You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We could use some Roslyn analyzers added to our new dev analyzers project around [MethodImpl(MethodImplOptions.NoInlining)] and StackTraceHelper, found while investigating #931:
NoInlining should not be used on interface methods or abstract methods. This attribute is not inherited, so it has no effect there.
NoInlining should not be used on empty-bodied methods. Since the test code would never be called from an empty-bodied method (since that calls nothing), it can never match to the stack trace. So we should not harm performance by preventing inlining these methods.
NoInlining should be used on methods referenced in a call to StackTraceHelper.DoesStackTraceContainMethod (perhaps just the two-argument version for now that also validates the owning class), and the method body is not empty.
The text was updated successfully, but these errors were encountered:
Is there an existing issue for this?
Task description
We could use some Roslyn analyzers added to our new dev analyzers project around
[MethodImpl(MethodImplOptions.NoInlining)]
and StackTraceHelper, found while investigating #931:StackTraceHelper.DoesStackTraceContainMethod
(perhaps just the two-argument version for now that also validates the owning class), and the method body is not empty.The text was updated successfully, but these errors were encountered: