2 Copyright (C) 2000-2022 Free Software Foundation, Inc.
3 Contributed by Andy Vaught
5 This file is part of GCC.
7 GCC is free software; you can redistribute it and/or modify it under
8 the terms of the GNU General Public License as published by the Free
9 Software Foundation; either version 3, or (at your option) any later
12 GCC is distributed in the hope that it will be useful, but WITHOUT ANY
13 WARRANTY; without even the implied warranty of MERCHANTABILITY or
14 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
17 You should have received a copy of the GNU General Public License
18 along with GCC; see the file COPYING3. If not see
19 <http://www.gnu.org/licenses/>. */
21 /* Since target arithmetic must be done on the host, there has to
22 be some way of evaluating arithmetic expressions as the host
23 would evaluate them. We use the GNU MP library and the MPFR
24 library to do arithmetic, and this file provides the interface. */
28 #include "coretypes.h"
32 #include "target-memory.h"
33 #include "constructor.h"
37 /* MPFR does not have a direct replacement for mpz_set_f() from GMP.
38 It's easily implemented with a few calls though. */
41 gfc_mpfr_to_mpz (mpz_t z
, mpfr_t x
, locus
*where
)
45 if (mpfr_inf_p (x
) || mpfr_nan_p (x
))
47 gfc_error ("Conversion of an Infinity or Not-a-Number at %L "
53 e
= mpfr_get_z_exp (z
, x
);
56 mpz_mul_2exp (z
, z
, e
);
58 mpz_tdiv_q_2exp (z
, z
, -e
);
62 /* Set the model number precision by the requested KIND. */
65 gfc_set_model_kind (int kind
)
67 int index
= gfc_validate_kind (BT_REAL
, kind
, false);
70 base2prec
= gfc_real_kinds
[index
].digits
;
71 if (gfc_real_kinds
[index
].radix
!= 2)
72 base2prec
*= gfc_real_kinds
[index
].radix
/ 2;
73 mpfr_set_default_prec (base2prec
);
77 /* Set the model number precision from mpfr_t x. */
80 gfc_set_model (mpfr_t x
)
82 mpfr_set_default_prec (mpfr_get_prec (x
));
86 /* Given an arithmetic error code, return a pointer to a string that
87 explains the error. */
90 gfc_arith_error (arith code
)
97 p
= G_("Arithmetic OK at %L");
100 p
= G_("Arithmetic overflow at %L");
102 case ARITH_UNDERFLOW
:
103 p
= G_("Arithmetic underflow at %L");
106 p
= G_("Arithmetic NaN at %L");
109 p
= G_("Division by zero at %L");
111 case ARITH_INCOMMENSURATE
:
112 p
= G_("Array operands are incommensurate at %L");
114 case ARITH_ASYMMETRIC
:
115 p
= G_("Integer outside symmetric range implied by Standard Fortran"
118 case ARITH_WRONGCONCAT
:
119 p
= G_("Illegal type in character concatenation at %L");
123 gfc_internal_error ("gfc_arith_error(): Bad error code");
130 /* Get things ready to do math. */
133 gfc_arith_init_1 (void)
135 gfc_integer_info
*int_info
;
136 gfc_real_info
*real_info
;
140 mpfr_set_default_prec (128);
143 /* Convert the minimum and maximum values for each kind into their
144 GNU MP representation. */
145 for (int_info
= gfc_integer_kinds
; int_info
->kind
!= 0; int_info
++)
148 mpz_init (int_info
->huge
);
149 mpz_set_ui (int_info
->huge
, int_info
->radix
);
150 mpz_pow_ui (int_info
->huge
, int_info
->huge
, int_info
->digits
);
151 mpz_sub_ui (int_info
->huge
, int_info
->huge
, 1);
153 /* These are the numbers that are actually representable by the
154 target. For bases other than two, this needs to be changed. */
155 if (int_info
->radix
!= 2)
156 gfc_internal_error ("Fix min_int calculation");
158 /* See PRs 13490 and 17912, related to integer ranges.
159 The pedantic_min_int exists for range checking when a program
160 is compiled with -pedantic, and reflects the belief that
161 Standard Fortran requires integers to be symmetrical, i.e.
162 every negative integer must have a representable positive
163 absolute value, and vice versa. */
165 mpz_init (int_info
->pedantic_min_int
);
166 mpz_neg (int_info
->pedantic_min_int
, int_info
->huge
);
168 mpz_init (int_info
->min_int
);
169 mpz_sub_ui (int_info
->min_int
, int_info
->pedantic_min_int
, 1);
172 mpfr_set_z (a
, int_info
->huge
, GFC_RND_MODE
);
173 mpfr_log10 (a
, a
, GFC_RND_MODE
);
175 int_info
->range
= (int) mpfr_get_si (a
, GFC_RND_MODE
);
180 for (real_info
= gfc_real_kinds
; real_info
->kind
!= 0; real_info
++)
182 gfc_set_model_kind (real_info
->kind
);
187 /* huge(x) = (1 - b**(-p)) * b**(emax-1) * b */
189 mpfr_init (real_info
->huge
);
190 mpfr_set_ui (real_info
->huge
, 1, GFC_RND_MODE
);
191 mpfr_set_ui (a
, real_info
->radix
, GFC_RND_MODE
);
192 mpfr_pow_si (a
, a
, -real_info
->digits
, GFC_RND_MODE
);
193 mpfr_sub (real_info
->huge
, real_info
->huge
, a
, GFC_RND_MODE
);
196 mpfr_set_ui (a
, real_info
->radix
, GFC_RND_MODE
);
197 mpfr_pow_ui (a
, a
, real_info
->max_exponent
- 1, GFC_RND_MODE
);
199 /* (1 - b**(-p)) * b**(emax-1) */
200 mpfr_mul (real_info
->huge
, real_info
->huge
, a
, GFC_RND_MODE
);
202 /* (1 - b**(-p)) * b**(emax-1) * b */
203 mpfr_mul_ui (real_info
->huge
, real_info
->huge
, real_info
->radix
,
206 /* tiny(x) = b**(emin-1) */
207 mpfr_init (real_info
->tiny
);
208 mpfr_set_ui (real_info
->tiny
, real_info
->radix
, GFC_RND_MODE
);
209 mpfr_pow_si (real_info
->tiny
, real_info
->tiny
,
210 real_info
->min_exponent
- 1, GFC_RND_MODE
);
212 /* subnormal (x) = b**(emin - digit) */
213 mpfr_init (real_info
->subnormal
);
214 mpfr_set_ui (real_info
->subnormal
, real_info
->radix
, GFC_RND_MODE
);
215 mpfr_pow_si (real_info
->subnormal
, real_info
->subnormal
,
216 real_info
->min_exponent
- real_info
->digits
, GFC_RND_MODE
);
218 /* epsilon(x) = b**(1-p) */
219 mpfr_init (real_info
->epsilon
);
220 mpfr_set_ui (real_info
->epsilon
, real_info
->radix
, GFC_RND_MODE
);
221 mpfr_pow_si (real_info
->epsilon
, real_info
->epsilon
,
222 1 - real_info
->digits
, GFC_RND_MODE
);
224 /* range(x) = int(min(log10(huge(x)), -log10(tiny)) */
225 mpfr_log10 (a
, real_info
->huge
, GFC_RND_MODE
);
226 mpfr_log10 (b
, real_info
->tiny
, GFC_RND_MODE
);
227 mpfr_neg (b
, b
, GFC_RND_MODE
);
230 mpfr_min (a
, a
, b
, GFC_RND_MODE
);
232 real_info
->range
= (int) mpfr_get_si (a
, GFC_RND_MODE
);
234 /* precision(x) = int((p - 1) * log10(b)) + k */
235 mpfr_set_ui (a
, real_info
->radix
, GFC_RND_MODE
);
236 mpfr_log10 (a
, a
, GFC_RND_MODE
);
237 mpfr_mul_ui (a
, a
, real_info
->digits
- 1, GFC_RND_MODE
);
239 real_info
->precision
= (int) mpfr_get_si (a
, GFC_RND_MODE
);
241 /* If the radix is an integral power of 10, add one to the precision. */
242 for (i
= 10; i
<= real_info
->radix
; i
*= 10)
243 if (i
== real_info
->radix
)
244 real_info
->precision
++;
246 mpfr_clears (a
, b
, NULL
);
251 /* Clean up, get rid of numeric constants. */
254 gfc_arith_done_1 (void)
256 gfc_integer_info
*ip
;
259 for (ip
= gfc_integer_kinds
; ip
->kind
; ip
++)
261 mpz_clear (ip
->min_int
);
262 mpz_clear (ip
->pedantic_min_int
);
263 mpz_clear (ip
->huge
);
266 for (rp
= gfc_real_kinds
; rp
->kind
; rp
++)
267 mpfr_clears (rp
->epsilon
, rp
->huge
, rp
->tiny
, rp
->subnormal
, NULL
);
273 /* Given a wide character value and a character kind, determine whether
274 the character is representable for that kind. */
276 gfc_check_character_range (gfc_char_t c
, int kind
)
278 /* As wide characters are stored as 32-bit values, they're all
279 representable in UCS=4. */
284 return c
<= 255 ? true : false;
290 /* Given an integer and a kind, make sure that the integer lies within
291 the range of the kind. Returns ARITH_OK, ARITH_ASYMMETRIC or
295 gfc_check_integer_range (mpz_t p
, int kind
)
300 i
= gfc_validate_kind (BT_INTEGER
, kind
, false);
305 if (mpz_cmp (p
, gfc_integer_kinds
[i
].pedantic_min_int
) < 0)
306 result
= ARITH_ASYMMETRIC
;
310 if (flag_range_check
== 0)
313 if (mpz_cmp (p
, gfc_integer_kinds
[i
].min_int
) < 0
314 || mpz_cmp (p
, gfc_integer_kinds
[i
].huge
) > 0)
315 result
= ARITH_OVERFLOW
;
321 /* Given a real and a kind, make sure that the real lies within the
322 range of the kind. Returns ARITH_OK, ARITH_OVERFLOW or
326 gfc_check_real_range (mpfr_t p
, int kind
)
332 i
= gfc_validate_kind (BT_REAL
, kind
, false);
336 mpfr_abs (q
, p
, GFC_RND_MODE
);
342 if (flag_range_check
!= 0)
343 retval
= ARITH_OVERFLOW
;
345 else if (mpfr_nan_p (p
))
347 if (flag_range_check
!= 0)
350 else if (mpfr_sgn (q
) == 0)
355 else if (mpfr_cmp (q
, gfc_real_kinds
[i
].huge
) > 0)
357 if (flag_range_check
== 0)
358 mpfr_set_inf (p
, mpfr_sgn (p
));
360 retval
= ARITH_OVERFLOW
;
362 else if (mpfr_cmp (q
, gfc_real_kinds
[i
].subnormal
) < 0)
364 if (flag_range_check
== 0)
366 if (mpfr_sgn (p
) < 0)
368 mpfr_set_ui (p
, 0, GFC_RND_MODE
);
369 mpfr_set_si (q
, -1, GFC_RND_MODE
);
370 mpfr_copysign (p
, p
, q
, GFC_RND_MODE
);
373 mpfr_set_ui (p
, 0, GFC_RND_MODE
);
376 retval
= ARITH_UNDERFLOW
;
378 else if (mpfr_cmp (q
, gfc_real_kinds
[i
].tiny
) < 0)
380 mpfr_exp_t emin
, emax
;
383 /* Save current values of emin and emax. */
384 emin
= mpfr_get_emin ();
385 emax
= mpfr_get_emax ();
387 /* Set emin and emax for the current model number. */
388 en
= gfc_real_kinds
[i
].min_exponent
- gfc_real_kinds
[i
].digits
+ 1;
389 mpfr_set_emin ((mpfr_exp_t
) en
);
390 mpfr_set_emax ((mpfr_exp_t
) gfc_real_kinds
[i
].max_exponent
);
391 mpfr_check_range (q
, 0, GFC_RND_MODE
);
392 mpfr_subnormalize (q
, 0, GFC_RND_MODE
);
394 /* Reset emin and emax. */
395 mpfr_set_emin (emin
);
396 mpfr_set_emax (emax
);
398 /* Copy sign if needed. */
399 if (mpfr_sgn (p
) < 0)
400 mpfr_neg (p
, q
, MPFR_RNDN
);
402 mpfr_set (p
, q
, MPFR_RNDN
);
411 /* Low-level arithmetic functions. All of these subroutines assume
412 that all operands are of the same type and return an operand of the
413 same type. The other thing about these subroutines is that they
414 can fail in various ways -- overflow, underflow, division by zero,
415 zero raised to the zero, etc. */
418 gfc_arith_not (gfc_expr
*op1
, gfc_expr
**resultp
)
422 result
= gfc_get_constant_expr (BT_LOGICAL
, op1
->ts
.kind
, &op1
->where
);
423 result
->value
.logical
= !op1
->value
.logical
;
431 gfc_arith_and (gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**resultp
)
435 result
= gfc_get_constant_expr (BT_LOGICAL
, gfc_kind_max (op1
, op2
),
437 result
->value
.logical
= op1
->value
.logical
&& op2
->value
.logical
;
445 gfc_arith_or (gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**resultp
)
449 result
= gfc_get_constant_expr (BT_LOGICAL
, gfc_kind_max (op1
, op2
),
451 result
->value
.logical
= op1
->value
.logical
|| op2
->value
.logical
;
459 gfc_arith_eqv (gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**resultp
)
463 result
= gfc_get_constant_expr (BT_LOGICAL
, gfc_kind_max (op1
, op2
),
465 result
->value
.logical
= op1
->value
.logical
== op2
->value
.logical
;
473 gfc_arith_neqv (gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**resultp
)
477 result
= gfc_get_constant_expr (BT_LOGICAL
, gfc_kind_max (op1
, op2
),
479 result
->value
.logical
= op1
->value
.logical
!= op2
->value
.logical
;
486 /* Make sure a constant numeric expression is within the range for
487 its type and kind. Note that there's also a gfc_check_range(),
488 but that one deals with the intrinsic RANGE function. */
491 gfc_range_check (gfc_expr
*e
)
499 rc
= gfc_check_integer_range (e
->value
.integer
, e
->ts
.kind
);
503 rc
= gfc_check_real_range (e
->value
.real
, e
->ts
.kind
);
504 if (rc
== ARITH_UNDERFLOW
)
505 mpfr_set_ui (e
->value
.real
, 0, GFC_RND_MODE
);
506 if (rc
== ARITH_OVERFLOW
)
507 mpfr_set_inf (e
->value
.real
, mpfr_sgn (e
->value
.real
));
509 mpfr_set_nan (e
->value
.real
);
513 rc
= gfc_check_real_range (mpc_realref (e
->value
.complex), e
->ts
.kind
);
514 if (rc
== ARITH_UNDERFLOW
)
515 mpfr_set_ui (mpc_realref (e
->value
.complex), 0, GFC_RND_MODE
);
516 if (rc
== ARITH_OVERFLOW
)
517 mpfr_set_inf (mpc_realref (e
->value
.complex),
518 mpfr_sgn (mpc_realref (e
->value
.complex)));
520 mpfr_set_nan (mpc_realref (e
->value
.complex));
522 rc2
= gfc_check_real_range (mpc_imagref (e
->value
.complex), e
->ts
.kind
);
523 if (rc
== ARITH_UNDERFLOW
)
524 mpfr_set_ui (mpc_imagref (e
->value
.complex), 0, GFC_RND_MODE
);
525 if (rc
== ARITH_OVERFLOW
)
526 mpfr_set_inf (mpc_imagref (e
->value
.complex),
527 mpfr_sgn (mpc_imagref (e
->value
.complex)));
529 mpfr_set_nan (mpc_imagref (e
->value
.complex));
536 gfc_internal_error ("gfc_range_check(): Bad type");
543 /* Several of the following routines use the same set of statements to
544 check the validity of the result. Encapsulate the checking here. */
547 check_result (arith rc
, gfc_expr
*x
, gfc_expr
*r
, gfc_expr
**rp
)
551 if (val
== ARITH_UNDERFLOW
)
554 gfc_warning (OPT_Wunderflow
, gfc_arith_error (val
), &x
->where
);
558 if (val
== ARITH_ASYMMETRIC
)
560 gfc_warning (0, gfc_arith_error (val
), &x
->where
);
564 if (val
== ARITH_OK
|| val
== ARITH_OVERFLOW
)
573 /* It may seem silly to have a subroutine that actually computes the
574 unary plus of a constant, but it prevents us from making exceptions
575 in the code elsewhere. Used for unary plus and parenthesized
579 gfc_arith_identity (gfc_expr
*op1
, gfc_expr
**resultp
)
581 *resultp
= gfc_copy_expr (op1
);
587 gfc_arith_uminus (gfc_expr
*op1
, gfc_expr
**resultp
)
592 result
= gfc_get_constant_expr (op1
->ts
.type
, op1
->ts
.kind
, &op1
->where
);
594 switch (op1
->ts
.type
)
597 mpz_neg (result
->value
.integer
, op1
->value
.integer
);
601 mpfr_neg (result
->value
.real
, op1
->value
.real
, GFC_RND_MODE
);
605 mpc_neg (result
->value
.complex, op1
->value
.complex, GFC_MPC_RND_MODE
);
609 gfc_internal_error ("gfc_arith_uminus(): Bad basic type");
612 rc
= gfc_range_check (result
);
614 return check_result (rc
, op1
, result
, resultp
);
619 gfc_arith_plus (gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**resultp
)
624 result
= gfc_get_constant_expr (op1
->ts
.type
, op1
->ts
.kind
, &op1
->where
);
626 switch (op1
->ts
.type
)
629 mpz_add (result
->value
.integer
, op1
->value
.integer
, op2
->value
.integer
);
633 mpfr_add (result
->value
.real
, op1
->value
.real
, op2
->value
.real
,
638 mpc_add (result
->value
.complex, op1
->value
.complex, op2
->value
.complex,
643 gfc_internal_error ("gfc_arith_plus(): Bad basic type");
646 rc
= gfc_range_check (result
);
648 return check_result (rc
, op1
, result
, resultp
);
653 gfc_arith_minus (gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**resultp
)
658 result
= gfc_get_constant_expr (op1
->ts
.type
, op1
->ts
.kind
, &op1
->where
);
660 switch (op1
->ts
.type
)
663 mpz_sub (result
->value
.integer
, op1
->value
.integer
, op2
->value
.integer
);
667 mpfr_sub (result
->value
.real
, op1
->value
.real
, op2
->value
.real
,
672 mpc_sub (result
->value
.complex, op1
->value
.complex,
673 op2
->value
.complex, GFC_MPC_RND_MODE
);
677 gfc_internal_error ("gfc_arith_minus(): Bad basic type");
680 rc
= gfc_range_check (result
);
682 return check_result (rc
, op1
, result
, resultp
);
687 gfc_arith_times (gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**resultp
)
692 result
= gfc_get_constant_expr (op1
->ts
.type
, op1
->ts
.kind
, &op1
->where
);
694 switch (op1
->ts
.type
)
697 mpz_mul (result
->value
.integer
, op1
->value
.integer
, op2
->value
.integer
);
701 mpfr_mul (result
->value
.real
, op1
->value
.real
, op2
->value
.real
,
706 gfc_set_model (mpc_realref (op1
->value
.complex));
707 mpc_mul (result
->value
.complex, op1
->value
.complex, op2
->value
.complex,
712 gfc_internal_error ("gfc_arith_times(): Bad basic type");
715 rc
= gfc_range_check (result
);
717 return check_result (rc
, op1
, result
, resultp
);
722 gfc_arith_divide (gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**resultp
)
729 result
= gfc_get_constant_expr (op1
->ts
.type
, op1
->ts
.kind
, &op1
->where
);
731 switch (op1
->ts
.type
)
734 if (mpz_sgn (op2
->value
.integer
) == 0)
740 if (warn_integer_division
)
744 mpz_tdiv_qr (result
->value
.integer
, r
, op1
->value
.integer
,
747 if (mpz_cmp_si (r
, 0) != 0)
750 p
= mpz_get_str (NULL
, 10, result
->value
.integer
);
751 gfc_warning_now (OPT_Winteger_division
, "Integer division "
752 "truncated to constant %qs at %L", p
,
759 mpz_tdiv_q (result
->value
.integer
, op1
->value
.integer
,
765 if (mpfr_sgn (op2
->value
.real
) == 0 && flag_range_check
== 1)
771 mpfr_div (result
->value
.real
, op1
->value
.real
, op2
->value
.real
,
776 if (mpc_cmp_si_si (op2
->value
.complex, 0, 0) == 0
777 && flag_range_check
== 1)
783 gfc_set_model (mpc_realref (op1
->value
.complex));
784 if (mpc_cmp_si_si (op2
->value
.complex, 0, 0) == 0)
786 /* In Fortran, return (NaN + NaN I) for any zero divisor. See
788 mpfr_set_nan (mpc_realref (result
->value
.complex));
789 mpfr_set_nan (mpc_imagref (result
->value
.complex));
792 mpc_div (result
->value
.complex, op1
->value
.complex, op2
->value
.complex,
797 gfc_internal_error ("gfc_arith_divide(): Bad basic type");
801 rc
= gfc_range_check (result
);
803 return check_result (rc
, op1
, result
, resultp
);
806 /* Raise a number to a power. */
809 arith_power (gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**resultp
)
816 result
= gfc_get_constant_expr (op1
->ts
.type
, op1
->ts
.kind
, &op1
->where
);
818 switch (op2
->ts
.type
)
821 power_sign
= mpz_sgn (op2
->value
.integer
);
825 /* Handle something to the zeroth power. Since we're dealing
826 with integral exponents, there is no ambiguity in the
827 limiting procedure used to determine the value of 0**0. */
828 switch (op1
->ts
.type
)
831 mpz_set_ui (result
->value
.integer
, 1);
835 mpfr_set_ui (result
->value
.real
, 1, GFC_RND_MODE
);
839 mpc_set_ui (result
->value
.complex, 1, GFC_MPC_RND_MODE
);
843 gfc_internal_error ("arith_power(): Bad base");
848 switch (op1
->ts
.type
)
852 /* First, we simplify the cases of op1 == 1, 0 or -1. */
853 if (mpz_cmp_si (op1
->value
.integer
, 1) == 0)
856 mpz_set_si (result
->value
.integer
, 1);
858 else if (mpz_cmp_si (op1
->value
.integer
, 0) == 0)
860 /* 0**op2 == 0, if op2 > 0
861 0**op2 overflow, if op2 < 0 ; in that case, we
862 set the result to 0 and return ARITH_DIV0. */
863 mpz_set_si (result
->value
.integer
, 0);
864 if (mpz_cmp_si (op2
->value
.integer
, 0) < 0)
867 else if (mpz_cmp_si (op1
->value
.integer
, -1) == 0)
869 /* (-1)**op2 == (-1)**(mod(op2,2)) */
870 unsigned int odd
= mpz_fdiv_ui (op2
->value
.integer
, 2);
872 mpz_set_si (result
->value
.integer
, -1);
874 mpz_set_si (result
->value
.integer
, 1);
876 /* Then, we take care of op2 < 0. */
877 else if (mpz_cmp_si (op2
->value
.integer
, 0) < 0)
879 /* if op2 < 0, op1**op2 == 0 because abs(op1) > 1. */
880 mpz_set_si (result
->value
.integer
, 0);
881 if (warn_integer_division
)
882 gfc_warning_now (OPT_Winteger_division
, "Negative "
883 "exponent of integer has zero "
884 "result at %L", &result
->where
);
888 /* We have abs(op1) > 1 and op2 > 1.
889 If op2 > bit_size(op1), we'll have an out-of-range
893 k
= gfc_validate_kind (BT_INTEGER
, op1
->ts
.kind
, false);
894 power
= gfc_integer_kinds
[k
].bit_size
;
895 if (mpz_cmp_si (op2
->value
.integer
, power
) < 0)
897 gfc_extract_int (op2
, &power
);
898 mpz_pow_ui (result
->value
.integer
, op1
->value
.integer
,
900 rc
= gfc_range_check (result
);
901 if (rc
== ARITH_OVERFLOW
)
902 gfc_error_now ("Result of exponentiation at %L "
903 "exceeds the range of %s", &op1
->where
,
904 gfc_typename (&(op1
->ts
)));
908 /* Provide a nonsense value to propagate up. */
909 mpz_set (result
->value
.integer
,
910 gfc_integer_kinds
[k
].huge
);
911 mpz_add_ui (result
->value
.integer
,
912 result
->value
.integer
, 1);
920 mpfr_pow_z (result
->value
.real
, op1
->value
.real
,
921 op2
->value
.integer
, GFC_RND_MODE
);
925 mpc_pow_z (result
->value
.complex, op1
->value
.complex,
926 op2
->value
.integer
, GFC_MPC_RND_MODE
);
937 if (gfc_init_expr_flag
)
939 if (!gfc_notify_std (GFC_STD_F2003
, "Noninteger "
940 "exponent in an initialization "
941 "expression at %L", &op2
->where
))
943 gfc_free_expr (result
);
944 return ARITH_PROHIBIT
;
948 if (mpfr_cmp_si (op1
->value
.real
, 0) < 0)
950 gfc_error ("Raising a negative REAL at %L to "
951 "a REAL power is prohibited", &op1
->where
);
952 gfc_free_expr (result
);
953 return ARITH_PROHIBIT
;
956 mpfr_pow (result
->value
.real
, op1
->value
.real
, op2
->value
.real
,
962 if (gfc_init_expr_flag
)
964 if (!gfc_notify_std (GFC_STD_F2003
, "Noninteger "
965 "exponent in an initialization "
966 "expression at %L", &op2
->where
))
968 gfc_free_expr (result
);
969 return ARITH_PROHIBIT
;
973 mpc_pow (result
->value
.complex, op1
->value
.complex,
974 op2
->value
.complex, GFC_MPC_RND_MODE
);
978 gfc_internal_error ("arith_power(): unknown type");
982 rc
= gfc_range_check (result
);
984 return check_result (rc
, op1
, result
, resultp
);
988 /* Concatenate two string constants. */
991 gfc_arith_concat (gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**resultp
)
996 /* By cleverly playing around with constructors, it is possible
997 to get mismaching types here. */
998 if (op1
->ts
.type
!= BT_CHARACTER
|| op2
->ts
.type
!= BT_CHARACTER
999 || op1
->ts
.kind
!= op2
->ts
.kind
)
1000 return ARITH_WRONGCONCAT
;
1002 result
= gfc_get_constant_expr (BT_CHARACTER
, op1
->ts
.kind
,
1005 len
= op1
->value
.character
.length
+ op2
->value
.character
.length
;
1007 result
->value
.character
.string
= gfc_get_wide_string (len
+ 1);
1008 result
->value
.character
.length
= len
;
1010 memcpy (result
->value
.character
.string
, op1
->value
.character
.string
,
1011 op1
->value
.character
.length
* sizeof (gfc_char_t
));
1013 memcpy (&result
->value
.character
.string
[op1
->value
.character
.length
],
1014 op2
->value
.character
.string
,
1015 op2
->value
.character
.length
* sizeof (gfc_char_t
));
1017 result
->value
.character
.string
[len
] = '\0';
1024 /* Comparison between real values; returns 0 if (op1 .op. op2) is true.
1025 This function mimics mpfr_cmp but takes NaN into account. */
1028 compare_real (gfc_expr
*op1
, gfc_expr
*op2
, gfc_intrinsic_op op
)
1034 rc
= mpfr_equal_p (op1
->value
.real
, op2
->value
.real
) ? 0 : 1;
1037 rc
= mpfr_greater_p (op1
->value
.real
, op2
->value
.real
) ? 1 : -1;
1040 rc
= mpfr_greaterequal_p (op1
->value
.real
, op2
->value
.real
) ? 1 : -1;
1043 rc
= mpfr_less_p (op1
->value
.real
, op2
->value
.real
) ? -1 : 1;
1046 rc
= mpfr_lessequal_p (op1
->value
.real
, op2
->value
.real
) ? -1 : 1;
1049 gfc_internal_error ("compare_real(): Bad operator");
1055 /* Comparison operators. Assumes that the two expression nodes
1056 contain two constants of the same type. The op argument is
1057 needed to handle NaN correctly. */
1060 gfc_compare_expr (gfc_expr
*op1
, gfc_expr
*op2
, gfc_intrinsic_op op
)
1064 switch (op1
->ts
.type
)
1067 rc
= mpz_cmp (op1
->value
.integer
, op2
->value
.integer
);
1071 rc
= compare_real (op1
, op2
, op
);
1075 rc
= gfc_compare_string (op1
, op2
);
1079 rc
= ((!op1
->value
.logical
&& op2
->value
.logical
)
1080 || (op1
->value
.logical
&& !op2
->value
.logical
));
1084 gfc_internal_error ("gfc_compare_expr(): Bad basic type");
1091 /* Compare a pair of complex numbers. Naturally, this is only for
1092 equality and inequality. */
1095 compare_complex (gfc_expr
*op1
, gfc_expr
*op2
)
1097 return mpc_cmp (op1
->value
.complex, op2
->value
.complex) == 0;
1101 /* Given two constant strings and the inverse collating sequence, compare the
1102 strings. We return -1 for a < b, 0 for a == b and 1 for a > b.
1103 We use the processor's default collating sequence. */
1106 gfc_compare_string (gfc_expr
*a
, gfc_expr
*b
)
1108 size_t len
, alen
, blen
, i
;
1111 alen
= a
->value
.character
.length
;
1112 blen
= b
->value
.character
.length
;
1114 len
= MAX(alen
, blen
);
1116 for (i
= 0; i
< len
; i
++)
1118 ac
= ((i
< alen
) ? a
->value
.character
.string
[i
] : ' ');
1119 bc
= ((i
< blen
) ? b
->value
.character
.string
[i
] : ' ');
1127 /* Strings are equal */
1133 gfc_compare_with_Cstring (gfc_expr
*a
, const char *b
, bool case_sensitive
)
1135 size_t len
, alen
, blen
, i
;
1138 alen
= a
->value
.character
.length
;
1141 len
= MAX(alen
, blen
);
1143 for (i
= 0; i
< len
; i
++)
1145 ac
= ((i
< alen
) ? a
->value
.character
.string
[i
] : ' ');
1146 bc
= ((i
< blen
) ? b
[i
] : ' ');
1148 if (!case_sensitive
)
1160 /* Strings are equal */
1165 /* Specific comparison subroutines. */
1168 gfc_arith_eq (gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**resultp
)
1172 result
= gfc_get_constant_expr (BT_LOGICAL
, gfc_default_logical_kind
,
1174 result
->value
.logical
= (op1
->ts
.type
== BT_COMPLEX
)
1175 ? compare_complex (op1
, op2
)
1176 : (gfc_compare_expr (op1
, op2
, INTRINSIC_EQ
) == 0);
1184 gfc_arith_ne (gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**resultp
)
1188 result
= gfc_get_constant_expr (BT_LOGICAL
, gfc_default_logical_kind
,
1190 result
->value
.logical
= (op1
->ts
.type
== BT_COMPLEX
)
1191 ? !compare_complex (op1
, op2
)
1192 : (gfc_compare_expr (op1
, op2
, INTRINSIC_EQ
) != 0);
1200 gfc_arith_gt (gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**resultp
)
1204 result
= gfc_get_constant_expr (BT_LOGICAL
, gfc_default_logical_kind
,
1206 result
->value
.logical
= (gfc_compare_expr (op1
, op2
, INTRINSIC_GT
) > 0);
1214 gfc_arith_ge (gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**resultp
)
1218 result
= gfc_get_constant_expr (BT_LOGICAL
, gfc_default_logical_kind
,
1220 result
->value
.logical
= (gfc_compare_expr (op1
, op2
, INTRINSIC_GE
) >= 0);
1228 gfc_arith_lt (gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**resultp
)
1232 result
= gfc_get_constant_expr (BT_LOGICAL
, gfc_default_logical_kind
,
1234 result
->value
.logical
= (gfc_compare_expr (op1
, op2
, INTRINSIC_LT
) < 0);
1242 gfc_arith_le (gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**resultp
)
1246 result
= gfc_get_constant_expr (BT_LOGICAL
, gfc_default_logical_kind
,
1248 result
->value
.logical
= (gfc_compare_expr (op1
, op2
, INTRINSIC_LE
) <= 0);
1256 reduce_unary (arith (*eval
) (gfc_expr
*, gfc_expr
**), gfc_expr
*op
,
1259 gfc_constructor_base head
;
1264 if (op
->expr_type
== EXPR_CONSTANT
)
1265 return eval (op
, result
);
1268 head
= gfc_constructor_copy (op
->value
.constructor
);
1269 for (c
= gfc_constructor_first (head
); c
; c
= gfc_constructor_next (c
))
1271 rc
= reduce_unary (eval
, c
->expr
, &r
);
1276 gfc_replace_expr (c
->expr
, r
);
1280 gfc_constructor_free (head
);
1283 gfc_constructor
*c
= gfc_constructor_first (head
);
1284 r
= gfc_get_array_expr (c
->expr
->ts
.type
, c
->expr
->ts
.kind
,
1286 r
->shape
= gfc_copy_shape (op
->shape
, op
->rank
);
1288 r
->value
.constructor
= head
;
1297 reduce_binary_ac (arith (*eval
) (gfc_expr
*, gfc_expr
*, gfc_expr
**),
1298 gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**result
)
1300 gfc_constructor_base head
;
1303 arith rc
= ARITH_OK
;
1305 head
= gfc_constructor_copy (op1
->value
.constructor
);
1306 for (c
= gfc_constructor_first (head
); c
; c
= gfc_constructor_next (c
))
1308 gfc_simplify_expr (c
->expr
, 0);
1310 if (c
->expr
->expr_type
== EXPR_CONSTANT
)
1311 rc
= eval (c
->expr
, op2
, &r
);
1313 rc
= reduce_binary_ac (eval
, c
->expr
, op2
, &r
);
1318 gfc_replace_expr (c
->expr
, r
);
1322 gfc_constructor_free (head
);
1325 gfc_constructor
*c
= gfc_constructor_first (head
);
1328 r
= gfc_get_array_expr (c
->expr
->ts
.type
, c
->expr
->ts
.kind
,
1330 r
->shape
= gfc_copy_shape (op1
->shape
, op1
->rank
);
1334 gcc_assert (op1
->ts
.type
!= BT_UNKNOWN
);
1335 r
= gfc_get_array_expr (op1
->ts
.type
, op1
->ts
.kind
,
1337 r
->shape
= gfc_get_shape (op1
->rank
);
1339 r
->rank
= op1
->rank
;
1340 r
->value
.constructor
= head
;
1349 reduce_binary_ca (arith (*eval
) (gfc_expr
*, gfc_expr
*, gfc_expr
**),
1350 gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**result
)
1352 gfc_constructor_base head
;
1355 arith rc
= ARITH_OK
;
1357 head
= gfc_constructor_copy (op2
->value
.constructor
);
1358 for (c
= gfc_constructor_first (head
); c
; c
= gfc_constructor_next (c
))
1360 gfc_simplify_expr (c
->expr
, 0);
1362 if (c
->expr
->expr_type
== EXPR_CONSTANT
)
1363 rc
= eval (op1
, c
->expr
, &r
);
1365 rc
= reduce_binary_ca (eval
, op1
, c
->expr
, &r
);
1370 gfc_replace_expr (c
->expr
, r
);
1374 gfc_constructor_free (head
);
1377 gfc_constructor
*c
= gfc_constructor_first (head
);
1380 r
= gfc_get_array_expr (c
->expr
->ts
.type
, c
->expr
->ts
.kind
,
1382 r
->shape
= gfc_copy_shape (op2
->shape
, op2
->rank
);
1386 gcc_assert (op2
->ts
.type
!= BT_UNKNOWN
);
1387 r
= gfc_get_array_expr (op2
->ts
.type
, op2
->ts
.kind
,
1389 r
->shape
= gfc_get_shape (op2
->rank
);
1391 r
->rank
= op2
->rank
;
1392 r
->value
.constructor
= head
;
1400 /* We need a forward declaration of reduce_binary. */
1401 static arith
reduce_binary (arith (*eval
) (gfc_expr
*, gfc_expr
*, gfc_expr
**),
1402 gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**result
);
1406 reduce_binary_aa (arith (*eval
) (gfc_expr
*, gfc_expr
*, gfc_expr
**),
1407 gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**result
)
1409 gfc_constructor_base head
;
1410 gfc_constructor
*c
, *d
;
1412 arith rc
= ARITH_OK
;
1414 if (!gfc_check_conformance (op1
, op2
, _("elemental binary operation")))
1415 return ARITH_INCOMMENSURATE
;
1417 head
= gfc_constructor_copy (op1
->value
.constructor
);
1418 for (c
= gfc_constructor_first (head
),
1419 d
= gfc_constructor_first (op2
->value
.constructor
);
1421 c
= gfc_constructor_next (c
), d
= gfc_constructor_next (d
))
1423 rc
= reduce_binary (eval
, c
->expr
, d
->expr
, &r
);
1427 gfc_replace_expr (c
->expr
, r
);
1431 rc
= ARITH_INCOMMENSURATE
;
1434 gfc_constructor_free (head
);
1437 gfc_constructor
*c
= gfc_constructor_first (head
);
1438 r
= gfc_get_array_expr (c
->expr
->ts
.type
, c
->expr
->ts
.kind
,
1440 r
->shape
= gfc_copy_shape (op1
->shape
, op1
->rank
);
1441 r
->rank
= op1
->rank
;
1442 r
->value
.constructor
= head
;
1451 reduce_binary (arith (*eval
) (gfc_expr
*, gfc_expr
*, gfc_expr
**),
1452 gfc_expr
*op1
, gfc_expr
*op2
, gfc_expr
**result
)
1454 if (op1
->expr_type
== EXPR_CONSTANT
&& op2
->expr_type
== EXPR_CONSTANT
)
1455 return eval (op1
, op2
, result
);
1457 if (op1
->expr_type
== EXPR_CONSTANT
&& op2
->expr_type
== EXPR_ARRAY
)
1458 return reduce_binary_ca (eval
, op1
, op2
, result
);
1460 if (op1
->expr_type
== EXPR_ARRAY
&& op2
->expr_type
== EXPR_CONSTANT
)
1461 return reduce_binary_ac (eval
, op1
, op2
, result
);
1463 return reduce_binary_aa (eval
, op1
, op2
, result
);
1469 arith (*f2
)(gfc_expr
*, gfc_expr
**);
1470 arith (*f3
)(gfc_expr
*, gfc_expr
*, gfc_expr
**);
1474 /* High level arithmetic subroutines. These subroutines go into
1475 eval_intrinsic(), which can do one of several things to its
1476 operands. If the operands are incompatible with the intrinsic
1477 operation, we return a node pointing to the operands and hope that
1478 an operator interface is found during resolution.
1480 If the operands are compatible and are constants, then we try doing
1481 the arithmetic. We also handle the cases where either or both
1482 operands are array constructors. */
1485 eval_intrinsic (gfc_intrinsic_op op
,
1486 eval_f eval
, gfc_expr
*op1
, gfc_expr
*op2
)
1488 gfc_expr temp
, *result
;
1495 gfc_clear_ts (&temp
.ts
);
1501 if (op1
->ts
.type
!= BT_LOGICAL
)
1504 temp
.ts
.type
= BT_LOGICAL
;
1505 temp
.ts
.kind
= gfc_default_logical_kind
;
1509 /* Logical binary operators */
1512 case INTRINSIC_NEQV
:
1514 if (op1
->ts
.type
!= BT_LOGICAL
|| op2
->ts
.type
!= BT_LOGICAL
)
1517 temp
.ts
.type
= BT_LOGICAL
;
1518 temp
.ts
.kind
= gfc_default_logical_kind
;
1523 case INTRINSIC_UPLUS
:
1524 case INTRINSIC_UMINUS
:
1525 if (!gfc_numeric_ts (&op1
->ts
))
1532 case INTRINSIC_PARENTHESES
:
1537 /* Additional restrictions for ordering relations. */
1539 case INTRINSIC_GE_OS
:
1541 case INTRINSIC_LT_OS
:
1543 case INTRINSIC_LE_OS
:
1545 case INTRINSIC_GT_OS
:
1546 if (op1
->ts
.type
== BT_COMPLEX
|| op2
->ts
.type
== BT_COMPLEX
)
1548 temp
.ts
.type
= BT_LOGICAL
;
1549 temp
.ts
.kind
= gfc_default_logical_kind
;
1555 case INTRINSIC_EQ_OS
:
1557 case INTRINSIC_NE_OS
:
1558 if (op1
->ts
.type
== BT_CHARACTER
&& op2
->ts
.type
== BT_CHARACTER
)
1561 temp
.ts
.type
= BT_LOGICAL
;
1562 temp
.ts
.kind
= gfc_default_logical_kind
;
1564 /* If kind mismatch, exit and we'll error out later. */
1565 if (op1
->ts
.kind
!= op2
->ts
.kind
)
1572 /* Numeric binary */
1573 case INTRINSIC_PLUS
:
1574 case INTRINSIC_MINUS
:
1575 case INTRINSIC_TIMES
:
1576 case INTRINSIC_DIVIDE
:
1577 case INTRINSIC_POWER
:
1578 if (!gfc_numeric_ts (&op1
->ts
) || !gfc_numeric_ts (&op2
->ts
))
1581 /* Insert any necessary type conversions to make the operands
1584 temp
.expr_type
= EXPR_OP
;
1585 gfc_clear_ts (&temp
.ts
);
1586 temp
.value
.op
.op
= op
;
1588 temp
.value
.op
.op1
= op1
;
1589 temp
.value
.op
.op2
= op2
;
1591 gfc_type_convert_binary (&temp
, warn_conversion
|| warn_conversion_extra
);
1593 if (op
== INTRINSIC_EQ
|| op
== INTRINSIC_NE
1594 || op
== INTRINSIC_GE
|| op
== INTRINSIC_GT
1595 || op
== INTRINSIC_LE
|| op
== INTRINSIC_LT
1596 || op
== INTRINSIC_EQ_OS
|| op
== INTRINSIC_NE_OS
1597 || op
== INTRINSIC_GE_OS
|| op
== INTRINSIC_GT_OS
1598 || op
== INTRINSIC_LE_OS
|| op
== INTRINSIC_LT_OS
)
1600 temp
.ts
.type
= BT_LOGICAL
;
1601 temp
.ts
.kind
= gfc_default_logical_kind
;
1607 /* Character binary */
1608 case INTRINSIC_CONCAT
:
1609 if (op1
->ts
.type
!= BT_CHARACTER
|| op2
->ts
.type
!= BT_CHARACTER
1610 || op1
->ts
.kind
!= op2
->ts
.kind
)
1613 temp
.ts
.type
= BT_CHARACTER
;
1614 temp
.ts
.kind
= op1
->ts
.kind
;
1618 case INTRINSIC_USER
:
1622 gfc_internal_error ("eval_intrinsic(): Bad operator");
1625 if (op1
->expr_type
!= EXPR_CONSTANT
1626 && (op1
->expr_type
!= EXPR_ARRAY
1627 || !gfc_is_constant_expr (op1
) || !gfc_expanded_ac (op1
)))
1631 && op2
->expr_type
!= EXPR_CONSTANT
1632 && (op2
->expr_type
!= EXPR_ARRAY
1633 || !gfc_is_constant_expr (op2
) || !gfc_expanded_ac (op2
)))
1637 rc
= reduce_unary (eval
.f2
, op1
, &result
);
1639 rc
= reduce_binary (eval
.f3
, op1
, op2
, &result
);
1642 /* Something went wrong. */
1643 if (op
== INTRINSIC_POWER
&& rc
== ARITH_PROHIBIT
)
1648 gfc_error (gfc_arith_error (rc
), &op1
->where
);
1649 if (rc
== ARITH_OVERFLOW
)
1652 if (rc
== ARITH_DIV0
&& op2
->ts
.type
== BT_INTEGER
)
1653 gfc_seen_div0
= true;
1660 gfc_free_expr (op1
);
1661 gfc_free_expr (op2
);
1665 /* Create a run-time expression. */
1666 result
= gfc_get_operator_expr (&op1
->where
, op
, op1
, op2
);
1667 result
->ts
= temp
.ts
;
1673 /* Modify type of expression for zero size array. */
1676 eval_type_intrinsic0 (gfc_intrinsic_op iop
, gfc_expr
*op
)
1679 gfc_internal_error ("eval_type_intrinsic0(): op NULL");
1684 case INTRINSIC_GE_OS
:
1686 case INTRINSIC_LT_OS
:
1688 case INTRINSIC_LE_OS
:
1690 case INTRINSIC_GT_OS
:
1692 case INTRINSIC_EQ_OS
:
1694 case INTRINSIC_NE_OS
:
1695 op
->ts
.type
= BT_LOGICAL
;
1696 op
->ts
.kind
= gfc_default_logical_kind
;
1707 /* Return nonzero if the expression is a zero size array. */
1710 gfc_zero_size_array (gfc_expr
*e
)
1712 if (e
== NULL
|| e
->expr_type
!= EXPR_ARRAY
)
1715 return e
->value
.constructor
== NULL
;
1719 /* Reduce a binary expression where at least one of the operands
1720 involves a zero-length array. Returns NULL if neither of the
1721 operands is a zero-length array. */
1724 reduce_binary0 (gfc_expr
*op1
, gfc_expr
*op2
)
1726 if (gfc_zero_size_array (op1
))
1728 gfc_free_expr (op2
);
1732 if (gfc_zero_size_array (op2
))
1734 gfc_free_expr (op1
);
1743 eval_intrinsic_f2 (gfc_intrinsic_op op
,
1744 arith (*eval
) (gfc_expr
*, gfc_expr
**),
1745 gfc_expr
*op1
, gfc_expr
*op2
)
1752 if (gfc_zero_size_array (op1
))
1753 return eval_type_intrinsic0 (op
, op1
);
1757 result
= reduce_binary0 (op1
, op2
);
1759 return eval_type_intrinsic0 (op
, result
);
1763 return eval_intrinsic (op
, f
, op1
, op2
);
1768 eval_intrinsic_f3 (gfc_intrinsic_op op
,
1769 arith (*eval
) (gfc_expr
*, gfc_expr
*, gfc_expr
**),
1770 gfc_expr
*op1
, gfc_expr
*op2
)
1778 result
= reduce_binary0 (op1
, op2
);
1780 return eval_type_intrinsic0(op
, result
);
1783 return eval_intrinsic (op
, f
, op1
, op2
);
1788 gfc_parentheses (gfc_expr
*op
)
1790 if (gfc_is_constant_expr (op
))
1793 return eval_intrinsic_f2 (INTRINSIC_PARENTHESES
, gfc_arith_identity
,
1798 gfc_uplus (gfc_expr
*op
)
1800 return eval_intrinsic_f2 (INTRINSIC_UPLUS
, gfc_arith_identity
, op
, NULL
);
1805 gfc_uminus (gfc_expr
*op
)
1807 return eval_intrinsic_f2 (INTRINSIC_UMINUS
, gfc_arith_uminus
, op
, NULL
);
1812 gfc_add (gfc_expr
*op1
, gfc_expr
*op2
)
1814 return eval_intrinsic_f3 (INTRINSIC_PLUS
, gfc_arith_plus
, op1
, op2
);
1819 gfc_subtract (gfc_expr
*op1
, gfc_expr
*op2
)
1821 return eval_intrinsic_f3 (INTRINSIC_MINUS
, gfc_arith_minus
, op1
, op2
);
1826 gfc_multiply (gfc_expr
*op1
, gfc_expr
*op2
)
1828 return eval_intrinsic_f3 (INTRINSIC_TIMES
, gfc_arith_times
, op1
, op2
);
1833 gfc_divide (gfc_expr
*op1
, gfc_expr
*op2
)
1835 return eval_intrinsic_f3 (INTRINSIC_DIVIDE
, gfc_arith_divide
, op1
, op2
);
1840 gfc_power (gfc_expr
*op1
, gfc_expr
*op2
)
1842 return eval_intrinsic_f3 (INTRINSIC_POWER
, arith_power
, op1
, op2
);
1847 gfc_concat (gfc_expr
*op1
, gfc_expr
*op2
)
1849 return eval_intrinsic_f3 (INTRINSIC_CONCAT
, gfc_arith_concat
, op1
, op2
);
1854 gfc_and (gfc_expr
*op1
, gfc_expr
*op2
)
1856 return eval_intrinsic_f3 (INTRINSIC_AND
, gfc_arith_and
, op1
, op2
);
1861 gfc_or (gfc_expr
*op1
, gfc_expr
*op2
)
1863 return eval_intrinsic_f3 (INTRINSIC_OR
, gfc_arith_or
, op1
, op2
);
1868 gfc_not (gfc_expr
*op1
)
1870 return eval_intrinsic_f2 (INTRINSIC_NOT
, gfc_arith_not
, op1
, NULL
);
1875 gfc_eqv (gfc_expr
*op1
, gfc_expr
*op2
)
1877 return eval_intrinsic_f3 (INTRINSIC_EQV
, gfc_arith_eqv
, op1
, op2
);
1882 gfc_neqv (gfc_expr
*op1
, gfc_expr
*op2
)
1884 return eval_intrinsic_f3 (INTRINSIC_NEQV
, gfc_arith_neqv
, op1
, op2
);
1889 gfc_eq (gfc_expr
*op1
, gfc_expr
*op2
, gfc_intrinsic_op op
)
1891 return eval_intrinsic_f3 (op
, gfc_arith_eq
, op1
, op2
);
1896 gfc_ne (gfc_expr
*op1
, gfc_expr
*op2
, gfc_intrinsic_op op
)
1898 return eval_intrinsic_f3 (op
, gfc_arith_ne
, op1
, op2
);
1903 gfc_gt (gfc_expr
*op1
, gfc_expr
*op2
, gfc_intrinsic_op op
)
1905 return eval_intrinsic_f3 (op
, gfc_arith_gt
, op1
, op2
);
1910 gfc_ge (gfc_expr
*op1
, gfc_expr
*op2
, gfc_intrinsic_op op
)
1912 return eval_intrinsic_f3 (op
, gfc_arith_ge
, op1
, op2
);
1917 gfc_lt (gfc_expr
*op1
, gfc_expr
*op2
, gfc_intrinsic_op op
)
1919 return eval_intrinsic_f3 (op
, gfc_arith_lt
, op1
, op2
);
1924 gfc_le (gfc_expr
*op1
, gfc_expr
*op2
, gfc_intrinsic_op op
)
1926 return eval_intrinsic_f3 (op
, gfc_arith_le
, op1
, op2
);
1930 /******* Simplification of intrinsic functions with constant arguments *****/
1933 /* Deal with an arithmetic error. */
1936 arith_error (arith rc
, gfc_typespec
*from
, gfc_typespec
*to
, locus
*where
)
1941 gfc_error ("Arithmetic OK converting %s to %s at %L",
1942 gfc_typename (from
), gfc_typename (to
), where
);
1944 case ARITH_OVERFLOW
:
1945 gfc_error ("Arithmetic overflow converting %s to %s at %L. This check "
1946 "can be disabled with the option %<-fno-range-check%>",
1947 gfc_typename (from
), gfc_typename (to
), where
);
1949 case ARITH_UNDERFLOW
:
1950 gfc_error ("Arithmetic underflow converting %s to %s at %L. This check "
1951 "can be disabled with the option %<-fno-range-check%>",
1952 gfc_typename (from
), gfc_typename (to
), where
);
1955 gfc_error ("Arithmetic NaN converting %s to %s at %L. This check "
1956 "can be disabled with the option %<-fno-range-check%>",
1957 gfc_typename (from
), gfc_typename (to
), where
);
1960 gfc_error ("Division by zero converting %s to %s at %L",
1961 gfc_typename (from
), gfc_typename (to
), where
);
1963 case ARITH_INCOMMENSURATE
:
1964 gfc_error ("Array operands are incommensurate converting %s to %s at %L",
1965 gfc_typename (from
), gfc_typename (to
), where
);
1967 case ARITH_ASYMMETRIC
:
1968 gfc_error ("Integer outside symmetric range implied by Standard Fortran"
1969 " converting %s to %s at %L",
1970 gfc_typename (from
), gfc_typename (to
), where
);
1973 gfc_internal_error ("gfc_arith_error(): Bad error code");
1976 /* TODO: Do something about the error, i.e., throw exception, return
1980 /* Returns true if significant bits were lost when converting real
1981 constant r from from_kind to to_kind. */
1984 wprecision_real_real (mpfr_t r
, int from_kind
, int to_kind
)
1989 gfc_set_model_kind (to_kind
);
1991 gfc_set_model_kind (from_kind
);
1994 mpfr_set (rv
, r
, GFC_RND_MODE
);
1995 mpfr_sub (diff
, rv
, r
, GFC_RND_MODE
);
1997 ret
= ! mpfr_zero_p (diff
);
2003 /* Return true if conversion from an integer to a real loses precision. */
2006 wprecision_int_real (mpz_t n
, mpfr_t r
)
2011 mpfr_get_z (i
, r
, GFC_RND_MODE
);
2013 ret
= mpz_cmp_si (i
, 0) != 0;
2018 /* Convert integers to integers. */
2021 gfc_int2int (gfc_expr
*src
, int kind
)
2026 result
= gfc_get_constant_expr (BT_INTEGER
, kind
, &src
->where
);
2028 mpz_set (result
->value
.integer
, src
->value
.integer
);
2030 if ((rc
= gfc_check_integer_range (result
->value
.integer
, kind
)) != ARITH_OK
)
2032 if (rc
== ARITH_ASYMMETRIC
)
2034 gfc_warning (0, gfc_arith_error (rc
), &src
->where
);
2038 arith_error (rc
, &src
->ts
, &result
->ts
, &src
->where
);
2039 gfc_free_expr (result
);
2044 /* If we do not trap numeric overflow, we need to convert the number to
2045 signed, throwing away high-order bits if necessary. */
2046 if (flag_range_check
== 0)
2050 k
= gfc_validate_kind (BT_INTEGER
, kind
, false);
2051 gfc_convert_mpz_to_signed (result
->value
.integer
,
2052 gfc_integer_kinds
[k
].bit_size
);
2054 if (warn_conversion
&& !src
->do_not_warn
&& kind
< src
->ts
.kind
)
2055 gfc_warning_now (OPT_Wconversion
, "Conversion from %qs to %qs at %L",
2056 gfc_typename (&src
->ts
), gfc_typename (&result
->ts
),
2063 /* Convert integers to reals. */
2066 gfc_int2real (gfc_expr
*src
, int kind
)
2071 result
= gfc_get_constant_expr (BT_REAL
, kind
, &src
->where
);
2073 mpfr_set_z (result
->value
.real
, src
->value
.integer
, GFC_RND_MODE
);
2075 if ((rc
= gfc_check_real_range (result
->value
.real
, kind
)) != ARITH_OK
)
2077 arith_error (rc
, &src
->ts
, &result
->ts
, &src
->where
);
2078 gfc_free_expr (result
);
2083 && wprecision_int_real (src
->value
.integer
, result
->value
.real
))
2084 gfc_warning (OPT_Wconversion
, "Change of value in conversion "
2085 "from %qs to %qs at %L",
2086 gfc_typename (&src
->ts
),
2087 gfc_typename (&result
->ts
),
2094 /* Convert default integer to default complex. */
2097 gfc_int2complex (gfc_expr
*src
, int kind
)
2102 result
= gfc_get_constant_expr (BT_COMPLEX
, kind
, &src
->where
);
2104 mpc_set_z (result
->value
.complex, src
->value
.integer
, GFC_MPC_RND_MODE
);
2106 if ((rc
= gfc_check_real_range (mpc_realref (result
->value
.complex), kind
))
2109 arith_error (rc
, &src
->ts
, &result
->ts
, &src
->where
);
2110 gfc_free_expr (result
);
2115 && wprecision_int_real (src
->value
.integer
,
2116 mpc_realref (result
->value
.complex)))
2117 gfc_warning_now (OPT_Wconversion
, "Change of value in conversion "
2118 "from %qs to %qs at %L",
2119 gfc_typename (&src
->ts
),
2120 gfc_typename (&result
->ts
),
2127 /* Convert default real to default integer. */
2130 gfc_real2int (gfc_expr
*src
, int kind
)
2134 bool did_warn
= false;
2136 result
= gfc_get_constant_expr (BT_INTEGER
, kind
, &src
->where
);
2138 gfc_mpfr_to_mpz (result
->value
.integer
, src
->value
.real
, &src
->where
);
2140 if ((rc
= gfc_check_integer_range (result
->value
.integer
, kind
)) != ARITH_OK
)
2142 arith_error (rc
, &src
->ts
, &result
->ts
, &src
->where
);
2143 gfc_free_expr (result
);
2147 /* If there was a fractional part, warn about this. */
2149 if (warn_conversion
)
2153 mpfr_frac (f
, src
->value
.real
, GFC_RND_MODE
);
2154 if (mpfr_cmp_si (f
, 0) != 0)
2156 gfc_warning_now (OPT_Wconversion
, "Change of value in conversion "
2157 "from %qs to %qs at %L", gfc_typename (&src
->ts
),
2158 gfc_typename (&result
->ts
), &src
->where
);
2162 if (!did_warn
&& warn_conversion_extra
)
2164 gfc_warning_now (OPT_Wconversion_extra
, "Conversion from %qs to %qs "
2165 "at %L", gfc_typename (&src
->ts
),
2166 gfc_typename (&result
->ts
), &src
->where
);
2173 /* Convert real to real. */
2176 gfc_real2real (gfc_expr
*src
, int kind
)
2180 bool did_warn
= false;
2182 result
= gfc_get_constant_expr (BT_REAL
, kind
, &src
->where
);
2184 mpfr_set (result
->value
.real
, src
->value
.real
, GFC_RND_MODE
);
2186 rc
= gfc_check_real_range (result
->value
.real
, kind
);
2188 if (rc
== ARITH_UNDERFLOW
)
2191 gfc_warning (OPT_Woverflow
, gfc_arith_error (rc
), &src
->where
);
2192 mpfr_set_ui (result
->value
.real
, 0, GFC_RND_MODE
);
2194 else if (rc
!= ARITH_OK
)
2196 arith_error (rc
, &src
->ts
, &result
->ts
, &src
->where
);
2197 gfc_free_expr (result
);
2201 /* As a special bonus, don't warn about REAL values which are not changed by
2202 the conversion if -Wconversion is specified and -Wconversion-extra is
2205 if ((warn_conversion
|| warn_conversion_extra
) && src
->ts
.kind
> kind
)
2207 int w
= warn_conversion
? OPT_Wconversion
: OPT_Wconversion_extra
;
2209 /* Calculate the difference between the constant and the rounded
2210 value and check it against zero. */
2212 if (wprecision_real_real (src
->value
.real
, src
->ts
.kind
, kind
))
2214 gfc_warning_now (w
, "Change of value in conversion from "
2216 gfc_typename (&src
->ts
), gfc_typename (&result
->ts
),
2218 /* Make sure the conversion warning is not emitted again. */
2223 if (!did_warn
&& warn_conversion_extra
)
2224 gfc_warning_now (OPT_Wconversion_extra
, "Conversion from %qs to %qs "
2225 "at %L", gfc_typename(&src
->ts
),
2226 gfc_typename(&result
->ts
), &src
->where
);
2232 /* Convert real to complex. */
2235 gfc_real2complex (gfc_expr
*src
, int kind
)
2239 bool did_warn
= false;
2241 result
= gfc_get_constant_expr (BT_COMPLEX
, kind
, &src
->where
);
2243 mpc_set_fr (result
->value
.complex, src
->value
.real
, GFC_MPC_RND_MODE
);
2245 rc
= gfc_check_real_range (mpc_realref (result
->value
.complex), kind
);
2247 if (rc
== ARITH_UNDERFLOW
)
2250 gfc_warning (OPT_Woverflow
, gfc_arith_error (rc
), &src
->where
);
2251 mpfr_set_ui (mpc_realref (result
->value
.complex), 0, GFC_RND_MODE
);
2253 else if (rc
!= ARITH_OK
)
2255 arith_error (rc
, &src
->ts
, &result
->ts
, &src
->where
);
2256 gfc_free_expr (result
);
2260 if ((warn_conversion
|| warn_conversion_extra
) && src
->ts
.kind
> kind
)
2262 int w
= warn_conversion
? OPT_Wconversion
: OPT_Wconversion_extra
;
2264 if (wprecision_real_real (src
->value
.real
, src
->ts
.kind
, kind
))
2266 gfc_warning_now (w
, "Change of value in conversion from "
2268 gfc_typename (&src
->ts
), gfc_typename (&result
->ts
),
2270 /* Make sure the conversion warning is not emitted again. */
2275 if (!did_warn
&& warn_conversion_extra
)
2276 gfc_warning_now (OPT_Wconversion_extra
, "Conversion from %qs to %qs "
2277 "at %L", gfc_typename(&src
->ts
),
2278 gfc_typename(&result
->ts
), &src
->where
);
2284 /* Convert complex to integer. */
2287 gfc_complex2int (gfc_expr
*src
, int kind
)
2291 bool did_warn
= false;
2293 result
= gfc_get_constant_expr (BT_INTEGER
, kind
, &src
->where
);
2295 gfc_mpfr_to_mpz (result
->value
.integer
, mpc_realref (src
->value
.complex),
2298 if ((rc
= gfc_check_integer_range (result
->value
.integer
, kind
)) != ARITH_OK
)
2300 arith_error (rc
, &src
->ts
, &result
->ts
, &src
->where
);
2301 gfc_free_expr (result
);
2305 if (warn_conversion
|| warn_conversion_extra
)
2307 int w
= warn_conversion
? OPT_Wconversion
: OPT_Wconversion_extra
;
2309 /* See if we discarded an imaginary part. */
2310 if (mpfr_cmp_si (mpc_imagref (src
->value
.complex), 0) != 0)
2312 gfc_warning_now (w
, "Non-zero imaginary part discarded "
2313 "in conversion from %qs to %qs at %L",
2314 gfc_typename(&src
->ts
), gfc_typename (&result
->ts
),
2323 mpfr_frac (f
, src
->value
.real
, GFC_RND_MODE
);
2324 if (mpfr_cmp_si (f
, 0) != 0)
2326 gfc_warning_now (w
, "Change of value in conversion from "
2327 "%qs to %qs at %L", gfc_typename (&src
->ts
),
2328 gfc_typename (&result
->ts
), &src
->where
);
2334 if (!did_warn
&& warn_conversion_extra
)
2336 gfc_warning_now (OPT_Wconversion_extra
, "Conversion from %qs to %qs "
2337 "at %L", gfc_typename (&src
->ts
),
2338 gfc_typename (&result
->ts
), &src
->where
);
2346 /* Convert complex to real. */
2349 gfc_complex2real (gfc_expr
*src
, int kind
)
2353 bool did_warn
= false;
2355 result
= gfc_get_constant_expr (BT_REAL
, kind
, &src
->where
);
2357 mpc_real (result
->value
.real
, src
->value
.complex, GFC_RND_MODE
);
2359 rc
= gfc_check_real_range (result
->value
.real
, kind
);
2361 if (rc
== ARITH_UNDERFLOW
)
2364 gfc_warning (OPT_Woverflow
, gfc_arith_error (rc
), &src
->where
);
2365 mpfr_set_ui (result
->value
.real
, 0, GFC_RND_MODE
);
2369 arith_error (rc
, &src
->ts
, &result
->ts
, &src
->where
);
2370 gfc_free_expr (result
);
2374 if (warn_conversion
|| warn_conversion_extra
)
2376 int w
= warn_conversion
? OPT_Wconversion
: OPT_Wconversion_extra
;
2378 /* See if we discarded an imaginary part. */
2379 if (mpfr_cmp_si (mpc_imagref (src
->value
.complex), 0) != 0)
2381 gfc_warning (w
, "Non-zero imaginary part discarded "
2382 "in conversion from %qs to %qs at %L",
2383 gfc_typename(&src
->ts
), gfc_typename (&result
->ts
),
2388 /* Calculate the difference between the real constant and the rounded
2389 value and check it against zero. */
2391 if (kind
> src
->ts
.kind
2392 && wprecision_real_real (mpc_realref (src
->value
.complex),
2393 src
->ts
.kind
, kind
))
2395 gfc_warning_now (w
, "Change of value in conversion from "
2397 gfc_typename (&src
->ts
), gfc_typename (&result
->ts
),
2399 /* Make sure the conversion warning is not emitted again. */
2404 if (!did_warn
&& warn_conversion_extra
)
2405 gfc_warning_now (OPT_Wconversion
, "Conversion from %qs to %qs at %L",
2406 gfc_typename(&src
->ts
), gfc_typename (&result
->ts
),
2413 /* Convert complex to complex. */
2416 gfc_complex2complex (gfc_expr
*src
, int kind
)
2420 bool did_warn
= false;
2422 result
= gfc_get_constant_expr (BT_COMPLEX
, kind
, &src
->where
);
2424 mpc_set (result
->value
.complex, src
->value
.complex, GFC_MPC_RND_MODE
);
2426 rc
= gfc_check_real_range (mpc_realref (result
->value
.complex), kind
);
2428 if (rc
== ARITH_UNDERFLOW
)
2431 gfc_warning (OPT_Woverflow
, gfc_arith_error (rc
), &src
->where
);
2432 mpfr_set_ui (mpc_realref (result
->value
.complex), 0, GFC_RND_MODE
);
2434 else if (rc
!= ARITH_OK
)
2436 arith_error (rc
, &src
->ts
, &result
->ts
, &src
->where
);
2437 gfc_free_expr (result
);
2441 rc
= gfc_check_real_range (mpc_imagref (result
->value
.complex), kind
);
2443 if (rc
== ARITH_UNDERFLOW
)
2446 gfc_warning (OPT_Woverflow
, gfc_arith_error (rc
), &src
->where
);
2447 mpfr_set_ui (mpc_imagref (result
->value
.complex), 0, GFC_RND_MODE
);
2449 else if (rc
!= ARITH_OK
)
2451 arith_error (rc
, &src
->ts
, &result
->ts
, &src
->where
);
2452 gfc_free_expr (result
);
2456 if ((warn_conversion
|| warn_conversion_extra
) && src
->ts
.kind
> kind
2457 && (wprecision_real_real (mpc_realref (src
->value
.complex),
2459 || wprecision_real_real (mpc_imagref (src
->value
.complex),
2460 src
->ts
.kind
, kind
)))
2462 int w
= warn_conversion
? OPT_Wconversion
: OPT_Wconversion_extra
;
2464 gfc_warning_now (w
, "Change of value in conversion from "
2466 gfc_typename (&src
->ts
), gfc_typename (&result
->ts
),
2471 if (!did_warn
&& warn_conversion_extra
&& src
->ts
.kind
!= kind
)
2472 gfc_warning_now (OPT_Wconversion_extra
, "Conversion from %qs to %qs "
2473 "at %L", gfc_typename(&src
->ts
),
2474 gfc_typename (&result
->ts
), &src
->where
);
2480 /* Logical kind conversion. */
2483 gfc_log2log (gfc_expr
*src
, int kind
)
2487 result
= gfc_get_constant_expr (BT_LOGICAL
, kind
, &src
->where
);
2488 result
->value
.logical
= src
->value
.logical
;
2494 /* Convert logical to integer. */
2497 gfc_log2int (gfc_expr
*src
, int kind
)
2501 result
= gfc_get_constant_expr (BT_INTEGER
, kind
, &src
->where
);
2502 mpz_set_si (result
->value
.integer
, src
->value
.logical
);
2508 /* Convert integer to logical. */
2511 gfc_int2log (gfc_expr
*src
, int kind
)
2515 result
= gfc_get_constant_expr (BT_LOGICAL
, kind
, &src
->where
);
2516 result
->value
.logical
= (mpz_cmp_si (src
->value
.integer
, 0) != 0);
2521 /* Convert character to character. We only use wide strings internally,
2522 so we only set the kind. */
2525 gfc_character2character (gfc_expr
*src
, int kind
)
2528 result
= gfc_copy_expr (src
);
2529 result
->ts
.kind
= kind
;
2534 /* Helper function to set the representation in a Hollerith conversion.
2535 This assumes that the ts.type and ts.kind of the result have already
2539 hollerith2representation (gfc_expr
*result
, gfc_expr
*src
)
2541 size_t src_len
, result_len
;
2543 src_len
= src
->representation
.length
- src
->ts
.u
.pad
;
2544 gfc_target_expr_size (result
, &result_len
);
2546 if (src_len
> result_len
)
2548 gfc_warning (OPT_Wcharacter_truncation
, "The Hollerith constant at %L "
2549 "is truncated in conversion to %qs", &src
->where
,
2550 gfc_typename(&result
->ts
));
2553 result
->representation
.string
= XCNEWVEC (char, result_len
+ 1);
2554 memcpy (result
->representation
.string
, src
->representation
.string
,
2555 MIN (result_len
, src_len
));
2557 if (src_len
< result_len
)
2558 memset (&result
->representation
.string
[src_len
], ' ', result_len
- src_len
);
2560 result
->representation
.string
[result_len
] = '\0'; /* For debugger */
2561 result
->representation
.length
= result_len
;
2565 /* Helper function to set the representation in a character conversion.
2566 This assumes that the ts.type and ts.kind of the result have already
2570 character2representation (gfc_expr
*result
, gfc_expr
*src
)
2572 size_t src_len
, result_len
, i
;
2573 src_len
= src
->value
.character
.length
;
2574 gfc_target_expr_size (result
, &result_len
);
2576 if (src_len
> result_len
)
2577 gfc_warning (OPT_Wcharacter_truncation
, "The character constant at %L is "
2578 "truncated in conversion to %s", &src
->where
,
2579 gfc_typename(&result
->ts
));
2581 result
->representation
.string
= XCNEWVEC (char, result_len
+ 1);
2583 for (i
= 0; i
< MIN (result_len
, src_len
); i
++)
2584 result
->representation
.string
[i
] = (char) src
->value
.character
.string
[i
];
2586 if (src_len
< result_len
)
2587 memset (&result
->representation
.string
[src_len
], ' ',
2588 result_len
- src_len
);
2590 result
->representation
.string
[result_len
] = '\0'; /* For debugger. */
2591 result
->representation
.length
= result_len
;
2594 /* Convert Hollerith to integer. The constant will be padded or truncated. */
2597 gfc_hollerith2int (gfc_expr
*src
, int kind
)
2600 result
= gfc_get_constant_expr (BT_INTEGER
, kind
, &src
->where
);
2602 hollerith2representation (result
, src
);
2603 gfc_interpret_integer (kind
, (unsigned char *) result
->representation
.string
,
2604 result
->representation
.length
, result
->value
.integer
);
2609 /* Convert character to integer. The constant will be padded or truncated. */
2612 gfc_character2int (gfc_expr
*src
, int kind
)
2615 result
= gfc_get_constant_expr (BT_INTEGER
, kind
, &src
->where
);
2617 character2representation (result
, src
);
2618 gfc_interpret_integer (kind
, (unsigned char *) result
->representation
.string
,
2619 result
->representation
.length
, result
->value
.integer
);
2623 /* Convert Hollerith to real. The constant will be padded or truncated. */
2626 gfc_hollerith2real (gfc_expr
*src
, int kind
)
2629 result
= gfc_get_constant_expr (BT_REAL
, kind
, &src
->where
);
2631 hollerith2representation (result
, src
);
2632 gfc_interpret_float (kind
, (unsigned char *) result
->representation
.string
,
2633 result
->representation
.length
, result
->value
.real
);
2638 /* Convert character to real. The constant will be padded or truncated. */
2641 gfc_character2real (gfc_expr
*src
, int kind
)
2644 result
= gfc_get_constant_expr (BT_REAL
, kind
, &src
->where
);
2646 character2representation (result
, src
);
2647 gfc_interpret_float (kind
, (unsigned char *) result
->representation
.string
,
2648 result
->representation
.length
, result
->value
.real
);
2654 /* Convert Hollerith to complex. The constant will be padded or truncated. */
2657 gfc_hollerith2complex (gfc_expr
*src
, int kind
)
2660 result
= gfc_get_constant_expr (BT_COMPLEX
, kind
, &src
->where
);
2662 hollerith2representation (result
, src
);
2663 gfc_interpret_complex (kind
, (unsigned char *) result
->representation
.string
,
2664 result
->representation
.length
, result
->value
.complex);
2669 /* Convert character to complex. The constant will be padded or truncated. */
2672 gfc_character2complex (gfc_expr
*src
, int kind
)
2675 result
= gfc_get_constant_expr (BT_COMPLEX
, kind
, &src
->where
);
2677 character2representation (result
, src
);
2678 gfc_interpret_complex (kind
, (unsigned char *) result
->representation
.string
,
2679 result
->representation
.length
, result
->value
.complex);
2685 /* Convert Hollerith to character. */
2688 gfc_hollerith2character (gfc_expr
*src
, int kind
)
2692 result
= gfc_copy_expr (src
);
2693 result
->ts
.type
= BT_CHARACTER
;
2694 result
->ts
.kind
= kind
;
2695 result
->ts
.u
.pad
= 0;
2697 result
->value
.character
.length
= result
->representation
.length
;
2698 result
->value
.character
.string
2699 = gfc_char_to_widechar (result
->representation
.string
);
2705 /* Convert Hollerith to logical. The constant will be padded or truncated. */
2708 gfc_hollerith2logical (gfc_expr
*src
, int kind
)
2711 result
= gfc_get_constant_expr (BT_LOGICAL
, kind
, &src
->where
);
2713 hollerith2representation (result
, src
);
2714 gfc_interpret_logical (kind
, (unsigned char *) result
->representation
.string
,
2715 result
->representation
.length
, &result
->value
.logical
);
2720 /* Convert character to logical. The constant will be padded or truncated. */
2723 gfc_character2logical (gfc_expr
*src
, int kind
)
2726 result
= gfc_get_constant_expr (BT_LOGICAL
, kind
, &src
->where
);
2728 character2representation (result
, src
);
2729 gfc_interpret_logical (kind
, (unsigned char *) result
->representation
.string
,
2730 result
->representation
.length
, &result
->value
.logical
);