SAF-XC164CS-16F40FBB Infineon Technologies, SAF-XC164CS-16F40FBB Datasheet - Page 24

• Manufacturer Part Number: SAF-XC164CS-16F40FBB
• Manufacturer: Infineon Technologies

example, shift and rotate instructions are always processed during one machine cycle
independent of the number of bits to be shifted. Also multiplication and most MAC
instructions execute in one single cycle. All multiple-cycle instructions have been
optimized so that they can be executed very fast as well: for example, a division
algorithm is performed in 18 to 21 CPU cycles, depending on the data and division type.
Four cycles are always visible, the rest runs in the background. Another pipeline
optimization, the branch target prediction, allows eliminating the execution time of
branch instructions if the prediction was correct.
The CPU has a register context consisting of up to three register banks with 16 wordwide
GPRs each at its disposal. The global register bank is physically allocated within the
on
the active global register bank to be accessed by the CPU at any time. The number of
register banks is only restricted by the available internal RAM space. For easy parameter
passing, a register bank may overlap others.
A system stack of up to 32 Kwords is provided as a storage for temporary data. The
system stack can be allocated to any location within the address space (preferably in the
on-chip RAM area), and it is accessed by the CPU via the stack pointer (SP) register.
Two separate SFRs, STKOV and STKUN, are implicitly compared against the stack
pointer value upon each stack access for the detection of a stack overflow or underflow.
The high performance offered by the hardware implementation of the CPU can efficiently
be utilized by a programmer via the highly efficient XC164CS instruction set which
includes the following instruction classes:
•
•
•
•
•
•
•
•
•
•
•
•
•
The basic instruction length is either 2 or 4 bytes. Possible operand types are bits, bytes
and words. A variety of direct, indirect or immediate addressing modes are provided to
specify the required operands.
Data Sheet
-
chip DPRAM area. A Context Pointer (CP) register determines the base address of
Standard Arithmetic Instructions
DSP-Oriented Arithmetic Instructions
Logical Instructions
Boolean Bit Manipulation Instructions
Compare and Loop Control Instructions
Shift and Rotate Instructions
Prioritize Instruction
Data Movement Instructions
System Stack Instructions
Jump and Call Instructions
Return Instructions
System Control Instructions
Miscellaneous Instructions
22
Functional Description
Derivatives
V2.3, 2006-08
XC164CS