Renesas SH7781 - Manual (Page 170)

6. Floating-Point Unit (FPU)
Rev.1.00 Jan. 10, 2008 Page 140 of 1658
REJ09B0261-0100
6.6 Graphics Support Functions
This LSI supports two kinds of graphics functions: new instructions for geometric operations, and
pair single-precision transfer instructions that enable high-speed data transfer.
6.6.1 Geometric Operation Instructions
Geometric operation instructions perform approximate-value computations. To enable high-speed
computation with a minimum of hardware, this LSI ignores comparatively small values in the
partial computation results of four multiplications. Consequently, the error shown below is
produced in the result of the computation:
Maximum error = MAX (individual multiplication result
×
2 –MIN (number of multiplier significant digits–1, number of multiplicand significant digits–1) ) + MAX (result value
× 2 –23 , 2 –149 )
The number of significant digits is 24 for a normalized number and 23 for a denormalized number
(number of leading zeros in the fractional part).
In a future version of the SH Series, the above error is guaranteed, but the same result between
different processor cores is not guaranteed.
(1) FIPR FVm, FVn (m, n: 0, 4, 8, 12)
This instruction is basically used for the following purposes:
• Inner product (m ≠ n):
This operation is generally used for surface/rear surface determination for polygon surfaces.
• Sum of square of elements (m = n):
This operation is generally used to find the length of a vector.
Since an inexact exception is not detected by an FIPR instruction, the inexact exception (I) bit in
both the FPU exception cause field and flag field are always set to 1 when an FIPR instruction is
executed. Therefore, if the I bit is set in the FPU exception enable field, FPU exception handling
will be executed.