Sema is arbitrarily scalarizing some operations, which means that when I
try to implement vectorized versions of those operations in a backend,
they are impossible to test due to Sema not producing them. Now, I can
implement them and then temporarily enable the new feature for that
backend in order to test them. Once the backend supports all of them,
the feature can be permanently enabled.
This also deletes the Air instructions `int_from_bool` and
`int_from_ptr`, which are just bitcasts with a fixed result type, since
changing `un_op` to `ty_op` takes up the same amount of memory.
This instruction is like `intcast`, but includes two safety checks:
* Checks that the int is in range of the destination type
* If the destination type is an exhaustive enum, checks that the int
is a named enum value
This instruction is locked behind the `safety_checked_instructions`
backend feature; if unsupported, Sema will emit a fallback, as with
other safety-checked instructions.
This instruction is used to add a missing safety check for `@enumFromInt`
truncating bits. This check also has a fallback for backends which do
not yet support `safety_checked_instructions`.
Resolves: #21946
The original motivation here was to fix regressions caused by #22414.
However, while working on this, I ended up discussing a language
simplification with Andrew, which changes things a little from how they
worked before #22414.
The main user-facing change here is that any reference to a prior
function parameter, even if potentially comptime-known at the usage
site or even not analyzed, now makes a function generic. This applies
even if the parameter being referenced is not a `comptime` parameter,
since it could still be populated when performing an inline call. This
is a breaking language change.
The detection of this is done in AstGen; when evaluating a parameter
type or return type, we track whether it referenced any prior parameter,
and if so, we mark this type as being "generic" in ZIR. This will cause
Sema to not evaluate it until the time of instantiation or inline call.
A lovely consequence of this from an implementation perspective is that
it eliminates the need for most of the "generic poison" system. In
particular, `error.GenericPoison` is now completely unnecessary, because
we identify generic expressions earlier in the pipeline; this simplifies
the compiler and avoids redundant work. This also entirely eliminates
the concept of the "generic poison value". The only remnant of this
system is the "generic poison type" (`Type.generic_poison` and
`InternPool.Index.generic_poison_type`). This type is used in two
places:
* During semantic analysis, to represent an unknown result type.
* When storing generic function types, to represent a generic parameter/return type.
It's possible that these use cases should instead use `.none`, but I
leave that investigation to a future adventurer.
One last thing. Prior to #22414, inline calls were a little inefficient,
because they re-evaluated even non-generic parameter types whenever they
were called. Changing this behavior is what ultimately led to #22538.
Well, because the new logic will mark a type expression as generic if
there is any change its resolved type could differ in an inline call,
this redundant work is unnecessary! So, this is another way in which the
new design reduces redundant work and complexity.
Resolves: #22494Resolves: #22532Resolves: #22538
Previously, stepping from the single statement within the loop would
always exit the loop because all of the code unrolled from the loop is
associated with the same line and treated by the debugger as one line.
This commit reworks how anonymous struct literals and tuples work.
Previously, an untyped anonymous struct literal
(e.g. `const x = .{ .a = 123 }`) was given an "anonymous struct type",
which is a special kind of struct which coerces using structural
equivalence. This mechanism was a holdover from before we used
RLS / result types as the primary mechanism of type inference. This
commit changes the language so that the type assigned here is a "normal"
struct type. It uses a form of equivalence based on the AST node and the
type's structure, much like a reified (`@Type`) type.
Additionally, tuples have been simplified. The distinction between
"simple" and "complex" tuple types is eliminated. All tuples, even those
explicitly declared using `struct { ... }` syntax, use structural
equivalence, and do not undergo staged type resolution. Tuples are very
restricted: they cannot have non-`auto` layouts, cannot have aligned
fields, and cannot have default values with the exception of `comptime`
fields. Tuples currently do not have optimized layout, but this can be
changed in the future.
This change simplifies the language, and fixes some problematic
coercions through pointers which led to unintuitive behavior.
Resolves: #16865
This commit introduces a new AIR instruction, `repeat`, which causes
control flow to move back to the start of a given AIR loop. `loop`
instructions will no longer automatically perform this operation after
control flow reaches the end of the body.
The motivation for making this change now was really just consistency
with the upcoming implementation of #8220: it wouldn't make sense to
have this feature work significantly differently. However, there were
already some TODOs kicking around which wanted this feature. It's useful
for two key reasons:
* It allows loops over AIR instruction bodies to loop precisely until
they reach a `noreturn` instruction. This allows for tail calling a
few things, and avoiding a range check on each iteration of a hot
path, plus gives a nice assertion that validates AIR structure a
little. This is a very minor benefit, which this commit does apply to
the LLVM and C backends.
* It should allow for more compact ZIR and AIR to be emitted by having
AstGen emit `repeat` instructions more often rather than having
`continue` statements `break` to a `block` which is *followed* by a
`repeat`. This is done in status quo because `repeat` instructions
only ever cause the direct parent block to repeat. Now that AIR is
more flexible, this flexibility can be pretty trivially extended to
ZIR, and we can then emit better ZIR. This commit does not implement
this.
Support for this feature is currently regressed on all self-hosted
native backends, including x86_64. This support will be added where
necessary before this branch is merged.
This commit modifies the representation of the AIR `switch_br`
instruction to represent ranges in cases. Previously, Sema emitted
different AIR in the case of a range, where the `else` branch of the
`switch_br` contained a simple `cond_br` for each such case which did a
simple range check (`x > a and x < b`). Not only does this add
complexity to Sema, which we would like to minimize, but it also gets in
the way of the implementation of #8220. That proposal turns certain
`switch` statements into a looping construct, and for optimization
purposes, we want to lower this to AIR fairly directly (i.e. without
involving a `loop` instruction). That means we would ideally like a
single instruction to represent the entire `switch` statement, so that
we can dispatch back to it with a different operand as in #8220. This is
not really possible to do correctly under the status quo system.
This commit implements lowering of this new `switch_br` usage in the
LLVM and C backends. The C backend just turns any case containing ranges
entirely into conditionals, as before. The LLVM backend is a little
smarter, and puts scalar items into the `switch` instruction, only using
conditionals for the range cases (which direct to the same bb). All
remaining self-hosted backends are temporarily regressed in the presence
of switch range cases. This functionality will be restored for at least
the x86_64 backend before merge.
instead of relying on the LLVM sancov pass. The LLVM pass is still
executed if trace_pc_guard is requested, disabled otherwise. The LLVM
backend emits the instrumentation directly.
It uses `__sancov_pcs1` symbol name instead of `__sancov_pcs` because
each element is 1 usize instead of 2.
AIR: add CoveragePoint to branch hints which indicates whether those
branches are interesting for code coverage purposes.
Update libfuzzer to use the new instrumentation. It's simplified since
we no longer need the constructor and the pcs are now in a continguous
list.
This is a regression in the fuzzing functionality because the
instrumentation for comparisons is no longer emitted, resulting in worse
fuzzer inputs generated. A future commit will add that instrumentation
back.
Implements the accepted proposal to introduce `@branchHint`. This
builtin is permitted as the first statement of a block if that block is
the direct body of any of the following:
* a function (*not* a `test`)
* either branch of an `if`
* the RHS of a `catch` or `orelse`
* a `switch` prong
* an `or` or `and` expression
It lowers to the ZIR instruction `extended(branch_hint(...))`. When Sema
encounters this instruction, it sets `sema.branch_hint` appropriately,
and `zirCondBr` etc are expected to reset this value as necessary. The
state is on `Sema` rather than `Block` to make it automatically
propagate up non-conditional blocks without special handling. If
`@panic` is reached, the branch hint is set to `.cold` if none was
already set; similarly, error branches get a hint of `.unlikely` if no
hint is explicitly provided. If a condition is comptime-known, `cold`
hints from the taken branch are allowed to propagate up, but other hints
are discarded. This is because a `likely`/`unlikely` hint just indicates
the direction this branch is likely to go, which is redundant
information when the branch is known at comptime; but `cold` hints
indicate that control flow is unlikely to ever reach this branch,
meaning if the branch is always taken from its parent, then the parent
is also unlikely to ever be reached.
This branch information is stored in AIR `cond_br` and `switch_br`. In
addition, `try` and `try_ptr` instructions have variants `try_cold` and
`try_ptr_cold` which indicate that the error case is cold (rather than
just unlikely); this is reachable through e.g. `errdefer unreachable` or
`errdefer @panic("")`.
A new API `unwrapSwitch` is introduced to `Air` to make it more
convenient to access `switch_br` instructions. In time, I plan to update
all AIR instructions to be accessed via an `unwrap` method which returns
a convenient tagged union a la `InternPool.indexToKey`.
The LLVM backend lowers branch hints for conditional branches and
switches as follows:
* If any branch is marked `unpredictable`, the instruction is marked
`!unpredictable`.
* Any branch which is marked as `cold` gets a
`llvm.assume(i1 true) [ "cold"() ]` call to mark the code path cold.
* If any branch is marked `likely` or `unlikely`, branch weight metadata
is attached with `!prof`. Likely branches get a weight of 2000, and
unlikely branches a weight of 1. In `switch` statements, un-annotated
branches get a weight of 1000 as a "middle ground" hint, since there
could be likely *and* unlikely *and* un-annotated branches.
For functions, a `cold` hint corresponds to the `cold` function
attribute, and other hints are currently ignored -- as far as I can tell
LLVM doesn't really have a way to lower them. (Ideally, we would want
the branch hint given in the function to propagate to call sites.)
The compiler and standard library do not yet use this new builtin.
Resolves: #21148
This replaces the constant `Zir.Inst.Ref` tags (and the analagous tags
in `Air.Inst.Ref`, `InternPool.Index`) referring to types in
`std.builtin` with a ZIR instruction `extended(builtin_type(...))` which
instructs Sema to fetch such a type, effectively as if it were a
shorthand for the ZIR for `@import("std").builtin.xyz`.
Previously, this was achieved through constant tags in `Ref`. The
analagous `InternPool` indices began as `simple_type` values, and were
later rewritten to the correct type information. This system was kind of
brittle, and more importantly, isn't compatible with incremental
compilation of std, since incremental compilation relies on the ability
to recreate types at different indices when they change. Replacing the
old system with this instruction slightly increases the size of ZIR, but
it simplifies logic and allows incremental compilation to work correctly
on the standard library.
This shouldn't have a significant impact on ZIR size or compiler
performance, but I will take measurements in the PR to confirm this.
I'm so sorry.
This commit was just meant to be making all types fully resolve by
queueing resolution at the moment of their creation. Unfortunately, a
lot of dominoes ended up falling. Here's what happened:
* I added a work queue job to fully resolve a type.
* I realised that from here we could eliminate `Sema.types_to_resolve`
if we made function codegen a separate job. This is desirable for
simplicity of both spec and implementation.
* This led to a new AIR traversal to detect whether any required type is
unresolved. If a type in the AIR failed to resolve, then we can't run
codegen.
* Because full type resolution now occurs by the work queue job, a bug
was exposed whereby error messages for type resolution were associated
with the wrong `Decl`, resulting in duplicate error messages when the
type was also resolved "by" its owner `Decl` (which really *all*
resolution should be done on).
* A correct fix for this requires using a different `Sema` when
performing type resolution: we need a `Sema` owned by the type. Also
note that this fix is necessary for incremental compilation.
* This means a whole bunch of functions no longer need to take `Sema`s.
* First-order effects: `resolveTypeFields`, `resolveTypeLayout`, etc
* Second-order effects: `Type.abiAlignmentAdvanced`, `Value.orderAgainstZeroAdvanced`, etc
The end result of this is, in short, a more correct compiler and a
simpler language specification. This regressed a few error notes in the
test cases, but nothing that seems worth blocking this change.
Oh, also, I ripped out the old code in `test/src/Cases.zig` which
introduced a dependency on `Compilation`. This dependency was
problematic at best, and this code has been unused for a while. When we
re-enable incremental test cases, we must rewrite their executor to use
the compiler server protocol.
This patch is a pure rename plus only changing the file path in
`@import` sites, so it is expected to not create version control
conflicts, even when rebasing.
This reverts commit a7de02e052.
This did not implement the accepted proposal, and I did not sign off
on the changes. I would like a chance to review this, please.
This was a "fake" type used to handle C varargs parameters, much like
generic poison. In fact, it is treated identically to generic poison in
all cases other than one (the final coercion of a call argument), which
is trivially special-cased. Thus, it makes sense to remove this special
tag and instead use `generic_poison_type` in its place. This fixes
several bugs in Sema related to missing handling of this tag.
Resolves: #19781
This commit changes how we represent comptime-mutable memory
(`comptime var`) in the compiler in order to implement the intended
behavior that references to such memory can only exist at comptime.
It does *not* clean up the representation of mutable values, improve the
representation of comptime-known pointers, or fix the many bugs in the
comptime pointer access code. These will be future enhancements.
Comptime memory lives for the duration of a single Sema, and is not
permitted to escape that one analysis, either by becoming runtime-known
or by becoming comptime-known to other analyses. These restrictions mean
that we can represent comptime allocations not via Decl, but with state
local to Sema - specifically, the new `Sema.comptime_allocs` field. All
comptime-mutable allocations, as well as any comptime-known const allocs
containing references to such memory, live in here. This allows for
relatively fast checking of whether a value references any
comptime-mtuable memory, since we need only traverse values up to
pointers: pointers to Decls can never reference comptime-mutable memory,
and pointers into `Sema.comptime_allocs` always do.
This change exposed some faulty pointer access logic in `Value.zig`.
I've fixed the important cases, but there are some TODOs I've put in
which are definitely possible to hit with sufficiently esoteric code. I
plan to resolve these by auditing all direct accesses to pointers (most
of them ought to use Sema to perform the pointer access!), but for now
this is sufficient for all realistic code and to get tests passing.
This change eliminates `Zcu.tmp_hack_arena`, instead using the Sema
arena for comptime memory mutations, which is possible since comptime
memory is now local to the current Sema.
This change should allow `Decl` to store only an `InternPool.Index`
rather than a full-blown `ty: Type, val: Value`. This commit does not
perform this refactor.
* Introduce `-Ddebug-extensions` for enabling compiler debug helpers
* Replace safety mode checks with `std.debug.runtime_safety`
* Replace debugger helper checks with `!builtin.strip_debug_info`
Sometimes, you just have to debug optimized compilers...
This commit eliminates the `dbg_block_{begin,end}` instructions from
both ZIR and AIR. Instead, lexical scoping of `dbg_var_{ptr,val}`
instructions is decided based on the AIR block they exist within. This
is a much more robust system, and also results in a huge drop in ZIR
bytes - around 7% for Sema.zig.
This required some enhancements to Sema to prevent elision of blocks
when they are required for debug variable scoping. This can be observed
by looking at the AIR for the following simple test program with and
without `-fstrip`:
```zig
export fn f() void {
{
var a: u32 = 0;
_ = &a;
}
{
var a: u32 = 0;
_ = &a;
}
}
```
When `-fstrip` is passed, no AIR blocks are generated. When `-fno-strip`
is passed, the ZIR blocks are lowered to true AIR blocks to give correct
lexical scoping to the debug vars.
The changes here incidentally reolve #19060. A corresponding behavior
test has been added.
Resolves: #19060
AstGen provides all function call arguments with a result location,
referenced through the call instruction index. The idea is that this
should be the parameter type, but for `anytype` parameters, we use
generic poison, which is required to be handled correctly.
Previously, generic instantiations and inline calls worked by evaluating
all args in advance, before resolving generic parameter types. This
means any generic parameter (not just `anytype` ones) had generic poison
result types. This caused missing result locations in some cases.
Additionally, the generic instantiation logic caused `zirParam` to
analyze the argument types a second time before coercion. This meant
that for nominal types (struct/enum/etc), a *new* type was created,
distinct to the result type which was previously forwarded to the
argument expression.
This commit fixes both of these issues. Generic parameter type
resolution is now interleaved with argument analysis, so that we don't
have unnecessary generic poison types, and generic instantiation logic
now handles parameters itself rather than falling through to the
standard zirParam logic, so avoids duplicating the types.
Resolves: #16566Resolves: #16258Resolves: #16753
Previously, they shared function index with the owner decl, but that
would clobber the data stored for inferred error sets of runtime calls.
Now there is an adhoc_inferred_error_set_type which models the problem
much more correctly.
Abridged summary:
* Move `Module.Fn` into `InternPool`.
* Delete a lot of confusing and problematic `Sema` logic related to
generic function calls.
This commit removes `Module.Fn` and replaces it with two new
`InternPool.Tag` values:
* `func_decl` - corresponding to a function declared in the source
code. This one contains line/column numbers, zir_body_inst, etc.
* `func_instance` - one for each monomorphization of a generic
function. Contains a reference to the `func_decl` from whence the
instantiation came, along with the `comptime` parameter values (or
types in the case of `anytype`)
Since `InternPool` provides deduplication on these values, these fields
are now deleted from `Module`:
* `monomorphed_func_keys`
* `monomorphed_funcs`
* `align_stack_fns`
Instead of these, Sema logic for generic function instantiation now
unconditionally evaluates the function prototype expression for every
generic callsite. This is technically required in order for type
coercions to work. The previous code had some dubious, probably wrong
hacks to make things work, such as `hashUncoerced`. I'm not 100% sure
how we were able to eliminate that function and still pass all the
behavior tests, but I'm pretty sure things were still broken without
doing type coercion for every generic function call argument.
After the function prototype is evaluated, it produces a deduplicated
`func_instance` `InternPool.Index` which can then be used for the
generic function call.
Some other nice things made by this simplification are the removal of
`comptime_args_fn_inst` and `preallocated_new_func` from `Sema`, and the
messy logic associated with them.
I have not yet been able to measure the perf of this against master
branch. On one hand, it reduces memory usage and pointer chasing of the
most heavily used `InternPool` Tag - function bodies - but on the other
hand, it does evaluate function prototype expressions more than before.
We will soon find out.
Previously, interned values were represented as AIR instructions using
the `interned` tag. Now, the AIR ref directly encodes the InternPool
index. The encoding works as follows:
* If the ref matches one of the static values, it corresponds to the same InternPool index.
* Otherwise, if the MSB is 0, the ref corresponds to an InternPool index.
* Otherwise, if the MSB is 1, the ref corresponds to an AIR instruction index (after removing the MSB).
Note that since most static InternPool indices are low values (the
exceptions being `.none` and `.var_args_param_type`), the first rule is
almost a nop.
