Texas Instruments TI-89 - Internal Integer Arithmetic Functions; Table 2.5: Integer Arithmetic Functions

82
Section 2: Compiler
TI
-
89 / TI
-
92 Plus Sierra C Assembler Reference Manual
Not for Distribution
Beta Version February 2, 2001
2.12.1. Internal Integer Arithmetic Functions
The 68000 compiler,
com68
, generates function calls to perform integer
arithmetic when in-line code would require an unacceptable amount of space. It
also generates function calls to perform some division and modulus operations.
The size and type of the operands determine whether a function call is
generated. For example, the division of two signed 16-bit operands will be
handled by in-line code, but the division of two signed 32-bit operands will be
handled by a function call. Table 2.5, Integer Arithmetic Functions, lists the
functions that are called by the 68000 compiler to perform the specified integer
arithmetic operations.
Function Name Operation
__du16u16 Divide 16-bit unsigned by 16-bit unsigned
__ds32s32 Divide 32-bit signed by 32-bit signed
__du32u32 Divide 32-bit unsigned by 32-bit unsigned
__ds16u16 Divide 16-bit signed by 16-bit unsigned
__ms16u16 Mod 16-bit signed with 16-bit unsigned
__mu16u16 Mod 16-bit unsigned with 16-bit unsigned
__ms32s32 Mod 32-bit signed with 32-bit signed
__mu32u32 Mod 32-bit unsigned with 32-bit unsigned
Table 2.5: Integer Arithmetic Functions
Arguments are passed to the functions listed in Table 2.5 in registers
d0
and
d1
;
they are not passed on the stack. Results are always 32 bits and are returned in
register
d1
(basically,
d1 = d1 / d0
or
d1 = d1 % d0
). The functions that operate
on 16-bit operands use only registers
d0
and
d1
. The functions that operate on
32-bit operands also use registers
d2
,
a0
, and
a1
.
The following example, which shows both a 68000 C file and the generated
assembly code, illustrates the generation of an internally generated function call:
extern int i, j; move.l _i,d1
void f( void ) move.l _j,d0
{ jsr __ds32s32
i = i / j; move.l d1,_i
} rts