Overflow Behavior

The following table describes the overflow behavior when a packed decimal number is assigned to a smaller target. An exception is not generated when:

A packed decimal is assigned to a smaller target with integral type.
A packed decimal is assigned to a smaller target of floating point type.

An exception is generated when a packed decimal is assigned to a smaller packed decimal target. You can suppress runtime errors by using the #pragma nosigtrunc directive in your ILE C source code.

Table 1. Handling Overflow from a Packed Decimal to a Smaller Target
From Field	To Field	Runtime Error
Packed Decimal	char, int, short, long, long long, bit	No
Packed Decimal	Packed Decimal	Yes
Packed Decimal	Float	No²
Packed Decimal	Double	No^{1, 2}
Note: There is no packed decimal number large enough to cause overflow when the packed decimal is assigned to a double. If you use the MI instruction `setca` to unmask a floating point exception, you receive an error message MCH1213 for a floating point inexact result.