3 <!-- This HTML file has been created by texi2html 1.54
4 from suif1.texi on 28 April 1999 -->
6 <TITLE>The SUIF Version
1 Library - Array Types
</TITLE>
7 <link href=
"suif1_64.html" rel=Next
>
8 <link href=
"suif1_62.html" rel=Previous
>
9 <link href=
"suif1_toc.html" rel=ToC
>
13 <p>Go to the
<A HREF=
"suif1_1.html">first
</A>,
<A HREF=
"suif1_62.html">previous
</A>,
<A HREF=
"suif1_64.html">next
</A>,
<A HREF=
"suif1_113.html">last
</A> section,
<A HREF=
"suif1_toc.html">table of contents
</A>.
17 <H2><A NAME=
"SEC63" HREF=
"suif1_toc.html#TOC63">Array Types
</A></H2>
24 A SUIF array type defines a new type that is a one-dimensional vector
25 with elements of another type. Multi-dimensional arrays are handled as
26 nested arrays (i.e. arrays of arrays) as in C. The size of an array is
27 specified by the upper and lower bounds. Each bound is either an
28 integer expression or unknown. If the bound is an integer expression,
29 it can either be an integer known at compile time, or a SUIF variable
30 symbol representing an integer value to be computed at run time.
42 The upper and lower bounds of an array type are stored in objects of the
43 <CODE>array_bound
</CODE> class. An array bound can be one of three things,
44 an integer constant, a SUIF variable symbol, or entirely unknown. The
45 <CODE>is_constant
</CODE>,
<CODE>is_variable
</CODE>, and
<CODE>is_unknown
</CODE> methods
46 check for these three conditions. If the bound is a constant, the
47 <CODE>constant
</CODE> method returns the integer value of the bound.
48 Likewise, if it is a variable, the
<CODE>variable
</CODE> method returns a
49 pointer to the variable symbol. The
<CODE>print
</CODE> method is also
50 available to print out the value of an array bound. The global
51 variable
<CODE>unknown_bound
</CODE> is a predefined array bound that can be
52 used when a bound is unknown.
63 A type node with a
<CODE>TYPE_ARRAY
</CODE> operator is stored in an
64 <CODE>array_type
</CODE> object. The
<CODE>array_type
</CODE> class contains three
65 fields: the lower and upper bounds and the element type. The
66 <CODE>lower_bound
</CODE>,
<CODE>upper_bound
</CODE>, and
<CODE>elem_type
</CODE> methods
67 retrieve the contents of these fields, and the
<CODE>set_lower_bound
</CODE>,
68 <CODE>set_upper_bound
</CODE>, and
<CODE>set_elem_type
</CODE> methods change their
69 values. The SUIF definition of arrays says that the elements will be
70 stored adjacently. If this would violate alignment restrictions, the
71 element types must be padded with extra space. For example, if some
72 machine has
24 bit integers with a
32 bit alignment requirement, the
73 24 bit integers may not be used directly as array elements; instead,
74 one would have to create a structure of size
32 bits that contains a
75 24 bit integer at offset zero and no other fields, and then use arrays
76 of this structure type.
80 The bounds of an array type tell two things: the number of elements
81 and the how the index of an array access instruction will be
82 interpreted. The number of elements determines the size of the array
83 -- the size will be the number of elements times the size of each
84 element. Any array type with an upper or lower bound that is anything
85 but an integer constant will not have a size that is known at compile
86 time, so such a type cannot be used anywhere a known size is required,
87 such as for a variable type, an expression type, or the type of an
88 object copied by a
<CODE>memcpy
</CODE> instruction.
93 The other time when array bounds are meaningful is in an array access
94 (a SUIF
<CODE>in_array
</CODE> instruction). There each index is required to
95 be between the given bounds or the result is undefined. The location
96 referred to is determined by subtracting the lower bound from the index
97 and then multiplying by the size of the next lower dimension, or
98 element size if there are no more dimensions in the array access
99 instruction. The size in this case is not necessarily known at
100 compile time and is considered to be the difference between the upper
101 and lower bounds plus one (the number of elements) times the size of
102 the next lower dimension, or the element size if this is the smallest
103 dimension. The bounds are evaluated after all sub-expressions of the
104 array access instruction have been evaluated and before the result of
105 the array access instruction itself. Any bounds needed in this
106 calculation must evaluate to integers when this array access
107 instruction is executed, so they must be integer constants or variable
108 symbols, and if variable symbols they must have defined values at this
109 point. Note that this implies that the array type for an array access
110 may not have unknown lower bounds and may have only one unknown upper
111 bound, that of the outermost dimension. The
<CODE>are_bounds_unknown
</CODE>
112 method checks if the upper bound of an array type is unknown.
116 Note also that all of this information from the array type is also
117 available in a different form from the
<CODE>in_array
</CODE> class, between
118 the
<CODE>bound
</CODE> expressions and
<CODE>offset_op
</CODE> expression. The
119 results are required to be the same at run time whether the array type
120 bound information is used or whether the
<CODE>bound
</CODE> and
121 <CODE>offset_op
</CODE> expressions are used, or else the resulting behavior
122 is undefined. This redundancy allows more flexibility in analyzing
123 array access patterns. Code transformation passes can treat the
124 <CODE>in_array
</CODE> sub-expressions like any other expressions and
125 propagate additional information in, and dependence analysis on
126 individual accesses can look at these general expressions. Code
127 dealing with objects and types only has to deal with the symbolic
132 Array bounds are subject to certain constraints. The upper bound may
133 never be less than the lower bound. If both are constants, this
134 constraint applies to those constants. If one or both are symbolic
135 bounds, this constraint must hold for every array access for which
136 this is the array type. Symbolic bounds are typically local variables
137 that are assigned once each at the very beginning of a procedure and
138 used only as bounds, so their values have the same scope as the types
139 they are used in. This corresponds directly to the declarations in
140 Fortran code of array types with non-constant bounds, but this
141 particular form is not required by SUIF, and heavily transformed code
142 may be somewhat different. What is required is that symbolic bounds
143 not be sub-variables or call-by-reference parameters, since these are
144 really just short-hand for other expressions (see section
<A HREF=
"suif1_89.html#SEC89">Features for Compiling Fortran
</A>).
149 <p>Go to the
<A HREF=
"suif1_1.html">first
</A>,
<A HREF=
"suif1_62.html">previous
</A>,
<A HREF=
"suif1_64.html">next
</A>,
<A HREF=
"suif1_113.html">last
</A> section,
<A HREF=
"suif1_toc.html">table of contents
</A>.