1<html lang="en"> 2<head> 3<title>Debugging Options - Using the GNU Compiler Collection (GCC)</title> 4<meta http-equiv="Content-Type" content="text/html"> 5<meta name="description" content="Using the GNU Compiler Collection (GCC)"> 6<meta name="generator" content="makeinfo 4.13"> 7<link title="Top" rel="start" href="index.html#Top"> 8<link rel="up" href="Invoking-GCC.html#Invoking-GCC" title="Invoking GCC"> 9<link rel="prev" href="Warning-Options.html#Warning-Options" title="Warning Options"> 10<link rel="next" href="Optimize-Options.html#Optimize-Options" title="Optimize Options"> 11<link href="http://www.gnu.org/software/texinfo/" rel="generator-home" title="Texinfo Homepage"> 12<!-- 13Copyright (C) 1988, 1989, 1992, 1993, 1994, 1995, 1996, 1997, 141998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 152010 Free Software Foundation, Inc. 16 17Permission is granted to copy, distribute and/or modify this document 18under the terms of the GNU Free Documentation License, Version 1.3 or 19any later version published by the Free Software Foundation; with the 20Invariant Sections being ``Funding Free Software'', the Front-Cover 21Texts being (a) (see below), and with the Back-Cover Texts being (b) 22(see below). A copy of the license is included in the section entitled 23``GNU Free Documentation License''. 24 25(a) The FSF's Front-Cover Text is: 26 27 A GNU Manual 28 29(b) The FSF's Back-Cover Text is: 30 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.--> 34<meta http-equiv="Content-Style-Type" content="text/css"> 35<style type="text/css"><!-- 36 pre.display { font-family:inherit } 37 pre.format { font-family:inherit } 38 pre.smalldisplay { font-family:inherit; font-size:smaller } 39 pre.smallformat { font-family:inherit; font-size:smaller } 40 pre.smallexample { font-size:smaller } 41 pre.smalllisp { font-size:smaller } 42 span.sc { font-variant:small-caps } 43 span.roman { font-family:serif; font-weight:normal; } 44 span.sansserif { font-family:sans-serif; font-weight:normal; } 45--></style> 46<link rel="stylesheet" type="text/css" href="../cs.css"> 47</head> 48<body> 49<div class="node"> 50<a name="Debugging-Options"></a> 51<p> 52Next: <a rel="next" accesskey="n" href="Optimize-Options.html#Optimize-Options">Optimize Options</a>, 53Previous: <a rel="previous" accesskey="p" href="Warning-Options.html#Warning-Options">Warning Options</a>, 54Up: <a rel="up" accesskey="u" href="Invoking-GCC.html#Invoking-GCC">Invoking GCC</a> 55<hr> 56</div> 57 58<h3 class="section">3.9 Options for Debugging Your Program or GCC</h3> 59 60<p><a name="index-options_002c-debugging-500"></a><a name="index-debugging-information-options-501"></a> 61GCC has various special options that are used for debugging 62either your program or GCC: 63 64 <dl> 65<dt><code>-g</code><dd><a name="index-g-502"></a>Produce debugging information in the operating system's native format 66(stabs, COFF, XCOFF, or DWARF 2). GDB can work with this debugging 67information. 68 69 <p>On most systems that use stabs format, <samp><span class="option">-g</span></samp> enables use of extra 70debugging information that only GDB can use; this extra information 71makes debugging work better in GDB but will probably make other debuggers 72crash or 73refuse to read the program. If you want to control for certain whether 74to generate the extra information, use <samp><span class="option">-gstabs+</span></samp>, <samp><span class="option">-gstabs</span></samp>, 75<samp><span class="option">-gxcoff+</span></samp>, <samp><span class="option">-gxcoff</span></samp>, or <samp><span class="option">-gvms</span></samp> (see below). 76 77 <p>GCC allows you to use <samp><span class="option">-g</span></samp> with 78<samp><span class="option">-O</span></samp>. The shortcuts taken by optimized code may occasionally 79produce surprising results: some variables you declared may not exist 80at all; flow of control may briefly move where you did not expect it; 81some statements may not be executed because they compute constant 82results or their values were already at hand; some statements may 83execute in different places because they were moved out of loops. 84 85 <p>Nevertheless it proves possible to debug optimized output. This makes 86it reasonable to use the optimizer for programs that might have bugs. 87 88 <p>The following options are useful when GCC is generated with the 89capability for more than one debugging format. 90 91 <br><dt><code>-ggdb</code><dd><a name="index-ggdb-503"></a>Produce debugging information for use by GDB. This means to use the 92most expressive format available (DWARF 2, stabs, or the native format 93if neither of those are supported), including GDB extensions if at all 94possible. 95 96 <br><dt><code>-gstabs</code><dd><a name="index-gstabs-504"></a>Produce debugging information in stabs format (if that is supported), 97without GDB extensions. This is the format used by DBX on most BSD 98systems. On MIPS, Alpha and System V Release 4 systems this option 99produces stabs debugging output which is not understood by DBX or SDB. 100On System V Release 4 systems this option requires the GNU assembler. 101 102 <br><dt><code>-feliminate-unused-debug-symbols</code><dd><a name="index-feliminate_002dunused_002ddebug_002dsymbols-505"></a>Produce debugging information in stabs format (if that is supported), 103for only symbols that are actually used. 104 105 <br><dt><code>-femit-class-debug-always</code><dd>Instead of emitting debugging information for a C++ class in only one 106object file, emit it in all object files using the class. This option 107should be used only with debuggers that are unable to handle the way GCC 108normally emits debugging information for classes because using this 109option will increase the size of debugging information by as much as a 110factor of two. 111 112 <br><dt><code>-gstabs+</code><dd><a name="index-gstabs_002b-506"></a>Produce debugging information in stabs format (if that is supported), 113using GNU extensions understood only by the GNU debugger (GDB). The 114use of these extensions is likely to make other debuggers crash or 115refuse to read the program. 116 117 <br><dt><code>-gcoff</code><dd><a name="index-gcoff-507"></a>Produce debugging information in COFF format (if that is supported). 118This is the format used by SDB on most System V systems prior to 119System V Release 4. 120 121 <br><dt><code>-gxcoff</code><dd><a name="index-gxcoff-508"></a>Produce debugging information in XCOFF format (if that is supported). 122This is the format used by the DBX debugger on IBM RS/6000 systems. 123 124 <br><dt><code>-gxcoff+</code><dd><a name="index-gxcoff_002b-509"></a>Produce debugging information in XCOFF format (if that is supported), 125using GNU extensions understood only by the GNU debugger (GDB). The 126use of these extensions is likely to make other debuggers crash or 127refuse to read the program, and may cause assemblers other than the GNU 128assembler (GAS) to fail with an error. 129 130 <br><dt><code>-gdwarf-</code><var>version</var><dd><a name="index-gdwarf_002d_0040var_007bversion_007d-510"></a>Produce debugging information in DWARF format (if that is 131supported). This is the format used by DBX on IRIX 6. The value 132of <var>version</var> may be either 2, 3 or 4; the default version is 2. 133 134 <p>Note that with DWARF version 2 some ports require, and will always 135use, some non-conflicting DWARF 3 extensions in the unwind tables. 136 137 <p>Version 4 may require GDB 7.0 and <samp><span class="option">-fvar-tracking-assignments</span></samp> 138for maximum benefit. 139 140 <br><dt><code>-gstrict-dwarf</code><dd><a name="index-gstrict_002ddwarf-511"></a>Disallow using extensions of later DWARF standard version than selected 141with <samp><span class="option">-gdwarf-</span><var>version</var></samp>. On most targets using non-conflicting 142DWARF extensions from later standard versions is allowed. 143 144 <br><dt><code>-gno-strict-dwarf</code><dd><a name="index-gno_002dstrict_002ddwarf-512"></a>Allow using extensions of later DWARF standard version than selected with 145<samp><span class="option">-gdwarf-</span><var>version</var></samp>. 146 147 <br><dt><code>-gvms</code><dd><a name="index-gvms-513"></a>Produce debugging information in VMS debug format (if that is 148supported). This is the format used by DEBUG on VMS systems. 149 150 <br><dt><code>-g</code><var>level</var><dt><code>-ggdb</code><var>level</var><dt><code>-gstabs</code><var>level</var><dt><code>-gcoff</code><var>level</var><dt><code>-gxcoff</code><var>level</var><dt><code>-gvms</code><var>level</var><dd>Request debugging information and also use <var>level</var> to specify how 151much information. The default level is 2. 152 153 <p>Level 0 produces no debug information at all. Thus, <samp><span class="option">-g0</span></samp> negates 154<samp><span class="option">-g</span></samp>. 155 156 <p>Level 1 produces minimal information, enough for making backtraces in 157parts of the program that you don't plan to debug. This includes 158descriptions of functions and external variables, but no information 159about local variables and no line numbers. 160 161 <p>Level 3 includes extra information, such as all the macro definitions 162present in the program. Some debuggers support macro expansion when 163you use <samp><span class="option">-g3</span></samp>. 164 165 <p><samp><span class="option">-gdwarf-2</span></samp> does not accept a concatenated debug level, because 166GCC used to support an option <samp><span class="option">-gdwarf</span></samp> that meant to generate 167debug information in version 1 of the DWARF format (which is very 168different from version 2), and it would have been too confusing. That 169debug format is long obsolete, but the option cannot be changed now. 170Instead use an additional <samp><span class="option">-g</span><var>level</var></samp> option to change the 171debug level for DWARF. 172 173 <br><dt><code>-gtoggle</code><dd><a name="index-gtoggle-514"></a>Turn off generation of debug info, if leaving out this option would have 174generated it, or turn it on at level 2 otherwise. The position of this 175argument in the command line does not matter, it takes effect after all 176other options are processed, and it does so only once, no matter how 177many times it is given. This is mainly intended to be used with 178<samp><span class="option">-fcompare-debug</span></samp>. 179 180 <br><dt><code>-fdump-final-insns</code><span class="roman">[</span><code>=</code><var>file</var><span class="roman">]</span><dd><a name="index-fdump_002dfinal_002dinsns-515"></a>Dump the final internal representation (RTL) to <var>file</var>. If the 181optional argument is omitted (or if <var>file</var> is <code>.</code>), the name 182of the dump file will be determined by appending <code>.gkd</code> to the 183compilation output file name. 184 185 <br><dt><code>-fcompare-debug</code><span class="roman">[</span><code>=</code><var>opts</var><span class="roman">]</span><dd><a name="index-fcompare_002ddebug-516"></a><a name="index-fno_002dcompare_002ddebug-517"></a>If no error occurs during compilation, run the compiler a second time, 186adding <var>opts</var> and <samp><span class="option">-fcompare-debug-second</span></samp> to the arguments 187passed to the second compilation. Dump the final internal 188representation in both compilations, and print an error if they differ. 189 190 <p>If the equal sign is omitted, the default <samp><span class="option">-gtoggle</span></samp> is used. 191 192 <p>The environment variable <samp><span class="env">GCC_COMPARE_DEBUG</span></samp>, if defined, non-empty 193and nonzero, implicitly enables <samp><span class="option">-fcompare-debug</span></samp>. If 194<samp><span class="env">GCC_COMPARE_DEBUG</span></samp> is defined to a string starting with a dash, 195then it is used for <var>opts</var>, otherwise the default <samp><span class="option">-gtoggle</span></samp> 196is used. 197 198 <p><samp><span class="option">-fcompare-debug=</span></samp>, with the equal sign but without <var>opts</var>, 199is equivalent to <samp><span class="option">-fno-compare-debug</span></samp>, which disables the dumping 200of the final representation and the second compilation, preventing even 201<samp><span class="env">GCC_COMPARE_DEBUG</span></samp> from taking effect. 202 203 <p>To verify full coverage during <samp><span class="option">-fcompare-debug</span></samp> testing, set 204<samp><span class="env">GCC_COMPARE_DEBUG</span></samp> to say ‘<samp><span class="samp">-fcompare-debug-not-overridden</span></samp>’, 205which GCC will reject as an invalid option in any actual compilation 206(rather than preprocessing, assembly or linking). To get just a 207warning, setting <samp><span class="env">GCC_COMPARE_DEBUG</span></samp> to ‘<samp><span class="samp">-w%n-fcompare-debug 208not overridden</span></samp>’ will do. 209 210 <br><dt><code>-fcompare-debug-second</code><dd><a name="index-fcompare_002ddebug_002dsecond-518"></a>This option is implicitly passed to the compiler for the second 211compilation requested by <samp><span class="option">-fcompare-debug</span></samp>, along with options to 212silence warnings, and omitting other options that would cause 213side-effect compiler outputs to files or to the standard output. Dump 214files and preserved temporary files are renamed so as to contain the 215<code>.gk</code> additional extension during the second compilation, to avoid 216overwriting those generated by the first. 217 218 <p>When this option is passed to the compiler driver, it causes the 219<em>first</em> compilation to be skipped, which makes it useful for little 220other than debugging the compiler proper. 221 222 <br><dt><code>-feliminate-dwarf2-dups</code><dd><a name="index-feliminate_002ddwarf2_002ddups-519"></a>Compress DWARF2 debugging information by eliminating duplicated 223information about each symbol. This option only makes sense when 224generating DWARF2 debugging information with <samp><span class="option">-gdwarf-2</span></samp>. 225 226 <br><dt><code>-femit-struct-debug-baseonly</code><dd>Emit debug information for struct-like types 227only when the base name of the compilation source file 228matches the base name of file in which the struct was defined. 229 230 <p>This option substantially reduces the size of debugging information, 231but at significant potential loss in type information to the debugger. 232See <samp><span class="option">-femit-struct-debug-reduced</span></samp> for a less aggressive option. 233See <samp><span class="option">-femit-struct-debug-detailed</span></samp> for more detailed control. 234 235 <p>This option works only with DWARF 2. 236 237 <br><dt><code>-femit-struct-debug-reduced</code><dd>Emit debug information for struct-like types 238only when the base name of the compilation source file 239matches the base name of file in which the type was defined, 240unless the struct is a template or defined in a system header. 241 242 <p>This option significantly reduces the size of debugging information, 243with some potential loss in type information to the debugger. 244See <samp><span class="option">-femit-struct-debug-baseonly</span></samp> for a more aggressive option. 245See <samp><span class="option">-femit-struct-debug-detailed</span></samp> for more detailed control. 246 247 <p>This option works only with DWARF 2. 248 249 <br><dt><code>-femit-struct-debug-detailed</code><span class="roman">[</span><code>=</code><var>spec-list</var><span class="roman">]</span><dd>Specify the struct-like types 250for which the compiler will generate debug information. 251The intent is to reduce duplicate struct debug information 252between different object files within the same program. 253 254 <p>This option is a detailed version of 255<samp><span class="option">-femit-struct-debug-reduced</span></samp> and <samp><span class="option">-femit-struct-debug-baseonly</span></samp>, 256which will serve for most needs. 257 258 <p>A specification has the syntax<br> 259[‘<samp><span class="samp">dir:</span></samp>’|‘<samp><span class="samp">ind:</span></samp>’][‘<samp><span class="samp">ord:</span></samp>’|‘<samp><span class="samp">gen:</span></samp>’](‘<samp><span class="samp">any</span></samp>’|‘<samp><span class="samp">sys</span></samp>’|‘<samp><span class="samp">base</span></samp>’|‘<samp><span class="samp">none</span></samp>’) 260 261 <p>The optional first word limits the specification to 262structs that are used directly (‘<samp><span class="samp">dir:</span></samp>’) or used indirectly (‘<samp><span class="samp">ind:</span></samp>’). 263A struct type is used directly when it is the type of a variable, member. 264Indirect uses arise through pointers to structs. 265That is, when use of an incomplete struct would be legal, the use is indirect. 266An example is 267‘<samp><span class="samp">struct one direct; struct two * indirect;</span></samp>’. 268 269 <p>The optional second word limits the specification to 270ordinary structs (‘<samp><span class="samp">ord:</span></samp>’) or generic structs (‘<samp><span class="samp">gen:</span></samp>’). 271Generic structs are a bit complicated to explain. 272For C++, these are non-explicit specializations of template classes, 273or non-template classes within the above. 274Other programming languages have generics, 275but ‘<samp><span class="samp">-femit-struct-debug-detailed</span></samp>’ does not yet implement them. 276 277 <p>The third word specifies the source files for those 278structs for which the compiler will emit debug information. 279The values ‘<samp><span class="samp">none</span></samp>’ and ‘<samp><span class="samp">any</span></samp>’ have the normal meaning. 280The value ‘<samp><span class="samp">base</span></samp>’ means that 281the base of name of the file in which the type declaration appears 282must match the base of the name of the main compilation file. 283In practice, this means that 284types declared in <samp><span class="file">foo.c</span></samp> and <samp><span class="file">foo.h</span></samp> will have debug information, 285but types declared in other header will not. 286The value ‘<samp><span class="samp">sys</span></samp>’ means those types satisfying ‘<samp><span class="samp">base</span></samp>’ 287or declared in system or compiler headers. 288 289 <p>You may need to experiment to determine the best settings for your application. 290 291 <p>The default is ‘<samp><span class="samp">-femit-struct-debug-detailed=all</span></samp>’. 292 293 <p>This option works only with DWARF 2. 294 295 <br><dt><code>-fenable-icf-debug</code><dd><a name="index-fenable_002dicf_002ddebug-520"></a>Generate additional debug information to support identical code folding (ICF). 296This option only works with DWARF version 2 or higher. 297 298 <br><dt><code>-fno-merge-debug-strings</code><dd><a name="index-fmerge_002ddebug_002dstrings-521"></a><a name="index-fno_002dmerge_002ddebug_002dstrings-522"></a>Direct the linker to not merge together strings in the debugging 299information which are identical in different object files. Merging is 300not supported by all assemblers or linkers. Merging decreases the size 301of the debug information in the output file at the cost of increasing 302link processing time. Merging is enabled by default. 303 304 <br><dt><code>-fdebug-prefix-map=</code><var>old</var><code>=</code><var>new</var><dd><a name="index-fdebug_002dprefix_002dmap-523"></a>When compiling files in directory <samp><var>old</var></samp>, record debugging 305information describing them as in <samp><var>new</var></samp> instead. 306 307 <br><dt><code>-fno-dwarf2-cfi-asm</code><dd><a name="index-fdwarf2_002dcfi_002dasm-524"></a><a name="index-fno_002ddwarf2_002dcfi_002dasm-525"></a>Emit DWARF 2 unwind info as compiler generated <code>.eh_frame</code> section 308instead of using GAS <code>.cfi_*</code> directives. 309 310 <p><a name="index-g_t_0040command_007bprof_007d-526"></a><br><dt><code>-p</code><dd><a name="index-p-527"></a>Generate extra code to write profile information suitable for the 311analysis program <samp><span class="command">prof</span></samp>. You must use this option when compiling 312the source files you want data about, and you must also use it when 313linking. 314 315 <p><a name="index-g_t_0040command_007bgprof_007d-528"></a><br><dt><code>-pg</code><dd><a name="index-pg-529"></a>Generate extra code to write profile information suitable for the 316analysis program <samp><span class="command">gprof</span></samp>. You must use this option when compiling 317the source files you want data about, and you must also use it when 318linking. 319 320 <br><dt><code>-Q</code><dd><a name="index-Q-530"></a>Makes the compiler print out each function name as it is compiled, and 321print some statistics about each pass when it finishes. 322 323 <br><dt><code>-ftime-report</code><dd><a name="index-ftime_002dreport-531"></a>Makes the compiler print some statistics about the time consumed by each 324pass when it finishes. 325 326 <br><dt><code>-fmem-report</code><dd><a name="index-fmem_002dreport-532"></a>Makes the compiler print some statistics about permanent memory 327allocation when it finishes. 328 329 <br><dt><code>-fpre-ipa-mem-report</code><dd><a name="index-fpre_002dipa_002dmem_002dreport-533"></a><br><dt><code>-fpost-ipa-mem-report</code><dd><a name="index-fpost_002dipa_002dmem_002dreport-534"></a>Makes the compiler print some statistics about permanent memory 330allocation before or after interprocedural optimization. 331 332 <br><dt><code>-fstack-usage</code><dd><a name="index-fstack_002dusage-535"></a>Makes the compiler output stack usage information for the program, on a 333per-function basis. The filename for the dump is made by appending 334<samp><span class="file">.su</span></samp> to the <var>auxname</var>. <var>auxname</var> is generated from the name of 335the output file, if explicitly specified and it is not an executable, 336otherwise it is the basename of the source file. An entry is made up 337of three fields: 338 339 <ul> 340<li>The name of the function. 341<li>A number of bytes. 342<li>One or more qualifiers: <code>static</code>, <code>dynamic</code>, <code>bounded</code>. 343</ul> 344 345 <p>The qualifier <code>static</code> means that the function manipulates the stack 346statically: a fixed number of bytes are allocated for the frame on function 347entry and released on function exit; no stack adjustments are otherwise made 348in the function. The second field is this fixed number of bytes. 349 350 <p>The qualifier <code>dynamic</code> means that the function manipulates the stack 351dynamically: in addition to the static allocation described above, stack 352adjustments are made in the body of the function, for example to push/pop 353arguments around function calls. If the qualifier <code>bounded</code> is also 354present, the amount of these adjustments is bounded at compile-time and 355the second field is an upper bound of the total amount of stack used by 356the function. If it is not present, the amount of these adjustments is 357not bounded at compile-time and the second field only represents the 358bounded part. 359 360 <br><dt><code>-fprofile-arcs</code><dd><a name="index-fprofile_002darcs-536"></a>Add code so that program flow <dfn>arcs</dfn> are instrumented. During 361execution the program records how many times each branch and call is 362executed and how many times it is taken or returns. When the compiled 363program exits it saves this data to a file called 364<samp><var>auxname</var><span class="file">.gcda</span></samp> for each source file. The data may be used for 365profile-directed optimizations (<samp><span class="option">-fbranch-probabilities</span></samp>), or for 366test coverage analysis (<samp><span class="option">-ftest-coverage</span></samp>). Each object file's 367<var>auxname</var> is generated from the name of the output file, if 368explicitly specified and it is not the final executable, otherwise it is 369the basename of the source file. In both cases any suffix is removed 370(e.g. <samp><span class="file">foo.gcda</span></samp> for input file <samp><span class="file">dir/foo.c</span></samp>, or 371<samp><span class="file">dir/foo.gcda</span></samp> for output file specified as <samp><span class="option">-o dir/foo.o</span></samp>). 372See <a href="Cross_002dprofiling.html#Cross_002dprofiling">Cross-profiling</a>. 373 374 <p><a name="index-g_t_0040command_007bgcov_007d-537"></a><br><dt><code>--coverage</code><dd><a name="index-coverage-538"></a> 375This option is used to compile and link code instrumented for coverage 376analysis. The option is a synonym for <samp><span class="option">-fprofile-arcs</span></samp> 377<samp><span class="option">-ftest-coverage</span></samp> (when compiling) and <samp><span class="option">-lgcov</span></samp> (when 378linking). See the documentation for those options for more details. 379 380 <ul> 381<li>Compile the source files with <samp><span class="option">-fprofile-arcs</span></samp> plus optimization 382and code generation options. For test coverage analysis, use the 383additional <samp><span class="option">-ftest-coverage</span></samp> option. You do not need to profile 384every source file in a program. 385 386 <li>Link your object files with <samp><span class="option">-lgcov</span></samp> or <samp><span class="option">-fprofile-arcs</span></samp> 387(the latter implies the former). 388 389 <li>Run the program on a representative workload to generate the arc profile 390information. This may be repeated any number of times. You can run 391concurrent instances of your program, and provided that the file system 392supports locking, the data files will be correctly updated. Also 393<code>fork</code> calls are detected and correctly handled (double counting 394will not happen). 395 396 <li>For profile-directed optimizations, compile the source files again with 397the same optimization and code generation options plus 398<samp><span class="option">-fbranch-probabilities</span></samp> (see <a href="Optimize-Options.html#Optimize-Options">Options that Control Optimization</a>). 399 400 <li>For test coverage analysis, use <samp><span class="command">gcov</span></samp> to produce human readable 401information from the <samp><span class="file">.gcno</span></samp> and <samp><span class="file">.gcda</span></samp> files. Refer to the 402<samp><span class="command">gcov</span></samp> documentation for further information. 403 404 </ul> 405 406 <p>With <samp><span class="option">-fprofile-arcs</span></samp>, for each function of your program GCC 407creates a program flow graph, then finds a spanning tree for the graph. 408Only arcs that are not on the spanning tree have to be instrumented: the 409compiler adds code to count the number of times that these arcs are 410executed. When an arc is the only exit or only entrance to a block, the 411instrumentation code can be added to the block; otherwise, a new basic 412block must be created to hold the instrumentation code. 413 414 <br><dt><code>-ftest-coverage</code><dd><a name="index-ftest_002dcoverage-539"></a>Produce a notes file that the <samp><span class="command">gcov</span></samp> code-coverage utility 415(see <a href="Gcov.html#Gcov"><samp><span class="command">gcov</span></samp>—a Test Coverage Program</a>) can use to 416show program coverage. Each source file's note file is called 417<samp><var>auxname</var><span class="file">.gcno</span></samp>. Refer to the <samp><span class="option">-fprofile-arcs</span></samp> option 418above for a description of <var>auxname</var> and instructions on how to 419generate test coverage data. Coverage data will match the source files 420more closely, if you do not optimize. 421 422 <br><dt><code>-fdbg-cnt-list</code><dd><a name="index-fdbg_002dcnt_002dlist-540"></a>Print the name and the counter upper bound for all debug counters. 423 424 <br><dt><code>-fdbg-cnt=</code><var>counter-value-list</var><dd><a name="index-fdbg_002dcnt-541"></a>Set the internal debug counter upper bound. <var>counter-value-list</var> 425is a comma-separated list of <var>name</var>:<var>value</var> pairs 426which sets the upper bound of each debug counter <var>name</var> to <var>value</var>. 427All debug counters have the initial upper bound of <var>UINT_MAX</var>, 428thus dbg_cnt() returns true always unless the upper bound is set by this option. 429e.g. With -fdbg-cnt=dce:10,tail_call:0 430dbg_cnt(dce) will return true only for first 10 invocations 431and dbg_cnt(tail_call) will return false always. 432 433 <br><dt><code>-d</code><var>letters</var><dt><code>-fdump-rtl-</code><var>pass</var><dd><a name="index-d-542"></a>Says to make debugging dumps during compilation at times specified by 434<var>letters</var>. This is used for debugging the RTL-based passes of the 435compiler. The file names for most of the dumps are made by appending 436a pass number and a word to the <var>dumpname</var>, and the files are 437created in the directory of the output file. Note that the pass 438number is computed statically as passes get registered into the pass 439manager. Thus the numbering is not related to the dynamic order of 440execution of passes. In particular, a pass installed by a plugin 441could have a number over 200 even if it executed quite early. 442<var>dumpname</var> is generated from the name of the output file, if 443explicitly specified and it is not an executable, otherwise it is the 444basename of the source file. These switches may have different effects 445when <samp><span class="option">-E</span></samp> is used for preprocessing. 446 447 <p>Debug dumps can be enabled with a <samp><span class="option">-fdump-rtl</span></samp> switch or some 448<samp><span class="option">-d</span></samp> option <var>letters</var>. Here are the possible 449letters for use in <var>pass</var> and <var>letters</var>, and their meanings: 450 451 <dl> 452<dt><code>-fdump-rtl-alignments</code><dd><a name="index-fdump_002drtl_002dalignments-543"></a>Dump after branch alignments have been computed. 453 454 <br><dt><code>-fdump-rtl-asmcons</code><dd><a name="index-fdump_002drtl_002dasmcons-544"></a>Dump after fixing rtl statements that have unsatisfied in/out constraints. 455 456 <br><dt><code>-fdump-rtl-auto_inc_dec</code><dd><a name="index-fdump_002drtl_002dauto_005finc_005fdec-545"></a>Dump after auto-inc-dec discovery. This pass is only run on 457architectures that have auto inc or auto dec instructions. 458 459 <br><dt><code>-fdump-rtl-barriers</code><dd><a name="index-fdump_002drtl_002dbarriers-546"></a>Dump after cleaning up the barrier instructions. 460 461 <br><dt><code>-fdump-rtl-bbpart</code><dd><a name="index-fdump_002drtl_002dbbpart-547"></a>Dump after partitioning hot and cold basic blocks. 462 463 <br><dt><code>-fdump-rtl-bbro</code><dd><a name="index-fdump_002drtl_002dbbro-548"></a>Dump after block reordering. 464 465 <br><dt><code>-fdump-rtl-btl1</code><dt><code>-fdump-rtl-btl2</code><dd><a name="index-fdump_002drtl_002dbtl2-549"></a><a name="index-fdump_002drtl_002dbtl2-550"></a><samp><span class="option">-fdump-rtl-btl1</span></samp> and <samp><span class="option">-fdump-rtl-btl2</span></samp> enable dumping 466after the two branch 467target load optimization passes. 468 469 <br><dt><code>-fdump-rtl-bypass</code><dd><a name="index-fdump_002drtl_002dbypass-551"></a>Dump after jump bypassing and control flow optimizations. 470 471 <br><dt><code>-fdump-rtl-combine</code><dd><a name="index-fdump_002drtl_002dcombine-552"></a>Dump after the RTL instruction combination pass. 472 473 <br><dt><code>-fdump-rtl-compgotos</code><dd><a name="index-fdump_002drtl_002dcompgotos-553"></a>Dump after duplicating the computed gotos. 474 475 <br><dt><code>-fdump-rtl-ce1</code><dt><code>-fdump-rtl-ce2</code><dt><code>-fdump-rtl-ce3</code><dd><a name="index-fdump_002drtl_002dce1-554"></a><a name="index-fdump_002drtl_002dce2-555"></a><a name="index-fdump_002drtl_002dce3-556"></a><samp><span class="option">-fdump-rtl-ce1</span></samp>, <samp><span class="option">-fdump-rtl-ce2</span></samp>, and 476<samp><span class="option">-fdump-rtl-ce3</span></samp> enable dumping after the three 477if conversion passes. 478 479 <dt><code>-fdump-rtl-cprop_hardreg</code><dd><a name="index-fdump_002drtl_002dcprop_005fhardreg-557"></a>Dump after hard register copy propagation. 480 481 <dt><code>-fdump-rtl-csa</code><dd><a name="index-fdump_002drtl_002dcsa-558"></a>Dump after combining stack adjustments. 482 483 <br><dt><code>-fdump-rtl-cse1</code><dt><code>-fdump-rtl-cse2</code><dd><a name="index-fdump_002drtl_002dcse1-559"></a><a name="index-fdump_002drtl_002dcse2-560"></a><samp><span class="option">-fdump-rtl-cse1</span></samp> and <samp><span class="option">-fdump-rtl-cse2</span></samp> enable dumping after 484the two common sub-expression elimination passes. 485 486 <dt><code>-fdump-rtl-dce</code><dd><a name="index-fdump_002drtl_002ddce-561"></a>Dump after the standalone dead code elimination passes. 487 488 <dt><code>-fdump-rtl-dbr</code><dd><a name="index-fdump_002drtl_002ddbr-562"></a>Dump after delayed branch scheduling. 489 490 <br><dt><code>-fdump-rtl-dce1</code><dt><code>-fdump-rtl-dce2</code><dd><a name="index-fdump_002drtl_002ddce1-563"></a><a name="index-fdump_002drtl_002ddce2-564"></a><samp><span class="option">-fdump-rtl-dce1</span></samp> and <samp><span class="option">-fdump-rtl-dce2</span></samp> enable dumping after 491the two dead store elimination passes. 492 493 <br><dt><code>-fdump-rtl-eh</code><dd><a name="index-fdump_002drtl_002deh-565"></a>Dump after finalization of EH handling code. 494 495 <br><dt><code>-fdump-rtl-eh_ranges</code><dd><a name="index-fdump_002drtl_002deh_005franges-566"></a>Dump after conversion of EH handling range regions. 496 497 <br><dt><code>-fdump-rtl-expand</code><dd><a name="index-fdump_002drtl_002dexpand-567"></a>Dump after RTL generation. 498 499 <br><dt><code>-fdump-rtl-fwprop1</code><dt><code>-fdump-rtl-fwprop2</code><dd><a name="index-fdump_002drtl_002dfwprop1-568"></a><a name="index-fdump_002drtl_002dfwprop2-569"></a><samp><span class="option">-fdump-rtl-fwprop1</span></samp> and <samp><span class="option">-fdump-rtl-fwprop2</span></samp> enable 500dumping after the two forward propagation passes. 501 502 <br><dt><code>-fdump-rtl-gcse1</code><dt><code>-fdump-rtl-gcse2</code><dd><a name="index-fdump_002drtl_002dgcse1-570"></a><a name="index-fdump_002drtl_002dgcse2-571"></a><samp><span class="option">-fdump-rtl-gcse1</span></samp> and <samp><span class="option">-fdump-rtl-gcse2</span></samp> enable dumping 503after global common subexpression elimination. 504 505 <br><dt><code>-fdump-rtl-init-regs</code><dd><a name="index-fdump_002drtl_002dinit_002dregs-572"></a>Dump after the initialization of the registers. 506 507 <br><dt><code>-fdump-rtl-initvals</code><dd><a name="index-fdump_002drtl_002dinitvals-573"></a>Dump after the computation of the initial value sets. 508 509 <dt><code>-fdump-rtl-into_cfglayout</code><dd><a name="index-fdump_002drtl_002dinto_005fcfglayout-574"></a>Dump after converting to cfglayout mode. 510 511 <br><dt><code>-fdump-rtl-ira</code><dd><a name="index-fdump_002drtl_002dira-575"></a>Dump after iterated register allocation. 512 513 <br><dt><code>-fdump-rtl-jump</code><dd><a name="index-fdump_002drtl_002djump-576"></a>Dump after the second jump optimization. 514 515 <br><dt><code>-fdump-rtl-loop2</code><dd><a name="index-fdump_002drtl_002dloop2-577"></a><samp><span class="option">-fdump-rtl-loop2</span></samp> enables dumping after the rtl 516loop optimization passes. 517 518 <br><dt><code>-fdump-rtl-mach</code><dd><a name="index-fdump_002drtl_002dmach-578"></a>Dump after performing the machine dependent reorganization pass, if that 519pass exists. 520 521 <br><dt><code>-fdump-rtl-mode_sw</code><dd><a name="index-fdump_002drtl_002dmode_005fsw-579"></a>Dump after removing redundant mode switches. 522 523 <br><dt><code>-fdump-rtl-rnreg</code><dd><a name="index-fdump_002drtl_002drnreg-580"></a>Dump after register renumbering. 524 525 <dt><code>-fdump-rtl-outof_cfglayout</code><dd><a name="index-fdump_002drtl_002doutof_005fcfglayout-581"></a>Dump after converting from cfglayout mode. 526 527 <br><dt><code>-fdump-rtl-peephole2</code><dd><a name="index-fdump_002drtl_002dpeephole2-582"></a>Dump after the peephole pass. 528 529 <br><dt><code>-fdump-rtl-postreload</code><dd><a name="index-fdump_002drtl_002dpostreload-583"></a>Dump after post-reload optimizations. 530 531 <dt><code>-fdump-rtl-pro_and_epilogue</code><dd><a name="index-fdump_002drtl_002dpro_005fand_005fepilogue-584"></a>Dump after generating the function pro and epilogues. 532 533 <br><dt><code>-fdump-rtl-regmove</code><dd><a name="index-fdump_002drtl_002dregmove-585"></a>Dump after the register move pass. 534 535 <br><dt><code>-fdump-rtl-sched1</code><dt><code>-fdump-rtl-sched2</code><dd><a name="index-fdump_002drtl_002dsched1-586"></a><a name="index-fdump_002drtl_002dsched2-587"></a><samp><span class="option">-fdump-rtl-sched1</span></samp> and <samp><span class="option">-fdump-rtl-sched2</span></samp> enable dumping 536after the basic block scheduling passes. 537 538 <br><dt><code>-fdump-rtl-see</code><dd><a name="index-fdump_002drtl_002dsee-588"></a>Dump after sign extension elimination. 539 540 <br><dt><code>-fdump-rtl-seqabstr</code><dd><a name="index-fdump_002drtl_002dseqabstr-589"></a>Dump after common sequence discovery. 541 542 <br><dt><code>-fdump-rtl-shorten</code><dd><a name="index-fdump_002drtl_002dshorten-590"></a>Dump after shortening branches. 543 544 <br><dt><code>-fdump-rtl-sibling</code><dd><a name="index-fdump_002drtl_002dsibling-591"></a>Dump after sibling call optimizations. 545 546 <br><dt><code>-fdump-rtl-split1</code><dt><code>-fdump-rtl-split2</code><dt><code>-fdump-rtl-split3</code><dt><code>-fdump-rtl-split4</code><dt><code>-fdump-rtl-split5</code><dd><a name="index-fdump_002drtl_002dsplit1-592"></a><a name="index-fdump_002drtl_002dsplit2-593"></a><a name="index-fdump_002drtl_002dsplit3-594"></a><a name="index-fdump_002drtl_002dsplit4-595"></a><a name="index-fdump_002drtl_002dsplit5-596"></a><samp><span class="option">-fdump-rtl-split1</span></samp>, <samp><span class="option">-fdump-rtl-split2</span></samp>, 547<samp><span class="option">-fdump-rtl-split3</span></samp>, <samp><span class="option">-fdump-rtl-split4</span></samp> and 548<samp><span class="option">-fdump-rtl-split5</span></samp> enable dumping after five rounds of 549instruction splitting. 550 551 <br><dt><code>-fdump-rtl-sms</code><dd><a name="index-fdump_002drtl_002dsms-597"></a>Dump after modulo scheduling. This pass is only run on some 552architectures. 553 554 <br><dt><code>-fdump-rtl-stack</code><dd><a name="index-fdump_002drtl_002dstack-598"></a>Dump after conversion from GCC's "flat register file" registers to the 555x87's stack-like registers. This pass is only run on x86 variants. 556 557 <br><dt><code>-fdump-rtl-subreg1</code><dt><code>-fdump-rtl-subreg2</code><dd><a name="index-fdump_002drtl_002dsubreg1-599"></a><a name="index-fdump_002drtl_002dsubreg2-600"></a><samp><span class="option">-fdump-rtl-subreg1</span></samp> and <samp><span class="option">-fdump-rtl-subreg2</span></samp> enable dumping after 558the two subreg expansion passes. 559 560 <br><dt><code>-fdump-rtl-unshare</code><dd><a name="index-fdump_002drtl_002dunshare-601"></a>Dump after all rtl has been unshared. 561 562 <br><dt><code>-fdump-rtl-vartrack</code><dd><a name="index-fdump_002drtl_002dvartrack-602"></a>Dump after variable tracking. 563 564 <br><dt><code>-fdump-rtl-vregs</code><dd><a name="index-fdump_002drtl_002dvregs-603"></a>Dump after converting virtual registers to hard registers. 565 566 <br><dt><code>-fdump-rtl-web</code><dd><a name="index-fdump_002drtl_002dweb-604"></a>Dump after live range splitting. 567 568 <br><dt><code>-fdump-rtl-regclass</code><dt><code>-fdump-rtl-subregs_of_mode_init</code><dt><code>-fdump-rtl-subregs_of_mode_finish</code><dt><code>-fdump-rtl-dfinit</code><dt><code>-fdump-rtl-dfinish</code><dd><a name="index-fdump_002drtl_002dregclass-605"></a><a name="index-fdump_002drtl_002dsubregs_005fof_005fmode_005finit-606"></a><a name="index-fdump_002drtl_002dsubregs_005fof_005fmode_005ffinish-607"></a><a name="index-fdump_002drtl_002ddfinit-608"></a><a name="index-fdump_002drtl_002ddfinish-609"></a>These dumps are defined but always produce empty files. 569 570 <br><dt><code>-fdump-rtl-all</code><dd><a name="index-fdump_002drtl_002dall-610"></a>Produce all the dumps listed above. 571 572 <br><dt><code>-dA</code><dd><a name="index-dA-611"></a>Annotate the assembler output with miscellaneous debugging information. 573 574 <br><dt><code>-dD</code><dd><a name="index-dD-612"></a>Dump all macro definitions, at the end of preprocessing, in addition to 575normal output. 576 577 <br><dt><code>-dH</code><dd><a name="index-dH-613"></a>Produce a core dump whenever an error occurs. 578 579 <br><dt><code>-dm</code><dd><a name="index-dm-614"></a>Print statistics on memory usage, at the end of the run, to 580standard error. 581 582 <br><dt><code>-dp</code><dd><a name="index-dp-615"></a>Annotate the assembler output with a comment indicating which 583pattern and alternative was used. The length of each instruction is 584also printed. 585 586 <br><dt><code>-dP</code><dd><a name="index-dP-616"></a>Dump the RTL in the assembler output as a comment before each instruction. 587Also turns on <samp><span class="option">-dp</span></samp> annotation. 588 589 <br><dt><code>-dv</code><dd><a name="index-dv-617"></a>For each of the other indicated dump files (<samp><span class="option">-fdump-rtl-</span><var>pass</var></samp>), 590dump a representation of the control flow graph suitable for viewing with VCG 591to <samp><var>file</var><span class="file">.</span><var>pass</var><span class="file">.vcg</span></samp>. 592 593 <br><dt><code>-dx</code><dd><a name="index-dx-618"></a>Just generate RTL for a function instead of compiling it. Usually used 594with <samp><span class="option">-fdump-rtl-expand</span></samp>. 595</dl> 596 597 <br><dt><code>-fdump-noaddr</code><dd><a name="index-fdump_002dnoaddr-619"></a>When doing debugging dumps, suppress address output. This makes it more 598feasible to use diff on debugging dumps for compiler invocations with 599different compiler binaries and/or different 600text / bss / data / heap / stack / dso start locations. 601 602 <br><dt><code>-fdump-unnumbered</code><dd><a name="index-fdump_002dunnumbered-620"></a>When doing debugging dumps, suppress instruction numbers and address output. 603This makes it more feasible to use diff on debugging dumps for compiler 604invocations with different options, in particular with and without 605<samp><span class="option">-g</span></samp>. 606 607 <br><dt><code>-fdump-unnumbered-links</code><dd><a name="index-fdump_002dunnumbered_002dlinks-621"></a>When doing debugging dumps (see <samp><span class="option">-d</span></samp> option above), suppress 608instruction numbers for the links to the previous and next instructions 609in a sequence. 610 611 <br><dt><code>-fdump-translation-unit </code><span class="roman">(C++ only)</span><dt><code>-fdump-translation-unit-</code><var>options</var> <span class="roman">(C++ only)</span><dd><a name="index-fdump_002dtranslation_002dunit-622"></a>Dump a representation of the tree structure for the entire translation 612unit to a file. The file name is made by appending <samp><span class="file">.tu</span></samp> to the 613source file name, and the file is created in the same directory as the 614output file. If the ‘<samp><span class="samp">-</span><var>options</var></samp>’ form is used, <var>options</var> 615controls the details of the dump as described for the 616<samp><span class="option">-fdump-tree</span></samp> options. 617 618 <br><dt><code>-fdump-class-hierarchy </code><span class="roman">(C++ only)</span><dt><code>-fdump-class-hierarchy-</code><var>options</var> <span class="roman">(C++ only)</span><dd><a name="index-fdump_002dclass_002dhierarchy-623"></a>Dump a representation of each class's hierarchy and virtual function 619table layout to a file. The file name is made by appending 620<samp><span class="file">.class</span></samp> to the source file name, and the file is created in the 621same directory as the output file. If the ‘<samp><span class="samp">-</span><var>options</var></samp>’ form 622is used, <var>options</var> controls the details of the dump as described 623for the <samp><span class="option">-fdump-tree</span></samp> options. 624 625 <br><dt><code>-fdump-ipa-</code><var>switch</var><dd><a name="index-fdump_002dipa-624"></a>Control the dumping at various stages of inter-procedural analysis 626language tree to a file. The file name is generated by appending a 627switch specific suffix to the source file name, and the file is created 628in the same directory as the output file. The following dumps are 629possible: 630 631 <dl> 632<dt>‘<samp><span class="samp">all</span></samp>’<dd>Enables all inter-procedural analysis dumps. 633 634 <br><dt>‘<samp><span class="samp">cgraph</span></samp>’<dd>Dumps information about call-graph optimization, unused function removal, 635and inlining decisions. 636 637 <br><dt>‘<samp><span class="samp">inline</span></samp>’<dd>Dump after function inlining. 638 639 </dl> 640 641 <br><dt><code>-fdump-statistics-</code><var>option</var><dd><a name="index-fdump_002dstatistics-625"></a>Enable and control dumping of pass statistics in a separate file. The 642file name is generated by appending a suffix ending in 643‘<samp><span class="samp">.statistics</span></samp>’ to the source file name, and the file is created in 644the same directory as the output file. If the ‘<samp><span class="samp">-</span><var>option</var></samp>’ 645form is used, ‘<samp><span class="samp">-stats</span></samp>’ will cause counters to be summed over the 646whole compilation unit while ‘<samp><span class="samp">-details</span></samp>’ will dump every event as 647the passes generate them. The default with no option is to sum 648counters for each function compiled. 649 650 <br><dt><code>-fdump-tree-</code><var>switch</var><dt><code>-fdump-tree-</code><var>switch</var><code>-</code><var>options</var><dd><a name="index-fdump_002dtree-626"></a>Control the dumping at various stages of processing the intermediate 651language tree to a file. The file name is generated by appending a 652switch specific suffix to the source file name, and the file is 653created in the same directory as the output file. If the 654‘<samp><span class="samp">-</span><var>options</var></samp>’ form is used, <var>options</var> is a list of 655‘<samp><span class="samp">-</span></samp>’ separated options that control the details of the dump. Not 656all options are applicable to all dumps, those which are not 657meaningful will be ignored. The following options are available 658 659 <dl> 660<dt>‘<samp><span class="samp">address</span></samp>’<dd>Print the address of each node. Usually this is not meaningful as it 661changes according to the environment and source file. Its primary use 662is for tying up a dump file with a debug environment. 663<br><dt>‘<samp><span class="samp">asmname</span></samp>’<dd>If <code>DECL_ASSEMBLER_NAME</code> has been set for a given decl, use that 664in the dump instead of <code>DECL_NAME</code>. Its primary use is ease of 665use working backward from mangled names in the assembly file. 666<br><dt>‘<samp><span class="samp">slim</span></samp>’<dd>Inhibit dumping of members of a scope or body of a function merely 667because that scope has been reached. Only dump such items when they 668are directly reachable by some other path. When dumping pretty-printed 669trees, this option inhibits dumping the bodies of control structures. 670<br><dt>‘<samp><span class="samp">raw</span></samp>’<dd>Print a raw representation of the tree. By default, trees are 671pretty-printed into a C-like representation. 672<br><dt>‘<samp><span class="samp">details</span></samp>’<dd>Enable more detailed dumps (not honored by every dump option). 673<br><dt>‘<samp><span class="samp">stats</span></samp>’<dd>Enable dumping various statistics about the pass (not honored by every dump 674option). 675<br><dt>‘<samp><span class="samp">blocks</span></samp>’<dd>Enable showing basic block boundaries (disabled in raw dumps). 676<br><dt>‘<samp><span class="samp">vops</span></samp>’<dd>Enable showing virtual operands for every statement. 677<br><dt>‘<samp><span class="samp">lineno</span></samp>’<dd>Enable showing line numbers for statements. 678<br><dt>‘<samp><span class="samp">uid</span></samp>’<dd>Enable showing the unique ID (<code>DECL_UID</code>) for each variable. 679<br><dt>‘<samp><span class="samp">verbose</span></samp>’<dd>Enable showing the tree dump for each statement. 680<br><dt>‘<samp><span class="samp">eh</span></samp>’<dd>Enable showing the EH region number holding each statement. 681<br><dt>‘<samp><span class="samp">all</span></samp>’<dd>Turn on all options, except <samp><span class="option">raw</span></samp>, <samp><span class="option">slim</span></samp>, <samp><span class="option">verbose</span></samp> 682and <samp><span class="option">lineno</span></samp>. 683</dl> 684 685 <p>The following tree dumps are possible: 686 <dl> 687<dt>‘<samp><span class="samp">original</span></samp>’<dd><a name="index-fdump_002dtree_002doriginal-627"></a>Dump before any tree based optimization, to <samp><var>file</var><span class="file">.original</span></samp>. 688 689 <br><dt>‘<samp><span class="samp">optimized</span></samp>’<dd><a name="index-fdump_002dtree_002doptimized-628"></a>Dump after all tree based optimization, to <samp><var>file</var><span class="file">.optimized</span></samp>. 690 691 <br><dt>‘<samp><span class="samp">gimple</span></samp>’<dd><a name="index-fdump_002dtree_002dgimple-629"></a>Dump each function before and after the gimplification pass to a file. The 692file name is made by appending <samp><span class="file">.gimple</span></samp> to the source file name. 693 694 <br><dt>‘<samp><span class="samp">cfg</span></samp>’<dd><a name="index-fdump_002dtree_002dcfg-630"></a>Dump the control flow graph of each function to a file. The file name is 695made by appending <samp><span class="file">.cfg</span></samp> to the source file name. 696 697 <br><dt>‘<samp><span class="samp">vcg</span></samp>’<dd><a name="index-fdump_002dtree_002dvcg-631"></a>Dump the control flow graph of each function to a file in VCG format. The 698file name is made by appending <samp><span class="file">.vcg</span></samp> to the source file name. Note 699that if the file contains more than one function, the generated file cannot 700be used directly by VCG. You will need to cut and paste each function's 701graph into its own separate file first. 702 703 <br><dt>‘<samp><span class="samp">ch</span></samp>’<dd><a name="index-fdump_002dtree_002dch-632"></a>Dump each function after copying loop headers. The file name is made by 704appending <samp><span class="file">.ch</span></samp> to the source file name. 705 706 <br><dt>‘<samp><span class="samp">ssa</span></samp>’<dd><a name="index-fdump_002dtree_002dssa-633"></a>Dump SSA related information to a file. The file name is made by appending 707<samp><span class="file">.ssa</span></samp> to the source file name. 708 709 <br><dt>‘<samp><span class="samp">alias</span></samp>’<dd><a name="index-fdump_002dtree_002dalias-634"></a>Dump aliasing information for each function. The file name is made by 710appending <samp><span class="file">.alias</span></samp> to the source file name. 711 712 <br><dt>‘<samp><span class="samp">ccp</span></samp>’<dd><a name="index-fdump_002dtree_002dccp-635"></a>Dump each function after CCP. The file name is made by appending 713<samp><span class="file">.ccp</span></samp> to the source file name. 714 715 <br><dt>‘<samp><span class="samp">storeccp</span></samp>’<dd><a name="index-fdump_002dtree_002dstoreccp-636"></a>Dump each function after STORE-CCP. The file name is made by appending 716<samp><span class="file">.storeccp</span></samp> to the source file name. 717 718 <br><dt>‘<samp><span class="samp">pre</span></samp>’<dd><a name="index-fdump_002dtree_002dpre-637"></a>Dump trees after partial redundancy elimination. The file name is made 719by appending <samp><span class="file">.pre</span></samp> to the source file name. 720 721 <br><dt>‘<samp><span class="samp">fre</span></samp>’<dd><a name="index-fdump_002dtree_002dfre-638"></a>Dump trees after full redundancy elimination. The file name is made 722by appending <samp><span class="file">.fre</span></samp> to the source file name. 723 724 <br><dt>‘<samp><span class="samp">copyprop</span></samp>’<dd><a name="index-fdump_002dtree_002dcopyprop-639"></a>Dump trees after copy propagation. The file name is made 725by appending <samp><span class="file">.copyprop</span></samp> to the source file name. 726 727 <br><dt>‘<samp><span class="samp">store_copyprop</span></samp>’<dd><a name="index-fdump_002dtree_002dstore_005fcopyprop-640"></a>Dump trees after store copy-propagation. The file name is made 728by appending <samp><span class="file">.store_copyprop</span></samp> to the source file name. 729 730 <br><dt>‘<samp><span class="samp">dce</span></samp>’<dd><a name="index-fdump_002dtree_002ddce-641"></a>Dump each function after dead code elimination. The file name is made by 731appending <samp><span class="file">.dce</span></samp> to the source file name. 732 733 <br><dt>‘<samp><span class="samp">mudflap</span></samp>’<dd><a name="index-fdump_002dtree_002dmudflap-642"></a>Dump each function after adding mudflap instrumentation. The file name is 734made by appending <samp><span class="file">.mudflap</span></samp> to the source file name. 735 736 <br><dt>‘<samp><span class="samp">sra</span></samp>’<dd><a name="index-fdump_002dtree_002dsra-643"></a>Dump each function after performing scalar replacement of aggregates. The 737file name is made by appending <samp><span class="file">.sra</span></samp> to the source file name. 738 739 <br><dt>‘<samp><span class="samp">sink</span></samp>’<dd><a name="index-fdump_002dtree_002dsink-644"></a>Dump each function after performing code sinking. The file name is made 740by appending <samp><span class="file">.sink</span></samp> to the source file name. 741 742 <br><dt>‘<samp><span class="samp">dom</span></samp>’<dd><a name="index-fdump_002dtree_002ddom-645"></a>Dump each function after applying dominator tree optimizations. The file 743name is made by appending <samp><span class="file">.dom</span></samp> to the source file name. 744 745 <br><dt>‘<samp><span class="samp">dse</span></samp>’<dd><a name="index-fdump_002dtree_002ddse-646"></a>Dump each function after applying dead store elimination. The file 746name is made by appending <samp><span class="file">.dse</span></samp> to the source file name. 747 748 <br><dt>‘<samp><span class="samp">phiopt</span></samp>’<dd><a name="index-fdump_002dtree_002dphiopt-647"></a>Dump each function after optimizing PHI nodes into straightline code. The file 749name is made by appending <samp><span class="file">.phiopt</span></samp> to the source file name. 750 751 <br><dt>‘<samp><span class="samp">forwprop</span></samp>’<dd><a name="index-fdump_002dtree_002dforwprop-648"></a>Dump each function after forward propagating single use variables. The file 752name is made by appending <samp><span class="file">.forwprop</span></samp> to the source file name. 753 754 <br><dt>‘<samp><span class="samp">copyrename</span></samp>’<dd><a name="index-fdump_002dtree_002dcopyrename-649"></a>Dump each function after applying the copy rename optimization. The file 755name is made by appending <samp><span class="file">.copyrename</span></samp> to the source file name. 756 757 <br><dt>‘<samp><span class="samp">nrv</span></samp>’<dd><a name="index-fdump_002dtree_002dnrv-650"></a>Dump each function after applying the named return value optimization on 758generic trees. The file name is made by appending <samp><span class="file">.nrv</span></samp> to the source 759file name. 760 761 <br><dt>‘<samp><span class="samp">vect</span></samp>’<dd><a name="index-fdump_002dtree_002dvect-651"></a>Dump each function after applying vectorization of loops. The file name is 762made by appending <samp><span class="file">.vect</span></samp> to the source file name. 763 764 <br><dt>‘<samp><span class="samp">slp</span></samp>’<dd><a name="index-fdump_002dtree_002dslp-652"></a>Dump each function after applying vectorization of basic blocks. The file name 765is made by appending <samp><span class="file">.slp</span></samp> to the source file name. 766 767 <br><dt>‘<samp><span class="samp">vrp</span></samp>’<dd><a name="index-fdump_002dtree_002dvrp-653"></a>Dump each function after Value Range Propagation (VRP). The file name 768is made by appending <samp><span class="file">.vrp</span></samp> to the source file name. 769 770 <br><dt>‘<samp><span class="samp">all</span></samp>’<dd><a name="index-fdump_002dtree_002dall-654"></a>Enable all the available tree dumps with the flags provided in this option. 771</dl> 772 773 <br><dt><code>-ftree-vectorizer-verbose=</code><var>n</var><dd><a name="index-ftree_002dvectorizer_002dverbose-655"></a>This option controls the amount of debugging output the vectorizer prints. 774This information is written to standard error, unless 775<samp><span class="option">-fdump-tree-all</span></samp> or <samp><span class="option">-fdump-tree-vect</span></samp> is specified, 776in which case it is output to the usual dump listing file, <samp><span class="file">.vect</span></samp>. 777For <var>n</var>=0 no diagnostic information is reported. 778If <var>n</var>=1 the vectorizer reports each loop that got vectorized, 779and the total number of loops that got vectorized. 780If <var>n</var>=2 the vectorizer also reports non-vectorized loops that passed 781the first analysis phase (vect_analyze_loop_form) - i.e. countable, 782inner-most, single-bb, single-entry/exit loops. This is the same verbosity 783level that <samp><span class="option">-fdump-tree-vect-stats</span></samp> uses. 784Higher verbosity levels mean either more information dumped for each 785reported loop, or same amount of information reported for more loops: 786if <var>n</var>=3, vectorizer cost model information is reported. 787If <var>n</var>=4, alignment related information is added to the reports. 788If <var>n</var>=5, data-references related information (e.g. memory dependences, 789memory access-patterns) is added to the reports. 790If <var>n</var>=6, the vectorizer reports also non-vectorized inner-most loops 791that did not pass the first analysis phase (i.e., may not be countable, or 792may have complicated control-flow). 793If <var>n</var>=7, the vectorizer reports also non-vectorized nested loops. 794If <var>n</var>=8, SLP related information is added to the reports. 795For <var>n</var>=9, all the information the vectorizer generates during its 796analysis and transformation is reported. This is the same verbosity level 797that <samp><span class="option">-fdump-tree-vect-details</span></samp> uses. 798 799 <br><dt><code>-frandom-seed=</code><var>string</var><dd><a name="index-frandom_002dseed-656"></a>This option provides a seed that GCC uses when it would otherwise use 800random numbers. It is used to generate certain symbol names 801that have to be different in every compiled file. It is also used to 802place unique stamps in coverage data files and the object files that 803produce them. You can use the <samp><span class="option">-frandom-seed</span></samp> option to produce 804reproducibly identical object files. 805 806 <p>The <var>string</var> should be different for every file you compile. 807 808 <br><dt><code>-fsched-verbose=</code><var>n</var><dd><a name="index-fsched_002dverbose-657"></a>On targets that use instruction scheduling, this option controls the 809amount of debugging output the scheduler prints. This information is 810written to standard error, unless <samp><span class="option">-fdump-rtl-sched1</span></samp> or 811<samp><span class="option">-fdump-rtl-sched2</span></samp> is specified, in which case it is output 812to the usual dump listing file, <samp><span class="file">.sched1</span></samp> or <samp><span class="file">.sched2</span></samp> 813respectively. However for <var>n</var> greater than nine, the output is 814always printed to standard error. 815 816 <p>For <var>n</var> greater than zero, <samp><span class="option">-fsched-verbose</span></samp> outputs the 817same information as <samp><span class="option">-fdump-rtl-sched1</span></samp> and <samp><span class="option">-fdump-rtl-sched2</span></samp>. 818For <var>n</var> greater than one, it also output basic block probabilities, 819detailed ready list information and unit/insn info. For <var>n</var> greater 820than two, it includes RTL at abort point, control-flow and regions info. 821And for <var>n</var> over four, <samp><span class="option">-fsched-verbose</span></samp> also includes 822dependence info. 823 824 <br><dt><code>-save-temps</code><dt><code>-save-temps=cwd</code><dd><a name="index-save_002dtemps-658"></a>Store the usual “temporary” intermediate files permanently; place them 825in the current directory and name them based on the source file. Thus, 826compiling <samp><span class="file">foo.c</span></samp> with ‘<samp><span class="samp">-c -save-temps</span></samp>’ would produce files 827<samp><span class="file">foo.i</span></samp> and <samp><span class="file">foo.s</span></samp>, as well as <samp><span class="file">foo.o</span></samp>. This creates a 828preprocessed <samp><span class="file">foo.i</span></samp> output file even though the compiler now 829normally uses an integrated preprocessor. 830 831 <p>When used in combination with the <samp><span class="option">-x</span></samp> command line option, 832<samp><span class="option">-save-temps</span></samp> is sensible enough to avoid over writing an 833input source file with the same extension as an intermediate file. 834The corresponding intermediate file may be obtained by renaming the 835source file before using <samp><span class="option">-save-temps</span></samp>. 836 837 <p>If you invoke GCC in parallel, compiling several different source 838files that share a common base name in different subdirectories or the 839same source file compiled for multiple output destinations, it is 840likely that the different parallel compilers will interfere with each 841other, and overwrite the temporary files. For instance: 842 843 <pre class="smallexample"> gcc -save-temps -o outdir1/foo.o indir1/foo.c& 844 gcc -save-temps -o outdir2/foo.o indir2/foo.c& 845</pre> 846 <p>may result in <samp><span class="file">foo.i</span></samp> and <samp><span class="file">foo.o</span></samp> being written to 847simultaneously by both compilers. 848 849 <br><dt><code>-save-temps=obj</code><dd><a name="index-save_002dtemps_003dobj-659"></a>Store the usual “temporary” intermediate files permanently. If the 850<samp><span class="option">-o</span></samp> option is used, the temporary files are based on the 851object file. If the <samp><span class="option">-o</span></samp> option is not used, the 852<samp><span class="option">-save-temps=obj</span></samp> switch behaves like <samp><span class="option">-save-temps</span></samp>. 853 854 <p>For example: 855 856 <pre class="smallexample"> gcc -save-temps=obj -c foo.c 857 gcc -save-temps=obj -c bar.c -o dir/xbar.o 858 gcc -save-temps=obj foobar.c -o dir2/yfoobar 859</pre> 860 <p>would create <samp><span class="file">foo.i</span></samp>, <samp><span class="file">foo.s</span></samp>, <samp><span class="file">dir/xbar.i</span></samp>, 861<samp><span class="file">dir/xbar.s</span></samp>, <samp><span class="file">dir2/yfoobar.i</span></samp>, <samp><span class="file">dir2/yfoobar.s</span></samp>, and 862<samp><span class="file">dir2/yfoobar.o</span></samp>. 863 864 <br><dt><code>-time</code><span class="roman">[</span><code>=</code><var>file</var><span class="roman">]</span><dd><a name="index-time-660"></a>Report the CPU time taken by each subprocess in the compilation 865sequence. For C source files, this is the compiler proper and assembler 866(plus the linker if linking is done). 867 868 <p>Without the specification of an output file, the output looks like this: 869 870 <pre class="smallexample"> # cc1 0.12 0.01 871 # as 0.00 0.01 872</pre> 873 <p>The first number on each line is the “user time”, that is time spent 874executing the program itself. The second number is “system time”, 875time spent executing operating system routines on behalf of the program. 876Both numbers are in seconds. 877 878 <p>With the specification of an output file, the output is appended to the 879named file, and it looks like this: 880 881 <pre class="smallexample"> 0.12 0.01 cc1 <var>options</var> 882 0.00 0.01 as <var>options</var> 883</pre> 884 <p>The “user time” and the “system time” are moved before the program 885name, and the options passed to the program are displayed, so that one 886can later tell what file was being compiled, and with which options. 887 888 <br><dt><code>-fvar-tracking</code><dd><a name="index-fvar_002dtracking-661"></a>Run variable tracking pass. It computes where variables are stored at each 889position in code. Better debugging information is then generated 890(if the debugging information format supports this information). 891 892 <p>It is enabled by default when compiling with optimization (<samp><span class="option">-Os</span></samp>, 893<samp><span class="option">-O</span></samp>, <samp><span class="option">-O2</span></samp>, <small class="dots">...</small>), debugging information (<samp><span class="option">-g</span></samp>) and 894the debug info format supports it. 895 896 <br><dt><code>-fvar-tracking-assignments</code><dd><a name="index-fvar_002dtracking_002dassignments-662"></a><a name="index-fno_002dvar_002dtracking_002dassignments-663"></a>Annotate assignments to user variables early in the compilation and 897attempt to carry the annotations over throughout the compilation all the 898way to the end, in an attempt to improve debug information while 899optimizing. Use of <samp><span class="option">-gdwarf-4</span></samp> is recommended along with it. 900 901 <p>It can be enabled even if var-tracking is disabled, in which case 902annotations will be created and maintained, but discarded at the end. 903 904 <br><dt><code>-fvar-tracking-assignments-toggle</code><dd><a name="index-fvar_002dtracking_002dassignments_002dtoggle-664"></a><a name="index-fno_002dvar_002dtracking_002dassignments_002dtoggle-665"></a>Toggle <samp><span class="option">-fvar-tracking-assignments</span></samp>, in the same way that 905<samp><span class="option">-gtoggle</span></samp> toggles <samp><span class="option">-g</span></samp>. 906 907 <br><dt><code>-print-file-name=</code><var>library</var><dd><a name="index-print_002dfile_002dname-666"></a>Print the full absolute name of the library file <var>library</var> that 908would be used when linking—and don't do anything else. With this 909option, GCC does not compile or link anything; it just prints the 910file name. 911 912 <br><dt><code>-print-multi-directory</code><dd><a name="index-print_002dmulti_002ddirectory-667"></a>Print the directory name corresponding to the multilib selected by any 913other switches present in the command line. This directory is supposed 914to exist in <samp><span class="env">GCC_EXEC_PREFIX</span></samp>. 915 916 <br><dt><code>-print-multi-lib</code><dd><a name="index-print_002dmulti_002dlib-668"></a>Print the mapping from multilib directory names to compiler switches 917that enable them. The directory name is separated from the switches by 918‘<samp><span class="samp">;</span></samp>’, and each switch starts with an ‘<samp><span class="samp">@</span></samp>’ instead of the 919‘<samp><span class="samp">-</span></samp>’, without spaces between multiple switches. This is supposed to 920ease shell-processing. 921 922 <br><dt><code>-print-multi-os-directory</code><dd><a name="index-print_002dmulti_002dos_002ddirectory-669"></a>Print the path to OS libraries for the selected 923multilib, relative to some <samp><span class="file">lib</span></samp> subdirectory. If OS libraries are 924present in the <samp><span class="file">lib</span></samp> subdirectory and no multilibs are used, this is 925usually just <samp><span class="file">.</span></samp>, if OS libraries are present in <samp><span class="file">lib</span><var>suffix</var></samp> 926sibling directories this prints e.g. <samp><span class="file">../lib64</span></samp>, <samp><span class="file">../lib</span></samp> or 927<samp><span class="file">../lib32</span></samp>, or if OS libraries are present in <samp><span class="file">lib/</span><var>subdir</var></samp> 928subdirectories it prints e.g. <samp><span class="file">amd64</span></samp>, <samp><span class="file">sparcv9</span></samp> or <samp><span class="file">ev6</span></samp>. 929 930 <br><dt><code>-print-prog-name=</code><var>program</var><dd><a name="index-print_002dprog_002dname-670"></a>Like <samp><span class="option">-print-file-name</span></samp>, but searches for a program such as ‘<samp><span class="samp">cpp</span></samp>’. 931 932 <br><dt><code>-print-libgcc-file-name</code><dd><a name="index-print_002dlibgcc_002dfile_002dname-671"></a>Same as <samp><span class="option">-print-file-name=libgcc.a</span></samp>. 933 934 <p>This is useful when you use <samp><span class="option">-nostdlib</span></samp> or <samp><span class="option">-nodefaultlibs</span></samp> 935but you do want to link with <samp><span class="file">libgcc.a</span></samp>. You can do 936 937 <pre class="smallexample"> gcc -nostdlib <var>files</var>... `gcc -print-libgcc-file-name` 938</pre> 939 <br><dt><code>-print-search-dirs</code><dd><a name="index-print_002dsearch_002ddirs-672"></a>Print the name of the configured installation directory and a list of 940program and library directories <samp><span class="command">gcc</span></samp> will search—and don't do anything else. 941 942 <p>This is useful when <samp><span class="command">gcc</span></samp> prints the error message 943‘<samp><span class="samp">installation problem, cannot exec cpp0: No such file or directory</span></samp>’. 944To resolve this you either need to put <samp><span class="file">cpp0</span></samp> and the other compiler 945components where <samp><span class="command">gcc</span></samp> expects to find them, or you can set the environment 946variable <samp><span class="env">GCC_EXEC_PREFIX</span></samp> to the directory where you installed them. 947Don't forget the trailing ‘<samp><span class="samp">/</span></samp>’. 948See <a href="Environment-Variables.html#Environment-Variables">Environment Variables</a>. 949 950 <br><dt><code>-print-sysroot</code><dd><a name="index-print_002dsysroot-673"></a>Print the target sysroot directory that will be used during 951compilation. This is the target sysroot specified either at configure 952time or using the <samp><span class="option">--sysroot</span></samp> option, possibly with an extra 953suffix that depends on compilation options. If no target sysroot is 954specified, the option prints nothing. 955 956 <br><dt><code>-print-sysroot-headers-suffix</code><dd><a name="index-print_002dsysroot_002dheaders_002dsuffix-674"></a>Print the suffix added to the target sysroot when searching for 957headers, or give an error if the compiler is not configured with such 958a suffix—and don't do anything else. 959 960 <br><dt><code>-dumpmachine</code><dd><a name="index-dumpmachine-675"></a>Print the compiler's target machine (for example, 961‘<samp><span class="samp">i686-pc-linux-gnu</span></samp>’)—and don't do anything else. 962 963 <br><dt><code>-dumpversion</code><dd><a name="index-dumpversion-676"></a>Print the compiler version (for example, ‘<samp><span class="samp">3.0</span></samp>’)—and don't do 964anything else. 965 966 <br><dt><code>-dumpspecs</code><dd><a name="index-dumpspecs-677"></a>Print the compiler's built-in specs—and don't do anything else. (This 967is used when GCC itself is being built.) See <a href="Spec-Files.html#Spec-Files">Spec Files</a>. 968 969 <br><dt><code>-feliminate-unused-debug-types</code><dd><a name="index-feliminate_002dunused_002ddebug_002dtypes-678"></a>Normally, when producing DWARF2 output, GCC will emit debugging 970information for all types declared in a compilation 971unit, regardless of whether or not they are actually used 972in that compilation unit. Sometimes this is useful, such as 973if, in the debugger, you want to cast a value to a type that is 974not actually used in your program (but is declared). More often, 975however, this results in a significant amount of wasted space. 976With this option, GCC will avoid producing debug symbol output 977for types that are nowhere used in the source file being compiled. 978</dl> 979 980 </body></html> 981 982