This is Info file gcc.info, produced by Makeinfo version 1.68 from the input file ../../../src/gcc-2.95.3/gcc/gcc.texi. INFO-DIR-SECTION Programming START-INFO-DIR-ENTRY * gcc: (gcc). The GNU Compiler Collection. END-INFO-DIR-ENTRY This file documents the use and the internals of the GNU compiler. Published by the Free Software Foundation 59 Temple Place - Suite 330 Boston, MA 02111-1307 USA Copyright (C) 1988, 1989, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000 Free Software Foundation, Inc. Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies. Permission is granted to copy and distribute modified versions of this manual under the conditions for verbatim copying, provided also that the sections entitled "GNU General Public License" and "Funding for Free Software" are included exactly as in the original, and provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one. Permission is granted to copy and distribute translations of this manual into another language, under the above conditions for modified versions, except that the sections entitled "GNU General Public License" and "Funding for Free Software", and this permission notice, may be included in translations approved by the Free Software Foundation instead of in the original English.  File: gcc.info, Node: Trampolines, Next: Library Calls, Prev: Varargs, Up: Target Macros Trampolines for Nested Functions ================================ A "trampoline" is a small piece of code that is created at run time when the address of a nested function is taken. It normally resides on the stack, in the stack frame of the containing function. These macros tell GNU CC how to generate code to allocate and initialize a trampoline. The instructions in the trampoline must do two things: load a constant address into the static chain register, and jump to the real address of the nested function. On CISC machines such as the m68k, this requires two instructions, a move immediate and a jump. Then the two addresses exist in the trampoline as word-long immediate operands. On RISC machines, it is often necessary to load each address into a register in two parts. Then pieces of each address form separate immediate operands. The code generated to initialize the trampoline must store the variable parts--the static chain value and the function address--into the immediate operands of the instructions. On a CISC machine, this is simply a matter of copying each address to a memory reference at the proper offset from the start of the trampoline. On a RISC machine, it may be necessary to take out pieces of the address and store them separately. `TRAMPOLINE_TEMPLATE (FILE)' A C statement to output, on the stream FILE, assembler code for a block of data that contains the constant parts of a trampoline. This code should not include a label--the label is taken care of automatically. If you do not define this macro, it means no template is needed for the target. Do not define this macro on systems where the block move code to copy the trampoline into place would be larger than the code to generate it on the spot. `TRAMPOLINE_SECTION' The name of a subroutine to switch to the section in which the trampoline template is to be placed (*note Sections::.). The default is a value of `readonly_data_section', which places the trampoline in the section containing read-only data. `TRAMPOLINE_SIZE' A C expression for the size in bytes of the trampoline, as an integer. `TRAMPOLINE_ALIGNMENT' Alignment required for trampolines, in bits. If you don't define this macro, the value of `BIGGEST_ALIGNMENT' is used for aligning trampolines. `INITIALIZE_TRAMPOLINE (ADDR, FNADDR, STATIC_CHAIN)' A C statement to initialize the variable parts of a trampoline. ADDR is an RTX for the address of the trampoline; FNADDR is an RTX for the address of the nested function; STATIC_CHAIN is an RTX for the static chain value that should be passed to the function when it is called. `ALLOCATE_TRAMPOLINE (FP)' A C expression to allocate run-time space for a trampoline. The expression value should be an RTX representing a memory reference to the space for the trampoline. If this macro is not defined, by default the trampoline is allocated as a stack slot. This default is right for most machines. The exceptions are machines where it is impossible to execute instructions in the stack area. On such machines, you may have to implement a separate stack, using this macro in conjunction with `FUNCTION_PROLOGUE' and `FUNCTION_EPILOGUE'. FP points to a data structure, a `struct function', which describes the compilation status of the immediate containing function of the function which the trampoline is for. Normally (when `ALLOCATE_TRAMPOLINE' is not defined), the stack slot for the trampoline is in the stack frame of this containing function. Other allocation strategies probably must do something analogous with this information. Implementing trampolines is difficult on many machines because they have separate instruction and data caches. Writing into a stack location fails to clear the memory in the instruction cache, so when the program jumps to that location, it executes the old contents. Here are two possible solutions. One is to clear the relevant parts of the instruction cache whenever a trampoline is set up. The other is to make all trampolines identical, by having them jump to a standard subroutine. The former technique makes trampoline execution faster; the latter makes initialization faster. To clear the instruction cache when a trampoline is initialized, define the following macros which describe the shape of the cache. `INSN_CACHE_SIZE' The total size in bytes of the cache. `INSN_CACHE_LINE_WIDTH' The length in bytes of each cache line. The cache is divided into cache lines which are disjoint slots, each holding a contiguous chunk of data fetched from memory. Each time data is brought into the cache, an entire line is read at once. The data loaded into a cache line is always aligned on a boundary equal to the line size. `INSN_CACHE_DEPTH' The number of alternative cache lines that can hold any particular memory location. Alternatively, if the machine has system calls or instructions to clear the instruction cache directly, you can define the following macro. `CLEAR_INSN_CACHE (BEG, END)' If defined, expands to a C expression clearing the *instruction cache* in the specified interval. If it is not defined, and the macro INSN_CACHE_SIZE is defined, some generic code is generated to clear the cache. The definition of this macro would typically be a series of `asm' statements. Both BEG and END are both pointer expressions. To use a standard subroutine, define the following macro. In addition, you must make sure that the instructions in a trampoline fill an entire cache line with identical instructions, or else ensure that the beginning of the trampoline code is always aligned at the same point in its cache line. Look in `m68k.h' as a guide. `TRANSFER_FROM_TRAMPOLINE' Define this macro if trampolines need a special subroutine to do their work. The macro should expand to a series of `asm' statements which will be compiled with GNU CC. They go in a library function named `__transfer_from_trampoline'. If you need to avoid executing the ordinary prologue code of a compiled C function when you jump to the subroutine, you can do so by placing a special label of your own in the assembler code. Use one `asm' statement to generate an assembler label, and another to make the label global. Then trampolines can use that label to jump directly to your special assembler code.  File: gcc.info, Node: Library Calls, Next: Addressing Modes, Prev: Trampolines, Up: Target Macros Implicit Calls to Library Routines ================================== Here is an explanation of implicit calls to library routines. `MULSI3_LIBCALL' A C string constant giving the name of the function to call for multiplication of one signed full-word by another. If you do not define this macro, the default name is used, which is `__mulsi3', a function defined in `libgcc.a'. `DIVSI3_LIBCALL' A C string constant giving the name of the function to call for division of one signed full-word by another. If you do not define this macro, the default name is used, which is `__divsi3', a function defined in `libgcc.a'. `UDIVSI3_LIBCALL' A C string constant giving the name of the function to call for division of one unsigned full-word by another. If you do not define this macro, the default name is used, which is `__udivsi3', a function defined in `libgcc.a'. `MODSI3_LIBCALL' A C string constant giving the name of the function to call for the remainder in division of one signed full-word by another. If you do not define this macro, the default name is used, which is `__modsi3', a function defined in `libgcc.a'. `UMODSI3_LIBCALL' A C string constant giving the name of the function to call for the remainder in division of one unsigned full-word by another. If you do not define this macro, the default name is used, which is `__umodsi3', a function defined in `libgcc.a'. `MULDI3_LIBCALL' A C string constant giving the name of the function to call for multiplication of one signed double-word by another. If you do not define this macro, the default name is used, which is `__muldi3', a function defined in `libgcc.a'. `DIVDI3_LIBCALL' A C string constant giving the name of the function to call for division of one signed double-word by another. If you do not define this macro, the default name is used, which is `__divdi3', a function defined in `libgcc.a'. `UDIVDI3_LIBCALL' A C string constant giving the name of the function to call for division of one unsigned full-word by another. If you do not define this macro, the default name is used, which is `__udivdi3', a function defined in `libgcc.a'. `MODDI3_LIBCALL' A C string constant giving the name of the function to call for the remainder in division of one signed double-word by another. If you do not define this macro, the default name is used, which is `__moddi3', a function defined in `libgcc.a'. `UMODDI3_LIBCALL' A C string constant giving the name of the function to call for the remainder in division of one unsigned full-word by another. If you do not define this macro, the default name is used, which is `__umoddi3', a function defined in `libgcc.a'. `INIT_TARGET_OPTABS' Define this macro as a C statement that declares additional library routines renames existing ones. `init_optabs' calls this macro after initializing all the normal library routines. `TARGET_EDOM' The value of `EDOM' on the target machine, as a C integer constant expression. If you don't define this macro, GNU CC does not attempt to deposit the value of `EDOM' into `errno' directly. Look in `/usr/include/errno.h' to find the value of `EDOM' on your system. If you do not define `TARGET_EDOM', then compiled code reports domain errors by calling the library function and letting it report the error. If mathematical functions on your system use `matherr' when there is an error, then you should leave `TARGET_EDOM' undefined so that `matherr' is used normally. `GEN_ERRNO_RTX' Define this macro as a C expression to create an rtl expression that refers to the global "variable" `errno'. (On certain systems, `errno' may not actually be a variable.) If you don't define this macro, a reasonable default is used. `TARGET_MEM_FUNCTIONS' Define this macro if GNU CC should generate calls to the System V (and ANSI C) library functions `memcpy' and `memset' rather than the BSD functions `bcopy' and `bzero'. `LIBGCC_NEEDS_DOUBLE' Define this macro if only `float' arguments cannot be passed to library routines (so they must be converted to `double'). This macro affects both how library calls are generated and how the library routines in `libgcc1.c' accept their arguments. It is useful on machines where floating and fixed point arguments are passed differently, such as the i860. `FLOAT_ARG_TYPE' Define this macro to override the type used by the library routines to pick up arguments of type `float'. (By default, they use a union of `float' and `int'.) The obvious choice would be `float'--but that won't work with traditional C compilers that expect all arguments declared as `float' to arrive as `double'. To avoid this conversion, the library routines ask for the value as some other type and then treat it as a `float'. On some systems, no other type will work for this. For these systems, you must use `LIBGCC_NEEDS_DOUBLE' instead, to force conversion of the values `double' before they are passed. `FLOATIFY (PASSED-VALUE)' Define this macro to override the way library routines redesignate a `float' argument as a `float' instead of the type it was passed as. The default is an expression which takes the `float' field of the union. `FLOAT_VALUE_TYPE' Define this macro to override the type used by the library routines to return values that ought to have type `float'. (By default, they use `int'.) The obvious choice would be `float'--but that won't work with traditional C compilers gratuitously convert values declared as `float' into `double'. `INTIFY (FLOAT-VALUE)' Define this macro to override the way the value of a `float'-returning library routine should be packaged in order to return it. These functions are actually declared to return type `FLOAT_VALUE_TYPE' (normally `int'). These values can't be returned as type `float' because traditional C compilers would gratuitously convert the value to a `double'. A local variable named `intify' is always available when the macro `INTIFY' is used. It is a union of a `float' field named `f' and a field named `i' whose type is `FLOAT_VALUE_TYPE' or `int'. If you don't define this macro, the default definition works by copying the value through that union. `nongcc_SI_type' Define this macro as the name of the data type corresponding to `SImode' in the system's own C compiler. You need not define this macro if that type is `long int', as it usually is. `nongcc_word_type' Define this macro as the name of the data type corresponding to the word_mode in the system's own C compiler. You need not define this macro if that type is `long int', as it usually is. `perform_...' Define these macros to supply explicit C statements to carry out various arithmetic operations on types `float' and `double' in the library routines in `libgcc1.c'. See that file for a full list of these macros and their arguments. On most machines, you don't need to define any of these macros, because the C compiler that comes with the system takes care of doing them. `NEXT_OBJC_RUNTIME' Define this macro to generate code for Objective C message sending using the calling convention of the NeXT system. This calling convention involves passing the object, the selector and the method arguments all at once to the method-lookup library function. The default calling convention passes just the object and the selector to the lookup function, which returns a pointer to the method.  File: gcc.info, Node: Addressing Modes, Next: Condition Code, Prev: Library Calls, Up: Target Macros Addressing Modes ================ This is about addressing modes. `HAVE_POST_INCREMENT' A C expression that is nonzero the machine supports post-increment addressing. `HAVE_PRE_INCREMENT' `HAVE_POST_DECREMENT' `HAVE_PRE_DECREMENT' Similar for other kinds of addressing. `CONSTANT_ADDRESS_P (X)' A C expression that is 1 if the RTX X is a constant which is a valid address. On most machines, this can be defined as `CONSTANT_P (X)', but a few machines are more restrictive in which constant addresses are supported. `CONSTANT_P' accepts integer-values expressions whose values are not explicitly known, such as `symbol_ref', `label_ref', and `high' expressions and `const' arithmetic expressions, in addition to `const_int' and `const_double' expressions. `MAX_REGS_PER_ADDRESS' A number, the maximum number of registers that can appear in a valid memory address. Note that it is up to you to specify a value equal to the maximum number that `GO_IF_LEGITIMATE_ADDRESS' would ever accept. `GO_IF_LEGITIMATE_ADDRESS (MODE, X, LABEL)' A C compound statement with a conditional `goto LABEL;' executed if X (an RTX) is a legitimate memory address on the target machine for a memory operand of mode MODE. It usually pays to define several simpler macros to serve as subroutines for this one. Otherwise it may be too complicated to understand. This macro must exist in two variants: a strict variant and a non-strict one. The strict variant is used in the reload pass. It must be defined so that any pseudo-register that has not been allocated a hard register is considered a memory reference. In contexts where some kind of register is required, a pseudo-register with no hard register must be rejected. The non-strict variant is used in other passes. It must be defined to accept all pseudo-registers in every context where some kind of register is required. Compiler source files that want to use the strict variant of this macro define the macro `REG_OK_STRICT'. You should use an `#ifdef REG_OK_STRICT' conditional to define the strict variant in that case and the non-strict variant otherwise. Subroutines to check for acceptable registers for various purposes (one for base registers, one for index registers, and so on) are typically among the subroutines used to define `GO_IF_LEGITIMATE_ADDRESS'. Then only these subroutine macros need have two variants; the higher levels of macros may be the same whether strict or not. Normally, constant addresses which are the sum of a `symbol_ref' and an integer are stored inside a `const' RTX to mark them as constant. Therefore, there is no need to recognize such sums specifically as legitimate addresses. Normally you would simply recognize any `const' as legitimate. Usually `PRINT_OPERAND_ADDRESS' is not prepared to handle constant sums that are not marked with `const'. It assumes that a naked `plus' indicates indexing. If so, then you *must* reject such naked constant sums as illegitimate addresses, so that none of them will be given to `PRINT_OPERAND_ADDRESS'. On some machines, whether a symbolic address is legitimate depends on the section that the address refers to. On these machines, define the macro `ENCODE_SECTION_INFO' to store the information into the `symbol_ref', and then check for it here. When you see a `const', you will have to look inside it to find the `symbol_ref' in order to determine the section. *Note Assembler Format::. The best way to modify the name string is by adding text to the beginning, with suitable punctuation to prevent any ambiguity. Allocate the new name in `saveable_obstack'. You will have to modify `ASM_OUTPUT_LABELREF' to remove and decode the added text and output the name accordingly, and define `STRIP_NAME_ENCODING' to access the original name string. You can check the information stored here into the `symbol_ref' in the definitions of the macros `GO_IF_LEGITIMATE_ADDRESS' and `PRINT_OPERAND_ADDRESS'. `REG_OK_FOR_BASE_P (X)' A C expression that is nonzero if X (assumed to be a `reg' RTX) is valid for use as a base register. For hard registers, it should always accept those which the hardware permits and reject the others. Whether the macro accepts or rejects pseudo registers must be controlled by `REG_OK_STRICT' as described above. This usually requires two variant definitions, of which `REG_OK_STRICT' controls the one actually used. `REG_MODE_OK_FOR_BASE_P (X, MODE)' A C expression that is just like `REG_OK_FOR_BASE_P', except that that expression may examine the mode of the memory reference in MODE. You should define this macro if the mode of the memory reference affects whether a register may be used as a base register. If you define this macro, the compiler will use it instead of `REG_OK_FOR_BASE_P'. `REG_OK_FOR_INDEX_P (X)' A C expression that is nonzero if X (assumed to be a `reg' RTX) is valid for use as an index register. The difference between an index register and a base register is that the index register may be scaled. If an address involves the sum of two registers, neither one of them scaled, then either one may be labeled the "base" and the other the "index"; but whichever labeling is used must fit the machine's constraints of which registers may serve in each capacity. The compiler will try both labelings, looking for one that is valid, and will reload one or both registers only if neither labeling works. `LEGITIMIZE_ADDRESS (X, OLDX, MODE, WIN)' A C compound statement that attempts to replace X with a valid memory address for an operand of mode MODE. WIN will be a C statement label elsewhere in the code; the macro definition may use GO_IF_LEGITIMATE_ADDRESS (MODE, X, WIN); to avoid further processing if the address has become legitimate. X will always be the result of a call to `break_out_memory_refs', and OLDX will be the operand that was given to that function to produce X. The code generated by this macro should not alter the substructure of X. If it transforms X into a more legitimate form, it should assign X (which will always be a C variable) a new value. It is not necessary for this macro to come up with a legitimate address. The compiler has standard ways of doing so in all cases. In fact, it is safe for this macro to do nothing. But often a machine-dependent strategy can generate better code. `LEGITIMIZE_RELOAD_ADDRESS (X, MODE, OPNUM, TYPE, IND_LEVELS, WIN)' A C compound statement that attempts to replace X, which is an address that needs reloading, with a valid memory address for an operand of mode MODE. WIN will be a C statement label elsewhere in the code. It is not necessary to define this macro, but it might be useful for performance reasons. For example, on the i386, it is sometimes possible to use a single reload register instead of two by reloading a sum of two pseudo registers into a register. On the other hand, for number of RISC processors offsets are limited so that often an intermediate address needs to be generated in order to address a stack slot. By defining LEGITIMIZE_RELOAD_ADDRESS appropriately, the intermediate addresses generated for adjacent some stack slots can be made identical, and thus be shared. *Note*: This macro should be used with caution. It is necessary to know something of how reload works in order to effectively use this, and it is quite easy to produce macros that build in too much knowledge of reload internals. *Note*: This macro must be able to reload an address created by a previous invocation of this macro. If it fails to handle such addresses then the compiler may generate incorrect code or abort. The macro definition should use `push_reload' to indicate parts that need reloading; OPNUM, TYPE and IND_LEVELS are usually suitable to be passed unaltered to `push_reload'. The code generated by this macro must not alter the substructure of X. If it transforms X into a more legitimate form, it should assign X (which will always be a C variable) a new value. This also applies to parts that you change indirectly by calling `push_reload'. The macro definition may use `strict_memory_address_p' to test if the address has become legitimate. If you want to change only a part of X, one standard way of doing this is to use `copy_rtx'. Note, however, that is unshares only a single level of rtl. Thus, if the part to be changed is not at the top level, you'll need to replace first the top leve It is not necessary for this macro to come up with a legitimate address; but often a machine-dependent strategy can generate better code. `GO_IF_MODE_DEPENDENT_ADDRESS (ADDR, LABEL)' A C statement or compound statement with a conditional `goto LABEL;' executed if memory address X (an RTX) can have different meanings depending on the machine mode of the memory reference it is used for or if the address is valid for some modes but not others. Autoincrement and autodecrement addresses typically have mode-dependent effects because the amount of the increment or decrement is the size of the operand being addressed. Some machines have other mode-dependent addresses. Many RISC machines have no mode-dependent addresses. You may assume that ADDR is a valid address for the machine. `LEGITIMATE_CONSTANT_P (X)' A C expression that is nonzero if X is a legitimate constant for an immediate operand on the target machine. You can assume that X satisfies `CONSTANT_P', so you need not check this. In fact, `1' is a suitable definition for this macro on machines where anything `CONSTANT_P' is valid.  File: gcc.info, Node: Condition Code, Next: Costs, Prev: Addressing Modes, Up: Target Macros Condition Code Status ===================== This describes the condition code status. The file `conditions.h' defines a variable `cc_status' to describe how the condition code was computed (in case the interpretation of the condition code depends on the instruction that it was set by). This variable contains the RTL expressions on which the condition code is currently based, and several standard flags. Sometimes additional machine-specific flags must be defined in the machine description header file. It can also add additional machine-specific information by defining `CC_STATUS_MDEP'. `CC_STATUS_MDEP' C code for a data type which is used for declaring the `mdep' component of `cc_status'. It defaults to `int'. This macro is not used on machines that do not use `cc0'. `CC_STATUS_MDEP_INIT' A C expression to initialize the `mdep' field to "empty". The default definition does nothing, since most machines don't use the field anyway. If you want to use the field, you should probably define this macro to initialize it. This macro is not used on machines that do not use `cc0'. `NOTICE_UPDATE_CC (EXP, INSN)' A C compound statement to set the components of `cc_status' appropriately for an insn INSN whose body is EXP. It is this macro's responsibility to recognize insns that set the condition code as a byproduct of other activity as well as those that explicitly set `(cc0)'. This macro is not used on machines that do not use `cc0'. If there are insns that do not set the condition code but do alter other machine registers, this macro must check to see whether they invalidate the expressions that the condition code is recorded as reflecting. For example, on the 68000, insns that store in address registers do not set the condition code, which means that usually `NOTICE_UPDATE_CC' can leave `cc_status' unaltered for such insns. But suppose that the previous insn set the condition code based on location `a4@(102)' and the current insn stores a new value in `a4'. Although the condition code is not changed by this, it will no longer be true that it reflects the contents of `a4@(102)'. Therefore, `NOTICE_UPDATE_CC' must alter `cc_status' in this case to say that nothing is known about the condition code value. The definition of `NOTICE_UPDATE_CC' must be prepared to deal with the results of peephole optimization: insns whose patterns are `parallel' RTXs containing various `reg', `mem' or constants which are just the operands. The RTL structure of these insns is not sufficient to indicate what the insns actually do. What `NOTICE_UPDATE_CC' should do when it sees one is just to run `CC_STATUS_INIT'. A possible definition of `NOTICE_UPDATE_CC' is to call a function that looks at an attribute (*note Insn Attributes::.) named, for example, `cc'. This avoids having detailed information about patterns in two places, the `md' file and in `NOTICE_UPDATE_CC'. `EXTRA_CC_MODES' A list of names to be used for additional modes for condition code values in registers (*note Jump Patterns::.). These names are added to `enum machine_mode' and all have class `MODE_CC'. By convention, they should start with `CC' and end with `mode'. You should only define this macro if your machine does not use `cc0' and only if additional modes are required. `EXTRA_CC_NAMES' A list of C strings giving the names for the modes listed in `EXTRA_CC_MODES'. For example, the Sparc defines this macro and `EXTRA_CC_MODES' as #define EXTRA_CC_MODES CC_NOOVmode, CCFPmode, CCFPEmode #define EXTRA_CC_NAMES "CC_NOOV", "CCFP", "CCFPE" This macro is not required if `EXTRA_CC_MODES' is not defined. `SELECT_CC_MODE (OP, X, Y)' Returns a mode from class `MODE_CC' to be used when comparison operation code OP is applied to rtx X and Y. For example, on the Sparc, `SELECT_CC_MODE' is defined as (see *note Jump Patterns::. for a description of the reason for this definition) #define SELECT_CC_MODE(OP,X,Y) \ (GET_MODE_CLASS (GET_MODE (X)) == MODE_FLOAT \ ? ((OP == EQ || OP == NE) ? CCFPmode : CCFPEmode) \ : ((GET_CODE (X) == PLUS || GET_CODE (X) == MINUS \ || GET_CODE (X) == NEG) \ ? CC_NOOVmode : CCmode)) You need not define this macro if `EXTRA_CC_MODES' is not defined. `CANONICALIZE_COMPARISON (CODE, OP0, OP1)' One some machines not all possible comparisons are defined, but you can convert an invalid comparison into a valid one. For example, the Alpha does not have a `GT' comparison, but you can use an `LT' comparison instead and swap the order of the operands. On such machines, define this macro to be a C statement to do any required conversions. CODE is the initial comparison code and OP0 and OP1 are the left and right operands of the comparison, respectively. You should modify CODE, OP0, and OP1 as required. GNU CC will not assume that the comparison resulting from this macro is valid but will see if the resulting insn matches a pattern in the `md' file. You need not define this macro if it would never change the comparison code or operands. `REVERSIBLE_CC_MODE (MODE)' A C expression whose value is one if it is always safe to reverse a comparison whose mode is MODE. If `SELECT_CC_MODE' can ever return MODE for a floating-point inequality comparison, then `REVERSIBLE_CC_MODE (MODE)' must be zero. You need not define this macro if it would always returns zero or if the floating-point format is anything other than `IEEE_FLOAT_FORMAT'. For example, here is the definition used on the Sparc, where floating-point inequality comparisons are always given `CCFPEmode': #define REVERSIBLE_CC_MODE(MODE) ((MODE) != CCFPEmode)  File: gcc.info, Node: Costs, Next: Sections, Prev: Condition Code, Up: Target Macros Describing Relative Costs of Operations ======================================= These macros let you describe the relative speed of various operations on the target machine. `CONST_COSTS (X, CODE, OUTER_CODE)' A part of a C `switch' statement that describes the relative costs of constant RTL expressions. It must contain `case' labels for expression codes `const_int', `const', `symbol_ref', `label_ref' and `const_double'. Each case must ultimately reach a `return' statement to return the relative cost of the use of that kind of constant value in an expression. The cost may depend on the precise value of the constant, which is available for examination in X, and the rtx code of the expression in which it is contained, found in OUTER_CODE. CODE is the expression code--redundant, since it can be obtained with `GET_CODE (X)'. `RTX_COSTS (X, CODE, OUTER_CODE)' Like `CONST_COSTS' but applies to nonconstant RTL expressions. This can be used, for example, to indicate how costly a multiply instruction is. In writing this macro, you can use the construct `COSTS_N_INSNS (N)' to specify a cost equal to N fast instructions. OUTER_CODE is the code of the expression in which X is contained. This macro is optional; do not define it if the default cost assumptions are adequate for the target machine. `DEFAULT_RTX_COSTS (X, CODE, OUTER_CODE)' This macro, if defined, is called for any case not handled by the `RTX_COSTS' or `CONST_COSTS' macros. This eliminates the need to put case labels into the macro, but the code, or any functions it calls, must assume that the RTL in X could be of any type that has not already been handled. The arguments are the same as for `RTX_COSTS', and the macro should execute a return statement giving the cost of any RTL expressions that it can handle. The default cost calculation is used for any RTL for which this macro does not return a value. This macro is optional; do not define it if the default cost assumptions are adequate for the target machine. `ADDRESS_COST (ADDRESS)' An expression giving the cost of an addressing mode that contains ADDRESS. If not defined, the cost is computed from the ADDRESS expression and the `CONST_COSTS' values. For most CISC machines, the default cost is a good approximation of the true cost of the addressing mode. However, on RISC machines, all instructions normally have the same length and execution time. Hence all addresses will have equal costs. In cases where more than one form of an address is known, the form with the lowest cost will be used. If multiple forms have the same, lowest, cost, the one that is the most complex will be used. For example, suppose an address that is equal to the sum of a register and a constant is used twice in the same basic block. When this macro is not defined, the address will be computed in a register and memory references will be indirect through that register. On machines where the cost of the addressing mode containing the sum is no higher than that of a simple indirect reference, this will produce an additional instruction and possibly require an additional register. Proper specification of this macro eliminates this overhead for such machines. Similar use of this macro is made in strength reduction of loops. ADDRESS need not be valid as an address. In such a case, the cost is not relevant and can be any value; invalid addresses need not be assigned a different cost. On machines where an address involving more than one register is as cheap as an address computation involving only one register, defining `ADDRESS_COST' to reflect this can cause two registers to be live over a region of code where only one would have been if `ADDRESS_COST' were not defined in that manner. This effect should be considered in the definition of this macro. Equivalent costs should probably only be given to addresses with different numbers of registers on machines with lots of registers. This macro will normally either not be defined or be defined as a constant. `REGISTER_MOVE_COST (FROM, TO)' A C expression for the cost of moving data from a register in class FROM to one in class TO. The classes are expressed using the enumeration values such as `GENERAL_REGS'. A value of 2 is the default; other values are interpreted relative to that. It is not required that the cost always equal 2 when FROM is the same as TO; on some machines it is expensive to move between registers if they are not general registers. If reload sees an insn consisting of a single `set' between two hard registers, and if `REGISTER_MOVE_COST' applied to their classes returns a value of 2, reload does not check to ensure that the constraints of the insn are met. Setting a cost of other than 2 will allow reload to verify that the constraints are met. You should do this if the `movM' pattern's constraints do not allow such copying. `MEMORY_MOVE_COST (MODE, CLASS, IN)' A C expression for the cost of moving data of mode MODE between a register of class CLASS and memory; IN is zero if the value is to be written to memory, non-zero if it is to be read in. This cost is relative to those in `REGISTER_MOVE_COST'. If moving between registers and memory is more expensive than between two registers, you should define this macro to express the relative cost. If you do not define this macro, GNU CC uses a default cost of 4 plus the cost of copying via a secondary reload register, if one is needed. If your machine requires a secondary reload register to copy between memory and a register of CLASS but the reload mechanism is more complex than copying via an intermediate, define this macro to reflect the actual cost of the move. GNU CC defines the function `memory_move_secondary_cost' if secondary reloads are needed. It computes the costs due to copying via a secondary register. If your machine copies from memory using a secondary register in the conventional way but the default base value of 4 is not correct for your machine, define this macro to add some other value to the result of that function. The arguments to that function are the same as to this macro. `BRANCH_COST' A C expression for the cost of a branch instruction. A value of 1 is the default; other values are interpreted relative to that. Here are additional macros which do not specify precise relative costs, but only that certain actions are more expensive than GNU CC would ordinarily expect. `SLOW_BYTE_ACCESS' Define this macro as a C expression which is nonzero if accessing less than a word of memory (i.e. a `char' or a `short') is no faster than accessing a word of memory, i.e., if such access require more than one instruction or if there is no difference in cost between byte and (aligned) word loads. When this macro is not defined, the compiler will access a field by finding the smallest containing object; when it is defined, a fullword load will be used if alignment permits. Unless bytes accesses are faster than word accesses, using word accesses is preferable since it may eliminate subsequent memory access if subsequent accesses occur to other fields in the same word of the structure, but to different bytes. `SLOW_ZERO_EXTEND' Define this macro if zero-extension (of a `char' or `short' to an `int') can be done faster if the destination is a register that is known to be zero. If you define this macro, you must have instruction patterns that recognize RTL structures like this: (set (strict_low_part (subreg:QI (reg:SI ...) 0)) ...) and likewise for `HImode'. `SLOW_UNALIGNED_ACCESS' Define this macro to be the value 1 if unaligned accesses have a cost many times greater than aligned accesses, for example if they are emulated in a trap handler. When this macro is non-zero, the compiler will act as if `STRICT_ALIGNMENT' were non-zero when generating code for block moves. This can cause significantly more instructions to be produced. Therefore, do not set this macro non-zero if unaligned accesses only add a cycle or two to the time for a memory access. If the value of this macro is always zero, it need not be defined. `DONT_REDUCE_ADDR' Define this macro to inhibit strength reduction of memory addresses. (On some machines, such strength reduction seems to do harm rather than good.) `MOVE_RATIO' The threshold of number of scalar memory-to-memory move insns, *below* which a sequence of insns should be generated instead of a string move insn or a library call. Increasing the value will always make code faster, but eventually incurs high cost in increased code size. Note that on machines with no memory-to-memory move insns, this macro denotes the corresponding number of memory-to-memory *sequences*. If you don't define this, a reasonable default is used. `MOVE_BY_PIECES_P (SIZE, ALIGNMENT)' A C expression used to determine whether `move_by_pieces' will be used to copy a chunk of memory, or whether some other block move mechanism will be used. Defaults to 1 if `move_by_pieces_ninsns' returns less than `MOVE_RATIO'. `MOVE_MAX_PIECES' A C expression used by `move_by_pieces' to determine the largest unit a load or store used to copy memory is. Defaults to `MOVE_MAX'. `USE_LOAD_POST_INCREMENT (MODE)' A C expression used to determine whether a load postincrement is a good thing to use for a given mode. Defaults to the value of `HAVE_POST_INCREMENT'. `USE_LOAD_POST_DECREMENT (MODE)' A C expression used to determine whether a load postdecrement is a good thing to use for a given mode. Defaults to the value of `HAVE_POST_DECREMENT'. `USE_LOAD_PRE_INCREMENT (MODE)' A C expression used to determine whether a load preincrement is a good thing to use for a given mode. Defaults to the value of `HAVE_PRE_INCREMENT'. `USE_LOAD_PRE_DECREMENT (MODE)' A C expression used to determine whether a load predecrement is a good thing to use for a given mode. Defaults to the value of `HAVE_PRE_DECREMENT'. `USE_STORE_POST_INCREMENT (MODE)' A C expression used to determine whether a store postincrement is a good thing to use for a given mode. Defaults to the value of `HAVE_POST_INCREMENT'. `USE_STORE_POST_DECREMENT (MODE)' A C expression used to determine whether a store postdeccrement is a good thing to use for a given mode. Defaults to the value of `HAVE_POST_DECREMENT'. `USE_STORE_PRE_INCREMENT (MODE)' This macro is used to determine whether a store preincrement is a good thing to use for a given mode. Defaults to the value of `HAVE_PRE_INCREMENT'. `USE_STORE_PRE_DECREMENT (MODE)' This macro is used to determine whether a store predecrement is a good thing to use for a given mode. Defaults to the value of `HAVE_PRE_DECREMENT'. `NO_FUNCTION_CSE' Define this macro if it is as good or better to call a constant function address than to call an address kept in a register. `NO_RECURSIVE_FUNCTION_CSE' Define this macro if it is as good or better for a function to call itself with an explicit address than to call an address kept in a register. `ADJUST_COST (INSN, LINK, DEP_INSN, COST)' A C statement (sans semicolon) to update the integer variable COST based on the relationship between INSN that is dependent on DEP_INSN through the dependence LINK. The default is to make no adjustment to COST. This can be used for example to specify to the scheduler that an output- or anti-dependence does not incur the same cost as a data-dependence. `ADJUST_PRIORITY (INSN)' A C statement (sans semicolon) to update the integer scheduling priority `INSN_PRIORITY(INSN)'. Reduce the priority to execute the INSN earlier, increase the priority to execute INSN later. Do not define this macro if you do not need to adjust the scheduling priorities of insns.