Version 2.1 (R2007b) Fixed-Point Toolbox™ Software Release Notes

Version 2.1 (R2007b) Fixed-Point Toolbox™ Software

This table summarizes what's new in Version 2.1 (R2007b):

New features and changes introduced in this version are:

• Support for Data Type Override

• New Bitwise Operator Functions

• bitget Function Updated

• abs Function Supports Complex Inputs

• divide Function Updated

• fi Constructor Applies Property/Value Pairs After Numeric Value

Support for Data Type Override

Fixed-Point Toolbox™ software now supports data type override in Embedded MATLAB™ subset. This facilitates fixed-point design and enables a single source for fixed- and floating-point code generation.

New Bitwise Operator Functions

Fixed-Point Toolbox software now provides the following functions:

• bitandreduce — Bitwise AND of consecutive range of bits

• bitconcat — Concatenate bits of two fi objects

• bitorreduce — Bitwise OR of consecutive range of bits

• bitrol — Bitwise rotate left

• bitror — Bitwise rotate right

• bitsliceget — Consecutive slice of bits

• bitsll — Bit shift left logical

• bitsra — Bit shift right arithmetic

• bitsrl — Bit shift right logical

• bitxorreduce — Bitwise exclusive OR of consecutive range of bits

• getlsb — Least significant bit

• getmsb — Most significant bit

Embedded MATLAB subset also supports these functions.

bitget Function Updated

The bitget function now behaves as follows:

• bitget returns a u1,0.

• bitget supports variable indexing. This means that the position of the bit to get can be a variable instead of a constant.

• The input fi object and the position of the bit to get can be vectors or scalars.

For more information, see the bitget reference page.

Compatibility Consideration

In prior releases, this function returned a uint8. The function now returns a u1,0. To get a uint8, use the uint8 function on the bitget output.

abs Function Supports Complex Inputs

You can now use the abs function to compute the absolute value of a complex fi object.

For more information, see the abs reference page.

divide Function Updated

The divide function now obeys the DataTypeOverride settings of the fipref object.

For more information, see the divide reference page.

Compatibility Consideration

In prior releases, this function did not obey the DataTypeOverride settings of the fipref object. For example, if the input was fi ScaledDouble, but the input numerictype object was fi Fixed, the output was fi Fixed. The output is now fi ScaledDouble.

fi Constructor Applies Property/Value Pairs After Numeric Value

When you call the fi constructor with both a numeric value and one or more property/value pairs that change the numeric value of the fi object, the fi constructor applies the property/value pairs after it sets the numeric value of the fi object.

For more information, see the fi reference page.

Compatibility Consideration

In prior releases, the fi constructor applied the property/value pairs before it set the numeric value. For example, the following code used to produce a fi object with a value of 0:

a = fi(0,1,16,13,'hex','6488')

This code now produces a fi object with a value of pi.

Version 2.2 (R2008a) Fixed-Point Toolbox™ Software

Version 2.0 (R2007a) Fixed-Point Toolbox™ Software