@bitSizeOf should error when no sensible result can be returned

[shawnl]

#2177 (comment)

This function measures the size at runtime. For types that are disallowed at runtime, such as
 {#syntax#}comptime_int{#endsyntax#} and {#syntax#}type{#endsyntax#}, the result is {#syntax#}0{#endsyntax#}.

I can't find a detailed reason why this is needed over erroring if we can't return a sensible result. Types are always comptime. Throwing an compile error when possible is part of the zen of zig, and prevents bugs.

[LemonBoy]

Its behaviour is consistent with @sizeOf and sort-of makes sense if you consider the comptime types as "immaterial" as they don't get materialized past the analysis phase.

I do agree that throwing an error is a good idea, having the invariant that the size (either in bit or bytes) of such a type is never zero reduces the number of corner cases one has to be aware of.

[shawnl]

I think some types validly have zero size. Look at the discussion of zero-width integers. However certain types have no size at all.

[rohlem]

I failed to find a corresponding issue for @sizeOf and @alignOf, so I'll just add here that
imo they should equally be an error (or return null?) when given a comptime-only type which categorically lacks such properties
(as they are often properties the compiler decides for code generation for the runtime target system).