s5u1c33000c Epson Electronics America, Inc., s5u1c33000c Datasheet - Page 111

• Manufacturer Part Number: s5u1c33000c
• Description: Cmos 32-bit Single Chip Microcomputer S5u1c33000c Manual
• Manufacturer: Epson Electronics America, Inc.

7.2.2 Floating-point Format
The C Compiler gcc33 supports the float type (single-precision, 32 bit) and the double type (double-precision, 64
bit) floating-point numbers conforming to IEEE standards.
The following shows the internal format of floating-point numbers.
Format of double-precision floating-point number
Format of single-precision floating-point number
Note
S5U1C33000C MANUAL
(C COMPILER PACKAGE FOR S1C33 FAMILY) (Ver. 4)
The real number of the double type consists of 64 bits, as shown below.
Bit 63:
Bits 62–52: Exponent part (11 bits)
Bits 51–0: Fixed-point part (52 bits)
When this type of value is stored in a register, it occupies two registers. For example, the result of a
floating-point calculation is stored in the R11 and R10 registers.
The following shows the relationship of the effective range, floating-point representation, and internal data
of the double type.
Maximum normalized number:
Minimum normalized number:
Maximum unnormalized number:
Minimum unnormalized number:
Infinity:
Negative infinity:
Values 0x7ff00000 00000001 to 0x7fffffff ffffffff and 0xfff00000 00000001 to 0xffffffff ffffffff are not
recognized as numeric values.
The real number of the float type consists of 32 bits, as shown below.
Bit 31:
Bits 30–23: Exponent part (8 bits)
Bits 22–0: Fixed-point part (23 bits)
The float type data can be stored in one register.
The following shows the relationship of the effective range, floating-point representation, and internal data
of the float type.
Maximum normalized number:
Minimum normalized number:
Maximum unnormalized number:
Minimum unnormalized number:
Infinity:
Negative infinity:
Values 0x7f800001 to 0x7fffffff and 0xff800001 to 0xffffffff are not recognized as numeric values.
The floating-point numbers in the gcc33 differ from the IEEE-based FPU in precision and functionality,
including the manner in which infinity is handled.
63 62
31 30
S
S
0:
0:
0:
0:
R11 register: Sign bit, exponent part, and 20 high-order bits of fixed-point part (51:32)
R10 register: 32 low-order bits of fixed-point part (31:0)
Exponent part
Exponent part
Sign bit (1 bit)
Sign bit (1 bit)
52 51
23 22
EPSON
Fixed-point part
Fixed-point part
CHAPTER 7: EMULATION LIBRARY
0
0
95