Annotation of GNUtools/cc/gcc.info-17, revision 1.1

1.1     ! root        1: This is Info file gcc.info, produced by Makeinfo-1.54 from the input
        !             2: file gcc.texi.
        !             3: 
        !             4:    This file documents the use and the internals of the GNU compiler.
        !             5: 
        !             6:    Published by the Free Software Foundation 675 Massachusetts Avenue
        !             7: Cambridge, MA 02139 USA
        !             8: 
        !             9:    Copyright (C) 1988, 1989, 1992, 1993 Free Software Foundation, Inc.
        !            10: 
        !            11:    Permission is granted to make and distribute verbatim copies of this
        !            12: manual provided the copyright notice and this permission notice are
        !            13: preserved on all copies.
        !            14: 
        !            15:    Permission is granted to copy and distribute modified versions of
        !            16: this manual under the conditions for verbatim copying, provided also
        !            17: that the sections entitled "GNU General Public License" and "Protect
        !            18: Your Freedom--Fight `Look And Feel'" are included exactly as in the
        !            19: original, and provided that the entire resulting derived work is
        !            20: distributed under the terms of a permission notice identical to this
        !            21: one.
        !            22: 
        !            23:    Permission is granted to copy and distribute translations of this
        !            24: manual into another language, under the above conditions for modified
        !            25: versions, except that the sections entitled "GNU General Public
        !            26: License" and "Protect Your Freedom--Fight `Look And Feel'", and this
        !            27: permission notice, may be included in translations approved by the Free
        !            28: Software Foundation instead of in the original English.
        !            29: 
        !            30: 
        !            31: File: gcc.info,  Node: Driver,  Next: Run-time Target,  Up: Target Macros
        !            32: 
        !            33: Controlling the Compilation Driver, `gcc'
        !            34: =========================================
        !            35: 
        !            36: `SWITCH_TAKES_ARG (CHAR)'
        !            37:      A C expression which determines whether the option `-CHAR' takes
        !            38:      arguments.  The value should be the number of arguments that
        !            39:      option takes-zero, for many options.
        !            40: 
        !            41:      By default, this macro is defined to handle the standard options
        !            42:      properly.  You need not define it unless you wish to add additional
        !            43:      options which take arguments.
        !            44: 
        !            45: `WORD_SWITCH_TAKES_ARG (NAME)'
        !            46:      A C expression which determines whether the option `-NAME' takes
        !            47:      arguments.  The value should be the number of arguments that
        !            48:      option takes-zero, for many options.  This macro rather than
        !            49:      `SWITCH_TAKES_ARG' is used for multi-character option names.
        !            50: 
        !            51:      By default, this macro is defined as
        !            52:      `DEFAULT_WORD_SWITCH_TAKES_ARG', which handles the standard options
        !            53:      properly.  You need not define `WORD_SWITCH_TAKES_ARG' unless you
        !            54:      wish to add additional options which take arguments.  Any
        !            55:      redefinition should call `DEFAULT_WORD_SWITCH_TAKES_ARG' and then
        !            56:      check for additional options.
        !            57: 
        !            58: `SWITCHES_NEED_SPACES'
        !            59:      A string-valued C expression which is nonempty if the linker needs
        !            60:      a space between the `-L' or `-o' option and its argument.
        !            61: 
        !            62:      If this macro is not defined, the default value is 0.
        !            63: 
        !            64: `CPP_SPEC'
        !            65:      A C string constant that tells the GNU CC driver program options to
        !            66:      pass to CPP.  It can also specify how to translate options you
        !            67:      give to GNU CC into options for GNU CC to pass to the CPP.
        !            68: 
        !            69:      Do not define this macro if it does not need to do anything.
        !            70: 
        !            71: `NO_BUILTIN_SIZE_TYPE'
        !            72:      If this macro is defined, the preprocessor will not define the
        !            73:      builtin macro `__SIZE_TYPE__'.  The macro `__SIZE_TYPE__' must
        !            74:      then be defined by `CPP_SPEC' instead.
        !            75: 
        !            76:      This should be defined if `SIZE_TYPE' depends on target dependent
        !            77:      flags which are not accessible to the preprocessor.  Otherwise, it
        !            78:      should not be defined.
        !            79: 
        !            80: `NO_BUILTIN_PTRDIFF_TYPE'
        !            81:      If this macro is defined, the preprocessor will not define the
        !            82:      builtin macro `__PTRDIFF_TYPE__'.  The macro `__PTRDIFF_TYPE__'
        !            83:      must then be defined by `CPP_SPEC' instead.
        !            84: 
        !            85:      This should be defined if `PTRDIFF_TYPE' depends on target
        !            86:      dependent flags which are not accessible to the preprocessor.
        !            87:      Otherwise, it should not be defined.
        !            88: 
        !            89: `SIGNED_CHAR_SPEC'
        !            90:      A C string constant that tells the GNU CC driver program options to
        !            91:      pass to CPP.  By default, this macro is defined to pass the option
        !            92:      `-D__CHAR_UNSIGNED__' to CPP if `char' will be treated as
        !            93:      `unsigned char' by `cc1'.
        !            94: 
        !            95:      Do not define this macro unless you need to override the default
        !            96:      definition.
        !            97: 
        !            98: `CC1_SPEC'
        !            99:      A C string constant that tells the GNU CC driver program options to
        !           100:      pass to `cc1'.  It can also specify how to translate options you
        !           101:      give to GNU CC into options for GNU CC to pass to the `cc1'.
        !           102: 
        !           103:      Do not define this macro if it does not need to do anything.
        !           104: 
        !           105: `CC1PLUS_SPEC'
        !           106:      A C string constant that tells the GNU CC driver program options to
        !           107:      pass to `cc1plus'.  It can also specify how to translate options
        !           108:      you give to GNU CC into options for GNU CC to pass to the
        !           109:      `cc1plus'.
        !           110: 
        !           111:      Do not define this macro if it does not need to do anything.
        !           112: 
        !           113: `ASM_SPEC'
        !           114:      A C string constant that tells the GNU CC driver program options to
        !           115:      pass to the assembler.  It can also specify how to translate
        !           116:      options you give to GNU CC into options for GNU CC to pass to the
        !           117:      assembler.  See the file `sun3.h' for an example of this.
        !           118: 
        !           119:      Do not define this macro if it does not need to do anything.
        !           120: 
        !           121: `ASM_FINAL_SPEC'
        !           122:      A C string constant that tells the GNU CC driver program how to
        !           123:      run any programs which cleanup after the normal assembler.
        !           124:      Normally, this is not needed.  See the file `mips.h' for an
        !           125:      example of this.
        !           126: 
        !           127:      Do not define this macro if it does not need to do anything.
        !           128: 
        !           129: `LINK_SPEC'
        !           130:      A C string constant that tells the GNU CC driver program options to
        !           131:      pass to the linker.  It can also specify how to translate options
        !           132:      you give to GNU CC into options for GNU CC to pass to the linker.
        !           133: 
        !           134:      Do not define this macro if it does not need to do anything.
        !           135: 
        !           136: `LIB_SPEC'
        !           137:      Another C string constant used much like `LINK_SPEC'.  The
        !           138:      difference between the two is that `LIB_SPEC' is used at the end
        !           139:      of the command given to the linker.
        !           140: 
        !           141:      If this macro is not defined, a default is provided that loads the
        !           142:      standard C library from the usual place.  See `gcc.c'.
        !           143: 
        !           144: `STARTFILE_SPEC'
        !           145:      Another C string constant used much like `LINK_SPEC'.  The
        !           146:      difference between the two is that `STARTFILE_SPEC' is used at the
        !           147:      very beginning of the command given to the linker.
        !           148: 
        !           149:      If this macro is not defined, a default is provided that loads the
        !           150:      standard C startup file from the usual place.  See `gcc.c'.
        !           151: 
        !           152: `ENDFILE_SPEC'
        !           153:      Another C string constant used much like `LINK_SPEC'.  The
        !           154:      difference between the two is that `ENDFILE_SPEC' is used at the
        !           155:      very end of the command given to the linker.
        !           156: 
        !           157:      Do not define this macro if it does not need to do anything.
        !           158: 
        !           159: `LINK_LIBGCC_SPECIAL'
        !           160:      Define this macro meaning that `gcc' should find the library
        !           161:      `libgcc.a' by hand, rather than passing the argument `-lgcc' to
        !           162:      tell the linker to do the search; also, `gcc' should not generate
        !           163:      `-L' options to pass to the linker (as it normally does).
        !           164: 
        !           165: `LINK_LIBGCC_SPECIAL_1'
        !           166:      Define this macro meaning that `gcc' should find the library
        !           167:      `libgcc.a' by hand, rather than passing the argument `-lgcc' to
        !           168:      tell the linker to do the search.
        !           169: 
        !           170: `RELATIVE_PREFIX_NOT_LINKDIR'
        !           171:      Define this macro to tell `gcc' that it should only translate a
        !           172:      `-B' prefix into a `-L' linker option if the prefix indicates an
        !           173:      absolute file name.
        !           174: 
        !           175: `STANDARD_EXEC_PREFIX'
        !           176:      Define this macro as a C string constant if you wish to override
        !           177:      the standard choice of `/usr/local/lib/gcc-lib/' as the default
        !           178:      prefix to try when searching for the executable files of the
        !           179:      compiler.
        !           180: 
        !           181: `MD_EXEC_PREFIX'
        !           182:      If defined, this macro is an additional prefix to try after
        !           183:      `STANDARD_EXEC_PREFIX'.  `MD_EXEC_PREFIX' is not searched when the
        !           184:      `-b' option is used, or the compiler is built as a cross compiler.
        !           185: 
        !           186: `STANDARD_STARTFILE_PREFIX'
        !           187:      Define this macro as a C string constant if you wish to override
        !           188:      the standard choice of `/usr/local/lib/' as the default prefix to
        !           189:      try when searching for startup files such as `crt0.o'.
        !           190: 
        !           191: `MD_STARTFILE_PREFIX'
        !           192:      If defined, this macro supplies an additional prefix to try after
        !           193:      the standard prefixes.  `MD_EXEC_PREFIX' is not searched when the
        !           194:      `-b' option is used, or when the compiler is built as a cross
        !           195:      compiler.
        !           196: 
        !           197: `MD_STARTFILE_PREFIX_1'
        !           198:      If defined, this macro supplies yet another prefix to try after the
        !           199:      standard prefixes.  It is not searched when the `-b' option is
        !           200:      used, or when the compiler is built as a cross compiler.
        !           201: 
        !           202: `LOCAL_INCLUDE_DIR'
        !           203:      Define this macro as a C string constant if you wish to override
        !           204:      the standard choice of `/usr/local/include' as the default prefix
        !           205:      to try when searching for local header files.  `LOCAL_INCLUDE_DIR'
        !           206:      comes before `SYSTEM_INCLUDE_DIR' in the search order.
        !           207: 
        !           208:      Cross compilers do not use this macro and do not search either
        !           209:      `/usr/local/include' or its replacement.
        !           210: 
        !           211: `SYSTEM_INCLUDE_DIR'
        !           212:      Define this macro as a C string constant if you wish to specify a
        !           213:      system-specific directory to search for header files before the
        !           214:      standard directory.  `SYSTEM_INCLUDE_DIR' comes before
        !           215:      `STANDARD_INCLUDE_DIR' in the search order.
        !           216: 
        !           217:      Cross compilers do not use this macro and do not search the
        !           218:      directory specified.
        !           219: 
        !           220: `STANDARD_INCLUDE_DIR'
        !           221:      Define this macro as a C string constant if you wish to override
        !           222:      the standard choice of `/usr/include' as the default prefix to try
        !           223:      when searching for header files.
        !           224: 
        !           225:      Cross compilers do not use this macro and do not search either
        !           226:      `/usr/include' or its replacement.
        !           227: 
        !           228: `INCLUDE_DEFAULTS'
        !           229:      Define this macro if you wish to override the entire default
        !           230:      search path for include files.  The default search path includes
        !           231:      `GCC_INCLUDE_DIR', `LOCAL_INCLUDE_DIR', `SYSTEM_INCLUDE_DIR',
        !           232:      `GPLUSPLUS_INCLUDE_DIR', and `STANDARD_INCLUDE_DIR'.  In addition,
        !           233:      `GPLUSPLUS_INCLUDE_DIR' and `GCC_INCLUDE_DIR' are defined
        !           234:      automatically by `Makefile', and specify private search areas for
        !           235:      GCC.  The directory `GPLUSPLUS_INCLUDE_DIR' is used only for C++
        !           236:      programs.
        !           237: 
        !           238:      The definition should be an initializer for an array of structures.
        !           239:      Each array element should have two elements: the directory name (a
        !           240:      string constant) and a flag for C++-only directories.  Mark the
        !           241:      end of the array with a null element.  For example, here is the
        !           242:      definition used for VMS:
        !           243: 
        !           244:           #define INCLUDE_DEFAULTS \
        !           245:           {                                       \
        !           246:             { "GNU_GXX_INCLUDE:", 1},             \
        !           247:             { "GNU_CC_INCLUDE:", 0},              \
        !           248:             { "SYS$SYSROOT:[SYSLIB.]", 0},        \
        !           249:             { ".", 0},                            \
        !           250:             { 0, 0}                               \
        !           251:           }
        !           252: 
        !           253:    Here is the order of prefixes tried for exec files:
        !           254: 
        !           255:   1. Any prefixes specified by the user with `-B'.
        !           256: 
        !           257:   2. The environment variable `GCC_EXEC_PREFIX', if any.
        !           258: 
        !           259:   3. The directories specified by the environment variable
        !           260:      `COMPILER_PATH'.
        !           261: 
        !           262:   4. The macro `STANDARD_EXEC_PREFIX'.
        !           263: 
        !           264:   5. `/usr/lib/gcc/'.
        !           265: 
        !           266:   6. The macro `MD_EXEC_PREFIX', if any.
        !           267: 
        !           268:    Here is the order of prefixes tried for startfiles:
        !           269: 
        !           270:   1. Any prefixes specified by the user with `-B'.
        !           271: 
        !           272:   2. The environment variable `GCC_EXEC_PREFIX', if any.
        !           273: 
        !           274:   3. The directories specified by the environment variable
        !           275:      `LIBRARY_PATH'.
        !           276: 
        !           277:   4. The macro `STANDARD_EXEC_PREFIX'.
        !           278: 
        !           279:   5. `/usr/lib/gcc/'.
        !           280: 
        !           281:   6. The macro `MD_EXEC_PREFIX', if any.
        !           282: 
        !           283:   7. The macro `MD_STARTFILE_PREFIX', if any.
        !           284: 
        !           285:   8. The macro `STANDARD_STARTFILE_PREFIX'.
        !           286: 
        !           287:   9. `/lib/'.
        !           288: 
        !           289:  10. `/usr/lib/'.
        !           290: 
        !           291: 
        !           292: File: gcc.info,  Node: Run-time Target,  Next: Storage Layout,  Prev: Driver,  Up: Target Macros
        !           293: 
        !           294: Run-time Target Specification
        !           295: =============================
        !           296: 
        !           297: `CPP_PREDEFINES'
        !           298:      Define this to be a string constant containing `-D' options to
        !           299:      define the predefined macros that identify this machine and system.
        !           300:      These macros will be predefined unless the `-ansi' option is
        !           301:      specified.
        !           302: 
        !           303:      In addition, a parallel set of macros are predefined, whose names
        !           304:      are made by appending `__' at the beginning and at the end.  These
        !           305:      `__' macros are permitted by the ANSI standard, so they are
        !           306:      predefined regardless of whether `-ansi' is specified.
        !           307: 
        !           308:      For example, on the Sun, one can use the following value:
        !           309: 
        !           310:           "-Dmc68000 -Dsun -Dunix"
        !           311: 
        !           312:      The result is to define the macros `__mc68000__', `__sun__' and
        !           313:      `__unix__' unconditionally, and the macros `mc68000', `sun' and
        !           314:      `unix' provided `-ansi' is not specified.
        !           315: 
        !           316: `STDC_VALUE'
        !           317:      Define the value to be assigned to the built-in macro `__STDC__'.
        !           318:      The default is the value `1'.
        !           319: 
        !           320: `extern int target_flags;'
        !           321:      This declaration should be present.
        !           322: 
        !           323: `TARGET_...'
        !           324:      This series of macros is to allow compiler command arguments to
        !           325:      enable or disable the use of optional features of the target
        !           326:      machine.  For example, one machine description serves both the
        !           327:      68000 and the 68020; a command argument tells the compiler whether
        !           328:      it should use 68020-only instructions or not.  This command
        !           329:      argument works by means of a macro `TARGET_68020' that tests a bit
        !           330:      in `target_flags'.
        !           331: 
        !           332:      Define a macro `TARGET_FEATURENAME' for each such option.  Its
        !           333:      definition should test a bit in `target_flags'; for example:
        !           334: 
        !           335:           #define TARGET_68020 (target_flags & 1)
        !           336: 
        !           337:      One place where these macros are used is in the
        !           338:      condition-expressions of instruction patterns.  Note how
        !           339:      `TARGET_68020' appears frequently in the 68000 machine description
        !           340:      file, `m68k.md'.  Another place they are used is in the
        !           341:      definitions of the other macros in the `MACHINE.h' file.
        !           342: 
        !           343: `TARGET_SWITCHES'
        !           344:      This macro defines names of command options to set and clear bits
        !           345:      in `target_flags'.  Its definition is an initializer with a
        !           346:      subgrouping for each command option.
        !           347: 
        !           348:      Each subgrouping contains a string constant, that defines the
        !           349:      option name, and a number, which contains the bits to set in
        !           350:      `target_flags'.  A negative number says to clear bits instead; the
        !           351:      negative of the number is which bits to clear.  The actual option
        !           352:      name is made by appending `-m' to the specified name.
        !           353: 
        !           354:      One of the subgroupings should have a null string.  The number in
        !           355:      this grouping is the default value for `target_flags'.  Any target
        !           356:      options act starting with that value.
        !           357: 
        !           358:      Here is an example which defines `-m68000' and `-m68020' with
        !           359:      opposite meanings, and picks the latter as the default:
        !           360: 
        !           361:           #define TARGET_SWITCHES \
        !           362:             { { "68020", 1},      \
        !           363:               { "68000", -1},     \
        !           364:               { "", 1}}
        !           365: 
        !           366: `TARGET_OPTIONS'
        !           367:      This macro is similar to `TARGET_SWITCHES' but defines names of
        !           368:      command options that have values.  Its definition is an
        !           369:      initializer with a subgrouping for each command option.
        !           370: 
        !           371:      Each subgrouping contains a string constant, that defines the
        !           372:      fixed part of the option name, and the address of a variable.  The
        !           373:      variable, type `char *', is set to the variable part of the given
        !           374:      option if the fixed part matches.  The actual option name is made
        !           375:      by appending `-m' to the specified name.
        !           376: 
        !           377:      Here is an example which defines `-mshort-data-NUMBER'.  If the
        !           378:      given option is `-mshort-data-512', the variable `m88k_short_data'
        !           379:      will be set to the string `"512"'.
        !           380: 
        !           381:           extern char *m88k_short_data;
        !           382:           #define TARGET_OPTIONS \
        !           383:            { { "short-data-", &m88k_short_data } }
        !           384: 
        !           385: `TARGET_VERSION'
        !           386:      This macro is a C statement to print on `stderr' a string
        !           387:      describing the particular machine description choice.  Every
        !           388:      machine description should define `TARGET_VERSION'.  For example:
        !           389: 
        !           390:           #ifdef MOTOROLA
        !           391:           #define TARGET_VERSION \
        !           392:             fprintf (stderr, " (68k, Motorola syntax)");
        !           393:           #else
        !           394:           #define TARGET_VERSION \
        !           395:             fprintf (stderr, " (68k, MIT syntax)");
        !           396:           #endif
        !           397: 
        !           398: `OVERRIDE_OPTIONS'
        !           399:      Sometimes certain combinations of command options do not make
        !           400:      sense on a particular target machine.  You can define a macro
        !           401:      `OVERRIDE_OPTIONS' to take account of this.  This macro, if
        !           402:      defined, is executed once just after all the command options have
        !           403:      been parsed.
        !           404: 
        !           405:      Don't use this macro to turn on various extra optimizations for
        !           406:      `-O'.  That is what `OPTIMIZATION_OPTIONS' is for.
        !           407: 
        !           408: `OPTIMIZATION_OPTIONS (LEVEL)'
        !           409:      Some machines may desire to change what optimizations are
        !           410:      performed for various optimization levels.   This macro, if
        !           411:      defined, is executed once just after the optimization level is
        !           412:      determined and before the remainder of the command options have
        !           413:      been parsed.  Values set in this macro are used as the default
        !           414:      values for the other command line options.
        !           415: 
        !           416:      LEVEL is the optimization level specified; 2 if -O2 is specified,
        !           417:      1 if -O is specified, and 0 if neither is specified.
        !           418: 
        !           419:      *Do not examine `write_symbols' in this macro!* The debugging
        !           420:      options are not supposed to alter the generated code.
        !           421: 
        !           422: 
        !           423: File: gcc.info,  Node: Storage Layout,  Next: Type Layout,  Prev: Run-time Target,  Up: Target Macros
        !           424: 
        !           425: Storage Layout
        !           426: ==============
        !           427: 
        !           428:    Note that the definitions of the macros in this table which are
        !           429: sizes or alignments measured in bits do not need to be constant.  They
        !           430: can be C expressions that refer to static variables, such as the
        !           431: `target_flags'.  *Note Run-time Target::.
        !           432: 
        !           433: `BITS_BIG_ENDIAN'
        !           434:      Define this macro to be the value 1 if the most significant bit in
        !           435:      a byte has the lowest number; otherwise define it to be the value
        !           436:      zero.  This means that bit-field instructions count from the most
        !           437:      significant bit.  If the machine has no bit-field instructions,
        !           438:      then this must still be defined, but it doesn't matter which value
        !           439:      it is defined to.
        !           440: 
        !           441:      This macro does not affect the way structure fields are packed into
        !           442:      bytes or words; that is controlled by `BYTES_BIG_ENDIAN'.
        !           443: 
        !           444: `BYTES_BIG_ENDIAN'
        !           445:      Define this macro to be 1 if the most significant byte in a word
        !           446:      has the lowest number.
        !           447: 
        !           448: `WORDS_BIG_ENDIAN'
        !           449:      Define this macro to be 1 if, in a multiword object, the most
        !           450:      significant word has the lowest number.  This applies to both
        !           451:      memory locations and registers; GNU CC fundamentally assumes that
        !           452:      the order of words in memory is the same as the order in registers.
        !           453: 
        !           454: `FLOAT_WORDS_BIG_ENDIAN'
        !           455:      Define this macro to be 1 if `DFmode', `XFmode' or `TFmode'
        !           456:      floating point numbers are stored in memory with the word
        !           457:      containing the sign bit at the lowest address; otherwise define it
        !           458:      to be 0.
        !           459: 
        !           460:      You need not define this macro if the ordering is the same as for
        !           461:      multi-word integers.
        !           462: 
        !           463: `BITS_PER_UNIT'
        !           464:      Define this macro to be the number of bits in an addressable
        !           465:      storage unit (byte); normally 8.
        !           466: 
        !           467: `BITS_PER_WORD'
        !           468:      Number of bits in a word; normally 32.
        !           469: 
        !           470: `MAX_BITS_PER_WORD'
        !           471:      Maximum number of bits in a word.  If this is undefined, the
        !           472:      default is `BITS_PER_WORD'.  Otherwise, it is the constant value
        !           473:      that is the largest value that `BITS_PER_WORD' can have at
        !           474:      run-time.
        !           475: 
        !           476: `UNITS_PER_WORD'
        !           477:      Number of storage units in a word; normally 4.
        !           478: 
        !           479: `MAX_UNITS_PER_WORD'
        !           480:      Maximum number of units in a word.  If this is undefined, the
        !           481:      default is `UNITS_PER_WORD'.  Otherwise, it is the constant value
        !           482:      that is the largest value that `UNITS_PER_WORD' can have at
        !           483:      run-time.
        !           484: 
        !           485: `POINTER_SIZE'
        !           486:      Width of a pointer, in bits.
        !           487: 
        !           488: `PROMOTE_MODE (M, UNSIGNEDP, TYPE)'
        !           489:      A macro to update M and UNSIGNEDP when an object whose type is
        !           490:      TYPE and which has the specified mode and signedness is to be
        !           491:      stored in a register.  This macro is only called when TYPE is a
        !           492:      scalar type.
        !           493: 
        !           494:      On most RISC machines, which only have operations that operate on
        !           495:      a full register, define this macro to set M to `word_mode' if M is
        !           496:      an integer mode narrower than `BITS_PER_WORD'.  In most cases,
        !           497:      only integer modes should be widened because wider-precision
        !           498:      floating-point operations are usually more expensive than their
        !           499:      narrower counterparts.
        !           500: 
        !           501:      For most machines, the macro definition does not change UNSIGNEDP.
        !           502:      However, some machines, have instructions that preferentially
        !           503:      handle either signed or unsigned quantities of certain modes.  For
        !           504:      example, on the DEC Alpha, 32-bit loads from memory and 32-bit add
        !           505:      instructions sign-extend the result to 64 bits.  On such machines,
        !           506:      set UNSIGNEDP according to which kind of extension is more
        !           507:      efficient.
        !           508: 
        !           509:      Do not define this macro if it would never modify M.
        !           510: 
        !           511: `PROMOTE_FUNCTION_ARGS'
        !           512:      Define this macro if the promotion described by `PROMOTE_MODE'
        !           513:      should also be done for outgoing function arguments.
        !           514: 
        !           515: `PROMOTE_FUNCTION_RETURN'
        !           516:      Define this macro if the promotion described by `PROMOTE_MODE'
        !           517:      should also be done for the return value of functions.
        !           518: 
        !           519:      If this macro is defined, `FUNCTION_VALUE' must perform the same
        !           520:      promotions done by `PROMOTE_MODE'.
        !           521: 
        !           522: `PARM_BOUNDARY'
        !           523:      Normal alignment required for function parameters on the stack, in
        !           524:      bits.  All stack parameters receive at least this much alignment
        !           525:      regardless of data type.  On most machines, this is the same as the
        !           526:      size of an integer.
        !           527: 
        !           528: `STACK_BOUNDARY'
        !           529:      Define this macro if you wish to preserve a certain alignment for
        !           530:      the stack pointer.  The definition is a C expression for the
        !           531:      desired alignment (measured in bits).
        !           532: 
        !           533:      If `PUSH_ROUNDING' is not defined, the stack will always be aligned
        !           534:      to the specified boundary.  If `PUSH_ROUNDING' is defined and
        !           535:      specifies a less strict alignment than `STACK_BOUNDARY', the stack
        !           536:      may be momentarily unaligned while pushing arguments.
        !           537: 
        !           538: `FUNCTION_BOUNDARY'
        !           539:      Alignment required for a function entry point, in bits.
        !           540: 
        !           541: `BIGGEST_ALIGNMENT'
        !           542:      Biggest alignment that any data type can require on this machine,
        !           543:      in bits.
        !           544: 
        !           545: `BIGGEST_FIELD_ALIGNMENT'
        !           546:      Biggest alignment that any structure field can require on this
        !           547:      machine, in bits.  If defined, this overrides `BIGGEST_ALIGNMENT'
        !           548:      for structure fields only.
        !           549: 
        !           550: `MAX_OFILE_ALIGNMENT'
        !           551:      Biggest alignment supported by the object file format of this
        !           552:      machine.  Use this macro to limit the alignment which can be
        !           553:      specified using the `__attribute__ ((aligned (N)))' construct.  If
        !           554:      not defined, the default value is `BIGGEST_ALIGNMENT'.
        !           555: 
        !           556: `DATA_ALIGNMENT (TYPE, BASIC-ALIGN)'
        !           557:      If defined, a C expression to compute the alignment for a static
        !           558:      variable.  TYPE is the data type, and BASIC-ALIGN is the alignment
        !           559:      that the object would ordinarily have.  The value of this macro is
        !           560:      used instead of that alignment to align the object.
        !           561: 
        !           562:      If this macro is not defined, then BASIC-ALIGN is used.
        !           563: 
        !           564:      One use of this macro is to increase alignment of medium-size data
        !           565:      to make it all fit in fewer cache lines.  Another is to cause
        !           566:      character arrays to be word-aligned so that `strcpy' calls that
        !           567:      copy constants to character arrays can be done inline.
        !           568: 
        !           569: `CONSTANT_ALIGNMENT (CONSTANT, BASIC-ALIGN)'
        !           570:      If defined, a C expression to compute the alignment given to a
        !           571:      constant that is being placed in memory.  CONSTANT is the constant
        !           572:      and BASIC-ALIGN is the alignment that the object would ordinarily
        !           573:      have.  The value of this macro is used instead of that alignment to
        !           574:      align the object.
        !           575: 
        !           576:      If this macro is not defined, then BASIC-ALIGN is used.
        !           577: 
        !           578:      The typical use of this macro is to increase alignment for string
        !           579:      constants to be word aligned so that `strcpy' calls that copy
        !           580:      constants can be done inline.
        !           581: 
        !           582: `EMPTY_FIELD_BOUNDARY'
        !           583:      Alignment in bits to be given to a structure bit field that
        !           584:      follows an empty field such as `int : 0;'.
        !           585: 
        !           586:      Note that `PCC_BITFIELD_TYPE_MATTERS' also affects the alignment
        !           587:      that results from an empty field.
        !           588: 
        !           589: `STRUCTURE_SIZE_BOUNDARY'
        !           590:      Number of bits which any structure or union's size must be a
        !           591:      multiple of.  Each structure or union's size is rounded up to a
        !           592:      multiple of this.
        !           593: 
        !           594:      If you do not define this macro, the default is the same as
        !           595:      `BITS_PER_UNIT'.
        !           596: 
        !           597: `STRICT_ALIGNMENT'
        !           598:      Define this macro to be the value 1 if instructions will fail to
        !           599:      work if given data not on the nominal alignment.  If instructions
        !           600:      will merely go slower in that case, define this macro as 0.
        !           601: 
        !           602: `PCC_BITFIELD_TYPE_MATTERS'
        !           603:      Define this if you wish to imitate the way many other C compilers
        !           604:      handle alignment of bitfields and the structures that contain them.
        !           605: 
        !           606:      The behavior is that the type written for a bitfield (`int',
        !           607:      `short', or other integer type) imposes an alignment for the
        !           608:      entire structure, as if the structure really did contain an
        !           609:      ordinary field of that type.  In addition, the bitfield is placed
        !           610:      within the structure so that it would fit within such a field, not
        !           611:      crossing a boundary for it.
        !           612: 
        !           613:      Thus, on most machines, a bitfield whose type is written as `int'
        !           614:      would not cross a four-byte boundary, and would force four-byte
        !           615:      alignment for the whole structure.  (The alignment used may not be
        !           616:      four bytes; it is controlled by the other alignment parameters.)
        !           617: 
        !           618:      If the macro is defined, its definition should be a C expression;
        !           619:      a nonzero value for the expression enables this behavior.
        !           620: 
        !           621:      Note that if this macro is not defined, or its value is zero, some
        !           622:      bitfields may cross more than one alignment boundary.  The
        !           623:      compiler can support such references if there are `insv', `extv',
        !           624:      and `extzv' insns that can directly reference memory.
        !           625: 
        !           626:      The other known way of making bitfields work is to define
        !           627:      `STRUCTURE_SIZE_BOUNDARY' as large as `BIGGEST_ALIGNMENT'.  Then
        !           628:      every structure can be accessed with fullwords.
        !           629: 
        !           630:      Unless the machine has bitfield instructions or you define
        !           631:      `STRUCTURE_SIZE_BOUNDARY' that way, you must define
        !           632:      `PCC_BITFIELD_TYPE_MATTERS' to have a nonzero value.
        !           633: 
        !           634:      If your aim is to make GNU CC use the same conventions for laying
        !           635:      out bitfields as are used by another compiler, here is how to
        !           636:      investigate what the other compiler does.  Compile and run this
        !           637:      program:
        !           638: 
        !           639:           struct foo1
        !           640:           {
        !           641:             char x;
        !           642:             char :0;
        !           643:             char y;
        !           644:           };
        !           645:           
        !           646:           struct foo2
        !           647:           {
        !           648:             char x;
        !           649:             int :0;
        !           650:             char y;
        !           651:           };
        !           652:           
        !           653:           main ()
        !           654:           {
        !           655:             printf ("Size of foo1 is %d\n",
        !           656:                     sizeof (struct foo1));
        !           657:             printf ("Size of foo2 is %d\n",
        !           658:                     sizeof (struct foo2));
        !           659:             exit (0);
        !           660:           }
        !           661: 
        !           662:      If this prints 2 and 5, then the compiler's behavior is what you
        !           663:      would get from `PCC_BITFIELD_TYPE_MATTERS'.
        !           664: 
        !           665: `BITFIELD_NBYTES_LIMITED'
        !           666:      Like PCC_BITFIELD_TYPE_MATTERS except that its effect is limited to
        !           667:      aligning a bitfield within the structure.
        !           668: 
        !           669: `ROUND_TYPE_SIZE (STRUCT, SIZE, ALIGN)'
        !           670:      Define this macro as an expression for the overall size of a
        !           671:      structure (given by STRUCT as a tree node) when the size computed
        !           672:      from the fields is SIZE and the alignment is ALIGN.
        !           673: 
        !           674:      The default is to round SIZE up to a multiple of ALIGN.
        !           675: 
        !           676: `ROUND_TYPE_ALIGN (STRUCT, COMPUTED, SPECIFIED)'
        !           677:      Define this macro as an expression for the alignment of a structure
        !           678:      (given by STRUCT as a tree node) if the alignment computed in the
        !           679:      usual way is COMPUTED and the alignment explicitly specified was
        !           680:      SPECIFIED.
        !           681: 
        !           682:      The default is to use SPECIFIED if it is larger; otherwise, use
        !           683:      the smaller of COMPUTED and `BIGGEST_ALIGNMENT'
        !           684: 
        !           685: `MAX_FIXED_MODE_SIZE'
        !           686:      An integer expression for the size in bits of the largest integer
        !           687:      machine mode that should actually be used.  All integer machine
        !           688:      modes of this size or smaller can be used for structures and
        !           689:      unions with the appropriate sizes.  If this macro is undefined,
        !           690:      `GET_MODE_BITSIZE (DImode)' is assumed.
        !           691: 
        !           692: `CHECK_FLOAT_VALUE (MODE, VALUE)'
        !           693:      A C statement to validate the value VALUE (of type `double') for
        !           694:      mode MODE.  This means that you check whether VALUE fits within
        !           695:      the possible range of values for mode MODE on this target machine.
        !           696:      The mode MODE is always `SFmode' or `DFmode'.
        !           697: 
        !           698:      If VALUE is not valid, you should call `error' to print an error
        !           699:      message and then assign some valid value to VALUE.  Allowing an
        !           700:      invalid value to go through the compiler can produce incorrect
        !           701:      assembler code which may even cause Unix assemblers to crash.
        !           702: 
        !           703:      This macro need not be defined if there is no work for it to do.
        !           704: 
        !           705: `TARGET_FLOAT_FORMAT'
        !           706:      A code distinguishing the floating point format of the target
        !           707:      machine.  There are three defined values:
        !           708: 
        !           709:     `IEEE_FLOAT_FORMAT'
        !           710:           This code indicates IEEE floating point.  It is the default;
        !           711:           there is no need to define this macro when the format is IEEE.
        !           712: 
        !           713:     `VAX_FLOAT_FORMAT'
        !           714:           This code indicates the peculiar format used on the Vax.
        !           715: 
        !           716:     `UNKNOWN_FLOAT_FORMAT'
        !           717:           This code indicates any other format.
        !           718: 
        !           719:      The value of this macro is compared with `HOST_FLOAT_FORMAT'
        !           720:      (*note Config::.) to determine whether the target machine has the
        !           721:      same format as the host machine.  If any other formats are
        !           722:      actually in use on supported machines, new codes should be defined
        !           723:      for them.
        !           724: 
        !           725:      The ordering of the component words of floating point values
        !           726:      stored in memory is controlled by `FLOAT_WORDS_BIG_ENDIAN' for the
        !           727:      target machine and `HOST_FLOAT_WORDS_BIG_ENDIAN' for the host.
        !           728: 
        !           729: 
        !           730: File: gcc.info,  Node: Type Layout,  Next: Registers,  Prev: Storage Layout,  Up: Target Macros
        !           731: 
        !           732: Layout of Source Language Data Types
        !           733: ====================================
        !           734: 
        !           735:    These macros define the sizes and other characteristics of the
        !           736: standard basic data types used in programs being compiled.  Unlike the
        !           737: macros in the previous section, these apply to specific features of C
        !           738: and related languages, rather than to fundamental aspects of storage
        !           739: layout.
        !           740: 
        !           741: `INT_TYPE_SIZE'
        !           742:      A C expression for the size in bits of the type `int' on the
        !           743:      target machine.  If you don't define this, the default is one word.
        !           744: 
        !           745: `MAX_INT_TYPE_SIZE'
        !           746:      Maximum number for the size in bits of the type `int' on the target
        !           747:      machine.  If this is undefined, the default is `INT_TYPE_SIZE'.
        !           748:      Otherwise, it is the constant value that is the largest value that
        !           749:      `INT_TYPE_SIZE' can have at run-time.  This is used in `cpp'.
        !           750: 
        !           751: `SHORT_TYPE_SIZE'
        !           752:      A C expression for the size in bits of the type `short' on the
        !           753:      target machine.  If you don't define this, the default is half a
        !           754:      word.  (If this would be less than one storage unit, it is rounded
        !           755:      up to one unit.)
        !           756: 
        !           757: `LONG_TYPE_SIZE'
        !           758:      A C expression for the size in bits of the type `long' on the
        !           759:      target machine.  If you don't define this, the default is one word.
        !           760: 
        !           761: `MAX_LONG_TYPE_SIZE'
        !           762:      Maximum number for the size in bits of the type `long' on the
        !           763:      target machine.  If this is undefined, the default is
        !           764:      `LONG_TYPE_SIZE'.  Otherwise, it is the constant value that is the
        !           765:      largest value that `LONG_TYPE_SIZE' can have at run-time.  This is
        !           766:      used in `cpp'.
        !           767: 
        !           768: `LONG_LONG_TYPE_SIZE'
        !           769:      A C expression for the size in bits of the type `long long' on the
        !           770:      target machine.  If you don't define this, the default is two
        !           771:      words.
        !           772: 
        !           773: `CHAR_TYPE_SIZE'
        !           774:      A C expression for the size in bits of the type `char' on the
        !           775:      target machine.  If you don't define this, the default is one
        !           776:      quarter of a word.  (If this would be less than one storage unit,
        !           777:      it is rounded up to one unit.)
        !           778: 
        !           779: `MAX_CHAR_TYPE_SIZE'
        !           780:      Maximum number for the size in bits of the type `char' on the
        !           781:      target machine.  If this is undefined, the default is
        !           782:      `CHAR_TYPE_SIZE'.  Otherwise, it is the constant value that is the
        !           783:      largest value that `CHAR_TYPE_SIZE' can have at run-time.  This is
        !           784:      used in `cpp'.
        !           785: 
        !           786: `FLOAT_TYPE_SIZE'
        !           787:      A C expression for the size in bits of the type `float' on the
        !           788:      target machine.  If you don't define this, the default is one word.
        !           789: 
        !           790: `DOUBLE_TYPE_SIZE'
        !           791:      A C expression for the size in bits of the type `double' on the
        !           792:      target machine.  If you don't define this, the default is two
        !           793:      words.
        !           794: 
        !           795: `LONG_DOUBLE_TYPE_SIZE'
        !           796:      A C expression for the size in bits of the type `long double' on
        !           797:      the target machine.  If you don't define this, the default is two
        !           798:      words.
        !           799: 
        !           800: `DEFAULT_SIGNED_CHAR'
        !           801:      An expression whose value is 1 or 0, according to whether the type
        !           802:      `char' should be signed or unsigned by default.  The user can
        !           803:      always override this default with the options `-fsigned-char' and
        !           804:      `-funsigned-char'.
        !           805: 
        !           806: `DEFAULT_SHORT_ENUMS'
        !           807:      A C expression to determine whether to give an `enum' type only as
        !           808:      many bytes as it takes to represent the range of possible values
        !           809:      of that type.  A nonzero value means to do that; a zero value
        !           810:      means all `enum' types should be allocated like `int'.
        !           811: 
        !           812:      If you don't define the macro, the default is 0.
        !           813: 
        !           814: `SIZE_TYPE'
        !           815:      A C expression for a string describing the name of the data type
        !           816:      to use for size values.  The typedef name `size_t' is defined
        !           817:      using the contents of the string.
        !           818: 
        !           819:      The string can contain more than one keyword.  If so, separate
        !           820:      them with spaces, and write first any length keyword, then
        !           821:      `unsigned' if appropriate, and finally `int'.  The string must
        !           822:      exactly match one of the data type names defined in the function
        !           823:      `init_decl_processing' in the file `c-decl.c'.  You may not omit
        !           824:      `int' or change the order--that would cause the compiler to crash
        !           825:      on startup.
        !           826: 
        !           827:      If you don't define this macro, the default is `"long unsigned
        !           828:      int"'.
        !           829: 
        !           830: `PTRDIFF_TYPE'
        !           831:      A C expression for a string describing the name of the data type
        !           832:      to use for the result of subtracting two pointers.  The typedef
        !           833:      name `ptrdiff_t' is defined using the contents of the string.  See
        !           834:      `SIZE_TYPE' above for more information.
        !           835: 
        !           836:      If you don't define this macro, the default is `"long int"'.
        !           837: 
        !           838: `WCHAR_TYPE'
        !           839:      A C expression for a string describing the name of the data type
        !           840:      to use for wide characters.  The typedef name `wchar_t' is defined
        !           841:      using the contents of the string.  See `SIZE_TYPE' above for more
        !           842:      information.
        !           843: 
        !           844:      If you don't define this macro, the default is `"int"'.
        !           845: 
        !           846: `WCHAR_TYPE_SIZE'
        !           847:      A C expression for the size in bits of the data type for wide
        !           848:      characters.  This is used in `cpp', which cannot make use of
        !           849:      `WCHAR_TYPE'.
        !           850: 
        !           851: `MAX_WCHAR_TYPE_SIZE'
        !           852:      Maximum number for the size in bits of the data type for wide
        !           853:      characters.  If this is undefined, the default is
        !           854:      `WCHAR_TYPE_SIZE'.  Otherwise, it is the constant value that is the
        !           855:      largest value that `WCHAR_TYPE_SIZE' can have at run-time.  This is
        !           856:      used in `cpp'.
        !           857: 
        !           858: `OBJC_INT_SELECTORS'
        !           859:      Define this macro if the type of Objective C selectors should be
        !           860:      `int'.
        !           861: 
        !           862:      If this macro is not defined, then selectors should have the type
        !           863:      `struct objc_selector *'.
        !           864: 
        !           865: `OBJC_SELECTORS_WITHOUT_LABELS'
        !           866:      Define this macro if the compiler can group all the selectors
        !           867:      together into a vector and use just one label at the beginning of
        !           868:      the vector.  Otherwise, the compiler must give each selector its
        !           869:      own assembler label.
        !           870: 
        !           871:      On certain machines, it is important to have a separate label for
        !           872:      each selector because this enables the linker to eliminate
        !           873:      duplicate selectors.
        !           874: 
        !           875: `TARGET_BELL'
        !           876:      A C constant expression for the integer value for escape sequence
        !           877:      `\a'.
        !           878: 
        !           879: `TARGET_BS'
        !           880: `TARGET_TAB'
        !           881: `TARGET_NEWLINE'
        !           882:      C constant expressions for the integer values for escape sequences
        !           883:      `\b', `\t' and `\n'.
        !           884: 
        !           885: `TARGET_VT'
        !           886: `TARGET_FF'
        !           887: `TARGET_CR'
        !           888:      C constant expressions for the integer values for escape sequences
        !           889:      `\v', `\f' and `\r'.
        !           890: 
        !           891: 
        !           892: File: gcc.info,  Node: Registers,  Next: Register Classes,  Prev: Type Layout,  Up: Target Macros
        !           893: 
        !           894: Register Usage
        !           895: ==============
        !           896: 
        !           897:    This section explains how to describe what registers the target
        !           898: machine has, and how (in general) they can be used.
        !           899: 
        !           900:    The description of which registers a specific instruction can use is
        !           901: done with register classes; see *Note Register Classes::.  For
        !           902: information on using registers to access a stack frame, see *Note Frame
        !           903: Registers::.  For passing values in registers, see *Note Register
        !           904: Arguments::.  For returning values in registers, see *Note Scalar
        !           905: Return::.
        !           906: 
        !           907: * Menu:
        !           908: 
        !           909: * Register Basics::            Number and kinds of registers.
        !           910: * Allocation Order::           Order in which registers are allocated.
        !           911: * Values in Registers::                What kinds of values each reg can hold.
        !           912: * Leaf Functions::             Renumbering registers for leaf functions.
        !           913: * Stack Registers::            Handling a register stack such as 80387.
        !           914: * Obsolete Register Macros::   Macros formerly used for the 80387.
        !           915: 
        !           916: 
        !           917: File: gcc.info,  Node: Register Basics,  Next: Allocation Order,  Up: Registers
        !           918: 
        !           919: Basic Characteristics of Registers
        !           920: ----------------------------------
        !           921: 
        !           922: `FIRST_PSEUDO_REGISTER'
        !           923:      Number of hardware registers known to the compiler.  They receive
        !           924:      numbers 0 through `FIRST_PSEUDO_REGISTER-1'; thus, the first
        !           925:      pseudo register's number really is assigned the number
        !           926:      `FIRST_PSEUDO_REGISTER'.
        !           927: 
        !           928: `FIXED_REGISTERS'
        !           929:      An initializer that says which registers are used for fixed
        !           930:      purposes all throughout the compiled code and are therefore not
        !           931:      available for general allocation.  These would include the stack
        !           932:      pointer, the frame pointer (except on machines where that can be
        !           933:      used as a general register when no frame pointer is needed), the
        !           934:      program counter on machines where that is considered one of the
        !           935:      addressable registers, and any other numbered register with a
        !           936:      standard use.
        !           937: 
        !           938:      This information is expressed as a sequence of numbers, separated
        !           939:      by commas and surrounded by braces.  The Nth number is 1 if
        !           940:      register N is fixed, 0 otherwise.
        !           941: 
        !           942:      The table initialized from this macro, and the table initialized by
        !           943:      the following one, may be overridden at run time either
        !           944:      automatically, by the actions of the macro
        !           945:      `CONDITIONAL_REGISTER_USAGE', or by the user with the command
        !           946:      options `-ffixed-REG', `-fcall-used-REG' and `-fcall-saved-REG'.
        !           947: 
        !           948: `CALL_USED_REGISTERS'
        !           949:      Like `FIXED_REGISTERS' but has 1 for each register that is
        !           950:      clobbered (in general) by function calls as well as for fixed
        !           951:      registers.  This macro therefore identifies the registers that are
        !           952:      not available for general allocation of values that must live
        !           953:      across function calls.
        !           954: 
        !           955:      If a register has 0 in `CALL_USED_REGISTERS', the compiler
        !           956:      automatically saves it on function entry and restores it on
        !           957:      function exit, if the register is used within the function.
        !           958: 
        !           959: `CONDITIONAL_REGISTER_USAGE'
        !           960:      Zero or more C statements that may conditionally modify two
        !           961:      variables `fixed_regs' and `call_used_regs' (both of type `char
        !           962:      []') after they have been initialized from the two preceding
        !           963:      macros.
        !           964: 
        !           965:      This is necessary in case the fixed or call-clobbered registers
        !           966:      depend on target flags.
        !           967: 
        !           968:      You need not define this macro if it has no work to do.
        !           969: 
        !           970:      If the usage of an entire class of registers depends on the target
        !           971:      flags, you may indicate this to GCC by using this macro to modify
        !           972:      `fixed_regs' and `call_used_regs' to 1 for each of the registers
        !           973:      in the classes which should not be used by GCC.  Also define the
        !           974:      macro `REG_CLASS_FROM_LETTER' to return `NO_REGS' if it is called
        !           975:      with a letter for a class that shouldn't be used.
        !           976: 
        !           977:      (However, if this class is not included in `GENERAL_REGS' and all
        !           978:      of the insn patterns whose constraints permit this class are
        !           979:      controlled by target switches, then GCC will automatically avoid
        !           980:      using these registers when the target switches are opposed to
        !           981:      them.)
        !           982: 
        !           983: `NON_SAVING_SETJMP'
        !           984:      If this macro is defined and has a nonzero value, it means that
        !           985:      `setjmp' and related functions fail to save the registers, or that
        !           986:      `longjmp' fails to restore them.  To compensate, the compiler
        !           987:      avoids putting variables in registers in functions that use
        !           988:      `setjmp'.
        !           989: 
        !           990: `INCOMING_REGNO (OUT)'
        !           991:      Define this macro if the target machine has register windows.
        !           992:      This C expression returns the register number as seen by the
        !           993:      called function corresponding to the register number OUT as seen
        !           994:      by the calling function.  Return OUT if register number OUT is not
        !           995:      an outbound register.
        !           996: 
        !           997: `OUTGOING_REGNO (IN)'
        !           998:      Define this macro if the target machine has register windows.
        !           999:      This C expression returns the register number as seen by the
        !          1000:      calling function corresponding to the register number IN as seen
        !          1001:      by the called function.  Return IN if register number IN is not an
        !          1002:      inbound register.
        !          1003: 
        !          1004: 
        !          1005: File: gcc.info,  Node: Allocation Order,  Next: Values in Registers,  Prev: Register Basics,  Up: Registers
        !          1006: 
        !          1007: Order of Allocation of Registers
        !          1008: --------------------------------
        !          1009: 
        !          1010: `REG_ALLOC_ORDER'
        !          1011:      If defined, an initializer for a vector of integers, containing the
        !          1012:      numbers of hard registers in the order in which GNU CC should
        !          1013:      prefer to use them (from most preferred to least).
        !          1014: 
        !          1015:      If this macro is not defined, registers are used lowest numbered
        !          1016:      first (all else being equal).
        !          1017: 
        !          1018:      One use of this macro is on machines where the highest numbered
        !          1019:      registers must always be saved and the save-multiple-registers
        !          1020:      instruction supports only sequences of consecutive registers.  On
        !          1021:      such machines, define `REG_ALLOC_ORDER' to be an initializer that
        !          1022:      lists the highest numbered allocatable register first.
        !          1023: 
        !          1024: `ORDER_REGS_FOR_LOCAL_ALLOC'
        !          1025:      A C statement (sans semicolon) to choose the order in which to
        !          1026:      allocate hard registers for pseudo-registers local to a basic
        !          1027:      block.
        !          1028: 
        !          1029:      Store the desired register order in the array `reg_alloc_order'.
        !          1030:      Element 0 should be the register to allocate first; element 1, the
        !          1031:      next register; and so on.
        !          1032: 
        !          1033:      The macro body should not assume anything about the contents of
        !          1034:      `reg_alloc_order' before execution of the macro.
        !          1035: 
        !          1036:      On most machines, it is not necessary to define this macro.
        !          1037: 
        !          1038: 
        !          1039: File: gcc.info,  Node: Values in Registers,  Next: Leaf Functions,  Prev: Allocation Order,  Up: Registers
        !          1040: 
        !          1041: How Values Fit in Registers
        !          1042: ---------------------------
        !          1043: 
        !          1044:    This section discusses the macros that describe which kinds of values
        !          1045: (specifically, which machine modes) each register can hold, and how many
        !          1046: consecutive registers are needed for a given mode.
        !          1047: 
        !          1048: `HARD_REGNO_NREGS (REGNO, MODE)'
        !          1049:      A C expression for the number of consecutive hard registers,
        !          1050:      starting at register number REGNO, required to hold a value of mode
        !          1051:      MODE.
        !          1052: 
        !          1053:      On a machine where all registers are exactly one word, a suitable
        !          1054:      definition of this macro is
        !          1055: 
        !          1056:           #define HARD_REGNO_NREGS(REGNO, MODE)            \
        !          1057:              ((GET_MODE_SIZE (MODE) + UNITS_PER_WORD - 1)  \
        !          1058:               / UNITS_PER_WORD))
        !          1059: 
        !          1060: `HARD_REGNO_MODE_OK (REGNO, MODE)'
        !          1061:      A C expression that is nonzero if it is permissible to store a
        !          1062:      value of mode MODE in hard register number REGNO (or in several
        !          1063:      registers starting with that one).  For a machine where all
        !          1064:      registers are equivalent, a suitable definition is
        !          1065: 
        !          1066:           #define HARD_REGNO_MODE_OK(REGNO, MODE) 1
        !          1067: 
        !          1068:      It is not necessary for this macro to check for the numbers of
        !          1069:      fixed registers, because the allocation mechanism considers them
        !          1070:      to be always occupied.
        !          1071: 
        !          1072:      On some machines, double-precision values must be kept in even/odd
        !          1073:      register pairs.  The way to implement that is to define this macro
        !          1074:      to reject odd register numbers for such modes.
        !          1075: 
        !          1076:      The minimum requirement for a mode to be OK in a register is that
        !          1077:      the `movMODE' instruction pattern support moves between the
        !          1078:      register and any other hard register for which the mode is OK; and
        !          1079:      that moving a value into the register and back out not alter it.
        !          1080: 
        !          1081:      Since the same instruction used to move `SImode' will work for all
        !          1082:      narrower integer modes, it is not necessary on any machine for
        !          1083:      `HARD_REGNO_MODE_OK' to distinguish between these modes, provided
        !          1084:      you define patterns `movhi', etc., to take advantage of this.  This
        !          1085:      is useful because of the interaction between `HARD_REGNO_MODE_OK'
        !          1086:      and `MODES_TIEABLE_P'; it is very desirable for all integer modes
        !          1087:      to be tieable.
        !          1088: 
        !          1089:      Many machines have special registers for floating point arithmetic.
        !          1090:      Often people assume that floating point machine modes are allowed
        !          1091:      only in floating point registers.  This is not true.  Any
        !          1092:      registers that can hold integers can safely *hold* a floating
        !          1093:      point machine mode, whether or not floating arithmetic can be done
        !          1094:      on it in those registers.  Integer move instructions can be used
        !          1095:      to move the values.
        !          1096: 
        !          1097:      On some machines, though, the converse is true: fixed-point machine
        !          1098:      modes may not go in floating registers.  This is true if the
        !          1099:      floating registers normalize any value stored in them, because
        !          1100:      storing a non-floating value there would garble it.  In this case,
        !          1101:      `HARD_REGNO_MODE_OK' should reject fixed-point machine modes in
        !          1102:      floating registers.  But if the floating registers do not
        !          1103:      automatically normalize, if you can store any bit pattern in one
        !          1104:      and retrieve it unchanged without a trap, then any machine mode
        !          1105:      may go in a floating register, so you can define this macro to say
        !          1106:      so.
        !          1107: 
        !          1108:      On some machines, such as the Sparc and the Mips, we get better
        !          1109:      code by defining `HARD_REGNO_MODE_OK' to forbid integers in
        !          1110:      floating registers, even though the hardware is capable of
        !          1111:      handling them.  This is because transferring values between
        !          1112:      floating registers and general registers is so slow that it is
        !          1113:      better to keep the integer in memory.
        !          1114: 
        !          1115:      The primary significance of special floating registers is rather
        !          1116:      that they are the registers acceptable in floating point arithmetic
        !          1117:      instructions.  However, this is of no concern to
        !          1118:      `HARD_REGNO_MODE_OK'.  You handle it by writing the proper
        !          1119:      constraints for those instructions.
        !          1120: 
        !          1121:      On some machines, the floating registers are especially slow to
        !          1122:      access, so that it is better to store a value in a stack frame
        !          1123:      than in such a register if floating point arithmetic is not being
        !          1124:      done.  As long as the floating registers are not in class
        !          1125:      `GENERAL_REGS', they will not be used unless some pattern's
        !          1126:      constraint asks for one.
        !          1127: 
        !          1128: `MODES_TIEABLE_P (MODE1, MODE2)'
        !          1129:      A C expression that is nonzero if it is desirable to choose
        !          1130:      register allocation so as to avoid move instructions between a
        !          1131:      value of mode MODE1 and a value of mode MODE2.
        !          1132: 
        !          1133:      If `HARD_REGNO_MODE_OK (R, MODE1)' and `HARD_REGNO_MODE_OK (R,
        !          1134:      MODE2)' are ever different for any R, then `MODES_TIEABLE_P (MODE1,
        !          1135:      MODE2)' must be zero.
        !          1136: 
        !          1137: 
        !          1138: File: gcc.info,  Node: Leaf Functions,  Next: Stack Registers,  Prev: Values in Registers,  Up: Registers
        !          1139: 
        !          1140: Handling Leaf Functions
        !          1141: -----------------------
        !          1142: 
        !          1143:    On some machines, a leaf function (i.e., one which makes no calls)
        !          1144: can run more efficiently if it does not make its own register window.
        !          1145: Often this means it is required to receive its arguments in the
        !          1146: registers where they are passed by the caller, instead of the registers
        !          1147: where they would normally arrive.
        !          1148: 
        !          1149:    The special treatment for leaf functions generally applies only when
        !          1150: other conditions are met; for example, often they may use only those
        !          1151: registers for its own variables and temporaries.  We use the term "leaf
        !          1152: function" to mean a function that is suitable for this special
        !          1153: handling, so that functions with no calls are not necessarily "leaf
        !          1154: functions".
        !          1155: 
        !          1156:    GNU CC assigns register numbers before it knows whether the function
        !          1157: is suitable for leaf function treatment.  So it needs to renumber the
        !          1158: registers in order to output a leaf function.  The following macros
        !          1159: accomplish this.
        !          1160: 
        !          1161: `LEAF_REGISTERS'
        !          1162:      A C initializer for a vector, indexed by hard register number,
        !          1163:      which contains 1 for a register that is allowable in a candidate
        !          1164:      for leaf function treatment.
        !          1165: 
        !          1166:      If leaf function treatment involves renumbering the registers,
        !          1167:      then the registers marked here should be the ones before
        !          1168:      renumbering--those that GNU CC would ordinarily allocate.  The
        !          1169:      registers which will actually be used in the assembler code, after
        !          1170:      renumbering, should not be marked with 1 in this vector.
        !          1171: 
        !          1172:      Define this macro only if the target machine offers a way to
        !          1173:      optimize the treatment of leaf functions.
        !          1174: 
        !          1175: `LEAF_REG_REMAP (REGNO)'
        !          1176:      A C expression whose value is the register number to which REGNO
        !          1177:      should be renumbered, when a function is treated as a leaf
        !          1178:      function.
        !          1179: 
        !          1180:      If REGNO is a register number which should not appear in a leaf
        !          1181:      function before renumbering, then the expression should yield -1,
        !          1182:      which will cause the compiler to abort.
        !          1183: 
        !          1184:      Define this macro only if the target machine offers a way to
        !          1185:      optimize the treatment of leaf functions, and registers need to be
        !          1186:      renumbered to do this.
        !          1187: 
        !          1188: `REG_LEAF_ALLOC_ORDER'
        !          1189:      If defined, an initializer for a vector of integers, containing the
        !          1190:      numbers of hard registers in the order in which the GNU CC should
        !          1191:      prefer to use them (from most preferred to least) in a leaf
        !          1192:      function.  If this macro is not defined, REG_ALLOC_ORDER is used
        !          1193:      for both non-leaf and leaf-functions.
        !          1194: 
        !          1195:    Normally, `FUNCTION_PROLOGUE' and `FUNCTION_EPILOGUE' must treat
        !          1196: leaf functions specially.  It can test the C variable `leaf_function'
        !          1197: which is nonzero for leaf functions.  (The variable `leaf_function' is
        !          1198: defined only if `LEAF_REGISTERS' is defined.)
        !          1199: 
        !          1200: 
        !          1201: File: gcc.info,  Node: Stack Registers,  Next: Obsolete Register Macros,  Prev: Leaf Functions,  Up: Registers
        !          1202: 
        !          1203: Registers That Form a Stack
        !          1204: ---------------------------
        !          1205: 
        !          1206:    There are special features to handle computers where some of the
        !          1207: "registers" form a stack, as in the 80387 coprocessor for the 80386.
        !          1208: Stack registers are normally written by pushing onto the stack, and are
        !          1209: numbered relative to the top of the stack.
        !          1210: 
        !          1211:    Currently, GNU CC can only handle one group of stack-like registers,
        !          1212: and they must be consecutively numbered.
        !          1213: 
        !          1214: `STACK_REGS'
        !          1215:      Define this if the machine has any stack-like registers.
        !          1216: 
        !          1217: `FIRST_STACK_REG'
        !          1218:      The number of the first stack-like register.  This one is the top
        !          1219:      of the stack.
        !          1220: 
        !          1221: `LAST_STACK_REG'
        !          1222:      The number of the last stack-like register.  This one is the
        !          1223:      bottom of the stack.
        !          1224: 

unix.superglobalmegacorp.com

This archive runs on limited infrastructure. Preserving old code on modern bandwidth. Automated agents are requested to crawl responsibly.