[STDCXX-423] LIMITS.cpp assumes integers with no padding bits

• Type: Bug
• Status: Open
• Priority: Trivial
• Resolution: Unresolved
• Affects Version/s: 4.1.2, 4.1.3, 4.2.0
• Fix Version/s: 4.3.0
• Component/s: Configuration

Environment

unknown

• Severity: Incorrect Behavior
• Assignee: Martin Sebor
• Reporter: Martin Sebor
• Created: 22/May/07 15:59
• Updated: 18/Mar/08 17:51

Description

The LIMITS.cpp config test tries to determine exact-width integer typese (the equivalent of things such as int32_t) but it fails to consider the fact that all integer types other than unsigned char may have padding bits that contribute to their size but not to their range (i.e., there may be two or more distinct bit patters that represent the same value, or bit patterns that do not correspond to any value at all). This is only an issue in theory since we know of no architecture with such a representation.