MTXMultVecArraySR

C Specification

#include <revolution/mtx.h>

void C_MTXMultVecArraySR ( const Mtx m, const Vec* srcBase, Vec* dstBase, u32 count ); void PSMTXMultVecArraySR ( const Mtx m, const Vec* srcBase, Vec* dstBase, u32 count );

#define MTXMultVecArraySR C_MTXMultVecArraySR // for debug build #define MTXMultVecArraySR PSMTXMultVecArraySR // for nondebug build

Arguments

`m`	Source matrix for multiply. `m` is unaffected by this function. `m` may be either an `Mtx` or an `MtxPtr`.
`srcBase`	Pointer to first `Vec` of source vector array. `srcBase` may be either a `VecPtr` or a `Point3dPtr`.
`dstBase`	Pointer to first `Vec` of destination vector array. `dstBase` may be either a `VecPtr` or a `Point3dPtr`.
`count`	Number of contiguous array elements to multiply. Acceptable if `srcBase=dstBase`. However, the condition `srcBase<dstBase<(srcBase+count-1)` must be avoided.

Return Values

None.

Description

Multiplies an array of count 3D vectors on the left with a 3x3 sub-matrix (Scale and Rotate component of 3x4 matrix m). For arrays of vectors, this function is more efficient than calling MTXMultVecSR in a loop as it saves the overhead of repeated function calls.

Note: Vectors are transformed one array element at a time through a single temporary vector, so it's acceptable if srcBase=dstBase. However, problems will result if dstBase points are within the range: (srcBase+count-1). In this case, the resulting vectors will overwrite part of the srcBase array before it can be used as a source value.

During a debug build, MTXMultVecArraySR is interpreted as C_MTXMultVecArraySR, which is a normal C language version. During a non-debug build for a Broadway processor, the paired-singles equivalent of this function, PSMTXMultVecArraySR, will be automatically substituted. You can also call C_MTXMultVecArraySR or PSMTXMultVecArraySR explicitly.

Revision History