m3x3 - 3x3 multiply

m3x3    dest.xyz, src0, src1

Name	Registers
dest	An output register or temporary register. Requires the mask { .xyz }.
src0	A temporary register, input register, or floating-point constant register.
src1	A temporary register, an input register, or a floating-point constant register. Specifies the starting register of a 3x3 matrix. (In other words, if src1 is r0, then the 3x3 matrix is stored in r0, r1, and r2.)

You cannot specify a floating-point constant register for both src0 and src1.
You cannot specify an input register for both src0 and src1 at the same time.
You cannot specify the same register for dest and src0.

If you set dest to the registers that immediately follow src1 (this is src2, src3 in Macro Expansion below), there is no way to anticipate the result because the content of those registers will be updated during the execution of the expanded instructions.

Stores the result of multiplying a 3x3 matrix and a 3-component vector in dest.

src2 = Next_Index_Of ( src1 )
src3 = Next_Index_Of ( src2 )
dest.x = ( src0.x * src1.x ) + ( src0.y * src1.y ) + ( src0.z * src1.z )
dest.y = ( src0.x * src2.x ) + ( src0.y * src2.y ) + ( src0.z * src2.z )
dest.z = ( src0.x * src3.x ) + ( src0.y * src3.y ) + ( src0.z * src3.z )

dp3     dest.x, src0, src1
dp3     dest.y, src0, src2  // src2 = next index of src1
dp3     dest.z, src0, src3  // src3 = next index of src2

m3x3    r0.xyz, r1, c0
// This is expanded as follows
// dp3     r0.x, r1, c0
// dp3     r0.y, r1, c1
// dp3     r0.z, r1, c2

	1	2	3	4	5	6	7	8
dp3	read	MUL	ADD	ADD	post	write
dp3		read	MUL	ADD	ADD	post	write
dp3			read	MUL	ADD	ADD	post	write

2011/12/20

Initial version.