2 <IndexTerm><Primary>language, GHC</Primary></IndexTerm>
3 <IndexTerm><Primary>extensions, GHC</Primary></IndexTerm>
4 As with all known Haskell systems, GHC implements some extensions to
5 the language. To use them, you'll need to give a <Option>-fglasgow-exts</Option>
6 <IndexTerm><Primary>-fglasgow-exts option</Primary></IndexTerm> option.
10 Virtually all of the Glasgow extensions serve to give you access to
11 the underlying facilities with which we implement Haskell. Thus, you
12 can get at the Raw Iron, if you are willing to write some non-standard
13 code at a more primitive level. You need not be “stuck” on
14 performance because of the implementation costs of Haskell's
15 “high-level” features—you can always code “under” them. In an
16 extreme case, you can write all your time-critical code in C, and then
17 just glue it together with Haskell!
21 Executive summary of our extensions:
28 <Term>Unboxed types and primitive operations:</Term>
31 You can get right down to the raw machine types and operations;
32 included in this are “primitive arrays” (direct access to Big Wads
33 of Bytes). Please see <XRef LinkEnd="glasgow-unboxed"> and following.
39 <Term>Multi-parameter type classes:</Term>
42 GHC's type system supports extended type classes with multiple
43 parameters. Please see <XRef LinkEnd="multi-param-type-classes">.
49 <Term>Local universal quantification:</Term>
52 GHC's type system supports explicit universal quantification in
53 constructor fields and function arguments. This is useful for things
54 like defining <Literal>runST</Literal> from the state-thread world. See <XRef LinkEnd="universal-quantification">.
60 <Term>Extistentially quantification in data types:</Term>
63 Some or all of the type variables in a datatype declaration may be
64 <Emphasis>existentially quantified</Emphasis>. More details in <XRef LinkEnd="existential-quantification">.
70 <Term>Scoped type variables:</Term>
73 Scoped type variables enable the programmer to supply type signatures
74 for some nested declarations, where this would not be legal in Haskell
75 98. Details in <XRef LinkEnd="scoped-type-variables">.
81 <Term>Calling out to C:</Term>
84 Just what it sounds like. We provide <Emphasis>lots</Emphasis> of rope that you
85 can dangle around your neck. Please see <XRef LinkEnd="glasgow-ccalls">.
94 Pragmas are special instructions to the compiler placed in the source
95 file. The pragmas GHC supports are described in <XRef LinkEnd="pragmas">.
101 <Term>Rewrite rules:</Term>
104 The programmer can specify rewrite rules as part of the source program
105 (in a pragma). GHC applies these rewrite rules wherever it can.
106 Details in <XRef LinkEnd="rewrite-rules">.
114 Before you get too carried away working at the lowest level (e.g.,
115 sloshing <Literal>MutableByteArray#</Literal>s around your
116 program), you may wish to check if there are libraries that provide a
117 “Haskellised veneer” over the features you want. See
118 <xref linkend="book-hslibs">.
121 <Sect1 id="primitives">
122 <Title>Unboxed types and primitive operations
124 <IndexTerm><Primary>PrelGHC module</Primary></IndexTerm>
127 This module defines all the types which are primitive in Glasgow
128 Haskell, and the operations provided for them.
131 <Sect2 id="glasgow-unboxed">
136 <IndexTerm><Primary>Unboxed types (Glasgow extension)</Primary></IndexTerm>
139 <para>Most types in GHC are <firstterm>boxed</firstterm>, which means
140 that values of that type are represented by a pointer to a heap
141 object. The representation of a Haskell <literal>Int</literal>, for
142 example, is a two-word heap object. An <firstterm>unboxed</firstterm>
143 type, however, is represented by the value itself, no pointers or heap
144 allocation are involved.
148 Unboxed types correspond to the “raw machine” types you
149 would use in C: <Literal>Int#</Literal> (long int),
150 <Literal>Double#</Literal> (double), <Literal>Addr#</Literal>
151 (void *), etc. The <Emphasis>primitive operations</Emphasis>
152 (PrimOps) on these types are what you might expect; e.g.,
153 <Literal>(+#)</Literal> is addition on
154 <Literal>Int#</Literal>s, and is the machine-addition that we all
155 know and love—usually one instruction.
159 Primitive (unboxed) types cannot be defined in Haskell, and are
160 therefore built into the language and compiler. Primitive types are
161 always unlifted; that is, a value of a primitive type cannot be
162 bottom. We use the convention that primitive types, values, and
163 operations have a <Literal>#</Literal> suffix.
167 Primitive values are often represented by a simple bit-pattern, such
168 as <Literal>Int#</Literal>, <Literal>Float#</Literal>,
169 <Literal>Double#</Literal>. But this is not necessarily the case:
170 a primitive value might be represented by a pointer to a
171 heap-allocated object. Examples include
172 <Literal>Array#</Literal>, the type of primitive arrays. A
173 primitive array is heap-allocated because it is too big a value to fit
174 in a register, and would be too expensive to copy around; in a sense,
175 it is accidental that it is represented by a pointer. If a pointer
176 represents a primitive value, then it really does point to that value:
177 no unevaluated thunks, no indirections…nothing can be at the
178 other end of the pointer than the primitive value.
182 There are some restrictions on the use of primitive types, the main
183 one being that you can't pass a primitive value to a polymorphic
184 function or store one in a polymorphic data type. This rules out
185 things like <Literal>[Int#]</Literal> (i.e. lists of primitive
186 integers). The reason for this restriction is that polymorphic
187 arguments and constructor fields are assumed to be pointers: if an
188 unboxed integer is stored in one of these, the garbage collector would
189 attempt to follow it, leading to unpredictable space leaks. Or a
190 <Function>seq</Function> operation on the polymorphic component may
191 attempt to dereference the pointer, with disastrous results. Even
192 worse, the unboxed value might be larger than a pointer
193 (<Literal>Double#</Literal> for instance).
197 Nevertheless, A numerically-intensive program using unboxed types can
198 go a <Emphasis>lot</Emphasis> faster than its “standard”
199 counterpart—we saw a threefold speedup on one example.
204 <Sect2 id="unboxed-tuples">
205 <Title>Unboxed Tuples
209 Unboxed tuples aren't really exported by <Literal>PrelGHC</Literal>,
210 they're available by default with <Option>-fglasgow-exts</Option>. An
211 unboxed tuple looks like this:
223 where <Literal>e_1..e_n</Literal> are expressions of any
224 type (primitive or non-primitive). The type of an unboxed tuple looks
229 Unboxed tuples are used for functions that need to return multiple
230 values, but they avoid the heap allocation normally associated with
231 using fully-fledged tuples. When an unboxed tuple is returned, the
232 components are put directly into registers or on the stack; the
233 unboxed tuple itself does not have a composite representation. Many
234 of the primitive operations listed in this section return unboxed
239 There are some pretty stringent restrictions on the use of unboxed tuples:
248 Unboxed tuple types are subject to the same restrictions as
249 other unboxed types; i.e. they may not be stored in polymorphic data
250 structures or passed to polymorphic functions.
257 Unboxed tuples may only be constructed as the direct result of
258 a function, and may only be deconstructed with a <Literal>case</Literal> expression.
259 eg. the following are valid:
263 f x y = (# x+1, y-1 #)
264 g x = case f x x of { (# a, b #) -> a + b }
268 but the following are invalid:
282 No variable can have an unboxed tuple type. This is illegal:
286 f :: (# Int, Int #) -> (# Int, Int #)
291 because <VarName>x</VarName> has an unboxed tuple type.
301 Note: we may relax some of these restrictions in the future.
305 The <Literal>IO</Literal> and <Literal>ST</Literal> monads use unboxed tuples to avoid unnecessary
306 allocation during sequences of operations.
312 <Title>Character and numeric types</Title>
315 <IndexTerm><Primary>character types, primitive</Primary></IndexTerm>
316 <IndexTerm><Primary>numeric types, primitive</Primary></IndexTerm>
317 <IndexTerm><Primary>integer types, primitive</Primary></IndexTerm>
318 <IndexTerm><Primary>floating point types, primitive</Primary></IndexTerm>
319 There are the following obvious primitive types:
335 <IndexTerm><Primary><literal>Char#</literal></Primary></IndexTerm>
336 <IndexTerm><Primary><literal>Int#</literal></Primary></IndexTerm>
337 <IndexTerm><Primary><literal>Word#</literal></Primary></IndexTerm>
338 <IndexTerm><Primary><literal>Addr#</literal></Primary></IndexTerm>
339 <IndexTerm><Primary><literal>Float#</literal></Primary></IndexTerm>
340 <IndexTerm><Primary><literal>Double#</literal></Primary></IndexTerm>
341 <IndexTerm><Primary><literal>Int64#</literal></Primary></IndexTerm>
342 <IndexTerm><Primary><literal>Word64#</literal></Primary></IndexTerm>
346 If you really want to know their exact equivalents in C, see
347 <Filename>ghc/includes/StgTypes.h</Filename> in the GHC source tree.
351 Literals for these types may be written as follows:
360 'a'# a Char#; for weird characters, use '\o<octal>'#
361 "a"# an Addr# (a `char *')
364 <IndexTerm><Primary>literals, primitive</Primary></IndexTerm>
365 <IndexTerm><Primary>constants, primitive</Primary></IndexTerm>
366 <IndexTerm><Primary>numbers, primitive</Primary></IndexTerm>
372 <Title>Comparison operations</Title>
375 <IndexTerm><Primary>comparisons, primitive</Primary></IndexTerm>
376 <IndexTerm><Primary>operators, comparison</Primary></IndexTerm>
382 {>,>=,==,/=,<,<=}# :: Int# -> Int# -> Bool
384 {gt,ge,eq,ne,lt,le}Char# :: Char# -> Char# -> Bool
385 -- ditto for Word# and Addr#
388 <IndexTerm><Primary><literal>>#</literal></Primary></IndexTerm>
389 <IndexTerm><Primary><literal>>=#</literal></Primary></IndexTerm>
390 <IndexTerm><Primary><literal>==#</literal></Primary></IndexTerm>
391 <IndexTerm><Primary><literal>/=#</literal></Primary></IndexTerm>
392 <IndexTerm><Primary><literal><#</literal></Primary></IndexTerm>
393 <IndexTerm><Primary><literal><=#</literal></Primary></IndexTerm>
394 <IndexTerm><Primary><literal>gt{Char,Word,Addr}#</literal></Primary></IndexTerm>
395 <IndexTerm><Primary><literal>ge{Char,Word,Addr}#</literal></Primary></IndexTerm>
396 <IndexTerm><Primary><literal>eq{Char,Word,Addr}#</literal></Primary></IndexTerm>
397 <IndexTerm><Primary><literal>ne{Char,Word,Addr}#</literal></Primary></IndexTerm>
398 <IndexTerm><Primary><literal>lt{Char,Word,Addr}#</literal></Primary></IndexTerm>
399 <IndexTerm><Primary><literal>le{Char,Word,Addr}#</literal></Primary></IndexTerm>
405 <Title>Primitive-character operations</Title>
408 <IndexTerm><Primary>characters, primitive operations</Primary></IndexTerm>
409 <IndexTerm><Primary>operators, primitive character</Primary></IndexTerm>
415 ord# :: Char# -> Int#
416 chr# :: Int# -> Char#
419 <IndexTerm><Primary><literal>ord#</literal></Primary></IndexTerm>
420 <IndexTerm><Primary><literal>chr#</literal></Primary></IndexTerm>
426 <Title>Primitive-<Literal>Int</Literal> operations</Title>
429 <IndexTerm><Primary>integers, primitive operations</Primary></IndexTerm>
430 <IndexTerm><Primary>operators, primitive integer</Primary></IndexTerm>
436 {+,-,*,quotInt,remInt,gcdInt}# :: Int# -> Int# -> Int#
437 negateInt# :: Int# -> Int#
439 iShiftL#, iShiftRA#, iShiftRL# :: Int# -> Int# -> Int#
440 -- shift left, right arithmetic, right logical
442 addIntC#, subIntC#, mulIntC# :: Int# -> Int# -> (# Int#, Int# #)
443 -- add, subtract, multiply with carry
446 <IndexTerm><Primary><literal>+#</literal></Primary></IndexTerm>
447 <IndexTerm><Primary><literal>-#</literal></Primary></IndexTerm>
448 <IndexTerm><Primary><literal>*#</literal></Primary></IndexTerm>
449 <IndexTerm><Primary><literal>quotInt#</literal></Primary></IndexTerm>
450 <IndexTerm><Primary><literal>remInt#</literal></Primary></IndexTerm>
451 <IndexTerm><Primary><literal>gcdInt#</literal></Primary></IndexTerm>
452 <IndexTerm><Primary><literal>iShiftL#</literal></Primary></IndexTerm>
453 <IndexTerm><Primary><literal>iShiftRA#</literal></Primary></IndexTerm>
454 <IndexTerm><Primary><literal>iShiftRL#</literal></Primary></IndexTerm>
455 <IndexTerm><Primary><literal>addIntC#</literal></Primary></IndexTerm>
456 <IndexTerm><Primary><literal>subIntC#</literal></Primary></IndexTerm>
457 <IndexTerm><Primary><literal>mulIntC#</literal></Primary></IndexTerm>
458 <IndexTerm><Primary>shift operations, integer</Primary></IndexTerm>
462 <Emphasis>Note:</Emphasis> No error/overflow checking!
468 <Title>Primitive-<Literal>Double</Literal> and <Literal>Float</Literal> operations</Title>
471 <IndexTerm><Primary>floating point numbers, primitive</Primary></IndexTerm>
472 <IndexTerm><Primary>operators, primitive floating point</Primary></IndexTerm>
478 {+,-,*,/}## :: Double# -> Double# -> Double#
479 {<,<=,==,/=,>=,>}## :: Double# -> Double# -> Bool
480 negateDouble# :: Double# -> Double#
481 double2Int# :: Double# -> Int#
482 int2Double# :: Int# -> Double#
484 {plus,minux,times,divide}Float# :: Float# -> Float# -> Float#
485 {gt,ge,eq,ne,lt,le}Float# :: Float# -> Float# -> Bool
486 negateFloat# :: Float# -> Float#
487 float2Int# :: Float# -> Int#
488 int2Float# :: Int# -> Float#
494 <IndexTerm><Primary><literal>+##</literal></Primary></IndexTerm>
495 <IndexTerm><Primary><literal>-##</literal></Primary></IndexTerm>
496 <IndexTerm><Primary><literal>*##</literal></Primary></IndexTerm>
497 <IndexTerm><Primary><literal>/##</literal></Primary></IndexTerm>
498 <IndexTerm><Primary><literal><##</literal></Primary></IndexTerm>
499 <IndexTerm><Primary><literal><=##</literal></Primary></IndexTerm>
500 <IndexTerm><Primary><literal>==##</literal></Primary></IndexTerm>
501 <IndexTerm><Primary><literal>=/##</literal></Primary></IndexTerm>
502 <IndexTerm><Primary><literal>>=##</literal></Primary></IndexTerm>
503 <IndexTerm><Primary><literal>>##</literal></Primary></IndexTerm>
504 <IndexTerm><Primary><literal>negateDouble#</literal></Primary></IndexTerm>
505 <IndexTerm><Primary><literal>double2Int#</literal></Primary></IndexTerm>
506 <IndexTerm><Primary><literal>int2Double#</literal></Primary></IndexTerm>
510 <IndexTerm><Primary><literal>plusFloat#</literal></Primary></IndexTerm>
511 <IndexTerm><Primary><literal>minusFloat#</literal></Primary></IndexTerm>
512 <IndexTerm><Primary><literal>timesFloat#</literal></Primary></IndexTerm>
513 <IndexTerm><Primary><literal>divideFloat#</literal></Primary></IndexTerm>
514 <IndexTerm><Primary><literal>gtFloat#</literal></Primary></IndexTerm>
515 <IndexTerm><Primary><literal>geFloat#</literal></Primary></IndexTerm>
516 <IndexTerm><Primary><literal>eqFloat#</literal></Primary></IndexTerm>
517 <IndexTerm><Primary><literal>neFloat#</literal></Primary></IndexTerm>
518 <IndexTerm><Primary><literal>ltFloat#</literal></Primary></IndexTerm>
519 <IndexTerm><Primary><literal>leFloat#</literal></Primary></IndexTerm>
520 <IndexTerm><Primary><literal>negateFloat#</literal></Primary></IndexTerm>
521 <IndexTerm><Primary><literal>float2Int#</literal></Primary></IndexTerm>
522 <IndexTerm><Primary><literal>int2Float#</literal></Primary></IndexTerm>
526 And a full complement of trigonometric functions:
532 expDouble# :: Double# -> Double#
533 logDouble# :: Double# -> Double#
534 sqrtDouble# :: Double# -> Double#
535 sinDouble# :: Double# -> Double#
536 cosDouble# :: Double# -> Double#
537 tanDouble# :: Double# -> Double#
538 asinDouble# :: Double# -> Double#
539 acosDouble# :: Double# -> Double#
540 atanDouble# :: Double# -> Double#
541 sinhDouble# :: Double# -> Double#
542 coshDouble# :: Double# -> Double#
543 tanhDouble# :: Double# -> Double#
544 powerDouble# :: Double# -> Double# -> Double#
547 <IndexTerm><Primary>trigonometric functions, primitive</Primary></IndexTerm>
551 similarly for <Literal>Float#</Literal>.
555 There are two coercion functions for <Literal>Float#</Literal>/<Literal>Double#</Literal>:
561 float2Double# :: Float# -> Double#
562 double2Float# :: Double# -> Float#
565 <IndexTerm><Primary><literal>float2Double#</literal></Primary></IndexTerm>
566 <IndexTerm><Primary><literal>double2Float#</literal></Primary></IndexTerm>
570 The primitive version of <Function>decodeDouble</Function>
571 (<Function>encodeDouble</Function> is implemented as an external C
578 decodeDouble# :: Double# -> PrelNum.ReturnIntAndGMP
581 <IndexTerm><Primary><literal>encodeDouble#</literal></Primary></IndexTerm>
582 <IndexTerm><Primary><literal>decodeDouble#</literal></Primary></IndexTerm>
586 (And the same for <Literal>Float#</Literal>s.)
591 <Sect2 id="integer-operations">
592 <Title>Operations on/for <Literal>Integers</Literal> (interface to GMP)
596 <IndexTerm><Primary>arbitrary precision integers</Primary></IndexTerm>
597 <IndexTerm><Primary>Integer, operations on</Primary></IndexTerm>
601 We implement <Literal>Integers</Literal> (arbitrary-precision
602 integers) using the GNU multiple-precision (GMP) package (version
607 The data type for <Literal>Integer</Literal> is either a small
608 integer, represented by an <Literal>Int</Literal>, or a large integer
609 represented using the pieces required by GMP's
610 <Literal>MP_INT</Literal> in <Filename>gmp.h</Filename> (see
611 <Filename>gmp.info</Filename> in
612 <Filename>ghc/includes/runtime/gmp</Filename>). It comes out as:
618 data Integer = S# Int# -- small integers
619 | J# Int# ByteArray# -- large integers
622 <IndexTerm><Primary>Integer type</Primary></IndexTerm> The primitive
623 ops to support large <Literal>Integers</Literal> use the
624 “pieces” of the representation, and are as follows:
630 negateInteger# :: Int# -> ByteArray# -> Integer
632 {plus,minus,times}Integer#, gcdInteger#,
633 quotInteger#, remInteger#, divExactInteger#
634 :: Int# -> ByteArray#
635 -> Int# -> ByteArray#
636 -> (# Int#, ByteArray# #)
639 :: Int# -> ByteArray#
640 -> Int# -> ByteArray#
641 -> Int# -- -1 for <; 0 for ==; +1 for >
644 :: Int# -> ByteArray#
646 -> Int# -- -1 for <; 0 for ==; +1 for >
649 :: Int# -> ByteArray#
653 divModInteger#, quotRemInteger#
654 :: Int# -> ByteArray#
655 -> Int# -> ByteArray#
656 -> (# Int#, ByteArray#,
659 integer2Int# :: Int# -> ByteArray# -> Int#
661 int2Integer# :: Int# -> Integer -- NB: no error-checking on these two!
662 word2Integer# :: Word# -> Integer
664 addr2Integer# :: Addr# -> Integer
665 -- the Addr# is taken to be a `char *' string
666 -- to be converted into an Integer.
669 <IndexTerm><Primary><literal>negateInteger#</literal></Primary></IndexTerm>
670 <IndexTerm><Primary><literal>plusInteger#</literal></Primary></IndexTerm>
671 <IndexTerm><Primary><literal>minusInteger#</literal></Primary></IndexTerm>
672 <IndexTerm><Primary><literal>timesInteger#</literal></Primary></IndexTerm>
673 <IndexTerm><Primary><literal>quotInteger#</literal></Primary></IndexTerm>
674 <IndexTerm><Primary><literal>remInteger#</literal></Primary></IndexTerm>
675 <IndexTerm><Primary><literal>gcdInteger#</literal></Primary></IndexTerm>
676 <IndexTerm><Primary><literal>gcdIntegerInt#</literal></Primary></IndexTerm>
677 <IndexTerm><Primary><literal>divExactInteger#</literal></Primary></IndexTerm>
678 <IndexTerm><Primary><literal>cmpInteger#</literal></Primary></IndexTerm>
679 <IndexTerm><Primary><literal>divModInteger#</literal></Primary></IndexTerm>
680 <IndexTerm><Primary><literal>quotRemInteger#</literal></Primary></IndexTerm>
681 <IndexTerm><Primary><literal>integer2Int#</literal></Primary></IndexTerm>
682 <IndexTerm><Primary><literal>int2Integer#</literal></Primary></IndexTerm>
683 <IndexTerm><Primary><literal>word2Integer#</literal></Primary></IndexTerm>
684 <IndexTerm><Primary><literal>addr2Integer#</literal></Primary></IndexTerm>
690 <Title>Words and addresses</Title>
693 <IndexTerm><Primary>word, primitive type</Primary></IndexTerm>
694 <IndexTerm><Primary>address, primitive type</Primary></IndexTerm>
695 <IndexTerm><Primary>unsigned integer, primitive type</Primary></IndexTerm>
696 <IndexTerm><Primary>pointer, primitive type</Primary></IndexTerm>
700 A <Literal>Word#</Literal> is used for bit-twiddling operations.
701 It is the same size as an <Literal>Int#</Literal>, but has no sign
702 nor any arithmetic operations.
705 type Word# -- Same size/etc as Int# but *unsigned*
706 type Addr# -- A pointer from outside the "Haskell world" (from C, probably);
707 -- described under "arrays"
710 <IndexTerm><Primary><literal>Word#</literal></Primary></IndexTerm>
711 <IndexTerm><Primary><literal>Addr#</literal></Primary></IndexTerm>
715 <Literal>Word#</Literal>s and <Literal>Addr#</Literal>s have
716 the usual comparison operations. Other
717 unboxed-<Literal>Word</Literal> ops (bit-twiddling and coercions):
723 {gt,ge,eq,ne,lt,le}Word# :: Word# -> Word# -> Bool
725 and#, or#, xor# :: Word# -> Word# -> Word#
728 quotWord#, remWord# :: Word# -> Word# -> Word#
729 -- word (i.e. unsigned) versions are different from int
730 -- versions, so we have to provide these explicitly.
732 not# :: Word# -> Word#
734 shiftL#, shiftRL# :: Word# -> Int# -> Word#
735 -- shift left, right logical
737 int2Word# :: Int# -> Word# -- just a cast, really
738 word2Int# :: Word# -> Int#
741 <IndexTerm><Primary>bit operations, Word and Addr</Primary></IndexTerm>
742 <IndexTerm><Primary><literal>gtWord#</literal></Primary></IndexTerm>
743 <IndexTerm><Primary><literal>geWord#</literal></Primary></IndexTerm>
744 <IndexTerm><Primary><literal>eqWord#</literal></Primary></IndexTerm>
745 <IndexTerm><Primary><literal>neWord#</literal></Primary></IndexTerm>
746 <IndexTerm><Primary><literal>ltWord#</literal></Primary></IndexTerm>
747 <IndexTerm><Primary><literal>leWord#</literal></Primary></IndexTerm>
748 <IndexTerm><Primary><literal>and#</literal></Primary></IndexTerm>
749 <IndexTerm><Primary><literal>or#</literal></Primary></IndexTerm>
750 <IndexTerm><Primary><literal>xor#</literal></Primary></IndexTerm>
751 <IndexTerm><Primary><literal>not#</literal></Primary></IndexTerm>
752 <IndexTerm><Primary><literal>quotWord#</literal></Primary></IndexTerm>
753 <IndexTerm><Primary><literal>remWord#</literal></Primary></IndexTerm>
754 <IndexTerm><Primary><literal>shiftL#</literal></Primary></IndexTerm>
755 <IndexTerm><Primary><literal>shiftRA#</literal></Primary></IndexTerm>
756 <IndexTerm><Primary><literal>shiftRL#</literal></Primary></IndexTerm>
757 <IndexTerm><Primary><literal>int2Word#</literal></Primary></IndexTerm>
758 <IndexTerm><Primary><literal>word2Int#</literal></Primary></IndexTerm>
762 Unboxed-<Literal>Addr</Literal> ops (C casts, really):
765 {gt,ge,eq,ne,lt,le}Addr# :: Addr# -> Addr# -> Bool
767 int2Addr# :: Int# -> Addr#
768 addr2Int# :: Addr# -> Int#
769 addr2Integer# :: Addr# -> (# Int#, ByteArray# #)
772 <IndexTerm><Primary><literal>gtAddr#</literal></Primary></IndexTerm>
773 <IndexTerm><Primary><literal>geAddr#</literal></Primary></IndexTerm>
774 <IndexTerm><Primary><literal>eqAddr#</literal></Primary></IndexTerm>
775 <IndexTerm><Primary><literal>neAddr#</literal></Primary></IndexTerm>
776 <IndexTerm><Primary><literal>ltAddr#</literal></Primary></IndexTerm>
777 <IndexTerm><Primary><literal>leAddr#</literal></Primary></IndexTerm>
778 <IndexTerm><Primary><literal>int2Addr#</literal></Primary></IndexTerm>
779 <IndexTerm><Primary><literal>addr2Int#</literal></Primary></IndexTerm>
780 <IndexTerm><Primary><literal>addr2Integer#</literal></Primary></IndexTerm>
784 The casts between <Literal>Int#</Literal>,
785 <Literal>Word#</Literal> and <Literal>Addr#</Literal>
786 correspond to null operations at the machine level, but are required
787 to keep the Haskell type checker happy.
791 Operations for indexing off of C pointers
792 (<Literal>Addr#</Literal>s) to snatch values are listed under
793 “arrays”.
799 <Title>Arrays</Title>
802 <IndexTerm><Primary>arrays, primitive</Primary></IndexTerm>
806 The type <Literal>Array# elt</Literal> is the type of primitive,
807 unpointed arrays of values of type <Literal>elt</Literal>.
816 <IndexTerm><Primary><literal>Array#</literal></Primary></IndexTerm>
820 <Literal>Array#</Literal> is more primitive than a Haskell
821 array—indeed, the Haskell <Literal>Array</Literal> interface is
822 implemented using <Literal>Array#</Literal>—in that an
823 <Literal>Array#</Literal> is indexed only by
824 <Literal>Int#</Literal>s, starting at zero. It is also more
825 primitive by virtue of being unboxed. That doesn't mean that it isn't
826 a heap-allocated object—of course, it is. Rather, being unboxed
827 means that it is represented by a pointer to the array itself, and not
828 to a thunk which will evaluate to the array (or to bottom). The
829 components of an <Literal>Array#</Literal> are themselves boxed.
833 The type <Literal>ByteArray#</Literal> is similar to
834 <Literal>Array#</Literal>, except that it contains just a string
835 of (non-pointer) bytes.
844 <IndexTerm><Primary><literal>ByteArray#</literal></Primary></IndexTerm>
848 Arrays of these types are useful when a Haskell program wishes to
849 construct a value to pass to a C procedure. It is also possible to use
850 them to build (say) arrays of unboxed characters for internal use in a
851 Haskell program. Given these uses, <Literal>ByteArray#</Literal>
852 is deliberately a bit vague about the type of its components.
853 Operations are provided to extract values of type
854 <Literal>Char#</Literal>, <Literal>Int#</Literal>,
855 <Literal>Float#</Literal>, <Literal>Double#</Literal>, and
856 <Literal>Addr#</Literal> from arbitrary offsets within a
857 <Literal>ByteArray#</Literal>. (For type
858 <Literal>Foo#</Literal>, the $i$th offset gets you the $i$th
859 <Literal>Foo#</Literal>, not the <Literal>Foo#</Literal> at
860 byte-position $i$. Mumble.) (If you want a
861 <Literal>Word#</Literal>, grab an <Literal>Int#</Literal>,
866 Lastly, we have static byte-arrays, of type
867 <Literal>Addr#</Literal> [mentioned previously]. (Remember
868 the duality between arrays and pointers in C.) Arrays of this types
869 are represented by a pointer to an array in the world outside Haskell,
870 so this pointer is not followed by the garbage collector. In other
871 respects they are just like <Literal>ByteArray#</Literal>. They
872 are only needed in order to pass values from C to Haskell.
878 <Title>Reading and writing</Title>
881 Primitive arrays are linear, and indexed starting at zero.
885 The size and indices of a <Literal>ByteArray#</Literal>, <Literal>Addr#</Literal>, and
886 <Literal>MutableByteArray#</Literal> are all in bytes. It's up to the program to
887 calculate the correct byte offset from the start of the array. This
888 allows a <Literal>ByteArray#</Literal> to contain a mixture of values of different
889 type, which is often needed when preparing data for and unpicking
890 results from C. (Umm…not true of indices…WDP 95/09)
894 <Emphasis>Should we provide some <Literal>sizeOfDouble#</Literal> constants?</Emphasis>
898 Out-of-range errors on indexing should be caught by the code which
899 uses the primitive operation; the primitive operations themselves do
900 <Emphasis>not</Emphasis> check for out-of-range indexes. The intention is that the
901 primitive ops compile to one machine instruction or thereabouts.
905 We use the terms “reading” and “writing” to refer to accessing
906 <Emphasis>mutable</Emphasis> arrays (see <XRef LinkEnd="sect-mutable">), and
907 “indexing” to refer to reading a value from an <Emphasis>immutable</Emphasis>
912 Immutable byte arrays are straightforward to index (all indices in bytes):
915 indexCharArray# :: ByteArray# -> Int# -> Char#
916 indexIntArray# :: ByteArray# -> Int# -> Int#
917 indexAddrArray# :: ByteArray# -> Int# -> Addr#
918 indexFloatArray# :: ByteArray# -> Int# -> Float#
919 indexDoubleArray# :: ByteArray# -> Int# -> Double#
921 indexCharOffAddr# :: Addr# -> Int# -> Char#
922 indexIntOffAddr# :: Addr# -> Int# -> Int#
923 indexFloatOffAddr# :: Addr# -> Int# -> Float#
924 indexDoubleOffAddr# :: Addr# -> Int# -> Double#
925 indexAddrOffAddr# :: Addr# -> Int# -> Addr#
926 -- Get an Addr# from an Addr# offset
929 <IndexTerm><Primary><literal>indexCharArray#</literal></Primary></IndexTerm>
930 <IndexTerm><Primary><literal>indexIntArray#</literal></Primary></IndexTerm>
931 <IndexTerm><Primary><literal>indexAddrArray#</literal></Primary></IndexTerm>
932 <IndexTerm><Primary><literal>indexFloatArray#</literal></Primary></IndexTerm>
933 <IndexTerm><Primary><literal>indexDoubleArray#</literal></Primary></IndexTerm>
934 <IndexTerm><Primary><literal>indexCharOffAddr#</literal></Primary></IndexTerm>
935 <IndexTerm><Primary><literal>indexIntOffAddr#</literal></Primary></IndexTerm>
936 <IndexTerm><Primary><literal>indexFloatOffAddr#</literal></Primary></IndexTerm>
937 <IndexTerm><Primary><literal>indexDoubleOffAddr#</literal></Primary></IndexTerm>
938 <IndexTerm><Primary><literal>indexAddrOffAddr#</literal></Primary></IndexTerm>
942 The last of these, <Function>indexAddrOffAddr#</Function>, extracts an <Literal>Addr#</Literal> using an offset
943 from another <Literal>Addr#</Literal>, thereby providing the ability to follow a chain of
948 Something a bit more interesting goes on when indexing arrays of boxed
949 objects, because the result is simply the boxed object. So presumably
950 it should be entered—we never usually return an unevaluated
951 object! This is a pain: primitive ops aren't supposed to do
952 complicated things like enter objects. The current solution is to
953 return a single element unboxed tuple (see <XRef LinkEnd="unboxed-tuples">).
959 indexArray# :: Array# elt -> Int# -> (# elt #)
962 <IndexTerm><Primary><literal>indexArray#</literal></Primary></IndexTerm>
968 <Title>The state type</Title>
971 <IndexTerm><Primary><literal>state, primitive type</literal></Primary></IndexTerm>
972 <IndexTerm><Primary><literal>State#</literal></Primary></IndexTerm>
976 The primitive type <Literal>State#</Literal> represents the state of a state
977 transformer. It is parameterised on the desired type of state, which
978 serves to keep states from distinct threads distinct from one another.
979 But the <Emphasis>only</Emphasis> effect of this parameterisation is in the type
980 system: all values of type <Literal>State#</Literal> are represented in the same way.
981 Indeed, they are all represented by nothing at all! The code
982 generator “knows” to generate no code, and allocate no registers
983 etc, for primitive states.
995 The type <Literal>GHC.RealWorld</Literal> is truly opaque: there are no values defined
996 of this type, and no operations over it. It is “primitive” in that
997 sense - but it is <Emphasis>not unlifted!</Emphasis> Its only role in life is to be
998 the type which distinguishes the <Literal>IO</Literal> state transformer.
1012 <Title>State of the world</Title>
1015 A single, primitive, value of type <Literal>State# RealWorld</Literal> is provided.
1021 realWorld# :: State# RealWorld
1024 <IndexTerm><Primary>realWorld# state object</Primary></IndexTerm>
1028 (Note: in the compiler, not a <Literal>PrimOp</Literal>; just a mucho magic
1029 <Literal>Id</Literal>. Exported from <Literal>GHC</Literal>, though).
1034 <Sect2 id="sect-mutable">
1035 <Title>Mutable arrays</Title>
1038 <IndexTerm><Primary>mutable arrays</Primary></IndexTerm>
1039 <IndexTerm><Primary>arrays, mutable</Primary></IndexTerm>
1040 Corresponding to <Literal>Array#</Literal> and <Literal>ByteArray#</Literal>, we have the types of
1041 mutable versions of each. In each case, the representation is a
1042 pointer to a suitable block of (mutable) heap-allocated storage.
1048 type MutableArray# s elt
1049 type MutableByteArray# s
1052 <IndexTerm><Primary><literal>MutableArray#</literal></Primary></IndexTerm>
1053 <IndexTerm><Primary><literal>MutableByteArray#</literal></Primary></IndexTerm>
1057 <Title>Allocation</Title>
1060 <IndexTerm><Primary>mutable arrays, allocation</Primary></IndexTerm>
1061 <IndexTerm><Primary>arrays, allocation</Primary></IndexTerm>
1062 <IndexTerm><Primary>allocation, of mutable arrays</Primary></IndexTerm>
1066 Mutable arrays can be allocated. Only pointer-arrays are initialised;
1067 arrays of non-pointers are filled in by “user code” rather than by
1068 the array-allocation primitive. Reason: only the pointer case has to
1069 worry about GC striking with a partly-initialised array.
1075 newArray# :: Int# -> elt -> State# s -> (# State# s, MutableArray# s elt #)
1077 newCharArray# :: Int# -> State# s -> (# State# s, MutableByteArray# s elt #)
1078 newIntArray# :: Int# -> State# s -> (# State# s, MutableByteArray# s elt #)
1079 newAddrArray# :: Int# -> State# s -> (# State# s, MutableByteArray# s elt #)
1080 newFloatArray# :: Int# -> State# s -> (# State# s, MutableByteArray# s elt #)
1081 newDoubleArray# :: Int# -> State# s -> (# State# s, MutableByteArray# s elt #)
1084 <IndexTerm><Primary><literal>newArray#</literal></Primary></IndexTerm>
1085 <IndexTerm><Primary><literal>newCharArray#</literal></Primary></IndexTerm>
1086 <IndexTerm><Primary><literal>newIntArray#</literal></Primary></IndexTerm>
1087 <IndexTerm><Primary><literal>newAddrArray#</literal></Primary></IndexTerm>
1088 <IndexTerm><Primary><literal>newFloatArray#</literal></Primary></IndexTerm>
1089 <IndexTerm><Primary><literal>newDoubleArray#</literal></Primary></IndexTerm>
1093 The size of a <Literal>ByteArray#</Literal> is given in bytes.
1099 <Title>Reading and writing</Title>
1102 <IndexTerm><Primary>arrays, reading and writing</Primary></IndexTerm>
1108 readArray# :: MutableArray# s elt -> Int# -> State# s -> (# State# s, elt #)
1109 readCharArray# :: MutableByteArray# s -> Int# -> State# s -> (# State# s, Char# #)
1110 readIntArray# :: MutableByteArray# s -> Int# -> State# s -> (# State# s, Int# #)
1111 readAddrArray# :: MutableByteArray# s -> Int# -> State# s -> (# State# s, Addr# #)
1112 readFloatArray# :: MutableByteArray# s -> Int# -> State# s -> (# State# s, Float# #)
1113 readDoubleArray# :: MutableByteArray# s -> Int# -> State# s -> (# State# s, Double# #)
1115 writeArray# :: MutableArray# s elt -> Int# -> elt -> State# s -> State# s
1116 writeCharArray# :: MutableByteArray# s -> Int# -> Char# -> State# s -> State# s
1117 writeIntArray# :: MutableByteArray# s -> Int# -> Int# -> State# s -> State# s
1118 writeAddrArray# :: MutableByteArray# s -> Int# -> Addr# -> State# s -> State# s
1119 writeFloatArray# :: MutableByteArray# s -> Int# -> Float# -> State# s -> State# s
1120 writeDoubleArray# :: MutableByteArray# s -> Int# -> Double# -> State# s -> State# s
1123 <IndexTerm><Primary><literal>readArray#</literal></Primary></IndexTerm>
1124 <IndexTerm><Primary><literal>readCharArray#</literal></Primary></IndexTerm>
1125 <IndexTerm><Primary><literal>readIntArray#</literal></Primary></IndexTerm>
1126 <IndexTerm><Primary><literal>readAddrArray#</literal></Primary></IndexTerm>
1127 <IndexTerm><Primary><literal>readFloatArray#</literal></Primary></IndexTerm>
1128 <IndexTerm><Primary><literal>readDoubleArray#</literal></Primary></IndexTerm>
1129 <IndexTerm><Primary><literal>writeArray#</literal></Primary></IndexTerm>
1130 <IndexTerm><Primary><literal>writeCharArray#</literal></Primary></IndexTerm>
1131 <IndexTerm><Primary><literal>writeIntArray#</literal></Primary></IndexTerm>
1132 <IndexTerm><Primary><literal>writeAddrArray#</literal></Primary></IndexTerm>
1133 <IndexTerm><Primary><literal>writeFloatArray#</literal></Primary></IndexTerm>
1134 <IndexTerm><Primary><literal>writeDoubleArray#</literal></Primary></IndexTerm>
1140 <Title>Equality</Title>
1143 <IndexTerm><Primary>arrays, testing for equality</Primary></IndexTerm>
1147 One can take “equality” of mutable arrays. What is compared is the
1148 <Emphasis>name</Emphasis> or reference to the mutable array, not its contents.
1154 sameMutableArray# :: MutableArray# s elt -> MutableArray# s elt -> Bool
1155 sameMutableByteArray# :: MutableByteArray# s -> MutableByteArray# s -> Bool
1158 <IndexTerm><Primary><literal>sameMutableArray#</literal></Primary></IndexTerm>
1159 <IndexTerm><Primary><literal>sameMutableByteArray#</literal></Primary></IndexTerm>
1165 <Title>Freezing mutable arrays</Title>
1168 <IndexTerm><Primary>arrays, freezing mutable</Primary></IndexTerm>
1169 <IndexTerm><Primary>freezing mutable arrays</Primary></IndexTerm>
1170 <IndexTerm><Primary>mutable arrays, freezing</Primary></IndexTerm>
1174 Only unsafe-freeze has a primitive. (Safe freeze is done directly in Haskell
1175 by copying the array and then using <Function>unsafeFreeze</Function>.)
1181 unsafeFreezeArray# :: MutableArray# s elt -> State# s -> (# State# s, Array# s elt #)
1182 unsafeFreezeByteArray# :: MutableByteArray# s -> State# s -> (# State# s, ByteArray# #)
1185 <IndexTerm><Primary><literal>unsafeFreezeArray#</literal></Primary></IndexTerm>
1186 <IndexTerm><Primary><literal>unsafeFreezeByteArray#</literal></Primary></IndexTerm>
1194 <Title>Synchronizing variables (M-vars)</Title>
1197 <IndexTerm><Primary>synchronising variables (M-vars)</Primary></IndexTerm>
1198 <IndexTerm><Primary>M-Vars</Primary></IndexTerm>
1202 Synchronising variables are the primitive type used to implement
1203 Concurrent Haskell's MVars (see the Concurrent Haskell paper for
1204 the operational behaviour of these operations).
1210 type MVar# s elt -- primitive
1212 newMVar# :: State# s -> (# State# s, MVar# s elt #)
1213 takeMVar# :: SynchVar# s elt -> State# s -> (# State# s, elt #)
1214 putMVar# :: SynchVar# s elt -> State# s -> State# s
1217 <IndexTerm><Primary><literal>SynchVar#</literal></Primary></IndexTerm>
1218 <IndexTerm><Primary><literal>newSynchVar#</literal></Primary></IndexTerm>
1219 <IndexTerm><Primary><literal>takeMVar</literal></Primary></IndexTerm>
1220 <IndexTerm><Primary><literal>putMVar</literal></Primary></IndexTerm>
1227 <Sect1 id="glasgow-ST-monad">
1228 <Title>Primitive state-transformer monad
1232 <IndexTerm><Primary>state transformers (Glasgow extensions)</Primary></IndexTerm>
1233 <IndexTerm><Primary>ST monad (Glasgow extension)</Primary></IndexTerm>
1237 This monad underlies our implementation of arrays, mutable and
1238 immutable, and our implementation of I/O, including “C calls”.
1242 The <Literal>ST</Literal> library, which provides access to the
1243 <Function>ST</Function> monad, is described in <xref
1249 <Sect1 id="glasgow-prim-arrays">
1250 <Title>Primitive arrays, mutable and otherwise
1254 <IndexTerm><Primary>primitive arrays (Glasgow extension)</Primary></IndexTerm>
1255 <IndexTerm><Primary>arrays, primitive (Glasgow extension)</Primary></IndexTerm>
1259 GHC knows about quite a few flavours of Large Swathes of Bytes.
1263 First, GHC distinguishes between primitive arrays of (boxed) Haskell
1264 objects (type <Literal>Array# obj</Literal>) and primitive arrays of bytes (type
1265 <Literal>ByteArray#</Literal>).
1269 Second, it distinguishes between…
1273 <Term>Immutable:</Term>
1276 Arrays that do not change (as with “standard” Haskell arrays); you
1277 can only read from them. Obviously, they do not need the care and
1278 attention of the state-transformer monad.
1283 <Term>Mutable:</Term>
1286 Arrays that may be changed or “mutated.” All the operations on them
1287 live within the state-transformer monad and the updates happen
1288 <Emphasis>in-place</Emphasis>.
1293 <Term>“Static” (in C land):</Term>
1296 A C routine may pass an <Literal>Addr#</Literal> pointer back into Haskell land. There
1297 are then primitive operations with which you may merrily grab values
1298 over in C land, by indexing off the “static” pointer.
1303 <Term>“Stable” pointers:</Term>
1306 If, for some reason, you wish to hand a Haskell pointer (i.e.,
1307 <Emphasis>not</Emphasis> an unboxed value) to a C routine, you first make the
1308 pointer “stable,” so that the garbage collector won't forget that it
1309 exists. That is, GHC provides a safe way to pass Haskell pointers to
1314 Please see <XRef LinkEnd="glasgow-stablePtrs"> for more details.
1319 <Term>“Foreign objects”:</Term>
1322 A “foreign object” is a safe way to pass an external object (a
1323 C-allocated pointer, say) to Haskell and have Haskell do the Right
1324 Thing when it no longer references the object. So, for example, C
1325 could pass a large bitmap over to Haskell and say “please free this
1326 memory when you're done with it.”
1330 Please see <XRef LinkEnd="glasgow-foreignObjs"> for more details.
1338 The libraries documentatation gives more details on all these
1339 “primitive array” types and the operations on them.
1344 <Sect1 id="glasgow-ccalls">
1345 <Title>Calling C directly from Haskell
1349 <IndexTerm><Primary>C calls (Glasgow extension)</Primary></IndexTerm>
1350 <IndexTerm><Primary>_ccall_ (Glasgow extension)</Primary></IndexTerm>
1351 <IndexTerm><Primary>_casm_ (Glasgow extension)</Primary></IndexTerm>
1355 GOOD ADVICE: Because this stuff is not Entirely Stable as far as names
1356 and things go, you would be well-advised to keep your C-callery
1357 corraled in a few modules, rather than sprinkled all over your code.
1358 It will then be quite easy to update later on.
1361 <Sect2 id="ccall-intro">
1362 <Title><Function>_ccall_</Function> and <Function>_casm_</Function>: an introduction
1366 The simplest way to use a simple C function
1372 double fooC( FILE *in, char c, int i, double d, unsigned int u )
1378 is to provide a Haskell wrapper:
1384 fooH :: Char -> Int -> Double -> Word -> IO Double
1385 fooH c i d w = _ccall_ fooC (“stdin”::Addr) c i d w
1391 The function <Function>fooH</Function> unbox all of its arguments, call the C
1392 function <Function>fooC</Function> and box the corresponding arguments.
1396 One of the annoyances about <Function>_ccall_</Function>s is when the C types don't quite
1397 match the Haskell compiler's ideas. For this, the <Function>_casm_</Function> variant
1398 may be just the ticket (NB: <Emphasis>no chance</Emphasis> of such code going
1399 through a native-code generator):
1409 = _casm_ “%r = getenv((char *) %0);” name >>= \ litstring ->
1411 if (litstring == nullAddr) then
1412 Left ("Fail:oldGetEnv:"++name)
1414 Right (unpackCString litstring)
1421 The first literal-literal argument to a <Function>_casm_</Function> is like a <Function>printf</Function>
1422 format: <Literal>%r</Literal> is replaced with the “result,” <Literal>%0</Literal>–<Literal>%n-1</Literal> are
1423 replaced with the 1st–nth arguments. As you can see above, it is an
1424 easy way to do simple C casting. Everything said about <Function>_ccall_</Function> goes
1425 for <Function>_casm_</Function> as well.
1429 The use of <Function>_casm_</Function> in your code does pose a problem to the compiler
1430 when it comes to generating an interface file for a freshly compiled
1431 module. Included in an interface file is the unfolding (if any) of a
1432 declaration. However, if a declaration's unfolding happens to contain
1433 a <Function>_casm_</Function>, its unfolding will <Emphasis>not</Emphasis> be emitted into the interface
1434 file even if it qualifies by all the other criteria. The reason why
1435 the compiler prevents this from happening is that unfolding <Function>_casm_</Function>s
1436 into an interface file unduly constrains how code that import your
1437 module have to be compiled. If an imported declaration is unfolded and
1438 it contains a <Function>_casm_</Function>, you now have to be using a compiler backend
1439 capable of dealing with it (i.e., the C compiler backend). If you are
1440 using the C compiler backend, the unfolded <Function>_casm_</Function> may still cause you
1441 problems since the C code snippet it contains may mention CPP symbols
1442 that were in scope when compiling the original module are not when
1443 compiling the importing module.
1447 If you're willing to put up with the drawbacks of doing cross-module
1448 inlining of C code (GHC - A Better C Compiler :-), the option
1449 <Option>-funfold-casms-in-hi-file</Option> will turn off the default behaviour.
1450 <IndexTerm><Primary>-funfold-casms-in-hi-file option</Primary></IndexTerm>
1455 <Sect2 id="glasgow-literal-literals">
1456 <Title>Literal-literals</Title>
1459 <IndexTerm><Primary>Literal-literals</Primary></IndexTerm>
1460 The literal-literal argument to <Function>_casm_</Function> can be made use of separately
1461 from the <Function>_casm_</Function> construct itself. Indeed, we've already used it:
1467 fooH :: Char -> Int -> Double -> Word -> IO Double
1468 fooH c i d w = _ccall_ fooC (“stdin”::Addr) c i d w
1474 The first argument that's passed to <Function>fooC</Function> is given as a literal-literal,
1475 that is, a literal chunk of C code that will be inserted into the generated
1476 <Filename>.hc</Filename> code at the right place.
1480 A literal-literal is restricted to having a type that's an instance of
1481 the <Literal>CCallable</Literal> class, see <XRef LinkEnd="ccall-gotchas">
1482 for more information.
1486 Notice that literal-literals are by their very nature unfriendly to
1487 native code generators, so exercise judgement about whether or not to
1488 make use of them in your code.
1493 <Sect2 id="glasgow-foreign-headers">
1494 <Title>Using function headers
1498 <IndexTerm><Primary>C calls, function headers</Primary></IndexTerm>
1502 When generating C (using the <Option>-fvia-C</Option> directive), one can assist the
1503 C compiler in detecting type errors by using the <Command>-#include</Command> directive
1504 to provide <Filename>.h</Filename> files containing function headers.
1514 typedef unsigned long *StgForeignObj;
1515 typedef long StgInt;
1517 void initialiseEFS (StgInt size);
1518 StgInt terminateEFS (void);
1519 StgForeignObj emptyEFS(void);
1520 StgForeignObj updateEFS (StgForeignObj a, StgInt i, StgInt x);
1521 StgInt lookupEFS (StgForeignObj a, StgInt i);
1527 You can find appropriate definitions for <Literal>StgInt</Literal>, <Literal>StgForeignObj</Literal>,
1528 etc using <Command>gcc</Command> on your architecture by consulting
1529 <Filename>ghc/includes/StgTypes.h</Filename>. The following table summarises the
1530 relationship between Haskell types and C types.
1537 <ColSpec Align="Left" Colsep="0">
1538 <ColSpec Align="Left" Colsep="0">
1541 <Entry><Emphasis>C type name</Emphasis> </Entry>
1542 <Entry> <Emphasis>Haskell Type</Emphasis> </Entry>
1547 <Literal>StgChar</Literal> </Entry>
1548 <Entry> <Literal>Char#</Literal> </Entry>
1552 <Literal>StgInt</Literal> </Entry>
1553 <Entry> <Literal>Int#</Literal> </Entry>
1557 <Literal>StgWord</Literal> </Entry>
1558 <Entry> <Literal>Word#</Literal> </Entry>
1562 <Literal>StgAddr</Literal> </Entry>
1563 <Entry> <Literal>Addr#</Literal> </Entry>
1567 <Literal>StgFloat</Literal> </Entry>
1568 <Entry> <Literal>Float#</Literal> </Entry>
1572 <Literal>StgDouble</Literal> </Entry>
1573 <Entry> <Literal>Double#</Literal> </Entry>
1577 <Literal>StgArray</Literal> </Entry>
1578 <Entry> <Literal>Array#</Literal> </Entry>
1582 <Literal>StgByteArray</Literal> </Entry>
1583 <Entry> <Literal>ByteArray#</Literal> </Entry>
1587 <Literal>StgArray</Literal> </Entry>
1588 <Entry> <Literal>MutableArray#</Literal> </Entry>
1592 <Literal>StgByteArray</Literal> </Entry>
1593 <Entry> <Literal>MutableByteArray#</Literal> </Entry>
1597 <Literal>StgStablePtr</Literal> </Entry>
1598 <Entry> <Literal>StablePtr#</Literal> </Entry>
1602 <Literal>StgForeignObj</Literal> </Entry>
1603 <Entry> <Literal>ForeignObj#</Literal></Entry>
1612 Note that this approach is only <Emphasis>essential</Emphasis> for returning
1613 <Literal>float</Literal>s (or if <Literal>sizeof(int) != sizeof(int *)</Literal> on your
1614 architecture) but is a Good Thing for anyone who cares about writing
1615 solid code. You're crazy not to do it.
1620 <Sect2 id="glasgow-stablePtrs">
1621 <Title>Subverting automatic unboxing with “stable pointers”
1625 <IndexTerm><Primary>stable pointers (Glasgow extension)</Primary></IndexTerm>
1629 The arguments of a <Function>_ccall_</Function> automatically unboxed before the
1630 call. There are two reasons why this is usually the Right Thing to
1640 C is a strict language: it would be excessively tedious to pass
1641 unevaluated arguments and require the C programmer to force their
1642 evaluation before using them.
1649 Boxed values are stored on the Haskell heap and may be moved
1650 within the heap if a garbage collection occurs—that is, pointers
1651 to boxed objects are not <Emphasis>stable</Emphasis>.
1660 It is possible to subvert the unboxing process by creating a “stable
1661 pointer” to a value and passing the stable pointer instead. For
1662 example, to pass/return an integer lazily to C functions <Function>storeC</Function> and
1663 <Function>fetchC</Function> might write:
1669 storeH :: Int -> IO ()
1670 storeH x = makeStablePtr x >>= \ stable_x ->
1671 _ccall_ storeC stable_x
1674 fetchH x = _ccall_ fetchC >>= \ stable_x ->
1675 deRefStablePtr stable_x >>= \ x ->
1676 freeStablePtr stable_x >>
1683 The garbage collector will refrain from throwing a stable pointer away
1684 until you explicitly call one of the following from C or Haskell.
1690 void freeStablePointer( StgStablePtr stablePtrToToss )
1691 freeStablePtr :: StablePtr a -> IO ()
1697 As with the use of <Function>free</Function> in C programs, GREAT CARE SHOULD BE
1698 EXERCISED to ensure these functions are called at the right time: too
1699 early and you get dangling references (and, if you're lucky, an error
1700 message from the runtime system); too late and you get space leaks.
1704 And to force evaluation of the argument within <Function>fooC</Function>, one would
1705 call one of the following C functions (according to type of argument).
1711 void performIO ( StgStablePtr stableIndex /* StablePtr s (IO ()) */ );
1712 StgInt enterInt ( StgStablePtr stableIndex /* StablePtr s Int */ );
1713 StgFloat enterFloat ( StgStablePtr stableIndex /* StablePtr s Float */ );
1719 <IndexTerm><Primary>performIO</Primary></IndexTerm>
1720 <IndexTerm><Primary>enterInt</Primary></IndexTerm>
1721 <IndexTerm><Primary>enterFloat</Primary></IndexTerm>
1725 Nota Bene: <Function>_ccall_GC_</Function><IndexTerm><Primary>_ccall_GC_</Primary></IndexTerm> must be used if any of
1726 these functions are used.
1731 <Sect2 id="glasgow-foreignObjs">
1732 <Title>Foreign objects: pointing outside the Haskell heap
1736 <IndexTerm><Primary>foreign objects (Glasgow extension)</Primary></IndexTerm>
1740 There are two types that GHC programs can use to reference
1741 (heap-allocated) objects outside the Haskell world: <Literal>Addr</Literal> and
1742 <Literal>ForeignObj</Literal>.
1746 If you use <Literal>Addr</Literal>, it is up to you to the programmer to arrange
1747 allocation and deallocation of the objects.
1751 If you use <Literal>ForeignObj</Literal>, GHC's garbage collector will call upon the
1752 user-supplied <Emphasis>finaliser</Emphasis> function to free the object when the
1753 Haskell world no longer can access the object. (An object is
1754 associated with a finaliser function when the abstract
1755 Haskell type <Literal>ForeignObj</Literal> is created). The finaliser function is
1756 expressed in C, and is passed as argument the object:
1762 void foreignFinaliser ( StgForeignObj fo )
1768 when the Haskell world can no longer access the object. Since
1769 <Literal>ForeignObj</Literal>s only get released when a garbage collection occurs, we
1770 provide ways of triggering a garbage collection from within C and from
1777 void GarbageCollect()
1784 More information on the programmers' interface to <Literal>ForeignObj</Literal> can be
1785 found in the library documentation.
1790 <Sect2 id="glasgow-avoiding-monads">
1791 <Title>Avoiding monads
1795 <IndexTerm><Primary>C calls to `pure C'</Primary></IndexTerm>
1796 <IndexTerm><Primary>unsafePerformIO</Primary></IndexTerm>
1800 The <Function>_ccall_</Function> construct is part of the <Literal>IO</Literal> monad because 9 out of 10
1801 uses will be to call imperative functions with side effects such as
1802 <Function>printf</Function>. Use of the monad ensures that these operations happen in a
1803 predictable order in spite of laziness and compiler optimisations.
1807 To avoid having to be in the monad to call a C function, it is
1808 possible to use <Function>unsafePerformIO</Function>, which is available from the
1809 <Literal>IOExts</Literal> module. There are three situations where one might like to
1810 call a C function from outside the IO world:
1819 Calling a function with no side-effects:
1822 atan2d :: Double -> Double -> Double
1823 atan2d y x = unsafePerformIO (_ccall_ atan2d y x)
1825 sincosd :: Double -> (Double, Double)
1826 sincosd x = unsafePerformIO $ do
1827 da <- newDoubleArray (0, 1)
1828 _casm_ “sincosd( %0, &((double *)%1[0]), &((double *)%1[1]) );” x da
1829 s <- readDoubleArray da 0
1830 c <- readDoubleArray da 1
1840 Calling a set of functions which have side-effects but which can
1841 be used in a purely functional manner.
1843 For example, an imperative implementation of a purely functional
1844 lookup-table might be accessed using the following functions.
1849 update :: EFS x -> Int -> x -> EFS x
1850 lookup :: EFS a -> Int -> a
1852 empty = unsafePerformIO (_ccall_ emptyEFS)
1854 update a i x = unsafePerformIO $
1855 makeStablePtr x >>= \ stable_x ->
1856 _ccall_ updateEFS a i stable_x
1858 lookup a i = unsafePerformIO $
1859 _ccall_ lookupEFS a i >>= \ stable_x ->
1860 deRefStablePtr stable_x
1864 You will almost always want to use <Literal>ForeignObj</Literal>s with this.
1871 Calling a side-effecting function even though the results will
1872 be unpredictable. For example the <Function>trace</Function> function is defined by:
1876 trace :: String -> a -> a
1879 ((_ccall_ PreTraceHook sTDERR{-msg-}):: IO ()) >>
1880 fputs sTDERR string >>
1881 ((_ccall_ PostTraceHook sTDERR{-msg-}):: IO ()) >>
1884 sTDERR = (“stderr” :: Addr)
1888 (This kind of use is not highly recommended—it is only really
1889 useful in debugging code.)
1899 <Sect2 id="ccall-gotchas">
1900 <Title>C-calling “gotchas” checklist
1904 <IndexTerm><Primary>C call dangers</Primary></IndexTerm>
1905 <IndexTerm><Primary>CCallable</Primary></IndexTerm>
1906 <IndexTerm><Primary>CReturnable</Primary></IndexTerm>
1910 And some advice, too.
1919 For modules that use <Function>_ccall_</Function>s, etc., compile with
1920 <Option>-fvia-C</Option>.<IndexTerm><Primary>-fvia-C option</Primary></IndexTerm> You don't have to, but you should.
1922 Also, use the <Option>-#include "prototypes.h"</Option> flag (hack) to inform the C
1923 compiler of the fully-prototyped types of all the C functions you
1924 call. (<XRef LinkEnd="glasgow-foreign-headers"> says more about this…)
1926 This scheme is the <Emphasis>only</Emphasis> way that you will get <Emphasis>any</Emphasis>
1927 typechecking of your <Function>_ccall_</Function>s. (It shouldn't be that way, but…).
1928 GHC will pass the flag <Option>-Wimplicit</Option> to <Command>gcc</Command> so that you'll get warnings
1929 if any <Function>_ccall_</Function>ed functions have no prototypes.
1936 Try to avoid <Function>_ccall_</Function>s to C functions that take <Literal>float</Literal>
1937 arguments or return <Literal>float</Literal> results. Reason: if you do, you will
1938 become entangled in (ANSI?) C's rules for when arguments/results are
1939 promoted to <Literal>doubles</Literal>. It's a nightmare and just not worth it.
1940 Use <Literal>doubles</Literal> if possible.
1942 If you do use <Literal>floats</Literal>, check and re-check that the right thing is
1943 happening. Perhaps compile with <Option>-keep-hc-file-too</Option> and look at
1944 the intermediate C (<Function>.hc</Function>).
1951 The compiler uses two non-standard type-classes when
1952 type-checking the arguments and results of <Function>_ccall_</Function>: the arguments
1953 (respectively result) of <Function>_ccall_</Function> must be instances of the class
1954 <Literal>CCallable</Literal> (respectively <Literal>CReturnable</Literal>). Both classes may be
1955 imported from the module <Literal>CCall</Literal>, but this should only be
1956 necessary if you want to define a new instance. (Neither class
1957 defines any methods—their only function is to keep the
1958 type-checker happy.)
1960 The type checker must be able to figure out just which of the
1961 C-callable/returnable types is being used. If it can't, you have to
1962 add type signatures. For example,
1970 is not good enough, because the compiler can't work out what type <VarName>x</VarName>
1971 is, nor what type the <Function>_ccall_</Function> returns. You have to write, say:
1975 f :: Int -> IO Double
1980 This table summarises the standard instances of these classes.
1984 <ColSpec Align="Left" Colsep="0">
1985 <ColSpec Align="Left" Colsep="0">
1986 <ColSpec Align="Left" Colsep="0">
1987 <ColSpec Align="Left" Colsep="0">
1990 <Entry><Emphasis>Type</Emphasis> </Entry>
1991 <Entry><Emphasis>CCallable</Emphasis></Entry>
1992 <Entry><Emphasis>CReturnable</Emphasis> </Entry>
1993 <Entry><Emphasis>Which is probably…</Emphasis> </Entry>
1997 <Literal>Char</Literal> </Entry>
1998 <Entry> Yes </Entry>
1999 <Entry> Yes </Entry>
2000 <Entry> <Literal>unsigned char</Literal> </Entry>
2004 <Literal>Int</Literal> </Entry>
2005 <Entry> Yes </Entry>
2006 <Entry> Yes </Entry>
2007 <Entry> <Literal>long int</Literal> </Entry>
2011 <Literal>Word</Literal> </Entry>
2012 <Entry> Yes </Entry>
2013 <Entry> Yes </Entry>
2014 <Entry> <Literal>unsigned long int</Literal> </Entry>
2018 <Literal>Addr</Literal> </Entry>
2019 <Entry> Yes </Entry>
2020 <Entry> Yes </Entry>
2021 <Entry> <Literal>void *</Literal> </Entry>
2025 <Literal>Float</Literal> </Entry>
2026 <Entry> Yes </Entry>
2027 <Entry> Yes </Entry>
2028 <Entry> <Literal>float</Literal> </Entry>
2032 <Literal>Double</Literal> </Entry>
2033 <Entry> Yes </Entry>
2034 <Entry> Yes </Entry>
2035 <Entry> <Literal>double</Literal> </Entry>
2039 <Literal>()</Literal> </Entry>
2041 <Entry> Yes </Entry>
2042 <Entry> <Literal>void</Literal> </Entry>
2046 <Literal>[Char]</Literal> </Entry>
2047 <Entry> Yes </Entry>
2049 <Entry> <Literal>char *</Literal> (null-terminated) </Entry>
2053 <Literal>Array</Literal> </Entry>
2054 <Entry> Yes </Entry>
2056 <Entry> <Literal>unsigned long *</Literal> </Entry>
2060 <Literal>ByteArray</Literal> </Entry>
2061 <Entry> Yes </Entry>
2063 <Entry> <Literal>unsigned long *</Literal> </Entry>
2067 <Literal>MutableArray</Literal> </Entry>
2068 <Entry> Yes </Entry>
2070 <Entry> <Literal>unsigned long *</Literal> </Entry>
2074 <Literal>MutableByteArray</Literal> </Entry>
2075 <Entry> Yes </Entry>
2077 <Entry> <Literal>unsigned long *</Literal> </Entry>
2081 <Literal>State</Literal> </Entry>
2082 <Entry> Yes </Entry>
2083 <Entry> Yes </Entry>
2084 <Entry> nothing!</Entry>
2088 <Literal>StablePtr</Literal> </Entry>
2089 <Entry> Yes </Entry>
2090 <Entry> Yes </Entry>
2091 <Entry> <Literal>unsigned long *</Literal> </Entry>
2095 <Literal>ForeignObjs</Literal> </Entry>
2096 <Entry> Yes </Entry>
2097 <Entry> Yes </Entry>
2098 <Entry> see later </Entry>
2106 Actually, the <Literal>Word</Literal> type is defined as being the same size as a
2107 pointer on the target architecture, which is <Emphasis>probably</Emphasis>
2108 <Literal>unsigned long int</Literal>.
2110 The brave and careful programmer can add their own instances of these
2111 classes for the following types:
2118 A <Emphasis>boxed-primitive</Emphasis> type may be made an instance of both
2119 <Literal>CCallable</Literal> and <Literal>CReturnable</Literal>.
2121 A boxed primitive type is any data type with a
2122 single unary constructor with a single primitive argument. For
2123 example, the following are all boxed primitive types:
2129 data XDisplay = XDisplay Addr#
2130 data EFS a = EFS# ForeignObj#
2136 instance CCallable (EFS a)
2137 instance CReturnable (EFS a)
2146 Any datatype with a single nullary constructor may be made an
2147 instance of <Literal>CReturnable</Literal>. For example:
2151 data MyVoid = MyVoid
2152 instance CReturnable MyVoid
2161 As at version 2.09, <Literal>String</Literal> (i.e., <Literal>[Char]</Literal>) is still
2162 not a <Literal>CReturnable</Literal> type.
2164 Also, the now-builtin type <Literal>PackedString</Literal> is neither
2165 <Literal>CCallable</Literal> nor <Literal>CReturnable</Literal>. (But there are functions in
2166 the PackedString interface to let you get at the necessary bits…)
2178 The code-generator will complain if you attempt to use <Literal>%r</Literal> in
2179 a <Literal>_casm_</Literal> whose result type is <Literal>IO ()</Literal>; or if you don't use <Literal>%r</Literal>
2180 <Emphasis>precisely</Emphasis> once for any other result type. These messages are
2181 supposed to be helpful and catch bugs—please tell us if they wreck
2189 If you call out to C code which may trigger the Haskell garbage
2190 collector or create new threads (examples of this later…), then you
2191 must use the <Function>_ccall_GC_</Function><IndexTerm><Primary>_ccall_GC_ primitive</Primary></IndexTerm> or
2192 <Function>_casm_GC_</Function><IndexTerm><Primary>_casm_GC_ primitive</Primary></IndexTerm> variant of C-calls. (This
2193 does not work with the native code generator—use <Option>-fvia-C</Option>.) This
2194 stuff is hairy with a capital H!
2206 <Sect1 id="multi-param-type-classes">
2207 <Title>Multi-parameter type classes
2211 This section documents GHC's implementation of multi-paramter type
2212 classes. There's lots of background in the paper <ULink
2213 URL="http://research.microsoft.com/~simonpj/multi.ps.gz" >Type
2214 classes: exploring the design space</ULink > (Simon Peyton Jones, Mark
2215 Jones, Erik Meijer).
2219 I'd like to thank people who reported shorcomings in the GHC 3.02
2220 implementation. Our default decisions were all conservative ones, and
2221 the experience of these heroic pioneers has given useful concrete
2222 examples to support several generalisations. (These appear below as
2223 design choices not implemented in 3.02.)
2227 I've discussed these notes with Mark Jones, and I believe that Hugs
2228 will migrate towards the same design choices as I outline here.
2229 Thanks to him, and to many others who have offered very useful
2234 <Title>Types</Title>
2237 There are the following restrictions on the form of a qualified
2244 forall tv1..tvn (c1, ...,cn) => type
2250 (Here, I write the "foralls" explicitly, although the Haskell source
2251 language omits them; in Haskell 1.4, all the free type variables of an
2252 explicit source-language type signature are universally quantified,
2253 except for the class type variables in a class declaration. However,
2254 in GHC, you can give the foralls if you want. See <XRef LinkEnd="universal-quantification">).
2263 <Emphasis>Each universally quantified type variable
2264 <Literal>tvi</Literal> must be mentioned (i.e. appear free) in <Literal>type</Literal></Emphasis>.
2266 The reason for this is that a value with a type that does not obey
2267 this restriction could not be used without introducing
2268 ambiguity. Here, for example, is an illegal type:
2272 forall a. Eq a => Int
2276 When a value with this type was used, the constraint <Literal>Eq tv</Literal>
2277 would be introduced where <Literal>tv</Literal> is a fresh type variable, and
2278 (in the dictionary-translation implementation) the value would be
2279 applied to a dictionary for <Literal>Eq tv</Literal>. The difficulty is that we
2280 can never know which instance of <Literal>Eq</Literal> to use because we never
2281 get any more information about <Literal>tv</Literal>.
2288 <Emphasis>Every constraint <Literal>ci</Literal> must mention at least one of the
2289 universally quantified type variables <Literal>tvi</Literal></Emphasis>.
2291 For example, this type is OK because <Literal>C a b</Literal> mentions the
2292 universally quantified type variable <Literal>b</Literal>:
2296 forall a. C a b => burble
2300 The next type is illegal because the constraint <Literal>Eq b</Literal> does not
2301 mention <Literal>a</Literal>:
2305 forall a. Eq b => burble
2309 The reason for this restriction is milder than the other one. The
2310 excluded types are never useful or necessary (because the offending
2311 context doesn't need to be witnessed at this point; it can be floated
2312 out). Furthermore, floating them out increases sharing. Lastly,
2313 excluding them is a conservative choice; it leaves a patch of
2314 territory free in case we need it later.
2324 These restrictions apply to all types, whether declared in a type signature
2329 Unlike Haskell 1.4, constraints in types do <Emphasis>not</Emphasis> have to be of
2330 the form <Emphasis>(class type-variables)</Emphasis>. Thus, these type signatures
2337 f :: Eq (m a) => [m a] -> [m a]
2344 This choice recovers principal types, a property that Haskell 1.4 does not have.
2350 <Title>Class declarations</Title>
2358 <Emphasis>Multi-parameter type classes are permitted</Emphasis>. For example:
2362 class Collection c a where
2363 union :: c a -> c a -> c a
2374 <Emphasis>The class hierarchy must be acyclic</Emphasis>. However, the definition
2375 of "acyclic" involves only the superclass relationships. For example,
2381 op :: D b => a -> b -> b
2384 class C a => D a where { ... }
2388 Here, <Literal>C</Literal> is a superclass of <Literal>D</Literal>, but it's OK for a
2389 class operation <Literal>op</Literal> of <Literal>C</Literal> to mention <Literal>D</Literal>. (It
2390 would not be OK for <Literal>D</Literal> to be a superclass of <Literal>C</Literal>.)
2397 <Emphasis>There are no restrictions on the context in a class declaration
2398 (which introduces superclasses), except that the class hierarchy must
2399 be acyclic</Emphasis>. So these class declarations are OK:
2403 class Functor (m k) => FiniteMap m k where
2406 class (Monad m, Monad (t m)) => Transform t m where
2407 lift :: m a -> (t m) a
2416 <Emphasis>In the signature of a class operation, every constraint
2417 must mention at least one type variable that is not a class type
2418 variable</Emphasis>.
2424 class Collection c a where
2425 mapC :: Collection c b => (a->b) -> c a -> c b
2429 is OK because the constraint <Literal>(Collection a b)</Literal> mentions
2430 <Literal>b</Literal>, even though it also mentions the class variable
2431 <Literal>a</Literal>. On the other hand:
2436 op :: Eq a => (a,b) -> (a,b)
2440 is not OK because the constraint <Literal>(Eq a)</Literal> mentions on the class
2441 type variable <Literal>a</Literal>, but not <Literal>b</Literal>. However, any such
2442 example is easily fixed by moving the offending context up to the
2447 class Eq a => C a where
2452 A yet more relaxed rule would allow the context of a class-op signature
2453 to mention only class type variables. However, that conflicts with
2454 Rule 1(b) for types above.
2461 <Emphasis>The type of each class operation must mention <Emphasis>all</Emphasis> of
2462 the class type variables</Emphasis>. For example:
2466 class Coll s a where
2468 insert :: s -> a -> s
2472 is not OK, because the type of <Literal>empty</Literal> doesn't mention
2473 <Literal>a</Literal>. This rule is a consequence of Rule 1(a), above, for
2474 types, and has the same motivation.
2476 Sometimes, offending class declarations exhibit misunderstandings. For
2477 example, <Literal>Coll</Literal> might be rewritten
2481 class Coll s a where
2483 insert :: s a -> a -> s a
2487 which makes the connection between the type of a collection of
2488 <Literal>a</Literal>'s (namely <Literal>(s a)</Literal>) and the element type <Literal>a</Literal>.
2489 Occasionally this really doesn't work, in which case you can split the
2497 class CollE s => Coll s a where
2498 insert :: s -> a -> s
2512 <Title>Instance declarations</Title>
2520 <Emphasis>Instance declarations may not overlap</Emphasis>. The two instance
2525 instance context1 => C type1 where ...
2526 instance context2 => C type2 where ...
2530 "overlap" if <Literal>type1</Literal> and <Literal>type2</Literal> unify
2532 However, if you give the command line option
2533 <Option>-fallow-overlapping-instances</Option><IndexTerm><Primary>-fallow-overlapping-instances
2534 option</Primary></IndexTerm> then two overlapping instance declarations are permitted
2542 EITHER <Literal>type1</Literal> and <Literal>type2</Literal> do not unify
2548 OR <Literal>type2</Literal> is a substitution instance of <Literal>type1</Literal>
2549 (but not identical to <Literal>type1</Literal>)
2562 Notice that these rules
2569 make it clear which instance decl to use
2570 (pick the most specific one that matches)
2577 do not mention the contexts <Literal>context1</Literal>, <Literal>context2</Literal>
2578 Reason: you can pick which instance decl
2579 "matches" based on the type.
2586 Regrettably, GHC doesn't guarantee to detect overlapping instance
2587 declarations if they appear in different modules. GHC can "see" the
2588 instance declarations in the transitive closure of all the modules
2589 imported by the one being compiled, so it can "see" all instance decls
2590 when it is compiling <Literal>Main</Literal>. However, it currently chooses not
2591 to look at ones that can't possibly be of use in the module currently
2592 being compiled, in the interests of efficiency. (Perhaps we should
2593 change that decision, at least for <Literal>Main</Literal>.)
2600 <Emphasis>There are no restrictions on the type in an instance
2601 <Emphasis>head</Emphasis>, except that at least one must not be a type variable</Emphasis>.
2602 The instance "head" is the bit after the "=>" in an instance decl. For
2603 example, these are OK:
2607 instance C Int a where ...
2609 instance D (Int, Int) where ...
2611 instance E [[a]] where ...
2615 Note that instance heads <Emphasis>may</Emphasis> contain repeated type variables.
2616 For example, this is OK:
2620 instance Stateful (ST s) (MutVar s) where ...
2624 The "at least one not a type variable" restriction is to ensure that
2625 context reduction terminates: each reduction step removes one type
2626 constructor. For example, the following would make the type checker
2627 loop if it wasn't excluded:
2631 instance C a => C a where ...
2635 There are two situations in which the rule is a bit of a pain. First,
2636 if one allows overlapping instance declarations then it's quite
2637 convenient to have a "default instance" declaration that applies if
2638 something more specific does not:
2647 Second, sometimes you might want to use the following to get the
2648 effect of a "class synonym":
2652 class (C1 a, C2 a, C3 a) => C a where { }
2654 instance (C1 a, C2 a, C3 a) => C a where { }
2658 This allows you to write shorter signatures:
2670 f :: (C1 a, C2 a, C3 a) => ...
2674 I'm on the lookout for a simple rule that preserves decidability while
2675 allowing these idioms. The experimental flag
2676 <Option>-fallow-undecidable-instances</Option><IndexTerm><Primary>-fallow-undecidable-instances
2677 option</Primary></IndexTerm> lifts this restriction, allowing all the types in an
2678 instance head to be type variables.
2685 <Emphasis>Unlike Haskell 1.4, instance heads may use type
2686 synonyms</Emphasis>. As always, using a type synonym is just shorthand for
2687 writing the RHS of the type synonym definition. For example:
2691 type Point = (Int,Int)
2692 instance C Point where ...
2693 instance C [Point] where ...
2697 is legal. However, if you added
2701 instance C (Int,Int) where ...
2705 as well, then the compiler will complain about the overlapping
2706 (actually, identical) instance declarations. As always, type synonyms
2707 must be fully applied. You cannot, for example, write:
2712 instance Monad P where ...
2716 This design decision is independent of all the others, and easily
2717 reversed, but it makes sense to me.
2724 <Emphasis>The types in an instance-declaration <Emphasis>context</Emphasis> must all
2725 be type variables</Emphasis>. Thus
2729 instance C a b => Eq (a,b) where ...
2737 instance C Int b => Foo b where ...
2741 is not OK. Again, the intent here is to make sure that context
2742 reduction terminates.
2744 Voluminous correspondence on the Haskell mailing list has convinced me
2745 that it's worth experimenting with a more liberal rule. If you use
2746 the flag <Option>-fallow-undecidable-instances</Option> can use arbitrary
2747 types in an instance context. Termination is ensured by having a
2748 fixed-depth recursion stack. If you exceed the stack depth you get a
2749 sort of backtrace, and the opportunity to increase the stack depth
2750 with <Option>-fcontext-stack</Option><Emphasis>N</Emphasis>.
2763 <Sect1 id="universal-quantification">
2764 <Title>Explicit universal quantification
2768 GHC now allows you to write explicitly quantified types. GHC's
2769 syntax for this now agrees with Hugs's, namely:
2775 forall a b. (Ord a, Eq b) => a -> b -> a
2781 The context is, of course, optional. You can't use <Literal>forall</Literal> as
2782 a type variable any more!
2786 Haskell type signatures are implicitly quantified. The <Literal>forall</Literal>
2787 allows us to say exactly what this means. For example:
2805 g :: forall b. (b -> b)
2811 The two are treated identically.
2815 <Title>Universally-quantified data type fields
2819 In a <Literal>data</Literal> or <Literal>newtype</Literal> declaration one can quantify
2820 the types of the constructor arguments. Here are several examples:
2826 data T a = T1 (forall b. b -> b -> b) a
2828 data MonadT m = MkMonad { return :: forall a. a -> m a,
2829 bind :: forall a b. m a -> (a -> m b) -> m b
2832 newtype Swizzle = MkSwizzle (Ord a => [a] -> [a])
2838 The constructors now have so-called <Emphasis>rank 2</Emphasis> polymorphic
2839 types, in which there is a for-all in the argument types.:
2845 T1 :: forall a. (forall b. b -> b -> b) -> a -> T a
2846 MkMonad :: forall m. (forall a. a -> m a)
2847 -> (forall a b. m a -> (a -> m b) -> m b)
2849 MkSwizzle :: (Ord a => [a] -> [a]) -> Swizzle
2855 Notice that you don't need to use a <Literal>forall</Literal> if there's an
2856 explicit context. For example in the first argument of the
2857 constructor <Function>MkSwizzle</Function>, an implicit "<Literal>forall a.</Literal>" is
2858 prefixed to the argument type. The implicit <Literal>forall</Literal>
2859 quantifies all type variables that are not already in scope, and are
2860 mentioned in the type quantified over.
2864 As for type signatures, implicit quantification happens for non-overloaded
2865 types too. So if you write this:
2868 data T a = MkT (Either a b) (b -> b)
2871 it's just as if you had written this:
2874 data T a = MkT (forall b. Either a b) (forall b. b -> b)
2877 That is, since the type variable <Literal>b</Literal> isn't in scope, it's
2878 implicitly universally quantified. (Arguably, it would be better
2879 to <Emphasis>require</Emphasis> explicit quantification on constructor arguments
2880 where that is what is wanted. Feedback welcomed.)
2886 <Title>Construction </Title>
2889 You construct values of types <Literal>T1, MonadT, Swizzle</Literal> by applying
2890 the constructor to suitable values, just as usual. For example,
2896 (T1 (\xy->x) 3) :: T Int
2898 (MkSwizzle sort) :: Swizzle
2899 (MkSwizzle reverse) :: Swizzle
2906 MkMonad r b) :: MonadT Maybe
2912 The type of the argument can, as usual, be more general than the type
2913 required, as <Literal>(MkSwizzle reverse)</Literal> shows. (<Function>reverse</Function>
2914 does not need the <Literal>Ord</Literal> constraint.)
2920 <Title>Pattern matching</Title>
2923 When you use pattern matching, the bound variables may now have
2924 polymorphic types. For example:
2930 f :: T a -> a -> (a, Char)
2931 f (T1 f k) x = (f k x, f 'c' 'd')
2933 g :: (Ord a, Ord b) => Swizzle -> [a] -> (a -> b) -> [b]
2934 g (MkSwizzle s) xs f = s (map f (s xs))
2936 h :: MonadT m -> [m a] -> m [a]
2937 h m [] = return m []
2938 h m (x:xs) = bind m x $ \y ->
2939 bind m (h m xs) $ \ys ->
2946 In the function <Function>h</Function> we use the record selectors <Literal>return</Literal>
2947 and <Literal>bind</Literal> to extract the polymorphic bind and return functions
2948 from the <Literal>MonadT</Literal> data structure, rather than using pattern
2953 You cannot pattern-match against an argument that is polymorphic.
2957 newtype TIM s a = TIM (ST s (Maybe a))
2959 runTIM :: (forall s. TIM s a) -> Maybe a
2960 runTIM (TIM m) = runST m
2966 Here the pattern-match fails, because you can't pattern-match against
2967 an argument of type <Literal>(forall s. TIM s a)</Literal>. Instead you
2968 must bind the variable and pattern match in the right hand side:
2971 runTIM :: (forall s. TIM s a) -> Maybe a
2972 runTIM tm = case tm of { TIM m -> runST m }
2975 The <Literal>tm</Literal> on the right hand side is (invisibly) instantiated, like
2976 any polymorphic value at its occurrence site, and now you can pattern-match
2983 <Title>The partial-application restriction</Title>
2986 There is really only one way in which data structures with polymorphic
2987 components might surprise you: you must not partially apply them.
2988 For example, this is illegal:
2994 map MkSwizzle [sort, reverse]
3000 The restriction is this: <Emphasis>every subexpression of the program must
3001 have a type that has no for-alls, except that in a function
3002 application (f e1…en) the partial applications are not subject to
3003 this rule</Emphasis>. The restriction makes type inference feasible.
3007 In the illegal example, the sub-expression <Literal>MkSwizzle</Literal> has the
3008 polymorphic type <Literal>(Ord b => [b] -> [b]) -> Swizzle</Literal> and is not
3009 a sub-expression of an enclosing application. On the other hand, this
3016 map (T1 (\a b -> a)) [1,2,3]
3022 even though it involves a partial application of <Function>T1</Function>, because
3023 the sub-expression <Literal>T1 (\a b -> a)</Literal> has type <Literal>Int -> T
3030 <Title>Type signatures
3034 Once you have data constructors with universally-quantified fields, or
3035 constants such as <Constant>runST</Constant> that have rank-2 types, it isn't long
3036 before you discover that you need more! Consider:
3042 mkTs f x y = [T1 f x, T1 f y]
3048 <Function>mkTs</Function> is a fuction that constructs some values of type
3049 <Literal>T</Literal>, using some pieces passed to it. The trouble is that since
3050 <Literal>f</Literal> is a function argument, Haskell assumes that it is
3051 monomorphic, so we'll get a type error when applying <Function>T1</Function> to
3052 it. This is a rather silly example, but the problem really bites in
3053 practice. Lots of people trip over the fact that you can't make
3054 "wrappers functions" for <Constant>runST</Constant> for exactly the same reason.
3055 In short, it is impossible to build abstractions around functions with
3060 The solution is fairly clear. We provide the ability to give a rank-2
3061 type signature for <Emphasis>ordinary</Emphasis> functions (not only data
3062 constructors), thus:
3068 mkTs :: (forall b. b -> b -> b) -> a -> [T a]
3069 mkTs f x y = [T1 f x, T1 f y]
3075 This type signature tells the compiler to attribute <Literal>f</Literal> with
3076 the polymorphic type <Literal>(forall b. b -> b -> b)</Literal> when type
3077 checking the body of <Function>mkTs</Function>, so now the application of
3078 <Function>T1</Function> is fine.
3082 There are two restrictions:
3091 You can only define a rank 2 type, specified by the following
3096 rank2type ::= [forall tyvars .] [context =>] funty
3097 funty ::= ([forall tyvars .] [context =>] ty) -> funty
3099 ty ::= ...current Haskell monotype syntax...
3103 Informally, the universal quantification must all be right at the beginning,
3104 or at the top level of a function argument.
3111 There is a restriction on the definition of a function whose
3112 type signature is a rank-2 type: the polymorphic arguments must be
3113 matched on the left hand side of the "<Literal>=</Literal>" sign. You can't
3114 define <Function>mkTs</Function> like this:
3118 mkTs :: (forall b. b -> b -> b) -> a -> [T a]
3119 mkTs = \ f x y -> [T1 f x, T1 f y]
3124 The same partial-application rule applies to ordinary functions with
3125 rank-2 types as applied to data constructors.
3138 <Title>Type synonyms and hoisting
3142 GHC also allows you to write a <Literal>forall</Literal> in a type synonym, thus:
3144 type Discard a = forall b. a -> b -> a
3149 However, it is often convenient to use these sort of synonyms at the right hand
3150 end of an arrow, thus:
3152 type Discard a = forall b. a -> b -> a
3154 g :: Int -> Discard Int
3157 Simply expanding the type synonym would give
3159 g :: Int -> (forall b. Int -> b -> Int)
3161 but GHC "hoists" the <Literal>forall</Literal> to give the isomorphic type
3163 g :: forall b. Int -> Int -> b -> Int
3165 In general, the rule is this: <Emphasis>to determine the type specified by any explicit
3166 user-written type (e.g. in a type signature), GHC expands type synonyms and then repeatedly
3167 performs the transformation:</Emphasis>
3169 <Emphasis>type1</Emphasis> -> forall a. <Emphasis>type2</Emphasis>
3171 forall a. <Emphasis>type1</Emphasis> -> <Emphasis>type2</Emphasis>
3173 (In fact, GHC tries to retain as much synonym information as possible for use in
3174 error messages, but that is a usability issue.) This rule applies, of course, whether
3175 or not the <Literal>forall</Literal> comes from a synonym. For example, here is another
3176 valid way to write <Literal>g</Literal>'s type signature:
3178 g :: Int -> Int -> forall b. b -> Int
3185 <Sect1 id="existential-quantification">
3186 <Title>Existentially quantified data constructors
3190 The idea of using existential quantification in data type declarations
3191 was suggested by Laufer (I believe, thought doubtless someone will
3192 correct me), and implemented in Hope+. It's been in Lennart
3193 Augustsson's <Command>hbc</Command> Haskell compiler for several years, and
3194 proved very useful. Here's the idea. Consider the declaration:
3200 data Foo = forall a. MkFoo a (a -> Bool)
3207 The data type <Literal>Foo</Literal> has two constructors with types:
3213 MkFoo :: forall a. a -> (a -> Bool) -> Foo
3220 Notice that the type variable <Literal>a</Literal> in the type of <Function>MkFoo</Function>
3221 does not appear in the data type itself, which is plain <Literal>Foo</Literal>.
3222 For example, the following expression is fine:
3228 [MkFoo 3 even, MkFoo 'c' isUpper] :: [Foo]
3234 Here, <Literal>(MkFoo 3 even)</Literal> packages an integer with a function
3235 <Function>even</Function> that maps an integer to <Literal>Bool</Literal>; and <Function>MkFoo 'c'
3236 isUpper</Function> packages a character with a compatible function. These
3237 two things are each of type <Literal>Foo</Literal> and can be put in a list.
3241 What can we do with a value of type <Literal>Foo</Literal>?. In particular,
3242 what happens when we pattern-match on <Function>MkFoo</Function>?
3248 f (MkFoo val fn) = ???
3254 Since all we know about <Literal>val</Literal> and <Function>fn</Function> is that they
3255 are compatible, the only (useful) thing we can do with them is to
3256 apply <Function>fn</Function> to <Literal>val</Literal> to get a boolean. For example:
3263 f (MkFoo val fn) = fn val
3269 What this allows us to do is to package heterogenous values
3270 together with a bunch of functions that manipulate them, and then treat
3271 that collection of packages in a uniform manner. You can express
3272 quite a bit of object-oriented-like programming this way.
3275 <Sect2 id="existential">
3276 <Title>Why existential?
3280 What has this to do with <Emphasis>existential</Emphasis> quantification?
3281 Simply that <Function>MkFoo</Function> has the (nearly) isomorphic type
3287 MkFoo :: (exists a . (a, a -> Bool)) -> Foo
3293 But Haskell programmers can safely think of the ordinary
3294 <Emphasis>universally</Emphasis> quantified type given above, thereby avoiding
3295 adding a new existential quantification construct.
3301 <Title>Type classes</Title>
3304 An easy extension (implemented in <Command>hbc</Command>) is to allow
3305 arbitrary contexts before the constructor. For example:
3311 data Baz = forall a. Eq a => Baz1 a a
3312 | forall b. Show b => Baz2 b (b -> b)
3318 The two constructors have the types you'd expect:
3324 Baz1 :: forall a. Eq a => a -> a -> Baz
3325 Baz2 :: forall b. Show b => b -> (b -> b) -> Baz
3331 But when pattern matching on <Function>Baz1</Function> the matched values can be compared
3332 for equality, and when pattern matching on <Function>Baz2</Function> the first matched
3333 value can be converted to a string (as well as applying the function to it).
3334 So this program is legal:
3341 f (Baz1 p q) | p == q = "Yes"
3343 f (Baz1 v fn) = show (fn v)
3349 Operationally, in a dictionary-passing implementation, the
3350 constructors <Function>Baz1</Function> and <Function>Baz2</Function> must store the
3351 dictionaries for <Literal>Eq</Literal> and <Literal>Show</Literal> respectively, and
3352 extract it on pattern matching.
3356 Notice the way that the syntax fits smoothly with that used for
3357 universal quantification earlier.
3363 <Title>Restrictions</Title>
3366 There are several restrictions on the ways in which existentially-quantified
3367 constructors can be use.
3376 When pattern matching, each pattern match introduces a new,
3377 distinct, type for each existential type variable. These types cannot
3378 be unified with any other type, nor can they escape from the scope of
3379 the pattern match. For example, these fragments are incorrect:
3387 Here, the type bound by <Function>MkFoo</Function> "escapes", because <Literal>a</Literal>
3388 is the result of <Function>f1</Function>. One way to see why this is wrong is to
3389 ask what type <Function>f1</Function> has:
3393 f1 :: Foo -> a -- Weird!
3397 What is this "<Literal>a</Literal>" in the result type? Clearly we don't mean
3402 f1 :: forall a. Foo -> a -- Wrong!
3406 The original program is just plain wrong. Here's another sort of error
3410 f2 (Baz1 a b) (Baz1 p q) = a==q
3414 It's ok to say <Literal>a==b</Literal> or <Literal>p==q</Literal>, but
3415 <Literal>a==q</Literal> is wrong because it equates the two distinct types arising
3416 from the two <Function>Baz1</Function> constructors.
3424 You can't pattern-match on an existentially quantified
3425 constructor in a <Literal>let</Literal> or <Literal>where</Literal> group of
3426 bindings. So this is illegal:
3430 f3 x = a==b where { Baz1 a b = x }
3434 You can only pattern-match
3435 on an existentially-quantified constructor in a <Literal>case</Literal> expression or
3436 in the patterns of a function definition.
3438 The reason for this restriction is really an implementation one.
3439 Type-checking binding groups is already a nightmare without
3440 existentials complicating the picture. Also an existential pattern
3441 binding at the top level of a module doesn't make sense, because it's
3442 not clear how to prevent the existentially-quantified type "escaping".
3443 So for now, there's a simple-to-state restriction. We'll see how
3451 You can't use existential quantification for <Literal>newtype</Literal>
3452 declarations. So this is illegal:
3456 newtype T = forall a. Ord a => MkT a
3460 Reason: a value of type <Literal>T</Literal> must be represented as a pair
3461 of a dictionary for <Literal>Ord t</Literal> and a value of type <Literal>t</Literal>.
3462 That contradicts the idea that <Literal>newtype</Literal> should have no
3463 concrete representation. You can get just the same efficiency and effect
3464 by using <Literal>data</Literal> instead of <Literal>newtype</Literal>. If there is no
3465 overloading involved, then there is more of a case for allowing
3466 an existentially-quantified <Literal>newtype</Literal>, because the <Literal>data</Literal>
3467 because the <Literal>data</Literal> version does carry an implementation cost,
3468 but single-field existentially quantified constructors aren't much
3469 use. So the simple restriction (no existential stuff on <Literal>newtype</Literal>)
3470 stands, unless there are convincing reasons to change it.
3478 You can't use <Literal>deriving</Literal> to define instances of a
3479 data type with existentially quantified data constructors.
3481 Reason: in most cases it would not make sense. For example:#
3484 data T = forall a. MkT [a] deriving( Eq )
3487 To derive <Literal>Eq</Literal> in the standard way we would need to have equality
3488 between the single component of two <Function>MkT</Function> constructors:
3492 (MkT a) == (MkT b) = ???
3495 But <VarName>a</VarName> and <VarName>b</VarName> have distinct types, and so can't be compared.
3496 It's just about possible to imagine examples in which the derived instance
3497 would make sense, but it seems altogether simpler simply to prohibit such
3498 declarations. Define your own instances!
3510 <Sect1 id="sec-assertions">
3512 <IndexTerm><Primary>Assertions</Primary></IndexTerm>
3516 If you want to make use of assertions in your standard Haskell code, you
3517 could define a function like the following:
3523 assert :: Bool -> a -> a
3524 assert False x = error "assertion failed!"
3531 which works, but gives you back a less than useful error message --
3532 an assertion failed, but which and where?
3536 One way out is to define an extended <Function>assert</Function> function which also
3537 takes a descriptive string to include in the error message and
3538 perhaps combine this with the use of a pre-processor which inserts
3539 the source location where <Function>assert</Function> was used.
3543 Ghc offers a helping hand here, doing all of this for you. For every
3544 use of <Function>assert</Function> in the user's source:
3550 kelvinToC :: Double -> Double
3551 kelvinToC k = assert (k &gt;= 0.0) (k+273.15)
3557 Ghc will rewrite this to also include the source location where the
3564 assert pred val ==> assertError "Main.hs|15" pred val
3570 The rewrite is only performed by the compiler when it spots
3571 applications of <Function>Exception.assert</Function>, so you can still define and
3572 use your own versions of <Function>assert</Function>, should you so wish. If not,
3573 import <Literal>Exception</Literal> to make use <Function>assert</Function> in your code.
3577 To have the compiler ignore uses of assert, use the compiler option
3578 <Option>-fignore-asserts</Option>. <IndexTerm><Primary>-fignore-asserts option</Primary></IndexTerm> That is,
3579 expressions of the form <Literal>assert pred e</Literal> will be rewritten to <Literal>e</Literal>.
3583 Assertion failures can be caught, see the documentation for the
3584 <literal>Exception</literal> library (<xref linkend="sec-Exception">)
3590 <Sect1 id="scoped-type-variables">
3591 <Title>Scoped Type Variables
3595 A <Emphasis>pattern type signature</Emphasis> can introduce a <Emphasis>scoped type
3596 variable</Emphasis>. For example
3602 f (xs::[a]) = ys ++ ys
3611 The pattern <Literal>(xs::[a])</Literal> includes a type signature for <VarName>xs</VarName>.
3612 This brings the type variable <Literal>a</Literal> into scope; it scopes over
3613 all the patterns and right hand sides for this equation for <Function>f</Function>.
3614 In particular, it is in scope at the type signature for <VarName>y</VarName>.
3618 At ordinary type signatures, such as that for <VarName>ys</VarName>, any type variables
3619 mentioned in the type signature <Emphasis>that are not in scope</Emphasis> are
3620 implicitly universally quantified. (If there are no type variables in
3621 scope, all type variables mentioned in the signature are universally
3622 quantified, which is just as in Haskell 98.) In this case, since <VarName>a</VarName>
3623 is in scope, it is not universally quantified, so the type of <VarName>ys</VarName> is
3624 the same as that of <VarName>xs</VarName>. In Haskell 98 it is not possible to declare
3625 a type for <VarName>ys</VarName>; a major benefit of scoped type variables is that
3626 it becomes possible to do so.
3630 Scoped type variables are implemented in both GHC and Hugs. Where the
3631 implementations differ from the specification below, those differences
3636 So much for the basic idea. Here are the details.
3640 <Title>Scope and implicit quantification</Title>
3648 All the type variables mentioned in the patterns for a single
3649 function definition equation, that are not already in scope,
3650 are brought into scope by the patterns. We describe this set as
3651 the <Emphasis>type variables bound by the equation</Emphasis>.
3658 The type variables thus brought into scope may be mentioned
3659 in ordinary type signatures or pattern type signatures anywhere within
3667 In ordinary type signatures, any type variable mentioned in the
3668 signature that is in scope is <Emphasis>not</Emphasis> universally quantified.
3675 Ordinary type signatures do not bring any new type variables
3676 into scope (except in the type signature itself!). So this is illegal:
3685 It's illegal because <VarName>a</VarName> is not in scope in the body of <Function>f</Function>,
3686 so the ordinary signature <Literal>x::a</Literal> is equivalent to <Literal>x::forall a.a</Literal>;
3687 and that is an incorrect typing.
3694 There is no implicit universal quantification on pattern type
3695 signatures, nor may one write an explicit <Literal>forall</Literal> type in a pattern
3696 type signature. The pattern type signature is a monotype.
3704 The type variables in the head of a <Literal>class</Literal> or <Literal>instance</Literal> declaration
3705 scope over the methods defined in the <Literal>where</Literal> part. For example:
3719 (Not implemented in Hugs yet, Dec 98).
3730 <Title>Polymorphism</Title>
3738 Pattern type signatures are completely orthogonal to ordinary, separate
3739 type signatures. The two can be used independently or together. There is
3740 no scoping associated with the names of the type variables in a separate type signature.
3745 f (xs::[b]) = reverse xs
3754 The function must be polymorphic in the type variables
3755 bound by all its equations. Operationally, the type variables bound
3756 by one equation must not:
3763 Be unified with a type (such as <Literal>Int</Literal>, or <Literal>[a]</Literal>).
3769 Be unified with a type variable free in the environment.
3775 Be unified with each other. (They may unify with the type variables
3776 bound by another equation for the same function, of course.)
3783 For example, the following all fail to type check:
3787 f (x::a) (y::b) = [x,y] -- a unifies with b
3789 g (x::a) = x + 1::Int -- a unifies with Int
3791 h x = let k (y::a) = [x,y] -- a is free in the
3792 in k x -- environment
3794 k (x::a) True = ... -- a unifies with Int
3795 k (x::Int) False = ...
3798 w (x::a) = x -- a unifies with [b]
3807 The pattern-bound type variable may, however, be constrained
3808 by the context of the principal type, thus:
3812 f (x::a) (y::a) = x+y*2
3816 gets the inferred type: <Literal>forall a. Num a => a -> a -> a</Literal>.
3827 <Title>Result type signatures</Title>
3835 The result type of a function can be given a signature,
3840 f (x::a) :: [a] = [x,x,x]
3844 The final <Literal>:: [a]</Literal> after all the patterns gives a signature to the
3845 result type. Sometimes this is the only way of naming the type variable
3850 f :: Int -> [a] -> [a]
3851 f n :: ([a] -> [a]) = let g (x::a, y::a) = (y,x)
3852 in \xs -> map g (reverse xs `zip` xs)
3864 Result type signatures are not yet implemented in Hugs.
3870 <Title>Pattern signatures on other constructs</Title>
3878 A pattern type signature can be on an arbitrary sub-pattern, not
3883 f ((x,y)::(a,b)) = (y,x) :: (b,a)
3892 Pattern type signatures, including the result part, can be used
3893 in lambda abstractions:
3897 (\ (x::a, y) :: a -> x)
3901 Type variables bound by these patterns must be polymorphic in
3902 the sense defined above.
3907 f1 (x::c) = f1 x -- ok
3908 f2 = \(x::c) -> f2 x -- not ok
3912 Here, <Function>f1</Function> is OK, but <Function>f2</Function> is not, because <VarName>c</VarName> gets unified
3913 with a type variable free in the environment, in this
3914 case, the type of <Function>f2</Function>, which is in the environment when
3915 the lambda abstraction is checked.
3922 Pattern type signatures, including the result part, can be used
3923 in <Literal>case</Literal> expressions:
3927 case e of { (x::a, y) :: a -> x }
3931 The pattern-bound type variables must, as usual,
3932 be polymorphic in the following sense: each case alternative,
3933 considered as a lambda abstraction, must be polymorphic.
3938 case (True,False) of { (x::a, y) -> x }
3942 Even though the context is that of a pair of booleans,
3943 the alternative itself is polymorphic. Of course, it is
3948 case (True,False) of { (x::Bool, y) -> x }
3957 To avoid ambiguity, the type after the “<Literal>::</Literal>” in a result
3958 pattern signature on a lambda or <Literal>case</Literal> must be atomic (i.e. a single
3959 token or a parenthesised type of some sort). To see why,
3960 consider how one would parse this:
3973 Pattern type signatures that bind new type variables
3974 may not be used in pattern bindings at all.
3979 f x = let (y, z::a) = x in ...
3983 But these are OK, because they do not bind fresh type variables:
3987 f1 x = let (y, z::Int) = x in ...
3988 f2 (x::(Int,a)) = let (y, z::a) = x in ...
3992 However a single variable is considered a degenerate function binding,
3993 rather than a degerate pattern binding, so this is permitted, even
3994 though it binds a type variable:
3998 f :: (b->b) = \(x::b) -> x
4007 Such degnerate function bindings do not fall under the monomorphism
4014 g :: a -> a -> Bool = \x y. x==y
4020 Here <Function>g</Function> has type <Literal>forall a. Eq a => a -> a -> Bool</Literal>, just as if
4021 <Function>g</Function> had a separate type signature. Lacking a type signature, <Function>g</Function>
4022 would get a monomorphic type.
4028 <Title>Existentials</Title>
4036 Pattern type signatures can bind existential type variables.
4041 data T = forall a. MkT [a]
4044 f (MkT [t::a]) = MkT t3
4061 <Sect1 id="pragmas">
4066 GHC supports several pragmas, or instructions to the compiler placed
4067 in the source code. Pragmas don't affect the meaning of the program,
4068 but they might affect the efficiency of the generated code.
4071 <Sect2 id="inline-pragma">
4072 <Title>INLINE pragma
4074 <IndexTerm><Primary>INLINE pragma</Primary></IndexTerm>
4075 <IndexTerm><Primary>pragma, INLINE</Primary></IndexTerm></Title>
4078 GHC (with <Option>-O</Option>, as always) tries to inline (or “unfold”)
4079 functions/values that are “small enough,” thus avoiding the call
4080 overhead and possibly exposing other more-wonderful optimisations.
4084 You will probably see these unfoldings (in Core syntax) in your
4089 Normally, if GHC decides a function is “too expensive” to inline, it
4090 will not do so, nor will it export that unfolding for other modules to
4095 The sledgehammer you can bring to bear is the
4096 <Literal>INLINE</Literal><IndexTerm><Primary>INLINE pragma</Primary></IndexTerm> pragma, used thusly:
4099 key_function :: Int -> String -> (Bool, Double)
4101 #ifdef __GLASGOW_HASKELL__
4102 {-# INLINE key_function #-}
4106 (You don't need to do the C pre-processor carry-on unless you're going
4107 to stick the code through HBC—it doesn't like <Literal>INLINE</Literal> pragmas.)
4111 The major effect of an <Literal>INLINE</Literal> pragma is to declare a function's
4112 “cost” to be very low. The normal unfolding machinery will then be
4113 very keen to inline it.
4117 An <Literal>INLINE</Literal> pragma for a function can be put anywhere its type
4118 signature could be put.
4122 <Literal>INLINE</Literal> pragmas are a particularly good idea for the
4123 <Literal>then</Literal>/<Literal>return</Literal> (or <Literal>bind</Literal>/<Literal>unit</Literal>) functions in a monad.
4124 For example, in GHC's own <Literal>UniqueSupply</Literal> monad code, we have:
4127 #ifdef __GLASGOW_HASKELL__
4128 {-# INLINE thenUs #-}
4129 {-# INLINE returnUs #-}
4137 <Sect2 id="noinline-pragma">
4138 <Title>NOINLINE pragma
4142 <IndexTerm><Primary>NOINLINE pragma</Primary></IndexTerm>
4143 <IndexTerm><Primary>pragma, NOINLINE</Primary></IndexTerm>
4147 The <Literal>NOINLINE</Literal> pragma does exactly what you'd expect: it stops the
4148 named function from being inlined by the compiler. You shouldn't ever
4149 need to do this, unless you're very cautious about code size.
4154 <Sect2 id="specialize-pragma">
4155 <Title>SPECIALIZE pragma
4159 <IndexTerm><Primary>SPECIALIZE pragma</Primary></IndexTerm>
4160 <IndexTerm><Primary>pragma, SPECIALIZE</Primary></IndexTerm>
4161 <IndexTerm><Primary>overloading, death to</Primary></IndexTerm>
4165 (UK spelling also accepted.) For key overloaded functions, you can
4166 create extra versions (NB: more code space) specialised to particular
4167 types. Thus, if you have an overloaded function:
4173 hammeredLookup :: Ord key => [(key, value)] -> key -> value
4179 If it is heavily used on lists with <Literal>Widget</Literal> keys, you could
4180 specialise it as follows:
4183 {-# SPECIALIZE hammeredLookup :: [(Widget, value)] -> Widget -> value #-}
4189 To get very fancy, you can also specify a named function to use for
4190 the specialised value, by adding <Literal>= blah</Literal>, as in:
4193 {-# SPECIALIZE hammeredLookup :: ...as before... = blah #-}
4196 It's <Emphasis>Your Responsibility</Emphasis> to make sure that <Function>blah</Function> really
4197 behaves as a specialised version of <Function>hammeredLookup</Function>!!!
4201 NOTE: the <Literal>=blah</Literal> feature isn't implemented in GHC 4.xx.
4205 An example in which the <Literal>= blah</Literal> form will Win Big:
4208 toDouble :: Real a => a -> Double
4209 toDouble = fromRational . toRational
4211 {-# SPECIALIZE toDouble :: Int -> Double = i2d #-}
4212 i2d (I# i) = D# (int2Double# i) -- uses Glasgow prim-op directly
4215 The <Function>i2d</Function> function is virtually one machine instruction; the
4216 default conversion—via an intermediate <Literal>Rational</Literal>—is obscenely
4217 expensive by comparison.
4221 By using the US spelling, your <Literal>SPECIALIZE</Literal> pragma will work with
4222 HBC, too. Note that HBC doesn't support the <Literal>= blah</Literal> form.
4226 A <Literal>SPECIALIZE</Literal> pragma for a function can be put anywhere its type
4227 signature could be put.
4232 <Sect2 id="specialize-instance-pragma">
4233 <Title>SPECIALIZE instance pragma
4237 <IndexTerm><Primary>SPECIALIZE pragma</Primary></IndexTerm>
4238 <IndexTerm><Primary>overloading, death to</Primary></IndexTerm>
4239 Same idea, except for instance declarations. For example:
4242 instance (Eq a) => Eq (Foo a) where { ... usual stuff ... }
4244 {-# SPECIALIZE instance Eq (Foo [(Int, Bar)] #-}
4247 Compatible with HBC, by the way.
4252 <Sect2 id="line-pragma">
4257 <IndexTerm><Primary>LINE pragma</Primary></IndexTerm>
4258 <IndexTerm><Primary>pragma, LINE</Primary></IndexTerm>
4262 This pragma is similar to C's <Literal>#line</Literal> pragma, and is mainly for use in
4263 automatically generated Haskell code. It lets you specify the line
4264 number and filename of the original code; for example
4270 {-# LINE 42 "Foo.vhs" #-}
4276 if you'd generated the current file from something called <Filename>Foo.vhs</Filename>
4277 and this line corresponds to line 42 in the original. GHC will adjust
4278 its error messages to refer to the line/file named in the <Literal>LINE</Literal>
4285 <Title>RULES pragma</Title>
4288 The RULES pragma lets you specify rewrite rules. It is described in
4289 <XRef LinkEnd="rewrite-rules">.
4296 <Sect1 id="rewrite-rules">
4297 <Title>Rewrite rules
4299 <IndexTerm><Primary>RULES pagma</Primary></IndexTerm>
4300 <IndexTerm><Primary>pragma, RULES</Primary></IndexTerm>
4301 <IndexTerm><Primary>rewrite rules</Primary></IndexTerm></Title>
4304 The programmer can specify rewrite rules as part of the source program
4305 (in a pragma). GHC applies these rewrite rules wherever it can.
4313 "map/map" forall f g xs. map f (map g xs) = map (f.g) xs
4320 <Title>Syntax</Title>
4323 From a syntactic point of view:
4329 Each rule has a name, enclosed in double quotes. The name itself has
4330 no significance at all. It is only used when reporting how many times the rule fired.
4336 There may be zero or more rules in a <Literal>RULES</Literal> pragma.
4342 Layout applies in a <Literal>RULES</Literal> pragma. Currently no new indentation level
4343 is set, so you must lay out your rules starting in the same column as the
4344 enclosing definitions.
4350 Each variable mentioned in a rule must either be in scope (e.g. <Function>map</Function>),
4351 or bound by the <Literal>forall</Literal> (e.g. <Function>f</Function>, <Function>g</Function>, <Function>xs</Function>). The variables bound by
4352 the <Literal>forall</Literal> are called the <Emphasis>pattern</Emphasis> variables. They are separated
4353 by spaces, just like in a type <Literal>forall</Literal>.
4359 A pattern variable may optionally have a type signature.
4360 If the type of the pattern variable is polymorphic, it <Emphasis>must</Emphasis> have a type signature.
4361 For example, here is the <Literal>foldr/build</Literal> rule:
4364 "fold/build" forall k z (g::forall b. (a->b->b) -> b -> b) .
4365 foldr k z (build g) = g k z
4368 Since <Function>g</Function> has a polymorphic type, it must have a type signature.
4375 The left hand side of a rule must consist of a top-level variable applied
4376 to arbitrary expressions. For example, this is <Emphasis>not</Emphasis> OK:
4379 "wrong1" forall e1 e2. case True of { True -> e1; False -> e2 } = e1
4380 "wrong2" forall f. f True = True
4383 In <Literal>"wrong1"</Literal>, the LHS is not an application; in <Literal>"wrong1"</Literal>, the LHS has a pattern variable
4390 A rule does not need to be in the same module as (any of) the
4391 variables it mentions, though of course they need to be in scope.
4397 Rules are automatically exported from a module, just as instance declarations are.
4408 <Title>Semantics</Title>
4411 From a semantic point of view:
4417 Rules are only applied if you use the <Option>-O</Option> flag.
4423 Rules are regarded as left-to-right rewrite rules.
4424 When GHC finds an expression that is a substitution instance of the LHS
4425 of a rule, it replaces the expression by the (appropriately-substituted) RHS.
4426 By "a substitution instance" we mean that the LHS can be made equal to the
4427 expression by substituting for the pattern variables.
4434 The LHS and RHS of a rule are typechecked, and must have the
4442 GHC makes absolutely no attempt to verify that the LHS and RHS
4443 of a rule have the same meaning. That is undecideable in general, and
4444 infeasible in most interesting cases. The responsibility is entirely the programmer's!
4451 GHC makes no attempt to make sure that the rules are confluent or
4452 terminating. For example:
4455 "loop" forall x,y. f x y = f y x
4458 This rule will cause the compiler to go into an infinite loop.
4465 If more than one rule matches a call, GHC will choose one arbitrarily to apply.
4471 GHC currently uses a very simple, syntactic, matching algorithm
4472 for matching a rule LHS with an expression. It seeks a substitution
4473 which makes the LHS and expression syntactically equal modulo alpha
4474 conversion. The pattern (rule), but not the expression, is eta-expanded if
4475 necessary. (Eta-expanding the epression can lead to laziness bugs.)
4476 But not beta conversion (that's called higher-order matching).
4480 Matching is carried out on GHC's intermediate language, which includes
4481 type abstractions and applications. So a rule only matches if the
4482 types match too. See <XRef LinkEnd="rule-spec"> below.
4488 GHC keeps trying to apply the rules as it optimises the program.
4489 For example, consider:
4498 The expression <Literal>s (t xs)</Literal> does not match the rule <Literal>"map/map"</Literal>, but GHC
4499 will substitute for <VarName>s</VarName> and <VarName>t</VarName>, giving an expression which does match.
4500 If <VarName>s</VarName> or <VarName>t</VarName> was (a) used more than once, and (b) large or a redex, then it would
4501 not be substituted, and the rule would not fire.
4508 In the earlier phases of compilation, GHC inlines <Emphasis>nothing
4509 that appears on the LHS of a rule</Emphasis>, because once you have substituted
4510 for something you can't match against it (given the simple minded
4511 matching). So if you write the rule
4514 "map/map" forall f,g. map f . map g = map (f.g)
4517 this <Emphasis>won't</Emphasis> match the expression <Literal>map f (map g xs)</Literal>.
4518 It will only match something written with explicit use of ".".
4519 Well, not quite. It <Emphasis>will</Emphasis> match the expression
4525 where <Function>wibble</Function> is defined:
4528 wibble f g = map f . map g
4531 because <Function>wibble</Function> will be inlined (it's small).
4533 Later on in compilation, GHC starts inlining even things on the
4534 LHS of rules, but still leaves the rules enabled. This inlining
4535 policy is controlled by the per-simplification-pass flag <Option>-finline-phase</Option><Emphasis>n</Emphasis>.
4542 All rules are implicitly exported from the module, and are therefore
4543 in force in any module that imports the module that defined the rule, directly
4544 or indirectly. (That is, if A imports B, which imports C, then C's rules are
4545 in force when compiling A.) The situation is very similar to that for instance
4557 <Title>List fusion</Title>
4560 The RULES mechanism is used to implement fusion (deforestation) of common list functions.
4561 If a "good consumer" consumes an intermediate list constructed by a "good producer", the
4562 intermediate list should be eliminated entirely.
4566 The following are good producers:
4578 Enumerations of <Literal>Int</Literal> and <Literal>Char</Literal> (e.g. <Literal>['a'..'z']</Literal>).
4584 Explicit lists (e.g. <Literal>[True, False]</Literal>)
4590 The cons constructor (e.g <Literal>3:4:[]</Literal>)
4596 <Function>++</Function>
4602 <Function>map</Function>
4608 <Function>filter</Function>
4614 <Function>iterate</Function>, <Function>repeat</Function>
4620 <Function>zip</Function>, <Function>zipWith</Function>
4629 The following are good consumers:
4641 <Function>array</Function> (on its second argument)
4647 <Function>length</Function>
4653 <Function>++</Function> (on its first argument)
4659 <Function>map</Function>
4665 <Function>filter</Function>
4671 <Function>concat</Function>
4677 <Function>unzip</Function>, <Function>unzip2</Function>, <Function>unzip3</Function>, <Function>unzip4</Function>
4683 <Function>zip</Function>, <Function>zipWith</Function> (but on one argument only; if both are good producers, <Function>zip</Function>
4684 will fuse with one but not the other)
4690 <Function>partition</Function>
4696 <Function>head</Function>
4702 <Function>and</Function>, <Function>or</Function>, <Function>any</Function>, <Function>all</Function>
4708 <Function>sequence_</Function>
4714 <Function>msum</Function>
4720 <Function>sortBy</Function>
4729 So, for example, the following should generate no intermediate lists:
4732 array (1,10) [(i,i*i) | i <- map (+ 1) [0..9]]
4738 This list could readily be extended; if there are Prelude functions that you use
4739 a lot which are not included, please tell us.
4743 If you want to write your own good consumers or producers, look at the
4744 Prelude definitions of the above functions to see how to do so.
4749 <Sect2 id="rule-spec">
4750 <Title>Specialisation
4754 Rewrite rules can be used to get the same effect as a feature
4755 present in earlier version of GHC:
4758 {-# SPECIALIZE fromIntegral :: Int8 -> Int16 = int8ToInt16 #-}
4761 This told GHC to use <Function>int8ToInt16</Function> instead of <Function>fromIntegral</Function> whenever
4762 the latter was called with type <Literal>Int8 -> Int16</Literal>. That is, rather than
4763 specialising the original definition of <Function>fromIntegral</Function> the programmer is
4764 promising that it is safe to use <Function>int8ToInt16</Function> instead.
4768 This feature is no longer in GHC. But rewrite rules let you do the
4773 "fromIntegral/Int8/Int16" fromIntegral = int8ToInt16
4777 This slightly odd-looking rule instructs GHC to replace <Function>fromIntegral</Function>
4778 by <Function>int8ToInt16</Function> <Emphasis>whenever the types match</Emphasis>. Speaking more operationally,
4779 GHC adds the type and dictionary applications to get the typed rule
4782 forall (d1::Integral Int8) (d2::Num Int16) .
4783 fromIntegral Int8 Int16 d1 d2 = int8ToInt16
4787 this rule does not need to be in the same file as fromIntegral,
4788 unlike the <Literal>SPECIALISE</Literal> pragmas which currently do (so that they
4789 have an original definition available to specialise).
4795 <Title>Controlling what's going on</Title>
4803 Use <Option>-ddump-rules</Option> to see what transformation rules GHC is using.
4809 Use <Option>-ddump-simpl-stats</Option> to see what rules are being fired.
4810 If you add <Option>-dppr-debug</Option> you get a more detailed listing.
4816 The defintion of (say) <Function>build</Function> in <FileName>PrelBase.lhs</FileName> looks llike this:
4819 build :: forall a. (forall b. (a -> b -> b) -> b -> b) -> [a]
4820 {-# INLINE build #-}
4824 Notice the <Literal>INLINE</Literal>! That prevents <Literal>(:)</Literal> from being inlined when compiling
4825 <Literal>PrelBase</Literal>, so that an importing module will “see” the <Literal>(:)</Literal>, and can
4826 match it on the LHS of a rule. <Literal>INLINE</Literal> prevents any inlining happening
4827 in the RHS of the <Literal>INLINE</Literal> thing. I regret the delicacy of this.
4834 In <Filename>ghc/lib/std/PrelBase.lhs</Filename> look at the rules for <Function>map</Function> to
4835 see how to write rules that will do fusion and yet give an efficient
4836 program even if fusion doesn't happen. More rules in <Filename>PrelList.lhs</Filename>.
projects / ghc-hetmet.git / blob