🔧 C#: Fix an Enum HasFlag Returning True for the Zero Value
Enum.HasFlag() checks whether every bit set in its argument is also set on the instance it’s called against – well-defined for a genuine non-zero flag, but when the argument is the enum’s zero value, the check becomes “are all the bits of zero set,” which is vacuously true for every possible instance. A guard clause written as someValue.HasFlag(MyFlags.None) to mean “nothing is set” returns true no matter what someValue actually contains, and the branch behind it runs unconditionally instead of only when nothing is actually set.
🔎 The Problem
[Flags]
public enum Permissions { None = 0, Read = 1, Write = 2, Execute = 4 }
Permissions value = Permissions.Read;
bool noFlags = value.HasFlag(Permissions.None); // true - always true!
if (noFlags) { /* this branch runs even though Read IS set */ }
✅ Fix: Never Call HasFlag With the Zero Member
- Never call
HasFlag()with the zero orNonemember as the argument; test for “nothing set” with a direct equality check instead –someValue == MyFlags.None– since that’s the only correct way to express it. - Add a unit test that calls
HasFlag(None)against both a zero value and a non-zero value and asserts on the surprising result, so the behavior is documented in the test suite rather than rediscovered as a bug report. - Consider the bitwise replacement pattern –
(someValue & MyFlags.SomeFlag) == MyFlags.SomeFlag– for any flags check where this ambiguity matters, since it has no equivalent zero-argument trap and is also measurably faster than boxing-basedHasFlag().
🚨 Why This Is Dangerous
- A guard clause meant to catch “no flags set” that actually always returns true can mask real flag combinations that should have taken a completely different code path, with the bug invisible until someone specifically tests the zero case.
HasFlag(None) isn’t asking what you think it’s asking – every value trivially has all of zero’s bits set, which is exactly zero bits.

