Conversion of types:

GDB handles conversion of data types during assignment, printing, and arithmetic and comparison operation.

(gdb) p 1.2df +1.2dd

This converts double data type (1.2) to _Decimal64 data type and performs addition operation with _Decimal64 data type (1.2dd) and prints value in _Decimal64 type. HP WDB handles exceptions such as overflow, infinity and division by zero for Decimal Floating Point data type.

(gdb) print 1.2dd / 0

(gdb) inf

HP WDB handles finite, infinite and NaN(Not a Number) values of decimal floating point data type.

NOTE: HP WDB will not support:

•Command line calls of intrinsic functions(Mathematical functions like cos, sin, log etc)

•Decimal Floating Point data type support for Fortran and C++

•HP-UX 11iv2 Integrity and HP 9000 systems

14.24Additional Support for binary floating point data type

14.24.1 Support for Binary Floating Point constants f, l

If the binary floating point constant contains the letter f or l then HP WDB recognizes it as float or long double binary floating point constant. Generally a floating point constant without f or l is considered as double binary floating point constant.

(gdb) p <num><f/l>

This prints the binary floating point constant depending upon its data type.

14.24.2 Support Binary Floating Point variables with format specifier

HP WDB versions 5.9 and later provide the following format specifiers for binary floating point variables, which print the variables in the specified binary floating point format:

(gdb) p/f <var-name>

This prints the binary floating point value as oat.

(gdb) p/db <var-name>

This prints the binary floating point value as double.

(gdb) p/l <var-name>

This prints the binary floating point value as long double.

216 HP-UX Configuration-Specific Information