Unlike unions and structs, enums are actually *encoded* into the
InternPool directly, rather than using the SegmentedList trick. This
results in them being quite compact, and greatly improved the ergonomics
of using enum types throughout the compiler.
It did however require introducing a new concept to the InternPool which
is an "incomplete" item - something that is added to gain a permanent
Index, but which is then mutated in place. This was necessary because
enum tag values and tag types may reference the namespaces created by
the enum itself, which required constructing the namespace, decl, and
calling analyzeDecl on the decl, which required the decl value, which
required the enum type, which required an InternPool index to be
assigned and for it to be meaningful.
The API for updating enums in place turned out to be quite slick and
efficient - the methods directly populate pre-allocated arrays and
return the information necessary to output the same compilation errors
as before.
This introduces a string table into InternPool as well as a curious new
field called `maps` which is an array list of array hash maps with
void/void key/value.
Some types such as enums, structs, and unions need to store mappings
from field names to field index, or value to field index. In such cases,
they will store the underlying field names and values directly, relying
on one of these maps, stored separately, to provide lookup.
This allows the InternPool to be serialized via simple array copies,
omitting all the maps, which are only used for optimizing lookup based
on field name or field value.
When the InternPool is deserialized it can be loaded via simple array
copies, and then as a post-processing step the field name maps can be
generated as extra metadata that is tacked on.
This commit provides two encodings for enums - one when the integer tag
type is explicitly provided and one when it is not. This is simpler than
the previous setup, which has three encodings.
Previous sizes:
* EnumSimple: 40 bytes + 16 bytes per field
* EnumNumbered: 80 bytes + 24 bytes per field
* EnumFull: 184 bytes + 24 bytes per field
Sizes after this commit:
* type_enum_explicit: 24 bytes + 8 bytes per field
* type_enum_auto: 16 bytes + 4 bytes per field
On a simple input file, this had a total savings of 21% in the
InternPool:
Before:
int_positive: 427 occurrences, 8975 total bytes
After:
int_positive: 258 occurrences, 5426 total bytes
int_u8: 169 occurrences, 845 total bytes
Temporarily used for some unfortunate allocations made by backends that
need to construct pointer types that can't be represented by the
InternPool. Once all types are migrated to be stored in the InternPool,
this can be removed.
The Key struct now has a Storage tagged union which can store a u64,
i64, or big int.
This is needed so that indexToKey can be implemented for integers stored
compactly in the data structure.
Instead of doing everything at once which is a hopelessly large task,
this introduces a piecemeal transition that can be done in small
increments at a time.
This is a minimal changeset that keeps the compiler compiling. It only
uses the InternPool for a small set of types.
Behavior tests are not passing.
Air.Inst.Ref and Zir.Inst.Ref are separated into different enums but
compile-time verified to have the same fields in the same order.
The large set of changes is mainly to deal with the fact that most Type
and Value methods now require a Module to be passed in, so that the
InternPool object can be accessed.
