| ------------------------------------------------------------------------------ |
| -- -- |
| -- GNAT COMPILER COMPONENTS -- |
| -- -- |
| -- T T Y P E S -- |
| -- -- |
| -- S p e c -- |
| -- -- |
| -- Copyright (C) 1992-2022, Free Software Foundation, Inc. -- |
| -- -- |
| -- GNAT is free software; you can redistribute it and/or modify it under -- |
| -- terms of the GNU General Public License as published by the Free Soft- -- |
| -- ware Foundation; either version 3, or (at your option) any later ver- -- |
| -- sion. GNAT is distributed in the hope that it will be useful, but WITH- -- |
| -- OUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY -- |
| -- or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License -- |
| -- for more details. You should have received a copy of the GNU General -- |
| -- Public License distributed with GNAT; see file COPYING3. If not, go to -- |
| -- http://www.gnu.org/licenses for a complete copy of the license. -- |
| -- -- |
| -- GNAT was originally developed by the GNAT team at New York University. -- |
| -- Extensive contributions were provided by Ada Core Technologies Inc. -- |
| -- -- |
| ------------------------------------------------------------------------------ |
| |
| -- This package contains constants describing target properties |
| |
| with Types; use Types; |
| with Set_Targ; |
| |
| package Ttypes is |
| |
| ------------------------------ |
| -- Host/Target Dependencies -- |
| ------------------------------ |
| |
| -- It is vital to maintain a clear distinction between properties of |
| -- types on the host and types on the target, since in the general |
| -- case of a cross-compiler these will be different. |
| |
| -- This package provides definitions of values that describe the properties |
| -- of the target types. All instances of target dependencies, including the |
| -- definitions of such packages as Standard and System depend directly or |
| -- indirectly on the definitions in the Ttypes packages. |
| |
| -- In the source of the compiler, references to attributes such as |
| -- Integer'Size will give information regarding the host types (i.e. |
| -- the types within the compiler itself). Such references are therefore |
| -- almost always suspicious (it is hard for example to see that the |
| -- code in the compiler should even be using type Integer very much, |
| -- and certainly this code should not depend on the size of Integer). |
| |
| -- On the other hand, it is perfectly reasonable for the compiler to |
| -- require access to the size of type Integer for the target machine, |
| -- e.g. in constructing the internal representation of package Standard. |
| -- For this purpose, instead of referencing the attribute Integer'Size, |
| -- a reference to Ttypes.Standard_Integer_Size will provide the needed |
| -- value for the target type. |
| |
| -- Two approaches are used for handling target dependent values in the |
| -- standard library packages. Package Standard is handled specially, |
| -- being constructed internally (by package Stand). Target dependent |
| -- values needed in Stand are obtained by direct reference to Ttypes |
| -- and Ttypef. |
| |
| -- For package System, there is a separate version for each target, with |
| -- explicit declarations of the required, constants. |
| |
| -- Historical note: Originally we had in mind dealing with target dependent |
| -- differences by referencing appropriate attributes. Ada 95 already |
| -- defines most of the required attributes, and GNAT specific attributes |
| -- were defined to cover the remaining cases (such as Storage_Unit). |
| -- The additional attributes that have been added to GNAT (Address_Size, |
| -- Storage_Unit, and Word_Size) are redundant with respect to the |
| -- corresponding references to System constants. For example in a program, |
| -- System.Address_Size and Standard'Address_Size yield the same value. |
| -- These attributes have been retained for compatibility purposes, but |
| -- serve no purpose at this stage. |
| |
| -- Note that during compilation there are two versions of package System |
| -- around. The version that is directly with'ed by compiler packages |
| -- contains host-dependent definitions, which is what is needed in that |
| -- case (for example, System.Storage_Unit referenced in the source of the |
| -- compiler refers to the storage unit of the host, not the target). This |
| -- means that, like attribute references, any references to constants in |
| -- package System in the compiler code are suspicious, since it is strange |
| -- for the compiler to have such host dependencies. If the compiler needs |
| -- to access the target dependent values of such quantities as Storage_Unit |
| -- then it should reference the constants in this package (Ttypes), rather |
| -- than referencing System.Storage_Unit, or Standard'Storage_Unit, both of |
| -- which would yield the host value. |
| |
| --------------------------------------------------- |
| -- Target-Dependent Values for Types in Standard -- |
| --------------------------------------------------- |
| |
| -- Note: GNAT always supplies all the following integer and float types, |
| -- but depending on the machine, some of the types may be identical. For |
| -- example, on some machines, Short_Float may be the same as Float, and |
| -- Long_Long_Float may be the same as Long_Float. |
| |
| Standard_Short_Short_Integer_Size : constant Pos := |
| Set_Targ.Char_Size; |
| |
| Standard_Short_Integer_Size : constant Pos := |
| Set_Targ.Short_Size; |
| |
| Standard_Integer_Size : constant Pos := |
| Set_Targ.Int_Size; |
| |
| Standard_Long_Integer_Size : constant Pos := |
| Set_Targ.Long_Size; |
| |
| Standard_Long_Long_Integer_Size : constant Pos := |
| Set_Targ.Long_Long_Size; |
| |
| Standard_Long_Long_Long_Integer_Size : Pos := |
| Set_Targ.Long_Long_Long_Size; |
| |
| Standard_Short_Float_Size : constant Pos := |
| Set_Targ.Float_Size; |
| |
| Standard_Float_Size : constant Pos := |
| Set_Targ.Float_Size; |
| |
| Standard_Long_Float_Size : constant Pos := |
| Set_Targ.Double_Size; |
| |
| Standard_Long_Long_Float_Size : constant Pos := |
| Set_Targ.Long_Double_Size; |
| |
| Standard_Character_Size : constant Pos := Set_Targ.Char_Size; |
| |
| Standard_Wide_Character_Size : constant Pos := 16; |
| Standard_Wide_Wide_Character_Size : constant Pos := 32; |
| -- Standard wide character sizes |
| |
| -- Note: there is no specific control over the representation of |
| -- enumeration types. The convention used is that if an enumeration |
| -- type has fewer than 2**(Character'Size) elements, then the size |
| -- used is Character'Size, otherwise Integer'Size is used. |
| |
| -- Similarly, the size of fixed-point types depends on the size of the |
| -- corresponding integer type, which is the smallest predefined integer |
| -- type capable of representing the required range of values. |
| |
| ------------------------------------------------- |
| -- Target-Dependent Values for Types in System -- |
| ------------------------------------------------- |
| |
| System_Address_Size : constant Pos := Set_Targ.Pointer_Size; |
| -- System.Address'Size (also size of all thin pointers) |
| |
| System_Max_Integer_Size : Pos := Standard_Long_Long_Long_Integer_Size; |
| |
| System_Max_Binary_Modulus_Power : Pos := |
| Standard_Long_Long_Long_Integer_Size; |
| |
| System_Max_Nonbinary_Modulus_Power : constant Pos := Standard_Integer_Size; |
| |
| System_Storage_Unit : constant Pos := Set_Targ.Bits_Per_Unit; |
| System_Word_Size : constant Pos := Set_Targ.Bits_Per_Word; |
| |
| System_Tick_Nanoseconds : constant Pos := 1_000_000_000; |
| -- Value of System.Tick in nanoseconds. At the moment, this is a fixed |
| -- constant (with value of 1.0 seconds), but later we should add this |
| -- value to the GCC configuration file so that its value can be made |
| -- configuration dependent. |
| |
| ----------------------------------------------------- |
| -- Target-Dependent Values for Types in Interfaces -- |
| ----------------------------------------------------- |
| |
| Interfaces_Wchar_T_Size : constant Pos := Set_Targ.Wchar_T_Size; |
| |
| ---------------------------------------- |
| -- Other Target-Dependent Definitions -- |
| ---------------------------------------- |
| |
| Maximum_Alignment : constant Pos := Set_Targ.Maximum_Alignment; |
| -- The maximum alignment, in storage units, that an object or type may |
| -- require on the target machine. |
| |
| System_Allocator_Alignment : constant Pos := |
| Set_Targ.System_Allocator_Alignment; |
| -- The alignment in storage units of addresses returned by malloc |
| |
| Max_Unaligned_Field : constant Pos := Set_Targ.Max_Unaligned_Field; |
| -- The maximum supported size in bits for a field that is not aligned |
| -- on a storage unit boundary. |
| |
| Bytes_Big_Endian : Boolean := Set_Targ.Bytes_BE /= 0; |
| -- Important note: for Ada purposes, the important setting is the bytes |
| -- endianness (Bytes_Big_Endian), not the bits value (Bits_Big_Endian). |
| -- This is because Ada bit addressing must be compatible with the byte |
| -- ordering (otherwise we would end up with non-contiguous fields). It |
| -- is rare for the two to be different, but if they are, Bits_Big_Endian |
| -- is relevant only for the generation of instructions with bit numbers, |
| -- and thus relevant only to the back end. Note that this is a variable |
| -- rather than a constant, since it can be modified (flipped) by -gnatd8. |
| |
| Target_Short_Enums : constant Boolean := Set_Targ.Short_Enums /= 0; |
| -- True if we are in short enums mode, where foreign convention |
| -- (in particular C and C++) enumeration types will be sized as in Ada, |
| -- using the shortest possibility from 8,16,32 bits, signed or unsigned. |
| -- A zero value means Short_Enums are not in use, and in this case all |
| -- foreign convention enumeration types are given the same size as c int. |
| |
| Target_Strict_Alignment : Boolean := |
| Set_Targ.Strict_Alignment /= 0; |
| -- True if instructions will fail if data is misaligned. Note that this |
| -- is a variable rather than a constant since it can be modified (set to |
| -- True) if the debug flag -gnatd.a is used. |
| |
| Target_Double_Float_Alignment : constant Nat := |
| Set_Targ.Double_Float_Alignment; |
| -- The default alignment of "double" floating-point types, i.e. floating |
| -- point types whose size is equal to 64 bits, or 0 if this alignment is |
| -- not lower than the largest power of 2 multiple of System.Storage_Unit |
| -- that does not exceed either the object size of the type or the maximum |
| -- allowed alignment. |
| |
| Target_Double_Scalar_Alignment : constant Nat := |
| Set_Targ.Double_Scalar_Alignment; |
| -- The default alignment of "double" or larger scalar types, i.e. scalar |
| -- types whose size is greater or equal to 64 bits, or 0 if this alignment |
| -- is not lower than the largest power of 2 multiple of System.Storage_Unit |
| -- that does not exceed either the object size of the type or the maximum |
| -- allowed alignment. |
| |
| end Ttypes; |
