-[N]TRAP [d][i][o][u]

Trap NDP Exceptions

Compile only. Default: -ntrap

The -trap option specifies how each of four numeric data processor (NDP) exceptions will be handled at execution time of your program.

NDP Exceptions

NDP Exception	Option Argument
Divide-by-Zero	d
Invalid Operation	i
Overflow	o
Underflow	u

Specify -trap with any combination of d, i, o, and u to instruct the NDP chip to generate an interrupt when it detects the specified exception(s) and generate an error message. At least one argument must be specified when the -trap option is used.

Notes

• Trapping causes the program to run with a different hardware configuration, which is set when the Fortran runtime library is initialized. Because the runtime is initialized only once, in the main, how the main is compiled determines how the program will run. If the main is compiled with trapping enabled, the entire program will be trapped. If the main is compiled with trapping disabled, no trapping will occur.
• The zero divide exception for two and four byte integers is always handled by hardware, and is not affected by the trap option.
• Trapping cannot be disabled when -quad is specified, except by using the NDPEXC subroutine or the ERRSET service subroutine (see the file ReadMe_Service_Routines.txt).

Visual Studio Property

The -trap option is set using the Fortran|Exception Trapping|Trap All Floating Point Exceptions property. Individual traps can also be set on this page. This option can be set on the property page for the project only.

See Also

Setting Build Options