Next: Pattern Ordering, Previous: Constraints, Up: Machine Desc
Here is a table of the instruction names that are meaningful in the RTL generation pass of the compiler. Giving one of these names to an instruction pattern tells the RTL generation pass that it can use the pattern to accomplish a certain task.
If operand 0 is a subreg with mode m of a register whose
own mode is wider than m, the effect of this instruction is
to store the specified value in the part of the register that corresponds
to mode m. Bits outside of m, but which are within the
same target word as the subreg are undefined. Bits which are
outside the target word are left unchanged.
This class of patterns is special in several ways. First of all, each of these names up to and including full word size must be defined, because there is no other way to copy a datum from one place to another. If there are patterns accepting operands in larger modes, `movm' must be defined for integer modes of those sizes.
Second, these patterns are not used solely in the RTL generation pass. Even the reload pass can generate move insns to copy values from stack slots into temporary registers. When it does so, one of the operands is a hard register and the other is an operand that can need to be reloaded into a register.
Therefore, when given such a pair of operands, the pattern must generate
RTL which needs no reloading and needs no temporary registers—no
registers other than the operands. For example, if you support the
pattern with a define_expand, then in such a case the
define_expand mustn't call force_reg or any other such
function which might generate new pseudo registers.
This requirement exists even for subword modes on a RISC machine where fetching those modes from memory normally requires several insns and some temporary registers.
During reload a memory reference with an invalid address may be passed
as an operand. Such an address will be replaced with a valid address
later in the reload pass. In this case, nothing may be done with the
address except to use it as it stands. If it is copied, it will not be
replaced with a valid address. No attempt should be made to make such
an address into a valid address and no routine (such as
change_address) that will do so may be called. Note that
general_operand will fail when applied to such an address.
The global variable reload_in_progress (which must be explicitly
declared if required) can be used to determine whether such special
handling is required.
The variety of operands that have reloads depends on the rest of the machine description, but typically on a RISC machine these can only be pseudo registers that did not get hard registers, while on other machines explicit memory references will get optional reloads.
If a scratch register is required to move an object to or from memory,
it can be allocated using gen_reg_rtx prior to life analysis.
If there are cases which need scratch registers during or after reload, you must provide an appropriate secondary_reload target hook.
The global variable no_new_pseudos can be used to determine if it
is unsafe to create new pseudo registers. If this variable is nonzero, then
it is unsafe to call gen_reg_rtx to allocate a new pseudo.
The constraints on a `movm' must permit moving any hard
register to any other hard register provided that
HARD_REGNO_MODE_OK permits mode m in both registers and
REGISTER_MOVE_COST applied to their classes returns a value of 2.
It is obligatory to support floating point `movm'
instructions into and out of any registers that can hold fixed point
values, because unions and structures (which have modes SImode or
DImode) can be in those registers and they may have floating
point members.
There may also be a need to support fixed point `movm'
instructions in and out of floating point registers. Unfortunately, I
have forgotten why this was so, and I don't know whether it is still
true. If HARD_REGNO_MODE_OK rejects fixed point values in
floating point registers, then the constraints of the fixed point
`movm' instructions must be designed to avoid ever trying to
reload into a floating point register.
secondary_reload.
Like `movm', but used when a scratch register is required to
move between operand 0 and operand 1. Operand 2 describes the scratch
register. See the discussion of the SECONDARY_RELOAD_CLASS
macro in see Register Classes.
There are special restrictions on the form of the match_operands
used in these patterns. First, only the predicate for the reload
operand is examined, i.e., reload_in examines operand 1, but not
the predicates for operand 0 or 2. Second, there may be only one
alternative in the constraints. Third, only a single register class
letter may be used for the constraint; subsequent constraint letters
are ignored. As a special exception, an empty constraint string
matches the ALL_REGS register class. This may relieve ports
of the burden of defining an ALL_REGS constraint letter just
for these patterns.
subreg
with mode m of a register whose natural mode is wider,
the `movstrictm' instruction is guaranteed not to alter
any of the register except the part which belongs to mode m.
This pattern is used by the autovectorizer, and when expanding a
MISALIGNED_INDIRECT_REF expression.
Define this only if the target machine really has such an instruction; do not define this if the most efficient way of loading consecutive registers from memory is to do them one at a time.
On some machines, there are restrictions as to which consecutive
registers can be stored into memory, such as particular starting or
ending register numbers or only a range of valid counts. For those
machines, use a define_expand (see Expander Definitions)
and make the pattern fail if the restrictions are not met.
Write the generated insn as a parallel with elements being a
set of one register from the appropriate memory location (you may
also need use or clobber elements). Use a
match_parallel (see RTL Template) to recognize the insn. See
rs6000.md for examples of the use of this insn pattern.
PUSH_ROUNDING is defined. For historical reason, this pattern may be
missing and in such case an mov expander is used instead, with a
MEM expression forming the push operation. The mov expander
method is deprecated.
NaN, then
it is unspecified which of the two operands is returned as the result.
HImode, and store
a SImode product in operand 0.
For machines with an instruction that produces both a quotient and a remainder, provide a pattern for `divmodm4' but do not provide patterns for `divm3' and `modm3'. This allows optimization in the relatively common case when both the quotient and remainder are computed.
If an instruction that just produces a quotient or just a remainder
exists and is more efficient than the instruction that produces both,
write the output routine of `divmodm4' to call
find_reg_note and look for a REG_UNUSED note on the
quotient or remainder and generate the appropriate instruction.
TARGET_SHIFT_TRUNCATION_MASK.
See TARGET_SHIFT_TRUNCATION_MASK.
ashlm3 instructions.
The sqrt built-in function of C always uses the mode which
corresponds to the C data type double and the sqrtf
built-in function uses the mode which corresponds to the C data
type float.
The cos built-in function of C always uses the mode which
corresponds to the C data type double and the cosf
built-in function uses the mode which corresponds to the C data
type float.
The sin built-in function of C always uses the mode which
corresponds to the C data type double and the sinf
built-in function uses the mode which corresponds to the C data
type float.
The exp built-in function of C always uses the mode which
corresponds to the C data type double and the expf
built-in function uses the mode which corresponds to the C data
type float.
The log built-in function of C always uses the mode which
corresponds to the C data type double and the logf
built-in function uses the mode which corresponds to the C data
type float.
The pow built-in function of C always uses the mode which
corresponds to the C data type double and the powf
built-in function uses the mode which corresponds to the C data
type float.
The atan2 built-in function of C always uses the mode which
corresponds to the C data type double and the atan2f
built-in function uses the mode which corresponds to the C data
type float.
The floor built-in function of C always uses the mode which
corresponds to the C data type double and the floorf
built-in function uses the mode which corresponds to the C data
type float.
The trunc built-in function of C always uses the mode which
corresponds to the C data type double and the truncf
built-in function uses the mode which corresponds to the C data
type float.
The round built-in function of C always uses the mode which
corresponds to the C data type double and the roundf
built-in function uses the mode which corresponds to the C data
type float.
The ceil built-in function of C always uses the mode which
corresponds to the C data type double and the ceilf
built-in function uses the mode which corresponds to the C data
type float.
The nearbyint built-in function of C always uses the mode which
corresponds to the C data type double and the nearbyintf
built-in function uses the mode which corresponds to the C data
type float.
The rint built-in function of C always uses the mode which
corresponds to the C data type double and the rintf
built-in function uses the mode which corresponds to the C data
type float.
The copysign built-in function of C always uses the mode which
corresponds to the C data type double and the copysignf
built-in function uses the mode which corresponds to the C data
type float.
The ffs built-in function of C always uses the mode which
corresponds to the C data type int.
(set (cc0) (compare (match_operand:m 0 ...)
(match_operand:m 1 ...)))
(set (cc0) (match_operand:m 0 ...))
`tstm' patterns should not be defined for machines that do
not use (cc0). Doing so would confuse the optimizer since it
would no longer be clear which set operations were comparisons.
The `cmpm' patterns should be used instead.
mem:BLKs with an
address in mode Pmode.
The number of bytes to move is the third operand, in mode m.
Usually, you specify word_mode for m. However, if you can
generate better code knowing the range of valid lengths is smaller than
those representable in a full word, you should provide a pattern with a
mode corresponding to the range of values you can handle efficiently
(e.g., QImode for values in the range 0–127; note we avoid numbers
that appear negative) and also a pattern with word_mode.
The fourth operand is the known shared alignment of the source and
destination, in the form of a const_int rtx. Thus, if the
compiler knows that both source and destination are word-aligned,
it may provide the value 4 for this operand.
Descriptions of multiple movmemm patterns can only be
beneficial if the patterns for smaller modes have fewer restrictions
on their first, second and fourth operands. Note that the mode m
in movmemm does not impose any restriction on the mode of
individually moved data units in the block.
These patterns need not give special consideration to the possibility that the source and destination strings might overlap.
stpcpy semantics. Operand 0 is
an output operand in mode Pmode. The addresses of the
destination and source strings are operands 1 and 2, and both are
mem:BLKs with addresses in mode Pmode. The execution of
the expansion of this pattern should store in operand 0 the address in
which the NUL terminator was stored in the destination string.
mem:BLK whose address is in mode Pmode. The
number of bytes to set is the second operand, in mode m. The value to
initialize the memory with is the third operand. Targets that only support the
clearing of memory should reject any value that is not the constant 0. See
`movmemm' for a discussion of the choice of mode.
The fourth operand is the known alignment of the destination, in the form
of a const_int rtx. Thus, if the compiler knows that the
destination is word-aligned, it may provide the value 4 for this
operand.
The use for multiple setmemm is as for movmemm.
mem:BLK with an address in mode
Pmode.
The fourth operand is the known shared alignment of the source and
destination, in the form of a const_int rtx. Thus, if the
compiler knows that both source and destination are word-aligned,
it may provide the value 4 for this operand.
The two memory blocks specified are compared byte by byte in lexicographic order starting at the beginning of each string. The instruction is not allowed to prefetch more than one byte at a time since either string may end in the first byte and reading past that may access an invalid page or segment and cause a fault. The effect of the instruction is to store a value in operand 0 whose sign indicates the result of the comparison.
mem referring to the first character of the string,
operand 2 is the character to search for (normally zero),
and operand 3 is a constant describing the known alignment
of the beginning of the string.
If the machine description defines this pattern, it also needs to
define the ftrunc pattern.
word_mode.
Operand 1 may have mode byte_mode or word_mode; often
word_mode is allowed only for registers. Operands 2 and 3 must
be valid for word_mode.
The RTL generation pass generates this instruction only with constants for operands 2 and 3 and the constant is never zero for operand 2.
The bit-field value is sign-extended to a full word integer before it is stored in operand 0.
word_mode) into a
bit-field in operand 0, where operand 1 specifies the width in bits and
operand 2 the starting bit. Operand 0 may have mode byte_mode or
word_mode; often word_mode is allowed only for registers.
Operands 1 and 2 must be valid for word_mode.
The RTL generation pass generates this instruction only with constants for operands 1 and 2 and the constant is never zero for operand 1.
The mode of the operands being compared need not be the same as the operands being moved. Some machines, sparc64 for example, have instructions that conditionally move an integer value based on the floating point condition codes and vice versa.
If the machine does not have conditional move instructions, do not define these patterns.
eq, lt or leu.
You specify the mode that the operand must have when you write the
match_operand expression. The compiler automatically sees
which mode you have used and supplies an operand of that mode.
The value stored for a true condition must have 1 as its low bit, or
else must be negative. Otherwise the instruction is not suitable and
you should omit it from the machine description. You describe to the
compiler exactly which value is stored by defining the macro
STORE_FLAG_VALUE (see Misc). If a description cannot be
found that can be used for all the `scond' patterns, you
should omit those operations from the machine description.
These operations may fail, but should do so only in relatively uncommon cases; if they would fail for common cases involving integer comparisons, it is best to omit these patterns.
If these operations are omitted, the compiler will usually generate code
that copies the constant one to the target and branches around an
assignment of zero to the target. If this code is more efficient than
the potential instructions used for the `scond' pattern
followed by those required to convert the result into a 1 or a zero in
SImode, you should omit the `scond' operations from
the machine description.
label_ref that
refers to the label to jump to. Jump if the condition codes meet
condition cond.
Some machines do not follow the model assumed here where a comparison
instruction is followed by a conditional branch instruction. In that
case, the `cmpm' (and `tstm') patterns should
simply store the operands away and generate all the required insns in a
define_expand (see Expander Definitions) for the conditional
branch operations. All calls to expand `bcond' patterns are
immediately preceded by calls to expand either a `cmpm'
pattern or a `tstm' pattern.
Machines that use a pseudo register for the condition code value, or where the mode used for the comparison depends on the condition being tested, should also use the above mechanism. See Jump Patterns.
The above discussion also applies to the `movmodecc' and `scond' patterns.
label_ref that refers to the label to jump to.
label_ref of the label to jump to. This pattern name is mandatory
on all machines.
const_int; operand 2 is the number of registers used as
operands.
On most machines, operand 2 is not actually stored into the RTL pattern. It is supplied for the sake of some RISC machines which need to put this information into the assembler code; they can put it in the RTL instead of operand 1.
Operand 0 should be a mem RTX whose address is the address of the
function. Note, however, that this address can be a symbol_ref
expression even if it would not be a legitimate memory address on the
target machine. If it is also not a valid argument for a call
instruction, the pattern for this operation should be a
define_expand (see Expander Definitions) that places the
address into a register and uses that register in the call instruction.
Subroutines that return BLKmode objects use the `call'
insn.
RETURN_POPS_ARGS is nonzero. They should emit a parallel
that contains both the function call and a set to indicate the
adjustment made to the frame pointer.
For machines where RETURN_POPS_ARGS can be nonzero, the use of these
patterns increases the number of functions for which the frame pointer
can be eliminated, if desired.
parallel
expression where each element is a set expression that indicates
the saving of a function return value into the result block.
This instruction pattern should be defined to support
__builtin_apply on machines where special instructions are needed
to call a subroutine with arbitrary arguments or to save the value
returned. This instruction pattern is required on machines that have
multiple registers that can hold a return value
(i.e. FUNCTION_VALUE_REGNO_P is true for more than one register).
Like the `movm' patterns, this pattern is also used after the RTL generation phase. In this case it is to support machines where multiple instructions are usually needed to return from a function, but some class of functions only requires one instruction to implement a return. Normally, the applicable functions are those which do not need to save any registers or allocate stack space.
For such machines, the condition specified in this pattern should only
be true when reload_completed is nonzero and the function's
epilogue would only be a single instruction. For machines with register
windows, the routine leaf_function_p may be used to determine if
a register window push is required.
Machines that have conditional return instructions should define patterns such as
(define_insn ""
[(set (pc)
(if_then_else (match_operator
0 "comparison_operator"
[(cc0) (const_int 0)])
(return)
(pc)))]
"condition"
"...")
where condition would normally be the same condition specified on the named `return' pattern.
__builtin_return on machines where special
instructions are needed to return a value of any type.
Operand 0 is a memory location where the result of calling a function
with __builtin_apply is stored; operand 1 is a parallel
expression where each element is a set expression that indicates
the restoring of a function return value from the result block.
(const_int 0) will do as an
RTL pattern.
SImode.
The table is a addr_vec or addr_diff_vec inside of a
jump_insn. The number of elements in the table is one plus the
difference between the upper bound and the lower bound.
This pattern requires two operands: the address or offset, and a label
which should immediately precede the jump table. If the macro
CASE_VECTOR_PC_RELATIVE evaluates to a nonzero value then the first
operand is an offset which counts from the address of the table; otherwise,
it is an absolute address to jump to. In either case, the first operand has
mode Pmode.
The `tablejump' insn is always the last insn before the jump table it uses. Its assembler code normally has no need to use the second operand, but you should incorporate it in the RTL pattern so that the jump optimizer will not delete the table as unreachable code.
This optional instruction pattern is only used by the combiner, typically for loops reversed by the loop optimizer when strength reduction is enabled.
const_int or const0_rtx if this cannot be
determined until run-time; operand 2 is the actual or estimated maximum
number of iterations as a const_int; operand 3 is the number of
enclosed loops as a const_int (an innermost loop has a value of
1); operand 4 is the label to jump to if the register is nonzero.
See Looping Patterns.
This optional instruction pattern should be defined for machines with
low-overhead looping instructions as the loop optimizer will try to
modify suitable loops to utilize it. If nested low-overhead looping is
not supported, use a define_expand (see Expander Definitions)
and make the pattern fail if operand 3 is not const1_rtx.
Similarly, if the actual or estimated maximum number of iterations is
too large for this instruction, make it fail.
doloop_end required for machines that
need to perform some initialization, such as loading special registers
used by a low-overhead looping instruction. If initialization insns do
not always need to be emitted, use a define_expand
(see Expander Definitions) and make it fail.
Operand 0 is always a reg and has mode Pmode; operand 1
may be a reg, mem, symbol_ref, const_int, etc
and also has mode Pmode.
Canonicalization of a function pointer usually involves computing the address of the function which would be called if the function pointer were used in an indirect call.
Only define this pattern if function pointers on the target machine can have different values but still call the same function when used in an indirect call.
Pmode. Do not define these patterns on
such machines.
Some machines require special handling for stack pointer saves and
restores. On those machines, define the patterns corresponding to the
non-standard cases by using a define_expand (see Expander Definitions) that produces the required insns. The three types of
saves and restores are:
alloca. Only
the epilogue uses the restored stack pointer, allowing a simpler save or
restore sequence on some machines.
When saving the stack pointer, operand 0 is the save area and operand 1
is the stack pointer. The mode used to allocate the save area defaults
to Pmode but you can override that choice by defining the
STACK_SAVEAREA_MODE macro (see Storage Layout). You must
specify an integral mode, or VOIDmode if no save area is needed
for a particular type of save (either because no save is needed or
because a machine-specific save area can be used). Operand 0 is the
stack pointer and operand 1 is the save area for restore operations. If
`save_stack_block' is defined, operand 0 must not be
VOIDmode since these saves can be arbitrarily nested.
A save area is a mem that is at a constant offset from
virtual_stack_vars_rtx when the stack pointer is saved for use by
nonlocal gotos and a reg in the other two cases.
STACK_GROWS_DOWNWARD is undefined) operand 1 from
the stack pointer to create space for dynamically allocated data.
Store the resultant pointer to this space into operand 0. If you
are allocating space from the main stack, do this by emitting a
move insn to copy virtual_stack_dynamic_rtx to operand 0.
If you are allocating the space elsewhere, generate code to copy the
location of the space to operand 0. In the latter case, you must
ensure this space gets freed when the corresponding space on the main
stack is free.
Do not define this pattern if all that must be done is the subtraction. Some machines require other operations such as stack probes or maintaining the back chain. Define this pattern to emit those operations in addition to updating the stack pointer.
On most machines you need not define this pattern, since GCC will already generate the correct code, which is to load the frame pointer and static chain, restore the stack (using the `restore_stack_nonlocal' pattern, if defined), and jump indirectly to the dispatcher. You need only define this pattern if this code will not work on your machine.
jmp_buf. You will not normally need to define this pattern.
A typical reason why you might need this pattern is if some value, such
as a pointer to a global table, must be restored. Though it is
preferred that the pointer value be recalculated if possible (given the
address of a label for instance). The single argument is a pointer to
the jmp_buf. Note that the buffer is five words long and that
the first three are normally used by the generic mechanism.
builtin_setjmp_setup. The single argument is a pointer to the
jmp_buf.
__builtin_eh_return,
and thence the call frame exception handling library routines, are
built. It is intended to handle non-trivial actions needed along
the abnormal return path.
The address of the exception handler to which the function should return
is passed as operand to this pattern. It will normally need to copied by
the pattern to some special register or memory location.
If the pattern needs to determine the location of the target call
frame in order to do so, it may use EH_RETURN_STACKADJ_RTX,
if defined; it will have already been assigned.
If this pattern is not defined, the default action will be to simply
copy the return address to EH_RETURN_HANDLER_RTX. Either
that macro or this pattern needs to be defined if call frame exception
handling is to be used.
Using a prologue pattern is generally preferred over defining
TARGET_ASM_FUNCTION_PROLOGUE to emit assembly code for the prologue.
The prologue pattern is particularly useful for targets which perform
instruction scheduling.
Using an epilogue pattern is generally preferred over defining
TARGET_ASM_FUNCTION_EPILOGUE to emit assembly code for the epilogue.
The epilogue pattern is particularly useful for targets which perform
instruction scheduling or which have delay slots for their return instruction.
The sibcall_epilogue pattern must not clobber any arguments used for
parameter passing or any stack slots for arguments passed to the current
function.
A typical conditional_trap pattern looks like
(define_insn "conditional_trap"
[(trap_if (match_operator 0 "trap_operator"
[(cc0) (const_int 0)])
(match_operand 1 "const_int_operand" "i"))]
""
"...")
Targets that do not support write prefetches or locality hints can ignore the values of operands 1 and 2.
This pattern must show that both operand 0 and operand 1 are modified.
This pattern must issue any memory barrier instructions such that all memory operations before the atomic operation occur before the atomic operation and all memory operations after the atomic operation occur after the atomic operation.
sync_compare_and_swapmode, except
it should act as if compare part of the compare-and-swap were issued via
cmpm. This comparison will only be used with EQ and
NE branches and setcc operations.
Some targets do expose the success or failure of the compare-and-swap
operation via the status flags. Ideally we wouldn't need a separate
named pattern in order to take advantage of this, but the combine pass
does not handle patterns with multiple sets, which is required by
definition for sync_compare_and_swapmode.
The “nand” operation is ~op0 & op1.
This pattern must issue any memory barrier instructions such that all memory operations before the atomic operation occur before the atomic operation and all memory operations after the atomic operation occur after the atomic operation.
If these patterns are not defined, the operation will be constructed from a compare-and-swap operation, if defined.
This pattern must issue any memory barrier instructions such that all memory operations before the atomic operation occur before the atomic operation and all memory operations after the atomic operation occur after the atomic operation.
If these patterns are not defined, the operation will be constructed from a compare-and-swap operation, if defined.
sync_old_op counterparts,
except that they return the value that exists in the memory location
after the operation, rather than before the operation.
In the ideal case, this operation is an atomic exchange operation, in which the previous value in memory operand is copied into the result operand, and the value operand is stored in the memory operand.
For less capable targets, any value operand that is not the constant 1
should be rejected with FAIL. In this case the target may use
an atomic test-and-set bit operation. The result operand should contain
1 if the bit was previously set and 0 if the bit was previously clear.
The true contents of the memory operand are implementation defined.
This pattern must issue any memory barrier instructions such that the pattern as a whole acts as an acquire barrier, that is all memory operations after the pattern do not occur until the lock is acquired.
If this pattern is not defined, the operation will be constructed from a compare-and-swap operation, if defined.
sync_lock_test_and_setmode. Operand 0 is the memory
that contains the lock; operand 1 is the value to store in the lock.
If the target doesn't implement full semantics for
sync_lock_test_and_setmode, any value operand which is not
the constant 0 should be rejected with FAIL, and the true contents
of the memory operand are implementation defined.
This pattern must issue any memory barrier instructions such that the pattern as a whole acts as a release barrier, that is the lock is released only after all previous memory operations have completed.
If this pattern is not defined, then a memory_barrier pattern
will be emitted, followed by a store of the value to the memory operand.
Pmode value from the memory
in operand 1 to the memory in operand 0 without leaving the value in
a register afterward. This is to avoid leaking the value some place
that an attacker might use to rewrite the stack guard slot after
having clobbered it.
If this pattern is not defined, then a plain move pattern is generated.
Pmode value from the
memory in operand 1 with the memory in operand 0 without leaving the
value in a register afterward and branches to operand 2 if the values
weren't equal.
If this pattern is not defined, then a plain compare pattern and conditional branch pattern is used.