gcc.1 22 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592
  1. .\ " =====================
  2. .\ " This version is 1.20.
  3. .\ " =====================
  4. .TH GCC 1 "19 April 1988" "Printed 04/20/88"
  5. .\ ".UC 4
  6. .SH NAME
  7. gcc \- GNU project C Compiler
  8. .SH SYNOPSIS
  9. .B gcc
  10. [ option ] ... file ...
  11. .SH DESCRIPTION
  12. The
  13. .I GNU\ C\ compiler
  14. uses a command syntax much like the Unix C compiler. The
  15. .I gcc
  16. program accepts options and file names as operands. Multiple
  17. single-letter options may
  18. .I not
  19. be grouped: `\fB\-dr\fR'
  20. is very different from `\fB\-d\ \-r\fR'.
  21. When you invoke
  22. .I GNU\ CC
  23. \, it normally does preprocessing, compilation,
  24. assembly and linking. File names with `\fB.c\fR'
  25. are taken as C
  26. source to be preprocessed and compiled; compiler output files plus any
  27. input files with names ending in `\fB.s\fR'
  28. are assembled; then the
  29. resulting object files, plus any other input files, are linked together to
  30. produce an executable.
  31. .br
  32. Command options allow you to stop this process at an intermediate stage.
  33. For example, the `\fB\-c\fR'
  34. option says not to run the linker.
  35. Then the output consists of object files output by the assembler.
  36. .br
  37. Other command options are passed on to one stage.
  38. Some options control the preprocessor and others the compiler itself.
  39. .br
  40. Here are the options to control the overall compilation process,
  41. including those that say whether to link, whether to assemble, and so on.
  42. .TP
  43. .BI \-o "\ \ file"
  44. Place linker output in file \fIfile\fR.
  45. This applies regardless to whatever sort of output is being produced,
  46. whether it be an executable file, an object file, an assembler file or
  47. preprocessed C code.
  48. If `\fB\-o\fR'
  49. is not specified, the default is to put an executable file
  50. in `a.out', the object file `\fIsource\fR.o', and preprocessed C
  51. on standard output.
  52. .TP
  53. .B \-c
  54. Compile or assemble the source files, but do not link.
  55. Produce object files with names made by replacing `\fB\.c\fR'
  56. or `\fB\.s\fR'
  57. with `\fB\.o\fR'
  58. at the end of the input file names.
  59. Do nothing at all for object files specified as input.
  60. .TP
  61. .B \-S
  62. Compile into assembler code but do not assemble.
  63. The assembler output file name is made by replacing `\fB\.c\fR'
  64. with `\fB\.s\fR'
  65. at the end of the input file name.
  66. Do nothing at all for assembler source files or object files specified
  67. as input.
  68. .TP
  69. .B \-E
  70. Run only the C preprocessor.
  71. Preprocess all the C source files specified
  72. and output the results to standard output.
  73. .TP
  74. .B \-v
  75. Compiler driver program prints the commands it executes as it runs
  76. the preprocessor, compiler proper, assembler and linker.
  77. Some of these are directed to print their own version numbers.
  78. .TP
  79. .BI \-B "prefix"
  80. Compiler driver program tries \fIprefix\fR as a prefix for each program
  81. it tries to run. These programs are `\fBcpp\fR', `\fBcc1\fR',
  82. `\fBas\fR' and `\fBld\fR'.
  83. For each subprogram to be run, the compiler driver first tries
  84. the `\fB\-B\fR' prefix, if any.
  85. If that name is not found, or if `\fB\-B\fR' was not specified,
  86. the driver tries two standard prefixes,
  87. which are `\fB/usr/lib/gcc-\fR' and `\fB/usr/local/lib/gcc-\fR'.
  88. If neither of those results in a file name that is found,
  89. the unmodified program name is searched for using the directories
  90. specified in your `\fBPATH\fR' environment variable.
  91. The run-time support file `\fBgnulib\fR' is also searched for using
  92. the `\fB\-B\fR' prefix, if needed.
  93. If it is not found there, the two standard prefixes above are tried,
  94. and that is all.
  95. The file is left out of the link if it is not found by those means.
  96. Most of the time, on most machines, you can do without it.
  97. .PP
  98. These options control the C preprocessor,
  99. which is run on each C source file before actual compilation.
  100. If you use the `\fB\-E\fR' option, nothing is done except C preprocessing.
  101. Some of these options make sense only together with `\fB\-E\fR'
  102. because they request preprocessor output that is not suitable
  103. for actual compilation.
  104. .TP
  105. .B \-C
  106. Tell the preprocessor not to discard comments. Used with the `\fB\-E\fR'
  107. option.
  108. .TP
  109. .BI \-I "dir"
  110. Search directory
  111. .I dir
  112. for include files.
  113. .TP
  114. .B \-M
  115. Tell the preprocessor to output a rule suitable for \fBmake\fR
  116. describing the dependencies of each source file.
  117. For each source file, the preprocessor outputs one \fBmake\fR-rule
  118. whose target is the object file name for that source file and
  119. whose dependencies are all the files `\fB#include\fR'd in it.
  120. This rule may be a single line or may be continued with `\fB\\\fR'-newline
  121. if it is long. `\fB\-M\fR' implies `\fB\-E\fR'.
  122. .TP
  123. .B \-MM
  124. Like `\fB\-M\fR' but the output mentions only the user-header files included
  125. with `\fB#include "file"\fR'.
  126. System header files included with `\fB#include <\fIfile\fB>\fR'
  127. are omitted.`\fB\-MM\fR' implies `\fB\-E\fR'.
  128. .TP
  129. .BI \-D "macro"
  130. Define macro \fImacro\fR
  131. with the empty string as its definition.
  132. .TP
  133. .B \-D\fImacro=defn\fR
  134. Define macro \fImacro\fR as \fIdefn\fR.
  135. .TP
  136. .BI \-U "macro"
  137. Undefine macro \fImacro\fR.
  138. .TP
  139. .B \-T
  140. Support ANSI C trigraphs.
  141. You don't want to know about this brain-damage.
  142. The `\fB\-ansi\fR' option also has this effect.
  143. .PP
  144. These options control the details of C compilation itself.
  145. .TP
  146. .B \-ansi
  147. Support all ANSI standard C programs.
  148. This turns off certain features of GNU C that are incompatible with
  149. ANSI C, such as the \fBasm\fR, \fBinline\fR and \fBtypeof\fR keywords, and
  150. predefined macros such as \fBunix\fR and \fBvax\fR that identify
  151. the type of system you are using.
  152. It also enables the undesirable and rarely used ANSI tragraph feature.
  153. The `\fB\-ansi\fR' option does not cause non-ANSI programs to be rejected
  154. gratuitously.
  155. For that object that, `\fB\-pedantic\fR' is required
  156. in addition to `\fB\-ansi\fR'.
  157. The macro \fB__STRICT_ANSI__\fR
  158. is predefined when the `-ansi' option is used.
  159. Some header files may notice this macro and refrain from declaring
  160. certain functions or defining certain macros that the ANSI standard
  161. doesn't call for; this is to avoid interfering with any programs
  162. that might use these names for other things.
  163. .TP
  164. .B \-traditional
  165. Attempt to support some aspects of traditional C compilers.
  166. Specifically:
  167. .br
  168. \(** All \fBextern\fR declarations take effect globally even if
  169. they are written inside of a function definition.
  170. This includes implicit declarations of functions.
  171. .br
  172. \(** The keywords \fBtypeof\fR, \fBinline\fR, \fBsigned\fR, \fBconst\fR
  173. and \fBvolatile\fR are not recognized.
  174. .br
  175. \(** Comparisons between pointers and integers are always allowed.
  176. .br
  177. \(** Integer types \fBunsigned short\fR and \fBunsigned char\fR
  178. promote to \fBunsigned int\fR.
  179. .TP
  180. .B \-pedantic
  181. Issue all the warnings demanded by strict ANSI standard C;
  182. reject all programs that use forbidden extensions.
  183. Valid ANSI standard C programs should compile properly with or
  184. without this option (though a rare few will require `\fB\-ansi\fR'.
  185. However, without this option, certain GNU extensions and
  186. traditional C features are supported as well.
  187. With this option, they are rejected.
  188. There is no reason to \fIuse\fR
  189. this option; it exists only to satisfy pedants.
  190. .TP
  191. .B \-O
  192. Optimize. Optimizing compilation takes somewhat more time, and a lot
  193. more memory for a large function.
  194. Without `\fB\-O\fR', the compiler's goal is to reduce the cost of
  195. compilation and to make debugging produce the expected results.
  196. Statements are independent: if you stop the program with a breakpoint
  197. between statements, you can then assign a new value to any variable or
  198. change the program counter to any other statement in the function and
  199. get exactly the results you would expect from the source code.
  200. Without `\fB\-O\fR', only variables declared \fBregister\fR
  201. are allocated in registers.
  202. The resulting compiled code is
  203. a little worse than produced by PCC without `\fB\-O\fR'.
  204. With `\fB\-O\fR', the compiler tries to reduce code size and execution time.
  205. Some of the `\fB\-f\fR' options described below turn specific
  206. kinds of optimization on or off.
  207. .TP
  208. .B \-g
  209. Produce debugging information in DBX format.
  210. Unlike most other C compilers,
  211. GNU CC allows you to use `\fB\-g\fR' with `\fB\-O\fR'.
  212. The shortcuts taken by optimized code may occasionally
  213. produce surprising results: some variables you declared may not exist
  214. at all; flow of control may briefly move where you did not expect it;
  215. some statements may not be executed because they compute constant
  216. results or their values were already at hand; some statements may
  217. execute in different places because they were moved out of loops.
  218. Nevertheless it proves possible to debug optimized output.
  219. This makes it reasonable to use the optimizer for programs that might
  220. have bugs.
  221. .TP
  222. .B \-gg
  223. Produce debugging information in GDB(GNU Debugger)'s own format.
  224. This requires the GNU assembler and linker
  225. in order to work.
  226. .TP
  227. .B \-w
  228. Inhibit all warning messages.
  229. .TP
  230. .B \-W
  231. Print extra warning messages for these events:
  232. .br
  233. \(** An automatic variable is used without first being initialized.
  234. These warnings are possible only in optimizing compilation, because
  235. they require data flow information that is computed only when
  236. optimizing.
  237. They occur only for variables that are candidates for register
  238. allocation. Therefore, they do not occur for a variable that is
  239. declared
  240. .B volatile,
  241. or whose address is taken, or whose size is other than
  242. 1,2,4 or 8 bytes. Also, they do not occur for structures,
  243. unions or arrays, even when they are in registers.
  244. Note that there may be no warning about a variable that is used
  245. only to compute a value that itself is never used, because such
  246. computations may be deleted by the flow analysis pass before the
  247. warnings are printed.
  248. These warnings are made optional because GNU CC is not smart
  249. enough to see all the reasons why the code might be correct
  250. despite appearing to have an error.
  251. .br
  252. \(** A nonvolantile automatic variable might be changed
  253. by a call to \flongjmp\fR.
  254. These warnings as well are possible only in optimizing compilation.
  255. The compiler only the calls to \fBsetjmp\fR.
  256. It cannot know where \fBlongjmp\fR
  257. will be called; in fact, a signal handler could cal it at any point
  258. in the code. As a result, you may get a warning even when there is
  259. in fact no problem because \fBlongjmp\fR
  260. cannot in fact be called at the place which would cause a problem.
  261. .br
  262. \(** A function can return either with or without a value.
  263. (Falling off the end of the function body is considered returning
  264. without a value.)
  265. Spurious warning can occur because GNU CC does not realize that
  266. certain functions (including \fBabort\fR
  267. and \fBlongjmp\fR) will never return.
  268. .TP
  269. .B \-Wimplicit
  270. Warn whenever a function is implicitly declared.
  271. .TP
  272. .B \-Wreturn-type
  273. Warn whenever a function is defined with a return-type that
  274. defaults to \fBint\fR. Also warn about any \fBreturn\fR
  275. statement with no return-value in a function whose return-type
  276. is not \fBvoid\fR.
  277. .TP
  278. .B \-p
  279. Generate extra code to write profile information suitable for the
  280. analysis program \fBprof\fR.
  281. .TP
  282. .B \-pg
  283. Generate extra code to write profile information suitable for the
  284. analysis program \fBgprof\fR.
  285. .TP
  286. .BI \-m "machinespec"
  287. Machine-dependent option specifying something about the type of target machine.
  288. These options are defined by the macro \fBTARGET_SWITCHES\fR
  289. in the machine description. The default for the options is also
  290. defined by that macro, which enables you to change the defaults.
  291. .IP
  292. These are the `\fB\-m\fR' options defined in the 68000 machine description:
  293. .TP 10
  294. .B \ \ \ \ \ \ \ \ \-m68020
  295. Generate output for a 68020 (rather than a 68000).
  296. This is the default if you use the unmodified sources.
  297. .TP 10
  298. .B \ \ \ \ \ \ \ \ \-m68000
  299. Generate output for a 68000 (rather than a 68020).
  300. .TP 10
  301. .B \ \ \ \ \ \ \ \ \-m68881
  302. Generate output containing 68881 instructions for floating point.
  303. This is the default if you use the unmodified sources.
  304. .TP 10
  305. .B \ \ \ \ \ \ \ \ \-msoft-float
  306. Generate output containing library calls for floating point.
  307. .TP 10
  308. .B \ \ \ \ \ \ \ \ \-mshort
  309. Consider type \fBint\fR to be 16 bits wide, like \fBshort int\fR.
  310. .TP 10
  311. .B \ \ \ \ \ \ \ \ \-mnobitfield
  312. Do not use the bit-field instructions.
  313. .B '\-m68000'
  314. implies
  315. .B '\-mnobitfield'.
  316. .TP 10
  317. .B \ \ \ \ \ \ \ \ \-mbitfield
  318. Do use the bit-field instructions.
  319. .B '\-m68020'
  320. implies
  321. .B '\-mbitfield'.
  322. This is the default if you use the unmodified sources.
  323. .TP 10
  324. .B \ \ \ \ \ \ \ \ \-mrtd
  325. Use a different function-calling convention, in which functions that
  326. take a fixed number of arguments return with the \fBrtd\fR
  327. instruction, which pops their arguments while returning. This saves
  328. one instruction in the caller since there is no need to pop the
  329. arguments there.
  330. This calling convention is incompatible with the one normally used on
  331. Unix, so you cannot use it if you need to call libraries compiled with
  332. the Unix compiler.
  333. Also, you must provide function prototypes for all functions that take
  334. variable numbers of arguments (including \fBprintf\fR); otherwise
  335. incorrect code will be generated for calls to those functions.
  336. In addition, seriously incorrect code will result if you call a
  337. function with too many arguments. (Normally, extra arguments are
  338. harmlessly ignored.)
  339. The \fBrtd\fR
  340. instruction is supported by the 68010 and 68020
  341. processors, but not by the 68000.
  342. .IP
  343. These are the `\fB\-m\fR' options defined in the VAX machine description:
  344. .TP 10
  345. .B \ \ \ \ \ \ \ \ \-munix
  346. Do not output certain jump instructions (
  347. .B aobleq
  348. and so on) that the Unix assembler
  349. for the VAX cannot handle across long ranges.
  350. .TP 10
  351. .B \ \ \ \ \ \ \ \ \-mgnu
  352. Do output those jump instructions; on the assumption
  353. that you will assemble with the GNU assembler.
  354. .TP 5
  355. .BI \-f "flag"
  356. Specify machine-independent flags. These are the flags:
  357. .TP 10
  358. .B \ \ \ \ \ \ \ \ \-ffloat-store
  359. Do not store floating-point variables in registers.
  360. This prevents undesirable excess precision on machines such as the 68000
  361. where the floating registers (of the 68881) keep more precision
  362. than a \fIdouble\fR is supposed to have.
  363. For most programs, the excess precision does only good, but a few
  364. programs rely on the precise definition of IEEE floating point.
  365. Use `
  366. .B \-ffloat-store'
  367. for such programs.
  368. .TP 10
  369. .B \ \ \ \ \ \ \ \ \-frno-asm
  370. Do not recognize \fBasm\fR, \fBinline\fR or \fBtypeof\fR
  371. as a keyword. These words may then be used as identifiers.
  372. .TP 10
  373. .B \ \ \ \ \ \ \ \ \-fno-defer-pop
  374. Always pop the arguments to each function call as soon as that
  375. function returns.
  376. Normally the compiler (when optimizing) lets arguments accumulate on the
  377. stack for several function calls and pops them all at once.
  378. .TP 10
  379. .B \ \ \ \ \ \ \ \ \-fcombine-regs
  380. Allow the combine pass to combine an instruction that copies one
  381. register into another.
  382. This might or might not produce better code when used in addition to `
  383. .B \-O'.
  384. .TP 10
  385. .B \ \ \ \ \ \ \ \ \-fforce-mem
  386. Force memory operands to be copied into registers before doing
  387. arithmetic on them.
  388. This may produce better code by making all
  389. memory references potential common subexpressions.
  390. When they are not common subexpressions,
  391. instruction combination should eliminate the separate register-load.
  392. .TP 10
  393. .B \ \ \ \ \ \ \ \ \-fforce-addr
  394. Force memory address constants to be copied into registers before
  395. doing arithmetic on them.
  396. This may produce better code just as `
  397. .B \-fforce-mem'
  398. may.
  399. .TP 10
  400. .B \ \ \ \ \ \ \ \ \-fomit-frame-pointer
  401. Don't keep the frame pointer in a register for functions that don't
  402. need one. This avoids the instructions to save, set up and restore
  403. frame pointers; it also makes an extra register available in many
  404. functions. \fBIt\ also\ makes\ debugging\ impossible.\fR
  405. On some machines, such as the VAX, this flag has no effect. This
  406. is because the standard calling sequence automatically handles
  407. the frame pointer and nothing is saved by pretending it doesn't exist.
  408. The machine-description macro \fBFRAME_POINTER_REQUIRED\fR
  409. controls whether a target machine supports this flag.
  410. .TP 10
  411. .B \ \ \ \ \ \ \ \ \-finline-functions
  412. Integrate all simple functions into their callers.
  413. The compiler heuristically decides which functions are simple enough
  414. to be worth integrating in this way.
  415. If all calls to a given function are integrated, and the function
  416. is declared \fBstatic\fR,
  417. then the function is normally not output as assembler code in its
  418. own right.
  419. .TP 10
  420. .B \ \ \ \ \ \ \ \ \-fkeep-inline-functions
  421. Even if all calls to a given function are integrated, and the
  422. function is declared \fBstatic\fR,
  423. nevertheless output a separate run-time callable version of
  424. the function.
  425. .TP 10
  426. .B \ \ \ \ \ \ \ \ \-fwritable-strings
  427. Store string constants in the writable data segment and don't uniquize them.
  428. This is for compatibility with old programs which assume
  429. they can write into string constants. Writing into string constants
  430. is a very bad idea; ``constants'' should be constant.
  431. .TP 10
  432. .B \ \ \ \ \ \ \ \ \-fno-function-cse
  433. Do not put function addresses in registers; make each instruction that
  434. calls a constant function contain the function's address explicitly.
  435. This option results in less efficient code, but some strange hacks
  436. that alter the assembler output may be confused by the optimizations
  437. performed when this option is not used.
  438. .TP 10
  439. .B \ \ \ \ \ \ \ \ \-fnostdinc
  440. Don't search the standard directories for include files.
  441. Only the directories you specify explicitly with the `\fB\-I\fR'
  442. option will be searched.
  443. .TP 10
  444. .B \ \ \ \ \ \ \ \ \-fnostdlib
  445. Don't use the standard system libraries and startup files
  446. when linking.
  447. Only the files you specify (plus `\fBgnulib\fR') will be passed to the linker.
  448. .TP 10
  449. .B \ \ \ \ \ \ \ \ \-fvolatile
  450. Consider all memory references through pointers to be volatile.
  451. .TP 10
  452. .B \ \ \ \ \ \ \ \ \-funsigned-char
  453. Let the type \fBchar\f be the unsigned, like \fBunsigned char\fR.
  454. Each kind of machine has a default for what \fBchar\fR
  455. should be. It is either like \fBunsigned char\fR
  456. by default of like \fBsigned char\fR
  457. by default. (Actually, at present, the default is always signed.)
  458. The type \fBchar\fR
  459. is always a distinct type from either \fBsigned char\fR
  460. or \fBunsigned char\fR,
  461. even though its behavior is always just like one of those two.
  462. .TP 10
  463. .B \ \ \ \ \ \ \ \ \-fsigned-char
  464. Let the type \fBchar\fR be the same as \fBsigned char\fR.
  465. .TP 10
  466. .B \ \ \ \ \ \ \ \ \-ffixed-\fIreg\fR
  467. Treat the register named \fIreg\fR as a fixed register; generated
  468. code should never refer to it (except perhaps as a stack pointer,
  469. frame pointer or in some other fixed role). \fIreg\fR
  470. must be the name of a register.
  471. The register names accepted are machine-specific and are defined in
  472. the \fBREGISTER_NAMES\fR
  473. macro in the machine description macro file.
  474. .TP 10
  475. .B \ \ \ \ \ \ \ \ \-fcall-used-\fIreg\fR
  476. Treat the register named \fIreg\fR
  477. as an allocatable register that is clobberred by function calls.
  478. It may be allocated for temporaries or variables
  479. that do not live across a call.
  480. Functions compiled this way will not save and restore the
  481. register \fIreg\fR.
  482. Use of this flag for a register that has a fixed pervasive role
  483. in the machine's execution model, such as the stack pointer or
  484. frame pointer, will produce disastrous results.
  485. .TP 10
  486. .B \ \ \ \ \ \ \ \ \-fcall-saved-\fIreg\fR
  487. Treat the register named \fIreg\fR
  488. as an allocatable register saved by functions.
  489. It may be allocated even for temporaries or
  490. variables that live across a call. Functions compiled this way
  491. will save and restore the register \fIreg\fR if they use it.
  492. Use of this flag for a register that has a fixed pervasive role
  493. in the machine's execution model, such as the stack pointer or
  494. frame pointer, will produce disastrous results.
  495. A different sort of disaster will result from the use of this
  496. flag for a register in which function values are may be returned.
  497. .TP
  498. .BI \-d "letters"
  499. Says to make debugging dumps at times specified by \fIletters\fR.
  500. Here are the possible letters:
  501. .TP 10
  502. .B \ \ \ \ \ \ \ \ r
  503. Dump after RTL generation.
  504. .TP 10
  505. .B \ \ \ \ \ \ \ \ j
  506. Dump after first jump optimization.
  507. .TP 10
  508. .B \ \ \ \ \ \ \ \ J
  509. Dump after last jump optimization.
  510. .TP 10
  511. .B \ \ \ \ \ \ \ \ s
  512. Dump after CSE (including the jump optimization that sometimes
  513. follows CSE).
  514. .TP 10
  515. .B \ \ \ \ \ \ \ \ L
  516. Dump after loop optimization.
  517. .TP 10
  518. .B \ \ \ \ \ \ \ \ f
  519. Dump after flow analysis.
  520. .TP 10
  521. .B \ \ \ \ \ \ \ \ c
  522. Dump after instruction combination.
  523. .TP 10
  524. .B \ \ \ \ \ \ \ \ l
  525. Dump after local register allocation.
  526. .TP 10
  527. .B \ \ \ \ \ \ \ \ g
  528. Dump after global register allocation.
  529. .TP 10
  530. .B \ \ \ \ \ \ \ \ m
  531. Print statistics on memory usage, at the end of the run.
  532. .SH FILES
  533. .ta \w'/usr/local/lib/gcc-gnulib 'u
  534. file.c input file
  535. .br
  536. file.o object file
  537. .br
  538. a.out loaded output
  539. .br
  540. /tmp/cc? temporary
  541. .br
  542. /usr/local/lib/gcc-cpp preprocessor
  543. .br
  544. /usr/local/lib/gcc-cc1 compiler
  545. .br
  546. /usr/local/lib/gcc-gnulib library need by GCC on some machines
  547. .br
  548. /lib/crt0.o runtime startoff
  549. .br
  550. /lib/libc.a standard library, see
  551. .IR intro (3)
  552. .br
  553. /usr/include standard directory for `#include' files
  554. .br
  555. .SH "SEE ALSO"
  556. B. W. Kernighan and D. M. Ritchie,
  557. .I The C Programming Language,
  558. Prentice-Hall,
  559. 1978
  560. .br
  561. B. W. Kernighan,
  562. .I
  563. Programming in C
  564. .br
  565. D. M. Ritchie,
  566. .I
  567. C Reference Manual
  568. .br
  569. adb(1), ld(1), dbx(1), as(1)
  570. .SH BUGS
  571. Bugs should be reported to bug-gcc@prep.ai.mit.edu. Bugs tend actually to be
  572. fixed if they can be isolated, so it is in your interest to report them
  573. in such a way that they can be easily reproduced according to get newer version.
  574. .SH COPYING
  575. Copyright (C) 1988 Richard M. Stallman.
  576. .br
  577. Permission is granted to make and distribute verbatim copies of
  578. this manual provided the copyright notice and this permission notice
  579. are preserved on all copies.
  580. .br
  581. Permission is granted to copy and distribute modified versions of this
  582. manual under the conditions for verbatim copying, provided also that the
  583. section entitled "GNU CC General Public License" is included exactly as
  584. in the original, and provided that the entire resulting derived work is
  585. distributed under the terms of a permission notice identical to this one.
  586. .br
  587. Permission is granted to copy and distribute translations of this manual
  588. into another language, under the above conditions for modified versions,
  589. except that the section entitled "GNU CC General Public License" may be
  590. included in a translation approved by the author instead of in the original
  591. English.