📄 stencil.h
字号:
#ifndef BZ_ARRAYSTENCIL_H
#define BZ_ARRAYSTENCIL_H
#ifndef BZ_ARRAY_H
#error <blitz/array/stencil.h> must be included via <blitz/array.h>
#endif
#include <blitz/array/stencilops.h>
BZ_NAMESPACE(blitz)
// NEEDS_WORK: currently stencilExtent returns int(1). What if the
// stencil contains calls to math functions, or divisions, etc.?
// Should at least return a number of the appropriate type. Probably
// return a sequence of quasi-random floating point numbers.
/*
* These macros make it easier for users to declare stencil objects.
* The syntax is:
*
* BZ_DECLARE_STENCILN(stencilname, Array1, Array2, ..., ArrayN)
* // stencil operations go here
* BZ_END_STENCIL
*/
#define BZ_DECLARE_STENCIL2(name,A,B) \
struct name { \
template<class T1, class T2, class T3, class T4, class T5, class T6, \
class T7, class T8, class T9, class T10, class T11> \
static inline void apply(T1& A, T2& B, T3, T4, T5, T6, T7, T8, T9, T10, T11) \
{
#define BZ_END_STENCIL_WITH_SHAPE(MINS,MAXS) } \
template<int N> \
void getExtent(TinyVector<int,N>& minb, TinyVector<int,N>& maxb) const \
{ \
minb = MINS; \
maxb = MAXS; \
} \
enum { hasExtent = 1 }; \
};
#define BZ_END_STENCIL } enum { hasExtent = 0 }; };
#define BZ_STENCIL_END BZ_END_STENCIL
#define BZ_DECLARE_STENCIL3(name,A,B,C) \
struct name { \
template<class T1, class T2, class T3, class T4, class T5, class T6, \
class T7, class T8, class T9, class T10, class T11> \
static inline void apply(T1& A, T2& B, T3& C, T4, T5, T6, T7, T8, T9, \
T10, T11) \
{
#define BZ_DECLARE_STENCIL4(name,A,B,C,D) \
struct name { \
template<class T1, class T2, class T3, class T4, class T5, class T6, \
class T7, class T8, class T9, class T10, class T11> \
static inline void apply(T1& A, T2& B, T3& C, T4& D, T5, T6, T7, \
T8, T9, T10, T11) \
{
#define BZ_DECLARE_STENCIL5(name,A,B,C,D,E) \
struct name { \
template<class T1, class T2, class T3, class T4, class T5, class T6, \
class T7, class T8, class T9, class T10, class T11> \
static inline void apply(T1& A, T2& B, T3& C, T4& D, T5& E, T6, T7, T8, \
T9, T10, T11) \
{
#define BZ_DECLARE_STENCIL6(name,A,B,C,D,E,F) \
struct name { \
template<class T1, class T2, class T3, class T4, class T5, class T6, \
class T7, class T8, class T9, class T10, class T11> \
static inline void apply(T1& A, T2& B, T3& C, T4& D, T5& E, T6& F, \
T7, T8, T9, T10, T11) \
{
#define BZ_DECLARE_STENCIL7(name,A,B,C,D,E,F,G) \
struct name { \
template<class T1, class T2, class T3, class T4, \
class T5, class T6, class T7, class T8, class T9, class T10, class T11> \
static inline void apply(T1& A, T2& B, T3& C, T4& D, T5& E, T6& F, T7& G, \
T8, T9, T10, T11) \
{
#define BZ_DECLARE_STENCIL8(name,A,B,C,D,E,F,G,H) \
struct name { \
template<class T1, class T2, class T3, class T4, \
class T5, class T6, class T7, class T8, class T9, class T10, class T11> \
static inline void apply(T1& A, T2& B, T3& C, T4& D, T5& E, T6& F, T7& G, \
T8& H, T9, T10, T11) \
{
#define BZ_DECLARE_STENCIL9(name,A,B,C,D,E,F,G,H,I) \
struct name { \
template<class T1, class T2, class T3, class T4, \
class T5, class T6, class T7, class T8, class T9, class T10, \
class T11> \
static inline void apply(T1& A, T2& B, T3& C, T4& D, T5& E, T6& F, T7& G, \
T8& H, T9& I, T10, T11) \
{
#define BZ_DECLARE_STENCIL10(name,A,B,C,D,E,F,G,H,I,J) \
struct name { \
template<class T1, class T2, class T3, class T4, \
class T5, class T6, class T7, class T8, class T9, class T10, class T11> \
static inline void apply(T1& A, T2& B, T3& C, T4& D, T5& E, T6& F, T7& G, \
T8& H, T9& I, T10& J, T11) \
{
#define BZ_DECLARE_STENCIL11(name,A,B,C,D,E,F,G,H,I,J,K) \
struct name { \
template<class T1, class T2, class T3, class T4, \
class T5, class T6, class T7, class T8, class T9, class T10, \
class T11> \
static inline void apply(T1& A, T2& B, T3& C, T4& D, T5& E, T6& F, T7& G, \
T8& H, T9& I, T10& J, T11& K) \
{
/*
* dummyArray is used to provide "dummy" padding parameters to applyStencil(),
* so that any number of arrays (up to 11) can be given as arguments.
*/
template<class T> class dummy;
struct dummyArray {
typedef dummy<double> T_iterator;
const dummyArray& shape() const { return *this; }
};
_bz_global dummyArray _dummyArray;
/*
* This dummy class pretends to be a scalar of type T, or an array iterator
* of type T, but really does nothing.
*/
template<class T>
class dummy {
public:
dummy() { }
dummy(T value)
: value_(value)
{ }
dummy(const dummyArray&)
{ }
operator T() const { return value_; };
template<class T2>
void operator=(T2) { }
_bz_typename multicomponent_traits<T>::T_element operator[](int i) const
{ return value_[i]; }
void loadStride(int) { }
void moveTo(int) { }
void moveTo(int,int) { }
void moveTo(int,int,int) { }
void moveTo(int,int,int,int) { }
void advance() { }
T shift(int,int) { return T(); }
private:
T value_;
};
/*
* The stencilExtent object is passed to stencil objects to find out
* the spatial extent of the stencil. It pretends it's an array,
* but really it's just recording the locations of the array reads
* via operator().
*/
template<int N_rank, class P_numtype>
class stencilExtent {
public:
typedef P_numtype T_numtype;
stencilExtent()
{
min_ = 0;
max_ = 0;
}
dummy<T_numtype> operator()(int i)
{
update(0, i);
return dummy<T_numtype>(1);
}
dummy<T_numtype> operator()(int i, int j)
{
update(0, i);
update(1, j);
return dummy<T_numtype>(1);
}
dummy<T_numtype> operator()(int i, int j, int k)
{
update(0, i);
update(1, j);
update(2, k);
return dummy<T_numtype>(1);
}
dummy<T_numtype> shift(int offset, int dim)
{
update(dim, offset);
return dummy<T_numtype>(1);
}
dummy<T_numtype> shift(int offset1, int dim1, int offset2, int dim2)
{
update(dim1, offset1);
update(dim2, offset2);
return dummy<T_numtype>(1);
}
dummy<_bz_typename multicomponent_traits<T_numtype>::T_element>
operator[](int)
{
return dummy<_bz_typename multicomponent_traits<T_numtype>::T_element>
(1);
}
void update(int rank, int offset)
{
if (offset < min_[rank])
min_[rank] = offset;
if (offset > max_[rank])
max_[rank] = offset;
}
template<class T_numtype2>
void combine(const stencilExtent<N_rank,T_numtype2>& x)
{
for (int i=0; i < N_rank; ++i)
{
min_[i] = minmax::min(min_[i], x.min(i));
max_[i] = minmax::max(max_[i], x.max(i));
}
}
template<class T_numtype2>
void combine(const dummy<T_numtype2>&)
{ }
int min(int i) const
{ return min_[i]; }
int max(int i) const
{ return max_[i]; }
const TinyVector<int,N_rank>& min() const
{ return min_; }
const TinyVector<int,N_rank>& max() const
{ return max_; }
template<class T>
void operator=(T)
{ }
// NEEDS_WORK: other operators
template<class T> void operator+=(T) { }
template<class T> void operator-=(T) { }
template<class T> void operator*=(T) { }
template<class T> void operator/=(T) { }
operator T_numtype()
{ return T_numtype(1); }
T_numtype operator*()
{ return T_numtype(1); }
private:
_bz_mutable TinyVector<int,N_rank> min_, max_;
};
/*
* stencilExtent_traits gives a stencilExtent<N,T> object for arrays,
* and a dummy object for dummy arrays.
*/
template<class T>
struct stencilExtent_traits {
typedef dummy<double> T_stencilExtent;
};
template<class T_numtype, int N_rank>
struct stencilExtent_traits<Array<T_numtype,N_rank> > {
typedef stencilExtent<N_rank,T_numtype> T_stencilExtent;
};
/*
* Specialization of areShapesConformable(), originally
* defined in <blitz/shapecheck.h>
*/
template<class T_shape1>
inline _bz_bool areShapesConformable(const T_shape1&, const dummyArray&)
{
return _bz_true;
}
BZ_NAMESPACE_END
#include <blitz/array/stencil.cc>
#endif // BZ_ARRAYSTENCIL_H
⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -