core: Reformat and modernize irrt.cpp
- Use anon namespace instead of static - Use using declaration instead of typedef - Align pointers to the type instead of the identifier
This commit is contained in:
parent
bc40a32524
commit
1c56005a01
@ -1,29 +1,30 @@
|
|||||||
typedef _BitInt(8) int8_t;
|
using int8_t = _BitInt(8);
|
||||||
typedef unsigned _BitInt(8) uint8_t;
|
using uint8_t = unsigned _BitInt(8);
|
||||||
typedef _BitInt(32) int32_t;
|
using int32_t = _BitInt(32);
|
||||||
typedef unsigned _BitInt(32) uint32_t;
|
using uint32_t = unsigned _BitInt(32);
|
||||||
typedef _BitInt(64) int64_t;
|
using int64_t = _BitInt(64);
|
||||||
typedef unsigned _BitInt(64) uint64_t;
|
using uint64_t = unsigned _BitInt(64);
|
||||||
|
|
||||||
// NDArray indices are always `uint32_t`.
|
// NDArray indices are always `uint32_t`.
|
||||||
typedef uint32_t NDIndex;
|
using NDIndex = uint32_t;
|
||||||
// The type of an index or a value describing the length of a range/slice is always `int32_t`.
|
// The type of an index or a value describing the length of a range/slice is always `int32_t`.
|
||||||
typedef int32_t SliceIndex;
|
using SliceIndex = int32_t;
|
||||||
|
|
||||||
|
namespace {
|
||||||
template <typename T>
|
template <typename T>
|
||||||
static T max(T a, T b) {
|
const T& max(const T& a, const T& b) {
|
||||||
return a > b ? a : b;
|
return a > b ? a : b;
|
||||||
}
|
}
|
||||||
|
|
||||||
template <typename T>
|
template <typename T>
|
||||||
static T min(T a, T b) {
|
const T& min(const T& a, const T& b) {
|
||||||
return a > b ? b : a;
|
return a > b ? b : a;
|
||||||
}
|
}
|
||||||
|
|
||||||
// adapted from GNU Scientific Library: https://git.savannah.gnu.org/cgit/gsl.git/tree/sys/pow_int.c
|
// adapted from GNU Scientific Library: https://git.savannah.gnu.org/cgit/gsl.git/tree/sys/pow_int.c
|
||||||
// need to make sure `exp >= 0` before calling this function
|
// need to make sure `exp >= 0` before calling this function
|
||||||
template <typename T>
|
template <typename T>
|
||||||
static T __nac3_int_exp_impl(T base, T exp) {
|
T __nac3_int_exp_impl(T base, T exp) {
|
||||||
T res = 1;
|
T res = 1;
|
||||||
/* repeated squaring method */
|
/* repeated squaring method */
|
||||||
do {
|
do {
|
||||||
@ -37,8 +38,8 @@ static T __nac3_int_exp_impl(T base, T exp) {
|
|||||||
}
|
}
|
||||||
|
|
||||||
template <typename SizeT>
|
template <typename SizeT>
|
||||||
static SizeT __nac3_ndarray_calc_size_impl(
|
SizeT __nac3_ndarray_calc_size_impl(
|
||||||
const SizeT *list_data,
|
const SizeT* list_data,
|
||||||
SizeT list_len,
|
SizeT list_len,
|
||||||
SizeT begin_idx,
|
SizeT begin_idx,
|
||||||
SizeT end_idx
|
SizeT end_idx
|
||||||
@ -55,11 +56,11 @@ static SizeT __nac3_ndarray_calc_size_impl(
|
|||||||
}
|
}
|
||||||
|
|
||||||
template <typename SizeT>
|
template <typename SizeT>
|
||||||
static void __nac3_ndarray_calc_nd_indices_impl(
|
void __nac3_ndarray_calc_nd_indices_impl(
|
||||||
SizeT index,
|
SizeT index,
|
||||||
const SizeT *dims,
|
const SizeT* dims,
|
||||||
SizeT num_dims,
|
SizeT num_dims,
|
||||||
NDIndex *idxs
|
NDIndex* idxs
|
||||||
) {
|
) {
|
||||||
SizeT stride = 1;
|
SizeT stride = 1;
|
||||||
for (SizeT dim = 0; dim < num_dims; dim++) {
|
for (SizeT dim = 0; dim < num_dims; dim++) {
|
||||||
@ -71,10 +72,10 @@ static void __nac3_ndarray_calc_nd_indices_impl(
|
|||||||
}
|
}
|
||||||
|
|
||||||
template <typename SizeT>
|
template <typename SizeT>
|
||||||
static SizeT __nac3_ndarray_flatten_index_impl(
|
SizeT __nac3_ndarray_flatten_index_impl(
|
||||||
const SizeT *dims,
|
const SizeT* dims,
|
||||||
SizeT num_dims,
|
SizeT num_dims,
|
||||||
const NDIndex *indices,
|
const NDIndex* indices,
|
||||||
SizeT num_indices
|
SizeT num_indices
|
||||||
) {
|
) {
|
||||||
SizeT idx = 0;
|
SizeT idx = 0;
|
||||||
@ -92,19 +93,19 @@ static SizeT __nac3_ndarray_flatten_index_impl(
|
|||||||
}
|
}
|
||||||
|
|
||||||
template <typename SizeT>
|
template <typename SizeT>
|
||||||
static void __nac3_ndarray_calc_broadcast_impl(
|
void __nac3_ndarray_calc_broadcast_impl(
|
||||||
const SizeT *lhs_dims,
|
const SizeT* lhs_dims,
|
||||||
SizeT lhs_ndims,
|
SizeT lhs_ndims,
|
||||||
const SizeT *rhs_dims,
|
const SizeT* rhs_dims,
|
||||||
SizeT rhs_ndims,
|
SizeT rhs_ndims,
|
||||||
SizeT *out_dims
|
SizeT* out_dims
|
||||||
) {
|
) {
|
||||||
SizeT max_ndims = lhs_ndims > rhs_ndims ? lhs_ndims : rhs_ndims;
|
SizeT max_ndims = lhs_ndims > rhs_ndims ? lhs_ndims : rhs_ndims;
|
||||||
|
|
||||||
for (SizeT i = 0; i < max_ndims; ++i) {
|
for (SizeT i = 0; i < max_ndims; ++i) {
|
||||||
const SizeT *lhs_dim_sz = i < lhs_ndims ? &lhs_dims[lhs_ndims - i - 1] : nullptr;
|
const SizeT* lhs_dim_sz = i < lhs_ndims ? &lhs_dims[lhs_ndims - i - 1] : nullptr;
|
||||||
const SizeT *rhs_dim_sz = i < rhs_ndims ? &rhs_dims[rhs_ndims - i - 1] : nullptr;
|
const SizeT* rhs_dim_sz = i < rhs_ndims ? &rhs_dims[rhs_ndims - i - 1] : nullptr;
|
||||||
SizeT *out_dim = &out_dims[max_ndims - i - 1];
|
SizeT* out_dim = &out_dims[max_ndims - i - 1];
|
||||||
|
|
||||||
if (lhs_dim_sz == nullptr) {
|
if (lhs_dim_sz == nullptr) {
|
||||||
*out_dim = *rhs_dim_sz;
|
*out_dim = *rhs_dim_sz;
|
||||||
@ -123,290 +124,291 @@ static void __nac3_ndarray_calc_broadcast_impl(
|
|||||||
}
|
}
|
||||||
|
|
||||||
template <typename SizeT>
|
template <typename SizeT>
|
||||||
static void __nac3_ndarray_calc_broadcast_idx_impl(
|
void __nac3_ndarray_calc_broadcast_idx_impl(
|
||||||
const SizeT *src_dims,
|
const SizeT* src_dims,
|
||||||
SizeT src_ndims,
|
SizeT src_ndims,
|
||||||
const NDIndex *in_idx,
|
const NDIndex* in_idx,
|
||||||
NDIndex *out_idx
|
NDIndex* out_idx
|
||||||
) {
|
) {
|
||||||
for (SizeT i = 0; i < src_ndims; ++i) {
|
for (SizeT i = 0; i < src_ndims; ++i) {
|
||||||
SizeT src_i = src_ndims - i - 1;
|
SizeT src_i = src_ndims - i - 1;
|
||||||
out_idx[src_i] = src_dims[src_i] == 1 ? 0 : in_idx[src_i];
|
out_idx[src_i] = src_dims[src_i] == 1 ? 0 : in_idx[src_i];
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
} // namespace
|
||||||
|
|
||||||
extern "C" {
|
extern "C" {
|
||||||
#define DEF_nac3_int_exp_(T) \
|
#define DEF_nac3_int_exp_(T) \
|
||||||
T __nac3_int_exp_##T(T base, T exp) {\
|
T __nac3_int_exp_##T(T base, T exp) {\
|
||||||
return __nac3_int_exp_impl(base, exp);\
|
return __nac3_int_exp_impl(base, exp);\
|
||||||
}
|
|
||||||
|
|
||||||
DEF_nac3_int_exp_(int32_t)
|
|
||||||
DEF_nac3_int_exp_(int64_t)
|
|
||||||
DEF_nac3_int_exp_(uint32_t)
|
|
||||||
DEF_nac3_int_exp_(uint64_t)
|
|
||||||
|
|
||||||
SliceIndex __nac3_slice_index_bound(SliceIndex i, const SliceIndex len) {
|
|
||||||
if (i < 0) {
|
|
||||||
i = len + i;
|
|
||||||
}
|
|
||||||
if (i < 0) {
|
|
||||||
return 0;
|
|
||||||
} else if (i > len) {
|
|
||||||
return len;
|
|
||||||
}
|
|
||||||
return i;
|
|
||||||
}
|
}
|
||||||
|
|
||||||
SliceIndex __nac3_range_slice_len(
|
DEF_nac3_int_exp_(int32_t)
|
||||||
const SliceIndex start,
|
DEF_nac3_int_exp_(int64_t)
|
||||||
const SliceIndex end,
|
DEF_nac3_int_exp_(uint32_t)
|
||||||
const SliceIndex step
|
DEF_nac3_int_exp_(uint64_t)
|
||||||
) {
|
|
||||||
SliceIndex diff = end - start;
|
|
||||||
if (diff > 0 && step > 0) {
|
|
||||||
return ((diff - 1) / step) + 1;
|
|
||||||
} else if (diff < 0 && step < 0) {
|
|
||||||
return ((diff + 1) / step) + 1;
|
|
||||||
} else {
|
|
||||||
return 0;
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// Handle list assignment and dropping part of the list when
|
SliceIndex __nac3_slice_index_bound(SliceIndex i, const SliceIndex len) {
|
||||||
// both dest_step and src_step are +1.
|
if (i < 0) {
|
||||||
// - All the index must *not* be out-of-bound or negative,
|
i = len + i;
|
||||||
// - The end index is *inclusive*,
|
}
|
||||||
// - The length of src and dest slice size should already
|
if (i < 0) {
|
||||||
// be checked: if dest.step == 1 then len(src) <= len(dest) else len(src) == len(dest)
|
return 0;
|
||||||
SliceIndex __nac3_list_slice_assign_var_size(
|
} else if (i > len) {
|
||||||
SliceIndex dest_start,
|
return len;
|
||||||
SliceIndex dest_end,
|
}
|
||||||
SliceIndex dest_step,
|
return i;
|
||||||
uint8_t *dest_arr,
|
}
|
||||||
SliceIndex dest_arr_len,
|
|
||||||
SliceIndex src_start,
|
SliceIndex __nac3_range_slice_len(
|
||||||
SliceIndex src_end,
|
const SliceIndex start,
|
||||||
SliceIndex src_step,
|
const SliceIndex end,
|
||||||
uint8_t *src_arr,
|
const SliceIndex step
|
||||||
SliceIndex src_arr_len,
|
) {
|
||||||
const SliceIndex size
|
SliceIndex diff = end - start;
|
||||||
) {
|
if (diff > 0 && step > 0) {
|
||||||
/* if dest_arr_len == 0, do nothing since we do not support extending list */
|
return ((diff - 1) / step) + 1;
|
||||||
if (dest_arr_len == 0) return dest_arr_len;
|
} else if (diff < 0 && step < 0) {
|
||||||
/* if both step is 1, memmove directly, handle the dropping of the list, and shrink size */
|
return ((diff + 1) / step) + 1;
|
||||||
if (src_step == dest_step && dest_step == 1) {
|
} else {
|
||||||
const SliceIndex src_len = (src_end >= src_start) ? (src_end - src_start + 1) : 0;
|
return 0;
|
||||||
const SliceIndex dest_len = (dest_end >= dest_start) ? (dest_end - dest_start + 1) : 0;
|
}
|
||||||
if (src_len > 0) {
|
}
|
||||||
__builtin_memmove(
|
|
||||||
dest_arr + dest_start * size,
|
// Handle list assignment and dropping part of the list when
|
||||||
src_arr + src_start * size,
|
// both dest_step and src_step are +1.
|
||||||
src_len * size
|
// - All the index must *not* be out-of-bound or negative,
|
||||||
);
|
// - The end index is *inclusive*,
|
||||||
}
|
// - The length of src and dest slice size should already
|
||||||
if (dest_len > 0) {
|
// be checked: if dest.step == 1 then len(src) <= len(dest) else len(src) == len(dest)
|
||||||
/* dropping */
|
SliceIndex __nac3_list_slice_assign_var_size(
|
||||||
__builtin_memmove(
|
SliceIndex dest_start,
|
||||||
dest_arr + (dest_start + src_len) * size,
|
SliceIndex dest_end,
|
||||||
dest_arr + (dest_end + 1) * size,
|
SliceIndex dest_step,
|
||||||
(dest_arr_len - dest_end - 1) * size
|
uint8_t* dest_arr,
|
||||||
);
|
SliceIndex dest_arr_len,
|
||||||
}
|
SliceIndex src_start,
|
||||||
/* shrink size */
|
SliceIndex src_end,
|
||||||
return dest_arr_len - (dest_len - src_len);
|
SliceIndex src_step,
|
||||||
}
|
uint8_t* src_arr,
|
||||||
/* if two range overlaps, need alloca */
|
SliceIndex src_arr_len,
|
||||||
uint8_t need_alloca =
|
const SliceIndex size
|
||||||
(dest_arr == src_arr)
|
) {
|
||||||
&& !(
|
/* if dest_arr_len == 0, do nothing since we do not support extending list */
|
||||||
max(dest_start, dest_end) < min(src_start, src_end)
|
if (dest_arr_len == 0) return dest_arr_len;
|
||||||
|| max(src_start, src_end) < min(dest_start, dest_end)
|
/* if both step is 1, memmove directly, handle the dropping of the list, and shrink size */
|
||||||
);
|
if (src_step == dest_step && dest_step == 1) {
|
||||||
if (need_alloca) {
|
const SliceIndex src_len = (src_end >= src_start) ? (src_end - src_start + 1) : 0;
|
||||||
uint8_t *tmp = reinterpret_cast<uint8_t *>(__builtin_alloca(src_arr_len * size));
|
const SliceIndex dest_len = (dest_end >= dest_start) ? (dest_end - dest_start + 1) : 0;
|
||||||
__builtin_memcpy(tmp, src_arr, src_arr_len * size);
|
if (src_len > 0) {
|
||||||
src_arr = tmp;
|
|
||||||
}
|
|
||||||
SliceIndex src_ind = src_start;
|
|
||||||
SliceIndex dest_ind = dest_start;
|
|
||||||
for (;
|
|
||||||
(src_step > 0) ? (src_ind <= src_end) : (src_ind >= src_end);
|
|
||||||
src_ind += src_step, dest_ind += dest_step
|
|
||||||
) {
|
|
||||||
/* for constant optimization */
|
|
||||||
if (size == 1) {
|
|
||||||
__builtin_memcpy(dest_arr + dest_ind, src_arr + src_ind, 1);
|
|
||||||
} else if (size == 4) {
|
|
||||||
__builtin_memcpy(dest_arr + dest_ind * 4, src_arr + src_ind * 4, 4);
|
|
||||||
} else if (size == 8) {
|
|
||||||
__builtin_memcpy(dest_arr + dest_ind * 8, src_arr + src_ind * 8, 8);
|
|
||||||
} else {
|
|
||||||
/* memcpy for var size, cannot overlap after previous alloca */
|
|
||||||
__builtin_memcpy(dest_arr + dest_ind * size, src_arr + src_ind * size, size);
|
|
||||||
}
|
|
||||||
}
|
|
||||||
/* only dest_step == 1 can we shrink the dest list. */
|
|
||||||
/* size should be ensured prior to calling this function */
|
|
||||||
if (dest_step == 1 && dest_end >= dest_start) {
|
|
||||||
__builtin_memmove(
|
__builtin_memmove(
|
||||||
dest_arr + dest_ind * size,
|
dest_arr + dest_start * size,
|
||||||
|
src_arr + src_start * size,
|
||||||
|
src_len * size
|
||||||
|
);
|
||||||
|
}
|
||||||
|
if (dest_len > 0) {
|
||||||
|
/* dropping */
|
||||||
|
__builtin_memmove(
|
||||||
|
dest_arr + (dest_start + src_len) * size,
|
||||||
dest_arr + (dest_end + 1) * size,
|
dest_arr + (dest_end + 1) * size,
|
||||||
(dest_arr_len - dest_end - 1) * size
|
(dest_arr_len - dest_end - 1) * size
|
||||||
);
|
);
|
||||||
return dest_arr_len - (dest_end - dest_ind) - 1;
|
|
||||||
}
|
}
|
||||||
return dest_arr_len;
|
/* shrink size */
|
||||||
|
return dest_arr_len - (dest_len - src_len);
|
||||||
}
|
}
|
||||||
|
/* if two range overlaps, need alloca */
|
||||||
int32_t __nac3_isinf(double x) {
|
uint8_t need_alloca =
|
||||||
return __builtin_isinf(x);
|
(dest_arr == src_arr)
|
||||||
|
&& !(
|
||||||
|
max(dest_start, dest_end) < min(src_start, src_end)
|
||||||
|
|| max(src_start, src_end) < min(dest_start, dest_end)
|
||||||
|
);
|
||||||
|
if (need_alloca) {
|
||||||
|
uint8_t* tmp = reinterpret_cast<uint8_t *>(__builtin_alloca(src_arr_len * size));
|
||||||
|
__builtin_memcpy(tmp, src_arr, src_arr_len * size);
|
||||||
|
src_arr = tmp;
|
||||||
}
|
}
|
||||||
|
SliceIndex src_ind = src_start;
|
||||||
int32_t __nac3_isnan(double x) {
|
SliceIndex dest_ind = dest_start;
|
||||||
return __builtin_isnan(x);
|
for (;
|
||||||
}
|
(src_step > 0) ? (src_ind <= src_end) : (src_ind >= src_end);
|
||||||
|
src_ind += src_step, dest_ind += dest_step
|
||||||
double tgamma(double arg);
|
) {
|
||||||
|
/* for constant optimization */
|
||||||
double __nac3_gamma(double z) {
|
if (size == 1) {
|
||||||
// Handling for denormals
|
__builtin_memcpy(dest_arr + dest_ind, src_arr + src_ind, 1);
|
||||||
// | x | Python gamma(x) | C tgamma(x) |
|
} else if (size == 4) {
|
||||||
// --- | ----------------- | --------------- | ----------- |
|
__builtin_memcpy(dest_arr + dest_ind * 4, src_arr + src_ind * 4, 4);
|
||||||
// (1) | nan | nan | nan |
|
} else if (size == 8) {
|
||||||
// (2) | -inf | -inf | inf |
|
__builtin_memcpy(dest_arr + dest_ind * 8, src_arr + src_ind * 8, 8);
|
||||||
// (3) | inf | inf | inf |
|
} else {
|
||||||
// (4) | 0.0 | inf | inf |
|
/* memcpy for var size, cannot overlap after previous alloca */
|
||||||
// (5) | {-1.0, -2.0, ...} | inf | nan |
|
__builtin_memcpy(dest_arr + dest_ind * size, src_arr + src_ind * size, size);
|
||||||
|
|
||||||
// (1)-(3)
|
|
||||||
if (__builtin_isinf(z) || __builtin_isnan(z)) {
|
|
||||||
return z;
|
|
||||||
}
|
}
|
||||||
|
}
|
||||||
|
/* only dest_step == 1 can we shrink the dest list. */
|
||||||
|
/* size should be ensured prior to calling this function */
|
||||||
|
if (dest_step == 1 && dest_end >= dest_start) {
|
||||||
|
__builtin_memmove(
|
||||||
|
dest_arr + dest_ind * size,
|
||||||
|
dest_arr + (dest_end + 1) * size,
|
||||||
|
(dest_arr_len - dest_end - 1) * size
|
||||||
|
);
|
||||||
|
return dest_arr_len - (dest_end - dest_ind) - 1;
|
||||||
|
}
|
||||||
|
return dest_arr_len;
|
||||||
|
}
|
||||||
|
|
||||||
double v = tgamma(z);
|
int32_t __nac3_isinf(double x) {
|
||||||
|
return __builtin_isinf(x);
|
||||||
|
}
|
||||||
|
|
||||||
// (4)-(5)
|
int32_t __nac3_isnan(double x) {
|
||||||
return __builtin_isinf(v) || __builtin_isnan(v) ? __builtin_inf() : v;
|
return __builtin_isnan(x);
|
||||||
|
}
|
||||||
|
|
||||||
|
double tgamma(double arg);
|
||||||
|
|
||||||
|
double __nac3_gamma(double z) {
|
||||||
|
// Handling for denormals
|
||||||
|
// | x | Python gamma(x) | C tgamma(x) |
|
||||||
|
// --- | ----------------- | --------------- | ----------- |
|
||||||
|
// (1) | nan | nan | nan |
|
||||||
|
// (2) | -inf | -inf | inf |
|
||||||
|
// (3) | inf | inf | inf |
|
||||||
|
// (4) | 0.0 | inf | inf |
|
||||||
|
// (5) | {-1.0, -2.0, ...} | inf | nan |
|
||||||
|
|
||||||
|
// (1)-(3)
|
||||||
|
if (__builtin_isinf(z) || __builtin_isnan(z)) {
|
||||||
|
return z;
|
||||||
}
|
}
|
||||||
|
|
||||||
double lgamma(double arg);
|
double v = tgamma(z);
|
||||||
|
|
||||||
double __nac3_gammaln(double x) {
|
// (4)-(5)
|
||||||
// libm's handling of value overflows differs from scipy:
|
return __builtin_isinf(v) || __builtin_isnan(v) ? __builtin_inf() : v;
|
||||||
// - scipy: gammaln(-inf) -> -inf
|
}
|
||||||
// - libm : lgamma(-inf) -> inf
|
|
||||||
|
|
||||||
if (__builtin_isinf(x)) {
|
double lgamma(double arg);
|
||||||
return x;
|
|
||||||
}
|
|
||||||
|
|
||||||
return lgamma(x);
|
double __nac3_gammaln(double x) {
|
||||||
|
// libm's handling of value overflows differs from scipy:
|
||||||
|
// - scipy: gammaln(-inf) -> -inf
|
||||||
|
// - libm : lgamma(-inf) -> inf
|
||||||
|
|
||||||
|
if (__builtin_isinf(x)) {
|
||||||
|
return x;
|
||||||
}
|
}
|
||||||
|
|
||||||
double j0(double x);
|
return lgamma(x);
|
||||||
|
}
|
||||||
|
|
||||||
double __nac3_j0(double x) {
|
double j0(double x);
|
||||||
// libm's handling of value overflows differs from scipy:
|
|
||||||
// - scipy: j0(inf) -> nan
|
|
||||||
// - libm : j0(inf) -> 0.0
|
|
||||||
|
|
||||||
if (__builtin_isinf(x)) {
|
double __nac3_j0(double x) {
|
||||||
return __builtin_nan("");
|
// libm's handling of value overflows differs from scipy:
|
||||||
}
|
// - scipy: j0(inf) -> nan
|
||||||
|
// - libm : j0(inf) -> 0.0
|
||||||
|
|
||||||
return j0(x);
|
if (__builtin_isinf(x)) {
|
||||||
|
return __builtin_nan("");
|
||||||
}
|
}
|
||||||
|
|
||||||
uint32_t __nac3_ndarray_calc_size(
|
return j0(x);
|
||||||
const uint32_t *list_data,
|
}
|
||||||
uint32_t list_len,
|
|
||||||
uint32_t begin_idx,
|
|
||||||
uint32_t end_idx
|
|
||||||
) {
|
|
||||||
return __nac3_ndarray_calc_size_impl(list_data, list_len, begin_idx, end_idx);
|
|
||||||
}
|
|
||||||
|
|
||||||
uint64_t __nac3_ndarray_calc_size64(
|
uint32_t __nac3_ndarray_calc_size(
|
||||||
const uint64_t *list_data,
|
const uint32_t* list_data,
|
||||||
uint64_t list_len,
|
uint32_t list_len,
|
||||||
uint64_t begin_idx,
|
uint32_t begin_idx,
|
||||||
uint64_t end_idx
|
uint32_t end_idx
|
||||||
) {
|
) {
|
||||||
return __nac3_ndarray_calc_size_impl(list_data, list_len, begin_idx, end_idx);
|
return __nac3_ndarray_calc_size_impl(list_data, list_len, begin_idx, end_idx);
|
||||||
}
|
}
|
||||||
|
|
||||||
void __nac3_ndarray_calc_nd_indices(
|
uint64_t __nac3_ndarray_calc_size64(
|
||||||
uint32_t index,
|
const uint64_t* list_data,
|
||||||
const uint32_t* dims,
|
uint64_t list_len,
|
||||||
uint32_t num_dims,
|
uint64_t begin_idx,
|
||||||
NDIndex* idxs
|
uint64_t end_idx
|
||||||
) {
|
) {
|
||||||
__nac3_ndarray_calc_nd_indices_impl(index, dims, num_dims, idxs);
|
return __nac3_ndarray_calc_size_impl(list_data, list_len, begin_idx, end_idx);
|
||||||
}
|
}
|
||||||
|
|
||||||
void __nac3_ndarray_calc_nd_indices64(
|
void __nac3_ndarray_calc_nd_indices(
|
||||||
uint64_t index,
|
uint32_t index,
|
||||||
const uint64_t* dims,
|
const uint32_t* dims,
|
||||||
uint64_t num_dims,
|
uint32_t num_dims,
|
||||||
NDIndex* idxs
|
NDIndex* idxs
|
||||||
) {
|
) {
|
||||||
__nac3_ndarray_calc_nd_indices_impl(index, dims, num_dims, idxs);
|
__nac3_ndarray_calc_nd_indices_impl(index, dims, num_dims, idxs);
|
||||||
}
|
}
|
||||||
|
|
||||||
uint32_t __nac3_ndarray_flatten_index(
|
void __nac3_ndarray_calc_nd_indices64(
|
||||||
const uint32_t* dims,
|
uint64_t index,
|
||||||
uint32_t num_dims,
|
const uint64_t* dims,
|
||||||
const NDIndex* indices,
|
uint64_t num_dims,
|
||||||
uint32_t num_indices
|
NDIndex* idxs
|
||||||
) {
|
) {
|
||||||
return __nac3_ndarray_flatten_index_impl(dims, num_dims, indices, num_indices);
|
__nac3_ndarray_calc_nd_indices_impl(index, dims, num_dims, idxs);
|
||||||
}
|
}
|
||||||
|
|
||||||
uint64_t __nac3_ndarray_flatten_index64(
|
uint32_t __nac3_ndarray_flatten_index(
|
||||||
const uint64_t* dims,
|
const uint32_t* dims,
|
||||||
uint64_t num_dims,
|
uint32_t num_dims,
|
||||||
const NDIndex* indices,
|
const NDIndex* indices,
|
||||||
uint64_t num_indices
|
uint32_t num_indices
|
||||||
) {
|
) {
|
||||||
return __nac3_ndarray_flatten_index_impl(dims, num_dims, indices, num_indices);
|
return __nac3_ndarray_flatten_index_impl(dims, num_dims, indices, num_indices);
|
||||||
}
|
}
|
||||||
|
|
||||||
void __nac3_ndarray_calc_broadcast(
|
uint64_t __nac3_ndarray_flatten_index64(
|
||||||
const uint32_t *lhs_dims,
|
const uint64_t* dims,
|
||||||
uint32_t lhs_ndims,
|
uint64_t num_dims,
|
||||||
const uint32_t *rhs_dims,
|
const NDIndex* indices,
|
||||||
uint32_t rhs_ndims,
|
uint64_t num_indices
|
||||||
uint32_t *out_dims
|
) {
|
||||||
) {
|
return __nac3_ndarray_flatten_index_impl(dims, num_dims, indices, num_indices);
|
||||||
return __nac3_ndarray_calc_broadcast_impl(lhs_dims, lhs_ndims, rhs_dims, rhs_ndims, out_dims);
|
}
|
||||||
}
|
|
||||||
|
|
||||||
void __nac3_ndarray_calc_broadcast64(
|
void __nac3_ndarray_calc_broadcast(
|
||||||
const uint64_t *lhs_dims,
|
const uint32_t* lhs_dims,
|
||||||
uint64_t lhs_ndims,
|
uint32_t lhs_ndims,
|
||||||
const uint64_t *rhs_dims,
|
const uint32_t* rhs_dims,
|
||||||
uint64_t rhs_ndims,
|
uint32_t rhs_ndims,
|
||||||
uint64_t *out_dims
|
uint32_t* out_dims
|
||||||
) {
|
) {
|
||||||
return __nac3_ndarray_calc_broadcast_impl(lhs_dims, lhs_ndims, rhs_dims, rhs_ndims, out_dims);
|
return __nac3_ndarray_calc_broadcast_impl(lhs_dims, lhs_ndims, rhs_dims, rhs_ndims, out_dims);
|
||||||
}
|
}
|
||||||
|
|
||||||
void __nac3_ndarray_calc_broadcast_idx(
|
void __nac3_ndarray_calc_broadcast64(
|
||||||
const uint32_t *src_dims,
|
const uint64_t* lhs_dims,
|
||||||
uint32_t src_ndims,
|
uint64_t lhs_ndims,
|
||||||
const NDIndex *in_idx,
|
const uint64_t* rhs_dims,
|
||||||
NDIndex *out_idx
|
uint64_t rhs_ndims,
|
||||||
) {
|
uint64_t* out_dims
|
||||||
__nac3_ndarray_calc_broadcast_idx_impl(src_dims, src_ndims, in_idx, out_idx);
|
) {
|
||||||
}
|
return __nac3_ndarray_calc_broadcast_impl(lhs_dims, lhs_ndims, rhs_dims, rhs_ndims, out_dims);
|
||||||
|
}
|
||||||
|
|
||||||
void __nac3_ndarray_calc_broadcast_idx64(
|
void __nac3_ndarray_calc_broadcast_idx(
|
||||||
const uint64_t *src_dims,
|
const uint32_t* src_dims,
|
||||||
uint64_t src_ndims,
|
uint32_t src_ndims,
|
||||||
const NDIndex *in_idx,
|
const NDIndex* in_idx,
|
||||||
NDIndex *out_idx
|
NDIndex* out_idx
|
||||||
) {
|
) {
|
||||||
__nac3_ndarray_calc_broadcast_idx_impl(src_dims, src_ndims, in_idx, out_idx);
|
__nac3_ndarray_calc_broadcast_idx_impl(src_dims, src_ndims, in_idx, out_idx);
|
||||||
}
|
}
|
||||||
}
|
|
||||||
|
void __nac3_ndarray_calc_broadcast_idx64(
|
||||||
|
const uint64_t* src_dims,
|
||||||
|
uint64_t src_ndims,
|
||||||
|
const NDIndex* in_idx,
|
||||||
|
NDIndex* out_idx
|
||||||
|
) {
|
||||||
|
__nac3_ndarray_calc_broadcast_idx_impl(src_dims, src_ndims, in_idx, out_idx);
|
||||||
|
}
|
||||||
|
} // extern "C"
|
Loading…
Reference in New Issue
Block a user