summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorJakub Jelinek <jakub@redhat.com>2021-12-31 16:59:38 +0100
committerJakub Jelinek <jakub@redhat.com>2021-12-31 16:59:38 +0100
commitf88bfae2382f16d9e2221d7f9905b285f2ef19f8 (patch)
tree3ce69ac89e89c26e0c2d118ad5b7fe6d48f9cf9d
parentb99cc1744a7bb5411ec1ab3200add33f35c85106 (diff)
gfortran: Introduce gfc_type_abi_kind
The following patch detects the powerpc64le-linux kind == 16 cases and for the -mabi=ieeelongdouble case (no matter whether it is the configured in default or just option used on the command line) uses _r17 or _c17 instead of _r16 or _c17 in the library API names. From what I can see, e.g. calls to sin on real(kind = 16) works fine with or without this patch (we call __builtin_sinl and the backend uses rs6000_mangle_decl_assembler_name which ensures __sinieee128 is called). What is clearly still broken is IO, where for real(kind=16) a a = 1.0 print *, a end we call _gfortran_transfer_real_write (&dt_parm.0, &a, 16); for both -mabi=ibmlongdouble and -mabi=ieeelongdouble I don't remember what was the agreement, do we want _gfortran_transfer_real_write (&dt_parm.0, &a, 17); for the ieeelongdouble case, or some new entrypoint for the abi_kind == 17 real/complex IO? Also, what about kind stored in array descriptors? Shall we use there the abi_kind or kind? I guess at least before the IO case is solved there is no point in checking the testsuite, too many things will be majorly broken... 2021-12-31 Jakub Jelinek <jakub@redhat.com> * gfortran.h (gfc_real_info): Add abi_kind member. (gfc_type_abi_kind): Declare. * trans-types.c (gfc_init_kinds): Initialize abi_kind. * intrinsic.c (gfc_type_abi_kind): New function. (conv_name): Use it. * iresolve.c (resolve_transformational, gfc_resolve_abs, gfc_resolve_char_achar, gfc_resolve_acos, gfc_resolve_acosh, gfc_resolve_aimag, gfc_resolve_and, gfc_resolve_aint, gfc_resolve_all, gfc_resolve_anint, gfc_resolve_any, gfc_resolve_asin, gfc_resolve_asinh, gfc_resolve_atan, gfc_resolve_atanh, gfc_resolve_atan2, gfc_resolve_bessel_n2, gfc_resolve_ceiling, gfc_resolve_cmplx, gfc_resolve_complex, gfc_resolve_cos, gfc_resolve_cosh, gfc_resolve_count, gfc_resolve_dble, gfc_resolve_dim, gfc_resolve_dot_product, gfc_resolve_dprod, gfc_resolve_exp, gfc_resolve_floor, gfc_resolve_hypot, gfc_resolve_int, gfc_resolve_int2, gfc_resolve_int8, gfc_resolve_long, gfc_resolve_log, gfc_resolve_log10, gfc_resolve_logical, gfc_resolve_matmul, gfc_resolve_minmax, gfc_resolve_maxloc, gfc_resolve_findloc, gfc_resolve_maxval, gfc_resolve_merge, gfc_resolve_minloc, gfc_resolve_minval, gfc_resolve_mod, gfc_resolve_modulo, gfc_resolve_nearest, gfc_resolve_or, gfc_resolve_real, gfc_resolve_realpart, gfc_resolve_reshape, gfc_resolve_sign, gfc_resolve_sin, gfc_resolve_sinh, gfc_resolve_sqrt, gfc_resolve_tan, gfc_resolve_tanh, gfc_resolve_transpose, gfc_resolve_trigd, gfc_resolve_xor, gfc_resolve_random_number): Likewise. * trans-decl.c (gfc_build_intrinsic_function_decls): Likewise.
-rw-r--r--gcc/fortran/gfortran.h8
-rw-r--r--gcc/fortran/intrinsic.c25
-rw-r--r--gcc/fortran/iresolve.c185
-rw-r--r--gcc/fortran/trans-decl.c10
-rw-r--r--gcc/fortran/trans-types.c25
5 files changed, 177 insertions, 76 deletions
diff --git a/gcc/fortran/gfortran.h b/gcc/fortran/gfortran.h
index 1846ee4fd3c..e2802e5108e 100644
--- a/gcc/fortran/gfortran.h
+++ b/gcc/fortran/gfortran.h
@@ -2643,7 +2643,7 @@ extern gfc_logical_info gfc_logical_kinds[];
typedef struct
{
mpfr_t epsilon, huge, tiny, subnormal;
- int kind, radix, digits, min_exponent, max_exponent;
+ int kind, abi_kind, radix, digits, min_exponent, max_exponent;
int range, precision;
/* The precision of the type as reported by GET_MODE_PRECISION. */
@@ -3499,6 +3499,12 @@ void gfc_intrinsic_init_1 (void);
void gfc_intrinsic_done_1 (void);
char gfc_type_letter (bt, bool logical_equals_int = false);
+int gfc_type_abi_kind (bt, int);
+static inline int
+gfc_type_abi_kind (gfc_typespec *ts)
+{
+ return gfc_type_abi_kind (ts->type, ts->kind);
+}
gfc_symbol * gfc_get_intrinsic_sub_symbol (const char *);
gfc_symbol *gfc_get_intrinsic_function_symbol (gfc_expr *);
gfc_symbol *gfc_find_intrinsic_symbol (gfc_expr *);
diff --git a/gcc/fortran/intrinsic.c b/gcc/fortran/intrinsic.c
index 3682f9ae21f..5452ac810bf 100644
--- a/gcc/fortran/intrinsic.c
+++ b/gcc/fortran/intrinsic.c
@@ -103,6 +103,27 @@ gfc_type_letter (bt type, bool logical_equals_int)
}
+/* Return kind that should be used for ABI purposes in libgfortran
+ APIs. Usually the same as ts->kind, except for BT_REAL/BT_COMPLEX
+ for IEEE 754 quad format kind 16 where it returns 17. */
+
+int
+gfc_type_abi_kind (bt type, int kind)
+{
+ switch (type)
+ {
+ case BT_REAL:
+ case BT_COMPLEX:
+ if (kind == 16)
+ for (int i = 0; gfc_real_kinds[i].kind != 0; i++)
+ if (gfc_real_kinds[i].kind == kind)
+ return gfc_real_kinds[i].abi_kind;
+ return kind;
+ default:
+ return kind;
+ }
+}
+
/* Get a symbol for a resolved name. Note, if needed be, the elemental
attribute has be added afterwards. */
@@ -167,8 +188,8 @@ static const char *
conv_name (gfc_typespec *from, gfc_typespec *to)
{
return gfc_get_string ("__convert_%c%d_%c%d",
- gfc_type_letter (from->type), from->kind,
- gfc_type_letter (to->type), to->kind);
+ gfc_type_letter (from->type), gfc_type_abi_kind (from),
+ gfc_type_letter (to->type), gfc_type_abi_kind (to));
}
diff --git a/gcc/fortran/iresolve.c b/gcc/fortran/iresolve.c
index 598c0409b66..73891554e61 100644
--- a/gcc/fortran/iresolve.c
+++ b/gcc/fortran/iresolve.c
@@ -191,7 +191,8 @@ resolve_transformational (const char *name, gfc_expr *f, gfc_expr *array,
f->value.function.name
= gfc_get_string (PREFIX ("%s%s_%c%d"), prefix, name,
- gfc_type_letter (array->ts.type), array->ts.kind);
+ gfc_type_letter (array->ts.type),
+ gfc_type_abi_kind (&array->ts));
}
@@ -206,7 +207,8 @@ gfc_resolve_abs (gfc_expr *f, gfc_expr *a)
f->ts.type = BT_REAL;
f->value.function.name
- = gfc_get_string ("__abs_%c%d", gfc_type_letter (a->ts.type), a->ts.kind);
+ = gfc_get_string ("__abs_%c%d", gfc_type_letter (a->ts.type),
+ gfc_type_abi_kind (&a->ts));
}
@@ -256,7 +258,8 @@ gfc_resolve_char_achar (gfc_expr *f, gfc_expr *x, gfc_expr *kind,
f->value.function.name
= gfc_get_string ("__%schar_%d_%c%d", is_achar ? "a" : "", f->ts.kind,
- gfc_type_letter (x->ts.type), x->ts.kind);
+ gfc_type_letter (x->ts.type),
+ gfc_type_abi_kind (&x->ts));
}
@@ -272,7 +275,8 @@ gfc_resolve_acos (gfc_expr *f, gfc_expr *x)
{
f->ts = x->ts;
f->value.function.name
- = gfc_get_string ("__acos_%c%d", gfc_type_letter (x->ts.type), x->ts.kind);
+ = gfc_get_string ("__acos_%c%d", gfc_type_letter (x->ts.type),
+ gfc_type_abi_kind (&x->ts));
}
@@ -282,7 +286,7 @@ gfc_resolve_acosh (gfc_expr *f, gfc_expr *x)
f->ts = x->ts;
f->value.function.name
= gfc_get_string ("__acosh_%c%d", gfc_type_letter (x->ts.type),
- x->ts.kind);
+ gfc_type_abi_kind (&x->ts));
}
@@ -293,7 +297,7 @@ gfc_resolve_aimag (gfc_expr *f, gfc_expr *x)
f->ts.kind = x->ts.kind;
f->value.function.name
= gfc_get_string ("__aimag_%c%d", gfc_type_letter (x->ts.type),
- x->ts.kind);
+ gfc_type_abi_kind (&x->ts));
}
@@ -312,7 +316,8 @@ gfc_resolve_and (gfc_expr *f, gfc_expr *i, gfc_expr *j)
}
f->value.function.name
- = gfc_get_string ("__and_%c%d", gfc_type_letter (i->ts.type), f->ts.kind);
+ = gfc_get_string ("__and_%c%d", gfc_type_letter (i->ts.type),
+ gfc_type_abi_kind (&f->ts));
}
@@ -334,7 +339,8 @@ gfc_resolve_aint (gfc_expr *f, gfc_expr *a, gfc_expr *kind)
/* The resolved name is only used for specific intrinsics where
the return kind is the same as the arg kind. */
f->value.function.name
- = gfc_get_string ("__aint_%c%d", gfc_type_letter (a->ts.type), a->ts.kind);
+ = gfc_get_string ("__aint_%c%d", gfc_type_letter (a->ts.type),
+ gfc_type_abi_kind (&a->ts));
}
@@ -359,7 +365,7 @@ gfc_resolve_all (gfc_expr *f, gfc_expr *mask, gfc_expr *dim)
f->value.function.name
= gfc_get_string (PREFIX ("all_%c%d"), gfc_type_letter (mask->ts.type),
- mask->ts.kind);
+ gfc_type_abi_kind (&mask->ts));
}
@@ -383,7 +389,7 @@ gfc_resolve_anint (gfc_expr *f, gfc_expr *a, gfc_expr *kind)
the return kind is the same as the arg kind. */
f->value.function.name
= gfc_get_string ("__anint_%c%d", gfc_type_letter (a->ts.type),
- a->ts.kind);
+ gfc_type_abi_kind (&a->ts));
}
@@ -408,7 +414,7 @@ gfc_resolve_any (gfc_expr *f, gfc_expr *mask, gfc_expr *dim)
f->value.function.name
= gfc_get_string (PREFIX ("any_%c%d"), gfc_type_letter (mask->ts.type),
- mask->ts.kind);
+ gfc_type_abi_kind (&mask->ts));
}
@@ -417,7 +423,8 @@ gfc_resolve_asin (gfc_expr *f, gfc_expr *x)
{
f->ts = x->ts;
f->value.function.name
- = gfc_get_string ("__asin_%c%d", gfc_type_letter (x->ts.type), x->ts.kind);
+ = gfc_get_string ("__asin_%c%d", gfc_type_letter (x->ts.type),
+ gfc_type_abi_kind (&x->ts));
}
void
@@ -426,7 +433,7 @@ gfc_resolve_asinh (gfc_expr *f, gfc_expr *x)
f->ts = x->ts;
f->value.function.name
= gfc_get_string ("__asinh_%c%d", gfc_type_letter (x->ts.type),
- x->ts.kind);
+ gfc_type_abi_kind (&x->ts));
}
void
@@ -434,7 +441,8 @@ gfc_resolve_atan (gfc_expr *f, gfc_expr *x)
{
f->ts = x->ts;
f->value.function.name
- = gfc_get_string ("__atan_%c%d", gfc_type_letter (x->ts.type), x->ts.kind);
+ = gfc_get_string ("__atan_%c%d", gfc_type_letter (x->ts.type),
+ gfc_type_abi_kind (&x->ts));
}
void
@@ -443,7 +451,7 @@ gfc_resolve_atanh (gfc_expr *f, gfc_expr *x)
f->ts = x->ts;
f->value.function.name
= gfc_get_string ("__atanh_%c%d", gfc_type_letter (x->ts.type),
- x->ts.kind);
+ gfc_type_abi_kind (&x->ts));
}
void
@@ -452,7 +460,7 @@ gfc_resolve_atan2 (gfc_expr *f, gfc_expr *x, gfc_expr *y ATTRIBUTE_UNUSED)
f->ts = x->ts;
f->value.function.name
= gfc_get_string ("__atan2_%c%d", gfc_type_letter (x->ts.type),
- x->ts.kind);
+ gfc_type_abi_kind (&x->ts));
}
@@ -507,10 +515,10 @@ gfc_resolve_bessel_n2 (gfc_expr *f, gfc_expr *n1, gfc_expr *n2, gfc_expr *x)
if (f->value.function.isym->id == GFC_ISYM_JN2)
f->value.function.name = gfc_get_string (PREFIX ("bessel_jn_r%d"),
- f->ts.kind);
+ gfc_type_abi_kind (&f->ts));
else
f->value.function.name = gfc_get_string (PREFIX ("bessel_yn_r%d"),
- f->ts.kind);
+ gfc_type_abi_kind (&f->ts));
}
@@ -546,7 +554,8 @@ gfc_resolve_ceiling (gfc_expr *f, gfc_expr *a, gfc_expr *kind)
? gfc_default_integer_kind : mpz_get_si (kind->value.integer);
f->value.function.name
= gfc_get_string ("__ceiling_%d_%c%d", f->ts.kind,
- gfc_type_letter (a->ts.type), a->ts.kind);
+ gfc_type_letter (a->ts.type),
+ gfc_type_abi_kind (&a->ts));
}
@@ -618,12 +627,15 @@ gfc_resolve_cmplx (gfc_expr *f, gfc_expr *x, gfc_expr *y, gfc_expr *kind)
if (y == NULL)
f->value.function.name
= gfc_get_string ("__cmplx0_%d_%c%d", f->ts.kind,
- gfc_type_letter (x->ts.type), x->ts.kind);
+ gfc_type_letter (x->ts.type),
+ gfc_type_abi_kind (&x->ts));
else
f->value.function.name
= gfc_get_string ("__cmplx1_%d_%c%d_%c%d", f->ts.kind,
- gfc_type_letter (x->ts.type), x->ts.kind,
- gfc_type_letter (y->ts.type), y->ts.kind);
+ gfc_type_letter (x->ts.type),
+ gfc_type_abi_kind (&x->ts),
+ gfc_type_letter (y->ts.type),
+ gfc_type_abi_kind (&y->ts));
}
@@ -659,8 +671,10 @@ gfc_resolve_complex (gfc_expr *f, gfc_expr *x, gfc_expr *y)
f->ts.kind = kind;
f->value.function.name
= gfc_get_string ("__cmplx1_%d_%c%d_%c%d", f->ts.kind,
- gfc_type_letter (x->ts.type), x->ts.kind,
- gfc_type_letter (y->ts.type), y->ts.kind);
+ gfc_type_letter (x->ts.type),
+ gfc_type_abi_kind (&x->ts),
+ gfc_type_letter (y->ts.type),
+ gfc_type_abi_kind (&y->ts));
}
@@ -677,7 +691,8 @@ gfc_resolve_cos (gfc_expr *f, gfc_expr *x)
{
f->ts = x->ts;
f->value.function.name
- = gfc_get_string ("__cos_%c%d", gfc_type_letter (x->ts.type), x->ts.kind);
+ = gfc_get_string ("__cos_%c%d", gfc_type_letter (x->ts.type),
+ gfc_type_abi_kind (&x->ts));
}
@@ -686,7 +701,8 @@ gfc_resolve_cosh (gfc_expr *f, gfc_expr *x)
{
f->ts = x->ts;
f->value.function.name
- = gfc_get_string ("__cosh_%c%d", gfc_type_letter (x->ts.type), x->ts.kind);
+ = gfc_get_string ("__cosh_%c%d", gfc_type_letter (x->ts.type),
+ gfc_type_abi_kind (&x->ts));
}
@@ -709,7 +725,7 @@ gfc_resolve_count (gfc_expr *f, gfc_expr *mask, gfc_expr *dim, gfc_expr *kind)
resolve_mask_arg (mask);
f->value.function.name
- = gfc_get_string (PREFIX ("count_%d_%c"), f->ts.kind,
+ = gfc_get_string (PREFIX ("count_%d_%c"), gfc_type_abi_kind (&f->ts),
gfc_type_letter (mask->ts.type));
}
@@ -810,7 +826,8 @@ gfc_resolve_dble (gfc_expr *f, gfc_expr *a)
f->ts.type = BT_REAL;
f->ts.kind = gfc_default_double_kind;
f->value.function.name
- = gfc_get_string ("__dble_%c%d", gfc_type_letter (a->ts.type), a->ts.kind);
+ = gfc_get_string ("__dble_%c%d", gfc_type_letter (a->ts.type),
+ gfc_type_abi_kind (&a->ts));
}
@@ -832,7 +849,8 @@ gfc_resolve_dim (gfc_expr *f, gfc_expr *a, gfc_expr *p)
}
f->value.function.name
- = gfc_get_string ("__dim_%c%d", gfc_type_letter (f->ts.type), f->ts.kind);
+ = gfc_get_string ("__dim_%c%d", gfc_type_letter (f->ts.type),
+ gfc_type_abi_kind (&f->ts));
}
@@ -850,7 +868,8 @@ gfc_resolve_dot_product (gfc_expr *f, gfc_expr *a, gfc_expr *b)
f->ts = temp.ts;
f->value.function.name
= gfc_get_string (PREFIX ("dot_product_%c%d"),
- gfc_type_letter (f->ts.type), f->ts.kind);
+ gfc_type_letter (f->ts.type),
+ gfc_type_abi_kind (&f->ts));
}
@@ -860,7 +879,8 @@ gfc_resolve_dprod (gfc_expr *f, gfc_expr *a ATTRIBUTE_UNUSED,
{
f->ts.kind = gfc_default_double_kind;
f->ts.type = BT_REAL;
- f->value.function.name = gfc_get_string ("__dprod_r%d", f->ts.kind);
+ f->value.function.name = gfc_get_string ("__dprod_r%d",
+ gfc_type_abi_kind (&f->ts));
}
@@ -951,7 +971,8 @@ gfc_resolve_exp (gfc_expr *f, gfc_expr *x)
{
f->ts = x->ts;
f->value.function.name
- = gfc_get_string ("__exp_%c%d", gfc_type_letter (x->ts.type), x->ts.kind);
+ = gfc_get_string ("__exp_%c%d", gfc_type_letter (x->ts.type),
+ gfc_type_abi_kind (&x->ts));
}
@@ -1044,7 +1065,8 @@ gfc_resolve_floor (gfc_expr *f, gfc_expr *a, gfc_expr *kind)
? gfc_default_integer_kind : mpz_get_si (kind->value.integer);
f->value.function.name
= gfc_get_string ("__floor%d_%c%d", f->ts.kind,
- gfc_type_letter (a->ts.type), a->ts.kind);
+ gfc_type_letter (a->ts.type),
+ gfc_type_abi_kind (&a->ts));
}
@@ -1135,7 +1157,8 @@ void
gfc_resolve_hypot (gfc_expr *f, gfc_expr *x, gfc_expr *y ATTRIBUTE_UNUSED)
{
f->ts = x->ts;
- f->value.function.name = gfc_get_string ("__hypot_r%d", x->ts.kind);
+ f->value.function.name = gfc_get_string ("__hypot_r%d",
+ gfc_type_abi_kind (&x->ts));
}
@@ -1311,7 +1334,8 @@ gfc_resolve_int (gfc_expr *f, gfc_expr *a, gfc_expr *kind)
? gfc_default_integer_kind : mpz_get_si (kind->value.integer);
f->value.function.name
= gfc_get_string ("__int_%d_%c%d", f->ts.kind,
- gfc_type_letter (a->ts.type), a->ts.kind);
+ gfc_type_letter (a->ts.type),
+ gfc_type_abi_kind (&a->ts));
}
@@ -1322,7 +1346,8 @@ gfc_resolve_int2 (gfc_expr *f, gfc_expr *a)
f->ts.kind = 2;
f->value.function.name
= gfc_get_string ("__int_%d_%c%d", f->ts.kind,
- gfc_type_letter (a->ts.type), a->ts.kind);
+ gfc_type_letter (a->ts.type),
+ gfc_type_abi_kind (&a->ts));
}
@@ -1333,7 +1358,8 @@ gfc_resolve_int8 (gfc_expr *f, gfc_expr *a)
f->ts.kind = 8;
f->value.function.name
= gfc_get_string ("__int_%d_%c%d", f->ts.kind,
- gfc_type_letter (a->ts.type), a->ts.kind);
+ gfc_type_letter (a->ts.type),
+ gfc_type_abi_kind (&a->ts));
}
@@ -1344,7 +1370,8 @@ gfc_resolve_long (gfc_expr *f, gfc_expr *a)
f->ts.kind = 4;
f->value.function.name
= gfc_get_string ("__int_%d_%c%d", f->ts.kind,
- gfc_type_letter (a->ts.type), a->ts.kind);
+ gfc_type_letter (a->ts.type),
+ gfc_type_abi_kind (&a->ts));
}
@@ -1498,7 +1525,8 @@ gfc_resolve_log (gfc_expr *f, gfc_expr *x)
{
f->ts = x->ts;
f->value.function.name
- = gfc_get_string ("__log_%c%d", gfc_type_letter (x->ts.type), x->ts.kind);
+ = gfc_get_string ("__log_%c%d", gfc_type_letter (x->ts.type),
+ gfc_type_abi_kind (&x->ts));
}
@@ -1508,7 +1536,7 @@ gfc_resolve_log10 (gfc_expr *f, gfc_expr *x)
f->ts = x->ts;
f->value.function.name
= gfc_get_string ("__log10_%c%d", gfc_type_letter (x->ts.type),
- x->ts.kind);
+ gfc_type_abi_kind (&x->ts));
}
@@ -1522,7 +1550,8 @@ gfc_resolve_logical (gfc_expr *f, gfc_expr *a, gfc_expr *kind)
f->value.function.name
= gfc_get_string ("__logical_%d_%c%d", f->ts.kind,
- gfc_type_letter (a->ts.type), a->ts.kind);
+ gfc_type_letter (a->ts.type),
+ gfc_type_abi_kind (&a->ts));
}
@@ -1579,7 +1608,7 @@ gfc_resolve_matmul (gfc_expr *f, gfc_expr *a, gfc_expr *b)
f->value.function.name
= gfc_get_string (PREFIX ("matmul_%c%d"), gfc_type_letter (f->ts.type),
- f->ts.kind);
+ gfc_type_abi_kind (&f->ts));
}
@@ -1605,7 +1634,8 @@ gfc_resolve_minmax (const char *name, gfc_expr *f, gfc_actual_arglist *args)
}
f->value.function.name
- = gfc_get_string (name, gfc_type_letter (f->ts.type), f->ts.kind);
+ = gfc_get_string (name, gfc_type_letter (f->ts.type),
+ gfc_type_abi_kind (&f->ts));
}
@@ -1689,7 +1719,8 @@ gfc_resolve_maxloc (gfc_expr *f, gfc_expr *array, gfc_expr *dim,
f->value.function.name
= gfc_get_string (PREFIX ("%s%d_%d_%c%d"), name, d_num, f->ts.kind,
- gfc_type_letter (array->ts.type), array->ts.kind);
+ gfc_type_letter (array->ts.type),
+ gfc_type_abi_kind (&array->ts));
if (kind)
fkind = mpz_get_si (kind->value.integer);
@@ -1806,7 +1837,8 @@ gfc_resolve_findloc (gfc_expr *f, gfc_expr *array, gfc_expr *value,
f->value.function.name
= gfc_get_string (PREFIX ("%s%d_%c%d"), name, d_num,
- gfc_type_letter (array->ts.type, true), array->ts.kind);
+ gfc_type_letter (array->ts.type, true),
+ gfc_type_abi_kind (&array->ts));
/* We only have a single library function, so we need to convert
here. If the function is resolved from within a convert
@@ -1868,11 +1900,13 @@ gfc_resolve_maxval (gfc_expr *f, gfc_expr *array, gfc_expr *dim,
if (array->ts.type != BT_CHARACTER)
f->value.function.name
= gfc_get_string (PREFIX ("%s_%c%d"), name,
- gfc_type_letter (array->ts.type), array->ts.kind);
+ gfc_type_letter (array->ts.type),
+ gfc_type_abi_kind (&array->ts));
else
f->value.function.name
= gfc_get_string (PREFIX ("%s%d_%c%d"), name, f->rank != 0,
- gfc_type_letter (array->ts.type), array->ts.kind);
+ gfc_type_letter (array->ts.type),
+ gfc_type_abi_kind (&array->ts));
}
@@ -1926,7 +1960,7 @@ gfc_resolve_merge (gfc_expr *f, gfc_expr *tsource,
f->ts = tsource->ts;
f->value.function.name
= gfc_get_string ("__merge_%c%d", gfc_type_letter (tsource->ts.type),
- tsource->ts.kind);
+ gfc_type_abi_kind (&tsource->ts));
}
@@ -2017,7 +2051,8 @@ gfc_resolve_minloc (gfc_expr *f, gfc_expr *array, gfc_expr *dim,
f->value.function.name
= gfc_get_string (PREFIX ("%s%d_%d_%c%d"), name, d_num, f->ts.kind,
- gfc_type_letter (array->ts.type), array->ts.kind);
+ gfc_type_letter (array->ts.type),
+ gfc_type_abi_kind (&array->ts));
if (fkind != f->ts.kind)
{
@@ -2082,11 +2117,13 @@ gfc_resolve_minval (gfc_expr *f, gfc_expr *array, gfc_expr *dim,
if (array->ts.type != BT_CHARACTER)
f->value.function.name
= gfc_get_string (PREFIX ("%s_%c%d"), name,
- gfc_type_letter (array->ts.type), array->ts.kind);
+ gfc_type_letter (array->ts.type),
+ gfc_type_abi_kind (&array->ts));
else
f->value.function.name
= gfc_get_string (PREFIX ("%s%d_%c%d"), name, f->rank != 0,
- gfc_type_letter (array->ts.type), array->ts.kind);
+ gfc_type_letter (array->ts.type),
+ gfc_type_abi_kind (&array->ts));
}
@@ -2108,7 +2145,8 @@ gfc_resolve_mod (gfc_expr *f, gfc_expr *a, gfc_expr *p)
}
f->value.function.name
- = gfc_get_string ("__mod_%c%d", gfc_type_letter (f->ts.type), f->ts.kind);
+ = gfc_get_string ("__mod_%c%d", gfc_type_letter (f->ts.type),
+ gfc_type_abi_kind (&f->ts));
}
@@ -2131,7 +2169,7 @@ gfc_resolve_modulo (gfc_expr *f, gfc_expr *a, gfc_expr *p)
f->value.function.name
= gfc_get_string ("__modulo_%c%d", gfc_type_letter (f->ts.type),
- f->ts.kind);
+ gfc_type_abi_kind (&f->ts));
}
void
@@ -2143,7 +2181,7 @@ gfc_resolve_nearest (gfc_expr *f, gfc_expr *a, gfc_expr *p)
f->ts = a->ts;
f->value.function.name
= gfc_get_string ("__nearest_%c%d", gfc_type_letter (a->ts.type),
- a->ts.kind);
+ gfc_type_abi_kind (&a->ts));
}
void
@@ -2187,7 +2225,8 @@ gfc_resolve_or (gfc_expr *f, gfc_expr *i, gfc_expr *j)
}
f->value.function.name
- = gfc_get_string ("__or_%c%d", gfc_type_letter (i->ts.type), f->ts.kind);
+ = gfc_get_string ("__or_%c%d", gfc_type_letter (i->ts.type),
+ gfc_type_abi_kind (&f->ts));
}
@@ -2265,7 +2304,8 @@ gfc_resolve_real (gfc_expr *f, gfc_expr *a, gfc_expr *kind)
f->value.function.name
= gfc_get_string ("__real_%d_%c%d", f->ts.kind,
- gfc_type_letter (a->ts.type), a->ts.kind);
+ gfc_type_letter (a->ts.type),
+ gfc_type_abi_kind (&a->ts));
}
@@ -2276,7 +2316,8 @@ gfc_resolve_realpart (gfc_expr *f, gfc_expr *a)
f->ts.kind = a->ts.kind;
f->value.function.name
= gfc_get_string ("__real_%d_%c%d", f->ts.kind,
- gfc_type_letter (a->ts.type), a->ts.kind);
+ gfc_type_letter (a->ts.type),
+ gfc_type_abi_kind (&a->ts));
}
@@ -2361,7 +2402,7 @@ gfc_resolve_reshape (gfc_expr *f, gfc_expr *source, gfc_expr *shape,
f->value.function.name
= gfc_get_string (PREFIX ("reshape_%c%d"),
gfc_type_letter (source->ts.type),
- source->ts.kind);
+ gfc_type_abi_kind (&source->ts));
else if (source->ts.type == BT_CHARACTER)
f->value.function.name = gfc_get_string (PREFIX ("reshape_char%d"),
kind);
@@ -2506,7 +2547,8 @@ gfc_resolve_sign (gfc_expr *f, gfc_expr *a, gfc_expr *b ATTRIBUTE_UNUSED)
{
f->ts = a->ts;
f->value.function.name
- = gfc_get_string ("__sign_%c%d", gfc_type_letter (a->ts.type), a->ts.kind);
+ = gfc_get_string ("__sign_%c%d", gfc_type_letter (a->ts.type),
+ gfc_type_abi_kind (&a->ts));
}
@@ -2536,7 +2578,8 @@ gfc_resolve_sin (gfc_expr *f, gfc_expr *x)
{
f->ts = x->ts;
f->value.function.name
- = gfc_get_string ("__sin_%c%d", gfc_type_letter (x->ts.type), x->ts.kind);
+ = gfc_get_string ("__sin_%c%d", gfc_type_letter (x->ts.type),
+ gfc_type_abi_kind (&x->ts));
}
@@ -2545,7 +2588,8 @@ gfc_resolve_sinh (gfc_expr *f, gfc_expr *x)
{
f->ts = x->ts;
f->value.function.name
- = gfc_get_string ("__sinh_%c%d", gfc_type_letter (x->ts.type), x->ts.kind);
+ = gfc_get_string ("__sinh_%c%d", gfc_type_letter (x->ts.type),
+ gfc_type_abi_kind (&x->ts));
}
@@ -2639,7 +2683,8 @@ gfc_resolve_sqrt (gfc_expr *f, gfc_expr *x)
{
f->ts = x->ts;
f->value.function.name
- = gfc_get_string ("__sqrt_%c%d", gfc_type_letter (x->ts.type), x->ts.kind);
+ = gfc_get_string ("__sqrt_%c%d", gfc_type_letter (x->ts.type),
+ gfc_type_abi_kind (&x->ts));
}
@@ -2803,7 +2848,8 @@ gfc_resolve_tan (gfc_expr *f, gfc_expr *x)
{
f->ts = x->ts;
f->value.function.name
- = gfc_get_string ("__tan_%c%d", gfc_type_letter (x->ts.type), x->ts.kind);
+ = gfc_get_string ("__tan_%c%d", gfc_type_letter (x->ts.type),
+ gfc_type_abi_kind (&x->ts));
}
@@ -2812,7 +2858,8 @@ gfc_resolve_tanh (gfc_expr *f, gfc_expr *x)
{
f->ts = x->ts;
f->value.function.name
- = gfc_get_string ("__tanh_%c%d", gfc_type_letter (x->ts.type), x->ts.kind);
+ = gfc_get_string ("__tanh_%c%d", gfc_type_letter (x->ts.type),
+ gfc_type_abi_kind (&x->ts));
}
@@ -3012,7 +3059,7 @@ gfc_resolve_transpose (gfc_expr *f, gfc_expr *matrix)
f->value.function.name
= gfc_get_string (PREFIX ("transpose_%c%d"),
gfc_type_letter (matrix->ts.type),
- matrix->ts.kind);
+ gfc_type_abi_kind (&matrix->ts));
break;
case BT_INTEGER:
@@ -3060,7 +3107,8 @@ gfc_resolve_trigd (gfc_expr *f, gfc_expr *x)
f->ts = x->ts;
f->value.function.name
= gfc_get_string (PREFIX ("%s_%c%d"), f->value.function.isym->name,
- gfc_type_letter (x->ts.type), x->ts.kind);
+ gfc_type_letter (x->ts.type),
+ gfc_type_abi_kind (&x->ts));
}
@@ -3188,7 +3236,8 @@ gfc_resolve_xor (gfc_expr *f, gfc_expr *i, gfc_expr *j)
}
f->value.function.name
- = gfc_get_string ("__xor_%c%d", gfc_type_letter (i->ts.type), f->ts.kind);
+ = gfc_get_string ("__xor_%c%d", gfc_type_letter (i->ts.type),
+ gfc_type_abi_kind (&f->ts));
}
@@ -3326,7 +3375,7 @@ gfc_resolve_random_number (gfc_code *c)
const char *name;
int kind;
- kind = c->ext.actual->expr->ts.kind;
+ kind = gfc_type_abi_kind (&c->ext.actual->expr->ts);
if (c->ext.actual->expr->rank == 0)
name = gfc_get_string (PREFIX ("random_r%d"), kind);
else
diff --git a/gcc/fortran/trans-decl.c b/gcc/fortran/trans-decl.c
index cb7f684d52c..ba34ec611f9 100644
--- a/gcc/fortran/trans-decl.c
+++ b/gcc/fortran/trans-decl.c
@@ -3602,8 +3602,9 @@ gfc_build_intrinsic_function_decls (void)
rtype = gfc_get_real_type (rkinds[rkind]);
if (rtype && itype)
{
- sprintf (name, PREFIX("pow_r%d_i%d"), rkinds[rkind],
- ikinds[ikind]);
+ sprintf (name, PREFIX("pow_r%d_i%d"),
+ gfc_type_abi_kind (BT_REAL, rkinds[rkind]),
+ ikinds[ikind]);
gfor_fndecl_math_powi[rkind][ikind].real =
gfc_build_library_function_decl (get_identifier (name),
rtype, 2, rtype, itype);
@@ -3614,8 +3615,9 @@ gfc_build_intrinsic_function_decls (void)
ctype = gfc_get_complex_type (rkinds[rkind]);
if (ctype && itype)
{
- sprintf (name, PREFIX("pow_c%d_i%d"), rkinds[rkind],
- ikinds[ikind]);
+ sprintf (name, PREFIX("pow_c%d_i%d"),
+ gfc_type_abi_kind (BT_REAL, rkinds[rkind]),
+ ikinds[ikind]);
gfor_fndecl_math_powi[rkind][ikind].cmplx =
gfc_build_library_function_decl (get_identifier (name),
ctype, 2,ctype, itype);
diff --git a/gcc/fortran/trans-types.c b/gcc/fortran/trans-types.c
index e5d36d5a58f..12b74207c99 100644
--- a/gcc/fortran/trans-types.c
+++ b/gcc/fortran/trans-types.c
@@ -363,6 +363,8 @@ gfc_init_kinds (void)
int i_index, r_index, kind;
bool saw_i4 = false, saw_i8 = false;
bool saw_r4 = false, saw_r8 = false, saw_r10 = false, saw_r16 = false;
+ scalar_mode r16_mode = QImode;
+ scalar_mode composite_mode = QImode;
i_index = 0;
FOR_EACH_MODE_IN_CLASS (int_mode_iter, MODE_INT)
@@ -428,6 +430,10 @@ gfc_init_kinds (void)
if (!targetm.scalar_mode_supported_p (mode))
continue;
+ if (MODE_COMPOSITE_P (mode)
+ && (GET_MODE_PRECISION (mode) + 7) / 8 == 16)
+ composite_mode = mode;
+
/* Only let float, double, long double and TFmode go through.
Runtime support for others is not provided, so they would be
useless. */
@@ -471,7 +477,10 @@ gfc_init_kinds (void)
if (kind == 10)
saw_r10 = true;
if (kind == 16)
- saw_r16 = true;
+ {
+ saw_r16 = true;
+ r16_mode = mode;
+ }
/* Careful we don't stumble a weird internal mode. */
gcc_assert (r_index <= 0 || gfc_real_kinds[r_index-1].kind != kind);
@@ -479,6 +488,7 @@ gfc_init_kinds (void)
gcc_assert (r_index != MAX_REAL_KINDS);
gfc_real_kinds[r_index].kind = kind;
+ gfc_real_kinds[r_index].abi_kind = kind;
gfc_real_kinds[r_index].radix = fmt->b;
gfc_real_kinds[r_index].digits = fmt->p;
gfc_real_kinds[r_index].min_exponent = fmt->emin;
@@ -496,6 +506,19 @@ gfc_init_kinds (void)
r_index += 1;
}
+ /* Detect the powerpc64le-linux case with -mabi=ieeelongdouble, where
+ the long double type is non-MODE_COMPOSITE_P TFmode but one can use
+ -mabi=ibmlongdouble too and get MODE_COMPOSITE_P TFmode with the same
+ precision. For libgfortran calls pretend the IEEE 754 quad TFmode has
+ kind 17 rather than 16 and use kind 16 for the IBM extended format
+ TFmode. */
+ if (composite_mode != QImode && saw_r16 && !MODE_COMPOSITE_P (r16_mode))
+ {
+ for (int i = 0; i < r_index; ++i)
+ if (gfc_real_kinds[i].kind == 16)
+ gfc_real_kinds[i].abi_kind = 17;
+ }
+
/* Choose the default integer kind. We choose 4 unless the user directs us
otherwise. Even if the user specified that the default integer kind is 8,
the numeric storage size is not 64 bits. In this case, a warning will be