21d1335b2a
2012-03-10 Tobias Burnus <burnus@net-b.de> * libgfortran.h (descriptor_dimension, GFC_DIMENSION_LBOUND, GFC_DIMENSION_EXTENT, GFC_DIMENSION_SET, GFC_DESCRIPTOR_DATA, GFC_ARRAY_DESCRIPTOR): Rename _lbound to lower_bound and data to base_addr. * intrinsics/cshift0.c (size0, cshift0): Handle data to base_addr rename. * intrinsics/date_and_time.c (date_and_time, secnds, itime_i4, itime_i8, gmtime_i4, gmtime_i8, ltime_i4, ltime_i8): Ditto. * intrinsics/dtime.c (dtime_sub): Ditto. * intrinsics/eoshift0.c (eoshift0): Ditto. * intrinsics/eoshift2.c (eoshift2): Ditto. * intrinsics/etime.c (etime_sub): Ditto. * intrinsics/iso_c_binding.c (c_f_pointer_u0): Ditto. * intrinsics/move_alloc.c (move_alloc): Ditto. * intrinsics/pack_generic.c (pack_internal, pack, pack_s_internal): Ditto. * intrinsics/random.c (arandom_r4, arandom_r8, arandom_r10, arandom_r16, random_seed_i4, random_seed_i8): Ditto. * intrinsics/reshape_generic.c (reshape_internal): Ditto. * intrinsics/spread_generic.c (spread_internal, spread_internal_scalar, spread, spread_scalar): Ditto. * intrinsics/stat.c (stat_i4_sub_0, stat_i8_sub_0, fstat_i4_sub, fstat_i8_sub): Ditto. * intrinsics/transpose_generic.c (transpose_internal): Ditto. * intrinsics/unpack_generic.c (unpack_bounds, unpack_internal, unpack1, unpack0, unpack0_char4): Ditto. * m4/bessel.m4 (bessel_jn_r'rtype_kind`, * bessel_yn_r'rtype_kind`): Ditto. * m4/cshift0.m4 (cshift0_'rtype_code`): Ditto. * m4/cshift1.m4 (cshift1): Ditto. * m4/eoshift1.m4 (eoshift1): Ditto. * m4/eoshift3.m4 (eoshift3): Ditto. * m4/iforeach.m4 (name`'rtype_qual`_'atype_code): Ditto. * m4/ifunction.m4 (name`'rtype_qual`_'atype_code): Ditto. * m4/ifunction_logical.m4 (name`'rtype_qual`_'atype_code): * Ditto. * m4/in_pack.m4 (internal_pack_'rtype_ccode`): Ditto. * m4/in_unpack.m4 (internal_unpack_'rtype_ccode`): Ditto. * m4/matmul.m4 (matmul_'rtype_code`): Ditto. * m4/matmull.m4 (matmul_'rtype_code`): Ditto. * m4/pack.m4 (pack_'rtype_code`): Ditto. * m4/reshape.m4 (reshape_'rtype_ccode`): Ditto. * m4/shape.m4 (shape_'rtype_kind`): Ditto. * m4/spread.m4 (spread_'rtype_code`): Ditto. * m4/transpose.m4 (transpose_'rtype_code`): Ditto. * m4/unpack.m4 (unpack0_'rtype_code`, unpack1_'rtype_code`): * Ditto. * runtime/bounds.c (count_0): Ditto. * runtime/in_pack_generic.c (internal_pack): Ditto. * runtime/in_unpack_generic.c (internal_unpack): Ditto. * generated/cshift0_r4.c: Regenerated. * generated/unpack_r8.c: Regenerated. * generated/cshift0_c10.c: Regenerated. * generated/in_unpack_i1.c: Regenerated. * generated/eoshift1_16.c: Regenerated. * generated/iany_i16.c: Regenerated. * generated/reshape_i4.c: Regenerated. * generated/parity_l16.c: Regenerated. * generated/in_unpack_r16.c: Regenerated. * generated/iparity_i1.c: Regenerated. * generated/minloc0_16_r10.c: Regenerated. * generated/minloc1_16_i8.c: Regenerated. * generated/bessel_r8.c: Regenerated. * generated/maxloc0_4_i8.c: Regenerated. * generated/maxloc1_8_i1.c: Regenerated. * generated/sum_c10.c: Regenerated. * generated/maxloc0_8_r10.c: Regenerated. * generated/unpack_r16.c: Regenerated. * generated/maxloc1_16_i1.c: Regenerated. * generated/minloc1_4_i1.c: Regenerated. * generated/unpack_i4.c: Regenerated. * generated/in_pack_i1.c: Regenerated. * generated/cshift0_i1.c: Regenerated. * generated/minloc0_16_i4.c: Regenerated. * generated/minloc1_8_r8.c: Regenerated. * generated/maxval_r8.c: Regenerated. * generated/unpack_i2.c: Regenerated. * generated/matmul_r16.c: Regenerated. * generated/maxloc0_16_r8.c: Regenerated. * generated/maxloc1_8_i16.c: Regenerated. * generated/maxloc0_4_r4.c: Regenerated. * generated/pack_c10.c: Regenerated. * generated/in_unpack_c8.c: Regenerated. * generated/maxloc1_8_i8.c: Regenerated. * generated/in_pack_c16.c: Regenerated. * generated/parity_l1.c: Regenerated. * generated/in_pack_i2.c: Regenerated. * generated/maxloc0_4_r8.c: Regenerated. * generated/pack_r4.c: Regenerated. * generated/matmul_r8.c: Regenerated. * generated/pack_i1.c: Regenerated. * generated/minloc0_16_r16.c: Regenerated. * generated/cshift0_c8.c: Regenerated. * generated/minloc1_4_i16.c: Regenerated. * generated/minloc0_4_r4.c: Regenerated. * generated/product_i2.c: Regenerated. * generated/maxval_r10.c: Regenerated. * generated/minloc0_16_r8.c: Regenerated. * generated/in_pack_i8.c: Regenerated. * generated/all_l8.c: Regenerated. * generated/minloc0_8_i16.c: Regenerated. * generated/minloc1_8_i2.c: Regenerated. * generated/matmul_l8.c: Regenerated. * generated/spread_r10.c: Regenerated. * generated/product_c4.c: Regenerated. * generated/reshape_r16.c: Regenerated. * generated/cshift0_i16.c: Regenerated. * generated/minloc1_8_r10.c: Regenerated. * generated/iparity_i8.c: Regenerated. * generated/minloc0_16_i8.c: Regenerated. * generated/minloc1_8_i1.c: Regenerated. * generated/maxloc0_16_r10.c: Regenerated. * generated/maxloc1_8_i4.c: Regenerated. * generated/maxloc0_16_i8.c: Regenerated. * generated/maxloc1_16_i16.c: Regenerated. * generated/reshape_i8.c: Regenerated. * generated/all_l16.c: Regenerated. * generated/cshift1_8.c: Regenerated. * generated/matmul_c10.c: Regenerated. * generated/pack_r16.c: Regenerated. * generated/transpose_r10.c: Regenerated. * generated/maxloc1_16_i2.c: Regenerated. * generated/minloc1_4_r4.c: Regenerated. * generated/reshape_r8.c: Regenerated. * generated/product_r10.c: Regenerated. * generated/spread_c10.c: Regenerated. * generated/reshape_r4.c: Regenerated. * generated/in_unpack_i4.c: Regenerated. * generated/minval_r4.c: Regenerated. * generated/minloc1_16_r16.c: Regenerated. * generated/norm2_r8.c: Regenerated. * generated/iany_i2.c: Regenerated. * generated/eoshift3_8.c: Regenerated. * generated/maxloc1_16_r8.c: Regenerated. * generated/maxloc0_16_r16.c: Regenerated. * generated/cshift0_i8.c: Regenerated. * generated/all_l1.c: Regenerated. * generated/minloc1_16_r10.c: Regenerated. * generated/spread_i16.c: Regenerated. * generated/minloc1_8_r16.c: Regenerated. * generated/spread_r16.c: Regenerated. * generated/matmul_c8.c: Regenerated. * generated/pack_r10.c: Regenerated. * generated/reshape_r10.c: Regenerated. * generated/sum_r10.c: Regenerated. * generated/count_8_l.c: Regenerated. * generated/minloc1_16_i16.c: Regenerated. * generated/maxloc0_16_i2.c: Regenerated. * generated/count_1_l.c: Regenerated. * generated/unpack_c16.c: Regenerated. * generated/any_l8.c: Regenerated. * generated/minloc1_16_i1.c: Regenerated. * generated/sum_r4.c: Regenerated. * generated/in_unpack_r8.c: Regenerated. * generated/product_r8.c: Regenerated. * generated/in_unpack_r4.c: Regenerated. * generated/unpack_i1.c: Regenerated. * generated/cshift0_r8.c: Regenerated. * generated/spread_i2.c: Regenerated. * generated/minloc1_4_r10.c: Regenerated. * generated/minloc1_16_i2.c: Regenerated. * generated/minloc0_16_i2.c: Regenerated. * generated/eoshift3_16.c: Regenerated. * generated/maxloc0_8_i4.c: Regenerated. * generated/maxloc1_16_r10.c: Regenerated. * generated/maxloc1_4_i16.c: Regenerated. * generated/maxloc1_4_r8.c: Regenerated. * generated/parity_l2.c: Regenerated. * generated/maxloc0_16_i16.c: Regenerated. * generated/product_r4.c: Regenerated. * generated/in_unpack_r10.c: Regenerated. * generated/maxloc1_8_i2.c: Regenerated. * generated/iparity_i2.c: Regenerated. * generated/maxloc1_8_r4.c: Regenerated. * generated/maxloc1_8_r8.c: Regenerated. * generated/iall_i4.c: Regenerated. * generated/maxloc0_4_i16.c: Regenerated. * generated/cshift0_i4.c: Regenerated. * generated/in_unpack_c10.c: Regenerated. * generated/unpack_c8.c: Regenerated. * generated/transpose_c8.c: Regenerated. * generated/in_unpack_c16.c: Regenerated. * generated/product_c16.c: Regenerated. * generated/minval_r8.c: Regenerated. * generated/minloc0_4_i1.c: Regenerated. * generated/minloc0_8_r16.c: Regenerated. * generated/in_unpack_i16.c: Regenerated. * generated/reshape_c16.c: Regenerated. * generated/pack_i4.c: Regenerated. * generated/minval_r10.c: Regenerated. * generated/maxloc0_4_r10.c: Regenerated. * generated/unpack_i8.c: Regenerated. * generated/minloc1_16_r8.c: Regenerated. * generated/pack_c16.c: Regenerated. * generated/maxval_r4.c: Regenerated. * generated/spread_r8.c: Regenerated. * generated/minloc0_16_r4.c: Regenerated. * generated/maxval_i4.c: Regenerated. * generated/matmul_r4.c: Regenerated. * generated/product_c8.c: Regenerated. * generated/maxloc1_4_i4.c: Regenerated. * generated/any_l4.c: Regenerated. * generated/cshift0_r16.c: Regenerated. * generated/maxloc1_16_r4.c: Regenerated. * generated/in_pack_r10.c: Regenerated. * generated/iall_i1.c: Regenerated. * generated/product_i1.c: Regenerated. * generated/maxloc0_4_i1.c: Regenerated. * generated/maxloc0_8_r16.c: Regenerated. * generated/minloc1_8_i16.c: Regenerated. * generated/reshape_c10.c: Regenerated. * generated/matmul_l4.c: Regenerated. * generated/cshift0_c16.c: Regenerated. * generated/spread_r4.c: Regenerated. * generated/sum_r16.c: Regenerated. * generated/in_unpack_c4.c: Regenerated. * generated/maxval_i8.c: Regenerated. * generated/maxloc0_8_r8.c: Regenerated. * generated/unpack_i16.c: Regenerated. * generated/norm2_r4.c: Regenerated. * generated/maxloc1_4_r16.c: Regenerated. * generated/in_pack_r4.c: Regenerated. * generated/sum_c8.c: Regenerated. * generated/parity_l8.c: Regenerated. * generated/minval_i2.c: Regenerated. * generated/transpose_c10.c: Regenerated. * generated/cshift0_r10.c: Regenerated. * generated/cshift0_i2.c: Regenerated. * generated/spread_c4.c: Regenerated. * generated/maxval_r16.c: Regenerated. * generated/unpack_c10.c: Regenerated. * generated/bessel_r4.c: Regenerated. * generated/minloc0_8_i1.c: Regenerated. * generated/product_c10.c: Regenerated. * generated/minloc1_4_i4.c: Regenerated. * generated/minloc1_4_i2.c: Regenerated. * generated/transpose_c4.c: Regenerated. * generated/in_pack_c8.c: Regenerated. * generated/minval_i8.c: Regenerated. * generated/maxloc1_4_i2.c: Regenerated. * generated/maxloc1_4_i1.c: Regenerated. * generated/iall_i8.c: Regenerated. * generated/iany_i8.c: Regenerated. * generated/sum_c16.c: Regenerated. * generated/spread_i8.c: Regenerated. * generated/sum_i8.c: Regenerated. * generated/pack_i2.c: Regenerated. * generated/unpack_r10.c: Regenerated. * generated/minloc0_4_i4.c: Regenerated. * generated/iany_i4.c: Regenerated. * generated/maxloc0_4_i4.c: Regenerated. * generated/in_pack_r8.c: Regenerated. * generated/spread_i1.c: Regenerated. * generated/maxval_i16.c: Regenerated. * generated/pack_c8.c: Regenerated. * generated/iparity_i16.c: Regenerated. * generated/matmul_c16.c: Regenerated. * generated/eoshift1_8.c: Regenerated. * generated/any_l16.c: Regenerated. * generated/maxloc1_4_i8.c: Regenerated. * generated/minloc0_8_r4.c: Regenerated. * generated/transpose_c16.c: Regenerated. * generated/reshape_c4.c: Regenerated. * generated/minloc0_8_r8.c: Regenerated. * generated/count_16_l.c: Regenerated. * generated/minloc0_16_i1.c: Regenerated. * generated/minloc1_16_r4.c: Regenerated. * generated/maxloc0_8_i1.c: Regenerated. * generated/matmul_i1.c: Regenerated. * generated/sum_i2.c: Regenerated. * generated/cshift1_4.c: Regenerated. * generated/minloc0_4_i2.c: Regenerated. * generated/bessel_r10.c: Regenerated. * generated/matmul_c4.c: Regenerated. * generated/parity_l4.c: Regenerated. * generated/in_unpack_i2.c: Regenerated. * generated/product_i16.c: Regenerated. * generated/maxloc0_16_i4.c: Regenerated. * generated/minloc0_4_r10.c: Regenerated. * generated/transpose_i16.c: Regenerated. * generated/matmul_l16.c: Regenerated. * generated/any_l1.c: Regenerated. * generated/minval_r16.c: Regenerated. * generated/shape_i16.c: Regenerated. * generated/in_pack_c4.c: Regenerated. * generated/minloc1_8_r4.c: Regenerated. * generated/iany_i1.c: Regenerated. * generated/matmul_i16.c: Regenerated. * generated/maxloc0_16_r4.c: Regenerated. * generated/spread_c8.c: Regenerated. * generated/product_i8.c: Regenerated. * generated/minloc1_8_i4.c: Regenerated. * generated/product_i4.c: Regenerated. * generated/minval_i16.c: Regenerated. * generated/pack_r8.c: Regenerated. * generated/matmul_r10.c: Regenerated. * generated/shape_i8.c: Regenerated. * generated/eoshift1_4.c: Regenerated. * generated/minloc1_16_i4.c: Regenerated. * generated/transpose_r16.c: Regenerated. * generated/spread_c16.c: Regenerated. * generated/maxloc0_8_i16.c: Regenerated. * generated/minval_i1.c: Regenerated. * generated/sum_i4.c: Regenerated. * generated/in_pack_r16.c: Regenerated. * generated/iall_i16.c: Regenerated. * generated/minloc0_8_r10.c: Regenerated. * generated/maxloc0_8_i8.c: Regenerated. * generated/transpose_i4.c: Regenerated. * generated/maxloc0_4_i2.c: Regenerated. * generated/iall_i2.c: Regenerated. * generated/maxloc0_8_r4.c: Regenerated. * generated/minloc0_4_r16.c: Regenerated. * generated/minloc1_4_r8.c: Regenerated. * generated/minloc0_4_i8.c: Regenerated. * generated/minval_i4.c: Regenerated. * generated/minloc1_8_i8.c: Regenerated. * generated/in_pack_i16.c: Regenerated. * generated/spread_i4.c: Regenerated. * generated/cshift0_c4.c: Regenerated. * generated/minloc0_4_i16.c: Regenerated. * generated/sum_r8.c: Regenerated. * generated/minloc0_16_i16.c: Regenerated. * generated/sum_i1.c: Regenerated. * generated/all_l2.c: Regenerated. * generated/sum_c4.c: Regenerated. * generated/shape_i4.c: Regenerated. * generated/maxval_i1.c: Regenerated. * generated/norm2_r10.c: Regenerated. * generated/unpack_c4.c: Regenerated. * generated/maxloc0_16_i1.c: Regenerated. * generated/all_l4.c: Regenerated. * generated/maxloc1_16_i8.c: Regenerated. * generated/sum_i16.c: Regenerated. * generated/maxloc1_8_r16.c: Regenerated. * generated/minloc1_4_r16.c: Regenerated. * generated/maxloc1_16_r16.c: Regenerated. * generated/transpose_i8.c: Regenerated. * generated/minloc0_8_i2.c: Regenerated. * generated/product_r16.c: Regenerated. * generated/in_pack_c10.c: Regenerated. * generated/maxloc0_4_r16.c: Regenerated. * generated/pack_c4.c: Regenerated. * generated/bessel_r16.c: Regenerated. * generated/norm2_r16.c: Regenerated. * generated/in_unpack_i8.c: Regenerated. * generated/maxloc1_16_i4.c: Regenerated. * generated/minloc1_4_i8.c: Regenerated. * generated/eoshift3_4.c: Regenerated. * generated/maxloc1_8_r10.c: Regenerated. * generated/matmul_i8.c: Regenerated. * generated/matmul_i4.c: Regenerated. * generated/reshape_i16.c: Regenerated. * generated/matmul_i2.c: Regenerated. * generated/transpose_r8.c: Regenerated. * generated/transpose_r4.c: Regenerated. * generated/cshift1_16.c: Regenerated. * generated/maxloc1_4_r10.c: Regenerated. * generated/iparity_i4.c: Regenerated. * generated/in_pack_i4.c: Regenerated. * generated/maxloc0_8_i2.c: Regenerated. * generated/unpack_r4.c: Regenerated. * generated/count_4_l.c: Regenerated. * generated/maxloc1_4_r4.c: Regenerated. * generated/minloc0_8_i8.c: Regenerated. * generated/any_l2.c: Regenerated. * generated/minloc0_8_i4.c: Regenerated. * generated/pack_i8.c: Regenerated. * generated/count_2_l.c: Regenerated. * generated/reshape_c8.c: Regenerated. * generated/maxval_i2.c: Regenerated. * generated/minloc0_4_r8.c: Regenerated. From-SVN: r185177
509 lines
12 KiB
C
509 lines
12 KiB
C
/* Implementation of the SUM intrinsic
|
|
Copyright 2002, 2007, 2009 Free Software Foundation, Inc.
|
|
Contributed by Paul Brook <paul@nowt.org>
|
|
|
|
This file is part of the GNU Fortran 95 runtime library (libgfortran).
|
|
|
|
Libgfortran is free software; you can redistribute it and/or
|
|
modify it under the terms of the GNU General Public
|
|
License as published by the Free Software Foundation; either
|
|
version 3 of the License, or (at your option) any later version.
|
|
|
|
Libgfortran is distributed in the hope that it will be useful,
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
GNU General Public License for more details.
|
|
|
|
Under Section 7 of GPL version 3, you are granted additional
|
|
permissions described in the GCC Runtime Library Exception, version
|
|
3.1, as published by the Free Software Foundation.
|
|
|
|
You should have received a copy of the GNU General Public License and
|
|
a copy of the GCC Runtime Library Exception along with this program;
|
|
see the files COPYING3 and COPYING.RUNTIME respectively. If not, see
|
|
<http://www.gnu.org/licenses/>. */
|
|
|
|
#include "libgfortran.h"
|
|
#include <stdlib.h>
|
|
#include <assert.h>
|
|
|
|
|
|
#if defined (HAVE_GFC_REAL_4) && defined (HAVE_GFC_REAL_4)
|
|
|
|
|
|
extern void sum_r4 (gfc_array_r4 * const restrict,
|
|
gfc_array_r4 * const restrict, const index_type * const restrict);
|
|
export_proto(sum_r4);
|
|
|
|
void
|
|
sum_r4 (gfc_array_r4 * const restrict retarray,
|
|
gfc_array_r4 * const restrict array,
|
|
const index_type * const restrict pdim)
|
|
{
|
|
index_type count[GFC_MAX_DIMENSIONS];
|
|
index_type extent[GFC_MAX_DIMENSIONS];
|
|
index_type sstride[GFC_MAX_DIMENSIONS];
|
|
index_type dstride[GFC_MAX_DIMENSIONS];
|
|
const GFC_REAL_4 * restrict base;
|
|
GFC_REAL_4 * restrict dest;
|
|
index_type rank;
|
|
index_type n;
|
|
index_type len;
|
|
index_type delta;
|
|
index_type dim;
|
|
int continue_loop;
|
|
|
|
/* Make dim zero based to avoid confusion. */
|
|
dim = (*pdim) - 1;
|
|
rank = GFC_DESCRIPTOR_RANK (array) - 1;
|
|
|
|
len = GFC_DESCRIPTOR_EXTENT(array,dim);
|
|
if (len < 0)
|
|
len = 0;
|
|
delta = GFC_DESCRIPTOR_STRIDE(array,dim);
|
|
|
|
for (n = 0; n < dim; n++)
|
|
{
|
|
sstride[n] = GFC_DESCRIPTOR_STRIDE(array,n);
|
|
extent[n] = GFC_DESCRIPTOR_EXTENT(array,n);
|
|
|
|
if (extent[n] < 0)
|
|
extent[n] = 0;
|
|
}
|
|
for (n = dim; n < rank; n++)
|
|
{
|
|
sstride[n] = GFC_DESCRIPTOR_STRIDE(array, n + 1);
|
|
extent[n] = GFC_DESCRIPTOR_EXTENT(array, n + 1);
|
|
|
|
if (extent[n] < 0)
|
|
extent[n] = 0;
|
|
}
|
|
|
|
if (retarray->base_addr == NULL)
|
|
{
|
|
size_t alloc_size, str;
|
|
|
|
for (n = 0; n < rank; n++)
|
|
{
|
|
if (n == 0)
|
|
str = 1;
|
|
else
|
|
str = GFC_DESCRIPTOR_STRIDE(retarray,n-1) * extent[n-1];
|
|
|
|
GFC_DIMENSION_SET(retarray->dim[n], 0, extent[n] - 1, str);
|
|
|
|
}
|
|
|
|
retarray->offset = 0;
|
|
retarray->dtype = (array->dtype & ~GFC_DTYPE_RANK_MASK) | rank;
|
|
|
|
alloc_size = sizeof (GFC_REAL_4) * GFC_DESCRIPTOR_STRIDE(retarray,rank-1)
|
|
* extent[rank-1];
|
|
|
|
retarray->base_addr = internal_malloc_size (alloc_size);
|
|
if (alloc_size == 0)
|
|
{
|
|
/* Make sure we have a zero-sized array. */
|
|
GFC_DIMENSION_SET(retarray->dim[0], 0, -1, 1);
|
|
return;
|
|
|
|
}
|
|
}
|
|
else
|
|
{
|
|
if (rank != GFC_DESCRIPTOR_RANK (retarray))
|
|
runtime_error ("rank of return array incorrect in"
|
|
" SUM intrinsic: is %ld, should be %ld",
|
|
(long int) (GFC_DESCRIPTOR_RANK (retarray)),
|
|
(long int) rank);
|
|
|
|
if (unlikely (compile_options.bounds_check))
|
|
bounds_ifunction_return ((array_t *) retarray, extent,
|
|
"return value", "SUM");
|
|
}
|
|
|
|
for (n = 0; n < rank; n++)
|
|
{
|
|
count[n] = 0;
|
|
dstride[n] = GFC_DESCRIPTOR_STRIDE(retarray,n);
|
|
if (extent[n] <= 0)
|
|
return;
|
|
}
|
|
|
|
base = array->base_addr;
|
|
dest = retarray->base_addr;
|
|
|
|
continue_loop = 1;
|
|
while (continue_loop)
|
|
{
|
|
const GFC_REAL_4 * restrict src;
|
|
GFC_REAL_4 result;
|
|
src = base;
|
|
{
|
|
|
|
result = 0;
|
|
if (len <= 0)
|
|
*dest = 0;
|
|
else
|
|
{
|
|
for (n = 0; n < len; n++, src += delta)
|
|
{
|
|
|
|
result += *src;
|
|
}
|
|
|
|
*dest = result;
|
|
}
|
|
}
|
|
/* Advance to the next element. */
|
|
count[0]++;
|
|
base += sstride[0];
|
|
dest += dstride[0];
|
|
n = 0;
|
|
while (count[n] == extent[n])
|
|
{
|
|
/* When we get to the end of a dimension, reset it and increment
|
|
the next dimension. */
|
|
count[n] = 0;
|
|
/* We could precalculate these products, but this is a less
|
|
frequently used path so probably not worth it. */
|
|
base -= sstride[n] * extent[n];
|
|
dest -= dstride[n] * extent[n];
|
|
n++;
|
|
if (n == rank)
|
|
{
|
|
/* Break out of the look. */
|
|
continue_loop = 0;
|
|
break;
|
|
}
|
|
else
|
|
{
|
|
count[n]++;
|
|
base += sstride[n];
|
|
dest += dstride[n];
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
|
|
extern void msum_r4 (gfc_array_r4 * const restrict,
|
|
gfc_array_r4 * const restrict, const index_type * const restrict,
|
|
gfc_array_l1 * const restrict);
|
|
export_proto(msum_r4);
|
|
|
|
void
|
|
msum_r4 (gfc_array_r4 * const restrict retarray,
|
|
gfc_array_r4 * const restrict array,
|
|
const index_type * const restrict pdim,
|
|
gfc_array_l1 * const restrict mask)
|
|
{
|
|
index_type count[GFC_MAX_DIMENSIONS];
|
|
index_type extent[GFC_MAX_DIMENSIONS];
|
|
index_type sstride[GFC_MAX_DIMENSIONS];
|
|
index_type dstride[GFC_MAX_DIMENSIONS];
|
|
index_type mstride[GFC_MAX_DIMENSIONS];
|
|
GFC_REAL_4 * restrict dest;
|
|
const GFC_REAL_4 * restrict base;
|
|
const GFC_LOGICAL_1 * restrict mbase;
|
|
int rank;
|
|
int dim;
|
|
index_type n;
|
|
index_type len;
|
|
index_type delta;
|
|
index_type mdelta;
|
|
int mask_kind;
|
|
|
|
dim = (*pdim) - 1;
|
|
rank = GFC_DESCRIPTOR_RANK (array) - 1;
|
|
|
|
len = GFC_DESCRIPTOR_EXTENT(array,dim);
|
|
if (len <= 0)
|
|
return;
|
|
|
|
mbase = mask->base_addr;
|
|
|
|
mask_kind = GFC_DESCRIPTOR_SIZE (mask);
|
|
|
|
if (mask_kind == 1 || mask_kind == 2 || mask_kind == 4 || mask_kind == 8
|
|
#ifdef HAVE_GFC_LOGICAL_16
|
|
|| mask_kind == 16
|
|
#endif
|
|
)
|
|
mbase = GFOR_POINTER_TO_L1 (mbase, mask_kind);
|
|
else
|
|
runtime_error ("Funny sized logical array");
|
|
|
|
delta = GFC_DESCRIPTOR_STRIDE(array,dim);
|
|
mdelta = GFC_DESCRIPTOR_STRIDE_BYTES(mask,dim);
|
|
|
|
for (n = 0; n < dim; n++)
|
|
{
|
|
sstride[n] = GFC_DESCRIPTOR_STRIDE(array,n);
|
|
mstride[n] = GFC_DESCRIPTOR_STRIDE_BYTES(mask,n);
|
|
extent[n] = GFC_DESCRIPTOR_EXTENT(array,n);
|
|
|
|
if (extent[n] < 0)
|
|
extent[n] = 0;
|
|
|
|
}
|
|
for (n = dim; n < rank; n++)
|
|
{
|
|
sstride[n] = GFC_DESCRIPTOR_STRIDE(array,n + 1);
|
|
mstride[n] = GFC_DESCRIPTOR_STRIDE_BYTES(mask, n + 1);
|
|
extent[n] = GFC_DESCRIPTOR_EXTENT(array, n + 1);
|
|
|
|
if (extent[n] < 0)
|
|
extent[n] = 0;
|
|
}
|
|
|
|
if (retarray->base_addr == NULL)
|
|
{
|
|
size_t alloc_size, str;
|
|
|
|
for (n = 0; n < rank; n++)
|
|
{
|
|
if (n == 0)
|
|
str = 1;
|
|
else
|
|
str= GFC_DESCRIPTOR_STRIDE(retarray,n-1) * extent[n-1];
|
|
|
|
GFC_DIMENSION_SET(retarray->dim[n], 0, extent[n] - 1, str);
|
|
|
|
}
|
|
|
|
alloc_size = sizeof (GFC_REAL_4) * GFC_DESCRIPTOR_STRIDE(retarray,rank-1)
|
|
* extent[rank-1];
|
|
|
|
retarray->offset = 0;
|
|
retarray->dtype = (array->dtype & ~GFC_DTYPE_RANK_MASK) | rank;
|
|
|
|
if (alloc_size == 0)
|
|
{
|
|
/* Make sure we have a zero-sized array. */
|
|
GFC_DIMENSION_SET(retarray->dim[0], 0, -1, 1);
|
|
return;
|
|
}
|
|
else
|
|
retarray->base_addr = internal_malloc_size (alloc_size);
|
|
|
|
}
|
|
else
|
|
{
|
|
if (rank != GFC_DESCRIPTOR_RANK (retarray))
|
|
runtime_error ("rank of return array incorrect in SUM intrinsic");
|
|
|
|
if (unlikely (compile_options.bounds_check))
|
|
{
|
|
bounds_ifunction_return ((array_t *) retarray, extent,
|
|
"return value", "SUM");
|
|
bounds_equal_extents ((array_t *) mask, (array_t *) array,
|
|
"MASK argument", "SUM");
|
|
}
|
|
}
|
|
|
|
for (n = 0; n < rank; n++)
|
|
{
|
|
count[n] = 0;
|
|
dstride[n] = GFC_DESCRIPTOR_STRIDE(retarray,n);
|
|
if (extent[n] <= 0)
|
|
return;
|
|
}
|
|
|
|
dest = retarray->base_addr;
|
|
base = array->base_addr;
|
|
|
|
while (base)
|
|
{
|
|
const GFC_REAL_4 * restrict src;
|
|
const GFC_LOGICAL_1 * restrict msrc;
|
|
GFC_REAL_4 result;
|
|
src = base;
|
|
msrc = mbase;
|
|
{
|
|
|
|
result = 0;
|
|
if (len <= 0)
|
|
*dest = 0;
|
|
else
|
|
{
|
|
for (n = 0; n < len; n++, src += delta, msrc += mdelta)
|
|
{
|
|
|
|
if (*msrc)
|
|
result += *src;
|
|
}
|
|
*dest = result;
|
|
}
|
|
}
|
|
/* Advance to the next element. */
|
|
count[0]++;
|
|
base += sstride[0];
|
|
mbase += mstride[0];
|
|
dest += dstride[0];
|
|
n = 0;
|
|
while (count[n] == extent[n])
|
|
{
|
|
/* When we get to the end of a dimension, reset it and increment
|
|
the next dimension. */
|
|
count[n] = 0;
|
|
/* We could precalculate these products, but this is a less
|
|
frequently used path so probably not worth it. */
|
|
base -= sstride[n] * extent[n];
|
|
mbase -= mstride[n] * extent[n];
|
|
dest -= dstride[n] * extent[n];
|
|
n++;
|
|
if (n == rank)
|
|
{
|
|
/* Break out of the look. */
|
|
base = NULL;
|
|
break;
|
|
}
|
|
else
|
|
{
|
|
count[n]++;
|
|
base += sstride[n];
|
|
mbase += mstride[n];
|
|
dest += dstride[n];
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
|
|
extern void ssum_r4 (gfc_array_r4 * const restrict,
|
|
gfc_array_r4 * const restrict, const index_type * const restrict,
|
|
GFC_LOGICAL_4 *);
|
|
export_proto(ssum_r4);
|
|
|
|
void
|
|
ssum_r4 (gfc_array_r4 * const restrict retarray,
|
|
gfc_array_r4 * const restrict array,
|
|
const index_type * const restrict pdim,
|
|
GFC_LOGICAL_4 * mask)
|
|
{
|
|
index_type count[GFC_MAX_DIMENSIONS];
|
|
index_type extent[GFC_MAX_DIMENSIONS];
|
|
index_type dstride[GFC_MAX_DIMENSIONS];
|
|
GFC_REAL_4 * restrict dest;
|
|
index_type rank;
|
|
index_type n;
|
|
index_type dim;
|
|
|
|
|
|
if (*mask)
|
|
{
|
|
sum_r4 (retarray, array, pdim);
|
|
return;
|
|
}
|
|
/* Make dim zero based to avoid confusion. */
|
|
dim = (*pdim) - 1;
|
|
rank = GFC_DESCRIPTOR_RANK (array) - 1;
|
|
|
|
for (n = 0; n < dim; n++)
|
|
{
|
|
extent[n] = GFC_DESCRIPTOR_EXTENT(array,n);
|
|
|
|
if (extent[n] <= 0)
|
|
extent[n] = 0;
|
|
}
|
|
|
|
for (n = dim; n < rank; n++)
|
|
{
|
|
extent[n] =
|
|
GFC_DESCRIPTOR_EXTENT(array,n + 1);
|
|
|
|
if (extent[n] <= 0)
|
|
extent[n] = 0;
|
|
}
|
|
|
|
if (retarray->base_addr == NULL)
|
|
{
|
|
size_t alloc_size, str;
|
|
|
|
for (n = 0; n < rank; n++)
|
|
{
|
|
if (n == 0)
|
|
str = 1;
|
|
else
|
|
str = GFC_DESCRIPTOR_STRIDE(retarray,n-1) * extent[n-1];
|
|
|
|
GFC_DIMENSION_SET(retarray->dim[n], 0, extent[n] - 1, str);
|
|
|
|
}
|
|
|
|
retarray->offset = 0;
|
|
retarray->dtype = (array->dtype & ~GFC_DTYPE_RANK_MASK) | rank;
|
|
|
|
alloc_size = sizeof (GFC_REAL_4) * GFC_DESCRIPTOR_STRIDE(retarray,rank-1)
|
|
* extent[rank-1];
|
|
|
|
if (alloc_size == 0)
|
|
{
|
|
/* Make sure we have a zero-sized array. */
|
|
GFC_DIMENSION_SET(retarray->dim[0], 0, -1, 1);
|
|
return;
|
|
}
|
|
else
|
|
retarray->base_addr = internal_malloc_size (alloc_size);
|
|
}
|
|
else
|
|
{
|
|
if (rank != GFC_DESCRIPTOR_RANK (retarray))
|
|
runtime_error ("rank of return array incorrect in"
|
|
" SUM intrinsic: is %ld, should be %ld",
|
|
(long int) (GFC_DESCRIPTOR_RANK (retarray)),
|
|
(long int) rank);
|
|
|
|
if (unlikely (compile_options.bounds_check))
|
|
{
|
|
for (n=0; n < rank; n++)
|
|
{
|
|
index_type ret_extent;
|
|
|
|
ret_extent = GFC_DESCRIPTOR_EXTENT(retarray,n);
|
|
if (extent[n] != ret_extent)
|
|
runtime_error ("Incorrect extent in return value of"
|
|
" SUM intrinsic in dimension %ld:"
|
|
" is %ld, should be %ld", (long int) n + 1,
|
|
(long int) ret_extent, (long int) extent[n]);
|
|
}
|
|
}
|
|
}
|
|
|
|
for (n = 0; n < rank; n++)
|
|
{
|
|
count[n] = 0;
|
|
dstride[n] = GFC_DESCRIPTOR_STRIDE(retarray,n);
|
|
}
|
|
|
|
dest = retarray->base_addr;
|
|
|
|
while(1)
|
|
{
|
|
*dest = 0;
|
|
count[0]++;
|
|
dest += dstride[0];
|
|
n = 0;
|
|
while (count[n] == extent[n])
|
|
{
|
|
/* When we get to the end of a dimension, reset it and increment
|
|
the next dimension. */
|
|
count[n] = 0;
|
|
/* We could precalculate these products, but this is a less
|
|
frequently used path so probably not worth it. */
|
|
dest -= dstride[n] * extent[n];
|
|
n++;
|
|
if (n == rank)
|
|
return;
|
|
else
|
|
{
|
|
count[n]++;
|
|
dest += dstride[n];
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
#endif
|