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.
The idea here is that there are two ways we can reference a function at runtime:
* Through a direct call, i.e. where the function is comptime-known
* Through a function pointer
This means we can easily perform a form of rudimentary escape analysis
on functions. If we ever see a `decl_ref` or `ref` of a function, we
have a function pointer, which could "leak" into runtime code, so we
emit the function; but for a plain `decl_val`, there's no need to.
This change means that `comptime { _ = f; }` no longer forces a function
to be emitted, which was used for some things (mainly tests). These use
sites have been replaced with `_ = &f;`, which still triggers analysis
of the function body, since you're taking a pointer to the function.
Resolves: #6256Resolves: #15353
* move `ptrBitWidth` from Arch to Target since it needs to know about the abi
* double isn't always 8 bits
* AVR uses 1-byte alignment for everything in GCC
This commit removes the `field_call_bind` and `field_call_bind_named` ZIR
instructions, replacing them with a `field_call` instruction which does the bind
and call in one.
`field_call_bind` is an unfortunate instruction. It's tied into one very
specific usage pattern - its result can only be used as a callee. This means
that it creates a value of a "pseudo-type" of sorts, `bound_fn` - this type used
to exist in Zig, but now we just hide it from the user and have AstGen ensure
it's only used in one way. This is quite silly - `Type` and `Value` should, as
much as possible, reflect real Zig types and values.
It makes sense to instead encode the `a.b()` syntax as its own ZIR instruction,
so that's what we do here. This commit introduces a new instruction,
`field_call`. It's like `call`, but rather than a callee ref, it contains a ref
to the object pointer (`&a` in `a.b()`) and the string field name (`b`). This
eliminates `bound_fn` from the language, and slightly decreases the size of
generated ZIR - stats below.
This commit does remove a few usages which used to be allowed:
- `@field(a, "b")()`
- `@call(.auto, a.b, .{})`
- `@call(.auto, @field(a, "b"), .{})`
These forms used to work just like `a.b()`, but are no longer allowed. I believe
this is the correct choice for a few reasons:
- `a.b()` is a purely *syntactic* form; for instance, `(a.b)()` is not valid.
This means it is *not* inconsistent to not allow it in these cases; the
special case here isn't "a field access as a callee", but rather this exact
syntactic form.
- The second argument to `@call` looks much more visually distinct from the
callee in standard call syntax. To me, this makes it seem strange for that
argument to not work like a normal expression in this context.
- A more practical argument: it's confusing! `@field` and `@call` are used in
very different contexts to standard function calls: the former normally hints
at some comptime machinery, and the latter that you want more precise control
over parts of a function call. In these contexts, you don't want implicit
arguments adding extra confusion: you want to be very explicit about what
you're doing.
Lastly, some stats. I mentioned before that this change slightly reduces the
size of ZIR - this is due to two instructions (`field_call_bind` then `call`)
being replaced with one (`field_call`). Here are some numbers:
+--------------+----------+----------+--------+
| File | Before | After | Change |
+--------------+----------+----------+--------+
| Sema.zig | 4.72M | 4.53M | -4% |
| AstGen.zig | 1.52M | 1.48M | -3% |
| hash_map.zig | 283.9K | 276.2K | -3% |
| math.zig | 312.6K | 305.3K | -2% |
+--------------+----------+----------+--------+
This was a special type tag used for hacky stuff in Semantic Analysis.
Move the hacky stuff to use a dedicated `Air.Inst.Ref` instead.
This way, `var_args_param` is not involved in the type system or intern
pool.
```
$ cat overflow.zig
test {
var a: [1]u65535 = undefined;
_ = a;
}
$ zig-out/bin/zig test overflow.zig
thread 290266 panic: integer overflow
zig/src/type.zig:3604:55: 0xada43d in intAbiAlignment (zig)
std.math.ceilPowerOfTwoPromote(u16, (bits + 7) / 8),
^
zig/src/type.zig:3598:42: 0xadd4ea in intAbiSize (zig)
const alignment = intAbiAlignment(bits, target);
^
zig/src/type.zig:3500:61: 0x92be91 in abiSizeAdvanced (zig)
return AbiSizeAdvanced{ .scalar = intAbiSize(bits, target) };
^
zig/src/type.zig:3385:62: 0x928933 in abiSizeAdvanced (zig)
switch (try payload.elem_type.abiSizeAdvanced(target, strat)) {
^
zig/src/type.zig:3268:32: 0x92c012 in abiSize (zig)
return (abiSizeAdvanced(ty, target, .eager) catch unreachable).scalar;
^
```
This is only noticed in a debug build of zig and silently does the wrong
thing and overflows in release builds.
This happened to `[x]u65529` and above because of the ` + 7` on a `u16`.
Fixes#15489
This also lays the groundwork for exposing the whether or not a function is
noinline in std.builtin.Fn as an `is_noinline: bool` field if we ever want to do that.
* Sema: upgrade operands to array pointers if possible when emitting
AIR.
* Implement safety checks for length mismatch and aliasing.
* AIR: make ptrtoint support slice operands. Implement in LLVM backend.
* C backend: implement new `@memset` semantics. `@memcpy` is not done
yet.
AmdgpuKernel and NvptxKernel are unified into a Kernel calling convention.
There is really no reason for these to be separate; no backend is allowed to
emit the calling convention of the other. This is in the same spirit as the
.Interrupt calling convention lowering to different LLVM calling conventions,
and opens the way for SPIR-V kernels to be exported using the Kernel calling
convention.
This also makes another breaking change to for loops: in order to
capture a pointer of an element, one must take the address of array
values. This simplifies a lot of things, and makes more sense than how
it was before semantically.
It is still legal to use a for loop on an array value if the
corresponding element capture is byval instead of byref.
There are still a few occurrences of "stage1" in the standard library
and self-hosted compiler source, however, these instances need a bit
more careful inspection to ensure no breakage.
I wasn't able to create a reduced test case for this but the reasoning
can be seen in `abiAlignmentAdvancedUnion` where if `strat` was lazy
`hasRuntimeBitsAdvanced` would be given `null` instead of `sema`
which would cause eager evaluation when it is not valid or desired.
These parameters are only ever needed when `std.builtin` is out of sync
with the compiler in which case panicking is the only valid operation
anyways. Removing them causes a domino effect of functions no longer
needing a `src` and/or a `block` parameter resulting in handling
compilation errors where they are actually meaningful becoming simpler.
These functions have a very error-prone API. They are essentially
`all(cmp(op, ...))` but that's not reflected in the name.
This renames these functions to `compareAllAgainstZero...` etc.
for clarity and fixes >20 locations where the predicate was
incorrect.
In the future, the scalar `compare` should probably be split off
from the vector comparison. Rank-polymorphic programming is great,
but a proper implementation in Zig would decouple comparison and
reduction, which then needs a way to fuse ops at comptime.
