPU does not handle dynamic relocations. By default, GCC
14468 will give an error when it generates code that requires a dynamic
14469 relocation. @option{-mno-error-reloc} disables the error,
14470 @option{-mwarn-reloc} will generate a warning instead.
14473 @itemx -munsafe-dma
14475 @opindex munsafe-dma
14477 Instructions which initiate or test completion of DMA must not be
14478 reordered with respect to loads and stores of the memory which is being
14479 accessed. Users typically address this problem using the volatile
14480 keyword, but that can lead to inefficient code in places where the
14481 memory is known to not change. Rather than mark the memory as volatile
14482 we treat the DMA instructions as potentially effecting all memory. With
14483 @option{-munsafe-dma} users must use the volatile keyword to protect
14486 @item -mbranch-hints
14487 @opindex mbranch-hints
14489 By default, GCC will generate a branch hint instruction to avoid
14490 pipeline stalls for always taken or probably taken branches. A hint
14491 will not be generated closer than 8 instructions away from its branch.
14492 There is little reason to disable them, except for debugging purposes,
14493 or to make an object a little bit smaller.
14497 @opindex msmall-mem
14498 @opindex mlarge-mem
14500 By default, GCC generates code assuming that addresses are never larger
14501 than 18 bits. With @option{-mlarge-mem} code is generated that assumes
14502 a full 32 bit address.
14507 By default, GCC links against startup code that assumes the SPU-style
14508 main function interface (which has an unconventional parameter list).
14509 With @option{-mstdmain}, GCC will link your program against startup
14510 code that assumes a C99-style interface to @code{main}, including a
14511 local copy of @code{argv} strings.
14513 @item -mfixed-range=@var{register-range}
14514 @opindex mfixed-range
14515 Generate code treating the given register range as fixed registers.
14516 A fixed register is one that the register allocator can not use. This is
14517 useful when compiling kernel code. A register range is specified as
14518 two registers separated by a dash. Multiple register ranges can be
14519 specified separated by a comma.
14523 @node System V Options
14524 @subsection Options for System V
14526 These additional options are available on System V Release 4 for
14527 compatibility with other compilers on those systems:
14532 Create a shared object.
14533 It is recommended that @option{-symbolic} or @option{-shared} be used instead.
14537 Identify the versions of each tool used by the compiler, in a
14538 @code{.ident} assembler directive in the output.
14542 Refrain from adding @code{.ident} directives to the output file (this is
14545 @item -YP,@var{dirs}
14547 Search the directories @var{dirs}, and no others, for libraries
14548 specified with @option{-l}.
14550 @item -Ym,@var{dir}
14552 Look in the directory @var{dir} to find the M4 preprocessor.
14553 The assembler uses this option.
14554 @c This is supposed to go with a -Yd for predefined M4 macro files, but
14555 @c the generic assembler that comes with Solaris takes just -Ym.
14559 @subsection V850 Options
14560 @cindex V850 Options
14562 These @samp{-m} options are defined for V850 implementations:
14566 @itemx -mno-long-calls
14567 @opindex mlong-calls
14568 @opindex mno-long-calls
14569 Treat all calls as being far away (near). If calls are assumed to be
14570 far away, the compiler will always load the functions address up into a
14571 register, and call indirect through the pointer.
14577 Do not optimize (do optimize) basic blocks that use the same index
14578 pointer 4 or more times to copy pointer into the @code{ep} register, and
14579 use the shorter @code{sld} and @code{sst} instructions. The @option{-mep}
14580 option is on by default if you optimize.
14582 @item -mno-prolog-function
14583 @itemx -mprolog-function
14584 @opindex mno-prolog-function
14585 @opindex mprolog-function
14586 Do not use (do use) external functions to save and restore registers
14587 at the prologue and epilogue of a function. The external functions
14588 are slower, but use less code space if more than one function saves
14589 the same number of registers. The @option{-mprolog-function} option
14590 is on by default if you optimize.
14594 Try to make the code as small as possible. At present, this just turns
14595 on the @option{-mep} and @option{-mprolog-function} options.
14597 @item -mtda=@var{n}
14599 Put static or global variables whose size is @var{n} bytes or less into
14600 the tiny data area that register @code{ep} points to. The tiny data
14601 area can hold up to 256 bytes in total (128 bytes for byte references).
14603 @item -msda=@var{n}
14605 Put static or global variables whose size is @var{n} bytes or less into
14606 the small data area that register @code{gp} points to. The small data
14607 area can hold up to 64 kilobytes.
14609 @item -mzda=@var{n}
14611 Put static or global variables whose size is @var{n} bytes or less into
14612 the first 32 kilobytes of memory.
14616 Specify that the target processor is the V850.
14619 @opindex mbig-switch
14620 Generate code suitable for big switch tables. Use this option only if
14621 the assembler/linker complain about out of range branches within a switch
14626 This option will cause r2 and r5 to be used in the code generated by
14627 the compiler. This setting is the default.
14629 @item -mno-app-regs
14630 @opindex mno-app-regs
14631 This option will cause r2 and r5 to be treated as fixed registers.
14635 Specify that the target processor is the V850E1. The preprocessor
14636 constants @samp{__v850e1__} and @samp{__v850e__} will be defined if
14637 this option is used.
14641 Specify that the target processor is the V850E@. The preprocessor
14642 constant @samp{__v850e__} will be defined if this option is used.
14644 If neither @option{-mv850} nor @option{-mv850e} nor @option{-mv850e1}
14645 are defined then a default target processor will be chosen and the
14646 relevant @samp{__v850*__} preprocessor constant will be defined.
14648 The preprocessor constants @samp{__v850} and @samp{__v851__} are always
14649 defined, regardless of which processor variant is the target.
14651 @item -mdisable-callt
14652 @opindex mdisable-callt
14653 This option will suppress generation of the CALLT instruction for the
14654 v850e and v850e1 flavors of the v850 architecture. The default is
14655 @option{-mno-disable-callt} which allows the CALLT instruction to be used.
14660 @subsection VAX Options
14661 @cindex VAX options
14663 These @samp{-m} options are defined for the VAX:
14668 Do not output certain jump instructions (@code{aobleq} and so on)
14669 that the Unix assembler for the VAX cannot handle across long
14674 Do output those jump instructions, on the assumption that you
14675 will assemble with the GNU assembler.
14679 Output code for g-format floating point numbers instead of d-format.
14682 @node VxWorks Options
14683 @subsection VxWorks Options
14684 @cindex VxWorks Options
14686 The options in this section are defined for all VxWorks targets.
14687 Options specific to the target hardware are listed with the other
14688 options for that target.
14693 GCC can generate code for both VxWorks kernels and real time processes
14694 (RTPs). This option switches from the former to the latter. It also
14695 defines the preprocessor macro @code{__RTP__}.
14698 @opindex non-static
14699 Link an RTP executable against shared libraries rather than static
14700 libraries. The options @option{-static} and @option{-shared} can
14701 also be used for RTPs (@pxref{Link Options}); @option{-static}
14708 These options are passed down to the linker. They are defined for
14709 compatibility with Diab.
14712 @opindex Xbind-lazy
14713 Enable lazy binding of function calls. This option is equivalent to
14714 @option{-Wl,-z,now} and is defined for compatibility with Diab.
14718 Disable lazy binding of function calls. This option is the default and
14719 is defined for compatibility with Diab.
14722 @node x86-64 Options
14723 @subsection x86-64 Options
14724 @cindex x86-64 options
14726 These are listed under @xref{i386 and x86-64 Options}.
14728 @node Xstormy16 Options
14729 @subsection Xstormy16 Options
14730 @cindex Xstormy16 Options
14732 These options are defined for Xstormy16:
14737 Choose startup files and linker script suitable for the simulator.
14740 @node Xtensa Options
14741 @subsection Xtensa Options
14742 @cindex Xtensa Options
14744 These options are supported for Xtensa targets:
14748 @itemx -mno-const16
14750 @opindex mno-const16
14751 Enable or disable use of @code{CONST16} instructions for loading
14752 constant values. The @code{CONST16} instruction is currently not a
14753 standard option from Tensilica. When enabled, @code{CONST16}
14754 instructions are always used in place of the standard @code{L32R}
14755 instructions. The use of @code{CONST16} is enabled by default only if
14756 the @code{L32R} instruction is not available.
14759 @itemx -mno-fused-madd
14760 @opindex mfused-madd
14761 @opindex mno-fused-madd
14762 Enable or disable use of fused multiply/add and multiply/subtract
14763 instructions in the floating-point option. This has no effect if the
14764 floating-point option is not also enabled. Disabling fused multiply/add
14765 and multiply/subtract instructions forces the compiler to use separate
14766 instructions for the multiply and add/subtract operations. This may be
14767 desirable in some cases where strict IEEE 754-compliant results are
14768 required: the fused multiply add/subtract instructions do not round the
14769 intermediate result, thereby producing results with @emph{more} bits of
14770 precision than specified by the IEEE standard. Disabling fused multiply
14771 add/subtract instructions also ensures that the program output is not
14772 sensitive to the compiler's ability to combine multiply and add/subtract
14775 @item -mserialize-volatile
14776 @itemx -mno-serialize-volatile
14777 @opindex mserialize-volatile
14778 @opindex mno-serialize-volatile
14779 When this option is enabled, GCC inserts @code{MEMW} instructions before
14780 @code{volatile} memory references to guarantee sequential consistency.
14781 The default is @option{-mserialize-volatile}. Use
14782 @option{-mno-serialize-volatile} to omit the @code{MEMW} instructions.
14784 @item -mtext-section-literals
14785 @itemx -mno-text-section-literals
14786 @opindex mtext-section-literals
14787 @opindex mno-text-section-literals
14788 Control the treatment of literal pools. The default is
14789 @option{-mno-text-section-literals}, which places literals in a separate
14790 section in the output file. This allows the literal pool to be placed
14791 in a data RAM/ROM, and it also allows the linker to combine literal
14792 pools from separate object files to remove redundant literals and
14793 improve code size. With @option{-mtext-section-literals}, the literals
14794 are interspersed in the text section in order to keep them as close as
14795 possible to their references. This may be necessary for large assembly
14798 @item -mtarget-align
14799 @itemx -mno-target-align
14800 @opindex mtarget-align
14801 @opindex mno-target-align
14802 When this option is enabled, GCC instructs the assembler to
14803 automatically align instructions to reduce branch penalties at the
14804 expense of some code density. The assembler attempts to widen density
14805 instructions to align branch targets and the instructions following call
14806 instructions. If there are not enough preceding safe density
14807 instructions to align a target, no widening will be performed. The
14808 default is @option{-mtarget-align}. These options do not affect the
14809 treatment of auto-aligned instructions like @code{LOOP}, which the
14810 assembler will always align, either by widening density instructions or
14811 by inserting no-op instructions.
14814 @itemx -mno-longcalls
14815 @opindex mlongcalls
14816 @opindex mno-longcalls
14817 When this option is enabled, GCC instructs the assembler to translate
14818 direct calls to indirect calls unless it can determine that the target
14819 of a direct call is in the range allowed by the call instruction. This
14820 translation typically occurs for calls to functions in other source
14821 files. Specifically, the assembler translates a direct @code{CALL}
14822 instruction into an @code{L32R} followed by a @code{CALLX} instruction.
14823 The default is @option{-mno-longcalls}. This option should be used in
14824 programs where the call target can potentially be out of range. This
14825 option is implemented in the assembler, not the compiler, so the
14826 assembly code generated by GCC will still show direct call
14827 instructions---look at the disassembled object code to see the actual
14828 instructions. Note that the assembler will use an indirect call for
14829 every cross-file call, not just those that really will be out of range.
14832 @node zSeries Options
14833 @subsection zSeries Options
14834 @cindex zSeries options
14836 These are listed under @xref{S/390 and zSeries Options}.
14838 @node Code Gen Options
14839 @section Options for Code Generation Conventions
14840 @cindex code generation conventions
14841 @cindex options, code generation
14842 @cindex run-time options
14844 These machine-independent options control the interface conventions
14845 used in code generation.
14847 Most of them have both positive and negative forms; the negative form
14848 of @option{-ffoo} would be @option{-fno-foo}. In the table below, only
14849 one of the forms is listed---the one which is not the default. You
14850 can figure out the other form by either removing @samp{no-} or adding
14854 @item -fbounds-check
14855 @opindex fbounds-check
14856 For front-ends that support it, generate additional code to check that
14857 indices used to access arrays are within the declared range. This is
14858 currently only supported by the Java and Fortran front-ends, where
14859 this option defaults to true and false respectively.
14863 This option generates traps for signed overflow on addition, subtraction,
14864 multiplication operations.
14868 This option instructs the compiler to assume that signed arithmetic
14869 overflow of addition, subtraction and multiplication wraps around
14870 using twos-complement representation. This flag enables some optimizations
14871 and disables others. This option is enabled by default for the Java
14872 front-end, as required by the Java language specification.
14875 @opindex fexceptions
14876 Enable exception handling. Generates extra code needed to propagate
14877 exceptions. For some targets, this implies GCC will generate frame
14878 unwind information for all functions, which can produce significant data
14879 size overhead, although it does not affect execution. If you do not
14880 specify this option, GCC will enable it by default for languages like
14881 C++ which normally require exception handling, and disable it for
14882 languages like C that do not normally require it. However, you may need
14883 to enable this option when compiling C code that needs to interoperate
14884 properly with exception handlers written in C++. You may also wish to
14885 disable this option if you are compiling older C++ programs that don't
14886 use exception handling.
14888 @item -fnon-call-exceptions
14889 @opindex fnon-call-exceptions
14890 Generate code that allows trapping instructions to throw exceptions.
14891 Note that this requires platform-specific runtime support that does
14892 not exist everywhere. Moreover, it only allows @emph{trapping}
14893 instructions to throw exceptions, i.e.@: memory references or floating
14894 point instructions. It does not allow exceptions to be thrown from
14895 arbitrary signal handlers such as @code{SIGALRM}.
14897 @item -funwind-tables
14898 @opindex funwind-tables
14899 Similar to @option{-fexceptions}, except that it will just generate any needed
14900 static data, but will not affect the generated code in any other way.
14901 You will normally not enable this option; instead, a language processor
14902 that needs this handling would enable it on your behalf.
14904 @item -fasynchronous-unwind-tables
14905 @opindex fasynchronous-unwind-tables
14906 Generate unwind table in dwarf2 format, if supported by target machine. The
14907 table is exact at each instruction boundary, so it can be used for stack
14908 unwinding from asynchronous events (such as debugger or garbage collector).
14910 @item -fpcc-struct-return
14911 @opindex fpcc-struct-return
14912 Return ``short'' @code{struct} and @code{union} values in memory like
14913 longer ones, rather than in registers. This convention is less
14914 efficient, but it has the advantage of allowing intercallability between
14915 GCC-compiled files and files compiled with other compilers, particularly
14916 the Portable C Compiler (pcc).
14918 The precise convention for returning structures in memory depends
14919 on the target configuration macros.
14921 Short structures and unions are those whose size and alignment match
14922 that of some integer type.
14924 @strong{Warning:} code compiled with the @option{-fpcc-struct-return}
14925 switch is not binary compatible with code compiled with the
14926 @option{-freg-struct-return} switch.
14927 Use it to conform to a non-default application binary interface.
14929 @item -freg-struct-return
14930 @opindex freg-struct-return
14931 Return @code{struct} and @code{union} values in registers when possible.
14932 This is more efficient for small structures than
14933 @option{-fpcc-struct-return}.
14935 If you specify neither @option{-fpcc-struct-return} nor
14936 @option{-freg-struct-return}, GCC defaults to whichever convention is
14937 standard for the target. If there is no standard convention, GCC
14938 defaults to @option{-fpcc-struct-return}, except on targets where GCC is
14939 the principal compiler. In those cases, we can choose the standard, and
14940 we chose the more efficient register return alternative.
14942 @strong{Warning:} code compiled with the @option{-freg-struct-return}
14943 switch is not binary compatible with code compiled with the
14944 @option{-fpcc-struct-return} switch.
14945 Use it to conform to a non-default application binary interface.
14947 @item -fshort-enums
14948 @opindex fshort-enums
14949 Allocate to an @code{enum} type only as many bytes as it needs for the
14950 declared range of possible values. Specifically, the @code{enum} type
14951 will be equivalent to the smallest integer type which has enough room.
14953 @strong{Warning:} the @option{-fshort-enums} switch causes GCC to generate
14954 code that is not binary compatible with code generated without that switch.
14955 Use it to conform to a non-default application binary interface.
14957 @item -fshort-double
14958 @opindex fshort-double
14959 Use the same size for @code{double} as for @code{float}.
14961 @strong{Warning:} the @option{-fshort-double} switch causes GCC to generate
14962 code that is not binary compatible with code generated without that switch.
14963 Use it to conform to a non-default application binary interface.
14965 @item -fshort-wchar
14966 @opindex fshort-wchar
14967 Override the underlying type for @samp{wchar_t} to be @samp{short
14968 unsigned int} instead of the default for the target. This option is
14969 useful for building programs to run under WINE@.
14971 @strong{Warning:} the @option{-fshort-wchar} switch causes GCC to generate
14972 code that is not binary compatible with code generated without that switch.
14973 Use it to conform to a non-default application binary interface.
14976 @opindex fno-common
14977 In C, allocate even uninitialized global variables in the data section of the
14978 object file, rather than generating them as common blocks. This has the
14979 effect that if the same variable is declared (without @code{extern}) in
14980 two different compilations, you will get an error when you link them.
14981 The only reason this might be useful is if you wish to verify that the
14982 program will work on other systems which always work this way.
14986 Ignore the @samp{#ident} directive.
14988 @item -finhibit-size-directive
14989 @opindex finhibit-size-directive
14990 Don't output a @code{.size} assembler directive, or anything else that
14991 would cause trouble if the function is split in the middle, and the
14992 two halves are placed at locations far apart in memory. This option is
14993 used when compiling @file{crtstuff.c}; you should not need to use it
14996 @item -fverbose-asm
14997 @opindex fverbose-asm
14998 Put extra commentary information in the generated assembly code to
14999 make it more readable. This option is generally only of use to those
15000 who actually need to read the generated assembly code (perhaps while
15001 debugging the compiler itself).
15003 @option{-fno-verbose-asm}, the default, causes the
15004 extra information to be omitted and is useful when comparing two assembler
15007 @item -frecord-gcc-switches
15008 @opindex frecord-gcc-switches
15009 This switch causes the command line that was used to invoke the
15010 compiler to be recorded into the object file that is being created.
15011 This switch is only implemented on some targets and the exact format
15012 of the recording is target and binary file format dependent, but it
15013 usually takes the form of a section containing ASCII text. This
15014 switch is related to the @option{-fverbose-asm} switch, but that
15015 switch only records information in the assembler output file as
15016 comments, so it never reaches the object file.
15020 @cindex global offset table
15022 Generate position-independent code (PIC) suitable for use in a shared
15023 library, if supported for the target machine. Such code accesses all
15024 constant addresses through a global offset table (GOT)@. The dynamic
15025 loader resolves the GOT entries when the program starts (the dynamic
15026 loader is not part of GCC; it is part of the operating system). If
15027 the GOT size for the linked executable exceeds a machine-specific
15028 maximum size, you get an error message from the linker indicating that
15029 @option{-fpic} does not work; in that case, recompile with @option{-fPIC}
15030 instead. (These maximums are 8k on the SPARC and 32k
15031 on the m68k and RS/6000. The 386 has no such limit.)
15033 Position-independent code requires special support, and therefore works
15034 only on certain machines. For the 386, GCC supports PIC for System V
15035 but not for the Sun 386i. Code generated for the IBM RS/6000 is always
15036 position-independent.
15038 When this flag is set, the macros @code{__pic__} and @code{__PIC__}
15043 If supported for the target machine, emit position-independent code,
15044 suitable for dynamic linking and avoiding any limit on the size of the
15045 global offset table. This option makes a difference on the m68k,
15046 PowerPC and SPARC@.
15048 Position-independent code requires special support, and therefore works
15049 only on certain machines.
15051 When this flag is set, the macros @code{__pic__} and @code{__PIC__}
15058 These options are similar to @option{-fpic} and @option{-fPIC}, but
15059 generated position independent code can be only linked into executables.
15060 Usually these options are used when @option{-pie} GCC option will be
15061 used during linking.
15063 @option{-fpie} and @option{-fPIE} both define the macros
15064 @code{__pie__} and @code{__PIE__}. The macros have the value 1
15065 for @option{-fpie} and 2 for @option{-fPIE}.
15067 @item -fno-jump-tables
15068 @opindex fno-jump-tables
15069 Do not use jump tables for switch statements even where it would be
15070 more efficient than other code generation strategies. This option is
15071 of use in conjunction with @option{-fpic} or @option{-fPIC} for
15072 building code which forms part of a dynamic linker and cannot
15073 reference the address of a jump table. On some targets, jump tables
15074 do not require a GOT and this option is not needed.
15076 @item -ffixed-@var{reg}
15078 Treat the register named @var{reg} as a fixed register; generated code
15079 should never refer to it (except perhaps as a stack pointer, frame
15080 pointer or in some other fixed role).
15082 @var{reg} must be the name of a register. The register names accepted
15083 are machine-specific and are defined in the @code{REGISTER_NAMES}
15084 macro in the machine description macro file.
15086 This flag does not have a negative form, because it specifies a
15089 @item -fcall-used-@var{reg}
15090 @opindex fcall-used
15091 Treat the register named @var{reg} as an allocable register that is
15092 clobbered by function calls. It may be allocated for temporaries or
15093 variables that do not live across a call. Functions compiled this way
15094 will not save and restore the register @var{reg}.
15096 It is an error to used this flag with the frame pointer or stack pointer.
15097 Use of this flag for other registers that have fixed pervasive roles in
15098 the machine's execution model will produce disastrous results.
15100 This flag does not have a negative form, because it specifies a
15103 @item -fcall-saved-@var{reg}
15104 @opindex fcall-saved
15105 Treat the register named @var{reg} as an allocable register saved by
15106 functions. It may be allocated even for temporaries or variables that
15107 live across a call. Functions compiled this way will save and restore
15108 the register @var{reg} if they use it.
15110 It is an error to used this flag with the frame pointer or stack pointer.
15111 Use of this flag for other registers that have fixed pervasive roles in
15112 the machine's execution model will produce disastrous results.
15114 A different sort of disaster will result from the use of this flag for
15115 a register in which function values may be returned.
15117 This flag does not have a negative form, because it specifies a
15120 @item -fpack-struct[=@var{n}]
15121 @opindex fpack-struct
15122 Without a value specified, pack all structure members together without
15123 holes. When a value is specified (which must be a small power of two), pack
15124 structure members according to this value, representing the maximum
15125 alignment (that is, objects with default alignment requirements larger than
15126 this will be output potentially unaligned at the next fitting location.
15128 @strong{Warning:} the @option{-fpack-struct} switch causes GCC to generate
15129 code that is not binary compatible with code generated without that switch.
15130 Additionally, it makes the code suboptimal.
15131 Use it to conform to a non-default application binary interface.
15133 @item -finstrument-functions
15134 @opindex finstrument-functions
15135 Generate instrumentation calls for entry and exit to functions. Just
15136 after function entry and just before function exit, the following
15137 profiling functions will be called with the address of the current
15138 function and its call site. (On some platforms,
15139 @code{__builtin_return_address} does not work beyond the current
15140 function, so the call site information may not be available to the
15141 profiling functions otherwise.)
15144 void __cyg_profile_func_enter (void *this_fn,
15146 void __cyg_profile_func_exit (void *this_fn,
15150 The first argument is the address of the start of the current function,
15151 which may be looked up exactly in the symbol table.
15153 This instrumentation is also done for functions expanded inline in other
15154 functions. The profiling calls will indicate where, conceptually, the
15155 inline function is entered and exited. This means that addressable
15156 versions of such functions must be available. If all your uses of a
15157 function are expanded inline, this may mean an additional expansion of
15158 code size. If you use @samp{extern inline} in your C code, an
15159 addressable version of such functions must be provided. (This is
15160 normally the case anyways, but if you get lucky and the optimizer always
15161 expands the functions inline, you might have gotten away without
15162 providing static copies.)
15164 A function may be given the attribute @code{no_instrument_function}, in
15165 which case this instrumentation will not be done. This can be used, for
15166 example, for the profiling functions listed above, high-priority
15167 interrupt routines, and any functions from which the profiling functions
15168 cannot safely be called (perhaps signal handlers, if the profiling
15169 routines generate output or allocate memory).
15171 @item -finstrument-functions-exclude-file-list=@var{file},@var{file},@dots{}
15172 @opindex finstrument-functions-exclude-file-list
15174 Set the list of functions that are excluded from instrumentation (see
15175 the description of @code{-finstrument-functions}). If the file that
15176 contains a function definition matches with one of @var{file}, then
15177 that function is not instrumented. The match is done on substrings:
15178 if the @var{file} parameter is a substring of the file name, it is
15179 considered to be a match.
15182 @code{-finstrument-functions-exclude-file-list=/bits/stl,include/sys}
15183 will exclude any inline function defined in files whose pathnames
15184 contain @code{/bits/stl} or @code{include/sys}.
15186 If, for some reason, you want to include letter @code{','} in one of
15187 @var{sym}, write @code{'\,'}. For example,
15188 @code{-finstrument-functions-exclude-file-list='\,\,tmp'}
15189 (note the single quote surrounding the option).
15191 @item -finstrument-functions-exclude-function-list=@var{sym},@var{sym},@dots{}
15192 @opindex finstrument-functions-exclude-function-list
15194 This is similar to @code{-finstrument-functions-exclude-file-list},
15195 but this option sets the list of function names to be excluded from
15196 instrumentation. The function name to be matched is its user-visible
15197 name, such as @code{vector<int> blah(const vector<int> &)}, not the
15198 internal mangled name (e.g., @code{_Z4blahRSt6vectorIiSaIiEE}). The
15199 match is done on substrings: if the @var{sym} parameter is a substring
15200 of the function name, it is considered to be a match.
15202 @item -fstack-check
15203 @opindex fstack-check
15204 Generate code to verify that you do not go beyond the boundary of the
15205 stack. You should specify this flag if you are running in an
15206 environment with multiple threads, but only rarely need to specify it in
15207 a single-threaded environment since stack overflow is automatically
15208 detected on nearly all systems if there is only one stack.
15210 Note that this switch does not actually cause checking to be done; the
15211 operating system must do that. The switch causes generation of code
15212 to ensure that the operating system sees the stack being extended.
15214 @item -fstack-limit-register=@var{reg}
15215 @itemx -fstack-limit-symbol=@var{sym}
15216 @itemx -fno-stack-limit
15217 @opindex fstack-limit-register
15218 @opindex fstack-limit-symbol
15219 @opindex fno-stack-limit
15220 Generate code to ensure that the stack does not grow beyond a certain value,
15221 either the value of a register or the address of a symbol. If the stack
15222 would grow beyond the value, a signal is raised. For most targets,
15223 the signal is raised before the stack overruns the boundary, so
15224 it is possible to catch the signal without taking special precautions.
15226 For instance, if the stack starts at absolute address @samp{0x80000000}
15227 and grows downwards, you can use the flags
15228 @option{-fstack-limit-symbol=__stack_limit} and
15229 @option{-Wl,--defsym,__stack_limit=0x7ffe0000} to enforce a stack limit
15230 of 128KB@. Note that this may only work with the GNU linker.
15232 @cindex aliasing of parameters
15233 @cindex parameters, aliased
15234 @item -fargument-alias
15235 @itemx -fargument-noalias
15236 @itemx -fargument-noalias-global
15237 @itemx -fargument-noalias-anything
15238 @opindex fargument-alias
15239 @opindex fargument-noalias
15240 @opindex fargument-noalias-global
15241 @opindex fargument-noalias-anything
15242 Specify the possible relationships among parameters and between
15243 parameters and global data.
15245 @option{-fargument-alias} specifies that arguments (parameters) may
15246 alias each other and may alias global storage.@*
15247 @option{-fargument-noalias} specifies that arguments do not alias
15248 each other, but may alias global storage.@*
15249 @option{-fargument-noalias-global} specifies that arguments do not
15250 alias each other and do not alias global storage.
15251 @option{-fargument-noalias-anything} specifies that arguments do not
15252 alias any other storage.
15254 Each language will automatically use whatever option is required by
15255 the language standard. You should not need to use these options yourself.
15257 @item -fleading-underscore
15258 @opindex fleading-underscore
15259 This option and its counterpart, @option{-fno-leading-underscore}, forcibly
15260 change the way C symbols are represented in the object file. One use
15261 is to help link with legacy assembly code.
15263 @strong{Warning:} the @option{-fleading-underscore} switch causes GCC to
15264 generate code that is not binary compatible with code generated without that
15265 switch. Use it to conform to a non-default application binary interface.
15266 Not all targets provide complete support for this switch.
15268 @item -ftls-model=@var{model}
15269 @opindex ftls-model
15270 Alter the thread-local storage model to be used (@pxref{Thread-Local}).
15271 The @var{model} argument should be one of @code{global-dynamic},
15272 @code{local-dynamic}, @code{initial-exec} or @code{local-exec}.
15274 The default without @option{-fpic} is @code{initial-exec}; with
15275 @option{-fpic} the default is @code{global-dynamic}.
15277 @item -fvisibility=@var{default|internal|hidden|protected}
15278 @opindex fvisibility
15279 Set the default ELF image symbol visibility to the specified option---all
15280 symbols will be marked with this unless overridden within the code.
15281 Using this feature can very substantially improve linking and
15282 load times of shared object libraries, produce more optimized
15283 code, provide near-perfect API export and prevent symbol clashes.
15284 It is @strong{strongly} recommended that you use this in any shared objects
15287 Despite the nomenclature, @code{default} always means public ie;
15288 available to be linked against from outside the shared object.
15289 @code{protected} and @code{internal} are pretty useless in real-world
15290 usage so the only other commonly used option will be @code{hidden}.
15291 The default if @option{-fvisibility} isn't specified is
15292 @code{default}, i.e., make every
15293 symbol public---this causes the same behavior as previous versions of
15296 A good explanation of the benefits offered by ensuring ELF
15297 symbols have the correct visibility is given by ``How To Write
15298 Shared Libraries'' by Ulrich Drepper (which can be found at
15299 @w{@uref{http://people.redhat.com/~drepper/}})---however a superior
15300 solution made possible by this option to marking things hidden when
15301 the default is public is to make the default hidden and mark things
15302 public. This is the norm with DLL's on Windows and with @option{-fvisibility=hidden}
15303 and @code{__attribute__ ((visibility("default")))} instead of
15304 @code{__declspec(dllexport)} you get almost identical semantics with
15305 identical syntax. This is a great boon to those working with
15306 cross-platform projects.
15308 For those adding visibility support to existing code, you may find
15309 @samp{#pragma GCC visibility} of use. This works by you enclosing
15310 the declarations you wish to set visibility for with (for example)
15311 @samp{#pragma GCC visibility push(hidden)} and
15312 @samp{#pragma GCC visibility pop}.
15313 Bear in mind that symbol visibility should be viewed @strong{as
15314 part of the API interface contract} and thus all new code should
15315 always specify visibility when it is not the default ie; declarations
15316 only for use within the local DSO should @strong{always} be marked explicitly
15317 as hidden as so to avoid PLT indirection overheads---making this
15318 abundantly clear also aids readability and self-documentation of the code.
15319 Note that due to ISO C++ specification requirements, operator new and
15320 operator delete must always be of default visibility.
15322 Be aware that headers from outside your project, in particular system
15323 headers and headers from any other library you use, may not be
15324 expecting to be compiled with visibility other than the default. You
15325 may need to explicitly say @samp{#pragma GCC visibility push(default)}
15326 before including any such headers.
15328 @samp{extern} declarations are not affected by @samp{-fvisibility}, so
15329 a lot of code can be recompiled with @samp{-fvisibility=hidden} with
15330 no modifications. However, this means that calls to @samp{extern}
15331 functions with no explicit visibility will use the PLT, so it is more
15332 effective to use @samp{__attribute ((visibility))} and/or
15333 @samp{#pragma GCC visibility} to tell the compiler which @samp{extern}
15334 declarations should be treated as hidden.
15336 Note that @samp{-fvisibility} does affect C++ vague linkage
15337 entities. This means that, for instance, an exception class that will
15338 be thrown between DSOs must be explicitly marked with default
15339 visibility so that the @samp{type_info} nodes will be unified between
15342 An overview of these techniques, their benefits and how to use them
15343 is at @w{@uref{http://gcc.gnu.org/wiki/Visibility}}.
15349 @node Environment Variables
15350 @section Environment Variables Affecting GCC
15351 @cindex environment variables
15353 @c man begin ENVIRONMENT
15354 This section describes several environment variables that affect how GCC
15355 operates. Some of them work by specifying directories or prefixes to use
15356 when searching for various kinds of files. Some are used to specify other
15357 aspects of the compilation environment.
15359 Note that you can also specify places to search using options such as
15360 @option{-B}, @option{-I} and @option{-L} (@pxref{Directory Options}). These
15361 take precedence over places specified using environment variables, which
15362 in turn take precedence over those specified by the configuration of GCC@.
15363 @xref{Driver,, Controlling the Compilation Driver @file{gcc}, gccint,
15364 GNU Compiler Collection (GCC) Internals}.
15369 @c @itemx LC_COLLATE
15371 @c @itemx LC_MONETARY
15372 @c @itemx LC_NUMERIC
15377 @c @findex LC_COLLATE
15378 @findex LC_MESSAGES
15379 @c @findex LC_MONETARY
15380 @c @findex LC_NUMERIC
15384 These environment variables control the way that GCC uses
15385 localization information that allow GCC to work with different
15386 national conventions. GCC inspects the locale categories
15387 @env{LC_CTYPE} and @env{LC_MESSAGES} if it has been configured to do
15388 so. These locale categories can be set to any value supported by your
15389 installation. A typical value is @samp{en_GB.UTF-8} for English in the United
15390 Kingdom encoded in UTF-8.
15392 The @env{LC_CTYPE} environment variable specifies character
15393 classification. GCC uses it to determine the character boundaries in
15394 a string; this is needed for some multibyte encodings that contain quote
15395 and escape characters that would otherwise be interpreted as a string
15398 The @env{LC_MESSAGES} environment variable specifies the language to
15399 use in diagnostic messages.
15401 If the @env{LC_ALL} environment variable is set, it overrides the value
15402 of @env{LC_CTYPE} and @env{LC_MESSAGES}; otherwise, @env{LC_CTYPE}
15403 and @env{LC_MESSAGES} default to the value of the @env{LANG}
15404 environment variable. If none of these variables are set, GCC
15405 defaults to traditional C English behavior.
15409 If @env{TMPDIR} is set, it specifies the directory to use for temporary
15410 files. GCC uses temporary files to hold the output of one stage of
15411 compilation which is to be used as input to the next stage: for example,
15412 the output of the preprocessor, which is the input to the compiler
15415 @item GCC_EXEC_PREFIX
15416 @findex GCC_EXEC_PREFIX
15417 If @env{GCC_EXEC_PREFIX} is set, it specifies a prefix to use in the
15418 names of the subprograms executed by the compiler. No slash is added
15419 when this prefix is combined with the name of a subprogram, but you can
15420 specify a prefix that ends with a slash if you wish.
15422 If @env{GCC_EXEC_PREFIX} is not set, GCC will attempt to figure out
15423 an appropriate prefix to use based on the pathname it was invoked with.
15425 If GCC cannot find the subprogram using the specified prefix, it
15426 tries looking in the usual places for the subprogram.
15428 The default value of @env{GCC_EXEC_PREFIX} is
15429 @file{@var{prefix}/lib/gcc/} where @var{prefix} is the prefix to
15430 the installed compiler. In many cases @var{prefix} is the value
15431 of @code{prefix} when you ran the @file{configure} script.
15433 Other prefixes specified with @option{-B} take precedence over this prefix.
15435 This prefix is also used for finding files such as @file{crt0.o} that are
15438 In addition, the prefix is used in an unusual way in finding the
15439 directories to search for header files. For each of the standard
15440 directories whose name normally begins with @samp{/usr/local/lib/gcc}
15441 (more precisely, with the value of @env{GCC_INCLUDE_DIR}), GCC tries
15442 replacing that beginning with the specified prefix to produce an
15443 alternate directory name. Thus, with @option{-Bfoo/}, GCC will search
15444 @file{foo/bar} where it would normally search @file{/usr/local/lib/bar}.
15445 These alternate directories are searched first; the standard directories
15446 come next. If a standard directory begins with the configured
15447 @var{prefix} then the value of @var{prefix} is replaced by
15448 @env{GCC_EXEC_PREFIX} when looking for header files.
15450 @item COMPILER_PATH
15451 @findex COMPILER_PATH
15452 The value of @env{COMPILER_PATH} is a colon-separated list of
15453 directories, much like @env{PATH}. GCC tries the directories thus
15454 specified when searching for subprograms, if it can't find the
15455 subprograms using @env{GCC_EXEC_PREFIX}.
15458 @findex LIBRARY_PATH
15459 The value of @env{LIBRARY_PATH} is a colon-separated list of
15460 directories, much like @env{PATH}. When configured as a native compiler,
15461 GCC tries the directories thus specified when searching for special
15462 linker files, if it can't find them using @env{GCC_EXEC_PREFIX}. Linking
15463 using GCC also uses these directories when searching for ordinary
15464 libraries for the @option{-l} option (but directories specified with
15465 @option{-L} come first).
15469 @cindex locale definition
15470 This variable is used to pass locale information to the compiler. One way in
15471 which this information is used is to determine the character set to be used
15472 when character literals, string literals and comments are parsed in C and C++.
15473 When the compiler is configured to allow multibyte characters,
15474 the following values for @env{LANG} are recognized:
15478 Recognize JIS characters.
15480 Recognize SJIS characters.
15482 Recognize EUCJP characters.
15485 If @env{LANG} is not defined, or if it has some other value, then the
15486 compiler will use mblen and mbtowc as defined by the default locale to
15487 recognize and translate multibyte characters.
15491 Some additional environments variables affect the behavior of the
15494 @include cppenv.texi
15498 @node Precompiled Headers
15499 @section Using Precompiled Headers
15500 @cindex precompiled headers
15501 @cindex speed of compilation
15503 Often large projects have many header files that are included in every
15504 source file. The time the compiler takes to process these header files
15505 over and over again can account for nearly all of the time required to
15506 build the project. To make builds faster, GCC allows users to
15507 `precompile' a header file; then, if builds can use the precompiled
15508 header file they will be much faster.
15510 To create a precompiled header file, simply compile it as you would any
15511 other file, if necessary using the @option{-x} option to make the driver
15512 treat it as a C or C++ header file. You will probably want to use a
15513 tool like @command{make} to keep the precompiled header up-to-date when
15514 the headers it contains change.
15516 A precompiled header file will be searched for when @code{#include} is
15517 seen in the compilation. As it searches for the included file
15518 (@pxref{Search Path,,Search Path,cpp,The C Preprocessor}) the
15519 compiler looks for a precompiled header in each directory just before it
15520 looks for the include file in that directory. The name searched for is
15521 the name specified in the @code{#include} with @samp{.gch} appended. If
15522 the precompiled header file can't be used, it is ignored.
15524 For instance, if you have @code{#include "all.h"}, and you have
15525 @file{all.h.gch} in the same directory as @file{all.h}, then the
15526 precompiled header file will be used if possible, and the original
15527 header will be used otherwise.
15529 Alternatively, you might decide to put the precompiled header file in a
15530 directory and use @option{-I} to ensure that directory is searched
15531 before (or instead of) the directory containing the original header.
15532 Then, if you want to check that the precompiled header file is always
15533 used, you can put a file of the same name as the original header in this
15534 directory containing an @code{#error} command.
15536 This also works with @option{-include}. So yet another way to use
15537 precompiled headers, good for projects not designed with precompiled
15538 header files in mind, is to simply take most of the header files used by
15539 a project, include them from another header file, precompile that header
15540 file, and @option{-include} the precompiled header. If the header files
15541 have guards against multiple inclusion, they will be skipped because
15542 they've already been included (in the precompiled header).
15544 If you need to precompile the same header file for different
15545 languages, targets, or compiler options, you can instead make a
15546 @emph{directory} named like @file{all.h.gch}, and put each precompiled
15547 header in the directory, perhaps using @option{-o}. It doesn't matter
15548 what you call the files in the directory, every precompiled header in
15549 the directory will be considered. The first precompiled header
15550 encountered in the directory that is valid for this compilation will
15551 be used; they're searched in no particular order.
15553 There are many other possibilities, limited only by your imagination,
15554 good sense, and the constraints of your build system.
15556 A precompiled header file can be used only when these conditions apply:
15560 Only one precompiled header can be used in a particular compilation.
15563 A precompiled header can't be used once the first C token is seen. You
15564 can have preprocessor directives before a precompiled header; you can
15565 even include a precompiled header from inside another header, so long as
15566 there are no C tokens before the @code{#include}.
15569 The precompiled header file must be produced for the same language as
15570 the current compilation. You can't use a C precompiled header for a C++
15574 The precompiled header file must have been produced by the same compiler
15575 binary as the current compilation is using.
15578 Any macros defined before the precompiled header is included must
15579 either be defined in the same way as when the precompiled header was
15580 generated, or must not affect the precompiled header, which usually
15581 means that they don't appear in the precompiled header at all.
15583 The @option{-D} option is one way to define a macro before a
15584 precompiled header is included; using a @code{#define} can also do it.
15585 There are also some options that define macros implicitly, like
15586 @option{-O} and @option{-Wdeprecated}; the same rule applies to macros
15589 @item If debugging information is output when using the precompiled
15590 header, using @option{-g} or similar, the same kind of debugging information
15591 must have been output when building the precompiled header. However,
15592 a precompiled header built using @option{-g} can be used in a compilation
15593 when no debugging information is being output.
15595 @item The same @option{-m} options must generally be used when building
15596 and using the precompiled header. @xref{Submodel Options},
15597 for any cases where this rule is relaxed.
15599 @item Each of the following options must be the same when building and using
15600 the precompiled header:
15602 @gccoptlist{-fexceptions}
15605 Some other command-line options starting with @option{-f},
15606 @option{-p}, or @option{-O} must be defined in the same way as when
15607 the precompiled header was generated. At present, it's not clear
15608 which options are safe to change and which are not; the safest choice
15609 is to use exactly the same options when generating and using the
15610 precompiled header. The following are known to be safe:
15612 @gccoptlist{-fmessage-length= -fpreprocessed -fsched-interblock @gol
15613 -fsched-spec -fsched-spec-load -fsched-spec-load-dangerous @gol
15614 -fsched-verbose=<number> -fschedule-insns -fvisibility= @gol
15619 For all of these except the last, the compiler will automatically
15620 ignore the precompiled header if the conditions aren't met. If you
15621 find an option combination that doesn't work and doesn't cause the
15622 precompiled header to be ignored, please consider filing a bug report,
15625 If you do use differing options when generating and using the
15626 precompiled header, the actual behavior will be a mixture of the
15627 behavior for the options. For instance, if you use @option{-g} to
15628 generate the precompiled header but not when using it, you may or may
15629 not get debugging information for routines in the precompiled header.
15631 @node Running Protoize
15632 @section Running Protoize
15634 The program @code{protoize} is an optional part of GCC@. You can use
15635 it to add prototypes to a program, thus converting the program to ISO
15636 C in one respect. The companion program @code{unprotoize} does the
15637 reverse: it removes argument types from any prototypes that are found.
15639 When you run these programs, you must specify a set of source files as
15640 command line arguments. The conversion programs start out by compiling
15641 these files to see what functions they define. The information gathered
15642 about a file @var{foo} is saved in a file named @file{@var{foo}.X}.
15644 After scanning comes actual conversion. The specified files are all
15645 eligible to be converted; any files they include (whether sources or
15646 just headers) are eligible as well.
15648 But not all the eligible files are converted. By default,
15649 @code{protoize} and @code{unprotoize} convert only source and header
15650 files in the current directory. You can specify additional directories
15651 whose files should be converted with the @option{-d @var{directory}}
15652 option. You can also specify particular files to exclude with the
15653 @option{-x @var{file}} option. A file is converted if it is eligible, its
15654 directory name matches one of the specified directory names, and its
15655 name within the directory has not been excluded.
15657 Basic conversion with @code{protoize} consists of rewriting most
15658 function definitions and function declarations to specify the types of
15659 the arguments. The only ones not rewritten are those for varargs
15662 @code{protoize} optionally inserts prototype declarations at the
15663 beginning of the source file, to make them available for any calls that
15664 precede the function's definition. Or it can insert prototype
15665 declarations with block scope in the blocks where undeclared functions
15668 Basic conversion with @code{unprotoize} consists of rewriting most
15669 function declarations to remove any argument types, and rewriting
15670 function definitions to the old-style pre-ISO form.
15672 Both conversion programs print a warning for any function declaration or
15673 definition that they can't convert. You can suppress these warnings
15676 The output from @code{protoize} or @code{unprotoize} replaces the
15677 original source file. The original file is renamed to a name ending
15678 with @samp{.save} (for DOS, the saved filename ends in @samp{.sav}
15679 without the original @samp{.c} suffix). If the @samp{.save} (@samp{.sav}
15680 for DOS) file already exists, then the source file is simply discarded.
15682 @code{protoize} and @code{unprotoize} both depend on GCC itself to
15683 scan the program and collect information about the functions it uses.
15684 So neither of these programs will work until GCC is installed.
15686 Here is a table of the options you can use with @code{protoize} and
15687 @code{unprotoize}. Each option works with both programs unless
15691 @item -B @var{directory}
15692 Look for the file @file{SYSCALLS.c.X} in @var{directory}, instead of the
15693 usual directory (normally @file{/usr/local/lib}). This file contains
15694 prototype information about standard system functions. This option
15695 applies only to @code{protoize}.
15697 @item -c @var{compilation-options}
15698 Use @var{compilation-options} as the options when running @command{gcc} to
15699 produce the @samp{.X} files. The special option @option{-aux-info} is
15700 always passed in addition, to tell @command{gcc} to write a @samp{.X} file.
15702 Note that the compilation options must be given as a single argument to
15703 @code{protoize} or @code{unprotoize}. If you want to specify several
15704 @command{gcc} options, you must quote the entire set of compilation options
15705 to make them a single word in the shell.
15707 There are certain @command{gcc} arguments that you cannot use, because they
15708 would produce the wrong kind of output. These include @option{-g},
15709 @option{-O}, @option{-c}, @option{-S}, and @option{-o} If you include these in
15710 the @var{compilation-options}, they are ignored.
15713 Rename files to end in @samp{.C} (@samp{.cc} for DOS-based file
15714 systems) instead of @samp{.c}. This is convenient if you are converting
15715 a C program to C++. This option applies only to @code{protoize}.
15718 Add explicit global declarations. This means inserting explicit
15719 declarations at the beginning of each source file for each function
15720 that is called in the file and was not declared. These declarations
15721 precede the first function definition that contains a call to an
15722 undeclared function. This option applies only to @code{protoize}.
15724 @item -i @var{string}
15725 Indent old-style parameter declarations with the string @var{string}.
15726 This option applies only to @code{protoize}.
15728 @code{unprotoize} converts prototyped function definitions to old-style
15729 function definitions, where the arguments are declared between the
15730 argument list and the initial @samp{@{}. By default, @code{unprotoize}
15731 uses five spaces as the indentation. If you want to indent with just
15732 one space instead, use @option{-i " "}.
15735 Keep the @samp{.X} files. Normally, they are deleted after conversion
15739 Add explicit local declarations. @code{protoize} with @option{-l} inserts
15740 a prototype declaration for each function in each block which calls the
15741 function without any declaration. This option applies only to
15745 Make no real changes. This mode just prints information about the conversions
15746 that would have been done without @option{-n}.
15749 Make no @samp{.save} files. The original files are simply deleted.
15750 Use this option with caution.
15752 @item -p @var{program}
15753 Use the program @var{program} as the compiler. Normally, the name
15754 @file{gcc} is used.
15757 Work quietly. Most warnings are suppressed.
15760 Print the version number, just like @option{-v} for @command{gcc}.
15763 If you need special compiler options to compile one of your program's
15764 source files, then you should generate that file's @samp{.X} file
15765 specially, by running @command{gcc} on that source file with the
15766 appropriate options and the option @option{-aux-info}. Then run
15767 @code{protoize} on the entire set of files. @code{protoize} will use
15768 the existing @samp{.X} file because it is newer than the source file.
15772 gcc -Dfoo=bar file1.c -aux-info file1.X
15777 You need to include the special files along with the rest in the
15778 @code{protoize} command, even though their @samp{.X} files already
15779 exist, because otherwise they won't get converted.
15781 @xref{Protoize Caveats}, for more information on how to use
15782 @code{protoize} successfully.